IMPACT OF TRELLIS SYSTEMS AND CROP LOAD ON FRUIT AND WINE QUALITY OF
THE SUPER COLD HARDY CULTIVAR ‘MARQUETTE’
By
Jacob Eli Emling

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Horticulture – Master of Science
2014

ABSTRACT
IMPACT OF TRELLIS SYSTEMS AND CROP LOAD ON FRUIT AND WINE QUALITY OF
THE SUPER COLD HARDY CULTIVAR ‘MARQUETTE’
By
Jacob Eli Emling
Super cold hardy (SCH) grapes are a new class of wine grapes that have been bred
recently by the University of Minnesota breeding program. This research was developed to better
understand the effects of training system and canopy architecture on the SCH cultivar
‘Marquette’. The objectives were to compare fruit chemistry, canopy growth, and three different
trellis systems; Geneva Double curtain (GDC), High Wire Cordon (HWC), and a new
experimental moving trellis (MT). Studies were performed at Horticulture Teaching and
Research Center (HTRC) and Southwest Michigan Research and Education Center (SWMREC)
during the 2012 and 2013 growing seasons. Due to frost events in 2012, two populations of
shoots and the fruit associated with each type of shoot were tracked throughout the growing
season. The results from these two populations had influences on canopy development and
structure, fruit chemistry, and on wine quality. The experiments conducted in 2013 focused on
the influences of canopy densities and crop level on the vines in regards to canopy structure and
fruit quality. Fruit and must chemical profiles were similar from seasonal and harvest samples
but statistical differences were found after fermentation. In conclusion, these studies proved that
training system and shoot type all have slight impacts on fruit quality during the growing season
but has large impacts in the wines produced.

Dedicated to my family, close friends, and peers that have helped make this work possible.

iii

ACKNOWLEDGEMENTS

This research was conducted at two locations within the State of Michigan, Southwest
Michigan Research and Education Center (SWMREC) and Horticulture Teaching and Research
Center (HTRC). All of this could not have been accomplished without the guidance or help of
several people. To these people I owe the deepest of gratitude.
First off, I would like to thank Dr. Paolo Sabbatini, not only for helping me accomplish
my dream of earning this degree, but also for the past several years of mentoring and guidance. I
would also like to thank my other two committee members, Dr. Ronald Perry and Dr. Rufus
Issacs. Both of these professors have helped in many ways and have shared a great deal of
knowledge throughout my graduate career which is reflected within these pages.
The majority of this work would not have been possible without the help of the following
people that were part of the MSU Viticulture and Enology Research group. Thanks go to our
research technician Mr. Patrick Murad. A great deal of appreciation goes to my fellow graduate
students, Letizia Tozzini, Dana Acimovic, and Shijian Zhuang for their willingness to help. I
want to thank all of the undergraduate students that have helped in the laboratory, with the
countless hours conducting field work, and in the winery. These include Emily Winsjansen, Kate
McDonald, Kyle Totzke, Peter Rigan, PJ Geiger, and especially Zach Emling. I also need to
thank the farm managers, Bill Chase of the HTRC and David Francis of SWMREC for their help
with day-to-day management of the vineyard and the various forms of help they provided over
the past few years. Very special thanks goes to Ms. Katie Knox for helping with all of the editing
and being a sounding board during the whole writing process.

iv

TABLE OF CONTENTS

LIST OF TABLES ................................................................................................................ viii
LIST OF FIGURES ............................................................................................................... xviii
CHAPTER I .......................................................................................................................... 1
LITERATURE REVIEW ..................................................................................................... 1
Grapes…………………………………………………………………….……….....1
Climate and Varieties in Michigan………………………………………………......2
Challenges in Michigan…………………………………………....………................7
Cold Hardiness Development of Super Cold Hardy Cultivar ‘Marquette’…....…...10
Viticultural Problems with Super Cold Hardy (SCH) Varieties…………………...13
Fruit Chemistry of SCH………………………………………………...……….....16
Trellis Systems and Impact on Fruit Quality of Super Cold Hardy Varieties……...18
Vertical Shoot Positioning (VSP)...........................................................................19
High Wire Cordon Trellis (HWC)..........................................................................19
Geneva Double Curtain Trellis (GDC)...................................................................20
Split Canopy (Y-Trellis)………………………………………………………...20
Conclusion……………………………...………………………………..………....23
CHAPTER II....................................................................................................................... 27
FROST EVALUATION OF EFFECTS OF TRELLIS SYSTEM ON DAMAGE AND
FRUIT QUALITY OF PRIMARY FRUIT VS. SECONDARY FRUIT IN VARIOS
LOCATIONS IN MICHIGAN IN 2012 ….................................................................. .......27
Introduction………………………………………………………………………... 27
Materials and Methods……………………………………………………………..36
Field experiment –Horticulture teaching and Research center ................................ 36
Plant Material ..................................................................................................... 36
Experimental Design .......................................................................................... 37
Frost Damage Evaluation ................................................................................... 38
Canopy Growth Measurements .......................................................................... 39
Berry Growth Measurements ............................................................................. 39
Sampling Procedure and Harvest Data Collection ............................................. 39
Basic Fruit Chemistry Measurements ................................................................ 40
Statistical Analysis ............................................................................................. 41
Field experiment –Southwest Michigan Research Education and Center ................ 46
Plant Material ...................................................................................................... 46
Experimental Design ........................................................................................... 46
Canopy Growth Measurements ........................................................................... 48
Sampling Procedure and Harvest Data Collection .............................................. 48
Basic Fruit Chemistry Measurements .................................................................. 49
v

Statistical Analysis ............................................................................................... 49
Results: 2012 HTRC .................................................................................................. 53
Weather during the Growing Season ................................................................... 53
Fruit Chemistry during the Ripening Process ..................................................... 54
Fruit Chemistry and Yield Components at Harvest ............................................ 55
Results on Phenology .......................................................................................... 55
Results: 2012 SWMREC........................................................................................... 65
Weather during the Growing Season .................................................................. 65
Fruit Chemistry during the Ripening Process .................................................... 65
Fruit Chemistry and Yield Components at Harvest ........................................... 66
Results on Phenology ......................................................................................... 67
Discussion ................................................................................................................... 76
Conclusion................................................................................................................... 78
APPENDICES ................................................................................................................... 79
APPENDIX A: FRUIT CHEMICAL COMPOSITION OVER TIME……………...80
APPENDIX B: TYPE THREE TABLES OF ANOVA HTRC……………………...84
APPENDIX C: TYPE THREE TABLES OF ANOVA – SWMREC…….…………86

CHAPTER III ....................................................................................................................... 89
SHOOT DENSITY, TRELLIS SYSTEM, AND CROPLOAD EFFECT ON VARIOUS
LOCATIONS IN MICHIGAN IN 2013................................................................ ...... ........89
Introduction………………………………………………………………………….89
Materials and Methods………………………………………………………………93
Shoot Density Field Experiment –Horticulture teaching and Research center .......... 93
Plant Material ...................................................................................................... 93
Experimental Design .......................................................................................... 93
Canopy Growth Measurements ........................................................................... 94
Temperature and Radiation Measurements ........................................................ 95
Sampling Procedure and Harvest Data Collection ............................................. 96
Basic Fruit Chemistry Measurements ................................................................ 97
Crop Load Experiment –Southwest Michigan Research Education and Center ........ 99
Plant Material ..................................................................................................... 99
Experimental Design .......................................................................................... 99
Canopy Growth Measurements ........................................................................... 100
Temperature and Radiation Measurements ........................................................ 100
Sampling Procedure and harvest Data Collection .............................................. 101
Basic Fruit Chemistry Measurements ................................................................ 103
Statistical Analysis ............................................................................................. 103
Results: 2013 HTRC ................................................................................................... 105
Weather during the Growing Season .................................................................. 105
Fruit Chemistry during the Ripening Process .................................................... 105
Fruit Chemistry, Yield Components, and Canopy Measurements at Harvest ... 107
Results: 2013 SWMREC............................................................................................... 120
Weather during the Growing Season ......................................................................... 120
Fruit Chemistry during the Ripening Process ............................................................ 120
vi

Fruit Chemistry, Yield Components, and Canopy Measurements at Harvest ........... 121
Light data obtained from Data Logger ....................................................................... 122
Discussion ...................................................................................................................... 132
Conclusion...................................................................................................................... 134
APPENDICES ................................................................................................................... 135
APPENDIX A: FRUIT CHEMICAL COMPOSITION OVER TIME……………….136
APPENDIX B: Type Three Tables of ANOVA – HTRC…………………………….140
APPENDIX E: Type Three Tables of ANOVA – SWMREC………………………...142
CHAPTER IV ....................................................................................................................... 146
EXPERIMENTIAL WINE PRODUCTION AND CHEMICAL AND SENSORY ANAYLSIS
FOR 2012 AND 2013 ........................................................................................................... 146
Introduction.......................................................................................................................146
Materials and Methods………………………………………………………………....147
Production Facility ..................................................................................................... 147
Enology Procedure ..................................................................................................... 147
Sensory Evaluation ..................................................................................................... 149
Chemical Analysis...................................................................................................... 151
Statistical Analysis ..................................................................................................... 151
Results ............................................................................................................................ 156
Discussion ...................................................................................................................... 170
Conclusion ..................................................................................................................... 171
APPENDICES ................................................................................................................... 172
APPENDIX A: TYPE THREE TABLES OF ANOVA FOR IN 2012 ................................ 173
APPENDIX B: TYPE THREE TABLES OF ANOVA FOR IN 2013 ................................ 176
LITERATURE CITED ......................................................................................................... 179

vii

LIST OF TABLES

TABLE 1. Approximate warmest temperature where 80-100% primary bud kill may be expected
to occur in midwinter. Elaborated from Wine Grape Production Guide for Eastern North
America. (2008). T. Wolf et al., Zabadal T., Sabbatini P., Elsner D., 2008. Wine Grape Varieties
for Michigan and Other Cold Climate Viticultural Regions. MSU Extension Bulletin CD-007..
………………….……………………………………………………………………………….... 6

TABLE 2. Common fruit quality parameters at the time of harvest as related to wine style. Data
for the table collected from Boulton et al.,1996 ………………………………………...………. 9
TABLE 3. Details about specific vines used in the experiment at the HTRC location. ……….. 44

TABLE 4. Growing Degree Days and precipitation weather data collected for 2012 from
MAWN located at HTRC. ………………………………………………………..……………. 45

TABLE 5. Major Phenological stage, Date of occurrence and Growing Degree Days for
Marquette grown at HTRC in 2012. …………………………………………………………… 45
TABLE 6. Bud evaluation after frost event at HTRC in 2012. ……………………….………. 45

TABLE 7. Monthly Growing degree days and precipitation data collected for 2012 from MAWN
located at SWMREC. …………………………………………………………………...……… 52

TABLE 8. Major Phenological stage, Date of occurrence and Growing Degree Days for
Marquette grown at SWMREC in 2012. …………………………….…………………………. 52

TABLE 9. Bud evaluation of primary buds by block after frost event at SWMREC in 2012.
………………………………………………………………………………………………...… 52

TABLE 10. Yield components at harvest by vine in field experiment conducted in 2012 at
HTRC. ……………………………………………………..…………………………...………. 57

viii

TABLE 11. Yield components at harvest by bud type in field experiment conducted in 2012 at
HTRC. ………………………………………………………………………………………….. 57
TABLE 12. Fruit chemical composition at harvest by vine from field experiment conducted at
HTRC. ………………………………………………………………………………………….. 58

TABLE 13. Fruit chemical composition at harvest by bud type from field experiment conducted
at HTRC………………………………………………………………………………………… 58

TABLE 14. Yield components at harvest by vine in Frost evaluation field experiment conducted
in 2012 at SWMREC. ………………………………………………………………………….. 68

TABLE 15. Yield components at harvest by bud type in Frost evaluation field experiment
conducted in 2012 at SWMREC. ……………………………………………...……………….. 68

TABLE 16. Fruit chemical composition at harvest by vine from Frost evaluation field
experiment conducted in 2012 at SWMREC. …………………………………………………. 69

TABLE 17. Fruit chemical composition at harvest by bud type from Frost evaluation field
experiment conducted in 2012 at SWMREC. ……………………………………………..…… 69

TABLE 18. Fruit chemical composition over time for High Wire Cordon trellis at HTRC field
experiment in 2012. ………………………………………………………………………...….. 80

TABLE 19. Fruit chemical composition over time for Geneva Double Curtain at HTRC field
experiment in 2012. ……………………………………………………………………….…… 81

TABLE 20. Fruit chemical composition over time for Moving Trellis at HTRC field experiment
in 2012. ……………………………………………………………………………………...…. 82

TABLE 21. Fruit chemical composition over time for SWMREC frost evaluation field
experiment in 2012. ………………………………………………………………….………….83

TABLE 22. Analysis of variance (ANOVA) of TSS at HTRC in 2012…………….…….……..84

ix

TABLE 23. Analysis of variance (ANOVA) of pH at HTRC in 2012……………..………..…. 84
TABLE 24. Analysis of variance (ANOVA) of TA at HTRC in 2012. …………………….…. 84
TABLE 25. Analysis of variance (ANOVA) of Total Phenolics at HTRC in 2012. …..………..85
TABLE 26. Analysis of variance (ANOVA) of Anthocyanin content at HTRC in 2012. …..….85

TABLE 27. Analysis of variance (ANOVA) of Anthocyanin content by TSS at HTRC in 2012.
……………………………………………………………………………………………...…….85

TABLE 28. Analysis of variance (ANOVA) of Cluster Weights at SWMREC in 2012. ..…… .86
TABLE 29. Analysis of variance (ANOVA) of Rachis Weights at SWMREC in 2012. ……... .86

TABLE 30. Analysis of variance (ANOVA) of Number of berries per cluster at SWMREC in
2012. …………………………………………………………………………………….……….86

TABLE 31. Analysis of variance (ANOVA) of Berry diameters per cluster at SWMREC in 2012.
……………………………………………………..……………..……………………...……….86
TABLE 32. Analysis of variance (ANOVA) of Yield at SWMREC in 2012. ……………….....86
TABLE 33. Analysis of variance (ANOVA) of Number of cluster at SWMREC in 2012. ….....86
TABLE 34. Analysis of variance (ANOVA) of TSS (oBrix) at SWMREC in 2012. …………...87
TABLE 35. Analysis of variance (ANOVA) of pH at SWMREC in 2012. ……………….…....87
TABLE 36. Analysis of variance (ANOVA) of TA at SWMREC in 2012………………….......87
TABLE 37. Analysis of variance (ANOVA) of Total Phenolics at SWMREC in 2012…….…..87
x

TABLE 38. Analysis of variance (ANOVA) of Anthocyanin content at SWMREC in 2012. .....87
TABLE 39. Analysis of variance (ANOVA) of Anthocyanin/Degrees Brix at SWMREC in 2012.
…………………………………………………………………………………………………....88
TABLE 40. Monthly weather data collected for 2013 from MAWN located at HTRC. …..…...98

TABLE 41. Major Phenological stage, Date of occurrence and Growing Degree Days for
Marquette grown at HTRC in 2013……………………………………………….…..…..……..98

TABLE 42. Monthly weather data collected for 2013 from MAWN located at SWMREC. .....104

TABLE 43. Major Phenological stage, Date of occurrence and Growing Degree Days for
Marquette grown at SWMREC in 2013. …………………………………………….………....104

TABLE 44. Canopy measurements at harvest in field experiment conducted in 2013 at HTRC.
Shoot density was either adjusted to three shoots per linear 30 centimeter of cordon (3S) or six
shoots per linear 30 centimeter of cordon (6S). ………………………………………………..110

TABLE 45. Yield components at harvest in field experiment conducted in 2013 at HTRC. Shoot
density was either adjusted to three shoots per linear 30 centimeter of cordon (3S) or six shoots
per linear 30 centimeter of cordon (6S). ……………………………………………....……….111

TABLE 46. Cluster morphology at harvest in field experiment conducted in 2013 at HTRC.
Shoot density was either adjusted to three shoots per linear 30 centimeter of cordon (3S) or six
shoots per linear 30 centimeter of cordon …………….……………………….………..…….. 112

TABLE 47. Fruit chemical composition at harvest from experiment conducted at HTRC in 2013.
Shoot density was either adjusted to three shoots per linear 30 centimeter of cordon (3S) or six
shoots per linear 30 centimeter of cordon (6S). ………………………………………………. 113

xi

TABLE 48. Canopy measurements at harvest in field experiment conducted at SWMREC in
2013. …………………………………………………..………………………………….…… 124
TABLE 49. Yield components in Crop load Field Experiment from SWMREC at harvest in
2013. ………………………..…………………………………………………………………. 124

TABLE 50. Fruit chemical composition at harvest for Crop load Experiment at SWMREC in
2013 ...…………………………………………………………………………………………. 124

TABLE 51. Fruit chemical composition over time for HTRC field experiment in 2013. High
Wire Cordon (HWC) adjusted to either three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
………………………………………………………………………..………………………... 136

TABLE 52. Fruit chemical composition over time for HTRC field experiment in 2013Geneva
Double Curtain (GDC) adjusted to either three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
…………………………………………………………………………………………...…….. 137

TABLE 53. Fruit chemical composition over time for HTRC field experiment in 2013.
Experimental Moving Trellis (MT) adjusted to either three shoots per 30 linear centimeters of
cordon is represented by (3S) and the six shoots per 30 linear centimeters of cordon is
represented by (6S)……………………………………………………………………...…….. 138

TABLE 54. Fruit chemical composition over time for SWMREC field experiment in 2013.
..................................................................................................................................................…139

TABLE 55. Analysis of variance (ANOVA) for Total soluble solids (reported as Degrees Brix)
at HTRC in 2013. …………………………………………….………………….……………. 140
TABLE 56. Analysis of variance (ANOVA) for pH at HTRC in 2013. …………….…..…… 140
TABLE 57. Analysis of variance (ANOVA) for Titritable acidity at HTRC in 2013. …….. 140
TABLE 58. Analysis of variance (ANOVA) Total phenolic content at HTRC in 2013. …... 140

xii

TABLE 59. Analysis of variance (ANOVA) for Anthocyanin content at HTRC in 2013. ….. 140

TABLE 60. Analysis of variance (ANOVA) for Anthocyanin by Degrees Brix accumulation
content at HTRC in 2013 …………………………………………………………...………… 141
TABLE 61. Analysis of variance (ANOVA) for Shoot length at SWMREC in 2013 …….… 142

TABLE 62. Analysis of variance (ANOVA) for Lateral length at SWMREC in 2013 ….…. 142

TABLE 63. Analysis of variance (ANOVA) for Number of leaves per shoot at SWMREC in
2013 ………………………………………………………………………...……….………… 142

TABLE 64. Analysis of variance (ANOVA) for Leaf area at SWMREC in 2013 ……..……. 142

TABLE 65. Analysis of variance (ANOVA) for Leaf area by total shoot length at SWMREC in
2013 …………………………………………………………………………………………… 142

TABLE 66. Analysis of variance (ANOVA) for Ravaz Index at SWMREC in 2013 …….…. 143

TABLE 67. Analysis of variance (ANOVA) Yield (kilograms per vine) at SWMREC in 201..143
TABLE 68. Analysis of variance (ANOVA) Clusters per vine at SWMREC in 2013 …….... 143
TABLE 69. Analysis of variance (ANOVA) Clusters weight at SWMREC in 2013 …….… 143
TABLE 70. Analysis of variance (ANOVA) Berries per cluster at SWMREC in 2013 …….. 143
TABLE 71. Analysis of variance (ANOVA) Single berry weight at SWMREC in 2013 ...…. 144
TABLE 72. Analysis of variance (ANOVA) Total soluble solids at SWMREC in 2013 …… 144
xiii

TABLE 73. Analysis of variance (ANOVA) pH at SWMREC in 2013 ………………….….. 144
TABLE 74. Analysis of variance (ANOVA) Titritable acidity at SWMREC in 2013 ………. 144
TABLE 75. Analysis of variance (ANOVA) Phenolic content at SWMREC in 2013 ………. 144
TABLE 76. Analysis of variance (ANOVA) Anthocyanin content at SWMREC in 2013 … 145

TABLE 77. Analysis of variance (ANOVA) Anthocyanin by Degrees Brix accumulation content
at SWMREC in 2013 …………………………………………………………………………. 145

TABLE 78. Description of amount of fruit collected per treatment and wine produced per
treatment. ………………………………………………………………...…………………… 153

TABLE 79. A list of commonly found sensory characters from Marquette grapes. Table courtesy
of Dr. Katie Cook at University of Minnesota. …………………………....……………..…… 154
TABLE 80. Order of wines presented to the untrained panel of experts. ………………...… 154

TABLE 81. Chemical analysis of all wines produced from the Frost experiments conducted in
2012 and the crop load and shoot density experiments in 2013. ……………………..………. 159

TABLE 82. Wine tasting results of visual and aromatics analysis from produced from fruit on
vines that were trained to High Wire Cordon at HTRC in 2012. ………………………..…… 160

TABLE 83. Wine Tasting results of taste analysis from produced from fruit on vines that were
trained to High Wire Cordon at HTRC in 2012. ……………………………………………… 161

TABLE 84. Wine tasting results of visual and aromatics analysis from produced from fruit on
vines that were trained to Geneva Double Curtain at HTRC in 2012. …………….…………. 162

TABLE 85. Wine Tasting results of taste analysis from produced from fruit on vines that were
trained to Geneva Double Curtain at HTRC in 2012. …………………..……………………. 163

xiv

TABLE 86. Wine tasting results of visual and aromatics analysis from produced from fruit on
vines that were trained to Moving Trellis at HTRC in 2012. …………………………...……. 164
TABLE 87. Wine tasting results of taste analysis from produced from fruit on vines that were
trained to Moving Trellis at HTRC in 2012. …………………………………………………. 165

TABLE 88. Wine tasting results of visual and aromatics analysis from 2013 HTRC with three
shoots per linear 30 centimeters of cordon. ………………………...………………………… 166

TABLE 89. Wine tasting results of taste analysis from 2013 HTRC with three shoots per linear
30 centimeters of cordon. …………………………………..………………………………… 167

TABLE 90. Wine tasting results of visual and aromatics analysis from 2013 HTRC with six
shoots per linear 30 centimeters of cordon. ………………………….……………………….. 168

TABLE 91. Wine Tasting results of taste analysis from 2013 HTRC with six shoots per linear 30
centimeters of cordon. …………………………………………………….………...…...……. 169

TABLE 92. Analysis of variance (ANOVA) for Sweetness from wine produced from fruit on
HWC collected in 2012 at HTRC. …………………………………..……………..…………. 173

TABLE 93. Analysis of variance (ANOVA) for Color Intensity from wine produced from fruit
on GDC collected in 2012 at HTRC. …………………………………………….…………… 173

TABLE 94. Analysis of variance (ANOVA) for Color Hue from wine produced from fruit on
GDC collected in 2012 at HTRC. ……………………..……………...………………………. 173

TABLE 95. Analysis of variance (ANOVA) for Alcohol from wine produced from fruit on GDC
collected in 2012 at HTRC. ………...………………………………………………………… 173

TABLE 96. Analysis of variance (ANOVA) for Acidity from wine produced from fruit on GDC
collected in 2012 at HTRC. ………..…………………………………………………………. 173

TABLE 97. Analysis of variance (ANOVA) for Astringency from wine produced from fruit on
GDC collected in 2012 at HTRC. ……………………………...………….………………….. 174

xv

TABLE 98. Analysis of variance (ANOVA) for Body from wine produced from fruit on GDC
collected in 2012 at HTRC. ……………………………….………………………………….. 174
TABLE 99. Analysis of variance (ANOVA) for Color Hue from wine produced from fruit on
MT collected in 2012 at HTRC. ……………………………..…………………….…………. 174

TABLE 100. Analysis of variance (ANOVA) for Dark Fruit from wine produced from fruit on
MT collected in 2012 at HTRC. ……………………………...….…………………………… 174

TABLE 101. Analysis of variance (ANOVA) for Floral from wine produced from fruit on MT
collected in 2012 at HTRC. ……………….………………………………………………….. 174

TABLE 102. Analysis of variance (ANOVA) for Acidity from wine produced from fruit on MT
collected in 2012 at HTRC. …….…………………………………………………………….. 174

TABLE 103. Analysis of variance (ANOVA) for Astringency from wine produced from fruit on
MT collected in 2012 at HTRC. …………………………...…………………………………. 175

TABLE 104. Analysis of variance (ANOVA) for Color Hue from wine produced from fruit
collected in 2013 at three shoots per 30 linear centimeters of cordon. ……………………….. 176

TABLE 105. Analysis of variance (ANOVA) for Aromatics Intensity from wine produced from
fruit collected in 2013 at three shoots per 30 linear centimeters of cordon. ………………….. 176

TABLE 106. Analysis of variance (ANOVA) for Dark Fruit from wine produced from fruit
collected in 2013 at three shoots per 30 linear centimeters of cordon. ……………………….. 176

TABLE 107. Analysis of variance (ANOVA) for Vegetal from wine produced from fruit
collected in 2013 at three shoots per 30 linear centimeters of cordon. ………………………. 176

TABLE 108. Analysis of variance (ANOVA) for Floral from wine produced from fruit collected
in 2013 at three shoots per 30 linear centimeters of cordon. ………………………………. 176

TABLE 109. Analysis of variance (ANOVA) for Musty from wine produced from fruit collected
in 2013 at three shoots per 30 linear centimeters of cordon. …………………………...…….. 177

xvi

TABLE 110. Analysis of variance (ANOVA) for Alcohol from wine produced from fruit
collected in 2013 at three shoots per 30 linear centimeters of cordon. ………………………. 177
TABLE 111. Analysis of variance (ANOVA) for Acidity from wine produced from fruit
collected in 2013 at three shoots per 30 linear centimeters of cordon. ………….…………… 177

TABLE 112. Analysis of variance (ANOVA) for Body from wine produced from fruit collected
in 2013 at three shoots per 30 linear centimeters of cordon. …………………………………. 177

TABLE 113. Analysis of variance (ANOVA) for Color Hue from wine produced from fruit
collected in 2013 at six shoots per 30 linear centimeters of cordon. ………………...……….. 177

TABLE 114. Analysis of variance (ANOVA) for Aromatics Intensity from wine produced from
fruit collected in 2013 at six shoots per 30 linear centimeters of cordon. ……………………. 177

TABLE 115. Analysis of variance (ANOVA) for Dark Fruit from wine produced from fruit
collected in 2013 at six shoots per 30 linear centimeters of cordon. ……………….………… 178

TABLE 116. Analysis of variance (ANOVA) for Musty from wine produced from fruit collected
in 2013 at six shoots per 30 linear centimeters of cordon. ……………………………………. 178

TABLE 117. Analysis of variance (ANOVA) for Body from wine produced from fruit collected
in 2013 at six shoots per 30 linear centimeters of cordon. ………………...………………….. 178

xvii

LIST OF FIGURES

FIGURE 1: Map of areas suitable for grape production based on low midwinter temperatures and
historical experiences from variety trials. Areas shaded in grey are suited for Super Cold Hardy
grapes. Areas in blue are suitable for Super cold hardy and V.labruscana. Areas in purple are
suitable for Super cold hardy and V.labruscana and French-American Hybrids. Areas shaded in
pink are suitable for Super cold hardy and V.labruscana and French-American Hybrids some
varieties of V. vinifera. Reproduced with permission from Perry, Sabbatini and Burns (2012)
……………………………………………………………………………………………………..5
FIGURE 2: Genetic map of the complex genetic parentage of cultivar ‘Marquette’. Image
courtesy of www.chateaustripmine.com ...................................................................................... 12
FIGURE 3. Images of Vertical Shoot Positioning (VSP) Images courtesy Double A Vineyards.
https://www.doubleavineyards.com/news.aspx?showarticle=35 ……………...………………. 21
FIGURE 4. Images of High wire Cordon (HWC) …………………………………...………… 21
FIGURE 5. Image of Geneva Double Curtain (GDC). Image courtesy of Kentucky State
University.
http://www.pawpaw.kysu.edu/viticulture/information/design_your_vineyard.htm
…………………………………………………………………………………………………... 22
FIGURE 6. Image of Lyre Trellis Image courtesy of Kentucky State University
http://www.pawpaw.kysu.edu/viticulture/information/design_your_vineyard.htm .................... 22
FIGURE 7. Growing Degree Day (base 10oC) accumulations as recoded by the MAWN station
located at Horticulture Teaching and Research Center, Holt MI during the spring of 2012. ..… 31
FIGURE 8. Growing Degree Day (base 10oC) accumulations as recoded by the MAWN station
located at Southwest Michigan Research and Education Center, Benton Harbor, MI during the
spring of 2012. …………………………………………………………………………………. 31
FIGURE 9. Daily maximum and minimum air temperatures from 3/15/12 to 5/7/12 as recoded
by the MAWN station located at HTRC and at SWMREC during the spring of 2012. ……..... 32
FIGURE 10. Maximum and minimum air temperatures as recoded by the MAWN station located
at Horticulture Teaching and Research Center, Holt MI during the growing season of 2012…. 33

xviii

FIGURE 11. Maximum and minimum air temperatures as recoded by the MAWN station located
at Southwest Michigan Research and Education Center, Benton Harbor, MI during the growing
season of 2012 …………………………………………………………..……………………… 34
FIGURE 12. Moving Trellis (MT) design and details. Image on the left shows the details of two
brackets with a vine trained in between. Note the image is not to scale. Four adjustable catch
wires were used with two wires on each arm. The arms were allowed to be opened to a 45o angle
when opened. ……………………………………………………………………………...…… 35
FIGURE 13. Plot map for the experimental vineyard at HTRC. Layout for each trellis system
being evaluated: Geneva Double Curtain (GDC), High Wire Cordon (HWC), and Moving (MT)
…………………………………………………………………………………………...……… 43
FIGURE 14. Plot map showing vine locations and blocks in field location at SWMREC used for
the 2012 field experiments. ………………………………………………………………..…… 51
FIGURE 15. Total Soluble Solids (oBrix) over time for HTRC field experiment in 2012 Graph A
is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a secondary
bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. …………………………………………………………………………………….. 59
FIGURE 16. pH concentration over time for HTRC field experiment in 2012 Graph A is High
Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a secondary
bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. …………………………………………………………………………………….. 60
FIGURE 17. Titratable Acidity concentration over time for HTRC field experiment in 2012
Graph A is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C
is the experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a
secondary bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s
pairwise comparison. ………………………………………………………….……………….. 61
FIGURE 18. Phenolic concentration over time for HTRC field experiment in 2012 Graph A is
High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a secondary
bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. ……………………………………………………………………………..……… 62
FIGURE 19. Anthocyanin concentration over time for HTRC field experiment in 2012 Graph A
is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a secondary
xix

bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. …………………………………………………………………………………….. 63
FIGURE 20. Anthocyanin by Brix concentration over time for HTRC field experiment in 2012
Graph A is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C
is the experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a
secondary bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s
pairwise comparison. ……………………………………………………………………..……. 64
FIGURE 21. Total Soluble Solids (oBrix) over time for SWMREC field experiment in 2012.
Primary bud is represented by (P) and a secondary bud is represented by (S). Data points not
followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
…………………………………………………………………………………………...……… 70
FIGURE 22. pH over time for SWMREC field experiment in 2012. Primary bud is represented
by (P) and a secondary bud is represented by (S). Data points not followed by the same letter are
significantly different at P ≤ 0.05 by LSD’s pairwise comparison. ………….………………… 71
FIGURE 23. Titratable acidity concentration over time for SWMREC field experiment in 2012.
Primary bud is represented by (P) and a secondary bud is represented by (S). Data points not
followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
……………………………………………………………………..……………………………. 72
FIGURE 24. Phenolics concentration over time for SWMREC field experiment in 2012. Primary
bud is represented by (P) and a secondary bud is represented by (S). Data points not followed by
the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise comparison. …...…… 73
FIGURE 25. Anthocyanin concentration over time for SWMREC field experiment in 2012.
Primary bud is represented by (P) and a secondary bud is represented by (S). Data points not
followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
………………………………………………………………………………………...………… 74
FIGURE 26. Anthocyanin by Brix concentration over time for SWMREC field experiment in
2012. Primary bud is represented by (P) and a secondary bud is represented by (S). Data points
not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. ……………………………………….……………………………………………. 75
FIGURE 27. Growing Degree Day for HTRC. Data retrieved from MAWN site located at
station. ………………………………………………………………………………….………. 91
FIGURE 28. Growing Degree Day for SWMREC. Data retrieved from MAWN site located at
station. …………………………………………………………………………………..……… 92

xx

FIGURE 29. Total soluble solids over time for HTRC field experiment in 2013. Graph A is High
Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
(*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison. ……………. 114
FIGURE 30. pH concentration over time for HTRC field experiment in 2013. Graph A is High
Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
(*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison. ……………… 115
FIGURE 31. Titratable acidity over time for HTRC field experiment in 2013. Graph A is High
Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
(*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison. ………….…… 116
FIGURE 32. Phenolic concentration over time for HTRC field experiment in 2013. Graph A is
High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
(*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison. ……...……….. 117
FIGURE 33. Anthocyanin concentration over time for HTRC field experiment in 2013. Graph A
is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
(*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison. ………….…… 118
FIGURE 34. Anthocyanin by Brix concentration over time for HTRC field experiment in 2013.
Graph A is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C
is the experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of
cordon is represented by (3S) and the six shoots per 30 linear centimeters of cordon is
represented by (6S). (*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
…………………………………………………………………………………………………. 119
FIGURE 35. Daily average PAR for the different crop load treatments at SWMREC in 2013.
Measurements collected from the beginning of July to harvest. Control was quantum sensors
placed above the canopy of the vines as not to be affected by shade. …………………….. 125

xxi

FIGURE 36. Total soluble solids concentration over time for SWMREC field experiment in
2013. Data points where no later present show no significantly different at P ≤ 0.05.Data points
not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. ……………………………………………………………………………..…….. 126
FIGURE 37. pH concentration over time for SWMREC field experiment in 2013. Data points
where no later present show no significantly different at P ≤ 0.05. Data points not followed by
the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise comparison. …….… 127
FIGURE 38. Titratable Acidity concentration over time for SWMREC field experiment in 2013.
Data points where no later is present show no significantly different at P ≤ 0.05.Data points not
followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
………………………………………………………………………………………………… 128
FIGURE 39. Phenolic concentration over time for SWMREC field experiment in 2013. Data
points where no later is present show no significantly different at P ≤ 0.05.Data points not
followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
……………………………………………………………………………………………..….. 129
FIGURE 40. Anthocyanin concentration over time for SWMREC field experiment in 2013. Data
points where no later is present show no significantly different at P ≤ 0.05.Data points not
followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
……………………………………………………………………………………...………….. 130
FIGURE 41. Anthocyanin by Brix concentration over time for SWMREC field experiment in
2013. Data points where no later is present show no significantly different at P ≤ 0.05.Data
points not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison………………………………………………………………………………….…. 131
FIGURE 42. Example of score sheet used for experimental wine evaluation. Selections of
categories were based off from Marquette grapes. Information courtesy of Dr. Katie Cook at
University of Minnesota. ……………………………………………………………………... 155
FIGURE 43. Results from comparing wines produced from fruit collected from vines trained to
High Wire Cordon in 2012 .HWC (P) represents wine produced from fruit collected from shoots
that originated from the primary bud. HWC (S) represents wine produced from fruit that
originated from a secondary shoot. ………………………..………………………………….. 160
FIGURE 44. Results from comparing wines produced from fruit collected from vines trained to
Geneva Double Curtain in 2012. GDC (P) represents wine produced from fruit collected from
shoots that originated from the primary bud. GDC (S) represents wine produced from fruit that
originated from a secondary shoot. ……………………………………………..……….……. 162
xxii

FIGURE 45. Results from comparing wines produced from fruit collected from vines trained to
Experimental Moving Trellis in 2012. MT (P) represents wine produced from fruit collected
from shoots that originated from the primary bud. MT (S) represents wine produced from fruit
that originated from a secondary shoot. ……………………...……………………………….. 164
FIGURE 46. Results from comparing wines produced from fruit collected from vines used in
various trellis systems with a shoot density of three shoots per 30 centimeters of linear cordon in
2013. Vines trained to High Wire Cordon (HWC), Geneva Double Curtain (GDC), and
Experimental Moving Trellis (MT). …………………………………………..……………… 166
FIGURE 47. Results from comparing wines produced from fruit collected from vines used in
various trellis systems with a shoot density of six shoots per 30 centimeters of linear cordon in
2013. Vines trained to High Wire Cordon (HWC), Geneva Double Curtain (GDC), and
Experimental Moving Trellis (MT). ………………………………………………………….. 168

xxiii

CHAPTER I
LITERATURE REVIEW

Grapes:
The cultivation of grapes has been documented from as early as 4000 B.C. (Jackson,
1994). Grapes have been used for many purposes, ranging from a food source to being used for
production of various types of fermented beverages. According to data published by the Food
and Agriculture Organization from their 2009 survey, nearly 7.5 million hectares (ha) are under
cultivation worldwide. Of those hectares that are in production, the United States is considered
the fourth largest producer with 380,000 ha behind Spain, France, and Italy (FAO, 2009). On the
domestic side, the state of Michigan has 6,070 ha in commercial production as of 2011. Out of
these, 5,062 are used for the juice industry with the majority of comprised of plantings of
Concord or Niagara (USDA, 2012). The rest of the acreage is used for wine production with 293
ha planted as hybrids (interspecific crosses of Vitis labruscana, Vitis vinifera or other species of
Vitis).
Historically, the underlying theme of grape production for both the juice and wine
industry has been in the cultivation and manipulation of the vines in order to produce fruit with
the desired quality and yield set for each specific cultivar. This task has led to various techniques
and practices over the last 6,000 years. Aspects like site location, planting density, trellis system,
and variety have all had an impact on how the fruit is produced and how the vines are managed.
Several research efforts in analyzing different production factors have led to improvements that
have allowed the areas suitable for sustainable grape production to increase. Traditionally, the
only areas thought to be suitable for grape production were between 10o and 20o annual
isotherms (Jackson, 1994). These particular areas have similar environmental conditions as
1

found in the native range of Vitis vinifera. In order to overcome these limitations, several efforts
have been pursued over the years to develop and select cultivars that withstand particular
geographical problems. Seasonally low winter temperatures and the length of the growing season
are two of the more prominent geographical problems associated with new viticultural areas in
US, such as Michigan and the Great Lakes region. With all of this recent effort in research and
grape variety development, collaborative effort has formed between various agricultural
institutions throughout the Midwest and Northern states of the United State. This project known
as the Northern Grape Project includes various researchers and universities throughout Michigan,
New York, Connecticut, Iowa, Illinois, North Dakota, South Dakota, Nebraska and Minnesota.

Climate and varieties in Michigan
Michigan’s proximity to the Great Lakes (large thermal mass helping to moderate
temperature over the surrounding landmass) and the effect it has on climate, allows for grape
production only in specific areas (Perry, Sabbatini, and Burns, 2012) using Vitis labruscana or
Vitis vinifera (Figure 1). These areas adjacent to the Great Lakes have higher winter minimum
temperatures often above the threshold for vine damage. Currently, the majority of the grapes
being grown in the state fall into the areas on the western side of Michigan with close proximity
to Lake Michigan (Figure 1) and most of the ideal sites are already being utilized for commercial
production. There are many challenges that the grape vine must overcome in Michigan’s climate:
(a) low winter temperatures, (b) length of the growing season, (c) pests and diseases, (d) and
spring frost (Stergios and Howell, 1977; Zabadal and Andresen, 1997). The selection of a grape
variety adapted to our climate is still one of the most important viticultural issue that the young
and developing Michigan industry is facing and is still mainly driven by market demand more
2

than specific variety performance related to specific climatic conditions. After site selection, a
wine grape grower’s selection of a cultivar to plant is an all-important decision, which may be so
critical as to determine the outright survival of vines at the specific vineyard site. A priority
evaluation of: (a) potential for winter injury injury/damage, (b) potential yield and quality of
grapes and subsequent processing of these into quality wines and (c) potential market demand
and pricing for the grapes and wine. All these must be satisfactorily addressed before planting
any cultivars in challenging environments. In a new viticultural area, but specifically in
Michigan, some growers may choose cultivars based on industry trends, conversations and
literature and the resulting process is often speculative.
Different grape varieties with increased cold hardiness have recently been released and
are beginning to be adopted in Michigan, which are now in the early phase of production. These
varieties have been developed with emphasis on increasing the cold hardiness of the buds on the
vines by using selective breeding with specific genotypes that are already acclimated to
conditions where the environment is not suitable for others (Luby, 1991).
Michigan has a long history of growing grapes for commercial purposes. The first vines
planted in the state, used mainly for wine production, were in Monroe County in 1863
(Hathaway and Kegerreis, 2010). From the very first plantings of grapes in the state of Michigan,
viticulturists have been experimenting with different varieties and cultivars that can be grown
successfully under the state’s environmental conditions. Out of all of these challenges, the
limiting factor that viticulturists in this region currently face is the cold temperatures that occur
during the winter months and the consequent damages to vines.
In order to overcome damage from extremely low temperatures and frequent economic
crop losses, different species of Vitis were evaluated based on their ability to overcome winter
3

stress (Mills et al, 2006). Lethal winter temperatures at which 80% to 100% of primary buds
would be destroyed are represented in Table 1. Temperatures in various locations throughout
Michigan exhibit temperatures during the winter that would cause bud and trunk damage to
various varieties of grapes. As illustrated by the data in Table 1, temperatures that would cause
damage to the buds of V. vinifera would not be damaging to other cold hardy hybrids. For this
reason, the new plant material known as Super Cold Hardy hybrids (SCH) are gaining interest in
Michigan compared to the varieties of French-American hybrids due to their increased winter
hardiness, reliable concurrent productivity, and fruit technological maturity (chemistry makeup
and flavor profile) that determine high quality wines. In particular, when compared to the other
SCH and French-American hybrids that are used for commercial production, Marquette does not
have the characteristic “foxy” aroma profile as found in the other hybrids, and always recognized
as a negative wine attribute (Mansfield and Vickers, 2009)
In a recent survey of vineyards in Michigan (USDA, 2012) the percentage of hybrids has
changed from 625 acres in 2006 to 675 acres in 2011, an increase of 50 acres (10%) in 5 years or
10 acres per year. Of the varieties reported in the survey, Marquette has had the largest increase
of planting since introduction in 2006. As of 2011 according to the survey, 12 acres of Marquette
are in commercial production with more acres being planted (USDA, 2012); therefore of the 50
acres increase in hybrids planting, more than 25% was with Marquette vines. These data support
recent findings of states surveyed that are involved with the Northern Grape project that 55% of
the SCH varieties being planted are of the cultivar ‘Marquette’ (Tucker, 2012).

4

FIGURE 1: Map of areas suitable for grape production in Michigan based on low midwinter
temperatures and historical experiences from variety trials. Areas shaded in grey are suited for
Super Cold Hardy grapes. Areas in blue are suitable for Super cold hardy and V.labruscana.
Areas in purple are suitable for Super cold hardy and V.labruscana and French-American
Hybrids. Areas shaded in pink are suitable for Super cold hardy and V.labruscana and FrenchAmerican Hybrids and some varieties of V. vinifera. Reproduced with permission from Perry,
Sabbatini and Burns (2012).

5

TABLE 1. Approximate warmest temperature where 80-100% primary bud kill may be expected
to occur in midwinter. Elaborated from Wine Grape Production Guide for Eastern North
America. (2008). T. Wolf et al., Zabadal T., Sabbatini P., Elsner D., 2008. Wine Grape Varieties
for Michigan and Other Cold Climate Viticultural Regions. MSU Extension Bulletin CD-007.
Cultivar

Temperature

(Vinifera)

Cultivar

Temperature

(Hybrids)
(oF)

(oC)

(oF)

(oC)

Muscat Ottonel

-6

-20

Traminette

-20

-28

Merlot

-9

-21

Vidal blanc

-22

-30

Pinot gris

-10

-23

Chardonel

-22

-30

Pinot noir

-10

-23

Chambourcin

-23

-30

Sauvignon blanc

-10

-23

Seyval

-23

-30

Gewurztraminer

-12

-24

Vignoles

-26

-32

Chardonnay

-13

-25

Frontenac*

-35

-37

Riesling

-14

-25

Frontenac gris*

-35

-37

Cabernet Franc

-17

-27

Marquette*

-35

-37

* Super Cold Hardy variety

6

Challenges in Michigan
In Michigan, there are currently several problems that have caused distress for
viticulturists throughout the state. Most of these challenges have been caused by using varieties
and cultural practices that were designed and developed in areas outside of Michigan’s unique
climatic conditions. Throughout the literature, the primary problem related to a successful
vineyard has to do with the specific location and design, which has the most influence on the
establishment and sustainability of that vineyard (Stergios and Howell, 1977; Reynolds and
Vanden Heuvel, 2009). The majority of these concerns (i.e. frost injury, cold hardiness, and
drought conditions) can be related to the individual variety being grown at the specific location
(Reynolds and Vanden Heuvel, 2009). Once the problem of finding a suitable variety that can
maintain sustainable production at a given location is solved, specific concerns about canopy
development and the impacts it has on fruit chemistry become an issue (Reynolds et al., 1995).
Specifically in Michigan, the length of the growing season and the heat accumulation
during the ripening process are also limiting factors. Most grapes require a growing season
length of 165 days between the last freezing event in the spring and the first freezing event in the
fall (Zabadal and Andresen, 1997). This amount of time is typical for early ripening varieties of
V. vinifera. An advantage of the super cold hardy (SCH) varieties is a decrease in the length of
growing season needed for proper ripening of the fruit (University of Minnesota, 2012). With
this reduction of the seasonal growing length, locations within the state of Michigan that are not
suitable for other varieties can successfully produce fruit from SCH vines. Fruit grown under
conditions of a shorter season and/or cooler location than required, have often a lower
concentration of soluble solids and higher concentration of organic acids than desired (Reynolds
et al., 1986; Reynolds and Vanden Heuvel, 2009). This under-ripe fruit is caused by several
7

conditions such as, low temperatures within the canopy (Keller, 2010), lack of light exposure to
the cluster zone, vines that are out of balance (over cropped or under cropped), poor site
conditions, or poor vine management. Frequently, the essence of this problem can be related to
the temperature and light found at the level of the cluster (Keller, 2010). Temperature is involved
with metabolic activity of organic acids, phenol formation, and sugar accumulation (Boulton et
al., 1996). These environmental conditions have a direct effect on fruit composition and overall
fruit quality.
Varieties can be sensitive to specific locations due to cold damage that occurs during the
dormant season or from the shortened growing seasons, which in turn produces inferior fruit.
Combinations of poor management practices, ill-suited training structures, and variety selection
cause these undesired results from unsuitable varieties which results in poor fruit quality.
Fruit quality has been defined in several ways depending on the specific characteristics
that are of concern in a given viticultural area. When discussing berry composition, three major
aspects are identified by viticulturists and enologists: soluble solids (oBrix), acid concentration
(titratable acidity) and concentrations of phenols (total phenolics and anthocyanin content)
(Boulton et al., 1996). Each one of these parameters has been thoroughly documented to be
affected by different environmental conditions (Boulton et al., 1996; Bordelon et al., 2008).
Depending on the use of the grapes, different chemical compositions of the berries are needed in
order to produce the style of wine that is desired. These typical chemical composition values are
general guidelines for producing wine according to common styles (Table 2) (Boulton et al.,
1996).

8

TABLE 2. Common fruit quality parameters at the time of harvest as related to wine style. Data
for the table collected from Boulton et al., 1996.
Wine Style

oBrix

Titratable acidity

pH

(g tartaric/L)
Sparkling base

18.0 - 20.0

7.0 - 9.0

2.8 - 3.2

White Table

19.5 - 23.0

7.0 - 8.0

3.0 - 3.3

Red Table

20.5 - 23.5

6.5 - 7.5

3.2 - 3.4

Sweet Table

22.0 - 25.0

6.5 - 8.0

3.2 - 3.4

Dessert

23.0 - 26.0

5.0 - 7.5

3.3 - 3.7

9

Cold Hardiness Development of Super Cold Hardy Cultivar ‘Marquette’
The concept of breeding grapes to improve aspects of fruit quality or vine performance
has been occurring since before the 17th century (Alleweldt and Possingham, 1988). Since the
1860’s the main focus has been on both disease resistance and improving bud survival to cold
temperatures (Alleweldt and Possingham, 1988). There are currently two main species of Vitis
that are being studied to help with making improvements with cold hardiness; V. amurensis and
V. riparia (Alleweldt and Possingham, 1988, Hemstad and Luby, 2000). The reason behind the
use of V. riparia is the wide native range of climates that it can be found growing in within the
eastern portion of North America (Hemstad and Luby, 2000). It has been documented that buds
on V. riparia vines growing in the most northern portions of North America have been able to
survive temperatures as low as -40oC (Hemstad and Luby, 2000). The capability of developing
these new SCH varieties is due to easy hybridization of V. riparia with V. vinifera (Hemstad and
Luby, 2000). However, these native species usually produce fruit that has inferior fruit quality
and is not suitable for wine production purposes. In order to improve the fruit quality of these
native species and make the fruit quality comparable to traditional European wine grapes,
breeders have focused on increasing the amount of parentage of the European V. vinifera
(Hemstad and Luby, 2000). A large amount of these hybrids have been developed using genetic
material from wild V. riparia growing in areas that experience extreme cold winter temperatures.
In the case of the SCH material, substantial genetic material for those hybrids has come from
locations in Minnesota and Manitoba (Luby, 1991). With the recent release of new cultivars from
the University of Minnesota’s grape breeding program, the possibilities of areas suitable for
grape production in Michigan have increased dramatically.

10

Marquette is the most recent red variety released by the breeding program at the
University of Minnesota in 2006. The original cross for this plant took place in 1989 and was
given the experimental label as MN-1211 (US Patent #20070089208 was filed in 2005),
currently referred to as ’Marquette’ (Figure 2). This particular variety is a cross between MN
1094 x Ravat 262 (Smiley, 2008). Phenotypic traits typical of European varieties (i.e. upright
growth habit, compact clusters, and mild disease resistance) are displayed in Marquette due to a
high percentage of V. vinifera in the lineage. One of the most ideal characteristics of this new
variety is the improved cold hardiness, and this cultivar has been reported to survive
temperatures as low as -37.8o C (Smiley, 2008). Another benefit of Marquette is the perceived
wine. This cultivar has flavors similar to that of V. vinifera. A list of sensory descriptors for wine
produced from Marquette has been developed by the University of Minnesota (Cook, 2013,
personal communication).

11

FIGURE 2: Genetic map of the complex genetic parentage of cultivar ‘Marquette’. Image
courtesy of www.chateaustripmine.com

12

Viticultural Problems with Super Cold Hardy (SCH) Varieties
Scarce information has been published on the viticultural and enological performance due
to their recent introduction of SCH varieties to the industry. The major problems associated with
SCH varieties are controlling canopy size, yield, and consequently canopy density. With the
recent release of the plant material, not much information is available about cultural practices
that allow the vine to produce the fruit to a quality that is suitable for wine production. This
includes information on trellis systems, canopy architecture, and crop level and the impacts all
three aspects have on fruit quality. However, there have been numerous publications on the
effects of factors that affect fruit quality in grapes, especially on the relationships of light
exposure and temperatures, which in turn have a direct impact on wine quality (Smart, 1985).
Light and temperature in the fruit zone are related to both the trellis system and the canopy
architecture (Dokoozlian and Kliewer, 1995). In order to understand this relationship several
items must be examined such as canopy density, size, and location of fruit zone in regards to the
rest of the plant canopy.
The first aspect of this relationship between fruit chemistry and exposure has to do with
the canopy architecture. Canopy density is an important way of examining the concern about the
vigor of SCH varieties and the impacts on fruit quality. As mentioned by Smart (1985), the
amount of leaf area within a defined space is known as canopy density. Achieving the proper
amount of cluster exposure in order to allow for light interception by the fruit (Reynolds et al,
1996) is controlled by the canopy density. This plays an important role in the chemical
parameters of the fruit throughout the season. The amount of the canopy and canopy density can
be used to look at the amount of shading that occurs around the fruiting zone. The ideal canopy
would not contain interior leaves and have a leaf area to surface ratio close to 1.0 (Smart, 1985).

13

This ratio means that no interior leaves of the canopy are shaded by other leaves. This concept by
Smart provides a metric for helping decrease the amount of leaf layers or the amount of leaves
within the same defined area of the canopy.
In order to achieve this ratio with the SCH grapes, the trellis design must be adapted
from what is currently being used. Current commercial practice has been to train Marquette to a
High Wire Cordon (HWC) system. This type of training system is used due to the perceived idea
of the recumbent or downward growth habit of hybrids (Jackson, 1994). It has been observed
that the SCH cultivar ‘Marquette’ tends to have an upright growth habit similar to V. viniferia
(Luby, 1991). Due to this field observation Marquette could be trained to a system that allows for
the shoots to grow upright similar to a trellis system like Vertical Shoot Positioning (VSP). But,
this system has been shown to not be the most efficient training system for canopy light
interception (Dokoozlian and Kliewer, 1995), but it is still the most widely used in cool climate
regions due to the ease of pruning. This research will provide results that should suggest to the
use a training system that has a higher efficiency in light interception at the level of leaves and
light exposure at the level of the fruit zone. Pervious work conducted on other grape varieties has
shown light exposure and light inception increase desirable fruit chemical parameters (Smart,
2008). It has been widely documented that a vertically divided canopy allows for a smaller ratio
of leaf area to canopy surface area than a horizontal divided canopy (Smart 1985, Gladstone and
Dokoozlian, 2003). Because this increases the volume allowed for canopy growth (Reynolds and
Heuvel, 2009), these divided canopies allow for ratios of leaf area to surface area that are closer
to the ideal ratio described by Smart in 1985. It has also been mentioned by the same literature
that the level of exposure to sunlight of the fruit is increased.

14

The factor of light of exposure at the cluster level is also influenced by the amount of
vigor or size of the canopy. When a vineyard is experiencing a high amount of vigor, the vines
will tend to have longer shoots with a vast amount of lateral growth (Keller, 2010). Excessive
lateral growth is detrimental because it increases the amount of leaf area in the canopy, which in
turn affects the amount of light reaching the clusters. This increase of fruit shading has a
negative impact on fruit quality (Dokoozlian et al., 1995).
Research to improve the light reaching the fruit includes studies on trellis and training
systems that have a differential impact on fruit exposure to sunlight and consequently on
temperature (Zoecklein et al, 2008). There is a strong relationship between light in the canopy
and temperature of the cluster, and this affects fruit chemistry (Bergqvist et al 2001). Increasing
cluster temperatures affects the acid profile of the fruit (Dokoozlian et al., 1995). All of this
evidence is important because it is alleged that due to the high amount of V. riparia in the
genetics of Marquette and other SCH varieties, the acidity of the fruit at the time of harvest is
high, which is a negative attribute for wine production.

15

Fruit Chemistry of SCH
With the vast amount of interspecific crosses made with several species of Vitis to create
SCH varieties such as Marquette, the specific chemistry reflects the genetic makeup (Figure 2).
Currently, there are no peer reviewed publications on the particular fruit composition of the
cultivar ‘Marquette’. However, there are observational notes from the breeders at the University
of Minnesota on the subject (Luby, 1991). The fruit produced by Marquette has characteristics
that have been described by Hemstad and Luby as having small clusters (89 g/cluster) with small
sized berries (averaging 1.1 g/ berry and 12 mm in diameter). The V. riparia parentage is
reflected in the chemical composition of the fruit, which tends to have high concentrations of
soluble solids (averaging 25.9° brix) with high levels of acid (12.0 g/l of titratable acidity),
which results in low pH juice of about 3.0). These values are considered to be far outside ideal
winemaking parameters for the production of red wine. Characteristic parameters such as the
dark color of the grape, the juice, the highly acclaimed quality of the wine produced, and the
short growing season required for proper ripening of the fruit has drawn attention from the grape
industry not only in Michigan but throughout the Midwest.
Many papers have been written about the chemical composition of other varieties that are
interspecific crosses of various species of Vitis, with some being similar to SCH varieties. The
main sugars in V. vinifera are glucose and fructose, but grapes that have been bred with V.
labruscana or V. rotundifolia have higher concentrations of sucrose (Liu et al, 2006). The two
dominant acids are malic and tartaric, which comprise 90% or more of the acids found in the
fruit (Liu et al, 2006). The ratio of the specific concentration of the acids varies in specific
cultivars (Boulton et al, 1996). According to Keller (2010), malic acid can be reduced in the field
by metabolic processes due to increased cluster temperatures up to 37 oC. Tartaric acid cannot be

16

degraded in the field (instead is managed in the cellar) and the lowering concentration of the acid
in the berries during ripening is generally due to a dilution affect (Keller, 2010; Boulton et al.,
1996). Acid is important in these SCH varieties because of the higher ratio of acids to sugar than
compared to other varieties. This explains the difficulty of producing wines from SCH grown
fruit. From research conducted on 98 different cultivars of grapes, the acid profile in the grapes
varied depending on the specific variety (Liu, 2006). From that data set, V. vinifera grapes
generally had a tartaric acid to malic acid ration of two to one while the hybrids of V. vinifera
crossed with other species of grapes had the acid ratio three parts tartaric acid to one malic acid
(Liu, 2006).

17

Trellis Systems and Impact on Fruit Quality of Super Cold Hardy Varieties
An extensive amount of research has been done on training systems for V. vinifera and
French-American Hybrids. Consistently, throughout the literature, all of the different systems
can be split into four categories (Reynolds et al., 1996). These categories include: 1) head/cane
2) head/spur 3) cordon/cane and 4) cordon/spur. Within each of these categories, a collection of
different training systems is filled to each depending on how the vines are pruned and supported
(Reynolds et al., 1995).
Trellis systems have significant effects on several aspects of the vine, which relate back
to the amount of light interception occurring in the canopy (Reynolds et al, 1996). In particular,
the trellis system can control the amount of light reaching the fruit. Specific quality parameters
(sugar concentrations, anthocyanin concentrations, total phenolic content and acid levels) have
been compared in fruit grown under well exposed conditions to fruit under shaded conditions
(Dokoozlian and Kliewer, 1995). Light interception by the canopy is influenced by several
factors including planting density, row orientation, trellis height, row width, site conditions, and
time of year (Reynolds et al., 1996).
Currently, there are several trellis systems being used for commercial production in the
Eastern United States. French-American hybrids, V. labruscana, and other downward growing
varieties are trained to High Wire Cordon (HWC), which is the most common in the Midwest
region (Bordelon et al, 2008). Michigan has several types of training systems used for
commercial purposes. Head/cane, cordon/cane, and cordon/spur are the most common categories
that fit into the description by Reynolds (2009). Geneva Double Curtain (GDC) and Scott Henry
(SH) are both used in Michigan, but only few acres exist. Details about VSP, HWC, and GDC
and split canopies will be described in the following pages.

18

Vertical Shoot Positioning (VSP)
The most widely used training systems for wine production in Michigan is known as
Vertical Shoot Positioning (VSP). This form of training is a type of cordon/spur system with
spurs spread throughout an established cordon in which the spurs usually have three buds left per
spur. Figure 3 shows the dimensions as well as the fruit zone in this type of training system. In
this system, the shoots and canopy are above the fruit zone. The upward growth habit of V.
vinifera lends itself to the use of VSP systems. The benefits of this system are the relatively easy
training and management practices used for this type of system. This type of trellis system is
common among different viticultural areas around the world, which has led to development of
fully mechanized systems for management of these vineyards. One drawback is that it is the most
inefficient training system in light interception (Dokoozlian and Kliewer, 1995). This is caused
by the angle of the sunlight in relation to vine canopy. In areas that are geographically flat (like
Michigan) this trellis system causes shading to occur while the sun is between the hours of 11 am
to 1 pm. In locations that are geographical hilly the VSP system allows for better light exposure
and easier following of the contour of the landscape.
High Wire Cordon Trellis (HWC)
The majority of acres under grape production in Michigan use a system known as HighWire Cordon (HWC). HWC is also a form of the cordon/spur system in which the spurs are left
along an established cordon while allowing the shoots to grow in a downward direction as seen
in Figure 4. The downward growth habit of V. riparia, V. labruscana, and most FrenchAmerican hybrids is just one reason for using the HWC. This training system is widely used in
Michigan because of the low cost of installation, mechanization of these vineyards, and the
simple management of this system.

19

Geneva Double Curtain Trellis (GDC)
The Geneva Double Curtain Trellis was first described by Nelson Shaulis in the 1960’s at
the New York State Agricultural Experiment Station (Shaulis et al., 1966). The concept is to
allow for a larger canopy and increased exposure to light on a particular vineyard site to help
improve fruit quality, yield, process of ripening, and overall plant health. As evidenced by
several studies, the use of a divided canopy system has various modifications and designs but all
rely on increasing canopy size by splitting the canopy to increase light interception. This system
is considered to be a divided canopy system since the canopies are divided horizontally. An
example of this trellis can be seen in Figure 5. A pitfall of this trellis system is the increased cost
of installation, increased labor, and the lack of mechanization options for these vineyards.
Split Canopy (Y- Trellis)
Some of the most uncommon styles of trellises are known as the Lyre, U-shape, and Vshape. The use of a divided canopy can help to improve the light captured by the canopy, which
has been shown to have effect on fruit quality and the process of ripening. Figure 6 shows
examples of these lesser used trellis systems. Several vineyard sites in Michigan are using a
divided canopy site, generally, these sites are using the split canopy trellis system to help control
vine vigor or increase the overall yield per vine.

20

FIGURE 3. Images of Vertical Shoot Positioning (VSP) Images courtesy Double A Vineyards.
https://www.doubleavineyards.com/news.aspx?showarticle=35

FIGURE 4. Images of High wire Cordon (HWC).

21

FIGURE 5. Image of Geneva Double Curtain (GDC). Image courtesy of Kentucky State
University. http://www.pawpaw.kysu.edu/viticulture/information/design_your_vineyard.htm

FIGURE 6. Image of Lyre Trellis Image courtesy of Kentucky State University
http://www.pawpaw.kysu.edu/viticulture/information/design_your_vineyard.htm

22

Conclusion
The first and most fundamental question that may be posed for a new viticultural region
concern is genotype adaptation to the environmental conditions of that region. The questions
posed and the answers produced may be beneficial on several levels. The most obvious is the
determination that a valuable, previously non-cultured cultivar is adapted to the region. Less
often appreciated is the finding that an "unknown" or "new" cultivar, breeding, or clonal
selection in not adapted and costly trial and error planting and culture by vineyard managers can
be avoided.
In parallel with the field adaptation of a wine grape genotype are the twin economic
factors of processed quality and economically acceptable levels of yield per unit land area. For a
new genotype to be acceptable, therefore, it must survive, and produce economically acceptable
levels of crop that can be processed into value added production that is competitive in the
marketplace. Key environmental factors for the Eastern U.S. and especially in the Great Lakes
region are frequency and severity of winter cold episodes, sufficient length and heat unit
accumulation of the growing season, and frequency of damaging spring freeze episodes.
Finally, there is a need for an assessment of relative susceptibility to prevalent pest
problems of the region. These may be closely associated with both economic yield and
subsequent processed quality. As chemical options for vine protection decline, this factor will
become of increasing importance. The most important vine adaptation is also the production in
the area of fruit with optimal technological maturity at harvest (Palliotti et al, 2013). Fruit
chemical composition at harvest is related to several aspects that include the amount of fruit on
the vine (yield), the amount of sunlight that reaches the fruit (exposure level) and the
temperatures related to this exposure level. Viticulture research has been done on various species
of Vitis and hybrids of these species related to different aspects of the impacts of trellis design,
23

canopy architecture, and yield. The importance of studying the fruit chemical composition of
SCH cultivars will allow for better determination and validation of cultural practices in the
vineyard.
Lack of viticultural data for SHC include aspects of yield to leaf ratio which in turn leads
to the size of canopy that is needed to achieve the proper fruit chemistry and also amount of light
exposure that is needed by the fruit. Also basic viticulture knowledge like amount of yield per
training system, canopy density in terms of shoot number for length of cordon, bud fertility and
finally how to control high level of organic acids that have been observed in SCH varieties, is
needed.
Applying concepts and practices that have been previously studied in other cultivars
could be a foundation for improving cultural practices for this new class of SCH varieties. In
order to generate valid viticulture information, several field experiments were designed to
investigate the effects of trellis system in SHC, since it has the largest impact on light
interception in the fruit zone (Dokoozlian et al, 1995). Three trellis systems have been chosen to
be evaluated comparatively. Those include the HWC, GDC, and the new experiential trellis
(Moving Trellis or MT). In order to help improve the fruit quality at the time of harvest, MT is
based off a Y-shaped trellis similar to the one described by Palliotti (2011) and Takeda (2003).
The benefit of this new system would be both increased yield and improved ripening which in
turn would yield lower acid and increase the fermentable sugars in the berries. The proposed
research is to increase the amount light that reaches the fruit zone and change the temperatures of
the clusters; which in turn will cause increased metabolism of the malic acid concentration in the
fruit and increased synthesis of total phenolic (specifically anthocyanin) concentration. These
two factors (light and temperature) will be affected most by the trellis systems of the vine.
24

Selecting a particular trellis design allows for manipulation of the canopy; in particular the
canopy density (number of shoots per meter of cordon). Changing the canopy density affects the
amount of leaf layers in the trellis (Smart, 2008). The number of leaf areas can show how much
shade or light will be received by the cluster. Canopies with high amounts of leaf layers will
have more shading than canopies with lower amount of leaf layers. By changing the amount of
leaf areas allow for the adjustment of the amount of light that reaches the fruiting zone. To be
able to accomplish this, the grapevine canopy structure needs to be managed in a way to increase
exposure to light that the clusters are receiving while at the same time intercepting enough light
for carbohydrate production. The literature has shown the positive effects of increasing light
exposure at the cluster level have on fruit composition (Dokoozlian and Kliewer, 1995). In order
to achieve this, a new trellis will be evaluated based on fruit quality and the ripening curves of
the different canopies. The concept that a vine with a higher amount of leaf area creates more
shading in the fruiting zone and thus creating under ripe fruit will also be investigated by
comparing the different trellises. These experiments will provide a basis for understanding the
impact of the trellis and cultural practices on aspects of vine growth, fruit ripening, and fruit
chemical composition.
An additional concern has arisen due to the early bud-break nature of the particular
variety ‘Marquette’, already reported by growers and in trade journals. With the effects of the
weather in Michigan over the last few seasons and the unusual warmth in the beginning (March
and April) has led to spring frost damage to the vines. These frost events have caused damage to
different amounts of meristem tissue within the compound buds on the vines. After reviewing the
literature, little if any information is known on aspects of phenology, morphology or fruit quality
between fruit that has been produced from a shoot that has grown from the primary bud as

25

compared to fruit that has been produced by the secondary bud that has developed as a result of
the primary bud being damaged or destroyed. This new information will be important to
locations that are practicing cool/cold climate viticulture and subjected to frequent frost events,
such as Michigan.

26

CHAPTER II
EVALUATION OF EFFECTS OF TRELLIS SYSTEM ON DAMAGE AND
FRUIT QUALITY OF PRIMARY FRUIT VS. SECONDARY FRUIT IN
VARIOUS LOCATIONS IN MICHIGAN AFTER FROST DAMAGE
Introduction
In the spring of 2012, unusual warming events occurred starting from the 6th to the 28th of
March. According to the National Weather service, the average temperature during the month of
March was over 10o Celsius for Southwest Lower Michigan (Marino, 2012). Historically in this
portion of Michigan, the average temperature during March is approximately 1.7o C, which is
below the active temperature for grape development (Keller, 2010). The days from March 6th to
the 28th were continuously above the historical temperatures by as much as 20 degrees. During
these climatic events, the vines at Southwest Michigan Research and Extension Center
(SWMREC) and the Horticultural Teaching and Research Center (HTRC) were exposed to 125.8
growing degree days (GDD base 10o C) and 107.0 GDD, respectively. The GGD and ten year
average of GDD for each location for each vineyard location during this time period (Figure 7
and Figure 8). These amounts of heat units were adequate to break ecodormancy and cause bud
break on 3/25/2012 and 3/28/2012. Historically bud break occurs for grapes in these two
locations around the end of April. From April 1st to April 29th, seven frost events were recorded
at the vineyard located at HTRC and six frost events at SWMREC (Figure 9). Based from the
data recorded by the weather stations at both locations, the vines were subjected to similar
situations and conditions (Figure 10; Figure 11). For SCH varieties, like Marquette, that contain
high amounts of genetic material in their genetic parentage from V. riparia, these varieties tend
to have early bud break compared to other species like V. labrusca and V. vinifera. This greatly

27

increased the probability of damage due to frost events during the early part of the growing
season.
Buds on grapes, regard less of being a V. vinifera or SCH variety are unique compared to
other fruit due to the complexity of the bud. The inside of a grape bud consists of three structures
known as the primary, secondary and tertiary primordial (Keller, 2010). Due these unique
structures the bud is considered a compound bud. The primary primordial when beginning to
grow will produce a fruitful shoot. This is the first of the three structures to begin to grow when
environmental conditions are favorable and will be referred to as the primary shoot throughout
this thesis. If this structure is damaged the secondary primordial will begin to grow to replace the
damaged primary primordial, which will also produce a fruitful shoot. Pervious works have
found that the clusters associated with this structure are smaller than the fruit produced on a
primary shoot (Keller, 2010). In the case where both the primary and secondary primordial are
damaged the third structure in the bud, the tertiary primordial will produce a shoot without
clusters.
After conducting a literature review, it was determined that no research has been
conducted on Marquette grapevines on the relationship between the fruit chemistry that results
from a cluster attached to a primary shoot and from a cluster that was produced by a secondary
shoot. There is even less information known about how the individual shoots and the overall
canopy growth and phenological development perform associated with the origins of the shoot
Increased interest in planting SCH cultivars in Michigan where injurious frosts are
experienced as in 2012, will demonstrate their future economic value. There is a pivotal need to
determine if these SCH varieties can provide sustainable vine growth and fruit production after
experiencing frost events. The experiments conducted during the growing season of 2012 gave
28

us insight into how this particular SCH cultivar reacts to challenging spring environmental
conditions experienced in Michigan.
Several trellis systems were chosen to be studied for this experiment. The three trellis
systems chosen were High Wire Cordon (HWC), which is the industry standard for hybrids,
Geneva Double Curtain, which was developed by Nelso Shaluis in the 1960’s to help improve
light exposure to the fruit zone, and a third experimental moving trellis (MT), which is a
combination of lyre trellis system and a vertical shoot positioning system that can be manually
folded to convert the trellis from a divided system into a single canopy system. The MT trellis
was designed based on the trellis system designed and used by Palliotti (2011) which allowed the
canopy of the grape vines to be moved as needed. This system increase light interception,
controls vigor, improves fruit quality at the time at harvest, and increased yield. This type of
trellis is being referred to as MT (for moving trellis) since the trellis is not a static system and
allows for the movement of the canopy by adjusting the position of the shoots contained in this
system. This type of trellis system will improve fruit quality and yield by bridging the gap of a
divided canopy system that can be managed like a VSP system. In this type of trellis the system
was able to function as a divided canopy and then could be closed or ‘compressed” in one single
vertical canopy. The dimensions and descriptions of these particular trellis systems can be found
in Figure 12. The moving trellis system was designed to be folded by hand in order to allow for
equipment to pass through the vineyard plot and to allow for easier manipulation of the canopy
when manual work was performed.
The objective of this work was to evaluate the impact of trellis systems on the aspects of
frost damage and the fruit produced after significant damage has occurred to the vines. Due to
the frost events that occurred, two populations of shoots were studied from the time of bud break
29

until harvest of the fruit. The different fruit produced was also evaluated for wine quality
between fruit that originated from the primary bud or fruit that originated from the secondary
bud.

30

FIGURE 7. Growing Degree Day (base 10oC) accumulations as recoded by the MAWN station
located at Horticulture Teaching and Research Center, Holt MI during the spring of 2012.

FIGURE 8. Growing Degree Day (base 10oC) accumulations as recoded by the MAWN station
located at Southwest Michigan Research and Education Center, Benton Harbor, MI during the
spring of 2012.
31

HTRC

SWMREC

40

40
Air temp max
Air temp min

Air temp max
Air temp min
30

Air temp (oC)

Air temp (oC)

30

20

10

0

20

10

0
BUD BREAK

3/28/2012
-10
3/12 3/19 3/26 4/2

BUD BREAK

4/9

4/16 4/23 4/30

3/25/2012
-10
3/12 3/19 3/26 4/2

5/7

4/9

4/16 4/23 4/30

5/7

FIGURE 9. Daily maximum and minimum air temperatures from 3/15/12 to 5/7/12 as recoded
by the MAWN station located at HTRC and at SWMREC during the spring of 2012.

32

HTRC

air temp max
air temp min

30

o

Air Temperature ( C)

40

20

10

0

-10

BUD BREAK

ANTHESIS

VERAISON

HARVEST

60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250

Julian Date
FIGURE 10. Maximum and minimum air temperatures as recoded by the MAWN station located
at Horticulture Teaching and Research Center, Holt MI during the growing season of 2012

33

SWMREC
40
air temp max
air temp min

o

Air Temperature ( C)

30

20

10

0

-10

BUD BREAK

ANTHESIS

VERAISON

HARVEST

60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 210 220 230 240 250

Julian Date
FIGURE 11. Maximum and minimum air temperatures as recoded by the MAWN station located
at Southwest Michigan Research and Education Center, Benton Harbor, MI during the growing
season of 2012

34

FIGURE 12. Moving Trellis (MT) design and details. Image on the left shows the details of two
brackets with a vine trained in between. Note the image is not to scale. Four adjustable catch
wires were used with two wires on each arm. The arms were allowed to be opened to a 45o angle
when opened.
35

Materials and Methods
Field Experiment – Horticulture Teaching and Research Center (HTRC)
Plant Material
Marquette, an interspecific hybrid species of Vitis, where planted in 2008 in the
experimental and teaching vineyard located at the Horticulture Teaching and Research Center
(HTRC) of Michigan State University, located at 3291 College Road in Holt, Michigan. The
experimental vineyard is located on the southern portion of Michigan State University’s campus,
and it is a one acre plot with 35 rows, 61 meters long. Two rows were planted with Marquette
vines, with a spacing of 3.1 meters between rows and 2.4 meters between vines and a consequent
planting density of 545 vines per acre. All of the vines were trained to High Wire Cordon (HWC)
with the use of multiple trunks to ensure survival from damage at the time of establishment. In
preparation for the experiments to be conducted during the growing season of 2012, two thirds of
the vines were converted into two different trellis systems in the winter of 2011, while the vines
were still dormant. After the conversion was completed each row contained three blocks with
seven vines within each block trained to either high wire cordon (HWC), which is considered the
industry standard for growing hybrids in this region, Geneva Double Curtain (GDC) and the
experimental moving trellis (MT). Guard vines were trained to the trellis systems that were
adjacent to the vines to allow for separation from each block. The MT system was opened and
closed a maximum of four times during the season.
After these frost events had occurred, two separate populations of shoots were left on the
vines. The populations included primary shoots that survived the frost events and secondary buds
that developed as a result of the primary shoots being damaged or killed by the exposure to low
temperatures. These two populations were studied in order to assess the habit of canopy growth,
berry development, and fruit quality over time. The parameters (total soluble solids (TSS), pH,
36

titratable acidy (TA), total phenolic compounds, and anthocyanin content) were used to assess
fruit quality and were determined by collecting representative samples across the several
treatments at SWMREC and HTRC.
Adopted viticultural practices (mineral nutrition, green pruning, and pest management)
for commercial vineyards in Michigan were used. Climatological data were recorded during the
experiment by an automated weather station located within the farm 100 m from the
experimental vineyard.
Experimental Design
The experiment at HTRC was designed as a split-plot design with two replications. The
main factor for this experiment is trellis system with three different types of trellises being used
as previously mentioned. Each row being a replication (Row 30 = Replication 1, Row 31 =
Replication 2) with the trellis systems converted during the winter of 2011. The vines were split
into the different trellis systems at the time of retraining with no training system being directly
across from one and another (Figure 13). The vines at the end of each row (position 1 and
position 25) were trained to the same trellis system that was next to the experimental block.
These vines, in addition to vines in position 9 and position 17, were treated as outside guard
vines and no data were collected from these vines. Details about each vine position, the trellis the
vine is trained to, the purpose of the vine regarded to the experiment, and the data collected from
each vine are described (Table 3).
After the several frost events, a second factor (bud type) was added to the experiment.
Each vine was evaluated and all of the buds on the vine were categorized as either primary bud
that survived frost events or secondary bud. Within each experimental vine, three shoots were

37

selected to represent the shoots that had grown from the primary bud. Additionally, three shoots
were selected that had grown from the secondary bud. These shoots were selected from different
positions of each vine to allow for a representative sample for each type of shoot. Throughout the
experiment the selected shoots are referred to as model shoots. These six shoots were tagged for
identification purposes and were used for representative samples of each population of shoot
type found in the canopies of the experimental vines. These shoots were also used for repeated
measurements throughout the growing season. Environmental data, such as growing degree days
(GDD) and daily precipitation (Table 4) were measured using the Michigan Automated Weather
Network (MAWN) automated weather station. Major phenological stages were also recorded and
are presented with the date of observation and Julian date (Table 5).
Frost Damage Evaluation
In the first week of May 2012, after the threat of frost events had ceased, all of the vines
within the experimental plot were evaluated for bud damage resulting from the series of frost
events. During this time, each vine was examined in order to count the number of buds on the
vine, the number of primary buds that were exposed to the frost events, and the number of buds
that were killed from the frost events. All surviving primary buds were marked using flagging
tape in order to keep track of the shoots and fruit as they developed during the growing season.
Three weeks later the secondary buds were also flagged to allow for identification of shoot and
fruit throughout the season. All of the bud counts were used to calculate the amount of damage
that occurred per trellis system (Table 6).

38

Canopy Growth Measurements
After fruit were set on the primary and secondary shoots, three shoots from each
population were tagged per vine to be used as model shoots that represent the vine. These tagged
shoots were tracked to allow for the collection of shoot length during the rest of the growing
season and used for all measurements until harvest. The canopies of the different trellises were
monitored weekly, and at that time measurements were taken. Shoot lengths were measured after
anthesis until just after veraison, and at that time the growing shoot tips had stopped growing. To
conduct these measurements a measuring tape was used to measure shoot length from the base of
the model shoot to the active growing tip.
Berry Growth Measurements
Berry growth was also measured weekly by using the same tagged model shoots on each
vine within the experiment. The measurements of berry growth were obtained by taking the
diameter of three random berries from the apical cluster on the tagged model shoots. A digital
caliper was used for data collection. These measurements were carried out until harvest.
Sampling Procedures and Harvest Data Collection
At the time of veraison starting on the fruit produced by a primary bud, berry samples
were collected to be analyzed in the laboratory at a later date. The procedure for sampling was
carried out by collecting 30 selected berries from tagged primary shoots that were not being used
as model shoots per vine in each treatment. Berries were collected from different parts of the
clusters to insure a representative sample per vine. These berries were then transferred into
plastic bags and placed into a -20o C freezer until analysis was performed. This same process was

39

repeated for fruit on the secondary shoots. Samples were collected weekly and amounted to three
until harvest.
At the time of harvest, each experimental vine was hand harvested. Fruit from the tagged
model shoots were collected and placed into plastic bags. All of the collected samples were
frozen and were used for chemical analysis in the laboratory at a later date. Once the samples
from the tagged shoots were collected, the rest of the vine was harvested with care to separate the
fruit depending on the type of shoot that produced the fruit. All of the clusters from each type of
shoot (primary and secondary) were counted and weighted in order to calculate yield per vine. At
harvest, a representative sample of fruit was collected per trellis system and per shoot type to be
used for the experimental wine making process. These samples consisted of a maximum of 6
kilograms of fruit per combination of trellis system and shoot type. A total of six different
combinations of wines were produced. Details about the experimental wine making process and
procedures can be found in Chapter 4. During the following winter of 2012, the vines were
pruned to a bud count of 100 buds and all of the one-year old wood was collected and weighed to
obtain pruning weights per vine and calculate the Ravaz Index (Ravaz, 1911) or RI as an index
of vine balance.
Basic Fruit Chemistry Measurements
Samples were allowed to warm to room temperature before being placed into a 50 ml
centrifuge tube and homogenized using a Brinkmann Homogenizer (Brinkmann Instruments,
Westbury, NY). After the sample was homogenized, approximately 1 g ± 0.05 g of the
homogenate was weighed and placed into a 15 ml centrifuge tube to be used for anthocyanin and
total phenolic assay as described by Iland et al. (2004). Using a UV-Vis spectrophotometer

40

(Model UV-1800, Shimadzu Corporation, Kyoto, Japan) absorbance readings were measured at
wavelengths of 280nm, 520nm, and 700nm as called for in the procedure.
The remaining sample of the homogenate in the 50 ml tube was centrifuged using a
Legend X1R centrifuge (Thermo Scientific, Waltham, MA) for 10 minutes at 4000 rpm.
Subsequently the supernatant was transferred to a new 50 ml centrifuge tube to be used for
performing the analysis of pH, titratable acidity, and total soluble solids of the clarified juice. A
small portion of clarified juice was placed into the well of a digital refractometer (ATA-3810
PAL-1 Puls Inc., Van Nuys, CA) and used for analysis of total soluble solids which was reported
as Degrees Brix. The pH of the juice was measured using a 370 Thermo Orion pH meter
(Thermo Fisher Scientific Inc., Logan, UT), which was placed into the centrifuge tube with
enough juice to cover the tip of the pH probe. Titratable acidity (TA) was determined by
measuring the left over clarified juice and then diluting the juice with 100 ml of deionized water
to be analyzed using Multi-T 2.2 digital titrator (laboratory Synergy Inc., Goshen, NY). This
solution was titrated using 0.1N sodium hydroxide (NaOH) to the endpoint of a pH of 8.2. The
amount of sodium hydroxide needed to reach this endpoint was recorded and the values recorded
were inputted into the equation to determine the TA as expressed as g/L (Iland et al. 2004).
Equation 1.
TA (g/L as tartaric acid) = 75*0.1(Normality of NaOH)*Titre value (ml) / (volume of juice used)

Statistical Analysis
Basic statistics, analysis of variance, regression, and correlation analysis were performed
using SAS version 9.3 for Windows (Cary, NC). Data were analyzed by using two-way ANOVA
appropriate for a fully randomized design for determination of the factors of trellis system
41

(HWC, GDC, MT) and bud type (Primary, Secondary) effects. Results were tested for normality
and homogeneity of variance prior to being subjected to F-test (p≤0.05).

42

FIGURE 13. Plot map for the experimental vineyard at HTRC. Layout for each trellis system
being evaluated: Geneva Double Curtain (GDC), High Wire Cordon (HWC), and Moving (MT)

43

TABLE 3. Details about specific vines used in the experiment at the HTRC location.
Location

Row 30

Row 31

Vine Position

Trellis Type

Vine Purpose

1

GDC

Guard

2 thru 8

GDC

Experimental

9

GDC/HW

Guard

10 thru 16

HW

Experimental

17

HW/MT

Guard

18 thru 24

MT

Experimental

25

MT

Guard

Primary shoots
Secondary shoots
No data collected, used as transitional
vine between trellis systems
Primary shoots
Secondary shoots
No data collected, used as transitional
vine between trellis systems
Primary shoots
Secondary shoots
No data collected

1

MT

Guard

No data collected

2 thru 8

MT

Experimental

9

MT/GDC

Guard

10 thru 16

GDC

Experimental

17

GDC/HW

Guard

18 thru 24

HW

Experimental

25

HW

Guard

44

Details & Data Collected
No data collected

Primary shoots
Secondary shoots
No data collected, used as transitional
vine between trellis systems
Primary shoots
Secondary shoots
No data collected, used as transitional
vine between trellis systems
Primary shoots
Secondary shoots
No data collected

TABLE 4. Growing Degree Days and precipitation weather data measured for 2012 from
MAWN located at HTRC.
Parameter
GDD (base
10oC)
Precipitation
(mm)

March
107.0

April
48.8

May
214.5

June
303.0

July
434.6

August
322.5

September
200.1

Total
1630.5

63.0

53.1

72.9

28.4

63.7

49.3

59.4

389.9

TABLE 5. Major Phenological stage, Date of occurrence and Growing Degree Days for
Marquette grown at HTRC in 2012.
HTRC - 2012 Major Phenological stage, Date of occurrence and Growing Degree Days
Bud
Anthesis Anthesis PeaPeaVeraison Veraison Harvest
break
(P)
(S)
size
size
(P)
(S)
(P)
(S)
Date
3/28/12 5/31/12
6/6/12
6/7/1 6/21/12
7/14/12
7/25/12
8/18/12
2
Julian
87
151
157
158
172
195
206
230
Date
GDD
107.0
370.6
402.0
409.6 570.0
879.2
1034.4
1292.1
(base
10oC)
Days
64
70
71
85
108
119
143
from
Bud
break

TABLE 6. Bud evaluation after frost event at HTRC in 2012.
Trellis Total Number
Percent of
System
of buds
Primaries alive
HWC
176 a*
19.9 a
GDC
196 a
17.5 a
MT
223 a
21.9 a
*Treatment means within columns not followed by the same letter are significantly different at P
≤ 0.05 by LSD’s pairwise comparison.

45

Field Experiment – Southwest Michigan Research Education and Center (SWMREC)
Plant Material
Vitis spp. ‘Marquette’ vines were planted in 2008 in the experimental vineyard located at
Southwest Michigan Research and Education Center (SWMREC) in Benton Harbor, MI. The
vines were planted in a random block design as part of the NE1020 hybrid variety trial. The
NE1020 is a Northeast Regional Research project that is a multi-state evaluation of wine grape
cultivars and clones which has several research institutions including Michigan State University.
Within the trial, the vines are grouped into blocks of four vines of the same variety in the same
row. Six replicate blocks of Marquette were established with a total of 24 vines in production
(Figure 14). Own-rooted vines were planted on a spacing of 2.7 m between rows with 2.1 m
between vines at a planting density of 691 vines per acre.
Previously collected soil sample tests show the vineyard is planted in Spinks sandy loam.
All vines had been trained to High wire cordon (HWC) with the use of multiple trunks to ensure
survival from winter damage. All vineyard management practices were conducted similar to the
rest of the experimental vineyard. Environmental measurements were gathered from the
Michigan Automated Weather Network (MAWN) weather station located at SWMREC. This
tool was used to collect data on daily temperatures, growing degree days (GDD), and daily
precipitation

(Table

7).

Other

environmental

data

can

be

accessed

from

http://www.agweather.geo.msu.edu/mawn/.
Experimental Design
The design for this particular experiment was set up as a random complete block design
(RCBD) with the use of one type of trellis system (High Wire Cordon) for all vines planted in the
vineyard. The treatment for this experiment was bud type which was caused by the frost events.
46

Due to the early warmth experienced in the beginning of the 2012 growing season, the initial
experiment was modified to adapt to the frost damage. The frost damage caused by the cold
temperatures in April caused two populations of buds (primary and secondary). These shoots that
resulted from either a primary bud or a secondary bud were flagged using the flagging tape and
were assigned as the treatments for this experiment. Each vine within the plot was evaluated on
the damage and the two populations of shoots were allowed on all vines.
Within the vineyard, six blocks were planted in the RCBD with each block containing
four vines in each replication. All of these vines were previously established in 2008. For this
experiment, only four blocks out of the six were used due to vine loss in the blocks on the south
side of the plot. Therefore, 16 total vines were used. The Marquette vines used each had a total
of six tagged shoots that were followed throughout the growing season. Within these six tagged
shoots, three of them were shoots that come from the primary bud that had survived the frost
events and the other three shoots originated from the secondary bud that replaced the primary
bud after being damaged by the frost. The tagged shoots were used to represent the two sets of
shoot populations within each vine. All of the vines in the four blocks used had model shoots.
The plot map for this experiment (Figure 13) shows where the panels of vines were located in
relation to the rest of the research plot. Phenological data were measured for major stages of
growth throughout the season. The data are reported with the GDD and Julian date of observed
occurrence (Table 8).
In the beginning of May 2012, all of the vines within the vineyard were examined to
record the amount of damage caused by the several frost events (Table 9). The primary shoots
that were not damaged by the frost were flagged to help identify the difference from the primary
shoots and the secondary shoots that would appear later in the season. After fruit set had finished
47

for both the clusters on primary shoots and the clusters on secondary shoots, three shoots from
each population were tagged per vine to be used as representative model shoots for the vine to be
tracked during the rest of the growing season. These shoots were randomly selected and used for
all measurements from fruit set to harvest.
Canopy Growth Measurements
Weekly monitoring of shoot length of the model shoots (which are the six shoots chosen
to represent the two populations of shoots growing within the vines canopy) started from anthesis
until veraison. The collection of shoot length measurements was interrupted at veraison because
the canopy was fully developed and shoot growth had ceased at that time. To conduct these
measurements a measuring tape was used to measure from the base of the model shoot to the
active growing shoot tip.
Sampling Procedures and Harvest Data Collection
At the time veraison completed on fruit that originated from the primary bud, berry
samples were collected to be analyzed in the laboratory at a later date. The procedure for
sampling was carried out by collecting 30 selected berries from tagged primary shoots that were
tagged differently than the model shoots. The fruit attached to the model shoots was not sampled
to allow for cluster morphology data to be collected on the whole cluster at the time of harvest.
Care was taken to select berries from different parts of the clusters to insure a representative
sample per vine. These berries were then transferred into plastic bags and placed into a -20o C
freezer until analysis was performed. This same process was repeated for fruit on the secondary
shoots. The whole process was continued weekly until harvest.

48

At harvest, fruit was collected from the tagged model shoots and placed into plastic
collection bags. All of the collected samples were used for chemical analysis in the laboratory at
a later date. Once the samples from the tagged shoots were collected, the rest of the vine was
harvested with care to separate the fruit depending on the type of shoot that had produced the
fruit. All of the clusters from each type of shoot; primary and secondary; were counted and
weighed to calculate yield per vine. This process was repeated until harvest was completed. At
the time of completion of harvest, a representative sample of fruit was collected per shoot type to
be used for experimental wine making process. This sample consisted of a maximum of 6
kilograms of fruit per shoot type. Two different experimental wines were produced; wine made
from only fruit produced on primary shoots and wine made from fruit produced from only
secondary fruit. Details about the experimental wine making process and procedures can be
found in Chapter 4. During the following winter of 2012, the vines were pruned to a bud count of
100 buds and all of the one-year old wood was collected and weighed to obtain pruning weights
per vine and calculate the Ravaz Index (Ravaz, 1911).
Basic Fruit Chemistry Measurements
At the time of analysis, previously collected samples were tested using the same set of
procedures and techniques as used to evaluate the samples collected from field experiment
conducted at the HTRC. Details about how this was accomplished can be found in the section of
Basic Fruit Chemistry Measurements under the heading of Field Experiment – HTRC for 2012
(page 40).
Statistical Analysis
Based off of the experimental design, the main experimental factor of bud type was
examined. The data collected from each shoot type (Primary or Secondary) was compared by
49

time which was the repeated measure unit. Data measured throughout the course of the growing
season was tested using a repeated measure analysis of variance (ANOVA). All of the
parameters measured were checked for normality and model testing performed. After all of the
data were deemed to fit the normality curve, ANOVA was performed using PROC MIXED. All
statistical analysis was conducted using SAS version 9.3 for Windows (Cary, NC). All
parameters were tested using Least Significant Difference of the means (LSD) with all pair-wise
comparisons performed with the alpha equal to 0.05. For parameters measured at the time of
harvest, similar statistical Analysis was performed with the exception of the repeated measure
unit removed from the model. All parameter testing were conducted using SAS version 9.3 using
LSD with all pair-wise comparisons performed with the alpha equal to 0.05.

50

FIGURE 14. Plot map showing vine locations and replications (Block #) in field location at
SWMREC used for the 2012 field experiments.

51

TABLE 7. Monthly Growing degree days and precipitation data measured for 2012 from
MAWN located at SWMREC.
Parameter
GDD (base
10oC)
Precipitation
(mm)

March
130.6

April
54.3

May
207.4

June
332.8

July
478.0

August
332.6

Sept.
208.8

Total
1744.3

48.8

52.1

87.6

18.3

75.2

120.7

100.1

502.6

TABLE 8. Major Phenological stage, Date of occurrence and Growing Degree Days for
Marquette grown at SWMREC in 2012.
Phenological
Stage
Date
Julian Date
GDD (base
10oC)
Days from
Bud break

Bud break

Anthesis (P)

Veraison (P)

Veraison (S)

Harvest

3/25/2012
84
125.8

5/25/2012
145
344.1

7/13/2012
194
939.9

7/25/2012
206
1127.3

8/30/2012
242
1521.3

-

61

110

122

158

TABLE 9. Bud evaluation of primary buds by block after frost event at SWMREC in 2012.
Trellis
System

Block #

Total Number
Percent of
of buds
Primaries alive
1
108 a*
49.7 a
2
142 a
59.9 a
3
144 a
51.3 a
HWC
4
152 a
57.3 a
5
129 a
60.8 a
6
107 a
41.2 a
*Means within columns not followed by the same letter are significantly different at P ≤ 0.05 by
LSD’s pairwise comparison.

52

Results – 2012 HTRC
Weather during the Growing Season
The abnormal warm temperatures that started on March 6th and continued until March
30th caused the Marquette vines to break eco-dormancy and begin growing earlier than
previously recorded for this location. This large amount of heat accumulation caused bud break
to occur on March 28th, when the vines had experienced 107 growing degree-days (Base 10o C).
The ten-year average for the occurrence of this amount of heat units typically occurs around late
April or early May (Figure 9).
After this warming event, the temperatures returned to the historical average seasonal
temperatures for this location. During the period of April 7th to April 24th the growing shoots that
originated from the primary bud were subjected to several frost events. The temperatures during
these frost events were observed to be as low as -5o C (Figure 7). Starting in May, the threat of
frost events occurring greatly diminished and at that time all of the Marquette vines were
evaluated based on the amount of damage that had occurred. At the time of evaluation, each
vine within each trellis system was checked to obtain a count of the number of shoots that
originated from the primary buds that were damaged by the frosts. The observations from the
field showed that there was no statistical difference between the three trellis systems based on
the amount of bud-shoot survival (Table 5).
Several days later the vines were re-examined in order to tag all of the shoots that started
growing from the secondary bud to replace the killed primary bud or shoot. These populations of
shoots and the fruit associated with each one were tracked throughout the growing season as
described in the material and methods sections. This was done to investigate the respective
effects that the fruit originating from a primary bud and the fruit originating from a secondary
53

bud had on aspects of canopy development and structure, fruit chemistry, and impacts on wine
quality.
Fruit Chemistry during the Ripening Process
From veraison on the fruit that originated from a primary bud, weekly sampling of the
fruit was measured to examine the ripening profile of each type of fruit produced (primary, or
secondary) and with respect to each trellis system (HWC, GDC, MT). After veraison had
finished on the fruit originating from primary shoots, there was a statistical difference of the
basic fruit chemistry between the primary fruit and the secondary fruit. This held true for all of
the parameters of total soluble solids (Figure 12; Figure 13), pH (Figure 14; Figure 15) and
titratable acidity (Figure 16; Figure 17). This trend continued up until the time of harvest, at
which time only the total soluble solids was not statistically significant (Figure 12). With respect
to phenolic content, a difference was observed at the first sampling date (Figure 18) for all trellis
systems. There was a statistical difference observed only on the Geneva Double Curtain trellis on
the 8/15/12 sampling date, but there was no difference between any of the samples found at the
time of harvest a week later (Figure 19). Anthocyanin concentration followed a comparable
pattern with only a difference between the samples occurring on the first two sampling dates.
Similarly, no difference was observed at the time of harvest (Figure 20; Figure 21). When the
rate of sugar accumulation was related to anthocyanin production, differences between samples
were observed in all three trellis systems on the first two sampling dates and only again on the
8/15/12 sampling date for the MT trellis system (Figure 22: Figure 23).

54

Fruit Chemistry and Yield Components at Harvest
Harvest was conducted with all of the vines being hand harvested in order to collect data
on both by vine bases and by bud type. This allowed for the fruit to be counted, weighted, and
sampled separately. The two sets of data were then combined to obtain information on a vine
basis. There were no significant differences in yield components when all six of the parameters
(Total Yield, Total Number of Clusters, Total Cluster Weight, Total Berries per Cluster, Pruning
Weights, and Ravaz Index) were compared to each other on a whole vine basis (Table 6). This
did not hold true when the data was separated by bud type for each trellis system. Yield, total
number of clusters per bud type, cluster weight, total berries per cluster, and single berry weight
were all affected when separated by bud type at the time of harvest (Table 7). This same trend
was also observed when comparing the fruit chemistry parameters by the whole vine per trellis
system. Total soluble solids, phenolic content and anthocyanin by brix ratio were affected when
all of the data was pooled together (Table 8). When the data was examined by trellis and then by
bud type, significant differences were observed in all aspects expect for anthocyanin
concentration and anthocyanin by brix ratio (Table 9). Data for all of the fruit chemistry
parameters by trellis system and sampling date can be found in Appendix B.
Results on Phenology
Throughout the growing season, the phenological stages of the two populations of fruit
were followed from bud break until harvest. At the time of anthesis, there was only six day
difference between the primary buds and secondary buds (Table 4) This difference in days
increased to fourteen during the time of anthesis and by the time the berries reached pea-size (3
to 4 mm in diameter). By the time of veraison there was an eleven day difference from the

55

primary fruit changing color compared to the secondary fruit. All of the fruit was harvested on
the same date as called for in the harvest procedures.

56

TABLE 10. Yield components at harvest by vine in field experiment conducted in 2012 at
HTRC.
Trellis
System

Total
Yield
(Kg/vine)

Total
Number of
clusters

Total
Total
Pruning
Ravaz
Cluster
berries
weights
Index
weight
per
(kg)
(g)
cluster
HWC
3.30*
67
62.9
60
0.93
3.50
GDC
3.20
69
53.4
54
1.02
3.75
MT
3.53
75
58.0
62
1.12
3.84
*Treatment means within columns not followed by the same letter are significantly different at P
≤ 0.05 by LSD’s pairwise comparison. Columns where no letters are present indicate a lack of
significant differences among treatments.

TABLE 11. Yield components at harvest by bud type in field experiment conducted in 2012 at
HTRC.
Trellis
System

Bud Type

HWC

Primary

GDC

MT

Yield

Number
of clusters

Cluster
weight (g)

Berries
per
cluster

Single
berry
weight (g)

0.31 a*

6a

59.2 ab

55 ab

1.03 a

Secondary

2.99 b

62 b

63.2 a

62 a

1.00 a

Primary

0.28 a

7a

53.4 ab

55 ab

0.94 ab

Secondary

2.92 b

62 b

58.5 ab

61 ab

0.97 ab

Primary

0.11 a

2a

45.4 b

45 b

0.99 ab

Secondary

3.41 b

73 b

61.0 ab

67 a

0.91 b

(Kg/Vine/
bud type)

*Treatment means within columns not followed by the same letter are significantly different at P
≤ 0.05 by LSD’s pairwise comparison. Columns where no letters are present indicate a lack of
significant differences among treatments.

57

TABLE 12. Fruit chemical composition at harvest by trellis system from field experiment conducted at HTRC.
Trellis
System
HWC

TSS
(oBrix)
19.5 b*

pH

TA (g/L)

3.38

9.22

Phenolics
(a.u./g)
0.90 b

Anthocyanin Anthocyanin
(mg/g)
/brix
0.91
0.047 ab

GDC

21.4 a

3.32

9.44

1.05 a

0.92

0.044 b

MT

19.7 b

3.35

9.77

0.96 b

1.01

0.051 a

*Treatment means within columns not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. Columns where no letters are present indicate a lack of significant differences among treatments.

TABLE 13. Fruit chemical composition at harvest by bud type from field experiment conducted at HTRC.
Trellis
System
HWC

GDC

MT

Primary

20.3 abc*

3.50 a

8.66 a

Phenolics
(a.u./g)
0.97 ab

Secondary

19.7 bc

3.32 b

9.86 b

1.01 ab

1.01

0.050

Primary

21.7 a

3.46 a

8.68 a

1.03 ab

0.93

0.045

Secondary

21.5 a

3.27 b

10.02 b

1.07 a

0.98

0.044

Primary

20.9 ab

3.47 a

8.14 a

0.94 b

1.07

0.045

Secondary

19.1 c

3.32 b

10.48 b

0.96 ab

0.99

0.052

Bud Type

TSS

pH

TA (g/L)

(oBrix)

Anthocyanin
(mg/g)
0.91

Anthocyanin
/brix
0.042

*Treatment means within columns not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. Columns where no letters are present indicate a lack of significant differences among treatments.

58

24

*

22

*

TSS (oBrix)

20
18
16

A

*

14
12

HWC (P)
HWC (S)

10
8
7/23/12

7/30/12

8/6/12

24

8/13/12

8/20/12

*

22

*

TSS (oBrix)

20
18
16

B

*

14
12

GDC (P)
GDC (S)

10
8
6
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

24

*

22

*

TSS (oBrix)

20
18
16

C

*

14
12
10

MT (P)
MT(S)

8
6
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

FIGURE 15. Total Soluble Solids (oBrix) over time for HTRC field experiment in 2012 Graph A
is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a secondary
bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s pairwise
comparison.

59

3.6

*
*

3.4

*

A

pH

3.2

3.0

*
HWC (P)
HWC (S)

2.8

2.6
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

3.6

*
3.4

*
*

3.2

B

pH
3.0

*
GDC (P)
GDC (S)

2.8

2.6
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

3.6

*
3.4

*

C

pH

3.2

*
3.0

MT (P)
MT(S)

2.8

2.6
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

FIGURE 16. pH concentration over time for HTRC field experiment in 2012 Graph A is High
Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a secondary
bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s pairwise
comparison.
60

26

*

24

HWC (P)
HWC (S)

22

TA (g/L)

20
18

A

*

16
14

*

12

*

10
8
6
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

26
24

*

GDC (P)
GDC (S)

22

TA (g/L)

20
18

B

*

16
14

*

12

*

10
8
6
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

26
24

*

MT (P)
MT(S)

22

TA (g/L)

20
18

*

C

16
14

*

12

*

10
8
6
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

FIGURE 17. Titratable Acidity concentration over time for HTRC field experiment in 2012
Graph A is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C
is the experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a
secondary bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s
pairwise comparison.
61

1.8

Phenolics (a.u./g)

1.6

*
HWC (P)
HWC (S)

1.4

A

1.2
1.0
0.8
0.6
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

1.8

Phenolics (a.u./g)

1.6

*

GDC (P)
GDC (S)

1.4

B

1.2

*
1.0
0.8
0.6
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

1.8
1.6

MT (P)
MT(S)

Phenolics (a.u./g)

*
1.4

C

1.2
1.0
0.8
0.6
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

FIGURE 18. Phenolic concentration over time for HTRC field experiment in 2012 Graph A is
High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a secondary
bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s pairwise
comparison.
62

1.4

HWC (P)
HWC (S)

Anthocyanin (mg/g)

1.2
1.0
0.8
0.6
0.4

A

*
*

0.2
0.0
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

1.4

GDC (P)
GDC (S)

Anthocyanin (mg/g)

1.2
1.0
0.8
0.6
0.4

B

*
*

0.2
0.0
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

1.4

MT (P)
MT(S)

Anthocyanin (mg/g)

1.2
1.0
0.8

C

0.6
0.4

*
*

0.2
0.0
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

FIGURE 19. Anthocyanin concentration over time for HTRC field experiment in 2012 Graph A
is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a secondary
bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s pairwise
comparison.
63

0.07

HWC (P)
HWC (S)

Anthocyanin / Brix

0.06
0.05
0.04
0.03

A

*

*
0.02
0.01
0.00
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

0.07

GDC (P)
GDC (S)

Anthocyanin / Brix

0.06
0.05
0.04
0.03

B

*

*
0.02
0.01
0.00
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

0.07

MT (P)
MT(S)

Anthocyanin / Brix

0.06

*

0.05
0.04

C

0.03
0.02

*

*

0.01
0.00
7/23/12

7/30/12

8/6/12

8/13/12

8/20/12

FIGURE 20. Anthocyanin by Brix concentration over time for HTRC field experiment in 2012
Graph A is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C
is the experimental Moving Trellis. In all graphs the primary bud is represented by (P) and a
secondary bud is represented by (S). (*) Means significantly different at P ≤ 0.05 by LSD’s
pairwise comparison.
64

Results – 2012 SWMREC
Weather during the Growing Season
Similar to the weather pattern observed at the HTRC vineyard location, the same
unseasonable warmth in March also affected this location in the Southwest part of the state, in
the heart of one the most important Michigan appellation, the Lake Michigan Shore. This
weather caused a large amount of heat units to accumulate unusually early in the season, causing
the vines to break eco-dormancy and start growing earlier than the historical average. Bud break
for this location occurred on March 25th, at which point 125.8 growing degree-days (base 10o C)
had occurred (Table 11). In the same pattern of weather that occurred at the HTRC location, the
weather in April returned toward the historical average range, which subjected the growing
shoots to several frost events (Figure 11). These frost events caused damage to the buds and
shoots that were growing at the time.
In the beginning of May, after the threat of frost had significantly decreased, the vines
were evaluated to determine the amount of damage that occurred in the experimental vineyard.
From the field data, all of the damage was similar between the six blocks of Marquette planted
throughout the variety trial at SWMREC. The average percentage of primary buds damaged was
50 percent (Table 12). At the time of evaluation, all of the primary shoots that survived were
flagged to allow for identification purposes throughout the rest of the growing season. The same
information that was measured at HTRC was also measured at SWMREC.
Fruit Chemistry during the Ripening Process
When the fruit that originated from the primary bud had started veraison, sampling of the
experimental vines occurred to collect both the fruit from the primary fruit (P) and the fruit that
65

originated from the secondary bud (S), which pushed after the primary bud was damaged or
killed. Differences in total soluble solids were observed for all sampling dates including the
sampling that occurred at harvest (Figure 24). A similar trend of statistically significant
differences between the fruit chemistry parameters was also observed for pH (Figure 25) and for
titratable acidity (Figure 26). Phenolic content was only statistically different at the first
sampling date and was not observed again (Figure 27). Anthocyanin concentration, however,
was different at the first sampling between the two populations of fruit and again at the time of
harvest (Figure 28). When the rate of total soluble solids was compared to anthocyanin
concentration, a significant difference was not observed throughout the season, but a difference
was identified at the time of harvest (Figure 29).
Fruit Chemistry and Yield Components at Harvest
In a similar fashion to the experiment conducted at HTRC, the SWMREC experiment
was hand harvested in order to get data by bud type and on a whole vine basis. The data of yield
components, when pooled together on per vine basis, revealed no differences between
experimental blocks. However, a difference was observed once the data were separated by bud
type. The average of all of the blocks can be found in Table 13.
Cluster weight, berries per cluster, rachis weight, and single berry weight were all
significantly different (Table 14) between bud types, but there was no difference between the
number of clusters per vine or yield. The fruit that was on the secondary clusters did have more
berries on each cluster and had a heavier overall cluster weight than the primary clusters. In
terms of the fruit chemistry at the time of harvest, no difference was found between experimental
blocks when the data was pooled together by vine (Table 15). However, every parameter but
phenolic content was affected by bud type (Table 16). The fruit of the secondary cluster had
66

lower sugar content, lower pH, increase in acid content, and decrease in anthocyanin content. All
of the fruit chemistry data by sampling date can be found in Appendix B.
Results on Phenology
In the same way that the HTRC vines had various phenological stages tracked to monitor
the differences between the two populations of shoots and fruit associated with the different bud
types. The difference at veraison was twelve days between the primary and secondary fruit
(Table 11). This was very similar to the difference observed at the HTRC location.

67

TABLE 14. Yield components at harvest by vine in Frost evaluation field experiment conducted in 2012 at SWMREC.
Yield
(Kg/Vine)

Number
of cluster

Cluster
weight
(g)

Berries
per
cluster

Rachis
weight
(g)

Single
berry
weight (g)

Berry
diameter
(mm)

111

50.7

54

2.70

0.87

10.98

4.62

TABLE 15. Yield components at harvest by bud type in Frost evaluation field experiment conducted in 2012 at SWMREC.
Bud Type

Yield
(Kg/Vine/Bud
Type)

Number
of
cluster

Cluster
weight
(g)

Berries
per
cluster

Rachis
weight
(g)

Average
Berry
weight
(g)

Average
Berry
diameter
(mm)

Primary

2.35*

61

37.3 a

48 a

2.26 a

0.75 a

10.67

Secondary

2.26

56

56.0 b

58 b

2.83 b

0.91 b

11.30

*Treatment means within columns not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. Columns where no letters are present indicate a lack of significant differences among treatments.

68

TABLE 16. Fruit chemical composition at harvest by vine from Frost evaluation field
experiment conducted in 2012 at SWMREC.
TSS

pH

TA
(g/L)
7.42

(oBrix)
26.3

3.55

Phenolics
(a.u./g)
1.02

Anthocyanin Anthocyanin
(mg/g)
/ brix
0.69
0.028

TABLE 17. Fruit chemical composition at harvest by bud type from Frost evaluation field
experiment conducted in 2012 at SWMREC.
Bud Type
Primary

TSS
(oBrix)
28.9 a*

Secondary

24.8 b

pH
3.61 a

TA
(g/L)
7.01 a

Phenolics
(a.u./g)
0.99

Anthocyanin
(mg/g)
0.49 a

Anthocyanin
/ brix
0.020 a

3.51 b

7.59 b

1.02

0.81 b

0.033 b

*Treatment means within columns not followed by the same letter are significantly different at P
≤ 0.05 by LSD’s pairwise comparison. Columns where no letters are present indicate a lack of
significant differences among treatments

69

SWMREC 2012
30

*

28
26

22

*

20

o

TSS ( Brix)

24

18

*

16
14
12
(P)
(S)

10
8
7/23

7/30

8/6

8/13

8/20

8/27

9/3

Date
FIGURE 21. Total Soluble Solids (oBrix) over time for SWMREC field experiment in 2012.
Primary bud is represented by (P) and a secondary bud is represented by (S). Data points (*)
different at P ≤ 0.05 by LSD’s pairwise comparison.

70

SWMREC 2012
3.8

*
3.6

*

pH

3.4

3.2

*
3.0

(P)
(S)

2.8

2.6
7/23

7/30

8/6

8/13

8/20

8/27

9/3

Date
FIGURE 22. pH over time for SWMREC field experiment in 2012. Primary bud is represented
by (P) and a secondary bud is represented by (S). Data points (*) different at P ≤ 0.05 by LSD’s
pairwise comparison.

71

SWMREC 2012
24
22

*

(P)
(S)

20

TA (g/L)

18
16

*

14
12
10

*

8
6
7/23

7/30

8/6

8/13

8/20

8/27

9/3

Date
FIGURE 23. Titratable acidity concentration over time for SWMREC field experiment in 2012.
Primary bud is represented by (P) and a secondary bud is represented by (S). Data points (*)
different at P ≤ 0.05 by LSD’s pairwise comparison.

72

SWMREC 2012
1.5

1.4

*

(P)
(S)

Phenolics (a.u./g)

1.3

1.2

1.1

1.0

0.9

0.8
7/23

7/30

8/6

8/13

8/20

8/27

9/3

Date
FIGURE 24. Phenolics concentration over time for SWMREC field experiment in 2012. Primary
bud is represented by (P) and a secondary bud is represented by (S). Data points (*) different at
P ≤ 0.05 by LSD’s pairwise comparison.

73

SWMREC 2012
1.0
0.9

*

Anthocyanin (mg/g)

0.8
0.7
0.6
0.5
0.4

*

0.3
0.2
(P)
(S)

0.1
0.0
7/23

7/30

8/6

8/13

8/20

8/27

9/3

Date
FIGURE 25. Anthocyanin concentration over time for SWMREC field experiment in 2012.
Primary bud is represented by (P) and a secondary bud is represented by (S). Data points (*)
different at P ≤ 0.05 by LSD’s pairwise comparison.

74

SWMREC 2012
0.045
0.040

*

o

Anthocyanin / Brix

0.035
0.030
0.025
0.020
0.015

(P)
(S)

0.010
0.005
7/23

7/30

8/6

8/13

8/20

8/27

9/3

Date
FIGURE 26. Anthocyanin by Brix concentration over time for SWMREC field experiment in
2012. Primary bud is represented by (P) and a secondary bud is represented by (S). Data points
(*) different at P ≤ 0.05 by LSD’s pairwise comparison.

75

Discussion
The hypothesis of trellis system effects on fruit quality proved to hold true in both of the
experiments conducted in 2012. At the HTRC location, the training system had a significant
impact on the basic fruit chemistry when the fruit that was harvested was pooled together (Table
8). The vines trained to GDC had statistically higher total soluble solids and a higher
concentration of phenolics as compared to the other two trellis systems. The thought that a trellis
system that allows for more light to reach the fruiting zone helps improve fruit quality (Bordelon,
2008) was proven to hold true for Marquette. This same result had been documented in another
SCH cultivar in recent years (Bavougian, 2012). In that particular study, the GDC trained vines
also had the highest amount of total soluble solids, but unlike the results of that study, the 2012
HTRC frost experiment showed no difference in pH or titratable acidity (Table 9). The results
which reveal no difference in vine size caused by the different training system is consistent with
the literature for the region (Howell et al., 1991). Among the various fruit chemistry parameters
that were examined throughout the time period between veraison and harvest, statistical
differences were found, but most of these differences disappeared at the time of harvest. Those
that were found could have been related to the fact that all of the fruit was harvested on the same
date and being Marquette a very early ripening variety for Michigan climate, the fruit lacking
behind in fruit chemistry, could have reach an higher concentration of sugar an polyphenols just
delaying the harvest 1 or 2 weeks; the fruit on secondary shoots would reach the same chemical
parameters as of the fruit on primary shoots that escaped the frost events.
The experiment conducted at SWMREC showed very similar results as the vines trained
to HWC at the HTRC location. The yield component results at this location (Table 14.) which
show that the fruit from the primary bud is smaller than the secondary bud lends to the concept

76

that the meristem tissue was damaged during the frost events. The primary fruit had fewer
berries on the clusters, smaller single berry weight, and an overall increased cluster weight when
compared to the secondary fruit produced from the vines. This same trend does not appear in the
literature.
The greatest addition to the understanding of cold climate viticulture was on the aspects
of the fruit chemistry of fruit on secondary shoots. For a variety such as Marquette, which ripens
early in the season, the season length at certain locations in Michigan is long enough to ripen this
secondary fruit to levels that are acceptable for commercial winemaking. The information about
the number of clusters produced by the vines that ripen during the growing season when the
vines have experienced more than 80% damage to the primary buds is also significant.

77

Conclusion
The unexpected warmth at the beginning of March 2012 caused an early budburst and the
vines suffered from several frost events. However, this unique weather event allowed for the
evaluation of how the SCH cultivar ‘Marquette’ reacts to such limiting environmental
conditions. To our knowledge, this is the first study on the subject as the plant material was not
released for commercial propagation until 2006. Across the US and recent plantings, very few if
any locations were able to collect this level of detail on information regarding how the vines
react to frost events and the effects that trellis systems have on the resulting fruit. The data from
the experiment conducted at the HTRC showed that in a frost event like the one that occurred in
the spring of 2012, severe damage will occur, independent of the training system adopted for the
vines. Results show that the vines were able to produce fruit on the secondary buds, which
allowed for a very significant amount of yield compared to other grape cultivars used in the
region. This information will enable growers to harvest fruit also after several frost events with
the potential of killing primary shoots. The work done at SWMREC confirms that the results and
information obtained from the experimental site at HTRC, also at SWMREC that the frost events
were reduced and the vines were able to carry more fruit at harvest coming from primary shoots.

78

APPENDICES

79

APPENDIX A: FRUIT CHEMICAL COMPOSITION OVER TIME
TABLE 18. Fruit chemical composition over time for High Wire Cordon at HTRC field experiment in 2012.

High Wire Cordon Trellis System (HWC)
Date

Bud Type

TSS

pH

Primary

15.1 a*

2.96 a

TA
(g/L)
15.34 a

Secondary

8.9 b

2.78 b

24.65 b

1.56 b

0.06 b

0.007 b

Primary

20.5 a

3.24 a

10.99 a

0.68

0.49 a

0.024 a

Secondary

16.5 b

2.99 b

16.09 b

0.69

0.26 b

0.015 b

Primary

22.7 a

3.37 a

8.87 a

0.92

1.22

0.056

Secondary

21.2 b

3.17 b

12.05 b

0.89

1.12

0.053

Primary

20.3 a

3.50 a

8.66 a

0.97

0.91

0.042

Secondary

19.7 a

3.32 b

9.86 b

1.01

1.01

0.050

(oBrix)

Phenolics
(a.u./g)
0.85 a

Anthocyanin Anthocyanin
(mg/g)
/ brix
0.32 a
0.021 a

7/24/2012

8/3/2012

8/14/2012

8/18/2012
*Treatment means within each date not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. Rows within each date where no letters are present indicate a lack of significant differences among treatments.

80

TABLE 19. Fruit chemical composition over time for Geneva Double Curtain at HTRC field experiment in 2012.

Geneva Double Curtain Trellis System (GDC)
Date

Bud Type

TSS

pH

Primary

14.7 a*

2.96 a

TA
(g/L)
14.85 a

Secondary

7.9 b

2.80 b

24.24 b

1.58 b

0.05b

0.007 b

Primary

20.0 a

3.16 a

10.69 a

0.72

0.50 a

0.025 a

Secondary

16.0 b

2.96 b

15.71 b

0.70

0.22 b

0.013 b

Primary

22.4 a

3.32 a

9.70 a

1.01 a

1.30

0.058

Secondary

20.9 b

3.11 b

12.14 b

0.87 b

1.10

0.052

Primary

21.7 a

3.46 a

8.68 a

1.03

0.93

0.045

Secondary

21.5 a

3.27 b

10.02 b

1.07

0.98

0.044

(oBrix)

Phenolics
(a.u./g)
0.91 a

Anthocyanin Anthocyanin
(mg/g)
/ brix
0.31 a
0.021 a

7/24/2012

8/3/2012

8/14/2012

8/18/2012
*Treatment means within each date not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. Rows within each date where no letters are present indicate a lack of significant differences among treatments.

81

TABLE 20. Fruit chemical composition over time for Moving Trellis at HTRC field experiment in 2012.

Moving Trellis System (MT)
Date

Bud Type

TSS

pH

Primary

14.3 a*

3.05 a

TA
(g/L)
14.71 a

Secondary

8.4 b

2.81 b

23.50 b

1.45 b

0.06 b

0.007 b

Primary

19.0 a

3.22 a

10.91 a

0.60

0.40 a

0.021 a

Secondary

13.8 b

2.95 b

16.57 b

0.64

0.12 b

0.009 b

Primary

21.7 a

3.38 a

8.81 a

0.87

0.95

0.044 a

Secondary

20.1 b

3.22 a

11.37 b

0.84

1.05

0.052 b

Primary

20.9 a

3.47 a

8.14 a

0.94

1.07

0.050

Secondary

19.1 a

3.32 b

10.48 b

0.96

0.99

0.052

(oBrix)

Phenolics
(a.u./g)
0.78 a

Anthocyanin Anthocyanin
(mg/g)
/ brix
0.25 a
0.017 a

7/24/2012

8/3/2012

8/14/2012

8/18/2012
*Treatment means within each date not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. Rows within each date where no letters are present indicate a lack of significant differences among treatments.

82

TABLE 21. Fruit chemical composition over time for SWMREC frost evaluation field experiment in 2012.

Primary

16.3 a*

3.05 a

TA
(g/L)
15.13 a

Secondary

10.1 b

2.73 b

22.28 b

1.40 b

0.11 b

0.011

Primary

20.6 a

3.33 a

11.04 a

1.08

0.89

0.042

Secondary

17.7 b

3.12 b

13.36 b

1.04

0.69

0.035

8/30/2012

Primary

28.9 a

3.61 a

7.01 a

0.99

0.49 a

0.019 a

(Harvest)

Secondary

24.8 b

3.51 b

7.59 b

1.02

0.81 b

0.033 b

Date

Bud Type

TSS

pH

(oBrix)

Phenolics
(a.u./g)
0.90 a

Anthocyanin
(mg/g)
0.30 a

Anthocyanin
/ brix
0.018

7/25/12

8/6/12

*Treatment means within each date not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. Rows within each date where no letters are present indicate a lack of significant differences among treatments.

83

APPENDIX B: TYPE 3 TABLES OF ANOVA FOR HTRC
TABLE 22. Analysis of variance (ANOVA) of TSS at HTRC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
2.96
1.75
Trellis
1
38
443.27
Shoot_Type
2
38
1.21
Trellis*Shoot_Type
3
212
925.83
Timing
6
212
5.89
Trellis*Timing
3
213
72.90
Shoot_Type*Timing
6
213
1.36
Trelli*Shoot_*Timing

Pr > F
0.3145
<.0001
0.3083
<.0001
<.0001
<.0001
0.2327

TABLE 23. Analysis of variance (ANOVA) of pH at HTRC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
35.9
0.53
Trellis
1
37.3
765.12
Shoot_Type
2
37.3
0.25
Trellis*Shoot_Type
3
213
687.12
Timing
6
213
1.41
Trellis*Timing
3
213
91.08
Shoot_Type*Timing
6
213
0.61
Trelli*Shoot_*Timing

Pr > F
0.5903
<.0001
0.7763
<.0001
0.2101
<.0001
0.7207

TABLE 24. Analysis of variance (ANOVA) of TA at HTRC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
34.3
5.38
Trellis
1
34.1
121.62
Shoot_Type
2
34
0.76
Trellis*Shoot_Type
3
213
171.90
Timing
6
213
0.17
Trellis*Timing
3
213
110.49
Shoot_Type*Timing
6
213
0.27
Trelli*Shoot_*Timing

84

Pr > F
0.0093
<.0001
0.4769
<.0001
0.9847
<.0001
0.9524

TABLE 25. Analysis of variance (ANOVA) of Total Phenolics at HTRC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
36.8
1.86
Trellis
1
38.3
61.72
Shoot_Type
2
38.2
1.30
Trellis*Shoot_Type
3
212
839.58
Timing
6
212
5.10
Trellis*Timing
3
212
17.16
Shoot_Type*Timing
6
212
2.98
Trelli*Shoot_*Timing

Pr > F
0.1705
<.0001
0.2852
<.0001
<.0001
<.0001
0.0082

TABLE 26. Analysis of variance (ANOVA) of Anthocyanin content at HTRC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
2.92
0.80
Trellis
1
39.3
319.17
Shoot_Type
2
39.2
0.12
Trellis*Shoot_Type
3
213
378.29
Timing
6
213
1.60
Trellis*Timing
3
213
2.00
Shoot_Type*Timing
6
213
0.83
Trelli*Shoot_*Timing

Pr > F
0.5269
<.0001
0.8877
<.0001
0.1494
0.1149
0.5500

TABLE 27. Analysis of variance (ANOVA) of Anthocyanin content by TSS at HTRC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
47.7
0.02
Trellis
6
183
3.92
Shoot_Type
12
183
1.11
Trellis*Shoot_Type
2
112
930.22
Timing
4
112
1.61
Trellis*Timing
2
112
25.25
Shoot_Type*Timing
4
112
2.81
Trelli*Shoot_*Timing

85

Pr > F
0.9845
0.0010
0.3571
<.0001
0.1772
<.0001
0.0290

APPENDIX C: TYPE 3 TABLES OF ANOVA FOR SWMREC
TABLE 28. Analysis of variance (ANOVA) of Cluster Weights at SWMREC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value Pr > F
1
13
26.77 0.0002
ShootType
TABLE 29. Analysis of variance (ANOVA) of Rachis Weights at SWMREC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value Pr > F
1
13
5.02
0.0431
ShootType
TABLE 30. Analysis of variance (ANOVA) of Number of berries per cluster at SWMREC in
2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value Pr > F
1
13
6.30
0.0261
ShootType
TABLE 31. Analysis of variance (ANOVA) of Berry diameters per cluster at SWMREC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value Pr > F
1
13
2.20
0.1621
ShootType
TABLE 32. Analysis of variance (ANOVA) of Yield at SWMREC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value Pr > F
1
3
0.03
0.8747
ShootType
TABLE 33. Analysis of variance (ANOVA) of Number of cluster at SWMREC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value Pr > F
1
3
0.19
0.6915
ShootType

86

TABLE 34. Analysis of variance (ANOVA) of TSS (oBrix) at SWMREC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
1
15
102.23
ShootType
2
108
348.34
DateCode
2
108
6.12
ShootType*DateCode

Pr > F
<.0001
<.0001
0.0030

TABLE 35. Analysis of variance (ANOVA) of pH at SWMREC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
1
15
131.70
ShootType
2
104
662.45
DateCode
2
104
16.51
ShootType*DateCode

Pr > F
<.0001
<.0001
<.0001

TABLE 36. Analysis of variance (ANOVA) of TA at SWMREC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
1
15
110.30
ShootType
2
102
534.92
DateCode
2
102
40.09
ShootType*DateCode

Pr > F
<.0001
<.0001
<.0001

TABLE 37. Analysis of variance (ANOVA) of Total Phenolics at SWMREC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
1
15
11.99
ShootType
2
108
4.86
DateCode
2
108
17.92
ShootType*DateCode

Pr > F
0.0035
0.0096
<.0001

TABLE 38. Analysis of variance (ANOVA) of Anthocyanin content at SWMREC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
1
15
0.12
ShootType
2
108
42.24
DateCode
2
108
10.91
ShootType*DateCode

87

Pr > F
0.7342
<.0001
<.0001

TABLE 39. Analysis of variance (ANOVA) of Anthocyanin/Brix at SWMREC in 2012.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
1
45.5
0.01
ShootType
2
50.8
12.34
DateCode
2
50.6
7.36
ShootType*DateCode

88

Pr > F
0.9269
<.0001
0.0016

CHAPTER III
SHOOT DENSITY, TRELLIS SYSTEM, AND CROPLOAD EFFECTS ON
MARQUETTE GRAPEVINES GROWN IN MICHIGAN IN 2013

Introduction
The spring of 2012, was abnormally early with extremely high temperatures during
March. In contrast, the spring of 2013 was similar to the ten year average for both experimental
locations, HTRC and SWMREC (Figure 27; Figure 28). The weather during this time prevented
any damaging frost from occurring at either field location. Due to the mild spring in 2013 that
avoided significant frost damage, the experiments conducted in 2012 were not repeated for a
second year of study. Instead experiments were adjusted to accommodate the different research
objectives for the 2013 growing season. Starting in the winter of 2012, field plots at the
Horticulture Teaching and Research Center (HTRC) and Southwest Michigan Research and
Education Center (SWMREC) were selected for the second year of study. The focus of the
research was on canopy density and how exposure of the fruiting zone to light affects the fruit
chemistry parameters and canopy development in Super Cold Hardy ‘Marquette’ vines.
Experiments were conducted to evaluate the effects of yield on fruit ripening and the
threshold production for Marquette trained to High Wire Cordon (HWC) system at SWMREC.
For the experimental plot at HTRC containing different trellis systems, the vines were subjected
to two levels of shoot density with the same level of crop to observe effects of canopy density on
fruit quality parameters and ripening processes. These two field experiments will continue to lay
a foundation for future precision viticulture research for super cold hardy (SCH) hybrids that will

89

help improve not only the grape’s fruit quality but also how these SCH varieties are managed to
optimize their production under the Michigan climate.

90

Growing Degree Days
HTRC, East Lansing, MI
1800

10-Year Ave.
(2002 - 2011)
2012
2013

1600

1400

GDD (Base 10oC)

1200

1000

800

600

400

200

0
3/1

5/1

7/1

9/1

11/1

Date

FIGURE 27. Growing Degree Day for HTRC. Data retrieved from MAWN site located at
station.

91

Growing Degree Day
SWMREC, Benton Harbor, MI
2000

10 Year Ave.
(2002-2011)
2012
2013

1800

1600

GDD (Base 10oC)

1400

1200

1000

800

600

400

200

0
3/1

5/1

7/1

9/1

11/1

Date

FIGURE 28. Growing Degree Day for SWMREC. Data retrieved from MAWN site located at
station.

92

Materials and Methods
Shoot Density Field Experiment – MSU Horticulture Teaching and Research Center
Plant Material
Vitis spp. ‘Marquette’ own-rooted vines were planted in 2008 in the experimental and
teaching vineyard located at the Horticulture Teaching and Research Center (HTRC). Similar to
the experiment conducted in 2012, the same vines located at HTRC were used. Specific details
can be found in chapter two regarding aspects of planting density and vine establishment (page
30). All sprayings and vineyard management practices were conducted in a similar fashion to the
rest of the experimental vineyard, with the exception that there was no hedging of actively
growing canopies on experimental vines. Climatic data, such as daily temperatures, growing
degree days (GDD) and daily precipitations were obtained from the Michigan Automated
Weather Network (MAWN) weather station located at HTRC, at 100 m from the experimental
vineyard

(Table

40).

Supplemental

data

can

be

accessed

from

http://www.agweather.geo.msu.edu/mawn/. Major phenological stages were also recorded and
are presented with the date of observation and the Julian date (Table 41).
Experimental design
This experiment was set up using a Split-plot design with two replications. Within each
replication were the three trellis systems (HWC, GDC, MT) that had been converted in the
winter of 2011. The main experimental factor was trellis system which contained the three trellis
systems within each replication. The second main experimental factor of shoot density (3 shoots
per 30 cm of linear length cordon (3S) or 6 shoots per linear length of cordon (6S)) was chosen
to study. In 2013, both factors were the found on each replication and were used to study the
effects of shoot density on vine canopy growth and fruit quality. The vines were assigned to the
two separate levels of shoot density after fruit set had concluded. In order to achieve this, the
93

permanent cordon was measured and the total shoots on the vine were counted. Each vine was
evaluated on the length of established cordon and the number of shoots actively growing with
clusters, in order to determine the shoot density per 30 cm linear length of cordon of each
experimental vine. Once these two parameters were measured, all of the vines were adjusted to
fit into one of the two treatments: (1) 3 shoots per 30 cm of linear cordon or (2) 6 shoots per 30
cm of linear cordon. Two replications of each shoot density were distributed between the vines
trained to the three different trellis systems in place. To achieve two levels of shoot densities,
shoots were removed retaining shoots that had two clusters per shoot. All shoots not containing
clusters (sterile shoots) were removed at the time of adjustment. These treatments were applied
to the experimental vines after fruit set ended (6/10/2013) throughout the vineyard.
When applying the treatments, a set of five randomly selected shoots were chosen to
represent the whole vine during the growing season. These shoots are referred to throughout this
research as model shoots. These sample shoots were tagged to allow for easier identification and
to allow for collection of data throughout the rest of the growing season. These tagged shoots
were used for repeated measurements of several parameters: shoot length, canopy density, and
berry growth. They were also used for chemical analysis and canopy growth assessment at the
time of harvest.
Canopy Growth Measurements
Weekly monitoring of the shoot length in model shoots was used to measure canopy
growth from anthesis and until the canopy was fully developed (veraison). At the time of
veraison, no significant change in canopy development had occurred so no additional
measurements were made until the tagged shoots were collected at the time of harvest. Shoot

94

length was measured with a flexible measuring tape was used to measure from the base of the
model shoot to the active growing shoot tip.
Temperature and Radiation Measurements
In order to measure the different levels of light and temperatures within the different
canopies caused by the combinations of different trellis systems and shoot densities, a set of
quantum sensors measuring the photosynthetically active radiation (PAR) and thermocouples
were placed on selected vines within each treatment and trellis system. One vine from each
combination of trellis and shoot density was used as a representative sample for the rest of the
vines in the plot. A total of six vines were used for this data collection. To achieve the PAR
measurements, three sensors (model SQ-110; Apogee Instruments, Logan, UT) were placed at
the level of the fruit zone within the canopy and three additional sensors were placed below the
fruiting zone for each trellis and treatment. Two PAR sensors were placed above the canopies to
collect direct radiation, which was compared with the automated weather station located at
HTRC.
The same vines used for collection of the light levels using the quantum sensors were
also outfitted with thermocouples. These temperature measurements were acquired by placing
three Fine-wire (American Wire Gauge [AWG] thermocouples (Type T) [Copper-Constantan])
in three of the apical clusters attached to model shoots. All of the sensors were attached to a
multiplexer (AM18/32A, Campbell Scientific, Logan, UT) that measured readings from the
sensors every 60 seconds and then recorded these measurements to the data logger (CR-10X;
Campbell Scientific, Logan, UT) on a bases of every twenty minutes. This setup that allowed for
continuous data collection was placed in the field starting from the beginning of veraison until
the day of harvest.
95

Sampling Procedures and Harvest Data Collection
At the time of veraison, berry samples were collected to be analyzed in the laboratory at a
later date. Sampling was conducted by collecting 30 selected berries from non-tagged shoots.
Non-tagged shoots were used in order not to affect the configuration of the clusters on the tagged
shoots. This was done to not affect berry diameter, rachis weight, or berry weight. Berries were
selected from different parts of the clusters to insure a representative sample per vine. These
berries were then transferred into plastic bags and placed into a -20oC freezer until analysis was
performed. The whole process was continued weekly until harvest. A total of five samplings had
elapsed from veraison to harvest.
At the time of harvest, each vine used in the experiment was hand harvested. Each vine
had the fruit collected from the tagged model shoots to be used for chemical analysis in the
laboratory. After the fruit was removed, the tagged shoots per vine were also harvested and
placed together per vine with care not to damage the shoot or damage any of the leaves on the
tagged shoots. These shoots were then placed into coolers and taken to the laboratory for
immediate analysis of viticultural parameters like leaf area, number of leaves per shoot, and
shoot length. Leaf area was measured by collecting the leaves from the harvested shoots and
placing the leaves one at a time thru a leaf area meter (LI-3050AHS, Lambda Instruments
Corporation, Nebraska). This allowed for leaf area per shoot to be obtained.
As soon as the tagged shoots were collected, the rest of the vine was harvested with care
in order to count and mass the yield per vine. This process was repeated through the plot until
harvest was completed. At the time of completion of harvest a representative sample of fruit was
collected from each trellis system and shoot density being tested to be used for experimental
wine making process. This sample consisted of 13.6 kilograms of fruit per combination of trellis
96

system and shoot density. A total of six samples were collected. This subsample was sorted by
hand to only allow clusters with less than five percent of rot to be used for the microfermentation. A difference in the percentage of rot observed in the vineyard between trellis
systems was found, to not allow rot to be a factor in the wines produced from this experiment,
the fruit was sorted by visually inspecting each cluster being selected and not allowing more than
five percent of the berries on the cluster to be damaged or rotten. Details about the experimental
wine making process and procedures can be found in Chapter 4. During the following winter of
2013, the vines were pruned to a bud count of 100 buds and all of the one-year old wood was
collected and weighed to obtain pruning weights per vine and this was used to calculate the
Ravaz Index or RI (Ravaz, 1911) which is an index of vine balance and calculated as yield per
vine (kg) divided by pruning weight (kg).
Basic Fruit Chemistry Measurements
Laboratory analysis of fruit samples conducted at the HTRC were performed as already
described in Chapter 2 under section of Basic Fruit Chemistry Measurements under the heading
of Field Experiment – HTRC for 2012 (page 40).

97

TABLE 40. Monthly weather data measured for 2013 from MAWN located at HTRC.
GDD (base
10oC)
Precipitation
(mm)

March
2.4

April
41.1

May
215.1

June
286.9

July
352.9

August
301.0

September
198.6

Total
1397.6

11.7

178.8

76.2

160.8

67.6

97.3

20.8

618.7

TABLE 41. Major Phenological stage, Date of occurrence and Growing Degree Days for
Marquette grown at HTRC in 2013.
Phenological
Stage
Date
Julian Date
GDD (base
10oC)
Days from
Bud break

Bud
break

Anthesis

Pea-size

Veraison

Harvest

5/4/2013 6/11/2013
124
162
79.7
324.6

7/3/2013
184
572.7

8/2/2013
214
915.1

9/11/2013
254
1307.7

61

91

131

-

38

98

Crop Load Field Experiment – Southwest Michigan Education and Research Center
Plant Material
In 2013 at SWMREC, the same research plot was used as in the field experiment in 2012.
All pest and disease control and vineyard management practices were conducted similar to the
rest of the experimental vineyard. Environmental measurements were gathered from the
Michigan Automated Weather Network (MAWN) weather station located at SWMREC. This
instrument was used to collect factors such as daily temperatures, growing degree days (GDD),
and daily precipitations. The monthly totals throughout the growing season can be found in
Table 42. The environmental data can be accessed online from the MAWN website at
http://www.agweather.geo.msu.edu/mawn/. Major vine phenological stages were also recorded
and are presented with the date of observation and the Julian date (Table 43).
Experimental design
The field experiment conducted at SWMREC used the same vines as used in the 2012
season with the addition of one more block added to the design of the experiment. The design for
this particular experiment was setup up as a random complete block design (RCBD) with the use
of one type of trellis system (High Wire Cordon) for all vines planted in the vineyard. Within the
vineyard, each block contains four vines. For this experiment only five blocks out of the six were
used due to losses of vines in one of the blocks on the south side of the plot; within these blocks
the weakest vine and most vigorous vines were not used in the experiment; the result was 18
vines were used. The following treatments were assigned: high, medium, and low crop on the
vines within the plot. After fruit set occurred throughout all of the experimental blocks located
within the plot, the number of clusters per each vine were counted. Depending on the amount of
fruit that was present at this time, vines were placed into three categories (high, medium, and
low). These vines were then placed into each category which contained six vines in each. In
99

order to achieve these three levels of crop load, adjustments were made to reduce the number of
clusters that were on the vine. Each of the Marquette vines used had a total of five tagged shoots
that were followed throughout the growing season. The five tagged shoots were chosen to
represent the population of shoots within each vine and will be referred to as model shoots.
Several parameters of vine vigor and fruit chemistry were measured weekly from bud bread to
harvest using these model shoots to represent the vine as a whole.
Canopy Growth Measurements
The model tagged shoots were used to track canopy growth starting after fruit set and
until the canopy was fully developed. At the time of veraison, the canopies within the
experimental plot had stopped growing and no additional measurements were made until the
tagged shoots were collected at the time of harvest to be used to provide data on a per vine basis.
In order to conduct shoot measurements, a flexible measuring tape was used to measure from the
base of the shoot to the end of the actively growing shoot tip. These measurements were
completed on a weekly basis until the mature canopy had developed.
Temperature and Radiation Measurements
Similar to the experimental setup at HTRC for 2013, different levels of light and
temperatures within the different crop loads were measured. To collect this information, a set of
quantum sensors measuring the photosynthetically active radiation (PAR) and thermocouples
were placed on selected vines within each of the three treatments. Two vines per each crop load
were used as representative samples for the rest of the vines in the plot. A total of six vines were
used for this data collection. To achieve the PAR measurements, three sensors (model SQ-110;
Apogee Instruments, Logan, UT) were placed at the level of the fruit zone within the canopy and

100

three additional sensors were placed below the fruiting zone for each trellis and treatment. Two
PAR sensors were placed above the canopies to collect direct radiation, which was compared to
the sensor located as part of the MAWN system on the automated weather station located
SWMREC.
The same vines used for collection of the light levels using the quantum sensors were
also outfitted with thermocouples. These temperature measurements were conducted by placing
three Fine-wire (American Wire Gauge [AWG] thermocouples (Type T)[Copper-Constantan]) in
three of the apical clusters attached to model shoots. All of the sensors were attached to a
multiplexer (AM18/32A, Campbell Scientific, Logan, UT) that measured readings from the
sensors every 60 seconds and then recorded these measurements to the data logger (CR-10;
Campbell Scientific, Logan, UT) on a bases of every twenty minutes. This setup that allowed for
continuous data collection was placed in the field starting from the beginning of veraison until
the day of harvest.
Sampling Procedures and Harvest Data Collection
At the time of veraison, berry samples were collected to be analyzed in the viticulture
laboratory of the Department of Horticulture. The procedure for sampling was by collecting 30
selected berries from non-tagged shoots. Non-tagged shoots were used in order not to affect the
configuration of the clusters on the tagged shoots. Special care was applied to select berries from
different parts of the clusters and consequently insure a representative sample per vine. These
berries were then transferred into plastic bags and placed into a -20o C freezer until analysis was
performed. The whole process was continued weekly until harvest and repeated four times before
harvest.

101

At harvest, each vine used in the experiment was hand harvested and fruit collected from
the tagged model shoots to be used for chemical analysis in the laboratory. After the fruit was
removed, the tagged shoots per vine were also harvested and placed together per vine with care
not to damage the shoot or damage any of the leaves on the tagged shoots. These shoots were
then placed into coolers and taken to the laboratory for immediate analysis of viticultural
parameters like leaf area, number of leaves per shoot, and shoot length. Leaf area was measured
by collecting the leaves from the harvested shoots and placing the leaves one at a time through a
leaf area meter (LI-3050AHS, Lambda Instruments Corporation, Nebraska). This allowed for
leaf area per shoot to be obtained.
The clusters harvested from the model shoots were separated in order to count number of
berries per cluster and measure rachis weight. As soon as the tagged shoots were collected, the
rest of the vine was harvested to calculate the total yield per vine. This process was repeated
through the plot until harvest was completed. At the time of completion of harvest a
representative sub-sample of fruit was collected and used for experimental wine making process.
This sample consisted of 27.2 kilograms of fruit per crop load treatment (high, medium, and
low). Details about the experimental wine making process will be discussed with the sensory
evaluation in Chapter 4.
During winter of 2013, the vines were pruned to a bud count of 100 buds and all of the
one-year old wood was collected and weighed to obtain pruning weights per vine and was used
to calculate the Ravaz Index (Ravaz, 1911) a viticultural index of vine balance calculated as
yield per vine (kg) divided by pruning weight (kg).

102

Basic Fruit Chemistry Measurements
Seasonal and harvest fruit samples were analyzed in the laboratory as already described
in Chapter 2 under section of Basic Fruit Chemistry Measurements under the heading of Field
Experiment – HTRC for 2012 (page 40).
Statistical Analysis
As in 2012, once all samples were processed in the laboratory, all the viticultural and
chemical data measured were first tested for normality. Once all parameters measured had been
found to be normal distribution, one way Analysis Of Variance (ANOVA) was performed using
PROC MIXED in SAS version 9.3 for Windows (Cary, NC). All pair-wise comparisons were
performed and tested using Tukey’s Test with the alpha equal to 0.05.

103

TABLE 42. Monthly weather data measured for 2013 from MAWN located at SWMREC.
GDD (base
10oC)
Precipitation
(mm)

March
2.6

April
52.6

May
221.8

June
290.1

July
362.0

August
336.6

Sept.
246.8

Total
1512.5

19.8

125.2

55.1

69.1

42.2

54.1

25.7

391.1

TABLE 43. Major Phenological stage, Date of occurrence and Growing Degree Days for
Marquette grown at SWMREC in 2013.
Phenological Stage
Date
Julian Date
GDD (base 10oC)
Days from
Bud break

Bud
break

Anthesis

Veraison

Harvest

5/1/2013
121
67.1
-

6/8/2013
159
322.9
38

7/31/2013
212
928.9
91

9/13/2013
256
1407.3
135

104

Results – 2013 HTRC
Weather during the Growing Season
Unlike the spring of 2012, the spring of 2013 was very similar to the ten-year average
when compared to temperature and GDD accumulation (Figure 33). This is evident by the date
of bud break, which occurred on May 4th (Table 50). The heat units at bud break had
accumulated to 124 growing degree days or GGD (base 10oC). Due to this weather during the
beginning of the spring, no frost events had occurred and consequently all of the vines in the
experiment showed no damage to the primary buds. With no frost damage and balance pruning
of the vines, number of secondary shoots and secondary clusters that developed was negligible
and was not examined in 2013. Overall weather patterns in 2013 were similar to the ten-year
average with respect to both precipitation and heat that accumulated over the duration of the
growing season. The total GDD that accumulated by the end of the season was 1397.6 as of the
end of September with the precipitation total being 618.7 mm (Table 49). The GDD accumulated
at the time of harvest was 1307.7 (Table 50).
Fruit Chemistry during the Ripening Process
Starting at veraison, weekly samples were measured until harvest from all treatments as
described in the material and methods section. There were no statistical differences found with
the weekly sampling for testing of total soluble solids from any of the treatments. Starting on
8/7/13, the initial readings were around 13 degree brix. The highest reading of 24.6 degree brix
was obtained at the time of harvest for the vines trained to GDC with three shoots per linear 30
centimeter of cordon (Table 54). The curves for all three trellis systems can be observed in
Figure 34 and for the individual trellis systems with the different canopy density in Figure 35. As

105

seen in both figures, the rate of sugar accumulation did not vary significantly between the
different trellises or the different canopy densities. Same trends between canopy density and
training systems treatments were evident when looking at the titratable acidity of the samples
collected throughout the ripening process (Figure 38; Figure 39).
Alternatively, a similar trend was not true for the parameter of pH (Figure 36 and Figure
37). A difference was observed on 9/4/13 with the vines trained to HWC between the two
canopy densities. The samples from the three shoots per 30 centimeter of linear cordon had an
average pH of 3.28 while the six shoots per linear 30 centimeters had a pH of 3.38. This was the
only difference other than the difference seen at the time of harvest.
The greatest amount of differences between the treatments appears in the total phenolic
content of the samples (Figure 40). The concentration of the phenolic content varied depending
on the trellis system, as well as between the different canopy densities (Figure 41). In the case of
HWC, a change in concentration was found on the first sampling date of 8/7/13 and again on
9/4/13. Overall phenolic content for the HWC vines were higher than the other two trellis
systems. No statistical difference was found in regards in phenolics between the two canopy
densities with the vines trained to the MT trellis system, but differences were found in the GDC
trained vines on 8/14/13 and on 9/4/13 (Figure 41).
As for anthocyanin concentration between the treatments, only one difference was found
on the sampling date of 9/4/13 on vines trained to HWC (Figure 42; Figure 43). On this date the
three shoots per 30 centimeters of linear cordon had a higher concentration of 1.95 moles per
gram of homogenate as compared to 1.32 moles per gram, which was observed in the six shoots
per 30 centimeters of linear cordon (Table 60). The data reveals that the vines trained to the MT
trellis had a more consistent anthocyanin accumulation than the other two trellis systems (Figure
106

43). Previous research has shown a positive correlation between anthocyanin concentration and
total soluble solid content. For that reason the anthocyanin concentration was plotted by brix
concentration for all of the sampling dates (Figure 44). Upon separating the data by trellis
system, a statistical difference was only observed from the vines trained to HWC trellis system.
A difference between the two canopy densities occurred on 8/7/13 and on 9/4/13 (Table 60). The
samples measured on the 9/4/13 sampling date for HWC vines showed the greatest difference
between the ratio of anthocyanin to brix. All of the data separated by date and by trellis systems
can be found in the tables in Appendix B.
Fruit Chemistry, Yield Components, and Canopy Measurements at Harvest
Fruit chemistry was compared across each trellis and shoot density for all of the same
parameters that were measured from the time of veraison until harvest. When comparing
different densities within the same trellis system, the total soluble solids were generally higher in
the lower shoot density. When comparing across trellis systems, the HWC trained vines had the
greatest accumulation. The HWC with three shoots and GDC with three shoots both had higher
pH and titratable acidity as compared to the six shoot vines on the same trellis systems. The
HWC with six shoots had the highest concentration of phenolics while the GDC with six shoots
had the lowest (Table 54). No statistical differences were found when comparing anthocyanin
concentration among treatments.
At the time of harvest, the five model shoots that were tagged earlier in the season were
collected to allow for analysis in the laboratory. When investigating the shoot lengths, no
difference was found between the canopy densities on the HWC trained vines and the GDC
canopy densities (Table 51). However, a difference was found when looking at the number of
lateral shoots. The GDC trellis vines had a lower number of lateral shoots in both canopy
107

densities. Across all of the other parameters (total length of laterals, total length of shoot, number
of leaves per shoot and leaf area), the most significant differences came from the vines trained to
GDC with six shoots per 30 centimeters of linear cordon. When comparing the ratio of leaf area
by total shoot length, the different canopy densities per trellis system did not have as significant
of an effect as that observed by the type of trellis system (Table 51).
For yield components, each vine was hand harvested as described in the material and
methods section to allow for fruit sorting, cluster counts, and total yield per experimental vine.
Yields per treatment did differ as the highest yielding vines were the GDC with six shoots per 30
centimeter of linear cordon and the lowest yielding vines were on HWC with the same shoot
density (Table 52). This was also reflected in the number of clusters per vine. In terms of the
percentage of rotten clusters per treatment, the vines trained to HWC with three shoots per 30
centimeters of linear cordon had an average of 25 percent. The lowest percent of rot was
recorded on the HWC with six shoots and the MT with three shoots. At the time of pruning in the
following winter, the MT with three shoots had an average of 1.17 kilograms per vine, while the
lowest was GDC with three shoots at 0.71 kilograms per vine. The Ravaz index appeared to be
similar by trellis system since the GDC vines at both shoot densities had the highest two ratios as
compared to all of the other treatments and trellis systems.
The last set of parameters measured at the time of harvest was on the cluster morphology.
The vines trained to GDC with three shoots per 30 centimeter of linear cordon had the highest
number of berries per cluster and heaviest rachis weight, which results in the heaviest cluster
weight as compared to the other treatments (Table 53). The GDC with six shoots per 30
centimeters of linear cordon had a statistically different number of berries and rachis weight as
compared to the GDC with six shoots. The MT with three shoots per 30 centimeter of linear
108

cordon had the heaviest single berry weight. The same trellis system with six shoots had the
smallest clusters, which were close to 40 grams smaller per cluster than the heaviest cluster
(Table 53).

109

TABLE 44. Canopy measurements at harvest in field experiment conducted in 2013 at HTRC. Shoot density was either adjusted to
three shoots per linear 30 centimeter of cordon (3S) or six shoots per linear 30 centimeter of cordon (6S).
Trellis
System

HWC

GDC

MT

Shoot
Shoot
Density length (cm)

Number
of
laterals
per
shoot

Total
Lateral
length
(cm)

Total
Shoot
length
(cm)

Number
of leaves
per
shoot

Leaf area
of model
shoots
(cm^2)

Leaf area/
Total
shoot
length
(cm^2/cm)

3S

117.61 a*

6.46 a

277.79 a

395.39 a

104 a

4548.68 a

12.0 a

6S

115.50 a

5.46 a

242.04 ab

357.54 a

83 ab

4243.28 ab

11.2 ab

3S

113.68 ab

4.75 ab

165.46 bc

279.14 ab

70 bc

3373.29 abc

13.0 a

6S

94.40 ab

2.97 b

98.37 c

192.77 b

52 c

2392.15 c

12.0 a

3S

97.97 ab

5.00 a

225.30 ab

323.27 a

71 bc

3188.98 abc

9.7 b

6S

86.79 b

4.86 a

201.86 ab

288.66 ab

64 bc

2861.01 bc

9.6 b

*Treatment means within columns not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. Columns where no letters are present indicate a lack of significant differences among treatments.

110

TABLE 45. Yield components at harvest in field experiment conducted in 2013 at HTRC. Shoot density was either adjusted to three
shoots per linear 30 centimeter of cordon (3S) or six shoots per linear 30 centimeter of cordon (6S).
Trellis
System

Shoot
Density

Tons /
acre

Yield/vine

HWC

3S

GDC

MT

(Kg/Vine)

Number
of cluster
per vine

Percent of
Rotten
Clusters

Pruning
weights
(kg)

Ravaz
Index

2.26 cd*

3.77 cd

63 cd

25.3 a

0.91 ab

5.1 cb

6S

1.30 d

2.17 d

54 d

> 0.1 b

0.89 ab

2.6 c

3S

3.79 b

6.32 b

110 b

4.9 b

0.68 b

12.0 b

6S

6.88 a

11.48 a

171 a

2.5 b

0.71 b

21.5 a

3S

2.57 c

4.28 c

83 c

> 0.1 b

1.17 a

3.9 c

6S

1.79 cd

2.98 cd

74 cd

4.6 b

0.94 ab

3.3 c

*Treatment means within rows not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
Columns where no letters are present indicate a lack of significant differences among treatments.

111

TABLE 46. Cluster morphology at harvest in field experiment conducted in 2013 at HTRC. Shoot density was either adjusted to three
shoots per linear 30 centimeter of cordon (3S) or six shoots per linear 30 centimeter of cordon (6S).
Trellis
System

Shoot
Density

Cluster
weight (g)

Number
of berries
per
cluster

Total
berry
weight (g)

Single
berry
weight (g)

Rachis
weight (g)

HWC

3S

84.91 bc*

83 bc

81.79 b

0.98 b

2.77 ab

6S

82.16 c

81 bc

79.22 b

0.96 b

2.56 ab

3S

112.74 a

101 a

109.55 a

1.11 ab

3.06 a

6S

101.34 ab

94 ab

98.76 a

1.07 ab

2.35 b

3S

81.14 c

75 c

78.44 b

1.29 a

2.52 b

6S

80.57 c

69 c

78.25 b

1.12 ab

2.25 b

GDC

MT

*Treatment means within rows not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
Columns where no letters are present indicate a lack of significant differences among treatments.

112

TABLE 47. Fruit chemical composition at harvest from experiment conducted at HTRC in 2013. Shoot density was either adjusted to
three shoots per linear 30 centimeter of cordon (3S) or six shoots per linear 30 centimeter of cordon (6S).
TSS

pH

TA
(g/L)

Phenolics
(a.u./g)

Anthocyanin
(mol/g)

Anthocyanin
/ Brix

24.5 a*

3.54 a

8.35 abc

1.25 a

1.90

0.078 ab

6S

24.4 ab

3.52 ab

8.81 ab

1.26 a

1.81

0.075 ab

3S

25.0 a

3.43.bc

7.72 cd

1.11 bc

1.66

0.067 b

6S

23.5 abc

3.33 c

7.53 d

1.05 c

1.70

0.071 ab

3S

23.2 bc

3.54 a

8.08 bcd

1.17 ab

1.82

0.079 a

6S

22.8 c

3.58 a

8.85 a

1.21 ab

1.83

0.081 a

Trellis
System

Shoot
Density

(oBrix)

HWC

3S

GDC

MT

*Treatment means within rows not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
Columns where no letters are present indicate a lack of significant differences among treatments.

113

26
24

20

A

o

TSS ( Brix)

22

18
16

HWC (3S)
HWC (6S)

14
12
8/5

8/12

8/19

8/26

9/2

9/9

9/16

26
24

TSS (oBrix)

22
20

B

18
16

GDC (3S)
GDC (6S)

14
12
8/5

8/12

8/19

8/26

9/2

9/9

9/16

26
24

TSS (oBrix)

22
20

C

18
16

MT (3S)
MT (6S)

14
12
8/5

8/12

8/19

8/26

9/2

9/9

9/16

FIGURE 29. Total soluble solids over time for HTRC field experiment in 2013. Graph A is High
Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
(*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
114

3.8
3.6

*

3.4
3.2

pH

A

3.0
2.8

HWC (3S)
HWC (6S)

2.6
2.4
8/5

8/12

8/19

8/26

9/2

9/9

9/16

3.8
3.6

*
3.4
3.2

pH

B

3.0
2.8

GDC (3S)
GDC (6S)

2.6
2.4
8/5

8/12

8/19

8/26

9/2

9/9

9/16

3.8
3.6
3.4
3.2

pH

C

3.0
2.8

MT (3S)
MT (6S)

2.6
2.4
8/5

8/12

8/19

8/26

9/2

9/9

9/16

FIGURE 30. pH concentration over time for HTRC field experiment in 2013. Graph A is High
Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
(*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
115

22

HWC (3S)
HWC (6S)

20

TA (g/L)

18
16

A

14
12
10
8
6
8/5

8/12

8/19

8/26

9/2

9/9

9/16

22

GDC (3S)
GDC (6S)

20

TA (g/L)

18
16

B

14
12
10
8
6
8/5

8/12

8/19

8/26

9/2

9/9

9/16

22

MT (3S)
MT (6S)

20

TA (g/L)

18
16

C

14
12
10
8
6
8/5

8/12

8/19

8/26

9/2

9/9

9/16

FIGURE 31. Titratable acidity over time for HTRC field experiment in 2013. Graph A is High
Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
(*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
116

1.4

HWC (3S)
HWC (6S)

Phenolics (a.u./g)

1.2

*

1.0

A

0.8

*

0.6

0.4

0.2
8/5

8/12

8/19

8/26

9/2

9/9

9/16

1.4

Phenolics (a.u./g)

GDC (3S)
GDC (6S)
1.2

B

1.0

*
*
0.8

0.6
8/5

8/12

8/19

8/26

9/2

9/9

9/16

1.4

Phenolics (a.u./g)

MT (3S)
MT (6S)
1.2

C

1.0

0.8

0.6
8/5

8/12

8/19

8/26

9/2

9/9

9/16

FIGURE 32. Phenolic concentration over time for HTRC field experiment in 2013. Graph A is
High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
(*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
117

2.1

*

Anthocyanin (mg/g)

1.8
1.5

A

1.2
0.9
0.6

HWC (3S)
HWC (6S)

0.3
8/5

8/12

8/19

8/26

9/2

9/9

9/16

2.1

Anthocyanin (mg/g)

1.8
1.5

B

1.2
0.9
0.6

GDC (3S)
GDC (6S)

0.3
8/5

8/12

8/19

8/26

9/2

9/9

9/16

2.1

Anthocyanin (mg/g)

1.8
1.5

C

1.2
0.9
0.6

MT (3S)
MT (6S)

0.3
0.0
8/5

8/12

8/19

8/26

9/2

9/9

9/16

FIGURE 33. Anthocyanin concentration over time for HTRC field experiment in 2013. Graph A
is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C is the
experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of cordon is
represented by (3S) and the six shoots per 30 linear centimeters of cordon is represented by (6S).
(*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
118

0.10

*

Anthocyanin / Brix

0.08

0.06

A

*

0.04

0.02

HWC (3S)
HWC (6S)

0.00
8/5

8/12

8/19

8/26

9/2

9/9

9/16

0.10

Anthocyanin / Brix

0.08

0.06

B

0.04

0.02

GDC (3S)
GDC (6S)

0.00
8/5

8/12

8/19

8/26

9/2

9/9

9/16

0.10

Anthocyanin /Brix

0.08

0.06

C

0.04

0.02

MT (3S)
MT (6S)

0.00
8/5

8/12

8/19

8/26

9/2

9/9

9/16

FIGURE 34. Anthocyanin by Brix concentration over time for HTRC field experiment in 2013.
Graph A is High Wire Cordon (HWC), graph B is Geneva Double Curtain (GDC), and graph C
is the experimental Moving Trellis. In all graphs the three shoots per 30 linear centimeters of
cordon is represented by (3S) and the six shoots per 30 linear centimeters of cordon is
represented by (6S). (*) Means significantly different at P ≤ 0.05 by LSD’s pairwise comparison.
119

Results – 2013 SWMREC
Weather during the Growing Season
The growing season for 2013 started in the spring to be very similar to the ten-year
average for this location (Figure 46). The relative cool temperatures during these few months
protected the vines from the buds breaking eco-dormancy before the last known date of frost
events for SWMREC. The cool spring temperatures and mild winter caused the amount of
damaged buds on the vines to be minimal. This allowed for the vines at this location to be used
for studying the effects of crop load adjustment on various canopy parameters and fruit
chemistry. Bud break was observed on May 1st with the 67.1 growing degree days to have been
accumulated by that date (Table 56). The weather pattern shifted starting in May with the
temperatures being above normal for the rest of the growing season until September (Figure 46).
By the end of the growing season, the accumulated amount of growing degree days was more
than 100 heat units than the ten-year average but was not as high as the previous season in 2012.
Fruit Chemistry during the Ripening Process
Starting at veraison, sampling of the fruit occurred on a weekly basis until harvest. On the
first sampling date a statistical difference was observed for the amount of total soluble solids
(Figure 48) between the three treatments. The vines adjusted to a low crop load had the highest
average concentration at 13.9o Brix compared to the high and medium treatments that had no
difference with 11.3 and 11.8 o Brix, respectfully (Table 63). The low crop vines continued to
have readings that were higher than the other two treatments on the second sampling date, but a
difference was not detected on the other sampling dates until the date of harvest (Figure 48).

120

For the parameter of pH of the juice from the samples measured, a statistical significance
was found on the second sampling date (8/15/13) with the low crop vines having the highest pH
of 3.09, medium vines having a pH of 3.03 and the high crop vines having the lowest average pH
of 3.02 (Table 63). No other difference was found between treatments except on the date of
harvest (Figure 49). Again at this time the vines with the low crop had the highest pH and the
vines with the high crop had the lowest pH. The last of the basic fruit chemistry parameters that
was measured was titratable acidity. No significant statistical differences appeared when testing
all of the treatments for this parameter over time (Figure 50).
Total phenolic content and anthocyanin concentration were also analyzed as described in
the material and methods section. During the sampling process the samples measured on 8/29/13
had statistical significance (Figure 51). This was the only time throughout the ripening process
that a difference between the three treatments was found. In a similar fashion, only one
difference was found when comparing the anthocyanin concentration over the course of time.
The only detectable occurrence happened on the first sampling date. At this time the high crop
and medium crop vines had the same values while the low crop vines had twice the concentration
(Figure 52). From previous research, the relationship of anthocyanin concentration compared to
the accumulation of total soluble solids appeared to be positively related. Therefore all samples
measured from this experiment had this parameter investigated. The resulting ratio of the
increase of anthocyanin concentration by degree brix showed that throughout the ripening
process the low cropping vines were more efficient at the time of veraison (Figure 53).
Fruit Chemistry, Yield Components, and Canopy Measurements at the Time of Harvest
At the time of harvesting the experimental vines, data was measured on a per vine basis
with the model shoots measured for canopy measurements. The only difference that was
121

statistically different at the time of harvest was the ratio of leaf area per shoot divided by total
shoot length. The low crop vines had the highest ratio followed by the high crop vines. The
medium cropping vines had the lowest ratio (Table 57). All other parameters measured from
analyzing the model shoots (shoot length, lateral length, number of leaves per shoot, and Ravaz
index) showed no difference. For yield components, the treatments were successful at producing
three statistically different yields for the treatments. The high crop vines had an average of 18
kilograms per vine compared to the 13 kilograms per vine for the medium crop and the 9.1
kilograms per vine for the low crop (Table 58). All other parameters (cluster weight, berries per
cluster, and single berry weight) were not statistically significant. Even though the yield
components were not affected except for the number of cluster per vine and the total yield per
vine by the treatments, differences were observed for total soluble solids, pH, and anthocyanin
by degree brix (Table 59). At harvest, the low crop vines had the highest total soluble solids at
25.8 oBrix which was statistically significant compared to the high crop vines (22.4 oBrix) and
medium crop vine (22.9 oBrix). The low crop vines also had the highest pH and the lowest rate of
anthocyanin to degree brix accumulation as compared to the other treatments. All other
parameters were not significant when compared between treatments (Table 59).
Light data obtained from Data logger
Results from the information measured from the data logger placed at this vineyard
location shows that there was a difference in light exposure for the clusters under the three
treatments. The high crop load vines had the highest amount of PAR values verses the other two
crop loads (Figure 47). The difference between the high crop load vines and the control sensors,
which were above the canopies of the vines to get environmental data, had a difference of only
approximately 50 to 80 mmols-1m-2 per hour on average. The medium and low crop load had
122

similar PAR readings at the cluster level with the medium crop vines having the lowest values
and the low crop vines with slightly higher PAR values.

123

TABLE 48. Canopy measurements at harvest in field experiment conducted at SWMREC in
2013.
Crop
load

Shoot
length
(cm)

Lateral
length
(cm)

Number
of leaves
per
shoot

Leaf
area of
model
shoots
(cm^2)

Leaf area/
Total
shoot
length
(cm^2/cm)

Ravaz
Index

High

104.3

25.97

33

1872.4

14.19 ab*

33.69

Medium

112.7

25.73

32

1820.8

12.96 b

25.48

Low

116.5

25.40

39

2099.6

14.68 a

21.94

*Treatment means within columns not followed by the same letter are significantly different at P
≤ 0.05 by Tukey’s HSD. Columns where no letters are present indicate a lack of significant
differences among treatments.

TABLE 49. Yield components in Crop load Field Experiment from SWMREC at harvest in
2013.
Treatment
Yield
Yield
Cluster/
Cluster
Berries /
Single
Tons/acre Kg/vine
vine
weight (g)
cluster
berry
weight (g)
High
13.81 a*
18.17 a
264.0 a
114.85
93.00
1.19
Medium
9.84 b
12.95 b
184.8 b
115.61
94.34
1.18
Low
6.92 c
9.10 c
114.3 c
109.24
91.40
1.17
*Treatment means within columns not followed by the same letter are significantly different at P
≤ 0.05 by Tukey’s HSD. Columns where no letters are present indicate a lack of significant
differences among treatments.

TABLE 50. Fruit chemical composition at harvest for Crop load Experiment at SWMREC in
2013.
Treatment
TSS
pH
TA
Phenolics Anthocyanin Anthocyanin
(oBrix)
(g/L)
(a.u./g)
(mol/g)
/ brix
High
22.4 b*
3.57 b
6.70 a
0.86
1.20
0.053 a
Medium
22.9 b
3.62 ab
6.93 a
0.82
1.13
0.049 ab
Low
25.8 a
3.67 a
6.78 a
0.79
1.14
0.044 b
*Treatment means within columns not followed by the same letter are significantly different at P
≤ 0.05 by Tukey’s HSD. Columns where no letters are present indicate a lack of significant
differences among treatments.

124

SWMREC 2013
1400
Control
Low Crop
Medium Crop
High Crop

1200

-1

-2

PAR (mmols m )

1000
800
600
400
200
0
600

800

1000

1200

1400

1600

1800

2000

2200

Time (hour)
FIGURE 35. Daily average PAR for the different crop load treatments at SWMREC in 2013.
Measurements measured from the beginning of July to harvest. Control was quantum sensors
placed above the canopy of the vines as not to be affected by shade.

125

SWMREC 2013
28
High Crop
Medium Crop
Low Crop

26
24

a

b
b

20

o

TSS ( Brix)

22

18
16
a
14
12

a
b
b
b
b

10
8/5

8/12

8/19

8/26

9/2

9/9

9/16

Date
FIGURE 36. Total soluble solids concentration over time for SWMREC field experiment in
2013. Data points where no letters present show no significantly different at P ≤ 0.05. Data
points not followed by the same letter are significantly different at P ≤ 0.05 by Tukey’s HSD.

126

SWMREC 2013
3.8

3.6

High Crop
Medium Crop
Low Crop

a
ab
b

pH

3.4

3.2
a
ab
3.0

2.8
8/5

b

8/12

8/19

8/26

9/2

9/9

9/16

Date
FIGURE 37. pH concentration over time for SWMREC field experiment in 2013. Data points
where no letters present show no significantly different at P ≤ 0.05. Data points not followed by
the same letter are significantly different at P ≤ 0.05 by Tukey’s HSD.

127

SWMREC 2013
18
High Crop
Medium Crop
Low Crop

16

TA (g/L)

14

12

10

8

6
8/5

8/12

8/19

8/26

9/2

9/9

9/16

Date
FIGURE 38. Titratable Acidity concentration over time for SWMREC field experiment in 2013.
Data points where no letters present show no significantly different at P ≤ 0.05. Data points not
followed by the same letter are significantly different at P ≤ 0.05 by Tukey’s HSD.

128

SWMREC 2013
0.9

Phenolics (a.u./g)

0.8

a

High Crop
Medium Crop
Low Crop

ab

0.7

b

0.6

0.5

0.4
8/5

8/12

8/19

8/26

9/2

9/9

9/16

Date
FIGURE 39. Phenolic concentration over time for SWMREC field experiment in 2013. Data
points where no letters present show no significantly different at P ≤ 0.05. Data points not
followed by the same letter are significantly different at P ≤ 0.05 by Tukey’s HSD.

129

SWMREC 2013
1.4

Anthocyanin (mg/g)

1.2

High Crop
Medium Crop
Low Crop

1.0

0.8

0.6
a
0.4
b
0.2

0.0
8/5

b

8/12

8/19

8/26

9/2

9/9

9/16

Date
FIGURE 40. Anthocyanin concentration over time for SWMREC field experiment in 2013. Data
points where no letters present show no significantly different at P ≤ 0.05. Data points not
followed by the same letter are significantly different at P ≤ 0.05 by Tukey’s HSD.

130

SWMREC 2013
0.07

0.06

Anthocyanin / Brix

a
0.05

ab
b

0.04
a
0.03

0.02

b

High Crop
Medium Crop
Low Crop

b
0.01
8/5

8/12

8/19

8/26

9/2

9/9

9/16

Date
FIGURE 41. Anthocyanin by Brix concentration over time for SWMREC field experiment in
2013. Data points where no letters present show no significantly different at P ≤ 0.05. Data
points not followed by the same letter are significantly different at P ≤ 0.05 by Tukey’s HSD.

131

Discussion
Several research efforts have conducted on improving the micro-climate around the
cluster (Dokoozlian and Kliewer, 1995; Smart, 1973; Smart, 1988). The impacts of light
reaching the fruit zone is related to the temperature at the cluster level (Smart, 2008), which in
turn reflects the chemical composition of the fruit. The hypothesis for the experiment conducted
at SWMREC was to investigate if controlling the crop load of the vine would have impacts on
the canopy architecture, cluster morphology or the fruit chemistry.
The data measured from this experiment conducted at SWMREC shows that there was an
effect of the crop load directly on vine growth and indirectly on fruit chemistry. The crop load
placed a stress on the shoots, particularly when analyzed as ratio between leaf area to shoot
length. A significant difference in the PAR measurements measured by the data logger (Figure
47) was also observed throughout the growing season. This information follows what has been
observed in other work (Dokoozlian, 1995). The relationship of light and canopy density effects
on fruit microclimate has been described by many researchers in the literature (Gladstone and
Dokoozlian, 2003). The results from the instruments in the field showed that the vines with
smaller canopies did have higher PAR readings at the cluster zone as a result of the crop load.
As for the effects of the shoot thinning on the different canopy architecture, the
experiment conducted at HTRC in 2013 has provided a basis for future work on canopy density
and the ideal ratio of shoots needed per length of cordon to allow for optimal exposure to the
fruit zone. As seen by the data collected, there is still room for improvement in reaching the ratio
of the canopy as described by Smart (1973). No definitive results were achieved to be able to
suggest which is an acceptable shoot density for a given trellis system. The results from this
experiment show that there is a trend that the vines with more shoots per length of cordon have
132

higher leaf area, longer shoots, and more leaves which can affect the microclimate of the fruit
zone. The most notable statistical difference was observed in the vines with less shoots per
cordon length. Such vines had higher cluster weights and more berries per cluster (Table 53).

133

Conclusion
The experiments conducted in 2013 offer a better understanding of the relationship
between trellis systems and shoot density and the impacts they have on the vines and fruit
produced. The information measured will be a starting point for further study, especially when
dealing with the experimental trellis which allows for the vines to be manipulated by either
opening or closing the trellis system. This would allow for better control of the microclimate of
the fruit zone. The results that point to the quality of the Marquette grapes is related to how the
vines are trained and that with better microclimate conditions a significant improvement in grape
quality can result.
When examining the information from the crop load experiment conducted at SWMREC,
the result of the high yields is the most important factor to consider. The high yielding vines at
13 tons per acre were able to produce fruit that had chemical properties suitable for commercial
wine production. This variety’s potential to be a high tonnage producing grape will help increase
the popularity in areas not suitable for other varieties. Being able to produce this amount of yield
per vine without the noticeable high acids as reported in other northern locations suggests that
the Marquette cultivar does responds well to growing sites that are favorable for general grape
production.

134

APPENDICES

135

APPENDIX A: FRUIT CHEMICAL COMPOSITION OVER TIME

TABLE 51. Fruit chemical composition over time for HTRC field experiment in 2013. High Wire Cordon (HWC) adjusted to either
three shoots per 30 linear centimeters of cordon is represented by (3S) and the six shoots per 30 linear centimeters of cordon is
represented by (6S).

High Wire Cordon Trellis System (HWC)
Date

Shoot
Density

(oBrix)

TSS

pH

TA
(g/L)

Phenolics
(a.u./g)

3S

Anthocyanin Anthocyanin
(mol/g)
/ brix

14.3*

2.86

17.91

0.58 a

0.79

0.040 a

6S

14.0

2.83

19.15

0.37 b

0.80

0.026 b

3S

16.6

2.93

15.72

0.82

0.96

0.057

6S

16.7

2.97

15.95

0.79

0.84

0.050

3S

18.9

3.08

12.76

0.87

1.23

0.065

6S

18.8

3.04

13.48

0.76

1.03

0.055

3S

20.7

3.23

10.48

1.01

1.75

0.084

6S

20.5

3.14

10.89

0.95

1.60

0.078

3S

21.3

3.28 a

9.64

1.17 a

1.95 a

0.092 a

6S

21.7

3.38 b

10.10

0.82 b

1.32 b

0.062 b

3S

24.5

3.54

8.35

1.25

1.90

0.078

6S

24.4

3.52

8.81

1.26

1.81

0.075

8/7/2013

8/14/2013

8/21/2013

8/28/2013

9/4/2013
9/11/2013
(Harvest)

*Treatment means within each date not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. Rows within each date where no letters are present indicate a lack of significant differences among treatments.
136

TABLE 52. Fruit chemical composition over time for HTRC field experiment in 2013. Geneva Double Curtain (GDC) adjusted to
either three shoots per 30 linear centimeters of cordon is represented by (3S) and the six shoots per 30 linear centimeters of cordon is
represented by (6S).

Geneva Double Curtain Trellis System (GDC)
Date

Shoot
Density

TSS
(oBrix)

pH

TA
(g/L)

Phenolics
(a.u./g)

Anthocyanin Anthocyanin
(mol/g)
/ brix

3S

13.8*

2.81

19.45

0.72

0.38

0.027

6S

13.1

2.76

18.93

0.79

0.40

0.030

3S

15.7

2.89

16.24

0.81 a

0.76

0.049

6S

15.5

2.82

15.49

0.70 b

0.74

0.048

3S

18.0

2.97

13.80

0.65

0.85

0.047

6S

18.3

2.98

12.93

0.70

0.92

0.050

3S

19.5

3.02

11.26

0.90

1.38

0.071

6S

20.7

3.04

10.46

0.90

1.37

0.066

3S

21.9

3.28

9.75

0.65

1.01

0.046

6S

21.2

3.21

9.00

0.90

1.42

0.067

3S

24.6

3.43 a

7.72

1.11

1.67

0.067

6S

23.5

3.33 b

7.53

1.05

1.70

0.071

8/7/2013

8/14/2013

8/21/2013

8/28/2013

9/4/2013
9/11/2013
(Harvest)

*Treatment means within each date not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. Rows within each date where no letters are present indicate a lack of significant differences among treatments.

137

TABLE 53. Fruit chemical composition over time for HTRC field experiment in 2013. Experimental Moving Trellis (MT) adjusted to
either three shoots per 30 linear centimeters of cordon is represented by (3S) and the six shoots per 30 linear centimeters of cordon is
represented by (6S)

Moving Trellis System (MT)
Date

Shoot
Density

TSS
(oBrix)

pH

TA
(g/L)

Phenolics
(a.u./g)

Anthocyanin Anthocyanin
(mol/g)
/ brix

3S

12.6*

2.87

18.33

0.77

0.22

0.017

6S

12.4

2.85

19.37

0.84

0.29

0.023

3S

14.8

2.98

17.02

0.76

0.50

0.036

6S

15.6

2.91

16.75

0.78

0.60

0.038

3S

17.5

3.08

13.14

0.69

0.83

0.046

6S

17.9

3.01

14.39

0.69

0.89

0.049

3S

18.5

3.20

12.02

0.74

1.18

0.063

6S

19.1

3.23

12.31

0.756

1.15

0.060

3S

19.8

3.41

10.28

0.83

1.35

0.069

6S

20.1

3.43

10.57

1.02

1.56

0.078

3S

23.2

3.54

8.08

1.16

1.89

0.079

6S

22.8

3.58

8.85

1.21

1.83

0.081

8/7/2013

8/14/2013

8/21/2013

8/28/2013

9/4/2013
9/11/2013
(Harvest)

*Treatment means within each date not followed by the same letter are significantly different at P ≤ 0.05 by LSD’s pairwise
comparison. Rows within each date where no letters are present indicate a lack of significant differences among treatments.

138

TABLE 54. Fruit chemical composition over time for SWMREC field experiment in 2013.
Date

8/9/2013

8/15/2013

8/20/2013

8/29/2013

Crop
Load
High

TSS

pH

11.3 b*

2.91

Medium

11.8 b

2.97

16.75

0.56

0.22 b

0.018 b

Low

13.9 a

2.94

15.21

0.60

0.40 a

0.029 a

High

12.6 b

3.02 b

13.08

0.59

0.37

0.030

Medium

12.5 b

3.03 ab

13.39

0.54

0.37

0.030

Low

14.9 a

3.09 a

12.82

0.60

0.57

0.038

High

14.2

3.08

12.35

0.58

0.59

0.041

Medium

15.6

3.17

12.09

0.68

0.74

0.047

Low

14.8

3.19

11.79

0.68

0.77

0.053

High

17.1

3.31

8.44

0.67 b

0.90

0.053

Medium

18.1

3.39

8.87

0.81 ab

1.06

0.058

Low

19.0

3.45

8.69

0.86 a

1.23

0.064

High

22.4 b

3.57 b

6.70

0.86

1.20

0.053 a

Medium

22.9 b

3.62 ab

6.93

0.82

1.13

0.049 ab

Low

25.8 a

3.67 a

6.78

0.79

1.14

0.044 b

(oBrix)

TA (g/L) Phenolics
(a.u./g)
15.03
0.51

Anthocyanin Anthocyanin
(mol/g)
/ brix
0.19 b
0.016 b

9/13/2013
(Harvest)
*Treatment means followed by the same letter within a column are not significantly different at the α=0.05 level at P ≤ 0.05 by
Tukey’s HSD. Rows within each date where no letters are present indicate a lack of significant differences among treatments

139

APPENDIX B: TYPE 3 TABLES OF ANVOA FOR HTRC

TABLE 55. Analysis of variance (ANOVA) for Total soluble solids (reported as Degrees Brix)
at HTRC in 2013.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
28.6
4.73
Trellis
1
28.6
1.69
Density
2
28.6
0.50
Trellis*Density

Pr > F
0.0168
0.2042
0.6106

TABLE 56. Analysis of variance (ANOVA) for pH at HTRC in 2013.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
30.5
7.87
Trellis
1
30.6
0.46
Density
2
30.5
1.19
Trellis*Density

Pr > F
0.0018
0.5043
0.3169

TABLE 57. Analysis of variance (ANOVA) for Titritable acidity at HTRC in 2013.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
29.7
3.03
Trellis
1
29.7
0.96
Density
2
29.7
0.68
Trellis*Density

Pr > F
0.0633
0.3359
0.5126

TABLE 58. Analysis of variance (ANOVA )Total phenolic content at HTRC in 2013.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
30.5
3.04
Trellis
1
30.6
0.03
Density
2
30.5
0.37
Trellis*Density

Pr > F
0.0625
0.8686
0.6923

TABLE 59. Analysis of variance (ANOVA) for Anthocyanin content at HTRC in 2013.
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
30.6
0.64
Trellis
1
30.6
0.04
Density
2
30.6
0.17
Trellis*Density

140

Pr > F
0.5334
0.8413
0.8424

TABLE 60. Analysis of variance (ANOVA) for Anthocyanin by Degrees Brix accumulation
content at HTRC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
30.5
1.49
Trellis
1
30.6
0.01
Density
2
30.5
0.27
Trellis*Density

141

Pr > F
0.2407
0.9069
0.7685

APPENDIX C: TYPE 3 TABLES OF ANVOA FOR SWMREC
TABLE 61. Analysis of variance (ANOVA) for Shoot length at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
2.28
Crop_Load
3
75
1.03
Vine
6
75
1.96
Vine*Crop_

Pr > F
0.1089
0.3983
0.0946

TABLE 62. Analysis of variance (ANOVA) for Lateral length at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
0.09
Crop_Load
3
75
0.19
Vine
6
75
1.00
Vine*Crop_

Pr > F
0.9098
0.9448
0.4245

TABLE 63. Analysis of variance (ANOVA) for Number of leaves per shoot at SWMREC in
2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
2.08
Crop_Load
3
75
0.22
Vine
6
75
2.00
Vine*Crop_

Pr > F
0.1321
0.9260
0.0879

TABLE 64. Analysis of variance (ANOVA) for Leaf area at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
1.65
Crop_Load
3
75
0.53
Vine
6
75
1.81
Vine*Crop_

Pr > F
0.1988
0.7158
0.1198

TABLE 65. Analysis of variance (ANOVA) for Leaf area by total shoot length at SWMREC in
2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
3.15
Crop_Load
3
75
0.75
Vine
6
75
0.83
Vine*Crop_

142

Pr > F
0.0482
0.5631
0.5297

TABLE 66. Analysis of variance (ANOVA) for Ravaz Index at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
1.24
Crop_Load
3
75
0.68
Vine
6
75
0.54
Vine*Crop_

Pr > F
0.3401
0.5873
0.7092

TABLE 67. Analysis of variance (ANOVA) Yield (kilograms per vine) at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
16.73
Crop_Load
3
75
0.58
Vine
6
75
0.13
Vine*Crop_

Pr > F
0.0014
0.6453
0.9654

TABLE 68. Analysis of variance (ANOVA) Clusters per vine at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
39.00
Crop_Load
3
75
0.20
Vine
6
75
0.82
Vine*Crop_

Pr > F
<.0001
0.8955
0.5466

TABLE 69. Analysis of variance (ANOVA) Clusters weight at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
0.28
Crop_Load
3
75
1.41
Vine
6
75
0.73
Vine*Crop_

Pr > F
0.7588
0.2373
0.6672

TABLE 70. Analysis of variance (ANOVA) Berries per cluster at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
0.08
Crop_Load
3
75
1.53
Vine
6
75
1.21
Vine*Crop_

143

Pr > F
0.9223
0.2031
0.3053

TABLE 71. Analysis of variance (ANOVA) Single berry weight at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
0.71
Crop_Load
3
75
1.33
Vine
6
75
1.20
Vine*Crop_

Pr > F
0.4926
0.2649
0.3081

TABLE 72. Analysis of variance (ANOVA) Total soluble solids at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
32.53
Crop_Load
3
75
4.16
Vine
6
75
3.03
Vine*Crop_

Pr > F
<.0001
0.0042
0.0150

TABLE 73. Analysis of variance (ANOVA) pH at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
5.26
Crop_Load
3
75
1.58
Vine
6
75
0.90
Vine*Crop_

Pr > F
0.0072
0.1869
0.4856

TABLE 74. Analysis of variance (ANOVA) Titritable acidity at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
1.15
Crop_Load
3
75
2.25
Vine
6
75
1.32
Vine*Crop_

Pr > F
0.3206
0.0709
0.2641

TABLE 75. Analysis of variance (ANOVA) Phenolic content at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
1.09
Crop_Load
3
75
3.17
Vine
6
75
1.96
Vine*Crop_

144

Pr > F
0.3401
0.0183
0.0939

TABLE 76. Analysis of variance (ANOVA) Anthocyanin content at SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
0.10
Crop_Load
3
75
3.06
Vine
6
75
2.96
Vine*Crop_

Pr > F
0.9016
0.0214
0.0169

TABLE 77. Analysis of variance (ANOVA) Anthocyanin by brix accumulation content at
SWMREC in 2013
Type 3 Tests of Fixed Effects
Effect
Num DF Den DF F Value
2
75
3.27
Crop_Load
3
75
2.97
Vine
6
75
2.62
Vine*Crop_

145

Pr > F
0.0434
0.0245
0.0307

CHAPTER IV
EXPERIMENTIAL WINE PRODUCTION AND CHEMICAL AND
SENSORY ANAYLYSIS FOR 2012 AND 2013

Introduction
The developments of new super cold hardy varieties (SHC) has been focused on being
able to produce grapes that are suitable for wine production in locations that are not ideal for
other varieties because of damage by cold stress. One of the advantages of SCH varieties is that
increased cold hardiness will allows area not suitable for production of classical FrenchAmerican or European varieties, to be potential suitable for other types of grapes. For this reason
SCH varieties are gaining interest for commercial production in cool-cold viticulture areas like
Michigan. As seen from the data collected by the recent survey conducted in part with the
Northern Grape Project, red varieties of SCH is the largest growing segment of hybrids being
planted in Michigan (Tuck and Gartner, 2012). This trend has also been observed in other states
in the Mid-Atlantic region, Midwest, and Mid-south regions of the United States. In particular,
the plantings in Michigan have favored the planting of Marquette, which is the fastest growing
segment of SCH variety coming into production since being released from the University of
Minnesota breeding program.
As with any new cultivar release, little information has been published on aspects of both
viticultural and enological performance of SCH varieties. The information that has been
published has been released within the last three years. In order to fully evaluate the new SCH
cultivar ‘Marquette’, all of the treatments that had been applied in the field were also analyzed at
the sensory level after the field samples underwent micro fermentation. At the time of this MS

146

thesis, there is currently no peer reviewed research on the impacts of different trellis systems,
crop loads, or from different locations for the cultivar ‘Marquette’. To be able to compare these
wines a sensory evaluation and chemical analysis were performed on wines produced in 2012
and 2013.

Materials and Methods
Production Facility
All of the wines were produced from grapes that were either collected from the field
experiments conducted in 2012 or 2013 at both research plots in SWMREC and HTRC. The
wines were produced at the experimental winery, Spartan Cellars, located on the property of the
Horticultural Teaching and Research Center, in Holt, MI on Michigan State University main
campus. At this facility, fruit processing, fermentations, and general wine making processes were
carried out following the micro-vinification protocol developed by the regional project NE1020
“Coordinated Wine Grape Variety Evaluations in The Eastern USA”. Chemical analysis and
sensory evaluations were conducted at the Viticulture and Enology laboratory located at Plant
and Soil Science Building, Michigan State University, East Lansing MI.
Enology Procedure
The wines were made following the NE-1020 protocol developed by Dr. Stephen Menke
at Colorado State University with only slight changes to fit the amount of material collected at
harvest and equipment available at Spartan Cellars. All of the fruit was hand harvested and
material used for fermentation was a sub sample collected from the overall treatment yield. The
allotted amount of fruit used for the each fermentation and collected per treatment in both years
147

and locations are described in detail (Table 78). All of the samples collected were sorted to allow
for less than five percent fruit rot in each fermentation vessel. After fruit sorting, all of the fruit
was placed into five gallon plastic buckets with attached lids and cooled to 10oC overnight to
allow field heat dissipation. After cooling, the fruit was crushed and de-stemmed using a
stainless steel motorized crusher/de-stemmer produced by Enoitalia (Florence, Italy). The must
was collected in small food grade plastic buckets after crushing and had potassium metabisulfite
added to ensure a concentration of 50 ppm of SO2. All of the treatments had 50 ml of sample
collected in centrifuge tubes that were used for chemical analysis later in the laboratory. For each
year all fruit processing was completed the same day to insure that that all experimental
treatments were subjected to the same cellar conditions. The equipment was washed and resanitized between treatments.
The fermentation was not started until the must had reached a temperature between 1825o C. During the time of must warming up, yeast was rehydrated using the protocols as
recommended by Lallemand four step process. The yeast chosen for all of the fermentation in
both 2012 and 2013 was RC212. This yeast is produced by Lalvin (Lallemand Inc., Santa Rosa,
CA, USA) and is selected by the BIVB (Bureau Interprofessionnel des Vins de Bourgogne).
Throughout the primary fermentation a temperature of 18-25o C was maintained and measured
with a thermometer placed inside the fermentation vessels. No nutrients or other additives were
used during the primary fermentation. Each fermentation vessel had a lid loosely placed over to
keep dust and insects out. A thin sheet was also placed over all of the buckets to help keep
insects away from the vessels. Throughout the primary fermentation, the cap was broken up
using a sanitized wooden paddle twice a day until pressing was completed.

148

Pressing of the must was conducted after the wine had shown less than 4o Brix when
tested with a hydrometer. The pressing of the must was completed using a #30 wooden basket
press with a ratchet mechanism which could hold a maximum capacity of 31 liters of must. After
pressing, 50 ml centrifuge tubes were used to collect wine samples to be analyzed later in the
laboratory. The wine was transferred to 3.8 liter glass carboys with an airlock attached to each
carboy. The wine was allowed to complete primary fermentation inside of this vessel. When
primary fermentation was completed, an additional 50 ppm of SO2 was added to each wine
before being racked into new sanitized 3.8 liter glass carboys.
All of the research wines underwent the process for clarification and settling. In order to
complete this process, bentonite was added to each glass carboy after the first racking postfermentation at a rate of 1.25 grams per 3.8 liters of wine. This was done by suspending the
bentonite in 100 ml of hot distilled water. The suspended solution was then added and
thoroughly mixed in the carboy. The wines were allowed to settle for approximately two months
in a walk-in cooler set between 15-20o C and after the wines had one more addition of 50 ppm of
SO2 right before bottling. During the bottling process 50 ml of wine was collected into glass jars
to be used for chemical analysis at a later date. Green 750 ml Bordeaux bottles were used and
filled to the appropriate volume and then sealed with a natural cork. All of the wines were stored
in a climate controlled cellar kept at a temperature of 10oC until chemical analysis and sensory
evaluation. Samples collected throughout different sampling times were stored in a -20o C freeze
to preserve the samples until laboratory chemical analysis.
Sensory Evaluation
A panel of twelve people from industry and academia (10 males and 2 females with ages
ranging from 23 to 67) was selected to judge all of the experimental wine. These judges were
149

selected based on experience and knowledge of cold hardy hybrids and the winemaking process.
In order to judge the experimental wines, sensory descriptors for Marquette wines were provided
by K. Cook of University of Minnesota (Table 79). From the list of sensory descriptions, several
of them were selected and used for the evaluations of the wines that were produced. These
aromatic characteristics were dark fruit, vegetal, pepper, floral, and musty. Other aspects of the
wine were chosen to be evaluated by the judges. The score sheet that was created and given to
the judges (Figure 42) was split into three categories. Tasting procedure and score sheet was
based on recent work on the subject of sensory analysis for wine (Etaio et al., 2010). All wines
were assigned a random three digit number to allow for no bias from the judges. Each sample of
wine was poured using a volumetric precision pourer into clear red wine classes. The volume of
wine sample used for each wine being tested was approximately 35 ml and severed at room
temperature.
The wines were split into five flights (Table 80) with each flight focused on an individual
research objective. Flight one was focused on detecting a difference between the different trellis
systems (HWC, GDC, MT) when the vines were trained to the same shoot density of six shoots
per 30 centimeter of linear cordon. Flight two was a similar test as flight one but with a
difference in the shoot density (three shoots per 30 centimeter of linear cordon). Both of these
flights were evaluating wines produced from fruit from the 2013 season. Flight three, four and
five were designed to detect differences between wines produced from fruit that came from the
same trellis but from fruit that originated from a primary bud or fruit that originated from a
secondary bud.

150

Chemical Analysis
For laboratory analysis, one 750 ml bottle was opened per each single experimental.
Wine samples were warmed to room temperature before being used for chemical analysis.
Samples were separated based on whether they were must samples or wine samples. The must
samples were analyzed using the same procedures as used for evaluating the harvest and seasonal
sampling of fruit used in chapters two and three (page 40). Those tests included total soluble
solids, pH, tritatable acidity, total phenolics, and anthocyanin content. For wine samples,
additional assessments were performed to measure alcohol content and volatile acidity (VA).
The test for VA and alcohol content was performed using the protocol of Iland and Patrick
(2004). A volume of 10 ml of wine was allotted for VA testing, which was placed into the cash
still and used for the calculation as expressed as grams per liter of acetic acid.
Alcohol content was determined using the technique of ebulliometry. A volume of 50 ml
of wine was allotted and placed into an ebulliometer and used to find the boiling point of the
wine sample. This value was compared to the boiling point of the same volume of distilled water
which was known as ebulliometer degree. A chart was used to compare the ebulliometer degree
to determine the alcohol content by percent by volume (Iland et al., 2004). The ebulliometer was
rinsed several times with distilled water and then with the wine being tested in order to allow for
accuracy of the measurements. This process was repeated for all of the wines produced. All
parameters measured from the wine were analyzed three times and the means used for
comparisons (Table 81).
Statistical Analysis
All data measured was tested to ensure normality and analysis of variance (ANOVA) was
performed using PROC MIXED in SAS version 9.3 for Windows (Cary, NC). All pair-wise
151

comparisons were performed and tested using Tukey’s Test with the alpha equal to 0.05. For
data analysis measured from the sensory evaluation, a fixed-effect ANOVA was conducted with
each term on the score sheet being tested separately. The wines were judged on a scoring system
based from 1 to 7 with 1 being a negative rating and 7 being a desirable rating. Mean ratings
from the judges for each category for each wine was compared to the wine(s) in the same flight.
All of the data analysis was done using SAS version 9.3 for Windows.

152

TABLE 78. Description of amount of fruit measured per treatment and wine produced per
treatment.
Year

Location

Trellis

Treatment *

Amount of fruit
Amount of
measured (kg)
wine (ml)
HTRC
HWC
Primary
6.0
1500
HTRC
HWC
Secondary
6.0
1500
HTRC
MT
Primary
6.0
1500
2012
HTRC
MT
Secondary
6.0
1500
HTRC
GDC
Primary
6.0
1500
HTRC
GDC
Secondary
6.0
1500
SWMREC HWC
Primary
6.0
1500
2012
SWMREC HWC
Secondary
6.0
1500
HTRC
HWC
3S
13.6
6000
HTRC
HWC
6S
13.6
6000
HTRC
MT
3S
13.6
6000
2013
HTRC
MT
6S
13.6
6000
HTRC
GDC
3S
13.6
6000
HTRC
GDC
6S
13.6
6000
SWMREC HWC
Low Crop
27.2
11340
2013 SWMREC HWC
Medium Crop
27.2
11340
SWMREC HWC
High Crop
27.2
11340
* Primary = Fruit that originated from a primary shoot, Secondary = Fruit that originated from a
secondary shoot, 3S = three shoos per 30 cm linear length of cordon, 6S = six shoots per 30 cm
linear length of cordon, Low Crop = approx.125 clusters per vine, Medium Crop = approx. 200
cluster per vine, High Crop = approx. 275 clusters per vine.

153

TABLE 79. A list of commonly found sensory characters from Marquette grapes. Table courtesy
of Dr. Katie Cook at University of Minnesota.
Descriptive Term
Artificial Banana
Black Currant
Cooked Berry
Dark Fruit
Grapefruit
Cooked Vegetable
Fresh Green
Woody
Hay
Pepper
Spice
Floral
Ethanol
Chemical
Caramel
White Mushroom
Dried Mushroom
Tamari

Definition
The aroma associated with artificial banana flavor
Fruity aromatic characteristic of black currants
A fruity aroma note associated with blackberries,
raspberries, blue berries, and strawberries.
Fruity aromatic characteristic of black fruit, plum,
cherry, and jam
Fruity aromatic characteristic of grapefruit.
An overall term that describes the aromatic of cooked
vegetables in general
A “green” aroma/flavor typical of fresh grass or fresh
green vegetables
A woody aroma note associated with cedar, oak,
described as toasted, smoky.
Aromatic associated with sweet dry grasses
Spicy, pungent aromatic characteristic of white pepper
Sweet spicy aroma characteristic of spices.
A sweet, fragrant aromatic associated with roses
A pungent aroma associated with ethanol
Note associated with kerosene, gasoline, solvents or
hydrocarbons.
An overall term reminiscent of chewy caramel
An aromatic associated with fresh raw mushrooms;
damp, earthy, and musty.
A dark, pungent, aromatic associated with dried mixed
mushrooms.
A caramelized note associated with tamari.

TABLE 80. Order of wines presented to the untrained panel of experts.
Flight
Number

Year

1

2013

2

2013

3

2012

4

2012

5

2012

Assigned Random
Number
#477 = HWC
#827 = GDC
#373 = MT
#834 = HWC
#314 = GDC
#773 = MT
#368 = Primary Fruit
#602 = Secondary Fruit
#805 = Primary Fruit
#980 = Secondary Fruit
#575 = Primary Fruit
#418 = Secondary Fruit
154

Information
High shoot density
(6 shoots/30cm of cordon)
Low shoot density
(3 shoots/30cm of cordon)
Trained on HWC
Trained to GDC
Trained to MT

FIGURE 42. Example of score sheet used for experimental wine evaluation. Selections of
categories were based off from Marquette grapes. Information courtesy of Dr. Katie Cook at
University of Minnesota.

155

Results
Prior to the sensory analysis, all of the wine was subjected to basic laboratory analysis as
described in the material and methods section. The wines had similar values for titratable acidity,
volatile acidity, and pH. Noticeable differences in the alcohol percentage; from 10.2% up to
14.7% (Table 81)1.
From the sensory analysis, several trends became apparent after running AOV on the data
measured from the score sheets. The wine produced in 2012 from the vines trained to High Wire
Cordon had no noticeable difference in either visual and aromatics components between the
primary fruit or the secondary fruit (Table 82). Statistically, the perception of sweetness was
different with the primary fruit wine scoring an average of 2.3 and the secondary fruit wine
scoring 1.8 (Table 83). The closer the number was to 1 the wine was perceived to be more dry
and the closer the wine was to 4 the wine was perceived as more semi-sweet (Figure 43). All of
this information can be seen in the spider graph (Figure 44) associated with these two wines.
When comparing the 2012 wine produced from the vines trained to Geneva Double
Curtain, several differences in color intensity, color hue, alcohol, acidity, astringency, and body
were found (Table 84; Table 85). The wine produced from the primary fruit had more perceived
intense color that was more toward the purple hue in color as compared to the secondary fruit.
There are statistical differences in the perceived alcohol, acidity, astringency, and body
categories between the primary fruit and the secondary fruit when examining the spider graph

1

Wine produced from the MT Primary fruit in 2012 had a high amount of volatile acidity caused

by an error in the winemaking process.

156

(Figure 44). This showed that the tasting panel perceived that these two wines were not of the
same quality. The last set of wines produced in 2012, from vines trained to the experimental
moving trellis showed differences in color hue, dark fruit and floral aromatics, acidity, and
astringency (Table 86, Table 87, Figure 45). As mentioned the MT wine produced in 2012 from
primary fruit was oxidized at the time of tasting. This resulted as the wine being judged as
having a major flaw which impacted all of the sensory components that were being examined in
this study.
The wines produced in 2013 focused on looking at the differences caused by the each
trellis system when the canopies were adjusted to the same canopy density. The first set of wines
was from canopies of three shoots per linear 30 centimeters of cordon (Figure 46). Comparing
the three trellis systems at the same canopy density, several differences in the perceived quality
of the wine do appear in the data. In the visual analysis, there was no difference in color
intensity, but there was a statistical difference in color hue. The GDC had a higher observed
amount of a purple hue (a score closer to 7 compared to a score of 1 being more deep red) as
compared to the MT and HWC wines (Table 88). There were also differences in aromatics
intensity with the MT having less intensity when compared to the HWC and GDC. The same MT
wine also had more intense musty and vegetal aromatics. There were also differences in alcohol,
acidity, and body (Table 89). The HWC wines had the highest perceived alcohol content, were
less acidic and more body when compared to the GDC or MT wines.
The same exact trends were found when comparing the three trellis systems that had the
canopies adjusted to six shoots per linear 30 centimeters of cordon (Figure 47). In aromatics,
differences were found again in the categories of intensity, dark fruit and musty (Table 90). The
wine produced from the HWC had a higher intensity of aromatics and higher dark fruit aromas
157

than the MT and GDC trellis. While the MT trellis had a higher intensity of musty aroma
compare to the HWC and the GDC. When the same wines were done with the taste analysis, the
only difference was found when looking at the body of the wines. The MT wine had the least
amount of body as compared to HWC and GDC (Table 91).

158

TABLE 81. Chemical analysis of all wines produced from the Frost experiments conducted in
2012 and the crop load and shoot density experiments in 2013.
Year

Location

Trelli Treatment *
Alcohol
pH
TA
VA
s
Percentage
(g/L)
(g/L)
2012
HTRC
HWC
Primary
14.0
3.48
7.0
0.2
2012
HTRC
GDC
Primary
13.8
3.60
7.0
0.4
2012
HTRC
MT
Primary
13.9
3.38
7.6
1.4
2012
HTRC
HWC
Secondary
12.4
3.55
6.9
0.1
2012
HTRC
GDC
Secondary
12.5
3.74
6.5
0.2
2012
HTRC
MT
Secondary
12.4
3.33
7.1
0.3
2013
HTRC
HWC
3S
13.6
3.50
6.6
<0.1
2013
HTRC
GDC
3S
10.2
3.38
6.3
<0.1
2013
HTRC
MT
3S
14.5
3.43
6.4
<0.1
2013
HTRC
HWC
6S
14.5
3.44
7.2
<0.1
2013
HTRC
GDC
6S
13.2
3.25
7.7
<0.1
2013
HTRC
MT
6S
11.2
3.54
7.4
<0.1
2013
SWMREC HWC
High Crop
14.0
3.68
8.19
<0.1
2013
SWMREC HWC Medium Crop
14.5
3.48
7.96
<0.1
2013
SWMREC HWC
Low Crop
14.7
3.56
10.34
<0.1
* Primary = Fruit that originated from a primary shoot, Secondary = Fruit that originated from a
secondary shoot, 3S = three shoos per 30 cm linear length of cordon, 6S = six shoots per 30 cm
linear length of cordon, Low Crop = approx.125 clusters per vine, Medium Crop = approx. 200
cluster per vine, High Crop = approx. 275 clusters per vine.

159

Color Intensity
7

Body

Color Hue

6
5

Astringency

Aromatics
Intensity

4
3
2

Acidity

Dark Fruit

1

Alcohol

Vegetal

Sweetness

Pepper
Musty

Floral

HWC (P)
HWC (S)

FIGURE 43. Results from comparing wines produced from fruit collected from vines trained to
High Wire Cordon in 2012 .HWC (P) represents wine produced from fruit collected from shoots
that originated from the primary bud. HWC (S) represents wine produced from fruit that
originated from a secondary shoot.

TABLE 82. Wine tasting results of visual and aromatics analysis from produced from fruit on
vines that were trained to High Wire Cordon at HTRC in 2012.
Visual
Aromatics
Color Aromatics Dark Vegetal Pepper Floral Musty
Trellis Fruit Color
Intensity Fruit
Type Intensity Hue
P
5.3 *
5.2
4.9
4.6
2.2
1.9
3.0
2.3
HWC
S
5.3
4.8
4.5
3.5
2.6
2.1
2.8
2.2
*Treatment means followed by the same letter within a column are not significantly different at
the α=0.05 level. Columns where no letters are present indicate a lack of significant differences
among treatments.

160

TABLE 83. Wine tasting results of taste analysis from produced from fruit on vines that were
trained to High Wire Cordon at HTRC in 2012.
Taste
Acidity

Sweetness Alcohol
Astringency
Body
Fruit
Type
P
2.3 a*
3.8
2.7
3.2
3.8
HWC
S
1.8 b
3.4
2.9
3.3
3.8
*Treatment means followed by the same letter within a column are not significantly different at
the α=0.05 level. Columns where no letters are present indicate a lack of significant differences
among treatments.
Trellis

161

Body

Color Intensity
7

Color Hue

6
5

Astringency

Aromatics Intensity

4
3
2

Acidity

Dark Fruit

1

Alcohol

Vegetal

Sweetness

Pepper
Musty

GDC (P)
GDC (S)

Floral

FIGURE 44. Results from comparing wines produced from fruit collected from vines trained to
Geneva Double Curtain in 2012. GDC (P) represents wine produced from fruit collected from
shoots that originated from the primary bud. GDC (S) represents wine produced from fruit that
originated from a secondary shoot.

TABLE 84. Wine tasting results of visual and aromatics analysis from produced from fruit on
vines that were trained to Geneva Double Curtain at HTRC in 2012.
Visual
Aromatics
Color Aromatics Dark Vegetal Pepper Floral Musty
Trellis Fruit Color
Intensity Fruit
Type Intensity Hue
P
5.7 a*
4.9 a
4.8
4.8
2.3
2.5
3.0
2.5
GDC
S
4.8 b
4.2 b
4.2
3.9
3.1
2.3
2.6
2.2
*Treatment means followed by the same letter within a column are not significantly different at
the α=0.05 level. Columns where no letters are present indicate a lack of significant differences
among treatments.

162

TABLE 85. Wine tasting results of taste analysis from produced from fruit on vines that were
trained to Geneva Double Curtain at HTRC in 2012.
Taste
Acidity

Sweetness Alcohol
Astringency
Body
Fruit
Type
P
2.3 *
4.2 a
3.4 a
3.7 a
4.3 a
GDC
S
2.2
3.5 b
2.2 b
3.1 b
2.6 b
*Treatment means followed by the same letter within a column are not significantly different at
the α=0.05 level. Columns where no letters are present indicate a lack of significant differences
among treatments.
Trellis

163

Body

Color Intensity
7

Color Hue

6
5

Astringency

Aromatics Intensity

4
3
2

Acidity

Dark Fruit

1

MT (P)
MT (S)

Alcohol

Vegetal

Sweetness

Pepper
Musty

Floral

FIGURE 45. Results from comparing wines produced from fruit collected from vines trained to
Experimental Moving Trellis in 2012. MT (P) represents wine produced from fruit collected
from shoots that originated from the primary bud. MT (S) represents wine produced from fruit
that originated from a secondary shoot.

TABLE 86. Wine tasting results of visual and aromatics analysis from produced from fruit on
vines that were trained to Moving Trellis at HTRC in 2012.
Visual
Aromatics
Color Aromatics Dark Vegetal Pepper Floral Musty
Trellis Fruit Color
Intensity Fruit
Type Intensity Hue
P
4.3 *
3.0 b
3.5
2.8 b
2.1
1.9
1.8 b
2.7
MT
s
4.8
4.4 a
4.1
3.9 a
1.9
2.3
2.8 a
2.3
*Treatment means followed by the same letter within a column are not significantly different at
the α=0.05 level. Columns where no letters are present indicate a lack of significant differences
among treatments.

164

TABLE 87. Wine tasting results of taste analysis from produced from fruit on vines that were
trained to Moving Trellis at HTRC in 2012.
Taste
Acidity

Sweetness Alcohol
Astringency
Body
Fruit
Type
P
2.6 *
3.6
4.3 a
4.8 a
2.5
MT
S
2.3
3.8
3.1 b
3.7 b
3.3
*Treatment means followed by the same letter within a column are not significantly different at
the α=0.05 level. Columns where no letters are present indicate a lack of significant differences
among treatments.
Trellis

165

Body

Color Intensity
7

Color Hue

6
5

Astringency

Aromatics Intensity

4
3
2

Acidity

Dark Fruit

HWC
GDC

1

MT
Alcohol

Vegetal

Sweetness

Pepper
Musty

Floral

FIGURE 46. Results from comparing wines produced from fruit collected from vines used in
various trellis systems with a shoot density of three shoots per 30 centimeters of linear cordon in
2013. Vines trained to High Wire Cordon (HWC), Geneva Double Curtain (GDC), and
Experimental Moving Trellis (MT).

TABLE 88. Wine tasting results of visual and aromatics analysis from 2013 HTRC with three
shoots per linear 30 centimeters of cordon.
Visual
Aromatics
Color
Color Aromatics Dark Vegetal Pepper Floral
Musty
Trellis
Intensity Hue
Intensity
Fruit
HWC
6.0 *
4.8 ab
4.9 a
4.3 a
2.0 b
2.9
3.6 a
2.3 b
GDC
6.1
5.0 a
4.5 a
3.7 ab
2.1 b
2.5
3.3 a
2.0 b
MT
5.8
4.3 b
3.6 b
2.9 b
3.5 a
2.4
2.2 b
3.6 a
*Treatment means followed by the same letter within a column are not significantly different at
the α=0.05 level. Columns where no letters are present indicate a lack of significant differences
among treatments.

166

TABLE 89. Wine tasting results of taste analysis from 2013 HTRC with three shoots per linear
30 centimeters of cordon.
Taste
Body
Trellis Sweetness Alcohol Acidity Astringency
HWC
2.3 *
4.2 a
3.2 ab
3.6
4.6 a
GDC
2.1
3.9 ab
2.8 b
3.3
4.3 a
MT
2.4
3.6 b
3.3 a
3.8
3.5 b
*Treatment means followed by the same letter within a column are not significantly different at
the α=0.05 level. Columns where no letters are present indicate a lack of significant differences
among treatments.

167

Body

Color Intensity
7

Color Hue

6
5

Astringency

Aromatics Intensity

4
3
2

Acidity

Dark Fruit

HWC
GDC

1

MT
Alcohol

Vegetal

Sweetness

Pepper
Musty

Floral

FIGURE 47. Results from comparing wines produced from fruit collected from vines used in
various trellis systems with a shoot density of six shoots per 30 centimeters of linear cordon in
2013. Vines trained to High Wire Cordon (HWC), Geneva Double Curtain (GDC), and
Experimental Moving Trellis (MT).

TABLE 90. Wine tasting results of visual and aromatics analysis from 2013 HTRC with six
shoots per linear 30 centimeters of cordon.
Visual
Aromatics
Color
Color Aromatics Dark Vegetal Pepper Floral
Musty
Trellis
Intensity
Hue
Intensity
Fruit
HWC
5.8 *
4.0 ab
4.9 a
4.8 a
2.1
2.6
3.2
2.1 b
GDC
5.7
4.5 a
4.5 ab
4.3 ab
1.9
2.0
2.8
2.6 ab
MT
5.6
3.8 b
4.1 b
3.4 b
2.8
2.4
2.9
3.3 a
*Treatment means followed by the same letter within a column are not significantly different at
the α=0.05 level. Columns where no letters are present indicate a lack of significant differences
among treatments.

168

TABLE 91. Wine tasting results of taste analysis from 2013 HTRC with six shoots per linear 30
centimeters of cordon.
Taste
Astringency
Body
Trellis Sweetness Alcohol Acidity
HWC
2.2 *
3.9
3.2
4.3
4.1 ab
GDC
2.3
3.8
3.2
4.1
4.3 a
MT
2.3
3.4
3.3
4.1
3.5 b
*Treatment means followed by the same letter within a column are not significantly different at
the α=0.05 level. Columns where no letters are present indicate a lack of significant differences
among treatments.

169

Discussion
The impacts of treatments in the 2012 frost experiment have shown that the trellis system
does have an impact on the wine quality produced by fruit from specific training systems. This
relationship between the trellis system and the microclimate that the cluster is ripening in during
the growing season has been studied by other researchers and is still being looked at today. A
notable result of the wine sensory analysis was that the wine that is produced from the secondary
fruit is of a quality that is not inferior to the wine produced from the primary fruit. This
information is particularly important for a cultivar such as Marquette that tends to have an early
bud break compared to other varieties. With this early bud break, it would not be hard to expect
damage to the buds to occur often over the lifespan of the vineyard. Frost events, neither
classified as good nor bad, cause changes in how the vines are managed during the growing
season and how the fruit will be processed to achieve wines of desired quality.
The amount of exposure of the cluster to sunlight and increased temperature has been
shown as the most important aspect affecting wine quality from these experiments. As seen by
the various spider graphs, each wine produced varies depending on the trellis system, canopy
density, fruit type or a combination of all three. The results from this data cannot suggest what
the ideal combination is due to the fact that both growing years (2012 and 2013) were very
different and each year could produce different results. Inherently, the trellis system that allows
for the most interception of sunlight to the fruiting zone will cause more dark fruit aromatics,
more intense color, more body, less astringency and less acidic wines.

170

Conclusion
The wine produced from the different treatments in 2012 and 2013 have brought new
information on the cultivar ‘Marquette’. The experiments conducted in 2012 on the effects of
frost and trellis system were the first of their kind for the SCH varieties. After a thorough search
of the literature there is no known information in a peer-reviewed journal on the aspects of wine
quality made from a primary cluster verses a secondary cluster. The information obtained from
the tasting has laid a foundation of scientific testing and basic knowledge of the SCH cultivar
‘Marquette’. Similar work will need to be conducted on the other SCH that have already been
released and that will be released in the future. Overall, the cultivar ‘Marquette’ shows promise
as a variety that can produce high yields, commercially acceptable quality of wine, and an
expansion to the areas in Michigan that can have sustainable grape production.

171

APPENDICES

172

APPENDIX A: TYPE 3 TABLES OF ANVOA FOR HTRC IN 2012

TABLE 92. Analysis of variance (ANOVA) for Sweetness from wine produced from fruit on
HWC collected in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
11
11.00 0.0069
Info
TABLE 93. Analysis of variance (ANOVA) for Color Intensity from wine produced from fruit
on GDC collected in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
11
7.86
0.0172
Info
TABLE 94. Analysis of variance (ANOVA) for Color Hue from wine produced from fruit on
GDC collected in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
11
6.06
0.0316
Info
TABLE 95. Analysis of variance (ANOVA) for Alcohol from wine produced from fruit on GDC
collected in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
11
12.57 0.0046
Info
TABLE 96. Analysis of variance (ANOVA) for Acidity from wine produced from fruit on GDC
collected in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
11
16.84 0.0017
Info

173

TABLE 97. Analysis of variance (ANOVA) for Astringency from wine produced from fruit on
GDC collected in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
11
6.49
0.0271
Info
TABLE 98. Analysis of variance (ANOVA) for Body from wine produced from fruit on GDC
collected in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
22
12.04 0.0022
Info
TABLE 99. Analysis of variance (ANOVA) for Color Hue from wine produced from fruit on
MT collected in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
11
6.81
0.0243
Info
TABLE 100. Analysis of variance (ANOVA) for Dark Fruit from wine produced from fruit on
MT collected in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
11
6.49
0.0271
Info
TABLE 101. Analysis of variance (ANOVA) for Floral from wine produced from fruit collected
on MT in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
11
16.50 0.0019
Info
TABLE 102. Analysis of variance (ANOVA) for Acidity from wine produced from fruit on MT
collected in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
11
6.49
0.0271
Info

174

TABLE 103. Analysis of variance (ANOVA) for Astringency from wine produced from fruit on
MT collected in 2012 at HTRC.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
1
11
8.19
0.0155
Info

175

APPENDIX B: TYPE 3 TABLES OF ANVOA FOR HTRC IN 2013

TABLE 104. Analysis of variance (ANOVA) for Color Hue from wine produced from fruit
collected in 2013 at three shoots per 30 linear centimeters of cordon.
Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
22
2.43
0.1112
Trellis
TABLE 105. Analysis of variance (ANOVA) for Aromatics Intensity from wine produced from
fruit collected in 2013 at three shoots per 30 linear centimeters of cordon.
Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
19.1
3.48
0.0516
Trellis
TABLE 106. Analysis of variance (ANOVA) for Dark Fruit from wine produced from fruit
collected in 2013 at three shoots per 30 linear centimeters of cordon.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
22
5.86
0.0091
Trellis
TABLE 107. Analysis of variance (ANOVA) for Vegetal from wine produced from fruit
collected in 2013 at three shoots per 30 linear centimeters of cordon.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
22
7.52
0.0032
Trellis
TABLE 108. Analysis of variance (ANOVA) for Floral from wine produced from fruit collected
in 2013 at three shoots per 30 linear centimeters of cordon.
Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
33
6.81
0.0033
Trellis

176

TABLE 109. Analysis of variance (ANOVA) for Musty from wine produced from fruit collected
in 2013 at three shoots per 30 linear centimeters of cordon.
Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
20
10.30 0.0008
Trellis
TABLE 110. Analysis of variance (ANOVA) for Alcohol from wine produced from fruit
collected in 2013 at three shoots per 30 linear centimeters of cordon.
Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
22
2.00
0.1585
Trellis
TABLE 111. Analysis of variance (ANOVA) for Acidity from wine produced from fruit
collected in 2013 at three shoots per 30 linear centimeters of cordon.
Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
22
1.75
0.1971
Trellis
TABLE 112. Analysis of variance (ANOVA) for Body from wine produced from fruit collected
in 2013 at three shoots per 30 linear centimeters of cordon.
Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
20.1
4.62
0.0224
Trellis
TABLE 113. Analysis of variance (ANOVA) for Color Hue from wine produced from fruit
collected in 2013 at six shoots per 30 linear centimeters of cordon.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
22
3.08
0.0662
Trellis
TABLE 114. Analysis of variance (ANOVA) for Aromatics Intensity from wine produced from
fruit collected in 2013 at six shoots per 30 linear centimeters of cordon.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
22
2.04
0.1543
Trellis

177

TABLE 115. Analysis of variance (ANOVA) for Dark Fruit from wine produced from fruit
collected in 2013 at six shoots per 30 linear centimeters of cordon.
Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
22
4.31
0.0263
Trellis
TABLE 116. Analysis of variance (ANOVA) for Musty from wine produced from fruit collected
in 2013 at six shoots per 30 linear centimeters of cordon.
Type 3 Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
22
2.22
0.1329
Trellis
TABLE 117. Analysis of variance (ANOVA) for Body from wine produced from fruit collected
in 2013 at six shoots per 30 linear centimeters of cordon.
Type III Tests of Fixed Effects
Effect Num DF Den DF F Value Pr > F
2
22
1.95
0.1654
Trellis

178

LITERATURE CITED

179

LITERATURE CITED

Alleweldt, G., and Possingham, J. V. (1988). Progress in grapevine breeding. Theoretical and
Applied Genetics, 75, 669-673.
Bavougian, C. M., Read, P. E., and Walter-Shea, E. (2012). Training system effects on sunlight
penetration, canopy structure, yield, and fruit characteristics of 'Frontenac' grapevine (Vitis spp.).
International Journal of Fruit Science, 12, 402-409.
Bergqvist, J., Dokoozlian, N., and Ebisuda, N. (2001). Sunlight exposure and temperature effects
on berry growth and composition of Cabernet Sauvignon and Grenache in the Central San
Joaquin Valley of California. American Journal of Enology and Viticulture, 52(1), 1-7.
Bordelon, B. P., Skinkis, P. A., and Howard, P. H. (2008). Impact of training system on vine
performance and fruit composition of Traminette. American Journal of Enology and Viticulture,
59(1), 39-46.
Boulton, R. B., Singleton, V. L., Bisson, L. F., and Kunkee, R. E. (1996). Principles and
practices of winemaking. New York: Springer.
Dokoozlian, N. K., and Kliewer, W. M. (1995). The light environment within grapevine
canopies. I. Description and seasonal changes during fruit development. American Journal of
Enology and Viticulture, 46(2), 209-218.
Dokoozlian, N. K., and Kliewer, W. M. (1995). The light environment within grapevine
canopies. II. Influence of leaf area density on fruit zone light environment and some canopy
assessment parameters. American Journal of Enology and Viticulture, 46(2), 219-226.
I. Etaio, M. Albisu, M. Ojeda, P.F. Gil, J. Salmerón, F.J. Pérez Elortondo, (2010) Sensory
quality control for food certification: A case study on wine. Method development, Food Control,
21 (4), 533-541.
Gladstone, E. A., and Dokoozlian, N. K. (2003). Influence of leaf area density and trellis/training
system on the light microclimate within grapevine canopies. Vitis, 42(3), 123-131.
Hathaway, L., and Kegerreis, S. (2010). The History of Michigan Wines: 150 Years of
Winemaking along the Great Lakes. The History Press.
Hemstad, P. R., and Luby, J. J. (2000). Utilization of Vitis riparia for the development of new
wine varieties with resistance to disease and extreme cold. International Symposium on
Grapevine Genetics and Breeding (pp. 487-490). Acta Horticulturae.
180

Howell, G. S., Miller, D. P., Edson, C. E., and Striegler, R. K., (1991). Influence of training
system and pruning severity on yield, vine size, and fruit composition of Vignoles grapevines.
American Journal of Enology and Viticulture, 42(3), 191-198.
Iland, P., Bruer, N., Edwards, G., Weeks, S., & Wilkes, E. (2004). Chemical analysis of grapes
and wine:techniques and concepts. Campbelltown, Australia: Patrick Iland Wine Promotions.
Jackson, R. S. (1994). Wine Science: Principles and Applications. San Diego, CA: Academic
Press.
Keller, M. (2010). The science of grapevines: anatomy and physiology. London: Elsevier.
Liu, H.-F., Wu, B.-H., Fan, P.-G., Li, S.-H., and Li, L.-S. (2006). Sugar and acid concentrations
in 98 grape cultivars analyzed by principal component analysis. Journal of the Science of Food
and Agriculture, 1526-1536.
Luby, J. J. (1991). Breeding cold-hardy fruit crops in Minnesota. HortScience, 26(5), 507-512.
Mansfield, A. K., and Vickers, Z. M. (2009). Characterization of the aroma of red Frontenac
table wines by descriptive analysis. American Journal of Enology and Viticulture, 60(4), 435441.
Marino, W. D. (2012). Spring 2012 Climate Summary for Southwest Lower Michigan. Grand
Rapids: National Weather Service.
Mills, L. J., Ferguson, J. C., and Keller, M. (2006). Cold-Hardiness Evaluation of Grapevine
Buds and Cane Tissues. American Journal of Enology and Viticulture, 57(2), 194-201.
Palliotti, A. (2011,). A New closing Y-shaped training system for grapevines. Australian Journal
of Grape and Wine Research, 18(1), 53-67.
Perry, R., Sabbatini, P., and Burns, J. (2012, March). Growing Wine Grapes in Michigan.
Retrieved from www.michiganwines.com/docs/Industry/wine_grape_suitability.pdf
Ravaz, M.L. 1911. L’effeuillage de la vigne. Ann. l’École Nat.
Agric. Montpellier. 11:216-244
Reynolds, A. G., Pool, R. M., and Mattick, L. R. (1986). Influence of cluster exposure on fruit
composition and wine quality of Seyval blanc grapes. Vitis, 85-95.
Reynolds, A. G., and Vanden Heuvel, J. E. (2009). Influence of grapevine training systems on
vine growth and fruit composition: A review. American Journal of Enology and Viticulture,
60(3), 251-268.
Reynolds, A. G., Wardle, D. A., and Naylor, A. P. (1995). Impact of training system and vine
spacing on vine performance and berry composition of Chancellor. American Journal of Enology
and Viticulture, 46(1), 88-98.
181

Reynolds, A. G., Wardle, D. A., and Dever, M. (1996). Vine performance, fruit composition, and
wine sensory attributes of Gewürztraminer in response to vineyard location and canopy
manipulation. American Journal of Enology and Viticulture, 47(1), 77-92.
Shaulis, N., Shepardson, E. S., and Jordan, T. D. (1966) The Geneva double curtain for concord
grapes. Geneva, New York: New York State Agricultural Experiment Station.
Smart, R. E. (1973). Sunlight interception by vineyards. American Journal of Enology and
Viticulture, 24(4), 141-147.
Smart, R. E. (1985). Principles of grapevine canopy microclimate manipulation with
implications for yield and quality: A review. American Journal of Enology and Viticulture,
36(3), 230-239.
Smart, R. E. (1988). Shoot spacing and canopy light microclimate. American Journal of Enology
and Viticulture, 39(4), 325-333.
Smart, R. R. (2008). Sunlight into wine. Winetitles.
Smiley, L. (2008). Iowa State University. Retrieved from Marquette:
viticulture.hort.iastate.edu/cultivars/Marquette.pdf
Stergios, B. G., and Howell, G. S. (1977). Effect of site on cold acclimation and deacclimation of
Concord grapevines. American Journal of Enology and Viticulture, 28(1), 43-48.
Stergios, B. G., and Howell, G. S. (1977). Effects of defoliation, trellis height, and cropping
stress on the cold hardiness of Concord grapevines. American Journal of Enology and
Viticulture, 28(1), 34-43.
Takeda, F., Hummell, A.K., and Peterson D. L. (2003) Primocane growth in ‘Chester Thornless’
blackberry trained to the rotatable cross-arm trellis. HortScience, 38(3), 373-376.
Tuck, B., and Gartner, W. (2012). Vineyards and Wineries in Michigan: A status and economic
contribution report with focus on Michigan wine grapes. University of Minnesota. University of
Minnesota Extension.
University of Minnesota. (2012, September 14). Grapes. Retrieved from http://grapes.umn.edu/
USDA. (2012). Fruit Rotational Survey 2011-2012. Retrieved April 2013, from
http://www.nass.usda.gov/Statistics_by_State/Michigan/Publications/Michigan_Rotational_Surv
eys/mi_fruit12.html

182

Zoecklein, B. W., Wolf, T. K., Pelanne, L., Miller, M. K., and Birkenmaier, S. S. (2008). Effect
of vertical shoot-positioned, Smart-Dyson, and Geneva Double-Curtain training systems on
Viognier grape and wine composition. American Journal of Enology and Viticulture, 59(1), 1115.

183