<%BANNER%>

Yield, Nutritive Value, and Persistence Responses of Bahiagrass Genotypes to Extended Daylength and Defoliation Management

Permanent Link: http://ufdc.ufl.edu/UFE0022750/00001

Material Information

Title: Yield, Nutritive Value, and Persistence Responses of Bahiagrass Genotypes to Extended Daylength and Defoliation Management
Physical Description: 1 online resource (216 p.)
Language: english
Creator: Interrante, Sindy
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: bahiagrass, daylength, defoliation, nutritive, persistence, photoperiod, tiller, tnc, value, yield
Agronomy -- Dissertations, Academic -- UF
Genre: Agronomy thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Existing bahiagrass (Paspalum notatum Fl?gge) cultivars are daylength-sensitive and have minimal cool-season production, resulting in high winter feeding costs in forage-based livestock systems. A new genotype is less daylength-sensitive and possesses greater cold tolerance, but its response under a range of defoliation treatments is unknown. Two experiments were conducted to study the effects of extended daylength and defoliation management on photoperiod-sensitive (diploids ?Pensacola? and ?Tifton 9? and tetraploids ?Argentine? and Tifton 7) and less photoperiod sensitive (PCA) Cycle 4 bahiagrass genotypes. Experiment 1 was a 2-yr field study in pots evaluating Pensacola, Tifton 9, and PCA under two daylengths (ambient and extended to 15 h) and two fertilization levels (low and high). Experiment 2 was a 2-yr field-plot study comparing Pensacola, Tifton 9, Argentine, Tifton 7, and PCA defoliated to 4- and 8-cm stubble heights every 7 and 21 d. In Experiment 1, PCA tended to have lower below-:above-ground ratio and less TNC content than the other genotypes. The morphological characteristics and herbage yield of PCA are more similar to those of upright-growing Tifton 9 than the more decumbent Pensacola, and PCA is likely less tolerant of defoliation than Pensacola. In Experiment 2, PCA had lower annual herbage yield than Argentine and Tifton 7 and lower or equal yield to Tifton 9 and Pensacola. Nutritive value was as great or greater for PCA than the other genotypes. Percent cover was low for PCA except when defoliated every 21 d to 8 cm. Argentine and Pensacola had an average of 98% more root + rhizome TNC than Tifton 7, Tifton 9, and PCA after 2 yr. From these data, it can be concluded that defoliation management of PCA will likely be more critical to its persistence than for Pensacola and Argentine in Florida. The erect growth habit of PCA, its preferential allocation of DM to above-ground growth, the lesser stored TNC for regrowth, and greater reduction in ground cover accompanying stress-inducing defoliation treatments, imply that less frequent and intense defoliation will be required to sustain stands of PCA than is the case for commonly used Pensacola or Argentine.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sindy Interrante.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Sollenberger, Lynn E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022750:00001

Permanent Link: http://ufdc.ufl.edu/UFE0022750/00001

Material Information

Title: Yield, Nutritive Value, and Persistence Responses of Bahiagrass Genotypes to Extended Daylength and Defoliation Management
Physical Description: 1 online resource (216 p.)
Language: english
Creator: Interrante, Sindy
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: bahiagrass, daylength, defoliation, nutritive, persistence, photoperiod, tiller, tnc, value, yield
Agronomy -- Dissertations, Academic -- UF
Genre: Agronomy thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Existing bahiagrass (Paspalum notatum Fl?gge) cultivars are daylength-sensitive and have minimal cool-season production, resulting in high winter feeding costs in forage-based livestock systems. A new genotype is less daylength-sensitive and possesses greater cold tolerance, but its response under a range of defoliation treatments is unknown. Two experiments were conducted to study the effects of extended daylength and defoliation management on photoperiod-sensitive (diploids ?Pensacola? and ?Tifton 9? and tetraploids ?Argentine? and Tifton 7) and less photoperiod sensitive (PCA) Cycle 4 bahiagrass genotypes. Experiment 1 was a 2-yr field study in pots evaluating Pensacola, Tifton 9, and PCA under two daylengths (ambient and extended to 15 h) and two fertilization levels (low and high). Experiment 2 was a 2-yr field-plot study comparing Pensacola, Tifton 9, Argentine, Tifton 7, and PCA defoliated to 4- and 8-cm stubble heights every 7 and 21 d. In Experiment 1, PCA tended to have lower below-:above-ground ratio and less TNC content than the other genotypes. The morphological characteristics and herbage yield of PCA are more similar to those of upright-growing Tifton 9 than the more decumbent Pensacola, and PCA is likely less tolerant of defoliation than Pensacola. In Experiment 2, PCA had lower annual herbage yield than Argentine and Tifton 7 and lower or equal yield to Tifton 9 and Pensacola. Nutritive value was as great or greater for PCA than the other genotypes. Percent cover was low for PCA except when defoliated every 21 d to 8 cm. Argentine and Pensacola had an average of 98% more root + rhizome TNC than Tifton 7, Tifton 9, and PCA after 2 yr. From these data, it can be concluded that defoliation management of PCA will likely be more critical to its persistence than for Pensacola and Argentine in Florida. The erect growth habit of PCA, its preferential allocation of DM to above-ground growth, the lesser stored TNC for regrowth, and greater reduction in ground cover accompanying stress-inducing defoliation treatments, imply that less frequent and intense defoliation will be required to sustain stands of PCA than is the case for commonly used Pensacola or Argentine.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Sindy Interrante.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Sollenberger, Lynn E.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0022750:00001


This item has the following downloads:


Full Text

PAGE 1

1 YIELD, NUTRITIVE VALUE, AND PERS ISTENCE RESPONSES OF BAHIAGRASS GENOTYPES TO EXTENDED DAYLENGT H AND DEFOLIATION MANAGEMENT By SINDY MARIE INTERRANTE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Sindy Marie Interrante

PAGE 3

3 To my parents, for all of their suppor t and encouragement through the years.

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank m y parents for thei r continued support and encouragement through the years. I would also like to thank all of my friends for their patience and encouragement, especially Jamie Foster, Becky Williams, and Ke lly and Mort Vineyard. Special thanks go to Summer Houghton for her support, patience, and understanding. Special thanks go to Dr. Lynn E. Sollenberg er, my supervisory committee chair. His guidance has been invaluable and greatly apprec iated. Thanks also go to my other supervisory committee members, Dr. Ann Blount, Dr. Tom Si nclair, Dr. Ken Boote, Dr. Rao Mylavarapu, and Dr. Carroll Chambliss (deceased), for their willingness to serve on the committee, their input and direction during the program, a nd for reviewing the dissertation. Particular thanks go to those who helped during field and labora tory activities. This includes fellow graduate students Jose Dubeux, Joao Vendramini, Kesi Liu, and Renee White. Thanks also go to Dwight Thomas and Sid Jones for their help a nd support at the Beef Research Unit, to Richard Fethiere for his help in the Forage Evaluation Support Laboratory, and to Dr. Sam Coleman and Ed Bowers for their assistan ce at the SubTropical Agricultural Research Station.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........9 LIST OF FIGURES.......................................................................................................................14 ABSTRACT...................................................................................................................................16 CHAP TER 1 INTRODUCTION..................................................................................................................18 2 LITERATURE REVIEW.......................................................................................................22 Introduction of Bahiagrass to Florida and Uses in Agriculture..............................................22 Production Characteristics of Bahiagrass............................................................................... 22 Environmental Adaptation............................................................................................... 22 Yield and Yield Distribution...........................................................................................23 Yield Responses to Fertilization...................................................................................... 24 Tolerance to Defoliation.................................................................................................. 25 Forage Nutritive Value and Quality................................................................................ 27 Limitations of Bahiagrass...................................................................................................... .29 Harvested Forage Yields................................................................................................. 29 Cool-Season Growth.......................................................................................................30 Recent Plant Breeding Advances............................................................................................ 31 Daylength Effects on Forage Plants.......................................................................................32 Forage Plant Responses to Defoliation a nd Mechanism s of Grazing Tolerance.................... 37 Summary and Project Objectives............................................................................................ 40 3 BAHIAGRASS GROWTH, PHOTOS YNTHESIS, AND MORPHOL OGY RESPONSES TO DAYLENGTH IN ESTABLISHMENT YEAR....................................... 42 Introduction................................................................................................................... ..........42 Materials and Methods...........................................................................................................44 Experimental Site............................................................................................................ 44 Treatments and Design.................................................................................................... 45 Response Variables.........................................................................................................47 Results and Discussion......................................................................................................... ..49 Rainfall............................................................................................................................49 Total-Season Herbage Yield............................................................................................ 49 Seasonal Herbage Yield..................................................................................................53 Herbage Nitrogen............................................................................................................54 Tiller Number..................................................................................................................57 Leaf Area.........................................................................................................................58

PAGE 6

6 Plant Height and Diameter.............................................................................................. 60 Leaf Photosynthesis......................................................................................................... 65 Summary and Conclusions.....................................................................................................66 4 BAHIAGRASS COMPONENT MASS, NITROGEN, AND TOTAL NONSTRUCTUR AL CARBOHYDRATE RESPONSES TO DAYLENGTH IN ESTABLISHMENT YEAR.................................................................................................... 70 Introduction................................................................................................................... ..........70 Materials and Methods...........................................................................................................72 Experimental Site............................................................................................................ 72 Treatments and Design.................................................................................................... 72 Response Variables.........................................................................................................75 Results and Discussion......................................................................................................... ..77 Rainfall............................................................................................................................77 Allocation of Dry Matter to Plant Parts...........................................................................78 Above-ground component mass...............................................................................78 Below-ground component mass...............................................................................82 Below-:Above-ground Ratio........................................................................................... 84 Reserve Pools..................................................................................................................86 Nitrogen concentration............................................................................................. 86 Total nonstructural carb ohydrate concentration .......................................................87 Nitrogen content.......................................................................................................90 Total nonstructural ca rbohydrate content ................................................................. 93 Summary and Conclusions.....................................................................................................94 5 YIELD, NUTRITIVE VALUE, AND PERSISTENCE RESPONSES OF BAHIAGRASS GENOTYPES TO DEFOLIATION MANAGEMENT ............................... 97 Introduction................................................................................................................... ..........97 Materials and Methods.........................................................................................................101 Experimental Site.......................................................................................................... 101 Treatments and Design.................................................................................................. 101 Response Variables....................................................................................................... 103 Results and Discussion......................................................................................................... 105 Rainfall..........................................................................................................................105 Total-Season Herbage Yield.......................................................................................... 105 Seasonal Herbage Yield................................................................................................109 Cool-Season Herbage Yield..........................................................................................115 Nutritive Value..............................................................................................................117 Weighted total-season crude protein concentration ............................................... 117 Seasonal crude protein concentration..................................................................... 119 Weighted total-season in vitro di gestible organic m atter yield.............................. 124 Seasonal in vitro digest ible organic m atter............................................................ 127 Cover.............................................................................................................................132

PAGE 7

7 Plant Component Mass.................................................................................................. 133 Reserve Pools................................................................................................................137 Nitrogen concentration........................................................................................... 138 Total nonstructural carb ohydrate concentration ..................................................... 138 Nitrogen content.....................................................................................................139 Total nonstructural ca rbohydrate content ............................................................... 141 Summary and Conclusions...................................................................................................143 6 TILLER RESPONSES OF BAHIAGRASS GENOTYPES TO DEFOLIATION MANAGEMENT .................................................................................................................146 Introduction................................................................................................................... ........146 Materials and Methods.........................................................................................................148 Experimental Site.......................................................................................................... 148 Treatments and Design.................................................................................................. 148 Response Variables....................................................................................................... 149 Results and Discussion......................................................................................................... 151 Tiller Number and Tiller Number Change.................................................................... 151 Tiller number.......................................................................................................... 151 Tiller number change............................................................................................. 153 Tiller Mass.................................................................................................................... .155 Tiller Dynamics.............................................................................................................157 Tiller appearance rate............................................................................................. 157 Tiller death rate...................................................................................................... 162 Net tiller appearance rate........................................................................................163 Summary and Conclusions...................................................................................................164 7 SUMMARY AND CONCLUSIONS...................................................................................166 Extended Daylength and Fertilizati on Experiment (Chapters 3 and 4) ................................ 167 Genotype Responses......................................................................................................167 Daylength Responses..................................................................................................... 168 Seasonal Responses.......................................................................................................168 Defoliation Management Res ponses (Chapters 5 and 6) ...................................................... 169 Herbage Yield and Nutritive Value............................................................................... 169 Persistence Responses................................................................................................... 170 Tiller Responses............................................................................................................ 170 Implications of the Research................................................................................................ 171 APPENDIX A SOURCES OF VARIATION: CHAPTER 3........................................................................ 173 B SOURCES OF VARIATION: CHAPTER 4........................................................................ 180 C SOURCES OF VARIATION: CHAPTER 5........................................................................ 187 D SOURCES OF VARIATION: CHAPTER 6........................................................................ 199

PAGE 8

8 LIST OF REFERENCES.............................................................................................................206 BIOGRAPHICAL SKETCH.......................................................................................................216

PAGE 9

9 LIST OF TABLES Table page 3-1 Bahiagrass herbage weighted total-season and seasonal herbage N concentrations as affected by genotype. .........................................................................................................55 3-2 Bahiagrass leaf area as affected by daylength X evaluation date interaction. ................... 59 3-3 Bahiagrass leaf area as affect ed by genotype in 2004-05 and 2005-06. ............................60 3-4 Bahiagrass leaf area as affect ed by daylength in 2004-05 and 2005-06. ........................... 61 3-5 Bahiagrass plant height as affected by daylength X evaluation date interaction and plant diam eter as affected by genot ype X evaluation date interaction.............................. 61 3-6 Bahiagrass plant height as aff ected by genotype in 2004-05 and 2005-06........................62 3-7 Bahiagrass height:diameter ratio as aff ected by genotype in Nove mber, April, and June 2004-05 and 2005-06.................................................................................................64 3-8 Bahiagrass height:diameter ratio as aff ected by daylength in Nove mber, April, and June 2004-05 and 2005-06.................................................................................................65 4-1 Bahiagrass leaf mass as affected by da ylength X evaluation date interaction................... 79 4-2 Bahiagrass component mass as aff ected by genotype in 2004-05 and 2005-06. ............... 81 4-3 Bahiagrass root mass as affected by da ylength X evaluation date interaction.................. 83 4-4 Bahiagrass below-:above-ground ratio as affected by genotype in 2004-05 and 200506........................................................................................................................................85 4-5 Bahiagrass below-:above-ground ratio as affected by daylength X evaluation date interaction. .........................................................................................................................86 4-6 Bahiagrass rhizome + stem base total nonstructural carbohydrate concentration as affected by daylength X eval uation date interaction.......................................................... 88 4-7 Bahiagrass rhizome + stem base total nonstructural carbohydrate concentration as affected by genotype X evaluation date interaction. ..........................................................90 4-8 Bahiagrass root and rhizome + stem base nitrogen and to ta l nonstructural carbohydrate (TNC) content as affected by genotype....................................................... 92 4-9 Bahiagrass root nitrogen content as affected by daylength X evaluation date interaction. .........................................................................................................................92

PAGE 10

10 4-10 Bahiagrass rhizome + stem base total nons tructural carbohydrate content as affected by daylength X evaluation date interaction. ...................................................................... 94 5-1 Bahiagrass total-season herbage yield as affected by genotypes X harvest frequency interaction. .......................................................................................................................107 5-2 Bahiagrass total-season herbage yield as affected by genotype in 2005, 2006, and 2007..................................................................................................................................108 5-3 Bahiagrass cool-season herbage yield as affected by genot ype X stubble height interaction. .......................................................................................................................116 5-4 Bahiagrass cool-season herbage yield as affected by genotype X harvest frequency interaction. .......................................................................................................................117 5-5 Bahiagrass weighted total-season crude protein concentration as affected by genotype X harvest frequency interaction. ...................................................................... 118 5-6 Bahiagrass weighted total-season crude pr otein concentration as affected by stubble height X harvest frequency interaction. ...........................................................................119 5-7 Bahiagrass weighted total-season in vitro digestible organic m a tter concentration as affected by genotype X harvest frequency interaction.................................................... 126 5-8 Bahiagrass weighted total-season in vitro digestible organic m a tter concentration as affected by genotype in 2005 and 2006........................................................................... 126 5-9 Bahiagrass cover as affected by genot ype X stubble height X harvest frequency interaction in October. ..................................................................................................... 132 5-10 Bahiagrass root + rhizome mass as affected by genotype X stubble height interaction in October 2006................................................................................................................134 5-11 Bahiagrass root + rhizome mass as a ffected by genotype X harvest frequency interaction in October 2006. ............................................................................................135 5-12 Bahiagrass stem base mass as affected by genotype X harvest frequency interaction in October 2006................................................................................................................135 5-13 Bahiagrass root + rhizome and stem base m ass as affected by genotype in October 2007..................................................................................................................................136 5-14 Bahiagrass root + rhizome total nonstructu ral carbohydrate concentr ation as affected by genotype X harvest frequency interaction in O ctober 2006....................................... 139 5-15 Bahiagrass root + rhizome and stem base nitrogen content as affected by genotype X harvest frequency interactions in October 2006. ............................................................. 140

PAGE 11

11 5-16 Bahiagrass root + rhizome nitrogen content as affected by genotype X stubble height interaction in October 2006. ............................................................................................141 5-17 Bahiagrass root + rhizome total nonstruc tural carbohydrate content as affected by genotype X harvest frequency interaction in October 2006. ........................................... 142 6-1 Bahiagrass tiller number as affected by genotype X stubble height X harvest frequency interaction. ...................................................................................................... 152 6-2 Bahiagrass tiller number change as aff ected by genotype during the first year of defoliation (2005) and from the beginning of the first year of defoliation (Spring 2005) to the end of the second year (Fall 2006).............................................................. 154 6-3 Bahiagrass tiller number change as a ffected by stubble height X harvest frequency interaction from the beginning of the firs t year of defoliation (Spring 2005) to the end of the second year (Fall 2006)................................................................................... 155 6-4 Bahiagrass tiller mass as affected by genotype X stubble height interactions on 27 July and 27 Sept. 2005. .................................................................................................... 157 6-5 Bahiagrass tiller appearance rate as a ffected by genotype X stubble height X harvest frequency interaction on 29 May 2006. ........................................................................... 160 6-6 Bahiagrass tiller appearance rate as affected by genotype X harvest frequency interaction on 25 July and genotype x stubble height interaction on 21 Aug. 2006. .......161 A-1 Sources of variation for bahi agrass total-season herbage yield. ...................................... 173 A-2 Sources of variation for bahiagrass total-season herbage yield by year .......................... 173 A-3 Sources of variation for ba hiagrass seasonal herbage yield............................................. 173 A-4 Sources of variation for bahiagrass seasonal herbage yield by year. ............................... 174 A-5 Sources of variation for bahiagrass weighted total-seas on nitrogen (N) concentration. 174 A-6 Sources of variation for bahiagrass wei ghted seasonal nitrogen (N) concentration. ....... 175 A-7 Sources of variation fo r bahiagrass tiller num ber............................................................ 175 A-8 Sources of variation for bahiagrass leaf area................................................................... 176 A-9 Sources of variation for bahiagrass leaf area by year. ..................................................... 176 A-10 Sources of variation fo r bahiagrass plant height. ............................................................. 176 A-11 Sources of variation for bahi agrass plant height by year. ................................................ 177 A-12 Sources of variation for bahiagrass plant diam eter.......................................................... 177

PAGE 12

12 A-13 Sources of variation for bahiagrass height:diam eter ratio............................................... 178 A-14 Sources of variation for bahiag rass height:diam eter ratio by year.................................. 178 A-15 Sources of variation for bahiagrass leaf photosynthesis. ................................................. 179 B-1 Sources of variation for ba hiagrass plant com ponent mass............................................. 180 B-2 Sources of variation for bahiagrass leaf m ass by year..................................................... 181 B-3 Sources of variation for bahi agrass stem base mass by year........................................... 181 B-4 Sources of variation for bahi agrass inflorescence m ass by year...................................... 181 B-5 Sources of variation for ba hiagrass below-:above-ground ratio. .....................................182 B-6 Sources of variation for bahiag rass below-:above-ground ratio by year. ........................182 B-7 Sources of variation for bahiagrass pl ant com ponent nitrogen (N) concentration.......... 183 B-8 Sources of variation for bahiagrass plan t com ponent total nonstructural carbohydrate (TNC) concentration........................................................................................................ 184 B-9 Sources of variation for bahiagrass rhizom e + stem base total nonstructural carbohydrate concentration (TNC) by year..................................................................... 185 B-10 Sources of variation for bahiagrass plant com ponent nitrogen (N) content.................... 185 B-11 Sources of variation for bahiagrass plan t com ponent total nonstructural carbohydrate (TNC) content..................................................................................................................186 C-1 Sources of variation for bahi agrass total-season herbage yield. ...................................... 187 C-2 Sources of variation for bahiagrass total-season herbage yield by year. ......................... 187 C-3 Sources of variation for ba hiagrass seasonal herbage yield............................................. 188 C-4 Sources of variation for bahiagrass seasonal herbage yield by year. ............................... 189 C-5 Sources of variation for bahi agrass cool-season herbage yield. ...................................... 190 C-6 Sources of variation for bahiagrass weighted total-seas on crude protein (CP) and in vitro digestible organi c m atter concentration.................................................................. 191 C-7 Sources of variation for bahiagrass weighted total-seas on crude protein (CP) and in vitro digestible organic m a tter concentration by year...................................................... 192 C-8 Sources of variation for bahiagrass he rbage crude protein (C P) concentration............... 192

PAGE 13

13 C-9 Sources of variation for bahiagrass herbage crude protein (C P) by year........................ 194 C-10 Sources of variation for bahiagrass he rbage in vitro diges tib le organic matter (IVDOM) concentration...................................................................................................194 C-11 Sources of variation for bahiagrass he rbage in vitro diges tib le organic matter (IVDOM) by year............................................................................................................ 196 C-12 Sources of variation for bahiagrass cover........................................................................ 196 C-13 Sources of variation for bahiagrass root + rhizom e and stem base mass in October 2007 and 2008..................................................................................................................197 C-14 Sources of variation for bahiagrass r oot + rhizom e and stem base nitrogen (N) concentration and cont ent in October 2007..................................................................... 198 C-15 Sources of variation for bahiagrass root + rhizom e and stem base total nonstructural carbohydrate (TNC) concentration a nd content in October 2007.................................... 198 D-1 Sources of variation fo r bahiagrass tiller num ber............................................................ 199 D-2 Sources of variation for ba hiagrass tiller num ber change................................................ 200 D-3 Sources of variation for bahiagrass tiller mass................................................................ 200 D-4 Sources of variation for bahiag rass tiller m ass by evaluation date.................................. 202 D-5 Sources of variation for bahiagrass tiller appearance rate (TA R) and tiller death rate (TDR)...............................................................................................................................203 D-6 Sources of variation for bahiagrass tille r appea rance rate (TAR), tiller death rate (TDR), and net TAR by year...........................................................................................204

PAGE 14

14 LIST OF FIGURES Figure page 3-1 Monthly rainfall data at the experimental site; av erage of 30-yr, 2004-05, and 200506........................................................................................................................................50 3-2 Bahiagrass total-season herbage yield as affected by genotype in 2004-05 and 200506........................................................................................................................................51 3-3 Bahiagrass total-season herbage yield as affected by daylength in 2004-05 and 200506........................................................................................................................................52 3-4 Bahiagrass herbage weighted seasonal herb age N concentration as affected by daylength X season interaction.......................................................................................... 57 4-1 Monthly rainfall data at the experimental site; av erage of 30-yr, 2004-05, and 200506........................................................................................................................................77 4-2 Bahiagrass rhizome + stem base total nonstructural carbohydrate concentration as affected by genotype in 2004-05 and 2005-06. ................................................................. 91 5-1 Monthly rainfall data at the experi mental site; average of 30-yr, 2005, 2006, 2007, and through April 2008. ................................................................................................... 106 5-2 Bahiagrass total-season herbage yield as affected by harvest f requency in 2005, 2006, and 2007.................................................................................................................109 5-3 Bahiagrass seasonal herbage yield as a ffected by genotype during evaluation periods in 2005. ............................................................................................................................111 5-4 Bahiagrass seasonal herbage yield as a ffected by genotype during evaluation periods in 2006. ............................................................................................................................113 5-5 Bahiagrass seasonal herbage yield as a ffected by genotype during evaluation periods in 2007. ............................................................................................................................114 5-6 Bahiagrass seasonal crude protein concentration as affected by genotype at evaluation periods in 2005. .............................................................................................. 121 5-7 Bahiagrass seasonal crude protein concentration as affected by genotype at evaluation periods for the 7d harvest frequency in 2006................................................ 122 5-8 Bahiagrass seasonal crude pr otein co ncentration as affect ed by genotype during eight evaluation periods at the 21-d harvest frequency in 2006...............................................125 5-9 Bahiagrass seasonal in vitro digestible organic matter concentr ation as affected by genotype during eight evaluation periods in 2005. ..........................................................129

PAGE 15

15 5-10 Bahiagrass seasonal in vitro digestible organic matter concentr ation as affected by genotype during eight evaluation periods in 2006. ..........................................................131 6-1 Bahiagrass tiller appearance rate as affected by evaluation date in 2005 and 2006. .......158 6-2 Bahiagrass tiller death rate as aff ected by evaluation date in 2005 and 2006. ................ 162 6-3 Bahiagrass net tiller appear ance rate as affected by evaluation date in 2005 and 2006. 163

PAGE 16

16 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy YIELD, NUTRITIVE VALUE, AND PERS ISTENCE RESPONSES OF BAHIAGRASS GENOTYPES TO EXTENDED DAYLENGT H AND DEFOLIATION MANAGEMENT By Sindy Marie Interrante December 2008 Chair: Lynn E. Sollenberger Major: Agronomy Existing bahiagrass (Paspalum notatum Flgge) cultivars are dayl ength-sensitive and have minimal cool-season production, resulting in high wi nter feeding costs in forage-based livestock systems. A new genotype is less daylength-sensitiv e and possesses greater cold tolerance, but its response under a range of defo liation treatments is unknown. Two experiments were conducted to study the effects of extended daylength a nd defoliation management on photoperiod-sensitive (diploids Pensacola and Tifton 9 and tetr aploids Argentine and Tifton 7) and less photoperiod sensitive (PCA) Cycle 4 bahiagrass ge notypes. Experiment 1 was a 2-yr field study in pots evaluating Pensacola, Tifton 9, and PCA under two daylengths (a mbient and extended to 15 h) and two fertilization levels (low and hi gh). Experiment 2 was a 2-yr field-plot study comparing Pensacola, Tifton 9, Argentine, Tifton 7, and PCA defoliated to 4and 8-cm stubble heights every 7 and 21 d. In Experiment 1, PCA tended to have lower below-:above-ground ratio and less TNC content than the other genotypes. The morphological characteristics and herbage yield of PCA are more similar to those of upright-growing Tifton 9 than the more decumbent Pensacola, and PCA is likely less tolerant of defoliation than Pe nsacola. In Experiment 2, PCA had lower annual herbage yield than Argentine a nd Tifton 7 and lower or equal yield to Tifton 9 and Pensacola. Nutritive value was as great or greater for PCA than the other genotypes. Percent

PAGE 17

17 cover was low for PCA except when defoliated ev ery 21 d to 8 cm. Argentine and Pensacola had an average of 98% more root + rhizome TNC than Tifton 7, Tifton 9, and PCA after 2 yr. From these data, it can be concluded that defoliation management of PCA will likely be more critical to its persistence than for Pensaco la and Argentine in Florida. The erect growth habit of PCA, its preferential allocation of DM to above-ground growth, the lesser stored TNC for regrowth, and greater reduction in ground cover accompanying stress-inducing defoliation treatments, imply that less freque nt and intense defoliation will be required to sustain stands of PCA than is the case for commonly used Pensacola or Argentine.

PAGE 18

18 CHAPTER 1 INTRODUCTION Bahiagrass ( Paspalum notatum Flgge) is the prim ary past ure grass for beef cattle ( Bos sp.) and horses ( Equus caballis ) in Florida and is grown on more than one million ha (Chambliss and Adjei, 2006). It is a warm-season perennial th at in Florida produces 85 to 90% of its total annual herbage yield during April through September (Mislevy a nd Everett, 1981; Kalmbacher, 1997). This seasonal growth response results in c ool-season forage deficits for livestock which necessitate feeding significant quantities of costly supplements. Limited bahiagrass herbage accumulation during the cool season has occurred even when temperature, soil moisture, and soil fertility we re adequate for substa ntially greater growth (Sinclair et al., 1997). This led to the conclusion that the reducti on in bahiagrass growth may be attributed at least in part to short daylength (Sinclair et al., 1997, 2003). The significant economic implications of seasonal bahiagrass forage shortfall have stimulat ed research aimed at increasing productivity during shor t-daylength months. This effort has involved genetic selection and development of bahiagrass cultivars that are cold-adapted and less sensitive to photoperiod (Blount et al., 2001). The bahiagrass breeding program, which is a mu lti-disciplinary and multi-state effort with scientists from the University of Florida and USDA-ARS scientists in Georgia and Florida, focused on selection for cold tolerance, late -season forage growth, and improved rooting in Pensacola-type bahiagrass (Blount et al., 2001). The re sulting selections have been termed PCA for photoperiod and cold adapted. Sele ction cycles have been grown alternating between Marianna and Ona, FL, allowing for tes ting over diverse environments to improve cold tolerance, photoperiod response, nematode a nd disease resistance, rooting/rhizome mass,

PAGE 19

19 seedling vigor and establishment, seasonal distri bution of forage production, and forage quality (Blount et al., 2001). Additional cool-season productivity must not come at the expens e of persistence, however, because the cost of pasture renovation is high and return per unit land area from grasslandlivestock systems are relatively low. Hirata et al. (2002) observed that the decrease in aboveground bahiagrass pasture productivity with decreasing daylength wa s due to greater allocation of photosynthate to roots and rhizomes. Thus one concern with genotypes that are more productive during the cool season is that increasing above-ground gr owth may negatively affect the allocation of carbohydrates to storage orga ns. Carbohydrate reserves as well as stored nitrogen (N), are important factors that influence pasture pers istence. Adequate carbohydrate reserves are important for winter survival (i.e., to support respira tion of plant organs; Ta et al., 1990) and for initiation of early spring growth in perennial plants (D hont et al., 2002). Lesser mass and nutrient content of warm-season grass stor age organs have been associated with lower herbage yields in the subsequent growing season reduced tolerance to defoliation, and eventual stand loss (Chaparro et al., 1996; Macoon et al., 2002). In bahiag rass, greater persistence has been associated with maintenance of high tiller, rhizome, and root densities (Hirata, 1993b), along with maintenance of green leaves for photosynthesis (Pakiding and Hirata, 2003). Sinclair et al. (2003) reported no adverse affects of prolonged c ool-season growth of Pensacola bahiagrass (induced by artificial lig ht) on subsequent warm-season growth. There was a strong trend ( P = 0.051) toward lower below-ground mass in April; however, this had disappeared by mid-June. In that study, harvest intervals were 5 wk during the cool season and 4 wk (to a 7.5-cm stubble) during the warm s eason, while fertilization was 67, 15, and 56 kg ha-1 of N, phosphorus (P), and potassium (K), respectively, after each harvest. Plots were also

PAGE 20

20 irrigated to avoid water deficits. These manage ment practices, i.e., extended regrowth period, high fertilization levels, and irri gation, are considerably more conduc ive to stand persistence than those used by most producers in Florida, thus, additional work is warranted to assess the impact of greater cool-season growth on subsequent bahiagrass productiv ity and persistence under more typical management. The recent availability of ba hiagrass germplasm (PCA cycles) selected for greater cold tolerance and cool-season growth provides opportunity for detailed study of the mechanisms resulting in these responses and fo r assessing persistence of this germplasm under a range of defoliation treatments. The research reported in this dissertation de rives from the problem of seasonality of bahiagrass production and the potential of ne w germplasm to address this problem. Two experiments were developed and carried out for 2 yr. The objectives of Experiment 1 were to determine the effects of extended daylength and N and K fertilization during establishment year on seasonal changes in: Herbage production, the amount and proportion of dry matter (DM) in root, rhizome, stem base, and leaf components, the N and total nonstructural ca rbohydrate (TNC) content of r oot and rhizome + stem base components, and leaf photosynthesis, leaf area, and tiller number. It is anticipated that these data, which are presented in Chapters 3 and 4, will allow us to determine the effects of extended daylength a nd fertilization management on C allocation and growth of a PCA type compared with current cultivars during establishment year, and to determine if the responses of a PCA type are fundamentally different from those of existing cultivars.

PAGE 21

21 The objectives of Experiment 2 were to dete rmine the effects of de foliation frequency and intensity on: Herbage production and nutritive value, the amount and proportion of DM in root + rhizome and stem base components, the N and TNC content of root +rhi zome and stem base components, tiller number, mass, and appearance and death rates, and bahiagrass cover. These data are presented in Chapters 5 a nd 6, and will allow us to determine if the productivity and nutritive value of plants selected for increased cool-season growth is equivalent to or greater than current cultivars, if this char acteristic puts these plan ts in greater danger of decreased stand longevity, and if these plants ar e more sensitive to defoliation management than existing cultivars.

PAGE 22

22 CHAPTER 2 LITERATURE REVIEW Introduction of Bahiagrass to Florida and Uses in Agriculture Bahiagrass is native to S outh America and is found extensively on light-textured soils in the tropical and warm temperate regions of th e Western Hemisphere (Gates et al., 2004). Common bahiagrass was introduced into Florid a in 1913 (Ball et al., 2002). In 1935, a more productive and vigorous bahiagra ss was found growing near ship docks in Pensacola, Florida. This genotype was eventually released by the Un iversity of Florida and the Soil Conservation Service as the cultivar Pensacola (Ball et al., 2002). Bahiagrass is commonly established as pastur e for beef cattle, low maintenance turf, and sod-based crop rotation (Gates et al., 2004). Bahiagrass can also be utilized as hay, but it tends to have low nutritive value when dry matter (DM) yi eld is sufficiently high to make a hay harvest practical (Kalmbacher, 1997). Production Characteristics of Bahiagrass Environmental Adaptation Bahiagrass is an aggressive, deep-rooted, warm -season per ennial that is adapted to a wide range of soil and management conditions in Florida. It thrives in deep, well-drained acidic soils, poorly drained soils, and clay loam soils (Kalmbach er, 1997). It is tolerant of low soil fertility and over-grazing (Ball et al., 2002) As a warm-season grass, wint er temperatures frequently limit bahiagrass range of adaptation to tropical and subtropical areas (Gates et al., 2004). Diploid Pensacola-types are generally more cold and frost-tolerant than the tetraploid Argentine-types (Chambliss and Adjei, 2006), and the more rhizomatous types endure low temperatures better (Pedreira a nd Brown, 1996b). While bermudagrass [ Cynodon dactylon (L.) Pers.] is dormant following freezing winter temper atures, bahiagrass can remain green at low

PAGE 23

23 temperatures above 0C (Gates et al., 2001). A ccording to Mislevy and Dunavin (1993), even though the upper surface of bahiagra ss top growth is usually killed by frost, considerable green forage can exist within the sward if 15 to 20 cm of top growth is achieved prior to a frost event. Yield and Yield Distribution Mislevy et al. (2005) report ed 3-yr annual dry biom ass yields ranging from 10.3 Mg ha-1 for Pensacola to 12.1 Mg ha-1 for Tifton 7 in Central Flor ida. Bahiagrass cultivars were harvested monthly and fertilized with 56 kg N ha-1 prior to the initial and after subsequent harvests. For bahiagrass cultivars fert ilized once annually with 56, 28, and 56 kg ha-1 of N, P, and K, respectively, and cut to a 5-cm stubble every 35 d in Southwest Florida, Muchovej and Mullahey (2000) reported 2-yr annual DM yields ranging from 3 Mg ha-1 for Paraguay to 4 Mg ha-1 for Tifton 7. Hirata (1993a) reported greater yields of Pensacola with increasing stubble heights in the summer (8 Mg ha-1 at 22-cm stubble vs. 4.5 Mg ha-1 at 2-cm), but greater yields with decreasing stubble heights in the fall (2 Mg ha-1 at 2-cm stubble vs. 1 Mg ha-1 at 22-cm) when harvested every 2 to 4 wk. Chambliss ( 2003) reported total av erage annual yield of Pensacola, Argentine, and Tifton 9 to be approximately 10.3 Mg DM ha-1. In the southeastern United States (US), bahiagrass is productive fr om April to October (Mislevy and Dunavin, 1993; Ball et al., 2002). Due to a longer grow ing season in South Florida, bahiagrass forage growth is somewhat more evenly distributed throughout the year than in North Florida (Chambliss and Adjei, 2006) but in general 85 to 90% of bahiagrass forage is produced from April through September (Mislevy and Ever ett, 1981; Kalmbacher, 1997). Stewart et al. (2007) reported herbage accumulation rates ( HAR, dry matter basis) of 30, 62, and 15 kg ha-1 d-1 and herbage masses (HM, dry matter basis) of 2.2, 2.7, and 3.3 Mg ha-1 in May, mid-July, and October, respectively, for continuously stocked Pensacola bahiagrass pastures receiving 120 kg N ha-1 yr-1 at a stocking rate (SR) of 2.8 animal units (AU) ha-1.

PAGE 24

24 Yield Responses to Fertilization Bahiagrass is generally responsive to N fertiliza tion, but is less responsive to P and K. Soil fertility studies on established, grazed bahiagrass pastures in Central and South Florida have shown very little yield response to P and K application even though soils tested low in these nutrients (Chambliss and Adjei, 2006). The lack of response to applied P is thought to be due to the ability of roots to extract P from deeper portions of the soil profile that are not sampled for routine soil analysis (Rechcigl a nd Bottcher, 1995). As a result, Un iversity of Floridas Institute of Food and Agricultural Sciences (IFAS) has not recommended P fe rtilizer application to grazed established bahiagrass pastures in Central and South Florida (Kidder et al., 2002). This recommendation has recently been revised beca use long-term withholding of P fertilizer has been associated with bahiagrass stand decline (Mylavarapu et al., 2007). The current recommendation is to maintain soil pH at 5.5 or greater to increase P av ailability, and if the soil tests low or very low in P and plant tissue P concentrati on is less than 0.15 g (100 g)-1, then P fertilizer is recommended (Mylavarapu et al., 2007). Bahiagrass can persist under low soil fertility but does show increased DM production with N fertilization (Gates et al., 2004). In a study conducted in Central Florida with Pensacola bahiagrass, Johnson et al. (2001) re ported forage DM yields averag ed across five harvests (8-cm stubble cut at 28-d intervals) for five N fer tilizer amounts. Pensacola yielded 0.8 Mg DM ha-1 harvest-1 with 0 kg N ha-1, with forage yield peaking at 1.4 Mg DM ha-1 with 39 kg N ha-1 applied after each harvest. Johnson et al. (2001) reported no bahiagrass DM yield increase with N fertilizer amounts over 39 kg N ha-1 (up to 167 kg N ha-1) applied after each harvest. Research conducted at the Southeast Research Station in Louisiana showed that rates of 0, 224, 336, and 448 kg N ha-1 yr-1 produced Pensacola DM yields of 4, 12, 14, and 17 Mg ha-1, respectively

PAGE 25

25 (Twidwell et al., 1998). In Tifton, GA, research showed rates of 56, 112, 224, and 448 kg N ha-1 produced 3-yr average Pensacola DM yields of 6, 8, 12, and 15 Mg ha-1 (Burton et al., 1997). Tolerance to Defoliation A desirable attribute of bahiagrass is excellent persistence under seve re grazing (Gates et al., 2004). G ates et al. (1999) reporte d bahiagrass DM yields were greater at low cutting height (1.5 cm) than high cutting height (10 cm) at 2or 4-wk regrowth intervals over 2 yr. These authors also reported that bahi agrass persistence (visual cover estimate) was not influenced by cutting interval (2, 4, and 8 wk) throughout 3 yr of the experiment and was not influenced by cutting height after the first y ear. In a 2-yr grazing experiment Gates et al. (1999) reported Pensacola maintained 60% cover under continuous stocking of yearling heifers, while Tifton 9 cover was reduced from 70 to 60% after 2 yr. Bahiagrass forms a dense sward that contributes to its grazing tole rance (Hirata, 1993b). Bahiagrass also maintains stable tiller density in terms of space and time, which contributes to its high persistence under grazing (Hirata and Pakiding, 2004). Research in Japan showed that density characteristics of Pensacola bahiagra ss, such as tiller number and stolon1 length, tended to increase as 2to 4-wk interval cutting heights decreased from 22 to 2 cm (Hirata, 1993b). When Pensacola was defoliated daily to re move all regrowing lami nae from an initial 2cm cutting height, initial tiller density was mainta ined for 4 to 6 wk of imposing these treatments before declining (Pakiding and Hirata, 2002). In an N fertilization and defoliation intensity experiment in Japan, Pakiding and Hirata (2003) reported that Pe nsacola responded to low N (5 1The published literature is not consistent in referring to ba hiagrass as possessing stolons or rhizomes (Gates et al., 2004), so in the literature review the structur e name will be that used by the cited author.

PAGE 26

26 g m-2 yr-1) and intense defoliation (2-cm stubble at monthly harvest intervals) with increased tiller longevity, tiller appearance rate (TAR), and tiller density compared to high N (20 g m-2 yr1) and 12and 22-cm stubble heights (SH). Pensaco la tiller formation was stimulated with high N (20 g m-2 split applied annually) in May-June in a Japanese study on plots harvested monthly to a 3-cm stubble when compared to plots rece iving 5 g N m-2 split applied annually (Islam and Hirata, 2005). Hirata and Pakiding (2004) reporte d Pensacola tiller density tended to be less spatially heterogeneous than herbage mass unde r rotational stocking, resulting in greater persistence compared to other pasture grasses. The importance of storage organs to defo liation tolerance has been well-documented (Chaparro et al., 1996; Macoon et al., 2002). Research on alfalfa (Medicago sativa L.) has documented the important contributions of C re serves to respiration during regrowth and N reserves to shoot regrowth after defoliation (Ta et al., 1990 ; Meuriot et al., 2005). When subjected to continuous severe defoliation, bahiagrass tillers depend considerably on storage organs such as stolons, rhizomes, and roots for energy for maintenance and growth of new leaves (Hirata and Pakiding, 2003). Stol on mass of bahiagrass has been reported up to 1060 g DM m-2 (Hirata, 1994). Although neither cutt ing interval nor cutting height influenced root mass, Gates et al. (1999) reported an average r oot mass of 27 g per 75 mm diameter by 75 mm depth soil core (6100 g DM m-2) for bahiagrass cut every 2, 4, or 8 wk to 1.5or 10-cm stubble heights in Tifton, GA. Hirata and Pakiding (2003) reported that bahiagra ss swards degraded when all regrowing laminae were removed every 1 to 4 d, resulting in a reducti on of laminae production, mass of stubble, and stolon mass. They concluded that stolons play a key role in bahiagrass defoliation tolerance.

PAGE 27

27 Forage Nutritive Value and Quality Ball et al. (2 001) define nutritive value as protein, mineral, and energy composition, availability of energy, and efficiency of energy utilization, while forage quality is defined as the ability of a forage to suppor t desired levels of animal perf ormance (e.g., daily gain or milk production). Forage quality is considered to be a function of voluntary animal intake and nutritive value, and is affected by variations in plant genotype, matu rity, season of growth, management, and anti-quality f actors (Adesogan et al., 2006). Bahiagrass forage provides ade quate nutrition for mature beef cattle, but growing animals such as calves may make only small daily ga ins on bahiagrass in th e late summer months (Chambliss and Adjei, 2006). Sollenberger et al. (1988) reported no difference between continuously stocked Pensacola bahi agrass and Floralta limpograss [ Hemarthria altissima (Poir.) Stapf & Hubb] pastures in terms of av erage daily gain (ADG) (2-yr average of 0.38 and 0.33 kg d-1, respectively) or gain ha -1 (2-yr average of 370 and 344 kg ha-1, respectively) of yearling steers in Gainesville, FL. In a study by Sollenberger et al. (1989) on rotationally stocked Mott elephantgrass ( Pennisetum purpureum Schum.) and Pensacola bahiagrass pastures with an average SR of 4 yearling steers ha-1, ADG (0.97 and 0.38 kg d-1, respectively) and gain ha -1 (483 and 318 kg ha-1, respectively) were greater for Mott than Pensacola. Stewart et al. (2007) reported greater heifer ADG (0.34 and 0.28 kg d-1, respectively) for Pensacola bahiagrass under low intensity management (40 kg N ha-1 yr-1, 1.4 AU ha-1 SR) than high intensity (360 kg N ha-1 yr-1, 4.2 AU ha-1 SR), but gain ha-1 increased from low to hi gh intensity (101 to 252 kg ha-1). Animal performance is affected by bahiag rass forage yield and nutritive value as influenced by growing season. In a 3-yr clipping study, Mislevy et al. (2 005) reported greatest bahiagrass forage crude protein (CP) and in vitro organic matter digestibility (IVOMD) in April (157 and 534 g kg-1, respectively), October (157 and 542 g kg-1, respectively), and December

PAGE 28

28 (177 and 587 g kg-1, respectively), while lowest CP and IVOMD were always found in June (113 and 467 g kg-1, respectively) and August (122 and 482 g kg-1, respectively). Stewart et al. (2007) reported that Pensacola bahiagrass herbage CP and IVOMD in continuou sly stocked pastures generally decreased from May th rough August. Hirata (1993a) repo rted greatest annual in vitro DM digestibility (IVDMD) of Pensaco la at 2-cm stubble height (570 g kg-1) and lowest IVDMD at 22-cm stubble (460 g kg-1). This author also reported IVDM D was greater in the spring (580 g kg-1 IVDMD at 2-cm stubble and 540 g kg-1 at 22-cm stubble) than autumn (530 g kg-1 IVDMD at 2-cm stubble and 430 g kg-1 at 22-cm stubble). Average dail y gain of yearling steers on continuously stocked Pensacola bahiagrass in Ona, FL was greatest in May (1.0 kg d-1) and decreased throughout the grazing season to 0.15 kg d-1 in October (Prates et al., 1974). Gain per hectare from the same study was also greatest in May (122 kg ha-1) and decreased throughout the grazing season to 10 kg ha-1 in October. The authors also re ported highest forage production (average 2270 kg OM ha-1) and SR (average 11 animals ha-1) in July and August. Stewart et al. (2007) reported seasonal ADG of 0.3, 0.5, and 0.0 kg d-1 in May, mid-July, and October, respectively, for yearling beef heifers on conti nuously stocked bahiagra ss pastures receiving 120 kg N ha-1 yr-1 with a stocking rate of 2.8 AU ha-1 in Gainesville, FL. Bahiagrass shows an increase in DM producti on and herbage CP with N fertilization but the impact on other measures of nutritional valu e has been reported to be small (Brown and Kalmbacher, 1998). Johnson et al. (2001) repor ted that N fertilization had no effect on bahiagrass IVOMD or acid detergent fiber (ADF), while neutral detergent fiber (NDF) decreased and total N increased with increasing N fertiliza tion. In contrast, Stewar t et al. (2007) reported both greater bahiagrass CP (140 vs. 99 g kg-1) and IVOMD (505 vs. 459 g kg-1) for high than low intensity management. Similarly, Newman et al. (2006) reported both greater CP (58 vs. 79 g kg-

PAGE 29

291) and IVOMD (432 vs. 471 g kg-1) for Pensacola bahiagrass when N fertilizer was increased from 80 to 320 kg N ha-1. In general, bahiagrass cultivar differences in nutritive value have been minimal. Research from Louisiana showed no differences in CP or ADF among Pensacola, Ar gentine, and Tifton 9 (average 115 g CP kg-1 and 320 g ADF kg-1), while Tifton 9 and Argentine had lower NDF (640 and 642 g NDF kg-1, respectively) than Pensacola (657 g NDF kg-1; Cuomo et al., 1996). Muchovej and Mullahey (2000) reported no di fference overall in IVOMD or NDF among Argentine, Paraguay, Pensacola, Tifton 7, and Tifton 9 (average 497 g OM kg-1 and 785 g NDF kg-1), but their data for CP concentration was more variable. There was no difference among cultivars late in the growing season (average 88 g CP kg-1), while Tifton 7 tended to have lowest CP for much of the early growing season. Maturity effects on bahiagrass nutritive value and quality are similar to those reported for other C4 grasses. In a summary of Dr. John Moores sheep digestion trials, Brown and Kalmbacher (1998) reported that as bahiagrass maturity increased fr om 4 to 8 wk, total digestible nutrients (TDN) decreased from 560 to 535 g kg-1 and forage intake decreased from 22.6 to 17.4 g kg-1 of body weight. Limitations of Bahiagrass Two well-known lim itations of bahiagrass are it s relatively low nutri tive value and its susceptibility to mole cricket damage. The main fo cus of this dissertation is issues related to bahiagrass yield and yield distribution, so this se ction of the literature review will address these topics only. Harvested Forage Yields Under sim ilar growing conditions, bahiagrass ge nerally produces lower yields than other perennial warm-season grasses in Florida, with annual DM yield of bahiagrass ranging from

PAGE 30

30 approximately 4 to 12.1 Mg ha-1 (Muchovej and Mullahey, 2000; Mislevy et al., 2005). Other warm-season perennial grasses commonly utilized in Florida include bermudagrass and stargrass ( Cynodon nlemfuensis Vanderyst). Chambliss et al. (2006) reported annual DM yields of bermudagrass ranged from 11.2 to 23.4 Mg ha-1, while those for stargrass ranged from 11.2 to 16.0 Mg ha-1 (Mislevy, 2006). Pensacola is the most widely grown bahiag rass cultivar (Chambliss and Adjei, 2006). It has a prostrate growth habit (B eaty et al., 1968) which contribu tes to its comparatively low harvested yields. The USDA-ARS began a bahi agrass breeding program in Tifton, GA in the early 1960s. Through recurrent restricted phenotypi c selection, he selected for increased aboveground yield from Pensacola populations, eventu ally releasing Tifton 9 in 1989 (Burton, 1989). Tifton 9 has a taller and more upright growth habit and sma ller basal diameter than Pensacola. Research has shown that Tifton 9 has less DM in rhizomes and more in leaves and upright stems than Pensacola (Werner and Burton, 1991). Tifton 9 generally produces 30 to 40% more forage per year than Pensacola, with total annual yields in Gainesville, FL, of 12.5 Mg ha-1 and 10.0 Mg ha-1, respectively (Chambliss, 2003). Cool-Season Growth Bahiagrass cool-season growth in the southeastern US is m inimal (Kalmbacher, 1997). In some situations, forage growth is limited by short daylength, despite there being adequate soil moisture, soil fertility, and warm temperatures (S inclair et al., 1997; Gates et al., 2004). Lack of bahiagrass forage production during the cool season is a major limitation to livestock production. The cost of purchased supplements is a major obsta cle to profitability of livestock enterprises. Development of new genotypes that overcome the seasonal shortfall of av ailable forage would have a major positive impact on th e Florida livestock industries.

PAGE 31

31 Recent Plant Breeding Advances The University of Florida ba hiagrass breeding program has used the original Pensacola germplasm and recurrent selecti on to develop late-season forage bahiagrasses that are cold tolerant and less photoperiod sensitive, allowing these populations to withstand late fall, winter, and early spring cold temperatures common in Florida (Blount et al., 2003). From germplasm in a North Florida Research and Education Center (NFREC)-Quincy photoperiod study (19992001), plants were selected for high cold toleranc e and excellent crown and top growth in late spring 1999. Vegetative cuttings were crossed and seed from this cycle (Cycle 1) were germinated at NFREC-Quincy. Seedlings were se lected for rapid seedling emergence and vigor, and planted at NFREC-Marianna in summer 1999. Selections were made for fall forage growth and cold tolerance, and vegeta tive cuttings were crossed in ea rly spring 2000 and planted in NFREC-Quincy. Seed from this cycle (Cycle 2) was germinated and seedlings were transplanted at the Range Cattle Research and Education Center (RCREC)-Ona, FL in summer 2000. Vegetative cuttings were taken from plants sele cted for photoperiod insensitivity, cold tolerance, and general appearance. These cuttings were ta ken back to NFREC-Marianna and crossed in spring-summer 2001. Seed from this cycle (Cycle 3) was tested for rapi d seedling emergence and seedling vigor at Tifton, GA. Plants selected from this cycle were planted at RCREC-Ona in fall 2001, and then superior plants we re selected and crossed in summer 2002 at NFREC-Marianna. Seed from this cycle (Cycle 4) was germinated and plants were selected for rapid seedling emergence and seedling vigor at NFREC-Quincy in late summer 2002. It is this fourth cycle, termed PCA (photoperiod cold adapted) Cycle 4, wh ich will be the focus of research described in this dissertation. Of greatest in terest is the relationship of photoperiod insensitivity and cold adaptation to forage yield and long-term persistence.

PAGE 32

32 Daylength Effects on Forage Plants Several seasonal responses of plants are synchronized by photoperiodism (physiological reaction of organism to length of day or night ) as sensed through phytochromes (photochromic protein photoreceptors). Plant responses such as stem growth/elongation, leaf growth and abscission, cold acclimation/development of fros t resistance, tillering, formation of storage organs (Salisbury and Ross, 1992), dormancy, and allocation of photosynthetic resources to root, stem, leaf, reproductive, or storag e structures are potenti ally controlled by th e perception of light signals by phytochromes (Smith, 2000). Plants possess innate circadian rhythms through which phytochromes selectively regulate gene expressi on, ensuring crucial developmental steps are initiated at appropriate points in the life cycle (Smith, 2000). In temperate legumes such as white clover ( Trifolium repens L.), short photoperiod favored storage of soluble carbohydrates, possibl y improving plants winter survival rates but decreasing forage yield (Boller and Nosber ger, 1983). Above-ground bahiagrass pasture productivity in subtropical areas decreased as da ylength decreased as plants allocated greater proportions of nonstructural carbohy drates to storage organs (Hir ata et al., 2002). Research has shown that plant DM production can be in creased substantiall y by extending photoperiod without increasing photosyntheti cally active radiati on (PAR) supply (Hay, 1990). Pasture grass breeders in the United Kingdom have conducted rese arch aimed at increasing spring and autumn production without sacrificing wi nter hardiness (Hay, 1990). This area of research was stimulated in the early 1960s when breeders c onsidered low incident light energy and winter temperatures to be the main climatic f actors limiting forage production (Cooper, 1964). Subsequently, a breeding program was developed using Mediterranean, Atlantic maritime, and continental populations of forage grasses that produced more DM in winter and early spring in Britain than previously used material. Treatments were imposed to examine 1) seasonal

PAGE 33

33 variations in leaf a nd tiller development a nd 2) the effects of temperature and photoperiod on leaf development (Cooper, 1964). Results from this re search suggest that 1) differences in winter growth were primarily due to differences in re sponse to variation in te mperature, which were produced largely by differences in rate of leaf expansion, and 2) the basic differences in cell division and extension between clim atic races was due to differences in the use of assimilates in the expansion of new leaf surface, not in a ny difference in photosynthetic activity (Cooper, 1964). In the US, low productivity during short-da ylength months of economically important pasture grasses such as bahiagra ss (Blount et al., 2003), sorghum [ Sorghum bicolor (L.) Moench; Rooney and Aydin, 1999], and pearl millet [ Pennisetum glaucum (L.) R. Br.; Rai et al., 1999] has led to selection of less photoperiod sensitive genotypes that extend growth into the cool season. Grasses exposed to extended daylength generally exhibit increased DM partitioned to leaves, increased leaf blade and sheath length, an d changes in growth habit from prostrate to erect (Hay, 1990). Increased relative growth rate of temperate past ure grasses has been attributed primarily to increased leafiness, which results in increased radiant ener gy interception (Hay, 1990). By artificially extending daylength during the winter months, Si nclair et al. (2001) reported that forage yield (5-week harvests during the time of shortest daylength) of Pensacola bahiagrass, Tifton 85 bermudagrass ( Cynodon spp.), Florakirk bermudagrass, and Florona stargrass were increased up to 6.2-fold more than the yield under natural daylength. In another study with these same four subtro pical species, Sinclair et al. (2003) reported increased growth of all grasses in the extended photoperiod treatment dur ing the short daylength months. In this study, Pensacola bahiagrass exhibited the greatest in crease in yield for at least five harvests in

PAGE 34

34 each season, with forage yield for many harves ts under the extended daylength treatment being more than double that of the natural dayl ength treatment (Sincl air et al., 2003). By extending growth into the winter mont hs, carbohydrate reserves may be reduced, which may result in depressed forage production of less photoperiod sensitive genotypes in the subsequent spring and summer (Sinclair et al., 20 03). In a 2-yr study with Pensacola bahiagrass, Tifton 85 and Florakirk bermudagr asses, and Florona stargrass fertilized at 67, 15, and 56 kg ha-1 harvest-1 of N, P, and K, respectively, Sinclair et al. ( 2003) reported 1) no de crease in growth following the extended photoperiod treatment in either season, 2) no difference in below-ground tissue mass throughout the season between extended and natural daylength treatments, and 3) no influence of extended daylength on total nonstr uctural carbohydrate (T NC) concentration of below-ground tissue. Commonly grown bahiagrass genotypes such as Pensacola and Tifton 9 are sensitive to short daylengths, which induce a change of growth pattern re sulting in a reduction in aboveground growth during the cool season. During this time, remaining live herbage continues to photosynthesize, but instead of using photosynthate to produce new above-ground tissue, plants generally store photosynthate as nonstructural carbohydrates in stem bases, stolons, and rhizomes (Thornton et al., 2000). At this physiolo gical growth stage, plants are effectively preparing for winter survival and early spring regrowth, when photosynthetic production is inadequate to meet growth demands (Thor nton et al., 2000). Carbohydrate reserves play important roles in regrowth and tissue maintenance, particularly after seve re defoliation (Pedreira et al., 2000) and in early spring growt h. In defoliated perennial ryegrass ( Lolium perenne L.), Morvan-Bertrand et al. (1999) repor ted that 91% of the C in the basal part of elongating leaves

PAGE 35

35 was derived from reserves assimilated prior to de foliation. It is this char acteristic of C storage that is thought to contribute to the pers istence of bahiagrass pastures in Florida. The accumulation of carbohydrate reserves depends on rates of photosynthesis and respiration in the plant. In general, carbohydrate reserves decrease when growth and respiration demands exceed photosynthetic rate and increase when growth and respiration demands are less than photosynthetic rate (Fulkerson and Donaghy, 2001). As seasonal temperatures decrease, DM distribution to leaf blad es typically decreases (Inosak a et al., 1973) and nonstructural carbohydrates stored in stems are translocated to roots and stolons be fore winter dormancy (Hirata et al., 2002). Stored N also plays an important role in ea rly shoot regrowth in the spring and following defoliation. Nitrogen stored in plant tissues can be remobilized and allocated to active shoot meristems for the production of new leaf tissue in order to restore canopy photosynthesis. Lattanzi et al. (2005) reported ~ 50% of stored C and > 80% of stored N were remobilized to supply new leaf growth in pere nnial ryegrass and dallisgrass ( Paspalum dilatatum Poir.). Avice et al. (1996) reported that 34% of total assimilated 15N was recovered in alfalfa shoots after 30 d of regrowth, indicating the signifi cance of stored N to the synthesis of amino acids and proteins in leaves and stems. Extended daylength has been associated with a reduction in emerged tillers per plant in temperate pasture grasses, which may be detrimen tal to pasture survival and persistence (Hay, 1990). Increased tiller appearance has been associated with an increased red:far-red ratio of incident light. As canopy density increases, th e proportion of incident radiation intercepted increases and the red:far-red ratio at plant base s decreases, thereby reduc ing tillering (Casal et al., 1985). Extended daylength results in incr eased above-ground growth due to increased

PAGE 36

36 interception of PAR (Hay, 1990). Most forage gra sses require sunlight striking the basal buds to initiate new tiller development (Casal et al., 1985; Christiansen and Svejcar, 1988; McKenzie, 1997). When exposure to extended daylength results in excess forage growth, sunlight is prevented from striking basal buds and tillering is retarded (Salisbury and Ross, 1992). Tiller formation is delayed in dense swards because the li ght transmitted to basal buds is rich in far-red light wavelengths, causing a d ecrease in the ratio of Pfr (phytochrome/pigment that absorbs farred light) to total amount of phytochrome (Salis bury and Ross, 1992). Conversely, increasing the ratio of Pfr to the total amount of phytochrome with light rich in red wavelengths promotes tillering (Salisbury and Ross, 1992). Genetic selection of greater yielding grasses has been uti lized as a way of improving pasture productivity and animal performance. However, some research suggests that these improvements in forage yield may be detrimental to pasture persistence. Ti fton 9 is a taller, more erect, higher yielding bahiagrass cultivar selected from populati ons of Pensacola bahiagrass, which is a prostrate cultivar that partitions a large proportion of its dry weight to rhizomes (Beaty and Tan, 1972). Pedreira and Brown (1996a) reported that selection for increased bahiagrass yields (e.g., Tifton 9) appeared to have resulted in increas ed allocation of dry matter to harvestable foliage and possibly a greater pr oduction of non-root biomass. Additionally, Pedreira and Brown (1996b) reported that selectio n for increased bahiagra ss yields (e.g., Tifton 9) resulted in taller plants with fewer rhiz omes, a greater tendency fo r winter injury, and a greater susceptibility to population shifts under close, frequent mowing. This research suggests that it may be a significant challenge to main tain pasture persistence with improvements in forage yield. As related to the objectives of the dissertation research, bahiagrass genetic

PAGE 37

37 populations that are less sensit ive to photoperiod and that exte nd growth into the cool-season could be less persistent than those which cease top growth earlier. Forage Plant Responses to Defoliation and Mechanisms of Grazing Tolerance Defoliation has several effects on pasture plants with the m ost notab le effect being the disruption of carbohydrate supply fo r plant growth by removing photosynthetic tissue (Chapman and Lemaire, 1993). While some plants may have ch emical or morphological features that enable avoidance of defoliation, forage plants by virtue of the nature of their use in production systems can not completely avoid herbivory. Thus, it is crit ically important for forage plants to have the ability to replace tissues lost to herbivores in order for plant survival, growth, and reproduction (Richards, 1993). The initi al physiological responses of forage plants to defoliation are therefore important in long-term pasture persistence. Richards (1993) describes seve ral important defoliation charact eristics that influence the physiology of recovery. The amount and type of tissue removed (e.g., mature leaves vs. meristematic tissue), stage of plant developmen t, rate of tissue removal (e.g., continuous vs. discrete defoliation), and plants abiotic (e.g., light, water, and nut rients) and biotic environments (e.g., proximity to defoliation-tolera nt plants) are important factors in determining the impact of defoliation on plants. Two distinct phases in the physiology of defo liation-tolerant plants after moderate to severe defoliation are described by Richards (1993); the first phase is a transient period of one to a few days (immediate effects) and the second phase is a readjustment of physiological activity and plant morphogenetic integration that occurs over several weeks. The immediate effects of defoliation result in the instantaneous reduction of photosynthesis, root respiration and growth, N fixation, and nutrient ab sorption, and cessation of transl ocation of stored C by phloem loading. Total non-structural carbohydrates of roots decline as roots continue to respire, thereby

PAGE 38

38 utilizing these reserves. Additi onally, photosynthate supply to roots is reduced immediately because of greater allocation to sh oot and leaf meristematic tissues that contributes to a rapid reestablishment of the photosynt hetic capacity of the canopy. As plants continue to recover from defoliati on (i.e., second phase of days to weeks after defoliation), rates of C and/or nutri ent acquisition increase to aid in recovery. In order to achieve a positive whole-plant C balance, photosynthate is allocated to plant growth and maintenance, but not to storage. Research has shown that the photosynthetic rates of foliage on defoliated plants may be higher than undefoliated foliage of the same age, a response known as compensatory photosynthesis (Richards, 1993) Compensatory photosynthesis involves a rejuvenation of leaves or an inhibition of norma l photosynthetic decline associated with old and dying leaves, and is most often the result of increased photosynt hetic capacity of the mesophyll (Hodgkinson, 1974). Compensatory photosynthesis can be induced by changes in light intensity and quality that result from modifications in microhabitat and modifica tions from physiological functions that are mediated by cytokinins (M anske, 1999). Richards (1993) states, The development of these compensatory processes coupled with prefer ential shoot growth allocation is the second phase of plant recovery from defo liation and results in a return to more normal plant function. The ability to rapidly re -establish the photosyntheti c capacity of the canopy after defoliation is an important characteristic of defoliation-tolerant plants, and the presence of active shoot meristems allows for rapid leaf expansion from existing cel ls (Mott et al., 1992; Paige, 1992). If severe and/or frequent defoliation continues for an extended period, plant morphology may change to aid in the mainte nance of growth and plant surviv al (i.e., phenotypic plasticity). Defoliation tends to result in greater tiller density and number, smaller leaves, and shorter

PAGE 39

39 petioles. For example, under increasing grazing pr essure (i.e., increasing stocking rate), grass populations tend to adjust their structure so that there is a high density of small tillers per unit area (Bircham and Hodgson, 1983; Grant et al., 1983; Christia nsen and Svejcar, 1988). Additionally, some species are able to adjust their meristematic position by changing their growth habit from upright to prostrate, effectiv ely orienting new tillers parallel to the soil surface as protection from defoliation (Hyder, 1973; Mans ke, 1999). Tillers in grazed pastures tend to be relatively short-lived, with a cont inual turnover of plant material (Davies, 1988). As the level of defoliation intensity increases, leaf size and peti ole length are reduced, re sulting in some leaf material positioned below grazing height and av ailable for photosynthesis (Sheath and Hodgson, 1989). Frequent, severe defoliation may have negative effects on the mass of storage organs such as stem bases, rhizomes, stolons, and roots, which may in turn affect long-term pasture persistence. Gates et al. (1999) reported greater spring reserves (as estimated by total etiolated initial spring growth) in bahiagrass plots cut th e previous growing season every 8 wk than in those cut every 2 or 4 wk, while cutting height did not affect spring reserves. Gates et al. (1999) reported greater root mass for Pensacola vs Tifton 9 (33.9 vs. 21.4 g per 75by 75-mm soil core), but neither cutting height (1.5 or 10 cm) nor cutting interval (2, 4, or 8 wk) affected root mass. For perennial ryegrass and red fescue ( Festuca rubra L.), with defoliation treatments of undefoliated, defoliated once to a 4-cm stubble, or defoliated weekly to a 4-cm stubble, Thornton and Millard (1997) reported that weekly defoliated perennial ry egrass and red fescue had lesser root mass (14.4 and 59.3 mg root DM plant-1, respectively) than undefo liated and once-defoliated perennial ryegrass (148 a nd 133 mg root DM plant-1, respectively) and red fescue (158 and 105 mg root DM plant-1, respectively). These authors concluded that increased defoliation frequency

PAGE 40

40 of both grasses had little effect on N derived from root uptake for leaf growth, but depleted their capacity for N remobilization, resulting in a reduction in the rate of leaf growth. Chaparro et al. (1996) reporte d that frequent, close defolia tion of Mott elephantgrass resulted in reduced light interception, rhizom e mass, rhizome TNC and N reserves, and number of tillers per plant. However, Pedreira et al. (2000) reported that increasing stubble height of Florakirk bermudagrass resulted in a reduction of rhizome mass and TNC pool of rhizome plus stubble. Pensacola bahiagrass was harvested every 2 to 4 wk at five stubble heights, 2, 7, 12, 17, and 22 cm (Hirata et al., 2002). They reported th at early in the growi ng season, lower cutting heights showed higher leaf densities in which ra diation flux density declined less sharply as it passed through layers of LAI. Th is indicates favorable light utili zation efficiency for gross plant production, which may contribute to hi gh tolerance to close defoliation. In a quantitative review on effect of defoliation on growth of several grass genera, Ferraro and Oesterheld (2002) reported that N availability played a signifi cant role in the magnitude of defoliation effect on total biomass production. Specif ically, plants grown at high N levels were more negatively affected by defoliation than plan ts grown at standard N levels, while plants grown at lower than standard N levels were not affected. Similar data ar e reported in a review by White (1973) and may be explained by N fertili zation stimulation of amino acid and amide compound synthesis. This occurs at the expe nse of carbohydrate reserv es because carbohydrate reserves are used as C skeleton s for protein synthesis. Hirata and Pakiding (2003) reported no effect of N fertilizer on tiller survival or stolon bi omass of Pensacola bahiagrass. Summary and Project Objectives Bahiagrass is the prim ary pasture grass used in cow-calf production in Florida, but forage production peaks in late June and July (Johnson et al., 2001) resulting in a late-season forage deficit and subsequent increased costs to producers associated with supplemental feeding. This

PAGE 41

41 reduction in bahiagrass growth may be attributed in part to short dayle ngth (Sinclair et al., 1997, 2003). Research aimed at increasing the cool-sea son productivity of bahiagrass has resulted in the selection of PCA bahiagrass genotypes that are photoperiod and cold-adapted (Blount et al., 2001). However, the production and persistence of these genotype s has not been evaluated under low-input fertilizer management and intensive defoliation management. Two experiments were conducted to address th e issues of persistence and production of this potential new PCA germplasm. The objectives of Experiment 1 were to determine the effects of extended daylength and N and K fertilization on seasonal changes in herbage production; root, rhizome, stem base, and leaf chemical co mposition and DM amounts and proportions; leaf photosynthesis; leaf area; and till er number. The objectives of Experiment 2 were to determine the effects of defoliation frequency and intensity on herbage production and nutritive value; root + rhizome and stem base chemical compositi on and DM amounts and pr oportions; tiller number, mass, and appearance and death rates; and bahiagrass cover. These data will allow us to determine the effects of various treatments on PCA-type bahiagrass compared to current genotypes and to determine if these plants are ultimately more sensitive to defoliation management than existing cultivars.

PAGE 42

42 CHAPTER 3 BAHIAGRASS GROWTH, PHOTOSYNTHES IS, AND MORPHOL OGY RESPONSES TO DAYLENGTH IN ESTABLISHMENT YEAR Introduction Bahiagrass ( Paspalum notatum Flgge) is the prim ary past ure grass for beef cattle ( Bos sp.) and horses ( Equus caballis ) in Florida and is grown on more than one million ha (Chambliss and Adjei, 2006). It is a warm-season perennial that in Florida produces 85 to 90% of total annual herbage yield during April through September (Mislevy a nd Everett, 1981; Kalmbacher, 1997). Bahiagrass generally produces lower yields than other perennial warm-season grasses in Florida, with annual DM yield of bahiagrass ranging from approximately 4 to 12 Mg ha-1 (Muchovej and Mullahey, 2000; Mislevy et al., 2 005). Chambliss (2003) re ported total average annual yield of Pensacola and Argentine bahiagrasses to be approximately 10.3 Mg dry matter (DM) ha-1. Tifton 9 bahiagrass was selected for gr eater yield and generally produces 30 to 40% more forage per year than Pensacola, with to tal annual yields in Ga inesville, FL of 12.5 and 10 Mg ha-1, respectively (Chambliss, 2003). Commonly used bahiagrass cultivars such as Pe nsacola and Tifton 9 are sensitive to short daylengths, which induce a change of growth pa ttern, resulting in a re duction in above-ground growth during the cool season. Li mited bahiagrass herbage accu mulation during the cool season has occurred even when temperature, soil mois ture, and soil fertility were adequate for substantially greater growth (Sin clair et al., 1997). Plants that exhibit this response typically continue to photosynthesize during this time, but instead of using photosynthate to produce new above-ground tissue, they are thought to store ph otosynthate as nonstructu ral carbohydrates in stem bases, stolons, and rhizomes (Thornton et al., 2000). The reserve storage compound of tropical and subtropical perennial grasses is primarily starch a nd is stored in the vegetative organs (Smith, 1972). Reserve storage can occur wh en leaf area is adequately high and provides

PAGE 43

43 photosynthate in excess of growth and maintenance needs, or when plant growth rate is low (i.e., in the autumn for subtropical grasses) or is constrained by resource ava ilability (e.g., water or other nutrients) (Thornton et al., 2000). These response patterns allow plants to effectively prepare for winter survival and early spring regrowth (Thornton et al., 2000). Frequent, severe defoliation may have negative effects on the mass of storage organs such as stem bases, rhizomes, stolons, and roots, which may in turn affect long-term pasture persistence (Ortega-S. et al., 1992; Chaparro et al., 1996). Root growth is generally reduced by defoliation as a result of the re duction of photosynthetically activ e tissue and prioritization to shoot growth of a limited supply of carbohydrate (R ichards, 1993). Lower cutting height or more frequent defoliation results in a greater reduction in root we ight (Youngner, 1972). Beaty et al. (1970) reported harvest frequency had little effect on Pensacola fo rage yield, but root and stolon mass were 50 to 75% greater for 6than for 1-wk harvests. Carbohydrate reserves are generally reduced by defoliation (Youngner, 1972). Gates et al. (1999) reported greater spring reserves (as estimat ed by total etiolated in itial spring growth) in bahiagrass plots cut the previous growing season every 8 wk than in those cut every 2 or 4 wk, while cutting height did not aff ect spring reserves. Chaparro et al. (1996) reported that frequent, close defoliation of Mott elephantgrass ( Pennisetum pupureum Schum.) resulted in reduced light interception, rhizome mass, rhizome total nonstructural car bohydrate (TNC) and N reserves, and number of tillers per plant. Pe nsacola bahiagrass crow n had an average TNC concentration of 121 g kg-1 while Argentine averaged 97 g TNC kg-1 when grazed every 2, 4, 6, and 8 wk in Ona, FL (Adjei et al., 1989). Af ter 3 yr of grazing every 2 wk, bahiagrass crown TNC during August to early Septem ber averaged less than 90 g kg-1, which was 40 g kg-1 lower than the overall average for th e 3-, 5-, and 7-wk grazing freq uencies (Mislevy et al., 1991).

PAGE 44

44 The significant economic implications of s easonal bahiagrass forage shortfall have stimulated research aimed at in creasing productivity during shortdaylength months. This effort has involved genetic selection and development of bahiagrass cultivars that are cold adapted and less sensitive to photoperiod (Blount et al., 2001). Additional cool-season productivity must not come at the expense of persistence, however, because the cost of past ure renovation is high and return per unit land area from grasslandlivestock systems are relatively low. There are no known studies comparing gr owth, physiological, and morphological responses to extended daylength of existing cult ivars with those of le ss photoperiod sensitive, cold-adapted bahiagrass genotypes. This inform ation is needed to explore physiological and morphological attributes that accompany incr eased yield of the less photoperiod sensitive genotypes. Therefore, the objectives of this st udy were to determine the effects of extended daylength and level of fertilization on total season and seas onal herbage yield, N concentration, leaf area and photosynthesis, till er number, and plant height and diameter of two diploid bahiagrass cultivars and a diploid cold-adapted ge notype that is less sensitive to photoperiod. Materials and Methods Experimental Site This study was conducted in pots in the field at the Beef Research Unit, northeast of Gainesville, FL, at 29 N latitude. Maxim um da ylength (20 June) from sunrise to sunset in Gainesville is about 14.45 h (Nautical Almanac, 2003). Single plants were planted in individual plastic tree pots (20-cm diameter by 46-cm depth) that were inserted into augered holes of the same dimensions in the field plots. Pots were filled by volume with well-mixed, uniform, screened topsoil. A double layer of landscaping cl oth was placed in the bottom of the pots to minimize root escape and soil loss. Pots were replanted at the beginni ng of Year 2. At the beginning of the experiment in Year 1, soil had an average pH of 7.0. Average Mehlich-I

PAGE 45

45 extractable soil P, K, Mg, and Ca con centrations were >122, 13, 97, and >3100 mg kg-1, respectively. At initiation of the second year of the experime nt, soil had an average pH of 6.9. Average Mehlich-I extractable soil P, K, M g, and Ca concentrations were >190, 68, 138, and >3780 mg kg-1, respectively. Treatments and Design Treatm ents were the factorial combinations of two daylength treatments, two fertilizer treatments, and three bahiagrass genotypes. They were arranged in f our replications of a completely randomized design. Sexual diploid bahiagrass genotypes evaluated were Pensacola, Tifton 9, and PCA (photoperiod and cold-adapted) Cycle 4. Pensacola and Tifton 9 are commonly utilized bahiagrass cultivars in Florida, while PCA Cycle 4 is a novel genotype sel ected to improve cold tolerance, photoperiod response, nematode a nd disease resistance, rooting/rhizome mass, seedling vigor and establishment, seasonal distri bution of forage production, and forage quality (Blount et al., 2001). PCA Cycle 4 has been approved for release by the University of Florida Agricultural Experiment Station as cv. UF-Riata (Blount et al., in review). Daylength treatments were ambient and exte nded to 15 h per d. Extended daylength was achieved using 12 quartz-halogen lamps (1500 W, 240 V). Three lamps spaced 3.7 m apart were mounted 2 m above the soil on posts at each of the long sides of the daylength treatment main plots (9.6 m x 8.6 m each) for a total of six lamps in each main plot. There were a total of two banks of lamps, each bank accommodating two ex tended daylength main plots. The lamp fixtures were positioned to shine light across the main plot treatment from the two opposing sides. The average photosynthetic photon flux de nsity (PPFD) measured after sunset at plant level in the extended dayl ength treatment was 24 mol m-2 s-1. Ambient daylength plots were located 5 m away from extended daylength plots. No PPFD was measured at night in the ambient

PAGE 46

46 treatment while the lamps were in use. Timers we re used to control when the lamps were on. The lamps were activated 30 min before sunset a nd turned off when the total photoperiod had reached 15 h from sunrise. Timers were adjusted as the season progressed to achieve the 15-h photoperiod. Fertilization treatments were termed hi gh and low based on N and K amounts applied. Fertilization treatments were in cluded in the experiment because some previous work looking at Pensacola bahiagrass responses to daylength was conducted using high levels of fertilizer input between August and April (Sinclai r et al., 2003), and it is not clear the exte nt to which this impacted treatment response. Treatments consiste d of distinct warm and cool-season fertilizer amounts. In the warm-season, the high treatment plots received 60 kg N and 45 kg K ha-1 harvest-1 and the low treatment received 20 kg N and 15 kg K ha-1 harvest-1. In the cool-season, high treatment plots recei ved 30 kg N and 22.5 kg K ha-1 harvest-1 and low treatment plots received 10 kg N and 7.5 kg K ha-1. Total annual fertilizer amount s during the trial were 330 kg N and 248 kg K ha-1 yr-1 for the high treatment and 110 kg N and 83 kg K ha-1 yr-1 for the low treatment. The IFAS recommended fertilizer amounts for bahiagrass grown only for hay are 90 kg N and 37 kg K ha-1 applied after each cutting (Mylavarapu et al., 2007). Plots consisted of 12 spaced plants in pots, one pl ant per pot, inserted into the soil in a 6 x 2 arrangement. Pots were spaced 0.4-m apart within rows and 0.5-m between rows for a plot size of 1.6 x 4.3 m. The plants used were grown from s eed that was sown in flats in the greenhouse in March. After seedlings had at least two leaves (approximately 4-cm tall), they were moved to individual speedling flats in May and grown until they were approximately 12-cm tall. They were then planted in the field in pots in July 2004. The same procedure was used for the plants used for Year 2 of the study. At the end of the first year of the experiment in June 2005, all

PAGE 47

47 remaining plants were removed from pots. A ll pots received the equivalent of 50 kg N ha-1 in early August of each year. Plots were staged on 27 Aug. 2004 and 5 Sept. 2005. At staging, all plants were cut to an 8-cm st ubble height, and fertilizer and da ylength treatments were imposed. Plots were subsequently harvested every 5 wk through November. During the cool season that followed, plots were harvested every 7 wk through April. Following the April harvest, plots were then harvested every 5 wk through June. Theref ore, this experiment was conducted from 27 Aug. 2004 to 17 June 2005 (Year 1) and 5 Se pt. 2005 to 26 June 2006 (Year 2). Response Variables Response variables m easured include total season and seasonal herb age yield, herbage N, leaf area and photosynthesis, tiller number, and plant height, diameter, and height:diameter ratio. Herbage yield was determined every 5 wk dur ing warm-season months and 7 wk during coolseason months by clipping herbage to an 8-cm st ubble height from five of the 12 pots per plot. This herbage was composited acros s the five harvested pots with in a plot. Remaining pots were also clipped to 8 cm and material was di scarded. Herbage samples were dried at 60 C to constant weight, weighed, and ground in a Wiley mill (M odel 4 Thomas-Wiley Laboratory Mill, Thomas Scientific, Swedeboro, NJ) to pass a 1-mm screen prior to analyses. For N analysis, samples were digested us ing a modification of the aluminum block digestion procedure of Gallaher et al. (1975). Sample weight was 0.25 g, catalyst used was 1.5 g of 9:1 K2SO4:CuSO4, and digestion was conducted for 4 h at 375C using 6 ml of H2SO4 and 2 ml H2O2. Nitrogen in the digestate was determined by semiautomated colorimetry (Hambleton, 1977). Photosynthesis and morphological measurements were taken in November, April, and June. Tillers per pot were counted in three pots and plant height and diameter were measured.

PAGE 48

48 Height:diameter ratio was calcula ted by dividing plant he ight by plant diameter Plant height was considered to be the average unextended leaf he ight of the plant. Plant basal diameter was measured along two perpendicular lines per pot (three pots per plot) that in tersected at the middle of the plant. The end of each line was the point at which the tiller emerged that was most distant from the center of the plant. The average of th ese two measurements of diameter per pot was used in the height:diameter ratio calculati on. Additionally, leaf photosynthesis was measured within 2 h of solar noon on cloud-free days with a portable leaf photosynt hesis system (LI-COR model LI-6200) on one leaf from two tillers pe r plot. The youngest fully expanded lamina from two tillers was clamped into a 0.25-L leaf chambe r, which was sealed with a rubber gasket. Area of leaf in the leaf chamber was determined by multiplying the length of the chamber by leaf width inside the chamber. Leaf area was measured with a leaf area me ter (LI-COR model LI-3100) on three plants per plot. Plants were cut to soil level and laminae were removed at the collar region and stored in plastic bags in the refrigerator prior to leaf area measurements. The plants used for leaf area measurement were the same ones used for de structive sampling described in Chapter 4. Statistical analyses were performed using Proc GLM of SAS (SAS Inst. Inc., 1996), and the LSMEANS procedure was used to compare tr eatment means. Daylength treatment was the main plot, and fertilizer treatment by genotype co mbinations the subplots. Data were analyzed with year in the model, and subsequent analys es were performed based on significance of year interactions. Significance was determined at P 0.05. Statistical analysis of leaf photosynthesis was conducted using leaf temperature and solar radiation as covariates.

PAGE 49

49 Results and Discussion Rainfall Rainfall in S eptember 2005 was four times the 30-yr average for Gainesville, FL, due to two hurricanes affecting the area within a few w eeks time (Figure 3-1). As a result of the unusually high rainfall, there was standing water in the plots and pots for several days after each event. Because this negatively impacted plant esta blishment and growth in late summer and early fall, overall vigor of all plants in Year 1 was considerably inferior to that of plants in Year 2. In general, rainfall was higher than average for most months of the first year of the experiment and less than average from March through June of the second year. To avoid loss of plants during severe droughts, irrigation was applied to plots on 22 Nov. 2004, 8 Jan., 8 Feb., and 19 Feb. 2005 at amounts of 15, 8, 13, and 8 mm, respectively. In Year 2, irrigation was applied on 17 Nov. 2005 and 4 May 2006 at amounts of 13 and 18 mm, respectively. Total-Season Herbage Yield Total-season herbage yield was determ ined by summing yields for seven harvest dates each year. Total-season herbage yield was aff ected by genotype, daylength, year, and genotype X year and daylength X year interactions, so data were analyzed by year (Table A-1). There was no genotype X daylength interaction ( P = 0.9) for total-season herbage yield, so both daylength sensitive and non-sensitive plants responded similarly to daylength in this study. When analyzed by year, there were no differences in total-season herbage yield in 2004-05 (Table A-2), likely due in part to stress imposed on all plants by excess moisture in fall 2004. Total-season herbage yield was affected by genotype and daylength in 2005-06. Lack of a fertilizer response was unexpected, but is likely a functi on of the time period of sampling, i.e., most harvests occurred during fall through spring when other factors may be more limiting to growth than soil nutrient status.

PAGE 50

50 Figure 3-1. Monthly rainfall data at the experimental site; average of 30-yr, 2004-05, and 200506. Cumulative trial period rainfall for th e 30-yr average, 2004-05, and 2005-06 was 909, 1399, and 845 mm, respectively. Total-season herbage yield of Tifton 9 and PC A Cycle 4 was greater than for Pensacola in 2005-06 (49 and 57% more forage than Pensacola respectively; Figure 3-2). The prostrate growth habit of Pensacola has been associated wi th a high percentage of foliage near soil level and below harvest cutting height (Beaty et al., 1968). Both PC A Cycle 4 (Blount et al., in review) and Tifton 9 (Werner and Burton, 1991) are upright-growing types. Tifton 9 was selected for increased forage yield above cu tting height and improve d seedling vigor from populations of Pensacola through recurrent restricted phenot ypic selection (RRPS; Burton, 1989). Werner and Burton (1991) obs erved Tifton 9 is taller and has more culms per plant and wider, longer leaves than Pensaco la, but the diameter of Tifton 9 plants is less than Pensacola. Chambliss (2003) reported Tifton 9 generally pr oduces 30 to 40% more harvested forage per year than Pensacola, with total annual yields in Gainesville, FL, of 12.5 Mg ha-1 and 10 Mg ha-1, 0 50 100 150 200 250 300 350 400 450 500 SepOctNovDecJanFebMarAprMayJun MonthRainfall (mm) 2004-05 2005-06 30-yr average

PAGE 51

51 respectively. In several studies at Ona and Im mokalee, FL, Tifton 9 yielded 16 to 34% more forage per year than Pensacola (Mislevy et al., 1991, 2005; Muchovej and Mullahey, 2000). In Athens, GA, Pedreira and Brown (1996a) reported Tifton 9 yielde d 12% more forage per year than Pensacola (8.7 and 7.8 Mg ha-1, respectively), and they suggested that the greater yields of Tifton 9 were due to a greater percentage of bi omass being present above cutting height, rather than greater total above-ground biomass yields. Figure 3-2. Bahiagrass total-season herbage yiel d as affected by genotype in 2004-05 and 200506. Means within year followed by the same letter do not differ by the LSMEANS test (P > 0.05). Data are means across two dayl engths, two fertilization amounts, and four replicates (n=16). Total-season herbage yield was affected by daylength in 2005-06 (Figure 3-3). Extended daylength resulted in 27% greater forage yield than ambient. In general, exposure to extended daylength results in an increase in the DM partitioned to leaves (Hay, 1990). Sinclair et al. (2001) reported 2-yr average Pensacola forage yields increased 196% under 15-h daylength (extended) compared to ambient daylength be tween October and March in Ona, FL. They attributed this mainly to great er leaf growth and a stimula tion of reproductive development, 0 10 20 30 40 50 60 701YearHerbage yield (g DM pot-1) Pensacola Tifton 9 PCA Cycle 4 2005-06 2004-05 a a a b a a

PAGE 52

52 resulting in more inflorescences and stems in th e harvested biomass. Newman et al. (2007) found that 2-yr average Argentine yields between Oct ober and April in South Florida and Puerto Rico increased 250 and 93%, respectively, under 15-h extended daylength compared to ambient daylength. Blount et al. (2001) reported 2-yr average Pensacola and Tifton 9 yields between September and January in Quincy, FL, increased 51 and 21%, respectively, under 15-h extended daylength compared to ambient. Similar re sults have been reported for switchgrass ( Panicum virgatum L.) yields under extended photoperiod. Van Esbroeck et al. (2004) reported average switchgrass yields between October and March in College Station, TX increased 49% under 18h extended daylength compared to ambient dayle ngth. They associated the increased DM yields under extended photoperiods with greater tiller weight and tiller number. Figure 3-3. Bahiagrass total-season herbage yiel d as affected by daylength in 2004-05 and 200506. Means within year followed by the same letter do not differ by the LSMEANS test (P > 0.05). Data are means across three genotypes, two fertilization amounts, and four replicates (n=24). 0 10 20 30 40 50 601YearHerbage yield (g DM pot-1) Ambient Extended 2005-06 2004-05 a a a b

PAGE 53

53 Seasonal Herbage Yield Herbage yield was determ ined for seven harv est dates during autumn through late spring for 2 yr. The first autumn harvest was not incl uded in the seasonal comparisons that follow. Herbage yields from harvests in November, la te December-early January, and February were summed and are reported as cool-season herbage yi eld. Herbage yields fro m harvests in April, May, and June were summed and are reported as warm-season herbage yield. Cooland warmseason herbage yields, collectively termed seasonal herbage yields, were compared in both years. Seasonal herbage yield was affected by ge notype, season, year, and genotype X year and daylength X year interactions, so data were anal yzed by year (Table A-3). There was no effect of genotype X daylength interaction ( P = 0.6) on seasonal herbage yield, which was somewhat surprising because of the reported differences in daylength responses among genotypes. There were no interactions involving season, and this also was somewhat surprising as it would be logical to expect that the relative ranking of photoperiod sensitive and less sensitive plants would be different during warm and c ool seasons. Yield responses to genotype X year interaction for the seasonal yield data were similar to those already reported for total-season yield and will not be presented. Greater yields ( P < 0.0001) were observed in the warm compared to cool season (19.7 and 7.3 g DM pot-1, respectively). These results were expected because bahiagrass is a C4 grass and produces 85 to 90% of total annual herbage yield during Apr il through September (Mislevy and Everett, 1981; Kalmbacher, 1997). Mislevy et al. (2005) reported 3-yr average Pensacola and Tifton 9 yields of 1.3 Mg ha-1 in April through June in Ona, FL, while 3-yr average yields from November through December were 0.5 Mg ha-1. Muchovej and Mullahey (2000) reported 2-yr average Pensacola and Tifton 9 yields of 0.7 Mg ha-1 in June in Immokalee, FL, while yields from November through December were 0.1 Mg ha-1. Mislevy et al. (1991) reported 3-yr average

PAGE 54

54 Pensacola and Tifton 9 yields of 9.2 Mg ha-1 in May to December in Ona, FL, while yields in December to May were 4.7 Mg ha-1. Herbage Nitrogen Herbage N concentration was determined for seven harvest dates for two years. Each year, N yield was calculated across harv ests and divided by total-season DM yield to give weighted total-season herbage N concentration. Weighted N concentrations also were calculated for the three cool-season harvests (Novemb er, late December-early Januar y, and February) and the three warm-season harvests (April, May, and June). As was the case for DM yield, there were similar trends for weighted total-season and seasonal N concentrations, so data will be presented and discussed together. Weighted total-season N concentration was affected by genotype, daylength, fertilization, and year (Table A-5). Weighted seasonal N concentration was affected by genotype, fertilization, year, and daylength X season interaction (Table A-6). Weighted total-season and seasonal N concentr ations were affected by genotype (Table 31). Pensacola had greater total-season and cool and warm-season N concentrations than Tifton 9 and PCA Cycle 4. PCA Cycle 4 had lesser total-season N concentra tion than the other genotypes. The more prostrate growth habit of Pensacola may have resulted in a lower proportion of low Ncontaining stems occurring above the cutting height, resulting in greater N concentration than in the more upright-growing Tifton 9 and PCA Cycl e 4. Pensacola had less total-season DM yield and consistently greater N con centration than Tifton 9 and PCA Cycle 4. As plant dry weight increases, the proportions of stru ctural and storage materials th at contain low N concentration increase, resulting in a decline in N concentration (N dilution eff ect) in the plant (Greenwood et al., 1990). Thus, lesser N concentration of Tifton 9 and PCA Cycle 4 can partially be accounted for by a dilution effect associated with greater DM harvested. In general, reported differences in N among bahiagrass cultivars have been mini mal. Mislevy et al. (1991) reported average

PAGE 55

55 Pensacola N concentration was 1.4 g kg-1 greater than for Tifton 9 across four grazing frequencies. Muchovej and Mullahey (2000) repo rted Pensacola had greater N concentration than Tifton 9 in July (13 and 12 g kg-1, respectively), but N concentrations were similar in June, August, September, October, November, and D ecember. They reported no differences in 2-yr mean N concentration between Pensacola and Tifton 9 (15.4 and 14.6 g kg-1, respectively). Mislevy et al. (2005) reported Pensacola had greater N con centration than Tift on 9 in April and June 1998 (29.8 and 23.2 g kg-1 in April, respectively, and 17.9 and 15.8 g kg-1 in June, respectively) and in April 1999 (23.4 and 21.6 g kg-1, respectively), but ther e were no differences in N concentration at any other evaluati on date in 1998, 1999, or 2000. Cuomo et al. (1996) reported no differences in N concentration between Pensacola and Tifton 9 (18.1 g kg-1 for both) averaged across three harvest frequencies. Blount et al. (in review) repor ted no differences in N concentration among PCA Cycle 4, Argentine, Pensacola, and Tifton 9 at Ona, FL, averaged across 2 yr (25.9, 23.4, 20.8, and 23.2 g kg-1, respectively). Table 3-1. Bahiagrass herbage weighted total-se ason and seasonal herbage N concentrations as affected by genotype. Data are means acr oss two daylengths, two fertilization amounts, two years, and four replicates (n=32). Season Genotype Total-season Cool Warm -------------------------------g N kg-1 ------------------------------Pensacola 22.1 a 24.2 a 22.6 a Tifton 9 20.6 b 23.2 b 21.0 b PCA Cycle 4 19.7 c 22.4 b 20.3 b SE 0.2 0.3 0.2 Means within a column followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Total-season weighted N concentrati on was also affected by daylength ( P < 0.0001) and was greater at ambient than extended daylength (22.8 and 19.8 g kg-1, respectively). Ambient daylength resulted in less totalseason DM yield in Year 2 and greater N concentration than

PAGE 56

56 extended daylength across years. The lesser N concentration associ ated with extended daylength can partially be accounted for by a dilution eff ect associated with greater DM harvested. Newman et al. (2007) reported bahiagrass N con centration was consistently lower (average decrease of 5.6 g kg-1) at three locations unde r extended than ambient daylength, which was attributed in part to a diluti on effect associated with greater DM harvested under extended daylength. Sinclair et al. (2003) reported decreased N concentrat ion of Pensacola under extended photoperiod that was associated with large incr eases in shoot growth (based on increased amounts of stem material with low N concentrat ion due to decreased le af:total ratio) and decreased relative ability to accu mulate N during the cool season. Weighted total-season and seasonal N concen tration were affected by fertilization ( P < 0.0001 and P < 0.0001, respectively). The high fertilizati on treatment resulted in greater totalseason and seasonal N concentration than low fertilization (total-season N of 21.4 and 20.2 g kg1, respectively, and seasonal N of 22.9 and 21.6 g kg-1, respectively). The response of herbage N to increasing N fertilizer has been well documented. Johnson et al. (2001) reported total N concentration in Pensacola forage increased linearly as N fertili zation increased from 0 to 157 kg N ha-1 cutting-1. Reiling et al. (2001) reporte d as N fertilization rate in creased, Pensacola forage N increased approximately 3.2 g kg-1 for every 39 kg N ha-1 cutting-1. Burton et al. (1997) reported bahiagrass forage N concentr ation increased from 10.6 to 17.0 g kg-1 as N fertilizer increased from 56 to 448 kg ha-1. Newman et al. (2006) reported Pensacola forage N increased from 9.3 to 12.6 g kg-1 as N fertilizer increased from 80 to 320 kg ha-1. Weighted total-season N concentr ation was affected by year ( P = 0.0004). Plants in 200405 had greater total-season N concentrati on than plants in 2005-06 (21.3 and 20.5 g kg-1, respectively). Total-season DM yields wa s less in 2004-05 than in 2005-06, while N

PAGE 57

57 concentrations were greate r in 2004-05 than in 2005-06. Th e relationship of lesser N concentration with greater DM yield has b een consistent throughout this experiment. Weighted seasonal N concentration was also affected by daylength X season interaction ( P < 0.0001), so data were analyzed by season. Both cooland warm-season N concentrations were greater at ambient than extende d daylength (Figure 3-4), but th e difference was greater in the cool season causing the interaction. The generally greater DM yield asso ciated with extended daylength may account for the consistently lesser seasonal N concentrations under extended than ambient daylength. Figure 3-4. Bahiagrass herbage weighted season al herbage N concentration as affected by daylength X season interaction. Means within a season followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Data are means across three genotypes, two fertilization amounts, two years, and four replicates (n=48). Tiller Number Tille r number was determined in November, April, and June in each of 2 yr. Tiller number was affected by evaluation date and year (Table A-7). 0 5 10 15 20 25 301SeasonWeighted nitrogen concentration (g N kg-1) Ambient Extended Warm Cool a b b a

PAGE 58

58 There were more tillers (P < 0.0001) in June than in November (68 and 42 tillers pot-1, respectively), with fewest tillers occurring in April (36 tillers pot-1). Greater yields were observed in the warm compared to cool season, and were associated with greater tiller number in June. Beaty et al. (1977) repor ted Pensacola had a maximum number of tillers in June (avg. 24 tillers dm-2), which gradually decreased until October (avg. 15 tillers dm-2). They reported an average of 11 tillers dm-2 in May. Bahiagrass produces 85 to 90% of total annual herbage yield during the warm season (Mislevy and Everett, 19 81; Kalmbacher, 1997) when tiller number is expected to be at its greatest. Lowest tiller number occurred in April, when plants were initiating spring growth after the winter. There were more tillers (P < 0.0001) in 2005-06 than in 2004-05 (78 and 19 tillers pot-1, respectively). Plants in 2005-06 we re more vigorous and had greater total-season yields than in 2004-05 (Figures 3-2 and 3-3) due to extremel y high rainfall in fall 2004 that was already described. This contributed to greater numbers of tillers per pot in Year 2. Leaf Area Leaf area was determ ined at the same times each year as tiller number. Leaf area was affected by evaluation date, year, and daylength X evaluation date, genotype X year, and daylength X year interactions, so data were analyzed by year for genotype and daylength responses (Table A-8). Leaf area was affected by daylength X evaluati on date interaction (T able 3-2). June leaf area was greater than those in both November and April for both ambient and extended daylengths. Hay and Heide (1983) reported extended daylength resu lted in up to a 323% increase in leaf area of Kentucky bluegrass ( Poa pratensis L.), which was associated with increased leaf blade and sheath length. t Mannetj e and Pritchard (1974) reported leaf area decreased as total light energy decreased from 14 to 11 h in several tropical grasses. Growth rates for bahiagrass

PAGE 59

59 are greater in summer than in the fall and spring, t hus it is expected that leaf area in June would be greater than in November a nd April. November and April leaf areas were similar within a daylength since bahiagrass growth rates are slower at these ti mes of year. There were no differences in leaf area between daylengths in November. Leaf area wa s greater under ambient than extended daylength in Apr il, while leaf area was greater under extended than ambient daylength in June. The reason for the opposite pattern of res ponse is not well understood. One possible explanation is that am bient daylength plants grew more vigorously in spring following shorter days in winter. Table 3-2. Bahiagrass leaf area as affected by da ylength X evaluation date interaction. Data are means across three genotypes, two ferti lization amounts, two years, and four replicates (n=48). Evaluation date Daylength November April June -------------------------------cm2 pot-1--------------------------------Ambient 379 a B 419 a B 493 b A Extended 368 a B 332 b B 590 a A SE 23 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Leaf area was affected by genotype in 2004-05, but there were no differences in 2005-06 (Table A-9; Table 3-3). Pensacola had greater leaf area than PC A Cycle 4 in 2004-05, with leaf area of Tifton 9 being intermediate. The grea ter leaf area of Pensaco la in 2004-05 was not associated with differences in he rbage yield (Figures 3-2 and 3-4) Forage yield was measured at an 8-cm stubble height, while leaf area was meas ured by clipping herbage at soil level. Although Tifton 9 has longer, wider leaves than Pensaco la (Werner and Burton, 1991) the upright growth habit of Tifton 9 results in more herbage above cutting height, while Pensacola has more herbage below cutting height (Beaty et al., 1968; Pedreira and Brown, 1996b) that is not included in the measurement of yield but was included in le af area. Pedreira and Brown (1996a) reported

PAGE 60

60 Pensacola had greater leaf ar ea index (LAI) than Tifton 9 (a vg. 1.53 and 0.77, respectively) 1 d after mowing, which was attributed to the more prostrate growth habit of Pensacola. In the current study, leaf area was great er for Pensacola in 2004-05 when harvested yield was similar for the three genotypes. In 2005-06, there were no differences among cultivars in leaf area probably because yields of Tifton 9 and PCA Cycle 4 were sufficiently greater than Pensacola to compensate for the differences in verti cal distribution of leaf in the canopy. Table 3-3. Bahiagrass leaf area as affected by genotype in 2004-05 and 2005-06. Data are means across two daylengths, two fertilization amounts, and four replicates (n=16). Year Genotype 2004-05 2005-06 ------------------------cm2 pot-1 -----------------------Pensacola 347 a 522 a Tifton 9 306 ab 571 a PCA Cycle 4 268 b 566 a SE 17 27 Means within a column followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Leaf area was affected by daylength in 200405 but not in 2005-06 (Table 3-4). Leaf area was greater under ambient than extended daylen gth in 2004-05. These results are inconsistent with other results published. t Mannetje and Pritchard (1974) re ported leaf area decreased as total light energy decreased from 14 to 11 h in several tropical grasse s. Hay and Heide (1983) reported extended daylength resulted in up to 323% increase in leaf area of Kentucky bluegrass, which was associated with increas ed leaf blade and sheath length. Plant Height and Diameter Plant height and diam eter were measured in November, April, and June for 2 yr. Plant height was affected by dayle ngth, genotype, evaluation date, ye ar, and daylength X evaluation date, daylength X year, and genot ype X year interactions, so da ta were analyzed by year to assess genotype and daylength responses (Table A-10). Plant height was greater under extended

PAGE 61

61 than ambient daylength in November and April, but was similar for both daylengths in June (Table 3-5). Plant height was greatest in June for both daylengths, which is expected for C4 grasses. Sinclair et al. (2004) reported a linear in crease in height of Pensacola as PPFD increased at low PPFD. Swanton et al. (2000) repor ted shoot height of barnyardgrass [ Echinochloa crusgalli (L.) Beauv.] increased with increasing photoperiod from 8 to 16 h. Table 3-4. Bahiagrass leaf area as affected by daylength in 2004-05 and 2005-06. Data are means across three genotypes, two fertilization amounts, and four replicates (n=24). Year Daylength 2004-05 2005-06 ------------------------cm2 pot-1-----------------------Ambient 332 a 529 a Extended 283 b 577 a SE 14 22 Means within a column followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Table 3-5. Bahiagrass plant height as affected by daylength X evaluation date interaction and plant diameter as affected by genotype X evaluation date interaction. Plant height data are means across three genotypes, two fe rtilization amounts, two years, and four replicates (n=48) and plant diameter da ta are means across two daylengths, two fertilization amounts, two years, and four replicates (n = 24). Evaluation date Daylength November April June ------------------------------height (cm) -------------------------------Ambient 15.2 b B 14.7 b B 22.6 a A Extended 19.4 a B 16.1 a C 22.3 a A SE 0.5 Genotype ----------------------------diameter (cm) -----------------------------Pensacola 13.7 a C 15.7 a B 17.9 a A Tifton 9 13.1 a C 14.7 b B 17.8 a A PCA Cycle 4 13.1 a B 13.8 c B 17.8 a A SE 0.3 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05).

PAGE 62

62 Plant height was affected by daylength X y ear interaction. Interaction occurred because there was no effect of dayle ngth on plant height in 2005-06 ( P > 0.05), while plant height was greater under extended than ambient dayle ngth (18.6 and 15.8 cm, respectively) in 2004-05 (Table A-11). Greater plant height under extended daylength is c onsistent with results reported by Sinclair et al. (2004) and Swanton et al. (2000). Plant height was also affected by genotype X year interaction. Tifton 9 and PCA Cycle 4 had similar plant height in 200405, which was greater than Pensaco la (Table A-11; Table 3-6). In 2005-06, PCA Cycle 4 had tallest plants, while Pensacola had s hortest plants. This supports the observation that growth habit differences exist among these genotypes and both Tifton 9 and PCA Cycle 4 are more erect in habit than Pens acola. Werner and Burton (1991) reported Tifton 9 was taller than Pensacola (104 and 81 cm, resp ectively) when measured from the ground to top of the tallest inflorescence. However, Pedrei ra and Brown (1996b) repor ted Pensacola and Tifton 9 had similar plant heights (23 and 29 cm, respec tively) 84 d after plantin g. PCA Cycle 4 is more similar in growth habit to Tifton 9 than to Pensac ola, and these data indica te that when grown as individual plants it is as upright or even more upri ght growing than Tifton 9. Table 3-6. Bahiagrass plant height as aff ected by genotype in 2004-05 and 2005-06. Data are means across two daylengths, two fertiliza tion amounts, and four replicates (n=16). Year Genotype 2004-05 2005-06 ------------------------cm -----------------------Pensacola 15.3 b 16.8 c Tifton 9 18.4 a 20.0 b PCA Cycle 4 17.8 a 22.1 a SE 0.4 0.4 Means within a column followed by the same letter do not differ by the LSMEANS test ( P > 0.05).

PAGE 63

63 Plant diameter was affected by genotype, eval uation date, year, and genotype X evaluation date interaction (Table A-12). Plant diameter was greater in 2005-06 than in 2004-05 (18.5 and 12.0 cm, respectively), which is consistent with observations of greater plant vigor in 2005-06. Plant diameter was also affected by genotype X evaluation date interaction (Table A-12; Table 3-5). There were no differe nces in plant diameter among genotypes in November or June, while Pensacola had greatest diameter and PCA Cycle 4 had lesser diameter in April. Plant diameter was greatest in June for all genotypes. Pensacola and Tifton 9 had lesser plant diameter in November than in April, while PCA Cycle 4 diameter was similar in November and April. Werner and Burton (1991) reported Pensacola was greater in basal diameter than Tifton 9 (55 and 51 cm, respectively). Height:diameter ratio was affected by genot ype, daylength, evaluation date, year, and genotype X evaluation date, daylength X evaluation date, genotype X year and daylength X year interactions, so data were analyzed by year to assess genotype and daylength responses (Table A-13). Height:diameter ratio was affected by genotype X evaluation date and daylength X evaluation date interactions in 2004-05 (Table A-14) and by daylength X evaluation date interaction in 2005-06 (Table A-14), so data were analyzed by evaluation date for genotype and daylength responses. Height:diameter ratio was affected by genot ype at all evaluation dates for both years (Table 3-7). Pensacola had lesser height:diameter ratio than all genotypes at all dates, with the exception that it was similar to PCA Cycle 4 in November 2004-05. The prostrate growth habit of Pensacola contributed to its lesser height:diameter ratio, while the upright growth habit of Tifton 9 and PCA Cycle 4 contribut ed to their greater ratio. Wern er and Burton (1991) reported Tifton 9 was taller than Pensacola but lesser in diameter. Calculated height:diameter ratio for

PAGE 64

64 Table 3-7. Bahiagrass height:diameter ratio as af fected by genotype in November, April, and June 2004-05 and 2005-06. Data are means acr oss two daylengths, two fertilization amounts, and four replicates (n=16). Evaluation date November April June November April June Year Genotype 2004-05 2005-06 --------ratio height:diame ter ---------------ratio height:diameter --------Pensacola 1.4 b 1.1 b 1.2 b 0.9 b 0.7 c 1.1 c Tifton 9 2.0 a 1.4 a 1.4 a 1.1 a 0.9 b 1.2 b PCA Cycle 4 1.7 ab 1.5 a 1.5 a 1.2 a 1.1 a 1.4 a SE 0.1 0.1 0.1 0.04 0.04 0.1 Means within a column followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Tifton 9 in the Werner and Burton (1991) st udy was greater than Pensacola (2.0 and 1.5, respectively). Pedreira and Brown (1996b) reported 2-yr average height:diameter ratios of Tifton 9 that were greater than Pe nsacola (1.5 and 0.6, respectively) Tifton 9 and PCA Cycle 4 had similar height:diameter ratios for all dates except April and June 200506, when PCA Cycle 4 had greater height:diameter ratios than Tifton 9. PCA Cycle 4 is more similar in growth habit to Tifton 9 than to Pensacola, and these data indicate that when grown as in dividual plants it is as upright or even more upright growing than Tifton 9. Height:diameter ratio was also affected by daylength in Nove mber of both years and June 2005-06 (Table 3-8). Extended daylength resulted in greater height:diameter ratio than ambient daylength in November. However, in June 2005-06, height:diameter ratio was greater under ambient than extended daylength. Sinclair et al. (2004) reported a linear in crease in height of Pensacola as PPFD increased at low PPFD. Swant on et al. (2000) reported shoot height of barnyardgrass increased with incr easing photoperiod from 8 to 16 h.

PAGE 65

65 Table 3-8. Bahiagrass height:diameter ratio as af fected by daylength in November, April, and June 2004-05 and 2005-06. Data are means acr oss three genotypes, two fertilization amounts, and four replicates (n=24). Evaluation date November April June November April June Year Daylength 2004-05 2005-06 ------ratio height:diame ter ------------ratio height:diameter ------Ambient 1.4 b 1.3 a 1.3 a 1.0 b 0.9 a 1.3 a Extended 1.9 a 1.4 a 1.4 a 1.2 a 0.9 b 1.1 b SE 0.1 0.1 0.04 0.03 0.03 0.04 Means within a column followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Leaf Photosynthesis Leaf photosynthesis was m easured in Nove mber, April, and June each year. Leaf photosynthesis was affected by genotype, evaluation date, and year (Table A-15). Pensacola and Tifton 9 had similar leaf phot osynthesis rates (23.1 and 22.7 mol CO2 m-2 s-1, respectively), which were greater th an PCA Cycle 4 (20.7 mol CO2 m-2 s-1). Pensacola had greater totalseason leaf N concentration than Tifton 9 and PCA Cycle 4 (22.1, 20.6, and 19.7 g kg-1, respectively). The lower leaf N of PCA Cycle 4 may have contribute d to its lesser leaf photosynthesis rate. Bolton and Brown (1980) and Taub and Lerdau (2000) reported net photosynthesis increased linearly as leaf N concentration increased. This relationship may result from the large proportion of leaf protei n accounted for by ribulose-1,5-bisphosphate carboxylase/oxygenase enzyme and the regulation of photosynthesis by this enzyme (Bolton and Brown, 1980). Pedreira and Brown (1996b) reported Pensacola had greater leaf photosynthesis than Tifton 9 (27 and 24 mol CO2 m-2 s-1, respectively), but had le sser dry weight 110 and 127 d after planting (19 and 25 g, respectively). They attributed Pensacolas greater photosynthesis to its smaller leaf laminae than Tifton 9 (Wer ner and Burton, 1991). Bhagsari and Brown (1986)

PAGE 66

66 reported a negative relationship between leaf la mina area and leaf photosynthesis for several species, but the explanation for this relationshi p is unknown. In this experiment, leaf length and width were not measured, but PC A Cycle 4s growth habit and a ppearance were similar to Tifton 9. Its presumed longer, wider leaves than Pe nsacola may explain in part the lower leaf photosynthesis rate of PCA Cycle 4 compared to Pensacola. Leaf photosynthesis rate was greatest in April (25.5), intermediate in November (23.0), and least in June (18.0 mol CO2 m-2 s-1). Leaf N concentration followed the same trend as leaf photosynthesis and was great est in April (23.5 g kg-1), intermediate in November (21.9 g kg-1), and least in June (19.6 g kg-1). Leaf photosynthesis rate was greater in 2004-05 (27.0 mol CO2 m-2 s-1) than in 2005-06 (17.4 mol CO2 m-2 s-1). Based on total-season weighted N concentration, plants in 2004-05 had greater leaf N (21.3 g kg-1) than plants in 2005-06 (20.5 g kg-1). Summary and Conclusions The photoperiod sensitivity of ba hiagrass has been described by several authors (Blount et al., 2001; Sinclair et al., 2001; Sinclair et al., 2003; Sinclair et al., 2004; N ewman et al., 2007). They found that bahiagrass yields could be increased during shortdaylength months by artificially extending daylength. Selection fo r bahiagrass genotypes with less daylength sensitivity and greater cool-season productivity has occurred (Blount et al., 2001). One such genotype is PCA Cycle 4. There are no published st udies that characterize the differences in growth, physiological, and mor phological responses between da ylength-sensitive and less sensitive types under ambient light, nor are there studies that show diffe rences in response to extended daylength among these genotypes. This in formation is needed to assess the inherent differences between daylength-sensitive and less sensitive types that may affect their performance in production systems under ambien t light conditions and to provide a better

PAGE 67

67 understanding of their responses to daylength. The objectives of this study were to determine the effects of daylength and level of fertilization on total-season a nd seasonal herbage yield, tissue N concentration, leaf area and phot osynthesis, tiller number, and height:diameter ratio of two diploid bahiagrass cultivars (Tifton 9 and Pensac ola) and a cold adapte d genotype that is less sensitive to photoperiod (PCA Cycle 4). Contrary to what was expected, there were no interactions of genotype with daylength. Thus, under the conditions of this study daylength-sensitive a nd less sensitive types responded similarly to daylength. The primary factors affec ting responses measured were the main effects of genotype and daylength and interactions of year with these factors. Year interactions occurred primarily because excessive ra in associated with two hurricanes during September 2004 had a negative impact on plant vigor that carried over throughout the remainder of the 2004-05 experimental year. Pensacola had lesser forage yields (sampled to an 8-cm stubble) than Tifton 9 and PCA Cycle 4 in Year 2, but had greater total-season and seasonal herbage N concentrations. Tifton 9 and PCA Cycle 4 had similar fora ge yields and seasonal herbage N concentrations. The generally lesser herbage N concentrations of Tifton 9 and PCA Cycle 4 are likely explained by a dilution of N associated with their greater yield. Leaf photosynthesis was greatest for Pensacola and least for PCA Cycle 4, and photosynt hesis generally followed similar trends to herbage N concentration. Pensacola had grea ter (Year 1) or similar (Year 2) leaf area than the other genotypes, so there must be factors other than leaf area and leaf photosynthesis responsible for the lower yield of Pensacola. Pensacola demonstrated a more prostrate grow th habit that resulted in generally lesser height:diameter ratios than the more upright growing Tifton 9 and PCA Cycle 4, which had

PAGE 68

68 similar height:diameter ratios and growth hab its. This more prostrate growth habit likely contributed to lower Pensacola yields because a greater amount of biomass would be below clipping stubble height (8 cm). This supports the observation of Beaty et al. (1968) that the prostrate growth habit of Pensacola was associated with a high percentage of foliage near the ground and below cutting height. Pedreira a nd Brown (1996a) also reported Pensacola had greater LAI 1 d after mowing than Tifton 9 and attributed it to its more prostrate growth habit. The prostrate growth habit of Pensacola may be associated with a greater degree of self-shading of leaves lower in the canopy relative to more upright-growing Tifton 9 and PCA Cycle 4. Photosynthesis was measured on leaves in full sun, so the greater the degree of self-shading in the canopy the less representative the photosynthesis measurement is of the average leaf in the canopy. Thus, the combination of more herbage be low harvest height a nd greater self-shading likely contributed to the lower yi eld of Pensacola (Year 2) in sp ite of greater leaf photosynthesis and greater or similar leaf area than Tifton 9 and PCA Cycle 4. In addition to explaining the yield response, th e growth habit of Pens acola suggests that it may depend less on reserve carbohydrates to regr ow after defoliation th an Tifton 9 and PCA Cycle 4 due to its potentially great er residual leaf area. This charac teristic of Pensacola may be a factor in its long-term persistence un der continuous stocking and close grazing. Extended daylength resulted in greater bahiagrass herbage yi elds than ambient daylength, regardless of genotype. This response is consistent with reports from other authors (Blount et al., 2001; Sinclair et al., 2001; Sinc lair et al., 2003; Sinclair et al., 2004; Newman et al., 2007) studying daylength-sensitive types. Associated with greater herbage yields under extended daylength were lesser total-season and seasonal herbage N concentrations.

PAGE 69

69 Based on this research, responses of PCA Cycle 4 space plants to daylength are not measurably different than those of two diploi d cultivars, Tifton 9 and Pensacola. PCA Cycle 4 has similar morphological character istics as Tifton 9 and is cons istently more upright growing than Pensacola. Herbage N concentration and leaf photosynthesis rates were lower for PCA Cycle 4 than Pensacola, but PCA Cycle 4 yield was greater than or similar to Pensacola. The more erect growth habit of PCA Cycle 4 likely pl ayed a role in this be cause a greater proportion of herbage was above the clipping stubble height and these plants were less prone to selfshading. Extrapolating from these data, it can be concluded that management of PCA Cycle 4 will likely be more critical than for Pensacola bahiagrass in Florida. Its erect growth habit implies that more intensive management practices such as rotational stoc king and greater control of stocking rate may be warranted to ma intain the persistence of PCA Cycle 4.

PAGE 70

70 CHAPTER 4 BAHIAGRASS COMPONENT MASS, NITR OGEN, AND TOT AL NONSTRUCTURAL CARBOHYDRATE RESPONSES TO DAYLENGTH IN ESTABLISHMENT YEAR Introduction Bahiagrass ( Paspalum notatum Flgge) is the prim ary past ure grass for beef cattle ( Bos sp.) and horses ( Equus caballis ) in Florida and is grown on more than one million ha (Chambliss and Adjei, 2006). It is a warm-season perennial that in Florida produces 85 to 90% of total annual herbage yield during April through September (Mislevy a nd Everett, 1981; Kalmbacher, 1997). Commonly used bahiagrass genotypes such as Pensacola and Tifton 9 are sensitive to short daylengths, which induce a change of growth pattern, re sulting in a reduction in aboveground growth during the cool season. Plants that exhibit this respons e typically continue to photosynthesize, but instead of using photosynthate to produce new above-ground tissue, they are thought to store photosynthate as nonstructural carbohydrates in stem bases, stolons, and rhizomes (Thornton et al., 2000). Reserve storage can occur when leaf area is large and provides photosynthate in excess of growth and maintenance needs, or when plant growth rate is low (i.e., in the autumn for subtropical grasses) or is constrained by resource av ailability (e.g., water or other nutrients), allowing plants to effectively prepare for winter survival and early spring regrowth (Thornton et al., 2000). Frequent, severe defoliation may reduce the mass of storage organs which may in turn affect long-term pasture persistence (Ortega-S. et al., 1992; Chaparro et al., 1996). Root growth is generally reduced by defoliation as a result of the reduction of photos ynthetically active tissue and prioritization to shoot growth of a limite d supply of carbohydrate (R ichards, 1993). Lower cutting height or more frequent defoliation may result in a reduction in root weight and carbohydrate reserves (Youngner, 1972). Beaty et al. (1970) reported ha rvest frequency had little effect on Pensacola forage yield, but root and st olon mass were 50 to 75% greater for 6than for

PAGE 71

71 1-wk harvests. Gates et al. (1999) reported greater spring reserves (as estimated by total etiolated initial spring growth) in bahiagrass plots cut th e previous growing season every 8 wk than in those cut every 2 or 4 wk, while cutting height di d not affect spring reserves. Chaparro et al. (1996) reported that frequent, close defoliation of Mott elephantgrass (Pennisetum pupureum Schum.) resulted in reduced lig ht interception, rhizome mass, rhizome total nonstructural carbohydrate (TNC) and N reserves, and number of tillers per plant. Pensacola bahiagrass crown had an average TNC concentration of 121 g kg-1 while Argentine averaged 97 g TNC kg-1 when grazed every 2, 4, 6, and 8 wk in Ona, FL (Adjei et al., 1989). After 3 yr of grazing every 2 wk, bahiagrass crown TNC during August to earl y September averaged less than 90 g kg-1, which was 40 g kg-1 lower than the overall average for the 3-, 5-, and 7-wk grazing frequencies (Mislevy et al., 1991). The significant economic implications of s easonal bahiagrass forage shortfall have stimulated research aimed at in creasing productivity during shortdaylength months. This effort has involved genetic selection and development of bahiagrass cultivars that are cold adapted and less sensitive to photoperiod (Blount et al., 2001). Additional cool-season productivity must not come at the expense of persistence, however, because the cost of past ure renovation is high and return per unit land area from grasslandlivestock systems are relatively low. There are no known studies comparing chemical composition and mass responses of bahiagrass storage organs to extended daylengt h for existing cultivar s with those of less photoperiod sensitive, cold-adapted bahiagrass genot ypes. This information is needed to explore physiological and morphological attributes that accompany increased yield of the less photoperiod sensitive genotypes, particularly those at tributes that relate to persistence in the field. Therefore, the objectives of this study were to determine the effect s of daylength and level

PAGE 72

72 of fertilization on the mass of above-ground plant components (leaves, stem bases, and inflorescences) and the mass and N and TNC concen trations of storage organs (roots, rhizomes, and stem bases) of two diploid bahiagrass cult ivars and a diploid cold -adapted, less photoperiod sensitive genotype. Materials and Methods Experimental Site This study was conducted in pots in the field at the Beef Research Unit, northeast of Gainesville, FL, at 29 N latitude. Maxim um da ylength (20 June) from sunrise to sunset in Gainesville is about 14.45 h (Nautical Almanac, 2003). Single plants were planted in individual plastic tree pots (20-cm diameter by 46-cm depth) that were inserted into augered holes of the same dimensions in the field plots. Pots were filled by volume with well-mixed, uniform, screened topsoil. A double layer of landscaping cl oth was placed in the bottom of the pots to minimize root escape and soil loss. Pots were replanted at the beginni ng of Year 2. At the beginning of the experiment in Year 1, soil had an average pH of 7.0. Average Mehlich-I extractable soil P, K, Mg, and Ca con centrations were >122, 13, 97, and >3100 mg kg-1, respectively. At initiation of the second year of the experime nt, soil had an average pH of 6.9. Average Mehlich-I extractable soil P, K, M g, and Ca concentrations were >190, 68, 138, and >3780 mg kg-1, respectively. Treatments and Design Treatm ents were the factorial combinations of two daylengths, two fertilizer levels, and three bahiagrass genotypes. They were arranged in four replications of a completely randomized design. Sexual diploid bahiagrass genotypes evaluated were Pensacola, Tifton 9, and PCA (photoperiod and cold-adapted) Cycle 4. Pensacola and Tifton 9 are commonly utilized

PAGE 73

73 bahiagrass cultivars in Florida, while PCA Cycle 4 is a novel genotype sel ected to improve cold tolerance, photoperiod response, nematode a nd disease resistance, rooting/rhizome mass, seedling vigor and establishment, seasonal distri bution of forage production, and forage quality (Blount et al., 2001). PCA Cycle 4 has been approved for release by the University of Florida Agricultural Experiment Station as cv. UF-Riata (Blount et al., in review). Daylength treatments were ambient and exte nded. Extended daylength was 15 h and was achieved using 12 quartz-halogen lamps (1500 W, 240 V). Three lamps spaced 3.7 m apart were mounted 2 m above the soil on posts at each of the long sides of the daylength treatment main plots (9.6 m x 8.6 m each) for a total of six lamps in each main plot. There were a total of two banks of lamps, each bank accommodating two ex tended daylength main plots. The lamp fixtures were positioned to shin e light across the main plot tr eatment from two opposing sides. The average photosynthetic photon flux density (PPFD ) measured after sunset at plant level in the extended daylength treatment was 24 mol m-2 s-1. Ambient daylength plots were located 5 m away from extended daylength plots. No PPFD wa s measured at night in the ambient treatment while the lamps were in use. Timers were used to control when the lamps were on. The lamps were activated 30 min before sunset and turned off when the total photoperiod had reached 15 h from sunrise. Timers were adjusted as the season progressed to achieve the 15-h photoperiod. Fertilization treatments were termed hi gh and low based on N and K amounts applied. Fertilization treatments were in cluded in the experiment because some previous work looking at Pensacola bahiagrass responses to daylength was conducted using high levels of fertilizer input (Sinclair et al., 2003), and it is no t clear the extent to which this impacted treatment response. Treatments consisted of distinct warm and cool -season fertilizer amounts. In the warm-season, the high treatment plots received 60 kg N and 45 kg K ha-1 harvest-1 and the low treatment

PAGE 74

74 received 20 kg N and 15 kg K ha-1 harvest-1. In the cool-season, high tr eatment plots received 30 kg N and 22.5 kg K ha-1 harvest-1 and low treatment plots r eceived 10 kg N and 7.5 kg K ha-1 harvest-1. Total annual fert ilizer amounts during the trial were 330 kg N and 248 kg K ha-1 yr-1 for the high treatment and 110 kg N and 83 kg K ha-1 yr-1 for the low treatment. The IFAS recommended fertilizer amounts for bahiagrass grown only for hay are 90 kg N and 37 kg K ha-1 applied after each cutting (Mylavarapu et al., 2007). Plots consisted of 12 spaced plants in pots, one pl ant per pot, inserted into the soil in a 6 x 2 arrangement. Pots were spaced 0.4-m apart within rows and 0.5-m between rows for a plot size of 1.6 x 4.3 m. The plants used were grown from s eed that was sown in flats in the greenhouse in March. After seedlings had at least two leaves (approximately 4-cm tall), they were moved to individual speedling flats in May and grown until they were approximately 12-cm tall. They were then planted in the field in pots in July 2004. The same procedure was used for the plants used for Year 2 of the study. At the end of the first year of the experiment in June 2005, all remaining plants were removed fr om pots and the soil returned to the pots. All pots received the equivalent of 50 kg N ha-1 in early August of each year. Pl ots were staged on 27 Aug. 2004 and 5 Sept. 2005. At staging, all plants were cut to an 8-cm stubble height, and fertilizer and daylength treatments were imposed. Plots were subseque ntly harvested every 5 wk through November. During the cool season that followed, plots were harvested every 7 wk through April. Following the April harvest, plots were then harveste d every 5 wk through June. The experiment was conducted from 27 Aug. 2004 to 17 June 2005 (Year 1) and 5 Sept. 2005 to 26 June 2006 (Year 2). Destructive sampling occurred three times per y ear in November, April, and June to evaluate the following response variables.

PAGE 75

75 Response Variables Response variables m easured include mass a nd N and TNC concentrations of roots, rhizomes, and stem bases, and mass of leaves a nd inflorescences. Herbage yield was determined every 5 wk during warm-season months and 7 wk during cool-season months by clipping herbage to an 8-cm stubble height from five of the 12 pots per plot (Chapter 3). Remaining pots not designated for destructive sampling were also clipped to 8 cm and material was discarded. Destructive sampling was conducted in November, April, and June on three pots per plot. Each plot consisted initially of 12 plants, so this allowed sufficient intact plants for each forage harvest. At each destructive sampling date, th e pots to be sampled were removed from the ground and herbage was clipped to soil level. La minae were removed at the collar region to separate herbage into leaf and stem base com ponents. Any inflorescences present were removed and placed into an inflorescence component. Root s plus rhizomes were separated from soil by washing with water over a 1-cm2 mesh screen. Roots were clipped from the rhizomes. All components were dried at 60 C to constant weight. Below-:a bove-ground ratio was calculated by dividing the sum of root + rhizome dry weight components by the sum of leaf + stem base + inflorescence dry weight components. Dried samples were ground in a Wiley mill (Model 4 Thomas-Wiley Laboratory Mill, Thomas Scientific, Swedeboro, NJ), first throug h a 4-mm screen and finally to pass a 1-mm screen prior to analyses. To reduce the overall nu mber of samples for analyses and because it seemed likely that rhizomes and stem bases were the major reserve storage organs (Adjei et al., 1989), these components were ground together (rhiz ome + stem base). The root component was ground separately.

PAGE 76

76 Root and rhizome + stem base samples (1296 to tal from two experiments) were analyzed for N and TNC concentration by near-infrared reflectance spectroscopy (N IRS) calibrated with wet chemistry data from 10% of the samples an alyzed using the methods of Gallaher et al. (1975) and Hambleton (1977) for N and a modified procedure described by Christiansen et al. (1988) for TNC. For N analysis, samples were di gested using a modification of the aluminum block digestion procedure of Gallaher et al. (1975 ). Sample weight was 0.25 g, catalyst used was 1.5 g of 9:1 K2SO4:CuSO4, and digestion was conducted for at least 4 h at 375C using 6 ml of H2SO4 and 2 ml H2O2. Nitrogen in the digestate was dete rmined by semiautomated colorimetry (Hambleton, 1977). For TNC analysis, sample s were incubated with invertase and amyloglucosidase to hydrolyze sucrose and starch, respectively, and then resultant total hexoses were analyzed by a reducing sugar assay. Spectral data were collect ed on all samples, an average of 32 scans for each sample, with a NIR Syst ems 6500 spectrophotometer (Foss Int., Laurel, MD) equipped with a static sample cup. Principle component analysis wa s conducted on the spectral da ta and on a subset selected for calibration using the > Select = procedure of the software InfraSoft In ternational (ISI, State College, PA) based on spectral dissimilarity of samples (Schenk and Westerhaus, 1991a). Reference laboratory data for N and TNC were compared with the spectral data for the calibration samples and equations were develope d with the ISI software using partial least squares regression (Schenk and Westerhaus, 1991b). The N mean, standard error of validation, and r2 for the equation used were: 10.6 g kg-1, 0.9 g kg-1, and 0.85, respectively. The TNC mean, standard error of validation, and r2 for the equation used were: 52.2 g kg-1, 13.4 g kg-1, and 0.86, respectively. These equations were then used to predict N and TNC for all samples, including those used for the calibration.

PAGE 77

77 Statistical analyses were performed using Proc GLM of SAS (SAS Inst. Inc., 1996), and the LSMEANS procedure was used to compare tr eatment means. Daylength treatment was the main plot, and fertilizer treatment by genotype co mbinations the subplots. Data were analyzed with year in the model, and subsequent analyses were performed based on significance of year in interactions. Significance was determined at P 0.05. Results and Discussion Rainfall Rainfall in S eptember 2004 was four times the 30-yr average for Gainesville, FL, due to two hurricanes affecting the area within a few w eeks time (Figure 4-1). As a result of the unusually high rainfall, there was standing water in the plots and pots for se veral days after each Figure 4-1. Monthly rainfall data at the experimental site; average of 30-yr, 2004-05, and 200506. Cumulative trial period rainfall for 30-yr average, 2004-05, and 2005-06 were 909, 1399, and 845 mm, respectively. 0 50 100 150 200 250 300 350 400 450 500 SepOctNovDecJanFebMarAprMayJun MonthRainfall (mm) 2004-05 2005-06 30-yr average

PAGE 78

78 event. Because this negatively impacted plant esta blishment and growth in late summer and early fall, overall vigor of all plants in Year 1 was considerably inferior to that of plants in Year 2. In general, rainfall was higher than average for most months of the first year of the experiment and less than average from March through June of the second year. To avoid loss of plants during severe droughts, irrigation was applied to plots on 22 Nov. 2004, 8 Jan., 8 Feb., and 19 Feb. 2005 at amounts of 15, 8, 13, and 8 mm, respectively. In Year 2, irrigation was applied on 17 Nov. 2005 and 4 May 2006 at amounts of 13 and 18 mm, respectively. Allocation of Dry Matter to Plant Parts Above-ground component mass Leaf mass. Leaf m ass was affected by evaluation date, year, and daylength X evaluation date and genotype X year interactions (Table B-1). There was no effect of genotype X daylength interaction ( P = 0.5) on leaf mass, thus leaf mass of both daylength sensitive and less sensitive genotypes responded similarly to daylength. Leaf mass was affected by daylength X evalua tion date interacti on (Table 4-1). Unlike what was anticipated, there was no leaf mass response to daylengt h in November or April. In June, leaf mass was greater under extended than normal daylength. Leaf mass was greater in June than in November or April for both da ylengths, a result that was expected because bahiagrass is a C4 grass and produces 85 to 90% of tota l annual herbage yield during April through September (Mislevy and Everett, 1981; Kalmbacher, 1997). November and April leaf masses were similar within a daylength since bahiag rass growth rates are slow er at these times of year. Cool temperatures in November and Apri l may have inhibited growth responses to the extended daylength treatment during these months, but this has not been the case in several prior studies. Blount et al. (2001) reported extended da ylength increased the to p growth of Pensacola in September, October, and January by 13, 69, and 67%, respectively, compared to ambient

PAGE 79

79 daylength in Quincy, FL. In the same study, top growth of Tifton 9 was not affected by daylength in September, but was increased by 46 and 29% in October an d January, respectively, under extended daylength. Blount et al. (in review) reported extende d daylength increased forage yield of Pensacola and Tifton 9 by 75 and 50%, respectively, compared to ambient daylength in October 1999. Sinclair et al. (2003) reported extended photoper iod increased the yield of Pensacola bahiagrass during the s hort-daylength months by as much as six fold in January 1999, while there was no effect of daylength treatment on yields during the long -daylength months at Ona, FL. Table 4-1. Bahiagrass leaf mass as affected by da ylength X evaluation date interaction. Data are means across three genotypes, two ferti lization amounts, two years, and four replicates (n = 48). Evaluation date Daylength November April June -------------------------------g DM pot-1 ------------------------------Ambient 5.6 a B 6.1 a B 8.6 b A Extended 5.1 a B 5.8 a B 10.1 a A SE 0.3 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Leaf mass was also affected by genotype X year interaction, so data were analyzed by year to study the genotype effect (Table 4-2). In 2004-05, Pensacola and Tifton 9 had greater leaf mass than PCA Cycle 4, while in 2005-06, Tifton 9 and PCA Cycle 4 had greater leaf mass than Pensacola (Table B-2). Leaf mass was greater in 2005-06 for all genotypes due to excessive rainfall and less vigorous plant development during autumn 2004-05. In related research, Blount et al. (2001) reported average t op growth of Tifton 9 was 52% greater than Pensacola, while Pedreira and Brown (1996b) reporte d Tifton 9 dry weight was greater than Pensacola (25 and 19 g plant-1, respectively) when cut to 3-cm stubble height. Werner and Burton (1991) reported Tifton 9 leaf weight was greater than Pensacola (670 and 220 g, re spectively) when cut to 4-cm

PAGE 80

80 stubble height. Thus, Tifton 9 having the greatest or similar to the greatest leaf mass in the current study is consistent with previous studi es. In contrast, in 2004-05, when excessive rainfall occurred in autumn, Pensacola did better than expected and PCA Cycle 4 did relatively worse than in 2005-06. It is not known if there are diffe rences in flooding tole rance between these two cultivars that led to this effect or if some other factor caused the genotype X year interaction. Stem base mass. Stem base mass was affected by da ylength, genotype, evaluation date, year, and genotype X year interaction, so data we re analyzed by year for genotype effect (Table B-1). There was no effect of ge notype X daylength interaction ( P = 0.6) on stem base mass. Stem base mass was affected by daylength ( P = 0.02) and evaluation date ( P < 0.0001). Stem base mass was greater under extended th an ambient daylength (4.4 and 3.9 g DM pot-1, respectively). Stem base mass was greatest in June (6.0), and was greater in November (3.6) than in April (2.9 g DM pot-1). Growth rates for bahiagrass are greater in summer than in the fall and spring, thus it is expected that stem base mass in June would be greater than in November and April. Stem base mass was also affected by genotype X year interaction (Table 4-2). There were no differences in stem base mass in 2004-05 (T able B-3), while in 2005-06, PCA Cycle 4 had greatest stem base mass and Tift on 9 was greater than Pensacola. Inflorescence mass. Inflorescence mass was affected by daylength, evaluation date, year, and genotype X year interaction, so data were analyzed by year fo r genotype effect (Table B-1). There was no effect of genot ype X daylength interaction ( P = 0.4) on inflorescence mass. Inflorescence mass was affected by daylength ( P = 0.005) and evaluation date ( P < 0.0001). Inflorescence mass was greater under extended than ambient daylength (3.1 and 2.5 g DM pot-1, respectively). Sinclair et al. (2001) reported reproductive development of Pensacola

PAGE 81

81Table 4-2. Bahiagrass component mass as affected by genotype in 2004-05 and 2005-06. Data are means across two daylengths, two fertilization amounts, three evaluation da tes, and four replicates (n = 48). Bahiagrass component Leaf Stem base Inflorescence Root Rhizome Genotype 2004-05 2005-06 2004-05 2005-06 2004-05 2005-06 2004-05 2005-06 2004-05 2005-06 ---------------------------------------------------------------g DM pot-1 -----------------------------------------------------------Pensacola 4.7 a 8.7 b 2.8 a 4.2 c 2.6 a 3.3 a 20.6 a 46.6 ab 23.9 a 87.3 a Tifton 9 4.5 a 9.8 a 3.2 a 5.5 b 1.8 a 5.2 a 19.0 a 51.4 a 19.2 b 91.2 a PCA Cycle 4 3.7 b 10.0 a 3.0 a 6.5 a 1.7 a 4.9 a 16.8 a 40.7 b 17.8 b 85.7 a SE 0.2 0.4 0.1 0.3 0.4 1.2 1.7 2.9 1.2 4.0 Genotype X year interaction ( P < 0.05) Means within a column followed by the same letter do not differ by the LSMEANS test ( P > 0.05).

PAGE 82

82 bahiagrass was stimulated under extended photoper iod so that more inflorescences and stems were components of the harvested mass in that treatment. Inflorescence mass in the current study was greatest in June (7.2) and similar in November and April (1.4 and 0.7 g DM pot-1, respectively). Greater inflorescence mass under ex tended daylength and in June is expected because Pensacola bahiagrass is a long-day plant th at flowers when daylength is 14 h or greater (Nada, 1980). Inflorescence mass was also affected by genot ype X year interaction (Table 4-2). When analyzed by year, however, there were no differe nces in inflorescence mass among genotypes for either year (Table B-4). In 2004-05, there were no inflorescences collected for any genotype in November or April, while all but six plots in June had inflorescences In 2005-06, over 50% of plots contained inflorescences in November and April, and all pl ots contained inflorescences in June. Marousky et al. (1991) reported Pensacola bahiagrass produced inflorescences earlier in the growing season and in greater number than Tifton 9, but in the current study there were no differences among genotypes. Below-ground component mass Root mass. Root m ass was affected by genotype, ev aluation date, year, and daylength X evaluation date (Table B-1). There was no e ffect of genotype X daylength interaction ( P = 0.9) on root mass. For consistency of data presenta tion with other components, the root mass of genotypes is presented by year in Table 4-2. Ac ross years, Tifton 9 had greater root mass than PCA Cycle 4 (35.2 and 28.7, respective ly), with the root mass of Pensacola intermediate (33.6 g DM pot-1). Gates et al. (1999) reporte d Pensacola bahiagrass produced a greater root mass than Tifton 9 (33.6 and 21.4 g, respectively). Pedreira and Brown (1996b) repo rted Pensacola had a greater percentage of dry wei ght in roots than Tifton 9 (40 and 27%, respectively). Root mass was greater in 2005-06 than in 2004-05 (46.2 and 18.8 g DM pot-1, respectively) in the current

PAGE 83

83 study, which was consistent with observations of inferior overall vigor of plants in 2004-05 with respect to unusually high rainfall. Root mass was also affected by daylength X ev aluation date interact ion (Table 4-3). As with leaf mass, there were no differences in root mass between daylengths in November or April. In June, root mass was greater under extended than normal dayl ength. Sinclair et al. (2003) reported no difference in Pensacola bahiagra ss below-ground tissue mass throughout the season (September to June) between extended photoperiod and ambient daylength. In the current study, root mass was greater in June th an in November or April for bot h daylengths, likel y a function of the fact that these plants were continuing to grow and occupy more of the pot volume as the experiment progressed. In addition, growth rates for bahiagrass are gr eater in summer than in the fall and spring, and this explains the increase in mass from April to June. Table 4-3. Bahiagrass root mass as affected by daylength X evaluation date interaction. Data are means across three genotypes, two ferti lization amounts, two years, and four replicates (n = 48). Evaluation date Daylength November April June -------------------------------g DM pot-1 ------------------------------Ambient 27.6 a B 24.7 a B 39.7 b A Extended 24.7 a B 28.6 a B 49.7 a A SE 2.5 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Rhizome mass. Rhizome mass was affected by evaluation date ( P < 0.0001) and year (P < 0.0001; Table B-1). There was no effect of genotype X daylength interaction ( P = 0.6) on rhizome mass. For consistency of data presentati on with other component s, the rhizome mass of genotypes is presented by year in Table 4-2. Th ere was no effect of genotype in the full model (including year), but when data were analyzed by year Pensacola had greatest rhizome mass in 2004-05 but there were no differences in 2005-06.

PAGE 84

84 Rhizome mass was greatest in June (83.9), and rhizome mass in April was greater than in November (43.6 and 35.1 g DM pot-1, respectively). Weather condi tions favor greater rhizome mass in summer than in the fall, and plants were more fully established by April than November, This was likely responsible for greater rhizome mass in April than November. Rhizome mass was greater in 2005-06 than in 2004-05 (88.1 and 20.3 g DM pot-1, respectively), which was consistent with observations of inferior overall vigor of plants in 200405 with respect to unu sually high rainfall. Below-:Above-ground Ratio Below-:above-ground ratio was affected by da ylength, genotype, evaluation date, year, and daylength X evaluation date, dayl ength X year, and genotype X year interactions, so data were analyzed by year (Table B-5). There was no effect of genotype X daylength interaction ( P = 0.6) on below-:above-ground ratio. Below-:above-ground ratio was affected by geno type in 2005-06 (Table B-6; Table 4-4), with Pensacola having the greatest ratio and Ti fton 9 having a ratio greater than PCA Cycle 4, indicating proportionally more DM allocated to the belowth an above-ground component in Pensacola than in Tifton 9 and PCA Cycle 4. There were no differences ( P = 0.2) in the ratio among genotypes in 2004-05. Gates et al. (1999) reported Pensaco la produced a greater root mass than Tifton 9 (33.6 and 21.4 g, respectively), but less mean forage yields (6.5 and 7.7 Mg ha-1, respectively). Pedreira and Brown (1996a) reported similar stubble (leaf + stem below cutting height) + rhizome weights for Pensacola and Tifton 9 (1000 and 810 g m-2, respectively), but Pensacola produced less mean forage yield than Tifton 9 (7.8 and 8.7 Mg ha-1, respectively). Pedreira and Brown (1996b) reported a trend toward a greater percen tage of dry weight in roots for Pensacola vs. Tifton 9 ( 40 and 27%, respectively).

PAGE 85

85 Table 4-4. Bahiagrass below-:above-ground rati o as affected by genotype in 2004-05 and 200506. Data are means across two fertilization am ounts and four replicates for each year (n = 8). Year Genotype 2004-05 2005-06 ------------ratio below-: above ground ------------Pensacola 5.6 a 8.7 a Tifton 9 5.0 a 7.4 b PCA Cycle 4 5.7 a 6.3 c SE 0.3 0.3 Means within year followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Below-:above-ground ratio was al so affected by daylength ( P < 0.0001) in 2005-06, with greater below-:above-ground rati o under ambient than extended daylength (8.3 and 6.7, respectively), indicating proporti onally more DM allocated to the belowthan above-ground component under ambient than extended da ylength. There were no differences (P = 0.5) in the ratio between daylength treatments in 2004-05. Below-:above-ground ratio was affected by dayl ength X evaluation date interaction (Table 4-5). There were no differences in below-:abov e-ground ratio between daylength treatments in April or June, while the ratio for ambient was greater than extended daylength in November. This response suggests that in au tumn varying daylength causes plants to alter DM allocation. Specifically, extended days result in greater allocation to above-ground plant parts. This preferential allocation could result in inferior growth rates in sp ring or reduced persistence. In the companion study reported in Chapter 3, how ever, there was no evidence that daylength affected subsequent cool-season yields, and Ta ble 4-5 shows that below-:above-ground ratio differences were no longer present in April or June. Below-:above-ground ratio for the ambient daylength treatment was greater in November and April than in June indicating a greater proportion of DM allocated to the belowthan above-ground component in the cool season as

PAGE 86

86 well as low herbage growth rates. Extended da ylength below-:above-ground ratio was greater in April than both November and June, a reflecti on of the lesser proportion of DM allocated belowground in November in this treatment and greater top growth in June. Table 4-5. Bahiagrass below-:above-ground ratio as affected by daylength X evaluation date interaction. Data are means across three genotypes, two fertilization amounts, two years, and four replicates (n = 48). Evaluation date Daylength November April June -----------------ratio below-:above-ground -----------------Ambient 7.2 a A 7.3 a A 5.8 a B Extended 5.5 b B 7.2 a A 5.7 a B SE 0.3 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Reserve Pools Nitrogen and TNC conc entration and content we re measured on 1) root and 2) rhizome + stem base components. Nitrogen concentration Neither root N nor rhizom e + st em base N concentrations were affected by genotype or interactions with genotype. Root N concentr ation was affected by daylength, fertilization, evaluation date, and year (Table B-7). Root N concentration was greater ( P = 0.046) under ambient than extended daylength (9.7 and 9.4 g kg-1, respectively), and was greater ( P = 0.0002) under high than low ferti lization (9.8 and 9.3 g kg-1, respectively). The response to fertilization was expected based on observations by Beaty et al. (1975), in which th e N concentration of Pensacola bahiagrass roots increased with increa sing N fertilization amount s. Roots appear to have accumulated more N under ambient daylength. When differences in root biomass occurred, biomass was greater under extended daylength (Table 4-3), and this likely contributed to lower N concentration in roots of this treatment. Sincla ir et al. (2003) reported no difference in Pensacola

PAGE 87

87 bahiagrass root N concentration between extend ed and ambient photoperiods from September to June. In the current study, root N concentration was greatest ( P < 0.0001) in April (10.2), and June root N concentration was great er than November (9.5 and 9.0 g kg-1, respectively). This trend is similar to that shown by Sinclair et al. (2003) in which Pensacola bahiagrass root N concentration increased from September to Febr uary (extended photoperiod) or April (ambient daylength), and decreased through June. Root N concentration was greater ( P < 0.0001) in 200405, when root mass was less, than 2005-06 (9.9 and 9.2 g kg-1, respectively). Rhizome + stem base N concentration was a ffected by daylength, fe rtilization, evaluation date, year, and fertilization X ev aluation date and fertilization X year interactions (Table B-7). The responses of rhizome + stem base N con centration followed similar trends as root N concentration. Rhizome + stem base N concentration was greater ( P < 0.0001) under ambient than extended daylength (12.4 and 12.0 g kg-1, respectively), and was greater ( P < 0.0001) under high than low fertilization (12.4 and 11.9 g kg-1, respectively). Stem base biomass was greater under extended than ambient daylength (4.4 and 3.9 g DM pot-1, respectively), and this likely contributed to lower N concentration in rhizome + stem base component of this treatment. In the current study, rhizome + stem base N concentration was greatest ( P < 0.0001) in April (12.7), and November rhizome + stem base N concentr ation was greater than June (12.1 and 11.7 g kg-1, respectively). Rhizome and stem base biomass was greatest in June, and this likely contributed to lower N concentration in the rh izome + stem base component in June. Rhizome + stem base N concentration was greater ( P < 0.0001) in 2004-05, when rhizom e + stem base mass was less, than in 2005-06 (12.8 and 11.5 g kg-1, respectively). Total nonstructural carbohydrate concentration Root TNC concentration was affected by evalua tion date (Table B-8) There was no effect of genotype X daylength interaction (P = 0.2) on root TNC concentration. Root TNC

PAGE 88

88 concentration was greatest in April (46.1 g kg-1) and was similar in November and June (32.8 and 33.8 g kg-1, respectively), which may be associated with greater leaf phot osynthesis in April but continuing slow top growth (Chapter 3). These seasonal pa tterns, especially the lower concentrations in November than April, suggest that roots are not a primary reserve storage structure for bahiagrass. Rhizome + stem base TNC concentration was affected by genotype, fertilization, evaluation date, year, and dayle ngth X evaluation date, genotype X evaluation date, genotype X year interactions (Table B-8), so data were analyzed by year to assess genotype effects. Daylength X evaluation date and genotype X evaluation date inte ractions are presented across years. There was no effect of ge notype X daylength interaction (P = 0.1) on rhizome + stem base TNC concentration. Rhizome + stem base TNC concentration wa s affected by daylength X evaluation date interaction (Table 4-6). In November, rhizom e + stem base TNC concentration for ambient daylength plants was greater than extended dayl ength, while in April extended rhizome + stem base TNC concentration was greater than ambi ent daylength. Greater TNC concentration under ambient daylength in November was accompanie d by greater allocation of mass to below-ground Table 4-6. Bahiagrass rhizome + stem base to tal nonstructural carbohydrate concentration as affected by daylength X evaluation date interaction. Data are means across three genotypes, two fertilization amounts, two y ears, and four replicates (n = 48). Evaluation date Daylength November April June -------------------------------g TNC kg-1 ------------------------------Ambient 144 a A 104 b C 115 a B Extended 131 b A 120 a B 121 a B SE 3 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05).

PAGE 89

89 plant parts (Table 4-5). Both res ponses suggest greater pr ioritization of reserve storage for plants growing under ambient than extended daylength. There were no differences in rhizome + stem base TNC concentration between daylengths in June. Rhizome + stem base TNC concentrations were greatest in November for both ambien t and extended daylengths, suggesting their importance as storage organs for bahiagrass. Under ambient daylength, rhizome + stem base TNC concentration was less in April than in June likely because spring growth depleted reserves which were then being restored as the growing season progresse d. There were no differences in rhizome + stem base TNC concentration betw een April and June under extended daylength. Sinclair et al. (2003) reported no differences in Pensacola bahiagrass root TNC concentration between extended and ambient photoperiods from September to June. Rhizome + stem base TNC concentration was also affected by genotype X evaluation date interaction (Table 4-7). Pensacola had greate r rhizome + stem base TNC concentration in November than the other genotypes, while Tifton 9 and PCA Cycle 4 had similar rhizome + stem base TNC concentrations. Tifton 9 had greater rhizome + stem base TNC concentration than PCA Cycle 4 in April, with Pensacola intermed iate. PCA Cycle 4s lesse r leaf photosynthetic rate (Chapter 3) may be associated with its lesser rhizome + stem base TNC concentration in November and April. There were no differences in rhizome + stem base TNC concentration among genotypes in June. Rhizome + stem base TN C concentrations were greater in November for all genotypes than in April or June. April a nd June rhizome + stem base TNC concentrations were similar for both Pensacola and Tifton 9. PCA Cycle 4 had lowest rhizome + stem base TNC concentration in April. As previously not ed, the rhizome + stem base component was accumulating TNC in November, especially for Pensacola, and concentrations were depleted significantly by April and June. This trend is consistent with the tr end reported by Sinclair et al.

PAGE 90

90 (2003) in which Pensacola below-ground TNC c oncentration decreased from September to December through April, and increased again in June At the onset of the cool season, bahiagrass likely stored TNC in the rhizome + stem base com ponent to prepare for winter survival and early spring regrowth (Thornton et al., 2000); during these periods TNC concentration decreased for all genotypes. Table 4-7. Bahiagrass rhizome + stem base to tal nonstructural carbohydrate concentration as affected by genotype X evaluation date interaction. Data are means across two daylengths, two fertilization amounts, two years, and four replicates (n = 32). Evaluation date Genotype November April June -------------------------------g TNC kg-1 ------------------------------Pensacola 153 a A 114 ab B 113 a B Tifton 9 131 b A 118 a B 115 a B PCA Cycle 4 128 b A 104 b C 116 a B SE 4 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). When analyzed by year, rhizome + stem base TNC concentration was affected by genotype in both years (Table B-9; Figure 4-2). In 2004-05 Pensacola had greater rhizome + stem base TNC concentration than Tifton 9 and PCA Cy cle 4. In 2005-06, Pensacola and Tifton 9 had greater rhizome + stem base TNC concentratio ns than PCA Cycle 4. PCA Cycle 4 appears to accumulate lesser TNC concentration in the rhiz ome + stem base component than Pensacola, likely associated with its lesser daylength sensitivity. Adjei et al. (1989) reported average crown TNC concentration of 115 g kg-1 for Pensacola bahiagrass stockpiled from December to May in Ona, FL. Nitrogen content Root N content was affected by genotype evaluation date, year, and daylength X evaluation date interaction (Table B-10). Th ere was no effec t of genotype X daylength

PAGE 91

91 Figure 4-2. Bahiagrass rhizome + stem base to tal nonstructural carbohydrate concentration as affected by genotype in 2004-05 and 2005-06. Means within a year followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Data are means across two daylengths, two fertilization amounts, and f our replicates for each year (n = 16). interaction ( P = 0.9) on root N content. Pensacola and Ti fton 9 had greater root N content than PCA Cycle 4 (Table 4-8). Differences among genotypes were due to root mass as N concentration was not affected by ge notype. Root N content was greater ( P < 0.0001) in 2005-06 than in 2004-05 (0.4 and 0.2 g pot-1, respectively), which was cons istent with observations of inferior overall vigor of plan ts in 2004-05 with respect to unusually high rainfall. Root N content was also affected by daylengt h X evaluation date inte raction (Table 4-9). There were no differences in root N content betw een daylengths in November or April, while root N content was greater under extended than am bient daylength in June. Root N content was greater in June than in November and April under both ambient and extended daylengths and was primarily a function of seas onal differences in root mass. 0 20 40 60 80 100 120 1401YearTotal nonstructural carbohydrate (g TNC kg-1) Pensacola Tifton 9 PCA Cycle 4 2005-06 2004-05 a b b a b a

PAGE 92

92 Table 4-8. Bahiagrass root and rhizome + st em base nitrogen and total nonstructural carbohydrate (TNC) content as affected by genotype. Data are means across two daylengths, two fertilization amounts, two year s, and four replicates for each column (n = 32). N Content TNC Content Genotype Root Rhizome + stem base Root Rhizome + stem base ----------g N pot-1 -------------------g TNC pot-1 ----------Pensacola 0.31 a 0.68 a 1.13 ab 7.30 a Tifton 9 0.32 a 0.69 a 1.23 a 7.16 a PCA Cycle 4 0.26 b 0.65 a 0.99 b 6.32 b SE 0.15 0.02 0.07 0.30 Genotype main effect ( P < 0.05). Genotype main effect ( P > 0.05). Means within a column followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Table 4-9. Bahiagrass root nitrogen content as affected by daylength X evaluation date interaction. Data are means across three genotypes, two fertilization amounts, two years, and four replicates (n = 48). Evaluation date Daylength November April June -------------------------------g N pot-1 ------------------------------Ambient 0.24 a B 0.23 a B 0.38 b A Extended 0.21 a B 0.26 a B 0.47 a A SE 0.02 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Rhizome + stem base N content was affected by fertilization, evalua tion date, year, and fertilization X year interaction (Table B-10). There was no effect of genotype ( P = 0.5) or genotype X daylength interaction ( P = 0.7) on rhizome + stem ba se N content. Increased fertilization resulted in greater rhizome + stem base N content ( P = 0.01; 0.71 and 0.64 g pot-1, for high and low respectively). Consistent with obse rvations of inferior over all vigor of plants in 2004-05, rhizome + stem base N content was greater ( P < 0.0001) in 2005-06 than 2004-05 (1.06 and 0.29 g pot-1, respectively). Rhizome + stem base N content was greatest (P < 0.0001) in June (1.02), and was greater in April than November (0.55 and 0.45 g pot-1, respectively).

PAGE 93

93 Total nonstructural carbohydrate content Root TNC c ontent was affected genotype, eval uation date, and year (Table B-11). There was no effect of genotype X daylength interaction ( P = 0.6) on root TNC content. Tifton 9 had greater root TNC content than PCA Cycle 4, with Pensacola intermediate (Table 4-8), and these responses were driven primarily by root mass differences among genotypes. The responses of root TNC content to evaluation date and year fo llowed the same trend as rhizome + stem base N content. Root TNC content was greatest (P < 0.0001) in June (1.5 g pot-1), and April root TNC content was greater than November (1.0 and 0.8 g pot-1, respectively), and th ese responses were driven primarily by root mass differences among ev aluation dates. Consistent with observations of inferior overall vigor of plants in 2004-05, root TNC content was greater ( P < 0.0001) in 2005-06 than 2004-05 (1.7 and 0.5 g pot-1, respectively). Rhizome + stem base TNC content was affect ed by genotype, evaluation date, year, and daylength X evaluation date in teraction, but there was no eff ect of genotype X daylength interaction ( P = 0.9; Table B-11). Pensacola and Tifton 9 had 13 to 16% greater rhizome + stem base TNC content than PCA Cycl e 4 (Table 4-8), suggesting that PCA Cycle 4 has less stored energy for regrowth than Tifton 9 and Pensacola. Rhizome + stem base TNC content was greater ( P < 0.0001) in 2005-06 than 2004-05 (11.0 and 2.9 g pot-1, respectively), due to inferior overall vigor of plants in 2004-05. Rhizome + stem base TNC content was also affected by daylength X evaluation date interaction (Table 4-10). In November, rhiz ome + stem base TNC content was greater under ambient than extended daylength, illustrating th e reduced priority for storage under extended days. However, in June, rhizome + stem base TNC content was greater under extended than ambient daylength, a function of greater growth for the extended day plants. There were no differences in rhizome + stem base TNC content between daylengths in April. Rhizome + stem

PAGE 94

94 base TNC content was greater in June than in November and April under both ambient and extended daylengths. Table 4-10. Bahiagrass rhizome + stem base tota l nonstructural carbohydrate content as affected by daylength X evaluation date interacti on. Data are means across three genotypes, two fertilization amounts, two years, and four replicates (n = 48). Evaluation date Daylength November April June -------------------------------g TNC pot-1 -------------------------------Ambient 5.7 a B 5.1 a B 9.8 b A Extended 4.4 b B 5.5 a B 11.1 a A SE 0.4 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Summary and Conclusions The photoperiod sensitivity of ba hiagrass has been described by several authors (Blount et al., 2001; Sinclair et al., 2001; Sinclair et al., 2003; Sinclair et al., 2004; N ewman et al., 2007). Bahiagrass herbage yield was observed to decline markedly with short days even when temperature and soil moisture were adequate to support greater yields. These authors found that bahiagrass yields could be in creased during short-daylength months by artificially extending daylength. Selection for bahiagra ss genotypes with less daylength sensitivity and greater coolseason productivity has occurred (Blount et al., 2001). One such genotype is PCA Cycle 4. There are no published studies that char acterize the differences in st orage organ mass and N and TNC content responses between daylen gth-sensitive and less sensitive types under ambient light, nor are there studies that show differences in res ponse to extended daylength among these genotypes. This information is needed to assess the inhere nt differences between daylength-sensitive and less sensitive types that may a ffect their performance in production systems under ambient light conditions and to provide a better understanding of their responses to daylength. The objectives of this study were to determ ine the effects of daylength a nd level of fertilization on plant

PAGE 95

95 component mass, below-:above-ground ratio, and N and TNC concentration and content of two diploid bahiagrass cultivars (Tifton 9 and Pensacola) and a cold adapted, less photoperiod sensitive genotype (PCA Cycle 4). Contrary to what was expected, there were no interactions of genotype with daylength. Thus, under the conditions of this study daylength-sensitive a nd less sensitive types responded similarly to daylength. The primary factors affec ting responses measured were the main effects of genotype, daylength, and evaluati on date and interactions of y ear with these factors. Year interactions occurred primarily because excessive rain associ ated with two hurricanes during September 2004 had a negative impact on plant vi gor that carried over throughout the remainder of the 2004-05 experimental year. Fertilization level had very limited impact in this study, primarily affecting those responses that included N concentration as a component. PCA Cycle 4 had greatest stem base mass in 2005-06, but had less root mass than Tifton 9 in 2005-06 and less rhizome mass than Pensacola in 2004-05. PCA Cycle 4 also had less below:above-ground ratio than the other genotypes in 2005-06, indica ting proportionally less DM allocated to the belowthan above-ground comp onent in PCA Cycle 4 than in Pensacola and Tifton 9. PCA Cycle 4 had less root N content than Pe nsacola and Tifton 9 due to less root mass. Pensacola and Tifton 9 had 13 to 16% greater rhizome + stem base TNC content than PCA Cycle 4 associated with lesser leaf photosynthe tic rate and less DM allocated below ground proportionally for PCA Cycle 4 than Pensacola and Tifton 9. These results suggest that PCA Cycle 4 stored less energy and N for regrowth than Pensacola and Tifton 9. There were seasonal patterns in TNC concentr ation in roots and rhizomes + stem bases. Root TNC concentration was greater in Apr il than in November and June. Lower TNC

PAGE 96

96 concentration of roots in November than April suggests that roots are not the primary reserve storage structure for bahiagrass. Conversely, rhizome + stem base TNC concentration was greater in November than in June and April, suggesting their importance as reserve storage organs for bahiagrass. The responses of storage organ mass and TNC concentration and c ontent to daylength were of particular interest in November. The below-:above-ground ra tio was greater under ambient than extended daylength in November, indicating that altering daylength in autumn causes plants to alter DM allocation. Specificall y, proportionally more DM was allocated to the belowthan above-ground plant component under ambient than extended daylength. The preferential allocation of DM to above-ground plant parts under extended daylength could result in inferior growth rates in spring or reduced persistence. However, there was no evidence that daylength affected subsequent cool-season yields (Chapter 3). Associated with greater allocation of DM to below-ground plant parts under extended daylength was greater rhizome + stem base TN C concentration and content under ambient than extended daylength in November. These response s suggest greater prio ritization of reserve storage for plants growing under ambient than extended daylength. From these data, it can be concluded that management of PCA Cycle 4 will likely be more critical than for Pensacola bahiagrass in Florida. It s preferential allocation of DM to aboveground plant parts over below-ground storage structures as well as its lesser stored N and TNC for regrowth than Pensacola and Tifton 9 implies that more intensive management practices such as rotational stocking and greater control of stocking rate may be warranted to maintain the persistence of PCA Cycle 4.

PAGE 97

97 CHAPTER 5 YIELD, NUTRITIVE VALUE, AND PERSIST ENCE RESPONSES OF BAHIAGRASS GENOTYPES TO DEFOLIATION MANAGEMENT Introduction Bahiagrass ( Paspalum notatum Flgge) is the prim ary past ure grass for beef cattle ( Bos sp.) and horses ( Equus caballis ) in Florida and is grown on more than one million ha (Chambliss and Adjei, 2006). It produces 85 to 90% of a nnual herbage yield during April through September (Mislevy and Everett, 1981; Ka lmbacher, 1997). Bahiagrass yields less than other commonlyused perennial warm-season grasses in Florida, with annual yields ranging from 4 to 12 Mg ha-1 (Muchovej and Mullahey, 2000; Mislevy et al., 2005). Chambliss (2003) reported dry matter (DM) yield of Pensacola and Argentine ba hiagrasses to be approximately 10 Mg DM ha-1, but Tifton 9 generally produces 30 to 40% more fora ge per year than Pensac ola. Annual yields of Tifton 9 and Pensacola in Gainesville, FL were 12.5 and 10 Mg ha-1, respectively (Chambliss, 2003). Seasonally, diploid bahiagrass es generally produce more earlyand late-seaso n herbage in northern Florida than tetraploid cultivars like Ar gentine that initiate grow th later in the spring (Chambliss and Adjei, 2006). Frequent clipping generally re duces total DM production of forage plants, while crude protein (CP) and digestibility may be increase d with increasing harvest frequency (Prine and Burton, 1956; Burton et al., 1963). Beat y et al. (1963) repor ted total yield of Pensacola increased as interval between harvests increased from 1 to 6 wk. This trend of increasing herbage accumulation with decreasing harvest frequency was also reported for Argentine, Pensacola, and Tifton 9 (Cuomo et al., 1996) in Louisiana and Pensacola and Tifton 9 in Florida and Georgia (Mislevy et al., 1991; Gates et al ., 1999). However, Adjei et al (1989) reported no effect of grazing frequency (2, 4, 6, and 8 wk) on DM yields of Argentine and Pensacola in South Florida.

PAGE 98

98 Harvesting to a short stubble height also tends to reduce total DM production of many forage plants. Burns et al. (2002) reported that more intensiv e defoliation of tall fescue ( Festuca arundinacea Schreb.) generally resulted in lesser annual DM yields. Published data with bahiagrass indicate that it may respond differently to stubble height. Gates et al. (1999) indicated that total DM yields of Pensaco la and Tifton 9 generally were greater at 1.5than 10-cm stubble height. Similarly, Pedreira and Brown (1996a) re ported that annual yields of Pensacola and Tifton 9 were greater at 3.5 than 10 cm. Similarly, Mislevy and Everett (1981) found that total DM yields of Pensacola and Arge ntine were greater at 5than 10-cm stubble, while Beaty et al. (1968) reported total DM yields of Pensacola generally increase d as stubble height decreased across six heights. In Japan, Hirata (1993a) obser ved a seasonal response pattern of Pensacola DM yield to stubble height. In th e summer, greater yields were obtained at 22than at 2-cm stubble height, while in the spring and autumn, the reverse was observe d. These data indicate that bahiagrass yields generally are not negatively affected by close defoliation. Defoliation also affects nutritive value. Forage grasses in general show a decrease in CP and in vitro digestible organic matter (IVDOM) con centrations with advancing maturity (Ball et al., 2001), primarily due to reproductive stem elongation (Coleman et al., 2004). Harvest frequency and season of growth play major roles in determining forage quality. Forage quality tends to decline as forages regrow due to accumulation of stems and deposition of lignin in leaves and stems (Adesogan et al., 2006). Fora ge regrowth in the summer may have lower quality due to increased lignin deposition associat ed with high temperatures, and in Florida due to increased growth rates and maturation associat ed with high rainfall (Adesogan et al., 2006). Brown and Mislevy (1988) reporte d summer yields of Pensacola we re greater than spring yields, but CP and IVDOM were lower. This scenario was due to higher temperatures and more

PAGE 99

99 favorable soil moisture conditions in the summer that promoted more rapid growth and maturity (Chambliss and Sollenberger, 1991). Arthington and Brown (2003) also re ported that increases in Pensacola maturity (i.e., 4vs. 10-wk regrow th) resulted in decrease d CP and digestibility. Chaparro and Sollenberger (1997) reported frequent defoliation to short stubble height resulted in greatest IVDOM of Mott elephantgrass ( Pennisetum purpureum Schum.), while herbage harvested infrequently to short stubble was least digestible. When defoliated frequently, most of the harvested material of elephantgrass consisted of leaf blad e and less leaf sheath and stem (Chaparro et al., 1995). When considering defoliation effects on forage plant persistence, frequent, severe defoliation may have negative effects on the mass of reserve storage organs including stem bases, rhizomes, stolons, and root s. Root growth is generally re duced by defoliation as a result of the reduction of photosynthetically active tissue and shortage of carbohydrates for root growth, with lower cutting height or more frequent defoli ation resulting in a grea ter reduction in root weight (Youngner, 1972). Karl and Doescher (1991) showed that defoliation by cattle resulted in a reduction of relative number of root s of grazed vs. ungrazed orchardgrass ( Dactylis glomerata L. Potomac). Beaty et al. ( 1970) reported harvest frequency had little effect on Pensacola forage yield, but root and stolon masses were 50 to 75% greater for 6-wk than for 1-wk harvest intervals. Gates et al. (1999) reported greater spring reserves (as estimated by total etiolated initial spring growth) in bahiagrass plots cut th e previous growing season every 8 wk than in those cut every 2 or 4 wk, while cutting height did not affect spring reserves. Gates et al. (1999) also found that root mass was greater for Pe nsacola vs. Tifton 9 (33.9 vs. 21.4 g per 75by 75mm soil core), but neither cutting height (1.5 or 10 cm) nor cutting interval (2, 4, or 8 wk) affected root mass.

PAGE 100

100 Carbohydrate reserves are generally reduced by defoliation (Youngner, 1972). Chaparro et al. (1996) reported that sustained frequent, close defoliation of M ott elephantgrass resulted in reduced light interception, rhizome mass, rhiz ome total nonstructural carbohydrate (TNC) and N reserves, and number of tillers per plant. Pens acola had an average TNC concentration of 121 g kg-1 while Argentine averaged 97 g TNC kg-1 when grazed every 2, 4, 6, and 8 wk at Ona, FL (Adjei et al., 1989). After 3 yr of grazing ev ery 2 wk, bahiagrass crown TNC during August to early September averaged less than 90 g kg-1, which was 40 g kg-1 lower than the overall average for 3-, 5-, and 7-wk grazing fre quencies (Mislevy et al., 1991). Commonly used bahiagrass genotypes such as Pe nsacola and Tifton 9 are sensitive to short daylengths, which induce a change of growth pa ttern resulting in a re duction in above-ground growth during the cool season. During this time, remaining live he rbage continues to photosynthesize, but instead of using photosynthate to produce new above-ground tissue, plants accumulate storage organ mass and store photosynt hate as nonstructural carbohydrates in stem bases, stolons, and rhizomes (Thornton et al., 20 00). At this physiological growth stage, plants are effectively preparing for wi nter survival and early spring regrowth, when photosynthetic production is inadequate to meet growth demands (Thornton et al., 2000). Limited bahiagrass herbage accumulation during the cool season has o ccurred even when temperature, soil moisture, and soil fertility were adequate for substantia lly greater growth (Sinclair et al., 1997). The significant economic implications of s easonal bahiagrass forage shortfall have stimulated research aimed at in creasing productivity during shortdaylength months. This effort has involved genetic selection and development of bahiagrass cultivars that are cold adapted and less sensitive to photoperiod (Blount et al., 2001). Additional cool-season productivity must not

PAGE 101

101 come at the expense of persistence, however, because the cost of past ure renovation is high and return per unit land area from grasslandlivestock systems are relatively low. There are no known studies evaluating yield, nu tritive value, and persistence responses of less photoperiod sensitive, cold-adapted (PCA) bahiagrass genotypes to defoliation. Therefore, the objectives of this study were to determine th e effects of defoliation frequency, referred to as harvest frequency in this dissertation, and defo liation intensity, referred to as stubble height, on total season and seasonal herbage yield, herbage nutritive value (CP and IVDOM concentrations), plant cover, a nd root + rhizome and stem base mass, N concentration, and TNC concentration of PCA bahiagrass ge notypes and existing cultivars. Materials and Methods Experimental Site This field study was conducted at the Beef Research Unit, northeast of Gainesville, FL, at 29 N latitude. Soils were classified as Spodosols (loam y, siliceous, subactive, thermic Grossarenic Paleaquults from the Plummer se ries or sandy, siliceous, hyperthermic Ultic Alaquods from the Pomona series) with average pH of 6.8. Average Mehlich-I extractable soil P, K, Mg, and Ca concentrations at the beginni ng of the experiment were 14, 25, 82, and 731 mg kg-1, respectively. Treatments and Design This experim ent evaluated the effects of tw o harvest frequencies and two stubble heights on bahiagrass genotype yield and persistence resp onses for 3 yr. Treatments (n = 20) were the factorial combinations of two harvest frequencies, two stub ble heights, and five bahiagrass genotypes arranged in three replications of a randomized complete block design. Sexual diploid and apomictic tetraploid ge notypes of bahiagrass were evaluated. Sexual diploid genotypes were Pensacola, Tifton 9, a nd PCA (photoperiod and cold-adapted) Cycle 4.

PAGE 102

102 Apomictic tetraploid genotypes evaluated were Argentine and Tifton 7. Pensacola, Tifton 9, and Argentine are commonly utilized bahiagrass cultivars in Florida, while PCA Cycle 4 is a novel genotype selected to improve cold tolerance, photoperiod response, nematode and disease resistance, rooting/rhizome mass, seedling vigor and establishment, s easonal distribution of forage production, and forage quality (Blount et al., 2001). PCA Cycle 4 has been approved for release by the University of Florida Agricultural Experiment Station as cv. UF-Riata (Blount et al., in review). Tifton 7 is a noncommercial apom ictic genotype. Harvest frequencies were 7 and 21 d, and stubble heights were 4 and 8 cm. Defoliation treatments were selected to impose a range of stress on the five genotypes because bahiag rass cultivars such as Pensacola are valued due to their tolerance of close, frequent defoliation. Plots were planted in the field in July 2004 using seedlings that had been growing in speedling flats. Plants were arranged in 2x 1-m plots in a ten-plant X fi ve-plant grid and were spaced 20-cm apart within and between rows Plots were undefoliated throughout 2004, and 50 kg N ha-1 was applied on 28 July and 28 August. During each of the subsequent trial years, plots were fertilized in March with 40 kg N, 18 kg P, and 66 kg K ha-1. Plots were staged in late April to early May and fertilized with 40 kg N ha-1 immediately after stagi ng. Six weeks later, plots were fertilized with 40 kg N, 18 kg P, and 66 kg K ha-1. Plots were then fertilized with 40 kg N ha-1 for the two remaining 6-wk cycles until the e nd of the trial. Total-season fertilizer amounts during the trial were 200 kg N, 36 kg P, and 132 kg K ha-1 yr-1, and are lower than the IFAS recommendation for bahiagrass grown only for hay (90 kg N and 37 kg K ha-1 applied after each cutting; Mylavarapu et al., 2007). The experiment was conducted during the growing seasons of 2005 (2 May to 18 October), 2006 (1 May to 16 October), and 2007 (1 May to 16 October).

PAGE 103

103 Samples to measure cool-season herbage produc tion were taken in November 2007 and April 2008. Response Variables Response variables m easured include total season and seasonal herb age yield, nutritive value, cover, and root + rhizome and stem base mass, N, and TNC concentration. Herbage yield was determined at harvest dates by clipping 0.25-m2 quadrats to the appropriate stubble height. Remaining plot area was mowed to the target stubble height and raked. Herbage samples were dried at 60 C to constant weight, weighed, and ground in a Wiley mill (Model 4 Thomas-Wiley Laboratory Mill, Thomas Scientific, Swedeboro, NJ ) to pass a 1-mm screen prior to analyses. Herbage samples of 7-d treatments from three c onsecutive harvests were composited prior to grinding so for nutritive value an alyses there was one sample per plot, regardless of harvest frequency, for each 21-d period. For N analysis, samples were digested us ing a modification of the aluminum block digestion procedure of Gallaher et al. (1975). Sample weight was 0.25 g, catalyst used was 1.5 g of 9:1 K2SO4:CuSO4, and digestion was conducted for at l east 4 h at 375C using 6 ml of H2SO4 and 2 ml H2O2. Nitrogen in the digestate was determined by semiautomated colorimetry (Hambleton, 1977). Crude protein was calculated as N multiplied by 6.25. In vitro digestible organic matter was determined using a modifi cation of the two-stage technique (Moore and Mott, 1974). Bahiagrass cover was determined visually by tw o observers three times per year (following first harvest, middle of the growing season, and after the last harvest) using a 1-m2 quadrat divided into 10-cm X 10-cm squares. The cover estimate was made on a 1-m2 area from the center of the plot. Cover was evaluated at the end of the growing season for 3 yr.

PAGE 104

104 Destructive sampling was performed three times per year (spring, summer, and fall) after scheduled herbage harvests usi ng a soil core (10-cm diameter X 18-cm depth). After removing the core, stem bases were cut to soil level. R oot + rhizomes were separated from the soil by washing with water over a screen with 1-cm2 mesh. Root + rhizome and stem base components were dried at 60 C to constant weight, weighed, and gr ound to pass a 1-mm screen prior to analyses. Root + rhizome and stem base samples (1296 total from two experiments) were analyzed for N and TNC concentration by near-infrared reflectance spectroscopy (NIRS) calibrated with data from 10% of the samples analyzed using the methods of Gallaher et al. (1975) and Hambleton (1977) for N and a procedur e described by Christiansen et al. (1988) for TNC. Samples were incubated with invertase and amyloglucosidase to hydrolyze sucrose and starch, respectively, and then resultant total he xoses were analyzed by a reducing sugar assay. Spectral data were collected on all samples, an average of 32 scans for each sample, with a NIR Systems 6500 spectrophotometer (Foss Int., Laur el, MD) equipped with a static sample cup. Principle component analysis wa s conducted on the spectral da ta and on a subset selected for calibration using the > Select = procedure of the software InfraSoft In ternational (ISI, State College, PA) based on spectral dissimilarity of samples (Schenk and Westerhaus, 1991a). Reference laboratory data for N and TNC were compared with the spectral data for the calibration samples and equations were develope d with the ISI software using partial least squares regression (Schenk and Westerhaus, 1991b). The N mean, standard error of validation, and r2 for the equation used were: 10.6 g kg-1, 0.9 g kg-1, and 0.85, respectively. The TNC mean, standard error of validation, and r2 for the equation used were: 52.2 g kg-1, 13.4 g kg-1, and 0.86, respectively. These equations were then used to predict N and TNC for all samples, including those used for the calibration.

PAGE 105

105 Statistical analyses were performed using Pr oc Mixed of SAS (SAS Inst. Inc., 1996), and the LSMEANS procedure was used to compar e treatment means. Genotype, stubble height, harvest frequency, year, and their interactions were considered fi xed effects and replicate and its interactions random effects. Year was consider ed fixed because of the potential for carryover effects of treatments from Years 1 and 2. Year was included in the model as a subplot treatment in a split-plot arrangement, with the treatment combinations being the main plots. The PDIFF function of the LSMEANS procedure was used to compare genotype means. Differences among stubble height and harvest fre quency means were based on F te sts. When treatment by year interaction was significant, data were analyzed and are reported by year. For responses measured multiple times per year, sampling date was treated as a repeated measure. Significance was determined at P 0.05. Results and Discussion Rainfall Annual rainf all in 2006 (843 mm) was much lo wer than the 30-yr average (1228 mm), 2005 (1540 mm), and 2007 (1171 mm; Figure 5-1). Ra infall in 2006 was below average for all months except September and October, and was considerably lower than 2005 and 2007 in August. May and June 2005 rainfall were above average and much higher than in 2006 and 2007. Total-Season Herbage Yield Total-season herbage yield was determ ined by summing yields across harvests each year. Total-season herbage yield was affected by ge notype, harvest frequency, year, and genotype X harvest frequency, genotype X year, and harvest fr equency X year interactions, so data were analyzed by year (Table C-1). Total-season herbage yield was affected by genotype X harvest frequency interaction (Table 5-1). When harvested every 7 d, Argentine had greatest total-seas on herbage yield while

PAGE 106

106 Figure 5-1. Monthly rainfall da ta at the experimental si te; average of 30-yr, 2005, 2006, 2007, and through April 2008. PCA Cycle 4 had lesser yields th an other genotypes. When harv ested every 21 d, the tetraploids Argentine and Tifton 7 had greate r total-season herbage yields th an the diploids. Tifton 7, Tifton 9, and PCA Cycle 4, the most upright-growing genotypes, had greater total-season herbage yields when harvested every 21 th an every 7 d, but there were no differences in yields between harvest frequencies for the more decumbent Argentine and Pensacola. Harvesting every 7 d likely resulted in proportionally greater periods of negative C balance during which the upright types utilized reserves to form new leaves (D avies, 1988). Previous studies have shown that Pensacola had a greater percentage of herbage near soil level, resu lting in greater residual leaf area following defoliation (Beaty et al., 1968). A dditionally, tiller mass wa s generally greater when harvested every 21 than 7 d (Chapter 6) wh ich may also contribute to greater yields with less frequent harvest for upright-growing types. The literature is not to tally consistent on the effect of harvest frequency on bahiagrass yield. During a 2-yr study, Beaty et al. (1970) reported little effect of harvest freque ncy (harvested every 1, 2, 3, 4, 5, or 6 wk) on total Pensacola 0 50 100 150 200 250 300 JanFebMarAprMayJunJulAugSepOctNovDec MonthRainfall (mm) 2005 2006 2007 2008 30-yr average

PAGE 107

107 Table 5-1. Bahiagrass total-season herbage yield as affected by genotypes X harvest frequency interaction. Data are means across two st ubble heights, three years, and three replicates (n = 18). Harvest frequency (d) Genotype 7 21 --------------------------------Mg DM ha-1 -------------------------------Argentine 11.8 a A 12.4 a A Tifton 7 10.2 b B 12.7 a A Pensacola 10.0 b A 10.8 b A Tifton 9 9.7 b B 11.4 b A PCA Cycle 4 8.4 c B 10.6 b A SE 0.4 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). yield, with greatest yields occurring with most frequent clipping in the first year and least frequent clipping in the second y ear. Gates et al. (1999) and Misle vy et al. (1991) found that total DM yields of Pensacola and Tifton 9 generally increased as harvest frequency decreased. Likewise, Beaty et al. (1963) noted that Pensacola yields were gr eater as frequency of harvest decreased; 6-wk intervals between harvests result ed in greater yields than 1-, 2-, 3-, or 4-wk intervals. Similarly, Cuomo et al. (1996) reported average total season forage yields of Argentine, Pensacola, and Tifton 9 increased as interval between harvests increased from 20 (10.6 Mg ha-1), to 30 (11.8 Mg ha-1), to 40 d (12.3 Mg ha-1), although their levels cover a range that is largely outside th at of the current study. When analyzed by year, total-season herbag e yield was affected by genotype in 2005 and 2007 and harvest frequency in all years (Table C2). In 2005, Argentine had greatest total-season herbage yield (Table 5-2), while PCA Cycle 4 had lesser yield than all genotypes. There were no differences in total-season herbage yields be tween Tifton 7 and Pensacola, and Tifton 9 had lesser yields than all genotypes except PCA Cycle 4. There were no differences ( P > 0.05) in total-season herbage yield among genotypes in 2 006. In 2007, Argentine and Tifton 7 had greater

PAGE 108

108 total-season herbage yields than Pensacola and PCA Cycle 4, while Tifton 9 total-season herbage yield was intermediate. In general, the tetraploid s tended to have greater total-season yields than the diploids. PCA Cycle 4 had lesser total-season herbage yield in 2005, but its yields were similar to Pensacola and Tifton 9 in 2007. These total-season herbage yields are simila r to those reported fo r bahiagrasses under similar environments and management. Pedreira and Brown (1996a) reported 3-yr mean totalseason DM yields of 7.8 and 8.7 Mg ha-1 for Pensacola and Tifton 9, respectively, in Athens, GA, when fertilized with 100 kg N ha-1 in spring and summer and cut to either 3.5 or 10 cm every 14 d. When fertilized with 56 kg N ha-1 prior to initial and after each harvest and cut to 7 cm every 30 d, Mislevy et al. (2005) reported 3-yr mean tota l-season DM yields of 11.2, 12.1, 10.3, and 11.9 Mg ha-1 for Argentine, Tifton 7, Pensacola, an d Tifton 9, respectively, at Ona, FL. Mislevy and Everett (1981) reported 2-yr mean total-season DM yields of 10.3 and 12.1 Mg ha-1 for Argentine and Pensacola, respectively, in I mmokalee, FL, when fertilized with 56 kg N ha-1 after each warm-season harvest and cut to 5 or 10 cm every 30 d. Table 5-2. Bahiagrass total-season herbage yi eld as affected by genotype in 2005, 2006, and 2007. Data are means across two stubble height s, two harvest frequencies, and three replicates for each year (n = 12). Year Genotype 2005 2006 2007 ------------------------Mg DM ha-1 ------------------------Argentine 12.3 a 12.1 a 11.9 a Tifton 7 10.8 b 12.0 a 11.5 a Pensacola 9.0 b 11.5 a 10.8 b Tifton 9 8.6 c 11.8 a 11.2 ab PCA Cycle 4 7.6 d 10.3 a 10.3 b SE 0.3 0.5 0.4 Means within a column followed by the same letter do not differ by the LSMEANS test ( P > 0.05).

PAGE 109

109 There also was harvest frequency X year inte raction for total-season herbage yield (Figure 5-2). Across genotypes, total-seas on herbage yields were greater each year when harvested every 21 than every 7 d, and interacti on occurred because magnitude of the difference was greater in 2005 and 2006 than in 2007. Reasons for the harves t frequency response were discussed earlier. Figure 5-2. Bahiagrass total-season herbage yiel d as affected by harvest frequency in 2005, 2006, and 2007. Means within a year followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Data are means across five genotypes, two stubble heights, and three replicat es for each year (n = 30). Seasonal Herbage Yield Herbage yield was calcu lated for all treatments for eight 21-d harves t periods during each of 3 yr. Herbage yield was affected by multiple two-way and three-way interactions, and a fourway interaction (Table C-3). The focus of analysis of interactions for herbage yield was changes in genotypes across the growing season, so the ge notype X evaluation period interaction received particular attention. When data were analyzed by year, herbage yield was affected by genotype X evaluation period interaction in a ll 3 yr, so data were analyzed by evaluation period within year. 0 2 4 6 8 10 12 14 FreqYearHerbage DM yield (Mg DM ha-1) 7 d 21 d 2006 2007 b a 2005 a a b b

PAGE 110

110 In 2005, herbage yield was affected by genotype during all evaluation periods (Table C-4; Figure 5-3). The decline in yields in autumn can be attributed primarily to declining daylength (Sinclair et al., 2003) and possibl y to below-average rainfall in September (more than 50 mm less than normal). The rate of decline in yiel d from 16 August forward through autumn was less for PCA Cycle 4 than the othe r genotypes. During 3 to 24 May, Tifton 9 and Pensacola had greater yields than Tifton 7 and Argentine, wi th Argentine having least yield overall. PCA Cycle 4 yield was intermediate among genotypes. Cham bliss and Adjei (2006) describe Argentine as less cold tolerant than Pensacola and Tifton 9, and as a result, it does not initiate growth as early in the spring. This resulted in lesser spring yields of Argentin e, and a similar response was observed for the other tetraploid, Tifton 7. Cuomo et al. (1996) reported similar first harvest results, with Argentine producing less forage (1.9 Mg ha-1) than Pensacola (3.1 Mg ha-1) or Tifton 9 (2.9 Mg ha-1) on 19 May in Louisiana. During 24 May to 14 June, Argentine and Pens acola had greatest yields, and PCA Cycle 4 yield was less than all genotype s except Tifton 9. During most s ubsequent harvest periods in 2005, the tetraploids yielded more than the dipl oids, and PCA Cycle 4 yield was less than or similar to the other diploids. C uomo et al. (1996) reported that other than the May harvest in Louisiana, they found no differences in forage yield among bahiagrass cultivars by harvest date. Mislevy and Everett (1981) and Ad jei et al. (1989) re ported no differences in summer yields of Pensacola and Argentine in Imm okalee and Ona, FL, respectively. Herbage yield was affected by genotype duri ng all but three evalua tion periods in 2006 (Table C-4; Figure 5-4). The overall increased yields during 12 June to 3 July and 24 July to 14

PAGE 111

111 Figure 5-3. Bahiagrass seasonal herbage yi eld as affected by genotype during eval uation periods in 2005. ** Genotype effect significant at P 0.01. Data are means across two stubble heights, two ha rvest frequencies, and three replicates for each evaluation period (n = 12). 0 0.5 1 1.5 2 2.5 33 to 2 4 M a y 24 May to 1 4 J un e 14 June to 5 July 5 to 2 6 J u l y 26 J u l y to 16 Au gu s t 16 August to 6 September 6 to 2 7 Se pte m b er 27 September t o 18 OctoberEvaluation periodHerbage yield (Mg DM ha-1) Argentine Tifton 7 Pensacola Tifton 9 PCA Cycle 4 ** SE=0.1 ** SE=0.1 ** SE=0.1 ** SE=0.1 ** SE=0.1 ** SE=0.1 ** SE=0.1 ** SE=0.1

PAGE 112

112 August were a response to N fertilization that wa s applied on 12 June and 24 July, respectively. The decline in yields in autumn can be attribut ed primarily to shorteni ng daylength (Sinclair et al., 2003) and possibly to belowaverage rainfall for much of the growing season. During 22 May to 12 June, yield was similar among the diploids a nd greater than the tetrap loids, while Argentine had least yield. During 12 June to 3 July, Tifton 9 and PCA Cycle 4 herbage yields were greater than Tifton 7 and Argentine, with Pensacola yiel d intermediate. The lesse r spring yields of the tetraploids extend longer into the growing season in 2006 than 2005 and are consistent with expected results due to their slower growth in the spring (C hambliss and Adjei, 2006). From 24 July to 14 August, Argentine and Pensacola herb age yields were greater than Tifton 9 and PCA Cycle 4, with Tifton 7 yield intermediate. Argentine had greater herbage yield than all genotypes except Tifton 7 from 14 August to 5 September and 5 to 25 September. PCA Cycle 4 yield was less than all genotypes except Pensacola during 14 August to 5 September, and was similar to other diploids from 5 to 25 September. In 2007, herbage yield was affected by genot ype during all but tw o evaluation periods (Table C-4; Figure 5-5). The overall increased yields during 12 June to 3 July, 24 July to 14 August, and 5 to 25 September were a respons e to N fertilization on 12 June, 24 July, and 5 September, respectively. There was a pronounced dec line in late summer and autumn yields in 2007. It does not appear to be a function of rainfa ll as monthly totals we re at or above 30-yr averages for July through October. From 1 to 22 May and 22 May to 12 June, the diploids had similar herbage yield that was greater than the tetraploids (there were no differences in yield between the tetraploids). The lesser spring yields of the tetraploids are consistent with the first 2 yr, and like Year 2 extended into June. Begi nning from the 24 July to 14 August period and

PAGE 113

113 Figure 5-4. Bahiagrass seasonal herbage yiel d as affected by genotype during evaluati on periods in 2006. NS, Genotype effects n ot significant, P > 0.05. Genotype effect significant at P 0.05. ** Genotype effect significant at P 0.01. Data are means across two stubble heights, two harvest frequencies, and three replicates for each evaluation period (n = 12). 0 0.5 1 1.5 2 2.51 to 2 2 M ay 22 May to 12 June 12 June to 3 J u ly 3 to 2 4 J u ly 24 J ul y to 1 4 A u gus t 14 August to 5 Sept e mbe r 5 to 25 Sep te mber 25 Septem b er to 16 Octobe r Evaluation periodHerbage yield (Mg DM ha-1) Argentine Tifton 7 Pensacola Tifton 9 PCA Cycle 4 NS SE=0.1 ** SE=0.1 SE=0.1 NS SE=0.1 ** SE=0.1 ** SE=0.1 ** SE=0.1 NS SE=0.1

PAGE 114

114 Figure 5-5. Bahiagrass seasonal herbage yiel d as affected by genotype during evaluati on periods in 2007. NS, Genotype effect no t significant, P > 0.05. ** Genotype effect significant at P 0.01. Data are means across tw o stubble heights, two harvest frequencies, and three replicates for each evaluation period (n = 12). 0 0.5 1 1.5 2 2.5 31 to 22 May 22 May t o 12 J u ne 12 Ju ne to 3 July 3 t o 24 July 24 Ju l y to 14 August 14 Au gu st t o 5 S ep t ember 5 to 2 5 September 25 September to 16 Octobe rEvaluation periodHerbage yield (Mg DM ha-1) Argentine Tifton 7 Pensacola Tifton 9 PCA Cycle 4 ** SE=0.1 ** SE=0.1 NS SE=0.1 NS SE=0.1 ** SE=0.1 ** SE=0.1 ** SE=0.1 ** SE=0.1

PAGE 115

115 continuing through the 14 August to 5 September period, this trend was reversed, as Argentine and Tifton 7 had greater herbage yields than th e diploids. During these periods, there were no differences in herbage yields between the tetraploids or among the diploids. From 5 to 25 September, Argentine herbage yield was greater than all genotypes except Tifton 7, while PCA Cycle 4 herbage yield was less than all genotype s except Pensacola. From 25 September to 16 October, Tifton 7 herbage yield was greater than PCA Cycle 4 and Tifton 9, while Tifton 9 herbage yield was less than all except PCA Cycle 4. Cool-Season Herbage Yield Herbage yield was determ ined for two additional harvest dates during the 2007-08 cool season, 30 Nov. 2007 and 21 Apr. 2008. Cool-season herbage yield was affected by harvest frequency, evaluation date, and genotype X st ubble height, genotype X harvest frequency, stubble height X harvest freque ncy, stubble height X evaluation date, and harvest frequency X evaluation date interactions (Table C-5). Because the responses of genotypes in the cool season are of primary interest, only the interactions of stubble height and harvest frequency with genotype will be discussed. Because the harvest frequency treatment had not been imposed since the last harvest in October 2007, harvest frequenc y effects are carry over effects from imposing that treatment during 3 yr. When harvested to 4 cm, Pensacola and PC A Cycle 4 had similar cool-season herbage yields that were greater than Argentine and Tifton 7 (Table 53). Cool-season herbage yield of Tifton 9 was intermediate. At 8-cm stubble, Pensacola and Tifton 9 had greater cool-season herbage yields than Argentine and Tifton 7, with yield of PCA Cycle 4 intermediate. Muchovej and Mullahey (2000) and Mislevy et al. (2005) reported no differe nces in herbage yields among Argentine, Pensacola, Tifton 7, and Tifton 9 at Immokalee and Ona, FL, respectively, when harvested in October and November. In December, Mislevy et al. (2005) reported similar yields

PAGE 116

116 for Pensacola, Tifton 7, and Tifton 9, which were gr eater than yields of Argentine at Ona, FL. Cool-season herbage yield of Tifton 7, Pensaco la, and Tifton 9 in the current study was not affected by stubble height. However, Argentin e and PCA Cycle 4 had greater cool-season herbage yield when harvested to 4than 8-cm stubble height. When harvested during the winter months (October March), Pensacola and Argen tine produced more forage at 5than 10-cm stubble (Mislevy and Everett, 1981). Table 5-3. Bahiagrass cool-season herbage yiel d as affected by ge notype X stubble height interaction. Data are means across two harves t frequencies, two evaluation dates, and three replicates (n = 12). Stubble height (cm) Genotype 4 8 ------------------Mg DM ha-1 ------------------Argentine 0.39 b A 0.20 b B Tifton 7 0.34 b A 0.29 b A Pensacola 0.49 a A 0.41 a A Tifton 9 0.42 ab A 0.50 a A PCA Cycle 4 0.59 a A 0.32 ab B SE 0.06 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). There was no difference in cool-season herbag e yield among genotypes in plots that were previously harvested every 7 d (Table 5-4). For those that were harvested every 21 d, Pensacola, Tifton 9, and PCA Cycle 4 had similar cool-season herbage yields that were greater than the yields of Argentine and Tifton 7. There was no effect of previ ous harvest frequency on coolseason herbage yield of Argentine and Tifton 7, while Pensacola, Tifton 9, and PCA Cycle 4 had greater yields when harveste d every 21 than 7 d. As a group, tetraploid bahiagrasses are generally less cold-tolerant and more susceptible to frost damage, accounting for their generally lesser yields in the cool season (Chambliss and Adjei, 2006). Greater bahiagrass cover and more residual leaf area and phot osynthetic material remaining for 21than 7-d harvest frequency at the

PAGE 117

117 end of regular harvests in Octobe r likely contributed to greater yields of Pensacola, Tifton 9, and PCA Cycle 4 at the 21-d harvest frequency. Table 5-4. Bahiagrass cool-season herbage yield as affected by genotype X harvest frequency interaction. Data are means across two st ubble heights, two evaluation dates, and three replicates (n = 12). Harvest frequency (d) Genotype 7 21 ------------------Mg DM ha-1 ------------------Argentine 0.24 a A 0.36 b A Tifton 7 0.29 a A 0.34 b A Pensacola 0.29 a B 0.62 a A Tifton 9 0.31 a B 0.61 a A PCA Cycle 4 0.28 a B 0.64 a A SE 0.06 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Nutritive Value Herbage CP and IVDOM concentrations were m e asured for the eight 21-d harvest periods during the growing seasons of 2005 and 2006 onl y. Herbage from the 7-d treatment was composited over the three harvests within a given 3-wk period to provide one sample that corresponds with the sample taken from each harves t of the 21-d treatment. Laboratory analyses for CP and IVDOM concentrations were conducted on these samples. Each year, CP and IVDOM yield was calculated across harvests and divided by totalseason DM yield (for CP) or OM yield (for IVDOM) to give weighted totalseason herbage CP and IVDOM concentrations. Weighted total-season cr ude protein concentration Weighted total-season C P concentration was a ffected by genotype, st ubble height, harvest frequency, year, and genotype X harvest fr equency, stubble height X harvest frequency, genotype X year, and stubble X year interactions (Table C-6), so da ta were analyzed by year to compare genotypes and stubbl e heights within year.

PAGE 118

118 Weighted total-season CP concentration was affected by genotype X harvest frequency interaction (Table 5-5). When harvested ever y 7 d, Argentine, Pensacola, Tifton 9, and PCA Cycle 4 had similar CP concentrations, which were greater than Tifton 7. There were no differences in CP among genotypes when harves ted every 21 d. All genotypes had greater CP concentration when harves ted every 7 than 21 d. Table 5-5. Bahiagrass weighted total-season crude protein concentr ation as affected by genotype X harvest frequency interaction. Data ar e means across two stubble heights, two years, and three replicates (n = 12). Harvest frequency (d) Genotype 7 21 ----------------------g kg-1 -----------------------Argentine 137 a A 117 a B Tifton 7 130 b A 112 a B Pensacola 137 a A 117 a B Tifton 9 141 a A 113 a B PCA Cycle 4 141 a A 115 a B SE 2.0 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Weighted total-season CP concentration was also affected by stubble height X harvest frequency interaction (Table 5-6). The CP con centration was greater at 4than 8-cm stubble height for both 7and 21-d harvest frequenc ies, and was greater at 7than 21-d harvest frequency for both 4and 8-cm stubble heights. Interaction occurred because the difference among frequencies was slightly greater for 4than 8-cm stubble. Misl evy and Everett (1981) reported Pensacola and Argentine bahiagrass harveste d in September had greater CP at 5than at 10-cm stubble height (145 and 135 g CP kg-1, respectively) in Immokalee, FL. When analyzed by year to assess interactions with year, weighted total-season CP concentration was affected by genotype and st ubble height in 2005 and stubble height in 2006 (Table C-7). In 2005, Argentine, Pensacola, Tifton 9, and PCA Cycle 4 had similar CP

PAGE 119

119 Table 5-6. Bahiagrass weighted total-season crude protein concentr ation as affected by stubble height X harvest frequency interaction. Da ta are means across five genotypes, two years, and three replicates (n = 30). Harvest frequency (d) Stubble height (cm) 7 21 ----------------------g kg-1 -----------------------4 143 a A 117 a B 8 132 b A 112 b B SE 1.0 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). concentration (119, 122, 123, and 122 g kg-1, respectively), which was greater (P < 0.0001) than that of Tifton 7 (113 g kg-1). In 2005, CP concentration was greater (P < 0.0001) at 4than 8-cm stubble height (123 and 117 g kg-1, respectively). The same trend occurred in 2006, with greater ( P < 0.0001) total-season CP concentration occurring at 4than 8-cm stubble height (138 and 126 g kg-1, respectively). Seasonal crude protein concentration Seasonal CP concentration was af fected by multiple two-way and three-way interactions including year (Table C-8). The focus of analysis of interactio ns for seasonal CP data was changes in genotypes across the growing season, so the genotype X evaluation period interaction received particular attention. When data were analyzed by year, there was genotype X evaluation period interaction in 200 5 (Table C-9) and genotype X harv est frequency X evaluation period interaction in 2006 (Table C-9). When analyzed by evaluation period in 2005, CP concentration was affected by genotype during six of eight periods (Figure 5-6). Mean genotype CP concentrations ranged from 149 in May to 101 g kg-1 in July. In general, herbage CP concentration of all genotypes decreased from spring to summer and increased a pproaching autumn. Similar trends were reported for Argentine, Pensacola, Tifton 9, and Tifton 7 by Mislevy et al. (2005) in Ona, FL from 1998-2000, and

PAGE 120

120 Muchovej and Mullahey (2000) in Immokalee, FL from 1995-1996. St ewart et al. (2007) reported that Pensacola CP in continuously st ocked pastures generally decreased from May through August. The increased herbage CP concentr ations from 26 July to 16 August and 6 to 27 September were primarily a function of N fertilizer that was applied on 26 July and 6 September, respectively. From 3 to 24 May and 16 August to 6 Sept ember, Argentine had greatest herbage CP concentration, and there was no difference in CP concentration among the other genotypes. From 24 May to 14 June, Argentine had greater herbag e CP concentration than all genotypes except PCA Cycle 4 and Tifton 9, while CP concentrati on of Tifton 7 was less than all but Pensacola. From 14 June to 5 July and 5 to 26 July, the di ploids had similar herb age CP concentrations which were greater than the tetraploids. Ar gentine and Pensacola ha d similar herbage CP concentration on 27 September to 18 October that was greater than all genotypes except Tifton 9. Argentine tended to have greater herbage CP concentrations than other genotypes early and late in the growing season, while Tifton 7 tended to have less CP concentration than most genotypes throughout the growing season. The herbage CP concentration of PCA Cycle 4 was generally similar to that of the other diploids. These tr ends are consistent with results published for bahiagrass cultivars in Florida when harvested every 35 and 30 d, respectively (Muchovej and Mullahey, 2000; Mislevy et al., 2005). Data for 2006 were analyzed by harvest freque ncy and are presented by evaluation period within harvest frequency. The 2006 seasonal tren ds for CP for the 7-d harvest frequency are presented in Figure 5-7. Compar ed to 2005, CP concentration d ecreased less from spring to summer and there were more distinct responses of CP concentration to N fertilization in 2006.

PAGE 121

121 Figure 5-6. Bahiagrass seasonal crud e protein concentration as affected by genot ype at evaluation peri ods in 2005. NS, Genotype not significant, P > 0.05. ** Genotype significant at P 0.01. Data are means across two stubble heights, two harvest frequencies, and three replicates for each evaluation period (n = 12). 70 90 110 130 150 1703 to 24 May 24 May to 14 J une 14 June to 5 July 5 to 26 Ju l y 26 J ul y to 16 Au gust 16 Au gu st to 6 Septe m b er 6 to 27 S eptember 27 September to 18 OctoberEvaluation periodHerbage crude protein (g kg-1) Argentine Tifton 7 Pensacola Tifton 9 PCA Cycle 4 ** SE=5 ** SE=4 ** SE=3 ** SE=2 NS SE=5 ** SE=2 NS SE=2 ** SE=2

PAGE 122

122 Figure 5-7. Bahiagrass seasonal crud e protein concentration as affected by genot ype at evaluation periods for the 7-d harvest frequency in 2006. NS, Genotype effect not significant, P > 0.05. ** Genotype effect significant at P 0.01. Data are means across two stubble heights and three replicates at each evaluation period (n = 6). 110 130 150 1701 to 22 M ay 22 May t o 1 2 J u ne 1 2 J u ne t o 3 J uly 3 to 24 J uly 2 4 Jul y to 14 A ug u s t 1 4 A ug u s t t o 5 S ep t e m be r 5 to 25 Septe m ber 2 5 Se p t e mb er t o 1 6 O cto be rEvaluation periodHerbage crude protein (g kg-1) Argentine Tifton 7 Pensacola Tifton 9 PCA Cycle 4 NS SE=4 NS SE=3 ** SE=3 ** SE=2 NS SE=3 NS SE=3 ** SE=4 ** SE=4

PAGE 123

123 The increased herbage CP concentrations on 12 June to 3 July, 24 July to 14 August, and 5 to 25 September were primarily a function of N fertilization on 12 June, 24 July, and 5 September, respectively. Herbage CP concentrat ion was affected by genotype during four of eight periods in 2006. In general, there were fe w differences in herbage CP concentration among the diploids, and Tifton 7 tended to have lesser CP than most genotypes when harvested every 7 d from late-summer through autumn. Argentine had greater CP concentration than other genotypes in June and July. These trends are c onsistent with results published for bahiagrass cultivars in Florida by Muchovej and Mullahey (2000) and Misle vy et al. (2005). The 2006 genotype herbage CP concentration se asonal trend for the 21-d harvest frequency is presented in Figure 5-8. Compared to 2005, he rbage CP concentration decreased less from spring to summer and there were more distinct re sponses of CP concentration to N fertilization in2006 than was observed for the 7-d frequency. Crude protein concentration was affected by genotype during five of eight periods in 2006. The most striking genotype differences were from 1 to 22 May and 12 June to 3 July when Argentin e had greatest herbage CP concentration. There were no differences in CP concentration among the other genotypes at these dates, and among the diploids in general there we re few differences in herbage CP concentration when harvested every 21 d throughout 2006. The herbage CP concentration seasonal trends in this experiment were generally similar to results reported by other authors. Sollenberger et al. (1988 and 1989) found that CP of Pensacola bahiagrass pastures decreased from early summer to mid-summer, then increased in late summer, and decreased again approaching the cool season near Gainesville, FL. At Ona, FL, Ezenwa et al. (2006) reported Pensacola bahiagrass herbage CP on rotationally stocked pastures increased from May to June, declined through the summer and increased approaching the cool season.

PAGE 124

124 Genotype differences in CP in the current study were small and not consistent over time, thus there seems to be little reason to expect regu lar and predictable herbage CP differences among these five genotypes. Weighted total-season in vitro d igestible organic matter yield Weighted total-season IVDOM concentration was affected by genotype, stubble height, harvest frequency, year, and genotype X harvest frequency, genotype X year, harvest frequency X year, and stubble X year inte ractions (Table C-6), so data were analyzed by year. Across years, total-season I VDOM was affected by genotype X harvest frequency (Table 5-7). When harvested every 7 d, Tifton 9 and PCA Cycle 4 had greater IVDOM concentration than the other genotypes. When harvested ev ery 21 d, IVDOM of Tifton 7 was less than all genotypes except Tifton 9. With the exception of Argentine, all genotypes had greater IVDOM concentration when harvested every 7 than 21 d. These results are consistent with those published by Mislevy et al. ( 1991) in which IVDOM of Pensaco la and Tifton 9 decreased as interval between defoliation increased. Adjei et al. (1989) reported Pens acola IVDOM decreased as grazing frequency decreased, but IVDOM of Argentine was not affected by grazing frequency. Mislevy and Dunavin (1993) reported greates t IVDOM for Pensacola and Argentine when grazed every 2 wk, compared to 3-, 4-, and 5-wk grazing frequencies. Chaparro and Sollenberger (1997) indicated that IVDOM of M ott elephantgrass increased linearly as defoliation frequency increased. A similar tr end was reported by Motazedian and Sharrow (1990) in which perennial ryegrass ( Lolium perenne L.)-subclover (Trifolium subterraneum L.) digestibility increased linearly as harvest frequency increased.

PAGE 125

125 Figure 5-8. Bahiagrass seasonal crud e protein concentration as affected by genot ype during eight evaluati on periods at the 21-d harvest frequency in 2006. NS, Genotype effect not significant, P > 0.05. Genotype effect significant at P 0.05. ** Genotype effect significant at P 0.01. Data are means across two stubble height s and three replicates at each evaluation period (n = 6). 80 100 120 140 1601 to 2 2 May 22 May to 12 June 12 J un e t o 3 J uly 3 to 24 July 24 J uly t o 1 4 A u gus t 14 August to 5 Septemb er 5 t o 2 5 S e p tember 25 Sept e m b er t o 16 Oc t ob erEvaluation periodHerbage crude protein (g kg-1) Argentine Tifton 7 Pensacola Tifton 9 PCA Cycle 4 SE=3 SE=2 ** SE=2 NS SE=3 NS SE=3 NS SE=3 ** SE=3 ** SE=4

PAGE 126

126 Table 5-7. Bahiagrass weighted total-season in vi tro digestible organic matter concentration as affected by genotype X harvest frequency interaction. Data are means across two stubble heights, two years, and three replicates (n = 12). Harvest frequency (d) Genotype 7 21 -----------------------------g kg-1 --------------------------------Argentine 545 b A 538 a A Tifton 7 544 b A 530 b B Pensacola 547 b A 538 a B Tifton 9 562 a A 537 ab B PCA Cycle 4 565 a A 534 a B SE 4 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). When analyzed by year, total-season IVDOM concentration was affected by genotype and stubble height in both years and harvest fre quency in 2005 (Table C-7). In 2005, Tifton 9 and PCA Cycle 4 had greater IVDOM concentration th an the other genotypes (Table 5-8). In 2006, PCA Cycle 4 had greater total-season IVDOM con centration than all genotypes, while Tifton 7 had less IVDOM concentrati on than all genotypes. Table 5-8. Bahiagrass weighted total-season in vi tro digestible organic matter concentration as affected by genotype in 2005 and 2006. Da ta are means across two harvest frequencies, two stubble heights, and th ree replicates for each year (n = 12). Year Genotype 2005 2006 -----------------------------g kg-1 --------------------------------Argentine 531 b 552 b Tifton 7 537 b 537 c Pensacola 533 b 552 b Tifton 9 549 a 551 b PCA Cycle 4 546 a 562 a SE 3 5 Means within year followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Herbage IVDOM concentration was greater ( P < 0.0001) when harvested to 4than 8-cm stubble height in both years (2005: 547 and 532 g kg-1, respectively; 2006: 563 and 539 g kg-1,

PAGE 127

127 respectively). These results are consistent with those of Hirata (1993a) in which Pensacola in vitro DM digestibility (IVDMD) was almost always greater at 2than at 22-cm stubble height (570 and 460 g kg-1, respectively). Gates et al. (1999), however, reported no effect of stubble height (1.5 and 10 cm) on IVDM D of Pensacola and Tifton 9. Seasonal in vitro dige stible organic matter Seasonal IV DOM concentration was affected by multiple twoand three-way interactions including genotype X evaluation period X year intera ction (Table C-10). The focus of analysis of interactions for seasonal IVDOM data was changes among genotypes across the growing season, so the genotype X evaluation period interaction r eceived particular attention. Data were analyzed by year to describe this interaction. The res ponses of IVDOM to stubbl e height and harvest frequency main effects followed the same trends as weighted total-season IVDOM concentration and will not be presented here. There were few differences in IVDOM concentration among genotypes in 2005 (Table C11; Figure 5-9). In general, IVDOM concentration of all genotypes decreased from spring to summer, and remained fairly constant approa ching autumn. Tropical gr asses show a fairly consistent pattern related to dry matter digest ibility (DMD), with maximum DMD occurring in early/late spring, lesser valu es in mid-summer and early autumn, and an increase in late autumn (Coleman et al., 2004). Forage regrowth in the su mmer tends to have lowe r nutritive value due to increased lignin deposition associated with hi gh temperatures, and increased growth rates and maturation associated with high rainfall as is common in Florida (Adesogan et al., 2006). Higher temperatures and more favorable soil moisture conditions in the summ er promote more rapid growth and maturity, resulting in greater yi elds and lower digestibility (Chambliss and Sollenberger, 1991). Sollenberger et al. (1989) reported greatest Pens acola digestibility in spring and autumn on rotationally stocked pastures, with marked decline of IVDOM during late

PAGE 128

128 summer. A similar trend was observed by Johnson et al. (2001) for Pensacola harvested every 28 d to an 8-cm stubble. Some studies have not shown the autumn rebound in IVDOM. For example, Ezenwa et al. (2006) reported Pensacola IVDOM increased from May to a maximum in July, then decreased to the lowest level in Oc tober on rotationally stocked pastures. Likewise, Sollenberger et al. (1988) reported bahiagrass IVDOM generally decreased from June through October on continuously stocked pastures. Research in Japan showed that Pensacola IVDMD was greatest in the spring (560 g kg-1) and lowest in autumn (480 g kg-1) (Hirata, 1993a). Looking specifically at differences in seasonal patterns among genotypes in 2005, genotype main effect was significant only on 5 to 26 July, with Tifton 9 having greater IVDOM concentration than all other genotypes except PCA Cycle 4. Pe nsacola and Tifton 7 had lesser IVDOM concentration than all ge notypes except Argentine. Publis hed differences in nutritive value among bahiagrass cultivar s are generally small and inconsistent (Kalmbacher, 1997; Mislevy and Dunavin, 1993; Chamb liss and Adjei, 2006). Cuomo et al. (1996) reported Tifton 9 had slightly greater IVTD than Pensacola and Argentine in Louisiana, while Twidwell et al. (1998) found that Argentine had slightly greater digestibility than Pensacola and Tifton 9 in Louisiana. In southern Florida, Adjei et al (1989) reported no differe nces in summer IVDOM between Pensacola and Argentine and Mislevy et al. (1991) indicated that Pensacola had greater IVDOM than Tifton 9 in late summ er, but there were no differences between them in the fall. In Georgia, Gates et al. (1999) found no differences in IVDMD between Pensacola and Tifton 9. There were more differences in seasona l IVDOM concentration among genotypes in 2006 than 2005 (Table C-11; Figure 5-10). In general, IVDOM concentration remained fairly constant from spring through summer, and increased approaching autumn. The increased IVDOM concentration on 12 June to 3 July and 5 to 25 Se ptember may have been due to N fertilizer that

PAGE 129

129 Figure 5-9. Bahiagrass seasonal in vitro digestible organic matter concentration as affected by genotype during eight evaluation periods in 2005. NS Ge notype not significant, P > 0.05. ** Genotype significant at P 0.01. Data are means across two harvest frequencies, two stubble height s, and three replicates (n = 12). 400 450 500 550 600 650 7003 to 2 4 May 2 4 M a y to 14 June 1 4 J une to 5 Ju ly 5 to 26 Jul y 26 Ju ly to 16 A u g us t 16 August to 6 S e p te m b e r 6 to 27 September 2 7 S e p te m b e r t o 1 8 Octo b e rEvaluation periodHerbage in vitro digestible organic matter (g kg1) Argentine Tifton 7 Pensacola Tifton 9 PCA Cycle 4 NS SE=5 NS SE=8 NS SE=8 ** SE=9 NS SE=14 NS SE=3 NS SE=3 NS SE=5

PAGE 130

130 was applied on 12 June and 5 September, respectiv ely. Newman et al. (2006) reported Pensacola 2-yr mean IVDOM increased from 432 to 471 g kg-1 when N fertilization increased from 80 to 320 kg N ha-1. Stewart et al. (2007) reported Pensaco la IVDOM increased as management intensity (including N level) increased from low to high. In vitro digestible OM was affected by genot ype during five of eight periods in 2006. On 1 to 22 May, 22 May to 12 June, and 25 September to 16 October, PCA Cycle 4 and Pensacola had greater IVDOM concentration than Argentine and Tifton 7. On 12 J une to 3 July, Argentine had greater IVDOM concentration than all genotype s except Tifton 9, while Pensacola and Tifton 7 had lesser IVDOM concentration than all ge notypes except PCA Cycle 4. On 14 August to 5 September, PCA Cycle 4 had greatest IVDOM con centration, while Tifton 7 had lesser IVDOM concentration than all genotypes except Argentin e. There was a tendency for the tetraploids to have lesser IVDOM than the di ploids throughout the growing seas on in 2006. This contrasts with results of Muchovej and Mullahey (2000) in which there were no differences in IVDOM among bahiagrass cultivars in June, July, August, and October, while Mislevy et al. (2005) reported generally few differences in IVDOM among bahiagrass cultivars in June, August, and October. Exceptions were Argentine and Tifton 7 had lesser IVDOM (547 and 549 g kg-1, respectively) than Pensacola and Tifton 9 (605 and 601 g kg-1, respectively) in October 1998 and Argentine had lesser IVDOM (513 g kg-1) than Pensacola (542 g kg-1) in October 2000. These IVDOM data suggest that PCA Cycle 4 herbage is at least as digestible and sometimes more digestible than the existing bahiagrass cultivars. Differences among cultivars tend to be relatively small, howev er, and their impact on cattle gains are unlikely to be large. The exception may be Tifton 7, which in terms of total-season weighted data in 2006 was 25 g kg-1

PAGE 131

131 Figure 5-10. Bahiagrass seasonal in vitro digestible organic matter concentration as affected by genotype during eight evaluation periods in 2006. NS, Ge notype not significant, P > 0.05. ** Genotype significant at P 0.01. Data are means across two harvest frequencies, two stubble height s, and three replicates (n = 12). 400 450 500 550 600 650 7001 to 2 2 M ay 22 May to 12 June 12 Jun e to 3 J ul y 3 to 2 4 J u ly 24 J uly to 14 Au g ust 14 August t o 5 S eptember 5 to 2 5 Se ptem be r 25 Sept e mber t o 16 Octo b erEvaluation periodHerbage in vitro digestible organic matter (g kg1) Argentine Tifton 7 Pensacola Tifton 9 PCA Cycle 4 SE=8 ** SE=6 ** SE=5 NS SE=8 NS SE=18 ** SE=6 NS SE=6 ** SE=8

PAGE 132

132 less digestible than Cycle 4, a difference that likel y may lead to detectable differences in animal response. Cover Cover data included in the disse rtation are thos e for the end of the 3 yr of defoliation and were quantified in October each year. When anal yzed across years, there were no interactions with year but the three-way interaction of genotype X harv est frequency X stubble height approached significance ( P = 0.10; Table C-12). The threeway interaction means were compared. They illustrated important differen ces among genotypes (Table 5-9). Argentine was the only genotype for which cover was unaffected by defoliation treatment. Cover of Pensacola, also a decumbent type, was lower for the 7-d fre quency, 4-cm stubble height treatment compared to the other defoliation treatments. Tifton 9 and PCA Cycle 4 were most sensitive to defoliation treatment and had much greater cover when c lipped every 21 d to 8 cm than for any other treatments. Table 5-9. Bahiagrass cover as affected by ge notype X stubble height X harvest frequency interaction in October. Data are means acr oss three replicates and 3 yr (n = 9). Harvest frequency (d) 7 21 Stubble height (cm) Genotype 4 8 4 8 ------------------------------% cover ------------------------------Argentine 90 a A 89 a A 89 a A 93 a A Tifton 7 43 b B 71 bc A 72 b A 83 a A Pensacola 47 b B 81 ab A 79 ab A 86 a A Tifton 9 34 b C 64 c B 65 bc B 82 ab A PCA Cycle 4 37 b C 47 d BC 56 c B 76 b A SE 6 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). In previous work, Pensacola st ands were subjected to frequent clipping and were gradually weakened due to a reduction in root and stolon mass, and subsequent reduction in nonstructural

PAGE 133

133 carbohydrates available for regr owth (Beaty et al., 1970; Youngner, 1972). This occurred only with frequent clipping to a ve ry short stubble in th e current study. Mislevy and Everett (1981) reported Pensacola and Argentine stands harves ted to 10 cm every 30 d had superior stand persistence and minimal weed encroachment than plants harvested to 5 cm. In contrast, in the current study, Argentine was not affected by shor t stubble heights. Adje i et al. (1989) reported that persistence, measured as percent common bermudagrass [ Cynodon dactylon (L.) Pers. var. dactylon] ground cover, of Argentine and Pe nsacola were similar (average 23% common bermudagrass ground cover) following 2 yr of grazing and averaged over four grazing frequencies (2, 4, 6, and 8 wk). Gates et al. (1999) reported no difference in persistence of bahiagrass between cutting heights of 1.5 and 10 cm in a 3-yr trial in Tifton, GA. At the end of the first year of the trial (cli pped every 2, 4, or 8 wk), Pensacol a exhibited greater persistence than Tifton 9, but there were no di fferences between them at the end of the third year. Clearly, in the current study Tifton 9 did not pe rform as well, and these results support a conclusion that the more upright-growing Tifton 9 and Cycle 4 require greater care in defoliation management than Pensacola and Argentine. Specifically, they dem onstrate a need for tall er stubble and for longer intervals between defoliation events. Plant Component Mass Bahiagrass root + rhizom e and stem base mass were measured in October 2006 and 2007, at the end of 2 and 3 yr of defoliation, resp ectively. In October 2006, root + rhizome mass was affected by genotype, and genotype X stubble height and geno type X harvest frequency interactions (Table C-13), while stem base mass was affected by genotype and genotype X harvest frequency interaction (Table C-13). Argentine had greatest root + rhizome mass when harvested to 4-cm stubble height, and there were no differences among the other genotypes (Table 5-10). When harvested to 8-cm

PAGE 134

134 stubble height, Argentine and Pe nsacola had greater root + rh izome mass than all genotypes except Tifton 9, and there were no differences between Tifton 7 and PCA Cycle 4. With the exception of Argentine, stubble height had no eff ect on root + rhizome mass of the genotypes. Table 5-10. Bahiagrass root + rhizome mass as affected by genotype X stubble height interaction in October 2006. Data are means across two harvest frequencies and three replicates (n = 6). Stubble height (cm) Genotype 4 8 --------------------g DM m-2 --------------------Argentine 2530 a A 1620 a B Tifton 7 1560 b A 1530 b A Pensacola 1740 b A 2040 a A Tifton 9 1920 b A 1550 ab A PCA Cycle 4 1620 b A 1510 b A SE 210 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). In October 2006, root + rhizome mass was al so affected by genotype X harvest frequency interaction (Table 5-11). Argen tine had greater root + rhizome mass than all genotypes except Pensacola when harvested every 7 d, while root + rhizome mass of Tifton 9 was less than all genotypes except Tifton 7. When harvested every 21 d, Pensacola and Tifton 9 had greater root + rhizome mass than all genotypes except Argentin e. Tifton 9 had greater root + rhizome mass when harvested every 21 than 7 d, while ther e was no effect of harvest frequency on root + rhizome mass of the other genotypes. Stem base mass was also affected by genotype X harvest frequency interaction in October 2006 (Table 5-12). There were no differences in stem base mass among genotypes when harvested every 21 d, however, Argentine had great er stem base mass than all genotypes except Pensacola when harvested every 7 d, while stem base mass of PCA Cycle 4 was less than all genotypes except Tifton 9. There was no effect of harvest frequency on stem base mass of

PAGE 135

135 Tifton7 and Pensacola; however, Tifton 9 and PCA Cycle 4 had greater stem base mass when harvested every 21 than 7 d, while Argentine had gr eater stem base mass when harvested every 7 than 21 d. Table 5-11. Bahiagrass root + rhizome mass as affected by genotype X harvest frequency interaction in October 2006. Data are m eans across two stubble heights and three replicates (n = 6). Harvest frequency (d) Genotype 7 21 --------------------g DM m-2 --------------------Argentine 2280 a A 1870 ab A Tifton 7 1480 bc A 1610 b A Pensacola 1900 ab A 1880 a A Tifton 9 1240 c B 2220 a A PCA Cycle 4 1510 b A 1610 b A SE 210 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Table 5-12. Bahiagrass stem base mass as affect ed by genotype X harvest frequency interaction in October 2006. Data are means across two st ubble heights and thr ee replicates (n = 6). Harvest frequency (d) Genotype 7 21 --------------------g DM m-2 --------------------Argentine 680 a A 500 a B Tifton 7 470 bc A 610 a A Pensacola 620 ab A 550 a A Tifton 9 410 cd B 490 a A PCA Cycle 4 290 d B 510 a A SE 60 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). In October 2007, there were no interactions with genotype on root+ rhizome or stem base mass (Table C-13). Bahiagrass genotype affected root + rhizome mass (P = 0.01) and stem base mass ( P = 0.002) (Table 5-13). Argentin e had greater root + rhizome and stem base mass than all genotypes except Pensacola, while PCA Cycle 4 had lesser stem base mass than all genotypes

PAGE 136

136 except Tifton 7. Gates et al. ( 1999) reported Pensacola produced a greater root mass than Tifton 9 (33.6 and 21.4 g, respectively) from cores measuring 75 mm in diameter X 75 mm deep on plots receiving 168, 18, and 70 kg ha-1 yr-1 of N, P, and K, respectively. Pedreira and Brown (1996b) reported that Pensacola had more rhizomes per plant th an Tifton 9. Average October rhizome + stubble masses were 1000 and 810 g DM m-2 for Pensacola and Tifton 9, respectively, when harvested every 14 d to 3.5or 10-cm stubble height and fertilized with 200 kg N ha-1 annually (Pedreira and Brown, 1996a). Beaty et al. (1970) indicated that av erage stolon and root mass were 650 and 170 g DM m-2, respectively, from cores measuring 15 cm in diameter X 20 cm deep for Pensacola fertilized with 112 kg N ha-1 annually and clipped to soil level every 1 to 6 wk. Likewise, Beaty and Tan (1972) repor ted Pensacola produced 500 and 1130 g DM m-2 of roots and stolons, respectively, on unfertilized deep sand in Georgia. Be aty et al. (1975) found that Pensacola produced 940 and 1400 g DM m-2 of roots at 0 and 224 kg N ha-1, respectively, in Georgia from 15-cm diameter by 20-cm depth soil cores. Table 5-13. Bahiagrass root + rhizome and stem base mass as affected by genotype in October 2007. Data are means across two stubble height s, two harvest frequencies, and three replicates for each component (n = 12). Plant component Genotype Root + rhizome mass Stem base mass ----------------------------g DM m-2 ----------------------------Argentine 1660 a 320 a Tifton 7 910 b 200 bc Pensacola 1260 ab 250 ab Tifton 9 1100 b 220 b PCA Cycle 4 950 b 170 c SE 180 30 Means within a column followed by the same letter do not differ by the LSMEANS test ( P > 0.05). In the current study, bahiagrass root + rh izome mass was also affected by harvest frequency ( P = 0.001) in October 2007. Root + rhizome ma ss was greater at 21than 7-d harvest

PAGE 137

137 frequency (1440 and 915 g DM m-2, respectively). Weekly clipping of perennial ryegrass to 4 cm reduced root biomass compared to clipping every 8 wk, which was attributed to a shortage of carbohydrates available for root gr owth (Dawson et al., 2000). Frequent defoliation may result in translocation of carbohydrate reserves from roots, rhizomes, and leaf bases to the shoot for new leaf development to compensate for deficien cy of photosynthate created by defoliation (Youngner, 1972). Grazing by cattle has also b een shown to reduce relative number of orchardgrass roots compared to ungrazed plants (Karl and Doescher, 1991). Gates et al. (1999) reported harvest frequency of bahiagrass had no e ffect on root mass when plots were harvested to a 7.5-cm stubble. The reduction in root + rh izome mass in the current study was associated with frequent clipping and probably due to re duction in stored carbohydr ates, which may have been translocated and utilized for growth of new leaves. Bahiagrass stem base mass was al so affected by stubble height ( P = 0.0002) in October 2007. Stem base mass was greater at 8than at 4-cm stubble height (279 and 182 g DM m-2, respectively). This was expected because stem base was defined as stem above the target stubble height. These root + rhizome and stem base mass data suggest that PCA Cycle 4 shows tendencies that may cause it to be less persiste nt than Argentine and Pensacola under defoliation, and may be indicative of lack of tolerance to severe defoliation. Argentine generally showed superior persistence traits under severe defoliation. Reserve Pools Nitrogen and TNC conc entration and content we re measured on 1) root + rhizome and 2) stem base components at the end of 2 yr (Oct ober 2006) to determine th e effects of defoliation treatments on bahiagra ss persistence.

PAGE 138

138 Nitrogen concentration Nitrogen concentration was not affected by ge notype for either stem base or root + rhizome component ( P > 0.05; Table C-14). Stem base and r oot + rhizome N concentration were affected similarly by harvest frequency ( P < 0.0001 and P = 0.02, respectively). There was greater N concentration when harvested ever y 7 than 21 d in stem base (13.9 and 12.9 g kg-1, respectively) and in root + rhizome (12.4 and 11.9 g kg-1, respectively). Macoon et al. (2002) reported rhizome N concentration of four Pennisetum sp. was greater when harvested every 6 than 12 wk (18.3 and 9.5 g kg-1, respectively). Stem base N concentration was also greater ( P = 0.004) when harvested to 4than 8-cm stubble height (13.7 and 13.1 g kg-1, respectively). Total nonstructural carbohydrate concentration Stem base TNC concentration was greater ( P = 0.004; Table C-15) when harvested every 21 than 7 d (48.6 and 42.0 g kg-1, respectively). Root + rhizome TNC concentration was affected by genotype X harvest frequency interaction (Tab le C-15). When harvested every 7 d, Tifton 7 had less root + rhizome TNC c oncentration than all genotype s except Tifton 9 (Table 5-14). Argentine had less root + rhizome TNC concen tration than all genot ypes except PCA Cycle 4 when harvested every 21 d. There was an effect of harvest frequency on genotype root + rhizome TNC concentration only for Tifton 7 (Table 5-14) which exhibited greater TNC concentration when harvested every 21 than 7 d. A simila r trend was observed for Tifton 9 and Cycle 4. Mislevy et al. (1991) reported little difference in bahiag rass crown TNC concentration among grazing frequencies (2 to 7 wk) during the fi rst grazing year in Ona, FL. However, they reported crown TNC concentration was always greater for the 3, 5-, and 7than the 2-wk grazing frequencies in the sec ond and third grazing years. Th ey concluded that even though bahiagrass persists well under fr equent close grazing, plants a ppeared stressed, as reflected by lower TNC, at the 2-wk grazing frequency. Ad jei et al. (1989) found no effect of grazing

PAGE 139

139 frequency (2 to 8 wk) on Argentine and Pensacola crown TNC concentration after 2 yr in Ona, FL, but Pensacola had greater crown TNC concen tration (stockpiled from December to May) than Argentine (115 and 95 g kg-1, respectively). In the current study, the frequency effect or a trend toward a frequency effect was limited to the more upright genot ypes, Tifton 9, Tifton 7, and PCA Cycle 4, suggesting a relationship be tween morphology and the ability to sustain similar levels of TNC concentration acr oss a range of harvest frequencies. Table 5-14. Bahiagrass root + rhizome total nonstr uctural carbohydrate concen tration as affected by genotype X harvest frequency interaction in October 2006. Data are means across two stubble heights and thr ee replicates (n = 6). Harvest frequency (d) Genotype 7 21 --------------------g kg-2 ---------------------Argentine 44.8 a A 45.9 b A Tifton 7 30.7 b B 69.7 a A Pensacola 56.4 a A 52.1 a A Tifton 9 39.2 ab A 54.6 a A PCA Cycle 4 41.2 a A 50.0 ab A SE 7.3 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Nitrogen content There were more effects of genotype and in teractions with genotype on N content than concentration. Both root + rhizom e and stem base N content were affected by genotype X harvest frequency interaction (Table C-14; Ta ble 5-15). Differences among genotypes were due to root + rhizome and stem base mass as N concentration was not affected by genotype. When harvested every 7 d, Argentine had greater root + rhizome and stem base N content than all genotypes except Pensacola. At the 7d harvest frequency, Tift on 9 had less root + rhizome N content than all genotypes except Ti fton 7, while PCA Cycle 4 had less stem base N content than all genotypes except Tifton 9. There were no differences in stem base N content

PAGE 140

140 among genotypes when harvested ev ery 21 d; however, Pensacola and Tifton 9 had greater root + rhizome N content than Tifton 7 and PCA Cycle 4. Root + rhizome and stem base N content of Pensacola and Tifton 7 were not affected by ha rvest frequency. Neither the root + rhizome N content of PCA Cycle 4 nor the stem base N content of Tifton 9 were affected by harvest frequency. Argentine had greater root + rhizome and stem base N content when harvested every 7 than 21 d. Tifton 9 had greater root + rhizom e N content and PCA Cycle 4 had greater stem base N content at 21than 7-d harvest frequency. Table 5-15. Bahiagrass root + rhizome and stem base nitrogen cont ent as affected by genotype X harvest frequency interactions in Oct ober 2006. Data are means across two stubble heights and three replicates for each component (n = 6). Plant component Root + rhizome Stem base Genotype Harvest frequency (d) 7 21 7 21 -------------------------------g N m-2 -------------------------------Argentine 27 a A 22 ab B 10 a A 6 a B Tifton 7 19 bc A 19 b A 6 bc A 8 a A Pensacola 23 ab A 23 a A 8 ab A 7 a A Tifton 9 15 c B 26 a A 6 cd A 6 a A PCA Cycle 4 19 b A 18 b A 4 d B 6 a A SE 2 1 Means within a plant component followed by th e same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test (P > 0.05). Root + rhizome N content was also affected by genotype X stubble height interaction (Table C-14; Table 5-16). Argen tine had greatest root + rhizome N content when harvested to 4cm stubble height, and there were no differences among the othe r genotypes. At 8-cm stubble height, Pensacola had greatest root + rhizome N content, and there we re no differences among the other genotypes. There was no effect of stubble height on root + rhizome N content of genotypes, with the exception of Argentine whic h exhibited greater root + rhizome N content when harvested to 4than 8-cm stubble height.

PAGE 141

141 Table 5-16. Bahiagrass root + rhizome nitrogen content as affected by genotype X stubble height interaction in October 2006. Data are mean s across two harvest frequencies and three replicates (n = 6). Stubble height (cm) Genotype 4 8 --------------------g N m-2 ---------------------Argentine 29 a A 19 b B Tifton 7 19 b A 18 b A Pensacola 21 b A 25 a A Tifton 9 23 b A 19 b A PCA Cycle 4 20 b A 17 b A SE 2 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Total nonstructural carbohydrate content Stem base TNC content was affected by ge notype and harvest frequency (Table C-15). Argentine, Pensacola, and Tifton 7 had simila r stem base TNC content (29, 27, and 25 g m-2, respectively) that was greater ( P = 0.004) than Tifton 9 and PCA Cycle 4 (18 and 17 g m-2, respectively). Differences among genotypes were due to stem base mass as TNC concentration was not affected by genotype. Argentine, Pensacola and Tifton 7 had an average of 47 and 59% greater stem base TNC conten t than Tifton 9 and PCA Cycle 4, respectively, suggesting that Tifton 9 and PCA Cycle 4 have less stored energy for regrowth than Argentine, Pensacola, and Tifton 7. Stem base TNC content was greater (P = 0.03) at 21than 7-d harvest frequency (25 and 21 g m-2, respectively), a function of greater stem base TNC c oncentration at 21 than 7 d. Root + rhizome TNC content was affected by stubble height and genotype X harvest frequency interaction (Table C-15). Root + rhizome TNC content was greater (P = 0.02) at 4than 8-cm stubble height (96 and 76 g m-2, respectively). Pedreira et al. (2000) reported TNC content in the stubble + rhizome component of Florakirk bermuda grass decreased linearly as stubble height increased, which was attributed to greater rhizome mass at lower stubble height.

PAGE 142

142 Argentine and Pensacola had gr eater root + rhizome TNC cont ent than the other genotypes when harvested every 7 d, while PCA Cycle 4 ha d lesser root + rhizome TNC content than all genotypes except Argentine when harvested ever y 21 d (Table 5-17). Argentine and Pensacola had an average of 98% greater root + rhizom e TNC content than Tifton 7, Tifton 9, and PCA Cycle 4, suggesting that the upri ght types Tifton 7, Tifton 9, and PCA Cycle 4 have less stored energy for regrowth than Argentine and Pensacola Root + rhizome TNC content of Argentine, Pensacola, and PCA Cycle 4 was not affected by harvest frequency, while Tifton 7 and Tifton 9 had greater root + rhizome TNC content when harvested every 21 than 7 d. Table 5-17. Bahiagrass root + rhizome total nonstructural carbohydrate c ontent as affected by genotype X harvest frequency interaction in October 2006. Data are means across two stubble heights and three replicates (n = 6). Harvest frequency (d) Genotype 7 21 --------------------g TNC m-2 ---------------------Argentine 110 a A 90 ab A Tifton 7 50 b B 110 a A Pensacola 100 a A 100 a A Tifton 9 50 b B 120 a A PCA Cycle 4 60 b A 80 b A SE 10 Means followed by the same letter, lower-c ase letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). These N and TNC data indicate that PCA Cycl e 4 generally has less stored N and TNC at the end of 2 yr than Argentine and Pensacola when defoliated frequently, but is generally similar to the other diploids when defoliated to a s hort stubble height. The stem base N content of Pensacola and Tifton 9 at the e nd of 2 yr was not affected by harvest frequency; however, PCA Cycle 4 had less stem base N content when harves ted every 7 than 21 d. These data suggest that PCA Cycle 4 may not be as to lerant to frequent defoliati ons as Pensacola, Tifton 9, and

PAGE 143

143 Argentine because PCA Cycle 4 had less stored energy for regrowth approaching the cool season. Summary and Conclusions Bahiagrass c ultivars including Pensacola, Tift on 9, and Argentine are sensitive to short daylengths, which induces a cha nge in growth pattern resulting in a reduction in above-ground growth during the cool season, even when temper ature, soil moisture, and soil fertility were adequate for substantially greater growth (S inclair et al., 1997). Th e significant economic implications of seasonal bahiagra ss forage shortfall have stimulat ed research aimed at increasing productivity during short-daylen gth months. This effort has involved genetic selection and development of bahiagrass genotypes that are co ld adapted and less sensitive to photoperiod (Blount et al., 2001). One such genotype is PCA Cycle 4. Additional cool-season productivity must not come at the expens e of persistence, however, because the cost of pasture renovation is high and return per unit land area from grasslandlivestock systems are relatively low. There is lim ited information available that characterizes the differences in plant productivity, nutritive valu e, and persistence res ponses between daylengthsensitive and less sensitive types to defoliation ma nagement. This information is needed to assess whether less daylength-sensitive types have merit in production systems. The objectives of this study were to determine the effects of defo liation management on total-season and seasonal herbage yield, nutritive value, cover, plant co mponent mass, and N and TNC concentration and content of stem base and root + rhizome compon ents of two diploid bahi agrass cultivars (Tifton 9 and Pensacola), two tetraploid s (Argentine and Tifton 7), and a cold adapted, less photoperiod sensitive genotype (PCA Cycle 4). In the current study, the tetraplo ids tended to have gr eater total-season he rbage yields than the diploids. Early in the growi ng season, the diploids tended to outyield the tetraploids, but the

PAGE 144

144 tetraploids generally yielded more than the dipl oids for the remainder of the growing season. The lesser spring yields of the tetrap loids are consistent with expect ed results due to their reported slower growth in the spring (Chambliss and Adjei, 2006). PCA Cycle 4 generally had similar total-season and cool-seas on herbage yields as the other dipl oids, however when harvested every 7 d, PCA Cycle 4 had less total-season herbage yi eld than all genotypes. Seasonal herbage yields of PCA Cycle 4 were generally less than or simi lar to the other diploids. Herbage yield of all genotypes declined approaching autumn, but the rate of decline often was slower for PCA Cycle 4 than the other genotypes. There were relatively few differences in nut ritive value among the genotypes, which is in agreement with literature on nutritive value of bahiagrass cultivars (Kalmbacher, 1997; Mislevy and Dunavin, 1993; Chambliss and Adjei, 2006). PC A Cycle 4 generally had similar total-season and seasonal CP concentrations as the other diploids. Argentine te nded to have grea test seasonal CP concentrations during earlyto mid-summ er, while the diploids generally had greater seasonal IVDOM concentrations than the tetraplo ids. PCA Cycle 4 had greater IVDOM than all genotypes but Tifton 9 in 2005 and all genotypes in 2006. Persistence under defoliation management wa s evaluated based on percent cover, plant component mass, and N and TNC content. Ar gentine cover was unaffected by defoliation treatments, but Tifton 9 and PCA Cycle 4 were much more sensit ive to defoliation management. Harvesting every 21 d to an 8-cm stubble resulted in greatest cover for both Tifton 9 and PCA Cycle 4 (82 and 76%, respectively), while cover for both genotypes was less than 40% if harvested every 7 d to 4 cm. PCA Cycle 4 generally had similar root + rhizom e mass as the other diploids at the end of 2 and 3 yr, but it often was less than Argentine. At the end of 2 yr, Argentine and Pensacola had

PAGE 145

145 an average of 98% more root + rhizome TN C than Tifton 7, Tifton 9, and PCA Cycle 4, and Argentine, Pensacola, and Tifton 7 had 59% more stem base TNC than PCA Cycle 4, suggesting PCA Cycle 4 had less stored energy for regrowth approachin g the cool season. From these data, it can be concluded that PCA Cycle 4 is at least as high in nutritive value and perhaps greater in digestibility than othe r genotypes. Management of PCA Cycle 4 will likely be more critical than for Pensacola and Argentine bahiagrass in Florida. Its upright growth habit, lesser stored TNC for regr owth, and greater reduction in plan t cover with close, frequent defoliation than Argentine imply that more inte nsive management practices such as rotational stocking and greater control of stocking rate may be required to maintain the persistence of PCA Cycle 4 under grazing.

PAGE 146

146 CHAPTER 6 TILLER RESPONSES OF BAHIAGRASS GENOTYPES TO DEFOLIATION MANAGE MENT Introduction Tiller dynam ics refers to seasonal patterns in the relative rates of til ler appearance, tiller death, and tiller number (Korte et al., 1982). It is an important aspect of grass growth, and knowledge of tiller dynamics forms a basis for understanding and analyzing the mechanisms behind the variation in yield and persistence of a grass sward with time, climate, and management (Islam and Hirata, 2005). Persistence of a grass pasture is dependent on the ability of plants to maintain a high tiller density and th e ability of individual tillers to maintain live leaves (Hirata and Pakiding, 2001) Under increasing grazing pre ssure (i.e., increasing stocking rate), grass populations tend to adju st their structure so that there is a high density of small tillers per unit area (Bircham and Hodgson, 1983; Grant et al., 1983; Christiansen and Svejcar, 1988). Tillers in grazed pastures tend to be relatively short-lived, with a continual turnover of plant material (Davies, 1988). Bahiagrass ( Paspalum notatum Flgge) is a warm-season perennial grass that forms a highly persistent sward that maintains tiller density and leaf mass under a wide range of management regimes (Beaty et al., 1970, 1977; Hi rata, 1993a, 1993b). Resear ch has shown that bahiagrass tiller density remains stable because tiller s are long-lived with low rates of tiller death despite low rates of tiller a ppearance (Hirata and Pa kiding, 2001). Persistence of bahiagrass has been associated with maintenance of high tiller, rhizome, and root densities (Hirata, 1993b). Research in Japan showed that density charac teristics of Pensacola bahiagrass, such as tiller number and stolon length, tended to increas e as 2to 4-wk interval cutting heights decreased from 22 to 2 cm (Hirata, 1993b). Paki ding and Hirata (2003) re ported that Pensacola responded to low N (5 g m-2 yr-1) and intense defoliation (2-c m stubble at monthly harvest

PAGE 147

147 intervals) with increased tiller longevity compared to high N (20 g m-2 yr-1) and 12and 22-cm stubble heights (SH). Tiller app earance rate (TAR) a nd tiller density were also greater under intense defoliation. Commonly used bahiagrass cultivars such as Pens acola and Tifton 9 are sensitive to short daylengths, which induce a change in growth pattern resulting in a reduction in above-ground growth at the beginning of the c ool season. During this time, remaining live herbage continues to photosynthesize, but instead of using photosynthate to produce new above-ground tissue, plants store photosynthate as nonstr uctural carbohydrates in stem bases, stolons, and rhizomes (Thornton et al., 2000). At this phys iological growth stage, plants are effectively preparing for winter survival and early spring regrowth, times when photosynthetic produc tion is inadequate to meet growth demands (Thornton et al., 2000). Limited bahiagrass herbage accumulation during the cool season has occurred even when temper ature, soil moisture, and soil fertility were adequate for substantially greater growth (Sinclair et al., 1997). The significant economic implications of s easonal bahiagrass forage shortfall have stimulated research aimed at in creasing productivity during shortdaylength months. This effort has involved genetic selection and development of bahiagrass cultivars that are cold-adapted and less sensitive to photoperiod (Blount et al., 2001). Additional cool-season productivity must not come at the expense of persistence, however, because the cost of past ure renovation is high and return per unit land area from grassland-livestock systems are relatively low. Tiller dynamics of new genotypes may be an important factor aff ecting their response to defoliation and their persistence. There are no known studies evaluating tiller responses of less photoperiod sensitive, coldadapted bahiagrass genotypes to defoliation. Theref ore, the objectives of this study were to

PAGE 148

148 determine the effects of defoliati on frequency, referred to as harv est frequency in this study, and defoliation intensity, referred to as stubble height on tiller number, tiller number change, tiller mass, TAR, and tiller de ath rate (TDR) of bahiagrass genotypes. Materials and Methods Experimental Site This field study was conducted at the Beef Res earch Unit, northeast of Gainesville, FL, at 29 N latitude. Soils were classified as Spodosols (loam y, siliceous, subactive, thermic Grossarenic Paleaquults from the Plummer se ries or sandy, siliceous, hyperthermic Ultic Alaquods from the Pomona series) with average pH of 6.8. Average Mehlich-I extractable soil P, K, Mg, and Ca concentrations at the beginni ng of the experiment were 14, 25, 82, and 731 mg kg-1, respectively. Treatments and Design This experim ent evaluated the effects of tw o harvest frequencies and two stubble heights on bahiagrass tiller responses in 2005 and 2006. Treatments (n = 20) were the factorial combinations of two harvest frequencies, two st ubble heights, and five bahiagrass genotypes arranged in three replications of a randomized complete block design. Sexual diploid and apomictic tetraploid ge notypes of bahiagrass were evaluated. Sexual diploid genotypes were Pensacola, Tifton 9, a nd PCA (photoperiod and cold-adapted) Cycle 4. Apomictic tetraploid genotypes evaluated were Argentine and Tifton 7. Pensacola, Tifton 9, and Argentine are commonly utilized bahiagrass cultivars in Florida, while PCA Cycle 4 is a novel genotype selected to improve cold tolerance, photoperiod response, nematode and disease resistance, rooting/rhizome mass, seedling vigor and establishment, s easonal distribution of forage production, and forage quality (Blount et al., 2001). PCA Cycle 4 has been approved for release by the University of Florida Agricultural Experiment Station as cv. UF-Riata (Blount et

PAGE 149

149 al., in review). Tifton 7 is a noncommercial apom ictic genotype. Harvest frequencies were 7 and 21 d, and stubble heights were 4 and 8 cm. Defoliation treatments were selected to impose a range of stress on the five genotypes. Some treatme nt combinations were quite intensive because bahiagrass cultivars such as Pensacola are valued due to their tolerance of close, frequent defoliation. Plots were planted in July 2004 using seedlings in speedling flats. Plan ts were arranged in 2 x 1 m plots in a 10 plant X 5 plant grid and were spaced 20-cm apart within and between rows. Plots were undefoliated throughout 2004, and 50 kg N ha-1 was applied on 28 July and 28 August. During each of the subsequent trial years, plots were fertilized in March with 40 kg N, 18 kg P, and 66 kg K ha-1. Plots were staged in late April to early May and fertilized with 40 kg N ha-1 immediately after staging. Six weeks later, plots were fertilized with 40 kg N, 18 kg P, and 66 kg K ha-1. Plots were then fertilized with 40 kg N ha-1 for the two remaining 6-wk cycles until the end of the trial. Total season fertilizer amounts during the trial were 200 kg N, 36 kg P, and 132 kg K ha-1 yr-1, and are considered lower than the IFAS recommendation for bahiagrass grown only for hay (90 kg N and 37 kg K ha-1 applied after each cutting; Mylavarapu et al., 2007). Response Variables Response variables m easured include tiller num ber, tiller number cha nge, tiller mass, and tiller appearance and de ath rates. Tiller number was determined monthly from May through October 2005 and 2006 by counting live and dead tillers inside anchored 15-cm diameter quadrats (2 quadrats plot-1). On the first evaluation date of e ach year, individual live tillers within one quadrat plot-1 were marked with plastic tags of the same color. On the subsequent evaluation date, counts were made of live ti llers (previous tags remain on), dead tillers (tags removed), and new live tillers (marked with diffe rent colored tags). Ti llers were classified as dead when all

PAGE 150

150 parts were completely brown. Counting and ma rking procedures cont inued at subsequent evaluation dates using a to tal of six different colors of tags Tiller number change was calculated as the difference in live tiller number between two evaluation dates for two time intervals of interest: 1) Spring 2005 to Fall 2005 and 2) Spri ng 2005 to Fall 2006. For example, tiller number change for the first time interval was calculate d by subtracting Spring 2005 tiller number from that in Fall 2005. Tiller mass was measured three times per growing season (late spring, summer, and autumn) each year. Samples were taken after a harvest event by clipping 20 randomly selected tillers per plot to ground level. Tillers were dr ied at 60C and weighed to determine tiller mass. Tiller appearance rate was calculated mont hly by dividing the number of new tillers marked at each evaluation date by 28 d. Tiller death rate was calculated monthly by summing the number of dead tillers count ed at each evaluation date a nd dividing by 28 d. Net TAR was calculated by subtracting TDR from TAR for each evaluation date. Statistical analyses were performed using Proc Mixed of SAS (S AS Inst. Inc., 1996). Genotype, stubble height, harvest frequency, year, and their interactions were considered fixed effects and replicate and its inte ractions random effects. Year wa s considered fixed because of the potential for carryover effects of treatments fr om Year 1 to Year 2. Year was included in the model as a subplot treatment in a split-plot arrangement, with th e treatment combinations being the main plots. The PDIFF function of the LS MEANS procedure was used to compare genotype means. Differences among stubble height and ha rvest frequency means were based on F tests. When treatment by year interacti on was significant, data were anal yzed and are reported by year. For responses measured multiple times per year sampling date was treated as a repeated measure. Significance was determined at P 0.05.

PAGE 151

151 Results and Discussion Tiller Number and Tiller Number Change Tiller number Tiller num ber was evaluated at three dates during the growing season (late spring, summer, and autumn) that also corresponded to tiller mass evaluation dates. Tiller number was affected by genotype, evaluation date, year, and stubble height X harvest fr equency, stubble height X year, harvest frequency X year, evaluation date X year, genotype X stubble height X harvest frequency, and stubble height X harvest frequency X evaluation date interactions (Table D-1). Interactions involving genotype we re of primary concern in this study so the genotype X stubble height X harvest frequency interaction will be presented and discussed. Tifton 7 had lesser tiller number than all genotypes when harvested to 4 cm every 7 d, and there were no differences among the other genotype s (Table 6-1). When harvested to 4 cm every 21 d, Tifton 7 and PCA Cycle 4 had lesser tiller number than all ge notypes except Tifton 9. There were no differences between Argentine and Pensacola. Tifton 7 and Tifton 9 had lesser tiller number than all genotypes except PCA Cycl e 4 when harvested to 8 cm every 7 d, and there were no differences between Argentine an d Pensacola. When harvested to 8 cm every 21 d, Tifton 7 had lesser tiller numb er than all genotypes except Tifton 9, and there were no differences among Argentine, Pensacola, and PC A Cycle 4. Harvest frequency had no effect on tiller numbers of Argentine, Tifton 7, and Pensac ola across stubble heights. Beaty et al. (1977) reported no differences in Pensacola tiller numbe rs between 2.5and 7.5-cm stubble heights. Tifton 9 had more tillers when harvested to 4 cm every 7 d than the other defoliation treatments, and there were no differences in tiller nu mber for Tifton 9 among the other defoliation treatments. PCA Cycle 4 had greater tiller number wh en harvested to 4 cm every 7 d and to 8 cm

PAGE 152

152 Table 6-1. Bahiagrass tiller number as affect ed by genotype X stubble height X harvest frequency interaction. Data are means acro ss two years, three ev aluation dates, and three replicates (n = 18). Stubble height (cm) 4 8 Harvest frequency (d) Genotype 7 21 7 21 --------------------------------# m-2 --------------------------------------Argentine 2440 a A 2050 a A 2170 a A 2100 a A Tifton 7 1660 b A 1790 b A 1650 b A 1620 b A Pensacola 2380 a A 2300 a A 2240 a A 2090 a A Tifton 9 2640 a A 1960 ab B 1670 b B 1960 ab B PCA Cycle 4 2300 a A 1570 b B 1990 ab AB 2040 a A SE 160 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). every 21 d than when harvested at 4 cm every 21 d. The results for Tifton 9 and PCA Cycle 4 at close, frequent harvests are consistent with research showing that as defoliation pressure increased, tiller number and density increased. Hirata (1993b) and Pakidi ng and Hirata (2003) reported Pensacola tiller density tended to incr ease with decreasing cutting height from 22 to 2 cm. Bircham and Hodgson ( 1983) reported tiller density of perennial ryegrass ( Lolium perenne L.), annual bluegrass ( Poa annua L.), and white clover (Trifolium repens L.) in a mixed-species sward was greatest when herbage mass was maintained at 700 kg OM ha-1 compared to herbage masses of 1000 and 1700 kg OM ha-1. They attributed the decline in tiller density with increasing herbage mass above 700 kg OM ha-1 primarily to competition. Gran t et al. (1983) reported white clover tiller density was greatest when swards were maintained between a 2and 3-cm stubble height and declined in swards maintained above and below this range. Christiansen and Svejcar (1988) reported Cau casian bluestem [ Bothriochloa caucasia (Trin.) C.E. Hubb.] tiller density was greater in heavily grazed (7.5 cm) than lightly grazed (34-cm stubble height) pastures. They attributed the increase in tiller number with heavy grazing to diffe rences in light intensity and

PAGE 153

153 quality at the base of the sward and the release of apical dominance. Increased tiller appearance has been associated with an increased red:far-red ratio of incident light. As canopy density increases, the proportion of inci dent radiation intercepted increas es and the red:far-red ratio at plant bases decreases, ther eby reducing tillering (Casal et al., 1985). Dallisgrass [ Paspalum dilatatum (Poir.)] and annual ryegrass [ Lolium multiflorum (Lam.)] grown at high densities (19 plants m-2 for dallisgrass and 116 plants m-2 for annual ryegrass) exhibi ted a reduction in tillering rate compared to plants grown at low densities (7 plants m-2 for dallisgrass and 40 plants m-2 for annual ryegrass) which was attribut ed to decreased light interception per plant and lower red:farred light (Casal et al., 1986). Tiller number change Tiller num ber change was assessed for both Spring 2005 to Fall 2005 and Spring 2005 to Fall 2006 time intervals. The response was affected by genotype from Spring 2005 to Fall 2005 and Spring 2005 to Fall 2006 (Table 6-2), and by stubble height X harvest frequency interaction from Spring 2005 to Fall 2006 (Table D-2; Tabl e 6-3). From Spring 2005 to Fall 2005, Tifton 9 and PCA Cycle 4 showed a reduc tion in tiller number, Argentine and Pensacola showed an increase in tiller number, and Tifton 7 tille r number change was intermediate. The three genotypes with a decrease in tiller s in Year 1 are all mo re upright-growing types (Pedreira et al., 1996; Blount et al., in review) that may show less short-term phenotypic plasticity in response to defoliation than lower growing t ypes like Pensacola and Argentine. Defoliated grasses exhibit tiller mass/density compensation in response to grazi ng intensity, with closer grazing resulting in a high tiller population density th at is associated with small tillers (Matthew et al., 1995). Sbrissia et al. (2003) suggested th at there are limits to plastic ity depending on grass species and growing environment. They found that during so me seasons a 5-cm grazing height was below

PAGE 154

154 Table 6-2. Bahiagrass tiller number change as a ffected by genotype during the first year of defoliation (2005) and from the beginning of the first year of defoliation (Spring 2005) to the end of the sec ond year (Fall 2006). Data ar e means across two harvest frequencies, two stubble heights, and three replicat es (n = 12). Time interval Genotype Spring 2005-Fall 2005 Spring 2005-Fall 2006 --------------------------# m-2 --------------------------Argentine 60 a 830 b Tifton 7 -100 ab 730 b Pensacola 40 a 1600 a Tifton 9 -320 b 950 b PCA Cycle 4 -360 b 960 b SE 120 180 Means within a column followed by the same letter do not differ by the LSMEANS test ( P > 0.05). the plasticity limit for Tifton 85 bermudagrass, i.e., both tiller number and mass were lowest in a sward grazed to a 5-cm height compared to taller swards. Chapman and Lemaire (1993) suggested the plasticity limitation was due to the inability of some grasses to reduce sheath length to maintain leaf material below a part icular defoliation height. Matthew et al. (1995) suggested that under very severe grazing some grasses may lack carbohydrate rese rves for tiller appearance to increase tiller population density and pastures may collapse. From the beginning of the first to the end of the second year, all genotypes increased tiller number and this response was greater for Pensacola than for the other cultiv ars. Thus the loss in tiller number during Year 1 for the more uprig ht-growing cultivars was compensated for by increases in Year 2, such that til ler number change in these swards was as great as in Argentine but remained less than Pensacola. This suggest s that morphological adaptation occurred over time, even among the upright-growing genotypes. Tiller number change was affected by stubble height by harvest frequency interaction in Spring 2005 to Fall 2006 (Table D-2; Table 6-3) From Spring to Fall 2006, there was no stubble

PAGE 155

155 height difference in tiller numb er change for the 21-d harvest frequency, while there was a greater increase in tiller number for 7-d harvest frequency at the 4than the 8-cm stubble height. Table 6-3. Bahiagrass tiller number change as a ffected by stubble heig ht X harvest frequency interaction from the beginning of the first year of defoliation (Spring 2005) to the end of the second year (Fall 2006). Data ar e means across five genotypes and three replicates (n = 15). Harvest frequency (d) Stubble height (cm) 7 21 ---------------------# m-2 ----------------------4 1480 a A 800 a B 8 870 b A 910 a A SE 170 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). There was no difference in tiller number change between harvest frequencies at the 8-cm stubble height, while there was a greater increase in tiller number for 4-cm stubble height at the 7vs. 21-d harvest frequency. The largest increase in tiller number occurred for the most intensive defoliation treatment (4-cm stubble height, 7-d ha rvest frequency). This longer-term response is consistent with the tiller dens ity component of the tiller mass/density compensation described by Matthew et al. (1995). Tiller Mass Tiller m ass was evaluated at three times dur ing each growing season (late spring, summer, and autumn). Tiller mass was affected by genotyp e, stubble height, harvest frequency, evaluation date, year, and genotype X stubble height, stubble height X eval uation date, stubble height X year, harvest frequency X evalua tion date, evaluation date X y ear, genotype X stubble height X evaluation date, and stubble height X harvest frequency X evaluation date X year interactions (Table D-3). Interactions involvi ng genotype were of primary concer n in this study so data were

PAGE 156

156 analyzed by evaluation date to further explore the genotype X stubble height X evaluation date interaction. On 2 June, tiller mass was affected by genotype (P < 0.0001; Table D-4). As a group, the tetraploids had greater tiller mass than the dipl oids. Tifton 7 had greate st tiller mass (2.9 g), while Pensacola, Tifton 9, and PCA Cycle 4 had lesser tiller mass (2.3, 2.2, and 2.1 g, respectively) than other genotypes. Tiller mass of Argentine (2.7 g) was intermediate. The responses associated with ploidy can be related to the larger cells and characteristically larger leaves and stems of tetraploids than diploids (Gates et al., 2004). Tiller mass was also affected by stubble height ( P < 0.0001; Table D-4) on 2 June, with greater tiller mass occurring at 8-cm than 4cm stubble height (3.1 and 1.8 g, respectively). These results are consistent with expectations because tillers were harvested from the stubble height to soil level and as su ch were twice as tall for the 8than the 4-cm stubble height. Tiller mass was affected by genotype X stubbl e height interaction on 27 July and 27 September (Table 6-4; Table D-4). On 27 Jul y, Argentine had greater tiller mass than all genotypes except Tifton 7 and Pensacola at th e 4-cm stubble height, while PCA Cycle 4 had lesser tiller mass than all genot ypes except Tifton 9. At the 8-cm stubble height, the tetraploids had greater tiller mass than the di ploids. All genotypes had greater ti ller mass at an 8than 4-cm stubble height. On 27 September, Argentine, Tifton 7, and Pensacola had greater tiller mass than Tifton 9 at a 4-cm stubble height. Tifton 9 had lesser tiller mass than all genotypes except PCA Cycle 4. At an 8-cm stubble height, the tetraploids had gr eater tiller mass than the diploids. PCA Cycle 4 had lesser tiller mass than all genotypes except Tifton 9. With th e exception of PCA Cycle 4, all

PAGE 157

157 genotypes had greater tiller mass associated with 8than 4-cm stubble height. There was no effect of stubble height on th e tiller mass of PCA Cycle 4. Table 6-4. Bahiagrass tiller mass as affected by genotype X stubble height interactions on 27 July and 27 Sept. 2005. Means for both ev aluation dates are across two harvest frequencies and three replicates (n = 6). Evaluation date 27 July 27 September Stubble height (cm) Genotype 4 8 4 8 --------------------------------------g --------------------------------------Argentine 1.3 a B 1.8 a A 2.1 a B 3.8 a A Tifton 7 1.0 ab B 2.0 a A 2.3 a B 3.7 a A Pensacola 1.0 ab B 1.2 b A 1.9 a B 2.8 b A Tifton 9 0.8 bc B 1.1 b A 1.6 b B 2.4 bc A PCA Cycle 4 0.6 c B 1.1 b A 1.7 ab A 2.1 c A SE 0.1 0.2 Means within an evaluation date followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). Tiller Dynamics Tiller appearance rate Tiller appearance rate w as qua ntified monthly in six months of both 2005 and 2006. This response was affected by genotype, evaluation date and genotype X stubble height, genotype X evaluation date, genotype X year, stubble height X year, evaluation date X year, genotype X stubble height X evaluation date, genotype X ha rvest frequency X year, genotype X evaluation date X year, genotype X stubble height X harves t frequency X evaluation date, and genotype X stubble height X harvest frequency X evaluation date X year interactions, so data were analyzed by year (Table D-5). Tiller appearance rate in 2005 was affected by genotype ( P = 0.02; Table D-6). Argentine had greater TAR than all genotypes ex cept Pensacola (4.9 and 4.7 tillers m-2 d-1, respectively). Tifton 9 and Tifton 7 had lesser TAR than all genotypes except PCA Cycle 4 (3.3, 2.5, and 3.4

PAGE 158

158 tillers m-2 d-1, respectively). Pensacola and PCA Cycle 4 had similar TAR (4.7 and 3.4 tillers m-2 d-1, respectively). Tiller appearance rate in 2005 wa s also affected by evaluation date (Table D-6; Figure 61). Tiller appearance rate was greatest on 29 May 2005 (8.9 tillers m-2 d-1) and decreased throughout the growing season, with rates as low as 2.2 and 1.6 tillers m-2 d-1 in September and October, respectively. Hirata and Pakiding (2001) reported peak TAR in late spring or early summer (8.6 and 9.0 tiller m-2 d-1, respectively) in Pensacola bahiagrass pastures grazed by cattle, and lowest TAR in autumn and winter (5.5 and 1.7 tillers m-2 d-1, respectively). Figure 6-1. Bahiagrass tiller a ppearance rate as affected by evaluation date in 2005 and 2006. Means within year followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Data are means across five genotype s, two harvest frequencies, two stubble heights, and three replicates (n = 60). Data for 2006 are not compared statistically because of treatment interactions with eval uation date. Interactions are explored in Tables 6.5 and 6.6. Tiller appearance rate in 2006 was affected by genotype, stubble height, and genotype X stubble height, genotype X harvest frequency, ge notype X evaluation date, harvest frequency X evaluation date, genotype X st ubble height X evaluation date, and genotype X stubble height X 0 1 2 3 4 5 6 7 8 9 10 29 May26 June24 July21 August18 September16 October Evaluation dateTiller appearance rate (# tillers m-2d-1) 2005 2006ab b bc c c

PAGE 159

159 harvest frequency X evaluation date interactions so data were analyzed by evaluation date within year (Table D-6). For comparison to 2005, evaluation date main effect means for 2006 are shown in Figure 6-1. Mean comparisons were not made due to interactions of other factors with evaluation date. In 2006, there was genotype X stubble height X harvest frequency interaction on 29 May, genotype X harvest frequency interaction on 25 Ju ly, and genotype X stubble height interaction on 21 August. On 29 May, Tifton 9 had greatest TA R when harvested to 4 cm every 7 d (Table 6-5). When harvested to 4 cm every 21 d, Pensac ola TAR was greater than all genotypes except Tifton 9, and Tifton 7 had lesser TAR than all genotypes except Argentine and PCA Cycle 4. Tetraploid bahiagrasses such as Argentine and Tift on 7 are slower to initiat e growth in the spring than diploids Pensacola and Tifton 9 (Chambliss and Adjei, 2006), and the lower TAR exhibited by Argentine and Tifton 7 in the late spring when harvested to 4 cm every 21 d is consistent with this growth characteristic. Th ere were no differences in TAR among genotypes when harvested to 8 cm every 7 or every 21 d, although the tetr aploids continued to rank numerically lowest. Tiller appearance rates of Argentine, Tifton 7, and PCA Cycle 4 were not affected by harvest frequency across stubble heights. Pensacola had greater TAR when harvested to 4 cm every 21 than the other defoliation treatments, while Tifton 9 had greater TAR when harvested to 4 cm every 7 d. Tiller appearance rate generally increa ses as stubble height de creases to a certain level. Pakiding and Hirata (2003) reported Pensacola had greater TAR at 2than at 12and 22cm cutting heights. Frequent in tense and continuous grazing have been shown to stimulate TAR resulting from increased tiller bud sites and greater light penetr ation to the base of the canopy (Christiansen and Svejcar, 1988; McKenzie, 1997) However, Korte et al (1982) reported TAR

PAGE 160

160 of a tetraploid hybrid ryegrass (Lolium x hybridum ) was not affected by stubble height (1-2 and 5-6 cm). Table 6-5. Bahiagrass tiller appearance rate as a ffected by genotype X stubble height X harvest frequency interaction on 29 May 2006. Data are means across three replicates (n = 3). Stubble height (cm) 4 8 Harvest frequency (d) Genotype 7 21 7 21 ------------------------------# tillers m-2d-1 -------------------------------Argentine 2.7 b A 3.4 bc A 4.0 a A 0.7 a A Tifton 7 4.0 b A 1.3 c A 0.0 a A 2.0 a A Pensacola 5.4 b B 13.5 a A 5.4 a B 3.4 a B Tifton 9 23.5 a A 10.1 ab B 6.7 a B 6.1 a B PCA Cycle 4 8.7 b A 3.4 bc A 4.0 a A 6.1 a A SE 2.5 Means followed by the same letter, lower-case letters within a column and upper-case letters within a row, do not differ by the LSMEANS test ( P > 0.05). On 26 June, TAR was affected by stubble height ( P = 0.01) and harvest frequency ( P = 0.02). Tiller appearance rate was greater at 4than 8-cm (6.8 and 3.3 tillers m-2 d-1, respectively) stubble height, while TAR was greater when harv ested every 7 than 21 d (6.6 and 3.5 tillers m-2 d-1, respectively). Tiller appear ance rate was affected by stubble height on 25 July ( P = 0.002). As in June, TAR was greater at 4than 8-cm (6.4 and 3.2 tillers m-2 d-1, respectively) stubble height. Frequent and intense clipping resulted in greater TAR than the more lax treatments. Tiller appearance rate on 25 July was also affected by genotype X harvest frequency interaction (Table 6-6). Ther e were no differences in TAR among genotypes when harvested every 7 d, however, Pensacola TAR was greater than all genotypes when harvested every 21 d. Tiller appearance rate was similar for the other genotypes. With the exception of Pensacola, there was no effect of harvest frequency on TAR. Pe nsacola had greater TAR when harvested every 21 than 7 d.

PAGE 161

161 Tiller appearance rate on 21 August was affected by harvest frequency ( P = 0.01), with greater TAR occurring at 7(7.5) than 21-d (3.7 tillers m-2 d-1) harvest frequency. Tiller appearance rate on 21 August was also affected by genotype X stubble height interaction (Table 6-6). Pensacola had greater TAR than all other ge notypes when harvested to 4-cm stubble. Tiller appearance rate was similar for the other genot ypes. There were no differences in TAR among genotypes when harvested to 8-cm stubble. With the exception of Pensacola, there was no effect of stubble height on TAR. Pensacola had greater TAR when harvested to 4than 8-cm stubble. On 18 September, TAR was affected by genotype ( P = 0.002). Tifton 9 and Pensacola had similar TAR (8.9 and 7.4 tillers m-2 d-1, respectively), which were greater than PCA Cycle 4, Argentine, and Tifton 7 (3.7, 3.5, and 2.4 tillers m-2 d-1, respectively). There were no differences in TAR among PCA Cycle 4, Argentine, and Ti fton 7. The lower late-season TAR for Argentine may explain in part its lesser late-seas on growth than the diploid cultivars. Table 6-6. Bahiagrass tiller appearance rate as affected by genotype X harvest frequency interaction on 25 July and genotype x stubble height interaction on 21 Aug. 2006. Data are means across two stubble heights and three replicates (n = 6) for 25 July and two harvest frequencies and three re plicates (n = 6) for 21 August. Evaluation date 25 July 21 August Harvest frequency (d) Stubble height (cm) Genotype 7 21 4 8 ------------------------------# tillers m-2d-1 -------------------------------Argentine 6.4 a A 4.0 b A 3.4 b A 7.7 a A Tifton 7 2.4 a A 5.4 b A 3.4 b A 3.4 a A Pensacola 5.4 a B 11.8 a A 14.5 a A 3.0 a B Tifton 9 5.0 a A 1.7 b A 8.1 b A 3.4 a A PCA Cycle 4 2.7 a A 3.0 b A 6.1 b A 3.4 a A SE 1.5 2.1 Means within an evaluation date followed by the same letter, lower-case letters within the same column and upper-case letters w ithin the same row, do not differ by the LSMEANS test ( P > 0.05).

PAGE 162

162 Tiller death rate Tiller death rate was qu antified monthly in six months of both 2005 and 2006. This response was affected by evaluation date, year, a nd evaluation date X year interaction (Table D5), so data were analyzed by year. In 2005 and 2006, TDR was affected by evaluation date (Table D-6; Figure 6-2). In 2005, TDR was greatest on 21 August, with a general trend of increasing TDR across the first half of the gr owing season. In 2006, TDR increased from spring through September. Tiller death rates were less in the spring and early su mmer and were greatest from mid-summer through fall, with rates as great as 6.2 tillers m-2 d-1 in September 2006. Tiller death rate tended to be greater in 2005 than in 2006. Hirata and Pakiding ( 2001) reported relative Figure 6-2. Bahiagrass tiller death rate as affected by evaluation date in 2005 and 2006. Means within year followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Data are means across five genotype s, two harvest frequencies, two stubble heights, and three replicates for each year (n = 60). TDR of Pensacola increased exponentially from less than 0.0005 to 0.003 tillers tiller-1 d-1 with increasing daily air temperature from 5 to 33C. Similar trends were reported by Pakiding and Hirata (2003). Hirata and Pa kiding (2001) reported peak TDR in summer with lowest TDR occurring in the autumn (8.3 and 1.9 tillers m-2 d-1, respectively) on ro tationally stocked -1 1 3 5 7 9 11 13 29 May26 June24 July21 August18 September16 October Evaluation dateTiller death rate (# tillers m-2d-1) 2005 2006 c bc c a c b c c b ab a a

PAGE 163

163 Pensacola pastures. In contrast, TDR in the current environment was as great or nearly as great in autumn as in mid-summer. Net tiller appearance rate To m ore clearly define the seasonal pattern of bahiagrass tiller dynamics, net TAR was calculated by subtracting TDR from TAR for each evaluation date and is presented by evaluation date within year. Net TAR (Figure 6-3) was affected by evaluation date in 2005 (P = 0.0001) and2006 (P = 0.001) (Table D-6). In 2005, net TAR was greatest in May, which was the only evaluation date in 2005 to show a net positive TAR. In 2006, net TAR remained positive from May through August, with rates as high as 5.0 tillers m-2 d-1 in May. Net TAR decreased throughout the growing season each year, with yearly rates as lo w as .4 tillers m-2 d-1 in August 2005 and .1 tillers m-2 d-1 in October 2006. Hirata and Pakiding (2001) and Pakiding Figure 6-3. Bahiagrass net tiller ap pearance rate as affected by evaluation date in 2005 and 2006. Means within year followed by the same letter do not differ by the LSMEANS test ( P > 0.05). Data are means across five ge notypes, two harvest frequencies, two stubble heights, and three replicates for each year (n = 60). -9 -7 -5 -3 -1 1 3 5 7 29 May26 June24 July21 August18 September16 October Evaluation dateNet tiller appearance rate (# tillers m2d-1) 2005 2006 a ab c bc cd d a b b d bc cd

PAGE 164

164 and Hirata (2003) reported the balance between TAR and TDR (TAR minus TDR) in Pensacola bahiagrass was generally positive in late spring (May) or early summer (June), and was usually close to zero or negati ve in other seasons. Summary and Conclusions Defoliated grasses exhibit tiller m ass/density compensation in response to defoliation intensity and this self-regulation is a means by which grass swards adapt to changes in defoliation intensity (Matthew et al., 1995) and increase likelihood of sward persistence. There are limits to this compensation, or phenotypic plasticity, associated with particular plant species and environments (Chapman and Lemaire, 1993). These relationships have not been studied for a range of bahiagrass genotypes, a nd the objectives of this study were to determine the effects of harvest frequency and stubble height on tiller number tiller number change, tiller mass, and tiller appearance and death rates of five bahiagrass genotypes repres enting two ploidy levels and a range of growth habits. Patterns emerged among genotypes in tiller number. The tetrap loids (i.e., Argentine and Tifton 7) showed no response of tiller number to defoliation treatment. The upright-growing Tifton 7 had the fewest or similar to the fewe st number of tillers for each defoliation treatment, while more decumbent Argentine had the most or similar to the most tillers for each defoliation treatment. Among the diploids, Pensacola responded much like Argentine, always having the greatest or similar to the greatest number of tillers and not responding to defoliation management. In contrast, tiller number of both Tifton 9 and PCA Cycle 4 varied widely across defoliation treatments (a range of up to 970 and 730 tillers m-2, respectively), and they had the greatest tiller number when defolia ted frequently and closely. Tifton 9, PCA Cycle 4, and Tifton 7 decreased in tiller number across treatments in Year 1, likely indicating less rapid morphol ogical adaptation (i.e., ability to increase ti ller number and

PAGE 165

165 decrease tiller mass under defoliation) to defoliation by these upright-growing types than the more decumbent Argentine and Pensacola. All cultivars showed a positive change in tiller number from the beginning of Year 1 to the en d of Year 2, suggesting that over longer time frames phenotypic plasticity was possible even for the upright-growing types. Tiller mass response can be described in term s of ploidy, with tetraploids tending to or having greater tiller mass than diploids. PCA Cy cle 4 generally had the least or not different from the least tiller mass. Tiller appearance rate varied throughout the ye ar, being greatest in spring and least in autumn. In Year 2, when there was genotype X evaluation date interacti on, the tetraploids had the lowest or similar to the lowest TAR in sp ring and autumn, likely reflecting their slower growth in cooler months or possibly a greater resp onse to shorter days. Tiller death rate was least in spring and greatest in mid-summer through autumn and was not affected by genotype. Net TAR decreased as the growing season progressed, was greatest and positive in the spring in Year 1 and in the spring and early summer in Year 2, and tended to be close to zero or negative throughout the remainder of the growing season. In conclusion, the well-known persistence of Argentine and Pensacola under defoliation was explained in part in this study by their ab ility to sustain a similar high number of tillers across a wide range of defoliation treatments. In addition, Argentine had the greatest or not different from the greatest tiller mass among th e genotypes studied. Tift on 9 and PCA Cycle 4, which were shown in Chapter 5 to be less tolerant of freque nt, close defoliation, did increase tiller number under this type of management, but tiller mass was th e least or not different from the least among the genotypes in this experiment. These very small tillers may be indicative of a weakening stand and genotypes that are not as tolerant of freq uent, close defoliation.

PAGE 166

166 CHAPTER 7 SUMMARY AND CONCLUSIONS Bahiagrass ( Paspalum notatum Flgge) is the prim ary past ure grass for beef cattle ( Bos sp.) and horses ( Equus caballis ) in Florida and is grown on more than one million ha (Chambliss and Adjei, 2006). Commonly used bahiagrass culti vars such as Pensaco la and Tifton 9 are sensitive to short daylengths, wh ich induce a change in growth pa ttern, resulting in a reduction in above-ground growth during the cool season. Limited bahiagrass herbage accumulation during the cool season has occurred even when temper ature, soil moisture, and soil fertility were adequate for substantially greater growth (Sinclair et al., 1997). The significant economic implications of s easonal bahiagrass forage shortfall have stimulated research aimed at increasing producti vity during short-daylength months. Additional cool-season productivity must not come at the expe nse of persistence, however, because the cost of pasture renovation is high and return per unit land area from gr assland-livestock systems are relatively low. Thus, the objectives of this study were i) to determine the effects of extended daylength and level of fertili zation on herbage responses a nd physiological and morphological attributes of photoperiod-sensitiv e and less sensitive bahiagrass genotypes (Chapter 3 and 4); and ii) to determine the effects of defoliation management on herbage yield, nutritive value, persistence, and tiller dynamics of photoperiod-sensitive a nd less sensitive ba hiagrass genotypes (Chapter 5 and 6). In order to accomplish these objectives, experiments were conducted from 2004-06 (Experiment 1) and 2005-07 (Experiment 2). In Experiment 1, two photoperiod-sensitive diploid bahiagrass cultivars (Pensacola and Tifton 9) and a cold-adapted, less photoperiod sensitive genotype (PCA Cycle 4) were evaluated under tw o daylength treatments (ambient and extended to 15 h) and two fertilization levels. In Experiment 2, two photoperiod-sensitive diploid

PAGE 167

167 bahiagrass cultivars (Pensacola and Tifton 9), two photoperiod-sensitive tetraploid bahiagrass genotypes (Argentine and Tifton 7), and a diploid cold-adapted, less photoperiod sensitive genotype (PCA Cycle 4) were evaluated under tw o stubble heights (4 and 8 cm) and two harvest frequencies (7 and 21 d). Extended Daylength and Fertilizatio n Experiment (Chapters 3 and 4) Contrary to what was expected, there were no interactions of genotype with daylength. Thus, under the conditions of this study daylength-sensitive a nd less sensitive types responded sim ilarly to daylength for all responses measured. Genotype Responses In Year 2 of the study when rainfall was m o re nearly normal, Tifton 9 and PCA Cycle 4 generally had similar herbage yield, which was gr eater than Pensacola. Pensacola demonstrated a more prostrate growth habit that resulted in ge nerally lesser height:diameter ratios than the more upright growing Tifton 9 and PCA Cycle 4, which had similar heig ht:diameter ratios and growth habit. PCA Cycle 4 had less root mass than Tift on 9 in 2005-06 and less rhizome mass than Pensacola in 2004-05. PCA Cycle 4 also had a lo wer below-:above-ground ratio than the other genotypes in 2005-06, indicating pr oportionally less DM allocated belowground in PCA Cycle 4 than in Pensacola and Tifton 9. Pensacola and Tifton 9 had 13 to 16% greater rhizome + stem base TNC content than PCA Cycle 4 associated with lesser leaf photosynthetic rate and less dry matter (DM) allocated below ground proportionally for PCA Cycle 4 than Pensacola and Tifton 9. These results suggest that PCA Cycle 4 stored less energy and N for regrowth than Pensacola and Tifton 9.

PAGE 168

168 Daylength Responses Extended daylength resulted in greater bahiagrass herbage yi elds than ambient daylength, regardless of genotype. This response is consistent with reports from other authors (Blount et al., 2001; Sinclair et al., 2001; Sinc lair et al., 2003; Sinclair et al., 2004; Newm an et al., 2007) studying daylength-sensitive types. Associated with greater herbage yields under extended daylength were lesser total-season an d seasonal herbage N concentrations. The responses of storage organ mass and TNC concentration and c ontent to daylength were of particular interest in November. The below-:above-ground ra tio was greater under ambient than extended daylength in November, indicating that altering daylength in autumn causes plants to alter DM allocation. Specificall y, proportionally more DM was allocated to the belowthan above-ground plant component under ambient than extended daylength. The preferential allocation of DM to above-ground plant parts under extended daylength could result in inferior growth rates in spring or reduced persistence. However, there was no evidence that extended daylength negatively affected subsequent cool-season yields. A ssociated with greater allocation of DM to below-ground plant parts un der ambient daylength was greater rhizome + stem base TNC concentration and content under ambient than ex tended daylength in November. These responses suggest greater prioritization of reserve storage for plants growing under ambient than extended daylength. Seasonal Responses Tifton 9 and PCA Cycle 4 had sim ilar seasona l herbage yields, which were 63 and 71% greater, respectively, than Pensacola in Year 2. Th e prostrate growth habit of Pensacola and the upright growth habits of Tifton 9 and PCA Cycl e 4 contributed to the differences in seasonal herbage yield among the genotypes. Pensacola ha d greater total-season and seasonal herbage N concentrations. The generally lesser herbage N concentrations of Tifton 9 and PCA Cycle 4 are

PAGE 169

169 likely explained by a dilution of N associated wi th their greater yield. Leaf photosynthesis was greatest for Pensacola and leas t for PCA Cycle 4, and photosynthe sis generally followed similar trends to herbage N concentration. There we re seasonal patterns in total nonstructural carbohydrate (TNC) concentration in root and rhizomes + stem base components. Root TNC concentration was greater in April than in N ovember and June. Lower TNC concentration of roots in November than April suggests that roots are not the primary reserve storage structure for bahiagrass. Conversely, rhizome + stem base TNC concentration was greater in November than in June and April, suggesti ng their importance as reserve storage organs for bahiagrass. Defoliation Management Responses (Chapters 5 and 6) Herbage Yield and Nutritive Value The tetraploids (Argentine and Tifton 7) generally had greater total-season herbage yields than the diploids, but early in the growing season, the diploids (PCA Cycle 4, Tifton 9, and Pensacola) generally outyielded the tetrap loids. The lesser cool-season yields of the tetraploids are consistent with previous observations of slower growth in the spring (Chambliss and Adjei, 2006). PCA Cycle 4 generally had similar total-s eason and cool-season herbage yields as the other diploids, however when harvested every 7 d PCA Cycle 4 had le ss total-season herbage yield than all genotypes. Seasona l herbage yields of PCA Cycle 4 were generally less than or similar to the other diploids. Herbage yield of all genotypes declined approaching autumn, but the rate of decline was generally less for PCA Cycle 4 than the others. There were few differences in herbage cr ude protein among genotypes. This is in agreement with literature published on nutritive va lue of bahiagrass cultivars (Kalmbacher, 1997; Mislevy and Dunavin, 1993; Cham bliss and Adjei, 2006). PCA Cycle 4 generally had similar seasonal patterns of CP concentrations as th e other diploids, but Argentine tended to have greatest early summer CP concentrations. PCA Cy cle 4 had greater total-season IVDOM than all

PAGE 170

170 genotypes but Tifton 9 in one year and was greater than all genotypes in the second year. Thus nutritive value of PCA Cycle 4 is at least as great, or greater than the existing bahiagrass cultivars. Persistence Responses Percent cover in October of the 3-yr study was lowest or equal to th e lowest for PCA Cycle 4 f or all defoliation treatments and was less than 60% except when defoliated every 21 d to 8 cm (the most lax defoliation treatment). Of genotype s tested, the cover response of PCA Cycle 4 was most like that of Tifton 9. Argentine percent co ver was not affected by defoliation treatment and was greatest or equal to the greatest for each defoliation treatment tested. Pensacola cover remained high except when defoliated every 7 d to a 4-cm stubble. PCA Cycle 4 generally had similar root + rhizome mass as the other diploids at the end of 2 and 3 yr, which was generally less than Argentine (tetraploid) root + rhizome mass. At the end of 2 yr, Argentine and Pensacola had an average of 98% more root + rhizom e TNC than Tifton 7, Tifton 9, and PCA Cycle 4. Argentine, Pensacola, and Tifton 7 had 59% more stem base TNC than PCA Cycle 4, suggesting PCA Cycle 4 had less stored energy for regrowth approachin g the cool season. Tiller Responses Argentine h ad the most or similar to the most tillers for each defoliation treatment. Pensacola responded much like Argentine, always having the greatest or si milar to the greatest number of tillers, and tiller number of Pensacola did not respond to defoliation management. In contrast, tiller number of both Tifton 9 and PC A Cycle 4 varied widely across defoliation treatments (a range of up to 970 and 730 tillers m-2, respectively), and they had the greatest tiller number when defoliated frequently and closely. Tifton 9, PCA Cycle 4, and Tifton 7 decrease d in tiller number acr oss treatments in 2005 (Year 1), likely indicating less ra pid morphological adaptation (i.e ., ability to increase tiller

PAGE 171

171 number and decrease tiller mass under defoliation) to defoliati on by these upright-growing types than the more decumbent Argentine and Pensacola All cultivars showed a positive change in tiller number from the beginning of 2005 to the end of 2006, suggesting that over longer time frames phenotypic plasticity was possible even for the upright-growing types. Tiller mass response can be described in terms of ploidy, w ith tetraploids having greatest or equal to the greatest tiller mass than diploids. PCA Cycle 4 gene rally had the least or not different from the least tiller mass. The tetraploids Argentine and Tifton 7 had th e lowest or similar to the lowest tiller appearance rate (TAR) in spring and autumn, likel y reflecting their slower growth in cooler months or possibly a greater response to shorter days. Across genotypes, tiller death rate was least in spring and greatest in mid-summer through autumn. Net TAR decreased as the growing season progressed, was greatest and positive in the spring of 2005 and in the spring and early summer in 2006, and tended to be close to zero or negative throughout the remainder of the growing season. Implications of the Research Based on th ese data, it can be concluded that management of PCA Cycle 4 will likely be more critical than for Pensacola and Argentin e bahiagrass in Florida. PCA Cycle 4s erect growth habit, apparent preferential allocati on of DM to above-ground plant parts over belowground storage structures, smaller tillers, and less stored N and TNC for regrowth than Pensacola and Argentine imply that greater control of defo liation by clipping, and likely by grazing, will be required to sustain PCA Cycle 4 stands for the long term. Specifically, rotational stocking may be required to ensure adequate regrowth period and greater control of stocking rate may be necessary to insure that stubble is maintained above critical li mits. Based on the limited number of treatments applied in this work, it is proposed that PCA Cycle 4 not be defoliated more

PAGE 172

172 frequently than every 21 d or more intensively th an to an 8-cm stubble. The obvious next step in assessing PCA Cycle 4 is to impose a range of gr azing treatments to asce rtain its tolerance of defoliation by grazing livestock.

PAGE 173

173APPENDIX A SOURCES OF VARIATION: CHAPTER 3 Table A-1. Sources of variation for bahiagrass total-season herbage yield. Source of variation Response variable Genotype Daylength Fertilization Year Genotype X daylength Genotype X fertilization Genotype X year Daylength X fertilization Daylength X year Fertilization X Year Total-season herbage yield P < 0.0001 P = 0.013 P = 0.056 P < 0.0001 P = 0.868 P = 0.343 P = 0.0003 P = 0.547 P = 0.002 P = 0.088 Table A-2. Sources of variation for bahi agrass total-season herbage yield by year Source of variation Response variable Genotype Daylength Fertilization Genotype X daylength Genotype X fertilization Daylength X fertilization Totalseason herbage yield 2004-05 P = 0.18 P = 0.374 P = 0.796 P = 0.9 P = 0.315 P = 0.198 2005-06 P = 0.0001 P = 0.003 P = 0.056 P = 0.78 P = 0.65 P = 0.233 Table A-3. Sources of variation for bahiagrass seasonal herbage yield. Source of variation Response variable Genotype DaylengthFertilization Season Year Genotype X daylength Genotype X fertilization Genotype X season Seasonal herbage yield P < 0.0001 P = 0.106 P = 0.102 P < 0.0001 P < 0.0001 P = 0.64 P = 0.387 P = 0.085

PAGE 174

174Table A-3. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X season Daylength X year Fertilizer X season Fertilizer X year Seasonal herbage yield P < 0.0001 P = 0.39 P = 0.132 P = 0.001 P = 0.169 P = 0.146 Table A-4. Sources of variation for bahi agrass seasonal herbage yield by year. Source of variation Response variable Genotype Daylength Fertilization Season Genotype X daylength Genotype X fertilization Genotype X season Daylength X fertilization Daylength X season Fertilization X season Seasonal herbage yield 2004-05 P = 0.132 P = 0.034 P = 0.813 P < 0.0001 P = 0.957 P = 0.251 P = 0.261 P = 0.131 P = 0.0003 P = 0.719 2005-06 P < 0.0001 P = 0.003 P = 0.062 P < 0.0001 P = 0.437 P = 0.683 P = 0.007 P = 0.078 P = 0.968 P = 0.136 Table A-5. Sources of variation for bahiagrass weighted total-se ason nitrogen (N) concentration. Source of variation Response variable Genotype Daylength Fertilization Year Genotype X daylength Genotype X fertilization Genotype X year Daylength X fertilization Daylength x year Fertilization x Year Weighted total season N concentration P < 0.0001 P < 0.0001 P < 0.0001 P = 0.0004 P = 0.527 P = 0.187 P = 0.082 P = 0.299 P = 0.967 P = 0.078

PAGE 175

175Table A-6. Sources of variation for bahiagrass weighted seasona l nitrogen (N) concentration. Source of variation Response variable Genotype Daylength Fertiliza tion Season Year Genotype X daylength Genotype X fertilization Genotype X season Weighted seasonal N concentration P < 0.0001 P < 0.0001 P < 0.0001 P < 0.0001 P < 0.0001 P = 0.201 P = 0.231 P = 0.452 Table A-6. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X season Daylength X year Fertilizer X season Fertilizer X year Weighted seasonal N concentration P = 0.33 P = 0.546 P < 0.0001 P = 0.072 P = 0.7 P = 0.174 Table A-7. Sources of variation for bahiagrass tiller number. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Tiller number P = 0.428 P = 0.071 P = 0.359 P < 0.0001 P < 0.0001 P = 0.686 P = 0.219 P = 0.52 Table A-7. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Tiller number P = 0.483 P = 0.501 P = 0.148 P = 0.102 P = 0.845 P = 0.765

PAGE 176

176Table A-8. Sources of variation for bahiagrass leaf area. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Leaf area P = 0.611 P = 0.978 P = 0.37 P < 0.0001 P < 0.0001 P = 0.658 P = 0.624 P = 0.305 Table A-8. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Leaf area P = 0.024 P = 0.905 P = 0.0004 P = 0.011 P = 0.752 P = 0.29 Table A-9. Sources of variation for bahiagrass leaf area by year. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Genotype X daylength Genotype X fertilization Genotype X evaluation date Daylength X fertilization Daylength X evaluation date Fertilization X evaluation date Leaf area 2004-05 P = 0.004 P = 0.011 P = 0.872 P < 0.0001 P = 0.2 P = 0.775 P = 0.193 P = 0.496 P = 0.005 P = 0.973 2005-06 P = 0.383 P = 0.132 P = 0.245 P = 0.017 P = 0.351 P = 0.681 P = 0.536 P = 0.581 P = 0.006 P = 0.624 Table A-10. Sources of variation for bahiagrass plant height. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Plant height P < 0.0001 P < 0.0001 P = 0.472 P < 0.0001 P < 0.0001 P = 0.833 P = 0.466 P = 0.078

PAGE 177

177Table A-10. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Plant height P = 0.002 P = 0.185 P < 0.0001 P = 0.004 P = 0.804 P = 0.097 Table A-11. Sources of variation for bahiagrass plant height by year. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Genotype X daylength Genotype X fertilization Genotype X evaluation date Daylength X fertilization Daylength X evaluation date Fertilization X evaluation date Plant height 2004-05 P < 0.0001 P < 0.0001 P = 0.061 P < 0.0001 P = 0.917 P = 0.376 P = 0.112 P = 0.711 P < 0.0001 P = 0.611 2005-06 P < 0.0001 P = 0.169 P = 0.5 P < 0.0001 P = 0.83 P = 0.881 P = 0.292 P = 0.12 P = 0.001 P = 0.999 Table A-12. Sources of variation for bahiagrass plant diameter. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Plant diameter P = 0.0003 P = 0.302 P = 0.453 P < 0.0001 P < 0.0001 P = 0.113 P = 0.409 P = 0.017 Table A-12. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Plant diameter P = 0.147 P = 0.924 P = 0.772 P = 0.863 P = 0.913 P = 0.893

PAGE 178

178Table A-13. Sources of variation for bahiagrass height:diameter ratio. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Height:diameter ratio P < 0.0001 P < 0.0001 P = 0.185 P < 0.0001 P < 0.0001 P = 0.158 P = 0.362 P = 0.035 Table A-13. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Height:diameter ratio P = 0.019 P = 0.138 P < 0.0001 P = 0.007 P = 0.951 P = 0.106 Table A-14. Sources of variation for ba hiagrass height:diameter ratio by year. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Genotype X daylength Genotype X fertilization Genotype X evaluation date Daylength X fertilization Daylength X evaluation date Fertilization X evaluation date Height:diameter ratio 2004-05 P < 0.0001 P < 0.0001 P = 0.075 P < 0.0001 P = 0.154 P = 0.421 P = 0.039 P = 0.447 P = 0.001 P = 0.886 2005-06 P < 0.0001 P = 0.138 P = 0.738 P < 0.0001 P = 0.296 P = 0.718 P = 0.686 P = 0.051 P < 0.0001 P = 0.979

PAGE 179

179Table A-15. Sources of variation for bahiagrass leaf photosynthesis. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Leaf photosynthesis P = 0.04 P = 0.103 P = 0.262 P < 0.0001 P < 0.0001 P = 0.084 P = 0.587 P = 0.682 Table A-15. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Leaf photosynthesis P = 0.665 P = 0.754 P = 0.467 P = 0.168 P = 0.607 P = 0.129

PAGE 180

180APPENDIX B SOURCES OF VARIATION: CHAPTER 4 Table B-1. S ources of variation for bahiagrass plant component mass. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Leaf mass P = 0.328 P = 0.495 P = 0.458 P < 0.0001 P < 0.0001 P = 0.483 P = 0.34 P = 0.86 Stem base mass P < 0.0001 P = 0.017 P = 0.708 P < 0.0001 P < 0.0001 P = 0.603 P = 0.891 P = 0.489 Inflorescence mass P = 0.09 P = 0.005 P = 0.89 P < 0.0001 P < 0.0001 P = 0.355 P = 0.947 P = 0.519 Root mass P = 0.03 P = 0.073 P = 0.991 P < 0.0001 P < 0.0001 P = 0.877 P = 0.679 P = 0.98 Rhizome mass P = 0.449 P = 0.304 P = 0.108 P < 0.0001 P < 0.0001 P = 0.635 P = 0.55 P = 0.889 Table B-1. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Leaf mass P = 0.003 P = 0.612 P = 0.005 P = 0.092 P = 0.871 P = 0.261 Stem base mass P = 0.0003 P = 0.285 P = 0.083 P = 0.48 P = 0.82 P = 0.83 Inflorescence mass P = 0.016 P = 0.409 P = 0.301 P = 0.911 P = 0.69 P = 0.978 Root mass P = 0.214 P = 0.644 P = 0.04 P = 0.757 P = 0.427 P = 0.409 Rhizome mass P = 0.435 P = 0.844 P = 0.399 P = 0.56 P = 0.675 P = 0.071

PAGE 181

181Table B-2. Sources of variation for bahiagrass leaf mass by year. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Genotype X daylength Genotype X fertilization Genotype X evaluation date Daylength X fertilization Daylength X evaluation date Fertilization X evaluation date L e a f m a s s 2004-05 P = 0.001 P = 0.214 P = 0.638 P < 0.0001 P = 0.757 P = 0.991 P = 0.153 P = 0.147 P = 0.01 P = 0.99 2005-06 P = 0.049 P = 0.165 P = 0.274 P < 0.0001 P = 0.212 P = 0.217 P = 0.735 P = 0.199 P = 0.002 P = 0.858 Table B-3. Sources of variation for bahiagrass stem base mass by year. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Genotype X daylength Genotype X fertilization Genotype X evaluation date Daylength X fertilization Daylength X evaluation date Fertilization X evaluation date Stem base mass 2004-05 P = 0.088 P < 0.0001 P = 0.826 P < 0.0001 P = 0.098 P = 0.412 P = 0.12 P = 0.379 P < 0.0001 P = 0.77 2005-06 P < 0.0001 P = 0.309 P = 0.722 P = 0.0002 P = 0.956 P = 0.692 P = 0.607 P = 0.365 P = 0.378 P = 0.858 Table B-4. Sources of variation for bahiagrass inflorescence mass by year. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Genotype X daylength Genotype X fertilization Genotype X evaluation date Daylength X fertilization Daylength X evaluation date Fertilization X evaluation date Inflorescence mass 2004-05 P = 0.201 P = 0.566 P = 0.583 P = 0.982 P = 0.435 P = 0.645 2005-06 P = 0.271 P = 0.129 P = 0.581 P < 0.0001 P = 0.357 P = 0.959 P = 0.445 P = 0.257 P = 0.264 P = 0.705

PAGE 182

182Table B-5. Sources of variation for bahiagrass below-:above-ground ratio. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Below:above-ground ratio P = 0.0004 P = 0.011 P = 0.486 P < 0.0001 P < 0.0001 P = 0.621 P = 0.97 P = 0.625 Table B-5. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Below:above-ground ratio P = 0.0002 P = 0.222 P = 0.013 P = 0.001 P = 0.564 P = 0.418 Table B-6. Sources of variation for bahiagrass below-:above-ground ratio by year. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Genotype X daylength Genotype X fertilization Genotype X evaluation date Daylength X fertilization Daylength X evaluation date Fertilization X evaluation date Below:above-ground ratio 2004-05 P = 0.186 P = 0.483 P = 0.931 P = 0.049 P = 0.667 P = 0.501 P = 0.072 P = 0.464 P = 0.036 P = 0.858 2005-06 P < 0.0001 P < 0.0001 P = 0.277 P < 0.0001 P = 0.086 P = 0.497 P = 0.728 P = 0.283 P = 0.012 P = 0.52

PAGE 183

183Table B-7. Sources of variation for bahiagra ss plant component nitr ogen (N) concentration. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Root N concentration P = 0.521 P = 0.046 P = 0.0002 P < 0.0001 P < 0.0001 P = 0.82 P = 0.648 P = 0.981 Rhizome + stem base N concentration P = 0.286 P < 0.0001 P < 0.0001 P < 0.0001 P < 0.0001 P = 0.937 P = 0.759 P = 0.219 Table B-7. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Root N concentration P = 0.723 P = 0.707 P = 0.743 P = 0.926 P = 0.523 P = 0.419 Rhizome + stem base N concentration P = 0.087 P = 0.338 P = 0.11 P = 0.069 P = 0.03 P = 0.002

PAGE 184

184Table B-8. Sources of variation for bahiagrass plant com ponent total nonstructural ca rbohydrate (TNC) concentration. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Root TNC concentration P = 0.78 P = 0.15 P = 0.263 P < 0.0001 P = 0.138 P = 0.176 P = 0.895 P = 0.945 Rhizome + stem base TNC concentration P = 0.0001 P = 0.281 P = 0.0002 P < 0.0001 P = 0.002 P = 0.134 P = 0.962 P = 0.007 Table B-8. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Root TNC concentration P = 0.26 P = 0.698 P = 0.775 P = 0.153 P = 0.869 P = 0.721 Rhizome + stem base TNC concentration P = 0.032 P = 0.62 P < 0.0001 P = 0.065 P = 0.679 P = 0.848

PAGE 185

185Table B-9. Sources of variation for bahiagrass rhizome + stem base total nonstructu ral carbohydrate concen tration (TNC) by year. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Genotype X daylength Genotype X fertilization Genotype X evaluation date Daylength X fertilization Daylength X evaluation date Fertilization X evaluation date Rhizome + stem base TNC concentration 2004-05 P = 0.0002 P = 0.592 P = 0.007 P < 0.0001 P = 0.005 P = 0.913 P = 0.004 P = 0.351 P = 0.215 P = 0.814 2005-06 P = 0.011 P = 0.017 P = 0.004 P = 0.008 P = 0.303 P = 0.833 P = 0.211 P = 0.775 P < 0.0001 P = 0.758 Table B-10. Sources of variation for bahiag rass plant component ni trogen (N) content. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Root N content P = 0.028 P = 0.073 P = 0.365 P < 0.0001 P < 0.0001 P = 0.859 P = 0.83 P = 0.975 Rhizome + stem base N content P = 0.492 P = 0.081 P = 0.008 P < 0.0001 P < 0.0001 P = 0.681 P = 0.52 P = 0.844 Table B-10. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Root N content P = 0.256 P = 0.407 P = 0.018 P = 0.91 P = 0.71 P = 0.304 Rhizome + stem base N content P = 0.363 P = 0.835 P = 0.593 P = 0.135 P = 0.257 P = 0.043

PAGE 186

186Table B-11. Sources of variation for bahiagrass plant component total nonst ructural carbohydrate (TNC) content. Source of variation Response variable Genotype Daylength Fertilization Evaluation date Year Genotype X daylength Genotype X fertilization Genotype X evaluation date Root TNC content P = 0.039 P = 0.239 P = 0.688 P < 0.0001 P < 0.0001 P = 0.612 P = 0.811 P = 0.851 Rhizome + stem base TNC content P = 0.045 P = 0.67 P = 0.774 P < 0.0001 P < 0.0001 P = 0.911 P = 0.545 P = 0.632 Table B-11. Continued Source of variation continued Response variable Genotype X year Daylength X fertilization Daylength X evaluation date Daylength X year Fertilizer X evaluation date Fertilizer X year Root TNC content P = 0.146 P = 0.86 P = 0.068 P = 0.149 P = 0.452 P = 0.678 Rhizome + stem base TNC content P = 0.131 P = 0.993 P = 0.007 P = 0.423 P = 0.549 P = 0.49

PAGE 187

187APPENDIX C SOURCES OF VARIATION: CHAPTER 5 Table C-1. S ources of variation for bahiagrass total-seas on herbage yield. Source of variation continued Response variable Stubble height X year Harvest frequency X year Genotype X stubble height X harvest frequency Genotype X stubble height X year Genotype X harvest frequency X year Stubble height X harvest frequency X year Genotype X stubble height X harvest frequency X year Total-season herbage yield P = 0.108 P = 0.009 P = 0.284 P = 0.504 P = 0.887 P = 0.485 P = 0.958 Table C-2. Sources of variation for bahi agrass total-season he rbage yield by year. Source of variation Response variable Genotype Stubble height Harvest frequency Genotype X stubble height Genotype X harvest frequency Stubble Height X harvest frequency Genotype X stubble height X harvest frequency Totalseason herbage yield 2005 P < 0.0001 P = 0.392 P < 0.0001 P = 0.428 P = 0.046 P = 0.457 P = 0.792 2006 P = 0.079 P = 0.104 P < 0.0001 P = 0.647 P = 0.212 P = 0.544 P = 0.394 2007 P = 0.036 P = 0.158 P = 0.018 P = 0.094 P = 0.269 P = 0.489 P = 0.422

PAGE 188

188Table C-3. Sources of variation fo r bahiagrass seasonal herbage yield. Source of variation Response variable Genotype Stubble height Harvest frequency Evaluation date Year Genotype X stubble height Genotype X harvest frequency Genotype X evaluation date Seasonal herbage yield P = 0.0001 P = 0.261 P = 0.013 P < 0.0001 P = 0.016 P = 0.064 P = 0.001 P < 0.0001 Table C-3. Continued Source of variation continued Response variable Genotype X year Stubble height X harvest frequency Stubble height X evaluation date Stubble height X year Harvest frequency X evaluation date Harvest frequency X year Evaluation date X year Genotype X stubble height X harvest frequency Seasonal herbage yield P < 0.0001 P = 0.725 P = 0.0002 P = 0.129 P < 0.0001 P = 0.011 P < 0.0001 P = 0.064 Table C-3. Continued Source of variation continued Response variable Genotype X stubble height X evaluation date Genotype X stubble height X year Genotype X harvest frequency X evaluation date Genotype X harvest frequency X year Genotype X evaluation date X year Stubble height X harvest frequency X evaluation date Stubble height X harvest frequency X year Harvest frequency X evaluation date X year Seasonal herbage yield P = 0.76 P = 0.551 P = 0.036 P = 0.915 P < 0.0001 P = 0.584 P = 0.514 P < 0.0001

PAGE 189

189Table C-3. Continued Source of variation continued Response variable Genotype X stubble height X harvest frequency X evaluation date Genotype X stubble height X harvest frequency X year Genotype X harvest frequency X evaluation date X year Stubble height X harvest frequency X evaluation date X year Genotype X stubble height X harvest frequency X evaluation date X year Seasonal herbage yield P = 0.344 P = 0.971 P = 0.386 P = 0.001 P = 0.673 Table C-4. Sources of variation for bahiagrass seasonal herbage yield by year. Source of variation Response variable Genotype Stubble height Harvest frequency Evaluation date Genotype X stubble height Genotype X harvest frequency Stubble height X harvest frequency Genotype X evaluation date Seasonal herbage yield 2005 P < 0.0001 P = 0.433 P = 0.022 P < 0.0001 P = 0.291 P = 0.01 P = 0.398 P < 0.0001 2006 P = 0.15 P = 0.238 P = 0.032 P < 0.0001 P = 0.647 P = 0.195 P = 0.542 P < 0.0001 2007 P = 0.105 P = 0.363 P = 0.137 P < 0.0001 P = 0.086 P = 0.272 P = 0.494 P < 0.0001

PAGE 190

190Table C-4. Continued Source of variation continued Response variable Stubble height X evaluation date Harvest frequency X evaluation date Genotype X stubble height X harvest frequency Genotype X stubble height X evaluation date Genotype X harvest frequency X evaluation date Stubble height X harvest frequency X evaluation date Genotype X stubble height X harvest frequency X evaluation date Seasonal herbage yield 2005 P < 0.0001 P < 0.0001 P = 0.71 P = 0.168 P = 0.264 P = 0.097 P = 0.649 2006 P = 0.344 P < 0.0001 P = 0.383 P = 0.809 P = 0.077 P = 0.042 P = 0.618 2007 P = 0.13 P < 0.0001 P = 0.434 P = 0.544 P = 0.335 P = 0.081 P = 0.295 Table C-5. Sources of variation for bahiagrass cool-seas on herbage yield. Source of variation Response variable Genotype Stubble height Harvest frequency Evaluation date Genotype X stubble height Genotype X harvest frequency Stubble height X harvest frequency Genotype X evaluation date Cool-seasonal herbage yield P = 0.076 P = 0.122 P = 0.025 P = 0.005 P = 0.049 P = 0.043 P = 0.038 P = 0.375 Table C-5. Continued Source of variation continued Response variable Stubble height X evaluation date Harvest frequency X evaluation date Genotype X stubble height X harvest frequency Genotype X stubble height X evaluation date Genotype X harvest frequency X evaluation date Stubble height X harvest frequency X evaluation date Genotype X stubble height X harvest frequency X evaluation date Cool-seasonal herbage yield P = 0.037 P = 0.001 P = 0.823 P = 0.244 P = 0.304 P = 0.409 P = 0.798

PAGE 191

191Table C-6. Sources of variation for bahiagra ss weighted total-season crud e protein (CP) and in vitro digestible organic matter concentration. Source of variation Response variable Genotype Stubble height Harvest frequency Year Genotype X stubble height Genotype X harvest frequency Genotype X year Stubble height X harvest frequency Weighted total-season CP concentration P = 0.039 P = 0.017 P = 0.002 P < 0.0001 P = 0.474 P = 0.005 P = 0.032 P = 0.005 Weighted total-season IVDOM concentration P = 0.035 P = 0.008 P = 0.014 P < 0.0001 P = 0.118 P = 0.009 P = 0.001 P = 0.066 Table C-6. Continued Source of variation continued Response variable Stubble height X year Harvest frequency X year Genotype X stubble height X harvest frequency Genotype X stubble height X year Genotype X harvest frequency X year Stubble height X harvest frequency X year Genotype X stubble height X harvest frequency X year Weighted total-season CP concentration P = 0.0003 P = 0.16 P = 0.223 P = 0.875 P = 0.204 P = 0.137 P = 0.959 Weighted total-season IVDOM concentration P = 0.015 P < 0.0001 P = 0.658 P = 0.811 P = 0.214 P = 0.99 P = 0.717

PAGE 192

192Table C-7. Sources of variation for bahiagra ss weighted total-season crud e protein (CP) and in vitro digestible organic matter concentration by year. Source of variation Response variable Genotype Stubble height Harvest frequency Genotype X stubble height Genotype X harvest frequency Stubble Height X harvest frequency Genotype X stubble height X harvest frequency Weighted totalseason CP concentration 2005 P < 0.0001 P < 0.0001 P < 0.0001 P = 0.388 P = 0.034 P = 0.161 P = 0.345 2006 P = 0.078 P < 0.0001 P < 0.0001 P = 0.893 P = 0.199 P = 0.007 P = 0.616 Weighted totalseason IVDOM concentration 2005 P < 0.0001 P < 0.0001 P < 0.0001 P = 0.802 P = 0.001 P = 0.145 P = 0.63 2006 P = 0.0002 P < 0.0001 P = 0.655 P = 0.22 P = 0.704 P = 0.253 P = 0.765 Table C-8. Sources of variation for bahiagra ss herbage crude protein (CP) concentration. Source of variation Response variable Genotype Stubble height Harvest frequency Evaluation date Year Genotype X stubble height Genotype X harvest frequency Genotype X evaluation date Herbage CP concentration P = 0.02 P = 0.007 P = 0.002 P < 0.0001 P = 0.005 P = 0.098 P = 0.0001 P < 0.0001

PAGE 193

193Table C-8. Continued Source of variation continued Response variable Genotype X year Stubble height X harvest frequency Stubble height X evaluation date Stubble height X year Harvest frequency X evaluation date Harvest frequency X year Evaluation date X year Genotype X stubble height X harvest frequency Herbage CP concentration P = 0.388 P = 0.012 P < 0.0001 P < 0.0001 P < 0.0001 P = 0.013 P < 0.0001 P = 0.038 Table C-8. Continued Source of variation continued Response variable Genotype X stubble height X evaluation date Genotype X stubble height X year Genotype X harvest frequency X evaluation date Genotype X harvest frequency X year Genotype X evaluation date X year Stubble height X harvest frequency X evaluation date Stubble height X harvest frequency X year Harvest frequency X evaluation date X year Herbage CP concentration P = 0.002 P = 0.816 P = 0.013 P = 0.04 P < 0.0001 P = 0.062 P = 0.459 P < 0.0001 Table C-8. Continued Source of variation continued Response variable Genotype X stubble height X harvest frequency X evaluation date Genotype X stubble height X harvest frequency X year Genotype X harvest frequency X evaluation date X year Stubble height X harvest frequency X evaluation date X year Genotype X stubble height X harvest frequency X evaluation date X year Herbage CP concentration P = 0.545 P = 0.893 P = 0.115 P = 0.064 P = 0.952

PAGE 194

194Table C-9. Sources of variation for bahiag rass herbage crude protein (CP) by year. Source of variation Response variable Genotype Stubble height Harvest frequency Evaluation date Genotype X stubble height Genotype X harvest frequency Stubble height X harvest frequency Genotype X evaluation date Herbage CP concentration 2005 P = 0.022 P = 0.051 P = 0.002 P < 0.0001 P = 0.513 P = 0.0001 P = 0.194 P < 0.0001 2006 P = 0.028 P = 0.011 P = 0.003 P < 0.0001 P = 0.203 P = 0.035 P = 0.025 P < 0.0001 Table C-9. Continued Source of variation continued Response variable Stubble height X evaluation date Harvest frequency X evaluation date Genotype X stubble height X harvest frequency Genotype X stubble height X evaluation date Genotype X harvest frequency X evaluation date Stubble height X harvest frequency X evaluation date Genotype X stubble height X harvest frequency X evaluation date Herbage CP concentration 2005 P = 0.149 P = 0.004 P = 0.369 P = 0.058 P = 0.286 P = 0.565 P = 0.735 2006 P < 0.0001 P < 0.0001 P = 0.143 P = 0.192 P < 0.0001 P < 0.0001 P = 0.111 Table C-10. Sources of variation for ba hiagrass herbage in vitro digestible organic matter (IVDOM) concentration. Source of variation Response variable Genotype Stubble height Harvest frequency Evaluation date Year Genotype X stubble height Genotype X harvest frequency Genotype X evaluation date Herbage IVDOM concentration P = 0.092 P = 0.009 P = 0.113 P < 0.0001 P = 0.014 P = 0.076 P = 0.103 P = 0.033

PAGE 195

195Table C-10. Continued Source of variation continued Response variable Genotype X year Stubble height X harvest frequency Stubble height X evaluation date Stubble height X year Harvest frequency X evaluation date Harvest frequency X year Evaluation date X year Genotype X stubble height X harvest frequency Herbage IVDOM concentration P < 0.0001 P = 0.626 P = 0.417 P < 0.0001 P < 0.0001 P < 0.0001 P < 0.0001 P = 0.697 Table C-10. Continued Source of variation continued Response variable Genotype X stubble height X evaluation date Genotype X stubble height X year Genotype X harvest frequency X evaluation date Genotype X harvest frequency X year Genotype X evaluation date X year Stubble height X harvest frequency X evaluation date Stubble height X harvest frequency X year Harvest frequency X evaluation date X year Herbage IVDOM concentration P = 0.324 P = 0.432 P = 0.372 P = 0.336 P = 0.0001 P = 0.035 P = 0.403 P < 0.0001 Table C-10. Continued Source of variation continued Response variable Genotype X stubble height X harvest frequency X evaluation date Genotype X stubble height X harvest frequency X year Genotype X harvest frequency X evaluation date X year Stubble height X harvest frequency X evaluation date X year Genotype X stubble height X harvest frequency X evaluation date X year Herbage IVDOM concentration P = 0.388 P = 0.673 P = 0.457 P = 0.093 P = 0.392

PAGE 196

196Table C-11. Sources of variation for bahiagrass herbage in vitro diges tible organic matter (IVDOM) by year. Source of variation Response variable Genotype Stubble height Harvest frequency Evaluation date Genotype X stubble height Genotype X harvest frequency Stubble height X harvest frequency Genotype X evaluation date Herbage IVDOM concentration 2005 P = 0.069 P = 0.033 P = 0.015 P < 0.0001 P = 0.788 P = 0.006 P = 0.778 P = 0.365 2006 P = 0.004 P = 0.009 P = 0.398 P = 0.0001 P = 0.054 P = 0.631 P = 0.393 P < 0.0001 Table C-11. Continued Source of variation continued Response variable Stubble height X evaluation date Harvest frequency X evaluation date Genotype X stubble height X harvest frequency Genotype X stubble height X evaluation date Genotype X harvest frequency X evaluation date Stubble height X harvest frequency X evaluation date Genotype X stubble height X harvest frequency X evaluation date Herbage IVDOM concentration 2005 P = 0.015 P < 0.0001 P = 0.666 P = 0.285 P = 0.332 P = 0.741 P = 0.052 2006 P = 0.824 P = 0.411 P = 0.636 P = 0.36 P = 0.402 P = 0.008 P = 0.808 Table C-12. Sources of varia tion for bahiagrass cover. Source of variation Response variable Genotype Stubble height Harvest frequency Year Genotype X stubble height Genotype X harvest frequency Genotype X year Stubble height X harvest frequency Cover P = 0.001 P = 0.038 P = 0.027 P = 0.023 P = 0.062 P = 0.025 P = 0.68 P = 0.093

PAGE 197

197Table C-12. Continued Source of variation continued Response variable Stubble height X year Harvest frequency X year Genotype X stubble height X harvest frequency Genotype X stubble height X year Genotype X harvest frequency X year Stubble height X harvest frequency X year Genotype X stubble height X harvest frequency X year Cover P = 0.081 P = 0.229 P = 0.099 P = 0.597 P = 0.998 P = 0.885 P = 0.69 Table C-13. Sources of variation for bahiagrass root + rhizome and stem base mass in October 2007 and 2008. Source of variation Response variable Genotype Stubble height Harvest frequency Genotype X stubble height Genotype X harvest frequency Stubble height X harvest frequency Genotype X stubble height X harvest frequency Root + rhizome mass October 2007 P = 0.02 P = 0.052 P = 0.166 P = 0.021 P = 0.007 P = 0.133 P = 0.284 October 2008 P = 0.014 P = 0.457 P = 0.001 P = 0.692 P = 0.764 P = 0.866 P = 0.798 Stem base mass October 2007 P = 0.014 P = 0.162 P = 0.352 P = 0.053 P = 0.017 P = 0.997 P = 0.01 October 2008 P = 0.002 P = 0.0002 P = 0.442 P = 0.607 P = 0.275 P = 0.751 P = 0.741

PAGE 198

198Table C-14. Sources of variation for bahiagrass root + rhizome and stem base ni trogen (N) concentration and content in October 2007. Source of variation Response variable Genotype Stubble height Harvest frequency Genotype X stubble height Genotype X harvest frequency Stubble height X harvest frequency Genotype X stubble height X harvest frequency Root + rhizome N concentration P = 0.233 P = 0.676 P = 0.017 P = 0.408 P = 0.466 P = 0.817 P = 0.33 Stem base N concentration P = 0.649 P = 0.004 P < 0.0001 P = 0.577 P = 0.373 P = 0.44 P = 0.543 Root + rhizome N content P = 0.01 P = 0.018 P = 0.32 P = 0.011 P = 0.002 P = 0.092 P = 0.174 Stem base N content P = 0.013 P = 0.342 P = 0.949 P = 0.064 P = 0.03 P = 0.871 P = 0.142 Table C-15. Sources of variation for bahiag rass root + rhizome and stem base total nonstructural carbohydrat e (TNC) concentrati on and content in October 2007. Source of variation Response variable Genotype Stubble height Harvest frequency Genotype X stubble height Genotype X harvest frequency Stubble height X harvest frequency Genotype X stubble height X harvest frequency Root + rhizome TNC concentration P = 0.68 P = 0.1 P = 0.011 P = 0.853 P = 0.037 P = 0.488 P = 0.258 Stem base TNC concentration P = 0.052 P = 0.347 P = 0.004 P = 0.587 P = 0.114 P = 0.054 P = 0.37 Root + rhizome TNC content P = 0.115 P = 0.018 P = 0.008 P = 0.156 P = 0.005 P = 0.107 P = 0.907 Stem base TNC content P = 0.004 P = 0.581 P = 0.034 P = 0.112 P = 0.416 P = 0.206 P = 0.48

PAGE 199

199APPENDIX D SOURCES OF VARIATION: CHAPTER 6 Table D-1. Sources of variation for bahiagrass tiller num ber. Source of variation Response variable Genotype Stubble height Harvest frequency Evaluation date Year Genotype X stubble height Genotype X harvest frequency Genotype X evaluation date Tiller number P = 0.004 P = 0.143 P = 0.242 P = 0.01 P = 0.008 P = 0.1 P = 0.417 P = 0.29 Table D-1. Continued Source of variation continued Response variable Genotype X year Stubble height X harvest frequency Stubble height X evaluation date Stubble height X year Harvest frequency X evaluation date Harvest frequency X year Evaluation date X year Genotype X stubble height X harvest frequency Tiller number P = 0.126 P = 0.005 P = 0.372 P = 0.005 P = 0.399 P = 0.004 P < 0.0001 P = 0.022 Table D-1. Continued Source of variation continued Response variable Genotype X stubble height X evaluation date Genotype X stubble height X year Genotype X harvest frequency X evaluation date Genotype X harvest frequency X year Genotype X evaluation date X year Stubble height X harvest frequency X evaluation date Stubble height X harvest frequency X year Harvest frequency X evaluation date X year Tiller number P = 0.375 P = 0.323 P = 0.429 P = 0.586 P = 0.318 P = 0.036 P = 0.123 P = 0.078

PAGE 200

200Table D-1. Continued Source of variation continued Response variable Genotype X stubble height X harvest frequency X evaluation date Genotype X stubble height X harvest frequency X year Genotype X harvest frequency X evaluation date X year Stubble height X harvest frequency X evaluation date X year Genotype X stubble height X harvest frequency X evaluation date X year Tiller number P = 0.549 P = 0.309 P = 0.538 P = 0.095 P = 0.137 Table D-2. Sources of variation for bahiagrass tiller number change. Source of variation Response variable Genotype Stubble height Harvest frequency Genotype X stubble height Genotype X harvest frequency Stubble Height X harvest frequency Genotype X stubble height X harvest frequency Tiller number change Spring 2005 to fall 2005 P = 0.039 P = 0.163 P = 0.498 P = 0.92 P = 0.261 P = 0.059 P = 0.506 Spring 2005 to fall 2006 P = 0.014 P = 0.118 P = 0.051 P = 0.526 P = 0.166 P = 0.029 P = 0.499 Table D-3. Sources of variati on for bahiagrass tiller mass. Source of variation Response variable Genotype Stubble height Harvest frequency Evaluation date Year Genotype X stubble height Genotype X harvest frequency Genotype X evaluation date Tiller mass P < 0.0001 P = 0.009 P = 0.049 P = 0.0003 P = 0.008 P = 0.001 P = 0.599 P = 0.132

PAGE 201

201Table D-3. Continued Source of variation continued Response variable Genotype X year Stubble height X harvest frequency Stubble height X evaluation date Stubble height X year Harvest frequency X evaluation date Harvest frequency X year Evaluation date X year Genotype X stubble height X harvest frequency Tiller mass P = 0.329 P = 0.913 P < 0.0001 P = 0.004 P = 0.008 P = 0.652 P < 0.0001 P = 0.618 Table D-3. Continued Source of variation continued Response variable Genotype X stubble height X evaluation date Genotype X stubble height X year Genotype X harvest frequency X evaluation date Genotype X harvest frequency X year Genotype X evaluation date X year Stubble height X harvest frequency X evaluation date Stubble height X harvest frequency X year Harvest frequency X evaluation date X year Tiller mass P = 0.047 P = 0.998 P = 0.9 P = 0.557 P = 0.663 P = 0.781 P = 0.552 P = 0.238 Table D-3. Continued Source of variation continued Response variable Genotype X stubble height X harvest frequency X evaluation date Genotype X stubble height X harvest frequency X year Genotype X harvest frequency X evaluation date X year Stubble height X harvest frequency X evaluation date X year Genotype X stubble height X harvest frequency X evaluation date X year Tiller mass P = 0.718 P = 0.357 P = 0.165 P = 0.027 P = 0.188

PAGE 202

202Table D-4. Sources of variation for ba hiagrass tiller mass by evaluation date. Source of variation Response variable Genotype Stubble height Harvest frequency Year Genotype X stubble height Genotype X harvest frequency Stubble height X harvest frequency Genotype X year Tiller mass 2 June P < 0.0001 P < 0.0001 P = 0.249 P < 0.0001 P = 0.242 P = 0.76 P = 0.509 P = 0.097 27 July P < 0.0001 P < 0.0001 P < 0.0001 P < 0.0001 P = 0.002 P = 0.901 P = 0.728 P = 0.164 27 September P < 0.0001 P < 0.0001 P = 0.0001 P < 0.0001 P = 0.016 P = 0.654 P = 0.861 P =0.985 Table D-4. Continued Source of variation continued Response variable Stubble height X year Harvest frequency X year Genotype X stubble height X harvest frequency Genotype X stubble height X year Genotype X harvest frequency X year Stubble height X harvest frequency X year Genotype X stubble height X harvest frequency X year Tiller mass 2 June P < 0.0001 P = 0.504 P = 0.777 P = 0.247 P = 0.255 P = 0.385 P = 0.89 27 July P = 0.947 P = 0.751 P = 0.166 P = 0.371 P = 0.571 P = 0.3 P = 0.685 27 September P = 0.477 P = 0.225 P = 0.731 P = 0.718 P = 0.305 P = 0.528 P = 0.126

PAGE 203

203Table D-5. Sources of variation for bahiagrass tiller appearance rate (T AR) and tiller death rate (TDR). Source of variation Response variable Genotype Stubble height Harvest frequency Evaluation date Year Genotype X stubble height Genotype X harvest frequency Genotype X evaluation date TAR P = 0.004 P = 0.069 P = 0.175 P < 0.0001 P = 0.1 P = 0.044 P = 0.412 P = 0.013 TDR P = 0.147 P = 0.12 P = 0.506 P = 0.001 P = 0.029 P = 0.745 P = 0.347 P = 0.204 Table D-5. Continued Source of variation continued Response variable Genotype X year Stubble height X harvest frequency Stubble height X evaluation date Stubble height X year Harvest frequency X evaluation date Harvest frequency X year Evaluation date X year Genotype X stubble height X harvest frequency TAR P = 0.013 P = 0.346 P = 0.309 P = 0.0002 P = 0.051 P = 0.147 P < 0.0001 P = 0.853 TDR P = 0.5 P = 0.059 P = 0.948 P = 0.864 P = 0.676 P = 0.623 P < 0.0001 P = 0.538 Table D-5. Continued Source of variation continued Response variable Genotype X stubble height X evaluation date Genotype X stubble height X year Genotype X harvest frequency X evaluation date Genotype X harvest frequency X year Genotype X evaluation date X year Stubble height X harvest frequency X evaluation date Stubble height X harvest frequency X year Harvest frequency X evaluation date X year TAR P = 0.026 P = 0.164 P = 0.387 P = 0.004 P = 0.0001 P = 0.848 P = 0.284 P = 0.061 TDR P = 0.933 P = 0.703 P = 0.591 P = 0.709 P = 0.307 P = 0.948 P = 0.32 P = 0.681

PAGE 204

204Table D-5. Continued Source of variation continued Response variable Genotype X stubble height X harvest frequency X evaluation date Genotype X stubble height X harvest frequency X year Genotype X harvest frequency X evaluation date X year Stubble height X harvest frequency X evaluation date X year Genotype X stubble height X harvest frequency X evaluation date X year TAR P = 0.022 P = 0.727 P = 0.063 P = 0.386 P = 0.008 TDR P = 0.307 P = 0.814 P = 0.188 P = 0.433 P = 0.1 Table D-6. Sources of variation for bahiag rass tiller appearance rate (TAR), tiller death rate (TDR), and net TAR by year. Source of variation Response variable Genotype Stubble height Harvest frequency Evaluation date Genotype X stubble height Genotype X harvest frequency Stubble height X harvest frequency Genotype X evaluation date TAR 2005 P = 0.016 P = 0.92 P = 0.457 P < 0.0001 P = 0.692 P = 0.055 P = 0.902 P = 0.126 2006 P = 0.013 P = 0.05 P = 0.139 P = 0.069 P = 0.035 P = 0.047 P = 0.236 P < 0.0001 TDR 2005 P = 0.274 P = 0.259 P = 0.496 P = 0.007 P = 0.577 P = 0.786 P = 0.067 P = 0.123 2006 P = 0.361 P = 0.155 P = 0.826 P = 0.0003 P = 0.948 P = 0.223 P = 0.474 P = 0.834 Net TAR 2005 P = 0.093 P = 0.238 P = 0.813 P = 0.0001 P = 0.476 P = 0.662 P = 0.07 P = 0.145 2006 P = 0.218 P = 0.126 P = 0.183 P = 0.001 P = 0.007 P = 0.092 P = 0.027 P = 0.004

PAGE 205

205Table D-6. Continued Source of variation continued Response variable Stubble height X evaluation date Harvest frequency X evaluation date Genotype X stubble height X harvest frequency Genotype X stubble height X evaluation date Genotype X harvest frequency X evaluation date Stubble height X harvest frequency X evaluation date Genotype X stubble height X harvest frequency X evaluation date TAR 2005 P = 0.571 P = 0.957 P = 0.895 P = 0.387 P = 0.138 P = 0.909 P = 0.368 2006 P = 0.223 P = 0.005 P = 0.747 P = 0.002 P = 0.229 P = 0.546 P = 0.003 TDR 2005 P = 0.489 P = 0.695 P = 0.659 P = 0.845 P = 0.266 P = 0.856 P = 0.108 2006 P = 0.67 P = 0.658 P = 0.702 P = 0.25 P = 0.302 P = 0.647 P = 0.063 Net TAR 2005 P = 0.371 P = 0.91 P = 0.572 P = 0.591 P = 0.304 P = 0.911 P = 0.912 2006 P = 0.085 P = 0.349 P = 0.779 P = 0.0003 P = 0.304 P = 0.275 P = 0.043

PAGE 206

206 LIST OF REFERENCES Adesogan, A.T., L.E. Sollenberger, and J.E. Moore. 2006. Forage quality. Florida C ooperative Extension Service, SS-AGR-93. Adjei, M.B., J.H. Frank, and C.S. Gardner. 2003. Survey of pest mole crickets (Orthoptera: Gryllotalpidae) activity on past ure in south-central Florid a. Fla. Entomol. 86:199-205. Adjei, M.B., P. Mislevy, R.S. Kalmbach er, and P. Busey. 1989. Production, quality, and persistence of tropical grasse s as influenced by grazing fr equency. Proc. Soil Crop Sci. Soc. Fla. 48:1-6. Adjei, M.B., G.C. Smart, Jr., J.H. Frank, and N. C. Leppla. 2006. Control of pest mole crickets (Orthoptera: Gryllotalpidae) in bahi agrass pastures with the nematode Steinernema scapterisci (Rhabditida: Steinernematidae ). Fla. Entomol. 89:532-535. Arthington, J. and W. Brown. 2003. Effect of matu rity on measures of quality and dry matter intake of four common Florid a pasture forages. p. 11-12. In Florida Beef Report. Univ. of Florida. Avice, J.C., A. Ourry, G. Lemaire, and J. Boucaud. 1996. Nitrogen and carbon flows estimated by 15N and 13C pulse-chase labeling during regrowth of alfalfa. Plant Physiol. 112:281290. Ball, D., M. Collins, G. Lacefield, N. Martin, D. Mertens, K. Olson, D. Putnam, D. Undersander, and M. Wolf. 2001. Understanding forage qua lity. American Farm Bureau Federation Pub. 1-01, Park Ridge, IL. Ball, D.M., C.S. Hoveland, and G.D. Lacefield. 2002. p. 3, 26, 136, 152. In Southern forages, 3rd ed. Potash and Phosphate Institute and Foundation for Agronomic Research, Georgia. Beaty, E.R., R.H. Brown, and J.B. Morris. 1970. Response of Pensacola bahiagrass to intense clipping. p. 538-542. In M.J.T. Norman (ed.) Proc. Intl. Grassl. Cong., 11th, Surfers Paradise, Queensland. 13-23 Apr. 1970. Univ. Qu eensland Press, St. Lucia, Queensland. Beaty, E.R., J.L. Engel, and J.D. Powell. 1977. Yi eld, leaf growth, and ti llering in bahiagrass by N rate and season. Agron. J. 69:308-311. Beaty, E.R., J.D. Powell, R.H. Brown, and W.J. Ethredge. 1963. Effect of nitrogen rate and clipping frequency on yield of Pens acola bahiagrass. Agron. J. 55:3-4. Beaty, E.R., R.L. Stanley, and J. Powell. 1968. E ffect of height of cut on yield of Pensacola bahiagrass. Agron. J. 60:356-358. Beaty, E.R. and K.H. Tan. 1972. Organic matter, N, and base accumulation under Pensacola bahiagrass. J. Range Manage. 25:38-40.

PAGE 207

207 Beaty, E.R., K.H. Tan, R.A. McCreery, and J. B. Jones. 1975. Root-herbage production and nutrient uptake and retention by bermudagrass and bahiagrass. J. Range Manage. 28:385389. Bhagsari, A.S. and R.H. Brown. 1986. Leaf photos ynthesis and its correla tion with leaf area. Crop Sci. 26:127-132. Bircham, J.S. and J. Hodgson. 1983. The influence of sward condition on rates of herbage growth and senescence in mixed swards under continuous stocking management. Grass and Forage Sci. 38:323-331. Blount, A.R., P. Mislevy, T.R. Sinclair, and K. H. Quesenberry. FL PCA CYCLE 4 bahiagrass, a novel diploid bahiagrass for southern past ures. J. Plant Registr. (in review). Blount, A.R., K.H. Quesenberry, P. Mislevy, R. N. Gates, and T.R. Sinclair. 2001. Bahiagrass and other Paspalum species: An overview of the plant breeding efforts in the Southern Coastal Plain. In Proc. 56th South. Past. Forage Crop Improve. Conf., Springdale, AR. Blount, A., K. Quesenberry, B. Myer, and R. Gates. 2003. The bahiagrass and paspalum breeding program. p. 25-33. In Proc. Sod Based Cropping Syst. Conf., Quincy, FL. Blount, A.R., T.R. Sinclair, R.N. Gates, K.H. Quesenberry, P. Misl evy, and R. Littell. 2001. Photoperiod response in Pen sacola bahiagrass. p. 487-488. In J. A. Gomide et al. (ed.) Proc. Intl. Grassl. Cong., 19th, Piracicaba, Sao Paulo, Brazil: Brazilian Society of Animal Husbandry (FEALQ). Boller, B.C. and J. Nosberger. 1983. Effects of temperature and photoperiod on stolon characteristics, dry matter partitioning, and nonstructural carbohydrat e concentration of two white clover ecotypes. Crop Sci. 23:1057-1062. Bolton, J.K. and R.H. Brown. 1980. Photosynthesis of grass species differing in carbon dioxide fixation pathways. Plant Physiol. 66:97-100. Brown, W.F. and R.S. Kalmbacher. 1998. Nutrit ional value of native range and improved forages: a perspective from centr al and south Florida. p. 79-87. In Proc. 47th Annual Beef Cattle Short Course, Gainesville, FL. Brown, W.F. and P. Mislevy. 1988. Influence of matu rity and season on the yields and quality of tropical grasses. p. 46-54. In Beef Cattle Research Report. Univ. of Florida. Burns, J.C., D.S. Chamblee, and F.G. Giesbrecht. 2002. Defoliation intensity effects on seasonlong dry matter distribution and nutritive va lue of tall fescue. Crop Sci. 42:1274-1284. Burton, G.W. 1989. Registration of Tifton 9 Pensacola bahiagrass. Crop Sci. 29:1326. Burton, G.W., R.N. Gates, and G.J. Gascho. 1997. Re sponse of Pensacola bahiagrass to rates of nitrogen, phosphorus and potassium fertilizers. Proc. Soil Crop Sci. Soc. Fla. 56:31-35.

PAGE 208

208 Burton, G.W., J.E. Jackson, and R.H. Hart. 1963. Effects of cutting frequency and nitrogen on yield, in vitro digestibility, and protein, fiber, and carotene content of Coastal bermudagrass. Agron. J. 55:500-502. Casal, J.J., V.A. Deregibus, and R.A. Sanchez. 1985. Variations in tiller dynamics and morphology in Lolium mutltiflorum Lam. Vegetative and reproduc tive plants as affected by differences in red/far-red irradiance. Casal, J.J., R.A. Sanchez, and V.A. Deregibus. 1986. The effect of plant density on tillering: The involvement of R/FR ratio and the proportion of radiation intercepte d per plant. Environ. Exp. Bot. 26:365-371. Chambliss, C.G. 2003. Tifton-9 Pe nsacola bahiagrass. Florida Cooperative Extension Service, SS-AGR-25. Chambliss, C.G., and M.B. Adjei. 2006. Bahiagra ss. Florida Cooperative Extension Service, SSAGR-36. Chambliss, C.G., F.A. Johnson, and M.B. Adje i. 2006. Bermudagrass production in Florida. Florida Cooperative Extension Service, SS-AGR-60. Chambliss, C.G., and L.E. Sollenberger. 1991. Ba hiagrass: The foundation of cow-calf nutrition in Florida. In Proc. 40th Annual Florida Beef Cattle Short Course. May 13, 1991, Gainesville, FL. Univ. of Flor ida: Animal Science Dept. 196p. Chaparro, C.J., and L.E. Sollenberger. 1997. Nutri tive value of clipped Mott elephantgrass herbage. Agron .J. 89:789-793. Chaparro, C.J., L.E. Sollenberger, and C.S. Jones, Jr. 1995. Defoliation effects on Mott elephantgrass productivity and leaf percentage. Agron. J. 87:981-985. Chaparro, C.J., L.E. Sollenberger, and K.H. Quesenberry. 1996. Light interception, reserve status, and persistence of clipped Mott elephantgrass swards. Crop Sci. 36:649-655. Chapman, D.F. and G. Lemaire. 1993. Morphogenetic and structural determinants of plant regrowth after defoliation. p. 95-104. In Proc. Int. Grassl. Cong., 17th Rockhampton, Australia. 8-12 Feb. 1993. Dunmore Press Ltd., Palmerston North, New Zealand. Christiansen, S., O.C. Ruelke, W.R. Ocumpa ugh, K.H. Quesenberry, and J.E. Moore. 1988. Seasonal yield and quality of Bigalta, Redal ta, and Floralta limpograss. Trop. Agric. 65:49-55. Christiansen, S., and T. Svejcar. 1988. Grazing effects on shoot and root dynamics and aboveand below-ground non-structural carbohydrates in Caucasian bl uestem. Grass and Forage Sci. 43:111-119. Coleman, S.W., J.E. Moore, and J.R. Wilson. 2004. Quality and Utilization. p. 267-308. In Warm-season (C4) grasses, Agronomy Monograph no. 45. Madison, WI.

PAGE 209

209 Cooper, J.P. 1964. Climatic variation in forage gra sses. I. Leaf developmen t in climatic races of Lolium and Dactylis J. Appl. Ecol. 1:45-61. Cuomo, G.J., D.C. Blouin, D.L. Corkern, J.E. McCoy, and R. Walz. 1996. Plant morphology and forage nutritive value of three bahiagrasses as affected by harvest frequency. Agron. J. 88:85-89. Davies, A. 1988. The regrowth of grass swards. p. 85-127. In M.B. Jones and A. Lazenby (ed).The grass crop: the physiological basi s of production. Chapman and Hall, London. Dawson, L.A., S.J. Grayston, and E. Paterson. 2000. Effects of grazing on the roots and rhizosphere of grasses. p. 61-84. In G. Lemaire et al. (ed.). Grassland ecophysiology and grazing ecology. CAB Intern ational, New York. Dhont, C., Y. Castonguay, P. Nadeau, G. Belanger, and F-P. Chalifour. 2002. Alfalfa root carbohydrates and regrowth potential in re sponse to fall harvests. Crop Sci. 754-765. Ezenwa, I.V., R.S. Kalmbacher, J.D. Arth ington, and F.M. Pate. 2006. Creeping signalgrass versus bahiagrass for cow and calf grazing. Agron. J. 98:1582-1588. Ferraro, D.O., and M. Oesterheld. 2002. Effect of defoliation on grass growth. A quantitative review. Oikos 98:125-133. Fulkerson, W.J., and D.J. Donaghy. 2001. Plant-so luble carbohydrate reserves and senescence key criteria for developing an effective gr azing management system for ryegrass-based pastures: a review. Aust. J. Exp. Agric. 41:261-275. Gallaher, R. N., C. O. Weldon and J. G. Futral. 1975. An aluminum block digester for plant and soil analysis. Soil Sci. Soc. Amer. Proc. 39:803-806. Gates, R.N., G.M. Hill, and G.W. Burton. 1999. Re sponse of selected and unselected bahiagrass populations to defoliatio n. Agron. J. 91:787-795. Gates, R.N., P. Mislevy, and F.G. Martin. 2001. Herbage accumulation of three bahiagrass populations during the cool season. Agron. J. 93:112-117. Gates, R.N., C.L. Quarin, and C.G.S. Pedreira. 2004. Bahiagrass. p. 651-680. In Warm-season (C4) grasses, Agronomy M onograph no. 45. Madison, WI. Grant, S.A., G.T. Barthram, L. Torvell, J. King, and H.K. Smith. 1983. Sward management, lamina turnover and tiller populati on density in continuously stocked Lolium perennedominated swards. Grass and Forage Sci. 38:333-344. Greenwood, D.J., G. Lemaire, G. Gosse, P. Cruz, A. Draycott, and J.J. Neeteson. 1990. Decline in percentage N of C3 and C4 crops with increasing plant mass. Ann. Bot. 66:425-436. Hambleton, L. G. 1977. Semiautomated method for simultaneous determination of phosphorus, calcium and crude protein in an imal feeds. J.A.O.A.C. 60:845-852.

PAGE 210

210 Hay, R.K.M. 1990. Tansley review no. 26: the in fluence of photoperiod on the dry-matter production of grasses and cereals. New Phytol. 116:233254. Hay, R.K.M., and O.M. Heide. 1983. Specific photoperiodic stimulation of dry matter production in a high-latitude cultivar of Poa pratensis Physiol. Plant. 57:135-142. Hirata, M. 1993a. Response of bahiagrass ( Paspalum notatum Flgge.) sward to cutting height. 2. Yield and in vitro digestibility of herbage. J. Japan. Grassl. Sci. 39:183-195. Hirata, M. 1993b. Response of bahiagrass (Paspalum notatum Flgge.) sward to cutting height. 3. Density of tillers, stolons and primary roots. J. Japan. Grassl. Sci. 39:196-205. Hirata, M. 1994. Response of bahiagrass (Paspalum notatum Flgge.) sward to nitrogen fertilization rate and cutting in terval. 1. Dry weight of plant and litter. J. Japan. Grassl. Sci. 40:313-324. Hirata, M., K. Eguchi, T. Takayoshi, S. Osafune, K. Okimatsu, and N. Yamamoto. 2002. Response of bahiagrass ( Paspalum notatum Flgge.) sward to cutting height. 4. Canopy structure and light extinc tion. Grassl. Sci. 48:326-331. Hirata, M., and W. Pakiding. 2001. Tiller dynamics in a bahiagrass ( Paspalum notatum ) pasture under cattle grazing. Tro p. Grassl. 35:151-160. Hirata, M., and W. Pakiding. 2003. Responses of bahi agrass to nitrogen and defoliation. J. Range Manage. 56:608-615. Hirata, M., and W. Pakiding. 2004. Spatiotemporal dynamics in herbage mass and tiller density in a bahiagrass (Paspalum notatum Flgge.) pasture under cattle grazing: Results from 4year monitoring in permanent quadrats. Grassl. Sci. 50:201-204. Hodgkinson, K.C. 1974. Influence of partial defo liation on photosynthesi s, photorespiration and transpiration by lucerne leaves of different ages. Aust. J. Plant Phys. 1:561-578. Hyder, D.N. 1973. Defoliation in rela tion to vegetative growth. p. 304-317. In V.B. Youngner and C.M. McKell (ed.) The biol ogy and utilization of grasses. Academic Press, Inc., New York. Inosaka, M., K. Ito, H. Numaguchi, and S. Hirakawa. 1973. Studies on winter survival of tropical grasses. J. Japan. Gr assl. Sci. 19:77-84. Islam, M.A., and M. Hirata. 2005. Leaf appear ance, death and detachment, and tillering in centipedegrass ( Eremochloa ophiuroides (Munro) Hack.) in comparison with bahiagrass ( Paspalum notatum Flgge.): A study at a small s od scale. Grassl. Sci. 51:121-127. Johnson, C.R., B.A. Reiling, P. Mislevy, and M.B. Hall. 2001. Effects of nitrogen fertilization and harvest date on yield, digestibility, fiber, an d protein fractions of tropical grasses. J. Anim. Sci. 79:2439-2448.

PAGE 211

211 Kalmbacher, R. 1997. Basic perennial grasses for south Florida: Bahiagrass and Limpograss. p. 4-8. In Range Cattle REC, Research Re port RC-1997-2, Univ. of Florida. Karl, M.G., and P.S. Doescher. 1991. Monitoring roots of grazed rangeland vegetation with the root periscope/mini-rhizotron tec hnique. J. Range Manage. 44:296-298. Kidder, G., C.G. Chambliss, and R. Mylava rapu. 2002. UF/IFAS standardized fertilization recommendations for agronomic crops. Florida Cooperative Extension Service Fact Sheet SL-129. Korte, C.J., J.G. Spall, and A.C.P. Chu. 1982. Pattern of autumn tilleri ng in a Grasslands 4708 tetraploid hybrid ryegrass sward follow ing two heights of mowing. New Zealand J. Agric. Res. 25:157-161. Lattanzi, F.A., H. Schnyder, and B. Thornton. 2005. The sources of carbon and nitrogen supplying leaf growth. Assessment of the role of stores with compartmental models. Plant Physiol. 137:383-395. Macoon, B., L.E. Sollenberger, and J.E. Moor e. 2002. Defoliation effects on persistence and productivity of four Pennisetum spp. genotypes. Agron J. 94:541-548. Mannetje, L. t, and A.F. Pritchard. 1974. The eff ect of daylength and temperature on introduced legumes and grasses for the tropics and subt ropics of coastal Australia. I. Dry matter production, tillering and leaf area. Aust. J. Exp. Agric. Anim. Husb. 14:173-181. Manske, L.L. 1999. Can native prairie be sustained under livestock grazing? In Z. Abouguendia (ed.), 5th Prairie Conservation and Endange red Species Conference. Saskatoon, Saskatchewan. Marousky, F.J., R.C. Ploetz, D.C. Clayton, and C.G. Chambliss. 1991. Flowering response of Pensacola and Tifton 9 bahiagrasse s grown at different latitudes. Proc. Soil and Crop Sci. Soc. Fla. 50:65-69. Matthew, C., G. Lemaire, N.R. Sackville Hamilton, and A. Hernandez-Garay. 1995. A modified self-thinning equation to desc ribe size/density relationships for defoliated swards. Ann. Bot. 76:579-587. McKenzie, F.R. 1997. Influence of grazing freque ncy and intensity on tiller appearance and death rates of Lolium perenne L. under subtropical conditions. Aust. J. Agric. Res. 48:337-342. Meuriot, F., M.-L. Decau, A. Mo rvan-Bertrand, M.-P. PrudHomme, F. Gastal, J.-C. Simon, J.J. Volenec, and J.-C. Avice. 2005. Contribu tion of initial C and N reserves in Medicago sativa recovering from defoliation: impact of cu tting height and residua l leaf area. Func. Plant Biol. 32:321-334. Mislevy, P. 2006. Stargrass. Florida C ooperative Extension Service, SS-AGR-62.

PAGE 212

212 Mislevy, P., G.W. Burton, A.R. Blount, and F.G. Martin. 2005. Dry biomass yield and nutritive value of bahiagrass cultivars in central Flor ida. Proc. Soil Crop Sci. Soc. Fla. 64:75-79. Mislevy, P., G.W. Burton, and P. Busey. 1991. Ba hiagrass response to grazing frequency. Proc. Soil Crop Sci. Soc. Fla. 50:58-64. Mislevy, P., and L.S. Dunavin. 1993. Manage ment and utilization of bermudagrass and bahiagrass in South Florida. p. 84-95. In Proc. Beef Cattle Short Course, 42, Gainesville, FL. 5-7 May 1993. Inst. of Food and Agric. Sci., Univ. of Florida. Gainesville. Mislevy, P., and P.H. Everett. 1981. Subtropical grass species res ponse to different irrigation and harvest regimes. Agron. J. 73:601-604. Moore, J. E., and G. O. Mott. 1974. Recovery of residual organic matter from in vitro digestion of forages. J. Dairy Sci. 57:1258-1259. Morvan-Bertrand, A., N. Pavis, J. Boucaud, and M.-P. Prudhomme. 1999. Partitioning of reserve and newly assimilated carbon in roots and leaf tissues of Lolium perenne during regrowth after defoliation: assessment by 13C steady-state labe ling and carbohydrate status. Plant Cell Environ. 22:1097-1108. Motazedian, I., And S.H. Sharrow. 1990. Defoliation frequency and intensity effects on pasture forage quality. J. Range Manage. 43:198-201. Mott, J.J., M.M. Ludlow, J.H. Richards, and A. D. Parsons. 1992. Effects of moisture supply in the dry season and subsequent defoliati on on persistence of the Savanna grasses Themeda triandra Heteropogon contortus, and Panicum maximum. Aust. J. Agric. Res. 43:241260. Muchovej, R.M., and J.J. Mullahey. 2000. Yield a nd quality of five bahiagrass cultivars in southwest Florida. Proc. Soil Crop Sci. Soc. Fla. 59:82-84. Mylavarapu, R., D. Wright, G. Kidder, and C.G. Chambliss. 2007. UF/IFAS standardized fertilization recommendations for agronom ic crops. Florida Cooperative Extension Service, SL129. Nada, Y. 1980. Photoperiodic response s in flowering of main tropical pasture grasses. J. Japan. Grassl. Sci. 26:157-164. Nautical Almanac. 2003. The nautical almanac fo r the year 2004. Naval Observatory. U.S. Govt. Print. Office, Wash. D.C. Newman, Y.C., T.R. Sinclair, A.S. Blount, M.L. Lugo, and E. Valencia 2007. Forage production of tropical grasses under extended daylengt h at subtropical and tropical latitudes. Environ. Exp. Bot. 61:18-24.

PAGE 213

213 Newman, Y.C., L.E. Sollenberger, K.J. Boote, H. Allen, Jr., J.M. Thomas, and R.C. Littell. 2006. Nitrogen fertilization affects bahiagrass responses to elevated atmospheric carbon dioxide. Agron. J. 98:382-387. Ortega-S., J.A., L.E. Sollenberger, J.M. Bennett, and J.A. Cornell. 1992. Rhizome characteristics and canopy light interception of grazed rhizoma peanut pastures. Agron. J. 84:804-809. Paige, K.N. 1992. Overcompensation in response to mammalian herbivory: from mutualistic to antagonistic interactions. Ecology 73:2076-2085. Pakiding, W., and M. Hirata. 2002. Tolerance of bahiagrass ( Paspalum notatum Flgge) to defoliation. Grassl. Sci. 48:100-104. Pakiding, W., and M. Hirata. 2003. Effects of nitroge n fertilizer rate and cutting height on tiller and leaf dynamics in bahiagrass ( Paspalum notatum Flgge.) swards: tiller appearance and death. Grassl. Sci. 49:193-202. Pedreira, C.G.S., and R.H. Br own. 1996a. Yield of selected and unselected bahiagrass populations at two cutting heights. Crop Sci. 36:134-137. Pedreira, C.G.S., and R.H. Brown. 1996b. Phys iology, morphology, and growth of individual plants of selected and unselected bahiagrass populations. Crop Sci. 36:138-142. Pedreira, C.G.S., L.E. Sollenberger, and P. Mislevy. 2000. Botanical composition, light interception, and carbohydrate reserve status of grazed Florakirk bermudagrass. Agron. J. 92:194-199. Prates, E.R., H.L. Chapman, Jr., E.M. Hodges, and J.E. Moore. 1974. Animal performance by steers grazing Pensacola bahiagrass pasture in relation to forage production, forage composition, and estimated intake. Proc. So il Crop Sci. Soc. Florida 34:152-156. Prine, G.M., and G.W. Burton. 1956. The effect of nitrogen rate and clipping frequency upon the yield, protein content and certa in morphological characteristics of Coastal bermudagrass ( Cynodon dactylon [L.] Pers.). Agron. J. 48:296-301. Rai, K.N., D.S. Murty, D.J. Andrews, and P.J. Bramel-Cox. 1999. Genetic enhancement of pearl millet and sorghum for the semi-arid tropi cs of Asia and Africa. Genome 42:617-628. Rechcigl, J.E., and A.B. Bottcher. 1995. Fate of phosphorus on bahiagrass ( Paspalum notatum ) pastures. Ecol. Eng. 5:247-259. Reiling, B., C. Johnson, P. Mislevy, and M.B. Hall. 2001. Effect of nitrogen fertilization and date of harvest of yield and forage quality factors of Pensacola bahiagrass, Tifton 85 bermudagrass, and Florona stargrass. In Florida Beef Report. Univ. of Florida. Richards, J.H. 1993. Physiology of plants recovering from defoliation. p. 85-94. In Proc. Int. Grassl. Cong., 17th Rockhampton, Australia. 8-12 Feb. 1993. Dunmore Press Ltd., Palmerston North, New Zealand.

PAGE 214

214 Rooney, W.L., and S. Aydin. 1999. Genetic control of a photoperiod-sensitive response in Sorghum bicolor (L.) Moench. Crop Sci. 39:397-400. Salisbury, F.B., and C.W. Ross. 1992. Plant Physiology, 4th ed. Wadsworth Pub. Co., CA. p. 459, 504. SAS Inst. Inc. 1996. SAS statistics user's guide Release version 6. SAS Inst. Inc., Cary, NC. Sbrissia, A.F., S.C. da Silva, C. Matthew, C.A.B. Carvalho, R.A. Carnevalli, L.F.M. Pinto, J.L. Fagundes, and C.G.S. Pedreira. 2003. Tiller si ze/density compensation in grazed Tifton 85 bermudagrass swards. Pesq. Agropec. Bras., Brasilia. 38:1459-1468. Schenk, J.S., and M.O. Westerhaus. 1991a. P opulation definition, sample selection, and calibration procedures for near infrared reflectance spectrosc opy. Crop Sci. 31:469-474. Schenk, J.S., and M.O. Westerhaus. 1991b. Population structuring of near infrared spectra and modified partial least square s regression. Crop Sci. 31:1548-1555. Sheath, G.W., and J. Hodgson. 1989. Plant-animal factors influencing legume persistence. pp. 361-372. In G.C. Marten et al. (ed.) Persistence of forage legumes. ASA/CSSA/SSSA, Madison, WI. Sinclair, T.R., J.M. Bennett, and J.D. Ray. 1997. Environmental limitatio n to potential forage production during the winter in Florida. Proc. Soil Crop Sci. Soc. Fla. 56:58-63. Sinclair, T.R., P. Mislevy, and J.D. Ray. 2001. S hort photoperiod inhibits winter growth of subtropical grasses. Planta 213:488-491. Sinclair, T.R., J.D. Ray, P. Mislevy, and L. M. Premazzi. 2003. Growth of subtropical forage grasses under extended photoperiod during short-daylength months. Crop Sci. 43:618623. Sinclair, T.R., J.D. Ray, L.M. Premazzi, and P. Mislevy. 2004. Photosynthetic photon flux density influences grass responses to ex tended photoperiod. Environ. Exp. Bot. 51:69-74. Smith, D. 1972. Carbohydrate reserves of grasses. p. 318-333. In The biology and utilization of grasses. V.B. Youngner and C.M. McKe ll (ed.). Academic Press, New York. Smith, H. 2000. Phytochromes and light signal pe rception by plants an emerging synthesis. Nature 407:585-591. Sollenberger, L.E., W.R. Ocumpaugh, V.P.B. Euclid es, J.E. Moore, K.H. Quesenberry, and C.S. Jones, Jr. 1988. Animal performance on continuously stocked Pensacola bahiagrass and Floralta limpograss pastur es. J. Prod. Ag. 1:216-220. Sollenberger, L.E., G.A. Rusland, C.S. Jones, Jr., K.A. Albrecht, and K.L. Gieger. 1989. Animal and forage responses on rotationally grazed Floralta limpograss and Pensacola bahiagrass pastures. Agron. J. 81:760-764.

PAGE 215

215 Stewart, R.L., Jr., L.E. Sollenberger, J.C.B. Dube ux Jr., J.M.B. Vendramini, S.M. Interrante, and Y.C. Newman. 2007. Herbage and animal responses to management intensity of continuously stocked bahiagrass pastures. Agron. J. 99:107-112. Swanton, C.J., J.Z. Huang, A. Shrestha, M. Tollenaar, W. Deen, and H. Rahimian. 2000. Effects of temperature and photoperiod on the phonol ogical development of barnyardgrass. Agron. J. 92:1125-1134. Ta, T.C., F.D.H. MacDowall, and M.A. Faris. 1990. Utilization of carbon and nitrogen reserves of alfalfa roots in supporting N2-fixation and shoot regrowth. Plant Soil 127:231-236. Taub, D.R., and M.T. Lerdau. 2000. Relationship be tween leaf nitrogen and photosynthetic rate for three NAD-ME and three NADP-ME C4 grasses. Am. J. Bot. 87:412-417. Thornton, B., and P. Millard. 1997. Increased defo liation frequency depletes remobilization of nitrogen for leaf growth in grasses. Ann. Bot. 80:89-95. Thornton, B., P. Millard, and U. Bausenwein. 2000. Reserve formation and recycling of carbon and nitrogen during regrowth of defoliated plants. p. 85-99. In G. Lemaire et al. (ed.). Grassland ecophysiology and grazing eco logy. CAB International, New York. Twidwell, E.K., W.D. Pitman, and G.J. Cuom o. 1998. Bahiagrass production and management. Louisiana State Agricultural Center. Van Esbroeck, G.A., M.A. Hussey, and M.A. Sanderson. 2004. Reversal of dormancy in switchgrass with low-ligh t photoperiod extension. Bioresour. Technol. 91:141-144. Werner, B.K., and G.W. Burton. 1991. Recurrent rest ricted phenotypic selec tion for yield alters morphology and yield of Pensacola bahiagrass. Crop Sci. 31:48-50. White, L.M. 1973. Carbohydrate reserves of gras ses: a review. J. Range Manage. 26:13-18. Youngner, V.B. 1972. Physiology of defoliation and regrowth. p. 293-303. In V.B. Youngner and C.M. McKell (ed.) The biol ogy and utilization of grasses. Academic Press, Inc., New York.

PAGE 216

216 BIOGRAPHICAL SKETCH Sindy Marie Interran te was born on 4 December 1970, in Dallas, Texas. She received a B.S. in marine biology (1994) at Texas A&M Univ ersity at Galveston. In 2000, Sindy enrolled in graduate school at Tarleton State University in Stephenville, Texas. She obtained a M.S. in agriculture under the supervision of Dr. Jim Muir at the Texas Ag ricultural Experiment Station in 2002. After completion of this degree, Sindy work ed as an Extension Associate at the Texas Cooperative Extension Service. In 2003, Sindy began a doctorate program in forage agronomy at the University of Florida. During her PhD program, Sindy received the University of Florida Gerald O. Mott Meritorious Graduate Student Award in Crop Science in 2006 and the Paul Robin Harris Award in 2007.