<%BANNER%>

Strip-Planting of Rhizoma Peanut in Bahiagrass Pastures to Increase Production and Sustainability of Low-Input Forage-Li...

Permanent Link: http://ufdc.ufl.edu/UFE0044883/00001

Material Information

Title: Strip-Planting of Rhizoma Peanut in Bahiagrass Pastures to Increase Production and Sustainability of Low-Input Forage-Livestock Systems in Florida
Physical Description: 1 online resource (165 p.)
Language: english
Creator: Castillo Garcia, Miguel Sebastian
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2013

Subjects

Subjects / Keywords: bahiagrass -- forage -- forage-livestock -- forages -- grass -- grass-legume -- legume -- low-input -- pasture -- peanut -- rhizoma -- strip -- strip-planting -- sustainability -- sustainable -- warm-season
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: Rhizoma peanut (Arachis glabrata Benth.; RP)is a tropical legume with potential to be grown in association with grasses like with bahiagrass (Paspalum notatum Flüggé)to increase/maintain productivity and sustainability of extensive low-input forage-livestock systems of the Southeastern USA. In this work, four studies were conducted in Florida to investigate the viability of planting ‘Florigraze’ RP in strips into existing bahiagrass pastures. Using this approach, the specific objectives were to evaluate: defoliation options during the year-of- and year-after establishment, weed management, N fertilizer, and seedbed preparation on RP establishment success. The results indicate that single application of herbicide glyphosate to kill bahiagrass followed by mowing the above-ground biomass to 5-cm stubble height before planting RP provides adequate seedbed for RP establishment and may reduce costs compared to conventional practices that include several passes of heavy equipment to plant RP in a fully prepared seedbed. Seedbed preparation in the strips should be followed by a single application of herbicides imazapic (0.07 kg a.i. ha-1) or imazapic + 2,4-D amine (0.07 and 0.28 kg a.i.ha-1, respectively) when broadleaf and grass weeds reach 5- to 10-cm height to provide extended control of weeds. Light environment at the RP canopy height was = 96% of the incident photosynthetically activeradiation (PAR) when imazapic and imazapic + 2,4-D amine herbicides were used compared to = 82% for the other treatments. The use of imazapic stunts bahiagrass growth temporarily providing time for RP establishment. Thus, by end-season when bahiagrass actively regrows, the strip will result in a bahiagrass-RP mixture. Greatest canopy cover and frequency were ~35 and 80%, respectively. Application of 50 kg N ha-1 following herbicide application increased RP canopy cover (+10 percentage points) and frequency (+15 percentage points). Spread measurements indicated that RP spread at a rate of ~36 cm yr-1 into the adjacent bahiagrass sward. During the year-of establishment pasture-utilization should be shifted from grazing to production of hay to prevent loss of RP plants in the strip due to animal preference. During the year-after establishment, grazing management strategies should be targeted to the strip planted to RP to prevent defoliation of the RP-bahiagrass mixture lower than 15-cm stubble.
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 Miguel Sebastian Castillo Garcia.
Thesis: Thesis (Ph.D.)--University of Florida, 2013.
Local: Adviser: Sollenberger, Lynn E.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2013
System ID: UFE0044883:00001

Permanent Link: http://ufdc.ufl.edu/UFE0044883/00001

Material Information

Title: Strip-Planting of Rhizoma Peanut in Bahiagrass Pastures to Increase Production and Sustainability of Low-Input Forage-Livestock Systems in Florida
Physical Description: 1 online resource (165 p.)
Language: english
Creator: Castillo Garcia, Miguel Sebastian
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2013

Subjects

Subjects / Keywords: bahiagrass -- forage -- forage-livestock -- forages -- grass -- grass-legume -- legume -- low-input -- pasture -- peanut -- rhizoma -- strip -- strip-planting -- sustainability -- sustainable -- warm-season
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: Rhizoma peanut (Arachis glabrata Benth.; RP)is a tropical legume with potential to be grown in association with grasses like with bahiagrass (Paspalum notatum Flüggé)to increase/maintain productivity and sustainability of extensive low-input forage-livestock systems of the Southeastern USA. In this work, four studies were conducted in Florida to investigate the viability of planting ‘Florigraze’ RP in strips into existing bahiagrass pastures. Using this approach, the specific objectives were to evaluate: defoliation options during the year-of- and year-after establishment, weed management, N fertilizer, and seedbed preparation on RP establishment success. The results indicate that single application of herbicide glyphosate to kill bahiagrass followed by mowing the above-ground biomass to 5-cm stubble height before planting RP provides adequate seedbed for RP establishment and may reduce costs compared to conventional practices that include several passes of heavy equipment to plant RP in a fully prepared seedbed. Seedbed preparation in the strips should be followed by a single application of herbicides imazapic (0.07 kg a.i. ha-1) or imazapic + 2,4-D amine (0.07 and 0.28 kg a.i.ha-1, respectively) when broadleaf and grass weeds reach 5- to 10-cm height to provide extended control of weeds. Light environment at the RP canopy height was = 96% of the incident photosynthetically activeradiation (PAR) when imazapic and imazapic + 2,4-D amine herbicides were used compared to = 82% for the other treatments. The use of imazapic stunts bahiagrass growth temporarily providing time for RP establishment. Thus, by end-season when bahiagrass actively regrows, the strip will result in a bahiagrass-RP mixture. Greatest canopy cover and frequency were ~35 and 80%, respectively. Application of 50 kg N ha-1 following herbicide application increased RP canopy cover (+10 percentage points) and frequency (+15 percentage points). Spread measurements indicated that RP spread at a rate of ~36 cm yr-1 into the adjacent bahiagrass sward. During the year-of establishment pasture-utilization should be shifted from grazing to production of hay to prevent loss of RP plants in the strip due to animal preference. During the year-after establishment, grazing management strategies should be targeted to the strip planted to RP to prevent defoliation of the RP-bahiagrass mixture lower than 15-cm stubble.
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 Miguel Sebastian Castillo Garcia.
Thesis: Thesis (Ph.D.)--University of Florida, 2013.
Local: Adviser: Sollenberger, Lynn E.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2013
System ID: UFE0044883:00001


This item has the following downloads:


Full Text

PAGE 1

1 STRIP PLANTING OF RHIZOMA PEANUT IN BAHIAGRASS PASTURES TO INCREASE PRODUCTION AND SUSTAINABILITY OF LOW INPUT FORAGE LIVESTOCK SYSTEMS IN FLORIDA By MIGUEL SEBASTI N CASTILLO GARCA A DISSERTATION PRESENTED TO THE GRADUAT E SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 201 3

PAGE 2

2 2013 Miguel Sebastin Castillo Garca

PAGE 3

3 To Jos Bol var and Tania Your love is infinite

PAGE 4

4 ACK NOWLEDGMENTS I would like t o start by expressing my sincere gratitude to my adviser Dr. Lynn E. Sollenberger, for his guidance, continued support and the opportunity to interact with his research group Tempo pace made us partners Secon dly, I am grateful to Dr. A .R Blount for her contributions to the research project (sometimes from her personal resources) as well as for her hospitality during visit s to the research station in Marianna I t hank Dr. J.A. Ferrell, Dr. C.L. Mackowiak, and Dr. M.J. Williams for their constructive criticisms, willingness to review the dissertation, and all the support provided for the successful completion of the research experiments. I am grateful to those that helped with field work. Mr. Richard Cone from Cone Family Farms, LLC, Greenville, FL, for providing the rhizoma peanut (RP) planting material; his generosity is appreciated. I wish to thank Dwight Thomas, at the Beef Research Unit of the University of Florida (UF) Gainesville, FL for his assistance in taking m easureme nts maintaining equipment, handling livestock and his constant sense of humor which made long working days easier. I express my appreciation to Sergio Morichetti and Michael Durham former and current members of the Ferrell research group, respectively, f or the technical assistance provided with spraying equipment. I would also like to thank all the past and present members of the Sollenberger forage research group: Kesi Liu, Daniel R. Pereira, Kenneth R. Woodard, Richard Fethiere, Chaein Na, Hermes H. Cue rvo, Nick Krueger, Kim M. Mullenix, and Marcelo O. Wallau I am grateful to you for making my time in G ainesville during the M.S. and Ph.D. programs a much riche r experience than it would have otherwise been. To Andre Aguiar and Eduardo Alava much apprecia tion is due for being friends and fellow forage managers. I wish you the best in your careers.

PAGE 5

5 Workday planning sessions would not have been possible wi thout my running mates Javier L pez and Francisco Loayza. I t hank you for the great camaraderie Thanks are also due to Carlos Ca as, Esteban Rios and Venancio Fernndez f or their friendship and long overnight conversations To my girlfrien d Catalina Aguirre, an avid dreamer and traveler thank you for all the support, patience and love that were always pr esent regardless of distance, and for keep ing us moving. Finally, I would like to thank my parents, Jos Bolvar and Tania, and my siblings, Jos Gabriel and Mar a del Rosario. Your support, concern and encouragement are infinite and powerful.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 10 LIST OF FIGURES ................................ ................................ ................................ ........ 11 LIST OF ABBREVIATIONS ................................ ................................ ........................... 13 ABSTRACT ................................ ................................ ................................ ................... 14 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 16 2 LITERATURE REVIEW ................................ ................................ .......................... 19 Overview of the Research Problem ................................ ................................ ........ 19 Grass Legume Mixtures for Graz ing Systems ................................ ........................ 20 Species Composition ................................ ................................ ........................ 20 Establishment ................................ ................................ ................................ ... 23 Persisten ce ................................ ................................ ................................ ....... 25 Legume Contribution to Forage Livestock Systems ................................ ......... 26 Forage nutritive value ................................ ................................ ................ 26 Animal responses ................................ ................................ ...................... 28 Nutrient cycling ................................ ................................ .......................... 29 The Southeastern USA Experience ................................ ................................ ........ 32 Bahiagrass ................................ ................................ ................................ .............. 36 Rhizoma Peanut ................................ ................................ ................................ ..... 38 Origin, Distribution, and Adaptation ................................ ................................ .. 38 Taxonomy and Morphology ................................ ................................ .............. 39 Breeding and Selection ................................ ................................ .................... 40 Pest and Diseases ................................ ................................ ........................... 41 Establishment ................................ ................................ ................................ ... 42 Responses to Clipping and Grazing ................................ ................................ 45 Above ground biomass ................................ ................................ .............. 46 Below ground biomass ................................ ................................ .............. 49 Weed Control ................................ ................................ ................................ ... 51 Potential for Use in Graze d Pastures in the Gulf Coast USA Region ............... 54 The Strip Planting Approach for Rhizoma Peanut in Existing Bahiagrass Pastures ................................ ................................ ................................ .............. 56 Add itional Benefits Attributed to Spatial Separation of Species .............................. 57 Summary ................................ ................................ ................................ ................ 58

PAGE 7

7 3 STRIP PLANTING A LEGUME INTO WARM SEASON GRASS PAST URE: DEFOLIATION EFFECTS DURING THE YEAR OF ESTABLISHMENT ................ 59 Overview of Research Problem ................................ ................................ .............. 59 Materials and Methods ................................ ................................ ............................ 61 Experimental Site ................................ ................................ ............................. 61 Land Preparation and Planting ................................ ................................ ......... 62 Treatments and Desig n ................................ ................................ .................... 63 Response Variables ................................ ................................ ................................ 65 Canopy Cover ................................ ................................ ................................ .. 65 Frequency ................................ ................................ ................................ ........ 65 Light Environment ................................ ................................ ............................ 66 Spread ................................ ................................ ................................ .............. 66 Bahiagrass Herbage Harvested ................................ ................................ ....... 67 Statistical Analysis ................................ ................................ ................................ .. 68 Results and Discussion ................................ ................................ ........................... 68 Canopy Cover ................................ ................................ ................................ .. 68 Frequency ................................ ................................ ................................ ........ 69 Light Environment ................................ ................................ ............................ 69 Spread ................................ ................................ ................................ .............. 70 Herbage Harvested ................................ ................................ .......................... 71 Implications of the Research ................................ ................................ ................... 72 4 GRAZING MANAGEMENT STRATEGIES AFFECT YEAR AFTER ESTABLISHMENT PERFORMANCE OF A LEGUME STRIP PLANTED INTO WARM SEASON GRASS PASTURE ................................ ................................ ..... 78 Overview of Research Problem ................................ ................................ .............. 78 Materials and Methods ................................ ................................ ............................ 80 Experimental Site ................................ ................................ ............................. 80 Treatments and Experimental Design ................................ .............................. 81 Response Variables ................................ ................................ ................................ 83 Canopy Cover and Frequency ................................ ................................ .......... 83 Spread ................................ ................................ ................................ .............. 84 Botanical Composition ................................ ................................ ...................... 84 Bahiagrass Herbage Harvested ................................ ................................ ....... 85 Statistical Analysis ................................ ................................ ................................ .. 85 Results and Discussion ................................ ................................ ........................... 86 Canopy Cover and Frequency ................................ ................................ .......... 86 Botanical Composition ................................ ................................ ...................... 88 Spread ................................ ................................ ................................ .............. 90 Bahiagrass Herbage Harvested ................................ ................................ ....... 90 Implications of the Research ................................ ................................ ................... 90

PAGE 8

8 5 STRATEGIES TO CONTROL COMPETITION TO STRIP PLANTED LEGUME IN A WARM SEASON GRASS PASTURE ................................ ............................. 95 Overview of Research Problem ................................ ................................ .............. 95 Materials and Methods ................................ ................................ ............................ 96 Experimental Site ................................ ................................ ............................. 96 Land Preparation a nd Planting ................................ ................................ ......... 97 Treatments and Design ................................ ................................ .................... 98 Response Variables ................................ ................................ ................................ 99 Ca nopy Cover ................................ ................................ ................................ .. 99 Frequency ................................ ................................ ................................ ...... 100 Light Environment ................................ ................................ .......................... 100 Canopy Height and Spread ................................ ................................ ............ 101 Year after Establishment Measurements ................................ ....................... 101 Statistical Analysis ................................ ................................ ................................ 102 Results and Discussion ................................ ................................ ......................... 103 Canopy Cover and Frequency ................................ ................................ ........ 103 Light Environment ................................ ................................ .......................... 106 Canopy Height and Spread ................................ ................................ ............ 107 Year after Establishment Measurements ................................ ....................... 107 Canopy cover and freque ncy ................................ ................................ ... 107 Botanical composition ................................ ................................ .............. 108 Spread ................................ ................................ ................................ ..... 109 Implications of the Research ................................ ................................ ................. 109 6 SEEDBED PREPARATION TECHNIQUES AND WEED MANAGEMENT STRATEGIES FOR STRIP PLANTING A LEGUME INTO WARM SEASON GRASS PASTURES ................................ ................................ ............................. 117 Overview of Research Problem ................................ ................................ ............ 117 Materials and Methods ................................ ................................ .......................... 119 Experimental Site ................................ ................................ ........................... 119 Planting Methodology ................................ ................................ ..................... 120 Treatments and Experimental Design ................................ ............................ 121 Response Variables ................................ ................................ .............................. 122 Sprout Emergence ................................ ................................ ......................... 122 Canopy Cover and Frequency ................................ ................................ ........ 123 Light Environment ................................ ................................ .......................... 124 Spread and Canopy Height ................................ ................................ ............ 124 Statistical Analysis ................................ ................................ ................................ 125 Results an d Discussion ................................ ................................ ......................... 125 Sprout Emergence ................................ ................................ ......................... 125 Canopy Cover and Frequency ................................ ................................ ........ 126 Light Environment ................................ ................................ .......................... 128 Spread and Canopy Height ................................ ................................ ............ 129 Implications of the Research ................................ ................................ ................. 130

PAGE 9

9 7 SUMMARY AND CONCLUSIONS ................................ ................................ ........ 139 Effects of Defoliation in the Establishment Year ................................ ................... 141 Grazing Management i n the Year after Establishment ................................ ......... 14 3 Weed Management Strategies in the Year of Establishment ................................ 144 Seedbed Preparation of the Plante d Strip ................................ ............................. 146 Implications of the Research ................................ ................................ ................. 147 Future Research Needs ................................ ................................ ........................ 148 APPENDIX : METHODOLOGY FOR CANOPY COVER AND FREQUENCY MEASUREMENTS ................................ ................................ ............................... 150 LIST OF REFERENCES ................................ ................................ ............................. 152 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 165

PAGE 10

10 LIST OF TABLES Table page 3 1 Rhizoma peanut (RP) percentage canopy cover and frequency of occurrence in August of the establishment year, June of the year after establishment, and spread at the end of the establishment year following planting in strips in bahiagrass pastures and subjected to d ifferent defoliation treatments ............... 73 4 1 Effect of Year 1 (2010) defoliation strategy on Year 2 (2011) canopy cover and frequency of rhizoma peanut planted in 2010 ................................ .............. 92 4 2 Year 1 (2010) defoliation strategy Y ear 2 (2011) grazing management interaction effect on rhizoma peanut (RP) botanical composition in strips planted to RP in existing bahiagrass in 2010 ................................ ...................... 92 4 3 Botanical composition in 2011 stri ps planted to RP i n 2010 into existing bahiagrass ................................ ................................ ................................ .......... 92 5 1 Rhizoma peanut (RP) canopy cover, frequency, and botanical composition in July of the year after establishment for RP strips planted in bahiagrass pastures and subjected to various weed management strategie s with and without N fertilizer ................................ ................................ ............................. 111

PAGE 11

11 LIST OF FIGURES Figure page 3 1 Monthly rainfall at the University of Florida Beef Research Unit, Gainesville, FL for 2010, 2011, and the 30 yr average. ................................ ......................... 74 3 2 Canopy cover of rhizoma peanut planted in strips in existing bahiagrass pastures. Data are means of 2 yr. ................................ ................................ ...... 75 3 3 Frequency of occurrence of rhizoma peanut planted in strips in existing bahiagrass past ures. Data are means of 2 yr ................................ ..................... 76 3 4 Light environment at the top of the rhizoma peanut canopy for strip planted rhizoma peanut in existing bahiagrass past ures. Data are means of 2 yr .......... 77 4 1 Ca nopy cover and frequency of rhizoma peanut in 2011 as affected by 2011 grazing management treatment following planting in strips in bahiagrass past ures in 2010 ................................ ................................ ................................ 93 4 2 Effect of year of establi shment defoliation management on spread measurements taken at the end of the establishment growing season year of 2010 (Y1) and 2011 th e year after establishment (Y2) ................................ ....... 94 5 1 Canopy cover and f requency of occurrence of rhizoma peanut planted in strips in existing bahiagrass pasture s. Data are means across 2 y r. ................ 112 5 2 Canopy cover and frequency of occurrence of rhizoma peanut plant ed in strips in existing bahiagrass pastures Data are means across 2 yr ................. 113 5 3 Incident photosynthetically active radiation (PAR) at rhizoma peanut canopy height Data are menas across 2 yr ................................ ................................ .. 114 5 4 Incident photosynthetically active radiation (PAR) reaching the rhizoma peanut (RP) canopy Data are means across 2 yr ................................ ............ 115 5 5 Rhizoma peanut canopy height measured at the end of the growing season in 2010 and 2011 ................................ ................................ .............................. 116 6 1 Monthly rainfall at the University of Florida Beef Research Unit, Gainesville, FL for 201 1, and the 30 yr average. ................................ ................................ 132 6 2 Sprout emergence of rhizoma peanut planted in strips in existing bahiagrass plots. Data are means of 2 yr (2011 and 2012) ................................ ................ 133 6 3 Canopy cover and frequency of occurrence of rhizoma peanut planted in strips in existi ng bahiagrass pastures in 2011 ................................ ................. 134

PAGE 12

12 6 4 Light environment at the top o f the rhizoma peanut canopy for strip planted rhizoma peanut in existing bahiagrass pastures in 2011. Effects of seedbed preparation (above) and competition control strategy (below) and their interaction with sampling date are shown.. ................................ ....................... 135 6 5 Light environment at the top of the rhizoma peanut canopy for strip planted rhizoma peanut in existing bahiagrass pastures in 2011. Effects of seedbed preparation and competition control strategy interacti on are shown ................. 136 6 6 Spread of rhizoma peanut (RP) into existing bahiagrass sod following planting in 2011. Spread is the distance from the center of the planted RP strip to the farthest point where identifiable RP plant parts (above ground) were found ................................ ................................ ................................ ........ 137 6 7 Rhizoma peanut canopy height measured on ce at the end of the growing season in 2011. Effects of seedbed preparation and com petition control strategy interaction are shown ................................ ................................ .......... 138 A 1 Quadrat to measure rhizoma peanut (RP) canopy cover and frequency. Each square is 10 by 10 cm. Shaded squares correspond to the squares where measurements were made ................................ ................................ .............. 151

PAGE 13

13 LIST OF ABBREVIATION S ADF Acid detergent fiber CP Crude protein DM Dry matter IVOMD In vitro organic matter digestibility NDF Neutral detergent fiber PAR Photosynthetically active radiati on RP Rhizoma peanut ( Arachis glabrata Benth.) TNC Total non structural carbohydrates

PAGE 14

14 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Do ctor of Philosophy STRIP PLANTING OF RHIZOMA PEANUT IN BAHIAGRASS PASTURES TO INCREASE PRODUCTION AND SUSTAINABILITY OF LOW INPUT FORAGE LIVESTOCK SYSTEMS IN FLORIDA By Miguel Sebastin Castillo Garca May 2013 Chair: Lynn E. Sollenberger Major: Agronom y Rhizoma peanut ( Arachis glabrata Benth.; RP) is a sub tropical legume with potential to be grown in association with grasses like bahiagrass ( Paspalum notatum Flgg ) to increase/maintain productivity and sustainability of extensive low input forage live stock systems of the southeastern USA. In this work, four studies were existing bahiagrass pastures. Using this approach, the specific objectives were to evaluate: 1) defoliation options during the year of and year after establishment, 2) weed management, 3) N fertilizer application to establishing RP, and 4) seedbed preparation effects on RP establishment success. The results indicate that a single application of the herbicide glyphosate to kill bahiagrass followed by mowing the above ground biomass to 5 cm stubble height before planting RP provides adequate seedbed for RP establishment and may reduce costs compared to conventional practices that include several pa sses of heavy equipment to plant RP in a fully prepared seedbed. Regardless of method of seedbed preparation in the strip, it should be followed by a single application of herbicides imazapic (0.07 kg a.i. ha 1 ) or imazapic + 2,4 D amine

PAGE 15

15 (0.07 and 0.28 kg a.i. ha 1 respectively) when broadleaf and grass weeds reach 5 to 10 cm height to provide extended control of weeds. Light environment at the RP canopy height was 96% of the incident photosynthetically active radiation (PAR) when imazapic and imazapic + 2,4 D amine herbicides were used compared to 82% for the other treatments. The use of imazapic stunts bahiagrass growth temporarily providing time for RP establi shment. Thus, it is the end of the season until bahiagrass actively regrows resulting in a bahiagrass RP mixture in the strips. Greatest RP canopy cover and frequency during the establishment year were ~35 and 80%, respectively and occurred in plots wher e imazapic and imazapic + 2,4 D amine were applied. Application of 50 kg N ha 1 following herbicide application of imazapic or imazapic + 2,4 D amine increased RP canopy cover (+10 percentage points) and frequency (+15 percentage points) in plots where wee ds had been controlled successfully. Measurements indicated that early in stand life RP spread at a rate of ~36 cm yr 1 into the adjacent bahiagrass sward. During the year of establishment, utilization of the establishing RP should be hay production to pre vent loss of RP plants in the strip under grazing due to animal preference for plants in the strip and resultant overgrazing. During t he year after establishment, grazing management strategies should be targeted to favor the RP in the strip i.e. maintaini ng at least a15 cm stubble in the planted strip regardless of the height of the adjacent bahiagrass. Based on these results it is concluded that strip planting of RP is an option for incorporating the legume into grass based pastures, but critical manageme nt factors for establishment success include avoiding excessive defoliation by grazing during the first 2 yr of stand development and controlling competition to RP in the strip by use of herbicides.

PAGE 16

16 CHAPTER 1 INTRODUCTION The capacity of legume s to fix at mospheric N through a symbiotic association with soil bacteria (Hirsch et al., 200 1), is a sine qua non characteristic that has encouraged the use of legumes as a source of high nutritive value forage for livestock (Muir et al., 2011), and as a sustainable alternative to expensive fertilizer N (Cherr et al., 2006). In addition, benefits of incorporating legume s in a gricultural ecosystems include acceleration of N cycling (Cra ine et al., 2002) and improvement of physical, chemical, and biological soil proper ties (Thomas, 1995). Inclusion of legumes in forage livestock systems has significant potential to contribute to animal production in the tropics (Rusland et al., 1988; Sollenberger et al, 1989; Shelton et al., 2005) by reduc ing cost associated wit h provi ding high quality forage and prevent ing pasture degradation while increasing/maintaining pasture productivity. Tropical regions are dominated by grasses that have the C 4 photosynthetic pathway and lower nutritive value compared to temperate grasses with th e C 3 photosynthetic pathway (Minson et al., 1981). W hile there is greater potential to increase livestock productivity in the tropics, it is precisely in warm climate environments w h ere forage legumes have contributed less to livestock production compared to temperate climate areas In temperate areas it is more common to find perennial grass legume association s in pasture based livestock pr oduction systems. A challenge in warm climates is that C 3 legumes are overwhelmed when competing with vigorous C 4 gras ses (Dunavin, 1992; Sollenberger and Collins, 2003). There are documented m anagement strategies (e.g., physical separation, grazing management, deferred utilization, timing of planting ) that limit the competition by C 3

PAGE 17

17 legumes from vigorous C 4 grasses (Les leighter and Shelton, 1986; Sollenberger et al., 1987; Whitebread et al., 2009), but Sollenberger and Kalmbacher (2005) indicated that in general the difficulties of consistently establishing and maintaining legumes with C 4 grasses have been underestimated They cite the specific example of sustaining warm season annual legumes over time This requires deferment of grazing during late summer to allow seed set and development of a soil seed bank for re establishment year after year; further, this period of g razing deferment coincides with the time whe n legumes can have the greatest potential benefit in terms of animal responses. Thus, annual legumes are particularly challenging to maintain suggesting an advantage for use of perennial legume species. One legu me with demonstrated persistence and productivity under a variety of uses (grazing, haying, cover crop), and possessing the ability to perennate for many years in the USA Gulf Coast is rhizoma peanut (RP; Arachis glabrata Benth.). The versatility of RP mak es it a top candidate for inclusion in grass legume associations in low input forage livestock systems that are typically dominated by monoculture s of C 4 grasses. To this point, however, the use of RP has been limited to production of high quality hay, due to the cost associated with vegetative establishment, management of weeds, and removing land from production for one or several y ears to allow RP establishment. Current ly, most RP grass associations exist because of failure to control grass weeds growing in fields that initially were intended to be pure stands of RP for the production of hay. An alternative approach for achieving mixed pastures is to plant RP in stri ps in existing grass pastures. It may take a period of time for RP to spread from

PAGE 18

18 the plant ed strip to surrounding areas; nevertheless, if this can be achieve d it would provide a relatively low cost option for establishment of mixed grass legume pastures. Thus, the focus of this research is developing technology for strip planting RP in to exist ing pastures. It is hypothesized that planting method, defoliation management and weed control d uring the year of establishment and grazing management in the year after establishment will be critical factors affecting success of the technique. T he genera l objective of th e dissertation research was to develop cost effective management strategies for successful establishment of RP bahiagrass mixtures. The specific objectives were to evaluate the effects on RP establishment success of: 1) defoliation managem ent options during the year of establishment (Chapter 3); 2) grazing management strategies in the year after establishment (Chapter 4); 3) weed management strategies and N fertiliz ation in the establishment year (Chapter 5); and 4) seedbed preparation and post plant competition control strategies (Chapter 6).

PAGE 19

19 CHAPTER 2 LITERATURE REVIEW Overview of the Research Problem Lack of maintenance fertilization and inadequate grazing management are the primary factors resulting in degradation of grasslands in lo w input systems in some warm climate environments (e.g., Brazil; Boddey et al., 2004; Miles et al., 2004). Degraded grasslands have limited potential to serve their primary function as a source of forage for livestock or to provide ecosystem services. Addi tion of N fertilizer, with N generally being the most limiting nutrient in grasslands, can certainly reduce the occurrence of degradation (Vitousek et al., 1997); however, the practicality and economic viability of this practice is questionable for many fo rage livestock systems (Graham and Vance, 2003; Vitousek et al., 2009). Due to their capacity to fix N 2 from the atmosphere and their higher nutritive value than tropical grasses (Muir et al., 2011), legumes are an alternative source of N for grasslands wi th potential to improve the likelihood of long term persistence, prevent pasture degradation, and increase animal production in warm climates (Sollenberger et al., 1989; Shelton et al., 2005). Rhizoma peanut ( Arachis glabrata Benth.; RP) is a perennial leg ume with documented persistence in the USA Gulf Coast. To date, RP has not been widely used in grazed pasture in low input production systems. The dissertation research will explore options for establishment of RP in existing grass pastures as a lower cost alternative to current establishment practices. The objectives of the literature review are to: 1) provide the reader with the framework from which the dissertation research evolved, 2) discuss previous trials to address the issue of sustainability in gra sslands and specifically establishment,

PAGE 20

20 management, and persistence of tropical grass legume pastures, and 3) discuss the potential of RP in the USA Gulf Coast. The review is directly targeted to the species and problems being investigated. It starts with an overview of grass legume mixtures for grazing systems in temperate and tropical environments and then focuses on bahiagrass ( Paspalum notatum Flgg ) and RP research. Grass Legume Mixtures for Grazing Systems Across environments and production systems, several factors can interact and ultimately determine the fate of a grass legume mixture in time. They include: environmental factors such as light, temperature, and rainfall; growth habit of each species in the mixture (a function of occupying different n iches in time and/or space); nodulation ability and capacity for N fixation; edaphic factors (pH, nutrient availability and form), frequency and intensity of defoliation by grazing animals; ability to survive drought periods; seed production capacity; and pest and disease tolerance. This section compares and contrasts temperate and warm season grass legume pastures, with a focus on tropical perennial grasslands under grazing in terms of species composition, establishment, management, and persistence. Specie s Composition In temperate environments, white clover ( Trifolium repens L.), is the main legume found in pastures and meadows due to its well developed stolon mass and prostrate growth habit (Sheath and H ay, 1989; Rochon et al., 2004). Neither red clover ( T pratense L.) nor alfalfa ( Medicago sativa L.) are considered as well adapted as white clover to repeated grazing because of observed reduced persistence and slower recovery following defoliation. Nevertheless, efforts to select traits related to grazin g tolerance in red clover (Hyslop et al., 1999) and alfalfa (Smith et al., 2000) have had

PAGE 21

21 some success. Morphological characteristics of interest for grazing tolerant legumes have included: a deep set, large crown and underground spreading ability of the p lant through rhizomes or horizontal (creeping) roots, like the ones found in the yellow flowered M sativa subsp. falcata (L.) Arcangeli (Piano et al., 1996; Pecetti and Piano, 2005). There are several other temperate legumes with the ability to successful ly perennate in association with grasses in temperate grazing systems, such as: strawberry clover ( T fragiferum L.), Kenya white clover ( T semipilosum Fres.), greater lotus ( Lotus penduculatus Cab.), and birdsfoot trefoil ( L corniculatus L.) (Gramshaw e t al., 1989). Associations occur frequently with perennial ryegrass (Lolium perenne L.), tall fescue ( L. arundinacea Schreb.), kikuyu ( Pennisetum clandestinum Nees ex Steud), orchardgrass ( Dactylis glomerata L.), and Kentucky bluegrass ( Poa pratensis L.) (Gramshaw et al.,1989; Matches, 1989; Sheath and Hay, 1989; Burns and Bagley, 1996; Harris et al., 1998). Perennial grazing systems in temperate regions have been mainly based on herbaceous forages (grass alone or in combination with legumes). In some coun tries of Western Europe and the British Isles, the route chosen to provide quality forage for livestock production was mainly through N fertilization. In contrast are the New Zealand and Australian experiences, which were based to a greater degree on explo iting the potential of legumes to fix atmospheric N (Templeton, 1976; Hodgson et al., 2005) in year long production systems and to match animal requirements with pasture growth (Sears, 1962) under intensive rotational stocking.

PAGE 22

22 Studies on plant population dynamics and effects of defoliation converge on the bases that the successful association of grasses and legumes growing intermingled in temperate regions (i.e., white clover + perennial ryegrass) is founded on the competitive ability of the legume in low N environments (Hill, 1990). The equilibrium of botanical composition in white clover grass system has been described as dynamic (clover is automatically every ~ 4 yr (Parsons et al., 2006). The working hypothesis described by clover is exploited by the grass) which results in a self regulating system. This occurs because as legume contribution increases the amount of N made available to the companion grass increases followed by an increase of grass competition to the legume. As legume contribution is negatively impacted by grass competition the amount of N available to the grass dec reases and is followed by a decrease in grass contribution. There are no similar studies of population dynamics conducted under tropical environments; the first limitation being the management practices required to establish and most importantly to maintai n grass legume associations compared to temperate environments. In tropical environments the occurrence of associations of grasses and legumes in perennial forage based grazing systems is less frequent and more diverse (types of species and establishment a pproaches) compared to temperate regions. Greater adoption has occurred in Asia and Australia than in Africa, USA, or Latin America (with the exception of Brazil; Valentim and Andrade, 2005). Main reasons for limited adoption are: lack of perceived benefit s of legumes (realization of economic benefits), failure of

PAGE 23

23 technology (diseases, persistence), and failure in approach (no extension support) (Shelton et al., 2005; Sollenberger and Kalmbacher, 2005). Neve rtheless, there exist success stories of productio n systems that have not only included herbaceous species but also shrub and tree legumes. Species used in grazing systems include: stylo ( Stylosanthes spp. ), leucaena [ Leucaena leucocephala (Lam.) de Wit], sesbania [ Sesbania sesban (L.) Merr.], clitoria ( C litoria ternatea L.), phasey bean [ Pueraria phaseoloides (Roxb.) Benth.], pintoi peanut ( A pintoi Krapov. & W.C. Greg.), rhizoma peanut ( A glabrata Benth.), aeschynomene ( Aeschynomene americana L.), and carpon desmodium ( Desmodium heterocarpon L). Grasse s growing in association include: bahiagrass ( P notatum Flgg), buffelgrass ( Cenchrus ciliaris L.), brachiarias ( Brachiaria spp. ), gramalote or guineagrass ( Panicum maximum Jacq.), and gambagrass ( Andropogon gayanus Kunth.) (Shelton el al., 2005). Establ ishment Successful establishment of grass legume mixtures has been achieved in both temperate and tropical environments. Establishing a mixture can occ ur by simultaneously planting seed of several species or by planting one species in an existing sward of the companion species (over seeding) (Vengris, 1965; Cook, 1980; Mueller and Chamblee, 1984; Cook et al., 1993; Cuomo et al., 2001; Schlueter and Tracy, 2012). In temperate environments, either method is commonly used. In tropical environments, some addit ional form of management is required to account for the additional plant competition that emerging C3 legumes will encounter from vigorous emerging or existing C4 grasses. Grasses have fibrous root systems that give them an advantage when competing for sha llow moisture and soil nutrients. Only when soil N is low are grasses at a disadvantage relative to legumes (Muir et al., 2011). Cook et al.

PAGE 24

2 4 (1993) have referred to the process of creating favorable conditions for plant Specific approaches reported in the literature to account for additional management required to establish tropical grass legum e mixtures include: physical separation of the grass and legume components, grazing management, deferred utilization, and timing of planting. Whitebread et al. (2009) proposed simultaneously planting mixtures of burgundy bean [ Macroptilium bracteatum (Nees & Mart.) Marechal & Baudet ] lablab bean [ Lablab purpureus ( L. ) Sweet ] and Clitorea ternatea ( L. ) in association with guinea grass ( Panicum maximum Jacq. ) yellow bluestem ( Dichanthium aristatum Poir C.E. Hubb ) and creeping bluegrass ( Bothriochloa inscu lpta Hochst. Ex A. Rich. A. Camus ) Their methodology was based on planting alternating rows of species in strips in a prepared seedbed. Thus, physical separation of the legume and the grass components was the strategy used to manage interspecific competit ion. For sod seeding aeschynomene into existing limpograss pastures [( Hemarthia altissima (Poir.) Stapf & Hubb.], Sollenberger et al. (1987) proposed grazing limpograss to a low stubble height (8 cm) until aeschynomene seedlings emerged and were ~ 5 cm tal l. Cattle were then removed to allow the legume to fully establish, delaying utilization of the mixed pasture until the legume was at least 20 cm tall. Thus, utilization of the pastures was delayed until late July (middle of the growing season). Associati ons of forage tree legumes, such as leucaena and sesbania, with grasses have also been successfully achieved. Due to the very slow growth at seedling stage of

PAGE 25

25 the legume species, management practices are based on providing a competition free environment (a bove and below ground) for growth of the legume until it has reached a certain height followed by planting the grass species. Thus, the tradeoff is minimizing utilization during the year of establishment and allowing for an extended establishment period be fore introducing the companion grass (Lesleighter and Shelton, 1986; Catchpoole and Blair, 1990; Shelton, 1994). Studies on the effect of land preparation techniques (seedbed preparation) and management of plant competition through the use of herbicides o r defoliation have provided critical information leading to the establishment of mixtures and strategies to lower costs. However, the success of these management practices is dependent upon environmental factors ( e.g. rainfall, temperature, soil fertility) In addition, the ultimate fate of these technologies and the degree of their adoption by producers depends in large part on the level and cost of management required to maintain a functional grass legume mixture under grazing in the long term (Sollenberg er and Kalmbacher, 2005). Persistence The capacity to regrow after defoliation and to survive from season to season and year to year are critical determinants of whether a legume will be a successful companion for grass spec ies. For perennial legumes disti nguishable as discrete plants (e.g. alfalfa), poor persistence can be clearly defined as reduction in plant density. Defining persistence is more difficult for non discrete plants that vegetatively propagate by stolons or rhizomes. For these plants, it is not the death of a stolon unit that is viewed as a decrease in persistence, but the relative difference between rate of stolon death and replacement (Sheath and Hay, 1989).

PAGE 26

26 In both temperate and tropical environments, low growing stoloniferous or rhizomat ous forage legumes are generally considered better suited to defoliation by grazing rather than erect g rowing species. Reasons include the growing points are lower in the sward canopy and are less likely to be grazed, in contrast with upright growth habit types that are better adapted to infrequent defoliation (Pitman et al., 1988; Frame, 2005). Legume Contribution to Forage Livestock Systems Greater voluntary intake, nutritive value, and nutrient availability of forage legumes compared to grasses, often l eads to greater animal performance when legumes are present (Dewhurst et al., 2009). Therefore, it is precisely in N restricted tropical grasslands that inclusion of legumes in forage based production systems has greater potential positive impact. Neverthe less, legumes remain an under exploited resource for tropical farming systems (Thomas, 1995; Pengelly et al., 2003; Shelton et al., 2005). The focus of this section of the review is specific examples of how inclusion of legumes in warm climate grasslands c an improve provision of forage for livestock, animal responses, nutrient cycling, and ultimately prevent grassland degradation and advance the goal of sustainable systems. Forage n utritive v alue In general, tropical grasses have lower nutritive value than temperate species for livestock production. Johnson et al. (2001) evaluated the effects of N fertilization (range from 0 to 157 kg N ha 1 cutting 1 ) on yield, digestibility, fiber and protein fractions of Cynodon spp stargrass ( C nlemfuensis quadratic effect of N on digestibility of bermudagrass (ranged from 571 to 599 g kg 1 ),

PAGE 27

27 linear effect for stargrass (ranged from 517 to 576 g kg 1 ), and no effect for bahiagrass (ranged from 521 to 526 g kg 1 ). Neutral detergent fiber (NDF) across N rates ranged from 777 to 757 g kg 1 for bermudagrass, 769 to 720 g kg 1 for stargrass, and 761 to 739 g kg 1 for bahiagrass. Acid detergent fiber (AD F) ranged from 330 to 328 g kg 1 for bermudagrass, 332 to 317 g kg 1 for stargrass, and 362 to 359 g kg 1 for bahiagrass; total N (% of forage DM) increased for all three species. Forage CP concentrations were from 98 to 178 g kg 1 for bermudagrass, 96 to 176 g kg 1 for stargrass, and 90 to 150 g kg 1 for bahiagrass. In contrast, temperate grasses like annual ryegrass fertilized with 280 kg N ha 1 yr 1 clippe d every 30 d had CP, digestible dry matter, and NDF concentration s of 232, 848, and 388 g kg 1 re spectively (Redfearn et al., 2002). Haby and Robinson (1997) reported that ryegrass CP commonly averages 150 to 200 g kg 1 with no N applied and increased to 280 g kg 1 at a N rate of 448 kg ha 1 Minson (1981) presented relative frequency data for crude f iber and CP from six tropical and seven temperate species. The author indicated that 75% of the samples from tropical grasses had between 290 and 370 g kg 1 crude fiber (mode of 350), compared to temperate grasses where 60% of the samples had between 170 a nd 290 g kg 1 crude fiber (mode of 230). In terms of CP, 67% of the samples were between 30 and 120 g kg 1 for tropical grasses (mode of 80), compared with temperate grasses where 68% of the samples were between 60 and 180 g kg 1 CP (mode of 90). Among le gumes (tropical vs. temperate), differences in nutritive value are not as marked as among grasses. The mode of both distributions was 16% for CP with 60% of temperate legumes between 120 and 180 g kg 1 and 60% of tropical legumes within 120 and 210

PAGE 28

28 g kg 1 CP. Tropical legumes, like temperate ones, are higher in protein and lower in NDF (Van Soest, 1982) compared to grasses. Muir et al. (2011) conducted a meta analysis for CP and digestibility of warm season herbaceous legumes and grasses sampled throughou t a complete growing season in the northern hemisphere. Species included that are of particular interest in Florida were: bahiagrass (two cultivars), bermudagrass (three cultivars), elephantgrass ( Pennisetum purpureum Schumach), limpograss, stargrass, aesc hynomene, carpon desmodium and rhizoma peanut (two cultivars). Results indicated that throughout the growing season CP concentration decreased at a similar rate for grasses and legumes; nevertheless, minimum CP concentration was lower for grasses compared to legumes (78 vs. 151 g kg 1 respectively). Digestibility ranged from 493 to 586 g kg 1 for grasses and 624 to 793 g kg 1 for legumes. Digestibility of grasses decreased at a greater rate than legumes throughout the growing season. Intake is expected to be limited when CP concentration is below 70 g kg 1 (Poppi and Mclellan, 1995). Thus, because of increased CP and digestibility, inclusion of legumes in grass dominated grasslands has the potential to provide higher nutritive value forage for livestock and ultimately increase animal responses. Animal r esponses Increased animal production has been reported in the literature for grazing systems using forage legumes growing in association with grasses. Fribourg et al. (1979) evaluated the effect of four levels of N fertilization (0, 112, 224, and 448 kg ha 1 ) bermudagrass fertilized with 112 kg N ha 1 and a mixture of orchardgrass ( Dactylis glomerata L.) and ladino clover ( T repens L.). The authors reported greater total beef

PAGE 29

29 production (kg ha 1 ) for the orchardgrass ladino clover mixture (561) compared with all other treatments except the highest fertilized bemudagrass (605). St r icker et al. (1979) conducted an experiment to determine i f calf production could be economically increased by use of N fertilizer and/or creep feeding of spring born calves grazing tall fescue ladino clover pastures. The authors concluded that when productive legumes can be maintained in fescue sods, the return from increased carrying capacity by applying N fertilizer was not sufficient to offset the additional cost. In experiments conducted in the s outheastern USA, Rusland et al. (1988) reported greater mean gain ha 1 (+113 kg) for a limpograss aeschynomene mix ture compared to N fertilized limpograss (>100 kg N ha 1 yr 1 ). Pitman et al. (1992) evaluated mixtures of bahiagrass aeschynomene and bahiagrass phasey bean vs. fertilized bahiagrass (0, 56, and 224 kg ha 1 yr 1 ). The authors concluded that in 2 yr (out of three), animals grazing bahiagrass aeschynomene had greater average daily gains and greater total gain compared to 0 N bahiagrass pastures. Greater animal responses due to inclusion of legume s is attributed to greater CP concentration, digestibility, a nd mineral composition of livestock diets, resulting in greater forage intake and animal performance (Kretschmer et al., 1973; Wilson and Minson, 1980; Minson, 1981; Marten, 1985). Nutrient c ycling Nutrient inputs and forage utilization are generally grea ter in temperate forage systems. As a consequence, in intensively managed temperate systems, the greatest contribution associated with more efficient recapture of nutrients may be minimizing loss of nutrients to the environment (Dubeux et al., 2007). In co ntrast, in many low input tropical pastures efficient recapture of nutrients is critical for increas ing productivity to meet increasing demand for beef and milk on already cleared land and to limit

PAGE 30

30 expansion of pastoralism into fragile environments such as tropical forests (Thomas, 1992). Nutrients in a grassland ecosystem reside temporarily and cycle among various (Chapin et al., 2002). Nutrient pools in forage systems include: soil organic matter, living plant biomass (above and below ground), plant residues (dead, relatively undecomposed plant tissues), living animal biomass, and soil nutrients. In the case of mineralization (defined as microbially mediated release of NH 4 + and NO 3 from soil organic matter and plant residues), biological N fixation, N returned by grazing animals, and fertilizer or atmospheric N inputs (Wedin and Russelle, 2006). Nitrogen minera lization and immobilization processes occur simultaneously in the soil, with the relative magnitudes determining whether the overall effect is net N mineralization or net N immobilization (Cabrera et al., 2005). Although biological transformations of N in soils are complex, mineralization largely depends on the quantity and quality (composition) of organic matter (OM) and reflects the influence of the environment, principally temperature and moisture, on biological activity (Goncalves and Carlyle, 1994). A commonly used index of substrate quality and mineralization immobilization potential is the C:N ratio. This is based on the premise that for the assimilation of C to occur, N also has to be assimilated in an amount determined by the C:N ratio of the microb ial biomass. If the amount of N present is larger than that required by the microbial biomass, mineralization occurs with the release of inorganic N. If the amount

PAGE 31

31 of N is equal to that required by the microbial biomass, there will be no net N mineralizati on. On the other hand, if the amount of N in the material is lower than that required by the microbial biomass, there will be immobilization (Cabrera et al., 2005). Research suggests that the break even point between net N mineralization and N immobilizati on can be found between C:N ratios of 20 to 40 (Whitmore, 1996). The existence of a range instead of a single value for the break even point has been related to variation in the C:N ratio of the decomposing microbial biomass as well as the existence of org anic components with different susceptibility to decomposition (Cabrera et al., 2005). Due to their greater N concentration, legume plants have potential to shift the balance toward N mineralization and provide greater plant available N compared to grasses (Thomas and Asakawa, 1993). Results reported by Sainju et al. (2006) indicate that long term productivity of RP increased soil C and N pools due its greater N contributions and its lower C:N ratio compared to perennial weeds. If all other environmental fa ctors are accounted for, the first limitation to N cycling in tropical agro ecosystems is the lack of N among pools. Thus legume s, through atmospheric N fixation, have the potential to introduce N to the grassland and to increase soil fertility (Drinkwater et al., 1998). Estimates of N fixed by tropical legumes vary depending on species, growing conditions (alone or in association) and method of estimation. Some estimates of N fixed by tropical legumes are: 370, 157, and 63 kg ha 1 for calopo ( Calopogonium mucunoides Desv. ) vigna ( Vigna sinensis L.), and greengram ( Phaseolus aureus Roxb.) (Agboola and Fayemi, 1972); Cadish et al. (1989) reported ranges from 44 to 132 kg ha 1 of N fixed (above ground) for Centrosema acutifolium Benth., C macrocarpum Benth., Zornia glabra Desv., tropical kudzu

PAGE 32

32 [ Pueraria phaseoloides (Roxb.) Benth. ] desmodium [ Desmodium heterocarpon (L.) DC subsp. ovalifolium (Prain) H. Ohashi ] macrocephala ( Stylosanth e s macrocephala M.B. Ferreira & Sousa Costa ) S guianesis (Aubl.) Sw., an d S capitata Vogel (Heichel and Henjum, 1991). With an estimated yield of ~10 Mg ha 1 yr 1 and N concentration from the above ground tissue of ~30 g N kg 1 dry matter, monocultures of RP grown in Florida potentially fix ~300 kg N ha 1 yr 1 To date, the re are no studies investigating N fixation and transfer of N when RP is grown in association with other species. Fixed legume N is transferred to companion grasses indirectly through legume root decay and legume litter decomposition (mineralization process es ). Defoliation by grazing or cutting accelerates the rate of turnover of root nodules, and grazing results in N from the grazed forage being returned to the pasture in the form of animal excreta. Research conducted by Thomas (1992) indicated that to main tain a sustainable tropical pasture (without causing a drain on soil organic N reserves), with generally low levels of utilization of 10 to 40%, biologically fixed N or plant litter are likely to have the greatest impact on variation in the amount of inter nally cycled N. This is compared with temperate climates where utilization is often much greater (70%), and where varia tion in the recovery of excreta N is likely to have the greatest effect on the requirement for N to balance the cycle. Thomas (1992) sugg ests that legume content of 20 to 45% of the herbage dry matter could provide the N requirements for a productive and sustainable pasture. The Southeastern USA Experience The foundation of grazing systems for the beef cattle industry in the lower latitude states of the southeastern USA (Texas, Louisiana, Mississippi, Alabama, Georgia and Florida) is perennial tropical and sub tropical forage species. Native as well as planted

PAGE 33

33 pastures furnish viable grazing (Chambliss and Lord, 2001). In Florida specifica lly, bahiagrass, limpograss, stargrass, and hybrid bermudagrasses are the most planted grass species. In general, temperature is the most important environmental determinant of seasonal forage limitations (quantity and quality) as well as the degree to whi ch perennial and annual species contribute to the system. In terms of quantity, unless winter annuals are introduced in the production system in much of the Lower South, available forage for grazing is severely limited during periods of cool weather (Novem ber April). In addition there are anticipated periods of shortfall due to spring (May through early or mid June) and fall (October through November) drought (Sollenberger and Chambliss, 1991). Within a growing season, forage quality is highest in spring and decreases as temperatures rise in mid summer through the fall (Wilson and Minson, 1980; Sollenberger and Chambliss, 1991). Energy concentration in bahiagrass forage, for example, decreases substantially as the season progresses, regardless of fertility or defoliation management; thus, it is not considered well suited to meet the nutritional requirements of young, growing animals or lactating dairy cows (Gates et al., 2004). The same pattern of decrease for bahiagrass in vitro organic matter digestibilit y (IVOMD) was demonstrated by Sollenberger et al. (1989). They reported a decrease from ~600 to < 500 g kg 1 during the August through September period for bahiagrass. In contrast, IVOMD of RP remained above 700 g kg 1 throughout the growing season. The au South result primarily from environmental effects on plant characteristics as opposed to a direct effect on the animal (i.e., heat stress). Kretschmer

PAGE 34

34 et al. (1973) evaluated CP of associations of S humilis HBK., hairy indigo ( lndigofera hirsuta L.), siratro ( Phaseolus atropurpureus DC.), greenleaf [ D intortum (Mill.) Urb.], carpon desmodium ( D heterocarpon L. DC.); and Glycine wightii Willd, grown in combination with digit grass ( Digitaria eriantha Steud.), bahiagrass and Setaria anceps Stapf. The authors concluded that inclusion of legumes in what were formerly grass pastures has the potential to improve forage quality even when compared to fertilized tropical grasses (up t o 126 kg N ha 1 yr 1 ). Crude protein yields for a 2 yr period were greatest for the grass greenleaf association (1475 kg ha 1 ) compared to 172 and 423 kg ha 1 for grass with and without fertilization, respectively. Two legumes that received attention in th e past were aeschynomene and carpon desmodium. Aeschynomene was the first palatable, highly nutritious legume determined to be adapted to seasonally wet, relatively infertile soils of the region. Carpon desmodium was the first perennial (in southern Florid a at least) identified, and although it lacks palatability and the nutritional qualities of aeschynomene it was found to be persistent under close grazing. While reviewing adoption/inclusion of legumes in grass swards in the region, Sollenberger and Kalmba cher (2005) explained that although partnership between cattlemen, seedsmen, research, and extension personnel was strong, the difficulty of consistently establishing and maintaining these legumes in pasture had been underestimated. Aeschynomene seedlings emerge in the spring, but late spring drought (April May) can be devastating to initial establishment and re establishment from natural reseeding in many years. Erratic stand establishment and poor seedling vigor under moisture stress have also limited use of carpon desmodium.

PAGE 35

35 Pitman et al. (1998) evaluated 50 tropical legume accessions, representing 33 species and 17 genera, for persistence under two stocking rates (5 and 2 head ha 1 ), as single row entries in bahiagrass pastures. Grazing was deferred for one year to allow for establishment. Grazing stubble height was 15 cm based on the bahiagrass component. Their results indicated that accessions with prostrate growth habit and seed production had the greatest potential for inclusion in grazing systems in Florida. Similar conclusions were reached by Muir and Pitman (1991). Hernndez Garay et al. (2004), likewise, reported similar results when looking at the growth habit of RP Florigraze and Arbrook subjected to grazing. The authors attributed the greater d ecrease in the proportion of RP in the herbage mass ( 23 vs. 3 percent age points for Arbrook vs. Florigraze, respectively) over 3 yr to the more upright growth habit of Arbrook compared to Florigraze, suggesting that Arbrook was less grazing tolerant than Florigraze. The majority of the experiments conducted in this region have utilized deferment of grazing during late summer, as a strategy to allow seed set and development of a soil seed bank of naturally reseeding legumes for re establishment year aft er year. This period coincides with the time where inclusion of legumes can have the greatest impact in animal responses due to decreasing quality of warm season grasses. Thus, use of legume species that perennate by means of regrowth from stolons or rhizo mes, as opposed to seeds, seem s to be better suited for regional grazing systems than annuals which require reseeding each year. This supports investigation of the potential of RP for grass legume mixtures. Regional literature on animal responses while gr a zing grass legume mixtures (growing intermingled) is limited. Most of the research has been done in Florida. While

PAGE 36

36 there is documented methodology to establish grass legume mixtures [i.e., for aeschynomene in limpograss pastures see Sollenberger et al. (1 987); for mixtures of aeschynomene, phasey bean ( Macroptilium lathyroides [ L.] Urb.) and carpon desmodium in bahiagrass pastures see Aiken et al.(1991b)], measurements were limited to portions of the growing season due to the management strategies required to allow establishment and/or to sustain persistence of the legume component of the pasture. For example, Rusland et al. (1988) reported greater animal responses ( i.e. average daily gain and gain ha 1 over a 3 yr period) for a limpograss aeschynomene asso ciation compared with fertilized limpograss. In that study, measurements were limited to mid summer to early fall allowing for establishment of the legume in the mixture in early summer. Additionally, the legume was over seeded each year. In another study, Aiken et al. (1991a) used yearling steers to graze a mixture of carpon desmodium, aeschynomene, and phasey bean growing intermingled with bahiagrass. Desmodium is persistent under grazing but difficult to establish, while aeschynomene and phasey bean are easier to establish but are short lived legumes. They reported that during the first summer aeschynomene and phasey bean contributed the most to the animal diets while during the second summer (year) contribution from carpon desmodium was greater while tha t of aeschynomene and phasey bean was limited. Bahiagrass Bahiagrass is the most widely planted forage species in the state of Florida and represents the foundation of the beef cattle industry. It is not the purpose of this section to provide a thorough r eview of the bahiagrass literature. Such work has been published by Gates et al. (2004). Rather, the objective is to discuss the potential benefits of growing associations of RP and bahiagrass.

PAGE 37

37 Two thirds of improved pastures are planted with bahiagrass in Florida, which accounts for almost 1 million hectares (Newman et al., 2011). Bahiagrass is a perennial with strong, shallow, horizontal rhizomes formed by short, stout internodes usually covered with old, dry leaf sheaths. Leaves are mostly crowded at the bahiagrass belongs to P. notatum var. saurae, and when compared to common bahiagrass, it is taller, spreads faster, has longer and narrower leaves, smaller spikelets, and can have more racemes per inflorescence. Most of the agricultural land area planted to bahiagrass is used for pasture in extensive cow calf production systems (Gates et al., 2004). Bahiagrass is particularly well suited to this use because of its persistence, even under low soil fertility, and tolerance of environmental stresses and severe grazing by livestock. Bahiagrass has good forage quality during spring, but forage quality drops in July and August due to the high temperatures and abundant rainfall. With low levels of ferti 1 yr 1 ) bahiagrass is moderately productive during spring and summer, but it offers very little fall growth. While other improved perennial grasses such as stargrass and bermudagrass are higher in quality and more productive than bah iagrass throughout the year, they also require more management in order to maintain good stands and prevent overgrazing (Chambliss and Loor, 2001; Johnson et al., 2001). Bahiagrass is relatively free of pests with the exception of mole cricket ( Scapteriscu s spp. ). In summary, while morphological and agronomic characteristics have proven bahiagrass to be the best adapted species grown in Florida for extensive production systems, it is the relative low nutritive value of bahiagrass that sets the upper limit to animal responses, especially in low input (N limited) forage livestock systems. This

PAGE 38

38 situation has long been recognized and provides an opportunity to investigate the inclusion of legumes, like RP, in mixtures with bahiagrass as a potential source of N for the pastures, to increase animal production, and ultimately to take a step toward sustainable systems that are not dependant on N fertilization. Rhizoma Peanut Origin, Distribution, and Adaptation The genus Arachis is native to South America. Arachis glabrata is indigenous to Argentina, Paraguay, Uruguay, Bolivia, and Brazil (Bogdan, 1977; Gregory and Gregory, 1979). It is scattered between the Paraguay and Paran River basin north of their junction in Corrientes, Argentina (Gregory et al., 1980). It is found from altitudes of 50 m and higher above sea level and has been reported at latitudes from 8 to 35 south. In the USA, cultivation of RP has been typically limited to the warm, humid climates with well drained soils in the Gulf Coast Region. Nevert heless, efforts exist to select and cultivate cold tolerant RP lines across the Gulf Coast Region and to higher latitudes. The first record of introduction of A. glabrata into the USA was an addition to the USDA National Plant Germplasm System in 1936 when a collection from Matto Grosso, limited due to slow establishment and low productiv ity (Williams et al., 2008). Commercial acceptance of rhizoma peanut did not occur until after the release of the cultivars Florigraze (PI 421707) in 1978 and Arbrook (PI 262817) in 1986 (Prine et al., 1981; 1986a ,b ; Williams et al., 2008) by the Universit y of Florida. In general, cultivars selected for use in Florida are adapted to well drained sandy soils (Prine et al., 1981; Prine et al., 1986 a, b).

PAGE 39

39 Niles et al. (1990) reported that establishment of Florigraze was negatively associated with increasing so il pH (pH range from 5.2 to 7.9). Reed and Ocumpaugh (1991) evaluated Fe deficiency chlorosis using Arbrook, Florigraze, a Florigraze off type, and 69 rhizoma peanut PIs on a Parrita soil (pH 8). Their results indicated that Line 35 and Arbrook were the hi ghest yielding; however, Line 35 was superior to Arbrook in resistance to Fe was originally collected near Trinidad, Paraguay (27 8S 5537W, elevation 50 m) was selected for persistence and ear ly spring production under cold, dry climatic conditions up to 34N latitude. Efforts to identify/select cold tolerant RP are an active area of research (Interrante et al., 2011). Taxonomy and Morphology Arachis glabrata was named rhizoma peanut because o f its rhizomes. It belongs to the section VIII Rhizomatosae Series Rhizomatosae Rhizomes are underground stems with buds in the axils. The rhizome grows year after year and shoots emerge from buds on the rhizome. Arachis glabrata is a self pollinated, dec umbent, long lived warm season perennial. Leaves are tetrafoliolate. Stipules are subulate, villous to glabrous, sometimes with bristles. Leaflets are oblong, elliptical or obovate, with the margin somewhat marked on the underside. The upper leaf surface i s usually glabrous, but younger leaves may exhibit some very short, scattered hairs. The lower leaf surface is described as having adpressed hairs to subglabrous, and frequently with hairs somewhat longer on the midvein. The hypanthium is well developed an d villous. The calyx is villous. Flower color is standard orange and rarely yellow (Krapovickas and Gregory, 2007).

PAGE 40

40 Breeding and Selection All of the cultivars of rhizoma peanut grown in the USA are either plant introductions or naturally occurring seedlin gs that were isolated and increased. Breeding of rhizoma peanut is limited, mainly because there is very little (or none in some of the released cultivars) seed production and propagation of the crop is done with vegetative tissue (rhizomes). Low occurrenc e of cross pollination by insects and parthenogenesis has been reported for RP, as is the case for Florigraze, which is thought to be a chance hybrid that originated from a naturally occurring plant between 1 yr old plots of P1118457 and P1151982 (Prine et al., 1986 a ). Nevertheless, a hybridization process, which could help to make improvements in the population under a more systematic approach, is practically non existent for RP. Four forage type cultivars of RP have been released by the University of Flor ida (Florida Agricultural Experiment Station). Cultivars Florigraze and Arbrook were released in 1978 and 1986, respectively (Prine et al., 1981; Prine et al., 198 6a,b; Prine et al., 1990), and UF Tito (PI 262826) and UF Peace (PI 658214) were released in 2008 (Quesenberry et al., 2010). Arbrook has larger stems and leaflets than Florigraze, is better adapted to excessively drained soils and makes more rapid upright growth, but it spreads laterally slower and is less cold tolerant than Florigraze (Prine et al., 1986b). UF Tito and UF Peace showed improved field tolerance to peanut stunt virus (PSV; member of the cucumo virus) compared to Florigraze. UF Peace had greater lateral spread than the others and UF Tito was reported to be more competitive with commo n bermudagrass (Quesenberry et al., 2010). In addition, two low growing ornamental types of RP were released as germplasm with potential for forage. These were called Ecoturf (PI 658529) and Arblick (PI 658528). Numerous experimental lines and plant

PAGE 41

41 introd uctions have been evaluated in Louisiana (Venuto et al., 1995; Venuto et al., 1997) under the same selection criteria, but none of them out yielded Florigraze. As pointed out by Prine et al. (2010), the selection process so far has mainly focused on yield, and accessions that were relatively low growing with excellent ground cover potential were not included in later stages of the selection process. Thus, the fact that selection was done mainly in an environment free of any type of competition is more simil ar to a hay production system rather than the grazed forage livestock system toward which RP is being targeted in the current research program. Other efforts for selecting RP have been conducted in Australia using additional criteria such as spread in pure (Bowman and Gogel, 1998) The authors reported that accession CIP 93483 (PI 231318), which has been released as cv. Prine, and CPI 93469 (PI 262833) have the ability to persist and spread on the north coast of New South Wal es. Pest and Diseases Rhizoma peanut is generally recognized as resistant to economically serious pests and diseases (i.e., early and late leaf spot caused by Cercospora arachidicola and Cercosporidium personatum respectively, and rust caused by Puccinia arachidis ; Ruttinger, 1989). Rhizoma peanut has been documented to be infested with cotton root rot ( Phymatotrichopsis omnivore ) (Barnes, 1990). Cotton root rot, also known as Texas root rot, is a naturally occurring fungal pathogen found throughout Texas and southern Oklahoma, prevalent in calcareous clay loam soils with a pH range of 7.0 to 8.5 in areas with high summer temperatures. Infected areas appear as circular patterns throughout the field where the rhizoma peanut initially dies back. Arbrook suffe rs greater losses from cotton root rot than Florigraze. Also, isolated cases of leaf spot caused by

PAGE 42

42 Phyllosticta sp Stemplylium sp. and Leptosphaerulina sp have been observed in Florida, but no long term damage has be en reported (French et al., 1994 ). To date, peanut stunt virus (PSV) may be the one most important disease in Florida affecting RP fields. Symptoms of PSV include: stunted plants, chlorosis, malformed leaves, and reduced foliage yield (Blount et al., 2002). The majority of RP fields in Flor ida are planted to Florigraze which is susceptible to PSV. Nevertheless, it is expected that acreage planted to cultivars UF Tito and Peace, reported as showing field resistance to PSV, will increase over time as an alternative to control PSV infection. E stablishment Establishment is a key phase in the life of a sward since it lays the foundation for future productivity, resistance to weed invasion, tolerance of stock trampling if grazed, and resistance to wheel traffic if cut for conservation. The aim is to provide the best conditions for germination of the sown forage seeds, for vigorous shoot and root development of the seedlings, and finally for the formation of a dense sward. Nevertheless, if RP is to be to be grown in association with grasses, then th e parameters of selection for establishment could potentially be modified. Rhizoma peanut is propagated vegetatively using rhizomes, most often with equipment designed for establishment of vegetatively propagated tropical grass pastures (e.g. conventional bermudagrass sprig digger/planter) (Williams et al., 1997; Williams et al., 2002). Establishment is also possible by broadcasting the rhizomes on the surface of a prepared seedbed followed by disking for soil incorporation. Research conducted in Florida p rovided recommendations for RP establishment. Reliable soil moisture (for 60 to 90 d after planting), a planting date with the longest frost free period (Williams, 1993; Williams et al., 1997), and chemical composition of

PAGE 43

43 planting material (Rice et al., 19 95) were reported as major determinants of the overall success of RP establishment. Patterns of carbohydrate and rhizome mass accumulation have led to recommendations that defoliation of nursery areas be minimized or avoided when rhizomes are to be used as planting material (Saldivar et al., 1992a,b). Rice et al. (1995) suggested that to minimize the risk of stand failure, RP producers should plant rhizomes with pre plant total non structural carbohydrates (TNC) and N concen trations 1 respectively. The minimum planting rate generally recommended is 8 10 Mg ha 1 (packed at ~ 79 kg m 3 ), and the rule of thumb is that 1 ha of well managed nursery area produces enough rhizomes to plant 15 to 20 ha of RP (Will iams et al., 2011). Canudas et al. (1989) evaluated planting rates (0.3, 0.6, 1.2, 2, and 3.4 Mg ha 1 ) of Arbrook and Florigraze RP rhizomes planted with a row spacing of 0.5 m. For the treatments where herbicides were used to control broadleaf and grass weeds, ground cover showed a quadratic response as planting rate increased. Ground cover measured in October of the plantin g year peaked at an approximate planting rate of 2.2 Mg ha 1 for Arbrook (80%) and 2.7 Mg ha 1 for Florigraze (60%). They reported t hat Arbrook out yielded Florigraze across their set of treatments (herbicide pla nting rate). There was a linear and quadratic response of RP dry matter (DM) harvested to planting rate for both cultivars in the year of and year after establishment. The au thors suggested that due to the cost of establishment, planting rates of 1 Mg ha 1 or lower are most likely to occur. Prine et al. (1986 a ) reported that rhizome yields range from 70 to 175 m 3 ha 1 and recommended a planting rate of 3.6 m 3 ha 1 or greater. Indeed, at that rate of planting a full stand of peanut is generally achieved by the beginning of the third year. Canudas et

PAGE 44

44 al. (1989) reported that during the year of establishment DM harvested increased linearly as planting rate increased (at a planting rate of 3 Mg ha 1 dry matter yield was ~2.5 and ~1.5 Mg ha 1 for Arbrook and Florigraze, respectively) compared to a quadratic effect during the year after establishment where maximum yield leveled off at a planting rate of 3 Mg ha 1 with ~11 and 9 Mg ha 1 for Arbrook and Florigraze, respectively. Williams (1993) evaluated planting dates (winter: December to March; and summer: June to August) and three pre plant tillage (well prepared, moderate grass competition, and no pre plant tillage) effects on RP e stablishment in existing bahiagrass swards. The swards were burned or mowed before treatments were applied in winter and summer. Regardless of treatment, first sprout emergence was 5.1 and 3.1 wk after winter and summer plantings, respectively. Sprout emer gence declined to zero 6 to 7 wk after first sprout emergence. The effect of pre plant tillage varied with planting date and year, but the trend was: well prepared > moderate grass competition = no pre plant tillage. She concluded that RP should be planted in well prepared fields during winter. Similar recommendations were given by French and Prine (1991) who suggested a planting time of February through March in Florida. Cultivar (Arbrook vs. Florigraze), planting date (February, April, June, August, and December), and location (Brooksville, Gainesville, and Quincy, FL) effects on emergence (at 2 to 12 wk post planting) and rate of cover of RP were investigated by Williams et al. (1997). The authors indicated that number of sprouts m 2 at 12 wk post planti ng ranged from 0 to >200 and, in most cases, the sprout counts obtained for April, June, August, and December planting dates were lower or equal to those from the February planting date, regardless of location or cultivar. On average, February

PAGE 45

45 plantings ac hieved >60% RP ground cover 26 wk earlier than any of the other planting dates. Also, correlations of sprout emergence and estimated ground cover were greatest (r = 0.69) for sprout emergence measured at 12 wk with ground cover measured at 26 wk post plant ing. Effects of planter type (no till vs. conventional sprig planter), ground preparation (undisturbed sod vs. rotovated), planting date (winter vs. summer), and herbicide (glyphosate [N (phosphonomethyl) glycine] vs. none), on establishment and survival o f RP were evaluated by Williams et al. (2002). The authors concluded that there was no planter type effect on RP establishment (at emergence and final sprout counts measured 12 wk post planting). Further, they reported that RP ground cover was greater for rotovated+herbicide (22%), compared to rotovated no herbicide (13%), herbicide no t rotovated (6%), and no herbicide not rotovated (2%). The authors concluded that there are options in terms of management practices to establish RP that can be adapted to pro duction goals; i.e., clean cultivation establishment for hay production or dairy cattle grazing, and sod planting without herbicide for less intensive situations. Responses to Clipping and Grazing Literature on responses to clipping and grazing is more pr evalent for monocultures of RP than for RP grass mixtures. Reports of RP grass mixture responses to clipping and grazing are limited, probably because establishment of pure stands of RP has been the assumed goal. It is likely that RP grass mixtures have oc curred primarily because of failure to control grass weeds in existing swards that initially were intended for production of RP hay. Thus, studies that look at the equilibrium/dynamics of RP grass systems (i.e. botanical composition, nutrient cycling) und er grazing, similar to the white

PAGE 46

46 clover grass systems described for temperate environments (Parsons et al., 2006), remain an area of research to be explored. One of the few studies that evaluated botanical composition in a grazed RP grass mixture as a fun ction of N fertilization (0 and 35 kg ha 1 ) and stocking rate (1.5 vs. 2.5 animals ha 1 ) was conducted by Valencia et al. (1999). The authors concluded that N fertilization increased grass herbage mass (+8 percent age points) and that over time there was a decrease in the RP component ( 9 percent age points ). The authors reported that there was no change in population of the weed Mexican tea ( Chenopodium ambrosioides L.) as a function of increased grass contribution due to N fertilization. Further, they attr ibuted the lack of botanical composition response to stocking rate to the stocking rates used being in the low range. The authors suggested that application of N may be more useful for weed control in RP grass swards when used as a preventative measure; al so, that longer term experiments (> 3 yr) would probably explain better the changes in botanical composition as a function of the treatments. Above ground biomass Under clipping, dry matter yields of RP harvested for hay range from ~7 to 13 Mg ha 1 yr 1 f or Florigraze (Prine et al., 1986 a ) and up to ~16 Mg ha 1 yr 1 for Arbrook (Beltranena et al., 1981; Prine et al. 1981, 1986 b 1990). Annual forage yields are reduced when RP is cut at intervals less than 6 wk (Beltranena et al., 1981). In a multi location study across Florida, DM yield (total of two harvests per year) of UF Tito was generally equal or greater than Florigraze, and UF Peace yield was similar to UF Tito (Quesenberry et al., 2010).

PAGE 47

47 Mislevy et al. (2008) evaluated harvest management (2.5 vs 1 0 cm stubble height) for several RP entries [Arbrook Select (local ecotype), Arbrook (released cultivar), PI 262839, PI 262826, Florigraze (released cultivar), Ecoturf (PI 262840), and PI 26283] on above ground biomass yield, nutritive value (IVOMD, CP), r oot mass, and persistence on Florida flatwood soils (Ona series). The plots were harvested every time the RP canopy reached 30 cm height. The authors reported that harvesting RP to 2.5 cm stubble height resulted in greater DM yields during the first 2 yr (average across entries was 7 Mg ha 1 ) compared to 10 cm stubble height (4 Mg ha 1 ). Dry matter yield was not different between stubble heights in Years 3 (4.2 Mg ha 1 for 2.5 cm stubble height, and 3.9 Mg ha 1 for 10 cm stubble height) and 4 (6.5 Mg ha 1 for both stubble heights). Also, the authors reported that entries Ecoturf, PI 262833, and Florigraze exhibited DM yield increases of +89%, +26%, and +54%, respectively, between Year 1 and 4, compared to the others which were similar or lower in Year 4. Th e authors reported that CP (3 yr averages) was greater for Ecoturf (192 g kg 1 ) and PI 262833 (189 g kg 1 ) compared to Florigraze (160 g kg 1 ) and Arbrook (15 0 g kg 1 ). The authors suggested that greater CP may have been the result of greater leaf to stem ratio for the Ecoturf and PI 262833 entries, although such measurements were not taken. For IVOMD, PI 262833 (799 g kg 1 ; 3 yr average) was greater than Arbrook Select (667 g kg 1 ) and Arbrook (669 g kg 1 ), with the other entries falling in between these v alues. Florigraze (3 yr average) IVOMD averaged 700 g kg 1 Overall, the authors reported that CP and IVOMD were not affected by stubble height. In spite of greater RP yields for the 2.5 cm stubble height during the first 2 yr, the authors recommended tha t RP producers on Florida Spodosols maintain a 10 cm stubble height to prevent invasion

PAGE 48

48 from weeds based on ground coverage data showing that after 4 yr the 10 cm stubble height had 38% greater RP cover than the 2.5 cm stubble height,. In a clipping study at three locations in Louisiana, Redfearn et al. (2001) evaluated the effects of harvest frequency (every 30 and 60 d) and N fertilization (0, 110, and 220 kg N ha 1 ) on CP, NDF, and in vitro true digestibility (IVTD) of Florigraze RP. The locations were conducted at latitude slightly below 31 N. The plots were clipped at 8 cm stubble height. Averaged across locations, CP concentration was 205, 207, and 205 g kg 1 for 0, 110, and 220 kg N ha 1 respectively, when harvested at a 30 d interval, and 165, 16 6, and 169 g kg 1 for 0, 110, and 220 kg N ha 1 respectively, when harvested at a 60 d interval. The authors concluded that responses of CP yield and nutritive value were influenced more by environment (primarily rainfall) than by N fertilization and harv est management. In a similar study conducted in Louisiana, Venuto et al. (1998) reported that DM yield of RP is not likely to be increased by N fertilization. Ortega et al. (1992) evaluated the effects of grazing frequency (grazing cycles of 7, 21, 42 and 63 d; a grazing cycle consisted of 2 d or fewer of grazing plus the resting period between grazing events) and grazing intensity (residual above ground dry matter after grazing of 500, 1500, and 2500 kg ha 1 ) on RP productivity (herbage accumulation) and p ersistence (botanical composition by weight). The study was conducted on a well established Florigraze RP pasture (5 yr old stand) with an average botanical composition at the start of the experiment of 90% RP and 10% common bermudagrass. Year 1 RP herbage accumulation ranged from 6130 to 10 240 kg ha 1 and increased linearly as length of grazing cycle and amount of residual dry matter increased. In Year 2, herbage accumulation increased as length of grazing cycle

PAGE 49

49 increased when residual dry matter was low, but length of grazing cycle had less effect as residual dry matter increased. Rhizoma peanut percentage was greatest with high residual dry matter and long grazing cycle, but values of 80% or greater in the second year were achieved with residual dry matt er as low as 1300 kg ha 1 when grazing cycle was 63 d, or with grazing cycle as short as 7 d when RDM was above 2300 kg ha 1 The authors concluded that to maintain 80% RP or greater when grazing cycle was 42 d required a residual dry matter after grazing of ~1700 kg ha 1 (~16 cm stubble height) or greater, but if grazing cycle was 21 d or shorter a residual dry matter of 2300 kg ha 1 (~20 cm) was required. In terms of nutritive value, RP grazed to 1800 kg ha 1 residual dry matter every 35 d had CP > 150 g kg 1 and IVOMD >700 g kg 1 (Sollenberger et al., 1989). Below ground biomass In a study by Mislevy et al. (2008), the authors reported that harvesting RP to 2.5 cm stubble height reduced below ground (root + rhizome) biomass compared to harvesting to 10 cm stubble height. Over a 4 yr period, below ground biomass was 242 and 135 g m 2 for 10 and 2.5 cm stubble height s respectively. The authors reported the same trend for all entries evaluated and suggested that lower below ground biomass for the 2.5 cm stu bble height treatment was the result of new shoot development at the expense of underground carbohydrate reserves, since most of the above ground photosynthetically active machinery (leaves) was removed. A series of studies conducted by Saldivar et al. (1 992a,b) near Gainesville, FL, evaluated the effect of defoliation frequency (2, 6, and 8 wk) on above and below ground biomass production, total non structural carbohydrate (TNC) and N concentration. The authors reported that durin g the year of establishm ent, accumulated

PAGE 50

50 total dry matter (above + below ground biomass) increased until September, when above ground (shoot) growth plateaued or declined while below ground (rhizome) growth continued. Shoot/rhizome ratios increased from zero at planting to about 1.5 to 2 by late summer, and then declined to about 0.5 in autumn. Further, the authors indicated that as defoliation frequency increased, rhizome dry matter production decreased by one half (8 wk) to two thirds (2 and 6 wk) compared to undefoliated plant s during the year of establishment. In general, belowground TNC and N concentration were more responsive to defoliation management compared to above ground tissue. The authors reported that when growth (active shoot emergence) was initiated, TNC concentrat ion in the rhizomes declined and generally ranged between 100 and 200 g kg 1 during the summer season. Toward the end of the growing season (autumn), TNC concentration increased again to levels of about 400 g kg 1 which was coupled with increasing rhizome mass. The authors indicated that N concentration in the rhizome declined in spring and then leveled off after sufficient shoot development was reached; but in general, patterns of N concentration were less marked compared to TNC. Under grazing conditions, Ortega et al. (1992) reported similar patterns of TNC as a function of defoliation intensity and frequency as reported by Saldivar et al. (1992a,b). The authors reported that lower values of RP rhizome mas were associated with low levels of residual dry m atter and short grazing cycle. They indicated that rhizome mass increased with increasing length of grazing cycle when residual dry matter was 1000 kg ha 1 and that at residual dry matter of 1700 kg ha 1 or greater, the effect of grazing cycle was neglig ible. Further, lower TNC values in the rhizomes were associated with

PAGE 51

51 lower residual dry matter as a function of less residual leaves left for photosynthesis after defoliation. Weed Control Weed control in RP swards is an active area of research. The liter ature reported has mainly focused on the use of herbicides for weed management during establishment and for maintenance of monoculture RP fields. To this point, chemical control has been the most effective practice to control competition from weeds. A shor t list of the most problematic weeds for forage managers in Florida for RP and bahiagrass fields include: common bermudagrass, nutsedges ( Cyperus sp. ), cogongrass [ Imperata cylindrical (L.) P. Beauv.], smutgrass [ Sporobolus indicus (L.) R. Br.], dogfennel [ Eupathorium capillifolium (Lam.) Small], Mexican tea, Florida pusley ( Ricardia scabra L.), tropical soda apple ( Solanum viarum Dunal); blackberry ( Rubus sp. ), and thistle [ Cirsium vulgare ( Savi ) Ten. ]. Canudas et al. (1989) evaluated the effect of post emergence herbicides paraquat (0.56 kg a.i. ha 1 ), sethoxydim (0.44 kg a.i. ha 1 ), basagran (1.12 kg a.i. ha 1 ) and dicamba + 2,4 DB amine (0.49 kg a.i. ha 1 ) to control broadleaf and grass weeds during establishment of Florigraze and Arbrook. In that experiment, seedbed preparation consisted of disking during January and application of herbicides vernol ate (2 kg a.i. ha 1 ) and benefin (1.7 kg a.i. ha 1 ) for pre emergence control of grass and broadleaf weeds. The authors reported that greatest canopy cover during the year of establishment was achieved when broadleaf and grass weeds were controlled; nevert heless, Arbrook plots in which broadleaf weeds were controlled had lower RP canopy cover than the no herbicide control. They attributed this response to growth habit differences between the RP cultivars; Arbrook with a more up right growth habit was better able to compete for

PAGE 52

52 light with broadleaf weeds than was Florigraze, while Florigraze was more competitive with the shorter grass weeds than with broadleaf weeds. The authors reported that by the end of the year of establishment RP yield was almost double in the broadleaf+grass weed controlled treatment compared to no application of herbicides. Differences between herbicide treatments in the second year were not as great as in the first, partially due to uniform control of broadleaf weeds over all plots and because peanut competes well for light and nutrients after it is established. The authors indicated that rope wick applications of herbicides like dicamba+2,4 DB amine were effective management practices for control of dogfennel in establishing peanut. No specific reference was made regarding injury of RP plants due to herbicides. In a RP field (>10 yr old) infested mainly with Mexican tea and cogongrass, Valencia et al. (1999) evaluated the application of glyphosate (1.12, 2.24, and 3.36 kg a.i. ha 1 ) a nd triclopyr (0.56, 1.12, and 1.68 kg a.i. ha 1 ) during the summer on dry matter yield and botanical composition of the sward. The authors reported that 2 mo after herbicide application Mexican tea dry matter yield decreased with increasing rate of glyphos ate (86% reduction at the highest rate compared to control); nevertheless, after 4 mo dry matter yield of Mexican tea increased from recovering treated plants, as opposed to emerging seedlings. Thus, the authors concluded that more than a single applicati on of glyphosate might be needed to control Mexican tea. There was no effect of glyphosate on cogongrass or other grasses, but there was a linear decrease in RP dry matter yield as rate of glyphosate increased. The injury (phytotoxic) effect of glyphosate on RP was observed even 4 mo after application. The authors reported no effect of increasing rates of triclopyr to control Mexican tea; at the low rates there was

PAGE 53

53 81% reduction in dry matter yield and this was consistent when measured 4 mo after treatment. Triclopyr did not control cogongrass, and it increased linearly with increasing rates of triclopyr. Ferrell et al. (2006) evaluated the use of post emergent herbicides (in kg a.i ha 1 ) 2,4 D amine (0.56); 2,4 DB (0.48); imazamox (0.06); imazapic (0.07); and hexazinone (0.28 and 0.56) on injury and yield of well established (10 yr old plots) Florigraze and Arbrook swards. Their results showed that 2,4 DB, imazamox, and imazapic can be applied either at 3 or 21 d after clipping RP with minimal risk of visua l injury or yield loss, as opposed to hexazinone which could only be applied at 3 d after clipping. Further, the authors reported that the effect of 2,4 D amine on RP injury and yield was cultivar dependent. For Arbrook, application of 2,4 D amine had no e ffect on yield with little visual injury at either 3 or 21 d after clipping. In contrast, when it was applied 21 d after clipping Florigraze there was 21% yield reduction compared to that observed when 2,4 D was applied 3 d after clipping. Yield from plots treated 3 d after clipping was no different than the control and the other herbicides. There are re sources to provide reasonably effective control of weeds growing in monoculture RP fields and in bahiagrass pastures. However, control of weeds in RP bahia grass mixtures may present a challenge because herbicides targeted to a specific weed species may injure either the desired grass or RP component of the mixture. In addition, herbicides used to control weeds in bahiagrass pastures may not be used in RP fie lds, and herbicides that have proven effective in control of weeds and cause little injury to the desired species may not be labeled for use in a specific state (e.g., imazamox in Florida in RP fields). Ferrell and Sellers (2012) compiled a list of

PAGE 54

54 herbici des to be used in RP fields in Florida. The list consists of the herbicides 2,4 D amine, imazapic, and clethodim. Further studies are needed to determine the economic/biological threshold for herbicide application for controlling weeds in RP grass swards for both grazing and hay production. Potential for Use in Grazed Pastures in the Gulf Coast USA Region In Florida and across the lower southeastern USA, RP use is mainly in hay production systems. Within these systems, the costs associated with vegetative establishment, management for weeds and water, and taking land out of production for one or more growing seasons to allow adequate establishment of the RP crop may be affordable (Adjei and Prine, 1976; Prine et al., 1986 a ; Rice et al., 1995). Rhizoma peanu t is rarely planted in association with or into existing pastures; RP grass mixtures exist because of failure to control weeds in fields that initially were intended for the production of pure RP hay. In contrast to these relatively high input systems, th e presence of even relatively small amounts of RP in low input, warm climate grazing systems, typically dominated by C4 grasses, may increase the nutritive value of the sward and the overall productivity of the system (Lascano, 1994), and reduce the need f or N fertilization. Prine (1980) found that F l origraze, once established, grows we ll in mixtures with digitgrass bermudagrass, or bahiagrass if no N was applied. Several experiments have reported the performance of animals in RP swards compared to monocu lture grasses (with and without fertilization). For example, Sollenberger et al. (1989) reported greater animal live weight gain and gain ha 1 for animals grazing RP swards compared with N fertilized bahiagrass. Similar studies have been reported in the li terature using pinto peanut (Lascano, 1994). Nevertheless, it is

PAGE 55

55 not known at this point the relative contribution of RP in terms of animal responses and N contribution to the sward as a function of varying quantities of RP in the RP grass mixture. Further it is not known if management practices may be able to account for the slow establishment characteristic of RP plants and allow utilization of the pastures planted to RP during the year of or year after establishment. In terms of nutritive value, all re leased RP lines have similar characteristics. Quesenberry et al. (2010) provided estimates for all the released lines in Florida. They indicated that Florigraze, Arbrook, Ecoturf, UF Peace, and UF Tito ranged from 170 to 200 g kg 1 for crude protein, and 3 40 to 400, 420 to 510, and 80 to 90 g kg 1 for acid detergent fiber (ADF), neutral detergent fiber (NDF) and lignin, respectively. Also, total digestible nutrients (TDN) ranged from 510 (Arbrook) to 570 (Ecoturf) g kg 1 ; relative feed value ranged from 10 5 (Arbrook) to 132 (Ecoturf) compared to the industry standard, alfalfa, which ranged from 110 to 160. In summary, RP has the potential to provide high quality forage for livestock in warm climates comparable to that provided by alfalfa (Romero et al., 198 7) in temperate regions. Several positive attributes of RP contribute to the perception of it having high potential in Florida and areas of the southern Gulf Coast USA. In terms of grazing systems, one of the most important characteristics is persistence. Persistence of RP has been reported under a wide range of management systems for hay, silage, grazing, and as an understory forage crop (Prine et al., 1981; Ortega et al., 1992; Johnson et al., 2002). Nevertheless, such studies and others that have evaluat ed animals responses have been conducted in monocultures of well

PAGE 56

56 Information is needed regarding the opportunity to utilize forage during the legume establishment phase, so that land is not totally removed from the grazi ng rotation. Novel approaches for overcoming the barriers to successful growth of legumes in association with grasses in warm climate pastures and to identify low cost, long term solutions to the problem of N limitation in low input systems are needed. A cost effective strategy to include RP in the diet of grazing animals may be very attractive to forage livestock systems that do not require feed value as high as that of pure RP stands (e.g., cow calf operations) and for which the cost of annual N fertiliz er inputs to a pure grass sward can be prohibitive. An approach that may have potential is to plant RP in strips in already existing bahiagrass pastures. The Strip Planting Approach for Rhizoma Peanut in Existing Bahiagrass Pastures Strip planting RP is p roposed as a cost effective approach for achieving mixed pastures without having to undertake costly deep tillage and herbicide applications of the entire sward when starting from an already existing bahiagrass pasture. Using this strategy, a legume is pla nted that ha s potential to spread into surrounding grass areas. In Florida, the only legume that has demonstrated sufficient persistence and spread to function in such a system is RP. It may take a pe riod of time for RP planted in strips to spread througho ut the entire pasture, but if this can be achieved it may provide a relatively low cost option for establishment of mixed legume grass pastures. The theory behind strip planting is based on the temporal physical separation of species to favor establishment of the less vigorous species. Such an approach has been widely used in Australia, especially in systems that use forage tree legumes such as leucaena and sesbania growing in association with grasses (Lesleighter and Shelton, 1986; Catchpoole and Blair, 19 90; Shelton, 1994). Additionally, by virtue of physical

PAGE 57

57 separation of the active grass component (growing in the edges of the strip) and the establishing RP in the strips, the strip planting approach provides opportunities to investigate a combination of s eedbed preparation strategies as well as cultural, chemical and mechanical weed m anagement practices that may lower inputs (i.e., number of herbicides used and/or number of applications per growing season) while allowing successful establishment of a RP b ahiagrass mixture. Additional Benefits Attributed to Spatial Separation of Species Although not the approach being evaluated in this dissertation, long term spatial separation of grasses and legumes has been evaluated. Recent literature from temperate envi ronments, mainly using white clover and ryegrass, has r eported greater animal response w hen the forage is presented to animals in separate grazing units (monoculture legume next to monoculture grass) rather than species growing intermingled. The basis for this response is the demonstrated preference of grazing animals for clover when given the opportunity to freely choose (Parson s et al., 1994). On average, studies have shown that cows freely chose approximately 70 to 80% clover and 20 to 30% ryegrass (Cosg rove et al., 1999; Chapman et al., 2007). This level of clover selection may be greater than the actual proportion offered to the animals grazing mixed pastures. As pointed out by Chapman et al. (2007), the challenge for grassland management is to present feed to animals at pasture in ways that allow them to meet their dietary preferences, while also allowing high rates of animal production per hectare. Marotti et al. (2001) reported that milk yield (kg d 1 ) was 11 and 28% greater when animals were offered a free choice of white clover and ryegrass each growing in monoculture compared to grass and legume growing intermingled and grass alone,

PAGE 58

58 respectively. Solomon et al. (2011) evaluated animal and sward responses as a function of stocking rate (low: 3 anima ls ha 1 and high: 6 animals ha 1 ) and four forage systems (spatially separated on a 50:50 ratio for grass and legume as function of paddock size, monoculture legume, monoculture grass, and grass and legume intermingled) using white clover and annual ryegr ass. The authors reported that at high stocking rate herbage mass was similar among forage systems components, but at low stocking rate, monoculture grass had the greatest herbage mass. Further, they indicated that average daily gain for animals was greate r on the spatially separated system compared to monoculture legume, but neither was different from monoculture grass and grass legume intermingled. Summary Based on the literature reported, there appears to be potential for strip planting RP in existing b ahiagrass pastures with the long term goal of achieving RP bahiagrass mixtures that increase/maintain productivity and sustainability of tropical forage livestock systems. This approach may provide opportunity to reduce establishment costs compared to prep aration of a seedbed followed by planting RP in pure stands, and may offer the opportunity to utilize grass forage from the sward during the legume establishment phase so that land is not totally removed from the grazing rotation. The following chapters de scribe studies investigating 1) the degree to which defoliation management of the overall sward during RP establishment affects the presence and subsequent success of the legume, 2) grazing management effects during the year after establishment, 3) the cha llenges of weed management and competition for nutrients and light when RP is planted in strips, and 4) planting options for peanut establishment in strips.

PAGE 59

59 CHAPTER 3 STRIP PLANTING A LEGUME INTO WARM SEASON GRASS PASTURE: DEFOLIATION EFFECTS DURING THE Y EAR OF ESTABLISHME NT Overview of Research Problem Lack of maintenance fertilization and poor grazing management are the primary factors resulting in degradation of grasslands in low input systems in some warm climate environments (Boddey et al., 2004; Miles et al., 2004) Due to their capacity to f ix N 2 from the atmosphere and their higher nutritive value compared to tropical grasses (Muir et al., 201 1 ) legumes may be an alternative source of N for grasslands (Thomas, 1995) improving the likelihood of long term persistence while maintaining and/or improving productivity and forage quality. Nevertheless, for age legumes have contributed less to livestock production systems in the tropics and sub tropics than in temperate regions. Often, C 3 legumes are overwhelmed when competing with vigorous C 4 warm climate grasses (Dunavin, 1992; Sollenberger and Collins, 200 3; Muir et al., 201 1 ). Research is critical to develop novel approaches for overcoming the barriers to successful growth of legumes in association with grasses in warm c limates and to identify low cost, long term solutions to the problem of N limitation in low input systems. One possible approach to legume establishment is strip planting in grass swards (Cook et al., 1993 ; Whitbread et al., 2009) Using this strategy, legumes are planted that have potential to spread into surrounding grass areas. It will take a period of time for legumes to spread throughout the entire pasture, but if this can be achieve d it may provide a relatively low cost option for establishment of mixed legume grass pastures. In the USA Gulf Coast Region the only legume that has demonstrated sufficient persistence and potential for spread to function in such a system is rhizoma peanu t.

PAGE 60

60 Rhizoma peanut is a warm season, vegetatively propagated, perennial legume that was introduced to Florida, USA from South America in the 1930s. Positive attributes include: drought tolerance (French, 1988), dry matter yields up to 10 to 12 Mg ha 1 yr 1 under natural rainfall conditions (Beltranena et al., 1981; Ocumpaugh, 1990), similar crude protein concentration and digestibility to alfalfa ( Medicago sativa L.) (Prine et al., 1981; Beltranena et al., 1981), and persistence under a wide range of manage ment systems for hay, silage, grazing, and as an understory forage crop (Prine et al., 1981; Ortega et al., 1992; Johnson et al., 2002). Four forage type cultivars of RP have been released by the University of Florida (Florida Agricultural Experiment Stati 1986, respectively (Prine et al., 1981; Prine et al., 1986 a,b ; Prine et al., 1990 (Quesenberry et al., 2010) Despite demonstrated potential of RP for grazing systems in the southeastern USA, it has not been used widely in pastures. High costs associated with vegetative establishment, management for weeds, and removal of land from production to allow adeq uate time for establishment (Adjei and Prine, 1976; Prine et al., 1986; Rice et al., 1995) have limited RP use primarily to high quality hay for dairy and equine rations, uses where RP production costs can be recovered in the sale of a high value commodity Unlike hay production systems, where the presence of forbs and grasses are undesirable due to reduction of the feed and market value of RP hay (Williams et al., 1991), low input forage livestock systems (e.g., cow calf operations) may not require as high a feed value. These systems would benefit from the presence of even relatively

PAGE 61

61 small amounts of RP in grazed pasture through increased nutritive value of the sward (Lascano, 1994) and reduced need for N fertilization. The premise of this experiment is th at strip planting RP in existing bahiagrass pastures may offer the opportunity to utilize grass forage during the legume establishment phase, so that land is not totally removed from the grazing rotation, while allowing successful establishment of the legu me. The specific objectives were to quantify the effect of a range of grazing and haying treatments on: 1) RP canopy cover, frequency of occurrence, and spread; 2) the light environment of establishing RP plants; and 3) bahiagrass herbage harvested or unut ilized. Materials and Methods Experimental Site The experiment was conducted for 2 yr (2010 and 2011) at the University of planted each year with RP. The site was chosen be cause of available well established pastures at this site have persisted for 30 yr, indicative of adaptation to the area. The soils at the experimental site were classified as Sparr fine sand ( loamy, siliceous, subactive, hyperthermic Grossarenic Paleudults) and Pomona san d (sandy, siliceous, hyperthermic Ultic Alaquods). Initial characterization of the surface soil (0 to 15 cm) indicated soil pH of 5.5 and Mehlich 1 extracta ble P, K, Ca, and Mg of 35, 44, 290, and 46 mg kg 1 respectively. Based on a recommended target pH of 6.0 for growth of RP, 1 Mg ha 1 of dolomitic lime [ (CaMg)(CO 3 ) 2 ] was applied to the experimental area before planting in 2010. Soil samples taken in 2011 confirmed the increase of soil pH to 6.2. Also, each year the area was fertilized at the beginning of the growing season with 60

PAGE 62

62 kg ha 1 of K, using muriate of potash (KCl, 600 g K 2 O kg 1 and 500 g Cl kg 1 ). Rainfall data ar e presented for both years (Fi gure 3 1). Total rainfall was 1103 and 1029 mm in 2010 and 2011, respectively. Last freeze events before planting in spring occurred on 8 and 14 March 2010 and 2011, respectively. First freeze events at the end of the growing season occurred on 10 and 14 N ov. 2010 and 2011, respectively. The timing of these freeze events w as typical for this location. Land Preparation and Planting In preparation for strip planting RP rhizomes into existing bahiagrass sod, strips were plowed in February with a moldboard plow and heavily disked several times to ensure grass and weed free planting area. The strips were 4 m wide and accommodated eight rows of RP, with spacing between rows of 0.5 m. The first and last rows of planted rhizomes were 0.25 m from the undisturbed edg e of bahiagrass sod. The planted strips were bounded on both sides by a 2.5 m strip of undisturbed bahiagrass sod. Florigraze RP rhizomes were planted in the prepared strip using a conventional Bermuda King sprig planter on 25 March 2010 and 5 April 2011. The planting material was obtained from a commercial farmer cooperator. The rhizomes were planted at a rate of 1000 kg ha 1 (packed at ~ 79 kg m 3 ) to approximately a 5 cm depth. After planting, the plots were cultipacked to ensure adequate soil rhizome co ntact. Planted RP strips were sprayed with herbicides Select Max (a.i. c lethodim; (E) 2 2[1 [[3 chloro 2 propenyl)oxy]imino]propyl]5 [2(ethylthio)propyl] 3 hydroxy 2 cyclohexen 1 one) and Impose (a.i. a mmonium salt of imazapic; +/ 2 [4,5 dihydro 4 met hyl 4 (1 methylethyl) 5 oxo 1 H imidazol 2 yl] 5 methyl 3 pyridinecarboxylicacid) at a rate of 0.10 kg a.i. ha 1 and 0. 07 a.i. ha 1 respectively. The herbicides were sprayed in

PAGE 63

63 a single application when weeds were 5 to 10 cm tall to control a broad spect rum of weeds (Ferrell and Sellers, 2012 ). Select Max was applied on 10 May 2010 and 10 June, 2011, and Impose was applied on 18 June 2010 and 5 July 2011. The application was done using a CO 2 pressurized backpack sprayer calibrated to deliver 187 L ha 1 at 310 kPa. The strips were sprayed using a 3.04 m wide boom, so that the bahiagrass at the edges of the strips was not sprayed. Toward the end of the growing season in both years, all plots were mowed to 10 cm stubble height to prevent seed dispersion fro m flowering plants of the weed Mexican tea ( Chenopodium ambrosioides L.). Irrigation was applied during April and May each year such that weekly rainfall plus irrigation equaled the 30 yr average weekly rainfall (18 and 20 mm per week in April and May, res pectively). Total irrigation applied in April and May 2010 was 67 and 0 mm, respectively, and in April and May 2011 was 60 and 50 mm, respectively Defoliation treatments were initiated in June and no further irrigation was provided Treatments and Design T reatments were four defoliation strategies: 1) Control (no defoliation of the planted RP strip during the establishment year with adjacent bahiagrass harvested for hay production every 28 d during the growing season to a 10 cm stubble height); 2) Hay Produ ction (RP strip and adjacent bahiagrass both harvested for hay production every 28 d to a 10 cm stubble height); 3) Simulated Continuous Stocking (pastures grazed weekly throughout the entire growing season to a 15 cm bahiagrass stubble height); and 4) Rot ational Stocking (pastures grazed every 28 d to a 15 cm bahiagrass stubble height). Bahiagrass stubble height was chosen as a reference point because RP plants were expected to be short early in the establishment period. A taller stubble height was

PAGE 64

64 chosen for the grazed treatments because animal preference was not certain before the study began, and it was not considered desirable to impose undue stress on the establishing peanut. The four treatments were replicated three times in a randomized complete bloc k design for a total of 12 experimental units. The area of an experimental unit was 9 m wide 15 m long, and each consisted of one 4 m wide strip of RP running the length of the plot and bounded on each side by 2.5 m strips of bahiagrass that also ran the length of the plot. Initiation of defoliation treatments was targeted for the end of the RP sprout emergence period, and actual timing was based on research conducted in Florida by Williams (1993) and Williams et al. (1997). They reported that RP sprout emergence began 3 to 5 wk after planting in summer and winter, respectively, and ceased by 7 wk after first sprout emergence was observed. Based on these data, defoliation treatments were initiated 7 wk after sprout emergence began. First sprout emergence occurred 4 wk after planting in both years, so treatments were applied for the first time at 11 wk after planting on 10 June 2010 and 21 June 2011. Mowing for the Control and Hay Production treatments was done using a riding lawn mower adjusted to leave a 10 cm stubble. After mowing, the clippings were removed from the plots using lawn rakes. For the grazed plots, animals used were 350 kg yearling cross bred beef heifers ( Bos sp.). The 9 m 15 m experimental units were individually fenced to maximize con trol over animal grazing. When defoliation occurred on a plot, the animals had access to both the planted RP strip and the bahiagrass bounding it. The methodology used was mob stocking meaning that a high stocking density (10 animals per plot) was used for a short grazing period (~ 0.5 to 1 h). While

PAGE 65

65 the animals were grazing, bahiagrass height was monitored frequently using a ruler and animals were removed from the plots when the average height of 10 measures per experimental unit reached 15 cm. Response Va riables Canopy Cover Rhizoma peanut canopy cover in the planted strip was estimated visually every 28 d for all treatments, and it was measured on the day after each defoliation event (except for the weekly events on the Simulated Continuous Stocking treat ment). A 1 m 2 quadrat (0.5 2 m) was placed in the center of the RP strip at two permanently marked locations in each experimental unit, so that canopy cover was estimated on the same areas over time. The 0.5 m side of the quadrat was oriented parallel t o the RP rows. Thus, the area enclosed by the quadrat included four rows of RP with the ends of the quadrat positioned so that they rested midway between the outermost RP row that was included in the quadrat and the RP row that was located just outside the quadrat. The quadrat was divided into 100, 10 by 10 cm squares (five rows of 20), and canopy cover was estimated visually by the same observer in 20 stratified 10 by 10 cm squares (four squares in each row of 20 squares) per quadrat and averaged to obt ain an overall cover per quadrat location. The average of two locations provided an estimate for each experimental unit ( Appendix A; Interrante et al., 2009) Frequency Frequency was determined on the same dates at the same quadrat locations that were used to estimate RP canopy cover. Presence or absence of RP was determined in 20 stratified 10 by 10 cm squares in each of two quadrat locations per plot. Frequency

PAGE 66

66 was calculated as the percentage of the total number of cells assessed where RP was present. The average of two locations provided an estimate for each ex perimental unit. Light Environment Ambient light environment at the top of the RP canopy was measured on the day before treatments were applied for the first time, 2 wk after application of treatments (middle of a regrowth period), and every 28 d thereaft er. This sampling time was chosen to represent average light environment during a regrowth period. Light environment was characterized using a SunScan Canopy Analysis System (Dynamax Inc., Houston, TX). The system consisted of a 1 m long quantum sensor th at was placed at the height of the RP canopy to measure transmitted photosynthetically active radiation (PAR), and an unshaded beam fraction sensor that was placed outside the plots to measure incident PAR. Thus, the light environment experienced by RP pla nts was characterized as percent of incident PAR that reached the RP canopy and was calculated by dividing the transmitted PAR by incident PAR level and multiplying by 100 to express it as a percentage. The light environment was calculated as the average o f four observations in each experimental unit. Spread Rhizoma peanut spread was measured once each yr on the day before the last clipping/grazing event of the season. A transect was positioned through the center of the RP strip running the length of the e ach plot. At the 5 and 10 m points along the 15 m transect, a line, perpendicular to the transect, was extended on each side. Spread was defined as the distance from the center of the planted RP strip to the farthest point where identifiable RP plant part s (above ground) were found. The average of the four measurements provided the estimate of RP spread for each experimental unit.

PAGE 67

67 Bahiagrass Herbage Harvested Herbage harvested was measured every 28 d prior to each grazing or clipping event for each treat ment except Simulated Continuous Stocking. In the Hay Production and Control treatments, a 1 2 m area was cut to a 10 cm stubble height using a sickle bar mower in the bahiagrass portion of the plot. The collected herbage was weighed fresh, and a subsamp le was dried at 60C until constant weight to determine dry matter concentration and to calculate herbage harvested. In the grazed treatments, two representative 0.25 m 2 quadrats were clipped to the target 15 cm stubble in the bahiagrass portion of the exp erimental unit. Less area was sampled in the grazed plots to minimize the impact of sampling on grazing time and behavior of the animals. In the Simulated Continuous Stocking treatment, sampling occurred biweekly before every second grazing event. This tre atment was sampled more frequently because a cage technique was used to restrict grazing from sampling units and cages should not remain in one area for an extended period lest forage mass or sward structure become very different than the surrounding pastu re. Thus on Simulated Continuous Stocking pastures, two 0.5 m 2 circular exclusion cages per plot were positioned at representative locations in the bahiagrass strip. A 0.25 m 2 area from the center of each caged area was harvested biweekly before grazing. T he cages were moved to a new location for the next 2 wk period as soon as grazing was completed on that plot. Herbage harvested was not measured in the RP strip because RP plants generally did not reach the target stubble height (10 and 15 cm for clipping and grazing, respectively) during the year of establishment.

PAGE 68

68 Statistical Analysis Data were analyzed as repeated measures using PROC GLIMMIX of SAS (SAS Institute, 2010). Collection date was considered a repeated measurement with an autoregressive covaria nce structure. Year and block were considered random effects. Year was considered random because a new set of plots was established each year. Treatments were fixed effects. Mean separations and pre planned contrasts were done based on the SLICEDIFF and L SMESTIMATE procedures of LSMEANS in SAS. Plots of model residuals were used to check normality, and in the case of non normal distributions, data transformations were used. Square root transformation was used for canopy cover and frequency. Treatments were considered different when P R esults and D iscussion Canopy Cover Defoliation method, sampling date, and their interaction affected canopy cover. From July through the remainder of the establishment year, grazing (Rotational or Simulated Continuous Stocking) reduced RP canopy cover compared to the Control and Hay Production treatments (Figure 3 2). The greatest RP canopy cover was achieved in August at 32 and 29% for the Control and Hay Production treatments, respectively, compared with 5 and 4% for Simulated Continuous and Rot ational Stocking, respectively (Table 3 1 ). When measured in late June of the year after establishment, defoliation method in the establishment year continued to affect canopy cover. Cover was similar for Control and Hay Production treatments (32 and 35%, respectively), and both were greater than the grazing treatments which did not differ from each other (Table 3 1; 7 and 8% for Simulated Continuous and Rotational Stocking, respectively).

PAGE 69

69 Decreasing RP canopy cover in the grazing trea tments can be attributed to apparent animal preference for RP and the other herbage that occurred in the strips planted to RP. When entering the pasture, animals first grazed closely the RP strips before beginning to graze the adjacent bahiagrass. Livestoc k selection for Arachis sp. has been reported previously for mixtures where the grass and legume components grew intermingled ( Bennett et al., 1999 ; Lascano, 2000; Valencia et al., 2001 ). Thus, while physical separation of the legume and grass compon ents of the mixture provide advantages to manage plant competition (Castillo Chapter 5 and 6 ), animal selection behavior can offset these advantages and negatively affect legume establishment. Frequency There were defoliation method, sampling date, and m ethod sampling date interaction effects. Rhizoma peanut frequency of occurrence in the planted strip followed the same pattern of response as canopy cover. By August of the establishment year, frequency was 67% for both Control and Hay Production treatme nts and 21% for both Simulated Continuous and Rotational Stocking (Figure 3 3) Measurements taken in late June of the year after establishment also followed the same trend as canopy cover. Control and Hay Production treatments were similar (82 and 76%, re spectively) and both were greater than either Simulated Continuous or Rotational Stocking treatments (26 and 32%, respectively) (Table 3 1) Thus, by 14 mo after planting, RP was present in ~ 80% of quadrats assessed in the planted strip if grazing did not occur compared to only ~ 30% where grazing occurred. Light Environment Light environment was considered to be an important response because previous research in Florida showed that it had a major impact on success of aeschynomene

PAGE 70

70 ( Aeschynomene americana L .) establishment in existing bahiagrass sods (Kalmbacher and Martin, 1983). In the current study, there was effect of sampling date and a strong trend ( P = 0.059) toward an effect of defoliation method on light environment. Greatest numerical differences a mong treatments occurr ed in August and September (Figure 3 4), so treatments were compared within those dates. In August, Control and Hay Production treatments were similar (87 and 90%, respectively; P = 0.38), Rotational Stocking (95%) was similar to Hay Production ( P = 0.08) but greater than Control ( P = 0.03), and Simulated Continuous Stocking (96%) was similar to Rotational stocking ( P = 0.74). In September there was a trend toward treatment differences ( P = 0.11) with 86 and 87% of incident PAR reachin g RP for Control and Hay Production treatments and 92% for both Rotational and Simulated Continuous Stocking treatments. Unlike the situation in which aeschynomene was overseeded into bahiagrass (although not strip planted), the current data indicate that light environment was not the critical factor influencing RP establishment and was clearly less important than defoliation method. This conclusion is supported by data showing that treatments (Simulated Continuous and Rotational Stocking) with the greatest or tending to have the greatest percentage of incident PAR in August and September had the lowest establishment year RP canopy cover. Spread At the end of the establishment year, average distance from the center of the planted strip to the most distant a bove ground RP plant part was 182 cm for Control and Hay Production treatments, 168 cm for Simulated Continuous Stocking, and 177 cm for Rotational Stocking (Table 3 1). Single degree of freedom comparisons were made using LSMESTIMATES to test the average of mechanical (Control and Hay Production)

PAGE 71

71 vs. Simulated Continuous and Rotational Stocking defoliation methods. The average of the mowed treatments was similar to Rotational Stocking ( P = 0.39) and 14 cm greater than Simulated Continuous Stocking ( P = 0.0 2). Given that the outer row of RP was planted 175 cm from the center of the strip, spread was minimal in the first year in all treatments. Simulated Continuous Stocking actually resulted in loss of plants in the outer row of the strip closest to bahiagras s resulting in a reduction in spread. Results indicate that defoliation management under the strip planting scheme is critical in order to allow potential spread of RP into the grass component of the pasture during the establishment year. Herbage Harvested Herbage harvested from the bahiagrass portion of the plots was 3.7, 3.4, 2.9, and 3.6 Mg ha 1 for the Control, Hay Production, Simulated Continuous Stocking, and Rotational Stocking treatments, respectively. Single degree of freedom comparisons indicate t hat herbage harvested from the Control and Hay Production treatments was similar to Rotational Stocking and there was a trend toward lower herbage harvested for Simulated Continuous vs. Rotational Stocking ( P = 0.21). While there were no significant differ ences in herbage harvested due to defoliation method, the trend toward lower herbage harvested in continuously vs. rotationally stocked pastures agrees with previous reports in the literature ( Jones, 1981 ; Parsons and Leafe, 1981 ; Parsons et al., 1988; Stewart et al., 2005) It is of significance to note that under the conditions of this experiment a producer who chooses not to use any type of defoliation during the establishment year would sacrifice approximately 3.5 Mg of bahiagrass forage for each hectare of bahiagrass strips in their pasture.

PAGE 72

72 Implications of the Research Grazing weekly (Simulated Continuous Stocking) or every 28 d (Rotational Stocking) reduced RP canopy cover and frequency. Greatest RP canopy cover during the year of establishment was achieved during August with 32 and 29% for the Control and Ha y Production treatments compared to 5 and 4% for Simulated Continuous and Rotational Stocking, respectively. Frequency measurements followed the same trend as canopy cover. A measurement early during the following growing season revealed that differences i n canopy cover and frequency carried over. Spread was lowest and there was a trend toward less herbage harvested in the Simulated Continuous Stocking treatment compared to the others. Competition for light was not an important factor affecting RP establish ment under the strip planting approach used in this study. The results indicate that defoliation management is critical during the year of establishment when strip planting RP into bahiagrass pastures. Due to apparent animal preference of forage in the le gume planted strips, production of hay is the best option for utilizing the grass forage during the year of establishment. It may be possible to decrease the negative impact of grazing by utilizing rest periods between grazing events longer than 28 d, but currently there are no data available to evaluate this option. Additional research is needed to evaluate the potential of longer rest periods between grazing events, and studies are needed to quantify the effect of grazing management during the year after establishment as well as the adaptation of other RP cultivars with different growth habits (i.e., more prostrate) to the strip planting approach.

PAGE 73

73 Table 3 1 Rhizoma peanut (RP) percentage canopy cover and frequency of occurrence in August of the establis hment year, June of the year after establishment, and spread at the end of the establishment year following planting in strips in bahiagrass pastures and subjected to different defoliation treatments. Data are means across three replicates and 2 yr (n = 6) Cover Frequency Spread Defoliation treatment Establishment year Year after establishment Establishment year Year after establishment Establishment year -----------------------------------------% -------------------------------------cm Control (C) 32 32 67 82 182 Hay Production (H) 29 35 67 76 182 Simulated Continuous (SC) 5 7 21 26 168 Rotational ( R) 4 8 21 32 177 Contrast P values § C + H vs. SC <0.0001 <0.0001 <0.0001 <0.0001 0.0274 C + H vs. R <0.0 001 <0.0001 <0.0001 <0.0001 0.3976 SC vs. R 0.5691 0.5628 0.9632 0.2730 0.1932 SE 3.3 5.0 4.7 5.2 7.9 C: Control treatment (no defoliation of the planted RP strip during the establishment year with adjacent bahiagrass harvested for hay produc tion every 28 d during the growing season to a 10 cm stubble height); H: Hay Production (RP strip and adjacent bahiagrass both harvested for hay production every 28 d to a 10 cm stubble height); SC: Simulated Continuous Stocking (pastures grazed weekly thr oughout the entire growing season to a 15 cm bahiagrass stubble height); R: Rotational Stocking (pastures grazed every 28 d to a 15 cm bahiagrass stubble height). Spread is the distance from the center of the planted RP strip to the farthest point where identifiable above ground RP plant parts were found. § Linear combinations of LSMEANS developed using the LSMESTIMATE statement of SAS. Standard error of a defoliation treatment mean.

PAGE 74

74 Fig ure 3 1. Monthly rainfall at the University of Florida Beef R esearch Unit, Gainesville, FL for 2010, 2011, and the 30 yr average.

PAGE 75

75 Fig ure 3 2. Canopy cover of rhizoma peanut planted in strips in existing bahiagrass pastures. Data are means of 2 yr. Errors bars represent treatment means (n = 6) one standard error

PAGE 76

76 Fig ure 3 3. Frequency of occurrence of rhizoma peanut planted in strips in existing bahiagrass pastures. Data are averages of 2 yr. Error bars represent treatment means (n = 6) one standard error.

PAGE 77

77 Fig ure 3 4. Light environment at the top of t he rhizoma peanut canopy for strip planted rhizoma peanut in existing bahiagrass pastures. Data are means across 2 yr. Error bars represent treatment means (n = 6) one standard error. PAR = Photosynthetically active radiation. //

PAGE 78

78 CHAPTER 4 GRAZING M ANAGEMENT STRATEGIES AFFECT YEAR AFTER ESTABLISHMENT PERFORMANCE OF A LEGUME STRIP PLANTED INTO WARM SEASON GRASS PASTURE Overview of Research Problem Rhizoma peanut ( Arachis glabrata Benth.; RP) is a warm season perennial forage legume that has documented persistence under grazing (Ortega et al., 1992) and ability to compete effectively with perennial grasses in mixed pastures (Williams, 1994). It is high in nutritive value, and in multi year grazing studies has resulted in gains of yearling beef steers ( B os sp.) of 0.97 kg d 1 (nearly pure stands of RP) (Sollenberger et al., 1989) and gains of 6 to 12 mo old dairy heifers ( Bos taurus ) of 0.6 kg d 1 (botanical composition of ~ 90% RP) (Hernandez Garay et al., 2004). Further, greater animal production (gain ha 1 ) of up to 130 kg has been reported for animals grazing RP compared to N fertilized (120 kg N ha 1 yr 1 ) bahiagrass ( Paspalum notatum Flgge) pastures (Sollenberger et al., 1989). The mechanism by which greater animal production occurred was sustained high nutritive value (i.e., crude protein and in vitro organic matter digestibility) of RP compared to bahiagrass, especially in the latter half of the growing season. In addition to its impact on forage nutritive value and increased animal response, the capacity of RP and other legumes to fix atmospheric N 2 make them an alternative to inorganic fertilizer as a source of N for grasslands (Muir et al., 2011) Thus, use of legumes in general and RP specifically should improve the likelihood of long term pasture persistence while maintaining and/or impro ving productivity of low input forage livestock systems (Thomas, 1994).

PAGE 79

79 Persistence of legumes in grazed warm climate grasslands has long been a factor limiting their use (Shelton et al., 2005). Ortega S et al. (1992) evaluated RP persistence under a wide range of levels of residual forage mass after grazing and length of rest season perennial gra sses) across a relatively wide range of levels of rest interval if the residual dry matter after a grazing event was ~ 1700 to 2300 kg ha 1 (15 to 20 cm stubble height). In spite of documented persistence and impact on animal performance in pastures, RP use in the USA Gulf Coast Region has been limited primarily to hay production due to the costs associated with vegetative establishment, management for weeds and water, and taking land out of production for one or more growing seasons to allow adequate est ablishment (Adjei and Prine, 1976; Prine et al., 1986 a,b ; Rice et al., 1995). In an effort designed to reduce establishment costs and expand use of RP in low input grazed grassland systems in the Gulf Coast Region, planting RP in strips into existing bahia grass pastures was proposed (Castillo, Chapter 3). It was shown that either rotational or simulated continuous stocking during the establishment year greatly reduced RP cover and frequency of occurrence. Defoliation by mowing or no defoliation were superio r to grazing. Establishment of RP is thought to require two growing seasons under most conditions (Prine et al., 1986 a,b ), but there are no data describing the effect of grazing in the year after planting on RP establishment and spread. The current experim ent investigates grazing management strategies in Year 2 (Y2) plots (year after establishment) where different defoliation strategies were imposed

PAGE 80

80 in Year 1 (Y1; year of establishment). The specific objectives were to evaluate: 1) the effect of year after establishment grazing management of bahiagrass swards strip composition, and 2) the interaction effects of defoliation strategies during the establishment year with year after est ablishment grazing management of bahiagrass swards strip planted to RP. Materials and Methods Experimental Site The experiment was conducted for 1 yr (2011) at the University of Florida Beef year is being conducted in 2012, but data are not yet available for inclusion in the dissertation. The experiment was imposed on plots that were subjected to different defoliation management options during the year of establishment. Establishment year resp onses were described in Chapter 3. The soils at the experimental site were classified as Sparr fine sand ( loamy, siliceous, subactive, hyperthermic Grossarenic Paleudults) and Pomona san d (sandy, siliceous, hyperthermic Ultic Alaquods). Initial characte rization of the surface soil (0 to 15 cm) indicated soil pH of 6.2 and Mehlich 1 extractable P, K, Ca, and Mg of 43, 46, 468, and 63 mg kg 1 respectively. Based on the recommendations for growth of RP, the area was fertilized at the beginning of the growi ng season with 30 kg ha 1 of K, using muriate of potash (KCl, 600 g K 2 O kg 1 and 500 g Cl kg 1 ). Rainfall and temperature data from the experimental period are presented in Chapter 3. Total rainfall in 2011 was 1029 mm compared to the 30 yr average of 123 8 mm. Last freeze event before planting occurred on 14 March. First freeze event at the end of the growing season occurred on

PAGE 81

81 10 Nov. 2011. These dates are typical for this location and do not differ to a large extent f ro m long term averages. Treatments an d Experimental Design The treatments for this study were imposed in 2011 on the plots used for the first year (2010) of the study described in Chapter 3. Each original 9 m wide 15 m long plot from the earlier study was divided into three sub plots of 9 m width and 5 m length. The 9 m width of the sub plot consisted of a 4 m wide strip planted to Florigraze in 2010 bounded on both sides by 2.5 m wide strips of bahiagrass sod. Treatments were the factorial combinations of four Year 1 (Y1) defoliation strate gies and three Year 2 (Y2) grazing management treatments, for a total of 12 treatments. Treatments were replicated three times in a randomized complete block design and were arranged as a split plot experiment with Y1 treatment as the main plot and Y2 trea tment as the sub plot. Main plot defoliation strategies imposed in 2010 were described in detail in Chapter 3. In summary, there were four treatments: 1) control (no defoliation of the planted RP strip with adjacent bahiagrass harvested every 28 d to 10 cm stubble height); 2) hay production (RP strip and adjacent bahiagrass both harvested every 28 d to 10 cm stubble height); 3) simulated continuous stocking (pastures grazed weekly to 15 cm bahiagrass stubble height); and 4) rotational stocking (pastures gra zed every 28 d to 15 cm bahiagrass stubble height). Year 2 grazing management treatments applied to subplots were: 1) simulated continuous stocking (SC; same as Treatment 3 from Y1); 2) rotational stocking with 28 d resting period (RS 28; same as treatment 4 from Y1); and 3) rotational stocking with 42 d rest period between grazing events (RS 42). At

PAGE 82

82 each grazing event in 2011, a ll sub plot treatments were grazed until the bahiagrass stubble height was 15 cm The four main plot treatments were included bec ause they resulted in varying levels of RP contribution and because differences carried over to the year after establishment. The results in Chapter 3 indicate that grazing during the establishment year greatly reduced RP contribution compared to control a nd hay production treatments. This provided opportunity to investigate grazing management options during Y2 under a wide range of initial levels of RP starting conditions. For the sub plot treatment, the general guidelines to determine treatments were bas ed on a study by Ortega S. et al. (1992). They evaluated the effects of grazing frequency and intensity on persistence and herbage accumulation of a well established (5 yr old) Florigraze RP stand that contained 10 to 20% perennial grasses at the start of the trial. The authors indicated that to maintain experiment (based on botanical composition by weight), the residual dry matter after a rotational stocking event should be approximately 1700 to 2300 kg ha 1 (15 to 20 cm stubble height). The me thodology used to impose the treatments was mob stocking, meaning that a high stocking density (3 animals per plot) was used for a short grazing period (~ 0.5 to 1 h). While the animals were grazing, height of the bahigrass sod growing at the edges of the RP strip was monitored frequently using a ruler and animals were removed from the plots when the average height of 10 measures per experimental unit reached 15 cm. Grazing started on 28 June when average sward height was 20 cm based on RP and bahiagrass a nd continued according to the treatment schedule throughout the growing

PAGE 83

83 season. After the 28 June grazing (first grazing event), there were three and two grazing events for the RS 28 and RS 42 treatments, respectively. Response Variables Measurements of RP contribution were canopy cover, frequency, spread, and botanical composition by weight. Longer term as opposed to monthly changes in RP contribution were considered of importance. Initial measurements (taken before Y2 treatments were applied) that served as point of comparison for data collected later in the year included canopy cover and frequency on 28 June and spread which was measured before the last clipping/grazing event at the end of the 2010 growing season (Chapter 3). Canopy Cover and Frequency Rhizoma peanut canopy cover in the planted strip was estimated visually near mid season (August) and at the end of the season (October). A 1 m 2 quadrat (0.5 2 m) was placed in the center of the RP strip at a permanently marked location in each experiment al unit, so that canopy cover was estimated on the same area over time. The 0.5 m side of the quadrat was oriented parallel to the RP rows. Thus, the area enclosed by the quadrat included four rows of RP with the ends of the quadrat positioned so that they rested midway between the outermost RP row that was included in the quadrat and the RP row that was located just outside the quadrat. The quadrat was divided into 100, 10 by 10 cm squares (five rows of 20), and canopy cover was estimated visually by the same observer in 20 stratified 10 by 10 cm squares (four squares in each row of 20) per quadrat and averaged to obtain an overall cover per quadrat location (Appendix A).

PAGE 84

84 Frequency of occurrence is a measurement of the relative distribution of RP in the strip. It was determined on the same dates at the same quadrat locations that were used to estimate canopy cover. Presence or absence of RP was determined in 20 stratified 10 by 10 cm squares per quadrat location. Frequency was calculated as the percenta ge of the total number of cells assessed where RP was present. Spread Rhizoma peanut spread was measured once on the day before the last grazing event of the season (17 Sept. 2011). A transect was positioned through the center of the RP strip running the length of the plot. At the 1.5 and 3.5 m points along the 5 m transect, a line, perpendicular to the transect, was extended on each side. Spread was defined as the distance from the center of the planted RP strip to the farthest point where identifiable a bove ground RP plant parts were found. The average of the four measurements provided the estimate of RP spread for each experimental unit. Botanical Composition Botanical composition by weight in the middle of each strip was estimated two times (near mid s eason in August and toward the end of the growing season in September) by clipping two 0.25 m 2 quadrats per plot to a 10 cm stubble height in the middle of each RP strip. Regrowth was at least 3 wk old when sampling occurred. Fresh herbage was collected an d separated into grass and RP components and dried at 60C until constant weight. Botanical composition was calculated by dividing weight of the RP component by the sum of all other herbage plus RP. There were no botanical composition data collected from t he SC treatment because RP plants did not reach the 10 cm sampling height.

PAGE 85

85 Bahiagrass Herbage Harvested Herbage harvested was measured prior to each grazing event for the two rotational stocking treatments. One 0.25 m 2 quadrat was clipped to 15 cm stubble in the bahiagrass portion of the experimental unit. The collected herbage was weighed fresh, and dried at 60C until constant weight to determine dry matter concentration and to calculate herbage harvested. In the simulated continuous stocking treatment, sampling occurred biweekly before every second grazing event. This treatment was sampled more frequently because a cage technique was used to restrict grazing from sampling units and because cages should not remain in one area for an extended period lest f orage mass or sward structure become very different than the surrounding pasture. Thus, on SC pastures, one 0.5 m 2 circular exclusion cage per plot was positioned at a representative location in the bahiagrass strip. A 0.25 m 2 area from the center of the c aged area was harvested biweekly before grazing. The cages were moved to a new location for the next 2 wk period as soon as grazing was completed on that plot. Herbage harvested was not measured in the RP strip because RP plants generally did not reach the target stubble height (15 cm) during Y2. Statistical Analysis Data were analyzed using PROC GLIMMIX of SAS (SAS Institute, 1996). Sampling date was considered a repeated measurement with an autoregressive covariance structure. Block was considered a rando m effect. The Y1 defoliation treatment and its interactions were included in the model to look at the cumulative effects of Y1 and Y2 treatments. In the analysis, Y1 defoliation treatment, Y2 grazing treatment, and their interactions were considered fixed effects.

PAGE 86

86 Interaction effects were analyzed with the SLICE procedure, and mean separation was based on the PDIFF and SLICEDIFF procedure of LSMEANS using SAS. Plots of model residuals were used to check normality, and in the case of non normal distribution s, data transformations were used. Square root transformation was used for canopy cover, frequency, and botanical composition data. Treatments were considered different when P P Results a nd Discussion To ensure clarity of data interpretation, it should be noted that although Y1 (2010) treatment effects are tested in this chapter, all measurements reported were taken in 2011, fall 2010. Canopy Cover and Frequency For RP canopy cover, there were effects of Y1 defoliation strategy, sampling date, and Y2 grazing management treatment sampling date interactio n. For RP frequency, treatment effects were similar to those for canopy cover, except for a trend toward Y2 grazing management treatment sampling date interaction ( P = 0.08). Averaged across sampling dates in 2011 canopy cover and frequency for Y1 contr ol and hay production treatments were not different (Table 4 1; 13 and 66%, respectively for control; 15 and 59%, respectively, for hay production). Canopy cover of Y1 hay production treatment was greater than Y1 simulated continuous (5%) and rotational st ocking (4%). Canopy cover in the two grazing treatments was not different. Frequency was greatest for Y1 control and hay production treatments compared to simulated continuous and rotational stocking (28 and 22%, respectively; Table 4 1).

PAGE 87

87 There was sampl ing date Y2 grazing management treatment interaction. Canopy cover in mid season was greatest for RS 42 (12%) compared to 6% for both SC and RS 28 (Fig. 4 1). By late season, canopy cover was not different among treatments and was 6, 6, and 4% for SC, RS 28, and RS 42, respectively. Frequency followed the same pattern as cover. There was a trend ( P = 0.08) for greater decrease ( 12 percentage points) in RP frequency for RS 42 compared with RS 28 ( 10) and SC ( 7) from the initial measurement to late seaso n, respectively. The results indicate that regardless of the grazing management treatment used in Y2, observed cattle preference for herbage in the strip and associated heavy defoliation of RP at every grazing event overrode the potential benefit of longer resting periods between grazing events for the RS 42 treatment Previous research reported canopy cover and frequency of 30 and 67%, respectively, in the establishment year for planted RP strips that were managed for hay production or not defoliated (Chap ter 3). These values compared favorably to less than 5% cover and 20% frequency when plots were grazed (either simulated continuous or rotational stocking) (Chapter 3). When RP is planted in a prepared seedbed in pure stand for hay production, it is typica l that canopy cover increases over time to provide complete cover by the end of the second or third year (Prine et al., 1986 a, b ). The decrease in RP cover and frequency in Y2 indicates that both continuous and rotational stocking as practiced in this exp eriment had a negative impact. The sharp decline in RP cover and frequency from mid to late season was unexpected. This may have been due to the cumulative effect of defoliation throughout the growing season and because toward the end of the growing seaso n, prioritization of RP growth shifts to below ground

PAGE 88

88 biomass (Saldivar et al., 1992 a ). In contrast, SC and RS 28 showed marked declines by mid season, likely a function of more frequent grazing of those treatments than RS 42. Lower RP canopy cover and fre quency with grazing were attributed previously to observed animal preference for forage in the planted strip during the year of establishment (Chapter 3). Animal preference for the RP strip continued in the current experiment during Y2. The result was that the planted strip was regularly grazed below the 15 cm target for bahiagrass. Thus, in order to achieve establishment of a RP bahiagrass mixture using the strip planting approach and with grazing in Y2, frequency and intensity of grazing need to be select ed to favor RP, with less attention paid to the height of bahiagrass outside the strip. Well established Florigraze RP has demonstrated persistence under grazing (Ortega S et al., 1992); nevertheless, even for established stands grazing intensity (i.e., s tubble height, residual dry matter) was a critical factor determining RP contribution. Ortega S et al. (1992) reported that very close defoliation of RP (i.e., to 500 kg ha 1 residual dry matter) reduced legume contribution, regardless of the length of th e resting period between subsequent defoliation events. Although establishing Florigraze does not thrive under close grazing, it may be that other RP cultivars with different growth habit and morphology (i.e., more prostrate vs. upright), are more tolerant of close grazing when planted in strips; however, that information is not available. Botanical Composition For botanical composition there were effects of Y1 defoliation strategy, Y2 grazing management, sampling date, sampling date Y2 defoliation manage ment interaction, and a trend ( P = 0.07) toward Y1 Y2 defoliation treatment interaction. In general, the

PAGE 89

89 Botanical composition data were analyzed by Y1 defoliation strategy to explore the trend ( P = 0.07) toward Y1 defoliation strategy Y2 grazing management interaction. The RP percentage was greater for Y2 treatment RS 42 than RS 28 for plots that in Y1 had been managed for hay production or stocked rotationally (Table 4 2). There was only a trend ( P = 0 .16) toward an increase in RP percentage with longer Y2 rest period in Y1 control plots, and there was no effect of Y2 management on Y1 plots that received the simulated continuous stocking treatment (Table 4 2). Analysis by Y2 grazing strategy indicated t hat for pastures grazed every 42 d in Y2 (RS 42) botanical composition was greatest for plots that during Y1 had received either the control or hay production treatments (Table 4 2). Within Y2 treatment RS 28, the Y1 control had greater percentage RP than Y1 rotational but was not different than any other Y1 treatment (Table 4 2). Because of the sampling date Y2 grazing management strategy interaction, data were analyzed by sampling date and by Y2 grazing management strategy. Rhizoma peanut contribution a t mid season was greater for RS 42 (12%) compared to RS 28 (2%; Table 4 3). At late season there were no differences among grazing treatments, with RP contribution increasing from mid to late season by 9 percentage units for RS 28 compared to a trend towa rd a decrease for RS 42 ( 4 percentage points). It is not completely clear why the contrasting trends for RS 2 8 and RS 42 treatments occurred from mid to late season since botanical composition measurements were taken after ~3 wk of regrowth following the previous grazing event for both treatments.

PAGE 90

90 Spread There was Y1 defoliation strategy effect on spread, but control and hay production treatments were not different (201 and 197 cm, respectively) T hey were greater than both Y1 rotational and simulate d continuous stocking (160 and 163 cm, respectively). Given that Y2 grazing management effects were not significant and there was no interaction of Y1 defoliation treatment Y2 grazing treatment, Y2 grazing treatment means were averaged across sub plots an d the year effect (Y1 vs Y2) of Y1 treatments on spread was compared. Year was included in the model as a fixed effect to allow estimation of RP spread into bahiagrass over time. Results indicated spread was greater in Y2 (~ 30 cm) for areas that were mana ged using the control and hay production treatments in Y1 compared to either of the Y1 grazing treatments (Fig. 4 2). A similar effect was reported by Castillo (Chapter 3 ) in plots that were managed for hay production during 2 yr. Bahiagrass Herbage Harve sted There were no treatment effects on bahiagrass herbage harvested. Herbage harvested averaged 1.7 Mg ha 1 yr 1 Implications of the Research For establishment year treatments, Year 2 RP canopy cover was not different for control and hay production (13 and 15%, respectively). The c ontrol was not different than simulated continuous (5%), but it was greater than rotational stocking (4%); the grazing treatments were not different. Year 2 RP frequency was not different for Y1 control and hay production t reatments (66 and 59%, respectively) with both being greater than simulated continuous (28%) and rotational stocking (22%). For Y2 grazing treatments, canopy cover and frequency were greatest for RS 42 in mid season (12 and

PAGE 91

91 49%, respectively); nevertheless by late season there were no differences among treatments (approx. 5% cover and 40% frequency). In general, the RP component was the end of the season. The results indicate that while defoliation strategy during the year of establishment set the starting point for RP contribution, grazing during the year after es tablishment can override the potential positive effects of the suggest that if RP planted in strips is to be grazed in Year 2, grazing management should be focused on management strategies targeted to the RP, as oppos ed to the bahiagrass growing along side the strips Specifically, grazing should be no more frequent that every 42 d, and cattle should be removed when the planted strip is grazed to a height of 15 20 cm.

PAGE 92

92 Table 4 1. Effect of Year 1 (2010) defoliation str ategy on Year 2 (2011) canopy cover and frequency of rhizoma peanut planted in 2010 Data are means across three replicates, two sampling dates, and three grazing management treatments (n = 18). Year 1 defoliation strategy Canopy cover Frequency ------------------------% -----------------------Control 13 ab 66 a Hay production 15 a 59 a Simulated continuous stocking 5 bc 28 b Rotational stocking 28 d 4 c 22 b SE 3 6 Means within a column not followed by the same letter are different ( P < 0. 05). Table 4 2. Year 1 (2010) defoliation strategy Year 2 (2011) grazing management interaction effect on rhizoma peanut (RP) botanical composition in strips planted to RP in existing bahiagrass in 2010. Data ar e means across three replicates and two s ampling dates (n = 6) from year 2011. Year 2 grazing management Year 1 defoliation strategy RS 28 RS 42 P value ------------------% -------------------Control 10 a 13 a 0.16 Hay production 5 ab 15 a 0.001 Simulated continuous 5 ab 6 b 0.50 Rotational stocking 3 b 7 b 0.04 SE 1 1 RS 28 = rotational stocking with a 28 d interval between grazing events; RS 42 = rotational stocking with a 42 d interval between grazing events. Means within a column not followed by the same letter a re different ( P = 0.05). Table 4 3. Botanical composition in 2011 strips planted to RP in 2010 into existing bahiagrass. Data are means across three replicates, two 2011 sampling dates, and four defoliation strategies from 2010 (n = 24) Sampling date Year 2 grazing management Mid season Late season P value ---------------------% ---------------------Rotational stocking 42 d 12 a 8 0.061 Rotational stocking 28 d 2 b 11 <0.0001 SE 1 2 Means not followed by the same letter are differen t within columns.

PAGE 93

93 Fig ure 4 1. Canopy cover and frequency of rhizoma peanut in 2011 as affected by 2011 grazing management treatment following planting in strips in bahiagrass pastures in 2010. Initial, mid and late season da tes correspond to June, August, and October, respectively, in 2011. Error bars represent treatment means (n = 12) one standard error.

PAGE 94

94 Fig ure 4 2. Effect of year of establishment defoliation management on spread measureme nts taken at the end of the establishment year of 2010 (Y1) and the year after establishment of 2011 (Y2). Spread is the distance from the center of the planted RP strip to the farthest point where identifiable above ground RP plant parts were found. Error bars represent treatment means (n = 3) 1 standard error. P < 0.05 P < 0.05 NS NS

PAGE 95

95 CHAPTER 5 STRATEGIES TO CONTROL COMPETITION TO STRIP PLANTED LEGUME IN A WARM SEASON GRASS PASTURE Overview of Research Problem Rhizoma peanut is a warm season, vegetatively prop agated, perenn ial legume with potential for incorporation into low input forage livestock systems in the USA Gulf Co ast Region (French et al., 1994 ). Positive attributes of RP include drought tolerance (French, 1988), dry matter yields up to 10 to 12 Mg ha 1 yr 1 under natural rainfall conditions (Beltranena et al., 1981; Ocumpaugh, 1990), similar crude protein concentration and digestibility to alfalfa ( Medicago sativa L.) (Prine et al., 1981; Beltranena et al., 1981), and persistence under a wide range of management sy stems for hay, silage, grazing, and as an understory forage crop (Prine et al., 1981; Ortega et al., 1992; Johnson et al., 2002). Further, due its capacity to fix N 2 from the atmosphere and higher nutritive value compared to tropical grasses (Muir et al., 2011) RP may also be an alternative source of N for grasslands, improving the likelihood of long term persistence while maintaining and/or improving pr oductivity of low input forage livestock systems (Thomas, 1994). In spite of these advantages, high costs associated with vegetative establishment, management for weeds, and taking land out of production for one or more growing seasons to allow adequate es tablishment of RP, have limited its commercial use primarily to production of high quality hay for dairy and equine rations (Adjei and Prine, 1976; Prine et al., 1986 a, b ; Rice et al., 1995). Establishment of RP is generally slow and competition from weeds has been reported to affect early growth of RP when planted in pure stand and growing in RP bahiagrass mixtures (Canudas et al., 1989; Williams, 1994; Valencia et al., 1999). Herbicides have been the most used and

PAGE 96

96 effective practice to control competition from weeds in newly planted RP fields. Ferrell and Sellers (2012) compiled a list of labeled herbicides for use in RP pastures. Planting RP in strips is currently being evaluated as an alternative strategy for introducing RP into existing bahiagras s past ures The goal is to reduce establishment cost to make use of RP feasible for low input systems like beef cow calf production. If less expensive establishment can be achieved, RP has demonstrated ability to persist and spread in mixtures with bahiagrass (O rtega S. et al., 1992). Under the strip planting approach, initial physical separation of the legume and grass provides opportunities for specialized cultural, chemical, and mechanical management practices that may lower inputs (e.g., number of herbicide a pplications per growing season) required for successful establishment of a RP bahiagrass mixture. Additionally, there is potential to utilize the bahiagrass forage during the establishment year, thereby minimizing the negative impact to the overall grazing program. There is little existing information describing the effects of various management strategies on establishment of strip planted RP. Thus, the objectives were to determine the effect of 1) weed management strategies in the planted strip and 2) a st arter application of N fertilizer on strip planted RP establishment and spread. M aterials and Methods Experimental Site The experiment was conducted for 2 yr (2010 and 2011) at the University of lle, FL, with a new area planted each year. The site was chosen because of available well established (at least pastures and because RP had persisted in adjacent grazed pastures for at least 30 yr, indicating adapt ation of RP to this growing

PAGE 97

97 environment. The soil was classified as Sparr fine sand ( loamy, siliceous, subactive, hyperthermic Grossarenic Paleudults). Initial characterization of the surface soil (0 to 15 cm) indicated soil pH of 5.5 and Mehlich 1 extract able P, K, Ca, and Mg of 35, 44, 290, and 46 mg kg 1 respectively. Based on a recommended target pH of 6.0 for growth of RP, 1 Mg ha 1 of dolomitic lime [ (CaMg)(CO 3 ) 2 ] was applied to the experimental area before planting in 2010. Soil samples taken before planting in 2011 confirmed the increase of soil pH to 6.1. The area was fertilized with 60 kg K ha 1 yr 1 using muriate of potash (KCl, 600 g K 2 O kg 1 and 500 g Cl kg 1 ) at the beginning of the growing season. Detailed rainfall and temperature data duri ng the years of the experiment were presented in Figure 3 1 (Chapter 3) To generalize, total rainfall was 1103 and 1029 mm in 2010 and 2011, respectively, compared to the 30 yr average of 1238 mm. First and last freeze events of the growing season occurre d on 8 March and 10 Nov. in 2010, and 14 March and 14 Nov. in 2011, respectively, and these dates did not differ to a large extent from long term averages. Land Preparation and Planting Prior to strip planting RP in the existing bahiagrass sod, strips wer e plowed in February with a moldboard plow and heavily disked several times to ensure grass and weed free planting area. The strips were 4 m wide and accommodated eight rows of RP, with spacing between rows of 0.5 m. The first and last rows of planted rhi zomes were 0.25 m from the undisturbed edge of bahiagrass sod. The planted strips were bounded on both sides by a 2.5 m strip of undisturbed bahiagrass sod. Florigraze RP rhizomes were planted in the prepared strip using a conventional Bermuda King sprig p lanter during late winter (25 Mar. and 5 Apr. 2010 and 2011, respectively). The planting material was obtained from a commercial farmer cooperator. The rhizomes

PAGE 98

98 were planted at a rate of 1000 kg ha 1 ( packed at ~ 79 kg m 3 ) to approximately a 5 cm depth. A fter planting, plots were cultipacked to ensure adequate soil rhizome contact. Irrigation was applied during April and May each year such that weekly rainfall plus irrigation equaled the 30 yr average weekly rainfall (18 and 20 mm per week in April and May respectively). No irrigation was provided thereafter. Treatments and Design Treatments were the factorial combinations of two N rates and six weed management strategies. Nitrogen rates were 0 and 50 kg N ha 1 yr 1 Mechanical and chemical weed managemen t strategies were evaluated. They were: 1) control (no herbicide or mowing in the planted strip), 2) mowing (entire plot clipped every 28 d to 10 cm stubble height simulating a bahiagrass hay production system), and the application of herbicides: 3) pendim ethalin at a rate of 0.93 kg a.i. ha 1 at planting, 4) clethodim at a rate of 0.10 kg a.i. ha 1 applied when grass weeds were 10 to 15 cm tall, 5) imazapic at a rate of 0.07 kg a.i. ha 1 when grass or broadleaf weeds were 5 to 10 cm tall, and 6) imazapic (0.07 29 kg a.i. ha 1 ) mixed with 2,4 D amine at a rate of 0.28 kg a.i. ha 1 when grass or broadleaf weeds were 5 to 10 cm tall. The 12 treatments were assigned to experimental units as a factorial arrangement in a randomized complete block design and we re replicated three times. Experimental units were 3 m long 9 m wide, with a 1 m border between the lengths of the plots. Nitrogen was applied once per year to the entire plot (i.e., both the strips planted with RP and the adjacent bahiagrass). Applicat ion of N occurred 2 wk after completion of herbicide treatments (18 May 2010 and 29 June 2011). The source of N was NH 4 NO 3 fertilizer (340 g N kg 1 ). Addition of 50 k g N ha 1 y r 1 was chosen because it

PAGE 99

99 approximates the average amount of N fertilizer applie d per year to grazed bahiagrass pastures in FL (Mackowiak et al., 2008). The herbicides and rates of application were based on previous research, and these specific herbicides were chosen because they are the only ones labeled for use in RP pastures in Fl orida (Ferrell and Sellers, 20 12 ). Pendimethalin is an exception to this criterion and is not labeled for use in RP. It was included as a treatment because of its use as a pre plant incorporated herbicide in plantings of annual peanut ( Arachis hypogea L.) to manage competition from annual grasses ( Prostko et al., 2001 ; Johnson et al., 2002; Mosler and Aerts, 2010). Herbicide treatments were applied once per growing season and only to the RP strips. The strips were sprayed using a 3.04 m wide boom u sing a CO 2 pressurized backpack sprayer calibrated to deliver 187 L ha 1 at 310 kPa. The m owing treatment was first applied ~ 11 wk after planting (9 June 2010 and 28 June 2011), coinciding with the anticipated end of the sprout emergence period, and followed the approach described by Castillo (Chapter 3 ). Timing was based on data reported by Williams (1993) and Williams et al. (1997), who indicated that sprout emergence continued for 7 wk after first sprouts emerged. In the control and all herbicide treatments, th e planted strip was not mowed during the growing season but the bahiagrass bordering the planted strip was mowed to 10 cm stubble every 28 d. This occurred at the same time as the entire plot of the Mowing treatment was clipped. Response Variables Canopy Cover Rhizoma peanut canopy cover in the planted strip was estimated visually for all treatments starting at the end of the shoot emergence period and every 28 d thereafter

PAGE 100

100 during the growing season. Cover was measured on the day after each defoliation eve nt of the Mowing treatment. A 1 m 2 quadrat (0.5 2 m) was placed in the center of the rhizoma peanut strip at a fixed location so that canopy cover was estimated each time in the same area. The 0.5 m side of the quadrat was oriented parallel to the RP row s. Thus, the area enclosed by the quadrat included four rows of RP with the ends of the quadrat positioned so that they rested midway between the outermost RP row that was included in the quadrat and the RP row that was located just outside the quadrat. T he quadrat was divided into 100, 10 by 10 cm squares (five rows of 20), and canopy cover was estimated in 20 stratified 10 by 10 cm squares (four squares in each row of 20 squares) per quadrat and averaged to obtain an overall cover per experimental unit (Appendix A; Interrante et al., 2009). Frequency Frequency is a measurement of the distribution of RP in the planted strip. It was determined on the same dates at the same quadrat locations that were used to estimate RP canopy cover. Presence or absence of RP was determined in the 20 stratified 10 by 10 cm squares and was calculated as the percentage of cells where RP was present divided by the total number of cells. Light Environment Ambient light environment at the top of the RP canopy was measured 2 wk before the end of the shoot emergence phase and every 28 d thereafter. Measurements on all experimental units were taken between 1200 and 1500 h Eastern Daylight Time on Day 14 of each of the 28 d regrowth periods of mowing treatment. Light environment was characterized using a SunScan Canopy Analysis System (Dynamax Inc., Houston, TX). The system consists of a 1 m long quantum sensor that was placed at the height of the

PAGE 101

101 RP canopy to measure transmitted photosynthetically active radiation (PAR), and an unshaded beam fraction sensor that was placed outside the plots to measure incident PAR. Thus, the light environment experienced by RP plants was characterized as percent of incident PAR that reached the RP canopy and was calculated by dividing the trans mitted PAR by incident PAR and multiplying by 100 to express it as a percentage. The average of three observations per experimental unit provided an estimate of light environment. Canopy Height and Spread Rhizoma peanut canopy height and spread were measu red on the day before the last clipping event of each year (29 Sept. 2010 and 17 Sept. 2011). Four measurements per plot were averaged to provide the estimate for each experimental unit. Canopy height measurements were intended to describe canopy developme nt and interaction with treatments and to address concerns as to whether the application of 2,4 D herbicide during the year of establishment altered RP growth. Canopy height was estimated using a ruler to measure the distance from the soil surface to the n on extended height of the RP canopy. Spread was defined as the distance from a transect running through the length of the center of the planted strip to the farthest point where above ground RP plant parts were found. Spread was measured on each side of th e transect at 1 and 2 m from the end of the plot for a total of four observations per experimental unit. Year after Establishment Measurements Canopy cover, botanical composition by weight, and spread of RP were measured the year after RP establishment w ith the purpose of estimating treatment carryover effects. During the year after establishment the entire plot of all treatments

PAGE 102

102 was clipped to 10 cm stubble height every 28 d, simulating a bahiagrass hay production system. Canopy cover and botanical compo sition were measured in the middle of the growing season (28 July 2011 and 31 July 201 2 ). Spread was measured a t the end of the growing season in 2011 and will be measured at the end of the growing season in 2012. Canopy cover and spread methodology were t he same as described earlier. Botanical composition by weight was estimated by clipping one 0.25 m 2 quadrat to a 10 cm stubble height in the middle of each RP strip. Fresh herbage was collected and separated into grass and RP components. They were dried se parately at 60C until constant weight, and botanical composition was calculated. Statistical Analysis Data were analyzed as repeated measures using PROC GLIMMIX of SAS (SAS Institute, 2010). Sampling date was considered the repeated measurement with an a utoregressive covariance structure. Year and block were considered random effects. Year was considered random because a new set of plots was established each year. Treatments and their interactions were fixed effects. In the case of second and third order interactions, simple effects were analyzed using the SLICE procedure of SAS. Mean separation was based on the PDIFF and SLICEDIFF procedure of LSMEANS. Plots of model residuals were used to check normality, and in the case of non normal distributions, dat a transformations were used. Square root transformation was used for canopy cover and botanical composition data. Treatments were considered different when P

PAGE 103

103 Re sults and D iscussion Canopy Cover and Frequency There were effects of weed management st rategy and sampling date. Second and third order interactions were significant with the exception that there was only a trend toward N application sampling date interaction ( P = 0.06). Weed management strategy effects on canopy cover were significant st arting from July (2 nd sampling date), where imazapic and imazapic + 2,4 D were similar (20 and 19%, respectively) and greater than clethodim (7%), pendimethalin (4%), mowing (2%) and control (4%). Greatest canopy cover was achieved in August and was simil ar for imazapic and imazapic + 2,4 D (27 and 34%, respectively) and greater than the other treatments. Cover did not change from July through the remainder of the growing season for treatments which did not receive imazapic, while those receiving imazapic increased (Figure 5 1). Toward the end of the growing season, canopy cover in i mazapic was 9% lower than Imazapic + 2,4 D; although, both treatments remained greater than the others. Broadleaf weeds present in the strips planted to RP were mainly Mexican tea ( Chenopodium ambrosioides L.) and c utleaf g round cherry ( Physalis angulata L.), and they were most prevalent in the control, pendimethalin, mowing, and clethodim treatments. Also, there was a pronounced shift in weed population pressure to sedges ( Cype rus spp. ) after application of clethodim as opposed to bahiagrass and broadleaf weeds in the control, pendimethalin and mowing treatments. Due to the N rate competition control strategy interaction ( P = 0.02), simple effects of N application on weed man agement were averaged only across sampling dates (Fig ure 5 2). Nitrogen fertilization increased RP canopy cover in imazapic (from

PAGE 104

104 13 to 21%; P = 0.04) and imazapic + 2,4 D treatments (from 15 to 26%; P = 0.01), but there was no effect of N on mowing and a trend ( P = 0.07) toward reduced canopy cover in control, pendimethalin, and clethodim with N application. The literature has reported contradictory conclusions when N is applied during establishment of RP. Negative effects on RP ground coverage, dry matt er production, and nodulation were reported by Adjei and Prine (1976). Consequently N application was not recommended when planting RP. It is likely, however, that the negative RP response to N was due to the very high N rates used (0, 168 and 336 kg ha 1 ) and also increased competition from weeds after N fertilization. Valentim (1987) reported little effect on nodule weight in RP after an application of 50 kg N ha 1 compared with greater negative effect when 100 kg N ha 1 was applied. Thomas (1994) repor ted similar results to those of Valentim (1987) when working with Arachis pintoi Krapov. & W.C. Greg., where levels greater than 100 kg N ha 1 inhibited nodulation when measured at 8 wk after planting. It has been suggested that 50 kg N ha 1 could be used as a starter dose without unduly affecting A. glabrata or pintoi infection and nodulation ( Valentim et al., 1986 ; Thomas 1994). Our results indicate that application of 50 kg N ha 1 had positive effects on RP canopy cover and frequency in the treatments w here competition from weeds was effectively controlled (imazapic and imazapic + 2,4 D; Figs. 1 and 2). Competition control using these two treatments was achieved by completely suppressing broadleaf weeds (combined action of imazapic and 2,4 D herbicides) and temporarily suppressing bahiagrass growth which allowed time for establishment of RP while preventing

PAGE 105

105 emergence of other weeds. Thus, by the end of the season, canopy closure was achieved by RP and the re growing bahiagrass. Frequency responses were si milar to RP canopy cover (Fig ure 5 1). Starting from July, the imazapic and imazapic + 2,4 D treatments were similar (46 and 49%, respectively) and greater than clethodim (34%), pendimethalin (24%), mowing (15%), and control (23%). Greatest frequency was a chieved toward the end of the growing season. Imazapic and imazapic + 2,4 D were similar (67 and 73%, respectively) and greater than the other treatments which remained below 35%. Nitrogen fertilization increased RP frequency in imazapic (from 43 to 57%; P = 0.05) and imazapic + 2,4 D treatments (from 45 to 61%; P = 0.02); there was no effect of N in the Mowing treatment, and there was a negative effect in control, pendimethalin, and clethodim treatments ( P = 0.04; P value corresponds to averaged control, p endimethalin and clethodim treatments with 0 vs. 50 kg N ha 1 yr 1 ). Rhizoma peanut canopy cover and frequency (~30 and 80%, respectively) were similar to the values reported when no defoliation or when production of hay were imposed during the year of es tablishment on strip planted RP without N application and treated with imazapic ( Chapter 3 ). When N was applied following a single application of imazapic or imazapic +2,4 D, canopy cover and frequency were ~41 and 80%, respectively. Thus, a single applica tion of imazapic or imazapic + 2,4 D, followed by an application of 50 kg N ha 1 yr 1 has the potential to improve establishment of RP planted in strips, provide N to the bahiagrass growing along the edges of the strips, and increase RP contribution to the planted strip by the end of the year of establishment.

PAGE 106

106 Light Environment Nitrogen application, weed management strategy, sampling date, and second order interactions were significant. Incident PAR at RP canopy height was similar at all sampling dates for imazapic and imazapic + 2,4 D treatments (above 96%) and was consistently greater than the other treatments until July when PAR for control was 77%, 81% for mowing, 75% for pendimethalin, and 82% for clethodim. Light environment decreased from July until September in all treatments except Mowing which remained constant at ~ 81%. In September, imazapic and imazapic + 2,4 D were similar to mowing (81, 84, and 84%, respectively), and these three were greater than control (68%), pendimethalin (57%), and cletho dim (70%) (Fig ure 5 3). Even though there was a decrease in incident PAR at RP canopy height after July for the imazapic and imazapic + 2,4 D treatments, the decline (~10%) did not preclude further increases in RP c anopy cover and frequency (Figure 5 1) an d appeared to be function of bahiagrass regrowth. Nitrogen application weed management strategy interaction was analyzed over collection dates. Application of N decreased the amount of incident PAR reaching the canopy of RP in all treatments except in i mazapic and imazapic + 2,4 D (Fig ure 5 4). Consequently, effects on light environment appear to be a critical factor affecting RP establishment response to N fertilization. The benefits of N application are apparent when competition from weeds is suppres sed for an extended period, as was the case in the imazapic and imazapic + 2,4 D treatments. In contrast, application of N can be deleterious for the establishment of RP planted in strips when other species are actively competing with RP for light.

PAGE 107

107 Canopy Height and Spread For RP canopy height there was effect of weed management strategy and a trend toward weed management strategy N application interaction ( P = 0.08). The interaction was due to greater (numerically) RP height with N application in contro l, pendimethalin, and clethodim (+5, +2, and +5 cm, respectively) and equal or lower height for the mowing, imazapic, and imazapic + 2,4 D treatments (0, 2, and 2 cm, respectively). Averaged across N applications, RP canopy height was greatest and simila r in the control and pendimethalin treatments (30 and 31 cm, respectively) followed by clethodim (26 cm), and lowest for the mowing, imazapic, and imazapic + 2,4 D treatments (10, 11, and 8 cm, respectively; Fig ure 5 5). The results indicate that in growin g environments where light is limiting, RP has the capacity to show phenotypic plasticity (stem elongation in this case) as a light capturing strategy. Further, there was no effect of applying 2,4 D amine herbicide on RP canopy height (imazapic vs. imazapi c + 2,4 D) during the year of establishment at the rate used in this study. There were no treatment effects on spread measurements during the year of establishment. Similar results were reported in Chapter 3 for non defoliated and hay production treatment s imposed during the year of establishment. Year after Establishment Measurements Canopy c over and f requency There were effects of weed management strategy and weed management strategy N application interaction for canopy cover. The interaction was due to greater (numerically) RP canopy cover with N fertilization in imazapic and imazapic + 2,4 D treatments (+6, and +9%, respectively) and equal or lower canopy cover with N fertilization in control, mowing, pendimethalin, and clethodim ( 5, 0, 3,and 5%,

PAGE 108

108 respectively). Due to the interaction effect the data were analyzed by N application rate. The effect of N application, even a year after application, was most apparent in imazapic and imazapic + 2,4 D treatments where competition from weeds was effective ly controlled (Table 5 1). Frequency followed the same trend as canopy cover. There were effects of weed management strategy and its interaction with N application ( P = 0.03). When no N was applied, RP frequency in imazapic and imazapic + 2,4 D (57% for both) treatments were similar to clethodim (45%) and greater than the others. Clethodim was similar to control and pendimethalin (37 and 35%, respectively). The lowest frequency occurred in the mowing treatment (13%). When N was applied during the year of establishment, imazapic and imazapic + 2,4 D had the greatest frequency (70 and 73%, respectively) compared to the other treatments which were all similar and lower than 25% (Table 5 1). Botanical c omposition There were plant competition control strategy effects on botanical composition but no interaction with N rate ( P = 0.42). Imazapic and imazapic + 2,4 D treatments (11% for both treatments) had greater RP than control (1%), mowing (1%), and pendimethalin (3%). Clethodim (5%) was intermediate and simi lar to all treatments (Table 5 1). Imazapic and imazapic + 2,4 D treatments were noteworthy for the presence of RP patches where growth was relatively decumbent and less upright. Thus a significant proportion of the RP that resulted in superior cover and f requency responses in these treatments remained lower than the 10 cm cutting height for the botanical composition measures. This conclusion is supported by the canopy height results from the year of

PAGE 109

109 establishment where RP height was lowest (Figure 5 5) in treatments where canopy cover and frequency were highest (Fig ure 5 1). Spread There were no treatment effects during the year after establishment. Thus, given that there were no treatment e ffects during either year (year of and year after establishment), spread was analyzed combining the 2 yr data of the corresponding set of plots. Year was included in the statistical model as a fixed effect to estimate RP spread into the bahiagrass over time. Results from 2011 indicated that on average RP spread 36 cm pe r year (year effect; P = 0.006). Implications of the Research Competition control strategy effects on canopy cover were significant starting from July. Greatest canopy cover was achieved in August and was similar for imazapic and imazapic + 2,4 D (27 and 34%, respectively) and greater than clethodim (7%), pendimethalin (4%), control (4%), and mowing (2%) treatments. Application of 50 kg N ha 1 yr 1 increased RP canopy cover in imazapic (from 13 to 21%; P = 0.04) and imazapic + 2,4 D treatments (from 15 to 26%; P = 0.01); there was no effect of N application on mowing, and a trend ( P = 0.07) toward reduced canopy cover in control, pendimethalin, and clethodim treatments when N was applied. Frequency values followed the same trend as canopy cover. Greatest frequency was observed toward the end of the growing season. Imazapic and imazapic + 2,4 D were similar (67 and 73%, respectively) and greater than the other treatments which remained below 35%. Nitrogen fertilization increased RP frequency in imazapic (fr om 43 to 57%; P = 0.05) and imazapic + 2,4 D treatment (from 45 to 61%; P = 0.02); there was no effect of N application for mowing and a negative effect of N for control,

PAGE 110

110 pendimethalin, and clethodim treatments ( P = 0.04). Incident PAR at RP canopy height was similar in all sampling dates for imazapic and imazapic + 2,4 D treatments (above 96%) and was consistently greater than the other treatments until July (control, 77%; mowing, 81%; pendimethalin, 75%; and clethodim, 82%). Year after establishment measu rements indicate that treatment effects in the establishment year carried over. Canopy cover, frequency and botanical composition measurements were consistently greatest in the imazapic and imazapic + 2,4 D treatments. Spread measurements indicate that RP grew into bahiagrass sod at a rate of 36 cm yr 1 In conclusion, the benefits of N application accrue when weed competition is suppressed, as in the case of i mazapic and i mazapic + 2,4 D treatments. In contrast, N application can be deleterious if competi tion is not controlled. Single application of imazapic and imazapic + 2,D 4 herbicides followed by application of 50 kg N ha 1 yr 1 has the potential to improve establishment of RP planted in strips, increase bahiagrass dry matter harvested in the unplante d areas cut for hay, and aid in achieving a RP bahiagrass mixture in the planted strip by the end of the establishment year.

PAGE 111

111 Table 5 1. Rhizoma peanut (RP) canopy cover, frequency, and botanical composition in July of the year after establishment for RP s trips planted in bahiagrass pastures and subjected to various weed management strategies with and without N fertilizer. Data are means across three replicates and 2 yr (n = 6). Treatment Canopy cover Frequency Botanical composition description kg N ha 1 yr 1 kg N ha 1 yr 1 0 50 0 50 -------------------------------------% ----------------------------------------Mowing 2 c 2 b 13 c 15 b 1 b Pendimethalin 6 bc 3 b 35 b 23 b 3 b Control 8 ab 3 b 37 b 19 b 1 b Clethodim 10 ab 5 b 45 ab 25 b 5 ab Imazapic 12 ab 18 a 57 a 69 a 11 a Imazapic + 2,4 D 16 a 25 a 58 a 73 a 11 a SE 2 2 9 9 2 Mowing: every 28 d to 10 cm stubble height simulating bahiagrass hay production; control: untreated, un defoliated strip planted to RP with only adjacent bahiagrass defoliated; and the application of herbicides; pendimethalin (Prowl) at a rate of 0.93 kg a.i. ha 1 at planting; clethodim (Select Max) at a rate of 0.10 kg a.i. ha 1 applied when grass weeds were 10 to 15 cm tall); imazapic (Impose) at a rate of 0.07 kg a.i. ha 1 when grass or broadleaf weeds were 5 to 10 cm tall; and imazapic (0.07 kg a.i. ha 1 ) mixed with 2,4 D amine at a rate of 0.28 kg a.i. ha 1 when grass or broadleaves were 5 to 10 cm tall. Numbers within columns not followed by the same letter are different ( P < 0.05).

PAGE 112

112 Fig ure 5 1. Canopy cover and frequency of occurrence of rhizoma peanut planted in strips in existing bahiagrass pastures. Data are means across 2 yr. Error bars represent treatment means averaged across nitrogen rates (n = 6) 1 standard error.

PAGE 113

113 Figure 5 2. Canopy cover and frequency of occurrence of rhizoma peanut planted in strips in existing bahiagrass pastures. Data are means across 2 yr. E rror bars represent treatment means averaged across sampling dates and years (n=24) 1 standard error. = significant N effect ( P

PAGE 114

114 Fig ure 5 3. Incident photosynthetically active radiation (PAR) at rhizoma peanut canopy height. Data are me ans across 2 yr. Error bars represent treatment means averaged across N rates (n = 6) 1 standard error. //

PAGE 115

115 Figure 5 4. Incident photosynthetically active radiation (PAR) reaching the rhizoma peanut (RP) canopy. Data are means across 2 yr. Error bars r epresent treatment means averaged across sampling dates and years (n = 30) 1 standard error. = significant N effect ( P

PAGE 116

116 Fig ure 5 5. Rhizoma peanut canopy height measured at the end of the growing season in 2010 and 2011. Data are means of 2 yr. Error bars represent treatment means averaged across years (n = 6) 1 standard error.

PAGE 117

117 CHAPTER 6 SEEDBED PREPARATION TECHNIQUES AND WEED MANAGEMENT STRATEGIES FOR STRIP PLANTING A LEGUME INTO WARM SEASON GRASS PASTURES Overview of Research Problem Rhizoma peanut is a warm season, vegetatively propagated, perennial legume with potential for incorporation into low input, pasture based livestock systems in the USA Gulf Coast Region (French et al., 1994; Castillo, Chapter 3). Drought tolerance (French, 1998), dry matter yields up to 12 Mg ha 1 under natural rainfall conditions (Beltranena et al., 1981; Ocumpaugh, 1990), sim ilar crude protein concentration and digestibility to alfalfa ( Medicago sativa L.) (Prine et al., 1981; Beltranena et al., 1981), and persistence under a wide range of management systems for hay, silage, grazing, and as a understory forage crop (Prine et a l., 1981; Ortega S. et al., 1992; Johnson et USA Gulf Coast region has demonstrated the versatility of RP. Currently RP is used primarily as a high value hay crop for horses ( Equus caballus ) and dairy cattle ( Bos taurus ) in the Southeast USA. Factors limiting use of RP in grazed pasture systems include high cost and slow rate of establishment and the need to remove the planted area from the grazing rotation during the e stablishment period. If RP is to become more widely used in low input pasture systems, lower cost establishment methods will be needed. A technology that may have potential to address this need is strip planting RP into existing perennial grass pastures. P revious research evaluated and demonstrated potential of strip planting RP in existing bahiagrass pastures (Chapters 3 and 5 ). This work showed that grazing during the establishment year negatively affects RP cover, frequency, and spread compared to hay pr oduction which is similar to no defoliation of the planted strip (Chapter 3). It also

PAGE 118

118 indicated that imazapic or i mazapic + 2,4 D can be used effectively to control post emergence weed competition during the establishment year and that application of 50 kg N ha 1 after herbicide treatment increases RP cover and frequency in those herbicide treatments that are effective in controlling broadleaf weeds and sedges ( Cyperus sp.) and stunting bahiagrass (Chapter 5 ). Currently, there is little information availab le regarding seedbed preparation prior to planting RP and its effect on establishment response. Williams et al. (2002) found that planter type (no till vs. conventional sprig planter) had no effect on rhizoma peanut establishment. They also investigated th e factorial combination of seedbed preparation (undisturbed sod vs. rotovated), planting date (winter vs. summer), and use of the herbicide glyphosate vs. none. The authors analyzed the data by year and attributed a large portion of the variation in respon ses to environmental conditions (rainfall and temperature). In general, they reported a trend toward greater sprout emergence when soil was disturbed (i.e., rotovated, tilled) compared to sod seeding. Williams (1993) eval uated pre plant tillage (plowed = bottom plowed a nd disked; disked = disked only; and sod = planted directly into grass sod). She reported a general ranking for RP sprout emergence of plowed > disked = sod and recommended planting RP in well prepared fields during winter. While these studi es provide some information on the effects of seedbed preparation prior to planting RP, there was significant variability in treatment responses depending on weather and none of the experiments studied what is now the most likely approach to be used by pro ducers, the combination of herbicides and tillage.

PAGE 119

119 Herbicides, types of implements, and number of passes with equipment over the strips required to achieve an adequate seedbed for RP establishment are critical determinants of establishment cost. These req uirements will affect whether a low cost option can be realized that may result in adoption of this technology by farmers, especially by those using RP in grazed pastures, e.g., pasture managers in the cow calf industry. The current experiment was designe d to investigate what combinations of pre plant seedbed preparation and post plant competition control strategies are most objectives were to quantify the effect of seedb ed preparation techniques on establishment of RP planted in strips in bahiagrass sod and to determine the effect of post plant, competition control strategies and their interaction with seedbed preparation. M aterials and Methods Experimental Site The exper iment was conducted during 2011 and 2012 at the University of Florida sprout emergence, only data from 2011 will be presented in this chapter. The site was chosen because of available well bahiagrass pastures and because nearby RP pastures at this site have persisted for 30 yr, indicative of adaptation of RP to the area. The soils at the experimental site were classified as Pomona (sandy, siliceous, hyperthermic Ultic Alaquods) and Myakka sands (s andy, siliceous, hyperthermic Aeric Haplaquods) Initial characterization of the surface soil (0 to 15 cm) indicated soil pH of 6.0 and Mehlich 1 extractable P, K, Ca, and Mg of 8, 16, 161, and 25 mg kg 1 respectively. Based on the recommendation for

PAGE 120

120 growing RP, the area was fertilized at the beginning of the growing season with 16 kg ha 1 of P from triple super phosphate (440 g P 2 O 5 kg 1 ), and 60 kg ha 1 of K from muriate of potash (KCl, 600 g K 2 O kg 1 and 500 g Cl kg 1 ). Rainfall data from the experimental period and the 30 yr average are presented in Fig ure 6 1 Total rainfall during 2011 was 1029 mm. Last freeze event before planting in spring occurred on 8 March 2011. First fre eze event at the end of the growing season occurred on 10 Nov. 2011. The dates of freeze events were typical for this location. Planting Methodology The strips into which RP was planted were 4 m wide bounded on each side by a 1 m wide strip of undisturbed bahiagrass sod. Strips were planted using a conventional three row Bermuda King sprig planter (Williams et al., 2002) with a spacing of 0.5 m between rows. Each strip accommodated a total of nine rows of RP. The first and last rows of planted rhizomes were 0.25 m away from the undisturbed edge of bahiagrass sod, and the three external rows on each side of the strip were planted first. The center three rows were planted last such that spacing was 0.25 m between the second and third rows on either side of the center row. Florigraze RP rhizomes were planted at a rate of 1000 kg ha 1 ( packed at ~ 79 kg m 3 ) to approximately 5 cm depth on 5 April 2011. After planting, the plots were cultipacked to firm the seedbed and ensure adequate soil rhizome contact. Irriga tion was applied during April and May such that weekly rainfall plus irrigation equaled the 30 yr average weekly rainfall (18 and 20 mm per week in April and May, respectively). Once the Mowing treatment was initiated in June no further irrigation was prov ided.

PAGE 121

121 Treatments and Experimental Design Treatments were the factorial combinations of four seedbed preparation techniques and four weed competition control strategies, for a total of 16. Treatments were allocated in a split plot arrangement of a randomize d complete block design and were replicated three times. The main plot factor was seedbed preparation technique, and the sub plot factor was competition control strategy. The area of an experimental unit was 3 m long and 6 m wide, with a 1 m border between the lengths of the plots. Seedbed preparation techniques were: 1) glyphosate ( 6.2 kg a.i. ha 1 ) followed by conventional tillage {G CT; bahiagrass sod sprayed with glyphosate [N (phosphonomethyl) glycine, in the form of its potassium salt] in October 20 10 followed by deep tillage with a moldboard plow and heavy disking during February 2011}; 2) conventional tillage (CT; bahiagrass sod tilled as in G CT treatment but no glyphosate applied to kill the bahiagrass); 3) no till (NT, bahiagrass sod sprayed wit h glyphosate in October 2010, followed by mowing remaining above ground biomass to 5 cm stubble height before planting RP); 4) sod lifted (SL; bahiagrass sod was lifted with a sod cutter to a depth of 8 cm below soil level and removed from the strip before planting RP). Seedbed preparation treatments were chosen because they represent available and commercially practical options for addressing bahiagrass competition to establishing RP and because they have potential to create a wide range in disturbance of the bahiagrass sod. Weed competition control strategies applied to the strips planted to RP were: 1) single application of i mazapic { i mazapic; (+/ 2 [4,5 dihydro 4 methyl 4 (1 methylethyl) 5 oxo 1 H imidazol 2 yl] 5 methyl 3 pyridinecarboxylicacid; Impos e} at a rate of 0.07 kg a.i. ha 1 when grass or broadleaf weeds were 5 to 10 cm tall, 2) i mazapic ( 0.07 kg a.i.

PAGE 122

122 ha 1 ) mixed with 2,4 D amine (dimethylamine salt of 2,4 dichlorophenoxyacetic acid ; 2,4 D amine Weed Killer ) ( i mazapic + 2,4 D) at a rate of 0. 28 kg a.i. ha 1 when grass or broadleaf weeds were 5 to 10 cm tall, 3) mowing (Mow; every 28 d to 10 cm stubble height simulating a bahiagrass hay production treatment), and 4) control (Control; no herbicide application, non defoliated). Herbicide s were a pplied on 14 June 2011 using a CO 2 pressurized backpack sprayer calibrated to deliver 187 L ha 1 The strips were sprayed using a 3.04 m wide boom so that the bahiagrass at the edges of the strips was not sprayed. Weed management strategies were based on t he results of studies conducted first in 2010 (Chapters 3 and 5 ). The Mow treatment was applied starting ~ 11 wk after planting (28 June 2011), coinciding with the anticipated end of the sprout emergence period, and every 28 d thereafter throughout the gr owing season. Timing of initiation was based on data reported by Williams (1993) and Williams et al. (1997), who indicated that sprout emergence continued for 7 wk after first sprouts emerged. In G CT, CT, and SL treatments, the planted strip was not defol iated during the growing season, but the bahiagrass bordering the planted strip was mowed to 10 cm stubble every 28 d. T his occurred at the same time the entire plot of the Mow treatment was clipped. Herbicide Clethodim ((E) 2 2[1 [[3 chloro 2 propenyl)oxy ]imino]propyl]5 [2(ethylthio)propyl] 3 hydroxy 2 cyclohexen 1 one; Select Max) was spot sprayed to control common bermudagrass [ Cynodon dactylon (L.) Pers.] growing in the strips planted to RP. Response Variables Sprout Emergence Sprout counts were measur ed with the objective of determining the effect of seedbed preparation techniques. Sub plot treatments had not yet been imposed during

PAGE 123

123 the sprout emergence period, thus only main plot effects were quantified. Sprout counts began at sprout emergence and occ urred every 2 wk through 6 wk after emergence. First sprout emergence occurred 4 wk after planting on 28 April and May 17 in 2011 and 2012, respectively. Sprout emergence was determined by counting the number of sprouts within three randomly located, perma nently marked 20 by 50 cm quadrats per main plot, so that evaluations were done in the same places over time. The 50 cm side of the quadrat was always placed parallel to the RP rows with the 20 cm side centered perpendicular to a row of RP. Total sprout e mergence in a plot was calculated as the average of the three quadrats per plot and is expressed as sprouts m 2 Canopy Cover and Frequency Rhizoma peanut canopy cover in the planted strip was measured visually every 28 d for all treatments on the day af ter each defoliation event of the Mow treatment. A 1 m 2 quadrat (0.5 2 m) was placed in the center of the RP strip at a permanently marked location in each experimental unit, so that canopy cover was estimated on the same area over time. The 0.5 m side of the quadrat was oriented parallel to the RP rows and was placed 1 m away from the edge of the bahiagrass sod. Thus, the area enclosed by the quadrat included a total of six rows of RP. The quadrat was divided into 100, 10 by 10 cm squares (five rows of 20), and canopy cover was estimated visually by the same observer in 20 stratified 10 by 10 cm squares (four squares in each row of 20) per quadrat and averaged to obtain an overall cover per experimental unit ( Appendix A; Interrante et al., 2009) Frequency of occurrence is a measurement of the relative distrib ution of RP in the strip. It was determined on the same dates and at the same quadrat locations that were used to estimate RP canopy cover. Presence or absence of RP was determined in 20

PAGE 124

124 stratified 10 by 10 cm squares. Frequency was expressed as a percent age and was calculated as the number of cells where RP was present divided by the total number of cells assessed with the quotient multiplied by 100. Light Environment Ambient light environment at the top of the RP canopy was measured 2 wk before the pro jected end of the shoot emergence phase (Williams, 1993; Williams et al., 1997) and every 28 d thereafter. Measurements on all experimental units were taken between 1200 and 1500 h Eastern Daylight Time on Day 14 of each of the 28 d regrowth periods of the Mow treatment. Light environment was characterized using a SunScan Canopy Analysis System (Dynamax Inc., Houston, TX) at three randomly selected locations per plot. The SunScan consists of a 1 m long quantum sensor that was placed at the height of the RP canopy to measure transmitted photosynthetically active radiation (PAR), and an unshaded beam fraction sensor that was placed outside the plots to measure incident PAR. Thus, the light environment experienced by RP plants was characterized as percent of i ncident PAR that reached the RP canopy and was calculated by dividing the transmitted PAR by incident PAR and multiplying by 100 to express it as a percentage. The average of four observations per experimental unit provided an estimate of light environment Spread and Canopy Height Rhizoma peanut spread and canopy height were measured on all plots the day before the last clipping event of the season (17 September) for the Mow treatment. To measure spread, a transect was positioned through the center of the RP strip running the length of the each plot. At the 1 and 2 m points along the 3 m transect, a line, perpendicular to the transect, was extended on each side. Spread was defined as the

PAGE 125

125 distance from the center of the planted RP strip to the farthest poin t where identifiable RP plant parts (above ground) were found. Canopy height was intended to describe canopy development and interaction with treatments. It was measured using a ruler to quantify the distance from the soil surface to the non extended heig ht of the RP canopy. Four measurements per plot were averaged to provide estimates of spread and canopy height for each experimental unit. Statistical Analysis Data were analyzed as repeated measures using PROC GLIMMIX of SAS (SAS Institute, 2010). Samplin g date was considered a repeated measurement with a first order autoregressive covariance structure. Block was considered a random effect. Treatments were fixed effects. Mean separations were based on the SLICE and SLICEDIFF procedures of LSMEANS Plots o f model residuals were used to check normality, and in the case of non normal distributions, data transformations were used. Square root transformation was used for canopy cover, and log based 10 was used for sprout counts. Treatments were considered diffe rent when P R esults and Discussion Sprout Emergence There were seedbed preparation and collection date effects. Number of sprouts m 2 averaged across collection dates, was greatest and not different for G CT (119) and CT (90), while CT was not different tha n NT (58) but greater than SL (54), and NT was not different than SL (Fig ure 6 2 ). Sprout emergence averaged 80 m 2 at Week 4 after emergence and did not increase significantly through Week 6 (88 sprouts m 2 ). Our results agree with previous reports in the literature indicating that the majority of RP sprouts emerged within 6 to 7 wk of initial emergence regardless of pre plant tillage and

PAGE 126

126 planting date (Williams, 1993; Williams et al., 1997). Those authors also reported a trend toward greater sprout emerge nce when soil was disturbed (i.e., rotovated, tilled) compared to sod seeding. Fewer sprouts emerging when RP was planted in existing sod w as attributed to factors including desiccation of rhizomes when appropriate soil rhizome contact was not achieved an d to inadequate soil moisture (Williams, 1993). That study was conducted under non irrigated conditions and treatment responses varied depending on rainfall. Under the conditions of the current experiment, lack of water may not have been a limiting factor because irrigation was provided during the sprout emergence period. Further, there was no active bahiagrass growth in the NT treatment strips because of glyphosate application the previous fall. It is not completely clear how the presence of decaying plan t material (above and below ground) affected sprout emergence in the NT treatment. Also, it is not totally clear why there was low sprout emergence in the SL treatment where there was virtually no competition from above or below ground plant parts. Howe ver, the similarity of the SL and NT responses and the superior performance of G CT and trend toward superior performance of CT suggest that tilled soil allows for more favorable rhizome placement during planting, perhaps achieving greater depth or better soil rhizome contact. Nevertheless, differences in number of emerged sprouts did not result in differences in RP canopy cover and frequency of occurrence at 9 mo after planting. Canopy Cover and Frequency There were weed management strategy, sampling date and weed management strategy sampling date interaction effects. Seedbed preparation did not have an effect ( P = 0.67) on RP canopy cover nor did it interact with competition control strategy ( P =

PAGE 127

127 0.31). Averaged across subplot treatments, RP canopy cov er and frequency were greatest toward the latter half of the growing season with canopy cover values of 25, 25, 19, and 21% and frequency of occurrence of 59, 58, 52, and 41% for G CT, NT, CT, and SL treatments, respectively. These canopy cover and frequen cy data are similar to those reported for strip planted RP when seedbed preparation was the same as that used for CT in this experiment ( Chapter 3). The results further suggest that a single application of glyphosate herbicide in the autumn followed by mow ing to remove some of the above ground biomass before planting (NT) is a viable seedbed preparation option for strip planting RP in existing bahiagrass pastures. Differences in RP cover due to post plant competition control strategy occurred early in the g rowing season (July) and cover generally peaked in August (Fig ure 6 3 ). Treatment responses were consistent across the season and segr egated in treatment pairs with imazapic (33%) and i mazapic + 2,4 D (35%) being not different from each other and greater t han Control (16%) and Mow (9%). The superior performance of these herbicide treatments has been consistent in several studies evaluat ing RP strip planting ( Chapters 3 and 5 ). Frequency followed the same trend as canopy cover. There were weed management st rategy, sampling date, and weed management strategy sampling date interaction effects. Weed management strategy effect was significant starting from August (Fig ure 6 3), and the treatments segregated in pairs and remained consistent until the end of the growi ng season. Imazapic (66%) and i mazapic + 2,4 D (70%) were not different from each other and greater than Mow (35%) and Control (40%). Similar results were reported by Castillo (Chapter 5 ).

PAGE 128

128 It is important to note that the cover and frequency data repo rted in this chapter are from 2011 only, with data from 2012 not yet included in the analysis. It is apparent that regardless of the seedbed preparation strategy used, post emergence application of i mazapic or i mazapic + 2,4 D are needed to control competi tion from weeds or bahiagrass and allow successful establishment of RP planted in strips. Light Environment There was trend toward a seedbed preparation effect ( P = 0.07) and there were effects of weed competition control strategy, sampling date, and all s econd order interactions. Seedbed preparation effect on light environment of RP canopy was significant starting in September with PAR in NT and SL being above 90% and not different but greater than CT and G CT (83 and 76%, respectively) (Fig ure 6 4 ). Diffe rences followed the same trend until the end of the growing season in October. There were competition control strategy effects starting in August with PAR in Mow (96%), i mazapic, and i mazapic + 2,4 D (100% for both) being not different from each other and all of them greater than Control (86%). The response remained the same through the end of the growing season in October. The decrease in P AR at RP canopy height for the i mazapic, and i mazapic + 2,4 D treatments during the latter part of the growing season (~ 12%; P < 0.05) was also reported by Castillo (Chapter 5 ), and it was attributed to regrowth of bahiagrass. Due to seedbed preparation and competition control strategy interaction, data were analyzed by seedbed preparation (Fig ure 6 5). There were no di fferences in light environment due to weed competition control strategy in NT and SL treatments. In G CT, light environment was similar and greatest for the i mazapic and i mazapic + 2,4 D treatments (94 and 90%, respectively), followed by Mow (82%), and low est for the

PAGE 129

129 Control (74%). In the CT treatment, PA R at RP canopy height for Mow, i mazapic, and i mazapic + 2,4 D was not different (88, 92, and 91%, respectively) and greater than the Control (73%). Results indicate that a single application of glyphosate i n fall followed by no tillage (Treatment NT) provides adequate light environment for establishment of RP plication of herbicides (i.e., i mazapic and i mazapic + 2,4 D) does not increase PAR to the RP canopy. Spread and Canopy Height There was seedbed preparation effect on RP spread (Fig ure 6 6). Spread was not different for G CT and CT (189 cm) and NT (178 cm), while SL (169 cm) was not different than NT but lower than G CT and CT. Given that the outer row of RP was planted 175 cm fr om the center of the strip, spread was minimal in the first year in all treatments. Similar results were reported by Castillo (Chapters 3 and 5 ). The SL treatment actually resulted in loss of plants in the outer row of the strip closest to bahiagrass resul ting in a reduction in spread. A possible explanation for this response is that the difference in the level of the planted strip and the adjacent bahiagrass sod, due to removal of sod and associated topsoil, may have prevented uniform cultipacking and ther efore reduced soil rhizome contact of the outer most rows. There were seedbed preparation technique, competition control strategy and interaction effects on RP canopy height. Interaction occurred because there were no competition control strategy effects for NT and SL seedbed preparation, while for G CT and CT there were differences. Canopy height of RP averaged across competition control strategies in the NT and SL treatments was 5 and 6 cm, respectively. For G CT, canopy height was greatest in the Contr ol (17 cm), followed by Mow (8 cm), which was not different than i mazapic + 2,4 D (7 cm), but was greater than i mazapic (4 cm). Height

PAGE 130

130 of RP was not different in imazapic and i mazapic + 2,4 D. For CT, RP height was greatest for the Control (16 cm), followe d by Mow (8 cm), which was greater than imazapic (4 cm) and i mazapic + 2,4 D (4 cm) (Fig ure 6 7). Canopy height of RP in this study followed a similar pattern as that described in Chapter 5 When RP grows in light limited environments (Control treatment in G CT and CT treatments), RP has the capacity to elongate stems (phenotypic plasticity) as a light capturing strategy. Implications of the Research Number of sprouts m 2 averaged across collection dates, was greatest and not different for G CT (119) and CT (90), while CT was not different than NT (58) but greater than SL (54), and NT was not different than SL (Fig ure 6 1). Apparent advantage of prepared seedbeds may be due to superior rhizome placement or soil rhizome contact. In spite of differences in s prout emergence, seedbed preparation technique had no effect on RP canopy cover (avg. of 23%) and frequency (avg. of 53%) and there was no interaction with competition control strategy. There were plant competition control strategy effects on RP canopy cov er and frequency, with i mazapic (33 and 66%) and i mazapic + 2,4 D (35 and 70%) being not different from each other and greater than Control (16 and 40%) and Mow (9 and 35%) for canopy cover and frequency, respectively. There was seedbed preparation comp etition control strategy interaction effect on light environment of RP canopy. Differences in light environment due to weed competition control strategy occurred only in G CT and CT treatments where the Control had the lowest PAR. Treatment effects on RP c anopy height followed the same trend and RP was tallest in the Control treatment for G CT and CT. Spread in the

PAGE 131

131 establishment year was minimal for all treatments but least for SL which actually experienced loss of plants on the edge of the planted strip. One year data indicate that a single application of glyphosate to control bahiagrass in the fall followed by winter mowing to break up residue before planting RP may be a viable seedbed preparation option for successful establishment of RP and may reduce establishment costs over strategies that include a completely prepared seedbed. Regardless of seedbed preparation strategy, application of i mazapic or i mazapic + 2,4 D should be used to control post plant competition from weeds or bahiagrass.

PAGE 132

132 Fig ure 6 1 Monthly rainfall at the University of Florida Beef Research Unit, Gainesville, FL for 2011 and the 30 yr average.

PAGE 133

133 Fig ure 6 2. Sprout emergence of rhizoma peanut planted in strips in existing bahiagrass plots. Data are means across 2 yr (201 1 and 2012). Error bars represent treatment means (n = 6) one standard error.

PAGE 134

134 Fig ure 6 3. Canopy cover and frequency of occurrence of rhizoma peanut planted in strips in existing bahiagrass pastures in 2011. Error s bars represent treatment means (n = 3) one standard error.

PAGE 135

135 Fig ure 6 4. Light environment at the top of the rhizoma peanut canopy for strip planted rhizoma peanut in existing bahiagrass pastures in 2011. Effects of seedbed preparation (above) and competition control strategy (below) and their interaction with sampling date are shown. Error bars represent treatment means (n = 3) one standard error. PAR = Photosynthetically active radiation. = significant interaction effe ct ( P interaction effect.

PAGE 136

136 Fig ure 6 5. Light environment at the top of the rhizoma peanut canopy for strip planted rhizoma peanut in existing bahiagrass pastures in 2011. Effects of seedbed preparation and competition control strategy interaction are shown. Error bars represent treatment means (n = 3) one standard error. PAR = Photosynthetically active radiation. NS NS

PAGE 137

137 Fig ure 6 6. Spread of rhizoma peanut (RP) into existing bah iagrass sod following planting in 2011. Spread is the distance from the center of the planted RP strip to the farthest point where identifiable RP plant parts (above ground) were found. Close st RP row was planted at 175 cm from the center of the strip. Err or bars represent treatment means (n = 3) one standard error. a a ab b

PAGE 138

138 Figure 6 7 Rhizoma peanut canopy height measured on c e at the end of the growing season in 2011. Effects of seedbed preparation and competition control strategy interaction are shown. Error bars represent treatment means (n = 3) 1 standard error. = significant interaction effect ( P significant interaction effect. NS NS

PAGE 139

139 CHAPTER 7 SUMMARY AND CONCLUSI ONS Rhizoma peanut ( Arachis glabrata Benth.; RP) is a tropical/subtropical, vegetatively propagated, perennial legume with demonstrated persistence and potential to pr ovide high quality forage for livestock across the Gulf Coast, USA. It has nutritive value comparable to alfalfa ( Medicago sativa L.), and is persistent under a wide range of management systems including production of hay, grazing, and as an understory for age crop. Thus, RP is a top candidate among forage legumes to develop grass legume mixtures that will increase/maintain production and sustainability of the typically low input monoculture bahiagrass ( Paspalum notatum Flgg) pastures used in the forage li vestock systems of the lower southeastern USA. Up to this time however, the main use of RP has been for the production of hay due to the high costs associated with establishment, management of weeds, and taking land out of prod uction for ~2 yr to allow es tablishment of RP. Rhizoma peanut has rarely been planted with grasses; grass RP mixtures exi s t due to inadequate control of grass weeds growing in the swards that initially were intended for the production of high quality hay. If RP is to contribute to fo rage livestock systems under grazing, it is critical to develop management s trategies that minimize costs associated with establishment and allow for utilization of the forage produced during the establishment period of RP An alternative approach to RP e stablishment is to plant RP in strips in existing bahiagrass pastur es. This approach may minimize inputs compared to intensive preparation of a seedbed prior to planting a mixed pasture. Using this approach, RP is physically separated from the actively gro wing grass component of the pasture. P hysical separation provides opportunities to investigate seedbed preparation

PAGE 140

140 techniques and weed management strategies that may be successfully used for RP but would not be appropriate for an associated grass It will take a pe riod of time for RP to spread fro m the planted strip to surrounding areas, but if this can be achieve d it may provide a relatively low cost option for the establishment and maintenance of mixed grass legume pastures. Experiments were conducted to assess the merits of strip planting of RP in existing bahiagrass pasture. Rhizoma peanut canopy cover, frequency of occurrence, light environment, botanical composition, canopy height, and spread were measured to determine relative success of RP establ ishment in four studies designed to: 1) evaluate the effect of defoliation management strategies during the year of establishment (Chapter 3); 2) investigate grazing management strategies during the year af ter establishment on plots that started with varyi ng levels of RP contribution (Chapter 4); 3) determine the effect of chemical and mechanical weed management strategies and N fertilizer during the establishment year (Chapter 5); and 4) quantify the effects of seedbed preparation and post plant weed manag ement strategies (Chapter 6). The studies were conducted at t he Beef Research Unit (BRU) of the University of Florida from some measurements taken in 2012 are not yet available, so Chapters 4 an d 6 contain 1 yr of data while Chapters 3 and 5 contain 2 yr. All the experiments utilized a different area, i.e., a different planting of RP, for each year of study. The experiment reported in Chapter 4 (grazing management effects in the yea r after establishment) used the plots that had been planted and evaluated the previous year for the work reported in Chapter 3.

PAGE 141

141 The strips planted with RP were 4 m wide and accommodated eight or nine rows of RP, with spacing between rows of 0.5 m. The fir st and last rows of planted rhizomes were 0.25 m from the undisturbed edge of bahiagrass sod. The planted strips were bounded on both sides by a 2.5 m (studies in Chapter s 3, 4, and 5) or a 1 m (Chapter 6) RP rhizomes were planted in the prepared strip using a conventional Bermuda King sprig planter in March or April each year. The planting material was obtained from a commercial farmer cooperator. The rhizomes were planted at a rate of 1000 kg ha 1 (packed at ~ 79 kg m 3 ) to approximately a 5 cm depth. After planting, the plots were cultipacked to ensure adequate soil rhizome contact. Effects of Defoliation in the Establishment Year The study in Chapter 3 was designed to evaluate options for utilization of the RP bahiagrass pasture during the year of establishment. The treatments were: 1) control (no defoliation of the planted RP strip with adjacent bahiagrass harvested for hay production every 28 d during the growing season to a 10 cm stubble height); 2) ha y production (RP strip and adjacent bahiagrass both harvested for hay production every 28 d to a 10 cm stubble height); 3) simulated continuous stocking (pastures grazed weekly to a 15 cm bahiagrass stubble height); and 4) rotational stocking (pastures gra zed every 28 d to a 15 cm bahiagrass stubble height). A combination of herbicides clethodim and imazapic were applied to control weeds growing in the strip planted to RP. The area of an experimental unit was 9 m wide 15 m long, and each consisted of one 4 m wide strip of RP running the length of the plot and bounded on each side by strips of bahiagrass. Initiation of defoliation treatments was targeted for the end of the RP sprout emergence period which occurred ~11 wk after planting.

PAGE 142

142 Canopy cover and fr equency of RP were greatest during August. Hay production and control treatments were not different with 32 and 29% canopy cover, respectively, and 67% frequency for both, and were greater compared to 5 and 4% canopy cover for simulated c ontinuous and r ota tional s tocking, respectively and 21% frequency for both. Spread was lowest and there was a trend toward less herbage harvested in the simulated c ontinuous s tocking treatment compared to the others. Competition for light was not the driving factor affecti ng RP establishment u sing these defoliation treatments and the strip planting approach. Treatments with greatest light level reaching the RP canopy generally were those with lowest RP cover and frequency, indicating that the effect of defoliation of establ ishing RP was the principal factor driving cover and frequency responses. Measurements of canopy cover and frequency taken in June during the following growing season (year after establishment) revealed that differences in canopy cover and frequency observ ed in the establishment year carried over. The results indicate that grazing, either simulated continuous or rotational stocking every 28 d, reduced RP contribution during the year of establishment as a function of apparent animal preference for forage in the legume planted strips and the resultant reduction in stubble height well below that of the adjacent bahiagrass strips. It is not known at this point whether rest periods longer than 28 d between grazing events, or the use of other RP cultivars with di fferent growth habits (i.e., more prostrate) may be able to overcome the negative impact of grazing during the year of establishment. It is clear, however, that RP establishment is negatively affected by grazing during the year of establishment, as practic ed in this experiment, and defoliation by clipping seems to be the most feasible option for using the forage produced.

PAGE 143

143 Grazing Management in the Year after Establishment In 2011, plots that were used in 2010 in the study in Chapter 3 were divided into thre e sub plots of 9 m width and 5 m length. The 9 m width of the sub plot consisted of a 4 m wide strip planted to RP (in 2010) bounded on both sides by strips of bahiagrass sod. This provided a range of starting conditions of RP contribution to further evalu ate grazing management effects on RP establishment in Chapter 4. Treatments in Chapter 4 were the factorial combinations of four Year 1 (Y1) defoliation strategies (Chapter 3) and three Year 2 (Y2) grazing management treatments, for a total of 12 treatment s. Longer term as opposed to monthly changes in RP contribution was considered of importance; therefore, measurements were taken once at mid and late season. Year 2 g razing management treatments were: 1) simulated continuous stocking (SC; same as T reatmen t 3 from Y1); 2) rotational stocking 28 d (RS 28; equal to T reatment 4 from Y1); and 3) rotational stocking 42 d (RS 42; pastures stocked rotationally every 42 d). All sub plot treatments were grazed to 15 cm bahiagrass stubble height. There was no Y1 defo liation strategy (during the year of establishment) Y2 grazing management (during the year after establishment) effect on RP contribution. One year data indicate that defoliation management during Y1 remained a critical factor affecting RP contribution d uring Y2. Canopy cover and frequency of Y1 control and hay production were not different (13 and 15% for canopy cover, and 66 and 59% for frequency, respectively). Canopy cover was greater for hay production than either simulated continuous (5%) or rotatio nal stocking (4%). Frequency was greatest for control and hay production compared to both of the grazing treatments which had RP frequen

PAGE 144

144 Year 2 RP canopy cover and frequency were greatest for RS 42 with 12%, compared to 6% for both SC and RS 28 at midseason. Nevertheless, by late season RP canopy cover was not different among treatments 6, 6, and 4% for SC, RS 28, and RS 42. Fr of the herbage harvested at mid or late season. Botanical composition of RP was greater in plots that were managed for hay production or rotationally stocked during the year of esta blishment. The results indicate that defoliation management by graziers should be focused on the strip planted to RP as opposed to the bahiagrass borders. If grazing is to occur during the year after establishment, each grazing event must be terminated soo ner than in the current study, likely when RP forage in the strip is ~15 to 20 cm tall. Weed Management Strategies in the Year of Establishment In Chapter 5 the objectives were to determine the effect of chemical and mechanical weed management strategies, N fertilizer and its interaction on establishment of RP planted in strips. Treatments were the factorial combinations of two N rates (0 and 50 kg ha 1 yr 1 ) and six weed management strategies; 1) control (no herbicides, no mowing), 2) mowing (every 28 d to 10 cm stubble height); and a single application of herbicides 3) pendimethalin ( 0.93 kg a.i. ha 2 ), 4) clethodim ( 0.10 kg a.i. ha 2 ), 5) imazapic ( 0.07 kg a.i. ha 2 ), or 6) imazapic ( 0.07 kg a.i. ha 2 ) + 2,4 D amine ( 0.28 kg a.i. ha 2 ).The herbicides an d rates of application were based on previous research, and these specific herbicides were chosen because they are the only ones labeled for use in RP pastures in Florida, with the exception of pendimethalin which was included as a treatment because of its use as a pre plant incorporated herbicide in plantings of annual peanuts ( Arachis hypogea L.).

PAGE 145

145 Rhizoma peanut canopy cover and frequency were greatest toward the middle of the growing season in August. Imazapic and imazapic + 2,4 D treatments were not d ifferent for canopy cover (27 and 34%, respectively) and frequency (67 and 73%, and 35% for cover and frequency, respectively. Application of 50 kg N ha 1 yr 1 following h erbicide application, increased RP canopy cover (+10%) and frequency (+15%) in imazapic and imazapic +2,4 D treatments. In contrast, N application had a negative effect on the remaining treatments because it increased competition to RP from grass and broad leaf weeds growing in the strip that were not effectively controlled by those herbicides or by mowing. Treatment effects carried over to the year after establishment. The RP canopy in imazapic and imazapic + 2,4 D treatments cons istently received above 96% of incident photosynthetically active radiation (PAR) until July, compared to growing season, incident PAR was not different for imazapic, imazapic + 2,4 D and mowing treatments (~83% ) but it was still greater compared to control (68%), pendimethalin (57%) and clethodim (70%). The decrease in PAR at the level of RP in the canopy for the i mazapic, and imazapic + 2,4 D treatments during the latter part of the growing season (~12%; P < 0.05) was attributed to regrowth of bahiagrass. A pplication of N decreased the amount of incident PAR in all treatments except for imazapic and imazapic + 2,4 D Canopy height measurements of RP taken toward the end of season were greatest and not differe nt for control and pendimethalin treatments (30 and 31 cm, respectively), followed by clethodim (26 cm), and lowest for the mowing, imazapic, and imazapic + 2,4 D treatments (10, 11, and 8 cm, respectively). The results indicate that in

PAGE 146

146 growing environment s where light is limiting, RP has the capacity to show phenotypic plasticity (stem elongation in this case) as a light capturing strategy. There were no treatment effects on RP spread during the year of establishment. Spread measurements taken the year aft er establishment indicated that on average RP spread 36 cm yr 1 Botanical composition from the year after establishment indicated that RP contribution was greatest and not different for imazapic and imazapic + 2,4 D treatments (11% RP component for both), compared to the rest of treatments where RP contribution was 5%. Seedbed Preparation of the Planted Strip Based on the results from Chapter 5, the treatments in which greatest contribution of RP was measured (imazapic and imazapic + 2,4 D amine) were used in the study in Chapter 6 to investigate what combinations of pre plant seedbed preparation and post plant weed management strategies are most effective for RP establishment in strips in existing bahiagrass sod. The t reatments were the factorial combinations of four seedbed preparation techniques and four weed man agement strategies. Seedbed preparation techniques were : 1) glyphosate (6.2 kg a.i. ha 2 ) followed by conventional tillage (G CT; bahiagrass sod sprayed with glyphosate in October 2010 followed by deep tillage with a moldboard plow and heavy disking during February 2011); 2) conventional tillage (CT; bahiagrass sod tilled as in G CT treatment but no glyphosate applied to kill the bahiagrass); 3) no till (NT, bahiagrass sod sprayed with glyphosate in October 2010, followed by mowing the remaining above groun d biomass to 5 cm stubble height before planting RP); 4) sod lifted (SL; bahiagrass sod was lifted with a sod cutter to a depth of 8 cm below soil level and removed from the strip before planting RP). Weed management strategies were: 1) single application of imazapic (as in

PAGE 147

147 T reatment 5, Chapter 5); 2) imazapic mixed with 2,4 D amine (as in T reatment 6, Chapter 5); 3) mowing (every 28 d to 10 cm stubble height simulating bahiagrass hay production); and 4) control (no herbicide application, non defoliated). Number of sprouts m 2 was greatest and not different for G CT (119) and CT (90), while CT was not different than NT (58) but greater than SL (54), and NT was not different than SL. Apparent advantage of prepared seedbeds may have been due to superior rhizo me placement or soil rhizome contact. Nevertheless, in spite of differences in sprout emergence, seedbed preparation technique had no effect on RP canopy cover and frequency One year data indicate that a single application of glyphosate to control bahiag rass in the fall followed by winter mowing to break up residue before planting RP is a viable seedbed preparation option for successful establishment of RP and may reduce establishment costs over strategies that include a completely prepared seedbed. Regar dless of seedbed preparation strategy, application of i mazapic or i mazapic + 2,4 D should be used to control post plant competition from weeds or bahiagrass. Implications of the Research Our results indicate that when Florigraze RP is strip planted into ex isting bahiagrass pastures, utilization should be limited to production of hay during the year of establishment. If producers cannot afford to remove land from the grazing rotation or change to a hay production system during the year of establishment, then grazing management strategies should be developed that allow for resting periods between grazing events of longer than 28 d. In addition, decisions about when cattle should enter/exit the pasture should be based on the height of herbage in the strip plant ed to

PAGE 148

148 RP as opposed to the height of the grass component. Based on data from Ortega S. et al. (1992), it is suggested that RP not be grazed closer than a 15 cm stubble. Data from the seedbed preparation study and the grazing management in the year after es tablishment s tudy reflect only 1 yr of research Thus, the second year should be included in the analysis before strong conclusions are reached. The preliminary data from grazing management study does suggest, however, that close grazing of the planted str ip in the year after establishment is detrimental, even when grazing is infrequent (i.e., 42 d rest period). These experiments demonstrate sufficient success to support a conclusion that the technology of strip planting merits continued evaluation. Howeve r, they have identified a number of challenges that need to be addressed. Questions that remain to be answered include: 1) Would use of glyphosate to kill the grass sod in the fall before planting reduce the level of bahiagrass competition in the year of a nd year after establishment?; 2) Is grazing in the year of establishment an option if the rest period between grazing events is longer than the 28 d evaluated in this research and if livestock are removed based on stubble height of the planted strip instea d of the companion grass?; 3) Is grazing in the year after establishment an option if the rest height of the planted strip instead of the companion grass?; 4) How long d oes it take to achieve full RP cover in the strip when various defoliation treatments are imposed?; and 5) What is the expected rate of lateral spread of RP in subsequent years? Future Research Needs meaning so mewhere between upright (i.e., Arbrook) and low growing (i.e., germplasm

PAGE 149

149 Ecoturf) types. In recent years, there is an increased list of available RP cultivars and germplasms with different growth habits Th ese contrasting grow th habits and associated morphological characteristic s ( canopy height, shoot/ rhizome + root ratio, leaf area index) have the potential to respond differently under grazing and the strip planting establishment approach. Future research should incorporate a wider range of the av ailable RP material, including Prine (CP I 93483 in Australia and equivalent to PI 231318 in USA), which was selected for persistence and spread in pure stands and in strip planting as a multi year system should be evaluated on more nearly farm scale pastures with refined grazing strategies based on the current research and with additional grass species. These longer term experiments would allow: 1) the questions posed above to be addressed; 2) measurement of ecosystem services provided (i.e., N fixation, changes is soil structure, C sequestration); and 3) more accurate assessment of economic returns from targeted technologies. Should willing producers be identified, th ese studies could be conducted on farm with producer participation in deciding which treatment options should be compared.

PAGE 150

150 APPENDIX METHODOLOGY FOR CANOPY COVER AND FREQUENCY MEASUREMENTS Description The quadrat consists of 100, 10 by 10 cm squares (five rows of 20). The total area covered by the quadrat is 1 m 2 (0.5 2 m). The quadrat was placed in the center of the RP strip at permanently marked locations in each experimental unit, so that canopy cover and frequency were estimated on the same area s over time. The 0.5 m side (length) of the quadrat was oriented parallel to the rhizoma peanut (RP ; Arachis glabrata Benth. ) planted rows (Figure A 1). Thus, the area enclosed by the quadrat included four rows of RP with the ends of the quadrat positioned so that they rested midway between the outermost RP row that was included in the quadrat and the RP row that was located just outside the quadrat. The squares s hade d in black ( total of 20) in Figure A 1 correspond to th ose where RP canopy cover and freque ncy were determined. Canopy c over Canopy cover was estimated visually by the same observer in the 20 stratified marked squares (Figure A 1) and averaged to obtain an overall cover per quadrat loc ation. In the studies where the quadrat was placed in two l ocations per experimental unit, the average of the two locations provided an estimate for each experimental unit. A similar approach was described by Interrante et al. (2009) to measure canopy cover of bahiagrass ( Paspalum notatum Flgge). Frequency Presence or absence of RP was determined in the same 20 stratified 10 by 10 cm squares used to estimate RP canopy cover. Frequency was calculated as the percentage of the total number of squares ( 20 were assessed per quadrat) where RP was present.

PAGE 151

151 Figure A 1. Quadrat to measure rhizoma peanut (RP) canopy cover and f requency. Each square is 10 by 10 cm. Shaded squares correspond to the squares w h ere measurements were made

PAGE 152

152 LI S T OF REFERENCES Adjei, M. B., and G. M. Prine. 1976. Establishment of perennial peanut. Proc. Soil Crop Sci. Soc. Fla. 35:50 53. Agboola, A.A ., and A.A. Fayemi. 1972. Fixation and excretion of Nitrogen by tropical legumes. Agron. J. 64:409 412. Aiken, G.E., W.D. Pitman, C.G. Chambliss, and K.M. Portier. 1991a. Responses of yearling steers to different stocking rates on a subtropical grass legu me pasture. J. Anim. Sci. 69:3348 3356 Aiken, G., W. D. Pitman, C. G. Chambliss, and K. M. Portier. 1991b. Plant responses to stocking rate in a subtropical gras s legume pasture. Agron. J. 83: 124 129. Barnes, L.W. 1990. Diagnostic report on cotton root rot, Phymatotrichum omnivorum in Arbrook rhizoma peanut. Texas Plant Disease Laboratory, College Station, TX, USA. Beltranena, R., J. Breman, and G.M. Prine. 1981. Yield and quality of Florigraze perennial peanut ( Arachis glabrata Benth.) as affected by cutt ing height and frequency. Proc. Soil Crop Sci. Soc. Fla. 40:153 156. Bennett, L., A. Hammond, M. Williams, C. Chase, and W. Kunkle. 1999. Diet selection by steers using microhistological and stable carbon isotope ratio analyses. J. Anim. Sci. 77:2252 2258 Blount, A.R., R.K. Sprenkel, R.N. Pitman, B.A. Smith, R.N. Morgan, W. Dankers, and T.M. Momol. 2002. Peanut stunt virus reported on perennial peanut in north Florida and southern Georgia. Agronomy facts. SS AGR 37. Agron. Dep. Univ. of Florida, Gainesvi lle, Florida, USA Boddey, R.M., R. Macedo, R.M. Tarr, E. Ferreira, O.C. de Oliveira, C. de P. Rezende, R.B. Cantarutti, J.M. Pereira, B.J.R. Alves, S. Urquiaga. 2004. Nitrogen cycling in Brachiaria pastures: the key to understanding the process of past ure decline. Agric. Ecosyst. Environ. 103:389 403 Bogdan, A.V. 1977. Tropical pastures and fodder plants. Logman Inc. New York. Bowman, A.M., and B.J. Gogel. 1998. Evaluation of perennial peanuts ( Arachis spp.) as forage on the New South Wales north coa st. Trop. Grassl. 32:252 258. Burns, J.C, and C.P. Bagley. 1996. Cool season grasses for pasture. In: L.E. Moser et al., editors, Cool season forage grasses. Agron. Monogr. 34. ASA CSSA SSSA. Madison, WI. p. 321 355.

PAGE 153

153 Cabrera, M.L., D.E. Kissel, and M.F. Vigil. 2005. Nitrogen mineralization from organic residues: Research opportunities. J. Environ. Qual. 34:75 79. Canudas, E.G, K.H. Quesenberry, L.E. Sollenberger, and G.M. Prine. 1989. Establishment of two cultivars of rhizoma peanut as affected by weed c ontrol and planting date. Trop. Grassl. 23:162 170. Catchpoole, D.W., and G. Blair. 1990. Forage tree legumes. I. Productivity and N economy of Leucaena, Gliricidia, Calliandra and Sesbania and tree/green panic mixtures. Aust. J. Agric. Res. 41:521 530. Chambliss, C.G., and. E. Lord. 2001. Florida cow calf management, 2 nd edition Forages. AN118. Department of Animal Sciences, Univ. of Florida, Gainesville. Chapin III, F.S., P.A. Matson, and H.A. Mooney. 2002. Principles of terrestrial ecosystem ec ology. Springer, New York. Chapman, D.F., A.J. Parsons, G.P. Cosgrove, D.J. Barker, D.M. Marotti, K.J. Venning, S.M. Rutter, J. Hill, and A.N. Thompson. 2007. Impacts of spatial patterns in pasture on animal grazing behaviour, intake and performance. Crop Sci. 47:399 415. Cherr, C.M., J.M.S. Scholberg, R. McSorley. 2006. Green manure approaches to crop production: a synthesis. Agron. J. 98:302 319. Cook, S.J. 1980. Establishing pasture species in existing swards: a review. Trop. Grassl. 14:181 187. Cook S.J., R.L. Clem, N.D. MacLeod, and P.A. Walsh. 1993. Tropical pasture establishment. 7. Sowing methods for pasture establishment in northern Australia. Trop. Grassl. 27:335 343. Cosgrove, G.P., G.C. Waghorn, and A.J. Parsons. 1999. Exploring the nutriti onal basis of preference and diet selection by sheep. Proc. NZ. Grassl. Assoc. 61:175 180. Craine, J.M., D. Tilman, D. Wedin, P. Reich, M. Tjoelker, and J. Knops. 2002. Functional traits, productivity and effects on nitrogen cycling of 33 grassland specie s. Func. Ecol. 16:563 574. Cuomo, G.J., D.G. Johnson, and W.A. Head. 2001. Interseeding kura clover and birdsfoot trefoil into existing cool season grass pastures. Agron. J. 93:458 462. Dewhurst R.J., L. Delaby, A. Moloney, T. Boland, and E. Lewis. 2009. Nutritive value of forage legumes used for grazing and silage. Ir. J. Agric. Food Res. 48:167 187.

PAGE 154

154 Drinkwater, L.E., P. Wagoner, and M. Sarrantonio. 1998. Legume based cropping systems have reduced carbon and nitrogen losses. Nature. 396:262 265. Dubeux J.C.B., L.E. Sollenberger, B.W. Mathews, J.M. Scholberg, and H.Q. Santos. 2007. Nutrient cycling in warm climate grasslands. Crop Sci. 47:915 928. Dunavin, L.S. 1992. Florigraze rhizoma peanut in association with warm season perennial grasses. Agron. J. 32:148 151. Ferrell, J.A., and B.A. Sellers. 2012. Weed control in perennial peanut. Agronomy Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Science, University of Florida, SS AGR 261. Ferrell, J.A., B.A. Sellers, C.R. Mudge, and C.A. Smith. 2006. Evaluation of postemergent herbicides on rhizoma peanut injury and yield. Online. Forage and Grazinglands. doi:10.1094/FG 2006 0308 01 RS. Frame, J. 2005. Forage legumes for temperature grasslands. Food and Agricultural O rganization of the United Nations (FAO), Rome, Italy. Science Publishers Inc., Enfield, NH, USA. French, E. C., and G. M. Prine. 1991. Perennial peanut establishment guide. Agronomy facts. SS AGR 35. Agron. Dep. Univ. of Florida, Gainesville, Florida, US A. French, E.C. 1988. Perennial peanut: A promising forage for dairy herd management in the tropics. p. C20 C41. Proc. Int. Conf. Livestock Tropics, Gainesville, FL. 19 25 June 1988. IFAS, Univ. of Florida, Gainesville. French, E.C., G.M. Prine, W.R. Oc umpaugh and R.W. Rice. 1994. Regional experience with forage Arachis in the United States. p. 169 186. In P.C. Kerridge and B. Hardy (ed.) Biology and Agronomy of Forage Arachis. Centro de Agricultura Tropical (CIAT), Cali, Colombia. Fribourg, H.A., J.B. McLaren, K.M. Barth, J.M. Bryan, and J.T. Connell. 1979. Productivity and quality of bermudagrass and orchardgrass ladino clover pastures for beef steers. Agron. J. 71:315 320. Gates, R.N., C. Quarin, and G.S. Pedreira, 2004. Bahiagrass. In: L.E. Moser et al., editors, Warm season (C 4 ) grasses. Agron. Monogr. 45. ASA CSSA SSSA. Madison, WI. p. 651 680. Goncalves, J.L.M., and J.C. Carlyle. 1994. Modeling the influence of moisture and temperature on net nitrogen mineralization in a forested sandy soil. Soil Biol. Biochem. 26:1557 1564.

PAGE 155

155 Graham, P.H., and C.P. Vance. 2003. Legumes: Importance and constraints to greater use. Plant Phys. 131:872 877. Gramshaw, D., J.W. Read, W.J. Collins, and E.D. Carter. 1989. Sown pastures and legume persistence: An Austral ian perspective. In: G.C. Marten et al., editors, Persistence of Legumes. ASA CSSA SSSA, Madison, WI. p. 1 22. Gregory, M.P., and W.C. Gregory. 1979. Exotic germplasm of Arachis L. interspecific hybrids. J. Hered. 70:185 193. Gregory, W.C., A. Kaprovicka s, and M.P. Gregory. 1980. Structure, variation, and classification on Arachis In: R.J. Summerfield and A.H. Bunting, editors, Advances in legume science. Royal Botanical Garden, Surrey, England. p. 468 481. Haby, V.A., and D.L. Robinson. 1997. Soil fer tility and liming practices for productio n of annual ryegrass. In: F.M. Rouquette, and L.R. Nelson, editors, Production and management of of Lolium for forage in the USA. Crop Sci. Soc. Am. Madison, WI. Harris, S.L., D.A. Clark, and P.J. Laboyrie. 1998. Birdsfoot trefoil an alternative legume for New Zealand dairy pastures. Proc. New Zealand Grassl. Assoc. 60:99 103. Heichel, G.H., and K.I. Henjum. 1991. Dinitrogen fixation, nitrogen transfer, and productivity of forage legume grass communities.Crop Sc i. 31:202 208. Hernndez Garay, A., L.E. Sollenberger, C.R. Staples, and C.G.S. Pedreira. 2004. heifers. Crop Sci. 44:1355 1360. competition, coexisten ce and coevolution and their impact on the breeding of forage crop mixtures. Theor. Appl. Genet. 79:168 176. Hirsch, A.M., M.R. Lum, and J.A. Downie. 2001. What makes the rhizobia legume symbiosis so special? Plant Phys. 127:1484 1492. Hodgson. J., K. C ameron, D. Clark, L. Condron, T. Fraser, M. Hedley, C. Holmes, P. Kemp, R. Lucas, D. Moot, S. Morris, P. Nicholas, N. Shadbot, G. Sheath, I. industries: Efficient use of grassland resour ces. In: S.G. Reynolds and J. Frame, editors, Grasslands: Developments, opportunities, perspectives. Food and Agriculture Organization of the United Nations (FAO). Science Publishers Inc., Enfield, NH, USA. p. 181 205.

PAGE 156

156 Hyslop, M.G., P.D. Kemp, and J. Hodg son. 1999. Vegetatively reproductive red clovers ( Trifolium pratense L.): An overview. Proc. New Zealand Grassl. Assoc. 61:121 126. Interrante, S.M., J.P. Muir, A.M. Islam, A.L. Maas, W.F. Anderson, and T.J. Bulter. 2011. Establishment, agronomic characte ristics, and dry matter yield of rhizoma peanut genotypes in cool environments. Crop Sci. 51:2256 2261. Interrante, S.M., L.E. Sollenberger, A.R. Blount, S.W. Coleman, U.R. White, and K. Liu. 2009. Defoliation management of bahiagrass germplasm affects co ver and persistence related responses. Agron. J. 101:1381 1387. Johnson, C.R., B.A. Reiling, P. Mislevy, and M.B. Hall. 2001. Effects of nitrogen fertilization and harvest date on yield, digestibility, fiber, and protein fractions of tropical grasses. J. Anim. Sci. 79:2439 2448. Johnson, S.E., L.E. Sollenberger, I.F. Andrade, and J.M. Bennett. 2002. Nutritive value of rhizoma peanut growing under varying levels of artificial shade. Agron. J. 94:1071 1077. Johnson, W.C. III, E.P Prostko, and B.G. Mullini x, Jr. 2010. Texas panicum ( Panicum texacum) control in strip tillage peanut ( Arachis hypogaea) production. Peanut Sci. 29:141 145. Jones, M.B. 1981. A comparison of sward development under cutting and continuous grazing management. p. 63 67. In C.E. Wrig ht (ed.) Plant Physiology and Herbage Production, Occasional Symposium No.13, British Grassland Society, Hurley. Jung, G.A., J.L. Griffin, R.E. Kocher, J.A. Shaffer, and C.F. Gross. 1985. Performance of switchgrass and bluestem cultivars mixed with cool s eason species. Agron. J. 77:846 850. Kalmbacher, R. S., and F. G. Martin. 1983. Light penetrating a bahiagrass canopy and its influence on establishing jointvetch. Agron. J. 75:465 468. Krapovickas, A. and W.C. Gregory. 2007. Taxonomy of the genus Arachi s (Leguminosae). Translated by D.E. Williams and C.E. Simpson. Bonplandia 16 (Supl.). Kretschmer, A.E., G.H. Snyder, G.J. Gascho, and J.B. Brolmann. 1973. Production of six tropical legumes each in combination with three tropical grasses in Florida. Agron J. 65:890 892. Lascano, C.E. 1994. Nutritive value and animal production of forage Arachis p. 109 121. In P. Kerridge and B. Hardy (ed.) Biology and Agronomy of Forage Arachis. Centro Internacional de Agricultura Tropical (CIAT), Cali, Colombia.

PAGE 157

157 Lasca no, C.E. 2000. Selective grazing on grass legume mixtures in tropical grasslands. p. 249 263. In G. Lemaire et al. (ed.) Grassland Ecophysiology and Grazing Ecology, CABI publishing, New York, NY. Lesleighter, L.C., and H.M. Shelton. 1986. Adoption of the shrub legume Leucaena leucocephala in Central and Southeast Queensland. Trop. Grassl. 20:97 106. Mackowiak, C.L, A.R. Blount, E.A. Hanlon, M.L. Silveira, M.B. Adjei, and R.O. Myer. 2008. Getting the most out of bahiagrass fertilization. Soil and Water Sc ience Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida, SL 249. Marotti, D.M., G.P. Cosgrove, D.F. Chapman, and A.J. Parsons. 2001. Novel methods of forage presentation to boost nutritio n and performance of grazing dairy cows. Aust. J. Dairy Tech. 56:159. Marten, G.C. 1985. Nutritional value of the legume in temperate pastures of the U.S. In: R.F. Barnes et al., editors, Forage legumes for energy efficient animal production. Proc. Trila teral Workshop, Palmerston, North, NZ. 30 Apr. 4 May 1984. USDA, ARS, Washington, DC. p. 204 212. Matches, A. 1989. A survey of legume production and persistence in the United States. In: G.C. Marten et al., editors, Persistence of Legumes. ASA CSSA SSSA, Madison, WI. p. 37 44. Miles, J.W., C.B. do Valle, I.M. Rao, and V.P.B. Euclides. 2004. Brachiariagrasses. p. 745 783. In L.E. Moser et al. (ed.) Warm season (C 4 ) grasses. Agron. Monogr. 45. ASA/CSSA/SSSA, Madison, WI. Minson, D.J. 1981. Nutritional dif ferences between tropical and temperate pastures. In: Morley F.H.W., editors, Grazing animals. Elsevier, Amsterdam. p. 143 157. Mislevy, P., M.J. Williams, A.S. Blount, and K.H. Quesenberry. 2008. Influence of harvest management of rhizoma peanut producti on, nutritive value, and persistence of Flatwood soils. Forage and Grazinglands. doi:10.1094/FG 2007 1108 01 RS. Mossler, M., and M.J. Aerts. 2010. Florida Crop/Pest management profiles: Peanuts. Pesticide Information Office, Agronomy Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida, CIR 1260. Mueller, J.P., and D.S. Chamblee. 1984. Sod seeding of Ladino clover and alfalfa as influenced by seed placement, seeding date, and grass suppress ion. Agron.J. 76:284 289.

PAGE 158

158 Muir, J.P., and W.D. Pitman. 1991. Grazing tolerance warm season legumes in peninsular Florida. Agron. J. 83:297 302. Muir, J.P., W.D. Pitman, and J.L. Foster. 201 1 Sustainable, low input, warm season, grass legume grassland mi xtures: mission (nearly) impossible? Grass Forage Sci. 66:301 315. Newman Y., Vendramini J., and A. Blount. 2011. Bahiagrass ( Paspalum notatum ): Overview and management. EDIS publication. SS AGR 332. Department of Agronomy, Institute of Food and Agricultu ral Sciences, University of Florida, Gainesville, FL, USA. Niles, W.L., E.C. French, P.E. Hildebrand, G. Kidder, and G.M. Prine. 1990. Establishment of Florigraze rhizome peanut ( Arachis glabrata Benth) as affected by lime, phosphorus, potassium, magnesiu m, and sulfur. Proc. Soil Crop Sci. Soc. Fla. 49:207 210. Ocumpaugh, W. R. 1990. Production and nutritive value of Florigraze rhizoma peanut in a semiarid climate. Agron. J. 82:179 182. Ortega S., J.A., L.E. Sollenberger, K.H. Quesenberry, C.S. Jones, a nd J.A. Cornell. 1992. Productivity and persistence of rhizoma peanut pastures under different grazing managements. Agron. J. 84:799 804. Parsons, A.J., and E.L. Leafe. 1981. Photosynthesis and carbon balance of a grazed sward. p. 69 71. In C.E. Wright (e d.) Plant Physiology and Herbage Production, Occasional Symposium No. 13, British Grassland Society, Hurley. Parsons, A.J., G.R. Edwards, D.F. Chapman, and R.A. Carran. 2006. How far have we Proc. New Zealand Grassl. Assoc. p. 7 13. Parsons, A.J., I.R. Johnson, and J.H.H. Williams. 1988. Leaf age structure and canopy photosynthesis in rotationally and continuously grazed swards. Grass Forage Sci. 43:1 14. Parsons, A.J., J.A. Newman, P.D. Penning, A. Harvey, and R.J. Orr 1994 Diet preferences of sheep: effects of recent diet, physiological state and species abundance. J.Anim. Ecol. 63:465 478. Pecetti, L., and E. Piano. 2005. Heritability of morphophysiological traits and inbreeding effects in grazing type lucerne. Plant Bre e ding. 124:176 179. Pengelly, B.C., A. Whitbread, P.R. Mazaiwana, and N. Mukombe. 2003. Tropical forage research for the future better use of research resources to deliver adoption and benefits to farmers. Trop. Grassl. 37:207 216.

PAGE 159

159 Piano, E., P. Valent ini, P., L. Pecetti, M. Romani. 1996. Evaluation of lucerne germplasm collection in relation to traits conferring grazing tolerance. Euphytica. 89:279 288. Pitman, W., C. Chambliss, and A. Kretschmer Jr. 1988. Persistence of tropical legumes on peninsular Florida flatwoods (Spodosols) at two stocking rates. Trop. Grassl. 22:27 33. Pitman, W.D., K.M. Portier, C.G. Chambliss, and A.E. Kretschmer. 1992. Performance of yearling steers grazing bahiagrass pastures with summer annual legumes or nitrogen fertiliz er in subtropical Florida. Trop. Grassl. 26:206 211. Poppi, D.P., and S.R. McLennan. 1995. Protein and energy utilization by ruminants at pasture. J. Anim. Sci. 73:278 290. Prine, G. M. 1980. Forage seed and rhizome growers needed in Florida. Proceedings of the 29th Beef Cattle Short Course. IFAS. University of Florida, Gainesville. p. 21 23. Prine, G.M., E.C. French, A.R. Blount, M.J. Williams, and K.H. Quesenberry. 2010. Registration of Arblick and Ecoturf rhizoma peanut germplasms for ornamental or fo rage use. J. Plant Reg. 4:145 148. Prine, G.M., L.S. Dunavin, J.E. Moore, and R.D. Roush. 1981. Florigraze rhizoma peanut a perennial forage legume. Circ. S 275. Florida Agric. Exp. Stn., Gainesville, Florida, USA. Prine, G.M., L.S. Dunavin, J.E. Moore, and R.D. Roush. 1986 a Registration of 1085. Prine, G.M., L.S. Dunavin, R.J. Glennon, and R.D. Roush. 1986 b Arbrook rhizoma peanut: A perennial forage legume. Circ. S 332. Florida Agric. Exp. Stn., Gainesvi lle, Florida, USA. Prine, G.M., L.S. Dunavin, R.J. Glennon, and R.D. Roush. 1990. Registration of 744. Prostko, E.P., C.W. III Johnson, B.G. Jr. Mullinix. 2001. Annual grass control with preplant incorporated an d preemergence applications of ethalfluralin and pendimethalin in peanut Arachis hypogea. Weed Sci. 15:36 41. Quesenberry, K.H., A.R. Blount, P. Mislevy, E.C. French, M.J. Williams, and G.M. Prine. anut cultivars with high dry matter yields, persistence, and disease tolerance. J. Plant Reg. 4:17 21.

PAGE 160

160 Redfearn, D.D., B.C. Venuto, and W.D. Pitman. 2001. Nutritive value responses of rhizoma peanut to nitrogen and harvest frequency. Agron. J. 93:107 112. Redfearn, D.D., B.C. Venuto, W.D. Pitman, M.W. Alison, and J.D. Ward. 2002. Cultivar and environmental effects on annual ryegrass forage yield, yield distribution, and nutritive value. Agron. J. 42:2049 2054. Reed, R.L., and W.R. Ocumpaugh. 1991. Screen ing rhizoma peanut for adaptation to calcareous soils 1. J. Plant Nutr. 14:163 174. Rice, R.W., L.E. Sollenberger, K.H. Quesenberry, G.M. Prine, and E.C. French. 1995. Defoliation effects on rhizoma perennial peanut rhizome characteristics and establishme nts performance. Crop Sci. 35:1291 1299. Rochon, J.J., C.J. Doyle, J.M. Greef, A. Hopkins, G. Molle, M. Sitzia, D. Scholefield, and C.J. Smith. 2004. Grazing legumes in Europe: a review of their status, management, benefits, research needs and future pros pects. Grass Forage Sci. 59:197 214. Romero, F., H.H. Van Horn, G.M. Prine, and E.C. French. 1987. Effect of cutting interval upon yield, composition and digestibility of Florida 77 alfalfa and Florigraze rhizoma peanut. J. Anim. Sci. 65:786 796. Rusland G.A., L.E. Sollenberger, K.A. Albrecht, C.S. Jones, and L.V. Crowder. 1988. Animal performance on limpograss aeschynomene and nitrogen fertilized limpograss pastures. Agron. J. 80:957 962. Ruttinger, A.E. 1989. Evaluation of perennial Arachis germplasm for agronomic performance, response to root knot nematode, and three peanut leaf spot diseases. M.S. thesis. University of Florida, Gainesville, Fl, USA. Sainju, U.M., T.H. Terrill, S. Gelaye, and B.P. Singh. 2006. Soil carbon and nitrogen pools under long term productivity of rhizoma peanut and perennial weeds management systems. Arch. Agron. Soil Sci. 52:45 59. Saldivar, A.J., W.R. Ocumpaugh, R.R. Gildersleeve, and G.M. Prine. 1992a. Growth analysis of 'Florigraze' rhizoma peanut: Shoot and rh izome dry matter production. Agron. J. 84:444 449. Saldivar, A.J., W.R. Ocumpaugh, R.R. Gildersleeve, and G.M. Prine. 1992b. Total J. 84:439 444. Schlueter, D., and B.F. Tra cy. 2012. Sowing methods effects on clover establishment into permanent pasture. Agron. J. 104:1217 1222.

PAGE 161

161 Schwinning, S., and A.J. Parsons. 1996. Interactions between grasses and legumes: understanding variability in species composition. In: Occasional S ymposium Sustainable Farming Syst./ Brit. Grassl. Soc. p.153 163 Sears, P.D. 1962. Exploitation of high production pastures in New Zealand. Proc. NZ Grassl. Assoc. 9: 57 63. Sheath, G.W., and R.J.M. Hay, 1989. Overview of legume persistence in New Zealan d. In: G.C. Marten et al., editors, Persistence of Legumes. ASA CSSA SSSA, Madison, WI. p. 23 36. Shelton, H.M. 1994. Establishment of forage tree legumes. In: R.C. Gutteridge and H.M. Shelton, editors, Forage tree legumes in tropical agriculture. CAB Int ernational, Wallingford, UK. p.132 142. Shelton, H.M., S. Franzel, and M. Peters. 2005. Adoption of tropical legume technology around the world: analysis of success. Trop. Grassl. 39:198 209. Smith Jr., S.R., J.H. Bouton, A. Singh, and W.P. McCaughey. 20 00. Development and evaluation of grazing tolerant alfalfa cultivars: A review. Can. J.Plant Sci. 80: 503 512. Solomon, J.K.Q., B. Macoon, D.J. Lang, J.A. Parish, and R.C. Vann. 2011. A novel approach to grass legume management. Crop Sci. 51:1865 1876. S ollenberger, L.E., and C.G. Chambliss. 1991. Regional and seasonal forage production limits. p. 25 34. In Beef Cattle Short Course Proc. Gainesville, FL, 1 3 May 1991. Inst. Food Agric. Sci., Univ. of Fla., Gainesville. Sollenberger, L.E., and M. Collins. 2003. Legumes for southern areas. p. 191 213. In R.F Barnes et al. (ed.) Forages, Volume 1 An introduction to grassland agriculture, Iowa State Press, Ames, Iowa. Sollenberger, L.E., and R.S. Kalmbacher. 2005. Aeschynomene and carpon desmodium: legumes for bahiagrass pasture in Florida. Trop. Grassl. 39: 227. Sollenberger, L.E., C.S. Jones, and G.M. Prine. 1989. Animal performance on dwarf elephantgrass and perennial peanut pastures. p. 1189 1190. In: XVI Int. Grassl. Cong. Nice, France. Sollenberger, L.E., K.H. Quesenberry, and J.E. Moore. 1987. Effects of grazing management on establishment and productivity of aeschynomene overseeded in limpograss pastures. Agron. J. 79:78 82.

PAGE 162

162 Stewart, R.L., Jr., J.C.B. Dubeux, Jr, L.E. Sollenberger, J.M.B. Vendram ini, and S.M. Interrante. 2005. Stocking method affects plant responses of Pensacola bahiagrass pas tures. Online. Forage and Grazinglands doi:10.1094/FG 2005 1028 01 RS. Stricker, J.A., A.G. Matches, G.B. Thompson, V.E. Jacobs, F.A. Martz, H.N. Wheaton, H.D. Currence, and G.F. Krause. 1979. Cow calf production on tall fescue ladino clover pastures with and without nitrogen fertilization or creep feeding: spring calves. J. Anim. Sci. 48:13 25. Templeton Jr., W.C. 1976. Legume nitrogen versus fertilizer ni trogen for cool season grasses. In: Biological N fixation in forage livestock systems: proceedings of a symposium. ASA Special Publication No. 28. ASA CSSA SSSA Madison, WI. p. 35 54. Thomas, R.J. 1992. The role of the legume in the nitrogen cycle of pro ductive and sustainable pastures. Grass Forage Sci. 47:133 142. Thomas, R.J. 1994. Rhizobium requirements, nitrogen fixation, and nutrient cycling in forage Arachis p. 84 94. In P. Kerridge and B. Hardy (ed.) Biology and Agronomy of Forage Arachis Centr o Internacional de Agricultura Tropical (CIAT), Cali, Colombia. Thomas, R.J. 1995. Role of legumes in providing N for sustainable tropical pasture systems. Plant Soil 174:103 118. Thomas, R.J., and N.M. Asakawa. 1993. Decomposition of leaf litter from t ropical forage grasses and legumes. Soil Biol. Biochem. 25:1351 1361. Valencia, E., M.J. Williams, and L.E. Sollenberger. 1999. Yield and botanical composition of rhizoma peanut grass swards treated with herbicides. Agron. J. 91:956 961. Valencia, E., M. J. Williams, C.C. Chase, Jr., L.E. Sollenberger, A.C. Hammond, R.S. Kalmbacher, and W.E. Kunkle. 2001. Pasture management effects on diet composition and cattle performance on continuously stocked rhizoma peanut mixed grass swards. J. Anim. Sci. 79:2456 24 64. Valentim, J.F. 1987. Effect of environmental factors and management practices on nitrogen fixation of rhizoma peanut and transfer of nitrogen from the legume to an associated grass. Ph.D. dissertation. University of Florida, Gainesville, FL, USA. Val entim, J.F., and C.M.S. Andrade. 2005. Forage peanut ( Arachis pintoi ): a high yielding and high quality tropical legume for sustainable cattle production systems in the western Brazilian Amazon. p. 329. In International Grassland Congress 20th. Dublin, Ire land.

PAGE 163

163 Valentim, J.F., O.C. Ruelke, and G.M. Prine. 1986. Yield and quality responses of tropical grasses, a legume and grass legume associations as affected by fertilizer nitrogen. Soil Crop Sci. Soc. Fla. Proc. 45:138 143. Van Soest, P.J. 1982. Nutritio nal ecology of the ruminant. Cornell University Press, Ithaca, NY. Vengris, J. 1965. Use of grass killing herbicides in establishing grass legume mixtures. Agron. J. 57: 59 61. Venuto, B.C., D.D. Redfearn, and W.D. Pitman. 1998. Rhizoma peanut responses to harvest frequency and nitrogen fertilization on Louisiana coastal plain solis. Agron. J. 90:826 830. Venuto, B.C., G.C. Cuomo, and W.M. Wlkins. 1995. Performance of rhizomatous perennial peanut in southern Louisiana. In: M.W. Alison, editor, Proc. Lou isiana Association of Agronomists. 29 30 Mar. 1995. LA, Agric. Ctr., Baton Rouge, LA. p. 54 63. Venuto, B.C., W.M. Elkins, R.W. Hintz, and R.L. Reed. 1997. Comparison of seed derived lines from 'Florigraze' rhizoma perennial peanut. Crop Sci. 37:1098 1102 Vitousek, P.M., J.D. Aber, R.W. Howarth, G.E. Likens, P.A. Matson, D.W. Schindler, W.H. Schlesinger, and D.G. Tilman. 1997. Human alteration of the global nitrogen cycle: sources and consequences. Ecol. Appl. 7:737 750. Vitousek, P.M., R. Naylor, T. Cr ews, M.B. David, L.E. Drinkwater, E. Holland, P.J. Johnes, J. Katzenberger, L.A. Martinelli, P.A. Matson, G. Nziguheba, D. Ojima, C.A. Palm, G.P. Robertson, P.A. Sanchez, A.R. Townsend, and F.S. Zhang. 2009. Nutrient imbalances in agricultural development. Science. 324:1519 1520. Wedin, D.A., and M.P. Russelle. 2006. Nutrient cycling in forage production systems. In: R.F. Barnes et al., editors, The Science of Grassland Agriculture. Volume II, 6 th ed. Blackwell publishing, Malden, MA. p. 137 148. Whitbrea d, A.M., C.A. Hall, and B.C. Pengelly. 2009. A novel approach to planting grass legume pastures in the mixed farming zone of southern inland Queensland, Australia. Crop Past. Sci. 60:1147 1155. Whitmore, A.P. 1996. Modeling the release and loss of nitroge n after vegetable crops. Neth. J. Agric. Sci. 44:73 86. Williams, M. 1993. Planting date and preplant tillage effects on emergence and survival of rhizoma perennial peanut. Crop Sci. 33:132 13 6

PAGE 164

164 Williams, M.J. 1994. Growth characteristics of rhizoma pean ut and nitrogen fertilized bahiagrass swards. Agron. J. 86:819 823. Williams, M.J., A.C. Hammond, W.E. Kunkle, and T.H. Spreen. 1991. Stocker performance on continuously stocked mixed grass rhizoma peanut and bahiagrass pastures. J. Prod. Agric. 4:19 24 Williams, M.J., C.A. Kelly Begazo, R.L. Stanley, Jr., K.H. Quesenberry, and G.M. Prine. 1997. Establishment of rhizoma peanut: Interaction of cultivar, planting date, and location on emergence and rate of cover. Agron. J. 89:981 987. Williams, M.J., E. Valencia, and L.E. Sollenberger. 2002. No till establishment o f rhizoma peanut. Agron. J. 94: 1350 1354. Williams, M.J., Y.C. Newman, and A.R. Blount. 2011. Rhizoma perennial peanut. Agronomy fact sheets. SS AGR 349. Agronomy Deparment, University of Flor ida, Gainesville, Florida, USA. Williams, M.J., T.R. Sinclair, P. Mislevy, K.H. Quesenberry, A.R. Blount, and S.W. Coleman. 2008. Photoperiod sensitivity of rhizoma peanut germplasm. Agron. J. 100:1366 1370. Wilson, J.R., and D.J. Minson. 1980. Prospects for improving the digestibility and intake of trop ical grasses. Trop. Grassl. 14: 253 259.

PAGE 165

165 BIOGRAPHI CAL SKETCH Miguel S. Castillo was born in 1984 in Loja, Ecuador. His interest in agriculture developed at early stages of his life while listenin g to family conversations and while working on a family farm dedicated to the production of horticultural crops and to a dairy herd He received a B.S. degree in agricultural science and production (2006) from Zamorano University, Honduras, C.A. In spring 2006 he came to the Everglades Research and Education Center of the University of Florida (UF), as a short term scholar. Miguel joined the Agronomy Department at UF as Research Assistant in spring received the Graduate School Fellowship from UF to continue with the Ph.D. program, a nd graduated with his Ph.D. in agronomy and a minor in soil and w ater s cience in spring 201 3 He conducted both graduate degrees (M.S. and Ph.D.) in Dr. Lynn E. Sollenberg research and training program with the long term goal of increasing/maintaining sustainability in agro eco systems.