Citation
Defoliation Frequency and Location Effects on Root-Rhizome Mass, Herbage Accumulation, and Canopy Characteristics of Rhizoma Peanut Entries Differing in Growth Habit

Material Information

Title:
Defoliation Frequency and Location Effects on Root-Rhizome Mass, Herbage Accumulation, and Canopy Characteristics of Rhizoma Peanut Entries Differing in Growth Habit
Creator:
Cooley, Katie Doone
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (120 p.)

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Agronomy
Committee Chair:
SOLLENBERGER,LYNN E
Committee Co-Chair:
BLOUNT,ANN RACHEL SOFFES
Committee Members:
DUBEUX,JOSE CARLOS
SILVEIRA,MARIA LUCIA

Subjects

Subjects / Keywords:
arachis -- belowground -- biomass -- carbon -- core -- cultivar -- defoliation -- drainage -- dynamics -- forage -- frequency -- glabrata -- ingrowth -- legume -- nutritive -- peanut -- rhizoma -- root -- soil -- value
Agronomy -- Dissertations, Academic -- UF
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Agronomy thesis, M.S.

Notes

Abstract:
Grasslands occupy nearly 40% of USA land area and, in addition to supplying feed for livestock, provide valuable services to society including storing carbon in soil, minimizing soil erosion, purifying water by removing excess nutrients, and providing wildlife habitat. Most grasslands in the southeastern US are comprised of warm-climate grass species that are dependent on nitrogen fertilizer and have relatively low nutritional value for livestock. Inclusion of legumes in mixture with grasses can increase nutritive value and benefit nutrient cycling and soil quality. Rhizoma peanut (RP; Arachis glabrata Benth.) is a long-lived legume in the US Gulf Coast Region, providing an array of ecosystem services. Most current RP cultivars were selected for upright growth and hay production, but newer introductions and selections exist that are lower growing with potential for grazing or ornamental use. Data describing key forage responses are lacking for these entries, and evaluation of productivity, nutritive value, and persistence is needed in different soil environments and under different cutting regimes. Also needed is a better understanding of root-rhizome accumulation, which is a large contributor to soil carbon storage. One study compared the response of 14 RP entries to two defoliation frequency treatments, including one harvest per year at season end vs. two harvests per year occurring at mid-season and end of the growing season. A single defoliation event per year resulted in significant reductions in herbage accumulation (HA) and nutritive value for most RP entries relative to harvests at both the middle and end of the growing season. Herbage accumulation was not affected by defoliation frequency for the entry Quincy-Beta, perhaps because of its disease tolerance. Greater defoliation frequency decreased root-rhizome mass and non-structural carbohydrate pool, a response of particular importance to producers using RP fields both for forage and as a source of rhizomes for planting material. A second study compared the 14 RP entries growing in a well-drained soil at Quincy, Florida and in a seasonally flooded soil at Hague, Florida. Well-drained soil at Quincy was more favorable to RP entries in general than seasonally-saturated soil at Hague, which led to many entries having greater HA at Quincy than Hague. The entry Ona 33 had nearly the same HA at both locations, and it is considerably better adapted to seasonally-saturated soils than all RP entries currently being used commercially. A third study reports on the design and use of a modified ingrowth core for measuring root accumulation rate. Increasing the mesh size of the fabric to 4 mm and using a wire frame cage to support the core allowed it to maintain its geometry over 100-d deployment periods and detect differences among RP entries. Finally, results of these studies suggest there is opportunity to select new RP cultivars that are adapted to wetter soils and are well suited for grazing and ornamental uses while maintaining high levels of forage production, nutritive value, and belowground biomass. ( en )
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.
Thesis:
Thesis (M.S.)--University of Florida, 2018.
Local:
Adviser: SOLLENBERGER,LYNN E.
Local:
Co-adviser: BLOUNT,ANN RACHEL SOFFES.
Statement of Responsibility:
by Katie Doone Cooley.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
LD1780 2018 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

DEFOLIATION FREQUENC Y AND LOCATION EFFEC TS ON ROOT RHIZOME M ASS, HERBAGE ACCUMULATION AND CANOPY CHARACT ERISTICS OF RHIZOMA PEANUT ENTRIES DIFFERING IN GROWTH HABIT By KATIE DOONE COOLEY A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNI VERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2018

PAGE 2

2018 Katie Doone Cooley

PAGE 3

3 To my child may you be brave like me when I bring you into this world

PAGE 4

4 ACKNOW LEDGEMENTS First and foremost, I would like to thank my husband, Ben Cooley for his love wisdom and humor Every moment together truly is a blessing. I would also like to thank my advisor, Dr. Ly nn Sollenberger for being an incredible mentor and human be ing. Admitting me into his program has made all the difference in my life. Thank you to my co chair Dr. Ann Blount for allowing me to join her experiments and for our conversations about the Virgin Islands and baby backpacks. Thank you to my committee memb ers, Dr. Maria Lucia Silveira and Dr. Jos Carlos Dubeux Jr. for their support and feedback during my studies. All of my committee members have demonstrated the exemplary scholarship and personal i ntegrity that, to me, distinguishes alumni of the forage an d soils programs at the University of Florida. I have been fortunate to have joined a well rounded lab group. I am grateful to Marta Moura Kohmann and Liliane Severino da Silva for their friendship I have never felt out of place during my studies, thanks to them. I hope we remain lifelong friends. Thank you to my lab mates, Parmeshwor Aryal and Erin Shepard for being wonderful colleagues. The dedication that Richard Fethiere has shown to my lab analyses has been a great help in my experiments and my profes sional development. Thank you to Dwight Thomas, whose contagious spirit made fieldwork so fun that I almost enjoyed limpograss harvests. I would also like to thank Paul Reith, Matthew Bailey and the staff at the North Florida Research and Education Center who have been of great assistance during my experiments. There are many people who helped me build strong personal and academic foundation s prior to my studies at the University of F lorida. Thank y ou to my teachers and friends from the Harvard Graduate S chool of Design, especially Gary Hilderbrand, Andrea Hansen Phillips,

PAGE 5

5 Steven Handel, Siobhan Aitchison, a nd Frank Hu for encouraging me to chart my own course. I will miss Randy Brown, Doug Horne, Beau Rogalin Pame la Rogalin and Debba Jean Lindstrom, ment ors and friends, who all passed away within a short time of one another during my time here. I hope this thesis would have made them proud. I am grateful to my parents, Derek and Kerrigan Hotchkiss f or their sacrifices and for providing honest examples of how life might be lived. Lastly, thank you to my sister, Maggie Hotchkiss for her wit, passion, and friendship.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDGEMENTS ................................ ................................ ................................ ............ 4 LIST OF TABLES ................................ ................................ ................................ .......................... 9 LIST OF FIGURES ................................ ................................ ................................ ...................... 10 ABSTRACT ................................ ................................ ................................ ................................ .. 12 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 14 2 LITERATURE REVIEW ................................ ................................ ................................ ....... 17 Use of R hizoma Peanut in Florida ................................ ................................ .......................... 17 Rhizoma Peanut Introduction and Available Cultivars ................................ .......................... 18 Rhizoma Peanut Responses to Defoliation ................................ ................................ ............. 19 Herba ge Accumulation ................................ ................................ ................................ ... 19 Nutritive Value ................................ ................................ ................................ ............... 23 Grasslands and Soil C Sequestration ................................ ................................ ...................... 24 Relationship of Below Ground Biomass and Root Growth with Soil C Accumulation in Grasslands ................................ ................................ ................................ ........................... 28 3 RHIZOMA PEANUT HERBAGE ACCUMULATION, NUTRITIVE VALUE, AND ROOT RHIZOME RESPONSES TO DE FOLIATION FREQUENCY ................................ 30 Overview ................................ ................................ ................................ ................................ 30 Materials and Methods ................................ ................................ ................................ ............ 32 Experim ental Sites ................................ ................................ ................................ .......... 32 Treatments and Experimental Design ................................ ................................ ............ 33 Response Variables ................................ ................................ ................................ ........ 33 Peanut rust, herbage accumulation, nutritive value, and canopy characteristics .... 33 Root rhizome mass and total non structural carbohydrate concentration and r r h i zome diameter ................................ ................................ ................................ .... 35 Statistical Analyses ................................ ................................ ................................ ......... 36 Results and Discussion ................................ ................................ ................................ ........... 36 Peanut Rust, Herbage Accumulation, and Sward Canopy Characteristics .................... 36 Peanut rust incidence ................................ ................................ ............................... 36 Herbage accumulation ................................ ................................ ............................. 37 Canopy h eight ................................ ................................ ................................ ......... 39 Herbage bulk density ................................ ................................ ............................... 40 Nutritive Value Responses ................................ ................................ ............................. 41

PAGE 7

7 He rbage crude protein ................................ ................................ ............................. 41 Herbage in vitro digestibility ................................ ................................ ................... 42 Below Grou nd Responses ................................ ................................ .............................. 44 Root r hizome mass ................................ ................................ ................................ .. 44 Rhizome diameter ................................ ................................ ................................ ... 45 Root rhizo me total non structural carbohydrate concentration ............................... 46 Root rhizome total non structural carbohydrate pool ................................ ............. 47 Implications of the R esearch ................................ ................................ ................................ ... 47 4 ABOVE AND BELOW GROUND RESPONSES OF RHIZOMA PEANUT EXPERIMENTAL LINES AND CULTIVARS WHEN GROWN AT TWO LOCATIONS ................................ ................................ ................................ ................................ ................. 58 Overview ................................ ................................ ................................ ................................ 58 Materials and Method s ................................ ................................ ................................ ............ 59 Experimental Sites ................................ ................................ ................................ .......... 59 Treatmen ts and Experimental Design ................................ ................................ ............ 60 Response Variables ................................ ................................ ................................ ........ 61 Herbage accumulation, nutritive value, and canopy characteristics ....................... 61 Root rhizome mass and total non structural carbohydrate concentration and rhizome diameter ................................ ................................ ................................ 62 Statistical Analyses ................................ ................................ ................................ ......... 62 Results and Discussion ................................ ................................ ................................ ........... 63 Herbage Accumulation and Sward Canopy Characteristics ................................ ........... 63 Herba ge accumulation ................................ ................................ ............................. 63 Canopy h eight ................................ ................................ ................................ ......... 65 Herbage bulk density ................................ ................................ ............................... 65 Nutriti ve Value Responses ................................ ................................ ............................. 66 Herbage crude protein ................................ ................................ ............................. 66 Herbage in vitro digestibility ................................ ................................ ................... 68 Below Ground Responses ................................ ................................ .............................. 69 Root rhizome mass ................................ ................................ ................................ .. 69 Rhizome diameter ................................ ................................ ................................ ... 70 Total non structural carbohydrate concentration ................................ .................... 70 Total non structural carbohydrate pool ................................ ................................ ... 71 Implications of the Res earch ................................ ................................ ................................ ... 72 5 A MODIFIED ROOT INGROWTH CORE DEVICE TO MEASURE ROOT ACCUMULATION RATE OF PERENNIAL FORAGE SPECIES ................................ ...... 84 Overview ................................ ................................ ................................ ................................ 84 Materials and Method s ................................ ................................ ................................ ............ 85 Design and Construction of the Ingrowth Core ................................ .............................. 85 Use of the Ingrowth Core in a Field Experiment ................................ ........................... 87

PAGE 8

8 Treatments and experimental design ................................ ................................ ....... 87 Placement of the ingrowth cores ................................ ................................ ............. 87 Response Variables ................................ ................................ ................................ ........ 88 Statistical Analysis ................................ ................................ ................................ ......... 89 Results and Discussion ................................ ................................ ................................ ........... 89 Implications ................................ ................................ ................................ ............................. 93 6 CONCLUSIONS ................................ ................................ ................................ ................... 104 Rhizoma Peanut Herbage Accumulati on, Nutritive Value, and Root Rhizome Responses t o Defoliation Frequency Chapter 3 ................................ ................................ ................... 105 Above and Below ground responses of Rhizoma Peanut Experimental Lines and Cultivars when Grown at Two L ocations Differing in Soil Characteristics Chapter 4 ................. 107 Design and Use of a Root Ingrowth Core Dev ice to Measure Root Accumulation Rate of Perennial Forage Species Chapter 5 ................................ ................................ ............... 108 Implications of Research ................................ ................................ ................................ ....... 108 Future Research Needs ................................ ................................ ................................ ......... 109 APPENDIX : HERBAGE AC CUMULATION, CANOPY HEIGHT, HERBAGE BULK DENSITY, HERBAGE CRUDE PROTEIN, AND HERBAGE IN VITRO DIGESTIBLE ORGANIC MATTER AT HARVEST FOR RH IZOMA PEANUT ENTRY CHICO AND THE AVERAGE OF ALL ENTRIES HARVESTED TWICE PER YEAR IN EACH OF 2 YR ................................ .......................... 110 WORKS CITED ................................ ................................ ................................ ......................... 111 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ...... 120

PAGE 9

9 LIST OF TABLES page 3 1 Monthly rainfall at the experimental site in Quincy, FL and the 30 yr average rainfall for Quincy. ................................ ................................ ................................ .............................. 49 3 2 Rhizoma peanut cultivars, germplasms, introductions, and select ions (referred to as entries) evaluated at the North Florida Research and Education Center (NFREC) at Quincy, FL. ................................ ................................ ................................ ....................... 50 3 3 Peanut rust ( Puccinia arachidis ) disease incidence observed on leaflets of 14 rhizoma peanut entries at June and October 2017 harvest dates at Quincy, FL. ............................ 51 3 4 Main effects of defoliation frequency on 14 rhizoma peanut entries during 3 yr of defoliation at Quincy, FL.. ................................ ................................ ................................ 55 4 1 Monthly rainfall at the experimental sites near Quincy and Hague, FL and the 30 year average rainfall for Quincy and Hague, FL. ................................ ................................ ..... 74 4 2 Main effects of location on responses for which there was no entry x location interaction. ................................ ................................ ................................ ................................ ........... 76 5 1 Herbage accumulation and canopy height at harvest for six rhizoma peanut entries harveste d twice per year in each of 2 yr, and root rhizome mass of the sa me entries measured at season end of 2 yr. ................................ ................................ ...................... 103 5 2 Ingrowth rate of root rhizome mass accumulation for six rhizoma peanut entrie s measured during 102 to 104 d during each of 2 yr in Hague, FL. ................................ .. 103

PAGE 10

10 LIST OF FIGURES page 3 1 Annual herbage accumulation from 14 rhizoma peanut entries cut once (1X) or twice (2X) per year in 2015, 2016, and 2017 at Quincy, FL. ................................ ..................... 52 3 2 Mean canopy height at ha rvest of 14 rhizoma peanut entries cut once or twice per yea r at Quincy, FL. ................................ ................................ ................................ ....................... 53 3 3 Herbage bulk density of 14 rhizoma peanut entr ies harvested once (1X) or twice (2X) per year at Quincy, FL. ................................ ................................ ................................ .......... 54 3 4 Mean herbage crude protein concentration of 14 rhizoma peanut entries cut once or twice per year for three years at Quincy, FL. ................................ ................................ ............. 54 3 5 Mean in vitro digestible organic matter concentr ation (IVDOM) of herbage from 14 rhizoma peanut entries cut once (1X) or twice (2X) per year for thr ee years at Quincy, FL. ................................ ................................ ................................ ................................ ..... 55 3 6 Root rhizome mass of 14 rhizoma peanut entries cut once or twi ce per year for three years at Quincy, FL. ................................ ................................ ................................ .......... 56 3 7 Mean rhizome diameter of 14 rhizoma peanut entries cut once or twice per year for three years at Quincy, FL. ................................ ................................ ................................ .......... 56 3 8 Mean root rhizome total nonstructural carbohydrate concentration (TNC) for 14 rhizoma peanut entries cut once or twice per year for three years at Quincy, FL. ......................... 57 3 9 Mean root rhizome total nonstructural carbohydrate (TNC) pool for 14 rhizoma peanut e ntries cut once or twice per year for three years at Quincy, FL. ................................ ..... 57 4 1 Annual herbage accumulat ion of 14 rhizoma peanut entries harvested twice per year in Quincy and Hague, FL during 2016 and 2017. ................................ ................................ 75 4 2 Canopy height of 14 rhizoma peanut entries harvested twice per year in Quincy and Hague, FL. ................................ ................................ ................................ ........................ 76 4 3 Herbage bulk density of 14 rhizoma peanut entries harvested twice per year in Quincy and Hague, FL. ................................ ................................ ................................ .................. 77 4 4 Her bage crude protein concentration of 14 rhizoma peanut entries harvested twice per year in Quincy and Hague, FL. ................................ ................................ ......................... 78

PAGE 11

11 4 5 Herbage in vitro digestible organic matter (IVDOM) concentration for 14 r hizoma peanut entries harvested two times per year in Quincy and Hague, FL. ................................ ...... 79 4 6 Root rhizome mass of 14 rhizoma peanut entries harvested two times per year in Quincy and Hague, FL. ................................ ................................ ................................ .................. 80 4 7 Rhizome diameter of 14 rhizoma peanut entries harvested by clipping two times per year in Quincy and Hague, FL. ................................ ................................ ................................ 81 4 8 Root rhizo me total nonstructural carbohydrate (TNC) concentration for 14 rhizoma peanut entries harvested for hay twice per year in Quincy and Hague, FL. ..................... 82 4 9 Root rhizome total nonstructural carbohyd rate (TNC) pool for 14 rhizoma peanut entries harvested twice per year for hay in Quincy and Hague, FL.. ................................ ........... 83 5 1 A completed root rhizome ingrowth core ................................ ................................ ......... 94 5 2 Cutting the cage wire material to size and shaping it into a cylinder for the ingrowth core. ................................ ................................ ................................ ................................ ........... 95 5 3 Bottom and top of the ingrowth core before the polyester mesh i s placed on it. .............. 96 5 4 Preparing the mesh covering of the ingrowth core. ................................ .......................... 97 5 5 Sewing the mesh covering of the ingrowth core wi th exterior grade thread. ................... 98 5 6 Finished ingrowth core. ................................ ................................ ................................ .... 99 5 7 Excavating soil for ingrowth core installation in a grass pas tur e. ................................ .. 100 5 8 Sieving and screening plant material (leaves, roots, rhizomes) from excavations. ........ 101 5 9 Replacing ingrowth core with soil screened of roots and rhizomes into plots. .............. 101 5 10 Cutting roots and rhizomes protruding from the ingrowth core. ................................ .... 101 5 11 Washing soil from ingrowth core to obtain root rhizome mass ................................ ..... 102

PAGE 12

12 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science DEFOLIATI ON FREQUENCY AND LOC ATION EFFECTS ON ROO T RHIZOME MASS, HERBAGE ACCUMULATION AND CANOPY CHARACT ERISTICS OF RHIZOMA PEANUT ENTRIES DIFFERING IN GROWTH HABIT By Katie Cooley December 2018 Chair: Lynn E. Sollenberger Cochair: Ann Blount Major: Agronomy Grasslands occupy nearly 40% of USA land area and, in addition to supplying feed for livestock, provide valuable services to society including storing carbon in soil, minimizing soil erosion, purifying water by removing excess nu trients, and providing wild life habitat. Most grasslands in the southeast ern US are comprised of warm climate grass species that are dependent on nitrogen fertilizer and have relatively low nutritional value for livestock Inclusion of legumes in mixture w ith grasses can increase nu tritive value and benefit nutrient cycling and soil quality Rhizoma peanut (RP ; Arachis glabrata Benth. ) is a long lived legume in the US Gulf Coast Region, providing an array of ecosystem services Most current RP cultivars wer e selected for upright grow th and hay production, but newer introductions and selections exist that are lower growing with potential for grazing or ornamental use Data describing key forage responses are lacking for these entries, and evaluation of produc tivity, nutritive value, an d persistence is needed in different soil environments and under different cutting regimes Also needed is a better understanding of root rhizome accumulation, which is a large contributor to

PAGE 13

13 soil carbon storage. One study compar ed the response of 14 RP en tries over three years to two d efoliation frequency t reatments including one harvest per year at season end vs. two harvests per year occurring at mid season and end of the growing season A single defoliation event per year res ult ed in significant reduct ions in annual herbage accumulation (HA) and nutritive value for most RP entries relative to harvests at both the mid dle and end of the growing season. Herbage accumulation was not affected by defoliation frequency for the entry Quincy Beta, perhaps because of its disease tolerance G reater defoliation frequency decreased root rhizome mass and non structural carbohydrate po ol a response of particular importance to producers us ing RP fields both for forage and as a source of rhizo mes for planting material. A second study compared the 14 RP entries growing for two years in a well drained soil at Quincy, Florida and in a seaso nally flooded soil at Hague, Florida. Well drain ed soil at Quincy w as more favorable to RP entries in general than seasonally saturated soil at Hague, which led to many entries having greater annual HA at Quincy than Hague. The entry Ona 33 had nearly the same HA at both locations and it is considerably better adapted to seasonally saturated soils than all RP en tries currently being used commercially. A third study reports on the design and use of a modified ingrowth core for measuring root accumulation ra te Increasing the mesh size of the fabric from 2 to 4 mm and using a wire frame cage to support the core all owed it to maintain its geometry over 100 d deployment periods and detect differences among RP entries. Finally, r esults of th ese stud ies suggest t here is opportunity to select new RP cultivars that are adapted to wetter soils and are well suited for grazi ng and ornamental uses while maintaining high levels of forage production, nutritive value, and below ground biomass.

PAGE 14

14 CHAPTER 1 INTRODUCTION Successful f orage production is challenging and requires thoughtful planning that integrates environmental and ec onomic variables including weather patterns soil characteristics, minimizing cost of inputs, and selection of adapted species and cultivars Increasing input cos ts and more erratic weather conditions associated with climate change can thwart planning, mak ing achievement of production goals more difficult. One strategy to address these challenges is to increase plant species diversity in pastures and hayfields (Tracy et al., 2018) T he presence of legume species in perennial grass lands may help to buffer en vironmental and economic fluxes through their contributions including N fixation which decreases dependence on N fertiliz er, greater forage nutritive value, and potentially greater animal production (Muir et al., 2011) Legumes contribute significantly t o ecosystem services provided by grasslands (Jensen et al., 2012). In addition to forage and animal products, important se rvices from legumes include enhanced rates of soil C sequestration (Jensen et al., 2012) and N fixation (Dubeux et al., 2017) Carbon sequestration is a critical ecosystem service W ell maintained grasslands occupy a large area and likely have a positive impact on the global C cycle (Scurlock and Hall 1998). Although much is known about grass dominated areas more studies are needed on legume based systems as C sink s in the Southeast US. Plants with large allocation of biomass t o below ground organs have significant potential to contribute to soil C (Rasse et al., 2005) W ell established stands of rhizoma peanut (RP; Arachis glabrata Ben th.) may have root rhizome mass of more than 20 Mg ha 1 ( Shepard et al., 201 8 ; Dubeux et al., 2017 ), so they are candidates to accumulate significant amounts of soil C. Variation among experimental lines in below ground biomass and root

PAGE 15

15 accumulation rate w ithin a species may be important to consider because of their potential impact o n soi l C accumulation Nitrogen fixation is a nother critical ecosystem servi ce provided by legumes owing to associated reduc tion in inorganic N inputs that results in sav ing s on fertilizer costs and reduc ed potential for environmental degradation In Florida, limited legume participation in pastures minimizes N fixation, but even a relatively small proportion of legume in grass based pastures may contribute significant amounts of N (Santos et al. 201 8 ) N itrogen fertilizer cost has ranged from approximately $0. 65 to $0.90 kg 1 N in recent years (Solle n berger personal communication) thus the savings from biologically fixed N are significant to producers Rhizoma peanut growin g in monoculture can fix 120 to 280 kg N ha 1 yr 1 (Dubeux et al., 201 7) Along with creating economic value for agriculture lessening the need for fertilizer has an environmental benefit I t reduces the likelihood of over application of fertilizer and su bsequent eutrophication of waterbodies from nutrient runoff or leaching In addition, l arge peaks in N 2 O fluxes can occur from N fertilized pastures following fertilization and cumulative N 2 O losses from heavily N fertilized grasslands can be up to four f old greater than from unfertilized legume grass pastures (Soussana et al., 2010; Klumpp et al., 2011). Of perennial forage legumes proposed for use in Florida, RP is perhaps the best option nted p ersistence in association with grasses ( Ortega S et al., 1992b; Mullenix et al., 2016a), and the herbage has excellent nutritive (Quesenberry et al., 2010). The Un iversity of Florida has released several RP genotypes in the past decade (Prine et al., 2010; Quesenberry et al., 2010) that are superior to Florigraze. Efforts

PAGE 16

16 continue to identify other superior RP lines, and currently there are several RP experime ntal l ines being studied that have potential for use in Florida forage livestock systems. Data describing key forage responses are lacking for these lines, and evaluation of productivity, nutritive value, and persistence are needed in different environment s and under different defoliation practices Because of potential benefit s to livestock production and delivery of regulating and supporting ecosystem services from using RP as a forage a dditional work is needed to describe the various RP experimental li nes in terms of their forage responses and potential contribution to ecosystem services. The research described in this thesis addresses t he themes of herbage and below ground biomass characteristics of numerous RP plant introductions growing at different locati ons under different harvest management practices The objective s we re to 1) quantify and compare the effects of defoliation frequency on herbage production, nutritive value, and root rhizome traits of selected RP introductions with those of existing cultiv ars and germplasms (Chapter 3) 2) characterize herbage production, nutritive value, and root rhizome traits of selected RP introductions with those of existing cultivars and germplasms when grown at two locations with markedly different soi l characteristi cs (Chapter 4) and 3) develop and test a prototype root ingrowth core device designed to quantify root accumulation rate of a group of RP entries ( Chapter 5).

PAGE 17

17 CHAPTER 2 LITERATURE REVIEW Use of Rhizoma Peanut in Florida Several factors h ave stimulated interest in rhizoma peanut (RP) and framed RP research efforts Rhizoma peanut herbage in vitro digestibility is compar able to alfalfa ( Medicago sativa L.) (Terrill et al., 1996) and RP is very persisten t and can spread in grass pastures under good manage ment ( Dunavin, 1992; Ortega S et al., 1992 b ) R h to fix atmospheric N ranging from 120 to 280 kg N ha 1 yr 1 (Dubeux et al., 2017), provides cost saving s potential and reduction in e nvironmental impact compared with N fertilizer use E stablishment is the biggest challenge to more widespread RP use because t he plant produces few viable seed s and must be propagated vegetatively With resultant high establishment costs and relatively low economic return per hectare when used for livesto ck grazing compared with hay production producers are wary to use it as a pasture forage Thus, RP is used primarily for hay production (Mullenix et al., 2014). T he factors limiting RP use have stimulated research; m any studies ha ve been conducted on impr oved establishment methods (Castillo et al., 2013a, 2013b, 2014; Mullenix et al., 2014) Another theme has been evaluation of germplasm under different management practices ( Hernndez Garay et al., 2004; Mullenix et al., 2016a, 2016b) In addition, t he int eraction of growth habit with management practices as it affects RP establishment, persistence and herbage accumulation has been a major topic in RP research (M ullenix et al., 2014, 2016a, 2016b ; Shepard et al., 2018 )

PAGE 18

18 Rhizoma Peanut Introduction and Av ailable Cultivars Rhizoma peanut was first introduced in to the US in 1936 when a germplasm of A. glabrata was brought to t he USDA from a collection in Ma to Grosso, Brazil (Quesenberry et al., 2010) Further collections continued through the 1960s and 1980s prompting research on the forage potential of RP in Florida (Prine et al., 2010). As the culmination of that effort, t he first RP cultivar release d was Fl origraze (PI 421707) Currently, t here are two germplasms (Ecoturf and Arblick ; Prine et al., 2010 ) and five commercially available cultivars ( Florigraze, Arbrook UF Tito UF Peace and Latitude 34 ) in the US The most widely used cultivar is Florigraze but Florigraze herbage accumulation is negatively affected by the peanut st u nt virus ( Cucum ovirus spp.; Prine et al., 2010) As such there is a clear need for a more diverse genetic base among RP cultivars (Blo unt et al., 2006) Arbrook (PI 262817) was released in 1986. It was developed from germplasm collections from Paraguay in 1960 and is we ll adapted to the drought prone sand y soil s of Florida (Prine et al., 2010) Its upright habit is favorable for hay prod uction, but it has a low tolerance to overgrazing ( Hernndez Garay et al., 2004 ; Prine et al., 2010 ) Arblick and Ecoturf germplasms wer e selected from RP accessions in the 1950s Both are low growing and adapted to inland Florida and the s outheastern US C oastal P lain Ecoturf has shown rapid establishment (Prine et al., 2010) and tolerance to grazing (Mullenix et al., 2016a S hepard et al ., 2018 ) N ew RP cultivars UF Tito (PI 262826) and UF Peace (PI 658214) were plant introductions from Paraguay in the 1950s and were released in 2008 Both have a slightly more upright habit than Florigraze and high herbage accumulation ( HA; Quesenberry et al., 2010). Except for Latitude 34, all of these RP cultivars and germplasms were released from the University of Florida Latitude 34 was selected for early

PAGE 19

19 spring production and persistence under dry, cool climatic conditions in Texas (Muir et al., 201 1 ). Rhizoma Peanut Responses to Defoliation Herbage Accumulation Herbage accumulation is described as the difference between post grazing herbage mass and the pre grazing herbage mass of th e subsequent grazing event or simply it is the change in herbage m ass from immediately following hay ing or grazing u ntil immediately prior to the next defoliation Rhizoma peanut HA was 8.3 to 12 Mg ha 1 yr 1 during a 4 yr clipping trial in north central FL (Prine et al., 2010), 10.8 Mg ha 1 yr 1 across 4 yr under clippi ng in south Florida (Mislevy et al., 2007), up to 13 Mg ha 1 yr 1 in northwest Florida (Dubeux et al., 2017), and 6.0 to 9.2 Mg ha 1 yr 1 during 2 yr of grazing in Gainesville (Mullenix et al., 2016a). Mullenix et al. (2016a) also found that HA under grazi ng was similar across Florigraze UF Peace, UF Tito, and Ecoturf. Grazing intensity and frequency can affect HA Ortega S et al. ( 1992b) found that Florigraze had 35% lower HA in Year 1 when grazed frequently to a short stubble compared with treatments i ncluding longer regrowth interva ls and greater residual biomass after grazing. By Year 2 the decline in HA was 70% for frequently and closely grazed treatments They attain high levels of light int erception before subsequent defoliation explains in part the low rhizoma peanut herbage accumulation observed. ( Ortega S et al. 1992b). Ecoturf responded differently to grazing (Shepard et al., 2018). When grazed frequently and closely, it assumed a prost rate growth habit with high herbage bulk density and post grazing leaf mass. This allowed the plants

PAGE 20

20 to reduce dependence o n reserves for regrowth and maintain high levels of HA, even when grazed weekly to a 4 cm stubble. Wh ile the effect of defoliation on HA varies, there is a point at which HA decreases with increasing defoliation frequency. In Florida, an experiment evaluat ed the effects of cutting frequency on Florigraze RP herbage accumulation during 24 wk each year for 2 yr (Beltranena et al., 1981) Plots were cut at 2 4 6 8 10 and 12 wk intervals. Herbage accumulation increased with increasing interval between defoliation events through 6 wk in Year 1 and 8 wk in Year 2 after which HA remained the same or de creased. I n Australia 11 accessions of RP (some were Arachis glabrata and others Arachis pintoi Krapov. & W.C. Greg. ) were harvested at two different frequencies (2 times yr 1 or 3 4 times yr 1 ) for 2 yr (Bowman et al., 1998). They reported that the two h arvests yr 1 treatmen t had greater herbage HA of approximately 1 Mg ha 1 yr 1 than the more frequent clipping treatment. Similar observations were made in an experiment evaluating HA and nutritive value of Florigraze RP harvested at 6, 9, or 12 wk during s ummer and fall in Flo rida (Romero et al., 1987) As interval between summer harvest s increased from 6 to 12 wk, RP HA inc reased from 3.1 to 8.2 Mg ha 1 The authors suggest ed that RP accumulation was l e a s t in summer with the most frequent harvest interval because 6 wk did not allow time for the pl ant to recover from defoliation However, i n the fall, HA increased only through 9 wk, after which it d ecreased It has been suggest ed that excess herbage, as the result of either tall stubble height or low stockin g rate can decrease HA (Adjei et al., 1980; Sanchez et al., 2018) In an experiment in South Florida Arachis pintoi Krapov. & W.C. Greg.) into Jiggs bermudagrass [ Cynodon dactylon (L. ) Pers.] was evaluated for pastures grazed at 15 and

PAGE 21

21 25 cm stubble heights in 2 yr (Sanchez et al., 2018). In pastures grazed to the taller stubble height, there was 50% slower HA rate. The authors attribute d this response to excess herbage, which can cau se self shading, reduced photosynth esis, and accumulation of senescent material. This is supported by a study evaluating the HA and nutritive value response of stargrass ( Cynodon sp.) digitgrass ( Digi taria eriantha Ste ud ), and bahiagrass ( Paspalum notatum Flugge) to stocking rate (Adjei e t al., 1980). There was a positive relationship between stocking rate HA, and herbage in vitro digestible organic matter (IVDOM) concentration The authors a ttribute d this response to greater amounts of no n photosynthetic residue remaining in the low inte nsity treatment and the subsequent increase in self shading and reduction in photosynthesis. Although these two experiments d id not evaluate defoliation frequency, long defoliation intervals typically resul t in excess herbage, similar to what would occur w ith tall stubble heights and low stocking rates. So, it may be possible that the same physiologi cal mechanisms (self shading, reduced photosynthesis and non photosynthetic residue ) affect the HA of pasture s managed with long intervals between defoliation events, low stocking rate, and tall stubble height s Below ground Biomass Persistence and Vegetative Propagation Well established stands of RP may have root rhizome mass of more than 20 Mg ha 1 ( Dubeux et al., 2017 ; Shepard et al., 2018 ), and this has be en associated with excellent persistence (Mullenix et al., 2016b) Species persistence under different defoliation management practices is often associated with changes in storage organ mass (Sollenberger et al., 2012). Ortega S et al. (1992 a ) found that Florigraze RP proportion in pre grazing herbage mass decreased from ~ 90 to 65% after one growing season of close, frequent grazing and to 30% after two growing seasons Correspondingly, r hizo me mass of Florigraze decreased from 4.0 to about

PAGE 22

22 0.5 Mg ha 1 du ring the same time period ( Ortega S et al., 1992b). In contrast to Florigraze, when E coturf RP was grazed weekly to 4 cm over two growing seasons root rhizome mass remained the same or increased and proportion of RP in pre grazing herbage mass stayed the same or decreased no more than four percentage units (Shepard et al., 2018) In anoth er grazing study, Ecoturf had greater root rhizome mass than Florigraze under a range of grazing treatments (4450 and 3490 kg ha 1 respectively; Mullenix et al., 2016b), and the authors suggested that greater residual leaf area of Ecoturf after grazing al lowed it to be less dependent on stored energy and to preserve root rhizome mass. Thus, root rhizome biomass and changes in biomass over time are useful indicators of the vigor and potential persistence of the sward. After defoliation, root rhizome carboh ydrate reserves are mobilize d to support above ground growth subsequently decreasing root rhizome mass, total non structural carbohydrate ( TNC ) concentration and TNC po ol ( Saldivar et al., 1992a; 1992b ) This relationship was described in a clipping study evaluating effects of defoliation frequency (undefoliated and defoliated at 1 4 and 8 wk intervals) o f Florigraze RP on above and below ground biomass (Williams 1 994). Starting in early June, r oot rhizome mass of defoliated treatments dec reas ed during the first 8 wk after defoliation, but they recovered to pre defoliation levels by the latter part of the growing season By the end of the season, below ground biomas s of the defoliated plots was 32% l ess than the non defoliated pl ots. Similarly, as grazing intensity increased, Florigraze root rhizome mass and TNC concentration decreased markedly during 2 yr (Rice et al., 1995). C arbohydrate reserves of RP affect past ure establishment as well as stand persistence (Rice et al., 1996). Thus the relationship between harvest frequency and root rhizome

PAGE 23

23 carbohydrate storage is critical to producers managing RP defoliation in order to produce rhizomes for use as planting mat erial. In an experiment studying t he effect of grazing intensity on establishment performance of Florigraze RP rhizome s it was found that rhizomes from leniently grazed stands had more shoot production and a faster rate of herbage accumulation than pastur es grazed mor e intensively (Rice e t al., 199 6 ). The authors suggested this was because severe grazing drains rhizome carbohydrate reserves. Nutritive Value Nutritive value is the digestibility, chemical composition, and nature of digested products of a for age In general, legumes have greater nutritive value than warm season grasses Sto bbs et al. (1975) found that when legumes we re present in pastures mixed with C 4 grasses, ruminants increase d intake, resulting in greater live weight gain. Legume s are valu abl e to ruminant nutrition because they have high levels of crude protein (CP) (Muir et al. 2011 ). L egumes may increase intake and animal performance given their overall greater digestibility and CP during the growing season, compared with warm season gra sse s (Muir et al., 2011) If grass diets provide insufficient levels of N and energy to ruminants, legume addition to diet s can increase N retention by livestock (Foster et al., 2009). For Florigraze, UF Tito and UF Peace cultivars and Ecoturf germplasm the effect of defoliation management on herbage CP and IVDOM concentration s was relatively small (Mullenix et al., 2016 a ). They found a 1 1 In a separate study with Ecoturf RP, herbage CP and IVDOM varied l ittle among grazing treatments, averaging 181 and 698 g kg 1 respectively (Shepard et al., 2018) The authors attributed this lack of response in part to the relatively decumbent growth habit of Ecoturf and

PAGE 24

24 suggested that this trait offered wide flexibil i ty in timing of grazing or harvesting this forage. Saldivar et al. (1990) drew a similar conclusion, observing that the greater leaf:stem ratio of Florigraze RP minimized the typically negative effect of increasing maturity on nutritive value. While the e ffect of harvest frequency on nutritive value of RP may not be as dramatic as for many other species there is generally a negative relationship between defoliation frequency and nutritive value. T he effects of cutting frequency on Florigraze nutritive val ue was evaluated by Beltranena et al. ( 1981). Nutritive value, as measured by CP and IVDOM, was greatest in the most frequently cut samples (2 wk) and decreased as cutting interval increased. I n a multi location study in Louisiana evaluating t he effect of harvest frequency on nutritive value, Florigraze RP was harvested every 30 or 60 d (4 or 2 times yr 1 respectively) (Redfearn et al., 2001). In this experiment, as in the others, the greatest herbage CP was observed in the most frequent defol iation treatm ent. Grasslands and Soil C Sequestration Increasing atmospheric C O 2 levels and their associated impacts on global climate have stimulated research in soil C sequestration of grasslands. G rassland soil can serve as a C sink by storing CO 2 that was removed from the atmosphere during photosynthesis Fertilization, organic amendments, tillage, crop selection and crop rotation affect the quality, quantity, and placement of C in the soil (Magdoff and Weill 2004). The soil organic C (SOC) pool is vulnerable to disruption by agricultural practices especially tillage Organic matter that is exposed and oxidized will deplete the SOC pool, diminishing soil quality and subsequent plant biomass productivity (Follet t 2001). Soil texture, total soil nitrogen (TSN), pl ant species, and environmental factors can affect SOC.

PAGE 25

25 M anagement practices that increase forage production also increase soil C (Conant et al. 2001; Allard et al., 2007; Ammon et al., 2007). Likewise, p oor management of forage s can inhibit C accumulation in grasslands Although Florida soils are inherently low in characteristics that build and protect SOC (Hassink, 1997), there is potential to increase SOC by improving management practices (Conant 2001). The e ffect of manag ement on rate of change of SOC and T S N at various soil depths of a oastal bermudagrass pasture was studied by Franzluebbers et al. ( 2009 ). Treatments included low (5.8 steers ha 1 ) and high (8.7 steers ha 1 ) stocking rates of grazed swards and unharvested and monthly hayed be rmudagrass Significantly more C accumulated in the upper 15 c m of soil in grazed as compared with hayed pastures with the greatest annual rate of change of SOC occurring in the low stocking rate treatm ent (1.17 M g C ha 1 yr 1 ). Stocking rates were also e valuated in a 26 yr study o f the ir effects o n soil C and N in bermudagrass pa stures in Texas (Wright et al., 2004). The authors found that l ow stocking rate (2 to 2.5 cow calf pairs ha 1 ) increased C and N more than high stocking rate The high stocking ra te (5 to 7.4 cow calf pairs ha 1 ) physically disturb ed the soil to a greater extent and resulted in more rapid turnover of plant residues and greater amounts of nutrient s cycling in animal excreta than t he low er stocking rate. The effect s of defoliation an d fertilizer input on C sequestration over 3 and 6 yr was evaluated in two studies by Ammann et al. (2007; 2009). In the 2007 study, the site was converted from crop rotation to grass clover mixtures and included two management regimes; an intensive regime where grass was frequently cut and N fert i lizer was applied at a rate o f 200 kg ha 1 yr 1 a nd an extensive regi me where there was no N input and cutting was less frequent. The extensive treatment lost a net of 57 g C m 2 yr 1 while the intensive treatme nt sequestered 147 g

PAGE 26

26 C m 2 yr 1 over the course of the 3 yr study. In the six yr study, the management treatments continued, and the results were largely similar with more net C sequestered with the intensive treatment due to greater plant productivity as compared with the extensive treatment ( Ammann et al. 2009). Carbon sequestration potential of grassland s may al so be affected by inclusion of legumes or by species richness or functional type Legumes may enhance C sequestration in pastures because of th eir ability to increase soil N through N fixation The decay resistant, humic N polymers formed by N inhibit decomposition of humified soil C, increasing SOC (Fog 1998; Resh et al 2002). Reviewing several studies, Jensen et al. (2012) observed that inco rporating perennial legumes into pastures increase s SOC, yet there are few st udies on this. When seeded in mixtures in Minnesota, Fornara and Tilman (2008) observed the N provided by legumes (four species tested) increased SOC by 100% over 12 yr and also i ncrease d root biomass of associated C 4 species (four species tested) Cong et al. (201 4) observed that N fixation of legumes increased the rate of C sequestration in temperate grasslands where productivity wa s limited by low N. In a study on the effects of haying on soil C pools within plots of various species richness, only when the legumes birdsfoot trefoil ( L otus corniculatus L. ) and white clover ( T rifolium repens L. ) were present did soil C and N accumulation increase relativ e to treatments including no legumes (DeDeyn et al., 2009) N either the number of species nor functional group richness change d either C or N levels. In a continuation of that study, DeDeyn et al. (2011) further observed the presence of the legume red clov er ( T. pr a tense L. ) improv ed soil C and N accumulation compared with treatments without red clover The authors attribute d this to the fact that high rates of C and N additions reduced soil respiration, improved soil structure and increased soil

PAGE 27

27 organic ma tter (SOM) con centration Ho wever, fine root production of legumes and its relationships with below ground plant biomass accumulation and SOC content are not well understood. Planted w arm season perennial grasses ( C 4 photosynthetic pathway) may also enhance grassland C seque stration potential compared with native grasslands Conant et al. (2001) observed that t heir extensive root system and permanent vegetative cover are conducive to SOC accumulation The effects of stocking rates and N fertilization rates on SOM were evaluated in a bahiagrass pasture in a 3 yr study ( Dubeux et al. 2006). The C and N concentration s in the SOM light density fraction were observed to increase as management intensity increased (Dubeux et al. 2006). C ool season grasses ( C 3 photosynthetic pathway) such as small grains, can maintain or improve SOC when planted as cover crops in crop rotation systems compared with the row crop alone without a cover crop Franzluebbers and Stuedemann (2008) observed that total particulate organic matter remained relatively constant over time (2.3 Mg ha 1 yr 1 ) with summer cropping of sorghum ( Sorghum bicolor ) followed by winter grazing of rye ( Secale cereal e L. ). This integrated crop livestock study w as conducted over 3 yr in Georgia C over crop rotations incorporating winte r forages could span up to 51 million ha in the US, potentially sequester ing 100 to 300 kg C ha 1 yr 1 (Lal et al., 1999). With mild winters in Florida, incorporation of winter fora ges into year round forage production systems may play a role in maintain in g or increas ing soil C in grasslands.

PAGE 28

28 Relationship of Below Ground Biomass and Root Growth with Soil C Accumulation in Grasslands Plants with large allocation of biomass below ground have significant potential to contribute to soil C (Rasse et al., 2005), but below ground processes that drive SOC accumulation are not well understood Root dynamics is considered to be a critical component of the C cycle in grasslands, but i t is difficult to measure or predict and can vary temporally and among and within spec ies. Bradford et al ( 2013 ) observed that stable SOC pools may be en hanced by fine root abundance which transfer C to the rhizosphere through exudation of amino acids, su gars, and polysaccharides. Phillips et al. (2011) show ed that transfer of C through fi ne root exudates from live and decaying roots can dramatically impact nutrient cycling and SOC pools. Within a grassland context, fine root decay and root exudates are li kely to regulate productivity and SOC accumulation and they are affected by defoliatio n ( Augustine et al ., 2011 ; Hafner et al., 2012 ). However, a deeper understanding of the relationship between below ground biomass dynamics with soil C accumulation in grasslands is needed but in order to establish these relationships better methods of mea suring root mass accumulation are needed The objectives of the studies reported in this thesis were to 1) quantify and compare the effects of defoliation frequency on herbage production, nutritive value, and root rhizome traits of selected RP introductio ns with those of existing cultivars and germplasms (Chapter 3), 2) characterize herbage production, nutritive value, and root rhizome traits of selected RP introducti ons with those of existing cultivars and germplasms when grown at two locations with marke dly different soil characteristics (Chapter 4), and 3) develop and test a prototype root

PAGE 29

29 ingrowth core device designed to quantify root mass accumulation rate of a gr oup of RP entries (Chapter 5).

PAGE 30

30 CHAPTER 3 RHIZOMA PEANUT HERBAGE ACCUMULATION, NUTRITIVE VALUE, AND ROOT RHIZOME RESPONSES TO DEFOLIATION FREQUENCY Overview Perennial legumes can contribute significantly to forage systems. Their generally high nutritive v alue and ability to fix N are valuable characteristics to producers and can reduce the pot ential for negative impacts on the environment (Muir et al., 2011; Jensen et al., 2012). Of perennial legumes proposed for use in pastures in Florida, rhizoma peanut (RP; Arachis glabrata Benth.) is perhaps the best option because it is well adapted to Flo persistence under grazing ( Ortega S et al., 1992b; Hernandez Garay et al., 2004; Shepard et al., 2018), can spread in mixtures with gr asses (Castillo et al., 2013a, 2013b; Mullenix et al., 2014), and has excellent nutritive value (Mullenix et al., 2016b; Shepard et al., 2018). the peanut stunt virus (Quesenberry et al., 2010). The University of Florida has released several forage RP genotypes in the past decade (Prine et al., 2010; Quesenberry et al., 2010) that are superior to Florigraze. Most released RP lines were selected for upright g rowth, favoring use in hay production, but there are germplasms and experimental lines that vary in growth habit, which may affect their optimal forage use. For example, tall nce, persistence, and herbage accumulation (HA) yield under hay management. Lower growing germpla sms Ecoturf and Arblick may be better suited to grazing or ornamental use (Prine et al., 2010). Ecoturf is productive and persistent under close, frequent graz ing (weekly to 4 cm) due to its ability to increase sward canopy bulk density and maintain residu al leaf area close to the soil surface (Shepard et al., 2018). Efforts continue to

PAGE 31

31 identify superior RP lines, and currently there are numerous RP plant introd uctions and selections with potential for use in Florida forage livestock systems. Data describin g key forage responses are lacking for these entries, and evaluation is needed of forage productivity, nutritive value, and stand persistence. Defoliation man agement is critical to the successful use of any forage, and RP is no exception. Although defolia tion intensity is the most important determinant of plant and animal response (Sollenberger et al., 2012), in RP hay systems cutting height varies relatively l ittle. Thus, response to defoliation frequency is important and may affect HA, nutritive value, a nd rhizome mass and chemical composition. Rhizoma peanut HA increased with increasing interval between grazing events up to 63 d for Florigraze (Ortega S. et a l., 1992b) and 42 d for Florigraze, UF Tito, UF Peace, and Ecoturf (Mullenix et al., 2016a). Nutr itive value was minimally affected by defoliation frequency (Mullenix et al., 2016a; Shepard et al., 2018), but frequent, close defoliation reduced rhizome mas s and total non structural carbohydrate (TNC) concentration of Florigraze (Ortega S. et al., 1992 a; Rice et al., 1995) and rhizome mass of several RP entries (Mullenix et al., 2016b). Response of Ecoturf, a phenotypically plastic germplasm that adapts to c lose grazing by developing a short, dense canopy, was different than Florigraze, as it maintained or increased herbage HA and rhizome mass under frequent, close grazing (Shepard et al., 2018). Previous research has shown that RP responses to defoliation f requency differ among entries of varying growth habit and that both above and below ground plant responses are affected (Mullenix et al., 2016a). Thus, both herbage characteristics, important to hay producers, and rhizome characteristics, important to tho se who vegetatively propagate RP and for long term

PAGE 32

32 stand persistence, are affected by defoliation frequency and growth characteristics of individual entries. The objective of this study was to quantify and compare the effects of defoliation frequency on he rbage production, nutritive value, and root rhizome traits of selected RP introductions with thos e of existing cultivars and germplasms. Relatively long intervals between defoliation events were evaluated in order to address two primary scenarios: i) the d ifficulty of timely hay harvest in Florida because of frequent rain events during the summer grow ing season; and ii) situations in which RP is harvested for hay during the growing season and also used as a source of rhizomes for vegetative propagation in t he subsequent dormant season. M aterials and Methods Experimental Sites The 3 yr experiment yr (2 015, 2016, and 2017) was conducted at the North Florida Research and Education Center (NFREC) in Quincy, FL (30.55 N, 84.60 W) on well established plots that were planted in July 2009. Prior to planting the plots for the current experiment, the area was occupied by bahiagrass ( Paspalum notatum Flugge). The soil at the site is a Norfolk loamy fine sand (fine loamy, kaolinitic, thermic Typic Kandiudult s ), described as a well drained soil. Soil samples were taken to a depth of 15 cm and were analyzed at the University of Florida Extension Soil Testing Laboratory. Soil pH was 6.0 and Mehlich 3 extractable P, K, and Mg were 38, 28, and 58 mg kg 1 respectively In April 2017, K was applied at a rate of 56 kg ha 1 and P was applied at a rate of 15 kg ha 1 No fe rtilizer was applied in 2015 and 2016.

PAGE 33

33 Treatments and Experimental Design Rainfall data for the site during the experime ntal period is shown in Table 3 1 Treatments were the factorial combinations of two defoliation frequencies and 14 RP introductions/sel ections, ger mplasms, and cultivars (Table 3 2 ), here forward referred to as entries. The design was a split plot arrangement of a randomized complete blo ck design, with five replications. Main plots were 2.5 m wide x 3 m long (7.5 m 2 ), with a 1.8 m alley b etween main plots. The entry main plots were split into two subplots (each 2.5 m x 1.5 m) to which defoliation frequency levels of one (1X; fall) or two (2X; summer and fall) harvests per season were allocated. Response Variables P eanut rust, herbage accu mulation, nutritive value, and canopy characteristics Harvests for the 2X treatment occurred on 26 July and 23 Nov. 2015, 14 July and 7 Nov. 2016, and 27 June and 20 Oct. 2017. The 1X treatment was cut only at the October or November date already mentioned for the 2X treatment. Harvest dates were chosen based on typical conditions in North Florida, where spring drought precludes significant growth of rain fed RP before June, and in the absence of irrigation most hay producers get only two harvests per year. The 1X treatment reflects that summer harvests are challenging due to frequent and unpredictable rainfall events and poor drying conditions. Thus, some producers in some years fail to harvest RP in summer due to lack of sufficient consecutive days of sati sfactory drying conditions. In addition, some hay producers also dig rhizomes from hay fields for use in planting new areas to RP. They are interested in knowing the impact of multiple vs. single harvests per

PAGE 34

34 year on root rhizome mass and TNC reserves, fac tors that affect the amount of area that can be planted and the likelihood of establishment success (Rice et al., 1995; 1996). At each harvest date, her bage from one (2015 and 2016) or two (2017) 0.25 m 2 quadrats was harvested to a 6 cm stubble height fro m each experimental unit. The remainder of the subplot was clipped with a flail mower to the target stubble height and all herbage removed. Samples were dried at 60C until constant weight, weighed, and HA was expressed as M g DM ha 1 yr 1 These samples we re ground to pass a 1 mm screen using a Wiley mill (Model 4 Thomas Wiley Laboratory Mill, Thomas Scientific, Swedeboro, NJ). Samples were then analyzed f or N and in vitro digestible organic matter (IVDOM) concentrations. Nitrogen was measured using the alu minum block digestion technique (Gallaher et al., 1975), and IVDOM was determined using a modification of the two stage technique (Moore and Mott, 1974). Crude protein (CP) was calculated as N concentration x 6.25. Considering the typically low coefficient of variation associated with measures of nutritive value and the cost of laboratory analyses, samples from only three of the five replicates were analyz ed for CP and IVDOM. Before each harvest event in 2017, canopy height was measured at 10 locations in e ach experimental unit to aid in quantifying differences in growth habit among entries and to allow calculation of herbage bulk density. Herbage bulk dens ity was calculated by dividing HA (kg ha 1 ) at a given harvest by the depth of the harvested canopy and expressed in kg ha 1 cm 1 Thus, if canopy height was 20 cm and the stubble height was 6 cm, bulk density was calculated by dividing HA by 14 cm. Peanut rust ( Puccinia arachidis ) disease was assesse d in each experimental unit on 27 June and 11 Oct. 2017. P ercent age of leaflet area infect ed by the disease was estimated visually, and the average infection in each plot was converted to the Horsfall

PAGE 35

35 Barratt (H B) scale of plant disease. The H B scale is a visual assessment of plant disease and is based on a se mi quantitative scale (Horsfall and Barratt 1945). Root rhizome mass and total non structural carbohydrate concentration and rhizome diameter Root rhizome mass was measured by taking three cores from randomly selected locations in the center of each expe rimental unit o n 13 Dec. 2016 and 12 Dec. 2017. Cores were 10 cm in diameter and 20 cm deep. The three samples per plot were composited and wash ed over a 2 mm mesh screen to remove soil. The sample was then dried at 60C to a constant weight and weighed. S amples were ashed at 500 o C for 4 h and mass was expressed on an organic matter basis to avoid potential effects of soil contamination. Mean r hiz ome diameter was quantified for each experimental unit. Four representative rhizomes were selected from each rhi zome sample and diameter was measured at the mid point of the length of the rhizome. Mean r hizome diameter was calculated as the average of the four measurements per plot. Root rhizome samples were ground in a Wiley mill to pass a 1 mm stainless steel scre en prior to analysis for TNC concentration. The TNC concentration was determined using a modification of the procedure of Christiansen et al. (1 988) that was described by Chaparro et al. (1996), and it was expressed as a proportion of root rhizome organic matter. This procedure uses amyloglucosidase and invertase to convert starch and oligosaccharides into monosaccharides and measures reducing sug ars with a photometric copper reduction method (Nelson, 1994). As described for herbage nutritive value, root rh izome samples from three of five replicates were analyzed for TNC concentration. The TNC pool was calculated as root rhizome mass times TNC conc entration.

PAGE 36

36 Statistical Analyses Data were analyzed using a mixed model with entry, defoliation frequency, and t heir interaction as fixed effects and block and year as random effects. Year was considered a repeated measure. Defoliation frequency means were separated using the F test, and RP entry means were ed different when P data are presented graphically in the form of box plots. The structure of box plots is such that the upper and lower hinges ( terminal ends of the box) relate to the first and third quartiles (or 25 th and 75 th percentiles), respectively. The thick solid line within the box and perpendicular to the sides of the box is the median. The upper and lower whiskers (solid l ines extending above and below the box) extend from the hinge to the largest and smallest values (at most 1.5 in terquartile range) of the hinge, respectively. Data points beyond the whiskers are considered outliers. Results and Discussion Peanut Rust, Herb age Accumulation and Sward Canopy Characteristics Peanut r ust i ncidence In the June sampling date, all lines ex cept Ecoturf showed no disease (H B rating of 1) (Table 3 3) Ecoturf 1X and 2X had a H B score of 4. All replicates of each entry had the same rating in June, so it was not possible to analyze those data statistically. In October, t here was no entry x def oliation frequency interaction for peanut rust disease ( P = 0.85), no r was there a defoliation effect ( P = 0.84). However, there was an effect of entry (Table 3 3; P < 0.0001). Ecoturf was moderately to severely affected (H B rating of 7) as over half its leaflets were diseased. Apalachee and UF Tito were moderately disease d (H B rating of 5) and Quincy Alpha

PAGE 37

37 was mildly diseased (H B rating of 4). Entries Quincy and Waxy Leaf had less than 3% of their leaves diseased (H B rating of 2), and their rating was not different than lines with no disease. Annual h erbage a ccumulation There was an entry x defoliation frequency interaction for HA ( P < 0.0001; Figure 3 1). The 2X treatment had greater HA than 1X for all entries ( P < 0.0001) except for Quincy Beta, for w hich there was no defoliation frequency effect. For many of the entri es (Quincy Beta being the primary exception), the 2X treatment had nearly twice the HA as 1X, as evidenced by defoliation treatment main effect means of 10.5 vs. 5.4 Mg ha 1 yr 1 respect ively. Among interaction means, HA of the 2X treatments of Arbrook (1 3.9 Mg ha 1 yr 1 ), UF Tito (12.9 Mg ha 1 yr 1 ), and UF Peace (12.3 Mg ha 1 yr 1 ) w ere greater than the 2X treatments of Arblick (9.33 Mg ha 1 yr 1 ), Cowboy (9.11 Mg ha 1 yr 1 ), Florigraze (8.43 Mg ha 1 yr 1 ), and Apalachee (7.68 Mg ha 1 yr 1 ). Least herbage accumulation was ~4 Mg ha 1 yr 1 for numerous entries harvested once per year, including 3.86 Mg ha 1 yr 1 for industry standard Florigraze. Among the 1X treatments, Quincy Beta had gre ater herbage accumulation (8.77 Mg ha 1 yr 1 ) than all entries except Arbrook (7.92 Mg ha 1 yr 1 ), Quincy (7.67 Mg ha 1 yr 1 ), UF Peace (6.53 Mg ha 1 yr 1 ), and Waxy Leaf (6.29 Mg ha 1 yr 1 ). Lack of difference between Quincy Beta with 2X and 1X frequencie s (10.4 vs. 8.77 Mg ha 1 yr 1 respectively) may be due to greater disease tolerance of Quincy Beta than other lines resulting in greater leaf retention for full season growth (Blount, personal communication). Herbage accumulation of Quincy Beta was greate r than that reported by Santos et al. (2017) in a 2 yr expe riment in which Quincy Beta (5.37 Mg ha 1 yr 1 ) and the germplasm Ecoturf RP were harvested by clipping four times per year. In that study, annual HA of Ecoturf was 5.00 Mg ha 1

PAGE 38

38 compared with 3.91 Mg ha 1 (1X treatment) and 10.5 Mg ha 1 (2X treatment) in the current experiment. Lesser annual HA for 1X in the current study can be explained in part due to moderate to severe levels of peanut rust infestation for Ecoturf that is thought to reduce effec tive leaf area and HA. Average annual HA for the 2X treatme nt falls within the range observed in previous experiments. Annual RP HA in the southeastern US was 7 to 11 Mg ha 1 (Terrill et al., 1996; Venuto et al., 1999). Reported RP HA in grazing or clippin g trials in Florida have included 13 Mg ha 1 yr 1 (Dubeux et al., 2017), 6.0 to 9.2 Mg ha 1 yr 1 (Mullenix et al., 2016), and 8.3 to 12 Mg ha 1 yr 1 (Prine et al., 2010). However, annual HA for the 1X tre atments was generally well below these means. The av erage of the two defoliation treatment means for Ecoturf, UF Tito, Quincy Alpha, and Apalachee, the four lines with significant peanut rust damage, were at the lower end of reported HA means. However, 2X means of Ecoturf and UF Tito are well within reporte d means. Previous studies have evaluated defoliation frequency effects on various RP entries. Six defoliation frequencies ranging from 2 to 12 wk were compared for Florigraze RP during 2 yr (Beltranena e t al., 1981). Herbage accumulation increased as inte rval between defoliation events increased up to 6 wk in the first year of study and up to 8 wk in the second year. The 12 wk interval treatment in that experiment was harvested twice per year, with an ave rage annual HA of 10.1 Mg ha 1 yr 1 comparable to t he response of the 2X treatment in the current study. When grazed, Florigraze HA increased as interval between grazing events increased if pastures were grazed to a residual herbage mass of 500 to 1500 kg ha 1 (Ortega S. et al., 1992). When residual herbag e mass was greater, there was little effect of defoliation frequency on HA. Averaged

PAGE 39

39 across four RP entries (UF Tito, UF Peace, Florigraze, and Ecoturf), HA was greater when pastures were grazed every 6 t han every 3 wk (Mullenix et al., 2016a). There is ev idence in the literature, however, that more frequent defoliation may result in greater HA of some RP entries under some conditions. For example, in a grazing experiment with Ecoturf, weekly close defolia tion (4 or 8 cm stubble) resulted in greater HA tha n defoliating every 4 or 7 wk (Shepard et al., 2018). Likewise, in a mowing experiment with Ecoturf, frequent, close mowing was associated with greater HA, more new shoots, and a dense canopy (Rouse et al ., 2004). Thus, the HA response to defoliation frequ ency is not uniform across RP entries. It appears that those with shorter growth habit or greater phenotypic plasticity are favored by more frequent defoliation, while more upright types are favored by le ss frequent defoliation. The response of Quincy Beta supports this conclusion, as it is among the shortest of the entries evaluated (described later in this chapter). However, in the current study both levels of defoliation frequency imposed are considered infrequent and the response to frequency was more l ikely related to greater disease incidence in the 1X treatment than to physiological mechanisms that are at play when plants are defoliated more frequently. Canopy h eight There was no entry x defoliation frequency interaction ( P = 0.071) or defoliation fr equency main effect ( P = 0.15), but there was an effect of entry ( P < 0.001) on canopy height (Figure 3 2). Arbrook was taller (29) than all other entries except UF Tito (26 cm), while Quincy Beta was the shortest ( 20 cm). Intermediate entries included Eco turf (2 2 cm), Ona 33 (2 3 cm), and UF Peace (24 cm), and they were not different from UF Tito or Quincy Beta. This follows the observation by Quesenberry et al. (2010) that UF Tito is an upright variety, w hile UF

PAGE 40

40 Peace is intermediate, and Ecoturf is more d ecumbent. When grazed every 3 or 6 wk, UF Tito was tallest, UF Peace and Florigraze were intermediate, and Ecoturf was shortest during early, mid and late season (Mullenix et al., 2016b). However, in th e current study there were no differences in canopy phenotypic plasticity, whereby it assumes a shorter canopy with greater bulk density when frequently and closely defoliated (Mullenix et al., 2016b; Shepard et al., 2018). Under infreque nt cutting like that utilized in this study, Ecoturf assumes a more intermediate growth habit. Herbage b ulk d ensity There was an entry x defoliation frequency interaction ( P = 0.003) for herbage bulk den sity (Figure 3 3). Interaction occurred because bulk density was similar between defoliation frequencies for all entries except UF Peace, where the bulk density was 410 kg ha 1 cm 1 for 2X and 244 kg ha 1 cm 1 for 1X. The defoliation frequency effect also approached significance for UF Tito ( P = 0.66; 360 a nd 251 kg ha 1 cm 1 for 2X and 1X, respectively). The treatment combinations with greatest bulk density were Quincy Beta harvested once or twice per year (56 7 and 452 kg ha 1 cm 1 ) and UF Peace harvested twice per year (410 kg ha 1 ). The ability of a fora ge plant to change its bulk density under different defoliation methods can be an indication of phenotypic plasticity. In 2 growing seasons of grazing every 3 or 6 wk, Ecoturf always had greater herbage bulk density (223 260 kg ha 1 cm 1 ) than UF Tito (105 150 kg ha 1 cm 1 ), UF Peace (117 188 kg ha 1 cm 1 ), and Florigraze (130 152 kg ha 1 cm 1 ) (Mullenix et al., 2016b). Ecoturf was also the only one of the fou r entries in which herbage bulk density was greater in both years for the 3 vs. the 6 wk grazing i nterval (260 vs. 175 and 268 vs. 223 kg ha 1 cm 1 in Years 1 and 2, respectively). UF Peace had greater herbage bulk density for

PAGE 41

41 the 3 wk treatment in the fi rst year of th at study only (156 vs. 117 kg ha 1 cm 1 ). In other experiments investigating the effe ct of Ecoturf RP defoliation, close, frequent cutting or grazing resulted in greater herbage bulk density (Rouse et al., 2004; Shepard et al., 2018). When gr azed weekly to 4 cm, Ecoturf bulk density reached nearly 500 kg ha 1 cm 1 and it declined when pos t grazing height was 8 cm and when interval between grazing increased to 4 or 7 wk (Shepard et al., 2018). As observed previously with Ecoturf canopy height in the current experiment, defoliation only once or twice a year is likely not frequent enough to a ffect bulk density significantly. Nutritive Value Responses Herbage c rude p rotein There was no entry x defoliation frequency interaction ( P = 0.146), however there was an effect of entry on herbage CP ( P < 0.0001) (Figure 3 4). Apalachee and Cowboy had gr eatest CP concentrations (153 and 154 g kg 1 respectively), and Arbrook, Quincy Beta, and Florigraze had the lowest levels (106, 123, and 124 g kg 1 respec tively). The latter three were lower in CP than all other entries except Quincy and Ona 33 (126 and 132 g kg 1 respectively). The average CP across entries was 138 g kg 1 Even entries with greatest CP concentration had values lower than observed in an ex periment comparing Arbrook and Florigraze under continuous stocking (Hernandez Garay et al., 2004). After three years in a moderately stocked pasture, average CP of Florigraze was greater than Arbrook (177 vs. 161 g kg 1 respectively). Those cultivars wer e ranked similarly in the current experiment (124 vs. 106 g kg 1 respectively ) There was also an effect of defoliation frequency ( P < 0.0001) (Table 3 4 ) on herbage CP in the current study, with the 2X treatment having greater herbage CP concentration than 1X

PAGE 42

42 (150 vs. 125 g kg 1 ). Generally, herbage CP of RP has followed the expected pattern for most forages, i.e., it decreases with increasing interval be tween defoliation events. Crude protein concentration decreased from 219 to 147 g kg 1 as interval between defoliation events increased from 2 to 12 wk (Beltranena et al., 1981) and CP of several ent ries averaged 170 g kg 1 when harvested three times per year (Dubeux et al., 2017) When Ecoturf plots were grazed every 4 or 7 wk, CP was greater for the 4 than 7 wk defoliation frequency (187 vs. 175 g kg 1 ) (Shepard et al., 2018). Ecoturf, UF Peace, UF Tito, and Florigraze nutritive value was compared at gr azing frequencies of 3 and 6 wk (Mullenix et al., 2016a). Defoliation frequency did not affect CP concentration of Ecoturf or Florigraze, but the generally more upright growing types, i.e., UF Peace a nd UF Tito, had greater CP when grazed at 3 instead of 6 wk frequencies. In that experiment, Florigraze CP concentration was less than all other entries at both regrowth intervals, while Ecoturf had the greatest CP concentration at 6 wk, but it was less t han only UF Peace at 3 wk. Herbage in v itro d igestibili ty There was an entry x defoliation frequency interaction for herbage IVDOM ( P < 0.0001) (Figure 3 5). When differences between defoliation frequencies occurred within an entry, the 2X treatment was a lways favored over the 1X treatment, but differences wer e limited to entries Ona 33 (642 vs. 577 g kg 1 ), Cowboy (645 vs. 547 g kg 1 ), Pointed Leaf (637 vs. 533 g kg 1 ), and Florigraze (645 vs. 550 g kg 1 ). The treatments with greatest IVDOM concentrations were Quincy Alpha and Apalachee (658 and 651 g kg 1 re spectively) when harvested twice per year, while the lowest concentrations were Pointed Leaf (533 g kg 1 ) and Cowboy (547 g kg 1 ) harvested once per year.

PAGE 43

43 Arbrook, UF Peace, and UF Tito were among th e most productive entries tested, but their IVDOM was re latively low and the range in IVDOM across defoliation frequencies was narrow. For the 2X and 1X treatments, respectively, Arbrook IVDOM was 578 and 562 g kg 1 UF Peace IVDOM was 595 and 579 g kg 1 and UF Tito IVDOM was 593 and 559 g kg 1 In contrast, g reater IVDOM, especially for the 1X treatment was observed for more decumbent growing types, with Waxy Leaf (21 cm), Arblick (21 cm), and Quincy Beta ( 20 cm) having IVDOM of 631, 607, and 599 g kg 1 w hen harvested only once per year. In general, both 1X an d 2X treatments can be considered infrequent defoliation, and IVDOM was often less than previously reported for RP in the literature. For example, Florigraze and Arbrook herbage from continuously stoc ked pastures had average IVDOM over 3 yr of 705 and 661 g kg 1 respectively (Hernndez Garay et al., 2004). Under rotational stocking with 6 wk intervals between grazing events in Florida (Sollenberger et al., 1989) and when harvested twice per season in Georgia (Terrill et al., 1996) Florigraze IVDOM was 650 g kg 1 or greater. When Florigraze, UF Tito, UF Peace, and Ecoturf were grazed at 3 or 6 wk frequencies, IVDOM ranged from 660 to 690 g kg 1 and there were no differences among entries and generally no difference between grazing frequencies. Ecoturf IVDOM was greater when grazed every 4 vs. 7 wk (718 g kg 1 vs. 687 g kg 1 respectively), but the range in IVDOM was relatively small (Shepard et al., 2018). Thus, those producers who are harvesting only o nce or twice per year will likely experience a significa nt reduction in IVDOM relative to more frequently harvested material, and the penalty may be most severe for the more upright growing and productive entries.

PAGE 44

44 Below Ground Responses Root r hizome m ass For root rhizome mass, there was no entry x defoliation interaction ( P = 0.347), but there was an effect of entry ( P = 0.0002) (Figure 3 6) and defoliation frequency ( P < 0.0001) (Table 3 4 ). Arbrook had greater root rhizome mass (9.01 Mg ha 1 ) than Apalac hee (5.63 Mg ha 1 ), Pointed Leaf (5.92 Mg ha 1 ), and Florigraze (5.22 Mg ha 1 ). Other entries with relatively large below ground mass included Quincy Alpha (8.67 Mg ha 1 ), UF Peace (8.17 Mg ha 1 ), and UF Tito (7.14 Mg ha 1 ). Plots cut once per year had 54% more root rhizome mass than those cut twice p er year (8.51 vs 5.52 Mg ha 1 respectively). Root rhizome mass data vary widely in the literature. Means from the current study are within the range reported for Florigraze under grazing (Ortega S. et al. 199 2b; Rice et al., 1995), less than those measur ed under clipping for several entries by Dubeux et al. (2017) and under grazing for Ecoturf by Shepard et al. (2018), and greater than those found under grazing of four entries by Mullenix et al. (2016b). Arbro ok, UF Peace, and Florigraze root rhizome mass w ere 17.3, 21.5, and 10.5 Mg ha 1 respectively, when clipped three times per year for 2 yr (Dubeux et al., 2017). Ecoturf root rhizome mass ranged from 16 to 20 Mg ha 1 for long established pastures that were defoliated by grazing every 1, 4, or 7 wk for 2 yr (Shepard et al., 2018), and it ranged from 3.2 to 4.5 Mg ha 1 for young stands that had been grazed every 3 or 6 wk for 2 yr (Mullenix et al., 2016b). In that study, Ecoturf and UF Tito had greater root r hizome mass than Florigraze and UF Peace. It i s well established that defoliation affects root rhizome mass of Florigraze RP. After 3 yr of grazing, average root rhizome mass was greatest (17.0 Mg ha 1 ) for an ungrazed control

PAGE 45

45 compared with pastures that w ere grazed every 63 d to a residual herbage ma ss of 2500 kg ha 1 (9.4 Mg ha 1 ) or every 21 d to a residual herbage mass of 500 kg ha 1 (2.3 Mg ha 1 ) (Rice et al., 1995). Similarly, after 2 yr of grazing, root rhizome mass was least when Florigraze pastures were grazed frequently and closely; it increa sed with increasing interval between grazing events when grazed closely, but it changed only slightly due to different intervals between grazing events when post grazing residual herbage mass was above 1700 kg ha 1 (Ortega S. et al., 1992a). For Ecoturf pa stures, grazing to a 4 cm stubble during 2 yr resulted in lesser root rhizome mass than grazing to 8 cm stubble (16.2 vs. 19.7 Mg ha 1 ), but there was no effect of regrowth intervals of 1, 4, and 7 wk. Data fro m the current and previous studies have implic ations for producers hoping to utilize their RP fields for both grazing or haying and as a source of planting material in the subsequent dormant period. Increasing the frequency of defoliation, even only from o ne to two cuts per year, in conjunction with s hort stubble heights is likely to reduce root rhizome mass that will decrease the amount of rhizomes that can be harvested from that area for use to plant new areas. Rhizome d iameter There was no entry x defol iation interaction ( P = 0.109), but there was an effect of entry ( P < 0.0001; Fig. 3.7) and defoliation frequency approached significance ( P = 0.055). The 1X treatment rhizome diameter was 3.23 mm vs. 3.07 mm for the 2X treatment, leading to the nearly sig nificant response. Arbrook root rhizome diamet er (4.07 mm) was greater than that of any other entry, and it was followed by UF Peace (3.70 mm) that was not different from Quincy Alpha (3.20 mm), Arblick (3.20 mm), Waxy Leaf (3.15 mm), Ona 33 (3.08 mm), and Ecoturf

PAGE 46

46 (3.05 mm). Cowboy (2.68 mm), Quincy B eta (2.80 mm), Pointed Leaf (2.80 mm), Florigraze (2.83 mm), Quincy (2.90 mm), and Apalachee (2.95 mm) rhizomes had lesser diameters than UF Peace. There are very limited data regarding rhizome diameter, but Pr ine et al. (1986) noted that Arbrook produced thicker rhizomes than all known RP lines at that time. The importance of rhizome diameter has not been studied by others, but it is likely to reduce the number of viable bud sites per unit mass of rhizome and t hus large diameter rhizomes should likely be a ssociated with greater planting rates. Root r hizome t otal n on s tructural c arbohydrate c oncentration There was no entry x defoliation interaction ( P = 0.297) or defoliation frequency effect ( P = 0.305; Table 3 4 ) on root rhizome TNC concentration, but there was an effect of entry ( P < 0.0001; Figure 3 8). UF Peace had greater TNC concentration (312 g kg 1 ) than Cowboy (234 g kg 1 ), Apalachee (232 g kg 1 ), Pointed Leaf (232 g kg 1 ), Arblick (228 g kg 1 ), Waxy Leaf (227 g kg 1 ), and Eco turf (202 g kg 1 ). Interestingly, the lowest TNC concentrations in this experiment are similar to the greatest concentrations reported for Florigraze by Ortega S et al. (1992a). At the end of a 2 yr experiment that evaluated a range o f grazing treatments o n Florigraze, they found TNC concentrations from 58 g kg 1 under frequent, close grazing to 210 g kg 1 when residual herbage mass was large (> 1500 kg ha 1 ) and regardless of grazing frequency. The least frequent grazing treatment in that study was three e vents per year, thus it was somewhat more intense than in the current study where Florigraze TNC was 241 g kg 1 There are several studies that report a reduction in RP root rhizome TNC concentration with more frequent defoliation. T hese include Ortega S et al. (1992a) and Rice et al. (1995) with Florigraze. Mullenix et al. (2016b) found no effect of 3 or 6 wk grazing frequencies on UF Tito,

PAGE 47

47 UF Peace, Ecoturf, or Florigraze, while Shepard et al. (2018) reported a relatively small lin ear increase in Ecotur f root rhizome TNC (198 to 214 g kg 1 ) as regrowth interval between grazing events increased from 1 to 7 wk. Perhaps the lack of response to defoliation frequency in this study was due to the narrow range of defoliation frequencies ev aluated and all levels could be considered infrequent. These data suggest that although grazing twice a year reduces root rhizome mass it is not frequent enough to diminish storage organ TNC concentration. Thus, 2X defoliation results in fewer rhizomes pro duced for subsequent p lanting activities, but the rhizome quality should still be good. Root r hizome t otal n on s tructural c arbohydrate p ool There was no entry x defoliation frequency interaction ( P = 0.458). Both RP entry ( P = 0.0001) (Figure 3 9) and defo liation frequency effe cts were significant ( P < 0.0001) (Table 3 4 ). The 1X treatment resulted in 45% greater TNC pool than 2X (2010 vs. 1390 kg TNC ha 1 ) due to an average 54% greater root rhizome mass for the 1X than the 2X treatment. Among entries, Arbl ick and Arbrook had the least and greatest TNC pools, respectively (1200 kg ha 1 vs 2720 kg ha 1 ). Arbrook TNC pool was also greater than Apalachee (1240 kg ha 1 ), Waxy Leaf (1360 kg ha 1 ), Pointed Leaf (1400 kg ha 1 ), Florigraze (1410 kg ha 1 ), Ecoturf (1 430 kg ha 1 ), a nd Cowboy (1500 kg ha 1 ), but not different than the other entries. Implications of the Research A single defoliation event near the end of the growing season result ed in significant reductions in annual HA, herbage CP, and herbage IVDOM for most RP entries relative to harvests at both the mid dle and end of the growing season. G reater disease pressure for the one harvest per year treatment likely cause d leaf drop that decrease d photosynthesis and HA of some

PAGE 48

48 entries Herbage accumulation was n ot affect ed by defoliation frequency for only the entry Quincy Beta, perhaps because of Quincy observed previously (Blount, personal communication). I t is also likely that a single harvest wa s associated wi th long p eriods of self shading that reduce d HA rate Defoliation either once or twice a year was not frequent enough to affect morphological mechanisms that would significantly alter canopy height or bulk density (i.e., phenotypic plasticity). The reducti on in her bage IVDOM with a single vs. two harvests was more severe for the mo st upright growing and productive entries. G reater defoliation frequency (2X vs. 1X) in conjunction with short stubble heights decreased harvestable root rhizome mass and root rhi zome TNC pool a response of particular importance to producers hoping to use RP fields both for grazing or haying and as a source of planting material the subsequent winter season However, planting material quality, evidenced by storage organ TNC, was si milar in both 1X and 2X defoliation treatments. In conclusion, rhizoma peanut hay producers should prioritize at minimum a summer and a fall harvest from their hay fields, even if summer weather conditions are not optimal for drying. P reservation as haylag e or bala ge, options that require less field drying, should be explored. Those producers wishing to dig rhizomes from hay fields should keep in mind that defoliation, especially more than one defoliation event per year will reduce the amount of rhizome mas s availab le for digging the following winter or spring. Finally, the impact of harvest frequency varies among entries of RP, with Quincy Beta being particularly tolerant of infrequent harvests in terms of both HA and herbage IVDOM. More upright growing ent ries (e.g ., Arbrook and UF Tito) and those most susceptible to leaf diseases (e.g., Ecoturf) are likely to show some of the greatest negative impacts associated with a single vs. multiple harvests per year.

PAGE 49

49 Table 3 1 Monthly rainfall at the experimental site in Quincy, FL and the 30 yr average rainfall for Quincy Data obtained from the Florida Automated Weather Network (FAWN) which has a recording station on site. 2015 2016 2017 30 yr a verage mm January 138 150 237 122 February 88 75 75 121 March 60 73 31 149 April 153 480 87 93 May 75 50 151 128 June 143 157 246 150 July 178 112 103 187 August 96 117 168 172 September 96 119 91 105 October 7 9 49 104 November 171 10 11 89 December 175 134 81 96 Total 13 80 1486 1330 1516

PAGE 50

50 Table 3 2 Rhizoma peanut cultivars, germplasms, introductions, and selections (referred to as entries) evaluated at the North Florida Research and Education Center (NFREC) at Quincy, FL. Entry Abbreviation in figures Description On a 33 33 An introduction from Paraguay. PI 262833. Notable for tolerance to short (2 d) water conditions at Ona. Arachis glabrata var. hagenbeckii. QS6W /Quincy Alpha Alpha Origin has been lost. Surviving plants from an introduction nursery planted at the N FREC in the early 1970s. Survived and spread over the past 50 yr. Selected for its ability to spread into bahiagrass pasture. Apalachee Apala Collected from a planting in Blountstown, FL. Arblick Arbli Collected near Bela Vista, Brazil. PI 658528. Releas ed as a germplasm in 2008. Low slow to establish. Arbrook Brook Developed from germplasm collections from Paraguay. PI 262817. Released in 1985. Recommended for droughty soils with warm winter temperature, not tolerant of poorl y drained soils. Erect growth, thick stems, and distinctive larger rhizomes. QS6W / Quincy Beta Beta Origin has been lost. Surviving plants from an introduction nursery planted at the NFREC in the early 1970s. Survived and spread over the past 50 yr. Selec ted for its aggressive growth habit and competitiveness, compared with Alpha and QS5W when planted with bahiagrass. Cowboy CowB Collected from volu nteer clone in Tifton, GA. PP 1. Originated from either a superior genetic recombination or outcross from an Arachis glabrata introduction. Most related to Florigraze. Ecoturf Eco Collected near Bela Vista, Brazil. PI 658529. Released as a germplasm in 2008. Low growing, quick to establish and tolerant of grazing. Susceptible to peanut stunt virus, peanut rust, and powdery mildew. Florigraze Flor Possible outcrossing between two plant introductions or a vigorous seedling from Arb. PI 421707 Released as a cultivar in 1978. Intermediate growth habit, susceptible to peanut stunt virus and powdery mildew. UF Peac e Peace An introduction from Paraguay. PI 262839. Released as a cultivar in 2008. Upright habit, high DM yield.

PAGE 51

51 Table 3 2. Continued Entry Abbreviation in figures Description Pointed Leaf Point From Brazil. Introduced in 2002. NRCS#9056068. Lo w growing, produces many flowers. Also QS5W/Quincy Qncy Origin has been lost. Surviving plants from an introduction nursery planted at the NFREC in the early 1970s. Survived and spread over the past 50 yr. Selected for i ts ability to spread into bahiagrass pa sture. UF Tito Tito An introduction from Paraguay. PI 262826. Released as a cultivar in 2008. Upright habit, high DM yield. Named in Professor of Agronomy at UF Waxy Leaf Waxy Collected from Corrientes, Argentina. Introduced in 2002. PI 262801. Low growing, produces few flowers and seeds. Table 3 3 Peanut rust ( Puccinia arachidis ) disease incidence observed on leaflet s of 14 rhizoma peanut entries at June and October 2017 harvest dates at Quincy, FL. Data presented are based on the Horsfall Barratt scale (1 to 12 ) of plant disease (Horsfall and Barratt ., 1945). Entry 27 June 2017 11 October 2017 Ona 33 1 1 c Quin cy Alpha 1 4 bc Apalachee 1 5 b Arblick 1 1 c Quincy Beta 1 1 c Arbrook 1 1 c Cowboy 1 1 c Ecotu r f 4 7 a Florigraze 1 1 c UF Peace 1 1 c Pointed Leaf 1 1 c Quincy 1 2 c UF Tito 1 5 b Waxy Leaf 1 2 c Data for 27 June 2017 could n ot be analyzed statistically because there was no variation among replicates for any treatment.

PAGE 52

52 Figure 3 1 Annual herbage accumulation from 14 rhizoma peanut entries cut once (1X) or twice (2X) per year in 2015, 2016, and 2017 at Quincy, FL. Data are e ntry x defoliation frequency means across years. indicates defoliation treatments within an entry are different ( P P > 0.05. SE = 0.67.

PAGE 53

53 Figure 3 2 Mean canopy height at harvest of 14 rhizoma peanut entries c ut once or twice per year at Q uincy, FL. Data are entry means across defoliation frequencies from the 2017 harvest year Mean canopy height was not measured in 2015 and 2016, thus it was not considered in this analysis Entry means that have the same letter above the boxplot are not di fferent ( P > 0.05). SE = 0.95.

PAGE 54

54 Figure 3 3 Herbage bulk density of 14 rhizoma peanut entries harvested once (1X) or twice (2X) per year at Quincy, FL. Data are entry x defoliation frequency means from the 2017 harvest year. indicates defoliation trea tments within an entry are different ( P 0.05); NS, P > 0.05. SE = 38.2. Figure 3 4 Mean herbage crude protein concentration of 14 rhizoma peanut entries cut once or twice per year for three years at Quincy, FL Data are entry averages across defoliat ion frequency treatments from harvests in 2015, 2016, and 2017. Entry means with the same letter above the boxplot are not different ( P ). SE = 6.01.

PAGE 55

55 Table 3 4 Main effects of defoliation frequency on 14 rhizoma peanut entries during 3 yr of defoli ation at Quincy, FL. Data are presented for all variable s for which there was no entry x defoliation frequency interaction. Response variable D efoliation Frequency (# yr 1 ) SE P 1 2 Canopy height (cm) 23.3 23.7 0.52 0.15 Herbage crude protein (g kg 1 ) 125 150 5.59 < 0.001 Root rhizome mass ( Mg ha 1 ) 8.5 5.5 0.61 < 0.001 Rhizome diameter (mm) 3.2 3. 1 0.09 0.06 TNC concentration (g kg 1 ) 247 255 8 .39 0.31 TNC pool (kg ha 1 ) 2010 1390 158 < 0.001 Figure 3 5 Mean in vitro digestible organic matter concentration (IVDOM) of herbage from 14 rhizoma peanut entries cut once (1X) or twice (2X) per year for three years at Quincy, FL. Data are means from herbage harvested in 2015, 2016, and 2017. Indicates defoliation treatments within an entry are different ( P P > 0.05. SE = 18.8.

PAGE 56

56 Figure 3 6 Root rhizome mass of 14 rhizoma peanut entries cu t once or twice per year for three years at Quincy, FL Data are entry means across defoliation frequency treatments and year s from annual sampling events that occurred at the end of the 2016 and 2017 (Years 2 and 3) growing seasons. Entry means that have the same letter above the boxplot are not different ( P >0.05). SE = 0.83. Figure 3 7 Mean rhizome diameter of 14 rhizoma peanut entries cut once or twice per year for three years at Quincy, FL. Data are entry means across defoliation frequency treatmen ts and years from annual sampling events that occurred at the end of the 2016 and 2017 (Years 2 and 3) growing seasons. Entry means with the same letter above the boxplot are not different ( P > 0.05). SE = 0.17.

PAGE 57

57 Figure 3 8 Mean root rhizome total nonstructural carbohydrate concentration (TNC) for 14 rhizoma peanut entries cut once or twice per year for three years at Quincy, FL. Data are entry means across defoliation frequency treatments and years from annual sampling events that occurred at the e nd of the 2016 and 2017 (Years 2 and 3) growing seasons. Entry means that have the same letter above the boxplot are not different ( P > ). SE = 15.7. Figure 3 9 Mean root rhizome total nonstructural carbohydrate (TNC) pool for 14 rhizoma peanut ent ries cut once or twice per year for three years at Quincy, FL. Data are entry means across defoliation frequency treatments and ye ars from annual sampling events that occurred at the end of the 2016 and 2017 (Years 2 and 3) growing seasons. Entry means tha t have the same letter above the boxplot are not different ( P > 0.05). SE = 256.

PAGE 58

58 CHAPTER 4 ABOVE AND BELOW GROUND RESPONSES OF RHIZOMA PEANUT EXPERIMENTAL LINES AND CULTIVARS WHEN GROWN AT TWO LOCATIONS Overview The development of legume based, grassland livestock systems is considered a key component for more sustainable ruminant production i n the future (Lscher et al., 2014; S chultze Kraft et al., 2018). Including perennial legumes in warm climate grasslands has numerous potential benefits. Legumes can increase forage nutritive value and animal performance (Rusland et al., 1988; Sollenberger et al., 1989), fix large amounts of atmospheric N (Dubeux et al., 2017), increase the amount of N mineralized from plant litter (Kohmann et al., 2018), increase soil C accumulation (De Deyn et al., 2009, 2011; Jensen et al., 2012), reduce methane (Archim de et al., 2011) and nitrous oxide (S oussana et al., 2010; Klumpp et al., 2011) emissions, and decrease the potential for other negative impacts on the environment (Jensen et al., 2012). Of perennial legumes proposed for use in pastures in the US Gulf Coa st region, rhizoma peanut (RP; Arachi s glabrata Benth.) is perhaps the best option because it is well adapted to sandy soils, persists under grazing ( Ortega S et al., 1992b; Shepard et al., 2018), can spread in mixtures with grasses (Mullenix et al., 2014 ), and has excellent nutritive value (Mullenix et al., negatively affected by the peanut stunt virus (Quesenberry et al., 2010) and is recommended for use only on moderately well to extremely well the second cultivar released from University of Florida, is specifically recommended for droughty, sandy soils (Prine et al., 1986). The University of Florida has relea sed additional

PAGE 59

59 forage RP genotypes in the past decade (Prine et al., 2010; Quesenberry et al., 2010) that are superior to Florigraze and Arbrook, but they too are adapted to well drained soils. Large areas of the US Gulf Coast region are occupied by seaso nally saturated soils, and there are very few legumes adapted to these environments. Efforts continue to identify superior RP lines, including those adapted to less well drained sites. Currently there are numerous RP plant introductions and selections with potential for use in Florida forage livestock systems. Data comparing key forage responses on well drained and poorly drained soils are lacking for these entries. The objective of this study was to quantify and compare the effects of location, and particu larly soil drainage on herbage produ ction, nutritive value, and root rhizome traits of selected RP introductions with those of existing cultivars. Materials and Methods Experimental Sites The experiment was conducted on well established RP plots at two locations: the North Florida Research a nd Education Center (NFREC) in Quincy, FL (30.55 N, 84.60 W) and the Agronomy Forage Research Unit in Hague, FL (29.80 N, 82.41 W). Plots were plant ed at 2009 in Quincy and 2010 in Hague. The soil at Quincy is a Norfolk loamy fine sand (fine loamy, kao linitic, thermic Typic Kandiudults). The Norfolk soil is characterized as well drained. At Hague, the soil is a Chipley sand (thermic, coated Aquic Quar tzipsamments). The Chipley series consists of very deep, somewhat poorly drained soils that formed in th ick deposits of sandy marine sediments. At Hague, the depth to the highest seasonal water table is 76 cm and occurs from June to September. At Quincy, t he highest seasonal water table occurs from January to May and is 155 cm. Given that the seasonal water table at Hague is only one half as deep as at

PAGE 60

60 Quincy an important difference between the two locations was soil drainage L ikelihood of saturated soils during the summer rainy season was much greater at Hague. Prior to planting the current experiment, the Hague site was occupied by spaced plant ings of tall fescue ( Festuca arundinacea Schreb.) and various clovers ( Trifolium sp.) and the Quincy site was oc cupied by bahiagrass ( Paspalum notatum Flgge). Soil samples were taken from both locations to a depth o f 15 cm and were analyzed at the University of Florida Soil Testing Laboratory. At Hague, soil pH was 5.7 and Mehlich 3 extractable P, K, and Mg were 10 8, 13, and 18 mg kg 1 respectively. Based on these results, dolomitic lime was applied at a rate of 110 0 kg ha 1 on 21 Feb. 2017. Potassium was applied at a rate of 50 kg ha 1 on 27 May 2017. At Quincy, soil pH was 6.0 and Mehlich 3 extractable P, K, and Mg were 38, 28, and 58 mg kg 1 respectively. In April 2017, K was applied at a rate of 56 kg ha 1 and P was applied at a rate of 15 kg ha 1 No fertilizer was applied at eithe r site in 2016. Treatments and Experimental Design The experiment was conducted for 2 yr at both locations (2016 and 2017). Rainfall data for the site s during the experime ntal period is shown in Table 4 1 Treatments at each location were introductions/se lections, germplasms, and cultivars of RP ( Table 3 1), here forward referred to as entries. There were 14 entries evaluated at Quincy and 15 at Hague. All entries were the same at the two locations except for the experimental line Chico, which appeared onl y at Hague. Data from Chico are not included in the results presented in this chapter, but they are summarized in Table A 1. The design at both locations was a randomized complete bloc k, with four replications at Hague and five replications at Quincy. Plot s were 2.5 m wide x 3 m long (7.5

PAGE 61

61 m 2 ) with a 1.8 m alley between plots at Hague, and 2.5 m wide x 1.5 m long (3.75 m 2 ) at Quincy. Response Variables Herbage a ccumulation, n utritive v a lue, and c anopy c haracteristics All plots were harvested twice per growi ng season, once each in summer and fall. Harvest dates at Hague were 18 Aug. and 16 Oct. 2016 and 13 July and 19 Oct. 2017. At Quincy, harvests occurred on 14 July and 7 Nov. 2016 and 27 June and 20 Oct. 2017. Harvest dates were chosen to represent typical hay harvest management in North Florida, where spring drought precludes significant growth of rain fed RP before June, and most hay producers operating without irrigation harvest only twice per year. At each harvest date, herbage was harvested to a 6 cm stubble height as described in Chapter 3. Remaining herbage was removed from the plot using a flail chopper. The forage samples were processed and analyzed for crude protein (CP) and i n vitro digestible organic matter (IVDOM) concentrations using the metho ds described in Chapter 3. Before each harvest event in 2017, canopy height was measured at 10 locations in each experimental unit to aid in quantifying differences in growth habit amo ng entries. Herbage bulk density was calculated by dividing herbage accu mulation (HA; kg ha 1 ) at a given harvest by the depth of the harvested canopy and expressed in kg ha 1 cm 1 Thus, if canopy height was 20 cm and the stubble height was 6 cm, bulk den sity was calculated by dividing HA by 14 cm. Values of bulk density repo rted are the averages of the two harvests per year.

PAGE 62

62 Root rhizome m ass and t otal n on structural c arbohydrate c oncentration and r hizome d iameter Root rhizome mass was measured by sampli ng three circular cores per experimental unit on 6 Dec. 2016 and 14 Dec. 2017 at Hague and 13 Dec. 2016 and 12 Dec. 2017 at Quincy. Cores were 10 cm in diameter and 20 cm deep. Sample processing, rhizome diameter measurement, and total non structural carbo hydrate (TNC) concentration analyses were conducted using the methods de scribed in Chapter 3. Statistical A nalyses Data were analyzed using a mixed model with entry, location, and their interaction as fixed effects and block and year as random effects. Yea r was considered a repeated measure. Location means were separated using the F test, and RP entry means were separated using P graphically in the form of box plots. The structure of box plots is such that the upper and lower hinges ( termina l ends of the box) relate to the first and third quartiles (or 25 th and 75 th percentiles), respectively. The thick solid line within and perpendicular to the sides of the box is the me dian. The upper and lower whiskers (solid lines extending above and below the box) extend from the hinge to the largest and smallest values (at most 1.5 interquartile range) of the hinge, respectively. Data points beyond the whiskers are considered outlier s.

PAGE 63

63 Results and Discussion Herbage Accumulation and Sward Canopy Characteristics Annual h erbage a ccumulation There was an entry x location interactio n for HA ( P < 0.0001) (Figure 4 1). Seven of the entries had greater HA at Quincy than at Hague. Conversely, none of the entries had greater HA at Hague than at Quincy (Figure 4 1). While acknowledging the presence of entry x location interaction, it is in formative to note that location main effect means were 10.5 and 7.28 Mg ha 1 yr 1 for Quincy and Hague, resp ectively. The generally superior performance at Quincy reflects soil drainage characteristics that were more favorable to RP. Among entries, notable exceptions to greater HA at the Quincy location were Ona 33, with HA of 11.0 and 10.5 Mg ha 1 yr 1 at Quinc y and Hague, respectively, and Waxy Leaf, which had HA of 9.8 and 8.6 Mg ha 1 yr 1 at the two locations, respectively. At Hague, where soils experie nce seasonal water logging in summer, Ona 33 (10.5 Mg ha 1 yr 1 ) had the same HA as the average of all entr ies at Quincy. Ona 33 at Hague had greater HA than Florigraze (4.8 Mg ha 1 yr 1 ), Arbrook (5. 8 Mg ha 1 yr 1 ), Pointed Leaf (5.9 Mg ha 1 yr 1 ), Apala chee (6.7 Mg ha 1 yr 1 ), Ecoturf (6.8 Mg ha 1 yr 1 ), and Quincy Alpha (6.9 Mg ha 1 yr 1 ) at that location. T hose that had the greatest advantage in HA at Quincy vs. Hague included Arbrook (13.7 vs. 5.8 Mg ha 1 yr 1 ), Florigraze (8.4 vs. 4.8 Mg ha 1 yr 1 ) and Pointed Leaf (10.5 vs. 5.9 Mg ha 1 yr 1 ). During 4 yr of defoliation on a poorly dra ined Spodosol, annual HA of Ona 33 increased 54% from Year 1 to Year 4, while annual HA of Arbrook decreased 41% over the same time period (Mislevy et al., 2007). Thus, these results support a conclusion that Ona 33 is much more tolerant of seasonal water logging than Arbrook. The mechanism for this response has not yet

PAGE 64

64 been elucidated. In the current study, the most recently released cultivars, UF Peace and UF Tito, performed relatively well at both locations (> 7.8 Mg ha 1 ), but all of the currently relea sed cultivars (Arbroo k, Florigraze, UF Peace, and UF Tito) and the released germplasm Ecoturf had greater HA at Quincy. On the Spodosol previously reference d UF Peace and UF Tito had greater HA than Ona 33 during the first year, but there was no differenc e in HA during Years 2 through 4 (Mislevy et al., 2007). At that location, annual HA of Ona 33 increased 54% from Year 1 to Year 4 compared with a decline of 5 and 6% for UF Peace and UF Tito, re s pectively, during the same period. Thus, it is unlikely that Ona 33 will have gre ater HA than UF Peace or UF Tito early in stand life or on well drained soils, but Ona 33 does appear to have greater tolerance of seasonal waterlogging that can occur in many Florida locations during summer. Annual RP HA in the south eastern US was in the range of 7 to 11 Mg ha 1 (Terrill et al., 1996; Venuto et al., 1999; Prine et al., 2010). In the current experiment, HA at the Quincy location was within or exceeded this range, but except for Ona 33 the HA at Hague was in the lower p art of this range or even below. In a previous experiment conducted at Hague, Florigraze, Arbrook, UF Peace, and UF Tito were harvested for 6 yr under a management regime similar to that of the current experiment (Freire et al. 2000). For 3 yr, samples we re clipped twice per year and the annual HA values were as follows: Arbrook (11.0 Mg ha 1 ); UF Tito (10.4 Mg ha 1 ); UF Peace (9.5 Mg ha 1 ); and Florigraze (9.2 Mg ha 1 ). These means for Arbrook and Florigraze were considerably above those reported at Hague in the current exper iment. For Florigraze, this can be explained based on the emergence of peanut stunt virus as a major disease concern in recent years but it is not clear if rainfall or other factors allowed Arbrook to perform better at this location i n the study of Freire et al. (2000) than in the current study.

PAGE 65

65 Canopy h eight There was an interaction of entry x location ( P = 0.0005) (Figure 4 2). Eight of 14 entries were taller at Quincy than at Hague, and this is reflected in location main effect mea ns of 24 and 1 8 cm, r espectively. Arbrook (29 cm) and UF Tito (2 7 cm) were among the tallest entries at Quincy, while at Hague UF Tito (2 4 cm) and Ona 33 were among the tallest (23 cm). This supports the observation by Quesenberry et al. (2010) that UF Tit o is an upright growi ng cultivar. At Hague, UF Peace (1 8 cm) and Ecoturf (15 cm) were shorter than UF Tito (2 4 cm) and Florigraze was not different (1 9 cm), but at Quincy, neither UF Peace, Ecoturf, nor Florigraze (22, 21, and 2 5 cm, respectively) differed in height from UF Ti to (2 7 cm). Although results show it is not different in height from intermediate types like UF Peace and Florigraz e when intervals are long between defoliation ev ents. Ecoturf was found previously to be phenotypically plastic, assuming a shorter canopy height under frequent, close defoliation (Mullenix et al., 2016; Shepard et al., 2018), but the interval between defo liation events in the current study was not suff iciently short to elicit this type of response. Herbage b ulk d ensity There was an entry x location interaction ( P < 0.0001) for herbage bulk density (Figure 4 3). Interaction occurred because bulk density was different between sites only for Apalachee, whe re the bulk density was 443 kg ha 1 cm 1 at Hague and 261 kg ha 1 cm 1 at Quincy. The site effect also approached significance for Quincy Alpha ( P = 0.09; 416 and 258 kg ha 1 cm 1 for Hague and Quincy, respec tively). Ona 33, Ecoturf, and Florigraze had nea rly the same herbage bulk density at both sites. At the Quincy location, entries Quincy Beta (453 kg ha 1 cm 1 ) and UF

PAGE 66

66 Peace (450 kg ha 1 cm 1 ) had greater herbage bulk density than Apalachee, Quincy Alpha, a nd Florigraze (259 269 kg ha 1 cm 1 ), while at H ague, Waxy Leaf, Apalachee, and Quincy Alpha had greater bulk density (404 443 kg ha 1 cm 1 ) than UF Tito (225 kg ha 1 cm 1 ), Ecoturf (203 kg ha 1 cm 1 ), and Arbrook (187 kg ha 1 cm 1 ). Quincy Beta bulk densi ty was among the greatest numerically at both lo cations, ranging from 355 kg ha 1 cm 1 at Hague to 453 kg ha 1 cm 1 at Quincy. It was also consistently among the shorter entries with heights ranging from 16.4 to 19.3 cm at the two locations. In terms of re lative bulk density between Ecoturf, Florigraze, UF Peace, and UF Tito, the results of this experiment were different than those previously reported (Mullenix et al., 2016b). They observed bulk density of Ecoturf (range of 175 to 268 kg ha 1 cm 1 ) to be gr eater than UF Peace (range of 117 to 188 kg ha 1 cm 1 ), Florigraze (range of 130 to 152 kg ha 1 cm 1 ), and UF Tito (range of 105 to 150 kg ha 1 cm 1 ) in each of 2 yr when grazed every 3 or 6 wk. When Ecoturf was grazed weekly to a 4 cm stubble, herbage bulk density was greater than when grazed every 4 o r 7 wk (Shepard et al., 2018). With long regrowth intervals in the current study, Ecoturf bulk density (328 375 kg ha 1 cm 1 ) was not different than UF Peace, UF Tito, or Florigraze at either location. In contrast, when defoliated frequently and closely, i t assume d a short, compact canopy structure with greater herbage bulk density (Mullenix et al., 2016a; Shepard et al., 2018). Nutritive Value Responses Herbage c rude p rotein There was no entry x location interaction ( P = 0.20) for herbage CP, however, the re were effects of location ( P < 0.0001) (Table 4 2 ) and entry ( P < 0.0001) (Figure 4 4). Herbage CP at Hague was greater than at Quincy (169 and 151 g k g 1 respectively), most likely because of

PAGE 67

67 greater HA at Quincy and associated dilution of N. Arbrook C P (129 g kg 1 ) was less than all entries except Quincy Beta (143 g kg 1 ) and Florigraze (141 g kg 1 ). Waxy Leaf CP (182 g kg 1 ) was greater than all othe r entries except Pointed Leaf (176 g kg 1 ), Cowboy (175 g kg 1 ), Arblick (171 g kg 1 ), Apalachee (169 g kg 1 ), and Ecoturf (167 g kg 1 ). Ona 33 (162 g kg 1 ), UF Tito (158 g kg 1 ), and UF Peace (158 g kg 1 ) were intermediate in herbage CP concentration. The se CP concentrations are generally within, or slightly lower than the CP reported in other defoliation studies in North Florida, where CP of a number of different entries ranged from 145 to 210 g kg 1 (Hernandez Garay et al., 2004; Mullenix et al., 2016a; Prine et al., 2010; Shepard et al., 2018). Length of regrowth period likely played a role in the somewh at lower CP observed for some entries in the current vs. previous studies. In previous experiments, RP swards were continuously stocked (Hernandez Garay et al. 2004), grazed every 3 to 6 wk (Mullenix et al. 2016a), or grazed every 1, 4, or 7 wk (Shepard et al., 2018), compared with the two harvests per year frequency employed in the current study. Crude protein of RP lines was evaluated when harvested thre e or four times per year in a 3 yr clipping study at the Range Cattle Research and Education Center in South Florida (Mislevy et al., 2007). Average CP was greatest for Ecoturf (191 g kg 1 ) followed by UF Peace (188 g kg 1 ), Ona 33 (178 g kg 1 ), UF Tito (1 76 g kg 1 ), Florigraze (166 g kg 1 ), and Arbrook (146 g kg 1 ). Herbage CP of most lines followed simila r patterns of response to that observed in the current study. An exception was Ecoturf, for which long regrowth intervals increase occurrence of peanut r ust leaf disease (Chapter 3) negatively affecting nutritive value. In contrast, when harvested more fr equently, disease presence is reduced or not observed and nutritive value is greater

PAGE 68

68 Herbage in v itro d igestibility There was an entry x location interaction for herbage IVDOM ( P = 0.0002). Interaction occurred because there were differences between loca tions only for Florig raze and Pointed Leaf (Figure 4 5); in both cases IVDOM was greater at Quincy than Hague. While acknowledging the pres ence of entry x location interaction, location main effect ( P < 0.0001) means were 639 and 612 g kg 1 for Quincy and Hague, respectively. Herbage IVDOM of UF Peace (617 vs. 614 g kg 1 at Hague and Quincy, respectively) and UF Tito (608 vs. 604 g kg 1 at H ague and Quincy, respectively) was not different between locations. Ecoturf (577 g kg 1 ) and Arbrook (588 g kg 1 ) IVD OM at the Quincy location were less than all other entries except UF Tito (604 g kg 1 ), UF Peace (614 g kg 1 ), and the entry called Quincy (634 g kg 1 ). Greater leaf disease incidence may have lowered Ecoturf IVDOM at Quincy compared with Hague (577 vs. 60 2 g kg 1 ). At Hague, there was no difference among entries and IVDOM ranged from 592 g kg 1 for Arbrook to 638 g kg 1 for Waxy Leaf. Previ ous studies have reported a range in RP entry IVDOM from 650 to 705 g kg 1 (Sollenberger et al., 1989; Terrill et al. 1996; Hernandez Garay et al., 2004; Shepard et al., 2018). In a previous experiment conducted near the research site in Quincy, average I VDOM was 712 g kg 1 for Arblick, Arbrook, Ecoturf, Florigraze, UF Peace, and UF Tito when harvested three times per y ear (Dubeux et al., 2017). In an experiment in south Florida, several RP entries were harvested three or four times per year and average IV DOM was greater (688 g kg 1 ) than means reported in the current experiment. As already described for herbage CP, the primary reason for lower IVDOM in the current study was long intervals between harvest events. In south Florida, Ona 33 IVDOM was greater t han Arbrook and UF Tito in 2 of 3 yr and greater than UF

PAGE 69

69 Peace and Ecoturf in 1 of 3 yr In the current study, Ona 33 IVDOM (653 g kg 1 ) was greater than Arbrook and Ecoturf at Quincy but not different (605 g kg 1 ) than any other entry at Hague. Below Gro und Responses Root r hizome m ass There was no entry x location interaction ( P = 0.158) for root rhizome mass, but ther e was an effect of location ( P = 0.001) (Table 4 2 ) and entry ( P = 0.0001) (Figure 4 6). Below ground biomass was greater at Hague compared with Quincy (6.48 vs 5.64 Mg ha 1 respectively). Arbrook (7.47 Mg ha 1 ) and Quincy Alpha (7.44 Mg ha 1 ) had greater root rhizome mass than Apalachee and Florigraze (4.72 and 4.27 Mg ha 1 respectively). The entries Quincy and Waxy Leaf (6.82 and 7.08 Mg ha 1 respectively) also had greater root rhizome mass than Florigraze. In a clipping study near Quincy, root rhizom e mass of UF Peace, Arbrook, and Florigraze was 21.5, 17.3, and 10.6 Mg ha 1 respectively, values well above those observed in the current study. Near the Hague site in an experiment evaluating a Florigraze RP pasture under a range of grazing treatments ( from severe to lax grazing), Rice et al. (1995) reported mean Florigraze roo t rhizome mass of 5.75 Mg ha 1 similar to the average across locations and entries of 6.06 Mg ha 1 in the current study. When 2 yr old plots of four entries were grazed every 3 or 6 wk (Mullenix et al., 2016b), root rhizome mass of Ecoturf, UF Tito, Flori graze, and UF Peace was 4.45, 4.11, 3.49, and 3.17 Mg ha 1 respectively. In the current experiment, the same four entries had root rhizome mass of 6.02, 5.73, 4.27, and 6.59 Mg ha 1 respectively. Greater values than observed in the Mullenix et al. (2016b ) study was likely due to a combination of factors including less

PAGE 70

70 frequent defoliation and greater pasture age at time of initiation of defoliation in the current study. Rhizome d iameter There was no entry x location interaction ( P = 0.613) for rhizome di ameter, but there were effects of location ( P < 0.0001; Table 4 2 ) and entry ( P < 0.0001) (Figure 4 7). Rhizome diameter was greater at Quincy than Hague (3.08 vs. 2.63 mm, respect ively), in spite of lesser root rhizome mass at Quincy. The effects of the g rowing environment or defoliation management on rhizome diameter are not known. Arbrook rhizome diameter was greater (4.8 mm) than all other entries. UF Peace (3.6 mm) rhizome diam eter was less than Arbrook but greater than all entries except Florigraze (2 .9 mm), Waxy Leaf (2.9 mm), and Quincy Beta (2.8 mm). Other entries ranged from a low of 2.2 mm for Cowboy to 2.78 mm for Ona 33. While there are limited data regarding rhizome dia meter, Prine et al. (1986) noted that Arbrook produced thicker rhizomes than those of any known RP line at that time. Total n on s tructural c arbohydrate c oncentration There was no entry x location interaction for root rhizome TNC ( P = 0.532), but there were effects of entry ( P < 0.0001) (Figure 4 8) and location ( P = 0.05) (Table 4 2 ). Root rhizome TNC concentration was marginally greater at Hague (269 g kg 1 ) than at Quincy (255 g kg 1 ), corresponding with greater root rhizome mass at Hague. Arbrook (343 g kg 1 ) had greater TNC concentrations than Florigraze (260 g kg 1 ) and Arblic k (251 g kg 1 ), while both Arbrook and UF Peace (308 g kg 1 ) had greater root rhizome TNC concentrations than Ona 33 (235 g kg 1 ), Waxy Leaf (235 g kg 1 ), Pointed Leaf (234 g kg 1 ) UF Tito (232 g kg 1 ), Cowboy (231 g kg 1 ), Apalachee (230 g kg 1 ), and Eco turf (224 g kg 1 ).

PAGE 71

71 These TNC concentrations are generally within the ranges or greater than those observed in previous experiments. At the end of a 2 yr experiment evaluating an ar ray of grazing treatments on Florigraze RP pastures in Gainesville, FL, TNC concentrations ranged from 58 for closely and frequently grazed treatments to 210 g kg 1 for leniently defoliated treatments ( Ortega S et al., 1992a). In another grazing experimen t at the same location, Shepard et al. (2018) reported Ecoturf RP pastures w ith a range of 198 to 314 g kg 1 for grazing frequencies from 1 to 7 wk in combination with residual stubble heights of 4 to 8 cm. When 2 yr old stands of Ecoturf, Florigraze, UF P eace, and UF Tito were defoliated every 3 or 6 wk, root rhizome TNC ranged f rom 115 to 142 g TNC kg 1 (Mullenix et al., 2016b). A TNC concentration above 130 g kg 1 was suggested for Florigraze rhizomes used as planting material (Rice et al., 1995), as tho se with lesser concentrations resulted in stand failure. All entry means for TNC concentration in the current experiment are well above that level, but it is not known what concentrations are o ptimal for entries other than Florigraze. These data clearly in dicate that Arbrook maintains a large amount of below ground biomass with very thick rhizomes and high concentrations of TNC. In contrast, entries like Apalachee and Cowboy have lesser rhizome mass, rhizome diameter, and TNC concentration. Total n on s truct ural c arbohydrate p ool There was no entry x location interaction ( P = 0.651) for TNC pool, but both RP entry ( P < 0.0001) (Figure 4 9) and location effects were significant ( P < 0.0001) (Table 4 2 ). Hague TNC pool was 26% greater than Quincy (1760 vs. 1390 kg ha 1 ; Table 4 2 ) due to an average 15% greater root rhizome mass and 5.5% greater TNC concentration at Hague than Quincy. Among entries, Arbrook had greater TNC pool (2350 kg ha 1 ) than Arb lick (1450 kg ha 1 ) and

PAGE 72

72 UF Tito (1300 kg ha 1 ), and both Arbroo k and Quincy Beta (2090 kg ha 1 ) were greater than Pointed Leaf (1260 kg ha 1 ), Ona 33 (1250 kg ha 1 ), Ecoturf (1250 kg ha 1 ), Cowboy (1200 kg ha 1 ), Florigraze (1130 kg ha 1 ), and Apalachee (1 080 kg ha 1 ). Entries that were not different than Arbrook incl ude d Quincy Beta (2090 kg ha 1 ), UF Peace (2020 kg ha 1 ), Quincy (2010 kg ha 1 ), Quincy Alpha (1970 kg ha 1 ), and Waxy Leaf (1740 kg ha 1 ). These values are less than those previously reported for a more than 10 yr old stand of Ecoturf defoliated at a rang e of grazing intervals and stubble heights (3040 to 4060 kg ha 1 ) (Shepard et al., 2018), but greater than 2 or 3 yr old stands of Ecoturf, Florigraze, UF Tito, and UF Peace grazed every 3 to 6 wk to remove 50 to 75% of herbage mass (< 1000 kg ha 1 ; Mulle nix et al., 2016b). These data support the previously mentioned characterization of Arbrook as an entry with a large root rhizome mass and supply of reserves. In contrast, the TNC pool of Apala chee and Cowboy is only 46 and 51% as great as that of Arbrook. Implications of the Research Soil drainage characteristics at Quincy were more favorable to RP entries in general than those at Hague, which led to many entries having greater HA at Quincy tha n Hague. Ona 33 had nearly the same HA at both locations, and i ts HA in poorly drained soils at Hague was nearly identical to the average HA of all other entries at the well drained Quincy location. These data support a conclusion that Ona 33 is considerab ly better adapted to seasonally saturated soils than nearly all other RP entries currently being used commercially or being evaluated for potential cultivar release. Previous studies have characterized Ecoturf as a low growing plant with greater herbage bu lk density than current cultivars, however, this experiment sho wed no difference in height or

PAGE 73

73 bulk density between Ecoturf and intermediate growth habit types like UF Peace and Florigraze. The most likely explanation for these differences is that Ecoturf i s phenotypically plastic, assuming a shorter canopy height with greater herbage bulk density under frequent, close defoliation, but not differing in growth characteristics from other intermediate types when intervals between defoliation events are long, li ke those in the current study. Greater HA and associated dilut ion of N at Quincy resulted in greater herbage CP at Hague. In general, the longer regrowth periods in this study than most previous studies resulted in lesser herbage CP and IVDOM concentratio ns in the current experiment. The longer defoliation interval l ikely had the greatest negative effect on nutritive value of Ecoturf, which had the greatest incidence of leaf rust. Unlike HA, root rhizome mass and TNC pool were greater at Hague, but rhizom e diameter was larger at Quincy. Arbrook and Quincy Alpha had g reater root rhizome mass than Apalachee and Florigraze. In general, Arbrook maintained a large amount of below ground biomass with very thick rhizomes and high concentrations of TNC. In contras t, entries like Apalachee and Cowboy have lesser rhizome mass, rhizome diameter, and TNC concentration. These data suggest that there may be potential to expand the current zone of adaptation of RP into wetter soil environments by evaluating additional ent ries and releasing new cultivars. Additionally, there are a wid e range of above and below ground growth characteristics available among entries being tested, providing opportunity for targeting entries to specific intended uses or production goals.

PAGE 74

74 Tabl e 4 1 Monthly rainfall at the experimental sites near Quincy and Hague, FL and the 30 year average rainfall for Quincy and Hague, FL. Data obtained from the Florida Automated Weather Network (FAWN) which has recording stations at each site. Quincy Hague 2016 2017 30 yr a verage 2016 2017 30 yr a verage mm January 150 237 122 72 73 85 February 75 75 121 118 34 81 March 73 31 149 72 21 110 April 480 87 93 96 102 68 May 50 151 128 64 84 63 June 157 246 150 248 477 181 July 112 103 187 122 150 154 August 117 168 172 124 166 162 Septe mber 119 91 105 192 344 112 October 9 49 104 28 51 73 November 10 11 89 2 87 52 December 134 81 96 54 79 60 Total 1486 1330 1516 1192 1668 1201

PAGE 75

75 Figure 4 1 Annual herbage accumulation of 14 rhizoma peanut entries harvested twice per year in Quinc y and Hague, FL during 2016 and 2017. Data are entry x location means across two years. indicates locations within an entry are different ( P 0.05); NS, P > 0.05. SE = 0. 97.

PAGE 76

76 Figure 4 2 Canopy height of 14 rhizoma peanut entries harvested twice per y ear in Quincy and Hague, FL. Data are entry x location means across harvests for 2017. indicates locations within an entry are different ( P P > 0.05. SE = 1.25. Table 4 2 Main effects of location on responses for which there was no entry x location interaction. Response variable Location SE P Hague Quincy Herbage crude protein (g kg 1 ) 169 151 5.95 < 0.001 Root rhizome mass (Mg ha 1 ) 6. 5 5.6 0.61 0.001 Rhizome diameter (mm) 2.6 3.1 0.14 < 0.001 TNC concentration (g kg 1 ) 269 255 8. 65 0.05 0 TNC pool (kg ha 1 ) 1760 1390 203 0.001

PAGE 77

77 Figure 4 3 Herbage bulk density of 14 rhizoma peanut entries harvested twice per year in Quincy and Hague, FL. Data are entry x location interaction means across two harvests in 2017. indicate s location means within an entry are different ( P NS, P > 0.05. SE = 36.1.

PAGE 78

78 Figure 4 4 Herbage crude protein concentration of 14 rhizoma peanut entries harvested twi ce per year in Quincy and Hague, FL. Data are entry means across locations and harvests in 2016 and 2017. Entry means that have the same letter above the boxplot are not diff erent ( P > 0.05). SE = 7.01.

PAGE 79

79 Figure 4 5 Herbage in vitro digestible organic matter (IVDOM) concentration for 14 rhizoma peanut entries harvested two times per year in Quincy and Hague, FL. Data are entry x location interaction means across harvests in 2016 and 2017. indicates defoliation treatments within an entry are di fferent ( P P > 0.05. SE = 1 8.0

PAGE 80

80 Figure 4 6 Root rhizome mass of 14 rhizoma peanut entries harvested two times per year in Quincy and Hague, FL. Data are entry means across locations and years for annual mass measurements taken at the end of the 2016 and 2017 growing seasons. Entry means that hav e the same letter above the boxplot are not different ( P >0.05). SE = 0.2

PAGE 81

81 Figure 4 7 Rhizome diameter of 14 rhizoma peanut entries harvested by clipping two times per year in Quincy and Ha gue, FL. Data a re entry means across locations and years for annual measurements taken at the end of the 2016 and 2017 growing seasons. Entry means that have the same letter above the boxplot are not different ( P > 0.05). SE = 0.2

PAGE 82

82 Figure 4 8 Root rhizome total nons tructural carbohydrate (TNC) concentration for 14 rhizoma peanut entries harvested for hay twice per year in Quincy and Hague, FL. Data are entry means across locations and years for annual measurements taken at the end of the 2016 and 2017 growing seasons Entry means that have the same letter above the boxplot are not different ( P > 0. ). SE = 16.3.

PAGE 83

83 Figure 4 9 Root rhizome total nonstructural carbohydrate (TNC) pool for 14 rhizoma peanut entries harvested twice per year for h ay in Quincy and Hague, FL. Data are entry means across locations and harvests in 2016 and 2017. Entry means that have the same letter above the boxplot are not different ( P ). SE = 256.

PAGE 84

84 CHAPTER 5 A MODIFIED ROOT INGROWTH CORE DEVICE TO MEASURE ROO T ACCUMULATION RATE OF P ERENNIAL FORAGE SPECIES Overview Agricultural grassland soils can mitigate climate change by sequestering a significant amount of atmospheric C. Management practices, such as defoliation and fertilization that promote forage production also increase soil C storage (Conant et al., 2001; Allard et al., 2007; Ammon et al., 2007). Of particular importance to the C cycle in grasslands is root production, as roots may contribute more SOC than above ground bioma ss (Rasse et al., 2005). However, root production i n grasslands is difficult to measure or predict. It is known that fine root exudates transfer C to stable SOC pools (Bradford et al., 2013), which can impact nutrient cycling and SOC storage (Phillips et a l., 2011). In grasslands that are frequently graze d or cut for hay, removal of above ground biomass may render roots the most significant contributor of C to soil. There is little information available on root accumulation rates under perennial forage cro ps and more is needed to advance methods of studyin g root production in grasslands, as existing methods are limited. While indirect techniques of measurement, like rhizotron technologies, are desirable because they are non destructive, the cost and time re quired for data processing may preclude their use. Also, damage sustained from hay harvesting machinery or grazing livestock makes rhizotron use challenging in defoliation studies. I ngrowth core s are another option for studying root accumulation They are more affordable than rhizotrons, as they are made f rom commercially available window screen However, the screen may not be ideal because serve as

PAGE 85

85 a barrier to small insects rhizomes or thick roots A study was con ducted to quantify the effect of mesh sizes (0.25, 0.5, 1.0, 1.5 and 2.0 mm) for measuring fine root growth of Poplar ( Populus nigra L.) ( Montagnoli et al., 2014). A considerable amount of fine root biomass was found outside bags that had a mesh size small er than 1.5 mm, demonstrating that small mesh openings restrict root growth into the core Another reason to improve this method is that w indow screen may not be strong enough to withstand the impact of grazing livestock. Livestock treading and subsequent resulting in inaccurate calculations. Given the constraints of current methods to study ro ot production, more techniques are needed to better measure root accumulation rate of agricultural grasslands. The obj ective of this study was to develop and test a prototype root ingrowth core device designed to quantify root accumulation rate of a group of rhizoma peanut (RP; Arachis glabrata Benth.) entries Materials and Methods Design and Construction of the Ingrowth Core The ingrowth cores were constructed of 4 mm polyester mesh wrapped around 8 cm diameter cage wire mesh (Figure 5 1 ). The purpose of the cage wire is to reinforce the sides of the core without blocking root and rhizome ingrowth. In previous below ground studies, it was observed that pressure from root and rhizome growth and insect movement disturbed polyester mesh bags used to measure b elow ground litter decomposition (Sollenberger, personal communication). It was suspec ted that similar activity might distort the geometry of the ingrowth core and that such distortion could lead to miscalculation of soil volume and root rhizome biomass ac cumulation. Thus, the cage wire serve d to maintain the form of the ingrowth core.

PAGE 86

86 To p repare the cage wire for the cores, wire was cut into 26 x 31 cm pieces with pliers (Figure 5 2). The individual wire s that form the cage wire a re spaced 2.5 4 cm apart. The wire was then bent into a cylinder on a hard surface that can withstand scratches from exposed wire ends. To secure the cylindrical shape, the exposed wire was folded from one side of the core over the other side of the core To save time and avoid sna gging the mesh fabric when it wa s pulled over the core in later steps, the exposed wir e was always folded toward the interior of the core If the core is to be installed in the soil at a 45 angle, the top of the ingrowth core should be tapered to allow it to lie flush with the surface of the soil To do so, 2.5 4 cm should be c ut from 5 units ( 12. 7 cm ) of wire from the top of the core (Figure 5 3). One 70 cm x 16 cm piece of polyester netting utility fabric and one 35 cm x 16 cm piece of architectural draft ing paper should be cut per ingrowth core (Figure 5 4). Based on the dimensions in Figure 5 5, guidelines should be drawn to serve as a template for sewing of the mesh bags. T he mesh fabric should be folded in half so that the template is inside the fabric and the template can be secured to the fabric with tape. T he top of the fabric can be removed (Figure 5 5). T he sides of the mesh fabric should be sown with a zig zag stitch and exterior grade thread. T he drafting paper should then be removed from the me sh and recycle d T he mesh can then be pulled onto the core (Figure 5 6). The exposed wire ends should be pulled through the top of the fabric and the wire ends folded into the inside of the core. This will secure the fabric to the core pan.

PAGE 87

87 Use of the Ingrowth Core in a Field Experiment A 2 yr experiment was conducted in 2017 and 2018 at the Agronomy Forage Evaluation Unit in Hague, FL (29.80 N, 82.41 W). The experiment site is the same Hague location as descri bed in Chapter 4 and d et ailed site characteristics are presented in that chapter Treatments and e xperimental d esign The t reatment s were six RP entries including existing cultivars a s well as selections under evaluat ion for potential cultivar release. Treatments we re arranged i n four replicates of a randomized complete block design. T he six entries represent a subset of the 1 5 entries used at the Hague site that are described in Chapter 4 They included experimental lines Apalachee, Chico, and Ona 33, the released germplasm Ecot urf, and released UF UF These entries were chosen to include a wide range in herbage accumulation canopy height, and root rhizome mass based on data collected at this site in 2016 and 2017 (Table 5 1) Limiting the entries t o six wa s necessary because of the time associated with construction, installation, and removal of the root ingrowth cores Placement of the i ngrowth c ores The ingrowth cores were constructed as described earlier in this chapter and were used to measure a ccumulation of root rhizome mass based on methodology described by Makkonen and Helm i saari (1999). Three ingrowth cores were placed i n each of four replicates of the six entries for a total of 72 cores. The ingrowth cores were installed at a 45 angle and measured to a depth of 23 cm Based on measurements of RP rhizome diameter taken in 2016 it was determined that mesh with a 4 mm opening would be large enough to allow root and rhizome growth of each of these entries into the core.

PAGE 88

88 In order t o install ea ch core, a 7.5 cm diameter x 28 cm deep core of soil, rhizomes, and roots w as excavated with an auger angled at 45 at three locations per plot (Figure 5 7) The soil and plant material from each excavation was screened through a 250 u m screen and plant ma terial discarded. W hen high soil moisture prevented sieving plant material with a 250 u m screen, material was passed through a 0.635 cm screen (Figure 5 8) Any remaining plant material was then removed by hand. The sieved soil was placed inside the constr ucted ingrowth cores, and the y were inserted into the excavated cores such that the upper most edge of the ingrowth core was just below the soil surface (Figure 5 9) Response Va riables Roots in perennial pastures have a lifespan of approximately 100 d (14 wk) (Van der Krift & Berendse, 2002). Since it is important that the time period the ingrowth cores are left below ground is shorter than the lifespan of the roots, the current experiment was conducted for approximately 100 d. During each year, all plots were harvested once during the period when cores were in place. Forage was cut to a 6 cm stubble on 13 July 2017 and 20 July 2018. Cores remained in place during 1 0 2 d from 9 June through 19 Sept 2017 and 104 d from 13 June through 26 Sept 2018. Prior to excavation, c ores were located with a metal detector. At the end of the root growth period, the ingrowth cores were excavated and any portion of the root protruding from the core was cu t and discarded (Figure 5 10 ) All soil w as washed from the roots on a n 840 u m s creen and the three cores per plot were composited (Figure 5 11 ) Root samples were ov en dried at 60 C to a constant mass. Prior to weighing, root samples were screened again to remove any soil and small stones. R oot rhizome biomass w as multipli ed by a constant (10 0 000/ [ pi*3.75^2 ] ) to convert the measurements to k g ha 1 Root rhizome mass net

PAGE 89

89 accumulation rate w as calculated by dividing root rhizome mass by number of days the cores were in place. Accumulation rate d ata are presented in kg ha 1 d 1 Statistical Analysis Root rhizome mass accumulation rate d ata w ere analyzed using a mixed model with entry as a fixed effect and block as a random effect. The entry means were separated using P 0.05. Data are presented graphically in the form of box plots The str ucture of box plots is such that the upper and lower hinges ( terminal ends of the box) relate to the first and third quartiles (or 25 th and 75 th percentiles), respectively. The thick sol id line within the box and perpendicular to the sides of the box is th e median. The upper and lower whiskers (solid lines extending above and below the box) extend from the hinge to the largest and smallest values (at most 1.5 interquartile range) of the h inge, respectively. Data points beyond the whiskers are considered out liers. Results and Discussion There was a significant effect of entry on root accumulation rate ( P = 0.01) (Table 5 2). Root accumulation rate was greate r in decumbent Chico (32.8 kg ha 1 day 1 ) than either Ona 33 or UF Peace (10.0 and 14.2 kg ha 1 d 1 re spectively). Apalachee, Ecoturf, and UF Tito root accumulation rates (22. 8, 22.4, and 23.0 kg ha 1 day 1 respectively ) were intermediate and not different than any other entry T hat the ingrowth core was able to detect differences among entries of the same species with similar below ground architecture suggests potential for the technology tubes (2 m m mesh, 5 cm diameter tubes, 20 30 cm long) we re used to estimate fine root

PAGE 90

90 production during 1 yr in organic soils of moist tussock tundra and wet sedge ecosystems (Nadelhoffer et al., 2002) The authors found fine root production was greater in wet sedge than moist tundra (2.06 kg ha 1 d 1 vs 1.53 k g ha 1 d 1 ). Root accumulation rate during summer in the current experiment was on average 20 times greater than over 1 yr in the Alaska environment A key assumption Nadelhoffer et al. (2002) made was that roo t growth rates in the core were similar to tho se in the adjacent soil They also assumed that because the dominant species in the grassland was annual ( Eriophorum vaginatum L. ) and non mycorrhizal, senesced roots were easily detectible because they were fl at and grey. Upon excavating the core, they fo und few senesced roots. Roots in perenni al pastures have a lifespan of approximately 100 d ( 14 wk) (Van der Krift & Berendse, 2002) Since it is important that the time period the i ngrowth cores are left below ground is shorter than the lifespan of the roots, the current experiment was conducted for approximately 100 d. A study comparing the ingrowth core technique with sequential soi l cores for four consecutive growing seasons in a Scots pine ( Pinus sylvestris L.) stand in Ferric Podzol soils found that root accumulation rate varied from 6.14 to 13.4 kg ha 1 d 1 for ingrowth cores and 9 .59 to 37.8 kg ha 1 d 1 for soil cores (Makonnen and Helmisaari, 1999) The range reported for the ingrowth core includes the lower part of the range in the current experiment. The authors outlined a number of problems with the ingrowth core. One is that the ingrowth growing environment is different than natural soil because there are no channels create d from decomposing roots. New rhizomes and roots might grow through the channels created by old, decomposing roots. Another challenge to ingrowth core use is that soil horizons in naturalized landscapes suc h as forests and unmanaged g rasslands cannot be re constructed after sieving. T o

PAGE 91

91 compensate, t he authors put a humus clod on top of the installed core In the current experiment soil is likely not as stratified because it w as tilled prior to planting in 2009. Due to management, using the ingrowth core on agricultural soils might not present th e same challenge s indicated by Makonnen and Helmisaari (1999) In a recent study, s equential soil cores were compared with ingrowth cores on p asture s mainly composed of perennial ryegrass ( Lolium perenne L. ) with nea rly equal parts smooth meadowgrass ( Poa pratensis L. ), timothy ( Phleum pratense L. ) and white clover ( Trifolium repens L. ) in Northern Germany (Chen, et al., 2015) I ngrowth cores were left in Eutric Luvisols for three p eriods: short term (4 wk) medium term (9 wk) and long term (27 wk). S equential soil cores were sampled by root auger at the same interval as medium term ingrowth cores. Because the ingrowth cores are a direct and more accurate representation of root accumulatio n, the soil cores were compared with the medium term ingrowth core s to evaluate their accuracy R oot accumulation rate in the sequential soil cores was underestimated by approximately half that of the medium term ingrowth cores ( 5 6 vs 118 kg ha 1 d 1 ). An ingrowth core experiment in agricultural soils in Bavaria, Germany studied gross root growth in potato ( Solanum tuberosum L. ) spring wheat ( Triticum aestivum L ) winter wheat ( Triticum aestivum L ), and winter barley ( Hordeum vulgare L. ) for approximate ly 2 wk at a time (Steingrobe et al., 2001). Although the study measured root production as a function of root length density, not root accumulation rate, it addressed two common concerns with ingrowth cores. One is whether root growth can be affected by d isparate soil conditions inside and outside t he ingrowth core In an e xperiment altering soil density, N, P, K and soil moisture it was reported that only very high N ( treatments fertilized with 80 kg N ha 1 ) and high soil density

PAGE 92

92 ( 1.64 g cm 3 ) increased and decreased root grow th respectively, of rape seed ( Brassica napus L. ) (Steingrobe et al. 2000) A second concern the authors addressed is whether root growth is reduced by compact ion in soil walls during ingrowth core installation. Th e experiment concl uded that installation did not alter the root growth pattern in and around the core. Other experiments evaluating the effect of rhizotrons on root growth patterns found that technology did not affect root growth patterns either (Ostonen et al., 2007; Rytte r and Rytter, 2012). There is some concern in the literature that sieving, as it affects N mineralization of moist soil could affect growing conditions (Hassink, 1994) However, the soils i n the current experimental site do not have many aggregates that c ould be broken up and are l ow in organic matter and N. So significant mineralization from sieving soil is less likely to occur. In a study using the mesh screen technology to measure fine root accumulation in Norway spruce ( Picea abies L. ) average root a ccumulation was 7. 9 kg ha 1 d 1 (Jentschke et al., 2001) In mesh screen studies, a blade is inserted into the soil and removed so a mesh screen can be installed in its place. It is excavated after a period of time and the roots that grow through the mesh are counted and weighed. Disadvantages to this technology are that soil can drop into the slit the blade makes or the mesh can be difficult to install as it can fold easily when pushed into the hole (Hirano et al. 2009) They attempted to improve this met hod by using two thin st ainless steel sheets to accompany the metal blade. The blade is removed, but the sheets stay in place until the mesh is installed, blocking any soil particles that might interfer e with the mesh insertion In all mesh screen methods, miscalculation in root accumulation may occur (Hirano et al., 2009). Because roots are cut prior to installation, advantageous root growth may be

PAGE 93

93 encouraged or cut roots may be stunted. Also, mesh screens or ingrowth cores cannot measure temporal informat ion on root growth, unli ke minirhizotron technology (Majdi et al., 2005). The current experiment attempts to overcome many of the challenges presented in the aforementioned studies In all of the ingrowth studies mentioned above, the cores were installed a t a 45 angle to capture maximum root growth, as compared with vertically installed cores. Soil compaction and high N is less of as they are low in nutrient s and composed mainly of fine sand. However, care was ta ken during installation to approximate soil conditions inside the ingrow t h core to th ose in the surrounding soil It was ensur ed that the soil inserted into the core was the same as that excavated from the hole and the soil was compacted to approximate the same bulk density. Imp lications Ona 33 and UF Peace had the slowest root accumulation rate, while decumbent Chico had the fastest. Apalachee, Ecoturf, and UF Tito were in the mid range of root accumulation rates. S ignificant differences were detected amo ng treatments, suggestin g the modifications made to the ingrowth core are likely suitable to be used i n subsequent studies. Choosing a mesh fabric with larger holes (4 mm) likely permitted a more accurate estimation of root rhizome accumulation, as the hol es were large enough to allow roots and rhizomes to grow into the core. Addition of t he cage wire superstructure over 100 d deployment periods in the soil

PAGE 94

94 Figure 5 1 A completed ro ot r hizome ingrowth c ore Photo co urtesy of the author.

PAGE 95

95 Figure 5 2 Cutting the cage wire material to size and shaping it into a cylinder for the ingrowth core. Photo courtesy of the author.

PAGE 96

96 Figure 5 3 B ottom and top of the ingrowth core before the polyester mesh is placed on it. Pho to courtesy of the author.

PAGE 97

97 Figure 5 4 Preparing the mesh covering of the ingrowth core. White material is polyester cargo netting utility fabric. Yellow material is a template made from transparent architectural drafting paper. Photo courtesy of the aut hor.

PAGE 98

98 Figure 5 5 Sewing the mesh cover ing of the ingrowth core with exterior grade thread. (Not pictured: pull drafting paper from mesh to remove) Photo courtesy of the author.

PAGE 99

99 Figure 5 6 Finished ingrowth core. Fabric is secured to cage wire by bei ng pulled through exposed ends. Exposed ends are then folded into the core. Photo courtesy of the author.

PAGE 100

100 Figure 5 7 Excavating soil for ingrowth core installation in a grass pasture. Photo courtesy of the author.

PAGE 101

101 Figure 5 8 Sieving and screening pl ant material (leaves, roots, rhizomes) from excavations. Photo courtesy of the author. Figure 5 9 Replacing ingrowth core with soil screened of roots and rhizomes into plots. Photo courtesy of the author. Figure 5 10 Cutting roots and rhizomes pro truding from the ingro wth core. Photo courtesy of the author.

PAGE 102

102 Figure 5 11 Washing soil from ingrowth core to obtain root rhizome mass Photo courtesy of the author.

PAGE 103

103 Table 5 1 Herbage accumulation and canopy height at harvest for six rhizoma peanut e ntries harvested twice per year in each of 2 yr, and root rhizome mass of the same entries measured at season end of 2 yr. Data are entry means across 2 yr. Entry means that are followed by the same letter are not different ( P > 0.05). Response variable En try P SE Ona 33 Apalachee Chico Ecoturf UF Peace UF Tito Herbage accumulation (Mg ha 1 ) 10.5 a 6.7 cdef 9.6 ab 6.8 cdef 7.9 bcde 7.8 bcde < 0.0001 0.47 Canopy height (cm) 22.9 ab 13.0 d 13.7 d 15.2 cd 17.5 bcd 23.9 a < 0.0001 1.21 Root rhizome m ass (Mg ha 1 ) 6.4 abc 5.7 bc 8.6 a 6.9 abc 6.9 abc 5.5 ab < 0.0001 1.29 Table 5 2 Ingrowth rate of root rhizome mass accumulation for six rhizoma peanut entries measured during 102 to 104 d during each of 2 yr in Hague, FL. Entry means that are follow ed by the same letter are not different ( P > 0.05). Response variable Entry P SE Ona 33 Apalachee Chico Ecoturf UF Peace UF Tito Ingrowth rate of accumulation ( k g ha 1 d 1 ) 10. 0 b 22.8 ab 32.8 a 22.4 ab 14.2 ab 23.0 ab 0.01 8.07

PAGE 104

104 CHAPTER 6 CONCL USIONS Rhizoma peanut ( RP; Arachis glabrata Benth.) is a perennial forage legume that has excellent nutritive value and demonstrated persistence when planted in monoculture or with grass species in the US Gulf Coast region tivar in use, but it has limitations including susceptibility to peanut stunt virus ( C ucumovirus spp.; Prine et al., 2010). The University of Florida has released several additional RP germplasms and cultivars in the past decade (Prine et al., 2010; Quesen berry et al., 2010) that are superior to Florigraze. Most RP cultivars were selected f or upright growth, favoring use for hay production, but there are germplasms and experimental lines that vary in growth habit and perhaps in tolerance to a greater range of soil drainage conditions. These lines may provide opportunity for greater use of RP under grazing or as an ornamental plant and perhaps increase the range of soil environments in which it can be grown Data describing key forage responses are lacking f or these entries, and evaluation is needed of forage productivity, nutritive value, an d stand persistence responses under different defoliation practices and in different soil environments The goals of the three projects described in this thesis are to e valuate forage responses of 14 RP introductions/selections, germplasms, and cultivars for use in the southeast US and to assess t he utility of a root ingrowth core device to measure root accumulation rate within a pasture context. The objective of the firs t project w as to quantify and compare the effects of defoliation frequency o n herbage production, nutritive value, and root rhizome characteristics of selected RP introductions with existing germplasms and cultivars. The second project was designed to char acterize herbage production, nutritive value, and root rhizome traits of sel ected

PAGE 105

105 RP introductions with those of existing cultivars and germplasms when grown at two locations with different soil characteristics one with well drained and another with seaso nally saturated soil. The objective of the third project was to develop and test a prototype root ingrowth core device designed to measure root rhizome accumulation rate of perennial forage species. Rhizoma Peanut Herbage Accumulation, Nutritive Value, an d Root Rhizome Responses to Defoliation Frequency Chapter 3 The experiment was conducted at the North Florida Research and Education Center in Quincy, FL from 2015 to 2017. Two defoliation frequencies were imposed on split plots of 14 RP entries and meas ures were made of above ground and below ground biomass, herbage nutritive v alue, and storage organ characteristics. The defoliation frequenc ies evaluated were one (1X; fall) or two (2X; summer and fall) harvests per season. Relatively long intervals betwe en defoliation events were chose n in order to address two primary scenarios: i) the difficulty of timely hay harvest in Florida because of frequent rain events during the growing season; and ii) situations in which RP is harvested for hay during the growin g season and also used as a source of rhizomes for vegetative propagation in the subsequent dormant season. Overall, the results suggested that greater self shading and disease pressure in the 1X treatment likely affected many of the response variable s H erbage accumulation was nearly twice as great in the 2X treatment for many e ntries However, herbage accumulation of the Quincy Beta entry was not affected by defoliation frequency, probably because it ha s greater disease tolerance than other entries Cano py height was not affected by defoliation frequency. While previous experime nts suggest that Ecoturf has a decumbent growth habit with high herbage bulk density this was not observed in the current study. Since all defoliation treatments in the current

PAGE 106

106 ex periment were relatively infrequent Ecoturf canopy height and bulk density were not affected by defoliation frequency and it was generally intermediate in these traits This supports the conclusion of Shepard et al. (2018) that Ecoturf is a phenotypicall y plastic RP line that ad o pts a short dense canopy under frequent, close defoliation, but when defoliated in frequently it does not demonstrate these characteristics. The 1X treatment was associated with lesse r crude protein and in vitro digestibility, most likely due to disease ind uced senescence and leaf drop in the 1X plots and the effects of greater maturity The 2X treatment caused root rhizome mass to decrease when compared with 1X, thus multiple defoliation events are not recommended for fields that w ill subsequently be used a s source fields for RP rhizomes. D efoliation frequency did not affect rhizome diameter, although Arbrook and UF Peace produced the thickest rhizomes. Defoliation frequency did not affect root rhizome non structur al carbohydrate concentrat ion. However, given its effect on root rhizome mass the non structural carbohydrate pool was almost twice as high in the 1X treatment as in the 2X. Based on this study it is concluded that RP hay producers should prioritize obtai ning at least a summer and a fall harvest from their hay fields, even if summer weather conditions are not optimal for drying. P reservation as haylage or balage, options that require less field drying, should be explored as an alternative when drying condi tions are challenging Tho se producers wishing to dig rhizomes from hay fields should keep in mind that defoliation, especially more than one defoliation event per year will reduce the amount of rhizome mass available for digging the following winter or sp ring. Finally, the impact of harvest frequency varies among entries of RP, with Quincy Beta being particularly tolerant of infrequent harvests while m ore upright

PAGE 107

107 growing entries (e.g., Arbrook and Tito ) and those susceptible to leaf diseases (e.g., Ec oturf) are likely to show some of the greatest negative impacts associated with a single vs. multiple harvests per year. Above and Below ground responses of Rhizoma Peanut Experimental Lines and Cultivars when Grown at Two Locations Differing in Soil Char acteristics Chapter 4 Th e experiment was conducted at two locations: the North Florida Research and Education Center in Quincy, FL and the Agronomy Forage Research Unit in Hague, FL. The Quincy location is characterized by well drained soils and the Hagu e location by seasonally s aturated soils. Above and below ground biomass and chemical composition responses were measured on 14 introductions/selections, germplasms, and cultivars of RP. More favorable s oil drainage at Quincy resulted in generally greater herbage accumulation than at Hague. However, herbage accumulation of the entry Ona 33 did not differ between sites, suggesting that it may be better suited to poorly drained soils than most RP lines Greater HA and associated dilution of N at Quinc y resul ted in greater herbage crude protein at Hague. In general, the longer regrowth periods in this study than most previous studies resulted in lesser herbage crude protein and in vitro digestible organic matter concentrations in the current experiment. The lo nger defoliation interval likely had the greatest negative effect on nutritive value of Ecoturf, which had the greatest incidence of leaf rust. Below ground characteristics varied among entries, with Arbrook maintain ing a large amount of below groun d bioma ss with very thick rhizomes and high concentrations of TNC. In contrast, entries like Apalachee and Cowboy ha d lesser rhizome mass, rhizome diameter, and TNC concentration.

PAGE 108

108 These data suggest that there may be potential to expand the current zone of adapta tion of RP into wetter soil environments by evaluating additional entries and releasing new cultivars. Additionally, there are a wide range of above and below ground growth characteristics available among entries being tested, providing opportunity for ta rgeting entries to specific intended uses or production goals. Design and Use of a Root Ingrowth Core Device to Measure Root Accumulation Rate of Perennial Forage Species Chapter 5 A modified design for a root ingrowth core was developed and test ed in t his study. The major differences in this ingrowth core design included a larger mesh size of the polyester fabric lining the ingrowth core and a wire mesh support structure to enhance the ability of the core to sustain its shape and volume throughou t the i ngrowth period. The cores were then tested in an experiment measuring root mass accumulation of six RP entries at the Hague location during 2 yr. Root accumulation rates varied among treatments with l ow growing Chico having more rapid root accumula tion ra te than Ona 33 and UF Peace. Entries Apalachee, Ecoturf, and UF Tito were intermediate in root mass accumulation rate Enlarging the fabric holes and reinforcing the walls with wire allowed rhizomes and roots to grow into the core, while maintaining the ge ometry of the core Significant differences were detected among treatments, suggesting this ingrowth core design may contribute to subsequent root rhizome mass accumulation studies, including those comparing single species with similar root architec ture. Implications of Research Twice a year defoliation is not frequent enough to reduce the quality of rhizomes for planting material, but it can reduce root rhizome mass Although once a year harvest may be

PAGE 109

109 beneficial to root rhizome mass for nursery pr oductio n, it lowers forage nutritive value and herbage accumulation of most RP entries, especially in upright growing lines Thus, RP hay producers should prioritize at minimum a summer and a fall harvest from their hay fields. Compared with other RP entri es curr ently being used commercially or being evaluated for potential release, Ona 33 is be tter adapted to seasonally saturated soils Ecoturf is likely phenotypically plastic, assuming a shorter canopy height with greater herbage bulk density when frequen tly def oliated, but not differing in growth characteristics from other intermediate types when intervals between defoliation events are long. Future Research Needs Given t hat the recommended minimum of twice per year harvest for RP hay production is compli cated b y summer conditions, options that require less field drying, like haylage or balage, should be explored As a range of above and below ground characteristics have been identified in RP entries, choice of RP entr y to be planted can be tailored to managemen t goals and desired production outcomes. Fu r thermore, additional entries and new cultivars may be tested and released that expand the current zone of RP adaptation into wetter soil environments. The root ingrowth core developed i n this thesis likely can be used to study root rhizome mass accumulation in a wide range of perennial pasture species and under numerous management practices As mild winters in Florida allow for year round forage production, the ingrowth core should be us ed to investigate the relat ionship between defoliation, root accumulation and seasonality

PAGE 110

110 APPENDIX ABOVE AND BELOW GROUND RESPONSE VARIABLES FOR THE RHIZOMA PEANUT ENTRY CHICO Herbage accumulation, canopy height, herbage bulk density, herbage crude protein, and herbage in vitro digestible organic matter at harvest for rhizoma peanut entry C hico and the average of all entries harvested twice per year in each of 2 yr at the Hague location R oot rhizome mass, root rhizome diameter, TNC concentration, and TNC pool of the same entries were measured at season end of 2 yr. Data are entry means across 2 yr Response variable Mean for Chico Overall Hague location mean Herbage accumulati on (Mg ha 1 ) 9.6 7.4 Canopy height (cm) 13.7 17.3 Herbage bulk density (kg ha 1 cm 1 ) 572 333 Herbag e crude protein (g kg 1 ) 196 17 0 Herbage in vitro digestible organic matter (g kg 1 ) 661 615 Root rhizome mass (Mg ha 1 ) 8.6 6.6 Root rhizome total nonstructural carbohydrates (g kg 1 ) 222 266 Root rhizome pool (kg ha 1 cm 1 ) 1970 178 0 Rhizome diamete r (mm) 2. 4 2.6 RP entries at Hague include d Ona 33, Q uincy Alpha, Apalachee, Arblick, Arbrook, Quincy Beta, Chico, Cowboy, Ecoturf, Florigraze, UF Peace, Pointed Leaf, Quincy UF Tito, and Waxy Leaf

PAGE 111

111 WORKS CITED Adjei, M. B., Mislevv, P., & Ward, C. Y. (1980). Response of tropical grasses to stocki ng rate. Agron. J. (72), 863 868. Allard, V., Soussana, J., Falcimagne, R., Berbigier, P., Bonnefond, J., Ceschia, E., . C., P. P. (2007). The role of grazing management for the net biome productivity and greenhouse gas budget (CO2, N2O, and CH4) of semi natural grassland. Agric. Ecosy s. Environ. (121), 47 58. Ammann, C. C. (2007). The carbon budget of newly established temperate grassland depends on management intensity. Agric. Ecosys. En viron (121), 5 20. Ammann, C., Spirig, C., Leifeld, J., & Neftel, A. (2005). Assessment of the nitr ogen and carbon budget of two managed temperate grassland fields. Agric. Ecosys. Environ (133), 150 162. Archimde, H., Eugne, M., Magdeleine, C., Boval, M. Martin, C., Morgavi, D., . M., D. (2001). Comparison of methane production between C3 and C4 grasses and legumes. Anim. Feed Sci. Technol (166), 59 64. Augustine, D. J., Dijkstra, F. M., Iii, E. W., & A., M. J. (2011). Rhizosphere interactions, carbo n allocation, and nitrogen acquisition of two perennial North American grasses in response to defol iation and elevated atmospheric CO2. Oecologia (165), 755 770. Baker, C., Blount, A., & K., Q. (1999). Peanut stunt virus infecting perennial peanuts in Flor ida and Georgia. Plant Pathology Circular 395. Beltranena, R., Breman, J., & Prine, G. (1981). Yi eld and quality of Florigraze rhizoma peanut (Arachis glabrata Benth.) as affected by cutting height and frequency. Proc. Soil and Crop Soc. Florida (40), 153 156. Blount, A., Sprenkel, R., Pittman, R., Smith, B., Morgan, R., Dankers, W., & Momol, T. (2006 ). Peanut stunt virus reported on perennial peanut in North Florida and Southern Georgia. Retrieved 2 12, 2015, from UF/IFAS EDIS. SS AGR 37. www.edis. ifas. ufl.edu. Bowman, A., Wilson, G., & Gogel, B. (1998). Evaluation of perennial peanuts (Arachis spp. ) as forage on the New South Wales north coast. Tropical Grasslands (32), 252 258. Bradford, M., Keiser, A., Davies, C., Mersmann, C., & Strickland, M. (2013 ). Empirical evidence that soil carbon formation from plant inputs is positively related to microbi al growth. Biogeochemistry (113), 271 281. Butler, T., Muir, J., Islam, M., & Bow, J. (2007). Rhizoma peanut yield and nutritive value are influenced by harv est technique and timing. Agron. J. (99), 1559 1563.

PAGE 112

112 Castillo, M., Sollenberger, L., Blount, A., Ferrell, J., Na, C., Williams, M., & Mackowiak, C. (2014). Seedbed preparation techniques and weed control strategies for strip planting rhizoma peanut into wa rm season grass pastures. Crop Sci (54), 1868 1875. C astillo, M., Sollenberger, L., Blount, A., Ferrell, J., Williams, M., & Mackowiak, C. (2013a). Strip planting a legume into warm season grass pasture: Defoliation effects during the year of establishment Crop Sci (53), 724 731. Castillo, M., Sollenberger, L., Ferrell, J., Blount, A., Williams, M., & Mackowiak, C. (2013b). Strategies to control competition to strip planted legume in a warm season grass pasture. Crop Sci (53), 2255 2263. Chaparro, C. J., S ollenberger, L. E., & Quesenberry, K. H. (1996). Ligh t interception, reserve status, and persistence of clipped Mott elephantgrass swards. Crop Science (36.3), 649 655. Chen, S., Lin, S., Reinsch, T., Loges, R., Hasler, M., & Taube, F. (2015). Comparison o f ingrowth core and sequential core methods for estim ating belowground net primary production in grass clover swards. Grass and Forage Science 515 528. Christiansen, S., Ruelke, O. C., R., O. W., & H., Q. K. (1988). Seasonal yield and quality of Trop. Agric. ( 65), 49 55. Conant, R., Paustian, K., & Elliott, E. (2001). Grassland management and conversion in to grassland: Effects on soil carbon. Ecol. Appl. (11), 343 355. Cong, W. F., Ruijven, J. v., Mommer, L ., Deyn, G. D., Berendse, F., & Hoffland, E. (2014). Plant species richness promotes soil carbon and nitrogen stocks in grasslands without legumes. J. Ecol. (102), 1163 1170. De Deyn, G., Shiel, R., Ostle, N., McNamara N., Oakley, S., Young, I., . Bar dgett, R. (2011). Additional carbon sequestration benefits of grassland diversity restoration. J. Appl. Ecol. (48), 600 608. DeDeyn, G., Quirk, H., Yi, Z., Oakley, S., Ostle, N., & Bardgett, R. (2009). Vegetation compo sition promotes carbon and nitrogen st orage in model grassland communities of contrasting soil fertility. J. Ecol. (97), 864 875. Dubeux Jr, J., Sollenberger, L., Comerford, N., Scholberg, J., Ruggieri, A., Vendramini, J., . Portier, K. (2006). Managem ent intensity affects density fraction s of soil organic matter from grazed bahiagrass swards. Soil Biol. Biochem (38), 2705 2711.

PAGE 113

113 Dubeux, J. J., Blount, A., Mackowiak, C., Santos, E., Neto, J. P., Riveros, U., . Ruiz Moreno, M. (2017). Biological N2 fi xation, belowground responses, and for age potential of rhizoma peanut cultivars. Crop Sci (57), 1027 1038. Dunavin, L. (1992). Florigraze rhizoma peanut in association with warm season perennial grasses. Agron. J. (84), 148 151. Fog, K. (1988). The effect of added nitrogen on the rate of decom position of organic matter. Biol. Rev. (63), 433 462. Follett, R. (2001). Soil management concepts and carbon sequestration in cropland soils. Soil Tillage Res (61), 77 92. Fornara, D. A., and D. Tilman. 2008. Plant fu nctional composition influences rates of soil carbon and nitrogen accumulation. J. Ecol. 96:314 322. Foster, J., Adesogan, A., Carter, J., Sollenberger, L., Blount, A., Myer, R., . Maddox, M. (2009). Annual legumes for forage systems in the United Sta tes gulf coast region. Agron. J. (101), 415 421. Franzluebbers, A., & Stuedemann, J. (2008). Early response of soil organic fractions to tillage and integrated crop livestock production. Soil Sci. Soc. Am. J. (72), 613 625. Franzluebbers, A., & Stuedemann, J. (2009). Soil profile organic carbo n and total nitrogen during 12 years of pasture management in the Southern Piedmont USA. Agric. Ecosyst. Environ. (129), 28 36. Freire, M., Kelly Begazo, C., & Quesenberry, K. (2000). Establishment, yield, and competit iveness of rhizoma perennial peanut ge rmplasm on a flatwoods soil. Proc. Soil and Crop Soc. Florida (59), 68 72. French, E., Prine, G., & Blount, A. (2006). Perennial peanut: An alternative forage of growing importance. Fla. Agric. Expt. Stn. Bull. SS AGR 99. Univ. Florida, Gainesville. Gal laher, R. N. (1975). An aluminum block digester for plant and soil analysis. Soil Sci. Soc. Am. J. (39), 803 806. Hafner, S., Unteregelsbacher, S., Seeber, E., Lena, B., Xu, X., Li, X., . Kuzyakov, Y. (2012). Effect of grazing on carbon stocks and assi milate partitioning in a Tibetan montane pasture revealed by 13CO2 pulse labeling. Global Change Bio (18), 528 538. Hassink, J. (1994). Effects of soil texture and grassland management on soil organic C and N and rates of C and N mineralization. Soil Biolo gy and Biochemistry 1221 1231.

PAGE 114

114 Hernndez Garay, A., Sollenberger, L., Staples, C., & Pedreira, C. (2004). Florigraze and Arbrook rhizoma peanut as pasture for growing Holstein heifers. Crop Sci. (44), 1355 1360. Hirano, Y., Noguchi, K., Ohashi, M., Hishi, T., Makita, N., Fujii, S., & Fi ner, L. (2009). A new method for placing and lifting root meshes for estimating fine root production in forest ecosystems. Plant Root 26 31. Horsfall, J. G., & Barratt, R. W. (1945). An improved grading system for measurin g plant disease. Annual Review o f Phytopathology 137 150. Interrante, S., Sollenberger, L., Blount, A., Coleman, S., White, U., & Liu, K. (2009). Defoliation management of bahiagrass germplasm affects cover and persistence related responses. Agron. J. (10 1), 1381 1387. Jensen, E. S., P eoples, M. B., Boddey, R. M., Gresshoff, P. M., Hauggaard Nielsen, H., Alves, B. J., & Morrison, M. J. (2012). Legumes for mitigation of climate change and the provision of feedstock for biofuels and biorefineries. A review. Agron. Sustainable Development ( 32), 329 364. Jentschke, G., Drexhage, M., Heinz Werner, F., Fritz, E., Schella, B., Lee, D., . Godbold, D. L. (2001). Does soil acidity reduce subsoil rooting in Norway spruce (Picea ablies)? Plant and Soil, 237 91 10 8. Klumpp, K., Bloor, J. M., Am bus, P., & Soussana, J. (2011). Effects of clover density on N2O emissions and plant soil N transfers in a fertilised upland pasture. Plant Soil (343), 97 107. Kohmann, M., Sollenberger, L., J.C.B. Dubeux, J., Silveira, M., Moreno, L., Silva, L. d., & Arya l, P. (2018). Nitrogen fertilization and proportion of legume affect litter decomposition and nutrient return in grass pastures. Crop Sci. (58), 2138 2148. Lscher, A., Mueller Harvey, I., Soussana, J., Rees, R., & Peyraud, J. (2014). Potential of legume b ased grassland livestock systems in Europe: A review. Grass Forage Sci (69), 206 228. Lal, R., Follett, R., Kimble, J., & Cole, C. (1999). Managing US cropland to sequester carbon in soil. Journal Soil Water Conserv (54), 374 381. Madritch, M. D., & Hunter M. D. (2002). Phenotypic diversity influences ecosystem functioning in an Oak sandhills community. Ecology 83:2084 2090. Magdoff, F., Weil, R., & Ray, R. (2004). Soil organic matter management strategies. In F. Magdoff, Soil Organic Matter in Sustainab le Agriculture (pp. 45 65). New York: CRC Press. Majdi, H., Pregitzer, K., Moren, A. S., Nylund, J. E., & Agren, G. I. (2005). Measuring fine root turnover in forest ecosystems. Plant and Soil 1 8.

PAGE 115

115 Makkonen, K., & Helmisa ari, H. (n.d.). Assessing fine r oot biomass and production in a Scots pine stand comparison of soil core and root ingrowth core methods. Plant and Soil (210), 43 50. Mislevy, P., Williams, M., Blount, A., & Quesenberry, K. (2007). Influence of Harvest Ma nagement on Rhizoma Perennial Pe anut Production, Nutritive Value, and Persistence on Flatwoods Soils. Forage and Grazinglands 5. doi:10.1094/FG 2007 1108 01 RS Montagnoli, A. M., Terzaghi, M., Scippa, G. S., & Chiatante, D. (2014). Heterorhizy can lead t o underestimation of fine root p roduction when using mesh based techniques. Acta Oecologica 84 90. Moore, J., & Mott, G. (1974). Recovery of residual organic matter from in vitro digestion of forages. J. Dairy Sci (57), 1258 1259. Muir, J., Pitman, W., & Foster, J. (2011). Sustainable, low input, warm season, grass legume grassland mixtures: mission (nearly) impossible? Grass and For. Sci. (66), 301 315. Mullenix, M., Sollenberger, L., Blount, A., Vendramini, J., Silveira, M., & Cas tillo, M. (2014). Growt h habit of rhizoma peanut affects establishment and spread when strip planted in bahiagrass pastures. Crop Sci. (54), 2886 2892. Mullenix, M., Sollenberger, L., Wallau, M., Blount, A., Vendramini, J., & Silveira, M. (2016a). Herbage accumulation, nutritive value, and persistence responses of rhizoma peanut cultivars and germplasm to grazing management. Crop Sci. (56), 907 915. Mullenix, M., Sollenberger, L., Wallau, M., Rowland, D., Blount, A., Vendramini, J., & Silveira, M. (2016b). Sward structure, light interception, and rhizome root responses of rhizoma peanut cultivars and germplasm to grazing management. Crop Sci. (56), 899 906. Nadelhoffer, K. J., Johnson, L., Laundre, J., Giblin, A., & Shaver, G. (2002). Measuring fine root tur nover in forest ecosyst ems. Plant and Soil 107 113. Nelson, N. (1944). A photometric adaptation of the Somogyi method for the determination of glucose. J. Biol. Chem (153), 375 379. Ortega S., J., Sollenberger, L., Bennett, J., & Cornell, J. (1992a). Rhizome characteristics an d canopy light interception of grazed rhizoma peanut pastures. Agronomy Journal (84.5), 804 809. Ortega S., J., Sollenberger, L., Quesenberry, K., Cornell, J., & C.S. Jones, J. (1992b). Productivity and persistence of rhizoma peanut pastures under differen t grazing managements. Agron. J. (84), 799 804.

PAGE 116

116 Ostonen, I., Puttsep, U., Biel, C., Alberton, O., Bakker, M. R., Lohmus, K., . Brunner, I. (2007). Specific r oot length as an indicator of environmental change. Plant Biosystems 426 442. Phillips, R., F inzi, A., & Bernhardt, E. (2011). Enhanced root exudation induces microbial feedbacks to N cycling in a pine forest under long term CO2 fumigation. Ecology Lette rs (14), 187 194. Poppi, D., & McLennan, S. (1995). Protein and energy utilization by ruminants at pasture. J. Anim. Sci (73), 278 290. Prine, G., Dunavin, L., Glennon, R., & Roush, R. (1986). Arbrook rhizoma peanut: A perennial forage legume. Fla. Agric. Stn. Circ. S 332. University of Florida, Gainesville, FL. Prine, G., Dunavin, L., Moore, J., forage legume. Fla. Agric. Stn. Circ. S 275. University of Florida, Gainesville, FL. Prine, G., Fr ench, E., Blount, A., Williams, M., & Quesenberry, K. (2010). Registration of Arblick and Ecotu rf rhizoma peanut germplasms for ornamental or forage use. J. Plt. Reg. (4), 145 148. Quesenberry, K., Blount, A., Mislevy, P., French, E., Williams, M., & Prine G. (2010). Registration of 'UF Tito' and 'UF Peace' rhizoma peanut cultivars with high dry ma tter yields, persistence, and disease tolerance. J. Plt. Reg (4), 17 21. Rasse, D., Rumpel, C., & Dignac, M. F. (2005). Is soil carbon mostly root carbon? Mechan isms for a specific stabilisation. Plant Soil (269), 341 356. Redfearn, D., Venuto, B., & Pitma n, W. (2001). Nutritive value responses of rhizoma peanut to nitrogen and harvest frequency. Agron J (93), 107 112. Resh, S., Binkle, D., & Parrotta, J. (2002). Greater soil carbon sequestration under nitrogen fixing trees compared with Eucalyptus species. Ecosystems (5), 217 231. Rice, R., Sollenberger, L., Quesenberry, K., Prine, G., & French, E. (1995). Defoliation e ffects on rhizoma perennial peanut rhizome characteristics and establishment performance. Crop Sci (35), 1291 1299. Rice, R., Sollenberger, L., Quesenberry, K., Prine, G., & French, E. (1996). Establishment of rhizoma perennial peanut with varied rhizome n itrogen and carbohydrate concentrations. Agron. J (88), 61 66. Romero, F., Horn, H. V., Prine, G., & French, E. (1987). Effect of cutting in terval upon yield,

PAGE 117

117 composition and digestibility of Florida 77 alfalfa and Florigraze rhizoma peanut. J. Anim. Sci (6 5), 786 796. Rouse, R., Roka, F., & Miavitz Brown, E. (2004). Guide for establishing perennial peanut as a landscape groundcover. Proc. Fla State Hort. Soc. (117), 289 290. Rusland, G., Sollenberger, L., Albrecht, K., C.S. Jones, J., & Crowder, L. (1988) Animal performance on limpograss aeschynomene and nitrogen fertilized limpograss pastures. Agron. J. (80), 957 962. Rytter, R. M., & Rytte r, L. (2012). Quantitative estimates of root densities at minirhizotrons differ from those in the bulk soil. Plant a nd Soil 205 220. Saldivar, A. J., Ocumpaugh, W. R., Gildersleeve, R. R., & Moore, J. E. (1990). Growth analysis of 'Florigraze' rhizoma pe anut: forage nutritive value. Agron. J., 82 473 477. Saldivar, A. J., Ocumpaugh, W. R., Gildersleeve, R. R., & Pri ne, G. E. (1992a). Total nonstructural carbohydrates and nitrogen of 'Florigraze' rhizoma peanut. Agron. J. 84 439 444. Saldivar, A. J., Ocumpaugh, W. R., Gildersleeve, R. R., & Prine, G. E. (1992b). Growth analysis of 'Florigraze' rhizoma peanut: Shoot and rhizome dry matter production. Agron. J. 84 444 449. Sanchez, J., Vendramini, J., Sollenberger, L., Silveira, M., Dubeux, J., Moriel P., . Olivera, F. L. (2018). Forage Characteristics of Bermudagrass Pastures Overseeded with Pintoi Peanut and Grazed at Different Stubble Heights. Crop Sci (58), 1 9. Santos, E., Dubeux, J., Sollenberger, L., Blount, A., Mackowiak, C., Dilorenzo, N. . Ruiz Moreno, M. (2018). Herbage responses and biological N2 fixation of bahiagrass and rhizoma peanut monocultures compared with their binary mixtures. Crop Sci, 58 1 15. Schultze Kraft, R., Rao, I., Peters, M., Clements, R., Bai, C., & Liu, G. ( 2018). Tropical forage legumes for environmental benefits: An overview. Trop. Grassl. Forrajes Tropicales, 6 1 14. doi:10.17138/TGFT(6)1 14 Scurlock, J. M., & Hall, D. O. (1998) The global carbon sink: a grassland perspective. Global Change Biology (4) 229 233. Shepard, E., Sollenberger, L., Kohmann, M., Silva, L. d., Dubeux, J., & Vendramini, J. (2018). Phenotypic plasticity and other forage responses to grazing management of Ecoturf rhizoma peanut. Crop Sci. (58), 2164 2173. Sollenberger, L., Agouridis, C., Vanzant, E., Franzluebbers, A., & Owens, L. (2012). Prescribed grazing on pasturelands. Conservation outcomes from pastureland and hayland practices. USDA NRCS 111 204.

PAGE 118

118 S ollenberger, L., C.S. Jones, J., & Prine, G. (1989) Animal performance on dwarf elephantgrass and rhizoma peanut pastures. In R. Desroches, Proc. Int. Grassl (pp. 1189 1190). Nice, France: Proc. Int. Grassl. Congress. Soussana, J., Talec, T., & Blanfort, B. (2010). Mitigating the greenhouse gas balance o f ruminant production systems through carbon sequestration in grasslands. Animal (4), 334 350. Spitaleri, R., Sollenberger, L., Schank, S., & Staples, C. (1994). Defoliation effects on agronomic performanc e of seeded Pennisetum hexaploid hybrids. Agron. J. (86), 695 698. Steingrobe, B., Schmid, H., & Claasen, N. (2000). The use of the ingrowth core method for measuring root production of arable crops influence of soil conditions inside the ingrowth core o n root growth. Journal of Plant Nutrition and Soil Science 617 622. Steingrobe, B., Schmid, H., & Claasen, N. (2001). The use of the ingrowth core method for measuring root production of arable crops influence of soil and root disturbance during instal lation of the bags on root ingrowth into the cores. European Journal of Agronomy 143 151. Stobbs, T. (1975). Factors limiting the nutritional value of grazed tropical pastures for beef and milk production. Trop. Grasslands (9), 141 150. Terrill, T., Gela ye, S., Mahotiere, S., Amoah, E., Miller, S., Gates R., & Windham, W. (1996). Rhizoma peanut and alfalfa productivity and nutrient composition in central Georgia. Agron. J. (88), 485 488. Tracy, B., Foster, J., Butler, T., Islam, M., Toledo, D., & Vendram ini, J. (2018). Resilience in forage and grazinglan ds. Crop Sci (58), 31 42. Van der Krift, T. A., & Berendse, F. (2002). Root life spans of four grass species from habitats differeing in nutrient ability. Functional Ecology 198 203. Venuto, B., Pitman, W., Redfearn, D., & Twidwell, E. (1999). Rhizoma Pe anut: A new forage option for Louisiana. Louisiana State Univ., Baton Rouge: Louisiana Agric. Exp. Stn. Circ. 136. Venuto, B., Redfearn, D., & Pitman, W. (1998). Rhizoma peanut responses to harvest freque ncy and nitrogen fertilization on Louisiana coastal plain soil. Agron. J. (90), 826 830. Whitman, A., Bailey, J. K., Schweitzer, J., Shuster, S. M., Bangert, R. K., LeRoy, C. J., . Wooley, S. (2006). A framework for community and ecosystem genetics: fr om genes to ecosystems. Nature Reviews Genetics 51 0 523.

PAGE 119

119 Williams, M. (1994). Reproductive resource allocation in rhizoma peanut. Crop Sci (34), 477 482. Williams, M., Quesenberry, K., Prine, G., & Olson, C. (2005). Rhizoma peanut more than a Proceedings of XX Internation al Grasslands Congress 228 229. Wilson, C., Vendramini, J., Sollenberger, L., & Flory, S. (n.d.). Root production in a subtropical grassland is mediated by cultivar and defoliation severity. Plant Soil (In Review). Wright, A. F. (2004). Long term manag ement impacts on soil carbon and nitrogen dynamics of grazed bermudagrass pastures. Soil Biol. Biochem (36 ), 1809 1816.

PAGE 120

120 BIOGRAPHICAL SKETCH Katie Cooley was born and raised in New England. She received her Bachelor of Arts in environmental a nalysis from S cripps College in Claremont, CA in May 2006 and her Master in Landscape Architecture from the Graduate Sc hool of Design at Harvard University in May 2013. Her interest in plant and soil science and passion for environmental stewardship motivated her to ear n an additional graduate degree in agricultural science In December 2018, she received her Master of Sci ence in a gronomy with a m inor in soil and water science