This item is only available as the following downloads:
1 STRATEGIES FOR INCREASING RHIZOMA PEANUT CONTRIBUTION TO PRODUCTIVITY AND ECOSYSTEM SERVICES OF LOW INPUT PASTURE SYSTEMS By MARY KIMBERLY MULLENIX A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORI DA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2013
2 2013 Mary Kimberly Mullenix
3 To my parents, Bill and Mary Cline
4 ACKNOWLEDGMENTS First, I would like to thank God for giving me the strength, knowledge, and perseverance to complete this degree. I especially would like to thank my major professor, Dr. Lynn E. Sollenberger, for his time, knowledge, and resource investment in helping guide, de velop, and implement my research program while at Florida. His leadership by example has enabled me to become a better scientist and communicator of research based information to academic and producer based audiences. I would also like to thank my committe e members Dr. Joo Vendramini, Dr. Maria Silveira, Dr. Ann Blount, and Dr. Grady Roberts for their advice and input in project development, implementation, and data analysis over the past three years. Thanks are also due to Mr. Dwight Thomas for his joyful demeanor and support of daily management of my research pro jects at the Beef Research Unit, and Dr. Kenny Woodard for his help and encouragement throughout my degree program. A special thanks is also extended to Mr. Jim Boyer, the farm crew at the Plant Science Research and Education Unit in Citra, planting, sampling, harvest management, and cattle handling for my soil carbon study. I would also like to thank Drs. Dian e Rowland and John Erickson for their willingness to let me use their laboratory resources during data collection for my experiments, and Mr. Richard Fethiere for his help and support in the Forage Lab. The continuous support rent and former graduate students Dr. Miguel Castillo, Dr. Chae In Na, Marcelo Wallau, Nick Krueger, and Daniel Pereira made much of the data collection for my experiments possible. I am also appreciative for the help from the numerous visiting scientists and graduate students who worked in the Sollenberger lab over the past three years.
5 Next, I would like to thank Dr. Ken Quesenberry and Mrs. Joyce Quesenberry for their personal and professional mentorship throughout my time in Gainesville. They have been a continuous resource of support throughout my program and an example of a successful marriage to me and my husband, Daniel Mullenix. I am forever grateful to my parents for their support and encouragement to pursue higher education, and for teaching and instilling a love and appreciation of agriculture within me. Finally, I would like to thank my husband, Daniel, for his continuing love, support, and personal sacrifice for me to be able to pursue my education at the doctoral level.
6 TABLE OF CONTENTS p age ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 10 LIST OF FIGURES ................................ ................................ ................................ ........ 11 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 15 2 LITERATURE REVIEW ................................ ................................ .......................... 19 Background ................................ ................................ ................................ ............. 19 History ................................ ................................ ................................ .............. 19 Genotypes of RP ................................ ................................ .............................. 20 Establishment ................................ ................................ ................................ ......... 23 Methods o f Establishment ................................ ................................ ................ 23 Factors Affecting Establishment ................................ ................................ ....... 24 Grazing Management ................................ ................................ ............................. 27 Mixed Legume/Grass Production Systems ................................ ............................. 32 Advantages and Challenges to Incorporating Legumes ................................ ... 32 Examples of Successful War m Season Grass Legume Pastures .................... 35 Rationale for the Use of Rhizoma Peanut Grass Pastures in the Southeast USA ................................ ................................ ................................ ............... 38 Ecosystem Service s and C Sequestration in Grasslands ................................ ....... 40 Grassland Management and Factors Affecting Soil Quality ................................ .... 40 Overview ................................ ................................ ................................ .......... 40 Grazing Management and Soil Quality ................................ ............................. 42 Hay Management and Soil Quality ................................ ................................ ... 45 Forage Manage ment and Effects on SOM Dynamics ................................ ...... 46 Summary ................................ ................................ ................................ ................ 49 3 GROWTH HABIT OF RHIZOMA PEANUT CULTIVARS AFFECTS ESTABLISHMENT AND SPREAD WHEN ST RIP PLANTED IN BAHIAGRASS SOD ................................ ................................ ................................ ........................ 51 Overview of Research ................................ ................................ ............................. 51 Materials and Methods ................................ ................................ ............................ 53 Experimental Site ................................ ................................ ............................. 53 Treatments and Experimental Design ................................ .............................. 54 Plot Establishment and Management ................................ ............................... 55 Response Variables ................................ ................................ ................................ 57
7 Shoot Emergence ................................ ................................ ............................. 57 Rhizoma Peanut Ground Cover ................................ ................................ ....... 58 Rhizoma Peanut Frequency ................................ ................................ ............. 58 Rhizoma Peanut Spread ................................ ................................ .................. 58 Bahiagrass Herbage Harvested ................................ ................................ ....... 59 Root rhizome to Shoot Ratio and Root rhizome Mass ................................ ..... 59 Soil Bulk Density and Total C, N, and C:N Ratio ................................ .............. 60 C Isotope Ratio Determination of C 3 and C 4 derived C and SOC Retained ..... 61 Statistical Analysis ................................ ................................ ................................ .. 62 Results and Discussion ................................ ................................ ........................... 63 Shoot Emergence ................................ ................................ ............................. 63 Rhizoma Peanut Ground Cover ................................ ................................ ....... 63 Rhizoma Peanut Frequency ................................ ................................ ............. 65 Rhizoma Peanut Spread ................................ ................................ .................. 66 Herbage Harvested ................................ ................................ .......................... 68 Root rhizome to Shoot Ratio and Root rhizome Mass ................................ ..... 68 Soil Bulk Density, Total C, N, and C:N Ratio ................................ .................... 69 C Iso tope Ratio Determination of C 3 and C 4 derived C and SOC Retained ..... 71 Implications of the Research ................................ ................................ ................... 72 4 SWARD CHARACTERISTICS OF RHI ZOMA PEANUT GENOTYPES UNDER A RANGE OF GRAZING MANAGEMENT STRATEGIES ................................ ...... 79 Overview of Research ................................ ................................ ............................. 79 Materials and Methods ................................ ................................ ............................ 80 Experimental Site ................................ ................................ ............................. 80 Treatments and Experimental Design ................................ .............................. 81 Plot Establishmen t and Management ................................ ............................... 82 Response Variables ................................ ................................ ................................ 84 Herbage Mass and Accumulation ................................ ................................ ..... 84 Nutritive Value ................................ ................................ ................................ .. 84 Pre grazing Sward Height ................................ ................................ ................ 85 Pre grazing Leaf to Stem Ratio ................................ ................................ ........ 86 Pre grazing Canopy Light Interception ................................ ............................. 86 Post grazing Residual Leaf Area Index ................................ ............................ 86 Sward Botanical Compositio n ................................ ................................ ........... 87 Rhizoma Peanut Ground Cover ................................ ................................ ....... 87 Weed Frequency ................................ ................................ .............................. 88 Rhizome Mass ................................ ................................ ................................ .. 88 Rhizome TNC and N ................................ ................................ ........................ 88 Statistical Analysis ................................ ................................ ............................ 89 Results and Discuss ion ................................ ................................ ........................... 89 Herbage Accumulation ................................ ................................ ..................... 89 Nutritive Value ................................ ................................ ................................ .. 91 Pre grazing Sward Height ................................ ................................ ................ 93 Pre grazing Leaf to Stem Ratio ................................ ................................ ........ 95 Pre grazing Canopy Light Interception ................................ ............................. 97
8 Post grazing Residual Leaf Area Index ................................ ............................ 97 Sward Botanical Composition ................................ ................................ ........... 98 Rhizoma Peanut Ground Cover ................................ ................................ ..... 100 Weed Frequency ................................ ................................ ............................ 101 Rhizome Mass ................................ ................................ ................................ 103 Rhizome TNC and N ................................ ................................ ...................... 104 Implications of the Research ................................ ................................ ................. 106 5 DEFOLIATION MANAGEMENT AND SOIL CARBON DYNAMICS OF VARIOUS PRODUCTION SYSTEMS BASED ON WARM SEASON PERENNIAL FO RAGES ................................ ................................ ....................... 116 Overview of Research ................................ ................................ ........................... 116 Materials and Methods ................................ ................................ .......................... 117 Experi mental Site ................................ ................................ ........................... 117 Treatments and Experimental Design ................................ ............................ 118 Plot Establishment and Management ................................ ............................. 120 Response Variables ................................ ................................ .............................. 122 Herbage Mass, Accumulation, and Harvested ................................ ............... 122 Herbage Nutritive Value ................................ ................................ ................. 123 Residual Litter Mass, C and N Concentration, and C:N Ratio ........................ 124 Root rhizome Mass ................................ ................................ ........................ 124 Soil C and N ................................ ................................ ................................ ... 124 Aggregate Size Distribution ................................ ................................ ............ 125 Statistical Analysis ................................ ................................ ................................ 126 Results and Discussion ................................ ................................ ......................... 126 Plant Responses ................................ ................................ ............................ 126 Herbage harvested ................................ ................................ .................. 1 26 Nutritive value ................................ ................................ .......................... 128 Residual litter mass ................................ ................................ .................. 130 Residual litter C and N concentration and C:N ratio ................................ 132 Root rhizome mass ................................ ................................ .................. 135 Soil Responses ................................ ................................ .............................. 135 Aggregate size distribution ................................ ................................ ....... 135 Particle size fraction C, N, and C:N ratio ................................ .................. 136 Implications of the Research ................................ ................................ ................. 140 6 CONCLUSIONS ................................ ................................ ................................ ... 152 Growth Habit of Rhizoma Peanut Cultivars Effects on Establishment and Spread When Strip Planted in Bahiagrass Sod Chapter 3 .............................. 153 Sward Characteristics of Rhizoma Peanut Culvitars Under a Range of Grazing Management Strategies Chapter 4 ................................ ................................ .. 154 Defoliation Management Effects on Soil Carbon Dynami cs of Year Round Production Systems Chapter 5 ................................ ................................ ........ 155 Implications of the Research ................................ ................................ ................. 156 Future Research Needs ................................ ................................ ........................ 157
9 LIST OF REFERENCES ................................ ................................ ............................. 159 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 173
10 LIST OF TABLES Table page 3 1 Spread of RP entries during the year of ( P = 0.0086) and year after establishment ( P = 0.2424). ................................ ................................ ................ 74 3 2 Defoliation management effects on RP spread during the year of ( P = 0.0023) and year after establishment ( P = 0.0127). ................................ ........... 74 3 3 Soil organic C, N, and C:N ratio in the surface 20 cm of strip planted rhizoma peanut from 2011 to 2013. ................................ ................................ ................. 75 3 4 Percentage of C 4 and C 3 derived C in the surface 20 cm of strip planted rhizoma peanut from 2011 to 2013. ................................ ................................ .... 75 4 1 Date x entry interaction for rhizoma peanut leaf to stem ratio ( P = 0.0315) and post grazing residual leaf area index ( P = 0.0384; RLAI). ........................ 107 4 2 Entry x frequency interaction ( P = 0.0235) for rhizoma peanut leaf to stem ratio. ................................ ................................ ................................ ................ 107 4 3 Entry x frequency interaction for percentage of rhizoma peanut ( P = 0.0200) and grass ( P = 0.0355) in total herbage mass . ................................ ................ 107 5 1 Description of management treatments for warm season perennial based forage production systems. ................................ ................................ .............. 142 5 2 Species x system interaction ( P = 0.0013) for total rye herbage harvested (kg DM ha 1 ) from overseeded bermudagrass and rhizoma peanut. ..................... 142 5 3 Year x system interaction ( P < 0.0001) for total rye herbage harvested (kg DM ha 1 ) from overseeded bermudagrass and rhizoma peanut. ..................... 143 5 4 Species x system interaction ( P < 0.0001) for total herbage harvested (kg DM ha 1 ) during the 2012 summer season. ................................ ..................... 143 5 5 Year x system interaction for total C concentration in various soil aggregate size fractions ................................ ................................ ................................ .... 144 5 6 Year x system interaction for total N concentration in various soil aggregate size fractio ns ................................ ................................ ................................ .... 144
11 LIST OF FIGURES Figure page 3 1 Total monthly rainfall for 2011 and 2012 and 30 year average rainfall for the experimental location. ................................ ................................ ......................... 76 3 2 Date x entry interaction for rhizoma peanut ground cover (%) and frequency (%) during the year of establishment. ................................ ................................ 77 3 3 Date x defoliation treatment interaction for rhizoma peanut ground cover (%) and frequency (%) during the year of establishment. ................................ ......... 78 4 1 Season x entry interaction ( P = 0.0124) for herbage accumulation (kg DM ha 1 ) of RP genotypes.. ................................ ................................ ......................... 108 4 2 Season x grazing fre quency ( P = 0.0183) interaction for rhizoma peanut herbage accumulation.). ................................ ................................ ................... 109 4 3 Season x entry interaction ( P = 0.0004) for pre grazing sward height of RP genotypes under different levels of grazing management.. .............................. 110 4 4 Season x grazing frequency interaction ( P < 0.0001) for pre grazing rhizoma peanut sward height.. ................................ ................................ ....................... 111 4 5 Date x grazing frequency interaction ( P < 0.0001) for ground cover of rhizoma peanut entries.. ................................ ................................ ................... 112 4 6 Entry x grazing frequency interaction ( P = 0.0005) for ground cover of rhizoma peanu t entries. ................................ ................................ .................... 113 4 7 Date x entry interaction ( P = 0.0110) for frequency of weed occurrence in rhizoma peanut under different grazing regimes.. ................................ ............ 114 4 8 Date x grazing frequency interaction ( P < 0.0001) for frequency of weed occurrence in rhizoma peanut swards.. ................................ ............................ 115 5 1 Monthly rainfall for 2011 and 2012 for the experimental location and the 30 yr average for Citra, FL. ................................ ................................ ........................ 145 5 2 Date x system interaction ( P = 0.0172) for crude protein (g kg 1 DM) of winter rye overseeded forage systems. ................................ ................................ ....... 146 5 3 Date x species interaction ( P = 0.0317) for crude protein (g kg 1 DM) of ............ 147 5 4 Herbage in vitro digestible organic matter (IVDOM) of warm season perennial forage systems in summer 2012. ................................ ..................... 148
12 5 5 Year x species interaction ( P = 0.0002) for residual litter mass (kg ha 1 ) of winter overseeded bermudagrass (BG) and rhizoma peanut (RP).). ................ 149 5 6 Species x system interaction ( P = 0.0068) for residual litter mass (kg DM ha 1 ) of winter overseeded bermudagrass (BG) an d rhizoma peanut (RP). .......... 150 5 7 Species x system interaction ( P = 0.0382) for C:N Ratio of residual litter for winter overseeded bermudagrass (BG) and rhizoma peanut (RP). .................. 151
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy STRATEGIES FOR INCREASING RHIZOMA PEANUT CONTRIBUTIO N TO PRODUCTIVITY AND ECOSYSTEM SERVICES OF LOW INPUT PASTURE SYSTEMS By Mary Kimberly Mullenix December 2013 Chair: Lynn E. Sollenberger Major: Agronomy Livestock production in the USA Gulf Coast is based on nitrogen fertilized grass pastures, but inc reasing nitrogen cost threatens the future viability of this production system. Rhizoma peanut (RP; Arachis glabrata Benth.) is a regionally adapted perennial legume with documented long term persistence and ability to spread in grass pastures. To date, RP has been used primarily for high value hay production because establishment costs are high. Expansion to grazed pasture is desired, but lower cost establishment methods are needed. These projects were part of a larger research effort with the goal of prov iding technologies leading to sustainable grass RP pastures with no requirement for nitrogen fertilizer. Specific project objectives were to: 1) determine if strip planting RP into grass pastures is a viable alternative to current practice; 2) quantify gra zing tolerance of RP cultivars within the context of strip planting; and 3) measure soil quality benefits of conversion to RP based vs. traditional grass systems. When strip planted, Florigraze and Ecoturf RP had favorable sprout emergence, ground cover, frequency, and spread during the establishment year compared with Arb lick and Peace RP. Rotational stocking of establishing pastures every 28 d decreased
14 establishment success compared with a hay production system. Strip planting resulted in an initial decrease in soil C and N across the 2 yr of the present study. When established RP cultivars were grazed using different management strategies, there were no differences in herbage accumulation among cultivars in the first year, but. favorable changes in sward characteristics with less frequent and intensive grazing suggest that these strategies may favor long term stand production. Lastly, RP was used as the base forage in various production systems to determine its effect on soil quality compared with a grass nitrogen system. Following 1 yr of imposing the year round forage system treatments soil C had increased in the macroaggregate (2000 250 m) fraction, which best reflects short term contributions of organic matter inputs fr om the systems. Finally, overseeding RP and bermudagrass with an early maturing rye in this study did not negatively impact production of the warm season perennials, indicating its utility as a winter forage option for producers in Florida
15 CHAPTER 1 INT RODUCTION Rhizoma peanut (RP; Arachis glabrata Benth.) is a warm season perennial legume that is well adapted to the lower Coastal Plain region of the USA. Contribution of RP to pasture based livestock enterprises has been minimal due to the high cost of e stablishment, limiting the use of most planted areas to production of high value hay. In recent years, greater input costs in beef cow calf enterprises ha ve increased interest in the use of legumes as an alternative N source in grass pastures. In the sout heastern USA, pasture based systems rely on N fertilizer and are increasingly vulnerable to high fertilizer costs (Rouquette and Smith, 2010). Incorporation of legumes into grass based forage systems may provide a sustainable alternative for synthetic N fe rtilization, increase animal production by providing a source of high nutritive value forage (Sollenberger et al., 1989), and maintain pasture productivity (Lascano et al., 1989). While productive and persistent grass legume associations are well document ed in temperate environments (Laidlaw et al., 2001; Frame and Laidlaw, 2011), the use of legumes in tropical ecosystems has been less successful (Shelton et al., 2005). Competition between aggressive C 4 grasses and the C 3 legume (Dunavin, 1992), difficulty in legume establishment and maintenance in grass swards (Sollenberger and Kalmbacher, 2005), and disease susceptibility (Shelton et al., 2005) have limited the contribution of forage legumes in tropical pastures. However, well documented persistence of RP under a range of management strategies (i.e., haying or grazing; Ortega S. et al., 1992a) and ability to spread and persist in grass mixtures (Dunavin,
16 1992; Castillo et al., 2013) make this legume an ideal candidate for use in livestock production system s based on grazed pasture. Alternative establishment strategies are needed to reduce establishment costs if RP is to make significant contributions to pasture based systems in the future. One approach is strip planting the legume into grass swards (Castil lo et al., 2013). Because RP is a long lived perennial with ability to move laterally via an extensive rhizome system, it has potential to spread into the surrounding grass areas and form a mixed pasture over time. The amount of time needed for such a mixt ure to form is unknown and is likely dependent upon the RP genotype selected for use and the defoliation management imposed during the RP establishment phase. In the past decade, several genotypes of RP have been developed and released from the University of Florida as dual purpose hay and grazing crops (Prine et al., 2010; Quesenberry et al., 2010). These genotypes exhibit phenotypic variation in their growth habit from low growing to upright types. Growth habit may affect the success of establishment usi ng the strip planting approach and also their response to a range of grazing management strategies. While studies have illustrated the yield and nutritive value potential of some of these entries under clipping (Mislevy et al., 2007; Quesenberry et al., 20 10), no studies have evaluated plant responses of genotypes with varying growth habits to a range of grazing management strategies. Because long term pasture productivity and persistence are of great importance in low input systems, these data are needed t o guide cultivar selection by producers. Another potential benefit of incorporating legumes into livestock production systems is to sustain or enhance the ecosystem services provided by grasslands
17 (Boddey et al., 2004). Ecosystem services go beyond providi ng a feed source for livestock, and particularly include those which contribute to sustained, effective functioning of the agroecosystem. For example, managed grasslands can increase soil C and long term sustainability of forage based ecosystems through th e improvement of soil quality, and RP has potential to contribute significantly in this regard in Florida (French et al., 2006). Tropical and temperate grasslands play a major role in the global C cycle and serve as an important C sink (Scurlock and Hall, 1998) with as much as 90% of the grassland C pool being belowground (Schuman et al., 2002; Liu et al., 2011a). Improved forage management strategies such as fertilization and proper grazing management increase aboveground biomass production for livestock, but can also contribute to belowground C pools (Liu et al., 2011b; Silveira et al., 2013). The mild climate of Florida can provide favorable growing conditions for both warm and cool season forages, and selection of species and defoliation strategies for use throughout the winter and summer seasons has been shown to influence contribution to soil C (Franzluebbers et al., 2000; Franzluebbers et al., 2002). Currently, no work has been done in Florida to compare the potential C contribution of legume and gr ass based, year round forage systems. Within this context, evaluation of the role of winter overseeding and utilization of forage by grazing or as hay is also important. Identifying year round forage systems that promote soil C sequestration while maintain ing aboveground production will enhance ecosystem function and sustainability of Florida livestock grassland agroecosystems. The overall objective of the dissertation research was to develop management strategies for increasing RP contribution to grazing systems in Florida and to determine
18 ecosystem services associated with management of legume based forage systems. Specific project objectives were to 1) determine the viability of strip planting RP into grass pastures and quantify rate of establishment a nd spread of RP genotypes under defoliation during the establishment year (Chapter 3); 2) evaluate the productivity, persistence, and nutritive value of recently released RP genotypes under differing levels of grazing intensity and frequency (Chapter 4), a nd 3) measure the effect on soil C sequestration of RP vs. bermudagrass ( Cynodon spp.) based year round forag e production systems (Chapter 5).
19 CHAPTER 2 LITERATURE REVIEW Background History Rhizoma peanut (RP; Arachis glabrata Benth.) is a warm season, perennial legume that is well adapted to Florida and the lower Coastal Plain region of the southern USA. In 1936, a collection of A. glabrata was first brought to the USDA National Plant G ermplasm System from Matto Grosso, Brazil (Quesenberry et al., 2010 ) Natural Resources Conservation Service. While major collections of A. glabrata continued through the 1960s and again in the 1980s, research efforts first began to evalua te potential forage use of the species in Florida in the 1960s (Prine et al., 2010). Rhizoma peanut is self pollinated and produces few viable seed (Prine et al., 1981). Unlike common peanut ( Arachis hypogea L.), RP is primarily propagated vegetatively by was developed to distinguish this species from Arachis pintoi Krapov. & W.C. Greg., also a perennial species, but stoloniferous in nature (Quesenberry et al., 2010). Because RP is slow to es tablish (Rice et al., 1995; Williams et al., 1997), most of the research with RP in Florida over the past 20 yr has focused on improved establishment methods and evaluation of new germplasm under different management practices (Quesenberry et al., 2010). M ultiple cultivars and germplasms of RP have been developed for use as forage crops and low input ornamentals. Although RP exhibits decumbent growth in nature (Muir et al., 2010), selection of RP accessions over the past 60 yr has led to the development of several entries that vary in growth habit.
20 Depending on the type and level of management, growth habit may play a significant role in persistence and production potential of a RP stand Evidence to support this hypothesis can be drawn from a comparison of erect growing Arbrook with intermediate growth habit Florigraze under continuous stocking Arbrook was l ess tolerant of continuous stocking than Florigraze (Hernndez Garay et al., 2004). Genotypes of RP Currently, there are five commercial cultivars ( Florigraze Arbrook USA. With the exception of Latitude 34, all of these were developed and released by the University of Florida. The first commercial cultivar, Florigraze (PI 421707), was released in 1979 and is characterized by an intermediate growth habit (Prine et al., 1981). Florigraze was selected as an off type, chance hybrid during early forage evaluation trials of RP lines in the 1960s (Quesenberry et al., 2010). In 1962, G. M. Prine observed et al., 1986). Upon the rel ease of Florigraze, commercial acceptance of RP began to increase. This is the most widely grown cultivar in Florida, with roughly 10,500 ha statewide, and increasing acreage being planted annually (French et al., 2006). In recent years, peanut stunt virus ( Cucumovirus sp.) has been reported in Georgia and Florida, which has been shown to negatively affect production of Florigraze RP (Blount, personal communication, 2013). Because such a large proportion of the RP acreage is planted to Florigraze, there is a need to diversify the genetic base by developing and evaluating new germplasm. Arbrook (PI 262817) originated from RP germplasm collected in Paraguay in 1960 and was released in 1986 as a cultivar adapted to the deep, droughty sands of
21 Florida (Pri ne et al., 2010). The cultivar was selected from row evaluation plots of PI 212817 at the Arcadia and Brooksville, FL Plant Materials Centers and was thus given the name Arbrook (Prine et al., 1990). Arbrook is a larger plant type than Florigraze and exhib its an upright growth habit and shows favorable production in hay systems (Prine et al., 1986; Prine et al., 1990). Also, Arbrook has a more extensive, coarse root system than Florigraze, which gives it a greater drought tolerance compared to other commerc ially released cultivars. Although plant growth from rhizomes is slow, Arbrook provides more rapid upright growth than Florigraze. However, lateral spread from the plant is slower than Florigraze, which increases the time period to complete ground cover (P rine et al., 1990). Arbrook has been found to be less tolerant of continuous stocking than Florigraze (Hernndez Garay et al., 2004). Arblick and Ecoturf were both selected from early accessions of RP from the 1950s. Arblick was introduced to the USA as PI 262839 after collection by W.C. Gregory in Paraguay near the Brazil Paraguay border (Quesenberry et al., 2010). The entry was later given its name by the USDA Soil Conservation Service and was included in early evaluations of RP. Ecoturf (PI 262840) was a lso collected around the same time in Bela Vista, Brazil. Arblick and Ecoturf were included in the evaluation research leading to the selection and release of Florigraze; however, it was not until recently that Arblick and Ecoturf were released as germplas ms (Prine et al., 2010). Earlier research with these germplasms focused mainly on selection for dry matter yield and persistence when managed for forage (Prine et al., 2010). These lines were noted to be low growing types that may provide significant groun d cover and were released in 2008 for their potential ornamental use (Prine et al., 2010).
22 Cultivars UF Tito (PI 262826) and UF Peace (PI 658214) originated from plant introductions from Paraguay in the late 1950s. Both cultivars exhibit a growth habit s imilar to that of Florigraze and were released in 2008 after a 20 yr program of evaluation based on the observation of consistently high dry matter yields, persistence, and competitive ability with weedy grasses (Quesenberry et al., 2010). UF Tito was sel ected as the top producing line from a 10 yr evaluation trial by Friere et al. (2000). They stated that UF Tito had the greatest percentage of pure peanut and less invasion by common bermudagrass ( Cynodon dactylon L.) than other evaluated lines. Rate of sp read was also greater for UF Tito when compared to other lines (Friere et al., 2000). At the end of the evaluation, UF Tito had greater than 90% ground cover, indicating the competitiveness of the line. UF Peace was also selected from this evaluation as ha ving superior forage characteristics (Friere et al., 2000). UF Peace produced similar DM yields to UF Tito, although competitiveness with weeds and rate of spread were inferior to UF Tito. RP in the USA (Muir et al., 2010). The cultivar was selected from PI 262819, part of the germplasm collection housed at Stephenville, TX, and released by Texas AgriLife Research in 2009. The PI was originally collected near Trinidad, Paraguay and then take n to Stephenville, TX by C. E. Simpson (Muir et al., 2010). Latitude 34 was observed to have good forage potential, high pH tolerance, and persisted and spread aggressively in field trials at Beeville, TX and Stephenville, TX. Compared with other commercia lly available cultivars, Latitude 34 is noted for its relative cold and drought tolerance. Butler et al. (2006) observed persistence of Latitude 34 during years with freezing temperat ures
23 lasting as long as 48 h consecutively and where subsoil temperatures Additionally, Latitude 34 persisted and spread in field trials with as little as 454 mm of annual rainfall (Butler et al., 2006). Rhizoma peanut is often compared with alfalfa ( Medicago sativa L.) because of its yield potential (Andrews et al ., 1985), high nutritive value (Prine et al., 1981; Mislevy et al., 2007), and persistence under a variety of management conditions (Ortega S. et al., 1992a; Butler et al., 2007). When compared with alfalfa, average DM yield of Florigraze RP was 7.6 vs. 1 1.8 Mg ha 1 for alfalfa over a 3 yr period in central Georgia (Terrill et al., 1996). Mislevy et al. (2007) evaluated the influence of harvest management and RP entry on DM yield, nutritive value, root mass, and persistence on a moderately drained soil. Ov er the 4 yr trial, average DM yield of all RP entries was 11.8 and 8.9 Mg ha 1 when clipped to 2.5 and 10 cm stubble heights, respectively. Average crude protein (CP) concentration was 172 g kg 1 and IVDOM was 690 g kg 1 DM for all entries. During a 4 yr trial in Citra, FL, DM yield of Ecoturf, UF Tito, UF Peace, Florigraze, and Arbrook ranged between 8.3 and 12 Mg ha 1 illustrating the yield potential of RP (Prine et al., 2010). Establishment Methods of Establishment Rhizoma peanut is vegetatively propa gated from rhizomes because it produces very little viable seed (Quesenberry et al., 2010). Niles et al. (1989) suggested that producer reluctance to vegetatively establish RP ha d limited its production potential in Florida. Williams et al. (1997) stated t hat while small areas of land can be planted manually, larger areas must be planted with specialized equipment that is typically used for the establishment of vegetatively propagated tropical grasses. Rhizomes usually
24 form a 5 to 8 cm deep mat just below t he soil surface which can then be dug with a sprig harvester and planted as individual rhizome pieces or as sod pieces lifted with a sod lifter or by other means (Frenc h et al., 2006). The recommended planting time for RP is between January and March (Prin e at al., 1981), which is based on studies that have indicated highest sprout emergence during this period of time (Williams et al., 1993). Slow rate of cover is thought to be directly related to poor sprout emergence and survival of RP after planting (Wil liams et al., 1993). Because coverage rate appears to be dependent on the number of peanut sprouts per unit of land area that survive emergence and initial establishment, the suggested planting rate of RP has been increased to roughly 1000 kg rhizomes ha 1 (Williams et al., 1997). Additionally, planting should occur during a time of year with sufficient soil moisture and that maximizes the frost free period after planting (Williams et al., 1997). Factors Affecting Establishment A slow rate of establishmen t has been cited as the greatest factor limiting increased use of RP as forage (French, 1988; Saldivar et al., 1990). New germplasm evaluation of RP has focused on the development of RP genotypes with decreased time to establishment through increased sprou t emergenc e rate of spread, persistence, and competitiveness with weeds (Canudas et al., 1989; Quesenberry et al., 2010). While selection of genotypes is important, one of the most critical factors affecting establishment is adequate soil moisture at the time of planting. Williams et al. (1997) observed that reliable soil moisture, from rainfall or irrigation, is needed for 60 to 90 d post planting. In an establishment guide for RP, French and Prine (2002) suggest that irrigation be used when needed during the establishment phase if available and economical. Because the spring growing season in Florida can have limited rainfall
25 (Prine, 1981; Williams et al., 1993), adequate soil moisture may limit sprout emergence and rhizome survival in newly established s tands. Pre plant land preparation and time of planting may also influence rate of establishment. Williams et al. (1993) evaluated the effect of planting date and preplant tillage method on emergence and survival of RP. Three tillage intensities were evalu ated including plowed and well prepared for minimal grass competition, disked with moderate grass competition, and no preplant tillage where RP was planted directly into sod. After RP establishment, time to first sprout emergence was shorter for the plowed than disked or sod planted treatments. Based on these results, Williams et al. (1993) suggested that recommendations for establishing RP in well prepared fields during the winter should be cont inued. French et al. (2006) stated that land preparation shoul d begin in the late summer to allow for weed regrowth and subsequent elimination prior to a frost event. Preparing land early allows for adequate time for decomposition of plowed organic matter and provides time for accumulation of soil moisture and firmin g of the seedbed (French et al., 2006). Planting depth may also affect establishment potential of RP stands. Williams et al. (1993) observed a faster rate of sprout emergence in plowed plots where RP was shallowly planted and covered uniformly at the tim e of planting compared with disked and sod planted plots. In less well prepared plots, grass clumps prevented uniform coverage of rhizomes which were then susceptible to drying and dessication (Williams et al., 1993). French et al. (2006) recommended a 3.5 to 5 cm planting depth for sandy soils and a 2.5 cm depth for clay, and they indicated that different cultivars may have
26 more favorable emergence characteristics under deeper planting conditions than others depending on their rhizome characteristics. Ma nagement of rhizomes during the season prior to harvest and handling following harvest affect rhizome characteristics that are important to ensure successful establishment of RP (Rice et al., 1995; Venuto et al., 1999). Storage carbohydrate levels in rhizo mes have also been shown to influence establishment and are usually high during the late winter (Saldivar et al., 1992). Saldivar et al. (1992) noted that RP accumulates a high concentration (400 to 700 g kg 1 ) of total non structural carbohydrates (TNC) d uring the fall. If plants are to be used as a rhizome source for propagation, defoliation should be minimized during the preceding growing season in order to maximize TNC levels in the rhizomes (Saldivar et al., 1992). The effect of grazing management has been evaluated on rhizome chemical composition (Rice et al., 1995) and subsequent establishment success when these rhizomes were planted (Rice et al., 1996). Rhizomes with low TNC (62 g kg 1 ) resulted in stand failure during a drought year; however, rhizom es with initial TNC levels of at least 228 g kg 1 resulted in accumulation of rhizome and shoot mass following planting (Rice et al., 1996). Cost of establishment is an important consideration when choosing the amount of area to plant. It is estimated th at establishment of RP can cost as much as $1,200 per ha (Blount, personal communication, 2013). When compared with seeded forages, it is evident that the cost of vegetative establishment is expensive and may limit RP usage to production systems with the h ighest net returns (Williams et al., 2004). Subsequently, this constraint has limited the use of RP for grazing by beef cattle. Although RP provides long term productivity once established and an excellent source of high quality
27 forage, the cost associated with establishment, removing land from defoliation until RP is established, and weed control should be considered. Grazing Management Although several studies have described management of various RP entries under haying, fewer studies have considered th e influence of grazing management on RP. The response of plants to grazing is largely related to the type of grazing practice implemented. While many factors may influence plant responses to grazing, this review will focus mainly on plant growth habit, gra zing intensity, grazing frequency, and their interaction. Grazing intensity may be described in terms of stocking rate, grazing pressure, forage mass and allowance, or canopy height (Sollenberger et al., 2012). Plant response to various levels of grazing intensity is dependent on the species involved, frequency of grazing, and the environment (Sollenberger et al., 2012). Intensity of grazing has been suggested as a primary factor in determining plant productivity, persistence and sustainability in a given area (S ollenberger and Newman, 2007). Herbage accumulation (Ortega S et al., 1992a), sward botanical composition, animal performance (Stewart et al., 2005), and soil and water quality (Franzluebbers et al., 2000; Liu et al., 2011c) can be influence d greatl y by grazing intensity. Early clipping studies defined target stubble heights for Florigraze RP under defoliation. Prine et al. (1981) suggested an average height of at least 10 cm be maintained for Florigraze in order for adequate leaf tissue to be presen t to intercept light. In North Florida, Arbrook was found to be less tolerant of a 15 to 20 cm stubble height under continuous stocking than was Florigraze (Hernndez Garay et al., 2004). Under these conditions, Florigraze RP percentage in total forage ma ss remained relatively constant from Year 1 (90%) to
28 Year 3 (87%), whereas Arbrook percentage decreased from 89 to 66% over the 3 yr study as perennial grass contribution increased. Hernndez Garay et al. (2004) stated that the more upright growth habit of Arbrook makes it less tolerant of continuous stocking than Florigraze. Frequency of grazing and grazing method also play a large role in management of pasture systems, although stocking rate has been cited as having a greater influence on pasture product ivity than the grazing method used (Sollenberger and Newman, 2007). Sollenberger and Chambliss (1989) suggested that the most important tools for grazing management include the selection of the level of grazing intensity, as well as the frequency of grazin g. The combination of these management factors affects the productivity of forage per land unit, efficient use of forage by grazing animals, and persistence and productivity of a pasture system (Sollenberger and Chambliss, 1989). A grazing period of a defi ned length is often used along with a given level of intensity (i.e., stubble height, residual dry matter, etc.) to describe management regimes. The effect of frequency of defoliation of RP was first described under clipping by Prine et al. (1981) for the cultivar Florigraze. Based on these results, Prine et al. (1981) suggested a grazing period of no longer than 10 d under rotational stocking, with a 3 wk or longer rest period. If continuous stocking is to be used, the stocking rate should be low enough to maintain an average stubble height of 10 cm throughout the grazing period (Prine et al. 1981). In an animal performance trial, Sollenberger et al. (1989) rotationally stocked Florigraze RP pastures with a grazing period of 1 wk and rest periods of 5 wk. Steer 1
29 IVDOM; 170 to 220 g kg 1 CP) under these management conditions, illustrating the forage quality of Florigraze RP under rotational management. Studies have also il lustrated the interaction of grazing intensity with frequency. In Gainesville, FL, a 2 yr study evaluated the effects of three levels of postgraze residual dry matter (RDM, 500, 1500, and 2500 kg ha 1 ) with four intervals between grazing events (7, 21, 42, and 63 d) on the productivity and persistence of Florigraze RP pastures. Ortega S. et al. (1992a) reported that while RDM after grazing was the most important factor influencing RP herbage accumulation and botanical composition of acc umulated herbage, fre quency of grazing also had a pronounced effect. At low RDM, increasing the interval between grazing events increased herbage accumulation, but interval between grazing events had less effect as RDM increased. Herbage accumulation and peanut percentage were greatest when RDM was between 1500 and 2400 kg ha 1 greater in the stand, a postgraze RDM of 1500 kg ha 1 (a postgraze stubble height of 15 kg ha 1 ( 20 cm stubble height) if the grazing c ycle is less than 35 d (Ortega S et al., 1992a). Growth habit has been shown to play a role in the response of other forages to grazing and is an important consideration for RP use in pasture systems. Mathews et al. ( 1994) reported the effect of growth habit in a comparison of continuous and rotationally Cynodon dactylon (L.) Pers.] pastures that contained 10% of the decumbent growing common bermudagrass at the beginning of the trial. Du ring the 2 yr study, the proportion of common bermudagrass increased each year
30 under continuous stocking, but little change occurred in botanical composition under rotational stocking. Mathews et al. (1994) suggested that rotational stocking allowed the mo re erect growing Callie to shade common bermudagrass during the rest period. In contrast, continuous stocking may have allowed greater light penetration to the common bermudagrass, subsequently increasing its competitiveness with the hybrid bermudgrass. In 1995, Hernndez et al. conducted a trial with mixed palisadegrass [ Brachiaria brizantha (A. Rich.) Stapf] and pinto peanut under two continuous stocking rates (600 and 1200 kg liveweight ha 1 ). Across the 3 yr grazing trial, pinto peanut contributed 34 an d 6% of the dry matter production at the high and low stocking rates, respectively. The tolerance of pinto peanut to a high stocking rate was attributed to its prostrate, stoloniferous growth habit compared with the more upright growing palisadegrass (Soll enberger et al., 2012). This suggests that growth habit of the plant may play a significant role in response to a variety of prescribed grazing practices, but the only grazing study to address these issues with RP was that comparing Florigraze and Arbrook response to continuous stocking (Hernndez Garay et al., 2004). Several clipping studies have reported changes in RP growth habit under various clipping intensities and frequencies (Prine et al., 1981; Quesenberry et al., 2010), but no known studies have evaluated RP plant responses under grazing. As new entries continue to be released, recommendations are needed to provide produce rs with guidelines for use of RP in pasture systems. Plants within a given population have the ability to alter their morpholog y as a response to stress factors, also known as phenotypic plasticity (Nelson, 2000). Phenotypic plasticity is reversible and includes changes in size, structure, and spatial positioning of organs (Huber et al., 1999) in order
31 to adapt to changes in envir onment. Rhizoma peanut has been observed to exhibit this response under various frequencies and intensities of defoliation. When Florigraze RP was mowed every 2 wk, the plant grew in a rosette formation with leaves oriented flat on the ground (Prine et al. 1981). This is a survival mechanism and differs from the typical intermediate to upright growth habit of the cultivar. Under this level of management intensity, overall productivity of Florigraze is decreased. However, if Florigraze is managed to the rec ommended stubble height, it will continue to maintain an intermediate growth habit that responds well to less intensive management (Prine et al., 1981). Similar morphological and production responses were observed by Saldivar et al. (1990) when Florigraze RP was evaluated under clipping frequencies of 2, 6, and 8 wk. Plants in the 2 wk treatment had more prostrate growth habit with small leaves in mid August, and little biomass was removed with each harvest thereafter (Saldivar et al., 1990). In the 6 and 8 wk treatments, plants had elongated to a great extent above the defined clipping level, were more upright in growth habit, and most of the leaf area was removed with clipping. Mislevy et al. (2007) suggested that differences in nutritive value for RP en tries under two clipping management regimes were associated with leaf to stem ratio. Ecoturf and experimental line PI 262833, both low growing types, had higher CP concentration compared with the other entries, which was associated with a higher leaf to st em ratio. Observations during the trial indicated that other entries such as Arbrook were more erect in growth habit, contained fewer leaves and were more stemmy and these morphological traits were associated with a decrease in nutritive value (Mislevy et al., 2007). At the target stubble heights defined in this study (2.5 and
32 10 cm), defoliation regime may have favored an increased leaf to stem ratio in low growing types compared with more upright cultivars (Mislevy et al., 2007). Curl and Jones (1989) sta ted that frequent and intense defoliation of a grass legume mixture may favor temperate, prostrate legumes, and tropical legumes that are generally susceptible to shading by the grass component. Arblick and Ecoturf were evaluated for ornamental use under a 4 cm cutting height with mowing every 4 wk and a 8 cm heigh t mowed every 2 wk Poore r color and appearance ratings were observed for plots mowed every 2 wk to a 8 cm stubble height compared with the 4 cm stubble, 4 wk frequency. Plants under the 4 cm mowi ng height assumed a low growing canopy of leaves that remained intact between mowing events and reduced the appearance of stubble compared with the 8 wk treatment, illustrating a change in growth habit dependent upon management (French et al, 2001). Mixed Legume/Grass Production Systems Advantages and Challenges to Incorporating Legumes Legumes offer several potential contributions to pasture systems. The primary advantage is that they have the ability to fix atmospheric N (N 2 ), which increases the total N contribution and sustainability of the soil plant system. Transfer of fixed N from legumes to companion species does not occur directly, but through secondary processes. While excretion or leakage of N from living roots and nodules is minimal (Giller et al ., 1991), root cell and nodule sloughing, death and decay may contribute N to the grass component of mixed swards (Trannin et al., 2000). Additionally, transfer of N can occur from the legume to the grass via mycorrhizal hyphae (Haystead et al., 1988). Abo veground stubble on the soil surface may be decomposed and nutrients leached over time which may be available for use by companion species (Trannin et al., 2000).
33 Nitrogen can also be cycled to plants through urine and dung of grazing animals (Dubeux et al ., 2007). Russelle (1996) estimated that livestock only use 5 to 30% of ingested N in the diet for meat or milk production; therefore, contribution of urinary and fecal N can be a significant pool of nutrients for grazed swards. Total amount of N fixation can also be influenced by many factors including species, cultivar, soil nutrition, Rhizobium strain, season, environment and climate (Rouque tte and Smith, 2010). Legumes contribute to animal performance through increased nutritive value compared with mos t tropical C 4 grasses. Stobbs et al. (1975) stated that intake of legumes is usually greater than that of tropical grasses, and liveweight gain is increased. In a meta analysis of studies involving grass legume mixtures, Muir et al. (2011) reported that th e greatest contribution of legumes from a ruminant nutrition standpoint is crude protein. With a few exceptions, CP concentration of legumes typically does not fall below 70 g kg 1 at which point intake can become limited (Poppi and Mclel lan, 1995; Muir e t al., 2011). Additionally, the digestibility of legumes tends to be greater than for warm season grasses throughout the growing season, which when combined with increased CP may increase intake and animal performance (Muir et al., 2011). Foster et al. (20 09) evaluated the effect of supplementing bahiagrass ( Paspalum notatum hay with soybean ( Glycine max L.) meal or warm season legume hay on intake, digestibility, and N utilization by lambs. Lambs fed annual legume and RP hays were observed to have increased DM and N intake, digestibility, and improved microbial N synthesis, illustrating the potential of legumes to contribute to relatively low N diets (Foster et al., 2009). Legumes have been found to improve N retention by
34 ruminants when grass diets do not meet energy and N requirements (Foster et al., 2009). Although legumes are well noted for their forage potential, the contribution of legumes in tropical pastures has been limited. Thomas et al. (1995) states that while legumes may have the capacit y to help balance the N cycle in grazed pastures, several factors have limited the ir use in grass based systems. Perhaps the most cited factor is a lack of persistence in mixed species systems (Trannin et al., 2000 ; Shelton et a l., 2005). Trannin et al. (2 000) stated that legume persistence in mixed tropical grass legume pastures is often poor because of the strong competitiveness of the grass associated with its extensive root system, high N and P utilization, and relative tolerance under grazing. Difficul ty in establishing and maintaining legumes in grass based systems is also a driver in the lack of persistence (Shelton et al., 2005). Sollenberger and Kalmbacher (2005) observed a lack of adoption by producers of aeschynomene ( Aeschynomene a mericana L.) a nd desmodium ( Desmodium heterocarpon L. DC.) due to difficulty of establishment and in ability to be maintained in bahiagrass pastures. The majority of tropical forage legumes are annuals and must be reseeded every year. Adaptation of legumes to new environ ments has also been limited by their lack of disease and insect resistance, particularly in sub tropical and tropical environments (Shelton et al., 2005). Additionally, the nature of the grass legume relationship may be cyclical in nature, which can make management a challenge. As the percentage of legume increases in the sward, N contribution from the legume to the grass increases. Competition from the grass component increases, which negatively affects the legume.
35 Decreased N supply from the legume then negatively impacts growth potential of the grass, and the legume may begin to proliferate again (Trannin et al., 2000). Identifying management strategies which minimize competition from the grass component, but do not reduce contribution of the legume can be a challenge. Management strategies must be developed and producers must be educated for the benefits of legume grass pastures to be realized (Shelton et al., 2005). Examples of Successful Warm Season Grass Legume Pastures Shelton et al. (2005) estimat ed that the rate of adoption for the use of legumes in tropical systems has been greater in Asia, Australia, and Brazil than in Africa, the USA, or Latin America. Species that were able to provide multiple benefits had the greatest rate of success and acce ptability by producers. Gross economic progress was greatest where large scale adoption had occurred (Shelton et al., 2005). In Australia, early use of Styloanthes guianensis and Styloanthes humilis was widespread beginning in the mid 1900s, although disea S. humilis ) limited its use (Maass and Hawkins, 2004). Noble et al. (2000) stated that rapid adoption of pasture technology in Australia has resulted in oversowing of stylo on thousands of hectares of native pastures annually. Styloanthes scabra cv. Seca and tetraploid S. hamata cv. Verano account for 1 million ha (Noble et al., 2000). In Africa, 19,000 ha of stylo were cultivated in fodder banks by about 27,000 small land holders by the mid 1990s (Elbasha et al., 1999; Maass and Hawkins, 2004). Although success of this species has been prevalent in Australia and West Africa, Kalmbacher et al. (2002) stated that adoption in the USA has been limited. In Brazil, Valentim and Andrade (2005) stated that death of la rge areas of palisadegrass has led far mers in the Acre state to look for alternatives to maintain
36 productivity and profitability in their production systems. Tropical kudzu ( Pueraria phaseoloides Roxb.) was a n important forage legume used to restore pastur es in that region during the 1990s, with an estimated 480 000 ha present in 2005 (Valentim and Andrade, 2005). However, lack of compatibility of kuzdu with other introduced grasses such as African stargrass ( Cynodon nlemfuensis Vanderyst) promoted further i nvestigation into use of other species for mixed pasture systems. In 2000, Arachis pintoi cv. Belmonte was established vegetatively with stolons in African stargrass. Successful establishment and production from this system was realized by local producers, and planted acrea ge has increased annually to 65000 ha (Valentim and Andrade, 2005). Lascano et al. (2005) observed successful use of Arachis pintoi in a variety of brachiaria based pastures for dairy systems in Colombia, illustrating the potential succes s of forage peanut for tropical regions. While the use of cool season annual legumes has been prevalent throughout the USA (Ball et al., 2007; Rouquette and Smith, 2010, Muir et al., 2011), the number of success stories for warm season grass legume pastur es has been limited. Although the track record of legumes in warm climates is not stellar, RP use in the southeastern USA ence and ability to spread when growing in association with grasses, it is unique among legumes adapted to the region. Shelton et al. (2005) stated that one success has been the use of RP for high quality hay for horse and dairy markets. The introduction o f new genotypes has increased the planted acreage of RP throughout the Gulf Coast region of the USA and provided a more stable,
37 profitable enterprise when compared with other land uses (Williams et al., 2004 ; Shelton et al., 2005 ). Also, while grasses are rarely planted with RP, mixtures may occur because of a lack of control of grasses during RP establishment (Valencia et al., 1991). Valentim et al. (1986) observed that RP is competitive with perennial grasses such as bahiagrass, bermudagrass, and digitgr ass ( Digitaria umfolozi Hall), even at high levels of N application (Valentim et al., 1986). In general, Florigraze RP contributed over 50% of the total CP yield in these mixtures, illustrating the contribution of the legume to pasture system (Valentim e t al., 1986). Prine et al. (1981) reported that during 6 yr of close grazing, RP percentage in a mixed Florigraze bahiagrass pasture was relatively constant and RP was observed to have competitive ability with the grass component. Dunavin (1992) evaluated Hemarthria altissima (Poir.) Stapf and C. E. Hubb], yr trial at Jay, FL. During the first 4 yr, Florigraze competed well with e ach perennial grass, although the percentage of RP declined in all mixtures during the last 4 yr of the trial. Results at this northern location indicate that RP may compete well with perennial sods for several years, but may eventually be crowded out by t hick sodded grasses over time (Dunavin, 1992). In cases where tropical legumes have been the most successful, the target goal of the legume was well defined and producers were made aware of the potential benefits of the legume to their production system th rough education. Miles (2001) suggested that adoption of new practices is often low due to a lack of established relationships between farmers and public institutions. Decreased awareness about
38 legumes and their benefit appears to be a key factor in their rate of adoption (Shelton et al., 2005). In order for the introduction of new systems to be accepted by producers, understanding the socioeconomic needs, skills, and willingness of producers is an important consideration for the education efforts of land g rant institutions (Angle, 2011). Rationale for the Use of Rhizoma Peanut Grass Pastures in the Southeast USA Pasture systems in the southeastern USA are based largely on warm season perennial grasses, but forage nutritive value of these species rapidly de clines from mid summer to late fall (Sollenberger et al., 1989). Incorporation of a warm season legume into these grass pastures may increase forage nu tritive value as well as provide a source of N to otherwise low input systems. Lack of maintenance ferti lization, especially N, and inadequate grazing management are primary factors resulting in the degradation of pastureland in low input systems in warm climate environments (Boddey et al., 1997). Degraded pastureland has limited potential to serve as a sour ce of forage for livestock or to provide ecosystem services. Association of N fixing legumes with grasses offers an economic opportunity for improving pasture quality, productivity, and animal production in warm climate regions characterized by low soil f ertility (Lascano et al., 1989). The presence of even relatively small amounts of legume has increased forage nutritive value and productivity of the system, increased N cycling through cattle excreta, and reduced or eliminated the need for N fertilization (Boddey et al., 2004). Moreover, a combination of fluctuating feeder calf prices and costs of feed, fertilizer, and energy are challenging beef producers to place increased emphasis on grazed forages with significantly reduced inputs and improved managem ent (Prevatt, 2008). Because the cost of N fertilizer can be prohibitive in the maintenance of
39 perennial grass sods, the incorporation of legumes into these existing systems may reduce the need for N fertilization, as well as increase the nutritive value o f the pasture. This technology is needed because current production systems are based on N fertilized grasses and are increasingly vulnerable to high fertilizer cost (Rouquette and Smith, 2010). Currently, the majority of RP planted in the Southeast USA i s used for commercial hay production, although grazing trials have demonstrated excellent levels of animal performance (Sollenberger et al., 1989 ; Hernndez Garay et al., 2004). Because RP is relatively expensive to establish (Williams et al., 2004), it i s often not profitable for use solely as pasture by beef cow calf operations. In addition, the high nutritive value of pure stands of RP may exceed the nutrient requirements of most beef cattle (NRC, 1996). A more cost effective option for use of pure RP s tands in beef production systems may be to limit grazing to classes of animals that would benefit the most from high quality forage, such as calves and replacement heifers (Williams et al., 2004). Considering the high cost of establishment and the need to take establishing pastures out of the grazing rotation for 1 to 2 yr (Williams et al., 1993), alternative approaches are needed if RP is to become an important component of beef cattle production systems in the Southeast. One option for increased use of R P pastures for classes of livestock other than dairy animals (e.g., beef cow calf operations) may be to incorporate RP into perennial grass sods through strip planting. Strip planting entails preparation of a small area of land through tillage followed by planting with RP at the
40 2010), strip planting of RP with perennial grasses may eventually result in the formation of a sustainable grass legume mixture. Compared with establi shing RP in prepared seedbeds, this approach may also allow grazing to continue on these pastures during a significant portion of the establishment phase. Ecosystem Services and C Sequestration in Grasslands Grasslands have the capacity to provide a wide array of goods and services to society that are of economic, environmental, and social importance (Follet and Reed, 2010). It is generally accepted that the primary role of grasslands is t o provide feed for beef, dairy, and sheep industries in th e USA (Fol let and Reed, 2010). Ecosystem services are considered secondary services provided by grasslands and include the capacity of grasslands to act as a carbon sink (Conant et al., 2001), provide bioenergy (Casler et al., 2009), decrease erosion (Karlen et al., 2007), promote retention of nutrients and moisture in the soil (Woodard et al., 2002), act as vegetative buffer strips (Blanco Canqui et al., 2004), and provide aesthetic value for society. Although the list of ancillary benefits is quite extensive, the f ocus of this review will primarily be on the potential role of grasslands as a global C sink. Grassland Management and Factors Affecting Soil Quality Overview Increasing interest in reducing the impact of increasing atmospheric CO 2 has stimulated research evaluating the C sequestration potential of grasslands (Follet t 2001). Agricultural C sequestration is defined as the process through which agricultural practices remove CO 2 from the atmosphere, enhancing C storage in trees and soils, and preserving exis ting tree and soil organic carbon. Agricultural management practices can affect the quantity, quality, and placement of C in the soil via crop selection, crop
41 rotation, fertilization, organic amendments, and tillage type and frequency (Paustian et al., 199 7; Magdoff and Weil, 2004). In cropping systems, soil cultivation has disrupted the balance of the soil organic pool, causing organic matter to be exposed to oxidative processes (Rees et al., 2005). Ingram and Fernandes (2001) estimated that the oxidation of SOM in cultivated soils has contributed approximately 50 Pg C to the atmosphere. Depletion of the soil organic carbon (SOC) pool leads to degradation in soil quality and declining biomass productivity (Folle t t, 2001) which has implications for the ecosy stems involved. Other factors affecting the amount and rate of change in SOC include historical land use (Conant et al., 2001), soil texture (Hassink, 1997), plant species (Paustian et al., 1997), total soil nitrogen (TSN), and environmental factors such as temperature and rainfall (Conant et al., 2001). The role of forage management in C sequestration will be discussed in a later section. The importance of protection of SOM by silt and clay particles is well established (Hassink, 1997; Six et al, 2002). Hassink (1997) showed a relationship between SOM fractions and silt and clay associated C and soil texture. These findings formed the basis for the conclusion that the ability of soil to protect C is based on its association with these particles (Six et a l., 2002). In Florida, many of the soils are characterized by low clay plus silt concentrations, reducing the ir capacity to protect the SOM (Hassink, 1997), and likely limiting the amount o f C that can be sequestered. However, through changes in land manag ement, the potential exists to increase C in Florida soils. Species effects may also play a role in the C sequestration potential of grasslands. Warm season perennial grasses (C 4 photosynthetic pathway) have been
42 proposed for use in the improvement of SOC in grasslands because they produce extensive root systems and provide permanent vegetative cover (Conant et al., 2001).The role of cool season grasses (C 3 photosynthetic pathway), such as small grains, for improvement and maintenance of SOC as cover crops has long been established in crop rotation systems. In an integrated crop livestock system with summer cropping of sorghum ( Sorghum bicolor L ) and winter grazing of rye ( Secale cereal L.), total particulate organic matter remained relatively constant (2. 3 Mg ha 1 yr 1 ) across a 3 yr experiment in Georgia (Franzluebbers and Stuedemann, 2008). Lal et al. (1999) reviewed literature on crop rotations and concluded that potential exists to utilize cover crops on about 51 million ha in the USA, which could sequ ester an estimated 100 to 300 kg C ha 1 yr 1 Results illustrate the potential of various grass species to contribute to the maintenance of the soil C pool. Legumes also offer potential to promote the sustainability of agricultural systems not only throug h biological N fixation, but through the improvement of soil C as well. Continuous cropping systems with legumes can provide constant land cover and stimulate the retention of SOM (Boddey et al., 1997). Also, pasture degradation is commonly related to decr easing N availability caused by an accumulation of low quality plant litter and net immobilization of N due to the activity and number of soil microbes (Vendramini et al., 2007). Thus, one reason for the recommendation of establishing legumes in grass past ures is based on the assumption that legumes increase soil fertility through deposition of high quality litter (Dubeux et al., 2007). Grazing Management and Soil Quality Grazing management, which includes grazing intensity, frequency, and method, plays an important role in C and N sequestration potential of pasture systems.
43 Franzluebbers and Stuedemann (2002) stated that pasture management systems in the southeastern USA stimulate greater stratification of SOC than conservation tillage practices, illustrat ing the potential inputs provided by these systems. Literature describing the influence of stocking rate, a measur e of grazing intensity, indicates that is is the most prominent management factor in pasture studies in the southeastern USA where soil C and N were evaluated. However, evaluation has been less extensive in this region compared to more temperate environments of the USA (Ganjegunte et al., 2005). Franzluebbers et al. (2009) conducted a 12 yr study to evaluate management effects on the rate of ch ange in SOC and TSN throughout various depths in the soil profile in a Coastal bermudagrass (BG) pasture system. Management regimes included unharvested BG, hayed monthly, and grazing at low (5.8 steers ha 1 ) and high (8.7 steers ha 1 ) stocking rate. Grazi ng of pasture led to significantly greater levels of SOC in the surface 15 cm of soil than in ungrazed pastures. Additionally, the difference between low and high stocking rate became significant at the end of the 12 yr evaluation (21.6 vs. 19.9 g kg 1 SOC ). Total and particulate organic N were greater under the high stocking rate than under the low stocking rate at the end of 4 yr in the 0 to 6 cm depth (Franzluebbers and Stuedemann, 2002), but were not different throughout the soil profile between stocki ng rates at the end of 12 yr throughout the soil profile (Franzl uebbers and Stuedemann, 2009). Schuman et al. (1999) stated that generally less than 10% of grassland organic C is located in aboveground biomass, while the remainder is in root biomass or so il organic matter. Greater SOC in the grazed pastures was attributed to root biomass turnover, return of aboveground residues, and cycling of manures (Franzluebbers et al.,
44 2009). Also, grazing intensity and frequency affect aboveground production of forag es and promote nutrient cycling (Schuman et al., 2002), and they affect plant allocation of C by altering tillering and rhizome production and stimulating root exudation (Wright et al., 2004). Wright et al. (2004) stated that these above and below ground plant factors interact with grazing intensity and influence soil C and N mineralization rates. Over the 12 yr study, they found that the annual rate of change in SOC to a depth of 90 cm was greatest (1.17 Mg C ha 1 yr 1 ) for the low stocking rate treatment These results suggest that a moderate stocking rate will optimize soil organic C and N fractions compared with unharvested or hay management. Wright et al. (2004) evaluated grazing management impacts on soil C and N in BG pastures in Overton, TX. Managem ent included low (2.5 cow calf pairs ha 1 for cv. Coastal, 2 cow calf pairs ha 1 for common) and high grazing intensity (7.4 cow calf pairs ha 1 for cv. Coastal, 5 cow calf pairs ha 1 for common), fertilization, and winter overseeding with annual ryegrass ( Lolium multiflorum L.; RG) and clover ( Trifolium sp.; C). Grazing intensity played an important role in soil C and N sequestration, with the high grazing intensity resulting in a smaller increase in C and N over time compared with the low grazing intensit y. Soil organic C increased 67 and 39% from 7 to 26 yr at low grazing intensity for BG + RG and BG + C overseeded pastures, respectively. Differences in the two intensities was attributed to enhanced turnover of plant material, excreta, and physical disrup tion of the soil at high grazing pressure (Wright et al., 2004). A study by Liu et al. (2011b) evaluated the effect of a range of grazing intensities on the distribution of soil and plant nutrient pools in Tifton 85 bermudagrass. Pastures were stocked rot ationally every 28 d and grazed to a post grazing stubble height of 8,
45 16, or 24 cm Although the nutrient concentration of plant pools was not as affected by grazing intensity, there was a 17% increase in soil C content with increasing postgraze stubble h eight. Soil N concentration increased from 0.56 to 0.75 g kg 1 as stubble height increased from 8 to 24 cm, respectively. The authors suggested that changes in soil C and N were a function of the change in above and belowground biomass allocation and rete ntion of C in the plant soil system. These results illustrate that selecting an appropriate level of grazing intensity is an important management strategy to maximize plant and soil quality and can increase soil C content in Florida grassland ecosystems. The effect of grazing frequency and method is less well documented for systems in the southeastern USA. Because intensity and frequency often interact, frequency effects are usually reported in the literature along with a given level of management intensit y. Impacts of the intensity frequency interaction may include changes in soil bulk density, soil penetration, and water filtration, and nutrient cycling in grazed systems. Grazing method may affect soil OM through the influence on above and below groun d C pools. It has been suggested that plant related advantages attributed to rotational vs. continuous stocking include increased pasture carrying capacity, higher gain ha 1 improved persistence (Matches and Burns, 1995), and more uniform use of an extens ive pasture area (Hart et al., 1993). These pasture responses will influence nutrient cycling in grasslands through the amount and spatial distribution of plant litter and excreta, which ultimately impacts the soil OM pool (Dubeux et al., 2007). Hay Manag ement and Soil Quality There are an estimated 55 million acres of land harvested for hay production in the USA annually, representing $6.7 billion in cash receip ts (USDA NASS, 2013). Thus,
46 quantifying the role of ecosystem services of forage systems under various harvest management d is of interest. Franzluebbers and Stuedemann (2002) compared two grazed 15 year old and one 19 year old stands of Tifton 44 bermudagrass to two hayed 15 year year dagrass. Grazed pastures were continuously stocked for 5 mo yr 1 and hayed pastures were clipped three to four times annually. Total organic C was greater under grazed than hayed bermudagrasses (13 vs. 10.6 g kg 1 ) within the surface 20 cm. Differences am ong treatments was attributed to the return of feces and urine in the grazed treaments compared with those that were hayed. Additionally, surface litter C:N ratio was lower for the grazed pastures than hayed, illustrating the importance of quantity and qua lity of litter returned to these systems in C accumulation potential (Franzluebbers and Stuedemann, 2002). Franzluebbers and Stuedemann (2009) stated that haying removed forage from the field which was ultimately fed to cattle elsewhere, resulting in a re moval of C from the system. Although the hay would have the potential to sequester C through excreta of the animals consuming the feed elsewhere, this is a relatively inefficient system when compared with animals grazing directly on pasture (Franzluebbers and Stuedemann, 2009). Forage Management and Effects on SOM Dynamics Grazing intensity, frequency, and stocking method may influence the soil organic matter pool through the return of nutrients to the soil via plant litter or excreta (Dubeux et al., 2007) Nutrient retention in cattle body tissue and nutrient export through animal products represents less than 30% of the total nutrients ingested by cattle (Haynes and Williams, 1993). Thus, a large portion of the nutrients consumed by cattle is returned to
47 the soil via excreta (Vendramini et al., 2007). Nutrient return by excreta is non uniform in distribution, and tends to be concentrated in areas of shade and water (Mathews et al., 1996). Plant litter is distributed more evenly (Dubeux et al., 2007); howev er, litter from tropical C 4 grasses is often harder to degrade due to a high C:N ratio, which causes immobilization by soil microbes (Vendramini et al., 2007). Stocking rate plays an important role in determining the proportion of litter or excreta returne d to the system (Dubeux et al., 2007). Nutrients from animal excreta are more readily plant available upon deposition compared with low quality warm season C 4 plant residues often found in tropical systems (Dubeux et al., 2007). Although nutrient recovery is realized more quickly from excreta than plant litter, nutrient loss from the system can be significant depending on the management Thus, strategies are needed to increase the uniformity of excreta distribution and minimize nutrient losses from this nut rient pool to increase plant recovery efficiency (Mathews et al ., 2004; Dubeux et al., 2007). Removal of plant material through clipping influences the nutrient cycling dynamic as well. When herbage is removed for hay, less litter is available for degradat ion and return to the system. Although clipping frequency may increase nutritive value of potential litter, the act of removing clippings from the system negates the contribution of aboveground material to the system. Soil C and N pools are a function of the C and N concentration in the soil organic matter (SOM) and the amount of SOM present (Dubeux et al., 2006). The quantity and chemical composition of SOM is important to C and N cycling, as N is often the productivity limiting factor in grassland ecosyt ems (Ganjegunte et al., 2005). Soil OM includes plant, animal and microbial residues in all stages of decomposition (Post and
48 Kwon, 2000). Many of these organic compounds are closely associated with inorganic soil particles, and their turnover time is high ly dependent upon biological, chemical, and physical processes in the soil. Traditionally, SOM has been characterized through chemical fractionation (i.e. fulvic acid, humic acid, etc.); however, the use of these fractions to explain dynamics of agroecosy stems has been limited (Dubeux et al., 2006). Physical fractionation of SOM by size or density has become a well accepted method for characterizing SOM quality (Meijboom et al., 1995). Physically fractionated SOM relates to specific carbon pools, which are important for understanding soil carbon processes that occur in grasslands (Post and Kwon, 2000). Light fraction organic carbon (LFOC) is free, particulate plant and animal residues undergoing decomposition (Christensen, 1996). In systems with significant returns of plant litter, LFOC can accumulate readily, despite higher decomposition rates. Rate of turnover for this pool is on the order of months to a few years. The intermediate fraction is made of partially humified material. Soil OM can also be transf ormed by bacterial action and stabilized in clay or silt sized organomineral complexes, also known as heavy fraction organic carbon (HFOC), where the majority of SOC is found (Post and Kwon, 2000). For C sequestration strategies to be effective in the lon g term, it is likely that they must increase the slow and passive pools of SOM, such as that of the HFOC (Franzluebbers and Studemann, 2002). Evaluation of changes in these fractions may be useful in describing the contribution and dynamics of forage mana gement to SOM, but few studies have evaluated changes in these fractions in pasture management systems. Dubeux et al. (2006) evaluated changes in SOM of a bahiagrass pasture under various N fertilization
49 levels, stocking rates, and stocking methods. Over t he 3 yr study, bulk soil C and N was not affected by management intensity; however, effects were observed in the light density fraction of SOM that readily responds to changes in management (Dubeux et al., 2006). As management intensity increased (fertiliz ation level and SR), both C and N concentration were increased in the light density fraction, illustrating sh ort term changes in this pool. Carbon and N concentrations were greater for rotationally stocked treatments than the low stocking rate, continuousl y stocked treatments. Thus, increasing management intensity can contribute positively to soil fertility and C sequestration (Dubeux et al., 2006). Silveira et al. (2013) investigated the short term impacts of differing levels of grazing intensity (postgraz e stubble height of 8, 16, and 24 cm) and N fertilization (50, 150, and 250 kg N ha 1 yr 1 ) on soil C dynamics of rotationally yr study, particulate organic C and total C and N increased linearly with increasing stubble height in the < 53 m fraction. The authors suggested that C associated with this size fraction represents relatively short term changes in soil C for sandy soils, although similar evaluations in other soil types typically show the infl uence of management occurs more readily in macro and micro aggregates ( Therefore, understanding how grassland C is allocated to different soil size classes further help determine short and long term impacts of past ure management on soil C. Summary Rhizoma peanut is a perennial, forage legume well adapted to the lower Coastal Plain region of the southeastern USA and with potential for incorporation into beef cattle production systems. Mixed species swards with RP ma y increase pasture productivity over non fertilized grasses and provide a source of high nutritive value forage in
50 otherwise low input production systems. Time to establishment and establishment cost have been cited as the primary factors limiting the adop tion of RP in the Southeast, and alternative strategies are needed if RP is to make a significant contribution to the beef cattle industry in Florida. Strip planting of RP may be a viable method of establishment; however, further research is needed to dete rmine the effect of different RP growth habits on rate of establishment, response to RP defoliation management practices, and the effect of a range of grazing frequencies and intensities on persistence, productivity, ad nutritive value of established RP pa stures. Additionally, there is a need to define potential of RP and grass based systems to contribute to changes in soil quality and to assess the role of defoliation management on this response.
51 CHAPTER 3 GROWTH HABIT OF RHIZOMA PEANUT CULTIVARS AFFECTS ESTABLISHMENT AND SPREAD WHEN STRIP PLANTED IN BAHIAGRASS SOD Overview of Research Warm season perennial grasses such as bahiagrass ( Paspalum notatum Flgge) and bermudagrass [ Cynodon dactylon (L.) Pers.] form the basis of many grazing systems in the USA Gulf Coast Region (Ball et al., 2007). These grasses require N inputs for production and persistence, but rising cost of fertilizer (USDA NASS, 2013) makes N less affordable for many producers and prohibitive ly expensive for others. Failure to maintain ade quate N nutrition reduces forage accumulation, resulting in overgrazing and subsequent degradation of grasslands (Boddey et al., 2004), threatening the sustainability of grass based pasture livestock systems in the region. Association of N fixing legumes w ith grasses offers an opportunity for improving pasture quality, productivity, and animal production in regions characterized by low soil fer tility (Lascano et al., 1989). The incorporation of legumes into these existing systems may provide required N to l ivestock through consumption of high crude protein forage (Sollenberger et al., 1989), needed N to associated grasses through nutrient cycling from livestock waste and legume nodule sloughing (Dubeux et al., 2007), and critical economic relief to producers The contribution of legumes to pasture systems has been less pronounced in warm compared with temperate climates due to a general lack of persistence of tropical and subtropical legumes under grazing, few disease resistant cultivars, and limited adoptio n of technologies leading to successful use (Ortega et al., 1992 a; Shelton et al., 2005). However, in the southeastern USA rhizoma peanut (RP; Arachis glabrata Benth.) for the use of legumes by producers
52 (Shelton et al., 2005) Rhizoma peanut is a warm season perennial legume that is well adapted to the southern Gulf Coast region and has potential for incorporation into pasture based livestock systems. It is often compared with alfalfa ( Medicago sativa L.) because of its yield potential (Andrews et al., 1985), high nutritive value (Prine et al., 1981; Mislevy et al., 2007), and persistence under a variety of management conditions (Ortega S. et al., 1992a; Butler et al., 2007) spread when growing in association with grasses (Dunavin, 1992), it is unique among legumes adapted to the region. In spit e of its many desirable attributes, high cost of establishment of pure stands of RP (~ $1250 ha 1 ; Blount, personal communication, 2012) has made it uneconomical for use in livestock enterprises characterized by relatively low economic return per hectare s uch as beef cow ( Bos sp.) calf and other pasture based systems. Consequently, use of RP has primarily been limited to a high value hay crop for horses ( Equus caballus ) or dairy cows. Lower cost, alternative establishment strategies are needed if RP is to m ake significant contributions to grazing systems for livestock. One approach for lower cost incorporation of RP into grass pastures is strip planting (Castillo et al., 2013). Because RP is a long lived perennial with ability to move laterally via an extens ive rhizome system, it has potential to spread into the surrounding grass areas over time and form a mixed pasture. planting approach (Castillo et al., 2013). There are other recently releas ed genotypes of RP that may have potential for incorporation into existing grass pastures, but they have not
53 cultivars with an intermediate to upright growth habit that ma y exhibit high dry matter yields, persistence, and disease tolerance (Quesenberry et al., 2010). Prine et al. growing ecotypes for grazing or ornamental use. The range in growth habits repr esented by these genotypes may well affect their ability to spread and persist in grass pastures using a strip planting establishment approach. Furthermore, Florida grasslands have the potential to contribute to soil C sequestration (Silveira et al., 2013 ). Conversion of agricultural land from cultivation or native vegetation to improved grasslands has been shown to increase soil C sequestration (Conant et al., 2001). However, there is a need to understand how conversion among improved forage systems, such as shifting from C 4 to C 3 based systems, and forage management practices influence soil C and N in this environment. Therefore, the objectives of this study were to 1) evaluate differences among RP genotypes in their establishment ability and spread poten tial when strip planted in bahiagrass sods; 2) to determine the effect of defoliation management during the establishment year on these responses; and 3) to quantify the effects of converting from a C 4 to C 3 based pasture system on soil C and N dynamics. M aterials and Methods Experimental Site An establishment study was conducted during 2011 and 2012, with a new area planted each year, at the University of Florida Beef Research Unit in Gainesville, FL (29.72N, 82.35W). The site was chosen because of the p resence of well established moderately well drained soil. Soils at the site include Pomona fine sand (sandy, siliceous, hyperthermic Ultic Alaquods) and Plummer fine
54 sand (loamy, siliceous, subactive, thermic Grossarenic Palequu lts). Soil samples were taken to a depth of 15 cm and analyzed by the University of Florida Extension Soil Testing Laboratory. Initial characterization of the surface soil (top 15 cm) indicated a soil pH of 6.5 and Mehlich 1 extractable P, K, Mg and Ca of 11, 37, 107, and 659 mg kg 1 respectively. Treatments and Experimental Design Four RP genotypes and two defoliation management regimes were replicated three times in a split plot arrangement, with main plots allocated in a randomized, complete block des ign. Defoliation management regime served as the main plot, and RP entry as the sub plot for a total of 24 experimental units. Each experimental unit was 6 m wide 5 m long and consisted of a 4 m wide strip through the entire length of the plot into which RP was planted, bordered by a 1 m strip of bahiagrass on each side. Entries included Arblick, Ecoturf, Florigraze, and UF Peace. The entries were selected to represent a range in plant growth habit among existing RP cultivars and germplasms. Germplasms A rblick and Ecoturf are decumbent (Prine et al., 2010), whereas cultivars Florigraze and UF Peace have a more intermediate to upright growth habit (Quesenberry et al., 2010). The two defoliation treatments were selected based on preliminary results from a s trip planting study with Florigraze RP that has since been published ( Castillo et al., 2013) Defoliation treatments were hay production or rotational stocking. Hay production plots (bahiagrass and planted strip) were mechanically harvested every 28 d to a 10 cm bahiagrass stubble height usi ng a sickle bar mower. The 10 cm stubble was chosen to approximate the height used in bahiagrass hay production systems. The rotational stocking treatment was grazed every 28 d to a 15 cm bahiagrass stubble height. The 15 cm stubble was chosen because it w as not
55 certain to what degree the grazing animals would select herbage from the planted strip versus the bahiagrass bordering the planted strip. The intent was that a taller bahiagrass stubble for grazing than for hay production would minimize the likeliho od that RP would be overgrazed in the planted strip. The defoliation treatments were imposed both in the year of establishment and the year after establishment. For grazed treatments, mob stocking (Allen et al., 2011) was used to attain the target stubble height using 12, 350 kg yea rling cross bred beef heifers. Animals were assigned to an experimental unit for a short period of time (~1 hour), and grazing was monitored until a 15 cm bahiagra ss stubble height was reached. Animals were given access to both the bahiagrass bounding the planted strip and the RP within the strip. Defoliation management treatments were imposed starting 10 wk after first sprout emergence (13 wk after planting) on 15 June 2011 and 14 June 2012, respectively. Plot Establishment and Management A new set of plots was established in both 2011 and 2012. To prepare for planting, the 4 m wide strip in each plot was sprayed with glyphosate in October 2010 and 2011 at a rate of 3.4 kg a.i. ha 1 The application was done using a CO 2 pressurized backpack sprayer calibrated to deliver 187 L ha 1 at 310 kPa using a 3.04 m wide boom. In spring each year, the sprayed strips were prepared with a moldboard plow and heavily disked to ens ure a clean tilled planting area. Prior to planting, 10 cm deep furrows were made at 50 cm intervals in each plot (eight furrows per plot). Rhizomes of RP entries were harvested from existing planting stock nursery areas and were planted by hand, covered, and packed in the furrows at a rate of 1000 kg ha 1 on 17 and 18 March 2011 and 15 March 2012. Sprout emergence and successful establishment occurred in all plots in 2011; however, in 2012, all plots planted with Ecoturf failed to
56 establish. The reason for stand failure is unknown; however, the Ecoturf planting material had a total non structural carbohydrate (TNC) concentration of 135 and 57 g kg 1 DM in 2011 and 2012, respectively. In general, it is not recommended to use rhizomes with a pre plant TNC con centration less than 130 g kg 1 and stand failure has been associated with a concentration of less than 62 g kg 1 (Rice et al., 1995) Pre plant TNC concentration was 120 and 73 g kg 1 DM for Florigraze, 122 and 1 39 g kg 1 DM for Arblick, and 138 and 107 g kg 1 DM for UF Peace in 2011 and 2012, respectively. Because stand establishment was successful across a range of TNC values for Florigraze, there may be other contributing factors which led to stand failure of E coturf in 2012. Several methods were used to control weeds in the planted strip. Herbicides used were ammonium salt of imazapic (Impose; +/ 2 [4,5 dihydro 4 methyl 4 (1 methylethyl) 5 oxo 1H imidazol 2 yl] 5 methyl 3 pyridinecarboxylicacid) and 2,4 D (d imethylamine salt of 2,4 D dichlorophenoxyacetic acid) at rates of 0.07 kg a.i. ha 1 and 0.26 kg a.i. ha 1 respectively. Impose was applied on 11 May and 5 June 2011 and 6 July 2012 to control a broad spectrum of weeds when they were approximately 5 to 10 cm tall (Ferrell and Sellers, 2012). The herbicide 2,4 D was applied on 30 Aug. 2011 and 6 July 2012 for broadleaf weed control. At the end of the 2011 growing season, all plots were hand weeded to remove competition from Old World diamondflower [ Hedyot is corymbosa (L.) Lam]. It did not recur in the 2012 plots. Fertilization was based on soil test results, and P and K were applied over the entire experimental area in the forms of triple superphosphate and muriate of potash, respectively, at a rate of 30 kg P and 80 kg K ha 1 on 11 Apr. 2011 and 12 Apr. 2012 ( 25
57 days after planting) Irrigation was used during the establishment phase prior to initiation of the defoliation treatments. Amount of irrigation applied weekly was equal to the 30 yr average rainfa ll for that week less any rainfall that occurred during that week. Total irrigation applied in March, April, May, and June 2011 was 12.5, 60, 80, and 12.5 mm, respectively. In 2012, 65, 52, and 40 mm of irrigation were supplied in March, April, and May, re spectively. The final irrigation event occurred on 13 June 2011 and 21 May 2012 prior to initiation of the defoliation treatments. Monthly rainfall during the 2011 and 2012 calendar years and the 30 yr average for this location are shown (Figure 3 1). Resp onse Variables Shoot Emergence After first shoot emergence was observed for each genotype (3 wk after planting in 2011 and 2012), emergence data were collected weekly. Prior research has shown that sprout emergence of RP is complete by 7 wk after it begins (Williams, 1993, Williams et al., 1997) however, that work was conducted with Florigraze RP and no studies have evaluated shoot emergence patterns of other, newly released RP genotypes. To accommodate potential d ifferences among entries in duration of the sprout emergence period, measurements were taken for 10 wk. Counts were made in four 20 x 50 cm quadrats per plot, and counting locations were the same at each count date. Paired quadrats were placed at each of two fixed distances along a transect that ran parallel with the length of the plot and bisected the planted strip. A. The quadrat was placed such that the 50 cm side was parallel to the orientation of the RP rows, and the 20 cm side was centered on top of a row. Both defoliation treatment plots of each entry within a block were counted and because defoliation treatments had not yet been imposed there were a total of six observations per entry (three blocks times 2 defoliation
58 treatments). Shoot emergence wa s calculated as the average of the four locations in each plot. In 2011, first shoot emergence was observed on 5 April and data were collected weekly until 8 June. Sprout emergence was observed for the 2012 establishment plots beginning on 4 April and cont inued weekly until 6 June. Rhizoma Peanut Ground Cover Percent RP ground cover was estimated visually every 28 d prior to each defoliation event during the establishment year. During the year after establishment, ground cover was measured prior to the firs t and last defoliation event of the year. A 1 m 2 (2 0.5 m) quadrat was placed in the center of the RP strip at two marked locations along the length of the strip so that ground cover was estimated from the same areas at each date. The 0.5 m side of the q uadrat was oriented parallel to planted RP rows, and the quadrat encompassed four rows of RP. The quadrat was divided into 100, 10 by 10 cm squares to facilitate estimations. Percentage ground cover was visually estimated by the same observer in twenty of the 10 by 10 cm squares and averaged across estimates within a quadrat and across the two quadrats per plot to obtain an overall average for each experimental unit. Rhizoma Peanut Frequency Frequency of RP occurrence was evaluated at the same time and in the same twenty 10 by 10 cm squares per quadrat placement as percent cover. Two quadrats were sampled per plot and frequen cy was calculated as [ # of squares per plot containing RP/40] 100. Rhizoma Peanut Spread Rhizoma peanut spread was measured each year prior to the last defoliation event of the season. Spread of RP was estimated as the distance from the center of the
59 planted strip to the farthest location at the edge of the planted strip where distinguishable aboveground RP plant parts were found. A marked transect was positioned at the center of the planted strip parallel to the planted rows. At two locations along the transect, 1.5 and 3.5 m from one end of the plot, a line perpendicular to the original transect was extended on each side. Along the perpendicular line, distance from the middle of the planted strip to the last visible peanut plant material at the edge of the planted strip was measured, resulting in four measurements of spread per plot. Spread per experimental unit was expressed as the average of the four measurements per plot. Bahiagrass Herbage Harvested Herbage harvested was measured every 28 d prior to each grazing or haying event to determine the quantity of bahiagrass herbage that could be utilized for livestock feed during the e stablishment year of RP. Herbage harvested was measured in the bahiagrass portion of each plot by hand clipping two representative 0.25 m 2 quadrats to a stubble height of 15 cm for grazed and 10 cm for hayed plots. The harvested herbage was dried at 60C u ntil constant weight to determine dry matter harvested. Root rhizome to Shoot Ratio and Root rhizome Mass At the end of the 2012 growing season, root rhizome and shoot mass were measured at two random locations in each plot planted in 2011. The goal was to assess above vs. below ground partitioning of the different treatments and relate that to measures of establishment success. Above ground biomass within a 20 x 20 cm quadrat was harvested and dried at 60C until constant weight to determine dry matter h arvested. Roots and rhizomes under that quadrat were removed to a 20 cm depth, washed, and dried at 60C to constant weight. Root rhizome to shoot ratio was
60 calculated as the total dry weight of roots and rhizomes divided by the mass of above ground herbag e harvested. Soil Bulk Density and Total C, N, and C:N Ratio Soil measurements were taken at the time of establishment and after 2 yr of imposing treatments to quantify the effect of converting a well established C 4 grassland to a C 3 based system on soil C and N in the planted strip. Soil samples were collected by strata to a 70 cm sampling depth from each experimental unit at five random locations within the tilled strip Sampling occurred prior to planting and imposition of treatments in March 2011 and ag ain at the end of the 2 yr experiment (only the experimental area planted in 2011 was sampled) in March 2013. Before the establishment of this experiment, the pasture area con sisted of a 20 yr old stand of Pensacola bahiagrass. Two undisturbed soil cores w ere taken per plot fo r bulk density determination using a JMC Soil Sampler (Clements Associates, Inc., Newton, IA). Bulk density was calculated as the total dry weight of the soil divided by the volume of the coring device. Two additional soil cores were e xtracted from each experimental unit from the following strata using a soil auger: 0 to 10, 10 to 20, 20 to 40, and 40 to 70 cm. Sample collection below a 70 cm depth was limited by the presence of an argillic layer in the Spodosol, which created variabilit y in the amount of sample that could be collected past this depth. As the stratum above 70 cm was variable in depth from plot to plot, these data are not reported. Total C and total N were determined by dry combustion using a Flash EA 1112 C/N analyzer on samples ground in a ball mill for 5 min. The C:N ratio was calculated as the percentage of C within a sample divided by the percentage of N.
61 C Isotope Ratio Determination of C 3 and C 4 derived C and SOC Retained 13 C) were measured for a subset of soil samples to determine the origin of soil organic matter (SOM) inputs that occurred during the experiment. A subset from the samples taken in 2011 and 2013 were analyzed in ord er to determine change in C isotope following 2 yr of legume growth in the planted strips The soil samples analyzed were those from Florigraze plots, from the upper two soil layers (0 to 10 and 10 to 20 cm), from the two defoliation regimes (grazed or ha yed), and from all three replications of these treatment s (total 24 samples). Samples were analyzed using a Thermo Finnigan MAT DeltaPlus XL Isotope Ratio Mass Spectrometer (IRMS) interfaced via a Conflo III device to a Costech ECS 4010 elemental analyzer 13 C PDB standard as described by Boutton (1991). The ratio of 13 C/ 12 C is expressed as: 13 13 C/ 12 C ratio of sample)/( 13 C/ 12 C ratio of standard) 1] 1000. A negative value indicates a smaller proportion of 13 C in the sample compared with the standard. The more negative value for C 3 plants is related to discrimination in favor of 12 C by the pri mary carboxylation steps in the photosynthetic pathway catalyzed by ribulose bisphosphate carb oxylase. Because the primary carboxylation step differs for C 4 plants, there is typically greater discrimination for 13 C in C 3 plants (Lefroy et al., 1993). The ratio of carbon isotopes 13 C/ 12 C in a soil sample was compared to a plant standard (RP for C 3 or BG for C 4 respectively) to detect the photosynthetic pathway (C 3 vs. C 4 ) of the source plant. The contribution of soil C from C 3 or C 4 plants was determined according to the following equations by Follett et al. (2009): C 3 derived C(%) = ( 13 C sample 13 C C4 13 C C3 13 C C4 )*100
62 C 4 13 C C3 13 C sample 13 C C3 13 C C4 )*100 Carbon contribution from C 3 or C 4 plants was calculated as the percentage of C 3 or C 4 derived C multiplied by the total soil organic C (g kg 1 soil). The amount of SOC retained was determined to quantify how much C remained in the system following 2 yr of imposing treatments. The following equations were used to calculate the change in C across the experiment and expected residence time of C as described by Clay et al. (2007) and Silveira et al. (2013): SOC retained = [SOC final 13 C soil final 13 C plant 13 C SOCinitial 13 C plant ) 13 values represent 13 C associated with soil measurements from 2011 and 2013, respectively, and current RP plant material. Statistical Analysis Data were analyzed using PROC MIXED of SAS ( SAS Institute, 2010). In the year of establishment, e ntries, defoliation treatment, and their interactions were considered fixed effects, and years, blocks, and their interactions were considered random effects. In 2012, observations for Ecoturf plots that failed to establish were treated as missing data for the purpose of analysis of the two establishment years. Mean separation of entries was done using the PDIFF option of LSMEANS in SAS. Defoliation treatment means were separated using the F test. Date wa s considered a repeated measure for sprout emergence, ground cover, and frequency with an autoregressive covariance structure. For soil responses, data were a nalyzed within a given soil layer. Year, RP entry, and defoliation management regimes were conside red fixed effects and block was considered random. Treatments were considered different for plant responses when P when P for soil responses.
63 Results and Discussion Shoot Emergence Total shoot emergence was greater for Florigraze (79 sprouts m 2 ) than all other genotypes. Ecoturf and Arblick were similar (66 and 57 sprouts m 2 respectively), but had greater total emergence compared with UF Peace (30 sprouts m 2 ). Shoot emergence increased for up to 10 wk after planting, after which it was difficult to distinguish between individual shoots. Previous work with Florigraze suggested that emergence was c omplete at 7 wk after first emergence (Williams et al., 1993). However, differences in shoot emergence patterns have been previously reported in Florigraze and Arbrook (Williams et al., 1997), which suggests that the time to reach peak emergence may differ among genotypes. The results of the present study show that shoot emergence may continue for up to 10 wk, but there was no date x genotype interaction for emergence pattern ( P > 0.05). Rhizoma Peanut Ground Cover There were date x entry ( P = 0.0003) an d date x defoliation ( P = 0.0012) interactions for RP cover. Interaction occurred because Arblick and Florigraze had greater ground cover during June and July (Figure 3 2), while Florigraze and Ecoturf had the greatest ground cover from September through t he remainder of the season. All entries reached their peak percentage cover in August with the exception of Ecoturf which continued to increase through October. UF Peace had the least ground cover compared to the other lines and cover remained below 15% th roughout the entire establishment year. Cover of Florigraze and Ecoturf exceeded 30% by the end of the growing season. Interrante et al. (2011) observed similar cover for a monoculture of Florigraze during the year of establishment. Williams et al. (2008) reported 30 to 40%
64 ground cover for Arblick and Ecoturf during the year of establishment when the area p lanted was harvested every 9 wk. Prine et al. (2010) reported that Arblick generally is slower to establish than other forage type genotypes of RP, an observation which is supported by the results of the current study. There was also a date x defoliation management interaction (Figure 3 3) that occurred because the hayed treatment resulted in greater percentage cover in July and August of the establishment year, but there were only trends favoring the hayed treatment thereafter. The apparent advantage of t he hayed treatment likely occurred because grazed plots were defoliated more severely than clipped plots. This occurred even though the stubble height for the hayed treatment was 10 cm vs. a target stubble height (bahiagrass component) of 15 cm for grazing When cattle entered the rotationally stocked plots, they first went to the planted peanut strip and grazed it close to soil level before initiating grazing on the bahiagrass. Thus, by the time the bahiagrass was grazed to 15 cm the strip planted to RP ha d been grazed considerably lower than that. There was a much larger and highly significant advantage of hayed vs. grazed defoliation management in previous work with Florigraze in the same environment where the current work was done (Castillo et al., 2013) In that experiment, the greatest percentage of ground cover reported for the 28 d rotational stocking treatment was 4% vs. 29% for harvested treatments (Castillo et al., 2013). Lower RP cover in that experiment could be explained in part because measurem ents were taken following defoliation instead of before defoliation as in the current study, and in the present study, bahiagrass strips were sprayed and killed with glyphosate in the fall prior to planting. In the study by Castillo et al. (2013), no glyph osate was applied and bahiagrass likely
65 regenerated more rapidly following tillage of the strip than in the current study, likely resulting in greater competition to establishing RP. Ground cover differences were observed during the year after establishme nt among defoliation regimes. Hay production treatments had a greater percentage ground cover compared with rotational stocking (66 vs. 46%, respectively). This implies that defoliation management during the year after planting continues to affect RP estab lishment success. There was no effect of entry or entry x defoliation interaction Measurement of ground cover on these plots at the beginning of the 2013 growing season further ill ustrated an additional carryover effect of management on cover. In spring 2013, rotationally stocked treatments had 47% ground cover compared with 70% ( P = 0.0037) for clipped. Percentage of ground cover tended ( P = 0.0789) to be different among entries, w ith the low growing ecotypes Arblick and Ecoturf having increased ground cover (68%) compared with Florigraze and UF Peace (46%). Rhizoma Peanut Frequency There were date x entry ( P = 0.0002) and date x defoliation ( P = 0.0356) effects on RP frequency duri ng the establishment year. Entry differences in RP frequency followed a similar pattern to ground cover. Arblick and Florigraze had greater frequency in June compared with the other lines (Figure 3 2), but beginning in July, frequency of Ecoturf began to i ncrease and was similar to Arblick and Florigraze. Florigraze and Ecoturf continued to increase throughout the remainder of the season, and had a greater frequency of occurrence than Arblick and UF Peace. Frequency of UF Peace was lower than all other geno types from June to October, and did not exceed 30%. Increased late season cover and frequency associated with Florigraze and Eco turf
66 illustrate apparent greater potential for establishment success compared with other genotypes. Castillo et al. (2013) repor ted slightly lower values for strip planted Florigraze under the same management strategies (44%). At the beginning of the establishment year, RP frequency did not differ between defoliation management strategies, but it was lower for grazed treatments co mpared with those under hay production in July and August (Figure 3 3). This was likely due to the same reasons that RP cover was sometimes greater for hayed than grazed plots. No differences were observed between grazed and hayed treatments during Septemb er and October. These data illustrate that hay production treatments achieved a greater frequency early in the season, whereas those under rotational stocking took longer to achieve a similar level of occurrence. During the year after establishment, no di fferences ( P among treatments for percentage of RP frequency. Frequency of RP occurrence ranged from 77 to 95% for all entries. The average establishment period for RP to reach a full stand is 2 to 3 yr (Williams et al., 1993). Distri bution of RP reached its peak during this time frame as planted areas begin to fill in, which may explain the lack of differences among entries in the year after planting. Rhizoma Peanut Spread Differences in RP spread were observed among entries ( P = 0.0 086) and defoliation strategies ( P = 0.0023) (Tables 3 1 and 3 2, respectively). Ecoturf, Florigraze, and Arblick had greater total spread (25, 13, and 12 cm, respectively) into bahiagrass compared with UF Peace ( 3 cm). In an evaluation of RP genotypes, U F Peace and Florigraze were noted for superior spread potential and competitiveness with bermudagrass compared with other entries (Friere et al., 2000). However, these plots
67 were allowed to reach full establishment and then managed as pure stands under cli pping and encroachment into surrounding grass borders occurred over a 10 yr period. In the present study, the competition between well established grasses and the establishing legume may have reduced spread potential of UF Peace. The low growth habit of Ec oturf and Arblick may have protected RP from complete leaf removal during a defoliation event, resulting in greater spread potential for those entries. Haynes (1980) suggested that the sustainability of legumes in grazing systems is largely determined by g rowth habit and the ability to protect growing points. Legumes with creeping stems tend to escape serious damage by grazing animals, whereas those with an upright habit are more susceptible. Rotational stocking decreased RP spread (0 m) compared with treat ments under hay production (0.24 m). Spread may have been reduced in the rotationally stocked treatments because of animal selection as described in the previous section. Preference for legumes in mixed swards has been shown to affect their persistence in temperate and tropical forage systems (Lascano and Thomas, 1988; Schwinning and Parsons, 1996), and this may negatively impact establishing RP grass associations. When spread was measured at the end of the year after establishment, there were differences d ue to defoliation management regime ( P = 0.0127). Mean spread for hayed plots (Table 3 1) was 0.89 m compared with 0.55 m for those under rotational stocking. Although spread increased in the second year, these observations illustrate that the management s trategy utilized during the first 2 yr after planting can significantly impact the establishment success of RP. No differences were observed among entries
68 for spread during the second year of management, m into adjacent bahiagrass strips. Herbage Harvested Total herbage harvested did not differ among defoliation regimes and was 4.0 and 4.4 Mg DM ha 1 for grazed and hayed treatments, respectively. One management consider ation for strip planting is to determine whether the area can be utilized during the establishment phase for grazing or hay production without negati vely impacting RP establishment If the area is removed from production during this time, these values repr esent the amount of bahiagrass that would be sacrificed. Average production from bahiagrass in Florida ranges from 3,000 to 11,000 kg ha 1 depending upon N fertilization rate and environmental conditions (Stewart et al., 2005; Newman et al., 2011) Thus, the values reported in this study fall within the range of production for pastures in Florida, and their presence in the lower portion of this range reflects the absence of N fertilizer application. Root rhizome to Shoot Ratio and Root rhizome Mass There were no main effects of entry for root rhizome to shoot ratio ( P = 0.255) or root rhizome mass ( P = 0.499) due to large standard errors (0.29 for root rhizome to shoot ratio and 1010 kg ha 1 for root rhizome mass). Root rhizome to shoot ratios were 0.95, 1.4, 1.5, and 0.84 for Arblick, Ecoturf, Florigraze, and UF Peace, respectively. Root rhizome to shoot ratio comparisons of Florigraze and UF Peace ( P = 0.101) approached significance. Root rhizome mass of Arblick, Ecoturf, and Florigraze 1 compared with 3,830 kg ha 1 for UF Peace. The comparison of r oot rhizome mass of Ecoturf (5,750 kg ha 1 ) and UF Peace approached significance ( P = 0.178). En tires with root 1 correspond to those with the
69 greatest above ground spread. Interrante et al. (2011) suggested that measuring above ground RP spread can provide an indication of how far rhizomes have spread during the growing se ason. Florigraze, Ecoturf, and Arblick had the most favorable cover, frequency, and spread charact eristics throughout the study. These data illustrate a range in allocation of resources to root rhizome vs. shoot growth that may impact long term persistence of RP stands, and despite large variation in the response appear to favor Ecoturf and Florigraze. A correlation analysis detected a positive relationship (r = 0.41, P = 0.048) between ground cover and root rhizome mass. Thus, selecting genotypes for use i n the strip planting system with a combination of desirable above and below ground plant characteristics may lead to increased potential for contribution of RP to a mixed species sward. Soil Bulk Density, Total C, N, and C:N Ratio Soil characteristics wer e measured in the planted strip before and 2 yr after imposing treatments to assess the effect of planting RP on soil traits. Soil bulk density was affected by soil depth and was 1. 21 1. 61 and 1.69 g cm 3 for the 0 to 10 20 to 40 and 40 to 70 cm str ata, respectively In the 10 to 20 cm stratum, there was a decrease in bulk density (P = 0.0613) from 1.37 to 1.29 from 2011 to 2013, although the reason for the decrease at this depth is uncertain. Differences were observed in total C at varying soil dep ths after 2 yr of imposing treatments A year effect ( P = 0.0917 ; Table 3 3 ) was observed for SOC within the surface 10 cm of soil. Total C before planting in 2011 was greater (12.5 g C kg 1 soil) than in 2013 (11.7 g C kg 1 soil). In the 10 to 2 0 cm layer, soil C decreased across years ( P = 0.0002) from 9.1 g C kg 1 in 2011 to 6.6 g C kg 1 in 2013. Defoliation management affected soil C within this layer, and hayed treatments had 32% greater total C (9.3 g C
70 kg 1 soil) than grazed plots (6.4 g C kg 1 soil). Total C did not differ among treatments within the 20 to 40 cm and 40 to 70 cm layers. Average soil C concentration in the 20 to 40 cm layer was 5.2, and it was 4.7 g C kg 1 soil for the 40 to 70 cm depth. A decrease in soil C for the 0 t o 10 and 10 to 20 cm layers likely occurred due to tillage of the strip prior to planting with RP. Soil disturbance associated with this practice may have disrupted aggregates, and caused an initial decrease in soil C until RP plants began to establish. After planting, RP unde r hay production generally had greater establishment success during Year 1 (i.e. ground cover, frequency, spread), which may explain the greater total C associated with thi s management practice compared with rotational stocking. Tot al N differed within layers (Table 3 3) and the greater N contribution occurred in the top 20 cm of soil. The 0 to 10 cm layer had the greatest total N (0.38 g N kg 1 soil), but no differences were observed among treatments. Soil N decreased from 0.42 to 0.25 g N kg 1 soil across the 2 yr study (year effect, P = 0.0032) at the 10 to 20 cm depth which may be due to tillage prior to establishment of the experiment in 2011 Contribution of soil N was low and variable at depths of greater than 20 cm. Mean so il N was 0.13 and 0.07 g N kg 1 soil for the 20 to 40 and 40 to 70 cm depths, respectively. Soil C:N ratio was not different among treatments (Table 3 3) for the surface 10 cm (20.2) and the 10 to 20 cm depth (26.7). This ratio for the 0 to 10 cm lay er approaches the threshold C:N level of 20:1 that is typically when N immobilization begins to occur (Tisdale et al., 1985). A year x defoliation regime interaction was observed for the 20 to 40 cm depth. Interaction occurred because the C:N ratio was n ot different among defoliation strategies in 2011 (32.2. vs. 45.5 for grazing vs. hay
71 management, respectively), but C:N ratio was greater for grazed (50.3) than hayed (37.4) treatments in 2013. Mean C:N ratio for the 40 to 70 cm layer was 43.8 and did no t differ among treatments. The increased C:N ratio below the 10 cm depth is consistent with the low concentration and decreasing contribution of N moving deeper into the soil profile. C Isotope Ratio Determination of C 3 and C 4 derived C and SOC Retained S table isotope ratios increased from before i nitiation of treatments to 2 yr after initiation for the 0 to 10 ( P = 0.0029) and 10 to 20 cm ( P = 0.0049) soil layers. Isotope discrimination increased (value became more negative) from 2011 to 2013 for the s urface 10 cm of soil ( 20.4 to to 20 cm depth where 13 C values changed from Despite the relative short duration of the study these data illustrate the shift in the C source from a C 4 to a C 3 based plant population affected soil C pools A more negative isotope value refl ects the influence of RP on soil C following strip planting in previously established bahiagr ass swards ( material, respectively, in the present study) The changes in the present study are within the range reported in other experiments in which vegetation type shifted from C 4 to C 3 based plant communities and vice versa (Lefroy et al., 1993; Hobie and Werner, 2004). 13 C signatures did not differ between grazed and hayed Florigraze plots at the 0 to 10 or 10 to 20 cm depths. Mean C contribution from plants associated with the C 3 photosynthetic pathway was 5.8 g C kg 1 for the 0 to 10 cm layer, and 4.7 g C kg 1 for the 10 to 20 cm layer. Soil C contribution of C 4 plants differed from before treatments were imposed to 2 yr later ( P = 0.0098) in the surface 10 cm of soil (Table 3 4) Prior to imposing treatments, C 4 derived C accounted for 54.4% of soil C whereas th e
72 contribution decreased to 49.1% by 2013. Differences across the 2 yr were also prevalent for the 10 to 20 cm layer ( P = 0.0108), and C 4 derived C decreased from 48.1 to 39.1% of the soil C from 2011 to 2013. These results agree with the stable isotope d iscrimination data, which illustrated a shift toward C contribution from a C 4 to C 3 population. The change in isotope discrimination is likely associated with maintenance of C 3 C contribution, while C 4 derived soil C decreased. Finally, t he amount of reli c SOC retained across the 2 yr observation period did not differ between defoliation regimes and was 13.0 and 8.7 g C kg 1 soil for the 0 to 10 and 10 to 20 cm depths, respectively. Implications of the Research When planted in strips, Florigraze and Ec oturf generally had the greatest mean cover and frequency of occurrence throughout the study. Ground cover differences due to defoliation regime were more apparent during the year after establishment and early in the third year, with grazed plots having le ss RP cover compared with those under hay production. Spread of RP entries was reduced under grazing every 28 d compared with the hay production treatment during the establishment year, but differences were less pronounced in the year after establishment. Greater rhizome root mass was associated with entries with greater above ground cover and may play a role in rate of RP establishment. A decrease in total soil C from 2011 to 2013 illustrates the initial impact of strip planting RP and converting from a we ll established perennial grass pasture to an establishing mixed species pasture system. Surface soil C decreased for RP under rotational stocking, and contribution of C 4 derived C decreased across the 2 yr observation period.
73 These results indicate that d ifferences exist among commercially available RP entries in their ability to establish in strip planted swards, with these results favoring Ecoturf and Florigraze. Selection of defoliation management strategy is an important consideration that can impact t he success of RP establishment during the year of planting and in subsequent growing seasons. Rotational stocking of establishing strip planted RP following conversion from an established grass pasture may also cause an initial decrease in soil C. Hay prod uction following strip planting is a more favorable option for utilizing the grass component during the establishment phase while reducing removal of RP in the planted strip. Identification of alternative grazing management strategies are needed if RP entr ies are to be grazed during the establishment phase without negatively impacting stand establishment. Finally, evaluation of strip planting into warm season perennial grass pastures other than bahiagrass would provide needed information to beef cattle prod ucers considering the adoption of this technology. Identifying complementary growth habits of both the grass and legume component in these systems will likely be needed to achieve successful stand establishment.
74 Table 3 1. Spread of RP entries during the year of ( P = 0.0086) and year after establishment ( P = 0.2424). Entry Establishment year Year after establishment ------------------------------m ------------------------------Ecoturf 0.25 a 0.82 Florigraze 0.13 a 0.83 Arblick 0.12 a 0.72 UF Pea ce 0.04 b 0.51 SE 0.08 0.12 Standard error Within a column, means without common letters differ ( P < 0.05). Table 3 2. Defoliation management effects on RP spread during the year of ( P = 0.0023) and year after establishment ( P = 0.0127). Standard error Within a column, means without common l etters differ ( P < 0.05). Defoliatio n strategy Establishment year Year after establishment ------------------------------m ------------------------------Hay production 0.25 a 0.89 a Rotational stocking 0 b 0.55 b SE 0.10 0.08
75 Table 3 3. Soil organic C, N, and C:N ratio in the surface 20 cm of strip planted rhizom a peanut from 2011 to 2013. Standard error Within a column, means without common letters differ ( P < 0.05). Table 3 4. Percentage of C 4 and C 3 derived C in the surface 20 cm of strip planted rhizoma peanut from 2011 to 2013. Layer Year 0 to 10 cm 10 to 20 cm ----------------------C 4 derived C, % of total C -------------------2011 54.4 a 48.1 a 2013 49.1 b 39.1 b SE 3.5 3.5 -----------------------C 3 derived C, % of to tal C ------------------2011 45.5 b 50.8 b 2013 51.8 a 61.0 a SE 3.7 3.2 Standard error Within a column, means without common letters differ ( P < 0.05). Layer Year 0 to 10 cm 10 to 20 cm ---------------------------------SOC, g C kg 1 soil ------------------------------2011 12.5 a 9.1 a 2013 11.7 b 6.6 b SE 0.66 0.79 ---------------------------------SON, g N kg 1 soil --------------------------------2011 0.39 0.42 a 2013 0.37 0.25 b SE 0.03 0.04 ---------------------------------C:N Ratio -----------------------------------------2011 23.2 25.0 2013 17.3 28.4 SE 3.3 3.4
76 Figure 3 1. Total monthly rainfall for 2011 and 2012 an d 30 year average rainfall for the experimental location.
77 Figure 3 2. Date x entry interaction for rhizoma peanut ground cover (% ; P = 0.0003 ) and frequency (% ; P = 0.0002 ) during the year of establishment.
78 Figure 3 3. Date x defoliation treatment interaction for rhizoma peanut ground cover (% ; P = 0.0012 ) and frequency (% ; P = 0.0356 ) during the year of establishment.
79 CHAPTER 4 SWARD CHARACTERISTICS OF RHIZOMA PEANUT GENOTYPES UNDER A RANGE OF GRAZING MANAGEMENT STRATEGIES Overview of Research Recent research has explored options for increasing the contribution of rhizoma peanut (RP; Arachis glabrata Benth.) to pasture based livestock systems in the Southeast USA (Castillo et a l., 2013; Chapter 3). As new RP cultivars are released for use in pastures, it is necessary to evaluate their grazing tolerance to provide recommendations for management (Quesenberry et al., 2010). Although several studies have described management of vari ous RP entries under clipping, few experiments have determined the influence of grazing management on RP, particularly with more recently released cultivars and germplasms. The evaluation of RP under grazing has been conducted primarily with S. et a l., 1992a; Williams et al., 2004 ; Hernndez Garay et al., 2004). Florigraze is the most widely planted cultivar in Florida (French et al., 2006), but it has become susceptible to peanut stunt virus ( Cucumovirus sp.) (Prine et a l., 2010). The use of new cultivars may provide genetic diversity and resistance against potential disease spread from infected RP stands. Since 2010, the University of Pea represent a range in growth habit from decumbent to upright, which may affect their response to different grazing management strategies. Growth habit has been shown to play a role in the response of other forages to grazing (Mathews et al., 1994; Hernndez et al., 2004). Plants under defoliation can exhib it phenotypic plasticity, or environment induced effect s on plant morphology and
80 architecture that include changes in size, positi oning, and structure of above ground plant growth (Nelson, 2000). However, the effect of growth habit and the degree to which plasticity occurs in these cultivars and germplasms have not been evaluated under differing levels of grazing ma nagement The resp onse of plants under grazing is largely related to the type of grazing practice implemented. Intensity and frequency of grazing have been cited as key factors which influence plant responses (Sollenberger and Chambliss, 1989; Sollenberger and Newman, 2007) The evaluation of how grazing management practices interact with plant growth habit will provide an understanding of how to manage RP cultivars to maintain stand productivity, persistence, and quality for long term stand production. Thus, the objectives of this experiment are to determine the impact of grazing frequency and intensity on the productivity, persistence, and nutritive value of RP entries that vary in growth habit and to define the effect of growth habit on RP response to grazing management re gimes. Materials and Methods Experimental Site A 2 yr grazing experiment was conducted at the University of Florida Beef Research Unit in Gainesville, FL (29.72 N, 82.35W). The site had previously been planted to bermudagrass [ Cynodon dactylon (L.) Pers .] from 1993 to 2003 and summer annual legumes in 2004 and 2005. In the last 5 yr, the area was unmanaged and occupied by a mixture of bahiagrass ( Paspalum notatum Flgge) and common bermudagrass. Soils at the site are Pomona sand (sandy, silceous, hyperth ermic Ultic Alaquods) and Myakka fine sand (sandy, siliceous, hyperthermic Aeric Alaquods). Prior to planting, soil samples were collected to a 20 cm depth and analyzed by the
81 University of Florida Extension Soil Testing Laboratory. Initial characterizatio n of the site indicated soil pH of 6.7 and Mehlich 1 extractable P, K, Mg, and Ca of 5, 18, 75, and 458 mg kg 1 respectively. These levels are considered to be very low for P and K and very high for Mg. Treatments and Experimental Design Thirty six, 9 m 2 plots were established in two replicates of a randomized complete block design. The experiment included 16 treatments, consisting of the factorial com binations of four RP entries and four rotational stocking treatments (32 plots total). One extra plot was established for each of the four entries to use for calibration of measuring equipment and double sampling for herbage mass. Entries included Florigraze, UF Peace, UF Tito, and Ecoturf RP. They were chosen to represent a range in plant growth habit. Florig raze and UF Peace are intermediate in growth type, whereas UF Tito and Ecoturf are upright and decumbent in growth habit, respectively. Each entry was evaluated under two levels of grazing intensity and frequency. Intensities were 50 or 75% removal of the pre grazing canopy height, and frequencies were 3 or 6 wk. Proportion of herbage removed based on height was chosen as the measure of grazing intensity because widely varying growth habits among RP entries precluded the use of a standard post grazing stubb le height. The grazing frequencies were chosen to be within a range typically recommended for RP (6 wk; Ortega S. et al., 1992a) and which would apply significant stress (3 wk), particularly when used in combination with the 75% removal treatment. In 2012, g razing was initiated on 11 and 21 June for the 6 and 3 wk frequency treatments, respectively. The final grazing event of the season occurred on 3 Oct. 2012 for the 3 wk frequency (six grazing events in 2012) and 15 Oct. 2012 for the 6 wk frequency (four grazing events in
82 2012). The second year of the study began on 5 June 2013 for the 6 wk frequency and 14 June 2013 for the 3 wk frequency treatments but only first year data are reported in this chapter Plot Establishment and Management The experimental site was initially sprayed with glyphosate on 28 July 2010 at a rate of 4.5 kg a.i. ha 1 then plowed and heavily disked. Plots were planted between 11 and 16 Aug. 20 10 using 10 cm diameter plugs that were inserted into the soil on 30 cm centers for a tot al of 100 cores per plot. Cores were harvested from existing stands of the four entries and planted within 4 h of removal from the soil. This method was chosen to expedite the establishment process compared with the use of rhizomes alone Florigraze, UF Pe ace, and UF Tito plugs were dug from established stands located at th e Agronomy Forage Research Unit in Hague, FL. The Florigraze stand had been cores were taken from areas with less common bermudagrass to minimize transfer of grass rhizomes to the experimental site. The UF Peace and UF Tito stands were 2 yr old stands that had been established as increase plots for distribution of planting material to growers. Ecoturf plugs were dug from a 20 yr old mixed stand of RP, bahiagrass, and common ber mudagrass, but the area sampled was essentially free of grass. Four weeks after planting, plots were sprayed with imazapic (Impose TM MANA, Raleigh, NC) at a rate of 0.26 kg a.i. ha 1 to control emerging broadleaf weeds and sedge. Additional plant cores w ere added to plots during late September and early October in areas where establishment of RP was slow or not successful. All plots were sprayed with clethodim (Select Max, Valent, Walnut Creek, CA) at a rate of 0.10 kg a.i.
83 ha 1 on 14 Oct. 2010 to contro l common bermudagrass that had begun to grow. Additional herbicide applications occurred in 2011 and 2012 prior to beginning of the experiment to control annual broadleaf weed and grass encroachment and promote RP stand establishment. In 2011, imazapic was applied on 11 April at a rate of 0.26 kg a.i. ha 1 for control of winter annual weeds, and clethodim was applied on 5 July and 28 September at a rate of 0.10 kg a.i. ha 1 for bermudagrass control. Imazapic and 2,4 D were applied on 17 May 2012 (0.26 and 0 .60 kg a.i. ha 1 respectively) prior to the start of the experiment in June 2012. Following the start of the experiment, no additional herbicide applications were made. During the growing seasons of the establishment period, irrigation was applied weekly to supplement rainfall when it was below long term averages in order to ensure adequate soil moisture for establishment. Total annual and 30 yr average rainfall is presented in Table 1. During the establishment year, total irrigation applied was 70 mm in A ugust and 50 mm in September. No irrigation was applied in October 2010. In 2011, 70 mm was applied in both April and May, 13 mm in June, 25 mm in August, and 50 mm in September. No irrigation was applied during the experimental years of 2012 and 2013. Soi l samples were collected annually to a depth of 15 cm and fertilization was guided by recommendations of the University of Florida Extension Soil Testing Laboratory Plots were fertilized annually with 60 kg ha 1 of K 2 O and 30 kg ha 1 of P 2 O 5 in the forms of muriate of potash and triple superphosphate, respectively, on 16 Sept. 2010, 11 Apr. 2011, and 12 Apr. 2012
84 Response Variables Herbage Mass and Accumulation Herbage mass was determined before and after each grazing event using a double sampling tech nique (Frame, 1981). The indirect measure was a 0.25 m 2 aluminum disk meter, and the direct measure was hand clipping herbage from the same quadrat areas to a 2 cm stubble. The extra (non treatment) plot of each RP entry was used to calibrate the disk so a s to minimize destructive sampling and any possible carryover effects in the treatment plots. Three double samples were collected in the extra plot and served as calibration samples for the prediction equation. Calibration samples were taken in June, Augus t, and October at a time when the regrowth in the extra plots was 3 wk and again when it was 6 wk. Indirect sampling occurred in the treatment plots before and after each grazing event. Disk heights were taken at 10 locations per plot both pre grazing and post grazing, and the average of these observations was entered into a calibration equation to predict herbage mass. Herbage accumulation (HA) was calculated as the difference between post grazing herbage mass and pre grazing herbage mass of the next graz ing event. For the first grazing event of the year, pre grazing herbage mass was considered to be HA. Nutritive Value Hand plucked samples were taken at 10 locations per experimental unit prior to each grazing event for nutritive value determination. Locat ions were selected in a grid pattern to represent the entire plot, and individual samples were harvested using hand shears to the target stubble as defined by pre grazing height measurements. The 10 samples per plot were composited for analysis. In additio n, the samples from
85 consecutive 3 wk grazing events (i.e., Grazing Events 1 + 2, 3 + 4, and 5 + 6) were composited after drying for labor atory analyses. Samples from each grazing event of the 6 wk frequency were analyzed separately, providing one nutritive value sample every 6 wk for each experimental unit. Nutritive value analyses included crude protein (CP) and in vitro digestible organic matter (IVDOM) concentrations. Crude protein was estimated using a micro Kjeldahl technique for N (Gallaher et al., 1 975) and the two stage technique for IVDOM (Moore and Mott, 1974). A weighted total season herbage CP concentration was calculated as the CP concentration multiplied by herbage accumulation for each grazing event, summing these numbers across grazing event s, and dividing that sum by total herbage accumulation for the season. A weighted value for IVDOM was estimated using the same calculation. Early, mid and late season CP and IVDOM were also compared. For the 3 wk treatment, Grazing Events 1 and 2 were t ermed early, 3 and 4 were mid and 5 and 6 were late season. These time periods of the production season will be referred to as seasons throughout the remainder of the chapter. For the 6 wk treatment, Grazing Event 1 was early, 2 was mid and 3 and 4 wer e late season. For the seasonal comparisons, c onsecutive grazing events of the 3 wk treatment were composited as described earlier, and weighted CP and IVDOM were calculated for the third and fourth grazing events (late season) of the 6 wk treatment. Pre grazing Sward Height Sward height was measured before each grazing event at 15 locations per plot. Because of the definition of the grazing intensity treatment, pre grazing heights were used to determine target post grazing height for each entry at each gr azing event.
86 Height determinations were also used to compare entries in terms of relative growth potential and canopy structure under differing defoliation regimes. Data are reported by seas on and grazing events were averaged within early (June/early July ), mid (late July/August), and late season (September/October), and compared across grazing frequencies. Pre grazing Leaf to Stem Ratio Leaf to stem (L:S) ratio of the RP canopy was measured pre grazing twice during each grazing season in July and again i n August/September. A set of 10 hand plucked samples were harvested to the target post grazing stubble height for each entry. After collection, samples were separated into leaf, including petiole, and stem, dried, and weighed to determine L:S ratio of the grazed portion of the canopy. Pre grazing Canopy Light Interception Light measurements were taken at the same time as the leaf to stem ratio measurements using a SunScan SS1 (Dynamax Inc., Houston, TX) to provide information on the relative extent of cano py regeneration following grazing events. The unit consisted of a 1 m long quantum sensor that was used to measure photosynthetically active radiation (PAR) at the bottom of the RP canopy. A beam fraction sensor was placed outside of the plots in full sunl ight to measure incident PAR. Canopy light interception was calculated as the transmitted PAR divided by the incident PAR multiplied by 100 to obtain a percentage. Measures were taken the day prior to grazing between 1000 and 1200 h at four locations per p lot. Post grazing Residual Leaf Area Index Short post grazing stubble heights and the decumbent growth habit of Ecoturf RP limited the accuracy of light interception measurements as a means of characterizing
87 treatment effects on post grazing photosyntheti c capacity. Quantifying residual leaf area was chosen as an alternative approach. Leaf area was measured following a grazing event in July and again in August/September. Live, photosynthetically active leaves (including petiole) were removed from four, 20 by 20 cm quadrats per plot. Sites were selected to be representative of the average condition for that plot. Leaves were put in plastic bags and placed immediately on ice in a cooler for transport to the lab for area analysis. Samples were analyzed using a LI COR Biosciences, Lincoln, NE) rolling leaf area meter. Individual leaves were placed directly onto the meter with no overlap to estimate total leaf area. Total leaf area was divided by the area of the quadrat to determine leaf area index rema ining following a defoliation event. Sward Botanical Composition Botanical composition was quantified for each treatment prior to the first and last grazing each year. Percentages of RP, grass, and weeds were determined as the mass of each component prese nt divided by total pre grazing herbage mass. Measurements were made by hand clipping two representative 0.25 m 2 quadrats per plot to the target post grazing height and separating the fresh herbage into respective components. Component samples were dried a t 60C until constant weight and weighed. Rhizoma Peanut Ground Cover Percent RP cover was estimated prior to the first and last grazing event of each year. A 2 m by 0.5 m quadrat was used that was divided into 100, 10 by 10 cm squares. The quadrat was pl aced at two locations per plot, and for each location 20 squares were evaluated. Within each square, the percentage of ground cover by RP was estimated visually by a trained evaluator.
88 Weed Frequency To assist in characterizing changes in botanical compos ition, frequency of occurrence was measured for weeds (collectively, not by species). These measurements were done in conjunction with the cover estimates. At each placement of the 2 by 0.5 m frame within the plot, 20 individual 10 by 10 cm quadrats were evaluated for weed frequency. If any weed plant matter was present in the quadrat it plots in which weeds were present was the measure of weed frequency (WF). Rhizome Mass Rhizome root samples were taken immediately after the first and last grazing event in 2012. Samples were taken using a 10 cm soil coring device to a depth of 20 cm. Previous studies have shown this depth encompasses the entire rhizome mat plus ~10 cm of soil below (Saldivar et al., 1992; Rice et al., 1995). Four samples were taken per plot at each date. Above ground biomass was removed and the samples composited prior to washing over a 2 mm mesh, window type screen to remove soil. Samples were exposed to temperatures of 100C for 1 h to quickly stop respiration, and dried at 60C to constant weight. After drying, samples were weighed to determine belowground biomass. Rhizome TNC and N Rhizome s amples were ground to pass a 1 mm screen in a Wiley mill a nd analyzed to determine total non structural carbohydrate (TNC) and N concentration. The TNC concentration was determined by a modification of the procedure of Christiansen et al. (1988) that was described in detail by Chaparro et al. (1996). This procedu re uses invertase and amyloglucosidase to convert starch and oligosaccharides
89 into monosaccharides and measures reducing sugars with a photometric copper reduction method (Nelson, 1944). The N analysis was conducted using a micro Kjeldahl technique with a colorimetric analysis of N. Statistical Analysis Data were analyzed using PROC MIXED of SAS (SAS Institute, Cary, NC, 1996). The experiment was conducted in 2012 and 2013, but only 1 yr of data are included in the dissertation. Entry, defoliation frequency defoliation intensity, and their interactions were considered fixed effects and block and interactions with block were random. When a responses was measured at multiple dates, date was considered a repeated measures with an autoregressive covariance stru cture. Differences were declared when P P described when P Mean separation for entry effects was based on the PDIFF option of LSMEANS in SAS. Defoliation frequency and in tensity means were separated using the F test. Results and Discussion Herbage Accumulation No differences were observed among genotypes for total season HA, and the 1 for all entries. Average seasonal production of RP wa s 7 to 11 Mg DM ha 1 throughout the southeastern USA (Terrill et al., 1996; Venuto et al., 1998). During a 4 yr trial in Citra, FL, DM yield of Ecoturf, UF Tito, UF Peace, Florigraze, and Arbrook ranged between 8.3 and 12 Mg ha 1 illustrating the yield po tential of RP (Prine et al., 2010). Thus, this value is within the range reported by other studies conducted in this region. There was no effect of grazing intensity or frequency during the first year of the study on seasonal herbage accumulation
90 When eva luating treatment effects on seasonal dist ribution of HA interactions were observed for season x entry ( P = 0.0124) and season x frequency ( P = 0.0183). Differences among entries were apparent for early seaso n (Figure 4 1). Florigraze had greater HA than all other entries during this time period. During the midd le of the season, Florigraze (1 750 kg ha 1 ) and Ecoturf HA decreased (1 500 kg ha 1 ), and Ecoturf HA was less than that of the upright genotypes UF Tito (2 440 kg ha 1 ) and UF Peace (2 300 kg ha 1 ). Fl origraze HA was intermediate to both low and upright growing entries. No differences were observed among entries for late season HA. Ecoturf maintained a similar level of production from mid to late season, but there was a pronounced decrease in HA from mid to late season for more erect growing lines. Williams et al. (2008) noted that RP DM production decreases with decreasing daylength, and that the magnitude of response is greater in entries selected as forage types with greater DM production. In green house and field evaluations of RP lines under natural daylength or extended photoperiod, Ecoturf was less daylength sensitive in aboveground production than other forage ecotypes (Williams et al., 2008). The authors suggested that because Ecoturf has been selected as a low maintenance ornamental genotype, production potential is less sensitive to changes in daylength. In an evaluation of shoot and root growth of RP during the establishment year, Saldivar et al. (1992) observed shoot DM production began to p lateau or decline in September, while rhizome growth continued. Shoot to rhizome ratios increased from 0 at planting to 2 by late summer, but decreased to 0.5 in the fall. The authors attributed this to photosynthate partitioning from shoots to rhizomes, w hich further supports that herbage production of RP is responsive to photoperiod.
91 A date x frequency interaction (Figure 4 2) sh owed greater HA for the 6 wk (2 750 kg ha 1 ) than 3 wk frequency (2230 kg ha 1 ) from June through early July. From late July thro ugh August, HA decreased for the 3 wk frequency co mpared with the previous cycle (1 290 kg ha 1 ), but HA was maintained (2,720 kg ha 1 ) for the 6 wk frequency. Herbage accumulation was lowest during September and October for both frequencies, but HA remaine d greater for swards grazed every 6 wk vs. those defoliated every 3 wk. These results illustrate that the average HA per defoliation event was greater for the 6 wk frequency during Year 1 of the study, but total season HA did not differ because of greater number of grazing events for the 3 wk frequency. Further investigation of the effect s of grazing management strategies is being conducted in 2013 to determine the ir impact on a longer time scale. This is important because the impact of grazing management o n HA of Florigraze RP was much greater in the second year of imposing treatments than in the first (Ortega S. et al., 1992a). Nutritive Value Total 1 DM for all entries. There were differences among entries ( P < 0.0001), with Florigraze having a lower CP concentration (170 g kg 1 ) than UF Peace, Ecoturf, and UF Tito (~195 g kg 1 ). Grazing frequency affected CP ( P = 0.0369), and concentration was reduced for the 6 wk frequency (190 g kg 1 ) compared with the 3 wk treatment (200 g kg 1 ). T here was an entry x frequency interaction ( P = 0.0898) for CP. UF Peace and UF Tito had greater CP ( P = 0.0244 and P = 0.0400, respectively) when grazed every 3 wk (mean 210 g kg 1 ) than under a longer regrowth interval of 6 wk (mean 190 g kg 1 ). Florigraze had lower CP (170 g kg 1 for 3 wk and 6 wk frequency; P = 0.04096) compared with all other
92 entries, and Ecoturf w as not different ( P = 0.3248) between the 3 wk (200 g kg 1 ) and 6 wk frequency (195 g kg 1 ). A longer regrowth interval likely caused increased stem growth relative to leaf in more upright growing entries, which may have decreased CP concentrations. Altho ugh differences were observed among genotypes, these values exceed the protein requirements of all classes of beef cattle (NRC, 1996), and illustrate the high quality of RP forage. Romero et al. (1987) reported 190, 186, and 180 g CP kg 1 for Florigraze RP harvested every 6, 9, and 12 wk, respectively, during the summer growing season. Across a range of harvest methods, frequencies, and stubble heights, Butler et al. (2007) observed 186 to 204 g CP kg 1 illustrates the ability of RP to maintain nutritive value across a wide range of defoliation management practices. T here was a season x entry effect ( P < 0.0001). Interaction occurred because CP was greater for UF Peace (240 g kg 1 ) than all other genotypes from June to mid July. UF Tito and Ecoturf had lower CP (210 g kg 1 each) than UF Peace, but were greater than Florigraze (187 g kg 1 ) during this period. From mid July through August, UF Peace maintained a similar CP concentration to that in early season (230 g kg 1 ), but Ecoturf CP increased (220 g kg 1 ) compared with UF Tito and Florigraze (200 g kg 1 for both). During September and October, CP concentration decreased for all lines, and Ecoturf maintained a greater CP (175 g kg 1 ) compared with the more upright g enotypes (mean 155 g kg 1 ). A frequency effect approached significance ( P = 0.0535) for total season IVDOM, and IVDOM tended to be greater for the 6 than the 3 wk frequency (720 vs. 700 g kg 1 ).
93 These values are both in the upper range of digestibility t hat can be achieved in most forage livestock production systems in the southern region (Ball et al., 2007). In a 2 yr evaluation of RP across a range of harvest frequencies, IVDOM decreased from 750 early to 500 g kg 1 in the late season (Saldivar et al., 1990). Decreasing IVDOM was associated with a decline in leaf percentage across the season (Saldivar et al., 1990). Romero et al. (1987) observed a digestibility range of 570 to 610 g kg 1 for Florigra ze RP hay harvested every 6 to 12 wk. When comparing s easonal patterns of response, there was a season x frequency interaction ( P = 0.0115) for IVDOM. From June to October 2012, digestibility decreased more for the 3 wk (730 to 650 g kg 1 ) compared with the 6 wk frequency (730 to 700 g kg 1 ). While there were no differences between the 50 and 75% grazing intensity treatments from June through August, late season percentage IVDOM was greater (season x intensity interaction; P = 0.0955) for the 75 than the 50% treatment (690 vs. 670 g kg 1 respectively ; P = 0.0 227 ). Overall, the impact of grazing management on nutritive value was limited and RP nutritive value was high regardless of grazing treatment. Pre grazing Sward Height There were season x entry ( P = 0.0004) and season x frequency ( P < 0.0001) effects for RP pre grazing sward height (Figures 4 3 and 4 4, respectively). At the beginning of the growing season (June/early July), sward height differences were apparent, with UF Tito being taller (15 cm) than all other entries. Florigraze and UF Peace were simil ar in height (13 cm), but both were taller than Ecoturf (10 cm). Height differences among entries followed a similar pattern throughout the remainder of the growing season. Pre grazing sward height was tallest for all entries during late July
94 through Augus t and decreased from September to October. Quesenberry et al. (2010) described UF Tito as an upright ecotype, whereas UF Peace is more intermediate in growth habit like Florigraze. Anderson et al. (2012) evaluated the response of 16 RP selections to sun an d shade tolerance, and observed that sward height was more affected by inherent growth characteristics of RP selections than by various shade treatments. The height observations in the present study further support differences in growth habit among release d genotypes. The effect of grazing frequency on sward height differed across the season (Figure 4 2), but this interaction was primarily due to timing of the first grazing event. Height was greater for the 3 than the 6 wk frequency at the beginning of th e season because grazing was initiated 1 wk later for the 3 wk treatment than the 6 wk treatment. From mid July through August, entries grazed every 6 wk were tallest (22 vs. 16 cm), and both treatments achieved their greate st heights during this season. A t the end of the growing season, sward height decreased for both frequencies (16 and 12 cm for 6 and 3 wk, respectively), and treatments grazed less frequently achieved a taller height. Maroso et al. (2007) observed similar results for birdsfoot trefoil ( L otus corniculatus L.) clipped every 2 or 4 wk to a 4 and 8 cm stubble height s Grazing intensity also affected sward height, with the 50% canopy removal treatment having a greater ( P = 0.0082) pre grazing sward height (15 cm) compared with the 75% treatme nt (13 cm). Although these height differences were small, this illustrates that grazing intensity can impact subsequent sward regrowth potential. Regardless of plant growth habit, the longer regrowth interval and greater residual
95 stubble height favored tal ler sward heights compared to the more intensive management strategies. Pre grazing Leaf to Stem Ratio For two sampling dates in 2012, there were date x frequency ( P = 0.0230), date x entry ( P = 0.0315), and entry x frequency ( P = 0.0235) interactions for L:S ratio. In July, a greater L:S ratio was observed for the 6 than the 3 wk frequency (1.7 vs. 1.3, respectively) while in August/early September there was no difference (1.31 vs. 1.27, respectively). This response is unexpected based on typical response for other forages, but it likely reflects the relatively slow regrowth of RP immediately following a defoliation event that results in maintaining a high L:S even through 6 wk. Romero et al. (1987) observed that for fall harves ted RP, L:S ratio was less w ith a regrowth interval of 6 wk, than for 9 to 12 wk. The authors suggest that this was likely a function of defoliation stress, where a longer regrowth interval was needed to regenerate leaf area. The L:S response is consistent with the trend toward great er IVDOM at 6 than 3 wk defoliation frequency. Ranking of entries in L:S differed across dates (Table 4 1). Ecoturf had a greater L:S ratio (2.04) during July compared with the more erect growing ecotypes. UF Tito and Florigraze had a similar L:S ratio ( 1.75 and 1.67, respectively), and both of them were greater than UF Peace (1.41). In August/early September, although L:S ratio decreased for all entries, Ecoturf and UF Peace had the least change in L:S ratio (1.46 and 1.27, respectively) compared with Fl origraze (1.21) and UF Tito (1.17). The decrease in L:S ratio later in the season is likely associated with a decrease in regrowth as herbage accumulation decreased from the beginning to the end of the season, and the presence of residual stem following a defoliation event.
96 The entry x frequency interaction (Table 4 2) occurred because for the 6 wk defoliation frequency Ecoturf had the greatest L:S ratio (1.87) while for the 3 wk frequency Ecoturf (1.53) and UF Tito had similar L:S ratio (1.50), but both w ere greater than UF Peace (1.29) and Florigraze (1.33). Ecoturf was the only entry for which L:S was greater at a 6 than a 3 wk grazing frequency. The short canopy height of Ecoturf reduced the contribution from stem, and the amount of leaves associated w ith the dense growing canopy resulted in a greater L:S ratio. In more upright growing ecotypes, L:S was decreased because of increased stem growth. In clipping studies with RP harvested every 2 6 or 8 wk, leaves were 60 to 80% of the shoot component of Florigraze RP (Saldivar et al., 1990). Beltranena et al. (1981) reported that leaf fractions of established Florigraze RP declined seasonally when harvested under clipping every 56 d. Thus, increased stem contribution and less leaf regeneration later in t he season may decrease L:S ratio. Saldivar et al. (1990) noted that Florigraze RP harvested every 2 wk assumed a prostrate growth habit by mid season, and was characterized by small leaves that were not readily removed during a defoliation event. When clip ped every 6 or 8 wk, RP had elongated stems, and much of the leaf area was removed during harvest. This illustrates that plasticity of RP under defoliation may impact L:S ratio. Weijsched et al. (2007) observed morphological plastici ty among white clover ( Trifolium repens L.) selections when grown in competition with perennial ryegrass ( Lolium perenne L.) and managed under clipping every 48 d. Total shoot mass, internode length, petiole length, and total ramet number of clover decreased when clipped compar ed with the undefoliated control. The authors attribute this to a change in canopy structure from vertical to more horizontal, which illustrates the role of growth
97 habit and degree of plant plasticity in response to differing levels of defoliation manageme nt. Pre grazing Canopy Light Interception Light interception was affected by sampling date ( P < 0.0001), and pre grazing LI decreased from July (92%) to August/early September (84%). The greater LI occurred during the period of maximum herbage accumulatio n of RP. Percentage LI was greater ( P = 0.0219) for Ecoturf, UF Peace, and UF Tito (89, 90, 89%, respectively) than Florigraze (85%). Although differences occurred among entries, these values were relatively high throughout Year 1, and may explain the lack of difference in seasonal herbage accumulation among entries during 2012. Grazing frequency affected LI ( P < 0.0001), with the 3 wk frequency having reduced LI (85%) compared to 6 wk (91%). Although total HA was not different among grazing frequency treat ments, a shorter regrowth interval and lower percentage of LI for the 3 wk frequency decreased herbage accumulation within a grazing cycle compared with RP defoliated every 6 wk. Post grazing Residual Leaf Area Index A date x entry interaction ( P = 0.0384 ) occurred for post grazing RLAI (Table 4 1). UF Peace, UF Tito, and Ecoturf had greater RLAI (mean 1.34) than Florigraze (0.87) during July. However, in August, Ecoturf and UF Peace had greater RLAI (1.07 and 0.98, respectively) following a grazing event than UF Tito and Florigraze (0.75 for both entries). Florigraze had lower pre grazing canopy LI compared with the other entries from July to September, which may have contributed to a lower post grazing RLAI. Increasing the intensity of defoliation decreas ed ( P < 0.0001) the post grazing RLAI (0.89 for 75% removal vs. 1.22 for 50%). More photosynthetically active leaf area remaining following a grazing event likely enhances regrowth potential. With fewer
98 leaves for regrowth, defoliated plants rely upon rese rves to produce new aboveground growth (Richards, 1993). Ortega S. et al. (1992a) observed that canopy regrowth following a defoliation event was greater when RP post gr azing residual dry matter was 1 700 kg DM ha 1 or more, and that residual leaf area was a primary factor contributing to regrowth under less intense grazing strategies. In the present study, early and mid season post grazing he rbage mass ranged from 750 to 1 900 kg DM ha 1 for the 6 wk frequency, but 1 for the 3 wk treatme nt across the 50 and 75% removal intensity kg DM ha 1 for each frequency x intensity treatment. These values are lower than those recommended by Ortega S. et al. (1992a), indic ating that all treatments were putting significant stress on RP. It is useful to note that greater mid season height and HA for the 6 wk frequency were associated with greater residual DM as a function of the structure of the grazing management treatments. Sward Botanical Composition Composition was measured in June and October 2012 and June 2013. Percentage of sward components was affected by date for RP ( P = 0.0001), grass ( P = 0.0003), and weeds ( P = 0.0001). Percentage RP decreased from 91 to 83%, weed s increased from 5 to 11%, and grass increased from 4 to 7% from June to October 2012. In June 2013, stands consisted of 70% RP, while grass and weed presence increased to 10 and 19%, respectively, from the previous fall. Increased weed percentage from Oct ober 2012 to June 2013 may be partially associated with the increased presence of winter annual weeds. In addition, there was no herbicide applied during the
99 experimental period starting from June 2012, so weed encroachment was not controlled during this p eriod. Hernndez Garay et al. (2004) observed a decrease of 89 to 66% for continuously stocked Arbrook RP after 3 yr of grazing management, whereas RP percentage was maintained for Florigraze (90 to 87%) during this same time period. The authors suggest t hat the upright growth of Arbrook made it less tolerant of continuous stocking compared with Florigraze. This illustrates that choice of grazing management may have more long term impacts on RP stand percentage than observed in Year 1 of the present study. An entry x frequency interaction (Table 4 3) occurred for percentage of RP and grass ( P = 0.0200 and P = 0.0355, respectively). There were no differences in RP percentage due to defoliation frequency for UF Peace, UF Tito, and Ecoturf, but Florigraze plo ts had greater RP associated with the 6 wk (84%) compared with the 3 wk frequency (67%). More grass encroachment occurred in Florigraze plots grazed every 3 wk (17%) than those under the longer rest period (6%), but frequency did not impact grass percentag e for UF Peace, UF Tito, and Ecoturf. Although bermudagrass rhizomes associated with Florigraze planting material may account for some of the observed response, these data illustrate that a shorter rest period between grazing events increased competition between grass and RP. Ortega S. et al. (1992a) observed that grass contribution in RP pastures increased with decreasing intervals between grazing cycles and dec reasing residual herbage mass. In that study, g rass accumulation in Florigraze RP pastures was lowest with a regrowth interval of 21 d or greater, and residual herbage mass of 1 500 kg DM ha 1
100 Rhizoma Peanut Ground Cover There were sampling date x frequency ( P < 0.0001) and entry x frequency interactions ( P = 0.0005) for RP ground cover. In June 201 2 before grazing treatments were imposed, ground cover was greater for plots assigned to the 6 wk (90%) than the 3 wk (79%) grazing interval (Figure 4 5). However, by the end of the grazing season, ground cover increased to 93% for the 3 wk treatment and w as greater than that for the 6 wk treatment (87%). At the beginning of the 2013 grazing season, cover did not differ among defoliation frequencies (mean of 92% for both frequencies). For the entry x frequency effect (Figure 4 6), Ecoturf ground cover was g reater when defoliated on a 6 wk interval than every 3 wk and achieved a higher cover than all other lines. UF Peace grazed every 6 wk had greater cover than when grazed every 3 wk (96 vs. 86%, respectively). However, the upright growing UF Tito had less g round cover when grazed every 42 d compared with the 21 d frequency (84 vs. 91%, respectively). Percent cover of Florigraze did not differ between frequenc There was entry x frequency interaction ( P = 0.0785) for change in ground cover from June 2012 to June 2013. Florigraze and UF Peace grazed every 3 wk had a net positive increase in cover (+ 22 and 13%, respectively), while there was less change in cover for the 6 wk frequency (0% for both lines). An increase in ground cover was obser ved for both the 3 wk (+ 14%) and 6 wk (+ 7%) frequencies for UF Tito, whereas Ecoturf had the least change of all entries across frequencies (+ 4% for 3 wk and 0% for 6 wk frequencies, respectively). These results suggest that intermediate to upright grow ing ecotypes may be responding to more frequent grazing by altering their growth habit, resulting in a greater change in cover ratings. Grazing tolerant grasses have been characterized as shifting growth from more upright to decumbent to avoid defoliation
101 (Chapman and Lemaire, 1993). When managed to maintain a low level of herbage continuous stocking and four levels of herbage allowance (9, 35, 81, and 148 kg DM [100 kg] 1 BW ). Under increased grazing pressure, there was increased ground cover and a decrease in the amount of prehensible forage for animals. Thus, increasing forage ground cover can be associated with more intensive or frequent grazing, but the long term effects of frequent defoliation are not yet clear in the current study. Weed Frequency A date x entry interaction ( P = 0.0110; Figure 4 7) occurred for WF. In June frequency decreased for all lines at the end of the 2012 season; however, Florigraze continued to have greater WF (49%) th the start of Year 2 of the study (June 2013), Florigraze had greater weed occurrence than UF Tito (60 vs. 45%), but did not differ from UF Peace and Ecoturf (55% for each line, respectively). Increased weed enc roachment in June 2013 compared with fall of 2012 was likely due to the presence of winter annual weeds in all entries, but the presence of bermudagrass in Florigraze plots contributed toward the higher WF observed for that cultivar. A date x frequency in teraction ( P < 0.0001; Figure 4 8) occurred for WF. Before treatments were imposed, the 3 wk frequency had greater WF (53% ; P < 0.0001 ) than the 6 wk treatment (31%), but WF decreased for both frequencies by October 2012 (34 and 37% for the 3 and 6 wk tre atment, respectively). Weed encroachment increased from October 2012 to June 2013, but no differences were observed among frequencies
102 (mean 55%). Grazing intensity affected WF ( P = 0.0184), with less weed presence for the 50 than 75% canopy removal treatme nt (41 and 47%, respectively). There also was entry x intensity interaction for WF ( P = 0.0835) W eed frequency in Florigraze and Ecoturf was greater ( P = 0.0236 and P = 0.0391, respectively) when 75% of canopy height was removed (62 and 43%, respectively ) than when 50% of height was defoliated (52 and 32%, respectively). Weed frequency of UF Tito and UF Peace was less than Florigraze for both the 50% ( P = 0.0002 and P < 0.0001, respectively) and 75% removal treatment ( P = 0.0008 and P = 0.0007, respective ly) but they were similar to Ecoturf. Hernndez Garay et al. (2012) observed that for an alfalfa ( Medicago sativa L.) and orchardgrass ( Dactylis glomerata L.) mixed pasture under a high grazing frequency (rotational stocking every 25 to 35 d) and intensit y (target stubble 3 to 6 cm), forage mass could be increased, but there was greater risk for weed presence when managed more aggressively. Hoveland et al. (1996) conducted a 3 yr evaluation on the effect of harvest frequency on weed encroachment in grazing tolerant alfalfa cultivars in north Georgia. During Year 3, the authors reported that genotypes harvested every 2 wk had decreased persistence, production potential, and increased bermudagrass and crabgrass ( Digitaria sanguinalis L.) presence compared to a 4 or 6 wk harvest frequency (Hoveland et al., 1996). Although the observed effects for grazing frequency and intensity were more temporal in the present study (i.e., effects changed with season, etc.), these studies illustrate the potential for more lon g term effects of these management strategies on weed encroachment. When change in WF was evaluated from June 2012 to the beginning of Year 2, it increased (frequency effect; P < 0.0001) to a greater degree for the 6 (+ 19%) than for
103 the 3 wk frequency ( + 5%). This occurred because WF was initially greater for the 3 wk treatment, but decreased across the season in 2012, whereas WF increased during this time frame when defoliated every 42 d. Increase in WF among entries (entry effect; P = 0.0062) was great er for Ecoturf (+ 24%) and UF Peace (+ 23%) than UF Tito (+ 7%) and Florigraze (+ 5%). These data agree with the botanical composition data from June 2013 that illustrates a decrease in the percentage of RP since the beginning of the experiment in 2012. R hizome Mass Rhizome mass decreased (date effect; P = 0.0104) from 4,270 to 3,710 kg DM ha 1 from the beginning to the end of the 2012 grazing season. More frequent grazing decreased rhizome mass ( P = 0.0274), with means for the 3 wk frequency of 3,750 kg h a 1 compared with 4,230 kg ha 1 for the 6 wk frequency. Ortega S. et al. (1992b) observed no differences in Florigraze RP rhizome mass follo wing 1 yr of grazing, but during the spring of Year 2, there was a trend for decreasing mass for RP defoliated every 7 d with residual DM of 500 kg ha 1 By the end of Year 2, rh izome mass ranged from 400 to 4 100 kg ha 1 among treatments. Specifically, the lowest levels of rhizome mass were associated with short regrowth cycles (7 d) and low levels of residual herbage m ass (500 kg DM ha 1 ). Increasing the length of the regrowth cycle (> 7 d) increased rhizome mass w hen residual herbage mass was 1 000 kg ha 1 or less. Howe ver, for treatments exceeding 1 700 kg ha 1 residual herbage mass, the effect of less frequent grazing was diminished. Under more frequent and intense defoliation, RP reserve resources decreased in order to sustain forage DM accumulation. Liu et al. (2011a) noted decreased root rmudagrass from 9090 to 7 080 kg ha 1 as postgraze stubble height decreased from 24 to 8 cm. Mousel et al.
104 (2005) quantified root growth and reserve characteristics of big bluestem ( Andropogon gerardii Vitman) under different levels of grazing intensity and frequency. Root mass, surface area, and volume d ecreased more when grazed more frequently (< 40 d between rest periods) than under longer recovery periods. Thus, the effect of grazing management strategy can influence root reserves, and similar responses were observed in the present study. Differences w ere also observed among entries for rhizome mass ( P < 0.0001). Ecoturf and UF Tito had greater rhizome mass (4,750 and 4,350 kg ha 1 respectively) than UF Peace and Florigraze (3,200 and 3,660 kg ha 1 ). Although rhizome mass decreased across the season, t here were no differences among entries between dates (date x entry interaction, P > 0.10 ), which suggests that Ecoturf and UF Tito may partition more resources to below ground growth than UF Peace and Florigraze. These data are also within the range report ed for RP entries under clipping or rotational stocking every 28 d in Chapter 3. Rhizome TNC and N Rhizome TNC concentration was affected by date x entry interaction ( P = 0.0424). Grazing treatments did not affect TNC concentration during the first year of the study. In June 2012 before grazing treatments were imposed, no differences were observed among entries and TNC concentration ranged from 104 to 126 g TNC kg 1 DM. An interaction was observed because TNC concentration decreased from June to October (12 6 vs. 99 g TNC kg 1 ) for UF Tito, but TNC concentration was similar among all other lines across the season (mean 110 g TNC kg 1 DM in October). Although rhizome mass of UF Tito did not decrease across the season, decreased TNC
105 concentration in the fall ma y be attributed to reallocation of resources for canopy regeneration. Saldivar et al. (1992) reported that TNC concentration was low and variable in Florigraze under clipping every 2, 6, or 8 wk, but ranged from 100 to 200 g kg 1 during the summer growing season. Beginning in November, RP TNC concentration began to increase to levels > 400 g kg 1 Ortega S. et al. (1992b) observed 68 to 154 g kg 1 TNC for Florigraze RP rhizomes across a range of grazing frequencies (7 to 63 d regrowth) and intensities (res idual DM 500 to 2500 kg ha 1 ) following 1 yr of management. After 2 yr of grazing, TNC concentration ranged from 58 to 210 g kg 1 and was influenced by grazing cycle length and post grazing residual DM. Rhizome TNC concentration increased with increasing 1 but the effect of grazing frequency was neglible when post grazing herbage mass 1 During both years of the study, the lowest TNC concentration was associated with low r esidual herbage mass (500 kg ha 1 ). Thus, the TNC values reported in the present study are relatively low, but fall within this observed range following 1 yr of management. A date x entry interaction ( P = 0.0178) occurred for rhizome N concentration. Mean 1 for all lines in June 2012, and did not differ among entries. Florigraze maintained a similar N concentration across the season (18 g kg 1 1 DM for UF Tito, Ecoturf, and UF Peace from June to October. These values are similar to Ortega S. et al. (1992b) who reported 13 to 16 g kg 1 for RP rhizomes with post grazing resi dual herbage mass from 500 to 2 500 kg ha 1 Although N concentration changed across dates, there was no effect of grazing
106 f requency or intensity on rhizome N. While changes in the amount of rhizome storage carbohydrates occur more readily under different levels of defoliation (Saldivar et al., 1992; Rice et al., 1995), N concentration has been observed to be a more transient c ompound used in respiration that is not stored in a large concentration (Richards, 1993), which may make treatment differences less prevalent. Implications of the Research Grazing management strategies were observed to impact sward characteristics of RP g enotypes during Year 1 of the study. Although there were no differences among grazing treatments for total HA, similar or greater leaf to stem ratio and pre grazing light interception, and maintenance of a high percentage of ground cover with the 6 wk freq uency suggest an advantage for longer regrowth intervals. Greater RLAI, less WF, and trends for greater pre grazing sward height associated with the 50 vs. 75% removal level may favor RP persistence. Increased cover for the 3 wk frequency at the end of Yea r 1 may be associated with canopy structural adaptation of RP genotypes under short regrowth cycles. A second year of evaluation is being conducted through summer 2013 to quantify continuing changes in sward canopy characteristics of entries under these st rategies. The results indicate that while total HA was not affected by grazing management strategy in Year 1, changes in other above and below ground sward characteristics suggest that RP genotypes may favor a 6 vs. a 3 wk regrowth interval and that thes e differences may become more pronounced over time. Selection of RP genotypes that exhibit grazing tolerance through increased production, persistence, and nutritive value will likely increase the use of these entries in pastures by producers.
107 Table 4 1. Date x entry interaction for rhizoma peanut leaf to stem ratio ( P = 0.0315) and post grazing residual leaf area index ( P = 0.0384; RLAI). Data are means across two grazing intensities, two frequencies, and three replicates (n = 8). Leaf to stem ratio Po st grazing RLAI Entry July Aug./Sept. P value July Aug./Sept. P value Ecoturf 2.04 1.46 a <0.0001 1.33 a 1.07 a 0.0249 Florigraze 1.67 b 1.22 b 0.0003 0.87 b 0.73 b 0.2232 UF Peace 1.41 c 1.27 ab 0.2291 1.38 a 0.99 a 0.0011 UF Tito 1.75 b 1.18 b <0.0001 1.34 a 0.75 b <0.0001 SE 0.09 0.10 0.11 0.12 Table 4 2. Entry x frequency interaction ( P = 0.0235) for rhizoma peanut leaf to stem ratio. Data are means across two dates, two grazing intensities, and two replicates (n = 8). Table 4 3. Entry x frequency interaction for percentage of rhizoma peanut ( P = 0.0200) and grass ( P = 0.0355) in total herbage mass above the target stubble height. Data are means across two grazing intensities and tw o replicates (n = 4). Rhizoma peanut Grass Frequency (wk) Frequency (wk) 3 6 3 6 Entry -------------% -------------P value -------------% -------------P value Ecoturf 84 85 0.5461 8 b 3 0.7670 Florigraze 67 b 84 <0.0001 17 a 6 <0.0001 UF Peace 82 a 84 0.2970 5 b 3 0.8849 UF Tito 80 a 84 0.5607 7 b 6 0.9785 SE 4 4 2 2
108 Figure 4 1. Season x entry interaction ( P = 0.0124) f or herbage accumulation (kg DM ha 1 ) of RP genotypes. Early, mid and late season dates correspond to June/early July, late July/August, and September/October, respectively, in 2012. Data are means across two grazing intensities, two grazing frequencies, and two replicates (n = 8).
10 9 Figure 4 2. Season x grazing frequency ( P = 0.0183) interaction for rhizoma peanut herbage accumulation. Early, mid and late season dates correspond to June/early July, late July/August, and September/October, respectively, in 2012. Data are means across four rhizoma peanut entries, two grazing intensities, and two replicates (n = 16).
110 Figure 4 3. Season x entry interaction ( P = 0.0004) for pre grazi ng sward height of RP genotypes under different levels of grazing management. Early, mid and late season dates correspond to June/early July, late July/August, and September/October, respectively, in 2012. Data are means across two grazing frequencies, t wo grazing intensities, and two replicates (n = 8).
111 Figure 4 4. Season x grazing frequency interaction ( P < 0.0001) for pre grazing rhizoma peanut sward height. Early, mid and late season dates correspond to June/e arly July, late July/August, and September/October, respectively, in 2012. Data are means across four rhizoma peanut entries, two grazing intensities, and two replicates (n = 16).
112 Figure 4 5. Date x grazing frequency interaction ( P < 0.0001) for ground cover of rhizoma peanut entries. Data are means across four rhizoma peanut entries, two grazing intensities, and two replicates (n = 16).
113 Figure 4 6. Entry x grazing frequency int eraction ( P = 0.0005) for ground cover of rhizoma peanut entries. Data are means across two grazing intensities, three dates, and two replicates (n = 12).
114 Figure 4 7. Date x entry interaction ( P = 0.0110) for frequen cy of weed occurrence in rhizoma peanut under different grazing regimes. Data are means across three dates, two grazing intensities, and two replicates (n = 12).
115 Figure 4 8. Date x grazing frequency interaction ( P < 0.0001) for frequency of weed occurrence in rhizoma peanut swards. Data are means across four entries, two grazing intensities, and two replicates (n = 16).
116 CHAPTER 5 DEFOLIATION MANAGEMENT AND SOIL CARBON DYNAMICS OF VARIOUS PRODUCTION SYSTEMS BASED ON WARM SEASON PERENNIAL FORAGES Overview of Research In addition to their role as a source of feed for livestock, grasslands are increasingly being evaluated in terms of the ecosystem services they provide. An example of a particularly critical ecosystem f unction is the major role grasslands play in the global C cycle. They serve as an important C sink, with approximately 22% of global soil C stores existing under grassland (Soussana et al., 2004), and it is estimated that as much as 90% of C stored in gras sland ecosystems is below ground (Schuman et al., 2001). Soil C can accumulate when agricultural land is removed from annual row crop cultivation and planted with perennial grasses (Post and Kwon, 2000; Conant et al., 2001), but it has been suggested that there is also an important role of management in affecting the ability of a grassland to increase soil C sequestration (Silveira et al., 2013). In a meta analysis of grazing effects on soil C sequestration, McSherry and Ritchie (2013) reported that both v egetation type and grazing intensity strongly influence soil organic C storage potential of grasslands. In support of the conclusions of McSherry and Ritchie (2013), Franzluebbers and Stuedemann (2009) found that management of grass based forage systems, e .g., N fertilization, haying vs. grazing, and grazing at a range of stocking rates, affects their potential to store soil C. Thus, a key consideration for future research is whether forage systems and their management practices can be designed to maintain forage production while also increasing contribution to soil C and ecosystem sustainability.
117 There are significant gaps in the literature relative to the effect of grassland production systems on soil C. Specifically, little attention has been given to pe rennial legume based pasture systems. In addition, the effects of defoliation management, vegetation type, and their interaction on soil C have been less explored in sandy soil conditions like those of Florida relative to fine textured soils. This informat ion is needed to understand the potential for C sequestration in large areas of the USA Gulf Coast region. Thus, the objectives of this study were to quantify the effects of defoliation management and winter overseeding of forage systems based on rhizoma peanut ( Arachis glabrata Benth.; RP) or N Cynodon spp.) on herbage production and short term soil organic C and N responses after land conversion from row cropping. This experiment is expected to identify management tec hniques which promote sustainable grassland production systems and begin to quantify the capacity of USA Gulf Coast grassland systems as a C sink. Also, the research project will provide producer recommendations for winter overseeding in, and grazing manag ement practices for, RP and bermudagrass systems. Materials and Methods Experimental Site The experiment was planted in 2011 and treatments were imposed starting in fall 2011 and continuing through fall 2013 at the University of Florida Plant Science Resea rch and Education Unit in Citra, FL ( 29.24N, 82.10W ). For the purposes of this chapter, only data from the cool and warm seasons of 2011 2012 and the cool season of 2012 2013 will be reported.
118 The site was chosen because it was previously under long ter m (> 5 yr) crop Paspalum notatum Flgge) in March 2010. The area was managed extensively during the 2010 growing season, and because rainfall was limited during the summer at this location, the bahia grass stand did not establish well and much of the area was occupied by weeds or bare ground by the end of the 2010 growing season. On 1 Feb. 2011, a 0.6 ha portion of this area was plowed and heavily disked to ensure a well prepared seedbed for establishi ng the perennial forages upon which the systems were based. Soils at the site consist of a Placid fine sand (sandy, siliceous, hyperthermic Typic Humaquepts) and Tavares sand (hyperthermic, uncoated Typic Quartzipsamments). Prior to planting, soil samples were taken to a depth of 15 cm within each plot and analyzed by the University of Florida Extension Soil Testing Laboratory. Initial soil organic C to 15 cm for the experimental area at large was 9.4 g kg 1 Soil pH was 5.7 and Mehlich 1 extractable P, K, Mg, and Ca were 26, 17, 18, and 295 mg kg 1 respectively. Just prior to land preparation on 1 Feb. 2011, 910 kg ha 1 of dolomitic lime was applied. Treatments and Experimental Design Treatments were replicated three times in a split plot arrangement, with main plots allocated in a randomized complete block design. Perennial species was the main Within each main plot there were five production systems for a total of 10 treatm ents. The five production sy stems (Table 5 1) consisted of i ) summer hay produ ction of the warm season forage and no overseed ed cool season forage (SH No); ii ) hay production of the warm Seca le
119 cereal e L.) during winter (SH WH); iii ) grazing of the w arm season forage during summer and no overseed ed cool season forage (SG No); iv ) grazing of the warm season forage during summer and of overseeded rye during winter (SG WG); and v ) hay production of the warm season forage during summer followed by grazing of overseeded rye during winter (SH WG). The SH WG treatment was included because there are many RP and Tifton 85 hay producers in the region, and overseeding hay fields for winter grazing is a fe asible practice. It is important to determine if soil organic C (SOC) accumulation will be increased by overseeding rye and grazing during winter compared with allowing the hay field to remain dormant during winter. Overseeded plots were defoliated during winter according to practices associated with each production system. Winter grazing occurred when rye reached a 30 cm height and every 4 to 5 wk thereafter through the early spring. Grazing treatments were imposed beginning on 17 Jan. 2012 and 23 Jan. 20 1 3. Postgrazing target stubble height was 10 cm. In the hay treatment, rye was harvested when it reached 30 cm of height and every 4 wk until the boot stage was reached in the spring. The final harvest was made at this time. Clipping was initiated on 24 Jan 2012 and 23 Jan. 2013. For both grazed and clipped treatments, there were three defoliation events in 2012 and two in 2013. During the summer of 2012, hayed RP and BG areas were harvested every 4 wk to a cutting height of 10 cm. Experimental units under hay production were cut using a disk mower and raked to remove harvested material from the plot area. Grazed areas were also defoliated on a 4 wk frequency, but the target postgraze stubble height was 15 cm for both species. Animals grazed individual expe rimental units separately, and
120 grazing activity was monitored until the target stubble height was reached or when animals ceased grazing. Because animals were not fasted prior to grazing and were not accustomed to grazing small pastures, multiple grazing e vents over several days were sometimes needed to approach the target postgraze height, but it was not always was initiated on 23 May 2012 for BG plots and occurred mon thly until October. Clipping of BG hay treatments began on 25 May 2012 and occurred monthly until the end of the season in October. Rhizoma peanut plots were defoliated beginning in August 2012 when they first achieved a 20 cm stubble height. Grazing and c lipping treatments were imposed beginning on 7 Aug. and 14 Aug. 2012, and monthly thereafter until October. Plot Establishment and Management Warm season perennial species main plots were planted individually and each occupied an area of 35 x 13 m. Areas planted to RP were established using Florigraze rhizomes at a planting rate of 1000 kg ha 1 and a commercial planter (4 row planter with 50 cm between rows) on 5 Apr. 2011. On 2 June 2011, RP plots were fertilized to supply 15 kg P and 55 kg K ha 1 accordi ng to soil test recommendations. Tifton 85 BG was planted on 9 June 2011 using above ground stems. Stems were harvested from an existing foundation block of Tifton 85 at the research unit, transported to the experimental site, and spread evenly across the area by hand. Sprigs were then disked into the soil and the soil was firmed with a cultipacker. Bermudagrass main plots were fertilized with 45 kg N ha 1 using ammonium sulfate on 5 July 2011 and 45 kg N ha 1 using ammonium nitrate on 8 Aug. and 1 Sept. 20 11 to enhance grass establishment and spread. Main plots were divided into five 7 x 13 m subplots to accommodate the
121 five production system treatments. Arrangement of areas planted to the respective species is described below and in Figure 5 1. Overseede d treatments were planted with Florida 401 rye at a rate of 110 kg ha 1 on 8 Nov. 2011 and 13 Nov. 2012. Plots that were not overseeded to rye were not defoliated during winter and did not receive any fertilization. For all areas overseeded with rye, 30 kg N ha 1 was applied on 30 Nov. 2011 and 28 Nov. 2012 after seedlings had emerged. In winter 2012, grazed and hayed areas also received 30 kg N ha 1 on 24 January and 21 February following defoliation events. Phosphorus was applied at a rate of 15 kg ha 1 an d K was applied at 55 kg ha 1 on 21 Feb. 2012. During the winter 2013 season, 30 kg N ha 1 was applied on 23 January, but no additional N fertilizer was applied due to relative poor establishment and vigor of rye in that year. No P and K were applied in wi nter 2013. For BG plots, 45 kg N ha 1 as ammonium nitrate was applied on 3 Apr. 2012 to promote spring growth. Defoliation of grazed and hayed treatments was initiated on 23 and 25 May 2012, respectively, and the final grazing and clipping events of the se ason occurred on 8 and 9 Oct. 2012. Nitrogen fertilizer was applied monthly to all BG plots following a defoliation event at a rate of 45 kg N ha 1 for a warm season total of 225 kg N ha 1 yr 1 Rhizoma peanut and BG plots were fertilized with muriate of p otash at a rate of 55 and 75 kg K ha 1 respectively, on 4 June 2012 according to soil test recommendations. Herbicide applications occurred throughout the establishment period and during the experiment to control encroaching broadleaf weeds and grasses. Rhizoma peanut main plots were sprayed with ammonium salt of imazapic (0.07 kg a.i. ha 1 ; Impose TM
122 MANA, Raleigh, NC) on 23 May 2011, clethodim (SelectMax Valent, Walnut Creek, CA) on 6 July 2011, and imazapic (0.07 kg a.i. ha 1 ) + 2,4 D (0.26 kg a.i. h a 1 ; dimethylamine salt of 2,4 D dichlorophenoxyacetic acid) on 2 Apr. 2012. Aminopyralid and 2,4 D (GrazonNext TM ; rate 1.10 kg a.i. ha 1 ; Dow AgroSciences, Indianapolis, IN) were applied on 15 Sept 2011 and Chapparal TM [rate 0.12 kg a.i. ha 1 ; aminopyrali d and metsulfuron methyl; Dow AgroSciences) on 2 Apr. 2012 to BG plots. Irrigation was applied during establishment and occasionally throughout the experimental period to supplement rainfall when it was less than the 30 yr monthly average. During establis hment in 2011, total irrigation applied was 50, 38, 50, 20, and 25 mm during April, May, June, July, and November, respectively. During 2012, 25 mm was applied each month in April, May, and September. Response Variables Herbage Mass, Accumulation, and Har vested Herbage mass from both summer and winter forage components was measured prior to each grazing or haying event. In the hayed treatments, a 2 1 m area was harvested using a sickle bar mower to a 10 cm stubble height, weighed fresh, subsampled for D M determination, and dried. In the grazed treatments, pre grazing and post grazing herbage mass was determined for all grazing events. The double sampling technique used settling height of an aluminum disk as the indirect measure and hand clipping to 5 cm stubble as the direct measure for pre and post grazing samples during the winter and a 10 cm stubble in the summer. The disk was calibrated at each grazing event by taking both indirect and direct measures from two 0.25 m 2 quadrats per plot. Twenty indire ct measures were taken pre and post grazing at each grazing event using a stratified site selection approach (fixed distance between disk drops so that the
123 pasture was well represented). The average disk height of the 20 measures was inserted into a calib ration equation to predict herbage mass, and the resultant values were used to calculate herbage accumulation and harvested. Grass and legume herbage accumulation was measured as the difference between herbage mass after a defoliation event and prior to th e next defoliation event. Herbage harvested was calculated as the difference between pre grazing and post grazing herbage mass of the same grazing cycle. Herbage Nutritive Value For each grazed pastures, 10 hand plucked samples were taken to a 10 cm stubb le height prior to a grazing event during the winter and to a 15 cm height during the summer management season. Locations were selected in a grid pattern to represent the entire plot, and individual samples were harvested using hand shears to the target st ubble. The 10 samples were composited for subsequent grinding and laboratory analysis. For the haying treatments, a subsample was taken from the DM sample (clipped to 10 cm stubble) for determination of nutritive value. Samples constant weight. Nutritive value analyses included crude protein (CP) and in vitro digestible organic matter (IVDOM) concentrations. Crude protein was measured using a micro Kjeldahl technique for N (Gallaher et al., 1975) and th e two stage technique for IVDOM (Moore and Mott, 1974). In winter 2012, samples from each defoliation event were analyzed for CP and IVDOM. For the summer systems, samples from each species x system treatment were selected from the August and October harve st dates and analyzed. Data from each defoliation event for winter 2013 is reported for CP.
124 Residual Litter Mass, C and N Concentration, and C:N Ratio Litter in this study was defined as the post harvest stubble plus dead plant biomass on the soil surface Litter mass sampling occurred at the end of the 2012 and 2013 winter grazing/haying period and at the end of the 2012 summer grazing/hay season prior to overseeding rye. The purpose of this measurement was to quantify the amount of plant material remaini ng above ground that could potentially contribute to soil C. Four 0.25 m 2 quadrats were sampled per experimental unit. Sites for sampling were chosen to reflect average stubble mass Plant stubb le was harvested to soil level and collected with dead plant b iomass within a quadrat to provide a single sample. Plant litter was bagged and dried at 60C until dry. Litter C and N concentratio n were determined from winter 2012 and summer 2013 samples. Total organic C and total N were determined by dry combustion u sing a Flash EA 1112 C/N analyzer The amou nt of C and N in litter was calculated as the litter mass (kg DM ha 1 ) multiplied by the percentage of C and N, respectively. Finally, litter and stubble C:N ratio were calculated for each treatment. The C:N ratio was determined as the g kg 1 of C divided by the g kg 1 of N. Root rhizome Mass Root rhizome mass was quantified immediately after the last grazing event in October 2012 (following one winter and su mmer of management ). Five samples were taken per experime ntal unit using a 10 cm diameter soil coring device to a depth of 20 cm. Root cores were washed and dried to constant weight at 60C. Soil C and N Soil bulk density samples were collected from each experimental unit prior to treatments being imposed in 2 011 and for surface l ayers (0 to 10 and 10 to 20 cm ) at
125 the end of 1 yr of imposing treatments in fall 2012. In 2011, two undisturbed soil cores were collected to a depth of 100 cm in each experimental unit for bulk density determination. Three additional cores were augured from each experimental unit to provide sufficient sample for C and N analysis. The soil cores were taken to a depth of 100 cm and divided into layers of 0 to 10, 10 to 20, 20 to 40, 40 to 70, and 70 to 100 cm. Samples from these layers w ere dried at 105 C determined using soil from the undisturbed cores as the total dry weight of the soil divided by the volume of the coring device. For the initial soil samples in 2011, total organic C and total N of each so il layer were determined by dry combustion using a Carlo Erba NA 1500 C/N/S analyzer on subsamples ground in a ball mill for 5 min. Final sampling for bulk density and total C and N will occur at the end of summer 2013 following 2 yr of each management sys tem. Aggregate Size Distribution Distribution of soil size classes was determined for samples from experimental units under year round grazing or haying (Treatments SG WG and SH WH). A 50 g sample was taken before planting and 1 yr after initiation of tre atments (October 2012) from the 0 to 10 cm and 10 to 20 cm soil layers. Samples were sieved through a 2 mm screen and particles greater than 2 mm were discarded. Aggregate separation was done by wet sieving through two sieves (250 and 53 m) according to the procedure of size microaggregates, and silt plus clay size particles, respectively (C ambardella and Elliot, 1994; Six et al., 2004). After sieving, samples were dried at 60 C, weighed and analyzed for C and N using a Flash EA 1112 C / N analyzer
126 Statistical Analysis Data were analyzed using PROC MIXED of SAS (SAS Insti tute, Cary, NC, 1996) Data are reported for the winter seasons of 2012 and 2013 and the summer season of 2 012 only Species, management system, year, and their interactions were considered fixed effects. B lock and block species interaction were considered random effects. Sp ecies was the main plot and forage system the sub plot in the random ized complete block design. Means were compared usi ng the PDIFF option of LSMEANS. Differences were declared when P P Interactions were described when P Where multiple observations of a response occur across intervals of time, observation date was considered a repeated measure. Because of the presence of unequal variances, herbage harvested data were log transformed. Non tr ansformed means are reported, and transformed values were use d for statistical comparisons. Results and Discussion Plant Responses Herbage h arvested Across the 2 yr observation period for winter rye production, there were species x system (Table 5 2; P = 0.0013) and year x system (Table 5 3; P < 0.0001) interactions for herbage harvested When BG was overseeded with rye, SH WH had greater herbage harvested (1080 kg ha 1 ) than SG WG (480 kg ha 1 ) and SH WG (580 kg ha 1 ) treatments. Overseeded RP followed a similar pattern, and rye harvested for hay had greater total herbage harvested at the end of the season (1350 kg ha 1 ) than grazed treatments (270 and 248 kg ha 1 for SG WG and SH WG, respectively). Rye production decreased in 2013 compared with 2012. Diff erences among forage management
127 systems were apparent in Year 1, and hay treatments produced more herbage harvested (2060 kg ha 1 ) than those under rotational stocking (500 and 360 for SG WG and SH WG, respectively). No differences were observed among syst ems in Year 2, and mean rye harvested was 390 kg ha 1 Decreased rye production in Year 2 was primarily associated with weak stand establishment which reduced production and the number of defoliation events compared with Year 1. Moyer and Coffey et al. (20 00) suggested that early small grain growth could be suppressed by competition or inhibitory effects of bermudagrass sod when interseeded. The practice of interseeding also reduces production potential of small grains to a greater degree than when grown in monoculture in a prepared seedbed (Utley et al., 1976). Although rye herbage harvested in the present study is lower than reported yields of Florida 401 in the literature (Day et al., 2012), this experiment illustrates the potential for utilizing overseed ing of dormant warm season forage systems as a means to provide winter forage. A species x system interaction (Table 5 4; P < 0.0001) occurred during summer 2012. Bermudagrass herbage harvested was greater when harvested for hay every 28 d than when graz ed. Decreased herbage harvested for grazed BG in the present study is primarily a function of the inability to achieve the target postgraze stubble height during a grazing event. However, the reported values for herbage harvested for hay treatments are wit hin the range reported in the literature. Forage production potential of Tifton 85 bermudagrass varies with harvest frequency and intensity (Pedreira et al., 1999). Mislevy and Martin (1998) reported mean total season production of 13 700 kg ha 1 for Tifto n 85 under a 2 4 5 or 7 wk grazing frequency. In a 2 yr evaluation of the
128 effect of harvest frequency (21, 24, 27, or 35 d) and post harvest stubble height (8 or 16 cm) on Tifton 85 herbage production, Clavijo et al. (2010) observed herbage harvested of 7900 kg ha 1 for BG clipped every 27 d. First harvest of RP did not occur until August 2012, and was during the year after establishment, which explains reduced herbage production potential compared with other RP studies (Butler et al., 2007; Mislevy e t al., 2007). Differences were observed among RP based forage systems, and clipped RP had less total herbage harvested (mean 270 kg ha 1 ) than grazed treatments (mean 930 kg ha 1 ). Based on these data, there was no apparent reduction in herbage harvested in the summer for winter overseeded treatments compared with no overseeding. When Florida 401 was released, Pfahler et al. (1986) noted 50% greater herbage production during the early winter season compared with other varieties at the time of release, and similar production levels by mid season. Because of the early season production potential of this cultivar, this may have reduced competition with spring growth of warm season perennials and avoided a negative impact on herbage production. Nutritive v alu e In winter 2012, there was an effect of date ( P < 0.0001) and date x system ( P = 0.0172) on rye nutritive value. Crude protein decreased for rye across sampling dates (240 to 165 g kg 1 DM from January to March, respectively). The interaction occurred (Fi gure 5 2) because CP was not different among treatments in January (mean 235 g kg 1 ), but CP was greater for SG WG than SH WH in February (222 vs. 187 g kg 1 respectively), and was greater for both grazed treatments than the hay treatment in March (190 an d 185 for SH WG and SG WG, respectively, vs. 123 g kg 1 for SH WH).
129 There was a strong trend for differences in CP between rye overseeded into RP or BG ( P = 0.0559). For rye overseeded into BG, CP was 197 g CP kg 1 DM compared with 210 g CP kg 1 DM for ry e seeded into RP. This response is likely associated with greater uptake of soil N in the legume based system The nutritive value of rye varies with management, cultivar selected, and the production environment. Average CP of rye generally ranges from 120 to 230 g kg 1 DM across the season (NRC, 1996; Muir and Bow, 2009; Newell and Butler, 2012). There was a date x system interaction ( P < 0.0001) for rye IVDOM in winter 2012. At the beginning of the winter, there was greater IVDOM for grazed treatments th an clipped (7 75 vs. 673 g kg 1 ) prior to treatments being imposed, but by March 2012 IVDOM was not different among treatments (mean 680 g kg 1 ). Myer et al. (2008) reported mean IVDOM of 792 g kg 1 oat when sod seeded into dormant bahiagrass ( Paspalum notatum Flgge) or planted in clean tilled soils during a 2 yr evaluation in north Florida. These data reflect a similar level of digestibility for rye and small grain mixtures in a similar production environme nt. During summer 2012, a date x species interaction ( P = 0.0317; Figure 5 3) was observed for CP in BG and RP. Crude protein concentration increased from May to June (130 to 154 g kg 1 ) for BG, and remained at this level until August. Beginning in Septe mber, CP declined to 146 g kg 1 and decreased further to 104 g kg 1 by October. Rhizoma peanut CP was lower than BG from August to October, with the highest CP observed in August (139 g kg 1 ) and decreasing to 120 g kg 1 by September. Beltranena et al. (19 81) observed decreasing leaf percentage of RP across the season when harvested for hay production every 56 d, which may negatively impact nutritive value.
130 Also, the presence of weeds in the RP stand may have decreased CP values relative to BG. There were species x system ( P = 0.0014) and date x species interactions ( P = 0.0022 ) for IVDOM of summer forage production systems. A species x system effect (Figure 5 4) was observed because grazed RP had greater digestibility than hayed treatments whereas there wa s no effect of production system on BG. The date x species interaction occurred because in August RP had greater IVDOM than BG (698 vs. 552 g kg 1 ), however, IVDOM decreased by October for both species. Although digestibility decreased across the season, I VDOM of RP remained greater than BG (620 vs. 430 g kg 1 ). Eckert et al. (2010) observed slightly greater values with an IVDOM of 670 and 520 g kg 1 for the first cutting of RP hay compared with BG harvested every 5 wk, respectively. Saldivar et al. (1990) attributed decreasing RP digestibility across the season to a decreasing percentage of leaf and during early fall. Holt and Conrad (1985) observed a decrease in IVDOM of BG harvested for hay production with advancing season. These results also agree with J ohnson et al. (2001) who observed a quadratic effect of harvest date on IVDOM of bermudagrass where IVDOM was highest at first harvest in early June, but a reduction occurred by mid summer (July). Although IVDOM increased again for BG in August, IVDOM plat eaued throughout the remainder of the production season, illustrating a decrease in IVDOM relative to the first harvest date. Residual litter m ass There were year x species ( P = 0.0002), year x production system ( P = 0.0022), and species x production syst em ( P = 0.0068) interactions for residual litter mass at the end of the winter. The year x species interaction (Figure 5 5) occurred because overseeded BG and RP plots had greater residual herbage mass in 2013 than 2012
131 (Figure 5 5). At the end of the wint er of 2012, residual litter mass was similar between species (1900 vs. 1 600 kg ha 1 for BG and RP, respectively), whereas overseeded BG plots had greater residual litter mass (3 850 kg ha 1 ) compared with RP (2 400 kg ha 1 ) at the end of winter 2013. More re sidual litter in 2013 was associated with the accumulation of litter following more than 1 full year of imposing treatments compared with only a single winter season in 2012. Greater herbage harvested of BG than RP during the summer likely contributed to g reater litter mass at the end of winter season for BG plots. The year x production system interaction occurred because litter mass was greater for clipp ed, overseeded plots (2 150 kg ha 1 ) during winter 2012 compared with those that were grazed (1480 and 1 730 kg ha 1 for SG WG and SH WG systems, respectively). At the end of winter 2013, S G WG had greater litter mass (3 490 kg ha 1 ) compared with those t hat were clipped (2 800 kg ha 1 for SH WH), but SH WG was not dif ferent from either treatment (3 040 kg ha 1 ) A species x production system interaction (Figure 5 6) occurred because overseeded BG had greater residual litter for SG WG treatments (3 220 kg ha 1 ) than SH WH and SH WG rye (2870 and 2 550 kg ha 1 respectively). No differences were observed among syst ems for overseeded RP, and RP production systems had an av erage residual litter mass of 2 100 kg ha 1 Franzluebbers et al. (2000) reported that the practice of hay production reduces the amount of decomposable substrates added to the soil through the remov al of above ground plant material. Thus, the contribution of residual plant litter is often less than in grazed ecosystems.
132 Residual l it ter C and N concentration and C:N r atio Total C concentration in litter mass differed among years ( P < 0.0001), but ther e was no effect of warm season species or production system on C concentration in litter at the end of the winter season. Average total C concentration in litter mass was 364 and 421 g kg 1 DM in 2012 and 2013, respectively. Increased C concentration in 20 13 may have been to greater carryover of lower quality, more recalcitrant litter from the summer growing season in the second year vs. the first. Litter N concentration at the end of the winter seasons was affected by warm season perennial species ( P = 0. 0127) and was greater for RP than BG (27 vs. 24 g kg 1 DM). This coincides with the increased CP of rye overseeded into RP compared with BG. There also was a year x production system interaction ( P = 0.0467), and it occurred because litter from treatments that included overseeded rye under hay production had slightly lesser N concentration (27 g kg 1 DM) than grazed treatments (28 and 29 g kg 1 DM for SG WG and SH WG, respectively) in 2012, but concentrations were similar in 2013 (mean 22 g kg 1 DM for all treatments). Ruffo and Bollero (2003) evaluated rye and hairy vetch ( Vicia villosa Roth.) cover crops as a source of N for corn ( Zea mays L.). They observed a peak value of 1500 kg C ha 1 and 50 kg N ha 1 in rye residue prior to chemical burndown and plant ing of corn. These values are slightly greater than the C and N associated with residual litter mass in the present study because the rye was not defoliated in the cover crop experiment. The C:N ratio of plant litter from the winter season was affected by species x system ( P = 0.0382) and year x species ( P = 0.0228). A species x system interaction (Figure 5 7) occurred primarily because litter from winter haying of rye in BG plots resulted in a lower C:N ratio (15.7) than winter grazing treatments (17.0 for both SG WG
133 and SH WG). However, the C:N ratio of litter at the end of winter from overseeded RP was not different among grazed vs. hayed plots (14.5). There was an increase in C:N ratio from 2012 to 2013 for overseeded RP and BG. In 2012, residual litter C:N was 12.0 and 13.2 for overseeded RP and BG, respectively. By 2013, the ratio increased to 16.9 for RP and 19.7 for BG. These data are likely a function of the increased accumulation of lower quality litter mass arising from the summer of 2012. A year x production system interaction ( P = 0.0669) occurred because there was an increase in C:N ratio across years for all treatments. Fo r the SH WH treatment, C:N ratio increased from 12.7 in 2012 to 18.0 in 2013 ( P < 0.001) compared with 12.2 to 19.0, respecti vely, for SG WG treatment ( P < 0.0001) Additionally, C:N ratio increased from 13.0 to 18.1 for the SH WG production system ( P < 0.001) during this time. This suggests that litter quality decreased with increasing amount of litter following 1.5 yr of impos ing treatments. In plant material, mineralization is favored when C:N ratio is 20:1 or less, and immobilization occurs at a C:N ratio of 30:1 or greater (Dubeux et al., 2007 ; Lambers et al., 2008). However, although C:N ratio has been broadly used as an in dicator of OM susceptibility to decomposition, a number of biotic and abiotic factors can affect mineralization and immobilization processes outside of this ratio. Litter C:N increased at the end of summer 2012 for BG, but decreased in winter 2013 when ove rseeded with rye. This is likely because of the greater quality of rye than bermudagrass residue that was being returned to the system at that time. At the end of the summer in 2012, there were differences in residual litter mass among species ( P < 0.0005) Bermudagrass had greater residual litter mass (3990 kg DM ha 1 ) than RP (2190 kg DM ha 1 ). Greater productivity of BG contributed to greater
134 litter mass at the end of the season com pared with RP. There was a lso a species x system interaction ( P = 0.0613) Grazed bermudagrass plots had greater residual litter (3690 vs. 4050 kg DM ha 1 for SG WH and SG WG, respectively) than hayed (mean of 2650 kg DM ha 1 for SH WH and SH WG), but there were no differences in litter mass among RP based systems (mean 2 200 kg DM ha 1 ). Difficulty in achieving the target stubble height for grazed bermudagrass plots likely increased contribution to litter mass. Greater herbage mass under less intensive management strategies (i.e., low stocking rate) may cause more litter deposi tion because of a lower forage utilization rate and an increase in senescent herbage (T homas, 1992; Dubeux et al., 2007 ). During Year 1 of a 2 yr evaluation, Valria et al. (2013) observed less litter mass accumulation for signalgrass ( Brachiaria decumbens Stapf.) under a high stocking rate [3.9 or 5.8 AU (450 kg cattle live weight) ha 1 ; 2,090 and 2,210 kg OM ha 1 respectively ] than the lowest stocking rate of 2.0 AU ha 1 (2750 kg OM ha 1 ). However, litter mass in Year 2 was similar among treatments (3,13 0 kg OM ha 1 ). The authors attributed the lack of differences in the second year to the cumulative effect of N fertilization increasing net primary productivity and equilibrating grazing pressure at greater stocking rates. Liu et al. (2011b) evaluated the effect of postgraze stubble height on plant and soil pastures. As stubble height increased from 8 to 24 cm, there was a linear increase in plant litter (did not include residual stubble in the Liu experiment) dur ing the summer from 2,040 to 3,580 kg ha 1 These data also correspond with the herbage harvested, with lower productivity of RP during the first year of management likely explaining lesser residual litter mass deposition.
135 Concentration of C and N in plant litter at the end of the summer season of 2012 differed ( P < 0.0001 for both C and N, respectively) among species. Litter in RP systems had 358 g C kg 1 DM at the end of 1 yr of imposing treatments while BG litter had an average of 415 g C kg 1 The tota l N concentration was 15.3 and 18.6 g kg 1 DM, respectively, for RP and BG (species effect; P = 0.0027). Finally, the C:N ratio differed among RP and BG (species effect; P < 0.0001), and BG had a greater C:N ratio (27.5) than RP (19.5). A decreased C:N rat io for RP is typical for legumes relative to C 4 grasses (Thomas and Asakawa, 1993; Knops and Tillman, 2000). Franzluebbers et al. (2004) suggested that surface residues with a lower C:N ratio have the potential to be more rapidly mineralized and contribute to soil fertility in the surface soil (0 to 10 cm). T hus, RP litter likely had greater potential for mineralization than BG Root rhizome m ass There was no effect of production system on root rhizome mass when sampled at the end of 1 yr of imposing treat ments, however, there were differences among warm season species ( P = 0.0307). Bermudagrass had a greater root rhizome mass (5,910 kg ha 1 ) compared with RP (2,130 kg ha 1 ). The observed values for BG and RP are lower than reported in the literature when measured at simila r depths (Orte ga S. et al., 1992b; Rice et al., 1995; Liu et al., 2011a), but may reflect the relatively short establishment period and subsequent use of these warm season perennials during the year after planting. Soil Responses Aggregate size d istribut ion The aggregate size distribution of soil from the 0 to 10 and 10 to 20 cm depths did not differ among treatments. M a 1
136 of total soil. For the microaggregate fraction (53 250 m), there was a trend for differences among management systems ( P = 0.0635) in the 0 to 10 cm soil layer whereby SH WH had a greater contribution from this soil size class than SG WG (471 vs. 440 g kg 1 soil, respectively). The microaggregate fraction consisted of 464 g kg 1 of total soil in the 10 to 20 cm stratum and was not affected by treatment. Dubeux et al. (2006) evaluated the effect of four levels of grazing management (combinations of stocking rate and N fertilization) on soil fractions under bahiagrass The soil particle size distribution did not differ due to forage management regime, but the macro and microaggreg ate fractions (53 2 000 m) represented 990 g kg 1 of total soil. In th at Spodosol ( sandy siliceous, hyperthermic Ultic Alaquods) t he dominant fraction was coarse sand (250 2000 m) In the present study, the macro plus microaggregate fraction of the soil represented 960 g kg 1 of total soil. There were differences in particle size distribution of the < 53 m fraction across years ( P = 0.0047) for the 0 to 10 cm layer. At the beginning of the experiment in 2011, this fraction represented 22 g kg 1 total soil. However, it decreased to 17 g kg 1 soil by the end of the first full year of imposing treatments. These results suggest that aggregation occurred after one year of management, and is an important factor defining soil quality. In the 10 to 20 cm layer, the re was year x production system interaction ( P = 0.0522) because the soil particle size distribution decreased for clipped treatments in 2012 (24 to 15 g kg 1 ), but was maintained for grazed swards (21 to 18 g kg 1 ). Particle s ize fraction C, N, and C:N r atio Total C concentration differed among particle size classes (Table 5 5) There was a year x production system interaction ( P = 0.0057 ) observed for the > 250 m particle size class in the 0 to 10 cm layer. Prior to imposing treatments, SO C
137 concentrations in the bulk soil did not differ among management strategies and was 5.20 and 4.91 g kg 1 soil fraction for SH WH and SG WG treatments respectively. However, after 1 yr of imposing treatments, there was a decrease in SO C for the SH WH system (4.38 g kg 1 soil fraction ), while total C increased under year round grazing (SG WG; 6.98 g kg 1 ). Six et al. (2002) suggested that newly incorpor ated organic material is often associated with larger particle size fractions. Furthermore, under the warm and humid conditions at the experimental site, rapid decomposition of organic C inputs can occur, which contribute to soil C pools (Cambardella and E lliott, 1992). Greater C in the large particle fraction can be attributed in part to greater residual litter inputs associated with grazed vs. hayed systems. For the 53 to 250 m fraction, no differences were observed between warm season species or among p roduction system treatments within the 0 to 10 or 10 to 20 cm layers (5.64 and 5.65 g kg 1 respectively). In the surface 10 cm of soil, there were no differences in C contribution from the < 53 m size class following 1 yr of imposing treatments. Mean total C concentration associated with this fraction was 59.6 g kg 1 soil fraction for the 0 to 10 cm layer. This illustrates that although this particle size class makes up a smaller proportion of the total soil by weight, it serves an important source of C in this soil type. A year effect ( P = 0.0035) was observed for the C associated with the < 53 m size class for the 10 to 20 cm soil layer. Total C associated with this fraction decreased from prior to initiation of the experiment to 1 yr after initiat ion of treatments (69.2 vs. 53.7 g kg 1 soil fraction, respective ly). There was also a year x system effect ( P = 0.0964) for this soil depth. Total C of < 53 m particles was similar among year round hay production and grazing
138 treatments prior to treatment s being imposed (67.8 vs. 70.5 g kg 1 soil fraction, respectively), however, at the end of the first year, grazed systems had less total C concentration (47.4 g kg 1 ) compared with those harvested for hay (60.1 g kg 1 ). In less sandy soils, C associated w ith small particle size classes is considered to be relatively stable with a slow turnover rate (Six et al., 2000). However, Silveira et al. (2013) suggested that due to the low silt plus clay content in many Florida soils, the ability for chemical and phy sical protection of soil OM is limited, resulting in the accumulation of SO C in the fine particle size fraction that may more readily undergo degradation. Tisdall and Oades (1992) stated that the products of microbial decomposition (i.e., polysaccharides) t hat bind C in fine particle size fractions ( < 53 m class ) can be affected by land management strategy. Thus, physical disruption of the soil from tillage prior to planting may have caused an initial decrease in soil C in this experiment. D ecreased C under grazing may also be related to disturbance of the soil from cattle, particularly in the RP plots which had not achieved full ground cover during the first year of this study. Differences in total N concentration were found among soil particle size classe s (Table 5 6) P = 0.0503) within the 0 to 10 cm soil depth. At the beginning of the experiment, total N was similar among SH WH and SG WG systems (0.23 and 0.25 g kg 1 soil fraction, respectiv ely). After 1 yr of imposing treatments, total N concentration increased under year round grazing (0.42 g kg 1 ) compared with hay production (0.18 g kg 1 ). Increased soil N for grazed systems is likely a result of N return in livestock excreta and increase d litter deposition. Dubeux et al. (2006) suggested that the high C:N ratio of decaying plant and
139 root material may immobilize soil N during the decay process, and subsequently increase N concentration in the soil organic matter. No differences were observ ed for the 10 to 20 1 Total N concentration of soil particles from the 53 to 250 m size class did not differ among production systems at either soil depth. Average total N was 0.30 g kg 1 of this s ize class. The largest amount of N was associated with the < 53 m particle size fraction. Concentration of N in this fraction differed between years ( P = 0.0228) in the surface 10 cm of soil, and decreased from before treatment imposition to 1 yr later (4 .91 vs. 3.77 g kg 1 soil fraction, respectively). In the 10 to 20 cm soil layer, total N for the < 53 m size class averaged 4.21 g kg 1 soil fraction. These results agree with other studies conducted in Florida showing that the < 53 m fraction serves as an important source of C and N in the soil, although the concentration of this particle size fraction is relatively low in total soil (Dubeux et al., 2006; Silveira et al., 2013). The m acroaggregate fraction ct; P = 0.0380) for SH WH (28.9) than SG WG (19.8) in the top 10 cm of soil. A species x system interaction ( P = 0.0040) occurred for the 10 to 20 cm layer. Grazed BG had a lower C:N ratio (24.0) than SH WH (34.2), but grazed RP had a greater soil fractio n C:N ratio (34.9) than clipped treatments (20.5). There was a species x production system interaction ( P = 0.0635) for the 53 to 250 m fraction to a depth of 10 cm, although there was no difference in C:N ratio from 10 to 20 cm (mean of 25.2). The inter action occurred because grazed RP had a greater C:N ratio than clipped (29.1 vs. 18.0, respectively), but C:N was similar among BG systems (21.0 and 24.2 for SH WH and SG WG, respectively). In the < 53 m size class, C:N ratio did not differ among
140 treatmen ts within the 0 to 10 cm layer (14.3 and 14.5 for BG and RP, respectively), but there was a system effect ( P = 0.0150) for the 10 to 20 cm layer. Grazing increased the fraction soil C:N ratio (14.4) compared with those under year round hay production (13 .9), although the difference was fairly small. The C:N ratio of this size class is lower than the coarse and fine sand classes. This observation supports the decrease in C across time and increased contribution of N for this size fraction relative to the o thers. A smaller C:N ratio indicates that this fraction is potentially labile and may undergo degradation more readily when not in a physically protected form. Implications of the Research Year round forage production system affected the amount of herbage harvested and contribution of herbage to residual p lant litter The greater inherent productivity of BG compared with RP and use of grazing vs. hay harvest increased plant litter pools and the potential for C and N contribution. The greater C:N ratio of pl ant residue from RP based systems favored mineralization over that from BG systems, which may more readily impact short term soil C and N pools. Overseeding RP and BG with an early maturing rye did not influence subsequent herbage production of the warm se ason perennials during Year 1 of the study. Interseeding cool season annuals into these systems may provide an additional winter forage option for producers when planting in clean tilled soils is not an option. However, further evaluation should be conduct ed with other combinations of winter annual forages (i.e., legumes and grasses) to determine impacts on subsequent herbage production and soil C and N dynamics, particularly for RP for which there are few published studies. Although there was a decrease i maintenance of C in larger soil fractions illustrates newly added organic material in the
141 soil, and suggests that production systems used in the present study were contributing to the accumulation of soil C and N. Greater C accumulation in the > 250 m fraction for the summer winter grazing management system compared with year round hay production illustrates the short term contribution of nutrient return from plant litter and excreta from livestock. Contin uing evaluation of these management practices is necessary to determine the long term effects of grazing and hay production on soil quality and the capacity of Florida soils to retain C within different particle size classes.
142 Table 5 1. Description of ma nagement treatments for warm season perennial based forage production systems. Forage management options System abbreviation Management of summer perennial Overseeded with rye Management of rye SH No Hay harvest (28 d) No SH WH Hay harvest (28 d) Ye s Hay harvest (28 d) SG No Rotational stocking (28 d) No SG WG Rotational stocking (28 d) Yes Rotational stocking (28 d) SH WG Hay harvest (28 d) Yes Rotational stocking (28 d) no winter overseeding (SH No), summer and winter hay production (SH WH), summer grazing management no winter overseeding (SG No), summer and winter grazing management (SG WG), and summer hay wint er grazing management (SH WG). Table 5 2. Species x system interaction ( P = 0.0013) for to tal rye herbage harvested (kg DM ha 1 ) from overseeded bermudagrass and rhizoma peanut. Data are means of three overseeded production systems across three replicates within a species (n = 9). Warm season species System Bermudagrass Rhizoma peanut P va lue -----------------------kg ha 1 --------------------SH WH 1080 1360 a 0.0840 SG WG 480 b 270 b 0.0359 SH WG 580 b 350 b 0.0148 SE 104 Systems include summer and winter hay production management (SH WH), summer and winter grazing management (SG WG), and summery hay production winter grazing (SH WG).
143 Table 5 3. Year x system interaction ( P < 0.0001) for total rye herbage harvested (kg DM ha 1 ) from overseeded berm udagrass and rhizoma peanut. Data are means of three overseeded forage systems across three replicates within a species (n = 9). Year System 2012 2013 P value -----------------------kg ha 1 -----------------------SH WH 2060 360 <0.001 SG WG 3 60 b 390 0.5649 SG WH 510 b 410 0.1278 SE 102 Systems include summer and winter hay production management (SH WH), summer and winter grazing management (SG WG), and summer y hay production winter grazing (SH WG). Table 5 4. Species x system interaction ( P < 0.0001) for total herbage harvested (kg DM ha 1 ) during the 2012 summer season. Data are means of five forage production systems across three replicates within a specie s (n = 15). Systems include summer hay production no winter overseeding (SH No), summer and winter hay production management (SH WH), summer grazing management no winter overseeding (SG No) summer and winter grazing management (SG WG), and summer hay production winter grazing (SH WG). Species System BG RP P value ----------------------kg ha 1 ------------------------SH No 10 500 320 b <0.0001 SH WH 9710 a 280 b <0.0001 SG No 5580 b 1010 a <0.0001 SG WG 5080 b 860 a <0.0001 SH WG 12 200 a 210 b <0.0001 SE 621
144 Table 5 5. Y ear x system interaction for total C concentration in various soil aggregate size fractions prior to the beginning of the experi ment in 2011 and at the end of the first year of the study in 2012 a,b Within a column, means without common superscripts differ P < 0.05. Table 5 6. Year x system interaction for total N c oncentration in various soil aggregate size fractions prior to the beginning o f the experiment in 2011 and at the end of the fir st year of the study in 2012 Forage Management System 0 to 10 cm layer 10 to 20 cm layer SH WH SG WG P value SH WH SG WG P value Size Class ----------------------------------g kg soil fraction -----------------------------------2011 0.23 0.25 b 0.8078 0.23 0.20 0.7980 End of 2012 0.18 0.42 a 0.0045 0.23 0.21 0.8156 SE 0.05 0.05 53 to 250 m 2011 0.30 0.28 0.6546 0.21 0.26 0.5640 End of 2012 0.25 0.28 0.8652 0.26 0.28 0.7692 SE 0.05 0.01 2011 4.90 4.91 0.8888 4.21 4.40 0.8130 End of 2012 4.76 4.86 0.5546 3.91 4.48 0.6532 SE 0.34 0.47 a,b Within a column, means without common superscripts differ P < 0.05. Forage Management Sy stem 0 to 10 cm 10 to 20 cm SH WH SG WG P value SH WH SG WG P value Size Class -----------------------------------g C kg 1 soil fraction --------------------------------> 250 m 2011 5.20 4.91 b 0.2030 5.16 4.68 0.4976 End 2012 4.38 6.9 8 a 0.0052 5.05 5.25 0.3010 SE 0.56 0.65 53 250 m 2011 5.13 5.83 0.6008 5.03 5.66 0.7777 End 2012 5.51 5.25 0.8125 6.01 5.93 0.8190 SE 0.68 0.91 < 53 m 2011 60.8 63.2 0.7546 67.8 70.5 a 0.2221 End 2012 50.7 64.2 0.0910 60.1 47.4 b 0.0060 SE 6.29 6.03
145 Figure 5 1. Monthly rainfall for 2011 and 2012 for the experiment al location and the 30 yr average for Citra, FL.
146 Figure 5 2. Date x system interaction ( P = 0.0172) for crude protein (g kg 1 DM) of winter rye overseeded forage systems. Data are means across two overseeded warm s eason perennials and three replicates (n = 6). Systems include summer and winter hay production (SH WH), summer and winter grazing management (SG WG), and summer hay winter grazing management (SH WG).
147 Figure 5 3. Dat e x species interaction ( P = 0.0317) for crude protein (g kg 1 DM) of summer 2012. Within a species, data are means across five forage production systems and three replicates (n = 15).
148 Figure 5 4. Herbage in vitro digestible organic matter (IVDOM) of warm season perennial forage systems in summer 2012. Data are means across two sampling dates and three replicates for each system (n = 6). Systems inc lude summer hay production no winter overseeding (SH No), summer and winter hay production (SH WH), summer grazing management no winter overseeding (SG No), summer and winter grazing management (SG WG), and summer hay production winter grazing management ( SH WG).
149 Figure 5 5. Year x species interaction ( P = 0.0002) for residual litter mass (kg ha 1 ) of winter overseeded bermudagrass (BG) and rhizoma peanut (RP). Data are means across three overseeded forage systems and three replicates (n = 9).
150 Figure 5 6. Species x system interaction ( P = 0.0068) for residual litter mass (kg DM ha 1 ) of winter overseeded bermudagrass (BG) and rhizoma peanut (RP) after 2 yr of imposing winter mana gement treatments. Data are means across three replicates within a species (n = 3). Systems include summer and winter hay production (SH WH), summer and winter grazing management (SG WG), and summer hay production winter grazing management (SH WG).
151 Figure 5 7. Species x system interaction ( P = 0.0382) for C:N Ratio of residual litter for winter overseeded bermudagrass (BG) and rhizoma peanut (RP) forage systems. For each system, data are means across three replicates within a species (n = 3). Systems include summer and winter hay production (SH WH), summer and winter grazing management (SG WG), and summer hay production winter grazing management (SH WG).
152 CHAPTER 6 CONCLUSIONS Increasing input costs especially fuel a nd fertilizer, have made the incorporation of legumes an increasingly attractive option for beef cattle producers. R hizoma peanut (RP; Arachis glabrata Benth.) is a warm season perennial legume with documented persistence (Ortega S. et al., 1992) that is w ell adapted to Florida and has potential for incorporation into grazing systems. Because of the high cost of establishment, a lternative establishment strategies such as strip planting, are needed if RP is to make significant contributions to grazing based systems in the future. Multiple entries of RP have been developed and released as dual purpose hay and grazing crops, and they e xhibit a range in growth habit, which may interact with their ability to establish and persist under grazing management. Becau se long term pasture productivity and persistence are of great importance in low input systems, evaluation of performance of these lines under grazing is needed to guide genotype selection by producers. Finally a s increasing emphasis is placed on the abil ity of grasslands to provide ecosystem services, RP has potential to contribute significantly in this regard in Florida (French et al., 2006). Tropical and temperate grasslands play a major role in the global C cycle and serve as an important C sink (Scurl ock and Hall, 1998). Management of grass based forage systems ( e.g., haying vs. grazing and grazi ng at a range of stocking rates) has been shown to affect their potential to store C (Franzluebbers and Stuedemann, 2002; Franzluebbers and Stuedemann, 2009), but no work has been done with warm season, legume based swards in the Southeast USA. The overall objective of the se projects was to eval uate the effect of RP genotypes on rate of establishment, response to grazing, and contribution to ecosystem
153 services i n various production systems in Florida. Specifically, the projects in this dissertation were developed to : i ) determine the effect of growth habit and defoliation management of rhizoma peanut cultivars on establishment success and changes in soil C dynami cs when strip planted into bahiagrass ( Paspalum notatum Flgge) sod (Chapter 3); ii ) evaluate the effects of grazing frequency and intensity on recently released rhiz oma peanut genotypes in order t o determine grazing management recommendations (Chapter 4 ) ; and iii) investigate the contribution of rhizoma peanut or N fertilized grass based year round forage management systems on soil C and N dynamics in Florida (Chapter 5 ). Growth Habit of Rhizoma Peanut Cultivars Effects on Establishment and Spread When Str ip Planted in Bahiagrass Sod Chapter 3 The establishment study was conducted at the Beef Research Unit in Gainesville, FL (29.72N, 82.35W) in 2011 and 2012, and year after establishment effects are reported for 2012. Four RP genotypes were strip plante d into clean tilled soil in March of each year, and were defoliated under rotational stocking or hay management every 28 d beginning in June. Rhizoma peanut shoot emergence was greatest for Florigraze and Ecoturf, which led to a trend for more favorable g round cover, frequency, and spread for these genotypes than UF Peace and Arblick throughout the experiment. Defoliation management did not negatively impact ground cover during the establishment year, but there was a reduction in spread potential when swa rds were grazed compared with harvested for hay. Hay production decreased the intensity of peanut removal from the planted strip during the establishment phase, and led to more favorable spread characteristics. Animal preference for RP in the planted strip is a key consideration
154 when using strip planting, and it is likely that longer rest periods are needed between defoliation events as well as earlier removal of livestock in order to allow taller stubble in the planted strip. During the year after establis hment, differences among genotypes were less prevalent. However, defoliation strategy continued to impact success of establishment, with reduced ground cover and spread for grazed than hayed treatments. Finally, changes in soil dynamics occurred when conv erting from a C 4 to a mixed C 3 /C 4 pasture system. A decrease in soil C illustrates the initial impact of tillage and land management on soil quality, although the contribution of C and N in the planted strip from RP began to increase during the first 2 yr after planting RP. Sward Characteristics of Rhizoma Peanut Culvitars Under a Range of Grazing Management Strategies Chapter 4 The defoliation management study was conducted at the Beef Research Unit in Gainesville, FL (29.72N, 82.35W) in 2012, and dat a from a second year of study are being collected in summer 2013. Four rhizoma peanut genotypes (Florigraze, Ecoturf, UF Peace, and UF Tito) were managed under two grazing frequencies (3 or 6 wk interval) and two grazing intensities (50 or 75% pregraze ca nopy height removal). Grazing frequency and intensity are important considerations when selecting management criteria for forages, and they were observed to impact above and below ground sward characteristics during Year 1 of this study. Total herbage ac cumulation did not differ among entries, but greater accumulation occurred with a grazing frequency of 6 vs. 3 wk. When RP genotypes were grazed every 6 wk, there was i) maintenance of a high percentage of RP ground cover, ii) similar or greater pre grazin g herbage leaf to stem ratio, and iii) greater pre grazing light interception than when grazed every 3 wk.
155 Less post grazing residual leaf area and greater weed frequency for the 75% canopy removal level illustrated that greater grazing intensity can affec t changes in the sward canopy. These results suggest that there may be an advantage for longer regrowth intervals, and visual observations during Year 2 suggest that this affect may become more pronounced over time. Defoliation Management Effects on Soil Carbon Dynamics of Year Round Production Systems Chapter 5 The experiment was conducted at the Plant Science Research and Education Unit in Citra, FL (29.24N, 82.10W) during 2012. A second year of the study is currently being conducted, but those data were not reported. Five year round forage management systems were imposed on both RP and N fertilized bermudagrass (BG; Cynodon spp.), and their contributions to soil C and N were evaluated. Herbage production of BG was greater than RP and was primarily a function of the greater inherent productivity of this species at the fertilization rates used. Nutritive value differed across the season, but all systems provided relatively high in vitro digestible organic matter (IVDOM) and crude protein (CP) until Aug ust. Residual litter mass was greater for the N fertilized BG systems, further reflecting the greater production of BG than RP. Management systems that employed grazing vs. hay harvest had greater litter mass. Plant litter quality was greater for RP based systems, suggesting that there may be greater potential mineralization of this material than that from BG pastures. During the winter management season, total herbage production of rye was relatively low, but nutritive value was high. There was no effect of overseeding RP with an early maturing rye on subsequent yield of RP.
156 The increased quantity of plant litter remaining at the end of winter 2013 for grazed vs. hayed systems illustrates increased potential for C and N contribution from the litter pool. Finally, soil C and N changes occurred over one full season of management, and illustrated the contribution of newly added organic matter in soil macroaggregates (i.e., soil particle size class > 250 m). Implications of the Research Strip planting of RP is a viable approach for the establishment of grass legume mixtures. Favorable establishment characteristics of RP genotypes like Ecoturf in this study may increase the need for distribution of planting material of less commonly utilized germplasms to prod ucers in Florida. Following genotype selection, the choice of management practice during the establishment phase is critical. These results indicate that management for hay production may provide a way to effectively utilize the bahiagrass component of the system without sacrificing RP establishment success. For established stands, the choice of grazing management may not affect herbage accumulation in the first year of grazing, but sward characteristics such as species botanical composition, ground cover, weed frequency, light interception, and residual leaf area index are good indicators of overall sward health. These traits can be utilized to determine the short term impacts and perhaps predict longer term effects of defoliation management. Conversion fr om row cropping to warm season perennial forage systems can increase soil C and N in the short term. Grazing these swards during the winter and summer months has potential to increase these pools to a greater degree than hay production through increased pl ant litter deposition and excreta from livestock.
157 Future Research Needs The described experiments addressed how to increase RP contribution to low input production systems in Florida. Based on the results of these studies, there are many new questions t hat have emerged relative to how to incorporate and promote management of RP in these systems. Specifically, when strip planting RP, the choice of RP cultivar is an important consideration; however, the choice of a grass component should also be investigat ed to determine the best fit for establishing mixed pasture systems. The growth habit of the companion grass may affect establishment success. Additionally, the species of grass utilized (i.e., bermudagrass) is an area of interest among producers consideri ng this technology. Furthermore, on farm evaluations of this approach would further quantify the efficacy of management of strip planted areas on a larger production scale and under producer management. Utilizing effective grazing management strategies fo r RP will insure long term stand production potential. Although a second year of this evaluation is being conducted, continuing this work for a third year may provide further information on long term stand persistence that could be of value for making reco mmendations to producers. Previous evaluations with RP have shown that multiple years of evaluation are typically needed before these effects become prevalent (Hernndez Garay et al., 2004), and an additional year may provide this information. Overseeding RP with winter forages enables producers to have an additional forage resource when RP is dormant. Using cool season annuals may provide this option, but further evaluation of grasses, legumes, and mixtures should be conducted to quantify the impact on su bsequent production of RP in the summer months. Utilizing a
158 grass legume mixture may provide additional N to the system and serve as a source of high quality forage during the winter. Finally, long term evaluation of the effects of forage management pract ices on soil C and N is needed in Florida. It is important to define practices that not only optimize production and persistence of the forage component of the system, but also increase the contribution to soil quality and overall ecosystem health. As sust ainability continues to be a global issue, understanding the effect of these practices in subtropical environments will provide the basis for best management practices in this region.
159 LIST OF REFERENCES Allen, V.G., C. Batello, E.J. Berretta, J. Hodgson, M. Kothmann, X. Li, J. Mclvor, J.Milne, C. Morris, A. Peeters, and M. Sanderson. 2011. An international terminology for grazing lands and grazing animals. Grass and Forage Sci. 66:2 28. Andrews, J., C. Olson, and P.J. van Blokland. 1985. Cost, returns an d establishment bermuda grass.Univ. of Florida Food Resource Econ. Dep. Staff Paper., Inst. Food Agric. Sci., Gainesville, FL. Angle, J.S. 2011. Land grant school research key to US lead in food production. The Atlanta Journal Constitution. Atlanta, GA. Apolinrio, V.X., J.C.B. Dubeux, A.C.L. Mello, J.M.B. Vendramini, M.A. Lira, M.V.F. Santos, and J.P. Muir. 2013. Deposition and decomposition of signal grass pasture litter under varying nitrogen fertilizer and stocking rates. Agron. J. 4:999 1004. Ball, D.M., C.S. Hoveland, and G.D. Lacefield. 2007. Southern Forages. International Plant Institute, Norcross, GA. Beltranena, R., J. Breman, and G.M. Prine.1981. Yield and quality of florigraze rhizoma peanut ( Arachis glabrata Benth.) as affected by cutting height and frequency. Proc. Soil Sci. Soc. Fla. 40:153 156. Blanco Canqui, H., C. Gantzer, S. Anderson, E. Alberts, and A. Thompson. 2004. Grass barrier and vegetative filter stri p effectiveness in reducing runoff, sediment, nitrogen, and phosphorus loss. Soil Sci. Soc. Am. J. 68:1670 1678. Boddey, R., R. Macedo, R. Tarre, E. Ferreira, O. De Oliveira, C. de P Rezende, et al. 2004. Nitrogen cycling in Brachiari a pastures: the key t o understanding the process of pasture decline. Agric. Ecosyst. Environ. 103:389 403. Boddey, R.M., J.C. De Moraes S, B.J. Alves and S. Urquiaga. 1997. The contribution of biological nitrogen fixation for sustainable agricultural systems in the tropics. Soil Biol. Biochem. 29:787 799. Boutton, T.W. 1991. Stable carbon isotope ratios of natural materials I: Sample preparation and mass spectrometric analysis. p. 155 171. In Coleman, D.C and B. Fry (eds.) Carbon Isotope Techniques. Academic Press, New York. 13 C values of soil organic carbon and their use in documenting vegetation change in a subtropical savanna ecosystem. Geoderma 82:5 41.
160 Butler, T.J., J.P. Muir, M.A. Islam, and J.R Bow. 2007. Rhizoma peanut yield and nutritive value are influenced by harvest technique and timing. Agron. J. 99:1559 1563. Butler, T.J., W.R. Ocumpaugh, M.A. Sanderson, R.L. Reed, and J.P. Muir. 2006. Evaluation of rhizoma peanut genotypes for adaptatio n in Texas. Agron. J.98: 1589 1593. Cambardella, C.A., and E.T. Elliott. 1992. Particulate soil organic matter changes across a grassland cultivation sequence. Soil Sci. Soc. Am. J. 56:777 783. Canudas, E.G., K.H. Quesenberry, L.E. Sollenberger, and G.M. Prine. 1989. Establishment of two cultivars of rhizoma peanut as affected by weed control and planting rate. Tropical Grasslands. 23:162 170. Casler, M.D., E. Heaton, K.J. Shinners, H.G. Jung, P.J. Weimer, M.A. Liebig, R.B. Mitchell, and M.F. Digman. 2009 Grasses and legumes for cellulosic bioenergy.p. 205 219. In Wedin, F. and S.L. Fales (eds.) Grassland: Quietness and strength for a new American agriculture. American Society of Agronomy, Madison, WI. Castillo, M.S., L.E. Sollenberger, A.R. Blount, J.A. Ferrell, M.J. Williams and C.L. Mackowiak. 2013. Strip planting a legume into warm season grass pasture: Defoliation effects during the year of establishment. Crop Sci. 53:724 731. Chaparro, C.J., L.E. Sollenberger and K.H. Quesenberry. 1996. Light interc eption, reserve status, and persistence of clipped Mott elephantgrass swards. Crop science 36:649 655. Chapman, D.F. and Lemaire, G.1993. Morphogenetic and structural determinants of plant regrowth after defoliation. Proceedings of the XVII International Grassland Congress. Palmerston, New Zealand. p. 95 104. Christensen, B.T. 1996. Carbon in primary and secondary organomineral complexes. p. 97 165. In Carter, M. R. and B. A. Stewart (eds.) Structure and organic matter storage in agricultural soils. CRC P ress Inc., Boca Raton, FL. Christiansen, S., O.C. Ruelke, W.R. Ocumpaugh, K.H. Quesenberry, and J.E. Moore. 1988. Seasonal yield and quality of'Bigalta','Redalta'and'Floralta'limpograss. Tropical Agriculture. 65:49 55. Clay, D.E., C.E. Clapp, C. Reese, Z Liu, C.G. Carlson, H. Woodard, and A. Bly. 2007. Carbon 13 fractionation of relic soil organic carbon during mineralization effects calculated half lives. Soil Sci. Soc. Am. J. 71:1003 1009. Conant, R.T., K. Paustian, and E.T. Elliott. 2001. Grassland m anagement and conversion into grassland: effects on soil carbon. Ecological Applications 11:343 355.
161 Curll, M., and R. Jones. 1989. The Plant Animal Interface and Legume Persistence An Australian Perspective.p. 339 359. In Marten, G.C., A.G. Matches, R.F. Barnes, R.W. Brougham, R.J. Clements, and G.W. Sheath (eds).Persistence of forage legumes: American Society of Agronomy, Madison, WI. Day, J.L., A. E. Coy, and J. Gassett. 2012. 2011 2012 Small Grain Performance Tests. University of Georgia Cooperative E xtension Publ. No. 100 3. Dubeux, J.C.B., L.E. Sollenberger, B.W. Mathews, J.M. Scholberg, and H.Q. Santos. 2007. Nutrient cycling in warm climate grasslands. Crop Sci. 47:915 928. Dubeux Jr, J.C.B., L.E. Sollenberger, N.B. Comerford, J.M. Scholberg, A.C Ruggieri, J.M.B. Vendramini, S.M. Interrante, and K.M. Portier. 2006. Management intensity affects density fractions of soil organic matter from grazed bahiagrass swards. Soil Biol. Biochem. 38: 2705 2711. Dunavin, L.S. 1990. Cool season forage crops se eded over dormant rhizoma peanut. J. Prod. Agric. 3: 112 114. Dunavin, L.S. 1992. Florigraze rhizoma peanut in association with warm season perennial grasses. Agron. J. 84:148 151. Elbasha, E., P.K. Thornton, and G. Tarawali. 1999. An ex post economic imp act assessment of planted forages in West Africa. p. 68. In ILRI Impact Assessment Series 2, IRLI, Nairobi, Kenya. Ferrell, J.A., and B.A. Sellers. 2012. Weed control in perennial peanut. Agronomy Department, Florida Cooperative Extension Service, Institu te of Food and Agricultural Sciences, University of Florida, SS AGR 261. Follett, R. 2001. Soil management concepts and carbon sequestration in cropland soils. Soil Tillage Res. 61:77 92. Follett, R.F., and D.A. Reed. 2010. Soil carbon sequestration in gr azing lands: Societal benefits and policy implications. Rangeland Ecology and Management 63:4 15. Foster, J.L., A.T. Adesogan, J.N. Carter, L.E. Sollenberger, A.R. Blount, R.O. Myer, S.C. Phatak, and M.K. Maddox. 2009. Annual legumes for forage systems in the United States gulf coast region. Agron. J. 101:415 421. Frame, J. 1981. Herbage mass. p. 39 67. In Davies, A., S. A. Grant, and A.S. Laidlaw (eds.) Sward measurement handbook, 2 nd edition, British Grassland Society, Reading, UK. Frame, J., and A. S. Laidlaw. 2011. Managing white clover in mixed swards principles and practice. Pastos. 28:5 33.
162 Franzluebbers, A.J., J.A. Stuedemann, H.H. Schomberg, and S.R. Wilkinson. 2000. Soil organic C and N pools under long term pasture management in the Southern Pi edmont USA. Soil Biol. Biochem. 32:469 478. Franzluebbers, A.J., and J.A. Stuedemann. 2002. Particulate and non particulate fractions of soil organic carbon under pastures in the Southern Piedmont USA. Environ. Pollut. 116:S53 S62. Franzluebbers, A.J., a nd J.A. Stuedemann. 2008. Early response of soil organic fractions to tillage and integrated crop livestock production. Soil Sci. Soc. Am. J. 72:613 625. Franzluebbers, A.J., and J.A. Stuedemann. 2009. Soil profile organic carbon and total nitrogen during 12 years of pasture management in the Southern Piedmont USA. Agric. Ecosyst. Environ.129:28 36. Franzluebbers, A., S.R. Wilkinson, and J.A. Stuedemann. 2004. Bermudagrass management in the Southern Piedmont USA: X. Coastal productivity and persistence in response to fertilization and defoliation regimes. Agron. J. 96:1400 1411. Freire, M.J., C.A. Kelly Begazo and K.H. Quesenberry. 2000. Establishment, yield, and competitiveness of rhizoma perennial peanut germplasm on a flatwoods soil. Soil and Crop Sci Soc. Am. Fla. Proc. 59:68 72. French, E.C. Perennial peanut: A promising forage for dairy herd management in the tropics. p. C20 C41. In Proc. Int. Conf. Livestock Tropics, Gainesville, FL. 19 25 June 1988. IFAS, Univ. of Florida, Gainesville. French, E.C.,G.M. Prine, and A.R. Blount. 2006. Perennial peanut: An alternative forage of growing importance. Agronomy Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida, SS AGR 99. French, E. an d G. Prine. 2002. Agronomy Facts:Perennial peanut establishment guide. Agronomy Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida. SS AGR 35. Gallaher, R.N., C.O. Weldon and J.G. Futral. 1 975. An aluminum block digester for plant and soil analysis. Soil Sci. Soc. Am. J. 39:803 806. Ganjegunte, G.K., G.F. Vance, C.M. Preston, G.E. Schuman, L.J. Ingram, and P.D. Stahl. Soil organic carbon composition in a northern mixed grass prairie. Soil Sc i. Soc. Am. J. 69:1746 1756. Garay, A.H., P.A.M. Hernndez, J.Z. Esparza, H.V. Huerta, F.O. Gallardo, B.M.J. Torres, and M.E.V. Zebada. 2012. Characterization of forage yield of an alfalfa
163 orchardgrass pasture at different grazing frequencies and intensi ties. Revista Fitotecnia Mexicana 35:259 266. 44:1355 1360. Giller, K.E., J. Ormesher, a nd F.M. Awah. 1991. Nitrogen transfer from Phaseolus bean to intercropped maize measured using 15 N enrichment and 15 N isotope dilution methods. Soil Biol. Biochem. 23:339 346. Hart, S.C., M.K. Firestone, E.A. Paul, and J.L. Smith. 1993. Flow and fate of soil nitrogen in an annual grassland and a young mixed conifer forest. Soil Biol. Biochem. 25:431 442. Hassink, J., A.P. Whitmore, and J. Kubt. 1997. Size and density fractionation of soil organic matter and the physical capacity of soils to protect org anic matter. Developments in Crop Science 25:245 255. Haynes, R. 1980. Competitive aspects of the grass legume association. Adv. Agron 33:227 261. Haynes, R. and P. Williams. 1993. Nutrient cycling and soil fertility in the grazed pasture ecosystem. Adv. Agron. 49:119 199. Haystead, A., N. Malajczuk, and T. Grove. 1988. Underground transfer of nitrogen between pasture plants infected with vesicular arbuscular mycorrhizal fungi. New Phytol. 108:417 423. Hernndez, M., P. Argel, M.A. Ibrahim, and L.t. Man netje. 1995. Pasture production, diet selection and liveweight gains of cattle grazing Brachiaria brizantha with or without Arachis pintoi at two stocking rates in the Atlantic Zone of Costa Rica. Tropical Grasslands 29:134 141. Hobbie, E.A., and R.A. Wer ner. 2004. Intramolecular, compound specific, and bulk carbon isotope patterns in C 3 and C 4 plants: a review and synthesis. New Phytol. 161:371 385. Holt, E.C., and B.E. Conrad. 1986. Influence of harvest frequency and season on bermudagrass cultivar yield and forage quality. Agron. J. 78: 433 436. Hoveland, C., R. Durham, and J. Bouton. 1996. Weed encroachment in established alfalfa as affected by cutting frequency. J. Prod. Agric. 9:399 402. Huber, H., S. Lukcs and M.A. Watson. 1999. Spatial structure of stoloniferous herbs: an interpla y between structural blue print, ontogeny and phenotypic plasticity. Plant Ecology 141:107 115.
164 Ingram, J. and E. Fernandes. 2001. Managing carbon sequestration in soils: concepts and terminology. Agric. Ecosyst. Environ. 87:111 117. Interrante, S.M., J. P. Muir, A.M. Islam, A.L. Maas, W.F. Anderson, and T.J. Bulter. 2011. Establishment, agronomic characteristics, and dry matter yield of rhizoma peanut genotypes in cool environments. Crop Sci. 51:2256 2261. Johnson, C.R., B.A. Reiling, P. Mislevy, and M.B Hall. 2001. Effects of nitrogen fertilization and harvest date on yield, digestibility, fiber, and protein fractions of tropical grasses. J. Anim. Sci. 79:2439 2448. Kalmbacher, R.S., F.M. Pate, and F.G. Martin. 2002. Grazing evaluation of bahiagrass wi th evenia aeschynomene and stylosanthes. Soil Crop Sci. Soc. Florida Proc. 61:10 15. Kantar, M., C. Sheaffer, P. Porter, E. Krueger and T.E. Ochsner. 2011. Growth stage influences forage yield and quality of winter rye. Forage and Grazinglands. doi:10.10 94/FG 2011 0126 01 RS. Karlen, D., M. Mausbach, J. Doran, R. Cline, R. Harris and G. Schuman. 1997. Soil quality: A concept, definition, and framework for evaluation. Soil Sci. Soc. Am. J. 61:4 10. Knops, J. M., and D. Tilman. 2000. Dynamics of soil nit rogen and carbon accumulation for 61 years after agricultural abandonment. Ecology. 81:88 98. Laidlaw, A. S., and N. Teuber. 2001. Temperate forage grass legume mixtures: advances and perspectives. Proceedings of the Int. Grassland Congress. So Paulo, Bra zil. p. 85 92. Lal, R., R. Follett, J. Kimble and C. Cole. 1999. Managing US cropland to sequester carbon in soil. Journal Soil Water Conserv. 54:374 381. Lambers, H., J.A. Raven, G.R. Shaver and S.E. Smith. 2008. Plant nutrient acquisition strategies c hange with soil age. Trends in Ecology & Evolution 23:95 103. Lascano, C.E., and J. Estrada. 1989. Long term productivity of legume based and pure grass pastures in the Eastern Plains of Colombia. Proceedings of the XVI Int. Grassland Congress, Nice, Fran ce. p.1179 1180 Lascano, C.E., and D. Thomas. 1988. Forage quality and animal selection of Arachis pintoi in association with tropical grasses in the eastern plains of Colombia. Grass and Forage Sci. 43:433 439. Lascano, C.E., M. Peters, and F.J. Holmann 2005. Arachis pintoi in the humid tropics of Colombia: a forage legume success story. Proceedings of the XX International Grassland Congress. Dublin, Ireland. p. 327.
165 Lefroy, R.D., G.J. Blair, and W.M. Strong. 1993. Changes in soil organic matter with cropping as measured by organic carbon fractions and 13 C natural isotope abundance. Plant Soil. 155:399 402. Liu, K., L.E. Sollenberger, Y.C. Newman, J.M.B. Vendramini, S.M. Interrante, and R. White Leech. 2011a. Grazing management effects on productivity nutritive 360. Liu, K., L.E. Sollenberger, M.L. Silveira, Y.C. Newman, and J.M.B. Vendramini. 2011b. ermudagrass pastures: I. Mass, deposition rate, and chemical composition. Agron. J. 103:156 162. Liu, K., L.E. Sollenberger, M.L. Silveira, J.M.B. Vendramini, and Y.C. Newman. 2011c. Grazing intensity and nitrogen fertilization affect litter responses in Agronomy J. 103:163 168. Maass, B., M. Sawkins, and S. Chakraborty. 2004. History, relationships and diversity among Stylosanthes species of commercial significance. High yie lding anthracnose resistant Stylosanthes for agricultural systems. ACIAR Monograph No. 111. Australian Center for International Agricultural Research (ACIAR), Canberra, Australia. p.9 26. Magdoff, F., R. Weil, and R. Ray. 2004. Soil organic matter managem ent strategies. p.45 65. In Ma gdoff, F., and R.R. Weil (eds.). Soil Organic Matter in Sustainable Agriculture. CRC Press, New York. Maroso, R.P., S.M. Scheffer Basso and C.M. Carneiro. 2007. Regrowth of Lotus spp. with different growth habits. Revista Bras ileira de Zootecnia 36:1524 1531. Matches, A.G. and J.C. Burns. 1995. Systems of grazing management. p. 179 192. In R.F.Barnes (ed.) Forages: The science of grassland agriculture. Iowa State Univ. Press, Ames, IA. Mathews, B.W., L.E. Sollenberger, and C. R. Staples. 1994. Dairy heifer and bermudagrass pasture responses to rotational and continuous stocking. J.Dairy Sci. 77:244 252. Mathews, B.W., L.E. Sollenberger, and J. Tritschler. 1996. Grazing systems and spatial distribution of nutrients in pastures: soil considerations. p. 213 229. In R.E. Joost and C. A. Roberts (eds.). Nutrient Cycling in Forage Systems. Potash and Phosphate Institute, Norcross, GA. Mathews, B.W., S.C. Miyasaka, and J.P. Tritschler. 2004. Mineral nutrition of C 4 forage grasses. p. 217 265 In L.E. Moser et al. (ed.) Warm Season (C 4 ) Grasses. Am. Soc. Agron. Monograph No. 45. ASA, Madison, WI..
166 McSherry, M.E., and M.E. Ritchie. 2013. Effects of grazing on grassland soil carbon: a global review. Global Change Biol. 19:1347 1357. Me ijboom, F.W., J. Hassink, and M. Van Noordwijk. 1995. Density fractionation of soil macroorganic matter using silica suspensions. Soil Biol. Biochem. 27:1109 1111. Miles, J. W. 2001. Achievements and perspectives in the breeding of tropical grasses and le gumes. Proceedings of the XIX International Grassland Congress, Sao Paulo, Brazil. p. 541 542. Mislevy, P., and F.G. Martin. 1998. Comparison of Tifton 85 and other Cynodon grasses for production and nutritive value under grazing. Proc. Soil Crop Sci. Soc Fla. 57:77 82. Mislevy, P., M. J. Williams, A. S. Blount, and K. H. Quensenberry. 2007. Influence of harvest management on rhizoma perennial peanut production, nutritive value and persistence on flatwood soils. Forage and Grazinglands. doi:10.1094/FG 20 07 1108 01 RS. Moore, J., and G. Mott. 1974. Recovery of residual organic matter from in vitro digestion of forages. J. Dairy Sci. 57:1258 1259. Mousel, E., W.H. Schacht, C. Zanner, and L.E. Moser. 2005. Effects of summer grazing strategies on organic res erves and root characteristics of big bluestem. Crop Sci. 45:2008 2014. Moyer, J.L., and K.P. Coffey. 2000. Forage quality and production of small grains interseeded into bermudagrass sod or grown in monoculture. Agron. J. 92:748 753. Muir, J.P. 2009. Her bage, phosphorus, and nitrogen yields of winter season forages on high phosphorus soil. Agron J. 101:764 768. Perennial Peanut for Cool, Dry Climates. J. Plant Reg. 4:106 10 8. Muir, J., W. Pitman, and J. Foster. 2011. Sustainable, low input, warm season, grass legume grassland mixtures: mission (nearly) impossible? Grass and Forage Sci. 66:301 315. National Research Council. 1996. Nutrient Requirements of beef Cattle, 7 th revised ed. National Academy Press, Washington, D.C. Nelson, C.J. 2000. Shoot morphological plasticity of grasses: Leaf growth vs. tillering. p.101 126. In G. Lemaire et al. (ed.) Grassland ecophysiology and grazing ecology. CABI Pub.,New York.
167 Nelson, N. 1944. A photometric adaptation of the Somogyi method for the determination of glucose. J. Biol. Chem. 153:375 379. Newell, M.A. and T.J. Butler. 2013. Forage rye improvement in the southern united states: a review. Crop Sci. 53:38 47. Newman, Y.C. 2011. Bahiagrass ( Paspalum notatum ): Overview and management. Agron omy Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida. SS AGR 332. Niles, W.L., E.C. French, P.E. Hildebrand, G. Kidder, and G.M. Prine. 1990. Establishment of Florigraze rhizoma peanut ( Arachis glabrata Benth.) as affected by lime, phosphorus, potassium, magnesium and sulfur. Proc. Soil Crop Sci. Soc. Fla. 49:207 210. Noble, A. and P. Nelson. 2000. Sustainability of Stylosanthes based pasture systems in Northern Australia: Managing soil acidity. p. 60 69. In Lambert, J. (ed.) 1999 Peer Review of Resource and Whole Property Management Projects. North Australia Program Occassional Publication No. 10. Ortega S J.A., L.E. Sollenberger, J.M. Bennett, and J.A. Cornell. 1992a. Rhizome charact eristics and canopy light interception of grazed rhizoma peanut pastures. Agron. J. 84:804 809. Ortega S J.A., L.E. Sollenberger, K.H. Quesenberry, J.A. Cornell, and C.S. Jones. 1992b. Productivity and persistence of rhizoma peanut pastures under differ ent grazing managements. Agron. J. 84:799 804. Paustian, K., H.P. Collins, and E.A. Paul. 1997. Management controls on soil carbon. p. 15 49. In Soil organic matter in temperate agroecosystems: Long term experiments in North America. CRC Press, Boca Raton FL: Pedreira, C.G.S., L.E. Sollenberger, and P. Mislevy. 1999. Productivity and nutritive value of `florakirk' bermudagrass as affected by grazing management. Agron. J. 91:796 800. Sci. 26:836. Poppi, D.P., and S. McLennan. 1995. Protein and energy utilization by ruminants at pasture. J. Anim. Sci. 73:278 290. Post, W.M., and K.C. Kwon. 2000. Soil carbon sequestration and land use change: processes and potential. G lobal Change Biol. 6:317 327. Prevatt, W. 2008. Making adjustments to the cattle herd due to higher production costs. Alabama Cattleman. p.36.
168 Prine, G.M., L.S. Dunavin, J.E. Moore, and R.D. Roush. 1981. 'Florigraze' rhizoma peanut : a perennial forage l egume. Agronomy Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida, Circular S 275. Prine, G.M., L.S. Dunavin, J.E. Moore, and R.D. Roush. 1986. Registration of Crop Sci.26:5. Prine, G.M., L.S. Dunavin, R.J. Glennon, and R.D. Roush. 1990. Registration of Arbrook rhizoma peanut. Crop Sci. 30:743 744. Prine, G.M., E.C. French, A.R. Blount, M.J. Williams, and K.H. Quesenberry. 2010. Registration of Arblick and Ec oturf rhizoma peanut germplasms for ornamental or forage use. Journal of Plant Registrations. 4:145 148. Quesenberry, K.H., A.R. Blount, P. Mislevy, E.C. French, M.J. Williams, and G.M. Prine. 2010. Registration of 'UF Tito' and 'UF Peace' rhizoma peanut cultivars with high dry matter yields, persistence, and disease tolerance. Journal of Plant Registrations 4:17 21. Rees, R., I. Bingham, J. Baddeley, and C. Watson. 2005. The role of plants and land management in sequestering soil carbon in temperate arab le and grassland ecosystems. Geoderma 128:130 154. Rice, R., L. Sollenberger, K. Quesenberry, G. Prine and E. French. 1995. Defoliation effects on rhizoma perennial peanut rhizome characteristics and establishment performance. Crop Sci. 35:1291 1299. Ric e, R.W., L.E. Sollenberger, K.H. Quesenberry, G.M. Prine, and E.C. French. 1996. Establishment of rhizoma perennial peanut with varied rhizome nitrogen and carbohydrate concentrations. Agron. J. 88:61 66. Richards, J. H. 1993. Physiology of plants recover ing from defoliation. In M. J. Baker (ed.) Grasslands for our world. SIR Publishing, Wellington, New Zealand. Romero, F., H. Van Horn, G. Prine, and E. French. 1987. Effect of cutting interval upon yield, composition and digestibility of Florida 77 alfalf a and Florigraze rhizoma peanut. J. Anim. Sci. 65:786 796. Roth, L., F. Rouquette, and W. Ellis. 1990. Effects of herbage allowance on herbage and dietary attributes of Coastal bermudagrass. J. Anim. Sci. 68:193 205. Rouquette, F., and G. Smith. 2010. Re view: Effects of biological nitrogen fixation and nutrient cycling on stocking strategies for cow calf and stocker programs. The Prof. Anim. Sci. 26:131 141.
169 Ruffo, M.L., and G.A. Bollero. 2003. Modeling rye and hairy vetch residue decomposition as a func tion of degree days and decomposition days. Agron. J. 95:900 907. Russelle, M. 1996. Nitrogen cycling in pasture systems.p.203 212. In Joost, R.E. and C.A. Roberts (ed.) Nutrient Cycling in Forage Systems. Potash and Phosphate Institute and Foundation of Agronomic Research. Manhattan, KS. Saldivar, A.J., W.R. Ocumpaugh, R.R. Gildersleeve, and J.E. Moore. 1990. Growth analysis of Florigraze rhizoma peanut forage nutritive value. Agron. J. 82:473 477. Saldivar, A.J., W.R. Ocumpaugh, R.R. Gildersleeve, an d G.M. Prine. 1992. Total nonstructural carbohydrates and nitrogen of Florigraze rhizoma peanut. Agron. J. 84:439 444. Schuman, G., J. Reeder, J. Manley, R. Hart, and W. Manley. 1999. Impact of grazing management on the carbon and nitrogen balance of a mi xed grass rangeland. Ecol. Applic. 9:65 71. Schuman, G.E., H.H. Janzen, and J.E. Herrick. 2002. Soil carbon dynamics and potential carbon sequestration by rangelands. Environ. Pollut. 116:391 396. Schwinning, S., and A. Parsons. 1996. Analysis of the coe xistence mechanisms for grasses and legumes in grazing systems. J. Ecol. 84:799 813. Scurlock, J. M. O., and D. O. Hall.1998. The global carbon sink: a grassland perspective. Global Change Biology. 4: 229 233. Soussana, J. F., P. Loiseau, N. Vuichard, E. C eschia, J. Balesdent, T. Chevallier, and D. Arrouays. 2004. Carbon cycling and sequestration opportunities in temperate grasslands. Soil Use and Management. 20:219 230. National Agricultural Statistics Service. 2013. Crops and Plants. http://www.nass.usda .gov/. Accessed 23 Sept. 2013. Shelton, H.M., S. Franzel, and M. Peters. 2005. Adoption of tropical legume technology around the world: analysis of success. Tropical Grasslands 39:198 209. Silveira, M.L., K. Liu, L.E. Sollenberger, R.F. Follett, and J.M. B. Vendramini. 2013. Short term effects of grazing intensity and nitrogen fertilization on soil organic carbon pools under perennial grass pastures in the southeastern USA. Soil Biol. Biochem. 58:42 49.. Six, J., H. Bossuyt, S. Degryze, and K. Denef. 2004 A history of research on the link between (micro) aggregates, soil biota, and soil organic matter dynamics. Soil and Tillage Research. 79:7 31.
170 Six, J., R. Conant, E. Paul and K. Paustian. 2002. Stabilization mechanisms of soil organic matter: implicatio ns for C saturation of soils. Plant Soil. 241:155 176. Six, J., K. Paustian, E.T. Elliott and C. Combrink. 2000. Soil structure and organic matter: I. Distribution of aggregate size classes and aggregate associated carbon. Soil Sci. Soc. Am. J. 64:681 689. Sollenberger, L.E., C.T. Agouridis, E.S. Vanzant, A.J. Franzluebbers, and L.B. Owens. 2012. Prescribed grazing on pasturelands. In C. J. Nelson, editor Conservation outcomes from pasture and hayland practices: Assessment, recommendations, and knowledge g aps. Allen Press, Lawrence, KS. Sollenberger, L.E. and C.G. Chambliss. 1989. Grazing management of improved pastures. Proceedings of the UF/IFAS Beef Cattle Short Course. Gainesville, FL. p.42 44. Sollenberger, L.E., C.S. Jones, and G.M. Prine. 1989. Ani mal performance on dwarf elephant grass and rhizoma peanut pastures. XVI International Grassland Congress, Nice, France. p.1189 1190. Sollenberger, L.E., and R.S. Kalmbacher. 2005. Aeschynomene and carpon desmodium: legumes for bahiagrass pasture in Florid a. Proceedings of the XX International Grassland Congress. Dublin, Ireland. p.334. Sollenberger, L.E. and Y.C. Newman. 2007. Grazing management. p. 651 659. In R. F. Barnes, editor Forages: The science of grassland agriculture. Blackwell, Ames, IA. Solle nberger, L.E., K.H. Quesenberry, C.S. Jones and J.R. Cornell. 1992. Productivity and persistence of rhizoma peanut pastures under different grazing managements. Agron. J. 84:799 804. Stewart Jr, R.L., J.C.B. Dubeux Jr, L.E. Sollenberger, J.M.B. Vendramini and S.M. Interrante. 2005. Stocking method affects plant responses of Pensacola bahiagrass pastures. Forage and Grazinglands. doi: 10.1094/FG 2005 1028 01 RS. Stobbs, T. 1975. Factors limiting the nutritional value of grazed tropical pastures for beef and milk production. Tropical Grasslands 9:141 150. Terrill, T.H., S. Gelaye, S. Mahotiere, E.A. Amoah, S. Miller, R.N. Gates, and W.R. Windham. 1996. Rhizoma peanut and alfalfa productivity and nutrient composition in central Georgia. Agron. J. 88:485 488. Thomas, R. 1992. The role of the legume in the nitrogen cycle of productive and sustainable pastures. Grass and Forage Sci. 47:133 142.
171 Thomas, R. 1995. Role of legumes in providing N for sustainable tropical pasture systems. Plant Soil 174:103 118. Thoma s, R., and N. Asakawa. 1993. Decomposition of leaf litter from tropical forage grasses and legumes. Soil Biol. Biochem. 25:1351 1361. Tisdale, S.L., W.L. Nelson and J.D. Beaton. 1985. Soil fertility and fertilizers. Collier Macmillan Publishers, London. T isdall, J. M.., and J.M. Oades. 1982. Organic matter and water stable aggregates in soils. J. Soil Sci. 33:141 163. Trannin, W., S. Urquiaga, G. Guerra, J. Ibijbijen and G. Cadisch. 2000. Interspecies competition and N transfer in a tropical grass legume mixture. Biol. and Fert. of Soils. 32:441 448. Trimble, S.W. and A.C. Mendel. 1995. The cow as a geomorphic agent a critical review. Geomorphology. 13:233 253. Utley, P., W. Marchant and W. McCormick. 1976. Evaluation of annual grass forages in prepared seedbeds and overseeded into perennial sods. J. Anim. Sci. 42:16 20. Valencia, E., M.J. Williams, C.C. Chase, Jr., L.E. Sollenberger, A.C. Hammond, R.S. Kalmbacher, et al. 2001. Pasture management effects on diet composition and cattle performance on cont inuously stocked rhizoma peanut mixed grass swards. J. Anim. Sci. 79:2456 2464. Valentim, J. F., and C.M.S. Andrade. 2005. Forage peanut ( Arachis pintoi ): a high yielding and high quality tropical legume for sustainable cattle production systems in the We stern Brazilian Amazon. Proceedings of the XX International Grassland Congress, Dublin, Ireland. p.329. Valentim, J. F., O.C. Ruelke, and G. M. Prine. 1986. Yield and quality responses of tropical grasses, a legume and grass legume associations as affected by fertilizer nitrogen. Proc. Soil Crop Sci. Soc. Fla. 45:138 143. Vendramini, J.M.B., M.L. Silveira, J.C.B. Dubeux Jr and L.E. Sollenberger. 2007. Environmental impacts and nutrient recycling on pastures grazed by cattle. Revista Brasileira de Zootecnia 36:139 149. Venuto, B.C., W.D.Pitman, D.D. Redfearn, and E.K. Twidwell. 1999. Rhizoma peanut: a new forage option for Louisiana. Louisiana Agric. Exp. Stn. Circ. 136. Louisiana State Univ., Baton Rouge. Weijsched, J., R. Berentsen, H. de Kroon, H. Hube r 2008. Variation in petiole and internode length affects plant performance in Trifolium repens under opposing selection regimes. Evolutionary Ecology. 22:383 397.
172 Williams, M.J. 1993. Planting date and preplant tillage effects on emergence and survival of rhizoma perennial peanut. Crop Sci. 33:132 136. Williams, M., C. Chase and A. Hammond. 2004. Performance of cows and their calves creep grazed on rhizoma perennial peanut. Agron. J. 96: 671 676. Williams, M.J., C.A. Kelly Begazo, R.L. Stanley, K.H. Que senberry and G.M. Prine. 1997. Establishment of rhizoma peanut: Interaction of cultivar, planting date, and location on emergence and rate of cover. Agron. J. 89:981 987. Williams, M.J., T.R. Sinclair, P. Mislevy, K.H. Quesenberry, A.S. Blount and S.W. Co leman. 2008. Photoperiod sensitivity of rhizoma peanut germplasm. Agron. J. 100:1366 1370. Woodard, K.R., E.C. French, L.A. Sweat, D.A. Graetz, L.E. Sollenberger, B. Macoon, K.M. Portier, B.L. Wade, S.J. Rymph, G.M. Prine, and H.H. Van Horn. 2002. Nitrogen removal and nitrate leaching for forage systems receiving dairy effluent. J. Environ. Qual. 31:1980 1992. Wright, A.L., F.M. Hons, and J.E. Matocha Jr. 2005. Tillage impacts on microbial biomass and soil carbon and nitrogen dynamics of corn and cotton ro tations. Appl. Soil Ecol. 29:85 92.
173 BIOGRAPHICAL SKETCH Mary Kimberly Mullenix is from Newnan, Georgia and is the 5th generation of her family on a farm which consists of small commercial cow calf and hay production operation s She completed her B .S. in Animal Sciences in 2008 and M.S. in Anima l Sciences with an emphasis in ruminant n utrition in 2010 from Auburn University. In summer 2010, she received a Graduate School Fellowship from UF and began a doctoral program in the area of forage managem ent under the direction of Dr. Lynn E. Sollenberger. She completed her Ph.D. program in fall 2013 with a major in Agronomy and a minor in Agricultural Education and Communication. Her professional goals are to use her training to develop research base d solutions and educational programs to address issues facing beef cattle producers across the USA.