1 IMPROVING FRUIT YIELD AND NUTRIENT MANAGEMENT IN TOMATO PRODUCTION BY USING GRAFTING By DESIRE DJIDONOU A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 2012 Desire Djidonou
3 To my wife Nadia and my daughter Tiffany, with love for being with me in this journey
4 ACKNOWLEDGMENTS I would like to acknowledge my major profess or Dr. Xin Zhao for her support, guidance, and providing an assistantship for my Ph.D. program. I would also like to thank the members of my committee, Dr. K aren Koch Dr. E ric Simonne Dr. A my Simonne Dr. J ohn Erickson, and Dr. S ixue Chen for their contr ibuting practical discussions and providing their viewpoints and suggestions on different problems and concerns inherent to this project I want to further thank Glenn Zalman with whom I started this project before his retirement in 2010. Special thanks to Bee Ling Poh and Mike Alligood. Without their technical expertise in irrigation system, my field experiments in Live Oak would not have been so successful. I would also like to thank the personnel and staff of the North Florida Research and Education Cent er Suwannee Valley in Live Oak, FL for their kind support in providing maintenance of my research pl ots I would like to thank Dr. Zotarelli for letting me use the root sampling and processing equipment I thank Dr Rebecca Darnell, Valerie Jones, for thei r help in the nitrate reductase activity and soluble proteins analysis. Special thanks to Ning Zhu and all others students in Dr Chen Lab for helping me in any way or another with the plant hormones analysis. I would like to thank Kim Cordasco, Elena Lon Kan Adrian Berry James Lee and Francisco Loayaza for their assistance with the determination of the nutritional fruit quality attributes conducted in this project I also thank Charles Wenjing, Carmen, Steffen, and Chris for the excellent teamwork we de veloped and other students, Kaylene Sattano Joshua Adkins Clint Hunnicutt, Shaun Teo Dakson Sanon, Silvia Marino for help ing with the tomato grafting and harvest. Thanks go to Kenneth Lopiano, Danielle Boree, Dr Lynda Young, and Dr George Ho c hmuth for their great help and suggestions with statistical analysis. I extend my gratitude to all Horticultural
5 Science s Department staff for the administrative support and especially to Jeanne Tucker for her constant help with any computer issue. I am grateful to the University of Florida IFAS Research Innovation Funds which supported this Research and also the Southern Region Sustainable Agriculture Research and Education (SARE) for the Graduate Student Grant I also extend my gratitude to the Graduate School for the Doctoral Student Support Award which mainly supported my tuition and fees for the Fall 2012 semester.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ .......... 10 LIST OF FIGURES ................................ ................................ ................................ ........ 14 ABSTRACT ................................ ................................ ................................ ................... 15 CHAPTER 1 GENERAL INTR ODUCTION ................................ ................................ .................. 18 Statement of Problem, Rationale, and Justification ................................ ................ 18 Goal, Objectives, and Hypotheses ................................ ................................ .......... 25 2 APPLICATION OF GRAFTING IN FRUIT TREE AND VEGETABLE PRODUCTION BEYOND DISEASE CONTROL: A COMPARATIVE REVIEW ...... 27 Introduction ................................ ................................ ................................ ............. 27 Purposes of Fruit Tree Grafting ................................ ................................ .............. 29 Vegetative Propagation ................................ ................................ .................... 29 Avoidance of Juvenility ................................ ................................ ..................... 30 Size Control ................................ ................................ ................................ ...... 31 Scion Shoot Growth and Yield ................................ ................................ .......... 31 Tolerance or Resi stance to Abiotic Stresses ................................ .................... 33 Purposes of Vegetables Grafting ................................ ................................ ............ 34 Tolerance to Abiotic Stresses ................................ ................................ ........... 34 Growth Promotion and Crop Productivity Enhancement ................................ .. 39 Grafting and Fruit Quality ................................ ................................ ........................ 40 Impact of Grafting on Fruit Tree Quality ................................ ........................... 41 Rootstocks and Vegetable Quality ................................ ................................ ... 45 Physiological Basis of Rootstock Scion Interactions ................................ ............... 48 Hormonal Influence in Rootstock Scion Interactions ................................ .............. 53 Grafting and Gene Expression ................................ ................................ ................ 55 Concluding Remarks and Future Research Prospect ................................ ............. 57 3 INTERSPERSPECIFIC ROOTSTOCKS ENHANCE TOMATO YIELD WITHOUT COMPROMISING FRUIT QUALITY UNDER GREENHOUSE CONDITIONS ................................ ................................ ................................ ......... 61 Introduction ................................ ................................ ................................ ............. 61 Materials and Methods ................................ ................................ ............................ 63 Tomato Grafting and Plantin g ................................ ................................ ........... 63 Tomato Yield Measurements ................................ ................................ ............ 65
7 Plant Growth and Nutrient Accumulation ................................ .......................... 66 Fruit Quality Assessment ................................ ................................ .................. 67 Statistical Analyses ................................ ................................ .......................... 68 Results ................................ ................................ ................................ .................... 69 Tomato Yield Components ................................ ................................ ............... 69 Plant Growth ................................ ................................ ................................ ..... 69 Plant Tissue Nutrient Concentrations ................................ ............................... 70 Plant Nutrient Accumulation ................................ ................................ ............. 71 Fruit Quality Assessment ................................ ................................ .................. 72 Discussion ................................ ................................ ................................ .............. 73 Conclusions ................................ ................................ ................................ ............ 76 4 ROOTSTOCK EFFECTS ON NITROGEN ASSIMILATION AND ENDOGENOUS HORMONE STATUS IN TOMATO PLANTS UNDER GREENHOUSE CONDITIONS ................................ ................................ ............... 85 Introduction ................................ ................................ ................................ ............. 85 Materials and Methods ................................ ................................ ............................ 88 Plant Material, Growth Conditions, and Experimental Design .......................... 88 Gas Exchange Measurements ................................ ................................ ......... 88 Determination of In Vivo Nitrate Reductase Activity (NRA) .............................. 89 Nitrate and Organic N Determination ................................ ............................... 90 Soluble Protein Measurement ................................ ................................ .......... 90 Determin ation of Amino Acid Composition and Quantification of Phytohormones ................................ ................................ ............................. 91 Analysis of Aboveground Plant Biomass, Yield, and Root Characteristics ....... 91 Statistical Analyses ................................ ................................ .......................... 92 Results ................................ ................................ ................................ .................... 93 Plant Biomass, Yield and Yield components, and Root Characteristics ........... 93 Leaf Gas Exchange and Photosynthetic Activities ................................ ........... 94 Nitrate Reductase (NR) Activity and Nitrate Concentrations in the Plant Tissues ................................ ................................ ................................ .......... 95 Concentrations of Organic Nitrogen, Amino Acids, and Soluble Proteins ........ 96 Auxin, Cytokinin, and Gibberellic Acid Levels in Leaf and Root Tissues .......... 97 Relationship between Plant Growth Performance and Physiological Processes ................................ ................................ ................................ ..... 98 Discussion ................................ ................................ ................................ .............. 98 Conclusions ................................ ................................ ................................ .......... 104 5 YIELD, IRRIGATION WATER AND NITROGEN USE EFFICIENCY OF FIELD GROWN GRAFTED TOMATO WITH DRIP IRRIGATION ................................ .... 114 Introduction ................................ ................................ ................................ ........... 114 Materials and Methods ................................ ................................ .......................... 116 Grafting and T ransplant Production ................................ ................................ 116 Field Production ................................ ................................ ............................. 117 Yield I rrigation W ater Use Efficiency, and N itrogen U se E fficiency ............... 119
8 Statistical Analyses ................................ ................................ ........................ 119 Results ................................ ................................ ................................ .................. 120 Seasonal Rainfall at the Experimental Site ................................ ..................... 120 Total and Marketable Fruit Yields ................................ ................................ ... 120 Yield Components ................................ ................................ .......................... 122 Irrigation Water Use E fficiency (iWUE) ................................ ........................... 123 Nitrogen Use Efficiency (NUE) ................................ ................................ ....... 124 Discussion ................................ ................................ ................................ ............ 125 Grafting Influence on Fruit Yields of Tomato under Field Conditions ............. 125 Seasonal Variation of the Irrigation and N Rate Effects ................................ 127 Grafting Influence on Irrigation Water and Nitrogen Use Efficiency ................ 128 Conclusions ................................ ................................ ................................ .......... 130 6 BIOMASS, NITROGEN ACCUMULATIO N, AND ROOT DISTRIBUTION OF GRAFTED TOMATO AS AFFECTED BY NITROGEN FERTILIZATION .............. 141 Introduction ................................ ................................ ................................ ........... 141 Materials and Methods ................................ ................................ .......................... 143 Experimental Site and Design ................................ ................................ ........ 143 Plant Growth, Nitrogen Uptake and Efficiency ................................ ................ 145 Root Analysis ................................ ................................ ................................ 145 Statistical Analyses ................................ ................................ ........................ 146 Results ................................ ................................ ................................ .................. 147 Leaf Area Index and Plant Biomass ................................ ............................... 147 N Accumulation and N Uptake Efficiency ................................ ....................... 148 Root Length Density Distribution ................................ ................................ .... 149 Discussion ................................ ................................ ................................ ............ 150 Grafting Influence on Plant Growth and Nitrogen Use Efficiency ................... 150 Grafting Influence on the Root Length Density Distribution ............................ 153 Conclusions ................................ ................................ ................................ .......... 155 7 ESTIMATING NITROGEN FERTILIZAT ION REQUIREMENT FOR GRAFTED TOMATO GROWN IN THE FIELD ................................ ................................ ....... 162 Introduction ................................ ................................ ................................ ........... 162 Material and Methods ................................ ................................ ........................... 165 Field Experiment ................................ ................................ ............................ 165 Marketable Yield Response Functions ................................ ........................... 166 Optimal Nitrogen Rate Determina tion ................................ ............................. 167 Estimation of Uncertainty around Optimum N Rate of Fertilizer ..................... 170 Results ................................ ................................ ................................ .................. 171 Tomato Marketable Fruit Yield Model ................................ ............................. 171 Optimal Nitrogen Rate and Confidence Intervals ................................ ........... 172 Maximum Attain able Yields Predicted by Each Model and Confidence Bands ................................ ................................ ................................ .......... 175 Discussion ................................ ................................ ................................ ............ 176 Conclusions ................................ ................................ ................................ .......... 181
9 8 FRUIT QUALITY OF FIELD GROWN TOMATO AS AFFECTED BY GRAFTING WITH INTERSPECIFIC ROOTSTOCKS ................................ .............................. 190 Introduction ................................ ................................ ................................ ........... 190 Materials and Methods ................................ ................................ .......................... 192 Fruit Sampling and Preparation ................................ ................................ ...... 192 Determination of pH, Soluble Solids Content (SSC), and Total Titratable Acidity (TTA) ................................ ................................ ............................... 193 Determination of Ascorbic Acid Content ................................ ......................... 193 Determination of Carotenoids ................................ ................................ ......... 194 Determination of Total Phenolic Content and Antioxidant Activity .................. 195 Statistical Analyses ................................ ................................ ........................ 197 Results ................................ ................................ ................................ .................. 197 Soluble Solids Contents, pH, and Total Titratable Acidity .............................. 197 Ascorbic Acid and Carotenoid Contents ................................ ......................... 198 Total Phenolic Content and Antioxidant Capacity ................................ ........... 199 Yield and Yield Components ................................ ................................ .......... 200 Discussion ................................ ................................ ................................ ............ 200 Conclusions ................................ ................................ ................................ .......... 205 9 ECONOMIC ANALYSIS OF GRAFTED TOMATO PRODUCTION IN SANDY SOILS IN FLORIDA ................................ ................................ .............................. 210 Introduction ................................ ................................ ................................ ........... 210 Materials and Methods ................................ ................................ .......................... 212 Field Production o f Fresh Market Tomato ................................ ...................... 212 Costs of Grafted and Non Grafted Transplant Production .............................. 213 Base Production Cost Model for Fresh Market Tomato in Field ..................... 213 Partial Budget Analysis of Grafted and Non Grafted Tomato Production ....... 214 Results and Discussion ................................ ................................ ......................... 215 Costs of Tomato Transplant Production ................................ ......................... 215 Costs of Field Production of Fresh Market Tomato ................................ ........ 217 Marketable Fruit Yield and Gross Returns ................................ ...................... 217 Partial Budget Analysis of Grafted Tomato Production ................................ .. 219 Co nclusions ................................ ................................ ................................ .......... 222 10 GENERAL CONCLUSIONS ................................ ................................ ................. 230 Review and Synthesis of Findings ................................ ................................ ........ 230 Future Work ................................ ................................ ................................ .......... 232 LIST OF REFERENCES ................................ ................................ ............................. 234 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 259
10 LIST OF TABLES Ta ble page 2 1 Comparative summary of key research findings on the use of grafting in fruit trees and vegetables ................................ ................................ .......................... 60 3 1 M arketable yield, number of fruit, and average weight of tomatoes from from non .......... 77 3 2 Marketable yield, number of fruit, and average weight of fruit from non .......................... 78 3 3 Leaf area and dry weight of oragans from non (FL) plants, self ................................ ................................ ....... 79 3 4 Nutrient concentrations in tissues from non plants, self rootstoc ks ................................ ................................ .................... 80 3 5 Nutrient concentrations in tissue from non plants, self on of rootstocks. ................................ ................................ ................................ 81 3 6 Nutrient accumulated in tissues from non plants, self on either of rootstocks. ................................ ................................ ....................... 82 3 7 Nutrient accumulated in tissues from non plants, self on either of rootstocks. ................................ ................................ ....................... 83 3 8 Quality attributes of fruit from non self scions on rootstocks ................................ ................................ ......................... 84 4 1 Fruit yields and aboveground biomass of grafted and non grafted tomato plants. ................................ ................................ ................................ ............... 106 4 2 Total root length (RL), average root diameter (AD), r oot surface area (RS) and root volume (RV), specific root length (SRL), specific root surface area (SRSA), root mass: root volume ratio (M:V) of grafted and non grafted plants 107
11 4 3 Net CO 2 assimilation rate on a leaf area basis (P n ), stomatal conductance (G s ), sub stomatal CO 2 concentration (C i ), transpiration rate (Tr), intrinsic water use efficiency (iWUE), and specific leaf area (SLA). ............................... 108 4 4 Tissue concentrations of organic N, amino acids, and soluble proteins of grafted and non .... 109 4 5 Auxin, cytokinin gibberellic acids concentrations, and auxin:cytokinin ratio (AUX:CTK) in leaf and root tissues of grafted and non tomato plants at 42 and 72 DAT. ................................ ................................ ...... 110 4 6 Pearson correla tion of fruit yield (YLD), aboveground biomass (BS), specific root length (SRL), nitrate reductase activity (NRA), organic nitrogen (OGN), and concentrations of nitrat e (NIT), soluble proteins (PRO) ............................. 111 5 1 Analysis of variance for effects of irrigation regime, nitrogen fertilization rate, and grafting combination on total tomato fruit yield, marketable fruit yield, number of marketable fruit, ave rage weight of marketable fruit ....................... 132 5 2 Main effects of irrigation, nitrogen fertilization rate, and grafting combination on total yield, marketable yield, and nitrogen use efficiency of tomato in the 2010 trial in Live Oak, FL. ................................ ................................ ................. 133 5 3 Total and marketable yields of tomato as influenced by interactions between irrigation regime, N rate, and grafting treatments in the 2011 trial in Live Oak, FL. ................................ ................................ ................................ .................... 134 5 4 Number of marketable tomao fruit per plant as influenced by the interaction between irrigation regime and N rate in the 2010 and 2011 trials in Live Oak, FL. ................................ ................................ ................................ .................... 135 5 5 Average marketable weight of tomato fruit from grafted and non grafted plants as influenced by interaction between irrigation regime and N rate in the 2011 trial in Live Oak, FL. ................................ ................................ ........... 136 5 6 Nitrogen use efficiency (kg kg 1 ) of grafted tomato plants as influenced by interaction between irrigation treatment and nitrogen rate (left) and interaction between nitrogen rates and grafting combination (right). ................ 137 6 1 Analysis of variance of the effects of nitrogen fertilization and grafting on leaf area index (LAI), total aboveground biomass, accumulated nitrogen (N acc ), nitrogen uptake efficiency (NUE), in the 2010 and 2011 fiel d trials ................ 1 56 6 2 Effects of nitrogen fertilization and grafting on leaf area index (LAI), aboveground biomass, N accumulation, and nitrogen uptake efficiency (NUE) of tomato plants during t he 2010 and 2011 field trials at Live Oak, FL. 157
12 6 3 Effects of nitrogen fertilization rate and grafting on tissue nitrogen concentrations of tomato plants during the 2010 and 2011 field tri als at Live Oak, FL. ................................ ................................ ................................ ............ 158 6 4 Analysis of variance of the effects of nitrogen rate, grafting, root sampling position, and soil depth on root length density of tomato plants during the 2010 and 2011 field trials at Live Oak, FL. ................................ ....................... 159 6 5 Root length density interaction effects between N rates and graft in 2010; and N rates and soil depth in 2011during the field trials at Live Oak, FL. ......... 160 6 6 Root length density interaction effects between grafting and soil depth; and sampling position and soil depth during the 2010 and 2011 field trials at Live Oak, FL. ................................ ................................ ................................ ............ 161 7 1 Model selection criterion for the 5 different types of models using the 2010 data. ................................ ................................ ................................ ................. 183 7 2 Model selection criterion for the 5 different t ypes of models using the 2011 data. ................................ ................................ ................................ ................. 184 7 3 Estimated optimum nitrogen rates and associated uncertainty bounds with the exponential function in the 2010 and 2011 experiments. ............................ 185 7 4 Maximum attainable yield values with the exponential function with the 2010 and 2011 experimental data. ................................ ................................ ............ 186 8 1 Quality attributes of tom ato fruits from non ......... 206 8 2 Moisture, ascorbic acid and carotenoid contents of tomato fruits from non ................... 207 8 3 Total phenolic content and antioxidant activity of tomato fruits from non ................... 208 8 4 Total yield, marketable yield, fruit number, and average fruit weight of non ................... 209 9 1 Estimated costs of production of grafted and non grafted tomato transplants .. 223 9 2 Estimated costs per acre needed to produce transplants of non grafted (FL/MU). ................................ ................................ ................................ ........... 224
13 9 3 tomatoes with raised bed polyethylene mulch system in the spring seasons of 2010 and 2011 in Live Oak, FL. ................................ ................................ .... 225 9 4 Average marketable tomato fruit yields, harvest related costs, gross returns for non and gross returns relative to non grafted treatment. ................................ ......... 227 9 5 Added costs and reduced returns, total negative effects, reduced costs and added returns, total positive effects, and additional net returns incurred by ............................ 228 9 6 Comparisons of estimated gross returns, costs of transplants, harvest, and other production operations and marketing, and total net returns between field production of grafted and non graf .......... 229
1 4 LIST OF FIGURES Figure page 4 1 In vivo activity of nitrate reductase in leaf (A) and root (B) tissues of graf ted and non ................ 112 4 2 Nitrate concentrations in leaf (A) and root (B) tissues of grafted and non 42, 72, and 114 DAT. .............................. 113 5 1 Cumulative rainfall in Live Oak, FL during the tomato field trials in 2010 and 2011. ................................ ................................ ................................ ................ 138 5 2 Irriga tion water use efficiency as influenced by interaction between irrigation treatments and nitrogen fertilization rates ................................ ......................... 139 7 1 Examples of model fits for the 2010 data including the model fits for grafted model fits for non .... 187 7 2 Measured (dot) and fitted (lines) yields of non grafted and grafted tomato plants grown under 50 and 100% irrigation regime with the five yield response models ................................ ................................ .............................. 189
15 Abstract of Dissertation Presented to the Graduate School of the Univer sity of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy IMPROVING FRUIT YIELD AND NUTRIENT MANAGEMENT IN TOMATO PRODUCTION BY USING GRAFTING By Desire Djidonou December 2012 Chair: Xin Zhao Majo r: Horticultural Science s In addition to managing soil borne diseases, grafting with vigorous interspecific hybrid rootstocks has been shown to improve fruit yield in tomato ( Solanum lycopersicum L. ) production due in part to enhanced nutrient and water up take. Interest in tomato grafting is emerging among vegetable growers in Florida. While using grafting as an effective tool for disease control, it is also of great interest to determine its potential for integration into the Best Management Practices (BMP ) program developed in Florida. Yet, systematic research is needed to examine the horticultural and economic performance of grafted plants in site specific tomato production systems in Florida sandy soils. Following the greenhouse study which showed tomato yield enhancement by grafting with interspecific rootstocks, a 2 year field experiment was conducted in Live Oak, Florida during the spring seasons of 20 10 and 20 11 to determine the effects of interspecific rootstocks on 1) tomato yields, irrigation water and nitrogen use efficiency, 2) optimum N fertilization rate to maximize marketable fruit yield, 3) cost effectiveness of grafted tomato production, and 4) fruit quality attributes of
16 grafted tomato plants. A determinate tomato cultivar Florida 47 was gra fted onto each of S. lycopersicum S. habrochaites S. Knapp & D.M. Spooner). Non Plants were grown in twelve combinations of two irrigation re gimes (50 and 100%) and six nitrogen rates (56, 112 1 6 8, 224, 28 0 and 3 36 kg N ha 1 ) with 56 kg ha 1 as preplant application, using a split plot experimental design with 4 replications. In both seasons, grafting with the two rootstocks significantly impr oved the yields of Florida 47 Specifically in the 2010 season, the increase of total and marketable fruit yields by Beaufort reached 24% and 28%, respectively, in comparison with non grafted Florida 47 Similarly, Multifort increased total and mar ketable fruit yield by 29% and 33%, respectively. The two grafted combinations also showed significantly greater irrigation and nitrogen use efficiencies based on marketable yields as compared to the non grafted treatment. However, the minimum N rates requ ired to maximize the marketable grafted treatment. E stimated minimum N fertilization rates to reach maximum marketable fruit yields ranged from 211 to 240 kg N ha 1 for th e grafted plants vs. 163 to 180 kg N ha 1 f or the non grafted plants depending on the irrigation regime. The corresponding maximum marketable fruit yields were also higher in grafted tomato production ranging from 5 6 to 71 Mg ha 1 in comparison with 4 3 to 5 4 Mg ha 1 for non grafted plants. Estimated costs of grafted and non grafted transplants were $0.67 and $0.15 per plant, respectively; resulting in an additional cost of $ 3,015.63 per acre for using grafted transplants as compared to non grafted plants. Partial budget analyses performed in
17 this study showed that additional net returns associated with using grafted transplants ranged from $ 263.28 to $ 2,461.82 per acre, depending on the seasons and rootstocks. Grafted and non grafted plants did not demonst rate consistent differences in fruit quality attributes measured in this study although the fruit moisture content was increased due to grafting with interspefic rootstocks as observed in the field experiments. Levels of nitrate reductase activity, nitrate organic nitrogen, soluble proteins, and amino acids as well as concentrations of auxin, cytokinins, and gibberellic acids in leaf and root tissues measured at three growth stages in the greenhouse experiment did not reveal any major consistent difference s between the grafted and non grafted tomato plants.
18 CHAPTER 1 GENERAL INTRODUCTION Statement of Problem, Rationale, and Justification Florida offers a great diversity of agroecosystems that allow a wide range of production systems for several horticultu ral crops. The vegetable industry in Florida encompasses over 50 diverse crops grown over three seasons and on different soil types with different irrigation methods (Cantliffe et al., 2006) Specifically, Florida is the leading producer of fresh market tomatoes ( Solanum lycopersicum L.) accounting for approximately 51 % of the national fresh market tomato production, with an annual total cash value of ov er $4 00 millions during the 20 09 1 0 season (Usda, 2010) Grow i n g predominantly on sandy soil s with ver y low intrinsic water and nutrient retention capacities, commercial tomato producers in Florida rely heavily on intensive management with high inputs of fertilizers and irrigation water for achieving profitable yields (Hartz and Hochmuth, 1996; Hochmuth, 1992; Locascio, 2005) Frequent and excessive N fertilizer application when combined with high volume of irrigated water and occurrence of intense rainfall events under these sandy soil conditions may often result in N leaching below the active ro ot zone and may lead to various environmental issues including pollution of ground water (Prakash et al., 1999; Vzquez et al., 2006; Zotarelli et al., 2009b) With the increasing cost of fertilizers and more rigorous environmental regulations, there has been a need for implementation of more sustainable horticultural practices that allow for economical production while simultaneously ensuring protection of the environment (Obreza and Sartain, 2010) Therefore in Florida for example, the Best Management Practice (BMP) program has been developed to provide a variety of site specific practices that can optimize nutrient management such as fertigation with
19 drip irrigation, soil moisture sensing, crop rota tion, use of cover crops, and controlled release fertilizers (Fdacs, 2005) In essence, the overall goal of the implementation of any BMP program is to keep water and nutrien ts in the root zones for a more efficient uptake and use by the growing plants. However, various studies showed that concentrations of nitrate were still higher beyond the active vegetable root zone even with BMP practices (Simonne et al., 2006; Simonne et al., 2003; Zotarelli et al., 2008) In a most recent review on the current knowledge, gaps and future actions regarding successful implementation of BPM program for vegetable production, Simonne et al. (2010b) discussed the potential role of br eeding as a whole and vegetable grafting in particular on improving vegetable crop nutrient use efficiency. However, to date, there is limited information about grafted tomato production in relation to nutrient management especially in site specific field cultivation systems in the U.S. More studies are needed to assess the potential of grafting to enhance water and N use efficiency in tomato production and the feasibility of its implementation in BM P Since its beginning in the 1920 s in Japan and Korea, ve getable grafting is now widely used in controlling a number of soil borne diseases such as fusarium wilt, verticillium wilt, sou thern blight, and bacteria wilt, as well as root knot nematodes in many intensive production system s of solanaceous and cucurbit aceous vegetable s in Asia, Europe and the Mediterranean (Lee, 1994; Lee et al., 2010) Grafting with certain rootstocks has also been shown to imp rove plant tolerance to abiotic stresses such as high salt and low temperature (Fernandez Garcia et al., 2004b; Schwarz et al., 2010) Moreover, grafted plants may demonstrate enhanced vigor and fruit yield even under low pressure of biotic and abiotic stresses. In the case of tomato production, grafted
20 plants could exhibit an increase of marketable yield by 20% to 62% as compared with non grafted plants, depending upon the scion rootstock combinations and production conditions (Di Gioia et al., 2010; Lee an d Oda, 2003; Leonardi and Giuffrida, 2006; Pogonyi et al., 2005) In the U.S., interest in vegetable grafting is growing recently and grafted tomato research has been carried out under different production conditions with a focus on using resistant rootst ocks to manage soil borne disease (Lopez Perez et al., 2006; Mcavoy et al., 2012; Rivard and Louws, 2008; Rivard et al., 2010a) and root knot nematodes (Barrett et al., 2012a; Bausher, 2009) However, research information is scarce with respect to the response of grafted tomato plants to different irrigation regimes and nitrogen fertilization rates in comparison with non grafted plants, especially under Florida field conditions Research suggest s that fertilizer inputs may be reduced by about 33 50% in grafted vegetable production in contrast to the standard recommendation for non grafted plants owing to the improved nutrient uptake and growth vigor of gr afted plants (Colla et al., 2011; Lee and Oda, 2003; Salehi Mohammadi et al., 2009) Particularly with the sandy soil in Florida, t here is a need to establish the Crop Nutritional Requirement for N for grafted tomat o plants for drip irrigated tomato production under field condition. Despite the increasing interest in using grafted transplants for field tomato production in disease management, t he majority of users of grafted seedlings in the U.S. are currently green house hydroponic tomato growers (King et al., 2010; Kubota et al., 2008; Lee et al., 2010) The relative high costs of grafted transplants in comparison to non grafted transplants remains as the major barrier for wi de adoption of grafting in
21 field tomato production. Tomato growers are often concerned that the high investment associated with grafted transplants may not be offset by the increase in fruit yield to generate enough net return to grafted tomato production economically feasible (Ku bota et al., 2008; Lee et al., 2010) In addition to the costs of rootstock seeds, grafted transplant production requires other direct costs associated with supplies and materials as well as making the grafts, which ultimately increase the costs of grafte d transplants as opposed to the non grafted transplants (Barrett et al., 2012b; Rivard et al., 2010b) For example, estimated prices for grafted tomato transplants ranged from $0.59 to $1.88 as compared to $0.13 to $0.76 for non grafted plants in two different transplant production facilities in the U.S. (Rivard et al., 2010b) In general, efficient production practices can only be fully integra ted into endogenous agricultural management operations when growers can also sustain some economic benefits by using the techniques (Baggs et al., 2000) Hence, in addition to further evaluating the performance of different tomato rootstocks in improving plant resistance to soil borne and crop yields, it is also critical to assess the economic feasibility of the use of grafted tomato transplants as a viable component of the existing commercial to mato production system even beyond disease control. From a physiological perspective, the underlying mechanism by which the vigorous root system of certain rootstocks significantly influence plant vigor and fruit yield in vegetable crops has not been fully understood. In recent decades, evidence is available to suggest several hypotheses, including the involvement of nutrient and water movement (Pulgar et al., 2000; Ruiz et al., 1997) changes in endogenous growth hormone homeostasis (Seong et al., 2003) and molecules signaling that involves
22 translocation of small molecules like mRNAs across the graft union (Kudo and Harada, 2007; Omid et al., 2007) In effect, grafted plant can be perceived as an integrative system of a scion an d a rootstock in which there is a l ong distance movement of water, nutrients, plant hormones, sugars and assimilates amino acids proteins, and small RNAs through phloem and xylem and as such these processes are believed to fundamentally influence the pat tern and scale of the growth and development processes of the grafted plant. Specifically, studies by Ruiz and Romero (1999) demonstrated that grafted melon had in leaf tissues, lower concentrations of nitrate, higher levels of nitrate reductase (NR) activity, and lower contents of total free amino acids and soluble proteins compared to the non grafted control. Higher nitrate reductase and nitrite reductase activities were also found in leaves of grafted watermelon plants (Pulg ar et al., 2000) According to Ruiz et al. (2006a) the enhancement of nitrogen u se efficiency in grafted tobacco as compared to the non grafted plants is related to higher nitrate reductase activity and assimilation. In addition to the improved nitrogen assimilation as a result of grafting with specific rootstocks, enhanced endogenous hormone production has been proposed as another underlying basis for the enhanced growth and yield of grafted vegetables (Davis et al., 2008b) Different growth hormones including auxin, cytokinins, and gibberelli c acids are well known for their roles in r egulating different aspects of plant growth and development. Particularly with fruit trees, the relative concentrations of these growth hormones, e.g., auxin:cytokinin ratio, have been shown as one of the key physiological mechanisms by which dwarfing or v igorous rootstocks modify scion architecture of the grafted fruit trees. Specifically, it is shown on apple that dwarfing rootstocks control scion vigor by reducing the downward transport of indole 3
23 acetic acid (IAA) from the scion to the root which resul ts in a decrease in root growth and/or cytokinin biosynthesis and therefore limiting the supply of root produced cytokinins to the scion in the xylem vasculature (Lockard and Sch neider, 1981) It is likely that similar changes in hormonal balance may also exist and affect the rootstock scion interaction in the grafted vegetable plant including tomato Both tomato hybrid cultivars and interspecific tomato hybrids are used as roots tock s in grafted tomato production (King et al., 2010) In particular, the interspecific hybrid rootstocks are known for their vigorous growth characteristics and have been shown to increa se crop productivity in addition to their resistance to a variety of soil borne diseases. However, there is a lack of comprehensive understanding of the physiological modifications in grafted tomato plants and underlying mechanisms as a result of using the se vigorous rootstocks. Given the physiological changes that may occur in grafted plants relative to non grafted plants, it is possible that some of the quality attributes of fruit may also be affected. It is suggested that some fruit quality attributes i n grafted plants may be influenced by the rootstock as a result of the translocation of the metabolites inherent to the fruit quality from the root to the scion through xylem and/or change s in the physiological processes of the scion (Lee, 1994; Rouphael et al., 2010) In order to ensure successful marketing, grafted tomatoes must have both enhanced marketable fruit yields and premium fruit quality. Depending on the scion rootstock combinations used, previous studies ha ve yield mixed results regarding the change of fruit quality in grafted vegetables. While quality attributes of tomato fruit such as titratable acidity and soluble solids content did not show any significant difference s between grafted and self
24 rooted toma to plants (Khah et al., 2006) g rafting with certain rootstocks has been demonstrated in some instances to either decrease or increase the levels of lycopene, carotene, vitamin C, and antioxidant activity in t omato fruit (Davis et al., 2008a; Dorais et al., 2008; Fernandez Garcia et al., 2004a) .Melon plants grafted onto different Cucurbita spp. rootstocks exhibited a remarkable deterioration of taste and texture in some rootstock scion combinations (Traka Mavrona et al., 2000) In general, fruit sensory and nutritional quality attributes are complex and can be affected by various biotic and abiotic factors such as cu ltivar, maturity and ripeness at harvest, production practices, and environmental conditions before and after harvest (Dorais et al., 2008; Foolad, 2007) With respect to the grafted plants, Davis et al. (2008a) pointed ou t that the type of scion variety used determines primarily the size and quality of fruit in grafted plants, but the rootstock effects can drastically influence these characteristics as well. As many of the rootstocks used are often directly developed from wild relatives and related, although different species, it might not be surprising that some rootstocks show negative impacts on the quality attributes of the scion. It is often suggested that the rootstock/scion combinations should be carefully selected f or optimal fruit qualities and there is a need for more comparative studies of nutritional quality attributes as more rootstocks are becoming more available. This research project was conducted in response to the emerging interest in grafted tomato product ion among vegetable growers in Florida. A series of greenhouse and open field experiments were carried out to address the following questions: 1) whether the tomato growth and yield performance will be improved by using new interspecific rootstocks; 2) wha t would be the requirement for N fertilization to optimize
25 the marketable yield of grafted tomato production in sandy soils in Florida; 3) whether the use of grafted transplants in commercial tomato production is economically feasible considering the incre ase in both production costs and fruit yield improvement; 4) how grafting with interspecific rootstocks affect nitrogen and carbon assimilation and endogenous growth hormone status in tomato plants; and 5) how the fruit quality attributes may change as a r esult of grafting with vigorous rootstocks Goal, Objectives and Hypotheses T he overall goal of this research project was to determine the potential of grafting as an economically viable practice for integration into the BMP program for tomato production in sandy soils in Florida. Specific objectives are: 1. Assess the performance of grafted tomato plants onto different interspecific rootstocks in terms of their potential to accumulate biomass, improve nutrient accumulation and enhance marketable yield unde r greenhouse conditions. It is hypothesized that biomass accumulation, nutrient uptake and fruit yield can be improved by grafting tomato plants with vigorous rootstocks (Chapter 3) ; 2. Explore the effect of rootstocks on nitrogen metabolism and assimilation and hormone status in tomato plants under greenhouse conditions. It is hypothesized that grafted tomato plants with vigorous interspecific rootstocks have increased levels of nitrate reductase activity, nitrogen assimilation, and growth hormone production (Chapter 4); 3. Determine the influence of grafting with vigorous rootstocks on fruit yield, i rrigation water use efficiency, and nitrogen use efficiency under different N application rates and irrigation regimes in field tomato production in sandy soils Th e u nderlying hypothesis is that grafting with vigorous rootstocks increases fruit yield, i rrigation water use efficiency, and nitrogen use efficiency as compared to non grafted plants (Chapter 5); 4. Compare the growth, nitrogen concentration and accumulation N uptake efficiency, and root distribution of grafted and non grafted tomato in response to N fertilization. It is hypothesized that grafting with vigorous rootstocks increase biomass and nitrogen accumulation and also enhance root length density distrib ution as compared to non grafted tomato plants (Chapter 6 ); 5. Examine the Crop Nutritional Requirement for N for grafted tomato plants for drip irrigated tomato production under field condition in sandy soils. It is hypothesized
26 the N application rate requir ed to maxim ize fruit yield is different from the current recommended rate when tomato plants are grafted with vigorous interspecific rootstocks (Chapter 7 ); 6. Determine whether grafting with interspecific rootstocks can alter the quality attributes of tomato fruit. It is hypothesized that grafting does not adversely affect the major intrinsic quality measurements of fresh marketable tomato fruit such as pH, soluble solids content and titratable acidity (including their ratio), and carotenoid, vitamin C and ph enolic contents (Chapter 8 ) ; 7. Assess the economic feasibility for using grafted transplants in tomato production in Florida sandy soils to improve water and N fertilizer use efficiency It is hypothesized that integrating the grafting practice into the curr ent tomato production system in Florida will increase the net return despite the higher cost of grafted tomato transplants (Chapter 9).
27 CHAPTER 2 APPLICATION OF GRAFT ING IN FRUIT TREE AN D VEGETABLE PRODUCTI ON BEYOND DISEASE CONTR OL: A COMPARATIVE RE VIEW Introduction The technique of plant grafting is to insert a desirable scion into a resistant rootstock in a way that results in a graft combination that achieves physical union and grows as a single plant (Janick, 1986) Grafting involves the development of a composite genetic system by fusing two distinct genotypes, each of which conserves its own genetic make up throughout the life cycle of the resulting grafted plant (Mudge et al., 2009) Occasionally and exclusively in the fruit tree grafting, a third genetically distinct component, an interstock or interstem, is grafted between the rootstock and scion. It has been used with different fruit trees to modulate the tree size, fruit production and quality, and the aging of the tree (Gil Izquierdo et al., 2004) Once the two or three component s of the graft are placed in close physical proximity, development of the graft union begins. This merging involves the proliferation of new parenchyma cells from both root stock and scion producing the callus tissue that soon intermingles and interlocks. T hese cells fill the space between the two components and connect the scion and the rootstock together (Hartmann et al., 2002) Success of grafting has been credited to the deposition and subsequent polymerization of cell wall material that occurs in response to the wounding inherent to th e grafting process (Pina and Errea, 2005) According to Fernandez Garcia et al. (2004a) a successful graft union requires the formation of new connections between vascular strands at the callus graft interface via differentiation and lignificat ion. Failure of a graft union to successfully develop may be due to: a lack of cellular recognition, the growing stage of the respective plants, interference of the wounding response or growth regulators, or an
28 unfavorable grafting environment (Andrews and Marquez, 1993; Davis et al., 2008a) An experiment t hat used a non destructive method to assess the development of hydraulic connections in the graft union of tomato showed that the graft union is completely functional 6 8 days after grafting (Turquois and Malone, 1996) Originally, grafting was a technique of plant propagation and selection commonly reserved for perennial horticultural crops such as tree fruits and nuts that do not root easily from cuttings or can not be propagated clonally from seed, such as apples ( Ma lus domestica) citrus ( Citrus sp ) grapes ( Vitis sp ), pears ( Pyrus sp ) and peaches ( Prunus persica ) In essence, the graft is made during the dormant season, and healing occurs as the plant begins to grow in the spring. Within this group of perennial plan ts, the purposes of grafting with rootstocks are mainly to propagate scions of preferred cultivars, improve fruit tree tolerance to environmental stresses, and to control tree size (Webster and Wertheim, 2003) However, this practice of grafting plant parts is now used in production of certain vegetables and has become a widespread practice in many parts of the world. This method provides a n innovative tool to reduce the incidence of infection by soilborne pathogens such as Fusarium (Besri and R abat, 2005; Edelstein et al., 1999; Jifon et al., 2008) aid tolerance to environmental stresses such as low temperature, salt, drought and flooding (Colla et al., 2006a; Estan et al., 2005; Rivero et al., 2003b; Ro mero et al., 1997; Yetisir et al., 2006) enhance water and nutrient uptake, and to increase plant vigor (Lee, 1994; Lee and Oda, 2003; Leonardi and Giuffrida, 2006; Oda, 1999; Pulgar et al., 2000) Numerous specie s of vegetables crops notably in the Curcubitaceae and Solanaceae s uch as watermelons ( Citrullus lanatus (Thu n b.) melon ( Cucumis melo L.), cucumber ( Cucumis sativus L.) tomato
29 ( Solanum lycopersicon L.), eggplant ( Solanum melongena L.) and pepper ( Capsic um annuum L.) are commonly grafted to various rootstocks. Over the years, many studies have been conducted to evaluate horticultural aspects of grafting technology These findings were synthesized in exhaustive literature review s which explain, in anatomic al, nutritional, hormonal, or other physiological terms, rootstock effects on scion performance when grafting vegetables (Davis et al., 2008a; King et al., 2010; Lee, 1994; Lee and Oda, 2003; Lee et al., 2010) and f ruit trees (Castle, 1995; Mudge et al., 2009; Storey and Walker, 1999) This review analyzes the current state of knowledge on fruit tree grafting in comparison with the use of vegetable grafting and synthesize info rmation on purposes of grafting in each group of crops, effect of grafting on the quality of fruit in each group, physiological aspects and hormonal influence of rootstock scion interaction, and change in gene expression pattern after grafting. Purposes of F ruit T ree G rafting The art of grafting with specific rootstocks has been extensive ly use d on a wide range of fruit trees to achieve various horticultural and biological goals. Many uses of grafting of fruit tree include principally the vegetative propaga tion, avoidance of juvenility, size control, resistance or tolerance to biotic and abiotic stresses and to some extent fruit quality. Vegetative P ropagation Various r ootstocks were, and still are, used primarily as an aid to scion propagation. In various f ruit tree species, m any scion cultivars of horticultural values are extremely difficult to propagate clonally on their own roots and rootstocks offer a simple method for multiplying these cultivars (Mudge et al., 2009; Webster, 1997) However,
30 Webster, (1997) mentioned that these traditional p ropagation methods do not ensure disease free and healthy plants and they also depend on the season Avoidance of J uvenility In the fruit tree production, growers usually overcome a delay in flowering of a juvenile plant by grafting a mature scion tree ont o a seedling or mature rootstock, in that a mature scion always maintains its flowering state even after a grafting onto a juvenile rootstock (Mudge et al., 2009) Consequently, as a desirable characteris tic conferred by the grafting technique the resulting grafted plant starts fruiting at an earlier age Okudai (1980) observed that c itrus hybrid seedlings grafted onto satsuma mandarin ( Citrus unshiu ) interstocks or onto trifoliate orange ( Poncirus trifoliata L. Raf.) rootstocks formed flowers two to three years after grafting while Yoshida (1980) reported that citrus hybrid seedlings grown from germination witho ut grafting onto rootstock started flowering seven years after germination. Similar result was observed by Mitani (2008) with hybrid citrus seedlings graf ted onto shiikuwasha as rootstock It was concluded that grafting accelerated the onset of flowering of citrus hybrid seedlings and therefore is an effective method to shorten the juvenile period. It is reported that scions of apple grafted on 'M.9' roots tocks or interstocks generally ripen up to 1 week earlier than the same scion grafted on other more vigorous rootstocks and trees on 'M.27' usually ripen earlier than trees on 'M.26' in part due to difference in the vigor potential (Webster and Wertheim, 2003) In general, trees gro wn on less vigorous rootstocks (dwarfing rootstocks) tend to produce precociously and consistently from season to season (Webster, 19 94) In addition, r ootstocks may also influence floral precocity indirectly by their effects on scion branching This is particularly noted with rootstocks which induce the abundant
31 production of lateral branches (feathers) on the scion; feathered trees p roduce flowers more precociously than unfeathered trees (Webster, 1997) Size C ontrol Grafting is often used to control size of the grafted plants. As a result, certain rootstocks can induce dwarfing or vigor of the scion cultivar. For example, in apple, a same scion cultivar grafted onto various size controlling rootstocks can result in trees ranging from 2 to 10 m in height. The pur pose of d warfing rootstocks is to produce smaller trees that facilitate harvesting, pruning and high density planting. Examples of interspecific scion/stock grafting which result in dwarfing are pear on quince and orange ( Citrus sinensis ) on trifoliate ora nge ( Poncirus trifolia ) (Mudge et al., 2009) T he mechanism by which the rootstock exerts its size control on the scion is not well understood. It is h ypothes ized that reductions in nutrient and water movement and changes in hormone concentrat ion may in part explain the dwarfing (Jensen et al., 2003) Scion Shoot Growth and Y ield In addition to shortening time to first flowering f rom planting and increas ing flower number and as a direct effect of size control, certain rootstocks impart the benefits of reduced vegetative vigor, resulting in smaller trees that are easier to handle Research on rootstock scion interactions demonstrate d that rootstock s ha d more influence than scion on tree weight and growth rate in young apple trees but scion more strongly affected duration of growth (Tworkoski and Miller, 2007) In order to plant at high densities an d still take advantage of the strength of seedling rootstocks, rootstocks that induce dwarfing of the scion are generally preferred, whereby a dwarf interstem is placed betw een the scion and the rootstock A study by Di Vaio (2009) on apple
32 involving the dwarfing rootstocks used as interstock (M.27 and M.9) showed that the two dwarfing rootstocks induced a decrease in the tree size and the amount of pruning on mature trees of Annurca Rossa del Sud compared to the control trees (non grafted Annurca Rossa del Sud). It is reported that rootstocks influence on the vigor and size of scion tree s may be achieved in one or multiple ways including : delaying sc ion vegetative budburst and the onset of shoot extension in the spring; influencing rates of shoot extension throughout the season; affecting the timing of seasonal termination of extension shoot growth ; influencing branching habit; and channeling more of the tree s assimilates and minerals into fruit than into shoot growth (Webster, 1997) In general, it is observed that d warfing roo tstocks which produce smaller trees usually produce, at maturity, yields per tree less than those from larger trees on intermediate or vigorous rootstocks. However, the smaller tree stature allows trees to be planted at higher densities, thereby enabling m uch higher yields to be attained per unit of planted area (Webster, 1997) The performance of citrus rootstocks has been evaluated using rootstock seedlings and grafted plants As is the case in apple, rootstocks with less vigor enhancement on the scion appear to be of higher horticultural value. Indeed, it is observed that rusk citrange, a rootstock of moderate vigor, produced smalle r trees and better yield, fruit quality, and economic returns than Milam lemon, a vigorous rootstock (Wheaton et al. 1995) In the grape vine production, the e ffect of rootstocks on scion yield and vigor has also been well documented with m ost results show ing that rootstock significantly affects scion vigor and yields. As shown on apple and citrus, it is demonstrated that yields are
33 in general negative ly correlated with vine vigor (Parejo et al., 1993; Wolf and Pool, 1988) Similar pattern was also found by Ferree et al. (1996) with the graft as compared to the own rooted plants These results suggest that the e ffect of rootstocks on scion vigor and yield is specific to scion/rootstock combinations. It has also been report ed that rootstocks may induce ne gative or indifferent effects on scion vigor and yields. For example, significant effect on yields when compared to yields from the own rooted vines (Boselli et al., 19 92) According to Reynolds and Wardle (2001) yield per vine, clusters per vine, cluster weight and berry weight were not affected by r ootstocks. Tolerance or Resistance to Abiotic S tress es Certain rootstocks have been selected for tolerance or resistance to abiotic stresses such as salinity, wet or drought soils, alkaline or acidic soils, and extreme low temperature. Garcia Sanchez (2002) investigated effects of salinity on the growth and net gas exchange of CO 2 and water vapor in leaves of 2 year mandarin [( Citrus reticulata Blanco) x ( Citrus paradisi Macf. x C. reticulata )] trees grafted on either Cleopatra mandarin ( C. reticulata ) or Carrizo citrange ( Citrus sinensis L. Osb. x Poncirus tri foliata L.) rootstocks Cleopatra roots accumulated higher concentrations of Cl and Na + than Carrizo roots but and Na + and had higher CO 2 exchange rates than those on Carrizo. The greater salt resistance of rootstocks, such as Rangpur lime and Cleopatra mandarin, is associated with their capacity to reduce the accumulation of Cl in leaves by the process of Cl exclusion, e.g. regulation of root uptake and unloading of Cl at the root symplast/xylem inter face. Other rootstocks, like trifoliate orange, present a potential for limiting Na +
34 transport to the shoot at low salinity, e.g. reabsorption of Na + from the xylem stream in the root and basal stem (Storey and Walker, 1999) Syvertsen (1988) found no effect of salinity on gas exchange parameters of even though foliar concentrations of Cl were as high as 400 mM. More over, Walker et al. (2002) identified as salt tolerant rootstock interspecific r ootstocks from V. berlandieri x V. rupestris were considered to be d rought tolerant (Ezzahouani and Williams, 1995) In addition, it was observed that grafted plants using V. vinifera as rootstock maintained high rates of net CO 2 assimilation after 14 days of water deficit. The higher rate of photosynthesis was associated with significantly higher stomatal conductance and an increase in water use efficiency was also observed (Iacono et al., 1998) It is also reported that on apple, many cold tolerant selection of rootstocks have become available which mainly provide the roots and the trunks of the rootstock with greater tolerance to very low temperatures (Webster and Wertheim, 2003) Purposes of V egetables G rafting As in the case with fruits tree production described above certain purpos es for utilizing grafting techniques in vegetables are similar to those of fruit tree grafting and include tolerance to abiotic stresses, plant vigor, and yield increase, in addition to disease control which is the focus of this review. Tolerance to Abioti c S tress es Similar to fruit tree grafting, grafting techniques have been widely and successfully used to overcome abiotic stresses such as salinity, temperature extremes (cold and heat), drought and excessive soil moisture on numerous species of vegetable crops A
35 comprehensive review of the use of grafting to cope with these different abiotic stresses is recently presented by Schwarz et al. (2010) 1. Salinity: S alinity results in excessive accumulation of Na + and Cl in the plant, which induces a wide range of physiological and biochemical deteriorations that affect plant growth and significantly reduce yield. Salt stress can alter the water relations within the plants, as a result of osmotic effects due to specific ionic imbalance or energy availability inherent to carbohydrate concentrations which ultimately lead to inhibition of crop growth (Lazof and Bernstein, 1998) As shown on fruit tree grafting, r ootstocks are also capable of reduc ing the accumulation of Cl and Na + in grafted scion leaves. This is attributed to the exclusion or decreased absorption of Cl by the roots and replacement of total Na + by total K + in the leaves similar to the mechanism described by Storey and Walker (1999) (Davis et al., 2008a; Martinez Rodriguez et al., 2008; Rivero et al., 2003a) More specifically, g rafting tomato scions with tolerant rootstock genotypes has been highly effective at increasing significantly plant growth parameters (Fernandez Garcia et al., 2004b) With the us i as rootstocks under high levels of NaCl, Estan et al. (2005) observed an increase of up to 80% in yields compared to non grafted or self grafted tomato contro ls In addition, Yang et al. (2006) demonstrate d that grafted cucumber plants show greater net photosynthesis, stomatal conductance, and intercellular CO 2 concentrations under NaCl stress conditions than self rooted plants. It was suggested that greater salt tolerance of grafted watermelon is inherent to higher peroxidase activity and lower superoxide dismutase activity in the shoot (Liu et al., 2004a, b)
36 2. Temperature: Studies have also attempted to evaluate the influence of vegetable graftin g as a tool to mitigate adverse effects of extreme temperatures on crop performance. Abdelmageed et al. (2004) observed that tomatoes grafted with heat tolerant rootstocks showed increased vegetative growth and reduced chlorophyll fluorescence under heat stress T omato plants grafted to a more vigorous rootstock show greater resistance to thermal fluctuations, resulting in greater plant development and growth (Rivero et al., 2003b) These authors concluded that this advantage may be due to better H 2 O 2 accumulation in the different plant tissues, as related to a stronger enzymatic activity of glutathione peroxidase (GPX) and catala se (CAT) or of optimal acticity of the ascorbate/glutathione oxidation/reduction cycle. Other studies also demonstrated the use of grafting technique with specific rootstocks as a means to confer tolerance or resistance to adverse effects of cold temperatu re on plant growth and development. Gao et al. (2008) demonstrated that the increase in chilling injury index and electrolyte leakage rate was relatively lower in grafted than own rooted eggplant seedlings; indicating that grafted eggplant plants onto highl y cold resistant rootstocks could improve the tolerance of scion to low temperature. Vigh et al. (1985) and later Bulder et al. (1991) on cucumber suggested that the ability of the rootstock to tolerate cold temperatures is based upon lipid differ ences in the membranes in the roots of these genotypes Venema et al. (2008) examined the impact of grafting cultivated tomato onto the rootstock of a cold tolerant, high altitude accession of a related wild species and concluded that Solanum habrochaites LA 1777 provides a valuable germplasm pool to improv e the low temperature tolerance of existing commercial tomato rootstocks. In addition, Solanum lycopersicum x S. habrochaites rootstocks
37 provide tolerance to low soil temperatures (10 o C to 13 o C) for their grafted tomato scions, while eggplants grafted onto S. integrifolium x S. melongena rootstocks grew better at lower soil temperatures (18 o C to 21 o C) than non grafted plants (Okimura et al., 1986) While more study is still warranted to fully understand important physiological and molecular mechanisms that underlie the positive effect of this rootstock on the greater performance of the tomato scion at suboptimal temperatures, it is indicated that t he rootstocks used in this study were able to maintain absorption and upward movement of water and nutrients in the scion, a less reduced supply of root borne phytohormones to the scions. But the mechanisms that allow for these changes and its modulation a re not well understood. Under low temperature conditions, grafted cucumber showed higher net photosynthesis, stomatal conductance, and intercellular CO 2 concentrations than own rooted plants (Ahn et al., 1999; Zhang, 2002) F igleaf gourd ( Curcubita ficifolia Bouche) and bur c ucumber ( Sicos angulatus L.) are used as rootstocks for cucumber owing to their tolerance to low temperature (Venema et al., 2008) These rootstocks improve vegetative growth and early yield at suboptimal temperatur es and under conditions where only the roots are subjected to chilling (8 o C) temperatures The physiological basis underlying the higher level of low temperature tolerance of figleaf gourd rootstocks is related to the maintenance of a higher absorption ra te of water and nutrients and a higher cytokinin synthesis which stimulates root meristem activity and translocation of photosynthates to the roots 3. Flooding: Most vegetables are highly sensitive to flooding and genetic varia bility with respect to intri nsic tolerance or resistance is limited, particularly in tomato. In general, damage to vegetables by flooding is due to the reduction of oxygen in the root
38 zone which inhibits aerobic processes. Flooded plants accumulate endogenous ethylene that causes dam age to the plants (Drew, 1997) Low oxygen levels stimulate an increased production of an ethylene precursor, 1 aminocyclopropane 1 carboxylic acid (ACC), in the roots. Grafting has also been utilized in order to reduce the effects of flooding in areas where a wet season may occur as is the case on tomato (Black et al., 2003) 4. Water deficit: Water deficit is one of the major environmental constraints that limit plant growth and development. Resistance to water deficit is associated with both the maintenance of relatively high photosynthesis at a low leaf water potential, and maintenance of high leaf water potential in response to decrease in soil water potential (Xu and Ishii, 1996) Weng (2000) observed that the tomato plants grafted onto S olanum mammosum rootstock had a higher leaf water potential, net photosynthetic rate and leaf conductance than those grafted onto tomato rootstock under water deficit condition It was concluded from that study that higher water uptake was inherent to a decreased hydraulic conductivity which enhanced passive water uptake rather than active water uptake. Using reciprocal grafts technique, Holbrook et al. (2002) used the grafting technique to verify the hypothesis that ABA synthesized in the root leads to stomatal closure in response to drying soil using ABA deficient tomato mutants and their wild type parents. This study showed that in response to soil drying stomatal closure can happen even in the absence of leaf water deficit, and that ABA production by roots is not a prerequisite 5. Nutrient deficiency: Response of grafting with vigorous roots tocks to nutrient stress or deficiency is not well studied. According to Rivero (2003a) this is because the
39 selection of rootstocks is rarely based on characteristics related to nutrient uptake, but rather on resistance to diseases and tolerance to environmental stresses. However, the author acknowledged that information on the rootstock/scion nutritional status relationship could be of great importance in choosing rootstocks better adapted to soils that are deficient or toxic in one or more nutrients, as well as in developing nutrient management and fertilization programs f or the production of grafted vegetables in field or greenhouse. Growth Promotion and Crop Productivity E nhancement S everal studies have demonstrated that u nder environmental conditions with low levels of disease pressure and/or abiotic stress es, vegetable plant vigor and yield often increased as a result of grafting with vigorous rootstocks Leonardi and Giuffrida (2006) observed about 30% increase in marketable yield of grafted versus n on grafted tomato plants, as a result of greater macronu trient uptake. Tomato scions grafted onto vigorous rootstocks show much higher yield than non grafted plants when g rown in soil free of pathogens (Matsuzoe et al., 19 93) Lee and Oda (2003) reported that when tomato was there was a 39.3% increase in the number of fruit set and a 54.4% increase in marketable fruit yield and a significant decrease in the percentage of cull fruit s. Pogonyi et al. (2005) observed a 62% increase in tomato yield as a result of grafting. Y ield inc reases were observed in grafted eggplant as well. It was reported that grafting eggplant onto wild Solanum rootstock led to significant yield increases as com pared to self grafted controls (Ibrahim et al., 2001; Rahman et al., 2002) This increase was due to the increased fruit number (14%) and partly by the weight per fruit (45%). A two
40 range of 20 25% as compared to the non grafte d plants (Di Gioia et al. 2010) In greenhouse production, eggplant grafted onto tomato rootstock was observed to enhance yields with increased fruit size and number as compared to non grafted controls and those with eggplant rootstock (Passam et al., 2005) Salam et al. (2002) demonstrated a 3.5 times higher yield in grafted watermelon [ Citrullus lanatus (Thunb.) Matsumand Nakai] due to larger fruit size, and more fruit per plant. Yield increases by grafting were also reported in production of wate rmelon (Ruiz and Romero, 1999; Yetisir and Sari, 2003) and cucumber (Pavlou et al., 2002) Cucumber plants grafted onto pumpkin rootstocks had 27% more marketable fruit per plant than self rooted cucumber (Seong et al., 2003) Two squash interspecific hybrid increased both watermelon yield and fruit size by an average of 90% and 26%, respectively compared to non grafted watermelon plants (Miguel et al., 2004) On the other han d, some rootstocks may also reduce growth and yield of the scion plants. For example, working with tomato grafted on two species (tomato and scarlet eggplant) Oda et al. (1996) observed a depressive effect on growth and yield when scarlet eggplant was used; this was related to potential water deficiency caused by poor vascular connections at the graft union and/or small root systems. A comparative synthesis of these different purposes of fruit tree and vegetable production is presented in Table 2 1. Grafting and Fruit Q uality Q u ality of produce encompasses sensory properties (appearance, texture, taste and aroma), nutritive values illustrated by various chemical constituents and defects (Abbott, 1999; Bai and Lindhout, 2007) Beyond the benefits inherent to grafting as
41 described above, grafted plants need to maint ain premium fruit quality to ensure successful marketing. Numerous studies have attempted to evaluate the influence of various rootstocks on the nutritional quality of both tree fruits and fruiting vegetables but have yielded contrasting results. Impact of Grafting on Fruit Tree Q uality Ex ternal and internal fruit characteristics including size, shape, peel thickness, juice content, and juice soluble solid concentration of citrus fruits were reported to be affected by rootstocks (Castle, 1995) In a study involving combinations of different rootstocks interstocks and scions, G il I zquierdo et al (2004) observed that l emon juices from lemon trees grafted with sour orange ( Citrus aurantium L.) rootstock showed 1.8 fold higher flavonoid content than that obtained from lemon trees grafted with C macrophylla L. root stock In another study conducted by Castle et al. (1993) it also showed that t rees on sour orange ( C. aurantium L.) produce d fruits of greater quality because of their high soluble solids concentration, g reater flavor and fruit size, and rela tively thin peel ; while f ruit from trees on rough lemon ( C. jambhiri Lush.) rootstocks tended to be large, low in soluble solids and acids, and have thicker peels. Castle (1 995) argued that s ucrose and other sugars present in citrus juices are osmotica and thereby plant water status which can be influenced by the rootstock play primordial role o n these attributes As an example, it was demonstrated that fruit from trees on r ough lemon (vigorous rootstock) had the lowest juice content and soluble solids concentration, and the highest leaf water potentials leading to the suggestion that the influence of rootstock on the soluble solids content is rather inherent to the degree of dilution caused by the relatively higher water potential maintained in the scion variety (Albrigo, 1977)
42 Qua lity of a pple fruit is generally described in terms of peel color, flesh firmness, juice soluble solids concentration, and fruit size Castle (1995) reviewed different studies demonstrating the influence of various rootstock s on apple fruit quality in diff erent regions in the U.S. I nconsistent results were found regarding r ootstock effects on apple ripening and quality In another study, the variety M9 used as rootstock significantly increased flesh firmness, brix o and seed number based on the first year ev aluation However, during the second year of the same experiment, this same variety increased flesh firmness while there was no effect on fruit flesh color and brix o (Samad et al., 1999) In contrast, t grafted onto three different rootstocks ( Marubakaidou prunifoliu MM 106, and M 26) did not show any significant difference on the to tal soluble solids (Motosugi et al., 1995) This study also showed that w ith MM 106 and M 26 rootstocks, fruit from the trees fertilized with nitrate nitr ogen were larger, had better surface color, higher in soluble solids, sugars, and acidity than those fertilized with ammonium nitrogen implying that the response of fruit quality is also specific to the form of nitrogen. While, w ith M. prunifoliu rootstoc ks, however, the form of nitrogen supplied had no influence on the surface color, juice quality, and mineral content of the fruit In another study, Larsen (1985) evaluated t he effect of 9 rootstocks (M2, M7, M25, M26, MM104, MM106, MM109, MM111 and S eedling ) on fruit quality a t harvest and Wellspur Delicious (WS) and Gohlspur (GS) apples and of 3 rootstocks (M7, M26 and MM106) on fruit Red King Delicious (RK) and 'Golden Delicious' (GD) apple s during a 4 year period. This study revealed that f ruits from trees grafted on M26 were larger, developed earlier color and soluble solids concentrations, and maintained higher levels of acidity (at harvest and during storage)
43 in comparison with other rootstocks. Fruit from trees on M2 tended to have high SS while f ruit color from trees on MM104, MM106 and MM109 tended to be comparatively poor Scalzo et al. (2005) showed that the ant ioxidant potential in t he most widely grown apple cultivar, Golden Delicious, is strongly affected by the genotype of the rootstock onto which it is grafted. Analyzing the underlying mechanisms of rootstock effects on fruit quality, Castle (1995) noted that a pple and other deciduous fruits while contain ing a large amount of water accumulate starch which converts to sugar during ripening. Therefore, given that r ootstocks do not produce soluble solids, it is argued that sink/source relationship ( fruit leaf ) with respect to carbon partitioning within the plant essentially control s internal fruit quality in apple. The quality attributes of grape berries which eventually define wine quality, involves soluble solids (Brix), organic acids, pH, phe nolics, monoterpenes, and other components ( e.g. proline, arginine and other amino acids) (Jackson and Lombard, 1993) The se various compounds contributing to the flavor in wines are determined in part in the vineyard through a complex and poorly understood relationship between the natural environment, vineyard management practices, and vine genotypes, including the rootstocks (Lund and Bohlmann, 2006) Inve V. Champini Brien (1978) but higher pH, titratable acidity, malate and potassium. As a result, wines made from ber ries of grafted vines were less dense, had a duller hue than that from the berries of own rooted vines, and had lower values for total phenolics, anthocyanins and ionized
44 anthocyanins. These authors concluded that wine quality was negatively affected after grafting. Higher wine K + the own rooted vine were also found by Walker et al. (2000) Using 19 rootstocks and (1988) showed that own medium juice pH while the own conclud ed that t he rootstock effects on juice pH could be attributed to changes in potassium and sodium concentrations or in the tartrate/malate ratio Ozden et al. (2010) further observed interactions of rootstocks and water supply on the quality of the berry fruits. It was suggested that SO4 rootstock under non limiting water conditions is preferable due to its positi ve effect on grape quality parameters, while 1103P might be better choice under water limiting conditions Quality attributes of peach fruits were also examined in relation to grafting with different rootstocks. In a study involving 14 differen t rootstocks, Tsipouridis and Thomidis (2005) showed that GF677, PR204/84, KID 1, AN 1/6 and KID 2 presented greater potential in term of fruit quality evaluated such as Brix o acids and flesh firmness. Drogoudi and Tsipouridis (2007) investigate d the variability in the fruit antioxidant c ontent of six clingstone cultivars and three breeding selections of peach trees grafted on three rootstocks A clear cultivar effect was found on the frui t antioxidant content with less pronounced effects of rootstocks In addition, f irmness, soluble solid s and the soluble solids to total acidity ratio were only affected slightly by
45 different rootstocks. While the total acidity of the fruit varied significantly with rootstocks; the highest total acidities (Giorgi et al., 2005) Rootstocks and Vegetable Q uality Similar to quality of tree fruits, vegetable quality also involves various attributes related to physical properties and internal attributes characterizing flavor and health promoting compounds. These quality attributes may be influenced by grafting as a result of the translocation of the metabolites inherent to the fruit quality to the scion through xylem and/or change in the physiological processes of the scion (Lee, 1994; Rouphael et al., 2010) As shown on the tree fruits, literature presents several compelling and sometimes conflicting evidences on changes in vegetable fruits quality as result of grafting. Lee (1994) reported that the fruit size of watermelon grafted to rootstock with vigorous root systems is often significantly increased comp ared to fruit from non grafted plants. External quality attributes including fruit color, fruit fresh firmness and fruit fresh thickness did not show any significant difference between the grafted and self rooted tomato plants grown in the greenhouse with mycorrhiza application (Ulukapi and Onus, 2007) Colla et al. (2006b) observed that physical quality attributes such as fruit firmness and Hunter color values (L* and a*/b*) measured on the surface of the externa l pulp of melon were significantly higher in grafted plants as compared to non grafted plants grown under greenhouse conditions and under saline conditions Watermelon plants grafted onto Lagenaria rootstocks yielded fruits which were firmer by 24% than th e fruits from the non grafted plants independent of cultivar, rootstock and growing conditions (greenhouse vs. open field) (Yetisir and Sari, 2003) In their review, Rouphael et al. (2010) suggested that the effect of rootstock on vegetable
46 fruit firmness may be related to a number of anatomical and physiological changes in the fruits including cellular morphology, cell turgor, the chemical and mechanical properties of cell walls of fruit a nd water and nutritional status of scion. In addition to external quality attributes, rootstocks are also shown to have contrasting influences on the internal quality attributes of the fruit vegetable including tomato, cucumber, and watermelon. G rafting ha s been reported in some instances to increase lycopene, carotene, vitamin C, and antioxidant activity while such effects were dependent on species or varieties of rootstocks used (Davis et al., 2008a; Dorais et al. 2008; Fernandez Garcia et al., 2004b) For beef tomato, however, Dorais et al. (2008) reported from their unpublished studies that the lycopene content and the antioxidant activity of fruit from grafted and non grafted plants were not affected by grafting O ther quality attributes of tomato fruit such as ti tratable acidity and soluble solid content did not show any significant difference between grafted and non grafted plants (Ulukapi and Onus, 2007) The use of Beaufort rootstock enhanced the growth and production, but resulted in lower soluble solids values (Romano and Paratore, 2001) Possible explanations for the relative decrease of the quality parameters evaluated are that the improved productivity and subsequent yield increase caused a decreased concentration of the main fruit components (Marsic and Osvald, 2004; Pogonyi et al., 2005) Under saline conditions, Fernandez Garcia et al. (2004b) observed that depending on the grafted cultivar ascorbic acid of tomato fruit from grafted plants grown under 30 and 60 mM NaCl increased compared to non grafted plants, while grafting increased the lycopene and be ta carotene content of fruit grown under all three salinity treatments Moreover, Khah et al. (2006) observed no difference
47 in total soluble solids titratable acidity and soluble solid acidity ratios between non gra fted tomato cv. Big Red and grafted tomatoes onto He Man ( S. lycopersicum L. x S. habrochaites S. Knapp&D.M.Spooner) and ( S. lycopersicum L.) rootstocks both under open field and greenhouse conditions. More recently, Di Gioia et al. (2010) observed no significant differences in total soluble solids contents between non grafted tomato Cuore di Bue an heirloom 'oxheart' tomato, and grafted tomatoes onto two interspecific ( Solanum lycopersicum L. x Solanum habrochaites S. Knapp &D.M. Spooner ) rootstocks ('Beaufort F1' and 'Maxifort F1') grown under greenhouse conditions. However, these authors observed a decrease in vitamin C contents in a range of 14 20% in the fruit of plants grafted onto either rootstock. With respect to other solanaceous vegetables, Colla et al. (2008) observed that the nutritional qualities of grafted peppers such as fruit dry matter content, total soluble solid content, and titratable acidity were not affected by grafting with rootstocks. Among the cucurbitaceae species, Yetisir and Sari (2003) did not find any significant effect of grafting on the soluble solid content of watermelon fruit juice. However, in another study, l ycopene, dehydroascorbate and total vitamin C contents on fru its from grafted mini watermelon plants were higher by 40.5%, 13% and 7.3%, respectively, than th at from non grafted plants of mini watermelon The concentrations of spermidine and putrescine, small aliphatic polyamines found in all living organisms, indis pensable for growth and cell multiplication, were reduced by grafting by 24% and 59%, respectively (Proietti et al., 2008) Likewise, Huang et al. (2009) noticed an increase of vitamin C contents in cucumber fruits harvested from plants grafted onto figleaf gourd It has been shown that Cucurbita moschata used as rootstock caused a
48 reduction in texture and flavor in grafted Honey Dew fruits, despite ensuring resistance to Fusarium wilt (Imazu, 1949) Yetisir and Sari (2003) pointed out that the decrease in fruit quali ty does not represent a general phenomenon but depends on the specific scion rootstock interaction, and on the particular combination of growing conditions. T he total soluble solid concentration of melons grafted onto the pumpkin interspecific hybrid C. maxima Duchesne x C. moschata Duchesne) was reported to be lower than those from non grafted Cyrano ( C. melo L. var cantaloupensis Naud) (Colla et al., 2006b) Among the various reasons that may explain how rootstocks negatively affect scion fruit quality Rouphael et al. (2010) suggested that for instance watermelon fruits from grafts using interspecific hybrid rootstocks acquire color at about the same time as non grafted plants, but the sugars do not accumulate until later; this results in growers harvesting too soon (about the same time they would harvest non grafted plants), and the fruits tend t o have a lower pH and a flavor. Metabolically, it wa s observed that a cid invertase activity was high during the first third of watermelon fruit development, and then had little activity during the last 5 days of ripening. Sucrose synthase activity was moderately high in the first third of fruit development, and then remained low. Sucrose phosphate synthase increased during the second third of fruit development, and fell in the last five days of ripening. Therefore, t he low sugar content of grafted watermelon is correlated with low invertase, high sucrose syn thase activities, and low sucrose phosphate synthase activity Physiological Basis of Rootstock Scion I nteractions Selection of an appropriate graft combination is critical for the successful growth and development of grafted fruit trees and vegetable spec ies, in that evidences show that the scion rootstock interaction influences several physiological processes including
49 plant nutrition, water relations, and leaf gas exchange (Gonalves et al., 2006; Nielsen and Kappel, 1996) In addition to evaluating the performance of grafted plants, several studies were also conducted to understand how these different physiological processes are influenced within different graft combinations. Studies have demonstrated that underlying mechan isms for increased yield and vigor generally observed on grafted vegetable plants are attributable to increased water and nutrient uptake by vigorous rootstock genotypes. N utrient and water uptake, as well as the translocation of various substances such as ions, photosynthates, plant hormones, and alkaloids can be influenced by rootstocks or by grafting with vigorous rootstock Ruiz et al. (2007) observed that foliar contents of N, Na and K in grafted melon plants were determined by the rootstock genotype, and also these foliar contents explained the differences in yield observed between the grafted and non grafted plants. Total nitrogen (N) was found to be higher in grafted sweet pepper plants, but with lower microelements concentration (Del Amor et al., 2008) The enhanced uptake of minerals appeared to be closely related to the activity of enzymes responsible for absorption (Ahn et al., 1999) Studies by Ruiz and Romero (1999) demonstrated that grafted melon had in leaf tissue, lower concentrations of nitrate, higher nitrate reductase activity (NRA), and lower contents of total free amino acids and soluble proteins compared to the non grafted control. Higher nitrat e reductase and nitrite reductase activities were also found in leaves of grafted watermelon plants (Pulgar et al., 2000) These changes were attributed to differences in N utilization and assimilation between grafted and non grafted plants which resulted in higher yields. Among micronutrients, Rivero et al. (2004) observed that rootstock used on tomato showed greater capacity for Fe uptake and accumulation. Analyzing the graft
50 compatibili ty in Solanaceous plants in relation to plant nutrition, Kawaguchi et al. (2008) observed no significant differences in nitrogen concentration among the four graft combinations examined including tomato/tomato, tomato/pepper pepper/tomato and pepper/pepper. Furthermore, studies on melon grafting have shown that rootstocks can enhance some morphological and/or physiological characteristics of melon plants, leading to increas ed uptake of phosphorus (P) from soil and its transl ocation to the leaves of the scion (Ruiz et al., 1996) In another study on grafted melon plants P concentration s were similar between the scion and rootstock scion combina tions (Ruiz et al., 1997) Kawaguchi et al (2008) concluded that the rootstock species was the main factor affecting the absorption and translocation of phosphorus in the graft combinations of Solanaceous plants. With respect to grafted fruit trees, nutrient absorption and translocation in the scion can also be influenced by the rootstock g enotypes and are shown to be positively related to the growth of yield of the scion. Chaplin and Westwood (1980) did not find any effect of rootstock on the leaf mineral content of the scio n in grafted apple. However, Tagliavani et al. (1992) suggested that the vigor of both scion and root system had an important role in the uptake and translocation of nutrients in grafted fruit trees Sorgon et al (2006) compared the nitrogen use efficiency of citrus trees grafted on four rootstocks including 'Rough Lemon' ( C. jambhiri Lush), 'Sweet Orange' ( C sinensis (L.) Osbeck), 'Cleopatra Mandarin' ( C reshni Hort ex Tan.) and Sour Orange ( C aurantium L.). It was concluded that the 'Sour Orange' and 'Sweet Orange' were nitrate use efficient and inefficient rootstocks, respectively, while the 'Rough Lemon' and 'Cleopatra Mandarin' demonstrate greater and reduced genetic potential, respectively in term of
51 biomass accumulation I yengar et al (1982) studied the influence of various rootstocks on the leaf nutrient composition of mandarin cultivars Coorg and Kinnow and observed that differences due to rootstocks in the leaf nutrient composition with respect to N, K, Ca, Mg, Na and Mn were significant. In addition to nutrient accumulation, grafting with specific rootstocks may induce greater diversity in the performance of the scion in term of water use efficiency. Ho wever, there are relatively few studies on the impact of rootstocks on the water uptake and water use efficiency, particularly with the vegetable plants. Grafted 'Big Red' tomato onto 'He mans' rootstocks showed 50% higher water use efficiency (WUE, measur ed as amount of harvested fruit per volume of transpired water) as compared to the non grafted plants (Lykas et al., 2008) M any of the studies which attempted to demonstrate evidence of genetic variability of rootstocks in te rm of water use efficiency were conducted on the fruit tree species. When grapes were grown under well watered conditions and in the absence of salinity, the effect of rootstock on scion transpiration 13 C) is relatively small as related to variation among scion varieties (Gibberd et al., 2001; Virgona et al., 2003) These authors concluded that under such conditions, rootstocks affect yield through their influence on vine vigor and proposed that this effect would be apparent under water deficit conditions. Satisha et al. (2006) demonstrated that among the rootstocks evaluated, 'Dogridg e' was the most efficient water user, followed by 'Salt Creek' and 'VC clone'. Pou et al. (2008) show ed that the hybrid Richter 110 ( Vitis berlandieri x Vitis rupestris ) presents the potential of greater adaptation to drought. In a study involving 20 cultivars of grapevine grown under optimal and stress conditions, Bota et al. (2001)
52 observed that Escursach was found to be promising cultivar, presenting low water consumption as well as high carbon assimilation. Comparing different rootstocks of pistachio, Germ ana (1996) observed that Pistacia atlantica has higher transpiration and photosynthe tic activity than Pistacia terebinthus (11.5 mmol H 2 O m 2 s 1 against only 4.0 mmol H 2 O m 2 s 1 ), particularly in stressed plants, which could make it more suitable to drought tolerance. Solari (2006) confirmed that rootstock effects on the peach tree water relations are inherent in part, to differences in the tree hydraulic conductance associated with specific rootstocks. Mo reover, Soar et al. (2006) reported that rootstock effect on gas exchange of field grown grapevines is most likely due to differences in the relative potential of the rootstocks to higher uptake and translocation of water to the scion. Koundouras et al. (2008) evaluated the rootstock effects on the adaptive strategies of grapevine ( Vitis vinifera L. cv. Cabernet Sauvignon) under contrasting soils under no n limiting water supply due to its capacity to achieve a balanced would be better to grow in semi arid regions where water limitation occurs. The influence of the rootstoc k on nutrient and water uptake is attributed mainly to physical characteristics such as lateral and vertical development of the root, or greater uptake potential (Martnez Ballesta et al., 2010) Thus far, there is a need to evaluate root distribution of the root system of the grafted plants across various environment conditions. During g rafting process, there is development of callus tissues at the graft union, which allows water flow to occur from the rootstock to the scion. It is
53 demonstrated that in grafted tomato, the major hydraulic connections become functional over a period of 48 h from the fifth day after grafting (Turquois and Malone, 1996) Hormonal Influence in Rootstock Scion I nteractions In addition to influencing the nutritional and plant water status of the scion, there is evidence that hormonal status of the grafted plant is regulated by the rootstock scion interaction. Rootstocks differ in their inherent capacity to produce hormones and the ways in which these biologically active substances are transported within the plant. It is demons trated that a significant number of endogenous hormones can move from the rootstock through the vascular system of the graft union and affect scion growth after grafting (Ding et al., 2003; Lucas and Lee, 2004) These plant growth regulators, namely cytokinins in the roo t system are in part responsible for enhanced vigor and yield in grafted plants (Lee, 1994; Leonardi and Giuffrida, 2006; Ruiz et al., 2006b) Cytokinins are a group of plant hormones mainly synthesized in roots and moved upward in plants. They play a critical role in plant growth and development by regulating leaf senescence, apical dominance, root proliferation and nutritional signaling and also induction of photosynthesis gene expression (Hirose et al., 2008) Plants with vigorous root systems produce more cytokinins and the yield increase induced by a vigorous rootstock is closely a ssociated with the amount of cytokinins in the ascending xylem sap (Lee and Oda, 2003) The cytokinin activity in xylem exudate of grafted watermelon is affected by rootstocks and the shoot growth in watermelon was partly attributed to the higher quantity of cytokinins translocated to the shoot (Yamasaki et al., 1994) However, during the fruit formation and maturation stages, a decrease in the cytokinin concentrations was noticed in both the grafted and non grafted plants, suggesting that root cytokinin synthesis is regulated by the source/sink relation oc curring in the plant,
54 i.e. a strong assimilate demand in the fruits results in a reduction in the cytokinin synthesis. C ytokinin composition varies greatly between cucurbit species with only zeatin and dihydrozeatin found in cucumber, whereas large quantit ies of isopentenyladenine and isopentenyladenosine are found in squash and gourd C ucumber plants grafted on pumpkin rootstocks showed a 2.2 times increase in trans zeatin root content than non grafted cucumber (Seong et al., 2003) As in the case of vegetable grafti ng, evidence of hormonal influence was also shown with respect to the fruit tree plants. Kamboj et a l. (1999) evaluated t he concentrations of zeatin and zeatin riboside in shoot xylem sap and root pressure exudate obtai ned from three apple rootstocks ( M.27, M.9 and MM.106 ) and from trees of 'Fiesta' scion grafted onto the rootstocks. These authors observed that zeatin was the predominant cytokinin in xylem sap from the dwarfing rootstocks (M.27 and M.9), while zeatin rib oside was the predominant cytokinin in xylem sap from the more vigorous rootstock (MM.106). Cytokinin concentrations in root pressure exudate and shoot xylem sap increased with increasing vigor of the rootstock, despite whether the plants were non grafted rootstocks, or were grafted onto the rootstocks. Similar pattern was also observed in the shoot sap from the more vigorous rootstock (MM.106) with greater amounts of cytokinins than the more dwarfing rootstocks (M.9 and M.27). Moreover, a study on the horm onal relationship between grafted and non grafted peach trees showed a positive correlation between the growth potential of the rootstocks and its highest level of zea tin ribulose (Sorce et al., 2002)
55 Another hormone of importance in plant growth and development is auxin which is primarily synthesized in the shoot and can move downward to control root growt h It is reported that auxin is involved in the development of compatible graft union and is released from vascular strands of the stock and scion and induces the differentiation of vascular tissues, functioning as morphogenic substances (Aloni et al., 2010) Measurement of endogenous indole acetic acid (IAA) and abscisic acid (ABA) in the Eureka three citrus rootstock cultivar s was conducted by Nod a et al. (2000) They found that the IAA level in the new shoots was Flyin dwarfing rootstock). In contrast, the ABA level in the new shoots was highest These results led Sorce et al. (2002) to speculate t hat dwarfing rootstocks reduce the basipetal transport of IAA to the root, thereby decreasing the amount of root produced cytokinin and gibberellin transported to the scion Similarly, in their review on hormonal signaling in rootstock scion interactions, Aloni et al. (2010) a feedback loop exists in which a decrease in basipetal flow of IAA from the shoot stimulate s th e synthesis and export of cytokinins from the root. The upward movement of cytokinin in the xylem sap induces an increase in the synthesis and translocation of IAA out of the shoot apex which in turn reduce s cytokinin levels in the xylem sap. Grafting and Gene E xpression In addition to the hormonal communication and transport of other mobile macromolecules (photosynthates, amino acids) within different organs of the growing plant, several studies have shown that many proteins and mRNAs are also involved in long distance transport through phloem More specifically, Banerjee et al. (2006)
56 reported that after grafting, specific RNAs can also move long distances through the phloem across the graft union According to Harada (2010) grafting experiments were used to demonstrate that long distance transport of signals is involved in various physiological processes such as photoperiodic flowering, tuberization, nodulation, leaf development, shoot branching, and defense against pathogen s. Further, Kim et al. (2001) mentioned that signaling proteins or RNAs can move through the graft union, and small differences in the availa bility of those signals can alter growth of the scion. Investigating the phloem of apple plants for the presence of mRNAs that may have been transported over long distances from the rootstock, Kanehira et al. (2010) observed several mRNAs that have already been reported as phloem transported RNA in other plants, including SLR/IAA14 were fou nd to be transported from rootstock to scion through the graft union in apple. Furthermore, with grafting experiments, Xu et al. (2010) investigated the transport of endogenous gibberellic acid insensitive (GAI) mRNA in apple ( Malus x domestica cv. Fuji and Malus xiaojinensis ). Results showed that each GAI mRNA of scion and rootstock plants was detected i n the graft combinations as early as 5 days after grafting, indicating that the GAI mRNA moved in both upward and downward directions via the graft union. Jensen et al. (2003) compare M.9), which confer different levels of fire blight susceptibility to the scion. About 100 genes were identified that were differentia From a molecular perspective, Zhang et al. (2008) used eggplant as a scion and tomato as a rootstock to study the effects of rootstocks on the gene expression patter n
57 of the scions. They observed changes in gene expression pattern in a variety of genes in the eggplant scions influencing several physiological processes of the scion; these alterations in gene expression were induced by the effects of tomato rootstocks i n that they were absent from the non grafted and self grafted eggplant controls. Furthermore, Kudo et al. (2007) conducted heterografting experiments using potato ( Solanum tuberosum ) as scion and tomato ( Solan um lycopersicum ) as rootstock to test whether an RNA molecule responsible for changing leaf shape could be transported and function across the grafting union. The study revealed that a graft transmissible RNA from the tomato rootstock without any leaves co uld change leaf morphology of the potato scion. Interspecific grafts of Cucurbitaceae were used to study the mobility of structural P proteins in the phloem. It revealed that when Cucumis sativus L. scions were grafted onto Cucurbita rootstocks, at least n ine additional proteins were found in the scion exudate after grafting and these proteins corresponded exactly to those of the respective Cucurbita sp. rootstock (Golecki et al., 1999) indicating the long distance transport of mRNAs from the rootstock into the scion. Concluding Remarks and F uture Research P rospect While both fruit tree grafting and vegetable grafting are regularly performed to ensure tolerance or resistance to a wide array of biotic (pathogens and pests) and abiotic (environmental) stresses, size control (dwarfing of the scion) and the avoid ance of juvenility are unique benefits of grafting fruit tree. Across all the fruit trees where grafting is used, there is a negative correlation between the vigor of the rootstock and the yield recorded from the grafted plants on per plant basis. Converse ly, with respect to vegetable grafting, rootstock which induces greater vigor is preferable owing to the yield enhancement that usually results. The influence of rootstock scion interaction on the
58 quality attributes of the fruits is still elusive. In g ener al metabolic processes inherent to fruit quality attributes seem to be primarily species driven and are under scion genetic control However, t hese processes may also be influenced by environmental factors management practices and the rootstock genotypes It seems that in the fruit trees, rootstock with greater vigor potential which yields fruit with bigger size tends to induce a decrease in some internal chemical attributes, e.g., soluble solids content While on the vegetable fruit namely watermelon for instance a delay inherent to grafting in sugar accumulation in the fruit may explain a decrease in soluble solids content observed in grafted watermelon fruit. In essence, it is unclear how rootstocks exert their influence on fruit quality but water rela tions, nutrition, and plant growth regulators are undoubtedly among the most important factors involved Research efforts have been made to understand how rootstocks interact with the scions to cause anatomical, physiological, hormonal, and to some extent quality changes noticed. The findings help to gain in depth knowledge of several physiological and molecular processes that are the driving forces of the growth, yield formation and nutritional quality attributes of the grafted plants. However, several mec hanisms including gene expression and long distance transport of macromolecules and the direct interconnection with the performance of the grafted plants are still unclear. More research is still needed to demonstrate how the hormonal communication and lon g distance transport of mRNA influence the growth and development of grafted plants, namely with tomato as plant model. Further, the performance of available rootstocks for instance on tomato grafting on a wide range of scions across various growing syste ms n eeds to be evaluated in depth given that the effects of rootstocks on scion vigor and
59 yield are to some extent specific to scion/rootstock combinations. It is also necessary to demonstrate whether using the grafting with rootstock would lead to a reduc tion in the requirements of water and nutrients for optimum production especially under field conditions.
60 Table 2 1 Comparative summary of key research findings on the use of grafting in fruit trees and vegetables Key Areas Fruit Trees Vegetables Ci trus Apple Grapes Other Cucurbits: Watermelon, cucumber, melon Solanaceous: Tomato, eggplant, pepper, potato Response to Environmental stresses Tolerance and/or resistance to salinity, drought, low temperature, flooding Tolerance and/or resistance to sa linity, drought, heat and cold, flooding Avoidance of juvenility Allow precocity into fruit production Grafting delays reproductive stage Size control and vigor Dwarfing rootstocks are used to control the size Vigor is improved Yield Yield per plant is reduced but high density of planting enhances the yield on area basis Strong vigor results in high yield Physiology Nutrient Improved NUE, Higher N, K, Ca, Mg Leaf mineral content similar Enhanced nitrogen assimilation Higher nutrient accumulation but similar nutrient concentration Water Increased WUE, particularly in stress conditions WUE not explored Water uptake is improved Gas exchange Improved Carbon assimilation, stomatal conductance, transpiration and other parameter s Improved Carbon as Water uptake is improved similation, stomatal conductance, tanspiration Hormone Cytokinin and Auxin higher on rootstocks with high vigor Higher Cytokinin Auxin concentration Not well explored Quality Quality is negatively corre lated with vigor. Dwarfing rootstock tends to increase nutritional quality attributes, e.g. SSC Inconsistent findings, as related to internal quality attributes N UE: nitrogen use efficiency, WUE: water use efficiency, SSC: Soluble solids content
61 CHAPTER 3 INTERSPERSPECIFIC RO OTSTOCKS ENHANCE TOMATO YIELD WITHOUT COMPROMISING FRUIT QUALITY UNDER GREENHOUSE CONDITION S Introduction Grafting creates a new plant with combined desirable aboveground and belowground attributes from selected scion and rootstock p lants. In addition to its application for fruit trees, grafting has been successfully used in production of solanaceous and cucurbitaceous vegetables such as tomato ( Solanum lycopersicum L. ) eggplant ( S. melongena ), pepper ( Capsicum annuum ), watermelon ( C itrullus lanatus ), melon ( Cucumis melo ), and cucumber ( Cucumis sativus ) (Lee and Oda, 2003) In these species, grafting has been an effective tool for management of root knot nematodes ( Meloidogyne incognita ) and soilborne diseases including Fusarium wilt ( Fusarium oxysporum ), Verticillium wilt ( Verticil lium spp) and bacterial wilt ( Ralstonia solanacearum) (Davis et al., 2008b; Lee, 1994; Lee and Oda, 2003; Leonardi and Romano, 2004) The disease resistance acquired by grafted plants is mainly attributed to the di sease resistant rootstocks, although specific mechanisms are not yet well understood. V egetable grafting has thus evolved into a sustainable tool that fits within an integrated pest management program to tackle site specific problems and reduce pesticide i nputs (Bletsos et al., 2003; Iouannou, 2001; Kubota et al., 2008) As a viable sustainable alternative to soil fumigation, grafting has been widely adopted by vegetable growers in East Asia and Mediterranean countri es Recent r esearch in the U.S. has explore d effectiveness of grafting to successfully manage soil borne pathogens in both conventional and organic vegetable production systems (Kubota et al., 2008; Rivard and Louws, 2008)
62 Beyond disease resistance, efficiency of nutrient and water use also increases i n grafted vegetables, an effect often ascribed to the vigorous root system s of rootstock s (Lee, 1994; Passam et al., 2005) Modification of endogenous plant hormone (e.g., cytokinins) status by the rootstock is also suggested to play a n important role in promoting growth of the graft hybrid (Davis et al., 2008b; Edelstein, 2004; Lee, 1994; Lee and O da, 2003) The p erformance of grafted vegetables may vary substantially among rootstock scion combinations It was reported that p otential for yield increase in grafted melon was greatly dependent upon the rootstock genotypes (Ruiz et al., 1997) Both intraspecific and interspecific hybrid rootstocks are typically used for tomato production (Oda, 2007) G rafte d tomato plants with interspecific rootstock s exhibit more vigorous growth and higher yield compared to self grafted tomato scion plants and those grafted onto int ra specific rootstocks (Leonardi and Giuffrida, 2006) Interspecific r ootstocks ( S. lycopersicum S. habrochaites ) including Maxifort and Beaufort (De Ruiter Seeds Inc., Lakewood, CO, USA) are among the most popular tomato rootstocks in North America (Ki ng et al., 2010) In addition to high resistance to a range of soilborne pathogens and root knot nematodes, they are recognized by both greenhouse and field growers (Kubota et al., 2008; Rivard et al., 2010a) for improving plant vigor and yield of grafted tomato even under low disease pressure. Considering the increasing interest from tomato growers in the integrated use of grafting techniques for sustainable production (King et al., 2010; Rivard et al., 2010a) more research is needed to elucidate the influence of available tomato rootstocks on plant growth yield, nutrient uptake, and fruit quality beyond disease resistance Khah et
63 al. (2006) observed th at grafting might delay first flowering date and first harvest of tomato due to the physical stress incurred by grafting Knowing h ow rootstocks may affect early yield of tomato would be of great interest to market growers. Quality assessment of fruits fro m grafted vegetables is also critical especially when a new rootstock is used since some rootstocks may cause undesirable changes in sensory and nutritional qualit y attributes of fruits (Davis et al., 2008a; Davis et al., 2008b) Thus, especia lly in the U.S., effects of these interspecific rootstocks on soil borne disease control and yield increase are well known (Barrett et al., 2012a; Rivard and Louws, 2008) however less is known about the underlying effects on plant growth, nutrient uptake and fruit nutritional quality. Consequently, t he objectives of this study were to: 1) determine the biomass nutrient accumulation and yield performance of tomato plants grafted onto disease resistan t interspecific rootstocks under greenhouse conditions and 2) determine the effects of different rootstocks on quality attributes of tomato fruit. Materials and M ethods Tomato Grafting and P lanting T wo greenhouse studies evaluated first, the effects of a commonly used tomato and second, performance of this rootstock relative to three others RST 04 105 ( DP S eeds, Yuma, AZ S. lycopersicum ) ( Seminis Vegetable Seeds, Inc. Saint Louis, MO, USA) was used as the scion in both studies The f irst greenhouse study was conducted from October 2007 to April 2008. S eeds of and n into 128 cell styrofoam flats ( Speedling
64 Incorporated, Sun City, FL USA ) with Metro Mix 200 ( Sun Gro Horticulture Bellevue, WA, USA) in the greenhouse on 18 and 20 October 2007 respectively. Plants were graf ted using the splice method (Lee, 1994) on 21 November 200 7 when 5 to 6 true leaves were present G raft ed plants were placed in a closed healing chambe r equipped with two humidifiers, where temperature ranged from 22 o C to 32 o C and relative humidity was maintained above 80 % Light and ventilation were introduced gradually a fter a dark period of 3 days. O verall healing was completed with in 7 days with a survival rate above 90% Ten days after grafting, the g rafted plants / (scion/rootstock FL / MA) and n on grafted (FL) plants were transplanted into 11.4 L black plastic containers filled with Metro Mix 200. The experiment was arranged as a randomized complete block de sign with 4 replications (blocks) and 10 plants in each treatment per replication Fertilizers incl uding 20 N 8.8 P 16.6 K and 15.5 N 0 P 0 K 19 Ca were incorporated into each container every 5 to 14 days depending on plant growth and development A pplication rates rang ed from 4 to 28 g per plant. Plants were watered daily with 0.6 L to 2.0 L per plan t. The second greenhouse study was conducted from January to June 200 9. S eeds of the rootstocks including RST 04 105 were sow n on 2 January 2009 and seeds of the scion F lorida 47 were sow n on 6 January 2009 usi ng the same transplant production system described above Tomato plants were grafted on 29 January 2 00 9 T he chamber described above was equipped with an auto control air conditioning system for healing the grafts, where temperature was maintained at 25 3 o C Twelve days after grafting, control plants, including non grafted Florida 47 (FL) and self grafted Florida 47 (FL/FL) as well as completely healed
65 grafts from Florida 47 Maxifort ( FL/ MA), Florida 47 Beaufort ( FL/ BE), Florida 47 Multifor t ( FL/ MU) and Florida 47 RST 04 105 ( FL/ RS) were transplanted i nto 11.4 L black plastic containers filled with horticultural grade p erlite. A randomized complete block design with 4 replications was used with 6 plants in each treatment per replicati on A fertigation system using Dosatron injectors ( Dosatron International, Inc ., Clearwater, FL USA) and 1.89 L h 1 drippers ( N etafim irrigation, Inc. Fresno, CA USA) was established to deliver nutrient solution to each plant. Nutrient solution formulat ions developed by Hochmuth and Hochmuth (2001) were modified to provide solutions containing concentrated macro and micro nutrients as follows : 236 mM H 3 PO 4 321 mM KCl, 398 mM MgSO 4 237 mM KNO 3 55 CuSO 4 2 mM MnSO 4 54 ZnSO 4 1 mM Na 2 B 8 O 13 and 1 00 mM NaMo in stock A; 1581 mM CaNO3 and 79 mM Fe330 in stock B. By adjusting dilution rates, mixed nutrient solution s of appropriate concentrations were supplied to plants based on plant growth and development Twenty to thirty cycles were scheduled each day to fertigate each plant with a 1 or 2 minute delivery per cycle. Axial branches (s uckers ) below the first fruit cluster were pruned during both greenhouse experiments. T omato Y ield M easurements Tomato fruit were harvested at or after the breaker stage of ripeness. In the 2007 to 2008 study these fruits were harvested 12 times from 10 plants in each treatment and replication between 74 and 140 day s after transplanting (DAT) In the 2009 study fr uit were harvested 9 times from 4 plants each treatment and replication between 63 and 120 DAT. These harvests were reported as yields from early, mid and late period of the seasons with each period representing 4 or 3 harvests the first and second
66 studi es respectively. Analyses included fruit n umber average fruit weight and total m arketable yield per plant Plant G rowth and N utrient A ccumulation In the 2009 greenhouse experiment one plant from each treatment per replication was sampled at 28, 63 a nd 120 DAT, ( respectively representing anthesis, first fruit harvest and final fruit harvest) for destructive analys e s of p lant total leaf area and biomass accumulation. Each plant sampled was separated into leaf blade s petiole s stem s fruit, and root s After measuring the fresh weight of each pooled fraction representative sub samples ( approximately 300 g each) with ( 500 g for fruit) were used to determine fresh weight / dry weight ratios Dried weights were obtained after sub samples were dried in a forced air oven at 60C for 72 to 120 h to a constant weight Data were used to calculate total plant dry weight and that of component organs. Leaf blade area of the sub samples of leaf blade was measured using a LI COR 3100 leaf area meter (LI COR Inc., L incoln, NE, USA). Total leaf area was then estimate d for each plant. Nutrient accumulation was evaluated on plant tissue samples collected from each treatment per replication at 28, 63, and 120 DAT. Dried sub samples of each plant part (i.e., blade, petiol e, stem, and fruit) were ground using a Thomas Wiley mill (A.H. Thomas Co., Philadelphia, PA) until particles could pass through a 1 mm screen These were then analyzed for the concentrations of nitrogen, phosphorus, potassium, and calcium based on procedu res described by Mylavarapu and Kennelley (2004) Briefly, for N analysis, ground tissues were digested with a concentrated s ulfuric a cid After dig estion, t otal Kjeldahl N was determined using automated colorimetric analysis. Tissues were digested in a concentrated hydrochloric acid for determination of P, K, and
67 Ca. Inductively coupled plasma spectrophotometry was used to analyze K and Ca concentrat ions, while atomic absorption spectroscopy in combination with colorimetric analysis was used for P analysis. Finally, accumulation of N, P, K, and Ca into above ground plant parts was estimated based on the dry weights of blade, petiole, stem, and fruit f ractions and the collective nutrient levels in each of these plant parts Fruit Q uality A ssessment Fruit quality attributes were analyzed during the early and mid fruit harvest periods of the 2009 experiment. C omposite samples of five randomly selected ma rketable ripe fruit from each treatment were used for measurements of pH, total titratable acidity (TTA), soluble solid s content (SSC), a scorbic acid lycopene and carotene concentration s, and phenolic content. To ensure similar ripeness of fruit across different treatments, tomato fruit were first stored at 20 o C after harvest and fruit color was monitored using a Minolta CR 400 colorimeter ( Konica Minolta Sensing Americas, Inc ., Ramsey, NJ USA ). When the a* value (negative for green and positive for red) reached the positive peak level fruit were cut into quarters and frozen at 30 o C before extraction. The thawed and homogenized samples were centrifuged (20 m in; 1 7 6 00 gn ; 4C) The resulting supernatant was filtered through cheesecloth prior to analysis. The T T A was determined by titration using Titrino Metrohm (model 719 S, Switzerland) of 6.0 g of juice plus 50 mL of water with 0.1N sodium hydroxide solut ion until pH 8.2 was reached and the TA was expressed as percent citric acid. The pH of the diluted juice was determined automatically using the same equipment for T TA determination S oluble solids content was measured by a digital r efractometer (model ABB E Mark II, Cambridge Instruments Inc, U.S.A) and expressed as per cent FW of the juice
68 The concentration of ascorbic acid in tomato fruit w as measured spectrometrically according to the AOAC method 967.21 (Aoac International, 2000) (Horwitz, 2000) The modified m ethod of Nagata and Yamashita (1992) was used to determine concentrations of lycopene and carotene Bri efly, 1.0 g of homogenized fruit sample was mixed with 16 mL of acetone / hexane (2:3, v/v) solvent. For carotenoid extraction, t he mixture was placed in the freezer at 20 o C fo r 60 min followed by vortex sha k ing for 30 sec (fast speed). T wo phases were separated in the extract. An aliquot from the upper phase was measured for absorbance (Abs) at 663, 645, 505, and 453 nm by spectrophotometer, using the hexane solvent as a blank The final concentrations of lycopene and carotene in the fruit samples ( g/g FW) were calculated based on the following equations: l ycopene (mg/100 m L of extract) = 0.0458 A bs 663 + 0.204 A bs 645 + 0.372 A bs 505 0.0806 A bs 453 ; carotene (mg/100 m L of extract) = 0.216 A bs 663 1.22 A bs 645 0.304 A bs 505 + 0.452 A bs 453 P henolic contents in the hydrophilic and lipophilic fractions of fruit extract were determined as described previously by Toor and Savage (2005) Ga l lic acid was used as the standard and results were expressed as ga l lic acid equivalents per 100 g fresh weight. Statistical A nalyses Analysis of variance was performed for both greenhouse experiments, using t he Proc Mixed Procedure of the SAS system for Windows (version 9 2 Cary, N C USA) Data at e ach sampling was analyzed separately with a model including fixed effects of graft types and repetition (block) as a random term. Multiple comparisons among treatments were done using Fisher s LSD test and treatment s were considered different at P 0.05
69 R esults Tomato Y ield C omponents In both the 2007 2008 and 2009 experiments, total marketable yields of were significantly improved by grafting with interspecific rootstocks (Tables 3 1 and 3 2). The yield increase of grafted plants b egan at the mid and early harvests in 2008 and 2009, respectively. In the 2009 experiment, the four rootstocks tested resulted in similar total marketable fruit yields with no significant difference s among them. More specifically, in the 2007 2008 experime nt where only Maxifort was used as a rootstock, the total marketable yield was increased approximately 45% by using grafted plants. In the 2009 experiment, use of grafted plants increased total marketable yield by an average of 66% relative to either no n grafted or self grafted plants. The yield inc reases observed in both years were largely attributable to significantly higher number s of fruit per plant On the other hand, higher average fruit weight also contributed to the total yield increase s of FL/RS in 2009. Moreover in the 2009 experiment, increased average fruit weight was also observed in FL/MU and FL/RS at the early fruit harvest and for FL/MA and FL/MU at the late fruit harvest, while all four rootstocks had higher average fruit weight at mid f ruit harvest. Plant G rowth In the 2009 experiment, leaf area and t otal aboveground dry matter accumulat ion were similar between non grafted and grafted plants with interspecific rootstocks at 28 DAT, i.e., the anthesis stage (Table 3 3). More vigorous growth as a result of grafting with interspecific rootstocks was observed as plant development progressed. At 63 DAT, i.e., first fruit harvest, grafting with all the interspecific rootstocks led to a significant increase in leaf area. Specificall y, the average leaf area of
70 tomato plants grafted onto these rootstocks was 37% and 41% greater than that of non grafted and self grafted plants respectively. In addition, total aboveground biomass of FL/MA and FL/MU was significantly higher than FL/FL an d FL P lant growth declined toward the final fruit harvest (120 DAT), while FL/MA appeared to retain growth vigor, exhibiting significantly higher aboveground biomass and larger leaf area at final harvest relative to FL and FL/FL. The root dry weight of gr afted plants was significantly lower than that of non grafted plants at 28 DAT. Grafting with interspecific rootstocks led to similar root growth compared with non grafted plants at 63 and 120 DAT, whereas self grafting resulted in a significant reduction of root growth throughout the growing season. Plant T issue N utrient C oncentrations The macronutrient concentrations in plant tissues varied considerably between treatments (Table 3 4 and 3 5). At 28 DAT, FL/MA had significantly higher N and Ca concentrati ons in leaf blades than FL or FL/FL. Similarly, FL/MU and FL/BE had higher concentrations of K. In contrast, P concentrations in leaf blades were consistently lower in grafted plants than non grafted plants. Concentrations of N in the petiole were similar among all the treatments, whereas self grafted all interspecific rootstocks except FL/RS had significantly lower concentrations of P than non grafted Concentration of K in the petiole was significant ly higher in FL/RS while Ca concentrations were significantly higher in FL/MA and FL/MU as compared with FL and FL/FL. Grafting did not significantly affect the concentrations of K and Ca in the stem. FL/MU had a significantly higher N concentration than F L. With respect to P in the stem, FL/BE and FL/MA had lower concentrations of P than FL. Overall, N concentrations were highest in the leaf blades despite the different grafting combination s whereas K concentrations were highest in the petiole and stem
71 P lant tissue nutrient concentrations tended to decrease at 63 DAT relative to 28 DAT especially for N and K. Some differences in nutrient concentrations were observed between grafted and non grafted plants FL/BE and FL/RS had higher concentration s of N in the l eaf blades than did FL, whereas FL/RS had a lower N concentration in the fruit (Table 3 5) Some grafting treatments with interspecific rootstocks also showed higher concentrations of K and Ca compared with non grafted plants. For instance, K concent rations were significantly higher in the stem of all grafted versus non grafted plants, and in the FL/MA and FL/RS plants, K levels were also higher in leaf blades. Similarly, plants grafted onto the interspecific rootstocks had significantly higher concen trations of Ca in the ir petiole s and stem s compared to non grafted Florida 47 FL/MU also had a higher Ca concentration in leaf blade s compared with FL. T he concentration of P in leaf blade s and petiole s was also significantly lower in the interspecific grafted plants other than RS/FL. At 120 DAT, few detectable differences remained in nutrient concentrations between grafted and non grafted plants ( d ata not shown). Plant N utrient A ccumulation At the earliest growth stage (28 DAT), accumulation of N, K, a nd Ca showed no significant difference between non grafted plants and plants grafted with interspecific rootstocks However, some grafted plants showed significantly lower accumulation s of P than non grafted plants (Table 3 6) Overall, plant accumulation of N and K was higher than P and Ca regardless of grafting. At 63 DAT, enhanced nutrient uptake as a result of grafting with interspecific rootstocks was in accordance with vigorous growth of grafted plants (Tables 3 3 and 3 7). Plants grafted onto intersp ecific rootstocks showed significantly higher
72 accumulations of N in the leaves compared to non grafted Florida 47 while N accumulation in self grafted plants did not differ significantly from non grafted plants. Similar ly, total accumulated K and Ca in the leaves were also significantly higher with grafted plants relative to non grafted and self grafted plants In contrast P accumulation was similar between grafted and non grafted plants. As expected, aboveground accumulations of nutrients were more con centrated in leaves and fruit compared to stems. In general, fruit of plants grafted with interspecific rootstocks had significantly higher levels of Ca than those of non grafted plants. At the end of the production season, grafted plants with certain inte rspecific rootstocks maintained higher levels of N, K, and Ca accumulation than FL/FL and FL, while little influence of rootstocks on the accumulation of P was observed (data not shown). Fruit Q uality A ssessment In general, the use of interspecific rootst ocks did not affect fruit quality attributes V alues were similar for pH, SSC TTA and ascorbic acid, lycopene, carotene and phenolic contents for non grafted tomatoes and fruit from most grafted treatments (Table 3 8) However, some changes in fruit quality did result from certain rootstocks at the mid fruit harvest in contrast to the early harvest. The T TA of fruit from FL/BE and FL/MA was significantly lower than that of FL during the second fruit harvest. In addition, FL/MA had a lower level of fruit hydrophilic phenolics than did FL. On the other hand, FL/MU showed an increased value of fruit lipophilic phen olic content relative to FL. Values of ascorbic acid, lycopene, and phenolic contents were higher at the mid fruit harvest compared to the early fruit harvest independent to the treatments
73 D iscussion In these greenhouse studies, grafting with interspecif ic rootstocks led to a significant increase in marketable tomato yield in comparison with non grafted plants. The overall result is consistent with previous reports of y ield improvement from use of specific rootstocks in tomato production in other regions Leonardi and Giuffrida (2006) for example, observed a 30% higher marketable yield in grafted tomato plants with rootstock vs. n on grafted plants Pogonyi et al. (2005) also noted a n increase of approximately 62% Lemance F1 onto Beaufo rt rootstock In addition, Lee and Oda (2003) report ed that Seokwang tomato onto Kagemusha rootstock increased fruit set by 39% and marketable fruit yield by 54% P ercentage of cull fruit was also reduced Our results indicated that the enhancement in total marketable yield by the interspecific rootstocks was primarily due to the increase of fruit number per plant and increased average fruit weight. T he present study also show ed that the percentage of cull fruit of the grafted plants was significantly lower than that of the non and self grafted plants (data not shown). Further, use of both non grafted and self grafted controls in this study showed that yield improvement in grafted plants could be ascribed to the specific rootstocks used rather than the grafting process per se The yield increase in grafted tomato plants was accompanied by enhanced vigor in comparison with non grafted plants. The greater leaf area of grafted plants probably helped improve photosynthesis and accumulation of aboveground biomass which in turn could lead to improved fruit yield. The positive influence of rootstocks on enhanced vigor of the tomato scion observed in this study concurs with previous studies that have demonstrated a consistent increase in growth as a result of grafting with vigorous
74 rootstocks (Colla et al., 2010; Di Gioia et al., 2010; Leonardi and Giuffrida, 2006; Yetisir et al., 2007) Although the vigorous root system of rootstocks has been suggested to cause the enhanced aboveground growth (Lee et al., 2010) the dry weight of roots measured in the present study did not diffe r significantly between grafted and non grafted tomato plants The overall architecture of these roots w as not appraised here, however, and roots of grafted plants actually appeared visibly smaller during early growth. Future research will be able to test the possibility indicated here, that root architecture may be more important than root dry weight on extent of rootstock scion interactions. Evidence from other species also supports this suggestion (Eissenstat, 1991) Similar observations also indicated that the positive influence of rootstocks on the nutrient contents of the aboveground plant tissues may depend upon the physical characteristics of the root system, such as l ateral and vertical development (Lee, 1994; Martnez Ballesta et al., 2010) (2008) reported that use of Maxifort as a tomato rootstock resulted in greater levels and accumulations of N, P, Ca, Mg, S, Fe, Mn, Zn, Cu, and B per unit dry weight Similarly, Leonardi and Giuffrida (2006) observed compared with self grafted plants. This may be directly linked to the increase d growth and development by grafted plants (Lee, 1994; Lee e t al., 2010) U nder conditions of the present study some of the rootstock s improved the nutri ent status of tomato scion Tissues had higher concentrations and accumulations of N, K and Ca In particular, the enhancement of Ca uptake by grafted plants mig ht have contributed to the reduced incidence of blossom end rot in tomato fruit (data for cull fruit not shown). However, the
75 leaf concentrations of P tended to be lower in grafted plants especially with Maxifort Multifort and Beaufort rootstocks, relative to non grafted Florida 47 such a decrease was attenuated by t he middle of the growing cycle. Although grafted plants may offer benefits o f disease control and yield enhancement, grafting will not be economically advantageous for tomatoes unless plants maintain premium fruit quality for successful marketing. In the present work, measurements of flavor related attributes (pH, SSC and TTA) as well as nutritionally important antioxidants (ascorbic acid, lycopene, carotene and phenolics content) did not reveal significant influence of rootstocks during early harvest Later at mid fruit harvest, some differences were detectable with certain rootstocks. This indicates that assessment of rootstock effects on fruit quality may need to be included multiple harvests to identify possible effects of developmental and/or environmental factors. Previous studies have yielded inconsistent findings on changes in fruit quality attributes as affected by grafting with specific r ootstocks. S ome quality attributes were improved by grafting with certain rootstock s (Davis et al., 2008a; Dorais et al., 2008; Fernandez Garcia et al., 2004b) while neutral or negative effects were also noted (Dorais et al., 2008; Ulukapi and Onus, 2007) More specifically, Fernandez Garcia et al. (2004b) observed significantly higher levels of lycopen e and carotene when tomato cultivars Responses were evident in tomato fruit under both non saline and saline conditions Although (Pogonyi et al., 2005) rootstocks our study showed a decrease of TTA only, and this
76 was limited to mid fruit harvest and only when or were used as rootstocks. Di Gioia et al. (2010) detected a consistent decline in vitamin C content in compared to non grafted plants at various harvest dates. Flores et al. (2010) speculated that in grafted plant s metabolic processes inherent to fruit quality may be largely species driven and controlled by the scion. However, some of the fruit qua lity attributes may also be influenced by the rootstock as a result of metabolites hormones, and RNAs moving to the scion through xylem or phloem (Lee, 1994; Rouphael et al., 2010) C onclusion s This study demonstrated improved marketable fruit yields as a result of grafting with vigorous rootstocks in tomato production under greenhouse conditions with sufficient water and nutrient supplies. G reater leaf area and biomass in the above ground plant tissues were also observ ed in grafted tomato plants compared with non grafted and self grafted scion plants. Both increased fruit number per plant and higher average fruit weight contributed to yield enhancement by grafted plants depending upon the rootstock used. Among the roots tocks evaluated in this study, Maxifort and Multifort showed the great est potential for increasing growth and yield of rooted and self grafted plants In general, grafted plants with interspecific rootstocks showed grea ter accumulation of N, K and Ca compared with non grafted and self grafted plants. In contrast, concentrations of P were consistently higher in non grafted than grafted tomato plants. Overall, grafting with the interspecific rootstocks evaluated here did n ot consistent ly influence detectable attributes of fruit quality such as SSC, TTA, or levels of vitamin C, lycopene, carotenoid, or total phenolics.
77 Table 3 1 Marketable yield, number of fruit and average weight of tomatoes from grafted plants with Fl orida 47 ( FL/ MA) and from non grafted Florida 47 (FL) plants during early, mid and late harvest periods in the 2007 2008 greenhouse experiment. Treatment Marketable yield (g/plant) Number of fruit (# /plant ) Average f ruit weight ( g/fruit) (Early harvest) (74 to 88 DAT) FL/MA 604.6 4.1 150.1 FL 687.9 4.9 140.3 Significance a NS NS NS (Mid harvest) (92 to 109 DAT) FL/MA 1277.0 9.6 132.7 FL 877.1 6.9 127.4 Significance a P <0.01 P <0.05 NS (Late harvest) (114 to 140 DAT) FL/MA 1208.3 8.5 142.3 FL 557.8 4.3 126.1 Significance a P <0.05 P <0.05 NS (Total harvest) FL/MA 3089.9 22.2 139.3 FL 2122.8 16.1 131.3 Significance a P <0.05 P <0.05 NS a Comparison between grafted and non grafted tomato plants N S, non significant.
78 Table 3 2 Marketable yield, number of fruit and average weight of fruit from non grafted tomato Florida 47 (FL) plants sel f grafted Florida 47 (FL/FL) plants and from Florida 47 either Beaufort (BE/FL), Maxifort (MA/FL), Multifort (MU/FL), or RST 04 105 (RS/FL) Analyses are from the 2009 greenhouse experiment. Treatment Marketable yield (g/plant) Number of fruit/plant Average fruit weight (g/fruit) Early harvest FL/ B E 1788.3 a 8.8 a 212.0 ab FL/MA 1817.0 a 8.3 a 216.8 ab FL/ MU 1800.0 a 8.0 a 226.8 a FL/ RS 1873.0 a 8.3 a 223.8 a FL/FL 1381.3 b 8.0 a 182.5 c FL 1438.0 b 7.5 a 198.8 bc Mid harvest FL/ BE 2454.8 ab 15.3 a 163.8 b FL/MA 2788.3 a 16.5 a 180.3 ab FL/ MU 2640.5 a 15.0 a 179.5 ab FL/ RS 2109.3 b 10.8 b 193.0 a FL/FL 1346.8 c 10.5 b 130.8 c FL 1354.0 c 10.5 b 128.0 c Late harvest FL/ BE 351.8 ab 5.0 a 70.0 b FL/MA 3 84.0 ab 5.5 a 79.0 a FL/ MU 582.8 a 6.8 a 86.0 a FL/ RS 396.0 ab 6.0 a 71.5 ab FL/FL 125.3 b 4.0 a 16.8 c FL 65.0 b 1.5 a 53.0 b Total harvest FL/ BE 4594.8 a 28.5 a 162.0 ab FL/MA 4989.0 a 29.3 a 173.8 ab FL/ MU 5023.3 a 29.8 a 174.3 ab FL/ RS 4377.8 a 25.5 ab 177.0 a FL/FL 2852.8 b 22.3 ab 135.0 c FL 2856.8 b 19.3 b 149.8 bc Means within a column followed by the same letter are not significantly different at P 0.05
79 Table 3 3 Leaf area and dry weight of or agans from non grafted tomato Florida 47 (FL) plants sel f grafted Florida 47 (FL/FL) plants and grafted plants with Florida 47 scions on rootstock s of either Beaufort (BE/FL), Maxifort (MA/FL), Multifort (MU/FL), or RST 04 105 (RS/FL) as m easured during the production period in the 2009 greenhouse experiment. Treatment Leaf area (100 cm 2 /plant) Dry weight (g/plant) Root Stem Petiole Blade Total aboveground 28 DAT (anthesis) FL/ BE 27.2 ab 2.0 b 6.9 ab 4.9 ab 14.8 ab 26.8 ab FL/MA 32. 2 ab 1.8 b 6.1 ab 5.6 ab 17.0 ab 28.8 ab FL/ MU 26.91 ab 1.8 b 7.0 ab 5.6 ab 15.1 ab 27.8 ab FL/ RS 32.65 a 2.1 b 6.8 ab 5.4 ab 17.0 ab 29.2 ab FL/FL 22.10 b 1.8 b 4.7 b 3.6 b 12.0 b 20.4 b FL 35.06 a 3.6 a 8.5 a 6.7 a 20.4 a 35.7 a 63 DAT (First fruit harvest) FL/ BE 199.86 a 8.6 a 41.0 a 35.0 ab 78.7 ab 305.2 b FL/MA 217.44 a 8.3 a 44.9 a 40.2 a 87.1 a 352.8 a FL/ MU 216.21 a 8.7 a 42.4 a 41.7 a 85.9 a 341.7 a FL/ RS 205.58 a 9.1 a 44.8 a 38.5 a 84.9 a 323.5 ab FL/FL 148.40 b 6.1 b 35.0 a 27.7 b 62.8 c 255.2 c FL 153.27 b 8.1 ab 43.5 a 30.9 b 65.7 bc 302.3 b 120 DAT (Final fruit harvest) FL/ BE 91.23 ab 19.4 a 81.5 abc 45.4 ab 59.6 ab 186.6 ab FL/MA 126.89 a 16.2 a 102.6 a 58.5 a 79.8 a 241.0 a FL/ MU 74.85 bc 15 .9 a 87.0 ab 42.1 ab 55.3 bc 184.4 a b FL/ RS 55.51 bc 15.9 a 60.7 bc 36.0 b 37.4 bc 134.2 bc FL/FL 42.09 c 7.2 b 54.4 c 33.2 b 33.0 c 120.7 c FL 66.66 bc 16.8 a 78.5 abc 40.9 ab 48.2 bc 167.6 bc Means within a column followed by the same letter are not significantly different at P 0.05
80 Table 3 4 Nutrient concentrations in tissues from non tomato plants self (FL/FL) plants and grafted plants with scions on either or RST 04 105 rootstocks. Analyses are from 28 DAT in the greenhouse experiment from January to June 2009. Blade Petiole Stem N P K Ca N P K Ca N P K Ca Treatment (g kg 1 DW) FL/ BE 59. 4 ab 8.17 b 36.09 a 28.09 b 33.00 a 7.79 c 70.25 ab 17.08 ab 32.05 ab 8.42 b 57.90 a 11.89 a FL/MA 60. 7 a 8.45 b 30.21 b 32.01 a 33.00 a 7.87 c 70.07 ab 18.01 a 30.22 b 8.63 b 62.82 a 12.40 a FL/ MU 60. 5 ab 8.49 b 34. 59 a 31.23 a 33.30 a 8.43 bc 69.90 ab 18.48 a 34.05 a 9.08 ab 55.40 a 12.35 a FL/ RS 58.05 ab 8.71 b 34.68 a 27.63 b 31.75 a 8.65 ab 72.02 a 17.12 ab 32.05 ab 8.95 ab 62.20 a 11.84 a FL/FL 56.58 ab 8.51 b 2 7.50 b 29.53 ab 34.60 a 8.21 bc 62.86 c 15.8 6 b 32.95 ab 8.70 b 56.83 a 10.94 a FL 56.22 b 9.60 a 29.15 b 27.66 b 31.60 a 9.19 a 64.25 bc 15.32 b 29.58 b 10.22 a 58.70 a 11.86 a Means within a column followed by the same letter are not significantly different at P 0.05
81 Table 3 5 Nutrient concentrations in tissue f rom non tomato plants self (FL/FL) plants scions on either of or RST 04 105 Analyses are 63 DAT in the greenhouse experiment f rom January to June 2009. Treatment Blade Petiole Stem Fruit N P K Ca N P K Ca N P K Ca N P K Ca (g kg 1 DW) FL/ BE 52 .2 a 7. 3 b 2 8.2 ab 41.4 b 20.4 ab 7.6 c 46.4 ab 21.3 ab 20.4 bc 8.8 bc 34.4 a 12.8 ab 29.6 a 6.1 a 38.7 a 1.3 ab FL/MA 46.9 bcd 7.6 b 30. 1 a 45.8 ab 21.3 ab 7.7 c 45.9 ab 23.3 ab 20.8 bc 8.9 bc 34.4 a 13.5 ab 24.8 bc 5.6 a 34.3 b 1.3 ab FL/ MU 46.8 cd 7.1 b 26 .0 ab 48.1 a 19.3 ab 7.8 c 44.6 ab 23.6 a 19.4 c 8.3 c 32.8 a 13.7 a 26.1 bc 5.8 a 37.2 ab 1.3 ab FL/ RS 50. 5 abc 9.1 a 29.4 a 41.1 bc 20.7 ab 9.5 b 51.2 a 20.8 b 22.5 b 9.8 ab 33.6 a 12.1 bc 24.1 c 6.2 a 35.4 b 1.5 4 a FL/FL 50.6 ab 9.9 a 29.0 ab 35 .8 c 22.0 a 10.7 a 51.3 a 17. 1 c 26.2 a 11. 1 a 32.5 a 10.8 cd 24.8 bc 5.9 a 35.5 b 1. 1 b FL 43.9 d 9.4 a 24.1 b 40.8 bc 17.30 b 9.4 b 40.1 b 16.6 c 21. 5 bc 8.7 bc 25.7 b 10.3 d 26.9 ab 5.8 a 34.9 b 1.0 b Means within a column followed by the same l etter are not significantly different at P 0.05
82 Table 3 6 Nutrient accumulated in tissues f rom non tomato plants self (FL/FL) plants scions on either of or RST 04 105 Analyses are from 28 DAT in the greenhouse experiment from January to June 2009. Treatment Leaves (g /plant ) Stem (g /plant ) N P K Ca N P K Ca FL/ BE 1.05 ab 0.15 b 0.88 a 0.51 a 0.23 ab 0.05 b 0.40 ab 0.08 ab FL/MA 1.22 ab 0.18 ab 0.91 a 0.64 a 0.18 b 0.05 b 0.38 ab 0.07 ab FL/ MU 1.10 ab 0.17 b 0.91 a 0.56 a 0.23 ab 0.06 ab 0.39 ab 0.08 ab FL/ RS 1.17 ab 0.19 ab 0.98 2 a 0.56 a 0.23 ab 0.06 b 0 .42 ab 0.08 ab FL/FL 0.82 b 0.15 b 0.66 a 0.47 a 0.18 b 0.04 b 0.30 b 0.06 b FL 1.35 a 0.25 a 1.02 a 0.67 a 0.28 a 0.09 a 0.51 a 0.10 a Means within a column followed by the same letter are not significantly different at P 0.05
83 Table 3 7 Nutrient accumulated in tissues from non RST 04 105 63 DAT in the greenhouse experiment from January to June 2009. Treatment Leaves (g /plant ) Stem (g /plant ) Fruit (g /plant ) N P K Ca N P K Ca N P K Ca FL/ BE 4.83 a 0.84 b 3.84 ab 3.98 b 0.84 a 0.36 a 1.43 ab 0.53 ab 4.22 a 0.85 ab 5.46 ab 0.19 ab FL/MA 4.79 ab 0.98 ab 4.45 a 4.86 a 0.94 a 0.40 a 1.56 a 0.60 a 4.29 a 0.96 a 5.92 ab 0.23 a FL/ MU 4.84 a 0.94 b 4.08 a 5.09 a 0.81 a 0.34 a 1.36 ab 0.58 a 4.25 a 0.94 a 6.05 a 0.22 a FL/ RS 5.09 a 1 .14 a 4.43 a 4.29 b 1.01 a 0.44 a 1.50 ab 0.54 a 3.51 ab 0.90 ab 5.16 bc 0.22 a FL/FL 3.82 bc 0.91 b 3.27 bc 2.70 d 0.92 a 0.37 b 1.11 b 0.37 b 3.07 b 0.73 b 4.33 c 0.13 c FL 3.42 c 0.90 b 2.82 c 3.18 c 0.84 a 0.39 a 1.08 b 0.15 ab 4.15 a 0.88 ab 5 .36 ab 0.15 bc Means within a column followed by the same letter are not significantly different at P 0.05
84 Table 3 8 Quality attributes of fruit from non grafted Florida 47 (FL) tomato plants sel f grafted Florida 47 (FL/F L) plants and grafted plant with Florida 47 scions on rootstock s of either Beaufort (BE/FL), Maxifort (MA/FL), Multifort (MU/FL), or RST 04 105 (RS/FL) as measured during early and mid fruit harvests in the 2009 greenhouse experiment. Treatmen t pH TTA a (% citric acid) SSC b ( o Brix) Ascorbic Acid (mg/100g FW) Lycopene (ug/g FW) carotene (ug/g FW) Hydrophilic phenolics (mg GAE c /100 g FW) Lipophilic phenolics (mg GAE/100 g FW) (Early harvest) FL/ BE 4.45 a 0.49 a 4.3 a 16.5 a 33.9 a 9.1 ab 10.8 a 3.8 a FL/MA 4.44 a 0.47 a 4.5 a 16 .4 a 31.9 a 8.3 ab 12.4 a 4.5 a FL/ MU 4.43 a 0.49 a 4.6 a 15.5 a 35.1 a 9.1 ab 11.2 a 3.9 a FL/ RS 4.44 a 0.49 a 4.3 a 15.3 a 33.9 a 9.1 ab 11.9 a 4. 2 a FL/FL 4.43 a 0.46 a 4.4 a 16.8 a 29.8 a 7.9 b 10.8 a 4.2 a FL 4.41 a 0.46 a 4.6 a 15.7 a 37.4 a 9.7 a 12.2 a 3.9 a (Mid harvest) FL/ BE 4.43 a 0.42 2 b 5.5 a 24.2 abc 42.5 a 9.4 a 17.1 a 6.2 bc FL/MA 4.42 a 0.42 b 5.3 a 19.9 bc 38.9 a 8.1 a 13.1 b 4.9 d FL/ MU 4.30 a 0.50 ab 5.3 a 16.6 c 43.4 a 9.6 a 16.1 ab 7.2 a FL/ RS 4. 34 a 0.46 ab 5.4 a 22.7 bc 36.7 a 7.9 a 18.1 a 6.3 abc FL/FL 4.40 a 0.46 ab 5.6 a 30.7 a 39.6 a 8.2 a 17.3 a 6.4 ab FL 4.39 a 0.57 a 5.7 a 27.3 ab 37.3 a 8.2 a 17.9 a 5.5 cd a TTA: Total titratable acidity. b SSC: Soluble solids content. c GAE: Gallic acid equivalents. Means within a column followed by the same letter are not significantly different at P 0.05
85 CHAPTER 4 ROOTST OCK EFFECTS ON NITRO GEN ASSIMILATION AND ENDOGENOUS HORMONE STATUS IN TO MATO PLANTS UNDER GR EENHOUSE CONDITIONS Introduction A number of rootstocks have been developed to improve production efficiency of solanaceous and cucurbitaceous vegetable crops, incl uding melon, watermelon, cucumber, tomato, eggplant, and cucumber. These rootstocks are known to confer resistance or tolerance to many soil borne diseases and enhance plant tolerance to environmental stresses. Moreover, some of the vigorous rootstocks hav e been shown to increase growth vigor and consequently improving fruit yields (Davis et al., 2008b; King et al., 2010; Lee, 1994; Lee et al., 2010) With respect to plant nutrition, g rafting with vigorous rootstocks has been shown to alter positively the nutritional status of the grafted plant. For example, absorption and translocation of phosphorus, nitrogen, magnesium, iron, and calcium are reported to show an increase in grafted plants as compared to the non graft ed plants (Davis et al., 2008b; Lee, 1994) This enhancement in nutrient status o f grafted plants has often been attributed to the difference in root morphology or architecture. Root characteristics which are suggested to play an active role in nutrient and water uptake include root length and density, number of root hair and their len gth, and root surface area (Martnez Ballesta et al., 2010) Among the plant essenti al nutrients, nitrogen (N) is required in the greatest amount and is among the most limiting factors to plant growth. The N status in plant is generally dependent on both soil N availability and plant N uptake and assimilation potential (Natali et al., 2009) Once taken up by plants either in the form of nitrate (NO 3 ) or ammonium (NH 4 + ), the nitrogen assimilation pathway begins with the reduction of
86 nitrate into nitrite (NO 2 ) and then in ammonium which is fixed into glutamine and glutamate ultimately serving as substrates for transamination reactions to generate other proteinaceous amino acids in mos t plants (Marschner, 1995; Tischner, 2000) The first step of this N reduction process, i.e., the conversion of nitrate to nitrite is catalyzed by the enzyme nitrate reductase (NR) and is also known to be the most limiti ng step of nitrate reduction and thus the NR activity level is often related to the N uptake capacity of the plant (Campbell, 1996; Glass et al., 2002) With the enhanced root system traits of the vigorous rootstock, it is expected that grafted plants may offer greater potential for nitrogen uptake and reduction as compared to non grafted plants. Studies by Ruiz and Romero (1999) demonstrated that grafted melon had lower concentrations of nitrate, higher nitrate reductase activity (NRA), and lower contents of total free amino acids and soluble proteins in the leaf tissue as compared to the non grafted control. Higher nitrate reductase and nitrite reductase activities were also found in leaves of grafted watermelon plants (Pulgar et al., 2000) Ruiz et al (2006a) demonstrated that enhancement in nitrogen use efficiency traits in graft ed tobacco as compared to non grafted plants was related to higher nitrate reductase activity and assimilation. These changes were suggested to contribute to differences in N utilization and assimilation between grafted and non grafted plants. In addition to the improved nitrogen assimilation as a result of grafting with specific rootstocks, modified endogenous hormone status is also proposed as an underlying basis for the enhanced growth and yield of grafted plants (Davis et al., 2008b) Different growth h ormones including auxin, cytokinins, and gibberellins are involved in regulating different aspects of plant growth and development. For example, cytokinins are a group
87 of hormones predominantly synthesized in the root tips and transported via the xylem to the shoots where they can exert a stimulatory or inhibitory function inherent to different developmental processes including root growth and branching, control of apical dominance in the shoot, chloroplast development, and leaf senescence (Dodd and Beveridge, 2006; Hirose et al., 2008) Another hormone of great importance in plant growth and development is auxin which is mostly synthesized in the young shoot organs and moves downward to impact fundamental aspects of plant development, such as vascular differentiation and root apical dominance (Aloni et al., 2006) Particularly with fruit trees, the relative concentrations of these two growth hormones, i.e., auxin:cytokinin ratios, have been shown as one of the key physiological mechanisms for dwarfing or other scion architecture of th e grafted fruit trees modified by vigorous rootstocks. For example, dwarfing rootstocks could control the apple scion vigor by reducing the downward transport of indole 3 acetic acid (IAA) from scion to root which results in a decrease in root growth and/o r cytokinin biosynthesis and therefore reducing the level of root produced cytokinins supplied to the scion in the xylem vasculature (Lockard and Schneider, 1981) It is likely that similar changes in hormonal balance including auxin and cytokinins may also exist and affect the rootstock scion interaction in the grafted vegetable plants including tomato Current literature on these physiological processes is mostly related to Cuc urbitaceae species, while there is a lack of information on the physiological changes in relation to the rootstock effects on improving growth and yields of grafted tomato plants. The objectives of this study were to determine whether or not the vigorous i nterspecific tomato hybrid rootstocks affect the nitrate assimilation and plant growth
88 hormone status in the root and leaf tissues of grafted tomato plants, and how these physiological modifications are correlated with the growth parameters, especially pla nt biomass and fruit yield. Materials and Methods Plant Material, Growth Conditions, and Experimental D esign This experiment was carried out in 2010 in a greenhouse located on campus at the University of Florida, Gainesville FL An determinate tomato cult ivar Florida 47 ( Seminis Vegetable Seeds, Inc. Saint Louis, MO) was grafted onto two c ommercially available interspecific tomato hybrid rootstocks Beaufort Seeds Inc., Lakewood, CO, USA) to form two different graft treatments, i.e., FL/BE and FL/MU. Self FL/FL) and non controls (1994) was used. Seeds of the rootstocks were sow n hypocotyl diameter of the scion and rootstock plants during the grafting process. After healing and acclimatization processes, survived grafted plants from the two rootstock scion combinations and the self grafted and non were transplanted on 15 Jan. 2010 into 11.4 L black plastic pots filled with Fafard growing mix 2 (Conrad Fafard, Inc., Agawam, MA). During the growing cycle, plants were provided with sufficient water and nutrients using a controlled fertigation system (see Chapter 3). A randomized complete b lock design was used with 4 replications and 5 plants in e ach treatment per replication. Gas Exchange M easurements Net photosynthetic rates, stomatal conductance, concentration of intercellular CO 2 and transpiration rate were measured on two separate fully expanded uppermost leaves
89 of two intact plants in each treatment per replication at anthesis (42 days after transplanting (DAT)) and first harvest (72 DAT), using a LI 6400XT portable, open flow photosynthesis system (LI COR Inc., Lincoln, N E ) Measuremen ts were made between 10 :00 am and 1 2 :00 p m 2 s 1 photosynthetic photon flux density using the 6400 02 light emitting diode (LED) light source (LI COR). The sample chamber CO 2 1 air, and the flow rate of air t 1 Temperature was maintained at 30 1C and relative humidity was maintained between 60 and 70%, similar to environmental conditions inside the greenhouse. Each leaf was equilibrated in the leaf chamber for at least 1 min before a measurement was taken Determination of I n V ivo Nitrate Reductase A ctivity (NRA) NRA was measured in both leaf and root tissues at three different growth stages, i.e., 42, 72, and 114 DAT, on two randomly selected plants per treatm ent and replicate. The uppermost fully expanded leaves and root tips from each selected plant were sampled for the measurement. Because NRA is reported to exhibit diurnal fluctuations (Aslam and Rl Rains, 2001) leaf tissue material was collected between 10:00 and 11:00 am on a sunny day. At each sampling day, leaf discs (5 mm diameter) from the middle part weighing about 150 mg fresh weight (FW) were harvested on the selected leaves per plant for the assay. NRA was assayed using the modified in vivo method by Jaworskyi (1971) Briefly, leaf discs and root tips segments collected were kept in ice and then transported to the laboratory where they were placed in 10 ml tubes. Then, 2 m l of the assay solution containing 2% (v/v) of 1 propanol, 50% (v/v) of 200 mM KH 2 PO 4 (pH 7.5), 20% (v/v) of 150 mM KNO 3 and 28% (v/v) of deionized water were added to each tube. The tubes were then vacuum infiltrated for 5 min with the vacuum
90 released at every 2.5 min. Then, tubes were immediately incubated in a shaking water bath at 31C in darkness for 1 hour. Thereafter, solutions were filtered into new tubes using the filter paper. A triplicate 1 ml of aliquots from each sample was removed from the fi ltered assay solution for the determination of NO 2 Nitrite concentrations were determined by adding 1 ml of 1% sulfanilamide solution in 1.5N HCl followed by 0.02% N (1naphthyl) ethylenediamine dihydrochloride (NED) in 0.2N HCl. After addition of sulfani lamide and NED, samples were incubated at room temperature for 30 min. Absorbance was then measured spectrophotometrically at 540 nm with a Shimadzu UV 160 (Shimadzu Corp., Kyoto, Japan) and NRA was reported as M NO 2 produced per hour per gram of fresh weight of tissue material [ M NO 2 (g FW) 1 h 1 ]. Nitrate and O rganic N D etermination At each sampling date, leaf and root samples per treatment were collected and dried in a force air oven at 60 o C for 48 hours. The dried samples were used for the determination of nitrate by specific ion electrode and total N by combustion. O rganic N was then calculated as a difference between total N and nitrate S oluble P rotein M easurement At each sampling, fresh samples of 200 mg l eaf discs and root tips w ere weighed and put in labeled eppendorf tube s and immediately frozen in liquid nitrogen and stored at 80 o C. During the analysis, samples were crushed and ground into powder, dipping in liquid nitrogen every few seconds to a void the tissues to defrost. Then, each sample was re suspended in the buffer solution while continued to be ground until solution is well homogenized. The buffer solution was made of 40 mM Tris HCl, pH 7.6 and 1 mM DTT, and protease inhibitor cocktail Sa mples were then votexed and centrifuged at 12000 rpm for 10 min at 4 o C. Supernatant of each sample was used to determine the
91 concentration of soluble protein using the Bradford G 250 r eagent (Bradford, 1976) by measuring the absorbance spectrophotometrically at 540 nm with a Shimadzu UV 160 (Shimadzu Corp., Kyoto, Japan). Bovine serum albumin (BSA) was used as standard for calculating the soluble protein concent ration. Determination of A mino Acid Composition and Quantification of P hytohormones Along with the samplings for soluble proteins, 200 mg of leaf discs and root tips were separately sampled and immediately frozen in liquid nitrogen and stored at 80 C for the determination of amino acids and plant hormones. During the analysis, 200 L of 90% methanol were added to each frozen sample and were finely ground using GenoGrinder ( BT&C Inc., Lebanon, NJ). Samples were then vortexed, and centrifuged at 4C 12,000 rpm for 10 min. The resulting supernatant was transferred into a new glass tube. This step was repeated by adding more 90% methanol and centrifuge until a volume of 1 mL supernatant was collected out of each sample. Afterwards, 0.5 mL of chloroform and 0.6 mL of water were added to the supernatant and mix briefly. Samples were allowed for phase separation in cold room for about 14 hours and the upper aqueous phase was then removed into new tube and allowed to dryness in a speed vac. Each sample was again re dissolved by adding 30 L of the loading buffer and shaking for 20 min with Thermomixer. Then samples were centrifuged at 4C, 12,000 rpm for 15 min. The new supernatant was used for measuring free amino acids and plant growth hormones using the reverse ph ase high performance liquid chromatography (HPLC) and tandem mass spectrometry. A nalysis of Aboveground Plant B iomass, Y ield, and Root C haracteristics Plant growth was destructively evaluated on one plant from each treatment per replication at 72 DAT Eac h sampled plant was separated into leaf blade, petiole, stem,
92 fruit, and root. Leaf area was measured with a LI COR 3100 leaf area meter (LI COR Inc., Lincoln, NE). Root characteristics including total r oot length, average root diameter, and root surface a rea were evaluated on each plant using a root scanning apparatus (EPSON color image scanner LA1600 + Toronto, Canada), coupled with the image analysis software WinRhizo 2008a (Regent Instruments, Quebec, Canada). Then, d ry weights of different plant parts (leaf blade, petiole, stem, fruit, and root ) were determined by drying the samples at 60 o C in a forced air oven for 72 h while fruit samples were dried for 120 h. Specific leaf area (SLA, m 2 leaf g 1 leaf) were then calculated from the measured leaf area and leaf mass for each sampled plant. Y ield and yield components were evaluated by harvesting fruit at the breaker ripeness stage or later on 4 plants per treatment and replication starting 72 through 112 DAT Total and marketable yield s marketable fruit number per plant and average fruit weight were determined and average fruit weight calculated. Statistical A nalys e s Analysis of variance of the measurements w as performed using the Generalized Linear Mixed Model program of SAS statistical software (versio n 9.2, Cary, NC ). Pearson correlation coefficients were used to evaluate relationships among growth parameters (biomass and yield) and physiological traits evaluated (NRA, nitrate, amino acids, etc.) measured. T he SAS procedure DISCRIM was first used to assess the homogeneity of covariance matrices of the standardized variables for the two groups of plants: grafted plants (FL /BE and FL/ MU) and non grafted and self grafted control plants ( FL and FL / FL). Because Box s M test for homogeneity of covariance matrices of the two groups
93 was not significant, the two treatment groups were pooled together for the overall correlation analysis. Results Plant B iomass, Yield and Y ield components, and Roo t C haracteristics Analysis of the aboveground plant biomass measured at 72 DAT (first harvest) showed that the two rootstocks significantly affected the plant growth as compared to the non grafted plants. On average, the two grafted treatments had a signi ficantly higher level of aboveground biomass (leaf blade, petiole, stem, and fruit) by approximately 39% relative to that of the non grafted plants (Table 4 1). Self grafting did not show a significant impact on plant growth as compared to the non grafted treatment; however, the aboveground biomass did not differ significantly between the self grafted plants and Beaufort significantly enhanced the total and marketable fruit y and self grafted treatments (Table 4 1). Specifically, 57 and 61 % of the total and marketable yield s, respectively, in c omparison with the averaged yields of non grafted and the self The increase in fruit yield is attributed to significantly higher number of fruit in grafted plants as compared to non and self grafted plants. The positive influence of the two rootstocks was also demonstrated on the root traits evaluated with significantly greater values of total root length, average root diameter, and root volume in grafted plants as compared to non and self grafted plants. For instance, as compared to the control plants, total root length values of grafted plant (Table 4 2) For the root surface area, it was significantly higher in FL/MU than FL/FL
94 and FL, while F L/BE did not differ significantly from FL. Moreover, considerable differences were observed among the treatments in terms of the patterns of allocation of root length, surface area, and volume per unit root mass. Compared to the non grafted and self grafte and 73 % greater root length per unit root dry mass (SRL), respectively, as well as 23 and 2 6 % higher root surface area per unit root dry mass (SRSA). In contrast, the grafted plants showed significantly lower ratios of root dry mass to root volume (M:V), indicating decreased root tissue density, as compared to the non grafted and self grafted plants. As for the ratio of root length to root volume, it was significantly lower in FL/MU and FL/ BE than FL/FL, while it did not differ significantly between the grafted plants with the two rootstocks and the non grafted treatment. Leaf Gas E xchange and Photosynthetic A ctivities Gas exchange parameters evaluated on single leaf showed few differences among the treatments as a result of the use of interspecific rootstocks (Table 4 3). The mean net CO 2 assimilation rate (Pn) was in the range of 26 30 mol CO 2 m 2 s 1 although the self grafted plants showed the highest values (Table 4 3). Stomatal conduct ance (Gs) and the transpiration rate (Tr) showed a significant increase in plants grafted plants. However, this change in the transpiration rate did not translate into any signif icant difference in the instantaneous water use efficiency (iWUE) among the treatments. Values of the iWUE were in the range of 3.86 4.14 mol CO 2 mmol 1 HO 2 independent of the treatments. In contrast, the specific leaf area (SLA) evaluated on a whole plan t basis demonstrated significantly greater values in plant grafted onto and non grafted plants. On average,
95 incre ase in SLA of 44 and 53%, respectively. This enhancement in SLA in grafted plants is indicative of possible increase in photosynthetic activity evaluated on the whole canopy basis and could lead to greater assimilate production in grafted plants as compare d to non grafted control plants. N itrate R eductase (NR) A ctivity and Nitrate C onc entrations in the Plant T issues I n vivo NR activity values in the leaf tissues were much higher than those in the root (Fig ure 4 1 ) indicating that most N reduction processes took place in the leaf tissues in contrast to the root tissues However, no clear trend in NR activity was observed as a result of grafting over the growth cycle. At 42 DAT, the NR activity in the vel of the enzyme activity as compared to the non and self grafted plants but the difference was not statistically significant at P lowest value of the enzyme activity at the same growth stage. Furthermore, except for levels of NR activity w as observed at 72 DAT (first harvest) as compared to the levels recorded in the vegetative stage (42 DAT) in the leaves. This increase was in the range of 8 to 30% across the treatments. At 72 DAT, the non grafted plants showed significantly a higher level of NR non grafted plants did not differ sig final harvest (114 DAT), similar levels of leaf NR activity were observed among different treatments. Regardless of the growth stages, similar levels of the root NR activity were observed in the gra fted and non grafted plants. Across all growth stages, the root NR
96 activity was about 0.31 [mole of NO 2 (g of FW) 1 h 1 ] on average; about 13% of the levels exhibited in the leaf tissues. Similar to the levels of NR activity, n itrate concentration s were also much higher in the leaves as compared to the levels in the roots irrespective of grafting and growth stages (Fig ure 4 2). The concentrations of nitrates in the leaf and root tissues decreased over time, while few differences were found between grafted and non grafted treatments. At 72 DAT, the leaf nitrate level was significantly lower in the plants grafted Concentration s of Organic Nitrogen, Amino Acids, and Soluble P roteins Changes in the amount of reduced nitrogenous compounds did not show any clear effect of grafting with interspecific rootstocks. Significant difference in the leaf organic N levels was observed only between and non grafted plant at 72 DAT, with the highest levels of organic N in grafted vs. non grafted plants (Table 4 4). At each growth stage, the levels of organic N in the root tissue were lower than in the leaves. With respect to the levels of free amino acids and soluble proteins in the l eaves and roots, they were generally similar between grafted and non grafted plants free amino acids in the leaves as compared to all other treatments. At 42 DAT, the soluble protein content in the roots was significantly lower in plants grafted onto grafted treatment. There was a decreasing trend in the levels of amino acids and soluble proteins over time.
97 Auxin, Cytokinin, and G ib b erellic Aci d Levels in Leaf and Root T issues Analysis of the different hormones quantified in the leaf and root tissues revealed that the hormone levels were generally greater in the leaf tissues than in the root. Some differences between grafted and non grafted plan ts were observed depending on the growth stage and the type of hormone (Table 4 5). At 42 DAT, total auxin content in the treatments. The level of cytokinin in the leav es was also significantly higher in grafted grafted plants. At 72 DAT, plants grafted onto acid as compared to other treatments. The auxin to cytokinin ratio at 72 DAT was grafted Some significant changes in hormone levels in the root were observed with auxin and lower content of auxin as compared to non grafted and self grafted plants. At 72 DAT, a significant increase in the level of gibberelic acid was found with plants grafted onto and self grafted plants. In term of the relative content of individual types of cytokinins identified across the two sampling dates, there was a slight variation in the presence of hormones between the grafted and non grafted plants (data not shown). In the order of most to least abundant, different types of cytokinins identified within the leaf tissues of plant grafted N6 isopentenyladen os ine (N6 ISO1), trans zeatin (Z TN), kinetin (KIN) and zeatin ribulose (ZR), while in the control (non and self grafted) plants, the order is as follows: KIN, N6 ISO1, ZTN, and ZR. In contrast, relative
98 contents of individual types of auxins appeared to be similar between grafted and non grafted plants, with the most abundant to least abundant order as follows: indole 3 acetate (IAA), methyl indole 3 acetate (meIAA), indole 3 butyric acid (IBA), and indole 3 carboxy acid (ICA). Relationship between Plant Growth Performance and Physiologic al P rocesses Using pooled data of all the treatments, Pearson correlation analysis of the growth and physiological parameters evaluated in this study demonstrated considerable differences in terms of the type and significance of the relationships among dif ferent measurements (Table 4 6). Fruit yield and shoot biomass showed significantly positive correlations with the specific root length (r = 0.68 and r = 0.59) respectively. Although not statistically significant, positive relationships were also observed between these two growth parameters and the organic N. In contrast, negative correlations were observed between yield and biomass with nitrate reductase activity and nitrate concentrations in the leaves. Levels of the endogenous growth hormones in the leav es appeared to positively correlate with the fruit yield and biomass. A significant correlation between fruit yield and gibberelic acid was detected, while aboveground biomass significantly correlated with the level of cytokinins in the leaves. Discussion G rafting with specific rootstocks is often shown to induce enhancement in plant growth and fruit yield of the vegetable scion. O ur study with the tomato variety Florida yield compared to the non rel evant to water and nutrient uptake such as specific root length (SRL). As evi den ced
99 in studies on citrus trees (Eissenstat, 1991) rootstocks with high SRL and a greater root length density were able to extract water more rapidly and also take up ino rganic nutrients including nitrate more efficiently, in contrast to those with low SRL. Therefore with these root traits of the rootstocks in grafted tomato plants, an increase is expected in NO 3 absorption, upward transport, and accumulation in the scion thereby stimulating the nitrate reductase (NR) activity and therefore NO 3 assimilation. NR is an inducible enzyme whose activity is mostly induced by the level of nitrate concentrations in the medium and is thought to control the most limiting step in N assimilation (Stitt, 1999; Tischner, 2000) This is in line with results by Ruiz and Romero (1999) and Ruiz et al. (2006a) who ob served that the use of certain rootstocks in melon and tobacco plants increased foliar NO 3 reduction as compared to the non grafted plants. In our study, however, no clear influence of the specific rootstocks on the nitrogen assimilation was noticed throu ghout the growth cycle This contrasting trend was in part in agreement with results obtained by Colla et al. (2010) on melon which showed that at high nitrate conditions, there wa s no significant difference among grafted and non grafted treatments for NR activity, whereas under low nitrate concentrations, all grafted plants had higher activities of NR. This result could suggest that the positive influence of specific rootstocks on the physiological processes such as nitrogen assimilation is more pronounced when plants are grown under less than optimum environmental conditions. Given that sufficient supply of N was provided equally to the plants in our study, it is possible that the activity of NR may not differ greatly between grafted and non grafted plants.. This is reflected in the values of nitrate concentrations observed in this study which showed no consistent effect of rootstocks on leaf nitrate concentrations as
100 compared with the non grafted and self grafted plants. A negative correlation is often described between the levels of NR activity and the nitrate concentrations in the leaves, with plants characterized by higher rates of NO 3 reduction showing higher demand for NO 3 an d increased uptake from the medium. Our results showed a Pearson correlation coefficient of 0.01 between the nitrate concentrations and the level of NR activity; this may also account for the insignificant difference observed in nitrate concentrations and NR activities between the treatments. Levels of the leaf NR activity determine the levels of reduced N (organic N) available in the plant for growth and development, with high levels of NR activity resulting in increased level of reduced N. Previous stud ies have shown that the grafted plants presented higher total organic N concentrations as a result of enhanced NR activity (Pulgar et al., 2000; Ruiz and Romero, 1999) In our study, the increased level of organic N of grafted plants as compared to the non grafted plants was only observed proteins are the main end products of NO 3 assimilation (Barneix and Causin, 1996) High level of NR activity is shown to result in increased production and use in these nitrogenous compounds. In our study, few differences in concentrations of free amino acids and soluble proteins were observed between grafted and non grafted plants grafted and self grafted plants. Regardless of grafting, in this study levels of NR activity and other nitrate assimilation processes in the roots w ere lower than those in the leaves. This concurs with previous studies report ing higher levels of NR activity in leaves for several crop
101 species as compared with the roots (Black et al., 2002; Johnsen et al., 1991) Sanchez et al., (2004) In addition to improved nitrogen assimilation, changes in endogenous plant growth hormone levels, especially auxin, cytokinin s, and gibber ellic acids have been suggested to contribute to enhanced gro wth and yield of grafted plants The importance of these hormones in plant growth and development regulation in general has been extensively demonstrated in many previous works (Hirose et al., 2008) ; however, only few of them have studied the rootstock effects on the link between the horticultural benefits and hormo nal status in grafted vegetable s Yamasaki et al. (1994) observed that the cytokinin concentration, mainly ribosylzeatin, in xylem exudates was significantly greater in watermelon grafted onto squash rootstocks as compared to the non grafted watermelon plants. Simila rly, Seong et al (2003) have shown that cucumber plants grafted on pumpkin rootstocks had a 2.2 times higher trans zeatin content in the root and as a result, significantly higher dry matter than non grafted cucumber. These results were in general attributed to differenc es in root architecture characterized by greater quantity of root tips in rootstocks as compared to that of non grafted plants. Based on fruit trees, Kamjol et al. (1999) suggested that these changes in hormone levels a s a result of grafting with specific rootstocks may result from differences in the capacity of the roots to upload cytokinins into the transpiration stream, and the amount of cytokinins synthesized by the roots or the number of roots synthesizing cytokinin s. In th e present study, we found that rootstocks resulted in somewhat contrasting effects on the levels of endogenous hormones evaluated at the two growth stages. Partially in line with previous results, our study showed that concentrations of cytokinin s and gibberellic acid
102 in the leaves at 72 DAT were significantly greater in plant s grafted onto as compared to all other treatments, while 42 DAT the level of auxin in the leaves was n that of non grafted and self The positive correlation coefficients between the growth parameters (yield and biomass) and the concentrations of auxin and cytokinins, although not always signifi cant, seem to corroborate the importance of phytohormones, especially cytokinins and gibberelli c acids in regulating plant growth and development. Further, our results show a slight variation in the types of cytokinins observed within the grafted vs. non g rafted plants and over time, with N6 ISO1 of the highest level in the leaves of plants grafted and self grafted plants. The genotypic variation in relative content o f different cytokinin s was also observed between two lines of cotton (an early senescing line and a late senescing line) evaluated by Dong et al. (2008) Using the late senescing line (K2) as rootstock improved the accumulation of zeatin and its riboside (Z+ZR) as well as isopentenyl and adenine (iP+iPA) in leaves, and thus delayed leaf senescence. In contrast, the early senescence line (K1) rootstock re duced the accumulation of these hormones in the leaves, and therefore accelerated the leaf senescence. Similarly, the differential composition of cytokinins was also reported on grafted apples with the dwarfing and vigorous rootstocks. Kamboj et al. (1999) observed that zeatin was the zeatin riboside was the predominant cytokinin in xylem sap from the more invigorating rootstock MM. 106
103 Moreover, b oth auxin and cytokinin have been known for a long time to act either synergistically or antagonistically to regulate several important developmental processes in plants such as the formation and maintenance of branching and root growth (Su et al., 2011) In a recent review on hormonal signaling in rootstock scion interactions, Aloni et al. (2010) presented whereby a feedback loop exists in which a decrease in basipetal flow of auxin from the shoot stimulates the synthesis and export of cytokinins from the root. The u pward movement of cytokinin in the xylem sap induces an increase in the synthesis and translocation of auxin out of the shoot apex which in turn reduces cytokinin levels in the xylem sap. This concept involving the auxin to cytokinin ratio has been used as a key physiological mechanism to explain the dwarfing behavior of rootstocks of fruit trees Most recently, van Hooijdonk et al. (2011) demonstrated a putative mechanism Auxin (IAA) transport is decreased in the shoot leading to an increase in cytokinin level in the root which results in a change in the flow of assimilate partitioning within the plant parts leading to increased proportion in the root at the e xpense of the shoot. In contrast, evidence was reported for grapevines with rootstocks classified as high vigor showing increased levels in sap cytokinins (Nikolaou et al., 2000) This hormone message concept can be possibly used to account for the difference in the growth between grafted and non grafted vegetable plants. However, our results did not reveal a clear impact of the vigorous interspecific hybrid rootstocks on the auxin:cytokinin ratio in the
104 was significantly higher than those in non grafted plants and plants grafted onto grafted plants. F urther studies need to evaluate the flow of auxin in the shoot and cytokinins in the roots by continuously collecting xylem sap from the shoots and from the root pressure exudate in order to better establish the dynamic of auxin to cytokinin ratio and to evaluate how this ratio is correlate d with plant growth parameters. The influence of the grafting process in addition to the rootstoc k effects also need to further explored. Furthermore studies with fruit trees such as apple have also demonstrated that the relative contents of endogenous gibberellins acids (GAs) may influence the rootstock induced scion dwarfing with the suggestion th at GA concentration s increased with increasing rootstock vigor (Kamboj et al., 1999) It is suggested that decreased basipetal transport of auxin ( IAA ) may limit the amount of root produced gibberellins that is s ynthesized and supplied to the scion for shoot growth. In this study, plant s grafted on concentration of GA ( GA3 + GA4) in the leaf tissue at the reproductive stage ( 7 2 DAT) as compared with other treatments More int erestingly, the GA levels in the root tissues of the two vigorous rootstocks were significantly greater than non grafted and self grafted control plants. The correlation analysis also revealed a significantly positive relationship between yield and gibbere lli c acid levels Conclusion s In this greenhouse study showed significantly greater aboveground biomass and fruit yields as well as improved root characteristics such as root length compared to the non gr afted and self
105 However, differences in other physiological measurements, including nitrate reductase activity and levels of nitrate, organic nitrog en, amino acids, and soluble proteins, were less evident between grafted and non grafted plants and did not show any consistent trend. Moreover, analysis of endogenous hormones such as auxin, cytokinin, and gibberellins did not exhibit consistent effects o f the two vigorous rootstocks used. Future research is warranted to elucidate how different rootstocks may affect nitrate assimilation at varying supply levels of nitrogen to plants. The potential rootstock influence on hormonal modifications of grafted to mato plants also deserves more in depth examination.
106 Table 4 1 Fruit yield s and aboveground biomass of grafted and non grafted tomato plants Treatment Yield (kg /plant ) Marketable fruit (kg/plant) Biomass z (g /plant ) Total Marketable Mean fruit we ight (g) Fruit n umber FL/BE 6.00 a 5.63 a 163.31 a 35 a 433.77 ab FL/ MU 6.36 a 6.06 a 169.75 a 36 a 469.13 a FL / FL 4.07 b 3.75 b 160.32 a 24 b 379.15 bc FL 4.03 b 3.77 b 162.04 a 23 b 325.34 c z Aboveground biomass (leaf blade, petiole, stem, and fruit) was measured at first harvest, i.e., 72 DAT. Means within a column followed by the same letter are not significantly different at P 0.05 Florida 47 grafted onto Beaufort Florida 47 grafted onto Mul tifort FL/FL: self
107 Table 4 2 Total root length (RL), average root diameter (AD), root surface area (RS) and root volume (RV), specific root length (SRL), specific root surface area (SRSA), root mass: root volume ratio (M:V), and root length: root volume ratio (L:V) of grafted and non grafted tomato plants at 72 DAT. Treatment RL (m/plant) AD (mm/plant) RS (cm 2 /plant) RV (cm 3 /plant) SRL (cm/g) SRSA (cm 2 /g) M:V (g/cm 3 ) L:V (cm/cm 3 ) FL /BE 6339 a 1.648 a 2142.9 ab 87.8 a 809.8 a 275.4 a 0.09 b 72.46 b FL/MU 6822 a 1.826 a 2243.9 a 103.5 a 872.3 a 280.3 a 0.08 b 72.70 b FL/FL 3789 b 0.8446 b 1694.0 c 35.9 b 480.9 b 215.8 b 0.24 a 107.4 a FL 4292 b 0.3964 b 1877.8 bc 45.3 b 524.9 b 230. 1ab 0.18 a 94.11 ab Means within a column followed by the same letter are not significantly different at P Florida 47 grafted onto Beaufort Florida 47 grafted onto Multifort FL/FL: self
108 Table 4 3 Net CO 2 assimilation rate on a leaf area basis ( P n ) stomatal conductance ( G s ), sub stomatal CO 2 concentration (C i ) transpiration rate (Tr ) intri nsic water use efficiency (iWUE ), and specific leaf area (SLA ) of grafted and non grafted plants. Treatment P n (mol CO 2 m 2 s 1 ) G s (mol H 2 O m 2 s 1 ) C i (mol mol 1 ) Tr (mmol m 2 s 1 ) iWUE (mol CO 2 mmol 1 H 2 O) SLA (cm 2 g 1 ) FL/BE 28.80 ab 0.84 ab 287.77 a 6.97 ab 4.14 a 284.93 a FL/ MU 28.79 ab 0.89 a 290.23 a 7.45 a 3.86 a 303. 13 a FL / FL 30.56 a 0.97 a 289.25 a 7.49 a 4.12 a 194.97 b FL 26.26 b 0.74 b 288.16 a 6.56 b 4.02 a 201.35 b Florida 47 grafted onto Beaufort Florida 47 grafted onto Multifort FL/FL: self Means within a column followed by the same letter are not significantly different at P
109 Table 4 4 Tissue concentrations of organic N, amino acids, and soluble proteins of grafted and non grafted tomato plants at 42, 72, and 114 DAT. Treatment Organic N (mg g 1 DW) Amino Acids (mg g 1 FW) Sol uble Proteins (mg g 1 FW) 42 72 114 42 72 114 42 72 114 Leaves FL/BE 45.0 a 39.1 ab 42.7 a 1.3 a 0.96 a 0.10 a 21.46 a 20.63 a 16.08 a FL/MU 42.1 a 41.8 a 44.4 a 0.9 a 0.84 a 0.06 b 21.31 a 21.09 a 12.20 a FL/FL 47.1 a 39.7 ab 39.4 a 1.1 a 0. 71 a 0.05 b 21.82 a 22.30 a 10.76 a FL 40.1 a 35.7 b 38.7 a 1.0 a 0.95 a 0.05 b 22.08 a 22.73 a 9.92 a Roots FL/BE 24.7 a 19.8 a 20. 3 a 0.27 a 0.69 a 0.06 a 2.52 b 2.29 a 3.60 a FL/MU 26.8 a 18.6 a 18.5 a 0.24 a 0.45 a 0.08 a 3.55 ab 2.82 a 3.5 6 a FL/FL 26.7 a 21.4 a 20.7 a 0.3 a 0.71 a 0.09 a 3.46 ab 2.64 a 3.82 a FL 28.1 a 21.9 a 20.1 a 0.3 a 0.61 a 0.08 a 3.93 a 2.85 a 3.28 a Florida 47 grafted onto Beaufort Florida 47 grafted onto Multifort FL/FL: self grafted Means within a column followed by the same letter are not significantly different at P
110 Table 4 5 Auxin, cytokinin, gibberellic acids concentrations, and auxin:cytokinin ratio (AUX:CTK) in leaf and root tissues of grafted and non tomato plants at 42 and 72 DAT. Treatment Auxin (ng g 1 F W) Cytokinin (ng g 1 FW) Gibberellin (ng g 1 FW) AUX:CTK 42 72 42 72 42 72 42 72 Leaves FL/BE 1064.5 a 413.3 ab 0.46a 0.1 4 b 27.8 a 28.40 b 2730.3 a 3625.3 a FL/MU 428.6 b 343.8 ab 0.2 4 ab 0.58 a 24.5 a 50.15 a 2895.9 a 648.6 c FL/FL 492 .7 b 429.9 a 0.13 b 0.19 b 23.2 a 28.45 b 4013.8 a 2527.3 ab FL 558.6 b 231.8 b 0.18ab 0.15 b 26.4 a 25.43 b 3092.3 a 1693.8 bc Roots FL/BE 42.7 ab 52.1 a 0.06 a 0.05 a 5.1 a 5.82 a 758.6 a 966.2 a FL/MU 35.0 b 75.3 a 0.05 a 0.0 6 a 3.8 ab 6 .17 a 728.2 a 1267.6 a FL/FL 57.2 a 66.6 a 0.06 a 0.05 a 2.7 b 2.54 b 925.6 a 1312.1 a FL 57. 5 a 63.1 a 0.06 a 0.05 a 2.8 b 0.93 b 954.3 a 1182.5 a Florida 47 grafted onto Beaufort Florida 47 grafted onto Multifort FL/FL: self Means within a column followed by the same letter are not significantly different at P
11 1 Table 4 6 Pearson correlation of fruit yield (YLD), aboveground biomass (BS), specific root length (SRL), nitrate reductase activity (NRA), organic nitrogen (OGN), and concentrations of nitrate (NIT), soluble proteins ( PRO), amino acids (AA), auxins (AUX), cytokinins (CTK), and gibberrellic acids (GA) averaged across grafted and non YLD BS SRL NIT OGN NRA PRO AA AUX CTK GA YLD 1.00 0.41 BSs 0.41 1.00 SRL 0.68** 0.5 9* 1.00 NIT 0.50* 0.54 0.19 1.00 OGN 0.26 0.43 0.03 0.51 1.00 NRA 0.007 0.19 0.05 0.01 0.13 1.00 PRO 0.55* 0.05 0.47 0.09 0.14 0.19 1.00 AA 0.09 0.13 0.31 0.64 0.14 0.04 0.19 1.00 AUX 0.18 0.28 0.30 0.17 0.07 0.14 0.07 0.30 1.00 CTK 0.39 0.57* 0.48 0.65 0.49 0.34 0.11 0.34 0.25 1.00 GA 0.56* 0.18 0.32 0.21 0.32 0.71 0.47 0.04 0.14 0.07 1.00 *, ** Significan ce at P 0.05 and P 0.01, respectively
112 Fig ure 4 1 In vivo activity of nitrate reductase in leaf (A) and root (B) tissues of grafted and non grafted tomato plants at 42, 72, and 114 DAT. Florida 47 grafted onto Beaufort Florida 47 grafted onto Multifort and FL: non
113 Fig ure 4 2 Nitrate concentrations in leaf (A) and root (B) tissues of grafted and non grafted tomato plants at 42, 72, and 114 DAT. Florida 47 grafted onto Beaufort Flori da 47 grafted onto Multifort and FL: non
114 CHAPTER 5 YIELD, IRRIGATION WA TER AND NITROGEN USE EFFICIENCY OF FIELD GROWN GRAFTED TOMATO WITH DRIP IRRIGATION Introduction Grafting is currently practiced worldwide on many high value cucurbitaceous and solanaceous crops such as watermelon ( Citrullus lanatus ), melon ( Cucumis melo ), cucumber ( Cucumis sativus ), ( Solanum lycopersicum L.), eggplant ( S. melongena ), and pepper ( Capsicum annuum ) for both open field production and protected cul ture (Davis et al., 2008; Lee, 1994; Lee and Oda, 2003; Lee et al., 2010) Vegetable grafting has proven to be an innovative and effective technique for controlling soilborne diseases such as fusarium wilt (caused b y Fusarium oxysporum ), verticillium wilt (caused by Verticillium dahliae ), s outhern blight (caused by Sclerotium rolfsii ), and bacterial wilt (caused by Ralstonia solanacearum ) (Lopez Perez et al., 2006; Mcavoy et al., 2012; Rivard and Louws, 2008; Rivard et al., 2010; Rivard et al., 2012) as well as root knot nematodes (Barrett et al., 2012; Bausher, 2009) Grafting with certain rootstocks has also been shown to improve plant tolerance to abiotic stresses such as high salt and low temperature (Fernandez Garcia et al., 2004b; Schwarz et al., 2010) Previous studies have demonstrated that in addition to disease management, plant vigor and yield often increase as a result of grafting with vigorous rootstocks. In the case of tomato production, grafted plants can increase marketable yield by 20% to 62% over non grafted plants, depending on scion rootstock combinations and production conditions (Di Gioia et al., 2010; Lee and Oda, 2003; Leonardi and Giuffrida, 2006; Pogonyi et al., 200 5) The improved productivity of grafted plants has been attributed by some studies to the intrinsic vigor of the rootstock and the scion rootstock interaction, which in turn exerts positive influence on plant nutrient and water
115 absorption, endogenous hor mone balance, N assimilation, and photosynthetic processes (Aloni et al., 2010; Dong et al., 2008; Kato and Lou, 1989; Seong et al., 2003; Stegemann and Bock, 2009; Yamasaki et al., 1994) Given the physiological an d phenotypic modifications caused by grafting with selected, vigorous rootstocks, it is likely that irrigation and fertilization management for maximizing crop yield may differ between grafted vs. non grafted vegetable production. In addition to the enhanc ed fruit yields with grafted plants, plants grafted onto vigorous rootstocks also use irrigation water and fertilizer more effectively for producing marketable fruit yields (Colla et al., 2011; Rouphael et al., 2008) Recent research has addr essed the growth and yield performance of grafted plants in response to different levels of water or nutrients (Colla et al., 2010; Lykas et al., 2008; Rouphael et al., 2008) However, m ost of these studies focused on cucurbits, especially under greenhouse conditions, with an emphasis on either irrigation or fertilization management rather than a combination of both In the present study, responses to both nutrients and water are examined together and addressed by co mparing grafted vs. non grafted tomato production under field conditions. This information is important for aiding recommendations to producers who are increasingly adopting grafting for open field tomato production on a larger scale. Furthermore, conditio ns in Florida, where the commercial production of fresh market tomato occurs, include sandy soils with low water and nutrient retention capacities (Hartz and Hochmuth, 1996; Hochmuth, 1992; Locascio, 2005) The situ ation in Florida is compounded by recommendations for N fertilization based primarily on crop type rather than soil tests (Olson et al., 2010) Therefore, the objectives of this study were to 1) determine the influence of irrigation
116 regimes and N fertilization application rates on yield and yield com ponents of grafted tomato plants grown with drip irrigation in Florida sandy soils; and 2) characterize the influence of grafting with vigorous rootstocks on irrigation water and nitrogen use efficiency under these conditions. Materials and Methods Graftin g and T ransplant P roduction T he field grown, determinate tomato cultivar Florida 47 ( Seminis Vegetable Seeds, Inc., Saint Louis, MO) was used as the scion and grafted onto t wo commercially available, interspecific hybrid rootstocks Lakewood, CO) These two rootstocks are currently the most widely used tomato rootstocks in the United States. Rootstock seeds were sown on 21 Feb., 2010 and 19 Feb., 2011, 2 stem diameters at the time of grafting as t he rootstock cultivars tend to germinate and emerge more slowly Plants were splice grafted (Lee, 1994) on 16 Mar., 2010 and 20 Mar., 2011, when 5 6 true leaves were present. Grafted plants were immediately placed in a closed healing chamber equipped with two humidifiers and an auto control air conditioning system for healing the grafts, where temperature was maintained at 25 3 o C and relative humidity above 85% Light and ventilation were introduced gradually after a dark period of 4 days. Twelve days af grafted onto grafted onto U ) were completely healed and ready for transplanting to the field. Non grafted transplants provided the co ntrol treatment.
117 Field P roduction The field experiments were conducted during the spring seasons of 2010 and 2011 at the Suwannee Valley Agricultural Extension Center in Live Oak, FL (30.17 N, 82.59 W). The soil type was a Blanton Foxwort Alpin Complex s andy soil (Natural Resources Conservation Service, 2006) I n both years, field plots were disk ed and plow ed, five weeks before transplanting followed by soil fumigation using Telone C 35 (Dow AgroSciences, LLC; Indianapolis, IN) at the rate of 196.4 L ha 1 The field was fumigated to eliminate interference of soi lborne pest factors. Mehlich 1 soil test results conducted prior to field preparation showed that no additional P was needed while K fertilization at the rate of 205 kg ha 1 was required. Three weeks before transplanting, 13N 1.7P 10.8K (Mayo Fertilizer In c, Mayo, FL) fertilizer was applied at a rate providing 56 kg N ha 1, 7.3 kg P ha 1, and 46.5 kg K ha 1 to all plots during bed preparation. Grafted and non grafted plants were transplanted to raised beds with plastic mulch and drip irrigation on 29 Mar. 2 010 and 1 Apr. 2011. Beds were 0.71 m wide and spaced 1.52 m apart (from middle to middle) with 0.46 m in row spacing for open field tomato production. In both years, a split plot design with four replications was used. The whole plot treatments, i.e., 12 factorial combinations of two irrigation regimes and six N fertilization rates, were arranged in a randomized complete block design The subplot treatments included the two grafting treatments FL/BE and FL/M U and the non (FL) as contro l, all randomized within each whole plot. There were 12 plants for each treatment combination per replication in both 2010 and 2011. The two irrigation regimes included: 1) 100% irrigation regime based on the current University of Florida Institute of Food and Agricultural Sciences (UF/IFAS) recommendation for field production of round
118 tomatoes in sandy soils in Florida i.e., 9354 L/ha/day/string (Olson et al., 2010) and 2) 50% irrigation regime corresponding to 4677 L ha/day/string. The stake and weave method was used for trellising the tomato plan ts. T used in field tomato production to denote the growth stage of staked tomato plants; final harvest. The six N fertilization rate s were 56, 112, 168, 224, 280, and 336 kg ha 1 which represented 25%, 50 %, 75 %, 100%, 1 25 %, and 1 50 %, respectively, of the currently recommended total N application rate of 224 kg N ha 1 (a preplant application at 56 kg N ha 1 included) for field productio n of irrigated, round tomato in sandy soils in Florida (Olson et al., 2010) Except for the 56 kg N ha 1 rate which only included a preplant application of 13N 1.7P 10.8K a mmonium nitrate ( 34N 0P 0K Mayo Fertilizer Inc, Mayo, FL) was injected weekly through the drip tape starting one week after tr ansplanting (WAT) to provide the remaining amount of N for other fertilization rate treatments. The weekly injected amount s of N for each of these five N fertilization rates during 1 to 2 WAT, 3 to 4 WAT, 5 to 11 WAT, 12 WAT, and 13 WAT respectively, were as follows: 1) 3.0 4.0 5.0 4.0 and 3.0 kg ha 1 ; 2) 6. 1, 8.0 10.0 8.0 and 6.1 kg ha 1 ; 3) 9.1 12.0 15.0 12.0 and 9.1 kg ha 1 ; 4) 12.1 16.0 19.9 16.0 and 12.1 kg ha 1 ; and 5) 15.1 20.0 24.9 20.0 and 15.1 kg ha 1 Potassium chloride (Dyna Flo 0 N 0 P 15 K Chemical Dynamics Inc, Plant City, FL) was also applied through fertigation to provide each treatment with amount of K needed after accounting for the preplant application based on the soil test. The weekly injected amounts of K during the g rowing season were as follows: 11.8, 9.3, 14.3, 9.3, and 6.7 kg K ha 1 during 1 to 2 WAT, 3 to 4 WAT, 5 to 11 WAT, 12 WAT, and 13 WAT, respectively. Other cultural practices, including pest
119 control, followed current recommendations for commercial field tom ato production in Florida (Olson et al., 2010) Further, daily rainfall data collected by a weather station of the Florida Automated Weather Network (FAWN) located at Live Oak (FAWN, 201 1) were used to compare the 2010 and 2011 seasons in term s of rainfall distribution. It should be noted that in add ition to the non grafted scion plants (FL/FL) were added as a second set of controls These were included in the 100% irrigation and N rate (224 kg N ha 1 ) plots in order to examine the effect of graft injury and in itial growth reduction associated with the grafting process. Yield I rrigation W ater Use E fficiency and N itrogen U se E fficiency Mature green tomato fruit were harvested from 10 plants in each treatment combination per replication. Fruit were picked 80 and 88 days after transplanting ( DAT ) in 2010, and 75, 8 5 and 92 DAT in 2011. They w ere then graded as extra large, large, medium, and culls (small fruit and defective fruit). F ruit in each grade w ere counted and weighed. Total f ruit y ield, m arketable f ruit y ield, a verage f ruit w eight, and a verage n umber of f ruit per plant were estimated. Irrigation w ater use efficiency (iWUE) was estimated as the ratio of the m arketable fruit y ield to the amount of irrigation water applied during the production season N itro gen u se e fficiency (NUE) was estimated as the ratio of the m arketable fruit y ield to the amount of N supplied during the production season Statistical A nalyses Data from the 2010 and 2011 experiments w ere analyzed separately. Analysis of variance was cond ucted using the GLIMMIX procedure of SAS version 9.2 (SAS Institute, Cary, NC) Within each season, yield and yield components, i WUE, and NUE were analyzed with a model including main effects of irrigation regime, N fertilization
120 rate, and grafting treatme nt All models also included all possible interaction terms of these factors. An analysis of the conditional studentized residuals indicated no deviations from the normality and homoscedasticity assumptions and therefore data transformation was not needed 0.05) was used for multiple comparisons. Results Seasonal Rainfall at the Experimental Site Daily rainfall data collected by the weather station of FAWN located at the research center in Live Oak were used to compare the seasonal rainfa ll between the 2010 and 2011 trials (Florida Automated Weather Network, 2011) The rainfall p attern at the experimental site varied greatly between the 2010 and 2011 growing seasons (Fig ure 5 1). Total rainfall during the growing season was 272 mm in 2010 and 157 mm in 2011. Within the first 4 WAT, there was 12% higher rainfall in 2011 than in 2010. However, starting from the week 5 through week 11 an increased amount of rainfall by 326 % occurred in 2010 than in 2011, i.e., 186 mm in 2010 vs. 57 mm in 2011. Total and Marketable Fruit Y ield s Irrigation regim e N fertilization rate, and grafting all showed significant influence on tomato fruit yields in both 2010 and 2011 experiments (Table 5 1). In the 2010 trial, the 50% irrigation regime resulted in higher total and marketable fruit yields compared to the 1 00% irrigation regime. Increases averaged 15% and 19%, respectively (Table 5 2). Tomato yield responded to different N rates, with increased total and marketable fruit yields for increasing N rates up to 168 kg ha 1 Although N at 168 kg ha 1 increased yie lds above those at 56 kg N ha 1 no significant differences were observed with N rates higher than 168 kg ha 1 (Table 5 2). Grafting with the two rootstocks significantly
121 Averaged over the two rootstocks, the increase o f total and marketable fruit yields relative to those of non grafted reached 27% and 30%, respectively (Table 5 2). It was noted that the yield improvement in grafted plants was more pronounced starting in the second harvest especially with t he use of rootstock (data not shown). For the 2011 trial, the tomato yield response to N rates was dependent on irrigation regime as reflected by the significant interaction of irrigation regime by the N rate. With the 50% irrigation regime, t he total and marketable yields at 56 kg N ha 1 were significantly lower than those at higher N rates, but increasing N rate beyond 112 kg ha 1 did not enhance yields further (Table 5 3). In contrast, the 100% irrigation regime required higher N rates to ma ximize total fruit yield, which was similar above 224 kg N ha 1 Comparison of the yield response to the two irrigation regimes at each N rate did not show a clear pattern ; however, total yields did not differ except at 56 and 2 80 kg N ha 1 while marketabl e yields were similar except at 280 and 336 kg N ha 1 Similar to the 2010 experiment, total and marketable fruit yields of the grafted plants (FL/BE and FL/MU) were significantly higher than the non grafted plants (FL) in the 2011 trial, and a significant grafting N rate interaction was also observed for marketable fruit yield (Tables 5 1 and 5 rate at or above 112 kg N ha 1 and the yield increase reached 46% at 224 kg N ha 1 the c urrently recommended N rate for non responsiveness of marketable yield to N rates. For non yield was maximized at 112 kg N ha 1 and did not increase with more N In contrast, the
122 marketable fruit yields of the grafted plants continued to increase with N rates up to 280 kg ha 1 (Table 5 3). Yield C omponents The marketable yield components, including number of tomato fruit per plant and average fruit weight were further examined to identify the contributors to yield responses. In both 2010 and 2011 trials, grafted plants (FL/BE and FL/MU) had significantly more marketable fruit per plant than did non grafted plants (FL) (Table 5 1). Grafting increased fruit number by an average of 13% in 2010 and 27% in 2011 (data not shown). The number of marketable fruit per plant was also affected by a significant interaction between irrigation regime an d N rate in both years (Table 5 1). In 2010, under the 100% irrigation regime, the marketable fruit number per plant was significantly lower at 112 kg N ha 1 and most reduced at 56 kg N ha 1 (Table 5 4). Under the 50% irrigation regime, fruit number was lowest at 56 kg N ha 1 while increasing N rate beyond 112 kg ha 1 significantly increased the fruit number at 224 and 336 kg N ha 1 I n the 2011 trial t he marketable fruit number at different N rates under the two irrigation regimes followed the same response of marketable fruit yields (Tables 5 3 and 5 4). Consistent with the higher marketable y ield under the 50% irrigation regime in 2010, the average fruit weight was also significantly greater. S ignificant effects of grafting, N rate, and their interaction w ere also evident i n average fruit weight (Table 5 1 ). Grafting with the two rootstocks si gnificantly increased the average fruit weight of at most N rates tested but not at 112 kg ha 1 (data for 2010 not shown ). In the 2011 experiment, the impact of grafting on average fruit weight also reflected significant interaction effects as sociated with irriga tion regime and N rate (Table 5 1). Under the 50% irrigation regime, grafted plants (FL/BE and/or FL/MU) showed
123 significantly greater average fruit weight than the non grafted plant at each of the six N rates applied. In contrast, under the 100% irrigation regime, the average fruit weight did not differ between grafted and non 1 (Table 5 5). Under the recommended irrigation regime (100%) and N rate (224 kg ha 1 ), the average marketable fruit weight was increased by grafting by appropriately 17% in contrast to the non grafted plants. I rrigation W ater Use E fficienc y (iWUE) The iWUE relative to marketable fruit yields was affected by significant two way interactions between grafting, irrigation regime, and N rate. Results were similar in the 2010 and 2011 trials (Table 5 1). In both years, a marked decline of iWUE was observed when the irrigation regime was increased from 50% to 100% irrespective of grafting and N rate treatments (Figures 5 2A, 5 2B, 5 2C, a nd 5 2D) In 2010, the iWUE was maximized at 224 kg N ha 1 and 280 kg N ha 1 within the 50% and 100% irrigation treatments, respectively, while it did not differ significantly between N rates at and above 168 kg N ha 1 within each irrigation regime (Fig ure 5 2A). In 2011, under the 50% irrigation regime, the iWUE at 112 kg N ha 1 was significantly higher than that at 56 kg N ha 1 but the N rates above 112 kg ha 1 did not lead to any significant increase of iWUE. Within the 100% irrigation regime, the iWUE r eached the highest value at 280 kg N ha 1 but it did not differ significantly from values at 224 kg N ha 1 and 336 kg N ha 1 (Fig ure 5 2B). Furthermore, iWUE of the grafted plants were significantly higher relative to that of non grafted plants at both irr igation regimes in both years, although the difference tended to vary with the irrigation regime (Fig ure s 5 2C and 5 2D ). In 2010, the averaged iWUE values of the grafted plants were greater than that of the non grafted plants by 29% under the 50% irrigati on regime and by 32% under the 100% irrigation regime. In
124 2011, the average increase in iWUE as a result of grafting was 54% and 36% under 50% and 100% irrigation regimes, respectively. In general, both grafted and non grafted plants in each season exhibit ed a linear increase of iWUE with the increasing N rate from 56 to 168 kg ha 1 (Figure 5 2E and 5 2F). A further improvement of iWUE at 224 kg N ha 1 was observed in grafted plants but not in non N rate from 224 to 336 kg ha 1 did not result in any pronounced change of iWUE. Furthermore, the performance of the two rootstocks used also seemed to differ at certain N rates. In both years, grafting significantly enhanced the iWUE at different N rates except that there was no significant difference between the iWUE of FL/MU and non grafted (FL) at 56 kg N ha 1 while FL/BE and FL showed similar levels of iWUE at 168 kg N ha 1 in 2010. Nitrogen Use E fficiency (NUE) In the 2010 trial, the main effect of irrigation regime was sig nificant, while the significant effect of N rate on NUE was dependent on grafting (Table 5 1). NUE was 20% higher in the 50% irrigation regime compared to the 100% irrigation regime (data not shown). Moreover, compared with the non afted plants with the two rootstocks enhanced NUE significantly at each N rate except the 168 kg N ha 1 treatment (Table 5 6). On average, the increase in NUE due to grafting with the vigorous rootstocks were 81% and 23% with the 56 and 112 kg N ha 1 resp ectively, while ranging from 26% to 38% when increasing the N rate from 224 to 336 kg N ha 1 In addition, for both FL/BE and FL/MU, increasing the N rate from 56 to 336 kg N ha 1 consistently decreased the NUE. However, the decrease was mainly significant between the two lower N rates (56 and 112 kg ha 1 ) and the two higher N rates (280 and 336 kg ha 1 ). With the non grafted plants, except the N rate at 56 kg ha 1 NUE
125 decreased as the N rate increased from 112 to 336 kg N ha 1 Furthermore, in 2011, the m ain effect of grafting was significant, while the significant effect of N rate on NUE was related to the irrigation regime (Table 5 1). In 2011, compared with the non grafted e to that of non grafted plants (data not shown). Under both irrigation regimes regardless of the grafting treatment, the rate of 112 kg N ha 1 resulted in the highest NUE, while the lowest NUE was observed at 336 kg N ha 1 (Table 5 6). NUE values at certa in N rates were also influenced by the irrigation treatments. At 56 kg N ha 1 the NUE was significantly higher under the 50% irrigation regime while the opposite was observed at each of the higher N rates 1 Discussion Grafting Influence on Fruit Yields of Tomato under Field C onditions s at almost all N rates applied in this study. Moreover, grafted plants tended to be more responsive to the increase of N than non grafted plants particularly in the 2011 trial. Compared with non grafted tomato plants, it is likely that grafted plants may require a higher level of N for maximizing yield performance Fruit is an important sink for carbohydrates and amino acids in tomato plants (Valle et al., 1998) Given the greater number of fruit per plant and higher average fruit weight of graft ed plants, the sink strength of the grafted plants would have been at a higher level relative to that of the non grafted plants. This stronger fruit sink of grafted tomato may partly reflect a greater demand for nutrients especially nitrogen. F uture work i s warranted to elucidate the nutrient requirement for field production of grafted tomatoes with different combinations of scions and
126 rootstocks. In the present study, the two interspecific tomato hybrid rootstocks demonstrated an overall similar yield impr ovement. Although the rootstocks used are well known for their high resistance to several soilborne diseases, the increased yields observed here with fumigated soils could be attributed primarily to the vigorous characteristics of the rootstocks. The posit ive scion rootstock interactions would have improved plant growth and development. Yield enhancement from grafting, even in a low disease pressure, has been reported previously with solanaceous and cucurbitaceous vegetables (Fernandez Garcia et al., 2004a; Leonardi and Giuffrida, 2006; Proietti et al., 2008; Ruiz et al., 1997) In this study, fruit yields were similar for self grafted ( FL/FL ) and non grafted ( FL ) plants under the recommended irrigation regime and N r ate (data not shown), thus reinforcing the suggestion that yield increase could be attributed mainly to the specific rootstocks used rather than the grafting process per se. Lykas et al. (2008) reported an increase of tomato yield due to self grafting in a greenhouse hydroponic study However it was unclear how consistent the self grafting effect was as their study was not repeated in a second season Nevertheless, it has been suggested that the growth and yield enhancement in grafted vegetable plants is l argely contributed by the use of selected rootstocks rather than the grafting process per se (Davis et al., 2008; Lee and Oda, 2003; Lee et al., 2010) In general, data here showed that increases in both fruit numbe r per plant and average fruit weight contributed to the overall improvement of marketable fruit yield from grafted plants. These results concur with some previous reports (Lee and Oda, 2003; Passam et al., 2005) but Di Gioia et al. (2010) did not observe a significant change in yield components as a res ult of grafting. In our study, the cull fruit yield (small,
127 immature, and/or damaged fruit) did not differ significantly between grafted and non grafted tomato plants (data not shown). However, the percentage of cull fruit out of the total harvest was also reduced in grafted treatments by 28% on average relative to the non grafted plants, and thereby contributing partly to marketable fruit yield increase as well. Underlying mechanisms inherent to growth vigor and yield enhancement by grafted plants are ofte n attributed to enhanced nutrient and water uptake (Rouphael et al., 2008; Ruiz et al., 1997) and increased synthesis of endogenous hormones (Dong et al., 2008; Seong et al., 2003) Seasonal Variation of the Irrigation and N Rate E ffects Und er the two irrigation regimes tested in this study, i ncreasing N rate significantly increased fruit yields of both grafted and non grafted tomato plants to a maximum amount, which then leveled off with higher levels of N application. These results were con sistent with previous studies by Scholberg et al. (2000) Topcu et al. (2007) and Hebbar et al (2004) using tomatoes grown under diverse environmental conditions. The influence of irrigation regime and N rate on total and marketable yields varied between 2010 and 2011 seasons. In 2010, the 50% irrigation regime led to significantly higher fruit yields as compared to the 100% irrigation regime. In contrast, in 2011, yields under the 100% irrigation tended to be higher than those with the 50% irrigation at N rates above 224 kg ha 1 The inconsistent results over these two seasons could be largely related to the v ariation of environmental conditions especially the rainfall. The 2010 growing season had a higher lev el of rainfall than 2011 particularly during the plant reproductive stage. This period corresponded to the prime period of fruit development during which the nutrients applied through fertigation accounted for about 60% of the total nutrient supply after t ransplanting.
128 water holding capacity (0.0 7 0. 25 cm of water per cm of soil) and are more prone to nitrate (NO 3 ) leaching (Simonne et al., 2004, 2006) With the hea vier rainfall in 2010, NO 3 leaching problem in sandy soils at the experimental site could have been worse and therefore reduced N uptake by plants. A study by Zotarelli et al. (2009a) in Citra, FL reported similar inter annual differences in tomato yields when rainfall was greater along with higher temperatures in the middle of the growing season. Improving irrigation scheduling based on soil moisture sensors and crop needs will help reduce nitrate leaching in field tomato production in sandy soils (Simo nne et al., 2010; Zotarelli et al., 2009b) In this study there was no or reduced yield response to N rates higher than 168 kg ha 1 This is in line with results reported by Zotarelli et al. (2009a) who in a 3 yea r study did not find any significant difference in fruit yields of the same tomato cultivar Florida 47 in north Florida in response to three N rates ( 176, 220, and 330 kg ha 1 ). With the same tomato cultivar (non grafted) and experimental site as our stud y, Poh et al. (2011) observe d difference in yields as a result of inter annual environmental disparities. In one season, these authors obtained fruit yields of 44, 43, and 49 Mg ha 1 at 134, 179, and 224 kg N ha 1 respectively, with a significant difference between the highest N rat e and the other two lower N rates whereas such a trend was not found in the other season In another field study in central Florida on the specialty tomato cultivar Tasti Lee a significantly higher total marketable yield was found at 307 kg N ha 1 as com pared wtih those at 229 and 268 kg N ha 1 over two seasons (Santos et al., 2010) Grafting Influence on Irrigation Water and Nitrogen Use E fficienc y were more efficien t in water use for fruit production compared to the non grafted plants. These results concur with previous reports on rootstock effects on the iWUE relative to
129 marketable fruit yield. According to Lykas et al. (2008) mans improved WUE (measured as amount of harvested fruit per volume of transpired water) of greenhouse also exhibited a similar increase. Improvement of water use efficiency was a lso observed by Rouphael et al. (2008) on mini watermelon under field conditions. The enhanced WUE found in grafted plants in these studies was mainly due t o the improvement of fruit yields rather than reduced water use. In both 2010 and 2011 trials, the 100% irrigation regime resulted in a decrease in iWUE as opposed to the 50% irrigation regime despite the use of grafting with rootstocks. This indicates tha t the full irrigation regime probably caused an excess in water supply which did not favor fruit yield development. Similar findings were reported previously by Rouphael et al. (2008) on mini watermelon, Cabello et al. (2009) on melon and Kirnak et al. (2002) on eggplant. The enhancement in WUE in reduced irrigation regime s is often attributed to various adaptive physiological mechanisms which involve changes in water relations and gas exchange (Bloch et al., 2006; Rajabi et al., 2009) It is suggested that under drought stress water loss associated with plant carbon fixation can be reduced (Bloch et al., 2006) Increase in WUE of plants under reduced water supply may also relate to certain morphological changes in plants such as deeper root systems and canopy modification (Zhang et al., 1998) In addition to improved iWUE grafting with appropriate rootstocks may a lso induce greater diversity in terms of NUE as compared to non grafted plants. Our data show ed that when averaged over irrigation regimes and N rates, grafted plants were more nitrogen use efficient in marketable fruit production by about 3 5 and 4 2 % compa red to
130 non grafted plants in 2010 and 2011, respectively. An increase of NUE relative to yield by 11.8% in grafted melon was reported by Colla et al. (2010) in comparison with non grafted plants. In another study on mini watermelon, Colla et al. (2011) observed a 38% increase in NUE of grafted plants as compared with non grafted plants. Similar to the improvement of iWUE, higher NUE observed with grafted plants in this study was associated with greater yields in grafted plants than non grafted p lants. Although grafting, irrigation regime, and N rate showed relatively consistent effects on iWUE between 2010 and 2011, their influence on NUE varied with the production season. The significant interaction effect of irrigation regime and N application on NUE in 2011 suggested differential responses of plants to N fertilizer rates under different irrigation regimes. Except for the lowest N rate, increasing N rates under both irrigation regimes resulted in reduced efficiency of N use for yield production. Only one tomato scion cultivar was used in this study. Future research will need to involve more scions and rootstocks to identify the scion rootstock interactions and specific traits of rootstocks in relation to the improvement of iWUE and NUE in grafted tomato plants. Conclusions This study demonstrated that u se of grafted tomato plants could significantly improve fruit yields from field production systems used on sandy soils in Florida. The increase of marketable yields resulted from both more fruit per plant and higher average fruit weight Grafting with the two interspecific tomato hybrid rootstocks used here also led to significant enhancement in efficiency of water and nitrogen use. Overall, the two rootstocks performed similarly. Moreover, variation in environmental conditions particularly rainfall contributed greatly to yield response s to irrigation and N application. Future studies are warranted to explore the N requirement for grafted tomato production
131 under field conditions since data here ind icate that these needs are likely to differ from th ose of non grafted tomato es.
132 Table 5 1 Analysis of variance for effects of irrigation regime, nitrogen fertilization rate, and grafting combination on total tomato fruit yield, marketable fruit yield number of marketable fruit, average weight of marketable fruit, irrigation water use efficiency, and nitrogen use efficiency. Effect DF Total fruit yield Marketable fruit yield Marketable fruit Irrigation water use efficiency Nitrogen use efficiency Number of fruit per plant Average fruit weight 2010 2011 2010 2011 2010 2011 2010 2011 2010 2011 2010 2011 Irrigation (I) 1 *** ** *** ** *** NS ** ** ** *** *** NS N rate (N) 5 *** *** *** *** *** *** *** ** *** *** *** *** Graft (G) 2 *** *** *** *** *** *** *** *** *** *** *** *** IN 5 NS *** NS *** ** NS NS *** ** NS IG 2 NS NS NS NS NS NS NS ** *** NS NS NG 10 NS NS NS NS NS ** NS ING 10 NS NS NS NS NS NS NS ** NS NS NS NS NS,*,**,*** Non significant or significant at the P
133 Table 5 2 Main effects of irrigation, nitrogen fertiliz ation rate and grafting combination on total yield, marketable yield, and nitrogen use efficiency of tomato in the 2010 trial in Live Oak, FL. Treatment Yie ld (Mg ha 1 ) Nitrogen use efficiency (kg kg 1 ) Total Marketable Irrigation (%) 50 55.5 a 49.8 a 319.1 a 100 48.3 b 41.8 b 262.9 b N rate (kg ha 1 ) 56 23.4 c 18.4 c 327.9 b 75 42.8 b 36.3 b 483.9 a 148 57.0 a 49.9 a 337.3 b 224 63.8 a 57.4 a 256.2 c 298 61.9 a 55.4 a 185.8 d 372 62.5 a 57.5 a 154.6 d Graft FL/BE 54.7 a 48.9 a 314.5 a FL/MU 56.9 a 50.5 a 320.5 a FL 44.1 b 38.1 b 237.9 b Means within a column followed by the same letter do not differ at the P according to Florida 47 grafted onto Beaufort Florida 47 grafted onto Multifort and FL: non
134 Table 5 3 Total and marketable yield s of toma to as influenced by interactions between irrigation regime N rate, and grafting treatments in the 2011 trial in Live Oak, FL. N rate (kg ha 1 ) Total fruit yield (Mg ha 1 ) Marketable fruit yield (Mg ha 1 ) Marketable fruit yield (Mg ha 1 ) 50% 100% 50% 100% FL/BE FL/MU FL (Irrigation regime) (Irrigation regime) 56 25.4 Ba 18.6 Db 21.7 Ba 15.8 Da 21.7 Ca 20.2 Cab 14.4 Bb 75 54.4 Aa 51.5 Ca 49.4 Aa 45.9 Ca 51.8 Ba 53.2 Ba 38.0 Ab 148 59.3 Aa 57.3 Ca 53.2 Aa 52.9 BCa 58.3 ABa 57.3 ABa 43. 7 Ab 224 55.9 Aa 63.2 Ba 50.8 Aa 58.7 ABa 61.3 ABa 61.0 ABa 42.0 Ab 298 55.3 Ab 72.8 Aa 48.8 Ab 66.8 Aa 65.1 Aa 64.8 Aa 43.3 Ab 372 54.3 Aa 65.1 ABa 47.9 Ab 60.8 ABa 62.6 Aa 60.9 ABa 39.6 Ab Means followed by the same uppercase letter within a co lumn, and means followed by the same lowercase letters within a row are not significantly different according to at P 0.05 Florida 47 grafted onto Beaufort Florida 47 grafted onto Multifort
135 Table 5 4 Number of marketable tomao fruit per plant as influenced by the interaction between irrigation regime and N ra te in the 2010 and 2011 trials in Live Oak, FL. N rate (kg ha 1 ) 2010 Irrigation regime (%) 2011 Irrigation regime (%) 50 100 50 100 56 8 Ca 7 Ba 9 Ba 7 Da 75 16 Ba 12 Bb 19 Aa 17 Ca 148 19 ABa 18 Aa 20 Aa 19 BCa 224 22 Aa 18 Aa 20 Aa 21 ABa 298 19 ABa 20 Aa 19 Ab 23 Aa 372 22 Aa 18 Aa 18 Ab 22 ABa Means followed by the same uppercase letter within a column, and means followed by the same lowercase letters within a row are not significantly different according to at P 0.05
136 Table 5 5 Average marketable weight of tomato fruit from grafted and non grafted plants as influenced by interaction between irrigation regime and N rate in the 2011 trial in Live Oak, FL. Average fruit weigh t (g/fruit) N rate (kg ha 1 ) Graft 50% Irrigation regime 100% Irrigation regime 56 FL/BE 165.2 ab 161.7 a FL/MU 177.1 a 159.2 a FL 154.0 b 170.9 a 75 FL/BE 182.7 a 190.6 a FL/MU 185.6 a 203.1 a FL 165.6 b 174.3 b 148 FL/BE 188.7 ab 192.7 ab FL/MU 189.6 a 210.6 a FL 1 74.7 b 175.2 b 224 FL/BE 192.8 a 201.2 a FL/MU 178.3 ab 200.1 a FL 166.6 b 171.5 b 298 FL/BE 193.6 a 190.9 b FL/MU 182.0 a 211.5 a FL 157.0 b 195.0 b 372 FL/BE 190.1 a 202.5 a FL/MU 190.3 a 194.3 ab FL 156.9 b 180.2 b Means within a N rate and the three grafting in the columns h aving same letters do not differ at the P t. Florida 47 grafted onto Beaufort Florida 47 grafted onto Multifort and FL: non
137 Table 5 6 Nitrogen use efficiency (kg kg 1 ) of grafted tomato plants as influenced by interaction between irrigation treatment and nitrogen rate (left) and interaction between nitrogen rates and grafting combination (right) in Live Oak, FL during the spring of 2011. N rate (kg ha 1 ) Irrigation regime (%) Graft 50 100 FL /BE FL/ MU FL 56 388.54 B a 282.3 B C c b 387.34 BCa 360.37 Bab 258.59 BCb 75 658.86 A a 612.2 Aa 690.43 Aa 709.46 Aa 506.69 Ab 148 359.89 BC a 358.1 Ba 394.11 Ba 387.5 Ba 295.37 Bb 224 226.96 CD b 262.32 Ca 273.95 CDa 272.39 BCa 187.58 CDb 298 163.71 D b 224.29 CDa 218.77 Da 217.74 Ca 145.5 DEb 372 129.03 D b 163.5 Da 168.36 Da 163.76 Ca 106.66 Eb Means followed by the same uppercase letter within a column, and means followed by the same lowercase letters within a row are not significantly different according to at P 0.05 Florida 47 grafted onto Beaufort Florida 47 grafted onto Multifort and FL: non
138 Fig ure 5 1 Cumulative rainfall in Live Oak, FL during the tomato field trials in 2010 and 2011.
139 Fig ure 5 2 Irrigation water use efficiency as influenced by interaction between irrigation treatments and nitrogen fertilization rates (A), by interaction between irrigation and grafting treatments (B), and by interaction between nitroge n fertilization rates and grafting treatments (C) in 2010. Treatment values at each irrigation regime or each nitrogen rate followed by the same letter are not significantly different at P Florida 47 grafted onto Beaufort Florida 47 grafted onto Multifort and FL: non
141 CHAPTER 6 BIOMASS, NITROGEN AC CUMULATION, AND ROOT DISTRIBUTION OF GRAF TED TOMATO AS AFFE CTED BY NITROGEN FER TILIZATION Introduction As one of the essential plant nutrients, nitrogen (N) has the greatest influence on the growth and development of most crops, being a constituent of proteins, amino acids, chlorophyll, nucleic acids, and cell wal ls (Fageria, 2009; Neeteson, 1995) N fertilization is critical for shoot and root growth and it is often applied in greatest quantity and most frequently, followed by K fertilization, in many plant production systems, inclu ding the high value crops like tomato ( Solanum lycopersicum L. ) (Hartz and Bottoms, 2009; Hochmuth and Hanlon, 2011; Tei et al., 2002) However, due to the high mobility of N in the soil profile and the soil atmosph ere interface through leaching, volatilization, and denitrification, N fertilization management practices based on high levels of N application often resulted in various environmental concerns (Hallberg, 1989; Tan et al., 2002; Vzquez et al., 2006; Zotarelli et al., 2009a) Best management practices (BMPs) including cultural practices have been developed to improve fertilizer use efficiency by plants and minimize the adverse impact of nutrient loss from the producti on site on environmental quality. For example, on sandy soils with low intrinsic water and nutrient retention capacities, a variety of site specific practices are being used to optimize nutrient management including fertigation with drip irrigation, contro lled released fertilizers, soil moisture sensing which allow reducing, splitting or adjusting irrigation rates and/or frequency crop rotation, and cover cropping (Simonne et al., 2010b) In addition to improving management practices, it is suggested that the selection and use of crop varieties with inherently high N use efficiency (NUE) could also help, to some extent, alleviate the environmental concern associated with N loss in the production
142 system while ensuring stability of high yields (Lynch, 1998) Crop varieties with vigorous physical root traits that may allow increased nutrient uptake would increase yields especially in low fertility soils (Rengel and Marschner, 2005) Nutrient efficient varieties in crops such as wheat, maize, rice have been reported (Lynch, 1998; Rengel and Damon, 2008; Wiesler et al., 2001) Research is also taking place to identify the varietal differences i n N use efficiency in vegetable crops (Benincasa et al., 2011) Alternatively, vegetable grafting with vigorous rootstock is also suggested as a viable option for improving crop nutrient use efficiency. Alt hough primarily us ed as an effective tool to manage various soil borne diseases in many intensive and continuous vegetable cropping systems (Lee, 1994; Lee et al., 2 010) grafting with vigorous rootstocks has also been shown to enhance nutrient uptake (Leonardi and Giuffrida, 2006; Ruiz et al., 1997) improve water use efficiency (Rouphael et al., 2008b) and consequently result in higher yield. Specifically, it was demonstrated that grafted tomato plan ts could exhibit an increase of marketable yield by 20 62% as compared with non grafted plants, depending upon the production conditions and scion rootstock interactions (Di Gioia et al., 2010; Lee and Oda, 2003; Leo nardi and Giuffrida, 2006; Pogonyi et al., 2005) Similar grafting effects on yield improvement were observed on different cucurbit crops such as melon and watermelon (Ibrahim et al., 2001; Passam et al., 2005; Rahm an et al., 2002) For instance, Proietti et al. (2008) reported that total and marketable yield s of mini watermelon were higher by 46% and 64% as a result of the use of certain rootstocks. In comparison with non Minirossa watermelon, graftin g Minirossa onto Vita increased N use efficiency (yield/applied N rate), N uptake efficiency (plant N content/applied N rate), and physiological N utilization
143 efficiency (yield/plant N content) by 38%, 21%, and 17%, respectively (Colla et al., 2011) This improved efficiency of nutrient uptake and use in graf ted plants may be related to the modification of root architecture and distribution due to the use of certain rootstocks. Some of the r oot characteristics that may play an active role in nutrient and water uptake include root length and density, number of root hairs and their length and surface area, and intrinsic uptake ability (Martnez Ballesta et al., 2010) Oztekin et al. (2009) evaluated the root characteristics of grafted and non grafted tomato plants grown in solution culture and observed that root traits such as root density and number of root hair were improved with grafted plants in comparison to the self grafted plants. However, their findings did not provide practical implications for field production systems. To date, few studies have been conducted to understand root distribution characteristics of grafted plants as compared with non grafted plants in field production in relation to N uptake. The objective of the present study was to compare the growth, plant nitrogen concentration and accumulation, N uptake efficiency, and root distribution between grafted and non grafted tomato plants in response to different N fertilization rates in sandy soils. We hypothesized that grafting with vigorous rootstocks could increase root length density distribution and plant biomass and nitrogen accumulation as compared with non grafted plants. Materials and Methods Experimental S ite and D esign Fie ld experiment s w ere conducted during the spring seasons of 2010 and 2011 at the Suwannee Valley Agricultural Extension Center in Live Oak, FL. The soil type is a Blanton Foxwort Alpin Complex sandy soil (Natural Resources Conservation Service, 2006) Detailed information about the grafted transplant production, field preparation,
144 and management practices, etc. was provided in Chapter 5. Briefly, a split plot design with four replications (blocks) was used in both years. The whole plot treatments consisted of six N fertilization rate s arranged in a randomized complete block design The six N fertilization rates were 56, 112 1 6 8, 224, 28 0 and 3 36 kg N ha 1 which represented 25%, 50 %, 75 %, 100%, 1 25 %, and 1 50 % respectively of the current University of Florida Institute of Food and Agr icultural Sciences ( UF/IFAS ) recommended total N application rate at 2 24 kg N ha 1 (a preplant application at 56 kg N ha 1 included) f or field production of irrigated round tomato in sandy soils in Florida (Olson et al., 2010) Irrigation was maintained at the current UF/IFAS recommendation for irri gation for field production of round tomato in sandy soils in Florida i.e., 9354 L / ha /day /string. The Florida 47 grafted onto Beaufort Florida 47 grafted onto Multifort ( FL/ M U ) and th e non There were 12 plants for each treatment combination per replication in both 2010 and 20 11. Beds were 0. 71 m wide and spaced 1.52 m apart (from middle to middle ) with 0.46 m in row spacing for open field tomato production Except for the 56 kg N ha 1 treatment which only included a preplant application of 13N 1.7P 10.8K a mmonium nitrate ( 34 N 0 P 0 K Mayo Fertilizer Inc, Mayo, FL) was injected weekly through the drip tape start ing one week after transplanting to provide remaining amount of N of each fertilization rate, following the recommended fertilization scheduling (Olson et al., 2010) P otassium chloride (Dyna Flo 0 0 15, Chemical Dynamics Inc, Plant City, FL) was also applied through fertigation to provide an equival ent supply of potassium in each N rate treatment based on the soil test Other cultural practices including pruning and pest
145 control followed the current recommendations for commercial field tomato production in Florida (Olson et al., 2010) Plant G rowth, N itrogen Uptake and E fficiency Maximum accumu lated biomass was destructively evaluated on one representative plant per treatment combination in each replication at 85 days after transplanting ( DAT ) in 2010 and at 82 DAT in 2011. Each sampled plant was cut at ground base and separated into leaf blade, petiole, stem, and fruit, and fresh weight of each plant part was taken. Representative samples ( approximately 300 g each) from leaf blade, petiole, and stem respectively, and a sample of fruit (approximately 500 g) were taken from each sampled plant and weighed. Leaf area was measured with a LI COR 3100 leaf area meter (LI COR Inc., Lincoln, NE) and l eaf area index (LAI, m 2 leaf m 2 land) was calculated A ll the sub samples from each plant were dried in a forced air drying oven at 60C for 72 to 120 h un til constant weight. T otal aboveground biomass was then determined. Dried sub samples of leaf blades, petiole, stem, and fruit were analyzed for t otal K jeldahl n itrogen (TKN) concentrations by combustion technique Shoot N accumulation was determined by mu ltiplying dry mass of leaf blade, petiole, stem, and fruit by the corresponding N concentration s. In addition, N uptake efficiency expressed as total N accumulated/N supply was also estimated. Root A nalysis At 93 DAT in 2010 and at 97 DAT in 2011, root sam ples were collected from the grafted and non grafted plants to examine the nitrogen fertilization effect on root growth and distribution. The root analysis study was focused on three N application rate treatments including 112 224, 3 36 kg N ha 1 followin g the root sampling method
146 previously described in Zotarelli et al. (2009b) Briefly, root s were sampled by taking soil cores at four different depths: 0 15, 15 30, 30 60, and 60 90 cm using a 5 c m diameter soil au ger and at two different positions around the plant ( i.e., at the plant base vs. at 15 cm distance from the plant) in the center of each treatment plot. Soil samples collected at each depth per treatment were stored at 4 o C prior to processing in the lab. During the sample processing, each sample was weighed and washed with running water using a fine sieve to collect root and other debris. Then, cleaned materials above the sieve were placed in a clear glass pan and tomato roots were carefully handpicked wit h tweezers and placed in petri dishes. Washed and cleaned roots per soil core were scanned using a root scanning apparatus (EPSON color image scanner LA1600 + Toronto, Canada) and analyzed with the image analysis software WinRhizo 2008a (Regent Instruments Quebec, Canada) to determine the total root length. R oot length density (RLD cm cm 3 ) for each soil depth was then estimated as the total root length (cm) per volume of soil (cm 3 ) of each soil depth. Also, root surface area density (RSAD, cm 2 cm 3 ) was also calculated as the total root surface (cm 2 ) per volume of soil (cm 3 ). The RLD and the RSAD described the exploratory capacity of the root system in terms of length and surface area by unit of soil volume. Statistical A nalyses Statistical analyses wer e performed with SAS proc GLIMMIX procedure ( SAS 9.3, SAS Institute, Cary, NC). Within each season, the root distribution data were analyzed with a heteroscedastic linear mixed effects model including the main effects two way interactions, three way inter actions, and four way interactions of the four factors of N rates, grafting, position, and depth. N rate within block and blocks were added to the models as random effects. However, models with block as a random effect led to
147 smaller values of ormation Criterion and were then considered for the analysis. Covariance structure with the residual option based on the effect of Student panel plots were used to check normality and homogeneity of residu als, and in th is study, data transformations were not used. Further, whenever the F tests for fixed effects were found significant, post hoc analysis were performed to identify the least square means that were significant ly different (P < 0.05) fixed effec ts were observed pairwise comparisons were made using the LSMEANS statement adjusted with TUKEY method. Results Leaf A rea I ndex and Plant B iomass The significant main effects of N fertilization and grafting treatments on plant growth parameters were obser ved in both the 201 0 and 2011 field trials (Table 6 1) In both seasons, i ncreasing N fertilization from 56 to 224 kg ha 1 resulted in a significant increase in leaf area index (LAI) and aboveground biomass. However, any further increase in N rate did not result in a significant increase in the LAI and aboveground biomass accumulation (Table 6 2). Grafted plants with the two rootstocks showed significantly higher levels of LAI and aboveground biomass as compared with the non grafted plants, while FL/BE and FL/MU performed similarly. Averaged over the N rates, grafted plants demonstrated an increase of LAI by about 33% and 35% over the non grafted control in 2010 and 2011, respectively. The pronounced increases in LAI as a result of grafting was accompanied b y improved dry matter accumulation, i.e., total aboveground biomass (leaf blades, petiole, stem, and fruit) of grafted plants was higher than non grafted plants by approximately 16% and 27% in 2010 and 2011, respectively.
148 N A ccumulation and N U ptake E ffic iency The N application rate and grafting showed significant influence on the accumulated plant N and the nitrogen uptake efficiency (NUE) in both field experiments (Table 6 1) In 2010, i ncreasing N rates from 56 to 3 36 kg ha 1 significantly improved N ac cumulation in the aboveground tissues from 20.97 to 77.52 kg ha 1 In the 2011 season, the change in accumulated N ranged from 13.35 to 69.79 kg ha 1 (Table 6 2). In both seasons, the N accumulation was also significantly higher in grafted plants compared with the non grafted control The two grafting combinations did not differ significantly in 2010, while FL/MU showed a significantly greater level of accumulated N than FL/BE in 2011 Averaged over the N fertilization rates, the accumulated N in the graft ed plants was increased by about 20 and 33% relative to the non grafted plant in 2010 and 2011, respectively. Analysis of the plant tissue nitrogen concentrations revealed greater effect of N application rate than th at of grafting. N rate significant ly af fected N concentrations in the leaf blade and stem in both years. Fruit N concentration in 2010 was also significantly impacted by N fertilization. In contrast, grafting only showed a significant impact on the N concentration in the leaf petiole (Table 6 3 ). The nitrogen uptake efficiency (NUE) was affected significantly by both the N application rate and grafting in the 2 year study. In 2010, the NUE at 5 6 kg N ha 1 was significantly higher than that at any other N rates, but increas ing N rates from 1 12 to 336 kg N ha 1 did not lead to any significant difference in the NUE. In 2011, the NUE at 112 kg N ha 1 was significantly higher than that at other N rates except 168 kg N ha 1 NUE consistently decreased for increase N rates from 112 to 336 kg N ha 1 Th i s reduction of NUE was about 43 % when the N rate was increased from 112 to 3 36 kg N
149 ha 1 (Table 6 3). With respect to the grafting effect on the percentage of nitrogen recovered in the aboveground plant tissues results slightly var ied with the growing sea son. In 2010, both FL/BE and FL/MU significantly improved NUE relative to non grafted plants by 2 2 % on average. However, in 2011 the significant increase in NUE (about 3 9 %) as a result of grafting was only observed in FL/MU, while FL/BE and FL showed simil ar levels of NUE (Table 6 2). Root Length Density D istribution During the two seasons, the root length density (RLD) varied significan tly with the N rates, grafting type soil depth, and sampling position. Specif i cally, N rate s effect was dependent on gra fting type in 2010 and on the soil depth in 2011 (Table 6 3) Except at the 112 kg N ha 1 g rafting with the two rootstocks significantly increased the RDL of at each of the two higher N rate s (224 and 3 36 kg N ha 1 ) by 58 and 118% (Table 6 4 ) Furthermore, irrespective of the N rates, RLD significantly decreased along the soil depth The majority of the root system was present in the upper soil depth s ; p articularly with up to 65% of root system being concentrated in the 0 15 cm soil depth in c ontrast to only about 7% found in the deepest depth (60 90 cm) Similarly, t he re was a significant interaction effect of grafting by soil depth on the RLD in both seasons In general, as with the N rates, RLD consistently decreas ed as the soil depth increa sed to 90 cm independent of the grafting effect. Across all the grafted and non grafted tomato plants, over half of the root system was concentrated between 0 and 15 cm of the soil profile, which accounted for 57 63% and 56 65% of the total root length den sity in 2010 and 2011, respectively (Table 6 5). At the 0 15 cm soil depth, RLD of the grafted plants was significantly greater than that of the non grafted plants by about 78% and 69% in 2010 and 2011, respectively. However, values of the RLD measured wit hin 15 30 cm
150 and 30 60 cm of the soil profile were similar between grafted and non grafted treatments except that FL/MU demonstrated a significantly higher level of RLD than FL at 30 60 cm in 2011 At the 60 90 cm soil depth, RLD was significantly greater in grafted plants than non grafted plants in 2010, whereas it did not differ between grafted and non grafted treatments in 2011 (Table 6 5). Soil sampling position demonstrated a significant impact on RLD which was also determined by the soil depth. RLD m easured at the plant base position (P1) were significantly higher than that at the position of 15 cm distance from the plant (P2), with the exception of 60 90 cm of the soil profile where RLD was similar between P1 and P2 in the 2011 trial (Table 6 5). On average, the root length density at P1 was greater than that at P2 by approximately 66%, while the reduction of RLD from P1 to P2 appeared to be more pronounced at the top soil profile (0 15 cm). In addition, the decrease of RLD along the 0 90 cm of the so il profile seemed to be more rapid at P1 than at P2. The root length density decreased as the soil depth increased. Discussion Grafting I nfluence on P lant Growth and Nitrogen Use E fficiency Positive effects of N fertilization rates on tomato growth paramet ers have been found in previous studies (Elia and Conversa, 2012; Scholberg et al., 2000) As Below (2002) reported, nitrogen nutrition influences leaf growth and leaf area duration as well as the photosynthetic rate per unit of leaf area and the size of the photosynthetic system, which generate assimilates for plant growth. The increased LAI and the resulting biomass at increased N rates as observed in this study can be attributed to accelerated aboveground growth as a result of increased carbon assimila tion. The genetic potential of both grafted and non grafted tomato plants for absorbing and
151 utilizing N was possibly reached at 224 kg N ha 1 give n that any further increase in N fertilization rates above 224 kg ha 1 did not result in any significant impro vement in LAI and aboveground biomass in this study. T he crop response to N fertilization rates seemed to be less pronounced in 2011 than in 2010. This was likely due to the difference in rainfall distribution between the two seasons. In general, the 2011 spring season was drier than 2010, which could have resulted in reduced N leaching out of the root zone in 2011 and thus greater uptake of N by root. Marked differences in growth and yield potential often exist among crop species and cultivars when grown u nder the same soil and climatic conditions (Lafever, 1981) I onto two interspecific hybrid rootstocks were significantly higher than the non grafted tomato plants. Such differences reflected the potential of vigorous rootstocks in enhancing growth and yield of grafted plants. These results are consistent with previous reports on grafted vegetable production (Di Gioia et al., 2010; Fernandez Garcia et al., 2004b; Leonardi and Giuffrida, 2006) In this study, enhanced growth of grafted plants also resulted in higher accumulated N as the two grafted treatments exhibited significantly higher acc umulated N in aboveground plant tissues as compared to the non grafted treatment. However, the enhancement in N accumulation of grafted plants relative to the non grafted plants was more related to greater accumulated biomass rather than an increase in N c oncentration of the plant tissues per se In contrast, Leonardi and Giuffrida (2006) reported significant differences in N concentration grafted control plants. A recent study showed no difference in N concentrations of tomato plant tissues in
152 response to in creasing N fertilization rates (Elia a nd Conversa, 2012) However, different from the N concentration responses to grafting, N concentrations in plant tissues demonstrated a greater responsiveness to N application rates in our study particularly the N concentrations in the leaf blade and stem Nitrogen uptake efficiency is one of the physiological parameters that help to evaluate the effectiveness of fertilizer N recovery due to N uptake by the plant. Analysis of the nitrogen uptake efficiency defined in this study as a percentage of N recove red in the aboveground biomass to the applied N showed a decreasing trend of this parameter in response to increasing N fertilization rate. This is in accordance with previous studies conducted on tomato (Elia and Co nversa, 2012; Scholberg et al., 2000; Singandhupe et al., 2003; Zotarelli et al., 2009a) and other crops such as zucchini (Zotarelli et al., 2008) and melon (Cabello et al., 2009) This opposite trend of nitrogen uptake efficiency to increasing N rates is greatly due to the incremental increase in nitrogen application not producing ever in creasing harvestable yields (Cabello et al., 2009) and also to the fact that under limited supply of available nitrogen in growing media, plants absorb more mineralized soil nitrogen to meet their demand (Jamil Mohammad, 2004) For instance, Sweeney et al ., (1987) reported that under Florida sand y soil conditions 2 to 4% of the organic soil N could become available for plant uptake during the growing season and as a result, N supply from mineralization appeared to be about 10 to 40 kg N ha 1 during the growing season. The grafting effect on nitro gen uptake efficiency demonstrated in the present study was independent of the N application rate. Overall, grafted plants showed significantly greater nitrogen uptake efficiency as compared to non grafted plants. In the 2011 season, NUE values of plants g
153 nitrogen uptake efficiency as a result of grafting with c ertain rootstocks has been found in previous studies on grafted melon (Colla et al., 2010) on mini watermelon (Colla et al., 2011) While more in depth research is needed to fully understand the contributing factors, these authors s uggested that increase of nitrate reductase activity has been shown to be partly responsible for NUE improvement in some grafted cucurbit crops. This could also be due to the increased growth and sink demand. Grafting Influence on the Root Length Density D istribution The more extensive and vigorous root system of grafted vegetables is often suggested to contribute considerably to the observable crop vigor and yield enhancement as a result of grafting with vigorous rootstocks (Davis et al., 2008b; Martnez Ballesta et al., 2010; Oztekin et al., 2009; Ruiz and Romero, 1999) However, research information is scarce regarding the examination of root characteristics of grafted plants especially under field production condi tions. In this study, we compared the root length density (RLD) of grafted tomato plants to non grafted plants at different soil depth and sampling positions under varying nitrogen rates. Regardless of grafting, the root length density was more concentrate d within the top 15 cm of the soil profile. This root distribution pattern is in agreement with previous studies on tomato plants (Lecompte et al., 2008; Machado et al., 2000; Zotarelli et al., 2009b) The concentra tion of roots in the upper layer of soils is closely related to the frequent application of nutrients and water which makes this top soil layer more conducive to the proliferation of root growth (Jackson and Bloom, 1990) Consistent with previous reports, a de creasing trend in RLD along the soil profile was also noted in our study. The
154 restriction of root distribution in the deeper soil profile is largely due to the greater soil bulk density and increased level of mechanical resistance and nutrient/water avail ability Interestingly, the grafting effect varied remarkably with the soil depth. The RLD values were significantly greater in grafted vs. non grafted plants at the 0 15 cm soil profile and at 30 60 cm or 60 90 cm, while similar levels of RLD were detecte d between grafted and non grafted plants within the soil depth at 15 30 cm. The increase of RLD due to grafting at the top and deeper layers of the soil profile may have contributed directly to the enhancement of aboveground N accumulation and N uptake eff iciency in grafted tomato plants found in this study. It provided the evidence that tomato plants grafted on certain rootstocks can possess a more extensive root system than non grafted plants, which allows for greater capability of roots to access and tak e up nutrients like nitrogen that can easily move beyond the active root zone. Such findings can be even more meaningful for sandy soils with poor water and nutrient holding capacity. Although the difference between the two rootstocks used did not consiste ntly stand out in this 2 year study, it is possible that significant variations in nutrient uptake related to root characteristics may exist among rootstocks with different genetic makeup. Future research on rootstock selection for these root traits will g reatly advance our understanding of the rootstock capacity in enhancing nutrient uptake efficiency. The three way interaction between nitrogen rate, grafting, and soil depth was significant during the 2010 season, indicating that under the highest nitrogen rate of 3 36 kg N ha 1 and at the soil depth (0 15cm), grafted plants consistently showed significantly higher values of RLD than non grafted plants (data not shown). Furthermore, RLD was
155 significantly greater under the N rate of 3 36 kg N ha 1 than at 112 kg N ha 1 The tendency of RLD to increase with higher N application rates has been reported previously by Sainju et al. (2001) who f ound that RLD was significantly greater at 90 and 180 kg N ha 1 as compared with 0 kg N ha 1 In contrast, the study by Jackson and Bloom (1990) did not show a clear relationship between tomato root distribution and soil N availability. By and large, changes in root a rchitecture often occur in response to variations in the availability and distribution of inorganic nutrients in the soil (Lopez Bucio et al., 2003) High level of N fertilization also tends to promote shoot growth at the expense of root development. As a result, there is often a decrease in the shoot:root ratio for most plant species in response to a reduced availability of N (gren and Franklin, 2003; Marschner, 1995) Conclusion s Gr owth vigor and nitrogen use and uptake efficiency were improved by grafting ot length density than the non grafted plants particularly within the top and deeper layers of the soil profile, which could contribute directly to the enhancement in nitrogen uptake efficiency of grafted plants. More studies are warranted to explore the g enetic variations among tomato rootstocks with respect to their impact on root architecture of grafted plants in relation to nutrient acquisition. Use of grafted transplants may be a promising approach to be considered in the site specific program of best management practices.
156 Table 6 1 Analysis of variance of the effects of nitrogen fertilization and grafting on leaf area index (LAI), total aboveground biomass, accumulated nitrogen (N acc ), nitrogen uptake efficiency (NUE), and N concentrations in lea f blade, petiole, stem, and fruit samples in the 2010 and 2011 field trials at Live Oak, FL. N concentration Effect LAI Aboveground biomass N acc NUE Leaf blade Petiole Stem Fruit 2010 2011 2010 2011 2010 2011 2010 2011 2010 2011 2010 2011 2010 2011 2010 2011 Nitrogen (N) *** *** *** *** *** *** *** *** *** NS NS ** NS Graft (G) *** *** *** *** *** *** *** *** NS NS NS NS NS NS NG NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS NS ,*,**,*** Non significant or significant at the P
157 Table 6 2 Effects of nitrogen fertilization and grafting on leaf area index (LAI), aboveground biomass, N accumulation, and nitrogen uptake efficiency (NUE) of tomato plants during the 2010 and 2011 field tria ls at Live Oak, FL. Treatment LAI (m 2 m 2 ) Aboveground z biomass (Mg ha 1 ) N accumulation y (kg ha 1 ) NUE (%) 2010 Nitrogen (kg ha 1 ) 56 0.74 c 0.90 d 20.97 e 37.3 4 a 112 1.84 b 1.36 c 31.95 d 28.53 bc 1 6 8 2.48 b 2.04 b 48.53 c 28 .88 bc 224 3.75 a 2.60 a 67.48 b 30.12 b 28 0 3.84 a 2.81 a 70.18 ab 25.06 cb 3 36 4.28 a 2.86 a 77.52 a 23.07 c Graft x FL/BE 3.07 a 2.19 a 55.24 a 3 0 21 a FL/MU 3.09 a 2.21 a 56.59 a 3 1 12 a FL 2.31 b 1.89 b 46.48 b 2 5 .1 7 b 2011 Nitrogen (kg ha 1 ) 56 0.44 d 0.61 c 13.35 d 23.85 b c 112 1.83 c 1.79 b 40.75 c 36.38 a 16 8 2.75 bc 2.25 ab 51.93 bc 30.91 a b 224 3.46 ab 2.18 ab 50.05 bc 22.34 b c 28 0 4.13 a 2.61 a 64.06 ab 22.34 b c 3 36 4.22 a 2.83 a 69.79 a 20.77 c Graft FL/BE 2.96 a 2.05 a 48.55 b 2 5 9 0 b FL/MU 3.18 a 2.36 a 56.76 a 3 0 60 a FL 2.27 b 1.73 b 39.65 c 2 2 06 b z Shoot and fruit were included in determination of total aboveground biomass. y A t otal of N accumulation in s hoot and fruit. x Florida 47 Beaufort Florida 47 grafted onto Multifort FL: non Florida 47 Means within a column followed by the same letters do not differ significantly at P ac cording to
158 Table 6 3 Effects of nitrogen fertilization rate and grafting on tissue nitrogen concentrations of tomato plants during the 2010 and 2011 field trials at Live Oak, FL. Nitrogen concentration (mg N g 1 DW) Treatment Leaves P etiole Stem Fruit 2010 Nitrogen 56 24.35 c 19.78 16.96 b 26.05 c 112 28.43 bc 18.46 17.89 ab 26.15 c 168 28.52 bc 18.08 16.68 b 27.76 ab 224 32.55 ab 20.51 18.44 ab 27.87 ab 280 31.75 ab 17.98 19.56 a 26.46 bc 336 33.97 a 22.53 20.16 a 28.04 a Graft z FL/BE 30.87 17.82 b 18.46 27.01 FL/MU 30.74 21.39 a 19.11 26.96 FL 28.18 19.46 a 17.27 27.23 2011 Nitrogen 56 27.75 b 12.90 12.62 c 23.60 112 29.40 ab 13.58 13.19 bc 24.24 168 31.40 ab 13.66 13.99 bc 23.72 224 32.77 ab 14.18 13.75 bc 23.54 280 32.73 ab 14.36 14.70 ab 25.10 336 34.46 a 14.03 16.20 a 24.66 Graft FL/BE 32.03 13.60 a 14.17 23.92 FL/MU 32.22 14.37 a 13.98 24.31 FL 30.01 13.38 a 14.07 24.20 z Florida 47 Beaufort Florida 47 grafted onto Multifort FL: non Florida 47 Means within a column followed by the same letters do not differ significantly at P according to
159 Table 6 4 Analysis of variance of the effects of nitrogen rate, grafting, root sampling position, and soil depth on root length density (RLD) and root surface area density (RSAD) of tomato plants during the 2010 and 20 11 field trials at Live Oak, FL. Effect P value for RLD P value for RSAD 2010 2011 2010 2011 Main Effects Nitrogen (N) 0.0099 0.0002 0.0769 0.7805 Graft (G) <.0001 0.0005 0.0019 0.0041 Position (P) <.0001 <.0001 <.0001 <.0001 Dep th (D) <.0001 <.0001 <.0001 <.0001 Two way interactions N G 0.0245 0.3585 0.3741 0.5451 N P 0.9854 0.9095 0.5563 0.1322 N D 0.1891 0.0356 0.1278 0.9819 G P 0.2044 0.7806 0.6000 0.6032 G D 0.0137 0.0354 0.1052 0.1146 P D < .0001 <.0001 <.0001 <.0001 Three way interactions N G P 0.3037 0.7786 0.8144 0.1744 N G D 0.2721 0.4462 0.4347 0.4052 N P D 0.9101 0.7442 0.9749 0.5269 G P D 0.7328 0.6129 0.5892 0.7672 Four way interaction N G P D 0.6686 0.9089 0.9068 0.5348
160 Table 6 5 Root length density interaction effects between N rates and graft in 2010 ; and N rates and soil depth in 2011 during the field trials at Live Oak, FL. N rates (kg ha 1 ) Graft Soil depth (cm) FL/BE FL/MU FL 0 15 15 30 30 60 60 90 112 0.43 Ab 0.48 Ab 0.40 Aa 0.99 Ab 0.41 Ba 0.14 Ca 0.11 Ca 224 0.57 Ab 0.60 Ab 0.37 Ba 1.81 Aa 0.65 Ba 0.20 Ca 0.15 Ca 336 0.65 Aa 0.79 Aa 0.33 Ba 1.43 Aab 0.49 Ba 0.15 Ca 0.13 Ca z Florida 47 Beaufort Florida 47 grafted onto Multifort FL: non Florida 47 Means within a row followed by the same uppercase letters and means within a column followed by the same lowercase letters do not differ significantly at P
161 Table 6 6 Root length density interaction effects between grafting and soil depth; and sampling position and soil depth during the 2010 and 2011 field trials at Live Oak, FL. Root length density (cm cm 3 ) Graft z Position y Soil depth (cm) FL/BE FL/MU FL P1 P2 2010 0 15 1.4 0 Aa 1.5 6 Aa 0.8 5 Ba 1.70 Aa 0.82 Ba 15 30 0. 47 Ab 0.52 Ab 0.3 7 Ab 0.60 Ab 0.33 Bb 30 60 0. 19 Ac 0.27 Ac 0.1 6 Ac 0.28 Ac 0.15 Bc 60 90 0.13 Ac 0.15 Ad 0.08 Bc 0.15 Ac 0. 09 Bc 2011 0 15 1.5 8 Aa 1. 66 Aa 0.99 Ba 1. 92 Aa 0.90 Ba 15 30 0. 54 A b 0.6 2 Ab 0.38 Ab 0.6 5 Ab 0. 38 B b 30 60 0.17 AB c 0.2 1 Ac 0.1 1 Bc 0.20 Ac 0.12 B c 60 90 0.12 Ac 0.15 Ac 0.1 2 Ac 0.15 Ac 0.11 A c z Florida 47 Beaufort Florida 47 grafted onto Multifort FL: non Florida 47 y P1: at the plant base; P2: at 15 cm distance from the tomato plant. Means within a row followed by the same uppercase letters and means within a column followed by the same lower case letters do not differ significantly at P
162 CHAPTER 7 ESTIMATING NITROGEN FERTILIZATION REQUIR EMENT FOR GRAFTED TOMATO GROWN IN THE FIELD Introduction U se of vigorous interspecific, hybrid rootstocks has shown great potential for enhancing growth and fruit yields and improving water and nitrogen management in production of several solanaceous vegetables and cucurbits (Colla et al., 2010; Di Gioia et al., 2010; Lee and Oda, 2003; Leonardi and Giuffrida, 2006) As a result fo r example, due to increased nitrogen use efficiency (NUE) or greater yields at comparable nitrogen (N) inputs, p revious studies suggested up to 33 50% reduction in fertilizer inputs for grafted watermelon production in contrast to the standard recommendati on for non grafted plants (Colla et al., 2011; Lee and Oda, 2003) Similar enhanced NUE have also been reported on grafted tomato plants with v igorous rootstocks (Leonardi and Giuffrida, 2006). However a lthough several studies were conducted on yield responses to fertilizer application and particularly N rates in non grafted tomato production, few studies have attempted to estimate the N require ment for maximizing the fruit yield of grafted tomatoes under field conditions Moreover, g iven the increasing interest in using grafted plants in field tomato production and the possibility of incorporating grafting into vegetable best management practice s ( BMP ) programs, research is needed to determine the N fertilization requirement for grafted tomato under specific production conditions. Nitrogen is vital for plant growth and development. It is generally the nutrient applied most frequently in higher am ounts than any other nutrient. This is particularly true for crops of high economic importance such as tomatoes grown in a commercial production system where maximizing yield and profit are the primary goal s In these
163 high value systems, growers rely on i ntensive management with more than recommended inputs of fertilizers and irrigation water, especially in growing regions characterized by sandy soils with low intrinsic water and nutrient retention capacities. G iven the dynamic and mobile nature of N in th e soil profile, N in excess of that used by the crop may lead to nitrate leaching and loss to surface and ground waters which represents a serious environmental concern (Obreza and Sartain, 2010; Poh et al., 2011; S imonne et al., 2004; Simonne et al., 2006; Zotarelli et al., 2009a) To account for both economic and environmental costs inherent to applied N fertilization in production systems, management practices that are more efficient in fertilizer and irrigation water use are often developed and provided to growers around the world. A key pillar of these recommendation packages is the development of the optimum N application rate, i.e., a minimum amount of N required to maximize the marketable fruit yield of a giv en crop in a given environment. For example, the recommended optimum N fertilizer application rate for commercial tomato production in sandy soils in Florida is 224 kg ha 1 (Olson et al., 2010) However, it is common for growers to use much higher rates to compensate for the risk of yield reductions under deleterious growing conditions (Cantlif fe et al., 2006; Scholberg et al., 2000) Optimal N rate recommendations are traditionally derived from field or greenhouse trials in which crop yield response to increasing N rates is measured. A number of mathematical models exist for estimating the op timum N rate. Tageldin and El Gizawy (2005) summarized the most commonly used models, including: the linear, splined linear plateau, quadratic and splined quadratic plateau, square root, and exponential. Specifically, comparing five of these different models, Cerrato and Bla ckmer (1990)
164 found that all models fit the data equally well based on the coefficient of determination (R 2 ) statistic but concluded that the quadratic plateau was the best to describe corn yield response to N rates based on the residuals of each model. Furthermore, W illcutts et al. (1998) used the logistic, linear plateau, and quadratic models a nd found that the logistic model best described the lettuce yield response to the applied N Anderson and Nelson (1975) noted that the optimum N rate estimation often varies with the models used. Although commonly used for de scribing crop response to N rate s quadratic models tend to overestimate the response if the maximum point on the curve is taken as the best fertilization rate (Ozores Hampton et al., 2012; Willcutts et al., 1998) Often times, the optimum N rate is determined as a point estimate based on the best fit model without accounting for the statistical uncertainty associated with estimating the parameters in the model. Hernandez and Mulla (2008) stressed the need for accounting for the variability which in general results from factors such as annual climate differences as well as within field soil differences They pointed out that confidence interva ls (CI) of the estimated optimum N rate should be evaluated and examined before making final N recommendations. A number of methods are available for the determination of CI associated with the estimated optimum N rates. Hernandez and Mulla (2 008) reviewed four different methods for estimating CI assuming a quadratic response model. In that case, Wald type CIs tended to overestimate the lower bound and underestimate the upper bound of the optimum rate as compared to profile likelihood based CI s. In addition, they found that bootstrap methods present greater potential in quantifying uncertainty in optimum N rates. More recently, Jaynes (2011) used a parameter bootstrap approach to determine
165 the expectation, confidence bands, and cum ulative probability distributions for the economic optimum nitrogen rate (EONR) using different models The author concluded that when optimum N rates are estimated from yield data, both point estimates and a measure of its statistical reliability should a lways be reported. T he objectives of this study were thus to: 1) compare the goodness of fit of five different models for marketable tomato fruit yield as a function of N rate, 2) estimate the optimum rate of N fertilizer and corresponding maximum marketab le fruit yield in field production of grafted tomato for sandy soils in Florida, and 3) assess the uncertainty associated with the estimated optimum N rate s Material and Methods Field E xperiment The field studies of grafted and non grafted tomato product ion were carried out at the Suwannee Valley Agricultural Extension Center in Live Oak, FL (30.17 N, 82.59 W) during the spring seasons of 2010 and 2011. D etails of the experi mental set up were described in Chapter 5 Briefly, in both years, experimental plots were arranged in a split plot design with four blocks (replication). The main/whole plot treatments were 12 combinations of two irrigation regimes and six N fertilization rates which were arranged in a randomized complete block. The two irrigation re gimes included: 100% irrigation regime of University of Florida Institute of Food and Agricultural Sciences ( UF/IFAS ) recommendation, i.e., 9354 L/ha/day/string, and 50% irrigation regime corresponding to a 50% reduction of the 100% irrigation regime. The six nitrogen rates were 56, 112 1 6 8, 224, 28 0 and 3 36 kg ha 1 with a preplant application at 56 kg ha 1 The subplot treatments included Florida 47 grafted onto two different Beaufort ( FL/BE ) and Multifort ( FL/MU ) and the non grafted Florida
166 47 (FL) used as a control, all randomized within each whole plot. Each experimental unit consisted of 12 plants A ll treatments were re randomized in 2011. Ammonium nitrate (34 0 0, Mayo Fertilizer Inc, Mayo, FL) was u sed as the nitrogen source. After pre plant fertilization, the remaining N fertilizers were injected weekly through the drip tape based on the fertigation scheduling suggested by Olson et al. (2010) P otassium chloride (Dyna Flo 0 0 15, Chemical Dynamics Inc, Plant City, FL) was also applied to provide equ al amount of K for each N rate treatment. Mature green tomatoes fruit were harvested from 10 plants in each experimental unit. Fruit were graded as extra large, large, medium, and culls and the tomato fruits in each grade were counted and weighed. Marketab le f ruit y ield (Mg ha 1 ), representing the total of extra large, large, medium fruits was calculated. Marketable Y ield R esponse F unctions The marketable yield was modeled as a function of nitrogen rates. Historically, models that parametrically describe th e expected yield (E(Y)) as a predetermined function of nitrogen (X) are used. In this study, five linear and non linear models commonly cited in horticultural literature were considered including: (1) quadrat ic, (2) square root, (3) linear plateau, (4) qua dratic plateau, and (5) exponential models (Cerrato and Blackmer, 1990; Willcutts et al., 1998) ( 7 1) ( 7 2) if X< if X (7 3)
167 ( 7 5) where Y is the marketable fruit yield (Mg ha 1 ) and X is the rate of N application (kg ha 1 ); 1 2 and 3 are constants obtained by fitting the model to the data ; x 0 is the critical rate of fertilization, which occurs at the intersection of the linear response and the plateau lines, and P is the plateau or maximum yield The errors were assumed to be independent and identically distributed from a 2 For each of the 6 combinations of irrigation (100% and 50% irrigation regimes) and grafting ( FL/ BE, FL/ MU, and FL) treatments, each of the five models was fit to the data using (non linear) least squares. The models were compared based on three measures commonly used in model selection including the coefficient of determination (R 2 ), the Akaike information criterion (AIC), and the root mean squared error (R MSE). Proc NLIN of SAS (SAS Institute, Cary, NC) was used to fit the functions to the different trials data sets. Optimal N itrogen R ate D etermination Each of the five models was used to estimate the optimum N rate. Because each of the 5 models except the l inear plus plateau model has 3 parameters, a total of 19 parameters (18 mean parameters plus one variance parameter) would be needed if a separate curve with a common variance was assumed for each of the 6 combinations of irrigation and grafting. In the in terest of parsimony, the significance of the additional parameters was tested using nested model F tests (Rawli ngs et al., 1998) The full model has one curve for each combination of irrigation and grafting. Because the two if X< if X (7 4)
168 grafted treatments ( FL/BE and FL/ MU) are different from the non grafted treatment (FL), reduced model with a separate curve for each of the now four combinations of irrigation and grafting. That is, we tested whether or not the two grafting treatments are different from each other at each irrigation regime. A n F test of the hypothesis below yielded an F value of 0. 97, 1.04, 1.00, 1.03, and 1.01 based on 6 and 126 numerator and denominator degrees of freedom, for the exponential, linear plus plateau, quadratic plus plateau, quadratic, and square root model, res pectively. As a result, we failed to reject the null hypothesis and then concluded that the two grafted treatments ( FL/BE and FL/ was chosen as the final model for the relationship b etween total yield and nitrogen rate. H o : 4 curves: One curve for each combination of irrigation and grafted/non grafted H a : 6 curves: One curve for each combination of irrigation and grafting Therefore, for example with the exponential model, the final mo del selected for the expected total marketable yield as function of irrigation, grafting, and nitrogen was as follows: ( 7 6 ) Thus, Y is assumed to be normally distributed with mean E(Y) and varian 2 An analysis of the studentized residuals showed no violations of these assumptions. if irrigation = 50% and grafting = FL if irrigation = 100% and grafting = FL
169 Using th ese final model s optimal nitrogen fertilization rates were then estimated by setting the first derivative of each function except the linear plus plateau eq ual to 0 and solving for X as follows. For the exponential model the optimal value for each combination of irrigation and grafting is ( 7 7 ) where p is the pre specified proportion of the asymptotic expected marketable yield (e .g., 0.95). For the quadratic model and the quadratic plus plateau model the first derivative is: therefore ( 7 8 ) As f or the square root model the first derivative is : therefor e ( 7 9 ) For the linear plus plateau model (Eq. [ 7 3 ] above ), optimum rates of fertilization were represented by the intersection between the linear response line and the plateau line Finally, because the regression parameters ar e unknown, estimated values optimal nitrogen.
170 Estimation of Uncertainty around O ptimum N Rate of Fertilizer Estimating the regression parameters introduce s a level of uncertaint y that should be accounted for when reporting the optimal nitrogen. We considered two methods to account for this uncertainty: a delta method approach, and a bootstrap approach. Because the optimal nitrogen values are functions of the regression parameter s, the multivariate delta method can be used to obtain 95% confidence intervals for the optimal values. Although the method accounts for the parameter uncertainty, the results are based on the asymptotic normality of the parameter estimates. However, the n ormal approximations tend to perform poorly when the model is highly nonlinear in its parameters (Hougaard, 1985) In particular, the distributions tend to be asymmetric (Jaynes, 2011) Although others have proposed profile likelihood based confidence intervals to account for this asymmetry when estimating the uncertainty in optimal nitrogen rates (Hernandez and Mulla, 2008) such methods require re parameterizing the model with a parameter equal to the optimal nitrogen rate. Due to the complex nature of our optimal nitrogen rates, we instead considered a bootstrap method to obtain uncertainty estimates (Efron and Tibshirani, 1986; Hernandez and Mulla, 2008) We used a residual re sampling algorithm to generate 2000 bootstrap samples (Efron and Tibshirani, 1986; Fox, 2008) First, the parameters were estimated using and the residuals, were obtained for 4 To create a bootstrap sample, a randomly selected residual is added to each fitted value and the regression is carried out using the synthetic response variable. That is, for each i randomly select j the synthetic response is defined as ( 7 10 )
171 The regression parameters were then estimated using the as the response variable for calculation of the optimal nitrogen values. The median, 2.5 and 97.5 percentiles were reported. It should be noted that the algorithm used to estimate the parameters converged in 1995 of the 2000 bootstrap samples for the 2010 dataset and 1977 of the 2000 bootstrap samples for the 2011 dataset. Results Tomato Marketable Fruit Yield M odel Overall, all five models fit the ma rketable fruit yield to N rates quite well in 2010 within the same combination of irrigation and grafting types (Fig ure 7 1). However, in 2011, there was no variation in expected yield for N values greater than 56 under the 50% irrigation regime. As a resu lt, none of the models considered here could accurately describe the yield response curve. Hence, for 2011 we only used the data under the 100% irrigation regime. Across different combinations of irrigation and grafting types, R 2 values ranged from 0.71 wi th the exponential or square root function to 0.86 with the linear plateau function in the 2010 season (Table 7 1) In contrast, in the 2011 season, these values ranged from 0.77 with the linear plateau to 0.90 0.91 with the exponential and square root fun ctions under the 100% irrigation regime (Table 7 2) However, within the same combination of irrigation and grafting type, there is little difference in R 2 values of these five models. Similarly, when the evaluation of the goodness of fit of the five model s was based on the AIC values, it also appeared that the linear plateau function present ed the lowest values of AIC across the irrigation regime s and grafting type s in 2010. More specifically, in the 2010 experiment, AIC values ranged from 172 for the line ar plateau or square root function to 191 for the quadratic plateau or exponential
172 model. In contrast in 2011, these values ranged from 164 to 188 for the square root or exponential model and the linear plateau model, respectively. This inconsistent patter n among the models was also observed for the RMSE value s The linear plateau function had the lowest model fit residuals i.e., the lowest RMSE value of 7.0 as compared to the exponential function with RMSE value of 10.3 across different combinations of i rrigation and grafting treatments in 2010, whereas the opposite was observed in 2011. In essence, over these two seasons, one single model could not just be selected as the best fit model for the expected total marketable yield as a function of nitrogen ap plication rate within each irrigation regime. Although all models performed relatively well, t he linear plateau model in 2010 and the exponential model in 2011 outperform ed all other models according to R 2 AIC, and RMSE across all treatment combinations. Optimal N itrogen R ate and Confidence I ntervals Within the same combination of irrigation regime and grafting, estimated optimal N fertilization rates varied considerably with t he type of model used (Table 7 3). In the 2010 experiment, with the exponential function and assuming 95% of the asymptote, the optimal N fertilization rates (minimum N rates to achieve maximum yields) were 207 and 294 kg N ha 1 for the non grafted and grafted plants, respectively, under the 50% irrigation regime. With the 100% irriga tion regime in 2010, these estimates N rates were 236 and 307 kg N ha 1 for the non grafted and grafted treatments, respectively. In contra s t, the linear plateau model which presented the best goodness of fit among the five models in 2010 resulted in the l owest estimated optimum N rates relative to those of the exponential model. With the linear plateau model, optimum N rates needed to maximize the marketable fruit yields under the 50% irrigation in 2010 season were 161 and 216 kg N ha 1 for non grafted and grafted plants, respectively. Similarly, the
173 quadratic plateau also led to slightly lower optimum rates relative to the exponential model. O f all these five models evaluated, the square root model and with few exception the quadratic model resulted in the highest optimum N rates, irrespective to irrigation and grafting combinations. Specifically, under the 100% irrigation regime in 2010, these estimated optimum N rates using the square root model were 276 and 325 kg N ha 1 for the non grafted and grafted t r eatments, respectively (Table 7 3). In comparison with the optimum N rates obtained by using the linear plateau model, under the 50% irrigation regime for example, the square root model in 2010 overpredicted the optimal N fertilization rates for grafted a nd non grafted plants by 115 and 11 2 kg N ha 1 respectively. These differences between optimum N rates of the quadratic model and the linear plateau were 71 and 110 kg N ha 1 for grafted and non grafted plants, respectively. Furthermore, similar considerab le differences in the optimum N rates predicted by these models were also observed in the 2011 data. The quadratic and the square root models led to the highest estimated optimum N rates whereas the quadratic plateau and the linear plateau resulted in the lowest rates. The differences in estimated N rates between models were 97 and 102 kg N ha 1 for the non grafted and grafted plants, respectively. Over the two seasons, t hese differences in the performance of each model to predict the estimated optimum N ra tes were furt her illustrated in the fig ure 7 2. Overall the estimated N rates predicted in 2011 by each of the models were consistently lower than those in 2010. Moreover, over these two seasons, estimated optimum N rates needed to maximize marketable fr uit yields of grafted plants were consistently higher than those of the non grafted plants. In addition the level of differences in optimum N rates between grafted and non grafted plants varied with the
174 type of models considered. For example, under the 10 0% irrigation regime in 2010, optimum N rates to maximize yields of grafted plants were consistently higher than those of the non grafted plants by 30, 8, 25, 5, and 18% with the exponential, linear plateau, quadratic plateau, quadratic, and square root mo del, respectively. Furthermore, there were also great differences among the five models evaluated in terms of the confidence interval associated with the estimated optimum N rates irrespective of the method of confidence intervals estimati on used in both s easons (Table 7 3). Overall, in 2010, the linear plateau function with the best goodness of fit yielded the smallest confidence bands among the five models. In contrast, the square root or the exponential model led to the largest confidence band around the ir respective estimated optimum N rates. Moreover, although the point estimates based on the bootstrap and delta methods were nearly identical for the same combination of irrigation and grafting, estimates of the uncertainty associated with these optimum N rates were quite different. For example, in 2010, with the 95% asymptote method to estimate the optimal nitrogen for grafted plants under the 100% irrigation regime using the exponential model the confidence intervals of the point estimate (307 kg N ha 1 ) were [21 4 kg N ha 1 512 kg N ha 1 ] and [17 6 kg N ha 1 438 kg N ha 1 ] for the bootstrap and delta methods, respectively. With the linear plateau model, these confidence intervals for the same treatment combination were only [170 kg N ha 1 214 kg N ha 1 ] and [163 kg N ha 1 212 kg N ha 1 ] for the bootstrap and delta methods, respectively (Table 7 3) In this case the confidence interval based on the delta method was symmetric with lower and upper limits given by a shift of approximately 25 units to the left and right of the point estimate. For the bootstrap method, however, the confidence interval was
175 asymmetric with about 19 units to the left of the point estimate and 25 units to the right of the point estimate. A similar trend was consistently observed with all the confidence intervals associated with the estimated opt imum N rates despite the model used. The se result s illustrated how the delta method produces confidence intervals that are symmetric around the estimate point whereas the bootstrap method produces asymmetric confidence intervals. Such asymmetry is expected in nonlinear models and this asymmetry cannot be accounted for using the Wald type confidence intervals based on the delta method. Maximum A ttainable Yields Predicted by Each M odel and C o nfidence B ands The relatively higher optimal N fertilization rates for the grafted plants also corresponded to higher maximum attainable yields for the grafted plants compared with the non grafted plants, irrespective of the irrigation regime and type of m odel used (Table 7 4). Specifically, in the 2010 season with the exponential model, the maximum fruit yields were 61 Mg ha 1 under the 100% irrigation regime and 71 Mg ha 1 under the 50% irrigation regime with the grafted plants, whereas these values were 4 5 and 52 Mg ha 1 with the non grafted plants. Similar differences in estimated yields of grafted and non grafted were also shown with other models and in the 2011 season as well. However, within the same combination of grafting type and irrigation regime, the magnitude of the differences in the estimated maximum yields among the five models seem ed narrow. For example, under the 50% irrigation regime in 2010, the differences between the highest and the lowest estimated maximum yields were 3.25 Mg ha 1 and 3 .95 Mg ha 1 for the non grafted and grafted plants, respectively. In 2011, under the 100% irrigation regime, these differences were 4.5 2 Mg ha 1 and 2.7 1 Mg ha 1 for the non grafted and grafted plants. In general, t he quadratic response model and to some
176 e xtent, the exponential model consistently predicted higher maximum fruit yield than did the linear plateau or the quadratic plateau model. Similar to the estimated optimum N rate results the bootstrap method led to an asymmetric confidence band around the estimated maximum marketable yields whereas the confidence intervals using delta method were almost symmetrical around the estimated yield s Specifically, in the 2010 experiment, on average, the delta method was symmetric with lower and upper limits given by a shift of about 4.6 units to the left and right of the point estimate. For the bootstrap method, however, the confidence interval was asymmetric with 4.1 units to the left of the point estimate and 6.0 units to the right of the point estimate. Discuss ion Fruit yield responses to increasing N fertilization application rates can be described by a wide range of mathematical functions (Wood, 1980) Examination of the goodness of fit of the relationships between tomato marketable fruit yields and N rates in this study revealed discrepancies among the five models considered over the two seasons. The disparities with respect to the ability of various models to describe yield responses to N application rates have been noted previously (Anderson and Nelson, 1975; Cerrato and Bl ackmer, 1990; Ozores Hampton et al., 2012) Using R 2 AIC, and RMSE as model selection criteria, our results suggest that the linear plateau had the best fit out of the five models considered in 2010 whereas the exponential or square root function perform ed better than the other models in 2011. This indicate d that the expected marketable yield exhibits an apparent plateau and does not change significantly for any further increase in the N rate once the maximum attainable yield is reached. In other words, t here is no significant evidence that suggests a decrease in
177 fruit yields for higher N rates used in this study This concurs with the findings of Neeteson and Wadman (1987) on sugar beet and potato, Cerrato and Blackmer (1990) and Bullock and Bullock (1994) on corn, and Harris et al. (1992) on potato. U se o f each of these five models resulted in considerable differences in the estimated optimum N fertilization rates. Over the two seasons and irrespective of the grafting types and irrigation regimes, the quadratic and the square root response models tended to predict the highest optimum N rates whereas the linear plateau almost always yielded the lowest optimum N rates. Such differences in N rate studies have been reported previously (Bullock and Bullock, 1994; Cerrato a nd Blackmer, 1990; Jaynes, 2011; Neeteson and Wadman, 1987) Specifically, Bachmaier and Gandorfer (2 012) showed that the quadratic model result ed in much higher estimates of the optimum N rates than the linear plateau model. Moreover, the authors stressed that although the quadratic model is characterized by symmetry around the maximum yield, there is n othing intrinsically symmetric about yield response, especially in the field conditions with considerable local heterogeneity. Estimates of optimum N rates using fitted yield response curves should always be accompanied by a measure of statistical uncertai nty associated with the estimate. Uncertainty estimates provide additional information necessary to derive more accurate N fertilizer recommendations for growers. As discussed by Babcock (1992) several factors may lead to uncertainty in optimal nitrogen rates, including measurement errors, spatial heterogeneity within the experimental site, and both inter and intra annual temporal v ariations of important environmental factors such as temperature and rainfall. Recently, the importance of uncertainty estimation has been emphasized in the
178 literature (Hernandez and Mulla, 2008) In our study, the delta method and a bootstrap method were used to estimate confidence intervals of estimated optimum N rates of the different models. Because the delta method is based on asymptotic normal approximations, the resulting uncertainty estimates were symmetric around the point e stimate. However, because of the nonlinearity of the models used, asymptotic normality of parameter estimates typically does not hold true. In contrast, the bootstrap method used to obtain uncertainty estimations produce asymmetric confidence intervals. It has been shown that the traditional Wald type confidence intervals that assume asymptotic normality of parameter estimates, such as those estimated using the delta method, are not appropriate for estimating confidence intervals for nonlinear models (Cook and Weisberg, 1990; Hernandez and Mulla, 2008) In the context of estimating optimum N rates, the Wald type confidence intervals tend to produce narrow confidence bands by overestimating the lower bound and underestimating the upper bound (Hernandez and Mulla, 2008) In addition, the high level of uncertainty in the estimation of optimum N rates is reflected in the magnitude of the length of the c onfidence intervals. In our study, the length of the confidence intervals with the bootstrap method ranged from 4 4 to 360 kg N ha 1 with the smallest length by the linear plateau model and the highest value by the square root model. With the delta method, these lengths ranged from 4 9 for the linear plateau to 271 kg N ha 1 for the exponential model. Confidence intervals around optimum N rates of similar wide lengths with the bootstrap methods were also previously reported (Bachmaier and Gandorfer, 2012) suggesting the difficulty and care that should be taken in formulating N input recommendation s to growers based on optimum N rate studies. Confidence intervals for
179 optimum N fertilization rates are wide when differences among replications are large and/or the curves are flat and do not sink significantly beyond the maximum yield (Neeteson and Wadman, 1987) The magnitude of the length of the confidence intervals around the estimated N rates could also provide some level of evidence in the ability of the different models to accurately predict the optimum N rate needed to maximize the marketable yields (Jaynes, 2011) In this study, in 2010, confidence intervals around the estimated N rates by the linear plateau an d the quadratic plateau were overall within the range of the N fertilization rates tested (56 to 336 kg N ha 1 ), indicating that these optima can be used for developing somewhat more realistic N fertil ization recommendations. In con tra s t, the opposite was true for the confidence intervals around the estimated optimum N rates of the quadratic square root and sometimes the exponential model s with the high endpoints of the confidence bands out of the range of the tested N rates, irrespective of the irrigati on regimes. Similar results were also observed with the 2011 data. When the lengths of these confidence bands were closely examined and the confidence bands out of the ranges of tested N rates were removed, we can suggest that over the two seasons, the rec ommended N rates could be 211 to 240 kg N ha 1 for grafted plants vs.163 to 180 kg N ha 1 for non grafted plants. Optimum N rates reported by Scholberg et al. (2000) using the quadratic function ranged from 125 to 575 kg N ha 1 across various tomato production regions in Florida. More recently, s Hampton et al. (2012) estimated optimum N rates of 172 and 298 kg N ha 1 to maximize marketable yields using the quadratic plateau model. Corresponding maximum marketable yields estimated by these authors were 100 and 86.4 Mg ha 1 respectively.
180 In general, the five models in our study led to quite similar predict ed maximum marketable yield under the same irrigation and grafting combination This result provides further evidence that some of the models, especially the quadratic and the square root models led to an overestimation of the optimum N rates. In our study maximum attainable marketable fruit yields ranged from 56 to 71 Mg ha 1 for the grafted plants and from 43 to 54 Mg ha 1 for the non grafted plants, depending on the irrigation regime. These ranges of yield are comparable with the previous report s by Zot arelli et al. (2009b) and Poh et al. (2011) on tomato cultivar Florida 47 grown in North Florida. However, with the quadrat ic function, Scholberg et al. (2000) have also reported maximum total yield ranging from 59 to 100 Mg ha 1 for tomato production in Florida. In an updated review summarizing several decades of research on fertilization and tomato yield across Florida, Hochmuth and Hanlon (2011) concluded that a yield increase was rarely obtained with N rates above 224 kg ha 1 It is noteworthy that in our study, within the same combination of irrigation regime and grafting type, the optimum N rates estimated in 2011 were all slightly lower than the rates estimated in 2010. However, the corresponding maximum attainable fruit yields were in the same range across the two seasons. It is likely that tomato fruit yields were less responsive to increasing N fertilization application in 2011 a s compared with the 2010 spring season. In fact, the 2011 production season was drier than 2010 due the lower rainfall level ; which may have led to l ess N leaching in 2 011 The results demonstrated that yield responses to increasing N inputs may be season specific and could be greatly influence d by the environmental conditions during the production season. Similar inter annual variations in tomato yield response to N fertilization as a
181 result of the change in environmental conditions were previously reporte d in Florida (Poh et al., 2011; Zotarelli et al., 2009b) The comparison between grafted and non grafted tomato production indicated that grafted tomatoes grown in sandy soils with drip fertigation require greater l evel of minimum N rate to reach the maximum marketable fruit yield potential as compared to non grafted plants despite the irrigation regime. However, this higher N rate also led to greater maximum yield relative to that of the non grafted plants, implying the greater responsiveness of grafted plants to N rates. The i ncrease in N application requirement of grafted tomato plants in contrast to the non grafted plants may be due to the higher demand of photoassimilates which should be produced in order to sust ain higher sink strength of the grafted plants relative to the non grafted plants In fact, grafted tomato plants have often been shown to significantly increase yield components such as number of fruit per plant and average fruit weights (Di Gioia et al., 2010; Leonardi and Giuffrida, 2006) In essence, the two vigorous interspecific tomato hybrid rootstocks requirement did no t differ between the two grafted treatments. To our knowledge, this is the first study on the level of N fertilization requirement for grafted tomato production in the U.S. With the growing interest among tomato growers in using grafted transplants, more s tudies are warranted to determine the yield response of different scion rootstock combinations to N application in order to provide recommendations for N fertilization program in grafted tomato production. Conclusions In this 2 year study with the goal to develop N fertilization requirement s for grafted tomato production, linear plateau, quadratic plateau, and exponential models overall
182 performed better in terms of deriving optimum N rates in comparison to the quadratic and square root models. In general, e stimated optimum N rates needed to maximize the marketable fruit yields were consistently higher with grafted plants as compared to the non grafted plants, irrespective of the irrigation regimes. In order to more accurately formulate N fertilization recomm endations, the uncertainty associated with the estimated optimal nitrogen rates should always be estimated The bootstrap method was shown to be more reliable than the delta method in estimating the uncertainty in this study. With the exception of the line ar plateau model, t he relatively wide confidence intervals estimated reflected the lack of understanding of all the factors that may have influence d optimum N rates. Considering the N rate estimations and associated confidence intervals using different mod els, the recommended N rates could be 211 to 240 kg N ha 1 for grafted plants vs.163 to 180 kg N ha 1 for non grafted plants. Further studies need to be conducted across various sites and seasons in order to more accurately derive optimum N recommendations with associated measures of uncertainty taking into consideration of the climatic and soil conditions. Moreover, future research is needed to determine the yield response of grafted plants to N application as affected by different scions and rootstocks in order to provide basis for N fertilization recommendations for field production of grafted tomatoes.
183 Table 7 1. Model selection criterion for the 5 different types of models using the 2010 data. Irrigation x Grafting y Type z R Squared AIC Root MSE 50 F L EXP 0.775 178.074 7.866 50 FL LIPL 0.788 176.677 7.641 50 FL QUPL 0.786 178.911 7.868 50 FL QUAD 0.728 182.645 8.652 50 FL SQRT 0.763 179.331 8.075 50 FL/BE EXP 0.753 180.865 8.337 50 FL/BE LIPL 0.768 179.408 8.088 50 FL/BE QUPL 0.766 181.595 8.32 0 50 FL/BE QUAD 0.762 180.032 8.194 50 FL/BE SQRT 0.759 180.294 8.239 50 FL/MU EXP 0.854 178.921 8.006 50 FL/MU LIPL 0.864 177.235 7.730 50 FL/MU QUPL 0.858 180.187 8.080 50 FL/MU QUAD 0.856 178.563 7.947 50 FL/MU SQRT 0.854 178.873 7.999 100 FL EX P 0.763 173.111 7.094 100 FL LIPL 0.769 172.529 7.008 100 FL QUPL 0.769 174.511 7.179 100 FL QUAD 0.757 173.720 7.184 100 FL SQRT 0.766 172.797 7.048 100 FL/BE EXP 0.712 190.969 10.291 100 FL/BE LIPL 0.736 188.852 9.847 100 FL/BE QUPL 0.730 191.416 10.209 100 FL/BE QUAD 0.737 188.716 9.819 100 FL/BE SQRT 0.717 190.538 10.199 100 FL/MU EXP 0.751 178.498 7.936 100 FL/MU LIPL 0.782 175.271 7.420 100 FL/MU QUPL 0.761 179.470 7.960 100 FL/MU QUAD 0.757 177.861 7.832 100 FL/MU SQRT 0.750 178.520 7.9 40 z EXP: Exponential; LIPL: Linear Plateau; QUPL: Quadratic Plateau; SQRT: Square Root; QUAD: Quadratic. y Florida 47 Florida 47 x Both the 50% and 100% irr igation regimes were considered in modeling.
184 Table 7 2 Model selection criterion for the 5 different types of models using the 2011 data. Irrigation x Grafting y Type z R Squared AIC Root MSE 100 FL EXP 0.843 167.633 6.329 100 FL LIPL 0.777 176.051 7.5 42 100 FL QUPL 0.830 171.501 6.742 100 FL QUAD 0.798 173.633 7.171 100 FL SQRT 0.830 169.539 6.585 100 FL/BE EXP 0.912 164.251 5.898 100 FL/BE LIPL 0.888 169.974 6.645 100 FL/BE QUPL 0.905 167.896 6.254 100 FL/BE QUAD 0.903 166.537 6.186 100 FL/BE SQRT 0.912 164.154 5.886 100 FL/MU EXP 0.843 181.284 8.411 100 FL/MU LIPL 0.790 188.263 9.727 100 FL/MU QUPL 0.825 185.837 9.089 100 FL/MU QUAD 0.836 182.282 8.587 100 FL/MU SQRT 0.847 180.661 8.302 z EXP: Exponential; LIPL: Linear Plateau; QUPL: Quad ratic Plateau; SQRT: Square Root; QUAD: Quadratic. y Florida 47 Florida 47 x Only the 100% irrigation regime was considered in modeling.
185 Table 7 3 Est imated optimum nitrogen rates and associated uncertainty bounds with the exponential function in the 2010 and 2011 experiments Irrigation (%) Grafting Model Nopt (kg ha 1 ) Bootstrap Wald Lower Upper Lower Upper 2010 50 Non grafted EXP 207.39 129. 07 363.20 108.84 308.48 50 Non grafted LIPL 160.77 130.13 195.11 112.51 188.56 50 Non grafted QUPL 203.74 148.52 291.77 138.83 266.27 50 Non grafted QUAD 270.54 239.31 348.37 226.06 313.66 50 Non grafted SQRT 273.37 225.93 467.71 194.83 350.18 50 Grafted EXP 294.32 213.74 456.13 186.04 401.71 50 Grafted LIPL 215.74 190.45 244.85 183.30 249.45 50 Grafted QUPL 265.62 217.60 327.43 212.93 319.68 50 Grafted QUAD 286.69 260.42 335.67 250.79 322.32 50 Grafted SQRT 330.51 266.28 526.74 229.33 43 1.87 100 Non grafted EXP 236.11 135.90 484.14 99.60 370.75 100 Non grafted LIPL 174.14 133.27 213.24 139.34 203.20 100 Non grafted QUPL 215.89 150.49 321.99 138.83 293.23 100 Non grafted QUAD 266.57 235.90 348.77 220.90 311.33 100 Non grafted SQRT 276.31 220.60 580.18 183.15 367.01 100 Grafted EXP 307.27 213.98 512.36 175.93 438.46 100 Grafted LIPL 189.07 170.33 214.23 163.46 212.49 100 Grafted QUPL 270.74 216.11 345.16 209.27 330.88 100 Grafted QUAD 280.07 254.82 330.26 244.68 316.32 100 Grafted SQRT 325.23 258.92 584.28 217.57 431.36 2011 100 Non grafted EXP 185.93 125.27 275.57 113.57 256.95 100 Non grafted LIPL 173.41 134.83 207.55 141.92 198.22 100 Non grafted QUPL 164.99 130.19 224.16 120.09 211.17 100 Non grafted QUAD 262.3 1 238.40 315.44 229.82 295.53 100 Non grafted SQRT 261.72 224.52 361.26 207.51 317.28 100 Grafted EXP 242.58 197.16 312.26 186.95 298.97 100 Grafted LIPL 189.74 173.57 214.64 169.78 210.29 100 Grafted QUPL 249.65 215.58 289.11 209.42 289.07 100 G rafted QUAD 273.79 255.56 300.52 251.86 295.59 100 Grafted SQRT 291.36 258.17 354.45 245.31 336.84 z Both the 50% and 100% irrigation regimes were considered. y Only the 100% irrigation regime was considered. For the delta method, the lower and upper b ounds correspond to 95% confidence intervals calculated using the delta method. For the bootstrap method, the point estimate is the median value obtained from the bootstrap samples. The lower and upper values for the bootstrap method correspond to the lowe r 2.5 and upper 2.5% of the bootstrap samples.
186 Table 7 4 Maximum attainable yield values with the exponential function with the 2010 and 2011 experimental data Irrigation (%) Grafting Model Ymax (Mg ha 1 ) Bootstrap Wald Lower Upper Lower Upper 2010 50 Non grafted EXP 52.14 46.72 60.61 45.87 57.94 50 Non grafted LIPL 50.81 46.75 54.72 46.36 54.22 50 Non grafted QUPL 50.76 46.48 55.61 46.11 55.13 50 Non grafted QUAD 54.01 49.96 58.76 49.12 58.10 50 Non grafted SQRT 52.76 48.65 59.19 47. 88 56.48 50 Grafted EXP 70.96 65.61 81.00 63.71 77.90 50 Grafted LIPL 67.50 64.30 71.01 64.25 70.66 50 Grafted QUPL 67.01 63.37 70.94 63.13 70.72 50 Grafted QUAD 68.54 65.38 72.06 65.11 71.62 50 Grafted SQRT 68.07 64.37 76.24 63.03 72.42 100 Non grafted EXP 44.56 39.06 55.30 37.55 51.24 100 Non grafted LIPL 43.01 39.06 47.05 38.04 47.11 100 Non grafted QUPL 42.91 38.49 48.15 38.19 47.27 100 Non grafted QUAD 45.59 41.33 50.10 40.74 49.75 100 Non grafted SQRT 44.56 40.14 52.27 39.48 48.15 100 Grafted EXP 61.16 55.50 72.22 53.47 68.49 100 Grafted LIPL 56.40 53.39 59.42 53.13 59.54 100 Grafted QUPL 57.32 53.76 61.45 53.42 61.05 100 Grafted QUAD 58.51 55.45 61.88 55.14 61.53 100 Grafted SQRT 57.88 54.24 66.94 53.14 61.99 2011 100 N on grafted EXP 49.81 45.72 54.69 45.17 54.09 100 Non grafted LIPL 49.56 45.58 53.66 44.70 53.67 100 Non grafted QUPL 48.20 44.64 51.52 44.56 51.60 100 Non grafted QUAD 52.72 49.03 56.59 48.53 56.49 100 Non grafted SQRT 51.31 47.76 54.96 47.37 54.4 1 100 Grafted EXP 70.73 66.70 75.51 66.36 74.90 100 Grafted LIPL 68.60 65.60 71.83 65.44 71.78 100 Grafted QUPL 68.87 65.95 72.24 65.63 72.09 100 Grafted QUAD 71.31 68.50 74.06 68.44 74.02 100 Grafted SQRT 69.66 67.05 72.52 66.66 72.24 z Both the 50% and 100% irrigation regimes were considered. y Only the 100% irrigation regime was considered. For the Delta method, the lower and upper bounds correspond to 95% confidence intervals calculated using the standard error of the point estimates and assum ing normality. For the bootstrap method, the point estimate is the median value obtained from the bootstrap samples. The lower and upper values for the bootstrap method correspond to the lower 2.5 and upper 2.5% of the bootstrap samples.
187 Fig ure 7 1. Ex amples of model fits for the 2010 data including the model fits for grafted Florida 47 (FL/MU) under the 50% irrigation regime, and the model fits for non Florida 47
188 Embedded in each plot a re the model selection criteria, coefficient of determination (R Squared), root mean squared error (Root MSE), and Akaike information criterion (AIC).
189 Fig ure 7 2 Measured (dot) and fitted (lines) yields of non grafted and grafted tomato plants grown u nder 50 and 100% irrigation regime with the five yield response models, including, exponential (EXP), linear plateau (LP), quadratic plateau estimated optimum N rates (Nopt) values for each model
190 CHAPTER 8 FRUIT QUALITY OF FIE LD GROWN TOMATO AS A FFECTED BY GRAFTING WITH INTERSPECIFIC ROOTST OCKS Introduction Tomato ( Solanum lycopersicum L.) is one of the major vegetable crops widely grown and consumed throughout the world for both pr ocessing and fresh consumption (Foolad, 2007) Tomatoes are rich sources of nutritional components with antioxidant carotene and vitamin C and other phenolics compounds (Burri et al., 2009; Jones et al., 2003) These antioxidant compounds have been shown to provide protection against harmful free radicals and have been ass ociated with reduced risk of some chronic diseases such as cancer and cardio vascular diseases (Giovannucci, 2002) Tomato is susceptible to more than 200 diseases caused by pathogenic fungi, bacteria, viruses, or nematodes (Lukyanenko, 1991) In diverse production regions around the world, a wide range of b iotic and abiotic stresses hinder the growth and development of this vegetable crop which may result in significant economic losses. In addition to breeding of resistant cultivars, integrated pest management practices have been developed among which grafti ng has been successfully used for controlling several soil borne diseases and root knot nematodes in tomato production especially under intensive cultivation (Lee et al., 2010; Rivard et al., 2010a) Disease resistant tomato hybrid rootstocks and interspecific tomato hybrid rootstocks ha ve been developed over the years (King et al., 2010) In order to ensure successful marketing grafted tomatoes must have both enhanced marketable fruit yields and premium fruit quality. Pr evious studies on fruit quality of grafted tomatoes have yielded variable results depending on the scion
191 rootstock combinations. In one study no significant differences in titratable acidity, soluble solids content was found between grafted and self rooted tomato plants, (Khah et al., 2006) while in other studies g rafting has been shown to increase lycopene, carotene, vitamin C, and antioxidant activity in tomato fruit (D avis et al., 2008a; Dorais et al., 2008; Fernandez Garcia et al., 2004a) Furthermore melon plants grafted onto different Cucurbita spp. rootstocks showed a remarkable deterioration of taste and texture in some combinations (Traka Mavrona et al., 2000) Overall, the influence of rootstocks on the quality attributes of grafted vegetables such as tomato and melon is still not conclusive and further studies are still needed. Fruit quality and nutritional quality attributes are complex issues and are affected by various biotic and abiotic factors such as cultivar, maturity level before harvest, production practices, and environmental conditions before and after harvest (Dorais et al., 2008; Foolad, 2007; Simonne et al., 2011) For example, an adequate supply of potassium enhances the titratable acidity of tomato fruit (Adams, 2002) while an abundant level of nitrogen supply may reduce the sugar, (Parisi et al., 2004) and vitamin C contents, (Simonne e t al., 2007) thereby, decreasing fruit quality. In grafted vegetables, as rootstock is shown to positively influence the nutrient and water status and other physiological processes, changes in some quality attributes may be expected as a result of graftin g with specific rootstocks. It is suggested that on a hybrid rootstock scion plant, some of the quality attributes may be influenced by the rootstock as a result of the translocation of the metabolites inherent to the fruit quality to the scion through xyl em and/or change in the physiological processes of the scion (Lee, 1994; Rouphael et al., 2010) According to Davis et al. (2008), grafting can result in changes in quality
192 and these changes may be due to not only the rootstocks but also the scion rootstock interactions. As most of the interspecific hybrid rootstocks used are directly developed from wild species, incompatibility between the rootstock and scion sometime arise as an issue and can also lead to the decrease or deterioration of some of the quality attributes (Davis et al., 2008a; Rouphael et al., 2010) Moreover, as these rootsto cks often tend to greatly increase plant vegetative growth and fruit yields (Di Gioia et al., 2010; Leonardi and Giuffrida, 2006) whether such a physiological change causes a decline of fruit quality needs to be we ll examined. With more new rootstocks becoming more available for grafted tomato production, there is a need for more comparative studies to better understand the fruit attributes of grafted tomato The main objective of this study was to assess changes i n pH, soluble solids content (SSC) and total titrable acidity (TTA), ascorbic acid, carotenoids, total phenolics, and antioxidant activity of tomato fruits as influenced by grafting with interspecific tomato hybrid rootstocks under field production conditi ons. Materials and Methods Fruit Sampling and P reparation Tomato fruit were sampled from a field experiment carried out in Live Oak, FL during the 2010 and 2011 spring growing seasons. Treatments considered in this study involved the determinate tomato cul tivar Florida 47 grafted onto two interspecific Beaufort (BE/FL) and Multifort ( FL/ MU) in comparison with non grafted Florida 47 (FL) and self grafted Florida 47 (FL/FL) Plants were grown in plastic mulched beds with drip irrigation wi th a total nitrogen (N) fertilization application at 224 kg N ha 1 and full irrigation regime. The experiment was arranged in a randomized complete block design with four replications (Blocks) and 12 plants per replication. F ruit
193 at the mature green stage of ripeness were harvested at 80 (June 16) and 75 (June 14) days after transplanting (DAT) in 2010 and 2011, respectively. Fruit were brought to the lab and 8 to10 fruit of similar size and greenness in each treatment were randomly selected and were stored at 20 o C to monitor the fruit ripeness. When fruit reached full ripeness, they were sliced and homogenized in a Waring blender for 1 min stored at 30 or 80 o C and were used for measurements of pH, total titratable acidity (TTA), soluble solids content (SSC), carotene, and lutein) concentrations, total phenolic content, and antioxidant activity. Determination of pH, Soluble Solids Content (SSC), and Total Titratable A cidity (TTA) Frozen homogenized fruit tissues were thaw ed, centrifuged at 4C and 17,600 g n for 20 min The supernatant was then filtered through cheese cloth prior to analysis. The TTA was determined by titration using Titrino Metrohm (model 719 S, Switzerland) of 6.0 g of juice plus 50 mL of water with 0.1 N sodium hydroxide solution until pH 8.2 was reached and the T TA was expressed as percent of citric acid (Roberts et al., 2002) The pH of the diluted juice was determined automatically by the Titrino equipped with a pH electrode. Soluble solids content was measured by a digital refractometer (model ABBE Mark II, Cambridge Instruments Inc, U.S.A) Determination of Ascorbic A cid C ontent The ascorbic acid content (vitamin C) in the tomato fruit w as measured spectrometrically according to the AOAC method 967.21 (Horwitz, 2000) In brief, 2 g of the fresh homogenized fruit tissues were mixed with 20 mL of acid mixture (6% metaphosphoric acid with 2N glacial acetic acid) and were centrifuged 4C and 17,600 g n for 20 min. The supernatant was filtere d through Whatman #4 filter paper before
194 the analysis. A triplicate 1 mL of the clear supernatant from each sample was pipetted into test tube and 0.05 mL of 0.2% 2,6 dichlorophenolindophenol (DCIP) was added to each tube and vortexed and then all test tub es were held at room temperature for 1 hour. Afterwards, one mL of 2% Thiourea reagent in 5% metaphosphoric acid and 0.5 mL of 2% dinitrophenyldrazine (DNPH) in 9N H 2 SO 4 reagent were added to each tube and mixed well. The reaction mixtures were subsequentl y incubated in water bath at 60 C for 3 h our then 2.5 m L of ice cold 90 % (v/v) H 2 SO 4 was slowly added to each tube. Thereafter, 250 L of each sample was pipetted into the microplate and their absorbance was measured at 5 4 0 nm using a microplate spectrop hotometer (BioTek Instruments, Inc., model Power Wave X52, Highland Park, Winooski, VT). T he vitamin C content of the samples was calculated based on the established standard curve. Determination of C arotenoids Methylene chloride was used to extract the ca roten carotene, and lutein) from the fruit tissues. Briefly, after thawing, 5.0 g of frozen homogenized fruit tissues was weighted into screw cap glass tube and was mixed with 10 mL of methylene chloride and the mixture was blended with the hom ogenizer at 3600 rpm for 1 min. The solution was then left to separate into distinct layers; and the bottom clear layer was collected into a 25 mL volumetric flask. This step was repeated several times until a total of 25 mL of clear solution was collected Afterward, 10 mL of each extracted solution was transferred into a beaker and was allowed to dry by flushing/evaporating with nitrogen using the N EVAP 116 nitrogen evaporator. The dried extracted sample was resuspended by adding few drops of Tetrahydrof uran (THF) to the beaker and was transferred into a 10 mL volumetric flask. The solution was brought to a final volume with the mobile phase (70:30, methanol:Tert Butyl methyl ether (MTBE)).
195 Concentrations of and carotene, and lutein were then determined using the reverse phase high performance liquid chromatography (HPLC) [Water Alliance system 2695 Separation Module, 996 Photo Diode Array Detector, auto injector (injection volumes = 10 to 30 L), and column temperature regulator; Water Milford A reversed phase C 30 polymeric analytical column (ProntoSil 250 mm x 4.6 mm I.D, 5 um particle diameter, MAC MOD Analytical, Inc., PA) was used for separations (Simonne et al., 2006) Determination of Total Phenolic C ont ent and Antioxidant A ctivity Phenolic conte nts in the hydrophilic and lipophilic fractions of fruit tissues were determined using the method described previously by Toor and Savage (2005) In brief, 25.0 g of the homogenized fresh fruit tissues was weighted, centrifuged at 4 C an d 17,600 g n for 20 min; and the supernatant was filtered through cheese cloth. The filtered supernatant was used for determination of the hydrophilic fraction of the total phenolic content Further, 25 mL of acetone was added to the pellet from each samp le and the mixture was shaken for 1 hour. Then, the mixture was centrifuged again 4 C and 17,600 g n for 20 min The resulting supernatant from this second centrifuge was used for determination of the lipophilic fraction of the total phenolic content. Th en, each extracted supernatant was appropriately diluted with deionized water. Then, a triplicate 0.4 mL of each diluted supernatant from each sample was pipetted into test tubes and 2.5 mL of freshly diluted 0.2% Folin Ciocalteu reagent was added to each tube. Further, the oxidation reaction was neutralized by adding 2.0 mL of 7.5% w/v saturated sodium carbonated, and the samples were vortexed for 15 sec. The reaction mixtures were then subsequently incubated in water bath at 45 C for 20 min. Then, 250 L of each sample was pipetted into the microplate and the absorbance was measured at 765 nm using a
196 microplate spectrophotometer (BioTek Instruments, Inc., model Power Wave X52, Highland Park, Winooski, VT). The phenolic content of each fraction of the samp les was subsequently calculated, using gallic acid as standard The same supernatants of hydrophilic and lipophilic extracts described above were used for the determination of total antioxidant capacity using the oxygen radical absorbance capacity (ORAC) a ssay previously described by Huang et al. (2002) .This assay uses Trolox (6 hydroxy 2,5,7,8 tetramethylchromane 2 carboxylic acid), a vitamin E analogue and a known antioxidant, as a standard. After appropriate dilution of each extracted fraction with phosphate buffer (75 mM, pH 7. 4), 25 L of each diluted sample and the standards were separately mixed together with 150 L of 0.4 uM fluorescein solution inside the microplate wells. The plate containing the mixtures was then allowed to equilibrate by incubating for 15 min inside the Synergy HT Multi Detection Microplate Reader (BioTek Instruments, Inc., Winooski, VT). Afterwards, the scavenging reaction was initiated by automatically adding 25 uL of AAPH ( 2 2' Azobis (2 amidinopropane) dihydrochloride ) solution for a final volume of 2 00 L. The fluorescence was then monitored kinetically with data taken every min for 35 min. The antioxidant activity, i.e., the AA PH free radical scavenging ability of each sample was subsequently calculated with respect to Trolox, which was used as a sta ndard reference to convert the inhibi 1 Trolox equivale nt antioxidant capacity (TEAC). All dete rminations were carried out in du plicate The dry matter (DM) content of tomatoes fruit was determined by following the AOAC method 920.151 (Horwitz, 2000)
197 Statistical A nalys e s Statistical analyses were performed with a linear mixed effects model using the Proc Glimmix program of the SAS system for Windows (version 9.2, Cary, NC). The data of each quality attribute measurement were analyzed with a model including the main effects of year and grafting and the interaction terms of the two main effects. P 0.05. Results Soluble Solids Contents, pH, and Total Titratable A cidity With the exception of soluble solids contents (SSC), no significant difference was found among fruit from the grafted and non grafted plants over the two seasons for other quality a ttributes such as pH, total titratable acidity (TTA) and the SSC:TTA ratios (Table 8 1). The averaged pH values of grafted and non grafted tomato fruit were 4.5 and 4.4, respectively. The total titratable acidity were as low as 0.25% for the fruit from the grafted plants, with no significant difference among the four treatments. However, SSC ranged greatly from 3.57% to 4.23% with significantly lower SSC in FL/BE, FL/MU, and FL t han the self grafted treatment On average, the SSC measured in the fruit of the self grafted plants were significantly higher by 15% as compared to those in the fruit from grafted plants with the two interspecific hybrid rootstocks. The lowest SSC was obs erved in FL/BE and FL/MU although it did not differ significantly from that of the non grafted treatment. However, this significant decrease in SSC by grafting did not translate into a significant difference in the SSC to acid ratio values between the graf ted and non grated tomato fruit. Values of t he SSC:TTA ratio which is often closely associated with the
198 tomato flavor the self grafted treatment. Seasonal comparisons of these quality related attributes show that pH and the TTA were significantly higher in 2010 as compared to the 2011 season whereas the opposite trend was observed for the SSC:TTA ratio values. Specifically, averaged over the four treatments, values of pH increa sed from 4.39 in 2010 to 4.49 in 2011 while TTA decreased from 0.29 in 2010 to 0.24% in 2011. In contrast, values of SSC did not reveal any significant difference between the two seasons. Ascorbic A cid and Carotenoid C ontents Ascorbic acid levels in the fr uit from grafted plants were slightly lower as compared to those in the fruit from non grafted plants. However, fruit from grafted plants with the two rootstocks showed significantly lower levels of ascorbic acid than the self grafted treatment. On average fruit from the self grafted plants exhibited an increase of about 18% in ascorbic acid concentration (on a dry weight basis) as compared to that of the fruit from plants grafted onto the interspecific rootstocks (Table 8 2). It should be noted that the d ifference was more pronounced when the values were expressed on the fresh weight basis (data not shown). Grafting with interspecific rootstocks had a significant effect on the fruit moisture content, with a significant higher level of in the fruit from pla nts grafted onto the two rootstocks than non grafted and self grafted plants (Table 8 2). Given this difference in the fruit moisture content, values of the antioxidants evaluated were expressed on a dry weight of fruit basis. In contrast, lycopene levels a significant increase as compared to the fruit from self grafted plants. Fruit lycopene concentrations did not differ significantly between the non grafted plants and plants
199 grafted onto the interspe cific rootstocks. Averaged lycopene levels ranged from 370 to about 460 g per g of fruit (dry weight basis) over the two seasons, Similar trends were carotene. With respect to lutein, FL/MU showed a significantly increase d level than FL/BE and FL/FL but did not differ significantly from the non grafted treatment. Lycopene was the dominant carotenoid in the red tomato carotene, and lutein values in fruit from the 2010 growing se ason were all significantly higher than those from the 2011 growing season. In contrast, lycopene contents nearly doubled from 315 in 2010 to 536 g per g of fruit (dry weight) in 2011. Total Phenolic C ontent and Antioxidant C apacity Total phenolic conten ts were predominantly from the hydrophilic fraction as opposed to the lipophilic fraction irrespective of the growing season or grafting treatment (Table 8 3). On average, the hydrophilic fraction of the total phenolic contents was consistently 78% greater than the lipophilic fraction. When averaged overall the treatments, the results showed that the hydrophilic and lipophilic fractions measured during the 2011 season were all significantly higher than the values obtained in 2010 season. Further, when avera ged over the two seasons, there was no significant difference among the treatments with respect to any of the two fractions or total phenolic contents, even though the total phenolic contents measured in fruit from grafted plants with the interspecific roo tstocks tended to be higher than those in the non grafted and self grafted plants. T he antioxidant activity ( expressed in AC per 100 g fruit dry weight ) showed that fruit from FL/BE and FL/MU exhibited the greatest levels of hydrophilic antioxidant a ctivity which was 24% higher than the fruit from non grafted plants. The lipophilic f raction showed a much lower level of antioxidant
200 activity compared with the hydrophilic fraction. FL/MU exhibited significantly lower lipophilic antioxidant activity than the non grafted treatment. Total antioxidant activity of fruit (hydrophilic and lipophilic) was significantly increased in plants grafted onto AC E/ 100 g DW) than the non grafted plants (5761 TE AC / 100 g DW). Yield and Yield C omponents The total and marketable fruit yields averaged over the two seasons were significantly higher in grafted plants with the interspecific rootstocks than non g rafted and self grafted controls (Table 8 4). On average, the increases in total and marketable fruit yields were 36 and 41%, respectively. The yield increase could be attributed to significantly higher number of fruit per plant and average fruit weight. A verage total fruit numbers ranged from 21 fruit per plant for the non grafted and self grafted controls to 24 fruit per plant for the plants grafted onto the two rootstocks with marketable fruit numbers ranging ranged from 17 to 21 fruit per plant, respect ively for the control treatments and grafted plants. However, with the exception of the marketable fruit yields and fruit number per plant were significantly higher in 2011 as compared to the 2010 season, other yield and yield component variables did not v ary significantly between the two seasons. Discussion Perception of sweetness and tartness of tomato fruits are related to the levels of soluble solids content (SSC) and total titrable acidity (TTA) and SSC:TTA ratios, which often define the flavor of the fruit. High SSC and TTA are greatly desirable, especially when the tomato fruits are produced for fresh market (Cuartero and Fernndez Muoz, 1999) Under the conditions of this study, despite the decrease of SSC in grafted plants
201 with the interspecific rootstocks as compared with the self grafted treatment, the SSC:TTA ratios did not differ significantly among the four grafted and non grafted treatments. It is unclear whether fruit from grafted tomato plants would exhibit an inferior flavor and thus could be less appreciated by consumers as compared to fruit from non grafted plants. However, several previous studies have reported an enhancement in the levels of these quality attributes a s a result of grafting (Flores et al., 2010) while others have noticed a decrease in these attributes in different environments (Ulukapi and Onus, 2007) or no effect all tog ether (Qaryouti et al., 2007) In our study, the fruit moisture content was significantly higher in fruit from grafted plants with interspecific rootstocks as compared t o non grafted and self grafted plants. The reduction in the SSC values of the fruit from grafted plants might be related to the dilution effect; however, such a reduction was only in contrast to the self grafted treatment rather than the non grafted contro l. Whether the grafting process has an additional effect on fruit quality beyond the influence of rootstocks deserves more in depth research. Although the definitive dilution effect as a result of grafting with interspecific rootstocks was not found in the present study, the dilution effect has been reported on citrus fruits from more vigorous rootstocks as compared to dwarf rootstocks. Albrigo (1977) demonstrated that fruit from trees grafted on rough lemon (vigorous rootstock) had the lowest soluble solids content and the highest leaf water potentials In tomato, a negative correlation between fruit yield and SSC has bee n reported. (Geor gelis et al., 2004) As shown in our study and previous studies by others (Di Gioia et al., 2010; Flores et al., 2010) grafting with specific rootstocks can result in enhanced tomato fruits yield and average fruit weight. The two interspecific rootstocks used in the present study also
202 possess vigorou s growth characteristics. Therefore, a possible decrease in the SSC values would be expected as a result of dilution in the fruit. Future research need to quantify the levels of reducing sugars including glucose and fructose present in the fruit in additio n to measuring SSC to determine the influence of grafting with certain rootstocks. A more complete analysis would also involve the evaluation of the activity of different enzymes ( invertases, sucrose synthases (SuSys) and sucrose phosphate synthases (SPSs) ) that are involve d in the a ccumulation of sugar in fruits to see if there is any influence in the activities of these enzymes as result of grafting with specific rootstocks. More importantly, precisely, gene expression analyses could also provide more pre cise insights on some of the changes in nutritional quality attributes as related to grafting with interspecific rootstocks. Additionally, tomato fruit quality is also related to the levels of various antioxidants such as ascorbic acid, carotenoids, and ph enolic compounds present in the fruit. Similar inconsistent patterns as a result of grafting with specific rootstocks have also been reported on these attributes. Di Gioia et al. (2010) observed a consistent decline in omato fruit from grafted plants as compared to non grafted plants at various harvest dates; our results are in agreement with this finding with a 13% decrease in the levels of ascorbic acid in fruit from grafted plants as compared to fruits from the self g rafted plants, averaged over two seasons. In contrast, Pogony et al. (2005) noted an increase in the level of vitamin C as a result of grafting. Di Gioia et al. (2010) and Vinkovic Vrcek et al. (2011) suggested that a significant decrease in content of total vitamin C a fter grafting could be a result of redistribution or accumulation of vitamin C in other parts of grafted plants and may also be inherent to
203 higher accumulated biomass in the shoot of grafted plants compared to non grafted plants. However, this assertion co uld not be applied to the carotenoids in tomato fruits. In effect, in our study, when averaged over the two seasons, the highest values of lycopene content were found in fruit from plants grafted onto the interspecific rootstocks, although the difference b etween grafted and non grafted treatments was not statistically significant. Similar increase in the carotenoid contents ( carotene ) was previously reported by Fernandez et al (2004b) on grafted tomato. It is likely that certain rootstocks have a positive influence on the processes leading to the biosynthesis and accumulation of carotenoids especially lycopene in tom ato fruit. Carotenoid synthesis during the ripening of tomato fruit involves a typical developmental transition of chloroplasts to chromoplasts; as photosynthetic membranes are degraded, chlorophyll carotene and lycopene, accumulate (Saltveit, 2005) It is bel ieved that the increase in lycopene synthesis and accumulation is in part related to the high K concentration in the fruit as demonstrated in tomato fruits by Fanasca et al (2006) Potassium is shown to be primordial to the protein synthesis and the activity of acetic t hiokinase (Trudel and Ozbun, 1971) an enzyme needed in the formation of acetyl CoA which is invo lved in the biosynthesis of isopentenyl diphosphate (IPP), first precursor of carotenoids and the mevalonic acid pathway. As a result of enhanced nutrient uptake, more chlorophyll might be accumulated in the photosynthesizing organs (leaves and fruit) of t he grafted plants as compared to the non grafted plants. Possibly, t his will lead to greater degradation of chlorophyll and more accumulation of carotenoid in the fruit of grafted plants than the non grafted
204 plants. However, available literature does not p resent evidence of such a link age between inter specific rootstock effect on the chlorophyll accumulation and the ca rotenoid changes in the fruits. It should be noted that the regulatory mechanisms that control the biosynthesis and accumulation of carotenoi ds and carotenogenesis are very complex and are not quite well understood (Bramley, 2002) In this study, the level of total phenolics, measured as the sum of the hydrophilic and lipophi lic phenolics fractions, remain consistently constant in tomato fruit across grafted and grafted plants,. These results do not agree with those reported by Vinkovic Vrcek et al. (2011) w ho observed a decrease in total phenolics as result of grafting with In our study, by using the oxygen radical absorbance capacity (ORAC) assay, the total (hydrophilic and lipophilic) antioxidant activity showed a significant increase in non grafted fruit. However, plants grafted with the two rootstocks did not differ statistically signi ficantly from the self grafted treatment. In this case, it would be difficult to differentiate the effect of rootstocks from the impact of grafting process. In contrast, Vinkovic Vrcek et al (2011) compared to the respective rootstocks and scion. However, the DPPH ( 2, 2 Diphenyl 1 picrylhydrazy l ) assay was used in their study as opposed to the ORAC used in our study. Significant differences in the antioxidant activity measurements with different methods have often been reported (Ou et al., 2002)
205 Furthermore, nutritional quality attributes are often greatly subject to the variations of the environmental conditions (Sco tt, 2002) Often times, the same tomato cultivars may exhibit changes of fruit quality attributes under different production conditions. In this 2 year study, the levels of most of the fruit quality measurements varied significantly between the 2010 and 2 011 production seasons. It has been pointed out the rootstock effects on fruit quality can influenced by different production environments (e.g. light intensity, air temperature) and cultural practices (e.g. soilless vs. soil culture, irrigation, and fert ilization), type of scion rootstock combinations used, and harvest date (Davis et al., 2008 ). Flores et al. (2010) speculated that in grafted plant s metabolic processes inherent to fruit quality attributes are generally species driven and are larg ely controlled by the scion. However, some of the fruit quality attributes may also be influenced by the rootstock as a result of the translocation of the metabolites to the scion through xylem and/or change in the physiological processes of the scion (Lee, 1994; Rouphael et al., 2010) Conclusions Overall, c omparative analysis of the tomato fruit quality attributes in this study did not reveal any major changes as a result of grafting with interspecific rootstocks alth ough the use of rootstocks resulted in a significant increase in fruit moisture content as compared with the non grafted and self grafted plants. Grafting with the vigorous interspecific rootstocks can improve growth and fruit yields of tomato without any major detrimental effect on the quality attributes and nutritional properties of fruit. The effect of grafting process vs. the rootstock impact need to be further examined.
206 Table 8 1 Quality attributes of tomato fruits from non grafted Florida 47 (FL ), sel f grafted Florida 47 (FL/FL), and grafted Florida 47 with the rootstock s Beaufort ( FL/BE ) Multifort ( FL/ MU) harvested at 80 DAT in 2010 and 75 DAT in 2011. pH SSC a ( o Brix) TTA b (% citric acid) SSC:TTA ratio Season 2010 4.39 a 3 .82 a 0.289 a 13.32 b 2011 4.49 b 3.75 a 0.243 b 15.52 a Treatment FL/BE 4.45 a 3.58 b 0.254 a 14.28 a FL/MU 4.44 a 3.57 b 0.268 a 13.48 a FL/FL 4.43 a 4.23 a 0.281 a 15.14 a FL 4.45 a 3.75 b 0.261 a 14.65 a a SSC: To tal soluble solids content. b TTA: Total titratable acidity. Means (n = 4) within a column followed by the same letter are not significantly different at P 0.05
207 Table 8 2 Moisture, a scorbic acid and carotenoid conte nts of tomato fruits from non grafted Florida 47 (FL), sel f grafted Florida 47 (FL/FL), and grafted Florida 47 with the rootstock s Beaufort ( FL/ BE) Multifort ( FL/ MU) harvested at 80 DAT in 2010 and 75 DAT in 2011. Moisture (%) Ascorbic acid (mg/100g DW) Lycopene (g/g DW) carotene (g/g DW) Lutein (g/g DW) Season 2010 95.12 a 440.13 a 315.46 b 83.60 a 69.67 a 2011 94.89 a 342.52 b 536.16 a 47.80 b 18.21 b Treatment FL/BE 95.36 a 368.23 b 461.36 a 70.58 a 41.29 b FL/MU 95.25 a 365.97 b 433. 82 ab 70.26 a 50.95 a FL/FL 94.58 b 434.13 a 370.48 b 59.05 b 40.87 b FL 94.83 b 396.98 ab 437.56 ab 62.90 ab 42.63 ab Means (n=4) within a column followed by the same letter are not significantly different at P 0.05
208 Table 8 3 Total phenolic content and antioxidant activit y of tomato fruits from non grafted Florida 47 (FL), sel f grafted Florida 47 (FL/FL), and grafted Florida 47 with the rootstock s Beaufort ( FL/ BE) Multifort ( FL/ MU) harvested at 80 DAT in 2010 and 75 DAT in 2011. Total phenolics (mg GAE/100 g DW) Antioxidant activity (mol TEAC/100 g DW) Hydrophilic Lipophilic Total Hydrophilic Lipophilic Total Season 2010 225.33 b 46.3 b 271.6 b 5705.26 a 1362. 2 a 7067.4 a 2011 288.04 a 64 .3 a 352.4 a 5205.15 a 1216.9 b 6422. 1 a Treatment FL/BE 279.50 a 56.9 a 336.4 a 5729.17 a 1335.5 a 7064. 7 ab FL/MU 286.87 a 50.6 a 337.4 a 6349.42 a 1149. 2 b 7498.6 a FL/FL 232.83 a 52. 7 a 285.5 a 5358.35 ab 1295.8 ab 6654. 2 ab FL 227.55 a 61.1 a 288.6 a 4383.88 b 1377.6 a 5761.5 b GAE: Gallic acid equivalents, TEAC: Trolox equivalent antioxidant capacity. Means (n=4) within a column followed by the same letter are not significantly different at P 0.05
209 Table 8 4 Total yield, marketable yield, fruit number, and average fruit weight of non grafted Florida 47 (FL), sel f grafted Florida 47 (FL/FL), and grafted Florida 47 with the rootstock s Beaufort ( FL/ BE) Multifort ( FL/ MU) in 2010 and 2011. Total yield Marketable yield Yield (Mg ha 1 ) Fruit number per plant Average fruit weight (g/fruit) Yield (Mg ha 1 ) Fruit number per plant Average fruit weight (g/fruit) Season 2010 54.9 a 22.1 a 172.9 a 48.2 b 17.4 b 191.6 a 2011 59.7 a 23.6 a 175.1 a 54.9 a 20.2 a 186.4 a Graft FL/BE 66.1 a 24.4 ab 188.2 a 60.7 a 21.0 a 201.5 a FL/MU 65.8 a 25.0 a 184.7 a 59.9 a 20.8 a 200.2 a FL/FL 48.3 b 20.6 c 164.1 b 42.9 b 16.7 b 177.9 b FL 48.8 b 21.3 bc 159.1 b 42.7 b 16.8 b 176.5 b Means (n=4) within a column followed by the same letter are not significantly different at P 0.05
210 CHAPTER 9 ECONOMIC ANALYSIS OF GRAFTED TOMATO PRODU CTION IN SANDY SOILS IN FLORIDA Introduction To date, grafting has been used successfully in vegetable production for disease control and yield improvement in many parts of the world especially in Asia and Europe (Lee and Oda, 2003; Lee et al., 2010) A number of rootstocks have been developed for managing various soil borne diseases and root knot nematodes in production of tomato, eggplant, pepper, cucumber, watermelon, and melon particularly in intensive cultural systems (King et al., 2010; Lee et al., 2010; Rivard et al., 2010a) Moreover, many of these rootstocks demonstrate tolerance to abiotic stresses and show great potential fo r enhancing crop vigor and productivity even under low disease pressure (Di Gioia et al., 2010; Fernndez Garca et al., 2002) In the United States, t he majority of users of grafted seedlings are currently greenhouse h ydroponic tomato growers, while grafting it is still a relatively new technique for open field producers (King et al., 2010; Kubota et al., 2008; Lee et al., 2010) With the phase out of methyl bromide soil fumigan t and new search for integrated disease management practices in field vegetable production, interest in vegetable grafting under field conditions is growing recently in the U.S. (Barrett et al., 2012a; Kubota et al., 2008; Rivard and Louws, 2008) However, high cost associated with using grafted transplants still remains the major concern limiting the adoption of grafting by vegetable growers especially large scale producers (Kubota et al., 2008; Lee et al., 2010) In addition to the costs of rootstock seeds, grafted transplant production requires direct costs associated with space, supplies and materials as well as making the grafts, which ultimately increases the costs of grafted vegetable production in comparison with
211 non grafted plants (Barrett et al., 2012b; Rivard et al., 2010b) For example, estimated prices for grafted tomato transplants ranged from $0.59 to $1.88 as opposed to $0.13 to $0.76 for non grafted plants in two different transplant p roduction operations in the U.S. (Rivard et al., 2010b) A recent study on grafted heirloom tomato production demonstrated the economic feasibility of using grafted plants for root k not nemade control when there is a high population of root knot nematodes in the field (Barrett et al., 2012b) While the cost effectivness of growing grafted tomato plants as a result of yield enhancement has been show n in greenhouse operations, little information is available as to whether grafting can be used economically beyond disease control in open field production. In general, efficient production practices can only be fully integrated into endogenous agricultura l management practices when growers can also sustain some economic benefits by using the techniques (Baggs et al., 2000) Considering the multifaceted benefits of vegetable grafting, a comprehen sive approach involving different production scenarios is needed to evaluate the economic feasibility of the use of grafted tomato transplants as a viable component of field tomato production systems. The main objective of this 2 year study was to determin e the costs and benefits of using grafted transplants for field production of fresh market tomato in fumigated sandy soils in Northeast Florida. A partial budget analysis of grafted vs. non grafted tomato production was performed to determine if the additi onal costs associated with grafting can be offset by the improved marketable fruit yield.
212 Materials and Methods Field P roduction of F resh M arket T omato Partial budget analysis conducted in this study was based on field experiments carried out during the 20 10 and 2011 spring growing seasons at the North Florida Research and Education Center Suwannee Valley in Live Oak, FL on a Blanton Foxwort Alpin Complex sandy soil type. T he field grown determinate tomato cultivar Florida 47 ( Seminis Vegetable Seeds, Inc. Saint Louis, MO) was used as scion which was grafted onto t wo c ommercially available interspecific hybrid rootstocks Beaufort ( FL/ B E ) and Multifort ( FL/ M U ) and the non grafted Florida 47 (FL) used as a control These two rootstocks are among the mos t widely used tomato rootstocks in the United States at present. In both years, five weeks before transplanting, field plots were disk ed and plow ed followed by soil fumigation using Telone C 35 (Dow AgroSciences, LLC; Indianapolis, IN) at the rate of 196. 4 L ha 1 Grafted and non grafted plants were transplanted in raised beds with plastic mulch and drip irrigation on 29 Mar. 2010 and 1 Apr. 2011. Beds were 28 inches wide and spaced 5 ft apart with 18 inch in row spacing for open field tomato production. I n both seasons, a randomized complete block design with four replications (blocks) was used. Plants were grown with the current University of Florida Institute of Food and Agricultural Sciences ( UF/IFAS ) recommendation for irrigation and nutrients program for commercial tomato production (Olson et al. 2010) Split application of fertilizer treatments through the drip tape was done weekly throughout the growing season and in both years. The plants were trained using the standard stake and weave system. Other cultural practices including pest control f ollowed current recommendations for commercial field tomato production in Florida (Olson et al., 2010) Mature green tomato fruit were harvested at 80 and 88 days after
213 transplanting ( DAT ) in 2010 and at 75, 8 5 and 92 DAT in 2011. F ruit w ere graded as extra large, large, medium, and culls (small fru it and defected fruit) and t he tomato fruit in each grade w ere counted and weighed Marketable fruit y ield (number of 25 lb cartons per acre ), representing the summation of extra large, large, medium fruits was then calculated. Costs o f G rafted and Non G ra fted Transplant P roduction In this study, the costs of grafted and non grafted transplants were estimated following the procedure described by Barrett et al. (2012b) All the materials, supplies, and labor associated with the p roduction of grafted and non grafted transplant were estimated to calculate the costs of non Base Production Cost Model for Fresh Market T omato in F ield F or the purpose of this study, a base cost model for growing, harvesting, and acre field with a raised bed polyethylene mulch production system was established, using the existing crop budget model for fresh market tomato produ ction in the Manatee/Ruskin area developed by the University of Florida Center for Agribusiness (2009) Costs of production inputs including transplants and fertilizers during the 2010 and 2011 trials were estimated and used in developing the actual production budget. The variable costs of irrigation were based on information provided by Pitts et al. (2002) A harvest and marketing charge of $3.39 per 25 lb carton was used to estimate the harvest and marketing costs for total marketable tomato yields in 2010 and 2011. It included costs of con tainers, selling, packing, harvesting and hauling, and organization fees.
214 Partial Budget Analysis of G rafted and Non G rafted Tomato P roduction Partial budget analysis is a standard economic analysis tool commonly used to determine the effects of a series o f changes to certain operations of the farming production system on the change of economic returns (Sydorovych et al., 2008) Th is eco nomic analysis approach compares the negative and positive effects of applying a new treatment relative to a base or standard treatment It can provide a good snapshot on the possible economic benefit that may accrue to growers adopting the new production practice. In the context of this study, using grafted tomato transplants is considered as the new treatment while using non grafted transplants is the standard practice. The typical components of the partial budget analysis as presented by Sydorovych et al (2008) were adapted as follows: 1. Negative effects: a) Added costs due to grafting b) Reduced returns due to grafting c) Total negative effects due to gr afting, i.e., the summation of the added and reduced returns 2. Positive effects: a) Reduced costs due to grafting b) Added returns due to grafting c) Total positive effects due to grafting, i.e., the summation of reduced costs and added returns 3. Total effects/net chan ge in revenue: calculated as the difference between total positive effects and total negative effects. It was assumed that added costs of grafted tomato production relative to the non grafted tomato production would be incurred if the costs of grafted tran splants and the harvest costs of grafted treatment were higher than the transplant and harvest costs of
215 non grafted treatment. On the other hand, reduced costs of grafted tomato production relative to the non grafted tomato production would be incurred if the costs of grafted transplants and the harvest costs of grafted treatment were lower than the transplant and harvest costs of non grafted treatment. Furthermore, added returns would be incurred if the grafted treatment resulted in higher yields as compar ed to the non grafted treatment, which could result in higher gross returns in grafted tomato production than non grafted tomato production. In contrast lower yields as a result of growing grafted plants would reduce returns and thereby lead to lower gross returns relative to the non grafted tomato production. Results and Discussion Costs of Tomato Transplant P roduction Grafting involves the combination of two different plants with distinctive traits to form a new plant. Hence, the use of grafted tomato tra nsplants is expected to result in additional costs inherent to labor and materials needed for grafting. In this study, t he cost associated with pared with the estimated cost of non grafted transplants at $0.15/plant (Table 9 1). On a per plant basis, our estimations show that the grafted tomato transplants were 467% more expensive than the non grafted plant. Such a price difference was consistent with that reported by Barrett et al. (2012b) Similarly, estimated costs in a commercial farming operation located in Ivanhow, North Carolina by Riv ard et al. (2010b) were $0.59 and $0.13 per grafted and non grafted tomato transplant respectively. In contrast, Besri and Rabat (2005) reported only a 2 fold increase of tomato transplant price with grafting in Morocco, i.e., $0.19 an d $0.38
216 per plant of non grafted and grafted tomato transplants, respectively. Costs of grafted vegetable transplants also vary with the crop type. According to Taylor et al. (2008) the p rice of grafted seedless watermelon transplants was $0.87/plant in contrast to $0.28/plant for non grafted transplants. As indicated in these previous studies, the increase of costs for grafted plants is often ascribed to the higher costs of rootstock seed s rather than the labor costs of grafting operation. In the present study, the overall materials needed for grafting accounted for about 73% of the total costs of grafted transplant production. With the development of more rootstock breeding programs in th e U.S., it is anticipated that the rootstock seed price may decrease and thus reducing the cost of grafted transplants. Advancement in grafting techniques including automation of the grafting operation may also help lower the cost of grafted transplant pro duction (Lee et al., 2010) With the plant density of 5808 tomato plants per acre, grafting with either 3, 015.63 per acre to the total pre 9 2). This was in line with the report of grafted transp lant costs in North Carolina which suggested that the use of grafted tomato transplants would increase the production costs by $2,275 per acre (O'connell et al., 2009) The increase of transplant costs as a result of using grafted plants further increased the portion of transplant costs in the total budget for field tomato production with the raised bed plasticulture system. For inst ance, t he cost of grafted transplants represented 2 6 % of the total production cost s as compared to only 6 % with the use of non grafted transplants in the 2010 field trial
217 Costs of Field Production of Fresh Market T omato In this study, the total costs per acre required to produce, harvest, and market bed polyethylene mulch system were estimated at $14,722.61 and $14,959.91 in the spring production seasons of 2010 and 2011, respectively (Table 9 3). These total estimated costs involved the variable and fixed costs associated with the pre harvest operations and the costs needed for harvest and marketing. Specifically, the pre harvest operational costs included the costs associated with transplants, ir rigation, fertilization, and other management practices, which made up about 66% of the total production costs. The harvest and postharvest operations including packaging and marketing amounted to $4,935.84/acre and $5,173.14/acre in 2010 and 2011, respect ively. Our estimated total production costs were slightly lower than the $15,451.77 developed by the University of Florida Center for Agribusiness. As the same template budget was used, this difference was partly due to the different estimated costs of tra nsplants and irrigation and nutrient management program. In addition, the harvest costs estimated in our study reflected the average marketable yields measured in the 2010 and 2011 field trials. The total estimated costs of field tomato in the present stud y were also comparable to the estimated cost of $16,926/acre for producing and harvesting tomatoes in soils with methyl bromide fumigation in North Carolina (Sydorovych et al., 2008) Marketable Fruit Y ield and Gross R eturns In this study, yield increases as a result of grafting with interspecific rootstocks were demonstrated in both the 2010 and 2011 field trials. In 2010, the highest m arketable yield expressed in 25 lb cartons per acre were 2023 and 1890 cartons with
218 opposed to 1456 cartons for the non grafted plants (Table 9 4). In 2011, these values were 2166, 2138, and 1526 cartons for FL/BE, FL/MU, and FL, respectively. As a result, projected total harvest costs were higher with the grafted treatments compared to the non grafted treatment during the two production seasons. For example, in 2010 the harvest cost in production of non 9 4). The tomato selling price s of $10.95 and $11.95 per 25 lb carton during the harves t period in 2010 and 2011 seasons respectively (U SDA 2011) were used in calculating the gross returns (crop value) of grafted and non grafted tomato production. In 2010, the gross returns reached 22,158.09/acre and $20,697.35/acre for production of grafted grafted tomato production was valued at $15,948.89/acre, which translated to additional gross returns relative to non returns were slightly higher in 2011 than in 2010 due to higher marketable fruit yields showed the gross returns of $19,972/acre and $10,776/acre in the spring seasons of 2005 and 2006, respectively (Gazula, 2009) It indicated that the gross returns of field tomato production can vary considerably depending on seasonal change of fruit yield and price levels during the marketing period.
219 Partial Budget Analysis of G rafted Tomato P roduction The negative and positive effects associated with the use of grafted tomato transplants relative to the non grafted plants were summarized in Table 9 5. In terms of the production costs, the negative effects (added costs) related to the use of grafted transplants involved the incr eased costs of transplants and the harvest. These total negative effects amounted to $4,937.24/acre and $4,485.17/acre $5,184.82/acre and $5,091.66/acre in 2011. On the other hand, the positive effects associated with using grafted transplants involved the added returns on the tomato fruit values, which ranged from $4,748.45/acre to $7,646.64/acre over the two spring production seasons (Table 9 5). The estimated gross re turns, total production costs, and net returns for grafted and non grafted treatments were presented in Table 9 6. Th e net return can be estimated as the difference between the gross returns and production costs The transplant and harvest costs changed as a result of using grafted transplants while other production costs remained constant regardless of the use of grafted transplants The net return for each grafted treatment relative to the non grafted treatment can also be calculated as the difference bet ween the total positive effects of grafting and the total negative effects of grafting. After accounting for grafting and harvest costs, the net returns of grafting relative to non grafting were $1,271.96/acre and $263.28/acre for grafted tomato productio net returns were $2,461.82/acre and $2,226.59/acre in 2011. This indicated that the increased crop value resulting from the significant improvement of marketable tomato frui t yield could offset the increased cost of grafted tomato production and even made it
220 more profitable than non grafted tomato production in fumigated soils. In this study, grafted tomato production as observed in the present study was lower than that re C onnell et al. (2009) in an organic, open field production of indeterminate tomato in North Carolina. These authors found that the ne t return of the top yielding scion 04 105 $59,635/acre as compared to $8,780/acre for non be noted that such a great difference in net returns was due to the high pressure of Southern blight ( caused by Sclerotium rolfsii ) The prevalence of this particular soil borne disease resulted in more than 80% of dead plants in non grafted plots within the first two weeks of fruit production, whereas the use of re sistant rootstock extended fruit production for seven weeks. Similarly, recent studies on cost benefit analyses of grafted watermelon (Taylor et al., 2008) and heirloom tomato produ ction (Barrett et al., 2012b) indicated that using grafted transplants is more cost effective than the non grafted treatment only in the field with high pressure of soil borne disease and root knot nematodes in the abse nce of soil fumigants. In our study, the comparison of grafted and non grafted tomato production was conducted in soils fumigated with Telone C35 which allowed for evaluation of the grafting benefit for yield improvement beyond disease control. In order t o provide a more complete analysis of economic feasibility for field production of grafted tomato, production systems with and without soil fumigation can be included in future studies
221 which will help elucidate the economic returns associated with effectiv eness of disease control vs. improvement of crop vigor and resulting yield enhancement. Furthermore, with the improved irrigation and nitrogen use efficiency demonstrated by grafted plants as compared to non grafted plants, future studies should also evalu ate the extent of cost savings that may be generated in term of expenditures relative to nutrient and irrigation management programs by using vigorous rootstocks compared to non grafted plants. Moreover, in a study involving two different rootstocks of mel on, it has also been shown that the total yields of grafted plants grown at different plant density (50, 60, 80% of the conventional plant density used by the growers) were significantly higher than the yield obtained from non grafted plants indicating tha t with the use of grafted melon plants, planting density may be reduced by 60% (Ricrdez Salinas et al., 2010) Similar studies should also be conducted with rootstocks availa ble for tomato grafting to find out whether reduced plant density can be used in the goal to reduce costs inherent to grafted transplants production. In this study, the tomato plants were grafted above the rootstock cotyledons. When the graft union is form ed above the rootstock cotyledons (cotyledons are left intact), regrowth of the rootstock, i.e., rootstock suckers, may be formed from the axillary meristems. Additional labor is often required to remove the rootstock suckers on the grafted tomato plants p rior to transplanting or during the first few weeks after transplanting (Bausher, 2011) Removal of rootstock suckers can add more cost to grafted tomato production. This extra cost was not considered in the present study as rootstock suckering can be avoided by grafting below the rootstock cotyledons (Bausher, 2011)
222 Conclusion s Growing grafted tomato demonstrated higher production costs due to the increased costs associated with grafted transplants and harvest. In the field production with fumigated soils, the yield improvement as a result of using grafted transplants led to higher returns that eventually increased the net return at as compared with the use of non grafted tra nsplants. More on farm trials in commercial growing conditions across various tomato production areas in Florida are still needed to help growers decide whether grafting can be integrated as an economic viable component into their exiting production system s. A complete set of scenarios for economic analysis needs to be considered to fully assess the costs and profits associated with using grafted plants for disease management, enhancement of irrigation water and fertilizer use efficiency, and yield improvem ent. Furthermore, the cost effectiveness of grafted vegetable production is expected to increase as future development of vegetable grafting research leads to reduced costs of rootstock seeds and grafted transplants.
223 Table 9 1 Estimated costs of prod uction of grafted and non grafted tomato transplants Item Grafted Nongrafted Labor y Material Labor y Material ($/1000 plants) x ($/1000 plants) x Seeds 83.26 74.94 254.41 Seedling production Potting soil 19.00 9.50 Flats w 16.72 7.31 Seed sowing and care 92.95 59.15 Grafted transplant production Grafting v 46.94 Silicon clips 46.66 Miscellaneous supplies 8.00 Post graft care 21.13 Healing chamber u Humidifier 5.49 Air conditioner 31.80 Building supplies 26.86 Assembly 16.90 Subtotal 177.92 492.20 59.15 91.75 Total 670.12 150.90 Cost/plant 0.67 0.15 z A dapted from Barrett et al., (2012b) y $8.45/h pay wage for all labor. x Estimate of costs based on Spring 2012 prices for a target 1000 grafted transplants. w 128 cell count transplant flats, straight line depreciated for 5 years estimated use. v 200 plants/grafter per hour graft rate. u Straight line depreciated for 5 years estimated use.
224 Table 9 2 Estimated costs per acre needed to produce transplants of non grafted (FL/MU). T reatment Labor costs Materials costs Total transplant costs Added costs relative to non grafted plants ($/acre) FL 343.54 532.88 876.42 0.00 FL/BE 1 033.3 6 2 858.69 3 892.05 3 015.6 3 FL/MU 1 033.3 6 2 858. 69 3 892.05 3 015.6 3
225 Table 9 3 Estimated costs z bed polyethylene mulch system in the spring seasons of 2010 and 2011 in Live Oak, FL. Operation 2010 2011 2010 2011 ( $/acre) ($/25 lb Carton) Fertilizer and Irrigation Costs Fertilizer 617.61 617.61 Irrigation Tubing 107.50 107.50 Pumping Costs 8.59 8.59 Labor Cost 12.77 12.77 Total Fertilizer and Irrigation Costs 746.47 746.47 Oper ating Costs Transplants 876.43 876.43 Fumigant 675.00 675.00 Fungicide 264.45 264.45 Herbicide 45.35 45.35 Insecticide 520.17 520.17 General Farm Labor 186.62 186.62 Machenery Variable Cost 1,095.28 1,095. 28 Tractor Driver Labor 373.77 373.77 Miscellaneous Costs Tie Plants 145.20 145.20 Scouting 35.00 35.00 Plastic Mulch 330.00 330.00 Stakes 112.00 112.00 Plastic String 100.00 100.00 String, Stake a nd Mulch Disposal 342.80 342.80 Farm Vehicles 33.34 33.34 Interest on Operating Capital 296.21 296.21 Total Operating Costs 6,178.09 6,178.09 Fixed Costs Land Rent 300.00 300.00 Machinery Fixed Cost 235.94 235.94 Farm Management and Overhead 3,072.74 3,072.74 Total Fixed Cost 3,608.68 3,608.68 TOTAL PREHARVEST COSTS 9,786.77 9,786.77 6.72 6.41 Harvest and Marketing Costs Tomato Carton 1,092.00 1,144.50 0.75 0.75 Sell 218.40 228.90 0.15 0.15 Pack 2,402.40 2,517.90 1.65 1.65 Harvest and Haul 1,092.00 1,144.50 0.75 0.75 Organization Fees 131.04 137.34 0.09 0.09 Total Harvest and Marketing Cost 4,935.84 5,173.14 3.39 3.39 Total Cost 14,722.61 14,959.91 10.11 9.80
226 z Cost s of fer tilizer, irrigation, and other preharvest operations for a 1 acre tomato field were calculated based on information provided by suppliers, Pitts et al., (2002) and estimated production costs in the Manatee/Ruskin area, 200 8 200 9 University of Florida Center for Agribusiness (Agr ibusiness)
227 Table 9 4 Average marketable tomato fruit yields, harvest related costs, gross returns for non and gross returns relative to non grafted treatment during the spring seaso ns of 2010 and 2011 with raised bed polyethylene mulch system in Live Oak, FL. Treatment z Average m arketable fruit yield (25 lb cartons/acre) Total estimated harvest costs Harvest cost relative to non grafting Gross returns Gross returns relative to non gr afting ($/acre) 2010 FL 1,457 4,935.84 0.00 15,948.89 0.00 FL/BE 2,024 6,857.45 1,921.61 22,158.09 6,209.20 FL/MU 1,890 6,405.38 1,469.54 20,697.35 4,748.45 2011 FL 1,526 5,173.14 0.00 18,235.93 0.00 FL/BE 2,166 7,342.33 2,169.19 25,882. 56 7,646.64 FL/MU 2,138 7,249.17 2,076.03 25,554.18 7,318.25 z FL: non
228 Table 9 5 Added costs and reduced returns, total negative effec ts, reduced costs and added returns, total positive effects, and additional net returns incurred by of fresh market tomato during the spring seasons of 2010 and 2011 with r aised bed polyethylene mulch system in Live Oak, FL. Treatment z Added cost of grafting Reduced returns of grafting Total negative effects of grafting Reduced costs of grafting Added returns of grafting Total positive effects of grafting Additional net retu rns of grafting relative to nongrafting ($/acre) 2010 FL 0.00 0.00 0.00 0.00 0.00 0.00 0.00 FL/BE 4,937.24 0.00 4,937.24 0.00 6209.20 6209.20 1271.96 FL/MU 4,485.17 0.00 4,485.17 0.00 4748.45 4748.45 263.28 2011 FL 0.00 0.00 0. 00 0.00 0.00 0.00 0.00 FL/BE 5184.82 0.00 5184.82 0.00 7646.64 7646.64 2461.82 FL/MU 5091.66 0.00 5091.66 0.00 7318.25 7318.25 2226.59 z FL: non
229 Table 9 6 Comparisons of estimated gross returns, costs of transplants, harvest, and other production operations and marketing, and total net returns between field production of grafted and non Treatment Gro ss returns Transplant costs Harvest costs Other production and marketing costs Total net returns Additional net returns relative to nongrafting ($/acre) 2010 FL 15,948.89 876.43 4,935.84 8,910.34 1,226.29 0.00 FL/BE 22,158.09 3,892.06 6,857. 45 8,910.34 2,498.24 1,271.96 FL/MU 20,697.35 3,892.06 6,405.38 8,910.34 1,489.57 263.28 2011 FL 18,235.93 876.43 5,173.14 8,910.34 3,276.02 0.00 FL/BE 25,882.56 3,892.06 7,342.33 8,910.34 5,737.84 2,461.82 FL/MU 25,554.18 3,892.06 7,24 9.17 8,910.34 5,502.61 2,226.59 FL: non
230 CHAPTER 10 GENERAL CONCLUSIONS Review and Synthesis of F indings In addition to the cultural pra ctices currently used in the BMP program, sustainable practices are still needed in order to further enhance fruit yield while reducing the potential for N leaching in the tomato production systems in Florida especially in sandy soils. Using vigorous inter specific tomato hybrid rootstocks, this research project was performed to determine the potential of grafting as an economically viable practice to optimizing water and N management in tomato production in sandy soils in Florida. A preliminary study was c onducted with the determinate 04 Overall, the use of rootst ocks resulted in an increase of total marketable fruit yield by up to 66 % as compared with non grafted scion plants (Chapter 3). The improved total number of fruit per plant, while higher average fruit weight was the main contributing 04 yield as the non grafted scion control. In addition to yield improvement, enhanced uptake of N, P, K, and Ca was also observed in grafted plants. However, the enhancement in nutrient accumulation was largely related to increased biomass accumulation rather than higher nutrient concentration (on a dry weight basis). A follow up 2 fertilization rates in field conditions demonstrated that the use of interspecific rootstocks
231 in field production of toma to in sandy soils significantly improved fruit yields as compared to the non grafted plants (Chapter 5). Grafting with the two interspecific tomato hybrid rootstocks also led to significant enhancement in irrigation water and nitrogen use efficiency in con trast to the non grafted plants. Furthermore, plant g rowth, nitrogen use and nitrogen uptake efficiency, and root length density were also improved by grafting Our estimation of the C rop Nutritional Requirement for N for grafted tomato plants demonstrated that estimated minimum N rates needed to maximize the marketable fruit yields were consistently higher with grafted plants as compared to the non grafted plants, irrespective of the i rrigation regimes (Chapter 7). These results indicated that enhancement in vigor and fruit yield is to some extent at the expense of greater nutrient demands especially nitrogen. With the higher sink strength (higher number of fruit per plant and average f ruit weight) in grafted plants than that in the non grafted plants, it is expected that the production of the source of assimilates will increase to maintain the higher level of the sink strength. We have also attempted to examine the economic feasibility of using grafted tomato transplants in the current production system of fresh market tomato in sandy soils in Florida. Based on the results of the partial budget analysis, it appeared that although grafting increased the total cost to produce, harvest, an d market tomato fruit, the resulting increase in marketable fruit generated significant gross return to offset costs associated with the use of grafted transplants and increased profitability as compared to non grafted transplants (Chapter 9).
232 In another greenhouse study aiming at exploring the underlying physiological basis for enhanced plant vigor and fruit yield t he interspecific rootstocks showed no consistent effect s on nitrogen assimilation measurements including l evels of nitrate reductase activity nitrate, organic nitrogen, amino acids and soluble proteins as compared to those of the non and self grafted plants. Similar ly, no clear pattern s in the levels of endogenous hormones including auxin, cytokinin s and gibberell i c acids w ere also observed (Chapter 4) With respect to fruit quality attributes such as soluble solids content, titratable acid, and levels of vitamin C, carotenoids, and phenolics, no major differences were observed between grafted and non grafted plant s, although grafting with the interspecific rootstocks increased the fruit moisture content. Meanwhile, seasonal variations of the fruit quality attributes appeared to be much greater than the rootstock effects (Chapters 3 and 8). Results from this proje ct demonstrated the great potential for using vigorous interspecific hybrid tomato rootstocks as a cost effective tool to improve plant growth and fruit yield and to enhance N use efficiency in tomato production in sandy soils with no consistently negative influence on the fruit quality attributes. Future W ork More in depth studies are warranted to include more tomato scion cultivars in the field experiments across different locations to further examine the rootstock scion interaction effects on plant grow th and fruit yield and develop the recommendations for N fertilization and water management in grafted tomato production in sandy soils. In addition, future research can be conducted to monitor the movement of nutrients, especially nitrate in the soil prof ile in order to substantiate the impact of grafting with vigorous rootstocks on reducing nitrate leaching out of the plant active root zone.
233 Moreover further research is needed to better elucidate how different rootstocks may modify the seasonal variation in nitrate assimilation and to evaluate potential rootstock differences in affecting xylem sap flow and changes of growth hormone status in grafted plants. Rootstock evaluations in the future also need to differentiate the influence of the grafting proces s per se, if any, from the rootstock effects on the physiological modifications in grafted plants.
234 LIST OF REFERENCES Abbott, J.A. 1999. Quality measurement of fruits and vegetables. Postharvest Biol. Technol. 15: 207 225. Abdelmage ed, A.H.A., N. Gruda, and B. Geyer. 2004. Effects of temperature and grafting on the growth and development of tomato plants under controlled conditions, p. 5, Deutscher Tropentag, Berlin, Germany. Adams, P., 2002. Nutritional control in hydroponics, p. 21 1 261. In: D., S. and HC, P. (eds.), Hydroponic production of vegetables and ornamentals. Embryo Publications Athens, Athens, Greece. gren, G.I. and O. Franklin. 2003. Root: Shoot ratios, optimization and nitrogen productivity. Ann. Bot. 92: 795 800. Agri business, U.o.F.C.f. 2009. Spring tomatoes: Estimated production costs in the manatee/ruskin area. 04/18/2012. http://www.fred.ifas.ufl.edu/iatpc/files/RuskinSpringTomato09.p df Ahmad, R., M. Naveed, M. Aslam, Z.A. Zahir, M. Arshad, and G. Jilani. 2008. Economizing the use of nitrogen fertilizer in wheat production through enriched compost. Renew. Agric. Food Syst. 23: 243 249. Ahn, S., Y. Im, G. Chung, B. Cho, and S. Suh. 19 99. Physiological responses of grafted cucumber leaves and rootstock roots affected by low root temperature. Sci. Hortic. 81: 397 408. Proceedings of the International S ociety of Citriculture 1:62 65. Aloni, B., R. Cohen, L. Karni, H. Aktas, and M. Edelstein. 2010. Hormonal signaling in rootstock scion interactions. Sci. Hortic. 127: 119 126. Aloni, R., E. Aloni, M. Langhans, and C. Ullrich. 2006. Role of cytokinin and au xin in shaping root architecture: Regulating vascular differentiation, lateral root initiation, root apical dominance and root gravitropism. Ann. Bot. 97: 883. Anderson, R.L. and L.A. Nelson. 1975. A family of models involving intersecting straight lines a nd concomitant experimental designs useful in evaluating response to fertilizer nutrients. Biometrics 31: 303 318. Andrews, P. and C. Marquez. 1993. Graft incompatibility. Hortic. Rev. 15: 183 232. AOAC International. 2000. Official methods of analysis of aoac international. In: Horowitz, W. (ed.), The AOAC International, Maryland, USA
235 Aslam, M.T. and D. RL Rains. 2001. Diurnal fluctuations of nitrate uptake and in vivo nitrate reductase activity in pima and acala cotton. Crop Sci. 41: 372. Babcock, B.A. 19 92. The effects of uncertainty on optimal nitrogen applications. Rev. Agric. Econ. 14: 271 280. Bachmaier, M. and M. Gandorfer. 2012. Estimating uncertainty of economically optimum n fertilizer rates. Int. J. Agron. 2012: 1 10. Baggs, E., C. Watson, and R. Rees. 2000. The fate of nitrogen from incorporated cover crop and green manure residues. Nutr. Cycl. Agroecosys. 56: 153 163. Bai, Y. and P. Lindhout. 2007. Domestication and breeding of tomatoes: What have we gained and what can we gain in the future? An n. Bot. 100: 1085 1094. Banerjee, A., M. Chatterjee, Y. Yu, S. Suh, W. Miller, and D. Hannapel. 2006. Dynamics of a mobile rna of potato involved in a long distance signaling pathway. The Plant Cell Online 18: 3443 3457. Barneix, A. and H. Causin. 1996. Th e central role of amino acids on nitrogen utilization and plant growth. J. Plant Physiol. 149: 358 362. Barrett, C.E., X. Zhao, and R. McSorley. 2012a. Grafting for root knot nematode control and yield improvement in organic heirloom tomato production. Hor tScience 47: 614 620. Barrett, C.E., X. Zhao, and A.W. Hodges. 2012b. Cost benefit analysis of using grafted transplants for root knot nematode management in organic heirloom tomato production. HortTechnology 22: 252 257. Barrett, D.M., C. Weakley, J.V. Di az, and M. Watnik. 2007. Qualitative and nutritional differences in processing tomatoes grown under commercial organic and conventional production systems. J. Food Sci. 72: C441 C451. Bausher, M. 2009. Commercial tomato rootstock performance when exposed t o natural populations of root knot nematodes in florida. HortScience 44: 1021 (Abstr.). Bausher, M.G. 2011. Grafting technique to eliminate rootstock suckering of grafted tomatoes. HortScience 46: 596 598. Benincasa, P., M. Guiducci, and F. Tei. 2011. The nitrogen use efficiency: Meaning and sources of variation case studies on three vegetable crops in central italy. HortTechnology 21: 266 273. Bertin, N., H. Gautier, and C. Roche. 2002. Number of cells in tomato fruit depending on fruit position and source sink balance during plant development. Plant Growth Reg ul. 36: 105 112.
236 Besri, M. and M. Rabat. 2005. Current situation of tomato grafting as alternative to methyl bromide for tomato production in morocco. San Diego, CA Proceeding of the 12 th annual inter national Res. conference on methyl bromide alternatives and emission reductions 47:1 3. Black, B.L., L.H. Fuchigami, and G.D. Coleman. 2002. Partitioning of nitrate assimilation among leaves, stems and roots of poplar. Tree Physiol. 22: 717 724. Black, L., D. Wu, J. Wang, T. Kalb, D. Abbass, and J. Chen. 2003. Grafting tomatoes for production in the hot wet season. Asian Vegetable Research Development Center Publication: 1 8. Bletsos, F., C. Thanassoulopoulos, and D. Roupakias. 2003. Effect of grafting on g rowth, yield, and verticillium wilt of eggplant. HortScience 38: 183 186. Bloch, D., C.M. Hoffmann, and B. Marlander. 2006. Impact of water supply on photosynthesis, water use and carbon isotope discrimination of sugar beet genotypes. Eur. J. Agron. 24: 21 8 225. Boselli, M., M. Fregoni, A. Vercesi, and B. Volpe. 1992. Variation in mineral composition and effects on the growth and yield of 'chardonnay' grapes on various rootstocks. Agricoltura Ricerca 14: 138 139. Bota, J., J. Flexas, and H. Medrano. 2001. G enetic variability of photosynthesis and water use in balearic grapevine cultivars. Ann. Appl. Biol. 138: 353 361. Bradford, M.M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Anal. Biochem. 72: 248 254. Bramley, P.M. 2002. Regulation of carotenoid formation during tomato fruit ripening and development. J. Exp. Bot. 53: 2107 2113. Bulder, H., A. Den Nijs, E. Speek, P. Van Hasselt, and P. Kuiper. 1991. The effect of low root temperature on growth and lipid composition of low temperature tolerant rootstock genotypes for cucumber. J. Plant Physiol. 138: 661 666. Bullock, D.S. and D.G. Bullock. 1994. Quadratic and quadratic plus plateau models for predicting optimal nitr ogen rate of corn: A comparison. Agron. J. 86: 191 195. Burri, B.J., M.H. Chapman, T.R. Neidlinger, J.S. Seo, and B.K. Ishida. 2009. Tangerine tomatoes increase total and tetra cis lycopene isomer concentrations more than red tomatoes in healthy adult huma ns. Int. J. Food Sci. Nutr. 60: 1 16. Cabello, M., M. Castellanos, F. Romojaro, C. Martnez Madrid, and F. Ribas. 2009. Yield and quality of melon grown under different irrigation and nitrogen rates. Agric. Water Manage. 96: 866 874.
237 Campbell, W.H. 1996. N itrate reductase biochemistry comes of age. Plant Physiol. 111: 355. Cantliffe, D., P. Gilreath, D. Haman, C. Hutchinson, Y. Li, G. McAvoy, K. Migliaccio, T. Olczyk, S. Olson, and D. Parmenter. 2006. Review of nutrient management systems for florida vegeta ble producers: A white paper from the uf/ifas vegetable fertilizer task force. Proc. Fla. State. Hort. Soc. 119:240 248. Castle, W. 1995. Rootstock as a fruit quality factor in citrus and deciduous tree crops. N. Z. J. Crop Hortic. Sci. 23: 383 394. Castle W., D.P.H. Tucker, A. Krezdorn, and C.O. Youtsey. 1993. Rootstocks for florida citrus: Rootstock selection, the first step to success. University of Florida, Institute of Food and Agricultural Sciences, Gainesville. Cerrato, M. and A. Blackmer. 1990. Com parison of models for describing corn yield response to nitrogen fertilizer. Agron. J. 82: 138 143. Chaplin, M. and M. Westwood. 1980. Effects of pyrus species and related genera 346. Co lla, G., C. Cardona Suarez, and M. Cardarelli. 2010. Improving nitrogen use efficiency in melon by grafting. HortScience 45: 559 565. Colla, G., Y. Roupahel, M. Cardarelli, and E. Rea. 2006a. Effect of salinity on yield, fruit quality, leaf gas exchange, a nd mineral composition of grafted watermelon plants. HortScience 41: 622 627. Colla, G., Y. Rouphael, C. Mirabelli, and M. Cardarelli. 2011. Nitrogen use efficiency traits of mini watermelon in response to grafting and nitrogen fertilization doses. J. Plan t Nutr. Soil Sci. 174: 933 941. Colla, G., Y. Rouphael, M. Cardarelli, D. Massa, A. Salerno, and E. Rea. 2006b. Yield, fruit quality and mineral composition of grafted melon plants grown under saline conditions. J. Horticult. Sci. Biotechnol. 81: 146 152. Colla, G., Y. Rouphael, M. Cardarelli, O. Temperini, E. Rea, A. Salerno, and F. Pierandrei. 2008. Influence of grafting on yield and fruit quality of pepper ( capsicum annuum l.) grown under greenhouse conditions. Acta Hortic. 782: 359 364. Cook, R.D. and S Weisberg. 1990. Confidence curves in nonlinear regression. Journal of the American Statistical Association 85: 544 551. Cuartero, J. and R. Fernndez Muoz. 1999. Tomato and salinity. Sci. Hortic. 78: 83 126.
238 Davis, A., P. Perkins Veazie, R. Hassell, A. Levi, S. King, and X. Zhang. 2008a. Grafting effects on vegetable quality. HortScience 43: 1670 1672. Davis, A., P. Perkins Veazie, Y. Sakata, S. Lopez Galarza, J. Maroto, S. Lee, Y. Huh, Z. Sun, A. Miguel, and S. King. 2008b. Cucurbit grafting. Crit. Rev. Plant Sci. 27: 50 74. De Koning, A. 1993. Growth of a tomato crop. Measurements for model validation. Acta Hortic.: 141 141. del Amor, F., J. Lopez Marin, and A. Gonzalez. 2008. Effect of photoselective sheet and grafting technique on growth, yield, and m ineral composition of sweet pepper plants. J. Plant Nutr. 31: 1108 1120. Di Gioia, F., F. Serio, D. Buttaro, O. Ayala, and P. Santamaria. 2010. Influence of rootstock on vegetable growth, fruit yield and quality in 'cuore di bue', an heirloom tomato. J. Ho rticult. Sci. Biotechnol. 85: 477 482. Di Vaio, C., C. Cirillo, M. Buccheri, and F. Limongelli. 2009. Effect of interstock (m. 9 and m. 27) on vegetative growth and yield of apple trees (cv. Annurca). Sci. Hortic. 119: 270 274. Ding, B., A. Itaya, and Y. Q i. 2003. Symplasmic protein and rna traffic: Regulatory points and regulatory factors. Curr. Opin. Plant Biol. 6: 596 602. Dodd, I.C. and C.A. Beveridge. 2006. Xylem borne cytokinins: Still in search of a role? J. Exp. Bot. 57: 1 4. Dong, H., Y. Niu, W. Li and D. Zhang. 2008. Effects of cotton rootstock on endogenous cytokinins and abscisic acid in xylem sap and leaves in relation to leaf senescence. J. Exp. Bot. 59: 1295 1304. Dorais, M., D. Ehret, and A. Papadopoulos. 2008. Tomato ( solanum lycopersicum ) health components: From the seed to the consumer. Phytochem. Rev. 7: 231 250. Drew, M. 1997. Oxygen deficiency and root metabolism: Injury and acclimation under hypoxia and anoxia. Annu. Rev. Plant Biol. 48: 223 250. Drogoudi, P. and C. Tsipouridis. 2007. Effects of cultivar and rootstock on the antioxidant content and physical characters of clingstone peaches. Sci. Hortic. 115: 34 39. Edelstein, M. 2004. Grafting vegetable crop plants: Pros and cons. Acta Hort. 659: 235 238. Edelstein, M., M. Ben Hur, and Z. Plaut. 2006. Response of grafted and non grafted melon plants to effluent and boron under greenhouse conditions.325 328.
239 Edelstein, M., R. Cohen, Y. Burger, S. Shriber, S. Pivonia, and D. Shtienberg. 1999. Integrated management of sudden wilt in melons, caused by monosporascus cannonbolus using grafting and reduced rates of methyl bromide Plant Dis. 83: 1142 1145. Efron, B. and R. Tibshirani. 1986. Bootstrap methods for standard errors, confidence intervals, and other measures of statistical accuracy. S tatist. Sci. 1: 54 75. Eissenstat, D.M. 1991. On the relationship between specific root length and the rate of root proliferation: A field study using citrus rootstocks. New Phytol. 118: 63 68. Elia, A. and G. Conversa. 2012. Agronomic and physiological re sponses of a tomato crop to nitrogen input. Eur. J. Agron. 40: 64 74. Estan, M.T., M.M. Martinez Rodriguez, F. Perez Alfocea, T.J. Flowers, and M.C. Bolarin. 2005. Grafting raises the salt tolerance of tomato through limiting the transport of sodium and ch loride to the shoot. J. Exp. Bot. 56: 703 712. Ezzahouani, A. and L. Williams. 1995. The influence of rootstock on leaf water potential, yield, and berry composition of ruby seedless grapevines. Am. J. Enol. Vitic. 46: 559 563. Fageria, N. 2009. The use of nutrients in crop plants. CRC Press, Boca Raton, FL430. Fanasca, S., G. Colla, Y. Rouphael, F. Saccardo, G. Maiani, E. Venneria, and E. Azzini. 2006. Evolution of nutritional value of two tomato genotypes grown in soilless culture as affected by macrocati on proportions. HortScience 41: 1584 1588. FDACS. 2005. Water quality/quantity best management practices for florida vegetable and agronomic crops, p. 171. In: Services, F. D. o. A. a. C. (ed.). FDACS, Tallahassee. Fernandez Garcia, N., M. Carvajal, and E. Olmos. 2004a. Graft union formation in tomato plants: Peroxidase and catalase involvement. Ann. Bot. 93: 53 60. Fernandez Garcia, N., V. Martinez, A. Cerda, and M. Carvajal. 2004b. Fruit quality of grafted tomato plants grown under saline conditions. J. H orticult. Sci. Biotechnol. 79: 995 1001. Fernndez Garca, N., V. Martnez, A. Cerd, and M. Carvajal. 2002. Water and nutrient uptake of grafted tomato plants grown under saline conditions. J. Plant Physiol. 159: 899 905. Ferree, D., G. Cahoon, M. Ellis, D. Scurlock, and G. Johns. 1996. Influence of eight rootstocks on the performance of 'white riesling' and 'cabernet franc' over five years. J. Am. Pomol. Soc. 50: 124 130.
240 Flores, F., P. Sanchez Bel, M. Esta, M. Martinez Rodriguez, E. Moyano, B. Morales, J. Campos, J. Garcia Abelln, M. Egea, and N. Fernndez Garcia. 2010. The effectiveness of grafting to improve tomato fruit quality. Sci. Hortic. 125: 211 217. Florida Automated Weather Network, F. 2011. Archived weather data for Live O ak, FL .7 15 20011. Foolad, M.R. 2007. Genome mapping and molecular breeding of tomato. Int. J. Plant Genom. 2007: 1 52. Fox, J. 2008. Applied regression analysis and generalized linear models. SAGE Publications, Inc, Los Angeles, California665. Gao, Q.H., K. Xu, and X.F. Wa ng. 2008. Effect of grafting on cold tolerance in eggplant seedlings. Acta Hortic. 771: 167 174. Garca Snchez, F., J. Jifon, M. Carvajal, and J. Syvertsen. 2002. Gas exchange, chlorophyll and nutrient contents in relation to na + and cl accumulation in sunburst' mandarin grafted on different rootstocks. Plant Sci. 162: 705 712. Gazula, A. 2009. Impact of irrigation and nutrient management programs on fruit yields, nitrogen load, and crop value of fresh market tomato grown with plasticulture in the era of best management practices. University of Florida, Gainesville, Ph.D. Georgelis, N., J. Scott, and E. Baldwin. 2004. Relationship of tomato fruit sugar concentration with physical and chemical traits and linkage of rapd markers. J. Am. Soc. Hort. Sci. 129: 839 845. Germana, C. 1996. The response of pistachio trees to water stress as affected by two different rootstocks. Acta Horti. 449:513 520. Gibberd, M., R. Walker, D. Blackmore, and A. Condon. 2001. Transpiration efficiency and carbon isotope discriminat ion of grapevines grown under well watered conditions in either glasshouse or vineyard. Aust. J. Grape Wine Res. 7: 110 117. Gil Izquierdo, A., M. Riquelme, I. Porras, and F. Ferreres. 2004. Effect of the rootstock and interstock grafted in lemon tree ( cit rus limon (l.) burm.) on the flavonoid content of lemon juice. J. Agric. Food Chem. 52: 324 331. Giorgi, M., F. Capocasa, J. Scalzo, G. Murri, M. Battino, and B. Mezzetti. 2005. The rootstock effects on plant adaptability, production, fruit quality, and nu trition in the peach (cv. 'suncrest'). Sci. Hortic. 107: 36 42. Giovannucci, E. 2002. A review of epidemiologic studies of tomatoes, lycopene, and prostate cancer. Exp. Biol. Med. 227: 852 859.
241 Glass, A., D. Britto, B. Kaiser, J. Kinghorn, H. Kronzucker, A Kumar, M. Okamoto, S. Rawat, M. Siddiqi, and S. Unkles. 2002. The regulation of nitrate and ammonium transport systems in plants. J. Exp. Bot. 53: 855. Golecki, B., A. Schulz, and G. Thompson. 1999. Translocation of structural p proteins in the phloem. T he Plant Cell Online 11: 127 140. Gonalves, B., J. Moutinho Pereira, A. Santos, A. Silva, E. Bacelar, C. Correia, and E. Rosa. 2006. Scion rootstock interaction affects the physiology and fruit quality of sweet cherry. Tree Physiol. 26: 93 104. Hale, C. a nd C. Brien. 1978. Influence of salt creek rootstock on composition and quality of shiraz grapes and wine. Vitis 17: 139 146. Hallberg, G.R., 1989. Nitrate in ground water in the united states, p. 35 74. In: Follett, R. (ed.), Nitrogen management and groun d water protection. Elsevier Science Publishers, New York. Harada, T. 2010. Grafting and rna transport via phloem tissue in horticultural plants. Sci. Hortic. 125: 545 550. Harris, P., 1992. Mineral nutrition, p. 162 213. In: Harris, P. (ed.), The potato c rop. Chapman and Hall Ltd, London, London, UK. Hartmann, H.T., D.E. Kester, F.T. Davies, and R.L. Geneve. 2002. Hartmann and kester's plant propagation: Principles and practices. Prentice Hall, Upper Saddle River, NJ849. Hartz, T. and G. Hochmuth. 1996. Fe rtility management of drip irrigated vegetables. HortTechnology 6: 168 171. Hartz, T.K. and T.G. Bottoms. 2009. Nitrogen requirements of drip irrigated processing tomatoes. HortScience 44: 1988 1993. Hebbar, S., B. Ramachandrappa, H. Nanjappa, and M. Prabh akar. 2004. Studies on npk drip fertigation in field grown tomato ( L ycopersicon esculentum mill.). Eur. J. Agron. 21: 117 127. Hernandez, J.A. and D.J. Mulla. 2008. Estimating uncertainty of economically optimum fertilizer rates. Agron. j 100: 1221 1229. H irose, N., K. Takei, T. Kuroha, T. Kamada Nobusada, H. Hayashi, and H. Sakakibara. 2008. Regulation of cytokinin biosynthesis, compartmentalization and translocation. J. Exp. Bot. 59: 75 88. Hochmuth, G. 1992. Fertilizer management for drip irrigated veget ables in florida. HortTechnology 2: 27 32.
242 Hochmuth, G. and E. Hanlon. 2011. A summary of n, p, and k research with tomato in florida, p. 38, Vegetable Nutrition Management Series. Soil and Water Science Department, Florida Cooperative Extension Service, I nstitute of Food and Agricultural Sciences, University of Florida, Gainesville FL. Hochmuth, G.J. and R.C. Hochmuth. 2001. Nutrient solution formulation for hydroponic (perlite, rockwool, nft) tomatoes in florida, p. 1 11, HS796. Univ. Fla. Coop. Ext. Serv Horticultural Sciences Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida, Gainesville. Holbrook, N., V. Shashidhar, R. James, and R. Munns. 2002. Stomatal control in tomato with aba defi cient roots: Response of grafted plants to soil drying. J. Exp. Bot. 53: 1503 1514. Horwitz, W. 2000. Official methods of analysis of aoac international. AOAC international Gaithersburg, MD. Hougaard, P. 1985. The appropriateness of the asymptotic distribu tion in a nonlinear regression model in relation to curvature. Journal of the Royal Statistical Society. Series B (Methodological): 103 114. Huang, D., B. Ou, M. Hampsch Woodill, J. Flanagan, and R. Prior. 2002. High throughput assay of oxygen radical abso rbance capacity (orac) using a multichannel liquid handling system coupled with a microplate fluorescence reader in 96 well format. J. Agric. Food Chem. 50: 4437 4444. Huang, Y., R. Tang, Q. Cao, and Z. Bie. 2009. Improving the fruit yield and quality of c ucumber by grafting onto the salt tolerant rootstock under nacl stress. Sci. Hortic. 122: 26 31. Iacono, F., A. Buccella, and E. Peterlunger. 1998. Water stress and rootstock influence on leaf gas exchange of grafted and ungrafted grapevines. Sci. Hortic. 75: 27 39. Ibrahim, M., M. Munira, M. Kabir, A. Islam, and M. Miah. 2001. Seed germination and graft compatibility of wild solanum as rootstock of tomato. J. Biol. Sci. 1: 701 703. Imazu, T. 1949. On the symbiotic affinity caused by grafting among cucurbit aceous species. J Jpn. Soc. Hortic. Sci. 18: 6 42. Iouannou, N. 2001. Integrating soil solarization with grafting on resistant rootstocks for management of soil borne pathogens of eggplant. J. Horticult. Sci. Biotechnol. 76: 396 401. Iyengar, B., C. Iyer, and V. Sulladamath. 1982. Influence of rootstocks on the leaf nutrient composition of two scion cultivars of mandarin. Sci. Hortic. 16: 163 169. Jackson, D. and P. Lombard. 1993. Environmental and management practices affecting grape composition and wine q uality a review. Am. J. Enol. Vitic. 44: 409 430.
243 Jackson, L. and A. Bloom. 1990. Root distribution in relation to soil nitrogen availability in field grown tomatoes. Plant Soil 128: 115 126. Jamil Mohammad, M. 2004. Utilization of applied fertilizer nitro gen and irrigation water by drip fertigated squash as determined by nuclear and traditional techniques. Nutr. Cycl. Agroecosys. 68: 1 11. Janick, J. 1986. Horticultural science. WH Freeman, New York746. Jaworski, E. 1971. Nitrate reductase assay in intact plant tissues. Biochem. Biophys. Res. Commun. 43: 1274. Jaynes, D.B. 2011. Confidence bands for measured economically optimal nitrogen rates. Prec. Agric. 12: 196 213. Jensen, P., J. Rytter, E. Detwiler, J. Travis, and T. McNellis. 2003. Rootstock effects on gene expression patterns in apple tree scions. Plant Mol. Biol. 53: 493 511. Jifon, J., K. Crosby, D. Leskovar, and M. Miller. 2008. Possible physiological mechanisms for resistance to vine decline diseases in grafted watermelons. Acta Hortic. 782: 329 334. Johnsen, K.H., B.C. Bongarten, and L.R. Boring. 1991. Effects of nitrate on in vivo nitrate reductase activity of seedlings from three open pollinated families of robinia pseudoacacia. Tree Physiol. 8: 381. Jones, C., P. Mes, and J. Myers. 2003. Chara cterization and inheritance of the anthocyanin fruit (aft) tomato. J. Hered. 94: 449 456. Kamboj, J., P. Blake, J. Quinlan, and D. Baker. 1999. Identification and quantitation by gc ms of zeatin and zeatin riboside in xylem sap from rootstock and scion of grafted apple trees. Plant Growth Regul 28: 199 205. Kanehira, A., K. Yamada, T. Iwaya, R. Tsuwamoto, A. Kasai, M. Nakazono, and T. Harada. 2010. Apple phloem cells contain some mrnas transported over long distances. Tree Genet. Genom. 6: 635 642. Kato, T and H. Lou. 1989. Effects of rootstock on the yield, mineral nutrition and hormone level in xylem sap in eggplant. J J pn. Soc Hortic Sci 58: 345 352. Kawaguchi, M., A. Taji, D. Backhouse, and M. Oda. 2008. Anatomy and physiology of graft incompatibil ity in solanaceous plants. J. Horticult. Sci. Biotechnol. 83: 581 588. Khah, E., E. Kakava, A. Mavromatis, D. Chachalis, and C. Goulas. 2006. Effect of grafting on growth and yield of tomato ( lycopersicon esculentum mill.) in greenhouse and open field. J. Appl. Hort. 8: 3 7.
244 Kim, M., W. Canio, S. Kessler, and N. Sinha. 2001. Developmental changes due to long distance movement of a homeobox fusion transcript in tomato. Science 293: 287 289. King, S., A. Davis, X. Zhang, and K. Crosby. 2010. Genetics, breedin g and selection of rootstocks for solanaceae and cucurbitaceae. Sci. Hortic. 127: 106 111. Kirnak, H., I. Tas, C. Kaya, and D. Higgs. 2002. Effects of deficit irrigation on growth, yield, and fruit quality of eggplant under semi arid conditions. Aust. J. A gric. Res. 53: 1367 1374. Koundouras, S., I. Tsialtas, E. Zioziou, and N. Nikolaou. 2008. Rootstock effects on the adaptive strategies of grapevine (vitis vinifera l. Cv. Cabernet sauvignon) under contrasting water status: Leaf physiological and structural responses. Agric., Ecosyst. Environ. 128: 86 96. Kubota, C., M. Mcclure, N. Kokalis Burelle, M. Bausher, and E. Rosskopf. 2008. Vegetable grafting: History, use, and current technology status in north america. HortScience 43: 1664 1669. Kudo, H. and T. Ha rada. 2007. A graft transmissible rna from tomato rootstock changes leaf morphology of potato scion. HortScience 42: 225 226. Lafever, H.N. 1981. Genetic differences in plant response to soil nutrient stress. J. Plant Nutr. 4: 89 109. Larsen, F., R. Fritts Jr, and K. Olsen. 1985. Rootstock influence on  delicious' and  golden delicious' apple fruit quality at harvest and after storage. Sci. Hortic. 26: 339 349. Lazof, D. and N. Bernstein. 1998. The nacl induced inhibition of shoot growth: The case for d isturbed nutrition with special consideration of calcium. Adv. Bot. Res. 29: 113 189. Lecompte, F., E. Bressoud, L. Pares, and F. Bruyne. 2008. Root and nitrate distribution as related to the critical plant n statu s of a fertigated tomato crop. J. Horticul t. Sci. Biotechnol. 83: 223 231. Lee, J. 1994. Cultivation of grafted vegetables. I. Current status, grafting methods, and benefits. HortScience 29: 235 239. Lee, J. and M. Oda. 2003. Grafting of herbaceous vegetable and ornamental crops. Hortic. Rev. 28: 61 124. Lee, J.M., C. Kubota, S. Tsao, Z. Bie, P.H. Echevarria, L. Morra, and M. Oda. 2010. Current status of vegetable grafting: Diffusion, grafting techniques, automation. Sci. Hortic. 127: 93 105.
245 Leonardi, C. and D. Romano. 2004. Recent issues on veget able grafting. Acta Hortic. 631: 163 174. Leonardi, C. and F. Giuffrida. 2006. Variation of plant growth and macronutrient uptake in grafted tomatoes and eggplants on three different rootstocks. Eur. J. Hortic. Sci. 71: 97 101. Liu, Y.Q., S.Q. Liu, and H.B Wang. 2004a. Effect of salt tolerant stock grafting on the salt tolerant characteristics of watermelon seedlings. Acta Agriculturae Shanghai 20: 62 64. Liu, Y.Q., S.Q. Liu, and H.B. Wang. 2004b. Effect of salt tolerant stock on growth yield and quality o f watermelon. Shandong Agri. Sci. 4: 30 31. Locascio, S.J. 2005. Management of irrigation for vegetables: Past, present, and future. HortTechnology 15: 482 485. Lockard, R. and G. Schneider. 1981. Phenols and the dwarfing mechanism in apple rootstocks. Act a Hortic. 120: 107 112. Lopez Bucio, J., A. Cruz Ramirez, and L. Herrera Estrella. 2003. The role of nutrient availability in regulating root architecture. Curr. Opin. Plant Biol. 6: 280 287. Lopez Perez, J., M. Le Strange, I. Kaloshian, and A. Ploeg. 2006 Differential response of mi gene resistant tomato rootstocks to root knot nematodes ( M eloidogyne incognita ). Crop Protect. 25: 382 388. Lucas, W. and J. Lee. 2004. Plasmodesmata as a supracellular control network in plants. Nat. Rev. Mol. Cell Bio. 5: 71 2 726. Lukyanenko, A., 1991. Disease resistance in tomato, p. 99 119. In: Kalloo, G. (ed.), Genetic improvement of tomato. Springer Verlag, Berlin, Germany. Lund, S. and J. Bohlmann. 2006. The molecular basis for wine grape quality a volatile subject. Scie nce 311: 804 805. Lykas, C., C. Kittas, and A. Zambeka. 2008. Water and fertilizers use efficiency in grafted and non grafted tomato plants on soilless culture. Acta Hortic. 801: 1551 1555. Lynch, J. 1998. The role of nutrient efficient crops in modern agr iculture. J. Crop Prod. 1: 241 264. Machado, R., M. Oliveira, and C. Portas. 2000. Effect of drip irrigation and fertilization on tomato rooting patterns. Acta Hortic. 537: 313 320. Marschner, P. 1995. Mineral nutrition of higher plants. Academic press, Ne w York 889p
246 Marsic, N. and J. Osvald. 2004. The influence of grafting on yield of two tomato cultivars ( lycopersicon esculentum mill.) grown in a plastic house. Acta Agric. Slov. 83: 243 249. Martnez Ballesta, M., C. Alcaraz Lpez, B. Muries, C. Mota Ca denas, and M. Carvajal. 2010. Physiological aspects of rootstock scion interactions. Sci. Hortic. 127: 112 118. Martinez Rodriguez, M.M., M.T. Estan, E. Moyano, J.O. Garcia Abellan, F.B. Flores, J.F. Campos, M.J. Al Azzawi, T.J. Flowers, and M.C. Bolarin. 2008. The effectiveness of grafting to improve salt tolerance in tomato when an "Excluder" Genotype is used as scion. Environ. Exp. Bot. 63: 392 401. Matsuzoe, N., H. Okubo, and K. Fujieda. 1993. Resistance of tomato plants grafted on solanum rootstocks to bacterial wilt and root knot nematode. J Jpn. Soc. Hortic. Sci. 61: 865 865. McAvoy, T., J.H. Freeman, S.L. Rideout, S.M. Olson, and M.L. Paret. 2012. Evaluation of grafting using hybrid rootstocks for management of bacterial wilt in field tomato producti on. HortScience 47: 621 625. Miguel, A., J. Maroto, A. San Bautista, C. Baixauli, V. Cebolla, B. Pascual, S. Lopez, and J. Guardiola. 2004. The grafting of triploid watermelon is an advantageous alternative to soil fumigation by methyl bromide for control of fusarium wilt. Sci. Hortic. 103: 9 17. Mitani, N., R. Matsumoto, T. Yoshioka, and T. Kuniga. 2008. Citrus hybrid seedlings reduce initial time to flower when grafted onto shiikuwasha rootstock. Sci. Hortic. 116: 452 455. Motosugi, H., Y. Gao, and A. Sug iura. 1995. Rootstock effects on fruit quality of  fuji'apples grown with ammonium or nitrate nitrogen in sand culture. Sci. Hortic. 61: 205 214. Mudge, K., J. J., S. Scofield, and E.E. Goldschmidt. 2009. A history of grafting. Hortic. Rev. 35: 437 493. Mylavarapu, R. and E. Kennelly. 2004. Uf/ifas extension soil testing laboratory (estl) analytical procedures and training manual. Circ. 1248. Univ. Of florida, gainesville, p. 19. Nagata, M. and I. Yamashita. 1992. Simple method for simultaneous determinat ion of chlorophyll and carotenoids in tomato fruit. J Jpn Soc Hortic Sci 39: 925 928. Natali, S.M., S.A. Saudo Wilhelmy, and M.T. Lerdau. 2009. Effects of elevated carbon dioxide and nitrogen fertilization on nitrate reductase activity in sweetgum and lob lolly pine trees in two temperate forests. Plant Soil 314: 197 210.
247 Natural Resources Conservation Service, N. 2006. Soil survey of suwannee county, florida.24 July 2011. ht tp://soildatamart.nrcs.usda.gov/Manuscripts/FL121/0/Suwannee.pdf Neeteson, J. and W. Wadman. 1987. Assessment of economically optimum application rates of fertilizer n on the basis of response curves. Nutr. Cycl. Agroecosys. 12: 37 52. Neeteson, J.J., 199 5. Nitrogen management for intensively grown arable crops and field vegetables, p. 295 325. In: Edward, P. B. (ed.), Nitrogen fertilisation and the environment. Marcel Dekker New York, USA. Nielsen, G. and F. Kappel. 1996. 'bing' sweet cherry leaf nutrit ion is affected by rootstock. HortScience 31: 1169 1172. Nikolaou, N., M. Koukourikou, and N. Karagiannidis. 2000. Effects of various rootstocks on xylem exudates cytokinin content, nutrient uptake and growth patterns of grapevine vitis vinifera l. Cv. Tho mpson seedless. Agronomie 20: 363 373. Noda, K., H. Okuda, and I. Iwagaki. 2000. Indole acetic acid and abscisic acid levels in new shoots and fibrous roots of citrus scion rootstock combinations. Sci. Hortic. 84: 245 254. O'Connell, S. 2008. Grafted tomat o performance in organic production systems: Nutrient uptake, plant growth and yield. North Carolina State University, Raleigh, Master of Science. O'Connell, S., S. Hartmann, C.L. Rivard, M.M. Peet, and F.J. Louws. 2009. Grafting tomatoes on disease resist ant rootstocks for small scale organic production.05 01. http://ofrf.org/funded/reports/o%27connell_07f30.pdf Obreza, T.A. and J.B. Sartain. 2010. Improving nitrogen and phosphorus ferti lizer use efficiency for florida's horticultural crops. HortTechnology 20: 23 33. Oda, M. 1999. Grafting of vegetables to improve greenhouse production, p. 11. Food and Fertilizer Technology Center (FFTC). Taipei City, CN. Oda, M. 2007. Vegetable seedling grafting in japan. Acta Hort. 759: 175 180. Oda, M., M. Nagata, K. Tsuji, and H. Sasaki. 1996. Effects of scarlet eggplant rootstock on growth, yield, and suar content of grafted tomato fruits. J Jpn. Soc. Hortic. Sci. 65: 531 536. Okimura, M., S. Matsuo, K. Arai, and S. Okitsu. 1986. Influence of soil temperature on the growth of fruit vegetable grafted on different stocks. Bulletin of the Vegetable and Ornamental Crops Research Station. Series C. Kurume (Japan) 9: 43 58.
248 Okudai, N., K. Yoshinaga, T. Takah ara, D. Ishiuchi, and I. Oiyama. 1980. Studies on the promotion of flowering and fruiting of juvenile citrus seedlings. Ii. Effect of grafting. Bulletin of the Fruit Tree Research Station 2: 15 28. Olson, S., W. Stall, G. Vallad, S. Webb, T. Taylor, S.A. S mith, E. Simonne, E. McAvoy, and B. Santos, 2010. Tomato production in florida, p. 291 312. In: Olson, S. and Simonne, E. (eds.), Vegetable production handbook for florida. IFAS, University of Florida, Gainesville Florida. Omid, A., T. Keilin, A. Glass, D. Leshkowitz, and S. Wolf. 2007. Characterization of phloem sap transcription profile in melon plants. J. Exp. Bot. 58: 3645. Ou, B., D. Huang, M. Hampsch Woodill, J. Flanagan, and E. Deemer. 2002. Analysis of antioxidant activities of common vegetables emp loying oxygen radical absorbance capacity (orac) and ferric reducing antioxidant power (frap) assays: A comparative study. J. Agric. Food Chem 50: 3122 3128. Ozden, M., H. Vardin, M. Simsek, and M. Karaaslan. 2010. Effects of rootstocks and irrigation leve ls on grape quality of vitis vinifera l. Cv. Shiraz. Afri. J. Biotech. 9: 3801 3807. Ozores Hampton, M., E. Simonne, F. Roka, K. Morgan, S. Sargent, C. Snodgrass, and E. McAvoy. 2012. Nitrogen rates effects on the yield, nutritional status, fruit quality, and profitability of tomato grown in the spring with subsurface irrigation HortScience 47: 1129 1133. Oztekin, G., F. Giuffrida, Y. Tuzel, and C. Leonardi. 2009. Is the vigour of grafted tomato plants related to root characteristics? J. Food Agric. Envion. 7: 364 368. Parejo, J., S. Minguez, J. Sella, and E. Espinas. 1993. Sixteen years of monitoring the cultivar xarel. Lo (vitis vinifera l.) on several rootstocks. Acta Hortic. 388: 123 128. Parisi, M., L. Giordano, A. Pentangelo, B. D'Onofrio, and G. Villa ri. 2004. Effects of different levels of nitrogen fertilization on yield and fruit quality in processing tomato. Acta Hortic. 700:129 132. Passam, H., M. Stylianou, and A. Kotsiras. 2005. Performance of eggplant grafted on tomato and eggplant rootstocks. E ur. J. Hortic. Sci. 70: 130 134. Pavlou, G., D. Vakalounakis, and E. Ligoxigakis. 2002. Control of root and stem rot of cucumber, caused by fusarium oxysporum f. Sp. Radicis cucumerinum, by grafting onto resistant rootstocks. Plant Dis. 86: 379 382. Peng, Y.Y., Y.W. Zhang, and J.N. Ye. 2008. Determination of phenolic compounds and ascorbic acid in different fractions of tomato by capillary electrophoresis with electrochemical detection. J. Agric. Food Chem. 56: 1838 1844.
249 Pina, A. and P. Errea. 2005. A revi ew of new advances in mechanism of graft compatibility incompatibility. Sci. Hortic. 106: 1 11. Pitts, D.J., A.G. Smajstrla, D.Z. Haman, and C. G.A. 2002. Irrigation costs for tomato production in florida. Electronic database information system,, p. 6, AE7 4. Agricultural and Biological Engineering Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida, Gainesville. Pogonyi, A., Z. Pek, L. Helyes, and A. Lugasi. 2005. Effect of grafting on the to mato's yield, quality and main fruit components in spring forcing. Acta Aliment. Hung. 34: 453 462. Poh, B.L., A. Gazula, E.H. Simonne, F. Di Gioia, R.C. Hochmuth, and M.R. Alligood. 2011. Use of reduced irrigation operating pressure in irrigation scheduli ng. I. Effect of operating pressure, irrigation rate, and nitrogen rate on drip irrigated fresh market tomato nutritional status and yields: Implications on irrigation and fertilization management HortTechnology 21: 14 21. Pou, A., J. Flexas, M. del Mar Al sina, J. Bota, C. Carambula, F. De Herralde, J. Galms, C. Lovisolo, M. Jimnez, and M. Ribas Carb. 2008. Adjustments of water use efficiency by stomatal regulation during drought and recovery in the drought adapted vitis hybrid richter 110 ( v. Berlandier i v. Rupestris ). Physiol. Plant. 134: 313 323. Prakash, O., A. Alva, and S. Paramasivam. 1999. Use of the urease inhibitor n (n butyl) thiophosphoric triamide decreased nitrogen leaching from urea in a fine sandy soil. Water, Air, Soil Pollut. 116: 587 59 5. Proietti, S., Y. Rouphael, G. Colla, M. Cardarelli, M. De Agazio, M. Zacchini, E. Rea, S. Moscatello, and A. Battistelli. 2008. Fruit quality of mini watermelon as affected by grafting and irrigation regimes. J. Sci. Food Agric. 88: 1107 1114. Pulgar, G ., G. Villora, D. Moreno, and L. Romero. 2000. Improving the mineral nutrition in grafted watermelon plants: Nitrogen metabolism. Biol. Plant. 43: 607 609. Qaryouti, M., W. Qawasmi, H. Hamdan, and M. Edwan. 2007. Tomato fruit yield and quality as affected by grafting and growing system. Acta Hortic. 741: 199 206. Rahman, M., M. Rashid, M. Salam, M. Masud, A. Masum, and M. Hossain. 2002. Performance of some grafted eggplant genotypes on wild solanum root stocks against root knot nematode. J Bio Sci 2: 446 448. Rajabi, A., E.S. Ober, and H. Griffiths. 2009. Genotypic variation for water use efficiency, carbon isotope discrimination, and potential surrogate measures in sugar beet. Field Crops Res. 112: 172 181.
250 Rawlings, J.O., S.G. Pantula, and D.A. Dickey. 1998. Applied regression analysis: A research tool. Springer Verlag New York 646p Rengel, Z. and P. Marschner. 2005. Nutrient availability and management in the rhizosphere: Exploiting genotypic differences. New Phytol. 168: 305 312. Rengel, Z. and P.M. Damon. 2008. Crops and genotypes differ in efficiency of potassium uptake and use. Physiol. Plant. 133: 624 636. Reynolds, A.G. and D.A. Wardle. 2001. Rootstocks impact vine performance and fruit composition of grapes in british columbia. HortTechnology 11 : 419 427. Ricrdez Salinas, M., M. Huitrn Ramrez, J. Tello Marquina, and F. Camacho Ferre. 2010. Planting density for grafted melon as an alternative to methyl bromide use in mexico. Sci. Hortic. 126: 236 241. Rivard, C. and F. Louws. 2008. Grafting to manage soilborne diseases in heirloom tomato production. HortScience 43: 2104 2111. Rivard, C., S. O'Connell, M. Peet, and F. Louws. 2010a. Grafting tomato with interspecific rootstock to manage diseases caused by sclerotium rolfsii and southern root knot nematode. Plant Dis. 94: 1015 1021. Rivard, C., S. O'Connell, M. Peet, R. Welker, and F. Louws. 2012. Grafting tomato to manage bacterial wilt caused by ralstonia solanacearum in the southeastern united states. Plant Dis. 96: 973 978. Rivard, C.L., O. Sydo rovych, S. O'Connell, M.M. Peet, and F.J. Louws. 2010b. An economic analysis of two grafted tomato transplant production systems in the united states. HortTechnology 20: 794 803. Rivero, R., J. Ruiz, and L. Romero. 2003a. Role of grafting in horticultural plants under stress conditions. J. Food Agric. Environ. 1: 70 74. Rivero, R.M., J.M. Ruiz, and L. Romero. 2003b. Can grafting in tomato plants strengthen resistance to thermal stress? J. Sci. Food Agric. 83: 1315 1319. Rivero, R.M., J.M. Ruiz, and L. Romer o. 2004. Iron metabolism in tomato and watermelon plants: Influence of grafting. J. Plant Nutr. 27: 2221 2234. Roberts, K., S. Sargent, and A. Foxx. 2002. Effects of storage temperature on ripening and postharvest quality of grape and mini pear tomatoes. P roc. Fla. State Hort. Soc 115:80 84. Romano, D. and A. Paratore. 2001. Effects of gr afting on tomato and eggplant. Acta Hort 559:149 154.
251 Romero, L., A. Belakbir, L. RAGALA, and J. Ruiz. 1997. Response of plant yield and leaf pigments to saline conditio ns: Effectiveness of different rootstocks in melon plants( cucumis memo l.). Soil Sci. Plant Nutr. 43: 855 862. Rouphael, Y., M. Cardarelli, E. Rea, and G. Colla. 2008a. Grafting of cucumber as a means to minimize copper toxicity. Environ. Exp. Bot. 63: 49 58. Rouphael, Y., M. Cardarelli, G. Colla, and E. Rea. 2008b. Yield, mineral composition, water relations, and water use efficiency of grafted mini watermelon plants under deficit irrigation. HortScience 43: 730 736. Rouphael, Y., D. Schwarz, A. Krumbe in, and G. Colla. 2010. Impact of grafting on product quality of fruit vegetables. Sci. Hortic. 127: 172 179. Ruhl, E., P. Clingeleffer, P. Nicholas, R. Cirami, M. McCarthy, and J. Whiting. 1988. Effect of rootstocks on berry weight and ph, mineral content and organic acid concentrations of grape juice of some wine varieties. Aust J. Exp. Agric. 28: 119 125. Ruiz, J. and L. Romero. 1999. Nitrogen efficiency and metabolism in grafted melon plants. Sci. Hortic. 81: 113 123. Ruiz, J., A. Belakbir, and L. Romer o. 1996. Foliar level of phosphorus and its bioindicators in cucumis melo grafted plants. A possible effect of rootstock. J. Plant Physiol. 149: 400 404. Ruiz, J., A. Belakbir, I. Lpez Cantarero, and L. Romero. 1997. Leaf macronutrient content and yield i n grafted melon plants. A model to evaluate the influence of rootstock genotype. Sci. Hortic. 71: 227 234. Ruiz, J., R. Rivero, L. Cervilla, R. Castellano, and L. Romero. 2006a. Grafting to improve nitrogen use efficiency traits in tobacco plants. J. Sci. Food Agric. 86: 1014 1021. Ruiz, J.M., R.M. Rivero, L.M. Cervilla, R. Castellano, and L. Romero. 2006b. Grafting to improve nitrogen use efficiency traits in tobacco plants. J. Sci. Food Agric. 86: 1014 1021. Sainju, U., B. Singh, and W. Whitehead. 2001. C omparison of the effects of cover crops and nitrogen fertilization on tomato yield, root growth, and soil properties. Sci. Hortic. 91: 201 214. Salam, M., A. Masum, S. Chowdhury, M. Dhar, M. Saddeque, and M. Islam. 2002. Growth and yield of watermelon as i nfluenced by grafting. OnLine J. Biol. Sci 2: 298 299.
252 Salehi Mohammadi, R., A. Kashi, H. YunChan, M. Babalar, and M. Delshad. 2009. Assessing the survival and growth performance of iranian melon to grafting onto cucurbita rootstocks. Korean J Hortic Sci. 27: 1 6. Saltveit, M., 2005. Fruit ripening and fruit quality, p. 145 170. In: Heuvelink, E. (ed.), Tomatoes, crop production science in horticulture. CAB INTERNATIONAL. Samad, A., D. McNeil, and Z. Khan. 1999. Effect of interstock bridge grafting (m9 dwar fing rootstock and same cultivar cutting) on vegetative growth, reproductive growth and carbohydrate composition of mature apple trees. Sci. Hortic. 79: 23 38. Sanchez, E., R.M. Rivero, J.M. Ruiz, and L. Romero. 2004. Changes in biomass, enzymatic activity and protein concentration in roots and leaves of green bean plants (phaseolus vulgaris l. Cv. Strike) under high nh4no3 application rates. Sci. Hortic. 99: 237 248. Santos, B.M., J.W. Scott, and M. Ramrez Snchez. 2010. In row distances and nitrogen fert 579 584. Satisha, J., G. Prakash, and R. Venugopalan. 2006. Statistical modeling of the effect of physio biochemical parameters on water use efficiency of grape varieties, rootstocks an d their stionic combinations under moisture stress conditions. Turk. J. Agric. For. 30: 261 271. Scalzo, J., A. Politi, N. Pellegrini, B. Mezzetti, and M. Battino. 2005. Plant genotype affects total antioxidant capacity and phenolic contents in fruit. Nutr ition 21: 207 213. Scholberg, J., B. McNeal, K. Boote, J. Jones, S. Locascio, and S. Olson. 2000. Nitrogen stress effects on growth and nitrogen accumulation by field grown tomato. Agron. J. 92: 159 167. Schwarz, D., Y. Rouphael, G. Colla, and J. Venema. 2 010. Grafting as a tool to improve tolerance of vegetables to abiotic stresses: Thermal stress, water stress and organic pollutants. Sci. Hortic. 127:162 171. Scott, J. 2002. A breeder's perspective on the use of molecular techniques for improving fruit qu ality. HortScience 37: 464 467. Seong, K., J. Moon, S. Lee, Y. Kang, K. Kim, and H. Seo. 2003. Growth, lateral shoot development, and fruit yield of white spined cucumber ( cucumis sativus cv. Baekseong 3) as affected by grafting methods. J Korean Soc Hor tic Sci 44: 478 482. Simonne, A.H., C. do Nascimento Nunes, and J.K. Brecht, 2011. T omato and other solanaceous fruits, p. 321. In: Terry, L. A. (ed.), Health promoting properties of fruit and vegetables. CABI Publishing, Oxfordshire, UK.
253 Simonne, A.H., J.M. Fuzere, E. Simonne, R.C. Hochmuth, and M.R. Marshall. 2007. Effects of nitrogen rates on chemical composition of yellow grape tomato grown in a subtropical climate. J. Plant Nutr. 30: 927 935. Simonne, E., D. Studstill, and R. Hochmuth. 2004. Understa nding water movement in mulched beds on sandy soils: An approach to ecologically sound fertigation in vegetable production. Acta Hortic. 700: 173 178. Simonne, E., M. Dukes, G. Hochmuth, B. Hochmuth, and D. Studstill. 2006. Monitoring nitrate concentration in shallow ground water below a vegetable field. Proc. Fla. State Hort. Soc. 119:226 230. Simonne, E., D. Studstill, B. Hochmuth, T. Olczyk, M. Dukes, R. Munoz Carpena, and Y. Li. 2010a. Drip irrigation: The BMP era an integrated approach to water and fer tilizer management for vegetables grown with plasticulture, p. 13, Vegetable Nutrtion Management Serie. Horticultural Sciences Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida., Gainesvil le. Simonne, E., C. Hutchinson, J. DeValerio, R. Hochmuth, D. Treadwell, A. Wright, B. Santos, A. Whidden, G. McAvoy, and X. Zhao. 2010b. Current knowledge, gaps, and future needs for keeping water and nutrients in the root zone of vegetables grown in flor ida. HortTechnology 20: 143 152. Simonne, E.H., D.W. Studstill, R.C. Hochmuth, G. McAvoy, M.D. Dukes, and S. Olson. 2003. Visualization of water movement in mulched beds with injections of dye with drip irrigation. Proc. Fla. State Hort. Soc. 116:88 91. Si ngandhupe, R., G. Rao, N. Patil, and P. Brahmanand. 2003. Fertigation studies and irrigation scheduling in drip irrigation system in tomato crop ( L ycopersicon esculentum l.). Eur. J. Agron. 19: 327 340. Soar, C., P. Dry, and B. Loveys. 2006. Scion photosyn thesis and leaf gas exchange in vitis vinifera l. Cv. Shiraz: Mediation of rootstock effects via xylem sap aba. Aust. J. Grape Wine Res. 12: 82 96. Solari, L., S. Johnson, and T. DeJong. 2006. Hydraulic conductance characteristics of peach ( prunus persica ) trees on different rootstocks are related to biomass production and distribution. Tree Physiol. 26: 1343 1350. Sorce, C., R. Massai, P. Picciarelli, and R. Lorenzi. 2002. Hormonal relationships in xylem sap of grafted and ungrafted prunus rootstocks. Sci. Hortic. 93: 333 342. Sorgon A., M. Abenavoli, P. Gringeri, and G. Cacco. 2006. A comparison of nitrogen use efficiency definitions in citrus rootstocks. Sci. Hortic. 109: 389 393. Stegemann, S. and R. Bock. 2009. Exchange of genetic material between cel ls in plant tissue grafts. Science 324: 649.
254 Stitt, M. 1999. Nitrate regulation of metabolism and growth. Curr. Opin. Plant Biol. 2: 178 186. Storey, R. and R. Walker. 1999. Citrus and salinity. Sci. Hortic. 78: 39 81. Su, Y.H., Y.B. Liu, and X.S. Zhang. 2 011. Auxin cytokinin interaction regulates meristem development. Mol. Plant 4: 616 625. Sweeney, D., D. Graetz, A. Bottcher, S. Locascio, and K. Campbell. 1987. Tomato yield and nitrogen recovery as influenced by irrigation method, nitrogen source, and mul ch. HortScience 22: 27 29. Sydorovych, O., C. Safley, R. Welker, L. Ferguson, D. Monks, K. Jennings, J. Driver, and F. Louws. 2008. Economic evaluation of methyl bromide alternatives for the production of tomatoes in north carolina. HortTechnology 18: 705 713. Syvertsen, J. and G. Yelenosky. 1988. Salinity can enhance freeze tolerance of citrus rootstock seedlings by modifying growth, water relations, and mineral nutrition. J. Am. Soc. Hort. Sci. 113: 889 893. Tageldin, M.H.A. and N.K.B. El Gizawy. 2005. Li near and nonlinear segmented models describing response of maize grain yield to nitrogen fertilization, The 11th Conference of Agronomy, Agronomy Department Faculty of Agriculture, Assiut University. Tagliavini, M., D. Scudellari, B. Marangoni, A. Bastia nel, F. Franzin, and M. Zamborlini. 1992. Leaf mineral composition of apple tree: Sampling date and effects of cultivar and rootstock. J. Plant Nutr. 15: 605 619. Tan, C., C. Drury, W. Reynolds, P. Groenevelt, and H. Dadfar. 2002. Water and nitrate loss th rough tiles under a clay loam soil in ontario after 42 years of consistent fertilization and crop rotation. Agric., Ecosyst. Environ. 93: 121 130. Taylor, M., B. Bruton, W. Fish, and W. Roberts. 2008. Cost benefit analyses of using grafted watermelon trans plants for fusarium wilt disease control. Acta Hortic. 782: 343 350. Tei, F., P. Benincasa, and M. Guiducci. 2002. Critical nitrogen concentration in processing tomato. Eur. J. Agron. 18: 45 55. Tischner, R. 2000. Nitrate uptake and reduction in higher and lower plants. Plant, Cell Environ. 23: 1005 1024. Toor, R. and G. Savage. 2005. Antioxidant activity in different fractions of tomatoes. Food Res. Int. 38: 487 494.
255 Topcu, S., C. Kirda, Y. Dasgan, H. Kaman, M. Cetin, A. Yazici, and M. Bacon. 2007. Yield r esponse and n fertiliser recovery of tomato grown under deficit irrigation. Eur. J. Agron. 26: 64 70. Traka Mavrona, E., M. Koutsika Sotiriou, and T. Pritsa. 2000. Response of squash (cucurbita spp.) as rootstock for melon (cucumis melo l.). Sci. Hortic. 8 3: 353 362. Trudel, M.J. and J.L. Ozbun. 1971. Influence of potassium on carotenoids content of tomato fruit. J. Amer. Soc. Hort. Sci. 96: 763 765. Tsipouridis, C. and T. Thomidis. 2005. Effect of 14 peach rootstocks on the yield, fruit quality, mortality, girth expansion and resistance to frost damages of may crest peach variety and their susceptibility on phytophthora citrophthora. Sci. Hortic. 103: 421 428. Turquois, N. and M. Malone. 1996. Non destructive assessment of developing hydraulic connections i n the graft union of tomato. J. Exp. Bot. 47: 701 707. Tworkoski, T. and S. Miller. 2007. Rootstock effect on growth of apple scions with different growth habits. Sci. Hortic. 111: 335 343. Uddin, M.Q., B. Kubat, and I. Dormling. 1985. Water stress induced changes in cytokinins content determined by immunological method in scots pine seedlings. Physiol. Plant. 64: A27 A27. Ulukapi, K. and A. Onus. 2007. Comparison of the productivity and quality of the grafted and ungrafted tomato plants grown in the greenh ouse with mycorrhiza application. Acta Hortic. 758: 345 350. USDA. 2010. Agricultural statistics data base. 05/24/2012. http://usda.mannlib.cornell.edu/MannUsda/vi ewDocumentInfo.do?documentID=12 10 USDA. 2011. Agricultural marketing service. 04/25/2012. http://www.marketnews.usda.gov/portal/fv?paf_dm=full&paf_gear_id=1200002&st artIndex=1&dr=1&rowDispl ayMax=50&displaySort=&navClass=&repType=shipPri ceDaily&termNavClass=&shipNavClass=&movNavClass=&locName=&locAbr=&co mmAbr=TOM&commName=TOMATOES&varName=&repDate=07%2F01%2F201 1&environment=&organic=&x=11&y=3 Valle, E.M., S.B. Boggio, and H.W. Heldt. 1998. Free amino acid composition of phloem sap and growing fruit of Lycopersicon esculentum Plant Cell Physiol. 39:458 461. van Hooijdonk, B., D. Woolley, I. Warrington, and S. Tustin. 2011. Rootstocks modify scion architecture, endogenous hormones, and root g 102.
256 Vzquez, N., A. Pardo, M. Suso, and M. Quemada. 2006. Drainage and nitrate leaching under processing tomato growth with drip irrigation and plastic mulching. Agric., Ecosys t. Environ. 112: 313 323. Venema, J.H., B.E. Dijk, J.M. Bax, P.R. van Hasselt, and J.T.M. Elzenga. 2008. Grafting tomato ( solanum lycopersicum ) onto the rootstock of a high altitude accession of solanum habrochaites improves suboptimal temperature toleranc e. Environ. Exp. Bot. 63: 359 367. Vinkovic Vrcek, I., V. Samobor, M. Bojic, M. Medic Saric, M. Vukobratovic, R. Erhatic, D. Horvat, and Z. Matotan. 2011. The effect of grafting on the antioxidant properties of tomato (solanum lycopersicum l.). Span. J. Ag ric. Res. 9: 844 851. Virgona, J., J. Smith, and B. Holzapfel. 2003. Scions influence apparent transpiration efficiency of vitis vinifera (cv. Shiraz) rather than rootstocks. Aust J Grape Wine Res 9: 183 185. Walker, R., P. Read, and D. Blackmore. 2000. Rootstock and salinity effects on rates of berry maturation, ion accumulation and colour development in shiraz grapes. Aust. J. Grape Wine Res. 6: 227 239. Walker, R.B., D. Blackmore, P. Clingeleffer, and C.L. Ray. 200 2 Rootstock effects on salt toleranc e of irrigated field grown grapevines ( vitis vinifera l. Cv. Sultana). I. Yield and vigor inter relationships Aust. J. Grape Wine Res. 8: 3 14. Webster, A. 1997. A review of fruit tree roots tock research and development. Acta Hortic 451:53 74. Webster, A and S. Wertheim, 2003. Apple rootstocks, p. 91 124. In: Ferree, D. C. a. I. J. W. (ed.), Apples: Botany, production, and uses. CABI, London UK. Webster, T. 1994. Rootstock and interstock effects on deciduous fruit tree growth and cropping a brief review. Compact Fruit Tree 27: 5 16. Weng, J.H. 2000. The role of active and passive water uptake in maintaining leaf water status and photosynthesis in tomato under water deficit. Plant Prod. Sci. 3: 296 298. Wheaton, T., J. Whitney, W. Castle, R. Muraro, H. Bro wning, and D. Tucker. 1995. Citrus scion and rootstock, topping height, and tree spacing affect tree size, yield, fruit quality, and economic return. J J. Am. Soc. Hort. Sci. 120: 861 861. Wiesler, F., T. Behrens, and W.J. Horst. 2001. The role of nitrogen efficient cultivars in sustainable agriculture. ScientificWorldJournal 1:61 69. Willcutts, J., A. Overman, G. Hochmuth, D. Cantliffe, and P. Soundy. 1998. A comparison of three mathematical models of response to applied nitrogen: A case study using lettuc e. HortScience 33: 833 836.
257 Wolf, T. and R. Pool. 1988. Effects of rootstock and nitrogen fertilization on the growth and yield of chardonnay grapevines in new york. Am. J. Enol. Vitic. 39: 29. Wood, J., 1980. The mathematical expression of crop response t o inputs, p. 263 271. In: Berne (ed.), Physiological aspects of crop productivity. International Potash Institute. Xu, H. and R. Ishii. 1996. Wheat cultivar differences in photosynthetic response to low soil water potentials. Jpn. J. Crop Sci. 65: 518 524. Xu, H., W. Zhang, M. Li, T. Harada, Z. Han, and T. Li. 2010. Gibberellic acid insensitive mrna transport in both directions between stock and scion in malus. Tree Genet. Genom. 6 : 1013 1019 Yamasaki, A., M. Yamashita, and S. Furuya. 1994. Mineral concent rations and cytokinin activity in the xylem exudate of grafted watermelons as affected by rootstocks and crop load. J J Jpn. Soc. Hortic. Sci. 62: 817 817. Yang, L.F., Y.L. Zhu, C.M. Hu, Z.L. Liu, and G.W. Zhang. 2006. Effects of nacl stress on the conten ts of the substances regulating membrane lipid oxidation and osmosis and photosynthetic charact eristics of grafted cucumber Acta Bot. 26: 1195 1200. Yetisir, H. and N. Sari. 2003. Effect of different rootstock on plant growth, yield and quality of waterme lon. Aust J. Exp. Agric. 43: 1269 1274. Yetisir, H., M. aliskan, S. Soylu, and M. Sakar. 2006. Some physiological and growth responses of watermelon [ citrullus lanatus (thunb.) matsum. And nakai] grafted onto lagenaria siceraria to flooding. Environ. Exp. Bot. 58: 1 8. Yetisir, H., S. Kurt, N. Sari, and M. Tok. 2007. Rootstock potential of turkish lagenaria siceraria germplasm for watermelon: Plant growth, graft compatibility, and resistance to fusarium. Turk. J. Agric. For 31: 381 388. Yoshida, T. 1980. R elationship between length of the juvenile period and tree growth in citrus seedlings. Bull. Fruit Tree Res Station B7: 1 13. Zhang, H.F. 2002. Effects of low root temperature on leaf and root physiological characteristics of grafted cucumber. Sci. Tech. Info. Gansu 31: 33 35. Zhang, J., X. Sui, B. Li, B. Su, J. Li, and D. Zhou. 1998. An improved water use efficiency for winter wheat grown under reduced irrigation. Field Crops Res. 59: 91 98. Zhang, Z., Y. Wang, L. Long, Y. Lin, J. Pang, and B. Liu. 2008. Tomato rootstock effects on gene expression patterns in eggplant scions. Russ. J. Plant Physl. 55: 93 100.
258 Zotarelli, L., M. Dukes, J. Scholberg, R. Muoz Carpena, and J. Icerman. 2009a. Tomato nitrogen accumulation and fertilizer use efficiency on a sandy soil, as affected by nitrogen rate and irrigation scheduling. Agric. Water Manage. 96: 1247 1258. Zotarelli, L., J.M. Scholberg, M.D. Dukes, R. Munoz Carpena, and J. Icerman. 2009b. Tomato yield, biomass accumulation, root distribution and irrigation wate r use efficiency on a sandy soil, as affected by nitrogen rate and irrigation scheduling. Agric. Water Manage. 96: 23 34. Zotarelli, L., M.D. Dukes, J.M. Scholberg, T. Hanselman, K. Le Femminella, and R. Munoz Carpena. 2008. Nitrogen and water use efficien cy of zucchini squash for a plastic mulch bed system on a sandy soil. Sci. Hortic. 116: 8 16.
259 BIOGRAPHICAL SKETCH Desire Djidonou earned an Agronomist Engineer Degree (Major: c rop p roduction) from the College of Agronomic Sciences of the National Unive rsity of Benin. He then worked 2 years on a cowpea research project to promote an Integrated Pest Management approach for controlling the pests and diseases that inflict this legume. Later, he worked as a seasonal vegetable scout with Glades Crop Care, Inc ., based in Jupiter, Florida. This company offers to Florida vegetable growers, an agricultural consultancy through integrated pest management programs. As a member of the consulting team, Desire carried out scouting responsibilities on vegetables such as tomatoes, peppers, potatoes and cucurbits. He also had the opportunity to be involved in contract research activities where he participated in the evaluation of pesticide efficacy in different cropping systems. In August of 200 8 Desire completed his Maste r of Science degree in the Agronomy Department of the University of Florida under the direction of K enneth J. Boote. His major field of study was ecology and physiology with an emphasis on crop modeling. His research the CROPGRO legu me model to simulate growth and fresh market yield of snap bean his Doctor of Philosophy degree in the H orticultural S cience s Department at the University of Florida