Rootstock Breeding for Resistance to the Peach Root-Knot Nematode (Meloidogyne Floridensis)

ROOTSTOCK BREEDING FOR RESISTANCE TO THE PEACH ROOT-KNOT NEMATODE (MELOIDOGYNE FLORIDENSIS)

MARY ANN DY MAQUILAN

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

2017

To my mother: Lilia Dy Maquilan

ACKNOWLEDGMENTS

I thank my advisors Dr. Mercy Olmstead and Dr. José Chaparro for their kindness, encouragement, and extensive patience with me throughout this work. The complementary contributions of my committee to ensure successful completion of my dissertation as well as their support in my personal and professional growth are gratefully acknowledged: To Dr. Mercy Olmstead for trusting me with this project, for granting me many opportunities to hone my research skills, and for the sustained support that extends beyond academia. To Dr. José Chaparro for helping me conceptualize and refine my research questions, for broadening my cognition beyond

Mendelian genetics, for his meticulous review of the early drafts of my manuscript, and for the valuable life lessons imparted. To Dr. Donald Dickson for the warm and sincere support, for giving me the needed nematology training and providing the resources to successfully conduct the nematode resistance evaluations. To Dr. James Olmstead for teaching me to logically work through my genetic marker data.

Special thanks to Dr. Thomas Beckman for providing seeds and cuttings of some of the rootstock materials for the field trial, for sharing his experience in rootstock breeding, and for the insightful discussions over the course of this work. To Dr. Andrew

Nyczepir for his feedback into the design and evaluation of my nematology-related experiments. To Dr. Dario Chavez for teaching me the necessary skills to be able to conduct molecular characterization studies. To Dr. Maria Mendes for the valuable technical guidance and assistance in nematology work. To Dr. Edzard van Santen for assistance in statistical analysis of the rootstock trial data. To Drs. Wayne Sherman and Paul Lyrene for bolstering my interest in plant breeding by sharing their knowledge and showing passion in their work.

The completion of this research could not have been possible without the field logistics support of Matthew Ross during the initial stages of my research and the untiring assistance of Moshe Doron in field and greenhouse during the data gathering stages. I thank Werner Collante for lending me some needed lab supplies and teaching me some molecular marker techniques. I also thank Ashley Kreynin, James Haddix,

Alec McCloud, and Lorenzo Collante who offered help as needed and went well beyond what was required of them by showing empathy to my plants, being willing to work long hours, and making work more enjoyable. I thank Mark Gal for the technical advice with regards to maintenance of plants in the greenhouse and trees in the field. Cecile Shine who helped me set up irrigation for my greenhouse plants and for sharing his experiences in growing plants. John Thomas for lending me nifty tools for pollen collection and cracking seeds. The staff at Citra PSREU especially Jim Boyer for coordinating assistance in the field, and Mark Kann for ensuring that the greenhouse was set up to provide optimum conditions for my resistance screening experiments. To my lab colleagues: Zilfina, Maraisa, Elizabeth, Carlos, Ben, Emily, Rachel, Daniel,

Silvia, Sai, and Weimin, for their kind assistance and wonderful companionship.

I thank my mother, Lilia, for being a role model of dedication and perseverance, and for her faith in me and pride in even my smallest successes that motivate me to keep moving toward my goals.

This dissertation is possible thanks to funding from the Florida Department of

Agriculture and Consumer Services under the Specialty Crop Block Grant Contract

Nos.18004 and 20727.

TABLE OF CONTENTS

page

ACKNOWLEDGMENTS ...... 4

LIST OF TABLES ...... 8

LIST OF FIGURES ...... 11

LIST OF ABBREVIATIONS ...... 13

ABSTRACT ...... 16

CHAPTER

1 LITERATURE REVIEW ...... 18

Rootstock Development for Southeastern United States ...... 18 Sources and Inheritance of Resistance ...... 21 Location of Resistance Genes and Associated Markers ...... 26 Host-Plant and Nematode Interaction ...... 28 Screening for Resistance ...... 32 The Peach Root-knot Nematode Threat in Florida ...... 33 Research Objectives ...... 35

2 INHERITANCE OF RESISTANCE TO MELOIDOGYNE FLORIDENSIS IN INTERSPECIFIC PEACH X PRUNUS KANSUENSIS PROGENIES ...... 37

Introduction ...... 37 Materials and Methods...... 42 Nematode Isolate ...... 42 Genetic Material ...... 42 Growth Conditions ...... 43 Resistance Evaluation ...... 44 Data Analyses ...... 45 Results ...... 46 Selection Criteria for Resistance ...... 46 Segregation Analyses of P. persica x P. kansuensis F2 Families ...... 47 Testcross of F1 Hybrids with P. persica ...... 48 Discussion ...... 49

3 GENETIC ANALYSES OF RESISTANCE TO THE PEACH ROOT-KNOT NEMATODE USING MICROSATELLITE MARKERS ...... 61

Introduction ...... 61 Materials and Methods...... 64 Plant Material ...... 64

Phenotyping ...... 65 Genotyping ...... 65 Map Construction ...... 69 Data Analyses ...... 71 Results ...... 73 Resistance Segregation ...... 73 Polymorphisms and Genotype Segregation at Microsatellite Loci ...... 74 Linkage Maps ...... 76 Marker-trait Association ...... 78 Localization of Major Resistance Locus on LG 2 ...... 81 Discussion ...... 82

4 HORTICULTURAL EVALUATION OF POTENTIAL ROOTSTOCKS FOR LOW- CHILL PEACH PRODUCTION DURING INITIAL YEARS OF FIELD ESTABLISHMENT ...... 95

Introduction ...... 95 Materials and Methods...... 98 Rootstocks and Tree Preparation ...... 98 Field Site and Experimental Design ...... 99 Cultural Management ...... 100 Data Collection ...... 100 Data Analyses ...... 103 Results ...... 104 Tree Growth ...... 104 Phenological Observations ...... 104 Productive Performance ...... 106 Fruit Quality ...... 106 Discussion ...... 107

5 CONCLUDING REMARKS ...... 120

SUPPLEMENTARY DATA

A SEGREGATION ANALYSES ...... 123

B LINKAGE ANALYSIS ...... 126

C ROOTSTOCK EVALUATION ...... 153

LIST OF REFERENCES ...... 158

BIOGRAPHICAL SKETCH ...... 176

LIST OF TABLES

Table page

2-1 Interspecific peach x Prunus kansuensis F2 and testcross progenies used in the segregation analyses for resistance to Meloidogyne floridensis...... 56

2-2 Host resistance classification system based on the presence of galls and egg masses on roots...... 57

2-3 Segregation for resistance to Meloidogyne floridensis in Prunus persica x Prunus kansuensis F2 and testcross progenies based on gall index and chi- square tests of segregation data under a single-gene model...... 57

3-1 Segregation in F2 and BC1F1 interspecific peach x Prunus kansuensis populations for resistance to Meloidogyne floridensis ‘MFGnv14’ isolate based on root galling index and test of the data to single-gene ratio...... 89

3-2 Associations between genotypes at UDP98-025 locus and Meloidogyne floridensis (‘MFGnv14’ isolate) resistance with three alleles segregating in F2 and BC1F1 populations derived from peach x Prunus kansuensis crosses...... 90

3-3 Co-segregation of Meloidogyne floridensis-resistant genotypes with alleles at the UDP98-025 marker locus...... 91

3-4 Allelic relationships at the Mf locus and corresponding phenotypic effects...... 91

4-1 Description of rootstocks used in the study...... 113

4-2 Growth of ‘UFSun’ peach trees budded on five different rootstocks during the first three years of field establishment (2014-2016) at Citra, FL...... 114

4-3 Phenological characteristics of ‘UFSun’ peach trees as influenced by the rootstock genotype during the third year of field establishment (2016) at Citra, FL...... 115

4-4 Productive performance of ‘UFSun’ peach trees on five different rootstocks during the first three years of field establishment (2014-2016) at Citra, FL...... 116

4-5 Fruit quality of ‘UFSun’ peach on five different rootstocks during the first three years of field establishment (2014-2016) at Citra, FL...... 117

A-1 Segregation analyses of peach Prunus persica x Prunus kansuensis F2 and testcross populations for resistance based on galling index...... 123

A-2 Segregation analyses of peach Prunus persica x Prunus kansuensis F2 and testcross populations for resistance based on egg mass index...... 124

A-3 Segregation of peach Prunus persica x Prunus kansuensis F2 and testcross populations for resistance based on reproduction factor...... 125

B-1 Origins of the microsatellite markers tested for polymorphism and number of markers included in the linkage maps...... 126

B-2 Characteristics of the polymorphic microsatellite markers used to map the Mf resistance locus in peach x Prunus kansuensis interspecific progenies...... 127

B-3 Description of the mapping populations...... 128

B-4 Microsatellite allele configurations of ‘Okinawa’, Prunus kansuensis, and their hybrids inferred from segregation patterns of marker band and electropherogram data in F2 progenies...... 129

B-5 Microsatellite allele configurations of ‘Flordaguard’, Prunus kansuensis, and their hybrids inferred from segregation patterns of marker band and electropherogram data in F2 progenies...... 131

B-6 Microsatellite allele configurations and genotype cross combinations for ‘UFSharp’ x (‘Okinawa’ x Prunus kansuensis) species-level backcrosses inferred from segregation patterns of marker band and electropherogram data in BC1F1 progenies...... 133

B-7 Microsatellite allele configurations and genotype cross combinations for ‘UFSharp’ x (‘Flordaguard’ x Prunus kansuensis) species-level backcrosses inferred from segregation patterns of marker band and electropherogram data in BC1F1 progenies...... 135

B-8 Microsatellite allele configurations and genotype cross combinations for ‘Okinawa’ x (‘Flordaguard’ x Prunus kansuensis) species-level backcrosses inferred from segregation patterns of marker band and electropherogram data in BC1F1 progenies...... 137

B-9 Microsatellite markers used to construct linkage maps and amplification status in each population...... 139

B-10 Chi-square test for segregation distortion in twelve interspecific F2 and BC1F1 populations derived from peach x Prunus kansuensis crosses...... 141

B-11 Combined linkage map based on five F2 mapping populations derived from crosses between ‘Okinawa’ x Prunus kansuensis (OK x PK) compared with Prunus ‘Texas’ x ‘Earlygold’ (TxE) reference map...... 145

B-12 Combined linkage map based on two F2 mapping populations derived from crosses between ‘Flordaguard’ x Prunus kansuensis (FG x PK) compared with Prunus ‘Texas’ x ‘Earlygold’ (TxE) reference map...... 145

B-13 Combined linkage map based on three BC1 mapping populations of peach (‘Flordaguard’ or ‘UFSharp’) x F1 (‘Okinawa’ x Prunus kansuensis) (FG/SH x [OK x PK]) compared with Prunus ‘Texas’ x ‘Earlygold’ (TxE) reference map. . 146

B-14 Combined linkage map based on two BC1 mapping populations of peach (‘UFSharp’) x F1 (‘Flordaguard’ x Prunus kansuensis) (SH x [FG x PK]) compared with Prunus ‘Texas’ x ‘Earlygold’ (TxE) reference map...... 146

C-1 Growth attributes of ‘UFSun’ plants on five different rootstocks before trial establishment...... 153

C-2 Pearson’s correlation coefficients among growth attributes of ‘UFSun’ plants before trial establishment...... 153

C-3 The ANOVA P values for traits considered in evaluating horticultural performance of ‘UFSun’ trees on five different rootstocks during three years of field establishment...... 154

C-4 Means and ranges for yield parameters of ‘UFSun’ peach trees on different rootstocks during the first three years of field establishment (2014-2016) at Citra, FL...... 155

LIST OF FIGURES

Figure page

2-1 Meloidogyne floridensis ‘MFGnv14’ isolate from ‘Flordaguard’ peach rootstock ...... 58

2-2 Scatterplot and Spearman rank correlation data between gall index and indicators of host efficiency, namely: (A) egg mass index and (B) reproduction factor, RF...... 59

2-3 Genetic model for the inheritance of Mf gene in Prunus kansuensis where Mf1 is dominant over mf2 and mf3, and mf2 is dominant over mf3...... 60

3-1 Histograms of the resistance response to a pathogenic isolate of Meloidogyne floridensis (‘MFGnv14’) among twelve interspecific peach x Prunus kansuensis progenies grouped according to the parental cross...... 92

3-2 Comparison of marker distances for linkage group 2 and approximate position of Meloidogyne floridensis resistance locus from combined maps and component maps within each cross type...... 93

4-1 Accumulation of chilling hours between 32 to 45 °F (air temperatures at 2 meters) and growing degree days (GDD, base temperature of 50 °F) during 2015 and 2016 at Citra, Florida ...... 118

4-2 Estimated proportion (%) of ‘UFSun’ fruits harvested at each date relative to total fruits per rootstock throughout the harvest season during second (2015) and third (2016) year of establishment at Citra, FL...... 119

B-1 Microsatellite SSR markers screened for polymorphism and selected for mapping the Mf resistance locus on each of the eight linkage groups (LG1- LG8) of the ‘Texas’ almond x ‘Earlygold’ (TxE) peach reference map...... 147

B-2 Representative gel images of genotyping for microsatellite marker UDP98- 025 in F2 and BC1F1 mapping populations ...... 148

B-3 Combined SSR linkage map based on two F2 mapping populations derived from ‘Flordaguard’ (FG) x Prunus kansuensis (PK)...... 149

B-4 Combined SSR linkage map based on five F2 mapping populations derived from ‘Okinawa’ (OK) x Prunus kansuensis (PK)...... 150

B-5 Combined SSR linkage map based on three BC1 mapping populations derived from peach (‘Flordaguard’ [FG] and ‘UFSharp’ [SH]) x F1 (‘Okinawa’ [OK] x Prunus kansuensis [PK])...... 151

B-6 Combined SSR linkage map based on two BC1 mapping populations derived from peach (‘UFSharp’ [SH]) x F1 (‘Flordaguard’ [FG] x Prunus kansuensis [PK]) ...... 152

C-1 Planting layout for the peach rootstock evaluation...... 156

C-2 Trunk cross-sectional area of ‘UFSun’ trees budded on five different rootstocks during three years (2014-2016) of establishment in the field at Citra, FL...... 157

LIST OF ABBREVIATIONS

ANOVA Analysis of variance

ARR Armillaria root rot

BC backcross bp base pair

°C degrees Celsius cDNA complementary DNA cm centimeter cM centiMorgan cu chill unit

DNA deoxyribonucleic acid

EST expressed sequence tag

°F degrees Fahrenheit

FDP Fruit developmental period

FG ‘Flordaguard’

FL Florida

FP ‘Flordaguard’ x Prunus kansuensis ft feet gal gallon

GA Georgia

GDD growing degree days

GI galling index

KW Kruskal-Wallis lbf pounds·force

LG linkage group

LOD logarithm of odds

LSMEANS least squares means

MAS marker assisted selection

MA Meloidogyne arenaria

MI Meloidogyne incognita

MF Meloidogyne floridensis

MJ Meloidogyne javanica no. number

NPK nitrogen, phosphorous, potassium

OP ‘Okinawa’ x Prunus kansuensis

PCR polymerase chain reaction

PK Prunus kansuensis

PSREU Plant Science Research and Education Unit

PTSL Peach tree short life

QTL quantitative trait locus

RF reproduction factor

RGR relative growth rate

RKN root-knot nematode

SAS Statistical Analysis Software

SD standard deviation

SDL segregation distortion loci

SH ‘UFSharp’

SSR simple sequence repeat

TC trunk circumference

TCSA trunk cross-sectional area

TE Tris-Ethylenediaminetetraacetic acid

TIR-NB-LRR toll/interleukin1-receptor-nucleotide-binding leucine-rich repeat

TxE ‘Texas’ almond x ‘Earlygold’ peach

UF University of Florida

USDA-ARS Unites States Department of Agriculture - Agricultural Research Service

≈ approximately

≥ greater than or equal to

≤ less than or equal to

“ inch

% percent

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

ROOTSTOCK BREEDING FOR RESISTANCE TO THE PEACH ROOT-KNOT NEMATODE (MELOIDOGYNE FLORIDENSIS)

Mary Ann Dy Maquilan

December 2017

Chair: Mercy A. Olmstead Co-chair: José X. Chaparro Major: Horticultural Sciences

The Florida peach industry faces a unique problem with the prevalence of peach root-knot nematode (RKN), Meloidogyne floridensis (MF), which precludes use of common rootstocks ‘Nemaguard’, ‘Guardian’, and ‘Nemared’ with resistance to other major RKN species. ‘Flordaguard’ was selected for resistance to MF and is the only rootstock recommended for commercial peach production in Florida because of its low- chill adaptation and tolerance to nonalkaline soils. Growers must have a wide range of rootstock choices to make peach growing economically viable. This research was conducted to characterize the genetic mechanism of resistance to MF that will influence rootstock breeding strategies and to evaluate potential nematode-resistant rootstocks for commercial utility in Florida.

Twelve F2 and BC1F1 populations from crosses between Prunus kansuensis

Rehder (Kansu peach) and Prunus persica (L.) Batsch (peach) were screened for resistance against a pathogenic isolate of MF. Phenotype and genotype data from interspecific populations revealed P. kansuensis as source of high resistance to MF that is dominantly or recessively inherited. Single marker analyses detected one

microsatellite marker UDP98-025 strongly associated with resistance to MF, corroborating the observed single-locus pattern of inheritance across multiple progenies. Linkage analyses placed the resistance locus at the subtelomeric region of linkage group 2 co-localizing with other previously reported RKN resistance genes in

Prunus.

Two new rootstocks, ‘MP-29’ (peach-plum hybrid) and ‘P-22’ (‘Guardian’ x

‘Flordaguard’ peach hybrid), have the combined resistance to predominant pests/diseases in southeastern US such as RKN, Armillaria root rot, and peach tree short life, thus offering potential for peach production in Florida. Rootstock adaptability and performance were evaluated based on growth, productivity and quality of the peach scion cultivar ‘UFSun’ during first three years of field establishment at Citra, Florida. ‘P-

22’ showed to be a promising alternative to ‘Flordaguard’, although further evaluations should be carried out in other low-chill peach production areas for performance validation. Findings from this research provide new insights that will aid in designing effective breeding strategies to incorporate MF resistance into new rootstocks with improved horticultural features. Having additional resistant rootstocks available to

Florida growers will encourage industry expansion and sustain peach production in the state.

CHAPTER 1 LITERATURE REVIEW

The use of resistant rootstocks has long been viewed as an important management strategy for root-knot nematodes (Meloidogyne spp.) that are damaging to stonefruit crops including peach (Prunus persica [L.] Batsch). This is manifested in the incorporation of root-knot nematode resistance (RKN) as an essential component in breeding and selection of rootstocks for commercial peach cultivation. This chapter describes the progress made in the genetic improvement of peach rootstocks directed towards combining multiple resistance and/or tolerance to pests and diseases (mainly root-knot nematodes) and favorable horticultural features into new rootstocks adapted to the production conditions in the southeastern US. The availability of genetic resources has facilitated the advancement of our understanding of the mechanism of resistance against RKNs in Prunus species and will shape future breeding strategies for developing new rootstocks with improved nematode resistance.

Rootstock Development for Southeastern United States

The earliest rootstocks used in California orchards such as ‘Shalil’, ‘Yunnan’,

‘Bokhara’ and Stribling’s ‘S-37’ were chosen primarily for their resistance to root-knot nematodes (RKN) (Day, 1953). In Florida, the problem of root-galling on peaches and its widespread occurrence had been perceived by the growers since late 1800s as a limiting factor in production (FSHS, 1888). This coincides with the rapid increase in commercial peach production throughout southeastern United States (US) (Rasmussen,

1978). Growers had raised concerns about RKN infestation and the lack of improved varieties that were resistant to RKNs contributing to the decline of the Florida peach industry in 1950s (Sharpe, 1957). Efforts had been made in the stonefruit breeding

program at the University of Florida (UF) to identify and improve resistance in peach rootstocks beginning with the importation of ‘Okinawa’ peach seeds from Japan to

Florida in 1953 and subsequent screening along with old varieties for resistance to the predominant RKN species, Meloidoygne incognita (Kofoid & White) Chitwood,

Meloidoygne arenaria (Neal) Chitwood and M. javanica (Treub) Chitwood (Sharpe,

1957).

‘Okinawa’ and ‘Nemaguard’ (FV 234-1) were the first rootstocks for peach to show high degree of resistance to both M. incognita (MI) and M. javanica (MJ) in

California (Lownsbery et al., 1959) and had also been widely used in Florida during the

1960s to manage such commonly occurring nematode species. Furthermore, these peach seedling rootstocks and a wild related species, P. davidiana (Carr.) Franch., had been used frequently as parents in the early phase of rootstock breeding at the

University of Florida (UF) (Sharpe, 1967). However, the MI- and MJ-resistant hybrids and some advanced selections had been found to be susceptible to an unknown nematode species in the breeding orchard at Gainesville, Florida in 1966. These observations ushered in several decades of conventional breeding for resistance against the new pathogenic species, which was later characterized as Meloidogyne floridensis (MF) (Handoo et al., 2004; Sharpe, 1967; Young and Sherman, 1977). This pathogen has also garnered considerable attention in European breeding programs; they used the Florida peach isolate from ‘Nemaguard’ galled roots in their evaluation of several Prunus germplasm to broaden the spectrum of resistance to RKN (Esmenjaud et al., 1997; Esmenjaud et al., 2009; Lecouls et al., 1997; Rubio-Cabetas et al., 1998).

As older orchards were replaced with new trees, peach tree short life (PTSL) and

Armillaria root rot (ARR) have emerged as the leading cause of premature tree mortality in southeastern US. Peach tree short life, characterized by sudden wilt of young trees during spring and early summer, were affecting the RKN-resistant rootstocks ‘Shalil’,

‘Yunnan’, ‘S-37’, and ‘Nemaguard’, and PTSL orchards were found to be associated with greater incidence of the ring nematode [Mesocriconema xenoplax (Raski) Loof & de Grisse] in Georgia and Carolinas (Nyczepir et al., 1985; Reilly et al., 1986; Ritchie and Clayton, 1981). Several years of selective breeding for combined resistance to

RKN and tolerance to PTSL at the US Department of Agriculture (USDA) stonefruit breeding program in Byron, Georgia led to the release of ‘Guardian’ (BY520-9), which is now the dominant peach rootstock in Georgia and South Carolina albeit lacking in ARR tolerance (Beckman and Pusey, 2001; Okie et al., 1994a). ‘Guardian’ displaced ‘Lovell’ peach rootstock, which had been prevalent in the commercial industry for over 20 years.

‘Lovell’ survives longer on PTSL sites than ‘Nemaguard’ but it is susceptible to RKNs and does not have the same degree of tolerance to PTSL-associated ring nematode and scion longevity as ‘Guardian’ (Nyczepir and Esmenjaud, 2008). Meanwhile, rootstock breeding efforts at UF had been focused on resistance to the endemic RKN species, M. floridensis, leading to the release of ‘Flordaguard’ in 1991 as a standard peach rootstock for low-chill peach production in Florida (Sherman et al., 1991).

New complex Prunus hybrids with promising horticultural features and improved root-knot nematode resistances developed from foreign breeding programs are also being introduced into the US. Some interspecific hybrids such as ‘Jaspi’ and ‘Julior’ from France, ‘Krymsk’ from Russia, ‘Cadaman’ from Hungary as well as advanced

Prunus pumila L. selection ‘Pumiselect’ from Germany are being tested in a wide range of sites in North America. This cooperative effort among researchers would obviate the need for long evaluation times in selective breeding and therefore hasten the release of new rootstocks for stonefruit growers. However, their inconsistent performance under

North American climatic and edaphic conditions have been reported that would limit their widespread utility; specifically, there are issues with regards to their compatibility with peach, host status to ring nematodes, vigor control, or fruit size (Reighard, 2000;

Reighard and Loreti, 2008; Reighard et al., 2011).

At present, breeding work for rootstock improvement is being carried out mainly at the USDA Agricultural Research Service (Byron, Georgia) for most phenotypic traits of interest while screening work for resistance to MF is being conducted at UF

(Gainesville, Florida). There has been emphasis on combining resistances to PTSL,

ARR, and RKNs that are endemic to southeastern peach production areas resulting in

‘MP-29’, a clonal peach-plum interspecific hybrid (Beckman et al., 2012). This rootstock has the additional resistance to MF and a superior level of tolerance to ARR compared to ‘Guardian’ (Beckman and Pusey, 2001; Nyczepir et al., 2006). Furthermore, advanced breeding lines of peach, plum, and peach x plum interspecific hybrids are being screened for disease resistance, vigor, and other horticultural traits for commercial utility across southeastern US (Beckman and Chaparro, 2015). Rootstocks continue to be improved as the industry faces more pressing challenges brought about by phasing out of soil fumigants and replanting old orchard sites.

Sources and Inheritance of Resistance

Sources of resistance to the four major root-knot nematodes (MA, MI, MJ, and

MF) have been identified in the subgenus Amygdalus (peach and almond), subgenus

Prunophora (plum and apricot), and inter- and intra-specific hybrids. Peach seedling varieties introduced in the US such as ‘Shalil’ from northern India, ‘Bokhara’ from

Russia, and ‘Yunnan’ from northern China were the first to be reported resistant. These had resistances to MI and MA, but not to MJ (Day, 1953; Esmenjaud et al., 1994;

Hansen et al., 1956; Lownsbery et al., 1959). Early studies on resistant peach rootstocks ‘Shalil’ and ‘Yunan’ have shown that their resistance to M. incognita was controlled by a single dominant allele (Lownsbery and Thomson, 1959; Weinberger et al., 1943). Later it was confirmed through ‘GF.557’ peach-almond hybrid that the resistance of ‘Shalil’ peach to MI and MA were conferred by a single dominant gene

(RMia) and the same gene could be involved in its resistance to M. hapla but not to MJ

(Claverie et al., 2004a; Esmenjaud et al., 1994).

‘Okinawa’ peach had additional resistance to M. javanica and was used as parent material at the UF stonefruit breeding program for incorporating RKN resistance into new rootstocks (Sharpe, 1957; Sherman and Lyrene, 1983; Young and Sherman,

1977). ‘Nemaguard’, an open-pollinated peach seedling with P. davidiana (Chinese wild peach) in its purported parentage was released by USDA in 1959 and has been a popular rootstock for peaches, plum and nectarine. ‘Nemaguard’ and its derivatives such as the ‘Nemared’ peach (an F3 seedling of ‘Nemaguard’ crossed with a redleaf selection) and the ‘Garfi’ almond x ‘Nemared’ hybrid from Spain share similar rootstock characteristics and resistance to MA, MI, and MJ (Esmenjaud et al., 1997; McKenry,

1989; Ramming and Tanner, 1983). Different modes of inheritance have been proposed for resistance to MI and MJ in ‘Okinawa’ and ‘Nemaguard’. From the F1 hybrids and F2 progenies from crosses with ‘Okinawa’, it was found that a single

dominant allele was involved in the resistance to MI, whereas two independent dominant alleles were acting against MJ (Sharpe et al., 1969). In another study, a two- gene model has been proposed based on segregation patterns of F2 population from intraspecific cross between ‘Nemared’ (homozygous resistant to MI and MJ) and ‘Lovell’

(homozygous susceptible). It was hypothesized that the resistance to MI (Georgia isolate) and MJ in ‘Nemared’ were respectively controlled by two dominant genes (Mi and Mij) and a single dominant gene (Mij). These different findings were attributed to differences in rootstock genotypes from seed-propagated materials, sample size of segregating populations, nematode isolates tested, and evaluation criteria (Lu et al.,

2000b). A source of high resistance to MI has also been identified in P. kansuensis, a wild relative of peach, and the resistance (designated as the PkMi gene) segregated in a manner indicative of a single, dominant gene in the backcross population ([P. kansuensis x ‘Bailey’ peach] x ‘Bailey’ peach) (Cao et al., 2011).

A source of RKN resistance to M. javanica was found in ‘Alnem’ bitter almond

(Prunus dulcis ssp. amara) through recurrent selection and progeny testing in the field.

The observed segregation patterns indicated that a single dominant gene controls resistance to MJ in the almond selections (Kochba and Spiegel-Roy, 1975). In another study using different genotypes of peach and almond, two different genetic systems were hypothesized in the subgenus Amygdalus: one that acts against MA and MI, and another system acting against MJ (Esmenjaud et al., 1997). After evaluating the interspecific progenies Myrobalan x almond, it was later confirmed that this gene, designated RMja, is species-specific to MJ and controls neither MI nor MF (Esmenjaud et al., 2009).

The resistances in ‘Okinawa’, ‘Nemared’, and ‘Nemaguard’ had been found to be overcome by MF, a root-knot nematode species endemic to Florida, and therefore were no longer used as rootstocks for peach production in the state. ‘Flordaguard’ rootstock, an F3 seedling of ‘Okinawa’ with P. davidiana lineage has the additional resistance to

MF and was released in 1991 for use as a standard rootstock for commercial low-chill peach production in RKN-infested, non-alkaline soils. Currently, only the ‘Flordaguard’ rootstock is recommended for stone fruit production in Florida. (Handoo et al., 2004;

Olmstead et al., 2015; Sharpe et al., 1969; Sherman and Lyrene, 1983; Sherman et al.,

1991). Prior to our work, there has been no information regarding the genetic nature of resistance to MF in ‘Flordaguard’.

In Myrobalan plum, host responses ranged from susceptible to highly resistant when tested with MA (Esmenjaud et al., 1992) and certain genotypes were found to be highly resistant to all the Meloidogyne species including MF, which overcomes the resistance in Amygdalus sources. Myrobalan plum is self-incompatible and thus, interspecific hybrids (Myrobalan x Amygdalus) were created to study the genetics of resistance. Resistant clonal accessions were hypothesized to contain homozygous dominant (P.1079), heterozygous (P.2175 and P.2980), recessive (P.2032) alleles for resistance to MA. The tested clones P.2175 and P.1079 were found to be resistant at high temperatures and under high levels of inoculum pressure. The genetic factor designated as Ma gene is responsible for broad-spectrum resistance to predominant

Meloidogyne spp., and minor species such as northern RKN M. hapla and tropical RKN

M. mayaguensis (Esmenjaud et al., 1996a; Esmenjaud et al., 1994; Esmenjaud et al.,

1997; Esmenjaud et al., 2009; Lecouls et al., 1997; Rubio-Cabetas et al., 1999). The

Ma gene from Myrobalan plum is the most well-characterized resistance gene in Prunus and is the second cloned gene for resistance to RKN after the Mi-1 gene from tomato

(Claverie et al., 2011; Milligan et al., 1998).

Significant efforts have been made in selecting for RKN resistance and creating hybrids to broaden the resistance spectrum in European breeding programs. Root-knot nematode resistance has been incorporated into peach, peach-almond hybrid, plum, and other interspecific Prunus rootstocks representing commercial releases and experimental genotypes from the breeding programs in Spain, France, and Italy.

(Felipe, 2009; Fernandez et al., 1994; Moreno, 2004; Pinochet et al., 1999). Although sources of high-level resistance have been identified in plum, physiological incompatibility has been observed in peach-plum combinations. Rootstocks commonly used for plums such as Myrobalan (Prunus cerasifera Ehr.) and Marianna 2624 (P. cerasifera x P. munsoniana) proved to be unsuitable as rootstocks for peaches and nectarines due to incompatibility problems, which are strongly dependent on the rootstock genotype (McKenry, 1989; Moing and Gaudillere, 1992; Salesses and Bonnet,

1992). A Myrobalan plum accession originating from Romania, P.2175, is graft- compatible with many peach cultivars and has high RKN resistance to the four major

RKNs. The peach-almond hybrid ‘GF.557’ developed in Italy is resistant to MI and MA but not to MJ and MF. Breeding work in France focused on the tri-species hybrid progeny P.2175 x (‘Garfi’ x ‘Nemared’)22 with pyramided RKN resistance genes from plum (Ma), peach (Mia) and almond (RMja) to limit the potential risk of resistance breakdown (Esmenjaud et al., 1994; Esmenjaud et al., 1997; Esmenjaud et al., 2009).

Location of Resistance Genes and Associated Markers

Several resistance genes have been identified in Prunus spp. and placed on common linkage groups using different molecular markers. The conserved syntenic regions across different Prunus maps (Dirlewanger et al., 2004b) should help us understand the genomic organization of RKN resistance genes. From ‘Lovell’ x

‘Nemared’ (L x N) intraspecific cross, the segregating F2 population revealed marker banding patterns of EAA/MCAT10 STS (sequence-tagged site) marker that correlated well with the phenotypic response to MI and MJ infection (or linked with nematode resistance loci Mi and Mij in ‘Nemared’) and could distinguish the homozygous resistant, heterozygous resistant, and homozygous susceptible genotypes. Using this marker, the genotypes of 18 peach rootstock genotypes were consistent with their known phenotypes, except for ‘Okinawa’, ‘Flordaguard’, and ‘Yunnan’, which might have different sources of genetic resistance. The two independent loci conferring resistance to MI (Mi locus) and to both MI and MJ (Mij locus) derived from ‘Nemared’ were tightly linked together on an AFLP-based map (Lu et al., 1999). Close linkage between two different resistance loci against MI and MJ on linkage group (LG) 2 was also observed in the ‘Juseitou’ x ‘Akame’ F2 population but were associated with other STS markers, suggesting that the resistance in ‘Juseitou’ might be controlled by genes different from those of ‘Nemared’ (Yamamoto and Hayashi, 2002; Yamamoto et al., 2001). The Mi locus, which controls resistance to MI race 1, was mapped to linkage group (LG) 2 of the Prunus genome in the ‘Harrow Blood’ x ‘Okinawa’ (HB x Oki) F2 population using pchgms1 microsatellite and CAPS (cleaved amplified polymorphic sequence) markers

(Gillen and Bliss, 2005). Comparison of common markers among published linkage maps revealed differences in recombination frequencies among the same markers in

different populations but showed a common placement of the Mi locus onto LG 2 including that of the L x N F2 population initially placed in LG 1 by Lu et al. (1998).

The STS markers OPAP4, OPS14a and OPA11 associated with the new resistance loci (designated Mia and Mja for resistance to MI and MJ, respectively) in

‘Juseitou’ (Yamamoto and Hayashi, 2002) also flanked the RMia gene in the peach sources ‘Nemared’ and ‘Shalil’ (GF.557) (Claverie et al., 2004a). A single dominant gene for resistance to both MI and MA in Nemared, designated RMiaNem, was mapped to the same region on LG2 of the ‘Garfi’ x ‘Nemared’ (GN) map (Dirlewanger et al., 2004a) and Prunus reference map about 10 cM with the SSR marker UDP98-025 (Claverie et al., 2004a; Dirlewanger et al., 2004b). A genetic linkage map has been constructed from the backcross population of RKN-resistant ‘Honggengansutao’ (P. kansuensis) and susceptible ‘Bailey’ (P. persica). From this map, another R gene involved in the resistance against MI (PkMi), localized on the same region with RMia suggesting that this could be an allele of RMia (Cao et al., 2011).

Among the RKN R genes that have been mapped in Prunus species, the Ma gene from Myrobalan plum has the most stable resistance and is not overcome by the four major RKN species, including the “peach RKN” Meloidogyne floridensis

(Esmenjaud et al., 1996a; Handoo et al., 2004). The Ma1 allele from Myrobalan plum accession P.2175 was located on the linkage group 7 at an approximate distance of 2 cM from the SSR marker pchgms6 (Claverie et al., 2004a). It also localized on the same region on the Myrobalan × almond-peach map (Dirlewanger et al., 2004b).

Another wide-spectrum dominant resistance gene in Japanese plum (Prunus salicina) accession J.222, designated Rjap, shared the same map position as Ma and co-

segregated with the SSR markers pchgms6 and CPPCT022 suggesting that location of the two R genes may be conserved in cultivated and wild plum species, including diploid and hexaploid species (Claverie et al., 2004a). Several other reliable markers have been developed for marker-assisted selection for Ma. These include: sequence- characterized amplified region (SCAR) markers, SCAL19 ((Lecouls et al., 1999),

SCAFLP2 (Lecouls et al., 2004) and SCAFLP4 (Claverie et al., 2004b), flanking (<1 cM) or cosegregating with the gene. The Ma locus has been fine-mapped to a 32-kb cluster of three Toll/interleukin receptor-nucleotide binding site-leucine-rich repeat (TIR-NBS-

LRR) candidate genes. It was hypothesized that the PL domains of the Ma gene may be involved in pathogen recognition (Claverie et al., 2011).

Majority of the putative resistance genes belong to the toll/interleukin1 receptor

(TIR) nucleotide-binding site (NBS) leucine-rich repeat (LRR) class and these have been found to contain SSRs. In the peach genome, 60% of the disease-resistance- associated TIR-NBS-LRR genes with post-LRR regions contain SSR sequences in the intron between LRR and post-LRR exons suggesting the potential utility of polymorphic

SSRs for detecting and directly monitoring the introgression of disease resistance genes in breeding programs (Claverie et al., 2011; Lalli et al., 2005; Van Ghelder and

Esmenjaud, 2016). Multiple reported linkages for RKN resistance in the upper part of

LG 2 and LG 7 indicate either clustering of genes or high allelic variability in these regions and, therefore represent useful targets for generating rootstocks with broader spectrum and more durable resistance against RKN.

Host-Plant and Nematode Interaction

The nematode’s life cycle can be divided into four stages: egg, four larval or juvenile stages, and an adult stage. The infective second-stage juveniles (J2s) hatch

from eggs in the soil and migrate toward the roots of the host plants. The J2s enter the root cortex near the elongation zone and migrate intercellularly toward the meristem before turning around into the stele and moving to differentiating vascular cells in the elongation zone. The invading J2s select a few suitable cells and stimulate them to develop into hypertrophied giant cells that acquire multiple nuclei through several cycles of acytokinetic mitosis. Subsequent rounds of endoreduplication generate nuclei with variable sizes and ploidy levels; such process is important for giant cell maturation and enhances expansion of feeding sites from which nematode juveniles draw nutrients and subsequently develop into females capable of producing progeny. Hyperplasy of surrounding cells cause the formation of root-knots or galls typical of infection by

Meloidogyne species (Abad et al., 2009; Engler and Gheysen, 2013; Kyndt et al., 2013).

The female nematode lays eggs that are held together in a protective gelatinous matrix.

Mature female nematodes can lay hundreds of eggs on the root surface, which hatch, in warm moist soil to continue the life cycle. Continued infection of galled tissue by second and later generations of nematodes causes the massive galls. The length of the life cycle depends on the temperature and varies from 4-6 weeks in summer to 10-15 weeks in winter (Williamson and Kumar, 2006).

The pre-parasitic stages of penetration and migration into the roots are facilitated by effector molecules produced in the nematode’s esophageal glands and secreted through their stylet. The secreted effectors also have nuclear localization signals that facilitate trafficking into the host nucleus and enable nematodes to induce the formation of metabolically active feeding structures with which they acquire nutrients. Vascular disruption, i.e. differentiation of root parenchyma cells into enlarged and specialized

feeding cells, impairs root function and can result in disease symptoms in the leaves that resemble water or nutrient deficiency (Bartlem et al., 2014; Grundler and Hofmann,

2011; Quentin et al., 2013).

Plant response to pathogens is based on the type of immune system which uses polymorphic nucleotide-binding site (NBS) and leucine-rich repeat (LRR) proteins encoded by R genes that directly or indirectly recognize the presence or activity of individual effectors encoded by Avr genes. The R-Avr interactions generally lead to highly regulated hypersensitive responses (HR) (Dangl and Jones, 2001; Davis et al.,

2008; Hewezi and Baum, 2013; Husey, 1989). Most of the known R-genes belong to a class of genes containing NBS-LRR domains. The NBS are involved in signaling and the LRR domains are involved in protein-protein interactions. The LRR domains are highly diverse and their structures relate to their role in recognizing specific Avr gene products to promote resistance (Marone et al., 2013). Major R genes involved in root- knot nematode resistance have been mapped and found to co-localize with the genes that code for proteins with NBS-LRR motifs. Among these are: Mi-1 gene from tomato

(Milligan et al., 1998), Ma gene from Myrobalan plum (P. cerasifera Ehr.)(Claverie et al.,

2004a), the RMia gene from peach (P. persica) (Duval et al., 2014), the Rjap gene from

Japanese plum (P. salicina), and the PkMi gene from wild peach (Cao et al., 2014a).

The Prunus RKN R genes vary in their pathogen specificities which could be related to the variations in the length of their post-LRR domains involved in pathogen recognition

(Cao et al., 2014a; Claverie et al., 2011). Durable and broad-spectrum resistance could be achieved by combining multiple R genes (e.g. Ma and RMia) as has been demonstrated in a tri-species hybrid of Prunus (Khallouk et al., 2013).

Resistance in plants to RKN works at several levels due to genetic differences and it can be characterized by reduced penetration, delayed nematode development and lower reproduction (Williamson and Roberts, 2009). Nematode development and/or reproduction can be suppressed or delayed by resistance of the host plant depending on the timing and localization of the hypersensitive response. The Mi-1 gene in tomato that confers resistance against RKN carrying the Cg-1 gene is a well-studied example of an effective immune response. In such incompatible interaction, the nematode juveniles (J2s) can migrate to the target host cells and reactive oxygen species accumulate around the J2s within 12-48 h post-infection. Nitric oxide activity and ROS accumulation in cells along the migratory path of invading J2s up to the feeding site was observed to be higher during an incompatible interaction compared to a compatible interaction. Unlike the tomato-RKN interaction, the resistance response during an incompatible interaction of cowpea with RKN mediated by the Rk gene is delayed such that no HR or ROS accumulation could be observed at 9 days post- infection (Kandoth and Mitchum, 2013). In the resistant coffee ‘UFV408-28, infection by

M. incognita could be blocked right after penetration or during migration and establishment stages, at 6 d, 7 d, and 8 d post-infection (Albuquerue et al., 2010). Post- infection defense mechanisms typically result in a reduced nematode penetration and an inhibited or early degeneration of galls. Different resistant cultivars of cotton and pepper exhibited varying degrees of necrotic responses and giant cell formation. In highly resistant cotton lines, no galls were formed and extensive necrosis could be observed around the penetrating nematodes (Barbary et al., 2014; Mota et al., 2013).

Histological studies on rice and plum showed that the resistance conferred by the R

genes is associated with cell necrosis and hypersensitive-like features occurring in either vascular cylinder, meristematic regions, or the cortex, thereby preventing feeding- site induction and development into third-stage juveniles (Cabasan et al., 2014;

Khallouk et al., 2011; Voisin et al., 1999).

Screening for Resistance

Successful deployment of genetic resistance against RKNs requires proper phenotypic screening of germplasm to identify sources of resistance and to incorporate resistance into adapted genotypes with the combined favorable agronomic features, which can be a lengthy and tedious process. Phenotypic evaluation takes three to four months and will require extensive greenhouse space. Genetic analyses of RKN resistance can also be complicated by the differences in rootstock genotypes, RKN isolates, J2 inoculum concentration, age of plant materials, F2 population size, evaluation criteria (e.g. eggs vs. galls), or definition of resistant versus susceptible.

Developing a method and criteria for evaluating resistance requires an understanding of the nematode species’ behavior in the plant and the plant’s response to the attack at the cellular or tissue level. Current screening procedures are based on galling (plant response to the nematode), which is subjective and does not always correlate with nematode egg production, which is a quantitative indicator of host susceptibility. In some cases, gall formation may occur without egg production, leading to a false diagnosis. In addition, resistant and susceptible rootstocks may show similar intensities of galling, but have variable egg production levels. Thus, an invidivual plant that is considered resistant based on galling intensity may actually contain thousands of eggs. Early genetic studies used only the number of root-knot nematode galls as the main criterion in determining a resistant or susceptible rootstock (Sharpe et al., 1969),

while more recent studies also take into account the number of egg masses formed on

Prunus root systems (Esmenjaud et al., 1997; Lu et al., 2000b). In addition, evaluations of the degree of root-knot nematode infection were based on a single population per root-knot nematode species (Esmenjaud et al., 1997; Marull et al., 1994; Pinochet et al.,

1999) or on naturally occurring RKN species in field soil (Sharpe et al., 1969). It has been reported that different races of Meloidogyne sp. could induce different pathogenic responses in the same host plant (Nyczepir, 1991) and that the age of plant material affects the resistance expression in some resistant Prunus rootstocks (Fernandez et al.,

1995).

The development and availability of molecular markers associated with RKN resistance and a saturated linkage map for Prunus will facilitate the integration of information across different genetic/genomic studies. Microsatellite markers are useful markers for comparative mapping studies because they are reproducible and transferrable across related species (Mnejja et al., 2010). The microsatellite markers can be sourced from the Prunus reference linkage map (Dirlewanger et al., 2004a) and the Genome Database for Rosaceae (http://www.rosaceae.org/peach/genome). The specific parameters of resistance must be defined, as proper identification and determination of “resistance” and “susceptible” classes has important consequences for the validity testing of molecular markers associated with resistance trait.

The Peach Root-knot Nematode Threat in Florida

The peach root-knot nematode M. floridensis (MF) was characterized as a new species based on morphology and unique esterase isozyme pattern (Carneiro et al.,

2000; Handoo et al., 2004). This species overcomes the resistance of most commercial rootstocks such as ‘Nemaguard,’ ‘Okinawa,’ ‘Nemared,’ and ‘Guardian’ (Nyczepir and

Beckman, 2000; Sharpe et al., 1969; Sharpe, 1957; Sherman and Lyrene, 1983). M. floridensis parasitizes most commercial peach rootstocks known to be resistant to MI and MJ, including ‘Okinawa’, ‘Nemared’, ‘Nemaguard’, and ‘Guardian’. The distribution of MF throughout the Southeast is unknown, but it has been detected in seven different

Florida counties (Brito et al., 2008; Brito et al., 2010). It infects peach, watermelon, vegetable crops such as tomato, pepper, tobacco, eggplant, squash as well as several weed species (Handoo et al., 2004; Kokalis-Burelle and Nyczepir, 2004; Rich et al.,

2008; Stanley et al., 2009).

‘Flordaguard’ rootstock, an F3 seedling with ‘Okinawa’ and P. davidiana in its pedigree, has the combined root-knot nematode resistance to MI, MA, and MJ inherited from ‘Okinawa’ and resistance to MF. It is effective as a rootstock for Florida due to its low-chill adaptation, good size control, and ability to withstand infection by MF, an endemic root-knot nematode. (Handoo et al., 2004; Olmstead et al., 2015; Sharpe et al., 1969; Sherman and Lyrene, 1983; Sherman et al., 1991). However, monoculture of a single type of resistance may eventually lead to increase in MF densities or development of a pathogenic, resistance-breaking population (Castagnone-Sereno,

2002; Castagnone-Sereno et al., 2013; Davies and Elling, 2015). Additional rootstock sources must be identified to broaden genetic base for MF resistance, increase available rootstock to peach growers, and preempt future problems of resistance breakdown.

Commercial peach orchards in Florida were visited to examine trees weakened by nematodes. Soil and root samples were collected on-site and some samples were also brought by the growers for nematode assay. Intense galling on the roots was

observed on roughly 5-10% of the trees in commercial orchards. Galling on the roots was usually accompanied by severe stunting of the aboveground tree, defoliation, and reduced foliage resulting in lower yields, smaller-sized fruits, and poor fruit quality.

Observations from these field sites indicated that ‘Flordaguard’ may not be a ‘non-host’ or could be more aptly described as ‘tolerant’. Still, this rootstock may express a certain level of resistance because the nematode galls are smaller than those observed on susceptible rootstocks. There is also a possibility that a heavily galled ‘Flordaguard’ rootstock could be an outcrossed genotype associated with seed-propagated rootstocks, or that it could be a manifestation of resistance breakdown associated with monoculture of a single source of resistance. Although ‘Flordaguard’ rootstock has been described as resistant to RKN including MF, the mechanism of resistance to MF has not been investigated because this pathogen was discovered as a distinct species more recently relative to other RKN species.

Research Objectives

This research project addresses gaps in our knowledge about the genetic mechanism of resistance to the peach root-knot nematode in ‘Flordaguard’ and attempts to identify other sources of resistance to potentially broaden the genetic base in our current rootstocks. These will be achieved by first developing a reliable method for classifying resistance in segregating populations and then studying the mode of inheritance (Chapter 2). The project also aims to identify microsatellite markers associated with MF resistance, determine the genetic location of the resistance locus, and confirm the number of resistance factors involved (Chapter 3). The horticultural performance of new and improved rootstocks developed in Byron, Georgia were tested

under Florida’s subtropical climate to determine their potential as commercial rootstocks for low-chill peach production (Chapter 4).

CHAPTER 2 INHERITANCE OF RESISTANCE TO MELOIDOGYNE FLORIDENSIS IN INTERSPECIFIC PEACH X PRUNUS KANSUENSIS PROGENIES

Introduction

Peach production is continually threatened by plant-parasitic nematodes, predominantly by polyphagous, tropical apomictic Meloidogyne root-knot nematodes

(RKN) M. incognita (MI), M. arenaria (MA), and M. javanica (MJ). These RKN species are known to thrive in warmer temperate, tropical, and subtropical regions where peaches are widely cultivated and their damage can result in poor tree growth, and decline in crop yield or quality (Nyczepir and Esmenjaud, 2008). Their preference for sandy soils and warm subtropical environment make peach production regions in southern Florida prone to nematode infestation.

The use of resistant rootstocks for stone fruits is seen as a viable alternative to chemical control of RKNs, which can cause major economic losses if not effectively managed (Nyczepir, 1991). In the US, peach rootstocks with RKN resistance such as

‘Okinawa’ and ‘Nemaguard’ were introduced in the 1950s and improved breeding selections of these materials were released in the early 80s and 90s, namely:

‘Nemared’, ‘Guardian’, and ‘Flordaguard’ (Okie et al., 1994a; Ramming and Tanner,

1983; Reighard and Loreti, 2008; Sharpe, 1957; Sherman et al., 1991). ‘Flordaguard’ offers a higher level of resistance to RKN than the other sources and has been the only commercially viable rootstock for peach production in Florida due to its low-chill adaptation and additional resistance against the endemic peach root-knot nematode, M. floridensis (MF) (Nyczepir et al., 2006; Olmstead et al., 2015; Sherman et al., 1991).

First detected in 1966 in Gainesville, Florida infecting ‘Okinawa’ and

‘Nemaguard’, MF was initially identified as M. incognita race 3 (Sharpe, 1967; Sherman

and Lyrene, 1983), but was later characterized as a new species (Handoo et al., 2004).

This pathogen is of scientific interest because it overcomes the resistances of commonly used peach rootstocks such as ‘Okinawa’, ‘Nemaguard’, ‘Nemared’ and

‘Guardian’ (Esmenjaud et al., 2009; Handoo et al., 2004; Lecouls et al., 1997; Nyczepir and Beckman, 2000). The species has a wide host range and its distribution is yet unknown. After it was fully described and with improved techniques for accurate identification of the species, more populations of MF have been identified on ornamental, vegetable, herb, and weed species in several counties in Florida (Brito et al., 2008; Brito et al., 2010; Church, 2005; Kaur et al., 2007; Kokalis-Burelle and

Nyczepir, 2004). Greenhouse tests also determined it to be capable of reproducing on tomato and cowpea carrying the nematode resistance genes Mi and Rk, respectively

(Brito et al., 2015).

The potential of nematodes to develop virulent, resistance-breaking pathotypes as a consequence of selection pressure from monoculture of resistant plant cultivars has been extensively reviewed (Castagnone-Sereno, 2002; Castagnone-Sereno et al.,

2013; Davies and Elling, 2015), and so there is concern for an imminent breakdown of plant resistance by the nematode pathogens from prolonged plantings of a single source of resistance such as ‘Flordaguard’ rootstock for peach production systems in

Florida. Given that the MF is a facultative meiotic parthenogen capable of sexual reproduction under conditions of stress and crowding, genetic variation arising from cross-fertilization would facilitate a rapid adaptive response to host-plant resistance

(Castagnone-Sereno, 2006; Handoo et al., 2004). Indeed, during our initial work to obtain a MF isolate for this study, we detected a pathogenic population of MF at an

experimental field in Gainesville, Florida where breeding lines on ‘Flordaguard’ peach rootstock have been successively planted for over 20 years. The problem of a nematode population breaking the resistance of a standard rootstock has again come to the fore, hence requiring an urgent effort to develop improved rootstocks that are also effective against pathogenic populations of MF. It is imperative to seek new sources and understand the underlying mechanism of resistance against MF in order to enhance breeding strategies. Having a diverse panel of rootstocks available to peach growers will preempt future problems of resistance breakdown and help keep the industry functional.

In the genus Prunus, major RKN R genes have been identified in: Myrobalan plum (Ma) and Japanese plum (Rjap) (Claverie et al., 2004a); peach cultivars

‘Nemared’, ‘Nemaguard’, ‘Shalil’ and ‘Juseitou’ (RMia) (Claverie et al., 2004a; Duval et al., 2014; Esmenjaud et al., 2009; Yamamoto and Hayashi, 2002); almond ‘Alnem’

(RMja) (Van Ghelder et al., 2010); and more recently, wild peach P. kansuensis (PkMi)

(Cao et al., 2011). Resistance found in the subgenus Amygdalus (peach and almond) are limited in spectrum and none of the domesticated sources are resistant against MF

(Handoo et al., 2004) whereas in the subgenus Prunophora, accessions of Japanese plum and Myrobalan plum have shown broad and high level of resistance to RKN including MF (Esmenjaud et al., 1997; Lecouls et al., 1997; Rubio-Cabetas et al., 1998).

The ease with which domesticated peach can be hybridized with other cultivars and wild species to improve disease resistances has greatly benefitted breeding programs in the US and Europe (Beckman and Chaparro, 2015; Esmenjaud et al.,

2009; Felipe, 2009; Pinochet et al., 1999). Interspecific hybrids have been used as

rootstocks for peach to create genetic variability that is otherwise limited in highly self- compatible peach (Akagi et al., 2016; Scorza and Okie, 1991; Verde et al., 2013). The commonly used nematode-resistant rootstocks in the US such as ‘Okinawa’,

‘Nemaguard’, ‘Nemared’, ‘Guardian’, and ‘Flordaguard’ were derived from parent material limited to P. persica and P. davidiana; thus, it is of interest to develop new rootstocks that inherit their resistance from other Prunus species. The Myrobalan plum

(P. cerasifera), an outcrossing diploid species, has been incorporated into European breeding programs and some of the clonal selections and interspecific hybrid materials

(plum x [peach-almond], containing R genes Ma, RMia, and RMja have the combined beneficial agronomic features as well as broad, durable resistance to root-knot nematodes including MF (Khallouk et al., 2013; Salesses et al., 1998). Despite the availability of such a rootstock material and the potential for introgression of the dominant Ma gene from P. cerasifera into our commercial peach rootstocks to broaden the resistances, European plum rootstocks have not been successfully utilized in

Southeastern US breeding programs because of their susceptibility to peach tree short life, Armillaria root rot, and plum leaf scald (Beckman et al., 1998; Beckman and Pusey,

2001; Okie et al., 1994b; Reighard, 2000). Additionally, incompatiblities have been reported for most peach and plum (P. cerasifera and P. salicina) graft combinations

(Moreno et al., 1993; Salesses and Bonnet, 1992; Zarrouk et al., 2006).

Sources of disease resistance genes can be found in wild species and crosses can be made to enhance the resistance in commercial peach rootstocks while maintaining genetic diversity in the breeding pool. Crossing peach with P. kansuensis would produce successful hybrids because of their genetic relatedness (Cao et al.,

2014b). Originating from northwest China where it is traditionally used as seedling rootstock for almond and peach (Reighard and Loreti, 2008; Scorza and Okie, 1991), P. kansuensis has been an important source of genetic diversity at the University of Florida

(UF) stonefruit breeding program (Chavez et al., 2014). Chaparro et al. (2009) evaluated the usefulness of P. kansuensis as a parent for genetic studies in peach.

Apart from the high polymorphism observed based on microsatellite markers (≈80% expected heterozygosity), crosses of P. kansuensis with ‘Okinawa’ or ‘Flordaguard’ have produced vigorous and highly fertile F1 hybrids, which would allow for screening large numbers of F2 progeny. Observations of several F1 interspecific hybrid populations planted in a nematode-infested site in Gainesville, Florida led to the speculation that P. kansuensis may contain a resistance gene against MF that segregates in the F2 progeny (J. X. Chaparro, pers. comm.). It is quite conceivable that the RKN resistance could come from P. kansuensis as it has been reported to be resistant to MI and that the resistance segregated in the backcross population ([P. kansuensis x ‘Bailey’ peach] x ‘Bailey’ peach) in a manner indicative of a single, dominant gene (Cao et al., 2011).

In the present study, the F2 progenies, derived from crosses of ‘Okinawa’ peach and ‘Flordaguard’ peach with P. kansuensis, were artificially inoculated with a high inoculum pressure of a pathogenic isolate of MF. We analyzed the segregation of resistance in the F2 progenies and testcrosses to determine the possible mode of inheritance for resistance to MF. Confirmation of MF resistance from P. kansuensis would open possibilities for broadening the resistance to RKN in our current rootstocks.

Materials and Methods

Nematode Isolate

We isolated a pathogenic population of root-knot nematode that was found causing severe galling and reproducing on ‘Flordaguard’ rootstock (Figure 2-1A). The species’ identification was confirmed to be M. floridensis based on the esterase and malate dehydrogenase profiles attained from single females via polyacrylamide gel electrophoresis (PAGE) (Handoo et al., 2004). Intra-population heterogeneity has been reported in Meloidogyne spp. (Dalmasso and Berge, 1978) and, therefore a single egg mass culture was obtained from the field population to ensure purity of the isolate. The population was multiplied on susceptible tomato (Solanum lycopersicum L.) ‘Rutgers’ over two successive generations. Six months from culture initiation, females were dissected arbitrarily from individual root galls. The enzyme phenotypes of 20 females were re-examined to verify purity of population before starting the experiment (Figure 2-

1B). Second-stage juveniles (J2s) were collected from infected root systems of tomato plants using 1% sodium hypochlorite extraction method (Hussey and Barker, 1973).

The pathogenic isolate collected from ‘Flordaguard’ peach at Gainesville, Florida in the spring of 2014 is hereinafter referred to as ‘MFGnv14’.

Genetic Material

Thirteen open-pollinated F2 and backcross populations were used in the segregation analyses. These populations originated from crosses of the Prunus persica

(L.) Batsch (peach) with Prunus kansuensis Rehder (Kansu peach) developed in 2005 by J. X. Chaparro for the UF stonefruit breeding program. The parental genotypes: P. kansuensis and P. persica cultivars ‘Okinawa’, ‘Flordaguard’, and ‘UFSharp’ represent different levels of resistance to M. floridensis (MF) and were crossed to generate

segregating populations. The response of P. kansuensis to M. floridensis has not yet been determined, although it has been previously found to be highly resistant (gall index

= 0) to M. incognita (Cao et al., 2011). ‘Okinawa’ peach is resistant to M. incognita and

M. javanica but susceptible to MF (Gillen and Bliss, 2005; Sharpe et al., 1969; Sherman and Lyrene, 1983); under artificial inoculation in the greenhouse, it showed susceptibility to the ‘MFGnv14’ isolate (D. Dickson, pers. comm.). ‘Flordaguard’ peach, the standard rootstock cultivar for Florida peach production, was previously reported to have resistance to MF (Nyczepir et al., 2006; Sherman et al., 1991). However, in the present study, ‘Flordaguard’ was designated as homozygous-susceptible because the

‘MFGnv14’ isolate is capable of infecting and reproducing on the rootstock. ‘UFSharp’, a peach scion cultivar developed in the UF stonefruit breeding program, sets a good number of flower buds (Chaparro and Sherman, 2006) and is known to be susceptible to all major RKNs including MF (J. X. Chaparro pers. comm.) making it a suitable female parent for the testcrosses.

Eight F2 populations were obtained from open-pollination of interspecific hybrid accessions: six ‘Okinawa’ (OK) peach x P. kansuensis (PK) F1 hybrids and two

‘Flordaguard’ (FG) peach x PK hybrids. Two OK x PK hybrids and two FG x PK hybrids were backcrossed to a P. persica cultivar, either ‘UFSharp’ (SH) or FG, as the female parent. Pollinations were conducted in the spring of 2014 in Gainesville, FL and in

Citra, FL (Table 2-1).

Growth Conditions

Seeds were extracted from the pit, air-dried, and stratified on moist perlite for two to three months at 5-8°C. The germinated seeds were planted a week after radicle emergence onto seedling trays using coarse vermiculite as substrate for root

establishment. Seedlings were repotted after two months into 4.5” x 4.5” square polythene pots containing 50:50 (v/v) steam-pasteurized sand-vermiculite mix. The sand-vermiculite medium has proven satisfactory for nematode infection and reproduction under greenhouse conditions, thus allowing separation of susceptible from resistant genotypes as previously reported (Lu et al., 2000a; Nyczepir et al., 1999) and used in a number of inheritance studies and host status tests (Lu et al., 2000b; Nyczepir and Beckman, 2000; Nyczepir et al., 2008). The potted plants were moved to an environmentally-controlled greenhouse at the UF Plant Science Research and

Education Unit (Citra, FL). Individual genotypes were distributed on the greenhouse benches regardless of their parental origin. Day-night air temperatures in the greenhouse were adjusted to 26-30°C and relative humidity to 55-60%. Plants were watered through microtubes running to each pot and fertilized with Osmocote® 15-9-12 as needed.

Resistance Evaluation

Four-month old plants were inoculated with the MF isolate at a concentration of

10,000 eggs per pot. Into each pot, MF eggs were deposited by pipetting 5 ml of the egg suspension (containing 500 eggs·mL-1) two inches deep into each of the four holes made around the stem base and then pressing the media into the pre-punched holes.

Three-week old tomato plants were also inoculated to check nematode viability. Four to five months after inoculation, the plants were harvested and the root systems were processed by immersing in water and gently shaking to free the roots of adhering sand and vermiculite. Clean root systems were stained using a red food coloring dye to enhance visibility of the translucent egg masses (Thies et al., 2002). The galls and stained egg masses were viewed under a magnifying lamp and counted. Following

estimation of numbers of root galls and egg masses, the eggs were extracted from the whole root system using a 1% sodium hypochlorite solution (Hussey and Barker, 1973).

The mean number of eggs per root system were estimated by suspending them in tap water adjusted to 50 ml volume. Using a hematocytometer to count egg density per 1 ml of egg suspension, the counts were averaged from three 1 ml-aliquots and multiplied by

50 to give an estimate of the number of eggs per root system.

The number of galls and egg masses per root system were converted to ordinal data based on a 0-5 rating scale (Taylor and Sasser, 1978) as presented in Table 2-2.

Nematode reproduction factor (RF = Pf/Pi) was also calculated to assess nematode reproduction in the host, where Pi = 10,000 eggs or J2s during inoculation and Pf = final numbers of nematode eggs and (or) J2s. For each segregating population from the interspecific crosses, plant genotypes were separated into different resistance levels based on gall indices, egg mass indices, and RF; the resultant ratios from these different parameters were compared to determine the optimal criterion for detecting resistance and an acceptable threshold that would properly classify genotypes as either resistant or susceptible.

Data Analyses

Data used for segregation analyses were from individuals verified as true progeny of the F1 hybrids and backcrosses of a specific female-male combination based on individual allele profiles of 32-36 polymorphic microsatellite markers distributed across the genome. About 27% were culled out because they were found to be outcrossed genotypes, with one or more discrepant microsatellite alleles. Hence, the analyzed data set consisted of 379 individuals from thirteen F2 and testcross families.

For each family, observed ratios of resistant and susceptible phenotypes were tested

against expected ratios for a single-gene model using chi-square test. Correlation analyses between gall index and nematode reproduction indicators such as egg mass index and RF were based on Spearman’s correlation for ordinal data. All analyses were performed using SAS (version 9.4; SAS Institute, Cary, North Carolina, USA).

Results

Selection Criteria for Resistance

Correlation analyses of the three rating systems, namely: gall index (GI), egg mass index (EI), and level of reproduction factor (RF), suggested that GI based on number of galls per whole root system is a reliable indicator of plant resistance. Very high correlation coefficients were obtained between GI and EI (r = 0.966) as well as RF level (r = 0.886). When the disease scores were plotted on a scatterplot (Figure 2-2), it became evident that plants with a GI ≤ 2 tend to have very low reproduction levels with

RF values ranging from 0 to 0.1. At GI = 2 there were one to five egg masses detected on five plants but the RF values were below 0.1.

Out of the 193 plants with GI above 2, 51 had GI values higher than EI indicating that not all detected galls contained egg masses (e.g. incipient galls) or that the egg masses may be small enough to go undetected. At GI > 2, the RF values were beyond the “non-host” threshold of 0.1 (Sasser et al., 1984) except for 17 plants where spuriously low egg counts were recorded. Low egg counts could be attributed to either poor egg recovery or prevalence of incipient galls without egg masses and egg masses with smaller diameters (fewer eggs). The eggs could have been heterogeneously distributed in the diluted suspension and sampling only three 1-mL aliquots might have resulted in further underestimation of egg numbers.

The high correlation of GI with RF indicates that gall counts could be used reliably to assess resistance and we set our breakpoint at GI = 2 because, at this level, the RF values were within the range of 0 to 0.1, consistent with the classical definition of resistance as: the ability of the host to severely restrict or prevent reproduction of the pathogen (Trudgill, 1991; Williamson and Hussey, 1996). Therefore, the 379 individuals from thirteen interspecific families were classified as either “resistant” if there were less than 10 galls or egg masses, or “susceptible” if there were more than 10 galls or egg masses. Under such criteria, 186 out of 379 individuals were categorized as resistant and 93% of the resistant individuals had no detectable galls and egg masses.

Segregation Analyses of P. persica x P. kansuensis F2 Families

The ‘MFGnv14’ isolate found infecting ‘Flordaguard’ peach roots was used in this study to determine the genetic basis of resistance to MF. Therefore, the genotype status of ‘Flordaguard’ (FG), like ‘Okinawa’ (OK), was presumed to be homozygous recessive for this specific isolate. The genotype status of P. kansuensis (PK) was unknown and its genotype was deduced from the segregation patterns in the F2 progeny. The segregation of resistance in the F2 progenies of eight crosses between PK and P. persica cultivars OK or FG in our greenhouse test are presented in Table 2-3.

The F2 progenies derived from the four OK x PK hybrids consisted mostly of resistant phenotypes, all supporting a segregation ratio of 3R:1S (P = 0.238 to 0.662) based on chi-square test for a single-gene model. Two F2 progenies derived from FG x PK hybrids also did not deviate from a 3R:1S ratio (P = 0.564 to 0.915).

Contrarily, two other hybrids ([OK x PK]2 and [OK x PK]3) revealed an excess of susceptible phenotypes, significantly deviating from the 3R:1S ratio. This unexpected ratio is possibly caused by either chance deviation or incorrect categorization because

of inconspicuous galls. Most of the individuals scored as susceptible had small swellings or bumps on the roots that could not be easily differentiated from knobs of lateral root primordia in contrast to the pronounced galling that could be easily differentiated in susceptible individuals of the other F2 progenies. The ambiguous phenotypes observed in the two progenies may also indicate incomplete resistance. An alternative hypothesis is that the discrepancies in the observed numbers of susceptible plants might have resulted from the lack of a dominant resistance allele in the two F1 hybrids and the presence of two variably expressed recessive alleles, one from peach and the other from P. kansuensis, which may explain the occurrence of a few resistant individuals. Growing out and testing more F2 plants would allow us to confirm the true proportion of susceptible individuals. Alternatively, a test cross could be performed to confirm the genotypes of the two F1 hybrids.

Testcross of F1 Hybrids with P. persica

To provide additional evidence for a single-gene mode of inheritance and to confirm the allelic state of the parents, we crossed the F1 hybrids with P. persica cultivars SH and FG. Most of the testcrosses were conducted using SH as the female parent because of its good vigor and prolific flowering. This scion cultivar is known to be susceptible to MF and, therefore presumed to be a homozygous recessive genotype similar to FG. The other homozygous recessive parent of F1, OK peach, could not be used because it was showing tree decline in the experimental field at Gainesville,

Florida. We attempted to do backcrosses with a single OK tree at USDA-ARS germplasm repository in Parlier, California (with the assistance of Dr. Craig Ledbetter) but most of the emasculated flowers were aborted due to the cold spells, resulting in

very low fruit counts. Only the five testcrosses that provided sufficient sample sizes were analyzed.

Based on chi-square tests, the observed segregation in three out of five testcross progenies did not deviate from a 1:1 ratio (P = 0.343 to 0.640), confirming our hypothesis based upon observations in the F2 that resistance is controlled by a single locus. In two other testcross progenies (SH x [OK x PK]3 and FG x [ OK x PK]3) that shared a common male parent [OK x PK]3, most, if not all, susceptible phenotypes were observed. The prevalence of susceptible types in both F2 and testcross progenies confirmed that the [OK x PK]3 F1 hybrid lacks the dominant resistance allele. Notably, there were seven resistant out of 35 individuals in the [OK x PK]3 F2 progeny fitting a 1R:

3S (P = 0.494) ratio. The 1R: 3S ratio could be simply explained by a recessive allele from P. kansuensis inherited by the [OK x PK]3 F1 parent that may have contributed to the resistance. While the segregation analysis of [OK x PK]2 F2 progeny revealed an excessive proportion of susceptible phenotypes, the observed segregation in the testcross progeny conformed to the expected 1:1 ratio with discrete phenotypes, thus confirming the heterozygosity of the [OK x PK]2 F1 parent and indicating that the deviation from the expected 3R:1S ratio in the F2 progeny could have been due to incorrectly categorized phenotypes.

Discussion

In this study, genetic analyses were successfully conducted due to the availability of a pathogenic isolate ‘MFGnv14’ and the fortuitous segregation for strong resistance in progenies from crosses of peach with P. kansuensis, in which the level of resistance to MF was initially unknown. Prior to this work, we attempted to screen the interspecific populations for resistance using a different MF isolate. However, the isolate was most

likely non-pathogenic on Prunus material because no disease symptoms were observed in the peach progenies four months after inoculation, except for the control tomatoes

(data not shown). This prompted the search for a pathogenic isolate and, subsequently, we established pure cultures of the new isolate (‘MFGnv14’) that was found causing extensive galling on ‘Flordaguard’ in the field. It was unclear whether the non- pathogenicity of the previous nematode isolate (accession N05-227-17B sourced from the UF Entomology and Nematology Department) to our Prunus materials was developed during prolonged culture on a susceptible tomato cultivar or the isolate was inherently non-pathogenic on Prunus because it was originally obtained in the field from an infected tomato plant (Smith et al., 2015), although it has been demonstrated that other MF populations resident in tomato or cucumber could also reproduce on peach cultivars such as ‘Nemaguard’ and ‘Lovell’ (Stanley et al., 2009).

The nematode population on ‘Nemaguard’ roots, initially identified as M. incognita race 3 (Sherman and Lyrene, 1983) and later characterized as a new species,

M. floridensis (Handoo et al., 2004), originated in the same breeding orchard at

Gainesville, Florida where the pathogenic ‘MFGnv14’ isolate was collected. The characterized MF isolate, labeled ‘Floride’ or ‘Gainesville’ by researchers at INRA

(Institut National de la Recherche Agronomique, France) was tested on Prunus rootstocks/accessions including some P. cerasifera accessions such as P. 2980,

P.2175, and P.1079 that showed complete resistance to the isolate (Esmenjaud et al.,

1997; Esmenjaud et al., 2009; Lecouls et al., 1997; Rubio-Cabetas et al., 1998). The same MF population from ‘Nemaguard’ was used in a peach rootstock microplot trial in

Byron, GA where it reproduced on ‘Nemaguard’, ‘Guardian’, and to a much lesser

degree on ‘Flordaguard’; thus, ‘Flordaguard’ was considered resistant/tolerant to MF

(Nyczepir et al., 2006).

Since its release as a public variety in 1991, ‘Flordaguard’ has been the only rootstock that could be suitably used by peach growers in Florida and in other non- alkaline soils infested with MF (Olmstead et al., 2015; Sherman et al., 1991). Field- detected pathogenicity of ‘MFGnv14’ on ‘Flordaguard’ rootstock could be perceived as a consequence of selection due to successive plantings of the resistant rootstock for over

20 years. Analogous cases of pathogenic field populations of MI infecting plants carrying R genes have been reported for tomato carrying the Mi gene (Eddaoudi et al.,

1997; Kaloshian et al., 1996) and cowpea carrying the Rk gene (Roberts et al., 1995).

Artificial selection experiments have demonstrated the possibility for RKN populations to shift from avirulence to virulence when cultured on resistant host as in the case of the mitotic parthenogen MI (Castagnone-Sereno et al., 1994; Jarquin-Barbarena et al.,

1991) and the meiotic parthenogen M. chitwoodi (Janssen et al., 1998). The present study attempted to find possible sources of resistance effective against the resistance- breaking isolate ‘MFGnv14’ while seeking to understand the genetic nature of resistance to MF using interspecific populations. To avoid progressive loss or decline of pathogenicity during culture on susceptible tomato (Jarquin-Barbarena et al., 1991;

Petrillo et al., 2006), the inoculum was used after two cycles of propagation on the tomato host.

The progenies segregated for resistance when the plants were inoculated with the ‘MFGnv14’ isolate at a high inoculum density (10,000 eggs or J2s per plant) and the observed phenotypes after four to five months ranged from completely resistant to

highly susceptible. Three different parameters, namely: gall index, egg mass index, and

RF, characterized the variation observed in the F2 and testcross populations. Since the nematode’s reproductive capacity in a host is directly related to resistance, the RF was used as a basis to determine the resistance threshold for galling index. A final egg count that is ten-fold lower (RF ≤ 0.1) than the initial nematode inoculum density indicates strong expression of resistance in the plant host by severely restricting nematode establishment and reproduction, thereby considering the plant a “non-host”

(Sasser et al., 1984; Williamson and Hussey, 1996). The progenies were divided into resistant and susceptible classes using a cutoff value of 2 on GI (or 10 galls per whole root system), because at this level, all plants had a non-host level of RF (≤ 0.1). From the 186 plants classified as resistant using the cutoff GI = 2, 173 displayed complete resistance (no galls and egg masses). Very similar segregation ratios were obtained when the progenies were categorized based on egg mass index (cutoff value = 2) and

RF (cutoff value = 0.1) (Supplementary Tables A-1 to A-3). These results show that resistance screening can be performed based on gall numbers that could be assessed directly without the need for additional steps (e.g. root staining to detect egg masses on the galls, macerating and bleaching the roots to extract the eggs, and counting eggs under the microscope), which are tedious and time-consuming besides being prone to human error. This is in line with a previous report by Esmenjaud et al. (1992) that gall index for MA in Myrobalan plum is highly correlated with eggs and later stages of nematode in the roots; therefore, in subsequent RKN resistance screening studies, gall index was used as a criterion to characterize the Ma gene in Myrobalan plum

(Esmenjaud et al., 1996b; Lecouls et al., 1997; Rubio-Cabetas et al., 1998), Rjap gene in

Japanese plum (Claverie et al., 2004a), RMia in ‘Nemared’ and ‘Shalil’ peach (Claverie et al., 2004a; Esmenjaud, 2009) and RMja gene in ‘Alnem’ almond (Van Ghelder et al.,

2010). In those studies, however, root galling was visually judged (i.e. rating system based on percentage of root system galled), and determination of resistance likely becomes subjective when intermediate phenotypes are observed. Other studies used gall counts for inheritance studies or determination of host status in ‘Lovell’,

‘Nemaguard’, ‘Guardian’, and ‘Nemared’ peach rootstocks (Lu et al., 2000b; Nyczepir and Beckman, 2000; Stanley et al., 2009) and these provided baseline information for defining resistance since absolute values can be compared. In this study, setting the resistance limit to ten galls per plant allowed the objective classification of resistance and was sufficient for determining the mode of inheritance of resistance to MF in the peach x P. kansuensis progenies.

The inheritance of resistance to MF appeared to involve a single locus as observed in the F2 and testcross progenies, and that the resistance was contributed by the P. kansuensis parent. The F2 progenies, derived from OK x PK and FG x PK crosses, displayed two different segregation patterns. Six of the eight F2 progenies ([OK x PK]1, [OK x PK]4, [OK x PK]5, [OK x PK]6, [FG x PK]1, and [FG x PK]6) segregated into

3R:1S, conforming to the expected ratio for single-gene inheritance. The F2 progenies of the other two hybrids ([OK x PK]2 and [OK x PK]3) had predominantly susceptible types suggesting that the genotypes of these hybrids could be different. Because OK and FG were known to be susceptible to MF, these two peach parents were presumed to be homozygous recessive. Under a single-gene model, the two different segregation patterns in the F2 progeny could arise only if the other parent, P. kansuensis, is

heterozygous and one allele is involved in the dominantly inherited resistance observed in [OK x PK]1, [OK x PK]4, [OK x PK]5, [OK x PK]6, [FG x PK]1, and [FG x PK]6 F2 progenies while the other allele is involved in the recessively inherited resistance observed in [OK x PK]3 F2 progeny.

The hypothesized heterozygous genotype of our P. kansuensis accession was validated in the test crosses of the OK x PK and FG x PK hybrids with the susceptible genotype SH or FG where the progenies from SH x [OK x PK]2, SH x [FG x PK]1, and

SH x [FG x PK]6 segregated into 1R:1S. The other two progenies (SH x [OK x PK]3 and

FG x [OK x PK]3) that shared a common male parent [OK x PK]3, again had predominantly susceptible types, indicating that [OK x PK]3 lacks the dominant resistance allele; both testcross progenies were expected to show no segregation for resistance because of the lack of a resistance allele in the other parent, SH or FG.

In the F2 progenies of [OK x PK]2 and [OK x PK]3, there were inconspicuous galls

(i.e. incipient/small galls with or without egg masses that were hardly distinguishable from knobs of lateral root primordia) found on several root systems, which may have caused the distorted segregation in favor of susceptible types. This can also be interpreted to indicate some form of partial resistance. Nonetheless, the genotypes of the two hybrids were resolved from the testcross progenies where the symptomless resistant types could be clearly differentiated from the heavily-galled, susceptible types.

Thus, [OK x PK]2 was presumed to be heterozygous because the testcross progeny segregated 1R:1S whereas [OK x PK]3 was homozygous recessive because of the predominantly susceptible types in the testcross progeny.

The observed 3R:1S segregation ratios in six F2 families suggested that resistance to MF was inherited as a single, dominant allele from the P. kansuensis parent. The observed testcross segregation ratios of 1R:1S in the three testcrosses further strengthen our hypothesis of a single major gene controlling resistance to MF

(hereinafter designated as the Mf gene). The Mf gene could be traced back to P. kansuensis in heterozygous state and, although P. kansuensis was not tested directly due to unavailability of seeds, it is expected to be completely resistant to MF whether in homozygous dominant or heterozygous allelic state. In P. cerasifera, the heterozygous accession, P.2175, exhibited similar levels and spectrum of resistance to major RKN species as the homozygous-resistant accession, P.1079 (Lecouls et al., 1997).

An accession of P. kansuensis in China was reported to also contain a R gene

(PkMi) that confers immunity to MI. The PkMi gene was hypothesized to be an allele of

RMia which controls resistance to MI and MA in peach sources (Cao et al., 2011), raising the possibility that P. kansuensis may have a broad spectrum of resistance similar to that observed for the Ma gene in P. cerasifera. Further investigation on the resistance spectrum of P. kansuensis is warranted considering the observed infestation of other RKN species in some commercial peach orchards in Florida. Moreover, one would expect within the MF species a large pathogenic and genetic variability that have been found to occur in other meiotic parthenogenetic Meloidogyne spp. such as M. chitwoodi, M. fallax, M. hapla, M. exigua, and M. graminicola (Bellafiore et al., 2015;

Janssen et al., 1997; Muniz et al., 2008; Van der Beek et al., 1998). Thus, it is important to evaluate resistance to MF against different isolates to determine if the Mf gene in P. kansuensis is isolate-specific or confers resistance to multiple nematode

species and isolates. Furthermore, testing P. kansuensis with other RKN species would allow us to confirm if resistance to MF segregates in the same way as resistance to MI, which will shed light into the relationship between the Mf gene in this study and the

PkMi gene identified by Cao et al. (2011).

Our interpretation regarding the mode of inheritance of Mf gene from P. kansuensis based on phenotypic segregation patterns in seven F2 progenies and five testcross progenies of peach x P. kansuensis is illustrated in Figure 2-3. The genotypes of our P. kansuensis accession and the F1 hybrid parents were deduced from the segregation ratios in F2 and testcross progenies. The observed phenotype segregation patterns indicate that the Mf gene in P. kansuensis segregates for two resistance alleles with resistance effects behaving in either a dominant or recessive manner.

Table 2-1. Interspecific peach x Prunus kansuensis F2 and testcross progenies used in the segregation analyses for resistance to Meloidogyne floridensis. Pollinations were conducted in the spring of 2014. Population a Pollination site No. of individuals analyzed

F2: [OK x PK]1 Gainesville, FL 36 [OK x PK]2 Gainesville, FL 34 [OK x PK]3 Gainesville, FL 35 [OK x PK]4 Gainesville, FL 29 [OK x PK]5 Gainesville, FL 22 [OK x PK]6 Gainesville, FL 28 [FG x PK]1 Gainesville, FL 29 [FG x PK]6 Gainesville, FL 25

Testcrosses (species-level backcrosses): FG x [OK x PK]3 Gainesville, FL 33 SH x [OK x PK]2 Citra, FL 32 SH x [OK x PK]3 Citra, FL 20 SH x [FG x PK]1 Citra, FL 18 SH x [FG x PK]6 Citra, FL 40 a Parent materials: Prunus persica cultivars ‘Flordaguard’ (FG), ‘Okinawa’ (OK), and ‘UFSharp’ (SH); wild peach P. kansuensis (PK).

Table 2-2. Host resistance classification system based on the presence of galls and egg masses on roots. Galling and Number of galls Phenotype classification egg mass indices a or egg masses 0 0 Resistant 1 1-2 Resistant 2 3-10 Resistant 3 11-30 Susceptible 4 31-100 Susceptible 5 100+ Susceptible a Rating system according to Taylor and Sasser (1978).

Table 2-3. Segregation for resistance to Meloidogyne floridensis in Prunus persica x Prunus kansuensis F2 and testcross progenies based on gall index and chi- square tests of segregation data under a single-gene model. Seedling progeny a Resistant Susceptible Test ratio Calculated Χ2 P c (R) b (S) b R:S F2:

[OK x PK]1 25 11 3:1 0.59 0.441 [OK x PK]2 4 30 3:1 72.51 <.001 [OK x PK]3 7 28 3:1 56.47 <.001 [OK x PK]4 19 10 3:1 1.39 0.238 [OK x PK]5 18 4 3:1 0.54 0.460 [OK x PK]6 22 6 3:1 0.19 0.662 [FG x PK]1 22 7 3:1 0.01 0.915 [FG x PK]6 20 5 3:1 0.33 0.564 Testcross: SH x [OK x PK]2 15 18 1:1 0.27 0.602 SH x [OK x PK]3 3 27 1:1 19.20 <.001 FG x [OK x PK]3 0 20 1:1 - - SH x [FG x PK]1 8 10 1:1 0.22 0.637 SH x [FG x PK]6 23 17 1:1 0.90 0.343 a Parents: P. persica cultivars ‘Okinawa’ (OK), ‘Flordaguard’ (FG), ‘UFSharp’ (SH); P. kansuensis wild peach (PK). b Individual genotypes were evaluated 120 days after inoculation with M. floridensis (‘MFGnv14’ isolate) at a concentration of 10,000 eggs or second-stage juveniles per plant. Plants were classified as resistant if GI ≤ 2, otherwise they were considered susceptible. c Probability (P) values for testing goodness-of-fit to the expected segregation ratios. Hypothesis is accepted if P ≥ 0.05.

B 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Figure 2-1. Meloidogyne floridensis ‘MFGnv14’ isolate from ‘Flordaguard’ peach rootstock. (A) Galling of fine feeder roots. (B) Confirmation of purity of M. floridensis isolate multiplied on tomato host based on the electrophoretic profile of malate dehydrogenase and esterase enzyme phenotypes of single females of M. floridensis (lanes 2 to 14) and M. javanica included as reference (lanes 1 and 15). Photos courtesy of Mary Ann D. Maquilan.

A r = 0.966 P = <.0001 n = 379

B r = 0.886 P = <.0001 n = 377

resistant

Figure 2-2. Scatterplot and Spearman rank correlation data between gall index and indicators of host efficiency, namely: (A) egg mass index and (B) reproduction factor, RF. Data combined from 13 peach x Prunus kansuensis populations. Scoring system for gall index and egg mass index were based on total counts per whole root system: 0 = no galls or egg masses, 1 = one to two galls or egg masses, 2 = three to 10, 3 = 11-30, 4 = 31-100, and 5 = more than 100 (Taylor and Sasser, 1978). RF: ratio of the final egg count to initial inoculum at 10,000 eggs per plant. RF levels 1, 2, and 3 correspond to <0.1, 0.1-1, and >1 values, respectively. Plants were classified as “resistant” if RF < 0.1 and gall index or egg mass index ≤ 2.

Parents a

mf2 mf2 Mf1 mf3 ♀ Prunus persica ♂ Prunus kansuensis ‘Okinawa’ (OK), ‘Flordaguard’ (FG) ‘UFSharp’ (SH) b

F1 hybrids

mf2 mf3 Mf1 mf2

[OK x PK]3 [OK x PK]1 [OK x PK]5 [FG x PK]1 [OK x PK]4 [OK x PK]6 [FG x PK]6

x x

F2 progenies

1 mf3 mf3 : 2 mf2 mf3 : 1 mf2 mf2 1 Mf1 Mf1 : 2 Mf1 mf2 : 1 mf2 mf2 1 R : 3 S 3 R : 1 S

Testcrosses b

mf2 mf2 x mf2 mf3 mf2 mf2 x Mf1 mf2

♀ SH x ♂ [OK x PK]3 ♀ SH x ♂ [OK x PK]2 ♀ FG x ♂ [OK x PK]3 ♀ SH x ♂ [FG x PK]1 ♀ SH x ♂ [FG x PK]6

Testcross progenies

1 mf2 mf3 : 1 mf2 mf2 1 Mf1 mf2 : 1 mf2 mf2 0 R : 1 S 1 R : 1 S

Figure 2-3. Genetic model for the inheritance of Mf gene in Prunus kansuensis where a Mf1 is dominant over mf2 and mf3, and mf2 is dominant over mf3. Recessive b allele from Prunus persica (mf2) does not confer resistance. ‘UFSharp’, a peach scion cultivar, was used as an alternate susceptible parent for the testcrosses because of its prolific flowering. R, resistant; S, susceptible.

CHAPTER 3 GENETIC ANALYSES OF RESISTANCE TO THE PEACH ROOT-KNOT NEMATODE USING MICROSATELLITE MARKERS

Introduction

Development of peach rootstocks in the US has historically been directed at improving nematode resistance in environmentally adapted varieties. The old nematode-resistant rootstocks such as ‘Shalil’, ‘Yunnan’ and ‘Bokhara’ had been found to be susceptible to Meloidogyne javanica (Hansen et al., 1956; Sharpe, 1967).

Continued efforts to broaden the resistance base of peach rootstocks while improving horticultural features have led to the development of rootstocks ‘Nemaguard’, ‘Nemared’

‘Guardian’, and ‘Flordaguard’ that were also effective against M. javanica (Okie et al.,

1994a; Ramming and Tanner, 1983; Reighard and Loreti, 2008; Sherman et al., 1991).

The rootstock cultivar ‘Flordaguard’, with ‘Okinawa’ and Prunus davidiana in its lineage, has been effective as a rootstock for peach production in Florida due to its low-chilling requirement, adaptability to non-alkaline soils, and ability to withstand infection by the endemic peach root-knot nematode (RKN), Meloidogyne floridensis (MF), which is not controlled by the resistance (R) gene RMia in ‘Nemared,’ ‘Nemaguard,’ and ‘Okinawa’ rootstocks (Handoo et al., 2004; Olmstead et al., 2015).

Recently, a resistance-breaking isolate of MF was found infecting ‘Flordaguard’ and was used in our previous study on the inheritance of resistance against MF.

Segregation analyses among F2 and BC1F1 interspecific populations from crosses of

‘Okinawa’ or ‘Flordaguard’ with Prunus kansuensis (PK) have established PK as a source of new alleles conferring high level of resistance against MF and revealed single-locus inheritance patterns congruent with what has been observed for other RKN

R genes identified in Prunus (Claverie et al., 2004a; Esmenjaud et al., 1996b; Kochba

and Spiegel-Roy, 1975; Lu et al., 2000b; Rubio-Cabetas et al., 1998; Van Ghelder et al.,

2010; Yamamoto et al., 2001). Our previous study also demonstrated the potential of interspecific hybridization with PK for broadening the genetic base of RKN resistance in current rootstocks, having been derived from P. persica and P. davidiana.

Breeding for improved RKN resistance becomes increasingly important as highly toxic soil fumigants or nematicides are being phased out and even more so for production areas where the highly adaptive meiotic RKN species such as MF

(Castagnone-Sereno, 2006; Handoo et al., 2004) becomes a problem. Choosing an efficient breeding strategy for the introgression of MF resistance from PK into well- adapted cultivars requires an understanding of the genetic control of resistance and an informed selection of parental allele combinations to increase the frequency of resistance in the progeny. Use of molecular markers tightly linked to the trait can accelerate the identification of desirable genotypes (Charcosset and Moreau, 2004;

Hospital, 2005) especially those exhibiting nematode resistance, for which screening can be tedious and time-consuming. In this study, we used simple sequence repeat

(SSR) microsatellite markers to identify genomic regions associated with MF resistance and utilize marker-trait linkage information to track the transmission of resistance alleles from PK into backcross progenies.

Microsatellite markers are ideal for genetic mapping studies because of their abundance in the genome, codominance, high polymorphism, and transferability across species (Aranzana et al., 2003; Aranzana et al., 2002; Claverie et al., 2004a; Mnejja et al., 2010; Tautz, 1989). Additionally, microsatellite markers have shown promise in marker-assisted selection of RKN resistance in several agronomic crops including rice,

soybean, cotton, and peanut (Chu et al., 2011; Concibido et al., 2004; Jenkins et al.,

2012; Miah et al., 2013). Microsatellite markers have been integrated in the saturated

Prunus reference map constructed from ‘Texas’ almond x ‘Earlygold’ peach F2 population (Dirlewanger et al., 2004b), a useful resource for selecting informative and uniformly distributed markers, and have been used in initial mapping studies to locate underlying RKN R genes in several Prunus intra- and interspecific progenies (Cao et al.,

2011; Claverie et al., 2004a; Dirlewanger et al., 2004a). Genetic mapping studies have located Prunus RKN R genes Ma, RMja and Rjap in LG 7 at ≈2 cM from SSR marker pchgms6 (Claverie et al., 2004a; Van Ghelder et al., 2010), as well as RMia, and PkMi in the subtelomeric region of LG 2 at ≈10 cM from the SSR marker UDP98-025 (Cao et al., 2011; Claverie et al., 2004a). Putative resistance genes, with the majority belonging to the toll/interleukin1 receptor (TIR) nucleotide-binding site (NBS) leucine-rich repeat

(LRR) class, have been found to contain SSRs; in the peach genome, 60% of the disease-resistance-associated TIR-NBS-LRR genes with post-LRR regions contain

SSR sequences in the intron between LRR and post-LRR exons suggesting the potential utility of polymorphic SSRs for detecting and directly monitoring the introgression of disease resistance genes in breeding programs (Claverie et al., 2011;

Lalli et al., 2005; Van Ghelder and Esmenjaud, 2016). Multiple reported linkages for

RKN resistance in the upper part of LG 2 and LG 7 indicate either clustering of genes or high allelic variability in these regions and, therefore represent useful targets for generating rootstocks with broader spectrum and more durable resistance against RKN, as demonstrated for the three-way plum x (almond x peach) progeny containing completely dominant Ma, RMja, and RMia genes (Khallouk et al., 2013).

This study aims to identify SSR markers linked with MF resistance, determine the genomic location of the resistance locus, and track resistance introgression based on genetic linkage with the SSR marker in early-generation (F2 and BC1F1) interspecific populations derived from crosses between peach and PK while analyzing recombination patterns across the genome. Results from this study will provide a useful basis for further genetic studies and marker-assisted introgression of novel sources of resistance alleles from PK into locally adapted peach rootstocks.

Materials and Methods

Plant Material

Twelve F2 and BC1F1 populations developed from interspecific crosses between

Prunus persica (L.) Batsch (peach) and P. kansuensis Rehder (Kansu peach) were evaluated for genetic segregation at select microsatellite marker loci. The P. persica cultivars ‘Flordaguard’ (FG) and ‘Okinawa’ (OK) are utilized as genetic material at the

University of Florida (UF) stonefruit breeding program to breed for low-chill adaptation.

‘Okinawa’ is a domesticated peach seedling rootstock originating from Japan (Sharpe,

1957). ‘Flordaguard’ is the standard rootstock used for subtropical peach production and has both ‘Okinawa’ peach and P. davidiana in its parentage (Sherman et al., 1991).

‘UFSharp’ (SH), a low-chill-adapted peach scion cultivar, was used as one of the backcross female parents because of its profuse flowering (Chaparro and Sherman,

2006). From our previous study (see Chapter 2), the P. persica genotypes OK, FG, and

SH were determined to be homozygous susceptible for ‘MFGnv14’, a resistance- breaking isolate of M. floridensis (MF). The observed differences in segregation patterns among populations derived from the P. persica x P. kansuensis (PK) cross suggested that our PK accession is heterozygous at the resistance locus. Seven F2

populations were developed from selfing five OK x PK hybrid selections and two FG x

PK hybrid selections. Five of these interspecific hybrids were backcrossed to either FG or SH. A total of 478 individuals from 12 populations, each consisting of 23-43 individuals, were initially genotyped and evaluated for resistance to MF. All progenies were grown and evaluated for MF resistance in the greenhouse, as previously described (see Chapter 2).

Phenotyping

The response of individual genotypes to MF was assessed by subjecting four- month-old seedlings to a high inoculum pressure of 10,000 eggs or second-stage juveniles from a pure population of the ‘MFGnv14’ isolate. After four months from inoculation, root systems were visually rated for the number of galls. Each genotype was assigned a root galling index (GI) of 0 = no galls, 1 = one to two galls, 2 = three to

10, 3 = 11 to 30, 4 = 31 to 100, or 5 = more than 100 (Taylor and Sasser, 1978). Root

GI was the criterion established in the previous study to separate resistant and susceptible phenotypes, where values above 2 indicate host susceptibility.

Genotyping

DNA extraction and amplification

Genomic DNA was extracted from the progenies and parents of the crosses following the modified CTAB procedure described by Chavez and Chaparro (2011) .

Genomic DNA was diluted to a final concentration of 20 ng·uL-1 in TE buffer for PCR amplification of simple sequence repeat (SSR) loci. The PCR reactions were conducted in a 16 uL volume; a total of 10 uL master mix including 2.25 uL of 10X Thermopol® reaction buffer (New England BioLabs), 1 uL of 2.5 mM dNTP, 0.2 U of Taq DNA polymerase (New England BioLabs) and 6.5 uL of DNA-grade water, was combined

with 2 uL each of 5 uM fluorescent-labeled forward primer and 5 uM reverse primer, and

2 uL of 20 ng·uL-1 of genomic DNA. The PCR amplification proceeded using the following temperature profile: 94°C for 3 min, then 40 cycles of (94°C for 45 s, specific primer annealing temperature for 30 s, 72°C for 1 min) finishing with 72°C for 7 min.

SSR analysis

The linkage map developed from an interspecific cross of ‘Texas’ almond with

‘Earlygold’ peach (TxE) was used as a reference map (Genomic Database for

Rosaceae ) for selecting the SSR marker loci.

Initially, eighty-one SSR markers spanning eight linkage groups (LG) were screened for stable amplification, polymorphism between P. persica and P. kansuensis parental genotypes, and transmission of alleles in selected hybrids. Of the 59 that showed polymorphisms, only 39 were selected for genotyping the mapping panel containing seven F2 and five BC1F1 families based on their location on the Prunus TxE reference map (Supplementary Table B-1). Three SSR loci were selected from the linkage maps of ‘Contender’ peach x ‘Fla.92-2C’ peach (Fan et al., 2010) and ‘Ferjalou Jalousia’ peach x ‘Fantasia’ peach (Dirlewanger et al., 2007) F2 populations to extend coverage in

LG 4 and LG 5. Details of the location, primer sequences, and annealing temperature of the selected markers are provided in Supplementary Table B-2. The selected SSR markers were spread out over the eight linkage groups with average spacing ranging from 7 cM in LG 7 to 25 cM in LG 6 (Supplementary Figure B-1). The number of selected markers on each linkage group ranged from three for LG 3 to eight for LG 1.

Such a marker density would be sufficient to screen each chromosome for QTL in early generation populations such as F2 and backcrosses where a relatively large linkage disequilibrium between loci is expected (Bus et al., 2009).

A total of 478 individuals from 12 interspecific peach x P. kansuensis populations were genotyped for 39 genome-wide SSR markers (see Supplementary Figure B-2 for representative gel images used in assessing genotypes). The highly polymorphic SSR loci revealed 131 genotypes having off-type alleles at one or more loci (1-21 SSRs) and were consequently removed from the mapping files. Excluding off-types, the final mapping data sets comprised of 18-40 individual genotypes from 12 populations, for a total of 345 genotypes (Supplementary Table B-3).

Genotypic data were obtained by visual scoring of allele bands separated by agarose gel electrophoresis or by analyzing fluorescent peaks generated by capillary electrophoresis. The PCR products were separated via gel electrophoresis on 3-4% agarose gels stained with Ethidium bromide (run for 3-4 hours at 210 volts) and markers segregating for alleles that differed by more than 5 base pairs (bp) were scored from the gel images. Bands that differed by less than 5 bp and those that were fuzzy or too faint to be reliably scored were resolved by the capillary electrophoresis sequencer (ABI

3730 DNA Analyzer, Applied Biosystems) at the UF Interdisciplinary Center for

Biotechnology Research (Gainesville, Florida). Genotypes were determined from electropherograms generated in GeneMarker software (v.1.5, Soft Genetics) using the

Liz600 size standard to determine the size of sample peaks.

Marker scoring

Marker scoring followed the genotype coding conventions of JoinMap® (Van

Ooijen, 2006) for F2 and BC1 population types. For F2 populations, individuals were scored as “A” when they were homozygous for the P. persica allele, “B” when homozygous for the P. kansuensis allele, and “H” when heterozygous. For backcross populations, only SSR markers heterozygous in the F1 parent or both parents were

used. Letters A-D or Ø (null) were designated to the allelic variants of each marker

(Supplementary Tables B-4 to B-8) and were later grouped based on their segregation patterns. The polymorphic markers segregated according to: AB x CD, EF x EG, and

LM x LL. Alleles represented by the letters B, F, and M originated from P. kansuensis while the letters A, C, D, E, G, and L were from P. persica. Considering the BC1F1 populations as species-level backcrosses, alleles from P. persica were lumped into a single “A” allele and genotypes with only P. persica alleles were treated as homozygotes whereas genotypes with the combined alleles of P. kansuensis and P. persica were re-coded as “H” when converting genotypic data to the BC-type format.

The phenotypic marker (resistance to M. floridensis) was scored as a dominant marker:

“C” for resistant phenotypes and “D” for susceptible phenotypes.

Error checking

Some markers yielded ambiguous genotyping results because of faint or diffused bands on the gel, spurious peaks or overlapping stutter peaks on the electropherogram, and presence of null alleles. To resolve genotype ambiguities and to confirm the presence of null or off-type alleles, PCR amplification and electrophoresis were repeated once or twice. The electrophoretic band patterns and the allele peak patterns produced from GeneMarker® were compared. Markers were checked for Mendelian inconsistencies that may have resulted from genotyping errors by chi-square tests using

JoinMap. The electrophoresis output was re-examined when significant (P < 0.05) distortions were observed. Markers that could not be reliably scored were removed before performing linkage analysis. Furthermore, genotype errors or failed PCR reactions, usually due to poor-quality of DNA sample, were marked as missing data and individuals with more than 10% missing data were removed from the mapping files to

avoid erroneous marker data. Some markers revealed multiple stutter peaks that made it difficult to distinguish the true allele peaks. The true allele composition was determined by comparing segregation patterns across individuals within the population.

Map Construction

Individual maps

Maps containing 34-36 SSR markers were constructed for each of the 12 F2 and backcross populations using the regression mapping algorithm in JoinMap. Marker datasets were coded following the format for F2 and BC population types. Marker data were subjected to chi-square tests to check for deviations from the 1:2:1 expected segregation in F2 or 1:1 in backcross population. The presence and magnitude of segregation distortion in mapping populations were assessed under the “locus genotypic frequency” command in JoinMap. Segregation-distorted markers were retained in the map if they did not dramatically affect mapping distances and marker order.

Markers were grouped by their independence LOD (logarithm [base 10] of odds) scores ranging from 2.0 to 10.0 with steps of 1 LOD unit. Marker groupings with the highest possible LOD scores were selected from the “groupings tree” and any ungrouped markers, including segregation-distorted markers, were assigned into the existing linkage groups based on their known locations on the Prunus TxE reference map. Within each linkage group, markers were ordered starting with the most informative pair and the remaining markers were positioned one by one in subsequent rounds using the default settings (minimum LOD threshold of 1.0; maximum recombination frequency of 0.40; goodness-of-fit jump threshold for removal of loci =

5.0) and a ripple performed after each marker addition. Generally, the marker order is

established after one round of fitting. The LOD threshold was successively reduced

(0.5, 0.1, and 0.01) when a set of markers could not be mapped due to insufficient linkage. In some datasets where only two informative markers were available for LG 3, the parameters were adjusted to the lowest stringency (recombination frequency = 0.5,

LOD > 0.01) to place the two distant markers. Recombination frequencies were converted to map distances in centiMorgan (cM) using Kosambi’s mapping function.

The marker order was checked for consistency with that of the Prunus TxE reference map. Conflicts in the order of the markers were resolved using the “start order” and “fixed order” map building functions in JoinMap. The tightly linked markers on LG 7 often generated negative map distances that could not be resolved using the

“fixed order” command. Redundant markers (where no recombination occurred between two adjacent markers) were eliminated if they caused conflicts in marker order that could not be resolved due to insufficient resolution of the mapping population.

Combined maps

Combined maps were constructed using individual maps from populations with common parental allele combinations. Multiple maps for each linkage group were integrated using the “combine groups for map integration” command in JoinMap.

Marker pairs were examined for heterogeneity of recombination frequencies under the

“heterogeneity test” tabsheet. Where significant linkage heterogeneity (P < 0.05) occurred, marker data were re-examined for genotyping errors or inconsistencies in marker order. The marker pairs with deviant recombination frequencies causing map inflation were excluded from map construction. In LG 7, linkage heterogeneity was usually caused by clustering of markers at the same position in some mapping populations, particularly between CPPCT039-UDP98-405 (2-4 cM interval on the

Prunus TxE reference map) and between CPPCT017-EPDCU3117 (3 cM interval on the TxE reference map). Linkage heterogeneity also occurred in LG 5 where only 2-3 markers were mapped. Linkage map calculations were based on mean recombination frequencies and combined LOD values of common markers mapped in more than one population. The regression mapping approach was used for determining marker order and similar parameters used for individual map construction were considered for the combined maps. Marker pairs failing to meet set criteria were excluded from the map construction. Linkage maps were drawn and aligned with MapChart version 2.3

(Voorrips, 2002).

Data Analyses

Statistical analyses were performed using the SAS software (version 9.4, SAS

Institute Inc., Cary, North Carolina, USA). The distribution of phenotypic data (root galling index) was assessed visually with histograms and checked for normality using the Proc Univariate procedure of SAS. The 12 F2 and backcross families were grouped based on common segregation patterns or parental allele combinations previously inferred from the inheritance study and the pooled data sets were checked for homogeneity of variances using the Levene’s test, which is more robust for departures from normality. Chi-square goodness of fit tests were performed on phenotype segregation data to determine their agreement with the expected Mendelian ratios using the Proc Freq procedure in SAS. All statistical tests used a P-value threshold of 0.05 unless otherwise noted.

Data sets from similarly grouped crosses were separately analyzed for the presence of QTLs and then combined and analyzed together to increase the statistical power of detecting QTLs. Because the data did not follow univariate normal distribution

and the data could not be transformed to fit the normality assumption for parametric tests, the non-parametric mapping function in MapQTL® (Van Ooijen, 2009) based on

Kruskal-Wallis (KW) test was used to explore individual associations between SSR markers and resistance based on root GI. A threshold level of P = 0.005 for genome- wide significance (Van Ooijen, 2006) was considered for identifying marker(s) closely associated with the GI trait in both individual and combined families. Exact P values for the significant marker were obtained by a Kruskal-Wallis test in SAS (Proc Npar1way).

The proportion of phenotypic variation (R2) accounted for by the significant marker locus was determined by a simple linear regression model for each family as well as the combined families. In addition, a two-way ANOVA (Proc GLM) was performed to investigate main and interaction effects between marker genotype and family. When no significant family effects and marker x family interaction effects were observed, the data sets were pooled to provide improved estimates of the phenotypic effect of shared QTL alleles. Following a significant KW test, the effects of the genotypes at the associated marker locus were further examined with post-hoc pairwise comparisons using the

Wilcoxon rank-sum test run with α = 0.005 for two groups (BC genotypes segregating

1:1 at a locus) and the KW test for three groups (F2 genotypes segregating 1:2:1) with

Bonferroni-adjusted critical P-value (ά = 0.0017) to account for multiple comparisons.

Ranks were assigned for GI values, ordered from lowest to highest, and the mean ranks of genotypes were compared. The genotype effects were expressed as mean GI values ± standard deviation (SD) followed by different letters denoting significantly different mean ranks among genotypes.

Results

Resistance Segregation

A total of 345 individuals from 12 F2 and BC1F1 families derived from the cross P. persica x P. kansuensis were tested for resistance to the pathogenic isolate of M. floridensis (‘MfGnv14’) as described in the previous study (see Chapter 2). Two different patterns of segregation for root galling index (GI) were observed among the seven F2 families, of which six showed 3:1 resistant (R) to susceptible (S) ratio (Pcombined

= 0.894) and one showed a 1R:3S ratio (P = 0.494). The segregation patterns suggest that a major resistance locus is involved with resistance determined by a dominant allele in the families segregating 3R:1S, and a recessive allele in one family segregating

1R:3S. When backcrossed to a homozygous-susceptible P. persica genotype

(‘Flordaguard’ [FG] or ‘UFSharp’ [SH]), the BC1F1 progenies segregated 1R:1S (Pcombined

= 0.916) as expected if the dominant resistance allele was inherited, and 0R:1S if the recessive resistance allele was inherited in the progeny. The segregation ratios in each population and mean GI values for the R and S phenotypes are presented in Table 3-1.

The variation in segregation patterns among the F2 progenies indicate the heterozygous allelic status of the wild peach parent, P. kansuensis (PK), at the locus conferring resistance to M. floridensis (MF) and the segregation of two different PK alleles into the F1 hybrids. Through backcrosses to a homozygous-susceptible P. persica (PP) parent, we have confirmed that PK is the progenitor of MF resistance, for which we assigned the symbols “Mf1” and “mf3” to represent dominant and recessive resistance alleles, respectively. The symbol “mf2” was used to represent susceptible allelic forms of P. persica.

Based on common segregation patterns, the different crosses could be classified according to the different allele combinations in the F1 parent as follows: Mf1 mf2 x Mf1 mf2 and mf2 mf3 x mf2 mf3 for selfed F1 hybrids; mf2 mf2 x Mf1 mf2 and mf2 mf2 x mf2 mf3 for backcrosses to P. persica. The phenotypic distributions of the different cross combinations for the GI trait are presented in Figure 3-1. The GI trait displayed non- normal, strongly skewed distributions with the Mf1 mf2 x Mf1 mf2 crosses (N = 169) heavily skewed towards resistance and the mf2 mf3 x mf2 mf3 cross (N = 35) heavily skewed towards susceptibility. The mf2 mf2 x Mf1 mf2 crosses (N = 91) displayed the expected bimodal distribution while the mf2 mf2 x mf2 mf3 crosses (N = 50) consisted almost entirely of susceptible types. The histograms further show that resistance to MF is controlled by a major locus and there are dominantly and recessively inherited resistance alleles. Combined information from populations having the same phenotypic distributions corresponding to similar allelic constitutions at the MF locus could be used to provide better estimates of allele effects and increase the power to detect QTLs due to the increased sample size (Mackay et al., 2009).

Polymorphisms and Genotype Segregation at Microsatellite Loci

Among the 39 SSR markers selected for mapping, 29 consistently showed informative polymorphisms across twelve peach x P. kansuensis populations while the other markers were informative only in certain progenies (Supplementary Table B-9).

Markers CPPCT026, EPPISF032, BPPCT038, and UDP98-405 were not informative in some F2 progenies and backcrosses because the F1 parents inherited the common alleles of P. persica (‘Okinawa’ [OK] or ‘Flordaguard’ [FG]) and P. kansuensis (PK), and similarly when the F1 parents were backcrossed to another P. persica genotype

(‘UFSharp’ [SH]). Markers UDP96-013 and CPPCT005 displayed polymorphisms in the

F2 but the PK alleles could not be traced in the backcross progenies due to shared alleles with the backcross parent (SH or FG). Marker CPDCT025 was monomorphic between OK and PK while CPSCT034 did not show segregation in some F2 and backcross progenies.

The rate and distribution of segregation-distorted loci varied among the different crosses (Supplementary Table B-10). Goodness of fit tests of 1:2:1 (F2) and 1:1

(backcross) ratios revealed approximately 12-27% significantly distorted loci in different genomic regions for OK x PK, 11-15% for FG x PK, 10-17% for SH x [OK x PK], 11-22% for SH x [FG x PK], and 3% for FG x [OK x PK]3. Such biases were frequently observed to be in favor of PP alleles with more homozygotes or heterozygotes than expected, although these may also indicate ambiguities in marker scoring. In LG 6, an excess of genotypes homozygous for PP alleles was commonly observed. A high level of distortion (P < 0.005) at CPPCT030 and BPPCT025 loci on LG 6 was common among selfed progenies but there was no significant distortion at CPPCT030 locus among backcrosses to PP. In all F2 progenies where the CPPCT030 locus appeared to be severely distorted, there were no genotypes homozygous for PK alleles except for one progeny that contained one such homozygote. Significant distortions were observed at

BPPCT030 locus on LG 2 and at UDP98-408 locus in SH x [OK x PK]2 progeny having more heterozygotes than expected. Moderate distortions were observed at CPPCT039 and CPPCT022 loci on LG 7 in [OK x PK]4 having more PK-homozygotes than expected.

For the four markers on LG 1 that showed unexpected homozygote frequencies, several independent rounds of PCR amplification and fragment analyses were

conducted to compare the outputs and verify the scores. False homozygotes due to null alleles, allele dropout, and adjacent-allele heterozygotes were corrected. For

CPSCT008, many genotypes tended to be misinterpreted as homozygous because the larger allele (234 bp) appears as a very low peak or a very faint band when the other allele (182 bp) is present. In the case of CPPCT029 and BPPCT028, the OK x PK and

FG x PK F1 parents appeared to be homozygous and there was no apparent segregation among the F2 progenies. However, the genotypes segregated 1:1:1:1 when backcrossed to SH, which is heterozygous for the marker; this could be explained by the F1 parent being heterozygous for the null allele. For BPPCT028, the heterozygous-null genotypes were distinguished from the homozygous genotypes by the extended range of their stuttering peaks. Segregating null alleles from marker

CPPCT006 were also observed but these could not be confirmed in most populations due to poor repeatability at the locus that is not attributable to poor-quality DNA. For

CPPCT026, difficulty in differentiating stutters of homozygotes from adjacent-allele heterozygotes (191/189 bp) was frequently encountered in the OK x PK F2 populations.

The marker CPPCT019B on LG 8 presented a similar problem (for all 12 populations), which is known to occur quite commonly for markers having a dinucleotide repeat motif

(Clarke et al., 2001; DeWoody et al., 2006). Linkage maps were constructed with little regard for segregation distortion. However, only the marker loci that gave consistent genotyping results were included in the mapping files.

Linkage Maps

Linkage information from multiple populations was combined to improve the recombination frequency estimates and to construct maps with more accurate marker distances as our common set of markers displayed varied degrees of polymorphism and

recombination in different populations. The combined maps represented the four parental crosses: two for F2 populations (OK x PK and FG x PK) and two for backcrosses to P. persica (SH / FG x [OK x PK] and SH x [FG x PK]). The four combined maps consisted of 34 to 36 markers covering 59 to 64% of the TxE reference map. The marker coverage of the combined maps varied by linkage group. More markers were tested for LG 1, 2, 7, and 8 than for LG 3, 4, 5, and 6 (Supplementary

Figures B-3 to B-6). Several markers had to be tested in LG 1 and LG 8 because of the difficulty in identifying usable markers based on polymorphism. Because LG 1 is longer and more saturated with SSR markers than LG 8, more markers were used to map LG

1 (5 to 8 markers at 8.7 to 15.0 cM intervals) compared to LG 8 (2 to 3 markers at 10.9 to 22.8 cM intervals). The region in LG 4 had the smallest coverage (28.6%) on the TxE map because two (EPPISF032 and CPDCT014) of the four markers were selected from different linkage maps. Greater coverage (> 70%) and sufficient marker density (5 to 8 markers with 6.2 to 14.1 cM intervals) were ensured for linkage regions in LG 2 and LG

7 where the root-knot nematode resistance genes for Prunus had been localized (Cao et al., 2011; Claverie et al., 2004a; Van Ghelder et al., 2010).

The total map lengths ranged from 248.5 to 351.2 cM with the average marker spacing per linkage group ranging from 6.2 to 25.1 cM, excluding LG 3 in OK x PK and

FG / SH [OK x PK] combined maps that were dramatically inflated by the two distant markers. Only three markers were used to map LG 3, being the smallest (48.4 cM) of the eight linkage groups in TxE map, and one marker (CPDCT025 placed between

BPPCT007 and CPDCT027) could not be used because it was monomorphic between

OK and PK. The effect of segregation-distorted loci on LG 6 was noticeable on the F2

linkage maps. Three tightly-linked, distorted loci (BPPCT008, BPPCT025, and

CPPCT030) in LG 6 produced map lengths shorter than the TxE map by 53% in OK x

PK and 56% in FG x PK combined maps. Comparison of total map lengths between F2 and backcross populations indicated increased recombination frequencies in the latter; the combined map lengths in backcross populations were longer by 28% in FG / SH

[OK x PK]) and by 30% in SH x [FG x PK) than those of their respective F2 counterparts,

OK x PK and FG x PK (Supplementary Tables B-11 to B-14).

Marker-trait Association

The Kruskal-Wallis (KW) test was performed to identify microsatellite regions in the genome that are significantly associated with the resistance to M. floridensis (MF) based on root GI values. A threshold of P < 0.005 was considered for genome-wide significance (Van Ooijen, 2009). Because of the non-normal distribution of the GI trait, we used a nonparametric distribution-free test, which analyzes differences in medians for ordinal data. The KW test suggested the presence of a quantitative trait locus (QTL) for MF resistance in LG 2 close to the UDP98-025 locus. Among the 34 to 36 SSR regions screened, no other marker locus showed a stronger association with the GI trait than UDP98-025. In addition, there were no markers in other linkage groups that showed a genome-wide level of significance. Segregation distortion was not evident at the significant marker locus as determined by chi-square tests of 1:2:1 ratio in F2 progenies and 1:1 in backcrosses. The phenotypic variance explained (PVE) by the

QTL ranged from 23.9 to 80.4% when resistance is dominantly conferred in the progeny and 8.6 to 22.4% when resistance is recessively conferred. Table 3-2 shows the degree of association of UDP98-025 marker with the resistance response to MF infection, mean GI values for the genotypes and the PVE (R2) in each progeny.

The F2 progenies were classified into two genotype cross combinations based on the assumption of a segregating dominant resistance allele (Mf1) or recessive resistance allele (mf3) inherited from PK. The F2 progenies were partitioned into their genotypes for the codominant SSR marker UDP98-025 (B, homozygous for PK allele;

H, heterozygous; A, homozygous for PP allele) which, under the assumption of tight linkage, should correspond to their putative genotypes at the MF resistance locus (B =

Mf1; H = Mf1 mf2 or mf2 mf3 depending on the F1 source; A = mf2 mf2). In case of a multi- progeny cross type, the GI data were checked for non-significant interaction effects between progeny and marker genotype before combining datasets.

Following a significant KW test (P < 0.005), nonparametric pairwise comparisons per cross type were made to determine differences among the genotypes with respect to the GI data. The analysis was conducted on the combined dataset to increase representation of resistant/susceptible individuals in corresponding genotypes at the marker locus and therefore, increase the statistical power to detect significant genetic effects. Rank-based analyses revealed that, in the Mf1 mf2 x Mf1 mf2 cross, the PK- homozygotes (Mf1 Mf1) and heterozygotes (Mf1 mf2) were significantly different from the

PP-homozygotes (mf2 mf2). The Mf1 and Mf1 mf2 genotypes, corresponded to low mean

GI values of 0.18 ± 0.69 and 0.66 ± 1.55, respectively, whereas the mf2 genotypes corresponded to high mean GI values (4.00 ± 1.97). These results indicate the involvement of the Mf1 allele in reducing the amount of root galling while mf2 showed no such contribution. The significant effect of the Mf1 allele on resistance could also be observed in the backcross progenies of the mf2 x Mf1 mf2 cross, in which the Mf1 mf2 genotypes corresponded to lower mean GI values (0.78 ± 1.67) compared to mf2

genotypes (4.37 ± 1.65). In the mf2 mf3 x mf2 mf3 cross, the difference in the levels of root galling between PK-homozygotes (3.27 ± 1.79) and PP-homozygotes (5.00 ± 0.00) was marginally significant (P = 0.006) but, when compared with the heterozygotes (4.33

± 1.23), the PK-homozygotes did not show significantly (P =0.454) reduced levels of root galling as would be expected if two copies of the recessive allele (mf3 mf3) confer resistance and not when only a single copy (mf2 mf3) is present. The recessive contribution of the mf3 resistance allele may have been underestimated because the mf3mf3 genotypes were underrepresented (only 4 resistant out of 11 individuals with the

PK-homozygous marker genotype) due to recombination between the UDP98-025 marker and the QTL. Transmission of the recessive resistance allele to the backcross progenies of the mf2 mf2 x mf2 mf3 cross yielded predominantly suceptible phenotypes associated with the mf2 mf3 and mf2 mf2 genotypes.

The strong linkage of UDP98-025 with the MF resistance locus allowed us to follow the transmission of the resistant/susceptible alleles in the backcross progenies and the results provide further support for the existence of three different alleles at the resistance locus: two resistance alleles from PK that are either dominant (Mf1) or recessive (mf3), and susceptible alleles (mf2) from PP. In the F2 and backcross progenies of F1 parents carrying the dominant resistance allele (Mf1), the UDP98-025 marker correctly predicted resistance among PK-homozygotes and heterozygotes (co- segregating with Mf1 Mf1 and Mf1 mf2 genotypes, respectively), as well as susceptibility among PP-homozygotes (co-segregating with mf2 mf2 genotypes) up to 100%. In [OK x

PK]3-derived F2 and backcross progenies where the resistance trait was recessively inherited, the marker has a low prediction accuracy (36%) in PK-homozygotes due to

the low frequency of resistant individuals but could still reliably predict susceptibility in heterozygotes and PP-homozygotes (co-segregating with the respective mf2 mf3 and mf2 mf2 genotypes) up to 100% (Table 3-3).

Localization of Major Resistance Locus on LG 2

Four to five polymorphic SSR markers were mapped on LG 2 (using the regression mapping method in JoinMap) in seven interspecific F2 progenies and three backcross progenies with map lengths ranging from 13.5 cM to 66.8 cM in F2 and 45.2 cM to 48.9 cM in the backcrosses. The SH x [OK x PK]3 and FG x [OK x PK]3 backcross progenies showed no segregation for resistance so these populations were not used to map the MF locus. Three markers (BPPCT013, UDP96-013, and BPPCT030) in LG 2 generally appeared to be tightly linked (and clustered more tightly in F2 maps than in

BC1F1 maps) thus, the variability in the map lengths for LG 2 resulted mainly from differences in recombination between UDP98-025 and the three tightly linked loci.

The estimated linkage distances between the MF locus and the UDP98-025 SSR locus were highly variable across the F2 populations indicating different recombination levels at the subtelomeric region in LG 2. In F2 populations with dominantly inherited resistance, the linkage distances between UDP98-025 and MF locus ranged from 0.8 cM in [OK x PK]4 (n = 29) to 13.6 cM in [OK x PK]1 (n = 36). Combining linkage information from the six F2 populations, UDP98-025 was linked at 7.2 cM to the MF locus. In the [OK x PK]3 population, the MF locus was located at a greater distance

(24.6 cM) from UDP98-025 as expected due to the recessively inherited trait. The

BC1F1 populations generally showed greater map distances, which did not seem to be influenced by the size of the population used for mapping; the MF locus was positioned

8.5 cM from the UDP98-025 locus in SH x [FG x PK]6 (n = 40) to 20.6 cM in SH x [OK x

PK]2 (n = 33). The combined BC1F1 linkage map (N = 90) showed the association of

UDP98-025 with the MF locus at 14.5 cM, which is twofold greater than that of the combined F2 linkage map (7.2 cM, N = 169) (Figure 3-2).

Discussion

Our previous study, which examined the inheritance of resistance to M. floridensis in 12 F2 and BC1F1 progenies derived from crosses between peach and P. kansuensis (PK), showed that resistance based on root galling index is inherited from the heterozygous PK accession in either a dominant or recessive manner. The F2 progenies segregated in a ratio of 1:3 or 3:1 between resistance and susceptibility consistent with Mendelian expectations of a single major gene controlling resistance, which was also validated in backcrosses segregating 0:1 or 1:1 between resistance and susceptibility. The F2 phenotypic distributions appeared strongly skewed and those of the backcrosses followed a bimodal distribution confirming that a major locus is involved. In this study, similar populations were further used for genetic linkage analyses to identify SSR markers associated with resistance to MF and to elucidate the genetic architecture of the trait. The interspecific-hybrid-derived populations consisting of 23 to 43 individuals each were genotyped with 34 to 36 SSR markers distributed at

6.2 to 25.1 cM intervals across eight linkage groups based on the Prunus TxE reference map. Based on the assumption of a single locus with major phenotypic effect, the population size and marker density for our mapping populations were deemed sufficient for initial QTL detection in the presence of large unrecombined linkage blocks characteristic of early-generation populations (Flint-Garcia et al., 2003; Mackay et al.,

2009).

The QTL analysis in multiple independent populations detected one SSR locus associated with the resistance QTL corroborating our previous findings of single-locus inheritance from the progeny tests. The QTL was localized in the same genomic region that harbored the genetic factors controlling M. incognita and M. arenaria in P. kansuensis, as well as in P. persica cultivars Shalil, Nemared, and Juseitou (Cao et al.,

2011; Claverie et al., 2004a; Yamamoto et al., 2005) and mapped close to the SSR marker UDP98-025 at the telomeric region of LG 2. The UDP98-025 marker amplified three different fragment sizes in PK and PP genotypes. The 115-bp allele in PK was associated with resistance while the 111-bp (OK and FG) and 127-bp (SH) allele sizes in P. persica genotypes were associated with susceptibility. The use of such a codominant SSR marker, which can distinguish heterozygotes, in examining genotypic differences based on GI values enabled validation of the proposed single-locus dominant/recessive model from the segregation analyses. The QTL was confirmed to be completely dominant when levels of root galling in PK-homozygotes were similar to the heterozygotes and significantly lower than those of the PP-homozygotes at the linked marker locus. The QTL was confirmed to be recessive when levels of root galling in PK-homozygotes were significantly lower than those of the heterozygotes, whereas the heterozygotes have levels similar to those of PP-homozygotes. Although the effect of the recessive QTL was not apparent in the F2 due to the low frequency of resistant individuals and some recombination occurring between the QTL and the marker locus, the recessive effect was confirmed in the backcrosses, which showed non-significant differences between the heterozygous and PP-homozygous marker genotypes. The study also demonstrated the advantage of observing multiple F2 interspecific

populations to capture the two possible allelic combinations in segregating F1 hybrids

(Chaparro et al., 1994). The OK x PK and FG x PK hybrids each inherited a resistance allele from PK (designated as Mf1 or mf3) and a susceptible allelic form from PP

(designated as mf2), and allelic relationships were inferred based on segregation patterns in the F2 and backcross populations. Table 3-4 illustrates the relationship of the three different alleles at the same locus and their effects on resistance. Multiple alleles conferring high resistances to major RKN have also been identified in Myrobalan plum (P. cerasifera) by examining segregation in several intraspecific F1 and F2 progenies of crosses between heterozygous and homozygous accessions (Esmenjaud et al., 1996b; Rubio-Cabetas et al., 1998).

The codominant SSR marker was useful for detecting the resistance alleles especially in the heterozygotes and for tracking the transmission of resistance alleles via linkage to the backcross progenies when the genetic model is known. However, its use as a diagnostic marker for MF resistance in a breeding program will be complicated if the PK accession is heterozygous at the resistance locus. If resistance is recessively inherited in the progeny, several individuals carrying the 115-bp marker allele may not have the resistance allele and therefore, will be incorrectly classified as resistant. It would be necessary to identify other co-dominant markers within the region that can differentiate the two resistance alleles in PK. However, if a homozygous-dominant PK accession is identified and used to make crosses with homozygous-susceptible PP genotypes, the UDP98-025 marker can predict resistance with high accuracy (≈98%) among F2 individuals that are homozygous for the 115-bp allele.

In this study, the distance of the linked SSR marker locus UDP98-025 from the

MF resistance locus spans a large chromosomal segment corresponding to an interval size of 7.2 cM in F2 and 14.5 cM in backcrosses. The same marker was placed ≈10 cM from the RMia gene on the TxE reference map (Dirlewanger et al., 2004b) and from the

PkMi gene using a BC1F1 population derived from a cross between peach ‘Bailey’ and

P. kansuensis ‘Honggengansutao’ (Cao et al., 2011). These RKN R genes located on

LG 2, along with those located on the telomeric region of LG 7 (Ma, Rjap, and RMja)

(Cao et al., 2011; Claverie et al., 2004a; Van Ghelder et al., 2010; Yamamoto et al.,

2005) have been characterized as a class of toll/interleukin1 receptor-(TIR) nucleotide- binding-(NB) leucine-rich repeat (LRR) genes (Cao et al., 2014a; Claverie et al., 2011;

Duval et al., 2014). The MF resistance gene in PK mapped in this study broadens the suite of R genes on the top region of chromosome 2, which has been found to contain a fivefold higher density of genes encoding disease-resistance-associated NB-LRR proteins than the peach genome average (Verde et al., 2013). Other codominant markers such as sequence-tagged sites and SSRs (AMP117 and AMP116) flanking

RMia or PkMi on LG 2 as well as resistance gene analogues have been used to examine polymorphisms among Prunus accessions (Cao et al., 2014a; Duval et al.,

2014; Lu et al., 1999; Yamamoto et al., 2005). The same markers may also be used to examine the level of allelic variation at the MF resistance locus among PK accessions and in Prunus rootstock breeding germplasm due to collinearity between Prunus genomes (Dirlewanger et al., 2004b). Whether the Mf, RMia, and PkMi are distinct genes that are tightly linked in a single locus or are part of a single, multi-allelic resistance locus is an important question that remains unresolved and a complexity not

directly addressed in this study. Further investigations involving different RKN species and isolates are needed to establish the resistance specificities of the Mf alleles and to determine their relationship with the possibly linked or interacting R genes identified in

LG 2.

The genetic maps constructed from the twelve interspecific peach x P. kansuensis populations showed variability in recombination rates across the genome inherent in the interspecific cross. The combined genetic maps provide information about the recombination landscape in the early-generation populations and will be useful for further linkage mapping studies in identifying genomic regions that may require increased number of markers or increased number of individuals to capture more recombination events when analyzing interspecific-hybrid-derived F2 and backcross populations. The map lengths for backcross populations were generally longer across eight linkage groups than those of F2 indicating that more recombination has occurred in the former, which can be utilized further to achieve a higher-resolution linkage map particularly at the target QTL interval on the top region of LG 2 where recombination distance was twofold longer.

Among F2 genetic maps, there was a general trend of reduced map length at LG

6 (shorter than the TxE reference map by 53 to 56%) attributed to the segregation- distorted markers at the distal region (BPPCT025 and CPPCT030). Segregation patterns in the F2 progenies consistently indicated higher frequency of peach alleles and a lack of PK-homozygotes. The CPPCT030 marker locus has been mapped within the distal region of LG 6, which is known to contain the self-incompatibility (S) locus in

Prunus (Dirlewanger et al., 2004b). Assuming linkage between CPPCT030 and S

locus, the absence of PK-homozygotes in peach x PK F2 progenies could be explained by pollen-pistil incompatibility due to similar S determinants from PK inherited in selfed progeny. The gametophytic self-incompatibility system, whereby matching S determinants in the pollen (S locus F-box protein, SFB) and pistil (S-RNase) inhibit pollen tube growth, is known to occur in most Prunus species and a mutation in the pollen SFB gene confers self-compatibility in peach (Tao and Iezzoni, 2010; Tao et al.,

2007). When the peach x PK F1 hybrids (putatively heterozygous at the S locus) were backcrossed to self-compatible peach, the homozygotes and heterozygotes segregated in the expected 1:1 ratio at the CPPCT030 locus indicating the absence of gametophytic selection. A similar phenomenon in which selection acted in favor of peach alleles and against homozygotes of the other parent was also observed in interspecific F2 populations derived from crosses of peach with almond cv. Texas (Joobeur et al.,

1998), almond cv. Padre (Bliss et al., 2002), and P. davidiana (Foulongne et al., 2003).

Gametophytic self-incompatibility was cited by the authors as a possible cause for the segregation bias at the distal marker in LG 6 against alleles of almond and P. davidiana, which are both known to be preferentially allogamous and highly heterozygous. A self- incompatibility mating system in PK that promotes outcrossing (Kao and Huang, 1994), would seem to explain the deficit of PK-homozygotes in F2 progenies examined in this study and is reflected in the heterozygosity at the CPPCT030 marker locus (239 / 249- bp alleles) as well as in ≈50% of the polymorphic SSR loci screened.

The genetic analyses have revealed the effects of the resistance alleles from PK as tracked by the linked, codominant SSR marker and are consistent with the suggested dominance/recessive model from the segregation analyses. Normal

segregation was generally observed at polymorphic SSR loci among progeny of the interspecific cross except at the locus associated with self-incompatibility in Prunus that consistently showed severe distortion favoring peach alleles in F2 progenies. No evidence for distorted segregation was detected in LG 2 indicating that the introgression of MF resistance from PK into cultivated peach is predictable by single-gene-based

Mendelian inheritance.

Table 3-1. Segregation in F2 and BC1F1 interspecific peach x Prunus kansuensis populations for resistance to Meloidogyne floridensis ‘MFGnv14’ isolate based on root galling index and test of the data to single-gene ratio. Mean GI values ± SD (N) c Proposed allelic Test ratio Chi square Generation Source population a N b combination of the R (GI ≤ 2) S (GI > 2) R : S d P-value d cross

F2 [OK x PK]1 36 0.00 ± 0.00 (25) 4.64 ± 0.50 (11) 3:1 0.441 Mf1 mf2 (selfed) [OK x PK]4 29 0.37 ± 0.68 (19) 4.70 ± 0.67 (10) 3:1 0.238 [OK x PK]5 22 0.00 ± 0.00 (18) 5.00 ± 0.00 (4) 3:1 0.460 [OK x PK]6 28 0.00 ± 0.00 (22) 4.83 ± 0.41 (6) 3:1 0.662 [FG x PK]1 29 0.00 ± 0.00 (22) 5.00 ± 0.00 (7) 3:1 0.915 [FG x PK]6 25 0.10 ± 0.45 (20) 4.60 ± 0.55 (5) 3:1 0.564 Combined 169 0.07 ± 0.34 (126) 4.77 ± 0.48 (43) 3:1 0.894

F2 [OK x PK]3 35 1.57 ± 0.79 (7) 4.82 ± 0.48 (28) 1:3 0.494 mf2 mf3 (selfed)

BC1F1 SH x [OK x PK]2 33 0.07 ± 0.26 (15) 4.72 ± 0.57 (18) 1:1 0.602 mf2 mf2 x Mf1 mf2 SH x [FG x PK]1 18 0.00 ± 0.00 (8) 4.80 ± 0.63 (10) 1:1 0.637 SH x [FG x PK]6 40 0.00 ± 0.00 (23) 4.94 ± 0.24 (17) 1:1 0.343 Combined 91 0.02 ± 0.15 (46) 4.82 ± 0.49 (45) 1:1 0.916

BC1F1 SH x [OK x PK]3 30 0.00 ± 0.00 (3) 4.93 ± 0.27 (27) 0:1 - mf2 mf2 x mf2 mf3 FG x [OK x PK]3 20 (0) 4.80 ± 0.31 (20) 0:1 - Combined 50 0.00 ± 0.00 (3) 4.91 ± 0.28 (47) 0:1 - a PK = P. kansuensis wild peach; OK = ‘Okinawa’ peach; FG = ‘Flordaguard’ peach; SH = ‘UFSharp’ peach. b Only true selfs or backcrosses verified through microsatellite marker profiling were included in the segregation analysis. Individual plants were evaluated 120 days after nematode inoculation at a population density of 10,000 eggs per plant. c GI, galling index; SD, standard deviation; N, number of individuals analyzed. Classification into resistant (R) and susceptible (S) types based on number of galls per whole root system, where 0 = no galls, 1 = one to two galls, 2 = three to 10, 3 = 11 to 30, 4 = 31 to 100, and 5 = more than 100; GI values ≤ 2 indicate resistance (Taylor and Sasser, 1978). d R resistant, S susceptible; chi square probability (P) values for testing goodness-of-fit to expected ratios of R and S at the 0.05 significance level.

Table 3-2. Associations between genotypes at UDP98-025 locus and Meloidogyne floridensis (‘MFGnv14’ isolate) resistance with three alleles segregating in F2 and BC1F1 populations derived from peach x Prunus kansuensis crosses. Genotype cross a a b 2 c combination N K P Mean galling index ± SD (N) by genotype R (%)

Mf1 mf2 x Mf1 mf2 B: Mf1 Mf1 H: Mf1 mf2 A: mf2 mf2 [OK x PK]1 36 12.065 0.0024 0.50 ± 1.41 (8) 0.72 ± 1.67 (18) 3.40 ± 2.37 (10) 32.8 [OK x PK]4 29 19.441 <.0001 0.57 ± 0.79 (7) 0.71 ± 1.33 (14) 5.00 ± 0.00 (8) 80.4 [OK x PK]5 22 15.042 0.0005 0.00 ± 0.00 (5) 0.36 ± 1.34 (14) 5.00 ± 0.00 (3) 71.6 [OK x PK]6 28 6.462 0.0395 0.00 ± 0.00 (6) 0.82 ± 1.85 (17) 3.00 ± 2.74 (5) 23.9 [FG x PK]1 29 14.567 0.0007 0.00 ± 0.00 (9) 0.71 ± 1.82 (14) 4.17 ± 2.04 (6) 52.0 [FG x PK]6 25 12.849 0.0016 0.00 ± 0.00 (9) 0.55 ± 1.29 (11) 3.80 ± 2.17 (5) 58.7 Combined d 169 75.881 <.0001* 0.18 ± 0.69 (44) b 0.66 ± 1.55 (88) b 4.00 ± 1.97 (37) a 49.3

mf2 mf2 x Mf1 mf2 H: Mf1 mf2 A: mf2 mf2 SH x [OK x PK]2 33 17.496 <.0001 1.30 ± 1.98 (20) 4.62 ± 1.39 (13) 47.1 SH x [FG x PK]1 18 9.491 0.0021 0.89 ± 1.83 (9) 4.44 ± 1.67 (9) 53.7 SH x [FG x PK]6 40 24.536 <.0001 0.24 ± 1.09 (21) 4.16 ± 1.86 (19) 64.0 Combined d 91 50.815 <.0001* 0.78 ± 1.67 (50) b 4.37 ± 1.65 (41) a 54.1

mf2 mf3 x mf2 mf3 B: mf3 mf3 H: mf2 mf3 A: mf2 mf2 [OK x PK]3 35 9.110 0.0105 3.27 ± 1.79 (11) 4.33 ± 1.23 (15) 5.00 ± 0.00 (9) 22.4

mf2 mf2 x mf2 mf3 H: mf2 mf3 A: mf2 mf2 SH x [OK x PK]3 30 5.035 0.0248 3.94 ± 1.98 (16) 5.00 ± 0.00 (14) 12.5 FG x [OK x PK]3 20 1.727 0.1888 4.82 ± 0.40 (11) 5.00 ± 0.00 (9) 9.1 Combined 50 6.757 0.0093 4.30 ± 1.59 (27) 5.00 ± 0.00 (23) 8.6 a Kruskal-Wallis K statistic and corresponding P-value. Significant differences (*) found at P <0.005 level or at P <.0017 level for multiple pairwise comparisons. b A: homozygote with both alleles from peach, H: heterozygotes, B: homozygote with both alleles from P. kansuensis. Combined means followed by different letters indicate significant differences among mean ranks of genotypic classes. c Percentage of the phenotypic variation explained by the UDP98-025 marker. d No significant interaction effects between family and marker genotypic class, based on two-way analysis of variance at 95% confidence level.

Table 3-3. Co-segregation of Meloidogyne floridensis-resistant genotypes with alleles at the UDP98-025 marker locus. No. of Genotype cross N Proportion (%) of predicted phenotypes * recombinants combination (%)

Mf1 mf2 x Mf1 mf2 B: Mf1 Mf1 % R H: Mf1 mf2 % R A: mf2 mf2 % S [OK x PK]1 36 7 (8) 87.5 15 (18) 83.3 8 (10) 80.0 6 (16.7) [OK x PK]4 29 7 (7) 100.0 12 (14) 85.7 8 (8) 100.0 2 (6.9) [OK x PK]5 22 5 (5) 100.0 13 (14) 92.9 3 (3) 100.0 1 (4.5) [OK x PK]6 28 6 (6) 100.0 14 (17) 82.4 3 (5) 60.0 5 (17.9) [FG x PK]1 29 9 (9) 100.0 12 (14) 85.7 5 (6) 83.3 3 (10.3) [FG x PK]6 25 9 (9) 100.0 10 (11) 90.9 4 (5) 80.0 2 (8.0) Combined 169 43 (44) 97.7 76 (88) 86.4 31 (37) 83.8 19 (11.2)

mf2 mf2 x Mf1 mf2 H: Mf1 mf2 % R A: mf2 mf2 % S SH x [OK x PK]2 33 14 (20) 70.0 12 (13) 92.3 7 (21.2) SH x [FG x PK]1 18 7 (9) 77.8 8 (9) 88.9 3 (16.7) SH x [FG x PK]6 40 20 (21) 95.2 16 (19) 84.2 4 (10.0) Combined 91 41 (50) 82.0 36 (41) 87.8 14 (15.4)

mf2 mf3 x mf2 mf3 B: mf3 mf3 % R H: mf2 mf3 % S A: mf2 mf2 % S [OK x PK]3 35 4 (11) 36.4 12 (15) 80.0 9 (9) 100.0 10 (28.6)

mf2 mf2 x mf2 mf3 H: mf2 mf3 % S A: mf2 mf2 % S SH x [OK x PK]3 30 13 (16) 81.2 14 (14) 100.0 3 (10.0) FG x [OK x PK]3 20 11 (11) 100.0 9 (9) 100.0 0 (0.0) Combined 50 24 (27) 88.9 23 (23) 100.0 3 (6.0)

* Genotype classes: A, homozygote with both alleles from peach, H, heterozygotes, B, homozygote with both alleles from Prunus kansuensis. Number of individuals observed per genotypic class and the expected number of individuals, enclosed in parentheses, if the marker is tightly linked with the resistance locus. R, resistant; S, susceptible.

Table 3-4. Allelic relationships at the Mf locus and corresponding phenotypic effects. Allele Relationship * Genotype Response to Meloidogyne floridensis

Mf1 > mf2 > mf3 Mf1 Mf1 Resistant Mf1 mf2 Resistant mf3 mf3 Resistant mf2 mf3 Susceptible mf2 mf2 Susceptible

* Mf1 and mf3 represent resistance alleles from Prunus kansuensis while mf2 represents susceptible allelic forms from Prunus persica.

Mf1 mf2 x Mf1 mf2 mf2 mf3 x mf2 mf3

N = 169 N = 35 169

mf2 mf2 x Mf1 mf2 mf2 mf2 x mf2 mf3

N = 91 N = 50

Figure 3-1. Histograms of the resistance response to a pathogenic isolate of Meloidogyne floridensis (‘MFGnv14’) among twelve interspecific peach x Prunus kansuensis progenies grouped according to the parental cross. The populations segregate for dominant (Mf1) or recessive (mf3) resistant alleles from P. kansuensis and recessive susceptible (mf2) alleles from peach. The top and bottom panels show the distributions of root galling index among seven F2 progenies and five BC1F1 progenies, respectively.

A Combined OP1 OP4 OP5 OP6 FP1 FP6

0.0 Mf 0.0 Mf 0.0 Mf 0.0 Mf 0.0 Mf 0.0 Mf 0.0 Mf 0.8 UDP98-025 1.7 UDP98-025

UDP98-025 5.9 UDP98-025 7.2 UDP98-025 6.6 BPPCT013 6.3 UDP98-025 11.3 UDP96-013 10.4 13.6 UDP98-025 13.5 BPPCT030

BPPCT013 20.8 21.5 BPPCT013 21.2 BPPCT013 UDP96-013 22.8 BPPCT013 23.4 UDP96-013 BPPCT030 24.4 UDP96-013 22.9 26.4 UDP96-013 27.4 BPPCT030 27.7 BPPCT030 28.8 BPPCT013 30.0 BPPCT030

BPPCT013 36.6 UDP96-013 37.2 37.1 CPSCT034 38.5 BPPCT030 UDP96-013 41.6 CPSCT034 40.9 BPPCT030

52.4 CPSCT034

60.4 CPSCT034

Figure 3-2. Comparison of marker distances for linkage group 2 and approximate position of Meloidogyne floridensis resistance locus from combined maps and component maps within each cross type. A. Mf1 mf2 x Mf1 mf2 : OP ‘Okinawa’ x P. kansuensis, FP ‘Flordaguard’ x P. kansuensis (letters followed by accession number of the F1 parent); B. mf2 mf3 x mf2 mf3; and C. mf2 mf2 x Mf1 mf2: OP2S ‘UFSharp’ x (‘Okinawa x P. kansuensis), FP1S ‘UFSharp’ x (‘Flordaguard’ x P. kansuensis), FP6S ‘UFSharp’ x (‘Flordaguard’ x P. kansuensis). Lines between linkage maps connect common markers.

B C OP3 Combined OP2S FP1S FP6S

0.0 Mf 0.0 Mf 0.0 Mf 0.0 Mf 0.0 Mf

8.5 UDP98-025

14.5 UDP98-025 17.3 UDP98-025 20.6 UDP98-025 24.6 UDP98-025

BPPCT013 30.7 BPPCT013 33.5 BPPCT013 32.6 UDP96-013 33.4 UDP96-013 34.9 UDP96-013 38.6 BPPCT030 42.5 BPPCT030 45.2 BPPCT030 45.9 CPSCT034 BPPCT013 48.4 UDP96-013 48.9 CPSCT034 49.2 BPPCT013

55.7 BPPCT030 54.8 BPPCT030

CPSCT034 66.8

Figure 3-2. Continued.

CHAPTER 4 HORTICULTURAL EVALUATION OF POTENTIAL ROOTSTOCKS FOR LOW-CHILL PEACH PRODUCTION DURING INITIAL YEARS OF FIELD ESTABLISHMENT

Introduction

Low-chilling requirement and early-ripening ability were the earliest main goals of the stonefruit breeding program at the University of Florida (UF) and continue to be important considerations in the development of suitable varieties for Florida’s subtropical climate and market niche for early-season peaches. To produce higher- quality peaches, the non-melting flesh character was bred into adapted germplasm and a series of low-chill, non-melting flesh varieties have been released since the early

1990s. These varieties were envisioned to promote demand for fresh market peaches because the non-melting trait facilitated harvesting of tree-ripe fruit with better quality and sufficient firmness for handling and shipping to distant markets. Early breeding work involved several crosses using the non-melting peach germplasm from Central

Mexico, southeastern United States, and southern Brazil as well as superior nectarine and melting flesh genotypes to progressively shorten the fruit developmental period

(FDP) from about 120-150 days to about 60-110 days. These fruit cultivars have cropped successfully on a nematode-resistant, red-leaf type ‘Flordaguard’ seedling rootstock (Andersen et al., 2001; Sherman et al., 1996; Sherman et al., 1990).

Breeding for earliness and higher fruit quality in peach also takes into account many horticultural traits for commercial utility; some advances have been made in cultivars that have increased flower bud density and fruit set, reduced incidence of blind nodes and bud drop, compact tree structure to reduce labor demands for the grower, improved resistance to bacterial spot and leaf rust, and improved rootstock resistance to root-knot nematodes (Andersen et al., 2001; Sherman and Lyrene, 2003).

Low-chill peach cultivars bred at UF have shown adaptation in other peach production regions worldwide where mean January temperatures range from 11.3 to

17.5 °C (52.3 to 63.5 °F) (Sharpe and Sherman, 1990). In central and south central

Florida where January temperatures generally range from 13 to 18 °C (55 to 65 °F), early season cultivars bloom by late January or early February and permit harvesting during early April through late May, attaining premium prices when there is practically no competition from domestic and foreign peach producers (Morgan and Olmstead,

2013b). The distinct marketing advantage offered by the low-chill-adapted cultivars producing high-quality fruit has encouraged some Florida citrus growers to consider peach as an alternative crop. The production acreage for this specialty crop has expanded from 234 acres in 2007 (USDA, 2009) to an estimated 1,400 in acres in 2016 mostly in central and south central Florida (M. Olmstead pers. comm.).

With several early-ripening fruit cultivars available for commercial production, the

Florida peach industry still relies on a single rootstock variety. The former rootstocks

‘Okinawa’, ‘Nemaguard’, and ‘Nemared’ had been found to be susceptible to an endemic root-knot nematode, Meloidogyne floridensis, also known as the peach root- knot nematode (RKN). Since its release as a public variety in 1991, ‘Flordaguard’ has been the only rootstock recommended for low-chill peach production in Florida because it is adapted to non-alkaline soils and performs well on sandy soils commonly infested with nematodes (Handoo et al., 2004; Olmstead et al., 2016; Olmstead et al., 2015;

Sherman et al., 1991). New rootstock cultivars are being developed to diversify the panel of improved rootstocks for an expanding peach industry. Primary criteria for selection of new rootstocks for southeastern peach production regions are vigor control

and combined resistances to root-knot nematodes, Armillaria root rot (ARR), and peach tree short life (PTSL), on which extensive breeding work has been done at the

Southeastern USDA Agricultural Research Station (ARS) in Georgia for over twenty years (Beckman and Chaparro, 2015). Two new rootstocks, ‘MP-29’ and ‘P-22’, developed by Dr. Thomas Beckman at the USDA-ARS (Byron, Georgia) were evaluated in this study. These rootstocks showed good resistance to root-knot nematodes in field plantings in Byron, Georgia as well as tolerance to PTSL and ARR, thus offering potential for commercial utilization in Florida. ‘MP-29’ is a clonal interspecific hybrid between a red-leafed peach rootstock breeding selection and ‘Edible Sloe’ plum

(Beckman et al., 2012). The ‘P-22’ clonal selection is a ‘Guardian’ x ‘Flordaguard’ peach hybrid and is under consideration for release.

Several reports have documented the contribution of rootstocks to the success of an orchard. Effects of rootstocks on peach scion have been seen on bloom date and fruit maturation (Beckman et al., 1992; Durner and Goffreda, 1992), on vigor and yield

(Ben Yahmed et al., 2016; Bussi et al., 2002; Bussi et al., 1995; Caruso et al., 1997;

Marra et al., 2013; Moreno et al., 1994), and on fruit quality (Font i Forcada et al., 2012;

Font i Forcada et al., 2013; Giorgi et al., 2005). From multi-state trials, location-specific effects of the rootstocks on tree size, fruit maturity, and fruit size have been reported

(Reighard et al., 2011), underscoring the need for testing new rootstock releases before recommending them for commercial planting in specific production areas.

In this study, we compared the horticultural performance of the new rootstocks in terms of tree growth, yield, and fruit quality to those of low-chill-adapted varieties during the first three years of orchard establishment to determine their potential for commercial

utilization in a subtropical environment. This study was carried out in Citra, Florida with a low-chill, non-melting peach scion (‘UFSun’) budded on five different rootstocks. In addition to the two new rootstocks, ‘Barton’, a feral-peach-seedling-derived rootstock from Australia with nematode resistance and favorable horticultural characteristics, was included in the experiment to test for local adaptation and assess its potential as breeding material. ‘Okinawa’ seedling rootstock, an obsolete rootstock due to its susceptibility to M. floridensis, and ‘Flordaguard’ seedling rootstock (industry standard) were included as reference cultivars for low-chill adaptation.

Materials and Methods

Rootstocks and Tree Preparation

All trees were budded with ‘UFSun’, a commercial low-chill (100 cu) clingstone and non-melting fresh market peach developed by the UF stonefruit breeding program.

‘UFSun’ is a popular early season cultivar with an average FDP of 80 d when budded on ‘Flordaguard’ rootstock (Rouse and Sherman, 2004). The five rootstocks chosen for the experiment were: ‘Flordaguard’ (standard rootstock cultivar for Florida peach production), ‘Okinawa,’ and ‘Barton’ seedling rootstocks, as well as ‘MP-29’ and ‘P-22’ clonal hybrid rootstocks. Their origin and characteristics as unbudded trees are described in Table 4-1. These rootstocks were obtained in 2012 as bare-root and container-grown plants from various nurseries. The clonal rootstocks ‘MP-29’ and ‘P-

22’ were provided as two-year-old unbudded and budded plants, respectively, by T. G.

Beckman from the stonefruit rootstock breeding program at USDA-ARS (Byron,

Georgia). Container-grown plants of ‘Barton’ were available as one-year-old seedlings from a screenhouse nursery in Gainesville, FL. ‘Barton’ and ‘MP-29’ were budded with

‘UFSun’ in the summer of 2012. ‘Flordaguard’ and ‘Okinawa’ were supplied as two-year-

old budded trees (Island Grove Ag Products, Hawthorne, FL). In the spring of 2014, trees were removed from the 5-gal pots and transplanted to the field. Prior to transplanting, trees were headed at 75-80 cm and pruned to an open-center form with four to five scaffold limbs.

Field Site and Experimental Design

Rootstock evaluations were conducted from 2014 to 2016 at the UF Plant

Sciences Research and Education Unit (PSREU) in Citra, FL (29.4°N, 82.1°W). The north central Florida region typically receives 200 to 350 chill hours between 0 to 7.2°C

(32 to 45 °F). The soil type at the research site is classified as Arredondo sand (95% sand, 2% silt, 3% clay) which is mostly well-drained (Thomas et al., 1979). The experiment was laid out in a randomized complete block design, consisting of ten replicate rows in north to south orientation. In each row, thirty ‘UFSun’ trees on five different rootstocks were arranged such that six trees on similar rootstock were planted adjacent to each other to constitute a single replicate. The trees were distributed such that the sizes within a rootstock variety were as similar as possible in the row. Three hundred trees were distributed at a close spacing of 4 ft between trees within the rows and 20 ft between the rows to maximize utilization of space during the first two years.

The experimental trees were planted more closely than the 10 to 15 ft within-row spacing in commercial orchards (Morgan and Olmstead, 2013a). In January 2016, every alternate tree in the row as well as weak, diseased trees were removed to reduce crowding as trees grew larger; this resulted in a spacing of 8 to 16 ft between trees within the row during the third year of field establishment (Supplementary Figure C-1).

Cultural Management

Individual trees were irrigated with under-tree microsprinklers (Max-Cone Fan part no. MAO36D1, Maxijet, Dundee, FL) and fertilized each year with a ring application of NPK fertilizer (10-10-10 plus micronutrients) at a rate of 1 lb per inch of rootstock trunk diameter (maximum of 3 lbs per tree) in February, June, and August. Overhead sprinklers were installed to provide frost protection for the crop when necessary. Each year, trees were summer-pruned (after harvest in early June) and dormant-pruned (mid-

December to early January) to maintain an open center and to keep tree height and canopy diameter uniform across the block. All trees were pruned to maintain a limb height of approximately 6 ft. Trees were fruit-thinned before the pit-hardening stage (≤

1-inch diameter) to a distance of 4 to 6 inches between fruit. Weed, pest and disease control were in accordance with the Southeastern peach, nectarine and plum pest management guide (Horton et al., 2015).

Data Collection

Vegetative growth measurements

Prior to establishment of the trial in March of 2014, vegetative growth was assessed as crown height, average crown diameter (calculated average of two perpendicular crown diameter measurements), stem circumference, and pruning weight.

Stem circumference was measured at 2 inches above and below the graft union using a measuring tape. Pearson correlation analyses revealed that these growth parameters were all strongly correlated (Supplementary Tables C-1 and C-2) indicating that stem circumference is a reliable measure of tree size. Accordingly, only rootstock/scion stem

(trunk) circumference were recorded for each year of establishment during the dormant period (early January) and pruning weights (early January and June) to evaluate

100

rootstock effects on scion vigor. Trunk circumference measurements were converted to trunk cross sectional area (TCSA) values by deriving the trunk diameter and then using the formula: TCSA = (trunk diameter / 2)2 x 3.1416. Because the rootstock planting materials came from different nurseries with different propagation methods (e.g. seedling vs. cuttings) and growing times, stem circumferences varied significantly among trees on different rootstocks (Supplementary Table C-1). Thus, growth comparisons were made based on relative growth rates of the trunk rather than absolute measurements. Trunk relative growth rate (RGR) or change in TCSA relative to initial trunk circumference at pre-transplanting was expressed as: (TCf – TCi)/ TCi, where TCf is the calculated TCSA from trunk circumference measurements during the first (2014), second (2015), or third year (2016) of establishment.

Trees were headed to establish the scaffold system for an open center in

January 2014 prior to transplanting and the pruning weights for each tree were recorded to indicate tree vigor at first year of pruning. Subsequent summer and dormant season prunings were recorded as cumulative pruning weights for second year (June 2014 and

January 2015) and third year (June 2015 and January 2016). Trees were pruned to a maximum height of 6 ft and a crown spread of up to 4 ft within the row during the first and second year. With the wider tree spacing in the third year, lateral branches were cut to a crown width of 6 ft. Each year, maintenance pruning was conducted to remove water sprouts, crossing branches and low-hanging shoots as well as dead and diseased limbs.

Bloom and harvest dates

Bloom progression on individual trees was monitored at least three times per week in the second (2015) and third year (2016). Percentage bloom was visually

101

estimated on a whole-tree basis and full bloom dates (≈50% blooms open) were recorded. During harvest period from late April to mid-May, tree yield (in grams) for each harvest date was recorded and the number of days from full bloom to mid-harvest

(i.e. when harvested fruit reached 50% of total number of fruit per tree) were calculated.

Heat accumulation, expressed in terms of growing degree days (GDD), 30 days after full bloom (GDD30, which coincides with early fruit development) and from full bloom to mid-harvest were calculated using the daily degrees calculator from AgroClimate

() with 50 °F as base temperature. In addition, air temperature data from 2014-2016 measured by a 2-meter sensor at the weather station located at the PSREU research site in Citra, FL were downloaded from the database of the Florida Automated Weather Network

() to plot the accumulated chill hours (32 to 45 °F) from October

1 to April 30.

Fruit harvest data

Fruits were picked when the ground color changed to yellow and started to soften

(approximately 2 pounds·force when tested with a fruit penetrometer). Fruits were counted and weighed at each harvest date in 2014, 2015, and 2016. Values for mean fruit weight, yield efficiency, and fruit load were calculated on an individual tree basis.

Mean fruit weight was calculated by dividing the total yield (harvested fruit weight in grams) by the total number of fruits. Yield efficiency (kg·cm-2) was estimated by dividing the total yield (in kg) by the derived TCSA of scion (in cm2). Fruit load was expressed as fruit number per cm2 of scion TCSA. In the third harvest season (2016), fruit weight and size distributions per harvest date were recorded. Using manual sizing rings, fruits

102

were sorted into the following size classes: ≤ 2 1/4”, 2 5/16”, 2 3/8”, 2 7/16”, 2 1/2”, 2

5/8”, 2 3/4”, 2 7/8”, and ≥ 3” in diameter.

Five representative fruits per tree were sampled at mid-harvest to assess fruit quality attributes of individual fruits such as: weight, diameter, total soluble solids (°Brix) and flesh firmness (pounds·force, lbf). Individual fruit size was estimated by averaging the cheek and suture diameter (mm) as measured by a digital caliper. Total soluble solids content was determined by a digital hand-held refractometer (Pocket

Refractometer PAL-1, Atago Co., Tokyo, Japan) using the juice squeezed from one cheek of the fruit. The fruit was peeled at the opposite cheek to expose approximately

1- to 1.5-cm diameter of flesh and firmness was tested using a hand-held fruit penetrometer (Model FT 30, Wagner Instruments, Massachusetts, USA) with 0.8 cm plunger tip.

Data Analyses

Data were fitted to linear mixed models using the PROC GLIMMIX statement in

SAS software (SAS Institute Inc., Cary, North Carolina, USA). The rootstock cultivar, year and cultivar x year interaction were considered as fixed effects while block, block x cultivar, block x year were treated as random effects. To adjust for missing observations, least-squares means were generated by the analyses using the

LSMEANS statement. Comparisons on growth, yield, phenology, and fruit quality among rootstocks were made for each year due to significant rootstock x year interactions for almost all studied parameters (Supplementary Table C-3). All significant differences were at P ≤ 0.05 level.

103

Results

Tree Growth

The rootstock and scion displayed different trunk growth rates across three years

(Table 4-2). During the first year, there was a greater increase in trunk growth rates of younger trees on ‘MP-29’ and ‘Barton’ than older trees on ‘Flordaguard’, ‘Okinawa’, and

‘P-22’. The trunk growth trends were similar during the second and third year except that the trunk growth rate of ‘Flordaguard’ rootstock became significantly lower than those of ‘Okinawa’ and ‘P-22’. The trunk size of ‘MP-29’ and ‘Barton’ rootstock increased in the second and third year at a rate comparable to that of ‘Flordaguard’ while the scion trunks generally increased at a greater rate in younger trees than in older trees.

The pruning weights did not vary as much as the trunk growth rates among the rootstocks and such is partly due to uniform constraints we had set on the limb height and crown diameter for all trees according to the allocated within-row spacing. As trees grew larger, there was a greater amount of pruning to constrain the trees within the allocated space. Trees on ‘MP-29’ had the lowest pruning weights in the third year which, in the preceding years, did not differ from ‘Flordaguard’. In all three years, ‘MP-

29’ consistently showed lower pruning weights than vigorous trees on ‘Okinawa’. Trees on ‘Barton’, which initially had similar scion trunk size as those on ‘MP-29’

(Supplementary Figure C-2), tend to produce more vigorous vegetative growth as shown by the greater pruning weights compared to those on ‘MP-29’ in the third year.

Phenological Observations

Bloom dates of ‘UFSun’ varied between February 24 to March 2 in 2016 (third year) based on the rootstock (Table 4-3). Trees on ‘P-22’, ‘Okinawa’, and ‘Flordaguard’

104

had the earliest peak bloom dates (February 24 to 26) followed by those on ‘Barton’

(February 28), whereas those on ‘MP-29 were the latest to reach peak bloom dates

(March 2). Similar trends in bloom time occurred during the previous year with peak bloom dates recorded between February 17 to 22. However, full bloom dates in 2016 were delayed by a week as a consequence of later and slower chilling accumulation, which also led to later harvest dates for that year. By January 31, trees have received

371 chill hours in 2015 and only 159 chill hours in 2016 (Figure 4-1). Harvest periods extended over four weeks beginning late April with harvests reaching 50% threshold for fruit yield (mid-harvest) between May 6 to 11 in 2015 and May 13 to 17 in 2016 depending on the rootstock (Table 4-3). Harvests in mid-May (2015 May 13 and 2016

May 17) accounted for the largest portion (40 to 55%) of total fruits for trees on

‘Flordaguard’, ‘Okinawa’, ‘Barton’, and ‘P-22’ (Figure 4-2). On the average, mid- harvest date for late-blooming trees on ‘MP-29’ was delayed by four days relative to early-blooming trees on ‘Okinawa’ and ‘P-22’. Trees on ‘MP-29’ had a more concentrated bloom and harvest period with 73 to 80% of total fruits picked at mid- harvest.

The periods between bloom and harvest (fruit developmental period) was significantly shorter on ‘MP-29’ than on ‘Flordaguard’, which seemed to be due to higher heat accumulation during thirty days after bloom (GDD30). The bloom time on ‘MP-29’ coincided with an increase of heat accumulation in early March (Figure 4-1), resulting in greater GDD30 values by 49 to 111 units compared to those on other rootstocks.

Regardless of the rootstock, ‘UFSun’ fruits reached maturity after accumulating a range of 1451 to 1513 growing degree days (base temperature of 50°F) (Table 4-3).

105

Productive Performance

Wide range of values were recorded for fruit number and total yield between individual trees of ‘UFSun’ regardless of the rootstock used (Supplementary Table C-4) possibly due genetic heterogeneity of seedling rootstocks and to size- or age-related variation. Nevertheless, the number of observations were sufficient and differences between overall means of the yield parameters between rootstocks were large enough to be deemed statistically significant. Fruit yields increased as the trees matured and, within each year there were yield differences associated with the rootstocks. The differences in yield appeared to be due more to differences in fruit size than on fruit number induced by the rootstocks. During the three years, there were no significant differences in fruit number between ‘Flordaguard’ and the three test rootstocks (‘P-22’,

‘MP-29’, and ‘Barton’) after adjusting for scion trunk cross-sectional area (fruit load)

(Table 4-4). Greater proportion of smaller fruits (≤ 2 1/4") was recorded in the third year

(2016) for ‘MP-29’ and ‘Barton’ at peak harvest dates (Figure 4-2), resulting in lower mean fruit weights compared to those on ‘Flordaguard’ and ‘P-22’ (Table 4-4). In all three years, trees on ‘MP-29’ consistently produced the lowest mean fruit weights whereas their total yields and fruit numbers did not deviate significantly from those on

‘Flordaguard’ (Table 4-4). Conversely, for trees on ‘Barton’, yield, yield efficiency, and mean fruit weight were lower than those on ‘Flordaguard’ during the third year. In general, trees on ‘P-22’ maintained similar productive performance as ‘Okinawa’ and

‘Flordaguard’ throughout the first three years (Table 4-4).

Fruit Quality

Maturity of ‘UFSun’ fruits was determined by skin color change (from yellow- green to red over yellow ground color) and through hand firmness (equivalent to 2 to 3

106

lbf when measured by a fruit penetrometer). Except for fruit samples from trees on ‘P-

22’ that were slightly softer (2.15 lbf) than those on ‘Flordaguard’ and ‘MP-29’ (2.32 to

2.35 lbf) in the second year, there were no significant differences in flesh firmness among fruits harvested from trees on various rootstocks each year (Table 4-5). Fruit samples from the first year (2014) generally had smaller mean diameter and smaller fruit weight but higher Brix values (13.84 to 14.49) because the trees had set few fruits in the first year. In the second year (2015), sampled fruits from trees on ‘MP-29’ had significantly lower weight and correspondingly smaller fruit diameter (r = 0.936), but with a higher average Brix value compared to those on other rootstocks. ‘Barton’ differed significantly from the other rootstocks for the same fruit quality attributes in the third year (2016). Inconsistent sample size (1-24 fruits per tree) in the first year due to the large variability in the amount of fruit harvest and random selection of five fruits per tree in third year failed to detect significant differences in mean fruit weights induced between ‘MP-29’ and ‘Flordaguard’ (Table 4-4). However, there is a general trend indicating a negative association between fruit size and Brix values. In particular, the lower mean fruit weight and mean diameter recorded for ‘MP-29’ in the second year and

‘Barton’ in the third year were typically associated with higher Brix values (12.83 and

12.47, respectively) compared to those recorded for other rootstocks (≈ 11 degrees

Brix).

Discussion

In our field trial at Citra, Florida, ‘UFSun’ trees on ‘Flordaguard’ rootstock produced fruits that reached maturity during late April or about 80 days from bloom similar to the FDP reported for test sites in Immokalee (south Florida) as well as in

Gainesville (north central Florida) (Rouse and Sherman, 2004). With ‘P-22’ as

107

rootstock, fruit maturation occurred at about the same time as those with low-chill- adapted reference cultivars such as ‘Flordaguard’ and ‘Okinawa’, suggesting the potential of ‘P-22’ as a rootstock for low-chill peach production.

The differences among rootstocks in terms of individual fruit weight were inconsistent with the differences observed in terms of mean fruit weights based on whole-tree harvests; thus, comparisons about rootstock performance based on information from a limited number of randomly selected fruit samples must be made cautiously. It has been demonstrated that large variability in fruit size exists within peach trees due to variability in local intra-canopy factors that affect fruit growth (e.g. carbon source availability, competition between vegetative and reproductive organs, vigor of the fruiting shoots, and fruit position on the fruiting shoot) (Basile et al., 2007).

Our data on individual fruit attributes merely suggest that the differences in Brix values between ‘Flordaguard’ and ‘MP-29’ or ‘Barton’ were due to differences in fruit size.

When fruit size and maturity (as indicated by flesh firmness) of samples were the same, no significant variations in Brix values among other rootstocks were evidenced. In the second and third year, ‘UFSun’ fruits on ‘P-22’ as well as those on ‘Flordaguard’ and

‘Okinawa’ attained Brix values close to 11 (from sampled fruits weighing approximately

107 g), similar to the reported fruit characteristics of ‘UFSun’ at Immokalee, Florida from four to six-year old trees on ‘Flordaguard’ rootstock (Rouse and Sherman, 2004).

The greater proportion of smaller fruits (≤ 2 1/4”) recorded for ‘MP-29’ in all three years, and for ‘Barton’ in the third year were independent of fruit load. ‘MP-29’, an interspecific peach x plum rootstock, exerted such influence on fruit size through its effect on delaying bloom time which possibly shortened the FDP. It has been previously

108

established that high temperatures during 30 days after bloom negatively affects fruit sizing potential (Lopez and DeJong, 2008) due to resource limitations when early fruit growth rates increase in response to warm springs (Lopez and Dejong, 2007). Other studies have also found that warmer temperatures during fruit development could affect peach fruit quality, particularly by increasing the soluble solids content, incidence of protruding tips and suture deformation, and skin color intensity (Wert et al., 2007; Wert et al., 2009), although fruit shape and skin color were not documented in this study.

Consistent with the reported negative correlation between heat accumulation during 30 days after bloom and number of days for fruit growth (Ben Mimoun and DeJong, 1999;

Kenealy et al., 2015; Topp and Sherman, 1989), the higher GDD30 values shortened the

FDP for trees on ‘MP-29’ to four days relative to trees on ‘Flordaguard’. Full bloom for trees on ‘MP-29’ occurred during early March coinciding with a greater increase in the number of GDD (Figure 4-1), thereby resulting in lower mean fruit weights. Peach x plum rootstocks have been associated with smaller mean fruit weights and earlier fruit maturity of ‘Redtop’, ‘Cresthaven’, and ‘Redhaven’ peach at certain trial sites in Canada and North America, although the relationship between fruit weight and FDP were not examined in particular (Reighard et al., 2011).

As trees matured and developed fuller canopies, the increased vegetative growth required more extensive summer pruning (June 2015) to prevent excessive shading within the canopy and to limit tree size within the allocated in-row spacing of 4 ft.

Although the pruning weights were similar for all trees in the third year except for those on ‘MP-29’ which maintained a smaller canopy, prunings for trees on ‘Flordaguard’ and

‘P-22’ consisted mostly of branches with smaller diameter than those of ‘Okinawa’ and

109

‘Barton’ (data not shown). Pruning of vigorous branches during summer for trees on

‘Barton’ and ‘Okinawa’ seedling rootstocks appeared to have produced vigorous vegetative regrowth and resulted in smaller mean fruit weights compared to those on

‘Flordaguard’ in the third year. Corresponding observations of increased vegetative growth and reduced yield due to extensive pruning for cordon-trained trees have been attributed to a higher rate of partitioning to leaves and stems from May to August

(summer) than from bloom to May (spring) (Grossman and DeJong, 1998) and resource limitations during spring would lead to reduction in fruit growth potential unless trees are thinned at bloom (Grossman and DeJong, 1995). With fruit loads being the same for trees on different rootstocks, such resource limitations during fruit growth from severely pruned trees on ‘Barton’ and Okinawa’ rootstocks could explain the lower mean fruit weights relative to those on ‘Flordaguard’. The preponderance of smaller fruits (≤ 2 1/4" diameter) for trees on ‘Barton’ (65%) and those on ‘Okinawa’ (51%), accounted for the lower mean fruit weights in the third year, although less markedly different for trees on

‘Okinawa’. Despite having attained similar scion trunk size as ‘Flordaguard’, the lower yields induced by ‘Barton’ in the third year reduced the yield efficiency which, in the previous years, did not differ significantly from ‘Flordaguard’. ‘Barton’ rootstock tended to induce vigorous vegetative growth as also indicated by the enlargement of the scion trunk to a TCSA comparable to those achieved in the third year by older trees on

‘Flordaguard’ and ‘P-22’ while ‘Okinawa’ maintained the largest scion TCSA. A higher level of suckering was also observed (data not shown) on ‘Barton’ rootstock. These observations indicate that the two vigorous rootstocks ‘Barton’ and ‘Okinawa’ may not be amenable to close tree plantings and we may have underestimated their actual yield

110

potential in the third year due to severe restraints imposed on their growth. In contrast, trees on ‘P-22’ have shown higher productivity in terms of mean fruit weight as those on

‘Flordaguard’ under a high planting density.

During the first three years of field establishment at Citra Florida, ‘P-22’ has shown significant promise as a rootstock for peach production in areas where

‘Flordaguard’ could be utilized or in areas where ‘UFSun’ and similar low-chill-adapted scion cultivars have been successfully produced. ‘P-22' showed moderate vigor and consistently conferred better productivity and fruit quality like its ‘Flordaguard’ parent.

Additionally, it permitted harvests as early as the standard low-chill rootstocks from late

April to mid-May with FDP of approximately 78 days, which would allow Florida peach growers to meet the narrow window for marketing early, premium-priced peaches.

‘MP-29’ produced trees with a yield efficiency and fruit load comparable to the standard low-chill rootstocks in the second and third years and with exceptionally low vigor in terms of pruning weight as trees matured – a characteristic useful for high-density planting systems. Good size control is a desirable trait which reduces labor costs for pruning. ‘MP-29’ rootstock was associated with yields comparable with ‘Flordaguard’ but, due to late blooms and as a consequence of shortened fruit developmental period, it tended to induce greater numbers of smaller fruits which may disqualify this rootstock for commercial, low-chill peach production. In the third year, mean fruit weights induced by ‘MP-29’ generally ranged from 62.6 to 102.6 g with 70% of whole-tree harvests in the

≤ 2 1/4" size category associated with higher Brix values. This may be acceptable for small-scale fresh peach producers in eastern US with different selling targets that regard sweetness as more important than size (Zhao et al., 2017). Trees on ‘MP-29’

111

had a propensity for a more uniform bloom relative to ‘Flordaguard’, suggesting that specific cultural practices such as bloom-thinning may be required to alleviate resource limitations due to the abundance of reproductive sinks, and thereby increase resource availability during early fruit growth (Grossman and DeJong, 1995) leading to fruit size increase. Because of the concentrated flowering and fruit maturity observed (73-80% of fruits picked at mid-harvest), further investigations should determine its sizing potential at different fruit loads. ‘Barton’ rootstock tends to produce a lot of root suckers and excess vegetative vigor requiring extensive pruning, which would make it less economically efficient. Between the two new rootstocks, ‘P-22’ appeared to be more adapted to Florida’s subtropical conditions or production system and warrants further evaluations to determine if similar trends in productivity and fruit quality can be sustained as trees enter into full production.

112

Table 4-1. Description of rootstocks used in the study. Chilling Growth habit Nematode Rootstock requirement Origin Reference and vigor resistance (cu)

‘Flordaguard’ 150 – 300 a Multiple parentage Long limbs with Combined Sherman (P. persica and P. whippy growth resistances from et al. davidiana), its ‘Okinawa’ (1991) released by UF in ancestor and 1991 resistance to M. floridensis

‘Okinawa’ 150 a Introduced as seed Upright, good Resistant to M. Sharpe et by the Florida vigor incognita, M. al. (1969); Agricultural arenaria, and Sharpe Experiment Station some isolates of (1957) from Ryukyu, M. javanica; Japan in 1953 susceptible to M. floridensis

‘Barton’ 400 a Coastal rootstock Upright, good Showed good J. X. derived from a vigor resistance in the Chaparro feral peach field (pers. seedling in comm.) northern NSW Australia

‘P-22’ 500 b Hybrid (‘Guardian’ Upright, good Performed T. G x ‘Flordaguard’) vigor; similarly to Beckman peach selection horticulturally ‘Flordaguard’ on (pers. from USDA-ARS similar to field soils in comm.) (Byron, Georgia) ‘Guardian’ and Georgia infested rootstock breeding ‘Lovell’ seedling with M. incognita- program rootstocks and M. floridensis

‘MP-29’ 750 c Hybrid between a Semi-spreading, Resistant to Beckman red-leafed peach moderate vigor Armillaria root rot et al. rootstock selection and peach tree (2012) ‘SL0014’ and short life; ‘Edible Sloe’ plum; Appeared to be released for resistant to M. grower trial in 2012 incognita and M. by USDA-ARS floridensis in field (Byron, Georgia) trials in Georgia and Florida Agricultural Experiment Station a Based on bloom time with standard cultivars at Gainesville, FL. b Based on bloom time with standard cultivars at Byron, GA. c Vegetative budbreak coinciding with peach cultivars requiring ≈750 hours of chilling below 7 °C (45 °F) at Byron, GA.

113

Table 4-2. Growth of ‘UFSun’ peach trees budded on five different rootstocks during the first three years of field establishment (2014-2016) at Citra, FL. Rootstock Growth parameters a genotype b Rootstock trunk RGR c Scion trunk RGR c Pruning weight (g) d

Year 1 Flordaguard 1.75 ± 0.26 b 1.95 ± 0.43 b 245.0 ± 147.9 b Okinawa 1.21 ± 0.25 b 1.18 ± 0.40 b 963.3 ± 142.1 a P-22 1.35 ± 0.24 b 1.38 ± 0.39 b 646.4 ± 140.2 ab MP-29 2.78 ± 0.26 a 4.44 ± 0.44 a 268.1 ± 149.9 b Barton 2.99 ± 0.24 a 4.14 ± 0.39 a 421.7 ± 140.8 ab

Year 2 Flordaguard 8.07 ± 0.74 a 8.69 ± 1.19 b 1121.0 ± 131.1 ab Okinawa 2.94 ± 0.64 b 3.12 ± 1.03 c 1512.1 ± 129.8 a P-22 3.65 ± 0.62 b 3.56 ± 0.99 c 1316.5 ± 129.3 ab MP-29 10.33 ± 0.75 a 17.82 ± 1.21 a 912.8 ± 131.4 b Barton 8.60 ± 0.62 a 13.47 ± 1.00 a 1405.9 ± 129.5 ab

Year 3 Flordaguard 25.56 ± 1.68 a 26.39 ± 3.07 b 10442.0 ± 537.2 a Okinawa 8.31 ± 1.44 b 9.47 ± 2.63 c 11254.0 ± 471.9 a P-22 10.06 ± 1.39 b 10.70 ± 2.53 c 10705.0 ± 479.6 a MP-29 21.94 ± 1.70 a 42.69 ± 3.11 a 5618.4 ± 526.6 b Barton 24.55 ± 1.40 a 37.17 ± 2.55 ab 9951.5 ± 496.6 a a Values are least-squares means of ten replicates ± standard error. Different letters following means within a column indicate significant differences at P ≤ 0.05 according to a generalized linear mixed model analysis. b Trees on five different rootstocks were planted at 4 ft within-row and 20 ft between-row spacing during the first and second year. The spacing was widened to 8-16 ft in the third year by removal of alternate trees as well as diseased trees. c RGR, relative growth rate or change in trunk cross-sectional area (TCSA) relative to initial trunk size at pre-transplanting (TCi), expressed as: (TCf – TCi)/ TCi, where TCf is the calculated TCSA from trunk circumference measurements during the first (2014), second (2015), or third year (2016) of establishment. d Data for second and third years are from combined summer and winter pruning weights. Year 1: prunings from January 2014 before transplanting to field soil; Year 2: prunings from June 2014 and January 2015; Year 3: prunings from June 2015 and January 2016).

114

Table 4-3. Phenological characteristics of ‘UFSun’ peach trees as influenced by the rootstock genotype during the third year of field establishment (2016) at Citra, FL. GDD from full Days from full Mid- Rootstock Bloom date a, d GDD c, d bloom to mid- bloom to mid- harvest b 30 harvest d harvest d

Flordaguard 26 Feb ± 1.2 bc 16 May 459 ± 13.4 b 1513 ± 19.1 a 80 ± 0.9 a Okinawa 24 Feb ± 1.2 c 13 May 416 ± 13.9 c 1456 ± 20.0 a 79 ± 0.9 ab P-22 25 Feb ± 1.2 c 13 May 441 ± 13.4 bc 1451 ± 19.1 a 78 ± 0.9 ab MP-29 02 Mar ± 1.2 a 17 May 527 ± 14.5 a 1494 ± 21.0 a 76 ± 1.0 b Barton 28 Feb ± 1.2 ab 14 May 478 ± 13.4 b 1460 ± 19.1 a 77 ± 0.9 ab a Dates were recorded when approximately 50% of flowers reached full bloom. b When the amount of harvest reached 50% of total yield (in grams) for the season, which also corresponds to peak yield. Mid-harvest dates are averages from 21-25 trees per rootstock. c Growing degree days after 30 days from full bloom (GDD30) at Citra, FL as calculated in AgroClimate () using 10 °C (50 °F) as base temperature. d Values are least-squares means of ten replicates ± standard error. Different letters following means within a column indicate significant differences at P ≤ 0.05 according to a generalized linear mixed model analysis.

115

Table 4-4. Productive performance of ‘UFSun’ peach trees on five different rootstocks during the first three years of field establishment (2014-2016) at Citra, FL. Performance parameters b Rootstock a Fruit load (fruit Yield efficiency genotype Scion TCSA Mean fruit no. / tree Yield (g) / tree Mean fruit weight (g) no. / cm2 TCSA) (kg/cm2 TCSA)

Year 1 Flordaguard 5.40 ± 0.69 b 1.65 ± 1.01 bc 0.66 ± 0.15 ab 111.95 ± 154.39 b 0.04 ± 0.01 a 34.74 ± 3.83 b Okinawa 12.13 ± 0.61 a 7.98 ± 1.01 a 0.73 ± 0.13 a 498.23 ± 151.70 a 0.05 ± 0.01 a 55.23 ± 3.58 a P-22 10.57 ± 0.58 a 8.02 ± 1.01 a 0.78 ± 0.13 a 451.16 ± 150.41 a 0.04 ± 0.01 a 52.81 ± 3.41 a MP-29 4.55 ± 0.69 b 0.69 ± 1.01 c 0.08 ± 0.16 b 51.75 ± 153.18 b 0.01 ± 0.01 b 16.79 ± 3.66 c Barton 6.77 ± 0.59 b 5.03 ± 1.01 ab 0.72 ± 0.13 a 280.63 ± 150.93 ab 0.04 ± 0.01 a 51.61 ± 3.46 a Year 2 Flordaguard 15.73 ± 1.02 bc 23.12 ± 4.72 b 1.54 ± 0.27 a 2042.13 ± 391.97 b 0.14 ± 0.02 a 88.11 ± 2.18 a Okinawa 22.62 ± 0.96 a 45.37 ± 4.48 a 2.20 ± 0.25 a 3992.01 ± 373.86 a 0.19 ± 0.02 a 89.18 ± 2.08 a P-22 19.37 ± 0.92 ab 36.11 ± 4.45 ab 1.96 ± 0.24 a 3018.80 ± 371.13 ab 0.16 ± 0.02 a 84.92 ± 2.07 a MP-29 14.05 ± 0.97 c 28.96 ± 4.72 ab 2.03 ± 0.25 a 1983.39 ± 392.86 b 0.14 ± 0.02 a 70.55 ± 2.20 b Barton 17.03 ± 0.93 bc 32.12 ± 4.48 ab 1.92 ± 0.24 a 2619.60 ± 373.89 b 0.16 ± 0.02 a 83.50 ± 2.08 a Year 3 Flordaguard 44.57 ± 1.71 b 86.39 ± 9.71 a 1.77 ± 0.25 a 8781.22 ± 887.78 a 0.18 ± 0.02 a 101.89 ± 2.26 a Okinawa 55.44 ± 1.60 a 78.73 ± 8.78 a 1.46 ± 0.23 a 7017.11 ± 803.74 ab 0.13 ± 0.02 ab 90.01 ± 2.09 bc P-22 47.91 ± 1.53 b 63.38 ± 9.12 a 1.25 ± 0.23 a 5901.07 ± 834.23 ab 0.12 ± 0.02 ab 94.71 ± 2.15 ab MP-29 33.34 ± 1.62 c 70.47 ± 9.94 a 1.94 ± 0.25 a 5616.58 ± 908.75 ab 0.16 ± 0.02 ab 81.75 ± 2.32 cd Barton 43.60 ± 1.54 b 52.64 ± 9.12 a 1.20 ± 0.23 a 4264.04 ± 834.24 b 0.10 ± 0.02 b 81.37 ± 2.15 d a Trees on five different rootstocks were planted at 4 ft within-row and 20 ft between-row spacing during the first and second year. The spacing was widened to 8-16 ft in the third year by removal of alternate trees as well as diseased trees to encourage growth of bigger trees. b Values are least-squares means of ten replicates ± standard error. Different letters following means within a column indicate significant differences at P ≤ 0.05 according to a generalized linear mixed model analysis. TCSA, trunk cross- sectional area.

116

Table 4-5. Fruit quality of ‘UFSun’ peach on five different rootstocks during the first three years of field establishment (2014-2016) at Citra, FL. Rootstock Individual fruit attributes a genotype Mean diameter (mm) b Fruit weight (g) Flesh firmness (lbf) c TSS (°Brix)

Year 1 Flordaguard 49.33 ± 1.07 a 62.18 ± 3.44 a 3.18 ± 0.21 a 13.84 ± 0.30 a Okinawa 48.66 ± 0.54 a 60.51 ± 1.78 a 2.82 ± 0.10 a 14.13 ± 0.32 a P-22 47.50 ± 0.54 a 56.94 ± 1.76 a 2.97 ± 0.10 a 14.49 ± 0.27 a MP-29 48.15 ± 1.58 a 58.00 ± 5.05 a 3.42 ± 0.34 a 13.96 ± 0.36 a Barton 49.19 ± 0.67 a 58.37 ± 2.18 a 2.90 ± 0.13 a 14.14 ± 0.30 a

Year 2 Flordaguard 56.22 ± 0.44 ab 94.09 ± 1.84 ab 2.32 ± 0.05 a 11.41 ± 0.21 b Okinawa 57.19 ± 0.36 a 97.47 ± 1.49 a 2.27 ± 0.05 ab 10.94 ± 0.19 b P-22 55.89 ± 0.41 ab 91.49 ± 1.71 ab 2.15 ± 0.05 b 11.00 ± 0.21 b MP-29 51.05 ± 0.46 c 75.39 ± 1.90 c 2.35 ± 0.06 a 12.83 ± 0.21 a Barton 55.31 ± 0.40 b 89.24 ± 1.67 b 2.25 ± 0.05 ab 11.12 ± 0.21 b

Year 3 Flordaguard 60.70 ± 0.64 a 113.34 ± 3.16 a 1.87 ± 0.11 a 11.07 ± 0.33 b Okinawa 60.39 ± 0.62 ab 113.55 ± 3.08 a 2.00 ± 0.11 a 11.57 ± 0.33 ab P-22 59.45 ± 0.61 ab 106.92 ± 3.07 a 2.19 ± 0.10 a 11.35 ± 0.32 ab MP-29 57.73 ± 0.68 b 100.31 ± 3.36 a 1.96 ± 0.12 a 11.52 ± 0.34 ab Barton 54.60 ± 0.64 c 86.22 ± 3.14 b 2.03 ± 0.11 a 12.47 ± 0.31 a a Values are least-squares means ± standard error, with each replicate consisting of at least five representative fruits. Different letters following means within a column indicate significant differences at P ≤ 0.05 according to a generalized linear mixed model analysis. TCSA, trunk cross-sectional area. TSS, total soluble solids content. b Average of two measured dimensions in cheek direction and suture direction. c Expressed in pounds-force (lbf) as measured by a handheld fruit penetrometer with 8 mm-tip plunger.

117

2015 2016 Accumulated chill-hours (2014-2015) Accumulated chill-hours (2015-2016) 600 497 499 510 445 500

371 400

284 276 262 300 228 244 196 212 159 200 118 77 35 100 18 5 10 0 0 1-15 16-31 1-15 16-31 1-15 16-31 1-15 16-28 1-15 16-31 November December January February March

2015 2016 Accumulated GDD (2015) Accumulated GDD (2016) 3000

2386 2500 1957 2253 2000 1626 1843 1274 1500 1508 920 1189 1000 644 905 358 267 603 500 132 203 377 270 0 1-15 16-31 1-15 16-28 1-15 16-31 1-15 16-30 1-15 16-31 January February March April May

Figure 4-1. Accumulation of chilling hours between 32 to 45 °F (air temperatures at 2 meters) and growing degree days (GDD, base temperature of 50 °F) during 2015 and 2016 at Citra, Florida. Data were derived from the Florida Automated Weather Network weather archives for Citra station. Chilling hours were calculated with no negation accounted for temperatures outside the range of 32 to 45 °F.

118

2015 Harvest 2016 Harvest 23 April 30 April 05 May 13 May 20 May 26 April 03 May 11 May 17 May 24 May 100 100 90 90 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 F O P M B F O P M B

100 100 90 90 ≤ 2 1/4" 80 80 2 5/16" to 2 7/16" 70 70 ≥ 2 1/2" 60 60 50 50 40 40 30 30 20 20 10 10 0 0 F O P M B F O P M B F O P M B F O P M B F O P M B F O P M B F O P M B F O P M B F O P M B F O P M B 23 April 30 April 05 May 13 May 20 May 26 April 03 May 11 May 17 May 24 May

Figure 4-2. Estimated proportion (%) of ‘UFSun’ fruits harvested at each date relative to total fruits per rootstock throughout the harvest season during second (2015) and third (2016) year of establishment at Citra, FL. Top panels show the cumulative percentage of fruits per rootstock by harvest date and bottom panels show the proportions at each harvest date relative to season totals per rootstock. Harvested fruits from 2016 were sorted by size (≤ 2 1/4", 2 5/16" to 2 7/16", and ≥ 2 1/2") using fruit sizing rings. Rootstocks include: F, ‘Flordaguard’; O, ‘Okinawa’; P, ‘P-22’; M, ‘MP-29’; and B, ‘Barton’.

119

CHAPTER 5 CONCLUDING REMARKS

An endemic Florida nematode, Meloidogyne floridensis (also known as the peach root-knot nematode) infects the majority of available nematode-resistant peach rootstocks, restricting the availability of peach trees and thus, industry expansion within the state. To make peach growing economically viable, growers must have a wide range of rootstock choices. The results of this project have practical implications for the ongoing breeding efforts to increase the panel of rootstocks available to low-chill peach growers. This project was conducted with a two-pronged approach to provide short- and long-term solutions for the industry. To fulfill our short-term goal of updating rootstock recommendations to Florida peach growers, the commercial potential of new rootstock releases for subtropical, low-chill peach production were evaluated in terms of vigor, phenology, yield, and fruit quality during the initial years of orchard establishment.

This project component involved two years of tree preparation and three years of evaluating horticultural performance in the field. During the three years of field establishment, ‘P-22’ showed to be a viable alternative rootstock for Florida growers; besides having RKN resistance and good agronomical features, it has tolerance to

Armillaria root rot and peach tree short life.

As part of the long-term goal of developing improved rootstocks to advance the peach industry, we characterized the genetic mechanism of resistance to the peach root-not nematode in multiple interspecific Prunus populations by locating the gene, confirming the number of loci involved and the mode of inheritance. The results of the study provide a useful basis for designing efficient breeding strategies for the introgression new sources of nematode resistance into locally-adapted rootstock

120

genotypes. We have identified a true resistant source in P. kansuensis that we could readily exploit to broaden the nematode resistances in the rootstock germplasm. We have also established a resistance screening procedure to improve the phenotyping accuracy, which is important not only for effectively selecting resistant breeding material, but also for marker-trait association studies and the eventual deployment of a marker-assisted strategy to efficiently breed for nematode resistance. New rootstocks can be developed with improved commercial utility by combining high nematode resistance from P. kansuensis and other useful agronomic features of ‘Okinawa’ or

‘Flordaguard’ for peach production areas having the same climate and soil environment as Florida.

The genetic analyses of twelve interspecific peach x P. kansuensis progenies using microsatellite markers allowed us to infer the chromosomal location of the Mf locus, which confers resistance to M. floridensis. The Mf locus was found to be in the same region on linkage group 2 where other Prunus RKN resistance genes have been reported, e.g. RMia from peach (‘Shalil’ and ‘Nemared’), and PkMi from wild peach conferring resistance against M. incognita and M. arenaria. The findings about the mode of inheritance and the resistance locus would direct our breeding strategies towards developing new nematode-resistant rootstocks by effectively selecting resistant parents from diverse germplasm and keeping track of the resistance in the progeny. It must be noted, however, that the use of a single MF population to evaluate resistance among the interspecific progenies does not allow for definitive conclusions to be made about whether or not the same alleles confer resistance against other MF populations.

There are possibly multiple alleles (or haplotypes) in the highly polymorphic resistance

121

locus from the wild-related PK and different interspecific peach x PK hybrids may carry different resistance alleles at the MF locus. In this context, it is important to understand the influence of different allele combinations on the level of MF resistance when considering interspecific hybridization and introgression of RKN resistances from PK into peach rootstocks. In Myrobalan plum (P. cerasifera), resistance variability to major

RKN was observed among the different accessions and diallel crosses of five accessions revealed different modes of inheritance for resistance when challenged with

M. arenaria (Esmenjaud et al., 1996b; Esmenjaud et al., 1994). Screening additional markers within the subtelomeric region of LG 2 should identify more closely linked markers that can distinguish the various polymorphisms associated with resistance to

MF in PK and therefore, facilitate the efficient introgression of the trait into existing peach rootstocks, particularly ‘Okinawa’ and Flordaguard’.

122

APPENDIX A SUPPLEMENTARY DATA FOR SEGREGATION ANALYSES

Table A-1. Segregation analyses of peach Prunus persica x Prunus kansuensis F2 and testcross populations for resistance based on galling index. Segregating Individuals Galling index c Observed ratio d Expected ratio d Goodness-of-fit e population a analyzed (N) b 0 1 2 3 4 5 R (≤ 2) S (> 2) R S Χ2 P

F2: [OK x PK]1 36 25 0 0 0 4 7 25 11 27 9 0.593 0.441 [OK x PK]2 34 4 0 0 3 10 17 4 30 25.5 8.5 72.51 <.001 [OK x PK]3 35 1 1 5 1 3 24 7 28 26.25 8.75 56.47 <.001 [OK x PK]4 29 14 3 2 1 1 8 19 10 21.75 7.25 1.391 0.238 [OK x PK]5 22 18 0 0 0 0 4 18 4 16.5 5.5 0.546 0.460 [OK x PK]6 28 22 0 0 0 1 5 22 6 21 7 0.191 0.662 [FG x PK]1 29 22 0 0 0 0 7 22 7 21.75 7.25 0.012 0.915 [FG x PK]6 25 19 0 1 0 2 3 20 5 18.75 6.25 0.333 0.564

Testcross: SH x [OK x PK]2 33 14 1 0 1 3 14 15 18 16.5 16.5 0.273 0.602 SH x [OK x PK]3 30 3 0 0 0 2 25 3 27 15 15 19.20 <.001 FG x [OK x PK]3 20 0 0 0 2 1 17 0 20 10 10 20 <.001 SH x [FG x PK]1 18 8 0 0 1 0 9 8 10 9 9 0.222 0.637 SH x [FG x PK]6 40 23 0 0 0 1 16 23 17 20 20 0.900 0.343 a PK = P. kansuensis wild peach; OK = ‘Okinawa’ peach; FG = ‘Flordaguard’ peach; SH = ‘UFSharp’ peach. b True selfs or testcrosses (species-level backcrosses) verified through microsatellite marker profiling. Individual plants (N) were evaluated 120 days after inoculation with Meloidogyne floridensis (MFGnv14 isolate) at a population density of 10,000 eggs per plant. c 0 = no galls, 1 = one to two galls, 2 = three to 10, 3 = 11 to 30, 4 = 31 to 100, and 5 = more than 100, where values greater than 2 indicate host susceptibility (Taylor and Sasser, 1978). d R = resistant; S = susceptible; expected ratios are 3R:1S for F2 populations and 1R:1S for testcross populations based on the assumption of a single major gene. e Chi square (Χ2) and probability (P) values for testing goodness-of-fit to expected ratios of R and S at 95% confidence 2 2 level. Observed data conform to expected segregation ratio if Χ values are less than the critical value: Χ 0.05, 2 = 5.99 for 2 F2 or Χ 0.05, 1 = 3.84 for testcross.

123

Table A-2. Segregation analyses of peach Prunus persica x Prunus kansuensis F2 and testcross populations for resistance based on egg mass index. Segregating Individuals Egg mass index c Observed ratio d Expected ratio d Goodness-of-fit e population a analyzed b 0 1 2 3 4 5 R (≤ 2) S (> 2) R S Χ2 P

F2: [OK x PK]1 36 25 0 0 1 3 7 25 11 27 9 0.593 0.441 [OK x PK]2 34 4 0 2 9 11 8 6 28 25.5 8.5 59.65 <.001 [OK x PK]3 35 5 1 1 3 11 14 7 28 26.25 8.75 56.47 <.001 [OK x PK]4 29 17 2 1 1 1 7 20 9 21.75 7.25 0.563 0.453 [OK x PK]5 22 18 0 0 0 0 4 18 4 16.5 5.5 0.545 0.460 [OK x PK]6 28 22 0 0 1 0 5 22 6 21 7 0.191 0.662 [FG x PK]1 29 22 0 0 0 0 7 22 7 21.75 7.25 0.012 0.915 [FG x PK]6 25 19 0 2 2 1 1 21 4 18.75 6.25 1.080 0.299

Testcross: SH x [OK x PK]2 33 15 0 2 2 0 14 17 16 16.5 16.5 0.030 0.862 SH x [OK x PK]3 30 3 0 0 1 4 22 3 27 15 15 19.20 <.001 FG x [OK x PK]3 20 0 0 0 0 2 18 0 20 10 10 20 <.001 SH x [FG x PK]1 18 8 0 1 0 0 9 9 9 9 9 0 1 SH x [FG x PK]6 40 23 0 0 0 1 16 23 17 20 20 0.900 0.343 a PK = P. kansuensis wild peach; OK = ‘Okinawa’ peach; FG = ‘Flordaguard’ peach; SH = ‘UFSharp’ peach. b True selfs or testcrosses (species-level backcrosses) verified through microsatellite marker profiling. Individual plants (N) were evaluated 120 days after inoculation with Meloidogyne floridensis (MFGnv14 isolate) at a population density of 10,000 eggs per plant. c 0 = no egg masses, 1 = one to two egg masses, 2 = three to 10, 3 = 11 to 30, 4 = 31 to 100, and 5 = more than 100, where values greater than 2 indicate host susceptibility (Taylor and Sasser, 1978). d R = resistant; S = susceptible; expected ratios are 3R:1S for F2 populations and 1R:1S for testcross populations based on the assumption of a single major gene. e Chi square (Χ2) and probability (P) values for testing goodness-of-fit to expected ratios of R and S at 95% confidence 2 2 level. Observed data conform to expected segregation ratio if Χ values are less than the critical value: Χ 0.05, 2 = 5.99 for 2 F2 or Χ 0.05, 1 = 3.84 for testcross.

124

Table A-3. Segregation of peach Prunus persica x Prunus kansuensis F2 and testcross populations for resistance based on reproduction factor. Segregating Individuals Reproduction factor c Observed ratio d Expected ratio d Goodness-of-fit e population a analyzed b RF < 0.1 0.1< RF <1 RF > 1 R (< 1) S (> 1) R S Χ2 P “non-host” “poor host” “host”

F2:

[OK x PK]1 36 27 6 3 27 9 27 9 0 1 [OK x PK]2 34 5 23 6 5 29 25.5 8.5 65.92 <.001 [OK x PK]3 36 9 12 15 9 27 27 9 48 <.001 [OK x PK]4 28 21 4 3 21 7 21 7 0 1 [OK x PK]5 22 18 2 2 18 4 16.5 5.5 0.546 0.460 [OK x PK]6 28 22 5 1 22 6 21 7 0.190 0.663 [FG x PK]1 29 22 5 2 22 7 21.75 7.25 0.012 0.915 [FG x PK]6 24 20 3 1 20 4 18 6 0.890 0.346

Testcross: SH x [OK x PK]2 33 16 6 11 16 17 16.5 16.5 0.030 0.862 SH x [OK x PK]3 30 8 10 12 8 22 15 15 6.53 0.011 FG x [OK x PK]3 20 0 4 16 0 20 10 10 20 <.001 SH x [FG x PK]1 18 9 2 7 9 9 9 9 0 1 SH x [FG x PK]6 40 26 7 7 26 14 20 20 3.60 0.058 a OK = ‘Okinawa’ peach; PK = Prunus kansuensis wild peach; FG = ‘Flordaguard’ peach; SH = ‘UFSharp’ peach. b True selfs or testcrosses (species-level backcrosses) verified through microsatellite marker profiling. Individual plants (N) were evaluated 120 days after inoculation with Meloidogyne floridensis (MFGnv14 isolate) at a population density of 10,000 eggs per plant. c Reproduction factor (RF): ratio of the final egg count to initial inoculum at 10,000 eggs per plant. Plants were classified as resistant (R) if RF values fall below 0.1, which are considered “non-hosts”, as described in Sasser et al., 1984. d R = resistant; S = susceptible; expected ratios are 3R:1S for F2 populations and 1R:1S for testcross populations based on the assumption of a single major gene. e Chi square (Χ2) and probability (P) values for testing goodness-of-fit to the expected ratios of R and S at 95% confidence 2 2 level. Observed data conform to the expected segregation ratio if Χ values are less than the critical value: Χ 0.05, 2 = 5.99 2 for F2 or Χ 0.05, 1 = 3.84 for testcross.

125

APPENDIX B SUPPLEMENTARY DATA FOR LINKAGE ANALYSIS

Table B-1. Origins of the microsatellite markers tested for polymorphism and number of markers included in the linkage maps. Series Name Species Origin Tested Polymorphic a Selected b Reference BPPCT P. persica Genomic 18 15 11 Dirlewanger et al. (2002) CPDCT P. dulcis Genomic 13 8 4 Mnejja et al. (2005) CPPCT P. persica Genomic 19 15 12 Aranzana et al. (2002) CPSCT P. salicina Genomic 11 5 3 Mnejja et al. (2004) EPDCU P. dulcis cDNA 6 4 1 Mnejja et al. (2010) EPPIS P. persica EST 1 1 1 Vendramin et al. (2007) PMS P. avium Genomic 1 1 1 Cantini et al. (2001) Struss et al. (2002) PCHCMS, P. persica Genomic 3 1 0 Sosinski et al. (2000) PCHGMS UDP P. persica Genomic 9 9 6 Testolin et al. (2002); Cipriani et al. (1999) Total 81 58 39 a Markers that revealed clear polymorphisms between Prunus persica and Prunus kansuensis parental genotypes and segregating in F1 hybrids. b Informative markers showing consistent primer amplification patterns, are easily scored, and span an interval of approximately 10-20 cM throughout the genome with reference to the Prunus ‘Texas’ almond x ‘Earlygold’ peach linkage map.

126

Table B-2. Characteristics of the polymorphic microsatellite markers used to map the Mf resistance locus in peach x Prunus kansuensis interspecific progenies. Position a Marker Primer pair sequences (F: forward; R: reverse) Annealing LG cM Temp. (°C) LG 1 1.3 CPPCT016 F: aattccctatggaaattaga R: cgcatattataggtaggaaa 50 9.0 CPSCT008 F: tggatccaatccaagagtctg R: gcagcaagttgttcttggttc 62 23.1 CPPCT027 F: gagcagttcataagttggaacaa R: cgataaagattttgactgcatga 55-57 29.2 CPSCT027 F: cccatgctcctgtggtaagt R: tttagaatcccaaccccaca 62 33.9 CPPCT026 F: agacgcagcacccaaactac R: cattacatcaccgccaacaa 55 47.3 BPPCT027 F: ctctcaagcatcatgggc R: tgttgcccggttgtaatatc 57 65.1 CPPCT029 F: ccaaattccaaatctcctaaca R: tgatcaactttgagatttgttgaa 55 77.4 BPPCT028 F: tcaagttagctgaggatcgc R: gagcttgcctatgagaagacc 57 LG 2 9.6 UDP98-025 F: gggaggttactatgccatgaag R: cgcagacatgtagtaggacctc 57 25.0 BPPCT013 F: acccacaaatcaagcatatcc R: agcttcagccaccaagc 57-59 27.8 UDP96-013 F: attcttcactacacgtgcacg R: ccccagacatactgtggctt 57 38.0 BPPCT030 F: aattgtacttgccaatgctatga R: ctgccttctgctcacacc 57 48.6 CPSCT034 F: aggtggacaatagccgtgat R: tttccagaccctgagaaagc 62 LG 3 11.2 BPPCT007 F: tcattgctcgtcatcagc R: cagatttctgaagttagcggta 57-59 36.4 CPDCT025 F: gacctcatcagcatcaccaa R: ttccctaacgtccctgacac 62 46.4 CPDCT027 F: tgaggagagcactggaggag R: caaccgatccctctagacca 62 LG 4 10.4 CPPCT005 F: catgaactctactctcca R: tggtatggactcaccaac 57 28.3 UDP96-003 F: ttgctcaaaagtgtcgttgc R: acacgtagtgcaacactggc 62 62.7 EPPISF032 b F: tcccccacagatatttcagc R: gtcgaggagagagggctttt 57 78.2 CPDCT014 c F: tgcaaagaaaaacggagagg R: gaaactcagtggcacaatcg 57-62 LG 5 5.2 BPPCT026 F: atacctttgccacttgcg R: tgagttggaagaaaacgtaaca 57-62 20.1 BPPCT017 F: ttaagagtttgtgatgggaacc R: aagcataatttagcataaccaagc 57 32.9 BPPCT038 F: tatattgttggcttcttgcatg R: tgaaagtgaaacaatggaagc 57 94.4 CPDCT022 c F: tgatcggcgtctcctttatc R: aaagcaagcaggcaaatgaa 57-62 LG 6 30.1 BPPCT008 F: atggtgtgtatggacatgatga R: cctcaacctaagacaccttcact 57 56.4 BPPCT025 F: tcctgcgtagaagaaggtagc R: cgacataaagtccaaatggc 57 80.2 CPPCT030 F: tgaatattgttcctcaattc R: ctctaggcaagagatgaga 52-55 LG 7 14.1 CPPCT039 F: gcaccagttcttcgtcatctc R: gcatgcataaaacctttattgg 57 18.6 CPPCT022 F: caattagctagagagaattattg R: gacaagaagcaagtagtttg 50 22.3 UDP98-405 F: acgtgatgaactgacaccca R: gagtctttgctctgccatcc 57 23.7 UDP98-408 F: acaggcttgttgagcatgtg R: ccctcgtgggaaaatttga 57 38.9 CPPCT033 F: tcagcaaactagaaacaaacc R: ttgcaatctggttgatgtt 50 47.8 PMS2 F: cactgtctcccaggttaaact R: cctgagcttttgacacatgc 55 61.8 CPPCT017 F: tgacatgcatgcactaaacaa R: tgcaaatgcaatttcataaagg 57 64.7 EPDCU3392 F: cttttcatgggttcctcacc R: atcaaccagttcacgcacaa 57 LG 8 7.8 CPPCT019B F: aattcaatgtcaagacaca R: tcatcaaaataaatatccagt 50 14.1 BPPCT006 F: gcttgtggcatggaagc R: ccctgtttctcatagaactcacat 57-59 24.8 CPPCT006 F: aattaactccaacagctcca R: atggttgcttaattcaatgg 59 44.5 UDP98-409 F: gctgatgggttttatggttttc R: cggactcttatcctctatcaaca 59 a Marker positions on linkage groups (LG 1 - LG 8) according to Prunus ‘Texas’ almond x ‘Earlygold’ peach reference map (Dirlewanger et al., 2004b) b Mapped in ‘Contender’ peach x ‘Fla.92-2C’ peach F2 progeny (Fan et al., 2010) c Mapped in ‘Ferjalou Jalousia’ peach x ‘Fantasia’ peach F2 progeny (Dirlewanger et al., 2007).

127

Table B-3. Description of the mapping populations. No. of individuals (records Population available) for QTL analysis Mapping population a Size (N) Off-types b Genotype Phenotype F2: [OK x PK]1 43 7 36 36 [OK x PK]3 43 7 36 35 [OK x PK]4 43 14 29 29 [OK x PK]5 43 20 23 22 [OK x PK]6 43 15 28 28 [FG x PK]1 39 10 29 29 [FG x PK]6 43 18 25 25

BC1F1: SH x [OK x PK]2 42 9 33 33 SH x [OK x PK]3 43 13 30 30 FG x [OK x PK]3 23 3 20 20 SH x [FG x PK]1 30 12 18 18 SH x [FG x PK]6 43 3 40 40 Total 478 131 347 345 a PK = Prunus kansuensis wild peach; OK = ‘Okinawa’ peach; FG = ‘Flordaguard’ peach; SH = ‘UFSharp’ peach. b Individuals were considered as off-types when discrepant alleles were observed at 1- 21 microsatellite loci and, therefore excluded from mapping analysis.

128

Table B-4. Microsatellite allele configurations of ‘Okinawa’, Prunus kansuensis, and their hybrids inferred from segregation patterns of marker band and electropherogram data in F2 progenies. Linkage Parents * F2 population * Marker name group OK PK [OK x PK]1 [OK x PK]3 OK x PK]4 [OK x PK]5 [OK x PK]6 LG 1 CPPCT016 191 | 191 173 | 177 191 | 173 191 | 173 191 | 177 191 | 173 191 | 173 CPSCT008 182 | 182 234 | 234 182 | 234 182 | 234 182 | 234 182 | 234 182 | 234 CPPCT027 107 | 107 82 | 89 107 | 89 107 | 89 107 | 82 107 | 82 107 | 89 CPSCT027 137 | 137 167 | 169 137 | 169 137 | 169 137 | 169 137 | 167 137 | 169 CPPCT026 191 | 191 191 | 189 191 | 191 191 | 189 191 | 189 191 | 189 191 | 189 BPPCT027 244 | 244 238 | 254 244 | 254 244 | 238 244 | 238 244 | 238 244 | 238 CPPCT029 170 | 170 170 | 170 170 | Ø 170 | Ø 170 | Ø 170 | Ø 170 | Ø BPPCT028 163 | 163 158 | 158 163 | Ø 163 | Ø 163 | Ø 163 | Ø 163 | Ø LG 2 UDP98-025 111 | 111 115 | 115 111 | 115 111 | 115 111 | 115 111 | 115 111 | 115 CPDCT044 245 | 245 170 | 176 245 | 172 245 | 170 245 | 172 245 | 172 245 | 170 BPPCT013 178 | 178 208172 | 208192 178 | 208192 178 | 208192 178 | 208176 178 | 208176 178 | 208192 UDP96-013 195 | 195 203 | 233 195 | 233 195 | 203 195 | 233 195 | 233 195 | 233 BPPCT030 156 | 156 150 | 173 156 | 150 156 | 150 156 | 150 156 | 173 156 | 150 CPSCT034 237 | 237 245 | 245 237 | 245 237 | 245 237 | 237 237 | 237 237 | 245 LG 3 BPPCT007 140 | 140 109 | 109 140 | 109 140 | 109 140 | 109 140 | 109 140 | 109 CPDCT025 183 | 183 183 | 183 183 | 183 183 | 183 183 | 183 183 | 183 183 | 183 CPDCT027 162 | 162 146 | 152 162 | 146 162 | 146 162 | 152 162 | 152 162 | 152 LG 4 CPPCT005 171 | 171 157 | 132 171 | 132 171 | 157 171 | 157 171 | 132 171 | 132 UDP96-003 134 | 134 120 | 120 134 | 120 134 | 120 134 | 120 134 | 120 134 | 120 EPPISF032 198 | 198 198 | 186 198 | 198 198 | 186 198 | 186 198 | 198 198 | 198 CPDCT014 108 | 108 91 | 108 108 | 91 108 | 93 108 | 93 108 | 91 108 | 91 LG 5 93 BPPCT026 149 | 149 158 | 158 149 | 158 149 | 158 149 | 158 149 | 158 149 | 158 BPPCT017 148 | 148 162 | 162 148 | 162 148 | 162 148 | 162 148 | 162 148 | 162 BPPCT038 154 | 154 148 | 162 154 | 162 154 | 162 154 | 162 154 | 162 154 | 148 CPDCT022 146 | 146 157 | 157 146 | 157 146 | 157 146 | 157 146 | 157 146 | 157

129

Table B-4. Continued. Linkage Marker name Parents * F2 population * group OK PK [OK x PK]1 [OK x PK]3 OK x PK]4 [OK x PK]5 [OK x PK]6 LG 6 CPPCT008 157 | 157 157 | 157 157 | 157 157 | 157 157 | 157 157 | 157 157 | 157 BPPCT008 154 | 154 105 | 105 154 | 105 154 | 105 154 | 105 154 | 105 154 | 105 BPPCT025 187 | 187 181 | 183 187 | 181 187 | 181 187 | 183 187 | 181 187 | 181 CPPCT030 191 | 191 239 | 249 191 | 239 191 | 239 191 | 239 191 | 249 191 | 239 LG 7 CPPCT039 98 | 98 104 | 108 98 | 104 98 | 104 98 | 104 98 | 104 98 | 104 CPPCT022 260 | 260 290 | 292 260 | 292 260 | 292 260 | 292 260 | 292 260 | 292 UDP98-405 103 | 103 99 | 99 103 | 99 103 | 99 103 | 99 103 | 99 103 | 99 UDP98-408 100 | 100 89 | 89 100 | 89 100 | 89 100 | 89 100 | 89 100 | 89 CPPCT033 142 | 142 143 | 160 142 | 160 142 | 143 142 | 160 142 | 160 142 | 160 PMS2 126 | 126 140 | 150 126 | 150 126 | 140 126 | 150 126 | 150 126 | 150 CPPCT017 165 | 165 169 | 169 165 | 169 165 | 169 165 | 169 165 | 169 165 | 169 EPDCU3392 117 | 117 123 | 123 117 | 123 117 | 123 117 | 123 117 | 123 117 | 123 LG 8 CPPCT019B 173 | 175 173 | 171 173 | 175 173 | 175 173 | 175 173 | 175 173 | 175 BPPCT006 126 | 126 106 | 106 126 | 106 126 | 106 126 | 106 126 | 106 126 | 106 UDP96-019 213 | 213 214 | 214 213 | 214 213 | 214 213 | 214 213 | 214 213 | 214 CPPCT006 Ø | Ø 184 | 184 Ø | 184 Ø | 184 Ø | 184 Ø | 184 Ø | 184 UDP98-409 123 | 123 134 | 134 123 | 134 123 | 134 123 | 134 123 | 134 123 | 134 EPDCU3117 165 | 165 165 | 165 165 | 165 165 | 165 165 | 165 165 | 165 165 | 165 * PK = P. kansuensis wild peach; OK = ‘Okinawa’ peach; Ø, null allele.

130

Table B-5. Microsatellite allele configurations of ‘Flordaguard’, Prunus kansuensis, and their hybrids inferred from segregation patterns of marker band and electropherogram data in F2 progenies. Linkage Parents * F1 hybrid * Marker name group FG PK [FG x PK]1 [FG x PK]6 LG 1 CPPCT016 191 | 199 173 | 177 199 | 177 199 | 177 CPSCT008 175 | Ø 234 | 234 175 | 234 175 | 234 CPPCT027 105 | 105 82 | 89 105 | 89 105 | 89 CPSCT027 137 | 137 167 | 169 137 | 169 137 | 169 CPPCT026 191 | 174 191 | 189 191 | 189 174 | 189 BPPCT027 244 | 244 238 | 254 244 | 238 244 | 238 CPPCT029 170 | 170 170 | 170 170 | Ø 170 | Ø BPPCT028 163 | 163 158 | 158 163 | Ø 163 | Ø LG 2 UDP98-025 111 | 111 115 | 115 111 | 115 111 | 115 CPDCT044 215 223 | 247 170 172 | 176 192 215 223 | 172 223 247 | 172 192 BPPCT013 178 | 182 208 | 208 178247 | 208192 182 | 208 UDP96-013 195 | 195 203 | 233 195 | 203 195 | 233 BPPCT030 156 | 156 150 | 173 156 | 173 156 | 150 CPSCT034 237 | 237 245 | 245 237 | 237 237 | 245 LG 3 BPPCT007 140 | 140 109 | 109 140 | 109 140 | 109 CPDCT025 181 | 185 183 | 183 185 | 183 181 | 183 CPDCT027 162 | 162 146 | 152 162 | 146 162 | 146 LG 4 CPPCT005 157 | 148 157 | 132 148 | 132 148 | 132 UDP96-003 125 | 129 120 | 120 129 | 120 129 | 120 EPPISF032 198 | 200 198 | 186 198 | 198 198 | 186 CPDCT014 113 | Ø 91 93 | 108 Ø | 91 108 113 | 93 LG 5 BPPCT026 129 | 137 158 | 158 137 | 158 129 | 158 BPPCT017 148 | 158 162 | 162 148 | 162 148 | 162 BPPCT038 156 | 160 148 | 162 156 | 148 156 | 148 CPDCT022 146 | 146 157 | 157 146 | 157 146 | 157

131

Table B-5. Continued. Linkage Marker name Parents * F1 hybrid * group FG PK [FG x PK]1 [FG x PK]6 LG 6 CPPCT008 157 | 164 157 | 157 157 | 157 157 | 157 BPPCT008 133 | 133 105 | 105 133 | 105 133 | 105 BPPCT025 189 | 189 181 | 183 189 | 183 189 | 183 CPPCT030 193 | 193 239 | 249 193 | 249 193 | 239 LG 7 CPPCT039 98 | 91 104 | 108 91 | 104 91 | 108 CPPCT022 260 | 214 290 | 292 214 | 292 260 | 290 UDP98-405 103 | 99 99 | 99 99 | 99 103 | 99 UDP98-408 100 | 102 89 | 89 102 | 89 100 | 89 CPPCT033 142 | 142 143 | 160 142 | 160 142 | 160 PMS2 126 | 128 140 | 150 128 | 150 126 | 150 CPPCT017 165 | 175 169 | 169 175 | 169 165 | 169 EPDCU3392 117 | 120 123 | 123 120 | 123 117 | 123 LG 8 CPPCT019B 173 | 181 173 | 171 173 | 181 173 | 181 BPPCT006 114 | 114 106 | 106 114 | 106 114 | 106 UDP96-019 215 | 215 214 | 214 215 | 215 215 | 215 CPPCT006 186 | 186 184 | 184 186 | 184 186 | 184 UDP98-409 123 | 123 134 | 134 123 | 134 123 | 134 EPDCU3117 165 | 165 165 | 165 165 | 165 165 | 165 * PK = Prunus kansuensis wild peach; FG = ‘Flordaguard’ peach; Ø, null allele.

132

Table B-6. Microsatellite allele configurations and genotype cross combinations for ‘UFSharp’ x (‘Okinawa’ x Prunus kansuensis) species-level backcrosses inferred from segregation patterns of marker band and electropherogram data in BC1F1 progenies. SH x [OK x PK]2 * SH x [OK x PK]3 * Linkage Marker name Male parent Female parent Cross type Male parent Female parent Cross type group [OK x PK]2 SH [OK x PK]3 SH LG 1 CPPCT016 191 | 177 189 | 193 AB x CD 191 | 173 189 | 193 AB x CD CPSCT008 182 | 234 182 | 182 AB x AA 182 | 234 182 | 182 AB x AA CPPCT027 107 | 82 105 | 80 AB x CD 107 | 89 105 | 80 AB x CD CPSCT027 137 | 167 137 | 137 AB x AA 137 | 169 137 | 137 AB x AA CPPCT026 191 | 191 174 | 176 (NI) 191 | 189 174 | 176 AB x CD BPPCT027 244 | 254 244 | 244 AB x AA 244 | 238 244 | 244 AB x AA CPPCT029 170 | Ø 173 | 185 AØ x BC 170 | Ø 173 | 185 AØ x BC BPPCT028 163 | Ø 163 | 159 AØ x AB 163 | Ø 163 | 159 AØ x AB LG 2 UDP98-025 111 | 115 127 | 127 AB x CC 111 | 115 127 | 127 AB x CC CPDCT044 245 | 172 219 | 227 (NI) 245 | 170 219 | 227 (NI) BPPCT013 178 | 208176 182223 | 182247 AB x CC 178 | 208192 182223 | 182247 AB x CC UDP96-013 195 | 233 203 | 203 AB x CC 195 | 203 203 | 203 AB x BB BPPCT030 156 | 173 171 | 171 AB x CC 156 | 150 171 | 171 AB x CC CPSCT034 237 | 237 237 | 235 (NI) 237 | 245 237 | 235 AB x AC LG 3 BPPCT007 140 | 109 140 | 138 AB x AC 140 | 109 140 | 138 AB x AC CPDCT025 183 | 183 181 | 181 (NI) 183 | 183 181 | 181 (NI) CPDCT027 162 | 146 162 | 162 AB x AA 162 | 146 162 | 162 AB x AA LG 4 CPPCT005 171 | 132 148 | 173 AB x CD 171 | 157 148 | 173 AB x CD UDP96-003 134 | 120 119 | 136 AB x CD 134 | 120 119 | 136 AB x CD EPPISF032 198 | 198 202 | 202 (NI) 198 | 186 202 | 202 AB x CC CPDCT014 108 | 91 113 | 113 AB x CC 108 | 93 113 | 113 AB x CC LG 5 BPPCT026 149 | 158 129 | 129 AB x CC 149 | 158 129 | 129 AB x CC BPPCT017 148 | 162 176 | 176 AB x CC 148 | 162 176 | 176 AB x CC BPPCT038 154 | 162 148 | 148 AB x CC 154 | 162 148 | 148 AB x CC CPDCT022 146 | 157 146 | 161 AB x AC 146 | 157 146 | 161 AB x AC

133

Table B-6. Continued. SH x [OK x PK]2 * SH x [OK x PK]3 * Linkage Marker name Male parent Female parent Cross type Male parent Female parent Cross type group [OK x PK] 2 SH [OK x PK] 3 SH LG 6 CPPCT008 157 | 157 157 | 157 (NI) 157 | 157 157 | 157 (NI) BPPCT008 154 | 105 133 | 150 AB x CD 154 | 105 133 | 150 AB x CD BPPCT025 187 | 183 173 | 193 AB x CD 187 | 181 173 | 193 AB x CD CPPCT030 191 | 239 191 | 188 AB x AC 191 | 239 191 | 188 AB x AC LG 7 CPPCT039 98 | 104 98 | 98 AB x AA 98 | 104 98 | 98 AB x AA CPPCT022 260 | 292 283 | 289 AB x CD 260 | 292 283 | 289 AB x CD UDP98-405 103 | 99 103 | 103 AB x AA 103 | 99 103 | 103 AB x AA UDP98-408 100 | 89 100 | 98 AB x AC 100 | 89 100 | 98 AB x AC CPPCT033 142 | 143 144 | 151 AB x CD 142 | 143 144 | 151 AB x CD PMS2 126 | 140 128 | 128 AB x CC 126 | 140 128 | 128 AB x CC CPPCT017 165 | 169 175 | 187 AB x CD 165 | 169 175 | 187 AB x CD EPDCU3392 117 | 123 120 | 120 AB x CC 117 | 123 120 | 120 AB x CC LG 8 CPPCT019B 173 | 175 173 | 181 (NI) 173 | 175 173 | 181 (NI) BPPCT006 b 126 | 106 114 | 124183 AB x CD 126 | 106 114 | 124183 AB x CD UDP96-019 213 | 214 215 | 217 AB x CD 213 | 214 215 | 217 AB x CD CPPCT006 Ø | 184 186 | 186 ØA x BB Ø | 184 186 | 186 ØA x BB UDP98-409 123 | 134 123 | 125 AB x AC 123 | 134 123 | 125 AB x AC EPDCU3117 165 | 165 156 | 156 (NI) 165 | 165 156 | 156 (NI) * PK = P. kansuensis wild peach; OK = ‘Okinawa’ peach; SH = ‘UFSharp’ peach; Ø, null allele; NI, not informative.

134

Table B-7. Microsatellite allele configurations and genotype cross combinations for ‘UFSharp’ x (‘Flordaguard’ x Prunus kansuensis) species-level backcrosses inferred from segregation patterns of marker band and electropherogram data in BC1F1 progenies. SH x [FG x PK]1 * SH x [FG x PK]6 * Linkage Marker name Male parent Female parent Cross type Male parent Female parent Cross type group [FG x PK]1 SH [FG x PK]6 SH LG 1 CPPCT016 199 | 177 189 | 193 AB x CD 199 | 177 189 | 193 AB x CD CPSCT008 175 | 234 182 | 182 AB x CC 175 | 234 182 | 182 AB x CC CPPCT027 105 | 89 105 | 80 AB x AC 105 | 89 105 | 80 AB x AC CPSCT027 137 | 169 137 | 137 AB x AA 137 | 169 137 | 137 AB x AA CPPCT026 191 | 189 174 | 176 AB x CD 174 | 189 174 | 176 AB x AC BPPCT027 244 | 238 244 | 244 AB x AA 244 | 238 244 | 244 AB x AA CPPCT029 170 | Ø 173 | 185 AØ x BC 170 | Ø 173 | 185 AØ x BC BPPCT028 163 | Ø 163 | 159 AØ x AB 163 | Ø 163 | 159 AØ x AB LG 2 UDP98-025 111 | 115 127 | 127 AB x CC 111 | 115 127 | 127 AB x CC CPDCT044 215 | 172 219 | 227 (NI) 223 | 172 219 | 227 (NI) BPPCT013 178223 | 208192 182223 | 182247 AB x CC 182247 | 208192 182223 | 182247 AB x AA UDP96-013 195247 | 203 203 | 203 AB x BB 195 | 233 203 | 203 AB x CC BPPCT030 156 | 173 171 | 171 AB x CC 156 | 150 171 | 171 AB x CC CPSCT034 237 | 237 237 | 235 (NI) 237 | 245 237 | 235 (NI) LG 3 BPPCT007 140 | 109 140 | 138 AB x AC 140 | 109 140 | 138 AB x AC CPDCT025 185 | 183 181 | 181 AB x CC 181 | 183 181 | 181 AB x AA CPDCT027 162 | 146 162 | 162 AB x AA 162 | 146 162 | 162 AB x AA LG 4 CPPCT005 148 | 132 148 | 173 AB x AC 148 | 132 148 | 173 AB x AC UDP96-003 129 | 120 119 | 136 AB x CD 129 | 120 119 | 136 AB x CD EPPISF032 198 | 198 202 | 202 (NI) 198 | 186 202 | 202 AB x CC CPDCT014 Ø | 91 108 113 | 113 AB x CC 113 | 93 113 | 113 AB x CC LG 5 BPPCT026 137 | 158 129 | 129 AB x CC 129 | 158 129 | 129 AB x AA BPPCT017 148 | 162 176 | 176 AB x CC 148 | 162 176 | 176 AB x CC BPPCT038 156 | 148 148 | 148 AB x BB 156 | 148 148 | 148 AB x BB CPDCT022 146 | 157 146 | 161 AB x AC 146 | 157 146 | 161 AB x AC

135

Table B-7. Continued. SH x [FG x PK]1 * SH x [FG x PK]6 * Linkage Marker name Male parent Female parent Cross type Male parent Female parent Cross type group [FG x PK]1 SH [FG x PK]6 SH LG 6 CPPCT008 157 | 157 157 | 157 (NI) 157 | 157 157 | 157 (NI) BPPCT008 133 | 105 133 | 150 AB x AC 133 | 105 133 | 150 AB x AC BPPCT025 189 | 183 173 | 193 AB x CD 189 | 183 173 | 193 AB x CD CPPCT030 193 | 249 191 | 188 AB x CD 193 | 239 191 | 188 AB x CD LG 7 CPPCT039 91 | 104 98 | 98 AB x CC 91 | 108 98 | 98 AB x CC CPPCT022 214 | 292 283 | 289 AB x CD 260 | 290 283 | 289 AB x CD UDP98-405 99 | 99 103 | 103 (NI) 103 | 99 103 | 103 AB x AA UDP98-408 102 | 89 100 | 98 AB x CD 100 | 89 100 | 98 AB x AC CPPCT033 142 | 160 144 | 151 AB x CD 142 | 160 144 | 151 AB x CD PMS2 128 | 150 128 | 128 AB x AA 126 | 150 128 | 128 AB x CC CPPCT017 175 | 169 175 | 187 AB x AC 165 | 169 175 | 187 AB x CD EPDCU3392 120 | 123 120 | 120 AB x AA 117 | 123 120 | 120 AB x CC LG 8 CPPCT019B 173 | 181 173 | 181 (NI) 173 | 181 173 | 181 (NI) BPPCT006 114 | 106 114 | 124183 AB x AC 114 | 106 114 | 124183 AB x AC UDP96-019 215 | 215 215 | 217 (NI) 215 | 215 215 | 217 (NI) CPPCT006 186 | 184 186 | 186 AB x AA 186 | 184 186 | 186 AB x AA UDP98-409 123 | 134 123 | 125 AB x AC 123 | 134 123 | 125 AB x AC EPDCU3117 165 | 165 156 | 156 (NI) 165 | 165 156 | 156 (NI) * PK = P. kansuensis wild peach; FG = ‘Flordaguard’ peach; SH = ‘UFSharp’ peach; Ø, null allele; NI, not informative.

136

Table B-8. Microsatellite allele configurations and genotype cross combinations for ‘Okinawa’ x (‘Flordaguard’ x Prunus kansuensis) species-level backcrosses inferred from segregation patterns of marker band and electropherogram data in BC1F1 progenies. b FG x [OK x PK]3 population Linkage Marker name Male parent Female parent Cross type group [OK x PK]3 FG LG 1 CPPCT016 191 | 173 191 | 199 AB x AC CPSCT008 182 | 234 175 | Ø AB x CØ CPPCT027 107 | 89 105 | 105 AB x CC CPSCT027 137 | 169 137 | 137 AB x AA CPPCT026 191 | 189 191 | 174 AB x AC BPPCT027 244 | 238 244 | 244 AB x AA CPPCT029 170 | Ø 170 | 170 (NI) BPPCT028 163 | Ø 163 | 163 (NI) LG 2 UDP98-025 111 | 115 111 | 111 AB x AA CPDCT044 245 | 170 215 223 | 247 (NI) BPPCT013 178 | 208192 178 | 182 AB x AC UDP96-013 195 | 203 195 | 195 AB x AA BPPCT030 156 | 150 156 | 156 AB x AA CPSCT034 237 | 245 237 | 237 AB x AA LG 3 BPPCT007 140 | 109 140 | 140 AB x AA CPDCT025 183 | 183 181 | 185 AA x BC CPDCT027 162 | 146 162 | 162 AB x AA LG 4 CPPCT005 171 | 157 157 | 148 AB x BC UDP96-003 134 | 120 125 | 129 AB x CD EPPISF032 198 | 186 198 | 200 AB x AC CPDCT014 108 | 93 113 | Ø AB x CØ LG 5 BPPCT026 149 | 158 129 | 137 AB x CD BPPCT017 148 | 162 148 | 158 AB x AC BPPCT038 154 | 162 156 | 160 AB x CD CPDCT022 146 | 157 146 | 146 AB x AA

137

Table B-8. Continued. FG x [OK x PK]3 population * Linkage Marker name Male parent Female parent Cross type group [OK x PK]3 FG LG 6 CPPCT008 157 | 157 157 | 164 AA x AB BPPCT008 154 | 105 133 | 133 AB x CC BPPCT025 187 | 181 189 | 189 AB x CC CPPCT030 191 | 239 193 | 193 AB x CC LG 7 CPPCT039 98 | 104 98 | 91 AB x AC CPPCT022 260 | 292 260 | 214 AB x AC UDP98-405 103 | 99 103 | 99 AB x AB UDP98-408 100 | 89 100 | 102 AB x AC CPPCT033 142 | 143 142 | 142 AB x AA PMS2 126 | 140 126 | 128 AB x AC CPPCT017 165 | 169 165 | 175 AB x AC EPDCU3392 117 | 123 117 | 120 AB x AC LG 8 CPPCT019B 173 | 175 173 | 181 AB x AC BPPCT006 126 | 106 114 | 114 AB x CC UDP96-019 213 | 214 215 | 215 AB x CC CPPCT006 Ø | 184 186 | 186 ØA x BB UDP98-409 123 | 134 123 | 123 AB x AA EPDCU3117 165 | 165 165 | 165 (NI) * PK = P. kansuensis wild peach; FG = ‘Flordaguard’ peach; OK = ‘Okinawa’ peach; Ø, null allele; NI, not informative.

138

Table B-9. Microsatellite markers used to construct linkage maps and amplification status in each population.

Polymorphic/scored (+) in population * Linkage Marker SH x SH x FG x group OK x PK FG x PK [OK x PK] [FG x PK] [OK x PK] 1 3 4 5 6 2 3 1 6 1 6 3 LG 1 CPPCT016 + + + + + + + + + + + + CPSCT008 ++ ++ ++ ++ ++ ++ ++ ++ ++ D D D CPPCT027 + + + + + + + + + + + + CPSCT027 + + + + + + + + + + + + CPPCT026 NI D D D D NI + + + + + + BPPCT027 + + + + + + + + + + + + CPPCT029 M M M M M +Ø +Ø M M +Ø +Ø M BPPCT028 NS NS NS NS NS +Ø +Ø NS NS +Ø +Ø M LG 2 UDP98-025 + + + + + + + + + + + + BPPCT013 + + + + + + + + + + + + UDP96-013 + + + + + + NI + + NI + + BPPCT030 + + + + + + + + + + + + CPSCT034 + + NS NS + NS ++ NS + NS ++ + LG 3 BPPCT007 + + + + + + + + + + + + CPDCT025 M M M M M NS NS + + + + NI CPDCT027 + + + + + + + + + + + + LG 4 CPPCT005 + + + + + + + + + + + NI UDP96-003 + + + + + D + + + + + + EPPISF032 NI + + NI NI NI + NI NI NI + + CPDCT014 + + + + + + + + + + + + LG 5 BPPCT026 + + + + + + + + + + + + BPPCT017 + + + + + + + + + + + + BPPCT038 + + + + + + + + + NI NI + CPDCT022 + + + + + + + + + + + +

139

Table B-9. Continued.

Polymorphic/scored (+) in population Linkage Marker SH x SH x FG x group OK x PK FG x PK [OK x PK] [FG x PK] [OK x PK] 1 3 4 5 6 2 3 1 6 1 6 3 LG 6 BPPCT008 + + + + + + + + + + + + BPPCT025 + + + + + D + + + + + + CPPCT030 + + + + + + + + + + + + LG 7 CPPCT039 + + + + + + + + + + + + CPPCT022 + + + + + + + + + + + + UDP98-405 + + + + + + + NI + NI + NI UDP98-408 + + + + + + + + + + + + CPPCT033 + + + + + + + + + + + + PMS2 + + + + + + + + + + + + CPPCT017 + + + + + + + + + + + + EPDCU3392 + + + + + + + + + + + + LG 8 CPPCT019B D D D D D D D ++ ++ D D D BPPCT006 + + + + + + + + + + + + CPPCT006 ++ ++ ++ ++ ++ D D D D D ++ + UDP98-409 + + + + + + + + + + + + * D: difficult to score after repeated amplification, excluded from linkage files to avoid erroneous marker data; NI: not informative; NS: not segregating; M: monomorphic; +Ø: null alleles from Prunus kansuensis revealed in the backcross population; ++: marker scores verified after several rounds of amplification.

140

Table B-10. Chi-square test for segregation distortion in twelve interspecific F2 and BC1F1 populations derived from peach x Prunus kansuensis crosses. Linkage [OK x PK]1 [OK x PK]3 Marker group a h b - X2 P a h b - X2 P LG 1 CPPCT016 15 14 7 0 5.33 * 8 20 8 0 0.44 - CPSCT008 12 17 7 0 1.50 - 8 21 7 0 1.06 - CPPCT027 13 15 8 0 2.39 - 7 22 7 0 1.78 - CPSCT027 14 13 9 0 4.17 - 9 20 7 0 0.67 - BPPCT027 13 18 4 1 4.66 - 10 24 2 0 7.56 ** LG 2 UDP98-025 10 18 8 0 0.22 - 9 16 11 0 0.67 - BPPCT013 11 21 3 1 5.06 * 13 15 8 0 2.39 - UDP96-013 10 18 4 4 2.75 - 13 15 8 0 2.39 - BPPCT030 12 20 4 0 4.00 - 10 20 6 0 1.33 - CPSCT034 12 17 7 0 1.50 - 10 21 5 0 2.39 - LG 3 BPPCT007 10 18 8 0 0.22 - 12 17 7 0 1.50 - CPDCT027 6 22 8 0 2.00 - 7 23 5 1 3.69 - LG 4 CPPCT005 11 18 7 0 0.89 - 9 19 8 0 0.17 - UDP96-003 7 20 9 0 0.67 - 9 22 5 0 2.67 - EPPISF032 10 21 5 0 2.39 - CPDCT014 8 13 15 0 7.55 * 8 25 3 0 6.83 ** LG 5 BPPCT026 10 18 8 0 0.22 - 9 19 8 0 0.17 - BPPCT017 16 16 4 0 8.44 ** 8 21 7 0 1.06 - BPPCT038 18 12 6 0 12.00 **** 13 13 10 0 3.28 - CPDCT022 15 15 6 0 5.50 * 12 15 9 0 1.50 - LG 6 BPPCT008 12 24 0 0 12.00 **** 14 16 6 0 4.00 - BPPCT025 16 19 0 1 14.89 ***** 14 17 5 0 4.61 * CPPCT030 18 16 0 2 19.18 ******* 18 18 0 0 18.00 ****** LG 7 CPPCT039 9 18 9 0 0.00 - 6 19 11 0 1.50 - CPPCT022 11 14 11 0 1.78 - 6 19 11 0 1.50 - UDP98-405 11 14 11 0 1.78 - 7 21 8 0 1.06 - UDP98-408 12 13 11 0 2.83 - 8 21 7 0 1.06 - CPPCT033 12 16 8 0 1.33 - 14 16 6 0 4.00 - PMS2 13 14 9 0 2.67 - 14 15 7 0 3.72 - CPPCT017 18 10 8 0 12.67 **** 14 15 7 0 3.72 - EPDCU3392 14 10 7 5 7.06 ** 13 10 6 7 6.17 ** LG 8 BPPCT006 14 15 7 0 3.72 - 9 22 5 0 2.67 - CPPCT006 15 14 4 3 8.09 ** 6 20 3 7 4.79 * UDP98-409 14 19 3 0 6.83 ** 6 23 7 0 2.83 - P, Significance levels: *, <0.1 **, <0.05 ***, <0.01 ****, <0.005 *****, <0.001 ******, <0.0005 *******, <0.0001.

141

Table B-10. Continued. Linkage [OK x PK]4 [OK x PK]5 [OK x PK]6 Marker group a h b - X2 P a h b - X2 P a h b - X2 P LG 1 CPPCT016 5 15 9 0 1.14 - 7 14 2 0 3.26 - 11 11 6 0 3.07 - CPSCT008 6 16 7 0 0.38 - 6 14 3 0 1.87 - 10 10 8 0 2.57 - CPPCT027 4 19 6 0 3.07 - 8 12 3 0 2.22 - 12 13 3 0 5.93 * CPSCT027 4 18 7 0 2.31 - 8 11 3 1 2.27 - 12 12 3 1 6.33 ** BPPCT027 5 22 2 0 8.38 ** 12 10 1 0 10.91 **** 12 14 2 0 7.14 ** LG 2 UDP98-025 8 14 7 0 0.10 - 4 14 5 0 1.17 - 5 17 6 0 1.36 - BPPCT013 10 12 7 0 1.48 - 5 12 6 0 0.13 - 5 18 5 0 2.29 - UDP96-013 11 11 6 1 3.07 - 6 12 5 0 0.13 - 5 18 5 0 2.29 - BPPCT030 11 14 4 0 3.41 - 6 13 4 0 0.74 - 6 16 6 0 0.57 - CPSCT034 6 15 7 0 0.21 - LG 3 BPPCT007 4 13 12 0 4.72 * 7 10 6 0 0.48 - 8 14 6 0 0.29 - CPDCT027 7 13 9 0 0.59 - 3 14 6 0 1.87 - 6 12 10 0 1.71 - LG 4 CPPCT005 11 11 7 0 2.79 - 3 17 3 0 5.26 * 9 10 8 1 1.89 - UDP96-003 7 16 6 0 0.38 - 3 17 3 0 5.26 * 12 7 7 2 7.46 ** EPPISF032 8 13 8 0 0.31 - CPDCT014 5 22 2 0 8.38 ** 4 13 6 0 0.74 - 7 10 9 2 1.69 - LG 5 BPPCT026 5 12 12 0 4.24 - 8 13 2 0 3.52 - 4 17 7 0 1.93 - BPPCT017 5 12 12 0 4.24 - 6 14 3 0 1.87 - 7 16 5 0 0.86 - BPPCT038 6 13 10 0 1.41 - 7 12 4 0 0.83 - 10 12 4 2 2.92 - CPDCT022 5 13 11 0 2.79 - 4 13 6 0 0.74 - 11 11 6 0 3.07 - LG 6 BPPCT008 8 16 5 0 0.93 - 10 13 0 0 9.09 ** 9 15 3 1 3.00 - BPPCT025 11 16 1 1 7.71 ** 16 7 0 0 25.78 ******* 8 14 1 5 5.35 * CPPCT030 12 17 0 0 10.79 **** 14 9 0 0 18.13 ****** 12 16 0 0 10.86 **** LG 7 CPPCT039 7 9 13 0 6.66 ** 8 8 7 0 2.22 - 4 15 9 0 1.93 - CPPCT022 9 8 12 0 6.45 ** 6 11 6 0 0.04 - 3 16 9 0 3.14 - UDP98-405 9 9 11 0 4.45 - 6 11 6 0 0.04 - 3 14 10 1 3.67 - UDP98-408 9 10 10 0 2.86 - 5 12 6 0 0.13 - 3 14 10 1 3.67 - CPPCT033 9 13 7 0 0.59 - 7 13 3 0 1.78 - 2 16 10 0 5.14 * PMS2 6 14 9 0 0.66 - 7 11 5 0 0.39 - 3 14 10 1 3.67 - CPPCT017 8 11 10 0 1.97 - 6 13 4 0 0.74 - 5 15 8 0 0.79 - EPDCU3392 7 11 10 1 1.93 - 4 14 5 0 1.17 - 5 13 7 3 0.36 - LG 8 BPPCT006 9 13 7 0 0.59 - 6 12 5 0 0.13 - 10 13 5 0 1.93 - CPPCT006 11 8 3 7 7.45 ** 6 9 4 4 0.47 - 10 11 2 5 5.61 * UDP98-409 10 13 6 0 1.41 - 5 12 6 0 0.13 - 6 17 5 0 1.36 - P, Significance levels: *, <0.1 **, <0.05 ***, <0.01 ****, <0.005 *****, <0.001 ******, <0.0005 *******, <0.0001.

142

Table B-10. Continued. Linkage SH x [OK x PK]2 SH x [OK x PK]3 FG x [OK x PK]3 Marker group a h - X2 P a h - X2 P a h - X2 P LG 1 CPPCT016 17 16 0 0.03 - 11 17 2 1.29 - 9 11 0 0.20 - CPSCT008 22 11 0 3.67 * 23 7 0 8.53 **** CPPCT027 15 18 0 0.27 - 15 15 0 0.00 - 12 8 0 0.80 - CPSCT027 11 22 0 3.67 * 15 11 4 0.62 - 13 7 0 1.80 - CPPCT026 19 11 0 2.13 - 10 10 0 0.00 - BPPCT027 15 18 0 0.27 - 20 10 0 3.33 * 10 10 0 0.00 - CPPCT029 15 18 0 0.27 - 21 7 2 7.00 *** BPPCT028 16 17 0 0.03 - 23 7 0 8.53 **** LG 2 UDP98-025 13 20 0 1.48 - 12 16 2 0.57 - 9 11 0 0.20 - BPPCT013 11 22 0 3.67 * 13 17 0 0.53 - 11 9 0 0.20 - UDP96-013 11 22 0 3.67 * 11 9 0 0.20 - BPPCT030 9 24 0 6.82 *** 12 18 0 1.20 - 11 9 0 0.20 - CPSCT034 13 17 0 0.53 - 12 8 0 0.80 - LG 3 BPPCT007 9 24 0 6.82 *** 15 15 0 0.00 - 5 15 0 5.00 ** CPDCT027 18 15 0 0.27 - 11 19 0 2.13 - 8 12 0 0.80 - LG 4 CPPCT005 19 14 0 0.76 - 11 17 2 1.29 - UDP96-003 14 14 2 0.00 - 7 13 0 1.80 - EPPISF032 17 13 0 0.53 - 7 13 0 1.80 - CPDCT014 18 14 1 0.50 - 16 14 0 0.13 - 7 10 3 0.53 - LG 5 BPPCT026 20 12 1 2.00 - 12 18 0 1.20 - 13 7 0 1.80 - BPPCT017 18 15 0 0.27 - 12 17 1 0.86 - 11 9 0 0.20 - BPPCT038 21 12 0 2.45 - 13 17 0 0.53 - 10 10 0 0.00 - CPDCT022 17 16 0 0.03 - 12 18 0 1.20 - 10 10 0 0.00 - LG 6 BPPCT008 15 18 0 0.27 - 19 9 2 3.57 * 12 8 0 0.80 - BPPCT025 19 7 4 5.54 ** 11 9 0 0.20 - CPPCT030 14 19 0 0.76 - 19 9 2 3.57 * 9 11 0 0.20 - LG 7 CPPCT039 12 21 0 2.45 - 12 18 0 1.20 - 8 12 0 0.80 - CPPCT022 11 22 0 3.67 * 13 17 0 0.53 - 6 14 0 3.20 * UDP98-405 10 20 3 3.33 * 13 16 1 0.31 - UDP98-408 10 23 0 5.12 ** 14 16 0 0.13 - 7 12 1 1.32 - CPPCT033 15 18 0 0.27 - 11 18 1 1.69 - 8 12 0 0.80 - PMS2 10 19 4 2.79 * 11 16 3 0.93 - 8 12 0 0.80 - CPPCT017 15 19 0 0.27 - 9 21 0 4.80 ** 12 8 0 0.80 - EPDCU3392 15 17 1 0.12 - 13 17 0 0.53 - 10 6 4 1.00 - LG 8 BPPCT006 20 13 0 1.48 - 9 20 1 4.17 ** 8 12 0 0.80 - CPPCT006 8 12 0 0.80 - UDP98-409 17 16 0 0.03 - 10 19 1 2.79 * 9 11 0 0.20 - P, Significance levels: *, <0.1, ** <0.05 ***, <0.01 ****, <0.005 *****, <0.001 ******, <0.0005 *******, <0.0001.

143

Table B-10. Continued. Linkage Marker [FG x PK]1 [FG x PK]6 SH x [FG x PK]1 SH x [FG x PK]6 group a h b - X2 P a h b - X2 P a h - X2 P a h - X2 P LG 1 CPPCT016 11 12 6 0 2.59 - 6 17 2 0 4.52 - 12 6 0 2.00 - 20 20 0 0.00 - CPSCT008 7 16 6 0 0.38 - 7 17 1 0 6.12 ** CPPCT027 6 18 5 0 1.76 - 9 12 4 0 2.04 - 10 8 0 0.22 - 16 24 0 1.60 - CPSCT027 7 16 5 1 0.86 - 9 12 4 0 2.04 - 11 7 0 0.89 - 15 21 4 1.00 - CPPCT026 8 17 4 0 1.97 - 11 12 2 0 6.52 ** 11 7 0 0.89 - 20 20 0 0.00 - BPPCT027 10 16 3 0 3.69 - 10 13 2 0 5.16 * 8 10 0 0.22 - 21 19 0 0.10 - CPPCT029 10 8 0 0.22 - 19 21 0 0.10 - BPPCT028 12 5 1 2.88 * 22 17 1 0.64 - LG 2 UDP98-025 6 14 9 0 0.66 - 5 11 9 0 1.64 - 9 9 0 0.00 - 19 21 0 0.10 - BPPCT013 6 14 9 0 0.66 - 5 14 6 0 0.44 - 12 6 0 2.00 - 21 14 5 1.40 - UDP96-013 5 14 9 1 1.14 - 5 14 6 0 0.44 - 23 17 0 0.90 - BPPCT030 6 14 9 0 0.66 - 4 15 6 0 1.32 - 11 7 0 0.89 - 25 15 0 2.50 - CPSCT034 4 17 4 0 3.24 - 24 16 0 1.60 - LG 3 BPPCT007 9 10 10 0 2.86 - 4 15 6 0 1.32 - 3 15 0 8.00 **** 17 23 0 0.90 - CPDCT025 12 9 8 0 5.28 * 6 17 2 0 4.52 - 5 13 0 3.56 * 21 19 0 0.10 - CPDCT027 8 16 5 0 0.93 - 9 12 4 0 2.04 - 5 13 0 3.56 * 25 15 0 2.50 - LG 4 CPPCT005 5 18 6 0 1.76 - 9 13 3 0 2.92 - 13 3 2 6.25 ** 30 9 1 11.31 ***** UDP96-003 7 14 8 0 0.10 - 10 11 4 0 3.24 - 13 3 2 6.25 ** 29 11 0 8.10 **** EPPISF032 8 12 5 0 0.76 - 30 10 0 10.00 **** CPDCT014 7 12 6 0 0.12 - 8 8 2 0.00 - 28 12 0 6.40 ** LG 5 BPPCT026 8 9 12 0 5.28 * 2 15 8 0 3.88 - 10 8 0 0.22 - 18 22 0 0.40 - BPPCT017 7 13 9 0 0.59 - 3 15 7 0 2.28 - 9 9 0 0.00 - 21 19 0 0.10 - BPPCT038 10 10 9 0 2.86 - 3 18 4 0 4.92 * CPDCT022 9 12 8 0 0.93 - 5 15 5 0 1.00 - 13 5 0 3.56 * 21 19 0 0.10 - LG 6 BPPCT008 13 10 6 0 6.17 ** 11 11 3 0 5.48 * 14 4 0 5.56 ** 22 18 0 0.40 - BPPCT025 16 7 6 0 14.66 ***** 12 12 1 0 9.72 *** 15 3 0 8.00 **** 23 14 3 2.19 - CPPCT030 17 11 1 0 19.34 ******* 17 8 0 0 26.36 ******* 10 8 0 0.22 - 25 15 0 2.50 - LG 7 CPPCT039 7 15 7 0 0.03 - 7 11 7 0 0.36 - 8 9 1 0.06 - 18 22 0 0.40 - CPPCT022 6 16 7 0 0.38 - 7 11 7 0 0.36 - 9 9 0 0.00 - 18 22 0 0.40 - UDP98-405 7 10 7 1 0.67 16 24 0 1.60 - UDP98-408 7 14 8 0 0.10 - 7 10 8 0 1.08 - 8 8 2 0.00 - 18 22 0 0.40 - CPPCT033 9 16 4 0 2.03 - 5 13 7 0 0.36 - 10 8 0 0.22 - 21 19 0 0.10 - PMS2 7 16 6 0 0.38 - 6 12 7 0 0.12 - 10 8 0 0.22 - 18 22 0 0.40 - CPPCT017 8 15 6 0 0.31 - 5 13 7 0 0.36 - 6 12 0 2.00 - 19 21 0 0.10 - EPDCU3392 7 16 6 0 0.38 - 5 13 7 0 0.36 - 6 12 0 2.00 - 19 20 1 0.03 - LG 8 CPPCT019B 13 11 5 0 6.10 ** 10 10 4 1 3.67 - BPPCT006 13 10 6 0 6.17 ** 10 11 4 0 3.24 - 15 3 0 8.00 **** 25 15 0 2.50 - CPPCT006 12 3 3 5.40 ** 22 18 0 0.40 - UDP98-409 9 16 4 0 2.03 - 7 16 2 0 3.96 - 10 7 1 0.53 - 23 17 0 0.90 - P, Significance levels: *, <0.1, ** <0.05 ***, <0.01 ****, <0.005 *****, <0.001 ******, <0.0005 *******, <0.0001.

144

Table B-11. Combined linkage map based on five F2 mapping populations derived from crosses between ‘Okinawa’ x Prunus kansuensis (OK x PK) compared with Prunus ‘Texas’ x ‘Earlygold’ (TxE) reference map. Linkage No. of Marker interval TxE TxE OK x PK TxE – OK OK x PK group mapped marker distance map x PK mean loci coverage covered length difference distance (%) a (cM) (cM) (cM) interval (cM) b LG 1 5 CPPCT016 – BPPCT027 52.9 46.0 46.7 0.7 11.7 LG 2 5 UDP98-025 – CPSCT034 77.5 39.0 38.3 -0.7 9.6 LG 3 2 BPPCT007 – CPDCT027 72.7 35.2 47.0 11.8 47.0 LG 4 4 c CPPCT005 – UDP96-003 c 28.6 17.9 16.7 -1.2 16.7 LG 5 4 c BPPCT026 – BPPCT038 c 56.4 27.7 32.6 4.9 16.3 LG 6 3 BPPCT008 – CPPCT030 59.9 50.1 23.3 -26.8 11.6 LG 7 8 CPPCT039 – EPDCU3392 71.7 50.6 45.1 -5.5 6.4 LG 8 3 BPPCT006 – UDP98-409 54.4 30.4 24.3 -6.1 12.2 Total 34 58.6 296.9 273.9 a Marker coverage per linkage group calculated as: (TxE distance covered ÷ TxE linkage group map length) * 100; Total coverage calculated as: (TxE distance covered ÷ TxE linkage group map length of 507.5 cM) * 100. b Average distance between adjacent markers. c Distances of three markers (EPPISF032 and CPDCT014 at LG4, and CPDCT022 at LG5 originally mapped in other populations), were outside the region spanned by the chromosomal segment and therefore not included in the calculations.

Table B-12. Combined linkage map based on two F2 mapping populations derived from crosses between ‘Flordaguard’ x Prunus kansuensis (FG x PK) compared with Prunus ‘Texas’ x ‘Earlygold’ (TxE) reference map. Linkage No. of Marker interval TxE TxE FG x PK TxE – FG FG x PK group mapped marker distance map x PK average loci coverage covered length difference distance (%) a (cM) (cM) (cM) interval per LG (cM) b LG1 6 CPPCT016 – BPPCT027 52.9 46.0 43.6 -2.4 8.7 LG2 5 UDP98-025 – CPSCT034 77.5 39.0 32.4 -6.6 8.1 LG3 3 BPPCT007 – CPDCT027 72.7 35.2 33.7 -1.5 16.9 LG4 4 c CPPCT005 – UDP96-003 c 28.6 17.9 9.9 -8.0 8.1 LG5 4 c BPPCT026 – BPPCT038 c 56.4 27.7 30.2 2.5 15.1 LG6 3 BPPCT008 – CPPCT030 59.9 50.1 21.8 -28.3 10.9 LG7 8 CPPCT039 – EPDCU3392 71.7 50.6 45.5 -5.1 6.5 LG8 3 CPPCT019B – UDP98-409 65.7 36.7 31.4 -5.3 15.7 Total 36 59.7 303.2 248.5 Notes: see Table B-11.

145

Table B-13. Combined linkage map based on three BC1 mapping populations of peach (‘Flordaguard’ or ‘UFSharp’) x F1 (‘Okinawa’ x Prunus kansuensis) (FG/SH x [OK x PK]) compared with Prunus ‘Texas’ x ‘Earlygold’ (TxE) reference map. Linkage No. of Marker interval TxE TxE FG/SH TxE – FG/SH group mapped marker distance x (OK x FG/SH x x (OK x loci coverage covered PK) (OK x PK) PK) (%) a (cM) map difference average length (cM) distance (cM) interval per LG (cM) b LG1 8 CPPCT016 – BPPCT028 87.5 76.1 82.8 6.7 11.8 LG2 4 UDP98-025 – CPSCT034 77.5 39.0 42.2 3.2 14.1 LG3 2 BPPCT007 – CPDCT027 72.7 35.2 78.0 42.8 78.0 LG4 4 c CPPCT005 – UDP96-003 c 28.6 17.9 24.1 6.2 24.1 LG5 4 c BPPCT026 – BPPCT038 c 56.4 27.7 28.6 0.9 14.3 LG6 3 BPPCT008 – CPPCT030 59.9 50.1 34.1 -16.0 17.1 LG7 7 CPPCT039 – EPDCU3392 71.7 50.6 38.6 -12.0 6.4 LG8 2 BPPCT006 – UDP98-409 54.4 30.4 22.8 -7.6 22.8 Total 34 64.4 327.0 351.2 Notes: see Table B-11.

Table B-14. Combined linkage map based on two BC1 mapping populations of peach (‘UFSharp’) x F1 (‘Flordaguard’ x Prunus kansuensis) (SH x [FG x PK]) compared with Prunus ‘Texas’ x ‘Earlygold’ (TxE) reference map. Linkage No. of Marker interval TxE TxE SH x TxE – SH SH x group mapped marker distance (FG x x (FG x (FG x loci coverage covered PK) PK) PK) (%) a (cM) map difference average length (cM) distance (cM) interval per LG (cM) b LG1 7 CPPCT016 – BPPCT028 87.5 76.1 90.3 14.2 15.0 LG2 5 UDP98-025 – CPSCT034 77.5 39.0 43.3 4.3 10.8 LG3 3 BPPCT007 – CPDCT027 72.7 35.2 50.2 15.0 25.1 LG4 4 c CPPCT005 – UDP96-003 28.6 17.9 15.5 -2.4 15.5 LG5 3 c BPPCT026 – BPPCT017 c 30.3 14.9 14.1 -0.8 14.1 LG6 3 BPPCT008 – CPPCT030 c 59.9 50.1 46.5 -3.6 23.2 LG7 8 CPPCT039 – EPDCU3392 71.7 50.6 43.1 -7.5 6.2 LG8 3 BPPCT006 – UDP98-409 54.4 30.4 21.8 -8.6 10.9 Total 36 61.9 314.2 324.8 Notes: see Table B-11.

146

LG1 LG2 LG3 LG4 LG5 LG6 LG7 LG8

LG1 TxE start LG2 TxE start LG3 TxE start LG4 TxE start LG5 TxE start LG6 TxE start LG7 TxE start LG8 TxE start 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.3 CPPCT016 UDP98-025 cpsct018 9.6 5.2 BPPCT026 2.5 epdcu3122 (cpdct032) 8.7 (cppct008) 9.5 (cpsct004) 7.8 CPPCT019B 7.1 cpdct042 12.5 cpdct044 CPPCT005 11.2 BPPCT007 10.4 14.1 CPPCT039 9.0 CPSCT008 20.2 bppct004 14.1 BPPCT006 18.6 CPPCT022 10.8 (cppct024a) bppct001 18.8 cppct035 20.9 (cpsct026) (cppct004b) bppct002 BPPCT017 22.3 20.8 udp96-019 11.6 20.1 UDP98-405 cppct010 23.6 cpsct044 24.1 bppct012 23.7 UDP98-408 13.6 (pchcms4) 25.0 BPPCT013 24.8 CPPCT006 UDP96-003 14.5 epdcu5100 UDP96-013 28.3 29.6 (bppct029) 27.8 30.1 BPPCT008 23.1 CPPCT027 (cpdct004) 32.9 BPPCT038 25.8 cpdct038 29.3 udp97-402 36.4 CPDCT025 CPSCT027 35.1 pchgms1 38.9 CPPCT033 29.2 udp96-005 36.3 (bppct024) 42.6 (cpdct023) 33.2 cppct003a 38.0 BPPCT030 46.4 CPDCT027 44.5 UDP98-409 PMS2 33.9 CPPCT026 39.4 (cpsct021) 48.4 LG3 TxE end 49.1 LG5 TxE end 47.8 46.7 epdcu3454 36.6 (cpsct024) 43.2 (cpsct031) 51.4 (pchcms2) 54.7 (epdcu3117) 37.2 (cpdct024) 45.4 (cpsct023) 56.4 BPPCT025 56.1 cpdct013d 55.9 LG8 TxE end (cpdct017) 48.6 CPSCT034 40.5 (cppct034) 50.3 LG2 TxE end 62.5 LG4 TxE end 61.8 CPPCT017 47.3 BPPCT027 62.7 EPPISF032 64.7 EPDCU3392 55.2 (bppct016) 65.1 CPPCT029 70.6 LG7 TxE end 66.5 (epdcu2862) 77.4 BPPCT028 78.2 CPDCT014 80.2 CPPCT030 83.7 LG6 TxE end 87.0 LG1 TxE end

94.4 CPDCT022

Figure B-1. Microsatellite SSR markers screened for polymorphism and selected for mapping the Mf resistance locus on each of the eight linkage groups (LG1-LG8) of the ‘Texas’ almond x ‘Earlygold’ (TxE) peach reference map. Marker names and positions (Kosambi map distances, cM) are indicated on the right and left side of bars, respectively. Markers selected for linkage analyses are denoted by bold, uppercase letters. Three markers (colored red and green) in LG 4 and LG5 were mapped to other interspecific Prunus linkage maps at distances beyond the length of the TxE map. Non-selected markers denoted by lowercase letters are characterized by one of the following: 1) non-polymorphic (enclosed in parentheses), 2) polymorphic but are spaced at very close intervals (bold letters), and 3) polymorphic but are difficult to score (italicized letters).

147

M M

PP PK F1 F2 genotypes

200bp-- A B H H B A A H H H A A B B B B H H H B H H H H A A H H H H H H H H X B H H A H H B B H -- H

100bp--

M M

PP PK F1 PP BC1F1 genotypes

200bp-- A B H A H A A A A A X H H H -- A H H A A H A H A X H H X A A H H H A H H H A H H H H H A H A

100bp--

Figure B-2. Representative gel images of genotyping for microsatellite marker UDP98-025 in F2 and BC1F1 mapping populations. Individuals were scored as “A” when only PP alleles (111 bp and/or 127 bp) were present, “B” when only PK alleles (115 bp) were present, and “H” when both PP and PK alleles were present. Outcrosses (X) and ambiguous genotypes (--) were excluded from the mapping analysis. PP, Prunus persica parent; PK, Prunus kansuensis parent; M, DNA size marker.

148

FG x PK LInkage Map

LG1 LG2 LG3 LG4 LG5 LG6 LG7 LG8

CPPCT039 0.0 CPPCT016 0.0 UDP98-025 0.0 BPPCT007 0.0 CPPCT005 0.0 BPPCT026 0.0 BPPCT008 0.0 0.0 CPPCT019B CPPCT022 3.1 UDP98-405 4.7 UDP98-408 6.2 CPSCT008 7.4 BPPCT006 9.9 UDP96-003 9.6 BPPCT025

15.0 BPPCT013 15.7 BPPCT017 16.2 UDP96-013 16.1 EPPISF032

19.5 BPPCT030 20.8 CPDCT025 21.3 CPPCT027 21.8 CPPCT030

26.1 CPSCT027 26.0 CPPCT033

30.2 BPPCT038 PMS2 31.4 UDP98-409 32.4 CPSCT034 31.8 33.7 CPDCT027 35.1 CPDCT014 34.6 CPDCT022

39.7 CPPCT026

43.6 BPPCT027 CPPCT017 45.5 EPDCU3392

Figure B-3. Combined SSR linkage map based on two F2 mapping populations derived from ‘Flordaguard’ (FG) x Prunus kansuensis (PK). Microsatellite markers were assigned into eight linkage groups (LG1-LG8) and ordered in accordance with the ‘Texas’ almond x ‘Earlygold’ peach linkage map. Numbers to the left side of the bars represent the estimated genetic distances (centiMorgan) of the microsatellite markers.

149

OK x PK Linkage Map

LG1 LG2 LG3 LG4 LG5 LG6 LG7 LG8

0.0 CPPCT016 0.0 UDP98-025 0.0 BPPCT007 0.0 CPPCT005 0.0 BPPCT026 0.0 BPPCT008 0.0 CPPCT039 0.0 BPPCT006

5.6 CPPCT022 7.6 CPPCT006 9.7 CPSCT008 10.2 BPPCT025 9.7 UDP98-405 11.6 UDP98-408

15.9 BPPCT017 16.9 BPPCT013 16.7 UDP96-003

20.0 UDP96-013

23.3 CPPCT030 24.0 CPPCT027 24.3 UDP98-409 25.2 BPPCT030 25.4 EPPISF032 26.2 CPSCT027 25.7 CPPCT033

31.2 CPDCT014 31.4 PMS2 32.6 BPPCT038

38.3 CPSCT034 40.0 CPDCT022

43.4 CPPCT017 45.1 EPDCU3392 46.7 BPPCT027 47.0 CPDCT027

Figure B-4. Combined SSR linkage map based on five F2 mapping populations derived from ‘Okinawa’ (OK) x Prunus kansuensis (PK). See notes in Figure B-2.

150

SH/FG x (OK x PK) Linkage Map

LG1 LG2 LG3 LG4 LG5 LG6 LG7 LG8

0.0 CPPCT016 0.0 UDP98-025 0.0 BPPCT007 0.0 CPPCT005 0.0 BPPCT026 0.0 BPPCT008 0.0 CPPCT039 0.0 BPPCT006 4.5 CPPCT022 6.1 CPSCT008 10.1 UDP98-408 14.0 BPPCT017 17.1 CPPCT033 19.0 BPPCT013 22.6 PMS2 22.8 UDP98-409 24.1 UDP96-003 23.2 BPPCT025 BPPCT038 29.8 BPPCT030 28.6

35.1 CPPCT027 34.1 CPPCT030 38.2 CPPCT017 38.6 EPDCU3392 42.5 CPSCT027 42.2 CPSCT034 44.6 CPDCT022 46.5 EPPISF032

59.0 CPPCT026 60.3 BPPCT027 63.8 CPDCT014

70.1 CPPCT029

78.0 CPDCT027 82.8 BPPCT028

Figure B-5. Combined SSR linkage map based on three BC1 mapping populations derived from peach (‘Flordaguard’ [FG] and ‘UFSharp’ [SH]) x F1 (‘Okinawa’ [OK] x Prunus kansuensis [PK]). See notes in Figure B-2.

151

SH x (FG x PK) Linkage Map

LG1 LG2 LG3 LG4 LG5 LG6 LG7 LG8

0.0 CPPCT016 0.0 UDP98-025 0.0 BPPCT007 0.0 CPPCT005 0.0 BPPCT026 0.0 BPPCT008 CPPCT039 0.0 BPPCT006 0.0 CPPCT022 UDP98-405 7.2 BPPCT025 3.0 4.8 UDP98-408 9.7 CPPCT006 14.1 BPPCT017 15.5 UDP96-003

21.8 UDP98-409 24.0 CPPCT033 BPPCT013 27.6 EPPISF032 28.2 29.3 PMS2 30.9 UDP96-013 CPPCT027 34.6 35.8 BPPCT030 34.7 CPDCT025

CPPCT017 CPSCT034 42.4 CPDCT014 43.1 43.3 EPDCU3392 46.2 CPSCT027 46.0 CPDCT022 46.5 CPPCT030 50.2 CPDCT027

57.9 CPPCT026

66.9 BPPCT027

79.2 CPPCT029

90.3 BPPCT028

Figure B-6. Combined SSR linkage map based on two BC1 mapping populations derived from peach (‘UFSharp’ [SH]) x F1 (‘Flordaguard’ [FG] x Prunus kansuensis [PK]). See notes in Figure B-2.

152

APPENDIX C SUPPLEMENTARY DATA FOR ROOTSTOCK EVALUATION

Table C-1. Growth attributes of ‘UFSun’ plants on five different rootstocks before trial establishment. Growth parameters a Rootstock Rootstock stem Scion stem b Crown height Ave. crown Pruning weight genotype circumference circumference (cm) diameter (cm) c (g) d (cm) (cm) Flordaguard 114.02 ± 14.0 85.20 ± 12.9 7.99 ± 0.8 7.06 ± 0.8 145.45 ± 72.0 Okinawa 156.21 ± 17.4 148.06 ± 18.2 14.28 ± 1.2 11.75 ± 1.2 693.89 ± 279.6 P-22 169.53 ± 13.4 143.56 ± 12.6 13.77 ± 0.8 11.15 ± 0.9 472.72 ± 147.5 MP-29 102.10 ± 24.8 101.02 ± 30.2 8.61 ± 1.5 6.55 ± 1.7 200.69 ± 156.9 Barton 149.36 ± 18.3 125.64 ± 19.4 10.98 ± 1.3 8.89 ± 1.2 400.75 ± 161.5 a Values are means of 10 replicates (average of six tree observations) ± standard deviation b Seedling rootstocks (‘Flordaguard’, ‘Okinawa’, and ‘Barton’) and clonal hybrid rootstocks (‘MP-29’ and ‘P-22’) were obtained from different nursery sources as budded or unbudded materials. Growth measurements were taken from plants raised in 5-gal nursery pots containing field soil for about 21 months, except for those on ‘MP-29’ and ‘Barton’ rootstocks, which were budded 19 months prior to growth measurements. c Calculated average of two perpendicular crown diameter measurements. d Plants were pruned to maintain an open-center system.

Table C-2. Pearson’s correlation coefficients among growth attributes of ‘UFSun’ plants before trial establishment. Rootstock stem Scion stem Crown height * Ave. crown No. circumference * circumference * diameter * observed

Scion stem circumference 0.926 286 Crown height 0.804 0.826 286 Ave. crown diameter 0.817 0.846 0.762 285 Pruning weight 0.634 0.656 0.553 0.636 273

* All values are statistically significant (P < 0.0001).

153

Table C-3. The ANOVA P values for traits considered in evaluating horticultural performance of ‘UFSun’ trees on five different rootstocks during three years of field establishment. Significance of effects b Source of variation a Rootstock Year R x Y Scion growth Trunk cross-sectional area (cm2) < .001 < .001 < .001 Trunk relative growth rate < .001 < .001 < .001 Pruning weight (g) < .001 < .001 < .001 Harvest Fruit yield (g) per tree 0.006 < .001 < .001 Yield efficiency (kg·cm-2 scion TCSA) 0.414 < .001 0.022 Mean fruit weight (g) < .001 < .001 < .001 Mean fruit no. per tree 0.083 < .001 < .001 Fruit load (fruit·cm-2 scion TCSA) 0.869 < .001 0.009 Fruit quality Mean diameter (mm) < .001 < .001 < .001 Flesh firmness (lbf) 0.451 < .001 0.053 Total soluble solids content (°Brix) 0.080 < .001 < .001 a All variables, except flesh firmness, were considered significant at α = 0.05 for rootstock x year (R x Y) interaction. b Analysis of variance was performed using the generalized linear mixed model incorporating block, block x rootstock, and block x year as random effects.

154

Table C-4. Means and ranges for yield parameters of ‘UFSun’ peach trees on different rootstocks during the first three years of field establishment (2014-2016) at Citra, FL. Rootstock No. trees Mean fruit no. / tree Total yield (kg) / tree Mean fruit weight (g) genotype observed Mean Max Mean Max Mean Min Max Min Min

Year 1 Flordaguard 46 1.65 0 - 10 111.95 0 - 605.3 34.74 0 - 84.0 Okinawa 54 7.98 0 - 30 498.23 0 - 1535.7 55.23 0 - 94.3 P-22 59 8.02 0 - 21 451.16 0 - 1179.2 52.81 0 - 80.1 MP-29 51 0.69 0 - 6 51.75 0 - 320.0 16.79 0 - 90.3 Barton 56 5.03 0 - 20 280.63 0 - 1169.1 51.61 0 - 122.7 Year 2 Flordaguard 53 23.12 2 - 56 2042.13 155.7 - 5311.2 88.11 63.6 - 118.5 Okinawa 59 45.37 1 - 121 3992.01 48.8 - 9498.9 89.18 48.8 - 120.1 P-22 60 36.11 2 - 99 3018.80 142.0 - 7478.0 84.92 70.7 - 106.3 MP-29 53 28.96 4 - 80 1983.39 282.0 - 4978.4 70.55 47.6 - 112.9 Barton 59 32.12 1 - 104 2619.60 77.3 - 7734.4 83.50 69.0 - 109.2 Year 3 Flordaguard 20 86.39 41 - 172 8781.22 4218.4 - 18438.5 101.89 75.3 - 124.7 Okinawa 25 78.73 46 - 202 7017.11 4263.8 - 17236.5 90.01 72.0 - 104.7 P-22 24 63.38 17 - 157 5901.07 1769.0 - 13517.0 94.71 82.9 - 108.1 MP-29 18 70.47 29 - 143 5616.58 2472.1 - 11226.4 81.75 62.6 - 102.6 Barton 25 52.64 17 - 108 4264.04 1428.8 - 8981.1 81.37 61.3 - 105.7

155

Block

1 2 3 4 5 6 7 8 9 10

6 M F O B M P B P F B 6 5 M F O B M P B P F B 5 4 M F O B M P B P F B 4 3 M F O B M P B P F B 3 2 M F O B M P B P F B 2 1 M F O B M P B P F B 1 6 P B M F B O O F O P 6 5 P B M F B O O F O P 5 4 P B M F B O O F O P 4 3 P B M F B O O F O P 3 2 P B M F B O O F O P 2 N 1 P B M F B O O F O P 1 6 B P B O F F M M B M 6 5 B P B O F F M M B M 5 4 B P B O F F M M B M 4 3 B P B O F F M M B M 3

Plant No. Plant 2 B P B O F F M M B M 2 1 B P B O F F M M B M 1 6 O M P P O M F O P F 6 S 5 O M P P O M F O P F 5 4 O M P P O M F O P F 4 3 O M P P O M F O P F 3 2 O M P P O M F O P F 2 1 O M P P O M F O P F 1 6 F O F M P B P B M O 6 5 F O F M P B P B M O 5 4 F O F M P B P B M O 4 3 F O F M P B P B M O 3 2 F O F M P B P B M O 2 1 F O F M P B P B M O 1

Figure C-1. Planting layout for the peach rootstock evaluation. ‘UFSun’ budded onto five different rootstocks: F = ‘Flordaguard’; O = ‘Okinawa’; B = ‘Barton’; P = ‘P-22’; M = ‘MP-29’. Experimental plot size: 200 ft x 130 ft. Colored letters represent trees with missing data due to tree mortality (red), non-budded trees (green), and broken main scaffold (blue). Trees retained in the third year are indicated by yellow boxes.

156

80 a a 70 b

Rootstock Scion ) 2 60 b a

b 50 b c b

40 c a a

30 sectional area(cm - b b a ab a b bc 20 a a bc c b a c c

10 b b b Trunkcross 0 F M B P O F M B P O F M B P O 2014 2015 2016

Figure C-2. Trunk cross-sectional area of ‘UFSun’ trees budded on five different rootstocks during three years (2014-2016) of establishment in the field at Citra, FL. Different letters on top of bars within the year cluster indicate significant differences at α = 0.05 according to a generalized linear mixed model analysis.

157

LIST OF REFERENCES

Abad, P., Castagnone-Sereno, P., Rosso, M., Engler, J.A., and Favery, B. 2009. Invasion, feeding and development, p. 163-181. In: R.N. Perry, M., M. and Starr, J. L. (eds.). Root-knot nematodes. CAB International, Wallingford, UK.

Akagi, T., Hanada, T., Yaegaki, H., Gradziel, T.M., and Tao, R. 2016. Genome-wide view of genetic diversity reveals paths of selection and cultivar differentiation in peach domestication. DNA Res. 23:271-282.

Albuquerue, E.V., Carneiro, R.M., Costa, P.M., Gomes, A.C., Santos, M., Pereira, A., Nicole, M., Fernandez, D., and Grossi de Sa, M.F. 2010. Resistance to Meloidogyne incognita expresses a hypersensitive-like response in Coffea arabica. Eur. J. Plant Pathol. 127:365-373.

Andersen, R., Sherman, W., and Williamson, J. 2001. Low chill peach and nectarine cultivars from the University of Florida breeding program: 50 years of progress. Proc. Fla. State Hort. Soc. 114:33-36.

Aranzana, M.J., Carbo, J., and Arus, P. 2003. Microsatellite variability in peach [Prunus persica (L.) Batsch]: cultivar identification, marker mutation, pedigree inferences and population structure. Theor. Appl. Genet. 106:1341-1352.

Aranzana, M.J., Garcia-Mas, J., Carbo, J., and Arus, P. 2002. Development and variability analysis of microsatellite markers in peach. Plant Breed. 121:87-92.

Barbary, A., Palloix, A., Fazari, A., Marteu, N., Castagnone-Sereno, P., and Dijan- Caporalino, C. 2014. The plant genetic background affects the efficiency of the pepper major nematode resistance genes Me1 and Me3. Theor. Appl. Genet. 127:499-507.

Bartlem, D.G., Jones, M.G., and Hammes, U.Z. 2014. Vascularization and nutrient delivery at root-knot nematode feeding sites in host roots. J. Exp. Bot. 65:1789- 1798.

Basile, B., Solari, L.I., and Dejong, T.M. 2007. Intra-canopy variability of fruit growth rate in peach trees grafted on rootstocks with different vigour-control capacity. J. Hort. Sci. Biotechnol. 82:243-256.

Beckman, T.G. and Chaparro, J.X. 2015. Peach rootstock development for the Southeastern United States. Acta Hort. 1084:171-178.

Beckman, T.G., Chaparro, J.X., and Sherman, W.B. 2012. 'MP-29', a clonal interspecific hybrid rootstock for peach. HortScience 47:128-131.

Beckman, T.G., Okie, W.R., and Meyers, S.C. 1992. Rootstocks affect bloom date and fruit maturation of 'Redhaven' peach. J. Amer. Soc. Hort. Sci. 117:377-379.

158

Beckman, T.G., Okie, W.R., Nyczepir, A.P., Pusey, P.L., and Reilly, C.C. 1998. Relative susceptibility of peach and plum germplasm to Armillaria root rot. HortScience 33:1062-1065.

Beckman, T.G. and Pusey, P.L. 2001. Field testing peach rootstocks for resistance to Armillaria root rot. HortScience 36:101-103.

Bellafiore, S., Jougla, C., Chapuis, E., Besnard, G., Suong, M., Vu, P.N., De Waele, D., Gantet, P., and Thi, X.N. 2015. Intraspecific variability of the facultative meiotic parthenogenetic root-knot nematode (Meloidogyne graminicola) from rice fields in Vietnam. C. R. Biologies 338:471-483.

Ben Mimoun, M. and DeJong, T. 1999. Using the relation between growing degree hours and harvest date to estimate run-times for PEACH: a tree growth and yield simulation model. Acta Hort. 499:107-114.

Ben Yahmed, J., Ghrab, M., Moreno, M.A., Pinochet, J., and Ben Mimoun, M. 2016. Performance of 'Subirana' flat peach cultivar budded on different Prunus rootstocks in a warm production area in North Africa. Sci. Hort. 206:24-32.

Bliss, F.A., Arulsekar, S., Foolad, M.R., Becerra, V., Gillen, A.M., Warburton, M.L., Dandekar, A.M., Kocsisne, G.M., and Mydin, K.K. 2002. An expanded genetic linkage map of Prunus based on an interspecific cross between almond and peach. Genome 45:520-529.

Brito, J.A., Kaur, R., Cetintas, R., Stanley, J.D., Mendes, M.L., McAvoy, E.J., Powers, T.O., and Dickson, D.W. 2008. Identification and isozyme characterisation of Meloidogyne spp. infecting horticultural and agronomic crops, and weed plants in Florida. Nematology 10:757-766.

Brito, J.A., Kaur, R., Cetintas, R., Stanley, J.D., Mendes, M.L., Powers, T.O., and Dickson, D.W. 2010. Meloidogyne spp. infecting ornamental plants in Florida. Nematropica 40:87-103.

Brito, J.A., Vau, S.J.S., and Dickson, D.W. 2015. Reproduction of Meloidogyne floridensis on selected root-knot nematode resistant plant species. J. Nematol. 47:226 (abstr.).

Bus, V.G.M., Esmenjaud, D., Buck, E., and Laurens, F. 2009. Application of genetic markers in rosaceous crops, p. 563-599. In: Folta, K. M. and Gardiner, S. E. (eds.). Genetics and genomics of rosaceae. Springer, New York, NY.

Bussi, C., Besset, J., and Girard, T. 2002. Effects of peach or hybrid rootstocks on growth and cropping of two cultivars of peach trees (Emeraude and Zephyr). Fruits 57:249-255.

Bussi, C., Huguet, J.G., Besset, J., and Girard, T. 1995. Rootstock effects on the growth and fruit yield of peach. Eur. J. Agron. 4:387-393.

159

Cabasan, M., Kumar, A., and De Waele, D. 2014. Comparison of migration, penetration, development and reproduction of Meloidogyne graminicola on susceptible and resistant rice genotypes. Nematology 14:405-415.

Cantini, C., Iezzoni, A.F., Lamboy, W.F., Boritzki, M., and Struss, D. 2001. DNA fingerprinting of tetraploid cherry germplasm using simple sequence repeats. J. Amer. Soc. Hort. Sci. 126:205-209.

Cao, K., Wang, L.R., Zhao, P., Zhu, G.R., Fang, W.C., Chen, C.W., and Wang, X.W. 2014a. Identification of a candidate gene for resistance to root-knot nematode in a wild peach and screening of its polymorphisms. Plant Breed. 133:530-535.

Cao, K., Wang, L.R., Zhu, G.R., Fang, W.C., Chen, C.W., and Zhao, P. 2011. Construction of a linkage map and identification of resistance gene analog markers for root-knot nematodes in wild peach, Prunus kansuensis. J. Amer. Soc. Hort. Sci. 136:190-197.

Cao, K., Zheng, Z.J., Wang, L.R., Liu, X., Zhu, G.R., Fang, W.C., Cheng, S.F., Zeng, P., Chen, C.W., Wang, X.W., Xie, M., Zhong, X., Wang, X.L., Zhao, P., Bian, C., Zhu, Y.L., Zhang, J.H., Ma, G.S., Chen, C.X., Li, Y.J., Hao, F.G., Li, Y., Huang, G.D., Li, Y.X., Li, H.Y., Guo, J., Xu, X., and Wang, J. 2014b. Comparative population genomics reveals the domestication history of the peach, Prunus persica, and human influences on perennial fruit crops. Genome Biol. 15:415. doi: 10.1186/s13059-014-0415-1.

Caruso, T., Inglese, P., Sidari, M., and Sottile, F. 1997. Rootstock influences seasonal dry matter and carbohydrate content and partitioning in above-ground components of 'Flordaprince' peach trees. J. Amer. Soc. Hort. Sci. 122:673-679.

Castagnone-Sereno, P. 2002. Genetic variability of nematodes: a threat to the durability of plant resistance genes? Euphytica 124:193-199.

Castagnone-Sereno, P. 2006. Genetic variability and adaptive evolution in parthenogenetic root-knot nematodes. Heredity 96:282-289.

Castagnone-Sereno, P., Danchin, E.G.J., Perfus-Barbeoch, L., and Abad, P. 2013. Diversity and evolution of root-knot nematodes, genus Meloidogyne: new insights from the genomic era. Annu. Rev. Phytopathol. 51:203-220.

Castagnone-Sereno, P., Wajnberg, E., Bongiovanni, M., Leroy, F., and Dalmasso, A. 1994. Genetic variation in Meloidogyne incognita virulence against the tomato Mi resistance gene - evidence from isofemale line selection studies. Theor. Appl. Genet. 88:749-753.

Chaparro, J., Carillo, O., Sherman, W., and Beckman, T. 2009. Evaluation of Prunus kansuensis as a genetic tester for peach. Acta Hort. 962:151-154.

160

Chaparro, J.X. and Sherman, W.B. 2006. 'UFSharp' peach. J. Am. Pomol. Soc. 60:95- 96.

Chaparro, J.X., Werner, D.J., Omalley, D., and Sederoff, R.R. 1994. Targeted mapping and linkage analysis of morphological isozyme, and RAPD markers in peach. Theor. Appl. Genet. 87:805-815.

Charcosset, A. and Moreau, L. 2004. Use of molecular markers for the development of new cultivars and the evaluation of genetic diversity. Euphytica 137:81-94.

Chavez, D.J., Beckman, T.G., Werner, D.J., and Chaparro, J.X. 2014. Genetic diversity in peach [Prunus persica (L.) Batsch] at the University of Florida: past, present and future. Tree Genet. Genomes 10:1399-1417.

Chavez, D.J. and Chaparro, J.X. 2011. Identification of markers linked to seedlessness in Citrus kinokuni hort. ex Tanaka and its progeny using bulked segregant analysis. HortScience 46:693-697.

Chu, Y., Wu, C.L., Holbrook, C.C., Tillman, B.L., Person, G., and Ozias-Akins, P. 2011. Marker-assisted selection to pyramid nematode resistance and the high oleic trait in peanut. Plant Genome 4:110-117.

Church, G.T. 2005. First report of the root-knot nematode Meloidogyne floridensis on tomato (Lycopersicon esculentum) in Florida. Plant Dis. 89:527 (abstr.).

Cipriani, G., Lot, G., Huang, W.G., Marrazzo, T.M., Peterlunger, E., and Testolin, R. 1999. AC/GT and AG/CT microsatellite repeats in peach [Prunus persica (L) Batsch]: isolation, characterisation and cross-species amplification in Prunus. Theor. Appl. Genet. 99:65-72.

Clarke, L.A., Rebelo, C.S., Gonçalves, J., Boavida, M.G., and Jordan, P. 2001. PCR amplification introduces errors into mononucleotide and dinucleotide repeat sequences. J. Clin. Pathol.: Mol. Pathol. 54:351-353.

Claverie, M., Bosselut, N., Lecouls, A.C., Voisin, R., Lafargue, B., Poizat, C., Kleinhentz, M., Laigret, F., Dirlewanger, E., and Esmenjaud, D. 2004a. Location of independent root-knot nematode resistance genes in plum and peach. Theor. Appl. Genet. 108:765-773.

Claverie, M., Dirlewanger, E., Bosselut, N., Van Ghelder, C., Voisin, R., Kleinhentz, M., Lafargue, B., Abad, P., Rosso, M.N., Chalhoub, B., and Esmenjaud, D. 2011. The Ma Gene for complete-spectrum resistance to Meloidogyne species in Prunus is a TNL with a huge repeated C-terminal post-LRR region. Plant Physiol. 156:779-792.

Claverie, M., Dirlewanger, E., Cosson, P., Bosselut, N., Lecouls, A.C., Voisin, R., Kleinhentz, M., Lafargue, B., Caboche, M., Chalhoub, B., and Esmenjaud, D. 2004b. High-resolution mapping and chromosome landing at the root-knot

161

nematode resistance locus Ma from Myrobalan plum using a large-insert BAC DNA library. Theor. Appl. Genet. 109:1318-1327.

Concibido, V., Diers, B., and Arelli, R. 2004. A decade of QTL mapping for cyst nematode resistance in soybean. Crop Sci. 44:1121-1131.

Dalmasso, A. and Berge, J.B. 1978. Molecular polymorphism and phylogenetic relationship in some Meloidogyne spp.: application to the taxonomy of Meloidogyne. J. Nematol. 10:323-332.

Dangl, J.L. and Jones, J.D.G. 2001. Plant pathogens and integrated defence responses to infection. Nature 411:826-833.

Davies, L.J. and Elling, A.A. 2015. Resistance genes against plant-parasitic nematodes: a durable control strategy? Nematology 17:249-263.

Davis, E.L., Hussey, R.S., Mitchum, M.G., and Baum, T.J. 2008. Parasitism proteins in nematode–plant interactions. Curr. Opin. Plant Biol. 11:360-366.

Day, L.H. 1953. Observations and experiments with plum, peach, apricot, and almond roots for stone fruits. Calif. Agr. Expt. Sta. Bul. 736.

DeWoody, J.A., Nason, J.D., and Hipkins, V.D. 2006. Mitigating scoring errors in microsatellite data from wild populations. Mol. Ecol. Notes 6:951-957.

Dirlewanger, E., Cosson, P., Boudehri, K., Renaud, C., Capdeville, G., Tauzin, Y., Laigret, F., and Moing, A. 2007. Development of a second-generation genetic linkage map for peach [Prunus persica (L.) Batsch] and characterization of morphological traits affecting flower and fruit. Tree Genet. Genomes 3:1-13.

Dirlewanger, E., Cosson, P., Howad, W., Capdeville, G., Bosselut, N., Claverie, M., Voisin, R., Poizat, C., Lafargue, B., Baron, O., Laigret, F., Kleinhentz, M., Arus, P., and Esmenjaud, D. 2004a. Microsatellite genetic linkage maps of myrobalan plum and an almond-peach hybrid - location of root-knot nematode resistance genes. Theor. Appl. Genet. 109:827-838.

Dirlewanger, E., Cosson, P., Tavaud, M., Aranzana, M.J., Poizat, C., Zanetto, A., Arus, P., and Laigret, F. 2002. Development of microsatellite markers in peach [Prunus persica (L.) Batsch] and their use in genetic diversity analysis in peach and sweet cherry (Prunus avium L.). Theor. Appl. Genet. 105:127-138.

Dirlewanger, E., Graziano, E., Joobeur, T., Garriga-Calder, F., Cosson, P., Howad, W., and Arus, P. 2004b. Comparative mapping and marker-assisted selection in Rosaceae fruit crops. PNAS 101:9891-9896.

Durner, E. and Goffreda, J. 1992. Rootstock-induced differences in flower bud phenology in peach. J. Amer. Soc. Hort. Sci. 117:690-697.

162

Duval, H., Hoerter, M., Polidori, J., Confolent, C., Masse, M., Moretti, A., Van Ghelder, C., and Esmenjaud, D. 2014. High-resolution mapping of the RMia gene for resistance to root-knot nematodes in peach. Tree Genet. Genomes 10:297-306.

Eddaoudi, M., Ammati, M., and Rammah, A. 1997. Identification of the resistance breaking populations of Meloidogyne on tomatoes in Morocco and their effect on new sources of resistance. Fundam. Appl. Nematol. 20:285-289.

Engler, J.A. and Gheysen, G. 2013. Nematode-induced endoreduplication in plant host cells: why and how? Mol. Plant Microbe Interact. 26:17-24.

Esmenjaud, D. 2009. Resistance to root-knot nematodes in Prunus: characterization of sources, marker-assisted selection and cloning strategy for the Ma gene from Myrobalan plum. Acta Hort. 814:707-714.

Esmenjaud, D., Lamassese, C.S., Salesses, G., Minot, J.C., and Voisin, R. 1992. Method and criteria to evaluate resistance to Meloigodyne arenaria in Prunus cerasifera Ehr. Fundam. Appl. Nematol. 15:385-389.

Esmenjaud, D., Minot, J.C., and Voisin, R. 1996a. Effects of durable inoculum pressure and high temperature on root galling, nematode numbers and survival of Myrobalan plum genotypes (Prunus cerasifera Ehr) highly resistant to Meloidogyne spp. Fundam. Appl. Nematol. 19:85-90.

Esmenjaud, D., Minot, J.C., Voisin, R., Bonnet, A., and Salesses, G. 1996b. Inheritance of resistance to the root-knot nematode Meloidogyne arenaria in Myrobalan plum. Theor. Appl. Genet. 92:873-879.

Esmenjaud, D., Minot, J.C., Voisin, R., Pinochet, J., and Salesses, G. 1994. Inter- and intraspecific resistance variability in Myrobalan plum, peach, and peach-almond rootstock using 22 root-knot nematode populations. J. Amer. Soc. Hort. Sci. 119:94-100.

Esmenjaud, D., Minot, J.C., Voisin, R., Pinochet, J., Simard, M.H., and Salesses, G. 1997. Differential response to root-knot nematodes in Prunus species and correlative genetic implications. J. Nematol. 29:370-380.

Esmenjaud, D., Voisin, R., Van Ghelder, C., Bosselut, N., Lafargue, B., Di Vito, M., Dirlewanger, E., Poessel, J.L., and Kleinhentz, M. 2009. Genetic dissection of resistance to root-knot nematodes Meloidogyne spp. in plum, peach, almond, and apricot from various segregating interspecific Prunus progenies. Tree Genet. Genomes 5:279-289.

Fan, S., Bielenberg, D.G., Zhebentyayeva, T.N., Reighard, G.L., Okie, W.R., Holland, D., and Abbott, A.G. 2010. Mapping quantitative trait loci associated with chilling requirement, heat requirement and bloom date in peach (Prunus persica). New Phytol. 185:917-930.

163

Felipe, A.J. 2009. 'Felinem', 'Garnem', and 'Monegro' almond x peach hybrid rootstocks. HortScience 44:196-197.

Fernandez, C., Pinochet, J., Esmenjaud, D., Gravato-Nobre, M.J., and Felipe, A. 1995. Age of plant material influences resistance of some Prunus rootstocks to Meloidogyne incognita. HortScience 30:582-585.

Fernandez, C., Pinochet, J., Esmenjaud, D., Salesses, G., and Felipe, A. 1994. Resistance among new Prunus rootstocks and selections to root-knot nematodes in Spain and France. HortScience 29:1064-1067.

Flint-Garcia, S.A., Thornsberry, J.M., and Buckler, E.S. 2003. Structure of linkage disequilibrium in plants. Annu. Rev. Plant Biol. 54:357-374.

Font i Forcada, C., Gogorcena, Y., and Moreno, M. 2012. Agronomical and fruit quality traits of two peach cultivars on peach-almond hybrid rootstocks growing on Mediterranean conditions. Sci. Hort. 140:157-163.

Font i Forcada, C., Gogorcena, Y., and Moreno, M.A. 2013. Fruit sugar profile and antioxidants of peach and nectarine cultivars on almond×peach hybrid rootstocks. Sci. Hort. 164:563-572.

Foulongne, M., Pascal, T., Arus, P., and Kervella, J. 2003. The potential of Prunus davidiana for introgression into peach [Prunus persica (L.) Batsch] assessed by comparative mapping. Theor. Appl. Genet. 107:227-238.

Florida State Horticultural Society (FSHS). 1888. Proc. I-IV Annu. Mtg., 1888-1891. (Proc. Fla. State Hort. Soc. 1-4).

Gillen, A.M. and Bliss, F.A. 2005. Identification and mapping of markers linked to the Mi gene for root-knot nematode resistance in peach. J. Amer. Soc. Hort. Sci. 130:24-33.

Giorgi, M., Capocasa, F., Scalzo, J., Murri, G., Battino, M., and Mezzetti, B. 2005. The rootstock effects on plant adaptability, production, fruit quality, and nutrition in the peach (cv. 'Suncrest'). Sci. Hort. 107:36-42.

Grossman, Y. and DeJong, T. 1995. Maximum fruit growth potential following resource limitation during peach growth. Ann. Bot. 75:561-567.

Grossman, Y. and DeJong, T. 1998. Training and pruning system effects on vegetative growth potential, light interception, and cropping efficiency in peach trees. J. Amer. Soc. Hort. Sci. 123:1058-1064.

Grundler, F.M.W. and Hofmann, J. 2011. Water and nutrient transport in nematode feeding sites, p. 423-439. In: Jones, J., Gheysen, G., and Fenoll, C. (eds.). Genomics and molecular genetics of plant-nematode interactions. Springer Netherlands, Dordrecht.

164

Handoo, Z.A., Nyczepir, A.P., Esmenjaud, D., van der Beek, J.G., Castagnone-Sereno, P., Carta, L.K., Skantar, A.M., and Higgins, J.A. 2004. Morphological, molecular, and differential-host characterization of Meloidogyne floridensis n. sp (Nematoda : Meloidogynidae), a root-knot nematode parasitizing peach in Florida. J. Nematol. 36:20-35.

Hansen, C.J., Lownsbery, B.F., and Hesse, C.O. 1956. Nematode resistance in peaches. Calif. Agric. 10(9):5-11.

Hewezi, T. and Baum, T.J. 2013. Manipulation of plant cells by cyst and root-knot nematode effectors. Mol. Plant Microbe Interact. 26:9-16.

Horton, D., Brannen, P., Bellinger, B., Lockwood, D., and Ritchie, D. 2015. Southeastern peach, nectarine, and plum pest management and culture guide. Bulletin 171. University of Georgia, Athens, GA.

Hospital, F. 2005. Selection in backcross programmes. Phil. Trans. R. Soc. B 360:1503- 1511.

Husey, R.S. 1989. Disease-inducing secretions of plant-parasitic nematodes. Annu. Rev. Phytopathol. 27:123-141.

Hussey, R.S. and Barker, K.R. 1973. Comparison of methods of collecting inocula of Meloidogyne spp, including a new technique. Plant Dis. Rep. 57:1025-1028.

Janssen, G.J.W., Scholten, O.E., van Norel, A., and Hoogendoorn, C. 1998. Selection of virulence in Meloidogyne chitwoodi to resistance in the wild potato Solanum fendleri. Eur. J. Plant Pathol. 104:645-651.

Janssen, G.J.W., Van Norel, A., Verkerk-Bakker, B., and Janssen, R. 1997. Intra- and interspecific variation of root-knot nematodes, Meloidogyne spp., with regard to resistance in wild tuber-bearing Solanum species. Fundam. Appl. Nematol. 20:449-457.

Jarquin-Barbarena, H., Dalmasso, A., de Guiran, G., and Cardin, M. 1991. Acquired virulence in the plant parasitic nematode Meloidogyne incognita. 1. Biological analysis of the phenomenon. Revue Nematol. 14:299-303.

Jenkins, J.N., McCarty, J.C., Wubben, M.J., Hayes, R., Gutierrez, O.A., Callahan, F., and Deng, D. 2012. SSR markers for marker assisted selection of root-knot nematode (Meloidogyne incognita) resistant plants in cotton (Gossypium hirsutum L). Euphytica 183:49-54.

Joobeur, T., Viruel, M.A., de Vicente, M.C., Jauregui, B., Ballester, J., Dettori, M.T., Verde, I., Truco, M.J., Messeguer, R., Batlle, I., Quarta, R., Dirlewanger, E., and Arus, P. 1998. Construction of a saturated linkage map for Prunus using an almond x peach F2 progeny. Theor. Appl. Genet. 97:1034-1041.

165

Kaloshian, I., Williamson, V.M., Miyao, G., Lawn, D., and Westerdahl, B. 1996. Resistance-breaking nematodes identified in California tomatoes. Calif. Agric. 50(6):18-19.

Kandoth, P.K. and Mitchum, M.G. 2013. War of the worms: how plants fight underground attacks. Curr. Opin. Plant Biol. 16:457-463.

Kao, T. and Huang, S. 1994. Gametophytic self-incompatibility: a mechanism for self/nonself discrimination during sexual reproduction. Plant Physiol. 105:461- 466.

Kaur, R., Brito, J.A., and Rich, J.R. 2007. Host suitability of selected weed species to five Meloidogyne species. Nematropica 37:107-120.

Kenealy, L., Reighard, G., Rauh, B., and Bridges, W. 2015. Predicting peach maturity dates in South Carolina with a growing degree day model. Acta Hort. 1084:749- 752.

Khallouk, S., Voisin, R., Portier, U., Polidori, J., Van Ghelder, C., and Esmenjaud, D. 2013. Multiyear evaluation of the durability of the resistance conferred by Ma and RMia genes to Meloidogyne incognita in Prunus under controlled conditions. Phytopathology 103:833-840.

Khallouk, S., Voisin, R., Van Ghelder, C., Engler, G., Amiri, S., and Esmenjaud, D. 2011. Histological mechanisms of the resistance conferred by the Ma gene against Meloidogyne incognita in Prunus spp. Phytopathology 101:945-951.

Kochba, J. and Spiegel-Roy, P. 1975. Inheritance of resistance to the root-knot nematode (Meloidogyne javanica Chitwood) in bitter almond progenies. Euphytica 24:453-457.

Kokalis-Burelle, N. and Nyczepir, A.P. 2004. Host-range studies for Meloidogyne floridensis. J. Nematol. 36:328 (abstr.).

Kyndt, T., Vieira, P., Gheysen, G., and Engler, G. 2013. Nematode feeding sites: unique organs in plant roots. Planta 238:807-818.

Lalli, D.A., Decroocq, V., Blenda, A.V., Schurdi-Levraud, V., Garay, L., Le Gall, O., Damsteegt, V., Reighard, G.L., and Abbott, A.G. 2005. Identification and mapping of resistance gene analogs (RGAs) in Prunus: a resistance map for Prunus. Theor. Appl. Genet. 111:1504-1513.

Lecouls, A., Bergougnoux, V., Rubio-Cabetas, M., Bosselut, N., Voisin, R., Poessel, J., Faurobert, M., Bonnet, A., Salesses, G., Dirlewanger, E., and Esmenjaud, D. 2004. Marker-assisted selection for the wide-spectrum resistance to root-knot nematodes conferred by the Ma gene from Myrobalan plum (Prunus cersaifera) in interspecific Prunus material. Mol. Breed. 13:113-124.

166

Lecouls, A.C., Rubio-Cabetas, M.J., Minot, J.C., Voisin, R., Bonnet, A., Salesses, G., Dirlewanger, E., and Esmenjaud, D. 1999. RAPD and SCAR markers linked to the Ma1 root-knot nematode resistance gene in Myrobalan plum (Prunus cerasifera Ehr.). Theor. Appl. Genet. 99:328-335.

Lecouls, A.C., Salesses, G., Minot, J.C., Voisin, R., Bonnet, A., and Esmenjaud, D. 1997. Spectrum of the Ma genes for resistance to Meloidogyne spp. in Myrobalan plum. Theor. Appl. Genet. 95:1325-1334.

Lopez, G. and DeJong, T. 2008. Using growing degree hours accumulated thirty days after bloom to help growers predict difficult fruit sizing years. Acta Hort. 803:175- 180.

Lopez, G. and Dejong, T.M. 2007. Spring temperatures have a major effect on early stages of peach fruit growth. J. Hort. Sci. Biotechnol. 82:507-512.

Lownsbery, B.F., Serr, E.F., and Hansen, C.J. 1959. Deciduous fruit and nut trees. Calif. Agric. 13(9):19-20.

Lownsbery, B.F. and Thomson, I.J. 1959. Progress in nematology related to horticulture. Proc. Amer. Soc. Hort. Sci. 74:730-746.

Lu, Z.X., Reighard, G.L., Nyczepir, A.P., and Beckman, T.G. 2000a. Inocula and media affect root-knot nematode infection of peach seedling roots. J. Am. Pomol. Soc. 54:76-81.

Lu, Z.X., Reighard, G.L., Nyczepir, A.P., Beckman, T.G., and Ramming, D.W. 2000b. Inheritance of resistance to root-knot nematodes (Meloidogyne sp.) in Prunus rootstocks. HortScience 35:1344-1346.

Lu, Z.X., Sosinski, B., Reighard, G.L., Baird, W.V., and Abbott, A.G. 1998. Construction of a genetic linkage map and identification of AFLP markers for resistance to root-knot nematodes in peach rootstocks. Genome 41:199-207.

Lu, Z.X., Sossey-Alaoui, K., Reighard, G.L., Baird, W.V., and Abbott, A.G. 1999. Development and characterization of a codominant marker linked to root-knot nematode resistance, and its application to peach rootstock breeding. Theor. Appl. Genet. 99:115-122.

Mackay, T.F.C., Stone, E.A., and Ayroles, J.F. 2009. The genetics of quantitative traits: challenges and prospects. Nat. Rev. Genet. 10:565-577.

Marone, D., Russo, M.A., Laido, G., Leonardis, A.M., and Mastrangelo, A.M. 2013. Plant nucleotide binding site-leucine-rich repeat (NBS-LRR) genes: active guardians in host defense responses. Int. J. Mol. Sci. 14:7302-7306.

167

Marra, F.P., Lo Bianco, R., La Mantia, M., and Caruso, T. 2013. Growth, yield and fruit quality of ‘Tropic Snow’ peach on size-controlling rootstocks under dry Mediterranean climates. Sci. Hort. 160:274-282.

Marull, J., Pinochet, J., Felipe, A., and Cenis, J.L. 1994. Resistance verification in Prunus selections to a mixture of thirteen Meloidogyne isolates and resistance mechanisms of a peach-almond hybrid to M. javanica. Fundam. Appl. Nematol. 17:85-92.

McKenry, M.V. 1989. Nematodes, p. 139-147. In: LaRue, J. H. and Johnson, R. S. (eds.). Peaches, plums, and nectarines: growing and handling for fresh market. Publication 3331. University of California Division of Agriculture and Natural Resources, Oakland, CA.

Miah, G., Rafii, M.Y., Ismail, M.R., Puteh, A.B., Rahim, H.A., Islam, K.N., and Latif, M.A. 2013. A review of microsatellite markers and their applications in rice breeding programs to improve blast disease resistance. Int. J. Mol. Sci. 14:22499-22528.

Milligan, S.B., Bodeau, J., Yaghoobi, J., Kaloshian, I., and Zabel, P. 1998. The root-knot nematode resistance gene Mi from tomato is a member of the leucine zipper, nucleotide-binding, leucine-rich repeat family of plant genes. The Plant Cell 10:1307-1319.

Mnejja, M., Garcia-Mas, J., Audergon, J.M., and Arus, P. 2010. Prunus microsatellite marker transferability across rosaceous crops. Tree Genet. Genomes 6:689-700.

Mnejja, M., Garcia-Mas, J., Howad, W., and Arus, P. 2005. Development and transportability across Prunus species of 42 polymorphic almond microsatellites. Mol. Ecol. Notes 5:531-535.

Mnejja, M., Garcia-Mas, M., Howad, W., Badenes, M.L., and Arus, P. 2004. Simple- sequence repeat (SSR) markers of Japanese plum (Prunus salicina Lindl.) are highly polymorphic and transferable to peach and almond. Mol. Ecol. Notes 4:163-166.

Moing, A. and Gaudillere, J.P. 1992. Carbon and nitrogen partitioning in peach/plum grafts. Tree Physiol. 10:81-92.

Moreno, M.A. 2004. Breeding and selection of Prunus rootstocks at the Aula Dei Experimental Station, Zaragoza, Spain. Acta Hort. 658:519-528.

Moreno, M.A., Moing, A., Lansac, M., Gaudillere, J.P., and Salesses, G. 1993. Peach Myrobalan plum graft incompatibility in the nursery. J. Hort. Sci. Biotechnol. 68:705-714.

Moreno, M.A., Tabuenca, M.C., and Cambra, R. 1994. Performance of Adafuel and Adarcias as peach rootstocks. HortScience 29:1271-1273.

168

Morgan, K. and Olmstead, M. 2013a. Peach orchard establishment and production planning budgets for Florida. Proc. Fla. State Hort. Soc. 126:35-42.

Morgan, K.L. and Olmstead, M.A. 2013b. A diversification strategy for perennial horticulture in Florida. HortTechnology 23:482-489.

Mota, F.C., Alves, G.C., Giband, M., Gomes, A.C., Sousa, F.R., Mattos, V.S., Barbosa, V.H., Barroso, P.A., Nicole, M., Peixoto, J.R., Rocha, M.R., and M., C.R. 2013. New sources of resistance to Meloidogyne incognita race 3 in wild cotton accessions and histological characterization of the defence mechanisms. Plant Pathol. 62:1173-1183.

Muniz, M.D.S., Campos, V.P., Castagnone-Sereno, P., Castro, J.M.D.E., Almeida, M.R.A., and Carneiro, R.M.D.G. 2008. Diversity of Meloidogyne exigua (Tylenchida: Meloidogynidae) populations from coffee and rubber tree. Nematology 10:897-910.

Nyczepir, A. 1991. Nematode management strategies in stone fruits in the United States. J. Nematol. 23:334-341.

Nyczepir, A.P. and Beckman, T.G. 2000. Host status of Guardian peach rootstock to Meloidogyne sp. and M. javanica. HortScience 35:772.

Nyczepir, A.P., Beckman, T.G., and Reighard, G.L. 1999. Reproduction and development of Meloidogyne incognita and M. javanica on Guardian peach rootstock. J. Nematol. 31:334-340.

Nyczepir, A.P., Beckman, T.G., and Reighard, G.L. 2006. Field evaluation of 'Guardian' peach rootstock to different root-knot nematode species. Acta Hort. 713:303-309.

Nyczepir, A.P., Bertrand, P.F., Miller, R.W., and Motsinger, R.E. 1985. Incidence of Criconemella spp. and peach orchard histories in short-life and non-short-life sites in Georgia and South Carolina. Plant Dis. 69:874-877.

Nyczepir, A.P., Brito, J.A., Dickson, D.W., and Beckman, T.G. 2008. Host status of selected peach rootstocks to Meloidogyne mayaguensis. HortScience 43:804- 806.

Nyczepir, A.P. and Esmenjaud, D. 2008. Nematodes. p. 505-535. In: Layne, D. and Bassi, D. (eds) The peach: botany, production and uses. CABI publishing, Wallingford, UK.

Okie, W.R., Beckman, T.G., Nyczepir, A.P., Reighard, G.L., Newall, W.C., and Zehr, E.I. 1994a. BY520-9, a peach rootstock for the Southeastern United States that increases scion longevity. HortScience 29:705-706.

169

Okie, W.R., Reighard, G.L., Beckman, T.G., Nyczepir, A.P., Reilly, C.C., Zehr, E.I., Newall, W.C., and Cain, D.W. 1994b. Field screening Prunus for longevity in the Southeastern United States. HortScience 29:673-677.

Olmstead, M., Chaparro, J., Andersen, R., Williamson, J., and Ferguson, J. 2016. Florida peach and nectarine varieties. Circ 1159. Gainesville: University of Florida Institute of Food and Agricultural Sciences. 21 August 2016. .

Olmstead, M., Chaparro, J., and Ferguson, J. 2015. Rootstocks for Florida stone fruit. HS1110. Gainesville: University of Florida Institute of Food and Agricultural Sciences. 5 January 2016. .

Petrillo, M.D., Matthews, W.C., and Roberts, P.A. 2006. Dynamics of Meloidogyne incognita virulence to resistance genes Rk and Rk2 in cowpea. J. Nematol. 38:90-96.

Pinochet, J., Calvet, C., Hernandez-Dorrego, A., Bonet, A., Felipe, A., and Moreno, M. 1999. Resistance of peach and plum rootstocks from Spain, France, and Italy to root-knot nematode Meloidogyne javanica. HortScience 34:1259-1262.

Quentin, M., Abad, P., and Favery, B. 2013. Plant parasitic nematode effectors target host defense and nuclear functions to establish feeding cells. Front. Plant Sci. 53:1-7.

Ramming, D.W. and Tanner, O. 1983. 'Nemared' peach rootstock. HortScience 18:376.

Rasmussen, G. 1978. How fresh peaches are marketed. Marketing Bul. No. 64. Agr. Marketing Serv., U. S. Dept. of Agr., Washington, D. C., August 1978.

Reighard, G. 2000. Peach rootstocks for the United States: are foreign rootstocks the answer? HortTechnology 10:714-718.

Reighard, G. and Loreti, F. 2008. Rootstock development. p. 193-220. In: Layne, D. and Bassi, D. (eds) The peach: botany, production and uses. CABI publishing, Wallingford, UK.

Reighard, G.L., Beckman, T., Belding, R., Black, B., Byers, P., Cline, J., Cowgill, W., Godin, R., Johnson, R.S., Kamas, J., Kaps, M., Larsen, H., Lindstrom, T., Newell, M., Ouellette, D., Pokharel, R., Stein, L., Taylor, K., Walsh, C., Ward, D., and Whiting, M. 2011. Six-year performance of 14 Prunus rootstocks at 11 sites in the 2001 NC-140 peach trial. J. Am. Pomol. Soc. 65:26-41.

Reilly, C.C., Nyczepir, A.P., Sharpe, R.R., Okie, W.R., and Pusey, P.L. 1986. Short life of peach trees as related to tree physiology, environment, pathogens, and cultural practices. Plant Dis. 70:538-541.

170

Rich, J.R., Brito, J., Kaur, R., and Ferrell, J.A. 2008. Weed species as hosts of Meloidogyne: a review. Nematropica 39:157-185.

Ritchie, D.F. and Clayton, C.N. 1981. Peach tree short life: a complex of interacting factors. Plant Dis. 65:462-469.

Roberts, P.A., Frate, C., Matthews, W.C., and Osterli, P. 1995. Interactions of virulent Meloidogyne incognita and Fusarium wilt on resistant cowpea genotypes. Phytopathology 85:1288-1295.

Rouse, R.E. and Sherman, W. 2004. 'UFSun' peach released for subtropical central and south Florida. Proc. Fla. State Hort. Soc. 117:239-241.

Rubio-Cabetas, M.J., Lecouls, A.C., Salesses, G., Bonnet, A., Minot, J.C., Voisin, R., and Esmenjaud, D. 1998. Evidence of a new gene for high resistance to Meloidogyne spp. in Myrobalan plum, Prunus cerasifera. Plant Breed. 117:567- 571.

Rubio-Cabetas, M.J., Minot, J.C., Voisin, R., Esmenjaud, D., Salesses, G., and Bonnet, A. 1999. Resistance response of the Ma genes from 'Myrobalan' plum to Meloidogyne hapla and M. mayaguensis. HortScience 34:1266-1268.

Salesses, G. and Bonnet, A. 1992. Some physiological and genetic aspects of peach/plum graft incompatibility. Acta Hort. 315:177-186.

Salesses, G., Dirlewanger, E., Esmenjaud, D., and Lecouls, A. 1998. Root-knot nematode resistance in Myrobalan plum: inheritance and rootstock breeding perspectives using marker-assisted selection. Acta Hort. 478:45-52.

Sasser, J.N., Carter, C.C., and Hartman, K.M. 1984. Standardization of host suitability studies and reporting of resistance to root-knot nematodes. N.C. State Univ. Dep. Plant Path., Raleigh, N.C., 7p.

Scorza, R. and Okie, W. 1991. Peaches (Prunus). Acta Hort. 290:177-234.

Sharpe, R. 1967. Root-knot nematode populations on peaches in Florida. Proc. Fla. State Hort. Soc. 80:342-344.

Sharpe, R., Hesse, C., Lownsbery, B., and Hansen, C. 1969. Breeding peaches for root-knot nematode resistance. J. Amer. Soc. Hort. Sci. 94:209-212.

Sharpe, R. and Sherman, W. 1990. Peach cultivars in Florida and their chilling requirements. Acta Hort. 279:191-197.

Sharpe, R.H. 1957. Okinawa peach shows promising resistance to root-knot nematodes. Fla. Agr. Expt. Sta. 657:320-322.

171

Sherman, W. and Lyrene, P. 2003. Low chill breeding of deciduous fruits at the University of Florida. Acta Hort. 622:599-605.

Sherman, W., Lyrene, P., and Sharpe, R. 1996. Low-chill peach and nectarine breeding at the University of Florida. Proc. Fla. State Hort. Soc. 109:222-223.

Sherman, W., Topp, B., and Lyrene, P. 1990. Non-melting flesh for fresh market peaches. Proc. Fla. State Hort. Soc. 103:293-294.

Sherman, W.B. and Lyrene, P.M. 1983. Improvement of peach rootstock resistant to root-knot nematodes. Proc. Fla. State Hort. Soc. 96:207-208.

Sherman, W.B., Lyrene, P.M., and Sharpe, R.H. 1991. Flordaguard peach rootstock. HortScience 26:427-428.

Smith, T., Brito, J.A., Han, H., Kaur, R., Cetintas, R., and Dickson, D.W. 2015. Identification of the peach root-knot nematode, Meloidogyne floridensis using mtDNA PCR-RFLP. Nematropica 45:138-143.

Sosinski, B., Gannavarapu, M., Hager, L.D., Beck, L.E., King, G.J., Ryder, C.D., Rajapakse, S., Baird, W.V., Ballard, R.E., and Abbott, A.G. 2000. Characterization of microsatellite markers in peach [Prunus persica (L.) Batsch]. Theor. Appl. Genet. 101:421-428.

Stanley, J.D., Brito, J.A., Kokalis-Burelle, N., Frank, J.H., and Dickson, D.W. 2009. Biological evaluation and comparison of four Florida isolates of Meloidogyne floridensis. Nematropica 39:255-271.

Struss, D., Boritzki, M., Karle, R., and Iezzoni, A.F. 2002. Microsatellite markers differentiate eight Giessen cherry rootstocks. HortScience 37:191-193.

Tao, R. and Iezzoni, A.F. 2010. The S-RNase-based gametophytic self-incompatibility system in Prunus exhibits distinct genetic and molecular features. Sci. Hort. 124:423-433.

Tao, R., Watari, A., Hanada, T., Habu, T., Yaegaki, H., Yamaguchi, M., and Yamane, H. 2007. Self-compatible peach (Prunus persica) has mutant versions of the S haplotypes found in self-incompatible Prunus species. Plant Mol. Biol. 63:109- 123.

Tautz, D. 1989. Hypervariability of simple sequences as a general source for polymorphic DNA markers. Nucleic Acids Res. 17:6463-6471.

Taylor, A.L. and Sasser, J.N. 1978. Biology, identification and control of root-knot nematodes (Meloidogyne species). N.C. State Univ. Dep. Plant Path. and USAID, Raleigh, N.C., p. 101.

172

Testolin, R., Marrazzo, T., Cipriani, G., Quarta, R., Verde, I., Dettori, M.T., Pancaldi, M., and Sansavini, S. 2002. Microsatellite DNA in peach (Prunus persica L. Batsch) and its use in fingerprinting and testing the genetic origin of cultivars. Genome 43:512-520.

Thies, J.A., Merrill, S.B., and Corley, E.L. 2002. Red food coloring stain: new, safer procedures for staining nematodes in roots and egg masses on root surfaces. J. Nematol. 34:179-181.

Thomas, B., Law, L., and Stankey, D. 1979. Soil survey of Marion County area, Florida. U.S. Department of Agriculture, Soil Conservation Service, Washington, D.C.

Topp, B. and Sherman, W. 1989. The relationship between temperature and bloom-to- ripening period in low-chill peach. Fruit Var. J. 43:155-158.

Trudgill, D.L. 1991. Resistance to and tolerance of plant parasitic nematodes in plants. Annu. Rev. Phytopathol. 29:167-192.

USDA. 2009. 2007 Census of agriculture, United States: summary and state data 1(51). 14 July 2016. .

Van der Beek, J.G., Poleij, L.M., Zijlstra, C., Janssen, R., and Janssen, G.J.W. 1998. Variation in virulence within Meloidogyne chitwoodi, M. fallax, and M. hapla on Solanum spp. Phytopathology 88:658-665.

Van Ghelder, C. and Esmenjaud, D. 2016. TNL genes in peach: insights into the post- LRR domain. BMC Genomics 17:317. doi: 10.1186/s12864-016-2635-0.

Van Ghelder, C., Lafargue, B., Dirlewanger, E., Ouassa, A., Voisin, R., Polidori, J., Kleinhentz, M., and Esmenjaud, D. 2010. Characterization of the RMja gene for resistance to root-knot nematodes in almond: spectrum, location, and interest for Prunus breeding. Tree Genet. Genomes 6:503-511.

Van Ooijen, J. 2006. JoinMap 4.1, Software for the calculation of genetic linkage maps in experimental populations of diploid species. Kyazma B. V., Wageningen, Netherlands.

Van Ooijen, J. 2009. MapQTL 6, Software for the mapping of quantitative trait loci in experimental populations of diploid species. Kyazma B. V., Wageningen, Netherlands.

Vendramin, E., Dettori, M.T., Giovinazzi, J., Micali, S., Quarta, R., and Verde, I. 2007. A set of EST-SSRs isolated from peach fruit transcriptome and their transportability across Prunus species. Mol. Ecol. Notes 7:307-310.

173

Verde, I., Abbott, A.G., Scalabrin, S., Jung, S., Shu, S., Marroni, F., Zhebentyayeva, T., Dettori, M.T., Grimwood, J., and Cattonaro, F. 2013. The high-quality draft genome of peach (Prunus persica) identifies unique patterns of genetic diversity, domestication and genome evolution. Nat. Genet. 45:487-496.

Voisin, R., Rubio-Cabetas, M.J., Minot, J.C., and Esmenjaud, D. 1999. Penetration, development and emigration of juveniles of the nematode Meloidogyne arenaria in Myrobalan plum (Prunus cerasifera) clones bearing the Ma resistance genes. Eur. J. Plant Pathol. 105:103-108.

Voorrips, R.E. 2002. MapChart: Software for the graphical presentation of linkage maps and QTLs. J. Hered. 93:77-78.

Weinberger, J.H., C., M.P., and Scott, D.H. 1943. Inheritance study of root-knot nematode resistance in peach varieties. Proc. Amer. Soc. Hort. Sci. 42:321-325.

Wert, T.W., Williamson, J.G., Chaparro, J.X., and Miller, E.P. 2007. The influence of climate on fruit shape of four low-chill peach cultivars. HortScience 42:1589- 1591.

Wert, T.W., Williamson, J.G., Chaparro, J.X., Miller, E.P., and Rouse, R.E. 2009. The influence of climate on fruit development and quality of four low-chill peach cultivars. HortScience 44:666-670.

Williamson, V.M. and Hussey, R.S. 1996. Nematode pathogenesis and resistance in plants. The Plant Cell 8:1735-1745.

Williamson, V.M. and Kumar, A. 2006. Nematode resistance in plants: the battle underground. Trends Genet. 22:396-403.

Williamson, V.M. and Roberts, P.A. 2009. Mechanism and genetics of resistance, p. 302-325. In: Perry, R. N., Moens, M., and Starr, J. (eds.). Root-knot nematodes. CAB International, Wallingford, UK.

Yamamoto, T. and Hayashi, T. 2002. New root-knot nematode resistance genes and their STS markers in peach. Sci. Hort. 96:81-90.

Yamamoto, T., Shimada, T., Imai, T., Yaegaki, H., Haji, T., Matsuta, N., Yamaguchi, M., and Hayashi, T. 2001. Characterization of morphological traits based on a genetic linkage map in peach. Breed. Sci. 51:271-278.

Yamamoto, T., Yamaguchi, M., and Hayashi, T. 2005. An integrated genetic linkage map of peach by SSR, STS, AFLP and RAPD. J. Japan Soc. Hort. Sci. 74:204- 213.

Young, M.J. and Sherman, W.B. 1977. Evaluation of peach rootstocks for root-knot nematode resistance. Proc. Fla. State Hort. Soc. 90:241-242.

174

Zarrouk, O., Gogorcena, Y., Moreno, M., and Pinochet, J. 2006. Graft compatibility between peach cultivars and Prunus rootstocks. HortScience 41:1389-1394.

Zhao, S., Yue, C., Luby, J., Gallardo, K., McCracken, V., McFerson, J., and Layne, D. 2017. U.S. peach producer preference and willingness to pay for fruit attributes. HortScience 52:116-121.

175

BIOGRAPHICAL SKETCH

Mary Ann Maquilan completed her bachelor’s degree in biology at the University of the Philippines in Mindanao. She forayed into the plant sciences domain with her undergraduate research on the conservation of endangered and indigenous plants through reproductive biology and in vitro propagation studies. Subsequently, she got involved in research projects focused on enhancing agricultural productivity in smallholder farms and her research interests naturally gravitated toward development of science-based solutions to address the needs of local growers. Under the auspices of the Fulbright-Philippine Agriculture Scholarship, she pursued a master’s degree in horticulture and agronomy at the University of California in Davis where she conducted research on somatic embryogenesis and genetic transformation to improve scion and rootstock genotypes of almond. After completion, she joined an interdisciplinary research team in the Philippines to develop bioengineered local varieties of papaya and eggplant for virus and insect resistance with the goal of making biotech crops available for small-scale farmers. At the University of Florida where she earned a Ph.D. in

Horticultural Sciences, she explored another research opportunity to broaden possibilities for crop improvement and experienced the challenging, yet exciting, intricacies of breeding perennial fruit crops. Her experience in the UF stone fruit breeding program was pivotal in her consideration of fruit breeding as a possible career.

176