 

.\
L.
I

My.

Em...
.. an“.

I
1.
.\ 19.3. a...)

. .z.....5.$..(tv:.

E...
(111
x: . ..

 

.: v

.1
:1. a1: 9

_
t:..

a...

2.3.5; . . ”mam... .17.. .,
.6. n. 5...: 3 :l. . .

a .I 79:...
1.! . .
1&{‘!:l ‘Qiuv |
. a... 4....
mun...» 3.”..an
. 11.1.
“wig. nun? .1 .1.

A.
.
§

J Jew. in
9.3. 2. .ﬂLﬂ ﬁahﬁ?
I}: 4 fl- . .I .HA’ . I

. £83.}?!-
340 III}...
. 5.1 .1: i
1:3 4.?) .l
..o vLIOa (u e9.}l..t... .15...»-
v 35.311251...» 3...?

Killialna EL:

5. .s.
.3. .

 

 

This is to certify that the
dissertation entitled

LINKAGE MAP CONSTRUCTION AND ANALYSIS OF FRUIT
SIZE IN SWEET (Pmnus avium L.) AND SOUR (Pmnus
cerasus L.) CHERRY

presented by

JAMES WINSTON OLMSTEAD

has been accepted towards fulﬁllment
of the requirements for the

Plant Breeding and Genetics -
Doctoral degree in Department of Horticulture

 

 

Major Profeésﬁ’r’s Signature
777% 3) £0 06

 

Date

MSU is an Afﬁrmative Action/Equal Opportunity Institution

 

LIBRARY
Michigan State
University

 

 

 

---—--—c-.o—-—-

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/CIRC/DateDue.indd-p.1

 

 

LINKAGE MAP CONSTRUCTION AND ANALYSIS OF FRUIT SIZE IN
SWEET (Prunus avium L.) AND SOUR (Prunus cerasus L.) CHERRY

By

James Winston Olmstead

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Plant Breeding and Genetics - Horticulture

2006

ABSTRACT

LINKAGE MAP CONSTRUCTION AND ANALYSIS OF FRUIT SIZE IN
SWEET (Prunus avium L.) AND SOUR (Prunus cerasus L.) CHERRY

By
James Winston Olmstead

Maximizing fruit size is critical for proﬁtable sweet (Prunus avium L.) and sour
(Prunus cerasus L.) cherry production, yet little is known about the genetic control of this
trait. Fruit size varies widely between cherry cultivars, and signiﬁcant variation exists
among genetically identical fruit due to environmental and cultural differences. A more
thorough understanding of the genetic control of fruit size may be used to design future
management and genetic improvement strategies to increase cherry fruit size.

This research examined the mesocarp cellular differences between ﬁve cultivars
representing a broad range of fruit size in sweet cherry. Both cell number and cell size
were signiﬁcantly different (P < 0.05) between cultivars. However, the relationship of
cell number with fruit weight and diameter was signiﬁcantly and positively correlated
while cell size was not correlated with either measure of fruit size. Cell number was
stable during the three years of this study and in two different locations. Differences in
cell number due to environmental variation were examined in fruit from three of the same
cultivars that were signiﬁcantly different (P < 0.001) in fruit size. In this case, fruit size
differences were attributed to a difference in cell size rather than cell number, conﬁrming
the identiﬁcation of cell number as the primary genetic component resulting in fruit size
differences between cultivars.

To study the genetic control of fruit size in cherry, linkage maps were constructed

for reciprocal crosses between the sweet cherry cultivars ‘NY 54’ and ‘EF’. The linkage

maps consist of 8 linkage groups (LG) for the ‘EF’ parent (479.1 cM) and 10 LG for the
‘NY 54’ parent (308.9 CM). The average distance between marker loci and largest gaps
are 7 cM and 29 cM for ‘EF’ and 8 cM and 34 cM for ‘NY 54’, respectively. Fourteen
of the sweet cherry linkage groups could be aligned with the reference Prunus map based
on shared SSR markers.

QTL analysis of fruit size traits was performed using the ‘NY 54’ X ‘EF
population. For mesocarp length, one QTL (m1eng1h1 ) was identiﬁed on ‘EF’ linkage
group 6 (LG 6) and one on ‘NY 54’ LG (y) (mlengch), explaining 18.3% and 37.4% of
the total phenotypic variance, respectively. Three QTL were identiﬁed for mesocarp cell
length, on ‘EF’ LG 6 (clengthl) and ‘NY 54’ LG 6 (clengch) and LG (y) (cleng1h3).
The QTL explained 17.4, 16.8, and 16.8% of the phenotypic variation, respectively.

A targeted mapping approach, using SSR loci previously mapped to LG 6 in other
Prunus species was used to develop a linkage map for the ‘UF’ >< ‘Sureﬁre’ sour cherry
population. A QTL three cM from the S locus explaining 26.4% of the phenotypic
variation was identiﬁed in the ‘UF’ X ‘Sureﬁre’ population. Additionally, a fruit shape
QTL was also located on LG 6, co-segregating with the CPSCT012 marker and

explaining up to 22.6% of the phenotypic variation for fruit shape.

To my wife Mercy Olmstead, and my parents Peter and Patricia Olmstead.

iv

ACKNOWLEDGMENTS

I wish to thank my major professor, Dr. Amy Iezzoni, and the members of my
dissertation committee, Dr. Greg Lang, Dr. Dechun Wang, and Dr. Kyung-Hwan Han,

for their guidance and support during this research project.

TABLE OF CONTENTS

LIST OF TABLES .................................................................................... ix
LIST OF FIGURES ................................................................................... xi
CHAPTER 1--—-INTRODUCTION AND REVIEW OF LITERATURE .................... 1
Introduction ................................................................................... 2
Importance of Fruit Size in Production ........................................ 2

Factors Inﬂuencing Fruit Size 4
Prunus Linkage Map Construction 7
9

Prunus Reference Map ...........................................................
QTL Analyses in Prunus ....................................................... 11
Importance of LG 6 ............................................................................... 12
Utility of Simple Sequence Repeat Markers ......................................... 14
Objectives ............................................................................................................ 15
Literature Cited ............................................................................. 18

CHAPTER 2----GENOTYPIC DIFFERENCES IN SWEET CHERRY (Prunus avium L.)
FRUIT SIZE ARE PRIMARILY A FUNCTION OF CELL

NUMBER .............................................................................................. 27

Introduction 30

Materials and Methods ................................................................... 32
Plant Material ........................................................................................ 32
Flower and Fruit Sampling Scheme for Mesocarp Cell Number and
Size Comparisons Among the Five Cultivars ....................................... 33
Flower Bud Thinning Treatments to Determine the Inﬂuence of
Crop Load on Mesocarp Cell Number and Size ................................... 34
Fruit Measurement and Sectioning ....................................................... 35
Microscopic Analysis of Mesocarp Tissue ................................. 35
Image Analysis .................................................................. 36
Statistical Analysis ............................................................ 37
Results ...................................................................................... 37
Comparison of Fruit and Pit Measurements Among Five Sweet Cherry
Selections .............................................................................................. 37
Stability of Mesocarp Cell Number and Size for Selections
Subjected to Different Climatic and Cultural Conditions ..................... 39
Duration of Mesocarp Cell Division for the Five Sweet Cherry
Selections ........................................................................ 42

Discussion 43
Conclusions 49
Literature Cited 51
Tables and Figures ........................................................................ 55

vi

CHAPTER 3---- CONSTRUCTION OF A GENETIC LINKAGE MAP FOR THE ‘NY
54’ x ‘EMPEROR FRANCIS’ SWEET CHERRY (Prunus avium L.)
POPULATION .............................................................................. 67

Introduction 69
Materials and Methods 71
Plant Material .................................................................. 71
DNA Isolation and Marker Analysis ....................................... 72
Chi Square Analysis and Linkage Map Construction ..................... 76
Results...... ........................................................................................................ 76
Marker Segregation ............................................................ 76
Linkage Map Construction ................................................... 78
Discussion ................................................................................. 79
Conclusions 85
Literature Cited 86
Tables and Figures ........................................................................ 92

CHAPTER 4----TARGETED MAPPING OF FRUIT SIZE AND SHAPE QTL IN SOUR
CHERRY (Prunus cerasus L.) ......................................................... 104
Introduction 107
Materials and Methods .................................................................. 110
Plant Material and Phenotypic Analysis ............................................. 110
DNA Isolation and Marker Analysis .................................................. 111
Single Marker Analysis of Variance ................................................... 113
Linkage Analysis and Map Construction ............................................ 113
QTL and Statistical Analysis ............................................... 114
Results .................................................................................... 114
Marker Analysis and LG 6 Construction ............................................ 114
Fruit Size and Shape ........................................................................... 116

Single Marker Analysis to Test the Association of Fruit Traits With the
SLocus ................................................................................................ 118
QTL Analysis ...................................................................................... 118
Discussion 119
Conclusions 123
Literature Cited 125
Tables and Figures ...................................................................... 130

CHAPTER 5----QTL ANALYSIS OF FRUIT SIZE TRAITS FOR THE ‘NY 54’ X

‘EMPEROR FRANCIS’ SWEET CHERRY (Prunus avium L.)
POPULATION ............................................................................ . 140
Abstract ................................................................................... 141
Introduction .............................................................................. 143
Materials and Methods .................................................................. 144
Plant Material ................................................................ 144
Linkage Map Construction ................................................. 145

vii

Phenotypic Analysis ......................................................... 145

QTL and Statistical Analysis .............................................. 146

Results and Discussion ............................................................... 146
Conclusions ............................................................................. 150
Literature Cited ........................................................................ 1 5 1

Tables and Figures ..................................................................... 154
CHAPTER 6----SUMMARY AND CONCLUSIONS ...................................... 160

viii

LIST OF TABLES

1. Comparison of mean whole fresh fruit size, pit size, and mesocarp cell number per
radial section and size measurements for ‘Selah’, ‘Emperor Francis’ (‘EF’), ‘NY 54’,
‘Bing’, and ‘Regina’ sweet cherries from WSU-IAREC at harvest
maturity ....................................................................................................................... 57

2. Comparison of mean cell numbers at maturity for ‘Selah’, ‘Emperor Francis’ (‘EF’),
‘NY 54’, ‘Bing’, and ‘Regina’ sweet cherry fruit from 2003-2005 and at two
locations (WSU-IAREC and MSU-CHES) ................................................................ 60

3. Comparison of mean whole fresh fruit size, pit size, and mesocarp cell number (per
radial section) and size measurements for populations of large and small fruit from
‘Bing’, ‘Regina’, and ‘Selah’ sweet cherries at harvest maturity ................................ 61

4. Comparison of mean cell numbers at bloom, start of endocarp hardening, and
maturity for ‘Selah’, ‘Emperor Francis’ (‘EF’), ‘NY 54’, ‘Bing’, and ‘Regina’ sweet
cherry fruit ................................................................................................................... 63

5. Comparison of the duration and rate of cell division between ‘Bing’, ‘Regina’, and
‘Selah’ fruit from the period between bloom and endocarp hardening at WSU-IAREC
in 2005. The rate of cell division was calculated by dividing the increase in the
number of cells from bloom by the total accumulation of growing degree days
[GDD (4.4 C base)] from the point when bloom for that cultivar occurred ................ 66

6. Origins of simple sequence repeat (S SR) markers used in the development of the ‘NY
54’ X ‘Emperor Francis’ sweet cherry genetic linkage map ....................................... 93

7. Enzymes used for digest, selective nucleotide combinations used as primers, number
of polymorphic fragments, and number of mapped fragments generated by ampliﬁed
fragment length polymorphism (AF LP) analysis in the development of the ‘NY 54’ X
‘Emperor Francis’ sweet cherry genetic linkage map .................................................. 95

8. Number and type of markers for ‘Emperor Francis’ (‘EF’) and ‘NY 54’ parental
maps, map length, target map length from the ‘T’ X ‘E’ Prunus reference map
(Dirlewanger et al., 2004a), marker density, marker gap length, average linkage group
length, and average number of markers per linkage group ....................................... 100

9. Number and type of marker, length, target length from the ‘T’ X ‘E’ Prunus reference
map (Dirlewanger et al., 2004a), marker density and marker gap length for ‘Emperor
Francis’ (‘EF’) and ‘NY 54’ parental linkage groups ............................................... 101

10. Comparison of reciprocal differences in segregation of distorted markers present in
‘Emperor Francis’ that map to LG 6 .......................................................................... 103

ix

11. Broad sense heritability (H2), mean phenotypic values and standard deviations, and
progeny value range for sour cherry fruit weight, diameter, and length/width
percentage .................................................................................................................. 133

12. Mean fruit weight, diameter, and length/width percentage for ‘Uj fehértoi F ﬁrtos’
(‘UF’) X ‘Sureﬁre’ sour cherry progeny from each possible S-allele group resulting
from disomic inheritance for the years 2002-2004 ................................................... 136

13. QTL detected for sour cherry fruit weight and length/width ratio. All detected QTL
were located on the ‘Uj fehértoi F ﬁrtos’ linkage group (LG) 6a, corresponding to
linkage group 6 of the ‘T’ X ‘E’ reference Prunus map ............................................ 137

14. Comparison of sour cherry mean fruit weight (2002-2004) and mesocarp cell
numbers (2004) for ‘Ujfehértéi Fﬁrtos’ (‘UF’), ‘Sureﬁre’, and small and large
progeny individuals ﬁom the ‘UF’ X ‘Sureﬁre’ population. Five fruit were measured
for each year and trait ................................................................................................ 139

15. Broad sense heritability (H2), mean phenotypic values and standard deviations, and
progeny value range for sweet cherry fruit mesocarp radial cell number, mesocarp
radial length, and mesocarp mean cell length at endocarp hardening ....................... 155

16. QTL detected for mesocarp radial length and mean cell length at endocarp hardening
in ‘NY 54’ and ‘Emperor Francis’ (‘EF’) for the year 2005 .................................... 156

LIST OF FIGURES

. Diagram illustrating the area of sweet cherry ﬁ'uit samples sectioned for microscopic
analysis. Radial sections were prepared from the thickest part of the fruit mesocarp,
halfway between the point of stem attachment and the stylar scar, and 90 degrees
from the suture line.. ... 55

. Images of fruit from ‘Selah’ (A), ‘Emperor Francis’ (B), and ‘NY 54’ (C) sweet
cherries, illustrating the variation in fruit size and mesocarp thickness variation ....... 56

. Examples of ‘Selah’ (A), ‘Bing’ (B), ‘Regina’ (C), ‘Emperor Francis’ (‘EF’) (D), and
‘NY54’ (E) composite mesocarp images at fruit maturity (20x). Images were created
by aligning adjoining microscope ﬁeld-width images (n = 19, 15, 13, 13, and 5 for
‘Selah’, ‘Bing’, ‘Regina’, ‘EF’, and ‘NY 54’, respectively), and scaled relative to
each other for presentation. Scale bar = 200 um ........................................................ 58

. Linear correlation between mesocarp cell number and both ﬁ’esh fruit diameter and
weight (A), and mesocarp cell length and both fresh ﬁ'uit diameter and weight (B) for
all sweet cherry cultivars measured in 2004. The P-values calculated for the Pearson
correlation coefﬁcient (r) and R-square value for each comparison is indicated on

the plot ......................................................................................................................... 59

. Examples of mesocarp cell development at different stages of fruit development for
‘NY 54’ sweet cherry (45x). (A) bloom, (B) endocarp hardening, and (C) harvest
maturity. Images from endocarp hardening and harvest are composite images created
by aligning adjoining microscope ﬁeld-width images (n = 3 and 5, respectively), and
scaled relative to each other for presentation. Scale bar = 200 um ............................ 62

. Comparison of mesocarp cell number increase for ‘NY 54’, ‘Emperor Francis’ (‘EF’),
‘Bing’, ‘Regina’, and ‘Selah’ sweet cherry fruit from WSU-IAREC from bloom to
endocarp hardening. Sampling in 2004 (A) was performed on a weekly basis, while
2005 (B) sampling used 1-2 day intervals. Growing degree day accumulation (4.4 C
base) from the beginning of the sampling period is indicated on the lower x-axis.
Analysis was discontinued once the measured number of cells equaled that of the
harvest sample. Due to late spring ﬁ'eeze in 2005, no ‘EF’ fruit were available for
sampling ....................................................................................................................... 64

. Comparison of mesocarp cell number increase between ‘NY 54’, ‘Emperor Francis’
(‘EF’), ‘Bing’ and ‘Regina’ sweet cherry fruit from MSU-CHES from bloom to
endocarp hardening. Sampling in 2004 (A) and 2005 (B) was performed at 3-5 day
intervals. Growing degree day accumulation (4.4 C base) from the beginning of the
sampling period is indicated on the lower x-axis. Analysis was discontinued once the
measured number of cells equaled that of the harvest sample ..................................... 65

xi

10.

ll.

12.

13.

14.

15.

(A) Image of mature fruit from ‘Emperor Francis’ (‘EF’) (leﬁ) and ‘NY 54’ (right)
sweet cherry illustrating size variation between the two cultivars. (B) Selected
progeny from the ‘EF’ X ‘NY 54’ sweet cherry linkage mapping population
illustrating variation for fruit characteristics ............................................................... 92

Co-dominant simple sequence repeat (S SR) fragments obtained with primers for the
CPDCT022 marker. From leﬁ to right: (1) 50 base pair sizing ladder, (2) 10 base pair
sizing ladder, (3) ‘NY 54’, (4) ‘Emperor Francis’, (5-24) progeny. Arrows indicate
segregating fragments .................................................................................................. 94

Ampliﬁed fragment length polymorphism (AF LP) fragments obtained with the
selective primers EcoRI+AC and Msel +CTA. From left to right: (1) 10 base pair
sizing ladder, (2) 50 base pair sizing ladder, (3) ‘NY 54’, (4) ‘Emperor Francis’, (5-
29) progeny. Arrows indicate segregating fragments ................................................. 96

‘Emperor Francis’ (‘EF’) and ‘NY 54’ sweet cherry genetic linkage maps. Distance
in cM is given to the left of each linkage group. Marker names are on the right, with
the scored fragment size after the dash. Ampliﬁed fragment length polymorphism
(AF LP) combinations are named using the EcoRI / Msel selective primers. Linkage
groups corresponding to the ‘T’ X ‘E’ reference Prunus linkage map are numbered;
letters designate groups not aligned with the reference map. Anchor loci between the
parental linkage groups are connected by a solid line. Distorted loci are indicated by
* and ** at P < 0.05 and P < 0.01, respectively .......................................................... 97

Selected progeny from the “Ujfehértoi Filrtos’ X ‘Sureﬁre’ sour cherry population
illustrating variation for fruit size and shape ............................................................. 130

Alignment of the mapped homeologous linkage groups from ‘Uj fehértoi F ﬁrtos’
(‘UF’) sour cherry corresponding to LG 6 from the ‘T’ X ‘E’ reference Prunus map.
Only SSR markers from the ‘T’ X ‘E’ map are shown. Map distances in cM are
indicated to the left and marker names to the right of each vertical bar. Distorted loci
at the level of 0.1% level are denoted with a * following the name. Anchor loci
between ‘UF’ and ‘T’ X ‘E’ are connected by a dotted line ...................................... 131

Frequency distribution of sour cherry fruit weight (A) diameter (B), and length/width
percentage (C) traits measured on 126 progeny in the ‘Ujfehértoi F ﬁrtos’ (‘UF’) X
‘Sureﬁre’ population. The distribution is based on the mean value of each genotype
in years 2002-2004. Means for the parents ‘UF’ and ‘Sureﬁre’ are shown

by arrows ................................................................................................................... 132

Linear correlation between sour cherry progeny mean fruit weight and fruit diameter
(A), fruit weight and length/width percentage (B), and fruit diameter and length/width
percentage (C). The Pearson correlation coefﬁcient and R-square values for each
comparison are indicated on the plot ......................................................................... 134

xii

16. Frequency distribution of crop load rating (A), and linear correlation between sour
cherry progeny mean fruit weight and crop load rating (B) for 2004. Means for the
parents ‘Ujfehe'rtoi Furtos’ (‘UF’) and ‘Sureﬁre’ are shown by arrows on the
histogram. Individual progeny crop load was rated on a 1-10 scale, 1 being the
lowest crop load. The P-value calculated for the Spearman correlation coefﬁcient
and R-square value is indicated on the plot ............................................................... 135

17. Signiﬁcant QTL for sour cherry fruit shape (A) and fruit weight (B) on linkage group
(LG) 6a of ‘Ujfehe'rtéi F ﬁrtos’ (‘UF’) for years 2002-2004. Curves on the graphs
represent results from different years and correspond to the legend. Vertical lines on
the graph indicate LOD signiﬁcance scores for each year based on 1000
permutations. Bars between the graph and linkage group indicate l-LOD (ﬁlled) and
2-LOD (bars) interval for the QTL peak on the graph ............................................... 138

18. Frequency distribution of fruit mesocarp radial cell number (A), cell length (B), and
mesocarp radial length (C) measured at endocarp hardening for 67 progeny in the
‘NY 54’ X ‘Emperor Francis’ (‘EF’) population in 2005. Means for the parents
‘NY 54’ and ‘EF’ are shown by arrows ..................................................................... 154

19. Signiﬁcant QTL for fruit mesocarp cell length (A) and mesocarp radial length (B) on
‘Emperor Francis’ (‘EF’) LG 6. Vertical lines on the graph indicate the LOD
signiﬁcance threshold based on 1000 permutations. Bars between the graph and
linkage group indicate l-LOD (ﬁlled) and 2-LOD (bars) interval for the QTL peak
on the graph ............................................................................................................... 157

20. Signiﬁcant QTL for fruit mesocarp cell length at endocarp hardening on ‘NY 54’ LG
6. Vertical lines on the graph indicate the LOD signiﬁcance threshold based on 1000
permutations. Bars between the graph and linkage group indicate l-LOD (ﬁlled) and
2-LOD (bars) interval for the QTL peak on the graph .............................................. 158

21. Signiﬁcant QTL for fruit mesocarp cell length (A) and mesocarp radial length (B) at
endocarp hardening on ‘NY 54’ LG (y). Vertical lines on the graph indicate the LOD
signiﬁcance threshold based on 1000 permutations. Bars between the graph and
linkage group indicate l-LOD (ﬁlled) and 2-LOD (bars) interval for the QTL peak

on the graph ................................................................................................................ 159

xiii

CHAPTER ONE

Introduction and Review of Literature

INTRODUCTION

Sweet (Prunus avium L.) and sour (P. cerasus L.) cherries are produced in most
agricultural regions of the world where temperate crops can be grown. The genus Prunus
contains many economically important tree fruit and nut crops, including peach [P.
persica (L.) Batsch], almond [P. dulcis (Miller) D.A. Webb], European plum (P.
domestica L.), Japanese plum (P. salicina Lindl.), and apricot (P. armeniaca L.).
Approximately 1,850,000 mt of cherries were produced worldwide in 2005 (F AOSTAT
data, 2005). The United States produced 250,000 mt, 13.5% of the world total. Four
states (Washington, Oregon, California, and Michigan) produce over 95% of US. sweet
cherries, while Michigan alone accounts for over 65% of the US. sour cherry production
(NASS-USDA, 2005). Like other members of Prunus, sweet and sour cherries are
classiﬁed anatomically as a drupe, originating from a single carpel (Esau, 1977). The
pericarp, enlarged ovary tissue, is composed of three tissue types; the endocarp,
mesocarp, and exocarp. The scleriﬁed endocarp (stone or pit) contains the single seed.
The mesocarp is ﬂeshy, consisting of multiple layers of highly vacuolated parenchyma
cells. The specialized cuticular cells comprising the exocarp (skin) are typically only a

few cell layers thick.

Importance of Fruit Size in Production

The early progenitors of modern sweet cherry cultivars probably had small fruit
similar in size to wild sweet cherry forest trees. These wild sweet cherry types,
collectively referred to as Mazzard, have very small fruit averaging 1-2 g in weight. As a

result of selection and domestication, the fruit of modern cultivars can be over 10 g. This

increase is particularly striking, considering that many of the cultivars grown today are
only a few generations removed from the landraces from which they were originally
selected (Iezzoni et al., 1990). Improvements in cultural practices have contributed to the
increase in fruit size. For example, the average size of ‘Bing’ fruit achieved in
commercial production has increased in recent years, although the cultivar has been ﬁxed
genetically by vegetative propagation since its introduction in the 18705. However, the
10x or greater increase in fruit size, compared to wild members of the species also has a
signiﬁcant genetic component. Early sweet cherry breeders recognized relatively
uniform distribution of fruit size among progeny from different crosses suggesting a
quantitatively inherited trait (Fogle, 1961; Hansche et al., 1966; Lamb, 1953; Matthews,
1973)

Sour cherry is tetraploid (2n = 4x = 32), whereas most cultivated Prunus are
predominantly diploid (2n = 2x = 16). Sour cherry is believed to have arisen multiple
times through natural hybridization between ground cherry (P. ﬁuticosa Pall.; 2n = 4x =
32) and unreduced gametes from sweet cherry (2n = 2x = 16) (Beaver and Iezzoni, 1993;
Brettin et al., 2000; Olden and Nybom, 1968). Prunusfruticosa has small fruit; however,
some larger ﬁ'uited selections have been bred in Russia (Iezzoni et al., 1990). It is not
known if the origin of today’s sour cherry cultivars occurred through hybridization with
large or small-fruited sweet cherries. However, it is likely that certain landrace varieties
are the result of recent backcrossing with sweet cherry. Currently, the sour cherry
industry in the United States is based almost entirely on one genotype, ‘Montrnorency’, a
400-year-old selection from France that averages 4-6 g in fruit weight (Iezzoni, 1988,

2005).

Large fruit size is an essential component of proﬁtable fresh market sweet cherry
production. Currently, ﬁ'uit size is the primary criterion by which fresh cherries are
graded for sale, with ﬁ'uit averaging over 29 mm in diameter worth nearly twice as much
($/kg) as fruit less than 24 mm in diameter (Whiting et al., 2005, 2006). Therefore, sweet
cherry breeding efforts have long focused on the development of cultivars with larger
fruit. The fruit quality of many sour cherry cultivars is superior to ‘Montrnorency’, and
there has been increased producer interest in, and consumer acceptance of sour cherries
marketed for fresh consumption (Lang et al., 2003). As this market develops, a premium
will likely be placed on large sour cherries. Therefore, future breeding efforts may be
directed toward selection of larger sour cherry varieties for fresh market production.
However, little effort has been directed toward understanding the inﬂuence of cell

number and size on fruit development and ﬁnal size in cherry.

Factors Inﬂuencing Fruit Size

Various methods of quantifying fruit size have been employed in past research.
Fruit weight, length, diameter, and volume are all relatively simple measures of size.
However, the relationship between these measurements and cellular components of fruit
size is not clear. This is of particular importance in relation to quantitative trait loci
(QTL) identiﬁcation, where the reliability of phenotypic data is of utmost importance to
genetic analyses.

Fruits of Prunus, including cherry, generally exhibit a characteristic double
sigmoid growth curve, consisting of three developmental stages (Lilleland and Newsome,

1934). Stage I is characterized by rapid and exponential ﬁ'uit size increase, stage II by a

lag period of size increase coinciding with endocarp hardening, and stage III by a second
period of exponential size increase ending with harvest. These stages have been assigned
arbitrarily based on external link size increase measurements. However, the delineation
between the developmental phases is not distinct and does not necessarily coincide with
physiological development (Chalmers and van den Ende, 1973; Coombe, 1976; DeJong
and Goudriaan, 1989; Gage and Stutte, 1991).

Although the deﬁned developmental stages do not always correspond with
physiological development, anatomical and morphological changes in the ﬁ'uit are
thought to follow the general pattern of three stages of development (Coombe, 1976;
Gage and Stutte, 1991; Jackson and Coombe, 1966; Ragland, 1934; Tukey and Young,
1939; Yamaguchi et al., 2002b). Stage I, from anthesis to the beginning of endocarp
scleriﬁcation, is a period of rapid cell division and initial cell enlargement. By stage 11,
cell division has generally ceased, cell enlargement slows considerably, and endocarp
scleriﬁcation occurs, as well as a general thickening of parenchyma cell walls throughout
the mesocarp. Stage III is characterized by renewed cell enlargement, either radially or
tangentially, depending on cell location in relation to the endocarp or exocarp. Cell
layers closest to the endocarp enlarge in a radial direction, while exocarp cell layers
enlarge tangentially as fruit surface area increases with increasing size.

Larger sized fruit have been associated with increased cell numbers, increased cell
size, and increased intercellular spaces (Coombe, 1976). However, previous research
suggests that the role of intercellular space in Prunus ﬁnal fruit size increase is negligible
(Jackson and Coombe, 1966; Tukey and Young, 1939). The role of both mesocarp cell

number and cell size in relation to total fruit size has been examined. The bulk of this

research has been in apple (Malus domestica), where biennial bearing of many cultivars
has necessitated annual fruit thinning by hand or by chemical application. From this
body of work, evidence is conﬂicting as to whether cell number or cell size relates to
large fruit size. Increased fruit size within the same genotype after hand-thinning has
been attributed to differences in cell number (Bain and Robertson, 1951; Bergh,
1985,1990; Gofﬁnet et al., 1995; Martin et al., 1964; Westwood et al., 1967), cell size
(Al-Hinai and Roper, 2004; Atkinson et al., 2001 ), or a combination of both (Denne,
1960). However, between genotypes, differences in cell number and/or size have rarely
been documented. Only recently, the importance of increased cell division in
domesticated apple, as compared to related wild species, has been documented (Harada,
et a1. 2005).

In Prunus species, fruit mesocarp size has been associated with both cell number
and cell size. Although both cell number and size have been correlated with total fruit
size in experiments with a single cultivar (Bradley, 1959; Coombe, 1976; Jackson and
Coombe, 1966; Tukey and Young, 1939), relative differences in fruit cell number and
size between different cultivars have rarely been measured. When comparisons have
been made between cultivars of varying sizes, cell number was associated with overall
fruit size (Scorza et al., 1991; Yamaguchi et al., 2002a, 2002b, 2004). Alternatively,
signiﬁcant ﬁ'uit size differences within the same cultivar and without a corresponding
increase in cell number have been reported for other drupe species. In olive (Olea
europaea L.), development of fruit under irrigated and non-irrigated conditions, or after

water deﬁcit had been applied, resulted in overall fresh fruit size differences but no

signiﬁcant difference in cell number in the fruit (Costagli et al., 2003; Rapoport et al.,
2004)

Interestingly, these reports for Prunus species are similar to that of the model fruit
plant, tomato (Lycopersicon esculentum L.). In tomato, the gene underlying a major QTL
(wa. 2) contributing to fruit size difference between wild and cultivated species has been
cloned and shown to inﬂuence cell division and therefore ﬁnal fruit size (Frary et al.,

2000; Nesbitt and Tanksley, 2002).

Prunus Linkage Map Construction

Signiﬁcant improvement in average fruit size has been made in new cultivars in
all Prunus species. However, even with gain from selection being relatively high for this
quantitative trait, breeding long-lived perennial Prunus tree species is costly and time-
consuming. Long juvenility periods and garnetophytic self-incompatibility (de
Nettancourt, 1971) (peach being a notable exception), as well as large space
requirements, signiﬁcantly reduce progeny numbers that can be evaluated in a given time
period (Fogle, 1975). Marker-assisted selection (MAS) may hold the greatest promise
for improving selection efﬁciency in sweet cherry. However, efforts to develop suitable
linkage maps, and more importantly, identifying QTL for important agronomic traits in
sweet cherry, have lagged far behind other fruit crops.

Peach is the best genetically characterized member of Prunus. This is partly
because peach is the most economically important member of the genus, but also because
it is self—fertile and its juvenility period is shorter than most other Prunus species (3 years

versus 5-7 for sweet cherry). Self-fertility permits more amenable linkage mapping

populations such as backcross and F 2 to be used, rather than the F 1 pseudo-testcross
structure commonly used in cherry (Wang et al., 1998). However, self-fertility also has
limited the heterozygosity among cultivated peaches, and many Prunus linkage mapping
populations were developed from interspeciﬁc crosses between almond and peach
(Aranzana et al., 2003; Bliss et al., 2002; Dirlewanger et al., 2004a; F oolad et al., 1995;
Howad et al., 2005; Jauregui et al., 2001; Joobeur et al., 1998). Additional interspeciﬁc
crosses have been made between peach and related Prunus species such as P. davidiana
(Carr.) F ranch., P. ferganensis (Kost. & Rjab.) Y.Y. Yao, and P. cerasifera Ehrh. (Dettori
etal., 2001; Dirlewanger et al., 1996, 2004b; F oulongne et al, 2003).

One drawback to the use of linkage maps based on interspeciﬁc crosses has been
the high level of marker distortion, with up to 46% of the markers on these maps
deviating ﬁ'om the expected segregation ratios and indicating preferential inheritance of
certain genomic regions (Bliss et al., 2002; Foolad et al., 1995; Joobeur et al., 1998). In
many cases, the marker distortion has been attributed to garnetophytic selection due to
homologous pairing problems between species during meiosis. However, the presence of
an active garnetophytic self-incompatibility locus in many Prunus members also has been
implicated in the high percentage of markers exhibiting distorted segregation ratios (Bliss
et al., 2002; F oulongne et al., 2003; Joobeur et al., 1998; Lambert et al., 2004; Vilanova
et al., 2003).

Intraspeciﬁc linkage mapping populations also have been developed for other
Prunus members. Various peach and peach rootstock populations that segregate for
agronomic traits of interest such as tree architecture, peach-nectarine characters, and

nematode resistance, have been used to develop genetic linkage maps (Abbott et al.,

1998; Chaparro et al., 1994; Dirlewanger et al., 1998; Gillen and Bliss 2005; Lu et al.,
1998; Shimada etal., 2000; Yamamoto et al., 2005). In almond, mapping populations
were developed to identify bloom and self-incompatibility traits (Ballester et al., 1998,
2001; Joobeur et al., 2000). Similarly, in apricot, there are populations which segregate
for self-incompatibility and plum pox virus resistance (Hurtado et al., 2002; Lambert et
al., 2004; Vilanova et al., 2003).

In sour cherry, parental linkage maps have been developed for the ‘Rheinische
Schattenmorelle’ (‘RS’) X ‘Erdi Botenno’ (‘EB’) population (Wang et al., 1998). In
sweet cherry, Stockinger et a1. (1996) developed a RAPD marker-based linkage map of a
microspore-derived callus culture population. However, because of the marker system,
this map is not comparable with other Prunus linkage maps, and phenotypic analysis of
important fruit and tree traits are not possible for QTL studies. Similarly, Boskovic et a1.
(1997, 1998) reported isozyme-based interspeciﬁc maps of sweet cherry X P. incisa
Thunb. ex Murr. and sweet cherry X P. nipponica Matsum., but the marker density is not
conducive to QTL studies. Only Dirlewanger et al. (2004a) has developed a linkage map
from a ‘Regina’ X ‘Lapins’ sweet cherry cross that can be compared with current linkage
maps ﬁ'om other Prunus species using shared markers. However, ‘Regina’ (S1S3) and
‘Lapins’ (SIS4') are only partially compatible, and the loss of one pollen garnetophytic

class is likely to distort marker segregation ratios around the self-incompatibility locus.

Prunus Reference Map
One of the interspeciﬁc Prunus populations, 3 cross between ‘Texas’ (‘T’)

almond X ‘Earlygold’ (‘E’) peach, is considered the reference Prunus linkage map

because of the saturation of markers (0.92 cM average distance) and high number of
polymorphic simple sequence repeat (SSR) markers located on the map (Dirlewanger et
al., 2004a; Howad et al., 2005). The ‘T’ X ‘E’ population was developed originally by
J oobeur et a1. (1998) and consists of 88 F 2 progeny generated by selﬁng one individual
from the original cross. The ﬁrst generation of this map (Joobeur etal., 1998) consisted
of 246 RFLP and isozyme markers, covering a total distance of 491 cM over the expected
haploid chromosome number (x = 8) of linkage groups (LG) for diploid Prunus. As new
libraries of SSR markers were developed, they were subsequently added to the ‘T’ X ‘E’
map (Aranzana et al., 2003; Dirlewanger et al., 2004a). Additional RFLP probes from
Arabidopsis thaliana were also added (Dirlewanger eta1., 2004a). Currently, the map
consists of 562 markers covering 519 cM, with an average marker density of 0.92 cM,
and the largest gap of 7 cM (Dirlewanger et al., 2004a). The map distance is similar to
the predicted genome size of peach, 5.3X108 base pairs (Dickson et al., 1992). Most
recently, individuals from the ‘T’ X ‘E’ population were used in a selective bin mapping
strategy, whereby recombinational breakpoints are used to identify a small subset of
individuals that deﬁne a set of bins bounded by the breakpoints (Howad et al., 2005).
From this analysis, 264 additional SSRs were placed on the ‘T’ X ‘E’ map, although exact
linkage distances within each bin remain unknown (Howad et al., 2005). Twenty-eight
major agronomic genes have been integrated into the ‘T’ X ‘E’ map (Dirlewanger et al.,
2004a). The current ‘T’ X ‘E’ linkage map is available publicly through the Genome
Database for Rosaceae (GDR; http://www.rosaceae.org).

Since the adoption of the ‘T’ X ‘E’ map as the reference Prunus map, all linkage

group orientation and terminology have been assigned according to the ‘T’ X ‘E’

10

nomenclature. This has allowed comparison between the ‘T’ X ‘E’ map and several other
Prunus species (Dirlewanger et al., 2004a; Lambert et al., 2004). For the diploid Prunus
species, genome synteny appears to be the rule, not the exception (Dirlewanger et al.,
2004a). Only one major chromosomal rearrangement has been identiﬁed. A reciprocal
translocation between LG 6 and LG8 was identiﬁed in both an interspeciﬁc almond X
peach population and an intraspeciﬁc peach population (J auregui et al., 2001; Yamamoto
et al., 2001). Although the cultivars used in the development of these populations are
different, in each case a red-leaved peach was one of the parents. The gene for red vs.
green leaf color (Gr) is located close to the translocation breakpoint, and a relationship
between the cytogenetic and morphological phenotypes has not been excluded
(Dirlewanger et al., 2004a). Unfortunately, it is not known which of the parents in these

crosses had the standard or translocated conﬁgurations.

QTL Analyses in Prunus

Many vegetative, fruit, and disease resistance genes have now been mapped in
Prunus using these populations (Abbott et al., 1998; Ballester et al., 1998; Bliss et al.,
2002; Chaparro et al., 1994; Dirlewanger et al., 1996, 1998, 2004a; Gillen and Bliss
2005; Hurtado et al., 2002; Joobeur et al., 2000; Lambert et al., 2004; Lu et al., 1998;
Vilanova et al., 2003; Yamamoto et al., 2001). However, QTL analyses of Prunus
species has been documented only recently compared to other agronomic crops.
Dirlewanger et a1. (1996) identiﬁed QTL for powdery mildew resistance in a peach x P.
davidiana population designed speciﬁcally for that purpose. Subsequently, that

population was used for identiﬁcation of fruit quality traits such as bloom, maturity date,

11

ﬁ'uit and pit size, dry matter content, soluble solids content, individual sugar fractions,
organic acid fractions, titratable acidity, and ﬁ'uit and ﬂesh color (Quilot et al., 2004).
QTL for bloom date, maturity date, productivity, fresh weight, pH, titratable acidity,
soluble solid content, malic acid, citric acid, quinic acid, sucrose, glucose, fructose, and
sorbitol were identiﬁed in a peach intraspeciﬁc cross (Dirlewanger et al., 1999) and
candidate genes were subsequently identiﬁed for several of these sugar and organic acid
QTL (Etienne et al., 2002). Bloom date has been a priority for almond; QTL have been
identiﬁed for the trait (Ballester et al., 2001) and subsequent candidate genes were
located in similar positions as the late bloom QTL (Silva et al., 2005). QTL for bloom
date, pistil death, pollen germination, maturity date, fruit weight, and soluble solids
content were identiﬁed in sour cherry (Wang et al., 2000). Although only LG 2, LG 4,
LG 6, and LG 7 of sour cherry have been aligned with the same peach linkage groups
(Wang et al., 1998), it is possible to establish a degree of synteny between the two
species for some of these traits. For example, at least one QTL for bloom date, soluble
solids content, and maturity date was identiﬁed in both populations on the corresponding
LG 2, LG 4, and LG 6, respectively. To date, no QTL analyses have been performed in

sweet cherry.

Importance of Prunus LG 6

Prunus LG 6 appears to be of signiﬁcant importance to fruit size. On this linkage
group, Dirlewanger et a1. (1999) identiﬁed QTL for nearly all the fruit quality characters
measured in an intraspeciﬁc peach mapping population, including a QTL for fruit weight

explaining up to 47% of the total variation. In a different intraspeciﬁc peach population,

12

Yamamoto et al. (2001) identiﬁed four QTL for fruit weight. Three of these QTL were
located on LG3 in this map. However, shared markers with the ‘T’ X ‘E’ map
(Dirlewanger eta1., 2004a) indicate that LG3 in this map is the same as LG 6 and LG8 in
the ‘T’ X ‘E’ map. Two of the QTL map near the Dw brachytic dwarf locus which also is
located on LG 6 in a separate almond X peach population (Bliss et al., 2002). That the
linkage group in question consists of markers located on LG 6 and LG 8 of the ‘T’ X ‘E’
map is not without precedence. Jauregui et a1. (2001) found a similar situation in a cross
between ‘Garﬁ’ almond and ‘Nemared’ peach. In this case, a reciprocal translocation
was identiﬁed as the source of the exhibited linkage between LG 6 and LG8.
Coincidentally, the location of these fruit weight QTL appears to be near the self-
incompatibility locus (S). Sweet cherries possess a garnetophytic self-incompatibility
system. Until recently, most sweet cherry cultivars were self-incompatible, requiring co-
cultivation of at least two cultivars for adequate pollination. Although a few naturally
self-fertile cultivars have been described (Bargioni, 1996), self-fertility was not used
commonly until the release of the cultivar ‘Stella’. ‘Stella’ was developed from a cross
between ‘Lambert’ and the self-fertile seedling John Innes 2420. Gamma irradiated
‘Napoleon’ pollen, presumably creating a loss of function mutation for the pollen
recognition component of self-incompatibility, was used to fertilize ‘Emperor Francis’ to
create the John Innes 2420 seedling (Lewis and Crowe, 1954). Since the introduction of
‘Stella’ as a source of self-fertility (Lapins, 1970), all subsequent self-fertile cultivars
released have had ‘Stella’ in their pedigree. Although peach is self-fertile, almond has
the same self-incompatibility system as sweet cherry, and this locus has been mapped in

an almond X peach population (Bliss et al., 2002). On this map, the S locus is within 1.6

13

cM of the Dw locus. Therefore, several of the QTL described previously for fruit size in
peach are located in this area of LG 6. Linkage between QTL for fruit size and the S
locus may have great implications for future breeding efforts and marker assisted
selection. Because of the gametophytic incompatibility system in both almond and sweet
cherry, linkage distortion around the S locus is often observed in linkage maps (Bliss et
al., 2002; Joobeur et al., 1998; Foulongne et al., 2003; Lambert et al., 2004; Vilanova et
al., 2003). This occurs in cases where the paternal parent has a common S-allele with the
maternal parent. In these cases, haploid pollen containing the common S-allele is
rejected in the style of the ﬂower and cannot complete fertiliztion. This garnetophytic

selection is observed as distorted marker segregation ratios for those markers linked to

the S locus.

Utility of Simple Sequence Repeat Markers

Until recently, linkage mapping in Prunus was accomplished using
morphological, isozyme, RFLP (Restriction Fragment Length Polymorphism), RAPD
(Random Ampliﬁed Polymorphic DNA), and AF LP (Ampliﬁed Fragment Length
Polymorphism) markers. However, many labs have now developed SSR (Simple
Sequence Repeat) markers for use in Prunus (Aranzana et al., 2002; Cantini et al., 2001;
Cipriani et al., 1999; Clarke and Tobutt, 2003; Dirlewanger et al., 2002; Lopes et al.,
2002; Mnejja etal., 2004, 2005; Silva et al., 2005; Sosinski et al., 2000; Struss et al.,
2002; Testolin et al., 2000; Vaughan and Russell, 2004; Wang et al., 2002; Yamamoto et
al., 2002). SSR markers, short tandem repeats of single, di-, tri-, or tetranucleotide motifs

that are common throughout eukaryotic genomes (Ellegren, 2004; Weising et al., 1989),

14

are attractive because they are codominant, polymerase chain reaction (PCR) based,
repeatable, and often show a high degree of polymorphism. Because they are
codominant and repeatable, SSR markers are ideally suited for comparative mapping.

SSR markers developed from peach libraries have been added to maps of peach
and other Prunus species (Aranzana et a1, 2003; Bliss et al., 2002; Dirlewanger et al.,
2004a; Hurtado et al., 2002; Joobeur et al., 2000; Yamamoto et al., 2001). SSR markers
developed from peach libraries have shown ampliﬁcation in all of the main cultivated
Prunus species (peach/nectarine, sweet cherry, sour cherry, apricot, Japanese plum,
European plum, and almond) as well as wild Prunus species (P. serotina) (Aranzana et
al., 2002; Cantini et al., 2001; Cipriani et al., 1999; Dirlewanger et al., 2002; Downey and
Iezzoni, 2000; Horrnaza, 2002; Sosinski et al., 2000; Testolin et al., 2000; Wang et al.,
2002; Wunsch and Horrnaza, 2002;Yamamoto et al., 2002). SSR markers from peach
and other Prunus species have been used for ﬁngerprinting and genetic diversity analysis
of peach (Aranzana et al., 2002; Dirlewanger et al., 2002; Testolin et al., 2000), apricot
(Hormaza, 2002), sweet cherry (Wunsch and Hormaza, 2002; Dirlewanger et al., 2002),
and sour cherry (Cantini et al., 2001).

The transferability and reproducibility of SSR markers, as well as the extensive
collinearity of Prunus genomes, warrants their extensive use in any new linkage map

development.

OBJECTIVES
To fully exploit the genetic potential to increase fruit size in sweet and sour

cherry, a more thorough understanding of the control of this quantitative trait is needed.

15

For increased efﬁciency in the breeding process, a reductionist approach can be used,
whereby the total phenotypic variation for a quantitative trait such as ﬁ'uit size is reduced
to various components that inﬂuence the overall phenotype. In this manner, phenotypic
variance for the trait can be partitioned into testable units to determine those with the
most genotypic variance. This strategy is attractive for sweet and sour cherry, since the
available F1 population structures limit the ability to identify minor-effect QTL.

Two populations well-suited for the identiﬁcation of fruit size QTL were recently
developed in the Michigan State University sour cherry breeding and genetics program.
Reciprocal crosses between the sweet cherries ‘New York 54’ (‘NY 54’) and ‘Emperor
Francis’ (‘EF’) were used to develop a population of 617 individuals. ‘NY 54’ is a small
(1-2 g), acid, dark red, wild forest Mazzard selection, while ‘EF’ is a larger (6-8 g), sub-
acid, blushed yellow cherry cultivar. This intraspeciﬁc cross represents the change in
fruit size that occurred during domestication of wild sweet cherry. In sour cherry, the
phenotypic difference between the parents of the ‘Ujfehe'rtoi F tirtos’ (‘UF ’) X ‘Sureﬁre’
population was not great, but transgressive segregation for fruit size in the progeny from
the population suggested that different alleles for fruit size QTL were segregating (A.F.
Iezzoni, pers. com.)

The overall goal of this study was to understand the genetic bases for achieving
large fruit size in sweet and sour cherry. Experiments were designed to provide
knowledge that would be used to develop future genetic improvement strategies to
maximize fruit size in new cultivars. The speciﬁc objectives for this project included
identifying cherr'y fruit mesocarp histological differences that are associated with fruit

size differences, development of genetic linkage maps suitable for comparative mapping

l6

within Prunus, and identiﬁcation of the loci responsible for fruit size differences through

QTL analyses.

17

LITERATURE CITED

Abbott, A.G., S. Rjapakse, B. Sosinki, Z.X. Lu, K. Sossey-Alaoui, M. Gannavarapu, G.
Reighard, R.E. Ballard, W.V. Baird, R. Scorza, and A. Callahan. 1998.
Construction of saturated linkage maps of peach crosses segregating for

characters controlling fruit quality, tree architecture and pest resistance. Acta
Hort. 465: 41-49.

Al-Hinai, Y.K. and TR. ROper. 2004. Rootstock effects on growth, cell number, and
cell size of ‘Gala’ apples. J. Amer. Soc. Hort. Sci. 129:37—41.

Aranzana, M.J., A. Pineda, P. Cosson, E. Dirlewanger, J. Ascasibar, G. Cipriani, C.D.
Ryder, R. Testolin, A. Abbott, G.J. King, A.F. Iezzoni, and P. Arus. 2003. A set
of simple sequence repeat (SSR) markers covering the Prunus genome. Theor.
Appl. Genet. 106:819-825.

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

Atkinson, C.J., L. Taylor and G. Kingswell. 2001. The importance of temperature

differences, directly after anthesis, in determining growth and cellular
development of Malus fruits. J. Hortic. Sci. Biotech. 76:721-731.

Bain, J .M. and RN. Robertson. 1951. The physiology of growth in apple fruits. Aust. J.
Sci. Res. 4:75-91.

Ballester, J ., R. Boskovic, 1. Battle, P. Arus, F. Vargas, and MC. de Vicente. 1998.
Location of the self-incompatibility gene on the almond linkage map. Plant
Breed. 117:69-72.

Ballester, J ., R. Socias I Company, P. Arus, and MC. de Vicente. 2001. Genetic
mapping of a major gene delaying blooming time in almond. Plant Breed.
120:268-270.

Bargioni, G. 1996. Sweet Cherry Scions: Characteristics of the Principal Commercial
Cultivars, Breeding Objectives and Methods. In: Webster, AD. and Looney,
N.E. (eds) Cherries: Crop Physiology, Production and Uses. CAB International,
Cambridge, UK, pp. 73-112.

Beaver, J .A. and AP. Iezzoni. 1993. Allozyme inheritance in tetraploid sour cherry
(Prunus cerasus L.). J. Amer. Soc. Hort. Sci. 118:873-877.

Bergh, O. 1985. Effect of the previous season’s crop on cortical cell number of Malus
domestica cv. Starking apple ﬂower primordia, ﬂowers and fruit. S. Afr. J. Plant
Soil 2:191-196.

18

Bergh, O. 1990. Effect of time of hand-thinning on apple fruit size. S. Afr. J. Plant Soil
7:1-10.

Bliss, F .A., S. Arulsekar, M.R. Foolad, V. Becerra, A.M. Gillen, M.L. Warburton, A.M.
Dandekar, G.M. Kocsisne, and K.K. Mydin. 2002. An expanded genetic linkage

map of Prunus based on an interspeciﬁc cross between almond and peach.
Genome 45:520-529.

Boskovic, R. and KR. Tobutt. 1998. Inheritance and linkage relationships of
isoenzymes in two interspeciﬁc cherry progenies. Euphytica 103:273-286.

Boskovic, R., K.R. Tobutt, and F.J. Nicoll 1997. Inheritance of isoenzymes and their

linkage relationships in two interspeciﬁc cherry progenies. Euphytica 93: 129-
143.

Bradley, M.V. 1959. Mean cell size in the mesocarp of mature peaches of different
sizes. Proc. Amer. Soc. Hort. Sci. 73:120-124.

Brettin, T.S., R. Karle, E.J. Crowe, and A.F. Iezzoni. 2000. Chloroplast inheritance and
DNA variation in sweet, sour, and ground cherry. J. Hered. 91:75-79.

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

Chalmers, DJ. and B. van den Ende. 1973. A reapprisal of the growth and development
of peach fruit. Aust. J. Plant Physiol. 22623-34.

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

Cipriani, G., G. Lot, W.G. Huang, M.T. Marrazzo, E. Peterlunger, and R. Testolin. 1999.
AC/GT and AG/CT microsatellite repeats in peach [Prunus persica (L.) Batsch]:

isolation characterization and cross-species ampliﬁcation in Prunus. Theor. Appl.
Genet. 99:65-72.

Clarke, J .B. and KR. Tobutt. 2003. Development and characterization of polymorphic
microsatellites from Prunus avium ‘Napoleon’. Mol. Ecol. Notes 3:578-580.

Coombe, B.G. 1976. The development of ﬂeshy fruits. Ann. Rev. Plant Physiol.
27:507-528.

Costagli, G., R. Gucci and HF. Rapoport. 2003. Growth and development of fruits of

olive ‘Frantoio’ under irrigated and rainfed conditions. J. Hortic. Sci. Biotech.
78:119-124.

19

DeJong, TM. and J. Goudriaan. 1989. Modeling peach fruit growth and carbohydrate
requirements: Re-evaluation of the double-sigmoid growth pattern. J. Amer. Soc.
Hort. Sci. 114:800-804.

de Nettancourt, D. 1977. Incompatibility in angiosperrns. Springer, New York, NY.

Denne, MP. 1960. The grth of apple ﬁ'uitlets, and the effect of early thinning on ﬁ'uit
development. Ann. Bot. 24:397-406.

Dettori, M.T., R. Quarta, and I. Verde. 2001. A peach linkage map integrating RF LPs,
SSRs, RAPDs, and morphological markers. Genome 44:783-790.

Dickson, E.E., K. Arumuganathan, S. Kresovich, and J .J . Doyle. 1992. Nuclear DNA
content variation within the Rosaceae. Amer. J. Bot. 79:1081-1086.

Dirlewanger, E., A. Moing, C. Rothan, L. Svanella, V. Pronier, A. Guye, C. Plomion, and
R. Monet. 1999. Mapping QTL controlling fruit quality in peach [Prunus persica
(L.) Batsch]. Theor. Appl. Genet. 98:18-31.

Dirlewanger, E., E. Graziano, T. Joobeur, F. Garriga-Caldere, P. Cosson, W. Howad, and
P. Arus. 2004a. Comparative mapping and marker-assisted selection in Rosaceae
fruit crops. Proc. Natl. Acad. Sci. 101:9891-9896.

Dirlewanger, E., P. Cosson, M. Tavaud, M.J. Aranzana, C. Poizat, A. Zanetto, P. Arus,
and F. Laigrct. 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., P. Cosson, W. Howad, G. Capdeville, N. Bosselut, M. Claverie, R.
Voisin, C. Poizat, B. Lafargue, O. Baron, F. Laigret, M. Kleinhentz, P. Arus, and
D. Esmenjaud. 2004b. 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., T. Pascal, C. Zuger, and J. Kervella. 1996. Analysis of molecular
markers associated with powdery mildew resistance genes in peach (Prunus
persica (L.) Batsch) x Prunus davidiana hybrids. Theor. Appl. Genet. 93:909-
919.

Dirlewanger, E., V. Pronier, C. Parvery, C. Rothan, A. Guye, and R. Monet. 1998.
Genetic linkage map of peach [Prunus persica (L.) Batsch] using morphological
and molecular markers. Theor. Appl. Genet. 97:888-895.

Downey, S.L. and A.F. Iezzoni. 2000. Polymorphic DNA markers in black cherry

(Prunus serotina) are identiﬁed using sequences from sweet cherry, peach, and
sour cherry. J. Amer. Soc. Hort. Sci. 125:76-80.

20

Ellegren, H. 2004. Microsatellites: Simple sequences with complex evolution. Nat.
Rev. Genet. 5:435-445.

Esau, K. 1977. Anatomy of Seed Plants. 2nd ed. John Wiley and Sons, Inc., New York,
NY.

Etienne, C., C. Rothan, A. Moing, C. Plomion, C. Bodenes, L. Svanella-Dumas, P.
Cosson, V. Pronier, R. Monet, and E. Dirlewanger. 2002. Candidate genes and
QTLs for sugar and organic acid content in peach [Prunus persica (L.) Batsch].
Theor. Appl. Genet. 105: 145-149.

FAOSTAT data, 2005.

F ogle, H.W. 1961. Inheritance of some fruit and tree characteristics in sweet cherry
crosses. Proc. Amer. Soc. Hort. Sci. 78:76-85.

F ogle, H.W. 1975. Cherries. In: Janick, J. and IN. Moore (eds) Advances in Fruit
Breeding. Purdue University Press, West Lafayette, Ind, pp. 348-366.

Foolad, M.R., S. Arulsekar, V. Becerra, and RA. Bliss. 1995. A genetic map of Prunus
based on an interspeciﬁc cross between peach and almond. Theor. Appl. Genet.
91:262-269.

F oulongne, M., T. Pascal, P. Arus, and J. Kervella. 2003. The potential of Prunus
davidiana for introgression into peach [Prunus persica (L.) Batsch] assessed by
comparative mapping. Theor. Appl. Genet. 107:227-238.

Frary, A., T.C. Nesbitt, A. Frary, S. Grandillo, E. van der Knapp, B. Cong, J. Liu, J.
Meller, R. Elber, K.B. Alpert, and SD. Tanksley. 2000. wa.2: A quantitative
trait locus key to the evolution of tomato fruit size. Science 289:85-88.

Gage, J. and G. Stutte. 1991. Developmental indices of peach: An anatomical
framework. HortScience 26:459-463.

Gillen, A.M. and RA. Bliss. 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.

Gofﬁnet, M.C., T.L. Robinson and AN. Lakso. 1995. A comparison of ‘Empire’ apple
fruit size and anatomy in unthinned and hand-thinned trees. J. Hortic. Sci.
70:375-3 87.

Hansche, P.E., V. Beres, and RM. Brooks. 1966. Heritability and genetic correlation in
the sweet cherry. Proc. Amer. Soc. Hort. Sci. 88:173-183.

21

Harada, T., W. Kurahashi, M. Yanai, Y. Wakasa and T. Satoh. 2005. Involvement of
cell proliferation and cell enlargement in increasing the fruit size of Malus
species. Sci. Hortic. 105:447-456.

Horrnaza, J .I. 2002. Molecular characterization and similarity relationships among
apricot (Prunus armeniaca L.) genotypes using simple sequence repeats. Theor.
Appl. Genet. 104:321-328.

Howad, W., T. Yamamoto, E. Dirlewanger, R. Testolin, P. Cosson, G. Cipriani, A.J.
Monforte, L. Georgi, A.G. Abbott, and P. Arus. 2005. Mapping with a few
plants: Using selective mapping for microsatellite saturation of the Prunus
reference map. Genetics 171:1305-1309.

Hurtado, M.A., C. Romero, S. Vilanova, A.G. Abbott, G. Llacer, and ML. Badenes.
2002. Genetic linkage maps of two apricot cultivars (Prunus armeniaca L.), and
mapping of PPV (sharka) resistance. Theor. Appl. Genet. 105:182-191.

Iezzoni, A.F. 1988. ‘Montrnorency’ sour cherry. Fruit Var. J. 42:74-75.

Iezzoni, A.F. 2005. Acquiring cherry gerrnplasm from central and eastern Europe.
HortScience 40:304-308.

Iezzoni, A., H. Schmidt, and A. Albertini. 1990. Cherries (Prunus). In: Moore, J .N. and
J .R. Ballington Jr. (eds) Genetic Resources of Temperate Fruit and Nut Crops,
vol. 1, ISHS, Wageningen, the Netherlands. pp 111-173.

Jackson, D.I. and B.G. Coombe. 1966. The growth of apricot mu 1. Morphological
changes during development and the effects of various tree factors. Aust. J.
Agric. Res. 17:465-477.

Jauregui, B., M.C. de Vicente, R. Messeguer, A. Felipe, A. Bonnet, G. Salesses, and P.
Arus. 2001. A reciprocal translocation between ‘Garﬁ’ almond and ‘Nemared’
peach. Theor. Appl. Genet. 102:1 169-1176.

Joobeur, T., M.A. Viruel, M.C. de Vicente, B. Jauregui, J. Ballester, M.T. Dettori, I.
Verde, M.J. Truco, R. Messeguer, I. Batlle, R. Quarta, E. Dirlewanger, and P.
Arus. 1998. Construction of a saturated linkage map for Prunus using an almond
x peach F2 progeny. Theor. Appl. Genet. 97:1034-1041.

Joobeur, T., N. Periam, M.C. de Vicente, G.J. King, and P. Arus. 2000. Development of

a second generation linkage map for almond using RAPD and SSR markers.
Genome 43:649-655.

22

Lamb, RC. 1953. Notes on the inheritance of some characters in the sweet cherry
Prunus avium. Proc. Amer. Soc. Hort. Sci. 61 :293-298.

Lambert, P., L.S. Hagen, P. Arus, and J .M. Audergon. 2004. Genetic linkage maps of
two apricot cultivars (Prunus armeniaca L.) compared with the almond ‘Texas’ X

peach ‘Earlygold’ reference map for Prunus. Theor. Appl. Genet. 108:1120-
1130.

Lang, G.A., B. Behe, J. Sowa, A. Iezzoni, and E. Fallahi. 2003. Niche marketing of fruit
crops: Windows for proﬁtability? HortScience 38(5):746.

Lapins K. 1970. The Stella cherry. Fruit Var Hort Dig 24:19-20.

Lewis, D. and L.K. Crowe. 1954. The induction of self-fertility in tree fruits. J. Hort.
Sci. 29:220-225.

Lilleland, O. and L. Newsome. 1934. A growth study of the cherry fruit. Proc. Amer.
Soc. Hort. Sci. 32:291-299.

Lopes, M.S., K.M. Sefc, M. Laimer, and A. Da Camara Machado. 2002. Identiﬁcation
of microsatellite loci in apricot. Molecular Ecology Notes 2:24-26.

Lu, Z.-X., B. Sosinski, G.L. Reighard, W.V. Baird, and A.G. Abbott. 1998.
Construction of a genetic linkage map and identiﬁcation of AF LP markers or
resistance to root-knot nematodes in peach rootstocks. Genome 41:199-207.

Martin, D., T.L. Lewis and J. Cemy. 1964. Apple fruit cell numbers in relation to
cropping alternation and certain treatments. Aust. J. Agric. Res. 15:905-919.

Matthews, P. 1973. Some recent advances in sweet cherry genetics and breeding. Proc.
EUCARPIA Fruit Section Symp. 5. Topic Fruit Breeding. Canterbury: 84-107.

Mnejja, M., J. Garcia-Mas, W. Howad, and P.Arus. 2005. Development and
transportability across Prunus species of 42 polymorphic almond microsatellites.
Mol. Ecol. Notes 5:531-535.

Mnejja, M., J. Garcia-Mas, W. Howad, M.L. Badenes, and P. Arus. 2004. Simple
sequence repeat (S SR) markers of Japanese plum (Prunus salicina Lindl.) are
highly polymorphic and transferable to peach and almond. Mol. Ecol. Notes
4:163-166.

NASS-USDA. 2005. Cherry Production. June, 2005.
Nesbitt, T.C. and SD. Tanksley. 2002. Comparative sequencing in the genus

Lycopersicon: Implications for the evolution of fruit size in the domestication of
cultivated tomatoes. Genetics 162:365-379.

23

Olden, E.J. and N. Nybom. 1968. On the origin of Prunus cerasus L. Hereditas 59:327-
345.

Quilot, B., B.H. Wu, J. Kervella, M. Genard, M. Foulongne, ad K. Moreau. 2004. QTL
analysis of quality traits in an advanced backcross between Prunus persica and
the wild relative species P. davidiana. Theor. Appl. Genet. 109:884-897.

Ragland, CH. 1934. The development of the peach ﬁ'uit, with special reference to split-
pit and gumming. Proc. Amer. Soc. Hort. Sci. 31:1-21.

Rapoport, H.F., G. Costagli and R. Gucci. 2004. The effect of water deﬁcit during early

fruit development on olive ﬁ'uit morphogenesis. J. Amer. Soc. Hort. Sci.
129:121-127.

Scorza, R., L.G. May, B. Purnell and B. Upchurch. 1991. Differences in number and
area of mesocarp cells between small- and large-ﬁnited peach cultivars. J. Amer.
Soc. Hort. Sci. 116:861-864.

Shimada, T., T. Yamamoto, H. Hayama, M. Yamaguchi, and T. Hayashi. 2000. A
genetic linkage map constructed by using an intraspeciﬁc cross between peach
cultivars grown in Japan. J. Japan. Soc. Hort. Sci. 69:536-542.

Silva, C., J. Garcia-Mas, A.M. Sanchez, P. Arus, and M.M. Oliveira. 2005. Looking into
ﬂowering time in almond (Prunus dulcis (Mill) D.A. Webb): the candidate gene
approach. Theor. Appl. Genet. 110:959-968.

Sosinski, B., M. Gannavarapu, L.D. Hager, L.E. Beck, G.J. King, C.D. Ryder, S.
Rajapakse, W.V. Baird, R.E. Ballard, and A.G. Abbot. 2000. Characterization of

microsatellite markers in peach [Prunus persica (L.) Batsch]. Theor. Appl.
Genet. 101:421-428.

Stockinger E.J., C.A. Mulinix, C.M. Long, T.S. Brettin, and A.F. Iezzoni. 1996. A
linkage map of sweet cherry based on RAPD analysis of a microspore-derived
callus culture population. J. Hered. 87:214-218.

Struss, D., M. Boritzki, R. Karle, and A.F. Iezzoni. 2002. Microsatellite markers
differentiate eight Giessen cherry rootstocks. HortScience 37:191-193.

Testolin, R., T. Marrazzo, G. Cipriani, R. Quarta, I. Verde, M.T. Dettori, M. Pancaldi,
and S. Sansavini. 2000. Microsatellite DNA in peach (Prunus persica L. Batsch)
and its use in ﬁngerprinting and testing the genetic origin of cultivars. Genome
43:512-520.

Tukey, H.B. and J .0. Young. 1939. Histological study of the developing fruit of the
sour cherry. Bot. Gaz. 100:723-749.

24

Vaughan, SP. and K. Russell. 2004. Characterization of novel microsatellites and
development of multiplex PCR for large-scale population studies in wild cherry,
Prunus avium. Mol. Ecol. Notes 4:429-431.

Vilanova, S., C. Romero, A.G. Abbott, G. Llacer, and ML. Badenes. 2003. An apricot
(Prunus armeniaca L.) F2 progeny linkage map based on SSR and AF LP

markers, mapping pltun pox virus resistance and self-incompatibility traits.
Theor. Appl. Genet. 107:239-247.

Wang, D., R. Karle, and A.F. Iezzoni. 2000. QTL analysis of ﬂower and fruit traits in
sour cherry. Theor. Appl. Genet. 100:535-544.

Wang, D., R. Karle, T.S. Brettin, and A.F. Iezzoni. 1998. Genetic linkage map in sour
cherry using RF LP markers. Theor. Appl. Genet. 97:1217-1224.

Wang, Y., L.L. Georgi, T.N. Zhebentyayeva, G.L. Reighard, R. Scorza, and A.G. Abbott.
2002. High-throughput targeted SSR marker development in peach (Prunus
persica). Genome 45:319-328.

Weising, K., F. Weigand, A.J. Driesel, G. Kahl, H. Zischler, and J .T. Epplen. 1989.
Polymorphic simple GATA/GACA repeats in plant genomes. Nucleic Acids Res.
17:128.

Westwood, M.N., L.P. Batjer and HS. Billingsley. 1967. Cell size, number, and mu
density of apples as related to fruit size, position in cluster, and thinning method.
Proc. Amer. Soc. Hort. Sci. 91:51-62.

Whiting, M.D., D. Ophardt, and J .R. McFerson. 2006. Chemical blossom thinners vary
in their effect on sweet cherry fruit set, yield, fruit quality, and crop value.
HortTechnology 16:66-70.

Whiting, M.D., G. Lang and D. Ophardt. 2005 Rootstock and training system affect
cherry growth, yield, and fruit quality. HortScience 40:582-586.

Wunsch, A., and J .I. Hormaza. 2002. Molecular characterization of sweet cherry
(Prunus avium L.) genotypes using peach [Prunus persica (L.) Batsch] SSR
sequences. Heredity 89:56-63.

Yamaguchi, M. I. Sato, K. Takase, A. Watanabe and M. Ishiguro. 2004. Differences and
yearly variation in number and size of mesocarp cells in sweet cherry (Prunus
avium L.) cultivars and related species. J. Japan. Soc. Hort. Sci. 73:12-18.

Yamaguchi, M., T. Haji, M. Miyake and H. Yaegaki. 2002a. Studies on the varietal
differences and yearly deviation of mesocarp cell munbers and lengths and ﬁ'uit
weight among commercial peach [Prunus persica (L.) Batsch] cultivars and
selections, wild types, and their hybrids. J. Japan. Soc. Hort. Sci. 71:459-466.

25

Yamaguchi, M., T. Haji, M. Miyake and H. Yaegaki. 2002b. Varietal differences in cell
division and enlargement periods during peach (Prunus persica Batsch) fruit
development. J. Japan. Soc. Hort. Sci. 71:155-163.

Yamamoto, T., K. Mochida, T. lmai, Z. Shi, I. Ogiwara, and T. Hayashi. 2002.
Microsatellite markers in peach [Prunus persica (L.) Batsch] derived from an
enriched genomic and cDNA libraries. Molecular Ecology Notes 2:298-301.

Yamamoto, T., M. Yamaguchi, and T. Hayashi. 2005. An integrated linkage map of
peach by SSR, STS, AF LP and RAPD. J. Japan. Soc. Hort. Sci. 74:204-213.

Yamamoto, T., T. Shimada, T. lmai, H. Yaegaki, T. Haji, N. Matsuta, M. Yamaguchi,

and T. Hayashi. 2001. Characterization of morphological traits based on a
genetic linkage map in peach. Breeding Sci. 51 :271-278.

26

CHAPTER TWO

Genotypic differences in sweet cherry (Prunus avium L.) fruit size are primarily a

function of cell number

27

ABSTRACT

Large fruit size is critical for proﬁtable fresh sweet cherry (Prunus avium L.)
production. However, little is known about histological differences that contribute to
ﬁ'uit size differences in sweet cherry. Although fruit size varies widely between sweet
cherry cultivars, signiﬁcant variation exists among genetically identical fruit from the
same cultivar due to cultural and environmental differences. This research examined the
difference in mesocarp cell traits between ﬁve cultivars [‘Selah’, ‘Emperor Francis’
(‘EF’), ‘New York 54’ (‘NY 54’), ‘Bing’, and ‘Regina’] that represent a wide range of
fruit size in sweet cherry, as well as within genotype differences in fruit size due to
environmental variation.

The relative contributions of mesocarp cell number and size to ﬁnal fresh fruit
size were determined by analyzing equatorial sections of ‘Selah’, ‘EF’, ‘NY 54’, ‘Bing’,
and ‘Regina’. The cell number count in this dimension, representing the total radial cell
division at the widest diameter of cherry fruit, was signiﬁcantly different (P < 0.05)
between all cultivars except ‘Bing’ and ‘Regina’. The relationship of cell number with
fruit weight and diameter was signiﬁcantly and positively correlated (r2 = 0.72 and 0.59,
respectively), while cell length was not correlated with either measure of fruit size. Cell
number was stable during the three years of this study and at two different locations.
Differences in cell number due to environmental variation were examined in ﬁ'uit of
‘Bing’, ‘Regina’, and ‘Selah’ that differed signiﬁcantly (P < 0.001) in fruit size. Within
each cultivar, ﬁ'uit size differences were attributed to a difference in cell size rather than
cell number, conﬁrming cell number as the primary genetic component resulting in fruit

size differences. Cell division differences were most pronounced during the

28

developmental period between bloom and endocarp scleriﬁcation. This experiment
suggests that duration of cell division affects fruit size more than the rate of cell division.
Based on these results, ﬁ'uit mesocarp cell number is controlled genetically and has low
environmental variance. Therefore, this trait could be used for selection of improved fruit

size through breeding efforts.

29

INTRODUCTION

The mature cherry fruit is composed of a thin protective exocarp, a ﬂeshy
mesocarp, and an inedible stony endocarp or pit surrounding the seed (Esau, 1977). All
three tissue types arise from the ovary and the increase in fruit size results from a
coordinated series of cell divisions and expansions. Cherry, and other Prunus species,
exhibit a double sigmoid growth curve, consisting of three distinct growth stages
(Coombe, 1976; Chalmers and van den Ende, 1975; Lilleland and Newsome, 1934;
Nitsch, 1953). Stage I is characterized by rapid and exponential fruit growth following
anthesis, stage II by a lag period of fruit growth coinciding with endocarp (pit) hardening
and embryo development, and stage III by a second period of exponential fruit growth
ending with either harvest or physiological maturity. During Stage I, mesocarp growth
consists of both cell division and cell enlargement, while Stage III mesocarp growth is
predominantly due to cell expansion (Coombe, 1976; Gage and Stutte, 1991; Nitsch,
1953; Tukey and Young, 1939).

Sweet cherry fruit exhibit a dramatic range in ﬁ'uit size. Wild forms of forest
sweet cherry, which are generally used for wood or cherry rootstocks, have small (~l to 2
g) fruit that is ideal for dispersal since it consists predominantly of the pit containing the
seed. In contrast, cultivated sweet cherries produce fruit that weigh approximately 6 g
(exhibited by the old landrace varieties) to over 13 g for cultivated varieties. A previous
study of a diverse range of sweet cherry selections that exhibit differences in ﬁ'uit size
concluded that mesocarp thickness is determined primarily by cell number, however,
differences in cell size were also found (Yamaguchi et al., 2004). In peach [Prunus

persica (L.) Batsch], cell number and cell size were found to vary continuously in

30

cultivars of different fruit sizes indicating that fruit size is a quantitative character (Scorza
et al., 1991; Yamaguchi et al., 2002a).

Fruit diameter is the primary criterion upon which fresh cherries are graded for
sale. Fruit averaging over 29 mm in diameter are worth nearly twice as much ($/kg) as
ﬁ'uit less than 24 mm in diameter (Whiting et al., 2005, 2006). Therefore, our long term
goal is to determine the quantitative trait loci (QTL) that control fruit size, identify large
fruited alleles at these loci, and use marker assisted selection to increase the efﬁciency of
breeding large fruited sweet cherry cultivars. The improved fruit size of old landrace
varieties (~6 g) compared to sweet cherry forest trees (~1-2 g) represents a classic
example of fruit size increase associated with domestication (Janick, 2004; Tanksley,
2004). This common recurring increase in ﬁ'uit size that has accompanied the
domestication of many fruit and vegetable crops has been studied in most detail in tomato
(Lycopersicon esculentum Mill.), the fruit of which is a ﬂeshy carpel like in cherry
(Doganlar et al., 2002; Frary et al., 2000; Grandillo et al., 1999; Nesbitt and Tanksley,
2002; Tanksley, 2004). These studies suggest that the evolution of fruit size in tomato
likely represents the “stacking” of alleles at many loci. However, accumulated QTL
evidence for domestication traits in maize (Zea mays L.), rice (Oryza sativa L.), sorghum
(Sorghum bicoIor L.) and tomato supports the hypothesis that the majority of
domestication-associated anatomical changes can be attributed to a few loci with larger
effects (Paterson, 2002, Tanksley, 2004). The application of this hypothesis to sweet
cherry is consistent with the ﬁnding that cultivars with ﬁ'uit sizes of over 13 g have been
obtained from just three generations of breeding among 6 to 8 g landrace selections (Choi

and Kappel, 2004).

31

QTL analysis is a powerful method to identify those chromosomal regions
carrying genes contributing to trait variation, as this analysis requires no a priori
information other than the existence of variation and a means to quantify this variation
(Paterson, 2002). However, not all QTL can be detected with statistical signiﬁcance. In
many cases, the ability to quantify trait variation is the limiting factor, as QTL cannot be
resolved if signiﬁcant environmental and sampling variation obscures the resulting
phenotype. To increase our ability to identify fruit size QTL, we investigated the
components of mu size among potential parental selections to determine those traits that
would be most likely to identify QTL that contribute to an increase in fruit size. The
speciﬁc objectives were to determine: (1) the relative contribution of mesocarp ﬁ'uit cell
number and size differences to mature fruit weights among ﬁve sweet cherry selections,
(2) the environmental stability of these measurements, and (3) the relative timing of the

cell number increases.

MATERIALS AND METHODS
Plant material

Images in this dissertation are presented in color. Five sweet cherry cultivars
were selected to represent a wide range in average ﬁesh fruit size. ‘NY 54’ is a small-
fruited wild cherry selection, used commercially as a seed source for seedling P. avium
rootstock. ‘EF’ is a mid-sized old European cultivar of unknown origin, representing the
fruit size achieved through domestication. ‘Bing’ is a large-fruited, 130-year-old
selection, while ‘Regina’ is a large-fruited cultivar released in 1998. Finally, ‘Selah’ is a

very large-fruited cultivar introduced in 2000, which is among the largest current

32

cultivars. Experimental trees were located at Washington State University’s Irrigated
Agriculture Research and Extension Center (WSU-IAREC) in Prosser, Wash. (46.29 N,
119.73 W), and Michigan State University’s Clarksville Horticultural Station (MSU-
CHES) in Clarksville, Mich (42.87 N, 85.26 W). For comparative histology
measurements between genotypes, all except ‘NY 54’, ‘Bing’, and ‘Regina’ at MSU-
CHES were mature (> 20 yr) trees grafted on P. avium seedling rootstock and trained to
an open-center. ‘NY 54’, ‘Bing’, and ‘Regina’ at MSU-CHES were younger (3-5 yr),
grafted on ‘Gisela 6’ rootstock and trained to a central leader system. For histology
measurements within genotype, young (7 yr) ‘Selah’, ‘Bing’, and ‘Regina’ trees grafted
on P. avium seedling rootstock were used. Trees at WSU-IAREC were irrigated weekly
with under-tree sprinklers, while those at MSU-CHES were provided supplemental
irrigation by drip lines from mid-June until August. Standard orchard management
practices for each location (irrigation, fertilization, pest control, and dormant pruning)

were followed.

Flower and fruit sampling scheme for mesocarp cell number and size comparisons
among the ﬁve cultivars

To evaluate histological differences among the ﬁve cultivars (‘Selah’, ‘EF ’, ‘NY
54’, ‘Bing’, and ‘Regina’), well-exposed fruit were sampled randomly from the exterior
portion of the canopy. Five ﬁ'uit from each cultivar were analyzed. At WSU-IAREC,
‘Bing’ and ‘Regina’ were not sampled in 2003, and ‘Selah’ was not sampled from MSU-
CHES in 2004 or 2005. In 2003, samples at bloom, endocarp hardening, and harvest

maturity were taken at WSU-IAREC. Bloom samples were taken when 50% of the

33

ﬂowers on a treatment tree were open. Only ﬂowers that had recently opened ﬁrlly, as
judged by non-dehiscent anthers, were sampled. Endocarp hardening was when a
complete cut could not be made easily through the fruit. In 2004, samples at each
developmental stage were taken from WSU-IAREC and MSU-CHES, as well as weekly
samples during the period from bloom to endocarp hardening. In 2005, samples were
taken at one to two day intervals for all genotypes except ‘EF’ at WSU-IAREC. In 2005,
samples at the endocarp hardening stage were used to calculate cell numbers for ‘EF’ and
‘NY 54’ from WSU-IAREC and for all genotypes from MSU-CHES. To equalize the
potential temperature inﬂuence on cell division, growing degree day (GDD)
accmnulation using a 4.4 C base temperature was calculated for periods between

sampling dates.

Flower bud thinning treatments to determine the inﬂuence of crop load on mesocarp cell
number and size

To evaluate different-sized ﬁ'uits within genotype, three cultivars (‘Selah’, ‘Bing’,
and ‘Regina’) were subjected to whole tree pre-bloom thinning treatments at WSU-
IAREC in 2004 and 2005. Selah was not sampled in 2004. For the thinning treatments,
all spurs on a tree were hand-thinned to one ﬂower bud per spur prior to bloom. Control
trees were left unthinned. This thinning treatment had previously resulted in signiﬁcant
fruit size differences (Lang and Ophardt, 2000; Whiting and Lang, 2004). However, for
some genotypes, thinning did not result in signiﬁcant ﬁ'uit size difference between

treatments; in such cases, a large random sample of fruit was harvested at maturity, and

34

individual fruit were weighed to create pools of small and large size fruit from the same
genotype.
Fruit measurement and sectioning

The ﬁve fruit per cultivar or treatment were weighed individually and diameters
were measured at the widest point of the fruit (Fig. 1) using a digital caliper. The fruit
then were placed individually in storage vessels, immersed in a formalin-acetic acid-
alcohol solution (10:5:50 FAA; Ruzin, 1999) and stored until sectioning. Radial
mesocarp ﬂesh sections were obtained at the widest diameter of the ﬁ'uit (Fig. 1) by hand
sectioning with a double-edged razor blade. Cell division in drupes occurs in a radial
direction as the mesocarp develops (Tukey and Young, 1939). In addition, this plane of
measurement is equivalent to the dimension that commercially produced sweet cherries
are measured for size before sale. From bloom until endocarp hardening, tissue sections
were cut through the entire diameter of the fruit; from endocarp hardening on, tissue
sections were cut from the skin to the endocarp wall, consisting only of exocarp and
mesocarp tissue. After sectioning, mesocarp tissue was rehydrated with distilled water
before staining. For sections created aﬁer endocarp hardening, the pit weight and

diameter were measured.

Microscopic analysis of mesocarp tissue

Following tissue rehydration, the sections were stained for at least 24 h in a dilute
1:20 solution of 1 mg/ml acridine orange. Preliminary tests indicated that acridine orange
was suitable for staining mesocarp cell walls at all stages of development. Acridine

orange is a metachromatic ﬂuorescent dye that is excited at 500 nm and emits with peaks

35

in both green (526 nm) and red (650 nm) ranges (Lillie, 1977). After this staining period,
tissue sections were brieﬂy rinsed again in distilled water, and ﬂesh slide mounts using
distilled water were prepared immediately before microscopic evaluation. Unstained
samples of mesocarp tissue from each developmental stage were observed using the same
microscope parameters to ensure that the ﬂuorescence signal was not due to
autoﬂuorescence.

All microscopy was performed at the MSU Center for Advanced Microscopy,
using a Zeiss laser scanning confocal microscope and software (Zeiss LSM Pascal, Jena,
Germany). The following microscope parameters were used to collect ﬂuorescent
images: 488 nm argon laser line, 505-530 nm band pass ﬁlter, and 650 nm long pass
ﬁlter. Both 10x and 20x objectives were used for different stages of development.
Pinhole apertures of 70 um and 84 um were used with the 10x and 20x objectives,
respectively. For all but the earliest developmental date, multiple ﬁeld of view images
were necessary to scan through the entire mesocarp section. Images were captured
digitally as Tagged Image Format ﬁles with no compression using the integrated

microscope software, and stored on compact disc for later image analysis.

Image analysis

Individual ﬁeld of view images were ﬁrst aligned together into one composite
image using Adobe Photoshop 6.0 soﬁware (Adobe Systems Inc., San Jose, Calif).
Composite images were then calibrated to a deﬁned dimension using Sigma Scan Pro 5.0
software (Systat, Richmond, Calif). Once calibrated, the trace measurement function in

Sigma Scan Pro was used to draw and measure a line the length of the mesocarp section.

36

For each image, all of the cells touching the line were counted and this measurement was
subsequently used in all analyses, similar to that of Yamaguchi et al. (2002a, 2002b,
2004). Cell length in the sections was calculated by dividing the total mesocarp section

length by the number of cells counted in the same length.

Statistical analysis

Data were subjected to analysis of variance using SAS general linear model
procedure with the variance for subsamples used as the error term (SAS Institute, Inc.,
Cary, N.C.). SAS correlation procedure was used when appropriate to determine the
Pearson correlation coefﬁcient between related measures. All means were separated

using Tukey’s HSD or Fisher’s LSD.

RESULTS
Comparison of fruit and pit measurements among ﬁve sweet cherry selections

In 2003, fruit and pit measurements were made from ‘NY 54’ and ‘Selah’ as the
extremes in the range of fruit size diversity, and ‘EF’ as a domesticated selection (Table
1, Fig. 2). Mean ﬁ'uit weights and diameters were signiﬁcantly different (P < 0.05), with
‘EF’ exhibiting intermediate values. In 2004, ﬁ'uit ﬁom ‘Bing’ and ‘Regina’ also were
sampled. As in 2003, ‘Selah’, ‘EF’, and ‘NY 54’ exhibited signiﬁcant (P < 0.05)
differences in fruit weight and diameter; however, ‘Bing’ and ‘Regina’ had similar values
that were signiﬁcantly larger than ‘EF’ but signiﬁcantly less than ‘Selah’ (Table 1). This
is consistent with ‘Bing’ and ‘Regina’ representing a fruit size improvement over that

achieved through domestication.

37

In both years, the numbers of mesocarp cells counted in radial sections of ‘Selah’,
‘EF’, and ‘NY 54’ (Fig. 3) were signiﬁcantly different (P < 0.05) with ‘Selah’ having
~3x the number of mesocarp cells as ‘NY 54’ (Table 1). As with fruit weight and
diameter, the number of mesocarp cells for ‘EF’ was intermediate to that of ‘NY 54’ and
‘Selah’. In 2004, mean radial cell number for ‘Bing’ and ‘Regina’ were 48.3 and 43.8,
respectively (Table 1). As with ﬁ'uit weight and diameter, this cell number value for
‘Bing’ was signiﬁcantly larger than the value for ‘EF’, but signiﬁcantly smaller than the
value for ‘Selah’. Although the cell number for ‘Regina’ was statistically similar to that
of ‘Bing’, it was not signiﬁcantly different from that of ‘EF’.

Variation for cell length was less signiﬁcant than that for cell number, as the
selections fell into only two groups (Table 1). In both years, ‘EF’ had the longest
calculated cell lengths, which were similar to those from ‘Bing’ and ‘Regina’.
Interestingly, although ‘Selah’ ﬁ'uit were the largest overall and had the greatest number
of cells, the calculated cell length was not statistically different from ‘NY 54’ and was
shorter than that for ‘EF’, ‘Bing’, and ‘Regina’.

‘NY 54’ exhibited a signiﬁcantly smaller pit weight and diameter compared to the
other selections. The mean pit diameters for the four remaining selections were not
signiﬁcantly different. Pit weight did vary; however, this variation is more likely due to
seed development as early maturing selections have less developed seeds (Fogle, 1975).
The maturity order for these cultivars from earliest to latest, is ‘EF’, ‘Bing’ ~ ‘Selah’, and
‘Regina’.

Correlations were calculated for the fruit and cell measurements for the ﬁve sweet

cherry selections evaluated in 2004. Highly signiﬁcant positive relationships were

38

identiﬁed between cell number and both fruit diameter and fruit weight (r2 = 0.59 and
0.72, respectively, P < 0.001) (Fig. 4). However, cell length was not signiﬁcantly
correlated with either fruit diameter or fruit weight (Fig. 4). This supports the conclusion
that cell number and not cell size is the major cellular component contributing to the

genotypic differences in fruit size.

Stability of mesocarp cell number and size for selections subjected to diﬁfzrent climatic
and cultural conditions

To determine the stability of the cell number measurements for the ﬁve selections,
data was collected for three years (2003-05) and two locations (WSU-IAREC and MSU-
CHES). Analysis of variance indicated no signiﬁcant year X location interaction between
cell numbers for each cultivar. Likewise, no signiﬁcant within cultivar difference was
identiﬁed between the two locations. However, a signiﬁcant (P < 0.001) difference was
identiﬁed within the year main effect of the model. Therefore, the potential interaction
and location variance was pooled, and the analysis of variance was performed to identify
the year difference by mean separation. A signiﬁcant cell number difference for ‘EF’
was identiﬁed in 2003 (Table 2). In that year, within cultivar cell number counts
averaged higher than other years and locations. However, for all other cultivars, no
signiﬁcant within cultivar difference was found for cell number. This indicates that cell
division is under strong genetic control, and in general, is unaffected by different climatic
conditions.

To further examine the stability of mesocarp cell number, an analysis of within-

cultivar variation was done. Because sweet cherries are clonally propagated and the

39

measured area of the fruit is solely maternal tissue, a random sample of ﬁ'uit from the
same cultivar will be genetically identical. However, variation in fresh fruit size within
the same tree occurs due to physiological variables such as crop load and ﬁxed carbon
availability. In 2004, pre-bloom crop load adjustment was performed on ‘Bing’ and
‘Regina’ trees at WSU-IAREC to create differences in available carbon allocated to
individual fruit. Crop load adjustment was performed by hand-thinning whole trees to
one fruit bud per ﬁ'uiting spur. Similar treatments have been shown to result in
signiﬁcant increases in overall fruit size (Land and Ophardt, 2000; Whiting and Lang,
2004). Un-thinned control trees were used for comparison.

Crop load adjustment resulted in a signiﬁcant increase (P > 0.001) in overall fruit
weight and diameter for ‘Bing’ in 2004. However, due to low initial fruit set, the same
treatment on ‘Regina’ trees did not result in signiﬁcant fruit size differences in random
samples. Therefore, individual fruit from ‘Regina’ trees were weighed and separate pools
of small and large-sized fruit were created. The difference between the pools was at least
2 g, similar to the average weight differential for the ‘Bing’ treatments. In 2005, bud
. thinning treatments were applied to ‘Bing’, ‘Regina’, and ‘Selah’ trees. However, spring
frost damage resulted in non-signiﬁcant differences between treatments; therefore,
selected pools of different-sized fruit were used again for comparison.

In both 2004 and 2005, the mean fresh fruit weights and diameters for the small
versus large ﬁ'uit within each of the three cultivars were signiﬁcantly different (P <
0.001) (Table 3). Mesocarp cell numbers for a given cultivar were not signiﬁcantly
different for the large or small fruit samples. However, the calculated cell lengths were

signiﬁcantly different (P < 0.05) between all comparisons except ‘Selah’ in 2005. For

40

‘Bing’ and ‘Regina’, the larger fruit had signiﬁcantly longer mesocarp cell lengths
compared to small fruit. These results indicate the differences in fruit weights and
diameters between the large and small fruit of ‘Bing’, ‘Regina’, and ‘Selah’ were not due
to differences in mesocarp cell ntunber. Instead, fruit size increases within ‘Bing’ and
‘Regina’ were due to increases in cell length. Although cell lengths were not
signiﬁcantly different between large and small fruit from ‘Selah’, the trend of larger ﬁ'uit
size correlated with longer cell lengths was evident. Interestingly, the mean mesocarp
cell length for the ~13 g ‘Selah’ fruit were similar to those of the very small fruit from
‘NY 54’, and this cell length was unaffected by those environmental or cultural
conditions that resulted in larger fruit size. ‘Selah’ fruit may have been judged to be at
harvest maturity prior to the end of the fruit developmental period, resulting in
incomplete cell expansion.

Pit weight was signiﬁcantly different between the ‘Bing’ comparisons in both
2004 and 2005 (P < 0.001, P < 0.05, respectively), and ‘Regina’ in 2005 (P < 0.001)
(Table 3). Pit diameter was signiﬁcantly different between treatments for ‘Bing’ in 2004
(P < 0.001), and ‘Regina’ and ‘Selah’ in 2005 (P < 0.05, P < 0.0001, respectively). In
these cases, larger fruit had heavier pits with increased diameters. However, when the
percentage of total fruit diameter due to pit diameter was analyzed, it was apparent that
pits in small sized fruit contributed a greater percentage to the total fruit diameter (P <
0.01 for all cultivars), and thus smaller mesocarp ﬂesh diameter. In addition, these
differences were not consistent among cultivars and years. For example, the large
‘Regina’ fruit in 2004 and large ‘Selah’ ﬁuit exhibited mean diameters of 27.7 mm and

30 mm, respectively, yet their mean pit diameters were nearly equivalent, at 8.3 mm and

41

8.1 mm, respectively. This suggests that genetic increases in fruit size can occur without

an associated increase in pit diameter.

Duration of mesocarp cell division for the ﬁve sweet cherry selections

Examination of ﬂesh mesocarp sections of ‘Selah’, ‘EF’, ‘NY 54’, ‘Bing’ and
‘Regina’ at different stages of fruit development in 2003 and 2004 conﬁrmed the classic
general stone ﬁ'uit growth pattern (Chalmers and van den Ende, 1975; Coombe, 1976;
Gage and Stutte, 1991; Lilleland and Newsome, 1934; Nitsch, 1953; Tukey and Young,
1939). Mesocarp cell division occurred during the period from bloom until endocarp
ligrriﬁcation; generally, only cell enlargement occurred aﬁer endocarp hardening (Fig. 5,
Table 4). In 2003, samples taken from ‘EF’ ﬂowers just aﬁer opening had signiﬁcantly
fewer (P < 0.05) mesocarp cells than ‘Selah or ‘NY 54’ sampled at the same stage. In
2004, ‘NY 54’ and ‘EF’ had the fewest cells at bloom while ‘Selah’ had the most (Table
4). In 2004, the ranking of mean cell numbers at bloom was equivalent to the ranking
based upon mature fruit size (e.g., large to small, ‘Selah’, ‘Regina’ ~ ‘Bing’, ‘EF’, ‘NY
54’), although this was not the case in 2003.

In 2004, the GDD accumulation at WSU-IAREC until the total number of
mesocarp cells was reached ranged from 65 (‘Regina’) to 237 (‘Selah’). ‘Bing’, ‘EF’,
and ‘NY 54’ each took slightly less than 95 GDD to reach the total mesocarp cell
numbers. At MSU-CHES, the GDD accumulation until total cell numbers were reached
ranged from 22 (‘NY 54’) to 171 (‘Regina’). At WSU-IAREC, the GDD accumulation
until total cell numbers were reached ranged from 21 (‘NY 54’) to 139 (‘Selah’). More

frequent sampling, every three to ﬁve days as done in 2005, provided the ability to

42

calculate a cell division rate for each cultivar (Table 5). Because destructive sampling
was needed to measure cell numbers at each date, the differential between the mean
numbers of mesocarp cells at bloom was used to estimate the number of new cells added.
In contrast to the difference in duration of cell division for each cultivar, the calculated
cell division rate was only signiﬁcantly different (P < 0.05) for ‘NY 54’. However, the
calculated low rate of cell division in ‘NY 54’ is not biologically relevant since ‘NY 54’
essentially has its full complement of mesocarp cells at bloom time. Therefore, the
increase in mesocarp cell number associated with increased fruit size in the sweet cherry
selections was due to an increase in the duration of the cell division period, not more

rapid cell division.

DISCUSSION

Fruit size differences among the sweet cherry selections were determined
primarily by differences in mesocarp cell numbers. ‘NY 54’ essentially had its full
complement of mesocarp cells at bloom, whereas the larger-ﬁ'uited cultivars underwent
signiﬁcant cell number increases between bloom and endocarp hardening. This is similar
to results reported previously for sweet cherry (Yamaguchi et al., 2004), with the degree
of correlation between whole ﬁ'uit size measurements and mesocarp cell number
estimates in similar ranges. However, in the present study, the correlation between cell
size and overall fruit size was not signiﬁcant in a comparison between cultivars, while
Yamaguchi et al. (2004) report higher correlations between cell size and fruit size

between cultivars. This may be a result of the limited number of cultivars examined in

43

this study. For example, the small cell size combined with large ﬁ'uit size of ‘Selah’, and
large cell size in the smaller-fruited ‘EF’ may be unique among cherry genotypes.

An increase in the number of mesocarp cells corresponding to increased fruit size
also has been reported for comparisons between small and large-fruited peach cultivars
(Scorza et al., 1991; Yamaguchi et al., 2002a, 2002b). Collectively, these reports and the
present study indicate that the gene(s) involved in mesocarp cell number proliferation are
keys to understanding the genetic potential for increased ﬁ'uit size in sweet cherry.
Interestingly, these reports for Prunus species are similar to that of the model fruit plant,
tomato. In tomato, the gene underlying a major QTL (ﬁv2. 2) contributing to fruit size
difference between wild and cultivated species has been cloned and shown to inﬂuence
cell division and consequently fruit size (Frary et al., 2000; Nesbitt and Tanksley, 2002).

It is noteworthy, however, that ‘NY 54’ and ‘Selah’ mean mesocarp cell lengths
were similar, but were shorter than those for ‘EF’, ‘Bing’, and ‘Regina’. This suggests
that it might be possible to further genetically increase fruit size in sweet cherry by
combining the increased number of cell divisions in ‘Selah’ with the increased cell size
exhibited by ‘EF’, ‘Bing’, and ‘Regina’.

This study was undertaken as a component of a larger research plan to identify the
major QTL and genes involved in sweet cherry fruit size. ‘NY 54’ and ‘EF’ were
included in these analyses because they are the parents of a genetic linkage mapping
population developed at MSU. Therefore, histological differences between ‘NY 54’ and
‘EF’ mesocarp cells have the potential to segregate among progeny from the linkage
mapping population and can be used in a future QTL analysis. Furthermore, ‘NY 54’ is a

true wild example of P. avium (R.L. Andersen, pers. com.) and ‘EF’ can be considered

44

an early domesticate of sweet cherry in Europe. Hence, differences in cellular
development identiﬁed between the two are a direct result of early selection and
domestication by farmers. Because larger fruit size is considered one of the hallmarks of
early domestication (J anick, 2004), these differences are important to document from a
plant breeding standpoint. With many potentially valuable genes for traits such as pest,
disease, and stress tolerance available in wild members of the species and relatives, an
understanding of the traits originally selected during domestication will certainly speed
the recovery of suitable fruit size following gene introgression.

‘Selah’ was included in the study because it falls at the opposite end of the fruit
size spectrum from ‘NY 54’. ‘Selah’ is one of the largest-sized sweet cherry cultivars
released from the Washington State University stone fruit breeding program. ‘Bing’ and
‘Regina’ were also included in these experiments as production cultivars because they
either currently dominate U.S. production (‘Bing’), or provide valuable niche-market
alternatives (‘Regina’). Together with the goal of identifying useful genetic variation for
fruit size, an unanticipated but potentially valuable result of these experiments was to
clarify key stages of developmental activity relating to sweet cherry fruit growth that
horticulturists and physiologists may exploit to better maximize ﬁ'uit size of current
cultivars.

For a quantitative trait, such as ﬁ'uit size, to be efﬁciently selected in a breeding
program, there must be a high level of genetic variation coupled with low environmental
and sampling variation. In our study, mesocarp cell number exhibited an extraordinary
stability when subjected to different climatic conditions and cultural practices. During

the experimental time period, no differences for mature cell number were identiﬁed

45

between Washington and Michigan. The only within-cultivar difference in cell number
was identiﬁed for ‘EF’ in 2003. In that year, ‘EF’ had signiﬁcantly more cells than in
later samples.

In this study, although fruit size differed signiﬁcantly between pools of fruit from
the same cultivar, cell number was not different (Table 3). Consequently, calculated cell
size differences were apparent. Environmental effects on ﬁ'uit size appear to be
manifested through an increase or decrease in mean cell size. The magnitude of size
difference between pools of fruit from the same cultivar was similar to differences in
mean fruit size between cultivars. Both cell number and cell length were not
signiﬁcantly different between ‘Selah’ ﬁ'uit that averaged nearly ﬁve mm different in
diameter. However, these results are based on one year of data, and additional samples
will be necessary to determine whether the trend of increased cell size among larger
‘Selah’ ﬁ'uit is signiﬁcant. It is possible that Selah ﬁ'uit were judged to be at harvest
maturity based on color and taste before they were actually close to physiological
maturity. Pit diameter differences commonly were measured between different sized
ﬁ'uit from the same cultivar. The data suggest that larger diameter fruit have larger
diameter pits. However, when the percentage of total fruit diameter due to pit diameter
was analyzed, it was apparent that pits in small sized fruit contributed a greater
percentage to the total fruit diameter.

Differences in fruit size within the same cultivar without a corresponding increase
in cell number have been reported for other fruit species. In olive (Olea europaea L.),
development of fruit under water deﬁcit conditions, resulted in overall fresh fruit size

differences but no signiﬁcant difference in cell number in the fruit (Costagli et al., 2003;

46

Rapoport et al., 2004). However, Jackson and Coombe (1966) report that within and
between tree variation in fruit size for a single apricot (Prunus armeniaca L.) cultivar
was due to both cell number and size differences. Although a general tendency was
observed toward smaller cell sizes in small peach ﬁ'uit (Bradley, 1959), the correlation
between cell size and mesocarp size was not signiﬁcant, and the author concluded that
cell number differences must be involved.

In apple (Malus domestica Borkh.), conﬂicting evidence has been reported
concerning potential environmental effects on cell size and number in the same genotype.
Temperature differences applied to potted apple trees during the cell division period after
bloom resulted in larger cells but no increase in the number of cells (Atkinson, et al.,
2001). In contrast, Warrington et a1. (1999) found that early season temperature
differences affect cell division in apple cultivars. Bain and Robertson (1951) report that,
from a single cultivar, different fruit sizes were related to increased cell numbers in the
cortex. The biennial bearing nature of certain cultivars has been shown to decrease fruit
cell number in the year after heavy cropping (Bergh, 1985). Similarly, a higher chilling
unit accumulation resulted in a greater number of cells (Grebeye and Bergh, 2000).
Thinning of ﬂowers or fruit on apple trees of a single cultivar has repeatedly been shown
to increase cell number and/or cell size (Bain and Robertson, 1951; Bergh, 1990; Denne,
1960; Goffrnet et al., 1995; Westwood et al., 1967). The rootstock effect on apple cell
size or number in a single cultivar has been attributed either to cell size difference (Al-
Hinai and Roper, 2004) or number (Hirst and Flowers, 2000) differences in the fruit.
Although on the surface, experiments with different rootstocks can be considered as

potential environmental variation, the fact remains that the rootstock and scion are two

47

genetically different entities, and the extent to which rootstock-produced proteins may
inﬂuence scion growth and development has yet to be explored thoroughly. Apple fruit
ﬂesh results from cortical or accessory tissue in the ﬂower, while sweet cherry ﬂesh
begins as the true ovary wall. It is possible that the cortical region that comprises apple
fruit ﬂesh is under different genetic control than sweet cherry and Prunus in general.

The data from the present study indicate that cell number and not cell size in
sweet cherry has the greatest genotypic inﬂuence on ﬁ'uit size. Within a genotype, cell
number was constant while variations in fruit size resulted ﬁom increasing mesocarp cell
size. Together, these results indicate that mesocarp cell number is not greatly affected by
environmental variation, and is therefore an ideal trait for which to select in sweet cherry
breeding programs to increase ﬁ'uit size.

In Prunus, mesocarp cell division occurs during the period between anthesis and
endocarp hardening (Coombe, 1976; Gage and Stutte, 1991; Nitsch, 1953; Tukey and
Young, 1939). In this study, fruit from ‘Selah’, ‘EF’, ‘NY 54’, ‘Bing’, and ‘Regina’
followed this general growth pattern (Fig. 5, Table 4). Once endocarp hardening began,
the ﬁnal mesocarp cell number had nearly been reached, and difference in cell number
between cultivars was signiﬁcant (Table 4). At bloom, the endocarp cells are included in
radial sections through the fruit diameter (Fig. 5 A), but they are distinguishable from the
mesocarp parenchyma, as has been noted for peach (Masia et al., 1992), and were not
included in the cell number count at this developmental stage. The number of mesocarp
cells at bloom varied by year, and there was no clear relationship between the number of
cells at bloom and the ﬁnal cell number at harvest maturity (Table 4). This could be due

to the timing of sample collection at bloom. Samples from each different cultivar were

48

collected when the entire tree was judged to be at 50% full bloom by visual estimation.
At that point, only ﬂowers that were fully open but with anthers not yet dehiscent were
collected. Although this was deemed the most efﬁcient way to synchronize samples from
cultivars with divergent bloom times, the phenology estimate may not have been
accurate. Alternatively, the difference in cell numbers at bloom may be a cultivar-
speciﬁc trait that reﬂects differences in time or extent of ﬂoral differentiation. Difference
in cell number between small and large fruit size peach cultivars, before and at bloom,
has been previously documented (Scorza et al., 1991). The extent to which cell numbers
can be increased prior to mesocarp cell division after bloom warrants further
investigation in sweet cherry. However, the fact remains that to reach the ﬁnal cell
numbers observed in this study, mesocarp cells in Selah had to divide nearly twice as

often in the period between bloom and harvest as the other cultivars examined.

CONCLUSIONS

These experiments indicate mesocarp cell number is the major genetic
determinant of fruit size in sweet cherry. The number of mesocarp cell layers present
(from the endocarp to the exocarp) was remarkably stable in the three years of this study
and at two different locations with disparate environmental conditions. Cell number was
not affected by environmental or cultural variation, as illustrated by analyzing fruit of
different sizes from the same genotype. Collectively, these data suggest that cell number
difference would be an ideal trait to identify using QTL analysis. The low environmental

variance also would be advantageous for selection in a breeding program.

49

The variation in cell number between genotypes at bloom remains to be ﬁrlly
analyzed. Carbohydrate reserves are important for early season growth in sweet cherry
(Ayala, 2004), and reduction in stored carbohydrates has been shown to reduce cell
division in the following season in Japanese pear (Pyrus serotina) (Toyarna and Hayashi,
1956). Although the majority of cell division occurs in the period between bloom and
endocarp scleriﬁcation, horticultural treatments applied in the same season that ﬂower
buds are initiated and differentiate may increase the number of mesocarp cells prior to the

onset of cell division at bloom.

50

LITERATURE CITED

Al-Hinai, Y.K. and TR. Roper. 2004. Rootstock effects on growth, cell number, and
cell size of ‘Gala’ apples. J. Amer. Soc. Hort. Sci. 129237-41.

Atkinson, C.J., L. Taylor and G. Kingswell. 2001. The importance of temperature

differences, directly after anthesis, in determining growth and cellular
development of Malus fruits. J. Hortic. Sci. Biotech. 76:721-731.

Ayala, M. 2004. Carbon partitioning in sweet cherry (Prunus avium L.) on dwarﬁng
precocious rootstocks during fruit development. Ph.D. Dissertation, Michigan
State University.

Bain, J .M. and RN. Robertson. 1951. The physiology of growth in apple fruits. Aust. J.
Sci. Res. 4:75-91.

Bergh, O. 1985. Effect of the previous season’s crop on cortical cell number of Malus
domestica cv. Starking apple ﬂower primordia, ﬂowers and fruit. S. Afr. J. Plant
Soil 2:191-196.

Bergh, O. 1990. Effect of time of hand-thinning on apple ﬁ'uit size. S. Afr. J. Plant Soil
7:1-10.

Bradley, M.V. 1959. Mean cell size in the mesocarp of mature peaches of different
sizes. Proc. Amer. Soc. Hort. Sci. 73:120-124.

Chalmers, D.I. and B. van den Ende. 1975. A reappraisal of the growth and
development of peach fruit. Aust. J. Plant Physiol. 22623-634.

Choi, C., and F. Kappel. 2004. Inbreeding, coancestry, and founding clones of sweet
cherries from North America. J. Amer. Soc. Hort. Sci. 129:535-543.

Coombe, B.G. 1976. The development of ﬂeshy fruits. Ann. Rev. Plant Physiol.
27:507-528.

Costagli, G., R. Gucci and HF. Rapoport. 2003. Growth and development of fruits of
olive ‘Frantoio’ under irrigated and rainfed conditions. J. Hortic. Sci. Biotech.
78:119-124.

Denne, MP. 1960. The growth of apple fruitlets, and the effect of early thinning on fruit
development. Ann. Bot. 24:397-406.

Doganlar, S., A. Frary, M.-C. Daunay, R.N. Lester, and SD. Tanksley. 2002.

Conservation of gene function in the Solanaceae as revealed by comparative
mapping of domestication traits in eggplant. Genetics 161:1713-1726.

51

Esau, K. 1977. Anatomy of Seed Plants. 2nd ed. John Wiley and Sons, Inc., New York,
NY.

Fogle, H.W. 1975. Cherries. In: Janick, J. and J.N. Moore (eds) Advances in Fruit
Breeding. Purdue University Press, West Lafayette, Ind, pp. 348-366.

F rary, A. T.C. Nesbitt, A. F rary, S. Grandillo, E. van der Knaap, B. Cong, J. Liu, J.
Meller, R. Elber, K.B. Alpert and SD. Tanksley. 2000. fw2.2: A quantitative
trait locus key to the evolution of tomato fruit size. Science 289:85-88.

Gage, J. and G. Stutte. 1991. Developmental indeces of peach: an anatomical
framework. HortScience 26:459-463.

Gofﬁnet, M.C., T.L. Robinson and AN. Lakso. 1995. A comparison of ‘Empire’ apple
fruit size and anatomy in unthinned and hand-thinned trees. J. Hortic. Sci.
70:375-3 87.

Grandillo, 8., HM. Ku, and SD. Tanksley. 1999. Identifying the loci responsible for
natural variation in fruit size and shape in tomato. Theor. Appl. Genet. 99:978-
987.

Grebeye, E. and O. Bergh. 2000. The effect of winter chilling on cell division and
multiplication pre-anthesis and thus on ﬁnal fruit size of Royal Gala apples in
South Aﬁica. Acta Hort. 519:113-120.

Hirst, RM. and RR. Flowers. 2000. Rootstock effects on grth and cell size of ‘Gala’
apple fruit. Acta Hort. 517:189-194.

Jackson, D.I. and B.G. Coombe. 1966. The growth of apricot fruit I. Morphological
changes during development and the effects of various tree factors. Aust. J.
Agric. Res. 17:465-477.

Janick, J. 2004. Genetic alterations associated with the origins of fruit culture. Acta
Hort. 663:683-691.

Lang, GA. and DR. Ophardt. 2000. Intensive crop regulation strategies in sweet
cherries. Acta Hort. 514:227-233.

Lilleland, O. and L. Newsome. 1934. A growth study of the cherry fruit. Proc. Amer.
Soc. Hort. Sci. 32:291-299.

Lillie, RD. 1977. HI. Conn’s Biological Stains. 9th ed. Williams and Wilkins
Company, Baltimore, MD.

52

Masia, A., A. Zanchin, N. Rascio and A. Ramina. 1992. Some biochemical and
ultrastructural aspects of peach fruit development. J. Amer. Soc. Hort. Sci.
117:808-815.

Nesbitt, T.C. and SD. Tanksley. 2002. Comparative sequencing in the genus
Lycopersicon: implications for the evolution of ﬁ'uit size in the domestication of
cultivated tomatoes. Genetics 162:365-379.

Nitsch, J .P. 1953. The physiology of mu growth. Ann. Rev. Plant Physiol. 42199-236.

Paterson, AH. 2002. What has QTL mapping taught us about plant domestication?
New Phytologist 154:591-608.

Rapoport, HE, G. Costagli and R. Gucci. 2004. The effect of water deﬁcit during early
fruit development on olive fruit morphogenesis. J. Amer. Soc. Hort. Sci.
129:121-127.

Ruzin, SE. 1999. Plant Microtechnique and Microscopy. Oxford University Press,
New York, NY.

Scorza, R., L.G. May, B. Purnell and B. Upchurch. 1991. Differences in number and
area of mesocarp cells between small- and large-fruited peach cultivars. J. Amer.
Soc. Hort. Sci. 116:861-864.

Tanksley, SD. 2004. The genetic, developmental, and molecular bases of fruit size and
shape variation in tomato. Plant Cell 16:Sl81-Sl89.

Toyama, S. and S. Hayashi. 1956. Studies on the ﬁ'uit development of Japanese pears I.
On the ﬂesh cell-division, cell-enlargement and the relation between ﬂesh cell-
size and fruit size in some varieties. J. Hort. Ass. Japan 25:274-278.

Tukey, H.B. and J .0. Young. 1939. Histological study of the developing fruit of the
sour cherry. Bot. Gaz. 100:723-749.

Warrington, I.J., T.A. Fulton, E.A. Halligan and H.N. de Silva. 1999. Apple ﬁ'uit growth
and matruity are affected by early season temperatures. J. Amer. Soc. Hort. Sci.
124:468-477.

Westwood, M.N., L.P. Batjer and HS. Billingsley. 1967. Cell size, number, and fruit
density of apples as related to fruit size, position in cluster, and thinning method.
Proc. Amer. Soc. Hort. Sci. 91 :51-62.

Whiting, M.D., G. Lang and D. Ophardt. 2005 Rootstock and training system affect
cherry growth, yield, and fruit quality. HortScience 40:582-586.

53

Whiting, M.D., D. Ophardt, and J .R. McFerson. 2006. Chemical blossom thinners vary
in their effect on sweet cherry fruit set, yield, fruit quality, and crop value.
HortTechnology 16:66-70.

Whiting, MD. and GA Lang. 2004. ‘Bing’ sweet cherry on the dwarﬁng rootstock
‘Gisela 5’: thinning affects fruit quality and vegetative growth but not net CO2
exchange. J. Amer. Soc. Hort. Sci. 129:407-415.

Yamaguchi, M., T. Haji, M. Miyake and H. Yaegaki. 2002a. Studies on the varietal
differences and yearly deviation of mesocarp cell numbers and lengths and fruit
weight among commercial peach [Prunus persica (L.) Batsch] cultivars and
selections, wild types, and their hybrids. J. Japan. Soc. Hort. Sci. 71:459-466.

Yamaguchi, M., T. Haji, M. Miyake and H. Yaegaki. 2002b. Varietal differences in cell

division and enlargement periods during peach (Prunus persica Batsch) fruit
development. J. Japan. Soc. Hort. Sci. 71 :155-163.

Yamaguchi, M. I. Sato, K. Takase, A. Watanabe and M. Ishiguro. 2004. Differences and

yearly variation in number and size of mesocarp cells in sweet cherry (Prunus
avium L.) cultivars and related species. J. Japan. Soc. Hort. Sci. 73:12-18.

54

TABLES AND FIGURES

Figure 1. Diagram illustrating the area of sweet cherry fruit samples sectioned for
microscopic analysis. Radial sections were prepared from the thickest part of the ﬁuit
mesocarp, halfway between the point of stem attachment and the stylar scar, and 90

degrees from the suture line.

 

stylar scar

55

Figure 2. Images of fruit from ‘Selah’ (A), ‘Emperor Francis’ (B), and ‘NY 54’ (C)

sweet cherries, illustrating the variation in ﬁ'uit size and mesocarp thickness.

 

56

.23 v m as am: Maud—E. .3 3838 £53, Suwanee. SEEN

 

 

 

 

 

 

 

 

.32 25.? mod mwmd £.mm nus -- -- .. -- -- .. «Emma
£2 £2? 3.x. anemd awém no.5 -- : -- -- : -- 35
«ma 3.3 nod ow _ .o 3.2 32 3: 050m nmd nmmd o _ .m. 3.. vm>z
snow 3. 3 «ms 936 omNN and £3 3.: 3.5 m _ vd no. _N n56 mm
33 3.3. «ms amomd «ER «2: «a: «max 3: mwmd «mom NEWS :Eom
A83 ﬁwﬁ: .o: AEEV @ AEEV @ A83 598. .o: 2E5 va AEEV @ 32:30
:8 =8 d6 65 .ﬁc .S, =8 :8 St .3» d6 c3
Become—2 32m 83802 «E zen—
voom mocm

 

5:32: .822 a 023m? :5: 8E2“. .85 .aswom. Ea ..wam. ..E >2. 5.59 22.5 826m. ..ﬁam.

com 3582386 05m can 5:03 33.. Ba .5985: :8 98808 EB .35 an .oﬁm :Pc 5.0% 2055 :38 mo acmtaQEoU ._ 03a...

57

 

.61 com u :5 28m .cowﬁcomoa 8m 85¢ :03 8 023.2 @2QO can Ara—338%“: ..Vm >2. can ...mm. ..mﬁmom. ..wﬁm. nae—om.

co.“ m was .2 .2 .2 .3 H 5 momma: ﬁesta—om 288838 we???“ wcmswzu .3 @833 803 women: .953 .8338 33m «a women:

@5885 £838 Q .492. as .2: ram; 285 825. .Q annex. .9: .35. .3 22%. co anagram .m 053

58

.53 59.... ...o

 

 

 

 

 

 

 

£88. .3: .2: .2352 ...o

 

 

 

   
  

 

 

 

 

    

 

o8 m8 o8 8. 8. 8. 8. 8 8 8 o 2: 8 8 8 8 8 8 8 8 8 o
_ r p b . F p h p o p b . L P p _ p b o
0‘ O O T N 0‘ O 1 N
f v m. , v m.
o I o u y o “a
O ODD
00 #0 O T o m ' Q m
o o 0.. aoadll row W 000 rocdvl {or W
33...: , a. M 52...”. , N. m
«33:88....» , 3 38.9.3831 , 3
2 2
.5... 59.... ...o .588 3...: .2: .382 ..oo
08 8N o8 ...: 8. 8. 8. 8 8 8 o 2: 8 8 8 8 8 8 8 8 2 o
p . 8 p p _ p b b o F p b b _ p p b . f o
f m ... m 4..
o o 0 I or “M." 1 Dr W
a , 9 m I ...; m
s -. I 8 w , 8 w
III-iv! o .98. ouvdum 1mm m ooo Sadvm 1mm m
83.x. , 8 w. Boa-«m - 8 w:
£88333...» , 8 ( 353x885.» - 8 (
8 8

 

 

 

 

 

<

59

.83 2: :o @8865 mm acmtanQ :08 8m 02? 238$ 98 CV Homo—boon. noun—oboe
season 2.: 8w Bum—=23 mos—gi 2E. .voom 5 3.5988 982:3 85:0 .826 =a .8.“ Amv £303 98 “225% tam smut £8— 28

£95— =oo 8882: can A<v £903 28 88.5% E5 52% Son 28 858:: :3 9382: 50253 coca—oboe 80:5 .v 0.5me

Table 2. Comparison of mean cell numbers at maturity for ‘Selah’, ‘Emperor Francis’

(‘EF’), ‘NY 54’, ‘Bing’, and ‘Regina’ sweet cherry fruit from 2003-2005 and at two

locations (WSU-IAREC and MSU-CHES).

 

Cell no. per radial section

 

 

Year and Location Selah EFz NY54 Bing Regina
WA 2003 83.2 a 47.4 a 26.7 a -- --
WA 2004 78.8 a 41.4 b 28.6 a 48.3 a 43.8 a
WA 2005y 78.6 a 38.4 b 28.0 a 47.4 a 45.2 a

M12004 -- 40.0 b 28.6 a 44.8 a 47.6 a
MI 2005y -- 40.6 b 29.0 a 46.8 a 43.8 a

 

zMean separation within columns by Tukey’s HSD at P < 0.05.

yCell number for EF and NY54 in 2005 at WSU-IAREC and all samples from 2005 at
MSU-CHES were determined after endocarp hardening.

60

525882 .Sod
.86 .36 v m 8 858$:me ...»...3... 4:82.388 :89: ”Om: 9:23..— .3 8828 :m 9:258: :28: :8 :othQom :82N

 

 

 

 

 

 

mamN~ mcmdh «......NS mzavd .....iodm .....iwd l .. l .. .. l .3» >90—
Exam
m:\.m_ 83K 1.... :w 2mm: 15.3 .1186 _ -- -- -- -- -- -- :3 :wE
Exam
*0: Soﬁe ...ms .....de ...:mém find ...mm: m:w.mv 3o.» 836 1.... ..mm «1.5.5 :3 >5.
:Ewox
...o-m mcwév *N.w ......vod ***w.wm 3.2.12 ...v; 2:06: 25.x waved ......Rmhm ......«md— .39 :wE
«Ewom
...mﬁ 23:. 2mg. ..de ......oém tins .. :2 2mg. 2%.: 3.0%: ......w.v~ tabs 43%,:—
Em
*wom mcodv 25.x. *omd ....Ibﬁm ......*m.: *©o_ maﬁa: “25.x. ......hmd ...“...Lucm .1236 65%—mm:
Em
9:: .2. 2:5 @ ASE 3 :8: .8 3:5 8: :55 3 83:5
8:2 :8 =8 .8 .s, .8 .E .88. =8 =8 .8 .8 .8 ..E
98082 «E :3: 93082 .E E2:
moon voom

 

:2 3:062:82: 88 :8 90:08 3:2 .83 23:5: :8 9380:: :5 .87. a: .88 :9: 52.: 30:3 :88 mo smug—:00 .m 2an

5238.: 83.8: 8 8:223 8026 .aﬂom. :5 ..mEmom.

..mﬁm. Eat 2:5 :28 :8 092 m: 20:23::

61

Figure 5. Examples of mesocarp cell development at different stages of fruit
development for ‘NY 54’ sweet cherry (45x). (A) bloom, (B) endocarp hardening, and
(C) harvest maturity. Images from endocarp hardening and harvest are composite images
created by aligning adjoining microscope field-width images (11 = 3 and 5, respectively),

and are scaled relative to each other for presentation. Scale bar = 200 um.

 

62

Table 4. Comparison of mean cell numbers at bloom, start of endocarp hardening, and

maturity for ‘Selah’, ‘Emperor Francis’ (‘EF’), ‘NY 54’, ‘Bing’, and ‘Regina’ sweet

 

 

 

 

 

cherry fruit.
Cell no. per radial section
2003 2004
Cultivar BloomZ Pit harden Harvest Bloom Pit harden Harvest
Selah 23.5a 70.3a 83.2a 28.4a 76.2a 78.8a
EF 17.0b 40.2b 47.4b 24.0bc 38.2c 41 .4c
NY54 24.7a 26.0c 26.7c 23.4c 29.6d 28.6d
Bing -- -- -- 25.8abc 45.8b 48.3b
Regina -- -- -- 27.6ab 43.4b 43.8bc

 

zMean separation within columns by Tukey’s HSD at P < 0.05.

63

Ema—.0 :. 00:68:00: 0». 800008.: 00: 88:00: 82.0900 :9. .24 man .m8:03.. M.,—28:0. ﬁmmu. 58%. 5088... 8: $0.8. «$00:
0:03 3:: W08 €mCL>EwO #08 2008 8 8:008: 58:08:? 98:55 8 woo: 05 £8 0: a $00.3. Sam. 27:0 Noon Am: 2 T
N :8. 882:5. Gasman :0m80 :3 8088—30: 3.: n :88 W08 90 608885 0:. 50 88:35 :010: mm 8:820: 0: 80 522 x-
8%. >59?me £8 :8008880: 0:00 90 8088.0: 8882 0:. 00:0 3:80: 88 0:. 80 7:200. 8880. 6:0 8 88 0:35 @0000 8

N26. :0 6:. m8: £08 38820 :9. 8:55.

we
mo .
No 1
mo .
mo 1
Ac .
wo .
No .

 

 

 

 

 

Mean cell number (radius)
Moan cell number (radius)

8. 8.

 

 

 

 

 

o q q 4 J‘ a d a 4 q d 4 d a d d _ - a 4 a. d d at q u q q d o

 

03.1.6 :83. no: it 9 Quota: :33. no: it 9

64

358.0 q. 008::100: 0:. 800008.: 00: 80860.. 89.0000 :00 .Z< MAJ aTun—:03: «.8580. A.MmJ. $85. 8: .Wommam. 0200: 0505
m8: 908 zmcbzmm :08 2008 8 8:008: 73:085. 38:85 8 moi 9v 8: Nos: av 2:0 0: V: :3~ 8.0280. 085.85
:0580 :3~ 08:88:20: Q: 0 £58 :08 80 60588.5 0:. 80 88:55 :010: :0 8:380: 0: :0 820a 078:0. >850; 200

E08888: 0:00 90 800080: 0:802 0:. 00:0 095:0: 80: 0:. 80 3:200: 008:8.

 

8
00.
mo.
:0.
:04
0:
8.
Na.
8.
a.
8,

m;

o qq-qqqqqquuqqq-quqqqqquuuqdqddlﬁuqdqqqq«quuddqqqd
o
o20%no00)0ovzoce‘ozeozowzocnaobwezoavevo o 20 ea 00 v0 :0 00 )0 ea 00
03:5: :33. :0: it o. 933:3 :33. :03 it 9

 

65

 

Moan cell number (radius)
Moan cell number (radlus)

 

 

 

 

 

 

 

Table 5. Comparison of the duration and rate of cell division between ‘Bing’, ‘Regina’,
and ‘Selah’ fruit from the period between bloom and endocarp hardening at WSU-
IAREC in 2005. The rate of cell division was calculated by dividing the increase in the
number of cells from bloom by the total accumulation of growing degree days [GDD (4.4

C base)] from the point when bloom for that cultivar occurred.

 

 

GDD
Mean radial accumulation
Mean initial cell no. between bloom Post-bloom cell
radial cell increase after and max. radial division ratez

Cultivar no. at bloom bloom cell no. (no/GDD)

Bing 26.2 21.2 75 0.28a
Regina 28.4 16.8 59 0.2%

Selah 30.4 48.2 139 0.35a

 

zMean separation within coltunn by Tukey’s HSD at P < 0.05.

66

CHAPTER THREE

Construction of a genetic linkage map for the ‘NY 54’ x ‘Emperor Francis’

sweet cherry (Prunus avium L.) population

67

ABSTRACT

Genetic linkage maps were constructed from reciprocal crosses between the sweet
cherry (Prunus avium L.) (2n = 16) cultivars New York 54 (‘NY 54’) and Emperor
Francis (‘EF’). The linkage maps consist of 8 linkage groups (LG) for the ‘EF’ parent
and 10 LG for the ‘NY 54’ parent. The linkage maps for the two parents are 479.1 cM
and 308.9 cM for ‘EF’ and ‘NY 54’, respectively, and consist of 40 SSR, 47 AF LP, 3
SRAP, and 1 morphological marker. The average distance between marker loci is 7 cM
for ‘EF’ and 8 cM for ‘NY 54’. The largest gaps in the maps are 29 cM for ‘EF’ and 34
cM for ‘NY 54’. A total of 24% of the 1:1 markers exhibited distorted segregation ratios
(P < 0.05), many of which were linked together on the ‘EF’ LG 6. A comparison of the
‘NY 54’ x ‘EF’ reciprocal crosses revealed that distorted marker segregation occurred
only when ‘EF’ was used as the paternal parent, presumably resulting from garnetophytic
selection. Fourteen of the sweet cherry linkage groups could be aligned with the
reference Prunus map, ‘Texas’ (‘T’) almond >< ‘Earlygold’ (‘E’) peach, based on shared

SSR markers.

68

INTRODUCTION

Sweet cherry is a diploid (2n = 16) member of the genus Prunus, which contains
many of the economically important tree fruit and nut crops including peach [Prunus
persica (L.) Batsch], almond [P. dulcis (Miller) D.A. Webb], sour cherry (P. cerasus L.),
plum (P. salicina Lindl.), and apricot (P. armeniaca L.). A garnetophytic self-
incompatiblity (GSI) system is present in sweet cherry, typically preventing self-
fertilization and promoting outcrossing (de Nettancourt, 1971). The GSI system
combined with long juvenility periods and large space requirements signiﬁcantly reduce
progeny numbers that can be evaluated in a given time period in sweet cherry breeding
programs. Therefore, marker-assisted selection (MAS) for both qualitative and
quantitative traits, particularly those involved in fruit characteristics, holds great promise
for increasing the efﬁciency of sweet cherry breeding programs.

Currently, self-compatibility, controlled by a self-fertile allele at the S-locus, is
the only sweet cherry trait for which selection is routine using MAS (Dirlewanger et al.,
2004a). PCR-based primers that amplify multiple or speciﬁc S-RNase and SF B alleles at
this locus (Sonneveld et al., 2001; Tao et al., 1999; Yamane et al., 2001) provide an
efﬁcient method of compatible parental selection and identiﬁcation of progeny allelic
constitution years prior to, and at considerably less expense than, controlled crossing
studies in ﬁeld situations. The paucity of additional candidates for MAS results in part
from the lack of suitable populations segregating for traits of interest that can be used for
either genetic linkage map development or bulked segregant analysis (Michelmore,
1991). Additionally, GSI and long generation times of sweet cherry necessitate F.

mapping populations between presumably heterozygous individuals, a mapping

69

conﬁguration that theoretically limits the ability to identify quantitative trait loci (QTL)
because of the heterozygous background (Conner et al., 1998; Grattapaglia and Sederoff,
1994; Wang et al., 2000).

The status of genetic linkage map development in sweet cherry currently lags
behind other important Prunus crops. Stockinger et a1. (1996) developed a randomly
ampliﬁed polymorphic DNA (RAPD) marker-based linkage map of a microspore-derived
callus culture population. However, because of the marker system, this map is not
comparable with other Prunus linkage maps, and phenotypic analysis of horticulturally
important traits was not possible. Isozyme-based interspeciﬁc maps of sweet cherry ><
Prunus incisa Thunb. ex Murr. and sweet cherry x Prunus nippom'ca Matsum. were
reported by Boskovic et al. (1997, 1998), but only 27 markers were placed on the
combined map, a marker density far below that needed for QTL studies. Dirlewanger et
al. (2004a) constructed a linkage map from a ‘Regina’ X ‘Lapins’ sweet cherry cross
consisting of Prunus simple sequence repeat (SSR) markers suitable for comparative
mapping within Prunus species. However, the marker density and coverage was
insufﬁcient for QTL analyses.

In Prunus, interspeciﬁc hybridization between peach and almond has been used
effectively to generate marker diversity to facilitate the construction of highly saturated
linkage maps. The reference Prunus map is from the ‘Texas’ (‘T’) almond X ‘Earlygold’
(‘E’) peach cross, and consists of 562 markers covering 519 cM over the expected 8
linkage groups with a 0.92 cM average marker density (Dirlewanger et al., 2004a).
However, the use of interspeciﬁc crosses in this and other Prunus linkage maps has

resulted in up to 46% of the markers exhibiting distorted segregation ratios (Bliss et al.,

70

2002; F oolad et al., 1995; Joobeur et al., 1998). A common region of linkage distortion
is around the G81 locus (S), resulting from garnetophytic selection (Bliss et al., 2002;
F oulongne et al., 2003; Joobeur et al., 1998; Lambert et al., 2004; Vilanova et al., 2003).
Similarly, the ‘Regina’ >< ‘Lapins’ cross is partially incompatible, S1 S3 X 51 S4', thereby
limiting the analysis for the S-locus region to meiotic products ﬁ'om ‘Regina’ as all the
progeny would have the S4' allele from ‘Lapins’. Therefore, the ideal sweet cherry
mapping population would be an intraspeciﬁc, fully compatible cross where the progeny
segregate for many fruit and tree traits of interest. SSR markers already placed on
existing Prunus linkage maps would be the most informative marker system, taking
advantage of potential codominance, and collinearity of the Prunus genome (Dirlewanger
et al., 2004a).

The objective of this study was to construct a sweet cherry genetic linkage map
from the reciprocal crosses between ‘NY 54’ and ‘EF’ which would be comparable to
previously generated Prunus linkage maps as well as suitable for future QTL mapping

experiments.

MATERIALS AND METHODS
Plant material

Images in this dissertation are presented in color. The sweet cherry population
used for this study was developed from reciprocal crosses of ‘NY 54’ and ‘EF’ (Fig. 1).
‘NY 54’ was selected from wild P. avium forests in Germany and introduced at the New
York State Agricultural Experiment Station, Cornell University (R.L. Andersen, pers.

comm). ‘EF’ is a cultivated sweet cherry of unknown origin, grown primarily for

71

processed cherr'y products. In 2001, pollen was collected from ‘NY 54’ and ‘EF’ trees in
the National Research Support Project 5 (NRSPS) collection in Prosser, Wash. ‘NY 54’
was used as a maternal parent in Washington State, and pollen was transported to
Michigan for use in reciprocal crosses with ‘EF’ as the maternal parent. From the
crosses, 617 F1 individuals were planted at Michigan State University’s Clarksville
Horticultural Experiment Station (MSU-CHES) in Clarksville, Michigan in the spring of
2002. The seedlings were planted at 1.5 m and 6.1 m within and between row spacing,
respectively. Standard orchard management practices (irrigation, fertilization, and pest
and disease control) for MSU-CHES were followed.

From the total population, a linkage mapping subset of 190 individuals was
selected. This subset consisted of 86 individuals from the ‘NY 54’ X ‘EF’ reciprocal
cross, 103 individuals from the ‘EF’ X ‘NY 54’ reciprocal cross, and one individual with
no reciprocal cross information. Approximately equal numbers of progeny from each of
the four S-haplotype groups (48, S253; 49, S254; 47, S3S6; 46, S456) were included in the
mapping population. These four S-haplotype groups were shown previously to segregate

according to the expected 1:1:1:1 ratio (Ikeda et al., 2005).

DNA isolation and marker analyses

For DNA extraction, young, unfolded leaves from the parents and each progeny
individual were collected, placed immediately on dry ice, transported to the laboratory,
and placed directly in a -80°C freezer for at least 24 h. Leaves from each individual were
then lyophilized for 48 h and stored long-term at -20°C. DNA isolation was done using

the CTAB method described by Stockinger et al. (1996).

72

Genotypic data for S-allele segregation was published previously by Ikeda et al.
(2005). Brieﬂy, ‘EF’, ‘NY54’, and all progeny were genotyped for their S-RNase alleles
using the S—RNase gene speciﬁc PCR primer pair, Pru-C2 and PCE-R (Tao et al., 1999;
Yamane et al., 2001). Because the Sz-RNase-speciﬁc fragment is not clearly ampliﬁed
with the PruC2/PCE-R combination, the S2 allele speciﬁc PCR primer pair, PaS2-F and
PaS2-R was used for conﬁrmation of 5; presence (Sonneveld et al., 2001). Reaction
mixtures, PCR conditions, and PCR product visualization were as described by Ikeda et
al. (2005).

SSR markers developed from several Prunus species were used in the
development of the ‘NY 54’ X ‘EF’ linkage map (Table 1). The SSR markers used in
these analyses were derived from peach (“BPPCT”, Dirlewanger et al., 2002; “CPPCT”,
Aranzana et al., 2002; “UDP”, Cipriani et al., 1999; “MA”, Yamamoto et al., 2002; and
“Prp”, Silva et al., 2005), sweet cherry (“EMPA”, Clarke and Tobutt, 2003; “EMPaS”,
Vaughan and Russell, 2004; and “PMS”, Struss et al., 2002), sour cherry (“Pce”, Struss et
al., 2002; and “PS”, Sosinski et al., 2000), almond (“CPDCT”, Mnejja et al., 2005; and
“EPDCU”, P. Arus, pers. com.), and plum (“CPSCT”, Mnejja et al., 2004). A similar
temperature proﬁle, other than annealing temperature, was used for all PCR reactions:
94°C for 5 min, 35 cycles of 94°C (45 sec), X°C (45 sec), 72°C (90 sec), and a ﬁnal
extension step of 72°C for 5 min, where X = the published optimum annealing
temperature for each primer. For “EMPA” and “EMPaS” primers, a touchdown PCR
temperature proﬁle was used as described by Clarke and Tobutt (2003). The reaction
mixture contained 1X PCR buffer, 2.5 mM MgC12, 120 uM of each dNTP, 2.5 pmol of

each primer, 50 ng of genomic DNA and 0.3 U T aq polymerase (Invitrogen Corporation,

73

Carlsbad, Calif.) in a 12.5 ul reaction. PCR reactions were run in a MJ Research PTC
100 Peltier Thermal Cycler (Bio-Rad Laboratories, Hercules, Calif). PCR reactions
were stored at 4°C until use.

Ampliﬁed fragment length polymorphism (AF LP) analysis consisting of genomic
DNA digestion with EcoRI and Msel restriction enzymes, adapter ligation, pre-
ampliﬁcation, and selective ampliﬁcation were similar to Vos et al. (1995), with the
following modiﬁcations described by Hazen et al. (2002). Pre-ampliﬁcation of 2 ul of
restriction ligation genomic DNA product was combined with 25 ng each of EcoRI + A
and Msel + C oligonucleotides, 1X PCR buffer, 1.5 mM MgC12, 0.5 mM of each dNTP,
and 0.5 U T aq polymerase (Promega Corporation, Madison, Wisc.) in 20 ul total volume
and ampliﬁed in a MJ Research PTC 100 Peltier Thermal Cycler (Bio-Rad Laboratories,
Hercules, Calif). The temperature proﬁle used for pre-ampliﬁcation was 94°C for 2 min,
26 cycles of 94°C for 1 min, 56°C for l min, 72°C for l min, and a ﬁnal extension step of
72°C for 5 min. The pre-ampliﬁcation PCR product was diluted 6X with sterile water.
Selective ampliﬁcation used In] of the diluted pre-ampliﬁcation product with 25 ng of
EcoRl + AN primer, 30 ng of Msel + CNN primer, 1X PCR buffer, 1.5 mM MgC12, 0.2
mM of each dNTP, and 0.4 U T aq polymerase (Promega Corporation, Madison, Wisc.) in
20 ul total volume and ampliﬁed in a MJ Research PTC 100 Peltier Thermal Cycler (Bio-
Rad Laboratories, Hercules, Calif). The temperature proﬁle used for selective
ampliﬁcation was 94°C for 2 min, 12 cycles of 94°C for 30 sec, 65°C for 30 sec, 72°C for
1 min, 23 cycles of 94°C for 30 sec, 56°C for 30 sec, 72°C for 1 min, and a ﬁnal
extension step of 72°C for 2 min. Dinucleotide EcoRI + AN, and trinucleotide MseI +

CNN selective ampliﬁcation primers were used as the best compromise between number

74

of polymorphic bands per primer combination and ease and reliability of scoring (Lu et
aL,l998)

Sequence related ampliﬁed polymorphism (SRAP) primer combinations mel/em
and mel/em2 were used as reported by Li and Quiros (2001) with the following PCR
modiﬁcations. The reaction mixture contained 1X PCR buffer, 2.5 mM MgC12, 120 uM
of each dNTP, 2.5 pmol of each primer, 50 ng of genomic DNA and 0.3 U Taq
polymerase (Invitrogen Corporation, Carlsbad, Calif.) in a 12.5 pl reaction. The
temperature proﬁle used was 94°C for 2 min 5 cycles of 94°C for 45 sec, 35°C for 45 sec,
72°C for l min, 35 cycles of 94°C for 45 sec, 50°C for 45 sec, 72°C for l min, and a ﬁnal
extension step of 72°C for 7 min. PCR reactions were run in a MJ Research PTC 100
Peltier Thermal Cycler (Bio-Rad Laboratories, Hercules, Calif).

Fragment visualization was the same for SSR, AF LP, and SRAP markers. After
the addition of 4 pl formamide/dye solution, the PCR products were denatured at 94°C
for ﬁve min. The PCR products were visualized by electrophoresis on a 6% denaturing
polyacrylamide gel in a 50 cm Sequi-Gen GT vertical sequencing apparatus (Bio-Rad
Laboratories, Hercules, Calif.) for 2.5 h at 70 W with 1X TBE buffer. Following
electrophoresis, the gels were stained with the Silver Sequence DNA Sequencing System
(Promega Corporation, Madison, Wise.) and dried for 24 h. DNA ﬁagment sizes were
scored visually using 10 and 50 base pair ladders (Invitrogen Corporation, Carlsbad,

Calif).

75

Chi square analysis and linkage map construction

‘NY 54’, ‘EF’, and a total of 190 progeny were genotyped using 50 SSR markers,
8 AF LP primer combinations, and 2 SRAP primer combinations. Both dominant and
codominant SSR markers were genotyped in the ‘NY 54’ X ‘EF’ population. All AF LP
and SRAP fragments scored were dominant. For ease and reproducibility of scoring, all
markers were scored initially as dominant fragments whereby individual alleles for
codominant SSR markers were scored separately. Segregating fragments present in one
parent and absent in the other parent were tested to ﬁt a 1:1 ratio, while segregating
fragments present in both parents were tested to ﬁt a 3:1 ratio. Chi square goodness-of-ﬁt
tests were performed using ﬁmctions in Excel 2002 (Microsoft Corp., Redmond, Wash.).
Linkage analysis was performed with JoinMap 3.0 (Van Ooijen, and Voorrips, 2001),
using a minimum LOD score of 3.0 and a maximum recombination fraction of 0.4.
Linkage groups were constructed using MapChart 2.1 (V oorrips, 2002), with distances

presented in cM calculated by the Kosarnbi (1944) function.

RESULTS
Marker segregation

A total of 116 SSRs from various sources were screened for ampliﬁcation and
segregation in 190 progeny from the ‘NY 54’ X ‘EF’ reciprocal populations (Table 1). Of
these SSRs, 66 (57%) either did not amplify or did not segregate in this population. Six
of the surveyed SSR markers (5%) resulted in no ampliﬁcation product, while 60 (52%)
did not segregate in the ‘NY 54’ X ‘EF’ population. However, this high number of

monomorphic markers cannot be attributed solely to the use of primers developed in

76

other Prunus species, as a similar percentage (45%) of the SSR markers derived from P.
avium genomic libraries did not segregate. The remaining 50 (43%) SSRs were used to
genotype the progeny populations (Fig. 2). One of these SSRs (BPPCT021) was
removed ﬁom the analysis because the complex banding pattern was difﬁcult to interpret.
For AF LP markers, EcoRI dinucleotide and Msel trinucleotide selective primers were
used. Results from Lu et al. (1998) suggested that this conﬁguration offered the best
compromise between the number of polymorphic ﬁagments produced and ease of
scoring. From eight different EcoRI dinucleotide and Msel trinucleotide AF LP selective
primer combinations, a total of 61 polymorphic fragments were generated (Table 2). The
number of fragments identiﬁed varied from one (EcoRl + AA, Msel +CAA) to 17 (EcoRI
+ AT, Msel +CTC), with an average of eight polymorphic fragments per primer
combination (Fig. 3). This average fragment number is similar to the average (6.8)
reported for similar combinations in peach (Lu et al., 1998). Two SRAP primer
combinations were used during the development of the linkage map, generating seven
polymorphic loci. Fifteen (31%), 13 (21%), and four (57%) of the SSR, AF LP, and
SRAP markers, respectively, deviated signiﬁcantly (P < 0.05) from the expected Fl
segregation ratio. However, these markers were used in the initial linkage analysis.

‘NY 54’ exhibited less heterozygosity than ‘EF’ for SSR, AF LP, and SRAP
markers. Only 42% of SSR and AF LP markers used to genotype the population

segregated for alleles from ‘NY 54’ (Table 3).

77

Linkage map construction

An Fl pseudo-testcross mapping strategy was used to develop linkage maps for
both parents, ‘NY 54’ X ‘EF’. A total of 8 LG were constructed for ‘EF’ and 10 LG were
constructed for ‘NY 54’ (Fig. 4). Forty (82%) SSRs were placed on the linkage map,
while nine (18%) remained unlinked (Table 1). Forty-seven AF LP fragments (76%)
were placed on the linkage map, while 14 (24%) remained unlinked (Table 2). Three of
the SRAP markers (43%) were placed on the linkage map. The ‘NY 54’ X ‘EF’ linkage
map consists of a total of 91 markers; 40 SSR, 47 AF LP, and 3 SRAP markers and one
morphological (S) locus. Fifty-three dominant markers segregate on the ‘EF’ parental
map, while 23 segregate on the ‘NY 54’ parental map (Tables 3 and 4). Fifteen
codominant markers appear on both parental maps. Ten SSR, 10 AF LP, and 2 SRAP
markers placed on the linkage map deviated signiﬁcantly from the expected segregation
ratio. The total cM coverage is 479.] CM and 308.9 cM for the ‘EF’ and ‘NY 54’
parents, respectively (Tables 3 and 4). The average distance between markers is 7 and 8
cM for ‘EF’ and ‘NY 54’, respectively. The largest gaps in the linkage map are 29 cM
for ‘EF’ and 34 cM for ‘NY 54’.

Based on shared markers, homology between 12 of the 18 ‘NY 54’ and ‘EF’
linkage groups could be assigned. The use of SSR markers previously placed on the ‘T’
X ‘E’ reference Prunus map allowed tentative homology between the maps to be
established. At least one SSR marker on 14 of the linkage groups generated in this study
was located on the ‘T’ X ‘E’ map. These 14 linkage groups have been assigned group
numbers according to the ‘T’ X ‘E’ terminology (Fig. 4). Four linkage groups consisting

entirely of AF LP markers currently cannot be compared with the ‘T’ X ‘E’ map.

78

A large region (~ 50 cM) of LG 6 from the ‘EF’ parent consisted of linked
markers exhibiting distorted segregation ratios (Fig. 4). When chi-square goodness-of-ﬁt
tests were used to test for deviation ﬁom the expected segregation ratios within each
reciprocal cross, segregation distortion was only evident when ‘EF’ was used as the

pollen donor (Table 5).

DISCUSSION

The parents for this cross were selected for several reasons. ‘NY 54’ is a wild P.
avium selection, cultivated only for seedling rootstock production, while ‘EF’ is a
domesticated and cultivated variety. Therefore, although this cross is intraspeciﬁc, it is a
cross between a wild relative and a cultivated variety, presmnably maximizing the
available heterozygosity in P. avium. Since many potentially desirable traits are often
found in wild relatives (Tanksley and McCouch, 1997), future QTL studies using this
population may identify alleles not currently represented in the cultivated germplasm of
sweet cherry. Furthermore, ‘NY 54’ and ‘EF’ differed in many important fruit size and
quality characters, and identiﬁcation of QTL for these traits may provide insight into the
mechanism of domestication of sweet cherry. Finally, the presumed heterozygosity of
the two parents was predicted to maximize the SSR loci available for linkage mapping.
The other published intraspeciﬁc sweet cherry linkage map, a cross of ‘Regina’ X
‘Lapins’ (Dirlewanger et al., 2004a), two cultivated varieties, was presumed to have
lower overall heterozygosity.

‘EF’ is an extremely important cultivar in the history of sweet cherry

improvement. It was the maternal parent in crosses with irradiated ‘Napoleon’ pollen

79

that resulted in the self-fertile selection John Innes 2420 (Lewis and Crowe, 1954).
Because self-fertility is a highly desirable production trait, progeny from this original
cross have been heavily used in sweet cherry breeding programs. Therefore, both ‘EF’
and ‘Napoleon’ have contributed to all current self-compatible cultivars, and ‘EF’ has
contributed to 20% of the self-incompatible cultivars grown in North America (Choi and
Kappel, 2004). Because of the long juvenility period, only four to ﬁve generations have
been developed in the most advanced breeding programs since the introduction of self-
fertility (Kappel and Lay, 1997). The frequent appearance of ‘EF’ in modern sweet
cherry pedigrees, and the potential for continued marker-trait linkages due to a limited
number of meioses, suggest that marker and QTL alleles identiﬁed in this mapping
population would be informative even for current cultivars.

The size of the diploid Prunus reference map is 519 cM (Dirlewanger et al.,
2004a). The cM length for the ‘EF’ map was 479.1 cM, therefore approximating the
expected size. Despite the ~ 500 cM length for the ‘EF’ linkage map, it is incomplete,
with only six of the eight potential linkage groups identiﬁed. Although incomplete, the
total cM distance of the ‘EF’ map is only 40 cM less than the coverage of the ‘T’ X ‘E’
map (519 cM), and in the lower end of the 393 to 1,144 cM range of map distances
reported for diploid Prunus (Bliss et al., 2002; Chaparro et al., 1994; Dettori et al., 2001;
Dirlewanger et al., 1998, 2004a; Foolad et al., 1995; Hurtado et al., 2002; Joobeur et al.,
1998; Lambert et al., 2004; Vilanova et al., 2003; Viruel et al., 1995). The current
distance is consistent with the predicted genome size of sweet cherry, which is slightly
larger than peach (6.6X108 vs. 5.3X108) (Dickson et al., 1992), and two linkage groups

have yet to be identiﬁed. Unfortunately, the ‘NY 54’ map was only 309.8 cM, as it

80

exhibited less heterozygosity than ‘EF’ for the SSR loci. With a framework map
constructed, future marker selection will be targeted to reduce current gaps in the linkage
maps.

SSR markers were chosen as a framework for the ‘NY 54’ X ‘EF’ map, and higher
throughput marker systems, such as AF LP, were used to increase the map marker density.
The transferability of markers and collinearity of genomes between Prunus species
suggested that the use of SSR markers developed for other Prunus species would be
successful in sweet cherry (Cantini et al., 2001; Cipriani et al., 1999; Clarke and Tobutt,
2003; Dirlewanger et al., 2002, 2004a; Downey and Iezzoni, 2000; Hormaza, 2002;
Messina et al., 2004; Mnejja et al., 2004, 2005; Schueler et al., 2003; Sosinski et al.,
2000; Struss et al., 2003 Vilanova et al., 2003; Wang et al., 2002; Wunsch and Hormaza,
2002; Yamamoto et al., 2002). As predicted, only 5% of the surveyed SSR markers
resulted in no ampliﬁcation product and a high number of polymorphic fragments per
AF LP primer combination were generated by using EcoRI dinucleotide and Msel
trinucleotide selective primers.

A high percentage of distorted segregation ratios have been reported previously in
Prunus genetic linkage maps. This is particularly apparent in the linkage maps developed
from interspeciﬁc peach X almond crosses (Bliss et al., 2002; Foolad et al., 1995; Joobeur
et al., 1998), where the percent of loci exhibiting distorted segregation ratios range from
37% to 46%. Similarly, interspeciﬁc hybrids of peach X P. davidiana (Foulongne et al.,
2003) and the three-way cross of Myrobalan plum (P. cerasifera Ehrh) X (ahnond X
peach) (Dirlewanger et al., 2004b) exhibited high percentages of distorted marker

segregation ratios, 30% and 42%, respectively. Reported distorted segregation ratios

81

have been lower in intraspeciﬁc crosses; 15% to 18.5% in peach (Lu et al., 1998; Dettori
et al., 2001), 10% in almond (Joobeur et al., 2000), and 11% to 14% in apricot (Hurtado
et al., 2002; Vilanova et al., 2003). In the present study, 24% of all markers deviated
from the expected segregation ratio (Fig. 4). Both SSR and AF LP marker types had a
similar percentage of markers exhibiting distorted segregation.

The high percentage of markers exhibiting distorted segregation ratios has been
attributed to the interspeciﬁc nature of many of the Prunus crosses, and the presence of
garnetophytic self-incompatibility (Bliss et al., 2002; F oulongne et al., 2003; Joobeur et
al., 1998; Lambert et al., 2004; Vilanova et al., 2003). Therefore, an intraspeciﬁc, fully
compatible population for sweet cherry linkage mapping was a priority, as the importance
of the S-locus region for mu quality traits has been highlighted in several QTL studies in
Prunus (Dirlewanger et al., 1999; Etienne et al., 2002; Quilot et al., 2005; Yamamoto et
al., 2001). ‘NY 54’ (S256) and ‘EF’ (8384) are both self-incompatible cultivars, but the
cross between the two cultivars is fully compatible. Conversely, the ‘Regina’ X ‘Lapins’
sweet cherry mapping population (Dirlewanger et al., 2004a) is only partially compatible,
resulting in the potential loss of genomic regions linked to the S1 haplotype of the
‘Lapins’ S locus. Analysis of progeny segregation for the four potential S-haplotype
genotypic classes in the ‘NY 54’ X ‘EF’ population by Ikeda et al. (2005) conﬁrmed the
cross-compatibility. However, when ‘NY 54’ was used as the pollen parent there were a
signiﬁcant excess of individuals with the S2 allele from ‘NY 54’, and a deﬁcit of
individuals with the $6 allele from ‘NY 54’ (Ikeda et al., 2005). The apparent

competitive advantage of the S2 was not indicated by distorted segregation for the allele-

82

speciﬁc marker for that allele in this mapping population because the subset of progeny
used for mapping were comprised of equal number from each S—haplotype group.
Although the ‘NY 54’ X ‘EF’ population reported here did not show distorted
segregation ratios at or around the S locus, a signiﬁcant portion (~ 50 cM) of LG 6 from
the ‘EF’ parent consisted of linked markers exhibiting distorted segregation ratios (Fig.
4). Interestingly, a similar location of aberrant segregation ratios has been documented in
two previous studies. Dirlewanger et al. (2004b) identiﬁed a group of linked markers
exhibiting distorted segregation ratios in F 1 progeny of the three-way cross of Myrobalan
plum X (almond X peach). In the interspeciﬁc parent, distorted markers were located
throughout the linkage group corresponding to LG 6 of the ‘T’ X ‘E’ map. Map coverage
was not as complete for the Myrobalan plum parent, but distorted loci corresponding to
the central region of the homologous linkage group in that cross were also present. In
that study, distorted segregation was attributed to meiotic problems due to the
interspeciﬁc nature of the parent. Foulongne et al. (2003) identiﬁed loci in the same
location on LG 6 with distorted segregation ratios in an F2 population generated from a
peach X P. davidiana interspeciﬁc cross. An excess of the peach alleles were present
around two loci, UDP98-412 and UDP96-001. Both of these SSR loci also were
identiﬁed as distorted in the Myrobalan plum X (almond X peach) map reported by
Dirlewanger et al. (2004b). In the ‘NY 54’ X ‘EF’ sweet cherry population, UDP98-412
was not polymorphic, but it is located one to three cM from the S locus on the ‘T’ X ‘E’
linkage map (Dirlewanger eta1., 2004a). Gametophytic selection near the UDP98-412
locus was attributed to self-incompatibility in the P. davidiana parent in the peach X P.

davidiana interspeciﬁc cross (Foulongne et al., 2003). As expected, no distorted loci

83

were linked to the S locus in the ‘NY 54’ X ‘EF ’ sweet cherry population because it is
fully compatible. UDP96-001 segregates in both ‘NY 54’ and ‘EF’ and was located on
the opposite end of the linkage group as the S locus in each of these linkage maps. The
male sterility locus for peach was located 5.5 cM from UDP96-001 in the ‘T’ X ‘E’
linkage map (Dirlewanger et al., 2004a). Segregation at this locus may contribute to
distorted segregation ratios in an F2 population as reported by F oulongne et al. (2003),
but this would not explain the distorted segregation ratios that occurred in the F1 ‘NY 54’
X ‘EF’ population reported in this study.

Because the ‘NY 54’ X ‘EF’ population consists of reciprocal crosses, we were
able to examine the inﬂuence of gamete sources from each parent on the observed marker
distortion. Segregation distortion was only evident when ‘EF’ was used as the pollen
donor (Table 5). This observation is similar to that described by F oulongne et al. (2003),
in which gametophytic selection causing distorted segregation in a peach X P. davidiana
F2 population was assumed to occur only among male gametes. Since the ‘NY 54’ X
‘EF ’ cross is intraspeciﬁc, the cause of pollen garnetophytic selection cannot be attributed
to homologous chromosome pairing inconsistencies during interspeciﬁc hybridization.
These data, and the repeated observation of distorted segregation in this area of the
Prunus genome, suggests a genetic inﬂuence for the reduced ﬁtness of pollen gametes.
The linkage of the male-sterility (Ps) locus to markers exhibiting distorted segregation is
compelling. Although male sterility has not been documented in sweet cherry and cannot
fully explain the segregation distortion observed in the ‘NY 54’ X ‘EF’ cross, allele

combinations at the locus may inﬂuence pollen ﬁtness. Alternatively, this genomic

84

region may be important in meiosis and gamete formation, and other dysfunctional and

sub-lethal factors may be present.

CONCLUSIONS
This ﬁrst-generation genetic linkage map developed from the reciprocal cross

between ‘NY 54’ and ‘EF’ in this study provides a good starting point for future QTL
analyses. The parents were selected to maximize available P. avium heterozygosity for
important traits such as fruit size and color, while avoiding linkage distortion problems
identiﬁed in previous Prunus maps. The use of SSR markers common to other Prunus
maps allow for between species comparisons previously unavailable for sweet cherry.
Placement of additional SSR markers, ongoing at this time, will continue to reﬁne and

establish collinearity between sweet cherry and other Prunus species.

85

LITERATURE CITED

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

Bliss, F.A., S. Arulsekar, M.R. Foolad, V. Becerra, A.M. Gillen, M.L. Warburton, A.M.
Dandekar, G.M. Kocsisne, and K.K. Mydin. 2002. An expanded genetic linkage

map of Prunus based on an interspeciﬁc cross between almond and peach.
Genome 45:520-529.

Boskovic, R. and KR. Tobutt. 1998. Inheritance and linkage relationships of
isoenzymes in two interspeciﬁc cherry progenies. Euphytica 103:273-286.

Boskovic, R., K.R. Tobutt, and F .J . Nicoll 1997. Inheritance of isoenzymes and their
linkage relationships in two interspeciﬁc cherry progenies. Euphytica 93:129-
143.

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

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

Choi, C. and F. Kappel. 2004. Inbreeding, coancestry, and founding clones of sweet
cherries in North America. J. Amer. Soc. Hort. Sci. 129:535-543.

Cipriani, G. G. Lot, W.-G. Huang, M.T. Marrazzo, E. Peterlunger, and R. Testolin. 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, J .B. and KR. Tobutt. 2003. Development and characterization of polymorphic
microsatellites from Prunus avium ‘Napoleon’. Mol. Ecol. Notes 3:578-580.

Conner, F.J., S.K. Brown, and NF. Weeden. 1998. Molecular-marker analysis of
quantitative traits for growth and development in juvenile apple trees. Theor.
Appl. Genet. 96: 1027-1035.

de Nettancourt, D. 1977. Incompatibility in angiosperms. Springer, New York, NY.

Dettori, M.T., R. Quarta, and I. Verde. 2001. A peach linkage map integrating RFLPs,
SSRs, RAPDs, and morphological markers. Genome 44:783-790.

86

Dickson, E.E., K. Arumuganathan, S. Kresovich, and J .J . Doyle. 1992. Nuclear DNA
content variation within the Rosaceae. Amer. J. Bot. 79:1081-1086.

Dirlewanger, E., A. Moing, C. Rothan, L. Svanella, V. Pronier, A. Guye, C. Plomion, and
R. Monet. 1999. Mapping QTLs controlling fruit quality in peach (Prunus
persica (L.) Batsch). Theor. Appl. Genet. 98:18-31.

Dirlewanger, E., E. Graziano, T. Joobeur, F. Garriga-Caldere, P. Cosson, W. Howad, and
P. Arus. 2004a. Comparative mapping and marker-assisted selection in Rosaceae
fruit crops. Proc. Natl. Acad. Sci. 101:9891-9896.

Dirlewanger, E., P. Cosson, M. Tavaud, M.J. Aranzana, C. Poizat, A. Zanetto, P. Arus,
and F. Laigret. 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., P. Cosson, W. Howad, G. Capdeville, N. Bosselut, M. Claverie, R.
Voisin, C. Poizat, B. Lafargue, O. Baron, F. Laigret, M. Kleinhentz, P. Arus, and
D. Esmenjaud. 2004b. 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., V. Pronier, C. Parvery, C. Rothan, A. Guye, and R. Monet. 1998. A
genetic linkage map of peach (Prunus persica L. Batsch) using morphological,
RF LP, isoenzyme, RAPD, and AF LP markers. Theor. Appl. Genet. 97:888-895.

Downey, S.L. and A.F. Iezzoni. 2000. Polymorphic DNA markers in black cherry
(Prunus serotina) are identiﬁed using sequences from sweet cherry, peach, and
sour cherry. J. Amer. Soc. Hort. Sci. 125:76—80.

Etienne, C., C. Rothan, A. Moing, C. Plomion, C. Bodenes, L. Svanella-Dumas, P.
Cosson, V. Pronier, R. Monet, and E. Dirlewanger. 2002. Candidate genes and

QTLs for sugar and organic acid content in peach (Prunus persica (L.) Batsch).
Theor. Appl. Genet. 105:145-159.

Foolad, M.R., S. Arulsekar, V. Becerra, and FA. Bliss. 1995. A genetic map of Prunus
based on an interspeciﬁc cross between peach and almond. Theor. Appl. Genet.
91:262-269.

Foulongne, M., T. Pascal, P. Arus, and J. Kervella. 2003. The potential of Prunus
davidiana for introgression into peach [Prunus persica (L.) Batsch] assessed by
comparative mapping. Theor. Appl. Genet. 107:227-238.

Grattapaglia, D. and R. Sederoff. 1994. Genetic linkage maps of Eucalyptus grandis and

Eucalyptus urophylla using a pseudo-testcross: mapping strategy and RAPD
markers. Genetics 137:1121-1137.

87

Hazen, S.P., P. Leroy, and R. Ward. 2002. AP LP in Triticum aestivum L.: patterns of
genetic diversity and genome distribution. Euphytica 125:89-102.

Hormaza, J .J . 2002. Molecular characterization and similarity relationships among
apricot (Prunus armeniaca L.) genotypes using simple sequence repeats. Theor.
Appl. Genet. 104:321-328.

Hurtado, M.A., S. Vilanova, C. Romero, A.G. Abbott, G. Llacer, and ML. Badenes.

2002. Genetic linkage maps of two apricot cultivars (Prunus armeniaca L.) based
on molecular markers. Theor. Appl. Genet. 105:182-191.

Ikeda, K., K. Ushijima, H. Yamane, R. Tao, N.R. Hauck, A.M. Sebolt, and A.F. Iezzoni.
2005. Linkage and physical distances between the S-haplotype S-RNase and SFB
genes in sweet cherry. Sex. Plant Reprod. 17:289-296.

Joobeur, T., M.A. Viruel, M.C. de Vicente, B. Jauregui, J. Ballester, M.T. Dettori, I.
Verde, M.J. Truco, R. Messeguer, 1. Battle, R. Quarta, E. Dirlewanger and P.
Arus. 1998. Construction of a saturated linkage map for Prunus using an almond
X peach F2 progeny. Theor. Appl. Genet. 97:1034-1041.

Joobeur, T., N. Periam, M.C. de Vicente, G.J. King, and P. Arus. 2000. Development of
a second generation linkage map for almond using RAPD and SSR markers.
Genome 43:649-688.

Kappel, F. and W. Lay. 1997. Sweet cherry breeding in Canada from the early 19005 to
1994. Fruit Var. J. 51:233-238.

Kosambi, DD. 1944. The estimation of map distance from recombination values. Ann.
Eugen. 12:172-175.

Lambert, P., L.S. Hagen, P. Arus, and J .M. Audergon. 2004. Genetic linkage maps of
two apricot cultivars (Prunus armeniaca L.) compared with the almond ‘Texas’ X

peach ‘Earlygold’ reference map for Prunus. Theor. Appl. Genet. 108:1120—
1130.

Lewis, D. and L.K. Crowe. 1954. The induction of self-fertility in tree fruits. J. Hort.
Sci. 29:220-225.

Li, G. and CF. Quiros. 2001. Sequence-related ampliﬁed polymorphism (SRAP), a new
marker system based on a simple PCR reaction: its application to mapping and
gene tagging in Brassica. Theor. Appl. Genet. 103:445-61.

Lu, Z.-X., B. Sosinski, G.L. Reighard, W.V. Baird, and A.G. Abbott. 1998.
Construction of a genetic linkage map and identiﬁcation of AF LP markers or
resistance to root-knot nematodes in peach rootstocks. Genome 41 :199-207.

88

Messina, R., O. Lain, M.T. Marrazzo, G. Cipriani, and R. Testolin. 2004. New set of
microsatellite loci isolated in apricot. Mol. Ecol. Notes 4:432-434.

Michelmore, R.W., I. Paran, and RV. Kesseli. 1991. Identiﬁcation of markers linked to
disease-resistance genes by bulked segregant analysis: a rapid method to detect

markers in speciﬁc genomic regions by using segregating populations. Proc. Natl.
Acad. Sci. 88:9828-9832.

Mnejja, M., J. Garcia-Mas, W. Howad, and P.Arus. 2005. Development and
transportability across Prunus species of 42 polymorphic almond microsatellites.
Mol. Ecol. Notes 5:531-535.

Mnejja, M., J. Garcia-Mas, W. Howad, M.L. Badenes, and P. Arus. 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.

Quilot, B., J. Kervella, M. Genard, and F. Lescourret. 2005. Analysing the genetic

control of peach fruit quality through an ecophysiological model combined with a
QTL approach. J. Exp. Bot. 56:3083-3092.

Schueler, S., A. Tusch, M. Schuster, and B. Ziegenhagen. 2003. Characterization of
microsatellites in wild and sweet cherry (Prunus avium L.) — markers for
individual identiﬁcation and reproductive processes. Genome 46295-102.

Silva, C., J. Garcia-Mas, A.M. Sanchez, P. Arus, and M.M. Oliveira. 2005. Looking into
ﬂowering time in almond (Prunus dulcis (Mill) D.A. Webb): the candidate gene
approach. Theor. Appl. Genet. 110:959-968.

Sonneveld, T., T.P. Robbins, R. Boskovic, and KR. Tobutt. 2001. Cloning of six cherry
self-incompatibility alleles and development of allele-speciﬁc PCR detection.
Theor. Appl. Genet. 102: 1046-1055.

Sosinski, B., M. Gannavarapu, L.D. Hager, L.E. Beck, G.J. King, C.D. Ryder, S.
Rajapakse, W.V. Baird, R.E. Ballard, and A.G. Abbott. 2000. Characterization

of microsatellite markers in peach (Prunus persica (L.) Batsch). Theor. Appl.
Genet. 101:421-428.

Stockinger, E.J., C.A. Mulinix, C.M. Long, T.S. Brettin, and A.F. Iezzoni. 1996. A
linkage map of sweet cherry based on RAPD analysis of a microspore-derived
callus culture population. J. Hered. 87:214-218.

Struss, D., M. Boritzki, R. Karle, and A.F. Iezzoni. 2002. Microsatellite markers
differentiate eight Giessen cherry rootstocks. HortScience 37: 191-193.

89

Struss, D., R. Ahmad, S.M. Southwick, and M. Boritzki. 2003. Analysis of sweet cherry
(Prunus avium L.) cultivars using SSR and AFLP markers. J. Amer. Soc. Hort.
Sci. 128:904-909.

Tanksley, SD. and SR. McCouch. 1997. Seed banks and molecular maps: unlocking
genetic potential from the wild. Science 277:1063-1066.

Tao, R., H. Yamane, A. Sugiura, H. Murayama, H. Sassa, and H. Mori. 1999. Molecular
typing of S-alleles through identiﬁcation, characterization and cDNA cloning for
S-RNases in sweet cherry. J. Amer. Soc. Hort. Sci. 124:224-233.

Van Ooijen, J .W. and RE. Voorrips. 2001. JoinMap® 3.0, Software for the calculation
of genetic linkage maps. Plant Research lntemational, Wageningen, the
Netherlands.

Vaughan, SP. and K. Russell. 2004. Characterization of novel microsatellites and
development of multiplex PCR for large-scale population studies in wild cherry,
Prunus avium. Mol. Ecol. Notes 4:429-431.

Vilanova, S., C. Romero, A.G. Abbott, G. Llacer, and ML. Badenes. 2003. An apricot
(Prunus armeniaca L.) F2 progeny linkage map based on SSR and AF LP

markers, mapping plum pox virus resistance and self-incompatibility traits.
Theor. Appl. Genet. 107:239-247.

Viruel, M.A., R. Messeguer, M.C. de Vicente, J. Garcia-Mas, P. Puigdomenech, F.J.
Vargas, and P. Arus. 1995. A linkage map wih RFLP and isozyme markers for
almond. Theor. Appl. Genet. 91:964-971.

Voorrips, RE. 2002. MapChart: Software for the graphical presentation of linkage maps
and QTLs. J. Hered. 93:77-78.

Vos, R, R. Hogers, M. Bleeker, M. Rijans, T. van der Lee, M. Homes, A. Frijters, J. Pot,
J. Peleman, M. Kuiper, and M. Zabeau. 1995. AFLP: a new technique for DNA
ﬁngerprinting. Nucleic Acid Res. 23:4407-4414.

Wang, D., R. Karle, and A.F. Iezzoni. 2000. QTL analysis of ﬂower and ﬁ'uit traits in
sour cherry. Theor. Appl. Genet. 100:535-544.

Wang, Y., L.L. Georgi, T.N. Zhebentyayeva, G.L. Reighard, R. Scorza, and A.G. Abbott.
2002. High-throughput targeted SSR marker development in peach (Prunus
persica). Genome. 45:319-328.

Wunsch, A. and J .I. Hormaza. 2002. Molecular characterisation of sweet cherry (Prunus
avium L.) genotypes using peach (Prunus persica (L.) Batsch) SSR sequences.
Heredity 89:56—63.

90

Yamamoto, T., K. Mochida, T. lmai, Y.Z. Shi, I. Ogiwara, and T. Hayashi. 2002.
Microsatellite markers in peach (Prunus persica (L.) Batsch) derived from an
enriched genomic and cDNA libraries. Mol. Ecol. Notes. 2:298-301.

Yamamoto, T., T. Shimada, T. lmai, H. Yaegaki, T. Haji, N. Matsuta, M. Yamaguchi,
and T. Hayashi. 2001. Characterization of morphological traits based on a
genetic linkage map in peach. Breeding Sci. 51:271-278.

Yamane, H., R. Tao, A. Sugiura, N.R. Hauck, and A.F. Iezzoni. 2001. Identiﬁcation and

characterization of S-RNases in tetraploid sour cherry (Prunus cerasus L.). J.
Amer. Soc. Hort. Sci. 126:661-667.

91

TABLES AND FIGURES

Figure 1. (A) Image of mature ﬁ'uit ﬁ'om ‘Emperor Francis’ (‘EF’) (left) and ‘NY 54’

(right) sweet cherry, illustrating size variation between the two cultivars. (B) Selected
progeny from the ‘EF ’ X ‘NY 54’ sweet cherry linkage mapping population illustrating

variation for fruit characteristics.

i
E
E.
g
“a“:
E
E

a '
._

 

[umm
i

92

Table 1. Origins of simple sequence repeat (SSR) markers used in the development of

the ‘NY 54’ X ‘Emperor Francis’ sweet cherry genetic linkage map.

 

 

Marker Prunus Number of Number of
terminology species SSRs tested SSRs mapped Reference
BPPCT P. persica 17 Dirlewanger et al., 2002
CPDCT P. dulcis 9 Mnejja et al., 2005
CPPCT P. persica l8 Aranzana et al., 2002
CPSCT P. salicina 3 Mnejja et al., 2004
EMPA P. avium 7 Clarke and Tobutt, 2003
EMPaS P. avium 7 Vaughan and Russell, 2004
EPDCU P. dulcis 3 P. Arus (pers. comm.)
MA P. persica 2 Yamamoto et al., 2002
Pce P. cerasus 6 Struss et al., 2002
Feb P. persica 5 Sosinski et al., 2000
PMS P. avium 8 Struss et al., 2002
Prp P. persica 2 Silva et al., 2005
PS P. cerasus 6 Sosinski et al., 2000
UDP P. persica 23 Cipriani et al., 1999

 

93

Figure 2. Co-dominant simple sequence repeat (S SR) fragments obtained with primers
for the CPDCT022 marker. From left to right: (1) 50 base pair sizing ladder, (2) 10 base

pair sizing ladder, (3) ‘NY 54’, (4) ‘Emperor Francis’, (5-24) progeny. Arrows indicate

segregating fragments.

"N MVWVOI‘QCN

-
w“ w w U +
' : wh- w W U
.. _, Uh- v W V v ~‘_
.. w V’W" - v‘.
'- v” “* div W C!" +
v v * .31. “I! - up

94

Table 2. Enzymes used for digest, selective nucleotide combinations used as primers,
number of polymorphic fragments, and number of mapped ﬁ'agments generated by
ampliﬁed ﬁagment length polymorphism (AF LP) analysis in the development of the ‘NY

54’ X ‘Emperor Francis’ sweet cherry genetic linkage map.

 

 

Number of Number of mapped
EcoRI Msel polymorphic fragments ﬁagrnents
EAA CTT 6 5
EAA CAC 8 7
EAA CCC 6 4
EAA CCT 3 1
EAA CAA 1 0
EAT CTC 17 12
EAT CCC 9 7
EAC CTA 1 1 11

 

95

Figure 3. Ampliﬁed fragment length polymorphism (AF LP) fragments obtained with the
selective primers EcoRI+AC and Msel +CTA. From leﬁ to right: (1) 10 base pair sizing
ladder, (2) 50 base pair sizing ladder, (3) ‘NY 54’, (4) ‘Emperor Francis’, (5-29) progeny.

Arrows indicate segregating fragments.

I”

 

96

3. mm
Q “m 3392.883 v.8

2...<5.5% «.2 n mm

30.000835 ads

§t§<m v.8
0:”: Tag-cam 0.8
can 8N43<¢2m a «a opu.0<0!(<w «do

on 7.5% hum
8.1.00: mﬁv

. —.
«3.833. 2.3 «.3 39kg . 3. >2
Q >2 Ob wthnOPUOQO 0.3 8.308.503 Hun
\uﬁgvnaaa a...“ 5.3.33.5 n8
8«. 3 n.
.. 5.838... s. ««o\0 2 to: a . 87338.. 93
«ship—.OOQO 0.0? Dan's—.0206 6.0 9
33.35% ”.2 n >2 o2.~c..8a:a 2.. . .
8..<8!o% «.3 on" 825% n 2
838% ...:
2.523% H «.o 8.525% ... «265.5% a...
Saﬁuéﬁasm 82.8228 a... «8.3233 o... 835.5% a... 333.5% a... 3~.33..2w a...
Edaaﬁeooa VMA 3
8«.8..282..u a...
a e um

.bozuooame . 56 v m 98 mod v k E ...... EB .. .3 @28me 0.8 62 3:8me 6:: .58 a .3 8.82:8 2a 3.5% owed—:2

3:23 of :3an 62 852$ .92: 3:20.08 05 .33 8:33 8: 350% Saswmmoc 80:2 6838:: 98 29: ammo—e: ”382%
cusp—£8 .m. x r? 2: 2 wﬁuconmobg 3:on owed—=3 .808?” @3828 En: \ 28W 05 mama: 3:8: as 28:33:80 E.E.qo
Emiﬁoﬁba 5mg»— EoEwwa comm—93x .58 05 Lots 05m Beams—m v0.88 2: 5m? .29.. 05 so Ba 8:8: gov—SE drew owe:—

nomo he «2 05 2 cozw .04 Eu 5 8ng .85: owao—E— ouocow .323 803m .3 >2. 98 Aim; .mmocﬁm Sauna? .v oSwE

97

0 >2 0 mm
m ....w
00.270 ¢.:o||o 8.3.0 “”0.
«3.000200% «.2. 009<520% . 0. .
80.55% 0.2. 03.31.00... «.00 . 02.205000 « 00
0««..03<: 0.3 I 08.83:. 0.3
. «3.30.3830: 0.00
8.000% « «0 8.000!»% 0.00
0«v.0<0!<% 0.00 .
.. 03.05055. 0”? I .. 05.05::% « 3 03.03; 0. «c
.. 02.50% 0 3 . .
02.085000 0.? . 02 $05000 0.3 02.305000 «.8
«2.305000 0.00 . “WWWHMM “MM 03.05553. 1.0
03.000 0.«« .. . . «2.505000 0.00
5&0 . Awumq.m.u.w‘«».0.0=0% ”WM m >2 8.. 5.0% 0.8
.. . 022% . .
t «.e« 08.20800: 0 0«
0 085000 . 5.805000 0.«« . «5.000200% «.««
DNQUPUEFﬁ 0.0P c as. F.3PuNNOFUOkU '.h F
. 0. F. «00800: c «
«3.00.020 0.0 "VA
«2.30800: 0.0 .. «2.808% 0.0 .. 00.30.05 0.0 03.0523... 0.0

8.0080 .« 8:30

98

3 mm

. ONYOhUEhim «.2.
2. a ran-Fog hvv
. nonuo<0§<m o. 3

§0<0§m ad

3 >2

mpuéooitku cdu

«vuéhoitkm 0.1
2. ...tog ad

 

3 >2

nip-(hugﬁ fun
SVOOOELRM odN

suuipogua ad

a >2
cs. >2

8a.”: wan M.N'

OFQOHOEPE QNM

gqohoghﬁ 9.:

2.99.0536 0.9

uNVooog 5.:

238.55 a... 83525.“. a...

sum

oa-auannDon—am 0.2.

maﬁa. «60¢»..— v.00

outevﬁuwim aim
ounéhoitkm 0.;

37389.3 23
5 >2 .82.“...523 3n

3 ...«annDOn—mm 56» 3.03.. ..nauDOn—mm ads

«3.3 Vania 9.:
. acquuahonmo o6 v

on tun v- FMGNDUOAN 06 .. OCHNEOFOE :6

9:53 .N 2%:

99

Table 3. Number and type of markers for ‘Emperor Francis’ (‘EF’) and ‘NY 54’ parental
maps, map length, target map length from the ‘T’ x ‘E’ Prunus reference map
(Dirlewanger et al., 2004a), marker density, marker gap length, average linkage group

length, and average number of markers per linkage group.

 

Parental maps

 

 

 

Marker type Total EF NY 54
SSR 40 36 ‘7
AFLP 47 28 2°
SRAP 3 3 0
Total 90 67 37
Map statistics
Length in cM 479.1 3089
Target map length
(T X E CM) 519 519
Marker density
(markers/CM) 0.14 0.12
Average distance 7 1 3 3
between markers (cM) . .
Largest gap between
markers (CM) 290 34.0
Average cM/linkage 59 9 30 9
group . .
Average markers/ 8 4 3 7
linkage group . .

 

100

 

 

 

 

 

v.2 mdm ENN Nam _ m6 ﬁom 5mm Now 323 mam “mamas
. . . . . . . . A23 8536
mm as v: n: .m cc of :w HoxSEomﬁo><
. . . . . . . . gaining

2 o S o oo o oo o mm o 2 o oo o N_ o bmmcou coo—32
. . . . . . . . :6 m x b
_ av _ av w Ne m we v av v we 0 S o S 5&5— Howad,
«.5 Name 5mm N.Nw N6 0Q. 5mm You 923 ﬁwcoq
mutate: 32
m a N n N S N Z 38%
o _ o o o o o o mém
o m o m _ m _ m “in?
N m N v _ n _ 0 «mm
22 2m v>z Em >2 mm 52 Em about“:
828m owe—:5

 

.masoa 093?: 3:23 ,3 >2. Ea ABEL Eocen— hohoqu. com ﬁwco— mam Hod—.88 EB baa—ow coo—SE

Amvoom ..E .o coma—@5039 92: 8:20.?“ mzzfm .m, x L: 2: Eat ﬁgu— wowg .ﬁwaﬂ coo—.88 mo 09$ 93 cons—:2 .v 033.

101

 

 

 

 

 

9: mam «.2 «mm «3 - E. we 98 SW :23 3» imam
3 2: 3 N2 3 - 2: 3 on 3 cmﬁwaowmmmﬁ,
. . . . . . . . . 3028.35
«mo 25 2 o 23 2 o - 2 o 2 o 2 o one $53332
- - - - am an 92 S: 58 5.2 hmﬁwmowcww
a: 2mm odm «mm a? - 3m E v. : SH 93 £me
monmuﬁm 9&2
v m m m m o m a m. 2 36>
o o o o o o o N o o m<mm
v m m N v o o _ o b mmm<
o o o o _ o m m u x mmm
3mm E>z 3>z E>z ”>2 Em CE Em m>z Em 25 332
88% ammo—:3

 

3.283 .v 2an

102

.3 can: 858 1m ﬂ :ozmwohwom 38098 ”$5.8m £3 E 3895 a mnméHU—ébxm 8x38 how :58de as?

 

 

 

 

 

 

 

 

:bd 3d 02d no; 3. N am mvmd bmd god om.m Nom H mm mnméhU—ztkm
mood cod owed to om H ow omod _od mood 2.2 on H mm ofomoobmmo
Sod no.2 ooo._ ood on n on wood Nod mood and mm H mm NRQHUZCE
wood mod momd mod Vm ” 9» thd o_d ~ood 3.: om H mm madmoHUaEU
~ood no.2 oNod _od ow ” xv omod _od mood and mm H mm mmm-<,_.02\0<m
mood Gd mvmd nmd ow H mm omod NNd Sod _w.m_ on u om oo_-ouog<<m
_ood 3d— owod 2d mm ” ow omod _od mood owd G H mm noéoohommm
mood Nod owed 3d 31 3 wood mod mood N_.o on H 3 278895:
_mod de oomd we; 3 H mm mood mod _ood Kd— mm H mm mmméommﬁzm
02? 2:? 2:? 33> 0:8 853“ 02.2, 02? 33> 83> 2:2 3:88“ 8x32

i - ~x i - ~x ”858.5 i - ~x i - Nx ”3:035

.HN
ooaoaxm 32030 ooaoomxm 32890
Vm>mem mmem>Z

 

d 04 8 92: $5 .mmocﬁm “83:5. 5 :5on 883:: @2836 .«o coumwuumom E moocuuohmo 139580.50 833800 .m 2an

103

CHAPTER FOUR

Targeted mapping of fruit size and shape QTL in sour cherry (Prunus cerasus L.)

104

ABSTRACT

Fruit size and shape are important production traits in sour cherry (Prunus cerasus
L.), and certain size and shape parameters must be met for a new sour cherry cultivar to
be successful. Identiﬁcation of the genomic regions involved in variation for fruit size
and shape by quantitative trait loci (QTL) analysis would provide an early selection
method to increase the efﬁciency of sour cherry breeding. Previous QTL analyses in
both sour cherry and peach [P. persica (L.) Batsch] described fruit size QTL. In many
cases, these QTL were located on linkage group 6 (LG 6) in the same region as the
Prunus garnetophytic self-incompatibility locus (S) and the peach ﬂat/round fruit shape
locus (S *). The objective of this study was to conduct a targeted mapping and QTL
analysis for fruit size and shape traits using progeny from the cross between the sour
cherry cultivars Ujfehértoi Fﬁrtos (‘UF’) and ‘Sureﬁre’. Both homeologous LG 6 from
‘UF’ were developed using previously mapped SSR markers from other Prunus species
and aligned with the reference Prunus linkage map. The homeologous LG 6 were 49.1
cM (LG 6a) and 68.2 cM (LG 6b), respectively. Population distributions for progeny
ﬁ'uit weight, diameter, and length/width percentage approximated normal distributions
with transgressive segregation. Mean values for all three traits were signiﬁcantly
different (P < 0.05) among progeny S-allele groups. This indicates that the fruit traits did
not vary independently of S—genotype, suggesting that QTL for each trait may be linked
to the S locus. Using all the LG 6 markers, QTL were identiﬁed for fruit weight and fruit
length/width percentage, but not fruit diameter. The fruit weight QTL was only
signiﬁcant in 2004, but in that year it explained 26.4% of the phenotypic variation. As

predicted by the analysis of variance, the nearest marker was the Sd allele marker for the

105

S locus at a map distance of three cM on ‘UF’ LG 6a. The fruit length/width QTL was
signiﬁcant for all three years of the study, and explained between 10.5 and 22.6% of the

phenotypic variation. When all three years were combined, the QTL was co-located with

the CPSCT012 marker on ‘UF’ LG 6a.

106

INTRODUCTION

Sour cherry (Prunus cerasus L.) is a tetraploid (2n = 4x = 32) member of the
predominantly diploid (2n = 16) cultivated Prunus genus. Sour cherry is believed to have
arisen multiple times through natural hybridization between ground cherry (P. ﬁuticosa
Pall.; 2n = 4x = 32) and unreduced gametes from sweet cherry (P. avium L.; 2n = 16)
(Beaver and Iezzoni, 1993; Brettin et al., 2000; Olden and Nybom, 1968). Currently, the
sour cherry industry in the United States is based almost entirely on production of one
cultivar, ‘Montrnorency’, a 400-year-old selection from France (Iezzoni, 1988, 2005).
Primary utilization of fruit from ‘Montmorency’ is for processed cherry products. Thus,
adoption of ‘Montmorency’ as the major sour cherry cultivar was likely due to consistent
production and suitability for mechanical harvesting (Iezzoni, 2005), and not necessarily
for superior fruit quality characteristics.

Breeding for genotypes with fruit quality superior to ‘Montmorency’ is an
important goal of the Michigan State University (MSU) sour cherry breeding program.
One of the parents used in the breeding program is Ujfehértoi Fiirtos (‘UF’), due to its
excellent fruit quality (Iezzoni, 2005). Fruit of ‘UF’ is larger, ﬁrmer, and sweeter than
‘Montmorency’ fruit, and has been used for fresh market production (Lang et al., 2003).
As this market expands, a premium will also likely be placed on large sour cherries,
similar to sweet cherry fresh market production (Whiting et al., 2005, 2006). Therefore,
future breeding efforts may be directed toward selection of larger sour cherry varieties for
fresh market production.

In the MSU sour cherry breeding program, development of improved cultivars for

processing use is complicated by factors other than fruit quality. Existing harvest and

107

processing equipment designed for ‘Montmorency’ sour cherries requires any new
cultivar to be adaptable to existing technologies. For example, to avoid potential
breakage of endocarp (pit) ends during the pitting process, a small, round pit is desired
(A.F. Iezzoni, pers. comm). Although pit shape was not correlated with fruit shape in
peach (Quilot et al., 2004), observation of the MSU sour cherry breeding program
germplasm indicated that ﬁ'uit shape may be a good predictor of pit shape.

Because of the long juvenility period and extensive land use requirements
associated with breeding perennial tree crops such as sour cherry (Fogle, 1975), the MSU
sour cherry breeding program has a continued interest in development of molecular tools
useful for marker-assisted selection (MAS). Identiﬁcation of molecular markers for fruit
characteristics could have tremendous impact on speeding up the breeding and evaluation
cycle. However, many fruit quality characteristics are presumed to be quantitatively
inherited, resulting from the coordinated action of many potential genes affecting the
phenotypic expression of the trait. Currently, the primary method of evaluating the
number and relative signiﬁcance of the potential genes inﬂuencing a given trait is by
quantitative trait loci (QTL) analysis. In this type of analysis, phenotypic trait data are
combined with a genetic linkage map to identify regions of the genome signiﬁcantly
associated with mean differences in trait values.

For sour cherry, one QTL analysis has been reported for the ‘Rheinische
Schattenmorelle’ (‘RS’) X ‘Erdi Botermo’ (‘EB’) sour cherry population (Wang et al.,
2000). However, the map used for that analysis (Wang et al., 1998) was incomplete, and
the number of common markers with the Prunus reference map {‘Texas’ almond [Prunus

dulcis (Miller) D.A. Webb] X ‘Earlygold’ peach (‘T’ X ‘E’) (Dirlewanger et al., 2004)} is

108

low, preventing more general conclusions. Two fruit weight QTL accounting for over
29% of the phenotypic variation in that population were identiﬁed (Wang et al., 2000).
Few additional QTL analyses have been performed in peach (Dirlewanger et al., 1999;
Etienne et al., 2002; Quilot et al., 2005; Yamamoto et al., 2001). However, among these
few populations, there has been some consistency observed for ﬁ'uit weight QTL. For
example, in three different peach populations, one or more fruit size QTL have been
identiﬁed on linkage group 6 (LG 6) in the same region as the Prunus S locus
(Dirlewanger et al., 1999; Etienne et al., 2002; Quilot et al., 2005; Yamamoto et al.,
2001). The S locus has been an area of continued importance in sour cherry, as even
though most sour cherry cultivars are self-compatible, progeny from crosses can
segregate for self-incompatibility, an undesirable production trait (Lansari and Iezzoni,
1990). Interestingly, the peach S * is also located ~ 10 cM from the S locus (Dirlewanger
et al., 2004). Thus, three traits mapped in peach, but also important for sour cherry
production, are located in the same genomic region. The transferability of markers and
collinearity of genomes between Prunus species suggest that targeted mapping of this
region for the purpose of identifying QTL for fruit weight and shape in sour cherry could
be successful (Cantini et al., 2001; Cipriani et al., 1999; Clarke and Tobutt, 2003;
Dirlewanger et al., 2002, 2004; Downey and Iezzoni, 2000; Hormaza, 2002; Messina et
al., 2004; Mnejja et al., 2004, 2005; Schueler et al., 2003; Sosinski et al., 2000; Struss et
al., 2003 Vilanova et al., 2003; Wang et al., 2002; Wunsch and Hormaza, 2002;
Yamamoto et al., 2002). The objective of this study was to determine whether previously
identiﬁed QTL in peach, that are putatively co-located with the S and S * loci on LG 6,

are also present in sour cherry.

109

MATERIALS AND METHODS
Plant material and phenotypic analysis

Images in this dissertation are presented in color. The sour cherry population
used for this study was developed from the cross ‘UF’ X ‘Sureﬁre’ (Fig. 1). ‘UF’ was
selected from a Hungarian landrace and is sold in the US. as Balaton® (Iezzoni, 2005).
‘Sureﬁre’ was released from the New York State Agricultural Experiment Station,
Cornell University, and results from a cross between ‘Borchert Black Sour’ X ‘NY 6935’
(Cummins, 1994). From the cross, 197 F. individuals were planted at Michigan State
University’s Clarksville Horticultural Experiment Station (MSU-CHES) in Clarksville,
Michigan in spring 1998. Parental trees of ‘UF’ and ‘Sureﬁre’ were also located at
MSU-CHES. The seedlings were planted at 1.5 m and 6.1 m within and between row
spacing, respectively. The number of individuals used in this study was reduced to 126,
as 71 individuals either died, were identiﬁed as resulting from out-crossing or self-
crossing, or did not have any fruit, and were eliminated from the analyses. The seedling
trees were not pruned since establishment. Standard orchard management practices
(irrigation, fertilization, and pest and disease control) for MSU-CHES were followed.

Phenotypic measurements were performed for ‘UF’, ‘Sureﬁre’, and the 126
individuals ﬁ'om the population. Fruit measurements were made on ﬁve replicate fruit
from each individual at least twice during estimated harvest maturity. The harvest date
with the largest mean fruit weight was used for QTL analyses. Progeny were sampled for
three years (2002-2004). Individual fruit weight, length (polar diameter), and width
(cheek diameter) were measured. The length/width percentage, an indication of the

overall shape of the ﬁ'uit, was calculated from the measured fruit length and width. In

110

2004, a crop load rating was made for each progeny individual based on a 1-10 scale (1 =

low crop load, 10 = high crop load).

DNA isolation and marker analysis

For DNA extraction, young, unfolded leaves from the parents and each progeny
individual were collected, placed immediately on dry ice, transported to the laboratory,
and placed directly in a -80°C freezer for at least 24 h. Leaves from each individual were
then lyophilized for 48 h and stored long-term at -20°C. DNA isolation was done using
the CTAB method described by Stockinger et al. (1996).

Restriction fragment length polylmorphism (RF LP) analysis was used to genotype
parents and progeny for their S-allele haplotype. For RFLP analysis, six pg of DNA was
digested with HindIII (Boehringer Mannheim Biochemicals, Indianapolis, Ind.), run on a
1.0 % agarose gel for 36 h at 30 V, and transferred to a nylon membrane (Hybond-N+,
Amersham Biosciences, Piscataway, NJ.) according to Wang et al. (1998). PCR
ampliﬁed fragments of the Sé-RNase cDNA from sweet cherry (Tao et al., 1999) were
used as the probe. Probes were radiolabelled with 32P-dCTP (Amersham Pharrnacia
Biotech, NJ.) using the random primer hexamer-priming method described by Feinberg
and Vogelstein (1983). After hybridization at 60°C for 16 h and high stringency washes
(2 X 20 min with 2x SSC and 1% SDS followed by 2 X 30 min with 0.2x SSC and 0.5%
SDS at 60°C), radioactive signal was detected on X-ray ﬁlms.

Simple sequence repeat (SSR) markers developed from several Prunus species
that have been mapped to LG 6 of the ‘T’ X ‘E’ reference Prunus map (Dirlewanger, et

al., 2004) were selected for mapping in the ‘UF’ X ‘Sureﬁre’ population. Subsequently,

111

other SSR markers that were not mapped on the ‘T’ X ‘E’ map, but had been placed on
linkage groups in other Prunus maps that aligned to the ‘T’ X ‘E’ LG 6, were added to the
analyses. The SSR markers used in these analyses were derived from peach (“BPPCT”,
Dirlewanger et al., 2002; “CPPCT”, Aranzana et al., 2002; “UDP”, Cipriani et al., 1999;
and “MA”, Yamamoto et al., 2002; “Prp”, Silva et al., 2005), sweet cherry (“EMPA”,
Clarke and Tobutt, 2003; and “PS”, Sosinski et al., 2000), and plum (P. salicina Lindl.)
(“CPSCT”, Mnejja et al., 2004).

A similar temperature proﬁle, other than annealing temperature, was used for all
PCR reactions: 94°C for 5 min, 35 cycles of 94°C (45 sec), X°C (45 sec), 72°C (90 sec),
and a ﬁnal extension step of 72°C for 5 min, where X = the published optimum annealing
temperature for each primer. For “EMPA” primers, a touchdown PCR temperature
proﬁle was used as described by Clarke and Tobutt (2003). The reaction mixture
contained 1X PCR buffer, 2.5 mM MgC12, 120 pM of each dNTP, 2.5 pmol of each
primer, 50 ng of genomic DNA and 0.3 U T aq polymerase (Invitrogen Corporation,
Carlsbad, Calif.) in a 12.5 pl reaction. PCR reactions were run in a MJ Research PTC
100 Peltier Thermal Cycler (Bio-Rad Laboratories, Hercules, Calif.). PCR reactions
were stored at 4°C until use.

After the addition of 4 pl formamide/dye solution, the PCR products were
denatured at 94°C for ﬁve min. The PCR products were visualized by electrophoresis on
a 6% denaturing polyacrylamide gel in a 50 cm Sequi-Gen GT vertical sequencing
apparatus (Bio-Rad Laboratories, Hercules, Calif.) for 2.5 h at 70 W with 1X TBE buffer.
Following electrophoresis, the gels were stained with the Silver Sequence DNA

Sequencing System (Promega Corporation, Madison, Wise.) and dried for 24 h. DNA

112

fragment sizes were scored visually and fragment sizes were estimated relative to 10 and

50 base pair ladders (Invitrogen Corporation, Carlsbad, Calif).

Single marker analysis of variance

Based on RFLP proﬁles, all progeny could be placed in six S-allele groups
(S4 S3 513' 5d, Sa 513' Sd 51', Sa 513' Sr' Snun, S4 S3 513' Snuu, 54 Sa 513' 51', and Sa 513' Sd
Sm"). Two of these groups (S4 S, 813’ S1 ', and S, S13' Sd Snun) result from non-disomic
inheritance and occurred in only 10% of the progeny. Analysis of variance for progeny
fruit weight, diameter, and length/width percentage within each disomically-inherited S-
allele group was performed using the general linear model procedure in SAS (SAS

Institute, Cary, NC.) to determine potential linkage of each trait with the S-locus.

Linkage analysis and map construction

Due to the tetraploid genome and F1 pseudo-testcross population structure, all
SSR markers were scored as single-dose restriction fragments (Wang et al., 1998; Wu et
al., 1992), where the fragment was present in one but not both parents and ﬁt a 1:1
(presencezabsence) segregation ratio, or present in both parents and ﬁt a 3:1 segregation
ratio. ‘UF’, ‘Sureﬁre’, and a total of 126 progeny were genotyped using 17 SSR markers.
Linkage analysis was performed with JoinMap 3.0 (Van Ooijen, and Voorrips, 2001),
using a minimum LOD score of 3.0 and a maximum recombination fraction of 0.4.
Linkage groups were constructed using MapChart 2.1 (V oorrips, 2002), with distances

presented in cM calculated by the Kosambi (1944) frmction.

113

QTL and statistical analysis

QTL analyses were performed using Windows QTL Cartographer 2.0 (Wang et
al., 2005) using composite interval mapping (CIM). CIM was run with model 6 of the
program with the background markers selected using the forward and backward
regression method. The LOD threshold for declaring a QTL was determined by 1000
permutations for each trait at a signiﬁcance level of P < 0.05, a priori. Estimates of the
R-squared value indicating the explained phenotypic variance for each QTL and the
additive effect of the QTL were obtained from the QTL Cartographer output. Graphical
representations of the QTL were made using output from QTL Cartographer and
MapChart 2.1 (V oorrips, 2002). Analysis of variance, correlations, t-tests, and
heritability estimates were performed using the appropriate ﬁmction in SAS statistical

analysis soﬁware (SAS Institute, Cary, N.C.).

RESULTS
Marker analysis and LG 6 construction

A total of 27 SSR markers placed on the ‘T’ X ‘E’ reference Prunus LG 6, or
other Prunus LG 6 aligned with the ‘T’ X ‘E’ map, were tested for ampliﬁcation and
segregation in the ‘UF’ x ‘Sureﬁre’ population. Of those markers, 37% either did not
amplify a corresponding locus in the ‘UF’ x ‘Sureﬁre’ population or were monomorphic.
Of the 17 remaining markers, 14 ﬁt the expected 1:1 or 3:1 segregation ratios for single
dose restriction fragments (Wang et al., 1998; Wu et al., 1992), while three markers that

did not ﬁt the expected segregation ratio were not included in the analysis.

114

The S-allele genotype of ‘UF’ is S4 S.’ S, Sm... and ‘UF’ produces four gamete
types from regular pairing between homologous chromosomes (S4 S4 , S4 S,,..', S. ' S, ,
S. ' Sm...) and two gamete types from pairings between non-homologous chromosomes (5..
S. ', S, S,,..) (Hauck et al., submitted). For our analysis, only those progeny that resulted
from normal homologous pairing were included, and not the 10 % produced by non-
homologous pairing. As predicted, the segregation of the S-alleles (S4, S. ', and S4) in the
progeny all ﬁt the expected 1:1 segregation ratio. The S-allele phenotype of ‘Sureﬁre is
S4 S, S.3' (Hauck et al. 2006) Since ‘UF’ has a functional S4 allele, all ‘Sureﬁre’ pollen
gametes containing an S4 allele will be incompatible in the ‘UF’ style and pollen tube
growth will be arrested. As expected, the only ‘Sureﬁre’ pollen gamete type that
successfully fertilized ‘UF’ was S, S.3', resulting in four progeny types ﬁom regular
pairing between homologous chromosomes (S4 S, S.3' 8,, S, S. 3' S, S. ', S, S.3' S.' S,,.., S4
S, S. 3' S,,..). Therefore, it was not possible to place the S—locus on the ‘Sureﬁre’ linkage
map and only the meiotic products from ‘UF’ could be used to test the association of trait
variation with S-locus genotype.

Linkage analysis was performed with the 14 SSR markers and S-allele data for the
S4, S. ', and S4 self-incompatibility alleles segregating in the population generated from
RF LP analysis. Two homeologous linkage groups (LG 6a and LG 6b), consisting
entirely of SSR markers and the respective S-alleles at a 10 to 25 cM distance for ‘UF’
were generated and aligned with the ‘T’ X ‘E’ LG 6 (Fig. 2). The S4 S-allele was located
on LG 63, while the S4 and S.’ S-alleles were located on LG 6b. Only one LG 6 from
‘Surefrre’ was identiﬁed, consisting of two markers; however, the S locus could not be

mapped because no S-alleles segregated from ‘Sureﬁre’. LG 6a and LG 6b had lengths

115

of 49.1 cM and 68.9 cM, respectively, less than the 83.7 cM distance for LG 6 of the ‘T’
X ‘E’ reference map. Seven additional SSR markers remained unlinked. Approximately
10% of the progeny in this population had S-allele haplotypes indicating non-disomic
inheritance, an observation that has previously been made in sour cherry (Wang et al.,
1998). Due to the small number of individuals in these groups, the difficulty in
determining segregation ratios, and the potential for linkage map distance inﬂation, these
individuals were not included in linkage map development or QTL analysis. However,
the S, allele marker and the ﬂanking marker distal to it on ‘UF’ LG 6a still appeared
distorted, having a slightly skewed segregation ratio (P < 0.1) according to chi square
tests (Fig. 2). Distorted loci have often been identiﬁed in the S locus region, presumably
due to self-incompatibility (Bliss et al., 2002; Joobeur et al., 1998; Lambert et al., 2004;

Vilanova et al., 2003).

Fruit size and shape

Fruit weight, diameter, and length/width percentage were measured for three
consecutive years (2002-2004). To reduce variation in shape measurements, the
percentage of fruit length (polar diameter) divided by width (cheek diameter) was
calculated. Thus, the more ﬂat-oblate shape the fruit is, the lower the length/width
percentage. From 2002-2004, fruit weight, diameter, and shape measurements were
made for ﬁve fruit per individual. The ‘UF’ X ‘Sureﬁre’ population ﬁrst began ﬁ'uiting in
year 2002. Because of variability in precocity, only 76 individuals in the population had
fruit during 2002. In 2003 and 2004, 118 and 126 individuals were measured,

respectively.

116

All traits exhibited continuous variation typical of a quantitative trait with
polygenic inheritance (Fig. 3). Broad sense heritability (H2) for each trait was high
(Table 1) indicating consistency over the years of the study and a low genotype X
environment interaction. The parental means of ‘UF’ and ‘Sureﬁre’ were significantly
different (P < 0.001) for fruit weight and diameter, but not for fruit length/width
percentage. The average value of the parents was signiﬁcantly different than the progeny
mean value for fruit weight and fruit diameter (P < 0.0001) but not signiﬁcantly different
for the length/width percentage (Table 1). Transgressive segregation occurred for all
traits (Fig. 3). For fruit weight, the distribution of the progeny was skewed toward
smaller fruit, with 82% of the progeny averaging smaller fruit weight than the mid-parent
value. As expected, there was a strong positive linear correlation (P < 0.0001) between
fruit weight and fruit diameter (Fig. 4). Fruit weight and fruit length/width percentage
were not signiﬁcantly correlated (P= 0.324), but fruit diameter and fruit length/width
percentage were (P < 0.0001). However, the correlation was weak, with an R-squared
value of 0.042 (Fig. 4). Because fruit weight can be inﬂuenced by crop load level on
individual trees, linear correlation between mean fruit weight and tree crop load aﬁer fruit
set for each individual in the population was analyzed in 2004. Although the relationship
was signiﬁcant (P < 0.001), the low R-square value (0.163) and positive relationship
between increasing fruit weight and crop load indicated that small fruit size was not a

result of high crop load in this population (Fig. 5).

117

Single marker analysis to test the association of ﬁuit traits with the S locus

To determine whether potential QTL for fruit weight, diameter, and length/width
percentage were linked to the S-locus, analysis of variance was used to compare trait
phenotypic values of progeny within each potential disomically-inherited S-allele group.
In essence, this process is similar to single marker QTL analysis. Signiﬁcant differences
between trait mean values for S-haplotype groups (P < 0.001) indicated linkage between
the measured phenotype and the S locus (Table 2). However, this type of analysis does
not provide a linkage distance estimate from known markers, knowledge that is essential

for potential MAS strategies.

QTL analysis

Two QTL were identiﬁed on the ‘UF’ LG 6a, one for fruit weight (wt), and one
for fruit length/diameter percentage (shape) (Table 3, Fig. 6). The signiﬁcant QTL for
fruit weight (wt) was only identiﬁed in 2004, although there was a peak in the same
location for both 2002 and 2003 that did not reach the LOD signiﬁcance level. Similarly,
a QTL peak for fruit diameter was observed at the same map location as fruit weight for
all three years; however, the fruit diameter peak failed to reach the LOD signiﬁcance
level in any year. In 2004, the ﬁ'uit weight QTL explained 26.4% of the phenotypic
variation. The QTL had an effect in the opposite direction predicted by the paternal
phenotype, reducing fruit weight by 1.55 g. The QTL identiﬁed for fruit length/width
percentage (shape) was consistently identiﬁed in all three years of the study (Table 3,
Fig. 6). In years 2002-2004, the QTL explained 22.6, 17.5, and 10.5 % of the phenotypic

variation, respectively. Again, the QTL had an effect opposite as predicted by the

118

parental phenotype, increasing the fruit length/width percentage an average of 5%. The
S, allele speciﬁc marker for the S locus was the closest marker to the fruit weight QTL,
3.6 cM from the QTL peak. CPSCT012, an SSR marker derived from a genomic library
of Japanese plum (Mnejja et al., 2004), was the closest marker to the fruit length/width

percentage QTL, 0 to 4.9cM from the QTL peak depending on the year.

DISCUSSION

The objective of this study was to determine whether QTL identiﬁed in peach that
were co-located with the S and S * loci were also present in sour cherry. A targeted
mapping approach was used, whereby only SSR markers that have been mapped to the
linkage group containing the above loci in the ‘T’ X ‘E’ reference Prunus map (LG 6) and
homologous linkage groups from other Prunus populations were used.

A LG 6 totaling 34.4 cM containing the S locus had previously been constructed
from the ‘RS’ X ‘EB’ sour cherry population (Hauck et al., 2002). However, the ‘RS’ X
‘EB’ map is currently not comparable with the reference Prunus map because few
markers are common. In that study, the S, allele, later named S2,, (Hauck et al., 2006)
from the ‘RS’ parent was placed on LG 6. The only other S-alleles able to be mapped in
that population, S.3' and S6, both also from ‘RS’, were linked to each other but not to any
other marker on the ‘RS’ X ‘EB’ linkage map. In a previous QTL study using the ‘RS’ X
‘EB’ population, no fruit size QTL were located on this linkage group and fruit shape was
not measured (Wang et al., 2000).

Both ‘UF’ and ‘Sureﬁre’ are self-compatible sour cherries. However, self-

incompatible progeny can result from crosses between two self-compatible sour cherries

119

(Lansari and Iezzoni, 1990). If self-compatible and self-incompatible individuals
segregate in the population, the possibility exists that fruit weight could be inﬂuenced by
the level of crop load on the tree, with self-compatible individuals presumably having a
higher crop load. Because of this possibility, each individual in the population was rated
for crop level aﬁer fruit set in 2004. The correlation between fruit weight and crop load
was statistically signiﬁcant, but the R-squared value was only 0.163 (Fig. 5).
Furthermore, the relationship between fruit weight and crop load was positive, with ﬁ'uit
size increasing with crop load. Therefore, any putative QTL linked to the S locus are
likely true QTL, not simply artifacts of higher crop load.

From this study, a single QTL for both fruit weight and fruit length/width
percentage were identiﬁed on the ‘UF’ LG 6a (Table 3, Fig. 6). The QTL for fruit weight
(wt) was only signiﬁcant in year 2004. As predicted by analysis of variance, the QTL
peak was 3.6 cM from the nearest marker, S,, an allele speciﬁc marker for the S locus.
Although only signiﬁcant for one year of the study, this QTL explained 26.4% of the
phenotypic variation. More importantly, the effect was opposite of what was expected
from the ‘UF’ parental phenotype, reducing fruit weight an average of 1.55 g.
Interestingly, no signiﬁcant QTL for fruit diameter was identiﬁed, although there was a
strong positive correlation between fruit weight and diameter (Fig. 4). In each year of the
study, a QTL peak for fruit diameter at the same map location as that of the fruit weight
QTL was observed, but it failed to reach LOD signiﬁcance level in any year. Given the
strong correlation between ﬁ'uit weight and diameter, the fruit weight QTL identiﬁed may
inﬂuence both fruit weight and diameter. Although fruit weight and diameter were the

only measured traits used for QTL analysis, the ‘UF’ X ‘Sureﬁre’ population appears to

120

segregate for mesocarp cell number (see Chapters 2 and 5). ‘UF’ and ‘Sureﬁre’ have
statistically similar numbers of mesocarp cells, but progeny from the tails of the fruit
weight distribution are signiﬁcantly different (P < 0.05) for mesocarp cell number (Table
4). Further examination of this trait as a component of total fruit size in this population is
warranted.

The shape QTL for fruit length/width percentage was signiﬁcant in all years of
the study and explained between 10.5 and 22.6% of the phenotypic variation in a given
year. The peak for this QTL was nearest to the CPSCT012 marker, but varied slightly
from year to year. When phenotypic measurements from all three years were averaged,
the peak of the QTL co-located with the CPSCT012 marker. The peak for the shape
QTL was 26.2 cM from the S locus, within the 50% recombination range and explaining
the signiﬁcant association with the S locus when analysis of variance was performed.
Like the wt QTL, the effect of the shape QTL was opposite of the predicted parental
phenotype, increasing the fruit length/width percentage by 5 %. The fruit S * locus is
located ~ 10 cM from the S locus on the Prunus reference map (Dirlewanger et al., 2004).
Furthermore, the S locus is ~ 33 cM from the CPSCT012 marker on the Prunus reference
map, similar to the 26.2 cM distance observed in the present study (Fig. 2). In peach, the
S * gene is dominant for ﬂat-oblate fruit (Lesley, 1940). Although ﬂat-oblate fruited
progeny are present in the ‘UF’ X ‘Surefrre’ population, segregation for that character did
not indicate that it was controlled by a single locus. Because the shape QTL identiﬁed in
the ‘UF’ X ‘Sureﬁre’ population is further away from the S locus than the S* gene and
only explains 10.5 to 22.6% of the phenotypic variation, it is likely not the same locus.

However, the use of length/width percentage for phenotypic data presumably accounts

121

for variation that may be present even if the phenotype is measured as a dominant
agronomic character. Therefore, we cannot exclude the possibility that a modiﬁer gene
for the S * locus underlies the QTL discovered in the ‘UF’ X ‘Sureﬁre’ population.

Allelic effects opposite to those expected ﬁom the parental phenotype are not
uncommon in sour cherry. Wang et al. (2000) reported 50% of the QTL identiﬁed in the
‘RS’ X ‘EB’ cross to have effects opposite as predicted by the parental phenotype. As in
that study, this phenomenon likely explains the transgressive segregation seen for fruit
weight and fruit length/width percentage. Each parent likely contributed both favorable
and unfavorable alleles for QTL affecting the same trait, and since both the ‘RS’ X ‘EB’
and ‘UF’ X ‘Sureﬁre’ populations consist of F. individuals, recombination of these alleles
in the progeny generated transgressive phenotypes (Wang et al., 2000).

Although few QTL studies have been performed to date in Prunus, QTL for fruit
size have consistently been identiﬁed in the region where the S locus is presumably
located in peach, at the bottom of LG 6 (Dirlewanger et al., 1999; Etienne et al., 2002;
Quilot etal., 2005; Yamamoto etal., 2001). Because peach is self-compatible, this locus
is not included in many peach linkage maps. However, almond is self-incompatible, and
this locus has been placed on interspeciﬁc peach X almond maps. The peach S * locus is
located ~ 10 cM from the S locus, nearer to the center of LG 6 (Dirlewanger et al., 2004).
However, because this is a dominant locus in peach, and QTL analysis for fruit shape has
not been previously published. In the ‘Ferjalou Jalousia’ (‘J’) X ‘Fantasia’ (‘F’) peach
population, a QTL explaining between 22.6% and 51% (depending on the year) of the
phenotypic variance for fruit weight was identiﬁed (Dirlewanger et al., 1999; Etienne et

al., 2002). This QTL was co-located with the S * locus and had an effect in the same

122

direction of the parental phenotype. In the ‘Akame’ (‘A’) X ‘Juseitou’ (‘J’) peach cross,
two QTL were identiﬁed for fruit weight, near the Dw dwarf locus at the top of LG 6 and
the Gr red/green leaf locus in the central portion of LG 6 (Yamamoto et al., 2001). R-
square values indicating the phenotypic variance explained by these QTL were not
provided, but both QTL effect ﬁ'uit weight in the opposite direction as suggested by the
parental phenotype. In the backcross population developed from P. davidiana X
‘Summergrand’ peach, a QTL for pit diameter was identiﬁed at the PC60 marker locus, 4
cM from the S * locus. This QTL explained 41% of the phenotypic variation and had an

effect in the same direction as the parental phenotype.

CONCLUSIONS

This analysis demonstrates the utility of a targeted mapping approach for
identifying useful QTL. The targeted mapping approach for sour cherry takes advantage
of the transportability of SSR markers across and collinearity between Prunus genomes
(Dirlewanger etal., 2004). Identiﬁcation of the QTL for fruit weight and fruit
length/diameter percentage in this study should have immediate impact in sour cherry
breeding programs. For example, self-compatibility is one of the few characters selected
for using MAS (Dirlewanger et al., 2004). Identiﬁcation of the QTL in this study with a
strong negative effect linked to the S, allele of the S locus suggests selection against this
allele could have a positive impact on fruit size. Alternatively, the breeder would need to
create larger populations for the potential to break the linkage if the S, allele itself was
desired. Similarly, the close association of the fruit length/width percentage QTL with

the nearest marker may allow for efﬁcient marker-assisted selection. Since fruit and pit

123

shape were observed to be associated in sour cherry (A.F. Iezzoni, pers. comm), the
utility of this marker-trait association may be best utilized for early selection of desirable
pit shapes.

As with any QTL analysis, the true utility of the marker-trait associations
described in this study will be determined by the stability and repeatability of the
association in other populations. However, for at least the ﬁ'uit weight QTL, the fact that
similar QTL have been documented in other Prunus species is encouraging, particularly '

given the low density of the linkage group mapped in this study and the difficulty in

 

identiﬁcation of QTL in polyploid populations (Wang et al., 1998, 2000).

124

LITERATURE CITED

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

Beaver, J .A. and A.F. Iezzoni. 1993. Allozyme inheritance in tetraploid sour cherry
(Prunus cerasus L.). J. Amer. Soc. Hort. Sci. 118:873-877.

Bliss, F.A., S. Arulsekar, M.R. Foolad, V. Becerra, A.M. Gillen, M.L. Warburton, A.M.
Dandekar, G.M. Kocsisne, and K.K. Mydin. 2002. An expanded genetic linkage

map of Prunus based on an interspeciﬁc cross between almond and peach.
Genome 45:520-529.

Brettin, T.S., R. Karle, E.J. Crowe, and A.F. Iezzoni. 2000. Chloroplast inheritance and
DNA variation in sweet, sour, and ground cherry. J. Hered. 91 :75-79.

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

Cipriani, G., G. Lot, W.-G. Huang, M.T. Marrazzo, E. Peterlunger, and R. Testolin.
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, J .B. and KR. Tobutt. 2003. Development and characterization of polymorphic
microsatellites from Prunus avium ‘Napoleon’. Mol. Ecol. Notes 3:578-580.

Cummins, J .N. 1994. Register of new fruit and nut varieties: Brooks and Olmo List 36.
HortScience 29:942-969.

Dirlewanger, E., A. Moing, C. Rothan, L. Svanella, V. Pronier, A. Guye, C. Plomion, and
R. Monet. 1999. Mapping QTLs controlling fruit quality in peach (Prunus
persica (L.) Batsch). Theor. Appl. Genet. 98218-31.

Dirlewanger, E., E. Graziano, T. Joobeur, F. Garriga-Caldere, P. Cosson, W. Howad, and
P. Arus. 2004. Comparative mapping and marker-assisted selection in Rosaceae
fruit crops. Proc. Natl. Acad. Sci. 101:9891-9896.

Dirlewanger, E., P. Cosson, M. Tavaud, M.J. Aranzana, C. Poizat, A. Zanetto, P. Arus,
and F. Laigret. 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.

125

Downey, S.L. and A.F. Iezzoni. 2000. Polymorphic DNA markers in black cherry
(Prunus serotina) are identiﬁed using sequences from sweet cherry, peach, and
sour cherry. J. Amer. Soc. Hort. Sci. 125:76-80.

Etienne, C., C. Rothan, A. Moing, C. Plomion, C. Bodenes, L. Svanella-Dumas, P.
Cosson, V. Pronier, R. Monet, and E. Dirlewanger. 2002. Candidate genes and
QTLs for sugar and organic acid content in peach (Prunus persica (L.) Batsch).
Theor. Appl. Genet. 105:145-159.

F einberg, AD. and G. Vogelstein. 1983. A technique for radiolabelling DNA restriction
fragments to high speciﬁc activity. Anal. Biochem. 132:6-13.

F ogle, H.W. 1975. Cherries. In: Janick, J. and J .N. Moore (eds) Advances in Fruit
Breeding. Purdue University Press, West Lafayette, Ind, pp. 348-366.

Hauck, N.R., K. Ikeda, R. Tao, and A.F. Iezzoni. 200x. The mutated SI-haplotype in sour
cherry has an altered S-haplotype speciﬁc F-box protein gene. J. Hered.
(submitted).

Hauck, N.R., H. Yamane, R. Tao, and A.F. Iezzoni. 2002. Self-compatibility and
incompatibility in tetraploid sour cherry (Prunus cerasus L.). Sex. Plant Reprod.
15:39-46.

Hauck, N.R. H. Yamane, R. Tao, and A.F. Iezzoni. 2006. Accumulation of non-
functional S-haplotypes result in the breakdown of garnetophytic self-
incompatibility in tetraploid Prunus. Genetics 172: 1191-1198.

Hormaza, J .J . 2002. Molecular characterization and similarity relationships among
apricot (Prunus armeniaca L.) genotypes using simple sequence repeats. Theor.
Appl. Genet. 104:321-328.

Iezzoni, A.F. 1988. ‘Montrnorency’ sour cherry. Fruit Var. J. 42:74-75.

Iezzoni, A.F. 2005. Acquiring cherry germplasm ﬁ'om central and eastern Europe.
HortScience 40:304-308.

Joobeur, T., M.A. Viruel, M.C. de Vicente, B. Jauregui, J. Ballester, M.T. Dettori, I.
Verde, M.J. Truco, R. Messeguer, 1. Battle, R. Quarta, E. Dirlewanger and P.
Arus. 1998. Construction of a saturated linkage map for Prunus using an almond
X peach F2 progeny. Theor. Appl. Genet. 97:1034-1041.

Kosambi, DD. 1944. The estimation of map distance from recombination values. Ann.
Eugen. 12:172-175.

126

Lambert, P., L.S. Hagen, P. Arus, and J .M. Audergon. 2004. Genetic linkage maps of
two apricot cultivars (Prunus armeniaca L.) compared with the almond Texas X
peach Earlygold reference map for Prunus. Thoer. Appl. Genet. 108:1120—1130.

Lang, G.A., B. Behe, J. Sowa, A. Iezzoni, and E. Fallahi. 2003. Niche marketing of fruit
crops: Windows for proﬁtability? HortScience 38(5):746.

Lansari, A. and A. Iezzoni. 1990. A preliminary analysis of self-incompatibility in sour
cherry. HortScience 25: 1 636-163 8.

Lesley, J .W. 1940. A genetic study of saucer fruit shape and other characters in the
peach. Proc. Am. Soc. Hort. Sci. 37:218-222.

Messina, R., O. Lain, M.T. Marrazzo, G. Cipriani, and R. Testolin. 2004. New set of
microsatellite loci isolated in apricot. Mol. Ecol. Notes 4:432—434.

Mnejja, M., J. Garcia-Mas, W. Howad, and P.Arus. 2005. Development and
transportability across Prunus species of 42 polymorphic almond microsatellites.
Mol. Ecol. Notes 5:531-535.

Mnejja, M., J. Garcia-Mas, W. Howad, M.L. Badenes, and P. Arus. 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.

Olden, E.J. and N. Nybom. 1968. On the origin of Prunus cerasus L. Hereditas 59:327-
345.

Quilot, B., J. Kervella, and M. Genard. 2004. Shape, mass and dry matter content of
peaches of varieties with different domestication levels. Sci. Hortic. 99:397-393.

Quilot, B., J. Kervella, M. Genard, and F. Lescourret. 2005. Analysing the genetic

control of peach fruit quality through an ecophysiological model combined with a
QTL approach. J. Exp. Bot. 56:3083-3092.

Schueler, S., A. Tusch, M. Schuster, and B. Ziegenhagen. 2003. Characterization of
microsatellites in wild and sweet cherry (Prunus avium L.) — markers for
individual identiﬁcation and reproductive processes. Genome 46:95-102.

Silva, C., J. Garcia-Mas, A.M. Sanchez, P. Arus, and M.M. Oliveira. 2005. Looking into

ﬂowering time in almond (Prunus dulcis (Mill) D.A. Webb): the candidate gene
approach. Theor. Appl. Genet. 110:959-968.

127

Sosinski, B., M. Gannavarapu, L.D. Hager, L.E. Beck, G.J. King, C.D. Ryder, S.
Rajapakse, W.V. Baird, R.E. Ballard, and A.G. Abbott. 2000. Characterization
of microsatellite markers in peach (Prunus persica (L.) Batsch). Theor. Appl.
Genet. 101:421-428.

Stockinger, E.J., C.A. Mulinix, C.M. Long, T.S. Brettin, and A.F. Iezzoni. 1996. A
linkage map of sweet cherry based on RAPD analysis of a microspore-derived
callus culture population. J. Hered. 87:214-218.

Struss, D., R. Ahmad, S.M. Southwick, and M. Boritzki. 2003. Analysis of sweet cherry
(Prunus avium L.) cultivars using SSR and AF LP markers. J. Amer. Soc. Hort.
Sci. 128:904-909.

Tao, R., H. Yamane, A. Sugiura, H. Murayama, H. Sassa, and H. Mori. 1999. Molecular
typing of S-alleles through identiﬁcation, characterization and cDNA cloning for
S-RNases in sweet cherry. J. Amer. Soc. Hort. Sci. 124:224-233.

Van Ooijen, J .W. and RE. Voorrips. 2001. JoinMap® 3.0, Software for the calculation
of genetic linkage maps. Plant Research International, Wageningen, the
Netherlands.

Vilanova, S., C. Romero, A.G. Abbott, G. Llacer, and ML. Badenes. 2003. An apricot
(Prunus armeniaca L.) F2 progeny linkage map based on SSR and AF LP

markers, mapping plum pox virus resistance and self-incompatibility traits.
Theor. Appl. Genet. 107:239-247.

Voon'ips, RE. 2002. MapChart: Software for the graphical presentation of linkage maps
and QTLs. J. Hered. 93277-78.

Wang, D., R. Karle, and A.F. Iezzoni. 2000. QTL analysis of ﬂower and fruit traits in
sour cherry. Theor. Appl. Genet. 100:535-544.

Wang, D., R. Karle, T.S. Brettin, and A.F. Iezzoni. 1998. Genetic linkage map in sour
cherry using RFLP markers. Theor. Appl. Genet. 97: 1217-1224.

Wang, S., C. J. Basten, and Z.-B. Zeng. 2005. Windows QTL Cartographer 2.5.
Department of Statistics, North Carolina State University, Raleigh, NC.
(http://statgen.ncsu.edu/qtlcart/WQTLCart.htm).

Wang, Y., L.L. Georgi, T.N. Zhebentyayeva, G.L. Reighard, R. Scorza, and A.G. Abbott.
2002. High-throughput targeted SSR marker development in peach (Prunus
persica). Genome. 45:319-328.

Whiting, M.D., G. Lang and D. Ophardt. 2005. Rootstock and training system affect
cherry growth, yield, and fruit quality. HortScience 40:582-586.

128

Whiting, M.D., D. Ophardt, and J .R. McFerson. 2006. Chemical blossom thinners vary
in their effect on sweet cherry ﬁ'uit set, yield, fruit quality, and crop value.
HortTechnology 16:66-70.

Wu, K.K., W. Bumquist, M.E. Sorrells, T.L. Tew, P.H. Moore, and SD. Tanksley. 1992.
The detection and estimation of linkage in polyploids using single-dose restriction
ﬁ'agments. Theor. Appl. Genet. 83:294-300.

Wunsch, A. and J .I. Hormaza. 2002. Molecular characterisation of sweet cherry (Prunus
avium L.) genotypes using peach (Prunus persica (L.) Batsch) SSR sequences.
Heredity 89:56—63.

Yamamoto, T., K. Mochida, T. Irnai, Y.Z. Shi, I. Ogiwara, and T. Hayashi. 2002.
Microsatellite markers in peach (Prunus persica (L.) Batsch) derived from an
enriched genomic and cDNA libraries. Mol. Ecol. Notes. 2:298-301.

Yamamoto, T., T. Shimada, T. lmai, H. Yaegaki, T. Haji, N. Matsuta, M. Yamaguchi,

and T. Hayashi. 2001. Characterization of morphological traits based on a
genetic linkage map in peach. Breeding Sci. 51 :271-278.

129

TABLES AND FIGURES
Figure 1. Selected progeny from the ‘Ujfehe'rtoi Furtos’ X ‘Surefrre’ sour cherry

population illustrating variation for fruit size and shape.

 

130

Figure 2. Alignment of the mapped homeologous linkage groups from ‘Ujfehe'rtoi

F iirtos’ (‘UF’) sour cherry corresponding to LG 6 from the ‘T’ X ‘E’ reference Prunus
map. Only SSR markers from the ‘T’ X ‘E’ map are shown. Map distances in cM are
indicated to the leﬁ and marker names to the right of each vertical bar. Distorted loci at
the level of 0.1% level are denoted with a * following the name. Anchor loci between

‘UF’ and ‘T’ X ‘E’ are connected by a dotted line.

TxE LG6

n

7.0 NH, Ps7a2 UF LG6b
8.7 ”N CPPCT008 0.0 cpscroze 4e

 

 

17.5 UDP96-001
20.6 CPDCT0130

 

 

UF LG6“ 30.1 \ BPPCT008
\

T . -- -. .. - - - -. -. - -. .. 30.8 / epocroaza
0'0 BPPC 008 ° 33.3 \H/ BPPCT009

35.3 CPDCTO138

, ,. .. - -- -- ' ’ " 35.8 K cppcr015 25-7 MA027a
:2: ﬂ 2:321:14.
41:5 /‘\CPPc1023
44.7 CPPCT048

 

 

 

23.5 UDP96-010

 

55.4 .AP2M BPPCT025 f 45-1 Si
58.9 CPPcr047

 

 

 

37.1 Si 9: a.‘

 

."*-—. 72.0 UDP98-412

49.1 cppcroao 9: - ~ -- .. - -. - - .. - :1, 79.6 \=/ s. .’ 68.2 PrpFT
80.2 / \ CPPCTO3O

83.7 ’U\ CPPCT021

 

 

 

 

 

 

 

 

131

Ga 52552... :2... can:
8.. So :3 :3 8o 86 and 2d who 2.0 2.0

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. ........................ inﬂux. .3.92“....33
#1»ExamHum. my?HE... 4 o. m
Lanna. in 4...... a? ﬁn... w
in “yawn“..mmmmmmma. uni in» . m. m
mammamaaka ....... - cm .0.
Hunk”. W
2.33 - mm It
: U ,8 m
2:25 ..5 . mm
9.
Estate-E =2“. 5.: 3.22.323. :8:
08 man man now m9 m9 at 02 m9 3: on. o.» 3 ed ma... 3 3 o... m.» on on 3. 3
maﬁa”.ﬁﬁx. 0mm»atom.” ”+3.3.”xii. 3.3.3.”.33.x.” 3.3+”akin ............ 0 33+. 3+3. .H.H.H....H..:.H.H.H.....H.H.H.H.H...H.H.H.H.H. 03.x. $34”. .345. +3.3. +3.3. .m..x. o
”maymwmmmmmﬂmﬁﬂuxmmmmmwmmmmEmit“ .... an. may?45memymamx3.xii..m.mmmi”.19, 8
4“...wa H.N...mww... “ﬂaw. KNEW . N HHWWHMHM”wwwﬁw...“www.mwmmHF ”mum.” WWW... mum... many .1.” H I
* Maﬁaﬂaw.ii” in”: 9 m 3mmfromagingmumif.mﬁmuwﬁﬁ w
...: .mmmﬁ .. .2 m > maxim,inJuli. -2 .4
”HHHHHHJ...“ 1 ON 1 HHHHHHHHHWHuummmm. ..... H.N.“... 33...“. la
....H.H.H.H.H mu: .meﬁnnﬂgﬂumu - mp .09
”£3 -8 w ...: uh ............ w
- u v «
959:5 M— 8 A #44 - 8 .4
- 9. 2:25
9 on

 

 

.3283 .3 :32? 0.8 .9505? 28 35. 3:83 2: .8.“ 882 .voom-moo~ 88»
E 3.08% :08 Co 039 :88 05 so 323 m_ 5:33va 05. dots—smog .0595? x Ahab .mngm Envy—£5. 05 E .80on

03 :o @2385 33: ADV owﬁcoeom ﬁttinawao— van 6% 38:83 A<V Emma)» E5 40:23 58 no 22:554.? 35:62,,— .m 95me

132

 

 

m z: 83 8.0 4 035 8.0 4 43.0 8.0 4 3.4.0 85 £0 aging :2“.
SN 0.: 44 4 2m 2 4 EN no 4 0.3 85 £50 588% gen
2 E E 4 0.4 no 4 mm 3 4 3 43 as 3403 mam
5:232 858:4: 08084 2.0050. a: 05 £555: :3

00:00.60on

 

 

0mg .3035 Om “m 50.2

 

038020.“ 52.52%:0— was £30806 .5303

EE 90:0 5% Eu 0w:8 02g .30on 0:0 assuage 0.89.80 Ea 003? 23.05% 508 ANIV bmznﬂtﬁ 00:00 .00on ._ 030,—.

33

1

3.29.5.2“.
2: 3 o.» cs 3 an 3 an ad 3 o.o

_ L p p b r k p h

 

33 u a
~83 .. «a
:5: + £48.... a u

 

 

 

 

(98) Walla Ila-1:!

:5... 58...... ...E

 

 

 

 

 

 

gm 9% odu e..: 3: 3. O...
h p . h p Sso

386 v m
«3... a "a 4 8d

88... + £23.... a >

. o3
- one
, 8d
, 8.4
8..

3.29.5.2“.
0.2 3 3 S 3 on 3.

h ~ » b p h

od ON 0... 0.0

b p p

0.0

 

 

38... v a
.8... u «a
:3. + gnaw. u a

O

 

- 3
4 2:
, c9
, OS

I QmN

 

odn

 

(as) Input/tame: uma

(WW) “30w.” "'11:!

.83 2: so @8365 2a scam—$88

:an 8m 32? 053$ 28 530508 cots—oboe :oﬁmom 2F ADV owﬁcoohoq ﬂuriﬁmco— 98 “325% :45 98 A5 owﬁaoog

535%:2 98 2303 EE $3 888an 98 Emma? =3..— 52: ~30on page 58 5353 coca—oboe 80:5 .v 2sz

134

E 2.9.332“. 5.: 953. .23.. 86

 

 

    
 

 

 

 

 

 

 

  

 

  

 

 

 

 

 

 
 

 

 

 

 

0.9 9m Qm cs 0.0 o.m oé o.n od 0... oo 9 m w h m n v n N r
4 _ _ _ . _ _ 4 [ o JHNJ ...Hwﬁ‘mﬂrﬂ. ammuwu +333...Wia Human Maura... M magi o
o 80 .3 a: 8 o 000 ...-38. o I F d 33%. wwwmmmmmm ”wan”? wmmmmmmv ..ﬁ..ma. m mﬂmhmww.‘ m Wm...“ «may. 4 m
88 oo 8 .0 o 8 a: 1 N 6 Human ”wanna “$33; Hun. maﬁa“,
w ........................... 3...........AT OF M
, n M Mum”? ....ukhﬂu.“ .mmmmmmmmmm WWW... w
, e w. o... 9:. Km?mmmmmmmﬁm m. M
, m a .— m ,.5 in; ii, 8 w
... 15%...Gummy d
. c , o m inn.”Mai, mm m
0 O O O ‘ O : O 000 a 1 ...................
.83 m u n. «max, on m
. . 83...». - o m <4 in
. . . . 325+ 5:31. 4 a a mm
o 0 0 or ov

 

 

 

 

 

 

.83 05 :o 380%5 mm 33> o§Um-m ES E20508 cows—2.80
55.8on 2: Eu wen—=28 023i BE. .32 mos ~830— oﬁ maﬁa _ .038 AZ-— “ so 688 $3 32 mos >5on 3:232:
.EﬁwoﬁE of :o map—.8 E :39? 8a been? was A.,.SL .mwgm 63453.5. $5.8m oﬁ Sm £822 .38 By Amy maﬁa 98— no.8

98 Emma»; 3.5 :88 xcowoa gone 38 .8253 cots—oboe 80:: can A3 958 two. 208 mo cowsnmbmmu mocozcem .m oSwE

135

.36 v m 8 am; 8.5583 3 8:838 553» cos—«338 SEEN

 

 

 

 

136

o 885 a 33 0 £3 3 now an new a 8.2 o 3 o 3 n 3. :28 .28 “8 ,8
a 833 a £2 a “and o o. a a 8.8 o «.8 u 8 o E o 3 :28 ..m .28 an
a 82. a $2 a 82 a 98 a 98 o 8.8 a 3 a 3 B 3 8 um .28 an
a £3 a $3 a 82 a ma a 8.2 a oi a :V a 3 a S. ..8 .28 a8 ,8
25m 88 88 38 88 88 38 88 88 958 20:98
5 £23282 gem E25 53:36 :28 E 2%; gem

 

.voom-mocm who» 05 you 3.55:5 888% 88% was—82 95% 223% 0382. some Boa

~30on €23 58 .quSm. x Ari; Moran motmsoﬁo. 8m 03:822. ﬁgséwco— 28 £8256 .2303 E5 :32 .N 28;

5:0 Eco 8.? 222% a N

 

 

 

wed - 2: m SHUmmU o.m QB 92 $-38?
mod - mg.— NSHUmmU o.m o.~_ Wm _ A 3$§§
cod - QNN NSHUmmU 9N Q: Q: 3-3.3}. 325 ﬁgéﬁwcﬂ :3”—
mm._ - tom um YN m.mm o6. 3-3» 3 Emmy? 23,,—
Nm ”Soto 3% h832: GO: £28 5:32“ C2& Smac— SPA ‘50
3656 NM 3882 838232 xmoa DOA .5225

 

.92: SEEK 8:20.?“ .m. x L: 05 mo o 04 2 mcmucoamotg .3 ad 3on owe—c:

.623 6220:? 2: so 8282 203 do 852% =< 5:8 suséaﬁ Ea Emma :3 Eat 38 E 832% ﬁo .m 29¢

137

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8.3888 3”.
$.33

m
. . 3. in

9

a.” W
m ... 86588: 08
m w 879288 2:

m m .
o h. w h. m h. .m r“. w y. o ouréﬁoaam 3
54433221105 50::
904

 

'l

 

8832a ....
§~§5l

 

 

 

0‘- -. 0.. -. 0.. -- 0.. -. 0.. -. .-- -. 9.. -.

  

 

7}-|

2002

Shape 2003
She

Shap92004
H\\\\\ X\\:J—|

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SEnoBaao 3v
8 En
89883 08
meéohommo a9
muréoopomam o.o
«an: ..5

£3 2: 8 ﬁg. 45 2: .8 3:35 aha moi Ba 32:: moi

2865 macaw owe—c: Ea gnaw 05 5953 8am $533353 83 co woman .89» some how 888 3.82.“in DOA 83:3: saw—w

2: :o 8:: .momto> .uaowﬂ 05 2 9532.80 93 £3» =8..on Eat 838.. 58an macaw 05 so 33:0 .voom-mocm as?» 8m

f5; .85; Bumsﬁo. .8 3 Gd 3% 0mg: 8 av 2925 .3 Ba 2v 23% :3 923 58 5 do Esme; .c 2%;

138

Table 4. Comparison of sour cherry mean fruit weight (2002-2004) and mesocarp cell
numbers (2004) for ‘Ujfehértoi Fﬁrtos’ (‘UF’), ‘Surefire’, and small and large progeny
individuals from the ‘UF’ X ‘Sureﬁre’ population. Five fruit were measured for each

year and trait.

 

 

Mesocarp cell
3-yr average number (per

Genotype weight (g) radial section)2
2 (19) 1.72 28.8 a
2 (62) 1.96 32.0 a
Sureﬁre 5.12 37.6 b

UF 5.86 38.2 bc

2 (43) 7.25 41.6 c

 

zMean separation within colmnn by Tukey’s HSD at P < 0.05.

139

CHAPTER FIVE

QTL analysis of fruit size traits for the ‘NY 54’ X ‘Emperor Francis’ sweet cherry

(Prunus avium L.) population

140

ABSTRACT

Large fruit size is an essential production trait in sweet cherry (Prunus avium L.)
and an important goal of sweet cherry breeding programs. Identiﬁcation of the genomic
regions involved in variation for fruit size by quantitative trait loci (QTL) analysis would
provide an efﬁcient early selection method to increase the efficiency of sweet cherry
breeding. QTL analysis for fruit size traits was performed using the ‘New York 54’ (‘NY
54’) X ‘Emperor Francis’ (‘EF’) sweet cherry reciprocal populations. Fruit mesocarp cell
number and cell length, and mesocarp length were measured for 67 individuals in the
population. The parental means of ‘NY 54’ and ‘EF’ were signiﬁcantly different (P <
0.01) for all traits measured. Continuous variation was observed for all traits, although
the distribution was skewed toward the small fruit size exhibited by ‘NY 54’. No QTL
was identiﬁed for mesocarp cell number. However, ﬁve signiﬁcant QTL were identiﬁed
for mesocarp length and mesocarp cell length. All identiﬁed QTL affected the
phenotypic variance in the same direction as predicted by the parent.

For mesocarp length, one QTL (mIengthI) was identiﬁed on ‘EF’ linkage group 6
(LG 6) and one on ‘NY 54’ LG (y) (mlengch). The QTL mlength] explained 18.3% of
the total phenotypic variance. The closest marker to mlengthl was the AF LP marker,
EAT/MCCC-IOO, that was 0.1 cM from the LOD peak. The QTL mlengch explained
37.4% of the phenotypic variation, and was 3.9 cM from the nearest marker,
EAT/MCCC-ISO. Three QTL were identiﬁed for mesocarp cell length, on ‘EF’ LG 6
(clengthl) and ‘NY 54’ LG 6 (clengch) and LG (y) (clength3). The QTL explained
17.4, 16.8, and 16.8% of the phenotypic variation, respectively. The closest marker to

clengthl was CPPCT029-195, 0.1 cM from the LOD peak. The QTL clengch on ‘NY

141

54’ linkage group 6 was 0.1 cM from MAO40a-225, while clength3, on ‘NY54’ LG (y)
was co-located with EAT/MCCC-l 50. These QTL were identiﬁed with only one year of
phenotypic data on just 67 of the 190 progeny individuals genotyped for linkage map
construction, and therefore need to be veriﬁed in future years based on evaluation of all

190 progeny.

142

INTRODUCTION

Large fruit size is an essential component of fresh market sweet cherry (Prunus
avium L.) production as fruit averaging over 29 mm in diameter worth nearly twice as
much (Slkg) as fruit less than 24 mm in diameter (Whiting et al., 2005, 2006). Sweet
cherry ﬁ'uit size is a quantitative trait, presumed to be controlled by many separate loci
working in concert to produce the fruit size phenotype exhibited in a given cultivar.
Although the response to selection for increased fruit size has been relatively high in
sweet cherry breeding programs (F ogle, 1961; Hansche, 1966; Lamb, 1953; Matthews,
1973), little is known about the genetic control of fruit size.

Because of the long juvenility period and extensive land use requirements
associated with breeding perennial tree crops such as sweet cherry (Fogle, 1975), many
of the current cultivars are only a few generations removed from landraces from which
original domesticates were selected (Iezzoni et al., 1990). Marker-assisted selection
(MAS) could signiﬁcantly increase the efﬁciency of sweet cherry breeding, particularly
for fruit traits, where selection of favorable alleles based on DNA sequence rather than
phenotype would reduce the land use requirement and expense of maintaining seedling
individuals until the juvenility period has passed. Currently, the only trait for which
MAS is routine in sweet cherry breeding programs is self-compatibility conferred by the
mutated S4’ allele at the S-locus (Dirlewanger et al., 2004; Lewis and Crowe, 1954).

To implement marker-assisted selection for a quantitative trait, the relative
contribution of all the genes involved in expression of the phenotype must be described.
Quantitative trait loci (QTL) analysis is the method used to evaluate the number and

relative signiﬁcance of the potential genes inﬂuencing a given trait. In this type of

143

analysis, phenotypic trait data are combined with a genetic linkage map to identify
regions of the genome signiﬁcantly associated with mean differences in trait values. In
the ‘Rheinische Schattenmorelle’ (‘RS’) X ‘Erdi Botermo’ (‘EB’) sour cherry (P. cerasus
L.) population, two fruit weight QTL accounting for over 29% of the phenotypic
variation have been identiﬁed (Wang et al., 2000). An additional QTL for fruit weight
was identiﬁed in the ‘Ujfehértéi F iirtos’ (‘UF’) X ‘Sureﬁre’ sour cherry population (see
Chapter 4). More progress in QTL analysis has been made in peach [P. persica (L.)
Batsch], but there still have been relatively few analyses performed (Dirlewanger et al.,
1999; Etienne et al., 2002; Quilot et al., 2005; Yamamoto etal., 2001). However, among
these few populations, signiﬁcant QTL for fruit size measurements have consistently
been identiﬁed.

The phenotypic variability for fruit size exhibited by ‘New York 54’ (‘NY 54’)
and ‘Emperor Francis’ (‘EF’) (see Chapter 2) suggested that progeny from this
population would segregate for this trait. With the development of a genetic linkage map
for the ‘NY 54’ X ‘EF’ population (see Chapter 3), the ability to identify QTL for ﬁ'uit
size traits was possible. Therefore, the objective of this study was to identify potential

QTL for fruit size traits using the ‘NY 54’ X ‘EF’ sweet cherry population.

MATERIALS AND METHODS
Plant material

The sweet cherry population used for this study was developed from the
reciprocal crosses between ‘NY 54’ and ‘EF’. ‘NY 54’ was selected from wild P. avium

forests in Germany and introduced at the New York State Agricultural Experiment

144

Station, Cornell University (R.L. Andersen, pers. comm.) ‘EF ’ is a cultivated sweet
cherry of unknown origin, grown primarily for processed cherry products. From the
reciprocal crosses, 617 F1 individuals were planted at Michigan State University’s
Clarksville Horticultural Experiment Station (MSU-CHES). Details concerning

population development are provided in Chapter three.

Linkage map construction

Genotypic analyses and genetic linkage map development for the ‘NY 54’ X ‘EF’
population was described previously (see Chapter 3). Brieﬂy, simple sequence repeat
(SSR), ampliﬁed fragment length polymorphism (AF LP), sequence related ampliﬁed
polymorphism (SRAP), and S-RNase speciﬁc primers for alleles segregating at the self-
incompatibility locus were used to develop a 788 cM total distance F1 pseudo-testcross
linkage map for both ‘EF’ and ‘NY54’. SSR markers developed in several Prunus
species were used in the deve10pment of the ‘NY 54’ X ‘EF’ linkage map to facilitate
alignment and comparison with the Prunus reference map (Dirlewanger et al., 2004). A
total of 18 linkage groups were identiﬁed, with 12 of the 18 labeled according to the

reference map nomenclature based on shared markers.

Phenotypic analysis

A total of 67 individuals from the ‘NY 54’ X ‘EF’ population ﬂowered in 2005.
To ensure fruit set, all available ﬂowers were hand-pollinated with compatible pollen.
Whenever possible, at least ﬁve fruit from each individual were harvested aﬁer endocarp

hardening had occurred. The fruit from each individual were placed in storage vessels,

145

immersed in a formalin-acetic acid-alcohol solution (10:5:50 FAA; Ruzin, 1999) and
stored until sectioning. Radial mesocarp ﬂesh sections were created at the widest
diameter of the fruit as described previously (see Chapter 2). Microscopic analyses and
image analysis were used to determine the number of mesocarp cells per radial section,

mesocarp radial length, and mesocarp cell length (see Chapter 2).

QT L and statistical analysis

QTL analyses were performed using Windows QTL Cartographer 2.0 (Wang et
al., 2005) using composite interval mapping (CIM). CIM was run with model 6 of the
program using the forward and backward regression method. The LOD threshold for
declaring a QTL was determined by 1000 permutations for each trait at a signiﬁcance
level of P < 0.05, a priori. Estimates of the R-squared value indicating the explained
phenotypic variance for each QTL and the additive effect of the QTL were obtained from
the QTL Cartographer output. Graphical representations of the QTL were made using
output from QTL Cartographer and MapChart 2.1 (V oorrips, 2002). Analysis of
variance, correlations, and t-tests were performed using the appropriate function in SAS
statistical analysis software (SAS Institute, Cary, N.C.). Broad-sense heritability for each

trait was calculated using mean square values from analysis of variance (Fehr, 1987).

RESULTS AND DISCUSSION
The objective of this study was to identify potential QTL for fruit size traits using
the ‘NY 54’ X ‘EF’ sweet cherry population. This population was planted at MSU-CHES

in the spring of 2002, but the seedlings did not begin to fruit until 2005. In this year, only

146

67 individuals from the linkage mapping population had at least one ﬁ'uit available for
phenotypic measurements. Because of the lack of adequate fruit number in the ﬁrst
bearing year, the potential for animal predation, and the importance of mesocarp cell
number and mesocarp cell size to ﬁnal fruit size in Prunus (Chapter 2; also, Scorza et al.,
1991; Yamaguchi et al., 2002a, 2002b, 2004), emphasis was placed on the collection of
phenotypic data for these mesocarp cell number and cell size, and not ﬁnal fruit weight
and diameter. Thus, available fruit were harvested just aﬁer endocarp hardening had
occurred, prior to harvest maturity.

The parental means of ‘NY 54’ and ‘EF’ were signiﬁcantly different (P < 0.01)
for all traits measured (Table 1). However, the average value of the parents was only
signiﬁcantly different from the progeny mean value for mesocarp length (P < 0.05). All
traits exhibited continuous variation typical of a quantitatively inherited polygenic trait
(Fig. 1). For each trait, the distributions from the 67 progeny were skewed toward the
small values exhibited by ‘NY 54’. For mesocarp cell number, mesocarp length, and
mesocarp cell length, 87%, 94%, and 87% of the progeny averaged smaller than the mid-
parent value for the trait, respectively. Transgressive segregation occurred for all traits,
although progeny averaging greater than ‘EF’ were only measured for mesocarp cell
length (Figure 1). The broad sense heritability was high for each trait, although
measurements were made only in the year 2005 (Table 1). A positive linear correlation
existed between mesocarp cell number and mesocarp length (P < 0.0001) and between
mesocarp cell length and mesocarp length (P < 0.0001). A signiﬁcant negative
correlation was calculated between mesocarp cell number and mesocarp cell length (P <

0.05).

147

Signiﬁcant QTL were identiﬁed only for mesocarp length and mesocarp cell
length (Table 2). For mesocarp length, one QTL was identiﬁed on the bottom portion of
‘EF’ LG 6 (Fig. 2) and one on the bottom portion of ‘NY 54’ LG (y) (Fig. 4). The QTL
on ‘EF’ LG 6, mlengthl explained 18.3% of the total phenotypic variance and had an
effect as predicted by the parent, increasing mesocarp length by 0.31 mm. The closest
marker to mlengthl was the AF LP marker, EAT/MCCC-100, 0.1 cM ﬁ'om the LOD
peak. The QTL on ‘NY 54’ LG (y), mlengch, explained 37.4% of the phenotypic
variation, reduced mesocarp length by 0.41 mm, and was 3.9 cM from the nearest
marker, EAT/MCCC-ISO. Three QTL were identiﬁed for mesocarp cell length; on the
central portion of ‘EF’ LG 6 (Fig. 2), and at the bottom of ‘NY 54’ LG 6 (Fig. 3) and LG
(y) (Fig. 4). The QTL on ‘EF’ LG 6, clengthl, explained 17.4% of the phenotypic
variation and had an effect similar to that predicted by the parent, increasing cell length
by 7.62 pm. The closest marker to clength] was CPPCT029-195, 0.1 cM from the LOD
peak. The QTL on both ‘NY 54’ LG 6 and LG (y) explained 16.8% of the phenotypic
variation and reduced cell length by 7 pm. The QTL clengch on ‘NY 54’ LG 6 was 0.1
cM from MAO40a-225, while clength3, on ‘NY54’ LG (y) was co-located with
EAT/MCCC-lSO.

Fruit size QTL have been identiﬁed in other Prunus species. In sour cherry, two
fruit weight QTL accounting for over 29% of the phenotypic variation were identiﬁed in
the ‘RS’ X ‘EB’ population (Wang et al., 2000). However, there are no common markers
between the ‘RS X ‘EB’ and ‘NY 54’ X ‘EF’ linkage maps and comparison of the QTL is
not possible. In the ‘UF’ X ‘Sureﬁre’ sour cherry population, a QTL for fruit weight was

identiﬁed on the bottom of LG 6 (see Chapter 4), the same linkage group that QTL for

148

mean cell size and mesocarp length were identiﬁed in ‘EF’ and ‘NY 54’. Furthermore,
fruit weight QTL have been identiﬁed on both the bottom and central portions of LG 6 in
peach populations (Dirlewanger et al., 1999; Etienne et al., 2002; Quilot et al., 2005;
Yamamoto et al., 2001), similar to the location of mean cell length and mesocarp length
QTL identiﬁed on ‘EF’ LG 6 and ‘NY 54’ LG 6. This suggests potential conservation of
fruit size QTL in Prunus.

For all QTL identiﬁed in this study, the effect of the QTL was in the same
direction as predicted by the parental phenotype. However, this may not be an accurate
representation of fruit size QTL present, as the use of an F1 population limits the ability
to identify QTL in a heterozygous background (Conner et al., 1998; Grattapaglia and
Sederoff, 1994; Wang et al., 2000). In this case, for a QTL to be signiﬁcant, the effect
has to be sufﬁciently large enough to outweigh the variance of other potential loci
inﬂuencing the trait. If the parents are heterozygous for alleles at these loci, as they are
presumed to be in sweet cherry, both large and small fruit size QTL from both ‘NY 54’
and ‘EF’ are likely to segregate in the population (Wang et al., 2000).

The transgressive segregation and skewed population distribution toward small
fruit size suggests that small fruit size alleles are also present in ‘EF’. However, given
that only 35% of the ‘NY 54’ X ‘EF’ mapping population had ﬁ'uit for evaluation in the
ﬁrst bearing year, it is not possible to draw conclusions relative to the inheritance of these
traits. For example, if small fruit size is linked to precocious ﬂowering, it would
signiﬁcantly bias the phenotype of those seedlings available for analysis. Nonetheless,
the ability to identify QTL associated with fruit size with such a small population size is

encouraging. The stability of the QTL identiﬁed in this study is yet to be determined.

149

For example, because of the abbreviated fruit development period before the fruit were
sampled, the QTL identiﬁed for mesocarp length and mesocarp cell length may actually
be indicative of ripening date rather than overall ﬁ'uit size. The larger fruit from certain
individuals may simply have been closer to maturity. These questions will be answered
in the coming years when the full complement of ﬁ'uit traits can be analyzed in this

population.

CONCLUSIONS

In this preliminary analysis of fruit size traits for the ‘NY 54’ X ‘EF’ sweet cherry
population, QTL were identiﬁed in both parents for mesocarp length and mesocarp cell
length. Unfortunately, no QTL were identiﬁed for mesocarp cell number, a trait that has
been documented to inﬂuence fruit size in Prunus (Scorza et al., 1991; Yamaguchi et al.,
2002a, 2002b, 2004). This may be due to the limited number of individuals available for
QTL analysis in the first bearing year of the ‘NY 54’ X ‘EF’ sweet cherry population and
the skewed distribution toward the parent with fewer mesocarp cell numbers. However,
it is encouraging that QTL for both mesocarp length and mesocarp cell length were
identiﬁed on LG 6 of the parents in this population, given that fruit size QTL have
previously been located on this linkage group in other Prunus species (see Chapter 4;

Dirlewanger et al., 1999; Etienne et al., 2002; Quilot et al., 2005; Yamamoto et al., 2001).

150

LITERATURE CITED

Conner, P.J., S.K. Brown, and NF. Weeden. 1998. Molecular-marker analysis of
quantitative traits for growth and development in juvenile apple trees. Theor.
Appl. Genet. 96: 1027-1035.

Dirlewanger, E., A. Moing, C. Rothan, L. Svanella, V. Pronier, A. Guye, C. Plomion, and
R. Monet. 1999. Mapping QTLs controlling fruit quality in peach (Prunus
persica (L.) Batsch). Theor. Appl. Genet. 98:18-31.

Dirlewanger, E., E. Graziano, T. Joobeur, F. Garriga-Caldere, P. Cosson, W. Howad, and
P. Arus. 2004. Comparative mapping and marker-assisted selection in Rosaceae
fruit crops. Proc. Natl. Acad. Sci. 101:9891-9896.

Etienne, C., C. Rothan, A. Moing, C. Plomion, C. Bodenes, L. Svanella-Dumas, P.
Cosson, V. Pronier, R. Monet, and E. Dirlewanger. 2002. Candidate genes and

QTLs for sugar and organic acid content in peach (Prunus persica (L. ) Batsch).
Theor. Appl. Genet. 105:145-159.

Fehr, W.R. 1987. Principles of cultivar development. Macmillan, Inc., New York, New
York.

Fogle, H.W. 1961. Inheritance of some ﬁ'uit and tree characteristics in sweet cherry
crosses. Proc. Amer. Soc. Hort. Sci. 78:76-85.

Fogle, H.W. 1975. Cherries. In: Janick, J. and J .N. Moore (eds) Advances in Fruit
Breeding. Purdue University Press, West Lafayette, Ind, pp. 348-366.

Grattapaglia, D. and R. Sederoff. 1994. Genetic linkage maps of Eucalyptus grandis and
Eucalyptus urophylla using a pseudo-testcross: mapping strategy and RAPD
markers. Genetics 137:1121-1137.

Hansche, P.E., V. Beres, and RM. Brooks 1966. Heritability and genetic correlation in
the sweet cherry. Proc. Amer. Soc. Hort. Sci. 88:173-183.

Iezzoni, A., H. Schmidt, and A. Albertini. 1990. Cherries (Prunus). In: Moore, J .N. and
J .R. Ballington Jr. (eds) Genetic Resources of Temperate Fruit and Nut Crops,
vol. 1, ISHS, Wageningen, the Netherlands. pp 111-173.

Lamb, RC. 1953. Notes on the inheritance of some characters in the sweet cherry
Prunus avium. Proc. Amer. Soc. Hort. Sci. 61:293-298.

Lewis, D. and L.K. Crowe. 1954. The induction of self-fertility in tree fruits. J. Hort.
Sci. 29:220-225.

151

Matthews, P. 1973. Some recent advances in sweet cherry genetics and breeding. Proc.
EUCARPIA Fruit Section Symp. 5. Topic Fruit Breeding. Canterbury: 84-107.

Quilot, B., J. Kervella, M. Genard, and F. Lescourret. 2005. Analysing the genetic

control of peach fruit quality through an ecophysiological model combined with a
QTL approach. J. Exp. Bot. 56:3083-3092.

Ruzin, SE. 1999. Plant Microtechnique and Microscopy. Oxford University Press,
New York, NY.

Scorza, R., L.G. May, B. Pumell and B. Upchurch. 1991. Differences in number and
area of mesocarp cells between small- and large-ﬁ'uited peach cultivars. J. Amer.
Soc. Hort. Sci. 116:861-864.

Voorrips, RE. 2002. MapChart: Software for the graphical presentation of linkage maps
and QTLs. J. Hered. 93:77-78.

Wang, D., R. Karle, and A.F. Iezzoni. 2000. QTL analysis of ﬂower and fruit traits in
sour cherry. Theor. Appl. Genet. 100:535-544.

Wang, S., C. J. Basten, and Z.-B. Zeng. 2005. Windows QTL Cartographer 2.5.
Department of Statistics, North Carolina State University, Raleigh, NC.
(http://statgen.ncsu.edu/qtlcart/WQTLCart.htm).

Whiting, M.D., D. Ophardt, and J .R. McFerson. 2006. Chemical blossom thinners vary
in their effect on sweet cherry fruit set, yield, fruit quality, and crop value.
HortTechnology 16:66-70.

Whiting, M.D., G. Lang and D. Ophardt. 2005 Rootstock and training system affect
cherry growth, yield, and fruit quality. HortScience 40:582-586.

Yamaguchi, M. I. Sato, K. Takase, A. Watanabe and M. Ishiguro. 2004. Differences and
yearly variation in number and size of mesocarp cells in sweet cherry (Prunus
avium L.) cultivars and related species. J. Japan. Soc. Hort. Sci. 73:12-18.

Yamaguchi, M., T. Haji, M. Miyake and H. Yaegaki. 2002a. Studies on the varietal
differences and yearly deviation of mesocarp cell numbers and lengths and fruit
weight among commercial peach [Prunus persica (L.) Batsch] cultivars and
selections, wild types, and their hybrids. J. Japan. Soc. Hort. Sci. 71:459-466.

Yamaguchi, M., T. Haji, M. Miyake and H. Yaegaki. 2002b. Varietal differences in cell

division and enlargement periods during peach (Prunus persica Batsch) fruit
development. J. Japan. Soc. Hort. Sci. 71 :155-163.

152

Yamamoto, T., T. Shimada, T. Irnai, H. Yaegaki, T. Haji, N. Matsuta, M. Yamaguchi,
and T. Hayashi. 2001. Characterization of morphological traits based on a
genetic linkage map in peach. Breeding Sci. 51:271-278.

153

TABLES AND FIGURES

Figure 1. Frequency distribution of fruit mesocarp radial cell number (A), cell length
(B), and mesocarp radial length (C) measured at endocarp hardening for 67 progeny in
the ‘NY 54’ X ‘Emperor Francis’ (‘EF’) sweet cherry population in 2005. Means for the

parents ‘NY 54’ and ‘EF’ are shown by arrows.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2° ‘ A NY 54
>.
c
8. 15 . ,,,,,, v
2 : ; : : 1 :31313331333334 :5t3i3i3:3:1:3
9- .._.. -:~:.:-:- :::::
.6 .:.:.:,‘.:.:.{.j :.:.:.:.‘.' ..............
.. 10 ‘ 11131221}313‘E131ii33313i{3213131313} tar-fr:
“ .......... -1-i :-:-:-:-:-:: ':-:»:-:-:«:v:
'2 -;~:-:-:::~:?i3:::}:5;?}:;:1:313:33}:-‘tjziziffji:
3 5 12;}:2;111(3)].2..-§;§3;is§:§:§2§i:isizizizi2i:,,,_.,,,_ EF
0 »:-:7:3:5:?;3::?:5:?:?:?:':.71?:2T:'~:3:i‘it-:':3:3:3:1:‘?i?:?:?:1:$'1T3:31:T‘11't»:T‘F:i:1:T:?::-':?:7'3:3:3: --------------
22 25 27 29 31 33 35 37 39 41
Mean cell number (per radial section)
2° ‘ B V54
>.
= .':.:::::'.:::“:::::::::'::Z:
8» 15 ‘ 3:115:7132131252}
9 ......ft'iifizifii3:731:55!"
0- -:-:.:-:':-:--:-:-:-:- ::::
'5 ........
h 10 * 7323:2331:}:3:f:?:5:§. 3:53-3ij
e :-,»;-:-::.-.:,::. _:.:: ........
% 3:3:3:1:3:3:3: LIFE-71:3:57331133333"? ':-:-:-:-:':-:
2 s éi-iiiziiéiif :g:§:;:§:;:§:§ 3.2i2:i32i2?%i EF
0 :':':':’t':':' 1:535:11: :113‘3515: 55:33:32? 2:135:32? ‘?'?'3‘3'3‘3'7 7:33:3131'1- . 1
45 50 55 60 65 70 75 80 85 90
Mean cell lennth turn)
20 J
g C
8. 15 ~
8
a.
"6
h 10 “
o
a
E
z 5 r
o ’ 1

 

 

 

 

 

 

 

 

 

 

 

1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0
Mean radlal mesocarp length (mm)

154

 

 

a? an. o: a m. 5 Es a oi 4:. a mom :3 3.3 522 =8 :82
new a: 2:. a a: one a ”2 So a _S 85 9:5 3:2 88802
33 SN 8.9 u. com a: a 3:. N; a. in and cage .8 =8 98882
.5552 5:525 3&8; “a 3 >2 f: 3.535; as...

OmGomuﬁmOum

 

owcﬁ .98on Om “m :32

 

$55.22 88028 3 Swan— =8 :88 38808 28 Armam— Euﬁ 38808 .5983: :8 368 93808 55

.923 8026 How owcﬂ 02? .86on EB .mcoumgow 93:83 can 82.; 399823 :88 .ANZV b2595: 8:3 365 A 03am.

155

do Eco 2:? 2:63.. s ...

 

 

 

8.5 - Qw— om TUDOSE.<@ NM odm was A S>Z 3&2»?
cos - WE mwméovoemz NM Mao m.m o >2 gums»?
. . . . . Es 532
me n v 2 3738.950 w m _ mm 5 mm 0 mm 3&2»? =8 :82
:3 - vsm 926002.31"— o.m odm 0.2 A 3>Z mime»?
€25 595.
Ed 3: oo_-0002\._.<m m.m 5mm ”2 e um 3&2»? 98802
a 38% axe 338 m3 323 858g :23 5mg: anew H8
3850 NM 6832 Basia—z anon DOA 3285 omen—:5

 

.38 3% 2: 5 use Em;

288m 839:? can .vm >2. 5 wEEEﬁ 93265 3 Swan.— =3 :88 28 Emce— Evﬂ 98088 no“ 388% .50 .N 033.

156

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5.8.0 1 ..2. ‘ 5.8-0 1 :2.
0850.205 /1\ 0.2 08-50205 /1\ 0.2.
..... . . . mimxwtg 0.00 0,3039%“. 0.00
03-83% 0.8 2332.2 0.8
. . . h 3.338800: JuVmS 3.858800: /1\ 0.8
.. H ..n. 87000.55 \ / 0.8 87000220 \1/ 0.8
.. m a

.H m m

H .. m H
H .. ........ . 8.085000 / \ 0.3 m . 8.08580 /1\ 0.9.
H . #35555 \L/ 0.8 m . «505556 \1/ 0.8
m .. .. 8.088000 «.8 .. 02-880000 N8
H . 0350.205 0.8 . 3050.205. 0.8
n . .. 87000250 0.8 .. 8.000.230 FR
M . 5-805000 0. a . 5-880000 0. R
m .. 05-05220 0.2 08-05220 0.0.

m— m . 2 38800: E < f . 2 38800: E
. . . . .. . . . .. .. «2.50.0020 0.0 . . . . . . . . . .. «8.50.0020 0.0
mumumumwmum ”umuuummmum
:3 00: ...: no: 00: :0

€80 2: S 0:8 do 2.. 8: ates Ewe :03 05 80:5 00.: use? 9.20 00%: us... is» 2:
50.50: 000m 05:05.50: 82 :0 :08: 20:00:: 00:35:90 904 05 0:00:05 gnaw 05 :0 00:: _00_t0> .c O‘— Akmp 20:8:

00:09:? :0 @3020: 900005 .0 Amv 5w:0_ 300.. 900000.: ::0 A<v 5w:0_ :00 80080:: 0:5 :8 APO 0:005:3m .N 053.:

157

 

0000-0 0.:
«3.000220 X 0. :.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

V W 007502005 0.2.
A w' n..««.003§_ «.00
C
n
a
M
87000230 0.00
8.105230 0.00
8.085000 0.?
5.280000 0.00
2708:0000 «.«0
02085000 3«
«$3000.20 0.0
I I I ‘ I I I‘ I I d F” PI Psglon one
5 4 4 3 3 2 2 1 1 0 0
90‘— vm >2
904

.0080 20 8 0.80 00.0 20 80 0:33 age 000.«

0:0 900:5 Q04; 0:00:05 000% 0w0x:= 0:0 200% 05 :00250: 000m 0:232:00: co? :0 00000 20:00:: 00:00a_:wmm Q04 05

0:00:05 500% 05 :0 00:: _00_t0> .c 04 .3 >2. :0 3:030: 900005 00 5w:0_ :00 90080:: :05 :8 5.0 Hgomﬁmmm .m 05mE

158

 

 

03.12.0205 ...Nm 03.12.0265 fun

 

 

 

 

 

 

om TOOUEBKN QQN cm Toooitkm mdN
m h
m“ m.
0 .
n
s a
m“ m

 

a:
<

 

 

 

 

 

 

 

 

 

 

 

 

 

. . . . . . .J q a q om~.<.52\o<m 0.0 J Tq 4 0 . + . . oo~.<0.oszo<m 0.0
0505050505050. 0.505.050.5050.
6554433221100 54433221100

000 $0: 8 >z 000 30: 8 >2

.0080 20 8 0.30 0,5 20 80 02.25 @000 000-« Ea 30:8 00.: 3805 0:20 009.0:
0:0 000% 05 :003000 000m 8:003:80: 02: :0 080: 2080:: 0050c_:w_m Q04 05 0000005 gnaw 05 :0 80:: _00E0> .3

0: 0m >2. :0 w:_:0000: 900005 :0 R: «3:0. 300: 900808 0:0 03 0&5. :00 30080:: 00.0 :0.“ 15:0 0:00am:m_m .0 0SmE

159

CHAPTER SIX

SUMMARY AND CONCLUSIONS

160

In an increasingly competitive market, production of high-quality sweet (Prunus
avium L.) and sour (P. cerasus L.) cherries has become essential. For ﬂesh markets,
large fruit size is critical for profitable production. Increased and consistently large fruit
size is an area of continued horticultural and physiological research with existing
production cultivars. However, breeding efforts will continue to be an important avenue
to increase cherry fruit size. Unfortunately, the long generation time of perennial tree
fruit crops such as cherry and the quantitative nature of the fruit size trait make breeding
for improved fruit size inefficient. A better understanding of the cellular basis for fruit
size potential, both among and within cultivars, could increase selection efficiency in
cherry breeding programs. Further increases in efﬁciency would be realized if initial
selection for fruit size was based on genotypic markers rather than phenotypic expression
of the trait after the juvenility period has passed.

The research reported herein examined ﬁ'uit mesocarp cellular differences
between cultivars with a wide range of average fruit sizes and within fruit from single
cultivars exhibiting signiﬁcant size differences (Chapter 2). Mesocarp cell number
differences between cultivars were correlated with increasing fruit size, while mesocarp
cell size was not. However, differences in cell size were observed between cultivars. For
example, mesocarp cell sizes in fruit from ‘Selah’, the cultivar examined with the largest
fruit size, were not signiﬁcantly different than those in ‘New York 54’ (‘NY 54’), the
cultivar with the smallest fruit size. ‘Bing’, ‘Regina’, and ‘Emperor Francis’ (‘EF’), all
with ﬁ'uit sizes falling between ‘Selah’ and ‘NY 54’, had significantly larger cell sizes.
Mesocarp cell number was environmentally stable and did not differ when fruit thinning

treatments were applied; whereas mesocarp cell size contributed to the increase in fruit

161

size gained from reduced crop load. The low environmental variance exhibited for
mesocarp cell number makes it an obvious selection criterion for improved fruit size in
cherry breeding programs.

To further examine the genetic control of fruit size in sweet cherry, a linkage map
was constructed for reciprocal crosses between ‘NY 54’ and ‘EF’ (Chapter 3). These
parents were selected to represent the genetic differences accumulated during the
domestication of sweet cherry, by crossing a wild example (‘NY 54’) with an early
domesticate (‘EF ’). Simple sequence repeat (SSR) markers developed in other Prunus
species were used extensively to facilitate map comparison within Prunus. Although the
map is incomplete and only the ﬁrst year of fruit phenotypic data was available, a
preliminary quantitative trait loci (QTL) analysis identiﬁed fruit size QTL, predominantly
on LG 6 of both parents (Chapter 5). This linkage group represents an important
chromosome in Prunus, as it also contains the self-incompatibility locus (S), and ﬁ'uit
size QTL on LG 6 have been identiﬁed previously in peach [P. persica (L.) Batsch].
Fruit size QTL on LG 6 were examined further using a targeted mapping approach,
whereby only SSR loci previously mapped to LG 6 in other Prunus species were used to
develop a linkage map for the ‘Ujfehértoi Fiirtds’ (‘UF’) X ‘Sureﬁre’ sour cherry
population (Chapter 4). A QTL three cM from the S locus explaining 26.4% of the
phenotypic variation was identiﬁed in the ‘UF’ x ‘Surefire’ population. Additionally, a
fruit shape QTL was also located on LG 6, co-segregating with the CPSCT012 marker

and explaining up to 22.6% of the phenotypic variation for fruit shape.

162

AAAAAAAAAAAAAAAAAAAAAAAAAA

llllllWlllHllHlllllllIllllllllllllllHllllHllWill