kw ‘, with

 

 

Mpsr‘p;

ass-Ween; (:1

:4

 

 

lrfir
u

 

lbw-A ;

 

-\
a
w
v:
4
§

Ft?!

5

 

‘4;-

 

!\:§‘

\‘pq’CjQM‘

    

 

 

 

 

 

 

 

 

 

\.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 
 

 

 
 

 

   

 

   

 

 

 

 

I .
.,.

 

 

 

 

 

 

“4| ‘
.v.v-.~.

 

 

 

 

 

 

I

y \ «
u - 4 \
.

 

 

 

 

" will?“ twilliiillMxi ‘ 2 vi 0 l 3 70/

Date

 

 

 

 

293 0060

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the

thesis entitled

PRINCIPAL COMPONENT ANALYSIS AND ISOZYNE ANALYSIS
OF SWEET, SOUR, AND GROUND CHERRY GERMPLASM

presented by

Kimberly Howell Krahl

has been accepted towards fulfillment
of the requirements for

M. S . degree in Horticulture

 

 

W¢W

Major p fessor

August 3, 1989

 

O~7639

MS U is an Affirmative Action/Equal Opportunity Institution

PLACE ll RETURN BOX to remove this checkout from your record.
TO AVOID FINES mum on or baton due duo.

   
 
  

DATE DUE DATE DUE DATE DUE

 

 

 

 

QCHSSQW

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU In An Affirmative Action/Equal Opportunlty Institution

PRINCIPAL COMPONENT ANALYSIS AND ISOZYHE ANALYSIS OF SWEET, SOUR,
AND GROUND CHERRY GERMPLASH

by
Kimberly Howell Krahl

A THESIS

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

MASTER OF SCIENCE

Department of Horticulture

1989

(£304.03 '50

ABSTRACT

PRINCIPAL COMPONENT ANALYSIS AND ISOZYME ANALYSIS OF SWEET, SOUR, AND
GROUND CHERRY CERMPLASM

BY
Kimberly Howell Krahl

Principal component analysis (PCA) was employed to examine
variation between sweet cherry (£1333; axing), sour cherry (2.
ggrgggg), and ground cherry (E- frgtigggg) and within twelve sour
cherry families. Sweet and ground cherry, the putative progenitors of
sour cherry, occupy extreme opposite positions on the PC scattergram.
Sour cherry families occupy various intermediate positions. However,
ground cherry shares the extreme negative position on PCl with a sour
cherry family. Thus, using the characters in this study, it was not
possible to distinguish ground cherry from sour cherry. Continued gene
flow between these two tetraploid species may account for this
intergradation. The PCA reveals that a great range of morphological
variants are available for selection within the MSU sour cherry
germplasm collection.

An isozyme analysis was conducted to determine protocol and to
investigate possible polymorphisms between and within sweet, sour, and
ground cherry. Sufficient polymorphisms were revealed in several

enzyme systems to be used as genetic markers in future studies.

For Christie: friend, therapist, mentor. The literal truth is
that this would not have been possible without you. Thank you for
being on my side all the times I couldn't be there. Thank you for
loving me. I love you.

For little Fessie. My little buddy who came with me to graduate
school. I miss you and I love you. See you sometime.

And for Frank, whose love and laughter have seen me through on a

daily basis. I love you, C.

iii

ACKNOWLEDGMENTS

I thank Steve Krebs for helping me with the isozyme work. More
than that, I thank Steve for being who he is; a generous, kind, and

supportive person who has a true gift for teaching.
And many thanks to LuAnn Gloden, without whose kindness, generous

help, and expertise I would have been at a total loss when confronted

with getting this thesis in proper form.

iv

TABLE OF CONTENTS

Page

List of Tables ................................................. vi

List of Figures ................................................ vii
Chapte; One

Principal Component Analysis of Sweet, Sour and Ground Cherry... 1

Introduction .............................................. 2

Literature Review ......................................... 3

Cytology and Origin of Sour Cherry ..................... 3

Applications of Multivariate Analysis .................. 7

Materials and Methods ..................................... 19

Plant Material ......................................... 19

Characters Measured .................................... 19

Data Analysis .......................................... 23

Results ................................................... 23

Discussion ................................................ 33
Chapter Two

Isozyme Analysis of Sweet, Sour and Ground Cherry .............. 38

Introduction .............................................. 39

Literature Review ......................................... 39

Applications of Isozyme Analysis ....................... 39

Materials and Methods ..................................... 56

Results and Discussion .................................... 59

List of References ............................................. 75

Appendix
Segregation Numbers for Seed Tissue Zymograms .................. 84

Iable

LIST OF TABLES

W

Pedigrees of parents of the sour cherry families .......

Names, abbreviations, and progenies evaluated for
the clone and each family ..............................

Eigenvalues, proportions of variance, and cumulative
variances from the principal component analysis of the
cherry clones and families .............................

Eigenvectors of the first 3 principal component axes
from principal component analysis of the cherry
clones and families

000000000000000000000000000000000000

Clone and family means and ranges of morphological
characters. The range values are of the progeny
means averaged over samples ............................

Chapter Two

Staining solutions for enzymes ..: ......................

vi

20

22

24

26

28

6O

Figure

LIST OF FIGURES

Chapte; Opp
Positions of PC scores of sweet, sour and ground cherry

family means on the first three PC axes. Abbreviations
as in Table 1 ..........................................

Chapter Two

Zymograms of sour cherry, ground cherry and sweet
cherry leaf and seed tissue ............................

vii

64

CHAPTER ONE

Principal Component Analysis of Sweet, Sour and Ground Cherry

INTRODUCTION

Sour cherry is considered to have originated in Asia Minor and to
have spread west to Europe and east into what is now the Soviet Union.
Sour cherry has been grown for many centuries in Europe and Asia. The
French cultivar 'Montmorency' is one of the oldest of cultivated
varieties and was grown in Europe hundreds of years before there was
any commercial fruit production in the United States. It was brought
to this country during colonial times. Although Hedrick (1915)
reported that 270 sour cherry cultivars were once available in this
country, by 1915 'Montmorency' was well on its way to becoming
essentially the only commercial variety grown. For many years U.S.
sour cherry production has been a virtual monoculture of 'Montmorency',
and breeding programs have produced only a few cultivars ('Northstar',
'Meteor') which are primarily of interest to homeowners rather than
commercial growers. While 'Montmorency' has generally been a
dependable producer, it still has quite a few shortcomings. In
Michigan, the state producing approximately 80% of the U.S. sour cherry
crop, 'Montmorency' yields are dramatically reduced one year out of
three due to spring frost injury. The cultivar is susceptible to
several diseases, requiring many protective sprays, and only about 30%
of the flowers set fruit.

Sour cherry landraces and native cherry populations in Europe and
the Soviet Union have provided a great deal of diversity for European

breeding programs, and the resultant varieties exhibit some of this

diversity. The sour cherry germplasm collection at Michigan State
University, containing material collected in Yugoslavia, Bulgaria,
Hungary, and Poland plus germplasm from western Europe and the Soviet
Union, is much more diverse than collections previously available in
this country, and our breeding program therefore has high potential to
develop superior cultivars for commercial sour cherry production in the
United States.

A previous study using multivariate analysis to examine diversity
in this germplasm collection suggested that existing variations
reflected gradations between supposed progenitors of sour cherry --
sweet cherry and ground cherry. In an attempt to test this hypothesis,
the present study included representatives of sweet and ground cherry
in addition to sour cherry families exhibiting a range of variation.
The most obvious difference between sweet and ground cherry is gross
morphology. From leaf size to ultimate height, sweet cherry is
consistently larger than ground cherry. My purpose was to examine
several characters affecting habit and size to describe the range of

variability encompassed by the collection.

LITERATURE REVIEW
1 0 on h
firpppg peppsus L., sour cherry, is a member of the family
Rosaceae, subgenus Cerasus, section Eucerasus. A11 Epppps species have
a basic chromosome number of x-8. Sour cherry is a tetraploid with
2n-4x-32 (Darlington, 1928; Kobel, 1927; Okabe, 1928). Kobel (1927)
suggested that sweet cherry (B. agipm L.), a diploid species with

2n-2x-16, was one of the progenitors of sour cherry. He tentatively

suggested either ground cherry, B. fgppigpga Pall., a tetraploid
species (2n-4x-32), or P. fpppggpgpp Schneid. as the other progenitor.
Darlington (1928) stated, "it is not impossible" that sour cherry arose
due to hybridization between sweet cherry and a tetraploid species such
as ground cherry.

Although Raptopoulos (1941) concluded that sour cherry was an
autotetraploid, cytologists Kobel (1927), Prywer (1936), Hruby (1950),
Blasse (1957), and Berg (1958) independently reported it to be an
allotetraploid. Hruby (1950) cited the low percentage of quadrivalents
observed during meiosis as evidence of allotetraploidy. Examination of
250 pollen mother cells during meiosis revealed that 90% of the
chromosomes paired as bivalents, while only 1.2% grouped as
quadrivalents (Hruby, 1950). Kobel (1927), Prywer (1936), and Hruby
(1950) concluded that sour cherry originated by hybridization of two
ancestral diploid species which subsequently doubled. Hruby (1962)
concluded from his cytological work on sweet and sour cherry and
hybrids between the two that both specieS'were of hybrid origin and
that they shared one ancestor but not the second. He maintained that
while sweet cherry arose as a result of hybridization between two
similar genomes, sour cherry arose from sexual polyploidization between
two dissimilar genomes.

Olden and Nybom (1968) attempted to test experimentally the
hypothesis that sour cherry resulted from hybridization between sweet
and ground cherry by crossing these two species. Both diploid and
tetraploid sweet cherries were used, the tetraploid being produced
through colchicine treatment of a wild-type sweet cherry. Only

triploid seedlings were obtained using the diploid sweet cherry. These

resembled ground cherry in growth habit and leaf shape. The hybrid
tetraploid seedlings obtained by crossing ground cherry with the
tetraploid sweet cherry were morphologically very similar to wild and
cultivated sour cherries. They closely resembled sour cherry in such
characters as leaf shape and size, and leaf margin serrations and were
sexually compatible with sour cherry. Olden and Nybom (1968) reported
that leaf phenolics in the "resynthesized" hybrid and naturally
occurring sour cherry were quite similar in composition. They
concluded that the morphological and chemical evidence, along with the
fact that the natural distribution of sour cherry corresponds with the
area where the natural ranges of sweet and ground cherry overlap in
southeast Europe and southwest Asia, strongly supports the hypothesis
that sour cherry did in fact originate as a hybrid between the diploid
sweet cherry and the tetraploid ground cherry.

This hypothesis is also supported by evidence that unreduced
pollen grains occur in sweet cherry (Hruby, 1939; Galetta, 1959;
Iezzoni and Hancock, 1984). In addition, sour cherry exhibits
codominant expression of the MDH isozyme banding patterns of sweet and
ground cherry (Hancock and Iezzoni, 1988) and chloroplast RFLPs of
several sour cherry cultivars are similar to those of ground cherry (A.
Iezzoni, pers. comm.). Thus sour cherry probably resulted from
hybridization between the ancestors of ground cherry as the maternal
parent (contributing a normal, reduced gamete) and sweet cherry as the
paternal parent (contributing an unreduced gamete).

de Candolle (1882) hypothesized that sweet and sour cherry had
both originated in the area from the Caspian Sea to western Turkey.

Both species subsequently spread westward to Europe, but not to the

same extent. Sweet cherry "extended further and at an earlier epoch
and has become better naturalized". de Candolle concluded that sour
cherry probably arose from sweet cherry in prehistoric times. Present
day evolutionary theory supports de Candolle's conclusion; i.e., the
species with lower chromosome number is assumed to predate the species
with higher chromosome number. Vavilov (1951) hypothesized that sweet
cherry originated in the Near East center of origin (including the Near
East and Asia Minor, Transcaucasia, Iran and the highlands of
Turkmenistan). He suggested that the center of origin of sour cherry
might be Asia Minor. Watkins (1976) reasoned that if ground cherry was
an ancestor of sour cherry, then the sour cherry must have evolved
before sweet and sour cherry spread from western Asia to Europe,
because western and central Asia is the center of origin for ground
cherry.

Kolesnikova (1975) divided sour cherry into two ecogeographic
groups, Western European and Middle Russian, based on morphology and
winter hardiness. The Western European group, which included some Duke
cherries (sweet cherry x sour cherry), had better eating quality and
lower winter-hardiness. The majority of these varieties resembled
sweet cherry in growth habit (single-trunk tree form with larger
leaves). The Middle Russian ecotype had poorer eating quality but
improved winter hardiness. Many of these varieties resembled ground
cherry in having a multiple-stemmed shrub form with smaller leaves.

Yushev (1975,1977) used various morphological characters of leaves
and fruit to study 119 cultivars of sour cherry. He concluded that the
most significant traits for classifying cultivars were leaf form and

size, serrations, pubescence, and glands. He divided the sour cherry

cultivars studied into two groups, Western European and Eastern
European. The Western European group exhibited leaf characteristics.
typical of sweet cherry; seventy percent of the cultivars had large,
obovate leaves with large double serrations, large glandules and
persistent pubescence -- all characteristics typical of sweet cherry.
Also, 46% of the large fruited cultivars were of the Western European
type. Typically the fruit was heart-shaped with light colored juice
and medium to large pits. The Western European group generally
exhibited lower winter hardiness. The majority (56.53) of the Eastern
European group of cultivars had medium-sized ovate leaves with small
double crenations -- characteristics typical of ground cherry. This
group contained many cutivars with small to medium-sized pits, and
exhibited higher levels of winter-hardiness. Yushev hypothesized that
the similarity of leaf morphology of the Western European cultivars to
sweet cherry indicated that sweet cherry played a large part in the
origin of these varieties, whereas the similarity of leaf morphology
of the Eastern European cultivars to ground cherry indicated that
ground cherry participated widely in the origin of these varieties.
n u t v s

Multivariate analysis encompasses the statistical methods used to
analyze data when multiple variables or characters are scored for each
unit studied (Afifi and Clark, 1984). Various types of multivariate
analysis have been used in recent decades for taxonomic purposes.
Numerical analysis is used in an attempt to create taxonomic
classification that is quantifiable, repeatable, and objective.
Principal components analysis (PCA) and cluster analysis are types of

multivariate analysis that are employed to reveal underlying structure

in a multiple-variable data set. These methods require no prior
knowledge of the origins of the units. Both analyses can be used to
detect groups within the units. PCA tends to display the general
pattern of phenetic diversity at the expense of detail, and is best
employed to reveal relationships between groups. Conversely, cluster
analysis most clearly illustrates detailed patterns of phenetic
similarity at the expense of the overall pattern of diversity, and is
best employed-to examine relationships within groups or between units
in different groups.

PCA was first used by Hotelling (1933) in educational testing.
Since then PCA has been used in psychological, medical, and biological
applications (Seal,l964; Morrison,l976). Using PCA, every unit to be
studied is scored for a chosen number (n) of variables or characters.
Each variable or character can be imagined as representing an axis in
space. Each unit has a value on each variable axis. Since each unit
represents a unique combination of variable "scores", every unit can be
represented by a point in a multi-dimensional space which is defined by
n coordinate axes. The original data create an ellipsoidal cloud of
points (representing units) in n-dimensional space. PCA is used to
simplify the description of this set of correlated variables. It is a
method of transforming the original variables into new, uncorrelated
variables. These new variables are called principal components. The
major axis of the ellipsoid becomes principal component 1 (P01). The
second axis (PC2) is at right angles to the first. Each successive
axis is orthogonal to all others. The number of original axes eqhals
the number of principal components. Each PC is a linear function of

unit scores for the original n variables which have been transformed

such that the mean of each transformed variable is zero and the
standard deviation is unity. The total variance represented by the
original axes is preserved by the principal components. With
standardization of the data and rotation of the original axes, the
first axis becomes the major axis of the ellipsoidal cloud. This axis
accounts for the greatest amount of total variance and is called PC1.
Each successive PC is at right angles to all others and accounts for a
decreasing amount of the total variance. Thus PCA simplifies the
examination of a data set with multiple variables. Since the first few
PCs account for a large proportion of the total variance,
interpretation of the relationships between units usually can be based
on just the first few PCs. Principal components are interpreted by
examining the variables with high coefficients for that PC. A.high
coefficient for a variable on a PC is an indication of high correlation
between that variable and the PC. In other words, a variable with a
high coefficient on a particular PC makes a high contribution to the
variance of that PC. Given a high coefficient of a variable on a PC,
examination of that PC alone can proVide a picture of a high proportion
of the total variance for that particular variable.

PCA is best employed to depict overall diversity, phenetic
diversity in particular. Phenetics is the classification of
populations of organisms based on their relative phenotypic similarity
without regard to evolutionary.relationships. This similarity or non-
similarity is assessed on the basis of many equally—weighted variables.
In its strictest sense a PCA scattergram simply provides a picture of
phenetic dispersion. The extent to which the PCA scattergram can also

be interpreted as a measure of genetic dispersion is still open to

10

debate. Jardine (1971) states, "phenetic classification does have an
evolutionary basis because the observed phenetic dissimilarities of
populations have been established by evolutionary divergence". Sneath
and Sokal (1973) state, "constant evolutionary divergence leads to
overall phenetic divergence, though at different rates in different
lineages. Hence forms that are distantly related phylogenetically will
in general be phenetically distant and vice versa".

Adams and Wiersma (1978) proposed a method of using the phenetic
distance revealed in a PCA scattergram to calculate genetic similarity
between units. The "genetic distance” between units is determined from
the geometric (Euclidean) distance between the units. The more closely
related two genotypes are genetically, the closer they should be on the
scattergram. Adams (1977) applied this method to dry bean (Bhgpgplpg
gplggpip) cultivars of known pedigree. He was able to demonstrate that
the geometric distances between variables varied with genetic
distances. In typical understated fashion Adams noted, "the outcome
suggests, but of course does not 'prove', that 'distances' based upon
PC scores of cultivars have a substantial genetic implication".

Isleib and Wynne (1983) employed this method of estimating genetic
distance to evaluate heterosis in crosses between a high yielding,
adapted peanut breeding line and exotic peanut cultivars of diverse
origin. Due to the uncertain subspecific botanical classification of
some of the exotic parents, PCA was used to cluster the exotic parents
into five distinct morphologically similar groups. The phenetic
grouping largely agreed with accepted botanical classification. Isleib
and Wynne used the geometric distance from the adapted peanut to each

exotic parent as an estimate of genetic divergence between the two

ll ’

parents. For characters exhibiting dominance, heterosis in offspring
increased with morphological divergence between the parents.
Multivariate analysis has often been employed to correlate
phenetic variation with geographic and ecotypic origin. Hussaini et
a1. (1977) used PCA to study 640 genotypes from the world germplasm
collection of finger millet (Elgpgipg pp;apppa(L.) Gaertn.). The
material separated into 12 broad groups basically identifiable by
country of origin. The groups containing Indian genotypes exhibited a
clinal variation, with southern material at one extreme and eastern
material at the opposite extreme of the scattergram. Morishima (1969)
utilized PC and cluster analysis to study variation within prga
pgpgppis Moench. The Asian form of Q. ppgpppig may be the progenitor
of cultivated rice, 9. papiga L. The strains of Q. ppxpppip are
commonly classified into Asian, African, and American forms. Whether
or not any of these forms should be classified as a separate species
has been a matter of controversy. All of the 24 morphological
characters measured varied continuously among strains. None of the
geographical groups could be distinguished on the basis of a single
character. PCA based on the 24 morphological characters separated the
entries into three groups. The African group was distinct and
separate, whereas Oceanian and Asian material clustered together and
was overlapped by the American group. The Asian group displayed a
clinal variation, with annual types at one extreme and perennial types
at the other. In the American group the strains were arranged
according to latitudinal distribution from Cuba at one extreme to the
Amazon at the other. Adding data for interspecific F1 sterility to the

morphological data differentiated the Oceanian material from all other

12

groups. The Asian, African, and American groups were distinct and
separate, with no overlap.

Nevo et a1. (1979) used PCA to study genetic diversity in wild
barley (prggpm pppppppgpm), the recognized progenitor of cultivated
barley. Twenty-eight natural populations representing the entire
ecological range in Israel were sampled to study genetic variation in
these populations and its association with ecological variables. An
electrophoretic survey of 28 gene loci (in 15 enzyme systems) revealed
isozyme variation at 25 (-89%) of the loci. These data, along with
measurements for 4 morphological characters of the barley spikelet,
were used in the PCA. Nevo et al. found that the geographic patterns
of variation displayed for isozymes and spikelet morphology were
correlated with temperature and humidity. However, the variation for
isozyme loci and spikelet morphology were largely uncorrelated. Nevo
et a1. point out that the low level of correlation between isozymes and
spikelet morphology may indicate that the two have developed along
different lines. The wide divergence in land races for spikelet
morphology was not paralleled by similar isozyme diversity.

Martin and Adams (1987) measured 26 characters (including 5
phenological, ll morphological, 5 agronomic, and 5 qualitative traits)
for 375 lines from 15 landraces of dry bean (P.2plgapig) from northern
Malawi. A PC scattergram revealed great variation in quantitative and
qualitative traits. A clinal pattern was observed, with lines from
northern areas clustering at one extreme and those from southern areas
at the other. Whether this variation was due to environmental factors
or to consumer preference could not be determined. Estimated "genetic

distance" indicated greater between-area than within-area variability.

13

Rhodes et a1. (1971) utilized PC and cluster analysis based on 67
quantitative and qualitative characters to study 38 cultivars
representing the three races (Mexican, Guatemalan, West Indian) of
avocado (2gpgga pmggipppg Mill.) and their interracial hybrids. The PC
scattergram revealed that the three races formed separate and distinct
groups with interracial hybrids intermediate between the two parental
racial groups. Rhodes et a1. concluded that PCA revealed a clearer
picture of overall diversity while cluster analysis was more useful in
defining relationships between individual units.

Challice and Westwood (1973) employed PCA to analyze relationships
between 17 231p; species. They generated three analyses: one based
solely on 29 chemical characters (leaf phenolics), another based solely
on 22 botanical characters, and a third using all 51 characters. The
scattergram based upon both the chemical and botanical characters best
agreed with the known geographic distribution of the species. When
either the chemical or botanical characters were used alone, the PC
scattergram contained seriously misplaced species.

Jensen and Hancock (1982) performed multivariate analyses of three
species of California strawberry (Epagagia) from 32 wild populations,
and used PCA to classify the collection sites into six community types
based on ten environmental variables. PC and cluster analysis based on
19 morphological characters suggested that the three Eppggpig taxa were
morphologically distinct. The differentiation between the three taxa
was further verified by discriminant analysis. Discriminant analysis
of all 254 individual plants resulted in 77% correct classification of
individuals as to population of origin. However, 99.6% correct

classification was attained in assigning individuals to the correct

14

taxon. Discriminant analysis was also used to examine correlations
between Exagapia populations and community types. The two Epgggpig
taxa studied revealed morphological variation that corresponded to
population and community of origin.

Numerical taxonomic methods have also been employed to examine
taxa above the species level. Sneath (1976) reviewed these studies,
outlined difficulties and suggested certain stipulations to enhance the
use of numerical taxonomic techniques in classifying higher ranking
' taxa. France et a1. (1969) utilized PC and cluster analysis to examine
genera in the family Chrysobalanaceae. They conducted a preliminary
study of herbarium material of 254 species in the Hirtelleae tribe.
This preliminary study prompted them to question previously accepted
generic concepts. The diversity found in the genus Eppipppi, in
particular, suggested splitting into several more clearly defined taxa.
PC and cluster analysis conducted with data (21 quantitative and
qualitative morphological characters) from 140 species in the
Hirtelleae confirmed the heterogeneous nature of the genus. Further,
most of the tentative groupings indicated by the herbarium study were
confirmed by the taximetric analyses. Indecisive results were obtained
for only two of the 140 species studied. As a result of the study five
new genera were described by Prance.

Jensen (1977) utilized PC and cluster analysis to examine seven
oak species representing three series of oaks. The study included five
species in the Coccineae (scarlet oak) series plus two oak species
whose relationships to the scarlet oaks has been debated for many
years. The analysis, based on 36 quantitative and qualitative

characters, indicated little basis for recognizing the three

15

traditional series in the red and black oak subgenus Erythrobalanus.
The phenetic differences between the species within the scarlet oak
series were often greater than the differences between these species
and the two species representing different series. Jensen pointed out
that while these analyses are purely phenetic, the fact that various
multivariate methods repeat basic patterns indicates that these
patterns do indeed reflect evolutionary relationships.

To determine whether Cannabis is a monotypic genus or if more than
one species should be recognized, Small et al. (1976) analyzed plants
from 232 populations of wild and domesticated Qpppapig for 47
characters using cluster and discriminant analysis. (Discriminant
analysis is used to classify units into one of two or more previously
defined and distinct groups.) Cluster analysis, which weights all
characters equally without assuming any groupings, revealed no evidence
of distinct groupings that might warrant consideration as separate
species. Discriminant analysis was then used to group nonintoxicant
and semi-intoxicant populations together for comparison with the
intoxicant populations. Ninety-four percent correct classification of
populations was achieved based solely on morphological characters.
Further, the two groupings, the nonintoxicants and the semi-
intoxicants, within the less intoxicant category could be distinguished
with 75% accuracy. "Wild" plants could not be consistently
distinguished from domesticated plants; rather variation appeared to
continuous. Small et al. found that intoxicant plants could be
distinguished from less intoxicant plants, when grown under standard
conditions, by numerical analysis of many morphological characters of

plants. However, these characters were much too affected by the

16

environment for consistent identification of plants from various
sources.

PCA has been employed to examine accepted phylogenetic
relationships. It has also been utilized to examine interspecific
hybrids of known and unknown parentage. Rhodes et a1. (1968) employed
several different methods of cluster analysis to examine the genus
Qpppzpipa. The results were compared with accepted phylogenetic
relationships within the genus. This genus displays a high degree of
intraspecific uniformity and a large amount of interspecific
variability. Twenty-one species of gpppppipa were scored for 93
morphological characters. Rhodes et a1. decided that the cluster
analysis derived from product-moment correlation coefficients most
closely agreed with known cross compatibility, geographical
distribution, and ecological adaption for the species studied. Bemus
et a1. (1970), in a follow-up study, augmented the original 21
Qpppppipp species with 29 representatives of F1 hybrids, cultivars,
and unclassified accessions. Data from 93 quantitative and qualitative
characters were used in a cluster analysis. The additional 29 units
only slightly altered the original cluster patterns established with
the Cucurbita species alone. The F1 interspecific hybrids generally
clustered with one of the parent species. However, two interspecific
hybrids whose parents were widely divergent morphologically exhibited
little phenetic similarity with either parent and thus did not cluster
near either parent. F1 hybrids between wild and cultivated species
clustered near the wild parent. The cluster analysis resulted in a
grouping of species that closely agreed with known genetic

compatibility relationships.

17

Jensen and Estbaugh (1976a) studied three populations of red oak
with narrow geographic distributions and low taxonomic diversity
utilizing PC and cluster analysis to determine probable parentage of
putative oak hybrids. Sixty-three quantitative and qualitative
characters were scored. Each test population consisted of all the oaks
found in a small area, or three to five oak species plus their putative
hybrids. PCA was useful in determining probable parentage of naturally
occurring putative hybrids. In another study, Jensen and Estbaugh
(1976b) examined two oak populations with wide geographic distribution
and high taxonomic diversity. One population consisted of ten taxa of
red oak, the other of seven taxa, plus the putative hybrids in each
population. They concluded that in populations with many potential
parental taxa, especially when many are highly divergent, an estimation
of the parentage of hybrids relying solely on multivariate techniques
is difficult.

Heiser et al. (1965) employed cluster analysis to study fiplpppm
species, artificial hybrids, autoploids and alloploids to test the
efficacy of numerical taxonomic techniques with plants of known
relationship. The 75 units studied represented 17 species, 4
autoploids, 6 alloploids, 19 hybrids, and one plant of unknown origin.
Fifty-eight quantitative and qualitative characters were measured. The
cluster analysis disagreed in several cases with orthodox taxonomic
opinion about the position of some species. The interspecific hybrids
tended to cluster close to one of their parent species. The autoploids
clustered close to their diploid parent. The artificial alloploids
were morphologically distinct from both parental species and therefore

did not consistently cluster with either parent.

18

To date, only two PCAs have been done with the genus Pppppg.
Carter et a1. (1983) used multivariate analysis to examine geographic
variation and identify desirable seed sources for the important lumber

species, Prunus serotina (black cherry), after Pitcher (1982) had

 

concluded that phenotypic selection was ineffective. Seed was
collected from 33 stands of black cherry throughout its natural range
in the eastern United States. The resulting trees were evaluated at
ten years of age for phenological and morphological traits. Cluster
analysis based on leaf morphology clearly separated the southern from
the central and northern provenance trees. The seven provenances
identified as southern on the basis of leaf measurements included most
of the provenances that had been placed in the southern grouping on the
basis of leaf and flower phenology (Cech and Carter, 1979).

Hillig (1988) used PC and cluster analysis to examine a sour
cherry germplasm collection. Morphological traits of leaves and fruits
were employed to study 16 sour cherry cultivars plus hybrids and open-
pollinated seedlings of Eastern European varieties. Russian cultivars
were found to be morphologically diverse. The other cultivars plus the
hybrids and open-pollinated seedlings tended to cluster by area of
origin and known genetic relationship. PCA revealed that character
states typical of sweet cherry and ground cherry (the putative
progenitors of sour cherry) tended to fall at opposite ends of the PCs.
Thus, the PCs were interpreted as representing gradations between sweet

and ground cherry-like morphology.

19

MATERIALS AND METHODS
Elfln£_flflfifiliél

Pollen was collected in Spring 1983 from the following locations
in eastern Europe: Fruit Research Institute, Cacak, Yugoslavia; Fruit
Growing Research Institute, Plovdiv, Bulgaria; Research Institute for
Pomology, Pitesti, Romania; Enterprise in Extension in Fruit Growing
and Ornamentals, Budapest, Hungary; and the Research Institute of
Pomology, Skierniewice, Poland. The pollen was brought back to
Michigan and used in crosses with the sour cherry cultivars
'Montmorency', 'English Morello', and 'Wolynska', generating three
half-sib families (Tables 1 and 2).

Additionally, two representatives of the presumed progenitor
species of sour cherry were included in the study; 2. 32133 'Angela'
and 2. fruticpsa (ground cherry) o.p. Self-pollinated 'Montmorency'
seed was collected from a Michigan orchard.

All trees were planted in a completely randomized design at the
Clarksville Horticultural Experiment Station, Clarksville, Michigan.
The trees were three years old when evaluated in 1987.

Chapacperg Measuped

Data for sixteen vegetative characters were collected in August,
1987, after vegetative growth had ceased. Five leaves were collected
at random from the mid-length of the current year's growth from each
tree and evaluated for the following characters: leaf blade length and
width, petiole length and width, (widths measured at mid-lengths), the
number of swollen glands on the petiole and on the basal leaf edge, and
vein angle (measured between the mid-vein and the secondary vein

closest to the leaf blade mid-length). Pubescence on the lower leaf

20

 

H um>wuaao cmwummcsz

wonoo mouu
mo .chm uonoo xwmz
a pawn»: cmwumwcsm

.huuozo

usom use uooam cooaoon

pawn»: Henson: a

mu assesses smaawcm

m.~ sauna: caucuses

.muammmaxmqmum 4M

x Anhman exmaaoumv

humane undouw mu HmouH

.HmooH mo acowoua

.Q.o so ma ux>oaom

a.e.~ sauna; emacmaom

mcausnowz
mhmcpouopoam
w.m.o .cxm

coauou hanmadmoum
m .Hm>wuaso vac cm

Homc< mwmz x mm Anson

umboo xwmz x mpsmm

«use camu>
x meanness smaawem

ox>oaom
x mmmxmuwawomH>

Awhm>oxHHumx magmasofiz
x mameumEMm nauseoumv
humane upon

x humane unsouw

caocxc:

Hmuox uoouoz

mm:

Hm\ws =

mm\aa :

Swansea:
60> ouobuosum

oaamuoz smaswcm

 

HMQUCGHQM mm “USU—5:00

oouwwuom

um>wuaso

 

.moHHHEmw huuozo Hzom o£u mo unsound mo moouwwvom .H canoe

21

A.EEoo .muomv .> .me>omeuH
mama .cwusnowz
mama .cwusnowz

mama .HcouuoH

A.EEoo .muoav .< .wcouuoH

maea .xowuvom

mama .pouoo use sauce

chouuoH .4 on .5500 .muoav .H .Houmom<

HNMx‘i‘lanw

“WQUCQHQMOH HON QUOO HMOHHQESZH

 

q um>wuaso LmHHom csocxa: exm:%H03
.mwumoz ou
unawawm huo> ma Ao>uomcoo
m.::mawom .czmv Hench Hmcmm
c .um>wuH:o cow>mamow3> ma acumen Hmcuoume MJCHUwEdm

ow>mamow3> :H %Humoa
q .oomupcmH cmanumaumo mxmcwomano

a SmHHom managemoua caocxcs mammoz

ocsaaoo omwuoo
uo o>Humx omwuoo mo
Azowoun m zanwmmoa
m Hm>wuaso nonouh uao csocxc: hocouoaucoz

 

HmoQCouomom mucoaaoo oouwwoom Ho>auaso

 

.A.o.u:oov H «Home

22

Table 2. Names, abbreviations, and progenies
evaluated for the clone and each family.

 

No. of trees

 

Name Abbreviation evaluated

2. axipm Angela A 3
P. fruticosa o.p. F 3
Montmorency

x H 18/21 R1 10

x H 17/39 R2 7

x Meteor Korai MK 9

x M 63 H4 10

x Nefris Nf 9

x Fructbare VM 4

von Michurin

x mixture AV 6

x self M 10
Wolynska

x Sumadinka L1 10

x Oblacinska L2 8

English Morello
x Sumadinka E1 10

x Oblacinska E2 10

 

23

surface on the midvein and secondary veins was subjectively rated from
0 to 4, with 0 - maximum of three trichomes per leaf, to 4 - very
pubescent. Leaf area was measured utilizing a leaf area meter. Trunk
diameter was measured_with a caliper 8 centimeters from the soil line,
once in May and again in September, and trunk diameter increase was
calculated. Branch angle was measured for the first 5 lateral branches
from the base of the tree. Current year's terminal growth was measured
on 3 randomly chosen lateral branches. Internode length of the lateral
branches was measured at mid-shoot of the current year's growth, and
the number of leaves per lateral branch was recorded.
Data Analysis

Principal component analysis was performed using the PRINCOMP
procedure of the SAS statistical package (1985a). Family means were
used to create a correlation matrix from which standardized principal
component scores were extracted. Scatter plots of the first 3 PCs were
created with SAS/GRAPH (1985b). To determine which of the PCs
accounted for the greatest amount of variation for each trait, the
eigenvectors of the 3 PCs were compared for each trait. The trait
being considered was ascribed to the PC having the largest value. The
progressive increase or decrease of the clone or family means for each
trait along a given PC was followed to assign a trend to the PC for the

trait being considered.

RESULTS
The first 3 PCs account for, respectively, 62.9%, 12.6%, and 9.4%
of the variance between means or 84.9% of the total variation (Table

3). The contribution of the various traits to the eigenvectors is

24

 

 

Table 3. Eigenvalues, proportions of variance, and cumulative
variances from the principal component analysis of the
cherry clones and families.

Principal Proportion Cumulative

component Eigenvalue of variance variance

1 10.079 0.629 0.629
2 2.018 0.126 0.756
3 1.496 0.094 0.849
4 0.987 0.062 0.911
5 0.489 0.030 0.942
6 0.365 0.022 0.965

 

25

listed in Table 4. Almost two-thirds of the total variance is
accounted for by PC1. From plus to minus on PCl the family means
generally decrease for leaf length, leaf width, leaf area, petiole
length, lateral length, internode length, and trunk diameter increase.
From plus to minus along PC2 number of leaves per lateral, number of
petiole glands tend to decrease while pubescence increases. On PC3
from plus to minus there is a general decrease in branch angle,
serrations per centimeter, number of glands on the leaf blade basal
edge, and trunk diameter increase.

2. pgipm 'Angela' is located at the positive extreme of PCI and
distant from the ground cherry and the sour cherry families (Figure 1).
R1, the 'Montmorency' family with a Romanian cherry [sour cherry x
(sour x sweet cherry)] as the paternal parent does not cluster with the
other sour cherry families and is situated closer to 'Angela' than any
other family. Ground cherry (P. fruticosa) is not separate from the
sour cherry families, sharing the extreme negative position of P01 with
'Wolynska’ x 'Oblacinska' progeny. The 'Montmorency' family (VM) with
the Russian cultivar Fructbare von Michurin (ground cherry x sour
cherry) as the paternal parent is situated outside the sour cherry
cluster at the extreme of PC2. However, its position on P01 is within
the sour cherry cluster.

'Angela' had the longest internode and lateral lengths and the
largest leaf area, the greatest leaf length, leaf width, petiole
length, vein angle, pubescence on both mid and secondary veins, and the
lowest number of serrations per cm (Table 5). It was second to only
one other family (VM) for the greatest number of swollen petiole glands

and it had the lowest number of glands on the leaf blade edge. The

26

Table 4. Eigenvectors of the first 3 principal component axes
from principal component analysis of the cherry clones
and families.2

 

 

 

Eigenvectors
Character PCl PC2 PC3
Petiole length 0.304 -0.003 -0.041
Leaf length 0.309 -0.049 0.059
Leaf width 0.306 0.008 -0.028
Petiole width 0.252 0.210 -0.097
Vein angle 0.261 0.216 -0.186
Serrations per cm -0.234 0.311 0.298
Petiole glands 0.167 0.487 0.166
Blade glands -0.118 0.254 0.276
Pubescence (Mid-rib) 0.209 -0.405 0.327
Pubescence (Side-ribs) 0.240 -0.376 0.235
Branch angle -0.120 -0.053 0.699
Leaf area 0.300 -0.127 -0.029
Trunk diameter increase 0.273 0.064 0.285
Lateral length 0.297 0.189 0.010
Leaves per lateral 0.204 0.374 0.149
Internode length 0.279 -0.060 0.025

 

zPC1, PC2, and PC3 account for 62.9%, 12.6% and 9.4% of variance

between means, respectively.

27

.N manna cw we muoHumH>oubn<

.moxm om

moufiu umufim ozu so mcmoa xawamu humane vcsouw one know .uoozm mo monoum om mo macauumom

v0.0!

 

 

 

 

 

 

 

 

 

ucouma Hmcuoume mu mxmcxaoz

acumen Homecome mm caaouoz zmwawcm
ucouma Hecuooea ma zoceuoaucoz
causewfiaoa-maom accouoaucoz

dd g .M

mHowc< amMNm .M

 

 

.H oufiwwh

v 00.“

 

nun

28

 

 

 

 

 

 

 

 

n.~-o.a o.n a.ma-m as o..~-o.. n.na c.o~-a.na o.»~ o.na-o.n o.o engage-ago u
ode; o.~ one 3 981: «.ma 92-92 98 9%-"; 9.: assign u

oaaouo: sausage
n.~-o.a a.” an-m on a.on-a.n n.5H n.a~-o.a o.na ~.oa-o m.n axacaoaaao u
a.n-n.a o.~ «n-m as m.on-o.a ..n~ n.~n-m.oa n.o~ n.eH-s.n n.aa uaaaeaasm u

oanSOZ
a.n-n.a o.~ s~-o a” H.nn-o.n n.a~ a.nn-m.m o.a~ a.~a-n.a a.“ man. u
o.~-o.o m.a --m ca n.o.-u.n n.ma o.nn-n.o n.- n.oa-~.a m.m cases“: a
n.~-o.a n.~ n~-ma an n.oc-n.cu n.nn a.o~-o.na «.ma n.sa-a.o ..Ha manages: ca> .uansoaua x
n.~-n.a o.” oq-oa cm n.no-o.a n.on ~.mn-n.a a.- ~.--a.o n.o .«uuoz u
c.~-n.a a.“ an-n nu o.o.-o.n ~.an a.an-o.m a.n~ ~.o~-a.a c.aa mn\aa m u
n.~-c.a o.~ ~n-n an o.~m-o.. a.o~ H.H¢-m..a e.n~ m.an-n.c ”.44 no: u
o.~-o.a n.~ o~-oa ma a.on-~.aa m.a~ n..n-m.a ~.n~ n.c~-m.n o.oa «one: noougx u
n.~-c.~ n.~ os-a ma m.mo-a.a o.on a.om-o.~a ~.os m.mo-a.a o.cn .a~\oa a u

Accouoaunot

n.~-o.~ n.~ o~-~H «a o..~-~.ma o.ma ..on-a.ma n.n~ a.aa-o.m ~.o .a.o cmooassuu .m

o.n-n.n o.n nn-ma nu H.oo-o.oc o.an o.aa-a.mm ~.am a.n~-~.aa ..o~ .aaonnc. mmanu .m
as: x Hon—:53 u as... x Koo-:58 n 58:53 N hdelhonoao

A83 .133!— Aaov “No.3 2!: 030.35
53 ocean—3am non coerced nausea deuce..— eouo ueoq nod-'3 unis

 

Jeannie ~25 venom—2‘ e503 haemoua on» Mo and u25d> canon 23.

.nueuoeueso 103393.33 uo eons-u va- oouol him and 23.8 .n 93:.

29

 

 

 

 

 

 

 

0.0::«40 «.00 0.:nu.0 0.0 500.0 0.0 0.70.0 HA 0323038 N
.n.00u0.00 0.: «.00.. «.0 c.0nc.0 0.0 0.70.0 04 exudes—50 N
31.32 £033
«070.3 «.00 0.cu0.« 0.0 0.70.. 0.0 «170.0 0.0 0335038 n
0.00u«.n0 0.0: 0.0-0.0 0.0 5.00.0 5.5 «170.0 «A 0:50:50 N
338303
0.00.0.0« 5.00 0.0:0.« 0.. 0.30.0 0.x o.«n0.0 «A «a: u
0.00:0.0« 0.0.. 0.70.« 0... 0.3.0.0 0.0 «.«amd 04 unsani— u
«.00nu.00 0.: 0.7T: 0.0 0.0un.~ 0K 5.70.0 04 cans—~00: no: ouefiuaum u
1000.0: 0.0: 0.0-0.0 0.: 0.0-0.0 0.5 0.«u0.u 04 0.30.: N
n.00ns.0« 0.0: c.0uc.« 0.0 H.0u~.n 0.5 a.«ns.0 0A 022 a H
0.00-0.00 0.00 0.0-0.0 0.0 0.3-0.0 0K «.«u0.u 0.0 0% u
0.«0..n.00 0.«0 u.0n0.« 0... n.0u«.0 0.~ 0.7.0.0 04 «9.39 .3000! u
0.:-0.0« 0.«0 «0:50 «.0 n.«au0.c «.3 ~.«..«.a 0.« ««\0n n u
Usage:
0.000.; “.00 «.cncd «.0 0.0..«.0 0.0 0.70.0 «.u do 300330 .M
5.00:0.00 0.50 0.70.0 «.n 0.0T«.0H 0.: «.0u0.« 0.« .1392. ”Trad
and! u Hon—.3! u Marni N Hausa H ham—ImEEHU
Aloe .83 .3. :80
50!- 040300 533 need finned uooq 508A 0.3.3.0

 

$430303 0. can-H

30

 

 

 

 

 

 

 

 

0.«:«4 5.0 «0.0.0 0.0 «.«an0.0 0.0 0.«0n5.00 0.00 000.2038 M
0.0-«4 04 0.70.0 0.0 0.0.70.0 «.3 0.«01040 0.00 0450330 I
0.30.3: 530:0
0.«:«0 0.0 0.0-0.0 0.0 0.«.7«.H 0.0 0.00:0.«0 0.00 0322038 u
«.704 0.0 0.70.0 0.0 0.200 0.0 0.«0040 5.00 0:50.330 N
0308303
0.«..«.0 «4 0.00.0 0.0 0470.0 0.0 0.00-0.00 0.50 uaoe u
0.70.0 0.0 0.00.0 «.0 0.0700 «.0 0.50..«.00 «.00 9.9.038 u
0.70.0 0.« 0.70.0 0.0 0.0.70.3 «.3 0.004000 0.00 5553: 60> euofiosum N
0.«n«.0 0.« 0.70.0 0.0 0.«70.0 0.0." 0.00u«.«0 0.50 eauuez u
0.70.0 0.0 0.«:0.0 0.0 0.«au0.5 0.0 0.«0..0.00 0.50 00\5H m u
0.70.0 0.« 0.70.0 0.0 «.0700 5.0 0.«0040 0.00 09. u
0.0..«0 0.0 «270.0 0.0 0.3-0.0 0.0 n.50..«.00 0.00 «enou— uoouot 8
0.70.0 0.." 0.0.0.0 0.0 0.0-0.0 0.5 5.00:0.00 0.00 a«\0~ m x
fixings—~02
0.«-04 0.4 0.00.0 «.0 0.07«.0 «.3 0.00-0.00 0.00 .06 333030 .M
0.00.0 5.0 «.7«.0 5.0 0.00.0 0.0 000000 «.00 .3092. 852.6%
Hon—:5! u Huang... H N03190- m H0815... u h.— «30.00800
neg-83:00 5003
83.00 00.00 ova-.8 000300 ~00 0830.300 0.35 53>

 

.56.»:000 0 ounce

31

 

 

 

 

 

 

 

0.05:0.«0 5.00 0.«:0.0 0.« 0.«:0.n 0.« 1016000000 N
5.05-0.50 0.00 0.«:0.0 a.“ 0.«:0.0 0.« Incuuaasm n
0.333. 8.3030
0.00:0.00 0.00 «.«u0.0 0.0 0.«:0.0 0.« caucuonano M
«050.50 0.00 «.700 0.0 0.«..0.0 0.« annual—50 u
agenda:
0.00-0.00 0.00 «.nn0.0 0.0 «.«n0.0 0.0 «do. u
0.00u«.00 «.«0 «.nu0.0 0.0 0.«:0.0 5.0 ouznuua u
«.00u«.00 «.50 0.0-0.0 «.0 0.0-0.0 0.0 nausnonz 99> oudnuoaum u
0.05:0.00 0.00 0.«:0.0 5.0 0.au«.0 0.0 nuuuoz u
0.05-0.00 0.00 0.0-0.0 0.0 0.«:0.0 0.0 00\5H 0 u
0.00:0.00 0.00 0.«:0.0 0.0 0.«:0.0 «.0 008 u
«.00n0.00 «.«0 0.«:0.0 0.« 0.«:0.0 0.« aqua“ noouoz u
0.05:0.00 0.50 0.«:0.H 0.« 0.«:0.H 0.« H«\0~ 0 N
honouoaanox
0.«0-«.00 0.50 0.0:0.0 «.0 0.0-0.0 0.0 .n.o oncoausuu .M
«.50-«.00 0.00 0.0-0.« 0.« 0.«:0.« 0.« .uao0d<. a:«>~.M
unarsua x xdalcua N Musings u hauauh\oaoau
A0000 ncwp> n0uouuu nao> nuuvafl
000:: non-um no con-ononnm so concouonnm

 

.A.v.acbov 0 candh

32

family 'Montmorency' x H18/21 (R1) had the largest trunk diameter
increase and number of leaves per lateral, and was second only to
'Angela' for leaf area, lateral length (Table 4), leaf length and
width, and the least number of serrations per cm.

Ground cherry occupies the most extreme negative position on PCl
(Figure 1), sharing this position with one other family, 'Wolynska' x
'Oblacinska' (L2). Ground cherry is not the most extreme for most of
the characters measured. Two sour cherry families have fewer leaves
per lateral, 3 have a smaller trunk diameter increase and lateral
length, 1 family has a shorter leaf length and width, and 3 have a much
smaller leaf area. VM, the sour cherry family with a ground cherry
grandparent, is positioned outside the general sour cherry cluster, at
the extreme positive end of PC2. VM (with a ground cherry grandparent)
and R2 (with a ground cherry great-grandparent) are located very close
together on PC1.

The 'Wolynska' and 'English Morello' half-sib families both were
crossed to the same paternal parents, 'Sumadinka' and 'Oblacinska'.
Virtually all characters measured were smaller for the individuals in
half-sib families with ’Oblacinska' as the paternal parent than for
those in half-sib families with 'Sumadinka' as the paternal parent. In
fact, the 'Wolynska' x 'Oblacinska' family represents the extreme low
value among all the families studied for trunk diameter increase, leaf
area, leaf length, leaf width, and petiole length. The 'English
Morello' x 'Oblacinska' family represents the extreme low value among
all families for lateral length, number of leaves per lateral, and leaf
vein angle. 'Oblacinska' is a small statured selection of a

Yugoslavian landrace.

33

Two 'Montmorency' half-sib families are worthy of comparison --
'Montmorency' self-pollinated and 'Montmorency' cross-pollinated with a
mixture of cherry pollen from Romania (AV). For many of the characters
in Table 5, the means are similar; however, the 'Montmorency' selfed
progeny exhibited a smaller range of values and is often missing the
high or 'vigorous' end of the maximum values shown by the ’Montmorency'

outcrossed family.

DISCUSSION

Previous PCA of a diverse sour cherry seedling collection (Hillig,
1988) suggested that the values of the individual PCs might represent
gradations between character states typifying the putative progenitors
of sour cherry -- ground cherry and sweet cherry. This study, with the
inclusion of a sweet cherry cultivar and open-pollinated ground cherry,
is an attempt to more clearly define the wide variation exhibited by
sour cherry varieties. Additionally, this study includes measurements
of charaCters affecting growth habit and size which vary greatly within
the sour cherry collection.

Many morphological characters of sour cherry varieties are
intermediate between ground cherry and sweet cherry. The most obvious
difference between ground and sweet cherry is tree size. Sweet cherry
forms a large, upright-pyramidal tree with a height reaching to 18
meters, with large, thin, coarsely serrate obovate-elliptic leaves
(Bailey et al., 1976). The leaf is glabrous above and persistently
pubescent below, with a long petiole and prominent glands. Fruit of
cultivated varieties is large and sweet. Sweet cherry is the least

winter-hardy of the three species.

34

Ground cherry forms a spreading, multi-stemmed shrub about one
meter tall, with small, thick, finely serrate, elliptic leaves (Bailey
et al., 1976). The leaf is glabrous above and below, and has a short
petiole with few glands. The fruit is small to medium sized and
acidic. It is the most winter hardy and drought resistant of the three
species (Kolesnikova, 1975).

Sour cherry forms a shrub or small tree with medium-sized doubly
serrate ovate to obovate leaves. The fruit is usually medium sized and
tartly sweet (Bailey et al., 1976).

The sweet cherry 'Angela' was distinctly separate from the sour
cherry families, occupying the extreme positive position on PCI,
largely due to the high contributions of the following characters on
PC1: leaf length and width, leaf area, petiole length, lateral length,
internode length, and trunk diameter increase.

Although ground cherry occupies the extreme negative position on
PCl, it shares this position with one sour cherry family ('Wolynska' x
'Oblacinska'), and is not distinct and separate from the other sour
cherry families. In the Eastern European and Russian literature there
is no clear delineation between sour cherry and ground cherry.
Cultivars which are hybrids between sweet and sour cherry or sour and
ground cherry, and cultivars which are large fruited selections of
ground cherry are all considered to be "sour cherries". From the
characters measured in this study a clear distinction cannot be made
between ground cherry and sour cherry. Continued gene flow between
these two tetraploid species might account for the overlap between the
two. Kolesnikova (1975) has reported the occurrence of natural hybrids

between sour and ground cherry in the middle Volga region of Russia.

3S

C.D.Darlington (1928) stated, "if we endeavour, following the older
systematists, to distinguish among the cherries three species, 2.
ayium, 2. ggzaggg, and g. frutigggg, we find that the last two merge
into one another by imperceptible gradations”.

A basic limitation in this study is that only one sweet cherry
cultivar and open-pollinated progeny from three clones of ground cherry
were used in the experiment. Obviously these few selections cannot
exemplify the range of variation within either sweet or ground cherry.
If a more diverse sweet cherry population were used some individuals
might resemble sour cherry more closely than does 'Angela'. Similarly,
the ground cherries used in this study may represent cultivated ground
cherry clones rather than wild ground cherry types. Kolesnikova (1975)
describes several different groupings of both sweet cherry and ground
cherry based on morphological characteristics.

Although this study cannot begin to address the question of the
variability to be found within either sweet cherry or ground cherry, it
does indicate the potential within the sour cherry germplasm collection
at M80.

The sour cherry family (R1) with a sweet cherry great-grandparent
is separate from the other sour cherry families and toward the "sweet
cherry side" of the PC scattergram. R1 is second only to the sweet
cherry 'Angela' for greatest leaf area, leaf length, leaf width,
lateral length, petiole length, and vein angle. It slightly exceeds
'Angela' for the greatest number of leaves per lateral. This family's
exceptionally high trunk diameter increase suggests that sour cherry
clones may be selected that approach sweet cherry in size. Several of

the sour cherry varieties in Hungary (e.g. Pandy, 'Ujfehertoi Furtos',

36

and 'Erdi Nagygyumolcsu') approximate sweet cherry in tree size and are
planted at distances of 8 x 5 meters (Kollar, 1987).

In contrast, numerous "low vigor" sour cherry cultivars exist,
such as 'Oblacinska’ from Yugoslavia and 'Ilva' from Romania. They are
planted at distances of A x 2 m and 4 x 2.5 m, respectively (Iezzoni,
1984). Also, the dwarf sour cherry cultivar 'Kirsa' from Sweden
attains a mature height of only one meter and is spaced one meter
within the row (Iezzoni, pers. comm.). The two sour cherry families
with 'Oblacinska' as a paternal parent indicate that sour cherry clones
may be selected that approach the diminutive size of ground cherry. In
Sweden, meter-high dwarf sour cherry clones are presently being tested
in high density plantings. These dwarf sour cherries could be
harvested with an over-the-row harvester. Pruning and spraying
procedures would also be simplified. Such dwarf clones might also be
used in the United States. The dwarf clones could also prove to be
valuable as dwarfing rootstocks for sour cherry. Pick-your-own
operations and homeowners would undoubtedly welcome bush-type sour
cherries. The homeowner could grow a hedge of sour cherries which, in
addition to providing ornamental value (cherry blossoms in spring, red
fruit in summer, and glossy green leaves all season) and edible fruit,
could be much more easily protected from bird predation than the
standard size sour cherry tree.

Because sour cherry is a naturally outcrossing species, inbreeding
would be expected to reduce seedling vigor. The similarity between the
progeny means from 'Montmorency' self-pollinated and 'Montmorency'
outcrossed suggests that inbreeding depression does not have a major

effect in the SI generation for the vigor characters measured in our

37

study. However, individuals in the 'Montmorency' outcrossed progeny,

were more vigorous than those in the 'Montmorency' self-pollinated

progeny.

CHAPTER TWO

Isozyme Analysis of Sweet, Sour and Ground Cherry

38

INTRODUCTION

Isozymes have been widely employed to detect variability in plant
populations. The use of isozyme analysis enables the plant breeder to
screen plants at the seedling stage, thus conserving limited resources,
including time, money, land, and labor. A previous isozyme study of
cherry (Hancock and Iezzoni, 1988) revealed that MDH exhibited enough
polymorphisms to be used to identify interspecific cherry hybrids.
Codominant expression of MDH banding patterns of sweet cherry and
ground cherry in sour cherry provided additional evidence that sweet
and ground cherry (or, more accurately, the ancestral forms of these
two species) were the progenitors of sour cherry. The present study
was initiated to determine if other enzyme systems could be utilized
for the study of diversity and inheritance within and between the three

species, sweet cherry, ground cherry, and-sour cherry.

LITERATURE REVIEW

Applications of Isozyme Analysis

Isozymes, or isoenzymes, are the multiple molecular forms of
an enzyme occurring within a single organism and having similar or
identical catalytic activities (Scandlios, 1969). Isozyme
heterogeneity was first reported for esterase and lactate dehydrogenase
in 1957 (Hunter and Markert).

Starch gel electrophoresis is proven an efficient method to

separate enzymes into their different molecular forms. Electrophoresis

39

40

is fundamentally forced diffusion through an electrical field.

Proteins such as isozymes migrate at rates proportional to their net
charges and molecular weights, resulting in separation into bands which
can be revealed through staining.

Isozymes have proven to be extremely valuable biochemical markers.
The genetic control of isozyme polymorphism is chiefly monogenic
(Peirce and Brewbaker, 1973) and isozymes are often codominant in
inheritance (Moore and Collins, 1983).

Isozymes are being used in all areas of plant biology,including
identification of cultivars (Weeden and Lamb, 1985; Byrne and
Littleton, 1988; Hauagge et al., 1987; 1988; Bassiri, 1976; Blogg and
Imrie, 1982; Salinas et al., 1982; Weeden, 1984), species (Arulsekar et
al., 1985; Hancock and Iezzoni, 1988; Moore and Litz, 1984; Arulsekar
et al., 1986b), and interspecific hybrids (Byrne and Littleton, 1989;
Chaparro et al., 1989; Parfitt et al., 1985; Chaparro et al., 1987).
Isozymes have been used to estimate the degree of outbreeding or
inbreeding in populations (Allard et al., 1971; Brown and Allard,
1970), as well as genetic, taxonomic, and evolutionary relationships
between and within plant populations (Gottlieb, 1981; Hamrick, 1983;
Ladizinsky and Hymowitz, 1979; Crawford, 1983; Hancock and Bringhurst,
1981).

In Prunus several isozyme analyses have been conducted. Arulsekar
et al. (1986a) analyzed 290 peach cultivars for ten enzyme systems,
including aspartate amino transferase (AAT), acid phosphatase (AcP),
esterase (Est), malate dehydrogenase (MDH), phosphoglucomutase (PGM),
leucine aminopeptidase (LAP), alcohol dehydrogenase (ADH), shikimate

dehydrogenase (SkDH), and 6-phosphog1uconate dehydrogenase (6-PGD).

41

Malate dehydrogenase (MDH) was the only enzyme that showed
polymorphisms.

MDH displayed three distinct banding patterns in the 290 cultivars
analyzed. The F2 seedlings segregated for the three banding patterns
in a 1:2:1 ratio. This is consistent with a codominant, single locus,
Mendelian segregation. Segregation was not observed for four of the
bands. Therefore, the exact number of loci involved could not be
determined. Haploid plants derived from the heterozygous parents
exhibited only homodimeric banding patterns, suggesting that
heterodimers are formed between the products of the two alleles. MDH
functions as a dimeric enzyme in many plant species (Yang et al., 1977;
Sari Gorla et al., 1986; Orton, 1983; Morgan and Bell, 1983; Torres and
Bergh, 1980). The formation of two interaction bands in the
heterozygote in peach indicates that MDH has a polymeric subunit
structure in peach. Arulsekar et al. (1986a) concluded that MDH is a
useful biochemical marker in peach because the multiple banded types
show simple Mendelian inheritance.

In contrast to peach, 87 almond cultivars exhibited polymorphisms
in the enzyme systems AAT, LAP, PGM, MDH, 6-PGD, and PGI (Arulsekar et
al., 1986b). Although only 1/3 as many almond cultivars were analyzed,
they exhibited 11 times the variability found within the 290 peach
cultivars. Because the peach cultivars chosen for the study exhibited
much morphological variability, Arulsekar et al. (1986b) concluded that
morphological variability does not necessarily ensure isozyme
variability. Arulsekar et al. (1986b) point out that while peach is

predominantly self-fertile, almond is a self-incompatible, outcrossing

42

species. This could account for the vast difference in isozyme
variability between the two species.

Durham et a1. (1987) examined 38 enzyme systems in 59 peach
cultivars. Twelve of the enzyme systems produced well-resolved banding
patterns. Of these twelve, nine were monomorphic and only three
[diaphorase (DIA), malate dehydrogenase (MDH), and peroxidase (PX)]
were polymorphic.

Diaphorase (DIA) displayed three banding patterns in peach, two
single-banded and one double-banded. F2 progeny of the cultivars with
double-banded patterns segregated in a 1:2:1 ratio (singlezdouble:
single types) indicating that only one locus is involved and that the
two bands are allelic. No intermediate bands were found in
heterozygous plants so diaphorase appears to function as a monomer in
peach. Diaphorase functions as a monomer in soybean (Kiang and Gorman,
1983) and Camellia (Wendel and Parks, 1982).

Durham et a1. (1987) found three banding patterns for MDH in
peach. Although the patterns were similar to those found by Arulsekar
et al. (1986) they were not identical. Durham et a1. (1987) noted that
this discrepancy could easily reflect the use of different buffer
systems in the two studies. Durham and Arulsekar agreed as to the
genetic interpretation of MDH in peach. In addition to a four-banded
pattern like that found by Arulsekar et a1. (1986), Durham also found a
widely spaced, triple-banded type which Arulsekar et a1. (1986) did not
find. Upon selfing, both these types gave rise to progeny exhibiting
only the parental banding pattern. Durham concluded that these types
were homozygous at this MDH locus. The third banding pattern was a

seven-banded pattern, including all the bands represented in the two

43

homozygotes plus two intermediate bands. When Durham crossed the two
different homozygotes, all the resultant progeny exhibited the seven-
banded pattern, as expected when using parents homozygous for different
alleles at a single locus. When Durham backcrossed the heterozygous
type to one of the homozygous types, he recovered progeny in the
expected 1:1 ratio of parental patterns. Durham et a1. (1987)
concluded that the variability of MDH in peach is explained by two
alleles interacting at a single locus.

Peroxidase (PX) exhibited three well-resolved banding patterns in
the cathodal region of the gel for peach (Durham et al., 1987). These
patterns included two single-banded patterns and one double-banded
pattern which was a combination of the two single-banded patterns.
Peroxidase functions as a monomeric enzyme in other plant species
(Wijsman, 1983; Torres, 1983). In peach, selfed single-banded types
produced progeny with the parental banding pattern. When the two
single-banded types were crossed, all progeny exhibited the double-
banded pattern. Selfed double-banded types produced the expected 1:2:1
ratio (single:double:single types) in the progeny. Durham et a1.
(1987) concluded that the variability for peroxidase in peach could be
explained by the presence of a single locus with two alleles.

Chaparro et a1. (1987) used isozyme analysis to identify peach x
almond hybrids. They tested nine different enzyme systems, including
isocitrate dehydrogenase (IDH), esterase (Est), phosphohexose isomerase
(PHI), superoxide dismutase (SOD), acid phosphatase (AcP), PGM, MDH,
LAP, and 6-PGD. SOD, LAP, and AcP were difficult to resolve and were
not evaluated. IDH, MDH and PHI were also not included in the study

because peach and almond were monomorphic for these three enzymes. PGM

44

and 6-PGD proved to be useful in distinguishing hybrids. These enzymes
produced well-resolved banding patterns, and the patterns differed
between peach and almond.

Chaparro et a1. (1987) found two zones of activity for PGM, as
described by Parfitt et a1. (1985). (PGM-l indicates the more anodal
zone, while PGM-2 indicates the more cathodal zone.) Peach and almond
exhibited the same double-banded pattern for PGM-l. For PGM-2, almond
exhibited a double-banded pattern, whereas peach exhibited a single
band at the same position as the faster band in almond. Peach x almond
progeny exhibited both parental banding patterns. Segregating F2
almond x peach progeny displayed the expected 1:2:1 ratio
(singlezdouble;sing1e). PGM is a monomer in many other plants (Cheliak
and Pitel, 1985; Gottlieb, 1981). Based on this study PGM also
functions as a monomer in almond and peach.

6-PGD functions as a dimeric enzyme in corn (Goodman and Stuber,
1983), barley (Brown, 1983), and soybean (Kiang and German, 1983).
Peach and almond both displayed two zones°of activity for 6-PGD, but
the bands had differing mobilities in both regions. Both species
exhibited single-banded patterns at each locus. All peach x almond
progeny exhibited both parental bands and an intermediate band in both
the 6-PGD-1 and 6-PGD-2 regions. This finding is consistent with the
hypothesis that 6-PGD is a dimeric enzyme, displaying intermediate
bands that are heterodimers. F2 peach x almond progeny produced the
expected segregation at both the 6-PGD-1 and the 6-PGD-2 loci.

Chaparro et a1. (1987) concluded that PGM could be used to distinguish
only 50% of the peach x almond hybrids, whereas 6-PGD could distinguish

all hybrids since peach and almond share no common alleles for 6-PGD.

45

Hauagge et al. (1987a) studied the inheritance of five polymorphic
enzyme systems in almond x almond and in F1 and F2 peach x almond
hybrids. For the enzyme systems AcP, glyceraldehyde-phosphate
dehydrogenase (CAP), and 6-PGD Hauagge et al. (1987a) found two zones
of activity in almond but all loci were monomorphic. The enzyme
systems AAT, PGI, LAP, and PGM each exhibited two zones of activity.
Polymorphisms occurred at five of the eight loci.

In almond AAT-2 was monomorphic, exhibiting a single-banded
pattern. AAT-1 was polymorphic, exhibiting three banding patterns, two
single-banded patterns with either a "fast" or a ”slow" band and a
triple-banded pattern with bands at the "fast" and "slow" migration
positions plus an intermediate band. The peach cultivar used
exhibited the same AAT-2 band found in almond, and the same AAT-1 band
found in the almond "slow" banded homozygote. When triple-banded
almond cultivars were crossed with single-banded cultivars the progeny
exhibited the parental banding patterns in a 1:1 ratio. Crossing
parents with the triple-banded pattern resulted in the expected
Mendelian segregation of 1:2:1. Almond, peach, and segregating hybrids
within almond and between almond and peach all exhibited two zones of
activity for AAT in identical positions. F1 and F2 segregating
progenies indicated that the AAT-1 and AAT-2 loci were homologous
between almond and peach. Hauagge et al. (1987a) concluded that, in
almond and peach, AAT-1 functions as a dimeric enzyme composed of two
subunits with two alleles.

PGI also exhibited two zones of activity in almond (Hauagge et
al., 1987a). PGI-1 was monomorphic with a single-banded pattern,

whereas PGI-2 was polymorphic, exhibiting two single-banded patterns

46

with either a "fast" or a "slow" band and a triple-banded pattern with
bands at the ”fast" and "slow" migration locations plus an intermediate
band. Just as with AAT-1, when a homozygous single-banded parent was
crossed with a heterozygous triple-banded parent the progeny closely
fit the expected 1:1 ratio for parental banding patterns. Similarly,
crossed heterozygotes produced the expected 1:2:1 segregation ratio.
The peach cultivar used in the study exhibited the same banding pattern
at PGI-1 as that found in almond. At PGI-2 the peach cultivar
exhibited the "fast" homozygous pattern found in almond. The F1 and F2
progenies of almond x peach hybrids indicate that peach and almond have
homologous PGI loci.

LAP exhibited two zones of activity in almond (Hauagge et al.,
1987a). LAP-2 was monomorphic with a single-banded pattern and LAP-1
was polymorphic with two single-banded patterns and a double-banded
pattern (with bands at the same migration locations as those in the
single-banded patterns). Many of the almond crosses produced the
expected segregation ratios for a monomeric enzyme with two alleles.
However, several of the crosses involving either of two particular
almond cultivars deviated from expected ratios. Hauagge et al.
(1987a), proposed that in these two cultivars a null allele existed at
the LAP-1 locus. Null alleles for LAP have been reported in several
other plant species (Neale and Adams, 1981; Quiros and Morgan, 1981;
Wilson, 1976; Guries and Ledig, 1978; Adams and Joly, 1980). Peach
also exhibited two zones of activity for LAP, with a single band at
LAP-l and LAP-2, both at migration locations different from those found
in almond. Peach x almond hybrids combined the isozyme patterns of

both parental species, producing quadruple-banded patterns. The F2

47

peach x almond progenies did not exhibit segregation ratios for the LAP
loci consistent with independent assortment. The segregation ratio
exhibited closely matched a 1:2:1 segregation of the parental peach,
hybrid, and parental almond genotypes, with only a low percentage of
recombinants. This suggests that peach and almond both have linkage
between LAP-l and LAP-2. Hauagge et al. (1987a) concluded that since
almond and peach exhibit different migration distances for bands at
LAP-l and LAP-2, LAP provides a unique biochemical marker for tracing
gene introgression between the two species.

In almond PGM exhibited two zones of activity (Hauagge et al.,
1987a). Both loci were polymorphic, exhibiting single- and double-
banded patterns. Crosses between single- and double-banded parents
produced two classes of progeny with the parental banding patterns in a
1:1 ratio. Crosses between double-banded parents produced the expected
1:2:1 segregation ratio. Hauagge concluded that PGM behaves as a
monomeric enzyme in almond.

In a further study, Hauagge et al. (1987b) used the five
polymorphic loci, AAT-1, PGI-2, LAP-l, PGM-1 and PGM-2, to characterize
and group 76 almond cultivars and accessions. These loci permitted
exclusion of certain cultivars as parents. Hauagge found that the
isozyme analysis, combined with historical analysis, phenotypic
comparison, genetic studies, and pollen incompatibility studies,
provided additional evidence that the almond gene pool in California is
dominated by the descendants of the two major almond cultivars grown in
California. Additionally, isozyme differences showed that two
cultivars could not be bud mutations of one of the cultivars as

assumed.

48

Byrne and Littleton (1988) tested eight enzyme systems for
polymorphisms in 29 cultivars of Japanese plum. The enzymes tested
were, glutamate dehydrogenase (GDH), PGI, LAP, MDH, PGM, peroxidase
(PX), 6-PGD, and triosephosphate isomerase (TPI). The first six
enzymes are polymorphic in plum. 6-PGD and TPI were not useful since
both systems exhibited monomorphic banding patterns. GDH, although
variable, could not be resolved consistently, and was not used in this
study.

PGI exhibited two zones of activity in plum. PGI-l, although
variable, was not well resolved. PGI-2 was polymorphic and behaved as
a dimeric enzyme. Four PGI-2 banding patterns were observed in the
plums studied, two single-banded patterns and two triple-banded
patterns. Byrne and Littleton (1988) proposed three alleles at PGI-2,
with the intermediate bands in the triple-banded genotypes representing
interaction products. A

LAP exhibited two zones of activity, while LAP-2 displayed a
single invariate band. LAP-l was polymorphic and behaved like a
monomeric enzyme as in other plant species. Four banding patterns
were observed in the plum cultivars studied, two single-banded patterns
and two double-banded patterns. Byrne and Littleton (1988) proposed
the occurrence of three alleles at LAP-1 in plum.

MDH in plum exhibited two zones of activity, as reported in other
Prunus species. MDH-2, found to be mitochondrial in origin in peach
(Arulsekar et al., 1986) and cherry (Hancock and Iezzoni, 1988), was
monomorphic in plum. MDH-1, having cytosolic activity (Arulsekar et
al., 1986), was polymorphic in plum. MDH-1 exhibited three banding

patterns, a single-banded pattern and two triple-banded patterns. As

49

in other crops (Cheliak and Pitel, 1985), MDH functions as a dimeric
enzyme in plum.

PGM exhibited two zones of activity, both polymorphic. PGM-1
displayed four banding patterns, three double-banded and one triple-
banded. Byrne and Littleton (1988) proposed that the double-banded
patterns represent homozygous individuals, whereas the triple-banded
patterns represent heterozygous individuals. PGM exhibits double-
banded homozygotes in other species (Arus and Shields, 1983; Goodman et
al., 1980; Neal and Adams, 1981; Wendel and Parks, 1982). PGM-2
exhibited two banding patterns in plum, a single-banded pattern and a
double-banded pattern. Byrne and Littleton (1988) concluded that in
plum, as in other plant species (Cheliak and Pitel, 1985; Gottlieb,
1981), PGM functions as a monomeric enzyme.

Peroxidase (PX) produced anodal and cathodal banding patterns in
plum. Only the cathodal bands were consistently resolved. The
cathodal bands exhibited two patterns, a single-banded pattern and a
double-banded pattern. PX is monomeric in other plant species (Torres,
1983; Wijsman, 1983). Byrne and Littleton (1988) proposed that in
plum, as in peach (Durham, 1986) PX functions as a monomer with the two
bands representing alleles at a single locus, and that the PX bands are
consistent through the season. Therefore, PX was useful for plum
cultivar identification.

Using the five polymorphic enzyme systems, Byrne and Littleton
(1988) were able to resolve the original 29 plum cultivars into 19
groups, eleven of which conSisted of a single cultivar. The other
cultivars were divided into eight groups of two or three cultivars

each. The cultivars within these groups were then readily

50

characterized by vegetative characteristics. Byrne and Littleton
(1988) also used the presumptive genotypes generated by the isozyme
analysis to examine nine of the 29 plum cultivars and their reported
parents. For three of the cultivars neither parent was possible and
for two more the reported male parent was not possible. Four of the
nine cultivars exhibited genotypes consistent with that of their
reported parents. For the 29 plum cultivars examined, there was
approximately 50% inconsistency between the isozyme genotype and the
presumed parentage.

In another study Byrne and Littleton (1989) used five enzyme
systems: PGI, LAP, MDH, PGM, 6-PGD, and PX, to identify plum x apricot
hybrids. They analyzed 30 plum cultivars, 4O apricot cultivars and six
plum x apricot (plumcot) hybrids. PGI-2, a polymorphic dimeric enzyme
in plum (Byrne and Littleton, 1988), was monomorphic in the apricots
examined and exhibited a single-banded pattern at the same migration
distance as one of the bands found in some of the plum cultivars.
Therefore, Byrne and Littleton concluded that PGI was only useful in
uniquely identifying some plum x apricot hybrids, depending upon the
genotype of the plum parent.

LAP-1, a polymorphic monomeric enzyme in plum (Byrne and
Littleton, 1988), was monomorphic in apricot. In apricot, LAP-1
exhibited a single-banded pattern at the same migration distance as one
of the bands found in the plum cultivars. As in the case of PGI, LAP-1
can only identify some plumcots, depending on the genotype of the plum
parent.

MDH-1 was polymorphic in plum, while MDH-2 was monomorphic (Byrne

and Littleton, 1988). In apricot, both loci were polymorphic.

51

Although plum and apricot exhibited alleles in common, apricot also
possessed an allele at each locus not found in plum. If one or both of
the unique alleles were present in the apricot parent, then MDH could

be used to identify plum x apricot hybrids.

 

PGM, a monomeric enzyme in Prunus (Chaparro et al., 1987; Hauagge
et al., 1987), was polymorphic at both PGM-l and PGM-2 in plum (Byrne
and Littleton, 1988). Apricot was also found to be polymorphic at both
PGM loci. Although both species exhibit alleles in common at PGM-1 and
PGM-2, plum and apricot also have species-specific alleles at each
locus. Hybrid plumcots could be identified using PGM if either the
plum or the apricot parent used exhibited a species-specific allele.
6-PGD, a dimeric enzyme with two loci in 21332; (Byrne and Littleton,
1988; Chaparro et al., 1987; Hauagge et al., 1987), was monomorphic in
plum, exhibiting a single-banded pattern. Apricot was monomorphic at
6-PGD-l, exhibiting a single band at the same migration distance as
that for 6-PGD-l in plum. 6-PGD-2 was polymorphic in apricot,
exhibiting two banding patterns, a singlesbanded pattern at the same
migration location as exhibited in plum and a triple-banded pattern.
If the apricot parent was heterozygous at 6-PGD-2, then this enzyme
system could be used to identify hybrid plumcot progeny.

Byrne and Littleton (1988) concluded that while the five enzyme
systems, PGI, LAP, MDH, PGM, and 6-PGD, could be useful in identifying
plumcots, depending on the genotype of the plum and/or apricot parent,
only peroxidase could be used to uniquely identify plum x apricot
hybrids. Peroxidase was the only system in which plum and apricot

exhibited no alleles in common. Parfitt et a1. (1985) reported on the

use of two enzyme systems, phosphoglucose isomerase (PGI) and

52

phosphoglucomutase (PGM), to identify plum x peach (Prunus a1 x
21333; pgggiga) interspecific hybrids. Malate dehydrogenase (MDH),
leucine aminopeptidase (LAP), and esterase (Est) were also tested but
were not useful in identifying such hybrids.

Both PGI and PGM produced two zones of activity in plum and peach.
In other plant species PGM and PGI produce two zones of activity which
have been associated with individual loci (Gottlieb, 1981).

Parfitt et a1. (1985) analyzed eighteen plum cultivars, 62 peach
cultivars, and nine plum x peach interspecific hybrids for PGI and PGM.
Peach was found to be monomorphic for PGI-1 and PGI-2, producing a
single band for each locus. The peach PGI-2 band was at the same
migration distance as one of the plum PGI-2 bands. Plum was monomorphic
for PGI-1, producing a single band with a greater mobility than the
peach PGI-l band. Plum displayed three banding patterns for PGI-2 -- a
"fast" single-band pattern, a "slow" single-band pattern, and a triple-
banded pattern with the "fast" band, the "slow" band, and an
intermediate band. The triple-banded hybrid pattern indicates that
PGI-2 functions like a dimeric enzyme in plum, as in other plant
species (Arulsekar and Bringhurst, 1981; Gottlieb, 1977). Since the
peach and plum isozymes represented by the PGI-1 band had different
migration rates, PGI-1 was useful to uniquely characterize plum x peach
interspecific hybrids. Peach was monomorphic for PGM-1, showing a
double-banded pattern and for PGM-2, with a single-banded pattern. All
three peach bands had faster migration rates than those found in plum
PGM-1 and PGM-2 zones.

Plum was polymorphic for PGM-1 (two patterns, a double- and a

triple-banded pattern) and for PGM-2 (two patterns, a single- and a

53

double-banded pattern). Since plum and peach did not share any common
bands at either PGM-1 or PGM-2, both loci could be used to characterize
plum x peach hybrids. The interspecific hybrids all displayed banding
patterns that combined both parental types for PGM-1 and PGM-2.

In cherry, only three isozyme studies have been reported. Hancock
and Iezzoni (1988) analyzed 19 sour cherry (2. geragug L.) cultivars,
six sweet cherry (2. axium L.) cultivars, four ground cherry '

(B. fgggiggsg Pall.) clones, the open-pollinated progeny of two sour
cherry cultivars, plus the open-pollinated progeny of P. ingigg,

2. figbhirtglla, B. canescens, and 2. mahalgb for MDH variability. MDH
functions as a dimeric enzyme in many species (Orton, 1983; Sari Gorla
et al., 1986; Cheliak and Pitel, 1985) and as a multimeric enzyme in
peach (Arulsekar et al., 1986). R. axium and E. frutigggg share some
alleles for MDH. In addition, each also exhibited unique alleles.

R. geggsug has a banding pattern that exhibits codominant expression of
both the B. ayium and the 2. firutiggga bands. Hancock and Iezzoni's
findings (1988) supported the hypothesis that 2. £2133 and 2. fruticosa
are the progenitor species of 2. ggrgggs (Olden and Nybom, 1968).
Open-pollinated progeny of 2. mahalgb and E. canesggng exhibited
invariate multi-banded patterns. The open-pollinated progeny of

P. incisa, E. subhirtglla, and the two sour cherry cultivars each
exhibited two banding patterns. Hancock and Iezzoni (1988) concluded
that MDH exhibits enough polymorphism to be used to identify
interspecific hybrids in cherry.

Fernquist and Huntrieser (1986-1987) used isozyme analysis to
characterize five sour cherry cultivars and six wild sweet cherry

genotypes. 6-PGD exhibited two zones of activity in both the sour and

54

sweet cherries studied. In sour cherry 6-PGD-2 was invariate,
exhibiting a double-banded pattern. 6-PGD-l in sour cherry exhibited
three triple-banded patterns. The sweet cherry genotypes were
polymorphic for both loci. In sweet cherry 6-PGD-2 exhibited three
banding patterns, a single-, a double- and a triple-banded pattern.
6-PGD-l exhibited three single-banded patterns, plus a double- and a
triple-banded pattern. Fernquist and Huntrieser (1986-87) state that
they found sufficient polymorphism in 6-PGD, PGI, MDH, shikimate
dehydrogenase (SkDH), hexokinase (Hk), and PGM to uniquely characterize
sour cherry cultivars and sweet cherry genotypes using cross
examination of these enzyme systems (details not disclosed).

Kaurisch et a1. (1988) employed isozyme analysis to examine five
Emma: Species; P. aim. P. mam. 12. mm. 2. memos. and
B. fighhirggllg. The enzyme systems tested were aconitase (Acon),
isocitrate dehydrogenase (IDH), MDH, 6-PGD, SkDH, PGI, LAP, and PGM.
Useful polymorphisms were found within the following loci; Acon-2,
6-PGD-l, 6-PGD-2, IDH-2, and PGI-2. The six sweet cherry cultivars in
the study were monomorphic at 6-PGD-2, displaying a single "slow" band.
The four sour cherry cultivars were also monomorphic at 6-PGD-2,
displaying the single "slow" band. 2. canescgns and 2. subhirtella
were monomorphic at 6-PGD-2, displaying a single "fast" band. 2.
fzggiggga displayed a triple-banded hybrid.pattern, including the
”slow" and "fast" bands and an intermediate band. Using this locus
alone, 2. figugigggg could be distinguished from the other four Pluggg
species. 2. ayium and 2. cerasug could not be distinguished from each
other but they could be distinguished from 2. ganesceng and P.

5 (:6 a.

55

Sweet cherry was polymorphic at 6-PGD-1 and displayed two
patterns, a single "fast" banded pattern and a triple-banded hybrid
pattern. 2. gggggug exhibited the same patterns at 6-PGD-1.

B. gagggggng, 2. ggbhizgglla, and 2. frggiggga all exhibited a
monomorphic single "fast" banded pattern at 6-PGD-1. While these
patterns suggest a typical Mendelian distribution for a dimeric enzyme,
Kaurisch et a1. (1988) noted that they found variants at 6-PGD-1 in the
tetraploid cherry species (2. geragug and g. fruticosa) for which they
_ could offer no genetic explanation.

Sweet cherry exhibited three banding patterns at Acon-2, two
single-banded patterns with either a "fast" or a "slow" band and a
double-banded pattern with both bands. Aconitase exhibited banding
patterns typical of a monomeric enzyme. Sour cherry exhibited two
banding patterns at Acon-2, a single "slow" banded pattern and the
hybrid double-banded pattern. 3. gagescggg and P. fruticgsa were both
monomorphic at Acon-2, exhibiting the single "fast" banded pattern. 2.
subhirtella exhibited the hybrid double-banded pattern.

Sweet cherry and P. ggnegcens were monomorphic for IDH-2, with a
single "fast" banded pattern. P. ggzaggg displayed two patterns, a
single "fast" banded pattern and a triple-banded pattern. 2. frugiggga
and 2. gubhigtella were not tested for IDH.

Sweet cherry was polymorphic for PGI-2, with two banding patterns,
a single-banded pattern and a triple-banded pattern. P. frugicosa and
2. subhirtglla were found to be monomorphic at PGI-2, with the same
single band found in sweet cherry. Sour cherry also exhibited this
single-banded pattern in addition to two hybrid triple-banded patterns.

2. gagesgegs was monomorphic for PGI-2, with a single-banded pattern at

56

a slower migration rate than that found in the other species. Kaurisch
et a1. concluded that PGI functions as a dimeric enzyme in cherry, as
in other Prunus species (Hauagge et al., 1987; Parfitt et al., 1985).
Between the five species, Kaurisch et al. (1988) propose the occurrence
of four alleles at PGI-2.

Kaurisch et a1. (1988) observed polymorphisms in LAP, PGM, and
SkDH in the five Prunus species studied. However, the banding patterns
were not consistent, so these enzyme systems were not used in the
study.

For MDH, Kaurisch et a1. (1988) found that both sweet and sour
cherry exhibited the same four-banded pattern. 2. gangsgegg displayed a
distinctive double-banded pattern. 2. gubhirtella also displayed a

distinctive five-banded pattern. 2. fruticgsa was not tested for MDH.

MATERIALS AND METHODS

Seeds and leaves for isozyme analysis were collected from trees
growing at the Clarksville Horticultural Experiment Station,
Clarksville, Michigan, and the Hortidultural Research Center, East
Lansing, Michigan. Four sweet cherry cultivars, two ground cherry
selections, and 'Montmorency' sour cherry were analyzed for isozyme
variation. Mature fruit was collected during the summer of 1988. The
fleshy pericarp was removed and the endocarps were dipped in a Thiram
(fungicide) solution, dried overnight, and stored in polyethylene bags
at approximately 5C. Young, newly expanded leaves were collected
during the spring and summer of 1988. They were placed in polyethylene
bags and stored on ice at approximately BC. Also, one-year-old

branches were collected in February, forced in the laboratory at

57

ambient temperature for approximately one week, and the newly expanded
leaves used for isozyme analysis.

Seeds and leaves were ground by hand in porcelain mortars and
pestles which were cooled to OC. The extraction buffer consisted of
0.05 M Tris base, 0.007 M citric acid (monohydrate), 0.1% cysteine
hydrochloride, 0.1% ascorbic acid, 1.0% polyethylene glycol (Mr 3500)
and 1mM 2-mercaptoethanol, final pH approximately 8.0 (Arulsekar et
al., 1986). Approximately 0.3 ml extraction buffer was used to grind
each seed. One gram of insoluble polyvinyl polypyrrolidone (PVPP,
Sigma) was added to each 12 m1 extraction buffer for use with leaf
tissue. Approximately 0.3 m1 extraction buffer was used to grind 15 mg
of minced leaf tissue. After grinding, the plant tissue was held in
the mortars on ice prior to electrophoresis. Small squares of fine
nylon mesh were placed over each sample. 3 x 13 mm. paper wicks were
placed on top of the nylon mesh to absorb supernatant from the samples.

Four kinds of buffer systems were used depending on the enzymes
studied. A histidine gel buffer system (Arulsekar et al., 1986) was
used for MDH, ADH, SkDH, and PGM. The gel buffer consisted of 1.2
g/liter DL-histidine-HCL (monohydrate), pH adjusted to 7.0 with sodium
hydrochloride. The electrode buffer consisted of 18.2 g/liter Tris
base and 9.0 g/liter citric acid (monohydrate), pH appproximately 7.0.
A histidine-citric acid buffer system (Durham et al., 1987) was used
for 6-PGD, IDH, PGI, and Acon. The gel buffer consisted of 0.009 M
histidine and 0.003 M citric acid, pH adjusted to 5.7 with citric acid.
The electrode buffer consisted of 0.065 M histidine and 0.02 M citric
acid, pH adjusted to 5.7 with citric acid. A morpholine-citric acid

buffer system (Clayton and Tretiak, 1972) was used for 6-PGD, IDH, and

58

MDH. The gel and electrode buffer consisted of 0.04 M citric acid
(monohydrate), pH adjusted to 6.1 with N-(3-amino propyl) morpholine.
For gels the buffer solution was diluted 1:20 with deionized water. A
Tris base-citric acid gel buffer system (Kaurisch et al., 1988) was
used for LAP and PX. The gel buffer consisted of 0.015 M Tris base and
0.003 M citric acid, pH adjusted to 7.7 with citric acid. The
electrode buffer consisted of 0.031 M NaOH and 0.3 M boric acid, pH
adjusted to 7.5 with l N NaOH.

Eleven percent starch gels were prepared by placing 33 g of potato
starch (Sigma) and 300 ml of gel buffer in a dry l-liter vacuum flask.
The solution was vigorously swirled over a bunsen burner flame until it
began to boil. Vacuum was applied to the flask until large bubbles
formed. The gel was then poured onto the gel plate (length 22 cm.,
width 21 cm., depth 10 mm.) and any bubbles or undissolved starch
flakes were removed with a pipette. The gel was allowed to cool for
approximately 45 minutes and covered with polyethylene film. Next
morning the gel was cooled at approximately 5C for twenty minutes prior
to electrophoresis.

Just prior to electrophoresis the gels were removed from the
refrigerator. The gel was separated on all 4 sides from the gel plate
frame with a scalpel. A slot was cut in the gel at a 90 degree angle
to the direction of electrophoresis, approximately 5 cm. from one end
(referred to as the lower or cathodal end). The lower gel plate frame
piece was removed. The cut gel was pulled back slightly to allow
insertion of the paper wicks. The wicks were picked up with a forceps,
blotted on a paper towel, and placed in the slot. After all samples

were inserted, the shorter piece of gel was pushed against the larger

59

to ensure contact with the sample wicks. The lower gel frame piece was
replaced. All gel buffer systems were run at 40 mA for approximately 5
hours, being removed after the first 30 minutes. At the end of the
run, gels were sliced transversely. Slices were immersed in the
various staining solutions (Table l) and incubated at approximately 35C
in the dark until the isozyme bands became visible. The excess
staining solution was decanted. Gel slices were rinsed with 1.0%
acetic acid, and the rinse decanted. The gel slices were either rinsed
several times with water or immersed in 50% glycerol or 50% ethanol for

approximately 30 minutes and then stored in polyethylene bags.

RESULTS AND DISCUSSION

Preliminary isozyme gels in the early spring of 1988 revealed that
resolution for several enzyme systems was possible. However, as the
season advanced and summer temperatures regularly reached 32C activity
was almost completely lost. Despite using many different recipes for
stains, grinding buffers, electrode and gel systems it is apparent that
young, partially expanded cherry leaves are virtually mandatory for
electrophoretic success with cherries. More mature cherry leaves
accumulate phenolics which interact with proteins and inhibit enzyme
activity (Wendel and Parks, 1982). Therefore, in midsummer of 1988 the
isozyme study was continued with cherry seeds. Although the
pollinations were not controlled and a formal inheritance study could
not be conducted, the seed zymograms were useful in revealing the range
of polymorphisms for several enzyme systems.

MDH functions as a dimer in many crops (Yang et al., 1977; Orton,

1983; Morgan and Bell, 1983) and as a multimer in peach (Arulsekar et

 

60

Table l. Staining solutions for enzymes

 

Ingredients Amount

 

Glucose phosphate isomerase (PGI)z

NADP 10
PMS 5
MTT 15
l M Tris-HCl (pH 8.0) 10
0.1 M MgC12 6H20 10
DI water 80
0.18 M fructose-6-phosphate 2
Glucose-6-phosphate dehydrogenase 40

Alcohol dehydrogenase (ADH)z

NAD 50
PMS 5
MTT 15
0.1 M Tris-HCl (pH 7.5) 100
Ethanol (100% absolute) 3

Malate dehydrogenase (MDH)z

NAD 30
MTT 15
PMS 5
1 M Tris-HCl (pH 8.0) 20
1 M sodium L-malate 10
DI water 70

mg
mg
mg
ml
m1
ml
ml
units

mg
mg
mg
ml
ml

mg
mg
mg
ml
ml
m1

 

zArulsekar and Parfitt (1986)

61

Table 1 (cont'd.).

 

 

Ingredients Amount
6-phosphog1uconate
dehydrogenase (6-PGD)z
NADP 10 mg
MTT 15 mg
PMS 5 mg
6-phosphog1uconic acid 50 mg

(sodium salt)

1 M Tris-HCl (pH 8.0) 5 m1
DI water 95 m1
Phosphoglucomutase (PGM)z
NADP 10 mg
PMS 5 mg
MTT 15 mg
alpha-D-glucose-l-phosphate 75 mg
l M Tris-H01 (pH 8.0) 2 ml
DI water 93 ml
Glucose-6-phosphate dehydrogenase 40 units
Isocitrate dehydrogenase (IDH)y
NADP 10 mg
MTT 15 mg
PMS 2 mg
Isocitric acid,trisodium salt 100 mg
1.0 M Tris-HCl (pH 8.0) 10 m1
DI water 85 ml
1.0 M MgC12 5 m1

 

zArulsekar and Parfitt (1986)
ySoltis et a1. (1933)

62

Table l (cont'd.).

 

 

Ingredients Amount

Aconitase (Acon)y
NADP 15 mg
MTT 10 mg
PMS 2 mg
cis-aconitic acid 100 mg
1.0 M Tris-H01 (pH 8.5) 10 ml
DI water 90 ml
1.0 M Mg612 1 ml
Isocitrate dehydrogenase 7 units
Leucine aminopeptidase (LAP)y
L-leucine-beta-naphthy1amide 20 mg

dissolved in dimethyl formamide 5 ml
1.0 M phosphate buffer, pH 6.0 10 ml
DI water 90 m1
Black K salt or fast black K salt 50 mg
Peroxidase (PX)y
3-amino-9-ethyl carbazole 65 mg

dissolved in dimethyl formamide 5 ml
0.05 M sodium acetate buffer, pH 5.0 95 m1
Shikimate dehydrogenase (SkDH)y
NADP 10 mg
MTT 10 mg
PMS 2 mg
1.0 M Tris-HCl buffer, pH 8.5 10 ml
DI water 90 ml
Shikimic acid 100 mg

 

ySoltis et a1. (1933)

63

al., 1986). As previously reported (Hancock and Iezzoni, 1988), our
results confirm that sour cherry leaves exhibited a five-banded pattern
for MDH (Figure 1). Leaf tissue of sweet cherries 'Hedelfingen’ and
'Emperor Francis' had a triple-banded pattern, while the sweet cherry
'Angela' displayed a four-banded pattern. Both clones of ground cherry
exhibited four-banded patterns. Sweet and ground cherry, while sharing
some bands, also each exhibited unique bands. Sour cherry exhibited
codominant expression of the sweet and ground cherry bands. Cherry
seeds revealed the same banding patterns for each species plus often an
indefinite number of lower bands. Hancock and Iezzoni (1988) found
that sweet and sour cherry pollen displayed an additional four lower
bands. This indicates that cherry pollen and seed have at least one
additional MDH locus.

PGM functions as a monomer in many crops (Cheliak and Pitel,
1985), including plum and peach (Byrne and Littleton, 1988; Chapparo et
al., 1987). The two clones of ground cherry exhibited a single "slow"-
banded pattern for leaf tissue (Figure 1). The sweet cherries
'Hedelfingen', and 'Emperor Francis' (plus sweet cherry 'Napoleon')
exhibited a double-banded pattern for leaf and seed tissue. The sweet
cherry 'Angela' and the sour cherry 'Montmorency' (plus sour cherries
'Northstar', 'Meteor', and 'Montearly') exhibited a double-banded
pattern with a "slow" band at the same mobility as in ground cherry and
a "fast" band at the same mobility as in sweet cherry. Seed tissue of
ground cherry also exhibited this double-banded pattern, indicating
cross-pollination had occurred. Sweet cherry seed tissue of 'Van',
'Hedelfingen', and 'Emperor Francis' revealed the double~banded pattern

found in 'Hedelfingen' and 'Emperor Francis' leaf tissue.

 

64

 

.m:umuumm 0:00:00 mfi>~omw you Uoumomummm «10«0 mH 0:0 0:500 mH .ao:muoau:o= mo mammfiu Ummwx

.omummu no: mo: 0H00:< no mammwu comm>
.vmumou #0: was :m> mo osmmflu 00049

 

 

 

 

II II II x6000
II II. II
I I I I H I Sawmill
II II II II. “00H
II II.
II II II II IUM
II II _II
II II II. II II II. and:
I I I n
I I I I I I all:
II II II II
mflo:mum .
>0H00:< uoummfim :mm:fluaopm= N:w> «1000 0H «I500 mH >0:muoau:oz
flmwwm .M mwmmwwmmw .M mmwmmwm .N
wmmmmu 90030 000000 025000 000000 0:00

 

.osmmwu 600m 6:0 Mama auumno umoam 0:0

.>uuo:o 0:5000 .muum:o know no mamumoa>0 .H 00:000

65

 

.m:00uu0m 0:00:00 0500000 000 0000000000 :00:0000000 0:0 :0> .>0:00oau:o= no 050000 00003
.m:000000 0:00:00 0B>Nomw 000 0000000000 ~I0~0 0H 0:0 0:500 00 .00:00oau:ozuo 050000 00000
.000000 00: m0: 0000:< 00 050000 0000»

.000000 00: 003 :0> 00 050000 0000N

 

 

 

 

 

 

 

II II. II
I I I I I n I I I 380“
II. II II
I II II I I II :oo<
II
II II II II _II II 0000
II II II.
I I I I I I :00‘
l I m I gnaw“
II II II.
II II. II _flInI __IIII .mull 000
I I I HMS
II II II
II II II II _II II mmfl
000:000
00000:< 0000030 :00:0000000 N:0> 01000 0H Hiram 0H >0:00oau:o=
00000 .M mmmmflwmmw .N mdmmwmm .N

wmmmmo 80030

wmmmzo DZDOMU

Nmmmmu mDOm

 

.A.0.ucoov 0 000000

66

 

.mcumuumm

TCflvcmn maauomfi you umumomummm cm> can .munmn mH .Huumm «H .wocmuoaucoz mo momma» ummmx
.fimummu no: mmS Mammcd no mammfiu Ummmw

.flmummu uoc ma? cm> HO mfimmflu. “MEN

 

 

 

 

_II II. II, II
II II II II
II II IMII_ II:I- HImI. _lIflI xvflwm
II _II Illl. Ilnl. .IIII _llll MGM
_II
I HMS
II II II 1' II_ II.
:GH
II II II II .II II
mflocmum ,
hanmmcc uoummam cmmcauamvmm ucm> . munmn mH Huhwm «H >Ucmuoauco:
amwwm .N mwmwwmmmw .M mmwmmwm .N
wmmmmo Hausa >mmmmo 0290mm ummmmo zoom

 

.A.o.u:oov a musmwm

67

 

.mcnmuuma mcwccmn mamnomfl you noummmummm mummn mH can Hiram mH .aocmuoaucoz mo mammwu cmwmn

 

 

 

 

ll-l. .llllil. nl:lullil=llll
II II Illl_ II
I I I I I N80“
.llll _llllil. Ilnl. _llll
.Illl .Illl .l-II comic
_IIII .llll II _IIII-l_
II II DU
II _II II
II II. II mama
II
II II [I mwmlw
II II
II II. II
Hummm mH Huumm mH mocmuoaucoz
mmmmflwmmw .M mmmmumm .N
ammmmo 0290mm ummmmu moom

 

.A.c.ucoov H musmflm

.mcumuumm weancmn mshuomfl you vmummmummm mfiocmum uonmmam can cmocwuamcmm .:m> no mammfiu cmmmx
.cmummu uo: mm3 wamocd mo mammwu Ummmh
.vmumou no: mm: cm> no mammflu mama»

 

 

 

II II III
II
II wimm
II II .HIII
II II IIII. Dumlw
II-I_ _Illl _IllI-I_
II _II II
m use;
II II II awmlm
II II II
.II [I .ll

 

>mamoc< mflocmum uoummfim :mOCHMHwomm cm>

 

mmflfim .N

wmmmmu BMfiSm

 

.A.o.ucoov H musaflm

69

 

.Umummu no: mmz mammc< mo mammwu ummm»
.vmummu no: mm3 cm> mo mammwu wmoMN

 

 

 

 

 

II. II II. II II 600m
II .II II II II. II
II II II. II 4.. HI WW“
II II II II II Mama
II II. .ll IU II
_II II II [I II II me
mflocmum .
hmammcd uoummam cmwcfiuamumm Nam> mlmmm «H Huhwm mH >ocmhoaucoz
asw>m .N mmmmflmmmw .M mmmmmmw .m
ammmmu 8mm3m ammmmo ozoomo ammmmo moon

 

.A.u.ucoov a muzoflm

70

'Montmorency' seed exhibited the double-banded pattern found in leaf
tissue plus a single "slow"-banded pattern.

ADH functions as a dimer in other crops (Hart, 1970; Scandalios,
1974). It is known to increase in activity under anaerobic conditions.
Unlike most enzymes whose activity increases after germination, ADH
activity declines rapidly with germination. Thus, ADH zymograms from
seed tissue proved to be excellently resolved, while those from leaf
tissue were extremely faint. The leaf tissue for sour and ground
cherry appeared to exhibit an invariate triple-banded pattern
(Figure 1). Leaf tissue of sweet cherries 'Hedelfingen', 'Emperor ;

Francis', and 'Angela' (plus sweet cherry 'Napoleon') exhibited a

 

monomorphic single "fast"-banded pattern. In addition, all the sweet
cherry cultivars exhibited the same single "fast"-banded pattern for
seed tissue. The sour cherry 'Montmorency' and ground cherry IR 587-1
exhibited the single "fast"-banded homodimeric pattern plus a triple-
banded heterodimeric pattern. Ground cherry IR 323-2 proved to be most
interesting because it displayed both homodimeric patterns, the single
"slow" and "fast"-banded patterns plus the triple—banded heterodimeric
pattern. This segregation into three patterns was observed in a seed
population size of 70 (Appendix A). In addition it appeared that
dosage effects could be detected in many of the triple-banded
zymograms.

Acon functions as a monomer in several crops, including cherry
(Kaurisch et al., 1988). 'Montmorency' (plus sour cherries 'Meteor'
and 'Montearly') and the ground cherries exhibited a similar triple-
banded pattern for leaf tissue (Figure 1 ). Sweet cherries

'Hedelfingen' and 'Angela' exhibited a double-banded pattern for leaf

71

tissue. The bands were at the same migration distance as the "fast"
and "slow” bands in the sour and ground cherry zymograms. Sweet cherry
'Emperor Francis' exhibited a triple-banded pattern sharing the "slow"
and "fast" bands found in the other cultivars and clones, plus a unique
"slow“ band at a more cathodal position. (Sour cherry 'English
Morello' and sweet cherry 'Napoleon' also exhibited this pattern.)
Kaurisch et al. (1988) interpreted the consistent "fast” band found in
all cherry genotypes as representing one locus, and the variable lower
bands as representing a second locus. Seed tissue of the ground cherry
clones exhibited the same triple-banded pattern found in leaf tissue.
'Montmorency' seed exhibited the triple-banded pattern plus a double-
banded pattern. Seed of sweet cherries 'Van' and 'Hedelfingen' also
exhibited this double-banded pattern. Additionally, 'Van',
'Hedelfingen', and 'Emperor Francis' seed exhibited a triple-banded
pattern containing bands at the ”fast" and "slow” positions exhibited
in the sour and ground cherry zymograms plus a lower band at the same
position as the lowest band in the 'Emperor Francis' leaf tissue
zymogram.

IDH functions as a dimer in several crops, including cherry
(Kaurisch et al., 1988). Sweet cherry and ground cherry leaf tissue
exhibited a similar double-banded pattern (as did sour cherry 'Meteor')
(Figure l). 'Montmorency' exhibited a four-banded pattern, sharing the
two bands exhibited in sweet and ground cherry plus two lower bands
unique to 'Montmorency'. As with aconitase, Kaurisch et al. (1988)
interpreted the invariate "fast" band as representing a separate locus.
Seed tissue of sweet cherries 'Hedelfingen' and 'Emperor Francis' had

the same banding pattern as in the leaf tissue. However, the sweet

 

72

cherry 'Van', while sharing this banding pattern, also exhibited the
four-banded pattern as found in 'Montmorency'. Seed tissue of
'Montmorency' exhibited the four-banded pattern found in leaf tissue
plus a double-banded pattern. Conversely, seed tissue of both ground
cherry clones exhibited the double-banded pattern found in leaf tissue
plus the four-banded pattern.

6-PGD functions as a dimer in many plant species (Goodman and
Stuber, 1983; Brown, 1983; Kiang and German, 1983). The three cherry

species studied revealed two zones of activity for 6-PGD, as found by

“V

Kaurisch et al. (1988) (Figure l). 'Montmorency' leaf tissue exhibited

 

“I‘a'n

triple-banded patterns in both zones (as did sour cherries 'Montearly'
and 'Meteor'). Ground cherry IR 323-2 leaf tissue also had this E
pattern. Leaf tissue of ground cherry IR 587-1 exhibited the triple-

banded pattern common to 'Montmorency' and the other ground cherry

clone for the zone of slower mobility. In the zone of faster mobility

IR 587-1 exhibited a unique triple-banded pattern. The bands in this

zone shared the fastest and slowest bands of the other triple-banded

pattern. In addition, ground cherry IR 587-1 had a unique third band at

a slower mobility in the upper zone of activity. Leaf tissue of all

the sweet cherry cultivars (plus sweet cherry 'Napoleon') exhibited the

triple-banded pattern in the zone of faster mobility and a single-

banded pattern in the zone of slower mobility. 'Montmorency' seed

tissue exhibited a range of zymograms (Figure 1) including the

homodimeric "slow”-banded pattern in the zone of slower mobility and

both homodimeric ("slow" and "fast"-banded) patterns in the zone of

faster mobility plus the triple-banded pattern in both zones of

mobility and various combinations of the heterodimeric and homodimeric

73

banding patterns. The three banding patterns (both homodimeric
patterns and the heterodimeric pattern) in the zone of faster mobility
were observed in a seed population size of 73 (Appendix A). Ground
cherry IR 587—1 seed tissue exhibited triple-banded patterns in both
zones, as found in ground cherry IR 323-2 leaf tissue. In addition, IR
587-1 exhibited the "slow"-banded homodimeric pattern in both zones of
slower and faster mobility. Ground cherry IR 323-2 seed tissue
exhibited the triple-banded pattern in both zones, as found in leaf
tissue, plus the "slow"-banded homodimeric pattern in the zone of
slower mobility. Sweet cherry 'Van' seed tissue exhibited the same
pattern as found in the other sweet cherry cultivars' leaf tissue, plus
the triple-banded heterodimeric pattern in the zone of slower mobility
and the "fast"-banded homodimeric band found in the zone of faster
mobility. 'Hedelfingen' and 'Emperor Francis' seed tissue exhibited
the same pattern as found in leaf tissue, plus the "fast"-banded
homodimeric pattern in the zone of faster mobility. In addition it
appeared that dosage effects could be detected in many of the triple-
banded zymograms.

PGI functions as a dimer in many crops (Arulsekar and Bringhurst,
1981; Gottlieb, 1977). All the sweet cherry cultivars (plus sweet
cherry 'Napoleon') exhibited a monomorphic triple-banded pattern for
leaf and seed tissue (Figure l). 'Montmorency' leaf and seed tissue
exhibited this same pattern as did ground cherry IR 323-2. (Leaf
tissue of sour cherries 'Meteor' and 'Montearly' also exhibited this
pattern.) Ground cherry IR 587-1 revealed a single-banded pattern for
leaf tissue. (This pattern was also found in sour cherry cultivars

'English Morello' and 'Northstar'.) Seed tissue of IR 587-1 exhibited

 

74

the monomorphic triple-banded pattern as found in the other chery
genotypes.

Other enzyme systems in cherry with apparent polymorphisms but
inconsistent resolution include, LAP, SkDH, and PX.

It is apparent from the zymograms for the three species, sweet,
sour, and ground cherry, that a great deal of isozyme homology does
exist between the three species. This finding concurs with
Raptopoulos' hypothesis (1941) that all three species may have arisen
from a common diploid ancestor. This study also demonstrates that
abundant polymorphisms exist in several enzyme systems both within and
between the three cherry species. This is in sharp contrast to the
situation in another 2133;; species, peach. Arulsekar et al. (1986)
had to test ten and Durham (1987) thirty-eight enzyme systems to find,
respectively, one and three enzyme systems that revealed polymorphisms.
The wealth of polymorphisms found in cherry promises to provide

valuable genetic markers for future research.

LIST OF REFERENCES

 

LIST OF REFERENCES

Adams, M. W. 1977. An estimation of homogeneity in crop plants with
special reference to genetic vulnerability in the dry bean,

Phaggolug vulgarig L. Euphytica 26(3):665-679.

Adams, W. T. and R. J. Joly. 1980. Linkage relationships among twelve
allozyme loci in loblolly pine. J. Hered. 71:199-202.

Adams, W. T. and J. V. Wiersma. 1978. An adaptation of principal
components analysis to an assessment of genetic distance. Mich.
Agric. Exp. Stn. Res. Rep. 347.

Afifi, A. A. and V. Clark. 1984. Computer-aided multivariate analysis.
Lifetime Learning, Belmont, CA.

Allard, R. W., A. L. Kahler, and B. S. Weir. 1971. Isozyme
polymorphisms in barley populations. Barley Genetics Symp. Proc.
2D:1-13.

Arulsekar, S. and D. E. Parfitt. 1986. Isozyme analysis procedures for
stone fruits, almond, grape, walnut, pistachio, and fig. HortSci.
21(4):928-933;

Arulsekar, S., D. E. Parfitt, W. Beres, and P. E. Hansche. 1986a.
Genetics of malate dehydrogenase isozymes in the peach. J. Hered.
77:49-51.

Arulsekar, S., D. E. Parfitt, and D. E. Kester. 1986b. Comparison of
isozyme variability in peach and almond cultivars. J. Hered.
77:272-274.

Arulsekar, 8., D. E. Parfitt, and G. H. McGranahan. 1985. Isozyme gene
markers in Juglans species. J. Hered. 76:103-106.

Arus, P., and C. R. Shields. 1983. Cole crops (Braggigg glgxagga L.),
p. 339-350. In: S. D. Tanksley and T. J. Orton (eds.). Isozymes in
plant genetics and breeding, part B. Elsevier, Amsterdam.

Bailey, L. H., et al. 1976. Hortus third: a concise dictionary of
plants cultivated in the United States and Canada. Macmillan, New
York.

Barg, T. 1958. Cytological investigations on sour cherries (in
German). Gartenbauwiss. 23:200-208. [Hort. Abstr. 29:1175c; 1959].

75

76

Bassiri A. 1976. Barley cultivar identification by use of isozyme
electrophoretic patterns. Can. J. Plant Sci. 56:1-6.

Bemis W. P., A. M. Rhodes, T. W. Whitaker and S. G. Carmer. 1970.
Numerical taxonomy applied to Cucurbita relationships. Amer. J.
Bot. 57(4):404-412.

Blasse, W. 1957. Heredity analysis of interspecific hybrids between
Prunus gerasus L. and 2. azigm L. (in German). Arch. Gartenb.
5:104-172. [Plant Breed. Abstr. 29:668; 1959].

Blogg, D. and B. C. Imrie. 1982. Starch gel electrophoresis for soybean
cultivar identification. Seed Sci. and Technol. 10:19-24.

Brown A. H. D. 1983. Barley. p. 55-78. In: S. D. Tanksley and T. J.
Orton (eds.). Isozymes in plant genetics and breeding, part B.
Elsevier, Amsterdam.

Brown A. H. D. and R. W. Allard. 1970. Estimation of the mating system

in open-pollinated maize populations using isozyme polymorphisms.
Genetics 66:133-145.

Byrne D. H. and T. G. Littleton. 1988. Electrophoretic characterization
of diploid plums of the southwestern United States. J. Amer. Soc.
Hort. Sci. 113(6):918-924.

Byrne D. H. and T. G. Littleton. 1989. Interspecific hybrid
verification of plum x apricot hybrids via isozyme analysis.
HortSci. 24(1):132-134.

Carter, K. K., F. C. Cech, and D. H. DeHayes. 1983. Geographic
variation in Prunus sgggtina. Can. J. For. Res. 13(6):1025-1029.

Cech, F. C. and K. K. Carter. 1979. Geographic variation in black
cherry - ten year results of a West Virginia provenance test.
Proceedings of the First North Central Tree Improvement
Conference. Madison, WI. pp. 21-27.

Challice, J. S. and M. N. Westwood. 1973. Numerical taxonomic studies
of the genus Eyrus using both chemical and botanical characters.
Bot. J. Linn. Soc. 67:121-148.

Chaparro, J. X., R. E. Durham, G. A. Moore, and W. B. Sherman. 1987.
Use of isozyme techniques to identify peach x 'Nonpareil' almond
hybrids. HortSci. 22:300-302.

Chaparro, J. X., R. G. Goldy, B. D. Mowrey, and D. J. Werner. 1989.

Identification of 21:1; xinifgzg L. x Muscadinia Igtggdifiolia
small hybrids by starch gel electrophoresis. HortSci. 24(1):128-
130.

Cheliak, W. M. and J. A. Pitel. 1985. Inheritance and linkage of
allozymes in Larix larigina. Silvae Genet. 34:142-148.

77

Clayton, J. W., D. N. Tretiak. 1972. Amine-citrate buffers for pH
control in starch gel electrophoresis. J. Fish. Res. Board Can.
29:1169-1172.

Cociu, V. and T. Gozob. 1985. The sour cherry breeding program in
Romania. Acta Horticulturae 169:91-96.

Conkle, M. T., P. D. Hodgskiss, L. B. Hunnally, and S. C. Hunter. 1982.
Starch gel electrophoresis of conifer seeds: a laboratory manual.
USDA For. Ser. Tech. Rep. PSW-64, Pacific Southwest Forest and
Range Exp. Sta., Berkeley.

Crawford, D. J. 1983. Phylogenetic and systemic inferences from
electrophoretic studies, p. 257-288. In: S. D. Tanksley and T. J.
Orton (eds.). Isozymes in plant genetics and breeding, part A.
Elsevier, Amsterdam.

Darlington, C. D. 1928. Studies in Prunus I and II. J. Genet. 19:
213-256.

de Candolle, A. 1882. Origin of cultivated plants. Hafner, New York.

Durham, R. E. 1986. Isozyme banding patterns and inheritance studies in
peach (Prunus pgrgigg L.). MS thesis, Univ. of Florida,
Gainesville.

Fernquist, I. and I. Huntrieser. 1986-1987. Use of isozyme analysis for
the identification of fruit cultivars and genotypes. In: Breeding
of Fruit and Small Fruit at Balsgard. Annual Report for 1986-1987,
the Swedish University of Agricultural Sciences, Division of Fruit
Breeding. Balsgard, Sweden.

Galletta, G. J. 1959. Comparative breeding, morphological and
cytological studies of polyploid sweet cherries (Prunus avium L.).
PhD Diss., Univ. of California, Davis.

Goodman, M. M., C. W. Stalser, K. Newton, and H. H. Weissinger. 1980.
Linkage relationships of 19 enzyme loci in maize. Genetics 96:697-
710. -

Goodman, M. M. and C. W. Stuber. 1983. Maize, p. 1-34. In: S. D.
Tanksley and T. J. Orton (eds.). Isozymes in plant genetics and
breeding, part B. Elsevier, Amsterdam.

Gottlieb, L. D. 1981. Electrophoretic evidence and plant populations.
Prog. Phytochem. 7:1-46.

Guries, R. P. and F. T. Ledig. 1978. Inheritance of some polymorphic
isoenzymes in pitch pine (Pings zigida Mill.). Heredity 40:27-32.

Hamrick, J. L. 1983. The distribution of genetic variation within and
among natural plant populations, p. 335-348. In: C. M. Shonewald-
Cox, S. M. Chambers, 8. MacBryde, and L. Thomas (eds.). Genetics
and Conservations. Benjamin/Cummings, Menlo Park, N.J.

 

78

Hancock A. M. and A. F. Iezzoni. 1988. Malate dehydrogenase isozyme
patterns in seven Prunus species. HortSci. 23(2):381-383.

Hancock, J. F. and R. S. Bringhurst. 1978. Inter-populational
differentiation and adaptation in the perennial, diploid species
Eggggria vggca L. Amer. J. Bot. 65(7):795-803.

Hancock, J. F. and R. S. Bringhurst. 1981. Evolution in California
populations of diploid and octoploid Egégggia (Rosaceae): a
comparison. Amer. J. Bot. 68(1):1-5.

Hart,G. E. 1970. Evidence for triplicate genes for alcohol
dehydrogenase in hexaploid wheat. Proc. Nat'l. Acad. Sci.
66(4):1l36-1141.

r!

Hauagge, R., D. E. Kester, S. Arulsekar, D. E. Parfitt, and L. Liu.
1987a. Isozyme variation among California almond cultivars: I. S
Inheritance. J. Amer. Soc. Hort. Sci. 112(4):693-698.

Hauagge, R., D. E. Kester, and R. A. Asay. 1987b. Isozyme variation
among California almond cultivars: II. Cultivar characterization
and origins. J. Amer. Soc. Hort. Sci. 112(4):687-693.

 

Hedrick, U. P. 1915. The cherries of New York. J. B. Lyon, Albany, N.Y.

Heiser, C. B. Jr., J. Soria, and D. L. Burton. 1965. A numerical
taxonomic study of Sglagum species and hybrids. Amer. Naturalist
99(909):471—488.

Hillig, K. W. 1988. A multivariate analysis of a sour cherry germplasm
collection. MS Thesis, Mich. State Univ., East Lansing.

Hotelling, H. 1933. Analysis of a complex of statistical variables into
principal components. Jour. Educ. Psych. 24:417-441.

Hruby, K. 1939. The cytology of the Duke cherries and their
derivatives. J. Genet. 38:125-131.

Hruby, K. 1950. The cytology of tetraploid cherries. Studia Botan.
Cechosl. 11:87-97. [Plant Breed. Abstr. 26:3705; 1956].

Hruby, K. 1962. Chromosome behaviour and phylogeny of cultivated
nggfigg. Preslia 34:85-97.

Hunter, R. L. and C. L. Markert. 1957. Histochemical demonstration of

enzymes separated by zone electrophoresis in starch gels. Science
125:1294-1295.

Hussaini, S. H., M. M. Goodman and D. H. Timothy. 1977. Multivariate
analysis and the geographic distribution of the world collection
of finger millet. Crop Sci. 17:257-263.

Iezzoni, A. F. 1984. Sour cherry breeding in Eastern Europe. Fruit Var.
J. 38:121-125.

79

Iezzoni, A. F. 1985. Sour cherry germplasm and cultivar development in
Romania. Rept. Nat. Acad. Sci. 14 pgs.

Iezzoni, A. F. and R. L. Hamilton. 1985. Differences in spring floral
bud development among sour cherry cultivars. Hortscience
20(5):915-916. .

Isleib, T. G. and J. C. Wynne. 1983. Heterosis in testcrosses of 27
exotic peanut cultivars. Crop Sci. 23(5):832-841.

Jardine, S. and R. Sibson. 1971. Mathematical taxonomy. Wiley, New
York.

Jensen, R. J. 1977. Numerical analysis of the scarlet oak complex

(Qggxggg subgen. Erythrobalgngg) in the eastern United States:
relationships above the species level. Syst. Bot. 2(2):122-133.

Jensen, R. J. and W. H. Eshbaugh. 1976a. Numerical taxonomic studies of
hybridization in ngrgus, I. Populations of restricted areal
distribution and low taxonomic diversity. Syst. Bot. 1(1):l-10.

Jensen, R. J. and W. H. Eshbaugh. 1976b. Numerical taxonomic studies of
hybridization in Quegcus, II. Populations with wide areal
distributions and high taxonomic diversity. Syst. Bot. 1(1):ll-19.

Jensen, R. J. and J. F. Hancock Jr.. 1982. Multivariate relationships
among California strawberries. Bull. Torrey Bot. Club. 109(2):136-
147.

Kobel, F. 1927. Cytological studies on Prunoideae and Pomoideae (in
German). Arch. Jul. Klaus-Stift. 3:1-84.

Kolesnikova, A. F. 1975. Species and ecologo-geographical groups of
the sorts of cherry used in selection. Priokskogo Izd-vo, Orel,
U.S.S.R. [Plant Breed. Abstr. 47:10743; 1977].

Kollar, G. 1987. Cultivars, pruning systems, general cultural practices
for sour cherry in Hungary. Proc. Intl. Cherry Res. Conf.,
Traverse City, Mich., pp. 19-26.

Kaurisch, P., W. Gruppe and W. Kohler. 1988. Enzyme polymorphisms in
cherries (Prunus species) and species hybrids (2. x species).
Method, selected species and rootstock-scion interactions. Angew.
Botanik 62:41-52.

Kiang, Y. T. and M. B. Gorman. 1983. Soybeans, p. 295-328. In: S. D.
Tanksley and T. J. Orton (eds.). Isozymes in plant genetics and
breeding, part B. Elsevier, Amsterdam.

Ladizinsky, G. and T. Hymowitz. 1979. Seed protein electrophoresis in
taxonomic and evolutionary studies. Theor. Applied Genet. 54:145-
151.

80

Martin, G. B. and M. W. Adams. 1987. Landraces of Ehaaaalaa galgazia
(Fabaceae) in Northern Malawi. 1. Regional variation. Econ. Bot.
41(2):190-203.

Michurin, I. V. 1949. Selected works. Foreign Languages Publishing
House, Moscow.

Michurin, I. V. 1955. Izbrannie sochineniya. Gosudarst-vennoe Izd-vo,
Moscow.

Moore, G. A. and G. B. Collins, 1983. New challenges confronting plant
breeders, p; 36. In: S. D. Tanksley and T. J. Orton (eds.).
Isozymes in plant genetics and breeding, part A. Elsevier,
Amsterdam.

Moore, G. A. and R. E. Litz. 1984. Biochemical markers for aniaa

aaaaya, Q. aaalifilara, and plants from somatic embryos of their
hybrids. J. Amer. Soc. Hort. Sci. 109:213-218.

Morgan, G. F. and J. C. Bell. 1983. Eucalyptus, p. 423-441. In: S. D.
Tanksley and T. J. Orton (eds.). Isozymes in plant genetics and
breeding, part B. Elsevier, Amsterdam.

Morishima, H. 1969. Phenetic similarity and phylogenetic relationships

among strains of erza pagaaala, estimated by methods of numerical
taxonomy. Evol. 23(3):429-443.

Morrison, D. F. 1976. Multivariate statistical methods. McGraw-HillzNew
York.

Neale, D. B. and W. T. Adams. 1981. Inheritance of isozyme variants in

seed tissues of balsam fir (Abiaa halaamaa). Can. J. Bot. 59:1285-
1291.

Nevo E., D. Zohary, A. Brown and M Haber. 1979. Genetic diversity and
environmental association of wild barley (HQIQQBE aaanaagaam) in
Israel. Evol. 33(3):815-833.

Okabe, S. 1928. On the cytology of the genus Pragaa (in German). Sci.
Rep. Tohoku Imp. Univ. 4(3(:733-743. [Bio. Abstr. 3:4463; 1929].

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

Orton, T. J. 1983. Celery and celeriac (Aalam gzaxaalana L.), p. 351-
367. In: S. D. Tanksley and T. J. Orton (eds.). Isozymes in plant

genetics and breeding, part B. Elsevier, Amsterdam.

Parfitt, D. E., S. Arulsekar, and D. W. Ramming. 1985. Identification
of plum x peach hybrids by isoenzyme analysis. HortSci. 20(2):
246-248.

Pierce, L. C. and J. L. Brewbaker. 1973. Applications of isozyme
analysis in horticultural science. HortSci. 8(1):17-22.

 

81

Pitcher, J. A. 1982. Phenotype selection and half-sib family
performance in black cherry. For. Sci. 28:251-256.

Prance G. T., D. J. Rogers and F. White. 1969. A taximetric study of an

angiosperm family: generic delimitation in the Chrysobalanaceae.
New Phytol. 68:1203-1234.

Prywer, C. 1936. Cytological studies of some species of the genus
{ragga (in Polish). Acta Soc. Bot. Polon. 13:51-83. [Plant Breed.
Abstr. 8:634; 1937-1938].

Quiros, C. F. and K. Morgan. 1981. Peroxidase and leucine-

aminopeptidase in diploid MQQLQQSQ species closely related to
alfalfa: multiple gene loci, multiple allelism, and linkage.
Theor. Applied Genet. 60: 221- 228.

Raptopoulos, T. 1941. Chromosomes and fertility of cherries and their
hybrids. J. Genet. 42:91-113.

Rhodes, A. M., W. P. Bemis, T. W. Whitaker, and S. G. Carmer. 1968. A
numerical taxonomic study of gaaarhiaa. Brittonia 20:251-266.

Rhodes, A. M., S. E. Malo, C. W. Campbell and S. G. Carmer. 1971. A

numerical taxonomic study of the avocado (Earaaa amagiaana Mill.).

J. Amer. Soc. Hort. Sci. 96(3):391-395.

Salinas, J., M. P. de la Vega, and C. Benito. 1982. Identification of

hexaploid wheat cultivars based on isozyme patterns. J. Sci. Food
Agr. 33:221-226.

Sari Gorla, M., C. Frova, G. Binelli and E. Ottaviano. 1986. The
extent of gametophytic-sporophytic gene expression in maize.
Theor. Applied Genet. 72:42-47.

SAS Institute, Inc. 1985a. SAS user's guide: Statistics, version 5 ed.
SAS Institute, Inc., Cary, N.C.

SAS Institute, Inc. 1985b. SAS/GRAPH user's guide, version 5 ed. SAS
Institute, Inc., Cary, N.C.

Scandalios, J. G. 1969. Genetic control of multiple molecular forms of
enzymes in plants; a review. Biochem. Genet. 3:37-79.

Scandalios, J. G. 1974. Isozymes in development and differentiation.
Ann. Rev. Plant Physiol. 25:225-258.

Seal, H. 1964. Multivariate statistical analysis for biologists.
WileyzNew York.

Simmonds, N. W. (ed.). 1976. Evolution of crop plants. Longman, New
York.

hilt-inn

 

F._-__.._- __

82

Small E., P. Y. Jui and L. P. Lefkovitch. 1976. A numerical taxonomic
analysis of Qanaafiia with special reference to species
delimitation. Syst. Bot. 1(1):67-84.

Sneath, P. H. A. 1976. Phenetic taxonomy at the species level and
above. Taxon 254:437-450.

Sneath, P. H. A. and R. R. Sokal. 1973. Numerical taxonomy. Freeman,
San Francisco.

Soltis, D. E., C. H. Haufler, D. C. Darrow, and G. J. Gastony. 1983.
Starch gel electrophoresis of ferns: a compilation of grinding
buffers, gel and electrode buffers, and staining schedules. Amer.
Fern J. 73(1):9-27.

Torres, A. M. 1983. Fruit trees, p. 401-421. In: S. D. Tanksley and T.
J. Orton (eds.). Isozymes in plant genetics and breeding, part B.
Elsevier, Amsterdam.

Torres, A. M. and B. O. Bergh. 1980. Fruit and leaf isozymes as genetic
markers in avocado. J. Amer. Soc. Hort. Sci. 105:614-619.

Torres, A. M., R. K. Soost and T. Mau-Lastovicka. 1982. Citrus
isozymes. J. Hered. 73:335-339.

Vavilov, N. I. 1951. The origin, variation, immunity and breeding of
cultivated plants. Ronald Press, New York.

Weeden, N. F. 1984. Distinguishing among white seeded bean cultivars by
means of allozyme genotypes. Euphytica 33:199-208.

Weeden, N. F. and R. C. Lamb. 1985. Identification of apple cultivars
by isozyme phenotypes. J. Amer. Soc. Hort. Sci. 110(4):509-515.

Wendel, J. F. and C. R. Parks. 1982. Genetic control of isozyme
variation in Qamallia 1aaan1aa L. J. Hered. 73:197-204.

Wijsman, H. J. W. 1983. Petunia, p. 229-252. In: S. D. Tanksley and T.
J. Orton (eds.). Isozymes in plant genetics and breeding, part B.
Elsevier, Amsterdam.

Wilson, H. D. 1976. Genetic control and distribution of leucine
aminopeptidase in the cultivated chenopods (Chenopodium) and
related weed taxa. Biochem. Genet. 14:913-919.

Yang, N. S., J. C. Sorenson, and J. G. Scandalios. 1977. Genetic
control of mitochondrial malate dehydrogenases: Evidence for

duplicated chromosome segments. Proc. Natl. Acad. Sci. USA.
74:310-314. '

Yushev, A. A. 1975. Morphological characters of the leaf in sour cherry
and their use in the classification of varieties (in Russian). -
Byull. Vses. Ordena Lenina 54:34-40. [Plant Breed. Abstr. 47:3633;
1977].

 

83

Yushev, A. A. 1977. Morphological characters of the fruit in sour
cherry and their use in the classification of varieties (in
Russian). Byull. Vses. Ordena Lenina 75:27-31. [Plant Breed.
Abstr. 49:7429; 1979].

APPENDIX

 

 

84

 

mm Hoawooaon popcmp-saoams
we Hosapououofi popawn-oaawuu
m uoawpoaos popcmn-:ummm: A%UHHH903 woummm mo oconv
aum-o .zocouoaucoz.
m Hoawooso: popcon-=aoam:
we uoEHoououon popcon-oamauu
m poawpoaos oopcmn-=umwm= =a< N-mmm huuoso pcsouo
muonass auouuom
aoHumwmuwmm moapcom oakncm mnhuoaou

 

. ..»ocmuoau:oz. you aum-e mo sausages acumen we
econ on» you can «-mmm anyone padouw you :94 mo mawuwoamu oommfiu poem How mucosa: nowumwouwom .H< manna

maeuwoaan momma» town now muonadc couuowouwom H< xHazmmm<

"llll‘lllflllllllll