Unexploited germplasm, natural mutations, and selected in vitro techniques for citrus cultivar improvement

MISSING IMAGE

Material Information

Title:
Unexploited germplasm, natural mutations, and selected in vitro techniques for citrus cultivar improvement
Physical Description:
Unknown
Creator:
Bowman, Kim Dean, 1957-
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 26878457
oclc - 25526648
System ID:
AA00026793:00001


This item is only available as the following downloads:


Full Text












UNEXPLOITED GERMPLASM, NATURAL MUTATIONS,
AND SELECTED IN VITRO TECHNIQUES FOR CITRUS CULTIVAR IMPROVEMENT











By

KIM DEAN BOWMAN













A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1990















ACKNOWLEDGMENTS


Many people have provided assistance, support, encouragement, or

advice during the three years that I have been a student at the University

of Florida and the Citrus Research and Education Center; For this I am

eternally grateful. I especially wish to thank

My wife, Concessa, for love and understanding;

Frederick G. Gmitter, Jr., for financial support, advice, and

encouragement;

Gloria Moore, Jude Grosser, Paul Lyrene, and James Graham for

thoughtful suggestions and ideas;

Terri Zito for excellent photography and kindness;

Mary Ahnger, Bob Sorrell, and Ben Lye for statistical and

computational advice;

Walter Kender, Pamela Russ, Russell Rouseff, J.L. Chandler, and

Margie Wendell for friendship and assistance;

And my parents, Gary and Shirley, for providing me with the

opportunity.











ii















TABLE OF CONTENTS


Rage
ACKNOWLEDGMENTS............................................. ii

KEY TO ABBREVIATIONS... ...................................... v

ABSTRACT..................................................... vi

CHAPTER 1. INTRODUCTION .......................... .......... 1

CHAPTER 2. REVIEW OF THE LITERATURE
Introduction. .............................. 7
Forbidden Fruit............................. 7
Somaclonal Variation ........................ 16
Type of change
-morphology..................... 18
-polyploidy ..................... 20
-aneuploidy ..................... 20
-gross chromosomal............... 21
-simple genetic.................. 21
-organelle DNA .................. 22
-methylation .................... 23
Source of variation
-preexisting..................... 24
-dedifferentiation............... 28
-extended culture................ 29
-hormones in medium............ 30
Mechanisms for induction............... 31
Differences between reports............ 33
Value.................................. 37
Somaclonal Variation in Citrus......... 39
In Vitro Mutagenesis by Irradiation.......... 40
Fruit Sector Chimeras....................... 44
Evaluation of Resistance to Phytophthora..... 49

CHAPTER 3. REDISCOVERY OF CARIBBEAN FORBIDDEN FRUIT
AND EVALUATION OF ITS SIGNIFICANCE
FOR CITRUS BREEDING
Introduction..................... .... ..... 54
Materials and Methods ....................... 55
Results and Discussion ...................... 58


iii











page

CHAPTER 4. EVALUATION OF ORGANOGENIC CITRUS TISSUE
CULTURES AS A SOURCE OF GENETIC VARIATION
FOR CULTIVAR IMPROVEMENT
Introduction................................ 69
Materials and Methods....................... 71
Results and Discussion...................... 79

CHAPTER 5. CITRUS FRUIT SECTOR CHIMERAS AND THEIR
POTENTIAL VALUE AS A GENETIC RESOURCE
Introduction................................ 98
Materials and Methods....................... 99
Results and Discussion...................... 102

CHAPTER 6. RESPONSE OF STEMS FROM IN VITRO-GROWN
SEEDLINGS TO PHYTOPHTHORA PARASITICA
IN DUAL CULTURES
Introduction ............................ 115
Materials and Methods ....................... 117
Results and Discussion...................... 120

CHAPTER 7. SUMMARY AND CONCLUSIONS........................... 132

APPENDIX..................................................... 137

LITERATURE CITED.................................... ......... 169

BIOGRAPHICAL SKETCH.................................. .... 216




















iv















KEY TO ABBREVIATIONS


BA = N-(phenylmethyl)-1 H-purin-6-amine; benzyladenine.

CP1 = Callus proliferation medium.

EME Medium for growth of embryogenic callus.

GM1 = Medium for germination of seeds.

GTF = Grandis-type forbidden fruit.

Gy = Gray (a measure of radiation; 10 Gy = 1 krad).

i.d. = Inside diameter.

MT = Murashige and Tucker Citrus medium (Murashige and
Tucker, 1969).

NAA = 1-naphthaleneacetic acid.

RM1 = Root induction medium.

SCE = Sister chromatid exchange.

SIM = Shoot induction medium.

SPM2 = Fafard No. 2 (Conrad Fafard, Springfield, MA) soilless
potting mix.

SF = Shaddette-type forbidden fruit.

2,4-D = (2,4-dichlorophenoxy)acetic acid.

2,4,5-T = 2-(2,4,5-Trichlorophenoxy)propionic acid.








v















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

UNEXPLOITED GERMPLASM, NATURAL MUTATIONS,
AND SELECTED IN VITRO TECHNIQUES FOR CITRUS CULTIVAR IMPROVEMENT

By

KIM DEAN BOWMAN

December 1990

Chairman: Gloria A. Moore
Cochairman: Frederick G. Gmitter, Jr.
Major Department: Horticultural Science (Fruit Crops)

Breeding of Citrus by sexual hybridization has not been an effective

strategy for cultivar improvement because of long juvenile stages and

widespread apomixis. Investigations of unexploited genetic resources or

systems of evaluation may identify methods to increase the rate of genetic

advance. One potential germplasm resource, two sources of genetic

mutants, and one novel method of in vitro evaluation for disease

resistance were examined during this study. The potential value of each

project was critically reviewed.

A diverse population of Citrus, known as forbidden fruit, was

discovered in Saint Lucia. Historical records, morphological

characteristics, and isozyme evidence indicated that this population was

closely related to grapefruit. High degrees of zygotic embryony and

diversity within the population may make it a useful source of breeding

parents.


vi








Significant morphological variation was observed during initial

growth among Citrus plants regenerated from callus cultures by

organogenesis. Many of the aberrant plants were indistinguishable from

controls after a second cycle of vegetative regrowth. The frequency of

genetic mutation that could be measured as stable morphological change in

young plants was too low to allow efficient mutant recovery.

Fruit sector chimeras were observed among eight Citrus cultivars at

a frequency of 0.01 to 0.27 percent in commercial packinghouses. Mutants

were recovered from some sectored fruit as autotetraploid seedlings.

Several kinds of sectors with potential value as sources of genetic

mutations were described.

In vitro response of etiolated seedling shoots to Phytophthora

parasitica was measured as stem discoloration. The non-host, or

hypersensitive, response observed in Poncirus trifoliata and Severinia

buxifolia was clearly distinguished from the susceptible reaction of eight

Citrus selections and hybrids. However, two Citrus x Poncirus hybrids

with useful levels of field resistance to this pathogen exhibited a highly

susceptible reaction in vitro. This technique may prove useful in the

study of resistance mechanisms, but probably will not be effective for

population screening.












vii














CHAPTER 1

INTRODUCTION


The genus Citrus supplies several fruit crops of great importance

in tropical and subtropical regions of the world. The species of greatest

economic value as scions are Citrus sinensis (L.) Osbeck (sweet oranges),

C. reticulata Blanco (mandarins), C. xparadisi Macfadyen (grapefruits),

C. limon (L.) Burm.f. (lemons), C. aurantifolia (Christm.) Swingle

(limes), C. grandis (L.) Osbeck (pummelos), and a number of interspecific

hybrids, especially tangors (C. reticulata x C. sinensis) and tangelos (C.

xparadisi x C. reticulata). In addition, many other species (e.g. the

sour orange, C. aurantium L.), close relatives (e.g. Poncirus trifoliata

(L.) Raf. and Fortunella margarita (Lour.) Swingle), and hybrids (e.g.

'Carrizo' citrange, P. trifoliata x C. sinensis) are important as

rootstocks or specialty crops (Hodgson, 1967; Morton, 1987; Swingle and

Reece, 1967). This dissertation is focused on the genetic improvements

desired for the world's citrus industries, including simplified cultural

demands, improved fruit quality, increased yield, and reduced

susceptibility to cold, diseases and pests. Achieving significant genetic

improvement is hampered by several major obstacles: a long juvenile phase,

widespread nucellar polyembryony, the heterozygosity of most clones, and

the lack of screens for selection of favorable young plants.


1








2

Citrus spp. are woody perennials with an extended juvenile period

during which no flowering occurs. Citrus species vary considerably in the

length of the juvenile phase. Limes generally require only 2-3 years from

seeding to first fruiting, while sweet oranges may take 10 or more years

before the initiation of flowering (Furr et al., 1947; Sherman and Lyrene,

1983). The long life cycle delays evaluation of hybrid fruiting

characteristics and is a major factor limiting breeding progress. In

addition, most commercially important Citrus cultivars and species are

highly polyembryonic (Frost and Soost, 1968). These selections produce

seeds composed of multiple nucellar (apomictic) embryos. Nucellar embryos

are generally more well developed and vigorous than the single zygotic

embryo that may or may not survive in the polyembryonic clones. As a

result, most seedlings from polyembryonic species are genetically

identical to the maternal parent even when harvested from fruit produced

by controlled pollination. This factor severely limits sexual

hybridization within all Citrus except those species with monoembryonic

selections (C. medica L. [citrons], C. grandis, and part of C.

reticulata). High heterozygosity has been noted in most species of Citrus

(Cameron and Frost, 1968; Frost, 1926; Hagedoorn and Hagedoorn, 1914) and,

when combined with nucellar embryony and extensive vegetative propagation

of desirable genotypes, has resulted in the accumulation of many

deleterious recessive alleles in existing germplasm collections (Cameron

and Frost, 1968). Hybridization among these heterozygous types frequently

results in weak (Cameron and Frost, 1968; Frost, 1943; Torres, 1936)

and/or highly variable progeny (Cameron and Frost, 1968; Swingle, 1913).








3

The limitations placed on sexual hybridization by nucellar embryony,

the unpredictability of the results of successful hybridization, and the

long juvenile phase are problems that are exacerbated by the lack of

methods for evaluation of important characteristics in non-mature

seedlings (Soost and Cameron, 1975). Frequently, it is difficult to

identify the few true zygotic seedlings that may be obtained from a large

population of predominantly apomictic seedlings without waiting for

fruiting or using expensive and time-consuming biochemical techniques

(Bade et al., in press; Scora and Kumamoto, 1983; Soost et al., 1980;

Torres et al., 1978, 1985).

With this perspective, Citrus breeders have directed considerable

attention toward alternatives to sexual hybridization for genetic

improvement. The most notable successes have been in the fields of

somatic hybridization (Grosser and Gmitter, 1989, 1990; Grosser et al.,

1989; Kobayashi and Ohgawara, 1988; Ohgawara et al., 1989; Vardi et al.,

1987, 1989) and somatic mutagenesis (Hearn, 1984, 1986; Hensz, 1971, 1977,

1985; Russo et al., 1981; Spiegel-Roy and Kochba, 1973a; Spiegel-Roy et

al., 1985; Starrantino et al., 1988a, 1988b). More recent interest has

been directed toward the application of genetic engineering to Citrus

genetic improvement. Although Citrus molecular manipulation is still in

its infancy, the success of similar genetic engineering in other plant

species (Deshayes et al., 1985; Gasser and Fraley, 1989; Horsch et al.,

1985; Prols et al., 1989; Sanford, 1988; Saunders et al., 1989; Tomes et

al., 1990; Weising et al., 1988) has encouraged considerable optimism.

Characteristics such as virus (Nelson et al., 1987; Powell Abel et al.,

1986), insect (Fischhoff et al., 1987; Vaeck et al., 1987), and herbicide








4

(Comai et al., 1985; Shah et al., 1986) resistance, as well as others

(Hiatt et al., 1989; Mural et al., 1983; Ow et al., 1986; Smith et al.,

1990; Weising et al., 1988) have been engineered successfully into plants.

The rapid advance in understanding of codon usage (Campbell and Gowri,

1990) and regulatory sequences (Benfey and Chua, 1989), along with the

cloning of unique genes or determination of complete amino acid sequences

for other proteins influencing cold tolerance (Cutler et al., 1989; Lee,

1989), sweetness (Inglett and May, 1968; lyengar et al., 1979; Kennedy et

al., 1988; Ledeboer et al., 1984), insect resistance (Fishman et al.,

1984; Lazarovici et al., 1984; Quicke, 1988), bacterial resistance

(Casteels et al., 1989), and other characteristics (Gasser and Fraley,

1989; Weising et al., 1988) may indicate the much broader applicability

of these techniques. There are already some examples of successful

production of genetically transformed woody species (James et al., 1989;

McGranahan et al., 1988).

Unfortunately, each of these promising techniques appears to have

significant constraints. Somatic hybridization offers many opportunities

for useful genetic combinations, but will probably be limited by the

inability to access desirable traits controlled by recessive genes or to

eliminate dominant negative traits. In vivo somatic mutation of Citrus

has proven effective in the induction of three specific types of useful

genetic changes (i.e. increased pigmentation [Hensz, 1971, 1977, 1985],

decreased acidity [Yen, 1987], and decreased seediness [Hearn, 1984,

1986]) and will probably remain confined in scope. Plant genetic

engineering presents a number of ecological problems and social

complications (Doebley, 1990; Ellstrand and Hoffman, 1990; Hoffman, 1990;








5

Pimentel et al., 1989; Raffa, 1989; Wilson, 1990) that may interfere with

successful deployment of modified plants. In addition, specific molecular

modification of plant genomes will most likely be restricted to

characteristics under simple genetic control for some time to come, and

many of the limits confining conventional breeding (e.g. greater energy

spent on resistance to environmental factors = less energy to put into

yield) will also restrain genetic engineering. With the exception of

induced mutagenesis of mature budwood (in vivo mutagenesis) that avoids
the juvenile phase, none of these methods addresses the often more

significant problem of evaluating the "potentially" improved genotypes

that they produce. Somatic hybridization and genetic transformation

utilize plant regeneration from single cells and therefore revert to the

juvenile stage.

Considerable value may be derived from the development of additional

approaches to genetic improvement of Citrus, both in terms of inducing

genetic variation and evaluating germplasm. This dissertation is devoted

to the investigation of four such areas: the Caribbean forbidden fruit,

somaclonal variation/in vitro mutagenesis, fruit sector chimeras, and in

vitro screening for resistance to Phytophthora. The Caribbean forbidden

fruit is a recently identified, genetically diverse population of

grapefruit-like Citrus that should provide valuable breeding parents with

good fruit quality and a high degree of zygotic embryony. Induction of

variation in tissue cultures (by somaclonal variation or applied

mutagenesis) may provide alternative methods for inducing favorable

somatic mutations in existing cultivars. Fruit sector chimeras are

another source of somatic (as opposed to germ line) mutations that may








6

allow effeltive selection of desirable mutations at the outset of

experimentation, thus minimizing the high frequency of deleterious

mutations commonly associated with induced mutagenesis. Finally,

evaluation of clone resistance to Phytophthora in vitro could supply

important information about resistance mechanisms. If correlated with

field response, such a test could significantly expedite germplasm

evaluation, especially in conjunction with in vitro methods of inducing

genetic variation.















CHAPTER 2

REVIEW OF THE LITERATURE


Introduction

As a background for the dissertation research described herein, a

thoughtful review of the pertinent literature has been conducted in each

of the subject areas. Appropriate contemporary and historical literature

was examined to provide as much relevant information as possible on the

Caribbean forbidden fruit/grapefruit, somaclonal variation, gamma

irradiation of tissue cultures, fruit sector chimeras, and in vitro

evaluation of Phytophthora resistance. Following are concise, but

thorough and well documented summaries of these topics.


Forbidden Fruit

Citrus xparadisi (grapefruit) is the only widely recognized Citrus

species that is considered to have arisen in the western hemisphere.

Thorough searches of the Citrus regions in Asia and the Mediterranean have

revealed no grapefruit, except those introduced from the new world

(Hodgson, 1967; Kumamoto et al., 1987; Scora et al., 1982; Swingle and

Reece, 1967). Grapefruit is now considered to be an apomictically

reproducing hybrid species that originated in the West Indies during the

17th or 18th century. On the basis of historical, morphological, and

biochemical considerations, the most probable parents of grapefruit are


7








8

C. sinensis and C. grandis (Chapot, 1950; Scora and Kumamoto, 1983; Scora

et al., 1982). C. grandis (syn. C. maxima [Burm.] Merrell. and C.

decumana L.) and C. sinensis were present in the West Indies by 1692

(Scora et al., 1982; Sloane, 1696). Grapefruit probably originated

between this time and the 19th century by one or a series of natural

hybridizations between the two species. However, a more complex scenario

involving multiple hybridization events and other species is possible

within the present state of knowledge.

The grapefruits, as they are known in the United States citrus

industry, are a group of morphologically similar and very closely related

selections (Cooper, 1982; Hodgson, 1967; Ziegler and Wolfe, 1961). The

most common cultivars in Florida are 'Marsh' and 'Ruby Red' (Florida Dept.

Agr., 1988a, 1988b), two commercially seedless cultivars that differ

principally in flesh and rind coloration (Hodgson, 1967). 'Ruby Red' was

first identified in 1929 as a limb sport on one tree of 'Thompson' pink

grapefruit by A.E. Henninger in Texas (Hodgson, 1967; Ziegler and Wolfe,

1961). 'Thompson' was itself selected as a limb sport of 'Marsh' in 1913

(Hodgson, 1967; Ziegler and Wolfe, 1961). Many other red-fleshed limb

sports of 'Marsh' or 'Thompson' have been identified, but most were

considered to be essentially identical with 'Ruby Red' (Waibel, 1953;

Ziegler and Wolfe, 1961). Another cultivar that has achieved some

popularity/notoriety is 'Star Ruby', a darker-red fleshed cultivar that

originated from an irradiated seedling (probably mutated nucellar, not

zygotic) of the cultivar 'Hudson' (Hensz, 1971, 1977). 'Hudson' was a

limb sport of the cultivar 'Foster', that was in turn a limb sport of the

cultivar 'Walters' (Hodgson, 1967). Three new darker red-fleshed








9

cultivars ('Ray', 'Henderson' [or a nucellar seedling of 'Henderson' known

as 'Flame' in Florida], and 'Rio Red') that are attracting tremendous

attention in Florida, Texas, and California (Anonymous, 1988, 1989a,

1989b; Fairchild and Teague, 1987; Nauer et al., 1988) were produced as

natural or irradiation-induced bud sports from 'Ruby Red' (Hensz, 1978,

1981, 1985). The remaining commercial grapefruit cultivar is 'Duncan',
a seedy, highly flavored selection that is now declining in significance
and probably originated from a seedling planted in Florida about 1830.

'Duncan' was first introduced and propagated around 1892 (Hodgson, 1967;

Ziegler and Wolfe, 1961). A grove of grapefruit at Safety Harbor,

Florida, that was planted about 1823 by Count Don Phillippe either

directly produced 'Duncan' (Robinson, 1947; Ziegler and Wolfe, 1961), or

was the source of the seed that produced this cultivar (Hodgson, 1967).
This same grove of grapefruit probably also was the source, directly or

indirectly, for the cultivars 'Walters' and 'Marsh' (Cooper, 1982; Ziegler

and Wolfe, 1961).

Count Phillippe's grapefruit grove was planted by importation of

seed or trees from the Bahamas (Robinson, 1952; Anonymous, 1921; Ziegler

and Wolfe, 1961), Cuba (Webber, 1943), or somewhere else in the West

Indies. There are no other reports of grapefruit germplasm importation

into Florida, although Hume (1926) suggested that it may have been brought

to the state by the Spaniards between 1513 and 1821.

Other Citrus cultivars similar to grapefruit have been named
(including 'Oroblanco' [Soost and Cameron, 1980a, 1980b], 'Melogold'

[Soost and Cameron, 1985, 1986], 'Triumph', etc. [Hodgson, 1967]). Many

of these are thought or known to be the result of hybridization between








10

one of the aforementioned grapefruit cultivars and C. sinensis, C.

reticulata, or C. grandis. None of these grapefruit hybrids has been

enough like grapefruit to compete with 'Marsh', 'Ruby Red', or the other

red-fleshed sports for commercial importance (Hodgson, 1967).

Despite the uniqueness and commercial success of the popular

grapefruit cultivars, they are not without their faults, including

irregular fruit size, cold sensitivity, disease susceptibility, and

excessive fruit bitterness. Richard Hensz pointed out that

The improvements attributed to all of the new
grapefruit cultivars that have developed since
the beginning of commercial production have been
towards enhancement for consumer acceptance with
no real improvement in other horticultural
characteristics of the tree or fruit. Presently
grown seedless cultivars are no more vigorous,
disease resistant, productive, nor have better
eating quality than the white seedy types grown
in the early days of the grapefruit industry. In
fact, eliminating the seeds has resulted in the
production of fruit of comparably smaller size
and frequently lower yields.
(Hensz, 1977; p. 582)

The 'Star Ruby' cultivar (developed by R.A. Hensz) has been observed to

be less resistant than less darkly pigmented cultivars to many

environmental stresses, including Phytophthora foot rot, herbicides, and

sunburning (Hensz, 1977). Genetic improvement of grapefruit has been

frustrated by the long juvenile period (Frost, 1943; Furr et al., 1947),

high degree of nucellar embryony (Frost and Soost, 1968), and very limited

variability among the known selections (Hodgson, 1967). Attempts to

overcome these obstacles with existing selections have not been successful

(Furr and Reece, 1946; Snowball et al., 1988; Soost and Cameron, 1975;

Wantanabe et al., 1970).








11
Genetic improvement of grapefruit would benefit by the

identification of grapefruit selections with a shortened juvenile period,

and greater zygotic embryony and genetic diversity. The loss of genetic

diversity that results when population size is reduced to a small number

of individuals has been recognized (Carson, 1970, 1971; Mayr, 1963; Nei

et al., 1975; Prakash, 1972; Wright, 1931). Such a genetic "bottleneck"

may have occurred in the formulation of current populations of North

American grapefruit from the single Phillippe introduction (Bowman and
Gmitter, 1990a). Many of the characters desired for grapefruit genetic

improvement might be obtained from a diverse wild population, but no such

wild grapefruit have been identified or examined as sources of additional

germplasm. Maximum genetic diversity for crop species has frequently been

associated with the center of species origin (Clement, 1989; Harlan, 1971;

Hawkes, 1983; Vavilov, 1951). It is uncertain whether this trend would

apply to species (like grapefruit) that originated as interspecific
hybrids in the relatively recent past (200-300 years). Regardless,

centers of plant diversity not associated with place of origin have been

documented (Harlan, 1975; Peeters, 1988), and any existing wild

populations of grapefruit in the West Indies would seem to be potential

sources of unique alleles and genetic combinations.

The origin of grapefruit has been discussed by several authors on

the basis of the historical literature (Bowman and Gmitter, 1990a, 1990b;
Kumamoto et al., 1987; Robinson, 1952; Scora et al., 1982). The species

name now given to grapefruit, Citrus paradisi, was assigned in 1830 to a

kind of Citrus from Jamaica identified with the common name of "Forbidden

Fruit" (Macfadyen 1830, 1837). At that time, Macfadyen indicated that








12

there were two botanical varieties of this species in Jamaica, one known

as "Forbidden Fruit" and the other known as "Barbadoes Grape Fruit" (p.

304). Subsequent descriptions of Citrus in West Indian literature have

typically used the names grapefruit and forbidden fruit as synonyms

(Fawcett and Rendle, 1920; Freeman and Williams, 1927; Stehle et al.,

1937; Stout, 1982). The most recent report of forbidden fruit as a

distinct selection was from Bermuda (Britton, 1918), where it was

described as a particular small fruited selection of C. grandis.

About the same time as Macfadyen's application of the specific

epithet C. paradisi, a resident of Jamaica, Maycock (1830), stated "My

notes of the Grape-fruit and Forbidden-Fruit Trees, I am sorry to find,

are too imperfect to enable me to say with certainty that they are

specifically distinct, although I am inclined to think they are. I feel

quite certain they are not varieties of C. decumana" (p. 318). Tussac

(1824) also made a possible connection between the two selections when he

described the Jamaican forbidden fruit as "fruit defendu, ou smaller

schaddoc, petit chadec" (p. 74) with fruit borne in clusters like grapes.

Prior to 1830, the only other use of the name grapefruit in the literature

was by John Lunan, who described a variety of Jamaican shaddock "known by

the name of grape-fruit on account of its resemblance in flavour to the

grape" (Lunan, 1814; p. 171). These early descriptions, therefore,

suggest two possible origins for the name "grapefruit," the flavor and/or

the fruit clustering.

Jamaican forbidden fruit was described by several authors before

Macfadyen (Browne, 1756, 1789; Lunan, 1814; Tussac, 1824), although they

did not always agree on morphological characteristics. The first








13

description of forbidden fruit was from Barbados, an island some 1800 km

southeast of Jamaica (Hughes, 1750):

FORBIDDEN-FRUIT-TREE

The trunk, leaves, and flowers of this
Tree, very much refemble thofe of the Orange-
tree.
The Fruit, when ripe, is fomething longer
and larger than the largeft Orange; and exceeds,
in the Delicacy of its Tafte, the Fruit of every
Tree in this or any of our neighbouring Iflands.
It hath fomewhat the Tafte of a Shaddock;
but far exceeds that, as well as the beft Orange,
in its delicious Tafte and Flavour.
This is delineated in Plate VII. (p. 127)

Fourteen other kinds of Citrus found growing in Barbados at that

time were described by Hughes, and one of these, the Guiney orange, he

suggested should be called the "sour forbidden fruit." Hughes also

produced a drawing of the forbidden fruit tree that showed pyriform fruit,

no spines, and leaves without distinctively alate petioles. Hughes'

depiction of the Barbados forbidden fruit tree without spines or winged

petioles is at odds with the description of spines and alate petioles

given by Browne (1756) and Macfadyen (1830, 1837) for the forbidden fruit

of Jamaica. In addition, Macfadyen chose the pyriform fruit shape as a

characteristic that distinguished the "Barbadoes Grape Fruit" from the
"maliformis" Forbidden Fruit. Maycock (1830) indicated that Hughes had

no botanical expertise, and we are left with uncertainty as to whether

Hughes, in Barbados, was describing the same selection known as forbidden

fruit by Macfadyen in Jamaica. Kumamoto et al. (1987) indicated that they

believed the golden orange described by Hughes was actually what later

became known as grapefruit, and that "observers tended to confuse the

grapefruit with the forbidden fruit, believing them to be the same fruit








14

or varieties of the same species" (p. 100). Although some confusion may

have occurred among these early authors, there seems little reason to

question the close relationship between forbidden fruit and grapefruit

agreed upon by Macfadyen (1830, 1837) and Maycock (1830) without

examination of botanical specimens. Unfortunately, good herbarium

specimens (personal communication, R.A. Howard, Harvard) or living

selections of Jamaican and Barbados forbidden fruits do not exist.

Kumamoto et al. (1987) believed "the forbidden fruit tree became extinct,

not having been mentioned after the 19th century, [while] the grapefruit

flourished and spread throughout the West Indies" (p. 100). Both Hughes

(1750), by his forbidden fruit and sour forbidden fruit, and Macfadyen

(1830), by his two botanical varieties of forbidden fruit, indicated that

there was considerable variability within the "forbidden fruit." This

variation may simply be the result of normal intraspecific genetic

diversity, not confusion. Controversy over what constitutes a taxonomic

group within the genus Citrus has been common (Barrett and Rhodes, 1976;

DeCandole, 1824; Green et al., 1986; Hodgson, 1961, Hodgson et al., 1963a,

1963b, 1963c; Scora, 1988; Scora and Kumamoto, 1983; Swingle, 1914, 1943;

T. Tanaka, 1927a, 1935, 1954, 1961, 1977; Y. Tanaka, 1948) and is

obviously fueled by the nature of apomixis and the consequent lack of

traditional species groups in agamic complexes (Crow and Kimura, 1965;

Stebbins, 1950). However, the accepted interspecific origin (C. grandis

x C. sinensis) and monoembryonic ancestry (from C. grandis) of grapefruit

indicates that some genetic differences were probable among seedlings from

identical parentage. This natural genetic diversity may be sufficient to








15

explain the variation in forbidden fruit/grapefruit characteristics among

descriptions by early authors.
Although the grapefruit is generally believed to have originated in

the Caribbean, there are some indications that forbidden fruit (and

possibly grapefruit) may have been (or were derived from) one of the
selections of Citrus previously known in the Mediterranean region as
"apple of Adam" or "Adam's apple." Bonavia (1888) produced figures of a

kind of Citrus fruit being sold in England as "forbidden fruit." He noted
that this was not the same as the pummelo, and indicated that one specimen

had very smooth skin, a solid center, and was "sub-acid and sweet, and
slightly bitter" (Bonavia, 1888; plate XCII), while a second was smaller

and pyriform. Bonavia (1888) also wrote (in reference to the forbidden

fruit) "Gallesio [1811] says the Crusaders found the Pomo d'Adamo in

Palestine, and that it is not the Pompelmoess, the latter being a new
citrus introduced from the East Indies" (plate XCII). Many selections

of Citrus known as "apple of paradise" (Ferrari, 1646; Tolkowsky, 1938)
or "apple of Adam" (Ferrari, 1646; Risso and Poiteau, 1872; Tolkowsky,

1938; Volkamer, 1708-1714) were described in the early citriculture

literature from the Mediterranean, and one or more were probably

introduced into the West Indies (Risso and Poiteau, 1872). Tolkowsky

(1938) cites a Christian pilgrim who described about 1187 the "Adam's

apple" growing in Palestine as "trees which bear fruit called Adam's apple

(= the shaddock) wheron the marks of Adam's teeth may be right plainly
seen" (p. 139). Tolkowsky also indicated that in the Middle Ages the

citron and the shaddock were frequently known as "Adam's apple" or "apple

of paradise," because of a Jewish tradition that considered the citron to








16
be the forbidden fruit described in the religious accounts of the Garden

of Eden (Tolkowsky, 1938).


Somaclonal Variation

In the past several decades, there has been interest in applying

plant tissue culture systems to propagation and genetic improvement of

many crops. Most propagation techniques have made use of axillary or

other preexisting meristems for proliferation of shoots and subsequent

rooting. The propagation of clones in culture by way of preexisting

meristems (mericlones) frequently has not resulted in significantly more

variation than that observed with other methods of vegetative propagation

(Conger, 1981). However, as more sophisticated techniques of plant

culture, manipulation, and regeneration have been developed, it has been

observed that some systems of culture may result in regenerates that are

different in one or more significant characters from the original source.

Larkin and Scowcroft (1981) have applied the term "somaclonal variation"

to all such variability generated by plant cell culture and have suggested

that this variation is of tremendous potential for plant improvement.

Somaclonal variation has been the subject of many investigations and much

speculation over the past decade (Brown and Lorz, 1986; D'Amato, 1985;

DeWald and Moore, 1987; Evans, 1989; Evans et al., 1984; Evans and Sharp,

1986a, 1986b; Lee and Phillips, 1988; Orton, 1984; Reisch, 1983; Wersuhn,

1989). Evans et al. (1989) have been granted a United States patent for

some methods of generating somaclonal variation.

In evaluation of variation among plants regenerated from tissue

culture, it is important to distinguish between transient (epigenetic or








17
physiological) alterations in plant morphology and other more permanent

(epigenetic or genetic) changes in the regenerated plants. However, the

distinction between these two categories blurs somewhat in the cases of

variation in methylation patterns (see below). Temporary changes of

morphology and other characteristics have been commonly observed among

plants regenerated from culture. Tissue culture propagation of

strawberries produced temporary changes in many morphological characters

(Marcotrigiano et al., 1984; Sansavini and Gherardi, 1980; Swartz et al.,

1981). Shoemaker et al. (1985) observed an increase in susceptibility to

root-rotting fungi among tissue culture-propagated strawberries. However,

these plants reverted slowly to the normal resistant phenotype with

increasing time out of culture, and the authors detected no evidence for

genetic changes. The temporary changes in plants regenerated from tissue

culture may rapidly revert to normal (Detrez et al., 1989; Griesbach,

1989), or only after extended growth and/or cycles of vegetative

propagation (Lourens and Martin, 1987). These temporary changes may be

the result of stress or a physiological adaption to the culture

conditions, such as the habituation of tobacco pith tissues (Binns and

Meins, 1973). Generally, the temporary alterations in phenotype are not

of significance for propagation or cultivar improvement, except when they

impair the ability to discriminate true genetic mutants. Most published

reports of somaclonal variation have attempted to eliminate these

transient effects and focus on changes that appear to be genetic or at

least relatively long term.

Sugarcane was one of the first crops to obtain significant attention

because of the clonal variation observed among regenerated plants.








18

Selections were recovered from tissue culture with increased yield and

sucrose production (Heinz and Mee, 1971), as well as resistance to eyespot

disease (Heinz et al., 1977), Fiji disease, and downy mildew

(Krishnamurthi and Tlaskal, 1974). Subsequent selection of somaclones

from cultures of major commercial varieties has been reported to yield

variants with favorable changes in some morphological or horticultural

characteristic (Heinz et al., 1977; Krishnamurthi, 1977; Larkin and

Scowcroft, 1983; Liu and Chen, 1978; Maretzki, 1987). Many of these

sugarcane clones have been tested over several years with no apparent loss

of the acquired character (Orton, 1984), although other characters have

exhibited some phenotypic instability (Maretzki, 1987).

Similar success with somaclonal variation was also reported for

potato, which is a polyploid, vegetatively propagated crop like sugarcane.

Shepard et al. (1980) were able to regenerate plants with changes in tuber

skin color, tuber uniformity, maturity date, and numerous other characters

from leaf protoplasts of a popular potato variety, 'Russet Burbank'. Some

of these variants were reported to be horticultural improvements over the

original clone. Plants also were obtained with increased resistance to

early blight and late blight (Phytophthora infestans). Vegetative

propagation of these variants over several generations did not degrade the

acquired characteristics.

Type of change morphology. Changes in many types of morphological

or horticultural characteristics have been reported in somaclones,

including leaf shape (Burk and Matzinger, 1976; Freytag et al., 1989;

Osifo et al., 1989), foliage color (Osifo et al., 1989; Taliaferro et al.,

1989), fertility (Daub and Jenns, 1989; Griesbach 1989; McPheeters and








19
Skirvin, 1989; Orton, 1984; Taliaferro et al., 1989), plant size (Eapen

et al., 1989; Griesbach 1989; Jain et al., 1989; Taliaferro et al., 1989),

growth habit (Freytag et al., 1989), seed color (George and Rao, 1983),
time of flowering (Burk and Matzinger, 1976; Eapen et al., 1989; Ozias-

Akins, 1989), fruit number per plant (Eapen et al., 1989), cold resistance

(Galiba and Sutka, 1989; Lazar et al., 1988), insect resistance (Miles et

al., 1981), disease resistance (Behnke, 1979, 1980; Daub and Jenns, 1989;

Gengenbach et al., 1977; Hartman et al., 1984; Heath-Pagliuso et al.,

1988; Heinz et al., 1977; Latunde-Dada and Lucas, 1983; Orton, 1984;

Sacristan, 1982; Shahin and Spivey, 1986; Thanutong et al., 1983; Toyoda

et al., 1989), resistance to other factors (Grandbastien et al., 1989;

Kuehnle and Earle, 1989; McHughen, 1987; McHughen and Swartz, 1984; Ranch

et al., 1983; Schaeffer et al., 1989; Shahin and Spivey, 1986; Wakasa and

Widholm, 1987), alkaloid content (Burk and Matzinger, 1976), essential oil

composition (Mathur et al., 1988), nutritive quality (Schaeffer and

Sharpe, 1987; Schaeffer et al., 1989), and yield (Burk and Matzinger,

1976; Eapen et al., 1989; Jain et al., 1989; Mathur et al., 1988; Secor

and Shepard, 1981). Many reports of somaclonal variation do not indicate

the genetic basis of the observed changes, although some alterations of

the genetic material have been documented among somaclones. Genetic

changes may be demonstrated by inheritance of the altered character among

the sexual progeny, as was shown for somaclones of oat (Cummings et al.,

1976), barley (Breiman et al., 1987b), and tomato (Evans and Sharp, 1986a,

1986b). In other cases, cytological, biochemical, or molecular studies

have revealed a more specific genetic basis for the somaclonal variation.








20

Type of change polyploidy. Euploid increases in chromosome number
among plants regenerated from tissue culture have been reported for many

crops (D'Amato, 1977a, 1978), including bluestem grasses (Taliaferro et

al., 1989), tobacco (Ogura, 1976), tall fescue (Eizenga, 1989), alfalfa

(Saunders and Bingham, 1972) and wild rye (Park and Walton, 1989). Osifo

et al. (1989) produced callus from the cotyledons of a diploid potato

selection and reported that 70% of the regenerated somaclones were

tetraploid. Hexaploid and octoploid plants were found among the

regenerates from protoplasts of two tetraploid potato cultivars (Karp et

al., 1982), while Singh et al. (1972) obtained haploid plants from

cultures of a diploid legume species. However, examination of a large

number of plants regenerated from haploid, diploid, and triploid lines of

beet detected none with alterations in the original chromosome number

(Detrez et al., 1989). Interspecific hybrids have been noted to yield

polyploids after tissue culture (D'Amato, 1985); Cappadocia and Ramulu
(1980) obtained 21 amphidiploids among 57 regenerated plants from the

hybrid between two species of tomato.

Type of change aneuploidy. Aneuploidy has been reported most

frequently in somaclones from species that are polyploid (D'Amato, 1985);

diploids are much more likely to become inviable as a result of loss or

gain of one or a few chromosomes. Aneuploid somaclones have been

identified in celery (Murata and Orton, 1983; Orton, 1983), alfalfa
(Groose and Bingham, 1984; Johnson et al., 1984), tobacco (Ogura, 1976),

potato (Karp et al., 1982), Haworthia (Ogihara, 1981) and many grasses

(Ahoowalia, 1983; Eizenga, 1989; Reed and Conger, 1985; Taliaferro et al.,

1989). Eizenga (1989) observed 59 aneuploids among the 166 tall fescue








21
somaclones she examined. Meiotic analysis of the tall fescue somaclones

indicated that the aneuploids were monosomic, double monosomic, or triple

monosomic but not nullisomic (Eizenga, 1989).

Type of change gross chromosomal. Alterations in gross
chromosomal structure, or macromutations, have been found among plants

regenerated from tissue culture, as well as among in vivo sports

(Peloquin, 1981). Chromosome translocations, inversions, duplications

and/or deletions have been reported in somaclonal variants from oat (McCoy

et al., 1982), ryegrass (Ahloowalia, 1978), potato (Shepard, 1982), maize

(Benzion and Phillips, 1988; Lee and Phillips, 1987) and other species

(Armstrong et al., 1983; Lapitan et al., 1984; Lee and Phillips, 1988).

McCoy et al. (1978) noted the formation of tripolar divisions, ring

chromosomes, and heteromorphic bivalents at meiosis of some plants from

culture and considered non-homologous crossovers and deletions to be
responsible for these abnormalities. Benzion and Phillips (1988)

reported that over 12% of the somaclones regenerated from maize callus

cultures during a 22 month period contained some type of cytological

abnormality, and many of these were caused by the breakage of a chromosome

between the centromere and a heterochromatic knob.

Type of change simple genetic. Changes in one or a small number
of genes have been identified in somaclones by segregation among progeny
from controlled crosses in wheat (Larkin et al., 1984), maize (Edallo et

al., 1981), rice (Sun et al., 1983; Fukui, 1983), soybean (Freytag et al.,

1989), tobacco (Prat, 1983), lettuce (Engler and Grogan, 1984), and twelve

different variants of tomato (Evans and Sharp, 1983). Shahin and Spivey

(1986) produced protoplasts from a tomato cultivar susceptible to Fusarium








22

oxysporur f.sp. lycopersici, race 2, and reported the regeneration of

plants resistant to the pathogen both with and without in vitro selection

pressure. Segregation patterns among progeny of these tomato somaclones

indicated that some of the somaclones were heterozygous and some

homozygous for a single dominant resistance gene (Shahin and Spivey,

1986). Relatively minor genetic mutations in some somaclones have also

been suggested by changes in isozyme banding patterns (Allichio et al.,

1987; Jackson and Dale, 1989; Orton, 1983; Taliaferro et al., 1989) and

restriction fragment length polymorphisms (Roth et al., 1989). One maize

somaclone with a variant Adhl allele was shown to have resulted from a

single base substitution in the nuclear gene (Brettell et al., 1986a).

Some cases of chromosomal mutations leading to expression of recessive

genes may result from the genetic nullification of an obscuring dominant

gene, rather than from the de novo generation of the recessive gene

(Little, 1989).

Breiman et al. (1987a) reported changes in the number of rDNA

spacers in somaclones from one line of wheat after only a few weeks of

culture, and claimed that the studies of Appels and Dvorak (1982) and

Saghai-Maroof et al. (1984) indicated equivalent changes should take many

generations in vivo. Other workers have reported deficiencies in rRNA

genes among somaclones from triticale (Brettell et al., 1986b) and potato

(Landsmann and Uhrig, 1985).

Type of change organelle DNA. Changes in mitochondrial DNA

(mtDNA) have been observed in somaclones from maize (Gengenbach et al.,

1981; Umbeck and Gengenbach, 1983), sugar beet (Brears et al., 1989),

potato (Kemble and Shepard, 1984), and wheat (Hartmann et al., 1987, 1989;








23

Rode et al., 1987). Somaclones were recovered from Texas male sterile

lines of maize (susceptible to Helminthosporium maydis race T) with male

fertility or resistance to the toxin of H. maydis (both characters are

controlled by mitochondrial gene[s]) (Brettell et al., 1980; Umbeck and

Gengenbach, 1983). Brears et al. (1989) reported a mtDNA change in a

sugar beet somaclone. The mtDNA change was a specific reversion of the

RFLP pattern in the male sterile source genotype to the pattern of the

normal (fertile) genotype, although the somaclone remained male sterile

(Brears et al., 1989). Hartmann et al. (1989) examined the mitochondrial

genome of wheat somaclones and also observed specific rearrangements to

other known mitochondrial types, as well as some novel mitochondrial DNA

patterns.

No rearrangements were observed in the chloroplast genome of sugar

beet somaclones (Brears et al., 1986, 1989), and it generally has been

considered that the chloroplast genome rarely undergoes rearrangement.

However, Evans and Sharp (1986) reported identification of changes in

chloroplast DNA among some regenerated tomato plants. In gametoclonal

studies, regeneration of plants from anther culture was reported to yield

selections with deletions from the chloroplast DNA (Day and Ellis, 1984).

Type of change methylation. Alterations in the methylation of DNA

have been reported among cultured plant cells (Anderson et al., 1990;

Quemada et al., 1987). The level of DNA methylation has been suggested

to influence or indicate the expression of some genes, with reduced

methylation generally corresponding to greater expression (Amasino et al.,

1984; Bird, 1986; Cedar, 1988; Flavell et al., 1986; Hepburn et al., 1983;

Holliday, 1987, 1989; Jablonka and Lamb, 1989; Watson et al., 1987). The








24

observation that treatment of plant tissues with a demethylating agent,

5-azacytidine, increases expression of some genes (Amasino et al., 1984;

Hepburn et al., 1983) has shown that reduced methylation may promote gene

expression. However, the relative degrees of methylation for rRNA genes

in shoots, roots, and callus cultures of petunia differed widely and were

not correlated with relative gene expression or growth (Anderson et al.,

1990). The fact that higher plants contain many copies of rRNA genes, and

only a small portion of them are expressed at one time (Rogers and

Bendich, 1987) may make methylation of rDNA a poor indicator of total

relative gene expression. Quemada et al. (1987) suggested that changes

in methylation may be a cause of habituation in cell cultures.

Nevertheless, changes in gene methylation have been proposed as a possible

source of morphological variation in plants regenerated from tissue

culture (Brown and Lorz, 1986; Lee and Phillips, 1988; Quemada et al.,

1987). Patterns of methylation have been suggested to be heritable

(Holliday, 1989; Jablonka and Lamb, 1989), but it is unclear whether

methylation is a determiner or a consequence of gene activity. Generally,

ribosomal genes in petunia shoots were highly methylated, those in callus

cultures were variable in the degree of methylation (although the level

of methylation for each cell line remained stable over time), and

regenerated shoots typically returned to the degree of methylation

observed in the original explants (Anderson et al., 1990).

Source of variation preexisting. Evidence for morphological and

genetic variation among plants regenerated from tissue culture is

abundant. However, at least a portion of this variation is derived

directly, without change, from the genetic and heritable epigenetic








25
differences (methylation?) that are common among cells within a single

plant (D'Amato, 1975, 1977a). Cells in meristems generally do not change

in genetic composition over long periods of time (D'Amato, 1985), as

demonstrated by the large number of species that have been propagated

uniformly by meristem culture (Conger, 1981). However, many different

tissues can be successfully used for culture initiation (Murashige, 1974),

including differentiated tissue that does not have a high degree of

genetic fidelity. Explants from non-meristematic parts of the plant have

been considered most likely to yield variation in somaclones (Meins, 1983;

Reisch, 1983). Nearly 90 percent of the flowering plant species studied

have been found to exhibit some level of endoreduplication of the DNA

within the cells from differentiated tissues (D'Amato, 1977b, 1985). This

results in an elevated level of ploidy that may frequently vary over a

wide range within a single tissue (2x, 4x, 8x, or even higher). Plants

with differentiated tissues that are composed of a mixture of cells with

different ploidy levels are called polysomatic (D'Amato, 1985).

Replication of chromosomes without separation of the chromatids

(endomitosis) may occur to form polytene chromosomes, as in the cotyledons

of pea (Karp and Bright, 1985). Other species, such as Crepis, have been

found to maintain the same level of ploidy throughout their meristematic

and somatic tissues (Brossard, 1978).

Variation in chromosome numbers other than in genome sets
(aneuploidy) is not common in vivo but has been observed in the meristems

and throughout the tissues of several plant species (D'Amato, 1985).

Vaarama (1949) studied the tremendous range in aneuploidy found in a








26

tetraploid black currant (4x=32). Within the root tips, cells could be

identified containing from 4 to 32 chromosomes.

Sub-chromosomal nuclear and organellar variation also occurs in

plant tissues undergoing differentiation (D'Amato, 1952, 1985), although

it is frequently difficult to identify. Examples are generally limited

to those that have a visual effect, such as the white and green leaf

sectors that result from different chloroplast types (Poethig and Sussex,

1985). The color of stamen hairs in Tradescantia is controlled by a

single gene that has been observed to mutate from heterozygous to

homozygous recessive at a frequency greater than 0.04% under normal

conditions (Dolezel and Novak, 1984). Some cases of variation in plants

regenerated from culture have been directly related to the isolation of

one or more layers from the chimeric source plant (Hall et al., 1986a,

1986b, 1986c; McPheeters and Skirvin, 1989; Skeene and Barlass, 1983).

Chromosomal crossing-over in somatic tissues has been observed among

several crop species (Evans and Paddock, 1976) and is a mechanism (as is

sister chromatid exchange [see below]) by which genetic variation can be

produced among the cells within an individual. Much of the variation in

plants regenerated from pollen or cells undergoing meiosis (sometimes

called gametoclonal variation) probably results largely from the

segregation of chromosomes and organelles in the meiotic processes and is

best considered as a different phenomenon than somaclonal variation,

although the two have frequently been discussed together (Evans and Sharp,

1986a; Evans et al., 1984).

An indication of the significance of explant source tissues to
variation obtained in culture can be found by comparing cell populations








27
or regenerates obtained from different tissues. Wardell and Skoog (1973)

reported tobacco pith cells varied in their DNA content, depending upon

the height in the plant where the samples were taken. Plants regenerated

from the cells lower in the stem had a different morphology from those

regenerated from cells collected at the higher levels. Osifo et al.

(1989) observed a much greater frequency of polyploidy among potato

somaclones regenerated from cotyledon-derived callus than those

regenerated from leaf-derived callus. Feher et al. (1989) reported that

the chromosome number of tetraploid alfalfa somaclones regenerated by

embryogenesis was directly related to the chromosome number of the

original explant. Explants composed of somatic cells with 30 chromosomes

produced only somaclones with 30 chromosomes. Aneuploid explants with 29

chromosomes and euploids with 2n-32 resulted in somaclones with a

diversity of chromosome numbers, including a high percentage of mixoploids

(Feher et al., 1989). However, in some cases, even the presence of

genetic diversity within explant tissues does not lead to genetic

variation among regenerated plants. Polyploid cells are common in sugar

beet leaf tissue, but no polyploid, aneuploid or chimeric plants were

observed among a large number of somaclones regenerated via adventitious

shoots from petiole tissue (Detrez et al., 1989). The authors suggested

that this was the result of selection against polyploid cells during the

formation of in vitro adventitious buds.

In some cases, the tissue culture process seems to select for

genotypes that are rare in the explant tissue. Ramulu et al. (1989)

examined the ploidy levels of somaclones regenerated from protoplasts of

monohaploid, dihaploid, and diploid potato selections. No haploid and








28
only a low frequency of diploid somaclones were recovered from any of

these sources. Most of the somaclones from all three ploidy sources were

tetraploid (Ramulu et al., 1989). Haploid cultures of Crepis (Sacristan,

1971) and Datura (Furner et al., 1978) became diploid during extended

periods of culture.

Source of variation dedifferentiation. At least a part of the

variation observed among plants obtained from tissue culture may be

produced or stimulated by the tissue culture process itself. One

indicator of this is the changes in chromosome number that have been

reported to occur in plant cell cultures. Studies by Binarova and Dolezel

(1988) indicated that there was a slight increase in the frequency of

cells at higher ploidy during initiation of embryogenic alfalfa suspension

cultures, but frequencies reverted to normal after a short period in

culture. Similar types of changes during culture initiation (and

dedifferentiation) have been reported for alfalfa (Feher et al., 1989),

tobacco (Bennici and Caffaro, 1985), onion (Dolezel and Novak, 1985), and

pine (Franklin et al., 1989). Furner et al. (1978) observed that

polyploidy was produced by the stimulation of premitotic endoreduplication

during the dedifferentiation process. Using explants with genetic

markers, Barbier and Dulieu (1983) reported that most of the culture-

induced changes occurred during the first few cell divisions. Cionini et

al. (1978) studied the callus induction phase in broad bean, and observed

that the DNA composition of cells sometimes underwent an irregular

division process, called fragmentation. When mitosis and cytokinesis

followed, the resulting cells exhibited various degrees of aneuploidy.

Selective amplification of certain DNA sequences has been reported in one








29

tobacco species during initial phases of dedifferentiation (Durante et

al., 1983). Such differential DNA replication may affect fragmentation,

mutation, and gene expression.

Source of variation extended culture. In addition to those

changes induced during dedifferentiation, other variation may occur slowly

over extended periods of culture (Barbier and Dulieu, 1983). Chromosome

polyploidization may arise by endoreduplication (D'Amato, 1985), fusion

of spindles in cells with two nuclei (Mitra and Steward, 1961), or spindle

failure (Bayliss, 1973). An increase in the frequency of cells at higher

ploidy levels commonly has been reported for long term cell cultures

(Bayliss, 1980; Berlyn et al., 1986). A steady increase in ploidy level

was observed for cell cultures of Coulter pine, so that six weeks after

culture initiation about 80% of the cells had 4C DNA (diploid cells in G2,

or tetraploid cells), and regenerated buds contained many cells at the 8C

level (Patel and Berlyn, 1982). However, other studies with pine cultures

have indicated relatively little change in cell ploidy over time (Franklin

et al., 1989; Konar and Nagmani, 1972; Renfroe and Berlyn, 1984; Salmia,

1975), and there was no abnormality in the amount of DNA found in

adventitious shoots regenerated from 3-6 month old spruce cultures (Hakman

et al., 1984). Degree of aneuploidy may also increase in cultures with

age. Murashige and Nakano (1965) noted the higher frequency of aneuploidy

found in tobacco cultures as they grew older. Wheat somaclones obtained

from long term cultures were reported to be more likely to have

mitochondrial DNA rearrangements than plants obtained from short term

cultures (Hartmann et al., 1989).








30

Some of the diversity in frequencies of somaclonal variation

reported by researchers working on the same crop may be the result of

specific differences in culture technique. Alfalfa plants regenerated

from callus cultures were reported to be morphologically and

karyologically normal by some workers (Binarova and Dolezel, 1988; Kao and

Michayluk, 1980; Mezentsev, 1981), while other reports indicated

considerable morphological and karyological deviance among somaclones

(Groose and Bingham, 1984; Johnson et al., 1984; Nagarajan and Walton,

1987). Studies by Binarova and Dolezel (1988) indicated that if short

subculture intervals were used for alfalfa suspension cultures, the normal

ploidy level and the capacity for embryogenesis were maintained. When the

period between subcultures was extended to greater than seven days, there

was an increase in the frequency of polyploid cells in the cultures and

a corresponding loss in embryogenic potential (Binarova and Dolezel,

1988). This may be related to the differentiation process that begins

once cell cultures reach the stationary growth phase. Cell

differentiation frequently includes cell wall thickening, vacuolization,

and DNA endoreduplication (Kibler and Neumann, 1980).

Source of variation hormones in medium. Hormone composition of

the media was reported to influence the degree of endoreduplication in pea

cortex cells during dedifferentiation (Libbenga and Torrey, 1973). Auxins

like 2,4-D have been suggested to have an important effect on the

induction and frequency of somaclonal variation, including ploidy level

stability (Bayliss, 1980). Relatively high concentrations (at least 25

mg-1-') of 2,4-D and 2,4,5-T (another auxin) were found to induce

cytological abnormalities in onion root tips (Croker, 1953). Lazar et al.








31

(1983) observed that tissue culture medium with moderate levels of 2,4,5-T

(1 mgl-') resulted in a higher frequency of aberrant plants than the same

medium with a low 2,4,5-T concentration. However, Binarova and Dolezel

(1988) noted considerable ploidy stability in alfalfa suspension cultures

initiated on medium with a high level of 2,4-D, and other researchers have

found that growth regulator components of the media had no effect on the

karyological stability of cultures (Hanisch Ten Cate and Sree Ramulu,

1987). The relative ability of a chemical to induce sister chromatid

exchanges (SCE) is considered to be a good indicator of mutagenic action

(Latt, 1974; Perry and Evans, 1975; Uggla and Natarajan, 1982). Murata

(1985, 1989) made use of this fact to examine the potential mutagenic

effects of NAA, 2,4-D, 2,4,5,-T, and kinetin on wheat cell cultures.

Concentrations of 2-5 mgl-' of 2,4,5-T significantly increased the

frequency of SCE in the wheat cell cultures. However, NAA, 2,4-D, and

kinetin were not observed to have any effect on the frequency of SCE,

although inclusion of kinetin reduced the effects of 2,4,5-T.

Mechanisms for induction. One mechanism of somaclonal variation may

be the formation of SCEs or some other types of mitotic recombination.

Larkin and Scowcroft (1981) suggested that asymmetric crossing-over or

recombination between homologous chromosomes could lead to some of the

macromutations that had been observed in somaclones. Limited evidence for

the occurrence of mitotic recombination among plants regenerated from

tissue culture has been presented by several authors (Barbier and Dulieu,

1983; Dulieu and Barbier, 1982; Larkin et al., 1984; Lorz and Scowcroft,

1983; McCoy, 1980).








32

Chromosome fragmentation among plant cell cultures and somaclones

has been observed to commonly occur in heterochromatic regions (Armstrong

et al., 1983; Brettell et al., 1986; Lapitan et al., 1984; McCoy et al.,

1982; Murata and Orton, 1984) or near the nucleolus organizer region (Lee

and Phillips, 1987; Sacristan, 1971). Heterochromatin is known to

replicate late in the mitotic cycle (Lima-De-Faria, 1969), so the

formation of a bridge in the heterochromatic region and subsequent

chromosome breakage has been proposed as one mechanism of somaclonal

variation (Benzion and Phillips, 1988; Johnson et al., 1987; Lee and

Phillips, 1987, 1988; McCoy et al., 1982; Sacristan, 1971).

Nucleotide pool imbalances may produce genetic changes in

unicellular organisms (Kunz, 1982; Kunz and Haynes, 1982) and animals

(Weinberg et al., 1981), including SCEs, aneuploidy, and macromutations

in the nuclear and organellar genomes. Lee and Phillips (1988) have

proposed that cell culture may cause critical imbalances in the nucleotide

pools of plant cells that lead to similar genetic changes, and thus

somaclonal variation.

Another mechanism that has been suggested for the induction of

somaclonal variation is the activation of transposable elements by

stresses on the plant genome resulting from the tissue culture process

(Chaleff, 1983; Evola et al., 1985; Groose and Bingham, 1986a, 1986b;

Larkin and Scowcroft, 1981; Peschke et al., 1987) or some other specific

stress-induced rearrangement process (Benzion et al., 1986; Brettell et

al., 1986b; Murata and Orton, 1983). Roth et al. (1989) suggested that

the RFLP changes observed in inbred soybean suspension cultures may have

been caused by a mechanism that evolved to produce genetic variation in








33
response to stress. The changes that were observed at all loci were

almost entirely alternative alleles found in other soybean cultivars. It

was suggested that this may indicate that a specific recombinational or

rearrangement event occurred, such as gene conversion, transposon

movement, or inversion. Equivalent programmed gene rearrangements have

been documented in mammals and some unicellular organisms (Borst and

Greaves, 1987). Breiman et al. (1987a) proposed that the alterations they

observed in intergenic spacers of regenerated somaclones were the result

of tissue culture stress disrupting the mechanisms that normally maintain

the stability of repeated DNA sequences. It is thought that repeated

sequences play an important role in the adaption of the genome to change

(Flavell, 1985). Peschke et al. (1987) initiated cultures from maize

explants without transposable element activity and reported the recovery

of active Ac transposable elements in regenerated plants. Transposable

elements have been best studied in maize, but are found in many other

plant species as well (Grandbastien et al., 1989; Groose and Bingham,

1986a, 1986b). Maize transposable elements can be activated by different

types of shocks or stresses to the genome (McClintock, 1978, 1984), such

as chromosome breakage (Freeling, 1984), chromosome rearrangements

(Freeling, 1984; McClintock, 1978; Peterson, 1986), and alterations in DNA

methylation (Chandler and Walbot, 1986). All three of these processes

occur in plant tissue cultures (see above).

Differences between reports. Frequency of variants obtained from

culture has varied widely in the different cases where it has been

examined. Daylily has been observed to produce little somaclonal

variation (Krikorian et al., 1981), and Griesbach (1989) was able to








34

identify only one stable variant among 1000 daylily somaclones regenerated

by organogenesis from long-term callus cultures. In contrast, Eapen et

al. (1989) reported alterations in mustard plant height, pod number, seed

weight per plant, and yield in the second generation of nearly all plants

regenerated from protoplasts when compared with the parental genotype.

Frequently different researchers report entirely opposite results for the

same crop and similar methods. Secor and Shepard (1981) examined 65

protoplast-derived somaclones of potato and determined that every one was

statistically different from the source cultivar in at least one of the

35 characteristics measured (the authors indicated that this observation

was quite significant, but the statistical significance is questionable),

while a previous study (Wenzel et al., 1979) reported that there was no

variation among protoplast-derived potato plants. Lourens and Martin

(1987) observed that morphological variability among sugarcane somaclones

was much less than previously reported (Heinz et al., 1977; Heinz and Mee,

1971; Heinz et al., 1969) when deviant somaclones were vegetatively

propagated and examined after a second cycle of growth. Some of the

discrepancies within the literature may be caused by confusion between

temporary epigenetic changes and permanent genetic mutations. In

addition, kinds and amount of variation are dependent upon the characters

measured (Daub and Jenns, 1989). Three somaclonal variants for gliadin

proteins in wheat (Maddock et al., 1985) were not observed to be deviant

when examined for a variety of morphological characteristics (Maddock and

Semple, 1986).

In other cases, dissimilar results may be associated with large

differences between species, or between cultivars within a species, in








35
response to tissue culture (Bright and Jones, 1985) and in the frequency

of somaclonal variation (Breiman et al., 1987; Brettell et al., 1986; Karp

and Bright, 1985). Maddock and Semple (1986) speculated that the much
lower frequency of somaclonal variation they observed (Maddock et al.,

1985; Maddock and Semple, 1986) than Larkin et al. (1984) among

regenerated wheat plants was caused, at least partly, by cultivar

differences. Binarova and Dolezel (1988) reported a high degree of

karyological stability and maintenance of embryogenic capacity for one

line of alfalfa in suspension culture, while Atanasov and Brown (1984)

observed a considerable shift in cell ploidy and loss of embryogenesis in
a second line maintained under similar conditions.

Method of regeneration may also affect the frequency of variation

among somaclones. Regeneration by adventitious shoots usually has

produced a greater frequency of somaclonal variants than regeneration from

axillary meristems (Conger, 1981; Larkin and Scowcroft, 1981).

Organogenic regeneration generally has been considered to result in more

somaclonal variation than embryogenesis (Armstrong and Phillips, 1988;

Vasil, 1987), although this has not always proven true. Regeneration by

organogenesis from long-term callus cultures resulted in very little

variation in daylily (Griesbach, 1989), while regeneration of mustard

plants by direct somatic embryogenesis was reported to produce over 95%

somaclonal variants (Eapen et al., 1989). A study comparing the variation

among maize plants regenerated by the two methods indicated that the

frequency of variation was great for plants from embryogenesis, but even

greater for those from organogenesis (Armstrong and Phillips, 1988). The

process of plant regeneration may apply some selection against genetically








36

aberrant cell types that survive in culture but are not viable in vivo

(D'Amato, 1985; Vasil, 1988). Feher et al. (1989) observed that

embryogenesis occurred normally in alfalfa cultures containing high

percentages of aneuploid and/or polyploid cells, but that the resulting

embryos were mostly of the normal chromosome number. Polyploid cells are

common in sugar beet leaf tissue, but no polyploid, aneuploid or chimeric

plants were observed among a large number of somaclones regenerated by

adventitious shoots from petiole tissue (Detrez et al., 1989); the authors

suggested that this was the result of selection against polyploid cells

during the formation of in vitro adventitious buds.

As mentioned above, culture age may also affect variation among

regenerated plants. The frequency of somaclonal variation typically has

been observed to increase with aging of the callus from which plants are

regenerated (Armstrong and Phillips, 1988; Lee and Phillips, 1987). No

meiotic aberrations were observed among maize somaclones regenerated from

cultures 3-4 months old, but nearly half of the somaclones regenerated

from similar cultures 8-9 months old contained cytological aberrations

(Lee and Phillips, 1987). Benzion and Phillips (1988) presented evidence

that the increased frequency of cytogenetically abnormal maize somaclones

from older cultures was "due to mutational events that occurred [at a

constant rate] throughout culture development with a subsequent

maintenance and accumulation of aberrant cells over time" (p. 318).

However, callus age has not always resulted in a high frequency of

variants; daylily callus more than 20 months old produced few somaclonal

variants (Griesbach, 1989).








37

Value. Despite numerous reports of interesting or potentially

useful mutants obtained by somaclonal variation, relatively little benefit

has been evidenced in commercial varieties or breeding lines. One problem

with some clones derived by tissue cultures, especially those derived by

in vitro selection, is that what appear to be favorable changes in cell

characteristics or even greenhouse plant response may not translate into

improved field performance. A salt-tolerant somaclone of flax obtained

by in vitro selection (McHughen and Swartz, 1984) was shown to be more

productive than its parent cultivar under greenhouse conditions (McHughen,

1987). However, extensive field testing at both saline and nonsaline

sites did not indicate that the salt-tolerant line was more productive

than the parent cultivar under either condition (Rowland et al., 1988,

1989). In the case of sugarcane, it has been suggested that the

tremendous amount of variation already present in the germplasm renders

the minor variation obtained from culture relatively insignificant

(Maretzki, 1987). Another limitation is that mutations produced by

somaclonal variation are random and, like all types of random mutations,

are almost always deleterious or undesirable (Micke et al., 1987; Stadler,

1928a, 1928b).

The novelty of the mutations obtained determines the relative value

of any mutagenic method, including somaclonal variation. Some novel

mutations have been reported among somaclones (Evans, 1989; Evans et al.,

1989; Griesbach, 1989). However, the types of variation in potato

somaclones (Sanford et al., 1984), tobacco somaclones (Daub and Jenns,

1989) and wheat somaclones (Maddock and Semple, 1986) were not found to

be different from those that occurred naturally under field conditions or








38
were already known in germplasm collections. Sanford et al. (1984)

pointed out that sub-clonal selection in the field had been used for

potato improvement (Davidson and Lawley, 1953; Easton and Nagle, 1981;

McIntosh, 1945) but was generally proven less effective than sexual

hybridization. Gavazzi et al. (1987) compared tomato mutants derived from

somaclonal variation with those produced by ethyl methane sulphonate (EMS)

treatment in vivo, and determined that somaclonal variation produced a

higher frequency of variants. In addition, the types of mutations

produced by somaclonal variation and EMS methods were different;

Verticillium resistance was only recovered from EMS treatments (Gavazzi

et al., 1987). Montagno et al. (1989) compared somaclonal variation in

tomato with that induced by mutation with ionizing radiation. The two

methods produced similar types of mutations, although radiation (at 6.5

krad) produced heritable changes at more than twice the frequency of

somaclonal variation.

There is little doubt that somaclonal variation does occur among

regenerates from many tissue culture systems. However, the ultimate value

of somaclonal variation (or any method of genetic improvement) is

determined by the relative expense (time and money) incurred and the

relative genetic advancement achieved. Maddock and Semple (1986) have

aptly pointed out that "where genetic variation per se can be obtained

readily by standard sexual crossing procedures, random tissue culture-

induced variability may, therefore, not be of great value if large-scale

screening of plant populations is necessary" (p. 1076). The relative

value of somaclonal variation increases when sexual hybridization is

difficult or concurrent in vitro selection is possible.








39
Somaclonal variation in Citrus. Somaclonal variation in plants
regenerated by embryogenesis from Citrus has been evaluated by several

researchers (Gmitter, 1985; Kobayashi, 1987; Navarro et al., 1979, 1985;

Vardi, 1977; Vardi et al., 1982) with somewhat inconclusive results. One

of nine plants regenerated from embryogenic orange cultures by Vardi

(1977) was tetraploid, but the author did not indicate whether this plant

was from normal or X-irradiated cultures. In later studies, Vardi et al.

(1982) reported that nucellar callus lines from seven Citrus cultivars

maintained normal chromosome numbers, while an eighth nucellar callus line

from 'Villafranca' lemon (C. limon [L.] Burm.) became predominantly

tetraploid. Vardi et al. (1982) indicated that plants with normal

morphology were regenerated from each of these callus cultures. Navarro

et al. (1979) reported that regenerated sweet orange (polyembryonic)

somaclones were morphologically normal. In contrast, Navarro et al.

(1985) identified about 28% of the somaclones as morphologically abnormal

after regeneration from nucellus explants of 'Clementine' (a monoembryonic

cultivar) by somatic embryogenesis. The observation of uniformity among

somaclones obtained from the same explant in this latter study (Navarro

et al., 1985) may indicate that the variation was preexistent in the

explant and was not the result of the tissue culture process. Kobayashi

(1987) examined morphological, cytological, and biochemical

characteristics of 25 orange somaclones regenerated by embryogenesis from

protoplasts and determined that they were "identical" to the nucellar

seedlings. Gmitter (1985) identified morphologically abnormal somaclones

among the plants regenerated by embryogenesis, but did not observe any








40

cytological or isozymic variation. No published studies have

systematically examined Citrus populations regenerated via organogenesis.


In Vitro Mutagenesis by Irradiation
Mutation was first employed as a method of genetic improvement over

sixty years ago (Muller, 1927; Stadler, 1928a, 1928b). Induced mutations

do not necessarily duplicate natural genetic variation (Allard, 1960;

Herskowitz, 1962; Stubbe, 1967). Most mutants will be lethal or unusable

because they are clearly deleterious (Brock, 1971; Hansel, 1967).
However, this has not prevented the utilization of mutagenesis as a

breeding strategy; by the end of 1986, over 700 cultivars that were

developed by mutation breeding had been released (Gottschalk and Wolff,

1983; Konzak, 1984; Micke and Donini, 1982; Micke et al., 1985; Micke et

al., 1987). Mutation in plant materials has most commonly been induced

by chemical mutagens or radiation. Chemical mutagens were not utilized

in this dissertation research (unless one considers 2,4-D as a mutagen;

see above), but the types and usage have been discussed by several authors

(IAEA, 1977; Kleinhofs et al., 1974; Levy and Ashri, 1975; Moustafa et

al., 1989; Neale, 1976).

Radiation has many different effects on living tissues (Coggle,

1983; Dertinger and Jung, 1970; Wierbicki et al., 1986), and one of these

effects is the induction of genetic mutations (Coggle, 1983; Lea, 1955;

Savage, 1989; United Nations, 1986). High energy forms of radiation,

including X- and gamma-rays, are typically known as ionizing radiation

because of their ability to cause electrons to be driven from atoms which

they contact. Both X- and gamma radiation have been investigated as








41

methods of inducing mutations for crop genetic improvement (Broertjes,

1982; Eapen, 1976; Hearn, 1984, 1986; Hensz, 1977; Huitema et al., 1989;

Liu and Deng, 1985; Moustafa et al., 1989; Russo et al., 1981; Sharma et
al., 1989; Van Harten et al., 1981; Wang et al., 1988; Werry, 1981). The

mutagenic effects of ionizing radiation on plant cells have been studied

by cytology (Kuehnert, 1962), morphology (Alexander et al., 1971;

Kuehnert, 1962; Yu and Yeager, 1960), and isozymes (Gulin et al., 1989),

and were quantified by the measurement of sister chromatid exchanges
(Kuglik et al., 1987, 1989). Some mechanisms of gamma-induced mutation
in plants may be similar to those described for yeast (Baranowska et al.,
1987; Chepurnoi, et al., 1989) or may involve the induction of double

stranded breaks and imperfect repair (Charlton et al., 1989). Irradiation

for induction of mutations can be easily applied to plant material in

vivo, but the resulting mutations have been frequently chimeric,

complicating their identification and isolation (Broertjes, 1982; D'Amato,

1965). Irradiation of callus cultures may avoid these problems because

embryos or adventitious shoots regenerated from tissue cultures often

develop from single cells (Bhatia et al., 1986; Broertjes et al., 1968;

Broertjes and Keen, 1980; Micke et al., 1987; Van Harten et al., 1981).

The special advantages of in vitro mutagenesis when it can be used in

combination with in vitro selection have been noted (Chaleff, 1981;

Huitema et al., 1989; IAEA, 1985; Ingram, 1983; Ingram and MacDonald,

1986; Werry, 1981), although unsolved problems (such as identifying in

vitro selection methods that correlate with field responses) have

generally remained obstacles to the successful utilization of these

techniques (Lorz and Brown, 1986; Wersuhn, 1988, 1989).








42

Low levels of ionizing radiation sometimes have been reported to

stimulate plant growth or morphogenesis, a phenomenon termed radiation

hormesis (Hell, 1983; Miller and Miller, 1987; Nuttall et al., 1968;

Sagan, 1989; Sax, 1963; Sheppard and Chubey, 1990; Sidrak and Suess,

1973). This effect has not been observed in other cases after the same

levels of irradiation (Montagno et al., 1989; Wolff, 1989). Whether the

phenomenon of hormesis has any bearing on mutation induction is unclear.

Sensitivity to irradiation has been observed to vary by a factor of

up to 500 among different species (Balito et al., 1989; Sparrow and

Woodwell, 1962), and differences also have been reported between tissues

and organs of a single species (Bajaj et al., 1970). Large nuclear

volumes and rapid growth rates generally correlate with greater

radiosensitivity, although higher ploidy levels tend to be more resistant

(Balito et al., 1989; Eapen, 1976; Galun and Raveh, 1975; Sparrow et al.,

1961; Sparrow and Woodwell, 1962). Gamma irradiation of tissue cultures

for genetic improvement has been investigated in several crops (Broertjes,

1982; Eapen, 1976; Moustafa et al., 1989), although as Balito et al.

(1989) noted, only limited information is available on the effects of

gamma irradiation on plant tissue cultures or regenerated plants.

Moustafa et al. (1989) observed that regeneration of maize plants from

callus was considerably more sensitive to gamma irradiation than was the

further growth of the callus. A dose of 100 Gy gamma radiation reduced

maize callus growth about 50%, while a similar reduction in plant

regeneration occurred with a treatment of only 25 Gy.

Clones with increased flesh and rind color, fewer seeds, and

decreased flesh acidity have been obtained by irradiation of Citrus








43

budwood or nucellar seeds (Hearn, 1984, 1986; Hensz, 1971, 1977, 1985;

Russo et al., 1981; Spiegel-Roy and Kochba, 1973a; Spiegel-Roy et al.,

1985; Starrantino et al., 1988a, 1988b; Yen, 1987; Zubrzycki and Diamante

de Zubrzycki, 1982). Despite the striking achievements of in vivo

irradiation relative to other approaches to Citrus genetic improvement,

only limited effort has been devoted to irradiation of Citrus tissue

cultures (Chang et al., 1984; Kochba and Spiegel-Roy, 1982; Legave et al.,

1989; Liu and Deng, 1985; Spiegel-Roy and Kochba, 1973b; Vardi, 1977).

A dramatic stimulation of Citrus embryo formation from embryogenic

cultures after gamma or X-irradiation has been reported by several workers

(Nito et al., 1989; Spiegel-Roy and Kochba, 1973b; Vardi, 1977). Liu and

Deng (1985; Chang et al., 1984) treated organogenic callus cultures of

several sweet orange cultivars with gamma radiation from a Cobalt-60

source. The LD-50 dose was determined to be 6-7 krad by percentage of

callus survival, and gamma irradiated callus was noted to have a reduction

in bud formation and bud growth when compared to controls. Increased

frequencies of chromosome aberrations and polyploids were observed among

the cells of the irradiated callus (Liu and Deng, 1985). Vardi (1977)

reported an LD-50 of 3.4 krad X-radiation for sweet orange protoplasts and

obtained several plants from the irradiated callus, including one

tetraploid. Nito et al. (1989) studied the effect of gamma radiation on

callus growth and embryoid formation in cultures of 'Valencia' (Citrus

sinensis [L.] Osbeck), 'Yoshida' (C. sinensis), Calamondin (C. madurensis

Lour.), Yuzu (C. junos Sieb. ex Tan.), and 'Ishizuka Wase' satsuma (C.

unshiu Marc.), and reported LD-50 values of 10-20, 20-50, 20-50, 1-5, and

10-20 krad, respectively. These LD-50 values are similar to those








44

reported by Spiegel-Roy and Kochba (1973b), but are difficult to reconcile

with the values of 3-7 krad reported by other researchers (although a

difference between embryogenic and organogenic cultures may contribute)

for irradiation of Citrus callus cultures (Chang et al., 1984; Legave et

al., 1989; Liu and Deng, 1985; Vardi, 1977).


Fruit Sector Chimeras

A chimera may be defined as an organism or organ containing two or

more genetically distinct tissue types (Cramer, 1954; Tilney-Bassett,

1986; Vaughn, 1983). Chimeras have been described in many plant species

(Chevreau et al., 1989; Pratt, 1983; Tilney-Bassett, 1986; Vaughn, 1983)

and have been classified in a number of different ways. Although normal

grafted plants are not considered to be chimeras because the two cell

types are completely separated except for the graft union (Cameron and

Frost, 1968; Neilson-Jones, 1969), chimeras may develop from grafted

plants when the two dissimilar genotypes grow together and produce

"hybrid" organs containing intermingled cells of both genotypes (generally

segregated by histogenic layers).

Most species of dicotyledonous plants (including Citrus [Cameron and

Frost, 1968]) behave as though there are three histogenic layers in the

apical meristem (Gifford, 1954). The two layers of the tunica (the L-I

and L-II) generally form the epidermal and subepidermal (including

gametophytic and nucellar) tissues, respectively (Pratt, 1983). The

innermost layer of the apical meristem is called the corpus (L-III) and

produces the central tissues of most organs including much of the vascular

system (Pratt, 1983). The fruit of Citrus spp. are rather unusual in that








45
the rind is derived predominantly from the L-II layer, but the flesh

(modified epidermal hairs) is derived mostly from L-I (Cameron and Frost,

1968; Cameron et al., 1964).

Chimeras produced by grafting are termed graft chimeras, or

synthetic chimeras (T. Tanaka, 1927b; Winkler, 1907), and may be

distinguished from those produced by nature, called autogenous (Cameron
and Frost, 1968; Neilson-Jones, 1969). Natural chimeras are sometimes

further classified according to type of genetic difference. Cytochimeras,

or chromosomal chimeras, are composed of tissues of different ploidys

(e.g. diploid and tetraploid) (Dermen, 1947; Einset et al., 1947; Marks,

1953) or distinct cell karyotypes (Brumfield, 1943), while genic chimeras

are composed of tissues with relatively simple genetic differences

(Chevreau et al., 1989; Dayton, 1966; Einset and Pratt, 1959). Spatial
distribution in the plant tissues (i.e. histogenic layer origin) is

perhaps the most common method of classifying chimeras (Swanson, 1957).
A periclinal chimera is formed when genetically different cell types are

completely segregated by histogenic layers (e.g. L-I, L-II, or L-III) and

all of each layer is of a single cell type (Dermen, 1960; Doodeman and

Bianchi, 1985; Hall et al., 1986a, 1986b; Jones, 1934; Marcotrigiano,

1986; T. Tanaka, 1927b). A chimera composed of a sector of tissue

including a portion of all histogenic layers is termed a sectorial chimera

(Dermen, 1947; Neilson-Jones, 1969; Varghese and Robbelen, 1984). The
mericlinal chimera is distinguished from the other two types by the

alternate genotype being found only in a part of one cell layer (usually

the epidermis) (Jorgensen and Crane, 1927; Vaughn and Wilson, 1980).

Chimeras composed of irregular combinations of the above types are known








46

as mixed or mosaic chimeras (Cameron and Frost, 1968). Chimeras,

especially periclinal chimeras, of some plant species have been separated

into the component genotypes as non-chimeric plants by irradiation

(Johnson, 1980; Pereau-Leroy, 1974; Sagawa and Mehlquist, 1957), tissue

culture (Dommergues and Gillot, 1973; Johnson, 1980), root cuttings

(Decourtye, 1987), or production of adventitious shoots (Dommergues and

Gillot, 1973; Stewart and Dermen, 1970).

Chimeras are common in Citrus, and one of the first plant chimeras

ever described was the "bizzarria," a synthetic chimera produced by a

failed graft of citron (C. medica L.) and sour orange (C. aurantium L.)

about 1644 (Darwin, 1921; Gallesio, 1811; Nati, 1674, 1929; Risso and

Poiteau, 1818; T. Tanaka, 1927b). Bud sports have occurred frequently on

many Citrus cultivars (Bono et al., 1981; Devaux, 1981; Hensz, 1981;

Iwamasa and Nishiura, 1981; Iwamasa et al., 1981; Mendel, 1981; Russo,

1981; Shamel, 1943; Soost et al., 1961) and probably originate as

periclinal, sectorial or mericlinal chimeras. Mendel (1981) suggested

"that varieties showing a tendency to produce bud mutations are chimeras

themselves (e.g. the 'Shamouti' orange)" (p. 86).

Many Citrus cultivars are known to be relatively stable periclinal

chimeras (Cameron and Frost, 1968; Cameron et al., 1964; Iwamasa and

Nishiura, 1970). The pink-fleshed grapefruit 'Thompson' arose as a bud

sport on a tree of the white-fleshed grapefruit, 'Marsh' (Hodgson, 1967).

The cultivar 'Thompson' contains pink flesh (derived from L-I) but

unpigmented rind (derived from L-II), while a budsport derived from it,

'Ruby Red', has both red flesh and pigmented rind. 'Thompson' nucellar

seedlings (derived from L-II) produce only fruit with white rind and white








47
flesh. Nucellar seedlings from 'Ruby Red' produce only red-fleshed fruit

with pigmented rind. 'Thompson' is probably a periclinal chimera
containing a genetic mutation for red pigmentation in the L-I layer.

'Ruby Red' is nonchimeric for the red pigmented cell type and probably
arose as a layer substitution in 'Thompson' (Cameron et al., 1964; Soost

and Cameron, 1975).

Other Citrus chimeras that have been described include, 'Burgundy'
grapefruit (Olson et al., 1966), 'Foster' grapefruit (Cameron et al.,

1964), 'Golden Buckeye' navel orange (Shamel et al., 1925), 'Kobayashi
Mikan' (Cameron and Frost, 1968; Iwamasa et al., 1977; Samura and
Nakahara, 1928), 'Gailiangcheng' orange (Lu, 1978, 1982), 'Suzuki Wase'

satsuma (Iwamasa and Nishiura, 1970; Iwamasa et al., 1977), and several
variegated chimeras with white and green leaves (Cameron and Frost, 1968;
Shamel, 1932). Nishiura and Iwamasa (1970) suggested that recovery of

two color forms among nucellar seedlings of the satsuma selection

'Dobashibeni Unshu' may indicate that it is a sectorial chimera composed

of orange-yellow and red pigmented components. Citrus cytochimeras have

been observed with diploid, tetraploid, and octoploid layers (Barrett,

1974; Frost and Krug, 1942).

Fruit sector chimeras with changes in rind color (Cameron and Frost,
1968; Coit, 1915; Frost, 1926; Iwamasa et al., 1977; Nishiura and Iwamasa,

1970), rind thickness (Cameron and Frost, 1968; Coit, 1915; Frost, 1926;
Shamel et al., 1918), rind texture (Cameron and Frost, 1968; Shamel et
al., 1918), flesh color (Cameron and Frost, 1968), ripening season (Coit,
1915), or rind injury resistance (W. Grierson, personal communication)
have been observed from trees of many Citrus cultivars; sometimes








48
individual trees produced the chimeras at a high frequency (Frost, 1926).

Some 'Ruby' orange fruit with sectors having altered rind color were noted

to contain adjacent pulp segments with similar color changes (Cameron and

Frost, 1968). Coit (1915) noted that citrus workers commonly considered

the sectored fruit to be the result of cross-pollination, but that such

fruit were actually the result of a mutation in the tree or at the base

of the ovary. Frost (1926) reported that the cause of some sectorial

chimeras appeared to be "differential mitosis" or chromosome non-

disjunction because occasionally chimeric fruit had "two adjacent sectors,

of similar width, whose rind varies in opposite directions from the normal

condition" (p. 394). Nishiura and Iwamasa (1970) suggested that the fruit

sector chimeras observed on 'Dobashibeni Unshu' were an indication that

this particular selection of satsuma was a sectorial chimera. The

frequency of fruit chimeras with one or more thickened rind sectors,

described as "ridging" or "coxcombing," was observed to be about 0.1% in
navel orange, 0.1% in 'Valencia' orange, 0.2% for lemon, and 0.2% for

grapefruit in normal California orchards (Sinclair and Lindgren, 1943).

However, fumigation of Citrus trees with hydrocyanic acid (Lindgren and

Sinclair, 1941; Sinclair and Lindgren, 1943) or chlorpyrifos (M.L. Arpaia,

personal communication) during the early stages of fruit bud development

(January or February) was found to cause a dramatic increase in the

frequency of ridged fruit. Sinclair and Lindgren (1943) suggested that

hydrocyanic acid fumigation may cause "a genetic change, such as a

doubling of the number of chromosomes" (p. 104), in one or a small number

of cells in the developing ovary, and that this mutation could then be

propagated by mitosis to form the ridged sectors. Iwamasa et al. (1977)








49
isolated seeds from beneath the normal orange and mutant yellow sectors

of one chimeric fruit of 'Fukuhara' orange. Trees that developed from

seed borne in the orange sector produced unsectored orange-colored fruit,

while those that developed from seed borne in the yellow sector produced

unsectored yellow fruit. It was stated that a similar event had been

previously observed by Iwamasa and Nishiura (1970). Iwamasa et al. (1977)

suggested that the nucellar seed and rind sectors had developed from the

same L-II histogen and that sectors with improved fruit color might

therefore be used to produce better cultivars.


Evaluation of Resistance to Phytophthora

Phytophthora spp. cause serious soilborne diseases of many important
world crops. Several Phytophthora species can invade Citrus, including

P. parasitica Dast., P. citrophthora (R.E. Sm.& E.H. Sm.) Leonian, P.

hibernalis Carne, P. syringae Kleb., P. palmivora (Butler), and P.

citricola Saw. The most important of these species in Florida is P.

parasitica, which is responsible for root rot, foot rot, and gummosis

(Timmer and Menge, 1988), as well as causing damping-off of young Citrus

seedlings (Timmer, 1988). Citrus feeder roots are invaded principally by

zoospores from infected soil through wounds or undamaged root tips.

Zoospore germination and mycelial growth lead to a subsequent decay of the

root cortex that may result in tree decline (Timmer and Menge, 1988).

Phytophthora can invade the trunk of Citrus trees through openings in the

bark produced by wounding or natural cracking. The ensuing foot rot or

gummosis is characterized by necrosis of the cambium and rotting of the

bark. Foot rot is the name applied when the disease affects the tree








50
slightly above, at, or slightly below ground level, while the term

gummosis is more frequently used when the disease occurs higher in the

trunk. Trees affected by foot rot/gummosis may show symptoms of gum

exudation, dieback, weak growth, and/or death (Lutz and Menge, 1986;

Timmer and Menge, 1988).

Phytophthora can be controlled in the nursery by hot water treatment

of seeds, fumigation of soil, or soil drenches with appropriate fungicides
(Timmer and Menge, 1988). Some degree of control in young and mature

orchards is possible by fungicide application as foliar sprays of systemic

compounds (Davis, 1981; Sandler et al., 1989), drenches (for root rot),

or trunk paints (for foot rot and gummosis), as well as by sanitation

(Sandler et al., 1989; Timmer, 1977; Timmer and Menge, 1988). The most

economical control is planting resistant rootstocks with the bud union

well above soil level. Cultivars used as rootstocks vary widely in their

susceptibility to the diseases caused by Phytophthora (Broadbent et al.,

1971; Carpenter and Furr, 1962; Graham, in press; Hutchison and Grimm,

1972; Klotz et al., 1968; Klotz and Calavan, 1978; Smith et al., 1987;

Timmer and Menge, 1988). Although a general correlation between relative

resistance to root rot and foot rot has been observed for many Citrus

selections, there are reports of substantial differences in response of

some selections to the two diseases (Carpenter and Furr, 1962; Furr and

Carpenter, 1961; Grimm and Hutchison, 1973, 1977). 'Carrizo' citrange and

sour orange are tolerant to foot rot (Castle et al., 1989; Timmer and

Menge, 1988), yet were reported to be susceptible to root rot (Graham, in

press). Relatively little is known about the mechanism(s) of resistance

to Phytophthora in Citrus, although the involvement of coumarin








51
phytoalexins has been suggested (Afek and Sztejnberg, 1986, 1988; Afek et

al., 1986; Vernenghi et al., 1987). Isolate specificity has not been

reported for Phytophthora within Citrus. However, Matheron and Matejka

(1990) determined that isolates of P. parasitica isolated from tomato,

petunia, and five other unrelated hosts were not virulent on rough lemon

(Citrus jambhiri Lush.), and similar species specificity for isolates was

observed by other workers (Wheeler and Boyle, 1971). Hutchison (1985)
reported that resistance to P. parasitica in Citrus is probably controlled

by a multiple gene system. Because of the great differences in

characteristics of the affected tissues and etiologies (Lutz and Menge,

1986; Timmer and Menge, 1988), it seems probable that one or more

mechanism(s) of resistance are not common to both root rot and foot rot.

Several methods of evaluating resistance of young Citrus to root rot
and foot rot have been described. Soil flooding with zoospore suspensions

has frequently been used to evaluate susceptibility to root rot (Cameron

et al., 1972; Carpenter and Furr, 1962; Grimm and Hutchison, 1973;

Whiteside, 1974), although the results have not always been reliable

(Carpenter and Furr, 1962; Furr and Carpenter, 1961; Grimm and Hutchison,

1973). Tsao and Garber (1960) reported that soil infestation with

mycelium grown in liquid cultures produced reliable results.

Chlamydospores may provide a more uniform source of inoculum (Farih et

al., 1981; Graham and Egel, 1988), and Graham (in press) described a
procedure for chlamydospore inoculation and susceptibility assay.

Controlled evaluations of foot rot/gummosis resistance have

typically been completed by wounding trunk tissue and inoculating with

agar disks from mycelial cultures (Klotz et al., 1968; Smith et al.,








52

1987). General ratings of field resistance have commonly relied upon

observations of foot rot/gummosis susceptibility because it is the most

visible Phytophthora disease, despite a high degree of variability in

damage observed at different times of testing (Smith et al., 1987;

Whiteside, 1971).

Although several different methods of evaluating susceptibility are

available, screening Citrus plants for resistance to Phytophthora has

proven to be time-consuming and relatively imprecise in rootstock

development programs (Broadbent, 1971; Carpenter and Furr, 1962; Furr and

Carpenter, 1961; Grimm and Hutchison, 1973; Hutchison, 1985; Smith et al.,

1987). The development of a more rapid and reliable method for

determination of Phytophthora resistance would be of great benefit to

breeding programs.

Tissue culture techniques have become increasingly important for

programs involved in the genetic improvement of Citrus (Gmitter and Moore,

1986; Grosser and Gmitter, 1989; Grosser et al., 1988a, 1988b, 1989;

Hidaka and Kajiura, 1988; Kobayashi and Ohgawara, 1988; Kobayashi et al.,

1983, 1988; Kochba and Spiegel-Roy, 1982; Kochba et al., 1972; Navarro et

al., 1985; Ohgawara et al., 1985; Sim et al., 1988; Vardi and Galun, 1988;

Vardi et al., 1986b, 1987, 1989). Some success has been achieved in

characterizing or selecting for plant resistance to diseases in vitro

(Behnke, 1979, 1980; Brisset et al., 1988; Buiatti and Scala, 1985; Chang

et al., 1989; Dunbar and Stephens, 1989; Wenzel et al., 1985; Willmot et

al., 1989), and the advantages of in vitro selection have been noted

(Chaleff, 1981, 1983; Huitema et al., 1989; IAEA, 1985; Ingram, 1983;

Ingram and MacDonald, 1986; Werry, 1981). In vitro resistance of plant








53

tissues to Phytophthora in dual cultures (plant tissue and the pathogen

together in axenic culture) has been correlated with field performance

for: avocado (Dolan and Coffey, 1986; Zilberstein and Pinkas, 1987) and

crops of seven other genera (McComb et al., 1987) to P. cinnamomi Rands;

potato (Ingram, 1967) and tomato (Warren and Routley, 1970) to P.

infestans (Mont.)DBy.; alfalfa (Miller et al., 1984) to P. megasperma

Drechs; tobacco (Helgeson et al., 1976; Tedford et al., 1990) to P.

parasitica Dast. var. nicotianae (Breda de Haan) Tucker; and apple

(Barritt et al., 1990; Jeffers et al., 1981; Utkhede, 1986; Utkhede and

Quamme, 1988) to P. cactorum (Leb. & Cohn) Schroet. Isolate specificity

was noted in the reaction of papaya shoot cultures to P. palmivora (Butl.)

Butl. (Sharma and Skidmore, 1988). Measurement of lesions formed on

excised apple twigs after inoculation with P. cactorum has proven to be

a good indicator of relative resistance (Barritt et al., 1990; Jeffers et

al., 1981; Utkhede, 1986; Utkhede and Quamme, 1988). Preliminary

investigations on the response of Citrus cultivars to culture filtrate

from P. citrophthora were not particularly encouraging (Vardi et al.,

1986a), although Tusa et al. (1988) obtained a fair correlation between

in vitro tolerance to culture filtrate and in vivo tolerance to bark

inoculations in sour orange selections. No investigations of in vitro

response of Citrus cultivars in dual culture with P. parasitica have been

reported.














CHAPTER 3
REDISCOVERY OF CARIBBEAN FORBIDDEN FRUIT AND
EVALUATION OF ITS SIGNIFICANCE FOR CITRUS BREEDING


Introduction

A single introduction of grapefruit (Citrus xparadisi Macfadyen)

into Florida from the West Indies in 1823 has been identified in the

literature as the original source of all known grapefruit germplasm.

Essentially all grapefruit cultivars in the United States are bud sports

or nucellar seedlings derived from this single introduction. On the basis

of historical records as well as morphological and biochemical

characteristics, the grapefruit is considered to have originated in the

West Indies as one or a series of chance hybridizations between sweet

orange (C. sinensis) and pummelo (C. grandis) during the 17th or 18th

centuries (Scora et al., 1982; Scora and Kumamoto, 1983). In the first

literature describing the grapefruit, it is closely associated with

another kind of Citrus known as "Forbidden Fruit." Kumamoto et al. (1987)

considered this form to be extinct. Subsequently, we identified a

heterogeneous population of grapefruit-like Citrus growing in Saint Lucia

(West Indies) that is known as forbidden fruit by some of the local

residents (Bowman and Gmitter, 1990b). This population may provide a

source of genetic diversity for grapefruit and hybrid cultivar development

(Bowman and Gmitter, 1990a).


54








55

The objectives of this project were to:

1. Describe the Citrus populations known in Saint Lucia as

forbidden fruit;

2. Investigate and describe the probable relationship of

these populations to the forbidden fruit mentioned in the

early Caribbean literature and to the grapefruit;

3. Survey the history of grapefruit and its connection with

the Caribbean forbidden fruit;

4. Identify some of the needs of grapefruit genetic

improvement and limitations imposed by currently available

grapefruit germplasm;

5. Import (under quarantine) budwood and seeds of some of the

Caribbean forbidden fruit selections for characterization and

testing in Florida.


Materials and Methods

An extensive search for selections of the Caribbean forbidden fruit

was initiated in Saint Lucia during 1986 and 1987. A travel grant was

obtained (with F.G. Gmitter as co-investigator) from The Center For

Tropical Agriculture/ International Programs (University of Florida)

during 1987 to allow for the collection of seeds and budwood from the

previously identified forbidden fruit selections. This trip was completed

by F.G. Gmitter and K.D. Bowman during December, 1987, and included

careful examination of five Saint Lucia forbidden fruit selections,

collection of seeds and budwood, and discussions with Saint Lucians who

were able to identify these trees as forbidden fruit. The budwood








56

selections were turned over to the Florida Department of Plant Industry,

Gainesville, for shoot-tip budding, thermotherapy, and virus indexing.

Additional seeds were obtained from one of these selections and another

similar Saint Lucian selection in February and March, 1990. Seeds of

morphologically similar selections were obtained from Saint Vincent and

Trinidad in 1990.

Seeds of all selections were planted in soilless potting mix, and

plants were grown in a greenhouse. Seeds of selections SF23, SF24, and

SF25 were surface sterilized by immersion and agitation in 70% ethyl

alcohol for 10 minutes followed by 1.05% sodium hypochlorite plus 2 drops

polyoxyethylene-20-sorbitan monolaurate (Tween 20: Fisher Scientific,

Pittsburgh) per 100 ml for 20 minutes. Seeds were rinsed five times (five

minutes each) in sterile distilled water and placed on the surface of 10

ml agar-solidified half-strength MT basal medium with full strength iron

(Murashige and Tucker, 1969) and 25 gl-'1 sucrose, adjusted to pH 5.7

(called GM1 medium through remainder of this text), that had been

solidified in the bottom of 25 x 150 ml glass culture tubes (Bellco Glass,

Inc., Vineland, NJ). The tubes were covered with "kap-uts" (Bellco Glass)

translucent plastic closures.

Shoot tip cuttings (mericlones) from in vitro seedlings of

selections SF23-1 and SF24-1 were excised and placed with the cut surface

in RM1 rooting medium. RM1 medium was composed of agar-solidified half-

strength MT basal medium with full-strength iron (Murashige and Tucker,

1969), 25 g-1-' sucrose, 500 mgl-'1 activated charcoal, and 0.02 mg-l- NAA,

and adjusted to pH 5.8, as described by Grosser et al. (1989). Epicotyl

and internode segments (1 cm long) were excised from SF23-1 and SF24-1 and








57

placed horizontally on the surface of a callus proliferation media,

composed of solidified MT basal (Murashige and Tucker, 1969) with 50 g'l1

sucrose, 0.22 mg--' kinetin, 0.5 mg-l-' 2,4-D and adjusted to pH 5.8

(called CP1 media throughout this text), in 15 x 100 mm Petri plates.

After 6 weeks on callus proliferation medium, segments with

associated callus were transferred to a shoot induction medium (SIM

through the remainder of this text) composed of solidified MT basal

(Murashige and Tucker, 1969) with 25 g-I sucrose, 0.5 g1l-' malt extract,

0.25 gl1-' casein hydrolysate, 5 mg1l-' BA, 0.01 mgl-' 2,4-D and adjusted

to pH 5.8, in 20 x 100 mm Petri plates (SIM medium recipe, personal
communication from J.W. Grosser). Adventitious shoots (somaclones) that
formed on callused explants were excised and placed on rooting medium

(RM1). Rooted mericlones and somaclones were transferred to soilless

potting mix in controlled environment chambers. Mericlones were

morphologically characterized with other control genotypes.

Leaf samples from imported budwood and seedling selections were used

for isozyme analyses on horizontal gels containing 10% Connaught starch

(Fisher Scientific) and sometimes supplemented with 0.15% electrophoresis

grade agarose. Extracts from leaf samples were made by crushing small

rectangles of chromatography paper (wicks) into the abaxial side of leaf

tissue until they were green with cell juices immediately prior to gel

loading (Torres et al., 1978). Isozymes were separated using a pH 5.7

histidine-citrate buffer (Cardy et al., 1981) or Tris-citrate buffer

(Gmitter, 1985) with 3 hours electrophoresis at 50 ma constant current and

4C. Activity stains used were PGM (phosphoglucomutase), PGI

(phosphoglucose isomerase), PER (peroxidase), and PGD (6-phosphogluconate








58
dehydrogenase) on the histidine-citrate buffer system, or GOT (glutamate

oxaloacetate transaminase), and SDH (shikimic acid dehydrogenase) on the

tris-citrate buffer system. Stain recipes were as described by Vallejos

(1983) with slight modifications (see also Almansa et al., 1989; Kephart,

1990; Normand, 1988).


Results and Discussion
The origin and history of contemporary grapefruit cultivars has been

investigated through the literature (Figure 3.1). Essentially all the

popular cultivars were derived from a single set of germplasm introduced

to Florida about 1823. This germplasm and the grapefruit hybrid species

itself probably originated someplace in the West Indies, where it was

closely associated with another name or kind of Citrus, the forbidden

fruit. Natural populations of grapefruit-like Citrus have not been

evaluated as potential sources of germplasm. Initial observations of a

diverse semi-wild population of Citrus called forbidden fruit in Saint

Lucia, West Indies, indicated that it might be a useful source of genetic

material and prompted more thorough evaluations and germplasm collections.

Forbidden fruit budwood is now growing under quarantine at the DPI

facility in Gainesville and will be released after it is certified virus

free. Seedlings of five selections are growing in the growth chamber,

greenhouse, and/or field. Two manuscripts on the history, rediscovery,

and characteristics of the forbidden fruit and its significance for

grapefruit genetic improvement have been published (Bowman and Gmitter,

1990a, 1990b). Some of this published information is repeated here.








59


West Indies Tree(s)

Sseed or plant(s)

Safety Harbor, FL
AOO 01823 ft*
Marsh 1860 Walters 1887
limb | limb
sport sport

limb limb
sport sport
'im I ---------- I ------e
Ruby ~^^ Hudson
Sirradiated
limb budwood irradiated
sports I seed
Rio Red
Ray/Henderson Star Ruby


Figure 3.1. Probable pedigree of grapefruit cultivars
(modified from Bowman and Gmitter, 1990a; p. 42).




Many Saint Lucians who were interviewed had never heard of forbidden

fruit or only knew it was a kind of Citrus similar to grapefruit.

However, two morphologically distinct kinds of Citrus were identified as

forbidden fruit by some Saint Lucian citizens. The first of these

morphologically resembled the species Citrus grandis and examination of

seed samples indicated monoembryony. Only five trees of this kind could

be located, although other trees were mentioned by residents. Two of

these were known to have been vegetatively propagated by grafting from one

of the other three trees and another was very unhealthy. The two healthy,

mature (and probably seedling) trees were used as specimens for derivation

of the following description. One of these trees was located on the








60

Raveneau Estate, Souci, Saint Lucia and identified as forbidden fruit by

the resident estate owner, Fitz Raveneau. This tree was quite large and

overgrown by vines and adjacent vegetation. Fruit sampled the first week

of December, 1987, were very sour. Fitz Raveneau indicated that rind from

the fruit of this tree was once used to make a type of candy. The second

grandis-type tree used as a specimen to construct the description was

growing on the Fond Doux Estate, Soufriere, Saint Lucia and was identified

as forbidden fruit by Vernon, the resident estate manager. The original

trunk of this tree was very large but oriented horizontally on the soil

surface as a result of a hurricane in 1980. The "tree" that remained was

composed of one large vertical branch that appeared to be healthy. Fruit

of this tree were sweet and had a pleasant flavor when sampled between

January-March, 1986 and 1987, as well as again in December, 1987. A

composite description of these two trees from Bowman and Gmitter (1990b)

is reproduced below.

Saint Lucia Forbidden Fruit, grandis-type (GF)
Trees to 12 m tall, trunk upright, branches spreading.
Leaves evergreen, alternate; blade elliptic to ovate, 9-18 x
5-9 cm, glabrous above, glabrous with sparsely puberulent
veins below, apex obtuse to acute, margins entire to crenate,
base obtuse to cuneate, venation pinnate; petiole sub-alate to
broadly alate, articulated, oblanceolate or obovate,
occasionally with emarginate apex, 2-9 cm long. Young twigs
green, angled, bearing single axillary spines, 0-25 mm long;
mature twigs cylindrical, often spineless; older branches
apparently spineless. Flowers not observed. Fruit borne
singly or in pairs, a semi-sweet to acidic yellow hesperidium,
oblate to pyriform, 20-25 cm x 15-20 cm; basal area slightly
depressed, apex truncate; pedicel medium to very large;
flavedo coarsely glandular and turning yellow at full
maturity; albedo spongy, white, 1-4 cm thick; flesh
greenish-yellow to pale white, 11-14 locules, with large,
easily separated pulp-vesicles; central axis irregular,
hollow. Seeds none to 100 per fruit, 15-20 mm x 6-12 mm,
cream-color, obovoid, flattened, angular, ridged, rough,
monoembryonic; inner seedcoat light brown, dark chalazal spot;
cotyledons white. (Bowman and Gmitter, 1990b; p. 168)








61

A second kind of Citrus found growing sparsely throughout Saint

Lucia was identified as the forbidden fruit of the past by MacDonald

Ferdinand (born 1893) of Praslin, Micoud, Saint Lucia. Trees of this kind

were somewhat smaller than the grandis-type and were more commonly known

by Saint Lucians as "shaddette." Shaddette fruit varied considerably in

characteristics from tree to tree, but were generally smaller and more

similar to grapefruit or sweet orange than those of the grandis-type.

Four shaddette trees were selected as specimens for the description
reproduced from Bowman and Gmitter (1990b) below. Two of these trees were

located near Fond-Saint-Jacques, Soufriere, one on the Fond Doux Estate,

Soufriere, and the fourth in Blanchard, Micoud.

Saint Lucia Forbidden Fruit, shaddette-type (SF)
Trees to 10 m tall, trunk upright, branches spreading.
Leaves evergreen, alternate; blade elliptic to ovate, 6-10 x
4-6 cm, venation pinnate, blade glabrous above and below, apex
obtuse to acute or occasionally emarginate, margins entire to
crenate, base obtuse; petiole sub-alate to alate, articulated,
oblanceolate or obovate, sometimes with emarginate apex, 2-3
cm long. Young twigs green, angled, bearing single axillary
spines, 2-5 mm; mature twigs cylindrical, sometimes spineless;
older branches spineless or sometimes very stoutly spined.
Flowers in small racemes or one to several in leaf axils,
perfect, actinomorphic; calyx cup-shaped with 4-5 lobes;
petals 4-5, white to cream; stamens many; stigma flattened
globose. Fruit a semi-sweet hesperidium, oblate, globose, or
pyriform, 9-15 x 9-12 cm, sometimes hanging in distinct
clusters; basal area frequently with shoulder or neck,
depressed, lobed; apex truncate to rounded; flavedo moderately
glandular and turning yellow at maturity; albedo white to
pinkish, 5-20 mm thick; flesh pale yellow to pale pink, 11-14
locules, medium to large semi-coherent pulp vesicles; central
axis solid. Seeds 5-30 per fruit, 15-16 mm x 7-9 mm, white,
angular-ovate, moderately rough, monoembryonic and
polyembryonic; inner seedcoat light to medium brown, dark
brown chalazal spot; cotyledons white.
(Bowman and Gmitter, 1990b; pp. 168-170)

Both types of forbidden fruit were sometimes known by a second name

in the local French creole dialect. This name was spoken as though it








62

were spelled "fwee dayfwandee" in English and literally translates to

"fruit forbidden." The name "forbidden fruit" was considered by Saint

Lucians to have originated from the biblical story of the Garden of Eden.

The morphological descriptions of the field specimens indicated

grandis-type forbidden fruit (GF) closely resembled selections of Citrus

grandis, a monoembryonic species. Because of the monoembryony, all

seedlings of C. grandis are genetically unique and some morphological

variation among seedlings is expected. Consequently, the two plant sample

used to construct the morphological description of GF above may be

inadequate to encompass normal genetic variation within this kind of

forbidden fruit.

Vegetative morphology of mature shaddette-type forbidden fruit (SF)

trees in Saint Lucia was very similar to that of nearby grapefruit trees.

However, as noted in the description above, fruit varied considerably from

tree to tree, with some trees bearing more pummelo-like fruit, others

grapefruit-like fruit, and yet others somewhat orange-like fruit. The

four selections chosen as models for construction of the description above

probably do not adequately sample all the variation in characteristics

possible within the shaddettes. Examinations of small numbers of seeds

from several selections indicated that at least some selections contained

both monoembryonic and polyembryonic seeds.

In vitro seedling growth, callus induction on internodal explants,

adventitious shoot formation, and shoot rooting were evaluated for one

seedling selection of SF23 (SF23-1) and one seedling of SF24 (SF24-1).

Both seedlings were observed to grow well in culture, produce large

amounts of callus from explants, and form large numbers of adventitious








63

shoots from the callus. It is worth noting that adventitious shoot

formation was much more prolific on explants from both shaddette

selections than on explants from 'Valencia', 'Hamlin', and 'Ridge

Pineapple' sweet oranges or 'Duncan' grapefruit. Excised shoots of both

shaddette seedling selections were observed to root readily on RM1 medium.

Samples of shoots from 'Carrizo', 'Hamlin', 'Valencia', and SF23-1 were

examined for rooting at 20, 24, 28, and 32 days after placement on RM1

medium in culture boxes (Fig. 3.2). One month after placement on RM1, 33%

of SF23-1 shoots, 27% of 'Carrizo' shoots, 18% of 'Hamlin' shoots, and 0%

of 'Valencia' shoots had formed roots. After 44 days on RM1 medium, one

of 23 'Valencia' shoots (4%) had formed a root.










40- m Valencia (23 shoots) l Hamlin (57 shoots)
Li Carrizo (81 shoots) B SF23-1 (18 shoots)

% 30
R
0
0
t 20
e
d

10-



20 24 28 32
Days on RM1

Figure 3.2. Frequency of shoot rooting during the first
month on RM1 medium.








64
Meristem clones propagated by tissue culture were established in

soilless potting mix and morphologically characterized (Tables A.1-A.6).

Morphology of both shaddette clones differed considerably from 'Hamlin'

and 'Valencia'. The shaddette seedlings closely resembled the 'Duncan'

and 'Marsh' clones for the characteristics spine length, spine

length/diameter, oil gland density, and internode length.

Isozyme analysis was completed on one of the two specimen GF clones
(the other did not survive propagation in quarantine), as well one

standard selection of sweet orange, grapefruit, and pummelo (previous work

has indicated almost no variation within these species [F.G. Gmitter,

personal communication; Torres et al., 1978]). Allelic constitutions of

these clones were deduced from banding patterns of PGM, PGI, PER, SDH, and

GOT activity stains (Table 3.1). Banding patterns of PGM, PGI, SDH, and

GOT for samples from GF13 were identical with those observed for common

Florida selections of pummelo, C. grandis (Table 3.1). Clone GF13 was

observed to differ from common pummelo selections and resemble grapefruit

in its heterozygosity (FS) at the peroxidase locus. These isozyme data

are consistent with the possibility that GF13 is a hybrid between

grapefruit (or shaddette) and pummelo, but not a simple hybrid between

sweet orange and pummelo (because GF13 is homozygous FF at the GOT locus).

However, a thorough survey of C. grandis has not been completed, and it

is possible the F allele for peroxidase is present in some pure pummelo

selections.

Isozymic characterization of the three shaddette clones (SF23, SF24,
and SF25) was completed on plant material in quarantine. Allelic

constitution of these clones and standard selections of sweet orange,








65


Table 3.1. Isozyme genotypes of seven Citrus selections.

PGM PGI PER SDH GOT1
Sweet Orange FS MS FF IS SS

Pummelo SS SS SS SS FF, FS

Grapefruit SS SS FS IS FS

SF23 SS SS FS IS FS

SF24 SS SS FS IS FS

SF25 SS SS FS IS FS

GF13 SS SS FS SS FF

SF = shaddette-type forbidden fruit
GF = grandis-type forbidden fruit





grapefruit, and pummelo were deduced from banding patterns of PGM, PGI,

PER, SDH, and GOT activity stains (Table 3.1). The three shaddette clones

and grapefruit had identical banding patterns for PGM, PGI, PER, SDH and

GOT. Some preliminary evidence for differences in banding patterns of

shaddette clones and grapefruit were obtained with PGD (personal

communication, F.G. Gmitter). The shaddette selections were

differentiated from pummelo by being heterozygous at PER (FS) and SDH (IS)

loci. Shaddettes were distinguished from sweet orange by PGM, PGI, PER,

and GOT. Allelic differences between the GF13 clone and the shaddette

selections were detected with SDH and GOT as an increase in heterozygosity

among the shaddettes.








66
Although the three shaddette clones were quite similar to grapefruit

in morphologies and isozyme banding patterns, shaddette seedlings showed

a much higher frequency of isozyme variation at the PER locus than

seedlings from three grapefruit cultivars (Table 3.2; Gmitter et al., in

preparation). Similar isozyme variation has been observed among shaddette

seedlings at other isozyme loci (F.G. Gmitter, personal communication).

The appearance of altered isozyme genotypes indicates that seedlings are

zygotic in origin. An average 27 percent of shaddette seedlings showed

recombination at the PER locus by the presence of FF or SS homozygotes.

Because the seed parent clones are FS at peroxidase (Table 3.1), selfed

zygotic progeny should segregate at a ratio of 1FF:2FS:ISS. Therefore,

only about half of the zygotics would be detected as homozygotes at a

single locus. Ignoring the possibility of cross pollination, the

projected frequency of zygotics in a population of shaddette seedlings may

be about 54 percent. This is in great contrast to the 0 to 8 percent

zygotics predicted on the basis of the peroxidase segregation for the

grapefruit cultivars.

Although the phylogenetic relationships cannot be determined

precisely, isozyme evidence supports the morphological indications of a

close relationship between the grapefruit and the Saint Lucian forbidden

fruit. The connection between the two is further corroborated by the

close association of the names in historical literature. Bowman and

Gmitter (1990b) have proposed that the Saint Lucian shaddette and

grapefruit should be considered as two different groups of selections

within the hybrid species C. xparadisi. The name "forbidden fruit" was

apparently derived from an early C. grandis selection called "Adam's








67

Apple." The name forbidden fruit has subsequently been applied

sporadically to many (all?) different selections of C. grandis and C.

xparadisi in the West Indian historical literature. The hypothesized

species and nomenclatural relationships, as well as an indication of the

genetic diversity within populations, are presented pictorially in Figure

3.3. Bowman and Gmitter (1990a) suggested that the shaddette may be a

valuable germplasm resource, a proposal strengthened by the isozyme

evidence presented here.







Table 3.2. PER zymotypes and percent zygotics identified
among seedlings of grapefruit and three
shaddette (SF) selections
(Gmitter et al., in preparation).

Number Percent
Seedlings PER Locus Zygotics
Examined SS FS FF Identified

Duncan 51 0 49 2 4%

Foster 50 0 50 0 0

Marsh 50 0 50 0 0

SF23 41 12 29 0 29

SF24 39 8 27 4 31

SF25 23 5 18 0 22

SF = shaddette-type forbidden fruit











68








C. reticulata famb-
hiri




C.
aurantiumn
C. sinenais







Grapte
C. xparadisi fruj
Apple C- grandis
of
Shaddette Adam


\ ^_ ~-- p(Fruit .




Figure 3.3. Speculative taxonomic and nomenclatural relationships between
forms related to C. xparadisi (modified from Scora, 1975; p. 373).














CHAPTER 4

EVALUATION OF ORGANOGENIC CITRUS TISSUE CULTURES
AS A SOURCE OF GENETIC VARIATION
FOR CULTIVAR IMPROVEMENT

Introduction
Numerous reports of variation among plants regenerated from tissue

cultures have been recorded in the past two decades (D'Amato, 1985; Evans,

1989; Evans et al., 1984; Evans and Sharp, 1986; Orton, 1984). The term
"somaclonal variation" was coined by Larkin and Scowcroft (1981) to

describe this phenomenon. In some cases, useful mutations have been

reported in somaclonal variants, including higher yield (Heinz and Mee,

1971), male sterility (Griesbach, 1989), disease resistance (Heinz et al.,

1977; Krishnamurthi and Tlaskal, 1974; Orton, 1984; Shepard et al., 1980),

insect resistance (Miles et al., 1981), dwarfing (Griesbach, 1989), and

a wide variety of other morphological characteristics (Burk and Matzinger,

1976; Cummings et al., 1976; Eapen et al., 1989; George and Rao, 1983; Liu

and Chen, 1978). The severe limitations on sexual hybridization of Citrus

imposed by long juvenility and extensive nucellar polyembryony make

somaclonal variation an attractive alternative method of generating useful

genetic variation. Only a few preliminary investigations of somaclonal

variation have been completed in Citrus, with somewhat conflicting results

(Gmitter, 1985; Kobayashi, 1987; Navarro et al, 1985). No thorough



69








70

investigations of somaclonal variation arising from organogenic

regeneration of Citrus cultures have been reported.

If somatic variation from Citrus tissue cultures is not frequent

enough to make it productive in cultivar improvement, mutagenic treatments

may be employed to increase genetic variation. Irradiation of budwood and

nucellar seeds have both been successful in Citrus genetic improvement

(Hearn, 1984, 1986; Hensz, 1977, 1985), and there have been some

investigations of irradiation of Citrus tissue cultures (Chang et al.,

1984; Kochba and Spiegel-Roy, 1982; Liu and Deng, 1985; Spiegel-Roy and

Kochba, 1973b; Vardi, 1977; Zubrzycki and Diamante De Zubrzycki, 1982).

However, little information is available on the effect of gamma radiation

on Citrus organogenic and embryogenic callus cultures and the resulting

plants.

This study investigated the potential for somaclonal variation in

Citrus genetic improvement and studied the effects of gamma radiation on

callus cultures. The somaclones were produced by organogenesis from

internodal explants and compared with control plants produced by in vitro

cuttings from apical and axillary meristems (mericlones) of the same plant

material. The objectives were to:

1. Regenerate plants by organogenesis from callused

Citrus explants in vitro;

2. Compare the juvenile morphology of these plants and plants

produced by meristem cuttings from the same seedlings

and/or from other seedlings of the same cultivar;

3. Determine the effect of gamma radiation on

Citrus explants.








71

Materials and Methods
Production of somaclones. 'Hamlin' and 'Valencia' sweet oranges (C.
sinensis [L.] Osbeck) were chosen for the study of somaclonal variation

because they are common cultivars that might benefit by some types of

somatic mutations. The cultivar 'Carrizo' citrange (C. sinensis x P.

trifoliata [L.] Raf.), known to be responsive in culture, also was

included because the two sweet orange cultivars did not produce shoots

readily from callus during preliminary studies. Seeds of 'Carrizo',

'Hamlin', 'Valencia', 'Duncan' grapefruit (C. xparadisi Macfadyen),
shaddette selections SF23-1 and SF24-1 (presumably C. xparadisi; see

Chapter 3) and 'Meiwa' kumquat (Fortunella crassifolia Swing.) were

surface sterilized by immersion and agitation in 70% ethyl alcohol for 10

minutes followed by a solution of 1.05% sodium hypochlorite plus two drops

polyoxyethylene-20-sorbitan monolaurate (Tween 20: Fisher Scientific,

Pittsburgh) per 100 ml for 20 minutes. The seeds were rinsed in sterile

distilled water five times (five minutes each rinse) and individual seeds

placed on the surface of 10 ml germination medium (GM1; see Chapter 3)

solidified in the bottom of 25 x 150 mm glass culture tubes (Bellco Glass,

Inc., Vineland, NJ). The tubes were covered with translucent plastic Kap-

uts closures (Bellco Glass) and sealed with Nescofilm (Nippon Shoji Kaisha

Ltd., Osaka, Japan). In vitro seedling cultures were maintained at 27C

under constant fluorescent lighting.

After seeds germinated, 1 cm epicotyl and internode segments (called
"segments" through the remainder of this chapter) were excised and placed

horizontally on the surface of a callus proliferation medium, called CP1

(see Chapter 3). Some 'Carrizo' segments were sacrificed for








72

determination of the starting mass of apical and basal halves of the

explants. 'Carrizo' segments that were to be used for the 21 day mass

determinations were marked at their midpoint with filter-sterilized India

ink prior to placement on CP1 medium. Severed seedling apices and/or

nodes were retained as meristem clones (mericlones) and placed in rooting

medium (RM1, see Chapter 3) in 20 x 100 mm Petri plates or Magenta GA-7

culture boxes (Magenta Corp., Chicago).

After 3 weeks on CP1, marked segments were removed for fresh and dry

weight determinations. Segments that were to be maintained on CP1 medium

for 6 weeks were transferred to fresh CP1 medium at three weeks. After

4 or 6 weeks on callus proliferation medium, explant segments with

associated callus were transferred to a shoot induction medium, called SIM

(see Chapter 3). Adventitious shoots (somaclones) that formed on callused

explants were excised and placed on rooting medium (RM1) or minigrafted

on etiolated 'Carrizo' seedlings (see below).

Rooted adventitious shoots and meristem clones were transferred to

soilless potting mix and maintained in high humidity and moderate light

for several weeks before being transferred to a controlled environment

chamber where plants were grown prior to morphological evaluation.

Similar methods were used to produce twelve 'Valencia' mericlones, five

'Valencia' somaclones, one minigrafted 'Duncan' somaclone, one 'Meiwa'

mericlone, one shaddette SF23-1 mericlone, and three shaddette SF24-1

mericlones for use as genetically diverse controls (Table 4.1). One

seedling from a ribbed fruit chimera of 'Marsh' grapefruit (C. xparadisi)

was used as another genetically different, unifoliolate control. Seven

etiolated 'Carrizo' seedlings were decapitated between the cotyledons and








73

the first node and allowed to regrow by adventitious shoots as additional

trifoliolate controls. All plants for this study were grown in 153 ml of

a soilless potting mix (SPM2) containing peat moss, vermiculite, and

perlite (Fafard Mix No. 2: Conrad Fafard, Inc., Springfield, MA) using

pots 20.5 cm long and 4 cm inside diameter at the top.





Table 4.1. Clones used in somaclonal variation studies.


Approximate
number measurements
Clone Root Number per plant per
Cultivar type stock plants characteristic

Carrizo Seedling Self 7 6

Carrizo Mericlone Self 4 6

Carrizo Somaclone Self 47 6

Duncan Somaclone Carrizo 1 5

Hamlin Mericlone Self 14 5

Hamlin Somaclone Self 29 5

Hamlin Somaclone Carrizo 8 5

Marsh Seedling Self 1 5

Meiwa Mericlone Self 1 3

SF23-1 Mericlone Self 1 5

SF24-1 Mericlone Self 3 6

Valencia Mericlone Self 12 5

Valencia Somaclone Self 3 5

Valencia Somaclone Carrizo 2 5








74
'Carrizo' seedlings used for minigrafting were grown in SPM2 and
covered with tall unlighted tubes for maximum nonchlorophyllous shoot

elongation and then decapitated by a horizontal cut between the cotyledons

and first leaf node (1-5 cm above surface of SPM2). Shoot tips 1-3 mm in

length were excised from the explant and immediately placed on the

decapitated 'Carrizo' stock so that the cut surfaces of the two units were

in contact (shoot tips were much larger than those used in previously-

described micrografting [Nauer et al., 1983; Navarro et al., 1975; Oiyama

and Okudai, 1986; Tusa et al., 1978]). The stems of the etiolated

'Carrizo' seedlings and the minigrafted adventitious shoots were of

similar diameter.

After transfer to SPM2, all mericlones, somaclones, and minigrafted

somaclones (also the 'Marsh' and 'Carrizo' seedlings) were grown in a

controlled environment chamber with a diurnal cycle of 16 hr light at 30C

and 8 hr dark at 25C. Plants were watered every two days, alternating

between tap water and a solution of 20-10-20 fertilizer plus

micronutrients at 300 mg-1-' (Peter's Professional Peat Lite Special 20-

10-20 Water Soluble: W.R. Grace & Co., Fogelsville, PA). Insects were

controlled on three occasions by spray application of kinoprene (Enstar

5E: Zoecon Corp., Palo Alto, CA) plus oil or fluvalinate (Mavrik: Zoecon

Corp.) within 10 days of a cutback.

Evaluation of plant morphology. Plants were cutback and allowed to
regrow from axillary buds prior to the morphological evaluations so that

the transient effects of tissue culture would be minimized and all organs

would be the same age at the time of evaluation. Vegetative morphologies

of 'Hamlin' mericlones and somaclones, 'Valencia' mericlones and








75
somaclones, one 'Meiwa' mericlone, one shaddette SF23-1 mericlone, three

shaddette SF24-1 mericlones, one 'Duncan' somaclone, and one 'Marsh'

seedling were evaluated 6-9 weeks after plants were cut back 10 cm above
the soil level (or 10 cm above the graft union for the minigrafted

'Hamlin' somaclones and minigrafted 'Duncan' somaclone). Single shoots

were allowed to grow on each plant and any additional axillary budbreak

was removed. Characteristics measured during the first evaluation were

leaf length, leaf width, internode length, spine length, spine diameter

(at the spine midpoint), abaxial oil gland density, abaxial leaf color,

and adaxial leaf color. Leaf, internode, and spine lengths, as well as
leaf width, were measured with a steel ruler in millimeters. Spine

diameter was measured by a dial caliper to the nearest thousandth of an

inch and converted to the nearest tenth of a millimeter for calculations.

Abaxial oil gland density was estimated by averaging counts from a

microscope-video camera image of two 20 mm2 fields per leaf. Abaxial and

adaxial leaf colors were determined on the standard X, Y, and Z color axes

by a HunterLab Citrus Colorimeter model D25 (HunterLab, Reston, VA)

modified with a black-plastic faceplate containing a 19 mm circular

aperture. Conversions of color measurements to the L, A, and B visual

color axes were completed using standard mathematical formulae issued by

HunterLab and similar to those described by Ting and Rouseff (1986).

Morphological characterization, in most cases, was based on five nodes
from each plant, beginning with the first node more than 3 cm above the

start of the regrowth after the cutback and including each of the next 5

consecutive nodes. Organs obviously damaged or within 3 cm of a shoot

apex were not included.








76
A second morphological evaluation of 'Hamlin' mericlones and
selected somaclones was completed 14-16 weeks after a third shoot cutback.
Single axillary shoots were allowed to grow on each plant, and any

additional axillary budbreak was removed. Characteristics measured during

the second evaluation were leaf length, leaf width, spine length, spine
diameter (at the spine midpoint), abaxial leaf color, and adaxial leaf

color. The second morphological characterization, in most cases, was

based on five nodes from each plant, beginning with the first node more
than 3 cm above the start of the second growth flush after the third
cutback and including each of the next 5 consecutive nodes. Organs
obviously damaged or within 3 cm of a shoot apex were not included.

Morphological evaluations of 'Carrizo' mericlones, somaclones, and
adventitious seedlings were accomplished 20-22 weeks after the first

cutback (for leaflet length and chirality) or 19-20 weeks after the second

cutback (for spine length and diameter). Characteristics measured were
chirality (bud to the left or right of the spine at each node), left,

main, and right leaflet lengths, spine length, and spine diameter (at the

spine midpoint). Spine L/D ratio and leaflet ratio (left plus right

leaflet lengths divided by main leaflet length) were calculated.

Characterization of leaflet length and chirality were based on six nodes

from each plant, beginning with the first node more than 8 cm above the

start of the regrowth after the first cutback and including each of the
next six consecutive nodes. Characterization of spine length and diameter

was based on six nodes from each plant, beginning with the first node more
than 8 cm above the start of the second growth flush after the second








77
cutback and including each of the next six consecutive nodes. Organs

obviously damaged or within 8 cm of a shoot apex were not included for any

measurements.

Irradiation of embryoqenic cultures. Irradiation of embryogenic

callus cultures was investigated as a method of inducing a higher
frequency of mutations among plants from tissue culture. All available

embryogenic callus suspension lines were obtained from J.W. Grosser,

including 'Hamlin', 'Valencia', 'Valencia Rohde Red' (Stewart et al.,

1975), 'Ridge Pineapple' (C. sinensis), and 'Milam' (C. jambhiri Lush.).

Suspension cultures 7 days old from each cultivar were plated out onto EME

medium, composed of solidified MT basal (Murashige and Tucker, 1969) with

50 g-1-' sucrose, 0.5 g-1-1 malt extract and adjusted to pH 5.7, in 15 or

20 x 100 mm Petri plates. After 18 days, plates were exposed to 0, 3, 6,

12, or 24 krad from a Cobalt-60 source. Callus or embryos were

transferred to EME supplemented with 20 ml1-' coconut water about 12

weeks after irradiation. Germinating embryos were transferred to SPM2 and

established in growth chambers.

Irradiation of organogenic cultures. Irradiation of organogenic

explants or callus cultures was investigated as a method of inducing a

higher frequency of mutations among plants from tissue culture. 'Carrizo'

and 'Hamlin' were chosen for the same reasons described in the somaclonal

variation project. In this study, 'Ridge Pineapple' was substituted for

'Valencia' because the latter had proved very difficult to regenerate by

organogenesis. The hybrid US-119 ([P. trifoliata x C. xparadisi] x C.

sinensis) also was included as a trifoliolate type (and thus perhaps very

responsive to culture) that might benefit by disruption of some metabolic








78

pathways affecting flavor components (one of the types of mutations likely

to occur as a result of irradiation). US-119 is of interest as a cold

hardy and relatively good-tasting scion type that could still benefit by

the loss of some remaining trifoliolate flavor characteristics.

Seeds of 'Carrizo' citrange, 'Hamlin', 'Ridge Pineapple', and US-119

were surface sterilized by immersion in 70% ethyl alcohol for 10 minutes

followed by a 1.05% solution of sodium hypochlorite (20% bleach) plus 2

drops Tween per 100 ml for 20 minutes. Disinfested seeds were then rinsed

in sterile distilled water 5 times (5 minutes each rinse), and individual

seeds were placed on the surface of 10 ml germination medium solidified

in the bottom of 25 x 150 mm glass culture tubes covered with Kap-uts.

All seedling cultures were maintained at 27C under a diurnal cycle of 16

hr fluorescent light and 8 hr dark. After seeds germinated, 7 mm segments

were excised from the stems and placed flat on the surface of CP1 medium

in 20 x 100 mm Petri plates for callus induction. All callus cultures

were maintained at 27C and under continuous fluorescent lighting.

Selected segments were weighed prior to irradiation and exposed to 0, 3,

6, 12, or 24 krad from a Cobalt-60 source six days or 41 days following

explant placement on CP1 medium. After six weeks on CP1, explants and

associated callus were transferred to SIM medium in 20 x 100 mm Petri

plates. Combined callus and explants of material irradiated 6 days after

explant placement were weighed 36-38 days after irradiation. Health of

callus cultures was scored about 70 days after irradiation using a

subjective five level rating scale for organogenic potential: 1 = dead,
2-4 = intermediate, and 5 = adventitious shoots present.








79

Results and Discussion
Subjective observation did not reveal differences between epicotyl

and internodal segments in callus growth, adventitious shoot formation,

or subsequent rooting of these shoots. However, substantial differences

were evident in the responses of explants from the different cultivars.

Explants from 'Carrizo' and the shaddette selections SF23-1 and SF24-1

produced callus more rapidly and shoots more profusely than explants of

'Hamlin' and 'Valencia'. Shoots of all these selections rooted to some
degree in RM1 medium. 'Duncan' was distinguished by the production of

large amounts of callus and adventitious shoots but an exceptional

recalcitrance to rooting on RM1.

Callus growth from explants of all the cultivars examined exhibited

considerable polarity, and usually the basal portion of each 1 cm segment

produced substantially greater amounts of callus than the apical portion

of the segment. An experiment was conducted to measure this differential

callus growth. Fresh weight of 1 cm 'Carrizo' segments was found to be

equally distributed between the apical 0.5 cm and basal 0.5 cm (N=12) when

measured immediately following excision from the in vitro seedlings (Fig.

4.1). When maintained on CP1 medium, a much larger amount of callus

growth occurred on the basal half of the explant as compared with the

apical portion. The average fresh weight of the tissue that developed

from the basal half of the explant was about five times greater than that
from the apical half after 21 days on CP1 medium (N=8). Dry weight of

these segment halves at the end of 21 days on CP1 averaged 19.4 mg for the

basal section and 5.4 mg for the apical section. The polar growth

response was similar regardless of the segment length or how many pieces








80





I Aplcal5mm Ei Baal mm

Day 0





Day 21



0 20 40 60 80 100
Fresh Weight (mg)


Figure 4.1. Differential callus growth in apical and basal
portions of 1 cm 'Carrizo' segments on CP1 medium.





into which a single segment was divided. This effect may be the result

of differential hormone (e.g. 2,4-D) uptake from the medium by the apical

and basal ends of the explant. Although explants were oriented

horizontally on the surface of the medium, there may have been a

significant effect of xylem or phloem conductance polarity as determined

during seedling development. Similar polarity of callus proliferation was

noted in explants from 'Hamlin' and 'Valencia', although the total volume

of callus produced and the differential was considerably less.

Shoot induction from callused explants was never observed on CP1

medium and typically did not occur until three or more weeks after

transfer to SIM medium. This was probably a reflection of the need for








81
the hormone BA in the medium for shoot induction (Duran-Vila et al.,

1989). Sporadic adventitious shoots were obtained from callused segments

of 'Hamlin' and 'Valencia' on SIM medium, but most explants from these

cultivars produced no adventitious shoots, even after 6-12 months on SIM

medium. Although callused explants from 'Carrizo' and shaddette SF24-1

did not always produce adventitious shoots, they were more likely to

produce shoots than 'Hamlin' or 'Valencia'. In addition, organogenesis

from responsive individual explants of 'Carrizo' or SF24-1 was not

sporadic, because scores of adventitious shoots could typically be

obtained from the same explant. The 130 rooted 'Carrizo' somaclones
obtained during this study were all produced by four explants from one

nucellar 'Carrizo' seedling. These explants rapidly regenerated new

shoots after each harvest and remained healthy and prolific until the

experiment was terminated 12 months after first shoot induction. The

nature of shoot induction on 'Carrizo' and shaddette explants suggests

that each shoot may not be produced by an individual adventitious

initiation event. Although initial shoot production must be regarded as

adventitious because such shoots did not develop from preformed buds, the

prolific nature of shoot induction from certain parts of the explant may

result from the development of new meristematic regions in the callus

material. Numerous buds and shoots may then develop from these meristems

without additional adventitious events. If this is the case, the

possibility for derivation of somaclonal variation from such regions is

probably significantly less than it would be if each shoot developed from
a separate adventitious initiation event. The formation of such an








82
"adventitious meristematic region" would, however, increase the potential

of these techniques for micropropagation and/or genetic transformation.

Rooting of shoots on RMI was easily achieved for 'Carrizo',

shaddette SF23-1, and 'Hamlin', but was considerably slower for 'Valencia'

(see Fig. 3.2). Different rooting media compositions were not evaluated

for increased efficacy in this study, but the results of Duran-Vila et al.

(1989) indicated that higher NAA concentrations may produce better results

for unresponsive genotypes. A high percentage of the rooted shoots

survived transfer to soilless potting mix (SPM2), especially if shoots and

leaves were pruned prior to planting and high humidity was maintained in

the acclimation chamber for several weeks following transfer. However,

extending the period of high humidity or pruning roots prior to planting

was distinctly detrimental to plant survival and establishment.

Minigrafting short adventitious shoots directly from the callus mass
onto etiolated 'Carrizo' seedlings was investigated as a more rapid method

of regenerating somaclones and one that might overcome selective forces

eliminating certain types of mutants (e.g. mutations affecting rooting)

from the regenerated plant population. Some success was achieved in the

minigrafting of 'Hamlin' and 'Valencia' adventitious shoots, and ten of

the forty-three somaclones included in the evaluated population were

produced by this method. The one successfully regenerated 'Duncan'

somaclone was also produced by minigrafting. These plants are indicated

by a "Z" suffix in Tables 4.2 and A.1-A.17.

Many of the rooted adventitious shoots (somaclones) produced during
this study exhibited deviant morphology during in vitro culture and the

first few months of establishment in soilless potting mix. Elongated








83

leaves, extremely short internodes, and deformed shoots were observed

among a number of the 'Hamlin' somaclones during the in vitro rooting and

in vivo establishment phases. Similar aberrant morphologies, including

unifoliolate leaves, were observed in some 'Carrizo' somaclones during
these phases. To reduce the non-genetic (and ephemeral) alterations that

would interfere with identification of true genetic mutations,

morphological characterization was delayed until plants were well

established in soil and was completed on healthy shoots grown entirely

after a severe cutback in vivo. In addition, the principle control plants

were produced from the same seedlings (or supposed genetically identical

seedlings) by in vitro meristem cuttings (mericlones) that were rooted,

established in potting mix, and grown under conditions identical to those

used for the somaclones.

Morphological characteristics were chosen for use as indicators of
genetic variation among the clones (mericlones and somaclones) on the

basis of observed preliminary morphological abnormalities, convenience,

and a presumed diversity of genetic controllers. In general, no practical

techniques have been described for characterizing horticulturally and/or

economically important traits in individual young Citrus plants (although

evaluation of one such method was a goal of the research described in

Chapter 6). However, if somaclonal variation is caused by random

mutations, the frequency of mutations affecting several diverse

morphological characters should indicate the mutation frequency affecting

horticulturally important traits.

In the first morphological evaluation of the 'Hamlin' somaclones,
fourteen 'Hamlin' mericlones were used as the reference population. These








84

mericlones were derived from ten seedlings that in turn were obtained from

the fruit of one 'Hamlin' tree. These seedlings were expected to be of

nucellar origin and genetically identical because 'Hamlin' is highly

polyembryonic and has been observed only rarely to produce zygotic

seedlings. Twenty-three of the 'Hamlin' somaclones were produced from

these same seedlings or other seedlings obtained from the same tree at the

same time. The remaining fourteen 'Hamlin' somaclones were produced from

seedlings of a second 'Hamlin' tree considered to be morphologically

indistinguishable from the first.

Normal probability plots (quantile-quantile) and the Kolmogorov test

in the SAS univariate procedure indicated that the combined measurements

of the fourteen 'Hamlin' mericlones (60 to 70 repetitions for each trait)

were normally distributed for all of the evaluated characteristics

(Prob.>D was greater than 0.01 for each character; Prob.>D was greater

than 0.15 except for internode length, spine length, leaf length, and

spine ratio). Therefore, these measurements were combined to obtain the

"HAMLIN MERICLONE" used as the control in each of the statistical

comparisons.

Student's t-tests were used to compare the morphological

characteristics of the two subpopulations of 'Hamlin' somaclones that were

obtained from different seed sources (indicated as "HA" and "HX" in the

tables). Because all of the mericlones available for inclusion in the

HAMLIN MERICLONE were from the "HA" population, any genetic differences

between the two seed sources may have affected comparisons of the "HX"

somaclones with the "HA" mericlones. The two populations had similar

means (Prob.>t was at least 0.05) for all characters except leaf width








85
(Prob.>t 0.0023), abaxial leaf color B (Prob.>t 0.0383), and leaf L/W

ratio (Prob.>t = 0.0003). Comparison of the variances for each

characteristic in the two populations by the F statistic indicated that

there was no evidence for differences between the variances of the two

populations for most traits (H: Variances are equal, Prob.>F was at least

0.1). Although some mean differences were indicated, no alternative

control was available and the seed source was necessarily ignored in

subsequent statistical comparisons. However, this factor must be

considered as a possible uncontrolled source of variation that may have

influenced the characteristics of individual "HX" somaclones or the

resulting population statistics.

Student's t-tests were also used to compare the morphological
characteristics of the 'Hamlin' somaclones on their own roots (29 clones)

with the 'Hamlin' somaclones minigrafted on 'Carrizo' roostocks (8

clones). Differences in the effects of the two root types on

morphological characteristics of somaclones would influence determinations

of aberrant plants, and all of the mericlones included in the HAMLIN

MERICLONE were on their own roots. Comparisons of these two

subpopulations of 'Hamlin' somaclones by Student's t-tests indicated that

the means were different (at a = 0.05) for all of the morphological

characteristics except internode length, leaf length, and leaf ratio.

Root type significantly affected the morphological characteristics of the

somaclones (e.g. 'Hamlin' on 'Carrizo' rootstocks tended to produce

longer, thicker spines, and wider, less darkly colored leaves). However,

this factor was ignored in subsequent statistical comparisons because no

'Hamlin' mericlones on 'Carrizo' rootstocks had been produced during the








86

generation phase of the study. The effect of rootstock must, therefore,

be regarded as another uncontrolled source of variation in the statistical

comparison of clone morphologies.

Twelve 'Valencia' mericlones (grouped as a single genotype with 50

to 60 repetitions), one 'Duncan' somaclone, one shaddette SF23-1

mericlone, three shaddette SF24-1 mericlones (grouped as a single

genotype), one 'Marsh' seedling, and one 'Meiwa' mericlone were used in

comparisons with the HAMLIN MERICLONE for each of the morphological

traits. Duncan's mean separation (a = 0.01) indicated that at least one

clone was measurably different from 'Hamlin' in: adaxial leaf colors A,

B, L, and A/B ratio; abaxial leaf colors B, L, and A/B ratio; spine

length; spine diameter; spine length/diameter ratio; leaf width; leaf

length/width ratio; oil gland density; and internode length (Tables 4.2

and A.1-A.6). The consistency with which these genotypes are clearly

distinguished from 'Hamlin' gives an indication that genetic differences

do, in fact, lead to measurable differences in many of these traits.

However, it should be noted that the degree of genetic variation

represented within this reference group was greater than would be expected

to occur as a result of most somaclonal mutation. The cultivar 'Valencia'

is most closely related to 'Hamlin' and is only separated from it by four

characteristics with Duncan's test at a = 0.05.

The first characterization of each of the 37 'Hamlin' somaclones and

5 'Valencia' somaclones is summarized in the first half of each column in

Table 4.2. More complete results are presented in Tables A.7-A.12 where

a mean measured value is indicated when it differed significantly from

HAMLIN MERICLONE (Duncan's separation at a = 0.01). Analysis of variance








87

probability of a greater F value is indicated at the end of each table

column. More somaclones differed from HAMLIN MERICLONE than would be

expected by chance for most of the morphological characters. For example,

twelve of the forty-two somaclones differed from the mericlone in adaxial

color L by Duncan's separation at a = 0.05 (see Table A.7), while only 2-3

outliers are predicted by definition of the a = 0.05 test. However, in

general these aberrants were not distinctly different from all other

somaclones but seemed to represent the tails of a broader normal

distribution than that predicted by the mericlone controls. These

observations suggest either an increased genetic diversity or some

undefined nongenetic variation within the somaclone population.

In an attempt to determine whether the aberrations observed in these

somaclones were caused by genetic mutations or nongenetic factors, the

mericlones, the somaclones indicated to be aberrant during the first test,

and some of the "normal" somaclones were cut back and regrown twice before

a second morphological characterization. Measurements of oil gland

density and internode length detected no variants during the first test

(Table 4.2 and A.12.) and so were not included in the second

characterization. Some mericlones and somaclones failed to regrow after

the cutbacks and could not be reevaluated. For most of the characters,

there were substantial differences between the mean values of the HAMLIN

MERICLONE in the first and second morphological evaluations. This was

probably caused by advanced maturity and the increased stress resulting
from extended growth in relatively small rooting volumes. Nevertheless,

it seems reasonable to assume that the resulting alterations were similar

for each plant and should not have substantially affected relative








88

Table 4.2. Summary for morphological comparisons of HAMLIN MERICLONE
with 6 other cultivars and 42 somaclones. See Tables A.1-A.17 for
more complete data.

Adaxi Color Abaxi Color Spine Leaf In Oil
Clone A BLL R_ BLL LRLE DR2 LE WR3 Lgl

Val M 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Duncan 0 0 0 0 0 0 0 + + 0 + 0 0 -
SF23-1 + - 0 0 0 0 0 0 0 0 -
SF24-1 0 0 0 0 + 0 + 0 0 0 0 0
Marsh + + 0 0 + + + 0 0 0 0 0 0
Meiwa 0 0 + +

HA142E121 + 0 + - 0 0 0 0 0 0 0 0 0
HA142E215 00 00 00 00 00 00 00 00 0- 0- -- 00 00 00 0 0
HA142E222 0 0 0 0 0 0 0 0 0 0 0
HA4I221 0 0 0 0 0 0 0 0 0 0 0 0 0 0
HA5E121 0 0 0 0 0 0 0 0 0 0 0 0 0
HA5I122 0 0 0 0 0 0 0 0 0 0 + 0 0
HA6E114Z 0 + + 0 0 0 + 0 0 0 0 0 0 0 0 0
HA7E12Z 0- 0+ 0+ 0+ 00 0+ 0+ ++ 00 00 00 00 00 00 0 0
HA9E13Z 00 00 0000 00 00 +0 +0 00 00 00 00 00 0 0 0
HX1XI1 00 00 000000 00 00 00 00 -0 00 00 00 00 0 0
HX1X12 00 00 -0 -0 00 -0 -0 -0 -0 -0 -- 00 00 00 0 0
HX101E111 00 00 00 00 00 00 00 00 00 0+ 00 00 -0 00 0 0
HX115E12Z 0- 0+ 0+ 0+ 00 0+ 0+ 0+ 00 00 00 00 00 00 0 0
HX115E14Z 00 00 00 0000 00 +0 +0 00 00 00 00 00 00 0 0
HX15E11 0+ 00 00 00 00 00 00000 00 00 00 00 00 00 0 0
HX20E11 -0 +0 +0 00 00 +0 +0 00000 00 00 00 00 0 0 0
VA13E11 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
VA30E11 0 0 0 0 0 0 0 0 0 0 0 0 0 + 0 0
VA33E11 0 0 0 0 0 0 + 0 0 0 0 0 + 0 0
VX91E11Z 0 + + 0 0 0 + 0 0 0 0 0 0 0 0 0
VX91E12Z 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Other Hamlin Somaclones
5 other 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0 0
16 other 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

First and second columns in a category are for first and second test,
respectively. HM = HAMLIN MERICLONE; Val M = Valencia mericlone; HA or
HX = Hamlin somaclones; VA or VX = Valencia somaclones; SF = shaddette;
Z = minigrafted on Carrizo; Adaxi = adaxial leaf surface; Abaxi = abaxial
leaf surface; A, B, and L = those respective color axes as measured by
HunterLab colorimeter; RI = A/B; LE = length; D = diameter; R2 = spine
LE/D; W = width; R3 = leaf LE/W; In Lg = internode length; Oil Gl = oil
gland density. 0 = value not different from HM with a = 0.01; + = value
different and greater than HM with a = 0.01; = value different and less
than HM with a = 0.01.








89

comparisons. Several somaclones were distinctly separated from the

MERICLONE by adaxial leaf color during the retest (Tables 4.2, A.13, and

A.15). Significantly, the three somaclones with the greatest differences
in the second evaluation were indicated to be normal for the same

characteristics during the first evaluation (Tables 4.2 and A.7), and the

two somaclones that were most clearly distinguished from the MERICLONE for

these characteristics during the first test were indicated to be normal

in the second morphological evaluation. No somaclones were found to be

different from the HAMLIN MERICLONE for leaf length, leaf width, and leaf

shape during the second test. Two somaclones differing from the MERICLONE

for spine ratio during the first test, HA142E215 and HX1X12, differed

similarly for this same characteristic during the second test (Table 4.2).

One somaclone differing from the MERICLONE for abaxial leaf color ratio

during the first test (HA7E12Z) differed similarly during the second test.

These were the only three examples of confirmed morphological differences

between 'Hamlin' somaclones and the HAMLIN MERICLONE. Six somaclones

determined to be aberrant during the first morphological evaluation could

not be reevaluated because of poor growth. It cannot be definitively

determined whether the three somaclones that were confirmed

morphologically aberrant were genetically different from 'Hamlin'. Growth

to fruiting and/or extensive molecular characterization may produce more

conclusive evidence. These results indicate that the frequency of

detectable morphological mutants among 'Hamlin' somaclones produced by

these techniques was less than 8-24% (3-9 of 37).

Forty-seven 'Carrizo' somaclones were evaluated for morphological

characteristics and compared with the Z2 MERICLONE (two plants) derived








90

from the same seedling. Mericlones from two other 'Carrizo' seedlings (Zl

and Z3) and seven 'Carrizo' plants regenerated from decapitated seedlings

were used as additional controls.

The left, main, or right leaflet lengths and leaflet ratios of the

'Carrizo' somaclones were not observed to differ significantly from those

predicted by the Z2 MERICLONE. Only one to three of the 47 somaclones

were significantly different from the Z2 MERICLONE at a = 0.05, and one

or none was different at a = 0.01 (Tables 4.3, A.18, and A.20). It was

concluded that there was no evidence for additional variation in leaflet

lengths or leaflet ratios induced by the tissue culture process. In

addition, it should be noted that one of the two other mericlones tested,

the Z3 MERICLONE, was also separated from the Z2 MERICLONE (at a = 0.01)

by left and right leaflet lengths, indicating that morphological variation

among seedlings may be at least as great as morphological variation among

somaclones.

The number of 'Carrizo' somaclones with aberrant spine lengths and

spine L/D ratios were similar to those with aberrant leaflet lengths and

not different from those predicted by the Z2 MERICLONE distribution

(Tables 4.3 and A.19). In contrast, 19 of 41 somaclones (46%) were

separated from Z2 MERICLONE on the basis of spine diameter with a = 0.05

(Table A.19). However, the significance of this finding is questionable

because three of the six controls were also separated from the Z2

MERICLONE by Duncan's test on the basis of spine diameter. Perhaps the

number of repetitions of the control Z2 MERICLONE (six from one plant) was

insufficient to provide a good estimate of the morphological variability

observed within the clone. Nonetheless, there was no evidence of a








91






Table 4.3. Summary for morphological comparisons of
CARRIZO Z2 MERICLONE with 2 other mericlones,
7 seedlings and 47 somaclones. See Tables A.18-A.20
for complete results.

Leaf Spine Ch
Clone LL ML RL R4 LE D R2 Re

Z1 mericlone 0 0 0 0 0 0 0 0
Z3 mericlone + 0 + 0 0 + 0 0
Z11 seedling 0 0 0 0 0
Z12 seedling 0 0 0 0 0
Z13 seedling 0 0 0 0 0
Z14 seedling 0 0 0 0 0 + 0 0
Z15 seedling 0 0 0 0 0 0 0 0
Z17 seedling 0 0 0 0 0 0 0 0
Z19 seedling 0 0 0 0 0 0 0 0
Z2E1CB317 0 0 0
Z2E235 0 0 0 0 0 + 0 0
Z2E236 0 0 0 0 0 + 0 0
Z2E239 0 0 0 0 - 0
Z2E240 0 0 0 0 0 + 0 0
Z21153 0 0 0 0 0 + 0 0
Z21157 0 0 0 0 0 0 0
Z21166 0 0 0 0 0 + 0 0
Z21167 0 0 0 0 - 0
Other Z2 Somaclones
33 other 0 0 0 0 0 0 0 0
5 other 0 0 0 0 0

MZ2 = Carrizo 2 Mericlone; Z2 = Carrizo 2 somaclone;
LL= left leaflet length; ML = main leaflet length;
RL right leaflet length; R4 = (LL+RL)/ML; LE = length;
D = diameter; R2 = spine LE/D; Ch Re = chirality reversals.
0 = value not different from MZ2 with a = 0.01;
+ value different and greater than MZ2 with a = 0.01;
- value different and less than MZ2 with a = 0.01.








92
greater frequency of morphological variants among the 'Carrizo' somaclones

than among the population of 'Carrizo' seedlings.

The final morphological characteristic measured on the 'Carrizo'

somaclones was chirality at the nodes (i.e., whether the bud was to the

left or right of the spine). The actual chirality of a shoot was probably

not an important indicator of genetic variation, because it is known that

different seedlings of the same genotype, and even different growth

flushes of the same plant, vary in their direction of spiral (Schneider,

1968; Schroeder, 1953). However, mericlones were never observed to

possess nodes with different chirality in the same growth flush and

observations of such chirality reversals in some somaclones were
considered a possible indicator of morphological instability that was

probably not linked with genetic alterations. Numbers of chirality

reversals in six nodes (maximum reversals=5) were counted for the

'Carrizo' somaclones (Table A.20.) and compared to a table of binomial

distributions. Only two (of 47) somaclones had more chirality reversals

than that predicted at a = 0.05 (greater than three reversals). The

number of chirality reversals in the somaclonal population was not

different from expected.

The morphological evaluations of 'Hamlin' and 'Carrizo' somaclones

provided some evidence that more individual somaclones than expected were

outside the normal distribution predicted by the mericlones. This may
indicate the occurrence of mutations or a greater variance for some

characteristics in populations of Citrus somaclones regenerated by

organogenesis than in mericlone populations from the same explant

material. An increase in the variability of quantitative traits has been








93

reported for somaclones of other species (Daub and Jenns, 1989; Larkin,

1985; Ozias-Akins et al., 1989).

When individual 'Hamlin' variants were retested after a substantial

regrowth period, most of these somaclones were morphologically normal.

These observations indicate that at least part of the enhanced variability

in the somaclone populations was nongenetic variation of unknown origin.

The 'Carrizo' somaclone population contained more aberrant plants than

expected from the normal distribution predicted from the control mericlone

for only one character (spine diameter). The seedling 'Carrizo'

population demonstrated a similar frequency of deviation for the same

characteristic. The frequency of morphological variants among 'Carrizo'

somaclones may be no greater than the frequency of variants among nucellar

seedlings.

The extensive morphological evaluation of these 89 somaclones did

not provide evidence that the production of Citrus plants by the described

organogenic procedures produced frequent or striking morphological

variants or mutations. Minor genetic alterations may occur in individual

Citrus somaclones and rare mutations may alter gross morphological

features. One tetraploid from a confirmed diploid seedling was recovered

in a 'Hamlin' somaclone population produced during subsequent studies and

not included in the morphological analysis. Other Citrus plants

regenerated from tissue culture have differed morphologically from the

cultivar from which they were derived (J.W. Grosser, personal

communication). However, the results of the present study indicated that

such morphological variation in 'Hamlin' only occurred at a low frequency,

and in 'Carrizo' did not occur at a frequency greater than that observed




Full Text

PAGE 1

UNEXPLOITED GERMPLASM, NATURAL MUTATIONS, AND SELECTED IN VITRO TECHNIQUES FOR CITRUS CULTIVAR IMPROVEMENT By KIM DEAN BOWMAN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1990

PAGE 2

ACKNOWLEDGMENTS Many people have provided assistance, support, encouragement, or advice during the three years that I have been a student at the University of Florida and the Citrus Research and Education Center; For this I am eternally grateful. I especially wish to thank My wife, Concessa, for love and understanding; Frederick G. Gmitter, Jr., for financial support, advice, and encouragement; Gloria Moore, Jude Grosser, Paul Lyrene, and James Graham for thoughtful suggestions and ideas; Terri Zito for excellent photography and kindness; Mary Ahnger, Bob Sorrell, and Ben Lye for statistical and computational advice; Walter Kender, Pamela Russ, Russell Rouseff, J.L. Chandler, and Margie Wendell for friendship and assistance; And my parents, Gary and Shirley, for providing me with the opportunity. ii

PAGE 3

TABLE OF CONTENTS page ACKNOWLEDGMENTS ii KEY TO ABBREVIATIONS v ABSTRACT vi CHAPTER 1. INTRODUCTION 1 CHAPTER 2. REVIEW OF THE LITERATURE Introduction 7 Forbidden Fruit 7 Somaclonal Variation 16 Type of change -morphology 18 -polyploidy 20 -aneuploidy 20 -gross chromosomal 21 -simple genetic 21 -organelle DNA 22 -methyl ati on 23 Source of variation -preexisting 24 -dedifferentiation 28 -extended culture 29 -hormones in medium 30 Mechanisms for induction 31 Differences between reports 33 Value 37 Somaclonal Variation in Citrus 39 In Vitro Mutagenesis by Irradiation 40 Fruit Sector Chimeras 44 Evaluation of Resistance to Phytophthora 49 CHAPTER 3. REDISCOVERY OF CARIBBEAN FORBIDDEN FRUIT AND EVALUATION OF ITS SIGNIFICANCE FOR CITRUS BREEDING Introduction 54 Materials and Methods 55 Results and Discussion 58 iii

PAGE 4

page CHAPTER 4. EVALUATION OF ORGANOGENIC CITRUS TISSUE CULTURES AS A SOURCE OF GENETIC VARIATION FOR CULTIVAR IMPROVEMENT Introduction 69 Materials and Methods 71 Results and Discussion 79 CHAPTER 5. CITRUS FRUIT SECTOR CHIMERAS AND THEIR POTENTIAL VALUE AS A GENETIC RESOURCE Introduction 98 Materials and Methods 99 Results and Discussion 102 CHAPTER 6. RESPONSE OF STEMS FROM IN VITRO-GROWN SEEDLINGS TO PHYTOPHTHORA PARASITICA IN DUAL CULTURES Introduction 115 Materials and Methods 117 Results and Discussion 120 CHAPTER 7. SUMMARY AND CONCLUSIONS 132 APPENDIX 137 LITERATURE CITED 169 BIOGRAPHICAL SKETCH 216 iv

PAGE 5

KEY TO ABBREVIATIONS BA = N-(phenylmethyl)-l H-purin-6-amine; benzyl adenine. CP1 = Callus proliferation medium. EME = Medium for growth of embryogenic callus. GM1 = Medium for germination of seeds. GTF = Grandis-type forbidden fruit. Gy = Gray (a measure of radiation; 10 Gy = 1 krad). i .d. = Inside diameter. MT = Murashige and Tucker Citrus medium (Murashige and Tucker, 1969). NAA = 1-naphthaleneacetic acid. RM1 = Root induction medium. SCE = Sister chromatid exchange. SIM = Shoot induction medium. SPM2 = Fafard No. 2 (Conrad Fafard, Springfield, MA) soilless potting mix. SF = Shaddette-type forbidden fruit. 2,4-D = (2,4-dichlorophenoxy)acetic acid. 2,4,5-T = 2-(2,4,5-Trichlorophenoxy)propionic acid. v

PAGE 6

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy UNEXPLOITED GERMPLASM, NATURAL MUTATIONS, AND SELECTED IN VITRO TECHNIQUES FOR CITRUS CULTIVAR IMPROVEMENT By KIM DEAN BOWMAN December 1990 Chairman: Gloria A. Moore Cochairman: Frederick G. Gmitter, Jr. Major Department: Horticultural Science (Fruit Crops) Breeding of Citrus by sexual hybridization has not been an effective strategy for cultivar improvement because of long juvenile stages and widespread apomixis. Investigations of unexploited genetic resources or systems of evaluation may identify methods to increase the rate of genetic advance. One potential germplasm resource, two sources of genetic mutants, and one novel method of in vitro evaluation for disease resistance were examined during this study. The potential value of each project was critically reviewed. A diverse population of Citrus, known as forbidden fruit, was discovered in Saint Lucia. Historical records, morphological characteristics, and isozyme evidence indicated that this population was closely related to grapefruit. High degrees of zygotic embryony and diversity within the population may make it a useful source of breeding parents. vi

PAGE 7

Significant morphological variation was observed during initial growth among Citrus plants regenerated from callus cultures by organogenesis. Many of the aberrant plants were indistinguishable from controls after a second cycle of vegetative regrowth. The frequency of genetic mutation that could be measured as stable morphological change in young plants was too low to allow efficient mutant recovery. Fruit sector chimeras were observed among eight Citrus cultivars at a frequency of 0.01 to 0.27 percent in commercial packinghouses. Mutants were recovered from some sectored fruit as autotetraploid seedlings. Several kinds of sectors with potential value as sources of genetic mutations were described. In vitro response of etiolated seedling shoots to Phytophthora parasitica was measured as stem discoloration. The non-host, or hypersensitive, response observed in Poncirus trifoliata and Severinia buxifolia was clearly distinguished from the susceptible reaction of eight Citrus selections and hybrids. However, two Citrus x Poncirus hybrids with useful levels of field resistance to this pathogen exhibited a highly susceptible reaction in vitro. This technique may prove useful in the study of resistance mechanisms, but probably will not be effective for population screening. vii

PAGE 8

CHAPTER 1 INTRODUCTION The genus Citrus supplies several fruit crops of great importance in tropical and subtropical regions of the world. The species of greatest economic value as scions are Citrus sinensis (L.) Osbeck (sweet oranges), C. reticulata Blanco (mandarins), C. xparadisi Macfadyen (grapefruits), C. limon (L.) Burm.f. (lemons), C. aurantifolia (Christm.) Swingle (limes), C. grandis (L.) Osbeck (pummelos), and a number of interspecific hybrids, especially tangors (C. reticulata x C. sinensis) and tangelos (C. xparadisi C. reticulata) In addition, many other species (e.g. the sour orange, C. aurantium L.), close relatives (e.g. Poncirus trifoliata (L.) Raf. and Fortunella margarita (Lour.) Swingle), and hybrids (e.g. 'Carrizo' citrange, P. trifoliata x C. sinensis) are important as rootstocks or specialty crops (Hodgson, 1967; Morton, 1987; Swingle and Reece, 1967). This dissertation is focused on the genetic improvements desired for the world's citrus industries, including simplified cultural demands, improved fruit quality, increased yield, and reduced susceptibility to cold, diseases and pests. Achieving significant genetic improvement is hampered by several major obstacles: a long juvenile phase, widespread nucellar polyembryony, the heterozygosity of most clones, and the lack of screens for selection of favorable young plants. 1

PAGE 9

2 Citrus spp. are woody perennials with an extended juvenile period during which no flowering occurs. Citrus species vary considerably in the length of the juvenile phase. Limes generally require only 2-3 years from seeding to first fruiting, while sweet oranges may take 10 or more years before the initiation of flowering (Furr et al 1947; Sherman and Lyrene, 1983). The long life cycle delays evaluation of hybrid fruiting characteristics and is a major factor limiting breeding progress. In addition, most commercially important Citrus cultivars and species are highly polyembryonic (Frost and Soost, 1968). These selections produce seeds composed of multiple nucellar (apomictic) embryos. Nucellar embryos are generally more well developed and vigorous than the single zygotic embryo that may or may not survive in the polyembryonic clones. As a result, most seedlings from polyembryonic species are genetically identical to the maternal parent even when harvested from fruit produced by controlled pollination. This factor severely limits sexual hybridization within all Citrus except those species with monoembryonic selections (C. media L. [citrons], C. grandis, and part of C. reticulata) High heterozygosity has been noted in most species of Citrus (Cameron and Frost, 1968; Frost, 1926; Hagedoorn and Hagedoorn, 1914) and, when combined with nucellar embryony and extensive vegetative propagation of desirable genotypes, has resulted in the accumulation of many deleterious recessive alleles in existing germplasm collections (Cameron and Frost, 1968). Hybridization among these heterozygous types frequently results in weak (Cameron and Frost, 1968; Frost, 1943; Torres, 1936) and/or highly variable progeny (Cameron and Frost, 1968; Swingle, 1913).

PAGE 10

3 The limitations placed on sexual hybridization by nucellar embryony, the unpredictability of the results of successful hybridization, and the long juvenile phase are problems that are exacerbated by the lack of methods for evaluation of important characteristics in non-mature seedlings (Soost and Cameron, 1975). Frequently, it is difficult to identify the few true zygotic seedlings that may be obtained from a large population of predominantly apomictic seedlings without waiting for fruiting or using expensive and time-consuming biochemical techniques (Bade et al in press; Scora and Kumamoto, 1983; Soost et al 1980; Torres et al., 1978, 1985). With this perspective, Citrus breeders have directed considerable attention toward alternatives to sexual hybridization for genetic improvement. The most notable successes have been in the fields of somatic hybridization (Grosser and Gmitter, 1989, 1990; Grosser et al., 1989; Kobayashi and Ohgawara, 1988; Ohgawara et al., 1989; Vardi et al., 1987, 1989) and somatic mutagenesis (Hearn, 1984, 1986; Hensz, 1971, 1977, 1985; Russo et al 1981; Spiegel-Roy and Kochba, 1973a; Spiegel-Roy et al., 1985; Starrantino et al., 1988a, 1988b). More recent interest has been directed toward the application of genetic engineering to Citrus genetic improvement. Although Citrus molecular manipulation is still in its infancy, the success of similar genetic engineering in other plant species (Deshayes et al 1985; Gasser and Fraley, 1989; Horsch et al 1985; Prols et al 1989; Sanford, 1988; Saunders et al 1989; Tomes et al., 1990; Weising et al 1988) has encouraged considerable optimism. Characteristics such as virus (Nelson et al., 1987; Powell Abel et al., 1986), insect (Fischhoff et al 1987; Vaeck et al 1987), and herbicide

PAGE 11

4 (Comai et al 1985; Shah et al 1986) resistance, as well as others (Hiatt et al 1989; Murai et al 1983; Ow et al 1986; Smith et al 1990; Weising et al 1988) have been engineered successfully into plants. The rapid advance in understanding of codon usage (Campbell and Gowri, 1990) and regulatory sequences (Benfey and Chua, 1989), along with the cloning of unique genes or determination of complete amino acid sequences for other proteins influencing cold tolerance (Cutler et al., 1989; Lee, 1989), sweetness (Inglett and May, 1968; Iyengar et al 1979; Kennedy et al., 1988; Ledeboer et al 1984), insect resistance (Fishman et al., 1984; Lazarovici et al 1984; Quicke, 1988), bacterial resistance (Casteels et al 1989), and other characteristics (Gasser and Fraley, 1989; Weising et al 1988) may indicate the much broader applicability of these techniques. There are already some examples of successful production of genetically transformed woody species (James et al 1989; McGranahan et al 1988) Unfortunately, each of these promising techniques appears to have significant constraints. Somatic hybridization offers many opportunities for useful genetic combinations, but will probably be limited by the inability to access desirable traits controlled by recessive genes or to eliminate dominant negative traits. In vivo somatic mutation of Citrus has proven effective in the induction of three specific types of useful genetic changes (i.e. increased pigmentation [Hensz, 1971, 1977, 1985], decreased acidity [Yen, 1987], and decreased seediness [Hearn, 1984, 1986]) and will probably remain confined in scope. Plant genetic engineering presents a number of ecological problems and social complications (Doebley, 1990; Ellstrand and Hoffman, 1990; Hoffman, 1990;

PAGE 12

5 Pimentel et al 1989; Raffa, 1989; Wilson, 1990) that may interfere with successful deployment of modified plants. In addition, specific molecular modification of plant genomes will most likely be restricted to characteristics under simple genetic control for some time to come, and many of the limits confining conventional breeding (e.g. greater energy spent on resistance to environmental factors = less energy to put into yield) will also restrain genetic engineering. With the exception of induced mutagenesis of mature budwood (in vivo mutagenesis) that avoids the juvenile phase, none of these methods addresses the often more significant problem of evaluating the "potentially" improved genotypes that they produce. Somatic hybridization and genetic transformation utilize plant regeneration from single cells and therefore revert to the juvenile stage. Considerable value may be derived from the development of additional approaches to genetic improvement of Citrus, both in terms of inducing genetic variation and evaluating germplasm. This dissertation is devoted to the investigation of four such areas: the Caribbean forbidden fruit, somaclonal variation/in vitro mutagenesis, fruit sector chimeras, and in vitro screening for resistance to Phytophthora. The Caribbean forbidden fruit is a recently identified, genetically diverse population of grapefruit-like Citrus that should provide valuable breeding parents with good fruit quality and a high degree of zygotic embryony. Induction of variation in tissue cultures (by somaclonal variation or applied mutagenesis) may provide alternative methods for inducing favorable somatic mutations in existing cultivars. Fruit sector chimeras are another source of somatic (as opposed to germ line) mutations that may

PAGE 13

6 allow effective selection of desirable mutations at the outset of experimentation, thus minimizing the high frequency of deleterious mutations commonly associated with induced mutagenesis. Finally, evaluation of clone resistance to Phytophthora in vitro could supply important information about resistance mechanisms. If correlated with field response, such a test could significantly expedite germplasm evaluation, especially in conjunction with in vitro methods of inducing genetic variation.

PAGE 14

CHAPTER 2 REVIEW OF THE LITERATURE Introduction As a background for the dissertation research described herein, a thoughtful review of the pertinent literature has been conducted in each of the subject areas. Appropriate contemporary and historical literature was examined to provide as much relevant information as possible on the Caribbean forbidden fruit/grapefruit, somaclonal variation, gamma irradiation of tissue cultures, fruit sector chimeras, and in vitro evaluation of Phytophthora resistance. Following are concise, but thorough and well documented summaries of these topics. Forbidden Fruit Citrus *paradisi (grapefruit) is the only widely recognized Citrus species that is considered to have arisen in the western hemisphere. Thorough searches of the Citrus regions in Asia and the Mediterranean have revealed no grapefruit, except those introduced from the new world (Hodgson, 1967; Kumamoto et al 1987; Scora et al., 1982; Swingle and Reece, 1967). Grapefruit is now considered to be an apomictically reproducing hybrid species that originated in the West Indies during the 17th or 18th century. On the basis of historical, morphological, and biochemical considerations, the most probable parents of grapefruit are 7

PAGE 15

8 C. sinensis and C. grandis (Chapot, 1950; Scora and Kumamoto, 1983; Scora et al., 1982). C. grandis (syn. C. maxima [Burm.] Merrell. and C. decumana L.) and C. sinensis were present in the West Indies by 1692 (Scora et al., 1982; Sloane, 1696). Grapefruit probably originated between this time and the 19th century by one or a series of natural hybridizations between the two species. However, a more complex scenario involving multiple hybridization events and other species is possible within the present state of knowledge. The grapefruits, as they are known in the United States citrus industry, are a group of morphologically similar and very closely related selections (Cooper, 1982; Hodgson, 1967; Ziegler and Wolfe, 1961). The most common cultivars in Florida are 'Marsh' and 'Ruby Red' (Florida Dept. Agr., 1988a, 1988b), two commercially seedless cultivars that differ principally in flesh and rind coloration (Hodgson, 1967). 'Ruby Red' was first identified in 1929 as a limb sport on one tree of 'Thompson' pink grapefruit by A.E. Henninger in Texas (Hodgson, 1967; Ziegler and Wolfe, 1961). 'Thompson' was itself selected as a limb sport of 'Marsh' in 1913 (Hodgson, 1967; Ziegler and Wolfe, 1961). Many other red-fleshed limb sports of 'Marsh' or 'Thompson' have been identified, but most were considered to be essentially identical with 'Ruby Red' (Waibel, 1953; Ziegler and Wolfe, 1961). Another cultivar that has achieved some popularity/notoriety is 'Star Ruby', a darker-red fleshed cultivar that originated from an irradiated seedling (probably mutated nucellar, not zygotic) of the cultivar 'Hudson' (Hensz, 1971, 1977). 'Hudson' was a limb sport of the cultivar 'Foster', that was in turn a limb sport of the cultivar 'Walters' (Hodgson, 1967). Three new darker red-fleshed

PAGE 16

9 cultivars ('Ray', 'Henderson' [or a nucellar seedling of 'Henderson' known as 'Flame' in Florida], and 'Rio Red') that are attracting tremendous attention in Florida, Texas, and California (Anonymous, 1988, 1989a, 1989b; Fairchild and Teague, 1987; Nauer et al 1988) were produced as natural or irradiation-induced bud sports from 'Ruby Red' (Hensz, 1978, 1981, 1985). The remaining commercial grapefruit cultivar is 'Duncan', a seedy, highly flavored selection that is now declining in significance and probably originated from a seedling planted in Florida about 1830. 'Duncan' was first introduced and propagated around 1892 (Hodgson, 1967; Ziegler and Wolfe, 1961). A grove of grapefruit at Safety Harbor, Florida, that was planted about 1823 by Count Don Phillippe either directly produced 'Duncan' (Robinson, 1947; Ziegler and Wolfe, 1961), or was the source of the seed that produced this cultivar (Hodgson, 1967). This same grove of grapefruit probably also was the source, directly or indirectly, for the cultivars 'Walters' and 'Marsh' (Cooper, 1982; Ziegler and Wolfe, 1961). Count Phillippe's grapefruit grove was planted by importation of seed or trees from the Bahamas (Robinson, 1952; Anonymous, 1921; Ziegler and Wolfe, 1961), Cuba (Webber, 1943), or somewhere else in the West Indies. There are no other reports of grapefruit germplasm importation into Florida, although Hume (1926) suggested that it may have been brought to the state by the Spaniards between 1513 and 1821. Other Citrus cultivars similar to grapefruit have been named (including 'Oroblanco' [Soost and Cameron, 1980a, 1980b], 'Melogold' [Soost and Cameron, 1985, 1986], 'Triumph', etc. [Hodgson, 1967]). Many of these are thought or known to be the result of hybridization between

PAGE 17

10 one of the aforementioned grapefruit cultivars and C. sinensis, C. reticulata, or C. grandis. None of these grapefruit hybrids has been enough like grapefruit to compete with 'Marsh', 'Ruby Red', or the other red-fleshed sports for commercial importance (Hodgson, 1967). Despite the uniqueness and commercial success of the popular grapefruit cultivars, they are not without their faults, including irregular fruit size, cold sensitivity, disease susceptibility, and excessive fruit bitterness. Richard Hensz pointed out that The improvements attributed to all of the new grapefruit cultivars that have developed since the beginning of commercial production have been towards enhancement for consumer acceptance with no real improvement in other horticultural characteristics of the tree or fruit. Presently grown seedless cultivars are no more vigorous, disease resistant, productive, nor have better eating quality than the white seedy types grown in the early days of the grapefruit industry. In fact, eliminating the seeds has resulted in the production of fruit of comparably smaller size and frequently lower yields. (Hensz, 1977; p. 582) The 'Star Ruby' cultivar (developed by R.A. Hensz) has been observed to be less resistant than less darkly pigmented cultivars to many environmental stresses, including Phytophthora foot rot, herbicides, and sunburning (Hensz, 1977). Genetic improvement of grapefruit has been frustrated by the long juvenile period (Frost, 1943; Furr et al 1947), high degree of nucellar embryony (Frost and Soost, 1968), and very limited variability among the known selections (Hodgson, 1967). Attempts to overcome these obstacles with existing selections have not been successful (Furr and Reece, 1946; Snowball et al 1988; Soost and Cameron, 1975; Wantanabe et al 1970)

PAGE 18

11 Genetic improvement of grapefruit would benefit by the identification of grapefruit selections with a shortened juvenile period, and greater zygotic embryony and genetic diversity. The loss of genetic diversity that results when population size is reduced to a small number of individuals has been recognized (Carson, 1970, 1971; Mayr, 1963; Nei et al., 1975; Prakash, 1972; Wright, 1931). Such a genetic "bottleneck" may have occurred in the formulation of current populations of North American grapefruit from the single Phillippe introduction (Bowman and Gmitter, 1990a). Many of the characters desired for grapefruit genetic improvement might be obtained from a diverse wild population, but no such wild grapefruit have been identified or examined as sources of additional germplasm. Maximum genetic diversity for crop species has frequently been associated with the center of species origin (Clement, 1989; Harlan, 1971; Hawkes, 1983; Vavilov, 1951). It is uncertain whether this trend would apply to species (like grapefruit) that originated as interspecific hybrids in the relatively recent past (200-300 years). Regardless, centers of plant diversity not associated with place of origin have been documented (Harlan, 1975; Peeters, 1988), and any existing wild populations of grapefruit in the West Indies would seem to be potential sources of unique alleles and genetic combinations. The origin of grapefruit has been discussed by several authors on the basis of the historical literature (Bowman and Gmitter, 1990a, 1990b; Kumamoto et al 1987; Robinson, 1952; Scora et al 1982). The species name now given to grapefruit, Citrus paradisi, was assigned in 1830 to a kind of Citrus from Jamaica identified with the common name of "Forbidden Fruit" (Macfadyen 1830, 1837). At that time, Macfadyen indicated that

PAGE 19

12 there were two botanical varieties of this species in Jamaica, one known as "Forbidden Fruit" and the other known as "Barbadoes Grape Fruit" (p. 304). Subsequent descriptions of Citrus in West Indian literature have typically used the names grapefruit and forbidden fruit as synonyms (Fawcett and Rendle, 1920; Freeman and Williams, 1927; Stehle et al., 1937; Stout, 1982). The most recent report of forbidden fruit as a distinct selection was from Bermuda (Britton, 1918), where it was described as a particular small fruited selection of C. grandis. About the same time as Macfadyen's application of the specific epithet C. paradisi, a resident of Jamaica, Maycock (1830), stated "My notes of the Grape-fruit and Forbidden-Fruit Trees, I am sorry to find, are too imperfect to enable me to say with certainty that they are specifically distinct, although I am inclined to think they are. I feel quite certain they are not varieties of C. decumana" (p. 318). Tussac (1824) also made a possible connection between the two selections when he described the Jamaican forbidden fruit as "fruit defendu, ou smaller schaddoc, petit chadec" (p. 74) with fruit borne in clusters like grapes Prior to 1830, the only other use of the name grapefruit in the literature was by John Lunan, who described a variety of Jamaican shaddock "known by the name of grape-fruit on account of its resemblance in flavour to the grape" (Lunan, 1814; p. 171). These early descriptions, therefore, suggest two possible origins for the name "grapefruit," the flavor and/or the fruit clustering. Jamaican forbidden fruit was described by several authors before Macfadyen (Browne, 1756, 1789; Lunan, 1814; Tussac, 1824), although they did not always agree on morphological characteristics. The first

PAGE 20

13 description of forbidden fruit was from Barbados, an island some 1800 km southeast of Jamaica (Hughes, 1750): FORBIDDEN-FRUIT-TREE The trunk, leaves, and flowers of this Tree, very much refemble thofe of the Orangetree. The Fruit, when ripe, is fomething longer and larger than the largeft Orange; and exceeds, in the Delicacy of its Tafte, the Fruit of every Tree in this or any of our neighbouring Iflands. It hath fomewhat the Tafte of a Shaddock; but far exceeds that, as well as the beft Orange, in its delicious Tafte and Flavour. This is delineated in Plate VII. (p. 127) Fourteen other kinds of Citrus found growing in Barbados at that time were described by Hughes, and one of these, the Guiney orange, he suggested should be called the "sour forbidden fruit." Hughes also produced a drawing of the forbidden fruit tree that showed pyriform fruit, no spines, and leaves without distinctively alate petioles. Hughes' depiction of the Barbados forbidden fruit tree without spines or winged petioles is at odds with the description of spines and alate petioles given by Browne (1756) and Macfadyen (1830, 1837) for the forbidden fruit of Jamaica. In addition, Macfadyen chose the pyriform fruit shape as a characteristic that distinguished the "Barbadoes Grape Fruit" from the "maliformis" Forbidden Fruit. Maycock (1830) indicated that Hughes had no botanical expertise, and we are left with uncertainty as to whether Hughes, in Barbados, was describing the same selection known as forbidden fruit by Macfadyen in Jamaica. Kumamoto et al (1987) indicated that they believed the golden orange described by Hughes was actually what later became known as grapefruit, and that "observers tended to confuse the grapefruit with the forbidden fruit, believing them to be the same fruit

PAGE 21

14 or varieties of the same species" (p. 100). Although some confusion may have occurred among these early authors, there seems little reason to question the close relationship between forbidden fruit and grapefruit agreed upon by Macfadyen (1830, 1837) and Maycock (1830) without examination of botanical specimens. Unfortunately, good herbarium specimens (personal communication, R.A. Howard, Harvard) or living selections of Jamaican and Barbados forbidden fruits do not exist. Kumamoto et al (1987) believed "the forbidden fruit tree became extinct, not having been mentioned after the 19th century, [while] the grapefruit flourished and spread throughout the West Indies" (p. 100). Both Hughes (1750), by his forbidden fruit and sour forbidden fruit, and Macfadyen (1830), by his two botanical varieties of forbidden fruit, indicated that there was considerable variability within the "forbidden fruit." This variation may simply be the result of normal intraspecif ic genetic diversity, not confusion. Controversy over what constitutes a taxonomic group within the genus Citrus has been common (Barrett and Rhodes, 1976; DeCandole, 1824; Green et al 1986; Hodgson, 1961, Hodgson et al., 1963a, 1963b, 1963c; Scora, 1988; Scora and Kumamoto, 1983; Swingle, 1914, 1943; T. Tanaka, 1927a, 1935, 1954, 1961, 1977; Y. Tanaka, 1948) and is obviously fueled by the nature of apomixis and the consequent lack of traditional species groups in agamic complexes (Crow and Kimura, 1965; Stebbins, 1950). However, the accepted interspecific origin (C. grandis x C. sinensis) and monoembryonic ancestry (from C. grandis) of grapefruit indicates that some genetic differences were probable among seedlings from identical parentage. This natural genetic diversity may be sufficient to

PAGE 22

15 explain the variation in forbidden fruit/grapefruit characteristics among descriptions by early authors. Although the grapefruit is generally believed to have originated in the Caribbean, there are some indications that forbidden fruit (and possibly grapefruit) may have been (or were derived from) one of the selections of Citrus previously known in the Mediterranean region as "apple of Adam" or "Adam's apple." Bonavia (1888) produced figures of a kind of Citrus fruit being sold in England as "forbidden fruit." He noted that this was not the same as the pummelo, and indicated that one specimen had very smooth skin, a solid center, and was "sub-acid and sweet, and slightly bitter" (Bonavia, 1888; plate XCII), while a second was smaller and pyriform. Bonavia (1888) also wrote (in reference to the forbidden fruit) "Gallesio [1811] says the Crusaders found the Porno d'Adamo in Palestine, and that it is not the Pompelmoess, the latter being a new citrus introduced from the East Indies" (plate XCII). Many selections of Citrus known as "apple of paradise" (Ferrari, 1646; Tolkowsky, 1938) or "apple of Adam" (Ferrari, 1646; Risso and Poiteau, 1872; Tolkowsky, 1938; Volkamer, 1708-1714) were described in the early citriculture literature from the Mediterranean, and one or more were probably introduced into the West Indies (Risso and Poiteau, 1872). Tolkowsky (1938) cites a Christian pilgrim who described about 1187 the "Adam's apple" growing in Palestine as "trees which bear fruit called Adam's apple (= the shaddock) wheron the marks of Adam's teeth may be right plainly seen" (p. 139). Tolkowsky also indicated that in the Middle Ages the citron and the shaddock were frequently known as "Adam's apple" or "apple of paradise," because of a Jewish tradition that considered the citron to

PAGE 23

16 be the forbidden fruit described in the religious accounts of the Garden of Eden (Tolkowsky, 1938). Somaclonal Variation In the past several decades, there has been interest in applying plant tissue culture systems to propagation and genetic improvement of many crops. Most propagation techniques have made use of axillary or other preexisting meri stems for proliferation of shoots and subsequent rooting. The propagation of clones in culture by way of preexisting meristems (mericlones) frequently has not resulted in significantly more variation than that observed with other methods of vegetative propagation (Conger, 1981). However, as more sophisticated techniques of plant culture, manipulation, and regeneration have been developed, it has been observed that some systems of culture may result in regenerates that are different in one or more significant characters from the original source. Larkin and Scowcroft (1981) have applied the term "somaclonal variation" to all such variability generated by plant cell culture and have suggested that this variation is of tremendous potential for plant improvement. Somaclonal variation has been the subject of many investigations and much speculation over the past decade (Brown and Lorz, 1986; D'Amato, 1985; DeWald and Moore, 1987; Evans, 1989; Evans et al 1984; Evans and Sharp, 1986a, 1986b; Lee and Phillips, 1988; Orton, 1984; Reisch, 1983; Wersuhn, 1989). Evans et al (1989) have been granted a United States patent for some methods of generating somaclonal variation. In evaluation of variation among plants regenerated from tissue culture, it is important to distinguish between transient (epigenetic or

PAGE 24

17 physiological) alterations in plant morphology and other more permanent (epigenetic or genetic) changes in the regenerated plants. However, the distinction between these two categories blurs somewhat in the cases of variation in methylation patterns (see below). Temporary changes of morphology and other characteristics have been commonly observed among plants regenerated from culture. Tissue culture propagation of strawberries produced temporary changes in many morphological characters (Marcotrigiano et al 1984; Sansavini and Gherardi, 1980; Swartz et al., 1981). Shoemaker et al (1985) observed an increase in susceptibility to root-rotting fungi among tissue culture-propagated strawberries. However, these plants reverted slowly to the normal resistant phenotype with increasing time out of culture, and the authors detected no evidence for genetic changes. The temporary changes in plants regenerated from tissue culture may rapidly revert to normal (Detrez et al 1989; Griesbach, 1989), or only after extended growth and/or cycles of vegetative propagation (Lourens and Martin, 1987). These temporary changes may be the result of stress or a physiological adaption to the culture conditions, such as the habituation of tobacco pith tissues (Binns and Meins, 1973). Generally, the temporary alterations in phenotype are not of significance for propagation or cultivar improvement, except when they impair the ability to discriminate true genetic mutants. Most published reports of somaclonal variation have attempted to eliminate these transient effects and focus on changes that appear to be genetic or at least relatively long term. Sugarcane was one of the first crops to obtain significant attention because of the clonal variation observed among regenerated plants.

PAGE 25

Selections were recovered from tissue culture with increased yield and sucrose production (Heinz and Mee, 1971), as well as resistance to eyespot disease (Heinz et al., 1977), Fiji disease, and downy mildew (Krishnamurthi and Tlaskal, 1974). Subsequent selection of somaclones from cultures of major commercial varieties has been reported to yield variants with favorable changes in some morphological or horticultural characteristic (Heinz et al 1977; Krishnamurthi, 1977; Larkin and Scowcroft, 1983; Liu and Chen, 1978; Maretzki, 1987). Many of these sugarcane clones have been tested over several years with no apparent loss of the acquired character (Orton, 1984), although other characters have exhibited some phenotypic instability (Maretzki, 1987). Similar success with somaclonal variation was also reported for potato, which is a polyploid, vegetatively propagated crop like sugarcane. Shepard et al (1980) were able to regenerate plants with changes in tuber skin color, tuber uniformity, maturity date, and numerous other characters from leaf protoplasts of a popular potato variety, 'Russet Burbank' Some of these variants were reported to be horticultural improvements over the original clone. Plants also were obtained with increased resistance to early blight and late blight {Phytophthora infestans). Vegetative propagation of these variants over several generations did not degrade the acquired characteristics. Type of change morphology Changes in many types of morphological or horticultural characteristics have been reported in somaclones, including leaf shape (Burk and Matzinger, 1976; Freytag et al., 1989; Osifo et al., 1989), foliage color (Osifo et al 1989; Taliaferro et al 1989), fertility (Daub and Jenns, 1989; Griesbach 1989; McPheeters and

PAGE 26

19 Skirvin, 1989; Orton, 1984; Taliaferro et al 1989), plant size (Eapen et al., 1989; Griesbach 1989; Jain et al., 1989; Taliaferro et al., 1989), growth habit (Freytag et al., 1989), seed color (George and Rao, 1983), time of flowering (Burk and Matzinger, 1976; Eapen et al 1989; OziasAkins, 1989), fruit number per plant (Eapen et al., 1989), cold resistance (Galiba and Sutka, 1989; Lazar et al., 1988), insect resistance (Miles et al., 1981), disease resistance (Behnke, 1979, 1980; Daub and Jenns, 1989; Gengenbach et al., 1977; Hartman et al., 1984; Heath-Pagl iuso et al., 1988; Heinz et al 1977; Latunde-Dada and Lucas, 1983; Orton, 1984; Sacristan, 1982; Shahin and Spivey, 1986; Thanutong et al., 1983; Toyoda et al., 1989), resistance to other factors (Grandbastien et al., 1989; Kuehnle and Earle, 1989; McHughen, 1987; McHughen and Swartz, 1984; Ranch et al., 1983; Schaeffer et al 1989; Shahin and Spivey, 1986; Wakasa and Widholm, 1987), alkaloid content (Burk and Matzinger, 1976), essential oil composition (Mathur et al 1988), nutritive quality (Schaeffer and Sharpe, 1987; Schaeffer et al 1989), and yield (Burk and Matzinger, 1976; Eapen et al 1989; Jain et al 1989; Mathur et al 1988; Secor and Shepard, 1981). Many reports of somaclonal variation do not indicate the genetic basis of the observed changes, although some alterations of the genetic material have been documented among somaclones. Genetic changes may be demonstrated by inheritance of the altered character among the sexual progeny, as was shown for somaclones of oat (Cummings et al., 1976), barley (Breiman et al., 1987b), and tomato (Evans and Sharp, 1986a, 1986b). In other cases, cytological, biochemical, or molecular studies have revealed a more specific genetic basis for the somaclonal variation.

PAGE 27

20 Type of change polyploidy Euploid increases in chromosome number among plants regenerated from tissue culture have been reported for many crops (D'Amato, 1977a, 1978), including bluestem grasses (Taliaferro et al., 1989), tobacco (Ogura, 1976), tall fescue (Eizenga, 1989), alfalfa (Saunders and Bingham, 1972) and wild rye (Park and Walton, 1989). Osifo et al (1989) produced callus from the cotyledons of a diploid potato selection and reported that 70% of the regenerated somaclones were tetraploid. Hexaploid and octoploid plants were found among the regenerates from protoplasts of two tetraploid potato cultivars (Karp et al., 1982), while Singh et al (1972) obtained haploid plants from cultures of a diploid legume species. However, examination of a large number of plants regenerated from haploid, diploid, and triploid lines of beet detected none with alterations in the original chromosome number (Detrez et al 1989). Interspecific hybrids have been noted to yield polyploids after tissue culture (D'Amato, 1985); Cappadocia and Ramulu (1980) obtained 21 amphidiploids among 57 regenerated plants from the hybrid between two species of tomato. Type of change aneuploidy Aneuploidy has been reported most frequently in somaclones from species that are polyploid (D'Amato, 1985); diploids are much more likely to become inviable as a result of loss or gain of one or a few chromosomes. Aneuploid somaclones have been identified in celery (Murata and Orton, 1983; Orton, 1983), alfalfa (Groose and Bingham, 1984; Johnson et al 1984), tobacco (Ogura, 1976), potato (Karp et al., 1982), Haworthia (Ogihara, 1981) and many grasses (Ahoowalia, 1983; Eizenga, 1989; Reed and Conger, 1985; Taliaferro et al., 1989). Eizenga (1989) observed 59 aneuploids among the 166 tall fescue

PAGE 28

21 somaclones she examined. Meiotic analysis of the tall fescue somaclones indicated that the aneuploids were monosomic, double monosomic, or triple monosomic but not nullisomic (Eizenga, 1989). Type of change gross chromosomal Alterations in gross chromosomal structure, or macromutations, have been found among plants regenerated from tissue culture, as well as among in vivo sports (Peloquin, 1981). Chromosome translocations, inversions, duplications and/or deletions have been reported in somaclonal variants from oat (McCoy et al., 1982), ryegrass (Ahloowalia, 1978), potato (Shepard, 1982), maize (Benzion and Phillips, 1988; Lee and Phillips, 1987) and other species (Armstrong et al 1983; Lapitan et al 1984; Lee and Phillips, 1988). McCoy et al (1978) noted the formation of tripolar divisions, ring chromosomes, and heteromorphic bivalents at meiosis of some plants from culture and considered non-homologous crossovers and deletions to be responsible for these abnormalities. Benzion and Phillips (1988) reported that over 12% of the somaclones regenerated from maize callus cultures during a 22 month period contained some type of cytological abnormality, and many of these were caused by the breakage of a chromosome between the centromere and a heterochromatic knob. Type of change simple genetic Changes in one or a small number of genes have been identified in somaclones by segregation among progeny from controlled crosses in wheat (Larkin et al., 1984), maize (Edallo et al., 1981), rice (Sun et al 1983; Fukui, 1983), soybean (Freytag et al 1989), tobacco (Prat, 1983), lettuce (Engler and Grogan, 1984), and twelve different variants of tomato (Evans and Sharp, 1983). Shahin and Spivey (1986) produced protoplasts from a tomato cultivar susceptible to Fusarium

PAGE 29

22 oxysporum f.sp. lycopersici race 2, and reported the regeneration of plants resistant to the pathogen both with and without in vitro selection pressure. Segregation patterns among progeny of these tomato somaclones indicated that some of the somaclones were heterozygous and some homozygous for a single dominant resistance gene (Shahin and Spivey, 1986). Relatively minor genetic mutations in some somaclones have also been suggested by changes in isozyme banding patterns (Allichio et al., 1987; Jackson and Dale, 1989; Orton, 1983; Taliaferro et al 1989) and restriction fragment length polymorphisms (Roth et al 1989). One maize somaclone with a variant Adhl allele was shown to have resulted from a single base substitution in the nuclear gene (Brettell et al 1986a). Some cases of chromosomal mutations leading to expression of recessive genes may result from the genetic nullification of an obscuring dominant gene, rather than from the de novo generation of the recessive gene (Little, 1989). Breiman et al (1987a) reported changes in the number of rDNA spacers in somaclones from one line of wheat after only a few weeks of culture, and claimed that the studies of Appels and Dvorak (1982) and Saghai-Maroof et al (1984) indicated equivalent changes should take many generations in vivo. Other workers have reported deficiencies in rRNA genes among somaclones from triticale (Brettell et al., 1986b) and potato (Landsmann and Uhrig, 1985). Type of change organelle DNA Changes in mitochondrial DNA (mtDNA) have been observed in somaclones from maize (Gengenbach et al., 1981; Umbeck and Gengenbach, 1983), sugar beet (Brears et al 1989), potato (Kemble and Shepard, 1984), and wheat (Hartmann et al., 1987, 1989;

PAGE 30

23 Rode et al., 1987). Somaclones were recovered from Texas male sterile lines of maize (susceptible to Helminthosporium maydis race T) with male fertility or resistance to the toxin of H. maydis (both characters are controlled by mitochondrial genefs]) (Brettell et al., 1980; Umbeck and Gengenbach, 1983). Brears et al (1989) reported a mtDNA change in a sugar beet somaclone. The mtDNA change was a specific reversion of the RFLP pattern in the male sterile source genotype to the pattern of the normal (fertile) genotype, although the somaclone remained male sterile (Brears et al 1989). Hartmann et al (1989) examined the mitochondrial genome of wheat somaclones and also observed specific rearrangements to other known mitochondrial types, as well as some novel mitochondrial DNA patterns. No rearrangements were observed in the chloroplast genome of sugar beet somaclones (Brears et al 1986, 1989), and it generally has been considered that the chloroplast genome rarely undergoes rearrangement. However, Evans and Sharp (1986) reported identification of changes in chloroplast DNA among some regenerated tomato plants. In gametoclonal studies, regeneration of plants from anther culture was reported to yield selections with deletions from the chloroplast DNA (Day and Ellis, 1984). Type of change methvlation Alterations in the methylation of DNA have been reported among cultured plant cells (Anderson et al., 1990; Quemada et al 1987). The level of DNA methylation has been suggested to influence or indicate the expression of some genes, with reduced methylation generally corresponding to greater expression (Amasino et al., 1984; Bird, 1986; Cedar, 1988; Flavell et al., 1986; Hepburn et al 1983; Holliday, 1987, 1989; Jablonka and Lamb, 1989; Watson et al 1987). The

PAGE 31

24 observation that treatment of plant tissues with a demethylating agent, 5-azacytidine, increases expression of some genes (Amasino et al 1984; Hepburn et al 1983) has shown that reduced methyl ation may promote gene expression. However, the relative degrees of methylation for rRNA genes in shoots, roots, and callus cultures of petunia differed widely and were not correlated with relative gene expression or growth (Anderson et al 1990). The fact that higher plants contain many copies of rRNA genes, and only a small portion of them are expressed at one time (Rogers and Bendich, 1987) may make methylation of rDNA a poor indicator of total relative gene expression. Quemada et al (1987) suggested that changes in methylation may be a cause of habituation in cell cultures. Nevertheless, changes in gene methylation have been proposed as a possible source of morphological variation in plants regenerated from tissue culture (Brown and Lorz, 1986; Lee and Phillips, 1988; Quemada et al., 1987). Patterns of methylation have been suggested to be heritable (Holliday, 1989; Jablonka and Lamb, 1989), but it is unclear whether methylation is a determiner or a consequence of gene activity. Generally, ribosomal genes in petunia shoots were highly methylated, those in callus cultures were variable in the degree of methylation (although the level of methylation for each cell line remained stable over time), and regenerated shoots typically returned to the degree of methylation observed in the original explants (Anderson et al 1990). Source of variation preexisting Evidence for morphological and genetic variation among plants regenerated from tissue culture is abundant. However, at least a portion of this variation is derived directly, without change, from the genetic and heritable epigenetic

PAGE 32

25 differences (methyl ati on?) that are common among cells within a single plant (D'Amato, 1975, 1977a). Cells in meri stems generally do not change in genetic composition over long periods of time (D'Amato, 1985), as demonstrated by the large number of species that have been propagated uniformly by meristem culture (Conger, 1981). However, many different tissues can be successfully used for culture initiation (Murashige, 1974), including differentiated tissue that does not have a high degree of genetic fidelity. Explants from non-meristematic parts of the plant have been considered most likely to yield variation in somaclones (Meins, 1983; Reisch, 1983). Nearly 90 percent of the flowering plant species studied have been found to exhibit some level of endoredupl ication of the DNA within the cells from differentiated tissues (D'Amato, 1977b, 1985). This results in an elevated level of ploidy that may frequently vary over a wide range within a single tissue (2x, 4x, 8x, or even higher). Plants with differentiated tissues that are composed of a mixture of cells with different ploidy levels are called polysomatic (D'Amato, 1985). Replication of chromosomes without separation of the chromatids (endomitosis) may occur to form polytene chromosomes, as in the cotyledons of pea (Karp and Bright, 1985). Other species, such as Crepis, have been found to maintain the same level of ploidy throughout their meristematic and somatic tissues (Brossard, 1978). Variation in chromosome numbers other than in genome sets (aneuploidy) is not common in vivo but has been observed in the meristems and throughout the tissues of several plant species (D'Amato, 1985). Vaarama (1949) studied the tremendous range in aneuploidy found in a

PAGE 33

26 tetraploid black currant (4x=32). Within the root tips, cells could be identified containing from 4 to 32 chromosomes. Sub-chromosomal nuclear and organellar variation also occurs in plant tissues undergoing differentiation (D'Amato, 1952, 1985), although it is frequently difficult to identify. Examples are generally limited to those that have a visual effect, such as the white and green leaf sectors that result from different chloroplast types (Poethig and Sussex, 1985). The color of stamen hairs in Tradescantia is controlled by a single gene that has been observed to mutate from heterozygous to homozygous recessive at a frequency greater than 0.04% under normal conditions (Dolezel and Novak, 1984). Some cases of variation in plants regenerated from culture have been directly related to the isolation of one or more layers from the chimeric source plant (Hall et al 1986a, 1986b, 1986c; McPheeters and Skirvin, 1989; Skeene and Barlass, 1983). Chromosomal crossing-over in somatic tissues has been observed among several crop species (Evans and Paddock, 1976) and is a mechanism (as is sister chromatid exchange [see below]) by which genetic variation can be produced among the cells within an individual. Much of the variation in plants regenerated from pollen or cells undergoing meiosis (sometimes called gametoclonal variation) probably results largely from the segregation of chromosomes and organelles in the meiotic processes and is best considered as a different phenomenon than somaclonal variation, although the two have frequently been discussed together (Evans and Sharp, 1986a; Evans et al 1984). An indication of the significance of explant source tissues to variation obtained in culture can be found by comparing cell populations

PAGE 34

27 or regenerates obtained from different tissues. Warden and Skoog (1973) reported tobacco pith cells varied in their DNA content, depending upon the height in the plant where the samples were taken. Plants regenerated from the cells lower in the stem had a different morphology from those regenerated from cells collected at the higher levels. Osifo et al (1989) observed a much greater frequency of polyploidy among potato somaclones regenerated from cotyledon-derived callus than those regenerated from leaf-derived callus. Feher et al (1989) reported that the chromosome number of tetraploid alfalfa somaclones regenerated by embryogenesis was directly related to the chromosome number of the original explant. Explants composed of somatic cells with 30 chromosomes produced only somaclones with 30 chromosomes. Aneuploid explants with 29 chromosomes and euploids with 2n=32 resulted in somaclones with a diversity of chromosome numbers, including a high percentage of mixoploids (Feher et al 1989). However, in some cases, even the presence of genetic diversity within explant tissues does not lead to genetic variation among regenerated plants. Polyploid cells are common in sugar beet leaf tissue, but no polyploid, aneuploid or chimeric plants were observed among a large number of somaclones regenerated via adventitious shoots from petiole tissue (Detrez et al 1989). The authors suggested that this was the result of selection against polyploid cells during the formation of in vitro adventitious buds. In some cases, the tissue culture process seems to select for genotypes that are rare in the explant tissue. Ramulu et al (1989) examined the ploidy levels of somaclones regenerated from protoplasts of monohaploid, dihaploid, and diploid potato selections. No haploid and

PAGE 35

28 only a low frequency of diploid somaclones were recovered from any of these sources. Most of the somaclones from all three ploidy sources were tetraploid (Ramulu et al., 1989). Haploid cultures of Crepis (Sacristan, 1971) and Datura (Furner et al., 1978) became diploid during extended periods of culture. Source of variation dedifferentiation At least a part of the variation observed among plants obtained from tissue culture may be produced or stimulated by the tissue culture process itself. One indicator of this is the changes in chromosome number that have been reported to occur in plant cell cultures. Studies by Binarova and Dolezel (1988) indicated that there was a slight increase in the frequency of cells at higher ploidy during initiation of embryogenic alfalfa suspension cultures, but frequencies reverted to normal after a short period in culture. Similar types of changes during culture initiation (and dedifferentiation) have been reported for alfalfa (Feher et al 1989), tobacco (Bennici and Caffaro, 1985), onion (Dolezel and Novak, 1985), and pine (Franklin et al., 1989). Furner et al (1978) observed that polyploidy was produced by the stimulation of premitotic endoredupl ication during the dedifferentiation process. Using explants with genetic markers, Barbier and Dulieu (1983) reported that most of the cultureinduced changes occurred during the first few cell divisions. Cionini et al (1978) studied the callus induction phase in broad bean, and observed that the DNA composition of cells sometimes underwent an irregular division process, called fragmentation. When mitosis and cytokinesis followed, the resulting cells exhibited various degrees of aneuploidy. Selective amplification of certain DNA sequences has been reported in one

PAGE 36

29 tobacco species during initial phases of dedifferentiation (Durante et al., 1983). Such differential DNA replication may affect fragmentation, mutation, and gene expression. Source of variation extended culture In addition to those changes induced during dedifferentiation, other variation may occur slowly over extended periods of culture (Barbier and Dulieu, 1983). Chromosome polyploidization may arise by endoredupl ication (D'Amato, 1985), fusion of spindles in cells with two nuclei (Mitra and Steward, 1961), or spindle failure (Bayliss, 1973). An increase in the frequency of cells at higher ploidy levels commonly has been reported for long term cell cultures (Bayliss, 1980; Berlyn et al 1986). A steady increase in ploidy level was observed for cell cultures of Coulter pine, so that six weeks after culture initiation about 80% of the cells had 4C DNA (diploid cells in G2, or tetraploid cells), and regenerated buds contained many cells at the 8C level (Patel and Berlyn, 1982). However, other studies with pine cultures have indicated relatively little change in cell ploidy over time (Franklin et al., 1989; Konar and Nagmani, 1972; Renfroe and Berlyn, 1984; Salmia, 1975), and there was no abnormality in the amount of DNA found in adventitious shoots regenerated from 3-6 month old spruce cultures (Hakman et al., 1984). Degree of aneuploidy may also increase in cultures with age. Murashige and Nakano (1965) noted the higher frequency of aneuploidy found in tobacco cultures as they grew older. Wheat somaclones obtained from long term cultures were reported to be more likely to have mitochondrial DNA rearrangements than plants obtained from short term cultures (Hartmann et al 1989).

PAGE 37

30 Some of the diversity in frequencies of somaclonal variation reported by researchers working on the same crop may be the result of specific differences in culture technique. Alfalfa plants regenerated from callus cultures were reported to be morphologically and karyologically normal by some workers (Binarova and Dolezel, 1988; Kao and Michayluk, 1980; Mezentsev, 1981), while other reports indicated considerable morphological and karyological deviance among somaclones (Groose and Bingham, 1984; Johnson et al 1984; Nagarajan and Walton, 1987) Studies by Binarova and Dolezel (1988) indicated that if short subculture intervals were used for alfalfa suspension cultures, the normal ploidy level and the capacity for embryogenesis were maintained. When the period between subcultures was extended to greater than seven days, there was an increase in the frequency of polyploid cells in the cultures and a corresponding loss in embryogenic potential (Binarova and Dolezel, 1988) This may be related to the differentiation process that begins once cell cultures reach the stationary growth phase. Cell differentiation frequently includes cell wall thickening, vacuolization, and DNA endoredupl ication (Kibler and Neumann, 1980). Source of variation hormones in medium Hormone composition of the media was reported to influence the degree of endoredupl ication in pea cortex cells during dedifferentiation (Libbenga and Torrey, 1973). Auxins like 2,4-D have been suggested to have an important effect on the induction and frequency of somaclonal variation, including ploidy level stability (Bayliss, 1980). Relatively high concentrations (at least 25 mg-T 1 ) of 2,4-D and 2,4,5-T (another auxin) were found to induce cytological abnormalities in onion root tips (Croker, 1953). Lazar et al

PAGE 38

31 (1983) observed that tissue culture medium with moderate levels of 2,4, 5-T (1 mg-1" 1 ) resulted in a higher frequency of aberrant plants than the same medium with a low 2, 4, 5-T concentration. However, Binarova and Dolezel (1988) noted considerable ploidy stability in alfalfa suspension cultures initiated on medium with a high level of 2,4-D, and other researchers have found that growth regulator components of the media had no effect on the karyological stability of cultures (Hanisch Ten Cate and Sree Ramulu, 1987). The relative ability of a chemical to induce sister chromatid exchanges (SCE) is considered to be a good indicator of mutagenic action (Latt, 1974; Perry and Evans, 1975; Uggla and Natarajan, 1982). Murata (1985, 1989) made use of this fact to examine the potential mutagenic effects of NAA, 2,4-D, 2,4,5,-T, and kinetin on wheat cell cultures. Concentrations of 2-5 mg-1" 1 of 2,4, 5-T significantly increased the frequency of SCE in the wheat cell cultures. However, NAA, 2,4-D, and kinetin were not observed to have any effect on the frequency of SCE, although inclusion of kinetin reduced the effects of 2, 4, 5-T. Mechanisms for induction One mechanism of somaclonal variation may be the formation of SCEs or some other types of mitotic recombination. Larkin and Scowcroft (1981) suggested that asymmetric crossing-over or recombination between homologous chromosomes could lead to some of the macromutations that had been observed in somaclones. Limited evidence for the occurrence of mitotic recombination among plants regenerated from tissue culture has been presented by several authors (Barbier and Dulieu, 1983; Dulieu and Barbier, 1982; Larkin et al 1984; Lorz and Scowcroft, 1983; McCoy, 1980).

PAGE 39

32 Chromosome fragmentation among plant cell cultures and somaclones has been observed to commonly occur in heterochromatic regions (Armstrong et al., 1983; Brettell et al., 1986; Lapitan et al 1984; McCoy et al 1982; Murata and Orton, 1984) or near the nucleolus organizer region (Lee and Phillips, 1987; Sacristan, 1971). Heterochromatin is known to replicate late in the mitotic cycle (Lima-De-Faria, 1969), so the formation of a bridge in the heterochromatic region and subsequent chromosome breakage has been proposed as one mechanism of somaclonal variation (Benzion and Phillips, 1988; Johnson et al 1987; Lee and Phillips, 1987, 1988; McCoy et al 1982; Sacristan, 1971). Nucleotide pool imbalances may produce genetic changes in unicellular organisms (Kunz, 1982; Kunz and Haynes, 1982) and animals (Weinberg et al 1981), including SCEs, aneuploidy, and macromutations in the nuclear and organellar genomes. Lee and Phillips (1988) have proposed that cell culture may cause critical imbalances in the nucleotide pools of plant cells that lead to similar genetic changes, and thus somaclonal variation. Another mechanism that has been suggested for the induction of somaclonal variation is the activation of transposable elements by stresses on the plant genome resulting from the tissue culture process (Chaleff, 1983; Evola et al 1985; Groose and Bingham, 1986a, 1986b; Larkin and Scowcroft, 1981; Peschke et al 1987) or some other specific stress-induced rearrangement process (Benzion et al 1986; Brettell et al., 1986b; Murata and Orton, 1983). Roth et al (1989) suggested that the RFLP changes observed in inbred soybean suspension cultures may have been caused by a mechanism that evolved to produce genetic variation in

PAGE 40

33 response to stress. The changes that were observed at all loci were almost entirely alternative alleles found in other soybean cultivars. It was suggested that this may indicate that a specific recombi national or rearrangement event occurred, such as gene conversion, transposon movement, or inversion. Equivalent programmed gene rearrangements have been documented in mammals and some unicellular organisms (Borst and Greaves, 1987). Breiman et al (1987a) proposed that the alterations they observed in intergenic spacers of regenerated somaclones were the result of tissue culture stress disrupting the mechanisms that normally maintain the stability of repeated DNA sequences. It is thought that repeated sequences play an important role in the adaption of the genome to change (Flavell, 1985). Peschke et al (1987) initiated cultures from maize explants without transposable element activity and reported the recovery of active Ac transposable elements in regenerated plants. Transposable elements have been best studied in maize, but are found in many other plant species as well (Grandbastien et al 1989; Groose and Bingham, 1986a, 1986b). Maize transposable elements can be activated by different types of shocks or stresses to the genome (McClintock, 1978, 1984), such as chromosome breakage (Freeling, 1984), chromosome rearrangements (Freeling, 1984; McClintock, 1978; Peterson, 1986), and alterations in DNA methylation (Chandler and Walbot, 1986). All three of these processes occur in plant tissue cultures (see above). Differences between reports Frequency of variants obtained from culture has varied widely in the different cases where it has been examined. Daylily has been observed to produce little somaclonal variation (Krikorian et al., 1981), and Griesbach (1989) was able to

PAGE 41

34 identify only one stable variant among 1000 daylily somaclones regenerated by organogenesis from long-term callus cultures. In contrast, Eapen et al (1989) reported alterations in mustard plant height, pod number, seed weight per plant, and yield in the second generation of nearly all plants regenerated from protoplasts when compared with the parental genotype. Frequently different researchers report entirely opposite results for the same crop and similar methods. Secor and Shepard (1981) examined 65 protoplast-derived somaclones of potato and determined that every one was statistically different from the source cultivar in at least one of the 35 characteristics measured (the authors indicated that this observation was quite significant, but the statistical significance is questionable), while a previous study (Wenzel et al., 1979) reported that there was no variation among protoplast-derived potato plants. Lourens and Martin (1987) observed that morphological variability among sugarcane somaclones was much less than previously reported (Heinz et al 1977; Heinz and Mee, 1971; Heinz et al 1969) when deviant somaclones were vegetatively propagated and examined after a second cycle of growth. Some of the discrepancies within the literature may be caused by confusion between temporary epigenetic changes and permanent genetic mutations. In addition, kinds and amount of variation are dependent upon the characters measured (Daub and Jenns, 1989). Three somaclonal variants for gliadin proteins in wheat (Maddock et al., 1985) were not observed to be deviant when examined for a variety of morphological characteristics (Maddock and Semple, 1986). In other cases, dissimilar results may be associated with large differences between species, or between cultivars within a species, in

PAGE 42

35 response to tissue culture (Bright and Jones, 1985) and in the frequency of somaclonal variation (Breiman et al., 1987; Brettell et al., 1986; Karp and Bright, 1985). Maddock and Semple (1986) speculated that the much lower frequency of somaclonal variation they observed (Maddock et al., 1985; Maddock and Semple, 1986) than Larkin et al (1984) among regenerated wheat plants was caused, at least partly, by cultivar differences. Binarova and Dolezel (1988) reported a high degree of karyological stability and maintenance of embryogenic capacity for one line of alfalfa in suspension culture, while Atanasov and Brown (1984) observed a considerable shift in cell ploidy and loss of embryogenesis in a second line maintained under similar conditions. Method of regeneration may also affect the frequency of variation among somaclones. Regeneration by adventitious shoots usually has produced a greater frequency of somaclonal variants than regeneration from axillary meristems (Conger, 1981; Larkin and Scowcroft, 1981). Organogenic regeneration generally has been considered to result in more somaclonal variation than embryogenesis (Armstrong and Phillips, 1988; Vasil, 1987), although this has not always proven true. Regeneration by organogenesis from long-term callus cultures resulted in very little variation in daylily (Griesbach, 1989), while regeneration of mustard plants by direct somatic embryogenesis was reported to produce over 95% somaclonal variants (Eapen et al., 1989). A study comparing the variation among maize plants regenerated by the two methods indicated that the frequency of variation was great for plants from embryogenesis, but even greater for those from organogenesis (Armstrong and Phillips, 1988). The process of plant regeneration may apply some selection against genetically

PAGE 43

36 aberrant cell types that survive in culture but are not viable in vivo (D'Amato, 1985; Vasil, 1988). Feher et al (1989) observed that embryogenesis occurred normally in alfalfa cultures containing high percentages of aneuploid and/or polyploid cells, but that the resulting embryos were mostly of the normal chromosome number. Polyploid cells are common in sugar beet leaf tissue, but no polyploid, aneuploid or chimeric plants were observed among a large number of somaclones regenerated by adventitious shoots from petiole tissue (Detrez et al., 1989); the authors suggested that this was the result of selection against polyploid cells during the formation of in vitro adventitious buds. As mentioned above, culture age may also affect variation among regenerated plants. The frequency of somaclonal variation typically has been observed to increase with aging of the callus from which plants are regenerated (Armstrong and Phillips, 1988; Lee and Phillips, 1987). No meiotic aberrations were observed among maize somaclones regenerated from cultures 3-4 months old, but nearly half of the somaclones regenerated from similar cultures 8-9 months old contained cytological aberrations (Lee and Phillips, 1987). Benzion and Phillips (1988) presented evidence that the increased frequency of cytogenetically abnormal maize somaclones from older cultures was "due to mutational events that occurred [at a constant rate] throughout culture development with a subsequent maintenance and accumulation of aberrant cells over time" (p. 318). However, callus age has not always resulted in a high frequency of variants; daylily callus more than 20 months old produced few somaclonal variants (Griesbach, 1989).

PAGE 44

37 Value Despite numerous reports of interesting or potentially useful mutants obtained by somaclonal variation, relatively little benefit has been evidenced in commercial varieties or breeding lines. One problem with some clones derived by tissue cultures, especially those derived by in vitro selection, is that what appear to be favorable changes in cell characteristics or even greenhouse plant response may not translate into improved field performance. A salt-tolerant somaclone of flax obtained by in vitro selection (McHughen and Swartz, 1984) was shown to be more productive than its parent cultivar under greenhouse conditions (McHughen, 1987). However, extensive field testing at both saline and nonsaline sites did not indicate that the salt-tolerant line was more productive than the parent cultivar under either condition (Rowland et al 1988, 1989). In the case of sugarcane, it has been suggested that the tremendous amount of variation already present in the germplasm renders the minor variation obtained from culture relatively insignificant (Maretzki, 1987). Another limitation is that mutations produced by somaclonal variation are random and, like all types of random mutations, are almost always deleterious or undesirable (Micke et al., 1987; Stadler, 1928a, 1928b). The novelty of the mutations obtained determines the relative value of any mutagenic method, including somaclonal variation. Some novel mutations have been reported among somaclones (Evans, 1989; Evans et al., 1989; Griesbach, 1989). However, the types of variation in potato somaclones (Sanford et al 1984), tobacco somaclones (Daub and Jenns, 1989) and wheat somaclones (Maddock and Semple, 1986) were not found to be different from those that occurred naturally under field conditions or

PAGE 45

38 were already known in germplasm collections. Sanford et al (1984) pointed out that sub-clonal selection in the field had been used for potato improvement (Davidson and Lawley, 1953; Easton and Nagle, 1981; Mcintosh, 1945) but was generally proven less effective than sexual hybridization. Gavazzi et al (1987) compared tomato mutants derived from somaclonal variation with those produced by ethyl methane sulphonate (EMS) treatment in vivo, and determined that somaclonal variation produced a higher frequency of variants. In addition, the types of mutations produced by somaclonal variation and EMS methods were different; Verticil Hum resistance was only recovered from EMS treatments (Gavazzi et al., 1987). Montagno et al (1989) compared somaclonal variation in tomato with that induced by mutation with ionizing radiation. The two methods produced similar types of mutations, although radiation (at 6.5 krad) produced heritable changes at more than twice the frequency of somaclonal variation. There is little doubt that somaclonal variation does occur among regenerates from many tissue culture systems. However, the ultimate value of somaclonal variation (or any method of genetic improvement) is determined by the relative expense (time and money) incurred and the relative genetic advancement achieved. Maddock and Semple (1986) have aptly pointed out that "where genetic variation per se can be obtained readily by standard sexual crossing procedures, random tissue cultureinduced variability may, therefore, not be of great value if large-scale screening of plant populations is necessary" (p. 1076). The relative value of somaclonal variation increases when sexual hybridization is difficult or concurrent in vitro selection is possible.

PAGE 46

39 Somaclonal variation in Citrus Somaclonal variation in plants regenerated by embryogenesis from Citrus has been evaluated by several researchers (Gmitter, 1985; Kobayashi, 1987; Navarro et al 1979, 1985; Vardi, 1977; Vardi et al 1982) with somewhat inconclusive results. One of nine plants regenerated from embryogenic orange cultures by Vardi (1977) was tetraploid, but the author did not indicate whether this plant was from normal or X-irradiated cultures. In later studies, Vardi et al (1982) reported that nucellar callus lines from seven Citrus cultivars maintained normal chromosome numbers, while an eighth nucellar callus line from 'Villafranca' lemon (C. limon [L.] Burm.) became predominantly tetraploid. Vardi et al (1982) indicated that plants with normal morphology were regenerated from each of these callus cultures. Navarro et al (1979) reported that regenerated sweet orange (polyembryonic) somaclones were morphologically normal. In contrast, Navarro et al (1985) identified about 28% of the somaclones as morphologically abnormal after regeneration from nucellus explants of 'Clementine' (a monoembryonic cultivar) by somatic embryogenesis. The observation of uniformity among somaclones obtained from the same explant in this latter study (Navarro et al., 1985) may indicate that the variation was preexistent in the explant and was not the result of the tissue culture process. Kobayashi (1987) examined morphological, cytological, and biochemical characteristics of 25 orange somaclones regenerated by embryogenesis from protoplasts and determined that they were "identical" to the nucellar seedlings. Gmitter (1985) identified morphologically abnormal somaclones among the plants regenerated by embryogenesis, but did not observe any

PAGE 47

40 cytological or isozymic variation, systematically examined Citrus populations No published regenerated via studies have organogenesis. In Vitro Mutagenesis bv Irradiation Mutation was first employed as a method of genetic improvement over sixty years ago (Muller, 1927; Stadler, 1928a, 1928b). Induced mutations do not necessarily duplicate natural genetic variation (Allard, 1960; Herskowitz, 1962; Stubbe, 1967). Most mutants will be lethal or unusable because they are clearly deleterious (Brock, 1971; Hansel, 1967). However, this has not prevented the utilization of mutagenesis as a breeding strategy; by the end of 1986, over 700 cultivars that were developed by mutation breeding had been released (Gottschalk and Wolff, 1983; Konzak, 1984; Micke and Donini, 1982; Micke et al 1985; Micke et al., 1987). Mutation in plant materials has most commonly been induced by chemical mutagens or radiation. Chemical mutagens were not utilized in this dissertation research (unless one considers 2,4-D as a mutagen; see above), but the types and usage have been discussed by several authors (IAEA, 1977; Kleinhofs et al 1974; Levy and Ashri, 1975; Moustafa et al., 1989; Neale, 1976). Radiation has many different effects on living tissues (Coggle, 1983; Dertinger and Jung, 1970; Wierbicki et al., 1986), and one of these effects is the induction of genetic mutations (Coggle, 1983; Lea, 1955; Savage, 1989; United Nations, 1986). High energy forms of radiation, including Xand gamma-rays, are typically known as ionizing radiation because of their ability to cause electrons to be driven from atoms which they contact. Both Xand gamma radiation have been investigated as

PAGE 48

41 methods of inducing mutations for crop genetic improvement (Broertjes, 1982; Eapen, 1976; Hearn, 1984, 1986; Hensz, 1977; Huitema et al 1989; Liu and Deng, 1985; Moustafa et al 1989; Russo et al 1981; Sharma et al., 1989; Van Harten et al., 1981; Wang et al 1988; Werry, 1981). The mutagenic effects of ionizing radiation on plant cells have been studied by cytology (Kuehnert, 1962), morphology (Alexander et al 1971; Kuehnert, 1962; Yu and Yeager, 1960), and isozymes (Gulin et al 1989), and were quantified by the measurement of sister chromatid exchanges (Kuglik et al 1987, 1989). Some mechanisms of gamma-induced mutation in plants may be similar to those described for yeast (Baranowska et al., 1987; Chepurnoi, et al., 1989) or may involve the induction of double stranded breaks and imperfect repair (Charlton et al., 1989). Irradiation for induction of mutations can be easily applied to plant material in vivo, but the resulting mutations have been frequently chimeric, complicating their identification and isolation (Broertjes, 1982; D'Amato, 1965). Irradiation of callus cultures may avoid these problems because embryos or adventitious shoots regenerated from tissue cultures often develop from single cells (Bhatia et al., 1986; Broertjes et al., 1968; Broertjes and Keen, 1980; Micke et al., 1987; Van Harten et al., 1981). The special advantages of in vitro mutagenesis when it can be used in combination with in vitro selection have been noted (Chaleff, 1981; Huitema et al 1989; IAEA, 1985; Ingram, 1983; Ingram and MacDonald, 1986; Werry, 1981), although unsolved problems (such as identifying in vitro selection methods that correlate with field responses) have generally remained obstacles to the successful utilization of these techniques (Lorz and Brown, 1986; Wersuhn, 1988, 1989).

PAGE 49

42 Low levels of ionizing radiation sometimes have been reported to stimulate plant growth or morphogenesis, a phenomenon termed radiation hormesis (Hell, 1983; Miller and Miller, 1987; Nuttall et al., 1968; Sagan, 1989; Sax, 1963; Sheppard and Chubey, 1990; Sidrak and Suess, 1973). This effect has not been observed in other cases after the same levels of irradiation (Montagno et al., 1989; Wolff, 1989). Whether the phenomenon of hormesis has any bearing on mutation induction is unclear. Sensitivity to irradiation has been observed to vary by a factor of up to 500 among different species (Balito et al., 1989; Sparrow and Woodwell, 1962), and differences also have been reported between tissues and organs of a single species (Bajaj et al 1970). Large nuclear volumes and rapid growth rates generally correlate with greater radiosensitivity, although higher ploidy levels tend to be more resistant (Balito et al 1989; Eapen, 1976; Galun and Raveh, 1975; Sparrow et al., 1961; Sparrow and Woodwell, 1962). Gamma irradiation of tissue cultures for genetic improvement has been investigated in several crops (Broertjes, 1982; Eapen, 1976; Moustafa et al 1989), although as Balito et al (1989) noted, only limited information is available on the effects of gamma irradiation on plant tissue cultures or regenerated plants. Moustafa et al (1989) observed that regeneration of maize plants from callus was considerably more sensitive to gamma irradiation than was the further growth of the callus. A dose of 100 Gy gamma radiation reduced maize callus growth about 50%, while a similar reduction in plant regeneration occurred with a treatment of only 25 Gy. Clones with increased flesh and rind color, fewer seeds, and decreased flesh acidity have been obtained by irradiation of Citrus

PAGE 50

43 budwood or nucellar seeds (Hearn, 1984, 1986; Hensz, 1971, 1977, 1985; Russo et al., 1981; Spiegel-Roy and Kochba, 1973a; Spiegel-Roy et al 1985; Starrantino et al., 1988a, 1988b; Yen, 1987; Zubrzycki and Diamante de Zubrzycki, 1982). Despite the striking achievements of in vivo irradiation relative to other approaches to Citrus genetic improvement, only limited effort has been devoted to irradiation of Citrus tissue cultures (Chang et al., 1984; Kochba and Spiegel-Roy, 1982; Legave et al 1989; Liu and Deng, 1985; Spiegel-Roy and Kochba, 1973b; Vardi, 1977). A dramatic stimulation of Citrus embryo formation from embryogenic cultures after gamma or X-irradiation has been reported by several workers (Nito et al., 1989; Spiegel-Roy and Kochba, 1973b; Vardi, 1977). Liu and Deng (1985; Chang et al 1984) treated organogenic callus cultures of several sweet orange cultivars with gamma radiation from a Cobalt-60 source. The LD-50 dose was determined to be 6-7 krad by percentage of callus survival, and gamma irradiated callus was noted to have a reduction in bud formation and bud growth when compared to controls. Increased frequencies of chromosome aberrations and polyploids were observed among the cells of the irradiated callus (Liu and Deng, 1985). Vardi (1977) reported an LD-50 of 3.4 krad X-radiation for sweet orange protoplasts and obtained several plants from the irradiated callus, including one tetraploid. Nito et al (1989) studied the effect of gamma radiation on callus growth and embryoid formation in cultures of 'Valencia' (Citrus sinensis [L.] Osbeck), 'Yoshida' (C. sinensis), Calamondin (C. madurensis Lour.), Yuzu (C. junos Sieb. ex Tan.), and 'Ishizuka Wase' satsuma (C. unshiu Marc), and reported LD-50 values of 10-20, 20-50, 20-50, 1-5, and 10-20 krad, respectively. These LD-50 values are similar to those

PAGE 51

44 reported by Spiegel-Roy and Kochba (1973b), but are difficult to reconcile with the values of 3-7 krad reported by other researchers (although a difference between embryogenic and organogenic cultures may contribute) for irradiation of Citrus callus cultures (Chang et al., 1984; Legave et al., 1989; Liu and Deng, 1985; Vardi, 1977). Fruit Sector Chimeras A chimera may be defined as an organism or organ containing two or more genetically distinct tissue types (Cramer, 1954; Tilney-Bassett, 1986; Vaughn, 1983). Chimeras have been described in many plant species (Chevreau et al., 1989; Pratt, 1983; Tilney-Bassett, 1986; Vaughn, 1983) and have been classified in a number of different ways. Although normal grafted plants are not considered to be chimeras because the two cell types are completely separated except for the graft union (Cameron and Frost, 1968; Neil son-Jones, 1969), chimeras may develop from grafted plants when the two dissimilar genotypes grow together and produce "hybrid" organs containing intermingled cells of both genotypes (generally segregated by histogenic layers). Most species of dicotyledonous plants (including Citrus [Cameron and Frost, 1968]) behave as though there are three histogenic layers in the apical meristem (Gifford, 1954). The two layers of the tunica (the L-I and L-II) generally form the epidermal and subepidermal (including gametophytic and nucellar) tissues, respectively (Pratt, 1983). The innermost layer of the apical meristem is called the corpus (L-I 1 1 ) and produces the central tissues of most organs including much of the vascular system (Pratt, 1983). The fruit of Citrus spp. are rather unusual in that

PAGE 52

45 the rind is derived predominantly from the L-II layer, but the flesh (modified epidermal hairs) is derived mostly from L-I (Cameron and Frost, 1968; Cameron et al., 1964). Chimeras produced by grafting are termed graft chimeras, or synthetic chimeras (T. Tanaka, 1927b; Winkler, 1907), and may be distinguished from those produced by nature, called autogenous (Cameron and Frost, 1968; Neil son-Jones, 1969). Natural chimeras are sometimes further classified according to type of genetic difference. Cytochimeras, or chromosomal chimeras, are composed of tissues of different ploidys (e.g. diploid and tetraploid) (Dermen, 1947; Einset et al 1947; Marks, 1953) or distinct cell karyotypes (Brumfield, 1943), while genie chimeras are composed of tissues with relatively simple genetic differences (Chevreau et al., 1989; Dayton, 1966; Einset and Pratt, 1959). Spatial distribution in the plant tissues (i.e. histogenic layer origin) is perhaps the most common method of classifying chimeras (Swanson, 1957). A periclinal chimera is formed when genetically different cell types are completely segregated by histogenic layers (e.g. L-I, L-II, or L-III) and all of each layer is of a single cell type (Dermen, 1960; Doodeman and Bianchi, 1985; Hall et al 1986a, 1986b; Jones, 1934; Marcotrigiano, 1986; T. Tanaka, 1927b). A chimera composed of a sector of tissue including a portion of all histogenic layers is termed a sectorial chimera (Dermen, 1947; Neil son-Jones, 1969; Varghese and Robbelen, 1984). The mericlinal chimera is distinguished from the other two types by the alternate genotype being found only in a part of one cell layer (usually the epidermis) (Jorgensen and Crane, 1927; Vaughn and Wilson, 1980). Chimeras composed of irregular combinations of the above types are known

PAGE 53

46 as mixed or mosaic chimeras (Cameron and Frost, 1968). Chimeras, especially periclinal chimeras, of some plant species have been separated into the component genotypes as non-chimeric plants by irradiation (Johnson, 1980; Pereau-Leroy, 1974; Sagawa and Mehlquist, 1957), tissue culture (Dommergues and Gillot, 1973; Johnson, 1980), root cuttings (Decourtye, 1987), or production of adventitious shoots (Dommergues and Gillot, 1973; Stewart and Dermen, 1970). Chimeras are common in Citrus, and one of the first plant chimeras ever described was the "bizzarria," a synthetic chimera produced by a failed graft of citron (C. medica L.) and sour orange (C. aurantium L.) about 1644 (Darwin, 1921; Gallesio, 1811; Nati, 1674, 1929; Risso and Poiteau, 1818; T. Tanaka, 1927b). Bud sports have occurred frequently on many Citrus cultivars (Bono et al., 1981; Devaux, 1981; Hensz, 1981; Iwamasa and Nishiura, 1981; Iwamasa et al 1981; Mendel, 1981; Russo, 1981; Shamel, 1943; Soost et al 1961) and probably originate as periclinal, sectorial or mericlinal chimeras. Mendel (1981) suggested "that varieties showing a tendency to produce bud mutations are chimeras themselves (e.g. the 'Shamouti' orange)" (p. 86). Many Citrus cultivars are known to be relatively stable periclinal chimeras (Cameron and Frost, 1968; Cameron et al 1964; Iwamasa and Nishiura, 1970). The pink-fleshed grapefruit 'Thompson' arose as a bud sport on a tree of the white-fleshed grapefruit, 'Marsh' (Hodgson, 1967). The cultivar 'Thompson' contains pink flesh (derived from L-I) but unpigmented rind (derived from L-II), while a budsport derived from it, 'Ruby Red', has both red flesh and pigmented rind. 'Thompson' nucellar seedlings (derived from L-II) produce only fruit with white rind and white

PAGE 54

47 flesh. Nucellar seedlings from 'Ruby Red' produce only red-fleshed fruit with pigmented rind. 'Thompson' is probably a periclinal chimera containing a genetic mutation for red pigmentation in the L-I layer. 'Ruby Red' is nonchimeric for the red pigmented cell type and probably arose as a layer substitution in 'Thompson' (Cameron et al., 1964; Soost and Cameron, 1975). Other Citrus chimeras that have been described include, 'Burgundy' grapefruit (Olson et al 1966), 'Foster' grapefruit (Cameron et al 1964), 'Golden Buckeye' navel orange (Shamel et al., 1925), 'Kobayashi Mikan' (Cameron and Frost, 1968; Iwamasa et al 1977; Samura and Nakahara, 1928), 'Gail iangcheng' orange (Lu, 1978, 1982), 'Suzuki Wase' satsuma (Iwamasa and Nishiura, 1970; Iwamasa et al 1977), and several variegated chimeras with white and green leaves (Cameron and Frost, 1968; Shamel, 1932). Nishiura and Iwamasa (1970) suggested that recovery of two color forms among nucellar seedlings of the satsuma selection 'Dobashibeni Unshu' may indicate that it is a sectorial chimera composed of orange-yellow and red pigmented components. Citrus cytochimeras have been observed with diploid, tetraploid, and octoploid layers (Barrett, 1974; Frost and Krug, 1942). Fruit sector chimeras with changes in rind color (Cameron and Frost, 1968; Coit, 1915; Frost, 1926; Iwamasa et al 1977; Nishiura and Iwamasa, 1970), rind thickness (Cameron and Frost, 1968; Coit, 1915; Frost, 1926; Shamel et al 1918), rind texture (Cameron and Frost, 1968; Shamel et al., 1918), flesh color (Cameron and Frost, 1968), ripening season (Coit, 1915), or rind injury resistance (W. Grierson, personal communication) have been observed from trees of many Citrus cultivars; sometimes

PAGE 55

48 individual trees produced the chimeras at a high frequency (Frost, 1926). Some 'Ruby' orange fruit with sectors having altered rind color were noted to contain adjacent pulp segments with similar color changes (Cameron and Frost, 1968). Coit (1915) noted that citrus workers commonly considered the sectored fruit to be the result of cross-pollination, but that such fruit were actually the result of a mutation in the tree or at the base of the ovary. Frost (1926) reported that the cause of some sectorial chimeras appeared to be "differential mitosis" or chromosome nondisjunction because occasionally chimeric fruit had "two adjacent sectors, of similar width, whose rind varies in opposite directions from the normal condition" (p. 394). Nishiura and Iwamasa (1970) suggested that the fruit sector chimeras observed on 'Dobashibeni Unshu' were an indication that this particular selection of satsuma was a sectorial chimera. The frequency of fruit chimeras with one or more thickened rind sectors, described as "ridging" or "coxcombing, was observed to be about 0.1% in navel orange, 0.1% in 'Valencia' orange, 0.2% for lemon, and 0.2% for grapefruit in normal California orchards (Sinclair and Lindgren, 1943). However, fumigation of Citrus trees with hydrocyanic acid (Lindgren and Sinclair, 1941; Sinclair and Lindgren, 1943) or chlorpyrifos (M.L. Arpaia, personal communication) during the early stages of fruit bud development (January or February) was found to cause a dramatic increase in the frequency of ridged fruit. Sinclair and Lindgren (1943) suggested that hydrocyanic acid fumigation may cause "a genetic change, such as a doubling of the number of chromosomes" (p. 104), in one or a small number of cells in the developing ovary, and that this mutation could then be propagated by mitosis to form the ridged sectors. Iwamasa et al (1977)

PAGE 56

49 isolated seeds from beneath the normal orange and mutant yellow sectors of one chimeric fruit of 'Fukuhara' orange. Trees that developed from seed borne in the orange sector produced unsectored orange-colored fruit, while those that developed from seed borne in the yellow sector produced unsectored yellow fruit. It was stated that a similar event had been previously observed by Iwamasa and Nishiura (1970). Iwamasa et al (1977) suggested that the nucellar seed and rind sectors had developed from the same L-II histogen and that sectors with improved fruit color might therefore be used to produce better cultivars. Evaluation of Resistance to Phytophthora Phytophthora spp. cause serious soilborne diseases of many important world crops. Several Phytophthora species can invade Citrus, including P. parasitica Dast., P. citrophthora (R.E. Sm.& E.H. Sm.) Leonian, P. hibernalis Carne, P. syringae Kleb., P. palmivora (Butler), and P. citricola Saw. The most important of these species in Florida is P. parasitica, which is responsible for root rot, foot rot, and gummosis (Timmer and Menge, 1988), as well as causing damping-off of young Citrus seedlings (Timmer, 1988). Citrus feeder roots are invaded principally by zoospores from infected soil through wounds or undamaged root tips. Zoospore germination and mycelial growth lead to a subsequent decay of the root cortex that may result in tree decline (Timmer and Menge, 1988). Phytophthora can invade the trunk of Citrus trees through openings in the bark produced by wounding or natural cracking. The ensuing foot rot or gummosis is characterized by necrosis of the cambium and rotting of the bark. Foot rot is the name applied when the disease affects the tree

PAGE 57

50 slightly above, at, or slightly below ground level, while the term gummosis is more frequently used when the disease occurs higher in the trunk. Trees affected by foot rot/gummosis may show symptoms of gum exudation, dieback, weak growth, and/or death (Lutz and Menge, 1986; Timmer and Menge, 1988). Phytophthora can be controlled in the nursery by hot water treatment of seeds, fumigation of soil, or soil drenches with appropriate fungicides (Timmer and Menge, 1988). Some degree of control in young and mature orchards is possible by fungicide application as foliar sprays of systemic compounds (Davis, 1981; Sandler et al., 1989), drenches (for root rot), or trunk paints (for foot rot and gummosis), as well as by sanitation (Sandler et al., 1989; Timmer, 1977; Timmer and Menge, 1988). The most economical control is planting resistant rootstocks with the bud union well above soil level. Cultivars used as rootstocks vary widely in their susceptibility to the diseases caused by Phytophthora (Broadbent et al., 1971; Carpenter and Furr, 1962; Graham, in press; Hutchison and Grimm, 1972; Klotz et al 1968; Klotz and Calavan, 1978; Smith et al., 1987; Timmer and Menge, 1988). Although a general correlation between relative resistance to root rot and foot rot has been observed for many Citrus selections, there are reports of substantial differences in response of some selections to the two diseases (Carpenter and Furr, 1962; Furr and Carpenter, 1961; Grimm and Hutchison, 1973, 1977). 'Carrizo' citrange and sour orange are tolerant to foot rot (Castle et al 1989; Timmer and Menge, 1988), yet were reported to be susceptible to root rot (Graham, in press). Relatively little is known about the mechanism(s) of resistance to Phytophthora in Citrus, although the involvement of coumarin

PAGE 58

51 phytoalexins has been suggested (Afek and Sztejnberg, 1986, 1988; Afek et al., 1986; Vernenghi et al., 1987). Isolate specificity has not been reported for Phytophthora within Citrus. However, Matheron and Matejka (1990) determined that isolates of P. parasitica isolated from tomato, petunia, and five other unrelated hosts were not virulent on rough lemon {Citrus jambhiri Lush.), and similar species specificity for isolates was observed by other workers (Wheeler and Boyle, 1971). Hutchison (1985) reported that resistance to P. parasitica in Citrus is probably controlled by a multiple gene system. Because of the great differences in characteristics of the affected tissues and etiologies (Lutz and Menge, 1986; Timmer and Menge, 1988), it seems probable that one or more mechanism(s) of resistance are not common to both root rot and foot rot. Several methods of evaluating resistance of young Citrus to root rot and foot rot have been described. Soil flooding with zoospore suspensions has frequently been used to evaluate susceptibility to root rot (Cameron et al., 1972; Carpenter and Furr, 1962; Grimm and Hutchison, 1973; Whiteside, 1974), although the results have not always been reliable (Carpenter and Furr, 1962; Furr and Carpenter, 1961; Grimm and Hutchison, 1973). Tsao and Garber (1960) reported that soil infestation with mycelium grown in liquid cultures produced reliable results. Chlamydospores may provide a more uniform source of inoculum (Farih et al., 1981; Graham and Egel 1988), and Graham (in press) described a procedure for chlamydospore inoculation and susceptibility assay. Controlled evaluations of foot rot/gummosis resistance have typically been completed by wounding trunk tissue and inoculating with agar disks from mycelial cultures (Klotz et al., 1968; Smith et al.,

PAGE 59

52 1987). General ratings of field resistance have commonly relied upon observations of foot rot/gummosis susceptibility because it is the most visible Phytophthora disease, despite a high degree of variability in damage observed at different times of testing (Smith et al., 1987; Whiteside, 1971). Although several different methods of evaluating susceptibility are available, screening Citrus plants for resistance to Phytophthora has proven to be time-consuming and relatively imprecise in rootstock development programs (Broadbent, 1971; Carpenter and Furr, 1962; Furr and Carpenter, 1961; Grimm and Hutchison, 1973; Hutchison, 1985; Smith et al., 1987). The development of a more rapid and reliable method for determination of Phytophthora resistance would be of great benefit to breeding programs. Tissue culture techniques have become increasingly important for programs involved in the genetic improvement of Citrus (Gmitter and Moore, 1986; Grosser and Gmitter, 1989; Grosser et al 1988a, 1988b, 1989; Hidaka and Kajiura, 1988; Kobayashi and Ohgawara, 1988; Kobayashi et al., 1983, 1988; Kochba and Spiegel-Roy, 1982; Kochba et al., 1972; Navarro et al., 1985; Ohgawara et al 1985; Sim et al., 1988; Vardi and Galun, 1988; Vardi et al., 1986b, 1987, 1989). Some success has been achieved in characterizing or selecting for plant resistance to diseases in vitro (Behnke, 1979, 1980; Brisset et al 1988; Buiatti and Scala, 1985; Chang et al., 1989; Dunbar and Stephens, 1989; Wenzel et al 1985; Willmot et al., 1989), and the advantages of in vitro selection have been noted (Chaleff, 1981, 1983; Huitema et al 1989; IAEA, 1985; Ingram, 1983; Ingram and MacDonald, 1986; Werry, 1981). In vitro resistance of plant

PAGE 60

53 tissues to Phytophthora in dual cultures (plant tissue and the pathogen together in axenic culture) has been correlated with field performance for: avocado (Dolan and Coffey, 1986; Zilberstein and Pinkas, 1987) and crops of seven other genera (McComb et al., 1987) to P. cinnamomi Rands; potato (Ingram, 1967) and tomato (Warren and Routley, 1970) to P. infestans (Mont.)DBy.; alfalfa (Miller et al 1984) to P. megasperma Drechs; tobacco (Helgeson et al 1976; Tedford et al 1990) to P. parasitica Dast. var. nicotianae (Breda de Haan) Tucker; and apple (Barritt et al., 1990; Jeffers et al 1981; Utkhede, 1986; Utkhede and Quamme, 1988) to P. cactorum (Leb. & Cohn) Schroet. Isolate specificity was noted in the reaction of papaya shoot cultures to P. palmivora (Butl.) Butl (Sharma and Skidmore, 1988). Measurement of lesions formed on excised apple twigs after inoculation with P. cactorum has proven to be a good indicator of relative resistance (Barritt et al 1990; Jeffers et al., 1981; Utkhede, 1986; Utkhede and Quamme, 1988). Preliminary investigations on the response of Citrus cultivars to culture filtrate from P. citrophthora were not particularly encouraging (Vardi et al., 1986a), although Tusa et al (1988) obtained a fair correlation between in vitro tolerance to culture filtrate and in vivo tolerance to bark inoculations in sour orange selections. No investigations of in vitro response of Citrus cultivars in dual culture with P. parasitica have been reported.

PAGE 61

CHAPTER 3 REDISCOVERY OF CARIBBEAN FORBIDDEN FRUIT AND EVALUATION OF ITS SIGNIFICANCE FOR CITRUS BREEDING Introduction A single introduction of grapefruit (Citrus *paradisi Macfadyen) into Florida from the West Indies in 1823 has been identified in the literature as the original source of all known grapefruit germplasm. Essentially all grapefruit cultivars in the United States are bud sports or nucellar seedlings derived from this single introduction. On the basis of historical records as well as morphological and biochemical characteristics, the grapefruit is considered to have originated in the West Indies as one or a series of chance hybridizations between sweet orange (C. sinensis) and pummelo (C. grandis) during the 17th or 18th centuries (Scora et al., 1982; Scora and Kumamoto, 1983). In the first literature describing the grapefruit, it is closely associated with another kind of Citrus known as "Forbidden Fruit." Kumamoto et al (1987) considered this form to be extinct. Subsequently, we identified a heterogeneous population of grapefruit-like Citrus growing in Saint Lucia (West Indies) that is known as forbidden fruit by some of the local residents (Bowman and Gmitter, 1990b). This population may provide a source of genetic diversity for grapefruit and hybrid cultivar development (Bowman and Gmitter, 1990a). 54

PAGE 62

55 The objectives of this project were to: 1. Describe the Citrus populations known in Saint Lucia as forbidden fruit; 2. Investigate and describe the probable relationship of these populations to the forbidden fruit mentioned in the early Caribbean literature and to the grapefruit; 3. Survey the history of grapefruit and its connection with the Caribbean forbidden fruit; 4. Identify some of the needs of grapefruit genetic improvement and limitations imposed by currently available grapefruit germplasm; 5. Import (under quarantine) budwood and seeds of some of the Caribbean forbidden fruit selections for characterization and testing in Florida. Materials and Methods An extensive search for selections of the Caribbean forbidden fruit was initiated in Saint Lucia during 1986 and 1987. A travel grant was obtained (with F.G. Gmitter as co-investigator) from The Center For Tropical Agriculture/ International Programs (University of Florida) during 1987 to allow for the collection of seeds and budwood from the previously identified forbidden fruit selections. This trip was completed by F.G. Gmitter and K.D. Bowman during December, 1987, and included careful examination of five Saint Lucia forbidden fruit selections, collection of seeds and budwood, and discussions with Saint Lucians who were able to identify these trees as forbidden fruit. The budwood

PAGE 63

56 selections were turned over to the Florida Department of Plant Industry, Gainesville, for shoot-tip budding, thermotherapy, and virus indexing. Additional seeds were obtained from one of these selections and another similar Saint Lucian selection in February and March, 1990. Seeds of morphologically similar selections were obtained from Saint Vincent and Trinidad in 1990. Seeds of all selections were planted in soilless potting mix, and plants were grown in a greenhouse. Seeds of selections SF23, SF24, and SF25 were surface sterilized by immersion and agitation in 70% ethyl alcohol for 10 minutes followed by 1.05% sodium hypochlorite plus 2 drops polyoxyethylene-20-sorbitan monolaurate (Tween 20: Fisher Scientific, Pittsburgh) per 100 ml for 20 minutes. Seeds were rinsed five times (five minutes each) in sterile distilled water and placed on the surface of 10 ml agar-solidified half-strength MT basal medium with full strength iron (Murashige and Tucker, 1969) and 25 g-1" 1 sucrose, adjusted to pH 5.7 (called GM1 medium through remainder of this text), that had been solidified in the bottom of 25 x 150 ml glass culture tubes (Bellco Glass, Inc., Vineland, NJ). The tubes were covered with "kap-uts" (Bellco Glass) translucent plastic closures. Shoot tip cuttings (mericlones) from in vitro seedlings of selections SF23-1 and SF24-1 were excised and placed with the cut surface in RM1 rooting medium. RM1 medium was composed of agar-solidified halfstrength MT basal medium with full-strength iron (Murashige and Tucker, 1969), 25 g-1" 1 sucrose, 500 mg-1 1 activated charcoal, and 0.02 mg-V 1 NAA, and adjusted to pH 5.8, as described by Grosser et al (1989). Epicotyl and internode segments (1 cm long) were excised from SF23-1 and SF24-1 and

PAGE 64

57 placed horizontally on the surface of a callus proliferation media, composed of solidified MT basal (Murashige and Tucker, 1969) with 50 g-T 1 sucrose, 0.22 mg-1" 1 kinetin, 0.5 mg-T 1 2,4-D and adjusted to pH 5.8 (called CP1 media throughout this text), in 15 x 100 mm Petri plates. After 6 weeks on callus proliferation medium, segments with associated callus were transferred to a shoot induction medium (SIM through the remainder of this text) composed of solidified MT basal (Murashige and Tucker, 1969) with 25 g-1" 1 sucrose, 0.5 g-T 1 malt extract, 0.25 g-1" 1 casein hydrolysate, 5 mg-T 1 BA, 0.01 mg-1" 1 2,4-D and adjusted to pH 5.8, in 20 x 100 mm Petri plates (SIM medium recipe, personal communication from J. W. Grosser). Adventitious shoots (somaclones) that formed on call used explants were excised and placed on rooting medium (RM1). Rooted meri clones and somaclones were transferred to soilless potting mix in controlled environment chambers. Meri clones were morphologically characterized with other control genotypes. Leaf samples from imported budwood and seedling selections were used for isozyme analyses on horizontal gels containing 10% Connaught starch (Fisher Scientific) and sometimes supplemented with 0.15% electrophoresis grade agarose. Extracts from leaf samples were made by crushing small rectangles of chromatography paper (wicks) into the abaxial side of leaf tissue until they were green with cell juices immediately prior to gel loading (Torres et al 1978). Isozymes were separated using a pH 5.7 histidine-citrate buffer (Cardy et al 1981) or Tris-citrate buffer (Gmitter, 1985) with 3 hours electrophoresis at 50 ma constant current and 4C. Activity stains used were PGM (phosphoglucomutase), PGI (phosphoglucose isomerase), PER (peroxidase), and PGD (6-phosphogluconate

PAGE 65

58 dehydrogenase) on the histidine-citrate buffer system, or GOT (glutamate oxaloacetate transaminase), and SDH (shikimic acid dehydrogenase) on the tris-citrate buffer system. Stain recipes were as described by Vallejos (1983) with slight modifications (see also Almansa et al., 1989; Kephart, 1990; Normand, 1988). Results and Discussion The origin and history of contemporary grapefruit cultivars has been investigated through the literature (Figure 3.1). Essentially all the popular cultivars were derived from a single set of germplasm introduced to Florida about 1823. This germplasm and the grapefruit hybrid species itself probably originated someplace in the West Indies, where it was closely associated with another name or kind of Citrus, the forbidden fruit. Natural populations of grapefruit-like Citrus have not been evaluated as potential sources of germplasm. Initial observations of a diverse semi-wild population of Citrus called forbidden fruit in Saint Lucia, West Indies, indicated that it might be a useful source of genetic material and prompted more thorough evaluations and germplasm collections. Forbidden fruit budwood is now growing under quarantine at the DPI facility in Gainesville and will be released after it is certified virus free. Seedlings of five selections are growing in the growth chamber, greenhouse, and/or field. Two manuscripts on the history, rediscovery, and characteristics of the forbidden fruit and its significance for grapefruit genetic improvement have been published (Bowman and Gmitter, 1990a, 1990b). Some of this published information is repeated here.

PAGE 66

59 West Indies Tree(s) 1 seed or plant(s) Safety Harbor, FL 1823 C^Marsh 1 860 limb sport 1 1 <3 Walters 1887 Thompson~^> limb I sport I C^^uby~^> Duncan 1830^* 1 limb sport C^^Foster^_^> limb sport limb I sports I V irr irradiated budwood Hudson C^jRio Red J^) 1 irradiated seed Ray/Henderson <^~Star Ruby~^> Figure 3.1. Probable pedigree of grapefruit cultivars (modified from Bowman and Gmitter, 1990a; p. 42). Many Saint Lucians who were interviewed had never heard of forbidden fruit or only knew it was a kind of Citrus similar to grapefruit. However, two morphologically distinct kinds of Citrus were identified as forbidden fruit by some Saint Lucian citizens. The first of these morphologically resembled the species Citrus grandis and examination of seed samples indicated monoembryony. Only five trees of this kind could be located, although other trees were mentioned by residents. Two of these were known to have been vegetatively propagated by grafting from one of the other three trees and another was very unhealthy. The two healthy, mature (and probably seedling) trees were used as specimens for derivation of the following description. One of these trees was located on the

PAGE 67

60 Raveneau Estate, Souci, Saint Lucia and identified as forbidden fruit by the resident estate owner, Fitz Raveneau. This tree was quite large and overgrown by vines and adjacent vegetation. Fruit sampled the first week of December, 1987, were very sour. Fitz Raveneau indicated that rind from the fruit of this tree was once used to make a type of candy. The second grandis-type tree used as a specimen to construct the description was growing on the Fond Doux Estate, Soufriere, Saint Lucia and was identified as forbidden fruit by Vernon, the resident estate manager. The original trunk of this tree was very large but oriented horizontally on the soil surface as a result of a hurricane in 1980. The "tree" that remained was composed of one large vertical branch that appeared to be healthy. Fruit of this tree were sweet and had a pleasant flavor when sampled between January-March, 1986 and 1987, as well as again in December, 1987. A composite description of these two trees from Bowman and Gmitter (1990b) is reproduced below. Saint Lucia Forbidden Fruit, grandis-type (GF) Trees to 12 m tall, trunk upright, branches spreading. Leaves evergreen, alternate; blade elliptic to ovate, 9-18 x 5-9 cm, glabrous above, glabrous with sparsely puberulent veins below, apex obtuse to acute, margins entire to crenate, base obtuse to cuneate, venation pinnate; petiole sub-alate to broadly alate, articulated, oblanceolate or obovate, occasionally with emarginate apex, 2-9 cm long. Young twigs green, angled, bearing single axillary spines, 0-25 mm long; mature twigs cylindrical, often spineless; older branches apparently spineless. Flowers not observed. Fruit borne singly or in pairs, a semi-sweet to acidic yellow hesperidium, oblate to pyriform, 20-25 cm x 15-20 cm; basal area slightly depressed, apex truncate; pedicel medium to very large; flavedo coarsely glandular and turning yellow at full maturity; albedo spongy, white, 1-4 cm thick; flesh greenish-yellow to pale white, 11-14 locules, with large, easily separated pulp-vesicles; central axis irregular, hollow. Seeds none to 100 per fruit, 15-20 mm 6-12 mm, cream-color, obovoid, flattened, angular, ridged, rough, monoembryonic; inner seedcoat light brown, dark chalazal spot; cotyledons white. (Bowman and Gmitter, 1990b; p. 168)

PAGE 68

61 A second kind of Citrus found growing sparsely throughout Saint Lucia was identified as the forbidden fruit of the past by MacDonald Ferdinand (born 1893) of Praslin, Micoud, Saint Lucia. Trees of this kind were somewhat smaller than the grandis-type and were more commonly known by Saint Lucians as "shaddette." Shaddette fruit varied considerably in characteristics from tree to tree, but were generally smaller and more similar to grapefruit or sweet orange than those of the grandis-type. Four shaddette trees were selected as specimens for the description reproduced from Bowman and Gmitter (1990b) below. Two of these trees were located near Fond-Saint-Jacques, Soufriere, one on the Fond Doux Estate, Soufriere, and the fourth in Blanchard, Micoud. Saint Lucia Forbidden Fruit, shaddette-tvpe (SF) Trees to 10 m tall, trunk upright, branches spreading. Leaves evergreen, alternate; blade elliptic to ovate, 6-10 x 4-6 cm, venation pinnate, blade glabrous above and below, apex obtuse to acute or occasionally emarginate, margins entire to crenate, base obtuse; petiole sub-alate to alate, articulated, oblanceolate or obovate, sometimes with emarginate apex, 2-3 cm long. Young twigs green, angled, bearing single axillary spines, 2-5 mm; mature twigs cylindrical, sometimes spineless; older branches spineless or sometimes very stoutly spined. Flowers in small racemes or one to several in leaf axils, perfect, actinomorphic; calyx cup-shaped with 4-5 lobes; petals 4-5, white to cream; stamens many; stigma flattened globose. Fruit a semi-sweet hesperidium, oblate, globose, or pyriform, 9-15 x 9-12 cm, sometimes hanging in distinct clusters; basal area frequently with shoulder or neck, depressed, lobed; apex truncate to rounded; flavedo moderately glandular and turning yellow at maturity; albedo white to pinkish, 5-20 mm thick; flesh pale yellow to pale pink, 11-14 locules, medium to large semi-coherent pulp vesicles; central axis solid. Seeds 5-30 per fruit, 15-16 mm x 7-9 mm, white, angular-ovate, moderately rough, monoembryonic and polyembryonic; inner seedcoat light to medium brown, dark brown chalazal spot; cotyledons white. (Bowman and Gmitter, 1990b; pp. 168-170) Both types of forbidden fruit were sometimes known by a second name in the local French Creole dialect. This name was spoken as though it

PAGE 69

62 were spelled "fwee dayfwandee" in English and literally translates to "fruit forbidden." The name "forbidden fruit" was considered by Saint Lucians to have originated from the biblical story of the Garden of Eden. The morphological descriptions of the field specimens indicated grandis-type forbidden fruit (GF) closely resembled selections of Citrus grandi s, a monoembryonic species. Because of the monoembryony, all seedlings of C. grandis are genetically unique and some morphological variation among seedlings is expected. Consequently, the two plant sample used to construct the morphological description of GF above may be inadequate to encompass normal genetic variation within this kind of forbidden fruit. Vegetative morphology of mature shaddette-type forbidden fruit (SF) trees in Saint Lucia was very similar to that of nearby grapefruit trees. However, as noted in the description above, fruit varied considerably from tree to tree, with some trees bearing more pummel o-l ike fruit, others grapefruit-like fruit, and yet others somewhat orange-like fruit. The four selections chosen as models for construction of the description above probably do not adequately sample all the variation in characteristics possible within the shaddettes. Examinations of small numbers of seeds from several selections indicated that at least some selections contained both monoembryonic and polyembryonic seeds. In vitro seedling growth, callus induction on internodal explants, adventitious shoot formation, and shoot rooting were evaluated for one seedling selection of SF23 (SF23-1) and one seedling of SF24 (SF24-1). Both seedlings were observed to grow well in culture, produce large amounts of callus from explants, and form large numbers of adventitious

PAGE 70

63 shoots from the callus. It is worth noting that adventitious shoot formation was much more prolific on explants from both shaddette selections than on explants from 'Valencia', 'Hamlin', and 'Ridge Pineapple' sweet oranges or 'Duncan' grapefruit. Excised shoots of both shaddette seedling selections were observed to root readily on RM1 medium. Samples of shoots from 'Carrizo', 'Hamlin', 'Valencia', and SF23-1 were examined for rooting at 20, 24, 28, and 32 days after placement on RM1 medium in culture boxes (Fig. 3.2). One month after placement on RM1, 33% of SF23-1 shoots, 27% of 'Carrizo' shoots, 18% of 'Hamlin' shoots, and 0% of 'Valencia' shoots had formed roots. After 44 days on RM1 medium, one of 23 'Valencia' shoots (4%) had formed a root. Figure 3.2. Frequency of shoot rooting during the first month on RM1 medium.

PAGE 71

64 Meristem clones propagated by tissue culture were established in soilless potting mix and morphologically characterized (Tables A.1-A.6). Morphology of both shaddette clones differed considerably from 'Hamlin' and 'Valencia'. The shaddette seedlings closely resembled the 'Duncan' and 'Marsh' clones for the characteristics spine length, spine length/diameter, oil gland density, and internode length. Isozyme analysis was completed on one of the two specimen GF clones (the other did not survive propagation in quarantine), as well one standard selection of sweet orange, grapefruit, and pummelo (previous work has indicated almost no variation within these species [F.G. Gmitter, personal communication; Torres et al 1978]). Allelic constitutions of these clones were deduced from banding patterns of PGM, PGI, PER, SDH, and GOT activity stains (Table 3.1). Banding patterns of PGM, PGI, SDH, and GOT for samples from GF13 were identical with those observed for common Florida selections of pummelo, C. grandis (Table 3.1). Clone GF13 was observed to differ from common pummelo selections and resemble grapefruit in its heterozygosity (FS) at the peroxidase locus. These isozyme data are consistent with the possibility that GF13 is a hybrid between grapefruit (or shaddette) and pummelo, but not a simple hybrid between sweet orange and pummelo (because GF13 is homozygous FF at the GOT locus). However, a thorough survey of C. grandis has not been completed, and it is possible the F allele for peroxidase is present in some pure pummelo selections. Isozymic characterization of the three shaddette clones (SF23, SF24, and SF25) was completed on plant material in quarantine. Allelic constitution of these clones and standard selections of sweet orange,

PAGE 72

65 Table 3.1. Isozyme genotypes of seven Citrus selections. PGM PGI PER cni i SDH G0T1 Sweet Orange FS MS FF IS SS Pummel o ss SS ss ss FF, FS Grapefruit ss SS FS IS FS SF23 ss ss FS IS FS SF24 ss ss FS IS FS SF25 ss ss FS IS FS GF13 ss ss FS SS FF SF = shaddette-type forbidden fruit GF = grandis-type forbidden fruit grapefruit, and pummelo were deduced from banding patterns of PGM, PGI, PER, SDH, and GOT activity stains (Table 3.1). The three shaddette clones and grapefruit had identical banding patterns for PGM, PGI, PER, SDH and GOT. Some preliminary evidence for differences in banding patterns of shaddette clones and grapefruit were obtained with PGD (personal communication, F.G. Gmitter). The shaddette selections were differentiated from pummelo by being heterozygous at PER (FS) and SDH (IS) loci. Shaddettes were distinguished from sweet orange by PGM, PGI, PER, and GOT. Allelic differences between the GF13 clone and the shaddette selections were detected with SDH and GOT as an increase in heterozygosity among the shaddettes.

PAGE 73

66 Although the three shaddette clones were quite similar to grapefruit in morphologies and isozyme banding patterns, shaddette seedlings showed a much higher frequency of isozyme variation at the PER locus than seedlings from three grapefruit cultivars (Table 3.2; Gmitter et al., in preparation). Similar isozyme variation has been observed among shaddette seedlings at other isozyme loci (F.G. Gmitter, personal communication). The appearance of altered isozyme genotypes indicates that seedlings are zygotic in origin. An average 27 percent of shaddette seedlings showed recombination at the PER locus by the presence of FF or SS homozygotes. Because the seed parent clones are FS at peroxidase (Table 3.1), selfed zygotic progeny should segregate at a ratio of 1FF:2FS:1SS. Therefore, only about half of the zygotics would be detected as homozygotes at a single locus. Ignoring the possibility of cross pollination, the projected frequency of zygotics in a population of shaddette seedlings may be about 54 percent. This is in great contrast to the 0 to 8 percent zygotics predicted on the basis of the peroxidase segregation for the grapefruit cultivars. Although the phylogenetic relationships cannot be determined precisely, isozyme evidence supports the morphological indications of a close relationship between the grapefruit and the Saint Lucian forbidden fruit. The connection between the two is further corroborated by the close association of the names in historical literature. Bowman and Gmitter (1990b) have proposed that the Saint Lucian shaddette and grapefruit should be considered as two different groups of selections within the hybrid species C. *paradisi. The name "forbidden fruit" was apparently derived from an early C. grandis selection called "Adam's

PAGE 74

67 Apple." The name forbidden fruit has subsequently been applied sporadically to many (all?) different selections of C. grandis and C. xparadisi in the West Indian historical literature. The hypothesized species and nomenclatural relationships, as well as an indication of the genetic diversity within populations, are presented pictorially in Figure 3.3. Bowman and Gmitter (1990a) suggested that the shaddette may be a valuable germplasm resource, a proposal strengthened by the isozyme evidence presented here. Table 3.2. PER zymotypes and percent zygotics identified among seedlings of grapefruit and three shaddette (SF) selections (Gmitter et al., in preparation). Number Percent Seedl ings PER Locus Zygotics Examined SS FS FF Identified Duncan 51 0 49 2 4% Foster 50 0 50 0 0 Marsh 50 0 50 0 0 SF23 41 12 29 0 29 SF24 39 8 27 4 31 SF25 23 5 18 0 22 SF = shaddette-type forbidden fruit

PAGE 75

68 Figure 3.3. Speculative taxonomic and nomenclatural relationships between forms related to C. xparadisi (modified from Scora, 1975; p. 373).

PAGE 76

CHAPTER 4 EVALUATION OF ORGANOGENIC CITRUS TISSUE CULTURES AS A SOURCE OF GENETIC VARIATION FOR CULTIVAR IMPROVEMENT Introduction Numerous reports of variation among plants regenerated from tissue cultures have been recorded in the past two decades (D'Amato, 1985; Evans, 1989; Evans et al 1984; Evans and Sharp, 1986; Orton, 1984). The term "somaclonal variation" was coined by Larkin and Scowcroft (1981) to describe this phenomenon. In some cases, useful mutations have been reported in somaclonal variants, including higher yield (Heinz and Mee, 1971), male sterility (Griesbach, 1989), disease resistance (Heinz et al., 1977; Krishnamurthi and Tlaskal, 1974; Orton, 1984; Shepard et al 1980), insect resistance (Miles et al., 1981), dwarfing (Griesbach, 1989), and a wide variety of other morphological characteristics (Burk and Matzinger, 1976; Cummings et al 1976; Eapen et al 1989; George and Rao, 1983; Liu and Chen, 1978). The severe limitations on sexual hybridization of Citrus imposed by long juvenility and extensive nucellar polyembryony make somaclonal variation an attractive alternative method of generating useful genetic variation. Only a few preliminary investigations of somaclonal variation have been completed in Citrus, with somewhat conflicting results (Gmitter, 1985; Kobayashi, 1987; Navarro et al 1985). No thorough 69

PAGE 77

70 investigations of somaclonal variation arising from organogenic regeneration of Citrus cultures have been reported. If somatic variation from Citrus tissue cultures is not frequent enough to make it productive in cultivar improvement, mutagenic treatments may be employed to increase genetic variation. Irradiation of budwood and nucellar seeds have both been successful in Citrus genetic improvement (Hearn, 1984, 1986; Hensz, 1977, 1985), and there have been some investigations of irradiation of Citrus tissue cultures (Chang et al., 1984; Kochba and Spiegel-Roy, 1982; Liu and Deng, 1985; Spiegel-Roy and Kochba, 1973b; Vardi, 1977; Zubrzycki and Diamante De Zubrzycki, 1982). However, little information is available on the effect of gamma radiation on Citrus organogenic and embryogenic callus cultures and the resulting plants. This study investigated the potential for somaclonal variation in Citrus genetic improvement and studied the effects of gamma radiation on callus cultures. The somaclones were produced by organogenesis from internodal explants and compared with control plants produced by in vitro cuttings from apical and axillary meristems (mericlones) of the same plant material. The objectives were to: 1. Regenerate plants by organogenesis from call used Citrus explants in vitro; 2. Compare the juvenile morphology of these plants and plants produced by meristem cuttings from the same seedlings and/or from other seedlings of the same cultivar; 3. Determine the effect of gamma radiation on Citrus explants.

PAGE 78

71 Materials and Methods Production of somaclones 'Hamlin' and 'Valencia' sweet oranges (C. sinensis [L.] Osbeck) were chosen for the study of somaclonal variation because they are common cultivars that might benefit by some types of somatic mutations. The cultivar 'Carrizo' citrange (C. sinensis x p. trifoliata [L.] Raf.), known to be responsive in culture, also was included because the two sweet orange cultivars did not produce shoots readily from callus during preliminary studies. Seeds of 'Carrizo', 'Hamlin', 'Valencia', 'Duncan' grapefruit (C. *paradisi Macfadyen), shaddette selections SF23-1 and SF24-1 (presumably C. xparadisi; see Chapter 3) and 'Meiwa' kumquat {Fortunella crassifolia Swing.) were surface sterilized by immersion and agitation in 70% ethyl alcohol for 10 minutes followed by a solution of 1.05% sodium hypochlorite plus two drops polyoxyethylene-20-sorbitan monolaurate (Tween 20: Fisher Scientific, Pittsburgh) per 100 ml for 20 minutes. The seeds were rinsed in sterile distilled water five times (five minutes each rinse) and individual seeds placed on the surface of 10 ml germination medium (GM1; see Chapter 3) solidified in the bottom of 25 x 150 mm glass culture tubes (Bellco Glass, Inc., Vineland, NJ). The tubes were covered with translucent plastic Kaputs closures (Bellco Glass) and sealed with Nescofilm (Nippon Shoji Kaisha Ltd., Osaka, Japan). In vitro seedling cultures were maintained at 27C under constant fluorescent lighting. After seeds germinated, 1 cm epicotyl and internode segments (called "segments" through the remainder of this chapter) were excised and placed horizontally on the surface of a callus proliferation medium, called CP1 (see Chapter 3). Some 'Carrizo' segments were sacrificed for

PAGE 79

72 determination of the starting mass of apical and basal halves of the explants. 'Carrizo' segments that were to be used for the 21 day mass determinations were marked at their midpoint with filter-sterilized India ink prior to placement on CP1 medium. Severed seedling apices and/or nodes were retained as meristem clones (mericlones) and placed in rooting medium (RM1, see Chapter 3) in 20 100 mm Petri plates or Magenta GA-7 culture boxes (Magenta Corp., Chicago). After 3 weeks on CP1, marked segments were removed for fresh and dry weight determinations. Segments that were to be maintained on CP1 medium for 6 weeks were transferred to fresh CP1 medium at three weeks. After 4 or 6 weeks on callus proliferation medium, explant segments with associated callus were transferred to a shoot induction medium, called SIM (see Chapter 3). Adventitious shoots (somaclones) that formed on call used explants were excised and placed on rooting medium (RM1) or minigrafted on etiolated 'Carrizo' seedlings (see below). Rooted adventitious shoots and meristem clones were transferred to soilless potting mix and maintained in high humidity and moderate light for several weeks before being transferred to a controlled environment chamber where plants were grown prior to morphological evaluation. Similar methods were used to produce twelve 'Valencia' mericlones, five 'Valencia' somaclones, one minigrafted 'Duncan' somaclone, one 'Meiwa' mericlone, one shaddette SF23-1 mericlone, and three shaddette SF24-1 mericlones for use as genetically diverse controls (Table 4.1). One seedling from a ribbed fruit chimera of 'Marsh' grapefruit (C. xparadisi) was used as another genetically different, unifoliolate control. Seven etiolated 'Carrizo' seedlings were decapitated between the cotyledons and

PAGE 80

73 the first node and allowed to regrow by adventitious shoots as additional trifoliolate controls. All plants for this study were grown in 153 ml of a soilless potting mix (SPM2) containing peat moss, vermiculite, and perlite (Fafard Mix No. 2: Conrad Fafard, Inc., Springfield, MA) using pots 20.5 cm long and 4 cm inside diameter at the top. Table 4.1. Clones used in somaclonal variation studies. Approximate number measurements Cul ti var Clone tvDe Root stock Number d1 ants per plant per characteristic Carrizo Seedl ing Self 7 6 Carrizo Meri clone Self 4 6 Carrizo Somaclone Self 47 6 Duncan Somaclone Carrizo 1 5 Haml in Mericlone Self 14 5 Haml in Somaclone Self 29 5 Haml in Somaclone Carrizo 8 5 Marsh Seedl ing Self 1 5 Meiwa Mericlone Self 1 3 SF23-1 Mericlone Self 1 5 SF24-1 Mericlone Self 3 6 Valencia Mericlone Self 12 5 Valencia Somaclone Self 3 5 Valencia Somaclone Carrizo 2 5

PAGE 81

74 'Carrizo' seedlings used for minigrafting were grown in SPM2 and covered with tall unlighted tubes for maximum nonchlorophyllous shoot elongation and then decapitated by a horizontal cut between the cotyledons and first leaf node (1-5 cm above surface of SPM2). Shoot tips 1-3 mm in length were excised from the explant and immediately placed on the decapitated 'Carrizo' stock so that the cut surfaces of the two units were in contact (shoot tips were much larger than those used in previouslydescribed micrografting [Nauer et al 1983; Navarro et al 1975; Oiyama and Okudai, 1986; Tusa et al 1978]). The stems of the etiolated 'Carrizo' seedlings and the minigrafted adventitious shoots were of similar diameter. After transfer to SPM2, all mericlones, somaclones, and minigrafted somaclones (also the 'Marsh' and 'Carrizo' seedlings) were grown in a controlled environment chamber with a diurnal cycle of 16 hr light at 30C and 8 hr dark at 25C. Plants were watered every two days, alternating between tap water and a solution of 20-10-20 fertilizer plus micronutrients at 300 mg-1 1 (Peter's Professional Peat Lite Special 2010-20 Water Soluble: W.R. Grace & Co., Fogelsville, PA). Insects were controlled on three occasions by spray application of kinoprene (Enstar 5E: Zoecon Corp., Palo Alto, CA) plus oil or fluvalinate (Mavrik: Zoecon Corp.) within 10 days of a cutback. Evaluatio n of plant morphology Plants were cutback and allowed to regrow from axillary buds prior to the morphological evaluations so that the transient effects of tissue culture would be minimized and all organs would be the same age at the time of evaluation. Vegetative morphologies of 'Hamlin' mericlones and somaclones, 'Valencia' mericlones and

PAGE 82

75 somaclones, one 'Meiwa' mericlone, one shaddette SF23-1 mericlone, three shaddette SF24-1 mericlones, one 'Duncan' somaclone, and one 'Marsh' seedling were evaluated 6-9 weeks after plants were cut back 10 cm above the soil level (or 10 cm above the graft union for the minigrafted 'Hamlin' somaclones and minigrafted 'Duncan' somaclone). Single shoots were allowed to grow on each plant and any additional axillary budbreak was removed. Characteristics measured during the first evaluation were leaf length, leaf width, internode length, spine length, spine diameter (at the spine midpoint), abaxial oil gland density, abaxial leaf color, and adaxial leaf color. Leaf, internode, and spine lengths, as well as leaf width, were measured with a steel ruler in millimeters. Spine diameter was measured by a dial caliper to the nearest thousandth of an inch and converted to the nearest tenth of a millimeter for calculations. Abaxial oil gland density was estimated by averaging counts from a microscope-video camera image of two 20 mm 2 fields per leaf. Abaxial and adaxial leaf colors were determined on the standard X, Y, and Z color axes by a HunterLab Citrus Colorimeter model D25 (HunterLab, Reston, VA) modified with a black-plastic faceplate containing a 19 mm circular aperture. Conversions of color measurements to the L, A, and B visual color axes were completed using standard mathematical formulae issued by HunterLab and similar to those described by Ting and Rouseff (1986). Morphological characterization, in most cases, was based on five nodes from each plant, beginning with the first node more than 3 cm above the start of the regrowth after the cutback and including each of the next 5 consecutive nodes. Organs obviously damaged or within 3 cm of a shoot apex were not included.

PAGE 83

76 A second morphological evaluation of 'Hamlin' mericlones and selected somaclones was completed 14-16 weeks after a third shoot cutback. Single axillary shoots were allowed to grow on each plant, and any additional axillary budbreak was removed. Characteristics measured during the second evaluation were leaf length, leaf width, spine length, spine diameter (at the spine midpoint), abaxial leaf color, and adaxial leaf color. The second morphological characterization, in most cases, was based on five nodes from each plant, beginning with the first node more than 3 cm above the start of the second growth flush after the third cutback and including each of the next 5 consecutive nodes. Organs obviously damaged or within 3 cm of a shoot apex were not included. Morphological evaluations of 'Carrizo' mericlones, somaclones, and adventitious seedlings were accomplished 20-22 weeks after the first cutback (for leaflet length and chirality) or 19-20 weeks after the second cutback (for spine length and diameter). Characteristics measured were chirality (bud to the left or right of the spine at each node), left, main, and right leaflet lengths, spine length, and spine diameter (at the spine midpoint). Spine L/D ratio and leaflet ratio (left plus right leaflet lengths divided by main leaflet length) were calculated. Characterization of leaflet length and chirality were based on six nodes from each plant, beginning with the first node more than 8 cm above the start of the regrowth after the first cutback and including each of the next six consecutive nodes. Characterization of spine length and diameter was based on six nodes from each plant, beginning with the first node more than 8 cm above the start of the second growth flush after the second

PAGE 84

77 cutback and including each of the next six consecutive nodes. Organs obviously damaged or within 8 cm of a shoot apex were not included for any measurements. Irradiation of embrvogenic cultures Irradiation of embryogenic callus cultures was investigated as a method of inducing a higher frequency of mutations among plants from tissue culture. All available embryogenic callus suspension lines were obtained from J.W. Grosser, including 'Hamlin', 'Valencia', 'Valencia Rohde Red' (Stewart et al 1975), 'Ridge Pineapple' (C. sinensis), and 'Milam' (C. jambhiri Lush.). Suspension cultures 7 days old from each cultivar were plated out onto EME medium, composed of solidified MT basal (Murashige and Tucker, 1969) with 50 g-1" 1 sucrose, 0.5 g-T 1 malt extract and adjusted to pH 5.7, in 15 or 20 x 100 mm Petri plates. After 18 days, plates were exposed to 0, 3, 6, 12, or 24 krad from a Cobalt-60 source. Callus or embryos were transferred to EME supplemented with 20 ml-l" 1 coconut water about 12 weeks after irradiation. Germinating embryos were transferred to SPM2 and established in growth chambers. Irradiati on of organogenic cultures Irradiation of organogenic explants or callus cultures was investigated as a method of inducing a higher frequency of mutations among plants from tissue culture. 'Carrizo' and 'Hamlin' were chosen for the same reasons described in the somaclonal variation project. In this study, 'Ridge Pineapple' was substituted for 'Valencia' because the latter had proved very difficult to regenerate by organogenesis. The hybrid US-119 ([P. trifoliata C. *paradisi] x C. sinensis) also was included as a trifoliolate type (and thus perhaps very responsive to culture) that might benefit by disruption of some metabolic

PAGE 85

78 pathways affecting flavor components (one of the types of mutations likely to occur as a result of irradiation). US-119 is of interest as a cold hardy and relatively good-tasting scion type that could still benefit by the loss of some remaining trifoliolate flavor characteristics. Seeds of 'Carrizo' citrange, 'Hamlin', 'Ridge Pineapple', and US-119 were surface sterilized by immersion in 70% ethyl alcohol for 10 minutes followed by a 1.05% solution of sodium hypochlorite (20% bleach) plus 2 drops Tween per 100 ml for 20 minutes. Disinfested seeds were then rinsed in sterile distilled water 5 times (5 minutes each rinse), and individual seeds were placed on the surface of 10 ml germination medium solidified in the bottom of 25 x 150 mm glass culture tubes covered with Kap-uts. All seedling cultures were maintained at 27C under a diurnal cycle of 16 hr fluorescent light and 8 hr dark. After seeds germinated, 7 mm segments were excised from the stems and placed flat on the surface of CP1 medium in 20 x 100 mm Petri plates for callus induction. All callus cultures were maintained at 27C and under continuous fluorescent lighting. Selected segments were weighed prior to irradiation and exposed to 0, 3, 6, 12, or 24 krad from a Cobalt-60 source six days or 41 days following explant placement on CP1 medium. After six weeks on CP1, explants and associated callus were transferred to SIM medium in 20 x 100 mm Petri plates. Combined callus and explants of material irradiated 6 days after explant placement were weighed 36-38 days after irradiation. Health of callus cultures was scored about 70 days after irradiation using a subjective five level rating scale for organogenic potential: 1 = dead, 2-4 = intermediate, and 5 = adventitious shoots present.

PAGE 86

79 Results and Discussion Subjective observation did not reveal differences between epicotyl and internodal segments in callus growth, adventitious shoot formation, or subsequent rooting of these shoots. However, substantial differences were evident in the responses of explants from the different cultivars. Explants from 'Carrizo' and the shaddette selections SF23-1 and SF24-1 produced callus more rapidly and shoots more profusely than explants of 'Hamlin' and 'Valencia'. Shoots of all these selections rooted to some degree in RM1 medium. 'Duncan' was distinguished by the production of large amounts of callus and adventitious shoots but an exceptional recalcitrance to rooting on RM1. Callus growth from explants of all the cultivars examined exhibited considerable polarity, and usually the basal portion of each 1 cm segment produced substantially greater amounts of callus than the apical portion of the segment. An experiment was conducted to measure this differential callus growth. Fresh weight of 1 cm 'Carrizo' segments was found to be equally distributed between the apical 0.5 cm and basal 0.5 cm (N=12) when measured immediately following excision from the in vitro seedlings (Fig. 4.1). When maintained on CP1 medium, a much larger amount of callus growth occurred on the basal half of the explant as compared with the apical portion. The average fresh weight of the tissue that developed from the basal half of the explant was about five times greater than that from the apical half after 21 days on CP1 medium (N=8). Dry weight of these segment halves at the end of 21 days on CP1 averaged 19.4 mg for the basal section and 5.4 mg for the apical section. The polar growth response was similar regardless of the segment length or how many pieces

PAGE 87

80 0 20 40 60 80 100 Fresh Weight (mg) Figure 4.1. Differential callus growth in apical and basal portions of 1 cm 'Carrizo' segments on CP1 medium. into which a single segment was divided. This effect may be the result of differential hormone (e.g. 2,4-D) uptake from the medium by the apical and basal ends of the explant. Although explants were oriented horizontally on the surface of the medium, there may have been a significant effect of xylem or phloem conductance polarity as determined during seedling development. Similar polarity of callus proliferation was noted in explants from 'Hamlin' and 'Valencia', although the total volume of callus produced and the differential was considerably less. Shoot induction from call used explants was never observed on CP1 medium and typically did not occur until three or more weeks after transfer to SIM medium. This was probably a reflection of the need for

PAGE 88

81 the hormone BA in the medium for shoot induction (Duran-Vila et al 1989). Sporadic adventitious shoots were obtained from callused segments of 'Hamlin' and 'Valencia' on SIM medium, but most explants from these cultivars produced no adventitious shoots, even after 6-12 months on SIM medium. Although callused explants from 'Carrizo' and shaddette SF24-1 did not always produce adventitious shoots, they were more likely to produce shoots than 'Hamlin' or 'Valencia'. In addition, organogenesis from responsive individual explants of 'Carrizo' or SF24-1 was not sporadic, because scores of adventitious shoots could typically be obtained from the same explant. The 130 rooted 'Carrizo' somaclones obtained during this study were all produced by four explants from one nucellar 'Carrizo' seedling. These explants rapidly regenerated new shoots after each harvest and remained healthy and prolific until the experiment was terminated 12 months after first shoot induction. The nature of shoot induction on 'Carrizo' and shaddette explants suggests that each shoot may not be produced by an individual adventitious initiation event. Although initial shoot production must be regarded as adventitious because such shoots did not develop from preformed buds, the prolific nature of shoot induction from certain parts of the explant may result from the development of new meristematic regions in the callus material. Numerous buds and shoots may then develop from these meri stems without additional adventitious events. If this is the case, the possibility for derivation of somaclonal variation from such regions is probably significantly less than it would be if each shoot developed from a separate adventitious initiation event. The formation of such an

PAGE 89

82 "adventitious meristematic region" would, however, increase the potential of these techniques for micropropagation and/or genetic transformation. Rooting of shoots on RM1 was easily achieved for 'Carrizo', shaddette SF23-1, and 'Hamlin', but was considerably slower for 'Valencia' (see Fig. 3.2). Different rooting media compositions were not evaluated for increased efficacy in this study, but the results of Duran-Vila et al (1989) indicated that higher NAA concentrations may produce better results for unresponsive genotypes. A high percentage of the rooted shoots survived transfer to soilless potting mix (SPM2), especially if shoots and leaves were pruned prior to planting and high humidity was maintained in the acclimation chamber for several weeks following transfer. However, extending the period of high humidity or pruning roots prior to planting was distinctly detrimental to plant survival and establishment. Minigrafting short adventitious shoots directly from the callus mass onto etiolated 'Carrizo' seedlings was investigated as a more rapid method of regenerating somaclones and one that might overcome selective forces eliminating certain types of mutants (e.g. mutations affecting rooting) from the regenerated plant population. Some success was achieved in the minigrafting of 'Hamlin' and 'Valencia' adventitious shoots, and ten of the forty-three somaclones included in the evaluated population were produced by this method. The one successfully regenerated 'Duncan' somaclone was also produced by minigrafting. These plants are indicated by a "Z" suffix in Tables 4.2 and A.1-A.17. Many of the rooted adventitious shoots (somaclones) produced during this study exhibited deviant morphology during in vitro culture and the first few months of establishment in soilless potting mix. Elongated

PAGE 90

83 leaves, extremely short internodes, and deformed shoots were observed among a number of the 'Hamlin' somaclones during the in vitro rooting and in vivo establishment phases. Similar aberrant morphologies, including unifoliolate leaves, were observed in some 'Carrizo' somaclones during these phases. To reduce the non-genetic (and ephemeral) alterations that would interfere with identification of true genetic mutations, morphological characterization was delayed until plants were well established in soil and was completed on healthy shoots grown entirely after a severe cutback in vivo. In addition, the principle control plants were produced from the same seedlings (or supposed genetically identical seedlings) by in vitro meristem cuttings (mericlones) that were rooted, established in potting mix, and grown under conditions identical to those used for the somaclones. Morphological characteristics were chosen for use as indicators of genetic variation among the clones (mericlones and somaclones) on the basis of observed preliminary morphological abnormalities, convenience, and a presumed diversity of genetic controllers. In general, no practical techniques have been described for characterizing horticulturally and/or economically important traits in individual young Citrus plants (although evaluation of one such method was a goal of the research described in Chapter 6). However, if somaclonal variation is caused by random mutations, the frequency of mutations affecting several diverse morphological characters should indicate the mutation frequency affecting horticulturally important traits. In the first morphological evaluation of the 'Hamlin' somaclones, fourteen 'Hamlin' mericlones were used as the reference population. These

PAGE 91

84 meri clones were derived from ten seedlings that in turn were obtained from the fruit of one 'Hamlin' tree. These seedlings were expected to be of nucellar origin and genetically identical because 'Hamlin' is highly polyembryonic and has been observed only rarely to produce zygotic seedlings. Twenty-three of the 'Hamlin' somaclones were produced from these same seedlings or other seedlings obtained from the same tree at the same time. The remaining fourteen 'Hamlin' somaclones were produced from seedlings of a second 'Hamlin' tree considered to be morphologically indistinguishable from the first. Normal probability plots (quantile-quantile) and the Kolmogorov test in the SAS univariate procedure indicated that the combined measurements of the fourteen 'Hamlin' mericlones (60 to 70 repetitions for each trait) were normally distributed for all of the evaluated characteristics (Prob.>D was greater than 0.01 for each character; Prob.>D was greater than 0.15 except for internode length, spine length, leaf length, and spine ratio). Therefore, these measurements were combined to obtain the "HAMLIN MERICLONE" used as the control in each of the statistical comparisons. Student's t-tests were used to compare the morphological characteristics of the two subpopulations of 'Hamlin' somaclones that were obtained from different seed sources (indicated as "HA" and "HX" in the tables). Because all of the mericlones available for inclusion in the HAMLIN MERICLONE were from the "HA" population, any genetic differences between the two seed sources may have affected comparisons of the "HX" somaclones with the "HA" mericlones. The two populations had similar means (Prob.>t was at least 0.05) for all characters except leaf width

PAGE 92

85 (Prob.>t = 0.0023), abaxial leaf color B (Prob.>t = 0.0383), and leaf L/W ratio (Prob.>t = 0.0003). Comparison of the variances for each characteristic in the two populations by the F statistic indicated that there was no evidence for differences between the variances of the two populations for most traits (H: Variances are equal, Prob.>F was at least 0.1). Although some mean differences were indicated, no alternative control was available and the seed source was necessarily ignored in subsequent statistical comparisons. However, this factor must be considered as a possible uncontrolled source of variation that may have influenced the characteristics of individual "HX" somaclones or the resulting population statistics. Student's t-tests were also used to compare the morphological characteristics of the 'Hamlin' somaclones on their own roots (29 clones) with the 'Hamlin' somaclones minigrafted on 'Carrizo' roostocks (8 clones). Differences in the effects of the two root types on morphological characteristics of somaclones would influence determinations of aberrant plants, and all of the mericlones included in the HAMLIN MERICL0NE were on their own roots. Comparisons of these two subpopulations of 'Hamlin' somaclones by Student's t-tests indicated that the means were different (at a = 0.05) for all of the morphological characteristics except internode length, leaf length, and leaf ratio. Root type significantly affected the morphological characteristics of the somaclones (e.g. 'Hamlin' on 'Carrizo' rootstocks tended to produce longer, thicker spines, and wider, less darkly colored leaves). However, this factor was ignored in subsequent statistical comparisons because no 'Hamlin' mericlones on 'Carrizo' rootstocks had been produced during the

PAGE 93

86 generation phase of the study. The effect of rootstock must, therefore, be regarded as another uncontrolled source of variation in the statistical comparison of clone morphologies. Twelve 'Valencia' mericlones (grouped as a single genotype with 50 to 60 repetitions), one 'Duncan' somaclone, one shaddette SF23-1 mericlone, three shaddette SF24-1 mericlones (grouped as a single genotype), one 'Marsh' seedling, and one 'Meiwa' mericlone were used in comparisons with the HAMLIN MERICLONE for each of the morphological traits. Duncan's mean separation ( = 0.01) indicated that at least one clone was measurably different from 'Hamlin' in: adaxial leaf colors A, B, L, and A/B ratio; abaxial leaf colors B, L, and A/B ratio; spine length; spine diameter; spine length/diameter ratio; leaf width; leaf length/width ratio; oil gland density; and internode length (Tables 4.2 and A.1-A.6). The consistency with which these genotypes are clearly distinguished from 'Hamlin' gives an indication that genetic differences do, in fact, lead to measurable differences in many of these traits. However, it should be noted that the degree of genetic variation represented within this reference group was greater than would be expected to occur as a result of most somaclonal mutation. The cultivar 'Valencia' is most closely related to 'Hamlin' and is only separated from it by four characteristics with Duncan's test at a = 0.05. The first characterization of each of the 37 'Hamlin' somaclones and 5 'Valencia' somaclones is summarized in the first half of each column in Table 4.2. More complete results are presented in Tables A.7-A.12 where a mean measured value is indicated when it differed significantly from HAMLIN MERICLONE (Duncan's separation at a = 0.01). Analysis of variance

PAGE 94

87 probability of a greater F value is indicated at the end of each table column. More somaclones differed from HAMLIN MERICLONE than would be expected by chance for most of the morphological characters. For example, twelve of the fortytwo somaclones differed from the meri clone in adaxial color L by Duncan's separation at = 0.05 (see Table A. 7), while only 2-3 outliers are predicted by definition of the a = 0.05 test. However, in general these aberrants were not distinctly different from all other somaclones but seemed to represent the tails of a broader normal distribution than that predicted by the mericlone controls. These observations suggest either an increased genetic diversity or some undefined nongenetic variation within the somaclone population. In an attempt to determine whether the aberrations observed in these somaclones were caused by genetic mutations or nongenetic factors, the meri clones, the somaclones indicated to be aberrant during the first test, and some of the "normal" somaclones were cut back and regrown twice before a second morphological characterization. Measurements of oil gland density and internode length detected no variants during the first test (Table 4.2 and A. 12.) and so were not included in the second characterization. Some meri clones and somaclones failed to regrow after the cutbacks and could not be reevaluated. For most of the characters, there were substantial differences between the mean values of the HAMLIN MERICLONE in the first and second morphological evaluations. This was probably caused by advanced maturity and the increased stress resulting from extended growth in relatively small rooting volumes. Nevertheless, it seems reasonable to assume that the resulting alterations were similar for each plant and should not have substantially affected relative

PAGE 95

88 Table 4.2. Summary for morphological comparisons of HAMLIN MERICLONE with 6 other cultivars and 42 somaclones. See Tables A.1-A.17 for more complete data. Adaxi Color Abaxi Color Spine Leaf In Oil Clone A B L Rl A B L Rl LE D R2 LE W R3 Lg Gl Val M 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Duncan 0 0 0 0 0 0 0 + + 0 + 0 0 SF23-1 + — 0 0 0 0 0 0 0 0 — SF24-1 0 0 0 0 + 0 + 0 0 0 0 -~ 0 Marsh + + 0 0 + + + 0 0 0 0 0 0 Meiwa 0 0 + — + HA142E121 + 0 + 0 0 0 0 0 0 0 0 0 HA142E215 00 00 00 00 00 00 00 00 0000 00 00 0 0 HA142E222 0 0 0 0 0 0 0 0 0 0 0 HA4I221 0 0 0 0 0 0 0 0 0 0 0 0 0 0 HA5E121 0 0 0 0 0 0 0 0 0 0 0 0 0 HA5I122 0 0 0 0 0 0 0 0 0 0 + 0 0 HA6E114Z 0 + + 0 0 0 + 0 0 0 0 0 0 0 0 0 HA7E12Z 00+ 0+ 0+ 00 0+ 0+ ++ 00 00 00 00 00 00 0 0 HA9E13Z 00 00 00 00 00 00 +0 +0 00 00 00 00 00 00 0 0 HX1X11 00 00 00 00 00 00 00 00 00 -0 00 00 00 00 0 0 HX1X12 00 00 -0 -0 00 -0 -0 -0 -0 -0 00 00 00 0 0 HX101E111 00 00 00 00 00 00 00 00 00 0+ 00 00 -0 00 0 0 HX115E12Z 00+ 0+ 0+ 00 u+ u+ u+ uu uu 00 00 00 00 0 n u HX115E14Z 00 00 00 00 00 00 +0 +0 00 00 00 00 00 00 0 0 HX15E11 0+ 00 00 00 00 00 00 00 00 00 00 00 00 00 0 0 HX20E11 -0 +0 +0 00 00 +0 +0 00 00 00 00 00 00 00 0 0 VA13E11 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 VA30E11 0 0 0 0 0 0 0 0 0 0 0 0 0 + 0 0 VA33E11 0 0 0 0 0 0 0 + 0 0 0 0 0 + 0 0 VX91E11Z 0 + + 0 0 0 + 0 0 0 0 0 0 0 0 0 VX91E12Z 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Other Hamlin Somaclones 5 other 00 00 00 00 00 00 00 00 00 00 00 00 00 00 0 0 16 other 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 First and second columns in a category are for first and respectively. HM = HAMLIN MERICLONE; Val M = Valencia mericlone; HA or HX = Hamlin somaclones; VA or VX = Valencia somaclones; SF = shaddette; Z = minigrafted on Carrizo; Adaxi = adaxial leaf surface; Abaxi = abaxial leaf surface; A, B, and L = those respective color axes as measured by HunterLab colorimeter; Rl = A/B; LE = length; D = diameter; R2 = spine LE/D; W = width; R3 = leaf LE/W; In Lg = internode length; Oil Gl = oil gland density. 0 = value not different from HM with a = 0.01; + = value different and greater than HM with a = 0.01; = value different and less than HM with a = 0.01.

PAGE 96

89 comparisons. Several somaclones were distinctly separated from the MERICLONE by adaxial leaf color during the retest (Tables 4.2, A. 13, and A. 15). Significantly, the three somaclones with the greatest differences in the second evaluation were indicated to be normal for the same characteristics during the first evaluation (Tables 4.2 and A. 7), and the two somaclones that were most clearly distinguished from the MERICLONE for these characteristics during the first test were indicated to be normal in the second morphological evaluation. No somaclones were found to be different from the HAMLIN MERICLONE for leaf length, leaf width, and leaf shape during the second test. Two somaclones differing from the MERICLONE for spine ratio during the first test, HA142E215 and HX1X12, differed similarly for this same characteristic during the second test (Table 4.2). One somaclone differing from the MERICLONE for abaxial leaf color ratio during the first test (HA7E12Z) differed similarly during the second test. These were the only three examples of confirmed morphological differences between 'Hamlin' somaclones and the HAMLIN MERICLONE. Six somaclones determined to be aberrant during the first morphological evaluation could not be reevaluated because of poor growth. It cannot be definitively determined whether the three somaclones that were confirmed morphologically aberrant were genetically different from 'Hamlin'. Growth to fruiting and/or extensive molecular characterization may produce more conclusive evidence. These results indicate that the frequency of detectable morphological mutants among 'Hamlin' somaclones produced by these techniques was less than 8-24% (3-9 of 37). Forty-seven 'Carrizo' somaclones were evaluated for morphological characteristics and compared with the Z2 MERICLONE (two plants) derived

PAGE 97

90 from the same seedling. Mericlones from two other 'Carrizo' seedlings (Zl and Z3) and seven 'Carrizo' plants regenerated from decapitated seedlings were used as additional controls. The left, main, or right leaflet lengths and leaflet ratios of the 'Carrizo' somaclones were not observed to differ significantly from those predicted by the 12 MERICLONE. Only one to three of the 47 somaclones were significantly different from the Z2 MERICLONE at a = 0.05, and one or none was different at = 0.01 (Tables 4.3, A. 18, and A. 20). It was concluded that there was no evidence for additional variation in leaflet lengths or leaflet ratios induced by the tissue culture process. In addition, it should be noted that one of the two other mericlones tested, the Z3 MERICLONE, was also separated from the 12 MERICLONE (at a = 0.01) by left and right leaflet lengths, indicating that morphological variation among seedlings may be at least as great as morphological variation among somaclones. The number of 'Carrizo' somaclones with aberrant spine lengths and spine L/D ratios were similar to those with aberrant leaflet lengths and not different from those predicted by the Z2 MERICLONE distribution (Tables 4.3 and A. 19). In contrast, 19 of 41 somaclones (46%) were separated from Z2 MERICLONE on the basis of spine diameter with = 0.05 (Table A. 19). However, the significance of this finding is questionable because three of the six controls were also separated from the Z2 MERICLONE by Duncan's test on the basis of spine diameter. Perhaps the number of repetitions of the control Z2 MERICLONE (six from one plant) was insufficient to provide a good estimate of the morphological variability observed within the clone. Nonetheless, there was no evidence of a

PAGE 98

91 Table 4.3. Summary for morphological comparisons of CARRIZO Z2 MERICLONE with 2 other mericlones, 7 seedlings and 47 somaclones. See Tables A.18-A.20 for complete results. Leaf Spine Ch CI one LL ML RL R4 LE D_ £2 Re Zl mericlone 0 0 0 0 0 0 0 0 Z3 mericlone + 0 + 0 0 + 0 0 Zll seedling 0 0 0 0 0 Z12 seedling 0 0 0 0 0 Z13 seedling 0 0 0 0 0 Z14 seedling 0 0 0 0 0 + 0 0 Z15 seedling 0 0 0 0 0 0 0 0 Z17 seedling 0 0 0 0 0 0 0 0 Z19 seedling 0 0 0 0 0 0 0 0 Z2E1CB317 0 0 0 Z2E235 0 0 0 0 0 + 0 0 Z2E236 0 0 0 0 0 + 0 0 Z2E239 0 0 0 0 0 Z2E240 0 0 0 0 0 + 0 0 Z2I153 0 0 0 0 0 + 0 0 Z2I157 0 0 0 0 0 0 0 Z2I166 0 0 0 0 0 + 0 0 Z2I167 0 0 0 0 0 Other Z2 Somaclones 33 other 0 0 0 0 0 0 0 0 5 other 0 0 0 0 0 MZ2 = Carrizo 2 Mericlone; Z2 = Carrizo 2 somaclone; LL= left leaflet length; ML = main leaflet length; RL = right leaflet length; R4 = (LL+RL)/ML; LE = length; D = diameter; R2 = spine LE/D; Ch Re = chirality reversals 0 = value not different from MZ2 with a = 0.01; + = value different and greater than MZ2 with a = 0.01; = value different and less than MZ2 with = 0.01.

PAGE 99

92 greater frequency of morphological variants among the 'Carrizo' somaclones than among the population of 'Carrizo' seedlings. The final morphological characteristic measured on the 'Carrizo' somaclones was chirality at the nodes (i.e., whether the bud was to the left or right of the spine). The actual chirality of a shoot was probably not an important indicator of genetic variation, because it is known that different seedlings of the same genotype, and even different growth flushes of the same plant, vary in their direction of spiral (Schneider, 1968; Schroeder, 1953). However, mericlones were never observed to possess nodes with different chirality in the same growth flush and observations of such chirality reversals in some somaclones were considered a possible indicator of morphological instability that was probably not linked with genetic alterations. Numbers of chirality reversals in six nodes (maximum reversal s=5) were counted for the 'Carrizo' somaclones (Table A. 20.) and compared to a table of binomial distributions. Only two (of 47) somaclones had more chirality reversals than that predicted at a = 0.05 (greater than three reversals). The number of chirality reversals in the somaclonal population was not different from expected. The morphological evaluations of 'Hamlin' and 'Carrizo' somaclones provided some evidence that more individual somaclones than expected were outside the normal distribution predicted by the mericlones. This may indicate the occurrence of mutations or a greater variance for some characteristics in populations of Citrus somaclones regenerated by organogenesis than in mericlone populations from the same explant material. An increase in the variability of quantitative traits has been

PAGE 100

93 reported for somaclones of other species (Daub and Jenns, 1989; Larkin, 1985; Ozias-Akins et al., 1989). When individual 'Hamlin' variants were retested after a substantial regrowth period, most of these somaclones were morphologically normal. These observations indicate that at least part of the enhanced variability in the somaclone populations was nongenetic variation of unknown origin. The 'Carrizo' somaclone population contained more aberrant plants than expected from the normal distribution predicted from the control meri clone for only one character (spine diameter). The seedling 'Carrizo' population demonstrated a similar frequency of deviation for the same characteristic. The frequency of morphological variants among 'Carrizo' somaclones may be no greater than the frequency of variants among nucellar seedl ings. The extensive morphological evaluation of these 89 somaclones did not provide evidence that the production of Citrus plants by the described organogenic procedures produced frequent or striking morphological variants or mutations. Minor genetic alterations may occur in individual Citrus somaclones and rare mutations may alter gross morphological features. One tetraploid from a confirmed diploid seedling was recovered in a 'Hamlin' somaclone population produced during subsequent studies and not included in the morphological analysis. Other Citrus plants regenerated from tissue culture have differed morphologically from the cultivar from which they were derived (J.W. Grosser, personal communication). However, the results of the present study indicated that such morphological variation in 'Hamlin' only occurred at a low frequency, and in 'Carrizo' did not occur at a frequency greater than that observed

PAGE 101

94 in seedling populations. This evidence suggests that producing Citrus somaclones by organogenesis is not an efficient means of generating genetic variation that is obvious in juvenile plants. However, induced mutagenesis in vitro by gamma irradiation or other mutagens may increase the frequency of mutation in tissue culture material to a more useful level Preliminary experiments were conducted to evaluate the use of gamma radiation for mutagenesis of embryogenic and organogenic Citrus cultures. 'Ridge Pineapple', and 'Hamlin' embryogenic callus cultures produced normal embryos after exposure to 3, 6, or 12 krad of gamma radiation from a Cobalt-60 source. 'Milam' callus that had been irradiated with 24 krad produced green embryoid structures that did not develop normally nor regenerate plants. Plants from 'Hamlin' callus that received 3 and 6 krad, and plants from 'Ridge Pineapple' callus that received 12 krad appeared healthy and were transferred to soil. Morphological evaluations of these plants were not undertaken. Gamma irradiation of segments greatly affected explant response of 'Ridge Pineapple', 'Hamlin', 'Carrizo', and US119. Initial fresh weights of 7 mm explants averaged 13 g for 'Ridge Pineapple' (N=14) and 15 g for US119 (N = 36). Average explant fresh weight of the four cultivars after 40-42 days on CP1 medium was 226 g (N = 32), while irradiation with 12 krad six days following transfer to the medium reduced the average explant weight of the four cultivars after 40-42 days to 67 g (N = 28) (Figure 4.2). Significant cultivar differences were observed (Table 4.4) with eight repetitions per treatment (except US119 at 24 krad, where N = 4). Callus growth from selection US119 was much more sensitive to gamma

PAGE 102

95 irradiation than that from 'Carrizo'. Exposing callus that had been growing on CP1 medium for 46 days to gamma radiation affected the initiation of shoots by the callus after transfer to SIM medium. This is demonstrated by the organogenic ratings of callus 70 days post-irradiation and 65 days after transfer to SIM medium (Figure 4.3). A subjective rating scale was used to evaluate health of cultures: 1 = all callus dead, 2-4 = intermediate, and 5 = adventitious shoots present. Both methods of evaluation indicated that a dose of 3-5 krad would give about a 50% reduction in measured response. This radiation dose may be the most efficient to use for the production of mutants from organogenic Citrus cultures. The results of this investigation did not indicate that regeneration by organogenesis from call used epicotyl or internodal explants resulted in substantial morphological variability among the resulting Citrus plants. A more definitive indication of the value of somaclonal variation for mutation breeding and the problems it might create for genetic engineering of Citrus must await fruiting of somaclone populations. The preliminary studies with embryogenic and organogenic cultures indicated that exposure to moderate levels of gamma radiation results in significant reduction in callus growth, organogenesis, and embryogenesis, but plants may still be regenerated. Gamma irradiation of Citrus callus cultures may be a valuable method of increasing the frequency of mutations.

PAGE 103

96 Figure 4.2. Effect of gamma radiation on callus fresh weight increase (N = 8 for all treatments except US119 at 24 krad, where N = 4). Table 4.4. Fresh weight of callused explants 40-42 days postirradiation. Cultivar means were separated using Duncan's multiple range test within krad levels ( = 0.05) as indicated by letters following values. Krad Carrizo Haml in Ridqe US119 Mean Prob.>F 0 157 b 210 b 271 a 268 a 226 0.0005 3 145 145 145 152 147 0.9855 6 119 a 110 ab 89 be 81 c 100 0.0263 12 80 a 71 a 72 a 46 b 67 0.0319 24 90 a 59 b 57 b 31 c 63 0.0001

PAGE 104

97 gure 4.3. Effect of gamma radiation on organogenic rating: 1 = all callus dead, 2-4 = intermediate, and 5 = adventitious shoots present.

PAGE 105

CHAPTER 5 CITRUS FRUIT SECTOR CHIMERAS AND THEIR POTENTIAL VALUE AS A GENETIC RESOURCE Introduction Many of the commercially important Citrus cultivars, such as 'Redblush' grapefruit (C. *paradisi Macf.; probably the same as 'Ruby Red') and 'Rohde Red Valencia' sweet orange (C. sinensis [L.] Osbeck), arose as somatic mutations in previously existing selections (Hodgson, 1967; Mendel, 1981; Stewart et al 1975). This fact, along with long juvenility and extensive nucellar polyembryony, has prompted increasing interest among Citrus breeders in utilizing natural and induced somatic mutations to obtain genetically superior cultivars (Hearn, 1986; Hensz, 1981; Russo et al 1981). Citrus fruit sector chimeras appear to be a type of somatic mutation (Cameron and Frost, 1968; Frost, 1943; Shamel 1943; Toxopeus, 1933) but have received very little attention as a potential source of favorable genetic changes. Iwamasa et al (1977) reported that the yellow-rinded sector of one chimeric fruit of the cultivar 'Fukuhara' produced seedlings that bore fruit with entirely yellow rinds. The normal orange-rinded sector of the same chimeric fruit produced seedlings that bore fruit with entirely orange rinds. Similar sectored fruit have been observed among many Citrus cultivars, and some of these bear mutant sectors that appear to have improvements in rind color or an increase in resistance to various 98

PAGE 106

99 pests or diseases (personal communications, G.A. Moore and W. Grierson). Reports of an increased frequency for at least one type of sector mutation in response to treatment of flower buds with hydrogen cyanide (Sinclair and Lindgren, 1943) or chlorpyrifos (M.L. Arpaia, personal communication) suggest that the phenomenon may be manipulated by chemical applications to optimize the recovery of favorable mutations. This study was initiated to investigate the potential for sectored Citrus fruit as a source of mutations to be used in genetic improvement. The objectives of this study were to: 1. Determine the approximate frequency and types of fruit sector mutations of some common Citrus cultivars by observations of graded fruit at commercial packinghouses; 2. Evaluate the potential for occurrence of desirable sector mutations; 3. Determine whether any plants regenerated from chimeric sectors were genetically abnormal. Materials and Methods Chimeric Citrus fruit were collected between November, 1988, and April, 1989, at two commercial fresh-fruit packinghouses near Lake Alfred, Florida. The cultivars sampled were 'Pineapple', 'Hamlin', and 'Valencia' sweet oranges (C. sinensis), 'Orlando' tangelo (C. xparadisi x C. reticulata Blanco), 'Marsh' and 'Redblush' grapefruit (C. xparadisi), 'Temple' (probably C. reticulata x C. sinensis) and 'Murcott' (probably C. reticulata C. sinensis). Preliminary observations indicated that most sectored Citrus fruit were sorted into the number two grade and

PAGE 107

100 eliminations categories by graders during normal packinghouse operations, and relatively few chimeras would be obtained from the number one grade or culls (see Soule and Grierson [1986] for a description of grade standards). Total numbers of fruit run through the packinghouse (all categories) during a sampling period were calculated on the basis of packinghouse records (number of field boxes run) and periodic sample counts (average number of fruit per box). To obtain an estimate of the frequency of sectored fruit among the different cultivars, all of the chimeric fruit were counted from the number two grade and eliminations categories while a known number of fruit was being run through the packinghouse. This method of sampling chimeric fruit was considered to yield a good estimate of the frequency and types of chimeras recovered from each cultivar sampled. For some cultivars, elimination and number two grade fruit were combined during packinghouse operations and only combined data were obtained. During the counting of chimeric fruit at the packinghouses, interesting or potentially useful fruit were collected and returned to the lab for more careful examination and/or plant regeneration. Rind; flesh, and juice color were determined for normal and mutant sectors of selected chimeric fruit with a HunterLab Citrus Colorimeter model D25 (HunterLab, Reston, VA) using standard procedures (Redd et al., 1986). Rind puncture resistance was determined by an Instron model 1122 (Instron Corp., Canton, MA). Juice acidity and soluble solids were estimated by titration and refractometry, respectively (Redd et al., 1986).

PAGE 108

101 Seeds extracted from mutant and normal sectors of some fruit were planted in soilless potting mix. When potentially useful chimeric sectors lacked viable seeds, undeveloped or aborted seeds were surface sterilized with a 1.05% solution of sodium hypochlorite and placed on EME medium (see Chapter 4). A small percentage of the undeveloped seeds from most sampled fruit eventually began to develop and produced viable embryos. Embryos that germinated by root and/or shoot elongation were subsequently transferred to soilless potting mix. Seedling ploidy levels were determined by chromosome counts from root tips using a modification of the hematoxylin stain technique described by Sass (1958). The modified staining technique was developed by Professor X.X. Ling and is described below. Young, white root tips were collected from plants growing in controlled environment chambers with a 16 hour photoperiod. Fresh root tips were washed in distilled water and placed in a saturated aqueous solution of 1,4-dichlorobenzene for 2-3 hours at room temperature. Then they were blotted dry and fixed in Carnoy's fluid (15 ml absolute [100%] ethyl alcohol: 5 ml glacial acetic acid) for 2-24 hours at room temperature. Following fixation, the root tips were hydrolyzed for 20 minutes in 5N HC1 at room temperature and rinsed in distilled water. Then the tissues were transferred to a 4% solution of ferric ammonium sulfate for 2-4 hours at room temperature and thoroughly washed five times (five minutes each) in distilled water. Finally, tissues were stained in a 5% hematoxylin solution (0.5 g hematoxylin dissolved into 5 ml 95% ethanol then 95 ml boiling water added to the hematoxyl in-alcohol solution) for 2-4 hours and washed with distilled water twice. Samples were examined

PAGE 109

102 by light microscopy at 400 to 1000 magnification after squashing onto a glass slide with 45% acetic acid. Chromosome number could be readily determined in cells undergoing active division. Results and Discussion Sectored chimeric fruit were defined as those fruit with one or more abnormal sectors and were readily distinguished from other types of mottled, scarred, or deformed fruit by straight margins diverging from the pedicel, widening near the equator of the fruit, and converging again to the stylar scar. Fruit with more than one distinct abnormal sector on the same fruit were not uncommon and were counted as a single chimera. These multiple chimeric fruit were classified according to the predominant sector change. Frequently, multiple chimeric sectors on the same fruit were either similar to each other in type or were opposite in effect (e.g. dark-red and light yellow sectors on normal pink grapefruit; see discussion below) Substantial differences were observed in the frequency of chimeras among the eight cultivars that were examined (Table 5.1.). Chimeras were more frequent among the sweet orange cultivars 'Valencia', 'Pineapple', and 'Hamlin' than in the two grapefruit cultivars 'Redblush' and 'Marsh'. The mandarin hybrids were the most variable group, producing the highest ('Orlando' 0.271%), the mean ('Temple' 0.11%), and the lowest ('Murcott' 0.009%) frequencies of fruit sector chimeras. Repeated samples were not available for statistical comparison of frequencies for the different cultivars. However, three samples of eliminations 'Hamlin' fruit from one block and a fourth sample from a second block were examined in the

PAGE 110

103 Table 5.1. Number and frequency of fruit sector chimeras from eight cultivars. Cultivar Sample ^i 7P x 10 3 Number fruit with rh impair n I ill i 1 1 I (lie l I \* sectors Percent 1*11 1 MIC I 1 \* fruit Orlando 139 377 0.271 Valencia 337 664 0.197 Pineapple 430 720 0.167 Tempi e 181 200 0.110 Haml in 399 327 0.082 Redblush 161 66 0.041 Marsh 117 20 0.017 Murcott 519 46 0.009 packinghouse. Numbers of chimeric fruit in the three first block were 145 in 399,000 (total fruit run), 142 in 360,000, and 98 in 220,000, while the sample from the second block yielded 82 chimeras from a sample of 211,000 total fruit run. These represented frequencies of 0.036%, 0.039%, 0.045%, and 0.039%, respectively. The great similarity in frequency among the samples from the same cultivar suggests the possibility that chimera frequency may be a cultivar characteristic. Comparison of chimeras counted in number two grade (Table 5.2) with those observed in the eliminations (Table 5.3) indicated that the latter category contained more of the mutant fruit. However, both groups yielded large numbers of chimeras of all types and would probably be equally valuable in the search for useful mutations. Collection of sectored fruit

PAGE 111

104 Table 5.2. Number of fruit sector chimeras of different types in number two grade. Sample Number fruit with chimeric sectors size Red/ Black/ Cultivar x 10 3 Gioas z Sunken Orange Green Brown Other Orlando 139 46 10 1 34 86 Pineapple 430 39 45 5 11 71 102 Hamlin 399 57 17 0 12 90 6 Redblush 161 4 13 4 16 7 Murcott 519 1 10 3 15 0 z Gigas = sectors with raised, thickened rind. Table 5.3. Number of fruit sector chimeras of different types in eliminations. Sample Number fruit with chimeric sectors size Red/ Black/ Cultivar x 1Q 3 Gigas z Sunken Orange Green Brown Other Orlando 139 16 2 1 0 46 144 Pineapple 430 69 73 7 5 120 173 Hamlin 399 54 6 0 4 76 5 Redblush 161 4 5 3 3 12 4 Murcott 519 1 10 0 17 7 z Gigas = sectors with raised, thickened rind.

PAGE 112

105 from grading lines at commercial packinghouses was noted to be a much more efficient way to obtain large numbers of sector chimeras than searching among fruit on the trees. Arrangements could probably be made with cooperative packinghouse managers for instructing graders to collect particular types of sectored fruit. This would make it possible for the researcher to recover even extremely rare mutations with a minimum of labor. Many different types of sector changes were observed in chimeric Citrus fruit. Sectors having alterations in rind color were most common and included black, brown, green, yellow, white, darker-red, and darkerorange. Numbers of fruit in some of these categories are indicated in Tables 5.2 to 5.4. Darker-orange or darker-red sectors were particularly Table 5.4. Number of fruit sector chimeras of different types in combined eliminations and number two grade. Sample Number fruit with chimeric sectors size Red/ Black/ Cultivar x 1Q 3 Gigas* Sunken Orange Green Brown Other Valencia 337 102 86 19 61 189 207 Temple 181 21 35 6 34 74 30 Marsh 117 6 0 0 0 6 8 2 Gigas = sectors with raised, thickened rind.

PAGE 113

106 interesting because these characteristics would be favorable changes in many sweet orange and grapefruit cultivars. These darker pigmented sectors were observed among the fruit of 'Orlando', 'Pineapple', 'Redblush', 'Valencia', and 'Temple' (Tables 5.2 to 5.4). Some 'Hamlin' fruit with dark-orange sectors were recovered during a later study not reported in this text. One dark-red sectored 'Redblush' grapefruit was examined by several laboratory methods (Table 5.5). This particular chimera had both darker-red and lighter-yellow sectors on the otherwise normal faint-pink rind of the fruit. The rind of the mutant dark-red sector was much more highly pigmented than the normal part of the fruit, as indicated by the nearly 6-fold increase in the A/B ratio measured by a HunterLab Colorimeter. Flesh color of this chimeric sector was darkerred as well; this was documented by the near doubling of the flesh A/B ratio and the Citrus Redness (Ting and Rouseff, 1986) parameter of the juice color. Rind of the dark-red sector was also substantially more resistant to puncture than the normal part of the fruit. Other characteristics of the dark-red sector (soluble solids, acid, and brix/acid ratio) were not significantly different from the normal part of the fruit. The light-yellow sector of this fruit was too narrow to allow measurement of any parameters other than puncture resistance. However, it is interesting to note that this sector was less resistant to puncture than the normal part of the fruit, a change opposite to that observed in the adjacent dark-red sector. This may be additional support for the hypothesis that the mechanism of mutation in this case was somatic segregation (see discussion below).

PAGE 114

107 Table 5.5. Summary of internal and external characteristics of mutant and normal sectors of 'Redblush' chimera 11. Normal Dark-red Yellow sector sector sector Rind color A/B ratio 0.13 0.72 nd (s.d.) (0.06) (0.04) Flesh color A/B ratio 0.77 1.25 nd (s.d.) (0.08) (0.14) Juice color CR 17.7 31.5 nd CY 31.4 36.8 nd N 28.3 31.5 nd Soluble solids brix 7.9 8.1 nd (s.d.) (0.12) Acid 1.15 1.08 nd (s.d.) (0.13) TCB/acid ratio 7.0 7.7 nd Puncture resistance 2.4 3.3 1.9 (s.d.) (0.3) (0.3) (0.4) s.d. = standard deviation; nd = no data; A/B = color ratio (Ting and Rouseff, 1986); CR = citrus redness by Hunter colorimeter (Ting and Rouseff, 1986); CY = citrus yellow by Hunter colorimeter (Ting and Rouseff, 1986); N = color score (Redd et al 1986); TCB = total corrected brix (Redd et al 1986).

PAGE 115

108 Green sectored fruit were another type of color chimera that was observed among all cultivars except 'Marsh'. Green sectors observed in packinghouse fruit may be caused by resistance to degreening, or a lack of carotenoid or lycopene pigment development. In general, these would be unfavorable mutations. However, green sectors may also signify changes in the fruit maturation process. There could be some benefits in the development of clones that matured later but were otherwise identical to existing selections. One chimeric fruit with a very wide green sector was discovered on a navel orange tree in the field. This sector could be compared with normal fruit from the same tree as well as with the normal part of the chimeric fruit (Table 5.6). Navel orange is a cultivar group in which selections with different maturity dates have proven very valuable in the Australia (Tolley, 1989) and California citrus industries (Pehrson and Ivans, 1988). The A/B ratio of the rind from the green sector was measurably different from the normal part of the chimeric fruit (Table 5.6), while the rind color of the normal sector was essentially the same as that of the normal fruit from the same tree. The A/B ratio of the flesh in the green sector was less than that of the normal part of the chimeric fruit, but the differences were not as great as in the rind. Other characteristics of the green sector indicated less maturity (less juice color, more acid, lower soluble solids) than normal fruit from the same tree, but in general the internal characteristics of the normal sector of this particular fruit more closely resembled the green sector than it resembled the normal fruit from the same tree. One possible explanation for this observation is that the different flesh sections of a Citrus fruit may be associated in the expression of the characteristics

PAGE 116

109 Table 5.6. Summary of internal and external characteristics of mutant and normal sectors of navel chimera 11. Navel chimera 11 Normal Green Normal fruit sector sector from same tree Rind color A/B ratio 0.41 -0.07 0.43 (s.d.) (0.13) (0.18) (0.03) Flesh color A/B ratio 0.15 0.08 0.20 (s.d.) (0.07) (0.04) (0.03) Juice color CR 29.8 26.5 35.4 CY 68.7 64.6 73.4 N 35.0 34.0 36.4 Soluble solids brix 10.8 10.6 11.5 (s.d.) (0.14) (1.25) Acid 0.95 0.92 0.58 (s.d.) (0.04) (0.10) TCB/acid ratio 11.6 11.7 20.0 Puncture resistance 1.8 2.3 1.9 (s.d.) (0.1) (0.2) (0^2) s.d. = standard deviation; A/B = color ratio (Ting and Rouseff, 1986); CR = citrus redness by Hunter colorimeter (Ting and Rouseff, 1986); CY = citrus yellow by Hunter colorimeter (Ting and Rouseff, 1986); N = color score (Redd et al 1986); TCB = total corrected brix (Redd et al 1986).

PAGE 117

110 of maturity, and a mutation expressed in 75% of the flesh tissues may effectively alter the maturity of the entire fruit. Alternatively, the mutation may not be restricted to the flesh tissues beneath the mutant rind sector and may be present throughout the flesh. The most common color changes observed in sector chimeras were to black or brown (Tables 5.2 to 5.4). Many of these chimera mutations appeared to result in increased rind susceptibility to one or more pests or diseases. This study did not recover any sectored fruit where the mutation appeared to result in increased pest resistance. However, chimeric sectors with resistance to melanose (Diaporthe citri [Fawc] Wolf) and rust mite (Phyllocoptruta oleivora Ashmead) have been observed by other workers (W. Grierson, personal communication) and may be an important genetic resource. Many fruit sector chimeras were not observed as changes in color, but rather as changes in rind texture or thickness. The two most common of these mutations (Figures 5.2 to 5.4) led to increased rind thickness (termed "gigas") and decreased rind thickness (termed "sunken"). Seedlings recovered from gigas sectors of three 'Valencia' and one 'Orlando' chimeric fruit were tetraploid (2n=4x=36). No seedlings were obtained from normal sectors of these same chimeric fruit. However, seedlings obtained from normal sectors of other gigas chimeric fruit were diploid (2n=2x=18). Because the cultivars 'Valencia' and 'Orlando' are polyembryonic and seedlings are normally of nucellar origin, these observations suggest that gigas chimeras are, in fact, cytochimeras with the gigas sectors composed of tetraploid cells. Tetraploids obtained from mutations in nucellar tissue of polyembryonic cultivars are expected to

PAGE 118

Ill be the result of somatic doubling (autotetraploids) and this has been confirmed by identical isozyme banding patterns for the tetraploids and their counterpart diploid cultivars (data not shown). The generation of tetraploid seedlings from one type of sector chimera has provided evidence that mutations expressed in the rind of sectored fruit may be recovered in seedlings. Seedlings recovered from gigas sectors of monoembryonic cultivars (e.g. 'Temple') should be of zygotic origin. These zygotes should result from the fusion of a diploid (unreduced, ln=2x=18) egg and either pollen from the same flower sector (unreduced, ln=2x=18) or pollen from elsewhere (reduced, ln=lx=9 or unreduced, ln=2x=18). It will be interesting to see whether plants from gigas sectors of monoembryonic cultivars are diploid (unfertilized diploid egg or haploid egg plus haploid pollen), triploid (diploid egg + haploid pollen), allotetraploid (diploid egg + diploid foreign pollen), or autotetraploid (diploid egg + diploid self pollen). Triploid selections may be valuable as seedless cultivars, and tetraploid selections of existing cultivars may be useful as parents for the production of other seedless triploid hybrids (Soost and Cameron, 1968, 1980). Sunken sectors were often observed adjacent to gigas sectors on chimeric fruit, suggesting the possibility of an origin by somatic segregation. If this were the case, such sectors could be haploid (2n=lx=9) or aneuploid. Viable haploid or aneuploid Citrus plants would be of interest for some types of breeding and genetic studies. The origin of some fruit sector chimeras may be somatic segregation because occasionally two adjacent, contrasting sectors (twin sectors) were

PAGE 119

112 observed, such as dark-red and light-yellow sectors on normal slightlypink 'Redblush' grapefruit rind (Table 5.5), or the gigas-sunken twin sectors. In other cases, sector chimeras appeared to result from spontaneous polyploidization, as with the gigas chimeric fruit (without twin sunken sectors) from which tetraploid seedlings were recovered. It is also likely that some sector chimeras were the result of point mutations, transposable element activity, chromosome rearrangements, or other genetic mutations. Most reports of Citrus fruit sector chimeras suggested that they are sectorial, implying that all cells within the affected sector are of the mutant type (Cameron and Frost, 1968; Iwamasa et al 1977). It is likely that at least some fruit sector chimeras are mericlinal, with only the cell layer producing the rind being affected. The lack of darker flesh color in some dark-rinded chimeric fruit may be caused by the restriction of the mutation to the L-II layer; the flesh in these mericlinal chimeric fruit remains genetically and colorimetrically normal because it is derived from L-I tissue (Soost and Cameron, 1975). This effect is similar to the differential pigmentation in these tissues of the periclinal grapefruit cultivars 'Thompson', 'Foster', and Burgundy' (Soost and Cameron, 1975). Fortunately, Citrus nucellar embryos are derived from the L-II layer, as is the fruit rind (Soost and Cameron, 1975). Therefore, rind mutations that are mericlinal chimeras should be recovered in all three histogenic layers in seedlings. Seedlings grown from chimeric fruit are expected to be non-chimeric, whether normal or mutant, because zygotic and nucellar Citrus embryos originate from single cells (Bacchi, 1943; Osawa, 1912).

PAGE 120

113 Width of mutant sectors varied from less than one millimeter to almost the entire fruit. Wider sectors may indicate that the mutation occurred earlier during fruit ontogeny and thus is more likely to be found in the seed. However, sector width probably also is affected by the positive or negative influence of the mutation on cellular metabolism and tissue physiology. Boundaries of chimeric rind sectors aligned perfectly with internal fruit septa in some cases, but more often there was no alignment. It is unclear whether this characteristic has any bearing on type of chimera (mericlinal or sectorial) or likelihood of mutant recovery. Tetraploid seedlings were recovered from both aligned and unaligned gigas chimeras. The value of gigas fruit sector chimeras as a source of autotetraploid seedlings from commercial polyembryonic Citrus cultivars was demonstrated in this study. The reports of increased gigas sector formation following hydrogen cyanide fumigation (Sinclair and Lindgren, 1943) or chlorpyrifos application (M.L. Arpaia, personal communication) suggest that this phenomenon may be manipulated to maximize polyploid recovery. Tetraploids are valuable in breeding triploid seedless hybrids (Soost and Cameron, 1968, 1980). Observation of several other types of sector mutants that appear to be superior to the normal phenotype points out the potential usefulness of this resource. Mutants with darker fruit rind coloration would greatly enhance visual appeal of fresh fruit from many Citrus cultivars, thus reducing the need for ethylene degreening (McCornack and Wardowski, 1977) and application of dye (Kaplan, 1986) for consumer acceptance. Selections with increased flesh and juice color or altered maturity date could significantly increase grower income and

PAGE 121

114 improve marketability of Florida and California citrus. Finally, reported observations of pest and disease resistant sector chimeras suggest that this genetic resource may have tremendous potential for the production of selections that yield high quality fruit with minimal pesticide appl ication.

PAGE 122

CHAPTER 6 RESPONSE OF STEMS FROM IN VITRO-GROWN SEEDLINGS TO PHYTOPHTHORA PARASITICA IN DUAL CULTURES Introduction Phytophthora parasitica Dast. is a soil -borne pathogen of Citrus in most tropical and subtropical regions of the world. P. parasitica infection frequently results in extensive necrosis of the feeder roots of susceptible Citrus rootstocks (root rot) and may also cause extensive damage to the cambial tissue under the bark, called foot rot or gummosis (Timmer and Menge, 1988). Clones with resistance/tolerance to both diseases caused by Phytophthora are available within Citrus and among related genera (Carpenter and Furr, 1962; Grimm and Hutchison, 1977), but each of these selections has other undesirable attributes that limit their usefulness as rootstocks in commercial orchards (Castle, 1987; Castle et al, 1989). Although chemical control of the diseases is possible (Davis, 1981; Menge, 1986; Sandler et al., 1989; Timmer, 1985), a breeding program for Citrus should ideally include at least moderate resistance or tolerance to the Phytophthora diseases as one of the desirable attributes of any newly developed rootstock. Unfortunately, resistance to P. parasitica is probably controlled by a multiple gene system (Hutchison, 1985), and no rapid, simple test for accurate determination of Citrus susceptibility to the pathogen is known. Although several different tests of Phytophthora resistance/tolerance have been described for Citrus 115

PAGE 123

116 (Cameron et al., 1972; Carpenter and Furr, 1962; Graham, in press; Grimm and Hutchison, 1973; Smith et al 1987; Tsao and Garber, 1960; Whiteside, 1974), these procedures are very time-consuming and results do not always agree with observed field susceptibility. Observations of seasonal fluctuations in plant susceptibility (Matheron and Matejka, 1989) further complicate interpretation of short-term field or greenhouse testing. Some success has been achieved in characterizing or selecting for plant resistance to diseases in vitro (Behnke, 1979, 1980; Buiatti and Scala, 1985; Wenzel et al 1985). In vitro resistance of plant tissues to Phytophthora in dual cultures has been found to correlate with field performance for several crop species (Barritt et al., 1990; Dolan and Coffey, 1986; Helgeson et al 1976; Ingram, 1967; Jeffers et al 1981; McComb et al., 1987; Miller et al 1984; Tedford et al 1990; Utkhede and Quamme, 1988; Warren and Routley, 1970; Zilberstein and Pinkas, 1987). Investigations on the response of Citrus cultivars to culture filtrate from P. citrophthora were not particularly encouraging (Vardi et al., 1986). However, no investigations of in vitro response of Citrus cultivars in dual culture with P. parasitica have been reported. This project was initiated to investigate the possibility of using the in vitro reaction of Citrus seedling shoots to the pathogen as an indicator of field susceptibility to Phytophthora or as a technique for the study of resistance mechanisms under carefully controlled environmental conditions. The test conditions were not intended to imitate the response to either disease (root rot or foot rot) induced by P. parasitica on Citrus, but rather to examine general tissue response to the pathogen. The specific objectives of the study were to:

PAGE 124

117 1. Determine whether P. parasitica can invade excised shoots of Citrus seedlings in dual cultures; 2. Determine the response of seedling shoots from several Citrus selections to P. parasitica in vitro; 3. Compare these responses with the accepted field susceptibilities of mature trees. Materials and Methods Pure isolates of Phytophthora parasitica that had been recovered from Citrus orchards near Ft. Pierce, Florida (called P. parasitica 'Riverland') and Bartow, Florida (called P. parasitica 'Hall') were obtained from J.H. Graham. These pure P. parasitica cultures were maintained on corn meal agar (BBL Microbiology Systems, Becton Dickinson and Co., Cockeysville, MD) or as in vitro stem infections on Citrus shoots. Rate of P. parasitica 'Hall' growth on corn meal agar at 27C in continual darkness was estimated by measurement of distance from point of inoculation to culture margin at 2, 3, 4, 5, and 6 days after inoculation. Plant tissue culture medium for dual culture tests was inoculated with agar plugs from the actively growing margin of pure P. parasitica colonies on corn meal agar or by segments of in vitro-infected seedling shoots from other dual cultures. The completely defined plant tissue culture media RM1 and GM1 (see Chapter 3) supported the growth of a luxuriant surface lawn of P. parasitica during these tests. Experiments 1-3. Seeds of 'Valencia' and 'Hamlin' sweet oranges (Citrus sinensis [L.] Osbeck), 'Carrizo' citrange (C. sinensis Poncirus trifoliata [L.] Raf.) and 'Swingle' citrumelo (P. trifoliata x c.

PAGE 125

118 xparadisi Macf.) were surface sterilized by immersion and agitation in 70% ethyl alcohol for 10 minutes followed by 20% bleach (1.05% sodium hypochlorite plus two drops polyoxyethylene-20-sorbitan monolaurate [Tween 20: Fisher Scientific, Pittsburgh] per 100 ml) for 20 minutes. The seeds were rinsed in sterile distilled water five times (five minutes each rinse) and individual seeds placed on the surface of 10 ml GM1 medium (see Chapter 3) solidified in the bottom of 25 x 150 mm glass culture tubes (Bellco Glass, Inc., Vineland, NJ). The tubes were covered with translucent plastic Kap-uts closures (Bellco Glass, Inc.). Seedling and dual cultures were maintained at 27C under constant fluorescent lighting. At about three months (Experiment 1) or two weeks (Experiment 2) after seed germination, shoots were excised above the cotyledons and the base inserted several millimeters into solidified RM1 medium in Magenta GA-7 boxes (Magenta Corp., Chicago) that had been inoculated with P. parasitica 'Riverland'. In the preliminary study (Experiment 1), two 'Valencia' shoots were placed in inoculated medium and two were placed in uninoculated medium as controls. In Experiment 2, shoots of the cultivars 'Swingle' (2 shoots), Carrizo' (1 shoot), and 'Hamlin' (2 shoots) were excised and the bases of the stems were wounded by scalpel incisions before placement in the medium. Shoots were placed so that the wounded stem tissue remained in contact with the surface of the medium (where the Phytophthora grew most vigorously). The lengths of the resulting stem discoloration were recorded periodically during the following 12-24 days. Lesion growth rate was estimated by the slope of the linear regression of lesion length on time after placement for each cultivar.

PAGE 126

119 For Experiment 3, six shoots from each of the cultivars 'Swingle', 'Carrizo', and 'Hamlin' were excised above the cotyledons 3-4 weeks after seed germination, and wounded at the base by making an elongated diagonal cut from one side of the stem to the other with a scalpel. The base of each shoot was then placed in solidified RM1 medium so that the wounded stem tissue remained in contact with the surface of the medium. The RM1 medium was inoculated with P. parasitica 'Riverland' a few days before stem placement. The lengths of the resulting stem discoloration were recorded 4, 6, 8, 10, and 12 days after stem placement. Experiment 4 Outer seed coats were removed from seeds of 'Ridge Pineapple' and 'Valencia' sweet oranges (C. sinensis), 'Marsh' grapefruit (C. xparadisi), Sour Orange selection Florida DPI-WH 13/12 (C. aurantium L.), 'Cleopatra' mandarin (C. reticulata Blanco), Volkamer lemon (C. vol kameri ana Ten. & Pasq., or C. limon [L. ] Burm.f. 'Volkamer'), 'Carrizo' citrange, 'Swingle' citrumelo, P. trifoliata selection Florida DPI-WH 9/6, and Severinia buxifolia (Poir.) Tenore. Embryos were surface sterilized by a 30 second wash in 70% ethyl alcohol followed by immersion and agitation in 15% bleach (0.8% sodium hypochlorite plus 2 drops Tween per 100 ml) for 15 minutes. The seeds were then rinsed in sterile distilled water five times (five minutes each rinse), and individual seeds were placed on the surface of 10 ml GM1 medium solidified in the bottom of 25 x 150 mm glass culture tubes. The tubes were covered with translucent plastic Kap-uts and sealed with Nescofilm (Nippon Shoji Kaisha Ltd., Osaka, Japan). Seedling and dual cultures were maintained at 27C in continuous darkness.

PAGE 127

120 After shoots of each cultivar exceeded 70 mm in length (5-6 weeks from seed placement in tubes), they were excised above the cotyledons and 3-5 vertical slices made in the basal end of each shoot, leaving longitudinal flaps of stem tissue attached to the base of the stem. The basal end of each stem was then inserted several millimeters into solidified GM1 medium at the actively growing edge of a P. parasitica 'Hall' culture in a Magenta box. The medium in each Magenta box was inoculated with P. parasitica 'Hall' about five days prior to shoot placement in the medium. Shoots were arranged so that one shoot of each of the ten cultivars was in each of six Magenta boxes. The lengths of the resulting stem discoloration were recorded 2, 3, 4, 5, and 6 days after placement. Results and Discussion Phytophthora parasitica 'Hall' grew on corn meal agar as an expanding circle that spread out from the point of inoculation at a relatively uniform rate. Measurements of the distance (in mm) from point of inoculation to the margins of the Phytophthora growth were completed at 2, 3, 4, 5, and 6 days after inoculation on three Petri plates of corn meal agar. The rate of Phytophthora growth on this medium could then be estimated by linear regression of distance on time. The slope of the regression line indicated that the P. parasitica 'Hall' growth rate was about 3.8 mm-day" 1 on BBL corn meal agar (at 27C). During Experiment 1, P. parasitica 'Riverland' was observed to invade excised green shoots of three-month-old 'Valencia' seedlings in dual cultures. Invasion of the stems by Phytophthora could be observed

PAGE 128

121 as a distinct browning that moved from the agar surface (and shoot base) towards the shoot apex at a relatively constant rate. The mycelium of the fungus was rarely visible on the exterior of green shoots during the in vitro testing, and then only as small tufts at the nodes or the severed shoot apex. Growth of the mycelium from browned shoots on corn meal agar following the test indicated the presence of viable Phytophthora inoculum throughout the browned stems. Other green 'Valencia' shoots placed in uninoculated medium did not become discolored within 24 days. The rate of shoot browning or discoloration in response to Phytophthora infection was selected as one characteristic that may differ among genotypes and that may represent one aspect of relative cultivar susceptibility to the pathogen. This rate was estimated by measurement of the distance that the lesion (or discoloration) on a stem had progressed from the point of inoculation at several different periods of time after infection. Linear regression of these measurements produced a slope that should be a robust estimate of rate. This procedure also yielded a second characteristic for the regression line, the y-intercept, that may indicate other aspects of plant resistance to Phytophthora (such as the time after infection that the plant response was initiated). The y-intercept was not considered in the following discussions. Linear regression of lesion length (in mm) on days after placement of shoots on the inoculated medium produced estimates of 2.0 mm-day" 1 for the slope of three-month-old, green 'Valencia' shoots. A later attempt to infect excised green 'Valencia' shoots was not successful and was interpreted as a failure of the Phytophthora to invade the stem without an exposed wound surface. This prompted the wounding of the stem tissue at the medium surface for the ensuing experiments (2-4).

PAGE 129

122 During Experiment 2, P. parasitica 'Riverland' caused the progressive browning of wounded two-week-old green stems of the cultivars 'Hamlin' (two shoots), 'Carrizo' (one shoot), and 'Swingle' (two shoots). Estimates of slope were obtained by linear regression of lesion length (in mm) on time after placement for each cultivar. Mean slope values are shown in Table 6.1. If the rate of green stem browning in response to infection by Phytophthora (as measured by slope) is an indication of cultivar susceptibility, then the results of Experiment 2 indicated that 'Carrizo' was most susceptible, 'Hamlin' intermediate, and 'Swingle' the most resistant. Table 6.1. Mean regression slopes for excised green shoot browning from Phytophthora and separation by Duncan's multiple range test (o = 0.05). Experiment 2 Experiment 3 Cultivar Regression sloDe Regression Duncan slope qroupinq Swingle 1.5 mmday" 1 1.5 mmday" 1 a Haml in 2.7 2.6 ab Carrizo 8.0 3.8 b In Experiment 3, P. parasitica 'Riverland' caused the browning of wounded threeto four-week-old green stems of the cultivars 'Hamlin (six shoots), 'Carrizo' (six shoots), and 'Swingle' (six shoots). Slope for lesion growth rate was estimated by linear regression of lesion length (in mm) on time (4, 6, 8, 10, and 12 days) after placement in the medium for

PAGE 130

123 each shoot. The individual estimates of slope for each of the six shoots per cultivar were used as replicates in an analysis of variance (ANOVA) procedure and Duncan's mean separation. The ANOVA procedure for slope indicated significant cultivar differences (the probability of a greater F was 0.015). Duncan's mean separations for calculated slope values are shown in Table 6.1. On the basis of rate of green stem tissue browning in response to Phytophthora infection, (i.e., slope) during Experiment 3, 'Carrizo' appeared to be the most susceptible, 'Hamlin' was intermediate, and 'Swingle' most resistant. This ranking of the response of green shoots to Phytophthora for the three cultivars was the same as in Experiment 2. Browning response of etiolated seedlings following inoculation with P. parasitica 'Hall' varied considerably among the ten genotypes examined in Experiment 4. Regression of discoloration length on time indicated the response of most individual shoots was closely approximated by a straight line. The SAS PROC RSQUARE procedure was used to evaluate whether there was evidence that the cultivar responses had significant quadratic or cubic components. The six replicates for each cultivar were grouped together for this analysis. A linear model (independent variable = time in days; dependent variable = length in mm) produced the greatest or close to the greatest r 2 for each of the ten genotypes. Addition of quadratic (time 2 ) or cubic (time 3 ) components to the linear model produced insignificant increases in the r 2 value: The response was linear. The linear equations determined to best fit the response of each cultivar are given below, along with r 2 values.

PAGE 131

124 Carrizo y -17 + 10. 4x r 2 = 0.81 Swingle y -16 + 9.4x r 2 = 0.81 Valencia y = -6 + 7.7x r 2 = 0.55 Marsh y -6 + 6. Ox r 2 0.50 Sour y -4 + 5.9x r 2 0.27 Cleopatra y = -5 + 5.8x r 2 = 0.75 Ridge y -1 + 4.1x r 2 0.34 Vol kamer y 3.6x r 2 0.48 Poncirus y 4 + 0.2x r 2 0.04 Severinia y 3 + O.lx r 2 0.01 Graphical representation of these equations is shown in Figure 6.1. A great difference between the response of the two Citrus relatives, Poncirus and Severinia, and the Citrus cultivars and hybrids was noted. AN0VA was performed, and LSD mean separations of cultivars were calculated, by using the slope of each of the six shoots per cultivar as replicates. The ANOVA procedure indicated highly significant cultivar differences for slope (the probability of a greater F was about 0.0001). Calculated values for slopes are shown along with LSD mean separations in Fig. 6.2. At least three distinct types of etiolated shoot response to Phytophthora could be identified among the ten cultivars that were examined. The genotypes most resistant to Phytophthora were the Citrus relatives Poncirus trifoliata and Severinia buxifolia. These two selections exhibited very little discoloration of their stem tissue as indicated by the derived linear regression slope values of 0.2 mm-day" 1 for Poncirus and 0.1 mm-day" 1 for Severinia. This indication of very high

PAGE 132

125 50 40 30 mm 20 10 0 Carrizo Swingle Valencia Sour Marsh Cleopatra Ridge Volkamer Poncirus Severinia 6 8 Days Figure 6.1. Estimated linear response, as discoloration length, of the ten cultivars to Phytophthora 2-6 days after inoculation.

PAGE 133

126 mm/day LSD (T) Grouping Sev Pon Volk Ridg Cleo Sour Mar Val be be be cd Swin d Carr d Figure 6.2. Mean regression slopes for excised etiolated shoot discoloration from Phytophthora and LSD comparisons ( a = 0.05). resistance or immunity to P. parasitica for Severinia was in agreement with field performance and the results of other tests (Carpenter and Furr, 1962; Grimm and Hutchison, 1977; Klotz and Calavan, 1978). Poncirus trifoliata is generally considered highly resistant or immune to Phytophthora as well (Timmer and Menge, 1988), although some variation in resistance has been reported among different P. trifoliata selections (Carpenter and Furr, 1962; Grimm and Hutchison, 1977; Univ. Calif., 1978). It appeared that the selection of P. trifoliata used in this test (Florida DPI-WH 9/6) was highly resistant to etiolated stem infection. TheC/tn/scultivars Volkamer lemon, 'Ridge Pineapple', 'Cleopatra', sour orange, 'Marsh', and possibly 'Valencia' formed a group of selections

PAGE 134

127 with intermediate response to Phytophthora. These cultivars produced a response with regression slopes in the range of 3.6 to 7.7 mm-day" 1 Although some statistically significant differences were detected within the group, retesting will be necessary to conclude that these differences were meaningful. Generally, Volkamer lemon, 'Cleopatra', and sour orange are considered to possess low to moderate levels of Phytophthora resistance, while sweet oranges ('Valencia' and 'Ridge Pineapple') and grapefruit ('Marsh') are considered highly susceptible (Castle, 1987; Timmer and Menge, 1988). However, considerable variability in Phytophthora susceptibility has been observed among different sour (Hutchison and Grimm, 1972) and sweet orange selections (Smith et al., 1987), so it is quite uncertain what would be the field performance of the clones used in the in vitro test. The failure of the in vitro test to discriminate clearly among these six intermediate cultivars may indicate that there was little difference among these cultivars in response to the specific test conditions. However, mechanisms of tolerance or resistance only activated in green seedling shoots or not measured by this test (such as feeder root regeneration and resistance to initial injury) may play important roles in relative resistance or tolerance of these cultivars to Phytophthora under field conditions. The third type of response observed among the tested selections under these in vitro conditions was extreme susceptibility to Phytophthora. The response of 'Valencia' was overlapping with this group and the intermediate response described above. The rapid discoloration of etiolated stems of 'Carrizo' and 'Swingle' following inoculation with Phytophthora was indicated by regression slopes of 10.4 and 9.4 mm-day" 1

PAGE 135

128 respectively. This result was unexpected because 'Swingle' has been considered highly tolerant or resistant to Phytophthora (Castle, 1987; Carpenter and Furr, 1962; Graham, in press; Timmer and Menge, 1988), while 'Carrizo' has been observed to have a lower, but still significant, degree of resistance or tolerance (Castle, 1987; Grimm and Hutchison, 1977; Timmer and Menge, 1988). The three in vitro tests involving these two cultivars (Experiments 2-4) did, however, agree with their relative field performance in that 'Swingle' was consistently more resistant than 'Carrizo' to discoloration from Phytophthora. The observed great susceptibility of 'Swingle' and 'Carrizo' during in vitro testing of etiolated shoots was especially surprising because these two cultivars are hybrids of Poncirus trifoliata (indicated to be highly resistant in this test) and were considered to have obtained their resistance to Phytophthora from this parent. The fact that both hybrids produce a similar response, however, reduces the probability that these unexpected results were due to experimental error. It is possible that this test of in vitro-grown etiolated shoots avoids the resistance mechanism(s) of 'Carrizo' and 'Swingle'. The resistance mechanisms in these selections appeared to be different from the predominant one(s) in Poncirus and may require light or greater seedling maturity for activation. The observation that non-etiolated 'Hamlin' sweet orange (considered highly susceptible to Phytophthora) was more resistant than non-etiolated 'Carrizo' (considered moderately resistant) in Experiments 2 and 3 supports the hypothesis that maturity may be an important factor controlling activation of resistance in 'Carrizo'. Graham (in press) reported a reduction in root rot among in vivo seedlings with increasing

PAGE 136

129 age. However, a possible role for light in the activation process was suggested by the fact that green shoots of the cultivars 'Swingle' and 'Carrizo' both appeared considerably less susceptible to Phytophthora than when etiolated, as indicated by discoloration rate. It may be that both light and tissue maturity are involved in the activation of Phytophthora resistance in 'Carrizo' and 'Swingle'. Yet another explanation for the extreme sensitivity of 'Carrizo' and 'Swingle' to P. parasitica under these test conditions is that the resistance to Phytophthora observed in 'Carrizo' and 'Swingle' under field and other test conditions may be due to an ability to inhibit initial invasion of the plant tissues at wound sites. However, if the pathogen is able to penetrate this first line of defense under conditions especially favorable to Phytophthora (e.g. in vitro inoculation of severely wounded stems), and the host has limited physiological resources (e.g. no photosynthesis or root reserves), the remaining tissues will have little capacity to resist rampant in planta proliferation of Phytophthora. Several aspects of these in vitro tests should be noted. First, two different isolates of P. parasitica were used because of contamination of all 'Riverland' cultures a few weeks before the initiation of Experiment 4. To avoid a long delay in reisolation of pure 'Riverland' cultures, a second isolate, 'Hall', was substituted in the experiment. Although different isolates of P. parasitica are not known to vary in virulence on Citrus (personal communication, J.H. Graham), it cannot be ruled out that isolate differences may have influenced cultivar responses. Second, the three cultivars used in the in vitro tests of green shoots (Experiments 2 & 3) proved to be the three most susceptible among those examined in the

PAGE 137

130 etiolated test (assuming 'Hamlin' sweet orange = 'Valencia' sweet orange); Only preliminary conclusions can therefore be reached concerning the comparison of green to etiolated tissues. Third, growth of Phytophthora on excised shoots was usually restricted to the interior of the stem, and measurement of cultivar response was based entirely on stem discoloration (etiolated shoots) or browning (green shoots). Occasional small tufts of mycelium observed at nodes or the excised apex of infected discolored shoots, especially common in etiolated 'Carrizo' stems, suggested that discoloration or browning was providing a reasonable indication of Phytophthora movement through the tissues. In addition, short segments cut from aerial sections of browned shoots were reliable sources of Phytophthora inoculum. However, plant response (as discoloration or browning) was the measured parameter, and was not necessarily the same as the degree of Phytophthora colonization of the stem tissue. This work demonstrated that two isolates of Phytophthora parasitica can successfully invade wounded excised shoots of Citrus seedlings in dual cultures. Rates of shoot discoloration were determined, in vitro, for etiolated seedling shoots of ten Citrus cultivars and relatives, and for green seedling shoots of three cultivars. For many cultivars, the observed relative responses to Phytophthora were in agreement with accepted field responses and other testing methods (e.g. Poncirus, Severinia, Volkamer lemon, 'Cleopatra', sour orange, 'Marsh', and 'Valencia'). In other cases, the observed response to Phytophthora was nearly opposite to known field resistance (e.g. 'Swingle'). The failure of the method to produce results entirely in agreement with field performance, as well as the unexpectedly large amount of work it requires,

PAGE 138

131 makes it unlikely to be useful as a screening procedure in breeding efforts. However, other types of tissue may yield results that correlate better with field response. In addition, the method may provide a precise way of measuring some plant-pathogen interactions (e.g. hypersensitivity) under carefully controlled conditions for investigations of resistance mechanisms.

PAGE 139

CHAPTER 7 SUMMARY AND CONCLUSIONS The introductory chapter pointed out the achievements and potential of several approaches to Citrus genetic improvement. Regardless of the ultimate value of these methods, it seems certain that they will not solve all the problems facing contemporary citri culture. In addition, most of the popular approaches to cultivar development focus on methods of generating genetic changes and devote relatively few resources to improving techniques of germplasm evaluation. Considerable benefit could be derived from the development of other, as yet unexploited, sources of genetic variation and methods of evaluation. The research described in this dissertation has completed one stage of investigations on five topics in this category: a diverse semi-wild population with zygotic embryony as a source of potential breeding parents, the morphologic and potential genetic variation in plants regenerated from tissue culture, the response of tissue cultures to gamma irradiation as a source of mutations, fruit sector chimeras and recovery of mutant plants, and a method of testing in vitro response of Citrus selections to P. parasitica. The results from each of these projects were discussed in the respective chapters. Programs in genetic improvement must consider many factors, including ease of propagation, horticultural performance, yield, and consumer acceptance. The breeder's job is complicated further by strict 132

PAGE 140

133 limits of time and financial resources, making it impossible for one program to pursue more than a few potential avenues of genetic improvement during the career of a breeder. These limitations are more severe with crops having long juvenile periods, like Citrus. Consequently, difficult choices must be made about program direction(s) early in a breeder's career with only scant information and limited experience. Ideally, these decisions will be based on the costs of and potential returns from all the available options. Initial investigations, such as those described in this dissertation, may provide valuable information on which to base decisions about the relative expenses (time and money) and potential profit (in genetic advancement) of various approaches to genetic improvement and cultivar development. It may be prudent to devote a few years to evaluation of an approach under carefully controlled conditions (or with model genotypes) before proceeding with long term investments of time and money. With this perspective, a general assessment will be presented of the relative value of the several avenues for Citrus "breeding" that were described in this dissertation. The connection between the Caribbean forbidden fruit and the grapefruit has been the source of considerable controversy over nearly two centuries. Although the rediscovery of the forbidden fruit in Saint Lucia sheds considerable insight into the sources of the confusion, it will probably not put the controversy to rest. The isozyme and morphological data indicated that forbidden fruit and grapefruit are very closely related, although the former is considerably more diverse in characteristics and has a much higher degree of zygotic embryony. These same two characteristics (along with a possible greater degree of pollen

PAGE 141

134 fertility) make the forbidden fruit a promising parent for use in sexual hybridization, regardless of its actual taxonomic standing. In this respect, the forbidden fruit is not a new breeding direction, but a small (although potentially valuable) addition to the available Citrus breeding germplasm. A second aspect of the forbidden fruit potential that deserves mention is the possibility for the development of a new Citrus fruit type for fresh marketing. Most Saint Lucian forbidden fruit selections have good flavor characteristics that resemble grapefruit but are distinct in some respects as well. Many of the forbidden fruit selections appear somewhat intermediate between grapefruit and pummel o, and may present a pummel o-l ike form that could win acceptance in the American marketplace. The same attraction to the "forbidden" that resulted in biblical mankind's expulsion from the Garden of Eden may, with proper promotional titillation, provide an urge to consume the contemporary forbidden fruit that would reserve for it a place on the specialty produce counter. Generation of variation by tissue culture (somaclonal variation) has been the subject of research reports totaling in the hundreds over the past decade. Many of these reports have expounded a tremendous potential for the technique in pursuit of genetic improvement. Without question the phenomenon does occur in many plant species and in Citrus. However, the morphological evaluation of 'Carrizo' and 'Hamlin' somaclones regenerated by organogenesis indicated that stable aberrant plants are not frequently produced. Considering the relatively great investment of time required to produce somaclones by this method, and the expectation that most mutations will be deleterious, this probably will not be a cost effective method of obtaining useful genetic variation. For this reason,

PAGE 142

135 preliminary information was gathered on gamma irradiation of tissue cultures as a supplemental source of mutagenesis. A much higher frequency of genetic variants can be expected among regenerates from gamma irradiated tissue cultures. Irradiation or other mutagenic treatments are, therefore, advisable for use with tissue cultures as a source of Citrus genetic mutants. Sectored chimeric citrus fruit were easily recovered from packinghouses. Observations of potential genetic improvements in some sector mutants and successful recovery of mutations in nucellar seedlings from sectors suggest that sectored fruit may be a valuable source of genetic mutants. This resource is particularly attractive because of the ability to select desirable mutants at the start of the research involvement (the packinghouse). Preliminary contacts with packinghouse management indicated that it would be possible to have the packinghouse graders preselect any desired type of sectored fruit. The breeder could then select from these elite fruit the more promising candidates for propagation and field trials. The types of mutations would, of course, generally be restricted to those that could be observed in the fruit rind. However, these are the kinds of genetic changes that are often of most economic significance and impossible to select for at the seedling stage. Preliminary indications are that fruit sector chimeras are a very inexpensive (in both time and money) source of "selected mutations" that have considerable promise for genetic improvement of existing cultivars. One class of sector chimeras has already provided tetraploid clones that may be used in breeding.

PAGE 143

136 The response of etiolated seedling shoots to P. parasitica in vitro appeared to be a poor method of selecting genotypes resistant to foot rot or root rot in a breeding program, at least among trifoliate hybrids. The observation that the field-resistant cultivar, 'Swingle', exhibited a susceptible response under the test conditions indicated that the test results did not correlate well with field response. Preliminary evidence suggested that green seedling shoots may give a more useful indication of resistance, but the results were inconclusive. At any rate, this method appeared to be too expensive (in labor) to use with large numbers of genotypes. However, the technique may hold promise for controlled studies of resistance mechanisms. Success in breeding/genetic improvement, in contrast to "pure" science, must be measured in practical terms. Techniques producing spectacular results or novel genotypes do not equate with breeding progress. Similarly, mundane procedures or an accident of nature may result in a genetic combination that will dominate an industry for decades or centuries. The 'Washington' navel and 'Valencia' sweet oranges are examples of cultivars with tremendous commercial value that arose by chance (Hodgson, 1967). A program for genetic improvement of Citrus should employ the most efficient and promising techniques to achieve the desired goals. However, a Citrus breeder should recognize that favorable genetic combinations will be determined primarily by chance even in the most carefully designed and executed programs for genetic improvement. The ability to recognize useful genetic combinations wherever they occur and accurately select the best from the millions will probably remain the ultimate determinant of the pace of genetic advancement.

PAGE 144

APPENDIX Table A.l. Adaxial leaf color of selected control genotypes with mean comparison to HAMLIN MERICLONE by Duncan's multiple range test. Adaxial leaf color Clone N A B L HAMLIN MERICLONE 62 -13.58 17 .12 37 .00 Valencia 58 -14.81* 19 .67* 39 .94 Duncan/Z 5 -12.90 16 .08 35 .42 SF23-1 5 -10.24** 10 .77** 29 .67** SF24-1 16 -12.23* 13 .92* 33 .35* Marsh 5 -15.10** 20 .69** 42 .50** Meiwa 0 nd nd nd = mean significantly different from HAMLIN MERICLONE with a = 0.05; ** = mean significantly different with a = 0.01; N = number of observations; nd = no data. 137

PAGE 145

138 Table A. 2. Abaxial leaf color of selected control genotypes with mean comparison to HAMLIN MERICLONE by Duncan's multiple range test. Abaxial leaf color Clone N A B L HAMLIN MERICLONE 62 -14 .25 21 .34 49 .91 Valencia 57 -14 .23 22 .07 52 .27* Duncan/Z 5 -14 .27 22 .39* 50 .52 SF23-1 5 -14 .22 21 .36 48 .24 SF24-1 16 -14 .36 22 .49** 51 .05 Marsh 5 -14 .62 23 .84** 55 .21** Meiwa 0 nd nd nd = mean significantly different from HAMLIN MERICLONE with a = 0.05; ** = mean significantly different with = 0.01; N = number of observations; nd = no data.

PAGE 146

139 Table A. 3. Leaf color ratios of selected control genotypes with mean comparison to HAMLIN MERICLONE by Duncan's multiple range test. Leaf color as A/B Clone N Adaxial Abaxial HAMLIN MERICLONE 62 -0.8016 -0.6684 Valencia 57 -0.7604 -0.6455* Duncan/Z 5 -0.8056 -0.6375** SF23-1 5 -0.9522** -0.6657 SF24-1 16 -0.8810** -0.6387** Marsh 5 -0.7328* -0.6135** Meiwa 0 nd nd = mean significantly different from HAMLIN MERICLONE with a = 0.05; ** = mean significantly different with a = 0.01; N = number of observations; nd = no data.

PAGE 147

140 Table A. 4. Spine size and shape of selected control genotypes with mean comparison to HAMLIN MERICLONE by Duncan's multiple range test. Clone _N_ Spine lenqth Spine dia. Spine L/D HAMLIN MERICLONE 67 26 mm 1.2 mm 23 Valencia 60 30 1.2 25 Duncan/Z 6 18** 1.4** 13** SF23-1 8 15** 1.1 14** SF24-1 20 14** 1.1 12** Marsh 7 13** 1.0 11** Meiwa 3 14** 1.2 11** = mean significantly different from HAMLIN MERICLONE with a = 0.05; ** = mean significantly different with a = 0.01; N = number of observations; nd = no data.

PAGE 148

141 Table A. 5. Leaf size and shape of selected control genotypes with mean comparison to HAMLIN MERICLONE by Duncan's multiple range test. Clone N Leaf lenqth Leaf width Leaf L/W HAMLIN MERICLONE 67 43 mm 27 mm 1.6 Valencia 60 46 27 1.7 Duncan/Z 5 50 34** 1.5 SF23-1 5 35* 21* 1.7 SF24-1 18 37 23 1.6 Marsh 5 44 26 1.7 Meiwa 3 46 18** 2.5** = mean significantly different from HAMLIN MERICLONE with a = 0.05; ** = mean significantly different with a = 0.01; N = number of observations; nd = no data.

PAGE 149

142 Table A. 6. Leaf oil gland density and internode length of selected control genotypes with mean comparison to HAMLIN MERICLONE by Duncan's multiple range test. Clone N Oil glands per 20 mm 2 Internode lenoth HAMLIN MERICLONE 65 29 17 mm Valencia 60 32 18 Duncan/Z 5 10** 13 SF23-1 8 10** 8** SF24-1 20 16* g** Marsh 5 21 10* Meiwa 3 131** 8* = mean significantly different from HAMLIN MERICLONE with a = 0.05; ** = mean significantly different with a = 0.01; N = number of observations; nd = no data.

PAGE 150

143 Table A. 7. Adaxial leaf color of somaclones with mean comparison to HAMLIN MERICLONE by Duncan's multiple range test. Adaxial leaf color Clone N A B L HAMLIN MERICLONE 62 -13.58 17.12 37.00 HA142E121 2 -9.73** 10.95** 29.83** HA142E122 5 S s S HA142E124 5 S s S HA142E211 4 S s HA142E215 5 s s s HA142E221 1 s s s HA142E222 1 s s HA142E251 3 s s s HA142E252 3 s s s HA151E121 3 s s s HA151E122 5 s s s HA4E121 4 s s HA4I221 4 s s s HA5E121 1 s HA5I122 3 s s s HA5I123 5 s s s HA5I124 3 s HA6E112/Z 5 s s s HA6E114/Z 3 22.86** 46.26** HA6E116/Z 5 s HA7E11/Z 5 s s s HA7E12/Z 5 s s s

PAGE 151

144 Table A. 7. (continued) Adaxial leaf color Clone N A B L HA9E13/Z 3 S HX1X11 3 S s s HX1X12 3 s 29.83** HX101E111 3 S s s HX111E11 5 S s s HX115E12/Z 3 S s s HX115E14/Z 5 S HX14E12 4 S s s HX15E11 5 S s s HX20E11 5 -16.33** 23.15** 46.26** HX21E11 5 s s S HX22I11 4 S s S HX50E13 5 S s S HX60E11 5 HX61E11 3 S s s VA13E11 1 s s s VA30E11 4 s s s VA33E11 5 s s VX91E11/Z 5 22.50** 44.33** VX91E12/Z 5 s s S ANOVA GLM PR>F (df=42) 0.0001 0.0001 0.0001 S = mean not different from HAMLIN MERICLONE with = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with a = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 152

145 Table A. 8. Abaxial leaf color of somaclones with mean comparison to HAMLIN MERICLONE by Duncan's multiple range test. Abaxial leaf color Clone N A B L HAMLIN MERICLONE 62 -14.25 21.34 49.91 HA142E121 2 -13.16** 19.49** 45.39** HA142E122 5 S S S HA142E124 5 S s S HA142E211 4 S s s HA142E215 5 S s s HA142E221 1 S s s HA142E222 0 nd nd nd HA142E251 3 S S S HA142E252 3 S s s HA151E121 3 s s s HA151E122 5 s s s HA4E121 4 s s s HA4I221 4 s s HA5E121 1 s s s HA5I122 3 s s s MM J 1 1 C J 0 c 0 s s HA5I124 3 s s s HA6E112/Z 5 s s s HA6E114/Z 3 s 54.03** HA6E116/Z 5 s s HA7E11/Z 5 s s s HA7E12/Z 5 s

PAGE 153

146 Table A. 8. (continued) Abaxial leaf color Clone JL A B L HA9E13/Z 3 S 54.46** HX1X11 3 S s S HX1X12 3 S 19.39** 45.23** HX101E111 3 S S S HX111E11 5 S S HX115E12/Z 3 s s s HX115E14/Z 5 s 54.22** HX14E12 4 s s S HX15E11 5 s s S HX20E11 5 23.14** 55.29** HX21E11 5 s S S HX22I11 4 s S S HX50E13 5 s S S HX60E11 5 s s HX61E11 3 s s s VA13E11 1 s s o VA30E11 4 s s s VA33E11 5 s VX91E11/Z 5 s 54.93** VX91E12/Z 5 s ANOVA GLM PR>F (df-41) 0.0001 0.0001 0.0001 S = mean not different from HAMLIN MERICLONE with a = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with a = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 154

147 Table A. 9. Leaf color ratios of somaclones with mean comparison to HAMLIN MERICLONE by Duncan's multiple range test. Leaf color as A/B Clone JL Adaxial Abaxial HAMLIN MERICLONE 62 -0.8016 -0.6684 HA142E121 2 S S HA142E122 5 s S HA142E124 5 s S HA142E211 4 s s HA142E215 5 s s HA142E221 1 s HA142E222 1 s nd HA142E251 3 s S HA142E252 3 s S HA151E121 3 HA151E122 5 s s HA4E121 4 s s HA4I221 4 s s HA5E121 1 -0.9929** s HA5I122 3 S s HA5I123 5 c o c o HA5I124 3 s s HA6E112/Z 5 s s HA6E114/Z 3 s HA6E116/Z 5 s HA7E11/Z 5 s s HA7E12/Z 5 s -0.6166**

PAGE 155

148 Table A. 9. (continued) Leaf color as A/B Clone N Adaxial Abaxial HA9E13/Z 3 -0.6127** HX1X11 3 s S HX1X12 3 -0.9290** -0.7190** HX101E111 3 s S HX111E11 5 S HX115E12/Z 3 S S HX115E14/Z 5 -0.6241** HX14E12 4 s S HX15E11 5 s S HX20E11 5 S HX21E11 5 s s HX22I11 4 s s HX50E13 5 s s HX60E11 5 s s HX61E11 3 s s VA13E11 1 s s VA30E11 4 s VA33E11 5 s -0.6280** VX91E11/Z 5 VX91E12/Z 5 s ANOVA GLM PR>F (df=42) 0.0001 0.0001 S = mean not different from HAMLIN MERICLONE with a = 0.05; = mean significantly different with o = 0.05; ** = mean significantly different with = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 156

149 Table A. 10. Spine size and shape of somaclones with mean comparison to HAMLIN MERICLONE by Duncan's multiple range test. Spine Spine Spine Clone N lenqth dia. L/D HAMLIN MERICLONE 67 26 mm 1 .2 mm 23 HA142E121 4 S S HA142E122 6 S S s HA142E124 5 S s s HA142E211 5 s S s HA142E215 5 s S 15** HA142E221 2 s S S HA142E222 2 s S HA142E251 5 s S s HA142E252 5 s S s HA151E121 5 s s s HA151E122 5 s s s HA4E121 5 s s s HA4I221 5 s 0.9** s HA5E121 3 0.9** 15** HA5I122 5 10** 0.8** 16** HA5I123 0 c c o c o C 0 HA5I124 4 s s s HA6E112/Z 5 s s s HA6E114/Z 5 s s s HA6E116/Z 5 s s s HA7E11/Z 5 s s s HA7E12/Z 5 s s s

PAGE 157

150 Table A. 10 (continued) Clone N Spine lenqth Spine dia. Spine L/D HA9E13/Z 5 S S S HX1X11 4 0.9** HX1X12 3 0.8** 16** HX101E111 5 S S S HX111E11 5 S s S HX115E12/Z 5 s s s HX115E14/Z 5 s s s HX14E12 5 s s s HX15E11 5 s s s HX20E11 5 s s s HX21E11 5 s s s HX22I11 5 s s s HX50E13 5 s s s HX60E11 5 s s s HX61E11 5 s s s VA13E11 3 s s s VA30E11 5 s s VA33E11 5 s s s VX91E11/Z 5 VX91E12/Z 5 s s s ANOVA GLM PR>F (df=42) 0.0001 0.0001 0.0001 S = mean not different from HAMLIN MERICLONE with a = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with a = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 158

Table A. 11. Leaf size and shape of somaclones with comparison to HAMLIN MERICLONE by Duncan's multiple range test. Clone N Leaf lenqth Leaf width Leaf L/W HAMLIN MERICLONE 67 43 mm 27 mm 1.6 HA142E121 4 S S HA142E122 5 S S s HA142E124 5 s S s HA142E211 5 s s s HA142E215 5 s s s HA142E221 2 s HA142E222 2 s 19** s HA142E251 5 s S s HA142E252 r5 s S s ll*i r i r* 1 oi HA151E121 5 s S s HA151E122 5 s s s 1 1 A A r 1 11 HA4E121 5 s s s U A A TOO! HA4I221 5 s 18** s imrri ai HA5E121 3 s s HA5I122 5 s s 1.9** HA5I123 5 s s s HA5I124 3 s s s HA6E112/Z 5 s s s HA6E114/Z 5 s s s HA6E116/Z 5 s s s HA7E11/Z 5 s s s HA7E12/Z 5 s s s

PAGE 159

152 Table A. 11 (continued) Leaf Leaf Leaf Clone N lenath width L/W HA9E13/Z 5 S S S HX1X11 3 S S S HX1X12 3 S S s HX101E111 5 19** S HX111E11 5 S S S HX115E12/Z 5 s S s HX115E14/Z 5 s s s HX14E12 5 s s s HX15E11 5 s s s HX20E11 5 s s s HX21E11 5 s s s HX22I11 5 s s s HX50E13 5 s s s HX60E11 5 s s s HX61E11 5 s s VA13E11 3 s s VA30E11 5 s s 1.9** VA33E11 5 s s 1.9** VX91E11/Z 5 s s s VX91E12/Z 5 s s s ANOVA GLM PR>F (df=42) 0.0001 0.0001 0.0001 S = mean not different from HAMLIN MERICLONE with „ = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with <* = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 160

153 Table A. 12. Leaf oil gland density and internode length of somaclones with mean comparison to HAMLIN MERICLONE by Duncan's multiple range test. Oil glands Internode Clone N per 20 mm 2 lenqth HAMLIN MERICLONE 65 29 17 mm HA142E121 3 S S HA142E122 5 s s HA142E124 4 s s HA142E211 5 s s HA142E215 5 s s HA142E221 1 s s HA142E222 1 s s HA142E251 5 s s HA142E252 5 s s HA151E121 4 s s HA151E122 4 s s HA4E121 5 s s HA4I221 3 s s HA5E121 2 s s HA5I122 5 s s HA5I123 1 s s HA5I124 4 s s HA6E112/Z 5 s s HA6E114/Z 5 s s HA6E116/Z 5 s s HA7E11/Z 4 s s HA7E12/Z 5 s s

PAGE 161

154 Table A. 12. (continued) Oil glands Internode Clone N per 20 mm 2 length HA9E13/Z 5 S S HX1X11 3 S S HX1X12 2 S HX101E111 5 S S HX111E11 5 S S HX115E12/Z 5 S S HX115E14/Z 5 S S HX14E12 5 S S HX15E11 5 S S HX20E11 5 S S HX21E11 5 S S HX22I11 5 S S HX50E13 5 S S HX60E11 5 S S HX61E11 5 S S VA13E11 3 S S VA30E11 5 S S VA33E11 4 S S VX91E11/Z 5 S S VX91E12/Z 5 S S ANOVA GLM PR>F (df=42) 0.0002 0.0030 S = mean not different from HAMLIN MERICLONE with a = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with a = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 162

155 Table A. 13. Adaxial leaf color of selected somaclones during second test with mean comparison to MERICLONE by Duncan's multiple range test. Adaxial leaf color Clone N A B L MERICLONE 57 -10.36 11.61 30.69 HA142E215 5 S S S HA4E121 5 S S S HA7E11/Z 5 s S S HA7E12/Z 5 -14.08** 18.90** 39.36** HA9E13/Z 5 S S S HX1X11 5 S S S HX1X12 3 S S S HX101E111 5 S S s HX111E11 5 S s s HX115E12/Z 5 -13.42** 17.93** 38.18** HX115E14/Z 5 S S S HX15E11 3 -6.72** HX20E11 5 S s S HX22I11 5 S s s HX61E11 5 s s ANOVA GLM PR>F (df=15) 0.0001 0.0001 0.0001 Number of 15 somaclones Differing at .05 Differing at .01 3 3 3 2 4 2 S = mean not different from HAMLIN MERICLONE with a = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 163

156 Table A.M. Abaxial leaf color of selected somaclones during second test with mean comparison to MERICLONE by Duncan's multiple range test. Abaxial leaf color Clone N A B L MERICLONE 57 -12.89 19.64 47.54 HA142E215 5 S S S HA4E121 5 S S S HA7E11/Z 5 S s HA7E12/Z 5 S 22.14** 53.90** HA9E13/Z 5 s S S HX1X11 5 s S s HX1X12 3 s s s HX101E111 5 s s s HX111E11 5 s c c o HX115E12/Z 5 s 22.33** 54.36** HX115E14/Z 5 s S S HX15E11 3 s s s HX20E11 5 s s s HX22I11 5 s s s HX61E11 5 s ANOVA GLM PR>F (df=15) 0.0258 0.0001 0.0001 Number of 15 somaclones Differing at .05 Differing at .01 1 0 3 2 3 2 S = mean not different from HAMLIN MERICLONE with = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with a = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 164

157 Table A. 15. Leaf color ratios of HAMLIN MERICLONE and selected somaclones during second test with mean comparison to MERICLONE by Duncan's multiple range test. Leaf color as A/B ratio Clone N Adaxial Abaxial MERICLONE 57 -0.9099 -0.6577 HA142E215 5 S S HA4E121 5 S S HA7E11/Z 5 S S HA7E12/Z 5 -0.7464** -0.5821** HA9E13/Z 5 S S HX1X11 5 S s HX1X12 3 S s HX101E111 5 s s HX111E11 5 •J Q o c o HX115E12/Z 5 -0.7500** -0.5792** HX115E14/Z 5 S S HX15E11 3 s s HX20E11 5 s s HX22I11 5 s s HX61E11 5 s s ANOVA GLM PR>F (df-15) 0.0001 0.0001 Number of 15 somaclones Differing at .05 Differing at .01 2 2 2 2 S = mean not different from HAMLIN MERICLONE with a = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with a = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 165

158 Table A. 16. Spine size and shape of selected somaclones during second test with mean comparison to MERICLONE by Duncan's multiple range test. Spine size and shape Clone N Lenqth Diameter L/D ratio MERICLONE 56 29 mm 1.1 mm 26 HA142E215 5 15** 0.8** 19** HA4E121 5 S S HA7E11/Z 5 s HA7E12/Z 5 s HA9E13/Z 5 s S s HX1X11 4 S s s HX1X12 3 s 18** HX101E111 5 s 1.4** S HX1 1 1 F1 1 c j c o c o c o HX115E12/Z 5 s s s HX115E14/Z 5 s s HX15E11 3 HX20E11 5 s s s HX22I11 5 s s s HX61E11 5 s s s ANOVA GLM PR>F (df=15) 0.0001 0.0001 0.0001 Number of 15 somaclones Differing at .05 Differing at .01 5 1 6 2 4 2 S = mean not different from HAMLIN MERICLONE with a = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with a = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 166

159 Table A. 17. Leaf size and shape of selected somaclones during second test with mean comparison to MERICLONE by Duncan's multiple range test. Leaf size and shape Clone N Lenqth Width L/W ratio MERICLONE 58 63 mm 39 mm 1.6 HA142E215 5 s HA4E121 5 S S S HA7E11/Z 5 S s s HA7E12/Z 5 s s s HA9E13/Z 5 S s s HX1X11 5 S s s HX1X12 3 S s s HX101E111 5 S s s HY1 1 1 Fl 1 nAl 1 1 t 1 1 c D c c 0 s HX115E12/Z 5 s s s HX115E14/Z 5 s s s HX15E11 3 s o c o c o HX20E11 5 s s s HX22I11 5 s s s HX61E11 5 s s s ANOVA GLM PR>F (df=15) 0.2578 0.0987 0.0081 Number of 15 somaclones Differing at .05 Differing at .01 0 0 1 0 1 0 S = mean not different from HAMLIN MERICLONE with a = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with a = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 167

160 Table A. 18. Leaflet lengths of Carrizo mericlones, adventitious seedlings, and somaclones with mean comparison to Z2 MERICLONE by Duncan's multiple range test. Leaflet length Clone N Left Main Riqht Z2 MERICLONE 12 22 mm 36 mm 22 mm Zl mericlone 6 S S S Z3 mericlone 6 28** S 28** Zll seedling 6 S S S Z12 seedling 6 s s S Z13 seedling 6 s s S Z14 seedling 6 S s s Z15 seedling 6 S s s Z17 seedling 6 s s s Z19 seedling 6 S s s Z2E11 6 S s s Z2E1CB111 6 S s s Z2E1CB112 6 s s s Z2E1CB113 6 s s s Z2E1CB114 6 s s s Z2E1CB231 6 s s s Z2E1CB232 *— k 1— A v *J w W k 6 c o c o Z2E1CB315 6 s s s Z2E1CB316 6 s s s Z2E1CB317 6 16** 16** Z2E21 6 s s s Z2E22 6 s s s Z2E211 6 s s s

PAGE 168

161 Table A. 18. (continued) Leaflet lenath Clone N Left Main Riant Z2E212 6 S S S Z2E233 6 s S s Z2E234 6 s S s Z2E235 6 s s s Z2E236 6 s s s Z2E237 6 s s Z2E238 6 s s s Z2E239 6 s s s Z2E240 6 s s s Z2E241 6 s s s Z2E242 6 s s s Z2E243 6 s s s Z2E244 6 s s s Z2I11 6 s s s Z2I152 6 s s s Z2I153 6 s s s Z2I154 6 s s s Z2I155 6 s s s Z2I156 6 s s s Z2I157 6 s s s Z2I158 6 s s Z2I160 6 s s s Z2I161 6 s s

PAGE 169

162 Table A. 18, (continued) Clone N Z2I162 6 Z2I163 6 Z2I164 6 Z2I165 6 Z2I166 6 Z2I167 6 Z2I168 6 o Z2I269 6 Z2I270 6 Z2I271 6 ANOVA GLM PR>F (df=56) Number of somaclones Differing at .05 Differing at .01 Leaflet lenqth Left Main Riqht S S S S S S S s S S s S S s S s c o <: o c •j C o s s s s s s s s s s s s 0001 0.0001 0. 0001 3 1 2 1 0 1 S = mean not different from Z2 MERICL0NE with a = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with a = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 170

Table A. 19. Spine size and adventitious seedlings, and to Z2 MERICLONE by Duncan's Clone N Z2 MERICLONE 6 Zl mericlone 6 Z3 mericlone 6 Z14 seedling 6 Z15 seedling 6 Z17 seedling 6 Z19 seedling 6 Z2E11 6 Z2E1CB111 6 Z2E1CB112 6 Z2E1CB113 6 Z2E1CB114 6 Z2E1CB231 6 LLLIUDCJL c 0 Z2E1CB315 6 Z2E1CB316 6 Z2E21 6 Z2E22 6 Z2E211 6 Z2E212 6 163 shape of 'Carrizo' mericlones, somaclones with mean comparison multiple range test. Spine Spine Spine length diameter L/D 22 mm 1.2 mm 18 S S S 1.5** S S 1.4** S S S S S S S S S S S S S S S S S S S S S S s s s s s s s s s s s s s s s s s s s s s s

PAGE 171

164 Table A. 19. (continued) Clone N Z2E233 6 Z2E234 6 Z2E235 6 Z2E236 6 Z2E237 6 Z2E238 6 Z2E239 6 Z2E240 6 Z2E241 6 Z2E242 6 Z2E243 6 Z2I152 6 Z2I153 4 Z2I154 6 Z2I155 6 Z2I157 6 Z2I158 6 Z2I160 6 Z2I161 6 Z2I162 6 Spine lenoth Spine diameter Spine L/D S S S S S S S 1.4** S S 1.4** S S s S S s 11** 1.0** 11** S 1.4** s S S s S S s s s s s s s 1.4** s s S s s S s 1.0** s S s s S s s S s s s s

PAGE 172

165 Table A. 19. (continued) Clone N Z2I163 6 Z2I164 6 Z2I165 6 Z2I166 6 Z2I167 1 Z2I168 6 Z2I21 6 Z2I271 6 ANOVA GLM PR>F (df=47) Number of somaclones Differing at .05 Differing at .01 Spine Spine Spine length diameter L/D S S S S S S S S S S 1.3** S 10** 0.8** 12** S S S S S S S S 0.0001 0.0001 0.0012 3 19 3 2 8 2 S = mean not different from Z2 MERICLONE with a = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with o = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 173

Table A. 20. Leaflet ratio and chirality reversals of 'Carrizo' mericlones, adventitious seedlings, and somaclones with mean comparison to Z2 MERICLONE by Duncan's multiple range test. Clone N Leaflet ratio Chiral ity reversal s 12 MERICLONE 12 1.2 0 Zl meri clone 6 S 0 Z3 meri clone 6 S 0 Zll seedling 6 S 3 Z12 seedling 6 S 0 Z13 seedling 6 S 1 Z14 seedling 6 s 2 Z15 seedling 6 S 0 Z17 seedling 6 S 0 Z19 seedling 6 S 2 Z2E11 6 S 1 Z2E1CB111 6 S 0 Z2E1CB112 6 s 3 Z2E1CB113 6 s 3 Z2E1CB114 fi c •J n Z2E1CB231 6 s 4* Z2E1CB232 6 s 2 Z2E1CB315 6 s 1 Z2E1CB316 6 s 3 Z2E1CB317 6 0 Z2E21 6 s 0

PAGE 174

167 Table A. 20. (continued) Clone N Leaflet ratio Chiral ity reversal s Z2E22 6 S 0 Z2E211 6 S 2 Z2E212 6 S 0 Z2E233 6 S 0 Z2E234 6 S 0 Z2E235 6 S 2 Z2E236 6 s 1 Z2E237 6 s 0 Z2E238 6 s 0 Z2E239 6 0 Z2E240 6 s 4* Z2E241 6 s 0 Z2E242 6 s 0 Z2E243 6 s 3 Z2E244 <; D c o u Z2I11 6 s 0 Z2I152 6 s 0 Z2I153 6 s 0 Z2I154 6 s 0 Z2I155 6 s 0

PAGE 175

168 Table A. 20. (continued) Clone Z2I156 Z2I157 Z2I158 Z2I160 Z2I161 Z2I162 Z2I163 Z2I164 Z2I165 Z2I166 Z2I167 Z2I168 Z2I21 Z2I269 Z2I270 Z2I271 _N_ 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 5 Leaflet ratio S S S S S S S S S S S S S S S S Chiral ity reversal s 0 0 1 0 2 0 1 1 0 0 0 0 0 1 0 2 ANOVA GLM PR>F (df=56) 0.0006 Number of somaclones Differing at .05 2 Differing at .01 0 2 0 S = mean not different from Z2 MERICLONE with a = 0.05; = mean significantly different with a = 0.05; ** = mean significantly different with a = 0.01; nd = no data; N = number of observations; df = degrees of freedom.

PAGE 176

LITERATURE CITED Ahloowalia, B.S. 1978. Novel ryegrass genotypes regenerated from embryo-callus culture, p. 162. In: Abstracts, 4th International Ass. Plant Tiss. Cult. Meeting, Calgary, Alberta. Ahloowalia, B.S. 1983. Spectrum of variation in somaclones of triploid rye-grass. Crop Sci 23:1141-1147. Afek, U. and A. Sztejnberg. 1986. A citrus phytoalexin, 6,7-dimethoxycoumarin, as a defense mechanism against Phytophthora citrophthora, and the influence of fosetyl-AL and phosphorous acid on its production. Phytoparasitica 14:246. Afek, U. and A. Sztejnberg. 1988. Accumulation of scoporone, a phytoalexin associated with resistance of citrus to Phytophthora citrophthora. Phytopathology 78:1678-1682. Afek, U., A. Sztejnberg, and S. Carmely, 1986. 6,7-Dimethoxycoumarin, a citrus phytoalexin conferring resistance against Phytophthora gummosis. Phytochemistry 25:1855-1856. Alexander, L.J., G.L. Oakes, and C.A. Jaberg. 1971. The production of two needed mutations in tomato by irradiation. J. Heredity 62:311-315. Allard, R.W. 1960. Principles of Plant Breeding. John Wiley and Sons, New York. Allicchio, R., C. Antonioli, L. Graziani, R. Roncarati, and C. Vannini. 1987. Isozyme variation in leaf-callus regenerated plants of Solanum tuberosum. Plant Science 53:81-86. Almansa, M.S., L.A. del Rio, C.F. Alcaraz, and F. Sevilla. 1989. Isoenzyme pattern of superoxide dismutase in different varieties of citrus plants. Physiologia Plantarum 76:563-568. 169

PAGE 177

170 Amasino, R.M., A.L.T. Powell, and M.P. Gordon. 1984. Changes in T-DNA methylation and expression are associated with phenotypic variation and plant regeneration in a crown gall tumor line. Molec. Gen. Genet. 197:437-446. Anderson, S., A.C. Lewis-Smith, and S.M. Smith. 1990. Methylation of ribosomal RNA genes in Petunia hybrida plants, callus cultures and regenerated shoots. Plant Cell Reports 8:554-557. Anonymous. 1921. Romantic story of Dr. Phillipe, Safety Harbor's first settler. Tampa Sunday Tribune. May 1:3A, 6A. Anonymous. 1988. The Rio Red raises red flags. Citrograph 73:77-78. Anonymous. 1989a. Super red debuts in California. Citrograph 74:115-116. Anonymous. 1989b. Color the comeback red. Citrograph 74:244-245. Appels, R. and J. Dvorak. 1982. The wheat ribosomal DNA spacer region: Its structure and variation in populations and among species. Theor. Appl Genet. 63:337-348. Armstrong, C.L. and R.L. Phillips. 1988. Genetic and cytogenetic variation in plants regenerated from organogenic and friable embryogenic tissue cultures of maize. Crop Sci. 28:363-369. Armstrong, K.C., C. Nakamura, and W.A. Keller. 1983. Karyotype instability in tissue culture regenerants of triticale (x Triticosecale Wittmack) cv. 'Welsh' from 6-month-old callus cultures. Z. Pflanzenzuecht. 91:233-245. Atanasov, A. and D.C.W. Brown. 1984. Plant regeneration from suspension culture and mesophyll protoplasts of Nedicago sativa L. Plant Cell, Tissue and Organ Culture 3:149-162. Bacchi, 0. 1943. Cytological observations in Citrus: III. Megasporogenesis, fertilization, and polyembryony. Bot. Gaz. 105:221-225. Bade, J.B., F.G. Gmitter Jr., and K.D. Bowman. In press. Simplified method of volatile leaf oil analysis for identification of Citrus cultivars. HortScience.

PAGE 178

171 Bajaj, Y.P.S., A.W. Saettler, and M.W. Adams. 1970. Gamma irradiation studies on seeds, seedlings, and callus tissue cultures of Phaseolus vulgaris L. Radiation Bot. 10:119-124. Balito, L.P., K. Hattori, and Y. Futsuhara. 1989. Effects of gamma-ray irradiation on the growth of calli in Nicotiana species. Japan. J. Breed. 39:29-37. Baranowska, H., D. Zaborowska, and J. Zuk. 1987. Decreased UV mutagenesis in an excision-deficient mutant of yeast. Mutagenesis 2:1-6. Barbier, M. and H. Dulieu. 1983. Early occurrence of genetic variants in protoplast cultures. Plant Sci Letters 29:201-206. Barrett, H.C. 1974. Colchicine-induced polyploidy in Citrus. Bot. Gaz. 135:29-41. Barrett, H.C. and A.M. Rhodes. 1976. A numerical taxonomic study of affinity relationships in cultivated Citrus and its close relatives. Syst. Bot. 1:105-136. Barritt, B.H., R.P. Covey, and M.A. Dilley. 1990. In vitro testing of the reaction of apple rootstocks to Phytophthora cactorum. Fruit Var. J. 44:23-25. Bayliss, M.W. 1973. Origin of chromosome number variation in cultured plant cells. Nature 246:529-530. Bayliss, M.W. 1980. Chromosomal variation in plant tissues in cultures. Int. Rev. Cytol Suppl 11A: 1 13-144. Behnke, M. 1979. Selection of potato callus for resistance to culture filtrates of Phytophthora infestans and regeneration of resistant plants. Theor. Appl Genet. 55:69-71. Behnke, M. 1980. General resistance to late blight of Solanum tuberosum plants regenerated from callus resistant to culture filtrates of Phytophthora infestans. Theor. Appl. Genet. 56:151-152. Benfey, P.N. and N.H. Chua. 1989. Regulated genes in transgenic plants. Science 244:174-181. Bennici, A. and L. Caffaro. 1985. Karyological behavior during the first phases of dedifferentiation and habituation in Nicotiana bigelovii. Protoplasma 124:130-136. Benzion, G. and R.L. Phillips. 1988. Cytogenetic stability of maize tissue cultures: A cell line pedigree analysis. Genome 30:318-325.

PAGE 179

172 Benzion, G., R.L. Phillips, and H.W. Rines. 1986. Case histories of genetic variability in vitro: Oats and maize, pp. 435-448. In: I.K. Vasil (ed.), Cell Culture and Somatic Cell Genetics of Plants, Vol. 3. Academic, New York. Berlyn, G.P., R.C. Beck, and M.H. Renfroe. 1986. Tissue culture and the propagation and genetic improvement of conifers: problems and possibilities. Tree Physiology 1:227-240. Bhatia, C.R., G.S.S. Murty, C. Mouli, and D.M. Kale. 1986. Regeneration of Ml plants from 'de-embryonated' cotyledons to modify diplontic selection, pp. 419-427. In: Nuclear Techniques and In-Vitro Culture for Plant Improvement, Proceedings of a Symposium, Vienna 1985. International Atomic Energy Agency, Vienna. STI/PUB/698. Binarova, P. and J. Dolezel. 1988. Alfalfa embryogenic cell suspension culture: Growth and ploidy level stability. J. Plant Physiol. 133:561-566. Binns, A. and F. Meins. 1973. Habituation of tobacco pith cells for factors promoting cell division is heritable and potentially reversible. Proc. Nat. Acad. Sci. USA 70:2660-2662. Bird, A. P. 1986. CpG-rich islands and the function of DNA methylation. Nature 321:209-213. Bonavia, E. 1888. The Cultivated Oranges and Lemons Etc. of India and Ceylon With Researches Into Their Origin and The Derivation of Their Names, and Other Useful Information. 2 Vols. W.H. Allen & Co. London. Bono, R., L. Fernandez de Cordova, and J. Soler. 1981. 'Arrufatina, 'Esbal' and 'Guillermina, three Clementine mandarin mutations recently discovered in Spain. Proc. Int. Soc. Citriculture, Tokyo 1:94-96. Borst, P. and D.R. Greaves. 1987. Programmed gene rearrangements altering gene expression. Science 235:658-667. Bowman, K.D. and F.G. Gmitter, Jr. 1990a. Caribbean forbidden fruit: Grapefruit's missing link with the past and bridge to the future? Fruit Var. J. 44:41-44. Bowman, K.D. and F.G. Gmitter, Jr. 1990b. Forbidden Fruit (Citrus sp., Rutaceae) Rediscovered in Saint Lucia. Econ. Bot. 44:165-173. Bowman, K.D. and F.G. Gmitter, Jr. 1990c. Isozyme characterization of Caribbean forbidden fruit (Citrus sp.) clones and seedlings. Abstract No. 633, ASHS Annual Meeting, Tucson.

PAGE 180

173 Brears, T., G.J. Curtis, and D.M. Lonsdale. 1989. A specific rearrangement of mitochondrial DNA induced by tissue culture. Theor. Appl Genet. 77:620-624. Brears, T., C.L. Schardl, and D.M. Lonsdale. 1986. Chloroplast genome organization in sugar beet and maize. Plant Mol. Biol. 6:171-178. Breiman, A., T. Felsenburg, and G. Galun. 1987a. Nor loci analysis in progenies of plants regenerated from scutellar callus of bread wheat. A molecular approach to evaluate somaclonal variation. Theor. Appl. Genet. 73:827-831. Breiman, A., D. Rotem-Abarbanell A. Karp, and H. Shaskin. 1987b. Heritable somaclonal variation in wild barley (Hordeum spontaneum) Theor. Appl. Genet. 74:104-112. Brettell, R.I.S., E.S. Dennis, W.R. Scowcroft, and W.J. Peacock. 1986a. Molecular analysis of a somaclonal mutant of maize alcohol dehydrogenase. Mol. Gen. Genet. 202:235-239. Brettell, R.I.S., M.A. Pallotta, J. P. Gustafson, and R. Appels. 1986b. Variation at the Nor loci in triticale derived from tissue culture. Theor. Appl. Genet. 71:637-643. Brettell, R.I.S., R. Thomas, and D.S. Ingram. 1980. Reversion of Texas male-sterile cytoplasm maize in culture to give fertile, T-toxin resistant plants. Theor. Appl. Genet. 58:55-58. Brickell, CD., E.G. Voss, A.F. Kelly, and F. Schneider (eds.). 1980. International Code of Nomenclature for Cultivated Plants-1980. Bohn, Scheltema & Holkema, Utrecht, Netherlands. Bright, S.W.J, and M.G.K. Jones. 1985. Cereal Tissue and Cell Culture. Ni jhoff/Junk, Dordrecht. Brisset, M.N., J. P. Paulin, and M. Duron. 1988. Feasibility of rating fire blight susceptibility of pear cultivars (Pyrus communis) on in vitro microcuttings. Agronomie 8:707-710. Britton, N.L. 1918. Flora of Bermuda. Charles Scribner's Sons, New York. Broadbent, P., L.R. Fraser and Y. Waterworth. 1971. The reaction of seedlings of Citrus spp. and related genera to Phytophthora citrophthora. Proc. Linnean Soc. New South Wales 96:119-127. Brock, R.D. 1971. The role of induced mutations in plant improvement. Radiation Bot. 11:181-196.

PAGE 181

174 Broertjes, C. 1982. Significance of in vitro adventitious bud techniques for mutation breeding of vegetatively propagated crops, pp. 1-10. In: Induced Mutations in Vegetatively Propagated Plants II. IAEA, Vienna. Broertjes, C, B. Haccius, and S. Weidlich. 1968. Adventitious bud formation on isolated leaves and its significance for mutation breeding. Euphytica 17:321-344. Broertjes, C. and A. Keen. 1980. Adventitious shoots: Do they develop from one cell? Euphytica 29:73-88. Brossard, D. 1978. Microspectrophotometric and ultrastructural analyses of cell differentiation without endopolyploidization: The pith of Crepis capillaris. Protoplasma 93:369-380. Brown, P.T.H. and H. Lorz. 1986. Molecular changes and possible origins of somaclonal variation, pp. 148-159. In: J. Semal (ed.), Somaclonal Variations and Crop Improvement. Martinus Nijhoff, Boston. Browne, P. 1756. The Civil and Natural History of Jamaica. Osborne and Shipton, London. Browne, P. 1789. The Civil and Natural History of Jamaica. 2nd Edition. White and Son, London. Brumfield, R.T. 1943. Cell-lineage studies in root meristems by means of chromosome rearrangements induced by x-rays. Amer. J. Bot. 30:101-110. Buiatti, M. and A. Scala. 1985. Host-parasite interaction in vitro in tomato and carnation, pp. 85-90. In: A. Schafer-Menuhr (ed.), In Vitro Techniques, Propagation and Long Term Storage. Martinus Nijhoff /Dr. W. Junk Publ Boston. Burk, L.G and D.F. Matzinger. 1976. Variation among anther-derived doubled haploids from an inbred line of tobacco. J. Heredity 67:381-384. Cameron, J.W. and H.B. Frost. 1968. Genetics, breeding and nucellar embryony. pp. 325-370. In: W. Reuther, L.D. Batchelor, and H.J. Webber (eds.), The Citrus Industry, Vol. 2. Revised Edition. Univ. of California Press, Berkeley. Cameron, J.W., L.J. Klotz, T.A. DeWolfe, and R.K. Soost. 1972. Estimates of the resistance of Citrus X Poncirus hybrids to feeder root infection by Phytophthora spp., by a greenhouse seedling test. Plant Dis. Reporter 56:927-931. Cameron, J.W., R.K. Soost, and E.O. Olson. 1964. Chimeral basis for color in pink and red grapefruit. J. Heredity 55:23-28.

PAGE 182

175 Campbell, W.H. and G. Gowri 1990. Codon usage in higher plants, green algae, and cyanobacteria. Plant Physiol. 92:1-11. Cappadocia, M. and K.S. Ramulu. 1980. Plant regeneration from in vitro cultures of anthers and stem internodes in an interspecific hybrid, Lycopersicum esculentum x L. peruvianum and cytogenetic analysis of the regenerated plants. Plant Sci. Letters 20:157. Cardy, B.J., C.W. Stuber, and M.M. Goodman. 1981. Techniques for starch gel electrophoresis of enzymes from maize (Zea mays L.). pp. 1-31. Institute of Statistics Mimeograph Series No. 1317. North Carolina State University, Raleigh. Carpenter, J.B. and J.R. Furr. 1962. Evaluation of tolerance to root rot caused by Phytophthora parasitica in seedlings of Citrus and related genera. Phytopathology 52:1277-1285. Carson, H.L. 1970. Chromosome tracers of the origin of species. Science 168:1414-1418. Carson, H.L. 1971. Speciation and the founder principle. Stadler Genetics Symposia 3:51-70. Casteels, P., A. Christophe, F. Jacobs, M. Vaeck, and P. Tempst. 1989. Apidaecins: antibacterial peptides from honeybees. EMBO J. 8:2387-2391. Castle, W.S. 1987. Citrus rootstocks. pp. 361-399. In: R.C. Rom and R.F. Carlson (eds.), Rootstocks For Fruit Crops. John Wiley & Sons, New York. Castle, W.S., D.P.H. Tucker, A.H. Krezdorn, and CO. Youtsey. 1989. Rootstocks for Florida Citrus. IFAS Bull. SP42. Univ. of Florida, Gainesville. Cedar, H. 1988. DNA methylation and gene activity. Cell 53:3-4. Chaleff, R. 1981. Genetics of Higher Plants. Applications of Cell Culture. Cambridge Univ. Press, Cambridge. Chaleff, R.S. 1983. Isolation of agronomically useful mutants from plant cell cultures. Science 219:676-682. Chandler, V.L. and V. Walbot. 1986. DNA modification of a maize transposable element correlates with loss of activity. Proc. Natl. Acad. Sci. USA 83:1767-1771. Chang, L.S., A.F. Iezzoni, and G. Adams. 1989. Excised-shoot assay for tolerance of peach to Leucostoma persoonii. HortScience 24:1011-1012

PAGE 183

176 Chang, W.C., G.B. Liui, S.S. Deng, and X.C. Li. 1984. The role of callus tissues in citrus breeding works. Proc. Int. Soc. Citriculture, Sao Paulo, Brazil 1:25-27. Chapot, H. 1950. Pamplemousses, Pomelos ou Grape-fruits et Tangelos. Rev. Int. de Botan. Appl 327-328:62-75. Charlton, D.E., H. Nikjoo, and J.L. Humm. 1989. Calculation of initial yields of singleand double-strand breaks in cell nuclei from electrons, protons and alpha particles. Int. J. Radiat. Biol. 56:1-19. Clement, C.R. 1989. A center of crop genetic diversity in western Amazonia. Bioscience 39:624-631. Chepurnoi, A. I., N. Mikhova-Tsenova, and K. Bruntskova. 1989. Mechanisms of formation of gamma-induced mutants of Saccharomyces cerevisiae. Genetika 25:1179-1188. Chevreau, E., L. Decourtye, and R.M. Skirvin. 1989. A review of pear chimeras: Their identification and separation into pure types. HortScience 24:32-34. Cionini, P.G., A. Bennici, and F. D'Amato. 1978. Nuclear cytology of callus induction and development in vitro. I. Callus from Vicia faba cotyledons. Protoplasma 96:101-112. Coggle, J.E. 1983. Biological Effects of Radiation. 2nd Edition. Taylor & Francis Inc., New York. Coit, J.E. 1915. Citrus Fruits. Macmillan, New York. Comai, L., D. Facciotti, W.R. Hiatt, G. Thompson, R.E. Rose, and D.M. Stalker. 1985. Expression in plants of a mutant aroA gene from Salmonella typhimurium confers tolerance to glyphosate. Nature 317:741-744. Conger, B.V. 1981. Cloning Agricultural Plants Via In Vitro Techniques. CRC Press, Boca Raton, FL. Cooper, William C. 1982. In Search of the Golden Apple. Vantage Press, New York. Cramer, P.J.S. 1954. Chimeras. Bibl. Genet. 16:193-381. Croker, B.H. 1953. Effects of 2,4-dichlorophenoxyacetic acid and 2,4,5-trichlorophenoxyacetic acid on mitoses in Allium cepa. Bot. Gaz. 114:274-283. Crow, J.F. and M. Kimura. 1965. Evolution in sexual and asexual populations. Amer. Nat. 99:439-450.

PAGE 184

177 Cummings, D.P., C.E. Green, D.D. Stuthman. 1976. Callus induction and plant regeneration in oats. Crop Sci. 16:465-470. Cutler, A.J., M. Saleem, E. Kendall, L.V. Gusta, F. Georges, and G.L. Fletcher. 1989. Winter flounder antifreeze protein improves the cold hardiness of plant tissues. J. Plant Physiol. 135:351-354. D'Amato, F. 1952. Polyploidy in the differentiation and function of tissues and cells in plants. A critical examination of the literature. Caryologia 4:311-357. D'Amato, F. 1965. Chimera formation in mutagen-treated seeds and diplontic selection, pp. 302-316. In: The Use of Induced Mutations in Plant Breeding, Report of the FAO/IAEA Technical Meeting, Rome 1964. Pergamon Press, Oxford. D'Amato, F. 1975. The problem of genetic stability in plant tissue and cell cultures, pp. 333-348. In: O.H. Frankel and J.G. Hawkes (eds.), Crop Genetic Resources for Today and Tomorrow. Cambridge Univ. Press, New York. D'Amato, F. 1977a. Cytogenetics of differentiation in tissue and cell culture, pp. 343-357. In: J. Reinert and Y.P.S. Bajaj (eds.), Applied and Fundamental Aspects of Plant Cell, Tissue and Organ Culture. Springer-Verlag, Berlin. D'Amato, F. 1977b. Nuclear Cytology in Relation to Development. Cambridge University Press, New York. D'Amato, F. 1978. Chromosome number variation in cultured cells and regenerated plants, pp. 287-295. In: T.A. Thorpe (ed.), Frontiers of Plant Tissue Culture. Univ. Calgary, Calgary, Alberta. D'Amato, F. 1985. Cytogenetics of plant cell and tissue cultures and their regenerates. CRC Critical Reviews in Plant Sciences 3:73-112. Darwin, C. 1921. Variation of Animals and Plants Under Domestication. 2nd Edition. 10th imp. 2 Vols. John Murray, London. Daub, M.E. and A.E. Jenns. 1989. Field and greenhouse analysis of variation for disease resistance in tobacco somaclones. Phytopathology 79:600-605. Davidson, T.M.W. and D.N. Lawley. 1953. Experimental evidence of clonal variation affecting yield in potatoes. Empire J. Expt. Agric. 21:137-140.

PAGE 185

178 Davis, R.M. 1981. Phytophthora foot rot control with the systemic fungicides metalaxyl and fosetyl aluminum. Proc. Int. Soc. Citriculture, Tokyo 1:349-351. Day, A. and T.H.N. Ellis. 1984. Chloroplast DNA deletions associated with wheat plants regenerated from pollen: Possible basis for maternal inheritance of chloroplasts. Cell 39:359-368. Dayton, D.F. 1966. The pattern of distribution of anthocyanin in red pears. Proc. Amer. Soc. Hort. Sci. 89:110-116. DeCandole, A. P. 1824. Prodromus Systematis Natural is Regni Vegetabilis. 17 Vols. Treuttel & Wirtz, Paris. Decourtye, L. 1987. L'origine ontogenique des racines, puis des bourgeons, neoformes sur des chimeres de poirier (Pyrus communis L.) et de pommier {Malus pumila Mill.). Agronomie 7:27-32. Dermen, H. 1947. Periclinal cytochimeras and histogenesis in cranberry. Amer. J. Bot. 34:32-43. Dermen, H. 1960. Nature of plant sports. Amer. Hort. Mag. 39:123-173. Dertinger, H. and H. Jung. 1970. Molecular Radiation Biology. (translated by R.P.O. Huber and P. A. Gresham) Springer-Verlag, Berl in. Deshayes, A., L. Herrera-Estrella, and M. Caboche. 1985. Liposomemediated transformation of tobacco mesophyll protoplasts by an Escherichia coli plasmid. EMBO J. 4:2731-2737. Detrez, C, Sangwan, R.S., and B.S. Sangwan-Norreel 1989. Phenotypic and karyotypic status of Beta vulgaris plants regenerated from direct organogenesis in petiole culture. Theor. Appl Genet. 77:462-468. Devaux, R. 1981. New cultivars of Clementine mandarin in Morocco. Proc. Int. Soc. Citriculture, Tokyo 1:101-102. DeWald, S.G. and G.A. Moore. 1987. Somaclonal variation as a tool for the improvement of perennial fruit crops. Fruit Var. J. 41:54-57. Doebley, J. 1990. Molecular evidence for gene flow among lea species. Bioscience 40:443-448. Dolan, T.E. and M.D. Coffey. 1986. Laboratory screening technique for assessing resistance of four avocado rootstocks to Phytophthora cinnamomi. Plant Dis. 70:115-118.

PAGE 186

179 Dolezel, J., S.L. Lucretti, and F.J. Novak. 1987. The influence of 2,4-dichlorophenoxyacetic acid on cell cycle kinetics and sister-chromatid exchange frequency in garlic (Allium sativum L.) meristem cells. Biologia Plantarum 29:253-257. Dolezel, J. and F.J. Novak. 1984. Effect of plant tissue culture media on the frequency of somatic mutations in Tradescantia stamen hairs. Z. Pflanzenphysiol 114:51-58. Dolezel, J. and F.J. Novak. 1985. Karyological and cytophotometric study of callus induction in Allium sativum L. J. Plant Physiol. 118:421-429. Dolezel, J. and F.J. Novak. 1986. Sister chromatid exchanges in garlic (Allium sativum L.) callus cells. Plant Cell Reports 5:280-283. Dommergues, P. and J. Gillot. 1973. Obtention de clones genetiquement homogenes dans toutes leur couches ontogenetiques a partir d'une chimere d'oeillet americain. Ann. Amelior. Plant. 23:83-95. Doodeman, M. and F. Bianchi. 1985. Genetic analysis of instability in Petunia hybrida. 3. Periclinal chimeras resulting from frequent mutations of unstable alleles. Theor. Appl Genet. 69:297-304. Dulieu, H. and M. Barbier. 1982. High frequency of genetic variant plants regenerated from cotyledons of tobacco, pp. 211-227. In: E.D. Earle and Y. Demarly (eds.), Variability in Plants Regenerated from Tissue Culture. Praeger, New York. Dunbar, K.B. and C.T. Stephens. 1989. An in vitro screen for detecting resistance in Pelargonium somaclones to bacterial blight of geranium. Plant Dis. 73:910-912. Durante, M. C. Geri, J. Grisvard, E. Guille, R. Parenti, and M. Buiatti. 1983. Variation in DNA complexity in Nicotiana glauca tissue cultures. I. Pith tissue dedifferentiation in vitro. Protoplasma 114:114-118. Duran-Vila, N., V. Ortega, and L. Navarro. 1989. Morphogenesis and tissue cultures of three citrus species. Plant Cell, Tissue and Organ Culture 16:123-133. Eapen, S. 1976. Effect of gammaand ultraviolet-irradiation on survival and totipotency of haploid tobacco cells in culture. Protoplasma 89:149-155.

PAGE 187

180 Eapen, S., V. Abraham, M. Gerdemann, and 0. Schieder. 1989. Direct somatic embryogenesis, plant regeneration and evaluation of plants obtained from mesophyll protoplasts of Brassica juncea. Annals of Botany 63:369-372. Easton, G.D. and M.E. Nagle. 1981. Selecting superior clones of 'Russet Burbank' potato from Verticillium wilt resistant giant hill plants. Am. Potato J. 58:501. (Abstract) Edallo, S., C. Zucchinali, M. Perenzin, and F. Salamini. 1981. Chromosomal variation and frequency of spontaneous mutation associated with in vitro culture and plant regeneration in maize. Maydica 26:39-56. Einset, J., M.W. Blaser, and B. Imhoff. 1947. Chimerical sports of apple. J. Hered. 38:371-376. Einset, J. and C. Pratt. 1959. Spontaneous and induced apple sports with misshapen fruit. Proc. Amer. Soc. Hort. Sci. 73:1-8. Eizenga, G.C. 1989. Meiotic analyses of tall fescue somaclones. Genome 32:373-379. Ellstrand, N.C. and C.A. Hoffman. 1990. Hybridization as an avenue of escape for engineered genes. Bioscience 40:438-442. Engler, D.E. and R.G. Grogan. 1984. Variation in lettuce plants regenerated from protoplasts. J. Hered. 75:426-430. Evans, D.A. 1989. Somaclonal variation Genetic basis and breeding applications. Trends in Genetics 5:46-51. Evans, D.A. and J.E. Bravo. 1986. Phenotypic and genotypic stability of tissue cultured plants, pp. 73-94. In: R.H. Zimmerman, R.J. Griesbach, F.A. Hammerschlag, and R.H. Lawson (eds.), Tissue Culture as a Plant Production System for Horticultural Crops. Martinus Nijhoff Publishers, Dordrecht. Evans, D.A. and E.F. Paddock. 1976. Comparison of somatic crossing over frequency in Nicotiana tabacum and three other crop species. Can. J. Genet. Cytol 18:57-65. Evans, D.A. and W.R. Sharp. 1983. Single gene mutations in tomato plants regenerated from tissue culture. Science 221:949-951. Evans, D.A. and W.R. Sharp. 1986a. Somaclonal and gametoclonal variation, pp. 97-132. In: D. Evans, W. Sharp, and P. Ammirato (eds.), Handbook of Plant Cell Culture, Vol. 4. Techniques and Applications. Macmillan Publishing, New York. Evans, D.A. and W.R. Sharp. 1986b. Applications of somaclonal variation. Bio/Technology 4:528-532.

PAGE 188

181 Evans, D.A., C.E. Flick, and W.R. Sharp. 1989. Generation of somaclonal non-Mendel i an variants. United States Patent No. 4,827,079. Evans, D.A., W.R. Sharp, and H.P. Medina-Filho. 1984. Somaclonal and gametoclonal variation. Amer. J. Bot. 71:759-774. Fairchild, G.F. and P.W. league. 1987. The post-freeze U.S. grapefruit industry from a Florida and Texas perspective. Proc. Fla. State Hort. Soc. 100:99-103. Farih, A., J. A. Menge, P.H. Tsao, and H.D. Ohr. 1981. Metalaxyl and efosite aluminum for control of Phytophthora gummosis and root rot on citrus. Plant Dis. 65:654-657. Fawcett, W., and A.B. Rendle. 1920. Flora of Jamaica. Vol. IV: Dicotyledons. British Museum, London. Feher, F., M. Hangyel Tarczy, I. Bocsa, and D. Dudits. 1989. Somaclonal chromosome variation in tetraploid alfalfa. Plant Science 60:91-99. Ferrari, G.B. 1646. Hesperides; Sive, De Malorum Aureorum Cultura et Usu Libri Quatuor. Hermanni Scheus, Rome. Fischhoff, D.A., K.S. Bowdish, F.J. Perlak, P.G. Marrone, and S.M. McCormick. 1987. Insect tolerant transgenic tomato plants. Bio/Technology 5:807-813. Fishman, L., N. Primor, and E. Zlotkin. 1984. The selective permeability of fleshfly (Sarcophaga falculata) midgut to an orally toxic cobra venom cardiotoxin. J. Exp. Biol. 108:441-452. Flavell, R.B. 1985. Repeated sequences and genome change. pp. 139-152. In: B. Hohn and E.S. Dennis (eds.), Genetic Flux in Plants. Springer, Wien, New York. Flavell, R.B., M. 0'Dell, W.F. Thompson, M. Vincentz, R. Sardana, and R.F. Barker. 1986. The differential expression of ribosomal RNA genes. Phil. Trans. R. Soc. Lond. B 314:385-397. Florida Department of Agriculture and Consumer Services. 1988a. Commercial Citrus Inventory. Fla. Agric. Stat. Ser., Orlando. Florida Department of Agriculture and Consumer Services. 1988b. Unpublished data summary. Div. of Plant Ind. Bureau of Citrus Budwood Registration. Winter Haven, FL. Franklin, C.I., R.L. Mott, and T.M. Vuke. 1989. Stable ploidy levels in long-term callus cultures of loblolly pine. Plant Cell Reports 8:101-104.

PAGE 189

182 Freeling, M. 1984. Plant transposable elements and insertion sequences. Ann. Rev. Plant Physiol. 35:277-298. Freeman, W.G., and R.O. Williams. 1927. The Useful and Ornamental Plants of Trinidad and Tobago. Government Printing Office, Port-of-Spain, Trinidad. Freytag, A.H., A. P. Rao-Arelli, S.C. Anand, J. A. Wrather, and L.D. Owens. 1989. Somaclonal variation in soybean plants regenerated from tissue culture. Plant Cell Reports 8:199-202. Frost, H.B. 1926. Polyembryony, heterozygosis and chimeras in citrus. Hilgardia 1:365-402. Frost, H.B. 1943. Genetics and breeding, pp. 870-888. In: H.J. Webber and L.D. Batchelor (eds.), The Citrus Industry, Vol.1. Univ. of California Press, Berkeley. Frost, H.B. and C.A. Krug. 1942. Diploid-tetraploid periclinal chimeras as bud variants in citrus. Genetics 27:619-634. Frost, H.B. and R.K. Soost. 1968. Seed reproduction: development of gametes and embryos, pp. 290-324. In: W. Reuther, L.D. Batchelor, and H.J. Webber (eds.), The Citrus Industry, Vol. 2. Revised Edition. Univ. of California Press, Berkeley. Fukui, K. 1983. Sequential occurrence of mutations in a growing rice callus. Theor. Appl Genet. 65:225-230. Furner, L.J., J. King, and O.L. Gamborg. 1978. Plant regeneration from protoplasts isolated from a predominantly haploid suspension culture of Datura innoxia. Plant Sci. Letters 11:169-176. Furr, J.R. and J.B. Carpenter. 1961. Program for breeding citrus rootstocks tolerant to Phytophthora root rot. Proc. Fla. State Hort. Soc. 74:18-23. Furr, J.R., W.C. Cooper, and P.C. Reece. 1947. An investigation of flower formation in adult and juvenile citrus trees. Amer. J. Bot. 34:1-8. Furr, J.R. and P.C. Reece. 1946. Identification of hybrid and nucellar citrus seedlings by a modification of the rootstock color test. Proc. Amer. Soc. Hort. Sci. 48:141-146. Galiba, G. and J. Sutka. 1989. Frost resistance of somaclones derived from Triticum aestivum L. winter wheat calli. Plant Breeding 102:101-104. Gallesio, G. 1811. Traite du Citrus. Louis Fantin, Paris.

PAGE 190

183 Galun, E. and D. Raveh. 1975. In vitro culture of tobacco protoplasts: Survival of haploid and diploid protoplasts exposed to X-ray radiation at different times after isolation. Radiation Bot. 15:79-82. Gasser, C.S. and R.T. Fraley. 1989. Genetically engineering plants for crop improvement. Science 244:1293-1299. Gavazzi, G., C. Tonelli, G. Todesco, E. Arreghini, F. Raffaldi, F. Vecchio, G. Barbuzzi, M.G. Biasini, and F. Sala. 1987. Somaclonal variation versus chemically induced mutagenesis in tomato (Lycopersicon esculentum L.). Theor. Appl Genet. 74:733-738. Gengenbach, B.G., J. A. Connelly, D.R. Pring, and M.F. Conde. 1981. Mitochondrial DNA variation in maize plants regenerated during tissue culture selection. Theor. Appl. Genet. 59:161-167. Gengenbach, B.G., C.E. Green, and CM. Donovan. 1977. Inheritance of selected pathotoxin resistance in maize plants regenerated from cell cultures. Proc. Natl. Acad. Sci. USA 74:5113-5117. George, L. and P.S. Rao. 1983. Yellow seeded variants in in vitro regenerants of mustard (Brassica juncea Coss. var. Rai-5). Plant Sci. Letters 80:327-330. Gifford, E.M. Jr. 1954. The shoot apex in angiosperms. Bot. Rev. 20:477-529. Gmitter, F.G., Jr. 1985. Citrus embryogenesis in vitro: Culture initiation, plant regeneration, and phenotypic characterization. PhD Dissertation: University of Florida, Gainesville. Gmitter, F.G., Jr. and G.A. Moore. 1986. Plant regeneration from undeveloped ovules and embryogenic calli of Citrus: embryo production, germination, and plant survival. Plant Cell, Tissue and Organ Culture 6:139-147. Gottschalk, W. and G. Wolff. 1983. Induced Mutations in Plant Breeding. Monographs on Theoretical and Applied Genetics No. 7. Springer-Verlag, Berlin. Graham, J.H. In press. Evaluation of tolerance of citrus rootstocks to Phytophthora root rot in chlamydospore-infested soil. Plant Dis. Graham, J.H. and D.S. Egel 1988. Phytophthora root rot development on mycorrhizal and phosphorus-fertilized nonmycorrhizal sweet orange seedlings. Plant Dis. 72:611-614.

PAGE 191

184 Grandbastien, M.A., C. Missonier, J. Goujaud, J. P. Bourgin, A. Deshayes, and M. Caboche. 1989. Cellular genetic study of a somatic instability in a tobacco mutant: In vitro isolation of valine-resistant spontaneous mutants. Theor. Appl Genet. 77:482-488. Green, R.M., A. Vardi, and E. Galun. 1986. The plastome of Citrus. Physical map, variation among Citrus cultivars and species and comparison with related genera. Theor. Appl. Genet. 72:170-177. Griesbach, R.J. 1989. Selection of a dwarf Hemerocallis through tissue culture. HortScience 24:1027-1028. Grimm, G.R. and D.J. Hutchison. 1973. A procedure for evaluating resistance of citrus seedlings to Phytophthora parasitica. Plant Dis. Reporter 57:669-672. Grimm, G.R. and D.J. Hutchison. 1977. Evaluation of Citrus spp., relatives, and hybrids for resistance to Phytophthora parasitica Dastur. Proc. Int. Soc. Citriculture, Orlando 3:863-865. Groose, R.W. and E.T. Bingham. 1984. Variation in plants regenerated from tissue culture of tetraploid alfalfa, heterozygous for several traits. Crop Sci. 24:655-658. Groose, R.W. and E.T. Bingham. 1986a. An unstable anthocyanin mutation recovered from tissue culture of alfalfa (Medicago sativa). 1. High frequency of reversion upon reculture. Plant Cell Reports 5:104-107. Groose, R.W. and E.T. Bingham. 1986b. An unstable anthocyanin mutation recovered from tissue culture of alfalfa (Medicago sativa). 2. Stable nonrevertants derived from reculture. Plant Cell Reports 5:108-110. Grosser, J.W. and F.G. Gmitter, Jr. 1989. Somatic hybridization of Citrus with wild relatives for germplasm enhancement and cultivar development. HortScience 25:147-151. Grosser, J.W. and F.G. Gmitter, Jr. 1990. Strategies and methodologies for the application of protoplast fusion to citrus cultivar improvement. Plant Breeding Reviews 8:339-374. Grosser, J.W., F.G. Gmitter, Jr., and J.L. Chandler. 1988a. Intergeneric somatic hybrid plants of Citrus sinensis cv. Hamlin and Poncirus trifoliata cv. Flying Dragon. Plant Cell Reports 7:5-8. Grosser, J.W., F.G. Gmitter, Jr., and J.L. Chandler. 1988b. Intergeneric somatic hybrid plants from sexually incompatible woody species: Citrus sinensis and Severinia disticha. Theor. Appl. Genet. 75:397-401.

PAGE 192

185 Grosser, J.W., G.A. Moore, and F.G. Gmitter, Jr. 1989. Interspecific somatic hybrid plants from the fusion of 'Key' lime (Citrus aurantifolia) with 'Valencia' sweet orange (Citrus sinensis) protoplasts. Scientia Horticulture 39:23-29. Gulin, V.V., V.A. Kalchenko, H.Y. Saidov, M.R. Mirzaev, and L.T. Tsupina. 1989. Ionizing radiation induces electrophoretic variants of glutamate-oxaloacetate-transaminase in cotton. Genetika 25:321-327. Hagedoorn, A.C. and A.L. Hagedoorn. 1914. Another hypothesis to account for Dr. Swingle's experiments with Citrus. Amer. Nat. 48:446-448. Hakman, I., S. von Arnold, and A. Bengtsson. 1984. Cytofluorometric measurements of nuclear DNA in adventitious buds and shoots of Picea abies regenerated in vitro. Physiol. Plant. 60:321-325. Hall, H.K., D. Cohen, and R.M. Skirvin. 1986a. The inheritance of thornlessness from tissue culture-derived 'Thornless Evergreen' blackberry. Euphytica 35:891-898. Hall, H.K., M.H. Quazi, and R.M. Skirvin. 1986b. Isolation of a pure thornless loganberry by meristem tip culture. Euphytica 35:1039-1044. Hall, H.K., R.M. Skirvin, and W.F. Braam. 1986c. Germplasm release of 'Lincoln Logan', a tissue culture-derived genetic thornless 'Loganberry'. Fruit Var. J. 40:134-135. Hanisch Ten Cate, C.H. and K. Sree Ramulu. 1987. Callus growth, tumour development and polyploidization in the tetraploid potato cultivar Bintje. Plant Science 49:209-216. Hansel, H. 1967. Model for a theoretical estimate of optimal mutation rates per Ml-nucleus with a view to selecting beneficial mutations in different M-generations. pp. 125-129. In: Induced Mutations and their Utilization. Akademie-Verlag, Berl in. Harlan, J.R. 1971. Agricultural origins: Centers and noncenters. Science 174:468-474. Harlan, J.R. 1975. Crops and Man. American Society of Agronomy and Crop Science Society of America, Madison, WI. Hartman, C.L., T.J. McCoy, and T.R. Knous. 1984. Selection of alfalfa (Medicago sativa) cell lines and regeneration of plants resistant to the toxin (s) produced by Fusarium oxysporum f.sp. medicagim's. Plant Sci. Letters 34:183-194.

PAGE 193

186 Hartmann, C, J. DeBuyser, Y. Henry, D. Falconet, B. Lejeune, A. Benslimane, F. Quetier, A. Rode. 1987. Time-course of mitochondrial genome variation in wheat embryogenic somatic tissue cultures. Plant Sci 53:191-198. Hartmann, C, Y. Henry, J. DeBuyser, C. Aubry, and A. Rode. 1989. Identification of new mitochondrial genome organizations in wheat plants regenerated from somatic tissue cultures. Theor. Appl Genet. 77:169-175. Hawkes, J.G. 1983. The Diversity of Crop Plants. Harvard University Press, Cambridge, MA. Hearn, C.J. 1984. Development of seedless orange and grapefruit cultivars through seed irradiation. J. Amer. Soc. Hort. Sci. 109:270-273. Hearn, C.J. 1986. Development of seedless grapefruit cultivars through budwood irradiation. J. Amer. Soc. Hort. Sci. 111:304-306. Heath-Pagliuso, S., J. Pullman, and L. Rappaport. 1988. Somaclonal variation in celery: screening for resistance to Fusari urn oxysporum f.sp. apii. Theor. Appl. Genet. 75:446-451. Heinz, D.J. and G.W.P. Mee. 1971. Morphologic, cytogenetic, and enzymatic variation in Saccharum species hybrid clones derived from callus tissue. Amer. J. Bot. 58:257-262. Heinz, D.J., G.W.P. Mee, L.G. Nickel 1 1969. Chromosome numbers of some Saccharum species hybrids and their cell suspension cultures. Amer. J. Bot. 56:450-456. Heinz, D.J., M. Krishnamurthi L.G. Nickel 1 and A. Maretzki 1977. Cell, tissue, and organ culture in sugar cane improvement, pp. 1-17. In: J. Reinert and Y.P.S. Bajaja (eds.), Applied and Fundamental Aspects of Plant Cell, Tissue and Organ Culture. Springer-Verlag, Berlin. Helgeson, J. P., G.T. Haberlach, and CD. Upper. 1976. A dominant gene conferring disease resistance to tobacco plants is expressed in tissue cultures. Phytopathology 66:91-96. Hell, K.G. 1983. Survival of Nicoti ana tabacum L. cv. Wisconsin-38 plants regenerated from gammairradiated tissue cultures. Envir. Exp. Bot. 23:139-142. Hensz, R.A. 1971. 'Star Ruby', a new deep-red-fleshed grapefruit variety with distinct tree characteristics. J. Rio Grande Valley Hort. Soc. 25:54-58.

PAGE 194

187 Hensz, R.A. 1977. Mutation breeding and the development of the 'Star Ruby' grapefruit. Proc. Int. Soc. Citriculture, Orlando 2:582-585. Hensz, R.A. 1978. 'Ray Ruby' grapefruit, a mutant of 'Ruby Red' with redder flesh and peel color. J. Rio Grande Valley Hort. Soc. 32:39-41. Hensz, R.A. 1981. Bud mutations in citrus cultivars in Texas. Proc. Int. Soc. Citriculture, Tokyo 1:89-91. Hensz, R.A. 1985. 'Rio Red', a new grapefruit with a deep-red color. J. Rio Grande Valley Hort. Soc. 38:75-76. Hepburn, A.G., L.E. Clarke, L. Pearson, and J. White. 1983. The role of cytosine methylation in the control of nopal ine synthase gene expression in a plant tumor. J. Mol Appl Genet. 2:315-329. Herskowitz, I.H. 1962. Genetics. Little, Brown and Company, Boston. Hiatt, A., R. Cafferkey, and K. Bowdish. 1989. Production of antibodies in transgenic plants. Nature 342:76-78. Hidaka, T. and I. Kajiura. 1988. Plantlet differentiation from callus protoplasts induced from Citrus embryo. Scientia Horticulturae 34:85-92. Hodgson, R.W. 1961. Taxonomy and nomenclature in citrus. Intern. Org. Citrus Virol. Proc. 2:1-7. Hodgson, R.W. 1967. Horticultural varieties of citrus, pp. 431-591. In: W. Reuther, H.J. Webber, and L.D. Batchelor (eds.), The Citrus Industry, Vol. 1. Revised Edition. Univ. of California Press, Berkeley. Hodgson, R.W., R. Singh, and D. Singh. 1963a. Some little known Indian citrus species. Calif. Citrograph 48:188, 211-214. Hodgson, R.W., R. Singh, and D. Singh. 1963b. Some little known Indian citrus species. Calif. Citrograph 48:288-294. Hodgson, R.W., R. Singh, and D. Singh. 1963c. Some little known Indian citrus species. Calif. Citrograph 48:357-358, 360-364. Hoffman, C.A. 1990. Ecological risks of genetic engineering of crop plants. Bioscience 40:434-437. Holliday, R. The inheritance of epigenetic defects. Science 238:163-170.

PAGE 195

188 Holliday, R. 1989. A different kind of inheritance. Scientific American 260(6) :60-73. Horsch, R.B., J.E. Fry, N.L. Hoffmann, D. Eichholtz, S.G. Rogers, and R.T. Fraley. 1985. A simple and general method for transferring genes into plants. Science 227:1229-1231. Hughes, G. 1750. The Natural History of Barbados. Book V. Printed for the author, London. (Reprinted in 1972 by Arno Press, New York.) Huitema, J.B.M., W. Preil, G.C. Gussenhoven, and M. Schneidereit. 1989. Methods for the selection of low-temperature tolerant mutants of Chrysanthemum morifolium Ramat. by using irradiated cell suspension cultures: I. Selection of regenerants In vivo under suboptimal temperature conditions. Plant Breeding 102:140-147. Hume, H.H. 1926. The Cultivation of Citrus Fruits. Macmillan, New York. Hutchison, D.J. 1985. Rootstock development screening and selection for disease tolerance and horticultural characteristics. Fruit Var. J. 39:21-25. Hutchison, D.J. and G.R. Grimm. 1972. Variation in Phytophthora resistance of Florida rough lemon and sour orange clones. Proc. Florida State Hort. Soc. 85:38-39. IAEA. 1977. Manual on Mutation Breeding. 2nd Edition. International Atomic Energy Agency, Vienna. IAEA. 1985. Mutation Breeding for Disease Resistance Using In-Vitro Culture Techniques. International Atomic Energy Agency, Vienna. TEC-DOC-342. Inglett, G.E. and J.F. May. 1968. Tropical plants with unusual taste properties. Econ. Bot. 22:326-331. Ingram, D.S. 1967. The expression of R-gene resistance to Phytophthora infestans in tissue cultures of Solarium tuberosum. J. Gen. Microbiol. 49:99-108. Ingram, D.S. 1983. Prospects and challenges for the future, pp. 163-188. In: Use of Tissue Culture and Protoplasts in Plant Pathology. Academic Press, Australia. Ingram, D.S. and M.V. MacDonald. 1986. In vitro selection of mutants, pp. 241-257. In: Nuclear Techniques and In-Vitro Culture for Plant Improvement, Proceedings of a Symposium, Vienna 1985. International Atomic Energy Agency, STI/PUB/698.

PAGE 196

189 Iwamasa, M. and M. Nishiura. 1970. Evidence for the chimeral nature of vegetative reversion in 'Suzuki Wase,' an early-ripening satsuma mandarin. SABRAO Newsletter 2:109-114. Iwamasa, M. and M. Nishiura. 1981. Recent citrus mutant selections in Japan. Proc. Int. Soc. Citriculture, Tokyo 1:96-99. Iwamasa, M. M. Nishiura, N. Okudai, and D. Ishiuchi. 1977. Characteristics due to chimeras and their stability in citrus cultivars. Proc. Int. Soc. Citriculture, Orlando 2:571-574. Iwamasa, M., N. Nito, S. Yamaguchi, T. Kuriyama, T. Ehara, and T. Nakamuta. 1981. Occurrence of very early mutants from the wase (early ripening) satsumas. Proc. Int. Soc. Citriculture, Tokyo 1:99-101. Iyengar, R.B., P. Smits, H. Van der Wei, F.J.G. Van der Ouderaa, J.H. Van Brouwershaven, P. Ravenstein, G. Richters, and P.D. Van Wassenaar. 1979. The complete amino acid sequence of the sweet protein thaumatin I. Eur. J. Biochem. 96:193-204. Jablonka, E. and M.J. Lamb. 1989. The inheritance of acquired epigenetic variations. J. Theor. Biol. 139:69-83. Jackson, J. A. and P.J. Dale. 1989. Somaclonal variation in Lolium multiflorum L. and L. temulentum L. Plant Cell Reports 8:161-164. Jain, R.K., D.R. Sharma, and J.B. Chowdhury. 1989. High frequency regeneration and heritable somaclonal variation in Brassica juncea. Euphytica 40:75-81. James, D.J., A.J. Passey, D.J. Barbara, and M. Bevan. 1989. Genetic transformation of apple (Malus pumila Mill.) using a disarmed Ti-binary vector. Plant Cell Reports 7:658-661. Jeffers, S.N., H.S. Aldwinckle, and P. A. Arneson. 1981. Excised twig assay for the study of apple tree crown rot pathogens in vitro. Plant Dis. 65:823-825. Johnson, L.B., D.L. Stuteville, S.E. Schlarbaum, and D.Z. Skinner. 1984. Variation in phenotype and chromosome number in alfalfa protoclones regenerated from nonmutagenized calli. Crop Sci. 24:948-951. Johnson, R.T. 1980. Gamma irradiation and In vitro induced separation of chimeral genotypes in carnation. HortScience 15:605-606. Johnson, S.S., R.L. Phillips, and H.W. Rines. 1987. Possible role of heterochromatin in chromosome breakage induced by tissue culture in oats {Avena sativa L.). Genome 29:439-446.

PAGE 197

190 Jones, W.N. 1934. Plant Chimeras and Graft Hybrids. Methuen & Co., London. Jorgensen, C.A. and M.B. Crane. 1927. Formation and morphology of Solanum chimeras. J. Genet. 18:247-273. Kao, K.N. and M.R. Michayluk. 1980. Plant regeneration from mesophyll protoplasts of alfalfa. Z. Pflanzenphysiol 96:135-141. Kaplan, H.J. 1986. Washing, waxing, and color-adding, pp. 379-395. In: W.F. Wardowski, S. Nagy and W. Grierson (eds.), Fresh Citrus Fruits. AVI Publishing, Westport, Conn. Karp, A. and S.W.J. Bright. 1985. On the causes and origin of somaclonal variation. Oxford Surv. Plant Mol Cell. Biol. 2:199-234. Karp, A., R. Nelson, E. Thomas, and S.J. Bright. 1982. Chromosome variation in protoplast-derived potato plants. Theor. Appl Genet. 63:265-272. Kemble, R.J. and J.F. Shepard. 1984. Cytoplasmic DNA variation in a potato protoclonal population. Theor. Appl. Genet. 69:211-216. Kennedy, L.M., L.R. Saul, R. Sefecka, and D.A. Stevens. 1988. Hodulcin: selective sweetness-reducing principle from Hovenia dulcis leaves. Chemical Senses 13:529-543. Kephart, S.R. 1990. Starch gel electrophoresis of plant isozymes: A comparative analysis of techniques. Amer. J. Bot. 77:693-712. Kibler, R. and K.H. Neumann. 1980. On cytogenetic stability of cultured tissue and cell suspensions of haploid and diploid origin, pp. 59-65. In: F. Sala, B. Parisi, R. Cella and 0. Ciferi (eds.), Plant Cell Cultures: Results and Perspectives. Elsevier, Essex, UK. Kleinhofs, A., C. Sander, R.A. Nilan, and C.F. Konzak. 1974. Azide mutagenicity Mechanism and nature of mutants produced, pp. 195-199. In: Polyploidy and Induced Mutations in Plant Breeding. International Atomic Energy Agency, Vienna. Klotz, L.J., W.P. Bitters, T.A. DeWolfe, and M.J. Garber. 1968. Some factors in resistance of citrus to Phytophthora spp. Plant Dis. Reporter 52:952-955. Klotz, L.J. and E.C. Calavan. 1978. Gum diseases of citrus in California. University of California Leaflet 2593.

PAGE 198

191 Kobayashi, S. 1987. Uniformity of plants regenerated from orange (Citrus sinensis Osb.) protoplasts. Theor. Appl Genet. 74:10-14. Kobayashi, S. and T. Ohgawara. 1988. Production of somatic hybrid plants through protoplast fusion in citrus. J. Agr. Review Quarterly 22:181-188. Kobayashi, S., T. Ohgawara, E. Ohgawara, I. Oiyama, and S. Ishii. 1988. A somatic hybrid plant obtained by protoplast fusion between navel orange (Citrus sinensis) and satsuma mandarin. Plant Cell, Tissue and Organ Culture 14:63-69. Kobayashi, S., H. Uchimaya, and I. Ikeda. 1983. Plant regeneration from 'Trovita' orange protoplasts. Jpn. J. Breed. 33:119-122. Kochba, J. and P. Spiegel-Roy. 1982. Progress in selection for sodium chloride, 2,4-D dichlorophenoxy acetic acid (2,4-D) and streptomycin tolerance in Citrus sinensis ovular callus lines, pp. 77-89. In: Induced Mutations in Vegetatively Propagated Plants II. Proceedings from Coimbatore, India, 11-15 February 1980. IAEA, Vienna. Kochba, J., P. Spiegel-Roy, and H. Safran. 1972. Adventive plants from ovules and nucelli in Citrus. Planta 106:237-245. Konar, R.N. and R. Nagmani. 1972. Chromosome number in callus cultures of Pinus gerardiana Wall. Curr. Sci 41:714-715. Konzak, C.F. 1984. Role of induced mutations, pp. 216-292. In: P.B. Vose and S.G. Blixt (eds.), Crop Breeding, a Contemporary Basis. Pergamon Press, Oxford. Krikorian, A.D., S.A. Staicu, and R.P. Kann. 1981. Karyotype analysis of a daylily clone reared from aseptically cultured tissues. Ann. Bot. 47:121-131. Krishnamurthi M. 1977. Sugarcane improvement through tissue culture. Proc. ISSCT 16(l):23-28. Krishnamurthi, M. and J. Tlaskal. 1974. Fiji disease resistant Saccharum officinarum var. Pindar subclone from tissue culture. Proc. Int. Soc. Sugar Cane Tech. 15:130-137. Kuehnert, C.C. 1962. Cytological and morphological changes induced in tomato as a result of thermal neutron irradiation. Radiation Bot. 2:81-88. Kuehnle, A.R. and E.D. Earle. 1989. In vitro selection for methomyl resistance in CMS-T maize. Theor. Appl. Genet. 78:672-682.

PAGE 199

192 Kuglik, P., J. Slotova, and Z. Karpfel 1987. Incidence of sister chromatid exchanges (SCEs) and chromosomal aberrations in root meristem cells of Vicia faba after gamma irradiation. Stud. Biophys. 121:75-80. Kuglik, P., J. Slotova, and Z. Karpfel. 1989. The effect of continuous gamma irradiation on formation of sister chromatid exchanges (SCEs) and chromosomal aberrations in meristematic cells of Vicia faba. Biologia Plantarum 31:133-138. Kumamoto, J., R.W. Scora, H.W. Lawton, and W.A. Clerx. 1987. Mystery of the forbidden fruit: historical epilogue on the origin of the grapefruit, Citrus paradisi (Rutaceae). Econ. Bot. 41:97-107. Kunz, B.A. 1982. Genetic effects of deoxyribonucleotide pool imbalance. Environ. Mutagen. 4:695-725. Kunz, B.A. and R.H. Haynes. 1982. DNA repair and genetic effects of thymidylate stress in yeast. Mutat. Res. 93:353-375. Landsmann, J. and H. Uhrig. 1985. Somaclonal variation in Solanum tuberosum detected at the molecular level. Theor. Appl Genet. 71:500-505. Lapitan, N.L.V., R.G. Sears, and B.S. Gill. 1984. Translocations and other karyotypic structural changes in wheat x rye hybrids regenerated from tissue culture. Theor. Appl. Genet. 68:547-554. Larkin, P.J. 1985. In vitro culture and cereal breeding. pp. 273-296. In: S.W.J. Bright and M.G.K. Jones (eds.), Cereal Tissue and Cell Culture. Martinus Nijhoff/Dr. W. Junk Publ., Dordrecht. Larkin, P.J. and W.R. Scowcroft. 1981. Somaclonal variationA novel source of variability from cell cultures for plant improvement. Theor. Appl. Genet. 60:197-214. Larkin, P.J. and W.R. Scowcroft. 1983. Somaclonal variation and eyespot toxin tolerance in sugarcane. Plant Cell, Tissue and Organ Culture 2:111-121. Larkin, P.J., S.A. Ryan, R.I.S. Brettell, and W.R. Scowcroft. 1984. Heritable somaclonal variation in wheat. Theor. Appl. Genet. 67:443-455. Latt, S. 1974. Sister chromatid exchange, indices of human chromosome damage and repair: Detection by fluorescence and induction by mitomicine-C. Proc. Natl. Acad. Sci. USA 71:3162-3166.

PAGE 200

193 Latunde-Dada, A.O. and J. A. Lucas. 1983. Somaclonal variation and reaction to Verticillium wilt in Medicago sativa L. plants regenerated from protoplasts. Plant Sci. Letters 32:205-211. Lazar, M.D., T. H.H. Chen, L.V. Gusta, and K.K. Kartha. 1988. Somaclonal variation for freezing tolerance in a population derived from norstar winter wheat. Theor. Appl Genet. 75:480-484. Lazar, M.D., G.B. Collins, and W.E. Vian. 1983. Genetic and environmental effects on the growth and differentiation of wheat somatic cell cultures. J. Hered. 74:353-357. Lazarovici, P., M. Menashe, E. Zlotkin. 1984. Toxicity to Crustacea due to polypeptide-phospholipase (EC 3.1.1.4) interaction in the venom of a chactoid scorpion [Scorpio maurus palmatus]. Arch. Biochem. Biophys. 229:270-286. Lea, D.E. 1955. Actions of Radiations on Living Cells. 2nd Edition. Cambridge University Press, New York. Ledeboer, A.M., C.T. Verrips, and B.M.M. Dekker. 1984. Cloning of the natural gene for the sweet-tasting plant protein thaumatin. Gene 30:23-32. Lee, M. and R.L. Phillips. 1987. Genomic rearrangements in maize induced by tissue culture. Genome 29:122-128. Lee, M. and R.L. Phillips. 1988. The chromosomal basis of somaclonal variation. Ann. Rev. Plant Physiol. 39:413-437. Lee, R.E. Jr. 1989. Insect cold-hardiness: To freeze or not to freeze. Bioscience 39:308-313. Legave, J.M., D. Tisne-Agostini et C. Jacquemond. 1989. Effets physiologiques induits par irradiation gamma aigue de rameaux greffons de Clementinier, Citrus reticulata Blanco. Fruits 44:329-333. Levy, A. and A. Ashri 1975. Ethidium bromide an efficient mutagen in higher plants. Mutat. Res. 28:397-404. Libbenga, K. and J. Torrey. 1973. Hormone-induced endoredupl ication prior to mitosis in cultured pea root cortex cells. Amer. J. Bot. 60:293-299. Lima-De-Faria, A. 1969. DNA replication and gene amplification in heterochromatin. pp. 234-282. In: A. Lima-De-Faria (ed.), Handbook of Molecular Cytology. North Holland, Amsterdam.

PAGE 201

194 Lindgren, D.L. and W.B. Sinclair. 1941. The ridging of citrus fruits as influenced by fumigation of citrus trees with HCN. J. Econ. Ent. 34:477. Little, J.B. 1989. In vitro models of carcinogenesis: Expression of recessive genes by chromosomal mutations. Environ. Health Perspect. 81:63-66. Liu, G.B. and S.S. Deng. 1985. Induced mutations in citrus shoot tip culture in vitro. Fruit Var. J. 39:38-43. Liu, M.C. and W.H. Chen. 1978. Tissue and cell culture as aids to sugar cane breeding. 2. Performance and yield potential of callus derived lines. Euphytica 27:273-282. Lorz, H. and P. Brown. 1986. Variability in tissue culture derived plants possible origins: Advantages and drawbacks, pp. 513534. In: W. Horn, C. Jensen, W. Odenbach, and 0. Schieder (eds.), Genetic Manipulation in Plant Breeding. Walter de Gruyter, Berlin. Lorz, H. and W.R. Scowcroft. 1983. Variability among plants and their progeny from protoplasts of Su/su heterozygotes of Nicotiana tabacum. Theor. Appl Genet. 66:67-75. Lourens, A.G. and F.A. Martin. 1987. Evaluation of in vitro propagated sugarcane hybrids for somaclonal variations. Crop Sci. 27:793-796. Lu, L.X. 1978. A preliminary study on the variation of improved sweet orange (Gail iangcheng) [In Chinese with English summary]. Acta Genetica Sinica 5(1): 49-51 Lu, L.X. 1982. Investigations on the genetic behaviors of variable characters and chimera structure in Gail iangcheng (an improved sweet orange) [In Chinese with English summary]. J. Fujian Agr. College 9(3):47-53. Lunan, J. 1814. Hortus Jamaicensis, Vol. 2. St. Jago de la Vega Gazette, Spanish Town, Jamaica. Lutz, A. and J. A. Menge. 1986. Phytophthora root rot. Citrograph 71:33-39. Macfadyen, J. 1830. Some remarks on the species of the genus Citrus which are cultivated in Jamaica. Bot. Misc. 1:295-304. Macfadyen, J. 1837. The Flora of Jamaica; A Description of the Plants of That Island. Longman, Orme, Brown, Green, and Longmans, London.

PAGE 202

195 Maddock, S.E. and J.T. Semple. 1986. Field assessment of somaclonal variation in wheat. J. Exp. Bot. 37:1065-1078. Maddock, S.E., R. Risiott, S. Parmar, M.G.K. Jones, and P.R. Shewry. 1985. Somaclonal variation in the gliadin patterns of grains of regenerated wheat plants. J. Exp. Bot. 36:1976-1984. Marks, G.E. 1953. Genetical studies in pears: IV. Giant bud sports. J. Hort. Sci. 28:141-144. Marcotrigiano, M. 1986. Experimentally synthesized plant chimeras 3. Qualitative and quantitative characteristics of the flowers of interspecific Nicotiana chimeras. Ann. Bot. 57:435-442. Marcotrigiano, M., H.J. Swartz, S.E. Gray, D. Tokarcik, and J. Popenoe. 1984. The effect of benzyl amino purine on the In vitro multiplication rate and subsequent field performance of tissue-culture propagated strawberry plants. Adv. Strawberry Prod. 3:23-25. Maretzki, A. 1987. Tissue culture: Its prospects and problems. pp. 343-384. In: D. Heinz (ed.), Sugarcane Improvement Through Breeding. Elsevier, New York. Matheron, M.E. and J.C. Matejka. 1989. Temporal changes in susceptibility of citrus phloem tissue to colonization by Phytophthora citrophthora and P. parasitica. Plant Dis. 73:408-411. Matheron, M.E. and J.C. Matejka. 1990. Differential virulence of Phytophthora parasitica recovered from citrus and other plants to rough lemon and tomato. Plant Dis. 74:138-140. Mathur, A.K., P.S. Ahuja, B. Pandey, A.K. Kukreja, and S. Mandal 1988. Screening and evaluation of somaclonal variations for quantitative and qualitative traits in an aromatic grass, Cymbopogon winterianus Jowitt. Plant Breeding 101:321-334. Maycock, J.D. 1830. Flora Barbadensis: A Catalogue of Plants, Indigenous, Naturalized, and Cultivated in Barbados. James Ridgway, London. Mayr, E. 1963. Animal Species and Evolution. Harvard Univ. Press., Cambridge, MA. McClintock, B. 1978. Mechanisms that rapidly reorganize the genome. Stadler Genet. Symp. 10:25-48. McClintock, B. 1984. The significance of responses of the genome to challenge. Science 226:792-800.

PAGE 203

196 McComb, J. A., J.M. Hinch, and A.E. Clarke. 1987. Expression of field resistance in callus tissue inoculated with Phytophthora cinnamomi. Phytopathology 77:346-351. McCornack, A. A. and W.F. Wardowski 1977. Degreening Florida citrus fruit: Procedures and physiology. Proc. Int. Soc. Citriculture, Orlando 1:211-215. McCoy, T.J. 1980. Cytogenetic stability of tissue cultures and regenerated plants of oats (Avena sativa L.) and corn (Zea mays L.). PhD Dissertation: University of Minnesota, St. Paul. McCoy, T.J., R.L. Phillips, and D. Cummings. 1978. Cytogenetic variability in plants regenerated from tissue cultures of oats, p. 101. 4th Intl. Congr. Plant Tissue Cell Culture, Calgary. McCoy, T.J., R.L. Phillips, and H.W. Rines. 1982. Cytogenetic analysis of plants regenerated from oat {Avena sativa) tissue cultures: High frequency of partial chromosome loss. Can. J. Genet. Cytol 24:37-50. McGranahan, G.H., C.A. Leslie, S.L. Uratsu, L.A. Martin, and A.M. Dandekar. 1988. Agrobacterium-mediated transformation of walnut somatic embryos and regeneration of transgenic plants. Bio/Technology 6:800-804. McHughen, A. 1987. Salt tolerance through increased vigor in a flax line (STS-II) selected for salt tolerance in vitro. Theor. Appl Genet. 74:727-732. McHughen, A. and M. Swartz. 1984. A tissue-culture derived salttolerant line of flax {Linum usitatissimum) J. Plant Physiol. 117:109-117. Mcintosh, T.P. 1945. Variations in potato varieties. Scot. Agric. 25:125-132. McPheeters, K. and R.M. Skirvin. 1989. Somaclonal variation among ex vitro 'Thornless Evergreen' trailing blackberries. Euphytica 42:155-162. Meins, F.J. 1983. Heritable variation in plant cell culture. Annu. Rev. Plant Physiol. 34:327-346. Mendel, K. 1981. Bud mutations in citrus and their potential commercial value. Proc. Int. Soc. Citriculture, Tokyo 1:86-89. Menge, J. A. 1986. Use of new systemic fungicides on citrus. Citrograph 71:245-252.

PAGE 204

197 Mezentsev, A.V. 1981. [Mass regeneration of plants from the cells and protoplasts of alfalfa] (In Russian). Dokl. Vses. Akad. S Kh. Nauk im. V.I. Lenina 4:22-23. Micke, A. and B. Donini. 1982. Use of induced mutations in improvement of seed propagated crops, pp. 2-9. In: Induced Variability in Plant Breeding, Proceedings of an International Symposium, Wageningen 1981, Wageningen Centre for Agricultural Publishing and Documentation. Micke, A., B. Donini, and M. Maluszynski. 1987. Induced mutations for crop improvement a review. Trop. Agric. (Trinidad) 64:256-278. Micke, A., M. Maluszynski, and B. Donini. 1985. Plant cultivars derived from mutation induction or the use of induced mutants in cross breeding. Mutation Breeding Review No. 3. International Atomic Energy Agency, Vienna. Miles, J., J. Chaplin, L. Burk, and A. Baumhover. 1981. Registration of 1-35 tobacco germplasm. Crop Sci 22:1160-1164. Miller, M.W. and W.M. Miller. 1987. Radiation hormesis in plants. Hlth. Phys. 52:607-616. Miller, S.A., L.C. Davidse, and D.P. Maxwell. 1984. Expression of genetic susceptibility, host resistance, and nonhost resistance in alfalfa callus tissue inoculated with Phytophthora megasperma. Phytopathology 74:345-348. Mitra, J. and F. Steward. 1961. Growth induction in cultures of Haplopappus gracilis. II. The behavior of the nucleus. Amer. J. Bot. 48:358-368. Montagno, T.J., R.D. Lineberger, and S.Z. Berry. 1989. Somaclonal and radiation induced variation in Lycopersicon esculentum. Envir. Exp. Botany 29:401-408. Morton, J.F. 1987. Fruits of Warm Climates. J.F. Morton, Miami. Moustafa, R.A.K., D.R. Duncan, and J.M. Widholm. 1989. The effect of gamma radiation and N-ethyl-N-nitrosourea on cultured maize callus growth and plant regeneration. Plant Cell, Tissue and Organ Culture 17:121-132. Muller, H.J. 1927. Artificial transmutation of the gene. Science 66:84-87. Murai, N. D.W. Sutton, M.G. Murray, J.L. Slightom, and D.J. Merlo. 1983. Phased in gene from bean is expressed after transfer to sunflower via tumor-inducing plasmid vectors. Science 222:476-482.

PAGE 205

198 Murashige, T. 1974. Plant propagation through tissue culture. Ann. Rev. Plant Physiol. 25:135-166. Murashige, T. and R. Nakano. 1965. Morphogenetic behavior of tobacco tissue cultures and implications of plant senescence. Amer. J. Bot. 52:819-827. Murashige, T. and D.P.H. Tucker. 1969. Growth factor requirements of citrus tissue culture. Proc. First Int. Citrus Symp. 3:1155-1161. Murata, M. 1985. [Frequencies of sister chromatid exchanges in cultured cells of wheat] (in Japanese). Jpn. J. Genet. 60:617. Murata, M. 1989. Effects of auxin and cytokinin on induction of sister chromatid exchanges in cultured cells of wheat {Triticum aestivum L.). Theor. Appl Genet. 78:521-524. Murata, M. and T.J. Orton. 1983. Chromosome structural changes in cultured celery cells. In Vitro 19:83-89. Murata, M. and T.J. Orton. 1984. Chromosome fusions in cultured cells of celery. Can. J. Genet. Cytol 26:395-400. Nagarajan, P. and P.D. Walton. 1987. A comparison of somatic chromosomal instability in tissue culture regenerants from Medicago media Pers. Plant Cell Reports 6:109-113. Nati, P. 1674. Florentina Phytologica Observatio de Malo Limonia Citrata-aurantia, Florentiae Vulgo la Bizzarria. Hippolyti de Naue, Florence. Nati, P. 1929. Memoria Sull'Agrume Bizzarria. Ristampa integrale del testo latino con traduz. e note del Dott. Attilio Ragionieri. Francesco Battiato. Catania. (Latin reprint of 1674 edition and Italian translation). Nauer, E.M., D.J. Gumpf, T.L. Carson, and J. A. Bash. 1988. New varieties for California. Citrograph 73:152-154, 156, 158. Nauer, E.M., C.N. Roistacher, T.L. Carson, and T. Murashige. 1983. In vitro shoot-tip grafting to eliminate citrus viruses and virus-like pathogens produces uniform budlines. HortScience 18:308-309. Navarro, L., J. Juarez, J.F. Ballester, J. A. Pina, and C. Ortega. 1979. Obtencion de plantas nucelares libres de virus de diversas variedades de agrios del Grupo Navel [C. sinensis (L.) Osbeck] por cultivo de ovulos In vitro. An. Inst. Nac. Invest. Agrar. Ser. Prot. Veg. 12:95-113.

PAGE 206

199 Navarro, L., J.M. Ortiz, and J. Juarez. 1985. Aberrant citrus plants obtained by somatic embryogenesis of nucelli cultured in vitro. HortScience 20:214-215. Navarro, L., C.N. Roistacher, and T. Murashige. 1975. Improvement of shoot-tip grafting in vitro for virus-free citrus. J. Amer. Soc. Hort. Sci. 100:471-479. Neale, S. 1976. Mutagenicity of nitrosamides and nitrosamidines in microorganisms and plants. Mutat. Res. 32:229-266. Nei, M. T. Maruyama, and R. Chakraborty. 1975. The bottleneck effect and genetic variability in populations. Evolution 29:1-10. Neil son-Jones, W. 1969. Plant Chimeras. 2nd Edition. Methuen & Co., London. Nelson, R.S., P. Powell Abel, and R.N. Beachy. 1987. Lesion and virus accumulation in inoculated transgenic tobacco plants expressing the coat protein gene of tobacco mosaic virus. Virology 158:126-132. Nishiura, M. and M. Iwamasa. 1970. Reversion of fruit color in nucellar seedlings from the Dobashibeni Unshu, a red color mutant of the satsuma mandarin. Bui. Hort. Res. Sta. Japan. Ser. B. (Okitsu) No. 10:1-5. Nito, N., J.T. Ling, M. Iwamasa, and Y. Katayama. 1989. Effects of gamma-irradiation on growth and embryogenesis of citrus callus. J. Japan. Soc. Hort. Sci. 58:283-287. Normand, F. 1988. Interet des marqueurs enzymatiques pour 1 'etude des agrumes. Fruits 43:569-577. Nuttall, V.W., L.H. Lyall, D.H. Lees, and H.A. Hamilton. 1968. Response of garden crop plants to low dose gamma irradiation of seeds. Can. J. PI. Sci. 48:409-410. Ogihara, Y. 1981. Tissue culture in Haworthia. 4. Genetic characterization of plants regenerated from callus. Theor. Appl Genet. 60:353-363. Ogura, H. 1976. The cytological chimeras in original regenerates from tobacco tissue cultures and in their offsprings. Jap. J. Gen. 51:161-174. Ohgawara, T., S. Kobayashi, S. Ishii, K. Yoshinaga, and I. Oiyama. 1989. Somatic hybridization in Citrus: Navel orange (C. sinensis Osb.) and grapefruit (C. paradisi Macf.). Theor. Appl. Genet. 78:609-612.

PAGE 207

200 Ohgawara, T., S. Kobayashi, E. Ohgawara, H. Uchimaya, and S. Ishii. 1985. Somatic hybrid plants obtained by protoplast fusion between Citrus sinensis and Poncirus trifoliata. Theor. Appl Genet. 71:1-4. Oiyama, I. and N. Okudai 1986. Production of colchicine-induced autotetraploid plants through micrografting in monoembryonic citrus cultivars. Jpn. J. Breeding 36:371-376. Olson, E.O., J.W. Cameron, and R.K. Soost. 1966. The Burgundy sport: further evidence of the chimeral nature of pigmented grapefruits. HortScience 1:57-58. Orton, T.J. 1983. Spontaneous electrophoretic and chromosomal variability in callus cultures and regenerated plants of celery. Theor. Appl. Genet. 67:17-24. Orton, T.J. 1984. Genetic variation in somatic tissues: Method or madness? pp. 153-189. In: D. Ingram (ed.), Advances in Plant Pathology, Vol. 2. Academic Press, New York. Osawa, I. 1912. Cytological and experimental studies in Citrus. J. Coll. Agr. Univ. Tokyo 4:83-116. Osifo, E.O., J.K. Webb, G.G. Henshaw. 1989. Variation amongst callus-derived potato plants, Solanum brevidens. J. Plant Physiol. 134:1-4. Ow, D.W., K.V. Wood, M. DeLuca, J.R. DeWet, D.R. Helinski, and S.H. Howell. 1986. Transient and stable expression of the firefly luciferase gene in plant cells and transgenic plants. Science 234:856-859. Ozias-Akins, P., M. Dujardin, W.W. Hanna, and I.K. Vasil. 1989. Quantitative variation recovered from tissue cultures of an apomictic, interspecific Pennisetum hybrid. Maydica 34:123-132. Park, C.H. and P.D. Walton. 1989. Morphology and cytology of tissue culture derived octoploid of Elymus canadensis and its selfed progeny. Plant Breeding 102:208-214. Parlevliet, J.E. and J.W. Cameron. 1959. Evidence on the inheritance of nucellar embryony in citrus. Proc. Amer. Soc. Hort. Sci. 74:252-260. Patel, K.R. and G.P. Berlyn. 1982. Genetic instability of multiple buds of Pinus coulteri regenerated from tissue culture. Can. J. For. Res. 12:93-101. Peeters, J. P. 1988. The emergence of new centres of diversity: Evidence from barley. Theor. Appl. Genet. 76:17-24.

PAGE 208

201 Pehrson, J.E. and E.M. Ivans. 1988. Maturity and consumer acceptance in early navels. Citrograph 73:166-167. Peloquin, S.J. 1981. Chromosomal and cytoplasmic manipulations, pp. 117-150. In: K.J. Krey (ed.), Plant Breeding II. Iowa State Univ. Press, Iowa. Pereau-Leroy, P. 1974. Genetic interaction between the tissues of carnation petals as periclinal chimeras. Radiation Bot. 14:109-116. Perry, P. and H.J. Evans. 1975. Cytological detection of mutagencarcinogen exposure by sister chromatid exchange. Nature 258:121-125. Peschke, V.M., R.L. Phillips, and B.G. Gengenbach. 1987. Discovery of transposable element activity among progeny of tissue culture-derived maize plants. Science 238:804-807. Peterson, P. A. 1986. Mobile elements in maize. Plant Breed. Rev. 4:81-122. Pimentel, D., M.S. Hunter, J. A. LaGro, R.A. Efroymson, J.C. Landers, F.T. Mervis, C.A. McCarthy, and A.E. Boyd. 1989. Benefits and risks of genetic engineering in agriculture. Bioscience 39:606-614. Poethig, R. and I. Sussex. 1985. The cellular parameters of leaf development in tobacco: A clonal analysis. Planta 165:170-184. Powell Abel, P., R.S. Nelson, B. De, N. Hoffmann, S.G. Rogers, R.T. Fraley, and R.N. Beachy. 1986. Delay of disease development in transgenic plants that express the tobacco mosaic virus coat protein. Science 232:738-743. Prakash, S. 1972. Origin of reproductive isolation in the absence of apparent genie differentiation in a geographic isolate of Drosophila pseudoobscura. Genetics 72:143-155. Prat, D. 1983. Genetic variability induced in Nicotiana by protoplast culture. Theor. Appl Genet. 64:223-230. Pratt, C. 1983. Somatic selection and chimeras, pp. 172-185. In: J.N. Moore and J. Janick (eds.), Methods in Fruit Breeding. Purdue Univ. Press, West Lafayette, IN. Prols, M., J. Schell, and H.H. Steinbiss. 1989. Critical evaluation of electromediated gene transfer and transient expression in plant cells, pp. 367-376. In: E. Neumann, A.E. Sowers, and C.A. Jordan, (eds.), Electroporation and Electrofusion in Cell Biology. Plenum Publishing, New York.

PAGE 209

202 Quemada, H., E.J. Roth, and K.G. Lark. 1987. Changes in methylation of tissue cultured soybean cells detected by digestion with the restriction enzymes Hpall and Mspl. Plant Cell Reports 6:63-66. Quicke, D. 1988. Spiders bite their way towards safer insecticides. New Scientist 26 Nov:38-41. Raffa, K.F. 1989. Genetic engineering of trees to enhance resistance to insects. Bioscience 39:524-534. Ramulu, K.S., P. Dijkhuis, and S. Roest. 1989. Patterns of phenotypic and chromosome variation in plants derived from protoplast cultures of monohaploid, dihaploid and diploid genotypes and in somatic hybrids of potato. Plant Science 60:101-110. Ranch, J. P., S. Rick, J.E. Brotherton, and J.M. Widholm. 1983. Expression of 5-methyl tryptophan resistance in plants regenerated from resistant cell lines of Datura innoxia. Plant Physiol. 71:136-140. Redd, J.B., CM. Hendrix, Jr., and D.L. Hendrix. 1986. Quality Control Manual for Citrus Processing Plants, Vol. 1. Intercit, Safety Harbor, FL. Reed, J.N. and B.V. Conger. 1985. Meiotic analyses of tall fescue plants regenerated from callus cultures. Environ. Exp. Bot. 25:277-284. Reisch, B. 1983. Genetic variability in regenerated plants. pp. 749-769. In: D.A. Evans, W.R. Sharp, P.V. Ammirato, Y. Yamada (eds.), Handbook of Plant Cell Culture, Vol 1. Macmillan, New York. Renfroe, M.H. and G.P. Berlyn. 1984. Stability of nuclear DNA content during adventitious shoot formation in Pinus taeda L. tissue culture. Amer. J. Bot. 71:268-272. Risso, A. et A. Poiteau. 1818-1822. Histoire Naturelle des Orangers. Audot, Paris. Risso, A. et A. Poiteau. 1872. Histoire et Culture Des Orangers. Nouvelle edition. H. Plon et G. Masson, Paris. Robinson, T. R. 1947. Count Odette Phillippi— A correction to Florida's Citrus history. Proc. Fla. State Hort. Soc. 60:90-92. Robinson, T.R. 1952. Grapefruit and pummelo. Econ. Bot. 6:228-245.

PAGE 210

203 Rode, A., C. Hartmann, D. Falconet, B. Lejeune, F. Quetier, A. Benslimane, Y. Henry, J. de Buyser. 1987. Extensive mitochondrial DNA variation in somatic tissue cultures initiated from wheat immature embryos. Curr. Genet. 12:369-376. Rogers, S.O. and A.J. Bendich. 1987. Ribosomal RNA genes in plants: Variability in copy number and in the intergenic spacer. Plant Mol. Biol. 9:509-520. Roth, E.J., B.L. Frazier, N.R. Apuya, and K.G. Lark. 1989. Genetic variation in an inbred plant: Variation in tissue cultures of soybean [Glycine max (L.) Merrill]. Genetics 121:359-368. Rowland, G.G., A. McHughen, and C. McOnie. 1988. Field evaluation on nonsaline soils of a somaclonal variant of McGregor flax selected for salt tolerance in vitro. Can. J. Plant Sci. 68:345-349. Rowland, G.G., A. McHughen, and C. McOnie. 1989. Field performance at saline-affected sites of a somaclonal variant of McGregor flax selected for salt tolerance in vitro. Can. J. Plant Sci. 69:49-60. Russo, F. 1981. Bud variations in citrus cultivars in Italy. Proc. Int. Soc. Citriculture, Tokyo 1:91-94. Russo, F., B. Donini, and S. Starrantino. 1981. Mutagenesis applied for citrus improvement. Proc. Int. Soc. Citriculture, Tokyo 1:68-70. Sacristan, M.D. 1971. Karyotypic changes in callus cultures from haploid and diploid plants of Crepis capillaris (L.) Wallr. Chromosoma 33:273-283. Sacristan, M.D. 1982. Resistance responses to Phoma lingam of plants regenerated from selected cell and embryonic cultures of haploid Brassica napus. Theor. Appl Genet. 61:193-200. Sagan, L.A. 1989. On radiation, paradigms, and hormesis. Science 245:574, 621. Sagawa, Y. and G.A.L. Mehlquist. 1957. The mechanism responsible for some X-ray induced changes in flower color of the carnation, Dianthus caryophyllus L. Amer. J. Bot. 44:397-403. Saghai-Maroof, M.A., K.M. Soliman, R.A. Jorgensen, and R.W. Allard. 1984. Ribosomal DNA spacer-length polymorphism in barley: Mendel ian inheritance, chromosomal location, and population dynamics. Proc. Natl. Acad. Sci. USA 81:8014-8018. Salmia, M.A. 1975. Cytological studies on tissue culture of Pinus cembra. Physiol. Plant. 33:58-61.

PAGE 211

204 Samura, R. and K. Nakahara. 1928. Kobayashi Mikan, a periclinal chimera of Natsudaidai and satsuma (preliminary report). Nogyo Oyobi Engei (Japan) 3:1043-1047. Sandler, H.A., L.W. Timmer, J.H. Graham, and S.E. Zitko. 1989. Impact of fungicide applications on populations of Phytophthora parasitica, feeder root densities and fruit yields of citrus trees. Plant Dis. 73:902-906. Sanford, J.C. 1988. The biolistic process. Trends in Biotechnology 6:299-302. Sanford, J.C. N.F. Weeden, and Y.S. Chyi. 1984. Regarding the novelty and breeding value of protoplast-derived variants of Russet Burbank (Solanum tuberosum L.). Euphytica 33:709-715. Sansavini, S. and G. Gherardi 1980. Selezione clonale e stabilita genetica di fragole micropopagate. pp. 153-168. In: S.O.I. Incontro Nazionale Sulla Coltura Delia Fragola. Ferrara, Italy. Sass, J.E. 1958. Botanical Microtechnique. Iowa State Univ. Press, Ames. Saunders, J. A., B.F. Matthews, P.D. Miller. 1989. Plant gene transfer using electrofusion and electroporation. pp. 343-354. In: E. Neumann, A.E. Sowers, and C.A. Jordan, (eds.), Electroporation and Electrofusion in Cell Biology. Plenum Publishing, New York. Saunders, J.W. and E.T. Bingham. 1972. Production of alfalfa plants from cells grown in suspension cultures. Plant Sci. Letters 10:59-66. Savage, J.R.K. 1989. The production of chromosome structural changes by radiation. Experientia 45:52-59. Sax, K. 1963. The stimulation of plant growth by ionizing radiation. Radiation Bot. 3:179-186. Schaeffer, G.W. and F.T. Sharpe. 1987. Increased lysine and seed storage protein in rice plants recovered from calli selected with inhibitory levels of lysine plus threonine and S-(2-aminoethyl)-cysteine. Plant Physiol. 84:509-515. Schaeffer, G.W., F.T. Sharpe, Jr., and J.T. Dudley. 1989. Segregation for endosperm lysine in F2, F3 and F4 progeny from a cross of in vitro-selected and unselected cultivar of rice. Theor. Appl Genet. 77:176-183.

PAGE 212

205 Schneider, H. 1968. The anatomy of Citrus, pp. 1-85. In: W. Reuther, L.D. Batchelor, and H.J. Webber (eds.), The Citrus Industry, Vol. 2. Revised Edition. Univ. of California Press, Berkeley. Schroeder, C.A. 1953. Spiral ity in Citrus. Bot. Gaz. 114(3) :350-352. Scora, R.W. 1988. Biochemistry, taxonomy and evolution of modern cultivated citrus, pp. 277-289. In: R. Goren and K. Mendel (eds.), Proc. Sixth Intl. Citrus Congress. Balaban Publishers, Rehovot, Israel. Scora, R.W., and J. Kumamoto. 1983. Chemotaxonomy of the genus Citrus, pp. 343-351. In: P.G. Waterman and M.F. Grundon (eds.), Chemistry and Chemical Taxonomy of the Rutales. Academic Press, London. Scora, R.W., J. Kumamoto, R.K. Soost, and E.M. Nauer. 1982. Contribution to the origin of the grapefruit, Citrus paradisi (Rutaceae). Syst. Bot. 7:170-177. Secor, G.A. and J.F. Shepard. 1981. Variability of protoplastderived potato clones. Crop Sci. 21:102-105. Sekerka, V. 1977. Karyotype of a tissue culture of Allium cepa in vitro. ACTA FAC Rer. Natur. Univ. Comenianae, Physiol. Plant. 13:43-47. Shah, D.M., R.B. Horsch, H.J. Klee, G.M. Kishore, and J. A. Winter. 1986. Engineering herbicide tolerance in transgenic plants. Science 233:478-481. Shahin, E.A. and R. Spivey. 1986. A single dominant gene for Fusarium wilt resistance in protoplast-derived tomato plants. Theor. Appl Genet. 73:164-169. Shamel, A.D. 1932. A pink-fruited lemon. J. Hered. 23:23-27. Shamel, A.D. 1943. Bud variation and bud selection, pp. 915-952. In: H.J. Webber and L.D. Batchelor (eds.), The Citrus Industry, Vol. 1. Univ. of California Press, Berkeley. Shamel, A.D., C.S. Pomeroy, and R.E. Caryl. 1925. Bud selection in the Washington navel orange. Progeny tests of the Thomson, Washington, Corrugated, and Golden Buckeye strains originating as limb variations on a single tree. J. Hered. 16:232-241. Shamel, A.D., L.B. Scott, and C.S. Pomeroy. 1918. Citrus-fruit improvement. A study of bud variation in the Washington navel orange. USDA Bulletin No. 623. Government Printing Office, Washington, D.C.

PAGE 213

206 Shamel, A.D., L.B. Scott, C.S. Pomeroy, and C.L. Dyer. 1920. Citrusfruit improvement: a study of bud variation in the Eureka lemon. USDA Bulletin No. 813. Sharma, D.L., A.K. Gupta, and R.G. Saini. 1989. Induced mutations for leaf rust resistance. Cereal Research Communications 17(1) :31-34. Sharma, N.K. and D.I. Skidmore. 1988. In vitro expression of partial resistance to Phytophthora palmivora by shoot cultures of papaya. Plant Cell, Tissue and Organ Culture 14:187-196. Shepard, J.F. 1982. The regeneration of potato plants from leaf-cell protoplasts. Sci. Am. 246:154-166. Shepard, J.F., D. Bidney, and E. Shahin. 1980. Potato protoplasts in crop improvement. Science 208:17-24. Sheppard, S.C. and B.B. Chubey. 1990. Radiation hormesis of fieldseeded broccoli, parsnip and cauliflower. Can. J. Plant Sci. 70:369-373. Sherman, W.B. and P.M. Lyrene. 1983. Handling seedling populations, pp. 66-73. In: J.N. Moore and J. Janick (eds.), Methods in Fruit Breeding. Purdue Univ. Press, West Lafayette, IN. Shoemaker, N.P., H.J. Swartz, and G.J. Galletta. 1985. Cultivar dependent variation in pathogen resistance due to tissue culture-propagation of strawberries. HortScience 20:253-254. Sidrak, G.H. and A. Suess. 1973. Effects of low doses of gamma radiation on the growth and yield of two varieties of tomato. Radiation Bot. 13:309-314. Sim, G.E., C.S. Loh, and C.J. Goh. 1988. Direct somatic embryogenesis from protoplasts of Citrus mitis Blanco. Plant Cell Reports 7:418-420. Sinclair, W.B. and D.L. Lindgren. 1943. Ridges and sectors induced in the rind of citrus fruits by fumigation with hydrocyanic acid. Plant Physiol. 18:99-106. Singh, B., R. Harvey, K. Kao, and R. Miller. 1972. Selection pressure in cell population of Vicia cultured in vitro. Can. J. Genet. Cytol 14:65-70. Skeene, K. and M. Barlass. 1983. Studies on the fragmented shoot apex of grapevine. J. Exp. Bot. 34:1271-1280. Sloane, H. 1696. Catalogus Plantarum Quae In Insula Jamaica Sponte Proveniunt. Brown, London.

PAGE 214

207 Smith, C.J.S., C.F. Watson, P.C. Morris, C.R. Bird, G.B. Seymour, J.E. Gray, C. Arnold, G.A. Tucker, W. Schuch, S. Harding, and D. Grierson. 1990. Inheritance and effect on ripening of antisense polygalacturonase genes in transgenic tomatoes. Plant Molecular Biology 14:369-380. Smith, G.S., D.J. Hutchison, and C.T. Henderson. 1987. Screening sweet orange citrus cultivars for relative susceptibility to phytophthora foot rot. Proc. Fla. State Hort. Soc. 100:64-66. Snowball, A.M., E.A. Halligan, and M.G. Mull ins. 1988. Studies on juvenility in citrus. Proc. Sixth Intl. Citrus Congress, Tel Aviv, Israel 1:467-473. Soost, R.K. and J.W. Cameron. 1968. Tree and fruit characters of citrus triploid crosses. Hilgardia 31:569-579. Soost, R.K. and J.W. Cameron. 1975. Citrus, pp. 507-540. In: J. Janick and J.N. Moore (eds.), Advances in Fruit Breeding. Purdue Univ. Press, West Lafayette. Soost, R.K. and J.W. Cameron. 1980a. 'Oroblanco,' a triploid pummel o-grapefruit hybrid. HortScience 15:667-669. Soost, R.K. and J.W. Cameron. 1980b. 'Oroblanco,' a new grapefruit hybrid. Calif. Agriculture 34(11, 12) : 16-17. Soost, R.K. and J.W. Cameron. 1985. 'Melogold,' a new pummelograpefruit hybrid. Calif. Agriculture 40(1, 2):30-31. Soost, R.K. and J.W. Cameron. 1986. Melogold: A new grapefruit hybrid for California. Citrograph 71:115. Soost, R.K., J.W. Cameron, W.P. Bitters and R.G. Piatt. 1961. Citrus bud variation Old and new. Calif. Citrograph 46:176, 188, 190-193. Soost, R.K., T.E. Williams, and A.M. Torres. 1980. Identification of nucellar and zygotic seedlings of Citrus with leaf isozymes. HortScience 15:728-729. Soule, J. and W. Grierson. 1986. Maturity and grade standards. pp. 23-48. In: W.F. Wardowski, S. Nagy, and W. Grierson (eds.), Fresh Citrus Fruits. AVI Publishing, Westport, Conn. Sparrow, A.H., R.L. Cuany, J. P. Miksache and L.A. Schainer. 1961. Some factors affecting the responses of plants to acute and chronic radiation exposures. Radiation Bot. 1:10-34. Sparrow, A.H. and G.M. Woodwell. 1962. Prediction of the sensitivity of plants to chronic gamma radiation. Radiation Bot. 2:9-26.

PAGE 215

208 Spiegel-Roy, P. and J. Kochba. 1973a. Mutation breeding in citrus, pp. 91-105. In: Induced Mutations in Vegetatively Propagated Plants. IAEA, Vienna. Spiegel-Roy, P. and J. Kochba. 1973b. Stimulation of differentiation in orange (Citrus sinensis) ovular callus in relation to irradiation of the media. Radiation Bot. 13:97-103. Spiegel-Roy, S., A. Vardi, and A. Elhanati. 1985. Seedless induced mutant in lemon (Citrus limon). Mutation Breeding Newsletter 26:1-2. Stadler, L.J. 1928a. Mutations in barley induced by X-rays and radium. Science 68:186-187. Stadler, L.J. 1928b. Genetic effects of X-rays on maize. Proc. Natl. Acad. Sci. USA 14:69-75. Stebbins, G.L., Jr. 1950. Variation and Evolution in Plants. Columbia University Press, New York. Stafleu, F.A., V. Demoulin, W. Greuter, P. Hiepko, I. A. Linczevski, R. McVaugh, R.D. Meikle, R.C. Rollins, R. Ross, J.M. Schopf, and E.G. Voss. 1978. International Code of Botanical Nomenclature. Bohn, Scheltema and Holkema, Utrecht, Netherlands. Starrantino, A., P. Mannino, and F. Russo. 1988a. The genetic stability of three seedless clonal selections obtained by gamma rays of the seedy 'Monreal' Clementine. Proc. Sixth Int. Citrus Congress, Tel Aviv, Israel 1:175-182. Starrantino, A., F. Russo, B. Donini, and P. Spina. 1988b. Lemon mutants obtained by gamma irradiation of the nucellus cultured in vitro. Proc. Sixth Int. Citrus Congress, Tel Aviv, Israel 1:231-235. Stehle, H., M. Stehle, and L. Quentin. 1937. Flore de la Guadeloupe et Dependances. Tome II: Catalogue des Phanerogames et Fougeres. Imprimerie Catholique, Basse-Terre, Guadeloupe. Stewart, I., G.D. Bridges, A. P. Pieringer, and T.A. Wheaton. 1975. 'Rohde Red Valencia,' an orange selection with improved juice color. Proc. Florida State Hort. Soc. 88:17-19. Stewart, R.N. and H. Dermen. 1970. Somatic genetic analysis of the apical layers of chimeral sports in Chrysanthemum by experimental production of adventitious shoots. Amer. J. Bot. 57:1061-1071. Stoute, E.A. 1982. Grapefruit-forbidden fruit. Advocate Magazine, October 10. Bridgetown, Barbados.

PAGE 216

209 Stubbe, H. 1967. On the relationships between the spontaneous and experimentally induced form diversity and on some experiments on the evolution of cultivated plants, pp. 99-121. In: Induced Mutations and their Utilization. Akademie-Verlag, Berlin. Sun, Z., C. Zhao, K. Zheng, X. Qi, and Y. Fu. 1983. Somaclonal genetics of rice, Oryza sativa. Theor. Appl Genet. 67:67-73. Swanson, CP. 1957. Cytology and Cytogenetics. Prentice-Hall, Englewood Cliffs, NJ. Swartz, H.J., G.J. Galletta, and R.H. Zimmerman. 1981. Field performance and phenotypic stability of tissue culturepropagated strawberries. J. Amer. Soc. Hort. Sci. 106:667-673. Swingle, W.T. 1913. Variation in first-generation hybrids (imperfect dominance): its possible explanation through zygotaxis. Compt. Rend. Conf. Intern. Genet. IV [Paris, 1911] 1913:381-394. Swingle, W.T. 1914. Citrus and Poncirus. In: C.S. Sargent (ed.), Plantae Wilsonianae 2:141-151. Swingle, W.T. 1943. The botany of citrus and its wild relatives of the orange subfamily, pp. 129-474. In: H.J. Webber and L.D. Batchelor (eds.), The Citrus Industry, Vol. 1. University of California Press, Berkeley. Swingle, W.T. and P.C. Reece. 1967. The botany of citrus and its wild relatives, pp. 190-430. In: W. Reuther, H.J. Webber, and L.D. Batchelor (eds.), The Citrus Industry, Vol. 1. Revised Edition. University of California. Taliaferro, CM. S.M. Dabo, E.D. Mitchell, B.B. Johnson, and B.D. Metzinger. 1989. Morphologic, cytogenetic, and enzymatic variation in tissue culture regenerated plants of apomictic oldworld bluestem grasses (Bothriochloa sp.). Plant Cell, Tissue and Organ Culture 19:257-266. Tanaka, T. 1927a. Taxonomy of the citrus fruits of the Pacific region. Mem. Tanaka Citrus Exper. Sta. 1(1): 15-36. Tanaka, T. 1927b. Bizzarria a clear case of periclinal chimera. J. Genetics 18:77-85. Tanaka, T. 1935. History of dispute in the Citrus classification. Stud. Citrol. 7:1-39. Tanaka, T. 1954. Species Problem in Citrus (Revisio aurantiacearum, IX). Jap. Soc. Prom. Sci., Ueno, Tokyo.

PAGE 217

210 Tanaka, T. 1961. Citrologia. Semi-centennial commemoration papers on citrus studies. Citrologia Supporting Foundation, Osaka, Japan. Tanaka, T. 1977. Fundamental discussion of Citrus classification. Studia Citrologica 14:1-6. Tanaka, Y. 1948. An Iconograph of Japanese Citrus Fruits: A Monographic Study of Species and Varieties of Citrus Fruits Grown in Japan. 2 Vols. Yokendo Co. Ltd., Tokyo. Tedford, E.C., T.L. Miller, and M.T. Nielsen. 1990. A detached-leaf technique for detecting resistance to Phytophthora parasitica var. nicotianae in tobacco. Plant Dis. 74:313-316. Thanutong, P., I. Furusawa, M. Yamamoto. 1983. Resistant tobacco plants from protoplast-derived calluses selected for their resistance to Pseudomonas and Alternaria toxins. Theor. Appl Genet. 66:209-215. Tilney-Bassett, R.A.E. 1986. Plant Chimeras. E. Arnold, London. Timmer, L.W. 1977. Preventive and curative trunk treatments for control of Phytophthora foot rot of citrus. Phytopathology 67:1149-1154. Timmer, L.W. 1985. Effectiveness of metal axyl and fosetyl Al against Phytophthora parasitica on sweet orange. Plant Dis. 69:741-743. Timmer, L.W. 1988. Damping-off. pp. 13-14. In: J.O. Whiteside, S.M. Garnsey, and L.W. Timmer (eds.), Compendium of Citrus Diseases. APS Press, St. Paul, MN. Timmer, L.W. and J. A. Menge. 1988. Phytophthora-induced diseases, pp. 22-24. In: J.O. Whiteside, S.M. Garnsey, and L.W. Timmer (eds.), Compendium of Citrus Diseases. APS Press, St. Paul. Ting, S.V. and R.L. Rouseff. 1986. Citrus Fruits and Their Products: Analysis and Technology. Marcel Dekker, New York. Tolkowsky, S. 1938. Hesperides. A History of the Culture and Use of Citrus Fruits. John Bale, Sons & Curnow, London. Tolley, I. 1989. Another view of the "new" navel. Citrograph 74:98. Tomes, D.T., A.K. Weissinger, M. Ross, R. Higgins, B.J. Drummond, S. Schaaf, J. Malone-Schoneberg, M. Staebell, P. Flynn, J. Anderson, and J. Howard. 1990. Transgenic tobacco plants and their progeny derived by microprojectile bombardment of tobacco leaves. Plant Molecular Biology 14:261-268.

PAGE 218

211 Torres, A.M., T. Mau-Lastovicka, T.E. Williams, and R.K. Soost. 1985. Segregation distortion and linkage of Citrus and Poncirus isozyme genes. J. Hered. 76:289-294. Torres, A.M., R.K. Soost, and U. Diedenhofen. 1978. Leaf isozymes as genetic markers in citrus. Amer. J. Bot. 65:869-881. Torres, J. P. 1936. Polyembryony in citrus and study of hybrid seedlings. Philipp. J. Agr. 7:37-58. Toxopeus, H.J. 1933. Some cases of bud variation in Citrus observed on Java. Genetica 15:241-252. Toyoda, H., K. Shimizu, K. Chatani, N. Kita, Y. Matsuda, and S. Ouchi. 1989. Selection of bacterial wilt-resistant tomato through tissue culture. Plant Cell Reports 8:317-320. Tsao, P.H. and M.J. Garber. 1960. Methods of soil infestation, watering and assessing the degree of root infection for greenhouse in situ ecological studies with citrus phytophthoras. Plant Dis. Reporter 44:710-715. Tusa, N., F. De Pasquale, and L. Radogna. 1978. Research on the micro-grafting technology in citrus. Proc. Int. Soc. Citriculture, Sidney, Australia. 1978:143-145. Tusa, N. and G. Geraci 1988. Preliminary observations on correlation between In vivo and In vitro resistance of sour orange (Citrus aurantium L.) to Phytophthora gummosis. Proc. Sixth Int. Citrus Congress, Tel Aviv, Israel 2:775-778. Tussac, L.C.F.R.de. 1824. Flore des Antilles, Vol. 3. L'auteur, Paris. Uggla, A.H. and A.T. Natarajan. 1982. SCE in plants, pp. 217-233. In: A. A. Sandberg (ed.), Sister Chromatid Exchanges. Alan R. Liss, New York. Umbeck, P.F. and B.G. Gengenbach. 1983. Reversion of male-sterile T-cytoplasm maize to male fertility in tissue culture. Crop Sci. 23:584-588. United Nations. 1986. United Nations Scientific Committee on the Effects of Atomic Radiation. Genetic and Somatic Effects of Ionizing Radiation. 1986 Report to the General Assembly, with annexes. United Nations, New York. United States Department of Agriculture. 1989. Fruit and Tree Nuts Situation Outlook. (TFS-249) Econ. Res. Ser., Rockville, MD.

PAGE 219

212 Utkhede, R.S. 1986. In vitro screening of the world apple germplasm collection for resistance to Phytophthora cactorum crown rot. Scientia Horticulture 29:205-210. Utkhede, R.S. and H.A. Quamme. 1988. Use of the excised shoot assay to evaluate resistance to Phytophthora cactorum of apple rootstock cultivars. Can. J. Plant Sci. 68:851-857. Vaarama, A. 1949. Spindle abnormalities and variation in chromosome number in Ribes nigrum. Hereditas 35:136-162. Vaeck, M. A. Reynaerts, H. Hofte, S. Jansens, and M. De Beuckeleer. 1987. Transgenic plants protected from insect attack. Nature 328:33-37. Vallejos, C. 1983. Enzyme activity staining, pp. 469-516. In: S.D. Tanksley and T.J. Orton (eds.), Isozymes in Plant Genetics and Breeding, Part A. Elsevier, Amsterdam. Van Harten, A.M., H. Bouter, and C. Broertjes. 1981. In vitro adventitious bud techniques for vegetative propagation and mutation breeding of potato (SoTanum tuberosum cultivar Desiree): 2. Significance for mutation breeding. Euphytica 30:1-8. Vardi, A. 1977. Isolation of protoplast in citrus. Proc. Int. Soc. Citriculture, Orlando 2:575-578. Vardi, A., P. Arzee-Gonen, A. Frydman-Shani S. Bleichman, and E. Galun. 1989. Protoplast-fusion-mediated transfer of organelles from Microcitrus into Citrus and regeneration of novel alloplasmic trees. Theor. Appl Genet. 78:741-747. Vardi, A. and E. Galun. 1988. Recent advances in protoplast culture of horticultural crops: Citrus. Scientia Horticulturae 37:217-230. Vardi, A., A. Breiman, and E. Galun. 1987. Citrus cybrids: Production by donor-recipient protoplast-fusion and verification by mitochondrial -DNA restriction profiles. Theor. Appl. Genet. 75:51-58. Vardi, A., E. Epstein, and A. Breiman. 1986a. Is the Phytophthora citrophthora culture filtrate a reliable tool for the in vitro selection of resistant Citrus variants? Theor. Appl. Genet. 72:569-574. Vardi, A., D.J. Hutchison, and E. Galun. 1986b. A protoplast to tree system in Microcitrus based on protoplasts derived from a sustained embryogenic callus. Plant Cell Reports 5:412-414.

PAGE 220

213 Vardi, A., P. Spiegel-Roy, and E. Galun. 1982. Plant regeneration from Citrus protoplasts: Variability in methodological requirements among cultivars and species. Theor. Appl Genet. 62:171-176. Varghese, Y.A. and G. Robbelen. 1984. Sectorial chimeras as a proof of the de novo origin of mutants showing partial resistance to powdery mildew in barley. Z. Pflanzenzuchtg. 92:265-280. Vasil, I.K. 1987. Developing cell and tissue culture systems for the improvement of cereal and grass crops. J. Plant Physiol. 128:193-218. Vasil, I.K. 1988. Progress in the regeneration and genetic manipulation of cereal crops. Bio/Technology 6:397-402. Vaughn, K.C. 1983. Chimeras and variegation: Problems in propagation. HortScience 18:17-20. Vaughn, K.C. and K.G. Wilson. 1980. A dominant plastome mutation in Hosta. J. Hered. 71:203-206. Vavilov, N.I. 1951. The Origin, Variation, Immunity and Breeding of Cultivated Plants. Chronica Botanica, Waltham, MA. Vernenghi, A., F. Ramiandrasoa, S. Chuilon, et A. Ravise. 1987. Phytoalexines des Citrus: seseline proprietes inhibitrices et modulation de synthese. Fruits 42:103-111. Volkamer, J.C. 1708-1714. Nurnbergische Hesperides, oder Grundliche Beschreibung der edlen CitronatCitronenund PomeranzenFruchte. 2 Vols. Bei dem Authore, Nurnberg. Waibel, C. 1953. Varieties and strains of citrus originating in the lower Rio Grande Valley of Texas. Rio Grande Val Hort. Inst. Proc. 7:18-24. Wakasa, K. and J.M. Widholm. 1987. A 5-methyl tryptophan resistant rice mutant, MTR1, selected in tissue culture. Theor. Appl. Genet. 74:49-54. Wang, A.S., D.S.K. Cheng, J.B. Milcic, and T.C. Yang. 1988. Effect of X-ray irradiation on maize inbred line B73 tissue cultures and regenerated plants. Crop Sci. 28:358-362. Wardell, W.L. and F. Skoog. 1973. Flower formation in excised stem segments. III. Deoxyribosenucleic acid content in stem tissue of vegetative and flowering tobacco plants. Plant Physiol. 52:215-220.

PAGE 221

214 Warren, R.S. and D.G. Routley. 1970. The use of tissue culture in the study of single gene resistance of tomato to Phytophthora infestans. J. Amer. Soc. Hort. Sci. 95:266-269. Watanabe, H., H. Yamagata, and K. Syakudo. 1970. [Studies on the citrus generic polyembryony in relation to breeding: 2. Discrimination of the embryo fertilized by H-labeled pollen grains] (in Japanese). Jap. J. Breeding 20:141-145. Watson, J.C., L.S. Kaufman, and W.F. Thompson. 1987. Developmental regulation of cytosine methylation in the nuclear ribosomal RNA genes of Pi sum sativum. J. Mol Biol. 193:15-26. Webber, H.J. 1943. Cultivated varieties of citrus, pp. 475-668. In: H.J. Webber and L.D. Batchelor (eds.), The Citrus Industry, Vol. 1. University of California Press, Berkeley. Weinberg, G., B. Ullman, and D.W. Martin. 1981. Mutator phenotypes in mammalian cell mutants with distinct biochemical defects and abnormal deoxyribonucleoside triphosphate pools. Proc. Natl. Acad. Sci. USA 78:2447-2451. Weising, K., J. Schell, and G. Kahl 1988. Foreign genes in plants: Transfer, structure, expression, and applications. Annu. Rev. Genet. 22:421-477. Wenzel, G., B. Foroughi-Wehr, W. Friedt, F. Kohler, and T. Oo. 1985. Cell and tissue culture as supplementary tool in plant breeding: Exemplified in potato, oilseed rape and barley. Hereditas Suppl 3:15-25. Wenzel, G., 0. Scheider, T. Przewozny, S.K. Sopory, and G. Melchers. 1979. Comparison of single cell culture derived Solarium tuberosum L. plants and model for their application in breeding programs. Theor. Appl Genet. 55:49-55. Werry, P. S.T.J. 1981. Induction by ionizing radiation of genetic markers for the development of in vitro genetic manipulation as a tool in crop plant improvement, pp. 373-382. In: Induced Mutations A Tool in Plant Research. IAEA, Vienna. Wersuhn, G. 1988. Selecting in higher plant cell cultures. A discussion. Biochem. Physiol. Plants 183:361-369. Wersuhn, G. 1989. Obtaining mutants from cell cultures. Plant Breeding 102:1-9. Wheeler, J.E. and A.M. Boyle. 1971. Identification of four Phytophthora isolates previously unreported from Arizona. Phytopathology 61:1293-1296.

PAGE 222

215 Whiteside, J.O. 1971. Some factors affecting the occurrence and development of foot rot on citrus trees. Phytopathology 61:1233-1238. Whiteside, J.O. 1974. Zoospore-inoculation techniques for determining the relative susceptibility of citrus rootstocks to foot rot. Plant Dis. Reporter 58:713-717. Wierbicki, E., E.S. Josephson, A. Brynjolfsson, J.M. Coon, R.B. Maxcy, C. Merritt, Jr., I. A. Taub, M. Doyle, G.W. Froning, J.L. Goodenough, G.D. Griffin, A. A. Kader, W.J. Lipton, A. Mcintosh, S.O. Nelson, C.S. Perry, H.J. Reitz, M.G. Simic, J.S. Sivinski, J.W. Slavin, D.W. Smith, E.W. Tilton, and W.M. Urbain. 1986. Ionizing Energy in Food Processing and Pest Control: I. Wholesomeness of food treated with ionizing energy. Council for Agricultural Science and Technology Report No. 109. Ames, Iowa. Willmot, D.B., CD. Nickel 1 J.M. Widholm, and L.E. Gray. 1989. Evaluation of soybean resistance to Phialophora gregata culture filtrate in tissue culture. Theor. Appl Genet. 77:227-232. Wilson, H.D. 1990. Gene flow in squash species. Bioscience 40:449-455. Winkler, H. 1907. Ueber Pfropfbastarde und pflanzliche chimaren. Ber. d. deutsch. Bot. Gesellsch. 25:568-576. Wolff, S. 1989. Are radiation-induced effects hormetic? Science 245:575, 621. Wright, S. 1931. Evolution in Mendelian populations. Genetics 16:97-159. Yen, C.R. 1987. Assimilate Partitioning and Enzymes of Organic Acid Metabolism in Fruit of Calamondin and Low Acid Grapefruit. PhD Dissertation: University of Florida, Gainesville. Yu, S. and A.F. Yeager. 1960. Ten heritable mutations found in the tomato following irradiation with X-rays and thermal neutrons. Proc. Amer. Soc. Hort. Sci. 76:538-542. Ziegler, L.W. and H.S. Wolfe. 1961. Citrus Growing in Florida. University of Florida Press, Gainesville. Zilberstein, M. and Y. Pinkas. 1987. Detached root inoculation A new method to evaluate resistance to Phytophthora root rot in avocado trees. Phytopathology 77:841-844. Zubrzycki, H.M. and A. Diamante de Zubrzycki 1982. Induction of mutations in citrus for the development of resistance to Xanthomonas citri (Hasse) Dowson. pp. 91-110. In: Induced mutations in Vegetatively Propagated Plants II. IAEA, Vienna.

PAGE 223

BIOGRAPHICAL SKETCH Kim Dean Bowman was born May 2 nd 1957, in Dayton, Ohio, and completed his primary and secondary education in the nearby Brookville school system. He graduated from Brookville High School in 1975 as salutatorian and a member of the National Honor Society. After two years' employment in landscaping and factories, he enrolled in Miami University, Oxford, Ohio. He received the Bachelor of Science degree from Miami University in 1981 summa cum laude with honors in botany. Kim Dean Bowman was awarded a Liberty Hyde Bailey Fellowship from Cornell University in 1981, where he began his graduate studies in the Department of Pomology. During his studies at Cornell University, he undertook research on breeding for resistance to the woolly apple aphid, completed the 1982 Athens Marathon in 2:54:17, and received the Master of Science degree. He obtained his most significant instruction in the ways of science during this period from James N. Cummins and John Sanford. In 1985, Kim Dean Bowman began two years of service as a Peace Corps volunteer in the country of Saint Lucia. He was employed as a crop extension specialist, lecturer, and coordinator of agricultural services for the Ministry of Agriculture in that beautiful country. While in Saint Lucia, he met and married the woman of his dreams, Concessa Jean Marie. Kim Dean Bowman entered the PhD program in horticultural science at University of Florida in 1987. His principal advisor during this period 216

PAGE 224

217 was Frederick G. Gmitter, Jr., the plant breeder at the Citrus Research and Education Center in Lake Alfred. He received the PhD degree from University of Florida in 1990 and subsequently accepted the post of senior museum scientist with the Citrus Clonal Protection Program and University of California, Riverside.

PAGE 225

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Gloria A. Moore, Chair Associate Professor of Horticultural Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Frederick G. Gmitter, Assistant Professor of Horticultural Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. $dde W. Grosser Associate Professor of Horticultural Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Paul M. Lyrene ^ Professor of Horticultural Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. |es H. Graham, Jr loci ate Professor of Soil

PAGE 226

This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate School and was accepted as partial fulfillment of the requirements for^the degree of Doctor of Philosophy. December 1990 Dean, Graduate School