In vitro somatic embryogenesis and plant regeneration from mango (Mangifero indica L.) nucellar callus


Material Information

In vitro somatic embryogenesis and plant regeneration from mango (Mangifero indica L.) nucellar callus
Mangifera indica
Physical Description:
xvi, 162 leaves : ill. ; 28 cm.
DeWald, Stephen Gregory, 1950-
Publication Date:


Subjects / Keywords:
Mango -- Propagation -- In vitro   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1987.
Includes bibliographical references (leaves 148-161).
Statement of Responsibility:
by Stephen Gregory DeWald.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000950584
notis - AER2772
oclc - 16931239
System ID:

Full Text







This dissertation is dedicated to my cherished friend

and loving wife Maria, who has served as a constant source

of inspiration and support.


The author expresses his sincere thanks and gratitude

to the members of his advisory committee, Drs. Gloria A.

Moore, Wayne B. Sherman, Robert J. Knight, Jr., Prem S.

Chourey, and especially to his major advisor Dr. Richard E.


Many thanks go to his parents Stephen and Ruth Anne,

brother Joseph, and the rest of the DeWald family for their

support and encouragement.

Finally, the author extends his appreciation to the

Florida Mango Forum for their financial support and to the

graduate students, the service staff, and the faculty of the

University of Florida Fruit Crops Department, Gainesville,

and the Tropical Research and Education Center, Homestead,

for the many ways in which they have helped.



ACKNOWLEDGEMENTS ..........................................iii

LIST OF TABLES............................................. viii

LIST OF FIGURES...............................................xi

KEY TO ABBREVIATIONS...................................xiii

ABSTRACT......................................... .........xiv


I INTRODUCTION........................... ........

II LITERATURE REVIEW..............................3

Introduction........................ .......... 3
Mango: Taxonomy, Distribution, and
Breeding Efforts............................. 3
Polyembryony .... ............................... 9
Histological Studies of Mango Embryology...... 12
Factors Affecting Polyembryony................15
The Genetics of Polyembryony...................17
The Evolutionary Significance of Apomixis.....20
Plant Tissue and Cell Culture.................21
Somatic Embryogenesis..........................24
Factors Affecting In Vitro
Somatic Embryogenesis.......................28
Tissue Culture of Mango.......................37

IN MANGO. ......................................41

Introduction.................................. 41
Materials and Methods .........................42
Results and Discussion........................43

EMBRYOGENESIS IN MANGO........................72

Introduction................................... 72
Materials and Methods .........................73


Experiment 4-1: In Vitro Somatic
Embryogenesis from Mango Nucellar
Callus in Response to Genotype and
Plant Growth Regulators...................77
Experiment 4-2: In Vitro Somatic
Embryogenesis from 'Parris' Nucellar
Callus in Response to the Callus
Maintenance Medium and Major Salts
Experiment 4-3: In Vitro Somatic
Embryogenesis from 'Parris' Nucellar
Callus in Response to the Callus
Maintenance Medium Formulation
and Sucrose Concentration..................79
Experiment 4-4: Callus Production from
Nonembryogenic 'James Saigon'
Nucellar Callus in Response to the
Major Salts Formulation of the
Callus Maintenance Medium................. 80
Experiment 4-5: In Vitro Somatic
Embryogenesis from 'James Saigon'
Nucellar Callus in Response to the
Solidifying Agent Used in the
Embryogenesis Medium.......................81
Experiment 4-6: In Vitro Production
of 'James Saigon' Nucellar Callus in
Response to a Liquid or a Solid
Maintenance Medium........................82
Results and Discussion........................83
Experiment 4-1: In Vitro Somatic
Embryogenesis from Mango Nucellar
Callus in Response to Genotype and
Plant Growth Regulators................... 83
Experiment 4-2: In Vitro Somatic
Embryogenesis from 'Parris' Nucellar
Callus in Response to the Callus
Maintenance Medium and Major Salts
Formulations ....... ............. ........ 92
Experiment 4-3: In Vitro Somatic
Embryogenesis from 'Parris' Nucellar
Callus in Response to the Callus
Maintenance Medium Formulation
and Sucrose Concentration..................96
Experiment 4-4: Callus Production from
Nonembryogenic 'James Saigon'
Nucellar Callus in Response to
the Major Salts Formulation of the
Callus Maintenance Medium.................98


Experiment 4-5: In Vitro Somatic
Embryogenesis from 'James Saigon'
Nucellar Callus in Response to the
Solidifying Agent Used in the
Embryogenesis Medium......................98
Experiment 4-6: In Vitro Production
of 'James Saigon' Nucellar Callus in
Response to a Liquid or a Solid
Maintenance Medium.......................100
Conclusions .................................. 102

AND PLANTLET FORMATION.......................105

Materials and Methods ........................106
Experiment 5-1: In Vitro Somatic
Embryo Production and Maturation
from 'James Saigon' Nucellar Callus
in Response to Culture Regime............106
Experiment 5-2: In Vitro 'Parris'
Somatic Embryo Maturation in Response
to Sucrose Concentration and ABA in
a Liquid Embryo Maturation Medium......... 107
Experiment 5-3: In Vitro 'Parris'
Somatic Embryo Maturation in Response
to Sucrose Concentration and
Supplements to the Embryo Maturation
Experiment 5-4: In Vitro 'Parris'
Somatic Embryo Maturation in Response
to Solidifying Agent and ABA in the
Embryo Maturation Medium.................111
Experiment 5-5: In Vitro 'Parris'
Somatic Embryo Germination and Shoot
Formation in Response to the Embryo
Germination Medium Formulation...........113
Experiment 5-6: In Vitro 'Parris'
Somatic Embryo Germination and Shoot

Formation in Response to the Embryo
Germination Medium Formulation...........114
Results and Discussion....................... 116
Experiment 5-1: In Vitro Somatic
Embryo Production and Maturation
from 'James Saigon' Nucellar Callus
in Response to Culture Regime.............116
Experiment 5-2: In Vitro 'Parris'
Somatic Embryo Maturation in Response
to Sucrose Concentration and ABA in
a Liquid Embryo Maturation Medium.........118


Experiment 5-3: In Vitro 'Parris'
Somatic Embryo Maturation in Response
to Sucrose Concentration and
Supplements to the Embryo Maturation
Medium .............. .................... 120
Experiment 5-4: In Vitro 'Parris'
Somatic Embryo Maturation in Response
to Solidifying Agent and ABA in the
Embryo Maturation Medium.................122
Experiment 5-5: In Vitro 'Parris'
Somatic Embryo Germination and Shoot
Formation in Response to the Embryo
Germination Medium Formulation........... 124
Experiment 5-6: In Vitro 'Parris'
Somatic Embryo Germination and Shoot
Formation in Response to the Embryo
Germination Medium Formulation...........126
Conclusions .................................128

VI SUMMARY AND CONCLUSIONS......................134

Histological Investigations..................134
Tissue Culture Studies.......................136
Conclusions ....................................138

APPENDIX................................... ................... 140

LITERATURE CITED..........................................148

BIOGRAPHICAL SKETCH .....................................162


Table Page

4-1 Basal media major salts formulations used with
in vitro culture of mango...........................74

4-2 Murashige and Skoog minor salts formulation........ 74

4-3 Basal medium components and supplements used for
in vitro culture of mango..........................76

4-4 In vitro somatic embryogenesis from
'James Saigon' nucellar callus in response to
various plant growth regulators.....................84

4-5 In vitro somatic embryogenesis from 'Parris'
nucellar callus in response to various plant
growth regulators.................................85

4-6 In vitro somatic embryogenesis from
'Tommy Atkins' nucellar callus in response to
various plant growth regulators...................86

4-7 In vitro somatic embryogenesis from 'Heart'
nucellar callus in response to various plant
growth regulators.................................87

4-8 In vitro somatic embryogenesis from mango
nucellar callus for cultivars in experiment 1......88

4-9 In vitro somatic embryogenesis in response
to various complex organic addenda................. 88

4-10 In vitro somatic embryogenesis from 'Parris'
nucellar callus in response to the callus
maintenance medium and major salts formulations....93

4-11 In vitro somatic embryogenesis from 'Parris'
nucellar callus in response to the callus
maintenance medium formulation and sucrose

4-12 Callus production from nonembryogenic
'James Saigon' nucellar callus in response to
the major salts formulation of the callus
maintenance medium.................................99


Table Page

4-13 In vitro somatic embryogenesis from
'James Saigon' nucellar callus in response to
the gelling agent used in the embryogenesis
medium............................................. 99

4-14 In vitro production of 'James Saigon' nucellar
callus in response to liquid or solid callus
maintenance medium................................101

5-1 In vitro somatic embryo maturation rating
scale used in experiment 5-3......................110

5-2 In vitro somatic embryo maturation rating
scale used in experiment 5-4...................... 112

5-3 In vitro somatic embryo production
and maturation from 'James Saigon' nucellar
callus in response to the culture regime..........117

5-4 In vitro 'Parris' somatic embryo maturation
in response to sucrose concentration and ABA
in liquid embryo maturation medium................119

5-5 In vitro 'Parris' somatic embryo maturation
in response to sucrose concentration and
supplements to the embryo maturation medium........121

5-6 In vitro 'Parris' somatic embryo maturation
in response to the solidifying agent and ABA in
the embryo maturation medium.......................123

5-7 In vitro 'Parris' somatic embryo germination
and shoot formation in response to the embryo
germination medium formulation....................125

5-8 In vitro 'Parris' somatic embryo germination
and shoot formation in response to the embryo
germination medium formulation....................125

A-i Plant growth regulators (PGRs) used with
in vitro culture of mango.........................140

A-2 List of mango cultivars with successful
in vitro nucellar callus initiation...............141

A-3 In vitro somatic embryogenesis from 'Simmonds'
nucellar callus in response to various plant
growth regulators................................ 142

Table Page

A-4 In vitro somatic embryogenesis from 'Florigon'
nucellar callus in response to various plant
growth regulators................................143

A-5 In vitro somatic embryogenesis from
callus of 'Cambodiana' in response to
various plant growth regulators................... 144

A-6 In vitro somatic embryogenesis from
callus of 'Irwin' in response to various plant
growth regulators................................145

A-7 Peach embryo germination medium................... 146


Figure Page

3-1 A mango fruitlet approximately 75 days-post-
bloom bisected longitudinally with intact
ovule (megasporangium) ............................49

3-2 Longitudinal section through a polyembryonic
ovule approximately 30 days-post-bloom.............50

3-3 Micropylar region of a polyembryonic ovule.........51

3-4 Longitudinal section through the micropylar
half of a polyembryonic ovule approximately
45 days-post-bloom................................52

3-5 Globular stage adventive nucellar embryos..........53

3-6 Late globular stage adventive nucellar embryo......54

3-7 The initiation of embryogenic nucellar callus......55

3-8 Close-up view of embryogenic mango nucellar
callus initiation................................. 56

3-9 Suspension culture of rapidly proliferating
mango nucellar callus....................... ...... 57

3-10 The epidermis of a late globular stage somatic
embryo prior to redifferentiation and somatic

3-11 The epidermis of a late globular stage somatic
embryo beginning redifferentiate.................... 59

3-12 Early stages of epidermal somatic embryogenesis....60

3-13 Section through the periphery of a redif-
ferentiating globular somatic embryo or callus.....61

3-14 Four early globular stage somatic embryos..........62

3-15 A globular stage somatic embryo beginning to
differentiate an epidermis and showing signs
of polarity.................................. .......63

Figure Page

3-16 A suspension culture of budding 'Parris'
somatic embryos.....................................64

3-17 Close-up of a cluster of budding 'Parris'
somatic embryos grown in suspension culture........65

3-18 Cross section, at low magnification through a
cluster of budding somatic embryos grown in
suspension culture.................................66

3-19 A section through the periphery of the
central core of a budding cluster of somatic
embryos grown in suspension culture.................67

3-20 Globular stage somatic embryo grown in
suspension culture exhibiting aberrant
development ........................................68

3-21 Globular stage somatic embryos grown in
suspension culture with a poorly formed

3-22 Scanning electron micrograph of a cluster
of 'Parris' somatic embryos grown on solid

3-23 Longitudinal section through the base
of a somatic embryo grown on solid medium..........71

4-1 'Parris' somatic embryogenesis in response to
major salts formulations of experiment 4-2........104

5-1 'Parris' somatic embryo maturation in
response to solidifying agents and ABA of
experiment 5-4.....................................130

5-2 A culture of germinating 'Parris' somatic
embryos from experiment 5-6.......................131

5-3 Germinated 'Parris' somatic embryos ready
to be transferred to the soil.....................132

5-4 Young 'Parris' plants regenerated via
somatic embryogenesis growing in the soil.........133


ABA: abscisic acid

BA: 6-benzylaminopurine (N6 benzyladenine)

BM: basal medium

B-5 Gamborg et al., B-5 basal medium

CH: casein hydrolysate

GA: gibberellic acid

IAA: indole-3-acetic acid

IBA: indole-3-butyric acid

LCE: liquid coconut endosperm (coconut water)

M: monoembryonic seed

MS: Murashige and Skoog basal medium formulation

NAA: 2-naphthaleneacetic acid

Na2EDTA: disodium ethylenediaminetetraacetic acid

P: polyembryonic seed

YE: yeast extract

WPM: Lloyd and McCown woody plant medium

2iP: (2-isopentenyl)adenine

2,4-D: (2,4-dichlorophenoxy)acetic acid


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




May, 1987

Chairman: Richard E. Litz
Major Department: Horticultural Science (Fruit Crops)

Histological studies were conducted on polyembryonic

mango ovules and embryogenic nucellar callus to elucidate

the patterns of embryogenesis in these two systems. The

nucellar origin as well as the variable location of

adventive embryos within the micropylar portion of the

ovules was demonstrated. Adventive nucellar embryos,

characterized by densely staining cytoplasm and prominent

nuclei, were formed prior to the cellularization of the


Embryogenic callus was initiated from both the

micropylar and chalazal portions of excised nucelli. In

suspension culture this callus grew as compact, nonfriable

globular callus. Histological preparations of this callus

revealed that no true, undifferentiated, subculturable

tissue is produced. Proliferation occurs via secondary

budding from the epidermis of preformed somatic embryos.

Prior to somatic embryogenesis the epidermis

dedifferentiates and forms a layer, several cells thick, of

cytoplasmically dense cells with prominent nuclei. In

suspension culture, somatic embryos exhibited numerous

developmental abnormalities including poorly formed

epidermal layers, vascular system and meristems, noncompact

morphology, and secondary budding. More normal embryo

development occurred on solid medium. Somatic embryos were

shown to possess bipolar meristems with closed vascular

systems, thus unequivocally demonstrating the embryogenic

nature of plant regeneration from mango nucellar callus.

The morphogenic response of mango nucellar callus was

also investigated with emphasis on factors influencing

callus production, somatic embryogenesis, embryo maturation,

and germination. The addition of 2 mM glutamine, 5 uM 2,4-D,

5 uM kinetin, and vitamins to suspension multiplication

cultures improved the embryogenic efficiency of callus upon

transfer to a solid embryogenesis medium. Efficient somatic

embryogenesis occurred on a modified B-5 medium [lacking

(NH4)2SO4] supplemented with 5% (w/v) sucrose, 2 mM

glutamine, 11 uM NAA, 5 uM 2iP, vitamins, and amino acids.

The use of Gelrite gellan gum (0.19%, w/v) improved the

efficiency of somatic embryogenesis and maturation in

comparison to medium solidified with Difco Bacto-agar or

Sigma agar gum. Maturation and germination occurred on solid

media in which the sucrose concentration was sequentially

lowered and the synthetic plant growth regulators were

replaced with 20% (v/v) liquid coconut endosperm and 0.025%

(w/v) casine hydrolysate. Plant's were established in the

soil by watering with dilute basal salts.


Mango is one of the world's most important tropical

fruit crops. At the present time, production in Florida is

restricted to approximately 1000 ha of mature mango groves

(IFAS, 1983). However, despite its relatively small size in

Florida's fruit industry, mango production, both here and

abroad, should continue to expand (Bondad, 1980). One of the

major factors limiting the distribution and production of

this delicious tropical fruit crop is the lack of well

adapted, high yielding, disease resistant cultivars with

good quality fruit. Most of the industry throughout the

world is based on the use of chance seedlings maintained

clonally by means of various asexual propagation techniques.

Breeding attempts have been particularly discouraging in

comparison to many other agronomic crops.

Recent advances in biotechnology offer the plant

breeder new and exciting opportunities in cultivar

improvement and development. One of these new

biotechnologies is the field of plant cell and tissue

culture. Plant cell and tissue culture has been used

successfully to propagate large numbers of plants, eliminate

persistent pathogens, create new and economically useful

variation in existing well established genotypes, rescue

hybrid embryos from wide crosses, create homozygous breeding

lines via haploidy, produce economically important primary

and secondary plant metabolites, and elucidate complex

processes at both the cell and whole plant level. Plant

tissue culture and its associated technologies should be

particularly useful to breeders working with perennial

crops, where cultivar improvement has been slow and

difficult. Unfortunately, the woody perennials as a whole

have proved to be recalcitrant in vitro.

In vitro somatic embryogenesis from mango nucellar

cultures was first reported by Litz et al. (1982); however,

the recovery of plants from mango somatic embryos has not

been reported. The first objective of this research was to

examine histologically the processes of in vitro somatic

embryogenesis and adventive nucellar polyembryony in order

to determine more clearly the regeneration pathway in mango.

The second objective of this research was to investigate and

optimize in vitro parameters that affect somatic

embryogenesis, embryo development, and germination, thereby

developing an efficient in vitro plant regeneration system.



The purpose of this literature review is to introduce

the reader to the mango, its embryology, and in vitro

somatic embryogenesis. Polyembryony and somatic

embryogenesis are of particular importance to this

dissertation and will be discussed in detail. Somatic

embryogenesis in Citrus, is similar to that in mango. Citrus

In vitro culture is perhaps the best understood system for

somatic embryogenesis in woody plant species, and numerous

articles have been published since embryogenic cultures were

established by Stevenson in 1956. For these reasons the

literature pertinent to Citrus is emphasized throughout this


Mango: Taxonomy, Distribution, and Breeding Efforts

Mango is one of the choicest and most popular of the

tropical fruits. With an annual world production of nearly

14 million metric tons, mango is the fifth most important

fruit crop after grape, Musa, Citrus, and apples (FAO

Production Yearbook, 1984).

The culture of mango is most intensive in India, which

accounts for 2/3 of the total world production. Singh (1960)

reports that it has been under cultivation there for almost

6,000 years. Mangos are particularly well suited to India

where they .thrive in almost every region except altitudes

above 3,000 feet (Singh, 1960). In India, the mango is

referred to as the "king of fruits" and has played an

integral part in India's folklore and religions.

Mango, Mangifera indica L., is a member of the

Anacardiaceae, a large family of mostly tropical woody

perennial plants with inconspicuous flowers often produced

in large clusters, frequently bearing attractive and edible

fruit (Purseglove, 1968). The family contains some 64 genera

and includes several economically important fruit and nut

trees, e.g., cashew (Anacardium occidentale L.), a flesh

fruit (Bouea macrophylla Griff.), pistachio (Pistacia vera

L.), several edible spondias (Spondias cythera Sonn., S.

mombin L., S. purpurea L.), and some important plant pests,

e.g., poison ivy (Rhus toxicodendron L.) and the Brazilian

pepper tree (Schinus terebinthifolius Radd.) (Popenoe, 1920;

Purseglove, 1968).

The genus Mangifera contains some 41-62 invariably

arborescent species (Mukherjee, 1972; Singh, 1969)

characterized by long, leathery leaves with a

fiberous-resinous fruit. The trees are evergreen with

several distinctive annual growth cycles or flushes. The

brownish-red to red vegetative flushes are unique in that


they occur in sectors rather than all over the tree during

any one cycle (Popenoe, 1920).

Mango flowers are small and generally borne on terminal

panicles. The inflorescence is polygamous producing up to

4,000 staminate and perfect flowers per panicle. The fruit

is a large, fleshy drupe consisting of an edible mesocarp

and a fiberous, stony endocarp. Mango is often grouped into

2 major categories based on embryology: monoembryonic types

contain only a single embryo per seed, while the

polyembryonic types produce multiple (usually asexual)

embryos per seed. Almost without exception the Indian mangos

are monoembryonic, while the Indochinese and Philippines

types are polyembryonic (Singh, 1976). Young and Sauls

(1979) state that many monoembryonic Indian types produce

brightly colored fruit with an attractive blush and are

often susceptible to anthracnose, [Glomerella cingulata

Stonem (S. and V.S.)]. Polyembryonic types more commonly

produce a pale green to yellow, low fibered fruit and

generally possess higher levels of resistance to


Mukherjee (1950, 1972) determined that the basic

chromosome number in mango and its related species is 2n=40.

Mukherjee hypothesized that mango is a polyploid because it

has a high chromosome number with many nucleolar

chromosomes. Citing secondary meiotic associations, he

further hypothesized that a certain primitive mango type or

types probably arose through allopolyploidy and most likely

via amphidiploidy.

Most authors seem to agree that the center of origin

for mango is the Assam-Burma region (Indo-Burma), where

mangos can still be found growing wild (Singh, 1969). The

Mangifera genus probably originated in Burma, Thailand,

Indo-China, and the Malay Peninsula (Singh, 1960). A marked

center of diversity occurs in the region of

Indochina-Malaysia-Indonesia (Singh, 1976). Wild mangos

generally have fruit that is extremely fiberous-resinous and

may even be poisonous in some species of Mangifera.

Selection has been for fruits with succulence, low fiber,

high sugars, small stones, and low resin content (Singh,


Mangos were probably introduced from southern India to

Malaysia around 500 B.C., and to eastern Africa in the 10th

century by Arab merchants (Purseglove, 1968). Mukherjee

(1972) believes that mangos did not reach the Pacific

islands until the sixteenth century when the Portuguese

began opening trade routes. They are often given credit for

the extensive spread of mangos throughout the tropics

(Mukherjee, 1972; Popenoe, 1920). Their far-reaching trade

routes linked the Indian ports with the Persian gulf,

Pacific islands, and the African coast. Although the date is

uncertain, the Portuguese are reported to have brought the

Indian, monoembryonic type mangos to Rio de Janeiro around

1700 (Mukherjee, 1972). In Mexico, the earliest mango

introductions were probably polyembryonic, Indochinese

seedlings that arrived via the Spanish trade routes between

Manila and the Pacific ports of Mexico (Malo, 1977).

Henry Perrine probably first introduced mangos into the

United States from Campeche, Mexico in 1833 (Popenoe, 1920).

Rolfs (1915) divides the early history of mango cultivar

development in Florida into 2 periods. The first period is

characterized by the growing of seedling mangos by a few

dedicated and ingenious horticulturists and lasted from 1867

until the end of the 19th century (Knight, personal

communication). These mangos were polyembryonic types such

as 'Turpentine' and 'No. 11' (Young and Sauls, 1979). The

second period is characterized by the planting of grafted or

budded trees and began with the noteworthy introduction of

the monoembryonic 'Mulgoba' in 1889 (Rolfs, 1915).

'Mulgoba', an Indian cultivar brought to Florida by Van

Deman of the U.S. Department of Agriculture, was the female

parent of the open pollinated seedling 'Haden'. The

commercial potential of 'Haden' was at once realized and it

has served as the standard by which all other Florida

cultivars have been compared.

Recent surveys indicate close to 1,000 ha of commercial

mango orchards in South Florida (IFAS, 1983). In 1977 over

50% percent of the total acreage was devoted to 'Tommy

Atkins' and 'Keitt' (Malo, 1977). Other important cultivars

include 'Irwin', 'Kent', 'Van Dyke', 'Jubilee', 'Sensation',

'Palmer', and 'Haden'.

Mango breeding was initiated early in this century in

India and has been reviewed by Mukherjee et al. (1968),

Singh (1959), and Singh (1969). Unfortunately, the results

from these breeding efforts have been rather discouraging.

Singh (1959) lists some of the major problems associated

with the breeding of mango in India, e.g., long generation

times (6 years or more), one seed per fruit, high levels of

heterozygosity which makes hybrid performance unpredictable,

unique floral morphology and excessive fruit drop which

makes controlled hybridizations extremely difficult, and the

large field plots needed to grow segregating seedling


Mukherjee (1976) outlined 3 major methods for cultivar

improvement in mango. The first method is selection from

natural seedlings. This was the way that most of the early

Indian cultivars arose and it is still widely practiced in

developing nations. The second method is selection from

open-pollinated seedling progenies. This technique has been

successfully employed in Florida and Hawaii. Mukherjee

considers that this is still the most promising method for

cultivar development. The final method of improvement is

through controlled hybridization. The major problem seems to

be the overall difficulty in producing large populations of

hybrid seedlings. Hand pollinations have very low success

rates (0.3%) (Singh, 1969). These rates can be improved

somewhat if a relatively small number (40-50) of freshly

opened flowers per panicle are pollinated and the rest

removed (Mukherjee, 1976). Singh (1969) states that the most

important variable in obtaining large numbers of hybrid

seeds is the total number of panicles worked per tree and

not the total number of flowers. One method to allow

mass-pollination that has been employed in Florida with some

success is the "cage method." Parents are enclosed within a

screen tent, and pollination is accomplished by placing a

fly infested animal carcass within the enclosure (personal

communication, Knight).

Polyembryony can also restrict gene flow in controlled

hybridizations. The problem is not as pronounced in mango as

it is in Citrus, because most of the high quality Indian and

Florida cultivars are monoembryonic. Knight (1970)

recommends the increased usage of polyembryonic types and

cites favorable characteristics of fruit quality, higher

levels of disease resistance, regular high yields, and ease

of propagation. Spontaneous somatic mutations are another

source of improved cultivars and have given rise to



Polyembryony, the production of multiple embryos within

a single seed, is widespread but sporadically distributed

among the spermatophytes (Gustaffson, 1946). It was first

recorded by Leeuwenhoek in 1719 when he observed orange

seeds containing multiple embryos (Maheshwari, 1950).

Polyembryony is quite common in the gymnosperms but usually

involves some form of cleavage polyembryony, i.e., division

of the zygote or proembryo into 2 or more units (Maheshwari,

1950). It is less widespread in the angiosperms, but

exhibits more diversity in its expression. In the

angiosperms adventive embryos may originate from the egg

cell or synergids. This may occur within a single reduced or

unreduced embryo sac of an ovule, either following or

without fertilization. It may also occur from a fertilized

egg cell through the cleavage of the zygote or proembryo

(cleavage polyembryony) or directly from the cells of the

nucellus or inner integument of an ovule, i.e., adventitious

polyembryony (Brizicky, 1964). If the embryos are not a

result of the normal sexual process then polyembryony may be

correctly referred to as a form of apomixis, termed

agamospermy (Stebbins, 1941). Apomictic polyembryony can be

either gametophytic (diplospory or apospory, parthenogenesis

or apogamety) or sporophytic adventitiouss embryony)

(Webber, 1940; Maheshwari and Sachar, 1963).

Ernst (1918) and Schnarf (1929) (cited in Webber, 1940)

reviewed the old literature on polyembryony and

distinguished 2 main types: true and false. Their

distinction was based on whether or not the embryos arise

from or protrude into the same or different embryo sacs.

Webber (1940) considered these distinctions purely arbitrary

with no physiological or natural basis.

Polyembryony has been most widely studied in the genus

Citrus. Early reports indicated that the adventive embryos

arise in the nucellus shortly after the first division of

the zygote or sometime later (Bacchi, 1943; Rangan et al.,

1969). Substantial evidence now seems to indicate that

proembryos or proembryo initials are present in the

peripheral cells of polyembryonic nucelli prior to anthesis

(Esen and Soost, 1977; Kobayashi et al., 1979). Kobayashi et

al. (1981), studying some 74 Citrus cultivars, found

distinct primordial cells to be present at the time of

flowering in all of the polyembryonic cultivars, but did not

observe these cells in any of the monoembryonic cultivars

studied. Esen and Soost (1977) concluded that although

adventive nucellar embryogenesis seems to be independent of

pollination or fertilization, subsequent embryo development

and maturation require at least fertilization and division

of the polar nuclei endospermm). They added that previous

reports of adventive embryo development in the absence of

fertilization (Webber, 1930; Wright, 1936) were probably in


Survival is a function of the relative vigor of one

embryo and its location in respect to available nutrients

within the embryo sac (Esen and Soost, 1977). Numerous

studies in Citrus have shown that, in general, the location

of the nucellar embryos provides a competitive advantage for

their survival relative to that of the zygotic embryo (Ohta

and Furusato, 1957; Frost and Soost, 1968). Similarly, it

has been found that the more nucellar embryos initiated per

ovule the less likely it is that the sexual embryo will

survive to maturity (Frost, 1926). Frost (1926) indicated

that there is evidence of heterotic vigor in zygotic embryos

from wide crosses and that selection of parents can

significantly influence the survival chances of the zygote

relative to that of its asexual neighbors.

Histological Studies of Mango Embryology

Polyembryony in mango was first recorded by Schact in

1859 (Belling, 1908). The nucellar origin of the adventive

embryos in both Citrus and mango was elucidated by

Strasburger in 1878 (cited by Belling, 1908). Juliano and

Cuevas (1932) described megasporogenesis in the mango cv.

Pico. In 'Pico', an important polyembryonic Philippine

cultivar, very few perfect flowers of the panicle contain

functional megasporangia at the time of flowering. The

megasporangium ovulee) is formed from an outgrowth of the

inner carpellary wall. It grows at approximately the same

rate as the carpel until fertilization, at which time there

is a dramatic enlargement of the carpel walls. Unilateral

growth patterns give rise to a bitegmic, crassinucellate,

anatropous ovule contained within a single unicarpous pistil

(Juliano and Cuevas, 1932). A single hypodermal cell in the

nucellus functions as the archesporial cell. This

archesporial cell does not undergo meiosis directly, but

rather divides periclinally to form an outer primary

parietal cell and an inner primary sporagenous cell

megasporee mother cell). The primary parietal cell

subsequently forms an extensive nucellus while the megaspore

mother cell is pushed deeper into the developing ovule. The

megaspore mother cell differentiates and becomes distinct

from the rest of the nucellus, with its densely staining

cytoplasm and large polygonal shape (Juliano and Cuevas,


It is at this time, apparently, that certain cells of

the nucellus, especially in the region of the micropyle

surrounding the megaspore mother cell, become

distinguishable as proembryo initials by their densely

staining cytoplasm (Belling, 1908). The megaspore mother

cell enters into meiosis that results in the formation of a

linear tetrad of daughter cells. The chalazal-most megaspore

daughter cell becomes functional (thereby pushing itself

deeper into the nucellus) and undergoes megagametogenesis to

give rise to the normal 8-celled embryo sac. The polar

nuclei, at maturity, occupy a position very close to the egg

cell apparatus.

Belling (1908) reports that at the time of

fertilization, in the polyembryonic mango 'No. 11', dense

protoplasmic nucellar cells, separated from the embryo sac

by a layer of flattened cells, can be distinguished.

Immediately following double fertilization, while still in

the flower stage, the triploid endosperm nucleus starts

dividing. These divisions take place well before any

divisions of the zygotic or adventive embryos. In fact, the

first embryonic divisions cannot be detected until the

ovules are more then 3 mm long (fruit 7 mm). As the

endosperm develops, the nucleated protoplasm becomes

appressed to the embryo sac wall opposite the places where

the adventive embryos are forming. Belling (1908) contrasts

the cellularization of the endosperm in Citrus and mango

stating that it occurs much earlier in Citrus than in mango.

Sachar and Chopra (1957) studied endosperm development

in a large number of monoembryonic and polyembryonic mangos,

concluding that the endosperm is of the nuclear type.

Numerous free nuclei are formed and distributed along the

periphery of the embryo sac. The nuclei often fuse and

exhibit large variation in size and shape. Cellularization

occurs quite late and proceeds from the micropyle to the

chalazal end of the embryo sac. They found an extremely wide

range of variation with respect to the time of adventive

embryo initiation and development in polyembryonic mangos.

It waa not uncommon to observe proembryos adjacent to

cotyledonary embryos. The location of the adventive embryos

is also quite variable, but is generally more prevalent in

the micropylar half of the ovule. Nucellar embryos are

commonly observed fused at their radical, resulting in the

production of multiple shoots on a single root.

Sachar and Chopra (1957) also described embryo

development in the monoembryonic Indian cv. Desi. The zygote

remains in a resting state until hundreds of endosperm

nuclei are produced. Two fairly rapid divisions form a

3-celled proembryo. Both the apical and the basal cells

contribute to the formation of the embryo proper with no

organized suspensor.

Several studies have been conducted in mango to

ascertain the fate of the zygotic embryo in the

polyembryonic genotypes (Belling 1908; Juliano, 1934; 1937;

Sachar and Chopra, 1957). In some cultivars, the zygotic

embryo is more or less persistent, but in most cases it

eventually aborts (Sachar and Chopra, 1957). This may be due

to its inferior location within the embryo sac.

Factors Affecting Polyembryony

Because of the economic importance of polyembryony in

Citrus breeding, numerous studies have been conducted on

ways to influence agamospermy in this genus (Traub, 1936;

Furusato et al., 1957; Furusato and Ohta, 1969; Watanabe,

1985a; 1985b). Considerable year to year variation has been

documented and a number of environmental factors have now

been shown to significantly affect the percentage of zygotic

seedlings produced by polyembryonic Citrus cultivars.

Information concerning polyembryony in general is

complicated by the fact that monoembryony does not

necessarily imply sexual reproduction (Ozan et al., 1962).

Particularly in cases where the degree of polyembryony is

low, i.e., the mean number of embryos per seed is only

slightly greater than one, a monoembryonic seed may well

contain a single asexual embryo. To avoid this problem the

more recent studies involving polyembryony in Citrus are

performed using Poncirus trifoliata L. as the pollen parent,

since this species contains a single dominant gene coding

for the easily recognizable trifoliate leaf character.

Traub (1936) was one of the first researchers to report

the artificial control of polyembryony. He indicated that by

decreasing the food supply to fruit-bearing twigs he was

able to decrease the number embryos per seed. Furusato et

al. (1957) found that the mean number of embryos per seed

was significantly higher on the north side of the tree than

on the south side, in older trees than in younger (30 vs 5

years), and in years of high yields than in off years. The

percentage of zygotic to nucellar seedlings has also

reportedly been increased by high temperature greenhouse

incubations after fertilization (Nakatani et al., 1978) and

by flower and fruit bud treatments with plant growth

regulators (PGRs) including: maleic hydrazide (Furusato and

Ohta, 1969), GA (DeLange and Vincent, 1977), and

2,3,5-triiodobenzoic acid (Yoshida, 1979). Watanabe (1985a;

1985b) was unable to significantly reduce the number of

embryos per seed by use of X-rays, NAA, coumarin, or 2,4-D

in addition to any of the previously mentioned PGRs.

Furthermore, the use of high temperatures to reduce embryo

number has not been reproduced (Moore, personal

communication). There have been several references

concerning the use of radiation to reduce polyembryony

(Spiegel-Roy et al., 1972; Ikeda, 1981; Watanbe, 1985a;

1985b). Ikeda (1981) found that a low dose of gamma rays

(1-2 kR) applied for 20 hours to small floral buds (20-30

days pre-anthesis) selectively reduced the number of

nucellar embryos per seed. The developmental period at which

the radiation was supplied was a critical factor. Watanabe

(1985b) obtained a high percentage (86-100%) of

interspecific hybrids by growing artificially pollinated

Citrus trees in a gamma field. In another study Watanabe

(1985a) demonstrated histologically that adventive embryos

were preferentially inhibited by exposure to continuous

gamma irradiation (500 R/day) and that they seldom developed

past the quartet cell stage.

Tisserat and Murashige (1977a; 1977b) isolated a

"graft-transmissible" and "diffusable embryogenic repres-

sant" from monoembryonic citrus cultivars. They found that

the chalazal half of monoembryonic C. medical L. ovules

suppressed embryogenesis in several embryogenic plant

cultures. High levels of ethanol, IAA, ABA, and GA all

suppressed embryogenic activity.

In conclusion, from these studies it is quite clear

that the expression of polyembryony can be significantly

altered by environmental conditions (Furusato et al., 1957;

Watanabe 1985a; 1985b). Generally, conditions which stress

the tree have reduced the degree of polyembryony.

The Genetics of Polyembryony

Several studies attempting to elucidate the inheritance

of polyembryony in Citrus have been conducted (Parlevliet

and Cameron, 1959; Ozan et al., 1962; Cameron and Soost

1979). Parlevliet and Cameron (1959), utilizing several

segregating seedling populations derived from hand

pollinations, sampled some 700 Citrus hybrids. The

monoembryonic cultivars produced only monoembryonic progeny,

whereas the monoembryonic X polyembryonic (MxP) crosses

produced seedling populations exhibiting a wide range of P:M

offspring (8:19-10:0). They postulated a rather simple mode

of inheritance for polyembryony, stating that 1 or possibly

2 dominant or semidominant genes along with possible

modifying genes are responsible for polyembryony. Ozan et

al. (1962) analyzed some 2,000 hybrid seedlings derived from

crosses involving three apparently obligate monoembryonic

Citrus cultivars crossed with the polyembryonic P.

trifoliata. The total percentage of P seeds ranged from

2-8%, while the range of embryos per seed was 2-6. When the

P seeds were grown out, all evidence indicated that the

seeds were not apomictic in origin but rather the result of

cleavage of the zygotic embryo. In a more recent study,

Cameron and Soost (1979) analyzed a population of 121

seedlings derived from 14 crosses involving both M and P

parents. In their discussion, Cameron and Soost, clearly

indicate that their data are in contrast to previous reports

in which generally higher proportions of P offspring are

found. To account for this, they proposed that polyembryony

in Citrus is controlled by two complementary, dominant


In summary, the mode of inheritence for polyembryony in

Citrus is more complex than was first suspected and does not

appear to be controlled by a single or even two dominant

genes. Numerous crosses have shown that the genetic

background can significantly influence the expression of

polyembryony, implicating the possible role of modifying

genes. Some crosses of P x M Citrus produce only P

offspring, while another cross between monoembryonic parents

appears to have given rise to a polyembryonic offspring

(Cameron and Soost, 1979).

The situation in mango is also unclear due to the lack

of critical genetic studies. LeRoy (1947) proposed that

polyembryony in mango is controlled by 1 or more recessive

genes. Sturrock (1968) also concluded that polyembryony in

mango is recessive and probably controlled by a single gene,

but his conclusions were based on suspected cross

pollinations rather than controlled crosses. Litz and

Schaffer (1987) found that putrescine concentration is much

higher in non-embryogenic nucellar callus of monoembryonic

mangos than in non-embryogenic or embryogenic callus of

polyembryonic cultivars. Litz (1987) speculates that

blocking of S-adenosyl-L-methionine decarboxylase (SAMDC)

activity, which has been found associated with the

accumulation of putrescine and the production of ethylene

may cause ethylene production in mango nucelli. The ethylene

production, which has been shown to inhibit somatic

embryogenesis (Tisserat and Murashige, 1977c), would then

play an important regulating role in the control of

polyembryony in mango.

The Evolutionary Significance of Apomixis

The question often arises as to the evolutionary

significance of adventive embryony and other apomictic forms

of reproduction. Several hypotheses have been proposed

regarding the adaptive significance of apomictic plants

(Marshall and Brown, 1981). One hypothesis referred to as

the "escape from sterility" hypothesis proposes that there

is no selective advantage in agamospermous reproduction over

sexual reproduction. This view stresses that the principal

advantage of apomixis is to restore fertility to sexually

sterile individuals, generally derived from polyploidy and

wide hybridization. In the past this theory was widely

accepted because it was compatible with the theory that

sexual reproduction maximizes the rate of adaptive

evolution. It also provided a ready cause for the close

association of apomixis with polyploidy and interspecific

hybridization. Clausen (1954) was one of the first to

recognize formally that apomixis, particularly facultative

apomixis, does not automatically limit an individual's

capacity for adaptive evolution. He compared the adaptive

advantage of facultative agamospermy to the mass production

of automobiles, in which the production of new genotypes is

permitted at the same time that reproductive fidelity of the

best genotypes can be faithfully maintained. The hypothesis

often referred to as the "Henry Ford" or "Model T"

hypothesis has gained wide acceptance (Marshall and Brown,

1981). Williams (1975) has cited one shortcoming associated

with this hypothesis, i.e., it ignores the cost of meiosis.

Williams explains that the cost of meiosis is incurred for

sexual reproduction at both the individual and at the group

level. At the individual level only one half of the total

genomic complement is passed on to the offspring in sexual

reproduction, while at the population or group level

valuable resources are spent on the production of male

gametes. A third hypothesis has evolved out of recent

quantitative studies which consider the cost of meiosis

(Maynard-Smith, 1978). It has been called the "automatic

advantage" hypothesis because it proposes that apomictic

plants have a twofold advantage over sexual reproduction.

The first advantage is that the apomictic offspring carry

the full genetic complement of their mother. The second

advantage is that no resources are wasted on male gametes;

hence, the cost of meiosis is taken into consideration.

Plant Tissue and Cell Culture

Plant cell and tissue culture has evolved over the past

2 centuries as a result of observations concerning the

response of excised plant parts. Several botanists in the

mid-nineteenth century observed that plants often produce

callus, in response to wounding. The initiation and growth

of this dedifferentiated tissue has proved to be of

particular importance to in vitro culture of plant tissues

and cells.

Some of the first attempts to culture excised plant

tissues were made by the German botanist, Gottlieb

Haberlandt (1854-1945). Haberlandt is credited with the

theory of totipotency, which states that every plant cell,

under the proper conditions, is capable of giving rise to a

whole plant (Haberlandt, 1902). The first successful plant

tissue cultures were derived from excised root tips

(Knudson, 1916). Using a tomato root with an intact

meristem, White (1934) was able, by subculture, to establish

an in vitro continuous culture.

Further advances in the culture of excised plant parts

awaited the discovery of a complex group of natural and

synthetic plant growth substances. Went (1926), using

detached oat coleoptiles, discovered the first plant growth

regulator, indoleacetic acid (IAA). Its subsequent isolation

and chemical characterization (Kogl et al., 1934), led Snow

(1935) to recognize the importance of this naturally

occurring auxin for cell proliferation. The incorporation of

IAA in the culture medium resulted in the first continuous

plant tissue (not organs) cultures in 1939 by White (1939),

Nobecourt (1939), and Gautheret (1939).

Using a medium enriched with liqud coconut endosperm

(LCE), Van Overbeek et al. (1941) were able to culture

immature hybrid embryos of Datura species. The isolation of

the cytokinin, kinetin by Miller et al. (1955) lead Skoog

and Miller (1957) to the recognition that the

differentiation of shoots or roots in tobacco callus was

determined by the ratio of auxin:cytokinin in the culture

medium. Although differentiation in all plant tissue and

cell cultures is apparently not controlled so simply as in

the tobacco system, its dependence on the interplay of

complex plant growth substance and nutrients does still

appear to be valid.

Steward et al. (1958a) and Reinert (1958) were able to

regenerate whole plants via somatic embryogenesis from

carrot suspension cultures. The inherent uncertainty of the

totipotency concept as observed in liquid suspensions

cultures was overcome by Vasil and Hildebrandt (1965) using

tobacco hybrid cells cultured in a suspended drop of medium

enriched with LCE and NAA. Thus, the concept of totipotency,

as envisioned by Haberlandt almost 60 years earlier had been


Two major pathways are distinguished in the de novo

regeneration of plants in vitro, organogenesis and somatic

embryogenesis. In organogenesis, plants are regenerated via

the separate formation of shoots and roots, without

exhibiting the distinct developmental stages observed in

zygotic embryos. The organs are formed via meristems with

vascular connections to the explant, and are probably not of

single cell origin. The ratio of auxin:cytokinin in the

culture medium is often very important in determining root

or shoot production (Skoog and Miller, 1957). Typically, a

shoot is formed followed by a root, which may or may not

require subculture to a different nutrient medium (Flick et

al., 1983). In contrast, regeneration via somatic

embryogenesis is characterized by the absence of vascular

connection to the maternal tissue of the explant and has

been demonstrated to have a single cell origin. Somatic

embryos pass through developmental stages similar to those

of zygotic embryos, germinate and form plantlets (Ammirato,


Somatic Embryogenesis

Somatic embryogenesis is the process of embryo

initiation from cells that are not the products of gametic

fusion, i.e., somatic cells (Tisserat et al., 1979).

Tisserat et al. (1979) suggested that adventive and somatic

embryogenesis are valid and useful synonyms for asexual

embryogenesis when referring to the general case, but

apomixis and nucellar embryony should only be used to

describe specific naturally occurring events. In this

manuscript somatic embryogenesis will be used when referring

to the in vitro situation and adventive nucellar

embryogenesis will be used to refer to the in vivo event.

An embryo is an early stage in a plant's development.

It is a bipolar structure with a closed, discrete vascular

system. The closed vascular system implies no maternal

vascular connection, while the bipolarity arises from a

shoot and root meristem at opposite ends. Ammirato (1983)

cautions against the indiscriminate use of the term in vitro

somatic embryogenesis to describe cultures where no

histology has been performed and plantlets have not been

regenerated. Although many systems superficially resemble

somatic embryogenesis, histological studies are sometimes

needed to confirm the lack of vascular connections in the

somatic embryos.

Haberlandt's theory of totipotency (1902) was verified

by Reinert (1958) and Stewart et al. (1958a) who

independently reported somatic embryogenesis in tissue

cultures of carrot (Daucus carota). Since that time in vitro

somatic embryogenesis has been reported in a number of

monocots, dicots, and gymnosperms (Tisserat et al., 1979).

Several theories have been advanced to explain the

phenomenon of in vitro somatic embryogenesis based upon

observations of embryogenic carrot cultures, the model

system for somatic embryogenesis. Sharp et al. (1980) have

arbitarily distinguished 6 theories.

One of the most widely accepted theories postulates

that cells must undergo dedifferentiation before they can

attain embryological competency (Halperin, 1970). Indeed

this seems to be the case in carrot, where a recognizable

callus stage is a prerequisite to somatic embryogenesis.

However, not all embryogenic systems have a distinguishable

dedifferentiated stage, e.g., Citrus (Button et al., 1974)

and Ranunculus (Konar et al., 1972).

Another theory postulates that a cell must be

physiologically isolated from repressive factors of its

neighboring cells. This theory evolved mainly from

observations with the meiocytes of higher plants and has

been corroborated by observation with Citrus (Button et al.,


A third theory has been developed primarily to account

for observations which are not readily explained by the

theory of dedifferentiation. Street (1978) proposed that

dedifferentiation, i.e., callus formation, itself does not

guarantee embryological competence, but rather the

physiological status of the individual explant cells at the

time of culturing together with the subsequent culture

environment are the primary determinants. This theory is

strongly supported by the embryogenic qualities of nucelli

and other reproductive tissue in general.

A fourth theory emphasizes the importance of

intercellular communication and subsequent cytodifferentia-

tion for somatic embryogenesis (Halperin, 1967). It evolved

from observations on the importance of callus formation

prior to somatic embryogenesis, and has been supported by

observations with cultured protoplasts in which a cell mass

is formed prior to somatic embryogenesis.

The fifth theory hypothesizes that the physiological

status of the explant itself is the primary determinant of

embryological competence (Tisserat et al., 1979). The

culture environment only enhances or represses embryogenic


Finally, Sharp et al. (1980) have developed a working

hypothesis which attempts to integrate many aspects of the

previously discussed theories. They propose that in vitro

somatic embryogenesis follows 2 major developmental

patterns. In the first pattern somatic embryogenesis occurs

directly from cells in the explant that have been

predetermined to form somatic embryos. The culture

environment can be envisioned as permissively allowing these

predetermined cells to undergo the process of embryogenesis.

Examples of this pattern would be Citrus and mango nucellar

cultures. In contrast, the second pattern of somatic

embryogenesis hypothesizes that the explant must

dedifferentiate before it can be induced in a directed

manner to undergo embryogenesis. An example of indirect

somatic embryogenesis would be the carrot pith cultures. In

both cases the cultured cells are induced to undergo

embryogenesis, but in the first case it is a permissive

induction, while in the latter it is a directive induction.

Sharp's hypothesis appears to delineate a possible role

for plant growth regulators as agents that contribute to the

determination of the cells but that do not directly alter

their embryogenic competence. It also provides some

explanation for the disparities observed in the effects of

auxins on somatic embryogenesis in various plant culture

systems. In the direct pattern of embryogenesis, auxins are

viewed as merely cloning agents which produce multiple

copies of the predetermined cells while in indirect

embryogenesis auxins are envisioned as mitogenic substances

which redetermine the callus cells to an embryogenic state.

Factors Affecting In Vitro Somatic Embryogenesis

Somatic embryogenesis is a dynamic phenomenon

involving many genes and gene complexes, all controlled in a

highly coordinated fashion (Sung et al., 1984). Several

factors have a direct or indirect effect on somatic

embryogenesis, and a number of reviews have addressed this

(Raghavan, 1976a and 1976b; Reinert et al., 1977; Kohlenbach

1977; 1978; Tisserat et al., 1979; Ammirato, 1983; 1984;

Bhojwani and Razdan, 1983; Ozias-Akins and Vasil, 1985).

Plant genotype. Certain taxa have been more amenable to

in vitro somatic embryogenesis, e.g., Umbelliferae, but

within a given species or taxonomic group, individual

genotypes often show pronounced differences in their

embryogenic potential. In alfalfa (Kao and Michaluk, 1980),

clover (Keys et al., 1980), corn (Lu et al., 1982), carrot

(Steward et al., 1975), and indeed almost every culture

system where somatic embryogenesis has been investigated,

significant differences have been found among individual

cultivars (Ammirato, 1983).

Explant source. The explant source and its

physiological qualities are perhaps the most significant

factors in determining whether an embryogenic culture can be

initiated. Other in vitro influences merely enhance or

repress the embryogenic response (Tisserat et al., 1979).

For some plant species, e.g., D. carota, almost any explant

taken at any stage of development can be used to establish

embryogenic cultures. However, for other, more recalcitrant

plant species, choice of explant at an appropriate

developmental stage may be critical, e.g., sweet potato (Liu

and Cantliffe, 1984). Floral tissues, in general, have been

very useful for initiating embryogenic cultures in a number

of plant species (Ammirato, 1983). Ovules of several fruit

species have been successfully cultured, e.g., Citrus

(Kochba and Button, 1974), grape (Srinivasan and Mullins,

1980), apple (Eichholtz, 1979), mango (Litz et al., 1982),

papaya (Litz and Conover, 1982), jaboticaba (Litz, 1984c),

and Eugenia spp. (Litz, 1984a). Zygotic embryos have also

proved to be very useful for initiating embryogenic

cultures, e.g., cotton (Joshi and Johri, 1972), Ranunculus

sceleratus (Sachar and Guha, 1962), and barley (Norstog,

1970). Anthers and microspores are also frequently used,

e.g., Triticale (Sun et al., 1973), Ranunculus sceleratus

(Nataraja and Konar, 1970), Prunus avium (Zenkteler et al.,

1975), Coffea arabica (Sharp et al., 1973), and Datura (Guha

and Maheswari, 1964) The use of nonmaternal tissues, e.g.,

microspores and zygotic embryos, has limited the application

of somatic embryogenesis for propagation and for improvement

of clonally propagated crops. Meristems, e.g., coffee (Sharp

et al., 1973), Ranunculus (Nataraja and Konar, 1970), and

immature vegetative tissue, e.g., millets (Vasil and Vasil,

1981), sorghum (Brettell et al., 1980) are other important

sources of explant material.

The physiological condition of the explant at time of

culturing also significantly affects its embryogenic

potential. In many recalcitrant plant species there appears

to be a rather discrete developmental window, during which

the cells of the explant are receptive to outside influences

and somatic embryogenesis can be induced. With Citrus

Gmitter (1985) found that ovules of C. sinesis Osb. cv.

Hamlin became more responsive with increased fruit maturity.

The size and shape of the explant can also be important.

Explants below a certain minimum size may not have enough

viable cells in contact with the medium (Flick et al.,

1983). The polarity of the explant can also influence its

responsiveness. In Pennisetum, Vasil and Vasil (1981) found

that excised zygotic embryos were most responsive when both

the root and the shoot were in direct contact with the

culture medium. The ploidy level of the explant does not

seem to be a limiting factor, and somatic embryogenesis has

occurred from haploid, diploid, and polyploid tissues.

Plant growth regulators. Perhaps the most widely

studied variable affecting in vitro somatic embryogenesis

has been the kind and quantity of PGRs used in the culture

medium. Of the PGRs, auxin, particularly 2,4-D, has been the

most often cited as being useful for in vitro somatic

embryogenesis. In carrot, callus is maintained on

2,4-D-supplemented media. The callus is transferred to an

auxin-free medium in order to induce somatic embryogenesis

(Halperin and Wetherell, 1964; Halperin, 1966). In general,

the presence of auxin or of an auxin-like substance is

required for the initiation of embryogenic cultures, while

the quantitative lowering or qualitative changing of the

auxin source generally allows somatic embryo growth and

maturation to proceed (Ammirato, 1983). When in vitro

somatic embryogenesis has occurred without auxin in the

culture medium, it has involved tissues that were embryonic,

e.g., cotyledons of Ilex (Hu and Sussex, 1971) or

reproductive, e.g., ovules of Citrus (Kochba and

Spiegel-Roy, 1973). In Citrus and mango (Litz and Schaffer,

1987) it appears that somatic embryogenesis can occur

without the addition of PGRs, but the addition of exogenous

auxins does increase callus proliferation. The addition of

2,4-D to the culture medium is more effective for callus

proliferation in Citrus than either NAA or IAA (Murashige

and Tucker, 1969); however, Moore (1985) and Kochba and

Spiegel-Roy (1977) found that the addition of various auxins

significantly reduces somatic embryogenesis in several

Citrus tissue culture systems even at relatively low

concentrations (0.5 uM). Raghavan (1984a) points out that

many of the confusing results concerning the role of auxins

in somatic embryogenesis are related to the use of diverse

tissues (freshly isolated vs. habituated callus), the size

of the inoculum, and the concentration and type of auxin.

Sharp et al. (1980) make a strong distinction between the

auxin requirement for cultures that undergo indirect somatic

embryogenesis and those that undergo direct embryogenesis,

e.g., Citrus. They concluded that there is an absolute

requirement for exogeneous application of auxin in those

cultures undergoing indirect embryogenesis, and its role is

hypothesized to be one of a directive inducer, while in

systems undergoing direct embryogenesis, there does not seem

to be a strict auxin requirement. The role of auxin in these

systems is thought to be as a cloning agent for

pre-embryogenic determined cells.

Cytokinins are another class of PGRs routinely used in

embryogenic culture systems. Their quantitative interaction

with auxins and their role in developmental control are well

documented in organogenesis (Skoog and Miller, 1957). The

type and concentration of cytokinin often differ in media

used for callus multiplication, embryo induction, and embryo

maturation media. Evans et al. (1981) reported that

cytokinins have been employed in the initiation medium of

over 65% of all crop plants that have been cultured.

Abscisic acid (ABA) has been used successfully in a

large number of embryogenic plant culture systems. It is an

inhibitor of embryo development, whereby it allows embryo

maturation to proceed in a more normal fashion. This

normalizing effect has been reported in caraway (Ammirato,

1974), carrot (Kamada and Harada, 1981), Pennisetum (Vasil

and Vasil, 1981), and Citrus (Kochba et al., 1978).

According to Ammirato (1983) ABA also suppresses secondary

embryo proliferation and precocious germination.

Gibberellins have been used successfully in a few

instances to induce a normal pattern of embryo development

and stimulate germination and shoot elongation, e.g., Citrus

(Kochba et al., 1974), Panicum (Lu and Vasil, 1981), and

corn (Lu et al., 1982). Ethylene and other volatile gases

may exert a significant effect on closed tissue culture

systems but little research has been done to establish these

effects (Ozias-Akins and Vasil, 1985). Ethephon, which

liberates ethylene, was reported to stimulate embryogenesis

in Citrus cultures (Kochba et al., 1978).

Nitrogen. Elemental nitrogen occurs in the highest

concentration of all the macroelements included in most

culture media. The total nitrogen concentration of the

various nutrient media ranges from 3.3 mM in White's to 60.0

mM in MS (sulfur ranges from 1-5 mM while sodium ranges from

0.2-8.0 mM). Of all the mineral nutrients, the form in which

nitrogen (oxidized or reduced, organic or inorganic) is

supplied, probably has the most dramatic effect on both

growth and differentiation of cultured cells (Ozias-Akins

and Vasil, 1985). The importance of total nitrogen and the

form in which it is supplied to in vitro somatic

embryogenesis has been well established (Halperin and

Wetherell, 1965). A direct requirement for nitrogen either

as ammonium or in another of its reduced forms is necessary.

The time at which nitrogen is applied also appears to

be important. Halperin and Wetherell (1965) found that

cultures initiated on a medium supplied with nitrogen in the

nitrate form were less embryogenic than cultures initiated

on media containing the same total nitrogen, but supplied in

the ammonium form. Gamborg et al. (1968) found that when the

total nitrogen requirement of a suspension culture medium is

supplied in the form of the ammonium ion, the pH has a

tendency to drop and may ultimately restrict the

availability of nitrogen. Although the importance of

nitrogen and the form in which it is supplied has been well

established, there seems to be no general formula which

optimizes embryogenesis in all plant tissue culture systems.

Cultured cells can differ in their ability to utilize the

various forms of nitrogen. Carrot cells in vitro

characteristically have a very well developed glutamine

synthetase system (Caldas, 1971). Thus, the reduced forms of

nitrogen may not be as important as in other plant cultures.

The concentration of 2,4-D in the medium can also

significantly alter the culture's ammonium requirement

(Halperin and Wetherell, 1965; Halperin, 1966).

Many plant culture systems respond well to the addition

of reduced forms of nitrogen, i.e., amino acids, to the

culture medium. In general, glutamine and to a lesser

extent, glycine and aspartic acid have been the most useful

amino acids for increasing the efficiency of somatic

embryogenesis (Ammirato, 1983). Complex organic addenda such

as endosperm extracts, casein hydrolysate (CH), and malt and

yeast extracts serve as important sources of amino acids and

other reduced forms of nitrogen (Ozias-Atkins and Vasil,


Carbohydrates. Carbohydrates serve as a carbon source

for plant cultures, and because they are often present at

such high concentrations (1-12%) they are effective as an

osmoticum. The disaccharide sucrose is usually the most

effective sugar (Verma and Dougall, 1977), but other sugars

are also used. Maltose increases the germination of Citrus

aurantifolia somatic embryos (Miller, 1986). Lactose and

galactose increase somatic embryogenesis in various Citrus

spp. (Kochba et al., 1982), and sorbitol is effective in

Malus cultures (Chong and Taper, 1974). Inositol

(myo-inositol) has only rarely been shown to be essential in

plant tissue culture, e.g., in tobacco (Linsmaier and Skoog,

1965), but it does serve an important role in carbohydrate

metabolism (Loewus and Loewus, 1983) and is routinely

included in the culture medium.

Complex organic addenda. Complex, undefined organic

substances not only are important source of nitrogen, but

also are sources of PGRs, carbohydrates, and vitamins (Dix

and van Staden, 1982). They have also been useful for

somatic embryogenesis. In the early studies with carrot,

Steward et al. (1958a) thought that the addition of LCE to

the culture medium was essential for somatic embryogenesis;

however, efficient somatic embryogenesis has been achieved

on completely defined media (Steward et al., 1964). Many

early plant tissue culture systems using low salt media

relied on the addition of complex, undefined naturally

occurring substances to supplement the simple inorganic salt

formulations being used (Ozias-Atkins and Vasil, 1985).

Various combinations of PGRs, particularly auxins and

cytokinins, can often replace the requirement for complex

organic addenda. Casein hydrolysate was shown to increase

somatic embryogenesis and development in one of the earliest

report of Citrus tissue culture (Maheswari and Rangaswamy,

1958). Malt extract, which has been shown to contain

cytokinins, giberellins, and auxins (Dix and van Staden,

1982), is commonly added to embryogenic cultures of Citrus

(Kochba and Spiegel-Roy, 1973; Moore, 1985).

Other factors. Several undefined gelling agents have

been used to solidify culture media, the most common of

which is agar. The impurities present in agar often are an

important source of trace elements (Heller, 1953). Many

different commercial preparations of agar are now available

which vary significantly in quality. Gelrite gellan gum is a

self-gelling hydrocolloid that forms a transparent gel in

the presence of soluble solids (Pasqualetto et al., 1986).

A number of vitamins have also been used in plant

cultures. Thiamine is the one vitamin which has been

demonstrated repeatedly to significantly affect plant

regeneration (White, 1951; Murashige and Skoog, 1962;

Linsmaier and Skoog, 1965; Gamborg et al., 1968). Other

vitamins included in culture media are pyridoxine, nicotinic

acid, and glycine (Ozias-Akins and Vasil, 1985).

The osmolarity of the culture medium can also affect

somatic embryogenesis and embryo development. Litz (1986)

working with Carica and Wetherell (1984) with carrot found

that pretreatment exposures of embryogenic suspension

cultures to media with high osmolarities (0.3-1.0 M)

followed by subculture to media with lower osmolarities,

resulted in increased levels of somatic embryogenesis. These

results are thought to mimic the changes in osmolarity found

in the endosperm of developing ovules, in which high

concentrations are associated with the earliest stages of

embryogenesis and development (Raghavan, 1976b). High

osmotic concentrations are also thought to suppress organ

differentiation and prevent precocious germination in young

somatic embryos (Ammirato, 1983).

In Citrus, various culture stresses can increase

somatic embryogenesis. 'Shamouti' cultures become more

embryogenic when relatively long culture intervals are used

and when callus is briefly subcultured onto a medium devoid

of sucrose (Kochba and Button, 1974).

Tissue Culture of Mango

The in vitro potential of mango was clearly recognized

by Maheshwari and Rangaswamy (1958); however, no cultures

were actually initiated. The first successful reports of

tissue culture in mango were by Rao et al. (1982) and Litz

et al. (1982). Rao et al. (1982) reported callus initiation

from mango cotyledons on MS medium supplemented with 15%

LCE, 4% sucrose, 27 uM NAA, and 11.5-23 uM kinetin. Cultures

consisted of a dark compact callus that differentiated

prolific roots. Root proliferation continued with subsequent

subcultures, but no shoots or plantlet formation was


Litz et al. (1982) obtained callus initiation and

somatic embryogenesis from cultured ovules of polyembryonic

mangos collected approximately 2 months post-pollination.

They initiated cultures on a modified solid MS medium with

1/2 strength MS major salts, 6% sucrose, 0.8% Difco-Bacto

agar, 2.73 mM glutamine, 0.57 mM ascorbic acid with or

without 20% filter-sterilized LCE or 4.4-8.8 uM BA. The

nucelli were excised from the cultured ovules after 1-3

weeks and subcultured to fresh media. Embryogenic calli were

transferred to either solid or liquid media containing 20%

LCE. Cultures were incubated in a growth chamber at 25 C

with a 16 hr photoperiod (1000 lux). Liquid cultures were

maintained on a rotary shaker at 100 r.p.m. The most

efficient media for somatic embryogenesis was reported to be

liquid media with 20% LCE. The embryogenic response was

cultivar-dependent with 5 of the 9 polyembryonic cultivars

producing embryogenic callus. No root, shoot, or plantlet

production was reported from somatic embryos transferred to

LCE-free media.

There have been additional reports on in vitro somatic

embryogenesis of mango (Litz, 1984b; Litz et al., 1984; Litz

and Schafer, 1987; Litz, 1987). Nucelli and somatic embryos

were excised from cultured mango ovules 1-2 weeks after

culturing and transferred to a liquid medium containing 9 uM

2,4-D instead of LCE. Cultures were maintained in liquid for

3 months with monthly subcultures, and then transferred to

liquid MS medium without PGRs and with or without 0.5%

activated charcoal. Somatic embryos were transferred to a

solid germination medium supplemented with 0-22 uM BA

together with 0.01% CH, 0.01% ME, or 10% LCE. Activated

charcoal was reported to be deleterious to embryo

maturation. Limited root formation was observed on solid

media supplemented with 4.4-8.8 uM BA (Litz et al., 1984).

Excised nucelli from monoembryonic mango ovules when

cultured on a modified solid MS media (described previously)

supplemented with either 4.5-13.5 uM 2,4-D or to a lesser

extent 5.4-27.0 uM NAA, gave rise to a "loose and friable"

slowly growing callus (Litz, 1984b). Somatic embryos were

reported to have arisen indirectly from the nucelli via

callus formation. This observation contrasted with reports

of in vitro somatic embryogenesis in monoembryonic Citrus

where embryos were reported to have arisen directly from the

nucellus without intermediate callus formation (Rangan et

al., 1968). The cytokinins BA and 2iP were both reported by

Litz (1984b) to be ineffective in initiating callus. The

addition of 20% filter-sterilized LCE to the culture media

somewhat delayed a gradual necrosis that was observed in all

the embryos, while casein hydrolysate (CH), malt extract

(ME), and reducing agents were found to be ineffective.

Alterations in physical culture conditions such as light vs.

dark and solid vs. liquid were also reported to be

ineffective in the alleviation of the gradual necrosis.

Limited germination (root formation) was observed in liquid

media without PGRs. Litz and Schaffer (1987) reported that

although 2,4-D enhances callus formation from mango nucelli,

it does not significantly affect somatic embryogenesis.

There have been no histological studies to verify that

regeneration from mango nucellar callus is via somatic

embryogenesis, although the evidence is persuasive. Somatic

embryogenesis is most efficient in liquid cultures

supplemented with 20% LCE or 2,4-D. Polyembryonic and to a

lesser extent monoembryonic mangos both are capable of

somatic embryogenesis. The response is genotype-dependent.

Limited root growth has been reported but no plantlet or

shoot formation. Embryos exhibited a gradual and progressive

terminal necrosis.



The term somatic embryogenesis has been used throughout

this dissertation to refer to the initiation of embryos from

somatic tissues under in vitro conditions. The terms

adventive embryogenesis and adventive nucellar embryogenesis

have been used only to describe the natural in vivo

production of adventive embryos from cells of the nucellus.

Somatic embryos are generally thought to pass through all

the developmental stages exhibited by zygotic and adventive

embryos in vivo (Tisserat et al., 1979; Ammirato, 1983),

although minor differences in the first series of cell

divisions have been reported with carrot somatic embryos in

comparison to zygotic embryo development (McWilliams et al.,

1974). The most notable difference between somatic embryo

and zygotic embryo development in most plant cultures is the

lack of suspensors in the somatic embryos, e.g., Citrus

(Esan 1973), Daucus (Halperin and Jensen, 1967), and

Ranunculus (Konar et al., 1972). Cotyledon abnormalities

have often been associated with the development of somatic

embryos and include polycotyledony (Halperin, 1964), fused

cotyledons (Rashid and Street, 1973), unequally sized pairs

(Rao 1965), and fasciation (Rao 1965).

Materials and Methods

Flowers and fruitlets were gathered during the 1984 and

1985 season from the mango germplasm collection of the

University of Florida, Tropical Research and Education

Center in Homestead. Ovules over 1 cm in length were

dissected from the fruit and placed directly into the

fixative. The material was fixed and stored in FAA [90:5:5

of (95%) ethanol, formalin, glacial acetic acid].

Embryogenic calli were initiated from excised nucelli

of the polyembryonic 'Parris' on solid, modified MS medium

(callus initiation medium, Table 4-3). Embryogenic nucellar

callus was multiplied in suspension culture on a modified

B-5 medium supplemented with kinetin and 2,4-D (callus

maintenance medium C, Table 4-3). Embryo production,

maturation and germination occurred on a modified B-5 media

solidified with 0.19% Gelrite gellan gum and supplemented

with NAA and 2iP (for embryo production) or 20% liquid

coconut endosperm and 0.025% casein hydrolysate (for

maturation and germination). Callus and somatic embryos used

for histology were killed, fixed, and stored for 48 hours in

FAA (prepared with 70% ethanol).

All material used for light microscopy was removed from

FAA fixative, dehydrated with tertiary butyl alcohol, and

embedded in paraffin (Paraplast) according to Johansen

(1940). Sections were cut at 8-9 uM using an American

Optical rotary microtome and stained in safranin and fast


Clusters of cotyledonary somatic embryos produced on

solid media were used for scanning electron microscopy.

Embryos were harvested and fixed for 1 hour in a solution of

4.25% gluteraldehyde with 0.05 M sodium cacodylate buffer.

The material was then rinsed 3 times for 20 minutes in the

cacodylate buffer. Material was postfixed for 1 hour in a 2%

OsO4 solution with 0.05 M cacodylate buffer. Specimen

fixation was followed by three 20 minute rinses in distilled

H20 and dehydration was accomplished using a 6 step series

of ethanol solutions (25, 50, 75, 95, 100, and 100%). Any

remaining moisture was removed by critical point drying.

Material was gold coated under vacuum and observed with a

Philips scanning electron microscope.

Results and Discussion

Adventive Nucellar Embryony. The mango fruit is a drupe

which contains a large fibrous, hard endocarp enclosing the

seed. Figure 3-1 shows a young mango fruit approximately 75

days post-bloom. At this stage the tough, fiberous endocarp

has not yet developed around the ovule (megasporangium). The

fruit is easily cut open to expose the ovule which is borne

on a funiculus or stalk. The micropyle is just below the

point where the funiculus joins the ovule proper. The

chalazal region is at the opposite end of the ovule.

Mangos possess crassinucellate ovules which are

characterized by having well developed and persistent

nucelli. Figure 3-2 is representive of a polyembryonic mango

ovule approximately 30 days post-bloom. In this 'James

Saigon' ovule (3 mm length) the nucellus has already

undergone considerable degeneration. The nucellus appears as

a layer, several cells thick, consisting of mostly flattened

cells surrounding the entire embryo sac cavity. In this

ovule the nucellus is more developed and persistent near the

micropyle. The endosperm has not become cellularized at this

stage of development and is still in the free-nuclear

condition. It is characteristically a thin line of densely

staining nuclei and cytoplasm which is appressed against the

nucellus by the central vacuole. This free-nuclear endosperm

appears to be more concentrated in the micropylar region of

the ovule. The adventive nucellar embryos are a small

cluster of densely stained cells at the extreme micropylar

end of the embryo sac cavity (Figures 3-2; 3-3). In these

micrographs the micropyle is clearly defined. The epidermal

layer of the the integument lines the micropyle and

delineates the cells of the nucellus. In Figure 3-3 six

adventive embryos can be clearly distinguished. They range

from a two-celled proembryo to a twenty-four-celled early

globular stage adventive embryo. In the earliest stages of

adventive nucellar embryony, the proembryos contain cells

with rich densely staining cytoplasm, thin cell walls,

little vacuolation, and prominent nuclei. At approximately

45 days post bloom the adventive nucellar embryos have

enlarged and protrude into the embryo sac cavity as observed

in the 7 mm length 'Parris' ovule (Figure 3-4). At this

stage the endosperm has become fully cellularized. Adventive

nucellar embryos at the early to late globular stage are

demonstrated in Figures 3-5 and 3-6. The 0.5 mm length

adventive embryo in Figure 3-6 possesses a well developed

epidermis; however, no embryonic organs or vascularization

are distinguishable.

In Vitro Somatic Embryony. Embryogenic nucellar callus

can be initiated from both the micropylar and the chalazal

portions of excised nucelli cultured on solid callus

initiation medium (Table 4-3, see Figure 3-7). Whatever

factors limit adventive embryogenesis in the chalazal

portion of the nucellus in vivo are apparently not present

or are inhibited in vitro. This is in contrast with the

initiation of Citrus nucellar cultures as reported by

Kobayashi et al. (1979), in which the morphologically

distinct primordiumm cells" of the adventive embryos

proliferate to form embryogenic ovular callus in

polyembryonic Citrus cultures.

Embryogenic mango nucellar callus is compact,

nonfriable and globular in form (Figure 3-8). It consists of

numerous light-dark brown globules (0.5-2.0 mm diameter)

with cream-white somatic embryos 2-5 mm in length arising

from the periphery. When this callus is transferred to

liquid medium (callus maintenance medium, Table 4-3), it

proliferates rapidly as globular callus (Figure 3-9).

Somatic embryogenesis from globular callus grown in

suspension is restricted primarily to the epidermal layer of

the globular callus or somatic embryos. The epidermis of

globular callus or somatic embryos prior to the initiation

of embryogenesis is a smooth and continuous layer of cells

completely surrounding the somatic embryo. The cells are

highly vacuolated without densely staining cytoplasm (Figure

3-10). Eventually the epidermis begins to redifferentiate

and is no longer a continuous single cell layer (Figure

3-11). Numerous embryonic initials, i.e., cells with densely

staining cytoplasm are visible. As dedifferentiation

proceeds the layer of densely staining cells with prominent

nuclei becomes thicker (Figure 3-12), causing the somatic

embryos to have a rough and bumpy appearance at low

magnification. The process of dedifferentiation of somatic

embryos from the epidermal region of a late globular stage

somatic embryos is shown in Figure 3-13. Numerous distinct

somatic embryos from the single cell to the octet stage of

embryo development can be observed. Subsequently, more

advanced stages of somatic embryo development become evident

(Figures 3-14 and 3-15). In Figure 3-15 a globular stage

somatic embryo with a distinct epidermis is shown.

The embryogenic 'Parris' suspension cultures used in

the regeneration experiments of Chapters IV and V have a

rather distinctive growth morphology. Proliferating globular

'Parris' callus in suspension appears very differentiated.

Cultures consist of clusters of somatic embryos 0.5-1.5 cm

in diameter which have formed via the budding of secondary

embryos (Figures 3-16 and 3-17). These clusters of budding

somatic embryos contain a distinct central core (Figure

3-18), which consists of a late globular to early torpedo

stage somatic embryo. Numerous somatic embryos at various

stages of development radiate from the outer portions of the

epidermis-derived cell layer. The somatic embryos which form

the central core are still composed of living tissue.

However, the nonembryogenic epidermis-derived regions of the

somatic embryos become necrotic (Figure 3-18). The necrotic

portions of the epidermis along with some of the underlying

cell layers are sloughed off (Figure 3-19).

Somatic embryos produced in suspension culture exhibit

a number of developmental abnormalities, some of which are

evident at very early stages (Figures 3-20 and 3-21).

Globular somatic embryos without a distinct epidermis are

prevalent. Moreover, their overall morphology appears to be

less compact than typical adventive embryos (Figures 3-5 and


Fewer developmental abnormalities occur when somatic

embryos are produced on solid medium even though

polycotyledony and unequal sized cotyledons were present

(Figure 3-22). After the differentiation of cotyledons on

embryogenesis medium (Table 4-3) the bipolar nature of the

regenerated somatic embryos together with their closed


vascular system was apparent (Figure 3-23). Thus, the nature

of regeneration via somatic embryogenesis was unequivocally


- ~- I




= (U



r --




04 *
ul > 0

3 4- *H
O1 e
-I U


0 *H
4. O 0
0 0::
00 -i 4

O- O
Oj f0 "

0(0 .00

' .4 *


< X!*- U



o 0 0 >- ,CO
.a)- (I)

Saa) u-H )
*> 0
04 M -W W > 04,

0 >) >,- 1 0
0 0 4 (0
0 a) -0 U 0

0 Z 4 r.
>iE-4 U'- 4 0
04 a1MU aE
O *C- 0) *
0 W 1 Oa)


I 0 04 4 ::1
-i 4 O ,~ E --

4-) a ) H 0 4-

0 4 r 0 4 ..
o jI ,C n U -I
4) r-I Z 4-)

.H (a 4 E E-
oU E 0c2E OW 0 *

4-)X- a Q)20)
--'0 O .C 0

0 Q04 > C 0 0
i- 8( 020 r- c-

0 u


ir" s::

.4J -

> -H o u
(a Eao-


u a) Z

aE)4 -

E- 4-4

r- Q)

0 -1 Ha
04' U2
m I

Z 44

0 0)

oH 4Q)

Ca) 0 zi
rn r; tr

-H E a) o
Z 4a)M

O~O a) a)l

k ca) -


(0 N
E -H

OO aim
4 -1 Z*
I-l I 1- e--I
4-4 .Q rl

0 U
4CU 00

0 -Pn -0
O (0 0-

r.0 -1


CDE 0 -

oi-i o --

4-i > E

C -
0 C) C U II

0>> O

ro 0 0
r- -I >M
0 C i-i --
- -H Q ) Qi ()

0) : 3: 0 0
rC '0 )
0 0 r C c

-1 04 U) -H C)





,Q -4

4 E
M -


U 04


> 0

4J E


(a 0C

4 0 (4







a) I




r-4 r-I

4J Q)

w m 0 z
>i 4J 4-4 H
0-C O
00 a0 4

W C( 0 -Hi

c,-1 0 O *H
E-- *H 4 0 U-

r' *- O 0 H
H e -H1

S-i 0 .1 -
r-4 4- 0 t

(0 r-HH

U C ( OO- E
3 *H Q. .C-
r C M-i *4
4H 0

uo NO

C) ro p =
Ot 4C roC ,
0 U ) E-
>i-0 20 4 *

I *-l *H 0
0Q U iU r -1C 44
3) Q) (- CH3 I
4C -H *

C0 0 ()
)C)U00 I
C-H r--l i-

C 0 G (0
4-ui r. 4 E-
m, r r4 0 WE

-Q) (01 H
0) U 04--1 U 3
0----U a)





0 0


Lo m

30 O
r-4 O
I-i 4-4 -4
(0 0 (


0 0 w
0 1 0 0O
: 3 r 0

0 0) 0

0 ,-I ,- C r

CP .a e
M -P (H

C 4-

0 Z

Ij M C-
0 0
o o o

>- 3 -H -H*

0)(0 0 E
4-4 cn0
0 a)

) '0 4-

M 0 4 -H

0 *4-H 0- 04-




4 4.
I.. -,

ia~~. IIJ*-









.rl 4




M 4Q)
(4-J 0 0

co *-i 0

o0 0

4- (n 0

'4H (a'-

0 -H
0 4-)


)Q) 4
a) E r

E-i rn 4




*H M

-P 4 >
0 r- *


U0 4-)

'a c
-( C

en -H

m0 0
4-) U4 r

(0 0

Q) C1H

0W *- -

E0 -H
rl 0 1 i

E ,Q 0 to
*-I- U

w > 0 w)

0) C






> 0C



EE1 *
m0 O
I w
e 0(
ro Ua

aI-1 0 I

0 ^ 0


M 04 -l
(11 Q)

Na) >






*HU r-q



>1 0
w u w 0
W 44 4JP

04 0 0 (
4 0 U) 4
a) >1 1) (a
04 4 0 4
a) 4j
(1) U
u -Wt
I, ao-H
h: 1"~ 41
0 E -) r
4 0 M (10

(1) a) ,
U) (A ra

'' ~~4


-4 0:)
a I

-H W-

0 0


0 0

>i E E
a) -


0 0.0




4J E

4 C>a

I ~ X4-4 r1
34-4 1l
Aby-H 4

-H 4

Cr 0
-H 04

At >1 m


4- -Hi


C ~3~I~i~le~Qg4~le~(0 41
r-4 (1

0 04





H 0 U)
E >-d


-u a)

M -4- Q)
.H4 4-4

r- o

-P Q),
Ou- aa)


1%0 0.~

-4 a)4
U a) 4

E E~


C12Z >.C


U U)

.I1 41'
ru M

U) U 0

Z' .I -

G-r) 24

-H q
44i fn M
O r.E-
Q) U)
w m-

a) 0

OrO 4

I OQ)e1)

440 ) >i4J

0 Q -H Vo
E m :$



U) U)
4 4i 3:
4-4 U MU 0) Hr


4 )U) C4-
-O 0 U) (H

U0- U (1) L4 E

z4~ -r1 02 1

>t -Hr-l Q,
**i$- 4JU)U)a~
Ev E 0a u

-:2 .. .. .0 '0 0 '0 C C
-HE -i-E J$

u a) I >
)O6 U ) 4

U) 4 4-d E 4.J

c ::3 0 m 0 (a,
0 (1) U E C

A1. 0 44

(1 -H (1 V (
u 4 Q) r

-1. r.4J
-H 0 Q) 4 t7
9 4 U 4 a) Wf
t" l O-l E ra C
(d Tj Q)Q

0 4 4-) Q) (1

U 04 ~
-Hr E Q) E

a 4 0a) Q

u U) r kRf =

~I Q) 4 rO 0 Elr O~~

~' .4 U 11l-

U) U) Z -H 4
0 ro:3 u r T



U1c O
Ua) U
04 E a)
O .Q

.4 <) o
-- O, e

s ri U

0 0 0 -H
U -H .t;
a) *1 u -I
u (0 ) 4-

Ov, 0 I
n 3t m
a) 0 U 0-1 4
,C >|-H k t
Q M (0 ) 4-41

OE~ 00 0
40 44 0 ) 0
0 U -* 0

Q) -H 0
4 w 0 i-
-H (n 13 0)M

aj co 0
0 10 0 U

OH >i5 U C
*H 03 I CD (

a 4CO OW

O4 0 0 EO
0 : 0 r0
4 00I 00 >i *-


*H 0

w) u

EnU) 0

E: 0 -H
o Z~ -r

;>I Q)


(U E Q)
E 0

-P Q)o

0f r- -

Q) 4Q

~'YY Lt~b~~ Lr ru; ILo
U) u
m 4 r- M


0) x >

SQ) 4J 0

Fc~rd CN1

&A 4




-H 0
0a) r

02 E

Lam 11

R2 H
0 -4
>I )
4 -i Q
.0 E O
m c 4
a) U

Er a
E sc4
0 4-4
k /I 6
4 (3)
a) 0 ;.,
0)4. -r-
m rl a

4 0
(a 0
04 4~

F CN -r


44 ~

Figure 3-22 A scanning electron micrograph of a cluster
of 'Parris' somatic embryos grown on solid
medium (bar=500 um).





~C. it.t
"'* ~
i:'~ir;i;$c r:i


u a)
-HI 4- 0

(al 0 aU
0 U) Q
O U X *
3 (o -2
) E

4 U0

-0 ( 4
O o

a) C 4- II
0Q) 0

.Q) 0Q
U) O4 ,

ra 0
0 e U

oa )

"0 10 (a
CT *H 2

H 0.0


4-I M

zOO 0
U0 0

Z 4 (0

0 0
.CO Q4

E :4-4
a) -
r0 41
*H $T3

c J ( 1
mli C. --

*^tn (



The efficient regeneration of plants via somatic

embryogenesis is partially dependent on the production of

large quantities of uniform, embryogenically competent

callus. This callus must then be induced by means of the

culture environment to undergo efficient and preferably

synchronized embryogenesis. Control of somatic embryo

maturation and germination is a final prerequisite for

normal plantlet formation.

In chapters IV and V a series of experiments are

described that addresses somatic embryogenesis and plant

recovery in mango. In Chapter IV, callus production and

somatic embryogenesis are considered. In Chapter V

parameters that affect the maturation and germination of the

somatic embryos are considered.

Callus production parameters have been quantified by

both biomass production and by measuring the embryogenic

competence of the callus produced. Somatic embryogenesis

parameters have been quantified by total cotylodonary embryo

production. Ratings have also been used to evaluate the

somatic embryos based on absence of necrotic or vitrified

regions, large relative size (5-8 mm in length), single

cotyledons without fasciations, and no secondary embryo


Materials and Methods

Immature mango fruits of both monoembryonic and

polyembryonic cultivars were harvested 2-3 months post-

pollination from the mango germplasm collections of the

University of Florida Tropical Research and Education Center

in Homestead and from the U.S.D.A. Subtropical Horticultural

Research Unit in Miami, Florida. Surface-sterilization of

the fruitlets was accomplished using the following protocol:

10 minutes in 70% ethanol, 30 minutes in a 1% sodium

hypochlorite solution with 2-3 drops of Tween 20 or Tween

80, followed by 2 rinses in sterile distilled water. Ovules

were dissected from the fruitlets using sterile forceps and

scapels. Embryogenic callus was initiated from whole

immature ovules, immature ovule halves without embryo(s),

zygotic embryos, nucellar embryos, and excised nucelli. All

callus was initiated on a modified solid Murashige and Skoog

(MS) medium consisting of 1/2 concentration major salts

(Table 4-1), MS minor salts (Table 4-2), 6% sucrose, 0.8%

Difco Bacto-agar, 2.74 mM (400 mg 1-1) glutamine, 0.55 mM

(100 mg 1-1) inositol, 0.57 mM (100 mg 1-1) ascorbic acid,

1.2 uM (0.4 mg 1-1) thiamine HC1, and 4.5 uM (1 mg 1-1)

2,4-D (callus initiation medium, Table 4-3). All media were

Table 4-1. Basal media major salts formulations
used with in vitro culture of mango.

Formulation mM(mg 1-1)

Major Salts MSz 1/2 MS WPMy B-5x

NH4NO3 20.6(1650) 10.3(825) 5.0(400)
KNO3 18.8(1900) 9.4(950) 25.0(2500)
Ca(N03)2 4H20 2.4(556)
K2SO4 -- (990)
MgSO4 7H20 1.5(370) 1.8(185) 1.5(370) 1.0(250)
CaC12 2H20 3.0(440) 1.5(220) 0.7(96)
KH2PO4 1.3(170) 0.6(85) 1.3(170)
NaH2PO4 H20 1.1(150)

ZMurashige and Skoog, 1962.
YWoody Plant Media, Lloyd and McCown, 1982.
Gamborg B-5, Gamborg, et al., 1968, without

Table 4-2. Murashige and Skoog minor salts formulation.

Minor saltsz uM(mg 1-1) Minor salts uM(mg 1-1)

Na2EDTA 100(37.3) KI 5.0(0.830)
FeSO4 7H20 100(27.8) Na2MoO4 2H20 1.0(0.250)
MnSO4 4H20 100(22.3) CoSO4 6H20 0.1(0.025)
ZnSO4 7H20 30(8.6) CuS04 5H20 0.1(0.025)
H3B03 100(6.2)

ZMurashige and Skoog, 1962.

adjusted to pH 5.75 using HC1, KOH, or NaOH prior to
autoclaving at 121 C and 1.1 kg cm for 18 minutes. Plant

growth regulators (PGRs) were dissolved in 1N HC1, KOH, NaOH

or 95% ethanol according to Table A-i.

Explants were cultured on sterile medium in 60 X 15 mm

plastic petri dishes, stored in clear plastic boxes and

incubated either in growth chambers at a constant
temperature of 25 C, with a 16 hr photoperiod (40 umol s
m ) provided by cool-white fluorescent lamps or at ambient

room temperature without supplemental light. Embryogenic

callus was established from both monoembryonic and

polyembryonic mango cultivars (Table A-2).

Original explants were subcultured onto fresh medium

several days after the initial culturing to remove the

tissues from oxidation products that had accumulated in the

media. They were subcultured again 1 week later and

subsequently at 3-4 week intervals.

One hundred to 500 mg of excised proliferating callus,

visible 1-3 months after culturing, was transferred to

fresh, solid callus initiation medium (Table 4-3) or

directly into liquid medium for multiplication and

maintenance. Liquid cultures consisted of 50 ml of basal

medium (BM) in 125 ml Erlenmeyer flasks maintained at 100

r.p.m. on rotary shakers. Liquid callus cultures were

subcultured at 2 week intervals until the fresh weight of

the callus was approximately 10 g per flask. Cultures were

0 0
0 0
P r (N






0 (




a 0





L U) dP
I S i

Ul c) dP
I S i





o n

CN 0


o Ln

CN 0


o Ln

N 0

0 *

( N <



0 0








z E

4- O

E Q)
0 2

E 4





0 0 0 *,

2: dp
3 3 dP (N
L dp
co m r- 0 r-
* 0 I *
0o 0 Ln

0 0 0 0


0 0 0 0 0

0 Co

3 3:


o *
m n

o o *


0 0
rO *

o *



> 0
- 02
r-4 U3



-r- I

>1 -
04 04
- I





0 0


0 0


I a

0 0
t I


EH -

0 0

U 02


W (N

0 0

-) r.

0 0

4 -4


then subdivided into 2 flasks of 5 g callus per flask with

fresh medium.

Experiment 4-1: In Vitro Somatic Embryogenesis from Mango
Nucellar Callus in Response to Genotype and Plant Growth

Calli from 10 mango cultivars (8 polyembryonic, 2

monoembryonic), initiated 3 months previously on solid media

(Table 4-3), were multiplied in liquid medium of the

composition. Treatments consisted of 200 mg callus

inoculated onto 12 ml of sterile BM in 60 X 15 mm plastic

petri dishes supplemented with 1 or more combinations of the

following PGRs, 2,4-D, NAA, IBA, or IAA at 0, 5, 10, or 20

uM and kinetin, 2iP, or BA at 0, 5, 20, or 80 uM. In

addition to the various synthetic PGRs, 3 complex organic

growth addenda were used, 20% (v/v) liquid coconut endosperm

(LCE), 0.015% (w/v) casein hydrolysate (CH), and 0.05% (w/v)

yeast extract (YE). All PGRs were prepared according to

Table A-i.

A modified 10 X 10 X 13 unreplicated factorial

experimental design was used, with 10 cultivars and 130 PGR

treatment combinations plus 6 treatment combinations with

complex organic addenda. Petri dishes were stacked in clear

plastic boxes and incubated in a growth chamber at a

constant temperature of 25 C, with a 16 hr photoperiod

provided by low intensity illumination (40 umol s- m )

from cool-white fluorescent lamps. The total number of

somatic cotyledonary embryos per dish was determined 2

months after inoculation.

Experiment 4-2: In Vitro Somatic Embryogenesis from 'Parris'
Nucellar Callus in Response to the Callus Maintenance
Medium and Major Salts Formulations

One-year-old embryogenic 'Parris' callus, initiated and

maintained on a solid modified MS medium (callus initiation

medium, Table 4-3) was transferred to liquid medium for

callus multiplication. Based on the results from Experiment

4-1 2 liquid callus maintenance media formulations were

tested for their effects on somatic embryogenesis. Callus

maintenance medium A (Table 4-3) was supplemented with 4.5

uM (1 mg 1-1) 2,4-D and 18.6 uM (4 mg 1-1) kinetin. Callus

maintenance medium B (Table 4-3) was supplemented with 4.5
-i -1
uM (1 mg 1-1) 2,4-D, 4.6 uM (1 mg 1-1) kinetin, and 2.05 mM

(300 mg 1-) glutamine. Both callus maintenance media

formulations contained the same BM consisting of: 1/2 MS

major salts, MS minor salts, 5% sucrose, 0.55 mM (100 mg

1-1) inositol, and 1.2 uM (0.4 mg 1-1) thiamine HC1.

Four BM major salts formulations MS, 1/2 MS, Lloyd

McCown Woody Plant Medium (WPM), and a modified Gamborg B-5

(B-5) were tested for their effect on somatic embryogenesis

(Table 4-1). All embryogenesis media contained: MS minor

salts, 5% sucrose, 0.75% Difco Bacto-agar, 2.05 mM (300 mg

1-1) glutamine, 0.55 mM (100 mg 1-1) inositol, 53.3 uM (4 mg

1-1) glycine, 3.0 uM (1 mg 1-1) thiamine HCI, 4.1 uM (0.5 mg

1-1) nicotinic acid, 2.4 uM (0.5 mg 11) pyridoxine HC1,

10.8 uM (2 mg 1-1) NAA, and 4.9 uM (1 mg 1-1) 2iP

(embryogenesis medium, Table 4-3).

Callus (300 mg/inoculum) was inoculated onto 12 ml of

solid sterile embryogenesis medium in 60 X 15 mm plastic

petri dishes. Cultures were incubated at room temperature

(24-26 C) in a clear plastic box with no supplemental light.

The experimental design was a 2 X 4 factorial with 6

replicates. Treatments were scored 45 days after inoculation

by counting the total number of somatic cotyledonary embryos


Experiment 4-3: In Vitro Somatic Embryogenesis from 'Parris'
Nucellar Callus in Response to the Callus Maintenance
Medium Formulation and Sucrose Concentration

One-year-old embryogenic 'Parris' callus, initiated and

maintained on a solid modified MS medium (callus initiation

medium, Table 4-3), was transferred to liquid medium for

callus multiplication. Two callus maintenance media

formulations, A and B (Table 4-3) were tested for their

effects on somatic embryogenesis.

Six sucrose concentrations: 0, 2, 3, 4, 5, and 6% were

also tested for their effect on somatic embryogenesis. Only

sucrose concentrations and callus maintenance medium

formulations were altered between treatments. All

embryogenesis media contained the same BM components

(embryogenesis medium, Table 4-3) and were solidified with

0.75% (w/v) Difco Bacto-agar.

Callus (300 mg/inoculum) was inoculated onto 12 ml of

solid sterile embryogenesis medium in 60 X 15 mm plastic

petri dishes. Cultures were incubated at room temperature

(24-26 C) in clear plastic boxes with no supplemental light.

The experimental design was a 2 X 6 factorial with 5

replicates. Treatments were scored 45 days after inoculation

by counting the total number of somatic cotyledonary


Experiment 4-4: Callus Production from Nonembryogenic
'James Saigon' Nucellar Callus in Response to the Major
Salts Formulation of the Callus Maintenance Medium

One-year-old nonembryogenic 'James Saigon' callus was

initiated and maintained on a solid modified MS medium

(callus initiation medium, Table 4-3). Callus was then

multiplied in liquid BM of the same formulation but without

any supplemental PGRs. Only major salts were altered between

treatments. All callus maintenance media contained MS minor

salts, 5% sucrose, 2.05 mM (300 mg 1-1) glutamine, 0.55 mM

(100 mg 1-1) inositol, 53.3 uM (4 mg 1-1) glycine, 3.0 uM (1

mg 1-1) thiamine HC1, 4.1 uM (0.5 mg 1-1) nicotinic acid,

and 2.4 uM (0.5 mg 1-) pyridoxin HC1 (callus maintenance

medium C, Table 4-3).

One gram of callus was used to inoculate 50 ml of

liquid callus maintenance medium in 125 ml Erlenmeyer

flasks. Cultures were incubated at room temperature (24-26

C) in a transparent plastic box without supplemental light

at 100 r.p.m. on rotary shakers. The callus was subcultured

every 2 weeks during the first month and every week during

the second month. The experiment was of a completely

randomized design with 4 treatments and 5 replicates.

Treatments were scored 60 days after inoculation by weighing

the total callus production (fresh weight) per flask.

Experiment 4-5: In Vitro Somatic Embryogenesis from
'James Saigon' Nucellar Callus in Response to the
Solidifing Agent Used in the Embryogenesis Medium

Two year old embryogenic 'James Saigon' callus that was

initiated on a solid modified MS medium (callus initiation

medium, Table 4-3) was maintained on a solid modified MS

medium. The solid callus maintenance medium consisted of 1/4

MS major salts, 1/2 MS minor salts, 2.5% sucrose, 1.0 mM

(150 mg 11) glutamine, 0.27 mM (50 mg 1-1) inositol, 100 uM

(27.8 mg 1-1) FeSO4 7H20, and 1.2 uM (0.4 mg 1-1) thiamine

HC1. Embryogenic callus was transferred to a liquid modified

B-5 medium for callus multiplication (callus maintenance

medium C, Table 4-3).

Two gelling agents, Difco Bacto-agar at 0.7% and

Gelrite gellan gum at 0.18%, were tested for their effect on

somatic embryogenesis. Only the gelling agent was altered

between treatments. All embryogenesis media contained the

same BM components [embryogenesis medium (without 2,4-D),

Table 4-3].

Callus (300 mg/inoculum) was inoculated onto 25 ml of

solid sterile embryogenesis medium in 100 X 15 mm plastic

petri dishes. Cultures were incubated at room temperature

(24-26 C) in clear plastic boxes with no supplemental light.

The experimental design was a completely randomized design

with 2 treatments and 8 replicates. Treatments were scored

45 days after inoculation by counting the total number of

somatic cotyledonary embryos.

Experiment 4-6: In Vitro Production of 'James Saigon'
Nucellar Callus in Response to a Liquid or a Solid
Maintenance Medium

Two-year-old embryogenic 'James Saigon' callus that was

initiated on a solid modified MS medium (callus initiation

medium, Table 4-3) was maintained on a solid modified MS

medium consisting of 1/4 MS major salts, 1/2 MS minor salts,

2.5% sucrose, 1.0 mM (150 mg 1-1) glutamine, 0.27 mM (50 mg

1-1) inositol, 100 uM (27.8 mg 1-1) FeSO4 7H20, and 1.2 uM

(0.4 mg 1-1) thiamine HC1. Embryogenic callus was

transferred to a liquid modified B-5 medium (callus

maintenance medium C, Table 4-3) for multiplication.

Two culture regimes, liquid and solid, were tested for

their effects on callus production. Two gelling agents,

Gelrite gellan gum at 0.2% and Sigma agar gum at 0.7% were

used to solidify the callus maintenance medium. All callus

maintenance media contained the same BM components (callus

maintenance medium C, Table 4-3).

Callus (300 mg/inoculum) was inoculated onto 12 ml of

solid sterile callus maintenance medium in 60 X 15 mm

plastic petri dishes or into 50 ml of sterile liquid BM in

125 ml Erlenmeyer flasks. Solid callus cultures were

incubated at room temperature (24-26 C) in clear plastic

boxes without supplemental light. Liquid callus cultures

were also maintained at room temperature without

supplemental light on a rotary shaker at 100 r.p.m.. Solid

cultures were subcultured twice at 3 week intrervals

supplemental light on a rotary shaker at 100 r.p.m.. Solid

cultures were subcultured twice at 3 week intervals, while

liquid shaker cultures were subcultured twice at the

beginning of the 4th and 6th week after inoculation. The

experimental design was a completely randomized design with

3 treatments and 10 replicates. Treatments were scored 60

days after inoculation by weighing the total callus

production (fresh weight) per flask.

Results and Discussion

Experiment 4-1: In Vitro Somatic Embryogenesis from Mango
Nucellar Callus in Response to Genotype and Plant
Growth Regulators

The results are shown in Tables 4-6 through 4-9 and A-3

through A-6. The total number of cotyledonary embryos

produced per plate (from 200 mg of inoculum on 12 ml of BM),

when averaged within genotypes over all 130 treatments

ranged from 31.0 ('James Saigon') to 0.0 ('Gadong')[Table

4-8]. Treatment responses were highly genotype-dependent

with no single treatment or treatment combination giving the

optimal response for all genotypes tested. 'James Saigon'

produced embryos on almost all treatment combinations

tested, even on treatments lacking PGRs (Table 4-4). Trends

In 000 0n ~ 0QL0 0f Ln
(N ,-l ("CN -4 .-.-iLn "N NNN

X in In 0oo
m in Ln Ln

0 o00 iln o
.1 i m r ru- in

umr~ \O~








4) i)l






0 oom 0oo OON
r r-N-h Lr r 1-4

o OOLA Lnoo
r- Ln N C m -

0 0
(N *'

0 000 000 000
M mmCN o em m nc

0 LAoo





r- Ln Or-4 r moo
un CN

+ +

m LoN Lo N o Lo Io N -o

+ +

+ + ++
S 0 mo a n o Lo in CO
m r ~r-.- (Nkoo (N

0 mo00 moo 0oo0 Lmoo
1-4 (N 1-4 CN -4 N r-A C%4

+ +

r-~ n




LA {r












C 0)

4-* -H
U)l 4.

rUo' 2

0~ 0
0 U

0 0)

4.) co
U O 4
EU 0

0 !->
M1 O

400 0

a-4) E

U) 0 a)
C 0) 0

,3 *-I > !
04 0 W

-1 U) e

U CW a)

o o 4
0 0o Q, 0
- 0

N >1C

M- +
N >i X