Dwarfism, self-incompatibility, and female sterility in Vaccinium ashei (Reade)


Material Information

Dwarfism, self-incompatibility, and female sterility in Vaccinium ashei (Reade)
Vaccinium ashei
Physical Description:
v, 143 leaves : ill. ; 28 cm.
Garvey, Edward Joseph, 1952-
Publication Date:


Subjects / Keywords:
Blueberries -- Breeding   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1985.
Includes bibliographical references (leaves 135-142).
Statement of Responsibility:
by Edward Joseph Garvey.
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 - 000567450
notis - ACZ3902
oclc - 14579336
System ID:

Full Text








Sincere appreciation is extended to Dr. Paul Lyrene, committee

chairman, for his guidance of this research, sharing of knowledge,

patience, and understanding. I also extend my appreciation to the

members of my graduate committee: Dr. Wayne Sherman, Dr. Gloria Moore,

Dr. David Knauft for their guidance and input. I also thank Dr. Timothy

Crocker for his input.

I thank Dr. Carlos Munoz who taught me many of the laboratory

techniques I required. I also thank David Norden who assisted in the

field and greenhouse, and Anne Harper who assisted in the greenhouse,

lab, and with the photographic work.

Finally, I express my appreciation for my wife, Angela, and

daughter, Maria Elisabeth, for their love and support.



ACKNOWLEDGMENTS............................. *...... .......* ... *ii

ABSTRACT.................................................. .......... iv

SECTION I INTRODUCTION.......................... .....................1


Dwarfism in plants...................................***** ******* 3
Self-incompatibility....... o .............................*.....14
Female sterility in plants ...............................*******.... 23

IN Vaccinium ashei (Reade) ..............................27

Expression of dwarfism in V. ashei
Materials and Methods........................... ... ....... 28
Results and Discussion..................................... ..33
Growth of V. ashei W78-66 in vitro
Materials and Methods................................ *********47
Results and Discussion. .... .......... ....................... 50
Inheritance of dwarfism in V. ashei
Materials and Methods.................................*******. 59
Results and Discussion....................................... 61

SECTION IV SELF-INCOMPATIBILITY IN Vaccinium ashei (Reade)..........73

Introduction.....................-..*****************. 73
Materials and Methods............. .... .. ....... .. ..***** ..******** 73
Results and Discussion......... ........ ......................... 76

SECTION V FEMALE STERILITY IN Vaccinium ashei (Reade)...............84

Introduction.................... .... ....... .....******************* 84
Materials and Methods............ o*.. *.... ..**.... .. .** ..****.84
Results and Discussion.. oo ................................... ..87

SECTION VI SUMMARY AND CONCLUSIONS.............................. 103

APPENDIX .... .... ............. ....... .... ... ..... ..************** ....106

LITERATURE CITED.................................... .......... .....135

BIOGRAPHICAL SKETCH................................................ 143

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





Chairman: Paul M. Lyrene
Major Department: Horticultural Science (Fruit Crops)

Dwarfism and female sterility found in 2 anomalous wild V. ashei

plants from north Florida and self-incompatibility in selected V. ashei

plants from north Florida and south Georgia were investigated.

A dwarf plant of V. ashei (W78-66) was found in a native population

of V. ashei in Nassau County, Florida. The plant had the normal

chromosome number for V. ashei (2n=6x=72), had normal meiosis and

produced fertile gametes. Stomate size and density were normal; leaf

area was smaller than for most V. ashei cultivars but was normal for V.

ashei plants. W78-66 had shorter internodes than cultivars and native

plants, which were due to fewer cells per internode. W78-66 responded

to GA both in vitro and in vivo. W78-66 grew well in vitro,

maintaining its relatively short stature and internode size.

Dwarf and normal-height plants were found in the progeny of W78-66

x tall V. ashei cultivars. Short-internode selections from this

progeny, when intercrossed, produced normal height, dwarf, and extreme

dwarf plants; when crossed with tall cultivars, they produced dwarf and

normal height plants; when backcrossed to W78-66, they produced normal-

height, dwarf, and extreme dwarf plants.

Self-compatibility measured as fruit set, number of seeds per

pollination, and number of seeds per fruit was highly variable among 19

native clones from north Florida and south Georgia. Self-fruitfulness

in all clones was insufficient to allow block plantings of single clones

in a production field.

The female-sterile V. ashei clone (W78-122) had extremely large

leaves compared with other native and cultivated plants of V. ashei.

Despite heavy flowering for 4 observed years, fruit set never occurred.

Meiosis appeared normal, with 2n=6x-72 chromosomes. Male fertility,

measured as percent pollen germination on an artificial medium and fruit

set of unrelated V. ashei cultivars pollinated with W78-122, was normal.

V. ashei pollen germinated on the pistil and grew through the style of

W78-122. Dissected ovaries from flowers at anthesis contained only a

few healthy-looking ovules. Cleared ovules from flowers at anthesis

indicated abnormal megagametophyte development. Female sterility ranged

between 11 and 25% in 3 of 4 populations of fertile V. ashei cultivars x

W78-122. All progeny of the fourth cross were female fertile.



The beginning of the cultivation of blueberries dates only to the

early part of this century. Prior to this time all blueberries were

harvested from native stands. The successful development of the

blueberry industry has been closely tied with Vaccinium breeding

programs (Galletta, 1975). Dr. F.V. Coville of the United States

Department of Agriculture directed the first such breeding program. He

released the 3 cultivars -- 'Pioneer', 'Cabot', and 'Katherine' in

1920, and these served as the basis for the early blueberry industry

(Coville, 1921).

All commercially-grown blueberries are derived from species within

Vaccinium section Cyanococcus (Galletta, 1975). Cultivated blueberries

are divided into 3 major groups: 1) native lowbush, 2) improved

highbush, and 3) improved rabbiteye.

V. ashei, the rabbiteye, is the tallest of the cultivated blueberry

species. As a native plant, it is adapted to the upland regions and

stream beds of the southeastern United States. Major populations are

found along the Satilla River in southeastern Georgia, the Suwannee

River in northern Florida, and the Yellow River in the Florida Panhandle

(Camp, 1945).

V. ashei is not restricted to these areas. In the 1920s, 1000

hectares of blueberry seedlings, mostly V. ashei, were collected from

native stands and planted in Florida (Darrow, 1966). These blueberry

plantations, now abandoned, contain an enormous amount of potentially

valuable genetic variability (Lyrene and Sherman, 1979).

Utilization of this germplasm in a breeding program requires an

understanding of the traits for which variablility exists and

information about their inheritance. It was the objective of this

research to 1) investigate the nature and inheritance of dwarfism and

female sterility, as found in 2 anomalous wild V. ashei plants from

north Florida, 2) determine the nature and degree of self-

incompatibility in selected V. ashei plants from north Florida and south




Dwarfism in Plants

Mature plant height is the end result of many interacting

biochemical and physiological pathways and is influenced by both

genotype and environment. Environmentally-induced dwarf plants are

common in areas such as mountain slopes, salt marshes, and dry regions.

Plants reach their maximum height in areas most conducive to plant

growth, such as the wet tropics, and are progressively shorter as the

latitude or altitude increases. An example of man's exploitation of

environmentally-induced dwarfing is Bonsai, in which pruning and

shaping, combined with growing in small containers, maintain a normally

large plant at a small size for periods up to century or more.

Genetic dwarfs are common throughout the plant kingdom.

Nonvigorous dwarfs may arise due to semilethal genes or incompatibility

within the genome or between the genome and cytoplasm. On the other

hand, genes exist in the gene pools of many plant species that can

produce vigorous but dwarf plants. There are basically 2 types of

vigorous dwarfs. In one type, called semidwarfism, only the internodes

are shorter than normal. Short-internode dwarfs can be extremely

important in agriculture and horticulture. The second type of dwarfism

could be called complete dwarfism, because it results in plants in which

all organs are smaller than normal.

Genetically dwarf plants may be dwarf because of 1) fewer cells due

to fewer cell divisions, 2) smaller cells due to less cell elongation or

early secondary cell wall thickening, 3) a combination of 1 and 2.

Bindloss (1942) reported that there were fewer cell divisions in the

stem of a dwarf Lycopersicon esculentum L. compared with the normals.

In the dwarf 4963 of Zea mays L., O'Donald reported fewer epidermal

cells in the coleoptile than in normal plants (Pelton, 1964). The

epidermal and coleoptile parenchyma cells averaged shorter in Dwarf-l of

Zea mays than in normals (Pelton, 1964). Shorter epidermal cells have

also been reported for the root of 'wilty dwarf' of Lycopersicon

esculentum (Lee, 1958) and for leaves of a dwarf Lolium perenne L.

(Cooper, 1958).

A combination of fewer cell divisions and less cell elongation has

been reported for various tissues of 'Dwarf-1' versus normal sibs in Zea

mays. The combination was reported in the rib meristem which produced

the internodal tissue (Abbe and Phinney, 1940), in the parenchyma

(Hansen and Abbe, 1943 as cited by Pelton, 1964), in the cortex (Hansen,

1957), and in the axis of the mesocotyl and parenchyma of the first leaf

sheath (Skjegstad, 1958 as cited by Pelton, 1964).

Precocious secondary cell wall thickening was found in the dwarf

columbine (Aguilegea vulgares L. compactta) (Pelton, 1964). Early cell

wall thickening inhibited cell elongation, resulting in a dwarf, compact

growth habit. Brittle stems and erect flower buds were pleiotropic

effects of the same gene.

Hormones have been shown to play a vital role in the control of

growth of the whole plant as well as of individual organs. Three

classes of growth-promoting hormones that have been investigated in


reference to dwarfs are auxins, cytokinins, and gibberellins (Thimann,


Auxin is required for cell elongation (Thimann, 1977). Internode

tissues which are elongating most rapidly have the highest level of

endogenous auxin (Wareing and Phillips, 1978). Treatment with

indoleacetic acid did not increase the growth of dwarf Tephrosia vogelli

Hook. (Irving and Freyre, 1960), Lycopersicon esculentum (Plummer and

Tomes, 1958) and Ipomea nil L. (Pelton, 1964). Phinney (1956) reported

that 7 single gene dwarfs of Zea mays did not respond to several

different auxins or hydrolyzed casein and coconut milk. However, dwarfs

may be either sensitive or insensitive to exogenous applications of

plant hormones. Reduced sensitivity to exogenously applied auxin was

reported in 'Dwarf-l' in Zea mays (Overbeek, 1938 as cited by Teas,

1957). High rates of auxin destruction in the plant were found to be

the reason for short stature. Low auxin sensitivity in this mutant also

explains its insensitivity to applied auxins.

Interaction of auxins and cytokinins was shown by Sachs and Thimann

(1967) to control apical dominance. Applied cytokinins are known to

stimulate growth of dormant axillary buds (Sachs and Thimann, 1964;

Williams and Stahly, 1968). Shootiness, which is characteristic of many

dwarf types, results because of the growth of many axillary buds, and

may be due to high levels of endogenous auxin in the axillary buds

(Wareing and Phillis, 1978). Auxin can direct the transport of

cytokinins from the root where they are synthesized.

Optimum shoot number was obtained when 3 strains of 'McIntosh'

apples (Malus spp.) were grown in vitro at a cytokinin level of between

3-6 uM 6-benzylamino purine (BA) (Lane and Looney, 1982). When 10 uM BA


was used, only the extremely compact strain 'McIntosh Wijcik' and the

compact 'Macspur' continued to proliferate more new shoots than the

standard varieties,and since they all had a similar cytokinin optimum in

vitro, dwarfness is probably not due to differences in endogenous levels

of cytokinin. The abilities of these compact apple clones to metabolize

excess levels of cytokinin or to counter the high cytokinin levels with

higher or lower levels of other hormones were given as possible reasons

for the continued shoot proliferation at high cytokinin levels (Lane and

Looney, 1982).

Gibberellins (GA) were first studied by Japanese botanists. A rice

disease caused by the fungus Gibberell fujikuroi Idn. produces plants

which are much taller and thinner than the uninfected plants. Yabuta

and Sumiki (1938 as cited by Stowe and Yamake, 1959) isolated two

crystalline, biologically active compounds which they named gibberellins

A and B. Today 52 different GA structures are known. It is now clear

that gibberellins are naturally-occurring compounds in higher plants.

Gibberellins are found in all parts of higher plants, and the major GA

in intact plants is GA .

Brian and Hemming (1955) reported that only a minute quantity of

gibberellic acid (GA) was required to increase the growth rate of dwarf

peas to that of a normal variety. Since that time numerous studies have

been conducted on the effects of GA on plant growth. The activity of GA

is not limited to height; it also increases leaf expansion, branch

angle, parthenocarpy, bolting of long day plants, and sometimes

eliminates vernalization requirements and removes light inhibition to

seed germination (Weaver, 1972). But, the effect on dwarf plants is

probably the most dramatic.

Hedden, MacMillan, and Phinney (1978) list 3 processes -- metabolic

control, deactivation, and compartmentization -- which can regulate GA

levels in higher plants. Light, temperature, and other hormones also

are known to affect GA levels (Weaver, 1972).

Metabolic control involves enzymatic regulation at one of the

branch points in the GA biosynthetic pathway. The GA pathway has a

common intermediate with the phytol and carotenoid pathway. Hedden and

Phinney (1967) correlated the absence of GA-like substances in the

'Dwarf-5' mutant of Zea mays with low rates of the GA intermediate

ent-kaurene synthesis.

The 2B hydroxylation of GA's is probably the most widespread type

of deactivation of gibberellic acid (Hedden, MacMillan, and Phinney,

1978). Rappaport et al. (1974) used radioactive GA's to establish

direct evidence for 2B-hydroxylation by higher plants. The reaction is

not known to be reversible and results in a marked reduction in

biological activity.

Only recently has it been shown that the complete GA biosynthetic

pathway operates in plastids (Hedden, MacMillan, and Phinney, 1978).

The phytochrome red/far red light reaction has been shown to affect the

level of extractable GA substances from plastids (Stoddart, 1976).

Compartmentization of GA synthesis and phytochrome-mediated release may

be an important method of GA regulation (Cooke and Kendick, 1976).

While GA's have been shown to affect a number of growth processes,

the effect on stem elongation has received the most attention. Brian

and Hemming (1955) reported differences in height between treated and

untreated dwarf peas (Pisium sativum L.) of 500% after I application of

GA The response was generally limited to younger tissues which were

still growing and could be influenced by both genotype and environment.

Increased cell length has usually been reported when anatomical

examinations were made after GA treatment. Kurosawa (1926 as cited by

Stowe and Yamaki, 1959) showed that with rice (Oryza sativa L.) plants

infected with G. fujikuroi that the epidermal and parenchymatous cells

of leaves and internodes were longer and slightly reduced in radial and

tangential diameters compared with cells in uninfected plants.

Increased cell divisions have also been reported after treatment with

GA's (Stowe and Yamake, 1959). Katsumi et al. (1964) stimulated

seedling elongation in the 2 maize dwarfs, 'Dwarf-5' and

'Anther-ear-one, by treating with GA precursors, but 'Dwarf-I' and

'Dwarf-2' were not stimulated by the same treatments. Presumably,

determination of the nature of a dwarf is possible by noting whether or

not the dwarf plant elongates when treated with compounds representing

different steps in the GA biosynthetic pathway.

Response to exogenous GA varies with dosage, environment, and

frequency of application (Pelton, 1964). Phinney (1956) found that the

degree of response was positively correlated with dosage in 4 dwarf

mutants of Zea mays. Daily applications of very small dosages of GA3 to

'Dwarf-l' and normal seedlings increased growth to the lethal point.

All plants died, and the stem length of the dwarf was similar to treated

normals (Pelton, 1964).

Insensitivity to exogenously-applied GA was shown for the 3 "major"

genes for reduced height in Triticum aestivum L.: Rht 1, Rht 2, and

Rht 3 (Radley, 1970). Radley reported that the tissues of some Norin-10

dwarfs contained increased levels of endogenous GA over some tall plants

and concluded that dwarfness was due to a block in the utilization of

GA, not a block in the production of GA.

Genetic dwarfism has been found in numerous plant families. Pelton

(1964) listed 112 dwarf forms in 34 genera of 17 families of

angiosperms. Den Ouden and Boon (1965) listed numerous genetic dwarfs

in many species of conifers.

Vavilov (1951) used dwarfism to exemplify homologous variation in

plants, and he predicted that dwarf forms could be found throughout the

plant kingdom. Although dwarfism has been found and successfully used

in many crop plants, the genetic pathways to dwarfism appear to be

numerous and diverse, and uncertainty still exists over the inheritance

of plant height. Polygenic, oligogenic, and monogenic control, both

dominant and recessive, have been listed, often for the same crop.

Interactions with other characteristics such as number and angle of

branches, vigor and genetic background, have been postulated as possible

reasons for unclear results from some inheritance studies. High

heterozygosity, large plant size, a long juvenility period, and, in some

cases, polyploidy, have limited genetic studies of dwarfism in woody


In hexaploid bread wheats (T. aestivum), polygenic control of

height was indicated by early studies (Freeman, 1919; Clark, 1924;

Torrie, 1936). Reduced height was brought about by new combinations of

already-present genes (Gale and Law, 1977). Two independent major genes

for dwarfing Rht-1 and Rht-2, were introduced through the Japanese

Variety Norin-10 (Fisk and Qualset, 1973). Another dwarf gene, Rht-3,

from variety 'Tom Thumb', was found to be allelic to Rht-1 (Fisk and

Qualset, 1973). Allen, Vogel, and Peterson (1968), showed that the

Norin-10 and Brevor-14 varieties carried 2 genes determining

semidwarfism. Yet, in some varieties, such as Suvon-92, homozygosity

for only 1 of these 2 genes was required to produce the semidwarf

phenotype. The 2 genes are additive in their effects and the alleles

determining dwarfism at each locus are recessive (Fisk and Qualset,


Crosses between Norin-10 semidwarf wheat and conventional lines did

not result in discrete F-2 classes as had been expected. The F-2

frequency distribution for plant height showed continuous variation.

This was true whether or not populations were segregating for the major

genes (Gale and Law, 1977). The removal of vernalization and

photoperiodic effects by growing progeny of 8 segregating wheat lines in

a greenhouse was necessary to reveal the expression of almost complete

dominance of tall plants (Halloran, 1975).

Polygenec control of height as well as "major" qualitative genes

exist in rice (Oryza sativa) (Foster and Rutger, 1978). A single

recessive gene for dwarfness from the variety 'Dee-geo-woo-gen' was

reported by Aquino and Jennings (1966). At least 3 nonallelic

semidwarfing genes--sdl, sd2, and sd4--have been induced by irradiation

in the rice cultivar 'Calrose' (Mackill and Rutger, 1979).

Legg and Collins (1982) reported that while the inheritance of

short internodes from a short-internode tobacco (Nicotiana tobaccum L.)

was monogenic, expression of the trait was highly affected by year,

maternal parent, cultivar background, and percentage of germplasm from

the mutant parent. The variation in expression was only found when the

gene was heterozygous. When homozygous, it was totally epistatic over

any quantitative genes, and all homozygous recessive plants were fully


Several types of dwarf plants have also been described in peaches,

Prunus persica L. Batsch. Hesse (1975) listed 2 types of dwarf trees:

Brachytic dwarfs (dw) and bushy (bul and bu2). Lammerts (1945) studied

the inheritance of both dwarf types. All of the F-I progeny of a cross

between the brachytic dwarf and normal plants were normal in height. In

the F-2 a clear 3:1 normal:dwarf segregation was obtained. The dwarf

parent and progeny had very short internodes, and at the end of 2 years

were approximately 1/5th the height of the normal progeny.

Analysis of tree types has been conducted with peach plants having

varying levels of the dw gene in different genetic backgrounds (Scorza,

1984). Phenotypic variations resulted from interactions of internode

and branch length, bud break, and branch angle. The homozygous

recessive (dw1dw1) was commercially undesirable due to excessive

dwarfism which resulted in plants being too short and the foliage too


The bushy genes (bul,bu2) in peach were discovered in a

segregating population from "close pollination" of the commercial

variety 'Babcock' (Lammerts, 1945). Of the 329 seedlings from the cross,

308 were normal tall plants and 21 were "somewhat brachytic, bushy

types." These plants had shorter internodes and the branches were

thicker than the tall types. Nonallelism of the 2 dwarf types was

determined by intercrossing.

Reduction in tree size in apples has been correlated with short

internodes (Brown, 1975; Lapins, 1976). Dwarfism was originally found

as a spontaneous mutant in some apple cultivars and has also been

induced be mutagens (Lapins, 1974, 1976). Zagaja and Faust (1983)

correlated tree size and vigor. Continuous variation was found in tree

size due to vigor, number of branches, and internode length. They

concluded that because of the observed continuous variation, the traits

were under polygenic control. Decourte (1967 as cited by Brown, 1975)

showed that dwarfness (non-spur type) is controlled by the single

recessive gene n and a number of cultivars are Nn. Polygenic control

has been reported for the spur type growth habit in 59 of 60 progenies

of apple (Lapins, 1976). In this report, one progeny from the cross

'Golden Delicious' x 'McIntosh Wijeik' segregated 1:1 spur to non-spur

type, which indicated that the spur type was segregating for a dominant

gene conditioning the spur phenotype.

Dwarfism has also been reported in Prunus avium L. (Fogle,1961),

Pyrus spp. L. (Tukey, 1964), Rubus ideaus L. (Keep, 1969), and Citrus

spp. L. (Cameron and Frost, 1968).

Dwarfism, like most traits in Vaccinium, has not been thoroughly

investigated as to mode of inheritance. No dwarf genes in any of the

Vaccinium species are reported in the literature. There are, however,

definite differences in height among species. Height ranges from the

low, less than 30 cm tall, rhizomatomous, lowbush types to the very tall

(2-5m) rabbiteye and highbush types (Eck and Childers, 1966). Natural

selection for adaptation to different environments would tend to "fix"

height within a population. Advantages of short stature include: winter

survival under snow cover, increased drought tolerance, greater fire

tolerance, and decreased vulnerability to wind. On the other hand, in

the habitats populated by highbush blueberry species, greater stature

allows the plants to compete more effectively for sunlight.

Crosses between highbush and lowbush species have been made by

several investigators (Johnston, 1946; Meader, Smith and Yeager, 1954;

Darrow, Morrow and Scott, 1952). Johnston (1946) found that the

lowbush growth habit was almost completely dominant in crosses between

lowbush and highbush types. He reported that 97% of the progeny were

lowbush types.

The inheritance of bush type in hybrids of highbush and lowbush

blueberries was studied by Meader, Smith, and Yeager (1954) in New

Hampshire. All F, plants were intermediate in height and definitely

taller than the lowbush parents. The F2 population after 13 years in

the field was grouped into 3 classes: 1) lowbush (30 cm or less); 2)

intermediate (31 cm to 1 m); and 3) highbush (1 m or more). Segregation

for bush type of 954 seedlings was 246 lowbush, 699 intermediate, and 9

highbush. While dominance of the lowbush type is not readily apparent,

neither the intermediate height of the FI nor the artificial grouping of

the F2 argue against the findings of Johnston.

Darrow, Morrow, and Scott (1952) reported on the characteristics of

highbush x lowbush blueberies progenies grown in New Jersey, Maryland,

and North Carolina. Most of the FI hybrid population ranged from 75 to

100 cm in height and were relatively uniform in height. The BC1 plants

(half-high backcrossed to highbush) had a height ranging from half-high

(75 to 100 cm) to 2 m. Plants of a small F2 population were grown and

found to be vigorous and no higher than the half-high F1 plants. At the

time of measurement, plants were fairly uniform in height but were

starting to show variability in spreading habit.

Dwarfism as expressed in the Vaccinium selection W78-66, may be a

product of: interspecific hybridization between native highbush and

lowbush species; an extreme segregate from a normal curve distribution

in "height" genes; a rare allele that became homozygous; or a new


dominant mutation. It was the objective of this research to gain an

understanding of: the origin of dwarfism as expressed in W78-66; the

nature of the dwarf trait; and how the trait is inherited.


In 1793, Cristian Sprengel wrote, "It seems that nature is

unwilling that any flower should be fertilized by its own pollen"

(translated by Lovell, 1918 p.12). Similar observations have

subsequently been made by numerous authors as detailed by many reviewers

of pollination biology (Baker, 1979; Frankel and Galun, 1977;

Nettancourt, 1977; Lewis, 1949; and Schmid, 1975).

Mechanisms designed to prevent inbreeding can be divided into 2

groups based on their time of action-- prefertilization and post-

fertilization. Prefertilization barriers consist of dichogamy, dioecy,

and self-incompatibility. Postfertilization barriers include death of

the zygote, reduced viability, and possibly sterility of the resultant


Self-incompatibility has been defined in various ways (Brewbaker,

1957; Arasu, 1968; Lewis, 1949; Lundqvist, 1964). The definition of

self-incompatibility in higher plants as "the inability of a fertile

hermaphrodite seed-plant to produce zygotes after self-pollination"

(Lundqvist, 1964 p.221) avoids several weaknesses of other definitions.

Nonfunctional male and female gametes as well as other pre- and

postfertilization barriers are distinguished from self-incompatibility.

The mechanism of the self-incompatibility system in flowering

plants is the blockage or restriction of incompatible pollen growth to

the ovule. Self-incompatibility has been divided into sporophytic and

gametophytic systems. In sporophytic systems pollen's compatibility

phenotype is determined by the genotype of the pollen-producing parent.

In ganmetophytic incompatibility systems the genotype of the individual

microspore determines the recognition reaction. Time of gene action,

site of expression, and presence or absence of floral polymorphism are

often different the 2 incompatibility systems. The incompatibility

reaction is expressed in the female part of the flower at the time of

flower opening in both sporophytic and gametophytic systems

(Nettancourt, 1977).

Sporopollenin deposited at the time of tapetal breakdown into the

cavities of the pollen exine is responsible for the sporophytic control

of self-incompatibility (Heslop-Harrison, 1975; Dickinson and Lewis,


The incompatibility reaction in the ganmetophytic system occurs

during pollen tube growth in the style. In the ganmetophytic

incompatibility system, abnormal behavior of the generative and

vegetative nuclei of incompatible pollen tubes (Nettancourt, 1977) was

reported. In incompatible pollen tubes, the generative nucleus did not

divide, and the vegetative nucleus disappeared within the first few

hours after germination. The vegetative nucleus was reported by

Nettancourt (1977) to provide the genetic information for the

incompatibility reaction in pollen tubes. Thickened pollen tube walls,

an increase in the number of callose plugs, high respiration rate,

increased acid phosphatase and cytochrome oxidase activity, and altered

protein patterns have all been reported in incompatible pollen tubes

(Nettancourt, 1977).

The stigma, style, and ovule have been listed as sites of the

incompatibility reaction in different plants. Brewbaker (1959)

discovered that the site of the reaction was closely associated with

pollen cytology. The stigma was listed as the site of reaction of

species with trinucleate pollen (pollen which has undergone the second

mitotic division prior to anther dehiscence). This included grasses

with homomorphic gametophytic incompatibility and many species with the

sporophytic system. Incompatibility reactions within the style or ovary

were reported for 36 genera of higher plants. Of these, 33 had

binucleate pollen (pollen which has undergone only 1 mitosis at the time

of dehiscence of the anther). The ovary as the site of the

incompatibility reaction was, until recently, only reported in Theobroma

cacao L. (Cope, 1962). It has since been reported to be the site of the

reaction in Vaccinium (El-Agamy, 1979) and in several tropical trees

(Bawa, 1979).

In sporophytic incompatibility systems, polymorphisms in flower

morphology which reinforce outbreeding have developed in a number of

species. Distyly and tristyly are 2 such systems. Species with distyly

have 2 different flower types; those with tristyly have 3 flower types.

Style and anther filament lengths differ in the flower, and pollinations

of like flower types are incompatible. Differences in flower morphology

are not the main barriers to self seed production in these plants, but

are associated with the SI system (Darwin 1878). Differences in flower

morphology are not associated with the gametophytic incompatibility


The genetic control of self-incompatibility in plants having the

gametophytic system may be due to 1, 2, or even 3 loci, each having

several to many alleles (Nettancourt, 1977). The one, multiallelic

locus (S) system was first reported in Nicotiana sanderae Erd. (East and

Mangelsdorf, 1925). Compatible pollination takes place only when a

given S-allele in the pollen is met with different S-alleles in the

pistil. The 2 multiallelic loci (S and Z) system was discovered by

Lundqvist (1954) in rye (Secale cereale L.) and subsequently found in

other species of the Gramineae. The two loci are inherited

independently but are epistatic. Allelic identity at both loci in the

genotypes of the pollen and the pistil leads to incompatibility, while

identity at only 1 locus leads to a compatible pollination. At least 3

loci (S) have been estimated to be active in Ranunculus acris Hook. and

possibly 4 in Beta vulgaris L.(Lundqvist, 1975).

The sporophytic self-incompatibility system is characterized as

having a single locus (S) with several alleles responsible for genetic

control (Nettancourt, 1977). Since the reaction of the pollen is

determined by the genotype of the sporophytic tissue in which it was

formed, the reaction in the pollen as well as the stigma is controlled

by two S-alleles. The 2 S-alleles may react independently or they may

interact by one being dominant over the other, and these relationships

may exist in 1, both or neither the pollen or stigma. In addition,

dominance and independence relationships of the S-alleles in the pollen

and in the pistil may differ.

Sporophytic self-incompatibility is characteristic of the

Cruciferae, Compositae, and Rubiaceae (Lewis, 1979). Gametophytic

self-incompatibility is characteristic of the Rosaceae, Leguminosae,

Onagraceae, Scrophulariaceae, Solanaceae, Ranunculaceae, and

Chenopodiaceae (Lewis, 1979).


Permanent and temporary elimination of self-incompatibility are

possible. Doubling the chromosome number, induction of self-fertility

mutations, and transfer of compatibility alleles into an incompatible

line can result in the permanent elimination of self-incompatibility

(Frankel and Galun, 1977).

Tetraploids of Trifolium, Nicotiana, Prunus, and Petunia lost the

self-incompatiblity that was present in the diploids (Nettancourt,

1969). The production of self-compatible pollen by tetraploids in

Oenothera organensis Hook. is limited to pollen grains that carry 2

different S-alleles (Lewis, 1947 and 1949). The compatibility of the

S-heterozygous alleles was confirmed by Lewis and Modlibowska (1942).

It is thought that competitive interaction between the different

S-alleles leads to loss in expressivity for each of the 2 alleles


Annerstedt and Lundqvist (1967) showed that polyploidization does

not result in the loss of self-incompatibility in monocots. They

suggested that genetic drift, selection, or an excess supply of

S-metabolites in the pollen grain have enabled monocot self-

incompatibility systems to tolerate gene duplication. Genetic

background of the pollen grain, gene dosage, and stylar interactions are

reasons given for the disturbance of the self-incompatibility reaction

in triploids and aneuploids (Nettancourt, 1977).

Induction of self-compatibility using radiation breeding has had

variable success. Both pollen and stylar mutations have been achieved

(Nettancourt, 1977). The self-compatible allele in the sweet cherry

'Stella' (Prunus avium) resulted from irradiation of a self-incompatible

cherry (Fogle, 1961).

Transfer of self-incompatibility by crossing a highly self-

incompatible cultivar with a compatible cultivar has become an

important part of hybrid seed production in some Brassica species

(Odland and Noll, 1950). Ronald and Asher (1975) transferred

compatibility from garden clones of Chrysanthemum x morifolium Ramat.

into incompatible greenhouse cultivars. Compatibility appeared to be


Variation in degree of self-incompatibility has been reported

within a number of plant species including some Vaccinium species

(Meader and Darrow, 1947; Aalders and Hall, 1961; El Agamy, 1979).

Variation in the activity or penetrance of the S-alleles and epistatic

reactions with fertility genes are listed as possible reasons (Frankel

and Galun, 1977). According to these authors, in most systems studied

thoroughly, genes were found which interfered with strict

incompatibility. The heritability and the type of genetic control of

intraspecific variation in self-incompatibility seemed to be quite


Pseudo-self-compatibility (PSC) is a temporary disruption in the

self-incompatibility system. It can be found in the pollen and/or the

pistil. Plants with PSC have a functional incompatibility system but

produce seeds from self- or incompatible cross-pollinations in amounts

ranging from less than 1% to 100% of a compatible cross. PSC is thought

to be due to segregation of modifying genes. (Henny and Ascher, 1976;

Litzow and Ascher, 1983).

Temporary bypassing of the incompatibility barrier has been

achieved by irradiation of pollen mother cells, bud pollination, delayed

pollination, heat treatment of the pistil, application of growth

regulators, maceration and surgical removal of the stigma, shortening of

the style, double pollination (pollinating with mixed killed-compatible

and live-incompatible pollen or pollinating with compatible pollen

followed several hours later by pollinating with incompatible pollen).

In addition, pollen has been injected directly into the ovary. In vitro

techniques have also been used to overcome self-incompatibility. One

technique has been to remove the ovules and place them in sterile

culture with the incompatible pollen (Nettancourt, 1977).

Early pollination studies in Vaccinium revealed the presence of

self-incompatibility. F.V. Coville (1921) stated that self-pollinated

blueberry flowers yielded berries which were "fewer, smaller, and later

in maturing" than cross-pollinated ones. Pollination experiments

conducted over 3 years on 15 highbush varieties showed that cross-

pollination usually increased the crop sufficiently to warrant inter-

planting of 2 or more varieties, even though fruit set was not always

reduced by selfing (Meader and Darrow, 1947). Also, mature cross-

pollinated berries contained a higher number of well-developed seeds and

tended to be larger than the fewer-seeded, self-pollinated berries.

Flowers of 24 clones of the lowbush blueberry ( V. angustifolium

Ait.) were self-pollinated in the field during 1958 and 1959 (Aalders

and Hall, 1961). Methods used to prevent cross-pollination were not

reported. Fruit set was obtained on only 1 clone which had only 8

fruit. Of the 21 lowbush clones tested in the greenhouse, 12 were

completely self-incompatible and 9 had varying degrees of fruit set.

Aalders and Hall concluded that the same pattern of self-fertility

reported in highbush blueberries (Morrow, 1943) existed in lowbush

blueberries. There are moderately self-fertile clones, slightly

self-fertile clones, and self-sterile clones. The average levels of

self-fertility in lowbush and highbush blueberry appeared to be very


The 10 rabbiteye varieties (Vaccinium ashei) tested in 1943 (Meader

and Darrow, 1944) were placed into 3 groups based on their degree of

fruit set after hand selfing in the greenhouse. Seven were considered

relatively self-infertile and had a fruit set between

0 and 11.6%. Two varieties set approximately one-third of the flowers

self-pollinated compared to about 70% when crossed. The variety

'Blueboy' set fruit equally well when selfed or crossed, averaging about

75% set. In general, days to maturity increased and berry size and seed

content were reduced with selling versus cross-pollination. Similar

findings were made with Florida highbush and rabbiteye varieties

(El-Agany et al., 1981; Hellman and Moore, 1983).

To determine the site of the incompatibility reaction in highbush

and rabbiteye clones, pollen tube growth in the pistil was monitored by

El-Agamy (1979). He found that self-pollination of highbush and

rabbiteye clones gave no significant difference in pollen germination

and growth through the style. There were no visual differences in

callose deposition on pollen tube walls in compatible and incompatible

matings. Also, pollen tubes entered the ovary cavity in both

pollination types, and in several cases, pollen tube growth into an

ovule occurred in normally incompatible matings. The rate of growth of

pollen tubes in the style was shown to vary with type of cross

(incompatible vs. compatible) and temperature in which the plant was

growing (Knight and Scott, 1964). Incompatible pollen tubes grew much

slower than compatible pollen tubes at both temperatures tested (high,

16-270 C and low, 8-24o C). At the high temperatures incompatible tubes

grew much faster than at lower temperature.

Despite the incompatibility barrier to self-pollination in

Vaccinium, new cultivar development has relied upon a very narrow

genetic base. Hancock and Siefker (1982) reported that coefficients of

inbreeding (F) ranged from 0.00 to 0.25 for the 63 highbush blueberry

cultivars released by public agencies in the United States. Three

native selections ('Brooks', 'Sooy', and 'Rubel') contributed most of

the genes in present-day cultivars. The inbreeding coeffients of most

rabbiteye blueberry cultivars is also considerable, and most of the

cultivated germplasm can be traced to the 4 landraces, 'Ethel', 'Myers',

'Clara', and 'Black Giant' (Lyrene, 1983).

Plant vigor (height, shoot diameter, and survival) was greatly

reduced after one generation of selling of 2 cultivars and 6 native

selections of V. ashei compared with outcrossed populations (Lyrene,

1983). Two highbush cultivars demonstrated a significant negative

correlation between mean seedling fresh weight and level of inbreeding

(Rellman and Moore, 1983).

The high degree of self-incompatibility in most present cultivars

of blueberries may be due to chance selection of high self-

incompatiblity in the foundation clones, or alternatively, it may

reflect a generally high level of self incompatibility in V. ashei. The

objectives of this research were to measure the degree of

self-incompatibility in V. ashei selections from native populations in

north Florida and south Georgia.

Female Sterility in Plants

Sterile plants lack functional male and/or female gametes. The

sterility may be due to chromosome aberration (inversions, trans-

locations, deletions), aneuploidy (2n+l, 2n-1, etc.), odd-ploidy (lx,

3x, 5x, etc.), chromosome incompatibility after wide-hybridization, or

genic factors (nuclear or cytoplasmic). Incomplete chromosome pairing

during meiotic prophase can result in uneven distribution of chromosomes

and the loss of genetic information in the gametes. These gametes are

often nonfunctional. Chromosome aberrations, aneuploidy, odd-ploidy,

and wide-hybridization can decrease chromosome pairing, and fertility is

often reduced.

Genic factors causing male and female sterility have been reported

for a number of crops. Mutants were found in Zea mays which suppress

the pistillate inflorescence (sk, ba) or convert the normally staminate

tassel to pistillate (ts). All 3 genes are expressed only when

homozygous (Frankel and Galun, 1977). In Rubus idaeus (Lewis and John,

1968), a normally hermaphroditic plant, genes suppressing stamen

differentiation (m) and carpel differentiation (f) were found. Both

genes are recessive and act independently.

Genic sterility factors have also been reported in bananas (Musa

spp. L.) (Simmonds, 1953), citrus (Citrus spp. L.) (Frost and Soost,

1968; Osawa, 1912 cited by Jackson, 1972), peppers (Capsicum spp. L.)

(Wilson et al. 1982; Kormos, 1954; Bergh and Lippert, 1965), alfalfa

(Medicago spp. L.) (Bingham and Hawkins-Pfeiffer, 1984), peas (Pisium

sativum L.) (Blixt, 1978), and tomatoes (Lycopersicon esculentum) (Rick

and Butler, 1956). Male sterility in plants is reported more often than

female sterility.

Female sterility may be due to lack of viable female gametophytes

or other necessary structures. Breakdown of the integument and the

nucellus and subsequent starvation of the fertilized embryo sac is

associated with female sterility in Vaccinium angustifolium (Hall et

al., 1966). Slow development of the integument was observed in 1 type

of female sterile alfalfa (Bingham and Hawkins-Pfeiffer, 1984). The

inner integument was absent and the outer integument did not completely

close over the nucellus. As a result, both the nucellus and the female

gametophyte were ruptured and destroyed. Childers (1960) reported that

in another female sterile alfalfa, 85% of the ovules lacked a female

gametophyte and the remainder had abnormal ganetophytes. In Capsicum

annuum L., 6 types of female sterility have been reported (Wilson et

al., 1982). Two were due to abnormal ovules, 2 had a deformed style, 1

lacked integuments, and in 1 the cause of sterility was not reported.

In most cases of genic female sterility, male fertility was reported as


Megasporogenesis and megagametophyte development in Vaccinium have

been studied by several researchers (Bell, 1957; Stushnoff and Palser,

1970; Edwards, 1970). Ovules are barely visible and appear as small

white mounds on the placentae during the dormant season previous to

flowering (Stushnoff and Palser, 1970). The megaspore mother cell

develops directly from the micropylar archesporial cell. There is no

parietal tissue and the megaspore mother cell is surrounded only by the

nucellar epidermis. Regular meiotic divisions of the megaspore mother

cell usually result in a linear tetrad, with the 3 micropylar megaspores

disentegrating, leaving the chalazal megaspore to function. This cell

divides 3 times to produce an 8 nucleated mature embryo sac which

corresponds to the Polygonum type (Maheshwari, 1950 cited by Stushnoff

and Palser, 1969).

Ovular tissue continues to develop concomitant with

megasporogenesis and megagametogenesis. The tissues surrounding the

ovule continue to bend until at the time of meiosis the ovule is

completely anatropous. A single integument elongates, closing over the

tip of the nucellus to form the micropyle. The nucellus is completely

broken down by the late 4-nucleate stage of the megagametophyte

(Stushnoff and Palser, 1970). Megagametogenesis was reported to be

completed at anthesis in 3 tetraploid advanced selections (Fla 4-15, Fla

4-71, and Fla 6-164) from the the Florida Vaccinium breeding program

(Edwards, 1970).

The use of female sterility to increase hybrid seed production has

been proposed in alfalfa (Childers, 1960), and peppers (Bergh and

Lippert, 1965). Flower production declines in many plants as fruit set

and develop, while in the "styleless" mutant of pepper, new flowers are

produced for a longer period of time, providing a more continuous source

of pollen for cross-pollination (Bergh and Lippert, 1965). Mixed

planting of male and female sterile plants would allow harvest of all

fruit without danger of getting selfed seed.

In all reported cases of female sterility in pepper, the

malfunction has resulted from single recessive genes that are

pleiotropic (Bergh and Lippert, 1965). Female sterility in alfalfa was

due to a single recessive trait, tetrasomically inherited (Bingham and

Hawkins-Pfeiffer, 1984). Hall et al. (1966) reported that multiple

genetic factors may determine female sterility in Vaccinium



Male and female sterility have been reported in lowbush blueberries

(Hall et al., 1966; Hall and Aalders, 1961). It was initially

hypothesized that female sterility as expressed in the Vaccinium

selection W78-122 could have been due to: chromosome aberrations,

aneuploidy, odd-ploidy, chromosome incompatibility after wide

hybridization, or genic factors affecting floral and fruit development.

It was the objective of this research to determine the nature of the

female sterility trait in W78-122 and its inheritance.




The rabbiteye blueberry (Vaccinium ashei) is the tallest of the

commercially-grown Vaccinium species. Three to 6 meters is the normal

height for this species, but plants 9 meters tall have been reported

(Eck and Childers, 1966). Differences in average height exist among the

widely scattered populations of rabbiteye blueberries. West Florida

populations are normally tall (about 5 to 6 meters), while south Georgia

populations average 3 to 4 meters in height (Eck and Childers, 1966).

Natural selection for adaptation to different environments would

tend to "fix" height within a population. Advantages of short stature

include winter survival under snow cover for northern Vaccinium species,

increased drought tolerance, greater fire tolerance, and decreased

exposure to wind. The major disadvantage would be the inability of the

short plant to compete with larger neighboring plants for sunlight.

Long intervals between breeding cycles and large plant size are 2

limitations to efficient and rapid improvement in many perennial woody

plants. Dwarfism has been proposed as a way of reducing these problems

(Hanche and Beres, 1980). Dwarf plants are often precocious, with

shorter germination-to-flowering intervals. The use of dwarfing genes

could allow more rapid progress in a breeding program by making possible

more generations in a set amount of time. With dwarfs, more plants can


be grown per unit area. This allows the breeder to screen more plants,

which increases the probability that superior plants will be found.

The benefits of dwarfism are not limited to the breeder; they also

extend to the grower. Shorter intervals between planting and the first

harvest mean a quicker return on initial investment. Estimated

economic returns from standard size and genetically-dwarfed peach trees

after 7 years' growth in California were compared by Hanche and Beres

(1980). They reported that even though the initial investment was

greater for the dwarf trees, the investment could be recovered 2-3 years

earlier with dwarf trees than with standard size trees.

The objectives of the research described in this section were to

1) describe a short-statured, compact-growing V. ashei clone (W78-66) in

its native environment and at the Horticultural Research Unit (HRU) near

Gainesville, Fl., 2) determine the nature of its dwarfism, and 3) study

the inheritance of this trait.

Expression of dwarfism in V. ashei W78-66.

Materials and Methods

Phenotype in its native environment

The clone W78-66 was found by Dr. Paul Lyrene on June 27, 1978.

The plant was located at the edge of a predominantly pine forest facing

SR 121, 11.8 kilometers north of Highway 90, northwest of Baldwin,

Florida (Nassau County). Softwood cuttings and seeds were collected

from the plant in July, 1978. Cuttings were rooted under mist and

planted into the Vaccinium Germplasm Block (VGB) at the Horticulture

Researh Unit (HRU) near Gainesville, Florida. The open-pollinated seeds

were germinated the following winter, and seedlings were planted into a

high density nursery at the HRU during the spring of 1979.

On March 6, 1985, W78-66 was studied in its natural habitat with 2

objectives in mind: 1) to describe the phenotype of the original dwarf;

and 2) to see if the dwarf was qualitatively different from other V.

ashei plants in the population or was merely an extreme variant from a

normal-curve distribution. An abandoned blueberry plantation located

near the town of Glen St. Mary's, Florida, was also studied. This

plantation was probably planted between 1920 and 1940 using V. ashei

from the western Florida panhandle (Lyrene, 1979), but many of the

plants at the site were seedling progeny of the original plants. This

site was selected for comparison of plant types. Measurements collected

were 1) plant height, 2) colony diameter measured at 1 meter height, 3)

number of stalks per colony, 4) trunk diameter, and 5) type of colony


Phenotype at the HRU

Two ranets of W78-66 were planted 20 cm apart in the VGB at the HRU

near Gainesville, Florida in 1979. They have since grown together and

were treated as one plant. These ramets were compared with wild V.

ashei clones which were also planted in the VGB in 1979 and with 8

cultivars of V. ashei growing close by. The VGB includes clones of wild

V. ashei from throughout north Florida and southeast Georgia. These

clones were selected because of their good horticultural

characteristics. Internode measurements were collected for W78-66, for

13 other members of the VGB, and for 8 cultivars of V. ashei growing in

close proximity.

Internode lengths were determined for each plant by dividing the

measured length of a stem by the number of internodes making up the

stem. Ten stems were measured; all measurements were made on the third

growth flush from the apex. This flush of growth was usually a light

brownish-red color, signifying a moderate amount of lignification.

Chromosome count

Chromosome counts were made from mitotic cells of the apical

meristem. Shoots from forced vegetative buds were collected when less

than 1 cm long, fixed in 3:1 absolute ethanol:glacial acetic acid and

held in the freezer until used (at least 24 hr). The outer leaves of

the shoot tip were removed to expose the apical meristem. The apical

meristem and a small piece of stem tissue were placed in 25% HCL for 20

min, and transferred to a pectinase-cellulysin solution [.03g pectinase

plus .03g cellulysin (Calbiochem, San Deigo, CA) in 2 ml distilled

water] for about 15 min. The tissue was squashed in 1% acetocarmine,

destined in 45% acetic acid, and observed under phase contrast at


Stomate size and density

Stomate size and number per unit area were measured on W78-66, 3

rabbiteye cultivars, 2 F-1 plants from the cross W78-66 x 6x V. ashei

cultivar composite, and an extreme dwarf V. ashei (81-44) from the HRU.

This extreme dwarf was a seedling from a V. ashei cultivar polycross in

the rabbiteye breeding program. Leaf imprints were taken by applying

fingernail polish to the underside of a newly mature leaf from each

plant. The imprints were peeled off the leaves, placed on microscope

slides, and viewed under 250X magnification. Stomate length was

measured using an ocular micrometer. Thirty stomates were measured for

each leaf imprint. Stomate density was measured by counting the number

of stomates in 10 different fields of view at 250X magnification.

Leaf area

Leaf size was measured on W78-66, 6 other native V. ashei clones,

and on 4 rabbiteye cultivars to see how W78-66 differed from the other

clones. Ten newly mature leaves were randomly taken to represent all

sides of each plant. Two ramets of each cultivar and 1 ramet from each

clone in the VGB were used with the exception of W78-66 where 2 plants

were used. All plants were grown in an open field under full sunlight.

Leaf size was measured using a LiCor Portable Area Meter Model LI 3000.

Cell length and number

Internode measurements were collected for W78-66, for an extreme

dwarf F2 segregate from the cross W78-66 x V. ashei cultivar, for a

plant of normal height from the same cross, for another native V. ashei

plant, and for 2 cultivars of V. ashei. Stem segments from the second

oldest growth flush were collected. Internode measurements were made, a

1 cm segment of the internode was dissected using a Lancer Model 1300

Vibrotome, and longitudinal sections were placed on a microscope slide

in glycerine. Cell number per segment and cell size were determined for

each plant by counting the number of parenchyma cells in a column of

cells and computing the average diameter. Cell size and number of cells

per internode were subsequently calculated for each plant.

Gibberellic acid in vivo

The objective of this experiment was to determine whether the

dwarfism expressed by W78-66 is due to the absence or to low levels of

endogenous GA. Softwood cuttings of W78-66 and the V. ashei cultivar

'Tifblue' were rooted under mist during the fall of 1983. Rooted

cuttings were potted into 15 cm pots January, 1984. Soil mix was 1:1

Canadian peat moss:sand v/v. On March 21, 1984, 12 plants of each clone

were placed in full sun. All treatments were initiated at this time.

GA at 0, 100, 250, and 500 mg/liter in deionized water and 4 drops of

the surfactant "Tween" was applied to 3 plants of each clone. A

pressurized sprayer was used and all solutions were applied to the point

of runoff. A second application of GA3 at the same levels was applied

to the same plants in the same manner on June 21, 1984.

On August 15, 1984 plants were pruned, fertilized with Osmocote

18-6-12 at 42g per 15 cm pot, and placed in a greenhouse. The

photoperiod was artificially lengthened using 1:1 fluorescent:

incandescent lights to provide approximately 100 lux at plant level.

Lights were set to provide a 4-hr night break between 10:00 PM and 2:00

AM. Plants were sprayed 3 times with GA3 at 2-week intervals. GA3

levels were 0, 500, 1000, and 2000 mg/liter. There were 3 plants per

treatment. Plants remained in the greenhouse until December 17, 1984

when internode length of new growth on all plants was measured. Length

of the new growth and number of nodes per new shoot were collected for


Results and Discussion

Phenotype in its native habitat

The selection W78-66 was found not to be unique. Many V. ashei

plants were observed and 13 plants were measured at the Baldwin site and

13 more at the Glen St Mary's site (Table 3-1). Three plants at the

Baldwin site were found to have a very similar phenotype to W78-66. The

dwarf phenotype was characterized by the combination of several traits.

It was usually less than 2m tall and was extremely colonial, having many

small-diameter shoots arising perpendicular to the ground. Each of the

4 dwarf types had a colony diameter of about 6m, with W78-66 having a

colony diameter of 8m. The nondwarf types were usually 3m tall or

taller. The canopy diameter was 1-2 meters; the colonies had less than

20 shoots with the largest 3 being greater than 2cm in diameter. In

addition, most shoots arose from a narrow crown and angled outward.

While both the dwarf and tall colony types were found at both

sites, there were differences in colony characteristics. The nondwarf

types were much taller at the Glen St. Mary's site than at the Baldwin

site. The dwarf colony types at the Glen St. Mary's site were more open

and did not form the dense colonies that were found at the Baldwin site.

While it was helpful to divide the clones into dwarf and nondwarf

types, this division was arbitrary. Intermediate types were found at

both sites, and they showed variable expression in all the traits


Phenotype at the HRU

Compared to the standard V. ashei phenotype, W78-66 is a

short-internode dwarf with a distinctive growth habit (Figure 3-1). It

Table 3-1.

Characteristics of
northeast Florida.

27 V. ashei clones from 2 populations in

Height of Colony Diameter of Base
tallest stalks diameter Shootsz largest stalks N=narrowy
(m) (m) (no.) (cm) W-wide

Glen Saint Mary's Sitex


Dwarf type
Dwarf type
Dwarf type

W78-66 Site


W Dwarf
W Dwarf
W W78-66
W Dwarf




ZNumber of shoots

arising from either the base of the plant or from the

YShoots arising from a very narrow crown (N) or arising perpendicularly
from the ground up to 1.3 m from the center of the colony (W).

xPlants were descendents of a plantation made at Glen St. Mary's
Nursery, probably before 1940. Original plants are believed to be
from native populations from the western Florida panhandle.



Figure 3-1.

Phenotype of W78-66 and a normal height V. ashei
plant growing at the Horticulture Research Unit,
Gainesville, FL. 1-a. W78-66 (dwarf). 1-b.
W78-40 (normal height). 1-c. W78-66, typical
shoot. 1-d. W78-40, typical shoot.

is very vigorous and highly fertile, both when open pollinated and when

hand-crossed in the greenhouse with V. ashei cultivars. It has dark

green foliage and rather small leaves which are semievergreen. The

fruit are small and black, the plant very compact and spreading. As

with other V. ashei, the terminal 4-6 nodes of the growing shoots die

back at the end of each of several growth flushes the plant undergoes

each year. In W78-66, shoots are concentrated in the upper third of the

stem. Of these shoots, 2-3 are relatively large in diameter. The

shoots form wide angles with the main stem. In addition, the shoots are

of unequal length; the most apical shoots are shortest and those

arising more basally get progressively longer. Consequently, all shoots

from this main stem extend to approximately the same height. The main

stem tends to grow outward and the axillary shoots grow straight which

gives the plant a "layered" appearance.

The growth habits of all the cultivars and most of the other

members of the VGB are quite different from the dwarf. The main stems

tend to be upright. Lateral shoots arise from about the 4th node from

the terminal and are evenly located the entire length of the stem. Most

shoots are about the same length and have a small diameter. This gives

the plants an upright, open appearance. Within the VGB there are

intermediate types having one or several characteristics of W78-66 yet

having longer internodes and one or more non-dwarf traits.

A possible reason for the differences in phenotype between W78-66

in its native habitat and at the HRU is environmental. In its native

environment the plant was growing in the understory of tall pines, while

the HRU plants are in full sun. The dwarf, in its native habitat,

appears to be about 15 years old from seed. The ramets at the HRU were

started as rooted cuttings and are only about 6 years old. In both

places, the dwarf appears different from most other V. ashei plants

growing nearby. The dwarf at the HRU may take on the phenotype of the

original plant as it gets older.

Internode length was shortest for W78-66 compared with both the

cultivars and native V. ashei (Table 3-2). Cultivars tended to have

longer internodes than native plants, but there were exceptions.

Chromosome count

Approximately 72 chromosomes were counted in each of 25 somatic

cells of W78-66 undergoing mitosis, indicating that W78-66 is probably

hexaploid with 2n=6x=72.

Stomate size and density

Stomate size and density varied significantly among plants, but

both W78-66 and the extreme dwarf 81-44 were within the normal range of

V. ashei (Table 3-3).

Leaf area

The average surface area per leaf was significantly smaller for

W78-66 than for all of the other V. ashei cultivars measured (Table

3-4). The dwarf also had the smallest average leaf surface area of the

7 wild V. ashei clones measured in the VGB, although it was not

significantly smaller in leaf area than 2 of the wild clones. Area per

leaf was quite variable, though cultivars usually had larger leaf areas

than unimproved clones. There were exceptions, especially with W78-122,

which was originally selected from the woods because of its large leaves

and associated female sterility.

Table 3-2.

Internode lengths of 13 cultivars and 8 native V.
ashei clones growing at the HRU in Gainesville.

Internode Minimum Maximum
length lengthz lengthz
Clone Type (cm) SD (cm) (cm)

Aliceblue Cultivar 1.3ay .2 1.1 1.7
Bluegem 1.3a .2 .9 1.6
Beckyblue 1.3a .2 1.0 1.6
Tifblue 1.2ab .2 1.0 1.6
Bonita l.lb .1 .9 1.3
Climax l.lb .1 .9 1.3
Chaucer I.lb .1 .9 1.2
Delite 1.lb .1 .9 1.3

W78-66 Wild .6e .1 .5 .8
W78-41B 1.3a .2 1.1 1.9
W78-39 1.2ab .3 .9 1.9
W78-122 1.2ab .2 .9 1.4
Beulah 4 l.lb .2 .8 1.2
W78-37 l.lb .1 .8 1.3
W78-68 l.Obc .2 .8 1.3
W78-75 1.Obc .2 .8 1.3
W78-70 .9cd .1 .8 1.1
Beulah 2 .9cd .1 .8 1.0
W78-69 .9cd .1 .6 1.1
Beulah 1 .9cd .1 .7 1.0
W78-64 .8d .1 .7 1.1

zLowest or highest of 10 stems sampled per clone.

YMean separation by Duncan's Multiple Range Test, 5% level.

Table 3-3.

Stomate size and density in V. ashei grown in full sun at
the Horticulture Research Unit, Gainesville, Fl.

Stomatal pore length Stomatf
Plant Type (u) per mm

Bluebelle tall cultivar 45az 260ab
Aliceblue tall cultivar 37b 274a
Bonita tall cultivar 39b 270a
W78-66 dwarf 38b 250b
82-165 (tall cv x dwarf, F ) 41ab 241b
82-194 (tall cv x dwarf, F ) 39b 253b
81-44 extreme dwarf 38b 254b

ZMean separation within columns by Duncan's Multiple Range Test, 5%

Table 3-4. Average area per leaf of natives and cultivars of V. ashei.

Plant Mean Minimum Maximum
Plant Type (no.) (cm2) SD (cm2) (cm2)

Bluebelle Cultivar 20 10.7bz 1.7 7.8 13.9
Climax 20 10.6b 3.0 6.6 16.3
Fla K 20 16.9a 2.2 13.5 20.9
Beckyblue 20 12.Ob 3.0 9.1 18.1
W78-66 Wild 10 5.3c 1.4 3.0 6.5

W78-122 Wild 10 25.6az 5.3 19.2 36.9
W78-35 10 11.8b 2.6 8.3 15.8
W78-39 10 8.4c 1.4 5.1 10.3
W78-74 10 8.2c 1.4 5.6 10.2
W78-37 10 6.8cd 1.2 4.4 8.2
W79-34 10 6.0cd .8 4.0 6.7
W78-66 10 5.3d 1.4 3.0 6.9

ZMean separation within columns within groups
Test, 5% level.

by Duncan's Multiple Range

Small leaf size in W78-66 may be due to a pleiotropic effect or

close linkage with the gene or genes responsible for the short stature.

But, since small leaf size is not unique to the dwarf and is found in

several other native V. ashei, a more interesting question may be why

the cultivars have such large leaves.

Cell length and number

In general, short-internode plants had smaller pith parenchyma

cells than long-internode plants (Table 3-5 and Figure 3-2). The 2

dwarf types (W78-66, 84-172) had the smallest cells and the shortest

internodes. While there were minor differences in cell size between

the dwarf and nondwarf types, differences between the 2 types in cell

number per internode were very large, indicating that the

short-internode trait in W78-66 is due predominantly to fewer cells per

internode compared with nondwarf plants.

A seedling from a cross between 2 wild V. ashei clones (84-171),

was growing in the same test plot as the extreme dwarf F-2 segregate

(84-172) from the cross W78-66 x tall cultivar, and was included as a

check plant to compare with 84-172. This plant was selected as a

representative of nondwarf types. Cell size and internode length of

this nondwarf were approximately twice the size of the extremedwarf,

while number of cells per internode was about the same for the 2 plants.

This indicates that differences in internode length are due to

differences in cell size more than cell number in this situation.

Table 3-5.

Internode length, cell length, and cell number per internode
in dwarf versus nondwarf V. ashei.

Internode Cell Cells per
Plant Type length diameter internode
(cm) (Ui) (no.)

W78-66 dwarf .7 22 290
84-172 dwarf selection .5 18 260
Tifblue tall cultivar 1.5 27 530
Bluebelle tall cultivar 1.5 29 520
W78-122 tall selection 1.5 29 520
84-171 tall selection .9 33 280

~c 17~ cj

Figure 3-2.

Pith parenchyma cells in internodal regions from
newly mature stem segments from V. ashei W78-66
(top) and 'Tifblue' (bottom).

Gibberellic acid in vivo

Differences in phenotype were observed between the dwarf and

nondwarf clones and within treatment levels of the dwarf. The new

shoots of both clones had light green leaves. The oldest leaves of the

new shoots were normal size and color, but the newest leaves were

unusually small and chlorotic. Tip necrosis was present on most new

shoots of plants which were treated with GA3. A high degree of necrosis

of the new growth of the nondwarf cultivar 'Tifblue' did not allow

length of new shoots or number of nodes of new shoots to be recorded.

Internode length was determined as the mean internode length of 4

continuous internode regions in the mid-portion of the new stem. This

measurement was collected for both clones (Table 3-6). Internode length

of both clones was increased by exogenously applied GA3. Differences

between genotypes were significant. The "Dwarf" fit a cubic equation

only slightly better than a linear regression indicating that the effect

of GA3 conc. on internode length was approximately linear in the range


Length of new shoots, number of nodes per shoot, and internode

lengths were measured on W78-66. All parameters mentioned increased

with the increasing levels of GA3 (Table 3-7).

Stem and internode elongation in response to GA3 application has

been studied by many researchers (see Pelton, 1964). Chlorotic leaves

have also been reported after GA3 treatment. This is believed to be due

to failure of chlorophyll synthesis to keep pace with increased cell

expansion-not to gibberellin interference with chlorophyll metabolism

(Pelton, 1964). Plant necrosis was reported (Pelton, 1962 cited by

Pelton, 1964) after daily applications of GA Tip necrosis of the

treated 'Tifblue' plants in the present study may be a similar response.

Table 3-6. Internode length of dwarf
exogenously applied GA3.

and nondwarf types as affected by

Stem Internode length
Clone GA3 level segments (cm)
(no.) Mean SD

0 mg/liter 30 1.1 .81
500 24 1.6 .50
1000 10 1.3 .88
2000 10 1.7 .71

0 mg/liter 20 1.3 .43
500 19 1.4 1.10
1000 20 1.6 .87
2000 19 1.9 1.40

Linear **
GA level
Linear **
Quadratic ns
Cubic ns
Linear *

z**, Significant at .01 and .05

level respectively, and ns=not

Table 3-7. rew growth of W78-66 as affected by level of GA3.

Length of new Internode
shoots Nodes length
GA3 level (cm) (no.) (cm)

0 mg/liter 8.1 11.0 1.09
500 12.7 15.0 1.13
1000 15.6 18.0 1.21
2000 20.5 21.0 1.50

Linear ** ** **
Quadratic ns ns ns
Cubic ns ns us

ZMean of 4 internodes located in the middle of the new shoot, not the
total length of the new shoots/number of nodes.

**, significant at .01 and .05 level respectively, ns-not

The fact that internode length of the dwarf increased to levels

found in the nondwarf after treatment with GA3 suggests that W78-66 is

dwarf due either to insufficient levels of active GA being produced by

the plant or high levels of active GA deactivation.

Growth of Vaccinium ashei W78-66 in vitro.

Materials and Methods for Experiments 1-4

Experiment 1

The objectives of this experiment were to: 1) select seedlings of

the populations dwarf V. ashei 'W78-66' selfed and V. ashei 'Beckyblue'

x 'Premier' that grew well in vitro. 2) clone the vigorous seedlings

from each cross to be used in later experiments. On June 27, 1983 seeds

from each cross were surface-sterilized with 30% Clorox (1.6% sodium

hypochlorite) for 15 minutes, rinsed twice in sterile, deionized water

and placed in vials containing 10ml of a 0.55% agar gel. The agar was

autoclaved at 1.05 Kg/cm2 for 15 minutes.

Ten seeds were placed in each vial, and 50 vials were used per

cross. The vials were placed in front of a window to receive indirect

sunlight 9 hrs per day. Cold requirements were met by placing the vials

in an unlit refrigerator (7C) 15 hrs each day for 30 days.

On September 27, 1983 vigorous seedlings from each cross were

removed from the vials and cut into two-node cuttings. Cuttings were

placed into vials containing 10 ml of modified Anderson's Rhododendron

Medium (ARM) (Table 3-8) which had been previously autoclaved as

described for the agar water. Vials were placed on a bench where they

Table 3-8.

Composition of Modified Anderson's Rhododendron Medium




CaCl2 4H20

MgSO 7H120

NaH2PO4 H2(


FeSO4 H120

MnSO4 H20

ZnSO4 H20

Na2MoO H 2

CuSO4 5H 20


CoCl2 6H20

Thiamine HC1





Adenine Sulfate

Casein Hydrolysate

























pH adjusted to 5.7 with IN NaOH.

Autoclaved at 1.05 Kg/cm2 for 15

received approximately 1 klux light provided by cool white fluorescent

tubes for 15 hrs per day and were in the dark the other 9 hrs.

Internode lengths were measured from 5 clones of each cross on

March 5, 1984. Five stems from each clone were measured. Two-node

cuttings from each clone were then subcultured onto fresh modified ARM.

Shoot number per vial, height of the tallest shoot, and internode

measurements were collected for 3 clones from each cross on August 23,

1984. Clones were then subcultured by placing 2-node cuttings in vials

of modified ARM.

Experiment 2

The objective of this experiment was to determine the effect of

exogenously-applied GA on internode lengths of the dwarf and nondwarf

clones growing in vitro. The experiment was initiated August 27, 1984.

Two clones from each of the previously described populations were used.

The 4 levels of GA3 used were 1.0, 5.0, 10.0, and 100.0 mg/liter.

Two-node cuttings of each genotype were placed in vials containing 10

ml of modified ARM amended with the GA The GA3 amended media was

autoclaved as before, after it was dispensed into the vials. There were

10 replicates per treatment. Vials received approximately 1 klux

fluorescent light 16 hrs daily and were kept at about 25*C. Colony

size, height of the tallest shoot, and number of shoots per colony were

collected October 31, 1984.

Experiment 3

The objective of this experiment was to determine if W78-66 was

more tolerant than normal plants to higher than optimal levels of the

cytokinin 21P (6-gamma-gamma-dimethyl-allyl amino purine). Two clones

from the dwarf and 2 from the normal cultures developed in Experiment 2

were used, with 10 replicates per clone. Four levels of 21P were used:

5, 10, 20, and 40 mg/liter. One two-node cutting was placed in each

vial. Vials were given approximately 1 klux fluorescent light for 16

hours/day. Colony size, height of tallest shoot, and shoot number were

recorded for all surviving vials after 2 months.

Experiment 4

The objective of this experiment was to compare the in vitro growth

of W78-66 seedlings and normal V. ashei on a medium containing low

levels of cytokinin. Two clones from the dwarf, 2 from the normal

populations, and 3 levels of 21P (0.0, 0.1, and 1.0 mg/1) were used.

Two-node cuttings were placed in vials containing 10 ml of modified ARM

with the appropriate level of 21P. There were 8 replicates per

treatment. Plants were treated as in other experiments. On March 21,

1985 height of tallest shoot, number of shoots, colony diameter, and

internode length were recorded.

Results and Discussion

Experiment 1

Germination had begun in both populations by July 15, 1983, less

than 3 weeks after sowing. After 2 months, most seeds from the cross V.

ashei 'Beckyblue' x 'Premier' had germinated and produced tall, vigorous

seedlings with a well-branched root system and large internode lengths.

Seedlings from W78-66 selfed were much shorter than from the crossed

population; the shoots were vigorous with many short internodes, but

they often fell over due to having a weak root system. Germination

percentage was much reduced compared with seed from the 'Beckyblue' x

'Premier' cross. From approximately 100 seedlings, 7 seedlings with 3

cotyledons and 2 with variegated white and green cotyledons were found

from this population.

Inbreeding depression may account for much of the poor germination

with the dwarf selfed seeds. Selfing could also account for the 2 types

of anomalous seedlings: tricotyledons and variegated cotyledons. Selfed

seeds were used despite the inherent problems in order to obtain

seedlings with the maximum expression of the dwarf trait.

Vigorous seedlings which appeared to have short internodes were

selected from the dwarf selfed population for subculturing (Table 3-9).

Seedlings of the other cross were not selected for internode length, but

the ones used had internode lengths at least twice the size of the dwarf

population (Figure 3-3).

Shoot number did not differ significantly among clones of the

different populations (Table 3-10), but average height and internode

length of all the shoots in a vial were about half as great in the dwarf

populations as in the normal population (Table 3-10).

Experiment 2

Height increased significantly with increasing levels of GA3 in the

dwarf clone grown in vitro (Table 3-11). The nondwarf clone did not

differ in height among the different levels of GA3 (Table 3-11).

Surprisingly, axillary shoot development increased significantly with

increasing levels of GA3 with both plant types. Cytokinins and their

Table 3-9.

Internode length of 5 V. ashei dwarf clones (W78-66 selfed)
and 5 normal V. ashei clones (Beckyblue x Premier) in vitro.

Clone W78-66 (self ed) Beckyblue X Premier
(cm) (cm)
Mean SD Mean SD

1 .0la .08 .30az .03
2 .12a .03 .46a .12
3 .12a .02 .34a .08
4 .15a .04 .23a .02
5 .15a .01 .21a .02

X = .13b .03 .30a .11

zMean separation within columns by Duncan's Multiple Range Test, 5%

Table 3-10. Growth of dwarf and control seedlings of V. ashei in vitro.

Shoots per Height of Internode
vial tallest length
Clone (no.) shoot (cm) (cm)

W78-66 (selfed)
Clone 2
vial 1 27 az 2.Obz .16bz
vial 2 19a 2.5b .17b
Clone 3
vial 1 27a 3.Ob .16b
Clone 6
vial 1 23a 3.7b .16b

X = 24a 2.8b .16b
Beckyblue x Premier
Clone 1
vial 1 25a 6.0a .42a
vial 2 19a 5.5a .33a
Clone 4
vial 1 26a 4.0ab .25a
vial 2 29a 3.5b .28a

X = 25a 4.8a .32a

ZMean separation within columns

by Duncan's Multiple Range Test, 5%

t ext

Figure 3-3. Growth of W78-66 selfed (right) and 'Bluebelle x
'Premier' (left) after 3 months in vitro.

Table 3-11.

Growth of dwarf (W78-66 selfed)
'Premier') V. ashei clones in

and nondwarf ('Beckyblue' x
vitro as affected by GA3

Dwarf Nondwarf
GA3 Plants Height Shoot Plants Height Shoot
(no.) (cm) (no.) (no.) (cm) (no.)

0 mg/liter 8 .75 30 10 4.2 6.5
1 9 .89 17 5 3.5 11.0
5 9 1.26 28 7 3.3 10.0
10 5 1.48 36 7 3.3 16.0
100 8 1.58 65 9 3.6 33.0

Linear ** ** ** **
GA Level
Linear ** ** ** **
Quadratic ns ns ns ns
Cubic type ns ns ns ns

z**, Significant

at .01 and .05 level respectively, ns=not

interaction with auxins are reported to stimulate axillary bud

development in vitro (Murashige, 1974), while gibberellins are reported

to repress axillary bud development. Axillary shoot development in

potato can to be controlled by manipulation of levels of auxin,

gibberellin, and cytokinin (Kumar and Wareing, 1972 as cited in Wareing

and Phillips, 1978).

Abnormal growth of the dwarf and nondwarf plants in vitro was

observed only at the 2 highest levels of GA3. The effect was mainly on

leaf appearance. Leaves from plants receiving the 2 highest levels of

GA3 (10 and 100 mg/liter) were sometimes long and narrow. This leaf

type was found with both dwarf and normal genotypes.

Experiment 3

Height, shoot number per vial, and colony diameter were neither

significantly different between the dwarf and nondwarf, nor did they

change significantly with increasing levels of cytokinin (Table 3-12).

Few replications due to high plant loss at the high cytokinin levels and

large variability among replications are the probable reasons for the

lack of significance. Trends seemed to be indicated, but this

experiment was not sufficiently precise to show them clearly. Height

tended to decrease with higher levels of 21P in both dwarf and nondwarf

plants. Shoot number and colony size appeared greatest with 10 mg/liter

21P in the dwarf clones and decreased at higher levels. The nondwarf

had its highest shoot number but lowest colony diameter at 20 mg/liter


Cytokinin levels above 10 mg/liter appear to be detrimental to

overall growth of both plant types. Based on this experience selection

Table 3-12.

Growth of dwarf (W78-66 selfed) and nondwarf ('Beckyblue' x
'Premier') V. ashei clones in vitro as affected by high
cytokinin levels.

Vials Survival Height Shoot Colony size
21P level (no.) (Z) (cm) (no.) (cm)

5 mg/liter 15 .75 1.1 15 1.9
10 5 .25 1.1 40 2.4
20 3 .15 .9 27 1.3
40 2 .10 .6 13 .3

5 10 .50 2.3 36 2.3
10 5 .25 1.1 38 1.7
20 4 .20 1.0 47 1.4
40 0 .00 -- -- --
Linear ns ns ns
2IP level
Linear ns ns ns
Quadratic ns ns ns
Cubic ns ns ns

zWidth x depth.

Yns=not significant.

for dwarfism using high levels of cytokinin (Lane and Looney, 1982)

would not be an efficient method with the type of dwarfism exemplified

by W78-66.

Experiment 4

The effect of cytokinin levels was not statistically different

between the dwarf and nondwarf types in any of the parameters measured

(Table 3-13). Plant height and shoot number had linear relationships

with cytokinin levels. Height decreased and shoot number increased as

cytokinin level increased up to 1.0 mg/liter 21P. The colony diameter

and internode length relationship was curvilinear. Internode length

initially decreased and the number of shoots increased as cytokinin

level increased. Colony size initially increased and then decreased

with increasing cytokinin. While the pattern of growth was similar, the

dwarf was shorter and had shorter internodes than the nondwarf.

Consequently, dwarfness as exhibited in the W78-66 is probably not due

to differences in endogenous cytokinin levels nor is the dwarf more

sensitive to exogenously-applied cytokinin levels.

Inheritance of dwarfism in V. ashei W78-66.


The objective of this research was to observe how the dwarfism in

V. ashei W78-66 is inherited.

Table 3-13.

Growth in vitro of dwarf (W78-66 selfed) and nondwarf
('Beckyblue' x 'Premier') V. ashei clones as affected by
low cytokinin levels.

21P Colony Internode
level Vials Height Shoots size length
Genotype (mg/liter) (no.) (cm) (no.) (cm) (cm)





21P level








length x width.

Y**, Significant

at .01 and .05 level, respectively, ns=not

Materials and Methods

During the spring of 1979 cross-pollinations were made by Dr. P.M.

Lyrene, between the dwarf V. ashei (W78-66) and commercial cultivars of

V. ashei. In preparation for crossing, plants were placed in a

refrigerated room at 4C for 30 days after being dug from the field to

insure that the cold requirement had been met. The dwarf plant was used

as the female parent, and pollen from 4 V. ashei cultivars was bulked

and applied to depetaled flowers before anthesis in the greenhouse.

Berries were collected as they ripened and seed was extracted using a

food blender. Dried seed was stored at 4C for 5 months.

The seed was sown on moist, sterilized peat moss in the greenhouse

on November 7, 1980 and most germinated within 120 days. The seedlings

were transferred to flats of peat moss. After 4 months the plants were

transplanted from the greenhouse to a high-density field nursery. Plant

spacing was 25 cm between rows and 12 cm within the row. Plants were

fertilized and irrigated as required.

W78-66 was self-pollinated in the greenhouse during the spring of

1980. The resulting seeds and seedlings were treated as before.

Plants were selected from the F, population December, 1983 for use

in further crosses. Selected plants had relatively short internodes

compared with the rest of the population. Cross-pollinations were made

between the short internode F 's; backcrosses to both parents were made

and W78-66 was self-pollinated (Table 3-14). Berries were harvested

when ripe. Ten berries from each plant were weighed and number of plump

seeds per berry were counted. Seeds were removed from the remaining

berries, and all seeds were dried and treated as before. Seedlings were

planted into a high density nursery May 23, 1984.


Table 3-14. Pollination records, Spring 1983.

Fruit dev. Fruit Fruit Plump seed
Pollination Pollinations Harvest period harvested set per fruit
Population period (no.) duration (days)z (no.) (%) (no.)

(W78-66 x tall cultivar) x (W78-66 x tall cultivar) (Both parents selected for short
82-197 x 82-200 2/22-3/17 189 5/25-7/25 102 66 35 16
82-200 x 82-199 3/01-3/17 89 5/16-7/25 81 89 100 21
82-199 x 82-166 3/01-3/08 80 7/25 147 1 3 2
82-166 x 82-201 2/07-3/21 215 5/16-7/10 93 36 17 9
82-201 x 82-197 2/22-3/17 143 5/20-6/28 99 30 21 23
82-157 x 82-158 2/08-2/21 185 5/11-5/16 87 3 2 31
82-158 x 82-160 2/10-2/21 289 5/01-6/20 88 97 33 40
82-160 x 82-161 2/12-3/01 172 5/16-5/25 91 7 4 26
82-161 x 82-156 2/12-3/16 275 5/16-7/05 99 224 81 40
82-156 x 82-157 2/12-2/19 250 4/26-6/04 80 137 55 33

Tall cultivar x (W78-66 x tall cultivar) (Pollen parent selected for short internodes)

Aliceblue x 82-167 2/28-3/16 55 5/11-7/05 79 55 100 59
Bonita x 82-201 2/09-2/28 457 4/30-6/08 82 59 13 48
Bluebelle x 82-166 2/14-3/03 442 5/16-7/25 100 367 75 50

(W78-66 x tall cultivar) x W78-66 (Female parent selected for short internodes)

82-194 x W78-66 2/24-3/22 165 5/16-7/19 91 156 98 33
82-195 x W78-66 2/27-3/25 176 5/25-7/10 94 160 90 13
82-196 x W78-66 2/27-3/25 155 6/01-7/25 100 94 62 29

Tall cultivar x W78-66

Bonita x W78-66 3/29-4/04 422 4/30-6/08 82 8 2 33
Aliceblue x W78-66 2/24-2/27 55 5/01-6/01 86 43 78 32

W78-66 selfed 2/24-3-11 565 6/01-7/19 91 178 31 28

SInterval between 50% pollination

and 50% fruit harvest.

Height, stem diameter, and internode length were measured on the

1980, F1 population and the 1981, selfed population during September

1983. Height, number of shoots per plant, and internode length were

measured on the 1983 populations during September, 1984. Plant height

was measured as the height of the tallest stem. Internode length was

measured on newly mature wood of the tallest shoot averaged over 4-5

internod es.

For comparison, measurements were collected on 2 nondwarf hybrid V.

ashei populations and on populations obtained by selling 2 tall V. ashei

cultivars. Six nondwarf hybrid populations were also measured in the

1983 high-density nursery.

Results and Discussion

The F, population from W78-66 x 6x V. ashei cultivar composite

averaged significantly shorter and had shorter internodes than one of

the tall V. ashei x tall V. ashei populations (79-18 x 79 QP), but had

about the sane height and internode length as the other (79-17 x 79 EP).

(Table 3-15). Stem diameter did not differ significantly in the 3


The W78-66 selfed population averaged significantly smaller in all

parameters measured than both check populations ('Climax' x 'Beckyblue'

and 'Climax' x NC1830) and the other selfed population ('Premier'

self ed).

The 3 hybrid populations ('Climax' x 'Beckyblue', 'Climax' x

NC1830, and 'Premier' x V. ashei Late Composite) were compared with the

2 selfed populations (W78-66 selfed and 'Premier' selfed) to distinguish

the effects of crossing vs selling vs dwarfism. Height was

8. .
oo a oo> f Mo o oen
,- 1 -1 ,-1 1 -4 ,1 14 4
0 ,III II I II 0
be u4>o 0000u O

nNN V V 9 w i

0 .. 0 0

,-4 0
a {{0 4c o t g4o oa a
"- -g 0o CM r4 0 r4 Cn 5n

o o n .o m
-t ^U C~ 0 C' (' 0? h'
a 3 4
o r- r oo 0 o(0-i4 *- 0
. .
.-4 ~~ ~ ~ b r--~J- -4.4~ O 4 0
E"*0 ^ I I I 3
be 000 000
m r l Cg 4 1

U 3 0$
9 0 ^

aa 4 0 N o a
0 w 0

c *i** *o OA ,1 a -^ o
oo 0 .-0 0 V o 0
o o- 4 a o -a
54 4 -14 -4 0U r

00 0 N NOo 0% : .4 0
-'-4c O~4-z4c l ,-4 J 0 4%D -
I- -4---4-.- 4 4 4 4
SI I I I I I I I g 0

4- 0C)CO % -4 o Dw nIna wu w
0 ,-4f-4.T4-4 -4 I U i 4 0
> *3 *0 > 0.
-- o o< oa- rr4 0 a
0 a M" (nt Ln c14 ce) o % ^-n $4< toi
u C1 C*('4 -tC; CC0'4 en
cc N-4 N 4 er-i-4-4 N 0

0 X0 0. 0- x 0) 0. 4.4 1 V m

U. U0El4
s ~i "w 2 ag >
bo 1S. n Cptr C4 o w C4 14 aA x w D
9-4 b*O | b o a
U3 l, 00 cC7%r- co 0*hfen,-% C.4C, 3
O* 0^-U2 C'4 C'4C) C ^-4r-4lINC U 0 ,J U r

sp4 4 wy I- v

boft >.. : I >, t I 1 l ~ r- 4 Aj 0 010 0
0 N 0
o 4 .n en e n- A0
4 g 0 4 A 4 o
0o o U0 i r4 *
0 N l tSN CqrQ ((n 4 M -:T cncM N W0
01.4 w -4r
U44 P U -3 w- -
S24 Ai wC
04 0 x 0 .4
0 > >0 0 O

to w ag a ^o'aa a n ,
>% 4 .14 0-
*N o 44 W a% U

m~me uwchn O' pz m u m 3 3> a
t" r- 4 0 a
0W U C4-0 4D
I z- 10 1414 w .4 O

10 0 a4-
Oo w. 0"o0 0 W0 4n
'.I' .D I -4 00 V4 '.O-,4 %. 0w 0. 0 0
-4~ 9- 4-4 00 0 0 -4 U
.0a o II -,--'W .5c 8o 0) o^o 0
0 ONS ON- w -44 u r W U E -. Z Q
S 3r-.r-.~ f3 N > 4>

significantly less in the W78-66 selfed population than in the 'Premier'

selfed which was less than in the 3 crossed populations. These

measurements indicated that both inbreeding depression and short-stature

genes from W78-66 contributed to the dwarfism of the W78-66 selfed

progeny. Internode length was reduced in the W78-66 selfed population

compared with the 2 crossed populations ('Climax' x 'Beckyblue' and

'Climax' x NC1830). Inbreeding depression did not greatly affect

internode length based on comparisons among 2-year old populations

(Table 3-15). Therefore the reduction in internode length in the 1-year

old W78-66 selfed population compared to the 2 crossed populations was

probably due mainly to dwarfism.

Stem diameter was much smaller in all selfed populations than in

hybrid populations. Stem diameter did not differ statistically between

the 2 selfed populations (W78-66 selfed and 'Premier' selfed) of the

same age. Reduced stem diameter of the W78-66 selfed population

probably measures inbreeding depression, not effects of dwarfism.

The standard deviations for height and internode lengths were

relatively low for the W78-66 selfed population compared to the other

populations (Table 3-15) which suggests that W78-66 is not segregating

for a major gene affecting height and internode length.

High correlations were found between stem diameter and height in

most populations (Table 3-16). A low, nonsignificant correlation was

measured in the 'Climax' x NC 1830 population, which may be due to the

interspecific hybrid nature of the NC 1830 parent (V. constablaei x V.

ashei 'Premier').

Significant correlations between height and internode length were

seen only in one hybrid population (79-18 x 79QP and in the W78-66-

Table 3-16.

Correlations between height, stem diameter, and internode
length of different V. ashei populations measured
September, 1983.

Stem Stem
diameter diameter Height
Plants Height Internode Internode
Population Type (no.) Age length length

W78-66 x 6x Cv Comp 1 117 2yrs .77** .llns .14ns
79-17 x 79 EP 2 33 .67** .21ns .24ns
79-18 x 79 OP 2 33 .40** .18ns .46**
Premier selfed 3 107 "

Climax x Beckyblue 2 33 lyr .47** .17ns .16ns
Climax x NC1830 2 33 .17ns -.23ns .02ns
W78-66 selfed 4 114 .55** .21ns .30**
Premier selfed 3 114 .32**
Premier x Late Comp 2 82 .87**

Zl=dwarf x tall; 2= tall x tall; 3= tall selfed; 4= dwarf selfed.

** Significant at .01 and .05 level respectively, ns=not significant.

selfed population. Height of a plant is determined by the number of

nodes on the stem and the internode length.

Plant height measurements in the 1984 high density nursery were

greatest in populations not having W78-66 in their pedigree and least in

the (W78-66 x tall cultivar) x W78-66 population (Table 3-17). The one

exception was in the cross (79-69 x 83-17), in which a new dwarf

phenotype was found to be segregating, with about 15% of the seedlings

distinctly dwarfed. The parents of this cross were both tall wild V.

ashei clones, one of which came from Baldwin, FL (78-69) and the other

from south Georgia (83-17).

The populations having the most consanguinity with W78-66 had, in

general more shoots per plant, but there were some exceptions to this

trend. Shoot number and plant height were interrelated and the nature

of this association was different in the dwarf and nondwarf populations.

This association is discussed further in a later section.

Populations with W78-66 in their pedigrees generally had shorter

internodes than populations not related to W78-66 (Table 3-18). The

second check population, in which a new dwarf plant type was

segregating, was an exception to this pattern.

Scatter diagrams (Appendix 1 Fig. 1 through 14) show the

relationship between height and internode length in each of the 14

populations. In most cases, plant height was positively correlated with

internode length and many populations had similar r values, but the

nature of the relationship varied among populations.

The largest height by internode length correlation coefficients

were found in populations from selected dwarf F-I plants backcrossed to

W78-66 (Table 3-18). The lowest r values were found in 2 of the 4

Table 3-17. Growth of dwarf and nondwarf populations of V. ashei.

Height Shoots length
Seedlings (cm)_ (no.) (ca)
Population (no.) Mean SD Range Mean SD Range Mean SD Range

Populations not having W78-66 in the pedigree

78-41 x 83-4 79 40.1az 10.6 10-60 18bz 7.3 4-40 1.3bz .26 .8-2.1
78-69 x 83-17 79 35.1c 11.0 8-60 18b 9.3 5-54 l.lc .24 .6-1.8
83-44 x 83-24 80 41.Oa 12.3 13-71 14c 7.7 3-46 1.3b .32 .8-2.1
83-45 x 83-19 80 39.8ab 12.4 8-64 lid 5.3 2-28 l.5a .28 .6-2.5

Populations having W78-66 in the pedigree

Tall cultivar x (W78-66 x tall cultivar)

Aliceblue x 82-167 80 36.7bc 11.0 15-64 15c 6.8 4-45 1.3b .29 .4-1.8
Bonita x 82-201 80 34.6c 10.1 8-58 14c 6.0 3-31 l.lc .20 .5-1.5
Bluebelle x 82-166 80 36.6bc 10.4 9-72 20a 11.7 5-84 l.lc .23 .6-1.8

Tall cultivar x W78-66

Bluebelle x W78-66 82 36.2c 11.5 11-56 21a 11.2 4-54 1.lc .21 .7-1.6

(W78-66 x tall cultivar) x (W78-66 x tall cultivar)

82-201 x 82-197 94 31.1d 8.7 7-52 19ab 8.6 4-54 1.Ocd .17 .5-1.4
82-158 x 82-160 170 21.8f 8.5 4-45 19ab 11.4 1-76 1.Ocd .30 .3-1.7
82-160 x 82-161 92 23.2f 10.9 4-46 18b 9.1 3-45 .9d .28 .3-1.6
82-161 x 82-156 80 30.1de 10.2 7-49 20a 11.0 3-66 l.lc .22 .4-1.6

(W78-66 x tall cultivar) x W78-66

82-195 x W78-66 80 22.3f 10.3 3-68 20a 11.0 6-61 .8d .22 .3-1.4
82-196 x W78-66 94 27.4e 9.2 6-52 19ab 9.8 4-54 .9d .20 .4-1.5

zMean separation within columns by Duncan's Multiple Range Test, 5% level.

Table 3-18.

Correlations between height, shoot number and internode
length in various V. ashei populations, measured September,

Height Height Internode length
vs vs vs
Seedlings Shoot Internode Shoot
Population (no.) number length number

Populations not having W78-66 in the pedegree

78-41 x 83-4 79 .25* .21* .09ns
78-69 x 83-17 79 .22* .25* -.19ns
83-44 x 83-24 80 .37** -.03ns -.04ns
83-45 x 83-19 80 .33** .Olns -.llns

Populations having W78-66 in the pedigree

Tall cultivar x (W78-66 x tall cultivar)
Aliceblue x 82-167 80 .16* .36** -.08ns
Bonita x 82-201 80 .16ns .31** .03ns
Bluebelle x 82-166 80 .06ns .34** -.06ns

Tall cultivar x W78-66
Bluebelle x W78-66 82 .28** .22** -.05

(W78-66 x tall cultivar) x (W78-66 x tall cultivar)
82-201 x 82-197 94 .31** .32** .02*
82-158 x 82-160 170 -.llns .35** -.01ns
82-160 x 82-161 92 .14ns .24** -.llns
82-161 x 82-156 80 .26** .30** .04ns

(W78-66 x tall cultivar) x W78-66
82-195 x W78-66 80 .03ns .55** -.17ns
82-196 x W78-66 94 .25** .49** -.14ns

**, Significant at .01 and .05 level, respectively, ns=not

populations which did not have W78-66 anywhere in their pedigrees.

Moderate but significant r values were found in the remaining


Scatter diagrams also show the relationship between height and

number of shoots in the various populations (Appendix 1 Figure 15

through 28). Most significant correlations between these 2 variables

are found in the populations which do not have W78-66 in their

pedigrees, but all are below .37.

In certain F2 populations resulting from crossing selected short

internode F1 plants, both populations resulting from backcrossing short

internode Fl's to W78-66, and the W78-66 selfed population, a unique

plant phenotype segregated (Table 3-19 and Figure 4). These plants were

all less than 10 cm tall and had at least 25 extremely compact branches

and internode lengths of 4 mm or less. Some of these plants appeared to

be quite vigorous, but they were far shorter than W78-66 or any of the

highly-tillered Fl plants obtained by crossing W78-66 x tall cultivars.

Inheritance patterns in polyploids can be complex and depend upon

the behavior of the chromosomes that carry the genes responsible for the

trait being studied, and the number and nature of the different alleles

present at the different loci (Burnhan, 1962). Chromosomes may range

from being completely homologous with the corresponding chromosomes in

the different genomes to totally lacking homology in corresponding

chromosomes. Most natural polyploids range in chromosome homology

within the genome. This would imply that disomic, polysomic, and

intermediate ratios for different characters are possible in progenies

from the same plant (Hermsen, 1984).

Table 3-19.

Segregation ratios of the extreme dwarf phenotype in
different populations of V. ashei.

Extreme Ratio of Ratio of
Seedlings dwarf extreme tall
Population (no.) (no.) dwarf plants

Populations not having W78-66 in pedigree

78-41 x 83-4
78-69 x 83-17
83-44 x 83-24
83-45 x 83-19

Populations having W78-66 in the pedigree

Tall cultivars x (W78-66 x tall cultivars)

Aliceblue x 82-167
Bonita x 82-201
Bluebelle x 82-166

Tall cultivars x W78-66

Bluebelle x W78-66

(tall cultivars x W78-66) x (tall cultivars x W78-66)

82-201 x 82-197
82-158 x 82-160
82-160 x 82-161
82-161 x 82-156




(tall cultivars x W78-66) x W78-66

82-195 x W78-66
82-196 x W78-66


W78-66 selfedz






zPlants are in a different location and 1
the other populations listed.

year younger than plants from

YRatio could not be determined due to lack of clear distinction between

Figure 3-4. Progeny from crosses with and without W78-66 in the
pedigree. 4-a. Progeny of population without
W78-66 in the pedigree. 4-b, c. Progeny of
populations with W78-66 in the pedigree, ex=extreme
dwarf, t=tall.






Since V. ashei is hexaploid, it could exhibit disomic, tetrasomic,

and/or hexasomic inheritance in the dwarf trait. It is also possible in

tetrasomic or hexasomic inheritance to have nonrandom chromosome pairing

at meiosis which would alter segregation ratios. This was reported in

the tetraploid V. australe Small (Draper and Scott, 1971). Since some

plants in the F1 populations (W78-66 x tall cultivar) and backcross to

tall cultivars [(W78-66 x tall cultivars) x tall cultivars] have dwarf

characteristics, the dwarf trait is somewhat dominant in its expression

and is possibly controlled by 1 or 2 genes. If dwarfism as expressed in

W78-66 is controlled by only 1 or 2 genes, the extreme dwarf found in

many of the populations resulting from a cross between 2 short-

internode F 's, when a short-internode FI is backcrossed to a tall

cultivar, and when W78-66 is selfed, may be homozygous for the dwarf

trait, and therefore W78-66 would be heterozygous for the trait. Since

dwarfs have not been found in crosses between V. ashei cultivars

(Lyrene, per. comm.), they are probably homozygous for normal height.

More crosses need to be made, such as the continued backcrossing of

short-internode segregates with tall cultivars, and crossing of the

extreme dwarf segregates with W78-66, tall cultivars, and selling before

the inheritance pattern is better understood.




The objectives of this research were to measure and describe

self-incompatibility in native Vaccinium ashei plants selected from

north Florida and south Georgia.

Materials and Methods

In spring 1984, 19 rabbiteye blueberry plants from throughout north

Florida and south Georgia were self- and cross-pollinated (Table 4-1).

These plants had been selected during the previous 5-year period,

primarily on the basis of large fruit size and high fruit quality. The

plants had been propagated by cuttings, grown to flowering size in a

field nursery, and then grown in 12-liter pots for a year before they

were used in this study. In January, 1984, the plants were placed in a

refrigerator at 7C for 4 weeks to satisfy their chilling requirement,

and then were moved to a greenhouse maintained between 40C and 25*C for

flowering. Each plant was divided into 3 parts: 2 sections were

self-pollinated and 1 section was cross-pollinated. Approximately 200

flowers of each section were used. Both crossed and selfed flowers were

emasculated. The pollen used in cross-pollination of all native plants

was a bulk collected from the 4 V. ashei cultivars 'Bonita',

'Bluebelle', 'Climax', and 'Beckyblue'.

Table 4-1.



Location of the 19
native V. ashei clones.

County, State

Columbia, FL
Okaloosa, FL
Escambia, FL
Okaloosa, FL
Washington, I
Nassau, FL
Washington, I
Nassau, FL
Washington, I
Okaloosa, FL
Leon, FL
Okaloosa, FL
Okaloosa, FL
Okaloosa, FL
Okaloosa, FL
Okaloosa, FL
Ware, GA
Nassau, FL
Nassau, FL


Male fertility of the 19 native clones was measured by pollen

germination using the "hanging drop technique" (Stanley and Linskens,

1964) and fruit set after pollination of the two cultivars 'Bluegem' and

'Climax'. Germination medium was that of Goldy and Lyrene (1983) and

consisted of 100mg/liter H3B03, 300mg/liter Ca(N03)2 H20, 200 mg/liter

MgSO4 7 H20, 100mg/liter KNO3, and 10% sucrose. Female fertility was

measured when native plants were cross-pollinated with bulked cultivar


Seven of the native V. ashei clones were used to study pollen

germination and pollen tube growth after cross- and self-pollination.

For this study 5 styles were collected from each of the 7 clones at each

of 6 intervals after pollination (12, 24, 36, 48, 72, and 96 hr) for

both cross- and self-pollinations. Styles were stained with

water-soluable Aniline blue and examined by fluorescence microscopy (Kho

and Baer, 1968).

Fruit set, fruit weight (10 berries/treatment), number of seedlings

per fruit, number of germinated seeds/pollination, fruit development

period (interval of 50% pollination to 50% ripe fruit), and harvest

duration were recorded. Self-fruitfulness and self-fertility were

calculated for each clone.

Self-fruitfulness was defined as the number of fruit per

pollination when self-pollinated, divided by the number of fruit per

pollination when cross-pollinated. Self-fertility was measured by 1)

the number of seedlings produced per fruit when selfed, divided by the

number of seedlings produced per fruit when crossed, and 2) the number

of seedlings produced per pollination when selfed divided by the number

of seedlings produced per pollination when crossed.

Results and Discussion

Pollen germination varied from 8% to 73% among native plants, but

there was considerable variation in different samples from the sane

clone (Table 4-2). Pollen fertility, measured as the percent fruit set

after pollination onto V. ashei cultivars, varied from 29% to 89% among

the 19 native V. ashei clones. The low pollen germination of 83-42 (8%)

was not reflected in the pollination test, and females self-pollinated

set well. Fruit set varied from 3% to 82% for the 19 V. ashei clones

when they were pollinated with a bulk of pollen from 4 V. ashei

cultivars (Table 4-2).

While pollen had germinated on the stigmas of all plants by the end

of 12 hr, its rate of growth in the style was highly variable

(Table 4-3). All seven plants were not pollinated on the sane day and

the slow growth of the pollen tubes of 3 plants 78-72, 78-77, and 78-112

may have been due to low temperatures after pollination of these plants

(Knight and Scott, 1964).

Self-fruitfulness as measured by percent fruit set after selfing

divided by percent set after crossing varied from 0.0% to 68%

(Table 4-4). Five native plants had a relative fruit set greater than

50%, 9 were greater than 20%, and only 1 plant was completely


Relative self-fertility compared to cross-fertility (based on

number of seedlings/fruit after selfing divided by seeds per fruit after

crossing) varied from 0% to 110% (Table 4-4). Relative self-fertility as

the number of seedlings/pollination varied from 1% to 45% (Table 4-4).

Fruit set was not correlated with number of seed per fruit,

indicating that plants like 83-45, which have a high fruit set but few

Table 4-2.

Fertility of 19 clones of V. ashei from south Georgia and
north Florida measured by pollen germination on an
artificial medium, fruit set after pollination of known
fertile plants, and fruit set after cross-pollination.

Pollen Fruit set when Fruit set when
germination pollen put onto pollinated by
(%) V. ashei cultivarsy cultivarsx
Clone Mean Range (%) (%)

W78-72 73 53-90 89 78
W78-112 22 13-29 72 3
W78-77 22 14-29 56 31
W78-74 39 34-50 38 65
W78-39 73 54-77 66 82
83-42 08 05-14 53 47
83-44 41 27-52 67 62
83-40 40 26-53 61 77
83-43 43 15-76 56 75
W78-102 33 21-43 77 81
W78-34 43 20-65 53 45
83-47 59 52-68 57 60
W78-41 36 25-52 29 67
W78-40 35 16-57 38 40
83-46 57 40-77 37 55
83-45 33 27-39 58 45
83-41 66 30-85 76 52
W78-69 62 52-75 33 76
W78-66 23 10-39 63 52

zlO% Sucrose .01Z
tested per clone.

boron solution for 4 hours.

Five samples were

YNumber of fruit/pollination on V. ashei cultivars when the native
plants were the pollen parents. Approximately 30 flowers pollinated
to test each clone.

Number of fruit/pollination when native plants were pollinated with a
composite pollen from V. ashei cultivars. Approximately 200 flowers
were pollinated to test each clone.



Table 4-3.

Hours required for pollen germination and growth in cross
and self-pollinated pistils of 7 native V. ashei plants.

germination on
crossed selfed

12 hr

12 hr

Pollen tubes
mid-way in
crossed selfed

36 hr

36 hr

Pollen tubes
crossed selfed

48 hr

48 hr



Table 4-4.

Percent self-fruitfulness and percent self-fertility of 19
native plants of V. ashei from south Georgia and north

Clone Self-fruitfulness Self-fertility Self-fertility
(%) sf/sfx sp/spw
actual relative (%) (%)

W78-72 obv Oc Oc 0b
W78-112 # # # #
W78-77 2b 6c 60ab 4b
W78-74 3b 5c 59ab 2b
W78-39 3b ic 33b lb
83-42 3b 6c 33b 2b
83-44 3b 8bc 33b lb
83-40 6b 7bc ll0a 7b
83-43 9b 12bc 32b 6b
W78-102 13b 16abc 15b 3b
W78-34 16ab 37abc 12b 4b
83-47 19ab 31abc 29b 10b
W78-41 20ab 30abc 20b 7b
W78-40 24a 60ab 33b 20a
83-46 29a 51abc lOb 6b
83-45 31a 68a 10b 7b
83-41 35a 66a 44b 6b
W78-69 35a 46abc lib 7b
W78-66 36a 69a 23b 7b

Clone ** *

Not calculated

due to the high level of female sterility of the clone.

ZPercent fruit set when self-pollinated.

YNumber of fruit per pollination when selfed/number of fruit per
pollination when crossed.

XNumber of seedlings per fruit when selfed/number of seedlings per
fruit when crossed.

Number of seedlings per pollination when selfed/number of seedlings per
pollination when crossed.

VMean separation within columns by Duncan's Multiple Range Test, 5%
u**, *, significant at .01 and .05 level respectively, ns=not

viable seed per fruit, may have a greater ability to develop

parthenocarpic fruit than plants like 83-40, which had many viable

seed/berry. Increasing the number of seed/berry increased fruit size

and decreased the fruit development period (4-5).

In general, selfing resulted in reduced fruit set, smaller berries,

fewer seed/fruit, fewer seed/pollination, and a longer fruit development

period, but did not alter the duration of harvest (Table 4-6).

These results indicate that there is considerable variation for

self-compatibility in the native Florida rabbiteye germplasm. The low

level of self-compatibility in the present rabbiteye cultivars

(El-Agamy, 1979) may be attributed to chance selection of high

self-incompatibility in the 6 original plants (Lyrene, 1983).

Self-compatibility has been transferred in other crops by making the

appropriate crosses with compatible varieties (Ronald and Asher, 1975;

Odland and Noll, 1950) and may be possible with V. ashei.

Although there was considerable variation in self-fruitfulness

among the 19 native V. ashei clones tested, none would be successful

horticulturally if planted in solid blocks. In general, levels of

self-incompatibility in V. ashei appear substantially higher than in the

V. corymbosum cultivars that have been tested (El-Agamy, 1979; Meader

and Darrow, 1947; and Johnston, 1940).

Table 4-5. Fruit characteristics of 19 native rabbiteye blueberry
plants from south Georgia and north Florida as affected by
self and cross-pollination.

Fruit Fruit Seedlings Seedlings Fruit Harvest
set weight per fruit per pol. dev. duration
Clone (no.) (g) (no.) (no.) (days) (days)

selfed 0 --- -- 0.0 --- --
crossed (female)y 78 1.20 10 8.0 93 10.8
crossed (male)x 89 1.50 31 28.0 81 20.4
selfed 1 1.00 17 0.1 110 --
crossed (female) 3 .80 8 0.2 101 11.5
crossed (male) 72 1.80 18 12.0 95 21.3
selfed 2 .42 4 0.1 114 15.3
crossed (female) 31 1.00 6 2.0 100 17.4
crossed (male) 56 1.70 34 11.0 90 12.3
selfed 3 .60 3 0.1 116 14.8
crossed (female) 65 1.10 5 3.0 92 13.1
crossed (male) 38 1.50 34 20.0 104 21.6
selfed 3 .60 3 0.1 121 14.3
crossed (female) 82 1.20 9 7.0 94 13.3
crossed (male) 66 1.80 52 50.0 88 14.7
selfed 3 .40 5 0.2 90 5.9
crossed (female) 47 .80 15 7.0 84 5.9
crossed (male) 53 1.80 21 10.0 91 14.5
selfed 3 .80 4 0.1 103 7.9
crossed (female) 62 1.30 12 7.0 93 11.9
crossed (male) 67 1.70 24 17.0 85 16.3
selfed 6 1.10 10 0.5 82 10.5
crossed (female) 77 1.60 9 7.0 80 6.5
crossed (male) 61 1.80 12 9.0 91 10.7
selfed 9 .47 2 0.2 126 15.7
crossed (female) 75 1.10 5 4.0 85 22.5
crossed (male) 56 1.70 24 11.0 92 11.3
selfed 13 .40 2 0.3 102 12.6
crossed (female) 81 .80 13 11.0 105 15.2
crossed (male) 77 1.80 39 30.0 100 22.2
selfed 16 .56 1 0.2 101 15.3
crossed (female) 45 1.10 10 5.0 86 9.5
crossed (male) 53 1.80 30 17.0 88 11.5

Table 4-5. (continued)

Fruit Fruit Seedlings Seedlings Fruit Harvest
set weight per fruit per pol. dev. duration
Clone (no.) (g) (no.) (no.) (days) (days)

selfed 19 .85 3 0.6 124 15.4
crossed (female) 60 1.00 11 6.0 97 16.8
crossed (male) 57 2.10 30 16.0 81 11.8
selfed 20 .53 2 0.3 126 17.8
crossed (female) 67 1.10 6 4.0 102 23.1
crossed (male) 29 1.60 12 11.0 84 13.4
selfed 24 1.10 5 1.0 102 13.9
crossed (female) 40 1.50 13 5.0 99 15.5
crossed (male) 38 1.40 13 5.0 101 13.0
selfed 29 .80 3 0.7 89 11.9
crossed (female) 55 1.20 23 13.0 81 13.9
crossed (male) 37 1.70 22 8.0 95 9.1
selfed 31 .80 2 0.5 114 12.8
crossed (female) 45 .85 16 7.0 84 7.9
crossed (male) 58 2.00 40 34.0 82 11.8
selfed 35 .80 7 0.5 80 11.3
crossed (female) 52 .70 16 8.0 82 7.2
crossed (male) 76 1.70 20 14.0 106 22.7
selfed 35 .90 1 0.4 100 15.8
crossed (female) 76 1.30 8 6.0 81 14.1
crossed (male) 33 1.50 21 8.0 81 14.1
selfed 36 .50 4 0.6 102 10.4
crossed (female) 52 1.10 17 9.0 99 8.4
crossed (male) 63 1.50 25 12.0 90 12.5

ZlInterval between 50% pollination and 50% fruit harvest.

YCharacteristics of 78-72 when cross-pollinated by V. ashei cultivars.
XCharacteristics of V. ashei cultivar when pollinated by 78-72.

Table 4-6.

Average fruit characteristics of 19 native rabbiteye
blueberry plants as affected by self and cross-pollination.

Fruit Fruit Seedlings Seedlings Fruit Harvest
Treatment set weight per fruit per pol. development duration
(%) (ginm) (no.) (no.) (days) (days)

Selfed 15bu 0.7c" 4c" 0.5c" 106a" 13a"
Cross ed(f emalely 58a 1.lb lib 6.Ob 92b 13a
Crossed (male) 57a 1.7a 26a 17.Oa 91b 15a


Self vs Crossy ** ** ** ** ** ns
Self vs Crossx ** ** ** ** ** ns
Cross vs Crossx ns ** ** ** ** ns
Self vs Crossedv ** ** ** ** ** ns

Zlnterval between 50% pollination and 50% fruit harvest.

YBulk pollen from V. ashei cultivars used

to pollinate native V. ashei

Xpollen from native V. ashei clones used to pollinate V. ashei
w** *, significant at .01 and .05 level respectively, ns=not

Combined results from both types of cross pollination.

UMean separation within columns by Duncan's Multiple Range Test, 5%




On September 23, 1978, cuttings from W78-122, a very large,

vigorous blueberry plant, were collected in an abandoned rabbiteye

plantation 5 km north of Gainesville, Florida. The original planting is

believed to have been made between 1920 and 1930 with native seedlings

from west Florida (Lyrene and Sherman, 1979). W78-122 had unusually

large leaves and set no fruit even though it had flowered heavily. The

objectives of this experiment were to study the nature of this

large-statured, female-sterile plant.

Materials and Methods

Phenotype of W78-122 in its native environment

On August 10, 1983, height, trunk diameter (at 1 meter height), and

number of large trunks/plant were collected from 11 of the largest

plants (including W78-122) growing within 50m of W78-122. Ten

fully-expanded leaves from 20 large rabbiteye plants growing in the same

area were collected, and leaf area and leaf index (length/width) were

measured. Leaf area was measured using a LiCor Portable Area Meter

Model 3000. Presence or absence of stalked glands on the undersurface

of the leaves was noted.

Phenotype at the Horticultural Research Unit

On August 17, 1983, 10 leaves from each of 7 random Vaccinium

plants and W78-122 growing in the Germplasm Block at the Horticultural

Research Unit were collected, and leaf area was measured. Ten leaves

from each of 2 plants for each of 4 cultivars of V. ashei were collected

at the same time, and leaf area was measured.

Meiotic study and chromosome counts

Meiosis was studied using pollen mother cells from developing

inflorescence buds. Developing inflorescence buds were collected and

fixed in 3:1 absolute ethanol:glacial acetic acid for 24 hr and stored

in 70% ethanol until used. Anthers were removed from 4-5 flowers of an

inflorescence, squashed in 1% acetocarmine stain, destined in 45%

acetic acid, and viewed under phase contrast at 1000X.

Male fertility

Male fertility was measured as the percent pollen germination in an

artificial medium (as described in Section 4) and percent fruit set of

known fertile V. ashei cultivars when pollinated with W78-122 pollen.

Both measurements were compared with results from a number of other

native V. ashei plants.

Pollen germination and growth in pistil

The method used to study pollen germination and growth in the

pistil of W78-122 and the other native plants was the same as that

described in Section 4 of this paper.

Ovary and ovule morphology

Flower buds were collected from W78-122 at stages from immature to

anthesis. Flowers were dissected, size of the ovary was measured, and

number of viable-appearing ovules were counted. Ovules from immature

and mature ovaries were collected and cleared for direct observation.

Ovules were fixed in FPA50 (40%formaldehyde:propionic acid:50%ethanol,

5:5:90, v:v:v) for 24 hr (Herr, 1974a). Ovules were cleared using

Crane's (1978) technique. Fixed ovules were dehydrated in a 50-70-85-

100-100% ethanol series for 30 min each. They were then brought to 100%

methyl salicylate through 2 changes for 30 min each in a 1:1 and 1:3

ethanol:methyl salicylate mixture. Finally they were transferred to

100% methyl salicylate overnight, and mounted on a Raj slide (Herr,

1974b) in the same solution. Observation of the cleared ovules was done

with a Nikon Normarski differential interference contrast microscope at

a magnification of 400x.

Phenotype of F1

Cross-pollinations of W78-122 (pollen parent) with 4 different V.

ashei cultivars were conducted during the spring of 1980 by Dr. P. M.

Lyrene. In the spring of 1981, seedlings of the populations were

planted into the high-density nursery with seedlings from the breeding

program. Stem diameter, height, and internode length of all F, plants

and plants from 3 unrelated populations were measured in February 1983.

Ten leaves from: all F1 plants, plants from 3 populations which did not

have W78-122 in the pedigree, ramets of 2 V. ashei cultivars, and 3

ramets of W78-122 were collected and leaf area was measured as above.

All plants were growing in a high-density nursery.

Female fertility of the F1 plants was monitored for 3 years by

recording the plants which flowered and fruited. Plants which produced

flowers but never had fruit were considered female-sterile.

Results and Discussions

Phenotype in its native environment

Height, trunk diameter, and internode length were larger in W78-122

than in most of the other measured plants growing in the same vicinity

(Table 5-1). Leaf area was more than 2.5X greater in W78-122 than on

the average check plant (Table 5-2). In a normal population with mean

and variance like that of the 10 sampled "check plants", only 1 plant in

100,000,000 would be expected to have a leaf area as great as that of

W78-122. Leaf index, a measure of leaf shape, was not significantly

different among W78-122 and the check plants but was highly variable

(Table 5-2). The undersurface of the leaves of W78-122 was totally void

of stalked glands. While this is considered by some to be a

characteristic of V. ashei, 2 of the other native rabbiteye plants

sampled also lacked stalked glands on their leaf undersurface

(Table 5-2). Fruit has never been observed on W78-122 during 7 years

despite abundant flowering each year (Lyrene, per. comm).

Phenotype at the Horticultural Research Unit

A ramet of W78-122 growing at the HRU had a larger leaf area than

any of the native or cultivated V. ashei plants measured (Table 5-3).

This ramet has flowered for 4 years without setting fruit.

Superficially, the flowers appear normal, although rather small for V.

Table 5-1.

Height, trunk diameter and internode length of the 10
largest V. ashei plants growing with W78-122 in an abandoned
blueberry plantation near Gainesville, Florida.

Trunk diameter-
1 2 3 Mean


Internode length


zAverage diameter of the 3 largest trunks at a height of 1.5 m above

YMean separation within columns by Duncan's Multiple Range Test, 5%


Check 1



Table 5-2.

Leaf area and leaf index comparisons
with W78-122 growing in an abandoned
near Gainesville, Florida.

of 20 V. ashei plants
blueberry plantation

Leaf area (cm 2) Leaf indexy
Plant Mean SD Mean SD

Rep 1# 37.0 3.3 1.8 .12
Rep 2# 36.4 7.2 1.8 .08

Check 1 12.2 1.7 1.8 .15
2 14.3 2.3 2.0 .14
3 17.0 2.3 1.8 .10
4 14.4 2.4 1.7 .10
5 10.4 1.8 2.1 .30
6 11.3 1.3 1.8 .10
7 13.3 2.0 2.1 .10
8 12.7 2.0 1.8 .10
9 11.8 1.4 2.1 .10
10 14.6 1.1 1.7 .05
11 19.7 2.6 1.7 .20
12 13.0 1.8 1.8 .10
13 11.5 2.0 2.4 .20
14 24.2 4.1 1.7 .10
15 12.6 1.3 1.6 .10
16 11.3 1.4 1.8 .10
17 13.2 1.5 1.6 .20
18# 15.1 1.9 1.8 .10
19 13.3 2.3 2.0 .20
20# 16.5 2.7 2.2 .20

W78-122 (Combined) 36.7a 5.5 1.8ax .10
Checks (Combined) 14.1b 3.3 1.9a .21

ZSurface area measured using

a LiCor Portable Surface Area Meter Model


XMean separation within columns using Duncan's Multiple Range Test, 5%

Mean of 10 leaves measured on each plant. Two replications of 10
leaves were collected from W78-122.

#No stalked glands on the undersurface of leaves.

Table 5-3.

Leaf area comparisons of native and cultivated V. ashei with
W78-122 growing at the Horticulture Research Unit.

Leaf area
Clone Type Mean SD

W78-122 Native 25.6az 5.3

W78-34 6.Ode .8
W78-66 5.3e 1.4
W78-35 11.8c 2.6
W78-74 8.2d 1.4
W78-37 6.8de 1.2
W78-39 8.4d 1.4

Bluebelle Cultivar 10.7c 1.7
Climax 10.6c 3.0
Fla K 16.9b 2.2
Beckyblue 12.Oc 3.0

ZMean separation within columns using Duncan's Multiple Range test, 5%

ashei, with normal-appearing calyx, corolla, pistil, and anthers.

Anthesis appears to be normal. Flowering superficially appears to be

proceeding normally until a week or two after anthesis, when the flowers

abscise. Flower abscission occurs at about the same time in the

reproductive cycle as the abscission of unpollinated flowers on normal


Chromosome counts

Approximately 36 bivalent chromosomes were counted in 32 pollen

mother cells undergoing meiosis. This indicated that W78-122 is most

likely a member of the hexaploid species V. ashei with a chromosome

count of 2n=6x=72.

Male fertility

Male fertility, measured as percent pollen germination on an

artificial medium and fruit set of unrelated V. ashei cultivars

pollinated with W78-122, was normal when compared with 19 other native

V. ashei clones tested from north Florida and south Georgia (Table 5-4).

Pollen germination of W78-122 on the sucrose medium was similar to

pollen germination of 7 other native plants and much greater than 13

plants. The variability in percent pollen germination of W78-122 was

large, but it was also large for some of the other clones. Percent

fruit set when pollen from W78-122 was used to pollinate V. ashei

cultivars was higher than that of 5 other clones, less than 3, and

similar to the others. Male fertility in V. ashei appears to be

similar to that found in V. angustifolium (Hall and Aalders, 1961).

Table 5-4.

Male fertility compared for W78-122 and 19 other native
clones of V. ashei from south Georgia and north Florida
measured by pollen germination on an artificial medium and
fruit set after pollination of known fertile plants.

Pollen Fruit set when
germination pollen put onto
(%) V. ashei cultivarsy
Clone Mean Range (%)



z10% sucrose .01% boron solution for 4 hours, 5 samples per clone.

Number of fruit/pollination on V. ashei cultivars when the native
plants are the pollen parents. About 30 flowers pollinated for each

Pollen germination and growth in pistil

Pollen had germinatated on the stigma W78-122 and 7 other native V.

ashei plants within 12 hours (Table 5-5). While pollen tube growth

through the style and into the ovary was highly variable among the 8

clones, pollen tube growth of W78-122 was similar to pollen tube growth

of W78-112. Long intervals between sampling times may accentuate

differences in reported pollen tube growth rates. Sterility in this

plant is therefore probably not due to abnormalities of the stigma and

style which would block the pollen tube growth in the style and into the


Ovary and ovule morphology

Ovule degeneration occurred during the development of the flower

(Table 5-6). Dissected ovaries from young flowers of W78-122 contained

ovules with the normal number and appearance (Figure 5-1). Dissected

ovaries from flowers at anthesis contained a few, white, plump healthy

looking ovules, but the majority were abnormal. Some abnormal ovules

were white but misshapen, some clear and thin, and some very small and

white. Cleared ovules from flowers at anthesis had not completely bent

over and were not anatropous This seemed to indicate a cessation of

or delay in development. One abnormal ovule contained an embryo sac in

which only 2 nuclei could be seen instead of the mature megagametophyte

expected (Stushnoff and Palser, 1970).

It appears that female sterility in W78-122 is due to failure of

the megagametophyte to develop normally. Meiosis was studied only in

the microspore mother cell and was not attempted in the megaspore mother

cell. Since normal meiosis was observed in the microsporocyte, abnormal

Table 5-5.

Number of
growth in
native V.

hours from pollination to pollen germination and
cross-pollinated pistils of W78-122 and 7 other
ashei plants.

Pollen Pollen tubes Pollen tubes
Clone germination on midway in in
stigma style ovary


12 hr

36 hr

48 hr

Table 5-6.

Ovule number and appearance of immature and mature flowers
of W78-122 and the V. ashei cultivar 'Bluebelle'.

Immature flowers Mature flowersy
W78-122 ovules W78-122 ovules Bluebelle ovules
(no.) (no.) (no.)
Flower Normal Abnormal Normal Abnormal Normal Abnormal

1 50-60 0 13 11 53 0
2 1.. I 20 47
3 3 10 51 "
4 3 13 49 "
5 7 9 55 "
6 5 7 58 "
7 6 11 46
8 .. 8 3 47
9 5 20 54 "
10 17 11 50

ZFlowers from newly expanded infloresence bud, less than 3 mm long.

YFlowers at anthesis.