CHAPTER 1

APPLES
Jules Janick, James N. Cummins, Susan K. Brown, and Minou Hemmat

The apple is the most ubiquitous of temperate fruits and has been cultivated in
Europe and Asia from antiquity. It was known to the Greeks and Romans and
mentioned by Theophrastus in the third century B.C. Since then the apple has been
distributed into almost all parts of the world. The genetic variability found in the
apple has allowed adapted types to be selected for different environments, and
selection continues for new types to extend apple culture into both colder and warmer
regions. Orchards are now found in Siberia and northern China where winter
temperatures fall to 40C and in high elevations in Colombia and Indonesia
straddling the equator where two crops can be produced in a single year (Janick
1974). Present world production of apples (FAO 1995) is close to 49 million tonnes.
Apples are the fourth fruit crop in importance after all citrus (85 million t), grapes
(56 million t), and banana (53 million t). The leading countries for apple production
are presented by continent in Table 1.
Apples are popular because of the many ways that they can be consumed and
because of their convenience and durability. Apples may be eaten off the tree or
stored for up to a full year. Apples can be processed into sauce, slices, or juice and
are favored for pastries, cakes, tarts, and pies (Downing 1989). The pulp has been
processed into candies (fruit leathers) and used as a source of pectin. The juice can
be consumed fresh, either natural or filtered, fermented into alcoholic beverages
such as cider or wine, distilled into brandy, or transformed into vinegar. Apples
have become the symbol of wholesomeness: An apple a day keeps the doctor away
is a favorite aphorism, and apple pie has become a symbol of goodness along with
motherhood. Finally, crabapples, grown for their attractive flowers, foliage, and
fruit, are among the most popular of ornamental trees.
At one time each country and area had its own local cultivars. This is still the
case in some areas, but with expanding production and transport networks, and the
Fruit Breed, Volume I: Tree and Tropical Fruits, edited by Jules Janick and James N. Moore
ISBN 0-471-31014-X 1996 John Wiley & Sons, Inc.

APPLES

TABLE 1. Wor
ld Pr
oduction of Apples, 1994.
orld
Production
Principal apple
producing country
Asia
China
Iran
India
Japan
Europe
Eur
ope
France
Italy
Russian Federation
Romania
Poland
Germany
Spain
Hungary
Netherlands

Production
(million
tonnes)
20.4
12.0
1.6
1.2
1.0
13.2
2.2
2.1
1.7
1.4
1.4
0.9
0.7
0.7
0.7

Principal apple
producing country

Production
(million
tonnes)

Nor
th Amer
ica
North
America
United States
Mexico
Canada

6.0
4.9
0.5
0.5

ica
South Amer
America
Argentina
Brazil

2.7
1.0
0.8

Africa
South Africa
Egypt
Morocco

1.3
0.6
0.3
0.3

Oceania
New Zealand
Australia

0.9
0.5
0.3

Source: FAO (1995).

requirements for high yield of commercial quality, a few cultivars dominate all the
major apple growing areas. For example, the most widely grown cultivars by far
are Golden Delicious and Delicious and its red sports, both chance seedlings of
American origin. Golden Delicious has been widely and successfully used in
breeding, and its seedlings, which make up a high proportion of the new cultivars,
are rapidly changing the apple industry. Since the apple is a long-lived tree and
vegetatively propagated, cultivars known hundreds of years ago still exist. Apples
that date to antiquity have been collected in Italy (S. Sansavini, personal
communication). The large world collections and germplasm repositories are living
museums that show the development and improvement of the apple and contain a
vast reservoir of desirable genes. Old cultivars are little used by breeders because
their seedlings generally produce fruit inferior to that of the cultivars being grown
today, but they represent a living gene bank. Most breeding in recent years has been
among the best commercial cultivars. Interspecific hybridization to introduce new
genes to scion cultivars as well as rootstocks is being used, but this requires a lengthy
breeding process. To overcome these limitations of time and space, recent advances
in biotechnology involving gene transformation now make it possible to introduce
genes from almost any source into apple.

ORIGIN AND EARLY DEVELOPMENT

The apple, in company with most of the important temperate fruits-pear, plum peach,
apricot, cherry strawberry, raspberry, and blackberry-belongs to the Rosaceae or

ORIGIN AND EARLY DEVELOPMENT

rose family. Apple, pear, quince, medlar, and some less well-known genera have
been classified into the subfamily pomoideae, the pome fruits. These are
characterized by fruits consisting of two to five carpels enclosed in a fleshy covering.
The genus Malus has, according to most authorities, 25 to 30 species and several
subspecies of so-called crabapples, many of which are cultivated as ornamental
trees for their profuse blossom and attractive fruits (Table 2). Most of the species
intercross and, since self-incompatibility is common, trees produced from seed
obtained from a botanic garden or arboretum where collections of Malus are grown
are almost always interspecific or intercultivar hybrids. it is therefore very difficult
to be certain of the authenticity of specific names.
The cultivated apple is likely the result of interspecific hybridization and at present
the binomial Malus domestica has been generally accepted as the appropriate
scientific name (Korban and Skirvin 1984). The main ancestor of apple is now
considered to be Malus sieversii, which is wild from the Heavenly Mountains (Tien
Shan) at the boundary between western China and the former Soviet Union, to the
edge of the Caspian Sea (Morgan and Richards 1993). This species is diverse and
wild trees bearing the full range of forms, colors, and tastes are found in Kazakhstan
and other independent countries of Central Asia formed from the breakup of the
Soviet Union and especially around Alma Ata (Father of Apples). This is the area of
greatest diversity and the center of origin. Recent collection trips to central Asia
have verified that M. siever3ii is very diverse and has all the qualities present in M.
domestica (Forsline et al. 1994; Forsline 1995). Vavilov (1930) in his explorations
found Trony wild apples in woods in the Caucasus and Turkestan bearing firuits
within a wide range of sizes, some of quite good quality. Species that have contributed
to the genetic makeup of the apple include M. orientalis, which bears late-keeping
bitter fruit; M. sylvestris, the European crab, bearing small astringent, greenishyellow fruits, native to an area that extends from Britain across Europe to the Ballan and northern Turkey; and a number of species from eastern Asia, including M.
baccata, the hardy but small Siberian crab, M. mandshurica, the Manchurian crab;
and M. prunifolia, the larger Chinese crab. A list of Malus sections, primary species,
and species hybrids is presented in Table 2.
Cultivation of the apple seems to have been practiced by the Greeks and Romans
and, as a result of their travels and invasions, to have been spread by them throughout
Europe and Asia. Later cultivation was concentrated around the medieval
monasteries. Cultivars were selected and propagated in very early times, for grafting
was known at least 2000 years ago. By the end of the thirteenth century, many
named cultivars were known, and from this time we get the names Pearmain and
Costard. The history and romance of the apple has recently been summarized in
The Book of Apples by Morgan and Richards (1993).
Until the latter half of the twentieth century most of the worlds apple cultivars
were chance seedlings selected by fruit growers. More than 10,000 cultivars are
documented, yet only a few dozen are grown on a commercial scale worldwide
(Way et al. 1990). In 1983, the best known cultivars in the world were all chance
seedlings found in the eighteenth or nineteenth centuries, of which many were derived
in North America: Golden Delicious (6.3 million t, origin US), Delicious (3.0

APPLES

TABLE 2. Malus Sections and Pr

imar
y Species
Primar
imary

Sections

2n

Fruit
size (cm
Apomixis
diam) Calyxa

SECTION 1. Malus
Subsection A. Pumilae
Series a. Pumilae
M. asiatica Nakai
34
M. domestica Borkh.
34,51,68
M. micromalus Makino
34
M. orientalis Uglitzk. ex Juz.
M. prunifolia (Willd.) Borkh. 34
M. pumila Miller
34
M. sieversii (Lodeb.) M.

Roemer
M. spectabilis (Aiton) Borkh. 34,68
M. sylvestris Miller
34

Persistance
Carpel of ripe
no.
fruit

No
No
No

No
No

>2
>2
1

2
>2
>2

P
P
D

P
P
P

5
5

5
5
5

No
No

Yes
No
No

No
No

2
>2

P
P

5
5

Yes
No

34,68
34

No
No

1
1

D
D

4, 5
4, 5

Yes
Yes

34
51

No
Yes

1
1

D
D

3, 4, 5
3, 4

Yes
Yes

34

No

Yes

51

Yes

4, 5

Yes

Yes
?

1
1

D
D

3, 4, 5
3, 4, 5

Yes
Yes

34

No
No

1
1

D
D

2, 3, 4
3, 4, 5

Yes
Yes

51

Yes

1
1

D
D

4, 5

Yes
Yes

Yes

Series b. Yunnanenses
M. honanensis Rehder

M. ombrophila Hand.-Mazz.

1.5
>2

P
P

4
5

Yes
No?

Series b. Baccatae
M. baccata (L.) Borkh.
M. floribunda (Siebold)
ex. Van Houte
M. halliana Koehne
M. hupehensis (Pampan.)
Rehder
M. mandshurica (Maxim.)
V. Komarov
M. sikkimensis (Wenzig)
Koehne ex C. Schneider

Subsection B. Sieboldianae
M. sargentii Rehder
68
M. sieboldii (Regel) Rehder 34-85?
Subsection C. Kansuenses
Series a. Kansuenses
M. fusca (Raf.) C. Schneider
M. kansuensis (Batalin)
C. Schneider
M. komarovii (Sarg.) Rehder
M. toringoides (Rehder)
Hughes
M. transitoria (Batalin)
C. Schneider

ORIGIN AND EARLY DEVELOPMENT

TABLE 2. (Contin
ued).
(Continued).

Sections

M. prattii (Hemsley)
C. Schneider
M. yunnanensis (Franchet)
C. Schneider

2n

Fruit
size (cm
Apomixis
diam) Calyxa

Persistance
Carpel of ripe
no.
fruit

34

No

1.5

Yes

34

No

1.5

Yes

SECTION II. Sorbomalus

M. florentina (Zuccagni)
C. Schneider

34

No

P, D

3, 4, 5

Yes

SECTION III. Eriolobus

M. trilobata (Poiret)
C. Schneider

No

34

No

>2

No

Yes?
No

>2
>2

P
P

5
5

No
No

No

>2
>2

P
P

5
5

No
No

No

>2

No

SECTION IV. Chloromeles

M. angustifolia (Aiton)
Michaux
M. coronaria (L.) Miller
M. ioensis (Alph. Wood)
Britton

51(68)
34

SECTION V. Docyniopsis
M. doumeri (Bois) A. Chev.
M. melliana (Hand.-Mazz.)
Rehder
M. tschonoskii (Maxim.)
34
C. Schneider

Note. Malus species hybrids include

M. adstringens Zabel. (baccata pumila
M. arnoldiana (Rehd.) Sarg. (baccata floribunda)
M. astracanica Dum.-Cours. (pumila prunifolia)
M. atrosanguinea (Spaeth) Schneid. (halliana sieboldia)
M. dawsonia Rehd. (fusca pumila)
M. gloriosa Lem. (pumila niedzwetskyana scheideckeri)
M. hartwiggii Koehne (halliana baccata)
M. heterophylla Spach (coronaria pumila)
M. magdeburgensis Schoch. (spectabilis pumila)
M. platycarpa Rehd. (coronaria domestica)
M. purpurea (Barbier) Rehd. (niedzwetzkyana atrosanguinea )
M. robusta (Carr.) Rehd. (baccata prunifoia)
M. scheideckeri (Spaeth )Zab. (floribunda prunifolia)
M. soulardii (Bailey) Brit. (ioensis pumila)
M. sublobata (Dipp.) Rehd. (prunifolia sieboldiii)
M. zumi (Mats.) Rehd. (mandschurica sieboldii)
a
P, persistent; D, deciduous.
Source: Adapted from Way et al. (1990).

APPLES

million t, origin US) Coxs Orange Pippin (1.7 million t, origin England), Rome
Beauty (0.8 million t; origin US), Belle de Boskoop (0.7 million t, origin The
Netherlands), Granny Smith (0.6 million t, origin Australia), Jonathan (0.5 million
t, origin US), Reinette du Canada (0.4 million t, origin France), McIntosh (0.3
million t, origin Canada), and Worcester Pearmain (0.3 million t, origin England).
The introduction of apples into the United States as seeds and the replanting of
seeds obtained from cider mills resulted in literally tens of millions of seedlings
being grown and evaluated by fruit growers. The nurseryman Jonathan Chapman
of Leominster, Massachusetts, known as the legendary Johnny Appleseed, is credited
with distributing apple seeds hum the cider mills of Western Pennsylvania throughout
the wilderness of Pennsylvania, Ohio, and Illinois in the eighteenth century (Morgan
and Richards 1993). The evaluation of these seedlings was the major achievement
of United States pomology in the nineteenth and early twentieth centuries. The
Delicious apple (discovered in Iowa, 1872) still dominates United States production,
and Golden Delicious (found in West Virginia, 1905) dominates Europe. Other
cultivars are important regionally such as McIntosh in the northeastern United
States and eastern Canada, Jonathan in the Midwest, York Imperial in the
Shenandoah Valley, Granny Smith in Australia, Coxs Orange Pippin in England,
and Belle de Boskoop in The Netherlands.
The origin of controlled breeding of apples is attributed to Thomas Andrew
Knight (1759-1838) who produced the first cultivars of known parentage. This
technique continues to be the basis of all present day apple breeding programs but
the process until recently had been notorious unsuccessful as compared to other
fruits. In retrospect, the reason appears to be the poor choice of parents. In the first
half of the twentieth century the only important US cultivars derived from
controlled breeding were Cortland (McIntosh Ben Davis) and Idared
(Jonathan Wagener), despite the fact that hundreds of apple seedlings were
named. However, in the last 25 years, seedlings derived from controlled hybridization began to find a place in world apple production. These include Elstar, Gala,
Jonagold, Mutsu, and Pink Lady (all seedlings of Golden Delicious) and
Empire and Fuji (seedlings of Delicious). Of 25 cultivars derived from crosses
developed in Japan, nine (36%) are seedlings of Golden Delicious (Bessho et al.
1993). Ironically, Braeburn, the newest sensation from New Zealand, is a chance
seedling.
The introduction of resistance to apple scab incited by Venturia inaequalis (Cke.)
Wint. by incorporating the Vf gene from Malus floribunda 821 has indicated the
potential of interspecific crosses to introduce new genes (Crosby et al. 1992). The
exploitation of naturally occurring mutations (budsports) in adapted cultivars has
also been an important technique for improving apples. While most mutations are
selected as improved color sports, mutations that change fruiting habit to compact
spur-bearing farm have been very important in Delicious and Golden
Delicious. The new technology of gene transformation promises to be the next
technological breakthrough in apple breeding. This technique offers the possibility
of introducing insect and disease resistance to established apple cultivars in a single

MODERN BREEDING OBJECTIVES

step, although thorough testing will be required. Sonic genetic transformants of

apple are now in the testing stage.

MODERN BREEDING OBJECTIVES

The apple is grown as a composite tree consisting of a rootstock and a fruiting
scion, and occasionally a three-part tree that includes a genetically distinct trunk or
interstem. Thus, genetic improvement must involve both rootstock and scion. In
the 1980s there was considerable revival of interest in planting self-rooted trees, a
prospect made commercially possible by improved micropropagation methods. Loss
of the growth-regulating chemicals necessary to control tree growth in the orchard
has virtually eliminated interest in own-rooted apples but the concept is attractive
for very high-density orchards.
Rootstock problems tend to be either catastrophic or debilitating. For example,
fire blight of Malling (M.) 26 has caused huge losses in the United States, and the
tree decline encountered in many plantings of the Mark rootstock has caused serious
economic problems worldwide. A new wave of clonal rootstocks, capable of
surviving and thriving under a wide range of environmental conditions, inducing
precocity, productivity, and good fruit qualities in the scion cultivar, is urgently
needed in many parts of the world. In many areas of southern Europe there is
almost universal use of M.9 rootstock, but a slightly smaller tree is considered
desirable.
As a result of the long life of the standard orchards, only about 3% of bearing
trees are replanted each year. Thus the cultivar picture in the industry has changed
very slowly, despite the fact that there can be rapid change in newly planted cultivars.
For example, 26% of all apples planted in Washington State in 1992 were Fuji.
Although marketers of apples tend to be conservative and prefer to deal with no
more than a dozen cultivars, the appeal of new cultivars is very strong to growers
because prices can be very high for new cultivars that have captured the imagination
of buyers and the public, especially when the supply is limited. Nurserymen who
sell trees are constantly seeking an advantage by controlling the source of new
cultivars and of new sports of old cultivars. The objective of genetic improvement
is to increase marketability of the fruit (or tree in the case of ornamentals) and to
introduce traits that reduce production costs.

Increasing Marketability
Marketability of apple cannot be simply defined because of the many uses of the
fruit. There is, in fact, not one market but many, e.g., fresh, stored, or processed;
local market, commercial market or export. The market is dynamic, rather than
static, and quality criteria vary. This may be a real problem if the breeders concept
of quality differs from that of the marketplace. The largest market is for dessert
apples that are sold on the basis of appearance (size, color, shape, and freedom from
blemishes) and quality (taste and mouthfeel). Apples are produced in a narrow

APPLES

window (typically three months depending on maturity), but are consumed year
round. Thus, a suitable cultivar that will become important in the industry must be
able to be stored and have a suitable shelf life after storage. The world apple market
has been reviewed by ORourke (1994).
Traditionally, the breeder has aimed at the commercial market, now primarily
large supermarket chains. There are two other major consumers of new cultivars,
both serviced but casually by the breeder. the backyard fruit gardener (very important
in Europe) and the direct-to-consumer orchardist. Freiherr von Berlepsch and
Ontario have never made a place in the commercial market, but in Germany both
are important for these specialty outlets. In the United States, scab-resistant cultivars
have not yet made an impact on the commercial market but Liberty, Redfree,
and Jonafree are becoming increasingly important for hobbyists and organic
growers; newer introductions such as GoldRush, Enterprise . and Pristine will
enhance this trend.
It is clear that market targets must be identified by the breeder and that each
market offers a valid niche that should be filled. But the breeder is responsible for
introducing at least a rudimentary production system with each cultivar introduced.
Such an orchard system should include identification of potential faults of the
candidate and reasonable cultural practices to overcome these faults.
Fewer and fewer apple cultivars are destined exclusively for processing, although
in Europe, special high-tannin cultivars of apples are used to make cider, an alcoholic
beverage virtually unknown in the United States. Apple cultivars suitable for the
processing market require heavy productivity, large size, with quality determined
based on the processed product (typically high soluble solids, a specific sugar acid
ratio, yellow flesh color, nonbrowning flesh, and an apple taste. Outward
appearance is unimportant. There are also niche markets. These include small
markets for very early fruits and apples with special attributes such as all-russetted
types, unique flavors, and special size (from crabapples to oversized).
Conventional breeding has concentrated on quality traits. However, the definition
of fruit quality is in some ways difficult to explain, even ephemeral. The problem is
complicated because quality depends not only on genetic factors but is enormously
influenced by production practices, orchard management, and climate. The
traditional technique of increasing quality has centered therefore on improving size,
appearance, and storage characteristics. In the 1920s and 1930s there were more
that 40 apple breeding programs in North America and many more in other apple
producing regions. These programs, based on intercrossing existing cultivars, have
resulted in introductions such as Cortland, Empire, Fortune. Fuji, Gala,
Idared. Jonagold, Macoun, Melrose. Mutsu, Pink Lady, and Spartan.
Individual factors of quality will be discussed in more detail under Breeding for
Specific Characters.
Regional preferences have had a tremendous effect on defining quality and
determines which new cultivars will dominate the market. For example, the
preference for McIntosh in the northeastern United States is almost unique, not
shared by the rest of the U.S. and European market and selections based on

MODERN BREEDING OBJECTIVES

McIntosh have been summarily rejected outside of this area. Similarly, Asian
markets prefer mild to sweet, low-acid apples, while northern Europe and the U.S.
Midwest appreciate tartness.
Appearance is also subjective. The market demands blemish-free apples, and
any amount of russet (except for completely russetted apples, which still have a
small niche market outside of the United States) is considered a serious defect even
though it has no adverse effect on eating quality and in fact may have a positive
effect. In the past, green skin was considered suitable only for cooking apples, but
this has changed with the acceptance of Granny Smith. The dessert market is now
based on yellow, red, green, and pink, but other colors and color combinations are
possible. Bicolor (or tricolor apples such as Coxs Orange Pippin) are popular in
Europe, especially yellow apples with a distinct red blush, and red color over a
yellow ground color. Red and green or red over green which leads to a dark liverish
color typical of some Delicious strains, is considered unsuitable in Europe. The
United States prefers washed color while Europe prefers striped apples.
Fruit shape is important. Ovate or conic apples are preferred; globular apples
are acceptable, but flat, oblate apples are now considered unacceptable (for no rational
reason). The typey shape of the ovate Delicious with a lobed calyx end is highly
acceptable for this cultivar, and special sprays (gibberellin and auxin) are applied in
the eastern and midwestern United States to obtain suitably shaped fruit.
In general, the preference is for large apples, 70 to 85 min (2 3/4 to 3 1/4 inches)
in diameter; in Japan even larger apples are preferred. Apples smaller than 57 mm
(2 1/4 inches) in diameter, often termed schoolboys in the United States, are
unacceptable; apples below this size are called crabapples and have essentially no
value except for juice. Of course, many consumers prefer small apples, but this is
not reflected in the market price. Clearly, a special marketing and packaging program
for small apples is needed.

Reducing Production Costs

Introducing traits that reduce production costs is another way to achieve profitable
apple growing. One strategy is to achieve high yields through a combination of
adaptability, precocity, productivity, and harvest consistency from year to year.
Production costs can also be reduced by increasing the case of harvest (through
dwarfing rootstocks) and disease and insect resistance. Apples that are deficient in
these qualities can still be grown but must be propped up by cultural practices,
many of which are expensive and some of which are considered to be undesirable
for the environment and public health. For example, McIntosh, an apple with
unusual taste qualities, suffers from early drop, soft fruit, and poor color. For many
years this problem was solved by the grower with preharvest spays of an antigibberellin growth regulator daminozide, sold under the trade name Alar, which
kept the apple on the tree to increase color without loss of firmness. This chemical
prop was lost when daminozide was banned after high publicized but unsubstantiated
claims that it was carcinogenic. Although McIntosh has not disappeared, new

10

APPLES

plantings have sharply declined and the cultivar in being replaced by Empire and
other new cultivars. Attempts are underway to transform McIntosh with genes to
reduce ethylene accumulation to diminish this problem.
In marginal areas of apple production, breeding is essential to obtain adaptability.
Objectives include cold hardiness for climates with extreme winters, low chill
requirements for subtropical climates, and sunbwn resistance and/or very late
ripening for hot, and climates. Genetic disease and insect resistance we also required
for those areas where other means of control are nonexistent, unavailable, too
expensive, or unacceptable. Disease-resistance breeding has been an important
strategy to obtain adaptability and reduce production costs.
The consuming public buys apples on the basis of name, appearance, quality (so
far as can be judged), and price. While apples that can be produced more efficiently
might presumably be sold at a lower price, the consumer has no way of knowing
this. The breeder should be aware that apples with pest resistance, while of interest
to growers and to hobby and home orchardists, have had, until recently, little interest
to the average consumer and thew apples must compete with susceptible apples an
a strict quality basis. However, some consumers have become interested in organic
apples, fruit that is presumably grown without certain nonorganic pesticides, and
disease-resistant apples could become a potential selling point for this market. The
bottom fine, however, is that whatever the improvement, the final competition of
apples at the marketplace is based on fruit characters alone.

BREEDING TECHNIQUES
Floral Biology
The flowers of apple cultivars and seedlings vary considerably in size and petal
shape, and in color from white to deep pink. They are produced at the same time as
the leaves, and we borne in cymose clusters on fairly short pedicels, usually on spur
type growth, but in some instances from the termirial or lateral buds of the previous
seasons growth. The typical flower consists of five petals, a calyx of five sepals,
about 20 stamens and the pistil which divides into five styles. The ovary has five
carpels, each usually containing two ovules, so that in most cases, the maximum
seed content is 10 but some cultivars have more. Liberty and Northern Spy, for
example, usually have 12 to 18 seeds and the Ottawa 3 rootstock often has 20 to 30
seeds.

Pollination
Pollination is the mechanical transfer of pollen from the anthers to the stigmas and
is a prerequisite to the fertilization of the ovules and the development of seeds and
fruit. Most apples require cross-pollination, which necessitates the transfer of pollen

BREEDING TECHNIQUES

11

from the flowers of one tree to the stigmas of a genetically different one. In the
orchard, pollination is carried out by insects, particularly bees, in their search for
pollen and nectar. Most flowers are admirably designed to facilitate this process,
but in a few genotypes, such as Delicious, the stamens are long and the styles
short, which allows the bees to visit the flowers without making contact with the
stigmas. This character may lead to a considerable loss of the potential crop.
Pollen collection. The collection of pollen is the first part of the hybridization
process. Flowers are collected at the balloon stage just before the petals expand,
and before the anthers dehisce, although circumstances often demand that they are
collected a little earlier. Anthers are removed by rubbing flowers over a screen and
are deposited into petri dishes or similar containers to dehisce. Putting the collected
anthers in a paper boat and leaving them overnight under an incandescent lamp is
a satisfactory method. The pollen once dry can be used immediately. Since not all
cultivars flower at the same time (weeks my elapse between the flowering of the
earliest and the latest), pollen may be collected from the earlier parent and stored
until required. Pollen may be forced in the greenhouse by cutting flowering shoots
and leaving them in a vase with water. Dried pollen can be conveniently mailed or
dispatched in polyethylene packets, making it easy to share germplasm. Pollen in
loosely stoppered vials placed in a desiccator containing calcium chloride can be
stored for a year or more at 15C. Apple pollen can be preserved longer at cryogenic
temperatures (Hanna and Towill 1995).
Viability can be estimated microscopically after staining with acetocarmine
(Marks 1954) or germinated on a drop of sugar solution (2.5 to 20%) to calculate
germination. Most recommend 10% sucrose in aqueous solution or 5% agar at
20C for good pollen tube formation. The addition of 10 ppm boron to the medium
increases germination (Thompson and Baijer 1950).
Emasculation. Before crossing, the flowers of the seed parent are usually
emasculated at the balloon stage. Removal of the petals in the emasculation process
serves three purposes: (1) it prevents self-pollination (although self-incompatibility
makes this a rare occurrence); (2) it exposes the stigmas to facilitate cross pollination; and (3) it minimizes visitation by bees and pollinating insects attracted to the
petals, reducing contamination with unknown pollen. When hybridizations are being
made for genetic studies stricter precautions are needed to prevent contamination,
and emasculated flowers are often bagged. If crosses are part of a breeding program
with the sole purpose of producing improved cultivars, then 100% control is not
essential. Emasculation is usually achieved by flicking off the sepals, petals, and
stamens with the fingernails, which is easiest at the balloon stage. Only the styles
are left sticking up from the receptacle and ovary. Other emasculation methods
include the use of scissors with a notch cut in the blade (Barrett and Arisumi 1952).
At Cornell University, Geneva, pollinations are usually made immediately after
emasculation and again a day or so later, especially when weather conditions are
marginal.

12

APPLES

Apple trees produce flowers in grew abundance and only a small proportion of
them can set and produce fruit. Typically, two flowers per cluster are emasculated
and the others removed. Unemasculated flowers producing fruits can be identified,
because fruits from emasculated flowers have no calyx. It is convenient to emasculate
first, leaving an unemasculated flower as a flag to identify the cluster, this flag is
removed at the final pollination. Experiments have shown that insects do not visit
flowers when the stamens and petals have been removed, thus there is no need to
protect them from insects (Visser 195 1).
Other techniques can be utilized to achieve cross-pollination. One is to eliminate emasculation entirely to avoid injury, but to bag flowers (pollinated in the balloon
stage after manually opening the petals), to prevent visitation from insect pollinators.
The crossed cluster must be tagged because the telltale belly button appearance
of the calyx end of crossed fruit derived from emasculation is lacking. This technique
is common in Europe. Another technique is to enclose a seed tree in bee-proof
screening and to enclose a bouquet of open flowers of the desire pollen parent and
insert a small package of honeybees. But if too many bees are provided they will
harvest pollen form the target stigmas and lower fruit set. Another approach is to
provide bouquets of a single cultivar that contains a dominant marker gene that
can be identified in the seedling stage in an orchard of several cultivars. For example,
if bouquets of a disease-resistant apple such as Liberty or GoldRush are placed
in an orchard of Fuji and Gala, seeds collected from trees close to the bouquet
could be expected to have a small percentage of the appropriate hybrid that could be
identified by screening in the seedling stage. A final technique is to collect openpollinated seed from orchards that contain two cultivars of interest. For example,
Empire was selected from open-pollinated seeds obtained from an orchard
containing only McIntosh and Delicious in a year when all hand pollinations
failed.
Crossing. Pollination can be achieved by dipping a small, soft brush into a vial
containing the pollen and lightly brushing the stigmas. The brushes can be cleaned
by immersing them in 95% alcohol and drying them before applying a new pollen.
Another technique is to dip the eraser end of a pencil or the fingertip into the pollen
and touch the stigmas to transfer the pollen. The finger technique uses a lot of
pollen but the pollinator can see the yellow pollen and can feel the touch of the
stigmas. Either way is quick and effective. Some pollinate flowers immediately
after emasculation; others prefer to emasculate and return in a few days and make
the pollinations when the stigmas are more receptive. It is best to avoid pollination
when temperatures are below 50C because pollen tube growth may be inhibited.
The whole tree may be pollinated with one parent or a number of crosses may be
made on different parts of the tree, providing limbs are tagged carefully to prevent
mix-ups.
In an effort to improve seed set and seed germination, Keulemans et al. (1994)
examined the influence of the number of pollinated stigmas per flower, the number
of flowers per cluster, and pollinator. Seed set increased with the number of pollinated
stigmas on the flowers. Compared to one flower per cluster, seed set was slightly

BREEDING TECHNIQUES

13

reduced when three flowers per cluster were left. The pollinator had no clear effect
on final fruit set, but genotypic effects were noted. For all pollinators, seed
germination was negatively correlated with seed number. A delayed harvest reduced
seed germination.

Seed Handling
Crossed fruit should be harvested slightly before it ripens. Dry seeds can be stored
if they are not to be planted in the following year. Seed storage has been reviewed
by Ellis (1985). Apple seeds will not germinate unless stratified. This involves
keeping the seeds in moist conditions and subjecting them to a period of cold to
allow after-ripening, during which embryo changes occur. The easiest technique
when seeds are not stored is to leave seeds in the fruit and store them in cold but
above freezing temperatures where they stratify naturally, but moldy core can cause
problems. Seeds should be removed from any damaged or rotting fruit at harvest.
The usual technique in most breeding programs is to extract seeds slightly before
fruit maturity. This avoids losses due to pilfering or mix-ups due to droppinga
problem when more than one cross is made on a tree. At Cornell University, Geneva,
individual fruits are labeled well before ripening time on trees in which several
pollens have been used. After-ripening will proceed at temperatures between 0 and
10C, but the optimum is from 3 to 5C. The time required may vary from 6 to 14
weeks and depends to some extent on the temperature. Abbott (1955) has shown
that the temperature for after-ripening is critical. Above 17C the after-ripening
process reverses and the longer seeds are held above this temperature the longer the
period of cold required to allow them to germinate. Seeds can be stratified in
polyethylene bags containing moist filter paper or moistened peat moss, which is
slightly fungistatic. The stratification process is usually completed in 6 weeks but
should be checked periodically for radicle emergence. When 50% or more of seeds
have germinated they can be planted in seed boxes or trays with individual cells,
and placed under optimum conditions for seedling growth.
Rots may be troublesome during stratification, particularly if the seeds are not
clean when put in the trays. Rhizoctonia solani Kuhn, the principal culprit, grows
on the outside of the testa and in due course penetrates and causes the embryo to rot,
quickly spreading from seed to seed. It is always advisable to surface sterilize the
seeds in calcium hypochlorite solution (10 g in 140 mL water and filtered) for 5
min, and then wash them before stratification. At Cornell University, Geneva, seeds
are soaked overnight in a saturated suspension of Captan fungicide before stratifying
and a second fungicide treatment is often necessary after planting.

Juvenility
Apple trees undergo different phases of development between seed germination
and the adult fruiting tree. In the juvenile phase, when no flowers are produced,
plants may differ considerably from mature adult trees. The leaves are smaller and

14

APPLES

usually more finely serrate, and the shoots are thinner and are often produced at
right or obtuse angles to the main stem. Bud break is early and leaf fall is late. The
onset of the adult phase is marked by the development of flower buds. Between
these two phases there is a transition phase when the lower part of the plant is still
juvenile and the upper part is adult. Cuttings from most adult apple trees used as
scion cultivars are extremely difficult, if not impossible, to root. Cuttings from
young seedlings root readily but rooting is more difficult to achieve as the seedlings
age, suggesting that the transition is gradual. The duration of the juvenile phase
varies from 3 to 10 or more years, depending on the genotype of the seedlings and
the cultural practices.
Ways of shortening the duration of the juvenile phase have been investigated
(Kemmer 1953; Murawski 1955; Visser 1964). Methods attempted to induce adult
trees to flower include shoot priming, root priming, and bark ringing. However,
any method that restricts growth in the very young seedling does not shorten the
duration of the juvenile phase but in fact lengthens it and delays the onset of fruiting.
Cultural methods that check the growth of seedlings are effective in hastening
flowering only when a certain stage of development has been reached. Way (1971)
was able to induce seedlings to flower by bark ringing when they were 4 years old,
but 3-year-old seedlings did not respond.
Other cultural methods can effectively shorten the juvenile phase. One is to
grow the seedling as fast and vigorously as possible. This can best be achieved by
avoiding any check to growth in the early stages of development and, where this is
not possible, to keep disturbance to a minimum. Seedlings have flowered in the
greenhouse in I year by keeping them in a continuous state of growth (Aldwinckle
et al. 1976). Seedlings grown under optimum conditions in the greenhouse, closely
planted, watered, and fed, may reach 3 m in height in the first season compared with
I in for seedlings planted out in the field. These may be planted directly into
permanent field locations. The nursery should be avoided because it delays the
process. Seedlings should continue to grow freely and, since the seedlings first
come out of the juvenile condition at the top of the seedling, they should be left
unpruned until they flower and fruit (Brown 1964; Zimmerman 1971). Tydeman
and Alston (1965) have shown that the juvenile phase can be considerably shortened
by budding seedlings onto dwarfing rootstocks. Buds were taken from the upper
part of the main shoot in late surnmer of the second years growth and worked on
the dwarfing rootstocks MY and M.27, grown as closely planted cordons with all
lateral growth pruned back annually in late summer. Nine years after germination
88% of 902 budded seedlings had fruited compared with 49% of those on their own
roots. Virus-free rootstocks must be used to avoid infection.
Juvenility poses a different sort of problem for the rootstock breeder. In the
stoolbed, juvenile plants are usually quite thorny, root very easily, and exhibit earlier
bud break and later leaf fall. Evaluations in the rootstock program at Cornell
University (Geneva Station) are usually made in the third season, when these juvenile
characteristics have somewhat abated.

BREEDING TECHNIQUES

15

Seedling Selection
As soon as the seeds have germinated and the first true leaves are showing, the long
process of seedling evaluation may begin. Some of this selection occurs naturally.
A pale green lethal recessive is carried in Golden Delicious, M.9, and many
other cultivars (Klein et al. 1961) and 25% mortality can be expected when these
cultivars are intercrossed. Negative decisions should be made as early as possible
because every delay in decision-making adds to the costs of eventual introductions.
Even more important, delay limits the attention that should be given the best
candidates and so limits the numbers of such elite candidates that can be evaluated
(Cummins and Aldwinckle, 1988).
The seedling stage is the best time to select for resistance to diseases where
infection occurs on young leaves, such as scab and cedar apple rust (Gymnosporangium juniperi-virginianae Schw.). Inoculations with spore suspensions at
this early stage show infections on the young leaves in a few weeks and the susceptible
seedlings can be discarded. Selection for scab resistance alone can reduce the size
of the progeny by 50 to 80% depending on selection criteria. Fire blight inoculation
of seedlings 50 to 90 cm tall with strains of Ervinia amylovora (Burrill) Winslow et
al. will eliminate a large percentage of most progenies. Some other diseases such as
mildew (incited by Podosphaera leucotricha [Ell. & Ev.] Salm) are better left until
the seedlings are in their second year of growth before being assessed for resistance.
It is important to be able to discard inferior seedlings at the earliest possible stage of
growth.
Many growth characters can be evaluated in the early years of development of
the seedlings, such as tree weakness, the production of long spindly shoots, and
other defects in general growth that would prevent the seedling from ever making a
satisfactory tree. Any such defect should be enough to eliminate the seedling. Some
correlated characters can also be detected, such as late leafing, which is associated
with late flowering, and early selection can be made when this is one of the breeding
aims. Similarly, the rootstock breeder can use very early screening for susceptibility
to Phytophthora spp., to fire blight, and to woolly apple aphids to reduce the original
population of seedlings by 95% before commencing any horticultural evaluations.

Preselection
Searches are always being made for characters in the juvenile stage that are correlated
with fruit characters in the mature tree. Any positive correlation would be of
considerable value in allowing seedlings with undesirable fruit characters to be
discarded at an early stage. Such correlations must fulfill two requirements: (1)
The degree of correlation must be of a fairly high order, and (2) the character in the
juvenile plant must be reasonably easy to recognize, for any time-consuming series
of measurements or complicated chemical analyses greatly reduces the value of the
correlation as a practical means of preselection.

16

APPLES

A number of correlations have been claimed but few are of any value to the
breeder. Nybom (1959) found a correlation between the pH of the leaf sap and the
pH of the fruit juice, with leaves of the sweet type at pH 5.7 and the acid types 5.5
to 5.6. This has not proved to be a reliable method because the differences are so
small that the error between samples from the same seedling are often greater than
between seedlings. Another correlation that is often claimed is between red fruit
color and anthocyanin pigmentation of other parts of the tree such as 1-year-old
shoots or leaf petioles. The distribution of anthocyanin in the tree depends on so
many factors and there are so many exceptions to these correlations as to make this
of no value as a method for predetermining fruit color. Only in progenies from trees
that are pigmented throughout the tissue with purplish red anthocyanin, as in Malus
pumila niedzwetzkyana Schneid., can red-leafed seedlings be selected that will
produce trees with red leaves and shoots and fruit with purplish-red skin and flesh.
Some useful correlations have been established between late leafing and late
flowering, between length of juvenile period and fruiting age in the vegetative
propagules, and others that will be discussed under the appropriate character
headings. Selection for tree type might reasonably be expected to be selected for in
the early seedling stage but experience with evaluation of progeny of McIntosh
Wijcik, an extreme, highly spurred mutant heterozygous for the dominant columnar
gene (Co), indicates that this is dangerous. In the future, preselection will be made
by associating important traits with molecular markers. This is discussed in greater
detail under Biotechnology.

Fruit Evaluation
When the seedlings produce flowers and fruit, the most important part of seedling
evaluation begins: assessment for fruit quality. Increasingly, researchers are looking
for methods of quality assessment that will correlate with consumer preferences
(Redalen 1988). There are many criteria by which the fruit can be judged; to fall
below the standard in any of these will automatically eliminate the seedling. Size is
one that can be easily assessed. Fruits consistently below 60 min in diameter are
too small for a commercial dessert apple. Color is important, with preference for
clear yellow, bright red, bright green, or bicolor fruit. These should be free from
russet, although there may be a place for sour completely russet apples. Dull-looking
apples are not worthy of further consideration. Shape has become important.
Irregularly shaped apples and especially very oblate apples will not be accepted by
marketers. Flavor is of prime importance. Many seedlings have fruit with very
little acid and in consequence they are very sweet and insipid. This type is
unacceptable except in some Asian markets. Fruit may have a strongly aromatic or
distinct anise-like flavor and, while these may prove to be good home garden apples,
they are unacceptable as commercial apples because such flavors are not universally
appreciated. The ideal is sub-acid with a pleasant flavor but there is increasing
demand for apples with complex, spicy, and full flavor. Of particular importance is
the flavor after storage. Apples that develop an aldehyde or off-flavor will be
unacceptable. Apples that are too mild will often taste bland after storage because

BREEDING TECHNIQUES

17

acidity decreases. The texture and mouthfeel of the flesh should be firm and crisp
with plenty of juice. Crisp flesh should persist in storage; soft apples are
unacceptable. Overly thick skin can be a problem as well as skin that is too sensitive
to bruising, such as Sir Prize.
When the fruits of the seedlings have been examined and those that meet all the
requirements have been selected, the next stage of selection is for yield. Care must
be exercised to be sure that each selection is kept virus-free. For this a number of
trees need to be propagated on commercially important rootstocks and grown in
properly conducted trials with the best standard cultivars as controls and crop weights
determined and grade distribution assessed. At the same time, the selection can be
topgrafted on an existing tree so that fairly large samples of fruit can be obtained
and be subjected to storage trials and examined for susceptibility to storage diseases.
Where possible, the selected seedlings that show real promise should also be grown
in trials in a number of localities to see if their behavior is consistent under varying
conditions. In these trials, only those selections with heavy annual yields need be
considered further.
During the selection process it is important to be absolutely ruthless and to
eliminate from further consideration any seedling that does not reach the required
standards. Carrying out trials on selected seedlings that never make the grade is a
waste of time and money. No seedling is perfect in every character but the breeder
needs to strike a balance between eliminating everything and releasing too freely.
Because apples may perform differently in different locations, wide testing of
promising seedlings is essential.

Records
An important part of any breeding program is record-keeping. What records are
kept and how detailed they are depends entirely on the amount of information
required for future use. Descriptor lists for apple are available (Watkins and Smith
1982). Selection indices based on a number of characters have been proposed by
Blazek and Paprstein (1988). It is always tempting for the breeder to engage in
genetic studies, but many breeders have become bogged down with record-taking
at the expense of moving vigorously forward to the goal of cultivar improvement.
In general, it is more efficient to carry out genetic studies on specific crosses because
extracting voluminous data from all crosses is inefficient. There are a number of
ways of making field records. One convenient system is to use a predetermined
numerical classification, typically on a I to 5 or 0 to 9 scale to save time and space.
Verbal descriptions are not amenable to computerization. The details of a seedling
can be contained in one line across a scoring sheet.

Mutations and Chimeras

A bud sport is a mutation arising in a cell from which a bud develops. It produces a
shoot that differs, usually in only one character, from the plant on which it was

18

APPLES

produced. These mutations can affect any part of the plant. There are two important
types of mutations in apple: (1) those that produce single gene differences in some
character of the tree or fruit, and (2) those that alter ploidy. Mutations affecting the
appearance of the fruit are easily recognized and are the type most often found.
Increases in the amount of anthocyanin in the outer cell layers of the fruit skin are
most common and red sports of many of the popular cultivars have arisen. Some
cultivars such as Coxs Orange Pippin, Delicious, Estar, Gala, Jonagold,
Rome Beauty, and Winesap are prone to produce mutants of this type, while
other widely grown cultivars are stable and seldom seem to mutate. Not all the red
sports of one cultivar are necessarily identical; apart from differences in the area of
the fruit covered, the intensity of the pigmentation in the outer cell layers can differ
(Dayton 1959). Very often the mutation is limited to one cell layer in the apical
meristem; therefore, the plant is likely to be a periclinal (hand-in-glove) chimera.
The mutation is usually not heritable unless the second layer (L-2). which gives
rise to gametes, is involved (Pratt 1983). Bud sports with increased red color may
be extremely valuable and should be propagated and trialed to see if they are superior
to the original clone. Reduced russet, russet-free, and total russet occurs and some
of these traits may be desirable.
Mutations affecting growth habit, particularly the spur or compact types, which
produce compact or dwarfish, freely spurring trees, may be extremely valuable and
are being sought in all the important commercial cultivars. Mutations inferior to
the original clone my also occur and stock trees should be carefully observed so
that no inferior mutants are unwittingly propagated. This is the basis for bud selection
to maintain the integrity of the cultivar.
The rate at which single gene mutations occur can be increased by irradiation
with X-rays, gamma rays, or thermal neutrons. Bishop (1959) produced two dark
red sports of Cortland, two sports of Sandow with less color and more russet,
and a Golden Russet with considerably less russet than normal. Lapins (1965)
and others have produced compact mutations by means of irradiation. The effective
dosage is 3-5 krad for dormant scions and 2-4 krad for summer buds when X-rays
are used and 3.9-15.6 1012 thermal neutrons/cm2. Mutants from irradiated material
do not readily show themselves and the normal growth if allowed to develop will
suppress the mutant. When dormant scions have been treated and grafted, the basal
ten buds or so of the shoot emerging from each treated bud should be removed and
budded onto a dwarfing rootstock. The mutants are likely to show in the second
vegetative generation. Lacey and Campbell (1987) reviewed the production and
selection of mutant apples and described the tests and the experimental evidence
used to determine the nature of the mutated plants.
The other mutations important in apple breeding are the large or giant sports
(Einset and Imbofe 1947, 1949, 1951) due to changes in ploidy level, typically
chromosome doubling (diploid to tetraploid or triploid to hexaploid). These sports
are usually recognized first by their large f1ruits, which are sometimes twice as
large (in volume) as their diploid counterpart. The fruits, apart from the increase in
size, are usually flatter and more irregular in shape and have no commercial value.
Giant sports of a number of our most widely grown apple cultivars have arisen

BIOTECHNOLOGY

19

spontaneously and have been discovered by observant growers. Most of these

tetraploids are periclinal chimeras and can be grouped according to the arrangement
of the 2x and 4x tissues. The four layers in the apical meristem (L-1 to L-4) can be
classified on the basis of nuclear size as either 2-2-2-2, 2-2-2-4, 2-2-4-4, 2-4-4-4, or
very rarely 4-4-4-4 based on whether the cell is diploid or tetraploid. The breeding
behavior of these sports differs according to the location of the 4x tissue. Only
when L-2 is 4x will the tree breed as a tetraploid. Thus many giant sports, apart
from having tetraploid growth characters, genetically behave as diploids. The 2-24-4 cytochimeras look like a tetraploid but breed as a diploid (Pratt 1983).
It is possible, however, to produce homogeneous tetraploid plants from 2-2-4-4
sports by inducing shoots to grow from endogenous tissue or from root suckers
developed from the scion because as roots develop from L-3. One method is to
grow 1-year-old trees of the giant sports in large containers, cut them down to 30
cm, and remove all the buds and any growth that appears in the region of the removed
buds or from the rootstocks. This encourages the formation of sphaeroblasts in the
internodal regions. These will in due course crack and produce adventitious buds
which develop into shoots. Having been developed from the region of the phloem,
these shoots should be 4-4-4-4. Although practically all the 4x sports have arisen
spontaneously, it is also possible to encourage the formation of tetraploids by the
use of colchicine applied to the growing point (Hunter 1954). These induced
tetraploids are also likely to be cytochimeral.

BIOTECHNOLOGY
Micropropagation
Information on tissue culture of apples has been reviewed by Zimmerman (1984,
1991) and Skirvin (1986). Micropropagation using shoot proliferation has been
achieved by standard protocols. Shoot proliferation medium is based on Murashige
and Skoog salts with minimal organics, and the addition of cytokinins such as 4.4
M benzylamino purine (BA) or 0.1 M thidiazuron, the inclusion of 0.5 M
indolebutyric acid (IBA) and 1.4 M gibberellic acid (GA) and 87.6 mM sucrose
using agar as a gelling agent. Rooting medium is modified by halving the
concentrations of salts and sucrose, eliminating BA and GA, and varying the
concentration of IBA. A simpler method is to place shoots for 37 days in the dark
in a liquid medium containing 43.8 mM sucrose and 1.5 M IBA at 30C. (Zimmerman
and Fordham 1985). Protocols vary with genotype and many are proprietary.
Tissue culture has been seen as a means to propagate fruiting cultivars on their
own roots but results have been variable. Own-rooted apples trees are vigorous and
branch freely. Flowering may be delayed as compared to planting budded trees
(which are usually older). Many-branched nursery stock will produce a crop in the
second leaf. Tree size of own-rooted trees will obviously vary with each genotype.
Apple rootstocks have been micropropagated in Canada, Italy, France, and the
United States, but cost is not competitive with stoolbed production. Thus

20

APPLES

micropropagation has been used to establish new stoolbed plantings with diseaseresistant clones. It is essential to establish new cultures annually to assure that no
genetic change has occurred during the micropropagation process. The chance of
mix-up or genetic change is ever-present and some nurseries have avoided
micropropagated rootstocks.

Somatic Embryogenesis
Embryogenesis has been induced from nucellar tissues from seeds (Eichholtz 1979)
and calli from various in vitro-grown apple seedlings or from flower buds and excised
petals from field-grow trees (James et al. 1994). However, there have been no reports
of complete development of viable seedlings from adventitious embryony.

Embryo Rescue
Development of immature embryos (8 days following pollination) have been achieved
(Zhang and Lespinasse 1988), but rescue of interspecific or intergeneric: hybrids
has not been attempted.

Protoplast Culture
Viable protoplasts have been isolated from shoot cultures of shoots, callus, and cell
suspensions of various apple genotypes and rooted autotrophic plants were reported
from MY and Spartan (Patat-Ochatt et al. 1988). Wallin and Johansson (1989)
regenerated shoots of a columnar apple (A 1583) from leaf mesophyll protoplasts.
There are no reports of successful protoplast fusions and subsequent regeneration
in apple.

Regeneration
An efficient regeneration system is a prerequisite for apple transformation (Skirvin
et al. 1986). Important factors that influence regeneration include regeneration
media, age, genotype and orientation of the explant; and incubation conditions (James
et al. 1988; Welander 1988; Fasolo et al. 1989). Regeneration has been obtained
from shoots and leaf segments (Welander 1988), leaf disks (Elobeidy and Korban
1988), and leaf strips (James et al. 1988). Leaves are most commonly used because
of their abundance and ease of manipulation for Agrobacterium infection (Jams and
Dandekar 1991). For maximal transformation and regeneration efficiency of apple
leaf explants, the shoot culture must be between 20 and 40 days of age (Bondt et al.
1994). Hyperhydricity (vitrification) must be avoided. Exposure to darkness for I
to 4 weeks induces regeneration (Welander 1988; Fasolo et al. 1989; Korban et al.
1992, Durham and Korban 1994).

BIOTECHNOLOGY

21

Prior to transformation, the effect of aminoglycosides on regeneration and growth

of explants must be determined because cefotaxime is used in transformation studies
to control growth of surviving Agrobacterium after infection. The stimulatory effect
on morphogenesis has been observed in apples (Maheswaran et al. 1992).

Molecular Markers
Biochemical markers such as isozymes have been extensively studied in plants and
many important agricultural characters have been correlated with isozyme
polymorphism (Weeden 1989). As a result of the high heterozygosity and high
level of polymorphism that exist in apple, isozyme techniques have provided a reliable
method for cultivar identification in scion and rootstock (Vinterhalter and James
1983, 1986; Weeden and Lamb 1985, 1987; Menendez et al. 1986; Weller and
Costante 1987; Manganaris and Alston 1989; Samimy and Cummins 1992). Evidence for the allopolyploid nature of the apple genome was demonstrated using
apple pollen enzyme system (Chevreau et al. 1985). Evidence that the maternal
parent generally provides the diploid gamete in natural triploids was demonstrated
with the isozyme 6-phosphogluconate dehydrogenase (Chyi and Weeden 1984).
Genetics and linkage analysis of isozymes has been studied in apple (Weeden
and Lamb 1987; Manganaris 1989) and isozymes have been used in the development
of linkage maps to anchor homologous linkage groups and to establish marker points
(Hemmat et al. 1994). Isozyme ldh-2 has been shown to segregate with fruit color
on group 3 of the apple linkage map (Weeden et al. 1994). A close linkage was
demonstrated between the genes for endophosphatase (ENP-1), acid phosphatase
(ACP-1), and the pale green lethal (l) (Manganaris and Alston 1988).
A number of commercially important traits have been linked to isozyme loci.
Self-incompatibility has been linked to GOT-1 (Aat-c) (Manganaris and Alston 1989),
and scab resistance (Vf) is about 8 cM from PGM-1 on the apple linkage group
(Manganaris et al. 1994). Powdery mildew resistance has been identified by two
closely linked isozyme markers ACP-3 and GOT-2 (Aat-p) (Manganaris 1989).
Although 19 isozyme markers have been mapped on the apple linkage map, the
method has its limitations. Those isozymes that show a well-defined metabolic role
and have highly conserved isozyme numbers are particularly useful (Weeden 1989).
However, the number of isozymes that can be assayed is limited. Most of them are
monomorphic and are of limited value for gene tagging.
The development of polymerase chain reaction (PCR) techniques has provided
a new method of detecting DNA polymorphism by amplifying specific DNA fragments. Random amplified polymorphic DNA (RAPD) fragments have been useful
and reliable for genetic mapping and gene tagging in many different plant species
including apple (Weeden et al. 1992). Decamer random oligonucleotide primers
are usually used in this method (Williams et al. 1990). The advantage of the RAPD
technique is that it can provide numerous genetic markers in a short time, making it
the method of choice for mapping economically important plants. A genetic link

22

APPLES

age map constructed from over 400 markers, mainly RAPDs has been developed
for apple based on a cross of Rome Beauty with White Angel Olemmat et al.
1994). Maps from other progenies are under development (Conner et al. 1995;
King 1995). In a study of two half-sib apple progenies, Conner et al. (1995) reported
on the usefulness of molecular marker-bail association in different genetic and
environmental backgrounds. Approximately half of the marker-trait associations
were detected in both progenies.
RAPD markers can be very useful for cultivar identification. Landry et al. (1994)
used RAPD marker data in phylogeny analysis of 25 apple rootstocks and developed
a DNA fingerprinting system and identification key for apple rootstocks. Koller et
al. (1993) reported 14 RAPD markers resulting from two decamer primers could
discriminate 11 apple cultivars.
RAPD markers arc proving useful in plant breeding (Williams et al. 1990; Welsh
and McClelland 1990). They enable apple breeders to tag any genome region early
in the seedling stage (Weeden 1989). One important advantage of this technology
is that a genetic map is not necessary in order to identify a trait of interest. The
correct selection of a cross for identification of a particular trait, however, is an
important factor in the search for markers. Bulked segregant analysis (BSA) is the
quickest approach in identifying markers in any population that a gene or a trait of
interest is segregating (Michelmore et al. 1991). Using the BSA technique in apple,
several tightly linked RAPD markers were identified for the scab-resistance gene
(Vf) derived from Malus floribunda 821 by various groups (Durham and Korban
1994; Koller et al. 1994; and Yang and Krilger 1994; Hemmat et al. 1995, Tartarini,
in press). S. Tartarini (personal communication) using several RAPD markers close
to the Vf locus determined the portion of the original Malus floribunda 821 DNA
present in each of 30 cultivars.
A map of the scab resistance region has been developed (Hemmat et al. 1995),
showing three RAPD markers closer to the Vf loci than the isozyme marker PGM-1.
These markers could be of great value for resistance selection and screening. Many
other morphological and developmental traits, such as terminal bearing, reproductive
bud break, bloom time, and root sucker formation, have been mapped within several
centimorgans of a RAPD marker on the apple linkage map (Lawson et al. 1995).
Markers for anthocyanin fruit color have been identified and demonstrated to be
present in many red apple cultivars (R. Cheng, unpublished). A RAPD marker for
the columnar habit (Co), which suppresses branching, was found to contain a simple
sequence repeat (SSR) that is highly polymorphic in apple (Hemmat et al. 1995).
This SSR has been napped on linkage group 11 of the Rome Beauty White
Angel linkage map. A marker (OPAT20) for powdery mildew (Podosphaera
leucotricha) resistance has been reported (Markussen et al. 1995) and was mapped
within a genetic distance of 4 cM from Pl1 gene from M. robusta. This marker
could provide a convenient method for selection of the resistant genotype in
progenies. Other complex traits such as branching habit, flowering time, vegetative
bud break, and acidity and the genetic basis of these characters are under investigation
using molecular markers and isozymes that offer great potential for assisted selection
in quantitative traits (Hagens 1992; Conner et al. 1995).

BIOTECHNOLOGY

23

Transformation
Apple transformation was first reported by James et al. (1989) based on
Agrobacterium-mediated transformation of leaf disks of Greensleeves using the
binary vector pBin6. Detailed transformation protocols have been published by
James and Dandekar (1991). However, Agrobacterium transformation systems must
be optimized for each crop species and indeed for each cultivar. Explants are taken
from the plants established in culture and co-cultivated with Agrobacterium carrying
the vector plasmid. Transgenics are regenerated under selective conditions. Although
transgenic apple plants have been produced, until recently efficiency has been low,
even with highly regenerable cultivars (James et al. 1993a,b), but rates of 20% or
more have been achieved with several cultivars (H.S. Aldwinckle, personal
communication). Important factors that influence transformation include pre-culture
conditions, the strain of the bacteria used, wounding of leaf tissue prior to bacterial
treatment, induction of the bacteria, co-cultivation conditions, explant, genotype,
the regeneration system, and conditions to select out transgenics (James 1991; James
and Dandekar 1991; Welander and Maheswaran 1991; Bondt et al. 1994). The
transgenes incorporated into apple via Agrobacterium-mediated transformation are
stably incorporated and inherited in a Mendelian fashion (James et al. 1994).
In apple, several genes have been targeted for transformation using
Agrobacterium-mediated gene transfer, including those for insecticidal crystalline
proteins (ICP) from Bacillus thuringiensis and the cowpea. trypsin inhibitor protein
(cpTi) (Ely 1993; James et al. 1993a); lytic peptides for bacterial disease resistance
(Norelli et al. 1994); antisense genes for reduction of softening (Gray et al. 1994;
Yao et al. 1995); cross-protection for viral resistance (Mauren et al. 1992); and
chitinase genes for fungal resistance (H.S. Aldwinckle, personal communication).
In apples, qualitative genes for resistance to major pests are available and are potential
targets for future improvements (Brown 1992). These include several dominant
genes for resistance to apple scab, Venturia inaequalls (Shay et al. 1953); resistance
to mildew, Podosphaera leucotricha (Knight and Alston 1969); and resistance to
woolly apple aphid, Eriosoma lanigerum. The Co gene that confers columnar plant
habit is also desirable (Lapins and Watkins 1973). Techniques are available to isolate,
identify, and introduce the disease resistant genes into existing cultivars (Michelmore
et al. 1992a,b; Michelmore 1995; Martin 1996). Cultivars transformed include
Greensleeves (James et al. 1989), Delicious (Sriskandarajah et al. 1994), Royal
Gala (Yao et al. 1995), McIntosh (Bolar 1995), and the rootstocks M.26 (Lambert
and Tepfer 1992; Maheswaran et al. 1992) and M.7 (incorrectly reported as M.26)
(Norelli et al. 1994).
The incorporation of insecticidal activity of proteins made by Bacillus
thuringiensis would be an effective strategy to control codling moth, one of the
major pests of apple. Chemical control of secondary infections of codling moth
also kill predators of mites and aphids; thus genetic control of codling moth could
have a major effect on other apple pests. At least 40 genes encoding ICP effective
against Lepidopteran, Dipteran, and Coleopteran insects have been isolated and

24

APPLES

their DNA sequences compared (Hfte and Whiteley 1989). Purified preparations
of 1CPs encoded by cryIA(b) and cryIA(c) genes have been shown to be very toxic
against codling moth (Vail et al. 1991).
The cryIA(c) gene was introduced initially using the binary plasmid pWB139
containing the active fragment encoding the insecticidal activity of the cryIA(c)
gene from the HD-73 strain of B. thuriengiensis (Dandekar et al. 1992). Low levels
of expression were observed among 27 independent transformed lines which were
the first field introduction of transgenic apple. A synthetic gene was subsequently
introduced that had a codon bias compatible with dicots, expressed at levels sufficient
to kill all the feeding codling moth larvae (Dandekar, personal communication) and
over 400 transgenic plants are under test.
Transgenic M.7 apple rootstocks have been obtained expressing attacin E gene
from the giant silk moth, Hyalophoras cecropia (Norelli et al. 1994). Greenhouse
and field results indicate that one transgenic line has increased resistance to Erwinia
amylovora the incitant of fire blight. Attacin E is one of a class of proteins referred
to as lytic peptides effective against bacterial infections (Bowman and Hultmark
1987). However, these proteins are rapidly turned over in plants and accumulation
to high levels may be required to increase their efficacy. Field trials are in progress
to evaluate MX and Royal Gala transgenics with attacin E and three other lytic
proteins (H.E. Aldwinckle, personal communication).

BREEDING SYSTEMS
The majority of cultivated apples are functional diploids (2n = 34) and 17 bivalents
form at meiosis (Lespinasse et al. 1976; Zhang et al. 1988ab,c). Although most
characters of apple that have been studied are polygenically controlled, the characters
under single gene control all demonstrate diploid segregation. There have been
suggestions (Darlington and Moffett 1930) that they are complex polyploids, partly
tetraploid and partly hexaploid, with the basic number of x = 7, which is common in
Rosaceae. The basic chromosome numbers of 8 and 9 also occur in Rosaceae, and
Sax (1931) put forward the theory that all the Pomoideae were allopolyploids derived
from a doubled hybrid between two remote ancestral types, one with the basic number
of 8 and the other 9. Dermen (1949), while agreeing in general with Sax, considered
that, because of the great diversity between genera in the Pomoideae, each genus
must have arisen in this way independently. Electrophoretic studies of the inheritance
of seven pollen enzymes in apple have indicated bigenic disomic inheritance for
five enzymes and monogenic control for two enzymes, results compatible with
allopolyploidy (Chevreau et al. 1985).
Triploidy occurs in apple cultivars (2n = 51). These have arisen naturally from
the fertilization of unreduced gametes. Among the cultivated apples, triploids appear
to be more common than one would expect, accounting for about 10% of the
commonly grown cultivars, yet appearing in populations from diploid parents only
at the rate of approximately 0.3%. A number of triploids are of considerable
economic importance, such as Baldwin, Gravenstein, Rhode Island Greening,

BREEDING SYSTEMS

25

Blenheim Orange, Stayman Winesap, and more recently Jonagold and Mutsu.
In fact, all of the promising seedlings that are assumed to derive from Winesap,
such as Stayman Winesap and Mammoth Black Twig, are triploid. It would
appear that there is more chance of a spontaneous triploid seedling being outstanding
than a diploid. They are more vigorous and tend to have larger fruits, which might
account for some of their selective advantage. Triploids, however, produce very
poor pollen and this leads to pollination problems in the orchard since it requires a
diploid to pollinate the triploid, and another diploid to ensure fruit on the diploid
pollinizer. The triploids produce few viable seeds; nevertheless they are sufficiently
fertile to produce a good crop of fruit if properly pollinated. The lower content of
seed that they contain (because of aneuploidy) might also contribute to annual
bearing. They were considered useless as parents for further breeding since, when
selfed or crossed, practically all seedlings would be aneuploid, which are weak and
seldom develop into trees. Of 329 seedlings from open-pollinated triploids; 2 (0.6%)
were haploid, 6 (1.8%) were diploid, 4 (1.8%) were triploid, and 10 (3.0%) were
tetraploid, while 307 (92.8%) were aneuploid (Einset 1945). Although triploids as
pollen parents can be expected to produce very few 34 chromosome gametes, the 2x
gametes being balanced will have a selective advantage over aneuploid gametes.
Hokuto, a triploid seedling, was selected from 79 seedlings resulting from the
cross between diploid Fuji and triploid Mutsu as the pollen parent (H. Bessho,
personal communication).
Tetraploids (2n = 68) have been discovered as large-fruited sports on trees of
diploid cultivars. Usually such tetraploid chimeras are of no commercial value.
Shoots are typically thick and sparsely branched with short internodes, and fruits
are large, often flattened, and sometimes misshapen. At least two tetraploid forms
of Yellow Transparent were significant improvements over the normal diploid
cultivar and were planted on a limited commercial scale. The tetraploid giant sport
Perrine Transparent produced a semi-dwarf tree carrying regular crops of wellshaped apples, 60 to 65 min in diameter, in contrast to the very vigorous, highly
biennial diploid Yellow Transparent with fruit 40- to 50-mm-diameter fruit.
Spontaneous tetraploids usually occur as chimeras with one or more layers of diploid
tissue over the tetraploid. If L-2 is tetraploid, the giant sport will produce segregating
2x gametes and the progeny in crosses with diploids will be triploid. By crossing
those that breed as tetraploids with diploids, large progenies of triploids can be
produced. However, these are not quite the same nor as valuable as triploids derived
from an unreduced diploid egg cell and fertilized by 1x pollen.
The explanation for the superiority of these natural triploids given by Knight
and Alston (1969) is that in triploids produced from 4x 2x parents, the gametes
from both parents have undergone meiosis and therefore have a random reassortment
of genes. Natural triploids, however, usually arise from the fertilization of an
unreduced egg cell by a haploid pollen grain. The pollen has undergone full
segregation, but the maternal gamete will have either undergone no segregation or a
minimum of reassortment and therefore contributes to the resulting triploid twothirds of the genomic constitution of the maternal parent intact or slightly changed

26

APPLES

from a parent that was probably a very successful cultivar. This preserves the essential
quality of the desirable maternal parent, which may be enhanced by an additional
genome as well as by the triploid condition. Jonagold (a cross of diploid Golden
Delicious diploid Jonathan) is a triploid produced from nonreduction in Golden
Delicious, but it is unclear if the complete genome of Golden Delicious is intact
or if some reassortment has occurred, perhaps, from second-division restitution in
meiosis.
The problem in taking advantage of nonreduction to produce triploids is that the
frequency of their occurrence is low. Thus, many seedlings have to be grown to get
a reasonable number of triploid seedlings from which to select. The method used
by Knight and Alston (1969) is to save seeds from discarded fruit from a packinghouse
where the seed parent is known and perhaps even the pollenizer. By selecting the
largest seeds from the mass of seeds saved, growing seedlings, and selecting for leaf
size and vigor, a number of triploids can be obtained without growing thousands of
seedlings.
Tetraploids can also be produced from triploids pollinated by diploids. The
number of good seeds produced from such crosses is very small and most produce
aneuploid seedlings. Laubscher and Hurter (1960) made root-tip counts of 884
seedlings from triploid cultivars growing in a mixed orchard of diploids. In all,
eight tetraploids were obtained, or one in 110 seedlings. Not all triploids respond in
the same way and some produce many more tetraploids than others (Einset 1952).
The tetraploids are presumed to arise from unreduced 51-chromosome egg cells
fertilized by 17-chromosome pollen, and it is suggested that tetraploids produced in
this way may be better parents for breeding than those from colchicine-treated
diploids. The first tetraploid cultivar to have been bred in this way is the Swedish
Alpha 68. Hexaploids have also been obtained by colchicine treating triploids; 63-3 and 3-6-6 chimeras have been produced of the cultivars Paragon and Stayman
Winesap (Dermen 1965). One disadvantage of many of these polyploid seedlings
is that they have very long juvenile periods (some of the natural triploids may be an
exception) and long juvenility is correlated with slowness to come into bearing
after propagation of adult material.

Sterility and Incompatibility

There are two main causes of unfruitfulness in the apple: sterility and incompatibility.
Generational sterility can be brought about by the failure of any of the processes
concerned with the development of pollen, embryo-sac, embryo, and endosperm.
This is evident in triploids where chromosome imbalance in gametes results in only
a small percentage of good pollen being produced and poor seed set. Some diploids
also have high amounts of aborted pollen, probably due to lethal recessive genes
present in the heterozygous condition in the sporophyte.
Sexual incompatibility that is due to the failure of the pollen to grow down the
style and bring about fertilization is widespread in apple. Self-incompatibility is
particularly common, although cases of cross-incompatibility are also known. Apples

BREEDING SYSTEMS

27

have a gametophytic incompatibility system whereby the pollen tube growth is

arrested in the style. Pollen germination is not noticeably lower in incompatible
pollinations than in compatible ones and is unrelated to subsequent pollen tube
growth. Research on the S allele of apple should aid in understanding the system
(Battle et al. 1995; Broothaerts et al. 1995).
Almost all apple cultivars are self-incompatible to some extent, and some are
completely so. Even those that appear to be self-compatible set more fruit with
higher seed content when pollinated with a cross-compatible cultivar. Brittain (1933)
selfed Coxs Orange Pippin, Golden Russet, and Northern Spy on a large scaleseveral thousand flowers of each-and the mean fruit set was about 1.5%, compared
to 7% from cross,-pollination. Cross-incompatibility is not very common, but a
number of cases have been reported, usually between close relatives. Sports of a
cultivar are cross-incompatible with the original form and with its other sports, but
this is essentially self-incompatibility.
Triploid cultivars behave in much the same way as diploids, varying in the degree
of self-incompatibility and producing an increase in fruit set when pollinated with
diploids, but varying considerably when crossed with other triploids. Natural
tetraploids are also very variable, some being possibly self-fertile and others only
partly so because gametophytic incompatibility is typically voided in diploid gametes.
In a series of crosses (Brown 1975), the following results were obtained: 2x 4x
gave 9% fruit set; 4x 2x gave 3% fruit set; 3x 4x gave 7% fruit set. Other
information suggests that 4x 4x sets fruit freely.

Apomixis
Some plants, while appearing to produce seeds in the normal fashion, actually
reproduce from unfertilized eggs, from diploid cells of the nucellus, or from some
cell of the megagametophyte. This form of reproduction is known as apomixis.
Some plants produce both sexual and apornictic seed. If seeds are produced from
the diploid maternal cells of the nucellus, the resulting plants will be identical to the
maternal parent and show no segregation. Thus, this type of apomixis is a form of
vegetative propagation. If seeds form from haploid eggs, the resulting plants will
show segregation and may be haploid or, if doubled naturally or with colchicine,
will give rise to a completely homozygous individuals. There are various techniques
to produce haploids. Haploids may occur spontaneously (typically one in a thousand
seeds), but they are very weak and generally do not survive. They can be selected
by pollinating with a homozygous dominant seedling marker and selecting for
seedling recessives. Using this technique, 13 haploids were identified (Lespinasse
and Chevreau 1987) and some of these haploids have been doubled. Other techniques
suggested include (1) in vitro androgenesis. by anther culture, (2) in vitro
gynogenesis by unfertilized ovule culture, and (3) in situ parthenogenesis by
irradiating pollen followed by in vitro culture of immature embryos (Hofer and
Lespinasse 1996). While progress has been made, with these techniques, have not
as yet been successful at producing viable haploids in apple.

28

APPLES

Homozygous plants of apple represent excellent experimental material for

cytological and genetic studies. but the breeding value of haploids has been
overstated. A random haploid or doubled haploid has no real value except as a
genetic model. Two random double-haploid apples when crossed will produce
uniform F1 progeny but this genotype will probably be no better than a random
heterozygous selection obtained from a diploid diploid cross. There is no evidence
to support expectations that natural selection for viable haploids will be related to
horticulturally desirable parents, although one might assume some natural selection
for vigor since deleterious recessives will be lethal in haploids. Further, it will be
much more difficult to select promising hybrids because the chance of selecting
two promising haploids will be extremely low and they must be tested by
intercrossing, and then their progeny evaluated, an expensive technique. This is a
very inefficient way to produce apple selections because superior genotypes produced
from heterozygous diploid heterozygous diploid can be fixed by vegetative
propagation. Seedling apples will also be juvenile, which will be a problem. This
method might be a promising technique to produce seed-propagated homogeneous
rootstocks because they would be free of virus, but it seems unlikely that this would
be less expensive that the present stoolbed production of virus-free clones. However,
the limitations cited for fruiting cultivars applies also to rootstocks.
Facultative apomixis is characteristic of a number of Malus species that are
probably of hybrid origin, but does not appear to occur among the cultivated apples.
The apomictic species that have been investigated are polyploids. Malus sikkimensis
is a triploid; M. coronaria, M. hupehensis, M. lancefolia Rehd., M. platycarpa,
and M. toringoides (Relid.) Hughes are known in both triploid and tetraploid forms;
M. sargentii is tetraploid, and M. sieboldii is known in diploid, triploid, tetraploid,
and pentaploid forms. It is probable that the so-called diploid forms of M. sieboldii
are sexual hybrids with other species.
Under normal circumstances these species reproduce themselves freely by
apomictic seeds but most of them can produce sexual hybrids if crossed with
sexual diploids. Sax (1959) has demonstrated that the tetraploid M. sargentii when
pollinated with sexual diploids produced maternal tetraploids and hybrid triploids
and pentaploids. Seedlings from these apomictic species are not necessarily identical
and a certain amount of variation can be found. The importance of this character in
Malus species is that the seedlings of some are sufficiently uniform to be used as
rootstocks that are virus free. There are certain problems of stock-scion
incompatibility and sensitivity to virus which will be dealt with in the section on
apple rootstock breeding. Hanna Schmidt (1958) has demonstrated that it my be
practical to create apomictic parents that will produce clonal rootstocks by seed.
Some of her M. sargentii M.9 selections are facultative apomicts, producing 60
to 90% maternal clones. These could be identified by using a homozygous redleafed pollinator and roguing the red seedlings. Orchard evaluations of seedlings
from the M. sargentii M.9 selections has not been encouraging; yields have been
low and fruits small (Schmidt, 1988).
The inheritance of apomixis is not very straightforward. Sax (1959) considered
apomixis to be a dominant trait in the F1 hybrids from controlled crosses, whereas

BREEDING SYSTEMS

29

the evidence from Schmidt (1964, 1988) shows it to be recessive but varying
according to the ploidy of the species. Triploid hybrids of apomictic amphimictic
forms of M. sieboldii were in the ratio of 6 amphimicts: 1 apomict, but in crosses
between high chromosome number apomicts and low chromosome number
amphimicts, the proportion of apomicts was higher and increased proportionally to
the difference in chromosome number.

Parthenocarpy
The apple usually has 10 ovules but it is not necessary for them all to be fertilized
and develop into seeds for a fruit to be produced. Often a single seed is sufficient
for the development of the fruit; thus fruitfulness may still be maintained even when
a high degree of generational sterility is present. It is even possible in some cultivars
and in certain conditions for fruit to develop parthenocarpically without fertilization
and without weds. Fruits that arise in this way vary according to the cultivar: In
some they are small and often misshapen, and in others they are normal in size and
appearance. These fruits tend to ripen earlier and do not keep as well in storage as
seeded fruits.
In a classic, elegant experiment using removal of seeded fruit from apetalous
cultivars Spencer Seedless and Ohio 3, Chan and Cain (1967) demonstrated that
it is the presence of seeds inhibits subsequent flower formation and not the nutritional
competition of the developing seedless fruit. The critical period was 3 weeks after
pollination. The two apetalous cultivars bear annual crops of parthenocarpic seedless
fruits (because insects do not visit the apetalous flowers) but produce seeded fruits
following hand pollination.
The use of cultivars that produce parthenocarpic fruits consistently has been
recommended because of their ability to produce good yields in years when flowers
are damaged by late spring frosts or when conditions are unfavorable for pollination
(Thiele 1950). Some quite heavy-cropping cultivars have been described that produce
parthenocarpic fruit, but none seems to have been grown to any extent. Ewert, as
long ago as 1909, was advocating the planting of seedless apples as a safeguard
against failure of fertilization. Others have suggested breeding for this character,
but until recently little breeding was carried out. The combination of apetaly with
columnar apples has been achieved by crossing Wellington Bloomless with
McIntosh Wijcik and backcrossing the columnar normal flowered seedlings to the
apetalous Spencer Seedless (Tobutt 1994). This combination could be useful to
avoid bienniality and to achieve consistent production in high-density orchards
without pollination, allowing production to be independent of bee activity.

Inbreeding
Inbreeding in apples has generally been considered to be an unsuitable method of
breeding because of the expected loss of vigor. Vigor could be restored by
intercrossing inbreds but this is an unpromising breeding approach because

30

APPLES

TABLE 3. Rela
tion of Mor
tality and Vig
or With Inbr
eeding
Relation
Mortality
igor
Inbreeding
Inbreeding
coefficient
0.000
0.625
0.875

Mortality
(%)

Avg. height (m)

after 6 years

3.2
29.5
55.0

2.62
2.00
1.56

Source: Data from Brown (1975).

heterozygous individuals obtained by intercrossing two heterozygous parents can

be clonally propagated. However, because excess vigor requires expensive annual
pruning a slight reduction in vigor induced by inbreeding may be appropriate in
some areas.
Any program of inbreeding by selfing is difficult to pursue with apples because
of the high degree of self-incompatibility. A great many flowers have to be selfed to
produce a few fruits from which only a few weds are obtained. The average fruit set
from selfing a great many cultivars is about 2% and the number of seeds per fruit
averages one to two. Only about 30% of the seeds germinate and a great many of
these are weak and die, probably as a result of lethals becoming homozygous. The
end result is that only a very small proportion of the seeds produce viable seedlings
capable of flowering and fruiting. In these selfed progenies it is possible to get a
few good seedlings which, if not completely self-incompatible, can be selfed again.
By selecting the most vigorous and most self-fertile it has been possible to carry
this process through three selfed generations at the John Innes Institute, Norwich,
England. However, a comparison of seedlings derived from three progenies with
different degree of inbreeding of Coxs Orange Pippin indicated increased mortality
and decreased vigor as a result of inbreeding (Table 3). Karnatz (1988) reinvestigated
selfing as a breeding method and selected some promising selections from inbreeding
Golden Delicious.
Inbreeding is increased when parents have common ancestors. Thus, it is
important when making crosses between apparently widely different cultivars to
consider their pedigrees. Inbreeding may result from intercrossing new cultivars
because so many of them are based on Golden Delicious or Delicious.

Outbreeding and Backcrossing

Most breeding programs are based on outcrossing, usually between cultivars of
known merit, with the object of combining the good qualities of both parents in
some of the progeny. Much of the present breeding is aimed at specific objectives,
and complementary parents are selected that are expected to impart some specific
character to the offspring. In such cases a certain amount of backcrossing is used to
introduce one specific character into another background as in single gene resistance
to disease. It must be remembered that inbreeding occurs at the same rate in selfmg and backcrossing.

BREEDING SYSTEMS

31

Dominant single gene resistance (as Vf gene for scab resistance) in Malus species
has been transferred to the cultivated apple by a modified backcross procedure to
avoid inbreeding (Crosby et al. 1992). The method involves crossing the wild species
(heterozygous for VJ) with a large-fruited cultivar. The resistant Fig are heterozygous
and the best are selected and backcrossed to a good cultivar. Their progeny yields
50% seedlings carrying the dominant gene identified by scab resistance. The best
of these are again backcrossed to suitable cultivars and so on until all the suitable
qualities of the cultivated apple are recovered and the resistance from the species is
retained. Inbreeding is avoided by alternating different cultivars for the recurrent
quality parent, which eliminates loss of vigor and incompatibility problems.

Characters Controlled by Single Genes

In the early days of genetics, when the simple segregations from single gene
differences were being discovered, it was thought that plant breeding would be
revolutionized, and so it was for some plants. But when fruit trees were crossed the
results did not fit any of the simple segregations and no single gene characters were
known for a long time. All the important characters, such as fruit shape, size, and
color, were shown to be inherited quantitatively. In those days this was not easy to
understand, and Mendelian genetics was thought to have little to offer the fruit
breeder. Since then a number of single gene characters have been found, particularly
in relation to disease resistance, but also other characters such as albinism, apetaly,
lethals, and columnar growth (Table 4).
TABLE 4. Genes of the Apple
Preferred
symbola

Er
Sd1
Sd2
Sd3
Sdpr
Smh

Original
symbol Gene Effect

Source

Pest resistance
Eriosoma
Northern Spy
resistance
Dysaphis devecta
Coxs Orange
resistance
Pippin
Dysaphis devecta
Northern Spy
resistance
Dysaphis devecta
M. robusta
resistance
MAL 59/9
Dysaphis devecta
McIntosh
resistance
Dysaphis
M. robusta MAL
plantaginea
59/9
hypersensitivity

Reference

Knight et al. (1962)

Alston & Briggs
(1968)
Alston (1970b)
Alston (1970)b
Alston & Briggs
(1977)
Alston & Briggs
(1970)

32

APPLES

TABLE 4. (Contin
ued)
(Continued)
Preferred
symbola

Original
symbol Gene Effect

Source

Disease resistance
Glomerella
Golden Delicious
cingulata
susceptibility
Gymnosporangium Spartan
resistance
Phytophthora
Northern Spy
cactorum
resistance
Podosphaera
M. robusta
leucotricha
MAL 59/9
resistance
Podosphaera
M. zumi
leucotricha
MAL 68/5
resistance
Phyllosticta
Jonathan
solitaria
susceptibility
Idared

Gb
Gy-a
Gy-b
Pc
Pl1
Pl2
Ps1
Ps2
Vb

Venturia
inaequalis
resistance

M. Hansens
baccata 2

Vbj

Venturia
inaequalis
resistance
Venturia
inaequalis
resistance
Venturia
inaequalis
resistance
Venturia
inaequalis
resistance

M. baccata jackii
Dg27Tl

Vf
Vm
Vr

Bp1Bp2
CaaCab
Ma
Rf
Ru

AK

Reference
Thompson &
Taylor (1971)
Aldwinckle et al.
(1977)
Alston (1970b)

Knight & Alston

(1968)
Knight & Alston
(1968)
Mowry & Dayton
(1964)
Mowry & Dayton
(1964)
Dayton &
Williams (1968);
Williams &
Kuc (1969)
Dayton &
Williams (1968)

M. floribunda 821

Dayton &
Williams (1968)

M. micromalus
245-38

Dayton &
Williams (1968)

Russian sdlg.
R12740-7A

Dayton &
Williams (1968)

Fruit attributes
bitter pit
Coop 11
resistance
deciduous calyx
M. zumi
malic acid content Lord Lambourne
anthocyanin in
Worcester
fruit skin
Pearmain
russet
DArcy Spice

Korban &
Swiader (1984)
Henning (1947)
Nybom (1959)
Wilcox & Angelo
(1936)
Alston (1973)

BREEDING SYSTEMS

33

TABLE 4. (Contin
ued)
(Continued)
Preferred
symbola

Original
symbol Gene Effect

ape
atc
P1

al

P2

a2

P3
P4
P5
Pd

a3
d1
d2

S 1S 2
al

C1

bu1bu2
Co

tu

d1
d2
d3
d4
G

Flower or reproductive attributes

apetaly
Spencer Seedless
atrophied cololla
Melrose
pollen lethal
Reale d
Entraygues
pollen lethal
Transparente de
Croncels
pollen lethal
Germaine (loire)
pollen lethal
Carrey
pollen lethal
Cellini
double petals
Dorothea
incompatibility
Northern Spy
(multiple alleles)
Plant attributes
albinism
Rev. W. Wilks

burrknots
compact habit

Sb

dwarf
dwarf
dwarf
dwarf
dwarf re-growth
promoter
pale green lethal
chlorophyll
deficiency
necrotic bark
purple pigmentation in all
tissues
sieboldin present

weeping habit

ym1ym2ym3 a1a2a3

yellow mottle

C2

nb
Rt

Source

Northern Spy
McIntosh
(Wijcik)
Ottawa 521
Northern Spy
Ottawa 521
Starkrimson
Ottawa 521
Northern Spy
Calville Blanc
dHiver
Reinette du Mans
Baskatong

M. floribunda 821
M. baccata
gracilis
Worcester
Pearmain

Reference
Tobutt (1994)
Decourtye (1967)
Heilborn (1935)
Heilborn (1935)
Heilborn (1935)
Heilborn (1935)
Heilborn (1935)
Sampson &
Cameron (1965)
Knight et al. (1962)

Hall & Crane

(1933); Crane &
Lawrence (1933)
Decourtye (1967)
Lapins & Watkins
(1973)
Alston (1976)
Decourtye (1967)
Alston (1976)
Alston (1976)
Alston (1976)
Klein et al. (1961)
Williams (1958)
Lespinasse (unpub.)
Lewis & Crane
(1938); Sampson
& Cameron (1965)
Alston & Hunter
(1971)
Sampson &
Cameron (1965)
Sadamori et al.
(1964)

Indicates change from original symbol in accordance with nomenclature recommendations.

Source: Brown (1992).

34

APPLES

One of the first characters to be found in the apple that was controlled by a
single dominant gene was the complete anthocyanin pigmentation originating in
Malus pumila niedzwetzkyana Schneid. (Lewis and Crane 1938). Within the redleafed seedlings derived from crossing M. p. niedzwetzkyana with normal green
apples, there is considerable variation in the intensity of the pigmentation, suggesting
that other factors am involved. Nevertheless, in this case one can make a positive
classification into those that we pigmented and those that are not. Homozygous red
occurs rarely. Royalty is one of the few cultivars producing 100% red-pigmented
progeny. Most red-pigmented apples produce more non-red progeny than predicted.
A red green cross may yield as few as 15% red-pigmented progeny (J.N. Cummins,
unpublished).
Where a character is controlled by a dominant gene, it is very easy to transfer it
into other apples. The number of generations required depends entirely on the
quality of the apple in which the desired character is found. If it is already within
the cultivated apple, then possibly only one generation is required; if, on the other
hand, it is in a species of Malus, then several generations, possibly five or six, of
crossing and backcrossing may be required to eliminate the undesirable characters
of the species and retain the good qualities of the large-fruited cultivars and the
desired character of the species. However, if we assume even a minimum of 4 years
per generation, 5 backcrosses require a minimum of 20 years, a daunting prospect
for any but very young breeders. The possibility of genetic transformation could
shorten this interval considerably.
When the character is recessive (as apetalous), as are most mutations, then the
breeding becomes much more complicated and protracted. When a cultivar carrying a recessive character is crossed with a normal, the F1 is normal. The recessive
character reappears in a quarter of the F2 resulting from sib crossing or in half of the
backcross to the recessive parent.

Characters Controlled by Polygenes

Fruit size, shape, and productivity are under polygenic: control, which means that
when two cultivars are intercrossed there will be a wide and continuous range of
expression of all these characters in the seedlings. They will not segregate into
discrete categories. Most polygenic: characters behave independently. The range
of variation is related to the expression of the characters in the parents and the
progeny mean is always related to the parental mean. In some, given a strictly
additive system where there is no dominance, the progeny mean and the parental
mean will be the same; in others it may be above or below the parental mean. The
extent of the deviation and its direction will depend on dominance or epistasis and
on the difference between the dominant and recessive phenotype.
When the number of seedlings is plotted against the character values a simple
distribution curve is formed about the mean. In cases where no dominance is present
and the parental values are known, it is easy to predict the progeny mean and the
proportion of the progeny that will be above and below the parental values. Characters

BREEDING SYSTEMS

35

that behave in this way appear to be those that have not been subjected to great
selective pressure. Fruit shape is a character where there has been no selection at
either end of the scale for particularly flat apples or tall ones. There are cases where
there is a limit in the values of expression. Thus if July is considered the earliest
that an apple can ripen because this is the minimum time from flowering in which a
fruit can develop, then July is a fixed low value. The progeny from two earlyripening cultivars will show a one-sided distribution because of the July barrier and
will produce only seedlings that ripen in July and later but none earlier. In such
cases the progeny mean will be higher than the parental mean.
It frequently happens that the parental mean is considerably greater than the
progeny mean although the distribution forms the same simple curve. This occurs
with characters that have been subjected to extreme selective pressure over many
generations. Where selection is always at the extreme end of the curve, the progeny
mean will tend toward the middle of the range of variation for the character. An
example of this is fruit size, where strong selection has been away from the smallfruited crabs and progeny mean nay be as much as 40% less than the parental mean
(Brown 1960). However, the breeder should be aware that many characters are
distorted by metrical bias. Thus, crossing large and small apples often gives hybrids
that seem closer to the small parent suggesting dominance for small size. Plotting
fruit size on a log scale will often show that the hybrid is at the midpoint, suggesting
that the genes for size work on a logarithmic increase rather than on an arithmetic
increase.
In addition to having characters controlled by either a single major gene or by
polygenes, it is possible to have one system superimposed on the other. This can
present a rather confusing picture. Such a situation exists in the inheritance of
malic acid concentration in the fruit where medium to high acid is dominant over a
very low acid type, but within both types there is a range of variation typical of
polygenic control (Brown and Harvey 1971). In many quantitative characters where
many genes are involved, major control is often exerted by a relatively few genes,
each with a large effect. These few genes may be identified by linkage with molecular
markers. Thus, the mapping of the apple genome could advance selection even for
so-called quantitative traits.

Parental Contributions
Where detailed fruit and tree records of seedlings of many progenies have accumulated over many years, it is possible to use these data to calculate the contributions made by individual parents to their progenies. This assumes that the data for
the different characters have been recorded in such a way that it is possible to classify the seedlings into categories suitable for analysis. Gilbert (1967) describes a
method of analyzing a series of biparental crosses that have some parents in common by which the main effects of each parent on each of the characters can be
calculated by treating the data as an incomplete diallel cross. Since the combining
abilities are additive, the calculated main effects for each parent allow predictions

36

APPLES

to be made with considerable accuracy as to the progeny mean of untested combinations of parents.

Selection of Parents
Since most apple breeding is based on selection from progenies produced by crossing
extremely heterozgyous parents, consideration must be given to parent selection;
otherwise, breeding becomes haphazard and the outcome unpredictable. If the major
aim is to introduce a character that is controlled by a major gene or genes, then the
choice of at least one parent is limited to cultivars carrying these genes. However,
most of the important characters are polygenically controlled; for each there will be
a considerable range of variation and the breeder seeks the optimum expression of
each. Fortunately, the breeding behavior of highly heterozygous parental lines,
with respect to measurable characters having a low level of genetic dominance, is
often predictable. As has been shown, the mean value for the progenies falls with
reasonable consistency around the mean of the two parents. It should be the aim to
try to combine parents that between them have all the chosen characters present at
as near the optimum expression as possible. Thus if polygenic mildew resistance is
desired, then one and preferably both parents should have a high degree of resistance
in order that as many resistant seedlings as possible are available from which to
select for other characters. It is probably more realistic to choose parents with
complementary characters in order to select progeny that have the desirable attributes
of each parent.
Having chosen the parents, the breeder must determine the size of the progeny
required to be reasonably certain of obtaining a new apple having all the attributes
visualized when planning the cross. Using actual records, Williams (1959) calculated that the percentage of desirable seedlings that can be expected as the main
product of an apple breeding program for polygenically controlled characters is
seldom more than 40% and for every additional character the figure rapidly decreases.
Thus for a program in which the main objective is polygenically controlled mildew
resistance, size of fruit, season of maturity flavor, and color of skin, a reasonable
estimate would be 40, 20, 20, 10, and 10%, respectively. Assuming that these
characters behave independently, a progeny of 6250 (16/100,000) would, on the
average, yield one seedling possessing a combination of the five traits at an acceptable
level of expression; the progeny size would have to be even larger to expect the rare
recombinant to occur with a high probability of success. However, experience with
rootstock screening indicates that these estimates may be overly optimistic by an
order of magnitude (Cummins and Aldwinckle 1983). In addition to these definable
characters there are others for which allowance must be made, as well as some that
cannot be readily included in the calculation, such as minor characters of flavor or
texture which so often decide the ultimate success or failure of a selection. It is not
unreasonable to assume that an actual population as high as 30,000 will be required,
and to observe 30,000 seedlings would be practically impossible if the seedlings are

BREEDING SYSTEMS

37

to be evaluated at the fruiting stage. Nevertheless, in any serious breeding program,

provided that great care is taken in the selection of the parents, progenies as large as
practicable should be aimed at. Instead of growing all the seedlings to fruiting, as
many as possible should be eliminated in the early stages of growth. For example,
a high proportion might be eliminated because of an unacceptable level of mildew
susceptibility by the second season of growth. Results, however, can be improved
by increasing the number of cycles of selection.

Breeding Strategy
Bringhurst (1983) contributed a penetrating analysis of fruit breeding based on his
experiences as a breeder of strawberry and avocado. His conclusions seem particularly appropriate to apple breeding where efforts have only recently been more
effective than evaluation of naturally occurring variation.
The traditional strategy for fruit breeding has been to identify superior phenotypes, propagate the best selections, develop cultural practices that enhance the
performance of the selected cultivars, hybridize among the best selections, and then
continue the cycle. This breeding method may be considered a form of recurrent
mass selection in which the key concept is selection of the best individuals and
continual recombination over many cycles. The most productive breeders carry on
a number of objectives simultaneously and keep the program moving on with the
best selections becoming parents for the next generation. The assumption is made
that as the population improves, the chance of releasing superior genotypes will
increase. Results from a number of crops including apple indicate that this strategy
will succeed. Why then did most apple breeding programs fail to make an impact
on the industry? Bringhursts analysis includes the following reasons:
1. Insufficient support. Because of the tremendous selection pressure that
has resulted in our present cultivars it is extremely unlikely that casual
programs can be expected to produce superior. seedlings. This is borne
out by the evidence. Many programs throughout the United States have
not been successful and have been dropped; only long-term program
have been successful. This require continuous support and a long-term
strategy. With deficiency in funding, short-term breeding objectives my
be overemphasized to the detriment of long-term objectives. Progress in
fruit breeding tends to be incremental and there is no quick Jim.
2 Faulty strategy. Genetic progress in recurrent selection requires continuous generations of progress. Attempting to spend most resources on
making many crosses to find the best combinations and then repeating
appropriate crosses with very large progenies is usually an inefficient use
of time and resources. Errors are very costly. Intense examination of
unsuitable crosses is folly. No end of testing will compensate for a
dearth of good material from which to select.

38

APPLES

3. Misdirected emphasis. Undue emphasis on secondary traits is misdirected; the breeder needs to emphasize the most essential traits. For
example, breeding that concentrates on diseases and insects that can be
controlled or we of minor importance is inefficient. The power of the
cooperative apple breeding program of Purdue, Rutgers, and Illinois was
in its unceasing, single-minded emphasis on moving the Vf gene into
adapted types.
Other errors include faulty procedures, failure to measure things properly, failure to start at the most advanced level, failure to move through the generations,
failure to eliminate the losers (breeders must be ruthless), and failure in breadth of
program. Because a breeder of a tree crop such as apple has time for but a few
generations, several programs must proceed simultaneously. Finally, Bringhurst
points out that every cultivar has its own peculiar characteristics and may require
special cultural practices. Consequently, it is in the best interest to the breeder and
grower to consider those cultural practices that enhance each cultivars performance.

BREEDING FOR SPECIFIC CHARACTERS

Vigor
The study of the inheritance of vigor is difficult because several factors are involved.
For example, parent cultivars we usually grown on rootstocks that control vigor
while seedlings are typically self-rooted. In an early study, Spinks, (1936) concluded
that there was no correlation between the size and vigor of the seedlings and that of
the parents. Watkins and Spangelo (1970) in a diallel analysis of seedling vigor
(expressed as plant height) found that the progeny mean was equal to the midparent value in one group while in another set there was considerable divergence of
the progeny mean from the mid-parent value. This suggests that several factors
influence or contribute to seedling vigor. The chief difficulty arises in trying to
assess the vigor of parent cultivars. As a practical matter, differences in seedling
vigor can be adjusted by the selection of rootstocks.
The effect of inbreeding and aneuploidy in reducing vigor has already been
discussed. Interestingly, triploidy increases vigor while tetraploidy decreases vigor,
suggesting that the 3x condition is the most physiologically appropriate for apple.
Tree size control, is one of the most important uses of rootstocks. In England
the early programs were to increase vigor but the general trend now in apple production is to produce compact precocious trees for ease of harvest. To compensate for
the reduced tree size, planting density has been greatly increased in the modem
orchard.
There is no consistent semantic distinction between a clone that is dwarf when
grown on its own roots or on a vigorous stock and a clone that has dwarfing propensity
when used as a rootstock. The term compact has been used to describe a tree of
small stature regardless of whether its compactness might be caused by genetics, by

BREEDING FOR SPECIFIC CHARACTERS

39

a dwarfing stock, or by summer priming (Cummins and Aldwinckle 1983). In an

ungrafted plant, dwarfness is typified in the orchard by a spur-type fruiting cultivar
such as Oregon Spur Delicious, Delcon, and Nugget Golden Delicious.
However, these genotypes do not induce dwarfing when used as rootstock or
interstem. Typical dwarfing rootstocks such as M.8 M.9, Bud 9, P.22, and
Mark are all small low-growing bushes that differ from the compact fruiting
cultivars.
The physiological mechanisms of dwarfness and dwarfing are still not clearly
understood. A number of correlations however, have been claimed for predicting
dwarfing capacity of potential rootstocks. Hutchinson (1967) presented data from
studies of M.9 progenies that suggest that within a family the most dwarfing and
the most invigorating individuals can be identified by simple field observations of
vigor of ungrafted plants. Beakbane et al. (1939,1941,1947,1975) reported correlations between dwarfing propensity and various structural features. These include
the percentage of live tissue of the root cross section; the bark/wood ratio, the
percentage of medullary ray tissue in root transection; and high stomatal density of
leaves. Some rootstock breeding programs have employed the bark/wood ratio to
reduce populations or to select a reduced number of individuals most likely to be
dwarfing. Lockard and Schneider (1981) have proposed that the bark is the key to
the dwarfing mechanism and the dwarfing bark reduces translocation of auxin, sugars,
and other compounds. In the Cornell, Geneva program, orchard testing of grafted
trees has been used for estimation of vigor induction. The growth pattern of the
orchard tree appears to be well established by the fifth year of testing.
The relation between rooting of branches and dwarfing has been known since
the Middle Ages. While not all apples that root readily by adventitious roots from
mature shoots are necessarily dwarfing when used as rootstock, all of the most
common dwarfing rootstocks possess this character (Tukey 1964).
Dwarfing capacity appears to be under control of a number of genes (J.N.
Cummins, unpublished), but few studies on the subject have been published. In a
progeny of M.9 M.8 both very dwarfing, about one-quarter of the seedlings
were as dwarfing as the parents, but progeny midpoint was much more vigorous
than the mid-parental mean. Less than 5% of seedlings from crosses of M.9 with
vigorous parents have been as dwarfing as M.9 and as few as <1% in some crosses.
The picture is further complicated by the fact that fruit bearing itself is dwarfing
and few studies have examined this direct effect of precocity.

Cold Hardiness
In severe winter climates the most important character of the tree is cold hardiness,
for many cultivars are seriously injured and, in extreme cases, even killed. There is
a continual need for apples of quality that are more winter hardy so they may escape
damage in conditions of severe cold in existing apple growing areas and be extended
into colder areas. There are considerable differences in cold hardiness among various
cultivars and species and it is quite possible to breed for increased hardiness. The

40

APPLES

breeding and selection for cold hardiness in deciduous fruit crops was reviewed by
Quamme and Stushnoff (1983) and Saveljev (1988).
To survive freezing temperatures, plants have mechanisms either to avoid or to
tolerate ice within their tissues by freezing point depression and supercooling. Woody
species such as apple have the ability to maintain supercooled cellular water in
critical tissues to 40C. Deep supercooling involves freezing avoidance to the
point of spontaneous nucleation of the cellular solutions. Freezing in deep
supercooled tissue is lethal when it occurs and the degree to which deep supercooling occurs determines the poleward limits of production. Detection of the
supercooling point provides the basis for measuring the level of cold hardiness.
Tissues of apple, however, can survive to liquid nitrogen temperature (-196C) when
fully acclimated. This technique has been utilized to store shoots cryogenically for
germplasm preservation (Sakai 1984).
Several techniques can be used to evaluate winter hardiness. The traditional
technique was to rely on test winters, but this is unreliable. Much winter injury can
be traced to severe low temperatures in the late fall, that may injure trunk and scaffold branches before cold hardiness develops. Acclimation is triggered by environmental cues, including shortening day length, changes in light quality, and
temperature drops. Much freezing damage is due to unseasonably warm weather in
the winter that results when deacclimated tissues suffer subsequent freezing. It is
important to use objective and quantitative methods of testing cold hardiness because
field observation under natural conditions is not always reliable. These include
artificial freezing tests using portable chambers where damage is evaluated by tissue
browning or one of several objective methods of measuring cell viability.
Conductivity tests are based on the fact that shoots or other parts of the plant, after
having been frozen and soaked for a certain time in pure water, exude electrolytes
by leakage from the killed cells and to a less extent from living cells. The amount of
electrolyte exuded may be estimated by measuring the electrolytic conductivity of
the water. Exotherm analysis detects the heat of fusion by a deflection on the timetemperature profile (exotherm) during constant cooling. This approach has been
used to determine cultivar differences in wood hardiness of apple.
Inheritance of cold hardiness is under polygenic control with the frequency
distribution in progenies forming a normal curve about the mean. In most progenies there air some seedlings more hardy than either parent. Watkins and Spangelo
(1970), using a diallel analysis of seedling apples derived from parents that were
unusually hardy and evaluated with artificial freezing, demonstrated that additivity
was a major component of genetic variance for winter survival.
Tolerance of low winter temperatures has been the central objective of numerous breeders across Canada, the northern United States, Sweden, Poland, northern
China, and Russia (Saveljev 1988). These hardiness-centered programs have
generally been intended to extend the range of apple production for local
consumption, rather than to develop cultivars to compete on the broad commercial
market. Half a century ago, the hardy cultivars available for marginal areas included
Duchess of Oldenburg, Yellow Transparent, Wealthy, Charlamoff, and several

BREEDING FOR SPECIFIC CHARACTERS

41

Antonovka types, all of relatively low quality but capable of surviving, even thriving,
through the challenges of a severe winter. Breeders of winterhardy cultivars have
been especially successful; introductions include Fantazia, Fireside, Handson,
Honcycrisp, Lobo, Kaljo, and Mantet, Northern Lights. Winter-hardy
rootstocks derived from. northern breeding programs include Alnarp 2, Bemali,
the Budagovski series from Russia, Mark, and the P-series from Poland.

Winter Chilling Requirements

Apple trees require a certain amount of chilling during the dormant season for their
proper development. In subtropical climates this requirement is not always met and
prolonged dormancy or delayed foliation results. Cultivars with low chilling
requirements occur locally but most are of poor quality. Oppenheimer and Slor
(1968) in Israel crossed some of these with good quality cultivars and introduced
Anna, Ein Shemer, and Schlor, cultivars with low chilling requirements and
much improved quality. Early leafing was the criterion for selection; seedlings that
broke bud within three weeks of the earliest were grown. This character appears to
be polygenically controlled; about 30% of the F1 progeny and 40% of the backcrosses to Delicious and Jonathan were sufficiently early to be retained for further
testing. Recent discoveries in Mexico and South Africa of Golden Delicious sports
with low-chilling requirements suggest other, perhaps simpler mechanisms
(Gonzalez-Cepeda 1992). It might be possible to eliminate the chilling requirement
since some of the earliest to leaf were earlier after warm winters than after cooler
ones. Breeders have utilized germplasm introduced by Oppenheimer 40 years ago
to produce cultivars with minimal chilling requirements and with higher quality and
better production potential. Low chilling requirement of Anna was postulated to
be controlled by one major dominant gene and several minor genes (Hauagge and
Cummins 1991). Mollies Delicious has been used successfully as a source of low
chilling in the Brazilian breeding program (F. Denardi, personal communication)
Recent low-chill cultivars that are beginning to make an impact on the market include
Adina, EarliDel, Goldina, Princessa and SummerDel. Primicia is a new
scab-resistant, low-chill cultivar released in Brazil.

Season of Flowering
Since cross-pollination is essential for maximum fruit production in the apple,
synchronization of flowering is important. Cultivars that flower very early or very
late require special attention to provide efficient pollinators, but this problem has
been alleviated by using a range of crabapple pollinizers. It is convenient if new
cultivars flower midseason or late midseason for pollination to coincide with that of
most of the widely grown cultivars. Late flowering is considered important to avoid
disastrous spring frost damage. One objective in a number of breeding programs is
to produce cultivars that flower so late that practically all danger to the blossoms
from late frost is past. However, progress has been hampered because combining

42

APPLES

late flowering with early ripening because of die minimum time required from
flowering to the development of full fruit size and maturity.
The inheritance of time of flowering is polygenically controlled. A fair estimate
of mean flowering date of the progeny can be obtained from the mid-flowering time
of the parents. The seedling distribution about the mean appears to be normal and
the spread is such that seedlings flowering later than either parent will normally be
found.
Good correlation between the time of leaf emergence and season of flowering
was found by Tydeman (1964) and Murawski (1967). Thus, early or even midseason leafing seedlings could be discarded in their second year of growth.
Another contributory factor to crop losses due to late frost is the inherent
susceptibility of the flowers to injury. In general, flowers are most tender in the
later stages of development, for the styles are most susceptible to damage. Fortunately, not a the flower buds on a tree are at the same stage of development at one
time, so that some may escape damage while others are killed. Although there are
differences in sensitivity of blossom to frost damage, selection for this character
does not appear to have been carried out. Tetraploids and triploids such as Jonagold
and Mutsu, with larger cells, are more prone to damage than diploids. The hardiness
of the tree to winter injury and the hardiness of the flowers to spring frost damage
appear to be inherited independently.

Duration of the Juvenile Period

The length of the juvenile period of the seedlings is of prime importance because it
determines the number of years the trees have to be grown and occupy space before
the results can be assessed. Furthermore, the length of the juvenile period is correlated
with fruiting precocity and perhaps with productivity after propagation. Some parents
contribute to very short juvenile periods with many of their seedlings flowering in
three or four years from seeding, while Northern Spy progenies may take ten or
more years to fruit.
The inheritance of juvenility among cultivars is difficult to determine because
the length of the juvenile period of old cultivars is usually unknown. Studies of
various progenies indicate that the character is quantitatively inherited. Visser (1970)
demonstrated a good correlation between the vegetative phase of a cultivar (i.e., the
time from propagation to fruiting) and the contribution it makes to the length of the
juvenile period of its progeny. Evidence that seedlings must attain a certain size
before they reach the stage at which they can flower and fruit would suggest that
vigorous seedlings are the ones most likely to attain the optimum size in the shortest
time. In general, polyploids have much longer juvenile periods and their seedlings,
although they my attain much greater size than diploids, nay take several more
years to fruit. Recurrent selection for early flowering can be obtained naturally by
using early-flowering seedlings in segregating progeny as parents and by continually
discarding all seedlings that do not flower by the fifth or sixth year from planting.

BREEDING FOR SPECIFIC CHARACTERS

43

The ability to induce precocity is an important character in rootstock selection.

In Cornell, Geneva, this is carried out by using Delicious and Northern Spy,
notoriously nonprecocious, as test scions on rootstock candidates.

Spur Types
The spur types are characterized by compact habit, reduced internode length, limited
side branching on shoots, and prolific development of fruit spurs, producing trees
that are precocious in bearing and crop more heavily and regularly in the early
years. The spur types are sometimes referred to as dwarfs or compact mutants, but
it is possible to have compact trees that do not develop spurs freely. The spur bearing
habit occurs as bud sports. Compact strains of Delicious and other cultivars have
dramatically changed the apple industry by increasing overall orchard efficiency
and permitting use of more vigorous rootstocks. The inheritance of plant form
appears to be under polygenic control but it has been difficult to breed for the spur
habit in an efficient way because of continuous variation from trees with no or few
spurs to trees with many spurs (Lapins 1974; Blazek 1983). Although genetic dwarfs
appear superficially similar to compact types, not all dwarfs are spur types. The
dwarf character analyzed by Decourtye (1967) was controlled by a single recessive
gene and various cultivars, including Golden Delicious, are heterozygous. The
dwarfs analyzed by Alston (1976) were due to recessives in various combinations.
New spur-types can be produced either by inducing mutations in existing cultivars or by breeding. Mutations can be induced by irradiating dormant scions. The
optimum X-ray dose for apples appears to be 3 krad. The treated scions are then
grafted onto rootstocks, and the selection for compact types can begin during the
first seasons growth. The selection criteria are that the length/diameter ratio of the
shoots be less than normal and that the shoots have shorter internodes. Selection
for short internodes alone is unreliable. The trees are cut back into the original
graft, the lateral shoots are cut back to induce latent buds to develop, and the selection
method is repeated. This process may be repeated in the third year. These compact
selections can be further selected for free spurring and precocious fruiting. These
methods are described in detail by Visser et al. (1971).
Not all spur types that have been used in breeding behave in the same way.
Lapins (1969) reported that a radiation-induced spur mutant of McIntosh crossed
with Golden Delicious did not produce any compact seedlings, suggesting that
some of the mutants are chimeras and not necessarily found in the L-2 layer. In
contrast, 44% of progeny of the spontaneous compact mutant McIntosh Wijcik
when crossed with Golden Delicious had compact growth. Ibis extremely compact mutant is characterized by very short internodes and reduced lateral shoot grown
and increased spurriness (Kelsey and Brown 1992). Segregation data indicate that
genotype is heterozygous for a dominant mutant named Co for columnar, but modifier
genes influence the segregation ratios and forms obtained. In England, Co has been
used to produce columnar trees for the home garden, including an ornamental

44

APPLES

(Maypole) and a number of dessert types (Tobutt 1985). The Co gene has been
combined with apetalous in an attempt to produce parthenocarpic columnar-type
trees (Tobutt 1994). It has been suggested that the development of apetalous,
parthenocarpic, columnar trees with collar rot resistance would permit the use of
inexpensive self-rooted trees produced by tissue culture for very high density
orchards. This could dramatically change the way apples are grown if problems of
excessive vigor can be solved.
In some instances, mutation for spurriness has been associated with delayed
ripening of fruit and/or increase in tendency to russet. Several spur-type sports of
Golden Delicious, for example, ripen 4 to 7 days later than normal Golden Delicious, and are much more likely to be russetted. Interestingly, these spur-type
Golden Delicious frequently revert to the normal growth habit but retain the delayed
ripening and russetting attributes.

Fruit Size
Fruit size is one of the most critical characters in the selection of apple seedlings; if
fruits cannot attain the required size, the seedling should be eliminated from further
consideration. Fruit size is a somewhat variable character that can be influenced by
environmental conditions, and especially by fruit load. Thus, some defects in fruit
size can be adjusted by fruit thinning. If seedlings and parents are grown under
comparable conditions and fruit load there should be little difficulty in establishing
the standard.
There are a number of ways of expressing fruit size; the most common way is
diameter. The generally accepted minimum diameter is about 65 mm. In analyzing
the inheritance of size, the unit used by Brown (1960) was estimated volume (mL =
[(height(mm) radius (mm)2)/1000] 2.7. While not as accurate as measurement
by displacement, this method permits the use of records of fruit outlines. Classifying
seedlings into size groups indicated fruit size to be polygenically controlled with
normal distribution about the mean.
The contribution made by parents to size is not quite as straightforward as some
other characters. In some progenies a small percentage of the seedlings have fruits
larger than the larger parent and quite frequently 50% or more of the seedlings have
fruit smaller than the smaller parent. The parental contribution to size can be
calculated if data are available from a large number of progenies, by treating them
as an incomplete diallel using the method described by Gilbert (1967). This can be
much more accurate than trying to estimate the progeny mean from the parental
mean.
Fruit size in the apple has been greatly increased by rigorous selection over the
years from the small wild types to the size of the present-day commercial apple.
Where extreme selection pressure of this kind occurs there is a tendency for the
progeny mean to be smaller than the parental mean. This is home out by the analyses of many crosses where the mean fruit size of the progenies is smaller than the
average size of the parents by about 34%. This means that many cultivars that are
acceptable themselves produce progenies in which the fruit size of the majority of

BREEDING FOR SPECIFIC CHARACTERS

45

the seedlings is below commercial size. Some combinations of cultivars are unlikely
to yield any or very few seedlings with fruit of acceptable size. In general, the
larger the fruits of the parents, the greater will be the proportion of the seedlings
producing fruit of acceptable size. It may well be sound policy to select very largefruited seedlings, not necessarily acceptable as cultivars, for use as parents to ensure
an increase in the proportion of seedlings with good fruit size from which selection
for other desirable characters can be made.

Fruit Shape
Fruits of apple vary in shape from very flat, oblate to oblong. If the shape is translated into height/diameter and expressed as a percentage, then the very flat apples
are about 65% and the very tall ones about 100%. It is rare indeed for the height to
exceed the diameter. This procedure does not take into account those conic fruits
that taper or other irregularities of contour such as ribbing. Wilcox and Angelo
(1937) demonstrated that the tapering conical character is rarely observed in oblate
apples (1%) and is most common in the round oblong (32%). By dividing the range
of ratio percentages from 65 to 100 into 5% intervals eight shape classes are formed,
and by classifying the fruits of seedlings into these categories, it is possible to plot
the frequency distribution among the eight classes for the different progenies. The
majority of the seedlings fall between the parental values, although the distribution
frequency extends beyond the parents and forms a normal distribution curve about
the mean (Brown 1960).
By studying a fairly large number of progenies it is possible to calculate the
parental contribution to shape of the different cultivars used. In 35 progenies there
was a remarkable similarity between the actual progeny mean, the estimated progeny
mean (calculated from the sum of the calculated main effects of the parents), and
the mid-parent value (Brown 1975). It is possible to use the height/ diameter ratio
of the fruit of the parents to predict with considerable accuracy the progeny mean
and the range of shape classes expected. Because of the market preference against
oblate apples, the breeder will do well to eliminate all oblate apples from
consideration as parents.
The amount of ribbing of seedling fruits in a number of small progenies suggests that the progenies from angular-fruited parents will have a preponderance of
angular-fruited seedlings (Spinks 1936). Practically all apple progenies will have
seedlings with sonic angular, slightly angular, and nonangular fruit, but the proportion in each category differs based on of the parents.

Season of Ripening
Season of ripening is an important character, although transhemispheric shipment
and improvement in storage have affected the market demand for very early and
very late apples in some areas. However, improved early-ripening and late-ripening cultivars are sought in most countries. A number of studies (Brown 1960; Tancred
et al. 1995) indicate that time of ripening is a highly heritable, polygenic character

46

APPLES

with predominately additive genetic components of variance responsible for genetic

variation. Despite the existence of some specific combining ability and some
nonnormal family distributions, the best strategy to predict harvest date is to calculate
nz-an harvest date of parents. The progenies from crosses between early-ripening
parents have a small range of distribution, which is no doubt due to the limit of
earliness being fixed by the minimum time from flowering to fully developed fruit.
Thus when very early ripening parents are crossed, the progeny mean is quite close
to the parental mean. However, because most very early apples have poor quality,
crossing early early may not be the best strategy; it may be preferable to cross
early high-quality midseason apples and select transgressive, high-quality early
segregates.
Breeding for the very hot, dry, and long summer climate of Western Australia
that is followed by gradual cooling of night temperatures in late autumn has been
successfully carried out by John Cripps. In this challenging climate, even the lateripening Granny Smith (170 days from flowering) matured during the hot period
and quality was unacceptable. Selections of Golden Delicious Lady Williams
delayed the harvest date 2 to 8 weeks after Granny Smith; Pink Lady, Sundowner, and Big Time mature 2, 4, and 6 weeks after Granny Smith, respectively, after the hot season when cooler nights bring good sugar/acid balance and
attractive finish. These cultivars hold special promise for regions with similar
climates, such as South Africa, the southeastern United States and California, and
perhaps Egypt and elsewhere in the Middle East. Pink Lady is now being widely
planted in southern France and Washington State.

Fruit Color
Fruit color is determined primarily by the ground color of the skin and secondly by
the superimposed anthocyanin pigmentation. The ground color of the immature
fruit is dark green. As the fruit matures, (1) the green may fade until it has completely disappeared and the ground color of the fruit will then be in the range from
very pale cream to deep yellow; (2) the green may fade but not completely, producing
ground colors in the greenish-yellow to yellowish-green range; or (3) the green may
not fade at all, leaving a green ground color. The ground color is polygenically
controlled and the yellow range (which is related, at least to some extent, to flesh
color) and the green range may be controlled independently (Brown 1975).
Anthocyanin production in the fruit skin may be present or absent. If absent,
then the fruit is either all yellow or all green. If pigmentation is present, it can take
several forms from small red flecks to bold stripes and from a faint blush to solid
red. The intensity of the color can vary from very pale to very deep red and the area
can be from practically nil to complete coverage. A study of all aspects of fruit
color is complex and often confusing because the expression of all these characters
can be affected by the stage of maturity of the fruit, by the general environment, by
nutritional and cultural factors, and by the microenvironment within the area of the
tree.

BREEDING FOR SPECIFIC CHARACTERS

47

The production of anthocyanin is normally dominant over the lack of it and the
majority of red apples appear to be heterozygous since most when intercrossed
produce a few nonred seedlings (Crane 1953). The coloring most often takes the
form of striping, but blushes may also occur. These two characters are distinct and
these can be stripes, stripes on a blush, or blush alone. Wilcox and Angelo (1936)
found that by classifying seedlings into striped and nonstriped (which included the
blushed and unpigmented fruit), striping behaved more or less as a single maJor
gene producing either 100% striped, 1:1, or 3:1 segregations of striped and nonstriped. The blush may be pronounced but in some cases is ephemeral, being
extremely light-sensitive. Quite a number of normally yellow apples will, if exposed
to full sun, develop a slight blush and it is possible for a few presumed yellow
cultivars to produce a few seedlings with colored fruit. Normally when blushed
apples are intercrossed, the offspring will be either blushed or nonpigmented but
not striped (Spinks 1936).
The shade of red that develops depends to a very great extent on the ground
color. The most brilliant red is produced when the ground color is almost white and
the dullest brown when the ground color is green. Because dull, dark colors are
considered unattractive, ground color may be the most important factor in producing attractive colored fruit. The area of the fruit covered and the color intensity are
inherited quantitatively. In most families there is a continuous gradation in color
and the average amount in the progeny is proportional to the color of the two parents.
The anthocyanin pigment is in solution in cells of the epidermis and subepidermal layers, but not all cells are pigmented. Some apples have no pigment in the
epidermis and the intensity of color depends on the proportion of cells in each
subepidermal layer that contain pigment (Dayton 1959). Since overall red is a very
desirable character, red sports from many cultivars have been selected that have
either greater coverage or more intense red. Most of these sports have arisen naturally but some have been induced by irradiation. It is natural when breeding red
apples to select the red sports as parents in preference to the original cultivar, believing
the red sports to be better parents. Bergendal (1970) crossed Golden Delicious
with six cultivars and their red sports and grew a progeny from each. In all cases the
proportion of seedlings in each of the ground color classes was the same from original
and sport. In the amount of red, the results from four of the pairs showed no
difference, whereas in two the results were quite different.
Misic and Tesovic (1970) examined the pigmentation of the three outer-cell layers
of the fruit and showed that the number of pigmented cells in the epidermis of
Delicious and Richared are the same, but that in the subepidermis (L-2) there are
twice as many pigmented cells in Richared as in Delicious, confirming the results
of Dayton (1959). The mutation must have taken place in L-2. This indicates that
a histological examination of the three outer-cell layers of the fruit skin of cultivars
and their red sports would indicate whether a red sport will produce more redfruited seedlings or will breed the same as the original.
In the past, color sports were extremely important because of a quirk in the
patent law. The discoverer of the mutant could patent the new sport and owe no

48

APPLES

royalties to the originator of the original cultivar. Thus color sports became a legal
way to overcome patent protection. Many licensors of apples now write in the
contract that all sports belong to the originator. It is clear that some joint, equitable
arrangements need to be made between the originator of the cultivar and the
discoverer of improved mutations.
Studies are underway to separate the chimeral nature of Gala sports at the
University of Illinois (R. Skirvin, personal communication). Preliminary evidence
is that trees fruiting from a red sport of Gala regenerated from leaves produced
fruit that were extremely uniform and tended to be poorly colored, suggesting that
this technique my not be promising for chimeral engineering. Regeneration from
epidermal strips (Compton and Veilleux. 1992) may be a more promising technique.

Russet
Russet on the fruit varies from complete to free with all gradations in between.
Complete russet is popular in some countries because it is associated with increased
aromatic flavor. Many genotypes have russet confined to the stalk cavity or to the
calyx, and others have it in patches over the fruit. Russetting tendencies are inherited
but external factors such as low temperature, high humidity during early development,
and spray damage can also influence russet production. Golden Delicious is
particularly susceptible to russet and russet-reduced sports have been identified.
Patchy russet is considered a defect and disliked on apples like Golden Delicious,
but tolerated on Coxs Orange Pippin.
Information on the inheritance of russet is available from small progenies where
russetting was classified as none, slight (russet in the stalk cavity or calyx only),
moderate (in patches over the fruit), and complete (Brown 1975). In two progenies
where two completely russeted cultivars were crossed, approximately 50% of the
seedlings were completely russet and the others slight to moderate. Similarly, when
Coxs Orange Pippin, a moderate russet, was crossed with the completely russet
Egremont Russet, 50% of the seedlings had completely russet fruit. McIntosh,
when crossed with Egremont Russet produced no completely russet seedlings in a
rather small progeny. Coxs Orange Pippin crossed with the completely russet
Golden Russet produced only 25% completely russet seedlings. In a number of
progenies where moderate russets were crossed with slightly russet parents, a few
completely russet seedlings where produced. It is obvious from these results that
more than one factor is involved since all full russets do not behave in the same way.

Flesh Color
The flesh of the apple varies in color from white through cream to pale yellow; it
may be greenish white or it may even be tinged with red. These categories are not
discrete; one merges with another. The preferred color is clear yellow because this
is accepted by all consumers of the fresh fruit and preferred by processors of sauce
and slices. White flesh is accepted and even preferred in regions where McIntosh

BREEDING FOR SPECIFIC CHARACTERS

49

TABLE 5. Appr
oxima
te P
er
centa
ge Contr
ib
ution P
ar
ental Types
ppro
ximate
Per
ercenta
centag
Contrib
ibution
Par
arental
to Flesh Color in Seedlings.
Seedling phenotype
Parental
phenotype
White
Cream
Yellow
Green

Yellow

Cream

White

0
8
35
0

24
29
12
31

22
8
0
0

Green
4
5
3
19

Source: Data from Brown

is considered the standard of apple quality. Red flesh and even very slight bleeding
from the skin, are unacceptable.
Brown (1975) has developed a protocol to predict the expected behavior of parents
with respect to color. Adding the two parental contributions shown in Table 5 gives
an estimate of the distribution of seedlings expected in each color class (although
the value for green flesh is based on too few seedlings to be accurate).

Flesh Oxidation
Even more important than flesh color is flesh browning when exposed to air due to
the presence of polyphenyl oxidases. Apples that brown are quite unattractive in
salads or when served peeled. In the United States almost all producers of apple
sauce and apple slices depend on sulfite baths of the peeled apple to prevent
browning. It now seems probable that this treatment may become illegal, in which
case the process may have to confine processing to nonbrowning cultivars such as
GoldRush or NY 674.
Inhibition of flesh browning in apples, while important for the fresh market, can
be critical for processing. Mechanisms of resistance vary. Cultivars with resistance
to browning may have low levels of polyphenol oxidase, low levels of tannins, high
levels of ascorbic acid, or some combination of the above. There are no published
reports on the inheritance of flesh browning but this trait is under investigation at
Cornell University, Geneva. An advanced selection, NY 674, with low levels of
polyphenol oxidase and excellent resistance to browning has been crossed with a
range of cultivars varying in susceptibility; an analysis of their progenies should
separate the components of flesh browning and their inheritance.

Flesh Texture
Flesh texture is increasing in importance in fruit acceptability but there has been
little published work on the evaluation or on the inheritance of this character. This
is a complex character and there appears to be a relationship between flesh texture
and ethylene production. Cultivars differ widely in flesh texture and the persistence

50

APPLES

of this character in storage. McIntosh is considered unacceptably soft fleshed,

while two new cultivars, Honeycrisp and GoldRush have remarkable crispness.
The crispness of GoldRush, persists for a year in refrigerated air storage.

Fruit Flavor
Fruit flavor, one of the most important criteria in selection of apple seedlings, is
difficult to analyze since the constituents of flavor are a complex combination of
acids, sugars, tannin , and aromatic substances. A review of apple flavors by Yahia
(1994) indicated the presence of literally hundreds of volatiles but the precise role
played by individual components or by various combinations in flavor is not
understood. Although the appreciation of flavor is personal and variable (De gustibus
non disputandum est), the marketplace has made a decision on acceptability.
The basis of apple taste and flavor is acidity and sweetness; it is the balance
between these, irrespective of aroma, that primarily determines the acceptability of
the fruit. Apples that are high in acid and low in sugar are quite unpalatable, being
too acid; similarly, apples high in sugar and low in acid are too sweet and insipid.
The acid in the mature fruit is almost entirely malic acid, and is measured either as
percentage of malic acid in the fruit juice or as the pH of the juice. The main sugars
are fructose, sucrose, and glucose conveniently measured by refractive index as
percentage of total sugars in the fruit juice (Brown and Harvey 1971).
Acidity and sweetness are inherited independently. From an analysis of over
100 cultivars, the majority of successful dessert apples are in the groups of medium
acid/medium sugar, medium acid/high sugar, and low acid/medium sugar. The
inheritance of sweetness, measured as the sugar concentration in the fruit juice,
shows a quantitative pattern with a normal distribution of progenies about the mean
that is very close to the mean value of the two parents. The inheritance of acidity is,
however, more complicated in that two patterns are involved. A single gene control,
with medium, to high acidity being dominant over very low acidity, is superimposed
on a quantitative pattern. The very low acid seedlings, often referred to as sweets,
have from 0.1 to 0.3% malic acid in the fruit. In progenies where one or both
parents are homozygous acid (Ma Ma) all the seedlings will have normal acidity. In
other progenies, Ma ma Ma ma or Ma ma ma ma, a quarter or a half of the
seedlings will be of the sweet type. These very low acid types are on the whole
undesirable and have to be discarded.
In progenies where one or both parents are homozygous dominant, the distribution of acidity is normal with the progeny mean approximating the parental mean.
In progenies where sweets are segregating and the sweets are excluded, the mean of
the remaining acid portion is somewhat higher than the parental mean. When this is
known, a fair estimate of the progeny mean for malic acid concentration can be
made from the parental mean.
An unusual feature of this sweet character is that because of its unpleasant
flavor it would have a negative selection value, yet few cultivars are known that are

BREEDING FOR SPECIFIC CHARACTERS

51

homozygous for the dominant acid type and the majority of the cultivars studied are
heterozygous. This suggests that this is an example where selection for quantitative
economic characters has favored selection of heterozygotes for a neutral character
(Williams and Brown 1956).
By knowing the sugar concentration and the malic acid concentration in the
fruits of cultivars, parents can be selected that will produce progenies in which the
majority of the seedlings will combine desirable sugar and acid contents. If the Ma
genotype is known, the number of sweets that will have to be discarded can be
predicted. If there is a selective advantage in the heterozygotes, then Ma Ma ma
ma progenies would all be heterozygous and such crosses would be worth considering. A comparison between a number of cultivars, their color sports, and their
tetraploid sports showed no major differences in the sugar or acid concentration in
their fruits.

Fruit Defects
A number of fruit and flesh defects have an important effect on apple quality. These
include fruit cracking (Opara et al. 1996), calcium defects such as bitter pit (Ferguson
and Watkins 1989), and the flesh disorder watercore, which results in a glassy
appearance of the flesh (Marlow and Loescher 1984). Each of these defects is
strongly influenced by genotype and environment. It is important that susceptibility to these fruit defects be considered in the selection process.
Fruit cracking varies by type (from cuticle cracks to cracks that deeply penetrate
the flesh), intensity, and genotype. Cracking is also greatly affected by the water
status of the plant. Severe cracking of susceptible cultivars can lead to unacceptable
losses. The inheritance of cracking is undetermined.
Calcium defects can be controlled by Ca sprays or sometimes by fertilization
but susceptible cultivars can result in serious production problems. Bitter pit-resistant
seedlings had higher levels of Ca and B and lower levels of Mg and K in both fruit
peel and leaf tissues than susceptibles. Two genes, Bp-1 and Bp-2, controlling Ca
accumulation and distribution within the fruit were hypothesized (Korban and
Swiader 1984). Bitter pit is also strongly affected by rootstock, probably related to
Ca uptake or translocation. M.7 and MM.106 rootstocks reduced bitter pit in
Golden Delicious in South Africa.
Watercore appears to be due to a change in membrane integrity associated with
maturation and ripening, resulting in the accumulation of fluid in the intercellular
spaces and elevated sorbitol concentrations. It is associated with elevated sweetness in the fruit and is therefore preferred by the Japanese market. In some cultivars
the effect disappears in storage; in badly affected cases watercore leads to storage
breakdown resulting in huge economic losses. Most cultivars are susceptible to
varying degrees. Watercore can be severe in some new cultivars such as Fuji and
Williams Pride. Only a few cultivars may be considered completely resistant,
e.g., McIntosh is resistant and the disorder is uncommon in Golden Delicious.

52

APPLES

Disease Resistance
Over the centuries apples have been selected for high productivity, good flavor,
attractive appearance, and long storage life, but in the process the innate resistance
of species to many diseases (see Jones and Aldwinckle 1990) and pests has been
lost. Commercial orchards require a tremendous amount of chemical control to
produce high productivity and blemish-free fruit. However, resistances to most
diseases and pests can be found in various cultivars, suggesting that apples could be
improved genetically to contain multiple resistance for many diseases and pests.
Originally this concept was suggested as a technique to reduce grower costs; however, consumer fear of chemical pesticides has created a market for fruits grown
without pesticides by what has become known as the organic market. Organic
growers attempt to eliminate certain chemical sprays and utilize other techniques,
one of which is the use of genetic resistance. The problem is that it is highly unlikely
to produce an apple that is resistant to all of the natural shocks to which apple is
heir. At present, only very early disease-resistant apples, such as Pristine can
produce blemish-free f1ruits without fungicides. Disease resistance will be important
not only for organic growers but also for most home owners and small orchards. In
the future it may be possible to produce fruit without insecticides and miticides by
a combination of resistance, some of which may be introduced by genetic
transformation, and by other means such as traps and the introduction of beneficial
insects. A new approach called integrated pest management (IPM) strives to reduce
but not to eliminate the use of chemical sprays by various techniques, including
monitoring, the use of beneficial insects, more careful attention to pesticide
application to reduce the amount used, and the use of genetic resistance.
Many of those pathogens and pests for which host resistances have been discovered have shown capability for overcoming those resistances. For example, Vf
resistance to Venturia inaequalis (apple scab) has been overcome in Germany (Parisi
et al. 1993); major gene resistance to Podosphaera leucotrichum (powdery mildew)
is rapidly overcome, sometimes within a single growing season; some strains of
Erwinia amylovora, the fire blight incitant, are capable of causing disease in some
Malus clones previously thought to be resistant. Although Bt genes for bisect
resistance look promising in preliminary tests, experience with the Dipel type
insecticide, crystal spore suspensions of Bt, strongly suggests that insects will be
able to overcome such resistance in time (Cummins and Aldwinckle 1992). There
are other problems. A given disease may have more than one causal agent and
developing cultivars with resistance to one agent may not guarantee resistance to
others. For example, Phytophthora cactorum has long been considered the agent of
crown rot and collar rot of apples, but it now is clear that them are several
Phytophthora species that attack apple rootstocks and produce similar symptoms
and that there are specific host/pathogen interactions (Mircetich and Browne 1989;
Jeffers and Wilcox 1990).
The apple breeder should aim to minimize the likelihood of such breakdowns of
resistance. There seem to be two general options. One is to pyramid resistant genes

BREEDING FOR SPECIFIC CHARACTERS

53

and the other is to aim for tolerance rather than immunity, because this type of
resistance may be more durable (Kellerhals and Furrer 1994). Disease resistance
breeding is difficult but not hopeless. Many resistances have held up for a very long
time. For example, Delicious continues to display useful resistance to fire blight;
Malus floribunda 821 has shown scab resistance in the United States for over a 100
years, and woolly apple aphids continue to be held in check in many areas (but not
all) with the moderate resistance obtained from Northern Spy.
Apple scab. This disease, caused by the fungus Venturia inaequalis, is the most
serious disease of the apple worldwide. It attacks leaves (causing defoliation in
severe cases) and shoots and disfigures the fruit to the point of unsalability. It can
be controlled chemically but at considerable expense and difficulty. Breeding
resistant cultivars is a rational strategy that has been attempted on a large scale for
many years in many countries; new cultivars that are virtually field immune are now
available.
Two forms of resistance are available: polygenic and monogenic. The most
desirable and durable type of resistance is likely to be obtained from a combination
of both types. As with most diseases, some cultivars are much more resistant than
others, which allows selection for high, but not complete, resistance. A great many
biotypes of the pathogen exist and when cultures are grown from single spores and
tested on young leaves of cultivars some give positive and some negative results.
However, resistance that does not act against races is unreliable. An excellent review
of scab, the organism, and resistance strategy is found in MacHardy (1996).
Polygenic resistance is found in the cultivar Antonovka Poltobutanaja, which
at one time was field resistant to all known races of the fungus but is now attacked
by some races. This type of polygenic resistance is also found in selections of some
species, notably M. baccata, M. sargentii, M. sieboldii, and M. zumi calocarpa
(Shay et al. 1962). Forms of Antonovka and selections from its progenies have
been used extensively in breeding programs, particularly in Europe.
Modern breeding programs for scab resistance date to the discovery by L.F.
Hough of field immunity to scab in progenies derived from species crosses made
early in the century by C.S. Crandall at the University of Illinois (Crosby et al.
1992). A cooperative program was initiated by L.F. Hough at the University of
Illinois and later Rutgers University and J.R. Shay at Purdue University and joined
subsequently by E.B. Williams, J. Janick, J.A. Crosby (Purdue); D.F. Dayton and
S.S. Korban (Illinois); and J. Goffreda (Rutgers) among others. This program
spawned a number of independent disease-resistant programs, including those at
Cornell University in the United States and in Australia, Brazil, Canada,
Czechoslovakia, England, France, Germany, Hungary, Italy, and Romania. Species
such as M. micromalus, M. atrosanguinea, M. prunifolia, and others have been
introduced into the program, but the most advanced material carries the floribunda
resistance. Some of this material has been shown to contain resistance to a number
of other diseases, including fire blight, cedar-apple rust, or mildew with some
tolerance to insects and mites. Some selections, such as Enterprise and Liberty,
show multiple disease resistance. At the present time about 70 cultivars from various

54

APPLES

programs around the world have been introduced based on the Vf gene and a number
show promise as potentially important cultivars, including Enterprise, GoldRush,
Jonafree, Liberty, and Pristine. At present the most widely planted scab-resistant
cultivar is Judeline, a juice apple with more than a million trees in France.
The conventional method of testing for scab resistance is to grow seedlings in a
greenhouse at a temperature of 18 to 20C. When the first true leaves are showing
between the cotyledons, the seedlings are sprayed with a spore suspension of mixed
inoculum of the fungus and covered with polyethylene sheeting for 48 h or misted
in a chamber to ensure that the leaf surface does not dry. They are inoculated again
after 10 days as a safeguard. Sporulation is soon evident on the susceptible seedlings,
which can be discarded. Satisfactory infection can be produced only on very young
leaves. The inoculum is prepared by washing the conidia from laboratory cultures
of all races of V. inaequalis plus isolates collected from many sources to give a
spore suspension of a representative mixture of the pathogen. Races and isolates
may also be built up separately on young susceptible seedlings and the conidia
washed off the leaves. Conidial suspensions may be stored frozen for up to one
year. Resistance to leaf infection is generally well correlated with resistance to fruit
infection.
The resistance is scored according to symptoms on young leaves grown under
greenhouse conditions. 0 = no macroscopic evidence of infection; 1 = pinpoint
pits, no sporulation; 2 = irregular chlorotic or necrotic lesions and no sporulation; 3
= few restricted sporulating lesions; and 4 = extensive, abundantly sporulating
lesions. The class M has been added to indicate a mixture of necrotic, nonsporulating, and sparsely sporulating lesions (Shay and Hough 1952a). Only class 4 is
considered as field susceptibleall the others are classified as field resistant and do
not show symptoms of infection when grown outdoors where conditions for scab
are very favorable. Using a broad definition of resistance, crosses of heterozygous
resistant with homozygous susceptibles segregate 1 resistant : 1 susceptible but usable
resistance (completely nonsporulating, i.e., clones 0, 1, and 2) is lower, typically
30%. Through a series of intercrosses (tests of allelism) it was established that 11
of 25 sources of resistance were due to the same gene (Vf), found in M. floribunda
821. A total of six loci for qualitative resistance (defined gene pools) were identified
as follows:
Symbol

Vf
Vm
Vr
Vbj
Vb
Va

Original source

M. floribunda
M. micromalus
M. pumila R12740-7A (a Russian apple;
perhaps M. sieversii)
M. baccata jackii
Hansens baccata #2
Antonovka PI 172623

BREEDING FOR SPECIFIC CHARACTERS

55

Five different virulent races of Venturia inaequalis were identified up to 1993 and
race 6 was recently discovered in Germany (Parisi et al. 1993):
Race

Source

1
2

Worldwide
South Dakota, USA

3
4
5

Nova Scotia, Canada

Lafayette, Indiana, USA
Norwich, England

Ahrensburg, Germany

Susceptible material
Most of the worlds cultivars
M. baccata, Dolgo, Alexis, Bittercrab
segregates of R12740-7A, Geneva
Geneva
Segregates of R12740-7A
Micromalus pit type resistance,
M. atrosangunia 804
Prima (Vf cultivars) but not Evereste
M. Perpetu and M. floribunda 804

The isolates of race 6 examined to date are nonaggressive and have not increased in
Ahrensburg, but the future of Vf is clouded. It is not clear if race 6 is a recent
mutation or merely a biotype that has always existed. The appearance of this race
underscores the necessity of combining qualitative and quantitative genes and to
carefully eliminate any seedlings with type 3 reactions. The class 2 reaction has
always remained field immune until the appearance of race 6. Previous reports of
low levels of scab infection on Vf material had proven to be a result of mislabeling.
But Prima, the first scab-resistant cultivar, exhibited only a 3 reaction type, and
limited leaf sporulation of V. inaequalis can occur in some genotypes in some
environments. At present, literally millions of Vf trees in France (principally the
cultivar Judeline, a juice apple in France) remain scab-free.

Mildew. This scourge of apple is incited by Podosphaera leucotricha. Mildew is

one of the most important diseases of apple, especially in humid climates. Although
it usually does not damage the fruit to any extent, it debilitates the tree and can kill
seedlings. It is a significant problem in nurseries and stoolbeds. This disease,
which attacks the foliage and young shoots, has a very marked effect on the quantity
and the quality of the fruit produced. Mildew varies in its intensity according to the
climate, but is present wherever apples are grown. Some cultivars are very
susceptible, others are highly resistant, and between these two extremes there is a
complete range. Delicious, McIntosh and the Geneva 65 rootstock, for example,
are field resistant, while Idared, Jonathan and the Malling-Merton rootstocks are
extremely susceptible. It is doubtful if stable complete immunity can be found
within the cultivated apple. It is possible, however, to select some that are very
highly resistant and only rarely produce any infected shoots.
Mildew resistance can be inherited quantitatively with no evidence of dominance (Brown 1959). There is evidence that direct assessment of mildew resistance
among cultivars provides a sufficiently reliable guide to their breeding behavior,
but the differences in resistance are not clear until the seedlings are 2 years old.

56

APPLES

McIntosh and Lapins (1966) found large differences in susceptibility to mildew

among 21 clones of McIntosh grown from dormant scions X-rayed at doses from
3.75 to 5 krad. Most showed normal susceptibility, but a few were very susceptible
and some were quite resistant. The changes are attributed to mutations.
In addition to polygenic resistance, single gene immunity has been reported
from M. robusta and M. zumi (Knight and Alston 1968, 1972; Alston 1977), and
from an unknown interspecific: hybrid from Morton Arboretum 8, M. zumi calocarpa, M. sargentii, and M. baccata jackii (Dayton 1977; Korban and Dayton (1983).
However, these monogenic resistances have succumbed to different pathogenic races
of the fungus. Recently, there has been renewed interest in mildew-resistance
breeding using both conventional (Janse et al. 1994; Kruger 1994; and Schmidt
1994) and new methods (Battle and Alston 1994).

Apple canker. Branch- or stem-encircling cankers caused by Nectria galligena

Bres are on the increase and create considerable concern among fruit growers. On
heavy soil in particular, trees seem prone to attack. To date, few investigations have
been carried out on this disease, but at least it is known that there are considerable
differences between cultivars in their susceptibility. Some cultivars, particularly
among the cider apples, are claimed to be very highly resistant (Wormald 1955) as
are M.1 and M.12 among the rootstocks (Moore 1960). Where degrees of
resistance exist, breeding and selection could yield a high degree of resistance
combined with good commercial qualities. Canker resistance has been a major
objective in the Ahrensburg. Germany, breeding program (H. Schmidt, pers. comm.).
Valsa canker, caused by Valsa ceratosperma (Tode ex Fr.) Maire, is one of the
serious apple diseases in Japan and none of the cultivars presently grown are resistant.
Moderate resistance to valsa canker has been located in M. sieboldii (Bessho et al.
1994).

Fire blight. A serious bacterial disease caused by Erwinia amylovora, fire blight
can infect most of the Pomoideae. It can be a devastating disease of pears and,
although it is usually less severe on apples, no cultivar is immune. Once confined to
the United States, fire blight is spreading to most of the apple world. The bacterium
attacks flowers, fruit, shoots, and branches, and the susceptibility of any tree depends
on the condition of the growth, tune of flowering, and environmental conditions at
the time infective material is present. This can lead to conflicting reports on the
susceptibility or resistance of a cultivar, although this may also be due to the existence
of different biological races (Nonnecke 1948).
All commercial cultivars of apple are susceptible if inoculum. from a virulent
culture of the bacterium is injected into succulent tissue, but immunity is known in
some Maims species. There can be considerable resistance to the rapid advance of
the disease into the tissue in some cultivars and very high resistance has been found
in Robusta 5 (M. robusta) and M. fusca H-12. Splendour is nearly immune to the
disease. Delicious and Winesap are among the most resistant cultivars, while
Jonathan, Rome Beauty, and York Imperial are highly susceptible, as well as

BREEDING FOR SPECIFIC CHARACTERS

57

some of the newer cultivars such as Fuji and Gala. Interestingly, a number of
scab-resistant selections such as Prima, Priscilla, Enterprise, and Liberty have
a useful level of resistance; it is likely that resistance was transmitted fortuitously
from M. floribunda 821 along with scab resistance. Yet, a number of scab-resistant
selections are extremely susceptible; an otherwise promising selection, Co-op 25,
was not released for this reason.
There are a number of ways of assessing the susceptibility of seedlings but most
involve inoculation of the tip of actively growing terminal shoots and, after a given
time, measuring the distance the disease has invaded the shoots and comparing the
run of infection. It seems that a satisfactory resistance to fire blight, although not
immunity, can be obtained through careful choice of parents and active screening in
areas with erratic blight. In the U.S. Midwest, susceptible seedlings seldom escape
blight in the field and most susceptible selections become identified naturally.
Fire blight is extremely important in rootstocks because basipetal transmission
of the organism from flower and shoot infection through the trunk can kill the tree.
Several clonal rootstocks including Alnarp 2, M.9, M.26, and Ottawa 3 are
extremely susceptible to E. amylova. This poses a potential calamity in southern
Europe where M.9 is almost universal. In the Geneva program, screening for fire
blight resistance is a top priority. Fire blight may limit the deployment of very
susceptible cultivars, such as Gala, on M.26 in some regions, of the United States
including the Shenandoah Valley, Utah, and the Columbia Basin.
The possibility of obtaining resistance by genetic transformation is under
investigation at Cornell University, Geneva. Results indicate that substantial
resistance can by induced by transfer of the attacin E gene from the giant silk moth,
Hyalophoras cecropia (Norelli et al. 1994).

Crown rot. This disease is caused mainly by the fungus Phytophthora cactorum
(Leb. Cohn) Schroet. in many temperate regions but other species of Phytophthora
are involved in certain areas. Unfortunately, resistance to P. cactorum does not
necessarily indicate resistance to other species. Zoospores in the soil infect the bark
of apple trees at or new ground level and it is possible to avoid infection of the scion
by grafting about 60 cm above soil level. The disease is capable of killing quite
large trees by girdling the main stem especially in poorly drained areas. Since both
rootstock and the scion cultivar may be attacked, it is essential to have resistant
rootstocks and desirable to have resistant cultivars. Phytophthora resistance would
be essential in those selections where self-rooted trees could be of interest.
There appears to be both polygenic and monogenic resistance. McIntosh and
Mellor (1954), using a mixed inoculum to infect trees, suggested that resistance is
partially or completely donibmt. Resistant resistant progenies produced 7%
infected seedlings; resistant susceptible 24%, and susceptible susceptible 93%.
Alston (1970a) reported a single dominant resistance gene (Pc) in Northern Spy.
The moderate resistance to collar rot in some of the Mailing-Merton rootstocks is
derived from Northern Spy. Sources of resistance among dwarfing rootstocks

58

APPLES

include M.9, M.27, and Ottawa 3, but a higher level of resistance is transmitted
by M. sublobata Novole, M. angustafolia, and M. magdeburgensis Hartwig
(Cummins and Aldwinckle 1983). However, there are great differences in pathogenicity between fungal isolates.
A very rapid method for the preselection for collar rot resistance is described by
Watkins and Werts (1971) and later modified by Cummins and Aldwinckle (1983),
where young seedlings 2 to 3 weeks old are grown in a peat-sand mixture that is
inoculated by flooding with a mixture of zoospore suspensions. The effects of
attack are wen after 1 week and the complete results after 2 weeks. By this method,
only the resistant or tolerant ones are left to grow on for further testing. In the
Cornell University, Geneva, rootstock program, 15 to 30% of the seedlings inoculated
with P. cactonan and P. megasperma survive inoculations, flooding, and incubation.
Other methods of testing cultivars and seedlings usually depend on culturing the
fungus in the laboratory and then inoculating with mycelium, infected agar disks
into the stem of growing trees or into cutoff shoots in controlled conditions in the
greenhouse or laboratory and after a given time measuring the infection (Sewell
and Wilson 1959). These methods probably reflect only certain portions of a plants
response to Phylophthora in an orchard setting.

Apple rust. The cedar apple rust fungus, Gymnosporangium juniperi-virginianae Schw., attacks leaves and fruits of apple and has the red cedar (Juniperus
virginiana L.) as its alternate host. It has been quite serious in the eastern United
States. Quince rust, hawthorne rust, and a related rust species in Japan can be
serious problems. A number of high-quality cultivars are resistant. Moore (1940)
and Shay and Hough (1952b) considered resistance to be controlled by a single
dominant gene, with Arkansas Black and McIntosh being homozygous for
resistance; Delicious, Winesap, and Wolf River heterozygous; and Jonathan
and Rome Beauty fully susceptible. However, Mowry (1964), Aldwinckle et
al. (1977), and Chen and Korban (1987) consider resistance to be controlled by
two dominants with the double recessive susceptible. Priscilla was considered
to be homozygous dominant for resistance (Aldwinckle et al. 1977); Golden
Delicious is susceptible.
Apple blotch. Mowry and Dayton (1964) analyzed the progenies from controlled crosses for susceptibility to blotch caused by Phyllosticta solitaria E. & E.
Susceptibility was controlled by two completely dominant genes with duplicate
recessive epistatic interaction between the gene pairs. It was not possible to predict
with any accuracy the genotype of a cultivar from its expressed apple blotch
phenotype.

Glomerella leaf blotch. Golden Delicious was reported heterozgyous for

a dominant gene Gb which confers susceptibility to a new strain of Glomerella

BREEDING FOR SPECIFIC CHARACTERS

59

cingulata (Stonem.) Spauld. & Schrenk. Camilo (1989) found all isolates pathogenic
to every Malus clone except M. sieboldii 2892-22.

Storage rots. Considerable losses can occur from fruit rotting diseases in
storage, and species of Gloeosporium, particularly G. perennans Zellar & Childs,
are largely responsible. In a survey of over 200 cultivars, Alston (1967) found that
some are very susceptible and a few (about 5%) were highly resistant, with Cravert
Rouge the most resistant. A small progeny from Coxs Orange Pippin, which is
very susceptible, crossed with Cravert Rouge produced seedlings with a fair degree
of resistance. When compared with a progeny of Coxs Orange Pippin selfed
which had a mean of 4.7 lesions per fruit, the progeny from the cross with Cravert
Rouge had a mean of 0.8 lesions. Also among the most resistant cultivars in this
survey were Jonathan and its color sport Jonared, and a number of derivatives of
Jonathan also show good resistance to this pathogen. The method of estimating
the susceptibility is to dip the fruits in a spore suspension (50,000 spores/mL) and
store in paper bags at 5C for up to 5 months. Five or more fruits of each cultivar
need to be sampled.

Virus diseases. Diseases caused by viruses, viroids, and mycoplasm-like

organisms present a number of problems to the fruit grower and the propagator
since many (like tomato ringspot virus, rubbery wood, and flat limb) are quite serious
and my disable the tree; others affect the efficiency of the tree, and most cause a
reduction in crop. Viruses pass freely across graft unions, so it is important to have
virus-free rootstocks as well as virus-free fruiting cultivars. The responses of apple
cultivars to viruses differ widely. Some viruses are present in many cultivars without showing any symptoms. With others, such as vein-banding mosaic, the severity
of the symptoms varies from one cultivar to another, some being very severe, some
moderate, some slight, and some symptomless (Luckwill 1953). Other viruses
produce severe symptoms or none at all. Many of the viruses have been present but
undetected in cultivars until fairly recent times when indicator plants have been
found that produce visible symptoms (Fridlund 1989). Most virus can now be
detected by enzyme-linked immnunosorbant assays (ELISA).
Since so many cultivars carry viruses without showing symptoms, it would be
an almost impossible task to try to breed for resistance. There is evidence that
sensitivity my be passed to offspring. In the scab breeding program Russian R12740-7A, which is an indicator for chlorotic leaf spot, passed on the sensitive
reaction to some of its seedlings (Mink and Shay 1962). The vole-resistant apple
rootstock Novole, which is sensitive to apple stem-grooving virus, transmit this
sensitivity to more than half its seedlings. The incorporation of virus resistance into
plants is one of the successes of genetic engineering (Grumet 1994) and this approach
is under investigation in apple (Mauren et al. 1992). The rootstock MM.106 is
very susceptible to tomato ringspot virus; genetically engineered resistance is being
developed at Geneva (D. Gonsalves, personal communication).

60

APPLES

Pest Resistance
Resistance breeding to insect pests began in response to epizootics of woolly apple
aphid in Australia, New Zeal-I South Africa and the southeastern United States.
Woolly apple aphis (Eriosoma lanigerum Hausm.) is a very widespread pest of
apples that is capable of attacking not only the above-ground stem of the tree, causing
cankers, but also die roots, where large colonies can become established under certain
conditions. The above-ground infestations can be controlled by pesticides, but those
on the roots cannot. Woolly apple aphis can be quite a serious problem on nursery
stocks especially in the stoolbed. Breeders at the John Innes Institute and East
Malling Research Station in England and at the United States Department of
Agriculture began breeding for resistance in the early 1920s. Growers had recognized
for some time that Northern Spy and some of its derivatives, including Ivorys
Double Vigour, were resistant to the woolly apple aphid and were more or less
satisfactory for use as rootstocks. Subsequently, Cummins et al. (1981) also identified
Kola, M. halliana, M. hupehensis, M. robusta (Robusta 5), and M. tschonoskii
as resistant. Resistance found in Irish Peach, Malling 15 and M. micromalus is
not transmitted to progeny, at least not in the F1.
In England, Northern Spy was crossed with M.1, M.2, M.9, and M.15 to
produce progenies that were rigorously screened for resistance. The USDA breeders similarly resorted primarily to Northern Spy for resistance. The American
program produced Spy 227. much used today by virologists to detect apple stempitting virus. The Merton Immune series was produced from the English program,
of which Merton Immune 793 remains in use today, while MM.106 and MM.111
the Malling-Merton rootstocks, have gained world-wide acceptance.
Resistance has also been found to the rosy apple aphid (Dysaphis plantaginea
Pass.) and the rosy leaf-curling aphid, which are conditioned by different single
dominants (Table 4). Resistance to the rosy leaf curling aphid (Dysaphis devecta
Wlk.) is controlled by a single dominant gene, Smh, obtained in an open pollinated
selection of M. robusta. Alston and Briggs (1970) found three biotypes of the rosy
leaf-curling aphid and at least three resistant genes in the host. Coxs Orange
Pippin carries Sd1 and is resistant to biotypes 1 and 2; Northern Spy has Sd2 and is
resistant to biotype 1 only; and a selection from M. robusta with Sd3 is resistant to
all three biotypes.
Goonewardene et al. (1998; Goonewardene and Williams 1988) screened progenies, accessions, and scab-resistant selections of Malus and found several selections and progenies with multiple pen and disease tolerance or resistance from
observations in the field, laboratory, and greenhouse. The arthopods investigated
included apple maggot, Rhagoletis pomonella (Walsh); codling moth, Laspyresia
pomonella (L.); European red mite, Panonychus ulmi (Koch.); plum curculio,
Conotrachelus nenuplar (Herbst); and redbanded leafroller, Argyrotaenia velutinana (Walker). Insect resistance was typically compared by the results of differential feeding on June-drop apples as compared to a control. Seven apple selections
(E11-24, E14-32, E36-7, D7-47, E7-54, E29-56, and E31-10) have been released as
advanced germplasm lines with multiple resistance to pests and diseases

BREEDING FOR SPECIFIC CHARACTERS

61

(Goonewardene 1987; Goonewardene and Howard 1989) and are presently under
evaluation.
Results from the transformation of apple with Bt genes (see Transformation)
indicate that resistance to lepidopterous bisects, including codling moth and perhaps apple maggot, may be obtained. It is uncertain whether the ultimate strategy
will be to add them to new cultivars or to develop trap trees (e.g., by incorporating
into crab pollinators) that will be used to reduce the population levels of the insect
in commercial orchards (Dandekar, personal communication). Clearly, recent
advances in biotechnology are opening up a new em in insect resistance of apple.

Rootstock Breeding
The rootstock is a critical part of the apple tree, and rootstock breeding must be
considered with as much care as that of the scion cultivars. Rootstocks must satisfy
many special requirements and the testing period is long and difficult, for it is only
when the scion is grafted on the rootstock and has reached full cropping that the
potential of the rootstock can be assessed. World breeding programs have been
reviewed by Cummins and Aldwinckle (1983) and Ferree and Carlson (1987).
There are a number of special requirements for any rootstock: ease of propagation either vegetatively or by seeds; clean upright stern, easy to bud or graft; a root
system that will provide adequate anchorage to support the tree without staking;
and absence of stock-scion incompatibility, suckers, and burrknots (root primordia
on the trunk). Rootstocks should induce early and heavy cropping and be resistant
to a number of pests and diseases. Finally, specific rootstocks should offer a range
of tree size control. The objectives of the Cornell University, Geneva, apple rootstock breeding program are shown in Table 6, but specific objectives can be expected
to vary in other parts of the world. For example, rootstocks my have to be extremely
winter hardy, adapted to alkaline or waterlogged soil to increase calcium uptake to
avoid fruit defects such as bitter pit, or offer resistance to specific root diseases, e.g.,
black root rot (caused by Xylaria mali Fromme), limited to the southern United
States; Rosellinia root rot (Rosellinia necatrix (Hart.) Berl.), important in India and
California, and Sclerotum rolfsii Sacc., important in Israel and other subtemperate
regions. At present there is a wide range of vegetatively propagated apple rootstocks
that impart different degrees of vigor, ranging from those producing dwarf, compact
trees to those producing very large, vigorous trees, but there is a need for a greater
range in winter-hardy stocks and for resistance to many pests and diseases.
One of the problems of vegetatively propagated rootstocks is virus diseases,
which may be present in the rootstocks and then passed to the scion. One way of
overcoming this is to raise rootstocks from seeds. This practice has been in use for
a great many years and one of the reasons for the change to clonal stocks was the
great lack of uniformity among the seedlings, since those used as rootstocks displayed
the enormous variability found in most apple seedling populations. In recent years
apomictic seedlings, which are much more uniform, have been bred from some
Malus species. These, too, present problems; some exhibit extreme sensitivity to

62

APPLES

TABLE 6. Objecti
ves of Cor
nell Uni
ver
sity
va Apple Rootstoc
k Br
eeding
Objectiv
Cornell
Univ
ersity
sity,, Gene
Genev
Rootstock
Breeding
oject
Project
Pr

Attribute
Essential
Resistant to Erwinia amylovora (fire blight)
Resistant to Phytophthora spp. (crown rot)
Readily propagable
Liners smooth, relatively free of spines
Liners thrifty in nursery row; easily buddable
Induce heavy, early fruit production
Early hardening of scion and collar
Tolerant to low midwinter temperatures
Late-leafing in early spring
Free of burrknots
Reacting to TmRSV as does
Induce no more scion susceptibility
to E. amylovora than
Important
Resistant to woolly apple aphids
Roots structurally strong (not brittle)
Well anchored
Few or no suckers in orchard
Inducing early maturation of fruit
Resistant to pine voles
Hardy to very low temperatures in midwinter
Tolerant of chlorotic leaf spot, apple stem
grooving, and apple stem pitting viruses
Helpful
Leaves red or otherwise distinctive
Moderately resistant to Venturi inaequalis
Moderately resistant to Podosphaera leucotricha
Resistant to meadow vole
Special Objecti
ves
Objectiv
Tree size control: full range, from dwarfing
through vigorous
Some stocks tolerant of poorly drained soils
Some stocks tolerant of drought

Minimum
standard

M.7
M.9
MM.106
MM.106
MM.106
MM.106
M.2
MM.111
M.9
M.2
M.7

Ideal
standard

R5
Bud.9
A.2
O.3
A.2
M.9
M.9
A.2
K-14
Hibernal
Delicious

M.7
MM.111
M.26
M.7
MM.106
M.2
R5
M.26

R5
M.7
M.111
Hibernal
M.9
Novole
R5

M.9

M.9

Bud.9;O.3
M.7
M.9
Sugar Crab
M.9
Alnarp 2
M.13
MM.111

virus infection from the scion. They have the advantages that propagation is simple
and the seedlings are uniform and free from virus. A number of species have been
tried, particularly M. sieboldii, M. sargentii, M. hupehensis, M. sikkimensis, M.
toringoides, and hybrids between some of these. M. hupehensis and M sikkimensis
are autonomous apomicts and can produce apomictic seeds without pollination,

ACHIEVEMENTS AND PROSPECTS

63

which produce seedlings of the maternal type. Other species are facultative apomicts
and a small percentage of the seedlings may be hybrids. Since sour of these require
pollination to produce seeds, they can be pollinated with a homozygous red-leafed
plant and the undesirable hybrids immediately identified by the pigmentation of the
leaves. When virus-free clones are used as scions much of the incompatibility
disappears. Some M. sieboldii hybrids have shown a good affinity for virus-infected
scions and appear tolerant of latent virus infection (Schmidt 1970).
Evaluating vegetatively propagated rootstocks is by its nature a difficult and
long-term project. In the initial stages, each new seedling must be tested to determine its ability to root and to produce a good stem, its rate of multiplication, and
whether it has all the normal qualities that make a good healthy plant. If disease or
pest resistance is being incorporated, this can be determined at this stage. But final
testing of new candidates must be carried out in many areas to determine adaptability. In North America, regional programs (NC 140) have been organized to
evaluate rootstock-scion (stionic) interactions with location.

ACHIEVEMENTS AND PROSPECTS

The apple is an ancient crop that has been under rather continuous selection since
antiquity. Until very recently the apple industry has depended on nurseryman and
fruit-grower selections of scion cultivars and rootstocks with about a dozen genotypes dominating the commercial markets of the world. Deficiencies inherent in
these genotypes in many cases have been ameliorated through cultural practices to
prop them up, including specific rootstocks combinations, insect and disease control
via chemical pesticides, growth regulators, and special handling and storage
technology.
Plant breeding did not contribute significant new cultivars to the industry during
the first half of the twentieth century but in the last half, the results of a number of
breeding efforts around the world have begun to show an impact and the traditional
cultivars are slowly being replaced. There is now an increasing demand for new
selections produced by conventional breeding and a number of extremely promising
selections are under test throughout the world. Some of these contain resistance to
various diseases; a number have greatly improved storage life, improved textural
quality, winter hardiness, and/or different tree forms. Some of these genotypes
together with intensive methods of production, particularly high-density orchards,
can be expected to have a profound effect on the future apple industry of the world.
The selection of superior genotypes from genetically variable populations derived
from sexual recombinations is powerful because it is evolutionary and progress can
be cumulative, with improved individuals continually serving as parents for
subsequent cycles of breeding. Thus, genetic improvement of apple has made
substantial changes when the effort has been well supported and long term.
Successful programs are now underway in the United States, Japan, Australia, New
Zealand, and various countries of Europe.

64

APPLES

However, there have been restraints to conventional plant breeding that are
especially limiting in tree fruits with long juvenile period and that are represented
by unique, highly selected genotypes. These include the reliance on naturally
occurring variation which may be unavailable, or more likely found in very primitive and unadapted material; restraints due to the inability of the sexual system to
incorporate genes from nonrelated species and especially the inability of the sexual
system to incorporate small changes without recombination resulting in the loss of
desirable unique combinations; the difficulty of selection in detecting infrequent or
rare recombinations; and the dependence of conventional breeding upon time to
generate cycles of recombination, space to grow the necessary population to recover
superior recombinants, and resources to be able to select, identify, and evaluate
desirable recombinants.
Some of these limitations may be overcome by novel strategies from advances
in biotechnology. Recombinant DNA has been especially attractive to fruit breeders because it offers a way to overcome the limitations of the sexual system by
permitting the introduction of foreign genes heretofore unusable and more important makes it possible to introduce small discrete defined changes into established
genotypes. The use of molecular markers will facilitate selection and provide the
ability to identify clones with a great deal of precision for cultivar protection and
perhaps to tag clones with a genetic brand.
Recent advances by a new generation of fruit breeders have demonstrated that
these new technologies are feasible in apple. The transfer of genes effective against
fire blight and insertions of genes from Bacillus thuringensis that may control
lepidopterous pests such as codling moth have been achieved and are under test.
The future promises to be very exciting. However, it remains clear that these new
technologies will not replace but will merely complement conventional breeding.
Most of the advances in DNA transformation will in the end involve improvements
due to the insertions of single genes, which represent a conservative approach to
improvement. Truly transgressive changes must still rely on sexual recombination.
How will breeders meet the challenges that are required? At present most are
employed by government or state research stations or by universities where resources
to continue large programs are becoming restricted. The private sector, particularly
nurseries selling apple stock, has not gone into apple breeding, preferring to scavenge
naturally occurring variation and sports from growers and to distribute selections
from public sector breeding. What is clearly needed are new levels of cooperation
between breeding programs within and between countries and between the public
and private sector. The cooperative disease-resistant breeding program between
Purdue University, Rutgers University, and the University of Illinois, which has
been active for over 40 years is a good example, but the future is clouded by the
level of individual support from each university. In Europe, three new cooperative
efforts may stand as good models for the future: (1) Apple breeders at INRA and
private nurseries; (2) apple breeders at the Institute for Horticultural Research at
East Malling are strongly underwritten by a group of large European and American

LITERATURE CITED

65

nurseries; (3) apple breeding in Belgium with cooperation between University

breeders and private nurserymen.
The acceptance of plant patenting and the return of royalties (some of which nay
be returned to the breeding program) is now firmly entrenched in the public sectors,
since research funds for long-term program have become limited or disappeared.
But plant variety protection and the lure of royalties while engendering cooperation
between the public and private sector could have the unhappy consequence of
restricting, discouraging, or impeding cooperation among breeders. A new wrinkle
of cultivar licensing has occurred in parts of the fruit industry where new cultivars
are restricted to very large growers and royalties are returned in the form of a
surcharge on each box of fruit (Janick 1995). This could have a profound effect on
plant breeding.
Despite all of these obstacles and problems it is clear that the future of apple
breeding will continue to be exciting with the introduction of new technologies,
new genes, and new cultivars. Clearly, apples will continue to be one of the most
popular fruits of humans, and plant breeders will have a key role in this industry.

ACKNOWLEDGMENTS
The contributions of the late A.G. Brown, author of the original chapter, H.E.
Aldwinckle, J.P. Bolar, A.M. Dandekar, P.L. Forsline, R.M. Skirvin, and R.H.
Zimmerman are gratefully acknowledged.

LITERATURE CITED
Abbott, D. L. 1955. Temperature and the dormancy of apple seeds. p. 746-753. In: Rep.
Int. Congr. Scheveningen.
Aldwinckle, H. S., H. L. Gustafson, and R. C. Lamb. 1976. Early determination of
genotypes by forced flowering of test cross progenies. Euphytica 25:185-19 1.
Aldwinckle, H. S., R. C. Lamb, and H. L. Gustafson. 1977. Nature and inheritance of
resistance to Gymnosporangium juniperi-virginianae in apple cultivars.
Phytopathology 67:259-266.
Alston, F. H. 1967. Varietal response to Gloeosporium perennans in the apple. p. 132-134.
In: Rep. E. Malling Res. Sta. 1966.
Alston, F. H. 1970a. Resistance to collar rot, Phytophthora cactorum (Leb. and Cohn)
Schroet. in apple. p. 143-145. In: Rep. E. Malling Res. Sta. 1969.
Alston, F. H. 1970b. Integration of major characters in breeding commercial apples. p.
231248. In: Proc. Eucarpia Fruit Breeding Symp., Angers.
Alston, F. H. 1973. Apple scion varieties: p. 134. In: Preliminary selection. Annu. Rep.
E. Malling Res. Sta. 1972.
Alston, F. H. 1976. Dwarfing and lethal genes in apple progenies. Euphytica 25:505-514.
Alston, F. H. 1977. Practical aspects of breeding for mildew (Podosphaera leucotricha)
resistance in apples. p. 4-13. In: Proc. Eucarpia Symp. Tree Fruit Breeding, 1976,
Wageningen.

66

APPLES

Alston, F. H., and J. B. Briggs. 1968. Resistance to Sappaphis devecta (WIk.) in apple.
Euphytica 17:468-472.
Alston, F. H., and J. B. Briggs. 1970. Inheritance of hypersensitivity to rosy apple aphid
Dysaphis plantaginea in apple. Can. J. Genet. Cytol. 12-257-258.
Alston, F. H., and J. B. Briggs. 1977. Resistance genes in apple and biotypes of Dysaphis
devecta. Ann. Appl. Biol. 97:75-81.
Alston, F. H., and L. D. Hunter. 1971. Scab resistance and the hydroxy phloridzin,
sieboldin. p. 95. In: Annu. Rep. E. Malling Res. Sta. 1970.
Aronson, A. I. 1994. Bacillus thuringensis and its use as a biological insecticide. Plant
Breeding Rev. 12:19-45.
Barrett, H. C., and T. Arisumi. 1952. Methods of pollen collection, emasculation and
pollination in fruit breeding. Proc. Am. Soc. Hort. Sci. 59:259-262.
Battle, L., and F. H. Alston. 1994. Isoenzyme aided selection in the transfer of mildew
(Podosphaera leucotricha) resistance from Malus hupehensis to the cultivated apple.
p. 17-20. In: H. Schmidt and M. Kellerhals (eds.), Progress in temperate fruit breeding.
Kluwer Academic, Dordrecht, Netherlands.
Battle, L., F. H. Alston, and K. M. Evans. 1995. The use of the isoenzymic marker gene
Got-1 in the recognition of incompatibility S alleles in apple. Theor. Appl. Genet.
90:303-306.
Beakbane, A. B. 1941. Anatomical studies of stems and roots of hardy fruit trees, In: the
anatomical structure of some clonal and seedling apple rootstocks stem- and
root-grafted with a scion variety. J. Pomol. Hort. Sci. 18:344-367.
Beakbane, A. B., and P. K. Majumder. 1975. A relationship between stomatal density
and growth potential in apple rootstocks. J. Hort. Sci. 50:285-289.
Beakbane, A. B., and E. C. Thompson. 1939. Anatomical studies of stems and roots of
hardy fruit trees, 11: the internal structure of the roots of some vigorous and some
dwarfing rootstocks and the correlations of structure with vigor. J. Pom. 17:141-149.
Beakbane, A. B., and E. C. Thompson. 1947. Anatomical studies of stems and roots of
hardy fruit trees, 4: the root structure of some new clonal apple rootstocks budded
with Coxs Orange Pippin. J. Pom. 23:206-211.
Bergendal, P. 0. 1970. On the inheritance of red color in crosses between Golden Delicious and various red apple mutants. p. 181-194. In: Proc. Eucarpia Fruit Breeding
Symp., Angers.
Bessho, H., J. Soejima, Y. Ito, and S. Komori. 1993. Breeding and genetic analysis of
apple in Japan. In: T. Hayashi et al. (eds.). Techniques on gene diagnosis and breeding
in fruit trees. FTRS, Japan.
Bessho, H., S. Tsuchiya, and J Soejima. 1994. Screening methods for resistance to Valsa
canker. p. 49-52. In: H. Schmidt and M. Kellerhals (eds.). Progress in temperate fruit
breeding. Kluwer Academic, Dordrecht, Netherlands.
Bishop, C. J. 1959. Radiation-induced fruit colour mutations in apples. Can. J. Genet.
Cytol. 1:118-123.
Blazek, J. 1983. Inheritance and genetic variation of spurred growth habit in Apples.
Acta Hort. 140:55-62.
Blazek, J., and F. Paprustein. 1988. Complex number for evaluation of apple breeding
stock. Acts. Hort. 224:185-196.
Bolar, J. T. 1995. Examination of factors affecting the transformation of Marshall
McIntosh apple by Agrobacterium tumefaciens. Thesis, Cornell University, Ithaca, NY

LITERATURE CITED

67

Bondt, A. D., K. Eggermont, P. Druart, M. D. Vil, I. Goderis, J. Vanderlayden, and W. F.

Broekaert. 1994. Agrobacterium- mediated transformation of apple (Malus domestica
Borkh): an assessment of factors affecting gene transfer efficiency during early transformation steps. Plant Cell Rep. 13:587-593.
Bowman, H. G., and D. Hultmark. 1987. Cell-free immunity in insects. Annu. Rev.
Microbiol. 41:103-126.
Bringhurst, R. S. 1983. Breeding strategy. p. 147-153. In: J. Janick and J. N. Moore
(eds.). Methods in fruit breeding. Purdue University Press, West Lafayette, IN.
Brittain, W. H. 1933. Apple pollination studies in the Annapolis Valley, N. S., Canada.
1928-32. Can. Dept. Agric. Bull. 162.
Broothaerts, W., G. A. Janssens, P. Proost, and W. F. Broekaert. 1995. cDNA cloning
and molecular analysis of two self-incompatibility alleles from apple. Plant Mol.
Biol. 27:499-511.
Brown, A. G. 1954. Frost susceptibility of 2x and 4x Fertility pear. p. 7. In: 45th Annu.
Rep. John Innes; Inst.
Brown, A. G. 1959. The inheritance of mildew resistance in progenies of cultivated
apple. Euphytica 8:81-88.
Brown, A. G. 1960. The inheritance of shape, size and season of ripening in progenies
of the cultivated apple. Euphytica 9:327-337.
Brown, A. G. 1964. Fruit breeding. p. 37-38. In: 55th Annu. Rep. John Innes Inst.
Brown, A. G. 1966. New fruits from old. p. 10-24. In: Fruit present and future. Royal
Hort. Soc., London.
Brown, A. G. 1975. Apples, p. 3-37. In: J. Janick and J. N. Moore (eds.). Advances in
fruit breeding. Purdue University Press, West Lafayette, IN.
Brown, A. G., and D.M. Harvey. 197 1. The nature and inheritance of sweetness and
acidity in the cultivated apple. Euphytica 20:68-80.
Brown, S. K. 1992. Genetics of apple. Plant Breeding Rev. 9:333-365.
Camilo, A. P. 1989. Genetic resistance in Malus to Glomerella cingulata (Stoneman)
Spaulding and Von Schrenck: Sources of resistance, leaf infection, progeny evaluation,
and pathogenicity. Thesis, Cornell University, Ithaca, NY.
Chan, B. G., and J. C. Cain. 1967. The effect of seed formation on subsequent flowering
in apple. Proc. Am. Soc. Hort. Sci. 91:63--68.
Chen, H., and S. S. Korban. 1987. Genetic variability and the inheritance of resistance
to cedar-apple rust in apple. Plant Pathol. 36:168-174.
Chevreau, E., Y. Lespinasse, and M. Gallet. 1985. Inheritance of pollen enzymes and
polyploid origin of apple (Malus xdomestica Borkh). Theor. Appl. Genet. 71:268-277.
Chyi, Y. S., and N. F. Weeden. 1994. Relative isozyme banding pattern permit the identification of the 2n gamete parent for triploid apple cultivars. HortScience 19:818-819.
Compton, M. E., and R. E. Veilleux. 1992. Thin layer morphogenesis. Hort. Rev. 14:239264.
Conner, P. J., S. K. Brown, and N. F. Weeden. 1995. Examination of apple (Malus
xdomestica) fruit quality and tree and fruit morphology using molecular markers.
HortScience 30:832 (Abstr. 448).
Crane, M. B. 1953. The genetics and breeding of tree fruits p. 687-695. In: Rep. 13th
Int. Hort. Congr. London.
Crane, M. B. and W. J. C. Lawrence. 1933. Genetical studies in cultivated apples. J.
Genet. 28:265-296.

68

APPLES

Crosby, J. A., J. Janick, P. C. Pecknold, S. S. Korban, P. A. OConnor, S. M. Ries, J.

Goffreda, and A Voordeckers. 1992. Breeding apples for scab resistance: 1945-1990.
Fruit Var. J. 46:145-166.
Cummins, J. N. and H. S. Aldwinckle. 1983. Breeding apple rootstocks. Plant Breeding
Rev. 1:294-394.
Cummins, J. N. and H. S. Aldwinckle. 1988. New directions in rootstock breeding.
HortScience 23: 100-101.
Cummins, J. N. and H. S. Aldwinckle. 1992. Resistant host plants-The too-often absent
component of Integrated Fruit Production. Proc. Australian Soc. Hort. Sci. 1: 157163.
Cummins, J. N., P. L. Forshne, and J. D. Mackenzie. 1981. Woolly apple aphid
colonization on Malus cultivars. J. Am. Soc. Hort. Sci. 106:26-30.
Dandekar, A. M., G. H. McGranahan, and D. J. James. 1992. Transgenic woody plants.
p. 129-151. In: Shain-dow and R. Wu (eds.). Transgenic plants, Vol. 2. Academic,
New York.
Darlington, C. D., and A. A. Moffett. 1930. Primary and secondary chromosome balance
in Pyrus. J. Genet. 22:129-15 1.
Dayton, D. F. 1959. Red color distribution in apple skin. Proc. Am. Soc. Hort. Sci.
74:7281.
Dayton, D. F. 1977. Genetic immunity to apple mildew incited by Podosphaera
leucotricha. HortScience 12:225-226.
Dayton, D. F., and E. B. Williams. 1968. Independent genes in Malus for resistance to
Venturia inaequalis. Proc. Am. Soc. Hart. Sci. 92:89-.94.
Decourtye, L. 1967. A study of some characters controlled by simple genetic means in
apple trees (Malus sp.) and pear trees (Pyrus communis) (in French). Ann. Amelior.
Plantes 17:234--266.
Dermen, H. 1949. Are the pomes amphidiploid? A note on the origin of the Pomoideae.
J. Hered. 40:221-222.
Dermen, H. 1965. Colchiploidy and histological imbalance in triploid apple and pear.
Am. 1. Bot. 52:353-359.
Downing, D. L. (ed.). 1989. Processed apple products. Van Nostrand Reinhold, New York.
Durham, R. E., and S. S. Korban. 1994. Evidence of gene introgression in apple using
RAPD markers. Euphytica 79:109-114.
Eichholtz, D. A., H. A. Robitaille, and P. M. Hasegawa. 1979. Adventive embryony in
apple. HortScience 14:699-700.
Einset, J. 1945. The spontaneous origin of polyploid apples. Proc. Am. Soc. Hort. Sci.
46:91-93.
Einset, J. 1952. Spontaneous polyploidy in cultivated apples. Proc. Am. Soc. Hort. Sci.
59:291-302.
Einset, J., and B. Imbofe. 1947. Chromosome numbers of apple varieties and sports.
Proc. Am. Soc. Hort. Sci. 50:45-50.
Einset, J., and B. Imbofe. 1949. Chromosome numbers of apple varieties and sports, 2.
Proc. Am. Soc. Hort. Sci. 53:197-201.
Einset, J., and B. Imhofe. 195 1. Chromosome numbers of apple varieties and sports, 3.
Proc. Am. Soc. Hort. Sci. 58:103-108.
Ellis, R. H. 1985. Seed storage and germination of apple and pear. In: Long term storage
of major temperate fruits. Int. Board Plant Genetic Resources, Rome.

LITERATURE CITED

69

Elobeidy, A., and S. S. Korban. 1988. The effect of thisdiazuron on shoot regeneration
from apple leaf discs. HortScience 23:755.
Ely, S. 1993. The engineering of plants to express Bacillus thuringiensis d-endotoxins.
p. 105-124. In: P. F. Entwistle (ed.). Bacillus thuringiensis, an environmental
biopesticide. Wiley, New York.
FAO. 1995. 1994 production yearbook. United Nations, Rome.
Fasolo, F., R. H. Zimmerman, and I. Fordham. 1989. Adventitious shoot formation on
excised leaves of invitro, grown shoots of apple cultivars. Plant Cell Tissue Organ
Cult. 16:75-87.
Ferree, D. C., and R. F. Carlson. 1987. Apple rootstocks. p. 107-143. In: R. C. Rom and
R. F. Carlson (eds.) Rootstocks for fruit crops. Wiley, New York.
Ferguson, I. B., and C. B. Watkins. 1989. Bitter pit in apple fruit. Plant Breeding Rev.
11:229-288.
Forsline, P. L. 1995. Adding diversity to the national apple germplasm collection: collecting wild apple in Kazakhstan. New York Fruit Quart. 3(3):3-6.
Forsline, P. L., E. E. Dickson, and A. D. Djangalieu. 1994. Collection of wild Malus,
Vitus and other fruit species genetic resources in Kazakhstan and neighboring
republics. HortScience 29:433.
Fridlund, P. R. (ed.). 1989. Virus and viruslike diseases of pome fruits and simulating
noninfectious disorders. Coop. Ext. Publication SP003. Washington State University,
Pullman, WA.
Gilbert, N. 1967. Additive combining abilities fitted to plant breeding data. Biometrics
23:45-49.
Gonzalez-Cepeda, I. A. 1992. The Mexican apple industry as it exists today. Compact
Fruit Tree 25:95-98.
Goonewardene, H. F. 1987. El 1-24, E14-32 and E36-7 apple germplasm with multiple
pest resistance. HortScience 22:1346-1348.
Goonewardene, H. F., and P.H. Howard. 1989. E7-47, E7-54, E29-56, and E31-10 apple
germplasm with multiple pest resistance. HortScience 24:167-169.
Goonewardene, H. F., and E. B. Williams. 1988. Arthropod and disease resistance in
selections of Malus sp. originating from the cooperative apple breeding program.
Purdue Univ. Agric. Exp. Sta. Bull. 537, West Lafayette, IN.
Goonewardene, H. F., P. H. Howard, and J. Triscari. 1988. Arthropod and disease
resistance in progenies of Malus sp. originating from the USDA (ARS)-Purdue
program. Purdue Univ. Agric. Exp. Sta. Bull. 540, West Lafayette, IN.
Gray, J. E., S. Picton, J. J. Giovannoni, and D. Grierson. 1994. The use of transgenic
and naturally occurring mutants to understand and manipulate tomato fruit ripening.
Plant Cell Environ. 17:557-571.
Grumet, R. 1994. Development of virus resistant plants via genetic engineering. Hort.
Rev. 12:47-79.
Hagens, D. M. 1992. Genetic studies in Malus: integration of isozymes, DNA markers
and morphological traits. Thesis. Cornell University, Ithaca, NY.
Hall, A. D. and M. B. Crane. 1933. Ile apple. Martin Hopkinson, London.
Hanna, W. W. and L. E. Towill. 1995. Long-term pollen storage. Plant Breeding Rev.
133:170-207.
Hauagge, R. and J. N. Cummins. 1991. Genetics of length of dormancy period in Malus
vegetative buds. J. Am. Soc. Hort. Sci. 116:121-126.

70

APPLES

Heilborn, 0. 1935. Reductive division, pollen lethality and polyploidy in apple. Acta
Host. 11:129-184.
Hemmat, M., N. F. Weeden, A. G. Manganaris, and D. M. Lawson. 1994. Molecular
marker linkage map of apple. J. Hered. 85:4-11.
Hemmat, M., N. F. Weeden, and S. K. Brown. 1995. Molecular markers for scab resistant
(Vf) region in apple. HortScience 30:850 (Abstr. 914).
Henning, W. 1947. (Morphological, systematic and genetic investigations on the spades
and species-hybrids of the genus Malus.) Zchter 17:289-349 (Plant Breeding Abstr.
19:376).
Hofer, M., and Y. Lespinasse. 1996. Haploidy in apples. In: In vitro haploid production
in higher plants. Kluwer Academic, Dordrecht, Netherlands.
Hfte, H., and H. R. Whiteley. 1989. Insecticidal crystal proteins of Bacillus thuringiensis
Microbiol. Rev. 53:242-255.
Hunter, A. W. S. 1934. Tetraploidy in vegetative shoots of the apple induced by the use
of colchicine. J. Hered. 45:15-16.
Hutchinson, A. 1967. Inheritance of stature, stooling, rooting ability, and other
characteristics in progenies of Malling DC. p. 76-82. In: Annu. Rep. Hort. Inst. Ont.
1966.
James, D. J. 1991. Agrobacterium-mediated transformation of apple (Malus pumila Mill.).
p. 213-226. In: M.R. Abuja (ad.). Woody plant biotechnology. Plenum, New York.
James, D. J., and A. M. Dandekar. 1991. Regeneration and transformation of apple
(Malus pumila Mill.). p. 1-18. In: K. Lindsey (ad.). Plant tissue culture manual.
Kluwer Academic, Dordrecht, Netherlands.
James, D. J., A. J. Passey, and D. C. Deeming. 1984. Adventitious embryogenesis and
the in vitro culture of apple seed parts. J. Plant Physiol. 115:217-229.
James, D. J., A. J. Passey, and E. Rugini. 1988. Factors affecting high frequency plant
regeneration from apple leaf tissues cultured in vitro. J. Plant Physiol. 132:148-154.
James, D. J., A. J. Passey, D. J. Barbara, and M. W. Bevan. 1989. Genetic transformation
of apple (Malus psunila Mill.) using a disarmed Ti-binary vector. Plant Can Rep.
7:658661.
James, D. J., A. J. Passey, A. D. Webster, D. J. Barbara, P. Visa, A. M. Dandekar, and S.
Uratsu. 1993a. Transgenic apples and strawberries: advance in transformation,
introduction of genes for insect resistance and field studies of tissue cultured plants.
Acta Hort. 336:179-194.
James, D. J., S. Uratsu, J. Chang, P. Negri, P. Viss, and A. M. Dandekar. 1993b. Acetosyringone and osmoprotectants like betaine or proline synergistically enhance Agrobacterium-mediated transformation of apple. Plant Cell Rep. 12:559-563.
James, D. J., A. J. Passey, and S. A. Baker. 1994. Stable gene expression in transgenic
apple tree tissues and segregation of transgenes in the progeny: preliminary evidence.
Euphytica 77:119-121.
Janick, J. 1974. The apple in Java. HortScience 9:13-15.
Janick, J. 1995. Breeding to biotechnology: Keeping the competitive edge. p. 22-28. In:
Proc. 2nd Hort. Ind. Tech. Conf., 27-30 July, 1994 Austral. Soc. Hort. Sci.
Janse, J., J. J. Verhaegh, and A. P. M. den Mijs. 1994. Early selection for partial resistance
to powdery mildew, Podosphaera leucotricha (Ell. at Ev.) Salm. in apple progenies.
p. 13-15. In: H. Schmidt and M. Kellerhals (eds.). Progress in temperate fruit breeding.
Kluwer Academic, Dordrecht, Netherlands.

LITERATURE CITED

71

Jeffers, S., and W. F. Wilcox. 1990. Phytophthora crown, collar and root rots. p. 43-45.
In: A. L. Jones and H. S. Aldwinckle (eds.). Compendium of apple and pear diseases.
APS Press, St. Paul, MN.
Jones, A. L., and H. S. Aldwinckle (eds.). 1990. Compendium of apple and pear diseases.
APS Press, St. Paul, MN.
Karnatz, A. 1988. New apple cultivars after self-pollination. Acta Hort. 224:169-175.
Kellerhals, M., and B. Furrer. 1994. Approaches for breeding apples with durable disease
resistance. Euphytica 77:31-35.
Kelsey, D. F., and S. K. Brown. 1992. McIntosh Wijick: a columnar mutation of
McIntosh apple proving useful in physiology and breeding research. Fruit Var. 1.
46:83-87.
Kemmer, E. 1953. On the primary and the fertile stages of apple orchards (in German).
Zchter 23:122-127.
Keulemans, J., R. Eyssen, and G. Colda. 1994. Improvement of seed act and seed germination in apple. p. 225-228. In: H. Schmidt and M. Kellerhals (ads.). Progress in
temperate fruit breeding. Kluwer Academic, Dordrecht, Netherlands.
King, G. 1995. Preliminary European apple genetic linkage map. p. 5. In: European
Apple Issue 3.
Klein, L. G., R. D. Way, and R. C. Lamb. 1961. The inheritance of a lethal factor in
apples. Proc. Am. Soc. Hort. Sci. 77:50-53.
Knight, R. L., and F. H. Alston. 1968. Sources of field immunity to mildew (Podosphaera
leucotricha). Can. J. Genet. Cytol. 10:294-298.
Knight, R. L., and F. H. Alston. 1969. Developments in apple breeding. p. 125-132. In:
Annu. Rep. E. Malling Res. Sta. 1968.
Knight, R. L., and F. H. Alston. 1972. Progress at East Malling in breeding for resistance
to apple mildew Podosphaera leucotricha. p. 317-324. In: Annu. Rep. E. Malling
Res. Sta., 1971.
Knight, R. L., J. B. Briggs, A. M. Massee, and H. M. Tydeman. 1962. The inheritance of
resistance to woolly aphid, Eriosoma lanigerium (Hausmn.) in the apple. J. Hort.
Sci. 37:207-218.
Koller, B., A. Lehmann, J. M. McDermot, and C. Gessler. 1993. Identification of apple
cultivars using RAPD markers. Theor. Appl. Genet. 85:901-904.
Koller, B., L. Gianfranceschi, N. Seglias, J. McDermot, and C. Gessler. 1994. DNA
markers linked to Malus floribunda 821 scab resistance. Plant Mol. Biol. 26:597-602.
Korban, S. S., and D. F. Dayton. 1983. Evaluation of Malus germplasm for sources of
resistance to powdery mildew. HortScience 18:219-220.
Korban, S. S., and R. M. Skirvin. 1994. Nomenclature of the cultivated apple. HortScience
19:177-180.
Korban, S. S., and J. M. Swiader. 1994. Genetic and nutritional status in bitter pit-resistant
and susceptible seedlings. J. Am. Soc. Hort. Sci. 109:428-432.
Korban, S. S.9 P. A. OConnor, and A. Elobeidy. 1992. Effects of thidiazuron, napthaleneacetic acid, dark incubation and genotype on shoot organogenesis from Malus
leaves. J. Hort. Sci. 76:341-349.
Krger, J. 1994. Observations on different mildew sources used in apple breeding at
Ahrensberg. p. 7-12. In: H. Schmidt and M. Kellerbals (ads.), Progress in temperate
fruit breeding. Kluwer Academic, Dordrecht, Netherlands.

72

APPLES

Lacey, C. N. D., and A. I. Cambell. 1987. Selection, stability and propagation of mutant
apples. p. 351-361. In: A. J. Abbott and R. K. Atkin (eds.). Improving vegetatively
propagated crops. Academic, New York.
Lambert, C., and D. Tepfer. 1992. Use of Agrobacterium rhizogenes to create transgenic
apple trees having an altered organogenic response to hormones. Theor. Appl. Genet.
85:105-109.
Landry, B. S., R. Q. Li, W. Y. Cheung, and R. L. Granger. 1994. Phylogeny analysis of
25 apple rootstocks using RAPD markers and tactical gene tagging. Theor. Appl.
Genet. 89:847-852.
Lapins, K. 1965. Compact mutants of apple induced by ionizing radiation. Can. J. Plant
Sci. 45:117-124.
Lapins, K. 1969. Segregation of compact growth types in certain apple seedling progenies.
Can. J. Plant Sci. 49:765-768.
Lapins, K. O. 1974. Spur type growth habit in 60 apple progenies. J. Am. Soc. Hort. Sci.
99:568-572.
Lapins, K. O. and R. Watkins. 1973. Genetics of compact growth. p. 136. In: Annu.
Rep. E. Malling Res. Sta., 1972.
Laubscher, F. X., and Hurter, N. 1960. The chromosome numbers of seedling progeny
from triploid apples. S. African. J. Sci. 3:31-39.
Lawson, D. M., M. Hemmat, and N. F. Weeden. 1995. The use of molecular markers to
analyze the inheritance of morphological and developmental traits in apple. J. Am.
Soc. Hort. Sci. 120:532-537.
Lespinasse, Y., and E. Chevreau. 1997. Progressi nel miglioramento genetico del melo:
cloni autocompatibili e piante aploidi. Riv. Fruttic. Ortofloric. 69:25-29.
Lespinasse, Y., F. H. Alston, and R. Watkins. 1976. Cytological techniques for use in
apple breeding. Ann. Appl. Biol. 82:349-353.
Lewis, D.. and M. B. Crane. 1938. Genetical studies in apples, II. J. Genet. 37:119-128.
Lockard, R. G., and G. W. Schneider. 1981. Stock and scion growth relationships and
the dwarfing mechanism in apple. Hort. Rev. 3:315-375.
Luckwill, L. C. 1953. Virus diseases of fruit trees 4. Further observations on rubbery
wood, chat fruit and mosaic in apples. p. 40-46. In: Rep. Long Ashton Res. Sta.
MacHardy, W. E. 1996. Apple scab: biology, epidemiology and management. APS, St.
Paul, MN.
Maheswaran, G, M. Welander, J. F. Hutchinson, M. W. Graham, and D. Richards. 1992.
Transformation of apple rootstock M.26 with Agrobacterium tumefaciens. J. Plant
Physiol. 139:560-568.
Manganaris, A. G. 1989. Isozymes as genetic markers in apple breeding. Thesis, Wye
College, University of London.
Manganaris, A. G., and F. H. Alston. 1988. The acid phosphatase gene, ACP-1 and its
linkage with the endopeptidase ENP-1 and pale green lethal gene I in apple. Acta
Hort. 224:177-184.
Manganaris, A. G., and F. H. Alston. 1989. Glutamate oxaloacetate transaminase isoenzymes in apple cultivars and rootstocks. J. Hort. Sci. 64:9-15.
Manganaris, A. G., F. H. Alston, N. F. Weeden, H.S. Aldwinckle, H. L. Gustafson, and
S. K. Brown. 1994. Isozyme locus Pgm-1 is tightly linked to a gene (Vf) for resistance
in apple. J. Am. Soc. Hort. Sci. 119:1286-1288.

LITERATURE CITED

73

Marks, G. E. 1954. An acetocarmine-glycerol jelly for use in pollen fertility counts.

Stain Technol. 29:277.
Markussen, T., J. Krger, H. Schmidt, and F. Dunemann. 1995. Progress on mapping a
gene for powdery mildew resistance in apple. p. 6. In: European Apple, Issue No. 3.
Marlow, G. C. and W. H. Loescher. 1994. Watercore. Plant Breeding Rev. 6:189-252
Martin, G. B. 1996. Molecular in cloning of plant disease resistance genes. p. 1-32. In:
G. Stacey and N. Keen (eds.). Plant-microbe interactions, Vol 1. Chapman-Hall,
New York.
Mauren M. M. F., R. M. Manshardt, D. Gonsalves, J. L. Slightom, and J. C. Sanford.
1992. Virus resistant papaya plants derived from tissues bombarded with the coat
protein gene of papaya ringspot virus. Bio/Technology 10: 1466-1472.
McIntosh, D. L., and Lapins, K. 1966. Differences in susceptibility to apple powdery
mildew observed in McIntosh clones after exposure to ionizing radiation. Can. J.
Plant Sci. 46:619-623.
McIntosh, D. L., and F. C. Mellor. 1954. Crown rot of fruit trees in British Columbia,
III: resistance trials on apple seedlings obtained from controlled crosses. Can. J.
Agric. Sci. 34:539-541.
Menendez, R. A., F. E. Larsen, and R. Fritts. 1986. Fingerprinting apple cultivars by
electrophoretic isozyme banding patterns. J. Environ. Hort. 4:101-107.
Michelmore, R. W. 1995a. Molecular manipulation of disease resistance genes. Annu.
Rev. Phytopathol. 33:393-427.
Michelmore, R. W. 1995b. Isolation of disease resistance genes from crop plants. Current
Opinion Biotechnol. 6:145-152.
Michelmore. R. W., I. Paran, and R. V. Kesseli. 1991. Identification of markers linked to
disease resistance genes by bulked segregant analysis: a rapid method to detect
markers in specific genome regions using segregating populations. Proc. Natl. Acad.
Sci. (USA) 88:9828-9832.
Michelmore, R. W., R. V. Kesseli, D. M. Francis, I. Paran, M. G. Fortin, and C. H. Yang.
1992. Strategies for cloning plant disease resistance genes. p. 99-124. In: S. Guff,
McPherson, and Bowles (eds.). Molecular plant pathology: a practical approach,
Vol 2. IRL, Oxford, UK.
Mink, G. I., and J. R. Shay. 1962. Latent viruses in apple. Purdue Univ. Agric. Exp. Sta.
Res. Bull. 756.
Mircetich, J. S. M., and G. T. Browne. 1989. Phytophthora root and crown rot of
deciduous fruit trees: The influence of different rootstocks and soil moisture
conditions on severity of the disease. Compact Fruit Tree 22:115-123.
Misic, P., and Z. Tesovic. 1970. Distribution of anthocyanin in the skin of apple varieties
Jonathan, Red Delicious and their dark red mutants, 1. Sci. Agr. Res. Boegrad. 23:121126.
Moore, M. H. 1960. Apple rootstocks susceptible to scab, mildew and canker for use in
glasshouse and field experiments. Plant Path. 9:84-87.
Moore, R. C. 1940. A study of inheritance of susceptibility and resistance to apple cedar
rust. Proc. Am. Soc. Hort. Sci. 37:242-244.
Morgan, J. and A. Richards. 1993. The book of apples. Ebury Press, London.
Mowry, J. B. 1964. Inheritance of susceptibility to Gymnosporangium
juniperi-virginianae. Phytopathology 54:1363-1366.
Mowry, J. B., and D. F. Dayton. 1964. Inheritance of susceptibility to apple blotch. J.
Hered. 55:129-132.

74

APPLES

Murawski, H. 1955. Investigations of stages of apple seedlings as a basis for fruit research
(in German). Arch. Gartenb. 3:255-273.
Murawski, H. 1967. A contribution to apple cultivation research, 10: results of cultivation
of apple species with late leafing-out blooming (in German). Zchter 3:134-139.
Nonnecke, I. 1948. Fireblight resistance across the prairies in Malus. p. 31-32. In: Proc.
W. Can. Soc. Hort. Sci.
Norelli, J. L., H. S. Aldwinckle, L. D. Beltran, and J. M. Jaynes. 1994. Transgenic
Malling 26 [= M.7] apple expressing the attacin E gene has increased resistance
to Erwinia amylovora. Euphytica 77:123-128.
Nybom, N. 1959. On the inheritance of acidity in cultivated apples. Hereditas 45:332-350.
Opara, L. U., C. J. Studman, and N. H. Banks. 1996. Fruit skin splitting and cracking.
Hort. Rev. 18:(in press)
Oppenheimer, C., and E. Slor, 1968. Breeding of apples for a sub-tropical climate, 2.
Theor. Appl. Genet. 38:97-102.
ORourke, A. D. 1994. The world apple market. Food Product Press, New York.
Parisi, L., Y. Lespinasse, J. Guillaumes, and J. Kilger. 1993. A new race of Venturia
inaequalis virulent to apples with resistance due to the Vf gene. Phytopathology
83:533-537.
Patat-Ochatt, E. M., S. J. Ochatt, and J. B. Power. 1988. Plant regeneration from
protoplasts of apple rootstocks and scion varieties (Malus xdomestica Borkh.). J.
Plant Physiol. 133:460-463.
Pratt, C. 1983. Somatic selection and chimeras. p. 172-185. In: J. N. Moore and J.
Janick (eds.). Methods in fruit breeding. Purdue University Press, West Lafayette, IN.
Quamme, H. A., and C. Stushnoff. 1983. Resistance to environmental strew. p. 242-266.
In: J. Janick and J. N. Moore (eds.). Methods in fruit breeding. Purdue University
Press, West Lafayette, IN.
Redalen, G. 1988. Quality assessment of apple cultivars and selections. Acta Hort.
224:441-445.
Sadamori, S., T. Haniuda, S. Tsuchiya, and Y. Yoshida. 1964. (Studies on aberrant leaf
in apple, 1: observations on the frequency of aberrant leaf in apples and their
seedlings.) p. 1-7. In: Bull. Hort. Res. Sta. Morioka, Japan, Ser. C, No 2.
Sakai, A. 1994. Cryopreservation of apical meristems. Hort. Rev. 6:357-372.
Samimy, C., and J. N. Cummins. 1992. Distinguishing apple rootstocks by isozyme
patterns. HortScience 27:829-831.
Sampson, D. R., and R. F. Cameron. 1965. Inheritance of bronze foliage, extra petals
and pendulous habit in ornamental crab apples. Proc. Am. Soc. Hort. Sci. 96:717-722.
Save1jev, N. 1. 1989. Breeding material and genetic principles of breeding apple trees
for frost and cold resistance. Acta Hort. 224:165-168.
Sax, K. 193 1. The origin and relationship of the Pomoideae. J. Arnold Arbor. 12:3-22.
Sax, K. 1959. The cytogenetics of facultative apomixis in Maims species. J. Arnold
Arbor. 40:289-297.
Schmidt, H. 1964. A contribution to the cultivation of apomictic apple rootstocks, 1:
cytogenetic and embryologic research (in German). Z. Pflanzenz. 52:27-102.
Schmidt, H. 1970. A contribution to the cultivation of apomictic apple rootstocks, 3:
their affinity with cultured species in relation to virus content (in German). Z.
Pflanzenz. 63:214-226.
Schmidt, H. 1988. Criteria and procedures for evaluating apomictic rootstocks for apple.
HortScience 23:104-107.

LITERATURE CITED

75

Schmidt, H. 1994. Progress in combining mildew resistance from Malus robusta and
Malus zumi with fruit quality. p. 3-6. In: H. Schmidt and M. Kellerhals (eds.). Progress
in temperate fruit breeding. Kluwer Academic, Dordrecht, Netherlands.
Sewell, G. W. F., and Wilson, J. F. 1959. Resistance trials of some apple rootstock varieties
to Phytophthora cactorum (L. and C.) Schroet. J. Hort. Sci. 34:51-58.
Shay, J. R., and L. F. Hough. 1952a. Evaluation of apple scab resistance in selections in
Malus. Am. J. Bot. 39:289--297.
Shay, J. R., and L. F. Hough. 1952b. Inheritance of cedar rust resistance in apple. Phytopathology 42:19.
Shay, J. R., D. F. Dayton, and L. F. Hough. 1953. Apple scab resistance from a number
of Malus species. Proc. Am. Soc. Hort. Sci. 62:348-356.
Shay, J. R., E. B. Williams, and J. Janick. 1962. Disease resistance in apple and pear.
Proc. Am. Soc. Hort. Sci. 80:97-104.
Skirvin, R. M., M. Kouider, H. Joung, and S.S. Korban. 1986. Apple (Malus domestica
Borkh.). p. 183-198. In: Y. P. S. Bajaj (ed.). Biotechnology in agriculture and forestry,
1: trees 1. Springer, Berlin.
Spinks, G. T. 1936. Apple breeding investigations, 1: results obtained from certain families
of seedlings. p. 19-49. In: Rep. Long Ashton Res. Sta. 1936.
Sriskandarajah, S., P. B. Goodwin, and J. Speirs. 1994. Genetic transformation of the
apple scion cultivar Delicious via Agrobacterium tumefaciens. Plant Cell Tissue
Organ Cult. 36:317-329.
Tancred, S. J., A. G. Zeppa, M. Cooper, and J. K. Stringer. 1995. Heritability and patterns
of inheritance of the ripening date in apples. HortScience 30:325-328.
Tartarini, S. 1996. RAPD markers linked to the Vf gene for scab resistance in apple.
Theor. Appl. Genet. (in press)
Thiele, I. 1950. Fruitfulness of maiden pome trees as a growers project (in German).
Zchter 26:241-243.
Thompson, A. H., and Batjer, L. P. 1950. The effect of boron in the germinating medium
on pollen germination and pollen tube growth for several deciduous tree fruits. Proc.
Am. Soc. Hort. Sci. 56:227-230.
Thompson, J. M., and 1. Taylor. 1971. Genetic susceptibility to Glomerella leaf blotch
in apple. J. Hered. 62:303-306.
Tobutt, K. R. 1985. Breeding columnar apples at East Malling. Acta Hort. 159:63-68.
Tobutt, K. R. 1994. Combining apetalous parthenocarpy with columnar growth habit in
apple. Euphytica 77:51-54.
Tukey, H. B. 1964. Dwarfed fruit trees. Cornell University Press, Ithaca, NY.
Tydeman, H. M. 1964. The relation between time of leaf break and of flowering in apple
seedlings. p. 70-72. In: Rep. E. Malling Res. Sta., 1963.
Tydeman, H. M., and Alston, F. H. 1965. The influence of dwarfing stocks in shortening
the juvenile phase of apple seedlings. In: p. 97-98. Rep. E. Malling Res. Sta., 1964.
Vail, P. V., J. S. Tebbets, D. F. Hoffmann, and A. M. Dandekar. 1991. Response of production and postharvest walnut pests to Bacillus thuringiensis insecticidal crystal protein
fragments. Biol. Control 1:329-333.
Vavilov, N. I. 1930. Wild progenitors of the fruit trees of Turkistan and the Caucasus
and the problems of the origin of fruit trees. p. 271-286. In: Rep. 9th Int. Hort.
Congr. Royal Hort. Soc., London.
Vinterhalter, D. V., and D. T. James. 1983. The use of peroxidase polymorphism in the
identification of apple scion cultivars. Sci. Hort. 18:253-261.

76

APPLES

Vinterhalter, D. V., and D. T. James. 1986. The use of peroxidase polymorphism in the
identification of Malling and Malling Merton apple rootstocks. J. Hort. Sci. 110:509515.
Visser, T. 1951. Floral biology and crossing technique in apples and pears. Meded. Dir.
Tuinb. 14:707-726.
Visser, T. 1964. Juvenile phase and growth of apple and pear seedlings. Euphytica 13:119129.
Visser, T. 1970. Environmental and genetic factors influencing the juvenile period in
apple. p. 101-115. In: Proc. Angers Fruit Breeding. Symp.
Visser, T., J. J. Verhaegh, and D. P. de Vries. 1971. Preselection of compact mutants
induced by X-ray treatment in apple and pear. Euphytica 20:153-207.
Wallin, A., and L. Johansson. 1989. Plant regeneration from leaf mesophyll protoplasts
of in vitro cultured shoots of a columnar apple. J. Plant Physiol. 135:565-570.
Watkins, R., and R. A. Smith (eds.). 1982. Descriptor list for apple (Malus). Int. Board
Plant Genetic Resources, Rome.
Watkins, R., and L. P. S. Spangelo. 1970. Components of genetic variance for plant
survival and vigor of apple trees. Theor. Appl. Genet. 40:195-203.
Watkins, R., and J. M. Werts. 1971. Preselection for Phytophthora cactorum (Leb. and
Conn.) Schroet. resistance in apple seedlings. Ann. Appl. Biol. 67:153-156.
Way, R. D. 1971. Hastening the fruiting of apple seedlings. J. Am. Soc. Hort. Sci. 96:384389.
Way. R. D., H. S. Aldwinckle, R. C. Lamb, A. Rejman, S. Sansavini, T. Shen, R. Watkins,
M. M. Westwood, and Y. Yoshida. 1990. Apples (Malus). p.1-62. In: J. N. Moore and
J. R. Ballington Jr. (eds.). Genetic resources of temperate fruit and nut. Int. Soc.
Hort. Sci., Wageningen (Acta Hort. 290).
Weeden, N. F. 1989. Applications of isozymes; in plant breeding. Plant Breeding Rev.
6:1154.
Weeden, N. F., and R. C. Lamb. 1985. Identification of apple cultivars by isozyme phenotypes. J. Am. Soc. Hort. Sci. 110:509-515.
Weeden, N. F., and R. C. Lamb. 1987. Genetics and linkage analysis of 19 isozyme loci
in apple. J. Am. Soc. Hort, Sci. 112:865-872.
Weeden, N. F., G. M Timmerman, M. Hemmat, B. E. Kneen, and M. A. Lodhi. 1992.
inheritance and reliability of RAPD markers. p. 12-17. In: Application of RAPD
technology to plant breeding. Joint Plant Breeding Symp. Series. Crop Sci. Soc
Am./Am. Soc. Hort. Sci./Am. Genet. Assoc.
Weeden, N. F., M. Hemmat, D. M. Lawson, M. Lodhi, R. L. Bell, A. G. Manganaris, B.
Reish, S. K. Brown, and G. Ye. 1994. Development and application of the molecular
marker linkage maps in woody crops. p. 269-273. In: H. Schmidt and M. Kellerhals
(eds.). Progress in temperate fruit breeding. Kluwer Academic, Dordrecht, Netherlands.
Welander, M. 1988. Plant regeneration from leaf and stem segments of shoots raised in
vitro from mature apple trees. J. Plant Physiol. 132:738-744.
Welander, M., and G. Maheswaran. 1991. Regeneration and transformation experiments
in apple. p. 237-246. In: M. R. Ahuja (ed.). Woody plant biotechnology. Plenum,
New York.
Weller, D. L., and J. F. Costante. 1986. Peroxidase zymograms of 16 apple rootstocks.
Can. J. Plant Sci. 66:647-652.

LITERATURE CITED

77

Welsh, J., and M. McClelland. 1990. Fingerprinting genomes using PCR with arbitrary
primers. Nucleic Acids Res. 19:303-306.
Wilcox, A. N., and E. Angelo. 1936. Apple breeding studies, 1: fruit color. Proc. Am.
Soc. Hort. Sci. 33:108-113.
Wilcox, A. N., and E. Angelo. 1937. Apple breeding studies, 2: fruit shape. Proc. Am.
Soc. Hort. Sci. 34:9-12.
Williams, E. B., and J. Kuc. 1969. Resistance in Malus to Venturia inaequalis. Annu.
Rev. Phytopathol. 7:223-246.
Williams, J. G. K., A. R, Kubelik, K. J. Livak, and J. A. Rafalski. 1990. DNA polymorphisms amplified as arbitrary primers are useful genetic markers. Nucleic Acids Res.
18:6531-6535.
Williams, W. 1958. Dept. Plant Breeding. p. 6-13. In: Annu. Rep. John Innes Hort. Inst.
1957.
Williams, W. 1959. Selection of parents and family, size in the breeding of top fruits. p.
211-213. In: Rep. 2nd Congr. Eucarpia.
Williams, W., and Brown, A. G. 1956. Genetic response to selection in cultivated plants:
gene frequencies in Prunus avium. Heredity 10:237-245.
Wormald, H. 1955. Diseases of fruit and hops. Crosby Lockwood, London.
Yahia, E. M. 1994. Apple flavor. Hort. Rev. 16:197-234.
Yang, H., and J. Kroger. 1994. Identification of an RAPD marker linked to the Vf gene
for scab resistance in apples. Plant Breeding 112:323-329.
Yao, J. L., D. Cohen, R. Atkinson, K. Richardson, and B. Morris. 1995. Regeneration of
transgenic plants from the commercial apple cultivar Royal Gala. Plant Cell Rep.
14:407-412.
Zhang, Y. X., and Y. Lespinasse. 1988. Culture in vitro dovules non fecondes dembryons
preleve 8 jours apres pollinisation chez le pommier cultive (Malus xdomestica Borkh.).
Agronomie 8:837-842.
Zhang, Y. X., Y. Lespinasse, and E. Chevreau. 1988a. Obtention de plantes haploides de
pommier (Malus xdomestica Borkh.) issues de parthenogenese induite in situ par du
pollen irradie et culture in vitro des pepins immature. C. R. Acad. Sci. Paris 307:451457.
Zhang, Y. X., Y. Lespinasse, and G. Salesses. 1988b. Mise en evidence de quelques
anomalies meiotiques conduisant a la formation de gametes males non reduits chez
le pommier cultive (Malus xdomestica Borkh.). Cytologia 53:749-755.
Zhang, Y. X., G. Salesses, and Y. Lespinasse. 1988c. Etude cylogenetique de quelques
clones diploides et polyploides de pommier ( Malus xdomestica Borkh.).
Considerations quant a lorigine du pommier. Cytologia 53:739-748.
Zimmerman, R. H. 1971. Flowering in crab apple seedlings: methods of shortening the
juvenile phase. J. Am. Soc. Hort. Sci. 96:404 411.
Zimmerman, R. H. 1984. Apple. p. 369-395. In: W. R. Sharp, D. A. Evans, P. V. Ammirato,
and Y. Yamada (eds.). Handbook of plant cell culture, Vol. 2. crop species. Macmillian,
New York.
Zimmerman, R. H. 1991. Micropropagation of temperate zone fruit and nut crops. p.
231-246. In: P. C. DeBergh and R. H. Zimmerman (eds.). Micropropagation. Kluwer,
Netherlands.
Zimmerman, R. H., and 1. Fordham. 1985. Simplified method for rooting apple cultivars
in vitro. J. Am. Soc. Hort. Sci. 110:34-38.