A, Origin and Speciation

Blackberries are fruiting plants of the genus Rubus subgenus Rubus (formerly subgenus
Eubatus), Rosaceae (USDA 2006). This genus is characterized by considerable diversity, ranging
from tiny, prostrate plants to very large bushes over 5 m tall. Growth is usually herbaceous or
semi-woody and biennial, with a perennial crown and root system. Some species have annual
fruiting canes, and others have woody perennial growth as upright bushes or climbing lianes
(Allan 1928; Darrow 1937, 1967; Jennings 1988). Species may be rigidly erect, semi-erect
(arching), or trailing, and typically form bushes with new canes overgrowing those from the
previous year. Some members of the genus grow from a single crown and others produce root
suckers to form new crowns, or even grow new aerial tissues from, an underground rhizome each
year (Jennings 1988)().

Rubus fruit consist of a number of fleshy drupelets, each containing a single seed (pyrene)
around the central torus or receptacle. By horticultural definition, in blackberries the torus is
picked with the fruit, maintaining fruit integrity, whereas in raspberries (subgenus Idaeobatus),
the torus remains on the plants and fruit integrity is maintained by the presence of small hairs
holding the fruit together after removal.
„This horticultural differentiation between blackberry and raspberry species on the basis of the
above fruit characters provides taxonomic delineation between the two subgenera, Rubus
(formerly Eubatus) with many hundreds of defined species and the Idaeobatus with several
hundred species as outlined by Focke in his Species Ruborum (1910- 1914) and Sudre in his
RubiEuropae (1908-1913). Further subdivision in the genus Rubus by Focke gave a total of 12
subgenera with between 1 and 19 species, except for the Maiachobatus subgenera with 114
species. While there is some consistency between morphological characters within these
subgenera, these groupings may be somewhat arbitrary, as some species within subgenera are
widely separated geographically (Focke 1910—1914). While the division of the Rubus genus
into two major groups, blackberries and raspberries, is a valuable horticultural distinction,
molecular techniques have shown that it is taxo- nomically artificial. Numerous blackberry
species have characteristics in common with those of raspberry, especially in upright types.
Genes from raspberry are in their origin and indeed there is a probably continuous movement of
genes from raspberry to blackberry in nature (Alice 2002). The online database, Germplasm
Resources Information Network (USDA 2006) describes 13 subgenera for the genus Rubus, and
1
12 sections within Rubus subgenus Rubus.
In breeding new blackberry types, red raspberry (Rubus idaeus L.) has been important in the
development of ‘Brazos’ (Darrow 1967), and the wide range of blackberry types produced in the
University of Arkansas breeding program (Moore et al. 1977; Moore and Clark 1989a,b; Clark
and Moore 1999a,b). This is due to the fact that the raspberry x blackberry hybrid ‘Nessberry’
(Yarnell 1936) is in the background of ‘Brazos’. ‘Loganberry’ and ‘Phenomenal’ (Jennings
1988) are also “hybridberries” produced from raspberry x blackberry crosses and from
backcrossing these to blackberry, the cultivars Youngberry (Darrow 1925), presumably
Boysenberry (Wood et al. 1999), Olallie (Waldo 1968), Marion (Finn et al. 1997) and almost all
of the blackberry types produced in the United States Department of Agriculture-Agricultural
Research Service (USDA-ARS) Corvallis, Oregon breeding program were released. Other hybrid
berries include ‘Tayberry’ (Moore 1997b) and ‘Tummelberry’ (Jennings 1983) from the Scottish
Crop Research Institute and ‘Sun- berry’ from the U.K. (Keep et al. 1982). In the Horticulture
ap'd Food Research Institute of New Zealand (HortResearch) program-; selections from
Arkansas, Oregon, and Scotland, each with raspberries in their i background, have given rise to
the wide range of selections available today (Hall 1990, 1992; Hall et al. 2002, 2003).

Some early generation blackberry-raspberry hybrids did not pick easily either with or without
the receptacle, but it has been possible to select for blackberry type abscission in later
generations. Recommendations have been made for use of species from other subgenera such as
R. roseus Poir. or R. nubi- genus Kunth [R. macrocarpus Benth.) of the subgenus Lampohatus
(Popenoe 1920; Darrow 1937; Waldo 1950a; Darrow 1955a, 1967; Waldo 1968; Ourecky 1975;
USDA-ARS 2005), but there are no published records of success with such crosses.
Cultivated blackberries derived from sections Moriferi and Ursini of the Rubus subgenus, as well
as the hybrids between members of these sections and with members of other subgenera (the
hybridberries), are all considered in this review under the title “blackberry.”

In the treatment of the Moriferi, Sudre (1908-1913) separated the Moriferi into five
subsections, the Appendiculati, Caesii, Discolores, Silvatici, and Suberecti, This treatment is
useful in giving a descriptive index of the European blackberries, but does not recognize the two
main diversities within the group: the true blackberries and R. caesius L. and hybrids. This
problem was addressed by Gustafsson (1943), who divided the European blackberries into two

2
larger groups, the Moriferi veri with 11 subsections and R. caesius with all of its related types.
.). Rubus armeniacus is so common that most people presume it is native; rather it has
displaced the native R. ursinus in disturbed environments and is considered a noxious weed by
farmers, landowners, foresters, and by state and federal agencies. Rubus armeniacus has been
used in breeding and its offspring ‘Chehalem* is one of the parents of ‘Marion’.
Although much work has been done in Rubus taxonomy, especially in North America and
Europe, the true relationships among blackberry species or all Rubus species and groups of
species remain to be elucidated. In his 1990 review, Hall suggested that future studies of specia-
tion and taxonomy could examine the expression of genetic variability from a clone and the
mode of inheritance of traits through studies of breeding populations as described by Brainerd
and Peitersen (1920) and Gustafsson (1943). These techniques would be a definitive way of
determining species relationships in the wide range of blackberry ecotypes available. Much of
the diversity in the eastern North American and European blackberries may be represented by
only a few major genes (Brainerd and Peitersen 1920; Gustafsson 1943). An analysis of these
genes will be of value to taxonomists and breeders (Hall 1990), and further molecular work will
likely provide many answers on taxonomic relationships.

Table 2.1 Blackberry cultivars including information on parentage, ploidy level, presence or
absence of thorns, growth habit, location, year of release or publication, and reference.

Cultiv Parentage;
ar Information2; Origin

Chehalem Santiam x Himalaya; 6x; Y; L; Ore. 1948

Kotata ORUS743 x ORUS877; 7x; Y; L; Ore. 1984
Lincoln Logan Genetically thornless Logan; 6x; T(LL); L; N.Z.
Logan Aughinbaugh
1986 x Red Antwerp; 6x; Y; L; Calif.
Marion Chehalem x Olallie; 6x; Y; L; Ore. 1956
1881
Olallie Black Logan x Young; 6x; Y; L; Ore. 1950

3
Santiam Unknown; 9x-2; Y; L;
Tayberry Aurora x Complex; 6x+2; Y; L; Scotland 1979

In the 1880s, Judge James H. Logan released the first cultivars from a concerted breeding
effort (Logan 1955). He worked extensively with .selections derived from the native R. ursinus,
most notably ‘Aughinbaugh’. Red raspberry (‘Red Antwerp’) and eastern blackberry cultivars
(“Texas Early’/’Crandall’s Early’) were planted in proximity to the pistillate ‘Aughinbaugh’, and
the open-pollinated fruit of ‘Aughinbaugh’ were harvested and planted as seedlings. Serendipity
took hold, as the results were not what Judge Logan anticipated (Oregon Horticulture Society,
undated but possibly in the early 1920s): “It has been repeatedly stated in public prints that I
entertained the idea when I planted those seeds of a cross between the raspberry and blackberry.
I am sorry to disturb one of the supposed truths of history, but candor compels me to say that
such is not the case. I did not even deem such a cross possible, and did not know what I had done
until May, 1883, when the plant first fruited‘Loganberry’ and ‘Black Logan’ were the most
important of Logan’s cultivars and these went on to be important parents in the development of
trailing blackberry cultivars in the western U.S. Logan’s effort also encouraged Luther Burbank,
who released ‘Phenomenal’ in 1905 (Jennings 1981a). ‘Phenomenal’ was also called ‘Burbank’s
Logan’ and was considered nearly indistinguishable from ‘Loganberry’ (Dar- row 1918),
although some claimed it was more disease susceptible than ‘Loganberry’ (Jennings 1981a).
Byrnes M. Young, a businessman in Morgan City, Louisiana, who loved science and plants
(Christy 2004), had no success growing either ‘Loganberry’ or ‘Phenomenal’ (Darrow 1925). He
did however cross ‘Phenomenal’ with the better adapted ‘Austin Mayes’ to produce
‘Youngberry’ in 1905, which was then released in 1926 (Darrow 1937). Another important
cultivar of uncertain origin with a colorful history from this era is ‘Boysenberry’, often termed a
“hybridberry.” While there is a great deal of speculation and discussion about the origin
of‘Boysenberry’ (Knott 1935; Darrow 1937;
Stellar 1937; Thompson 1961; Jennings 1981a, 1988), Wood et al. (1999)

The thornless chimera of ‘Evergreen’ that was called Thornless Evergreen’ was found in
Stayton, Oregon in 1926. Although Thornless Evergreen’ commonly reverts to being thorny, it
was reliably very productive and became the mainstay of the industry until being largely

4
displaced by ‘Marion’ in the 1970s (Finn et al. 1997). McPheeters and Skirvin (1995,
The success of cultivars selected from the wild or from the various early private breeding efforts
provided the basis for the establishment of public breeding programs (Hall 1990). Darrow (1937)
cited the first public blackberry breeding program as the one begun in 1908 by the Texas
Agricultural Experiment Station, College Station. At this station, H. Ness, S. H. Yarnell, and H.
F. Morris successively developed “hybrid- berries” adapted to warm climates with low chilling
such as ‘Nessberry’ (a parent of the still-popular ‘Brazos’), which combined the trailing
blackberry R. trivialis with red raspberry (Darrow 1937)

. Despite a number of programs listed as existing in the early 1900s, Darrow (1937) noted that
“compared with other berries, little systematic breeding has been done with the blackberries.”

In the western U.S., George M. Darrow (1937) spent a few years in Oregon in the late 1920s and
early 1930s helping to firmly establish a breeding program there. Darrow and G.F. Waldo
swapped positions in 1932, with Darrow moving to Maryland and Waldo back to his native
Oregon. Waldo generated a breeding pool using primarily 'Loganberry’, ‘Youngberry’,
‘Mammoth’, ‘Himalaya’, and wild selections of R. ursinus (e.g., ‘Zielinski’), including perfect-
flowered cultivars (e.g., ‘Santiam’/ ’Ideal’), in which the maternal parent was R. ursinus and the
paternal parent was unknown but suspected to be ‘Loganberry’ (Waldo 1968). This germplasm
quickly generated the releases ‘Pacific’ (Waldo and Wiegand 1942), ‘Cascade’ (Waldo and
Wiegand 1942), ‘Chehalem’ (Waldo 1948), and ‘Olallie’ (Waldo 1950b). The next generation of
crosses produced ‘Marion’ (Waldo 1957), which is still the major culti- var in the Northwest. In
the late 1960s, Francis J. “Whitey” Lawrence took over the program from Waldo. He evaluated
some of the materials that Waldo started, for example ‘Kotata’ (Lawrence 1984),

‘Black Diamond’ has completed a circle of germplasm exchange between the USDA-ARS in
Oregon and HortResearch in New Zealand and highlights what is possible if germplasm
exchange is encouraged and not restricted. In 1980, a Boysenberry/blackberry/hybridberry breed-
ing program was initiated in Lincoln, New Zealand by Harvey K. Hall and soon thereafter
moved to Riwaka (HortResearch, Nelson Region). A germplasm pool was assembled that
initially consisted of all types of commercially available cultivars but was later supplemented
with selections from the Scottish Crop Research Institute (SCRI) and the USDA- ARS (Oregon).
Internally developed selections from early generation crosses and the diverse genetic background

5
that became available through the development of ‘Lincoln Logan’ also provided germplasm
(Hall et al. 1986c). This pool has been responsible for some noteworthy releases such as ‘Ranui’,
‘Waimate’, ‘Karaka Black’, and ‘Marahau’ (Hall and Stephens 1999; Clark and Finn 2002; Hall
et al. 2003). In 1991, a cross was made between ‘Kotata’ and NZ 8610L163, and the seed was
shared with the breeding program in Oregon. From this cross, which mixed the Oregon and New
Zealand germplasm, ‘Black Diamond’ was selected and in 2004-2005 more plants of it were sold
in thedNTorth- western U.S. than those of any other cultivar (P. Moore,-pers. com- mun.). The
germplasm pool developed in New Zealand represents a strong mix of the trailing and
erect/semi-erect germplasm pools, with a diverse range of ‘Lincoln Logan’ thornless selections
having been produced at nearly all ploidy levels from 3x to ~12x (Hall and Stephens 1999). The
New Zealand program has been able to incorporate genes from R. glaucus into new blackberry
types and has developed a wide range of hexaploid blackberries similar to ‘Marion’, ‘Karaka
Black’, and ‘Loganberry’, as well as some outstanding new types at the octoploid breeding level
and new 12x R. ursinus derivatives that are thornless and hermaphroditic. In recent years, this
program has focused on breeding new Boysenberry types, with considerable success being
achieved in producing thornless types with excellent fruit set, high yield, large fruit size,
improved fruit firmness, and suitability for fresh market sales (H.K. Hall, pers. observ.).
In a smaller breeding effort in the United Kingdom, Derek Jennings developed ‘Tayberry’,
‘Tummelberry’, and ‘Loch Ness’ during his time with the Scottish Crop Research Institute and
later ‘Adrienne’ and ‘Helen’ with private breeding programs (Jennings 1988,1989; Daubeny
1995; Clark and Finn 1999). ‘Tayberry’ is a hybrid between ‘Aurora’ blackberry and a tetraploid
raspberry selection, while ‘Tummelberry’ is a hybrid between ‘Tayberry’ and 69102/18, a sibling
of‘Tayberry’ (Jennings 1988). ‘Tayberry’, released in 1979, has gone on to develop a very
unique niche in the U.S. and U.K. processing markets, despite requiring hand harvest. Jennings
successfully combined the blackberry and raspberry germplasm. Crosses made in Scotland were
grown out in New Zealand to yield ‘Kaiteri’ and in Australia to yield ‘Murrindindi’ (Daubeny
1995). The SCRI program has been continued by Nikki Jennings and Rex Brennan and ‘Loch
Tay’ was released in 2004 (Kerby and Snape 2004).
There are a number of private programs smaller than Driscoll’s, The program of Harry
Swartz’s Five Aces Breeding, associated with the University of Maryland, has released
‘Chesapeake’ (Swartz et al. 2003) and the “hybridberry” ‘Anastasia’ (H,J. Swartz, pers.

6
commim.), and the program of J.W. Clark has released ‘Clark Gold’ (Clark 2004),
B. World Industry
The blackberry industry continues to expand around the world, and an estimated 20,035 ha are
planted and cultivated (wild plants harvested not included) (Strik et al. 2006). While it is hard
to point to one predominant reason for the continual expansion, some of the reasons include:
(1) blackberries are a new crop to many areas of the world, (2) they share many similarities to
red raspberry and as raspberry production develops in an area, blackberries often follow, (3)
blackberries are less expensive to produce than red raspberries because they do not have
to be replanted as often and commonly do not have as many pest control inputs, (4) improved
cultivars have been developed that ship better, extend the harvest season, taste better, are
thornless, etc,, and (5) the growing awareness of the nutraceutical value of blackberries; other
than black raspberries, blackberries have higher levels of anthocyanins than many other widely
available fruits.

1. North America

Western U.S. Two very large but somewhat dichotomous industries exist in the Pacific
Northwest and California. Oregon’s Willamette Valley is one of the leading production areas in
the world, with about 3100 ha of blackberries (OASS 2005). Over 2200 ha of this production are
in the processing cultivars ‘Marion’ and ‘Thornless Evergreen’ and about 400 ha are planted to
‘Boysenberry’ and ‘Loganberry’ (OASS 2005). The remaining hectares axe in newer thornless
processing cultivars such as ‘Black Diamond’ or in cultivars that are harvested for the fresh
market such as ‘Cherokee’, ‘Navaho’, ‘Chester Thornless’, ‘Siskiyou’, and ‘Obsidian’ (C.E.
Finn, pers. observ,). While there is a growing interest in blackberry production in Washington
(U.S.) and British Columbia (Canada), this is still a fairly small industry primarily focused on the
fresh market.
The California blackberry industry is mainly for fresh market. For many years, ‘Olallie’ was a
major cultivar in California and while it has largely disappeared from the wholesale marketplace,
there are many farms that depend on it for local, fresh-market sales.

However, the 2002 Census of Agriculture reported about 210-280 ha of blackberry production
(NASS 2004; Strik et al. 2006) in California. While California historically had a very large
processing industry based on ‘Boysenberry’, that industry has shrunk to 40-85 ha.

7
. The current production area is estimated to be about 2300-3000 ha (Strik et al. 2006, J. Lopez-
Medina, pers. commun.).
Canada. Blackberry production is a very minor industry in Canada, with scattered plantings
along the U.S. border in British Columbia, around the Great Lakes in Ontario, and in the
Maritime Provinces. While some of the production in British Columbia is for the wholesale fresh
market, most plantings are small and are for pick-your-own and roadside markets.

Australasia. Production of blackberries in this area is largely limited to Australia and New
Zealand and is primarily for fresh market production, with the exception of the New Zealand
‘Boysenberry’ industry (H.K. Hall, pers. observ.). In New Zealand, there are about 260 ha of
‘Boysenberry’ producing 3350 tonnes (t) annually (Strik et al. 2006).

In Spain and Italy, as in California, with the maturation of the red raspberry industry, an interest
in diversification into fresh market blackberries has developed.

South Africa, South Africa has limited blackberry production with total area estimated to be 140
ha. ‘Youngberry’ is grown for processing into jam and juice and makes up the majority of the
production, while the remainder is for local and export fresh production (Strik et al, 2006).

South America. South American production of cultivated blackberries is around 1600 ha.

II. BREEDING TECHNOLOGY

A. Floral Biology
Blackberry species usually have biennial canes that require a dormant period before fruiting.

There is a wide range of floral primordial development dates, ranging from ‘Boysenberry’ and
‘Marion’ buds in Oregon showing sepal primordial development in November and December
with further floral bud development continuing in January to April or May in later cultivars and
cooler locations. (Takeda et al. 2003).

Blackberry flowers have multiple ovaries, styles, and stigmas attached to a central receptacle,
which mature to form an aggregate fruit. The receptacle is surrounded by numerous stamens,
then petals and sepals (Darrow 1937). Flowers are usually borne on axillary racemes or panicles
forming fruiting laterals that develop on the biennial canes. Numbers of flowers per lateral vary
from one to many according to the species or cultivar (Darrow 1937).
In erect types such as ‘Darrow’, flowers are tightly clustered on short laterals (Slate 1958;

8
Slate and Watson 1963), but in trailing types such as ‘Boysenberry’ or ‘Flordagrand’ the flowers
are more widely separated on long hanging laterals (Creswell 1938; Shoemaker et al. 1958;
Shoemaker and Westgate 1966),
Most cultivated blackberries are hermaphroditic and self-fertile. However, wild species maybe
dioecious, as in the Ursini (Darrow 1937; Waldo 1950a), self-sterile as in erect species from the
eastern U.S. and trailing species from the southern U.S. (Darrow 1937; Sherman and Sharpe
1971). In Oregon, the dioecious habit of the native R. ursinus has been overcome with the
introduction of several hermaphroditic cultivars with complex origins (Waldo 1950a; Darrow
1967; Waldo 1968; Lawrence 1986b).

Infertility or partial infertility may also occur due to genetic factors (Darrow and Longley 1933;
Darrow 1937; Waldo 1950a; Darrow 1967), the effects of environment (Einset 1947; Slate and
Klein 1951; Haskell 1962, 1966), lack of pollinators, lack of attractiveness to polliiiators (C.E.
Finn, pers. observ.), or poor pollen production (Heilman and Clark 1984). In some seedlings of
R, ursinus hybrids, fruits have been found to set drupelets only near the base of the receptacle
(Hall 1990). This has evidently been due to the extreme length of the receptacle combined with
short stamens, or to the later development and receptivity of stigmas and ovaries on the end of
the receptacle. In each of the above types, daily hand pollinations while flowers were receptive
produced perfect fruits (Hall 1990).

Even when apparently perfect fruits are set, it is common for blackberries to have much less than
100% of the ovules develop into drupelets (Strik et al. 1996b). As in raspberries (Jennings 1975),
this may arise from low growth substance status, poor development of endosperm, and embryo
abortion, as well as the factors inducing sterility discussed above. Sometimes embryo
development may stop, but the development of drupelets continues and complete fruits may
result.

.). Seedless drupelets were also found when growth regulators such as 4-chlorophenoxy acetiq
acid (4CPA) were applied to flowers at the time of pollination, a technique commercially used on
‘Boysenberry’ in New Zealand (Kingston 1986; Kingston 1987].

9
Hybridization

When flowers have reached a stage where they are plump and anthers are turgid with mature but
undehisced pollen, they may be emasculated with a razor blade, scalpel, or sharp forceps,
removing the sepals, petals, and stamens (Harrow 1937; Ourecky 1975). Various techniques can
be used to emasculate flowers, but a balance must be struck between doing them quickly and
limiting the amount of damage to the flower. In New Zealand, a small, sharp-pointed forceps is
the preferred tool, as it is much quicker and can be easily used by unskilled hands. In Oregon,
where the approach to emasculation is to cut around the pedicel at the base of the flower, a
single-edge razor blade is the preferred tool because it is very quick, causes minimal flower
damage, and is an approach that can be taught quickly to new workers.

A single pollination 2-3 days after emasculation may be sufficient to achieve fruit set, but
multiple pollinations may produce more seed, especially when crosses among less closely related
genotypes are performed (Plate 2.1). Routinely, blackberry flowers are not covered with bags or
other material following emasculation, since insects do not usually visit emasculated flowers
(Finn 1996). In climates where it commonly rains or there are heavy dews during the flowering
season, bags are desirable because they allow one to get back into the field and work with dry
flowers quickly. (Plate 2.1).

Most blackberry crossing is done in the field. However, at some locations such as in the maritime
climates of Scotland (Jennings 1984) and New Zealand (Hall 1990), and in the continental
climate of Illinois (Hull 1961a, 1968), crosses are better done in
a greenhouse to enable synchronization of flowering times, the use of non-adapted parents, and
to achieve better fruit set, seed production, and germination. In addition, the subsequent harvest
of fruit can be completed before the fruiting season in the field commences (H.K. Hall, pers.
observ.). Pollen may be obtained from bagged flowers using the flower itself as a brush (Harrow
1937) or it may be extracted easily from flowers at the fully turgid stage just prior to opening,
using forceps, and stored for weeks (Perry and Moore 1985a) or many years before use (Ourecky
1975).

10
Pollen should be stored in a desiccator in a refrigerator when not in use. When using pollen in
the field, it should be carried in a cooler to prevent it from being damaged by high temperatures
and exposed to light.

In most cases, crosses between different blackberries produce hybrids, even when parents are
aneuploid or at different ploidy levels. Occasionally, crosses between genotypes derived from the
Ursini subsection of Rubus give rise to maternal types, presumably by apomixis. This is much
more common when blackberries from the European Rubus subsection are used as the female
parent.

Developed fruit, but not necessarily fully ripe fruit, can be harvested and seed extracted using a
variable speed blender with padded blades and liberal quantities of water. After blending and one
or more cycles of settling and decanting, clean seeds are obtained for drying and subsequent
germination (Morrow et al. 1954).

More recently, pectinase, an enzyme used in food processing, has been used to remove the
flesh from the seeds simply and without the potentially damaging process of using a blender. The
fruit are mashed with a few drops of pectinase and enough water to make a thick slurry. The fruit
are left at room temperature for 2-5 days or overnight at 37°C and then the flesh is decanted off
of the fruit as described above with the viable seeds sinking to the bottom. This is particularly
helpful when a firm-fruited female parent is used for crossing or where there is strong pulp
adherence to seed, such as with ‘Boysenberry’,

C. Seed Treatment and Germination

One of the significant challenges in blackberry breeding is attaining sufficient germination of
hybrid seeds. The most important treatments of seeds include scarification and subsequent
stratification.
Scarification of fully dry seeds (pyrenes) with concentrated sulfuric acid is the most common
method of treatment. Drying is usually accomplished by placing the seeds on a paper towel in an
ambient interior environment for 14 days or longer. Dry seed is essential before scarification, as
the heat generated by placing seed that is not dry into sulfuric acid will kill the embryo (Ourecky
1975).

11
With segregation of seedlings for the ‘Lincoln Logan’ gene for thornlessness, plants must be
grown to the third or fourth true leaf, 2-4 cm in height, before thorny seedlings can be removed.

. In New Zealand, 2500 seedlings of the first populations to explore the diversity of
hybridberries generated by crossing Aurora (8x+2) with ‘Cherokee’ or ‘Comanche’ upright
blackberries (4x) yielded only three or four upright plants, none of which was fully fertile. How-
ever, intercrossing and further breeding with these selections has enabled development of
populations with upright growth habit and almost full fertility (H.K. Hall, pers. observ.).

D. Selection Evaluation
After establishment of selections, evaluation for a range of characters is conducted. The major
evaluation period is during fruiting, with supplemental evaluations at other times of the year for
characteristics such as bloom date, winter hardiness, and foliar or cane disease incidence and
severity. Primary evaluation during fruiting commonly includes subjective data such as ratings
for maturity date, fruit size, shape, attractiveness, color, drupelet fertility, firmness/skin
toughness, ease of picking, freedom from disease, and flavor, along with plant characters such as
health and cane erectness (for erect types only), and lateral strength and length (particularly for
selections to be machine harvested).

While this can serve to help legally protect a patented cultivar, it more critically can be used to
help ensure correct cultivar identity throughout the propagation stream. This is especially critical
now that in vitro propagation is becoming the main method of commercial propagation and one
mistakenly collected meristem can, in a few months, be the source of thousands or tens of
thousands of incorrectly identified plants. In New Zealand, there were at least four cases of
somatic mutations, two of which were not recognized before being released to growers. This
caused substantial losses to growers in time and replanting costs (H.K. Hall, pers. observ.).
The majority of blackberry cultivars released today are protected, either by plant patents (U.S.),
plant breeders rights (most other countries), or trademarks. Exceptions to this approach are some
public programs such as the USDA-ARS, which continues to release cultivars without
propagation restrictions (as of 2006),

12
68 Program support from royalties is a primary incentive for protecting cultivars.

Most blackberries have not been labeled by cultivar name in the fruit marketplace, giving
consumers little ability to differentiate between them, in contrast to crops such as apples and
pears. Cultivars labeling has been practiced in the U.K., where poor-quality fruit of some
cultivars has damaged customer acceptance of blackberries (H.K. Hall, pers. observ.). Excep-
tions to this are ‘Marion’, ‘Olallie’, ‘Loganberry’, ‘Boysenberry’, and ‘Chehalem’, which have
been widely marketed with these cultivar names in processed products in the U.S., ‘Silvan’ in
Australia, and ‘Loch Ness’ and ‘Tayberry’ in the U.K. for the fresh market. If name recognition
is given to more fresh-market cultivars, marketing by more unique methods such as specialty
sales or as “club” cultivars might be considered.

E. Propagation
Asexual propagation ,of blackberries is not difficult, but certain guidelines must be followed to
avoid diseases and to maintain true-to-type plants. In many cases, the use of micropropagation is
preferred because of the ease of plant multiplication and the maintenance of virus-free stock.
However, in breeding program management, the initial propagation of selections is usually by
root or stem cuttings or tip layering. Tissue culture is more common when commercial quantities
of plants are needed.

The maintenance of virus-free plant material is paramount for propagating clean stock. This can
be challenging at times in breeding program management. Initially, a true-to-type plant is
virus'tested to ensure freedom from known viruses. If a plant is found to contain viruses, it is
typically put through in vitro heat therapy (thermotherapy) following established protocols
(Converse 1987), then kept in an isolated area to prevent infection from disease vectors.

Once freed of viruses, genotypes are kept as whole plants in insect-proof screen- houses and
grown in a sterile medium. In the U.S., state regulation of nurseries may require nurseries to re-
establish clean in-vitro stock from clean stock plants in a screenhouse every few years. State
regulations may also dictate requirements for establishment of propagation fields, which may
include being a certain distance from possible virus infection sources and soil fumigation
protocols (Crandall 1995). Hardwood cuttings are made in late fall and placed in soilless media
in flats with bottom heat, in an unheated greenhouse, with mist. Tip layers can be an easy source

13
of propagules, but they often cannot be counted on to provide a sufficient number of plants. They
are dug only when they clearly come from the genotype of interest and typically if other methods
of propagation have failed. As with trailing blackberries, erect blackberries are often microprop-
agated or propagated by more conventional methods such as root cuttings by commercial
nurseries. In a breeding program, the main method of propagation is by root cuttings (Caldwell
1984).

While tip layering, direct digging of root suckers, and crown division are also viable options,
they do not yield as many plants as do root cuttings. There are several ways to establish plants
from root cuttings, including direct field planting and forced adventitious shoot propagation
(Thompson et al. 2004). The adventitious shoot propagation method can yield numerous
propagules from one root, though the amount can vary based on cultivar and propagation media.
Rooting, survivability, and vigor are excellent with this technique (Thompson et al. 2004). Other
methods of propagating erect blackberries include softwood (Moore and Skirvin 1990) and
hardwood cuttings (Bray et al. 2004).

One- to two-node softwood stem cuttings with two or three leaves are typical and these root
readily when placed in perlite under mist (Moore and Skirvin 1990). Zimmerman et al. (1980)
propagated four genotypes of thornless blackberries using one-node cuttings. They discovered
that cane node location did not influence rooting, but rooting differed with genotype. Busby and
Himelrick (1999) found that rooting was affected by cultivar, IBA concentration, and method of
hormone application for four blackberry cultivars that were propagated by two-node cuttings.

1. Micropropagation. Tissue-culture propagation of blackberries is,an efficient way to rapidly

increase plant numbers. This technique is especially useful when material for cuttings is lacking
or the selection is not virus-free. Micropropagation of blackberry typically involves sequential
shoot tip proliferation via axillary budbreak. Shoots can be rooted on root-induction media.
Shoot tip culture is the most efficient method for rapid proliferation and avoids somaclonal
variation (Pelto and Clark 2000a). Broome and Zimmerman (1978) micropropagated 'Black
Satin’, ‘Dirksen Thornless’, ‘Smoothstem’, and selection US 64-39-2 and observed variation
among genotypes for explant proliferation and response to proliferation medium. Reed (1990)
evaluated 256 Rubus accessions for in-vitro regeneration and proliferation and found that 69% of
blackberry genotypes proliferated successfully on Murashige and Skoog (MS) (1962) basal salts.

14
Tissue culture of meristems or axillary buds can also be done, but these techniques have
drawbacks that make shoot tip culture more popular for propagation. Meristem culturing is often
essential in conjunction with thermotherapy, to eliminate viruses from the propagation stock
(Converse 1987).
A number of studies have attempted to improve in-vitro procedures to avoid somaclonal
variation (Harper 1978; Pyott and Converse 1981; Hu and Wang 1983; McPheeters et al. 1988;
Skirvin et al. 1994), to improve efficiency (Slivinski et al. 1984; Fernandez and Clark 1991;
Bobrowski et al. 1996; Pelto and Clark 2000b; Erig et al. 2002; Radmann et al. 2003), to
evaluate various explant material (Harper 1978), or to generate somatic embryogenesis (Cantoni
et al. 1993).

Regeneration protocols are also essential for genetic transformation. Graham and McNicol
(1990) reported that internodal stem segments had a higher success rate of regeneration than did
leaf discs for red raspberries ‘Autumn Bliss’ and SCRI selection 824E6, red raspberry x
blackberry hybrids ‘Tayberry’, ‘Tummelberry’, and ‘Sunberry’, and the blackberry ‘Loch Ness’.
Meng et al. (2004) developed a procedure to enhance regeneration of leaf and petiole explants of
‘Marion’.
Rosati (1986) found that tissue culturing of the periclinal chimera of ‘Loganberry’ was
effective in breaking down the L1-L3 layer stability within the plant, resulting in the
production of several plants that produced thornless seedlings from open-pollinated
seedling populations. Similar instability of LI chimeral thornless mutants of ‘Boysenberry’ has
been observed in New Zealand, resulting in both thorny types (L2-L3 displacing Ll) and sterile
types where the Ll tissues have displaced the L2 and L3 (H.K. Hall, pers, observ.).

Cultivar identification in Rubus using analytical techniques was first initiated when Haskell and
Garrie (1996) used paper chromatography to separate the flavanoid components of red raspberry
leaves.

Another study by Nybom (1966) used thin-layer chromatography on red raspberries and found
that it was successful for cultivar identification and determination of anthocyanin pigment

15
inheritance. Isozyme staining has been a popular technique for the identification of Rubus
genotypes. This is largely due to reports that culture environment has no effect on the isozyme
patterns of the genotypes.

Lopez-Medina et al. (1996) found that by using isozymes and protein markers, raspberry-
blackberry hybrids could be identified.

Minisatellite. Minisatellite DNA probes are useful in DNA fingerprinting of Rubus. The first
report of this technique utilizing Rubus was made by Nybom et al. (1989), who determined that it
was a useful tool to identify red raspberry and blackberry cultivars. Nybom and Schaal (1990)
reported minisatellite DNA probes to be appropriate for the understanding of-population
structure and species differentiation. A wider study among Rosaceous plants detected genetic
variation in Rubus as well as Malus and Primus (Nybom et al. 1990). Considerable inter- and
intraspecific variation existed among blackberries and raspberries, but there were also some
cases of identical “fingerprints.” Nybom and Hall (1991) later confirmed earlier results that
minisatellite DNA fingerprints were adequate for distinguishing cultivars and estimating their
genetic relatedness.

Randomly Amplified Polymorphic DNA. The search for a cheaper, faster alternative that uses
smaller amounts of DNA than RFLP led to the introduction of RAPD, a PCR-based analysis.
Several studies using RAPD markers on Rubus have been reported. The first study to utilize the
new analysis in Rubus was Parent și Pagé (1992).They found stability in markers even though
they used plant material from different propagation methods, seasons, and sources. Graham and
McNicol (1995) stated that RAPD marker data can identify cultivars, species, and interspecific
crosses. Another study with RAPD markers (Pamfil et al. 2000) on breeding anomalies among
interspecific crosses within a subgenus led to the conclusion that RAPD markers may not be
optimal for generating genetic similarity studies at the species level and caution should be used
with interpretation of genetic similarity at the interspecific level.

Internal Transcribed Spacer Region. Other DNA-based techniques developed in recent years are
being used to determine genetic variation within Rubus. A study utilized ITS of nuclear
ribosomal DNA to document natural hybridization in Rubus and confirm parental ploidy levels
(Alice et al. 1997).

16
These discoveries may later play a role in decisionmaking for marker-assisted selection
programs. Studies on Rubus involving nuclear ribosomal DNA ITS region sequences have
focused entirely at the species level (Alice and Campbell 1999; Alice et al. 200i; Alice 2002).

Blackberry Breeding with Molecular Markers. The use of molecular markers for blackberry
breeding has been limited, likely due to the small economic impact of blackberries worldwide
and the general challenges with using markers in polyploid fruit crops (Luby and Shaw 2001). ).
However, a growing global interest in blackberry production has fueled an increase in the area of
molecular marker analysis. Many of the molecular marker studies involving blackberry species
and cultivars have, focused on genetic similarity, phylogenetics, and taxonomic issues. The
prospect of molecular breeding in Rubus is in the early stages of development.

Multiple efforts also are underway to obtain blackberry genomic (N.V. Bassil, pers. commun.)
and EST sequences for SSR development. Molecular marker linkage to important traits such as
thornlessness and primocane fruiting have been attempted as well, but better primers would
likely improve results in the future.
Genetic Mapping. No genetic mapping has been done for blackberries, but some preliminary
research with SSR markers found primers that would be useful (Lewers et al. 2004; Stafne et al.
2005). More recently, SSR markers have been developed for Rubus. Graham et al. (2002)
developed the first SSR markers for Rubus from ‘Glen Moy’ red raspberry and tested the
markers on blackberries and blackberry x raspberry hybrids. They found that all 10 fluorescently
labeled primer pairs developed amplified polymorphisms, thus making them useful for further
molecular analysis such as genetic mapping.

Hybrid Confirmation. Waugh et al. (1990a) described how the understanding of genetic
relationships using molecular methods may be important for the future of Rubus breeding.
Chloroplast and nuclear DNA of several blackberry, blackberry x raspberry hybrids, and rasp-
berry genotypes were examined with RFLPs as a preliminary step to quantify variation. They
concluded that molecular analysis would increase the efficiency and precision in Rubus breeding.

A potentially important aspect of breeding is the molecular confirmation of hybrid genotypes.

Nybom (1995) evaluated interspecific crosses with facultatively apomictic blackberries with
restriction enzymes and minisatellite probes to assess apomictic and sexual seed set within the

17
population and to determine paternity. Antonius and Nybom (1995) described the evidence of
automixis in an interspecific Rubus breeding program with RFLP markers. They described how
the ability to determine plants derived through automixis would be useful because of their
homozygosity. Naess et al. (1998) used RAPD markers to verify hybrids from a study that
utilized irradiated pollen.

Lopez-Medina et al. (1996) utilized sodium dodecyl sulphate-polyacrylamide gel electrophoresis

(SDS-PAGE) to identify raspberry-blackberry hybrids. Lim et al. (1998) used GISH and FISH to
verify the presence of blackberry and raspberry chromosomes in blackberry (‘Aurora’) x
raspberry hybrids.

Taxonomy. Taxonomic relationships within Rubus subgenus Rubus are often complicated
because of widespread cross-fertility among species.

Recently, several molecular studies have been done to solve taxonomic questions that have
plagued botanists for many years. One of the first studies in this area was by Waugh et al.
(1990a, 1990b), who examined variability in 20 Rubus genotypes using heterologous cpDNA
sequences from barley and pea as probes and RFLP

Since the cpDNA displays infrequent mutation and a slow rate of sequence evolution, it is a good
candidate for analyzing evolutionary and taxonomic species history (Antonius-Klemola 1999).

Pamfil et al. (2000) produced a more inclusive analysis with 40 Rubus species. Using RAPD
markers, they found that R. loganobaccus and R. ursinus had a high degree of similarity
(>90%). The two species both belong to the Ursini section and are sometimes
undifferentiated.

Alice et al. (1997, 2001) also tested R. caesius and hybrids of R. caesius and R. idaeus with ITS
sequences. They determined that gene flow occurs in nature between R. caesius and R. idaeus
based on the evidence of detectable polymorphisms at most sites where the two species differ.
Alice (2002) found that, based on ITS sequence analysis, R. ursinus may be a hybrid with R.
macraei (.Idaeobatus) as its pollen parent.

Probably the most comprehensive study done on Rubus taxonomy was carried out by Alice
and Campbell (1999) with ITS sequences, which included 20 species of blackberries. They found

18
that blackberries were not monophyletic because of discrepancies with R. cuneifolius and its
closer relation to South American species than to European species. Ultimately, they concluded
that traditional morphological characteristics used for taxonomic identity such as leaf type and
thorn appearance have limited value for phylogenetic analysis in Rubus.
Antonius and Nybom (1995) first reported molecular evidence of automixis in raspberry x
blackberry hybrids.

Recombinant DNA and Gene Transfer. The use of recombinant DNA technology in Rubus has
been increasing, but most of the innovation has occurred with red raspberry. Some of the first
studies were done by McNicol and Graham (1989; 1992) and Graham and McNicol (1990), who
successfully transferred the cowpea trypsin inhibitor gene into Rubus (red raspberry, blackberry,
and raspberry-blackberry hybrids) through the use of Agrobacterium tumefaciens and binary
vector PBI121.X, which transferred the beta-glucuronidase (GUS) marker. The GUS marker was
used to confirm transfer of the gene into the transformed plants. Mathews et al. (1995)
transferred the S-adenosylmethionine hydrolase gene into raspberry, which can reduce the
plant’s ethylene production. Mathews in collaboration with R. Martin with the USDA-ARS
transferred the virus coat protein and movement genes to ‘Meeker’ red raspberry to successfully
impart resistance to Raspberry bushy dwarf virus (RBDV) (Martin, pers. commun.);

Protoplast Fusion. Phan et al. (1997) and Mezzetti et al. (1999) developed efficient protocols for
protoplast isolation and plant regeneration of red raspberry (‘Autumn Bliss’) and blackberry
(‘Hull Thornless’) cul- tivars. Mezzetti et al. (1999) fused the protoplast suspensions of the two
cultivars with polyethylene glycol (PEG) treatment. The resulting hybrid callus lines were
structurally differentiated for root formation and pre-embryonic growth.

III. BREEDING SYSTEMS

A. Genetic Structure and Polyploidy
The naturally occurring range of chromosome numbers in Rubus is from 2n=2x=14, the diploid
state, to 2n=14x=98 or possibly 2n=18x=126, including odd-ploids and aneuploids (Moore 1984;
Thompson 1995b; Meng and Finn 1999). The chromosomes are small (1-3 pm in length) and the
nuclear DNA content in diploid species ranges from 0.56 to 0.59 pg (Lim et al. 1998; Meng and
Finn 2002).
Genotypes representing many of the different Rubus ploidy levels have been used in breeding

19
and released as cultivars. While/there are commercial cultivars at nearly every ploidy level, some
generalizations can be made. The trailing southern dewberries (e.g., ‘Oklawaha’) are diploid.
The erect (e.g., ‘Navaho’) and semi-erect (e.g., ‘Chester Thornless’) cultivars and derivatives of
R. laciniatus (e.g., ‘Everthornless’) are all tetraploid. The trailing cultivars are hexaploid or
higher (e.g., ‘Mar- ion’=6x, ‘Siskiyou’=7x, ‘Douglass’=8x, ‘Black PearT=9x, ‘Tillamook’ =10x,
and ‘Zielinski’“12x) with several cultivars being fully fertile with less than full sets of
chromosomes (e.g., ‘Tayberry’=6x+2, ‘Aurora’ =8x+2, ‘Santiam’=8x+5 or 9x-2) (Meng and
Finn 2002).
While fertility is often poor if the hybrid is triploid, pentaploid, or an aneuploid with a
chromosome number less than hexaploid (Waldo 1950a; Lawrence 1986b), there are many
exceptions. Fertile triploid cultivars exist (Thompson 1995b). In the USDA- ARS Corvallis,
Oregon program, a number of fertile genotypes have been identified that are 5x, 6x, or 7x from
crosses made between ‘Navaho’, which is 4x, and ‘Kotata’, which is 7x, or ORUS 1122-1 or
‘Black Butte’, which are 6x (Meng and Finn 2002).

D. Interspecific Hybridization /

Interspecific hybridization has been the most widely^used source of variability for blackberry
improvement (Lawrence 1986b; Jennings 1988) and new species continue to be used (Finn et al.
1999b; Finn 2001; Finn et al. 2002). In cultivar improvement of erect blackberries primarily
derived from R. allegheniensis, other species have been the source of increased fruit size (Moore
1984). Blackberry spfecies are somewhat poorly defined; with apomictic clones having wide
distribution in Europe, and large numbers of natural hybrids in North America (Ourecky 1975;
Lawrence 1986b). In spite of the difficulty in identifying the species of superior blackberry
clones, or assigning them to a particular species combination, plants have been chosen and used
in crossing to produce further generations of variability (Darrow 1937).

Attempts to utilize blackberry species from the Moriferi and the Ursini, as well as raspberry
species and cultivars, for blackberry improvement have been successful (Darrow 1937; Waldo
and Darrow 1948; Waldo 1950a; Darrow 1967; Waldo 1968; Lawrence 1986b; Finn 2001),
While some attempts to use other Rubus species for cultivar improvement have failed (Darrow
1967; Sherman and Sharpe 1971; Jennings 1988) success has been achieved using R. glaucus in

20
New Zealand (H.K. Hall, pers. observ.).
Barriers affecting crossability, germination of seed, or fertility of progenies have been
encountered in crosses between closely or distantly related clones at similar or different ploidy
levels (Ourecky 1975; Lawrence 1986b). Diploid and triploid blackberry-raspberry hybrids are
often difficult to obtain due to poor seed set and germination (Einset and Pratt 1954; Darrow
1955a; Britton and Hull 1959) and the seedlings often die before reaching maturity or are
frequently infertile (Jennings 1988).
Great difficulties were experienced in intercrossing diploid blackberries and raspberries. In
Great Britain, attempts to cross R. ulmifolius Schott, with red raspberry failed (Crane et al. 1949),
while in the United States a cross between R. trivialis (=R. rubrisetus Rydb.) and ‘Brilliant’ red
raspberry produced 21 almost sterile hybrids (Ness 1919).
Crosses of tetraploid blackberries x diploid raspberries are easier to perform than diploid x
diploid crosses, but the triploid seedlings are infertile (Rietsema 1955; Jennings 1988). Hybrids
of tetraploid raspberries x tetraploid blackberries are easier to produce than diploid or triploid
crosses, provided that the raspberry is used as the seed parent (Einset and Pratt 1953; Jennings
1988), However, no seedlings have proved useful; their fruit are purple-black in color, difficult
to remove from the plant, and are usually astringent (Jennings 1988). In spite of these
difficulties, ‘Nessberry’ and ‘Veitchberry’ were used for further breeding (Crane 1936; Jennings
1988), ‘Nessberry’ was used in a cross with the tetraploid ‘Lawton’ blackberry, yielding
‘Brazos’ in the F2 generation (Darrow 1967; Jennings 1988).
). When the parents of a cross are even-ploid and the progeny are odd-ploid (e.g., 6xx 8x - 7x),
fertility is often low (Waldo 1950a). However, fruitful seedlings from such progenies can be
obtained in large populations, Toungberry’, the Oregon selection OSC 73 (‘Loganberry’ x
‘Austin Thornless’), and presumably ‘Boysenberry’ have been selected from such progenies
(Waldo 1950a; Britton and Hull 1956; Thompson 1961). Hybrids between hexa- ploid or
heptaploid and tetraploid parents have also produced some fertile progeny, but this cross was less
successful than those between higher ploidy types (Waldo 1950a; Meng and Finn 2002).
In some cases, fertile seedlings have been at the intermediate euploid level. ‘Kotata* (7x)
resulted from a cross between OSC 743 (8x+3) and OSC 877 (6x) (Fischer et al. 1941; Zielinski
and Galey 1951; Thompson 1961; Darrow 1967; Jennings 1988; Thompson 1995a).

21
). When the chromosome number of both parents and the progeny of a cross is even-ploid, the
fertility of seedlings is usually greater, especially when the hybrid contains pairs of like
genomes.

‘Loganberry’ was derived from a similar cross, and resulted from ‘Aughinbaugh’ pollinated
by an unreduced diploid pollen grain from ‘Red Antwerp’ red raspberry, giving the chromosome
constitution of V1V1V2V2ii, where i represents the raspberry genome (Thomas 1940). When
‘Mammoth’ and ‘Loganberry’ were intercrossed, the seedlings were sterile due to incom-
patibility between the ‘Crandall’ blackberry genome bj, and the raspberry genome i (Darrow and
Longley 1933; Thomas 1940). Similar crosses in New Zealand between ‘Loganberry’ and
‘Marion’ and ‘Loganberry’ and other 6x hybrid blackberries also had a high degree of sterility,
but a few selections were produced and these subsequently have been crossed back to 6x hybrid
blackberries giving fully fertile progenies (Hall and Brewer 1989; Hall and Stephens 1999; Hall
et al. 2002).
A number of attempts were made to produce a new ‘Loganberry’, resulting in a number of
‘Loganberry’-like seedlings in Oregon (Waldo and Darrow 1948; Waldo 1950a; Darrow 1967)
and eastern U.S., upright ‘Loganberry’-types in Illinois (Hull 1961a, 1968), although no cultivars
were released from this effort. Further attempts at producing a new ‘Loganberry’ have resulted in
the release of ‘Tayberry’ and ‘Tummel- berry’ from Scotland (Jennings 1980,1981b, 1983) and
‘Sunberry’ from England (Keep et al. 1982). ‘Sunberry’ produces fruits similar in size to
‘Loganberry’, but they are more rounded and a deeper color, and the plant is very vigorous
(Jennings 1988). ‘Tayberry’ and ‘Tummelberry’, a seedling of ‘Tayberry’ crossed with a full
sibling, benefited from the improved nature of the blackberry parent ‘Aurora’ and the raspberry
parent 626/67. Fruit size in both cultivars is considerably larger than ‘Loganberry’, although the
fruit are slightly more difficult to pick. Fruit is well presented on short laterals, and the plants are
productive (Jennings 1988).
Hybrids between hexaploids such as ‘Loganberry’ or ‘Marion’ and diploid raspberries are
largely infertile, and often the fertile segregants are difficult to harvest. However, such a cross
produced selection 21/27, one of the progenitors of selection 626/67, the tetraploid raspberry
parent of ‘Tayberry’ (Jennings 1988; D. L. Jennings, pers. commun.)

22
Crosses between the 8x ‘Aughinbaugh’ and the diploid raspberries ‘Red Antwerp' and ‘Cuthbert’
produced progeny from unreduced gametes from the raspberry parents (Jennings 1988).

Chromosome number in hybrids may also vary in less obvious ways. Chromosome loss in
somatically unstable plants is common, especially where autopolyploids have been produced by
treatment with colchicine (Britton and Hull 1956, 1957; Thompson 1962; Hull 1968). ‘Silvan’
blackberry, a hybrid of OSC 742 (8x+3) and ‘Marion’ (6x), is 6x (McGregor and Kroon 1984;
Thompson 1995a; Meng and Finn 2002), yet ‘Kotata’, from a similar background, is 7x
(Thompson 1995a). Similarly, a single population from the cross of 7x ‘Kotata’ with 4x
‘Navaho’ generated fertile selections that are 5x, 6x, and 7x (Meng and Finn 2002).

Blackberries are generally considered quite drought tolerant compared to raspberries and
breeding specifically for increased drought tolerance is not currently underway, as most
programs utilize irrigation as a component of culture.

Thornless Evergreen’ (Moore and Skirvin 1990) and ‘Silvan1 (McGregor and Kroon 1984) have
been reported to have drought tolerance. There is also a need to select for adaptation to
conditions of low atmospheric humidity in blackberries if large fruit are to be produced in arid
conditions. Most cultivars that produce large, attractive fruit under humid conditions show
reduced fruit size in low humidity conditions. This is clearly seen with ‘Boysenberry1, with very
large fruit in Nelson, New Zealand, medium-large fruit in Oregon, smaller fruit in the Central
Valley of California, and very small fruit from the semi-desert conditions of Chile (H.K, Hall,
pers. observ.).

Wind damage susceptibility has been observed in ‘Loganberry1 and ‘Olallie’ and their
offspring (H.K. Hall, pers. observ.). ‘Silvan’ was reported to have tolerance to wind (McGregor
and Kroon 1984),
. Likewise, ‘Marion1, ‘Silvan’, ‘Kotata’, ‘Waldo’, ‘BlackDiamond’, ‘Black Pearl’, and
‘Nightfall’ are well adapted to machine harvest production, as are nearly all USDA-ARS
selections for the processing market (C.E. Finn, pers. observ.). ‘Youngberry’ and ‘Boysenberry’
are also grown for machine harvest production in New Zealand, but care must be taken with

23
these cultivars not to over-fertilize with nitrogen early in the growing season or fruiting laterals
will be too long and harvest difficult.

Blackberries are remarkably free of serious disease and insect pest problems in much of their
range. In commercial settings where the crops are grown intensively in a monoculture, fungicide
applications are a standard management practice. Insecticides/acaricides are often applied for
specific problems such as raspberry crown borer [Pennisetia marginata (Harris)], red-necked
caneborer [Agrilus ruficollis (Fabricius)], redberry mite (Acalitus essigi Hassan), strawberry
weevil {Anthonomus signatus Say), brown and green stink bugs (Euschistus spp. and
Acrosternum hilare Say, respectively), Japanese beetle (Popillia japonica, Newman), thrips
(eastern and western flower thrips, Frankliniella tritici Fitch and F. occidentalis Pergande,
respectively), grass grub (Costelytra zealandia White), and foliar nematode [Aphelenchoides
ritzemabosi (Schwartz) Steiner], or as a “knockdown” to remove insects such as orange tortrix
[Argyrotaenia citrana Fernald) that can be a contaminant in machine- harvested fruit. However,
very often there is no standard insecticide program (Ellis et al. 1991). A few production regions
have very severe pests that require regular chemical spray programs. A good example is in New
Zealand where ‘Boysenberry’, ‘Marion’, and all other Rubus are attacked severely by raspberry
bud moth (Heterocrossa rubophaga Dug- dale) and/or blackberry bud moth (Eutorna
phaulacosma Meyrick). The leaf roller species, including Epiphyas postivittana Walker,
Planotor- trix exessana Walker, P. octo Dugdale, Ctenopseustis obliquana Walker,
Unlike raspberries, root rot [Phytophthora fragariae Hickman var. rubi Wilcox & Duncan) is
not generally a serious problem in blackberry; however, it can be a problem in some of the
raspberry-blackberry hybrids like ‘Loganberry’ (McKeen 1958).

Plant Habit

Blackberries are usually classified with three cane types: trailing, semi-erect, and erect (Strik
1992).

Trailing types are crown-forming and grow at or near ground level, and the canes must be
bundled and tied to a trellis. The greatest concentration of this type of plant is in Oregon and
includes cultivars such as ‘Marion*

24
Semi-erect plants also are crown-forming and require a trellis, with the mature canes growing
upward approximately 1 m before arching over to a horizontal orientation. Major semi-erect
cultivars are ‘Chester Thornless’, ‘Loch Ness’, and ‘Triple Crown’.

Erect-caned blackberries often sucker beneath the soil line and grow in a more upward direction.
In Arkansas and other areas of the U.S.,, the canes are usually very erect. However, support
wires on either side of the cane row are usually utilized to prevent canes from falling over in
wind or other disturbances. In Oregon, erect cultivars behave more similarly to semi-erects in
that they grow upward 1—2 m and are tied to a trellis. Canes in Oregon are usually fewer in
number and attain a greater height if not tipped (C.E. Finn, pers. observ.). Erect and semi-erect
cultivars respond positively to tipping of the canes, while trailing cultivars do not, and this
is reflected in how primocanes are managed. Erect cultivars include ‘Chickasaw', ‘Choctaw’,
‘Shawnee’, ‘Navaho’, ‘Arapaho’, and ‘Ouachita’.

The trailing habit offers some advantages for machine harvesting and winter hardiness. New
primocanes may be trained along the row on the ground below the catcher plates on a machine
harvester. Canes also may be removed when alternate year cropping is practiced or the growing
season is long enough to grow canes after harvest. With the primocanes out of the way on the
ground as opposed to being upright as they are in erect or semi-erect types, the berries fall
through the floricane vegetation more easily as the machine passes. Bell et al. (1995a) found that
leaving canes on the ground over the winter decreased the incidence of winter injury; however,
yield was reduced. Growers in Oregon take a calculated risk each August; if they train the canes
onto the trellis, they are at greater risk for winter injury but will get higher yields as opposed to
less chance of injury and lower yields with late-winter training.

Breeding erect thornless plants proved to be much more challenging. The recessive thornless
gene chosen for use in breeding was derived originally from ‘Merton Thornless’, an English
cultivar (Jennings 1988). The sources of this trait used in the Arkansas program were the semi-
erect, thornless, tetraploids ‘Thomfree’ and ‘Smoothstem’ and similar selections from the
USDA-ARS breeding program based at Carbondale, Illinois.

The trailing habit offers some advantages for machine harvesting and winter hardiness. New
primocanes may be trained along the row on the ground below the catcher plates on a machine

25
harvester. Canes also may be removed when alternate year cropping is practiced or the growing
season is long enough to grow canes after harvest. With the primocanes out of the way on the
ground as opposed to being upright as they are in erect or semi-erect types, the berries fall
through the floricane vegetation more easily as the machine passes.

The primary cane-related issue in trailing blackberries is thorniness (see Section IV.E).
However, cane flexibility; is important:1 also. Most trailing blackberries such as ‘Marion’
have flexible canes that can be untangled, bundled, and trained to the trellis with minimal
cane breakage.

Time of fruit ripening varies substantially, and is a plant trait that is of primary importance to
most blackberry breeders. Obviously;, the environment the plants are grown in has a major
impact on expression. In a moderate climate such as that in Oregon, the earliest trailing black-
berries ripen 2-3 weeks before the earliest erect cultivars, which in turn are 3-4 weeks ahead of
the semi-erect cultivars. This spread of seasons is not seen when the trailing types are grown in
Arkansas where .the temperatures rise rapidly in the spring and the earliest trailing cultivars are
very little, if any, earlier than the earliest erect cultivars (C.E, Finn and J.R. Clark, pers. observ.).
Apparently, under the cool springtime conditions in Oregon, the trailing types develop more
quickly. Early ripening is often of high priority,

A. Thornlessness

The canes of most blackberry species are sparsely to profusely covered from base to tip with
thorny protrusions (Plate 2.5). Botanically, these are termed spines and are derived entirely from
the tissues outside the vascular cortex, whereas thorns are subtended from the vascular tissues. In
the United Kingdom, the terms spines or prickles are frequently used (Jennings 1988; Coyner et
al. 2005), but, in North America, publications from the earliest times have commonly used the
terms “thorn,” ‘‘thorniness,’’ “thornless,” and “thornlessness” (Hedrick 1925; Darrow 1928a,b;
Coyner et al. 2005). In this review, spines and prickles will be referred to as thorns and the
absence of spines as thornlessness, according to the North American tradition as used in the
review of Coyner et al. (2005). Since blackberries were first cultivated, thomlessness has been
desired by gardeners, horticulturalists, and growers (Hedrick 1925; Darrow 1928a). Thornless
plants from the wild and thornless sports of cultivated clones have been selected from a wide

26
range of blackberry species (Hedrick 1925; Darrow 1928a, 1928b) and many of these were
adopted into cultivation. Included are the thornless variants of ‘Austin Mayes’, released as
‘Austin Thornless’, ‘Thornless Loganberry’ also known as ‘Bauer Thornless Loganberry’
(Jennings 1981a), Thornless variants were also found in other cultivated types such as
‘Boysenberry’, ‘Marion’, and ‘Youngberry’, but most of these were found to be sterile,
frequently producing imperfect flowers and no fruit (Darrow 1928a, 1937; Zych et al.1967; C.E.
Finn. pers. obsv.). In recent,years, a few fruitful thornless selections of ‘Boysenberry’ were
identified and three of these reached advanced trialing or commercialization: ‘Mapua’ and
‘Tasman’ in New Zealand, and ‘Clone 5’ in Oregon.
When the breeding behavior of early ‘Thornless Youngberry’, ‘Cory Thornless’, ‘Thornless
Loganberry’, ‘Thornless Evergreen’, and the productive thornless ‘Boysenberry’ clones was
investigated, all seedlings from these clones Were found to be thorny. Subsequent investigations
revealed that these thornless types were chimera!, bearing thornless genes only in the epidermal
(Ll) tissues (Darrow 1937; Coyner et al. 2005).

Recently, there has been a great deal of interest and research in the ‘Lincoln Logan’ source of
thomlessness as it has been developed and incorporated into advanced breeding material.

Further sources of dominant thornlessness have been developed from the chimeral blackberry
cultivar Thornless Evergreen’ and from the ‘Thornless Loganberry’ clone fL654’ (McPheeters
and Skirvin 1983; Skirvin 1983; Hall et al. 1986a, 1986b, 1986c; Jennings 1986, 1988; Rosati et
al. 1988; Coyner et al. 2005). In each case, Ll tissues from the thornless chimeral mother plant
were tissue cultured and whole plants regenerated (Coyner et al. 2005). Plant material from
tissue-culture- derived ‘Thornless Evergreen’ was initially released to collaborators as ‘Ill 6-6’
and plants of the genetically thornless ‘Loganberry’ as ‘Lincoln Logan’ (Hall et al. 1986c).

4. Breeding with ‘Lincoln Logan’ Thornless Gene SfL. Initial attempts to use the thornless
‘Loganberry’ as a source of thornlessness were thwarted by the chimeral nature of the plant.
However, in the 1950s this gene became available for use in breeding when Jack Hull raised a
thornless ‘Loganberry’ seedling as part of his Ph.D. studies (Hull 1958; Hall and Stephens 1999).
No record remains of the use of this plant for breeding, however. Further seedlings from
thornless ‘Loganberry’-type crosses were raised by Pasquale Rosati in the 1980s, showing the
instability of the L1-L2 layers of plants in tissue culture (Rosati 1986). While plants from this

27
population of seedlings were sent to Derek Jennings in the U.K., no successful use of this
material for. breeding ensued (D. Jennings, pers. commun.). Tissue culture was also used to
produce a clonal, non-chimeral, dominant-thornless ‘Loganberry’ type from the thornless
‘Loganberry’-type clone L654 in New Zealand (Hall et al. 1986a, 1986b, 1986c; Rosati et al.
1988). The non-chimeral ‘Loganberry’- type plant produced in New Zealand resulted from a
single adventitious bud that formed within callus tissue derived from epidermal cells of L654.
This plant was been released as ‘Lincoln Logan’ (Hall et al. 1986c) and has been used
extensively for breeding in New Zealand (Hall and Stephens 1999).
Like ‘Austin Thornless’, ‘Lincoln Logan’ had numerous deleterious traits that limited its value
as a breeding parent. These included: (1) a tendency towards a semi-erect growth habit with thin
canes that were inflexible and very easy to break when trained as trailing blackberries; (2) a
tendency toward fruit with red raspberry or ‘Loganberry’ character including red, dark red, or
purple color and raspberry flavors; (3) very strong infertility from crosses with hybrid blackberry
types; (4) greater disease susceptibility, particularly to foliar diseases, in Oregon; (5) less winter
hardiness than plants derived from ‘Austin Thornless’; and (6) a tendency towards tender skin
and small fruit.
Nevertheless, each of these negative features has been overcome, and the ‘Lincoln Logan’
gene SfL has proven very useful for breeding purposes, especially in New Zealand. Thornless
plants with gene SfL can be selected at a very early stage of growth, unlike ‘Austin Thornless’
hybrids described above (Hall and Brewer 1989). Crosses between ‘Lincoln Logan’ and red-
fruited types such as ‘Tayberry’ have produced a range of fully fertile seedlings with large,
attractive fruit, resulting in the release of the New Zealand cultivar ‘Waimate’ (Hall and
Stephens 1999).
Crosses of genetically thornless ‘Lincoln Logan’ with hexaploid blackberries produced largely
infertile seedlings, similar to the experience with thorny ‘Logan5 (Waldo 1950a; Hall 1990). A
single fruitful, red-fruited, thornless ‘Lincoln Logan5 x ‘Marion5 hybrid, selected in New
Zealand, was used to transfer the gene for thornlessness to black- fruited seedlings by
backcrossing (Hall 1990; Hall and Stephens 1999). Three subsequent generations of breeding
have produced some excellent thornless selections with black fruit color fully restored. Further
crosses have transferred gene S/L into a range of selections from 3x to ~12x with full fertility
and excellent thornless canes in all but the 3x types (Hall and Stephens 1999).

28
 Transfer of thornlessness from ‘Lincoln Logan' to ‘Boysenberry’ types took three generations
of backcrossing to restore ‘Boysenberry'-type color and flavor, as ‘Loganberry5 character
dominated in early generations (Hall and Stephens 1999). Transfer of thornlessness into 8x and
higher ploidy types also produced some red-fruited types, but there has been much less transfer
of ‘Loganberry’-type character than in ‘Boysenberry’ crosses. Octoploid seedlings with this
gene were particularly valuable, as almost every plant was fruitful, productive, vigorous and
healthy, and there was a tremendous range of variability (Hall and Stephens 1999).
However, the transfer of gene S/L to tetraploid blackberries was not as easy as with 6x hybrid
blackberries and higher ploidy types. Attempts to transfer gene SfL into 4x types by crossing
‘Lincoln Logan’ with diploid R. ulmifolius were totally unsuccessful, with no seed set from any
of the pollinations (H.K. Hall, pers. observ.). Crosses with 2x raspberry produced weak seedlings
that were almost completely sterile, although fertility was restored by crossing Fa selections with
4xblackberries. Further crosses among this material will be required to produce tetraploid black-
berries with commercial potential. The cross ‘Lincoln Logan’ x ‘Florda- grand5 (2x) was also
used in an attempt to produce fertile, thornless 4x types. This produced partially fertile hybrids
with excellent bud break, and high flower and fruit numbers, although no thornless selections of
consequence were identified (H.K. Hall, pers. observ.). This cross is worth repeating to provide
thornless plants with excellent adaptation to low-chill conditions. Crosses with 4x blackberries to
produce thornless 5x hybrids were successful, giving fruitful and fully fertile selections. Two
further generations of backcrossing to upright 4x types were required to restore full fertility.
These selections had excellent fruit quality, flavor, and appearance and they were markedly
superior to ‘Merton Thornless' types. Attempts to transfer SfL to 2x raspberries failed, as all ~
progeny were thorny (Hall and Stephens 1999).
While the inheritance thornlessness derived from ‘Lincoln Logan’ is clearly dominant, there is
again a clear difference in segregation ratios between progenies according to whether the SfL
parent was the pollen or seed parent. Ratios were -63% thornless and 37% thorny when the SfL
parent was the seed parent and -63% thorny and 37% thornless when the SfL parent was the
pollen parent (Hall and Stephens 1999). In some populations, the percentage of thornless
progeny was as low as 22%, These inheritance ratios and the failure to obtain any thornless
diploids from the 2x raspberry crossed with a 3x derivative of the 6x
‘Lincoln Logan’ may indicate that this form nf thornless ness is produced through the interaction

29
of two genes on separate chromosomes, perhaps at the same locus. *
‘Lincoln Logan’ thornlessness has proven very stable, with only a few canes showing thorny
sectors in over 100,000 plants. In tissue culture, only one clone has reverted to thorny in a small
percentage of plants propagated (Hall and Stephens 1999).

B. Fruit Quality

There is a wide, range of fruit flavors among blackberry genotypes. The distinct flavors of the
section Ursini are widely desired and bring a premium price in the processing market. Typically,
these flavors, as exemplified by ‘Marion’, are characterized as rich and aromatic with a good
balance of acidity and sweetness. High acidity is important for anthocyanin stability in a
processed product and when balanced with high soluble solids the berries have a full, intense
flavor. Blackberry flavor and how it varies from season-to-season has been well characterized in
the two most important processing cultivars ‘Marion’ and ‘Thornless Evergreen' (Klesk and Qian
2003a, 2003b; Qian and Wang 2005).

Fruit from 16 cultivars grown in New Zealand and Oregon were evaluated over two harvest
seasons. The cultivars included two erect (‘Navaho’, ‘Shawnee'), one semnerect (‘Hull
Thornless’), and ten trailing (‘Chehalem’, ‘Aurora’, ‘Waldo’, ‘Black Butte’, ‘Ranui’, ‘Silvan,
‘Siskiyou’, ‘Kotata’, ‘Marion’, and ORUS 1826) blackberry genotypes along with three
blackberry/raspberry hybrids (‘Boysenberry’, ‘Tayberry’, and ‘Loganberry’). Antioxidant
activity as determined by FRAP, total phenolics, total anthocyanins, as well as individual
anthocyanins were determined. Antioxidant activity and total phenolics were not significantly
different among pultivars and locations but the variation between years within location was
significant and the genotype x environment interactions were significant. They suggested that
antioxidant activity and total phenolics should be evaluated over several years and location. The
genotype x environment interaction was also significant for total and individual anthocyanins.
Correlations between anthocyanins and antioxidant activity were much lower (r=0.63) than they
were for total phenolics and antioxidant activity (r=0.97). These correlations are lower in the
case of anthocyanins and higher for total phenolics than those reported by Moyer et al. (2002);
perhaps reflecting the more narrowly focused germplasm pool in Connor et al. (2005a, 2005b).
In addition to fruit, Wang and Lin (2000) examined antioxidant activity in semi-erect blackberry
30
leaves and Bushman et al. (2004) examined antioxidant activity of Rubus spp. seeds and seed
oils as well as their chemical composition. While, the need for breeding for higher levels of
antioxidant activity in leaves, seeds, or seed oils is not easy to currently envision, variability
exists for improvement if such a need were realized. Overall, these studies indicate that genetic
variation for anthocyanins, total phenolics, and antioxidant levels exist and that breeding for
enhanced levels would likely be possible.

V. PROSPECTS FOR THE FUTURE

The world market is expanding for blackberries for the fresh, processed, and the newer
nutraceutical markets. In the coming decades, consumers will demand a better and more uniform
product and the industry will need to provide that product for the fresh and processed markets.
To meet the expanding needs and heightened expectations of growers, marketers, processors, and
consumers, breeding programs will be called on to produce an increasingly diverse array of
cultivars (Finn and Knight 2002). While strong programs at the turn of the 21st century were set
up to begin to address these needs, it is likely that there will be a need for either public and/or
private breeding programs to expand to fully meet these needs. >
Although the number of public blackberry breeding programs has been reduced from previous
levels, the potential for improvement is still very great. This is primarily due to the expanding
market, but also to recent genetic and technological advances. Fruit quality improve- ments have
been substantial in recent years and these improvements will play a key role in acceptance of
blackberry fruit by a wider consumer base. Plant improvements, including primocane fruiting,
increased thornless cultivar options, enhanced adaptation to a range of climates, and others not
yet exploited, will play key roles in the expansion of production. Increased focus on disease and
insect resistance will be required as production increases and pest pressures expand. Further
collection of plant and fruit-based royalties by breeding programs should ensure the financial
buoyancy of these programs and promise increased investment in breeding on the backs of new
successful cultivars.

31