Quote:This program is unique among apple rootstock programs around the world because of the
focus on disease resistance and orchard performance.

The USDA-ARS/Cornell University
Apple Rootstock Breeding and Evaluation Program

William C. (Bill) Johnson
USDA-ARS/Dept. of Horticultural Sciences
Cornell University
Geneva, NY

Presented at the 42ndAnnual IDFTAConference, February 20-24, 1999, Hamilton, Ontario, Canada.

INTRODUCTION
The Cornell apple rootstock breeding project was initiated in 1968 with the original objective to
develop rootstock genotypes with improved nursery and orchard characteristics that are much
better adapted to the abiotic stresses of New York and surrounding areas.During this era interest
in high density planting systems developed in New York because they dramatically increased
precocity and productivity.Partly because of this change in the direction of the industry, the
primary constraints facing growers in the northeastern US changed from the abiotic stresses such
as drought and cold tolerance to biotic stresses such as crown rot and fire blight.Therefore the
focus of the program also changed to the factors that were exacerbated by the weaker root
systems and higher plant densities of dwarfed tree plantings.Improved productivity and
precocity, combined with resistance to critical diseases and tree size control, remained the goals
of the project through the tenure of Dr. Jim Cummins, who retired in 1993.From 1994-1997 the
project was maintained at a reduced level by Dave Gill, who retired in 1997.Cornell University
was experiencing financial difficulties during this time, and few faculty positions were being
reinstated, so during this interim period there was a great deal of uncertainty about whether the
program would continue or be terminated.Preliminary trial data with some of the earliest
genotypes from the program generated strong interest among apple growers, particularly those
from regions where fire blight is a major problem.Because the genotypes developed from the
program had strong appeal throughout the nation, members of the apple industry began to seek
ways to resurrect the program.

As a result of industry encouragement, the United States Department of Agriculture–Agricultural
Research Service (USDA-ARS) decided to look for ways to support the program.In 1997 the
USDA-ARS and Cornell University developed a Specific Cooperative Agreement, mobilizing
the resources for the continued operation of the program.In February 1998 the ARS hired
Dr. William Johnson as a research geneticist for apple rootstock breeding and evaluation and as
an adjunct Assistant Professor for the Department of Horticultural Sciences at Cornell
University.Todd Holleran was hired in April 1998 as the research technician for the project.
Because the breeder is an ARS employee, the ARS was able to significantly modify the
responsibilities of the position compared to the job description of Dr. Cummins.In addition to
breeding of new varieties, the program now has an important new focus: the evaluation of
rootstock genotypes from other breeding programs.During the first year of the program we have
begun incorporating many new rootstock cultivars from Germany, Poland, Russia, the Czech
Republic, and France.

Through the Specific Cooperative Agreement (SCA) the USDA-ARS provides the salary for a
full-time scientist and a portion of the salary for a full-time technician as well as the use of some

ARS facilities.Cornell provides the remainder of the salary for the technician as well as full
access to the university resources of germplasm, fields, laboratories, vehicles, and libraries.
Cornell provides a small operating budget for the program, but the primary source of funds will
be from competitive research grants. The royalty revenues have not yet begun to flow back into
the program, but over time this will be a source of funding for the project.Prior to the SCA
Cornell University managed the business end of plant breeding programs through the Cornell
Research Foundation (CRF).CRF will continue to oversee the licensing and sales of Geneva
series apple rootstocks, but the agreement grants a sliding scale share of 0-50% of the proceeds
to the ARS.The fraction allocated to the ARS is determined on a cultivar basis, where royalties
from the previously released cultivars disburse none to ARS, and cultivars made from crosses
performed in the future disburse 50% of royalty proceeds to ARS.Of these royalty revenues a
small fraction of the Cornell side is invested back into the research program, while the entire
federal portion goes directly back into breeding and research.We expect that within 20 years
this will represent a major source of funding to conduct research so that the project will approach
self-sufficiency.

ROOTSTOCK BREEDING OVERVIEW
The apple rootstock breeding program functions like any breeding program, but with a few
added complications.The fundamental method employed by plant breeders is to generate large
populations of novel genotypes and then to progressively make the population size smaller by
removing the genotypes that are not commercially desirable. The complication that makes apple
rootstock breeding more of a challenge is that the final evaluations of the genotypes are as a part
of a compound system with both a rootstock and a scion genotype.Because the most expensive
phase of the breeding program is orchard trials, we attempt to reduce the population sizes as
much as possible in the seedling and initial stoolbed phases of the breeding program before new
rootstock genotypes are entered into orchard trials.Our final goal is to identify a group of
rootstock genotypes that are superior to the commercially available lines.After making crosses
between rootstock genotypes with complementary characteristics (e.g.,a disease-resistant
genotype crossed by a size-controlling commercially important rootstock cultivar) a large group
of seedlings from these crosses is then inoculated with a series of pathogens.Drs. Cummins and
Aldwinckle typically inoculated all seedlings first with a mixture of strains of the fire blight
bacterium, killing the majority of the plants in the process.The survivors were then placed in
waterlogged soils withPhytophthora cactorumcultures, and again a majority of the plants were
killed in the process.The survivors from these two screenings were then placed in the field and
grown as single tree stool plants.Woolly apple aphid populations were established in the fields
by chemically suppressing their natural enemies.After a few years to establish the single tree
stools, small numbers of liners from the plants with the best stoolbed characteristics are collected
from each plant.These are used to produce finished trees in the nursery that then go on to their
first orchard trial.

Rootstock cultivars require extensive evaluation through stoolbed trials, biotic and abiotic stress
screenings, and multiple orchard performance tests.Rootstock breeding programs are operated
only by public agencies due to the long development period (25 to 30 years minimum) required
to evaluate ultimate tree size and productivity in each breeding cycle, a time scale that precludes
commercial breeding projects. All rootstock genotypes will be assigned to a specific stage in this
more formalized program. The breeding program will continue to capitalize on its historic
strength in the development of novel, precocious rootstock genotypes with higher production
efficiencies and improved resistance to biotic and abiotic stresses.We will continue to develop
varieties with a range of vigor from fully dwarfing to near standard size, but there will be a

renewed emphasis on the traits of nursery propagability, lodging resistance, tolerance to extreme
temperatures, resistance to the soil pathogens of the sub-temperate regions of the US, and
tolerance to apple replant disorder.

APPLE ROOTSTOCK BREEDING AND EVALUATION PROTOCOL
Outlined below is the multi-stage process used to create rootstock genotypes and then evaluate
them to identify those most promising for commercial release.The stages, plant numbers, and
time frames presented describe how we expect the project to work ideally, without unanticipated
setbacks.In reality genotypes may sometimes require more time than is scheduled to move from
stage to stage, and any particular genotype could move more slowly through the breeding
program.It would be difficult, however, to move a genotype through the program more rapidly
than is described and still obtain adequate levels of testing.There are approximately
5000 rootstock genotypes presently in stages 2-9 (see below), and these will continue to be
evaluated by the program and to move through the appropriate stages.New rootstock genotypes
will be added annually to stage 1 through hybridizations and to stages 4 or 8 through introduction
from other breeding programs.Promotions to higher stages will be contingent upon desirable
characteristics that exceed commercially available cultivars.Updates on progress in the current
year and plans for the coming year are presented in italics.

STAGE 1
Parental selection, hybridization, disease screenings and stool tree establishment are carried out
in years 1-3 (5000 seedlings).Pairs of parents are carefully chosen for hybridization that have
complementary characteristics (for example, an easily propagated dwarfing parent might be
crossed with an exceptionally disease resistant parent).Seeds are collected from the fruit of
these crosses, and the seeds are germinated.We then attempt to infect seedlings with fire blight
bacteria (Erwinia amylovora) and crown rot fungus (Phytophthoraspp.), the rootstock diseases
that cause the most difficulty for US apple producers.We expect only 20% of our seedlings will
survive these screens, and we plant these populations in our fields to establish single plant stool
tree populations with enhanced disease resistance.

Thirty rootstock genotypes were chosen as potential parents for a crossing block. Each of these
will be grown for controlled pollinations beginning in 2001.

STAGE 2
Stool plant selection, nursery liner establishment and nursery tree growth occur in years 4-6
(1000 stool trees).Genotypes are propagated as single tree stool plants, and nursery liners are
harvested from genotypes that show adequate rooting, few spines, and non-brittle wood.Liners
are moved to a nursery for years 5 and 6, where finished trees are produced after budding with a
vigorous, non-precocious scion (Crispin).In years 5 and 6 stool trees are again evaluated for
resistance to fire blight and for infestation levels with woolly apple aphids, and susceptible
genotypes are discarded from the nursery and from the stool tree fields.Only approximately
100 genotypes per year move on to actual orchard tests.

We are presently evaluating (fall 1998) approximately 2000 unique genotypes of individual stool
trees for spines, brittleness, and rooting ability. An additional ~1000 genotypes will be evaluated
in 1999 or 2000.