Freezing Tolerance in the Triticeae - Translational Genomics for Crop Breeding: Abiotic Stress, Yield and Quality - page 113

ssp. spontaneum , representing the primary gene
pool. Accordingly, weak frost tolerance alleles
at QTLs on chromosomes 2H, 4H, and 6H were
contributed by the wild parent in the cross 'Arta'
x Hordeum v. ssp. spontaneum 41-1 (Baum et al.
2003).
A well known success story was the produc-
tion of the interspecific fertile allopolyploid trit-
icale in the 1960s, obtained by crossing wheat
and the most frost-tolerant species among small
grain cereals, that is, rye (Hommo 1994). Since
then triticale production grew and expanded,
until the most frost-resistant cultivar obtained,
'Aubrac,' was demonstrated to perform much
better at freezing temperatures than winter wheat
and, under the specific test conditions, as well
as a frost resistant rye (Rizza et al. 1997). The
few examples of using wild relatives to improve
tolerance to frost and other abiotic and biotic
stresses, and the large amount of diversity stored
in gene banks, leaves the impression that there is
still a lot of novel germplasm with potential for
exploitation in plant breeding. These few exam-
ples also pose questions about how to most effi-
ciently use gene bank accessions and mine their
sequence diversity in the genomic era. Despite
great efforts to coordinate their activities, the
quality of existing collections (availability of
pure, properly characterized and documented
materials) is still to be improved. Gene banks are
expected to significantly improve their usability
when they adopt systematic genotyping at key
adaptation genes, or even genome-wide genotyp-
ing by sequencing (Kilian and Graner 2012), of
pure germplasm stocks, coupled with their pre-
cise phenotyping. Molecular markers have begun
to unravel genetic diversity (Spooner et al. 2005).
For example, a set of 373 bread-wheat accessions
reflecting the maximum of both the number of
observed alleles and the number of geographical
origins was selected by screening nearly 4,000
bread-wheat accessions from 73 countries with
SSR markers, and by using passport data. Then,
the seeds of all of these core accessions were
made available, together with their DNAs stored
on a single 384-well plate, ready for different

molecular applications (Balfourier et al. 2007).
Not only will proper procedures for extraction
and long-term storage of DNA be required, the
quality of gene bank information systems also
still varies considerably, from poorly completed,
to functional online databases with descriptors,
molecular data, and geo-referenced data (Khoury
et al. 2010). From a breeding point of view, once
key loci for useful agronomic traits have been
cloned, genotypic characterization can uncover
further allelic variants that can be tested for supe-
rior performance. Allele mining of collections
can be accomplished in two major ways: (1)
a modified TILLING (targeting induced local
lesions in genome) method called EcoTilling
that detects the mutations naturally occurring
in the primary and secondary gene pools; and
(2) sequencing-based allele mining (Kumar et al.
2010).
Another approach to improving the efficiency
of genetic resource utilization is the focused
identification of germplasm strategy (FIGS), cur-
rently under development and evaluation. This
approach uses the information about the envi-
ronment where the accession was collected to
select the accessions most likely to contain the
traits that confer adaptation to that environ-
ment ( http://www.figstraitmine.org) . The rele-
vance of the FIGS approach in better target-
ing accessions held in gene banks for valuable
traits was demonstrated by El Bouhssini and col-
leagues (2011) when they compared random vs.
FIGS-based screening of the International Cen-
ter for Agricultural Research in the Dry Areas
(ICARDA) collection for Russian wheat aphid
resistance. While the screening of 5,000 ran-
domly selected bread wheats did not identify
resistant accessions, screening 510 accessions
selected by means of FIGS enabled identifica-
tion of 12 promising lines with different lev-
els of resistance. A good example of how ex-
situ germplasm collections can be mined for
useful alleles by FIGS is a large-scale project
that enabled the cloning of 7 new Pm3 pow-
dery mildew resistance alleles from a focused
subset of 1,300 wheat landraces. Doubling the