Exploring the use of
RAD markers in
tree breeding programmes Pablo Fuentes-Utrilla
Why trees?
1. There is a need to increase yield per hectare in the
forest industry
Source: Confor, Wood fibre availability and demand in Britain 2007-2025
Why trees?
1. There is a need to increase yield per hectare in the
forest industry
2. Adaptation to climate is of concern in tree breeding
programmes due to long life cycles of trees
Source: IPCC
Why trees?
1. There is a need to increase yield per hectare in the
forest industry
2. Adaptation to climate is of concern in tree breeding
programmes due to long life cycles of trees
3. Resistance against pests & diseases increasing
importance for tree production and conservation
Tree breeding cycle
growth
phenotypic
sellection
disease resistance
Tree breeding cycle
growth
phenotypic
sellection
propagation:
-clones
-F1
disease resistance
Tree breeding cycle
growth
phenotypic
sellection
propagation:
-clones
-F1
evaluation
disease resistance
Tree breeding cycle
growth
phenotypic
sellection
improved
material
propagation:
-clones
-F1
validation/
rejection
evaluation
disease resistance
Tree breeding cycle
growth
phenotypic
sellection
improved
material
propagation:
-clones
-F1
validation/
rejection
crosses
-F2
evaluation
disease resistance
Tree breeding cycle
growth
disease resistance
PROBLEM...
phenotypic
sellection
tree breeding cycles
improved
material
are
propagation:
-clones
-F1
crosses
-F2
...ZZZ
long (10-30 yr) & costly
evaluation
validation/
rejection
Marked-assisted selection in tree breeding
- Genetic markers are promising tools to develop QTLs
to shorten breeding cycles (marked-assisted selection)
- QTL mapping
-before, AFLP and SSR markers (10s to low 100s)
-now, SNP chips (some 100s, now some 1000s)
but... development is not cheap
- RAD can provide 1000s of markers across the
genome for
-rapid SNP recovery / genotyping
-QTL mapping
Objectives
- To develop the RAD technique in two forest species
with existing breeding programmes
One conifer: Picea sitchensis (Sitka spruce)
One angiosperm: Ulmus minor (Field elm)
- To evaluate the ability of RAD markers to detect
QTLs for traits of interest
Climatic adaptation
Wood production
Pathogen resistance
Tree material
Sitka spruce breeding programme
(Forestry Commission, UK)
- main aim: production
- breeding for wood production traits
Marked assisted selection trial (6 yr-old):
- 3 full-sib families / 1500 siblings per family
- replicates in 3 site (Dornoch, Llandovery, Devon)
Two traits measured:
HEIGHT recorded at age 5 in ALL families and sites => wood
production QTLs
FLUSHING DATE recorded at age 5 in ONE family and ALL sites
=> climatic adaptation QTLs
Tree material
Field elm breeding programme
(UPM, Spain)
- main aim: conservation
- breeding for resistance against Dutch elm
disease (DED)
Progeny trial for DED heritability (5 yr-old)
10 full-sib families/ 24-276 siblings per family
7 families Resistant x R, 1 family Susceptible x S, 2 families R x S
One trait measured:
RESISTANCE to DED => DED resistance QTLs
Inoculations carried out in May 2010 (prelim. results show
intrafamily variation).
Genomic background
No data available for P. sitchensis and U. minor.
Data from related taxa presented
SbfI
(CCTGCAGG)
Gbp
CG%
No rest
sites
Size rest
frag (kb)
No RAD
tags
ind/lane
20x cov*
ind/lane
40x cov*
Picea abies
18.875
39.9
107459
175.7
214918
3.5
1.7
Ulmus glabra
1.056
36.7
4041
261.4
8083
93
46
EcoRI
(GAATTC)
Gbp
CG%
No rest
sites
Size rest
frag (kb)
No RAD
tags
ind/lane
20x cov*
ind/lane
40x cov*
Picea abies
18.875
39.9
6125758
3.1
12251517
0.12
0.06
Ulmus glabra
1.056
36.7
356876
3.0
713751
2.1
1.1
*assuming 15 million reads per lane
Challenges
Handicap: large (elm) and very large (spruce) genomes
Possible solution: complexity reduction method by
second digestion.
Options:
- use a frequent cutter and select by fragment size
- use a methylation sensitive cutter
(non-coding plant DNA >>> methylated than coding
DNA)
- other suggestions?
Thanks for listening!
Questions? Comments?