Environmental Engineering Reference
In-Depth Information
1.3.3 t
rEES
aS
a
B
iofuEl
S
ourcE
Wood is perhaps the oldest sources of bioenergy in the world. As with perennial grasses, the potential
now exists to use wood from fast-growing tree species as a feedstock to produce liquid fuels.
Candidate trees include several species of pine (
Pinus
spp.),
Eucalyptus
spp., willows (
Salix
spp.), and
poplar (
Populus
spp.). Because of their importance in the timber and paper industries, considerable
genetic resources are already in place for many tree species.
Populus trichocarpa
, also known as
black cottonwood, has a genome size of 480 Mbp and a haploid chromosome number of 19 (Tuskan
et al. 2006). This is perhaps the smallest genome size among tree species, and so it was selected for
a genomic sequencing project. A consortium of researchers from governmental organizations and
universities from the United States, Canada, and several European countries was formed to conduct
the project (International
Populus
Genome Consortium 2010). The strategy used was whole genome
shotgun sequencing, in which thousands of randomly selected genomic clones are sequenced and
are then assembled in silico based on overlapping sequences. The assembled sequences were then
anchored to specific linkage groups on the basis of known locations of SSR markers identified in
the earlier stages of the project (Tuskan et al. 2004). Completion of the assembled sequence was
reported in 2006 by Tuskan et al. An international collaborative effort has also been formed to
sequence the
Eucalyptus
genome. The International
Eucalyptus
Genome Network (EUCAGEN)
includes researchers from both public and private sectors from every continent. The first draft
sequence of
Eucalyptus grandis
was recently reported (EUCAGEN 2010). A critical next phase
for these projects is to annotate the genome sequences; that is, to assign putative functions to
the genes. Such information should be useful to tree breeders to assist in selecting plants with
desirable traits. This is especially important given the long generation times for woody plants.
Genomics resources such as these should help to hasten the development of new cultivars with
improved biofuel traits.
1.4 conclusIons and Future ProsPects For BIoFuel croPs
There are many plant species that can be used for production of biofuels, from existing food and
feed crops to dedicated biomass crops like trees and perennial grasses. Existing crops have, in the
past, been selected mostly for yield of food and feed, and many of the emerging biomass crops
are barely domesticated at all. These crops will require specific improvements to make them more
amenable to biofuel production. Fortunately, considerable natural variation is present among most
cultivated crop species, and many of these crops also have a repository of genes in their wild rela-
tives that can be exploited for crop improvement. However, there are many challenges for plant
breeders in development of biofuel crops. Some of these species have very long generation times
and are highly outcrossing, preventing the creation of inbred lines for classical genetic studies. In
addition, the genomes of many biomass crops (e.g., sugarcane) are very large and complex. Still,
significant progress has been made in genetic improvement of biofuel feedstocks.
A combination of traditional and molecular tools has been used in biofuel crop improvement,
and the molecular tools are becoming increasingly important. Molecular markers are now routinely
used to assess the genetic diversity in germplasm collections, to map important quantitative traits
(QTL), and to select for desirable traits linked to those markers. Although progress has been made
by utilizing QTL analysis and MAS, it has been met with some limitations, especially for highly
polygenic traits. However, it is now possible to quickly genotype many markers, essentially covering
the entire genome, at relatively little cost. Thus, whole-genome selection has been proposed as the
next phase in marker-assisted plant breeding (Heffner et al. 2009). The procedure begins with col-
lecting phenotypic and genotypic data from a large collection of breeding material and germplasm
of interest. This “training population” is used to create a complex prediction model of trait-marker
associations. This information can then used to predict the breeding value of any individual on the
basis of all of its available genetic marker information. Data from further evaluations in the field can
