To secure a sustainable energy source for the future, we need to develop an alternative to fossil fuels. Cellulose-based biofuel production has great potential for development into a sustainable and renewable energy source. The thick secondary walls of xylem cells provide a natural source of cellulose. As a result of the extensive production of wood through cambial activity, massive amounts of xylem cells can be harvested from trees. How can we obtain a maximal cellulose biomass yield from these trees? Thus far, tree breeding has been very challenging because of the long generation time. Currently, new breeding possibilities are emerging through the development of high-throughput technologies in molecular genetics. What potential does our current knowledge on the regulation of cambial activity provide for the domestication of optimal bioenergy trees? We examine the hormonal and molecular regulation of wood development with the aim of identifying the key regulatory aspects. We describe traits, including stem morphology and xylem cell dimensions, that could be modified to enhance wood production. Finally, we discuss the potential of novel marker-assisted tree breeding technologies.
Wood, the secondary xylem of plants, enables water transport to the shoot and provides structural support. The main components of the thick secondary cell walls of xylem are cellulose, hemicellulose and lignin (Rowell et al., 2000). As a result of these cellulose-rich cell walls, wood has the potential to be used as a raw material for the production of ‘second-generation’ lignocellulosic biofuels. Wood does not yet represent a significant source for biofuels; currently, the major product of the first-generation biofuel industry is ethanol produced from sugarcane and maize starch (Sagar & Kartha, 2007; Fairley, 2011). This crop-based biofuel production has, however, led to direct competition between the use of natural resources and land for food and energy production (Searchinger et al., 2008). Second-generation biofuels have the potential for development into a sustainable energy source and an alternative to the nonrenewable fossil fuels (Solomon, 2010).
Wood is produced through the activity of vascular cambium, and our aim in this review is to explore how we can breed forest trees by fine tuning the activity of this meristem. Cambium develops when parenchymatic cells between the xylem and phloem in vascular bundles and between the bundles start to divide (Fig. 1). Cell divisions within the vascular bundles give rise to fascicular cambium and those between the bundles to interfascicular cambium. Together, these two domains of cell division form a continuous cylinder of vascular cambium. The most common organization of this meristem is a single bifacial (i.e. bidirectional) cambium that produces secondary phloem outwards and xylem inwards (Carlquist, 1988).
Wood consists mainly of longitudinally elongated, dead xylem cells with a hollow lumen and a thick secondary cell wall. The xylem cells include water-transporting tracheary elements (tracheids and vessel elements) and fibers. Angiosperm wood contains vessel elements and fibers, whereas tracheids form gymnosperm wood. The transversally elongated, mainly parenchymous, ray cells form a minor component of wood (Turner et al., 2007). The relative amount of lignocellulosic biomass varies between different xylem cell types: vessel elements with a wide diameter have a lower ratio of cell wall per unit cell volume than either fibers or tracheids. A special type of xylem, called reaction wood, is produced by a plant to straighten a leaning stem. Reaction wood has an increased fiber to vessel ratio, which leads to a high lignocellulosic biomass (Joyrez et al., 2001). Angiosperm reaction wood is enriched in cellulose and low in lignin content (Timell, 1969; Andersson-Gunnerås et al., 2006), whereas, in gymnosperms, the lignin content is increased (Timell, 1969).
With regard to second-generation biofuels, ethanol production is currently most advanced, whereas other potential products, including butanol and furans, are at earlier stages of development (Sanderson, 2011). Lignocellulosic ethanol production involves the pretreatment of woody material, hydrolysis for monosaccharide production (saccharification) and the production of ethanol through sugar fermentation (Ragauskas et al., 2006). A major obstacle in the conversion of lignocellulosic biomass to sugars is recalcitrance, the resistance of plant cell walls to microbial and enzymatic deconstruction (Himmel et al., 2007). Techniques for the efficient release of sugars from cellulose and the removal of lignin, the major nonsugar component of the cell wall, need to be developed before lignocellulosic biofuel production can be cost-effective (Himmel et al., 2007; Sanderson, 2011). Much research has been dedicated to the modification of the chemical composition of the secondary cell wall in order to facilitate biofuel production. As this topic is outside the scope of our review, we refer the reader to reviews covering current knowledge on this issue (Abramson et al., 2010; Lu et al., 2010; Simmons et al., 2010; Pu et al., 2011).
Currently, Populus species and willow, together with the monocot grasses Miscanthus and switchgrass, are grown as bioenergy feedstocks for lignocellulosic biomass production (Karp & Shield, 2008; Yuan et al., 2008; Somerville et al., 2010). In addition, Pinus and Eucalyptus have the potential to be developed into future biofuel feedstocks (Frederick et al., 2008; Romaníet al., 2010). Our focus in this review is on tree species which produce an extensive amount of wood through cambial growth. We concentrate on the cambial activity in the stem, as this is the most important and easily accessible organ for biomass harvesting in trees. The ideal form for a woody biomass feedstock would combine a large stem volume with high lignocellulosic biomass. In single-stem cultivation, the volume of woody biomass is determined by the stem height and diameter, together with the percentage of wood. In contrast with single-stem cultivation, fast-growing tree species, such as Populus and willow, are commonly grown in short-rotation coppice cycles (Tuskan, 1998; Aylott et al., 2008). In this cultivation form, the original main shoot is cut down to induce the production of several new shoots. These shoots can be harvested and the cycle is repeated every few years. As our focus is on cambial activity, we do not discuss the methods used to increase the apical growth rate or the shoot number; recent reviews have described the topic of general plant growth enhancement for biomass production (Demura & Ye, 2010; Harfouche et al., 2011).
In contrast with bioenergy feedstock species, most wood development research has taken place in model plant species for which extensive genetic information and well-established experimental protocols are available. One of the few exceptions is Populus species, which are both bioenergy crops and model tree species for basic plant science studies (Jansson & Douglas, 2007). Even herbaceous plants often produce secondary xylem, and can thus contribute to our understanding of this process. Despite its diminutive size relative to a tree, Arabidopsis still displays cambial-driven secondary thickening in several organs (Zhang et al., 2011). Another important model for wood development studies is the cell culture of Zinnia. In this system, leaf mesophyll cells are induced to dedifferentiate into cambial-like cells, which subsequently differentiate into tracheary elements under suitable hormonal treatments (Endo et al., 2008). This system allows the experimental testing of the requirements for different plant substances for xylem cell differentiation. We now take a look at the current knowledge on the regulation of wood development in plant species.
Several hormones contribute to the regulation of cambial activity
Most plant hormones seem to be able to promote cambial activity; auxin, cytokinin, gibberellin, ethylene and jasmonate have all been shown to have a stimulating effect (Fig. 1) (Elo et al., 2009). The classical hormone treatment studies are nowadays complemented by functional studies which take advantage of transgenic plants with modified hormonal signaling status. A common complication in these studies is that hormonal signaling is enhanced or suppressed in the entire plant, including the shoot apical meristem. Thus, the results may partially reflect an indirect consequence of an altered growth rate; they do not necessarily reveal the specific effect of modified hormone signaling on cambial activity. In the future, it will be important to use cambial-specific promoters for transgenic studies; this will allow hormonal signaling to be exclusively modified at cambial cells without any unnecessary effect on apical meristem activity.
A gradient of auxin (indole-3-acetic acid, IAA) has been detected across the cambial zone of both Populus and Pinus trees (Uggla et al., 1996; Tuominen et al., 1997), with the concentration peaking in dividing cambial cells. The auxin distribution correlates with the expression pattern of auxin signaling genes (Moyle et al., 2002). The hormone gradient is formed when auxin produced at the shoot apex (Sundberg & Uggla, 1997) is transported rootwards in the stem (Little & Savidge, 1987; Schrader et al., 2003; Björklund et al., 2007). Auxin transport is required for normal cambial activity, as its inhibition in Pinus stem suppresses secondary xylem formation (Sundberg et al., 1994), and an increase in the rate of auxin transport has been linked to enhanced secondary development in the Arabidopsis inflorescence (Ko et al., 2004). Further supporting the regulatory role of auxin signaling, transgenic Populus trees with a reduced auxin responsiveness display fewer cell divisions in the vascular cambium, resulting in compromised radial growth of the stem (Nilsson et al., 2008).
Interestingly, the level of cambial auxin signaling appears to be regulated separately both by the rate of auxin transport and by the cambial responsiveness to the transported hormone (Baba et al., 2011). The auxin responsiveness of the cambium decreases with the onset of dormancy, whereas basipetal auxin transport remains active (Baba et al., 2011) and cambial auxin levels remain stable (Uggla et al., 1998; Schrader et al., 2003, 2004). Thus, both the rate of auxin transport and the cambial responsiveness to this hormone represent potential breeding targets for the enhancement of the radial growth rate.
The rate of secondary xylem production has been shown to be increased in transgenic plants with either enhanced gibberellin signaling (Mauriat & Moritz, 2009) or biosynthesis (Eriksson et al., 2000; Biemelt et al., 2004; Dayan et al., 2010). However, these plants generally grew faster than the controls, so that the enhanced xylem production may have been partially a result of the increased growth rate. Enhanced gibberellin signaling has been shown to increase polar auxin transport in Populus stem by Björklund et al. (2007), indicating that this hormone may partially stimulate cambial activity by promoting polar auxin transport into cambial cells. Bioactive gibberellins, together with biosynthetic and signaling genes, are, however, present at the cambial zone of Populus (Israelsson et al., 2005), indicating a more direct role of gibberellin in the regulation of cambial activity. Further evidence for a direct role has been found in the Arabidopsis hypocotyl, where secondary xylem production is accelerated on initiation of flowering, when the plant switches from the vegetative to the reproductive stage (Sibout et al., 2008). Recently, Ragni et al. (2011) have shown that gibberellin acts as a mobile shoot-derived signal which activates the onset of extensive xylem production. Furthermore, this effect did not appear to be directly dependent on enhanced auxin transport, as treatment with an auxin transport inhibitor did not prevent xylem expansion (Ragni et al., 2011). Interestingly, transient induction of a gene encoding a gibberellin biosynthesis enzyme was observed in Populus during cambial reactivation in the spring (Druart et al., 2007). This indicates that gibberellin-driven activation may also play a role in the seasonal- or age-related regulation of cambial growth in tree species.
Cytokinin signaling has been shown to be required for cambial function. In Arabidopsis, cambial activity was completely missing in the root of a quadruple mutant lacking four key cytokinin biosynthetic enzymes (Matsumoto-Kitano et al., 2008). Transgenic Populus trees with impaired cytokinin signaling displayed compromised radial growth caused by a decreased number of cell divisions in the vascular cambium (Nieminen et al., 2008). In addition, genes encoding cytokinin receptors and cytokinin primary response genes were abundant in the cambial region of Populus stem (Nieminen et al., 2008). Taken together, these studies suggest that cytokinins are major hormonal regulators required for cambial cell proliferation during wood formation. Elevated cytokinin signaling could potentially enhance the cambial cell division rate and, in turn, the production of woody biomass. Indicating a further role for cytokinin in vascular development, Bishopp et al. (2011) have recently shown a role for cytokinin in directing the radial pattern of polar auxin transport during primary development in Arabidopsis root. Auxin flow at the cambial region might be similarly directed and patterned by cytokinin.
Ethylene signaling has been shown to promote the production of reaction wood; transgenic Populus trees with reduced ethylene response formed less wood in response to tilting of the stem than the controls (Love et al., 2009). In addition, ethylene treatment of a segment of Populus stem stimulated cambial growth in that area (Love et al., 2009), and ethylene biosynthetic and signaling genes were expressed in developing reaction wood (Andersson-Gunnerås et al., 2003, 2006). As reaction wood in angiosperms has an increased fiber and cellulose content, it might be beneficial to engineer trees with enhanced ethylene signaling to produce this type of wood even when growing in an upright position. Furthermore, inhibition of ethylene biosynthesis abolished tracheary element differentiation in the Zinnia cell culture system (Pesquet & Tuominen, 2011). This indicates that, in addition to reaction wood formation, ethylene may be required for certain basic processes during xylem development.
Recently, jasmonate has also been linked to the regulation of cambial activity. In Arabidopsis inflorescence stems, interfascicular secondary xylem development and total stem diameter were increased in plants with elevated jasmonate signaling and decreased in plants with reduced signaling (Sehr et al., 2010). Interestingly, jasmonate and ethylene pathways appear to be connected. Zhu et al. (2011) showed that jasmonate enhances the activity of transcription factors that induce the majority of ethylene response genes. It remains to be seen whether both jasmonate and ethylene activate similar downstream molecular processes in the regulation of cambial activity, and whether jasmonate plays a role in reaction wood development.
Hormonal synergy in the regulation of cambial activity
As hormones act as major switches that activate entire downstream regulatory pathways, the modification of their signaling might represent an efficient way to adjust general cambial activity. Surprisingly, the application of almost any hormone to a decapitated tree trunk appears to promote cambial cell division and wood production (Elo et al., 2009). Intuitively, hormones are not expected to be interchangeable with each other; they presumably have certain specific functions in the cambial meristem. Do the hormones share additive, synergistic or antagonistic relationships? We may need to have several overlapping hormonal signaling pathways in the same cell to induce a specific phenotypic effect. As already described, two hormones, cytokinin and gibberellin, have recently been implicated in the regulation of auxin transport (Björklund et al., 2007; Bishopp et al., 2011), and jasmonate and ethylene can induce, at least in part, similar downstream responses (Zhu et al., 2011). Extensive cross-talk between different hormones may act as an underlying mechanism of cambial development. Instead of modifying hormonal signaling, another breeding option is to use the downstream regulators as tools. By modifying the expression level and pattern of these transcription factors and other regulators, it may be possible to adjust certain highly specific features of wood development (Fig. 1).
Regulation of the xylem vs phloem ratio
One way to facilitate the production of lignocellulosic biomass is to breed trees with an increased xylem to phloem ratio. Currently, one transcription factor acting as a positive regulator of secondary phloem formation has been identified in Populus (Yordanov et al., 2010). Over-expression of PtaLBD1, a member of the LATERAL ORGAN BOUNDARIES DOMAIN (LBD) family, led to an increase in the production of secondary phloem and, subsequently, tree diameter. Furthermore, in trees over-expressing this gene, the expression of a Populus ortholog of a major Arabidopsis phloem identity regulator, ALTERED PHLOEM DEVELOPMENT (APL) (Bonke et al., 2003), was up-regulated. Interestingly, the number and width of xylem rays were also increased in the over-expressing trees. It remains to be seen whether this was caused directly by increased LBD activity or indirectly by the increased amount of phloem cells.
Receptor kinase regulators of cambial activity
Recently, a mobile phloem-produced peptide signal has been identified to function in the maintenance of cambial meristem. The CLE peptide TDIF/CLE41 was originally identified as a suppressor of tracheary element differentiation in the Zinnia cell culture system (Ito et al., 2006). In Arabidopsis, this peptide is secreted from the phloem and binds to a receptor-like kinase, PHLOEM INTERCALATED WITH XYLEM (PXY), in cambial cells (Fisher & Turner, 2007; Hirakawa et al., 2008). In the loss-of-function pxy mutant, xylem and phloem tissue organization is severely disturbed; the two tissues are no longer separated, but are intermixed (Fisher & Turner, 2007; Hirakawa et al., 2008). Furthermore, the position of the cell division plane in cambial cells is disturbed; the developing xylem cells have an abnormal curved morphology (Etchells & Turner, 2010a,b). Ectopic expression of CLE41 under the 35S promoter results in disorganized cell division, whereas phloem-specific over-expression accelerates normal cambial activity (Etchells & Turner, 2010a). The CLE41-PXY signaling system thus promotes the proliferation of cambial cells and maintains their meristematic activity by repressing their differentiation into xylem cells (Hirakawa et al., 2008, 2010). The CLE41 peptide confers the positional information via the PXY receptor necessary for the correct positioning of the plane of cell division in the middle of a dividing cambial cell (Etchells & Turner, 2010b).
The transcription factor WOX4, encoded by a WUSCHEL-related HOMEOBOX gene, has recently been recognized as one target of the CLE41-PXY signaling pathway. Its expression is induced by CLE41 and, in the Arabidopsis wox4 null mutant, the number of cambial cell divisions is reduced (Hirakawa et al., 2010; Ji et al., 2010). These results indicate that WOX4 promotes cambial cell divisions downstream of the CLE41-PXY pathway. Two WOX4 orthologs, PttHB2 and PttHB3, are expressed in the cambium of Populus (Hertzberg & Olsson, 1998), indicating that the CLE-PXY-WOX4 signaling system is active in tree cambium. Interestingly, a link has emerged between this signaling pathway and auxin (Suer et al., 2011). WOX4 expression is up-regulated by auxin, and auxin treatment is not able to stimulate cambial cell division activity in either the pxy or wox4 mutant (Suer et al., 2011). Thus, the CLE-PXY-WOX4 signaling system may play a role in the regulation of auxin responsiveness of the cambium.
In addition to PXY, two other leucine-rich receptor-like kinases, MORE LATERAL GROWTH1 (MOL1) and REDUCED IN LATERAL GROWTH1 (RUL1), with opposite effects on cambial activity have been identified recently in Arabidopsis (Agusti et al., 2011). Secondary growth was increased in the inflorescence stem of the mol1 loss-of-function mutant. By contrast, interfascicular cambial activity was reduced in the null rul1 mutant. Accordingly, MOL1 functions as a repressor, and RUL1 as an activator, of cambial activity (Agusti et al., 2011). Ligands binding to these receptors remain to be identified. Increasing the phloem production of mobile peptide ligands for cambial receptors could provide an opportunity to stimulate cambial activity in trees.
Class III HD Zip–KANADI interaction in the patterning of cambial activity
A novel group of regulators, the Class III HD Zip transcription factors, together with their antagonists, the KANADI transcriptional regulators, have recently become a focus of research. The HD Zip factors possess an miRNA 165/166 binding site, and miRNAs play a crucial role in regulating their expression. The Arabidopsis Class III HD Zip family is formed by five genes: Revoluta (REV/IFL), Phabulosa (PHB), Phavoluta (PHV), Corona (CNA/AtHB15) and AtHB8 (Emery et al., 2003).
In Arabidopsis, ectopic expression of the Class III HD Zip genes has been shown to promote xylem differentiation (Ilegems et al., 2010). Mutations in the auxin-regulated gene REVOLUTA (REV, described earlier as INTERFASCICULAR FIBERLESS1 or IFL1) resulted in the absence of interfascicular fibers in the stem and disrupted development of xylem fibers and vessel elements. By contrast, KAN1 can inhibit vascular tissue differentiation (Eshed et al., 2001; Kerstetter et al., 2001; Ilegems et al., 2010). Loss of KANADI function resulted in increased cambium activity; in the quadruple loss-of-function KANADI mutant, extra divisions of cambium cells were observed in the basal part of the Arabidopsis hypocotyl (Ilegems et al., 2010).
A close link between the Class III HD Zip genes and auxin signaling is suggested by their overlapping expression and accumulation patterns in plant tissues (Heisler et al., 2005) and by the induction of ATHB8, REV, PHV and ATHB15 expression by auxin (Baima et al., 1995; Zhou et al., 2007). Furthermore, Arabidopsis Class III HD Zip and KANADI multiple mutants showed atypical expression patterns of the auxin transporter Pin1 (Izhaki & Bowman, 2007). In the ifl Arabidopsis mutants, both basipetal auxin transport and cambial activity at the basal parts of inflorescence stems were reduced, indicative of a connection between auxin and REV action (Zhong & Ye, 2001).
In Arabidopsis, KANADI multiple loss-of-function mutants and REV, PHB and PHV gain-of-function mutants display phenotypes with an amphivasal vasculature, where the xylem surrounds the phloem. This suggests that similar vascular phenotypes can be achieved by controlled up-regulation of Class III HD Zip, as well as by down-regulation of counteracting KANADI genes. Thus, the Class III HD Zip–KANADI system is likely to play an important role in regulating the patterning of cambial activity.
Further indicating the role of HD Zip genes in the regulation of cambial patterning, a Populus ortholog of REV is expressed during secondary growth (Robischon et al., 2011). A vascular phenotype was observed in hybrid aspen over-expressing an miRNA-resistant version of the Populus REV homolog (PRE) under the 35S promoter (Robischon et al., 2011). In these lines, additional cambial regions develop in the stem cortex. Interestingly, in this ‘second’ cambium, the position of the differentiating tissues is inversed; phloem develops inwards and xylem outwards.
The Class III HD Zip–KANADI system could provide us with tools to design trees with multiple active cambial meristems; in theory, this might allow for faster growth. However, the presence of several cambial layers might weaken the physical strength of the tree stem. A more practical breeding strategy would probably be to increase wood production from a single cambial layer.
Relative abundance and morphology of different xylem cell types
The amount of lignocellulosic biomass in wood can be increased by an increase in the secondary cell wall volume. Thus, the fine tuning of the abundance, dimensions and cell wall thickness of various xylem cell types provides a potential avenue for bioenergy tree breeding. As fibers have thicker cell walls than vessel elements, maximizing their number would be beneficial. A few factors that regulate the differentiation of specific xylem cell types have already been identified. Ectopic expression of the transcription factors VASCULAR-RELATED NAC-DOMAIN 6 (VND6 ) and VND7 can induce xylem vessel element differentiation in both Arabidopsis and Populus leaves (Yamaguchi et al., 2008), indicating that these two factors represent major regulators of vessel identity.
Another breeding option is to reduce the diameter and increase the number of xylem cells. Several hormones have been linked to the control of xylem cell dimensions. Auxin increases the width and length of xylem vessels and fibers (Nilsson et al., 2008), whereas ethylene reduces them (Love et al., 2009), and gibberellin increases the fiber length (Eriksson et al., 2000). Future research will no doubt identify specific factors acting downstream of hormonal signals in the regulation of xylem cell morphology.
The morphology of xylem cells is also affected by the timing of their death. A thermospermine synthase ACAULIS5 (ACL5) controls xylem differentiation by preventing premature programmed cell death (Muñiz et al., 2008). In the hypocotyl of the Arabidopsis acl5 loss-of-function mutant, secondary growth is inhibited, xylem fibers are missing and vessel elements are small and have thin secondary cell walls. Accordingly, spermidine treatment prolongs tracheary element differentiation in the Zinnia cell culture system and increases the length and width of differentiating cells (Muñiz et al., 2008). Thus, the delay of xylem cell death does not appear to be a feasible strategy for tree breeding, as any potential increase in the secondary cell wall thickness would be counterbalanced by an increase in cell size.
Another approach is to directly increase the thickness of secondary cell walls. The transcription factors NAC SECONDARY WALL THICKENING PROMOTING FACTOR1 (NST1) and SECONDARY WALL-ASSOCIATED NAC-DOMAIN 1 (SND1) have been shown to regulate secondary cell wall thickening in Arabidopsis fibers (Zhong et al., 2006; Mitsuda et al., 2007). Several MYB transcription factors function downstream of SND1; they act by up-regulating the genes related to the synthesis of secondary cell wall components, including cellulose and lignin (Zhong et al., 2007; McCarthy et al., 2009).
As a result of their large size and long generation times, trees are not easily accessible for traditional breeding. The scope of tree breeding has been limited relative to food crops, which have been bred for millennia in some cases. The most common breeding method, used in Populus and Eucalyptus species, has been to induce hybrid vigor by crossing two species in the same genus (Potts & Dungey, 2004; Stanton et al., 2010). There is thus plenty of potential to domesticate trees for human needs through the selection of desired trait combinations (Nelson & Johnsen, 2008; Karp et al., 2011). New breeding possibilities are currently emerging through the development of high-throughput methodologies in molecular genetics.
For the biofuel industry, a major aim of tree domestication is to enhance cellulosic biomass production. In this article, we have described what is currently known about the regulation of wood development, and how this knowledge could be used to identify candidate genes for breeding purposes. Once a candidate gene for a favorable trait has been identified in a model species, its ortholog(s) can be cloned in cultivated species. Transgenic plants can be designed to either over-express or silence the gene of interest, and tested for their phenotypic properties in field trials. However, as genetically modified trees must obtain both regulatory permission and public acceptance, an easier approach may be to exploit natural variation in the candidate genes available in the tree species gene pools.
Tree cultivars can be screened for lines that either harbor a mutant form or that over- or under-express the gene of interest. These trees can then be crossed with other cultivars, and their progeny can be tested directly for the presence of the desired allele. In marker-assisted breeding, genotyping for an allele responsible for a desired feature enables the breeder to confirm its presence in the genome, even in the absence of any visible phenotypic traits. The exclusive selection of the progeny trees carrying the desired alleles will help to both accelerate and limit the scope of field trials, as the favorable traits in trees may take years to become visible (Resende et al., 2011).
We are not limited to model species in the search for candidate genes for breeding. Another option to identify regulators of adaptive traits is to map their quantitative trait loci (QTL) in cultivated species (Grattapaglia & Kirst, 2008; Gailing et al., 2009). This approach is becoming more popular with the increasing amount of genomic information available for trees. For species in which there are sufficient genomic data available, genetic linkage maps of molecular markers, most commonly single nucleotide polymorphisms (SNPs), can be determined for individual trees. Two trees with different adaptive features can be crossed and the progeny analyzed for both phenotypic traits and molecular markers. Statistical association between marker genotypes and phenotypic traits will define QTLs behind the desired traits. For example, QTLs have been identified for stem growth and wood properties in Eucalyptus (Kirst et al., 2004; Bundock et al., 2008; Freeman et al., 2009; Gion et al., 2011), Populus (Zhang et al., 2006; Rae et al., 2008; Rae et al., 2009) and Pinus (Kaya et al., 1999; Sewell et al., 2000, 2002; González-Martínez et al., 2007). These QTLs can be used in marker-assisted breeding even without an exact knowledge about the nature of the underlying gene function.
Another method of QTL identification, genome-wide association (GWA) mapping, is gradually becoming available for tree species. In this approach, phenotypic traits are analyzed for a large population of unrelated tree individuals which are genotyped for markers spanning the entire genomic sequence (Gailing et al., 2009). As the predominantly outcrossing tree populations are characterized by a high level of individual heterozygosity, a very large amount of markers will be needed to associate any loci in tree genomes with particular phenotypic traits (Gailing et al., 2009). However, with the development of dense genetic marker maps (Grattapaglia et al., 2011) and the improvement of sequencing technologies, this approach will provide new possibilities for tree breeding in the future.
We thank Sedeer El-Showk for correcting the English in the manuscript.