Tree Physiology 35, 1000–1006 doi:10.1093/treephys/tpv050 Letter José Antonio Cabezas1,2,†, Santiago C. González-Martínez1,3,4,†, Carmen Collada2,5,†, María Angeles Guevara1,2, Christophe Boury3,4, Nuria de María1,2, Emmanuelle Eveno3,4, Ismael Aranda1,2, Pauline H. Garnier-Géré3,4, Jean Brach3,4, Ricardo Alía1,2, Christophe Plomion3,4 and María Teresa Cervera1,2,6 of Forest Ecology and Genetics, INIA-CIFOR, 28040 Madrid, Spain; 2Unidad Mixta de Genómica y Ecofisiología Forestal, INIA/UPM, 28040 Madrid, Spain; UMR1202 BIOGECO, F-33610 Cestas, France; 4University of Bordeaux, UMR1202 BIOGECO, F-33170 Talence, France; 5Departamento de Biotecnología, ETSIM, Ciudad Universitaria s/n 28040 Madrid, Spain; 6Corresponding author (cervera@inia.es) 1Department 3INRA, Received December 1, 2014; accepted April 24, 2015; published online June 20, 2015; handling Editor Ron Sederoff We have carried out a candidate-gene-based association genetic study in Pinus pinaster Aiton and evaluated the predictive performance for genetic merit gain of the most significantly associated genes and single nucleotide polymorphisms (SNPs). We used a second generation 384-SNP array enriched with candidate genes for growth and wood properties to genotype mother trees collected in 20 natural populations covering most of the European distribution of the species. Phenotypic data for total height, polycyclism, root-collar diameter and biomass were obtained from a replicated provenance-progeny trial located in two sites with contrasting environments (Atlantic vs Mediterranean climate). General linear models identified strong associations between growth traits (total height and polycyclism) and four SNPs from the korrigan candidate gene, after multiple testing corrections using false discovery rate. The combined genomic breeding value predictions assessed for the four associated korrigan SNPs by ridge regression-best linear unbiased prediction (RR-BLUP) and cross-validation accounted for up to 8 and 15% of the phenotypic variance for height and polycyclic growth, respectively, and did not improve adding SNPs from other growth-related candidate genes. For root-collar diameter and total biomass, they accounted for 1.6 and 1.1% of the phenotypic variance, respectively, but increased to 15 and 4.1% when other SNPs from lp3.1, lp3.3 and cad were included in RR-BLUP models. These results point towards a desirable integration of candidate-gene studies as a means to pre-select relevant markers, and aid genomic selection in maritime pine breeding programs. Keywords: association genetics, candidate gene, genomic selection, maritime pine, SNP, wood formation. Global demand for timber and fibers derived from wood is continuously increasing over the years and it has been estimated that wood production from planted forests may increase considerably during the next decades (Carle and Holmgren 2008). Consequently, interest for breeding programs has revived in many forest tree species, including maritime pine (Pinus pinaster Aiton), for which biomass production and stem form are the †These main breeding targets. Maritime pine is the most important source of softwood in south-western Europe. Genetic improvement of pines is costly and time-consuming, since a single breeding cycle generally extends for over two decades (White and Carson 2004), as most traits cannot be accurately evaluated at early stages. Genomic selection (GS) holds great promise to accelerate tree breeding, by shortening breeding cycles authors contributed equally to this work. © The Author 2015. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com Downloaded from https://academic.oup.com/treephys/article/35/9/1000/1652124 by guest on 10 June 2022 Nucleotide polymorphisms in a pine ortholog of the Arabidopsis degrading enzyme cellulase KORRIGAN are associated with early growth performance in Pinus pinaster Association genetics for growth in maritime pine 1001 genes were selected from nucleotide diversity (Le Dantec et al. 2004, Pot et al. 2005, Eveno et al. 2008, Grivet et al. 2011) and gene expression (Dubos et al. 2003, Dubos and Plomion 2003, Le Provost et al. 2003) studies in P. pinaster and in other conifers (Lorenz et al. 2011), and positional candidate genes co-localizing with trait-QTLs (Chagné et al. 2003, Markussen et al. 2003, Pot et al. 2006). Studied trees were collected in 20 natural populations covering most of the distribution of the species (see Figure S1 available as Supplementary Data at Tree Physiology Online). Maritime pine survived the last glaciations in multiple refugia and several extant gene pools have been described in the species using neutral molecular markers (Bucci et al. 2007, Santos-del-Blanco et al. 2012, Jaramillo-Correa et al. 2015). Although population stratification leads to confounding effects in association analysis, increasing false-positive rates, these effects can be corrected incorporating the population structure to the model. Following Yu et al. (2006), we used STRUCTURE software (Pritchard et al. 2000) to detect the number of differentiated gene pools (totalling six, in agreement with the studies cited above) and compute the individual ancestry proportions allocated to each of them (i.e., the Q matrix) based on the 302 successfully genotyped SNPs (see Table S1 available as Supplementary Data at Tree Physiology Online). Phenotypic data were obtained from a replicated provenanceprogeny trial located in two sites with contrasting environments: Cestas (44°44′N, 00°46′W), a humid site in south-western France; and Cálcena (41°3′N, 01°43′W), a very dry site in north-eastern Spain (see Figure S1 available as Supplementary Data at Tree Physiology Online). Four growth-related traits were measured: total height, polycyclism, root-collar diameter and total dry biomass (see Materials and methods available as Supplementary Data at Tree Physiology Online). Best linear unbiased predictors (BLUPs) were computed for each trait using REML (see Table S3 available as Supplementary Data at Tree Physiology Online). Genetic correlations for growth traits across sites were not significant (see Table S4 available as Supplementary Data at Tree Physiology Online), thus they were analyzed separately for each provenance-progeny trial (as in González-Martínez et al. 2008). Single-marker association analyses were based on general linear models (GLMs) using STRUCTURE Q matrix as covariate to correct for population structure and permutation (106 permutations) to obtain unbiased P-values. Significant associations for growth traits after multiple testing corrections using false discovery rate (FDR) were only found in the humid site (Cestas), probably due to extreme environmental stress (expressed also by high mortality, ∼42%) in the dry site (Table 1). Four silent SNPs (two synonymous and two intronic) from one candidate gene, korrigan, had q-values (i.e., equivalent to significance values corrected for multiple testing) below 0.05 for height and polycyclic growth, each explaining 3–4% of the phenotypic variance (Table 1 and Figure 1). These SNPs were not in Hardy–Weinberg equilibrium (HWE) but it has been shown that HWE deviations Tree Physiology Online at http://www.treephys.oxfordjournals.org Downloaded from https://academic.oup.com/treephys/article/35/9/1000/1652124 by guest on 10 June 2022 and increasing selection accuracy (reviewed by Isik 2014). Early identification of individuals carrying the desired allele combinations would allow breeders to grow larger breeding populations, decrease maintenance and evaluation costs and select elite genotypes to be crossed in the next breeding cycle before they express the desired phenotype. Recent years have brought spectacular progress in the development of genotyping technologies for the identification of single nucleotide polymorphisms (SNPs). However, the implementation of genome-wide approaches to capture linkage disequilibrium (LD) between markers and quantitative trait nucleotides in natural or breeding populations of these species has been hampered by the combined effect of two factors. On one hand, by the rapid decay of LD typically found in most genes studied so far in forest trees (in the order of a few hundreds of base pairs, reviewed by Thavamanikumar et al. 2013). On the other hand, by the large size of conifer genomes, usually >10 Gb C−1 (reviewed in MacKay et al. 2012). Thus, association genetics and genomic prediction approaches in conifers have so far only relied on a limited set of SNPs (from a few dozen selected in candidate genes to a few thousand resulting from wider survey of polymorphisms within genes). Recently, proof-of-concept on GS carried out with low population sizes in pine (Resende et al. 2012b), spruce (Beaulieu et al. 2014) and eucalyptus (Resende et al. 2012) have proved to be successful for predicting breeding values. In this context, it has been suggested that pre-selection of markers based on prior information in terms of positional, expressional and functional candidate genes as well as knowledge about gene networks to take non-additive effects into account, could help improve the predictive ability of statistical models (Schulz-Streeck et al. 2011, Resende et al. 2012b). In the last decade, family-based QTL mapping has identified a few loci significantly associated with growth traits in maritime pine, albeit controlling small proportions of the total phenotypic variation (Brendel et al. 2002, Chagné et al. 2003, Markussen et al. 2003, Pot et al. 2006). To date, only two association studies have been developed for these traits although none evaluating a wide-range sample of the species [see Lepoittevin et al. (2012) for Aquitaine breeding population and Budde et al. (2014) for natural populations in eastern Spain]. In the path towards the development of genomic tools and strategies to complement current and future breeding programs for growth traits and biomass production in maritime pine, we carried out a candidate-gene-based association genetic study in this species. Furthermore, we evaluated and tested the predictive performance for genetic merit gain of the most significantly associated genes and SNPs. We used a 384-SNP array (Illumina BeadXpress, San Diego, USA; conversion rate of 78.65%), enriched with candidate genes for growth and wood properties (see Table S1 available as Supplementary Data at Tree Physiology Online) to genotype a collection of 394 trees (see Table S2 available as Supplementary Data at Tree Physiology Online). The 1002 Cabezas et al. Table 1. Single-marker and haplogroup association for growth-related traits. Only significant SNPs after multiple testing corrections using FDR (q-values <0.05) are reported. All significant SNPs were found in the korrigan gene. Alleles contributing the favorable effect at each locus are indicated in bold. Marker names (with ‘m’ prefix) and linkage groups (LG) as reported in Chancerel et al. (2011). Haplogroups as defined in Figure 1. MAF, minimum allele frequency; N, sample size. SNP ID/ haplogroup SNP motif 546 987 1395 1499 C/G T/C A/G A/G SNP site annotation nc syn syn nc LG MAF N Marker effects Polycyclic growth Height at age 3 years 12F 0.47 0.47 0.45 0.47 0.46 322 322 322 321 322 F P R2 17.839 17.839 19.142 18.229 17.834 4.20E−05 4.20E−05 1.20E−05 3.00E−05 4.59E−08 0.038 0.038 0.041 0.039 0.038 F P R2 17.349 17.349 19.864 18.031 17.346 1.20E−05 1.20E−05 2.00E−06 6.00E−06 7.13E−08 0.031 0.031 0.035 0.032 0.031 Figure 1. Schematic representation of the korrigan gene showing the positions of genotyped SNPs, SNPs associated with growth traits (following Table 1) and the five identified haplotypes (A–E). A horizontal line indicates the two haplogroups formed by the four SNPs associated with growth. have little effect on the overall rate of false positives (Chan et al. 2008). Although all four SNPs were almost in full LD (see Figure S2 available as Supplementary Data at Tree Physiology Online), best associations were found with korrigan-1395 SNP for both, polycyclic shoot growth (q-value = 0.0045) and height (q-value = 0.0136). Five korrigan haplotypes (named A–E, see Figure 1) could be identified by studying the phase relationships among the 12 SNPs genotyped in this gene (including the four significant ones), using PHASE v2.1 (Stephens et al. 2001). These same five haplotypes have been previously detected by Pot et al. (2005) using Sanger sequencing of haploid tissue (i.e., megagametophytes). The four significantly associated SNPs differentiate two haplogroups, one of them combining the alleles contributing the favorable effect on three very similar haplotypes. An association test developed using the same previous model but based on haplogroups led to highly significant results after multiple testing corrections by FDR (Table 1, Figure S3 available as Supplementary Data at Tree Physiology Online), with similar estimated effects to individual SNPs. We further investigated the utility of the detected associations for breeding purposes using the studied sample as a training population for genomic prediction. The very stringent corrections used in the association analyses to remove false positives (due to population structure, multiple testing issues, etc.) would have likely removed a substantial number of true-positives Tree Physiology Volume 35, 2015 (Atwell et al. 2010). Thus, for each targeted trait, SNPs from well-known functional candidate genes with significant P-values but with q-values >0.05 were also considered, if by doing so prediction accuracy increased. Genomic breeding value predictions (GEBVs) were estimated by ridge regression-best linear unbiased prediction (RR-BLUP), using the R-based package rrBLUP (Endelman 2011) (Figure 2). Regarding height and polycyclic shoot growth, the combined GEBVs for the four korrigan SNPs associated with these traits were similar to those from models including all genotyped korrigan SNPs (12 SNPs), and accounted for ∼8 and ∼15% of height and polycyclic growth phenotypic variance, respectively. Adding SNPs from other growth-related candidate genes did not improve the models. Regarding root-collar diameter and total biomass, the four korrigan SNPs accounted for only 1.6 and 1.1% of the phenotypic variance. However, these values increased to 15 and 4.1% when other SNPs from korrigan and those from lp3.1, lp3.3 and cad (three other candidate genes for growth and wood properties) were included in RR-BLUP models (most complete model included a total of 19 SNPs: 12 from korrigan, two from lp3.1, two from lp3.3 and three from cad). The predictive power of these SNPs for height, polycyclic growth and root-collar diameter were confirmed using a threefold cross-validation strategy (see Table S5 available as Supplementary Data at Tree Physiology Online). For total biomass, however, the predictive power was Downloaded from https://academic.oup.com/treephys/article/35/9/1000/1652124 by guest on 10 June 2022 snp203 m734 m727 snp236 haplogroup SNP position in GenBank accession JN013967 Association genetics for growth in maritime pine 1003 found to be variable, as expected from the low overall variance explained for this trait. Our association analysis allowed confirmation of previous results from QTL mapping experiments in maritime pine and expression analyses in this and other conifer species. The molecular function of korrigan is not fully known, but it is a membrane-bound endo-1,4β-d-glucanase involved in cellulose biosynthesis that appears to be part of the primary cell wall cellulose synthase complex in the plasma membrane [reviewed by Molhoj (2002), Somerville (2006), Taylor (2008), Vain et al. (2014)]. In Arabidopsis thaliana (L.) Heynh, mutations in KORRIGAN1 have been shown to impair cellulose biosynthesis, resulting in defective cell elongation and dwarfism (Vain et al. 2014). Population genetic analyses pointed to this gene as a potential target of natural and/or artificial selection in P. pinaster and Pinus radiata D. Don (Pot et al. 2005). Moreover, our results support the co-localization of korrigan with a wood-related QTL in a family-based P. pinaster mapping pedigree (Pot et al. 2006). Cad (cinnamyl alcohol dehydrogenase) is a key enzyme in the lignin biosynthesis pathway. It catalyzes the final step in the synthesis of monolignols by converting cinnamaldehydes to cinnamyl alcohols before oxidative polymerization in the cell wall. This gene presents natural mutations that produce abnormal lignin (Ralph et al. 1997). In the Arabidopsis double mutant Atcad4/Atcad5, the content of lignin in the stem is reduced by 60% with respect to the wild type and the level of coniferyl and sinapyl alcohol is reduced by 94% (Sibout et al. 2005). A 5% reduction of lignin was also detected in a Pinus taeda L. natural cad mutant (Ralph et al. 1997). However, loss of function or down-regulation of cad did not lead to gross morphological variations in Nicotiana tabacum L., Medicago truncatula Gaertn., Populus or Pinus (MacKay et al. 1995, Lapierre et al. 1999, 2004, Kim et al. 2003, Ralph et al. 2008, Zhao et al. 2013). Rather, they resulted in a red coloration of xylem tissue because of enrichment in coniferyl aldehyde and sinapyl aldehyde. The associations found in our study between growth-related variables and cad allelic variation in maritime pine are in agreement with those found in other tree species, such as Neolamarckia cadamba Robx. with wood density (Tchin et al. 2012) or P. taeda with wood specific gravity (involving various non-synonymous mutations; González-Martínez et al. 2007). Lp3 is a small gene family encoding for nuclear water-deficitinduced proteins highly homologus to asr (abcisic acid stress Tree Physiology Online at http://www.treephys.oxfordjournals.org Downloaded from https://academic.oup.com/treephys/article/35/9/1000/1652124 by guest on 10 June 2022 Figure 2. Correlation of GEBVs based on RR-BLUP and BLUPs obtained in the humid site (Cestas) for all data combined (see Table S5 available as Supplementary Data at Tree Physiology Online for a threefold cross-validation strategy showing similar predictive power). (a) 12 SNPs from korrigan gene and height at age 3 years; (b) 12 SNPs from korrigan gene and polycyclic growth; (c) 19 SNPs from korrigan (12 SNPs), lp3.1 (2 SNPs), lp3.3 (2 SNPs) and cad (3 SNPs) genes and root-collar diameter; (d) 19 SNPs from korrigan (12 SNPs), lp3.1 (2 SNPs), lp3.3 (2 SNPs) and cad (3 SNPs) genes and total biomass. Linear trends are also shown. 1004 Cabezas et al. Tree Physiology Volume 35, 2015 15 and 4.1% for root-collar diameter and total biomass when using 19 SNPs from four genes. A cross-validation strategy showed robust predictive power for all traits except biomass. These results point towards a desirable integration of candidategene studies (as a means to pre-select relevant marker) and the possible use of genomic prediction for maritime pine breeding in the near future. Supplementary data Supplementary data for this article are available at Tree Physiology Online. Acknowledgments The authors thank the experimental Units of INRA Pierroton and INIA for trial establishment and trait measurements. Experimental data from Cálcena used in this research is part of the Spanish Network of Genetic Trials (GENFORED, http://www.genfored. es). We thank all persons and institutions linked to the establishment and maintenance of field trials used in this study. Conflict of interest None declared. Funding This study was supported by funding from the Treesnips European Union project (no. 836501) and the BOOST-SNP French project (no. 07PFTV002). References Atwell S, Huang YS, Vilhjálmsson BJ et al. (2010) Genome-wide association study of 107 phenotypes in Arabidopsis thaliana inbred lines. Nature 465:627–631. Beaulieu J, Doerksen T, Clément S, MacKay J, Bousquet J (2014) Accuracy of genomic selection models in a large population of openpollinated families in white spruce. Heredity 113:343–352. Brendel O, Pot D, Plomion C, Rozenberg P, Guehl JM (2002) Genetic parameters and QTL analysis of δ13C and ring width in maritime pine. Plant Cell Environ 25:945–953. Bucci G, González-Martínez SC, Le Provost G, Plomion C, Ribeiro MM, Sebastiani F, Alía R, Vendramin GG (2007) Range-wide phylogeography and gene zones in Pinus pinaster Ait. revealed by chloroplast microsatellite markers. Mol Ecol 16:2137–2153. Budde KB, Heuertz M, Hernández-Serrano A, Pausas JG, Vendramin GG, Verdú M, González-Martínez SC (2014) In situ genetic association for serotiny, a fire-related trait, in Mediterranean maritime pine (Pinus pinaster). New Phytol 201:230–241. Canales J, Bautista R, Label P et al. (2014) De novo assembly of maritime pine transcriptome: implications for forest breeding and biotechnology. Plant Biotechnol J 12:286–299. Carle J, Holmgren P (2008) Wood from planted forests. A Global Outlook 2005–2030. For Prod J 58:6–18. Downloaded from https://academic.oup.com/treephys/article/35/9/1000/1652124 by guest on 10 June 2022 ripening) proteins (Padmanabhan et al. 1997, Wang et al. 2003), which are transcription factors whose likely targets are hexose transporters and abscicic acid (ABA) responsive genes. Differential expression of lp3 family genes has been related with drought stress in P. taeda (Lorenz et al. 2006, 2011), cold acclimation in Pinus sylvestris L. (Joosen et al. 2006) and xylem formation in P. pinaster (Le Provost et al. 2003, Paiva et al. 2008). Moreover, lp3 genes showed population signatures of selection in P. pinaster, pointing to a possible role in its response to drought (Eveno et al. 2008). Finally, significant associations between lp3 genes and latewood density have been identified in P. taeda (González-Martínez et al. 2007), suggesting their involvement in pine growth. The implementation of GS strategies has been proposed as a way to improve the efficiency of forest tree breeding by reducing the need for expensive field testing, shortening the breeding cycle and achieving more realized genetic gain per unit time and cost (Harfouche et al. 2012, Zapata-Valenzuela et al. 2012). Genomic selection allows obtaining predictions for the breeding values of each individual based on the summed effects of a panel of molecular markers in LD with the target traits. It has been estimated that in P. taeda the use of GS approaches combined with top-grafting, which drastically decrease the long juvenile phase from eight or more years to only one, could reduce the clonal testing cycle from 12–20 years to 4–7.5 years (Resende et al. 2012a, Zapata-Valenzuela et al. 2012, Westbrook et al. 2013). The recent availability of the first genome sequence for a pine species (P. taeda; Neale et al. 2014) and the assembly of a reference transcriptome for maritime pine (Canales et al. 2014) promise the generation of a very large number of molecular markers as already demonstrated by Chancerel et al. (2013). Even then, the huge size of the genomes of these species and the short extent of their LD will still make it unaffordable for quite some time to development studies covering complete conifer genomes, highlighting the utility of candidate-gene approaches such as the one developed here. Moreover, it has been suggested that pre-selection of SNPs can increase accuracy in genomic prediction modeling by reducing model over parameterization (Schulz-Streeck et al. 2011, Resende et al. 2012b). In this context, the use of a smaller set of markers (based on a first step of association testing) to predict the genetic merit of individuals could be an attractive strategy, since it would reduce the cost of genotyping and also simplify analytical approaches for routine GS analysis (Zapata-Valenzuela et al. 2012). Westbrook et al. (2013), estimating genomic predictions for oleoresin flow in P. taeda, found that models that included only the 20–30 SNPs most significantly associated with the trait were able to predict additive genetic variation with greater accuracy than models with either randomly selected SNPs or all 4854 polymorphic loci scored. In our work, using only four SNPs located in one gene, GEBVs accounted for up to 8 and 15% of the height and polycyclic growth phenotypic variance, and up to Association genetics for growth in maritime pine 1005 Pinus pinaster Aquitaine breeding population. Tree Genet Genomes 8:113–126. Le Provost G, Paiva J, Pot D, Brach J, Plomion C (2003) Seasonal variation in transcript accumulation in wood-forming tissues of maritime pine (Pinus pinaster Ait.) with emphasis on a cell wall glycine-rich protein. Planta 217:820–830. Lorenz WW, Sun F, Liang C, Kolychev D, Wang H, Zhao X, CordonnierPratt MM, Pratt LH, Dean JFD (2006) Water stress-responsive genes in loblolly pine (Pinus taeda) roots identified by analyses of expressed sequence tag libraries. Tree Physiol 26:1–16. Lorenz WW, Alba R, Yu YS, Bordeaux JM, Simões M, Dean JFD (2011) Microarray analysis and scale-free gene networks identify candidate regulators in drought-stressed roots of loblolly pine (P. taeda L.). BMC Genomics 12:264. doi:10.1186/1471-2164-12-264 MacKay JJ, Liu W, Whetten R, Sederoff RR, O’Malley DM (1995) Genetic analysis of cinnamyl alcohol dehydrogenase in loblolly pine: single gene inheritance, molecular characterization and evolution. Mol Gen Genet 247:537–545. Mackay J, Dean JFD, Plomion C et al (2012) Towards decoding the conifer giga-genome. Plant Mol Biol 80:555–569. Markussen T, Fladung M, Achere V et al. (2003) Identification of QTLs controlling growth, chemical and physical wood property traits in Pinus pinaster (Ait.). Silvae Genet 52:8–15. Molhoj M (2002) Towards understanding the role of membrane-bound endo-beta-1,4-glucanases in cellulose biosynthesis. Plant Cell Physiol 43:1399–1406. Neale DB, Wegrzyn JL, Stevens KA et al. (2014) Decoding the massive genome of loblolly pine using haploid DNA and novel assembly strategies. Genome Biol 15:R59. doi:10.1186/gb-2014-15-3-r59 Padmanabhan V, Dias DM, Newton RJ (1997) Expression analysis of a gene family in loblolly pine (Pinus taeda L.) induced by water deficit stress. Plant Mol Biol 35:801–807. Paiva JAP, Garcés M, Alves A et al. (2008) Molecular and phenotypic profiling from the base to the crown in maritime pine wood-forming tissue. New Phytol 178:283–301. Pot D, McMillan L, Echt C, Le Provost G, Garnier-Géré P, Cato S, Plomion C (2005) Nucleotide variation in genes involved in wood formation in two pine species. New Phytol 167:101–112. Pot D, Rodrigues JC, Rozenberg P, Chantre G, Tibbits J, Cahalan C, Pichavant F, Plomion C (2006) QTLs and candidate genes for wood properties in maritime pine (Pinus pinaster Ait.). Tree Genet Genomes 2:10–24. Pritchard JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics 155: 945–959. Ralph J, MacKay JJ, Hatfield RD, O’Malley DM, Whetten RW, Sederoff RR (1997) Abnormal lignin in a loblolly pine mutant. Science 277:235–239. Ralph J, Kim H, Lu F et al. (2008) Identification of the structure and origin of a thioacidolysis marker compound for ferulic acid incorporation into angiosperm lignins (and an indicator for cinnamoyl CoA reductase deficiency). Plant J 53:368–379. Resende MDV, Resende MFR, Sansaloni CP et al. (2012) Genomic selection for growth and wood quality in Eucalyptus: capturing the missing heritability and accelerating breeding for complex traits in forest trees. New Phytol 194:116–128. Resende MFR Jr, Muñoz P, Acosta JJ, Peter GF, Davis JM, Grattapaglia D, Resende MDV, Kirst M (2012a) Accelerating the domestication of trees using genomic selection: accuracy of prediction models across ages and environments. New Phytol 193:617–624. Resende MFR Jr, Muñoz P, Resende MDV et al. (2012b) Accuracy of genomic selection methods in a standard data set of loblolly pine (Pinus taeda L.). Genetics 190:1503–1510. Santos-del-Blanco L, Climent J, González-Martínez SC, Pannell JR (2012) Genetic differentiation for size at first reproduction through male Tree Physiology Online at http://www.treephys.oxfordjournals.org Downloaded from https://academic.oup.com/treephys/article/35/9/1000/1652124 by guest on 10 June 2022 Chagné D, Brown G, Lalanne C, Madur D, Pot D, Neale D, Plomion C (2003) Comparative genome and QTL mapping between maritime and loblolly pines. Mol Breed 12:185–195. Chan EKF, Hawken R, Reverter A (2008) The combined effect of SNPmarker and phenotype attributes in genome-wide association studies. Anim Genet 40:149–156. Chancerel E, Lepoittevin C, Le Provost G et al. (2011) Development and implementation of a highly-multiplexed SNP array for genetic mapping in maritime pine and comparative mapping with loblolly pine. BMC Genomics 12:368. doi:10.1186/1471-2164-12-368 Chancerel E, Lamy JB, Lesur I et al. (2013) High-density linkage mapping in a pine tree reveals a genomic region associated with inbreeding depression and provides clues to the extent and distribution of meiotic recombination. BMC Biol 11:50. doi:10.1186/1741-7007-11-50 Dubos C, Plomion C (2003) Identification of water-deficit responsive genes in maritime pine (Pinus pinaster Ait.) roots. Plant Mol Biol 51:249–262. Dubos C, Le Provost G, Pot D, Salin F, Lalane C, Madur D, Frigerio JM, Plomion C (2003) Identification and characterization of water-stressresponsive genes in hydroponically grown maritime pine (Pinus pinaster) seedlings. Tree Physiol 23:169–179. Endelman JB (2011) Ridge regression and other kernels for genomic selection with R package rrBLUP. Plant Genome 4:250–255. Eveno E, Collada C, Guevara MA et al. (2008) Contrasting patterns of selection at Pinus pinaster Ait. Drought stress candidate genes as revealed by genetic differentiation analyses. Mol Biol Evol 25:417–437. González-Martínez SC, Wheeler NC, Ersoz E, Nelson CD, Neale DB (2007) Association genetics in Pinus taeda L. I. Wood property traits. Genetics 175:399–409. González-Martínez SC, Huber D, Ersoz E, Davis JM, Neale DB (2008) Association genetics in Pinus taeda L. II. Carbon isotope discrimination. Heredity 101:19–26. Grivet D, Sebastiani F, Alía R, Bataillon T, Torre S, Zabal-Aguirre M, Vendramin GG, González-Martínez SC (2011) Molecular footprints of local adaptation in two Mediterranean conifers. Mol Biol Evol 28:101–116. Harfouche A, Meilan R, Kirst M, Morgante M, Boerjan W, Sabatti M, Scarascia Mugnozza G (2012) Accelerating the domestication of forest trees in a changing world. Trends Plant Sci 17:64–72. Isik F (2014) Genomic selection in forest tree breeding: The concept and an outlook to the future. New For 45:379–401. Jaramillo-Correa JP, Rodríguez-Quilón I, Grivet D et al. (2015) Molecular proxies for climate maladaptation in a long-lived tree (Pinus pinaster Aiton, Pinaceae). Genetics 199:793–807. Joosen RVL, Lammers M, Balk PA, Brønnum P, Konings MCJM, Perks M, Stattin E, van Wordragen MF, van der Geest AH (2006) Correlating gene expression to physiological parameters and environmental conditions during cold acclimation of Pinus sylvestris, identification of molecular markers using cDNA microarrays. Tree Physiol 26:1297–1313. Kim H, Ralph J, Lu F et al. (2003) NMR analysis of lignins in CAD-deficient plants. Part 1. Incorporation of hydroxycinnamaldehydes and hydroxybenzaldehydes into lignins. Org Biomol Chem 1:268–281. Lapierre C, Pollet B, Petit-Conil M et al. (1999) Structural alterations of lignins in transgenic poplars with depressed cinnamyl alcohol dehydrogenase or caffeic acid O-methyltransferase activity have an opposite impact on the efficiency of industrial kraft pulping. Plant Physiol 119:153–164. Lapierre C, Pilate G, Pollet B, Mila I, Leplé JC, Jouanin L, Kim H, Ralph J (2004) Signatures of cinnamyl alcohol dehydrogenase deficiency in poplar lignins. Phytochemistry 65:313–321. Le Dantec L, Chagné D, Pot D et al. (2004) Automated SNP detection in expressed sequence tags: statistical considerations and application to maritime pine sequences. Plant Mol Biol 54:461–470. Lepoittevin C, Harvengt L, Plomion C, Garnier-Géré P (2012) Association mapping for growth, straightness and wood chemistry traits in the 1006 Cabezas et al. Tree Physiology Volume 35, 2015 Vain T, Crowell EF, Timpano H et al. (2014) The cellulase KORRIGAN is part of the cellulose synthase complex. Plant Physiol 165:1521–1532. Wang JT, Gould JH, Padmanabhan V, Newton RJ (2003) Analysis and localization of the water-deficit stress-induced gene (lp3). J Plant Growth Regul 21:469–478. Westbrook JW, Resende MFR, Munoz P et al. (2013) Association genetics of oleoresin flow in loblolly pine: discovering genes and predicting phenotype for improved resistance to bark beetles and bioenergy potential. New Phytol 199:89–100. White TL, Carson MJ (2004) Breeding programs of conifers. In: White TL, Carson MJ (eds) Plantation forest biotechnology for the 21st century. Research Signpost, Trivandrum, India, pp 61–85. Yu J, Pressoir G, Briggs WH et al. (2006) A unified mixed-model method for association mapping that accounts for multiple levels of relatedness. Nat Genet 38:203–208. Zapata-Valenzuela J, Isik F, Maltecca C, Wegrzyn J, Neale D, McKeand S, Whetten R (2012) SNP markers trace familial linkages in a cloned population of Pinus taeda—prospects for genomic selection. Tree Genet Genomes 8:1307–1318. Zhao Q, Tobimatsu Y, Zhou R et al. (2013) Loss of function of cinnamyl alcohol dehydrogenase 1 leads to unconventional lignin and a temperature-sensitive growth defect in Medicago truncatula. Proc Natl Acad Sci USA 110:13660–13665. Downloaded from https://academic.oup.com/treephys/article/35/9/1000/1652124 by guest on 10 June 2022 versus female functions in the widespread Mediterranean tree Pinus pinaster. Ann Bot 110:1449–1460. Schulz-Streeck T, Ogutu JO, Piepho HP (2011) Pre-selection of markers for genomic selection. BMC Proc 5(Suppl 3):S12. doi:10.1186/17536561-5-S3-S12 Sibout R, Eudes A, Mouille G, Pollet B, Lapierre C, Jouanin L, Séguin A (2005) Cinnamyl alcohol dehydrogenase-C and -D are the primary genes involved in lignin biosynthesis in the floral stem of Arabidopsis. Plant Cell 17:2059–2076. Somerville C (2006) Cellulose synthesis in higher plants. Annu Rev Cell Dev Biol 22:53–78. Stephens M, Smith NJ, Donnelly P (2001) A new statistical method for haplotype reconstruction from population data. Am J Hum Genet 68:978–989. Taylor NG (2008) Cellulose biosynthesis and deposition in higher plants. New Phytol 178:239–252. Tchin BL, Ho WS, Pang SL, Ismail J (2012) Association genetics of the cinnamyl alcohol dehydrogenase (CAD) and cinnamate 4-hydroxylase (C4H) genes with basic wood density in Neolamarckia Cadamba. Biotechnology 11:307–317. Thavamanikumar S, Southerton SG, Bossinger G, Thumma BR (2013) Dissection of complex traits in forest trees: opportunities for markerassisted selection. Tree Genet Genomes 9:627–639.