Fangqun Ouyang?Jianwei Ma?Sanping An?Junhui Wang?Yuhui Weng

Introduction

Pinus tabuliformis Carr.is an important commercial species in north and northwest China,where about 2–3 hundred million seedlings are planted annually.Due to its economic and ecological importance(Fan et al.2006),genetic selection for this species has been practiced since the early 1970 s to improve growth rate and wood density(Chen et al.2006;Li et al.1998;Qin et al.2000;Wang et al.1992;Xu 1991).Significant gains have been realized in both traits(Fan et al.2006;Liu 2007).Clonal seed orchards of P.tabuliformis in China are used to massproduce seeds of high genetic quality,and tremendous achievements have been attained in seed lot production,quality evaluation,and reproductive system assessment(Li et al.2011,2012;Niu et al.2013;Yuan et al.2014,2016).The wood of P.tabuliformis has been used mainly for paper-making.There is limited knowledge regarding genetic variation in tree tracheid characteristics,despite the fact that they account for more than 90%of timber volume and despite their great impact on the properties of pulp and paper products(Hannrup and Danell 2001).Also,selection for tracheid dimensions has rarely been incorporated into tree-breeding programs.

Tracheids are the structural elements of trees and they influence the quality of the main wood-based products such as pulp and paper.Tracheid length,diameter and wall thickness not only affect tensile and tear strength but also impact the optical and printing properties of paper(Macdonald and Hubert 2002).Tracheid dimensions can be moderately to highly heritable(Cown et al.2004;Fries 2012;Hannrup et al.2004).As a consequence of the efforts to increase the productivity of the species,there is a trend to harvest increasingly large proportions of P.tabuliformis from fast-growing plantations.Such wood is characterized by wide annual rings and more earlywood(EW)than latewood(LW).Phenotypically,EW has thinner cell walls,longer tracheids and wider radial diameters than the corresponding LW,but the two tissue types have similar tracheid tangential diameters(Brandstrom 2001;Havimo et al.2009).Selecting for faster height growth by Scots pine(Pinus sylvestris)resulted in a slight increase in fibre width but a slight decrease in relative fibre length(Fries 2012).High negative genetic correlations between growth traits and wood properties suggest incorporating multiple traits selection are recommended for the future Scots pine breeding programs(Zhou et al.2014).Most breeding programs have included wood density as a substitute.In conifers,however,wood density is a complex trait that is determined partly by dimensions of tracheids(Roderick and Berry 2001).For P.tabuliformis,virtually nothing is known of the genetic variation in tracheid traits and how they affect wood density.

Using data collected from a 33-year old clonal seed orchard of P.tabuliformis,the objectives of this study were to:(1)calculate genetic variation and repeatability of tracheid traits;(2)investigate relationships of tracheid traits with growth-and wood density-related traits;(3)identify the key tracheid traits that determine wood density;and(4)estimate genetic gain achievable for tracheid traits when selecting for diameter at breast height(DBH).The results obtained should be useful in predicting the response of tracheid traits when selecting for DBH,which would benefit the design of optimal breeding strategies for the advanced breeding program of P.tabuliformis.

Materials and methods

Wood samples used in the study were collected in 2012(the trees were 33-year old)from a seed orchard of P.tabuliformis that was established in 1979.The orchard was located in Tianshui,Gansu Province,China(34°07′N,105°24′E).The elevation at the orchard is 1560–1800 m,with annual rainfall of 800 mm,annual average temperature of 7–9 °C,and a frost-free period of 153 days.The brown soil is fertile and about 50 cm thick,with pH from 5.5 to 6.5.

A total of 342 clones(grafts on rootstocks),which were phenotypically selected from natural and artificial stands of P.tabuliformis is,were planted in the orchard.The seed orchard was established using a randomized complete block design of four blocks.About four ramets per clone were planted in each block of the orchard at a spacing of 3 m × 3 m.343 ramets of 53 clones(average 2–8 ramets per clone)randomly selected were sampled for this study.Individual tree wood increment cores(inner diameter 7 mm)of these 90 clones(two trees per clone per block)were collected.DBH(cm)was measured for each sampled tree.

Wood density(WD)was measured using the drainage method(Hacke et al.2001)based on the Archimedes principle:we measured the sample weight of water displacement as the sample volume(Vmax)and oven-dried mass(mo)and calculated WD(g/cm3)as the ratio between moand Vmax.Ring width(RW)and latewood width were measured using a ring width tester(Alfred J.Amsler and Co.Schaffhouse.Switzerland)with accuracy of 0.01 mm.Latewood percent(LWP)was calculated as the ratio of latewood width over the total annual ring width(Denne et al.1997;Gagen et al.2004;Meko and Baisan 2001).

Tracheid traits were studied using the nitric acid and chromic acid mixture segregation Tayler staining method(Heimsch and Tschabold 1972).After staining with safranin,tracheid length was measured using a projection microscope(XST-2;Nanjing Jiangnan Novel Optics Co.,Ltd.,Nanjing,China).Fifty tracheids of each tree were randomly selected from LW and EW of the 31-year and 32-year rings for measuring tracheid length.Average tracheid length for each tree was calculated for EW and LW,and referred to as EW average tracheid length(EW_TL)and LW average tracheid length(LW_TL).In statistical analyses we used averages across rings because in a tree breeding program it is unlikely that trees will be selected by the tracheid length of individual rings.

The wood samples were cut into serial sections,using a wood slicing machine(1400 Slip type;Leitz,Wetzlar,Germany),diagonally from pith to bark.The obtained cross-sections(10–15 μm)were first dried,and then gradually dehydrated with ethanol.Finally,the edges of the coverslip were sealed using rubber cement.The images of 50 EW and 50 LW valid tracheids which were randomly selected from annual rings of an increment core wood sample were captured with a P4 computer-aided system consisting of a Nikon 80i microscope(Nikon,Tokyo,Japan)connected to a DS-Ri1 digital photography transducer(Nikon).Measurements were taken at no fewer than two points.Cross-sectional dimensions of individual tracheids were measured using TDY-5.2 Micro-color Image Analysis Software(Beijing Tiandiyu Science and Technology Co.,Ltd.,Beijing,China).At 640×480 magnif ication,the spatial calibration system corresponded to 0.1 μm/pixel.In total,12 tracheid dimensional properties,viz.tracheid length of EW(EW_TL),tracheid length of LW(LW_TL),cell wall thickness of EW(EW_WT),cell wall thickness of LW(LW_WT),radial lumen diameter of EW (EW_R_D1),radial lumen diameter of LW(LW_R_D1), tangential lumen diameter of EW(EW_T_D1), tangential lumen diameter of LW(LW_T_D1),radial central diameter of EW(EW_R_D2),radial central diameter of LW(LW_R_D2),tangential central diameter of EW(EW_T_D2),and tangential central diameter of LW(LW_T_D2)were measured and averaged by individual tree.We also calculated four ratios:EW_TL/EW_T_D2(EW_TL/T_D2),LW_TL/LW_T_D2(LW_TL/T_D2),EW_TL/EW_R_D1(EW_TL/R_D1),LW_TL/LW_R_D1(LW_TL/R_D1)(Table 1).

The data were analyzed by individual trait using oneway analysis of variance with clone effect(C)as a random effect and e as a vector of random residual errors.Clonal repeatability was estimated based on clonal mean(=1–1/F).Simple correlation between traits was calculated as the Pearson correlation coefficient.All of the above analyses were performed using the IBM SPSS statistics 20.0 software(Kirkpatrick and Feeney 2014).

To examine the predictability of wood density by tracheid traits,a multiple regression analysis was performed using the SAS(Sas 2010)Reg Procedure with the R-square selection method for CP model-building.

Genetic gains from selection were then calculated as the average genetic value of the selected population and expressed as the percent of the overall population mean(Hodge and White 1992).

Results

Table 1 lists mean values of all measured traits.Average growth traits were 15.38 cm for DBH,8.24 m for H and 3.04 mm for RW,and the average density-related traits were 0.38 g/cm3for WD and 27.88%for LWP.Across tissues(EW and LW),the averages were 3.23 mm,8.04,24.01,24.67,32.15,and 32.75 μm,respectively,for TL,WT,R_D1,T_D1,R_D2 and T_D2.In general,EW had larger diameter values than did LW,and this was particularly evident for R_D1(266%larger)and R_D2(120%).However,LW had larger ratios than EW for WT/R_D1(467%larger)and TL/T_D2(146%larger).Also,LW had longer,although comparable,tracheid average length and thicker cell wall than EW.

Table 1 lists coefficients of genetic variation(CVG)and heritabilities.The CVGof WD(8.32%)was the lowest among the measured traits,and RW had the highest CVG(41.56%).The CVGvalues for DBH and LWP were 27.04 and 34.80%.The CVGvaried with the tracheid traits,ranging from 9.49%(LW_T_D2)to 26.03%(LW_R_D1).No consistent patterns were obvious in CVGbetween EW and LW,with three traits(TL,R_D2 and T_D2)having slightly higher values in EW while the other three traits(WT,R_D1 and T_D1)had larger values in LW.

ANOVA showed that clones significantly affected DBH,H,RW,LWP and WD(Table 2).Clones significantly affected TL in LW but not in EW.The only exception was TL,where clones significantly affected WT,R_D1,T_D1,R_D2,T_D2,and the two ratios WT/R_D1 and TL/T_D2 in EW but had no significant effects in LW.was higher in LW(0.50)than in EW(0.20)for TL,whilewas higher in EW(0.27–0.46)for other tracheid traits and the two ratios(TL/T_D2 and WT/R_D1)than in LW(0.06–0.22)(Table 1).Also,tracheid lumen diameters appeared to be more heritable in the radial direction than the tangential direction,but comparablevalues were found for lumen and central diameters.of H,DBH,RW,LWPand WD were moderately higher than tracheid traits,the range was 0.39–0.50.

Table 1 Number,Mean,standard deviation(SD),coefficient of genetic variation(CVG),clonal-mean heritability(Hof measured traits

Table 1 Number,Mean,standard deviation(SD),coefficient of genetic variation(CVG),clonal-mean heritability(Hof measured traits

aDBH diameter at breast height(cm),H height(m),NSC crown breadth in south–north(m),WEC crown breadth in east–west(m),RW ring width(mm),LWP latewood percent(%),WD wood density(g/cm3),EW_TL and LW_TL earlywood and latewood average tracheid length(mm),EW_WT and LW_WT earlywood and latewood cell wall thickness(μm),EW_R_D1 and LW_R_D1 earlywood and latewood radial lumen diameter(μm),EW_T_D1 and LW_T_D1 earlywood and latewood tagential lumen diameter(μm),EW_R_D2 and LW_R_D2 earlywood and latewood radial central diameter(μm),EW_T_D2 and LW_T_D2 earlywood and latewood tangential central diameter(μm),EW_TL/T_D2 and LW_TL/T_D2 the ratio of earlywood average tracheid length and tangential central diameter.EW_WT/R_D1 and LW_WT/R_D1 the ratio of earlywood cell wall thickness and radial lumen diameter

Traita Samples number Mean SD CVG H 2 C DBH 343 15.38 cm 3.15 27.04 0.50 H 343 8.24 m 1.13 17.16 0.42 NSC 343 4.89 m 1.04 23.93 0.25 WEC 343 4.79 m 0.99 23.95 0.31 RW 339 3.04 mm 1.00 41.56 0.45 LWP 339 27.88% 7.93 34.80 0.39 WD 343 0.38 g/m3 0.03 8.32 0.49 EW_TL 343 3.06 mm 0.38 13.51 0.20 LW_TL 343 3.40 mm 0.31 12.02 0.50 EW_WT 340 6.54 μm 0.70 12.17 0.27 LW_WT 335 9.53 μm 1.47 16.26 0.11 EW_R_D1 340 37.70 μm 4.54 15.40 0.46 LW_R_D1 341 10.31 μm 2.49 26.03 0.17 EW_T_D1 340 29.68 μm 2.86 11.23 0.31 LW_T_D1 341 19.66 μm 3.21 18.09 0.22 EW_R_D2 340 44.23 μm 4.78 13.83 0.46 LW_R_D2 341 20.07 μm 2.51 13.77 0.21 EW_T_D2 340 36.22 μm 2.92 9.69 0.37 LW_T_D2 341 29.28 μm 2.61 9.49 0.14 EW_TL/T_D2 340 69.85 11.43 19.38 0.34 LW_TL/T_D2 341 171.61 24.22 15.40 0.19 EW_WT/R_D1 340 0.18 0.03 17.66 0.31 LW_WT/R_D1 341 1.02 0.44 44.35 0.06

DBH,H,NSC,WES,and RW were significantly positively correlated with one another(Table 3),and signif icantly negatively correlated with WD both at individual and clone levels.

Pearson correlation(rP)between tracheid traits(Table 4)varied greatly,but can be summarized as follows:(i)for a given trait,rPbetween EW and LW was consistently significantly positive at individual and clone levels(except for WT at clone level);(ii)rPwas positive for WT and TL at both tissue and individual level.However,at the clone level,LW_WT and TL in both tissues were negatively correlated;(iii)all tracheid diameters,regardless of tissue type,direction or diameter,were strongly and positively correlated at individual and clone levels(EW_T_D1 and LW_R_D2 were positively correlated);and(iv)the four ratio values(EW_WT/R_D1,LW_WT/R_D1,EW_TL/T_D2 and LW_TL/T_D2)were consistently significantly positively correlated both at individual and clone levels,and were consistently significantly negatively correlated with all tracheid diameters at individual and clone levels,and consistently significantly positively correlated with WT and TL at clone level.

Correlation coefficients for tracheid traits with growthor density-related traits at individual and clone level are listed in Tables 5 and 6.All tracheid diameters,regardless of tissue type,direction or diameter,were strong positively correlated with DBH,H,NSC,WEC,RW and P,

and strong negatively correlated with WD at individual and clone levels.WD and TL were significantly negatively correlated,but WD and WT were positively correlated at individual and clone levels LWP was negatively correlated with all tracheid traits(except LW_T_D1 at the individual level and LW_WT/R_D1 at clone level),but only two and four traits were significant at individual and clone levels.With few exceptions,the four ratio values(EW_WT/R_D1,LW_WT/R_D1,EW_TL/T_D2 and LW_TL/T_D2)were consistently significantly negatively correlated with DBH,H,NSC,WEC and RW,and significantly positively correlated with WD at individual and clone levels.

Table 2 ANOVA analysis of measured traits

The most important variables for predicting WD were LW_TL,EW_WT and R_D1 in both EW and LW(r2=0.22)(Table 7).Inclusion of additional variables to these three variables did not substantially improve the model fit.

Figure 1 shows genetic gains in tracheid traits from selection for DBH.Selecting the top 10%of the clones by DBH would improve DBH growth by 12.19%(wood density would be reduced by 0.14%)and produced similar responses in EW and LW for all tracheid traits,viz.a reduction of 0.94 and 3.69%,respectively,in tracheid length,and increases in tracheid diameters(from 0.36 to 5.24%)and double wall thickness(0.07 and 0.87%,respectively).The two ratios WT/R_D1 and TL/T_D2 across tissues(EW and LW)declined by 0.59 and 4.56%,respectively.The values increased or decreased by selecting for DBH gains were higher in EW than in LW.

Discussion

The mean values of TL,WT and T_D1 were comparable to those reported in earlier studies on P.tabuliformis and other hard pines in China(Xu 1991;Zhao et al.1999).For P.tabuliformis,our results suggest that EW has thinner tracheid cell walls,shorter tracheid length,and a proportionally larger radial lumen and central diameter than LW,confirming the findings reported for other conifers(Irbe et al.2015;Loustarinen 2012).In spring,high auxin levels induce the production of large tracheids while available photosynthates for use in the thickening of their walls are limited.As summer progresses,the auxin levels decline and more photosynthates become available,leading to small tracheid dimensions paired with thick cell walls in latewood(Zimmerman and Brown 1971).However,LW had the larger ratios of TL/T_D2 and WT/R_D1.

The estimated CVGandwere moderate(Table 1),suggesting that clonal variations in tracheid traits in P.tabuliformis is are under genetic control,and that improvements in these traits can be gained from clonal selection and deployment.Irbe et al.(2015)found signi ficant genetic variations in tracheid traits among Norway spruce clones.Tracheid dimensions in LW were found to be more strongly linked to environmental factors than thosein EW(Irbe et al.2015;Lenz et al.2010;Wimmer and Grabner 2000),and radial diameter was generally more heritable than tangential diameter(Ivkovich et al.2002),both findings agreeing well with our study(Table 1).

Table 3 Pearson correlations and their significance levels among growth traits and wood density in individual and clone level

Table 4 Pearson correlations and their significance levels among tracheid traits in individual and clone level

Mixed relationships between growth and tracheid traits have been reported in the literature on pine species.Tracheid length in white spruce(Picea glauca)is negatively correlated to growth traits(Beaulieu 2003).However,tracheid length showed positive correlations with height and relative branch diameter in P.sylvestris(Hannrup et al.2000).The relationship between TL and growth traits in P.tabuliformis were varied and relied on growth traits and individual or clone level(Tables 5 and 6),but with few exceptions,the four ratio values (EW_WT/R_D1,LW_WT/R_D1,EW_TL/T_D2 and LW_TL/T_D2)were consistently significantly negative correlated to all growth traits both at individual and clone levels.Average tracheid diameters of EW and LW moderately correlated with DBH,H,and RW(Irbe et al.2015).However,in the present study,all tracheid diameters were strongly positively correlated to DBH,H and RW.

Table 5 Pearson correlations and their significance levels of tracheid traits with growth traits and wood density in individual level

Table 6 Pearson correlations and their significance levels of tracheid traits with growth traits and wood density in clone level

Wood density is considered a secondary trait for selection in the current breeding program for P.tabuliformis(Yuan 2016).Wood density is a good indicator of the yield of pulp per unit volume but not of paper quality,which depends more on tracheid traits(Roderick and Berry 2001).Wood density was positively correlated with tracheid double wall thickness,but its relationships with tracheid length and diameters were significantly negative(Tables 5 and 6).These relationships are mainly encouraging in terms of genetic selection on wood density.However,Makinen et al.(2002)reported no clear relationship between cell wall thickness and wood density in both EW and LW.Relationships between cell wall thickness andwood density in both EW and LW of Norway spruce were weak(Irbe et al.2015;Makinen et al.2002).

Table 7 Regression models for predicting wood density using tracheid traits

Fig.1 Genetic gains in average tracheid length(TL),tangential lumen diameter(T_D1),tangential central diameter(T_D2),cell wall thickness(WT),radial central diameter(R_D2),and radial lumen diameter(R_D1)of earlywood(EW)and latewood(LW),the ratio of average tracheid length and tangential central diameter(TL/T_D2)and the ratio of cell wall thickness and radial lumen diameter(WT/R_D1)of earlywood(EW)and latewood(LW)from direct selecting the top 10%clones by DBH(gain on DBH was 12.19%and wood density was-0.14%)

In the current long-term breeding programs for P.tabuliformis,growth is considered as the selection trait of prime importance(Yuan 2016).Our results suggest that selection for larger DBH would reduce tracheid length and the ratios WT/R_D1 and TL/T_D2 in both EW and LW but would moderately increase tracheid diameters(Table 3;Fig.1).Reducing TL is disadvantageous,as shorter tracheids decrease the burst,tearing and tensile strength of paper(Wimmer et al.2002)and are associated with higher microfibril angles that reduce timber strength,stiffness,and dimensional stability (Macdonald and Hubert 2002).Increased tracheid diameters would reduce tensile energy absorption(McDonough et al.2012).Our results also show that enhancing DBH would increase double wall thickness in both EW and LW(Fig.1).An increase in cell wall thickness increases the tear index and compressive strength,but decreases the tensile index and reduces light scattering(Brandstrom 2001;Havimo et al.2009).Four ratio values(EW_WT/R_D1,LW_WT/R_D1,EW_TL/T_D2 and LW_TL/T_D2)were consistently significantly positively correlated with WD(Table 5).Reduction of the ratio TL/T_D2 was not beneficial for pulping.While long tracheids produce pulps with high strength and low optical properties,low tracheid diameters increase tearing resistance of paper.Thus,paper manufactured from wood of thin walled tracheids will be dense and well-formed.Strong negative correlations between growth traits or wood density and tracheid properties suggest the future breeding programs should adopt selection for multiple traits,including economic weights(Zhou et al.2014).

Conclusions

Results of this study suggest that tracheid traits possessed wide genetic variations for P.tabuliformis.Also,selection for increased DBH resulted in increased tracheid wall thickness and diameter but decreased tracheid length and the ratio of tracheid length to diameter,all of which are unfavorable for papermaking.The current long-term breeding program for P.tabuliformis simultaneously targets improvements in growth rate and wood density.Overall,our results suggest that benefits would result from including selection for tracheid traits into the current breeding strategy.

Beaulieu J(2003)Genetic variation in tracheid length and relationships with growth and wood traits in eastern white spruce(Picea glauca).Wood Fiberence 35(4):609–616

Brandstrom J(2001)Micro-and ultrastructural aspects of Norway spruce tracheids:a review.Iawa J 22(4):333–353

Chen JZ,Li Y,Li GF,Li J,Du QP,Wang XR(2006)Study on the genetic variations of half-sib progeny testing and advanced selections of Pinus taburaeformis.J Henan Agr Univ 40:146–160

Cown D,Ball R,Riddell M(2004)Wood density and microfibril angle in 10 Pinus radiata clones:distribution and influence on product performance.NZ J For Sci 34(3):293–315

Denne MP,Rao RV,Aebischer DP(1997)Latewood density in relation to wood fibre diameter,wall thickness,and fibre and vessel percentages in Quercus Robur L.Iawa J 18(2):127–138

Fan JF,Yang PH,Guo SJ,Liu YH(2006)Research progess of genetics and improvement on Pinus tabulaeformis carriere of Shanxi province.J Northwest Sci-Tech Univ Agr For(NatSci)34:45–50

Fries A(2012)Genetic parameters,genetic gain and correlated responses in growth,fibre dimensions and wood density in a Scots pine breeding population.Ann For Sci 69(7):783–794

Gagen M,Mccarroll D,Edouard JL (2004)Latewood width,maximum density,and stable carbon isotope ratios of Pine as climate indicators in a dry subalpine environment,French Alps.Arct Antarct Alp Res 36(2):166–171

Hacke UG,Sperry JS,Pockman WT,Davis SD,Mcculloh KA(2001)Trends in wood density and structure are linked to prevention of xylem implosion by negative pressure. Oecologia 126(4):457–461

Hannrup B,Danell o(2001)Relationships between wood density and tracheid dimensions in Pinus sylvestris L.Wood Fiberence 33(2):173–181

Hannrup B,Ekberg I,Persson A(2000)Genetic correlations among wood,growth capacity and stem traits in Pinus sylvestris.Scand J For Res 5(2):161–170

Hannrup B,Cahalan C,Chantre G,Grabner M,Karlsson B,Le Bayon I,Jones GL,Mulle U,Pereira H,Rodrigues JC,Rosner S,Rozenberg P,Wilhelmsson L,Wimmer R(2004)Genetic parameters of growth and wood quality traits in Picea abies.Scand J For Res 8(1):14–29

Havimo M,Rikala J,SirvioJ,Sipi M(2009)Tracheid cross-sectional dimensions in Scots pine(Pinus sylvestris)–distributions and comparison with Norway spruce(Picea abies).Silva Fenn 43(4):681–688

Heimsch C,Tschabold EE(1972)Xylem Studies in the Linaceae.Bot Gaz 133(3):242–253

Hodge GR,White TL(1992)Concepts of selection and gain prediction.In:Fins L,Friedman ST,Brotschol JV (eds)Handbook of quantitative forest genetics.Springer,Netherlands

Irbe I,Sable I,Noldt G,Grinfelds U,Jansons A,Treimanis A,Koch G(2015)Wood and tracheid properties of Norway spruce(Picea abies[L]Karst.)clones grown on former agricultural land in Latvia.Balt For 21(1):114–123

Ivkovich M,Namkoong G,Koshy M(2002)Genetic variation in wood properties of interior spruce.II.Tracheid characteristics.Can J Forest Res 32(12):2128–2139

Kirkpatrick LA,Feeney BC(2014)A simple guide to IBM SPSS:For Version 22.0.Belmont,Wadsworth

Lenz P,Cloutier A,MacKay J,Beaulieu J(2010)Genetic control of wood properties in Picea glauca-an analysis of trends with cambial age.Can J For Res 40(4):703–715

Li J,Li Y,Li GF,Zhang HF,Hou ZZ,Wang HY(1998)Progeny test of plus trees and advanced selection of Pinus tabulaeformis.J Beijing For Univ 20:19–24

Li W,Wang XR,Li Y(2011)Stability in and correlation between factors influencing genetic quality of seed lots in seed orchard of Pinus tabuliformis Carr.over a 12-year span.PLoS ONE 6(8):492

Li W,Wang XR,Li Y(2012)Variation in floral phenological synchronization in a clonal seed orchard of Pinus tabuliformis in Northeast of China.Silvae Genet 61(4):133–142

Liu YH(2007)Genetic variation and comprehensive selection of superior families of Pinus tabuleaformis Carr.J Central South Univ For Technol 27:21–25

Loustarinen K(2012)Tracheid wall thickness and lumen diameter in different axial and radial locations in cultivated Larix sibirica trunks.Silva Fenn 46(46):707–716

Macdonald E,Hubert J(2002)A review of the effects of silviculture on timber quality of Sitka spruce.Forestry 75(2):107–138

Makinen H,Saranpaa P,Linder S(2002)Wood-density variation of Norway spruce in relation to nutrient optimization and fibre dimensions.Can J For Res 32(2):185–194

McDonough TJ,Courchene CE,White DE,Schimleck L,Peter G(2012)Effects of loblolly pine tree age and wood properties on linerboard-grade pulp yield and sheet properties:Part 2.Tappi J 11:41–50

Meko DM,Baisan CH(2001)Pilot study of latewood-width of conifers as an indicator of variability of summer rainfall in the North American monsoon region.Int J Climatol 21(6):697–708

Niu SH,Li ZX,Yuan HW,Chen XY,Li Y,Li W(2013)Transcriptome characterisation of Pinus tabuliformis and evolution of genes in the Pinus phylogeny.BMC Genom 14:263–274

Qin CF,Li Y,Niu ZT,Li GF,Du QP,Wang HY(2000)Genetic variations of clonal traits and estimation of gamete pool composition in a clonal seed orchard of Pinus tabulaeformis.J Beijing For Univ 22:20–24

Roderick ML,Berry SL(2001)Linking wood density with tree growth and environment:a theoretical analysis based on the motion of water.New Phytol 149(3):473–485

Sas Institute,Inc(2010)SAS/STAT User’s Guide,Version 6.SAS Institute Inc,Cary,NC

Wang SW,Shen XH,Wang SM(1992)Study on variation of wood properties in Pinus tabulaeformis.J Beijing For Univ 14:87–92

Wimmer R,Grabner M(2000)A comparison of tree-ring features in Picea abies as correlated with climate.Iawa J 21(4):403–416

Wimmer R,Downes G,Evans R,Rasmussen G,French J(2002)Direct effects of wood characteristics on pulp and handsheet properties of Eucalyptus globulus.Holzforschung 56(3):244–252

Xu YM(1991)Specific gravity variation in Pinus tabulaeformis.J Huazhong Agr Univ 10:281–285

Yuan HW(2016)Key technologies and strategies for advanced-cycle genetic improvement of Pinus tabuliformis Carr.Ph.D.dissertation,Beijing Forestry University,139

Yuan HW,Li ZX,Fang P,Li W,Li Y(2014)Variation and stability in female strobili production of a first-generation clonal seed orchard of Chinese Pine(Pinus tabuliformis).Silvae Genet 63(1–2):41–47

Yuan HW,Niu SH,Zhon XQ,Du QP,Li Y,Li W(2016)Evaluation of seed production in a first-generation seed orchard of Chinese pine(Pinus tabuliformis).J Forestry Res 27(5):1–6

Zhao RJ,Guo JR,Yang PH,Xie B(1999)A study on morphological and comparison of the wood tracheid of the half-sib progency and parent of Pinus tabulaeformis Carr.J Northwest For Univ 14:6–9

Zhou H,Fries A,Wu HX(2014)High negative genetic correlations between growth traits and wood properties suggest incorporating multiple traits selection including economic weights for the future Scots pine breeding programs.Ann For Sci 71(4):463–472

Zimmerman MH,Brown CL(1971)Trees:structure and function,vol 1.Springer,New York,pp 61–67