1 Introduction

Loblolly Pine (Pinus taeda) is one of the most important exotic species in China and has been extensively planted on hilly lands and low mountains in the southern China for pulpwood and building lumber. It is estimated that its total planting areas were at least more than 1×10⁴ hm² in 1995. Having not been thought of the genetic properties in tree growth increment of Loblolly Pine in the past thirty years, the volume productive per hectare for these stands of Loblolly Pine was lower and trees stem form of the stands did not look good. According to the statistical data in 1992 its seed orchard in China was about 809 hm² and its seed stand was a little more or less 2 720 hm². Its seed output is far inferior to the demand of our forestry production and the property of the seed is genetically lower. Every year about 25 t of its seed was imported from the USA to our country (Peng, 1992). To improve this situation, the provenances trial of Loblolly Pine organized by Chinese Academic of Forestry was systematically carried out two times in 1981 and in 1983 in China respectively. The seeds of the trial in 1981 were collected from the natural stands and planted in seven provinces with different climatic and ecological condition. The 1983's were obtained from improved seeds and afforested separately in 13 provinces in China. In 1992 the preliminary results showed that there were statistically distinct differences among the different provenances in tree diameter and volume of Loblolly Pine, trees stem form, the interaction of seed sources in tree height with locations and resistance were significant (Peng, 1992). Now the reports about wood properties of exotic Loblolly Pine in China were not much, let alone the wood properties of its provenance (Xu et al., 1993). How many are the variant range in wood properties within tree and among trees? What do the wood variation patterns among and within its provenances have? What is its wood in China? How is the wood efficiently harnessed? Is it necessary to improve its wood properties? etc. In order to produce and use its wood efficiently, these questions should be progressively settled down.

In the indigenous range of Loblolly Pine, significant differences in tree growth increments, resistance and wood quality were found among the seed sources of Loblolly Pine (Zobel et al., 1989). Numerous investigations have been conducted to assess variation of wood properties for Loblolly Pine (Zobel, 1958; Zobel et al., 1966; 1989;Mitchel, 1963; Mcmillin, 1968; Charles, 1973; Pashin et al., 1980; Talbert et al., 1981; Mckinley et al., 1982; Megraw, 1985). The properties of pulp and paper made from its woodhave been reported (Barefoot et al., 1965; Zobel et al., 1989). There were also many other studies on wood properties of Loblolly Pine provenances (Zobel, 1958; Jackson et al., 1962; Mckinley et al., 1982;Byarm, 1988; Tauer, 1990). Much of their attention was paid to variation of specific gravity (or basic density) and tracheid length among and within provenances. According to their research results the wood properties of Loblolly Pine in nature stands, especially both specific gravity and tracheid length, usually follow a rather predictable pattern, but wood properties in its exotic plantation cannot be predicted. Therefore it is vital for us to directly determine the wood properties of exotic Loblolly Pine where it is grown.

The specific objective of this paper is to study the variation of wood anatomic features, wood density and chemical composition contents among provenances of exotic Loblolly Pine for pulpwood in a common environment, which will provide a background for further studying the wood variation in different environment. The results obtained will be used to evaluate the introduction effects of exotic Loblolly Pine, guide its tree breeding and its wood improvement, evaluate effects of silvicultural practices and tree growth rate on wood properties and guide reasonable and efficient utilization of the wood.

2 Materials and methods 2.1 The origin of the seeds, stand site and field plan

The trial stand of 31 provenances for Loblolly Pine was established in spring of 1983 in the Forest Farm of Fuyang County in Zhejiang Province (eastern of China). The Forest Farm is located at 119°58′E and 30°15′N. Its annual rainfall and mean temperature are 1 478 mm and 15.6℃. Its frostfree season is 221 days per year. The test stand is situated at the hill of about 150~170 m height above sea level in the Forest Farm, its site is red soil and its soil thickness is about 60~80 cm. The range of its pH was from 4.5 to 5.7.

The seeds of 30 provenances by Forest Service of the American were separately collected from improved stands that were distributed over twenty states (Tab. 1). The seed collected from the mother trees of Loblolly Pine planted in 1945 in Wuhan was tested together as the contrasting provenance. The geography and meteorology factors of various provenances in their local regions were presented in Tab. 1.

The field plan of provenances is designed in form of randomized block, double lines in a block are twenty trees and replication number is five. Plant and row spacing in the test stand is 1.5 m×2.5 m.

2.2 Methods of sample trees collection and cutting clear test specimens from sample trees

The test stand was 10 years old. Ten trees were collected from each provenance in the test stand (two trees per replication in each block) in light of the demands of ISO 4471-1982. The total number of sample trees collected was 310.

The height and diameter at breast height of each tree were first measured. After trees were felled, the diameters of 1/2 height of trees were also determined. Based on increments obtained, the volume of tree was calculated. A round-timber about 2.7 m length started at 1.0 m height from base to top was sawn from each tree for determining wood properties.

In the laboratory five clear sample-discs about 2~4 cm in thickness were sawn near the breast height from a round-timber. The specific location of these discs was judged according to the relative height of the tree so that the variation in wood properties could be controlled in a little range. The discs were made full use of to determine features of wood anatomy, basic density or specific gravity, chemical composition. 3~5 test specimens of each wood property above were cut from pith to bark at each disc on the condition of definite ring age.

According to the variant range of both latitude and longitude of Loblolly Pine, ten provenances were selected to study its variation in chemical constituents of wood and samples of each provenance were ten for chemical analyses.

2.3 Test methods and statistical analyses

Measurements of ring width and latewood width were taken directly on the smooth discs processed with the help of a stereomicroscope fitted with a micrometer. Based on the measurement the latewood percentage could be calculated. Basic density of 1~3 rings, 4~5 rings and 6~9 rings were determined by the maximum moisture method. Tracheids of 1~3 rings, 4~5 rings and 6~9 rings macerated separately with Juffery fluid were directly measured by projecting them on a screen (40 cm diameter) at 50×or 100×magnification with a 35 mm slide projector. Thirty readings were made per sample, selecting randomly unbroken tracheid. The values of tracheid width, tracheid diameter and tracheid wall thickness were acquired directly on 20 μm thick cross sections with microscope at 400× magnification, selecting thirty cells separately from earlywood to latewood to be determined. The MFA average of 1~3 rings, 4~5 rings and 6~9 rings were separately obtained on the bases of determining 30 15~20 μm thick tangential sections treated chemically with polarizing microscope at 100× magnification.

The values of wood properties below were obtained using the following test procedures: 1) Extractives content, ASTMD 1105-79;2) Lignin content, ASTEMD 1106-77;3) Cellulose content, ASTEMD 1103-77;4) Pentosan, China GB 2677-9-81.

The statistical analyses of data obtained were conducted using SAS Programmer and the IBM Programmer. These include analyses of variance and linear and multiple regression. The significance of each factor and the interactions between factors were ascertained at the statistic levels of α=0.05, 0.01 and 0.001 respectively. The variant range of the samples' values determined of each property were examined to be normally distributed with different means and with a common variance.

3 Results and discussion 3.1 The differences analyses of tree increments and wood properties among and within provenances 3.1.1 Trees growth characters

Tree height (TH), diameter at breast height (DBH), height-diameter ratio (HDR), ratio of diameter at middle height to DBH (MBDR), tree bole form (TBF) and tree volume were statistically tested at α=0.01 level, respectively. The results showed that the significant differences in tree growth characters except both of MBDR and TBF existed among provenances. This was consistent with the preliminary results in 1991(Peng, 1992). These meant that some superior provenances with both better growth characters and better wood properties could be selected out from trial stands of Loblolly Pine provenances. It was vital necessary for us to further study the variation of wood properties among and within its provenances.

Their broad sense hereditabilities (h_f²) were TH: 0.662, DBH: 0.720, HDR: 0.708 and tree volume: 0.695, in order. These proved that the four characters were under the genetic control of medium level. The environment factors in variance proportion accounted for 36.50%~45.57%, which reflected that site condition has great effects on tree growth of Loblolly Pine(Tab. 2). It should be emphasized in establishing plantation of Loblolly Pine for pulpwood and construction timber.

3.1.2 Tracheid morphological features

Variance analyses in Tab. 3 showed that the statistically distinct differences in tracheid length, tracheid width, trachied length-width ratio and tracheid diameter existed among 31 provenances of Loblolly Pine. The F-values range of 1~3 rings, 4~5 rings, 6~9 rings, 1~5 rings and 1~9 rings in tracheid length, tracheid width, tracheid length-width ratio and tracheid diameter were 3.35~4.91, 2.01~3.98, 2.49~3.36 and 2.29~6.74, separately. These values were all greater than F_α= 1.86 at α= 0.01 level. Their broad sense heritabilities (h_f²) range of tracheid length, tracheid width, tracheid length-width ratio and tracheid diameter were in order 0.695~0.752, 0.503~0.749, 0.564~0.858 and 0.598~0.702. These manifested that tracheid morphological features were genetically controlled above medium level. Among these heritabilities the h_f² value of tracheid length was the largest, the tracheid width was the smallest and the tracheid diameter was medial.These results reflected that tracheid length was controlled much genetically more than tracheid width and tracheid diameter, i.e., both tracheid diameter and tracheid width were much easiler affected by environment than tracheid length, which was vindicated by their environment variance percentage. For example, the environment factors of tracheid width and tracheid diameter in the total variance analyses accounted for 61.87% and 61.16%; that of tracheid length only had 51.68%. The heritabilities (h_f²) of tracheid morphological features of 1~3 rings, 1~5 rings and 1~9 rings in Tab. 3 increased gradually as trees aged. It is active and meaningful to guide wood improvement of Loblolly Pine.

3.1.3 Microfibril angle(MFA)

The results of variance analyses in Tab. 4 showed that there were statistically distinct differences of MFA of different rings range among provenance of Loblolly Pine (F>F_0.01= 1.86). This demonstrated that there were some meaningful effects in controlling variation of MFA through selecting superior seed sources based on the results of the test stand of its provenances. The MFA of Loblolly Pine was genetically controlled in medium level. Its broad sense heritabilities of 1~3 rings, 4~5 rings, 6~9 rings, 1~5 rings and 1~9 rings were 0.788 2, 0.748 1, 0.634 1, 0.743 3 and 0.656 6, respectively. These indicated that there was a tendency that the hereditability of MFA would descend as tree aged. It could be confirmed from the proportions of provenance and environment (block and error) in variance analyses. The percentage of environment factor would ascend and that of provenance would descend as tree aged.For example, within the ranges of 1~3 rings and 1~5 rings provenance variance was 61.9% and 47.5% separately and their environment variance was 38.1% and 52.5%, while within 1~9 rings the corresponding values of provenance and environment were 35.8% and 64.2% in order. These implied that genetic control of MFA was obviously affected by tree age and environment where trees were grown and as trees were elder the controlling effects genetically of MFA would lower. It is necessary to further study effect of sites and other environment factors on MFA of wood.

3.1.4 Latewood percent and basic density

Variance analyses in Tab. 5 showed there were distinct differences of both basic density and latewood percent among Loblolly Pine provenances (F>F_0.01=1.86). Their broad sense heritabilities were 0.722 and 0.644 separately. The environmental factors have significant effects on wood basic density and latewood percent of Loblolly Pine, which were confirmed from percentage of environment (basic density: 56.43% and latewood percent: 59.18%) in variance analyses in Tab. 6. These manifested that superior provenances in wood improvement of Loblolly Pine should be combined together with good sites selection and intensive culturing practices.

3.1.5 Wood chemical compositions

Because the wood chemical compositions were not easy to be determined and considering the broad natural range of Loblolly Pine provenances as well as representatives of samples, ten provenances were carefully chosen out to determine their wood chemical composition. The results in Tab. 6 showed there were no statistically momentous difference in contents of cellulose and lignin among provenances of Loblolly Pine except contents of both pentosan and benzene-alcohol extractives. Cellulose content among provenances of 10-year-old Loblolly Pine varied from 41.47% to 43.15%, while contents of lignin, pentosan and the extractives from 28.8%, 14.85% and 1.45% to 30.10%, 16.45% and 2.78%, respectively. The broad sense hereditabilities of wood chemical composition were: cellulose 0.088, lignin 0.003, pentosan 0.340 and benzene-alcohol extractives 0.307. These showed that wood cellulose content and lignin of juvenile Loblolly Pine were not genetically controlled, while the pentosan and the benzene-alcohol extractives were under weak genetic control. Further analyses indicated that variations in contents of main chemical constituents within a given provenances of Loblolly Pine were greater than those among provenances. Therefore the wood chemical compositions in selection of superior provenance of Loblolly Pine might not be considered.

3.2 Geographical variation of wood properties for short-rotation pulpwood

Method of multiple linear regression equation was used to analyze relationships of each of wood properties for pulpwood to latitude, longitude, annual mean temperature, annual rainfall and forstfree season of 30 Loblolly Pine provenances in indigenous regions, respectively. Here correlation coefficients analyzed of 15 provenances of Loblolly Pine from Eastern Coast of the American were listed in Tab. 7. The results showed that the latitude of provenances was significant negatively related to tracheid diameter and tracheid width and it had little negative effect on tracheid length. Of five factors, annual mean temperature was positively correlated with microfibril angle of S₂ layer of tracheid wall. This meant that wood from provenance of lower latitude districts had larger microfibril angle and its wood strength would decrease. Latitude had great significant effects on basic density and it was positively notably correlated with wood basic density. These results manifested that wood of the provenances from lower latitudes districts had lower basic density and weak strength and that wood of those from higher latitudes districts had higher basic density and stronger strength. This was consistent with geographic variation of microfibril angle and effects of microfibril angle on wood strength analyzed above. Geographic variation of wood basic density in the provenance trail stand (plantation) was consonant with the results of Byram (1988), Jackson et al.(1962) and Tauer (1990), but not agreeable to the results of its natural stands reported by Talbert et al. (1981), Zobel et al. (1989).

In Tab. 7 latitude has little distinct effects on tracheid length and tracheid length-width ratio. Tracheid of Loblolly Pine was very long and ratio of tracheid length-width was large. It does not seem important for these two items for pulpwood in selecting its superior provenances. It is suggested that tree growth increments are first considered and wood basic density second in its wood breeding. Some superior provenances with properties of both faster growth and middle basic density from lower latitude districts could be introduced to culture for better pulpwood.

The effects of provenance's longitude on wood properties were more complex than latitude. In a word, here is a tendency for strength of their wood to increase when provenances far from eastern coast were introduced to plant in districts near coast (for example in Zhejiang Province of China). On the condition of the difference of the latitude between originating and culturing areas being than 3° and the provenances of Loblolly Pine were introduced to plant near coast line (for example in Zhejiang Province of China), the originating longitude of these provenances was notably negatively correlated with basic density. The results should be stressed in introducing and culturing of Loblolly Pine provenances for wooden industrial material.

3.3 Effects of fast grown tree on wood properties

Usually tree growth rate could be expressed with its diameter or mean annual ring width of a given height within tree under condition of the same age. Method of linear regression equation was used to analyze relationships of diameters at breast height (DBH) of 310 trees with their wood properties sampled from discs near breast height of trees. The results in Tab. 8 proved that the growth rate had significantly negative effects on tracheid length-width ratio, tracheid double wall thickness-diameter ratio, basic density. And it also had notable positive effects on tracheid width, tracheid diameter, tracheid diameter-width ratio and tangential-radial shrinkage percent of wood. It was negatively related to tracheid length, tracheid wall thickness and contents of lignin, pentosan and bezen-alcohol extractives of wood, but not notable.

According to growth data of 31 provenances of Loblolly Pine, trees of provenances from districts of its lower latitudes grow faster than that from its higher latitudes. Wood basic density and strength of wood determined from higher latitudes areas were larger and stronger than those from the districts of lower latitudes (Xu et al., 1993). Specific analyzing the causes of faster growth rate of provenances from lower latitudes districts could be attributed to enlargement of tracheid diameter and tracheid width beside faster division rate of xylem cambium and increasing of tracheid number. Tracheid wall thickness, as an important anatomic feature of wood, did not proportionally increase with ascending of tracheid diameter and tracheid width and had a tendency to decrease. Theoretically this tendency will result in declining of wood density and wood strength and it is not beneficial to culturing building lumber with high strength. These were confirmed from relationships of wood density being notably negatively related to tracheid diameter (r=-0.458 0>r_0.05=0.145 2) and tracheid width (r=-0.383 3) and wood density being obviously positively related to wood strength. As to pulpwood descending of tracheid double thickness-diameter ratio and ascending of tracheid diameter-width ratio will be beneficial to collapsing of fibers in pulping and better folding of fibers and smoothness of paper. The reduction of pulping yield result from decreasing of wood density would be compensated by large increase of tree volume increments. Consequently some superior provenances having fast growth property with medium wood density could be selected to culture for pulpwood. For construction timber much more consideration given to growth rate of superior provenance might not be suitable and high wood density and high latewood percent in selection of superior provenances should be stressed.

4 Conclusions

The statistically significant differences in trial stand of 10-year-old exotic Loblolly Pine were found among provenances in tree growth increments, tree form factors and wood anatomic features for pulpwood. These suggested that selection of superior provenances in the trail stand be specific effective.

The wood properties studied except main chemical composition were genetically controlled in medium or more level and each of their broad sense heritabilities (h_f²) was greater than 0.50. There were no significant differences in contents of cellulose and lignin among 10 provenances studied. The heritabilities (h_f²) of the wood chemical constitutes were: cellulose 0.088, lignin 0.003, pentosan 0.340 and extractives 0.370 respectively. The environment factors also have a large notable effect on wood anatomical features. In culturing better wood resources for pulpwood, superior provenance of Loblolly Pine should be planted on better sites and vital intensive management practices should be handled.

The latitude of Loblolly Pine provenances in their native regions has a great effect on wood properties. It was significantly related negatively to diameter at breast height, tree volume, tracheid diameter and tracheid width, positively to basic density. The effects of provenances' longitude on wood properties were more complex than latitude. On the condition of the difference of the latitude between originating and culturing areas being than 3(and the provenances of Loblolly Pine were introduced to plant near coast line (for example in Zhejiang Province of China), the local longitude of these provenances was notably negatively correlated with basic density. The results are an active guiding meaningful to the introduction and culturing of Loblolly Pine provenances.

The growth rate of Loblolly Pine in the trail stand had significant negative effects on tracheid length-width ratio, tracheid double wall thickness-diameter ratio, and basic density. It also had notable positive effects on tracheid width, tracheid diameter, and tracheid diameter-width ratio.