红色的字体为初次提交后作修改的地方
Growth stress in Eucalyptus dunnii
邓恩桉的生长应力
Timothy N. Murphy1,3, Michael Henson2 and Jerome K. Vanclay1,4
1Sustainable Forestry Program, Southern Cross University, PO Box 157, Lismore, NSW 2480, Australia
2Forests NSW, PO Box J19, Coffs Harbour Jetty, NSW 2450, Australia
3Australian Forest Corporation, 75 Elliott Road, Lismore, NSW 2480, Australia
4Corresponding author, email: jvanclay@scu.edu.au
Revised manuscript received 28 April 2005

Summary
摘要

Growth stress in 9-y-old plantation-grown Eucalyptus dunnii was assessed by measuring longitudinal growth strain. Strain varied considerably (370–1560μm-1) and was sufficiently heritable (h2 = 0.3–0.5) that tree breeding may be an effective way to reduce the incidence of growth stress in this species. Although the formation of longitudinal growth strain appears to be under strong genetic control (P = 0.0015), there was a tendency for tall thin trees to exhibit higher stress than short thick trees (P = 0.025 for height:diameter ratio). Two provenances and three families identified in this study show potential as superior material for further tree breeding.
生长在人工林9年龄的邓恩桉其生长应力是通过测量纵向生长应变来进行评估的。应变幅度之大(370–1560μm-1)，遗传可能性之强(h2 = 0.3–0.5)，使得育种有可能成为减少该种属生长应力的有效途径。尽管表面上纵向生长应变很大程度受到基因的控制 (P = 0.0015)，却也存在着这样一种趋势：高瘦的植株比矮壮的植株有着更高的生长应力(P = 0.025  高度直径比)。本研究确定了2个种源及3个家系在进一步的树木育种方面是具有潜力的出众的材料。

Keywords: wood properties; growth stress; strain; heritability; Eucalyptus dunnii
关键词：木材性质，生长应力，应变，可遗传性，邓恩桉

Introduction
导言

Eucalyptus dunnii Maiden (Dunn’s white gum, Benson and Hager 1993) is a relatively new but increasingly important plantation species in eastern Australia. Over 10 000 ha of E. dunnii plantations have been established in New South Wales (NSW), and it remains one of the favoured species for planting, with some 40% of current plantings in north-eastern NSW and south-eastern Queensland using this species. Although it forms a crucial component of the long-term supply to the sawlog industry, this emerging plantation resource remains untried by industry.
在澳大利亚东部，邓恩桉(Dunn’s white gum, Benson and Hager 1993)的人工林种植虽然时间不长，其重要性却日益加大。新南威尔士州(NSW)已经种植了1万公顷以上的人工林，邓恩桉依然是最受欢迎种属之一。新南威尔士州东北部及昆士兰州东南部目前已有的种植面积中该种属约占40%。尽管邓恩桉是可锯木行业长期供应方面的重要组成成分，这种新兴的人工林材料在产业上却尚未成熟。

The Regional Forest Agreements in 2000 (Davey et al. 2002), and the transfer of further State Forests to National Parks in NSW in 2003, have increased the need for plantations to supply sawlogs to the forest industries. As a result of these decisions, and because of increasing demand, there is a shortage of high-quality sawlogs within NSW. The response by government has been to establish, and to encourage private investment in, hardwood plantations (e.g. the Plantations 2020 Vision, which aims to increase plantation area by 2 million ha by 2020).
2000年的《区域森林协议》(Davey et al. 2002)以及2003年的《进一步将新南威尔士州名下森林转为国家公园的决定》对林业产业提出了更高的种植要求，以保证可锯木的供给。由于这些决定的制定，加上不断增长的需求，新南威尔士州的高质量可锯木产出出现短缺，政府因此为硬木种植的个人投资开辟渠道并给予鼓励（例如，《人工林展望2020》制定了到2020年把人工林种植面积提高到200万公顷的目标）。

Eucalyptus dunnii, like most eucalypt species, suffers from growth stress that causes problems and financial losses in processing sawnwood. Fortunately, there is a great deal of natural variation in eucalypts in the propensity for growth stress (Swain et al. 2000; Yang et al. 2001), and the trait is heritable (Schacht et al. 1998), so there is scope to address this issue through tree breeding. The present study seeks to establish the extent of growth stress in a progeny trial of E. dunnii, to foreshadow the nature of the future mature timber resource and to provide guidance for further tree breeding efforts. While growth stress in 9-y-old trees is likely to be greater than that observed in mature sawlogs, the ranking of families with regard to this trait is likely to remain unchanged. Thus identifying families with high longitudinal growth strain at age 9 y, and removing these families from seed orchards and ongoing tree-breeding programs, will contribute to a better sawlog resource in the future.
正如其他桉类种属，邓恩桉的生长应力导致锯材加工的问题产生和经济损失。幸运的是，生长应力这种习性在天然的桉类树种中各有不同(Swain et al. 2000; Yang et al. 2001)，而且具有可遗传性 (Schacht et al. 1998)，因此，有可能通过树木育种来解决这个问题。目前的研究旨在通过邓恩桉的子代试验找出生长应力的范围，由此对未来的成熟木材资源作出预测和为将来更进一步的树木育种提供指导。因为9年龄植株的生长应力可能比在成熟锯材处测量到的要大，各家系在这种习性上的排序应该不会有所改变。因此，找出在9年龄具有高水平纵向生长应变的家系，并且把它们从种子园和进行中的树木育种计划中除去，将有助于改善未来的锯材资源。

Literature
文献资料

Growth stress is a major cause of degrade and processing problems in eucalypts (e.g. Jacobs 1945; Grzeskowiak 2001). Foresters and sawyers have long been aware of these problems, and have devised a range of techniques to gauge stress and its impact on wood processing (e.g. Jacobs 1945; Boyd 1950; Nicholson 1971; Yang and Hunter 2000; Yang et al. 2005). Most of these methods involve cutting the xylem under controlled conditions and recording the longitudinal strain that is induced as the stress is relieved.
生长应力是桉类木材质量下降和加工困难的主要原因(如 Jacobs 1945; Grzeskowiak 2001)。造林和加工业界对此早有认识，并采取了一系列技术措施对应力及其在木材加工过程中的影响进行测量估算(如 Jacobs 1945; Boyd 1950; Nicholson 1971; Yang and Hunter 2000; Yang et al. 2005)。这些方法中大部分是在控制条件下进行木质部切割并把应力释放后产生的纵向应变记录下来。

There is a great deal of natural variation in eucalypts in the propensity for growth stress (Swain et al. 2000; Yang et al. 2001, working with E. globulus), and the trait is heritable. For example, Schacht (1988) found that the heritability of growth stress in E. urophylla was 0.8, offering considerable scope to reduce this problem by conventional tree breeding. Few accounts of sawing studies of plantation-grown E. dunnii have been published, but results to hand suggest that as much as one-third of sawn material may suffer degrade attributable to growth stress (Matos et al. 2003).
生长应力这种习性在天然的桉类树种中各有不同(Swain et al. 2000; Yang et al. 2001，对蓝桉的研究)，而且具有可遗传性。例如，Schacht (1988)发现尾叶桉的生长应力遗传率为0.8，为通过传统育种解决该问题划定了重要的范围。尽管人工种植的邓恩桉在切割方面的研究结果发表的很少，即将提交的数据显示，经切割的木材多达三分之一会遭受生长应力导致的质量下降(Matos et al. 2003)。

Inferences may also be drawn from comparable studies with other eucalypt species. A study of 10-y-old plantation-grown E. globules (Yang et al. 2002) revealed that 30% of sawn boards were rejected due to excessive distortion, and that 40% of this distortion could be attributed to growth stress. Distortion varied greatly throughout the study material (Yang et al. 2001), and could not be attributed to site or provenance (P >0.1; Yang et al. 2002).
其他桉类种属作对比研究也可以得出结论。10年龄人工种植蓝桉的一项研究(Yang et al. 2002)显示，30%的锯材因为变形过大而被弃用，而该变形的40%应该与生长应力有关。各研究对象的变形情况彼此大不相同(Yang et al. 2001)，且不能归因于林址和种源(P >0.1; Yang et al. 2002)。

The now-standard method of measuring growth strain near the cambium of standing trees by releasing stress was pioneered by Nicholson (1971; Archer 1986), and has been refined into three variants using ‘CIRAD’, resistance and transducer strain gauges. Bailleres and Yang (2003) compared the Nicholson, CIRAD and resistance strain gauges in eucalypts, and reported advantages and disadvantages of all three methods.
通过释放应力测量立木形成层附近的生长应变是目前沿用的方法，其代表人物是Nicholson (1971; Archer 1986)，且被优化成为三种分别使用CIRAD，阻抗和传感器应变测量仪的方法。Bailleres and Yang (2003)以桉类为对象，在Nicholson法，CIRAD法和阻抗法之间作比较，并在报告中总结了这三种方法的优点与缺点。

The CIRAD or French method (Bailleres et al.1995, 1997; Raymond et al. 2002; Loup 2003) couples a Mitutoyo dial gauge with a jig devised by CIRAD, and enables strain to be assessed relatively easily. Advantages of this method are that only one hole is required to release stress, and that CIRAD provides a standardised ‘ready-for-use’ jig that helps to make measurements comparable between different agencies. It has become something of a standard in Australia (Raymond et al. 2002), but CIRAD scientists now appear to favour the transducer approach (Beismann et al. 2000; Fournier et al. 1994; Plomion et al. 2000). The ‘singlehole’ approach employed by the CIRAD method records an absolute value about twice that recorded by the other methods (Yoshida and Okuyama 2002).
CIRAD法也称为法国法(Bailleres et al.1995, 1997; Raymond et al. 2002; Loup 2003)，它把一个日本三丰标度盘仪与一把CIRAD设计的锯子结合起来，使得应变的测量相对容易。这种方法有其优点，释放应力仅需一个孔，而且CIRAD提供的这个既标准化又“可随时使用”的锯子使得不同的机构之间可以进行测量方面的比较。在澳大利亚，CIRAD法几乎形成了一个标准(Raymond et al. 2002)，只不过现在的CIRAD专家们更倾向于使用传感器法(Beismann et al. 2000; Fournier et al. 1994; Plomion et al. 2000)。CIRAD法采用 “单孔”方式记录到的绝对值是其他方法的两倍左右(Yoshida and Okuyama 2002)。

Resistance gauges are generally thin-film, general-purpose 120 ohm strain gauges measuring about 1 mm × 0.5 mm (such as Kyowa KDG-120-C1-11 gauges; Kikata 1972; Yoshida et al. 2000, 2002; Huang et al. 2002; Yoshida and Okuyama 2002). These gauges are glued to the xylem, and measured using a strain meter (e.g. Kyowa UCAM-1A) relying on a Wheatstone bridge. Strain is measured by making small cuts above and below the gauge to release the growth stress. Yoshida and Okuyama (2002) recommended that these cuts should be standardised at about 5– 10 mm deep and 3–10 mm from the gauge. The length of the cut should be about 1.5 times the distance from the gauge (Saurat and Gueneau 1976).
阻抗应变测量仪多为薄膜状结构，通用120欧姆，尺寸约1 mm × 0.5 mm(以Kyowa KDG-120-C1-11为例 ; Kikata 1972; Yoshida et al. 2000, 2002; Huang et al. 2002; Yoshida and Okuyama 2002)。这些仪器被粘固在木质部上，由一个依靠惠斯通电桥工作的应变表对其进行测量。通过在阻抗应变测量仪的上方和下方的施以小切口释放生长应力，应变就被测量到了。Yoshida and Okuyama (2002)推荐的做法是，把切口规范为深5~10mm，距离仪器3~10mm。切口的长度应为与仪器间距的1.5倍(Saurat and Gueneau 1976)。

Transducer gauges are a less intrusive analogue of the resistance gauges. The advantage of the transducer approach is that the gauge is attached to the tree with two probes, and need not be glued. Recent work with transducers (Fournier et al. 1994; Beismann et al. 2000; Plomion et al. 2000) has used a DD1 strain transducer manufactured by Hottinger Baldwin Messtechnik, a small (64 mm × 26 mm × 10 mm), light (20 g) and portable instrument (Yang et al. 2005 illustrate this instrument in their Fig. 3).
传感器应变测量仪与阻抗应变测量仪类似，但对植株的侵害性轻些。传感法的优点是仪器只需通过两个探测器附着在树上，无需粘固。最近进行的传感法研究(Fournier et al. 1994; Beismann et al. 2000; Plomion et al. 2000)使用Hottinger Baldwin Messtechnik生产的DD1传感器应变测量仪，它尺寸小(64 mm × 26 mm × 10 mm)，重量轻(20 g)，而且便于携带(Yang et al. 2005 在他们的图3里对其进行了描述)。

Growth strain may vary considerably within a tree, and more than one measurement is needed to reliably quantify strain. Raymond et al. (2002) examined a range of methods for assessing growth stress, by using strain gauges (up to 36 measurements taken at four aspects and nine heights), measuring board deflection, and by assessing end splitting. They concluded that two measurements of strain taken at or near breast height could provide an adequate and efficient estimate of growth stress that correlated well with other methods.
即使是在一棵树上，生长应力的变化也相当大，要测量到可靠的应变值，一组数据是不够的。Raymond et al. (2002)通过使用应变测量仪（4个方向9个高度多达36组数据）和通过评估末端开裂，对测量生长应变测量板挠曲的一系列方法进行研究。他们总结出两种在植株胸部或附近的进行应变测量的方法，能够对生长应变提供足够的有效的估计数据，还和其他方法紧密关联。

Study site
研究林址

The study was conducted in a 9-y-old progeny trial (Johnson and Arnold 1998) at Boambee State Forest (30°18′S, 153°03′E, 60 m asl), south-west of Coffs Harbour, during January and February 2004. Boambee experiences a mild temperate climate (average daily temperature 14–23°C), with a median rainfall of 1585 mm. The soils are yellow podsolics over shale geology.
2004年的1月和2月期间，本研究在一个9年龄子代试验林址(Johnson and Arnold 1998)里进行。该林址位于考夫斯港西南的Boambee国有森林(南纬30°18′, 东经153°03′, 海拔60 m)。那里气候温和(日平均温度 14–23°C)，1585 mm的中等降雨量，土壤为页岩发展而成的黄壤灰化土。

Eucalyptus dunnii has a limited distribution in and near the Border Ranges of NSW and Queensland (Boland et al. 1984; Benson and Hager 1993; Specht et al. 1995). The Boambee progeny trial contains seedlings raised from 219 open-pollinated (family) seedlots collected from individual E. dunnii trees in wild populations throughout most of this range. These seedlots were collected by the CSIRO Australian Tree Seed Centre between 1986 and 1991, and represent 21 different provenances, numbered 1–21 from north to south. Each seedlot was represented in the trial by six replicates of four-tree row plots arranged as incomplete block designs with one-dimensional blocking within replicates. Trees were spaced at 3.0 m (between rows) × 2.4 m (within rows).
在边界国家公园位于新南威尔士州和昆士兰州境内的那部分地区，邓恩桉的分布是有限的(Boland et al. 1984; Benson and Hager 1993; Specht et al. 1995)。Boambee子代试验林址里有着由219个自由授粉(家系)种批培育得来的实生植株，这些种批由澳大利亚联邦科学与工业研究组织的树种中心在1986年到1991年间收集自遍布该地区大部的野生群落里的邓恩桉个体，代表着21个种源，从北至南用1到21进行编号。在试验中，每一个种批按照不完全区组设计，以4棵一行的裂区重复6次进行种植，重复的裂区之间一维相邻。树木间距3.0 m（行间距）X2.4 m(行内树间距)。

谢谢layman的奖励。我会继续努力。

The study site was previously an E. grandis plantation, which was clearfelled at age 40 y. Logging debris was pushed into windrows and burnt in the spring of 1994. A winged ripper pulled by a crawler tractor ripped planting lines to a 50–60 cm depth between the stumps. Regrowth of woody weeds was sprayed with glyphosate herbicide in December 1994. The trials were handplanted in February 1995 with sun-hardened seedlings in Hiko V93 cells. Within a month of planting, each seedling received one 20 g Langley Tree Tablet (N 20%: P 4.4%: K 8.2%: S 6.0%), buried 10 cm deep about 15 cm away from the seedling. Thinning at age 4 y removed the poorest two trees from each four-tree row plot, based on a subjective estimate of tree volume and stem form (Johnson and Arnold 1998). At the time of last measure in late 2003, trees averaged 23 cm diameter (breast height, over bark) and 28 m in height.
研究林址之前是一片巨桉人工林，在40年龄时进行清伐。采伐剩余物推成行堆并在1994年春进行焚烧，然后用履带式拖拉机带动的翼状松土机把树桩间的栽种行犁至50–60 cm深。1994年冬天，用草甘磷除草剂喷洒重新长出来的木本杂草。1995年2月，试验林开始种植，手工栽种Hiko V93苗盘里的阳光硬化秧苗。栽种一个月内，在每一棵秧苗15cm开外10cm深的地方，埋置20g的Langley Tree Tablet（氮20%：磷4.4%：钾8.2%：硫6.0%）。4年龄进行间苗，根据对苗高和干型的主观估计，把4棵一行里面最弱的两棵移走(Johnson and Arnold 1998)。到2003年末实施最后一道措施止，植株胸径平均23cm（胸高，带树皮），树高平均28 m。

Method
方法

Fifty-two families from 15 provenances were selected for strain measurement on the basis of their superior growth. Strain in all the standing trees of these families and provenances (164 trees in all) was measured with resistance strain gauges in accordance with the procedures outlined by Yoshida and Okuyama (2002). Strain in some of the trees was also measured with the DD1 transducer (in addition to the standard resistance gauge measurement). The DD1 was not available in Australia at the time this study commenced, and was used on only the final 18 trees assessed in this study. Because of this small sample and haphazard selection, strain estimates from the DD1 were not used to rank families. The DD1 results are reported only because the instrument has not previously been compared with more established methods (Yang et al. 2005).
由于生长力超凡, 来自15个种源的52个家系被挑选进行应变测量。这些种源和家系的立木(共计164植株)的生长应变根据Yoshida and Okuyama (2002)规定的程序，通过阻抗应变测量仪进行测量。其中一些植株还使用了DD1传感器应变测量仪来测量生长应变（在标准的阻抗应变测量仪的基础上予以增加）。在试验开始之时，澳大利亚还没有DD1，本研究仅在最后的18植株使用了这种仪器。因为样本数量小而且带任意性，DD1的生长应变估值未被用来对家系进行排序。我们在报告中公布DD1的生长应变估值仅仅因为在此之前这种仪器还没有跟更成熟的方法进行过比较(Yang et al. 2005)。

Resistance strain gauges require a window to be cut into the bark so that the gauge can be glued to the cambium and initialised. Once that is done, growth stress is released by making small cuts above and below the gauge, and the strain is recorded on the gauge. The time-consuming part of this procedure is cutting the window in the bark, gluing the gauge in place, and allowing the glue to cure. The precision part — initialising the gauge, cutting the slot and recording strain — must be done carefully, in calm conditions. Thus it is prudent to cut the windows and place the gauges in advance (e.g. the afternoon before), and to read them the following day (e.g. at dawn, and for as long as calm conditions prevail).
阻抗应变测量仪的使用需要在树皮上开一个窗口，以便把仪器粘固在形成层上，再进行初始化。一旦做了这个工作，就可以通过在仪器的上方和下方实施小切口来释放生长应力，仪器则把应变记录下来。其中在树皮上开窗、粘固仪器、以及等待胶水起作用的步骤是费时的；而仪器初始化、做切口、和应变记录这些步骤是需要精确性的，必须在宁静的条件下小心对待。因此，稳妥的办法是，预先开窗和放置仪器（例如在前一天下午），在次日读取数据（例如在拂晓，只要环境条件总的是宁静的）。