Australian Forestry
Australian Forestry is published by the Institute of Foresters of Australia (IFA) for technical, scientific and professional communication relating to forestry in Australia and adjacent geographic regions. The views expressed in this journal are not necessarily those of the IFA. The journal is included in the Register of Refereed Journals maintained by the Australian Government Department of Education, Science and Training. The IFA gratefully acknowledges a grant provided by the Australian Government and State and Territory forestry and forest products agencies through the Forestry and Forest Products Committee to assist in the preparation and production of Australian Forestry. Managing Editor: Dr Brian Turner Production Editor: Mr Alan Brown Editorial Panel: Dr Stuart Davey Dr Ross Florence Dr Graeme Siemon Mr Neil (Curly) Humphreys Dr Ian Bevege Dr John Herbohn Dr Grant Wardell-Johnson Dr Humphrey Elliott Contributions Contributions to this journal are sought covering any aspect of forest ecology, forest management, forest policy and land use related to Australia and the South Pacific region. Contributions related to the performance of Australian tree genera elsewhere in the world are also welcome. Instructions to authors are given inside the back cover of each issue. Contributions should be sent to the Executive Director at the address below. This issue This issue of Australian Forestry presents the first five papers in a comprehensive series titled ‘Achievements in the genetic improvement of forest trees in Australia and New Zealand’, together with an introduction to the series. In the three decades since the Third World Consultation on Forest Tree Breeding, sponsored by IUFRO and FAO, was held in this country in 1977, tree improvement has become an essential tool in the cultivation of niche crops as well as major species. The editors wish to thank Garth Nikles for the assistance he has provided to Ross Florence in bringing this series to our readers. Journal subscriptions 2007 A$240 including GST within Australia A$270 per year for all other countries All correspondence relating to subscriptions should be addressed to: Executive Director, The Institute of Foresters of Australia, PO Box 7002, Yarralumla, ACT 2600, Australia. Phone: 61 2 6281 3992. Fax: 61 2 6281 4693. Email: [email protected] Web: http://www.forestry.org.au Cover The front cover features African mahogany, Khaya senegalensis. This is a promising plantation tree for northern Australia, tolerant of difficult sites and producing a high-value timber. The story behind the recognition of its promise, and work to realise its potential, are described by Garth Nikles on pages 68–69 of Volume 69 of Australian Forestry. He also selected the photos and prepared the captions. Photo 1. Butt logs (mainly), from selected trees in 32-y-old, unmanaged stands of African mahogany grown in the Darwin region of the Northern Territory (NT). They were used in a study of sawn timber recovery, timber drying schedules and wood properties, and for manufacturers’ evaluations of the timber. (Photo by courtesy of Don Reilly.) Photo 2. Award-winning set of chess table and chairs made from timber from a sample of the logs of Photo 1. It won three ‘Australian Furniture of the Year Awards, 2004’ sponsored by the Furniture Industry Association of Australia: the Queensland and national awards in the category ‘Excellence in furniture using Australian plantation timber’; and the Queensland ‘Best-of-the-best’ (across categories) award. (Photo by courtesy of Ray Burgess, Sight Photographics, via the manufacturer, Paragon Furniture, Brisbane.) Photo 3. A rooted cutting of African mahogany at 10 mo from planting in a clone test established in the Darwin region of the NT in January 2005. At 12 mo from planting, some trees were nearly 3.5 m high. (Photo by courtesy of Beau Robertson.) Photo 4. Pods on a graft of African mahogany in a clonal seed orchard planted near Darwin in December 2001. The pods are about 35 mm in diameter some 3 mo after flowering. (Photo by courtesy of Beau Robertson.) Photo 5. Part of an unmanaged stand of African mahogany aged 25 y derived from natural-stand seed. It was established on back-filled land following surface mining for bauxite at Weipa, Queensland. The central, dominant tree is 48 cm dbhob and one of 36 superior trees selected so far and used in the conservation and genetic improvement program at Weipa. (Photo by courtesy of Alan Bragg.) Photo 6. A pruned, 8-y-old stand of African mahogany, derived from unimproved seed, planted privately at 5 m × 2.5 m (800 trees ha–1), not thinned, near Bowen, in coastal central Queensland. The tree being measured had a dbhob of 17.1 cm. (Photo by courtesy of Geoff Dickinson.)
ISSN 0004-9158 ACN 083 197 586
Australian Forestry The Journal of the Institute of Foresters of Australia
Volume 70 Number 1, 2007
ISSN 0004-9158 Australian Forestry, Volume 70 Number 1, March 2007
Contents
Page
1 Guest editorial: The turn of the forestry wheel Roger Underwood 3 Achievements in the genetic improvement of forest trees in Australia and New Zealand: a series of papers presented in this and following issues of the journal The Editor 4 Achievements in forest tree genetic improvement in Australia and New Zealand 1: Eucalyptus pilularis Smith tree improvement in Australia Michael Henson and Helen J. Smith 11 Achievements in forest tree genetic improvement in Australia and New Zealand 2: Development of Corymbia species and hybrids for plantations in eastern Australia David J. Lee 17 Achievements in forest tree genetic improvement in Australia and New Zealand 3: Tree improvement of Eucalyptus dunnii Maiden Helen J. Smith and Michael Henson 23 Achievements in forest tree genetic improvement in Australia and New Zealand 4: Tree improvement for low-rainfall farm forestry C.E. Harwood, D.J. Bush, T. Butcher, R. Bird, M. Henson, R. Lott and S. Shaw 28 Achievements in forest tree genetic improvement in Australia and New Zealand 5: Genetic improvement of Douglas-fir in New Zealand C.J.A. Shelbourne, C.B. Low, L.D. Gea and R.L. Knowles 33 Effects of height on physical properties of wood of jhau (Casuarina equisetifolia) Md Qumruzzaman Chowdhury, A.Z.M. Manzoor Rashid, Md Shah Newaz and Mabubul Alam 37 Wood and fibre productivity potential of promising new eucalypt species for coastal Zululand, South Africa Robin A.W. Gardner, Keith M. Little and Athol Arbuthnot 48 Genetic variation in survival and growth of Eucalyptus globulus ssp. globulus in Ethiopia Tesfaye Hunde, Belachew Gizachew and Chris Harwood 53 Managing insect pests in Eucalyptus globulus (Labill.) plantations in Victoria using insecticide tablets at establishment N.G. Collett and J. McBeath 61 GIS-based tools for management of pine plantations, Queensland, Australia Micaela Preda, Kenneth A. Bubb and Malcolm E. Cox 70 Book review: Forestry and Environmental Change: Socioeconomic and Political Dimensions J. Schirmer 72 Referees 2006 Roger Underwood 1
Guest editorial
The turn of the forestry wheel
The last 10–15 years have seen two of the most dramatic changes consultations, draft documents, workshops, interstate consult- to Australian forestry since forest management by professionally ants, re-consultation, final plans, ministerial appeals and trained foresters began 90-odd years ago. decisions. A simple management plan for a patch of bush can take five years to prepare, and even then the final document is The first has been the transfer of extensive areas of native forests more about pressure group appeasement than providing a from state forests to national parks. State governments have blueprint for action. chalked this up as a political coupe and it has brought satisfaction to those in the community who regard national parks as the highest The most serious changes associated with the new agencies are form of land use.The change, however, has not been uniformly philosophical and cultural. In the new agencies, forests and welcomed, and there have been social and economic downsides. forestry are no longer the heart and soul of the business or a For example, the contraction in permitted forest uses (notably life-consuming passion. Their responsibilities are immense and the cessation of timber production and some other economic complex. Moreover, the upper echelons rarely consist of staff activities) has led to changed financial arrangements. Forest who have actually worked in the bush. It is now almost unknown managers unable to generate revenue have been forced to for departmental senior staff to have cut their teeth as young compete with health, education, and law and order for finite officers running a remote forest district, defending their patch treasury funds, generally without community support. This in turn from fires, regenerating cutover bush or carrying out silvicultural has led to contractions in forest staff and some operations. research. Nor have they learned what it is like to live in, come to understand, and contribute to life in a small rural community. In From the perspective of the forest, the tenure (i.e. state forest a modern bureaucracy, the name of the game is political survival. or national park) does not matter. What does matter is how the forests are managed. Is the new approach to forest management working? By this I mean are our forests in better shape now than 10–15 years ago? This brings me to the second great change: the revolution in the My view is that they are not, one indicator being the extensive administration of our public forests. In Queensland, New South crown decline in the once-beautiful and healthy wandoo and tuart Wales, Victoria and Western Australia the professionally-led forests, and elsewhere. Just as dramatic has been the deterioration forests departments or forestry commissions set up in the early in the standard of bushfire management, a fundamental require- years of the 20th century have disappeared. They have been ment for forest conservation in Australia. replaced by a mish-mash of agencies concerned primarily with environmental protection and wildlife conservation. In some Although no government or agency will admit it, bushfire cases, as in Western Australia, the department is overseen by a management in Australian forests is now at a new low point. The community body (made up of non-foresters) responsible for number and size of high-intensity forest fires clearly shows how forest policy and management plans. Administration of native badly things are going. In south-eastern and south-western forest timber production and plantations is the job of a small Australia there have been large and damaging forest fires every commercially-oriented agency with no land management summer for several years now. To some extent this is a result of responsibilities. Only in Tasmania does a government agency the de-staffing of the bush — the loss of experienced foresters survive whose focus is solely forests and forestry. and agency firefighters who were once found in every forest district, and their associates in the timber industry. There are These institutional changes have had many repercussions, also problems in the forest, such as failure to maintain access reaching right down into the forest. One retrograde example has for firefighters (indeed in some cases the deliberate closure of been the dissolution of the former system of forestry districts, access and fire trails), constraints on the use of bulldozers for each with its staff of professional foresters, field officers and fireline construction, and the new health and safety concept that forest workmen. The former forestry staff, who knew every track firefighters should not be put into dangerous situations. But the and tree in their patch, have been replaced by park rangers whose most troubling aspect is the philosophical change our new forest focus is management of people and recreation sites. The District managers have introduced. This has been to move away from Forester has become an environmental administrator in a remote preparedness and damage mitigation (including effective regional centre, and forest management is only a small part of programs of fuel reduction burning) in favour of suppression as the job. the key approach to bushfire control. The 2006/2007 fires in Another change has been the way in which forest management Victoria have demonstrated that firefighting in forests with heavy plans are developed. The former system was efficient and self- fuels is a task which exceeds the resources of the whole nation, correcting because the people preparing the plans had to even with international assistance. Yet the drip torch is being implement them. This has been replaced by a new juggernaut replaced by the aerial water bomber and the mild prescribed burn process based on ‘stakeholder analysis’. This involves public by the stand-replacement holocaust. In fact this ‘new’ philosophy
Australian Forestry 2007 Vol. 70 No. 1 pp. 1–2 2 The turn of the forestry wheel
(which denies Aboriginal burning and rejects the value of But a combination of burnt fingers, real-world experience, prescribed burning) is not new at all. It is a repetition of a strategy research into fire behaviour and a new understanding of the which failed dramatically in the USA, and also failed in Australian ecological role of fire in eucalypt forests saw an effective system forests when it was attempted here prior to the 1950s … another of fire protection developed and implemented by the 1950s. Now corporate memory lost when the old forestry departments were the wheel has turned full circle, with our forest managers again replaced. taking their fire wisdom from European ecologists and United States firemen while our forests are being burnt and reburnt by One of the things that most dismays me is that the new forest uncontrollable high-intensity bushfires. managers no longer seem to understand that fuel reduction burning in forests also provides less dangerous conditions for I am not arguing here for national parks to be changed back into firefighters and a higher level of bushfire safety for the wider state forests (although I confess I believe that the world needs community. Unstoppable wildfires originating deep in long- more well-managed multiple-use forests rather than more unburnt forests have now become the principal threat to national parks). My concerns are professional and concern the firefighter’s lives and community assets right across Australia’s well-being of our native forests, whatever their tenure. A number south-east and south-west. Furthermore, a policy of fighting of issues need to be addressed, but of these bushfire management forest fires only when they leave the forest increases the burden is the most urgent. To achieve an acceptable standard, significant of firefighting on volunteers and part-time firefighters. institutional, policy and philosophical changes are needed. These changes will not be initiated from within the agencies or come My heart bleeds for the small cadre of experienced foresters from the current state governments. It is, however, an issue in left in our forest management agencies who understand all this, which the Institute of Foresters could and should play a but whose voices are no longer heard or who are overridden by leadership role. the environmental protection and the health-and-safety experts. And I despair of those who blame bushfire problems on global Roger Underwood warming. If indeed the climate is becoming warmer and drier [email protected] then it is even more imperative that effective programs of bushfire preparedness and damage mitigation are adopted. Roger Underwood is a former District and Regional Forester and a Fellow of the Institute of Foresters of Australia. He is The introduction of formal forest management to Australia and chairman of the Bushfire Front, an independent organisation the creation of professionally-led and dedicated forestry of foresters with expertise in bushfire management, science agencies in the years around World War 1 were a response to and administration, and was the author of the IFA’s the realisation that our native forests were in a sorry state. The submissions to the 2003 House of Representatives Inquiry early foresters did not get it all right from the start — many into bushfires and the Senate’s 2005 inquiry into Australia’s initially had a European attitude to fire, and some were seduced national parks. by the all-out suppression approach of the US Forest Service.
Intense, uncontrolled fires have damaged enormous areas of Australian forests in the last five years.
Australian Forestry 2007 Vol. 70 No. 1 pp. 1–2 Garth Nikles 3
Achievements in the genetic improvement of forest trees in Australia and New Zealand: a series of papers presented in this and following issues of the journal
Australian Forestry has commissioned a series of papers The development of plantations in NZ can be regarded as designed to present to the profession at large an account of occurring in five phases: (i) a species-introduction and farm- achievements in the genetic improvement of plantation species, woodlot phase up to 1900; (ii) state forest plantation and directions for continuing progress. Because of the close development up till World War I; (iii) a major planting boom links between Australia and New Zealand and the notable during 1925–1935; (iv) planting concentrating on radiata and contributions New Zealand has made to plantation research and the development of silviculture from 1950 onwards; and (v) development, the series incorporates reports of advances from the post-1990 era of state forest sales and multiple ownership both countries. by many different companies, but with smaller enterprises doing most of the new planting. New Zealand plantations now Australia’s native hardwood (eucalypt) forests supplied the bulk total about 1.8 million ha. of Australia’s wood requirements from European settlement of the country until late in the 20th century; this production was The Australian and New Zealand plantation programs have supplemented by mainly softwood imports. The limitations of generated a substantial body of research and development leading the eucalypt forests and particularly those of the native softwood to knowledge, technology and practices for improving the yield resource were recognised late in the 19th century, and this lead of wood, and the quality of tree boles and wood properties for to the initiation of softwood planting. With the exception of specific end uses. With scientifically based management, overall hoop pine in the subtropics, these plantations have been based production is improving across a range of soil and environmental on exotic species, predominately radiata pine in temperate conditions and forest ownerships. regions and the slash/Caribbean pine complex in the subtropics. Programs for the genetic improvement of forest trees are one of Four phases can be recognised in the Australian plantation the key tools for enhancing plantation profitability and program: (i) early ‘softwood import replacement’ to the mid- sustainability. They incorporate assessment of genetic diversity in 1950s; (ii) increased planting extending to the mid-1970s in species; and selection and breeding to improve a range of growth, order to achieve national ‘self-sufficiency’ in wood production adaptation and wood quality traits. They commonly include inter- (based in part on Commonwealth funding of expanded plantation specific and or intra-specific hybridisation, evaluation of genetic programs); (iii) increasing private investment in softwood gains under specific soil and environmental conditions, and plantations through the 1980s and limited expansion of the then development of large-scale propagation technology including small area of eucalypt plantations; and (iv) beginning in the clonal forestry backed by strategies to minimise risk. 1990s, a shift to private investment as the dominant source of funds for plantation establishment and a strong emphasis on The series will describe tree improvement programs for the planting eucalypts. Australian plantation forests now occupy more-widely-planted softwood species: radiata pine (with about 1.75 million ha, two-thirds of which is softwood. separate papers covering achievements in Australia and New Zealand), hoop pine and the slash/Caribbean pine complex. Other In contrast to Australia, New Zealand plantations have relied softwoods to be examined include maritime pine in Western entirely on exotic species. These were planted on farms from Australia, and Douglas-fir and cypress in New Zealand. about 1860 for shelter, firewood and general amenity. Pilot- Achievements in hardwood tree improvement relate to scale planting for sawn timber began before 1900, and from plantations in both the temperate region (southern blue gum, around 1900 the state engaged in commercial afforestation, shining gum, eucalypts in New Zealand), and the subtropical often under the Justice Department to employ prisoners. Later region (blackbutt, Dunn’s white gum and Corymbia species and the state plantations became state forests, managed by the Forest hybrids). No account of Australian tree improvement would be Service. The first foresters favoured species such as Corsican complete without describing more recent work on species for pine, Austrian pine, European larch and Norway spruce from dry regions, species for saline conditions, and oil-producing Europe, although Douglas-fir, ponderosa pine and lodgepole pine species. An overview of forest tree improvement will be from North America, and some eucalypts, were also tried. Because presented at the end of the series. of its superior growth, however, radiata pine gradually became the preferred species. Eucalypts and other coniferous genera, The series commences in this issue with contributions on like cypresses, figured little until the 1970s when the NZ Forest programs for blackbutt, Dunn’s white gum, Corymbia species Service adopted a policy of developing small working circles and New Zealand Douglas-fir, and species for low-rainfall of special-purpose species. From 1970 onwards, one or two environments. companies established substantial eucalypt plantations for short- The Editor fibred pulp.
Australian Forestry 2007 Vol. 70 No. 1 p. 3 4 Eucalyptus pilularis tree improvement
Achievements in forest tree genetic improvement in Australia and New Zealand 1: Eucalyptus pilularis Smith tree improvement in Australia
Michael Henson1,2 and Helen J. Smith1
1Forests NSW, PO Box J19, Coffs Harbour Jetty, NSW 2450, Australia 2Email: [email protected]
Revised manuscript received 8 February 2007
Summary high-quality solid-wood products (Forests NSW MIS data). The volume of high-quality product was almost 20% of the total Eucalyptus pilularis Smith (blackbutt) is an important commercial product harvest for hardwoods from native forests and plantations plantation species on the north coast of NSW, with the first in the north coast area for 2005/2006. Regional Forest plantations being established in 1939. In 1997 Forests NSW Agreements (RFAs) between the NSW and Commonwealth established its first breeding population of E. pilularis and in the governments have led to the transfer of some north coast state subsequent 9 y the organisation has undertaken an aggressive forests to the national park estate. The reduction in area of improvement program for the species. Forests NSW has harvestable native forest has provided the impetus for plantations established 4.9 ha of clonal seed orchards of E. pilularis, and in to supply a greater proportion of the log quota needed to support parallel, is developing clones for commercial deployment through the industry in the future. selection and testing. The Forests NSW tree improvement program for E. pilularis places heavy emphasis on improving The first planting of E. pilularis established by Forests NSW the wood properties of young plantation-grown material so it can (formerly State Forests of NSW) was in 1939 in Whian Whian successfully replace wood from the native forest resource that is State Forest near Mullumbimby, and this was managed as a the feedstock for the established processing industry in the region. commercial plantation using silvicultural regimes to optimise growth, including being thinned in 1955 (Grant 1989). About Keywords: breeding programs; clones, seed orchards; wood density; 15 000 ha of E. pilularis plantations have been established by Eucalyptus pilularis; New South Wales Forests NSW, mainly in northern NSW. In recent years the program has aimed to supplement the supply of high-quality logs Introduction from native forests for the sawlog market. Eucalyptus pilularis has become an increasingly important species within this program Eucalyptus pilularis Smith (blackbutt) is a member of the due to its expected high wood quality. Since 1994, 4820 ha of subgenus Eucalyptus (the Monocalypts) section Pseudophloius E. pilularis have been established in new commercial plantations as defined by Brooker and Kleinig (1999). It occurs naturally in by Forests NSW (Forests NSW MIS data). In addition, sites coastal regions of New South Wales (NSW) and south-eastern originally planted by Australian Paper Manufacturers Ltd to Queensland from south of Bega, NSW, (37°30'S) to Mary- Eucalyptus grandis W.Hill ex Maiden, for pulp, are being borough, Queensland, (25°30'S), including Fraser Island, and at converted in the second rotation to plantations that specifically altitudes from near sea level to 600 m (Boland et al. 1984). It focus on log production for solid wood and veneer. Eucalyptus generally grows on gentle hill slopes between the sea and the pilularis is the species of choice for these plantations, and other coastal escarpment of the Great Dividing Range. It is a key species species are advised for planting only when factors site-limiting for high-quality log production from native forests in these for E. pilularis exist1. regions, and also a major plantation species within its natural range. The species is well known for site specificity — well Eucalyptus pilularis tends to suffer less leaf damage from insects drained soils of moderate to high fertility, and sites that are frost and fungal diseases than species in the subgenus Symphyomyrtus free with a reliable rainfall > 900 mm y–1 (Queensland Department (Simpson et al. 1997). This, along with its inherent wood of Primary Industries and Fisheries 2006). qualities, makes it an ideal species for growing in plantations in its natural range, where pest and disease pressures can be great. Eucalyptus pilularis is the principal species of hardwood timber The large range of insects that commonly attack Symphyomyrtus sawn in northern coastal NSW. It is a high-quality timber, with a number of uses including poles, sleepers, decking, flooring, furniture and building framework (Timber Development Association (NSW) Ltd 2003; Bootle 2005). In 2005/2006, about 1Kearney, D. (2005) Forests NSW North Coast hardwood plantation strategy 3 217 000 m of E. pilularis logs were harvested from state forests Part 1 — native forest operations estate. Forests NSW confidential report, Coffs in the northern coastal areas of NSW, of which over 60% was for Harbour, NSW.
Australian Forestry 2007 Vol. 70 No. 1 pp. 4–10 Michael Henson and Helen J. Smith 5 species (e.g. E. grandis and E. dunnii), often limit the commercial i) Growth — to optimise the productivity in terms of volume of viability of these species in plantations (Carnegie 2002). the plantation crop ii) Quality — to ensure that every plant established has the Deployment of E. pilularis into commercial plantations has been potential to produce an economic stem (i.e. one that is ‘fit for problematic. Seed production may be unreliable, with great purpose’) variation in the extent of flowering between sites, and large bud crops dropping buds before flowering (Florence 1964). Develop- iii) Adaptability — to reduce risk of economic losses through ment of seed for this species is lengthy, with the average time plantation failure, by breeding trees that are resistant or from bud initiation to flowering being 15 months or longer (Law tolerant to pests and/or diseases and are adapted to their target et al. 2000), and the interval between flowering and mature seed planting environment (e.g. frost and drought tolerant). being about 12 months (Florence 1964; Gunn 2001). Seed Forests NSW plans to complete a comprehensive breeding viability is often low, with average estimates at around 500–700 objectives study for eucalypt plantations and associated viable seeds per 10 g (Florence 1964; Gunn 2001). processing industries on the north coast of NSW in the next few Eucalyptus pilularis seedlings often have poor root quality, being years. One of the expected outputs of this study is economic affected by j-root and root coiling (Heagney et al. 2006). When weights for key traits for important commercial species, including seedlings are established in plantations, this phenomenon, in E. pilularis. conjunction with the shallow-rooting habit of the species, often The principal objective for improvement of E. pilularis will leads to instability, resulting in socketing, buttsweep and wind- always be increasing productivity of the forest estate by improving throw. While post-planting mortality is generally low, it may be growth and form traits, while maintaining desirable solid-wood up to 10% after heavy rainfall or in years of high winds (Carnegie properties. The weighting placed on quality and adaptability trait 1999). Plants derived from vegetative propagules, particularly objectives will vary depending on the species’ inherent wood minicuttings, have a root architecture that is different to that of properties and the exposure to risk in target environments. Given seedlings and do not have j-root deformation (Thomas and its high wood quality and the preference for the species by local Henson 2005). sawmillers, maintaining or improving solid-wood properties must Forests NSW commenced a limited tree improvement program have a high priority within the breeding program. Forests NSW, for E. pilularis in 1964 when 11 provenance trials were estab- in collaboration with Southern Cross University, is undertaking lished on the coast and northern escarpment of NSW (Johnson an extensive study of the wood properties of the E. pilularis and Stanton 1993). With the greater importance of E. pilularis breeding population, and the results will define the solid-wood as a plantation species in the post-1994 plantation program, a traits of importance for plantation-grown trees. pedigreed program was initiated in the mid-1990s to produce Recently, extending the planting range of E. pilularis has become superior genotypes for solid wood production2 (Johnson and an issue of importance as optimal sites for E. pilularis become Nikles 1997; Smith et al. 2004). Deployment issues have also rarer and less economic to plant, due to competition by other been addressed in recent years (Henson and Smith 2004; Smith land uses and the high land costs. Increasing the potential range et al. 2004). of sites for economically planting E. pilularis has become a key objective for the breeding program. Interspecific hybridisation Breeding objectives and strategy of selected E. pilularis clones with other members of the subgenus Eucalyptus — able to tolerate heavier and less well structured Breeding objectives clay soils and frosts — is being investigated to enhance its A breeding plan was written by Johnson and Nikles (1997) for adaptation to sites with these characteristics, while maintaining developing and deploying genetically improved material of the high-quality solid-wood product of E. pilularis. E. pilularis in NSW. The breeding objective was identified as ‘the efficient, cyclical establishment and management of a number Breeding strategy of breeding and propagation populations, for the accumulation The breeding strategy was initially one of genetic improvement of favourable genes, and the realisation of large genetic gains in within species, using a single population established from selected yield and quality of plantation sawlogs, poles and veneer logs on plus trees in native forests and older plantations. Forests NSW per-hectare, per-year and per-cost bases’. has a single breeding population consisting of open-pollinated The Forests NSW Tree Improvement Strategic Plan3 identifies families from 38 provenances, established in progeny trials on three general breeding objectives for all species under genetic three sites in northern NSW. improvement. These are: It is the target in the next 10 years that all Forests NSW commer- cial plantations of E. pilularis will be established using clones. To this end, two elite populations were selected and established at the Grafton Forest Technology Centre in 2005. Each population 2 Johnson, I.G. (2002) Blackbutt (E. pilularis) breeding seedling orchards — consists of ten clones from ten different provenances, and will growth and form at age 3 years, and selection for clonal seed orchards. State be used for controlled crosses, both within and between the elite Forests of NSW, confidential unpublished report, West Pennant Hills, NSW. populations, with the seed being used to establish clonal selection 70 pp. blocks. 3 Forests NSW Tree Improvement Program (2006) Strategic Plan 2006–2007. Forests NSW confidential internal report, Coffs Harbour, NSW. 6 pp.
Australian Forestry 2007 Vol. 70 No. 1 pp. 4–10 6 Eucalyptus pilularis tree improvement
The second generation of the breeding population will be Progeny trials established in 2009, with selected material from the three NSW open-pollinated progeny trials, the two elite populations, control- The progeny trials referred to in the breeding strategy section pollinated and open-pollinated clonal seed orchard seedlings, and were established during 1997 and 1998, on three Forests NSW with infusion of material from selected provenances. sites on the north coast of NSW (Fig. 1). These range from a high-quality fertile site with reliable rainfall near Hannam Vale Hybrids of Eucalyptus agglomerata Maiden and E. pilularis have (150 m asl), to a less fertile site near Clybucca (150 m asl), to a been found in the Forests NSW nursery, originating from seed more marginal site at Inglebar (295 m asl), an ‘inland’ site where collected from a thinned E. agglomerata progeny trial established drought is more likely to occur. Full site details are in Table 1. with an E. pilularis control seedlot (Johnson and Kaldor 1998). Exploitation of ‘natural’ hybrids as well the production of hybrids of selected E. pilularis clones with other stringybarks, through controlled pollination, is being investigated. Crossing of E. pilularis with E. agglomerata and Eucalyptus laevopinea R.Baker may produce germplasm tolerant of frost-prone sites, while hybridising with Eucalyptus pyrocarpa L.Johnson and Beerburrum Blaxell, Eucalyptus planchoniana F.Muell. and the white stringy- Mt Binga barks, including Eucalyptus nigra R.Baker and Eucalyptus tindaliae Blakely, may produce better growth on heavier clay soils. Field trials of hybrids between E. pilularis and E. planchoniana were established by Forests NSW as early as 1964.
Selecting and testing superior trees Inglebar
Selection of trees in base population Clybucca From late 1993 to 1995, intensive phenotypic selection of superior Hannam Vale trees was undertaken in high-quality natural stands and older commercial plantations within central and northern coastal NSW. The NSW south coast sites and sandy coastal sites such as Myall Lakes were not included in the selections due to their poor performance in previous provenance trials (Johnson and Stanton 1993). In all, 600 trees were selected, but resource restraints, logistical difficulties and natural factors reduced seed collection Figure 1. Provenances of Eucalyptus pilularis with families included to about 310 open-pollinated families from 34 provenances. in progeny trials (dots) and location of the five progeny trials (stars)
Table 1. Details of Eucalyptus pilularis progeny trial sites in New South Wales and Queensland
Attribute Hannam Vale Clybucca Inglebar Beerburrum Mt Binga State NSW NSW NSW Queensland Queensland Property Buttsworth LP Rolleston LP Inglebar Station Tripconys LA SF 611 Mt Binga SF 258 Latitude/Longitude 31º40'S,152º33'E 30º54'S,152º55'E 29º31'S,152º36'E 26º56'S,153º01'E 27º00'S,151º59'E Annual rainfall (mm) 1500 1260 1200 1505 941 Altitude (m) 150–170 150–170 285–305 15 600 Soils Deep yellow Deep red earths and Deep red and yellow Red lateritic kraznozem earths to yellow red podzolics, shallower podzolics podzolics yellow podzolics Planting date March 1997 March 1998 February 1998 March 1998 March 1998 Replicates 6 4 4 5 5 Design Row–column: Row–column: ICB, ICB, ICB, 18R × 17C 18R × 18C 22 blocks × 14 plots 95 blocks × 12 plots 95 blocks × 12 plots Plot size 4-tree row 4-tree row 4-tree row 4-tree row 4-tree row No. families 308 324 308 228 228 No. provenances 37 38 38 37 39 Total no. trees 7344 5184 4928 4560 4560 No. trees in control 120 48 48 60 0 seedlots No. trees in 7224 5136 4880 4500 4560 treatment seedlots
Australian Forestry 2007 Vol. 70 No. 1 pp. 4–10 Michael Henson and Helen J. Smith 7
Table 2. Family matrix for Eucalyptus pilularis progeny trials (number of families common between trials)
Site Site Hannam Vale Clybucca Inglebar Beerburrum Mt Binga Hannam Vale 308 265 258 135 151 Clybucca 324 308 166 171 Inglebar 308 160 162 Beerburrum 228 100 Mt Binga 228
The progeny trials incorporate seedling progeny from open- diameter. Joint analysis of assessment data from all five sites4 pollinated family seedlots, from 307 of the NSW selected plus has provided a maximum amount of information on the perform- trees and 52 plus tree selections from natural stands in Queens- ance of the families for both organisations. land. Five unimproved bulk seedlots were used as controls. Thirty- eight provenances are represented within the breeding population Growth and form traits in the Hannam Vale trial were reassessed with a maximum of 324 families at any one site. Each family is in October 2005 at age 8 years 8 months. Subsequently a range represented by either four or six replicates of four-tree plots. of wood properties was assessed by non-destructive methods on The trials are designed as incomplete blocks with a row–column selected trees, including Pilodyn penetration (all trees), acoustic layout. velocity, peripheral growth strain, circularity of the stem at breast height and crown diameter. A total of 1068 selected trees were Queensland Department of Primary Industries and Fisheries, cored for assessment of radial and tangential shrinkage. In August Horticulture and Forestry Science (then Queensland Forest 2006 the Hannam Vale trial was thinned to the best tree per plot. Research Institute) also established two progeny trials at Of the trees felled, 1116 were selected for detailed analysis of Beerburrum and Mt Binga (Fig. 1, Table 1) in 1998, containing wood properties, including spiral grain, acoustic stiffness and many of the families included in the NSW trials (Table 2). A heartwood:sapwood ratio. Five hundred and sixteen trees were total of 356 families drawn from the species range was planted, sent to Forest Enterprises Australia’s Hew Saw in Tasmania for including 248 families from NSW State Forests and 108 families commercial milling studies. The other 600 trees were from the south-eastern and south-western regions of Queensland destructively assessed for wood properties including shrinkage, (Collins 1998). One hundred families are common to both trials. collapse, basic density, hardness, modulus of elasticity, modulus At both sites, the progeny trial is planted in five replicates of of rupture, colour, cell wall thickness and wood chemistry four-tree line plots using a randomised incomplete block design (cellulose, lignin and hemi-cellulose). Leaf samples from all 1116 (Collins 1998). destructively sampled trees were collected for use in a molecular genetic association study seeking links between genes and wood Other resources properties.
In addition to the breeding populations, the two main NSW sites It is planned to assess the Clybucca progeny trial in 2007 for at Hannam Vale and Clybucca were planted with: growth and form, and to destructively assess the wood properties of a limited number of families. It is hoped that this will provide • Family selection blocks (FSB) of the expected best 115 NSW data on the effect of site on wood quality as well as on the and 10 Queensland families, to provide superior trees for clone importance of genotype × environment interaction for growth development and testing and wood property traits. • Extensive seedling seed orchards (ESSO) from bulked OP seedlots from 100 phenotypically superior NSW selections, planted for early production of improved seed Genetic parameters and selection criteria • Provenance resource stands (PRS) from bulked seedlots or Heritabilities for dbhob (3 y) on the three NSW sites ranged from provenances not represented in the breeding populations, to 0.18 to 0.24, with cross-site genetic correlations ranging from provide genetic diversity and an infusion population for the 0.73 to 0.90 between the three sites. Straightness at 3 y was second generation of the breeding populations (Johnson and slightly more heritable, ranging from 0.20 to 0.34 for the NSW Nikles 1997). sites, with cross-site genetic correlations from 0.83 to 0.98.
The ESSOs have not been used to date, although the FSBs at At 38 months the progeny trial at Clybucca was assessed for both Hannam Vale and Clybucca have been a source of superior severity of Mycosphaerella leaf disease, caused by Mycosphaerella trees propagated for clone development. cryptica (Cooke) Hansf. and Mycosphaerella marksii Carnegie
Assessment of trials
All three progeny trials in NSW and the two in Queensland were 4 Jarvis, S. (2003) Analysis of Eucalyptus pilularis trials. Commercial-in-con- assessed at 3 y for growth and form traits including height, fidence report for Queensland Forestry Research Institute and State Forests of diameter over bark at breast height (dbhob), stem form and branch NSW. Balmoral, Victoria. 25 pp.
Australian Forestry 2007 Vol. 70 No. 1 pp. 4–10 8 Eucalyptus pilularis tree improvement
Table 3. Estimated genetic gains for traits in Forests NSW Eucalyptus pilularis clonal seed orchards
Backward Forward Genetic gains (%) Site No. clones selects selects Height Diameter Straightness Urunga Stage 1 55 28 27 07.6 5.4 06.0 Urunga Stage 2 43 05 38 08.6 5.6 22.4 Urunga D block 09 00 09 08.8 6.6 26.4 Grafton 46 07 39 08.6 5.8 21.8 Elite 1 10 07 03 10.8 7.2 30.4 Elite 2 10 10 00 11.6 8.2 28.4
and Keane, and target spot, caused by Aulographina eucalypti Eucalyptus pilularis CSOs have been established by Forests NSW (Cooke and Mass.) von Ark and Muller, at the same time as growth on two sites: a coastal site near Urunga, south of Coffs Harbour, and form were assessed. These are the most damaging leaf fungi and a more inland site at Forests NSW Grafton Forest Technology in the NSW E. pilularis plantations (Stone et al. 1998). Carnegie Centre. Selection of superior individuals for the Stage 1 planting et al. (2004) found significant variation in disease and defoliation of the first orchard (Urunga) in November 2002 was based on among the provenances and families tested. The individual the 3-y growth data of the progeny trials at Hannam Vale and heritability estimates were moderate for Mycosphaerella damage Clybucca, and the form data for Hannam Vale. In all, 55 clones (h2 = 0.38), and low for defoliation (0.22) and Aulographina were planted in the CSO, including both original plus trees and damage (0.13). progeny trees from the breeding population. These trees were unrelated with only either parent or one progeny selected from 4 Jarvis developed a selection index based on volume, branch the better-performing trees2. thickness and stem diameter. Trees and parents from the five trials were ranked using this index, and selections were made for clonal Further selections were made after joint analysis of data from all seed orchards discussed later in this paper. five progeny trials in NSW and Queensland4. Stage 2 of the orchard was planted in two distinct blocks in March 2004; the Deployment strategy first block contains 38 forward selections and five backward selects and the second block (D) contains nine clones that are The long-term deployment strategy for E. pilularis in Forests the best forward selections from eight different provenances. NSW plantations is based on clonal forestry. In the interim, genetically improved seed is being produced in clonal seed CSOs are being established at multiple planting sites to reduce orchards for deployment as seedlings or as a source of seedling- the risk of loss of seed production and genetic material due to based mother plants for propagation of plants by cuttings. unforeseen circumstances or extreme weather conditions, such Improved material from the breeding program for E. pilularis as fire or hail damage. A second CSO was planted at the Grafton has been deployed by several methods, and the strategy is evolving Forest Technology Centre in April 2005. Ramets of another 46 as more improved germplasm and technologies are developed. clones were included in this orchard. In addition, two elite populations, each composed of ten clones from ten different provenances, were established for use in further breeding work, Deployment of plus tree seed particularly in the controlled crossing program. Predicted genetic Forests NSW commercial plantations were traditionally gains for the clonal seed orchards are presented in Table 3. established using bulked collections of preferred native forest The total CSO area for E. pilularis is currently 4.9 ha. Flower provenances. Identification of superior families in the breeding bud development was obvious in some ramets in the 2002 Stage 1 population2 allowed a strategy of ‘backward selection’ of parent planting at Urunga within 15 months of planting and, at trees to act as a source of open-pollinated plus-tree seedlots, when 24 months, 64% of the ramets were in bud. Flowering commenced seed was available. The genetic gain of commercial planting stock in 2005 and with it, intensive multi-season studies of the flowering produced from this source, over the bulked native collections, phenology of the clones. Ramets in Stage 1 of the Urunga orchard has been estimated to be 4.3% (Johnson and Nikles 1997). again commenced flowering in June 2006. About 71% of ramets were in bud, but some ramets aborted their buds prior to flowering, Clonal seed orchards a phenomenon described by Florence (1964). However, the first commercial crop of open-pollinated seed was harvested over Clonal seed orchards (CSOs) can produce seed with substantial summer 2006/2007. In addition, a small number of controlled gains over native seed. Although a CSO is expensive to establish, pollinations were successful and this seed is being grown for the time to seed production is greatly reduced compared to a further testing. seedling seed orchard (SSO). Grafts of scions from mature parts of the tree tend to flower within one or two seasons (Eldridge et The current assessment of the E. pilularis breeding population al. 1993); a SSO may take up to 8 y to flower (Johnson and for wood properties will be the basis for making further selections Nikles 1997).
Australian Forestry 2007 Vol. 70 No. 1 pp. 4–10 Michael Henson and Helen J. Smith 9 and roguing the CSOs to remove inferior clones. Grafting of 30 families identified from progeny trials), and field trials will ramets to establish a second CSO at Grafton, based on sawn- be established on two sites in autumn 2007 and a third site in wood properties, will take place in 2007/2008. spring 2007.
The clone trial established in 1999 near Coffs Harbour, and Family forestry composed of phenotypic selections from young plantations, was Family forestry is the mass vegetative propagation of seedlings assessed for height, dbhob and form at age 32 months (Johnson of selected pedigreed families, without testing and identification and Smith 2003) and five clones selected for further testing and of individual clones (Eldridge et al. 1993). The advantages of limited commercial deployment. At age 55 months, assessment clonal forestry have been well documented (Eldridge et al. 1993) of height, dbhob, straightness and solid wood properties, and include uniformity of both the plantations for management particularly density, stiffness and grain angle, resulted in selection and the harvested wood product, the opportunity to match clones of an additional five candidate clones for further testing and pilot- to sites and the ability to rapidly capture both additive and non- scale commercial deployment. Fifteen clones from the trial were additive genetic variation. Family forestry provides greater destructively assessed for wood property traits at 65 months. genetic diversity in the planting stock than clonal forestry, a These partially tested clones have been established as potted potential advantage for counteract the site specificity exhibited mother plants, along with the seedling-based mother plants for by E. pilularis clones when planted on NSW sites that may be the family forestry program. The productivity (rooting success sub-optimal for growth. Forests NSW has undertaken a program of harvested cuttings coupled with mother plant vigour) of these of deployment of genetically improved material by family forestry mother plants is far inferior when compared to that of the seedling- (Henson and Smith 2004; Smith et al. 2004). Propagation of based clones (Henson and Smith 2004) and maturity issues affect E. pilularis by minicuttings produces plants with better root their potential for plant production. To address these issues, a architecture than those propagated from seed (Thomas and study of the nutritional aspects of mother plant management is Henson 2005), while also providing greater genetic gain in the planned. planting stock. Acknowledgements Seed was collected, where possible, from the original ortets of the 30 families identified as the best-performing families over Thanks are due to Ian Johnson (NSW Department of Primary the five progeny trials. Potted seedling-based mother plants are Industries, Science and Research) for the establishment of the raised in flood-and-ebb trays, and flood irrigated with defined progeny trials and early assessment of these trials, and to Steve nutrient solutions to promote shoot development and rooting of Boyton (Forests NSW) for organising and coordinating later cuttings. Apical minicuttings are set to propagate planting stock, assessments of the trials and for providing data management and a strike rate of over 90% is achieved. Since initiation of the services. The authors would also like to thank the Forests and family forestry program, over 80 000 plants of E. pilularis have Wood Products Research and Development Corporation been propagated for commercial planting by Forests NSW and (FWPRDC), the Australian Centre for International Agricultural private forestry companies. Research (ACIAR) and Forests Enterprises Australia (FEA) for their support of the wood quality assessment of the Hannam Vale Currently E. pilularis seed produced by control pollinations progeny trial. Thanks to Southern Cross University and Ensis between selected parents is being grown for the family forestry Wood Quality for their assistance in the wood quality assessment program. of the Hannam Vale trial.
Clone development and testing References The first clones of E. pilularis were developed by Forests NSW Boland, D.J., Brooker, M.I.H., Chippendale, G.M., Hall, N., Hyland, in 1994 from coppice on stumps of harvested trees in old planta- B.P.M., Johnston, R.D., Kleinig, D.A. and Turner, J.D. (1984) tions and native forests. Plants of nine clones propagated from Forest Trees of Australia. 4th edn. CSIRO Publishing, Colling- two sites and seedling-based clones from six routine commercial wood, Victoria, Australia, 700 pp. seedlots were planted in the first clonal trial in Pine Creek State Bootle, K.R. (2005) Wood in Australia: Types, Properties and Uses. Forest in 1995. More of these clones were planted in 1997 on the 2nd edn. McGraw-Hill Australia, Sydney, Australia, 452 pp. same site as the progeny trial at Hannam Vale. Brooker, M.I.H. and Kleinig, D.A. (1999) Field Guide to Eucalypts. Volume 1 South-eastern Australia. 2nd edn. Bloomings Books, The identification of the better-performing families within the Hawthorn, Victoria, Australia, 353 pp. E. pilularis breeding population has allowed the development of Carnegie, A.J. (1999) Blow-down in young Eucalyptus pilularis improved clones for field testing. Seedling-based clones have plantations in New South Wales. Unpublished report, State Forests been developed from 40 plus-tree families and field-select clones of NSW, Forest Health Survey Unit, Locked Bag 23, Pennant from ‘family selection blocks’ (FSBs). These clones have been Hills, NSW 2120, 6 pp. supplemented by field-select clones developed from phenotypic Carnegie, A.J. (2002) Field Guide to Common Pests and Diseases in selections in young plantations within the Forests NSW estate. Eucalypt Plantations in NSW. State Forests of NSW, Research and Development Division, West Pennant Hills, NSW, Australia, Forests NSW currently has 169 E. pilularis seedling-based and 80 pp. field-select clones being tested over three sites, one planted in 1999 and two in 2002. An additional 493 candidate clones have been developed from FSBs and the family forestry program (top
Australian Forestry 2007 Vol. 70 No. 1 pp. 4–10 10 Eucalyptus pilularis tree improvement
Carnegie, A.J., Johnson, I.G. and Henson, M. (2004) Variation among Johnson, I.G. and Stanton, R.R. (1993) Thirty Years of Eucalypt Species provenances and families of blackbutt (Eucalyptus pilularis) in and Provenance Trials in New South Wales: Survival and Growth early growth and susceptibility to damage from leaf spot fungi. in Trials Established from 1961 to 1990. Forestry Commission of Canadian Journal of Forest Research 34, 2314–2326. NSW, Research Division, Research Paper No. 20. Sydney, NSW, Collins, S. (1998) Establishment of breeding seedling orchards of Australia, 92 pp. blackbutt (Eucalyptus pilularis) at Beerburrum and Mt Binga. Law, B., Mackowski, C., Schoer, L. and Tweedie, T. (2000) Flowering Commencement report Experiment 450/1 HWD. Queensland phenology of myrtaceous trees and their relation to climatic, Forest Research Institute, Gympie, Queensland, 20 pp. environmental and disturbance variables in northern New South Eldridge, K., Davidson, J., Harwood, C. and Van Wyk, G. (1993) Eucalypt Wales. Austral Ecology 25, 160–178. Domestication and Breeding. Oxford University Press, New York, Queensland Department of Primary Industries and Fisheries (2006) USA, 288 pp. Hardwoods Queensland Website http://www2.dpi.qld.gov.au/ Florence, R.G. (1964) A comparative study of flowering and seed hardwoodsqld/7706.html. production in six blackbutt (Eucalyptus pilularis Sm.) forest Simpson, J., Stone, C. and Eldridge, R. (1997) Eucalypt Plantation stands. Australian Forestry 28, 23–33. Pests and Diseases — Crop Loss Study. State Forests of NSW, Grant, T.C. (1989) History of Forestry in New South Wales 1788 to Forest Research and Development Division, Research Paper No. 1988. Star Printery, Erskineville, NSW, 320 pp. 35, Sydney, NSW, 47 pp. Gunn, B. (2001) Australian Tree Seed Centre Operations Manual. Smith, H.J., Johnson, I.G. and Henson, M. (2004) Improved family CSIRO Forestry and Forest Products, Canberra, Australia, 150 pp. forestry — the evolution of breeding and deployment strategies for Eucalyptus pilularis in New South Wales, Australia. In: Heagney, G., Thomas, D. and Henson, M. (2006) Alarming levels of J- Bailian, L. and McKeand, S. (eds) Forests Genetics and Tree root formation in eucalypt species planting stock. Unpublished Breeding in the Age of Genomics: Progress and Future. report, Forests NSW, PO Box J19, Coffs Harbour, NSW 2450, Proceedings of the IUFRO Joint Conference of Division 2, 1–5 20 pp. November, Charleston, South Carolina USA. North Carolina State Henson, M. and Smith, H. (2004) Borrowed technologies: adaption of University, North Carolina, USA, pp. 113–126. family forestry for eucalypts in New South Wales, Australia. In: Stone, C., Simpson, J.A. and Eldridge, R.H. (1998). Insect and fungal Borralho, N.M.G., Pereira, J.S., Marques, C., Coutinho, J., damage to young eucalypt plantings in northern New South Wales. Madeira, M. and Tomé, M. (eds) Eucalyptus in a Changing World. Australian Forestry 61, 7–20. Proceedings of the IUFRO Conference, 11–15 October, RAIZ, Instituto Investigação da Floresta e Papel, Aveiro, Portugal, p. 103. Thomas, D. and Henson, M. (2005) Comparison of cuttings and seedlings — some issues and recent findings. Unpublished report, Johnson, I.G. and Kaldor, C.J. (1998) Growth and form of Eucalyptus Forests NSW, PO Box J19, Coffs Harbour, NSW 2450, 6 pp. agglomerata provenances in New South Wales. Australian Forestry 61, 53–62. Timber Development Association (NSW) Ltd (2003) Australian Hardwood and Cypress Manual. Australian Hardwood Network, Johnson, I.G. and Nikles, D.G. (1997) Plan for Developing and http://www.australianhardwood.net. Deploying Genetically-Improved Varieties of Blackbutt (Eucalyptus pilularis Smith) in New South Wales. State Forests of NSW, Forest Research and Development Division, Technical Paper No. 63, Sydney, NSW, 101 pp. Johnson, I. and Smith, H. (2003) Growth and form of Eucalyptus pilularis field select clones and seedlings near Coffs Harbour, at age 34 months. Unpublished report, State Forests of NSW, Locked Bag 23, Pennant Hills, NSW 2120, 11 pp.
Australian Forestry 2007 Vol. 70 No. 1 pp. 4–10 David J. Lee 11
Achievements in forest tree genetic improvement in Australia and New Zealand 2: Development of Corymbia species and hybrids for plantations in eastern Australia
David J. Lee
Department of Primary Industries and Fisheries, Horticulture and Forestry Science, Locked Bag 16, Fraser Road, Gympie, Queensland 4570, Australia Email: [email protected]
Revised manuscript received 16 November 2006
Summary The spotted gums and cadaga belong to the genus Corymbia, one of the three genera encompassing all eucalypt species. A This paper describes the establishment of provenance seedling recent taxonomic revision of the spotted gums by McDonald and seed orchards of three spotted gums and cadaga (all species of Bean (2000) is followed throughout this paper. These authors Corymbia ex Eucalyptus). It also discusses the limitations of recognised four distinct spotted gum taxa: C. citriodora subsp. growing the spotted gums as pure species including: lack of mass citriodora (CCC; lemon-scented gum); C. citriodora subsp. flowering, susceptibility to a fungal shoot blight and low variegata (CCV; spotted gum); C. henryi (CH; large-leaved amenability to vegetative propagation. These limitations, together spotted gum); and C. maculata (CM; southern spotted gum). with observation of putative natural hybrids of the spotted gums Corymbia torelliana (CT; cadaga) is the other species of interest. with cadaga, and the early promise of manipulated hybrids, led The distributions of these species are described in Hill and to an intensive breeding and testing program. Many hybrid Johnson (1995). families have significant advantages in growth and tolerance to disease, insects and frost, and can be vegetatively propagated. This paper focuses on the development of breeding populations They also exhibit broad environmental plasticity, allowing the of the three spotted gum species that occur naturally in best varieties to be planted across a wider range of sites than the Queensland (CCC, CCV and CH), the establishment of a CT spotted gums, resulting in more land being suitable for plantation base population, and the development and early performance of development. hybrids between CT and the spotted gum species in Queensland. Reference is also made to tree improvement with Corymbia in Keywords: plantations; tree breeding; progeny; hybrids; eucalypts; NSW and in southern Australia. Corymbia; Corymbia torelliana; Corymbia citriodora; Corymbia henryi; Corymbia variegata; Ramularia; Quambalaria piterika Seed orchards and genetic bases of Corymbia Introduction species established in Australia
In Australia, there has been rapid expansion in recent years of There are three tree improvement programs for spotted gum in commercial plantations of hardwood species, especially of Australia. In NSW the focus in on CCV (Johnson et al. 2004), in Eucalyptus (Bureau of Rural Sciences 2005). In the tropics and southern Australia the Australian Low Rainfall Tree Improvement subtropics, the land most readily available for hardwood planta- Group (ALRTIG) program focuses on CM and CCV (Harwood tion development is in the 700–1000 mm mean annual rainfall et al. 2001, 2007), and in Queensland the emphasis is on CCV zone where, potentially, millions of hectares of plantations could (Lee et al. 2004) and Corymbia hybrids. Seed and information be planted on cleared land (Lee et al. 2001). Based on experience exchanges have taken place, but since each of these programs is gained in trial plantings in Queensland (Lee et al. 2001, 2005), targeting taxa for different edaphic and climatic conditions, spotted gums (Corymbia species) and hybrids between spotted integration may not be practicable. gums and cadaga (C. torelliana) were identified as promising The method used to initiate the spotted gum tree improvement taxa due to their superior survival and growth across a broad program in Queensland has been to establish successive range of edaphic and climatic conditions. The choice of spotted provenance seedling seed orchards (Nanson 1972). Provenances gums was also supported by the fact that they are the most (and families within provenances) were planted in 3–5-tree line commonly harvested native hardwoods in Queensland and have plots, which could be thinned to retain the best individuals for excellent wood properties (Underhill and Watts 2004). In northern seed production and as sources of select trees for establishment New South Wales (NSW), the spotted gums are also very of grafted seed orchards and breeding populations. This approach important species for plantation development (Bruskin 1999; was adopted because it (i) had the potential to quickly provide Johnson et al. 2004), and C. maculata has a role in farm forestry genetically improved planting stock to the plantation industry; in some low-rainfall areas of southern Australia (Harwood et (ii) allowed selection for different product streams (solid wood al. 2007).
Australian Forestry 2007 Vol. 70 No. 1 pp. 11–16 12 Corymbia species and hybrids for eastern Australia and pulpwood); (iii) provided information about genetic variation a natural disease of Corymbia and Angophora caused by the for important traits (phenology, growth and form plus pest and fungal pathogen Quambalaria pitereka (J.Walker & Bertus) disease resistance); and (iv) facilitated long-term genetic J.A.Simpson 2000 basionym: Ramularia pitereka, Sporothrix improvement of the species at a reasonable cost. pitereka and Sporotrichum destructor. Subsequently the seedling seed orchards showed significant variation between provenances Since the program began in 1996, broad genetic bases of the and families in tolerance to QSB. Hence as a basis of a plantation spotted gum species occurring in Queensland have been planted program it was possible to focus on provenances from the Gympie in provenance seedling seed orchards (PSSOs); these include 43 region in Queensland that had greater tolerance to the disease natural provenances of CCV, 11 provenances of CCC, 7 of CH (Lee et al. 1998). This reinstated the spotted gum plantation and 6 of CT (Table 1). Smaller numbers of superior trees may be programs. The plantation estate currently occupies 18 000 ha selected from these plantings to form sublines, each for (primarily CCV) across NSW and Queensland, with an annual improvement by within-family, recurrent selection over several planting program of about 2500 ha. generations (Shelbourne et al. 1991). All species and tested provenances of spotted gum are somewhat The improvement program with CCV in Forests NSW has susceptible to QSB (Dickinson et al. 2004), with severity at a followed a similar path to that in Queensland. Provenance specific site being closely associated with local climatic seedling seed orchards were planted in 1999, 2001 (two sites) conditions: the higher the annual rainfall the greater the disease and 2004 (four sites) (Johnson et al. 2004), and subsequently in incidence. No QSB-tolerant provenance of C. henryi has been 2006 (four sites). Two clonal seed orchards have also been found. Dickinson et al. also found that a provenance’s mean established, the first (with 39 clones) includes 37 clones of annual rainfall was a good indicator of the potential QSB Queensland origin, reflecting the superior Quambalaria shoot tolerance; high-rainfall provenances of CCV, such as Woondum blight tolerance and other desirable traits of some Queensland (south-east of Gympie), were more tolerant of QSB than low- provenances. Several clone tests have also been established rainfall provenances such as Presho, 120 km north of Roma including some semi-commercial clonal plantations (M. Henson, (Queensland). The most tolerant provenances, however, can have Forests NSW, pers. comm. 2005). up to 30% of the trees severely damaged by QSB under conditions favourable to the disease. This disease risk has great implications Impact of Quambalaria shoot blight on the spotted when deciding to grow spotted gum for either pulpwood or gums and C. torelliana sawlogs; many trees affected by QSB do not recover, and therefore reduce the effective stocking of the plantation. In contrast to the Planting programs using spotted gums began in the mid-1990s spotted gums, CT is recorded as not being susceptible to QSB in NSW and Queensland. However, these programs were curtailed — under natural conditions, or in trials (Self et al. 2002) or field following heavy losses due to Quambalaria shoot blight (QSB), planting (unpublished data).
Table 1. Genetic bases of Corymbia species established in provenance seedling seed orchards in Queenslanda
Details of PSSOs Species Trial identity Provenances and Latitude Longitude Rainfall (seed parentsb) (degrees) (degrees) (mm y–1) C. citriodora subsp. variegata 451 A HWD 12 (132) 25.78 152.83 1295 451 B HWD 22 (214) 27.50 152.26 0839 451 C HWD 15 (148) 25.77 152.63 1295 451 D HWD 16 (266) 25.76 152.67 1295 451 E HWD 09 (158) 25.51 152.31 1000 451 G HWD 12 (188) 25.67 152.52 0900 451 H HWD 19 (286) 25.52 151.47 0770 397 A HWD 010 (56) 24.28 151.50 1033 Total 43 (719) C. citriodora subsp. citriodorac 782 A TBS 090 (88+) 23.32 150.26 0763 683 HWD 05 (338+) 24.56 149.97 0686 Total 11 (388+) C. henryi 456 HWD 070 (47) 25.78 152.84 1295 C. torelliana 518 A HWD 050 (53) 17.13 145.43 1014 518 B HWD 050 (53) 17.33 145.50 1413 518 C HWD 05 (101) 26.99 152.00 0926 Total 06 (113)
aAdditional genetic resources for all species are available in taxa trials. bThis includes some bulks, so the number of seed parents represented in the PSSOs would be smaller than the number indicated here due to the limited sample planted. c782 A TBS is owned by a joint venture between Stanwell Corp and Forest Plantations Queensland.
Australian Forestry 2007 Vol. 70 No. 1 pp. 11–16 David J. Lee 13
Why develop Corymbia hybrids? Performance of the Corymbia hybrids relative to the spotted gums The spotted gum species tested to date are not amenable to propagation as rooted cuttings. Success rates of 10% rooted In replicated trials, many hybrid families have significantly out- cuttings has been achieved for CM (McComb and Wroth 1986), performed the best spotted gum provenance (CCV Woondum) 0.4–3.6% for CCC, CCV and CH (Catesby and Walker 1997) in height and diameter growth (P < 0.001; Fig. 1). Near Kingaroy and 2.5% for CCC (de Assis 2000). Thus the only current practical on a Red Ferrosol (Isbell 1996) soil the CCV Woondum option to commercially deploy improved germplasm from the provenance bulk seedlot reached 9.0 m in height and 7.8 cm in breeding program will be via seed. However, in the spotted gum diameter at height breast over bark (dbhob) at age 54 months. At seed orchards, heavy synchronous flowering has not occurred in this site, the top four Corymbia hybrid families CT × CCV135, the last 9 y. This mirrors poor flowering in natural stands of CT × CCV143, CT × CCC27, and CT × CH147 had heights of desired provenances of spotted gum, which in turn has limited 9.7–10.8 m (8–20% selection differential at this site), and dbhob the development of hardwood plantations in Queensland and of 9.6–10.9 cm (23–39% selection differential at this site). Most NSW as spotted gums (particularly CCV) are the species most of the hybrids were also significantly taller and larger in dbhob suited to a large portion of the sites available for plantation than the CT parental controls. development. In a trial near Gympie, frost resistance also varied significantly Interest in Corymbia hybrids stemmed from observations that among taxa at nine months (P < 0.001). Here the Corymbia hybrid amenity plantings of CT often included hybrids between CT and families had similar frost resistance to the CT parental controls, spotted gums. These hybrids had the good form of spotted gum, and both CT and the Corymbia hybrids (CT × CCC, CT × CCV were intermediate in some traits, and did not appear to be attacked and CT × CH) had living shoots significantly higher on the stem by QSB. Further, as CT (de Assis 2000; Lee et al. 2005) and its after frosting than the CCV Woondum bulk seedlot (Fig. 2). Other hybrids with CCC (de Assis 2000) were amenable to propagation traits with a similar pattern — where the better hybrid families as rooted cuttings, it was proposed that hybrids with the other show a clear benefit relative to the parental controls or follow spotted gum species might also be propagated commercially in the high parent for the trait — are detailed in Table 3. Commer- this manner. Another factor encouraging the development of cially the most important of these are height and diameter growth, Corymbia hybrids is that the wood properties of the parental frost tolerance, straightness, QSB tolerance, erinose mite species are similar; suggesting that wood of the hybrids would tolerance, and the seedling or coppice rootability of the Corymbia be just as acceptable. hybrids. Another important observation is that many Corymbia hybrid families are stable across disparate sites, for example, in A Corymbia hybrid controlled-cross-pollination program was three trials with 64–76 hybrid families under test, seven families initiated in 1999 and the first trial of Corymbia hybrids was ranked in the top 20 for volume across all sites at age 3 y (well planted in May 2001. Since then 18 progeny trials have been ahead of the CCV Woondum bulk seedlot; data not presented). planted at locations extending from Grafton in northern NSW (in conjunction with Forests NSW) to Mareeba in northern When all the benefits of the hybrids are considered relative to Queensland. In these trials 283 Corymbia hybrid families are the spotted gums and other taxa with plantation potential for being evaluated against pure species controls. Of these crosses, Queensland, it is clear that the focus of the breeding program 272 have CT as the female parent (Table 2), as the seed yield is must be on the Corymbia hybrids. Forests NSW is also attempting up to four times higher when the cross is made in this direction to produce hybrids and is studying the potential of hybrids in than when a spotted gum is used as the female parent (unpublished conjunction with the Queensland program. data).
Table 2. Number of Corymbia hybrid families being evaluated in progeny trials in Queensland and NSW
Male parent ( = pollen parent in cross) Female parent C. citriodora C. citriodora C. henryi C. maculata Corymbia hybrida subsp. variegata subsp. citriodora C. torellianab 162 36 30 15 29 aThese are spontaneous Corymbia hybrids and their pedigree is therefore unknown. Some of these hybrids have been used in crosses as the female parent. bHybrids with spotted gums as the female parent have been developed with limited success. These include five with C. citriodora subsp. variegata and two with C. henryi as the female parent.
Australian Forestry 2007 Vol. 70 No. 1 pp. 11–16 14 Corymbia species and hybrids for eastern Australia
Current status of Corymbia hybrid breeding from both plantations and seed orchards. As a broad genetic base program and future development of all spotted gum species has been developed (including collaboration with Forest NSW and Ensis), and genetic param- Breeding strategy eters can be estimated from progeny trial data, it is time to devise appropriate breeding strategies for the pure species and hybrids. The first series of hybrid progeny trials planted 2001–2006 in This important matter is now being addressed. In the meantime, Queensland were developed in an ad-hoc breeding program using developing grafted clone banks and seed orchards to facilitate CT trees selected at low intensity in amenity plantings. These breeding work are priorities. These will also facilitate a deployment were crossed with phenotypically selected CCC, CCV and CH
14 Height Dbhob 12
10
8
6
4
Mean height (m) and diameter (cm) 2
0 CT15 CT130 CT131 CT165 CTxCH147 CTxCCV22 CTxCCC27 CTxCCV135 CTxCCV143 CTxCCV138 CTxCCV140 CTxCCV146 CTxCCV145 CTxCCV134 CTxCCV137 CTxCCV139 CTxCCV151 CTxCCV152 CTxCCC136 CCV Woondum
Figure 1. Fifty-four-month height (5% LSD = 2.9) and diameter at breast height over bark (5% LSD = 3.6 of the top 15 of 22 Corymbia hybrid families ( CT × CCV, CT × CCC, CT × CH) ranked by mean dbhob) relative to CCV Woondum bulk (85-tree bulk) and C. torelliana (CT) controls in a trial on a Ferrosol soil near Kingaroy (780 mm MAR). Bars indicate standard errors.
2.5
2.0
1.5
1.0
0.5 Mean height (m) to first green shoot 0.0 CT15 CT131 CT169 CT236 CT165 CTxCH147 CTxCH160 CTxCCC27 CTxCCV134 CTxCCV135 CTxCCV139 CTxCCV146 CTxCCV148 CTxCCV156 CTxCCV159 CTxCCV162 CTxCCV163 CTxCCV230 CTxCCV235 CTxCCC136 CV Woondum C
Figure 2. Height to highest living shoot following severe frosts at Amamoor (5% LSD = 0.53) of top 15 of 38 Corymbia hybrids families (CT × CCC (dark grey), CT × CCV (black), CT × CH (light grey)) relative to CCV Woondum bulk seedlot (85-tree bulk, hatched) and CT (white) ranked by mean dbhob at age 3 y in a trial on a black Dermosol soil near Gympie (1090 mm MAR). Bars indicate standard errors.
Australian Forestry 2007 Vol. 70 No. 1 pp. 11–16 David J. Lee 15
Table 3. Relative performance for key traits of C. torelliana (CT), C. citriodora subsp. variegata (CCV) and Corymbia hybrids based on observations in 3-y-old progeny trials in Queensland (Mareeba to Gympie)
CT CCV Corymbia Trait (Kuranda provenance) (Woondum provenance) (F1) hybrid families Quambalaria shoot blight tolerance Very high (immune?) Moderate High Erinose mite tolerance Very high (immune?) Very low High Red-shouldered leaf beetle tolerance Low High Moderate to high Longicorn beetle tolerance Moderate High Moderate to high Height and diameter growth Low to moderate Low to high Low to very high Straightness Moderate Moderate Moderate to high Taper (low desirable) High Low Moderate Branch quality, including shedding Low High Moderate to high Canopy density (high results in good site capture) High Low to moderate Low to high Frost hardiness High Low to moderate High Environmental plasticity High Moderate High Seedling or coppice rootability Moderate to high Low Moderate to high Wood properties Unknown in plantations Showing potential for Showing potential for sawlogs and pulpwooda sawlogs and pulpwooda
aBased on unpublished results of tests carried out on 42-month-old hybrids and CCV.
crossing program based on selecting and reproducing the top for commercial deployment. These clones will be the first release hybrid families identified in the progeny trials. All work from the 7-y-old hybrid breeding program. Ongoing selection of associated with the breeding of hybrids will be integrated with families and clones for operational deployment, using more studies of silviculture and wood properties. information about growth, form, wood properties, rootability, and pest and disease resistance is planned. Deployment strategy Conclusions In order to deploy hybrids operationally, the best families or clones must be amenable to vegetative propagation as it is The development of the Corymbia hybrids is likely to underpin impractical to produce sufficient hybrid seed for direct establishment of a sustainable hardwood plantation sector in commercial use. There are two options for commercial Queensland and NSW, by providing germplasm with superior deployment of hybrids: (i) vegetative family forestry where early growth rates, pest and disease tolerance, environmental seedlings from hybrid families selected for superiority and plasticity and wood properties similar to those of the current best stability across sites are bulked up as rooted cuttings (Nikles alternatives (the spotted gum species). The breeding program is 1992; Henson and Smith 2004); and (ii) clonal forestry where in its infancy: field tests need to be run for several more years to individuals with proven clonal superiority are vegetatively achieve full confidence in the hybrids. In the interim, preliminary propagated (Shelbourne 1991). decisions on how to further develop the strategy for breeding the pure species and the hybrids are needed. Steps to gain the required Corymbia hybrid seedling families are now averaging 72% information are in hand. Meanwhile, Corymbia hybrids will soon rootability (64–86%, unpublished data), using mini-cutting be available to the plantation industry to meet current demands techniques (de Assis 2001). Similar results appear possible using for this material — despite, perhaps, small risks involved in stump coppice from 3-y-old trees. In the progeny trials there planting these hybrids on a commercial scale so soon after the appear to be families that are stable high performers across sites, first test plantings in 2001. with relatively low within-family variation. This offers the opportunity to deploy such families operationally if they are amenable to propagation as rooted cuttings (through vegetative Acknowledgements family forestry, for which a pilot program is underway with Forest I thank many staff at Department of Primary Industries and Plantations Queensland). Within the top families, and some poorer Fisheries, Horticulture and Forestry Science, in particular Peter families with high variances (e.g. CT × CCV151, Fig. 1), there Pomroy, John Oostenbrink, John Huth and Alan Ward for work are a few outstanding individuals that may have potential as resulting in the production and establishment of the pure species clones. In five hybrid progeny trials assessed at 3 y of age, 127 and hybrid populations now under test. I also thank Garth Nikles, superior trees (across 11 families) have been selected, based on who encouraged development of the pure species and hybrid growth, pest and disease tolerance, and form attributes. Currently breeding programs and who has guided many aspects of the work. 79 of these have been propagated successfully for clonal The contributions of collaborators at Sunshine Coast University, evaluation. The first clonal test was planted in April 2006. In Rhonda Stokoe and Helen Wallace, and Forest Plantations addition, 8 clones from good families (a subset of the 79 above) Queensland, whose assistance and work have underpinned much have been provided to a commercialisation partner (Dendrotech of this research program, are greatly appreciated. Provision of Pty Ltd), and are now being bulked up as partially-tested clones
Australian Forestry 2007 Vol. 70 No. 1 pp. 11–16 16 Corymbia species and hybrids for eastern Australia some seedlots of CCV and CH by Forests NSW is acknowledged. Lee, D.J., Nikles, D.G. and Walker, S.M. (1998) The genetic I thank Garth Nikles, John Huth and David Osborne for pre- improvement of native hardwood timber species in Queensland: publication review of the paper. case studies of three commercially important species. In: Proceedings of Managing and Growing Trees Training Conference. Kooralbyn, Queensland, 19–21 October 1998, References Department of Natural Resources and Mines, Queensland, Bruskin, S. (1999) New South Wales eucalypt plantation expansion — pp. 136–141. a silvicultural perspective. In: Ellis, R.C. and Smethurst, P.J. (eds) Lee, D.J., Nikles, D.G. and Dickinson, G.R. (2001) Prospects of Practising Forestry Today. 18th Biennial Conference of the eucalypt species, including interspecific hybrids from South Institute of Foresters of Australia. Institute of Foresters Inc., Africa, for hardwood plantations in marginal subtropical Hobart, Tasmania, pp. 152–156. environments in Queensland, Australia. Southern African Forestry Bureau of Rural Sciences (2005) Australia’s Forests at a Glance. Journal 190, 89–94. Department of Agriculture, Fisheries and Forestry, Canberra, Lee, D., Brawner, J. and Pomroy, P. (2004) Genetic variation in early 68 pp. growth and disease resistance of Corymbia citriodora subsp. Catesby, A. and Walker, S. (1997) An assessment of the relative variegata in southern Queensland, Australia. In: Borralho, amenability to vegetative propagation by leafy cuttings of 14 N.M.G., Pereira, J.S., Marques, C., Coutinho, J., Madeira, M. tropical and subtropical Eucalyptus and Corymbia species. In: and Tome, M. (eds) Eucalypts for a Changing World: Silviculture Kikkawa, J., Dart, P., Doley, D., Ishii, K., Lamb, D. and Suzuki, K. and Improvement of Eucalypts. International IUFRO Conference (eds) Overcoming Impediments to Reforestation: Tropical Forest of the SP2.08.03 on, Aveiro, Portugal, 11–15 October 2004. RAIZ, Rehabilitation in the Asia-Pacific Region. Bio-Refor, Brisbane, Instituto de Investigacao da Floresta e Papel, pp. 185–186. Australia, pp. 80–83. Lee, D.J., Debuse, V.J., Pomroy, P.C., Robson, K.J. and Nikles, D.G. de Assis, T.F. (2000) Production and use of eucalyptus hybrids for (2005) Developing Genetically Adapted Tree Varieties for industrial purposes. In: Dungey, H.S., Dieters, M.J. and Nikles, D.G. Marginal Areas of Northern Australia. RIRDC Publication No. (eds) Hybrid Breeding and Genetics of Forest Trees. Proceedings 04/186. Rural Industries Research and Development Corporation, of QFRI/CRC–SPF Symposium, Noosa, Queensland. Department Barton, Canberra, 58 pp. of Primary Industries, pp. 63–74. McComb, J.A. and Wroth, M. (1986) Vegetative propagation of de Assis, T.F. (2001) Evolution of technology for cloning Eucalyptus Eucalyptus resinifera and E. maculata using coppice cuttings and in large scale. In: Developing the Eucalypt of the Future. IUFRO micropropagation. Australian Forest Research 16, 231–242. International Symposium, Valdivia, Chile, 10–15 September 2001. McDonald, M.W. and Bean, A.R. (2000) A new combination in IUFRO, Santiago, pp. 1–16. Corymbia ‘section Politaria’: C. citriodora subsp. variegata Dickinson, G.R., Lee, D.J. and Huth, J.R. (2004) Early plantation (Myrtaceae). Austrobaileya 5, 735–736. growth and tolerance to ramularia shoot blight of provenances of Nanson, A. (1972) The provenance seedling seed orchard. Silvae the spotted gums taxa on a range of sites in southern Queensland. Genetica 21, 243–249. Australian Forestry 67, 122–130. Nikles, D.G. (1992) Influence of developments in breeding, propagation, Harwood, C., Bulman, P., Bush, D., Mazanec, R. and Stackpole, D. molecular markers, gene transfer and other new technologies on (2001) Australian Low Rainfall Tree Improvement Group genetic improvement strategies for forest trees in commercial Compendium of Hardwood Breeding Strategies. RIRDC plantation projects. In: Mass Production Technology for Publication No. 01/100. Rural Industries Research and Develop- Genetically Improved Fast Growing Forest Tree Species. IUFRO ment Corporation, Barton, ACT, 140 pp. meeting, AFOCEL, Bordeaux, France, 14–18 September 1992. Harwood, C.E., Bush, D.J., Butcher, T., Bird, R., Henson, M., Lott, R., Association Foret-Cellulose, Paris, pp. 137–150. and Shaw, S (2007) Tree improvement for low-rainfall farm Self, N.M., Aitken, E.A.B. and Dale, M.D. (2002) Susceptibility of forestry. Australian Forestry 70, 23–27. provenances of spotted gum to ramularia shoot blight. New Henson, M. and Smith, H. (2004) Borrowed technologies: adaptation Zealand Plant Protection 55, 68–72. of family forestry for eucalypts in NSW, Australia. In: Borralho, Shelbourne, C.J.A. (1991) Genetic gains from different kinds of N.M.G., Pereira, J.S., Marques, C., Coutinho, J., Madeira, M. breeding populations and seed or plant production population. and Tome, M. (eds) Eucalypts for a Changing World: Silviculture In: Schönau, A.P.G. (ed.) Intensive Forestry: The Role of and Improvement of Eucalypts. International IUFRO Conference Eucalypts. Proceedings of the IUFRO symposium, P2.02-01 of the SP2.08.03, Aveiro, Portugal, 11–15 October 2004. RAIZ, Productivity of Eucalypts, 2–6 September 1991. South African Instituto de Investigacao da Floresta e Papel, p. 103. Institute of Forestry, Durban, South Africa, pp. 300–317. Hill, K.D. and Johnson, L.A.S. (1995) Systematic studies in the Underhill, B. and Watts, H. (2004) The Queensland Forest Industry. eucalypts. 7. Revision of the bloodwoods genus Corymbia Department of Primary Industries and Fisheries, Brisbane, (Myrtaceae). Telopea 6, 185–504. Queensland, 128 pp. Isbell, R.F. (1996) The Australian Soil Classification. CSIRO Publishing, Melbourne, Victoria, 143 pp. Johnson, I., Carnegie, A. and Henson, M. (2004) Selecting for disease tolerant Corymbia citriodora subsp. variegata in New South Wales, Australia. In: Borralho, N.M.G., Pereira, J.S., Marques, C., Coutinho, J., Madeira, M. and Tome, M. (eds) Eucalypts for a Changing World: Silviculture and Improvement of Eucalypts. International IUFRO Conference of the SP2.08.03. Aveiro, Portugal, 11–15 October 2004. RAIZ, Instituto de Investigacao da Floresta e Papel, p. 184.
Australian Forestry 2007 Vol. 70 No. 1 pp. 11–16 Helen J. Smith and Michael Henson 17
Achievements in forest tree genetic improvement in Australia and New Zealand 3: Tree improvement of Eucalyptus dunnii Maiden
Helen J. Smith1,2 and Michael Henson1
1Forests NSW, PO Box J19, Coffs Harbour Jetty, NSW 2450, Australia 2Email: [email protected]
Revised manuscript received 1 February 2007
Summary important to the organisation, with over 8500 ha established — 32% of the total hardwood planted in the period (Forests NSW Eucalyptus dunnii Maiden has been widely planted in subtropical MIS data) — although the significance of the species in the annual Australia, with a total of almost 39 000 ha established in commer- planting program is now declining. Private forestry companies cial plantations in northern NSW and southern Queensland. in subtropical Australia have almost 30 000 ha established for Internationally, it has become recognised as a premium pulpwood diverse uses, and a recent planting rate of > 5000 ha y–1. species and its sawing properties have been considered superior to those of Eucalyptus grandis. Forests NSW established a Eucalyptus dunnii has been viewed as an alternative species to pedigreed tree improvement program for E. dunnii in 1995, and E. grandis W.Hill ex Maiden (flooded gum), versatile enough to this is the most advanced program for this species in Australia. be planted on a wide range of sites, although this perception has The main breeding population has been characterised for both led to the species being planted ‘off site’ in a number of cases. In pulpwood and sawlog traits, and improving the timber’s dimen- trial plantings in northern NSW, E. dunnii has equalled or sional stability during drying has been identified as a priority of outperformed E. grandis in growth on some sites (Johnson and the program. As clonal propagation of E. dunnii by cuttings is Stanton 1993), with the additional benefit of being better adapted not an economic means to deploy improved germplasm into to drier or frost-prone sites (Johnson and Arnold 2000). commercial plantations, Forests NSW has established three clonal Eucalyptus dunnii is commonly subjected to leaf defoliation by seed orchards for solid wood and pulp end uses, and plans to Christmas beetles but, when compared to E. grandis, generally establish a fourth in 2007. A range of E. dunnii hybrids with shows little susceptibility to wood borers such as cossid moth, high-value timber species has been produced with the aim of wood moth and longicorn beetles (Carnegie 2002) that cause improving the wood properties of the species. significant damage to trees and their timber (Phillips 1996). With Keywords: breeding programs; hybrids; wood properties; shrinkage; the expansion of E. dunnii plantations, the psyllid, Creiis lituratis, dimensional stability; pulpwood; propagation; Eucalyptus dunnii has become a significant insect defoliator of young plantations in north-eastern NSW and south-eastern Queensland and a serious Introduction problem, particularly where trees are stressed (Carnegie and Angel 2005). Recently it has been found that E. dunnii trees that Eucalyptus dunnii Maiden (Dunn’s white gum) is a member of are stressed, for instance by psyllid attack, are also susceptible the subgenus Symphyomyrtus section Maidenaria (Pryor and to cossid moth (A. Carnegie, NSW Department of Primary Indus- Johnson 1971). It is a species of limited natural distribution, having tries, West Pennant Hills, pers. comm., 2006). two disjunct populations in north-eastern New South Wales (NSW) and extending into south-eastern Queensland (Boland et al. 1984; Economically viable protocols for the vegetative propagation of Brooker and Kleinig 1999). The species occurs on moist, highly E. dunnii have yet to be developed, so current plantation establish- fertile soils of basaltic or alluvial origin, with a summer rainfall ment is by seedlings. Catesby and Walker (1997) considered the regime and a mean annual rainfall of 1100–1500 mm, and pre- species to be recalcitrant in vegetative propagation, with cuttings dominantly at altitudes of 400–650 m (Benson and Hager 1993). harvested from 40 seedling-based hedges having a mean rooting success of 3.2% at nine weeks, and 97.5% of clones having less Within NSW, Regional Forest Agreements (RFAs) between the than 30% rooting of cuttings. Results of initial Forests NSW State and Commonwealth governments have led to a reduction studies were similar1; subsequent studies using potted mother in the area of native forest that is available for harvest. Forests NSW (formerly State Forests of NSW) reintroduced a hardwood plantation program in 1994 with the aim of eventually supple- menting the wood supply from native forests. 1Smith, H.J. and Bacon, K.A. (2003) Vegetative propagation of plantation euca- lypts. A practical manual for the propagation of eucalypts for the North Coast Prior to 1994, Forests NSW had established only 93 ha of region of NSW. State Forests of NSW Confidential Internal Report, West Pen- E. dunnii plantations. Since that time the species has become nant Hills, NSW. 75 pp.
Australian Forestry 2007 Vol. 70 No. 1 pp. 17–22 18 Tree improvement of Eucalyptus dunnii plants, flood irrigated with defined nutrient solutions to optimise plantations and associated processing industries on the north coast production of minicuttings, have been unable to identify a method of NSW. It is expected that one of the outputs of this study will to economically propagate this species by cuttings (C. Moran, be economic weights for key traits to be improved in important Forests NSW Grafton, pers. comm., 2004). commercial species.
Eucalyptus dunnii timber from native forests has a basic density Improving productivity of 610 kg m–3 and is used for light construction such as building framework and joinery (Bootle 2005). The heartwood is of low A high priority for Forests NSW will always be improving forest durability and therefore not suitable for external use (Boland et productivity. The weighting placed on quality and adaptability al. 1984; Bootle 2005). Eucalyptus dunnii is becoming an inter- traits will vary depending on the species’ inherent wood properties nationally-favoured species for pulp, and extensive areas are being and risks in target plantation environments. The financial planted in southern China, and Central and South America, as importance of the increased productivity that can be achieved well as on the eastern seaboard of Australia. Plantation-grown using improved genetic material is indicated in a study by Henson E. dunnii has pulping properties similar to those of plantation- and Vanclay (2004): a plantation established on a high-quality grown E. grandis (Swain and Gardner 2003), but a higher pulp site using improved genetic planting stock had an internal rate of yield and superior paper-making properties (Backman and De return (IRR) of 13.0–13.6%, compared to an IRR of 11.4–11.6% León 1998). In studies on the Forests NSW breeding population, for a plantation established using wild planting stock on the same mean pulp yield was found to be 53.3% for the high-quality site. On the poorer site only improved material resulted in the Boambee site and 50.1% for the lower-quality Megan site plantation returning a profit. (Henson and Vanclay 2004).
The suitability of E. dunnii for solid wood products has been Improving wood properties addressed in a number of studies (Calori and Kikuti 1997; As timber from plantation-grown E. dunnii has high differential Dickson et al. 2003; Henson et al. 2004; Joe et al. 2004). Prob- shrinkage, reducing that shrinkage is currently a major objective lems include dimensional stability during drying, probably of the improvement program. Bandara (2006) found that mean resulting from high growth stresses and high differential annual increment (under bark, standing), Pilodyn penetration (a shrinkage. The species has a reputation for end-splitting, but measure of wood density) and tangential shrinkage are the best South African trials showed that there are family differences for selection criteria for the objective of maximising profit for a this trait; trees can be selected for little splitting and therefore vertically-integrated, grower–processor flooring production used for both mining and sawn timber production (Swain and system for E. dunnii. Gardner 2003). Similarly Dickson et al. (2003) found that 9- and 25-y-old plantation-grown trees produced timber that was Improving pest tolerance of acceptable hardness and strength for a range of solid-wood end uses, although wood quality varied significantly within and Creiis lituratus has caused substantial economic loses in E. dunnii among trees. Studies by Forests NSW and Southern Cross plantations on the north coast of NSW, with up to 500 ha classified University (Henson et al. 2004; Murphy et al. 2005; Thinley et as ‘failed’ or deemed ‘unsalvageable’ in both Forests NSW and al. 2005) have found that most wood property traits are under private plantations in the past 6 y (A. Carnegie, NSW Department moderate or high genetic control, suggesting there is the potential of Primary Industries West Pennant Hills, pers. comm., 2006). to improve such traits through tree improvement programs. While the pest outbreak may be an indirect result of prolonged and severe drought and severe damage to trees associated with Breeding objectives ‘off site’ planting, individual trees exhibiting tolerance to C. lituratus attack have been identified (Carnegie and Angel The Forests NSW Tree Improvement Strategic Plan2 identifies 2005) and propagated for inclusion in the breeding program. Intra- three general breeding objectives for all species under genetic specific hybridisation is being considered as an option to over- improvement. These are: come the C. lituratus problem. 1. Growth — to optimise productivity in terms of volume of the plantation crop Improving pulpwood properties 2. Quality — to ensure that every plant established has the Forests NSW has made clonal selections from the breeding potential to produce an economic stem (i.e. one that is ‘fit for population for pulpwood production and has developed a Near purpose’) Infrared Analysis model for pulpwood prediction in collaboration 3. Adaptability — to reduce risk of economic losses through with the Brisbane Forestry Research Centre (Muneri et al. 2005, plantation failure, by breeding trees that are resistant or 2007). Maximising pulp production and quality is not a key tolerant to pests and/or diseases and are adapted to their target objective for the improvement of E. dunnii in Forests NSW estate. planting environment (e.g., frost and drought tolerant). However, the organisation maintains a pulpwood deployment Forests NSW plans to complete, in the next few years, a population to provide seed to both domestic and international comprehensive study of breeding objectives for eucalypt pulpwood companies.
2 Forests NSW Tree Improvement Program (2006) Strategic Plan 2006–2007. Forests NSW Confidential Internal Report, Coffs Harbour, NSW. 6 pp.
Australian Forestry 2007 Vol. 70 No. 1 pp. 17–22 Helen J. Smith and Michael Henson 19
Breeding strategy replicates of four-tree plots. The trials were designed as incomplete blocks with a row–column layout. Currently Forests NSW has one main E. dunnii breeding population consisting of 219 open-pollinated families from Assessment of trials natural provenance collections. These are being tested on two sites in northern NSW in progeny trials established in 1995. It is The two progeny trials were first assessed in May 1998, at age planned to establish a second population as progeny trials across 39 months. Survival at both sites (86% and 93%) was high. The four sites in northern NSW in autumn 2007. This will consist of growth of the trees at the Megan site is poorer than at Boambee, 190 new families composed of new native provenance collections although the rankings of provenances at the two sites were (139 families), 31 families selected from provenance and progeny generally similar. Frosts, repeated heavy insect infestations and trials in NSW, and 20 families from trees selected in plantations. the relatively shallow soils at the Megan site may all have had a In addition, 77 families from the 1995 trials will be included on three of the sites, to link the two series of trials.
The long-term strategy will be to maintain a single multipurpose main population. Over the next few years, selections will be made in the 1995 progeny trials for the establishment of the second- generation main population, and selections from the 2007 trials Brisbane will be infused into the second generation over time. Two solid- wood elite populations will be selected along with a single pulpwood elite population. Each elite population will consist of 15 clones, and will be the focus of the pure species and hybrid breeding programs, with the aim of providing selected clonal material to be included in the deployment populations.
The production and testing of interspecific hybrids, with E. dunnii Casino as one of the parents, is being investigated as a strategy to improve wood quality and to produce genotypes better suited to poorer sites, which in turn may overcome the Creiis problem. Forests NSW has successfully crossed E. dunnii with E. pellita F.Muell., E. resinifera Smith and E. longirostrata (Blakely) L.A.S.Johnson Grafton & K.D.Hill with the aim of improving the sawlog value of the species. Hybrids of E. urophylla × E. dunnii created by Shell Forestry and tested in Queensland and South America have shown Megan E. dunnii Coffs Harbour great growth potential, and Forests NSW has recently imported progeny trial Boambee E. dunnii more of this material from South America. Natural hybrids of progeny trial E. dunnii × E. tereticornis have been identified in the nursery and it is hoped to test this material along with E. dunnii × E. longirostrata and E. dunnii × E. urophylla on drier sites.
Hybrid production may also help overcome the problems with Figure 1. Location of provenances included in Forests NSW breeding E. dunnii vegetative propagation of . Forests NSW established population (white squares) and the two progeny trial locations (stars) about 4 ha of potential open-pollinated E. dunnii × E. grandis in NSW hybrids in 2004. In 2005, 11 superior trees were selected and felled, and eight of the 11 trees were propagated from coppice. The rooting of minicuttings from these coppice-based mother plants is being investigated, and early results suggest that Table 1. Eucalyptus dunnii progeny trial site details propagation may be more successful than for E. dunnii (C. Moran, Site Name Boambee Megan Forests NSW Grafton, pers. comm., 2006). State NSW NSW Property Boambee SF Wild Cattle Creek SF Progeny trials Latitude/Longitude 30º18'S/153º03'E 30º17'S/152º47'E Annual rainfall (mm) 1900 1600 A breeding population, consisting of two progeny trials on Altitude (m) 60 730 contrasting Forests NSW sites in the Coffs Harbour–Dorrigo Soils Yellow podzolic Yellow podzolic region (Fig. 1), was established in February 1995. Boambee is a Planting date February 1995 March 1995 low-altitude (60 m asl) mild coastal ex-hardwood plantation site, Replicates 6 6 while the Megan site — formerly pasture — is at higher altitude Design Row–column Row–column (730 m asl) and colder (Johnson and Arnold 2000). Site and trial Plot size 4-tree row 4-tree row No. families 219 219 details are shown in Table 1. Both progeny trials were established a a with seedlings from 219 open-pollinated family seedlots collected Total no. of trees 5280 5280 a from 14 provenances. Each family was represented by six Includes one E. grandis bulk seedlot control
Australian Forestry 2007 Vol. 70 No. 1 pp. 17–22 20 Tree improvement of Eucalyptus dunnii detrimental affect on growth at that site (Johnson and Arnold material to operational plantations is through production of 2000). improved seed and deployment as seedlings.
Subsequent assessments were made at 75, 98, 102 and 108 months. Seedling seed orchards Measurements included growth (height, diameter at breast height over bark (dbhob)), stem straightness and stem taper, but the The Boambee progeny trial was selectively thinned in June 1999 primary focus has been on wood properties. At 75 months, bark- at age 52 months using the results of the 39-month assessment, to-bark increment cores at a height of 0.9 m were taken from five retaining the best two trees per four-tree plot. Flowering has not trees from each of 50 families and assessed for basic density and been widespread to date. Further thinning of the trial will take collapse (Arnold et al. 2004). At 98 months, destructively- place in early 2007 to retain the best tree per four-tree plot. This sampled trees from the better-performing families common to site will then be managed as an open-pollinated seedling seed the Boambee and Megan trials were assessed for a number of orchard (SSO). The Megan trial will be reassessed in 2007/2008 pulpwood traits. Samples from the more productive site at and is expected to be thinned on the basis of the assessment. Boambee had higher basic density and pulp yield, and longer fibres (Muneri et al. 2007). Clonal seed orchards
At 102 months, the Boambee trial was reassessed for growth, Forests NSW has established three joint-venture clonal seed tree form and density (using the Pilodyn). At 108 months, wood orchards (CSOs) with private landholders, and ramets for a further properties were assessed using increment coring (215 trees) and CSO have been grafted for planting in 2007. The earliest of these a range of non-destructive assessment tools including FAKOPP CSOs was established in November 2002, in collaboration with (a longitudinal stress wave acoustic device, Henson et al. 2004), CSIRO and a landholder, with 28 clones from eight provenances and growth strain was assessed using resistance gauges (Murphy selected for volume, form and wood properties (density and et al. 2005). One hundred and eighty one trees were destructively collapse assessed on cores) from the Boambee and Megan sampled and over 50 traits were assessed (Henson et al. 2004). progeny trials. Survival in this orchard has been poor, as grafted In addition 40 trees from 27 families were sawn and kiln dried, ramets of E. dunnii were found to suffer from later-age then scored for potential value-limiting defects (Harwood et al. incompatibility at the graft union. Similar graft incompatibility 2005). Results from this assessment have provided important has been reported in South Africa (Swain and Gardner 2003). information on the variation and genetic control of key wood Despite these setbacks, the CSO produced its first limited seed properties, as well as the efficiency and effectiveness of a range crop in summer 2005/2006. A small commercial seed crop has of assessment techniques. recently been harvested (December 2006). The estimated operational genetic gains from these clones, over the trial mean Genetic parameters at age 75 months, is 22% for dbhob at 6 y, but nothing for density based on the 9-y Pilodyn assessment. Operational genetic gain is Heritabilities for some key traits are listed in Table 2, but for calculated to be half of the additive genetic gain. more detailed information see Arnold et al. (2004) and Henson et al. (2004). Cross-site genetic correlation between Megan and The second CSO was established in 2005 after the destructive Boambee for dbhob at 6 y was 0.79 (se 0.14), suggesting that pulpwood evaluation (Muneri et al. 2005, 2007). A total of 50 there is no significant genotype × site interaction for growth traits. elite clones from 11 provenances selected for pulpwood traits Diameter (dbhob) and Pilodyn readings at 108 months were found and volume from the Boambee progeny trial are included in the to have a weak phenotypic correlation of 0.24 and a genetic orchard, together with a further eight pulpwood selects from a correlation of 0.07 (se ±0.14), suggesting there is no genetic progeny trial established by Dendrotech Pty Ltd in Eden. The correlation between density and growth. estimated operational genetic gain for the Boambee clones in this orchard, over the trial mean at age 102 months, is 15% for Deployment strategy dbhob and 7% for Pilodyn value (a positive gain in this value signifying an increase in density). As clonal propagation of E. dunnii is variable and uneconomic for commercial production, delivery of genetically improved
Table 2. Heritabilities for key traits of Eucalyptus dunnii from assessment of Boambee progeny trial+
Age (months) Trait n Mean CV (%) h2 (se) 102 Volume (m3) 2031 0.37 54 0.302 (0.077) 102 Pilodyn 6J (mm) 2031 12.36 13 0.510 (0.085) 108 Tangential shrinkage (%) 0179 11.7 18 0.695 (0.311) 108 Radial shrinkage (%) 0180 3.1 19 0.562 (0.305) 108 Differential shrinkage 0179 3.9 108 MOR (MPa) 0176 110.9 14 0.515 (0.320) 108 MOE (MPa) 0176 16 466 13 0.263 (0.296)
+.Data from Henson et al. (2004)
Australian Forestry 2007 Vol. 70 No. 1 pp. 17–22 Helen J. Smith and Michael Henson 21
The third CSO planted in 2006 has been established with clones Hancock Victoria Plantations have two E. dunnii progeny trials selected for solid-wood performance; it contains 11 clones established in 1989 at Stockdale and Jeeralang in Gippsland, selected for structural properties and 14 clones selected for Victoria. Both trials contain 57 families from four provenances, dimensional stability. Predicted family additive genetic gains from and one site has been assessed (dbh, density) and converted to the clones selected for structural properties are 4% for modulus an SSO (S. Elms, Hancock Victoria Plantations, Churchill, pers. of elasticity and 16% for hardness (expressing the gains as comm., 2007). improvement over the average of 47 assessed families). The clones selected for stability have an estimated family additive Conclusion genetic gain of 13% for reduced tangential shrinkage. The orchard as a whole has an operational genetic gain in dbhob of 16% over With the expanding establishment of E. dunnii plantations within the trial mean at 102 months. The fourth CSO to be established north-eastern NSW and south-eastern Queensland, for both solid in 2007 will contain the same clones. wood and pulp end uses, there is a continued need for the development of genetically-improved planting stock for specific Other organisations end uses. The Forests NSW tree improvement program for E. dunnii will soon move into the second generation, and continue Several organisations have established genetic trials for E. dunnii to develop and refine selections for the deployment populations that have subsequently been managed as seedling seed orchards. to service both the local solid-wood industry as well as the domestic and international pulp industry. The Forest Products Commission (FPC) of Western Australia established a progeny trial containing 80 families, from eight provenances, on two sites in south-western WA in July 1996. Acknowledgements These trials had some families in common with families in the Thanks are due to Ian Johnson (NSW Department of Primary Forests NSW progeny trials. Both sites were in the Manjimup Industries, Science and Research) for the establishment of the region, one with high rainfall and fertile soils, the other a dryland 1995 progeny trials and early assessment of these trials, and to site on sandy soils. Both trials were assessed in 2000. The more Steve Boyton (Forests NSW) for organising and coordinating fertile site was thinned to three trees per family (originally later assessments of the trials and for providing data management established with ten replicates per family in single-tree plots), services. We also thank Dane Thomas (Forests NSW), Angus while the second site has been retained at the original stocking. Carnegie (NSW Department of Primary Industries, Science and The thinned trial has been managed as an open-pollinated SSO, Research) and Bill Joe (Forests NSW) for reviewing this paper. and for the past two years has produced limited amounts of seed from many of the provenances (L. Barbour, Forest Products Commission WA, Wanneroo, pers. comm., 2006). References Arnold, R.J., Johnson, I.G. and Owen, J.V. (2004) Genetic variation in CSIRO Forestry and Forest Products established three joint- growth, stem straightness and wood properties in Eucalyptus venture progeny trials around Deniliquin, in southern NSW, in dunnii trials in northern New South Wales. Forest Genetics 11, 1996. The trials consist of a subset of about 120 of the families 1–12. represented in the 1995 series of the Forests NSW trials. With Backman, M. and De León, J.P.G. (1998) Pulp and paper properties of Deniliquin’s mediterranean climate the flood-irrigated sites do four-year-old eucalyptus trees for early species selection. In: not represent a typical commercial plantation environment for Proceedings 52nd Appita Annual General Conference. Carlton, E. dunnii. However, the sites have been found to promote Victoria.Volume 1, pp. 7–13. flowering and the normally shy-flowering E. dunnii has produced Bandara, K.M.A. (2006) Genetic improvement of solid wood product flowers at a relatively early age. Two of the three trials have value of subtropical eucalypts: a case study of Eucalyptus grandis been thinned to the best tree per plot (the trial was originally and E. dunnii. PhD thesis, Australian National University, established with four-tree line plots), and are used as SSOs, but Canberra, Australia, 215 pp. unfortunately the best trial was damaged by fire in 2005. CSIRO Benson, J.S. and Hager, T.C. (1993) The distribution, abundance and (now Ensis Genetics) established two second-generation progeny habitat of Eucalyptus dunnii (Myrtaceae) (Dunn’s white gum) in New South Wales. Cunninghamia 3, 123–125. trials in the same region in 2005. Boland, D.J., Brooker, M.I.H., Chippendale, G.M., Hall, N., Hyland, In 2000, Queensland Department of Primary Industries–Forestry B.P.M., Johnston, R.D., Kleinig, D.A. and Turner, J.D. (1984) established E. dunnii provenance/progeny trials on two sites in Forest Trees of Australia. 4th edn. CSIRO Publishing, Colling- southern Queensland at Mount Binga and Pechey. The trials wood, Victoria, Australia, 700 pp. contain 21 improved pedigreed seedlots imported from South Bootle, K.R. (2005) Wood in Australia. Types, Properties and Uses. Africa, with an additional 80 families being represented in bulk 2nd edn. McGraw-Hill Australia, Sydney, Australia, 452 pp. seedlots (Queensland Government Department of Primary Brooker, M.I.H. and Kleinig, D.A. (1999) Field Guide to Eucalypts. Industries 2003). One of the trials has been thinned to final Volume 1 South-eastern Australia. 2nd edn. Bloomings Books, stocking and is now producing small quantities of seed, while Hawthorn, Victoria, Australia, 353 pp. the other has been retained at high stocking for making selections and carrying out wood properties studies (D. Lee, Queensland Department of Primary Industries and Fisheries, Gympie, pers. comm., 2006).
Australian Forestry 2007 Vol. 70 No. 1 pp. 17–22 22 Tree improvement of Eucalyptus dunnii
Calori, J.V. and Kikuti, P. (1997) Physical and mechanical properties Joe, W., Dickson, R., Raymond, C., Ilic, J. and Matheson, C. (2004) of wood from Eucalyptus dunnii of 20 years of age. In: Prediction of Eucalyptus dunnii and Pinus radiata Timber Stiffness Proceedings of IUFRO Conference on Silviculture and Using Acoustics. Rural Industries Research and Development Improvement of Eucalypts Volume 3: Silviculture, Productivity Corporation Publication No. 04/013, Final Report Project No. and Utilization of Eucalypts. 24–29 August, Salvador, Brazil. PN99.2010, Australian Capital Territory. 121 pp. EMBRAPA, Colombo, Brazil, pp. 321–326. Johnson, I.G. and Arnold, R.J. (2000) Eucalyptus dunnii Provenance- Carnegie, A.J. (2002) Field Guide to Common Pests and Diseases in Family Trials in Northern New South Wales — Age Three-year Eucalypt Plantations in NSW. State Forests of NSW, Research Assessment. State Forests of New South Wales, Research and and Development Division, West Pennant Hills, NSW, Australia, Development Division, Research Paper No. 37, West Pennant 80 pp. Hills, Australia, 28 pp. Carnegie, A.J. and Angel, P. (2005) Creiis lituratus (Froggatt) Johnson, I.G. and Stanton, R.R. (1993) Thirty Years of Eucalypt Species (Hemiptera: Psyllidae): a new insect pest of Eucalyptus dunnii and Provenance Trials in New South Wales. Survival and Growth plantations in sub-tropical Australia. Australian Forestry 68, in Trials Established from 1961 to 1990. Forestry Commission of 59–64. NSW, Research Division, Research Paper No. 20, West Pennant Catesby, A.L. and Walker, S.M. (1997) An assessment of the relative Hills, Australia, 92 pp. amenability to vegetative propagation by leafy cuttings of 14 Muneri, A., Daido, T. and Henson, M. (2005) Near infrared tropical and sub-tropical Eucalyptus and Corymbia species. In: spectroscopy calibrations for pulp yield and basic density of 8 Kikkawa, J., Dart, P., Doley, D., Ishii, K., Lamb, D. and Suzuki, K. year old Eucalyptus dunnii using samples from contrasting sites. (eds) Overcoming Impediments to Reforestation: Tropical Forest In: Proceedings 59th Appita Annual General Conference. 16–19 Rehabilitation in the Asia-Pacific Region. Proceedings of the 6th May, Auckland, New Zealand, p. 431. Annual International Workshop of Bio-Refor, 2–5 December. Bio- Muneri, A., Daido, T., Henson, M. and Johnson, I. (2007) Variation in Refor, Brisbane, Queensland, Australia, Volume 1, pp. 80–83. pulpwood quality of superior Eucalyptus dunnii families grown Dickson, R.L., Raymond, C.A., Joe, W. and Wilkinson, C.A. (2003) in NSW. Appita Journal 60, 74–77. Segregation of Eucalyptus dunnii logs using acoustics. Forest Murphy, T.N., Henson, M. and Vanclay, J.K. (2005) Growth stress in Ecology and Management 179, 243–251. Eucalyptus dunnii. Australian Forestry 68, 144–149. Harwood, C., Bandara, K., Washusen, R., Northway, R., Henson, M. Phillips, C. (1996) Insects, Diseases and Deficiencies Associated with and Boyton, S. (2005) Variation in Wood Properties of Plantation- Eucalypts in South Australia. Primary Industries, South Australia, Grown Eucalyptus dunnii Relevant to Solid-Wood Products. Forest 160 pp. and Wood Products Research and Development Corporation, Final Pryor, L.D. and Johnson, L.A.S. (1971) A Classification of the Report Project No. PN04.3003, Victoria. 37 pp. Eucalypts. The Australian National University, Canberra, Henson, M. and Vanclay, J.K. (2004) The value of good sites and Australia, 102 pp. genotypes: an analysis of Eucalyptus dunnii plantations in NSW. Queensland Government Department of Primary Industries (2003) In: González, A. and Gabriel, J. (eds) The Economics and South East Queensland Forests Agreement Hardwood Plantations Management of High Productivity Plantations. Proceedings of Initiative. Performance Report September 1999 to June 2003. the IUFRO International Meeting, 27–30 September, Universidad Queensland Department of Primary Industries, Brisbane. de Santiago de Compostela, Lugo, Spain. Electronic Publication CD, http://www.centraldellibro.com/The-economics-and- Swain, T-L. and Gardner, R.A.W. (2003) A Summary of Current management-of-high-productivity-plantations-proceedings-of- Knowledge of Cold Tolerant Eucalypt Species (CTE’s) Grown in the-IUFRO-International-Meeting-Universidad-de-Santiago-de- South Africa. Institute for Commercial Forestry Research, ICFR Compostela-Lugo-27-30-de-septiembre-de-2004-es1918315.htm, Bulletin 03/2003, Scottsville, South Africa, 57 pp. 6 pp. Thinley, C., Palmer, G., Vanclay, J.K. and Henson, M. (2005) Spiral Henson, M., Boyton, S., Davies, M., Joe, W., Kangane, B., Murphy, T., and interlocking grain in Eucalyptus dunnii. Holz als Roh-und Palmer, G. and Vanclay, J. (2004) Genetic parameters of wood Werkstoff 63, 372–379. properties in a 9 year old E. dunnii progeny trial in NSW, Australia. In: Borralho, N.M.G., Pereira, J.S., Marques, C., Coutinho, J., Madeira, M. and Tomé, M. (eds) Eucalyptus in a Changing World. Proceedings of the IUFRO Conference, 11–15 October, RAIZ, Instituto Investigação da Floresta e Papel, Aveiro, Portugal, p. 183.
Australian Forestry 2007 Vol. 70 No. 1 pp. 17–22 C.E. Harwood, D.J. Bush, T. Butcher, R. Bird, M. Henson, R. Lott and S. Shaw 23
Achievements in forest tree genetic improvement in Australia and New Zealand 4: Tree improvement for low-rainfall farm forestry
C.E. Harwood1,2, D.J. Bush3, T. Butcher4, R. Bird5, M. Henson6, R. Lott7 and S. Shaw8
1Ensis, Private Bag 12, Hobart, Tasmania 7001, Australia 2Email: [email protected] 3Ensis, PO Box E4008, Kingston, ACT 2604, Australia 4Forest Products Commission, 120 Albany Highway, Albany, WA 6330, Australia 5Department of Primary Industries, Private Bag 105, Hamilton, Victoria 3300, Australia 6Forests NSW Northern Research, PO Box J19, Coffs Harbour Jetty, NSW 2450, Australia 7Rural Industries Research and Development Corporation, PO Box 4776, Kingston, ACT 2604, Australia 8Primary Industries and Resources, South Australia – Forestry, PO Box 162, Mount Gambier, SA 5290, Australia
Revised manuscript received 8 September 2006
Summary Introduction
The challenge for low-rainfall plantation forestry is to develop In recent decades there has been a growing awareness of wide- financially viable plantations using hardy species that produce scale environmental problems such as dryland salinity, and the commercial products as well as delivering environmental benefits. need to ameliorate this by planting deep-rooted perennial Since 1999, forest research agencies have co-operated on tree vegetation in the low-rainfall regions of southern Australia improvement for dryland southern Australia through the (Stirzaker et al. 2002). The aim of so-called commercial environ- Australian Low Rainfall Tree Improvement Group (ALRTIG). mental forestry plantations is to generate two income streams A small number of key species, including hardwoods (tree-sized for growers: an income imputed to the environmental benefits of eucalypt species), softwoods (Pinus brutia, P. pinaster and plantations through, for example, salinity or carbon credits for P. radiata) and two mallee eucalypt species yielding foliar essen- trees that control dryland salinity and/or sequester carbon, and a tial oils were selected for genetic evaluation and improvement. commercial return from the sale of timber or other forest products Selection was based on their adaptability, growth rates, yields of (Nambiar and Ferguson 2005). Genetic improvement is expected potential commercial products, and availability of reliable to enhance commercial product yield and quality, bringing information and genetic resources. plantations closer to overall financial viability. In addition to genetic improvement, improved silviculture, environmental Improvement for the hardwood and softwood species aims to service payments, and optimised harvesting, transport and pro- develop breeds that will maximise the net present value (stumpage cessing will all be needed for low-rainfall plantations to approach value less growing costs) of logs produced on rotations of 20–40 y. profitability and attract investment (Harwood et al. 2005). Breeding will be based on recurrent selection for general combining ability, primarily in open-pollinated breeding The main plantation species in southern Australia, Eucalyptus populations based on selections from the best natural provenances globulus and Pinus radiata, grow poorly and are at risk of death and planted stands. A total of 29 provenance-progeny trials of from drought in environments receiving less than about 600 mm the key species have been established across a range of dryland mean annual rainfall. Evaluation of tree species suited to sites. These trials are now yielding information on the genetic Australia’s dryland (400–600 mm mean annual rainfall) sheep– control of growth, stem form and branching traits. This will enable wheat belt has been on-going for many years in a number of production of genetically-improved planting stock through later State and Commonwealth organisations, but genetic improvement conversion of some trials to seedling seed orchards and develop- programs have been initiated for only a few such species. In South ment of clonal seed orchards. Interim seed production areas based Australia and Victoria, E. cladocalyx has been planted widely as on planted stands of known best provenances have been developed a low-rainfall farm tree species from the 1870s, while in Western for short-term production of somewhat-improved seed, and Australia P. pinaster was identified as a suitable low-rainfall genetic gain trials to compare these seed sources with unimproved species and a genetic improvement program commenced in the material have been established. 1940s (Butcher 2006).
Keywords: plantations; semiarid zones; farm forestry; seed sources; seed output; genetic improvement; seed orchards; Eucalyptus; Pinus; mallees
Australian Forestry 2007 Vol. 70 No. 1 pp. 23–27 24 Tree improvement for low-rainfall farms
ALRTIG genetic improvement strategies These species were judged sufficiently hardy for at least part of the target planting area and capable of producing sawlogs or other A multi-agency cooperative program, the Australian Low Rainfall commercial wood products. Strategies were developed to secure Tree Improvement Group (ALRTIG), was established in 1999. short-term supplies of somewhat-improved seed, and to com- The mission of ALRTIG is: to produce genetically improved mence improvement programs to deliver ongoing genetic gains. planting material for farm forestry in the low-rainfall areas of The strategies are based primarily on recurrent selection for southern Australia, and inform tree growers of its availability. general combining ability in open-pollinated breeding populations ALRTIG pooled pre-existing biological and scientific resources (Harwood et al. 2001). Some important elements of the strategies belonging to its core partners (state forestry agencies from NSW, for these hardwood species have already been implemented: South Australia, Victoria and Western Australia, and Ensis) in • analysis of partners’ pre-existing field trials to make recom- order to maximise impact through nationally-coordinated research mendations on best-bet interim seed sources, including best and development. Further details on ALRTIG’s organisation and natural provenances and planted stands modus operandi are provided by Bush et al. (2007). • conversion of some existing plantings into seed production ALRTIG has coordinated genetic improvement of three groups areas by selective thinning to meet short-term demand for of species: hardwoods, softwoods and short-rotation woody crops. improved seed Candidate species for improvement in these three groups were • for each key species, establishment of provenance/progeny reviewed in 1998–1999 and a small number of key species on trials testing a common set of 80–100 or so open-pollinated which to concentrate was selected using criteria of growth rates, progenies from known superior provenances and planted seed potential commercial products, availability of reliable information sources, across a range of dryland environments in southern on performance and availability of genetic resources. Australia • establishment of genetic gain trials to test the performance of Improvement objectives recommended interim improved seed sources against unim- proved natural-provenance controls The following general improvement objective was developed for the hardwood species: To develop breeds that will maximise the • research into vegetative propagation as a deployment option. net present value (stumpage value less growing costs) of logs Harwood et al. (2005) reviewed progress achieved to that date produced on rotations of 20–40 years, on a per-hectare basis. and presented additional biological information obtained on the Focus will be on production of a low-cost, high-value final crop species, but made no major recommendations for changes in the which can be used for solid wood applications including sawlogs, improvement strategy or methods for improvement. veneer and engineered wood products.
Leaf area index, stand water use and susceptibility of the planta- Softwoods tion to drought can be manipulated by thinning and pruning so Two key species were chosen initially. ALRTIG’s involvement that available site resources are directed to an optimum number with P. pinaster (maritime pine) is restricted to planting a series of trees per hectare (White et al. 2003). The aim of genetic of yield trials that compare the performance of various categories improvement programs will be to breed for improved profitability, of improved and unimproved genetic stock produced by the taking into account likely silvicultural regimes. In addition to Western Australian breeding program (Butcher 2007). Pinus improving growth rates, breeding will therefore aim to improve brutia, native to the eastern Mediterranean and western Asia, branching and stem form so as to reduce the silvicultural inputs has potential for very dry sites, including those with alkaline and/ and attendant costs required to produce valuable logs on the or shallow soils. It is amongst the most drought-resistant Pinus retained trees. Wood density, hardness and stiffness of the selected species. A genetic improvement strategy was prepared and hardwood species is sufficiently high for likely wood products commenced for P. brutia (Spencer and Boardman 2002). (Blakemore et al. 2002). At this stage there is insufficient market ALRTIG created two seed stands by thinning existing plantations information to identify wood properties requiring improvement. for short-term seed production. These will supplement seed Survival, growth (height and diameter at breast height), stem form available from a pre-existing clonal seed orchard in Western (apical dominance and stem straightness) and branching (light Australia. A survey of all known Australian P. brutia stands was branching and tendency for self-pruning) have been identified as undertaken, and a list of 161 candidate plus trees compiled. The selection traits, but economic weights for these traits have not top 39 of these trees were grafted into clonal seed orchards in yet been determined. Improvement objectives for the softwood the ACT, South Australia and Western Australia in 2003. Three species are broadly similar, and more straightforward because progeny trials testing open-pollinated seedlots collected from 70 markets for wood products from radiata pine and maritime pine of the candidate plus trees were established in 2002. New are already well established. For the mallee species, essential oil collections have been made from the species’ natural range to yield and quality will be major traits for improvement. expand the genetic base.
Hardwoods In 2002, the potential for creating a low-rainfall breed of P. radiata, distinct from the Australian breeds developed for Eucalyptus camaldulensis (river red gum), E. cladocalyx (sugar higher-rainfall environments, was examined. A strategy was gum), E. occidentalis (swamp yate), E. tricarpa and E. sideroxylon devised for creating a new breed, incorporating selections from (red ironbarks) and Corymbia maculata and C. citriodora subsp. the southern natural provenances (two Mexican islands and variegata (spotted gums) were selected for genetic improvement.
Australian Forestry 2007 Vol. 70 No. 1 pp. 23–27 C.E. Harwood, D.J. Bush, T. Butcher, R. Bird, M. Henson, R. Lott and S. Shaw 25
Cambria, USA), which may display greater drought tolerance. will suffice or whether genotype × environment interaction across The strategy (Matheson et al. 2007) has a number of components: the different regions is sufficient to warrant separate breeding • preliminary screening in field trials of selected material from populations. Seedling-based trials testing common sets of 24 provenance stands including Cambria, Guadalupe and Cedros open-pollinated families of each species on a total of five sites for genetic variation in drought resistance (two sites in Western Australia and one each in NSW, South Australia and Victoria) have been established. • establishment of longer-term field trials of candidate material on target site types In all, 29 progeny trials of hardwood species, ten yield trials of • formation of a breeding population using stored seed and P. pinaster, three progeny trials and three clonal seed orchards controlled crosses in existing stands of the Mexican and of P. brutia and five mallee genetic trials have been planted by Cambria provenances of interest ALRTIG partners over the period 2000–2002: a total of over 50 • possible integration of additional drought resistance through ha of genetic trials. interspecific hybridisation with closely related species such as P. attenuata, P. oocarpa or other pines if breeding barriers Progress to date can be identified and overcome. Eucalyptus cladocalyx progeny trials Work under the strategy, including establishment of field trials and controlled crossing programs, commenced in 2003–2004. Early results obtained from ALRTIG trials of one of the key hardwood species, E. cladocalyx, illustrate the progress already Short-rotation woody crops made in genetic evaluation and the potential for genetic improvement. Assessment data are presented from one of the Two mallee eucalypt species, E. polybractea (blue mallee) and E. cladocalyx progeny trials near Bordertown, South Australia, E. horistes (oil mallee), have been selected for a series of managed by PIRSA Forestry. Height and diameter at breast height genotype × environment trials in each of the southern mainland (dbh) were measured and stem forking assessed 28 months after states. Both of these species have high yields of cineole-rich leaf planting. Forking was scored as: 1 = a fork at ground level, 2 = a oils, and are grown on three-year coppice rotations. In Western fork in the first quarter of the stem, 3 = a fork in the second Australia it was already known that E. horistes was suited to quarter … through to 6 = no forks. Seed source means for height, relatively drier sites and E. polybractea relatively wetter ones, dbh and forking score are shown in Table 1. Each seedlot mean though their performance on eastern Australian site types was represents the mean of five or more open-pollinated progenies unknown. The study aims to determine (1) the relative suitability from that seedlot. Narrow-sense, individual tree, within- of each species on target site types in eastern Australia, and provenance heritabilities for each trait were calculated using a (2) whether single national breeding populations for each species coefficient of relationship of 0.4 within open-pollinated families.
Table 1. Twenty-eighth-month growth and stem form of seedlots in Eucalyptus cladocalyx progeny trial at Bordertown, South Australia
Stem form Diameter bh Height Region and seedlot (1–6 fork score) (mm) (dm) Kangaroo Island Flinders Chase NP 3.3 37.8 40.5 American River 3.3 34.4 38.9 Flinders Chase 3.6 42.3 43.5 Cygnet River 3.6 39.2 42.9 Planted stands Mt Burr 4.7 45.9 41.6 Kersbrook SPA* 5.0 43.7 41.0 Majorca 5.2 48.9 43.2 Wail 5.2 47.2 41.3 Lismore 5.3 41.6 39.6 Southern Flinders S. Wilmington 5.0 37.6 35.0 Wilmington 5.0 36.6 34.9 Wirrabara S.F. 5.1 41.1 38.7 Mt Remarkable 5.0 35.4 34.4 Average s.e.d.† 0.6 02.2 03.1 Heritability (s.e.) 0.21 (0.05) 0.21 (0.06) 0.25 (0.06)
*Kersbrook seed production area †s.e.d. = standard error of difference of seedlot means
Australian Forestry 2007 Vol. 70 No. 1 pp. 23–27 26 Tree improvement for low-rainfall farms
The Southern Flinders natural provenances have significantly kept under study in C. maculata seed orchards as it may affect better stem form (less forking) than those from Kangaroo Island. progeny performance. Conversely, the Kangaroo Island provenances are more vigorous than those from the Southern Flinders. The planted/improved Genetic gain trials seed sources (progeny from phenotypically superior mother trees selected from land race plantations originally derived from In 2003 a total of 20 genetic gain trials of four of the key hardwood Southern Flinders provenance, except for Kersbrook SPA, which species were established by collaborating tree-growing is a mixture of Southern Flinders and Kangaroo Island companies. These trials will enable the performance of recom- provenances) combine vigour equivalent to Kangaroo Island with mended interim ‘best bet’ seed sources to be compared with non- the superior stem form of Southern Flinders. Within-seed source improved benchmarks on collaborators own sites using their individual-tree heritabilities for height, dbh and stem form were preferred silvicultural regime. These trials used large plots and moderate (0.21–0.25) for the three traits. at least five replicates to obtain accurate rankings of seedlot performance. Early results (Bush et al. 2007) from the This result corroborates information on natural provenance C. maculata genetic gain trials indicate superior growth of the rankings derived from Partners’ earlier trials, supports ALRTIG’s seed orchard material under test, relative to natural provenances. recommendation to use material derived from the identified seed stands and seed production areas, and demonstrates the potential Planting rates and capacity to supply improved seed for further genetic improvement. With a total of nine provenance/ progeny trials of E. cladocalyx established across dryland Planting rates have fluctuated in recent years (Table 2). From southern Australia in 2001, breeders now have excellent genetic surveys of major nurseries, the total annual plantings for the resources available. Outstanding selections could be grafted into ALRTIG key species were about 1300 ha (hardwood species), clonal seed orchards, and some trials converted to seedling seed 1600 ha (softwood species) and 600 ha (mallee species) in 2004. orchards via selective thinning. The interim seed production areas of the key hardwood species –1 –1 Outcrossing rates in interim seed production areas have typically produced about 2 kg of seed ha y . Assuming typical seed numbers per kilogram of seed, nursery efficiencies It is well known that selfing will reduce the performance of and stocking rates, and thinning provenance/progeny trials to progeny in many Eucalyptus species (Eldridge et al. 1993). convert them into seedling seed orchards, annual plantings of Outcrossing rates were determined in two of the interim seed 2000 ha for sugar gum and swamp yate, 1000 ha for spotted gum production areas to see whether high levels of selfing might be a and somewhat less than 1000 ha for red ironbarks could be problem (Bush et al. 2007). A multi-provenance E. occidentalis supported. This capacity could be supplemented by clonal seed seed production area in South Australia was studied by examining orchards established from selections from the provenance/ genetic variability of progeny from ten trees using microsatellite progeny trials. Private-sector firms have established other markers. Mean estimated outcrossing rates of the ten parents seedling seed orchards of C. citriodora subsp. variegata, ranged from 0.69 to 1, with a mean of 0.92 (i.e. 92% outcrossing C. maculata, E. camaldulensis and E. cladocalyx. The WA Forest rate). Allozyme markers on progeny of ten trees were similarly Products Commission has independently developed seedling seed used to determine outcrossing rates in a C. maculata seed orchards of the two oil mallee species. The seed supply capacity production area developed from a provenance–progeny trial of of P. brutia in the thinned seed stands and clonal seed orchards C. maculata which also included C. citriodora subsp. variegata is yet to be established. and C. henryi progenies. Outcrossing rates for individual trees ranged from 0.50 to 1, and the mean multi-locus outcrossing rate Clonal propagation is another option for deploying improved E. camaldulensis was 0.76. These results suggest that inbreeding is unlikely to be germplasm. clones are already commercially Eucalyptus cladocalyx E. occidentalis a major problem in E. occidentalis seed orchards, but should be available. , and red
Table 2. Plantation establishment rates for ALRTIG key species
Number of seedlings (× 1000) Estimate of area 1 Species raised in major nurseries planted (ha) 2001 2004 2001 2004 Sugar gum 0402 0989 0400 1000 Swamp yate 0108 0010 0100 0100 Red ironbark 0102 0084 0100 0100 River red gum 0325 0141 0300 0100 Mallee 3695 1596 1400 0600 Spotted gum 0523 0253 0500 0300 Maritime pine 9200 2367 6200 1600 Brutian pine 0004 0009 0004 0009
1Assuming stocking rates of 2600 trees ha–1 (oil mallees) , 1500 trees ha–1 (maritime pine) and 1000 trees ha–1 (other species)
Australian Forestry 2007 Vol. 70 No. 1 pp. 23–27 C.E. Harwood, D.J. Bush, T. Butcher, R. Bird, M. Henson, R. Lott and S. Shaw 27 ironbarks can be propagated readily from stem cuttings taken References from seedling hedge plants, indicating that family forestry is Blakemore, P., Waugh, G., Northway, R. and Washusen, R. (2002) Wood technically feasible (C. Harwood, unpublished data). Products from Low Rainfall (400–600 mm/yr) Farm Forestry. RIRDC Publication No. 02/051. Rural Industries Research and In summary, progress has been made towards remedying a serious Development Corporation, Canberra. impediment to adoption of plantation forestry in low-rainfall Bush, D., Butcher, T., Harwood, C., Bird, R., Henson, M. and Shaw, S. southern Australia: making available germplasm that can yield a (2007) Australian Low Rainfall Tree Improvement Group Progress commercial product in addition to an environmental benefit. 1999–2005. Rural Industries Research and Development Improved planting material of the chosen key species will be Corporation, Canberra. available to support the development of sizeable plantation Butcher, T.B. (2007) Achievements in forest tree genetic improvement estates. In 1999, when the ALRTIG key species were selected, it in Australia and New Zealand: maritime pine and red pine tree was thought that new information might effect changes to the improvement programs in Western Australia. Australian Forestry key species. Thus far, the original key species have remained (in press). among the ‘best bets’ for low-rainfall forestry (Nambiar and Eldridge, K.G., Davidson, J., Harwood, C.E. and Van Wyk, G. (1993) Ferguson 2005). Eucalypt Domestication and Breeding. Oxford University Press, New York. Extension and communication Harwood, C.E. and Bush, D. (eds) (2002) Breeding Trees for the Low Rainfall Regions of Southern Australia. JVAP Research Update ALRTIG has communicated information and promoted the use Series No. 4. Joint Venture Agroforestry Program, Canberra. of genetically improved germplasm through written publications Harwood, C., Bulman, P., Bush, D., Stackpole, D. and Mazanec, R. for scientific and general audiences (Harwood and Bush 2002), (2001) Australian Low Rainfall Tree Improvement Group: presentations to the media, conferences and workshops, Compendium of Hardwood Breeding Strategies. RIRDC Publica- contributions to training courses for forestry and agroforestry tion No. 01/100. Rural Industries Research and Development practitioners, and a website (http://www.ffp.csiro.au/alrtig/) Corporation, Canberra. containing information on the key species and available best-bet Harwood, C.E., Bird, R., Butcher, T., Bush, D.J., Jackson, T., Johnson, I., seed sources. Communication among geographically-spread Stackpole, D. and Underdown, M. (2005) Update of Australian partners from different organisations, and leadership of ALRTIG, Low Rainfall Tree Improvement Group Breeding Strategies. were significant challenges. The role of a National Coordinator RIRDC Publication No. 05/023. Rural Industries Research and Development Corporation, Canberra. in the first four years of the program was found to be essential. Matheson, A.C., Spencer, D.J., Bush, D.J. and Porada, H. (2007) Australian Low Rainfall Tree Improvement Group: A Strategy to The future of low-rainfall tree improvement Develop a Low Rainfall Breed of Pinus radiata. Rural Industries Research and Development Corporation, Canberra. 22 pp. Over the next 1–2 y the ALRTIG genetic trials will be assessed Nambiar, E.K.S. and Ferguson, I. (eds) (2005) New Forests: Wood and data assembled for across-site statistical analysis. The extent Production and Environmental Services. CSIRO Publishing, of further investment in tree improvement will depend on the Melbourne. future demand for improved seed. The ALRTIG trials will also Spencer, D. and Boardman, R. (2002). Genetic improvement strategy be of use for studies in a wide range of disciplines including for Pinus brutia in southern Australia. In: Australian Low Rainfall population genetics, growth and yield modelling, hydrology and Tree Improvement Group Compendium of Softwood Breeding wood quality studies. Strategies. RIRDC Publication No. 02/028, Rural Industries Research and Development Corporation, Canberra, pp. 26–44. Acknowledgements Stirzaker, R., Vertessy, R. and Sarre, A. (eds) (2002) Trees, Water and Salt: An Australian Guide to Using Trees for Healthy Catchments We wish to acknowledge the strong contributions of former and Productive Farms. RIRDC Publication No. 01/086. Rural members of the ALRTIG National Steering Committee, Industries Research and Development Corporation, Canberra. particularly Ms Sharon Davis, Ms Deborah O’Connell, Dr Hans White, D.A., Battaglia, M., Macfarlane, C., Mummery, D., McGrath, Porada, Mr Des Stackpole and Mr Mick Underdown. Financial J.F. and Beadle. C.L. (2003) Selecting species for recharge support for the program from the Joint Venture Agroforestry management in Mediterranean south-western Australia — some ecophysiological considerations. Plant and Soil 257, 283–293. Program, the Department of Agriculture, Fisheries and Forestry, and the partner agencies including Ensis, Forests NSW, Primary Industries and Resources, South Australia – Forestry, the Victorian Department of Primary Industries and the Forest Products Commission of Western Australia is gratefully acknowledged.
Australian Forestry 2007 Vol. 70 No. 1 pp. 23–27 28 Genetic improvement of Douglas-fir
Achievements in forest tree genetic improvement in Australia and New Zealand 5: Genetic improvement of Douglas-fir in New Zealand
C.J.A. Shelbourne1, C.B. Low2, L.D. Gea2,3 and R.L. Knowles2
199 Acacia Rd, RD5 3076 Rotorua, New Zealand 2Ensis, Private Bag 3020, 49 Sala Street, 3010 Rotorua, New Zealand 3Email: [email protected]
Revised manuscript received 1 November 2006
Summary those of radiata pine), modest value and low profitability discouraged expansion, as did Swiss needlecast disease (caused Douglas-fir is New Zealand’s second most important plantation by the fungus Phaeocryptopus gaeumannii) (Anon. 1973). species and has been grown there since the late 1800s. Recently there has been increased commercial interest in the species, Since around 1988, there has been renewed commercial interest especially for higher-altitude, snow-prone sites in the southern in the species, owing to high values of Douglas-fir timber and South Island. Tree improvement began in 1957 with the establish- the recognition that on some higher-altitude, snow-prone sites, ment of extensive large-plot provenance trials. These have shown especially in Otago and Southland, Douglas-fir performs better that coastal fogbelt provenances from California and southern than radiata pine. This has been an important factor in the re- Oregon are superior in growth to those of Washington coastal evaluation of Douglas-fir, as has the realisation that Californian origin. The Washington provenances had been used for much of and southern Oregon coastal fog-belt provenances were growing the afforestation up to 1980. much faster than those of Washington origin. Also the superior stiffness of Douglas-fir, especially in the corewood zone, and A first breeding program, started in 1969 from stands of Washing- the useful degree of natural durability of its early-forming ton origin, became redundant following the provenance trial heartwood, can produce timber much better suited to structural findings and, in 1988, selection of 186 plus trees in provenance uses than radiata pine. trial plots of coastal Californian and Oregon provenances restarted the breeding program. In addition, open pollinated seed from Historical development of tree improvement 240 parents in 21 coastal Californian and Oregon populations was planted as a breeding population in 1994. An ambitious A program of provenance research in Douglas-fir was started in program of polycrossing and paircrossing of NZ selections failed 1955 by I.J. Thulin (Sweet 1965; Thulin 1967). Trials of 35 and to deliver sufficient crosses and seed, and recently a new breeding 45 provenances were established in 1957 and 1959, with large strategy based on open pollination in a clonal archive of the 1988 144-tree plots replicated on eight and 19 sites respectively. The selections has been instituted. first series was mostly of commercial seedlots from Washington and Oregon, and the second was from seedlots collected by NZ The breeding objectives have also been recently revised to include breeder Egon Larsen from Californian and Oregon coastal popu- volume yield, log quality and timber stiffness, selection criteria lations. Further provenance- and local seed-source trials were being diameter, bole straightness, light well-distributed branching planted in 1974. Early results of these trials at age 6 y, and and outerwood density and/or sound velocity. especially the 12-y results (M.D. Wilcox unpublished data) Keywords: breeding programs; history; open pollination; wood showed clearly that seedlots of coastal Washington origin were properties; Douglas-fir; New Zealand growing substantially slower than those from coastal provenances from southern Oregon and California (Thulin 1967; Sweet 1965). Introduction Recent studies based on 22 coastal fogbelt populations of California and Oregon (Low et al. 2007; Low and Shelbourne Douglas-fir is New Zealand’s second most important exotic 2007) have confirmed results from the earlier trials. conifer, though it falls a long way short of radiata pine in both Seed stands were initially developed to supply seed of known area planted (104 000 ha) and the current annual planting rate of origin and to incorporate some genetic improvement from local about 10 000 ha (Miller and Knowles 1994; Anon. 2002/2003). land races of probable Washington and southern Oregon origins. Douglas-fir has been grown in New Zealand since the end of the Seed stands were also later developed from broadly-based native 19th century but the main establishment periods have been 1900– provenance seedlots from Fort Bragg and from Santa Cruz on 1935, 1950–1970 and from 1990 onwards. There was little the Californian coast. interest in new planting of Douglas-fir during much of the 1970s and 1980s, as high establishment costs, long rotations (about twice
Australian Forestry 2007 Vol. 70 No. 1 pp. 28–32 C.J.A. Shelbourne, C.B. Low, L.D. Gea and R.L. Knowles 29
A breeding program was initiated by M.D. Wilcox in 1969 before seedlots collected in the USA (Superline B). The strategy was these provenance trial results became known. This was based on one of ‘recurrent selection for general combining ability’, with selection of 125 plus trees in stands aged 35–50 y of probable the creation of 15 sublines in Superlines A and B, each of around Washington origin. Trees were selected for superior diameter at 27 families. Sublines systematically included a variety of breast height (dbh); stem straightness; freedom from malforma- provenance origins, within which future mating would take place tion; and well-distributed, flat-angled and moderate-sized for breeding-population turnover, and between which pollination branches; and later wood density. Non-intensive phenotypic would occur in clonal seed orchards. This would ensure that selection and open-pollinated (OP) progeny testing resulted in although inbreeding would build up slowly within sublines over the planting of 125 OP progenies of the first series of selections generations, the production of seed from the seed orchard would in North and South Island forests in 1972, with a further 60 OP be from unrelated parents. progenies planted in 1973, selected also for wood density. The 1995 strategy involved complementary mating designs of By age 13 y the 1959 provenance trials were showing clearly polycrossing for estimation of breeding value and pair crossing that provenances from the fogbelt of the Californian and south to generate full-sib families for forward selection in Superline A. Oregon coasts were growing appreciably faster in volume than Superline B relied on OP mating, at least in the first generation, the NZ landrace and Washington provenances (M.D. Wilcox for advancement. The program for Superline A proved to be unpublished data). This finding indicated that the 1969 and 1971 unachievable for a variety of biological and operational reasons, selections in Washington material were unsuitable as the sole and after over a decade of effort sufficient seed of the projected basis for a breeding program. For this reason, and the lack of crosses has not been produced, either of polycrosses or pair- industry interest in the species at the time, the Douglas-fir crosses. This has necessitated the revision of the strategy in favour breeding program was put on hold for the next 14 y. of a single open-pollinated design that fulfils breeding value and the requirements of forward selection, and is more rapidly and The genetic improvement of Douglas-fir in NZ was reviewed in cheaply attainable. The two Superlines will be maintained in the 1988 (C.J.A. Shelbourne unpublished data) when prices of new strategy, but the use of open-pollinated mating will disrupt Douglas-fir sawlogs from unpruned stands had risen rapidly and the sublining. It is anticipated that control-pollinated orchards there was renewed interest by the industry in growing Douglas- may be adopted in NZ in the longer term, which would make fir. A new breeding program was initiated by making 186 new sublining unnecessary to achieve outcrossing in the orchards. selections in the earlier provenance trials countrywide, mainly from provenances from the coastal fogbelt of California and The most immediate deficiencies of Douglas-fir genetic improve- Oregon. Selections were also made in seed stands of Fort Bragg, ment prior to the change of strategy were the lack of progeny California provenance, and from stands of local land races of tests of the NZ-derived Superline A parents to provide breeding southern Oregon and Washington origin. A local seed company, values for selecting seed orchard clones, and the lack of a Proseed NZ Ltd, funded the assessment of the provenance trials, sufficient area of commercial seed orchards. The inability to the selection of plus trees (‘Superline A’ in the present breeding extensively cross the NZ Superline A selections (originally from strategy), and collection and grafting of scion material. Grafts of genetically-isolated populations) prevented the creation of an the 186 clones were subsequently planted as a clonal archive at adapted NZ landrace and breeding population. Proseed NZ Ltd’s Waikuku site (near Christchurch, South Island). This greater confidence in Douglas-fir led to the formation of an Selection criteria for Superline A candidates in the provenance FRI-Industry Douglas-fir Cooperative in April 1993, with base trials in 1988 were diameter at breast height, bole straightness, funding by the Ministry of Research, Science and Technology. absence of forking and ramicorn branching, and dense deep crowns. Selection of the parent trees of Superline B in the USA The restricted number of individuals and parentage of the coastal was for growth and form. Breeding objectives (the harvest-age fog belt provenances in the earlier trials, and the possibility of traits to be improved) were yield and improved log quality. In a some relatedness among the 186 selections from these, were of recent wood properties and sawing study of a stand of Douglas- concern. Hence in 1993 an FRI expedition was organised which fir of Fort Bragg, Californian origin, Knowles et al. (2003) have collected or acquired 240 OP progeny seedlots from 21 shown that lumber stiffness, the most important determinant of provenances in coastal Oregon and California (C.B. Low and quality for structural lumber, shows high variability among trees M.A. Miller unpublished data). These progenies constitute of Douglas-fir. New data on sonic velocity from the 1969 progeny Superline B (Fig. 1) and were planted in 1996 in three progeny/ tests show moderately high heritability for sonic velocity and provenance trials at one North Island and two South Island sites thus stiffness (L. Gea unpublished data). It was attempted in the at latitudes 38°, 41° and 45°S. These trials were first assessed at early 1970s to choose the breeding objectives and plus tree age 4 y (Low et al. 2007) and by 2008 should be ready for forward selection criteria through two individual-tree wood property and selection of the best trees for use in clonal seed orchards, and sawing studies. These determined the amount of tree-to-tree from which OP seed will be collected to establish a new genera- variation in mean plank stiffness (by machine stress grading) and tion of Superline B. its relationship to tree form and wood properties (C.J.A. Shelbourne, J.M. Harris, J.R. Tustin, I.D. Whiteside unpublished Breeding strategy data; M.D. Wilcox unpublished data). In both sawing studies tree- mean minimum stiffness as a 100 × 50 mm plank was well A strategy was formulated in 1995 (C.J.A. Shelbourne predicted through multiple stepwise regression on branch unpublished data) to include groups of selections from the NZ diameter, basic density and stem deviations (R2 = 0.62–0.79). A provenance trials and other stands (Superline A), and the OP recent study of eighteen 42-y-old trees from a densely-stocked
Australian Forestry 2007 Vol. 70 No. 1 pp. 28–32 30 Genetic improvement of Douglas-fir
Figure 1. Douglas-fir breeding and seed production
stand of Californian origin (Knowles et al. 2003) showed that involve collection of OP seed from each of the 186 NZ Superline most tree variation in timber stiffness (as a joist) was explained A clones in the Waikuku clonal archive, thus intermating and by density (r2 = 0.75), while branch diameter accounted for only turning over the generation for this Superline. Later, after the 24% of the variation in log mean timber stiffness. Mean tree 12-y assessment of the USA Superline B trials, collection of OP timber stiffness was also well predicted by microfibril angle, as seed from about 200 selections will turn over the generation of measured by SilviScan. this Superline.
Open-pollinated progeny testing is a practicable strategy to Use of open-pollinated seed for tree improvement progress the Douglas-fir breeding population, and to realise gains The revised and simplified long-term breeding strategy (Fig. 1) from orchard seed either by backward or forward selection. will be effected through future use of OP seed for progeny testing Provided at least one or two sites of the breeding population as well as breeding-population advancement. This will first progeny tests show early and good seed production, the mating
Australian Forestry 2007 Vol. 70 No. 1 pp. 28–32 C.J.A. Shelbourne, C.B. Low, L.D. Gea and R.L. Knowles 31 of selections by open pollination will reduce the length of the Predicted genetic gain in the breeding population breeding cycle by avoiding the costly and time-consuming steps and from seed orchards of grafted clonal archive establishment and control pollination. Provided balanced within-family selection is practiced, status On average, basal area of offspring of unselected parents from number (Ns) (Lindgren et al. 1996) of the breeding population the Californian fogbelt provenances near harvest age will show should be well conserved. This strategy utilises within-family 18% superiority over NZ seed sources originally of Washington genetic variance without the reduction in effective population origin (Kimberley and Knowles 2002). This was demonstrated size and status number that result from among-family selection, at age 29 y in replicated provenance trials with large plots on six thus ensuring better sustainability of the breeding program over widely-scattered sites with one to three plots per provenance. generations. As a part of research towards devising a new breeding strategy There is little disadvantage in using OP progenies of this species for radiata pine in NZ, genetic gains were deterministically versus polycross progenies for estimating breeding values and simulated over a range of trait heritabilities for selection in large for advancing a breeding population by balanced within-family Main and small Elite breeding populations and from seed orchards selection (Shelbourne et al. 2007). The main disadvantages of (Shelbourne et al. 2007). Gains from Main populations derived OP half-sib progenies versus full-sibs as a breeding population from open pollination, pollen mix crosses and full-sib crossing are the loss of full control of parentage which may result in using balanced within-family selection were highest for polycross selection of half-sib relatives from different families, leading to families (equivalent to open pollination in an isolated clonal increased inbreeding and reduced status number. In future, archive, e.g. Waikuku), somewhat lower from open-pollinated parental analysis through DNA markers might be used to avoid families in a thinned test and lowest for full-sib families. Gains selecting siblings from different OP families (Lambeth et al. from forward-selected orchards from OP, polycross and full-sib 2001). The same approach has been successfully used in a recent breeding populations were all very similar at different heritabili- study of (insect) open-pollinated breeding in a Eucalyptus nitens ties, and gains for backward selection of the parents of the forward clonal archive (Gea et al. 2007). selections were always less than from forward selection. Results of these simulations in large Main populations support the use of There are opportunities for gain from intensive phenotypic open pollination that forms the basis of the new Douglas-fir selection for stiffness in 15-y or older stands of second-generation strategy. Fort Bragg, Californian origin. These stands were established using seed collected in seed stands planted at the same time as The gene resources of coastal fogbelt populations of Douglas-fir the 1959 provenance trials, and are of a top-performing prove- have been well utilised to select a breeding population of about nance for growth. They form a resource from which intensive 400 parent trees that will eventually recombine the adaptation phenotypic selection for growth, form and sonic velocity can and variability of a wide variety of coastal populations. There realise higher gains than from other elements of the program. are preliminary indications from the performance of a few Conceptually, selections from this developed land race can form controlled crosses between Oregon and Californian parents that a small ‘Elite’ breeding population that can be managed by this outcrossing could generate some ‘hybrid vigour’. intensive among- and within-family selection for immediate gains, especially in stiffness, to supply seed orchard parents. Selection Conclusions may run down status number of the Elite population quickly without jeopardising the status number of the ‘Main’ breeding The Douglas-fir breeding program stalled because an ambitious population. program of polycross mating for estimation of breeding values, and pair crossing for generation turnover, in NZ-selected clones Future seed orchards from provenance trials failed to deliver sufficient crosses and seed. The new strategy uses open pollination in the breeding Open-pollinated clonal seed orchards of grafts are the most population for both generation turnover and for forward and efficient means of producing genetically-improved seed of backward selection of orchard parents. Douglas-fir in quantity, and they use technologies that are well developed. They have not been without problems in British The selections of Superline A from coastal fogbelt provenances Columbia and the Pacific Northwest; graft incompatibility has in the 1957 and 1959 provenance trials, archived at Waikuku, been serious in North America. However, compatible-rootstock have recently provided seed from open pollination within the families have been successfully developed there, and these should archive from natural crossing amongst a variety of coastal fog- be accessed for orchard establishment in NZ. belt populations, with some potential release of heterosis. Field trials of these families will rank the parents for roguing seed The immediate establishment of ‘Rolling Front’ clonal orchards orchards and will eventually provide another generation of the with the better Superline A clones and the new ‘stiffness’ Elite breeding population from mainly within-family selection. selections is a key component of the new strategy. These will be progressively rogued of poorer clones and extended with grafts USA Superline B of OP families from seed collected in the native of the best, as indicated by assessment of the OP progeny tests. populations of coastal California and Oregon is an unrelated set By 2008, forward selection within Superline B will add further of the same sort of material as in the 1959 provenance trials. fast-growing, high-stiffness selections to the clonal orchards. Seed collected from selections in these trials will recombine
Australian Forestry 2007 Vol. 70 No. 1 pp. 28–32 32 Genetic improvement of Douglas-fir genotypes of 21 different native populations, following some Gea, L., McConnochie, R. and Wynyard, S. (2007) Parental selection in the NZ environment. Selection of the best parents by reconstruction for breeding, deployment and seed orchard about 2008 and then open pollination within these progeny trials management of Eucalyptus nitens. New Zealand Journal of after thinning will parallel the collection of seed from the archive Forestry Science 37, 23–36. of Superline A, both of which will generate an improved Kimberley, M. and Knowles, L. (2002) Effect of provenance on growth provenance hybrid swarm. The propagation of the best of these and yield of Douglas-fir at the stand level. In: Proceedings, forward selections in a clonal seed orchard, along with those of Douglas-fir Cooperative Meeting, Christchurch 13–14 February 2002. Superline A, will produce commercial seed with a greatly increased potential for growth, log quality and stiffness. Knowles, R.L., Hansen, L.W., Downes, G., Kimberley, M.O., Gaunt, D.J., Lee, J.R. and Roper, J.G. (2003) Modelling within-tree and These two superlines form a solid and diverse basis for ongoing between-tree variation in Douglas-fir wood and lumber properties. and sustainable improvement of Douglas-fir in the long term, In: Proceedings, IUFRO All Division 5 Conference, Rotorua, NZ, 11–15 March 2003. provided they are managed skilfully through open pollinated mating and within-family selection, moderated by parentage Lambeth, C., Lee, B.C., O’Malley, D. and Wheeler, N. (2001) Polymix breeding with parental analysis of progeny: an alternative to full- reconstruction by DNA markers. sib breeding and testing. Theoretical and Applied Genetics 103, The new Elite population will derive largely from an adapted, 930–943. advanced-generation NZ land-race, originally of Fort Bragg, Lindgren, D., Gea, L.D. and Jefferson, P. (1996) Effective number and California, origin that has been a top performer for growth in all coancestry in breeding populations following within family trials. The intensive selection of plus trees that is possible in selection. Silvae Genetica 45, 52–59. these extensive stands will enable substantial gains in stiffness Low, C.B. and Shelbourne, C.J.A. (2007) Eight-year growth and form as well as growth and form to be quickly realised from orchards. of Douglas-fir provenances and New Zealand seed sources on higher altitude sites. New Zealand Journal of Forestry Science In future, intensive among- and within-family selection will (in preparation). provide a pre-production population from which to select orchard Low, C.B., Ledgard, N.J. and Shelbourne, C.J.A. (2007) Early growth clones. Rolling-front establishment of seed orchards of the best and form of coastal provenances and progenies of Douglas-fir at clones from these three populations will provide progressively- three sites in New Zealand. New Zealand Journal of Forestry improved seed for industry planting. Science (in preparation). Miller, J.T. and Knowles, F.B. (1994) Introduced Forest Trees in New Acknowledgements Zealand: Recognition, Role and Seed Source. Part 14. Douglas- fir Pseudotsuga menziesii (Mirb.) Franco. FRI Bulletin No. The re-starting of the Douglas-fir breeding program in 1988 was 124/14, Rotorua. supported by Peter Bolton of Proseed NZ, and the establishment Shelbourne, C.J.A. (1969) Tree Breeding Methods. Technical Paper and management of Waikuku orchard has been done by this 55, New Zealand Forest Service, Wellington. company since then. The Foundation of Research, Science and Shelbourne, C.J.A., Kumar, S., Burdon, R.D., Gea, L.D. and Dungey, H.S. Technology and the Douglas-fir Cooperative have supported the (2007) Deterministic simulation of gains for seedling and cloned breeding program since 1993 when the Cooperative was started. main and elite breeding populations of Pinus radiata and implica- We gratefully acknowledge the support of Proseed NZ, these tions for strategy. Silvae Genetica (in press). agencies and members of the Cooperative. Sweet, G.B. (1965) Provenance differences in Pacific Coast Douglas- fir. Silvae Genetica 14, 46–56. References Thulin, I.J. (1967) Provenance testing, Pseudotsuga menziesii. New Zealand Forest Service, Forest Research Institute Annual Report Anon. (1973) Douglas-fir management alternatives. Unpublished report 1966, p. 22. prepared by Kaingaroa Forest Rotorua Conservancy Staff. Anon. (2002/2003) New Zealand Forest Industry Facts and Figures.The Forestry Statistics Unit, Policy Information Group, Ministry of Agriculture and Forestry.
Australian Forestry 2007 Vol. 70 No. 1 pp. 28–32 Md Qumruzzaman Chowdhury, A.Z.M. Manzoor Rashid, Md Shah Newaz and Mabubul Alam 33
Effects of height on physical properties of wood of jhau (Casuarina equisetifolia)
Md Qumruzzaman Chowdhury1, A.Z.M. Manzoor Rashid1,2, Md Shah Newaz3 and Mabubul Alam1
1Department of Forestry, Shahjalal University of Science and Technology, Sylhet-3114, Bangladesh 2Email: [email protected], [email protected] 3Institute of Forestry and Environmental Sciences,Chittagong University, Chittagong, Bangladesh
Revised manuscript received 20 July 2006
Summary this species (Guha et al. 1970). The tree is used extensively for firewood (Burgess 1966). Moisture content, and tangential, radial and longitudinal shrinkage, and basic density were assessed in samples of wood from near Research interest has moved from taxonomy, species trials in the centre of the cross-section at three different heights in 25-y- different sites, and growth rates to suitability for shelterbelts, old jhau (Casuarina equisetifolia). The mean initial moisture tolerance to harsh environmental conditions, and yield studies. content was 41% in the butt log, 40% in the centre log and 48% However, wood properties have received little attention, in spite in the crown log. Mean tangential shrinkage was 10% in the butt, of increasing industrial use of the wood (Dudley and Fownes 9% in the centre and 11% in the crown. The corresponding 1992; El-Osta et al. 1992; Constantinides and Fownes 1994; Puri measurements for radial shrinkage were 6%, 5% and 5%, and for et al. 1994). longitudinal shrinkage 0.4%, 0.3% and 0.5%. There were no In Bangladesh, trees in jhau plantations are approaching their statistically significant differences in either moisture content or economic maturity in size and vigour, and must be used to the radial and longitudinal shrinkage with increasing height above greatest economic advantage. As information regarding wood ground, but differences in tangential shrinkage were significant properties is required for best utilisation, this study was intended at P < 0.05. The overall mean basic density was about 800 kg m–3, to evaluate the physical properties of this species. and this property varied a little with height in the tree. Keywords: wood properties; moisture content; shrinkage; density; Methods height; Casuarina equisetifolia Study area Introduction Samples were collected from a plantation on the Chittagong Casuarina equisetifolia L., named jhau in Bangladesh, is one of University campus, with coordinates of 91°50'E and 22°30'N. four species of the Casuarinaceae (a phylogenetically isolated The altitude is about 87 m above sea level, and the hills of this family and the sole representative of the order Casuarinales) which area are part of the South Fatikchari Range with mostly sandy naturally occur on subtropical and tropical coastlines in northern soils (Baten 1969). Annual rainfall is about 3000 mm, mostly Australia, and throughout northern Malaysia, Melanesia, falling between June and September. With minor variations, Polynesia and Vanuatu (Pinyopusarerk and House 1993). The humidity remains high at 70–85% throughout the year (Gafur et species has been introduced to many countries outside its natural al. 1979). range and used for soil stabilisation, reclamation work and coastal protection, as well as rehabilitation of degraded land (Midgley Selection of sample trees et al. 1983; El-Lakany et al. 1990). In Bangladesh this species has been planted for decades in coastal areas and near beaches Four plots, 10 m × 10 m (0.01 ha), were laid down at least two for sand dune stabilisation (Gafur et al. 1979). rows from the stand edge so as to exclude edge effects. The diameter at breast height over bark (dbhob) of all trees in each The tree has light red to dark reddish-brown heartwood, plot was measured, and five trees randomly selected from each sometimes darker in older trees, and usually-distinct, thick, buff- plot to give a total of 20 trees for testing. coloured sapwood with a fine to moderately-fine texture. The grain is typically straight or shallowly interlocked. The wide Specimen preparation medullary rays result in an attractive prominent figure on radial surfaces when logs are quartersawn. The wood is generally hard After felling, trees were docked to commercial-sized logs 5 m and dense, and tends to split when sawn (Burgess 1966). long. From each of the butt logs, a section 1 m long was cut with breast height as close to the log’s mid-point as possible. A 1-m Jhau is popular for use as underground piles for construction work, length was cut from the lower end of the centre and crown logs. as well as for fish-trap stakes and tool handles. Hardboard, A section about 500 mm long was cut from the largest internodal particleboard and woodchips for pulping are also produced from area clear of knots and other defects or injury in each of the 1-m
Australian Forestry 2007 Vol. 70 No. 1 pp. 33–36 34 Physical properties of wood of Casuarina equisetifolia logs. These samples were wrapped in polythene film for transport layer in fibres results in large shrinkage in the longitudinal to the laboratory, where they were processed a week after felling. direction. Typically a single specimen of 25 mm × 25 mm (in the stem cross-section) × 100 mm (in the stem longitudinal direction) was The magnitude of shrinkage and swelling is affected by the taken from the near the middle of a diametrical strip cut from amount of moisture gained or lost by wood when the moisture each section, a few millimeters from the pith and excluding content fluctuates between zero and fibre saturation point reaction wood. In the largest discs, more than one specimen was (Tsoumis 1991). Kollman and Cote (1968) explained that shrink- taken. If the specimen developed collapse during oven drying, it age differs in the three different directions due to the influence was replaced with a new sample from the same section. of wood rays and different arrangements of fibrils in the cell walls. In our study, radial and longitudinal shrinkages did not vary significantly with increasing height in the tree. However, Determination of physical properties tangential shrinkage did vary significantly with increasing height Moisture content was measured using the oven-dry method, and (Fig. 2), and the Tukey–HSD test showed that the crown wood shrinkage in the three linear directions was evaluated as dimen- had significantly greater shrinkage than wood from the two lower sional change from the green to oven-dry condition expressed as logs. This greater value could be due to the presence of more a fraction. Oven-dry measurements were taken after the specimens unrecognized tension wood in the crown samples than in those were dried to constant weight in an oven at 103 ± 2°C. The from lower positions. dimensions were measured in both green and dry conditions by a The effects of moisture content on different shrinkages (tangential, slide caliper, and mass measured on an electric balance to an radial and longitudinal) were assessed by linear regression (Figs 3, accuracy of 0.01 g. Basic density was determined from green 4 and 5). volume (using the water displacement method) and oven-dry mass. The coefficients of determination (r2) are 0.85, 0.80 and 0.86 respectively for the three relationships between shrinkage and Statistical analysis moisture content. The regression models showed a linear relationship, and the models are significant at P < 0.05. An ANOVA test was done to determine the significance of variation of physical properties (moisture content, density and shrinkage) at the three different heights (butt, centre and crown), followed by a post-hoc (Tukey–HSD) test if the ANOVA showed 60 significant differences. Relationships between moisture content 50 and shrinkage in different directions were evaluated by using a general linear model (GLM). 40
Results and discussion 30 20 The mean diameter over bark at breast height of the trees in the Moisture content (%) sample plots was 200 mm. 10
Heartwood of the 25-y-old trees was light red to reddish-brown. 0 Butt Centre Crown Moisture content is an important physical property of wood because it affects other physical and mechanical properties, and Figure 1. Average moisture content of jhau wood at different heights. consequently utilisation. In this trial, the mean moisture content The whiskers indicate the standard error of the means. of jhau was 41% in the butt log, 40% in the centre and 48% in the crown (Fig. 1). In comparison, Burgess (1966) found average moisture content to be 36.5% in Sabah-grown wood. The greater 14 2c moisture content in the crown samples is presumably due to the 1a Butt 12 greater proportion of juvenile wood in the upper bole, but the 1b Centre differences between heights were not significant. 10 a Crown 8 a a The mean tangential shrinkage was 10% in the butt log, 9% in 6 the centre and 11% in the crown section (Fig. 2). The correspond-
Shrinkage (%) 4 ing radial shrinkages were 6%, 5% and 5%, and longitudinal a a a shrinkages 0.4%, 0.3% and 0.5% respectively (Fig. 2). As 2 regularly discussed in the literature (e.g. Panshin and de Zeeuw 0 1980), tangential shrinkage is regularly about twice that of radial Tangential Radial Longitudinal shrinkage. Variation of shrinkage in different directions is due to Figure 2. Average shrinkage of jhau wood at different heights. Within the cellular structure and physical organisation of cellulose chain categories, different letters denote significant variation (ANOVA) at molecules within the cell walls (Schniewind 1989); the microfibril P < 0.05; in the tangential case different numbers denote significant angle of the S2 layer is an important factor (Koponen et al. 1989). variation (Tukey–HSD test) at P < 0.05. The whiskers indicate the Yamamoto et al. (1992) reported that the presence of a gelatinous standard error of the mean.
Australian Forestry 2007 Vol. 70 No. 1 pp. 33–36 Md Qumruzzaman Chowdhury, A.Z.M. Manzoor Rashid, Md Shah Newaz and Mabubul Alam 35
14 860 y = 0.1685x + 2.2011 12 r 2 = 0.85 840 10 )
8 –3 820 6 800 4 Density (kg m Tangential shrinkage (%) 2 780 0 0 102030405060 760 Moisture content (%) 740 Figure 3. Relationship between moisture content and tangential Butt Centre Crown shrinkage Figure 6. Average basic density of jhau wood at different heights. The whiskers indicate the standard error of the means.
9 y = 0.0888x + 2.4181 8 r 2 = 0.80 species in Cuba as 840 kg m–3, while Jain and Lala (1966) 7 reported Indian-grown wood as 740 kg m–3. As strength is directly 6 related to wood density (Shrivastava 1997), jhau will have 5 satisfactory strength properties. 4 3 Finally, a comparison was made with teak (Tectona grandis). 2 Teak is regarded in Bangladesh as a standard with which other Radial shrinkage (%) 1 locally-grown or imported timbers may be compared, e.g. the 0 strength properties of 40-y-old Chittagong-grown teak were used 0 102030405060by Yakub et al. (1978) to assess the relative suitability of other Moisture content (%) species for various purposes. The moisture content and shrinkage of jhau wood were similar to those of teak, but the density was Figure 4. Relationship between moisture content and radial shrinkage generally higher than that of teak.
Conclusion 0.7 y x = 0.0075 + 0.1436 Initial (green) moisture content of jhau varied little with increasing 0.6 r 2 = 0.86 height in the tree. Radial and longitudinal shrinkages did not vary 0.5 with increasing height, but tangential shrinkage did vary 0.4 significantly. Good correlations were found between the different shrinkages and initial (green) moisture content. Basic density, 0.3 which is considered an important indicator of strength, was 0.2 marginally higher in the butt log (about 800 kg m–3), but there was no significant variation with increasing height in the tree. Longitudinal shrinkage (%) 0.1
0 0 102030405060Acknowledgements Moisture content (%) We thank the Institute of Forestry and Environmental Sciences, Figure 5. Relationship between moisture content and longitudinal Chittagong University (IFESCU), for supplying the wood shrinkage samples, and the reviewers and editors of this journal for assistance in revising the manuscript.
References The mean basic density of jhau at different heights in the stem Baten, S.A. (1969) Revised working plan for the forest of Chittagong –3 (Fig. 6) was about 800 kg m , with no significant differences Division for the period of 1968–69 to 1977–78; Working Plans between heights. The density decreased by a small amount from Division-1, Dhaka (unpublished). butt to crown, possibly the result of a higher proportion of juvenile Burgess, P.F. (1966) Timbers of Sabah. Sabah Forest Records 6, 70–73. wood in the crown region (Zobel and Jett 1995). Constantinides, M. and Fownes, J.H. (1994) Nitrogen mineralization Our result that jhau is a high-density wood is supported by from leaves and litter of tropical plants: relationship to nitrogen, lignin and soluble phenol concentrations. Soil Biology and findings of other authors (e.g. Burgess 1966; Sosa-Suarez et al. Biochemistry 26, 49–55. 1990 ). Sosa-Suarez et al. (1990) measured basic density of the
Australian Forestry 2007 Vol. 70 No. 1 pp. 33–36 36 Physical properties of wood of Casuarina equisetifolia
Dudley, N.D. and Fownes, J.H. (1992) Preliminary biomass equations Pinyopusarerk, K. and House, A.P.N. (1993) Casuarina: An Annotated for eight species of fast-growing tropical trees. Journal of Tropical Bibliography of C. equisetifolia, C. junghuhniana and C. oligodon. Forest Science 5, 68–73. International Centre for Research in Agroforestry, Nairobi, Kenya, El-Osta, M.L., Megahed, M.M., El-Lakany, M.H. and Hegazy, S.S. 298 pp. (1992) Estimates of above-ground biomass and its distribution Puri, S., Sdingh, S. and Bhushan, B. (1994) Evaluation of fuelwood for Casuarina windbreak trees in the northwestern region of Egypt. quality of indigenous and exotic tree species of India’s semi-arid Alexandria Journal of Agricultural Research 37, 57–83. region. Agroforestry Systems 26, 123–133. El-Lakany, M.H., Turnbull, J.W. and Brewbaker, J.L. (1990) Advances Sosa-Suarez, C., Ibanez, A. and Hesse, R. (1990) Physical and strength in Casuarina Research and Utilization. Proceedings of the 2nd properties of Casuarina equisetifolia Forst. Part 1. Physical International Casuarina Workshop, 15–20 January 1990, Desert properties. Holztechnologie 30, 129–131. Development Center, AUC, Cairo, Egypt, 241 pp. Shrivastava, M. B. (1997) Wood Technology. Vikas Publishing House Gafur, M.A., Karim, A. and Khan, M.A. (1979) Phytosociological Pvt Ltd, Jangpura, New Delhi, India. studies of hills of Chittagong University Campus. Chittagong Schniewind, A.P. (1989) Concise Encyclopedia of Wood and Wood- University Studies, Part 2, Science 3, 11–28. Based Materials. Pergamon Press, 248 pp. Guha, S.R.D., Sharma, Y.K., Pant, R. and Dhandiyal, S.N. (1970) Tsoumis, G. (1991) Science and Technology of Wood. Chapman and Chemical, semi-chemical and mechanical pulps from Casuarina Hall, New York, 491 pp. equisetifolia. Indian Forester 96, 830–840. Yakub, M., Ali, M.O. and Bhattacharjee, D.K. (1978) Strength Jain, N.C. and Lala, M.K. (1966) Compressed wood from Indian Properties of Chittagong Teak (Tectona grandis) Representing timbers. Indian Forester 92, 730–738. Different Age Groups. Bulletin 4. Timber Physics Series, BFRI, Kollman, F.F. and Cote, W.A. (1968) Principles of Wood Science and Chittagong, 12 pp. Technology. Vol. 1. Springer-Verlag, Berlin, 592 pp. Yamamoto, H., Okuyama, T., Sugiyama, K. and Yoshida, M. (1992) Koponen, S., Toratti, T. and Kanerva, P. (1989) Modelling longitudinal Generation process of growth in stresses in cell walls IV. Action and elastic properties of wood. Wood Science and Technology of the cellulose microfibril upon the generation of the tensile 23, 55–63. stresses. Mokuzai Gakkaishi 38, 107–113. Midgley, S.J., Turnbull, J.W. and Johnston, R.D. (1983) Casuarina: Zobel, B.J. and Jett, J.B. (1995) Genetics of Wood Production. Springer- Ecology, Management and Utilization. Proceedings of an Verlag, Berlin, 337 pp. international workshop, Canberra, Australia, 17–21 August 1981. CSIRO Division of Forest Research, Canberra, 286 pp. Panshin, A.J. and de Zeeuw, C. (1980) Textbook of Wood Technology. 4th edn, McGraw-Hill, New York, 722 pp.
Australian Forestry 2007 Vol. 70 No. 1 pp. 33–36 Robin A.W. Gardner, Keith M. Little and Athol Arbuthnot 37
Wood and fibre productivity potential of promising new eucalypt species for coastal Zululand, South Africa
Robin A.W. Gardner1,2, Keith M. Little1 and Athol Arbuthnot3
1Institute for Commercial Forestry Research, PO Box 100281, Scottsville 3209, Pietermaritzburg, South Africa 2Email: [email protected] 3Mondi Business Paper, PO Box 1551, Richards Bay 3900, South Africa
Revised manuscript received 28 November 2006
Summary — although world paper consumption has increased five-fold — wood fibre still accounts for 90% of total fibre input. Non- In South Africa pressure is increasing to improve the quantity coniferous species, however, now contribute 30%, with an and quality of eucalypt wood being produced from the existing increasing fraction of this made up of eucalypts grown mainly in forestry land base, and to identify species that would enable the subtropics and tropics (Brown et al. 1997). Timber from profitable forestry on land currently considered marginal. In a eucalypt plantations has progressively replaced supplies from series of site × species interaction trials established on the natural tropical hardwood forests in a variety of solid-wood Zululand coastal plain during 1992, alternative eucalypt species markets and applications over the past decade (Flynn 2003). The were compared with a number of commercial eucalypt species main reasons for this shift are the decreased risk of diverse sources and clones at three sites. At rotation age of 7 y, the wood and of supply and increased demand for products derived from fibre productivity of the five most promising new species, namely ‘sustainably managed’ forests by major international retailing Corymbia henryi, C. citriodora ssp. citriodora, Eucalyptus chains (Dunne 2000; Flynn 2003). To meet the increasing global longirostrata, E. pilularis and E. tereticornis, was assessed at demand for wood, worldwide plantation hardwood timber outputs two of the sites. will need to be increased. This may be done either by increasing the amount of timber obtained from the existing land base, or by On the basis of merchantable wood production, E. longirostrata planting additional land (Brown and Hillis 1984; Kimmins 1994). Monto provenance showed excellent potential for high productivity (wet) sites on the Zululand coastal plain, and In South Africa, present and future land use policies are likely to C. citriodora ssp. citriodora and C. henryi excellent potential restrict the conversion of non-forested land to timber plantations. for marginal (moderately dry) sites. On the basis of fibre Factors that may contribute to increased timber yield from the production, E. longirostrata Monto and C. henryi showed existing land base include the use of site–species matching excellent potential for high productivity sites, and C. citriodora (Gardner 2001; Miranda and Pereria 2002), tree breeding and ssp. citriodora and C. henryi excellent potential for marginal sites. clonal propagation of pure species or hybrids (Denison and Kietzka 1993a; Duncan et al. 2000; Pierce and Verryn 2000) Eucalyptus longirostrata could be a good partner for E. grandis and improved silvicultural techniques (Du Toit et al. 2001; Smith and E. urophylla for the production of hybrids better adapted to et al. 2001; Little et al. 2002). Research involving the matching the range of environmental conditions in coastal Zululand than of a wide range of eucalypt species to site conditions within the either of the latter two species. In such hybrid combinations, the different geographic regions in South Africa has been carried high basic wood density of E. longirostrata could be of advantage out since the early 1950s (Gardner 2000). Of these regions, to the local hardwood woodchip export industry. Zululand (lying between 27° and 29°S latitude) is one of the Keywords: plantations; adaptation; species trials; site-species matching; most important plantation forestry areas in South Africa, productivity; wood properties; pulping; Eucalyptus; Corymbia; South contributing 20% of the country’s hardwood pulpwood (DWAF Africa 2005). Large-scale planting of eucalypts on the Zululand coastal plain began during the early 1920s and was based almost entirely Introduction on Eucalyptus grandis seedlings (Poynton 1979; Harrison 1993). To counter the incidence of disease associated with E. grandis Wood is one of man’s most important resources, and its combined with the desire to extend the areas planted to drier significance is increasing (Brown and Hillis 1984). Pulp is an regions, forestry had by 1980 become almost entirely based on important end use of wood, accounting for 522 million m3 or clones of either E. grandis or hybrids of E. grandis with 15% of total world wood consumption in 2004 (FAO 2005). In E. urophylla, E. camaldulensis or E. tereticornis (Denison and the 1950s, 95% of paper was made from wood fibre, and 90% of Kietzka 1993b; Gardner 2001). Despite meeting industry that fibre was obtained from coniferous trees. Forty years later objectives (reduced incidence of disease and increased drought
Australian Forestry 2007 Vol. 70 No. 1 pp. 37–47 38 Productivity of new eucalypts in Zululand tolerance and growth), the aforementioned species and hybrids to the driest and least productive site at False Bay (MAP 764 were not adequately adapted to the large areas of marginal mm) in the north (Table 1). Apart from False Bay having a much environments available (Gardner 2001). lower MAP than Terranera, the higher evaporation at False Bay exacerbates moisture stress at this site. As pulping data were To explore potential alternative eucalypt species for the region, available only for Terranera and False Bay (Gardner 2001), wood three site × species interaction trials were established on the and fibre productivity was investigated only in these two trials. coastal plain during 1992. At rotation age, 7 y after planting, the trees were assessed for various growth attributes at all three sites, Trial design and treatments and the best-performing species and controls across all sites were sampled at two sites, Terranera and False Bay, and tested for At each site, an experiment consisting of 15 eucalypt species pulping properties. The results of these assessments, reported by and clones (treatments) was laid out in a randomised complete- Gardner (2001), revealed that Corymbia henryi (Hill and Johnson block design. Some of the species and inter-specific hybrids 1995), C. citriodora ssp. citriodora (McDonald and Bean 2000), (clones) were represented by more than one seedlot or clone E. longirostrata (Chippendale 1988) (formerly E. punctata var. (Gardner 2001). Each experiment consisted of three replicates longirostrata), E. pilularis and E. tereticornis had the best overall of 30 treatments. Treatment plots consisted of 30 trees (5 rows × growth potential of all ‘new’ species tested. Further investigations 6 trees in each row) with the inner net plot of 12 trees being were then carried out at Terranera and False Bay to determine measured (3 rows × 4 trees in each row). For the variates ‘stem the potential wood and fibre productivity of each of the latter straightness’, ‘mean infection index’, ‘merchantable volume’, species in coastal Zululand conditions. The results of the latter ‘wood yield’ and ‘fibre productivity’, all measured trees were investigations are presented and discussed in this paper. used. Details of the origins of the seedlots and clones included in the investigations of wood and fibre productivity are presented Material and methods in Table 2. The Australian provenances (seedlots) were chosen for possible suitability to trial site conditions and availability of Study sites seed. The control treatments were local, improved seedlots of E. grandis and E. urophylla, the best local-performing natural- Mean annual precipitation (MAP) is the most important bio- stand provenance of E. camaldulensis (Petford) (Darrow 1983; climatic variable determining productivity of forest tree Eldridge et al. 1993) and local top-performing hybrid clones of plantations in sub-tropical, coastal Zululand (Gardner 2001), E. grandis × E. camaldulensis and E. grandis × E. urophylla where MAP decreases rapidly with increasing distance from the (Denison and Kietzka 1993a; Eldridge et al. 1993). coast and also from south to north (Schulze 1997). The predominant soil parent materials in the area, Quaternary Measurements and assessments aeolianite and alluvium, have given rise to soils with generally sandy textures (Herbert and Musto 1993). The soils most typical Tree growth attributes on the coastal plain are arenic lixisols and arenic kandiustults (FAO–UNESCO 1974). Survival was monitored through the duration of the experiment, and stem straightness, incidence of disease, height (ht) and The initial three trials were sited along a linear transect of the diameter at breast height (dbh) were assessed or measured at commercial forestry belt (Gardner 2001), from the wettest and rotation end (7 y of age). most productive site at Terranera (MAP 1197 mm) in the south,
Table 1. Site descriptions and planting details for trials at Terranera and False Bay in Zululand, South Africa
Trial name Site conditions Terranera False Bay Climatic description Warm, wet Hot, dry Latitude and longitude 28°42'04''S, 32°08'04''E 28°00'22''S; 32°19'51''E Altitude (m asl) 25 75 Rainfall distribution pattern Summer maximum Summer maximum Mean annual precipitation (mm)a 1197 764 Mean annual temperature (°C)b 21.7 21.9 Mean annual A-pan evaporation (mm)b 1479 1544 Soils: Taxonomyc Arenosol Arenosol Depth (m) > 1.2 > 1.2 Planting date 24 June 1992 14 May 1992 Spacing at planting (m) 2.7 × 2.7 3 × 2.5
a Eight-year mean for the period 1992–1999 b Schulze (1997) c FAO–UNESCO (1974)
Australian Forestry 2007 Vol. 70 No. 1 pp. 37–47 Robin A.W. Gardner, Keith M. Little and Athol Arbuthnot 39
Table 2. Origins of the seedlots and hybrid clones evaluated for wood and fibre productivity in trials at Terranera and False Bay in Zululand, South Africa
Origin Nature of treatment Seedlot/ No. of Altitude and treatment name clone no. Locality Latitude Longitude seed parents (m asl) Seedlots E. camaldulensis W 16720‡ Petford area, Qld 101 17°24' 145°2' 590 E. grandis W EG_1* Zululand (SO) 080 — — — E. longirostrata W,F 15602‡ 27 km SE of Gympie, Qld 002 26°18' 152°49' 380 E. longirostrata W,F 15637‡ NW of Monto, Qld 027 24°49' 150°57' 500 E. pilularis W,F 13537‡ 10 km W of Beerburrum, Qld 005 26°57' 152°52' 040 E. tereticornis W,F 13319‡ N of Woolgolga, NSW 006 29°55' 153°12' 030 E. urophylla W EU_1* Zululand (SO) 116 — — — C. citriodora ssp. citriodora W,F 18062‡ Expedition Range, Qld 002 24°37' 149°2' 400 C. henryi W,F 13572‡ S of Grafton, NSW 010 29°45' 152°58' 085 Hybrid clones E. grandis × E. camaldulensis W,F GC_2** Zululand — — — — E. grandis × E. urophylla W,F GU_2** Zululand — — — —
‡ Imported, unimproved seedlot obtained from CSIRO, Australia * South African commercial seedlot **South African commercial hybrid clone W Species or clone selected for merchantable wood volume determination F Species or clone selected for pulping property determination (note: E. pilularis sampled at Terranera only)
Using a four-point scoring system, stem straightness was assessed the final assessments at 7 y were selected for accurate determin- as follows: ation of stem volume. From each treatment plot, six trees, free • 1 = poor stem straightness with less than 25% of the bole from defects, were randomly selected. This equated to 18 trees straight and the remainder undulating or spiraling per treatment at each site. For each of these trees, height to a minimum over-bark stem diameter of 0.07 m (H0.07) was • 2 = poor to fair stem straightness where some portion of the determined, and under-bark diameter measured at each 1-m bole is straight, but the degree of undulation, kinking or interval from tree base to the point of H . For each 1-m stem spiraling in the stems makes the tree unsatisfactory on the whole 0.07 section, the under-bark volume (Vsec) was calculated using the • 3 = good stem straightness where besides some minor defects formula for a truncated cone. The sum of these was used to the bole is otherwise straight determine merchantable under-bark volume (Vm) per tree. log10 transformations of V , dbh and ht were used for determining • 4 = excellent, where the entire bole is straight with only one or m separate Schumacher and Hall (1933) volume models on an two very minor defects. individual species basis for each site. Using these derived data Trees were not generally downgraded for forking. However, a and stocking for relevant treatment plots, total ‘merchantable tree that scored 4 (excellent) was downgraded to score 3 (good) wood volume’ (Vmha) per species or clone per site was then if any forking occurred above breast height. These scores were calculated. combined, and the plot means used for analysis. Wood and pulping properties Each measured tree was rated for visual symptoms of four major stem diseases encountered in Zululand, Botryosphaeria dothidea, As the elected pulping laboratory could accommodate only a Coniothyrium (various spp.), Chrysoporthe austroafricana limited number of wood samples at the time, the wood and pulping (previously described as Cryphonectria cubensis) (Gryzenhout property tests were confined to the six ‘new’ species most et al. 2004) and Endothia gyrosa (0 = free of symptoms; 1 = promising across all three sites and two selected commercial slightly infected; 2 = moderately infected; 3 = severely infected). clones based on 4-y measurements (Gardner 2001). Thus, of the The combination of these scores on a treatment plot basis 11 taxa assessed for merchantable wood volume, only eight were provided a ‘mean infection index’ (MII) to give an overall assessed for wood and pulping properties (Table 2). For the latter indication of susceptibility to disease. MII was calculated by investigations, two trees, free of defects and closest to the mean averaging the four scores for individual disease infection for each plot diameter at breast height, were selected from each treatment seedlot or clone. plot (six trees per taxon per site). When felled, two discs 20 mm thick were cut at 1-m intervals along the length of the tree bole
Stem volume from the base to H0.07. Of the 15 species and clones that were planted at each site, only One set of discs was used to assess whole-tree density (TAPPI those 11 that exhibited good growth across all three sites during test method T258 om-89) and extractable content of the wood
Australian Forestry 2007 Vol. 70 No. 1 pp. 37–47 40 Productivity of new eucalypts in Zululand
(TAPPI test method T204 om-88). The product of the Specific consumption is important as it directly affects the merchantable wood volume per hectare (m3 ha–1) and the wood economics of pulping, and its effect is magnified where the raw density (kg m–3) divided by 1000 gives an indication of the wood wood is shipped for processing overseas. yield per hectare (t ha–1). Unbleached kraft-pulp tests were carried 3 –1 out at the Mondi Business Paper pulp-mill laboratory, Richard’s Using the data for merchantable wood volume (m ha ) and SPY –1 –1 Bay, South Africa. The cooking and analytical process is briefly (%), wood yield (t ha ) and fibre production (t pulp ha ) were described as follows: calculated for each species and clone.
To determine the extractable content, individual wedges from Statistical analyses each disc were chipped and Wiley milled in order to obtain a sample of air-dried sawdust to pass through a 0.40-mm screen Prior to analysis, Bartlett’s test (Snedecor and Cochran 1980) (TAPPI test method T257 cm-85). The ground wood from the was used to test the assumption of homogeneity of variance, wedges was combined to form a representative whole-tree bulk necessary to permit a valid analysis of variance. The variate ‘mean sample, and the ethanol-benzene (T204 om-88) and hot water infection index’ (MII) was transformed (( (x + 1) ), Gomez and (T207 om-88) extractable contents were then determined for each Gomez 1984) prior to analysis. The analysis of variance (based sample. on plot means) appropriate for a randomised complete block design was used to test for treatment effects (F-test) using The second set of discs from each tree was used for pulping tests. Genstat® for WindowsTM (Lane and Payne 1996). Only if the F- The discs were bulked for each taxon for each site (16 taxa value was significant were treatment differences further samples), and chipped by a guillotine-style laboratory chipper to investigated using least significant differences. These were produce chips of a uniform size. Samples were pulped in an calculated as the product of the standard error of the difference ‘Aurora’ electrically heated, batch type, re-circulating liquor of the means and the tabulated t-value (Steel and Torrie 1981) digester using the Kraft process. The cooking conditions used in for relevant degrees of freedom. this study were similar to those used in the batch digesters at the Richards Bay mill pre-2005. The cooking conditions when Results and discussion applied to the normal mix of woodchips at Richards Bay would result in a pulp the kappa number of which was approximately 20. Stem straightness
Pulping conditions were as follows: In both trials, the hybrid clones ranked highly with respect to • Active alkali charge (% Na2O) of oven-dry wood– 15% stem straightness (Table 3). This was anticipated, as both the • Sulphidity of the cooking liquor – 25% clones and their parent trees were the products of breeding programs where stem straightness was an important selection • Liquor : wood ratio – 4.6 ml : 1 g criterion. • Pulping cycle: Ambient to 170°C – 90 minutes Of all seedling treatments, C. citriodora ssp. citriodora and Cooking time at 170°C – H-factor 900 (± 50 minutes) C. henryi were the most consistent across sites in producing • Degassing was carried out at 90°C and at 110°C to remove straight stems. The stem straightness of E. longirostrata (Monto) gasses not condensable in water at such a rate that no liquor was strongly influenced by site, as the species scored above the trial was lost from the digester average at Terranera (3.3) but below the trial average at False Bay (2.5). The hybrid parent control species E. camaldulensis • Cooling to atmospheric pressure at end of cook – 15 minutes and E. urophylla were consistently poor at both sites. Since stem A spent (black) liquor sample was taken through a coil condenser straightness is moderately heritable, one would expect that this at the end of the cook, prior to cooling, and then analysed for trait in any of the above species or provenances could be improved residual alkali content (TAPPI test method T625 om-85). The through breeding (Zobel and Talbert 1984; Verryn 2000). Lee et kappa number of cooked pulp was determined using TAPPI test al. (2005a) reported improved stem straightness in C. torelliana method T236 cm-85. The kappa number is the volume (ml) of × C. citriodora ssp. variegata hybrids over and above that in 0.1N potassium permanganate solution consumed by one gram either parent. of moisture-free pulp. The results are corrected to 50% consumption of the permanganate added. Immediately after Disease removal from the digester, the pulp samples were de-fibreised using a blunt-bladed disintegrator. The disintegrated pulp was The mean infection index provides an indication of the overall screened through a Packer screen fitted with a 0.8-mm slotted susceptibility of the species to the four stem diseases most screen to remove the shives and uncooked material. The pulp prevalent in Zululand. A high incidence of disease is undesirable fibres that passed through the slots were used to calculate the as it may lower site productivity through either a reduction in screened pulp yield (SPY), while the retained material was used growth or increased mortality (Denison and Kietzka 1993a). In to calculate the percentage rejects. SPY, therefore, represented South Africa, bark is manually removed from the stem in-field mass of pulp produced per mass of oven-dry wood charged to (Herbert and Robertson 1991), and the presence of disease hinders the digester, expressed as a percentage. This indicates the amount this process (Nicol 1992). The most disease-prone treatment was of pulp produced relative to the amount of wood pulped. the local E. grandis commercial seedlot, which had by far the highest MII scores of all treatments at both Terranera (1.25) and Specific consumption (volume of wood consumed to produce False Bay (1.26) (Table 3). Although only E. grandis had MII one tonne of pulp) was calculated from SPY and basic density.
Australian Forestry 2007 Vol. 70 No. 1 pp. 37–47 Robin A.W. Gardner, Keith M. Little and Athol Arbuthnot 41
Table 3. Summary of the analyses of variance and data for end-of-rotation (7 y of age) measurements in trials at Terranera and False Bay in Zululand, South Africa
Stem straightness Mean infection index (MII) Merchantable volume (score, 1–4 (best)) (score, 0–3 (poorest)) (m3 ha–1) Terranera False Bay Terranera False Bay Terranera False Bay Analysis Source of variation df Mean squares Replicate 02 0.013 0.299 0.0024 0.0012 35** 481** Species 10 0.303** 0.834** 0.0148** 0.0136** 24.204** 9.063** Residual 20 0.032 0.074 0.0013 0.0024 3.202** 760**
Summary of data Treatment Treatment means ‡ E. camaldulensis 2.51d 2.13f 1.042a 1.057a 03,109e 00,88ef E. grandis 3.52abc 2.64cde 1.253b 1.262b 03,235bc 0,104def E. longirostrata (Gympie) 3.12c 2.81cd 1.040a 1.019a 03,171cde 0,129cde E. longirostrata (Monto) 3.33bc 2.46def 1.031a 1.034a 03,223bc 0,141cd E. pilularis 3.29bc 2.77cde 1.026a 1.070a 03,206cd 00,85f E. tereticornis 3.33bc 2.35e 1.080a 1.064a 03,173cde 0,141cd E. urophylla 3.14c 2.14f 1.024a 1.024a 03,202cde 00,85f C. citriodora ssp. citriodora 3.39abc 3.31ab 1.044a 1.050a 03,114de 0,187ab C. henryi 3.36bc 2.99bc 1.007a 1.038a 03,184cde 0,166bc E. grandis × E. camaldulensis 3.67a 3.59a 1.000a 1.096a 03,320b 0,214a E. grandis × E. urophylla 3.66a 3.58a 1.014a 1.079a 03,422a 0,228a Mean 3.30 2.80 1.051 1.072 03,214 0,142 sed+ 0.15 0.22 0.030 0.040 00,046 00,20 Coefficient of variation (units) 5.4 9.7 3.5 4.5 00,026 00,17
** Significant at P < 0.01 ‡ Within each column, values followed by the same letter are not significantly different (P < 0.05) according to Student’s t-test + Standard error of the treatment differences scores significantly different from the other treatments at each emphasise that the hybridisation of two eucalypt species does site, this species and E. tereticornis at Terranera and the E. grandis not necessarily result in the F1 hybrid having greater disease × E. camaldulensis and E. grandis × E. urophylla commercial resistance than either parent. To the contrary, the degree of clones at False Bay all had MII scores above the trial mean, resistance in the F1 is usually intermediate to varying degrees or suggesting a possibility of above-average disease susceptibility similar to that of the more susceptible parent. of these treatments. Eucalyptus camaldulensis, E. urophylla, E. longirostrata (both Gympie and Monto provenances), Survival C. citriodora ssp. citriodora and C. henryi all had MII scores below the trial mean at both sites, suggesting some possible The main factors influencing survival were sandblast during the resistance to the particular stem diseases assessed. The high first year after planting, and wind-throw, drought and stem disease infection scores in the E. grandis commercial seedlot at diseases during subsequent years (Gardner 2001). Survival up to both Terranera and False Bay suggest that the potential for 10 months after planting was good, with the mean survival for breeding for resistance to the stem diseases prevalent in Zululand Terranera and False Bay being 93% and 91% respectively is limited in the particular E. grandis local landrace (Eldridge et (Table 4). Most species recorded 80% or greater survival at both al. 1993). By contrast, the unimproved Gympie and Monto sites, although E. pilularis at Terranera, and E. urophylla and seedlots of E. longirostrata showed low disease infection scores C. citriodora ssp. citriodora at False Bay were exceptions, all at both Terranera (1.04 and 1.03) and False Bay (1.02 and 1.03), recording 78% survival at the respective sites. By 48 months, suggesting that resistance to the particular diseases is present in species originating from humid climates, such as E. urophylla the particular natural populations of this species. and E. pilularis, suffered increased mortality resulting from the 1992–1994 drought in the region (Darrow 1997). By rotation The results of the stem disease assessments highlight two impor- end, the mean survival figures for Terranera and False Bay were tant points. First, the high level of susceptibility to stem diseases 87% and 81% respectively. Eucalyptus grandis, E. pilularis, in the local, commercial E. grandis seedling material; secondly, E. urophylla and C. citriodora ssp. citriodora all recorded that gains in disease resistance may be possible through inter- survival below the trial mean at both sites. Eucalyptus pilularis specific hybridisation (e.g. the E. grandis × E. camaldulensis showed the poorest and second-poorest survival of all treatments hybrid clone at Terranera) and possibly via hybrids of at Terranera (64%) and False Bay (64%), respectively. At E. longirostrata with E. urophylla. Potts and Dungey (2004)
Australian Forestry 2007 Vol. 70 No. 1 pp. 37–47 42 Productivity of new eucalypts in Zululand
Terranera the survival of E. grandis was satisfactory (83%), but the only seedlots to produce merchantable wood volumes above when subjected to the more severe conditions at False Bay the the trial mean (214 m3 ha–1). Corymbia henryi was unable to species had the poorest survival (50%) of all 11 treatments. The capitalise on the favourable growing conditions at this site, high survival of E. camaldulensis and the E. grandis × producing only 184 m3 ha–1. The E. camaldulensis control E. camaldulensis commercial hybrid clone at Terranera (100% (108 m3 ha–1) produced the lowest merchantable volume of all and 100% respectively) and False Bay (97% and 97% species at Terranera. At False Bay, C. citriodora ssp. citriodora respectively) confirmed that E. camaldulensis is well adapted to (187 m3 ha–1) and C. henryi (166 m3 ha–1) were the only seedlots environmental conditions in coastal Zululand (Darrow 1983; to produce merchantable volumes above the trial mean Eldridge et al. 1993). The results suggest that, of all alternative (142 m3 ha–1), outperforming the E. grandis control (104 m3 ha–1). species tested, C. henryi and E. longirostrata (Monto) have the At the same site, C. citriodora ssp. citriodora (187 m3 ha–1), greatest ability to survive under the wide range of environmental C. henryi (166 m3 ha–1), E. longirostrata Monto (141 m3 ha–1) conditions of the coastal Zululand commercial forestry belt. and E. tereticornis (141 m3 ha–1) outperformed the two control species E. camaldulensis (88 m3 ha–1) and E. urophylla Merchantable volume (85 m3 ha–1). Corymbia citriodora ssp. citriodora appeared better adapted to the drier conditions at False Bay where the seedlot Merchantable volume (m3 ha–1) as calculated here is a good was the most productive of the new species, compared to being indicator of the potential productivity of a species or site, as it the second-lowest yielder at Terranera. In contrast, E. pilularis incorporates both stocking and the volume of the tree that can be appeared better adapted to the wetter conditions at Terranera used by wood processors. The Terranera site was more productive where the seedlot was third-most productive species (206 m3 ha–1) than False Bay, producing 50% more merchantable wood volume compared to being lowest-yielding seedlot at False Bay (Table 3). The commercial hybrid clones of E. grandis × (85 m3 ha–1). E. camaldulensis and E. grandis × E. urophylla outperformed all seedlots at either site, the latter producing the greatest The results highlighted the following: first, of all ‘new’ species eval- merchantable volume at both Terranera and False Bay (422 m3 uated, E. longirostrata (Monto), C. citriodora ssp. citriodora and ha–1 and 228 m3 ha–1, respectively). C. henryi showed the most promise for merchantable wood pro- duction on the coastal plain; secondly, although E. longirostrata At Terranera, the E. grandis control (235 m3 ha–1) and the appears better adapted to highly productive (wet) sites, the species unimproved E. longirostrata Monto seedlot (223 m3 ha–1) were
Table 4. Summary of the analyses of variance and data for tree survival at the end of the first and fourth growing seasons and at the end of the rotation (7 y of age) in trials at Terranera and False Bay in Zululand, South Africa
Terranera False Bay
10 months 48 months 84 months 10 months 48 months 84 months Analysis Source of variation df Mean squares Replicate 02 59* 82** 59* 044* 059* 280* Species 10 215* 357** 332* 250* 365* 732* Residual 20 75* 59** 84* 095* 138* 196
Summary of data Treatment Treatment means (%)‡ E. camaldulensis 100a 100a 100a 100a 100a 97.2a E. grandis 091.7ab 091.7ab 083.3bc 080.6bc 072.2c 50.0d E. longirostrata (Gympie) 100a 100a 088.9abc 097.2ab 097.2ab 75.0abc E. longirostrata (Monto) 100a 100a 094.4ab 100a 097.2ab 91.7ab E. pilularis 077.8b 066.7c 063.9d 083.3bc 077.8b 63.9c E. tereticornis 097.2a 094.4a 088.9abc 097.2ab 091.7ab 91.7a E. urophylla 086.1ab 080.6b 077.8cd 077.8c 072.2c 69.4bc C. citriodora ssp. citriodora 088.9ab 088.9ab 083.3bc 077.8c 077.8b 75.0abc C. henryi 100a 097.2a 091.7abc 097.2ab 097.2ab 91.7ab E. grandis × E. camaldulensis 100a 100a 100a 097.2ab 097.2ab 97.2a E. grandis × E. urophylla 080.6b 080.6b 080.6bc 091.7abc 091.7ab 91.7ab Mean 092.9 090.9 086.6 090.9 088.4 81.3 sed+ 007.08 006.27 007.50 007.96 009.58 11.45 Coefficient of variation (units) 009.3 008.4 010.6 010.7 013.3 17.2
* Significant at P < 0.05 ** Significant at P < 0.01 ‡ Within each column, values followed by the same letter are not significantly different (P < 0.05) according to Student’s t-test + Standard error of the treatment differences
Australian Forestry 2007 Vol. 70 No. 1 pp. 37–47 Robin A.W. Gardner, Keith M. Little and Athol Arbuthnot 43 also appears to have good potential for use as a hybrid parent in The Mondi Business Paper mill where the wood samples were inter-specific combinations with low-density phenotypes of processed requires a SPY of ≥51.5%, a kappa number of ≤22, E. grandis and E. urophylla for marginal (dry) sites; thirdly, and a pulpability factor of ≥2.34 (the latter factor is derived by C. citriodora ssp. citriodora (Expedition Range), unlike its close dividing SPY by kappa number, giving a good indication of relative C. henryi (Grafton), appears better adapted to marginal pulpwood quality without having to do multiple cooks and rather than highly productive growing conditions in Zululand. In interpolate to the desired kappa number of 20). Ideal basic wood the case of C. henryi, lack of drought tolerance may not be density for local kraft pulping is 480–520 kg m–3, while a higher provenance-specific, as in a later (1996-established) trial series density of 520–650 kg m–3 is preferred for woodchips exported in a hotter and drier sector of Zululand, the C. henryi Myrtle for kraft pulping. Creek provenance showed only moderate drought-tolerance and was outperformed by both the Monto and Chinchilla provenances At Terranera, only the E. grandis × camaldulensis clone –3 of E. longirostrata on the basis of survival and growth (Gardner (500 kg m ) fell within the ideal ‘basic wood density’ range for –3 2004, 2006). local kraft pulping. The densities of E. pilularis (476 kg m ) and E. grandis × E. urophylla clone (411 kg m–3) were both lower than the ideal range, whilst those of E. longirostrata Monto Wood and pulping properties (574 kg m–3), E. longirostrata Gympie (637 kg m–3), C. henryi Site appeared to substantially influence wood and pulping (581 kg m–3) and C. citriodora ssp. citriodora (637 kg m–3) were properties, with species and clones showing lower wood densities above the optimum range. The latter four taxa had basic densities and higher pulp yields at the wet site, Terranera, than at the far within the ideal range for woodchip exports. drier site, False Bay (Table 5). Similar effects of site on pulping At False Bay, only the E. grandis × urophylla clone (502 kg m–3) properties have been observed for E. grandis and E. grandis × fell within the optimum basic density range for local kraft pulping. E. camaldulensis and E. grandis × E. urophylla clones grown in Eucalyptus longirostrata Monto (632 kg m–3), E. tereticornis Zululand (Retief et al. 1997; Stanger 2004). (549 kg m–3) and the E. grandis × camaldulensis clone Species ranks for basic wood density were rather consistent (596 kg m–3) were within the optimum density range for export between sites. Although basic wood density in E. longirostrata woodchips. appeared strongly influenced by provenance, in general, Although high-density wood offers some important advantages C. citriodora ssp. citriodora, C. henryi and E. longirostrata were over low-density wood, such as improved digester productivity, consistently at the upper end of the basic density range, whereas paper quality can be negatively affected when wood density the E. grandis × E. urophylla clone and the E. tereticornis seedlot becomes excessive (Downes et al. 1997). The basic densities of were consistently at the lower end of the range for either site the three promising ‘new’ species, E. longirostrata, C. henryi (Table 5).
Table 5. Summary of the kraft pulping results for selected eucalypt species and hybrid clones at 7 y of age in trials at Terranera and False Bay in Zululand, South Africa
Basic density Screened pulp yield Pulpability factor Specific consumption Location and treatment Kappa no. (kg m–3 wood) (SPY) (%) (SPY kappa–1) (m3 wood t pulp–1) Terranera E. longirostrata (Gympie) 637 51.0 24.0 2.12 3.08 E. longirostrata (Monto) 574 52.9 21.4 2.47 3.29 E. pilularis‡ 476 52.0 20.8 2.50 4.04 E. tereticornis 465 50.3 24.5 2.06 4.28 C. citriodora ssp. citriodora 637 54.3 29.4 1.85 2.89 C. henryi 581 55.9 19.5 2.87 3.08 E. grandis × E. camaldulensis 500 51.4 22.5 2.29 3.89 E. grandis × E. urophylla 411 53.0 21.3 2.49 4.59 False Bay E. longirostrata (Gympie) 667 47.2 32.5 1.45 3.18 E. longirostrata (Monto) 632 48.2 31.3 1.54 3.28 E. tereticornis 549 47.3 34.6 1.37 3.85 C. citriodora ssp. citriodora 672 53.3 21.1 2.52 2.79 C. henryi 658 53.5 22.4 2.39 2.84 E. grandis × E. camaldulensis 596 49.7 28.7 1.73 3.38 E. grandis × E. urophylla 502 45.2 38.3 1.18 4.41 ‡ E. pilularis was sampled at Terranera only
Australian Forestry 2007 Vol. 70 No. 1 pp. 37–47 44 Productivity of new eucalypts in Zululand and C. citriodora ssp. citriodora, were all above the local ideal were within the required range of ≥2.34. At the dry False Bay range for kraft pulping at both trial sites. Wood density is a highly site, only C. citriodora ssp. citriodora (2.52) and C. henryi (2.39) heritable trait in Eucalyptus (Malan 1988; Eldridge et al.1993). fell within the required range. Corymbia henryi, therefore, was Therefore density in any of the aforementioned species should the only species or clone to perform consistently well across sites be amenable to change, if required, through breeding including on the basis of ‘pulpability’. hybridisation (Verryn 2000). ‘Specific consumption’, the volume of wood (in cubic metres) In the case of E. longirostrata, local experience has shown that required to produce one tonne of pulp, has important economic the species easily hybridises (controlled crosses) with low-density implications for both pulp-processing and woodchip exporting species such as E. grandis or E. urophylla (S. Verryn, CSIR, industries. In the case of the former, low specific consumption pers. comm. 2005). Therefore, in addition to within-species increases digester productivity, whilst in the latter, low specific con- selection, inter-specific hybridisation would probably be useful sumption reduces shipping costs. At Terranera, C. citriodora ssp. for lowering wood density in E. longirostrata should this be citriodora (2.89 m3 t–1), followed by E. longirostrata (Gympie) required. In the case of C. henryi or C. citriodora ssp. citriodora, (3.08 m3 t–1), C. henryi (3.08 m3 t–1) and E. longirostrata (Monto) neither species is known to hybridise with others outside the genus (3.39 m3 t–1) provided the greatest yields (Table 5). The E. grandis Corymbia (Hill and Johnson 1995); breeding for wood properties, × E. urophylla commercial clone provided the lowest yield at and indeed other traits, would probably be on an intra-generic 4.59 m3 t–1. At False Bay, C. citriodora ssp. citriodora basis. Successful controlled crossing of species within the (2.79 m3 t–1) was again best performer, followed by C. henryi Corymbia such as C. citriodora ssp. citriodora, C. citriodora (2.84 m3 t–1), E. longirostrata (Gympie) (3.18 m3 t–1) and ssp. variegata and C. torelliana, with the aim of improving a E. longirostrata (Monto) (3.28 m3 t–1). The E. grandis × range of traits such growth rate; disease, frost and drought E. urophylla commercial clone (4.41 m3 t–1) provided the lowest resistance; and root-strike in cuttings, has been reported by Lee yield of all species and clones at this site. et al. (2005a,b). On the basis of specific consumption across both sites, On the basis of ‘SPY’, at Terranera C. henryi (55.9%), C. citriodora ssp. citriodora, C. henryi and E. longirostrata could C. citriodora ssp. citriodora (54.3%), E. longirostrata (Monto) be rated highly when compared to the two commercial controls, (52.9%), E. pilularis (52.0%) and the E. grandis × E. urophylla the E. grandis × E. urophylla and E. grandis × E. camaldulensis commercial hybrid clone were within the required range of clones. ≥51.5%. At the same site, E. tereticornis produced the lowest SPY of all species and clones (50.3%). At False Bay, only Fibre productivity C. henryi (53.5%) and C. citriodora ssp. citriodora (53.3%) fell within the desired range, with the commercial E. grandis × On the basis of fibre productivity (tonnes screened pulp per urophylla clone producing the lowest SPY of all species and hectare) (Table 6), at Terranera E. longirostrata (Monto) clones (45.2%). Because the Terranera and False Bay sites (65.0 t ha–1) was the highest-yielding unimproved species represent the two extremes of growing conditions in coastal (followed by C. henryi at 57.8 t ha–1), significantly outperforming Zululand, and C. henryi and C. citriodora ssp. citriodora the two lowest-yielding species E. tereticornis (39.1 t ha–1) and performed well at both sites on the basis of SPY, the two species C. citriodora ssp. citriodora (37.1 t ha–1). The E. grandis × appear to be the best adapted of all species evaluated to the range E. urophylla commercial clone was the most productive treatment of forestry environmental conditions in the region. The same at this site at 88.8 t ha–1, significantly outperformimg all other flexibility is unlikely in the case of E. pilularis — pulp-tested at species and clones except for the E. grandis × camaldulensis Terranera only (SPY 52.0%) — as the species showed a high commercial clone (80.2 t ha–1). At False Bay, the unimproved level of drought-sensitivity and poor survival and growth at False C. citriodora ssp. citriodora seedlot was the most productive Bay (Gardner 2001). Regarding the commercial clones, at treatment at 65.3 t ha–1, significantly outperforming all other Terranera both the E. grandis × E. camaldulensis and E. grandis species and clones except for the E. grandis × E. camaldulensis × E. urophylla clone were within the required SPY range (51.4 commercial hybrid clone (59.9 t ha–1) and C. henryi (56.5 t ha–1). and 53.0% respectively), whilst at False Bay neither performed adequately (49.7 and 45.2% respectively). The reversal of ranking The results for fibre productivity demonstrated three important in SPY between the E. grandis × E. camaldulensis and E. grandis points regarding the potential of the ‘new’ species for coastal C. henryi × E. urophylla commercial hybrid clones at the two different Zululand. First, because ranked highly at both sites sites (Table 5) suggests specificity of site requirement of either investigated, the species appears to have a high fibre production clone. In the case of the E. grandis × E. urophylla clone, the potential for the range of sites encountered in coastal Zululand. C. henryi C. citriodora citriodora apparent lack of flexibility of the clone was further supported by Secondly, ’s close relative ssp. the fact that it was top-ranking treatment for merchantable wood appears highly site specific. At the dry False Bay site, this taxon volume at both Terranera and False Bay (Table 3), yet at False excelled on the basis of fibre productivity, whereas at the Bay the SPY and other pulping properties of the clone were significantly wetter Terranera site it was the poorest of all E. longirostrata markedly inferior to those of all other species and clones tested. treatments. Thirdly, (Monto) appears to have the greatest fibre productivity potential of all promising ‘new’ On the basis of ‘pulpability factor’, at the wet site Terranera, species for high-rainfall (traditionally highly productive) sites C. henryi (2.87), E. pilularis (2.50), E. longirostrata (Monto) on the Zululand coastal plain. (2.47) and the E. grandis × urophylla commercial clone (2.49)
Australian Forestry 2007 Vol. 70 No. 1 pp. 37–47 Robin A.W. Gardner, Keith M. Little and Athol Arbuthnot 45
Table 6. Summary of the analyses of variance and data for wood yield and fibre production of selected eucalypt species and clones in trials at 7 y of age at Terranera and False Bay in Zululand, South Africa
Location Wood yield Fibre production (t ha–1) (t ha–1) Terranera False Bay Terranera False Bay Terranera False Bay Analysis Source of variation df Mean squares Replicate 02 02 0236 0408 0061 100 Treatment 07 06 3724** 1184* 1024** 429** Residual 14 12 0486 0289 0140 074
Summary of data Treatment Treatment means ‡ E. longirostrata (Gympie) 104 079 53cd 37c E. longirostrata (Monto) 123 084 65bc 40c E. pilularis 094 0— 49cd — E. tereticornis 078 073 39d 35c C. citriodora ssp. citriodora 068 123 37d 65a C. henryi 103 106 58cd 56ab E. grandis × E. camaldulensis 156 120 80ab 60ab E. grandis × E. urophylla 168 104 89a 47bc Mean 112 098 59 49 sed+ 018 014 10 07 Coefficient of variation (units) 020 017 20 18 * Significant at P < 0.05 ** Significant at P < 0.01 ‡ Within each column, values followed by the same letter are not significantly different (P < 0.05) according to Student’s t-test + Standard error of the treatment differences
Conclusion a potential saw-timber species for high productivity sites in coastal Zululand. Eucalyptus longirostrata, C. henryi and C. citriodora ssp. citriodora have excellent potential for commercial forestry on On the basis of fibre production, E. longirostrata (Monto) and the Zululand coastal plain. C. henryi showed excellent potential for high productivity sites on the coastal plain, whilst C. citriodora ssp. citriodora and For merchantable wood production, E. longirostrata (Monto) C. henryi showed similar potential for marginal (moderately dry) showed excellent potential for high productivity (wet) sites and sites in the same region. C. citriodora ssp. citriodora and C. henryi similar potential for marginal (moderately dry) sites on the coastal plain. The saw- In the site × species interaction trials discussed, the unimproved timber properties of locally-grown E. longirostrata have yet to species were represented by single-provenance bulked-family be accurately determined. If these properties are similar to those seedlots that provided narrow samples of the genetic resources of the closely-related E. punctata and E. propinqua, it is likely of the species (Table 2). Other provenances of the same taxa that timber of the former species will be suitable only for heavy may perform differently on similar sites. The most reliable construction, building and treated poles, and not for high quality inference which can be made from the site × species trial results end-products (Poynton 1979; Brown and Hillis 1984). The results is that species that performed well warrant further, detailed of the Zululand trials tend to confirm earlier reports that timber investigation. Because of the overall promising performances of of E. longirostrata, C. citriodora ssp. citriodora and C. henryi is E. longirostrata and C. henryi in past and present ICFR site × of moderate to high basic density (Poynton 1979; FAO 1980a,b; species trials on the Zululand coastal plain, the ICFR established Lee et al. 2006). The excellent saw-timber properties of ‘forest provenance/progeny trials and breeding seed orchards of these tree’ species within the Corymbia, such as C. citriodora ssp. species in the region during 2001. The trials and seed orchards citriodora, C. maculata and C. henryi are well documented are being managed for rapid first-generation turnover and (Poynton 1979; Brown and Hillis 1984; Dickinson et al. 2004). incorporation of the improved genetic material into commercial Because of their relatively high merchantable timber yields at eucalypt breeding programmes in the region. Based on recent False Bay, C. citriodora ssp. citriodora and C. henryi appear to overseas experience, it would be prudent for South African have good potential for saw-timber production on moderately breeders, at this stage, to consider the establishment of a local dry sites in Zululand. Due to the relatively good performance of base breeding population of C. torelliana, for ultimate use in E. pilularis at Terranera, coupled with the species’ worldwide inter-specific hybrid combination with other species within the reputation for high quality saw-timber production (Poynton 1979; Corymbia such as C. henryi. In Australia, levels of resistance to Brown and Hillis 1984), E. pilularis should not be discounted as quambalaria shoot blight (Quambalaria pitereka) (Simpson 2000)
Australian Forestry 2007 Vol. 70 No. 1 pp. 37–47 46 Productivity of new eucalypts in Zululand and percentage root-strike in stem cuttings in C. citriodora ssp. Dunne, N. (2000). International trends in the timber furniture industry citriodora and C. henryi have been increased substantially via and the implications for South African furniture exporters. CSDS inter-specific hybridisation with C. torelliana (Lee et al. 2005a), Working Paper No. 25. University of Natal, Durban, pp. 14–24. a species possessing far greater Quambalaria resistance and Du Toit, B., Arbuthnot, A., Oscroft, D. and Job, A. (2001) The effects rooting potential than the former two (Dickinson et al. 2004; of remedial fertilizer treatments on growth and pulp properties of Lee et al. 2005b). Currently, Quambalaria pitereka is not Eucalyptus grandis stands established on infertile soils of the problematic to eucalypt plantation forestry in South Africa, Zululand coastal plain. Southern African Forestry Journal 192, 9–18. although recent cases of damage to E. grandis nursery cuttings and field-grown E. nitens seedlings by Quambalaria eucalypti DWAF (2005) Report on Commercial Timber Resources and Primary Roundwood Processing in South Africa 2003/2004. Department have been reported in the same country (FABI 2005). of Water Affairs and Forestry (DWAF), Pretoria, 239 pp. Eldridge, K., Davidson, J., Harwood, C.E. and Van Wyk, G. (1993) Acknowledgements Eucalypt Domestication and Breeding. Clarendon Press, Oxford, 288 pp. We thank Alpheos Ntombela and Denis Oscroft of the ICFR for FABI (2005) FABI Fourth Biennial Report. Forestry and Agricultural their valuable technical assistance regarding the establishment, Biotechnology Institute, Pretoria, pp. 3–8. maintenance and measurement of the trials; the staff at Mondi FAO (1980a) Pulping and Papermaking Characteristics of Fast- Business Paper Richards Bay mill research laboratory for their Growing Plantation Wood Species. Vol. 1. Food and Agriculture contribution regarding the wood pulping analyses; Marius Du Organization of the United Nations, FAO Forestry Paper Series Plessis (Mondi Business Paper) and Dr Colin Smith (ICFR) for No. 19/1, Rome, 396 pp. reviewing an early draft of the manuscript; and Dr Garth Nikles FAO (1980b) Pulping and Papermaking Characteristics of Fast- (QFRI) and an anonymous reviewer for their useful suggestions Growing Plantation Wood Species. Vol. 2. Food and Agriculture for later drafts of the manuscript. Organization of the United Nations, FAO Forestry Paper Series No. 19/2, Rome, 472 pp. References FAO (2005) FAOSTAT Database Collections. http://faostat.fao.org/. Last accessed February 2005. Brown, A.G. and Hillis, W.E. (1984) General introduction. In: Hillis, W.E. and Brown, A.G (eds) Eucalypts for Wood Production. FAO–UNESCO (1974) Soil Map of the World. Vol. 1. UNESCO, Paris. Academic Press, Sydney, pp. 2–5. Flynn, R. (2003) Eucalyptus: Having an Impact on the Global Brown, A.G., Nambiar, E.K.S. and Cossalter, C. (1997) Plantations Solidwood Industry. Wood Resources International, Washington, for the tropics: their role, extent and nature. In: Nambiar, E.K.S. USA. http://www.wri-ltd.com/marketcommentary.html. and Brown, A.G. (eds) Management of Soil, Nutrients and Water Gardner, R.A.W. (2000) Report on 1997 ICFR/CSIRO Collaborative in Tropical Plantation Forests. ACIAR, Canberra, pp. 1–23. Seed Collection Trip in Northern New South Wales, Australia, Chippendale, G.M. (1988) Eucalyptus. In: George, A.S. (ed.) Flora of and Subsequent Deployment of the Seedlots in South Africa. Australia, vol. 19, AGPS, Canberra, pp. 495–509. Institute for Commercial Forestry Research, ICFR Bulletin Series No. 15/2000, Pietermaritzburg, 30 pp. Darrow, W.K. (1983) Provenance-type trials of Eucalyptus camaldulensis and E. tereticornis in South Africa and South-West Africa: Eight- Gardner, R.A.W. (2001) Alternative eucalypt species for Zululand: year results. South African Forestry Journal 124, 13–22. Seven year results from site:species interaction trials in the region. Southern African Forestry Journal 190, 79–88. Darrow, W.K. (1997) Eucalypt Site-Species Trials in Zululand: Results at Four Years of Age. Institute for Commercial Forestry Research, Gardner, R.A.W. (2004) Selection of Eucalyptus species and ICFR Bulletin Series No. 3/1997, Pietermaritzburg, 26 pp. provenances for hot, dry conditions in northern, coastal Zululand in South Africa. In: Borralho, N.M.G., Pereira, J.S., Marques, Denison, N.P. and Kietzka, J.A. (1993a) The use and importance of C., Coutinho, J., Madeira, M. and Tome, M. (eds) Eucalyptus in hybrid intensive forestry in South Africa. South African Forestry a Changing World. Proceedings of IUFRO Conference, 11–15 Journal 165, 55–60. October, Aveiro, Portugal, RAIZ, Aveiro, Portugal, pp. 171–172. Denison, N.P. and Kietzka, J.A. (1993b) The development and Gardner, R.A.W. (2006) Investigating drought tolerant eucalypt species utilisation of vegetative propagation in Mondi for commercial and provenances for northern, coastal Zululand. In: Bender, C.M. afforestation programmes. South African Forestry Journal 166, (ed.) 6th Plant Breeding Symposium of the Southern African Plant 53–60. Breeders Association. Proceedings of SAPBA Symposium, 13–15 Dickinson, G.R., Lee, D.J. and Huth, J.R. (2004) Early plantation March, Langebaan. SAPBA, Bloemfontein, South Africa, p. 43. growth and tolerance to ramularia shoot blight of provenances of Gomez, K.A. and Gomez, A.A. (1984) Statistical Procedures for three spotted gum taxa on a range of sites in Queensland. Agricultural Research. John Wiley and Sons, New York, 680 pp. Australian Forestry 67, 122–130. Gryzenhout, M., Myburg, H., Van der Merwe, N.A., Wingfield, B.D. Downes, G.M., Hudson, I.L., Raymond, C.A., Dean, G.H., Michell, and Wingfield, M.J. (2004) Chrysoporthe, a new genus to A.J., Schimleck, L.R., Evans, R. and Muneri, A. (1997) Sampling accommodate Cryphonectria cubensis. Studies in Mycology 50, Plantation Eucalypts for Wood and Fibre Properties. CSIRO 119–142. Publishing, Collingwood, Australia, pp. 2–5. Harrison, E.R. (1993) The Saligna Story. Kingspress, Empangeni, South Duncan, E.A., van Deventer, F., Kietzka, J.E., Lindley, R.C. and Africa, 31 pp. Denison, N.P. (2000) The applied subtropical Eucalyptus clonal programme in Mondi forests, Zululand coastal region. In: Herbert, M.A. and Musto, J.W. (1993) The Sandy Forestry Soils on the MacLennan, L.A. (ed.) Forest Genetics for the Next Millennium. Zululand Coastal Plain. Institute for Commercial Forestry Proceedings of the IUFRO Symposium, 8–13 October 2000, Research, ICFR Bulletin Series No. 12/1993, Pietermaritzburg, Durban. ICFR, Pietermaritzburg, South Africa, pp. 95–97. 11 pp.
Australian Forestry 2007 Vol. 70 No. 1 pp. 37–47 Robin A.W. Gardner, Keith M. Little and Athol Arbuthnot 47
Herbert, M.A. and Robertson, M.A. (1991). Above-ground biomass Pierce, B.T. and Verryn, S.D. (2000) Five year results from a site by composition and nutrient content for Eucalyptus species in the clone interaction series of Eucalyptus grandis in the summer southeastern Transvaal. In: Schönau, A.P.G. (ed.) Intensive rainfall areas of South Africa. In: MacLennan, L.A. (ed.) Forest Forestry: The role of Eucalypts. Proceedings of the IUFRO Genetics for the Next Millennium. Proceedings of the IUFRO Conference, 2–6 September, Durban. SAIF, Pretoria, South Africa, Conference, 8–13 October, Durban. ICFR, Pietermaritzburg, pp. 662–674. South Africa, pp. 186–191. Hill, K.D. and Johnson, L.A.S. (1995) Systematic studies in the Potts, B.M. and Dungey, H.S. (2004) Interspecific hybridisation of eucalypts.7. A revision of the bloodwoods, genus Corymbia Eucalyptus: key issues for breeders and geneticists. New Forests (Myrtaceae) Telopea 6, 185–504. 27, 115–138. Kimmins, J.P. (1994) Identifying key processes affecting long-term site Poynton, R.J. (1979) Tree Planting in Southern Africa. Vol. 2. The productivity. In: Dyck, W.J., Cole, D.W. and Comerford, N.B. Eucalypts. DEA, Government Printer, Pretoria, 882 pp. (eds) Impacts of Forest Harvesting on Long-Term Site Retief, R.J., Male, J.R., Malan, F.S., Dyer, S.T., Conradie, D., Turner, Productivity. Chapman and Hall, London, pp. 119–150. P., Havenga, A. and Gama, D. (1997) First Report on the Effect Lane, P.W. and Payne, R.W. (1996) Genstat® for Windows™, An of Site Quality on the Wood and Pulp Properties of Eucalyptus Introductory Course. 2nd edn. Lawes Agricultural Trust, grandis Clonal Material. ENV/P/I 97045. Division of Water, Rothamsted Experimental Station. Environment and Forestry Technology, CSIR, Pretoria, 11 pp. Lee, D.J., Nikles, D.G., Pomroy, P.C., Brawner, J., Wallace H. and Schulze, R.E. (1997) South African Atlas of Agrohydrology and Stokoe, R. (2005a) Corymbia species and hybrids: a solution to Climatology. Report TT 82/96. Water Research Commission, Queensland hardwood plantations? In: Lee, D.J. (ed.) Corymbia Pretoria, South Africa, 276 pp. Research Meeting: Underpinning Development of a Profitable Schumacher, F.X. and Hall, F.D.S. (1933) Logarithmic expression of Hardwood Plantation Industry in Northern Australia by Research timber-tree volume. Journal of Agricultural Research 47, into Corymbia Species and Hybrids. Proceedings of the Inaugural 719–734. Meeting of Corymbia Researchers and the Plantation Industry, Simpson, J.A. (2000) Quambalaria, a new genus of eucalypt pathogens. 1–2 June, Gympie. Department of Primary Industries and Australasian Mycologist 19(2), 57–62. Fisheries, Brisbane, Australia, pp. 5–7. Smith, C.W., Little, K.M. and Norris, C.H. (2001) The effect of land Lee, D.J., DeBuse, V.J., Pomroy, P.C., Robson, K.J. and Nikles, D.G. preparation at re-establishment on the productivity of fast growing (2005b) Developing Genetically Adapted Tree Varieties for hardwoods in South Africa. Australian Forestry 64, 165–174. Marginal Areas of Northern Australia. Rural Industries Research and Development Corporation Publication No. 04/186. Kingston, Snedecor, G.W. and Cochran, W.G. (1980) Statistical Methods. 7th ACT, Australia, pp. 9–53. edn. The Iowa State College Press, Iowa, 507 pp. Lee, D.J., Lawson, S., Armstrong, M. and Dickinson, G.R. (2006) Stanger, T. (2004) Evaluation of the dissolving pulp properties of fifteen Hardwoods Advice: Plantation Species Profiles: Spotted Gums. sub tropical hybrid Eucalyptus clones. In: Borralho, N.M.G., Department of Primary Industries and Fisheries, Queensland Pereira, J.S., Marques, C., Coutinho, J., Madeira, M. and Tome, Government, Australia. M. (eds) Eucalyptus in a Changing World. Proceedings of the http://www2.dpi.qld.gov.au/hardwoodsqld/10142.html. IUFRO Conference, 11–15 October, Aveiro. RAIZ, Aveiro, Portugal, pp. 705–706. Little, K.M., Schumann, A.W. and Noble, A.D. (2002) Performance of a Eucalyptus grandis × E. camaldulensis hybrid clone as Steel, R.G.D. and Torrie, J.H. (1981) Principles and Procedures of influenced by a cowpea cover-crop. Forest Ecology and Statistics: A Biometrical Approach. 2nd edn. McGraw-Hill Management 168, 43–52. International Book Company, Singapore, 633 pp. Malan, F.S. (1988) Genetic variation in some growth and wood Verryn, S.D. (2000) Eucalypt hybrid breeding in South Africa. In: properties among 18 full-sib families of South African grown Dungey, H.S., Dieters, M.J. and Nickles, D.G. (eds) Hybrid Eucalyptus grandis: a preliminary investigation. South African Breeding and Genetics of Forest Trees. Proceedings of the QFRI Forestry Journal 146, 38–43. CRC – SPF Symposium, 9–14 April, Noosa, Queensland. Brisbane. Department of Primary Industries, Queensland, McDonald, M.W. and Bean, A.R. (2000) A new combination in pp. 435–439. Corymbia ‘section Politaria’: C. citriodora ssp. variegata (Myrtaceae). Austrobaileya 5, 735–736. Zobel, B. and Talbert, J. (1984) Applied Forest Tree Improvement. John Wiley & Sons, Inc., USA, 505 pp. Miranda, I. and Pereria, H. (2002) Variation of pulpwood quality with provenances and site in Eucalyptus globulus. Annals of Forest Science 59, 283–291. Nicol, N. (1992) Plantation Pathology: Notes for Identification of Common Pathological Problems in South African Forestry. Institute for Commercial Forestry Research, ICFR Bulletin Series No. 23/1992, Pietermaritzburg, 15 pp.
Australian Forestry 2007 Vol. 70 No. 1 pp. 37–47 48 Genetic variation in Eucalyptus globulus in Ethopia
Genetic variation in survival and growth of Eucalyptus globulus ssp. globulus in Ethiopia
Tesfaye Hunde1, Belachew Gizachew1 and Chris Harwood2,3,4
1Forestry Research Centre, PO Box 308, Addis Ababa, Ethiopia 2Ensis Genetics, Private Bag 12, Hobart, Tasmania 7001, Australia 3Cooperative Research Centre for Forestry, Private Bag 12, Hobart, Tasmania 7001, Australia 4Email: [email protected]
Revised manuscript received 4 December 2006
Summary recognised, with extensive intergrade zones between them. Eucalyptus globulus ssp. globulus is the most widely planted of A provenance/progeny trial of Eucalyptus globulus ssp. globulus the four subspecies (Eldridge et al. 1993). In this paper we follow was established at Ilalaa Gojo in the central highlands of Ethiopia Brooker and Kleinig (2006) in preferring to use the name E. in 1990 to identify superior provenances and establish a breeding globulus rather than E. globulus ssp. globulus. population for genetic improvement programs. The trial tested a total of 299 open-pollinated families from 52 localities within Eucalyptus globulus was one of the first eucalypt species 17 natural sub-races in Australia, and one local landrace from introduced to Ethiopia and is currently the most popular, and is Ethiopia. Nine years after planting, more than 61% of trees had now perceived as ‘indigenous’ among local communities. Com- survived and the overall mean height and diameter at breast height mencing in 1895, plantations were established to supply poles (dbh) of surviving trees were 21.2 m and 16.4 cm respectively. and fuelwood to the city of Addis Ababa and other major towns Significant (P < 0.001) differences in survival, height and dbh in the Central Highlands (Davidson 1995). Community woodlots per tree were found between sub-races; differences among local- and plantations are managed by groups of farmers. The total area ities within the sub-races were also significant. Eastern and of plantations of E. globulus exceeds 100 000 ha. Most of the Western Otways, Cape Patton and Strzelecki Ranges in Victoria plantations appear to be E. globulus ssp. globulus, on the basis were the fastest-growing sub-races, and also displayed above- that their inflorescences are single-fruited (Kirkpatrick 1975). average survival. Sub-races from the Tasmanian mainland Results of previous species elimination trials in Ethiopia have performed poorly; the Wilsons Promontory sub-race was the confirmed E. globulus to be the best-performing species in the slowest-growing and displayed poor survival. The Ethiopian central highlands, with mean annual productivity of plantations landrace performed well although it was not adequately 10–30 m3 ha–1 (Pohjonen and Pukkala 1990). The ever-increasing represented in the trial. Assuming a coefficient of relationship of demand for fuelwood and construction wood in Ethiopia, and 0.4 for open-pollinated families, individual-tree within-sub-race the continually declining wood supply from natural forests, heritabilities for height and dbh were 0.20 ± 0.03 and 0.16 ± 0.02 necessitate wood production from fast-growing plantation respectively. We conclude that there is substantial potential for species. Notwithstanding an ongoing public argument for and genetic improvement of E. globulus in Ethiopia. against the use of eucalypts as plantation species, E. globulus plantations are attractive from an economic viewpoint (Pohjonen Keywords: provenance trials; growth; survival; heritability; Eucalyptus and Pukkala 1988). globulus; Ethiopia Despite the importance of the species to Ethiopia, information Introduction on the genetic status of the Ethiopian populations is scant. Although the first introductions of E. globulus to Ethiopia were Eucalyptus globulus Labill., also commonly known as blue gum, successful, the provenance origins and genetic base of the is a forest tree occurring naturally in Tasmania, Victoria and New introductions are unknown and probably sub-optimal, and there South Wales (NSW), Australia (Brooker and Kleinig 2006). It could be a high degree of inbreeding in the land races that have was one of the first eucalypt species to be both validly named developed (Ethiopian Forestry Action Plan 1994; Davidson 1995). and brought into cultivation within Australia and overseas. The species is planted widely in southern Australia, Portugal, Spain, A tree improvement program for E. globulus in Ethiopia was Chile and the equatorial highlands of Africa and Latin America initiated by FAO in 1990 in collaboration with the Ethiopian for a range of uses including firewood, pulpwood and timber Forestry Research Centre. A provenance/progeny trial was (Eldridge et al. 1993). Eucalyptus globulus was amalgamated established in the central highlands, testing progenies from an with three closely related species by Kirkpatrick (1975) and the extensive range of provenances from across the natural range of sub-species globulus, maidenii, bicostata and pseudoglobulus E. globulus in Tasmania and Victoria. The objective of the
Australian Forestry 2007 Vol. 70 No. 1 pp. 48–52 Tesfaye Hunde, Belachew Gizachew and Chris Harwood 49 program was to establish a base population from which a breeding differences among localities within sub-races, by incorporating population could be selected for genetic improvement programs. localities within sub-races as an additional fixed effect in the This paper describes genetic variation in survival and growth, model and declaring open-pollinated families within localities 9 y after planting. rather than within sub-races as a random effect.
Narrow-sense heritabilities for height and dbh and their standard Materials and methods errors were estimated from the individual-tree data using ASReml The Australian seedlots used in the trial were from the 1987 and (Gilmour et al. 2002, Chapter 11), using a linear mixed model 1988 CSIRO collections (Gardiner and Crawford 1988). A total with replicates and sub-races as fixed effects and incomplete of 299 open-pollinated families from 52 CSIRO local provenance blocks (rows and columns within replicates), plots and open- seed collections (hereafter termed localities) were raised for pollinated families within sub-races as random effects. Herit- ability was computed as h2 = 1/r × σ2 /σ2, where h2 denotes planting in the trial. The localities are described in Appendix 1. f p Localities have been grouped into 17 sub-races according to the individual-tree heritability, r denotes coefficient of relationship, σ2 denotes variance of families within sub-races and σ2 denotes racial classification of E. globulus made by Dutkowski and Potts f p (1999). A bulk seed collection from a local Ethiopian landrace phenotypic variance, which was the sum of the variance was also included as a single treatment in the trial, giving a total components for open-pollinated families within sub-races, plots of 300 experimental treatments. and residual variance. To take account of selfing and relatedness within open-pollinated families, known to occur in E. globulus, The trial was established at Ilalaa Gojo, in the Ethiopian central a coefficient of relationship of 0.4 was assumed in the computation highlands (latitude 38°30'E, longitude 9°30'N, altitude 2400 m). of heritabilities (Williams et al. 2002; Potts et al. 2004). The site is located in the tepid cool humid plateau sub-agro- Heritabilities were also calculated on a within-locality basis using ecological zone (Ministry of Agriculture 1998) and is represent- a model in which locality was declared as a fixed effect. ative of the major E. globulus plantation sites in Ethiopia. The area receives a mean annual rainfall of 2400 mm and has a mean Results annual temperature of 14°C. Of the total annual rainfall, 70% occurs in the long rainy season between June and September, Overall growth was excellent, with a mean height of 21.2 m and and there is a second short rainy season in March–April. The mean dbh of 16.4 cm attained at age 9 y (Table 1). trial site is situated on relatively flat terrain between mountain slopes and gently ascending plateaux. The soil at the trial site is The fraction of living trees included in the assessment was 61%. reddish-brown in colour, of light clay texture in the upper horizons Overall survival was slightly higher than this, because a few of and heavy clay in the lower horizons, with pH 5–6. Nearby the plots were not fully stocked at the time of planting. The sub- vegetation not under agricultural cropping consists mainly of races differed significantly (P < 0.001) for survival, height and remnant natural forest, herbaceous flora and eucalypt plantations. dbh. The fastest-growing sub-races were Western and Eastern Otways, Cape Patton and Strzelecki Ranges from Victoria, all of The experimental design employed for the 300 treatments was which had heights of > 22 m and dbh >17 cm. These sub-races an incomplete block with ten replicates and two-dimensional also displayed above-average survival, except for Cape Patton incomplete blocking provided by 15 rows and 20 columns within which was slightly below average at 59%. The Strzelecki Ranges replicates. Each entry was represented by a square plot of four sub-race displayed the highest survival at 79%. The worst- trees in each replicate. Initial spacing was 3 m × 3 m. Replicates performing sub-race was Wilson’s Promontory, with height of were separated from one another by 6 m. Seedlings were raised 15.7 m, dbh of 11.5 cm and survival of 37%. Sub-races from in the nursery in plastic bags for about 5 months before planting southern and eastern Tasmania also performed poorly, all being out in July 1990. Prior to out-planting the site was fully cultivated well below the trial mean in survival, height and/or dbh. The using manual labour. Weed control was carried out manually as Ethiopian land race performed relatively well, with height and necessary. No fertiliser was applied. dbh close to those of the Otways and Strzelecki Ranges sub- races, but its performance was estimated from only 21 surviving Diameter at breast height over bark (dbh) was measured to the trees. Localities within sub-races also differed significantly in nearest 0.1 cm and total height was measured to the nearest 0.25 m their performance (P < 0.001 for survival, P = 0.01 for height for all surviving trees at 9 y of age, using diameter tapes and and P < 0.01 for dbh). height poles respectively. Estimated individual-tree within-sub-race heritabilities for height Data screening was carried out by scrutinising plot means and and dbh were 0.20 ± 0.03 and 0.16 ± 0.02. Individual-tree within- variances (Williams et al. 2002). Six trees < 5 m in height or locality heritabilities were lower, at 0.17 ± 0.03 and 0.13 ± 0.02 < 5 cm in dbh were omitted from the data set. The plot means for respectively. survival, height and dbh were analysed using the software package Genstat (Genstat 2003) to obtain sub-races means and the average standard errors of differences between sub-races. A linear mixed Discussion model was used, with replicates and sub-races as fixed effects, The trial had not been thinned up to the age of assessment at 9 y. and incomplete blocks (rows and columns within replicates) and Volume estimates (not presented here) indicate that the best- open-pollinated families within sub-races as random effects. The performing sub-races had a mean annual increment of 29 m3 ha–1, significance of differences between sub-races was tested using which is close to the maximum observed in Ethiopia (Pojonen Wald tests. A further analysis examined the significance of
Australian Forestry 2007 Vol. 70 No. 1 pp. 48–52 50 Genetic variation in Eucalyptus globulus in Ethopia
Table 1. Mean height, diameter and survival of Eucalyptus globulus sub-races at age 9 y at Ilalaa Gojo, Ethiopia
Height Diameter Survival Sub-race (no. of families tested) State (m) (dbh, cm) (%) Western Otways (52) Victoria 22.5 17.6 66.8 Strzelecki Ranges (32) Victoria 22.3 17.7 79.1 South Gippsland (5) Victoria 19.0 14.7 22.0 Wilsons Promontory (7) Victoria 15.7 11.5 37.2 Cape Patton (17) Victoria 22.2 17.5 58.8 Eastern Otways (23) Victoria 22.3 18.2 71.3 Western Tasmania 19) Tasmania 20.5 14.7 60.1 St Helens, Tasmania (2) Tasmania 19.6 14.1 76.8 Southern Furneaux (35) Tasmania 21.3 16.6 66.5 King Island (10) Tasmania 21.5 16.0 52.1 Flinders Island (37) Tasmania 20.9 16.0 67.2 SE Tasmania (26) Tasmania 19.8 14.6 45.5 Southern Tasmania (15) Tasmania 21.4 16.0 42.1 Dromedary (4) Tasmania 19.9 15.1 44.8 NE Tasmania (5) Tasmania 19.6 13.9 30.9 Inland NE Tasmania (5) Tasmania 19.6 14.5 42.1 Tasman Peninsula (5) Tasmania 20.3 15.7 35.6 Ethiopian land race (bulk) 22.1 18.0 52.5 Overall mean 21.2 16.4 60.8 Average standard error of difference of means 0.86 0.87 11.1 Significance of differences between sub-races P < 0.001 P < 0.001 P < 0.001 Significance of differences between localities within sub-races P = 0.01 P < 0.01 P < 0.001
and Pukkala 1990). Intense competition between adjacent plots The study reported here also demonstrated significant differences would have accentuated differences between fast-growing and among localities within sub-races in height, dbh and survival. slow-growing families, through suppression of the slower- This is to be expected, as the local collections within sub-races growing families. Support for this interpretation is provided by are separated by considerable geographic differences in many the much poorer survival of the slowest-growing sub-race, cases, although a few of the localities were repeat collections Wilsons Promontory. Eucalyptus globulus is known to have from virtually the same site (Appendix 1). In Tasmanian trials, variable but generally high levels of selfing and neighbourhood Jordan et al. (1994) similarly reported significant variation in inbreeding in natural populations (Hardner et al. 1996), and these growth among collection localities within twelve geographic races inbred individuals typically become suppressed when growing of E. globulus subsp. globulus. The variation in locality perform- under competition (Hardner and Potts 1995). Poorly performing ance accounts for some of the family-within-sub-race variance, individual trees within plots would thus be likely to die following leading to the reduction in individual-tree heritabilities for height suppression, contributing to the mortality observed in the trial. and dbh when these are calculated on a within-locality basis. Additional complicating factors were that the trial was not surrounded by external buffer rows and that replicates were The observed major differences among sub-races and localities separated from one another by 6 m, creating edge effects. The and the significant heritabilities for height and dbh within incomplete block design used enables adjustment for edge effects; localities in the present trial support the findings of other row incomplete-block variance components were more than twice researchers that there is substantial variation between provenances the magnitude of their standard errors for survival, height and dbh. of E. globulus in survival and growth parameters, and that there is an opportunity for genetic gain by selection and breeding. The differences between sub-races measured at age 9 y are Estimated within-sub-race, individual-tree heritabilities (0.20 for probably somewhat exaggerated by competition effects. However, height and 0.16 for dbh) are within the range found in open- the four best-performing sub-races (Western Otways, Eastern pollinated progeny trials of E. globulus (Eldridge et al. 1993; Otways, Strzelecki Ranges and Cape Patton) are well represented Potts et al. 2004). However, differences in family performance in the trial, together contributing 124 out of 300 progenies, so arising from variable levels of inbreeding tend to inflate these their superiority has been clearly established. The ranking of sub- estimates, even when a coefficient of relationship of 0.4, rather races for growth is in agreement with results from trials in north- than 0.25 (appropriate for half-sib families), is used (Hodge et western Tasmania of the same CSIRO seed collections (Jordan al. 1996). et al. 1994; Dutkowski and Potts 1999) in which Strzelecki Ranges and Otways were outstanding performers. In earlier trials The trial did not effectively resolve the performance of the in Victoria, Australia, E. globulus provenances from the Otway Ethiopian landraces of E. globulus relative to natural provenances Ranges also showed superior growth (Eldridge et al. 1993). of the species, as it tested only a single bulk from the Addis
Australian Forestry 2007 Vol. 70 No. 1 pp. 48–52 Tesfaye Hunde, Belachew Gizachew and Chris Harwood 51
Ababa region, represented by only 21 surviving individual trees. Genstat (2003) GenStat for Windows. Release 7.1. Sixth edition. VSN It is possible that the genetic status of the landrace varies International, Oxford. considerably in different parts of the country, as seed collection Gilmour, A.R., Gogel, B.J., Cullis, B.R., Welham, S.J. and Thompson, and seed distribution is not carried out in any systematic way. R. (2002) ASReml User Guide. Release 1.0. VSN International, Oxford. The genetic resources established in this trial have broadened Hardner, C.M. and Potts, B.M. (1995) Inbreeding depression and the genetic base of E. globulus ssp. globulus in Ethiopia available changes in variation after selfing in Eucalyptus globulus ssp. for use in breeding programs. The information on sub-race globulus. Silvae Genetica 44, 46–54. rankings presented here was used by the Ethiopian Forestry Hardner, C.M., Vaillancourt, R.E. and Potts, B.M. (1996) Stand density Research Centre to establish new seed production areas based influences outcrossing rate and growth of open-pollinated families on the superior Otways and Strzelecki Ranges sub-races. of Eucalyptus globulus. Silvae Genetica 45, 226–228. Hodge, G.R., Volker, P.W., Potts, B.M. and Owen, J.V. (1996) A Acknowledgements comparison of genetic information from open-pollinated and control-pollinated progeny tests in two eucalypt species. We thank the research staff of the Forestry Research Centre for Theoretical and Applied Genetics 92, 53–63. establishing and maintaining the trial, which was established with Jordan, G.J., Borralho, N.M.G., Tilyard, P. and Potts, B.M. (1994) support from the United Nations Food and Agriculture Identification of races in Eucalyptus globulus ssp. globulus based Organisation, under the guidance of John Davidson. The on growth traits in Tasmania and geographic distribution. Silvae Genetica 43, 292–298. Australian Centre for International Agricultural Research provided a training attachment for Mr Gizachew to visit Australia Kirkpatrick, J.B. (1975) Geographic Variation in Eucalyptus globulus. to carry out statistical analysis of trial data. Colin Matheson and Bulletin 47. Forestry and Timber Bureau, Canberra. David Bush are thanked for their review of the manuscript. Ministry of Agriculture (1998) Agro-ecological Zones of Ethiopia. Natural Resources Management and Regulatory Department, Addis Ababa. References Pohjonen, V. and Pukkala, T. (1988) Profitability of establishing Brooker, M.I.H. and Kleinig, D.A. (2006) Field Guide to Eucalypts. Eucalyptus globulus plantations in the Central Highlands of Vol. 1 South-eastern Australia. Third edition. Bloomings Books, Ethiopia. Silva Fennica 22, 307–321. Melbourne. Pohjonen, V. and Pukkala, T. (1990) Eucalyptus globulus in Ethiopian Davidson, J. (1995) Eucalyptus Tree Improvement and Breeding. forestry. Forest Ecology and Management 36, 19–31. Ministry of Natural Resources, Development and Environmental Potts, B.M., Vaillancourt, R.E., Jordan, G.J., Dutkoski, G., da Costa e Protection. Forestry Research Centre, Addis Ababa, Ethiopia, Silva, J., McKinnon, G., Steane, D., Volker, P., Lopez, G., 96 pp. Apiolaza, L., Li, Yongjun, Marques, C. and Borallho, N. (2004) Dutkowski, G.W. and Potts, B.M. (1999) Geographic patterns of genetic Exploration of the Eucalyptus globulus gene pool. In: Borallho, N., variation in Eucalyptus globulus ssp. globulus and a revised racial Pereira, J.S., Marwues, C., Coultinho, J., Madeira, M. and Tomé, M. classification. Australian Journal of Botany 47, 237–263. (eds) Eucalypts for a Changing World. Proceedings of IUFRO Eldridge, K., Davidson, J., Harwood, C. and van Wyk, G. (1993) conference, Aveiro, Portugal, 11–15 October 2004. RAIZ, Eucalypt Domestication and Breeding. Clarendon Press, Oxford. Instituto Investigação da Floresta e papel, Portugal, pp. 46–51. Ethiopian Forestry Action Plan. (1994) The Challenge for Development. Williams, E.R., Matheson, A.C. and Harwood, C.E. (2002) Final Report. Vol. II. Ministry of Natural Resources and Experimental Design and Analysis for Tree Improvement. Second Environmental Protection, Addis Ababa, 95 pp. edition. CSIRO Publishing, Melbourne. Gardiner, C.A. and Crawford, D.F. (1988) 1988 seed collections of Eucalyptus globulus subsp. globulus Labill. for tree improvement purposes. Unpublished report, Australian Tree Seed Centre, CSIRO, PO Box E4008, Kingston, ACT 2604, Canberra.
Australian Forestry 2007 Vol. 70 No. 1 pp. 48–52 52 Genetic variation in Eucalyptus globulus in Ethopia
Appendix 1. Eucalyptus globulus localities tested in the trial at Ilalaa Gojo, Ethiopia, grouped according to the sub-race classification of Dutkowski and Potts (1999)
CSIRO Sub-race No. of Latitude Longitude Altitude Locationa Stateb Sub-race namea seedlot no. families (S) (E) (m) 16223 19 km SW Apollo Bay PO Vic W Otways 1 03 38°77' 143°53' 200 16224 21.6 km SW Apollo Bay Vic W Otways 1 10 38°82' 143°57' 145 16225 17 km SW Apollo Bay PO Vic W Otways 1 02 38°78' 143°58' 200 16226 8.0 km SW Apollo Bay Vic W Otways 1 12 38°80' 143°62' 130 16227 9.5 km SW Apollo Bay Vic W Otways 1 02 38°78' 143°62' 150 16240 Otway State Forest Vic W Otways 1 18 38°75' 143°45' 150 16241 SW of Lavers Hill Vic W Otways 1 05 38°73' 143°30' 230 16319 Jeeralang North Vic Strzelecki Ra. 2 32 38°32' 146°55' 220 16398 1.5 km NW of Hedley Vic S Gippsland 3 04 38°63' 146°50' 020 16399 Wilsons Promontory Vic Wilsons Prom. 4 07 39°13' 146°42' 060 16400 8.5 km N of Toora PO Vic S Gippsland 3 01 38°62' 146°35' 180 16401 10.2 km W Kennett R. Vic Cape Patton 5 06 38°67' 143°80' 300 16402 5.4 km W Kennett R. Vic Cape Patton 5 06 38°65' 143°80' 250 16403 0.6 km W Kennett R. Vic Cape Patton 5 05 38°67' 143°83' 130 16405 12.1 km S Lorne PO Vic E Otways 6 07 38°60' 143°90' 200 16406 2.4 km NW Lorne PO Vic E Otways 6 05 38°52' 143°95' 210 16407 17.1 km W Lorne PO Vic E Otways 6 09 38°53' 143°93' 210 16408 6.5 km N Lorne PO Vic E Otways 6 01 38°50' 144°02' 060 16409 W of Lorne Vic E Otways 6 01 38°55' 143°97' 100 16410 Badgers Ck Quarry Rd Tas W Tasmania 7 07 41°98' 145°30' 120 16411 Near Binalong Bay Tas St Helens 8 02 41°27' 148°30' 120 16412 Little Henty River Tas W Tasmania 7 08 41°93' 145°20' 010 16413 W Cape Barren Is. Tas S Furneaux 9 02 40°40' 148°00' 060 16414 SW Cape Barren Is. Tas S Furneaux 9 01 40°45' 148°10' 030 16415 Clarke Is. Tas S Furneaux 9 03 40°53' 148°13' 040 16416 NE Cape Barren Is. Tas S Furneaux 9 04 40°32' 148°32' 060 16417 N Cape Barren Is. Tas S Furneaux 9 05 40°37' 148°22' 020 16418 W Cape Barren Is. Tas S Furneaux 9 02 40°40' 148°05' 220 16419 NW Cape Barren Is. Tas S Furneaux 9 08 40°35' 148°12' 020 16420 NW Cape Barren Is. Tas S Furneaux 9 05 40°37' 148°08' 060 16421 SW Cape Barren Is. Tas S Furneaux 9 05 40°43' 148°05' 040 16422 Maquarie Harbour Tas W Tasmania 7 04 42°33' 145°33' 020 16424 King Is. Tas King Is. 10 10 40°00' 144°00' 060 16425 S Flinders Is. Tas Flinders Is. 11 04 40°23' 148°13' 120 16426 NW Flinders Is. Tas Flinders Is. 11 02 39°77' 147°87' 020 16427 N Flinders Is. Tas Flinders Is. 11 04 39°75' 147°95' 040 16428 W Flinders Is. Tas Flinders Is. 11 02 39°85' 147°83' 020 16429 Central Flinders Is. Tas Flinders Is. 11 07 39°92' 147°95' 040 16430 Central Flinders Is. Tas Flinders Is. 11 01 39°92' 148°03' 020 16431 Central Flinders Is. Tas Flinders Is. 11 09 40°03' 148°02' 190 16432 E Flinders Is. Tas Flinders Is. 11 01 39°98' 148°18' 060 16433 Cent Flinders Is. Tas Flinders Is. 11 03 40°07' 148°07' 150 16434 S Flinders Is. Tas Flinders Is. 11 04 40°27' 148°17' 012 16470 Moogara Tas SE Tasmania 12 21 42°78' 146°92' 500 16471 NW of Dover Tas S Tasmania 13 05 43°27' 146°98' 190 16472 Ellendale Tas SE Tasmania 12 05 42°63' 146°70' 460 16473 NE of New Norfolk Tas Dromedary 14 04 42°72' 147°15' 300 16474 N of St Marys Tas NE Tasmania 15 05 41°57' 148°20' 400 16475 SW of Jericho Tas Inland NE Tasmania 16 05 42°42' 147°27' 500 16476 S of Geeveston Tas S Tasmania 13 07 43°20' 146°90' 250 16477 N of Geeveston Tas S Tasmania 13 03 43°13' 146°95' 200 16478 Koonya Tasman Pen. Tas Tasman Pen. 17 05 43°07' 147°83' 020 Local Addis Ababa Ethiopia Bulk 9°03' (N) 38°42' aCk = Creek; Is. = Island; Pen. = Peninsula; PO = Post office; R. = River; Ra. = Ranges bVic = Victoria; Tas = Tasmania
Australian Forestry 2007 Vol. 70 No. 1 pp. 48–52 N.G. Collett and J. McBeath 53
Managing insect pests in Eucalyptus globulus (Labill.) plantations in Victoria using insecticide tablets at establishment
N.G. Collett1,2 and J. McBeath3
1School of Forest and Ecosystem Science, The University of Melbourne, 123 Brown Street, Heidelberg, Victoria 3084, Australia 2Email: [email protected] 3Bayer CropScience Pty Ltd, 391–393 Tooronga Road, East Hawthorn, Victoria 3123, Australia
Revised manuscript received 8 January 2007
Summary In Australia, hardwood plantations are dominated by Eucalyptus globulus Labill. (blue gum) which accounts for 21% of the total We studied the effects of fertiliser and insecticide delivered by plantation area and 60% of the total hardwood plantings, while tablets on defoliation and growth of Eucalyptus globulus in in Victoria these figures are about 30% and 70% respectively plantations near Meredith in west-central Victoria to determine (Commonwealth of Australia 2005). Thus E. globulus forms a whether improved foliage retention and growth were conferred significant part of the hardwood plantation complex both across by the insecticide and/or fertiliser components of the tablet. Australia and within Victoria (United States Department of Agriculture 2003; National Forest Inventory 2004; Common- Trees at the study site were variably defoliated by three species wealth of Australia 2005). Most eucalypt plantations throughout of insect from March 2004 to September 2005. In the first year Australia (and Victoria) are in private ownership and grown on of the study, treatments containing imidacloprid insecticide had short rotations (10–25 y) to provide a source of wood fibre for significantly less defoliation in both the upper and lower crowns producing pulp and paper (United States Department of than did treatments containing no insecticide. Treatments Agriculture 2003). About 80% of hardwood chips are exported, containing both insecticide and fertiliser showed increased height with Japan taking over 90% of these (Australian Bureau of growth, most likely as a result of a combination of increased levels Agriculture and Resource Economics 2000). of available nutrients and protection from defoliating insects provided by the tablets leading to increased overall vigour. The A major threat to the productivity and economic viability of tablets appear to provide adequate protection to trees for about eucalypt plantations, however, is the risk of severe to total 1 y after planting, although some protection seems to have defoliation by insect pests during the first 4 y after planting. The remained until the 1.7 y assessment. Further observation is pests involved may be species of leaf-feeding moth (Order required to determine how long this protection lasts. Lepidoptera), beetles (Coleoptera), sawfly (Hymenoptera), locust (Orthoptera) or sap-sucking bugs, including psyllids and coreids The implications of the study are significant for eucalypt planta- (Hemiptera) acting either singly or in combination (Stone et al. tion managers. The application of insecticide and fertiliser in 1998; Collett and Neumann 2002). The causes of sudden insect tablets at the time of planting potentially provides a more effective pest outbreaks in young eucalypt plantings are still not fully and environmentally safer method of insect control than traditional understood, although several factors, acting singly or in combina- broadscale insecticide applications during the establishment tion, may be involved. These include: phase. This could potentially lead to improved growth, yield and general health in the longer term. 1. prolonged periods of warm, dry weather accelerating insect life cycles and enhancing survival, and as a consequence Keywords: plantations; defoliation; growth; insecticides; fertilizers; increasing population levels application methods; imidacloprid; Eucalyptus globulus 2. susceptibility of the eucalypt species or provenance planted to attack Introduction 3. a generally greater susceptibility of monocultures of low genetic diversity to attack Australia had 1.72 million ha of plantations in 2004, consisting of about 716 000 and 1 001 000 ha of hardwood and softwood 4. the absence of effective biocontrol agents species respectively. The hardwood estate has expanded rapidly, 5. slow initial growth delaying canopy closure and preventing the a 68% increase in area being recorded from 1995 to 2004. New rapid development of tougher, less susceptible mature foliage. hardwood plantings averaged of 74 000 ha annually from 2000 to 2004. The aim of the Australian government is to expand the These factors, either singly or in combination, can lead to the plantation estate to over 3 million ha by 2020 (National Forest onset of economically damaging outbreaks (Neumann 1993). Inventory 2004; Commonwealth of Australia 2005). Studies of the effects of defoliation on E. globulus growth have
Australian Forestry 2007 Vol. 70 No. 1 pp. 53–60 54 Managing insect pests using insecticide tablets found that single and repeated defoliation events, especially 144°1'08.31''E), or about 40 km south-east of Ballarat and 80 during autumn, adversely affect height, basal area and volume km west of Melbourne. The trial was established on an ex-pasture growth (Collett and Neumann 2002). Stone et al. (1998) found site used for sheep grazing up to early 2003, after when 46 ha that the application of pesticides alleviated insect defoliation, were planted with Eucalyptus globulus seedlings in four blocks. enabling the average height and general growth vigour of the The largest of the four blocks was about 28 ha; the trial was Symphyomyrtus sub-genera (including E. globulus) to established over an area of 1.2 ha in the north-eastern corner of significantly exceed that of the Monocalyptus sub-genera. Elliott this block at an elevation of 280–300 m on a ridge slope with a et al. (1993) found that 1-y-old E. regnans trees repeatedly westerly aspect. The area around the study site was mainly cleared damaged by chrysomelid leaf beetles lost about 45% and 52% of pasture used for sheep and cattle grazing. Along the northern their potential height and basal area increment respectively, while edge of the study site shelterbelts had been planted using a variety both Candy et al. (1992) and Wills et al. (2004) observed, in of native eucalypt and acacia species, while to the north-east of defoliation trials running for more than 10 y, that repeated the trial site there was a small E. globulus plantation about 6 y defoliation episodes significantly reduced growth and overall old. Native vegetation in the area consisted mainly of sparse economic viability of plantations. mixed-species forests in small stands and remnant roadside areas, dominated by eucalypts such as E. camaldulensis (Dehnh.) (river Timely control measures may be required to minimise the threats red gum), E. ovata (Labill.) (swamp gum), E. leucoxylon insect pests pose to plantations; the application of chemical (F.Muell.) (yellow gum), E. viminalis (Labill.) (manna gum) and insecticides is one of several techniques available to achieve this E. melliodora (Cunn. ex Schauer) (yellow box) (Costermans control (Neumann 1992). While the application of insecticides 1981; Boland et al. 1992). Average annual rainfall is about during outbreaks can provide effective control, the risks posed 670 mm, while the region has average minimum and maximum by ‘traditional’ spray applications (including the potential effects temperatures of 8° and 18°C respectively (Robinson et al. 2003). on non-target insect populations such as beneficial predator Soil pH in the study area is generally acid to neutral, while soil species), concerns relating to the use of chemicals in the environ- structure is generally soft to firm mottled-yellow duplex soils ment and risks to human health sometimes limit or preclude their with partial sub-surface bleach and a loamy clay surface texture use in plantations (Stone et al. 1998; Collett 2001). (Maher and Martin 1987; Robinson et al. 2003). To eliminate or minimise these risks, a new tablet formulation of an existing insecticide is currently under development. The tablets Trial design and treatment application containing insecticide are placed in holes under the seedlings at A 4 × 4 latin square design was used (four replicates × four planting, wherefrom the insecticide is absorbed systemically into treatments) with relatively large plots to take into account any the plant, imparting resistance to insect attack. The advantage of spatial variability across the site in soil properties and other site- such a targeted delivery system is that risks to human health, related factors. Treatments were: non-target insects and the environment in general are greatly reduced. The insecticide formulation being tested is imidacloprid, 1. an untreated control (no insecticide or fertiliser added to the an active ingredient already used in various horticultural, home- seedling-planting hole at establishment) garden and termite-control products. 2. a fertiliser-only treatment of one 2.5 g tablet containing NPKMg at 5.6, 5.4, 4.2 and 1.4% respectively added to each Imidacloprid belongs to a relatively new class of insecticides hole at planting called chloronicotinyls that exhibit high levels of insecticidal 3. an insecticide-only treatment of one 2.5 g tablet containing activity coupled with a high margin of safety — with currently 500 mg ai of imidacloprid added to each hole at planting approved use patterns — to mammals and other vertebrates. Imidacloprid is highly systemic in plants, a property very useful 4. an insecticide-plus-fertiliser treatment of one 2.5 g tablet when targeting sucking and chewing pests such as defoliating containing both components at the aforementioned rates added insects in eucalypt plantations. to each hole at planting.
While initial glasshouse and field trials have demonstrated the Each treatment plot of 32.0 m × 17.6 m contained of 81 –1 efficacy of the buried tablet mode of insecticide delivery, these E. globulus seedlings at a spacing of 4 m × 2.2 m (1100 trees ha ), trials involved simultaneous applications of imidacloprid and with an internal measurement/assessment plot of 49 seedlings fertiliser. There remained some uncertainty as to the apportioning (24.0 m × 15.4 m). Thus in total 196 E. globulus seedlings were of observed improvements in growth between the insecticide and assessed for each treatment and 784 seedlings for all four fertiliser components of the tablets. The aim of the present study treatments. was therefore to quantify the effects on defoliation and growth The trial site was ripped and mounded in June–July 2003, prior of the insecticide and fertiliser components of the tablets, and so to the trial establishment in August. Trees in the trial were planted to determine the true magnitude of the effect of the separate using standard establishment techniques with a Potti Putki® components of this tablet formulation on overall growth. seedling planting device whereby a slit is created by the device in the soil and the seedling dropped down a tube into the planting Materials and methods hole to be firmed in place by the heel of the boot. Where the treatment required the addition of a tablet, it was placed in the Study site planting hole prior to the seedling being planted. Follow-up The selected study site was located about south-east of the treatments included a post-planting herbicide application to the township of Meredith in west-central Victoria (37°3'41.94''S, trial site and surrounding plantation area in May 2004 (G. Ogston,
Australian Forestry 2007 Vol. 70 No. 1 pp. 53–60 N.G. Collett and J. McBeath 55
Treecorp, pers. comm. 2004). While the surrounding plantation information is specifically available on C. m-fuscum, this species had NPKS fertiliser (14.6:6:8:13.7) applied at a rate of 110 kg ha–1 is a member of a genus that has an extensive history of causing after planting, this was not applied to the trial site so as not to significant defoliation to a variety of eucalypt species throughout compromise the experimental treatments. Australia (Elliott et al. 1998).
While the three defoliators caused various levels of defoliation, Assessment the attacks were generally considered to be of low to moderate Tree height was measured using a telescopic height pole, and intensity, and trees substantially recovered lost foliage prior to defoliation due to insect herbivory in the upper and lower 50% the next attack. Although the three insect species varied in their of the total tree crown was scored using a scale with increments mode of attack, due to the relative uniformity of site conditions of 5%. The primary and, if present, any secondary causes of — including climate, planting stock used, establishment method defoliation were identified and recorded. Trees were also assessed and foliage type on trial trees available for attack — all trees for a variety of other factors that, if present, might affect tree within the trial had the same probability of being attacked at health and overall growth including any physical damage caused various stages of their growth, apart from any influence of the by animal browsing (i.e. rabbits and wallabies), any discoloura- experimental treatments. tion potentially caused by disease or nutritional factors, and During the assessment period from March 2004 to September foliage loss due to a range of abiotic factors including frost and 2005, apart from defoliation caused by the three major defoliating hail. The type of foliage present (juvenile, transitional, adult) insect pest species identified, no physical damage caused by was also recorded as this can directly relate to the susceptibility animal browsing (i.e. rabbits, wallabies, livestock, etc.) or abiotic of E. globulus to various insect pests and pathogens. The tree factors such as frost and hail was observed. No foliage crown was divided into upper and lower portions for scoring discoloration due to disease or nutritional factors was observed defoliation because insect defoliators sometimes display a throughout the study. All trees assessed had juvenile foliage only preference for particular sections of a tree’s crown (Collett and present during the March and July 2004 assessments. During the Neumann 2002). March 2005 assessment, transitional foliage had appeared in the The trial was assessed using the above criteria on four occasions upper crowns which developed into adult foliage by September throughout the study: in March and July 2004, and March and 2005, although the lower crowns and a significant proportion of September 2005. These assessments were scheduled to coincide the upper crown still carried juvenile foliage. with the periods of likely peak emergence of potential insect pests when damage to trees would be most evident but before the trees Dimensions and growth of the trial plantings had started to recover. Table 1 summarises the growth of the E. globulus for each of the four treatments from the commencement of the trial in August Analysis 2003 to September 2005. The height of all experimental trees at Percentage defoliation data were transformed using the arcsine the start of the trial (August 2003) averaged 14.0 cm (range: transformation. Different treatments within the study were 13.8–14.2 cm), while in March 2004 it averaged 85.8 cm (76.9– compared using analyses of variance (ANOVA with Fisher’s 97.7 cm). In August 2004 and March 2005 it averaged 127 cm PLSD for variables) to assess the effects of the various fertiliser (120–142 cm) and 354 cm (339–379 cm) respectively, while at and insecticide applications against the untreated control the end of the experiment in September 2005 (aged 2.1 y) the treatment on the extent of defoliation and tree height growth. trees averaged 422 cm (408–443 cm) (Table 1). Growth was The Kruskal–Wallis H test (Fowler et al. 1998, Zar 1999) was greatest between spring and summer, and least during autumn used to assess the significance of differences in defoliation and winter. between treatments. Effects of treatments on tree height
Results The patterns of growth of E. globulus in all treatments, expressed Major defoliating insect pest species identified during the as mean height from August 2003 to September 2005, were study generally sigmoid, thus conforming to generalised growth patterns for young forest trees. At the establishment of the trial in August From August 2003 to September 2005, three insect pest species 2003, height of none of the four treatments was significantly were identified as causing defoliation within the trial. These were different from that of any other (Table 1). Thereafter, the fertiliser- the lightbrown apple moth (Epiphyas postvittana Walker) in only treatment displayed growth generally consistent with that March 2004, autumn gum moth (Mnesampela privata Guenee) of the untreated control, apart from the March 2004 assessment in August 2004 and September 2005, and the eucalypt leaf beetle where the height of the fertiliser-only treatment was significantly (Chrysophtharta m-fuscum Boheman) in March 2005. E. postvittana greater than that of the untreated control. From then on, up to is a common leaf- and shoot-feeding caterpillar found throughout and including September 2005, there was no significant difference south-eastern Australia where it has caused significant damage in height between these two treatments. Height of the insecticide- to a range of plantation and orchard tree species, while M. privata only treatment was generally consistent with that of the fertiliser- is a major pest of juvenile E. globulus foliage and has caused only treatment up to and including August 2004, after when height significant damage in eucalypt plantations across south-eastern of the insecticide-only treatment was significantly greater than Australia in recent years (Elliott et al. 1998). While little that of the fertiliser-only treatment. The insecticide-only treatment
Australian Forestry 2007 Vol. 70 No. 1 pp. 53–60 56 Managing insect pests using insecticide tablets
Table 1. Tree heights (H), and levels of upper (U) and lower (L) crown defoliation of young Eucalyptus globulus following application of fertiliser and/or insecticide in tablets, and an untreated control
Mean height (cm), upper and lower-crown defoliation1 (%) ± s.e.2 (n = 180, 187) Treatment and variable Aug 2003 Mar 2004 Aug 2004 Mar 2005 Sep 2005 Establishment Assessment Assessment Assessment Assessment 1. Untreated control H3 14.0 ± 0.2a 76.9 ± 2.0a 119.6 ± 2.9a 338.8 ±6.5a 410.1± 7.3a U3 0.0 ± 0.0a 6.2 ± 0.6a 14.7 ± 0.8a 2.7 ± 0.2a 2.0 ±0.5a L3 0.0 ± 0.0a 7.8 ± 0.5a 16.3 ± 0.8a 4.9 ± 0.3ab 2.8 ±0.5a
2. Fertiliser only H 13.8 ± 0.2a 83.6 ± 2.1b 122.4 ± 2.6a 340.2 ± 5.3a 407.8 ± 6.4a U 0.0 ± 0.0a 5.9 ± 0.7a 14.7 ± 0.7a 2.7 ± 0.2a 1.8 ± 0.2a L 0.0 ± 0.0a 8.8 ± 0.7a 17.0 ± 1.0a 5.0 ± 0.3a 2.3 ± 0.2a
3. Insecticide only H 14.0 ± 0.2a 84.9 ± 1.7b 124.3 ± 2.5a 356.9 ± 5.3b 428.7 ± 6.0b U 0.0 ± 0.0a 1.7 ± 0.3b 9.1 ± 0.5b 2.2 ± 0.2ab 1.4 ± 0.2a L 0.0 ± 0.0a 2.9 ± 0.3b 9.7 ± 0.6b 4.2 ± 0.3bc 2.5 ± 0.2a
4. Fertiliser and.insecticide H 14.2 ± 0.2a 97.7 ± 1.7c 141.8 ± 2.4b 379.3 ± 5.1c 442.6 ± 6.1b U 0.0 ± 0.0a 0.8 ± 0.2b 8.0 ± 0.4b 1.8 ± 0.2b 1.4 ± 0.2a L 0.0 ± 0.0a 1.8 ± 0.2b 9.0 ± 0.5b 3.9 ± 0.2c 2.2 ± 0.2a
1 Within columns and for each variable, the means designated by different letters are significantly different (P <0.05) 2.s.e. = standard error 3. H = height, U = upper crown, L = lower crown
was significantly taller than the untreated control apart from Effects of treatments on lower-crown defoliation August 2004, when no significant difference was observed. The combined fertiliser-plus-insecticide treatment, however, At the establishment of the trial in August 2003, seedlings were displayed significantly greater height than the other three generally uniformly covered in healthy foliage with no insect treatments from March 2004 to March 2005, while recording no browsing observed: no treatment was significantly different from significant difference to the insecticide-only treatment in others in lower-crown defoliation (Table 1). As with the upper- September 2005. crown observations, no significant difference in levels of lower- crown defoliation was observed between the untreated control and fertiliser-only treatments throughout the study assessment Effects of treatments on upper-crown defoliation period (March 2004 – September 2005). The insecticide-only At the establishment of the trial in August 2003, seedlings were and fertiliser-plus-insecticide treatments also did not differ generally uniformly covered in healthy foliage and no insect significantly throughout the study, with similar levels of browsing damage was observed. No treatment was therefore defoliation being observed over all four assessment periods significantly different from others in observed upper-crown (March 2004 – September 2005). These observations confirm a defoliation (Table 1). Throughout the study, no significant general trend observed in the upper crown that when the untreated difference was observed between the untreated control and control and fertiliser-only ‘pair’ of treatments are compared to fertiliser-only treatments, with mean upper-crown defoliation the insecticide-only and fertiliser-plus-insecticide treatment pairs, levels of 6% and 6% respectively for the March 2004 assessment, treatments containing the insecticide component displayed 15% and 3% respectively for both treatments for the August 2004 significantly less defoliation due to assorted insect herbivory than and March 2005 assessments, and 2% and 2% for the September the untreated control or fertiliser-only treatments. This trend was 2005 assessment. The insecticide-only and insecticide-plus- consistent for the March and August 2004 assessments. In the fertiliser treatments did not significantly differ throughout the March 2005 assessment, however, the untreated control, while study, with similar levels of defoliation being observed over all having a higher mean level of defoliation than the insecticide- four assessment periods from March 2004 to September 2005. only treatment, did not differ significantly from it, while no When the untreated control and fertiliser-only ‘pair’ of treatments treatment differed significantly in the September 2005 assessment. are compared to the insecticide-only and fertiliser-plus-insecticide pair of treatments, however, treatments containing the insecticide Effects of treatments expressed as a function of defoliation component displayed significantly less defoliation due to insect category herbivory than the untreated control or fertiliser-only based treatments for the March and August 2004 assessments. During The defoliation scores across the four treatments show that most the March 2005 assessment, however, the insecticide-only trees defoliated in both the upper and lower crowns in March treatment did not differ significantly from any of the other three 2004 by E. postvittana were generally subjected to nil (0%) or ≤ treatments in mean levels of upper-crown defoliation, while in trace ( 10%) levels of defoliation, with only a small number September 2005, no treatment differed significantly. experiencing more than 10% defoliation. This is indicative of
Australian Forestry 2007 Vol. 70 No. 1 pp. 53–60 N.G. Collett and J. McBeath 57 low populations of this defoliator. In contrast, in the August 2004 defoliation, with >50% (upper crown) and 20% (lower crown) assessment for M. privata larvae, substantially more trees across of trees in each treatment recording nil (0%) defoliation. Thus the four treatments recorded defoliation levels of >10% (Figs 1 C. m-fuscum was at very low levels and causing only traces of and 2), reflecting a more sustained attack by a larger population. damage. This pattern of defoliation was also observed in In the March 2005 assessment for the defoliator C. m-fuscum, September 2005 when M. privata caused little defoliation (<10%) most trees across the four treatments recorded 10% or less in most trees across all four treatments.
100 Treatment One (untreated control) 90 Treatment Two (fertiliser only) Treatment Three (insecticide only) 80 Treatment Four (fertiliser and insecticide) 70
60
50
40
30
Fraction of trees within treatment (%) 20
10
0 0% 0% 0% 0% 10% 10% 20% 20% 30% 30% 40% 40% 10% 20% 30% 40% 10% 20% 30% 40% – – – – – – – – – – – – – – – – 41+% 41+% 41+% 41+% 1 1 1 1 11 11 21 21 31 31 11 21 31 11 21 31 March 2004 August 2004 March 2005 September 2005
Figure 1. Frequency of Eucalyptus globulus trees within the four treatments examined in various categories of defoliation based on percentage of defoliation to the upper crowns
100 Treatment One (untreated control) 90 Treatment Two (fertiliser only) Treatment Three (insecticide only) 80 Treatment Four (fertiliser and insecticide) 70
60
50
40
30
Fraction of trees within treatment (%) 20
10
0 0% 0% 0% 0% 10% 20% 30% 40% 10% 20% 30% 40% 10% 20% 30% 40% 10% 20% 30% 40% – – – – – – – – – – – – – – – – 41+% 41+% 41+% 41+% 1 1 1 1 11 21 31 11 21 31 11 21 31 11 21 31 March 2004 August 2004 March 2005 September 2005
Figure 2. Frequency of Eucalyptus globulus trees within the four treatments examined in various categories of defoliation based on percentage of defoliation to the lower crowns
Australian Forestry 2007 Vol. 70 No. 1 pp. 53–60 58 Managing insect pests using insecticide tablets
For the March 2004 assessment, 44, 45, 77 and 87% of trees in insecticide treatments, the fractions of trees within treatments the untreated control, fertiliser-only, insecticide-only and recording <10% defoliation in the upper crown were 71, 74, 91 fertiliser-plus-insecticide treatments respectively recorded 0% and 97% respectively (Table 2). The corresponding levels of (nil) defoliation for the upper crown. The corresponding levels upper-crown defoliation for August 2004 and March 2005 were of upper-crown defoliation for the same treatments in August 30, 20, 52, 57% and 93, 96, 97, 98% respectively, while for the 2004 and March 2005 were 1, 1, 3, 5%, and 53, 51, 60, 66% final assessment in September 2005 proportions of 97, 97, 99 respectively, while for the final assessment in September 2005 and 100% were recorded. For lower-crown defoliation in March nil defoliation was recorded in 73, 66, 72 and 72% of trees in the 2004, 64, 64, 90 and 97% of trees in the untreated control, above respective treatments (Table 2, Fig. 1). For lower-crown fertiliser-only, insecticide-only and, fertiliser-plus-insecticide defoliation in March 2004, 16, 17, 56 and 69% of trees in the treatments recorded <10% defoliation. In August 2004 and March untreated control, fertiliser-only, insecticide-only and fertiliser- 2005 the corresponding levels of lower-crown defoliation for plus-insecticide treatments recorded 0% (nil) defoliation, while the same treatments were 17, 16, 50, 50%, and 78, 81, 82, 85% the corresponding fractions of trees with 0% defoliation in the respectively, while for the final assessment in September 2005 lower crown for the same treatments in August 2004 were 0.0, proportions of 93, 97, 97 and 99% were recorded (Table 2). 0.5, 3 and 2% (Table 2, Fig. 2). In March 2005, 29, 22, 39 and 38% levels of 0% defoliation occurred respectively in the When the number of trees with <10% defoliation was compared aforementioned treatments, while in September 2005 correspond- to the number with 10% or more in the four different treatments ing levels of 62, 58, 53 and 57% were recorded. for the March and August 2004 assessments, significantly more trees in treatments containing insecticide (insecticide-only and When the number of trees with 0% (nil) defoliation was compared fertiliser-plus-insecticide) had <10% defoliation compared with to the number with defoliation levels of >1% in the four different treatments with no insecticide (the untreated control and fertiliser- treatments at the March 2004 assessment, significantly more trees only treatments) (Table 3). However, for the March and within treatments containing insecticide (insecticide-only and September 2005 assessments, no significant difference was fertiliser-plus-insecticide) had 0% (nil) defoliation when observed, indicating that most trees across all treatments compared to those in treatments not containing insecticide (the experienced very little insect defoliation. No differentiation untreated control and fertiliser-only treatments). However, at the between treatments was observed over the four observation August 2004 assessment, no significant difference was observed periods regardless of whether defoliation occurred in the upper (Table 3), indicating greater insect defoliating pressure on trees or lower crown. across all treatments for this period. For the March 2005 assess- ment, while no significant difference was observed between Discussion treatments for upper-crown defoliation, a significant difference was observed in levels of lower-crown defoliation. No significant This study is one of the first in Australia to evaluate the effects difference between the number of trees with 0% (nil) defoliation of applying insecticide in a tablet form at planting to eucalypts and those with >1% defoliation was observed in September 2005. in plantations as an alternative to more traditional broadscale spray applications with their potential adverse environmental When defoliation was assessed in March 2004 for the untreated effects and effects on non-target organisms. Previous studies by control, fertiliser-only, insecticide-only and fertiliser-plus-
Table 2. Fraction of Eucalyptus globulus trees recording a nil (0%)1 or low (<10%)2 defoliation score over from March 2004 to September 2005 after defoliation by three different insect pest species
Fraction of trees defoliated (%) LBAM3 AGM3 ELB3 AGM3 Treatment (E. postvittana) March 2004 (M. privata) August 2004 (C. m-fuscum) March 2005 (M. privata) September 2005
UC4 LC4 UC LC UC LC UC LC
0% <10% 0% <10% 0% <10% 0% <10% 0% <10% 0% <10% 0% <10% 0% <10% Untreated 44 71 16 64 1 30 0.0 17 53 93 29 78 73 97 62 93 control Fertiliser 45 74 17 64 1 20 0.5 16 51 96 22 81 66 97 58 97 only Insecticide 77 91 56 90 3 52 2.7 50 60 97 39 82 72 99 53 97 only Fertiliser + 87 97 69 97 5 57 1.6 50 66 98 38 85 72 100 57 99 insecticide
1 Per cent of trees within treatment with nil (0%) defoliation compared to trees with 1% or more defoliation 2 Per cent of trees within treatment with 9% or less defoliation compared to trees with 10% or more defoliation 3 LBAM = light brown apple moth, AGM = autumn gum moth, ELB = eucalypt leaf beetle 4 UC = upper crown, LC = lower crown
Australian Forestry 2007 Vol. 70 No. 1 pp. 53–60 N.G. Collett and J. McBeath 59
Neumann and Collett (1997a,b) examined the effects of different both the upper and lower crowns was significantly less in insecticides at varying rates using spray application methods in treatments with an imidacloprid component than in those with a controlled testing environment, while work by Candy et al. no imidacloprid. By the completion of the study, however, levels (1992) and Collett and Neumann (2002) examined the effects of of both upper and lower-crown defoliation were approaching artificial defoliation on tree growth. However, such studies have uniformity across all treatments, indicating that the protective been conducted in isolation to the extent of examining insecticide effects of the tablets containing imidacloprid were diminishing. application independently of tree growth, or examining the effects This could be due either to the eventual breakdown and disappear- of defoliation on tree growth in the absence of both defoliating ance of the tablets by the second year, and/or the growth of the insects and insecticide application. The current study addresses trees to such a size that the amount of imidacloprid available to this gap in our knowledge by examining a new environmentally- them no longer provided adequate and uniform protection from friendlier method of insecticide application in eucalypt herbivorous predators. plantations, coupled with assessments of tree defoliation by various defoliating agents and of height growth. Studies involving This observation is further strengthened when defoliation insecticides and associated defoliation in the field are subject to assessment scores were grouped in a 0% (nil) category or in the uncertainty that defoliating insect species may not appear at categories of 1–10%, 11–20% etc., and examined in more detail. the right time and place and in sufficient numbers to inflict In the first year of observation, regardless of grouping, treatments measurable levels of damage. This study was fortunate, however, containing imidacloprid retained more foliage and were subject in experiencing defoliation episodes of sufficient intensity at to less insect defoliation than were treatments without different times across the entire study area and its surrounds to imidacloprid. Although the large populations of the defoliating enable the various treatments to be thoroughly assessed. insect M. privata present in August 2004 damaged most trees in all treatments, the treatments containing imidacloprid still We have shown that in an E. globulus plantation from initial experienced significantly less defoliation (<10%) than treatments establishment up to 2.1 y of age and prior to canopy closure, with no imidacloprid. That the effects of the imidacloprid trees treated at planting with tablets containing imidacloprid (with appeared to be diminishing in the second year of the study is and without accompanying fertiliser) generally displayed indicated by the observation that levels of defoliation across all significantly better height growth and had more unbrowsed foliage treatments were trending towards statistically-similar levels. than treatments containing no imidacloprid (the untreated control and fertiliser-only treatments). This trend was particularly evident Treatment effects on tree heights require particular examination in the first year of the study, when insect-related defoliation in in that while it appears that increased height growth in
Table 3. Significance of differences between three different fertiliser/insecticide treatment datasets and an untreated control expressed by their mean rankings for the Kruskal-Wallis H-test, for three defoliation periods split into upper and lower crown assessments for a) trees with 0 % defoliation and b) trees with <10% defoliation
Treatment/mean rank (n = 180–188)1 Defoliation category, H statistic2 and 1. Untreated 4. Fertiliser + date and crown position 2. Fertiliser only 3. Insecticide only significance control insecticide
a) 0% (nil) defoliation3 Upper crown 444 441 319 284 082.9*** March 2004 Lower crown 461 456 311 261 125.3*** Upper crown 380 375 367 365 000.6NS August 2004 Lower crown 381 369 363 373 000.7NS Upper crown 385 395 360 339 007.8NS March 2005 Lower crown 381 406 346 346 010.3* Upper crown 361 386 364 363 001.6NS September 2005 Lower crown 357 370 388 359 002.5NS b) <10% (low) defoliation4 Upper crown 419 409 340 320 029.6*** March 2004 Lower crown 423 442 325 297 061.8*** Upper crown 406 443 328 310 048.9*** August 2004 Lower crown 433 440 308 308 066.7*** Upper crown 380 369 367 361 000.7NS March 2005 Lower crown 383 370 369 359 001.2NS Upper crown 375 372 365 362 000.4NS September 2005 Lower crown 381 366 369 360 000.9NS
1 n = range of the number of observations across the four treatments 2 Levels of significance are * = P<0.05, *** = P<0.001 and NS = not significant (Kruskal–Wallis H-test) 3 Ranking based on numbers of trees within each treatment with 0% (nil) defoliation 4 Ranking based on numbers of trees within each treatment with <10% defoliation
Australian Forestry 2007 Vol. 70 No. 1 pp. 53–60 60 Managing insect pests using insecticide tablets imidacloprid-only treated trees is a function of them experiencing Collett, N.G. (2001) Biology and control of psyllids, and the possible less defoliation and consequently growing more vigorously than causes or defoliation of Eucalyptus camaldulensis Dehnh. (river untreated trees, trees treated with fertiliser would also benefit in red gum) in south-eastern Australia — a review. Australian terms of their potential to grow more vigorously. This trend is Forestry 64, 88–95. confirmed early in the trial with the fertiliser and insecticide- Collett, N.G. and Neumann, F.G. (2002) Effects of simulated chronic only treatments displaying similar height growth patterns defoliation in summer on growth and survival of blue gum compared to the untreated trees which recorded inferior height (Eucalyptus globulus Labill.) within young plantations in northern Victoria. Australian Forestry 65, 99–106. growth. Of particular interest however, were the apparent positive synergies developed when fertiliser and imidacloprid were Commonwealth of Australia (2005) Australia’s Forests at a Glance. Bureau of Rural Sciences, Department of Agriculture, Fisheries combined in one treatment and subsequent height growth was and Forestry, Canberra, 68 pp. superior to that in all other treatments. This trend remained Costermans, L.F. (1981) Trees of Victoria – An Illustrated Field Guide. constant throughout the two-year study, indicating that not only Costermans Publishing, Frankston, Victoria, 84 pp. would trees treated with imidacloprid but those also containing a fertiliser component would substantially increase in height Elliott, H.J., Bashford, R. and Greener, A. (1993) Effects of defoliation by the leaf beetle, Chrysophtharta bimaculata, on growth of growth. Despite indications that the initial direct effects of the Eucalyptus regnans plantations in Tasmania. Australian Forestry insecticide and the fertiliser tablets wear off after one year post 56, 22–26. planting, height growth in both treatments containing imidacloprid Elliott, H.J., Ohmart, C.P. and Wylie, F.R. (1998) Insect Pests of was still superior to that in the other treatments at the conclusion Australian Forests: Ecology and Management. Inkata Press, of the study. This indicates that the treatments may have assisted Australia. in establishing healthier and vigorously growing trees with better Fowler, J., Cohen, L. and Jarvis, P. (1998) Practical Statistics for Field root systems and access to water and nutrients. Similar recent Biology. Second edn. John Wiley and Sons, England. studies undertaken by Forests New South Wales indicate that Maher, J.M. and Martin, J.J. (1987) Soils and Landforms of South- height growth responses in treatments using the tablets were Western Victoria. Part 1. Inventory of Soils and their Associated maintained for up to 5 y (C. Stone, Forests NSW, pers. comm., Landscapes. Research Report Series No. 40. Department of 2007). Re-assessment of the current study is therefore desirable Agriculture and Rural Affairs, Melbourne. at a later date to determine whether this height growth trend also National Forest Inventory (2004) National Plantation Inventory Update subsequently continues. — March 2004. Bureau of Rural Sciences, Canberra. When the aforementioned observations are considered in light Neumann, F.G. (1992) Regulation and usage of insecticides in Australian forestry from the mid-1960s to 1990. Australian of the practical relevance for commercial eucalypt plantations, Forestry 55, 48–64. the cost-benefit over time from the use at planting of the combined Neumann, F.G. (1993) Insect pest problems of eucalypt plantations in fertiliser/imidacloprid tablet could potentially be considerable. Australia, 3. Victoria. Australian Forestry 56, 370–374. The requirement for other insecticide and fertiliser application Neumann, F.G. and Collett, N.G. (1997a) Insecticide trials for control in the first one to two years could potentially be eliminated, saving of the steelblue sawfly (Perga affinis affinis), a primary defoliator the application costs in both instances and using the synergies of in young commercial eucalypt plantations in south-eastern the one-step tablet application at planting. Furthermore, the Australia. Australian Forestry 60, 75–83. growth benefits imparted by the tablet may result in greater yields Neumann, F.G. and Collett, N.G. (1997b) Insecticide trials for control over shorter rotation periods. Further study, however, is required of the autumn gum moth (Mnesampela privata), a primary to accurately quantify these potential benefits. defoliator in commercial eucalypt plantations prior to canopy closure. Australian Forestry 60, 130–137. Acknowledgements Robinson, N., Rees, D., Reynard, K., MacEwan, R., Dalhaus, P., Imhof, M., Boyle, G. and Baxter, N. (2003) A land resource assessment Many thanks to David Smith, Paul Clements, Carolien of the Corangamite region. Centre for Land Protection Research, Schoenborn and Glenn Ogston for assistance in setting up and Department of Primary Industries, Victoria. assessing the trial, and Treecorp Pty Ltd and the landholder Stone, C., Simpson, J.A. and Eldridge, R.H. (1998) Insect and fungal Mr Branagh for providing the study site and ongoing access. damage to young eucalypt trial plantings in northern New South Thanks also to the referees who provided valuable comments on Wales. Australian Forestry 61, 7–20. the manuscript. Stone, C., Simpson, J.A. and Gittins, R. (1998) Differential impact of insect herbivores and fungal pathogens on the Eucalyptus subgenera Symphyomyrtus and Monocalyptus and genus References Corymbia. Australian Journal of Botany 46, 723–734. Australian Bureau of Agriculture and Resource Economics (2000) United States Department of Agriculture (2003) Pest Risk Assessment Australian Forest Products Statistics, June Report. ABARE, of the Importation into the United States of Unprocessed Logs Canberra. and Chips of Eighteen Eucalypt Species from Australia. USDA Boland, D.J., Brooker, M.I.H., Chippendale, G.M., Hall, N., Hyland, Forest Service General Technical Report FPL-GTR-137, 206 pp. B.P.M., Johnston, R.D., Kleinig, D.A. and Turner, J.D. (1992) Wills, A.J., Burbidge, T.E. and Abbott, I. (2004) Impact of repeated Forest Trees of Australia. CSIRO, Canberra. defoliation on jarrah (Eucalyptus marginata) saplings. Australian Candy, S.G., Elliott, H.J., Bashford, R. and Greener, A. (1992) Forestry 67, 194–198. Modelling the impact of defoliation by the leaf beetle Zar, J.H. (1999) Biostatistical Analysis. Fourth edn. Prentice Hall Chrysophtharta bimaculata (Coleoptera: Chrysomelidae), on International, New Jersey. height growth of Eucalyptus regnans. Forest Ecology and Management 54, 69–87.
Australian Forestry 2007 Vol. 70 No. 1 pp. 53–60 Micaela Preda, Kenneth A. Bubb and Malcolm E. Cox 61
GIS-based tools for management of pine plantations, Queensland, Australia
Micaela Preda1,2, Kenneth A. Bubb3 and Malcolm E. Cox4
1Queensland University of Technology, School of Natural Resource Sciences, GPO Box 2434, Brisbane, Queensland 4001, Australia 2Email: [email protected] 3DPI Forestry, MS 483 Fraser Rd, Gympie, Queensland 4570, Australia 4Queensland University of Technology, School of Natural Resource Sciences, GPO Box 2434, Brisbane, Queensland 4001, Australia
Revised manuscript received 22 January 2007
Summary Introduction
Two examples of GIS-based multiple-criteria evaluations of Geographic information systems (GIS) have developed rapidly plantation forests are presented. These desktop assessments use over the last several decades and they are now widely used for available topographical, geological and pedological information basic spatial analyses (i.e. measurement and overlay tasks), carto- to establish the risk of occurrence of certain environmentally graphic modelling (i.e. landuse suitability), or the latest artificial detrimental processes. The first case study is concerned with the intelligence applications (Malczewski 2004). Despite their great risk that chemical additives (i.e. simazine) applied within the capabilities, however, GIS are not widely employed to their full forestry landscape may reach the drainage system. The second potential. In many activities where spatial data are available, GIS case study assesses the vulnerability of forested areas to are still used only for basic manipulation and visualisation of landslides. georeferenced datasets (e.g. Pettit and Pullar 1999). Regardless of the sophistication of the method or approach employed, the The subject of the first multiple-criteria evaluation (MCE) was a ultimate goal of any GIS-based analysis is to provide support for 2 4 km logging area, which had been recently site-prepared for a making spatial decisions (Malczewski 1999). As a result, the Pinus plantation. The criteria considered relevant to the integration of geographic information systems with multiple assessment were proximity to creeks, slope, soil depth to the criteria decision-making methods (MCDM) has been extensively restrictive layer (i.e. potential depth to a perched water table) explored by many researchers over the last 15 y (e.g. Carver and soil erodability (based on clay content). The output of the 1991; Jankowski 1995; Malczewski 1996, 1999; Hajkowicz 2002). MCE was in accordance with field observations, showing that this approach has the potential to provide management support Modelling using GIS-based multiple criteria approaches has by highlighting areas vulnerable to waterlogging, which in turn increasingly been used in the analysis of land suitability (Pereira can trigger overland flow and export of pollutants to the local and Duckstein 1993; Laaribi et al. 1996; Ceballos-Silva and stream network. Lopez-Blanco 2003; Malczewski 2004), in ecological applications (Store and Kangas 2001; Rouget et al. 2003; The subject of the second evaluation was an Araucaria plantation Gkaraveli et al. 2004) or management of natural resources which is prone to landslips during heavy rain. The parameters including water (Tkach and Simonovic 1997; Schumann and included in the assessment were drainage system, the slope of Geyer 1999; Nath et al. 2000; Bhuyan et al. 2003). Multicriteria the terrain and geological features such as rocks and structures. analyses have also been adapted to map the risk of certain A good correlation between the MCE results and field observa- phenomena or processes occurring, such as natural hazards (Chen tions was found, suggesting that this GIS approach is useful for et al. 2001), erosion (Bantayan and Bishop 1998; Dragan et al. the assessment of natural hazards. 2003; Sivertun and Prange 2003) and pollution of water supplies (Foster and McDonald 2000; Sivertun and Prange 2003). More Multiple-criteria evaluations are highly flexible as they can be recently, GIS have been used in conjunction with specialised designed in either vector or raster format, depending on the type hydrological models to assist with data integration, mapping and of available data. Although tested on specific areas, the MCEs visualisation (Bhuyan et al. 2003; Jain et al. 2004; Ropke et al. presented here can be easily used elsewhere and assist both 2004). management intervention and the protection of the adjacent environment by assessing the vulnerability of the forest landscape In the context of extensive GIS use in many domains, there is to either introduced chemicals or natural hazards. little published information on forestry applications, although Keywords: geographical information systems; forest management; this industry has been expanding worldwide over the last 25 y. decision making; evaluation; assessment; risk; landscape; spatial According to the Food and Agriculture Organisation of the United variation Nations there were almost 124 million ha of plantation forests in
Australian Forestry 2007 Vol. 70 No. 1 pp. 61–69 62 GIS tools for managing pine plantations the world in 1995, of which 25% were established in tropical Although tested on specific locations, the GIS-based assessments regions (Brown 2000). Australia manages 1.3 million ha of can be easily used elsewhere and assist both management plantations; this area is expected to increase over the next decade, intervention and the protection of the adjacent environment. making Australia a major timber producer in the Pacific Rim Significantly, these evaluations do not preclude the use of region (NAFI 2004). specialised modelling techniques or programs. In addition, multiple-criteria evaluations can be easy to use and are of value To date, GIS-based analysis or modelling of the forestry landscape as precursors to modelling as they can identify areas of concern. has been mostly concerned with ecological aspects (e.g. Kangas et al. 2000; Nagashima et al. 2002), although increasing attention is being given to soil characteristics and their role in forest growth Case study 1 — Chemical treatment (e.g. Payn et al. 1999). Recently, GIS-assisted remote sensing Problem description techniques have been used to process data on forest growth or to carry out soil surveys (Coops and Waring 2001; Balzter et al. The subject of the GIS-based analysis was a 4 km2 logging area 2003; Riano et al. 2004). Other aspects of concern in silviculture located in tropical northern Queensland, Australia (Fig. 1). The are related to hydrological issues such as water availability, quality area had been recently site-prepared for a Pinus plantation using and flow, as well as waterlogging, which can be significant in a combination of cultivation techniques — spot mounding on low-lying coastal areas. Such processes are typically modelled slopes >5% and continuous mounds on slopes < 5%. The soil using specialised hydrological software packages (e.g. type was classified as a Podzolic, characterised by a shallow loam Koesmarno 1997; Bubb and Croton 2002; Ticehurst et al. 2003), (0.5–1.0 m) over heavy clay with low hydraulic conductivity; many without the assistance of GIS. Furthermore, GIS-based this promotes the formation of a perched watertable in the upper multicriteria analyses are yet to be widely used for forestry profile. The residual herbicide simazine was manually applied applications, although GIS databases are now commonly along the planting band at 5 kg ha–1 (treated area) to individual available and contain information that could be processed at a sections of the logging area, between 23 December 2003 and higher level. In Australia, for example, collection and recording 29 February 2004. Simazine was used for control of perennial of detailed information on topography, soil character and weeds dominated by thick swards of tropical pasture grasses. silviculture for all government-managed plantation forests has This period was characterised by irregular rainfall events, which become standard practice. Such information is primarily used along with normal breaks in field operations resulted in a mosaic for simple visualisation, despite its scientific significance for of treated areas throughout the area. management and decision making. As part of the associated monitoring program, monthly grab- This paper presents two forestry examples of GIS-based multiple- samples were taken from the creek which drains the area (Fig. 1). criteria evaluations which use available topographical, geological Simazine concentrations above the aquatic ecosystem trigger- and pedological information to establish the risk of occurrence values of 3.2 µg L–1 (ANZECC and ARMCANZ 2000) were of certain environmentally detrimental processes. The first case detected in samples taken on 16 January 2004 and 5 February study is concerned with the risk that chemical additives applied 2004, whereas samples in the seven months following were below within the forestry landscape may reach the drainage system; the this value. These initial findings resulted in a field inspection second assesses the vulnerability of forested areas to landsliding.
Monitoring site N
M a r y v a le C r e e k
Forestry compartments
Stream
0 1 kilometre
Figure 1. Location of the study area showing forestry compartments, drainage system and the monitoring site
Australian Forestry 2007 Vol. 70 No. 1 pp. 61–69 Micaela Preda, Kenneth A. Bubb and Malcolm E. Cox 63 and review of the simazine application procedures along with a control over the transport of additives to the drainage system, proposal to undertake a GIS-based desktop study in order to and rainfall will influence only the time and quantity of pollution. identify the areas likely to export simazine offsite. Existing data such as topography, soil characteristics, rainfall records and As this MCE was directed to the modelling of natural parameters, history of simazine applications were made available and used the layers included were proximity to creeks, slope, depth to the in this investigation. For the purpose of this study, the term restrictive soil layer (i.e. potential depth to a perched water table) ‘vulnerability’ refers to the likelihood that an introduced chemical and soil erodability, which was based on clay content (i.e. little (i.e. simazine) will be lost from a certain area of land and will clay equates to high risk). Other factors potentially influencing enter the drainage system. overland flow such as climate, vegetative cover and stream characteristics (Ward and Robinson 1990; Manning 1992; Simazine is moderately persistent with a half-life of 1–20 weeks Wanielista et al. 1997) were regarded as quite homogenous (Meakins et al. 1995; Vink and Van der Zee 1997; Bubb 2001; throughout the study area and therefore not considered. The final Calderon et al. 2004). This high variability is most likely due to weighted score was the result of the summation of all these the simazine response to various field conditions; for example parameters; it was not aimed at quantifying the actual overland warm temperatures, high soil moisture, low pH and the presence flow but at visualisation of sites highly vulnerable to chemical of sandy material can all increase the rate of simazine breakdown transport due to waterlogging and overland flow potential. (Meakins et al. 1995; Thomas and Piper 2001). Apparently there is little lateral movement of the chemical within the soil profile, Parameters and their weights and many studies have observed simazine in the top 20–100 cm of soil (e.g. Close et al. 1998; Bubb 2001; Thomas and Piper The drainage system layer of the GIS was used to create a new 2001). The low solubility of this herbicide can cause slow release, layer containing a 100-m buffer centred on the creek (100 m each and the peak of its concentration in soil and groundwater can side), this being the area most likely to waterlog and produce occur 10–20 days after application. Rainfall, however, can overland flow that would reach the creek; the remaining area influence this process, as about 90% of simazine is transported further from the stream network was considered of low in dissolved form (Garmouma et al. 1997; Bubb 2001). Rainfall vulnerability to waterlogging (Fig. 2a). As surface water flow is simulations showed that the simazine concentration in runoff can the main mechanism of exporting pollutants, this layer of be quite high immediately after application, but can decrease by proximity to creeks was given a weight of 40%. a factor of 5 and 10 times after rain events which are simulated one and two weeks after application, respectively (Liu and The slope layer was generated from 0.25-m elevation contours O’Connell 2003). Bubb (2001) found large simazine ‘spikes’ in and was divided into five categories of slope (Fig. 2b). Field streamwater to be associated primarily with the first runoff/flood observations showed that the gentler slopes of low-lying coastal event following application to plantation areas; attenuated stream areas are more likely to waterlog and cause overland flow. This concentrations (order of magnitude) were measured in following process is referred to as saturation overland flow; it occurs when events, for up to six weeks after application. These characteristics the soil profile is almost or completely saturated, effectively are common to many herbicides and are relevant to the design of reducing surface infiltration (Ward and Robinson 1990). the GIS assessment. Therefore, in the MCE analysis, gentler slopes were considered more susceptible to export of soil additives. The weight of this data layer was set at 40%. Method The soil erodability layer, and depth to the restrictive layer, were Design and rationale point layers and therefore treated similarly; using radial function The GIS method employed for the assessment of the study area interpolation polygonal objects were created to enable inter- was a multiple-criteria evaluation (MCE) as described by Carver section with the other polygon files. It was assumed that the higher (1991). The geoprocessing was carried out in a vector format the erodability of surficial soils, the higher was the risk of simazine and consisted of the intersection of several layers of data; the adsorbed onto particles reaching the creeks during rain events. software employed was ArcView 3.1. The data were normalised This property, however, was given a low weighting of 10% in the and a final weighted score of vulnerability was calculated and analysis, as simazine is known to be primarily transported in plotted. dissolved form (Bubb 2001; Thomas and Piper 2001). If the MCE were to be run for phosphorus, for example, a higher weighing We considered that an assessment of the morphology and soil for erodability should be applied as this nutrient is mainly characteristics of the area would enable evaluation of the transported by suspended particles. It was also assumed that the vulnerability to waterlogging, which in turn can trigger overland greater the depth to a restrictive layer of low permeability, the flow and consequently export of additives. Rainfall duration and less likely is the terrain to waterlog and produce overland flow. intensity are always used in hydrological assessments and This layer was also given a 10% weighting. modelling (e.g. Foster and McDonald 2000; Ticehurst et al. 2003), but rainfall was not included in this MCE as it was Results and discussion considered that precipitation is not directly linked to the propensity of a region to release pollutants. Previous work in Vulnerability mapping these landscape settings has found any major influence on runoff to be via the process of saturated overland flow, whereby a The intersection of the four layers produced a new file containing saturated soil profile is the precursor to runoff. As a consequence, several thousand polygons, each characterised by a value of slope, the natural setting (topography and soil character) is the main a value defining its proximity to the creek (i.e. closer or further
Australian Forestry 2007 Vol. 70 No. 1 pp. 61–69 64 GIS tools for managing pine plantations
(a) Proximity to creeks (b) Slope (%)
Forestry compartments 0–5 10–15 > 20 100-m buffer 5–10 15–20 Streams
(c) Erodability (on a scale from 1 (low) to 9 (high)) (d) Depth to the restrictive layer (cm)
# #
### ###
# # # # # # # # # # # # # # # # # # # ##### # # # #####
# # # # # # # # # # # # # # # #
# # # ##### # # # #####
# # # # # # # # # # # # # # # #
# ########## # ##########
# # # # # # # # # # # # # # # #
# # # # # # # # # # # # # # # # # # # # # ## # # # ## # # # # # # # # # # # # # # # # # # # # # # # #
3–4 30–53 Soil sampling points 4–5 53–77 5–6 Streams 77–100
Figure 2. Parameters of the MCE: (a) proximity to creeks in the form of a 100–m wide buffer and (b) slope derived from elevation surveys; (c) erodability and (d) depth to the restrictive layer are based on field analyses and measurements, and are presented as interpolated surfaces.
than 100 m from creek), an erodability value and a depth to the The main advantage of this type of analysis is that the weighted restrictive layer. These values were, however, of different orders scores can be easily recalculated and the relationship between of magnitude and, most importantly, were measured in different parameters reassessed when more data are collected; the main units. In order to further manipulate the data, these values were disadvantage is that the MCE analysis assumes a linear relation- standardised, weighted and added to form a new column of ship between parameters, which may not always be the case. weighted scores (Carver 1991). Plotting these scores enabled the However, field observations and measurements can assist visualisation of the combined influence of all parameters identification of any non-linear relationships and improve the considered. The results were in the form of a map of vulnerability worth of the final weighted score. to waterlogging, overland flow generation, and simazine or nutrient export to waterways (Fig. 3). The objects of this map were then merged back into compartments to provide plantation managers with an easy-to-use decision-making tool (Fig. 4).
Australian Forestry 2007 Vol. 70 No. 1 pp. 61–69 Micaela Preda, Kenneth A. Bubb and Malcolm E. Cox 65
MCE analysis versus field measurements
The analysis showed that, in addition to areas immediately adjacent to watercourses, the south-eastern corner of this new plantation area is the most vulnerable to waterlogging (Fig. 4). This finding is consistent with field observations.
The application of simazine began on 23 December in the south- eastern compartments (Fig. 5) but it is likely that this had little effect on the water quality of the nearby creeks as there was no rain for 12 days after treatment. However, the 34-mm rain event that followed probably caused the simazine transport and the high ‘spike’ in the streamwater surface runoff detected on 16 January. Although the treatments continued on 13, 14 and 15 January, the lack of rain at those times and the fact that those compartments are at limited risk (Fig. 4) suggests that the transport of simazine to the stream was also limited. We conclude that the simazine detected at the monitoring site (Fig. 1) on 16 January 2004 was the result of the 34 mm of rain that fell on 5 January and runoff- transported simazine from the south-eastern compartments that were treated on 23 and 24 of December 2003. < 0.4 0.4 –0.6 A high simazine spike was again detected at the monitoring site > 0.6 Streams on 5 February. It is assumed that this was caused by the 53-mm
Figure 3. Map of vulnerability showing the MCE solution, including (a) proximity to creeks (40%), slope (40%), erodability (10%) and depth to the restrictive layer (10%). The flat areas close to creeks are more Monitoring site likely to export pollutants due to both proximity to drainage and low slope. Medium slopes with small depths to the restrictive layer are also 16.01 of high vulnerability (score > 0.6), followed by flat areas of high M 3.02 a 3.02 13.01 ry erodability (score = 0.4–0.6). The lowest vulnerability (score <0.4) is v a associated with steep slopes located away from the stream network 3.02 16.0116. le C r 01 24.12 e 2.02 e k 16.0116.01 24.12
15.0115.01
24.12
14.0114. 01 23.12
23.12
0 1 kilometre
(b) 60 Simazine 53 mm 50 application 16.12.04 5.02.04 40 Water monitoring 34 mm 30 23–24.12.03 20 13–16.01.04 2–3.02.04 Rainfall (mm) 10 0 < 0.4 0.4 –0.6 1/01/04 7/01/04 6/02/04 > 0.6 Streams 20/12/03 26/12/03 13/01/04 19/01/04 25/01/04 31/01/04 Date
Figure 5. (a) Map of the logging area and (b) diagram showing the Figure 4. Vulnerability map of weighted scores averaged to dates of simazine application and those of water quality measurements, compartment level in relation to rainfall events
Australian Forestry 2007 Vol. 70 No. 1 pp. 61–69 66 GIS tools for managing pine plantations rain event that occurred on 3 February, with most contamination Method being from herbicide recently applied in the northern compartments on 2 and 3 February (Fig. 5). According to the Design and rationale MCE analyses, a significant proportion of the areas around the The MCE approach was directed to the assessment of major watercourse are of high risk, which could explain the high vulnerability to landslides. Considering that the plantation forests concentration of simazine. However, the February spike was much had not significantly modified the terrain in the area, the main lower than that in January; this difference is significant when factors taken into account in this assessment were natural and considering that the quantity detected in February was added to related to the well-established causes of mass wasting (i.e. that already in the system from the December and January landslides), which include: applications. This difference is explained by the MCE, which showed that the compartments treated at the beginning of • adverse bedding or rock structures — exposed bedding or joint February were of lower vulnerability than those treated in planes are surfaces on which rock masses can slide. December. • oversteepened slopes — the resisting mass at the base is lessened and potential failure planes, such as stratification, may Case study 2 — Slope stability be exposed; slopes > 25° are the most likely to slide, regardless of the geological material involved. Problem description • increased water content — the water does not necessarily act as a lubricant, but increases pore water pressure and changes 2 The subject of the second evaluation was a 120 km Araucaria the effective stress between mineral grains; in addition, water plantation located in subtropical Queensland. As the area assessed adds weight to the slope (e.g. Summerfield 1991). is extensive and composed of several state forests, the methodology will be demonstrated using only two logging areas The GIS method used was a multiple-criteria evaluation (MCE) (Fig. 6). The terrain is rugged, of high relief and subject to adapted from Carver (1991). The analysis was carried out in a landslides during heavy rain. Although landslips have been raster format and included the transformation of the vector data mapped in some areas, a regional GIS-based evaluation of this into raster format, their reclassification in terms of vulnerability natural hazard was required. to landsliding, and finally the calculation of the weighted score of vulnerability. The program used was ArcView 8.3.
Parameters and their weights
Based on the theoretical background and considering the available data, the design of the MCE included the drainage system, the slope of the terrain and geological features such as rocks and structures. Several combinations of weightings for different data Forestry compartments layers were tested, but the one that best matched the situation in the field was drainage 30%, slope 50%, geology 10% and Streams structure 10%.
As water can be an important control of landslide occurrence, the density of the drainage system in the region was included in the analysis. For the purpose of the analysis, it was considered that the denser the stream network, the more likely is sliding to occur.
Slope was based on the 20-m elevation map of the region. It was considered that slopes > 25° are the most vulnerable to landslides. In addition, on slopes steeper than 35°, tree weight can also decrease stability (Selby 1993).
Geology was based on the 1:100 000 geological map (Natural Resources, Mines and Energy 2004). For the purpose of the analysis, the geological units were ranked into five categories of vulnerability to landslides, from 5 (high) to 1 (low): • 5 — metamorphic rocks ± sedimentary rocks consisting of schists or other highly foliated rocks 0 1 kilometre • 4 — sedimentary rocks interbedded with volcanics and unconsolidated alluvium Figure 6. Location of the study area showing forestry compartments • 3 — sedimentary rocks including very fine and very coarse and drainage system material
Australian Forestry 2007 Vol. 70 No. 1 pp. 61–69 Micaela Preda, Kenneth A. Bubb and Malcolm E. Cox 67
• 2 — sedimentary rocks, generally sandstones When compared to the situation in the field (Fig. 8), the model • 1 — intrusive and volcanic rocks with massive structures appeared to be ‘the worst case scenario’ because not all the areas (granite, granodiorite, rhyolite, andesite, basalt). highlighted by the vulnerability evaluation were known to experience landslides. This discrepancy may be due to the fact Structure was also based on the 1 : 100 000 geological map. This that existing slips are discovered only prior to clearfelling and layer included all known structural features of the area: geological could be hidden for many decades during the 50-y rotation. boundaries, various types of faults, lineaments and shear zones, Nevertheless, the good correlation between field observations as they all represent weak areas where landslides can be initiated. and the desktop analysis revealed new applications for the MCE. As with the drainage system, this layer was used to produce a Such models are not only predictive (i.e. they identify areas of density map, where the denser these features, the more vulnerable risk in terms of natural hazards) but also can help the user, while the area. refining the weighting scheme to match the situation in the field, to recognize the main controls over a particular hazard or process. Results and discussion In the case presented here, the MCE version that matched the field situation was slope 50% + drainage 30% + geology 10% + Vulnerability mapping versus field observations structure 10%; this weighting suggests that topography and the presence of water are the main controls over landslide generation. The intersection of the four layers produced a new raster map For other areas where geology is more varied and/or there is a (30-m resolution) displaying the vulnerability to landslides large density of structural features, the weighting scheme would (Fig. 7). In general, half of southern section and most of the probably need to be changed to accommodate these characteristics. northern logging area are of concern under wet conditions. General conclusions and future work
The multiple-criteria evaluation used in this study has the potential to provide management support by identifying areas vulnerable to pollution or natural hazards. Additionally, it is highly flexible as it can be designed in either vector or raster format, depending on the type of available data.
As the topographical and pedological data used in the first analysis < 0.4 are routinely collected from areas being prepared for plantations, 0.4–0.6 this MCE method can be easily applied elsewhere and also > 0.6 provides a precursor for hydrological modelling. As more field data are collected, the weighting of attributes can be refined to provide an increasingly accurate assessment.
< 0.4 0.4–0.6 > 0.6
known areas of landsliding
Figure 8. Comparison between the MCE solution and the areas known to have experienced landslides during heavy rain. There is a good correlation between the model and the field observations, although not Figure 7. Map of vulnerability to landslides showing the MCE solution, all the areas considered to be of high risk are known to be hazardous. based on the density of the drainage system (30%), slope (50%), geological units (10%) and structural features (10%)
Australian Forestry 2007 Vol. 70 No. 1 pp. 61–69 68 GIS tools for managing pine plantations
The geological and structural data used in the second analysis Balzter, H., Skinner, L., Luckman, A. and Brooke, R. (2003) Estimation are currently available for the entire state of Queensland, although of tree growth in a conifer plantation over 19 years from multi- the scale or resolution may not always be appropriate for the satellite L-band SAR. Remote Sensing of Environment 84, type of assessment required. However, the most important param- 184–191. eters for many applications are altitude and slope, as these Bantayan, N. and Bishop, I.D. (1998) Linking objective and subjective topographic features govern most natural hazards and human- modelling for landuse decision-making. Landscape and Urban induced processes. As high-resolution digital elevation data Planning 43, 35–48. become available, more detailed assessments can be obtained. Bhuyan, S.J., Koelliker, J.K., Marzen, L.J. and Harrington, Jr. J.A. (2003) An integrated approach for water quality assessment of a From the first case study it can be concluded that MCE Kansas watershed. Environmental Modelling and Software 18, assessments can be a valuable tool for planning forest treatment, 473–484. and they should be run before any application of herbicides, Brown, C. (2000) The Global Outlook for Future Wood Supply from nutrients or other soil additives. Supplementary criteria could be Forest Plantations. Working Paper No. GFPOS/WP/03. Forestry added in the analyses. For example, in this study only a buffer Policy and Planning Division, Food and Agriculture Organisation, Rome. zone 100 m wide was applied to the creek layer; additional buffer zones of decreasing risk at greater distances from the creek would Bubb, K. (2001) Simazine concentrations in soil, groundwater and produce a more refined assessment. This is especially the case stream water following application to Pinus plantations in the coastal lowlands of south-east Queensland. Australian Forestry for large areas, where the magnitude of streams should also be 64, 8–13. incorporated. For example, a region drained by a higher-order Bubb, K.A. and Croton, J.T. (2002) Effects on catchment water balance stream should have a higher risk associated with it than an area from the management of Pinus plantations on the coastal lowlands drained by an ephemeral low-order stream. of south-east Queensland, Australia. Hydrological Processes 16, 105–117. The second case study shows that natural hazards can be efficiently predicted using the MCE approach. In addition, the Calderon, M.J., Ortega, M., Hermosin, M.C., Garcia-Baudin, J. and Cornejo, J. (2004) Hexazinone and simazine dissipation in forestry comparison of our results with the real situation in the field field nurseries. Chemoshere 54, 1–8. showed that such models can also help to identify the main controls over a certain process. For example, more field mapping Carver, S. (1991) Integrating multi-criteria evaluation with geographic information systems. International Journal of Geographical should be directed to the detection of schistose rocks, which are Information Systems 5, 321–339. known to create problems in the area. Also important is the Ceballos-Silva, A. and Lopez-Blanco, J. (2003) Delineation of suitable determination of the dip of the strata in relation to the slope of areas for crops using a multi-criteria evaluation approach and land- the terrain, the monitoring of the water table and the characterisa- use/cover mapping: a case study in Central Mexico. Agricultural tion of the weathering profile, as these elements could help refine Systems 77, 117–136. the assessment of the local rocks in relation to vulnerability to Close, M.E., Pang, L., Watt, J.P.C. and Vincent, K.W. (1998) Leaching mass wasting. of picloram, atrazine and simazine through two New Zealand soils. Geoderma 84, 45–63 In conclusion, the MCE approach offers flexibility in design and shows substantial potential to assist general management by Chen, K., Blong, R. and Jacobson, C. (2001) MCE-RISK: integrating multicriteria evaluation and GIS for a risk decision-making in assessing the vulnerability of the forest landscape to either natural hazards. Environmental Modelling and Software 16, introduced chemicals or natural hazards. 387–397. Coops, N.C. and Waring, R.H. (2001) The use of multiscale remote Acknowledgements sensing imagery to derive regional estimates of forest growth capacity using 3-PGS. Remote Sensing of Environment 75, We thank the Queensland Department of Primary Industry– 324–334. Forestry for funding and Col Reugebrink, Peter Jessen, Tu Dragan, M., Feoli, E., Fernetti, M. and Zerihun, W. (2003) Application Nugyant, Adrian Knight and Ian Grayson for making data of a spatial decision support system (SDSS) to reduce soil erosion available, as well as providing technical advice and support. We in northern Ethiopia. Environmental Modelling and Software 18, acknowledge Mike Robinson’s constructive feedback, as well as 861–868. his input in formulating spatial problems. Len Cranfield Foster, F.A. and McDonald, A.T. (2000) Assessing pollution risks to (Geological Survey of Queensland) provided additional informa- water supply intakes using geographical information systems tion on the geological formations of the study area. (GIS). Environmental Modelling and Software 15, 225–734. Garmouma, M., Blanchard, M., Chesterikoff, A., Ansart, P. and References Chevreuil, M. (1997) Seasonal transport of herbicides (triazines and phenylureas) in a small stream draining an agricultural basin: ANZECC and ARMCANZ (Australian and New Zealand Environment Melarchez (France). Water Research 31, 1489–1503. and Conservation Council, and Agriculture and Resource Gkaraveli, A., Good, J.E.G. and Williams, J.H. (2004) Determining Management Council of Australia and New Zealand) (2000) priority areas for native woodland expansion and restoration in Australian and New Zealand Guidelines for Fresh and Marine Snowdonia National Park, Wales. Biological Conservation 115, Water Quality. Volume 1. The Guidelines. National Water 395–402. Management Strategy Paper No. 4. Environment Australia, Department of Environment and Heritage, Canberra.
Australian Forestry 2007 Vol. 70 No. 1 pp. 61–69 Micaela Preda, Kenneth A. Bubb and Malcolm E. Cox 69
Hajkowicz, S. (2002) Regional priority setting in Queensland: a multi- Pereira, J.M.C. and Duckstein, L. (1993) A multiple criteria decision- criteria evaluation framework. Consultancy report for the making approach to GIS-based land suitability evaluation. Queensland Department of Natural Resources and Mines. Folio International Journal of Geographical Information Systems 7, No. 02/1311, CSIRO Land and Water, Brisbane, CLW- 407–424. [email protected]. Pettit, C. and Pullar, D. (1999) An integrated tool based upon multiple- Jain, M.K., Kothyari, U.C. and Ranga Raju, K.G. (2004) A GIS based criteria evaluation of spatial information. Computers, Environment distribution rainfall-runoff model. Journal of Hydrology 299, and Urban Systems 23, 339–357. 107–135. Riano, D., Chuvieco, E., Condes, S., Gonzales-Matesanz, J. and Ustin, Jankowski, P. (1995) Integrating geographical information systems and S.L. (2004) Generation of crown bulk density for Pinus sylvestris multiple criteria decision-making methods. International Journal L. from lidar. Remote Sensing of Environment 92, 345–352. of Geographical Information Systems 9, 251–273. Ropke, B., Bach, M. and Frede, H-G. (2004) DRIPS — a DSS for Kangas, J., Store, R., Leskinen, P. and Mehtätalo, L. (2000) Improving estimating the input quantity of pesticides for German river basins. the quality of landscape ecological planning by utilising advanced Environmental Modelling and Software 19, 1021–1028. decision-support tools. Forest Ecology and Management 132, Rouget, M., Richardson, D.M. and Cowling, R.M. (2003) The current 157–171. configuration of protected areas in the Cape Floristic Region, Koesmarno, H.K. (1997) OFSIM: A simulation shell for forest South Africa — reservation bias and representation of biodiversity management. Environmental Modelling and Software 12, 75–83. patterns and processes. Biological Conservation 112, 129–145. Laaribi, A., Chevallier, J.J. and Martel, J.M. (1996) A spatial decision Selby, M.J. (1993) Hillslope Materials and Processes. 2nd edn. Oxford aid: a multicriterion evaluation approach. Computers, Environment University Press, Oxford. and Urban Systems 20, 351–366. Schumann, A.H. and Geyer, J. (1999) New ways of application of GIS Liu, F. and O’Connell, N. (2003) Movement of simazine in runoff water in hydrological modelling. In: Water 99 Joint Congress: 25th and weed control from citrus orchard as affected by reduced rate Hydrology and Water Resources Symposium, 2nd International of herbicide application. Bioresource Technology 86, 253–258. Conference on Water Resources and Environment Research. 6–8 Malczewski, J. (1996). A GIS-based approach to multiple criteria group July, Brisbane. Institution of Engineers Australia, Barton, ACT decision-making. International Journal of Geographical Sivertun, A. and Prange, L. (2003) Non-point critical area analysis in Information Systems 10, 955–971. the Gisselo watershed using GIS. Environmental Modelling and Malczewski, J. (1999) GIS and Multicriteria Decision Analysis. John Software 18, 887–898. Wiley and Sons, Inc., New York, 391 pp. Store, R. and Kangas, J. (2001) Integrating spatial multi-criteria Malczewski, J. (2004) GIS-based land-use suitability analysis: a critical evaluation and expert knowledge for GIS-based habitat suitability overview. Progress in Planning 62, 3–65. modelling. Landscape and Urban Planning 55, 79–93. Manning, J. (1992) Applied Principles of Hydrology. 2nd edn. Summerfield, M.A. (1991) Global Geomorphology: An Introduction Macmillan Publishing Company, New York, 294 pp. to the Study of Landforms. Longman Scientific and Technical, Wiley and Sons, Inc., New York. Meakins, N.C., Bubb, J.M. and Lester J.N. (1995) The mobility, partitioning and degradation of atrizine and simazine in the salt Thomas, B. and Piper, T. (2001) Persistence of Simazine in the Soil in marsh environment. Marine Pollution Bulletin 30, 812–819. Western Australia. Farmnote No. 85/2001. Department of Agriculture, Perth. www.agric.wa.gov.au. Nagashima, K., Sands, R., Whyte, A.G.D., Bilek E.M. and Nakagoshi, N. (2002) Regional landscape change as a consequence of Ticehurst, J.L., Cresswell, H.P. and Jakeman, A.J. (2003) Using a plantation forestry expansion: an example in the Nelson region, physically based model to conduct a sensitivity analysis of New Zealand. Forest Ecology and Management 163, 245–261. subsurface lateral flow in south-east Australia. Environmental Modelling and Software 18, 729–740. Nath, S.S., Bolte, J.P., Lindsay, G.R. and Aquillar-Manjarrez, J. (2000) Applications of geographical information systems (GIS) for spatial Tkach, R.J. and Simonovic, S.P. (1997) A new approach to multi-criteria decision support in aquaculture. Agricultural Engineering 23, decision making in water resources. Journal of Geographic 233–278. Information and Decision Analysis 1, 25–44. NAFI (National Association of Forest Industries) (2004). Forestry Vink, J.P.M. and Van der Zee, S.E.A.T.M. (1997) Pesticide Facts: An Overview. http://www.nafi.com.au/briefings/ biotransformation in surface waters: multivariate analyses of index.php3?brief=2. environmental factors at field sites. Water Resource 31, 2858–2868. Natural Resources, Mines and Energy (2004) Geoscience Data. DVD+R. Queensland geological mapping (polygonised vector) Wanielista, M., Kersten, R. and Eaglin, R. (1997) Hydrology: Water data — regional and 1:100 000 sheet areas. Geological Survey of Quantity and Quality Control. John Wiley and Sons Inc., New Queensland, Brisbane. York, 567 pp. Payn, T.W., Hill, R.B., Höck, B.K. and Skinner, M.F. (1999) Potential Ward, R.C. and Robinson, M. 1990. Principles of Hydrology. 3rd edn. for the use of GIS and spatial techniques as tools for monitoring McGraw-Hill Book Company, London. changes in forest productivity and nutrition, a New Zealand example. Forest Ecology and Management 122, 187–196.
Australian Forestry 2007 Vol. 70 No. 1 pp. 61–69 70 Book review
Book review
Forestry and Environmental Change: Socioeconomic and Political Dimensions
J.L. Innes, G.M. Hickey and H.F. Hoen (eds) Report No. 5 of the IUFRO Task Force on Environmental Change CABI Publishing in association with The International Union of Forestry Research Organizations (IUFRO), 1995, vii + 265 pages ISBN 085 1 990029 £49.95
This book aims to present theories and models that can assist in c is based on having ecological reserve zones, intensive production understanding the socioeconomic changes in the forest sector that zones, and multiple/extensive use zones. In Chapter 7, Csoka both influence and are influenced by environmental change in reviews changes in forest management, law and policies, forest the world’s forests. A wide range of authors present papers cover, and forest sector employment in countries in transition in discussing various tools and approaches for forest management, Europe. While providing a broad overview, the short space and and analysing socioeconomic change in some forestry regions. broad topic makes this a fairly general account. The contributors focus in particular on the use of economic tools Certification is covered in Chapter 8, with a brief history and such as discount rates, zoning approaches to managing forest land, description of the two major global forest certification schemes and spatial-modelling-based decision-making systems that can (Forest Stewardship Council, FSC, and Programme for the be used by forest managers. Endorsement of Forest Certification, PEFC), as well as of ISO The book provides a useful overview of current thinking in terms 14001 certification. Various individual schemes are also of tools used to help plan and decide on forest management, described. The chapter provides a useful introduction to the particularly in North America and Europe. Disappointingly, promise, problems and challenges of certification. however, the ‘socio’ part of the ‘socioeconomic’ in the book’s Chapter 9 examines the impacts of declaration of large-scale title is barely addressed. There is a strong focus on neoclassical nature reserves on the global forest products market. Perez-Garcia economic theory, planning approaches and tools rather than on reviews historical demand and supply and estimates future presenting other research to help explain how and why the changes in wood consumption, concluding that the establishment socioeconomic dimensions interact with environmental dimen- of nature reserves does not necessarily result in increased cost of sions of the world’s forests. supply for the forest products sector. In Chapter 10, Sedjo also Chapter 1 sets the scene by reviewing the changing physical, examines the establishment of large-scale nature reserves, economic, social, policy and cultural environments faced by those focusing on case studies in the USA and Finland. His analysis practising forestry worldwide in recent decades. In Chapter 2, concentrates on the economic impact of reserves, arguing that Dudley provides a useful review of the many different attempts actual economic effect is less than expected because of the currently underway to categorise forest quality, and the availability availability of alternative timber supplies. of data for different measures, but stops short of suggesting a In Chapter 11, Asan describes the ‘forest function’ criteria used way forward for measuring forest quality. as a basis for sustainable forest management (SFM) in Turkey. Chapters 3, 4 and 5 focus on economic theory. In Chapter 3, This is a multiple-use approach to forest management that uses Farmer and Randall engage in some reasonably technical spatial mapping to overlay forest functions and then determine discussions of the theories underlying benefit–cost analysis. They management strategies. Chapter 12 similarly uses a country case discuss the use of discounting of future returns when estimating study — Malaysia — to demonstrate a GIS-based approach to benefits from forestry, which can usefully inform the economic forest conservation planning. In this approach, criteria and analysis of returns from forestry versus shorter-term land uses. indicators for sustainable forest management are used as the basis In Chapter 4, Price also discusses discounting, raising a number of a multiple-criteria decision-making model that uses spatial of problems for practitioners to consider including issues mapping to identify and prioritise areas to be managed for surrounding social preferences for consuming sooner rather than different purposes. later; risk and physical change in a resource over time. In Chapter 5, Chapter 13 considers the social dimensions of forest change. Hofstad presents economic theories that aim to explain why Romm emphasises that the nature and conceptualisation of forests deforestation occurs. Game theory is then invoked to attempt to is defined as much by social as other forces, with shifting values further explain deforestation cycles. and globalisation affecting who has the power to influence forest The book then shifts to discussing various policy and legislative management. The implications of this shift, however, are not approaches to forest management. Nitschke and Innes review the drawn out in terms of how the forester can adapt to and meet the use of zoning as a forest management tool in different countries challenges of this new ‘no boundaries’ system of people and in Chapter 6, and promote the TRIAD approach in which zoning groups with influence and interest in forestry.
Australian Forestry 2007 Vol. 70 No. 1 pp. 70–71 Book review 71
Finally, in Chapter 14 the well-known limitations of linear- programming-based decision-making software are highlighted, and new approaches including fuzzy modelling, heuristic algorithms and spatial techniques are reviewed. The problems of hard systems approaches and the shift to participatory forest management are described, and the chapter explores how multi- criteria decision-making approaches can be adapted to assist in participatory forest management. Overall, the book provides an interesting snapshot of various approaches to forest planning and modelling of economic change in the forest sector, while not providing more diverse perspectives to challenge the neoclassical economic and technocratic views that underpin many of the contributions. If we are to truly understand socioeconomic change in the world’s forests, more attention should be given to the broad range of social science research being undertaken globally, which has a lot to say on why and how the world’s forests — and forest managers — are changing so rapidly.
J. Schirmer The Australian National University Email: [email protected]
Australian Forestry 2007 Vol. 70 No. 1 pp. 70–71 72 Referees
Referees 2006
The editors of Australian Forestry are pleased to acknowledge the important contributions of referees to the journal in 2006.
Penny Baalman Jerry Vanclay Eric Bachelard Ernst Kemmerer John Bain Simon Lawson Chris Beadle David Lee Phil Blakemore Jerry Leech Elaine Davison Phil Matthew David De Little John McGrath Bernie Dell Nick O’Brien Ross Dickson Charlma Phillips David Doley Martin Rayner Bob Eldridge Gary Richards Ken Eldridge Dave Smorfitt Jane Elek Peter Snowdon Kathleen Herbohn Mike Stukely Gary Hopewell Phil West Trevor Innes Ross Wylie
Australian Forestry 2007 Vol. 70 No. 1 p. 72 Australian Forestry
Notice to contributors Scope Australian Forestry provides a vehicle for formal publications in all areas of interest to the Institute of Foresters of Australia. The editors welcome original scientific and technical contributions, reviews, suitably argued opinions, short communications and reviews of books on topics related to forest research, forest management and forest policy in Australia and the South Pacific region. Papers relating to any aspect of knowledge or management of forest species and land use matters of interest to the region will be particularly valued. It is a condition of submission of original papers that they have not been accepted for publication elsewhere. The Editorial Panel evaluates the suitability of material for publication. All papers submitted to the journal will be reviewed by referees nominated by the Panel. Criteria for the assessment of papers centre on the suitability of the topic for publication in Australian Forestry, technical and scientific quality of the work, clarity and conciseness of presentation, appropriateness of the conclusions on the basis of the evidence presented, and suitability of the material for printing. Manuscripts Contributions should be printed with double spacing on one side of the paper only. Three copies of each paper should be submitted. Following review, please submit the revised manuscript as one hard copy and an electronic file or files. The latter should be in Word or RTF, and should include tables and illustrations. The latter should also be supplied in their original files (e.g. in Excel) or as photographs. Low-resolution files are not acceptable. Title The title should be brief and where possible begin with a word suitable for indexing. Summary The summary should follow immediately after the author’s address and should concisely describe the nature of the work, the principal results and conclusions. It should normally not exceed 5% of the length of the paper and should be suitable for reproduction by abstracting journals. If possible provide several keywords drawn from the CAB thesaurus; if this is not available it will be helpful to look at the keywords assigned to similar articles in Forestry Abstracts or in previous issues of this journal. Presentation Authors should follow the pattern of presentation used in this issue. Tables and figures should be numbered with arabic numerals. Photographs should be identified as figures, and numbered in a single sequence with line drawings. Photographic prints should be of high quality on gloss paper, with a full range of tones and good contrast. The preferred placement of tables and figures in the text should be marked with e.g. ‘Figure 1 near here’. Actual tables and figures should be gathered at the end of the paper. Illustrations Graphical material may be submitted in two forms: computer-generated graphics or line drawings. Diagrams should be provided in postscript (ps) or encapsulated postscript (eps) format if possible. If diagrams are supplied only as hard copy, they should be camera ready, with labels applied at an appropriate size, printed by high quality (e.g. laser) printer. Where possible, figures (including all labels) will be printed within a single column width. References References should be cited in the text by author and year of publication. Three or more authors should be referred to by the senior author et al. The reference list should cite the full title of the publication. Please follow the style used in recent issues of this journal. Submissions Please forward your submission to: Executive Director, The Institute of Foresters of Australia, PO Box 7002, Yarralumla ACT 2600, Australia. Copyright protection For the protection of authors and the IFA, a simple form of copyright agreement has been developed and is available on request from the Executive Director. A copy of the agreement will be sent to the senior author or his or her agent with the typeset proofs; it should be signed by the senior author or agent on behalf of all co-authors and returned to the office of the IFA. Under the terms of this agreement the IFA will obtain an interest in copyright only if the article is published by the Institute. Offprints A copy of the final paper in portable document format (pdf), and guidelines for the distribution of copies, will be provided to authors after publication of their paper. Correspondence on editorial matters should be addressed to the Managing Editor, Dr Brian Turner, C/- The Institute of Foresters of Australia.