New Forests: Wood Production and Environmental Services

Chapter 1

Introduction

Sadanandan Nambiar and Ian Ferguson

This book is about growing new forests in harsh environments for wood products and environmental services, the relevant multidisciplinary knowledge acquired over the last 15 years, and needs and priorities for the future. It builds on a smaller volume, ‘Restoring tree cover in the Murray–Darling Basin’, published earlier (Nambiar et al 2000).

In Australia during the last decade, increasing recognition of priorities in natural resource management has led to several key national initiatives. These initiatives have re-enunciated goals for the establishment of plantations and farm forests to provide economic and environmental benefits. The community has recognised that further expansion of traditional plantation forestry may have to be, in part, in new land base located in areas of low rainfall. At the same time, both the extent and the impact of salinity on agricultural production and water quality have become a matter of national concern. Dryland salinity, largely caused by large-scale clearing of woodlands and related agricultural practices, has degraded about 2.5 million hectares of agricultural land. The extent and severity of this threat to land and water is increasing rapidly.

The planting of trees and other perennial vegetation on a sufficient scale in strategic parts of the landscape provides one, but by no means the only, option for mitigating land and water salinity and improving productivity of the land. Other environmental services that can be provided by tree planting and similar revegetation include carbon sequestration, biological diversity, and aesthetic values at local and landscape levels.

Large-scale reforestation, like most activities that humans undertake, may also incur environmental disservices and other costs. Potential reductions in downstream water flows, and losses of agricultural production in the planted area, need to be weighed against the potential net benefits from environmental services and any commercial commodities — wood and an array of non-timber tree products — that can be produced.

The broad challenge is to ascertain where such tree planting might best be pursued, what species and technologies should be used for establishment and later management, how productivity can be improved, what mix of environmental services and commercial goods is optimal, and whether the likely net benefits justify the change in land use and the requisite investment.

While many regions of Australia face natural resource issues, water and soil management is brought to a focus in the Murray–Darling Basin. This vast and complex area of the Australian continent encompasses a wide range of soil and biophysical environments, and suffers from significant land and water degradation.

Characteristics of the Murray–Darling Basin which differ from those associated with commercial plantation development elsewhere in Australia include:

generally lower rainfall, and complex hydrology associated with salinity

more fragmented, and less favourable, locations relative to potential markets for commercial products

far greater emphasis on the environmental services that might be supplied

a need to mobilise joint public and private funding of the requisite level of investment.

These new forests to target landscapes affected by salinity will be established beyond the current boundaries of forest, on agricultural lands that generally have rainfall less than 650 mm year–1. Clearly a better understanding is needed of the productivity — the response to environment and management — of these new forests, planted on various scales and for multiple purposes. Forestry in such a low rainfall zone on sometimes degraded land is a new venture: we have no significant experience of developing extensive forestry in such a harsh environment.

The afforestation of the basin, on a scale sufficient to achieve catchment and regional change in the ecosystem, and on an economically sustainable basis, is a major task. The investment required is unlikely to be available from the traditional public purse, although government support and promotion can be catalytic. Investment is not easy to attract, because risks are perceived to be relatively high and returns are expected only some years in the future. Nevertheless, we need to begin, building on the research results now available and the modest activities already underway.

Some risks can be reduced by integrating existing knowledge of commercial plantations with new information about tree growing for environmental services and wood (or other commercial tree products) in lower-rainfall areas. These risks arise from the relatively harsh biophysical environment; inadequate information on the choice, long-term productivity and management of suitable species; inexperience of landowners in forestry, with attendant doubts about longterm sustainability; and uncertainties attached to the processing of wood and other products, and concerning access to markets. The need for sound information — on environmental impacts, productivity, management, and the prospects for products — is urgent.

In the late 1980s, discussions among research providers identified the lack of information concerning commercial tree growing as a means of ameliorating rising water-tables and salinity in the southern Murray–Darling Basin. Recognising that multidisciplinary research was essential and that no one research provider possessed all the desirable skills, a collaborative centre, the Trees for Profit Research Centre, was established in 1991.

The centre’s mission was to develop technology for commercial tree growing and associated wood processing for areas threatened by rising water-tables and salinity, and thus to achieve productive land use, optimise returns from commercial wood production integrated with agriculture, and control the adverse effects of salinity and rising groundwaters.

The centre’s research strategy spanned land and water care, tree growing and management, products and processing, economics and marketing, and technology transfer.

The geographic focus for the centre’s activities was the southern Murray–Darling Basin, with emphasis on the irrigation areas around Shepparton and Deniliquin, where a series of major plantation trials (pilot sites) were established for research and demonstration purposes. These sites attracted intense interest from landowners and from local and some wood companies, and stimulated farm forestry programs in Victoria and New South Wales.

The centre’s activities were complemented by other major initiatives by CSIRO in collaboration with relevant state and other agencies. The major ones include Wagga Wagga Effluent Plantation Project, Rejuvenating the Murray–Darling Basin with a Forest Products Industry, Trees for Saline Land, and Australian Low Rainfall Tree Improvement Group.

Results of these and other research activities, and the progressive application of the results through farm forestry over the last decade, provide the present basis for an approach to new plantation forests, integrating wood production and environmental services where increasing salinity is a major land degradation problem.

The wide range of sites involved in these experimental studies introduced a number of difficulties, including those of transport of personnel and resources, negotiating legal access to the land, and treatment and maintenance at the times and to the specifications desired. As with most fieldwork, there were successes and failures. Some of the latter occurred because we were breaking new ground in every sense, and the lessons from plantation development elsewhere did not necessarily apply in this new and harsher environment. These experiments and demonstrations were critical, however, for gaining an understanding of planting and growing eucalypts in the basin. Many of the results are applicable to other regions in Australia and elsewhere where land degradation and water deficits are common.

This book is not simply about growing commercial tree plantations primarily for wood production, and the net benefits therefrom. It is about new forests that can supply environmental services, such as salinity mitigation and carbon sequestration, in combination with commercial wood or non-timber products in an environment, beyond the boundaries of traditional forestry.

We have aimed to integrate the experience from many trial plantings and research with relevant information from elsewhere, and to disseminate our analysis of the problems and prospects of developing new forests in the Murray–Darling Basin to supply environmental services and wood. The experimental work that is the basis of the information presented here has been focused on the low-rainfall, agricultural, inland zone of the Murray–Darling Basin. This focus was adopted to make best use of the limited resources available. Where possible, however, the scope of most chapters has been expanded to synthesise generic information applicable to other regions in Australia and elsewhere.

We have attempted to provide a comprehensive account relevant to the development of these new forests, covering soils, the biophysical environment, water use and irrigation strategies, silviculture, pests and diseases, wood quality and products, and economic and policy implications.

No successful reforestation program is possible without a favourable balance between costs and returns. These returns may come through several pathways — the sale of wood or other commercial products and/or the provision of environmental services. There would be greater incentive for planting trees on a significant scale if the trees themselves provided direct economic returns to the grower in addition to environmental services (e.g. salinity mitigation). If this can be achieved, the new forests over time will increasingly provide a basis and a resource for business activities that sustain local economies and employment.

Reference

Nambiar EKS, Cromer R, Brown AG (Eds) (2000) ‘Restoring tree cover in the Murray–Darling Basin’. (CSIRO Forestry and Forest Products: Canberra)

Chapter 2

Environment, species selection and productivity

prediction

Trevor Booth

Abstract

Tree growers need to be able to select suitable trees (species, provenances, hybrids or clones) for new plantings in the Murray–Darling Basin and to be able to predict the likely productivity of sites under consideration for plantation development. This chapter outlines the biophysical environment of the basin, the process for selecting suitable trees, and methods for predicting potential productivity. Site evaluation charts and statistical methods for predicting potential productivity are described and then two process-based simulation models, ProMod and 3-PG, are outlined. The data required to run each model and its basic structure are described. Example outputs are provided and relevant applications discussed. Great progress has been made in tree growth modelling, particularly in the last 10 years. However, growth modelling will be most valuable when models can be calibrated and validated using information from an extensive network of mature trials across the targeted planting areas.

Introduction

There is increasing interest in growing trees in the Murray–Darling Basin, and greater focus is being placed on areas of low to medium rainfall (500–800 mm) than in the past (Nambiar et al. 2000; Consortium 2001). Some key questions for land managers, extension officers and growers considering plantation and farm forestry establishment in areas beyond those currently planted are:

Which tree (species, provenance, hybrid or clone) to grow?

Where will this particular tree grow?

How well will this particular tree grow on this site?

This chapter provides a brief overview of biophysical conditions within the Murray–Darling Basin, then describes methods to assist tree species selection and growth prediction in such environments. Various approaches to modelling tree growth are outlined, and examples of how such tools can be applied are described. This includes use of both broad-scale and site-specific predictions for assisting land management decisions. Finally, the need for further improvements in tree selection and growth prediction is considered.

Biophysical environment of the Murray–Darling Basin

Soils and climate are two of the most important factors influencing the suitability of specific sites for particular trees. The following two sections are adapted from Booth (2000).

Soils

Soils in the Murray–Darling Basin have many of the limitations commonly found in Australia. They are often low in nutrient status and commonly deficient in phosphorus and nitrogen. Deficiencies of nutrients such as zinc, iron, boron and copper also occur in some areas. Soils are often hard-setting, with poor water-infiltration characteristics. Salinity problems are increasing. In many areas, tree clearing and poor irrigation practices have caused watertables to rise. In dryland areas, tree removal, particularly in upland (recharge) areas, has also often led to rising watertables, particularly in lowland (discharge) areas. In both irrigated and dryland areas, rising groundwater has dissolved naturally-occurring salts and is rendering large areas unproductive. Areas south of the Murray River and in the eastern part of the basin have been particularly severely affected by dryland salinity. Wind erosion and water-induced erosion are also endemic problems. Profitable and sustainable farming systems can be established, but care is needed to recognise the limitations of particular soil types.

The characteristics and broad distribution of soils of the Murray–Darling Basin have been described in a CSIRO book, ‘Soils: an Australian viewpoint’ (Butler et al. 1983). More recently the distribution of soils has been related to a new Australian soil classification. Figure 2.1 (Plate 1, p. 105) shows the distribution of the 13 main soil orders across the Murray–Darling Basin. The following brief outlines of the main features of these soil orders are taken from a description of the classification by Isbell (1996).

Figure 2.1 (Plate 1, p. 105) is based on the Atlas of Australian Soils, which maps soils across Australia at a scale of 1:2 000 000. It is useful for showing the broad patterns of soil occurrence, but more detailed maps are needed for assessing the potential of different catchments for plantation forestry. The recently completed Murray–Darling Basin soil information strategy prepared by CSIRO has produced more detailed soil maps at scales of 1:100 000 and 1:250 000. The study used existing soil survey information, as well as digital elevation models. These were analysed to help estimate likely soil conditions where data were unavailable. A summary of the work is available on the Bureau of Rural Sciences web site (http://www.brs.gov.au/mdbsis) and detailed information has been made available in the ‘Basin-in-a-Box’ set of nine CD-ROMs (Murray–Darling Basin Commission 2002 or see http://www.georeality.com).

Climate

There is a general tendency for both high-pressure and low-pressure weather systems to move from west to east across southern areas of the Australian continent. Rainfall patterns in the Murray–Darling Basin are influenced by these movements and particularly their interactions with higher-elevation areas in the eastern part of the basin. Figure 2.2 shows mean annual rainfall patterns across the basin. Annual rainfall in the western parts of the basin is generally 200 mm or less and increases in the east to 700 mm or more.

Rainfall seasonality is an important influence on tree growth. Areas north of Bourke generally have pronounced summer rainfall with rainfall in the November–April period being at least 30% greater than rainfall in the May–October period. On the other hand, areas in the Murray– Darling Basin south of Mildura generally receive pronounced winter rainfall with at least 30% greater rainfall in May–October than in November–April.

The seasonality of rainfall interacts with rainfall intensity, evaporation rates and soil characteristics to determine how much water is available to plants. Annual evaporation rates recorded from open pans exceed annual rainfall over most of the Murray–Darling Basin except in some limited parts of the Eastern Highlands. For example, evaporation rates are typically more than 1500 mm per year in western areas of the basin where annual rainfall is less than 600 mm (Figure 2.2).

Annual mean temperature patterns across the basin are shown in Figure 2.3, with temperatures being mainly affected by latitude and elevation. North-western areas of the Murray– Darling Basin average annual mean temperatures around 20°C in contrast to areas of the basin to the south and east, which average about 14°C. These differences in average temperatures over the whole year are also reflected in extremes. For example, average daily maximum temperatures in the hottest month (January) are over 35˚C in Bourke, but less than 28˚C in Canberra. Similarly, the number of days of frost is less than 100 over most of the north-west area, but may be over 200 in high-elevation areas of the south-east. The north-west areas of the basin also receive more solar radiation (about 20 MJ m–2 day–1) than areas to the south-east (about 17 MJ m–2 day–1).

Figure 2.2 Mean annual rainfall patterns across the Murray–Darling Basin. Rainfall data from the ANU’s ESOCLIM package, and Murray–Darling Basin boundary data courtesy of AUSLIG

Average values of climatic parameters, such as those shown in Figures 2.2 and 2.3, convey little about the enormous climatic variability that is common in the Murray–Darling Basin. Most people are now familiar with the influence of ‘El Nino’ events on rainfall in Australia. Warming of the ocean off Peru in South America affects air circulation patterns over the Pacific Ocean, reducing rainfall in Australia. Severe El Nino droughts were experienced in Australia in 1982–83, 1997–98 and 2002–04. The Bureau of Rural Sciences analysed historical records of rainfall across Australia and produced a Rainfall Reliability Wizard system that can predict the chance of obtaining certain amounts of rainfall for a particular month or season at a selected location (see Bureau of Rural Sciences web site, http://www.brs.gov.au/rainfall).

It is suggested that increases in greenhouse gases, such as carbon dioxide, will influence temperature and rainfall patterns across Australia. For example, in inland areas, mean annual temperatures may rise by up to 2.2˚C by the year 2030 and by up to 6.7˚C by the year 2070 (CSIRO 2001; Pittock 2003). Changes in rainfall patterns are more difficult to predict and are likely to vary in different seasons and regions.

Predicting the effect of these changes on tree growth is a complex task, as plant growth will be influenced not only by temperature and rainfall changes but also by the increase in carbon dioxide in the atmosphere and its potential effect on water-use efficiency. Plants take in carbon dioxide for photosynthesis but simultaneously lose water from pores in their leaves. As there will be more carbon dioxide in the atmosphere, the plants will not need to open these pores so widely and so will lose less water. As a result they may grow more with the same amount of rainfall (Pittock 2003).

Figure 2.3 Mean annual temperature pattern in the Murray–Darling Basin

Species selection

Whatever future climatic conditions are experienced, the first question to consider when planting trees is ‘What is their purpose?’ Trees may be planted, for example, to produce commercial timber or pulp, to mitigate problems of salinity and waterlogging, to provide shade and shelter, or to restore biodiversity; they may also be planted for aesthetic reasons.

When the purpose or purposes of plantation establishment have been decided it is useful to seek information about species already growing in the targeted area. The Plantation Information Network of the Bureau of Rural Sciences (see web site http://data.brs.gov.au/mapserv/plant/index.phtml) provides information on species used in existing major plantations in national plantation inventory regions, such as the Murray Valley and Central Tablelands of New South Wales. Private forestry development committees have been established in significant plantation regions, including areas within the basin such as Murray Riverina and the Central Tablelands. The aims of these committees are ‘to enhance the economic development potential of their region through increasing the commercial plantation estate to allow industry to continue to expand and remain world competitive. The committees have produced reports describing the potential for major species such as P. radiata (radiata pine) and E. globulus (blue gum) (see, for example, a report on north-eastern Victoria at web site http://www.plantationsnorth-east.com.au/). They have also established or are aware of local trials and plantings of lesserknown species (see, for example, Kyeamba Valley Landcare Group and Murray Riverina Farm Forestry 2003).

Jovanovic and Booth (2002) have analysed the climatic environments of 27 species suitable for farm forestry, and mapped areas likely to be climatically suitable throughout Australia. The full report is downloadable from the Rural Industries Research and Development Corporation web site (http://www.rirdc.gov.au/reports/AFT/02-095.pdf). Species considered include wellknown species such as P. radiata and E. globulus as well as lesser-known species such as E. argophloia (Chinchilla or western white gum). Climatic factors are particularly useful for showing broad patterns of suitability, as climatic conditions can be easily estimated for any location in Australia (McMahon et al. 1996).

Figure 2.4 (a) Natural distribution of Eucalyptus cladocalyx (b) Climatically suitable areas for E. cladocalyx. From Jovanovic and Booth (2002), reproduced courtesy of the Joint Venture Agroforestry Program

Climatically suitable areas for particular species are identified according to the following six factors: mean annual rainfall (mm); rainfall regime (summer, uniform or winter); dry season length (months); mean maximum temperature of hottest month (°C); mean minimum temperature of coldest month (°C); and mean annual temperature (°C). For example, Figure 2.4 shows the natural distribution and climatically suitable areas for planting E. cladocalyx (sugar gum). This species can produce furniture-grade timber and has potential in areas with 400–700 mm annual rainfall.

Though the report by Jovanovic and Booth (2002) provides a useful overview of suitable areas, the information needs to be treated with care and should be reviewed in relation to local soil and topographic conditions. For example, local topography may result in moderate frosts that may make a particular site unsuitable for E. cladocalyx. Climatic information can assist in identifying potentially suitable species, but soil requirements must also be considered. Information on the soil requirements of many species can be found in the CABI Forestry Compendium CD-ROM (CAB International 2003). For example, the description for E. cladocalyx indicates that it thrives on a wide variety of soils — gravels, clay loams, sandy loams, and sands — although it grows poorly on very fine sandy soils. It tolerates shallow soils, has a wide pH range and has been grown successfully on soils with a high lime content. Together, climate and soil information provide a basis for selecting species for field trials and testing.

Predicting productivity

Site evaluation charts and statistical analyses

In addition to the importance of species selection, estimating likely growth rates is critical to determine whether planting a particular species in a specific location will be worthwhile. For as long as tree plantations have been established in Australia, there has been interest in developing methods to predict how well trees will grow on potential plantation sites. For example, the Green Triangle area in south-eastern South Australia and western Victoria is now the major centre for plantation forestry in Australia. As knowledge of the requirements of P. radiata in south-eastern South Australia increased during the 1940s and 1950s, detailed site evaluation systems were developed that took into account not only climate, particularly rainfall, but also native vegetation, including both tree and principal understorey species, as well as major soil types. Using these factors, site evaluation charts were developed. These indicated whether a particular site was suitable for plantation establishment and whether it required additional management inputs (e.g. fertilisation), and also predicted likely growth rates in terms of seven site quality classes (Lewis et al. 1976). Site quality indices provided a valuable baseline for appropriate management practices including thinning and fertiliser application. However, site evaluation charts and site quality indices can only be constructed when there is already considerable experience of planting a particular tree in a specific area. Using natural vegetation as a biological indicator of site productivity is now of little use as plantations are established on previously cleared agricultural land. Site evaluation charts also provide no indication of the reliability of the predictive relationships.

Accordingly, statistical analyses were developed to relate tree growth to more detailed measures of environmental conditions. For example, Turvey et al. (1990) described how the productivity of P. radiata at 181 locations in New South Wales, South Australia, Victoria and Tasmania could be related to a soil classification system. Part of this classification was based on somewhat similar factors to those used to create Figure 2.1 (Plate 1, p. 105), such as profile characteristics, though it also considered soil factors particularly important for tree growth, such as depth to an impeding layer. A multiple regression model accounted for 75% of the variation in wood volume. Inions (1992) related the growth of E. globulus at 56 plots in Western Australia to site, soil chemical, soil physical and climatic conditions; and accounted for about 75% of the variance in site index. Another study used tree-growth and environmental data collected at 189 plots at more than 90 locations in south-eastern mainland Australia (Booth et al. in press). It analysed the growth of E. grandis (rose or flooded gum), E. globulus, E. camaldulensis (river red gum) E. saligna (Sydney blue gum) and Corymbia maculata (spotted gum) and related tree growth to species, climate and soil factors. The model, which accounted for about 65% of the variance in site index at 10 years, included four dummy variables (to account for differences between species compared to E. grandis, which was used as the control); two climatic variables; four growth index variables generated by the GROWEST model (see section below); and three soil variables.

When statistical methods were first used for growth prediction, researchers used simple variables such as mean annual rainfall. As the work progressed, factors were included that reflect environmental influences in more realistic ways. For example, instead of using just mean annual rainfall, mean annual rainfall divided by evaporation may provide an improved indication of tree growth, and a monthly water balance calculation may produce an even better result (see, for example, McAlpine 1970). However, there is a limit to the realism that can be achieved within statistical relationships. Computer simulation models use more complex calculations to mimic environmental effects on plant growth more effectively. Such models are often known as ‘process-based models’ as they attempt to simulate the processes controlling growth.

Forest growth models

The following sections describe three simulation models that have been used in Australia to simulate forest growth. From simple beginnings in the 1970s using generic plant growth models such as GROWEST, models have been developed specifically to simulate tree growth. The ProMod and 3-PG models described here represent two process-based models that have been widely applied to simulate forest growth in Australia. Both are readily available to users and continue to be developed and improved.

GROWEST

The GROWEST model (Fitzpatrick and Nix 1970) is a simple plant growth simulation model that formed a precursor to later growth models for trees. Plant growth is related to three climatic factors: temperature, solar radiation, and moisture. The temperature index is defined by three cardinal temperatures: a minimum, below which the plant will not grow; an optimum, at which plant growth is greatest; and a maximum, above which the plant will not grow. The solar radiation index increases from zero up to a maximum, above which increasing radiation no longer increases the index. The moisture index is based on a simple water-balance equation, which calculates the amount of water available in the soil each week as a result of rainfall and evapotranspiration (McAlpine 1970). GROWEST calculates a weekly growth index by multiplying the temperature, solar radiation and moisture indices together. The growth index accounted for variations in tree growth across 18 P. radiata sites in south-eastern Australia much better (r=0.74) than simply using mean annual rainfall (r=0.47) (Booth and Saunders 1980), and the model was used to estimate the climatic suitability of hundreds of locations across south-eastern Australia.

Though GROWEST provided a useful index for predicting tree growth, it is a general plant growth model and was not designed to simulate features important for commercial forestry, such as stem volume growth. The ProMod (Battaglia and Sands 1997; Sands et al. 2000) and 3-PG (Landsberg and Waring 1997; Landsberg et al. 2001) models were designed specifically to simulate tree growth. Both were designed from the outset to be run using simple, readily available data and to be suitable for assisting management decisions, such as land evaluation for plantation development.

ProMod

ProMod can estimate the potential of plantations to grow on cleared agricultural land in terms of net primary productivity, which is the total amount of biomass produced over a given time (i.e. gross primary production minus carbon lost through respiration). ProMod converts this into measures commonly used by forest managers, such as site index or peak mean annual increment. Site index is related to predominant tree height at a standard age (usually 20 years), whilst mean annual increment is generally measured in m³ ha–1 year–1 and is the average annual increase in volume of individual trees or stands up to a nominated time. Peak mean annual increment is reached at the age when mean annual increment has its maximum value on a specific site at a given stocking. The standard ProMod model does not predict the development of the stand over time, though Battaglia et al. (1999) have described a hybrid version that can simulate stand development. Battaglia and Sands (1997) and Sands et al. (2000) have described ProMod in detail. The following description outlines some of the main features of the model in relatively simple terms.

ProMod is run using monthly climatic data. The monthly rainfall is randomly distributed within each month, and internal calculations are made on a daily basis. Results are averaged over several random distributions of daily rainfall. If the user has information on the latitude, longitude and elevation of a site, monthly mean climatic data can be estimated for anywhere in Australia using the ESOCLIM program within the ANUCLIM package (McMahon et al. 1996, see Australian National University’s Centre for Resource and Environmental Studies web site http://cres.anu.edu.au/outputs/anuclim/doc/Contents.html). If necessary, data on slope and aspect can be used to make corrections for solar radiation received by the site before the data are entered into ProMod. Data on soil depth, texture, stoniness, and drainage and a simple rating of soil fertility are also required.

ProMod includes four modules: canopy development, biomass production, soil water balance, and site quality estimation. Photosynthetically-active radiation is the ultimate limit to plant growth at all sites around the world, and