SEA Working Paper 01/03

An economic assessment of the role of commercial tree crops to achieve greenhouse gas neutrality in predominantly grazing systems of south-western Australia

Liz Petersen^a, Steven Schilizzi^b and David Bennett^c

^aAgriculture Western Australia, South Perth, WA
^bAgricultural and Resource Economics, Faculty of Agriculture, University of Western Australia, Nedlands, WA
^cNRMC Pty Ltd, Fremantle, WA

Abstract

The accreditation of tree crops as carbon sinks under the Kyoto Protocol is a contentious issue. The findings of this study show that in the presence of greenhouse gas emission restrictions, the accreditation of tree crops can allow predominantly grazing systems of south-western Australia to remain profitable where the farms would otherwise fail. It is argued that a Protocol that encourages tree planting is more likely to be successful, has other benefits such as salinity abatement, and is more likely to encourage greater innovation of green technologies than one that disallows such accreditation.

1. Introduction

There is little doubt that atmospheric concentrations of greenhouse gases have increased markedly due to human activity; such as the burning of fossil fuels, land clearing and agricultural production. These increases in atmospheric greenhouse gases are having a discernible impact on climatic conditions especially global warming (IPCC 2001). Concern for this human induced climate change and the realisation that addressing the issue requires an international co-operative effort lead to the establishment of the United Nations Framework Convention for Climate Change. One of the most significant initiatives of the parties to the Convention is the Kyoto Protocol, an agreement that requires ratifying countries to restrict emissions to a specified percentage of 1990 emissions (UNFCCC 1997). Australia is a signatory to, but has not ratified, the Protocol. If ratified, Australia will have to restrict emissions to 108 percent of 1990 levels in the first commitment period of 2008 to 2012. Most Australian emissions have their source in the burning of fossil fuels (53 percent in 1990, 55 percent in 1996); however, agriculture is the second biggest contributor (16 percent in 1990, 20 percent in 1996). Ruminant livestock are the greatest source of emissions in the agricultural sector, contributing 70 percent of agricultural emissions¹.

This paper aims to contribute to the environmental policy debate by assessing the viability of commercial trees for greenhouse gas abatement on mixed cropping enterprises in Western Australia. The analysis specifically focuses on the Great Southern region of Western Australia, a region with relatively high-rainfall compared with other agricultural regions in the state. Petersen et al. (2001) found that, in the absence of carbon sinks, there are few economically feasible management options for greenhouse gas abatement in this region due to the dependence of the system on ruminant livestock (typically 85 percent of the system is grazed). Any abatement policy would rapidly cause the present system to become unprofitable unless swift technological change provided alternative enterprises or reduced emission levels in current practices.

The introduction of commercial trees is considered to be an example of such a technological change; however, the use of forests and other plantations as carbon sinks under the Kyoto Protocol is a contentious issue (as the Sixth Conference of the Parties to the United Nations Framework Convention on Climate Change at The Hague (COP6) demonstrated). The properties of trees as a carbon sink are not yet well understood and are difficult to define and measure. In addition certain signatories to the Protocol are resisting the inclusion of trees as sinks on the grounds that it allows parties to continue the burning of fossil fuels at a higher rate.

A criticism of using trees as a carbon sink is that the effect is only temporary. Carbon will be released into the atmosphere at harvest and with the treatment of the harvested timber. Even if the trees are not harvested they will eventually die and return the sequestered CO₂ back into the atmosphere. However, it is also argued that if tree crops or Landcare plantings are maintained by replanting, the carbon pool will continue to be maintained, although the rate of increase in the carbon pool will lessen (Shea 1999). Furthermore, tree crops could permanently contribute to reductions in atmospheric CO₂ levels if they are used to generate energy, replacing fossil fuels that are the main contributors to greenhouse gas levels, and to replace products that take a lot of energy to produce, i.e. replacing steel, aluminium and cement with wood products. However, irrespective of the time period for which the CO₂ is sequestered, it is generally argued that accrediting tree crops as carbon sinks will ‘buy time’ so that new and more efficient technologies (such as alternative energy sources) can be developed (Shea 1999).

An external benefit of reforestation is the reduction in the extent of dryland salinity. Along with biodiversity preservation, dryland salinity is the most important environmental problem facing Western Australia and has received much public attention (WA Government 1998). Additionally, while it already has salinity problems, Campbell et al. (2000) quote the Great Southern region as having the greatest potential for the highest proportion of land affected by salinity in the state. Secondary dryland salinity is directly linked to Australia’s history of land clearing that started with European settlement. It is caused by the increase of recharge into the soil profile causing the water table to rise and eventually to intercept the soil surface bringing salt with it (Wood 1925). The impact of reforestation on recharge abatement in the Great Southern is also investigated in this paper as an extra benefit derived from reforestation to balance greenhouse emissions.

The paper proceeds as follows. Section 2 is a presentation of the methodology of the study giving a brief discussion of the tree crop species analysed and the modelling techniques used. Results are presented with discussion in Section 3. The final section draws some conclusions.

2. Methods

The instrument of analysis is a linear programming model of a steady-state single-period representation of a farming system. Named MIDAS (Model of an Integrated Dryland Agricultural System), the model was originally developed for the Merredin region of Western Australia, but has since been calibrated for several other regions. MIDAS includes the relevant biological complexities and interactions between enterprises in a typical wool or wheat belt farming system by employing a whole-farm modelling framework (Pannell 1996). MIDAS has been used to analyse issues concerning greenhouse gas abatement (Petersen et al. 2001), farm management (Schmidt and Pannell 1996), agricultural policy (Morrison and Young 1991) and research (Pannell 1999).

The Great Southern was chosen as, due to its soil-climate constraints, it supports a predominantly grazing farming system. Ruminant livestock contribute far more greenhouse gases than crops, especially non-irrigated crops. This region is approximately one million hectares in size with approximately 1000 farms of an average of 1100 hectares (ABS 1997). Readers are referred to Morrison and Young (1991) and Young (1995) for detailed expositions of the nature and structure of the Great Southern MIDAS model (GSM) (which excludes accounting for greenhouse gas emissions). Petersen et al. (2001) presents a detailed description of the developments made to the standard version of GSM to model include greenhouse gas emissions. A brief description of GSM and the modelling of greenhouse gas emissions is presented in Section 2.1, and a detailed discussion of the inclusion of tree crops in GSM is documented in Section 2.2.

2.1 Brief description of the Great Southern MIDAS

The Great Southern region of Western Australia typically has a Mediterranean climate with the majority of the annual rainfall (approximately 500 – 600mm) falling between April and the beginning of November. The production enterprises included in GSM are livestock (sheep) and crops (cereals, lupins, field peas and canola) with an average of 15 percent of land cropped in both the standard solution and observed field data. The model farm is highly mechanised which represents the nature of the Great Southern farming systems.

2.1.1 Soil types

The soil types are modelled in five land management units (see Table 1). The LMUs display a range of fertility with the saline (LMU1) and waterlogged (LMU2) soils (25% of farm area) being the least fertile, and the sandy gravels (LMU4) being the most fertile (50% of farm area). Rotational options for the LMUs are presented in Table 2. LMUs 1 and 2 (25 percent of land) are generally not suitable for cropping and, although there are allowances in GSM, cropping on these soils is only rarely selected. Canola production is only suitable on the heavier soils LMUs 4 and 5 (70% of farm area). To increase the model’s accuracy as a representation of reality, a number of interdependencies are represented in GSM. The three main interdependencies are the rotational benefits between phases in a rotation; the grazing of stubble by sheep; and the subsequent grazing of remnant grain in the paddock after harvest

.

Table 1: GSM soil types

Table 2: Rotational options in GSM

*P=Pasture, C=Cereal, L=Lupin, S=Fodder Crop, N=Canola

2.1.2 Greenhouse gas emissions

Greenhouse gases are assumed to have four main sources: sheep in the form of methane (CH₄), nitrogenous fertiliser application in the form of nitrous oxide (N₂O), fuel use in the form of carbon dioxide (CO₂) and stubble burning (which creates a range of greenhouse compounds). All of these emission are modelled according to the National Greenhouse Gas Inventory (NGGI) published by the Australian Greenhouse Office (NGGI 1998a; NGGI 1998b; NGGI 1998c; NGGI 1998d). Emissions are converted to carbon dioxide equivalents (CO₂-e) through multiplication by their average global warming potentials. These relative potentials are presented in Table 3. For a detailed exposition on the modelling of greenhouse gas emissions in the GSM the reader is referred to Petersen et al. (2001).

Table 3: Global-warming potential of greenhouse gasses relative to carbon dioxide

*Non-Methane Volatile Organic Compounds
Source: (AGO 1999)

2.1.3 Model characteristics

The objective of the model is to maximise farm profit where profit is defined to be net return to capital and management. It equates to residual income from production receipts after depreciation, operating overheads and opportunity costs have been deducted (the latter associated with farm assets exclusive of land). GSM is based solely on expected values (the first moment of the probability distribution) and therefore assumes risk-neutral decision-making. Model output indicates optimal (that is, profit-maximising) enterprise or rotational activity levels given CO₂-e emissions from each source.

2.2 Modelling of commercial tree crops

Two species of commercial trees are considered to be suitable for the Great Southern region of Western Australia. Pinus Pinaster, or maritime pine, is native to Portugal but has been grown in Australia for more than 80 years. In the Great Southern 1858 hectares have been planted to date with a state-wide target of 150,000 hectares over 10 years from 1996 (Grzyb et al. 2000). Maritime pines are grown for commercial timber production for a variety of wood products including structural-grade saw-logs and timber for medium density fibreboard, veneer, posts, poles, furniture and paper (Shea et al. 1998). The second species of tree crop suitable for the Great Southern region is oil mallee. Oil mallees are a group of Eucalyptus species with high oil content in their leaves that grow in the mallee form (many trunks forming a spreading habit). Unlike maritime pines, oil mallees are native to Australia, being present in the original native bushland. Oil mallees have potential as a short rotation tree crop producing eucalyptus oil for the pharmaceutical industry as inhalants, soaps, gargles, lozenges, perfumery and disinfectants; and for the industrial industry as solvents and hand cleaners (Boland 1991; Barton and Knight 1997). The above ground biomass also has potential as a fuel for electricity production and for the manufacture of charcoal and activated carbon (Shea et al. 1998, RIRDC 1999). The mallees are harvested by cutting them off a few centimetres above the ground (Abbott 1989). The trees re-sprout from underground lignotubers using energy stored in these tubers (Canadell and Lopez-Soria 1998).

Commercial plantings of oil mallee started in 1994 and more than 2.64 million trees have been planted to 1999 in the Great Southern. 1.5 million were to be planted in 2000 and 2.5 million are targeted for planting in 2001 (Grzyb et al. 2000). As yet the oil mallee industry is in its infancy with a 20,000 tonne demonstration-scale plant in Narrogin expected to consume most of the established mallees within 400 kilometres of the plant over the next few years. Upon successful completion of this demonstration, the plant is expected to expand to 100,000 tonne capacity in the subsequent five years, and up to nine full-scale plants (100,000 tonne capacity) are likely to be built in the low to medium rainfall agricultural region of south-western Australia (RIRDC 1999).

Maritime pines and oil mallees were included in GSM for this analysis as a source of income from their timber and eucalyptus oil products, and as a sink for greenhouse gases. Maritime pines require sandy, free draining soil (Shea et al. 1998). Hence, their use is limited to LMU3 (deep sands) that constitutes just five percent of GSM. Oil mallees perform poorly on saline or waterlogged soils; hence they are modelled on LMUs 3, 4 and 5 in GSM. The productivity of the mallees is assumed to be equivalent on each LMU with the assumption that different species will be grown on each soil type for maximum productivity. This assumption is realistic given that oil mallees are native and hence, well adapted to the soils and climate of the Great Southern region (Cooper 1999). Further, Wildy (2000) found that soil nutrient level had no effect on oil mallee growth rates in the Western Australian wheat belt, giving weight to this argument.

2.2.1 Economics and carbon sequestration rates

Establishment costs of the trees are listed in Table 4. It is assumed that the farmer uses available farm labour to establish the trees, hence, no contracting labour requirement is included². Oil mallees have almost double the establishment costs of maritime pines due to the higher cost of purchasing seedlings (27c for oil mallees compared with 21c for maritime pines), a greater planting density (2667 seedlings per hectare for oil mallees compared with 1600 seedlings per hectare for maritime pines), and greater weed control costs. However, oil mallees have lower pest management requirements than the pines (Abbott 1989).

Table 4: Establishment costs ($/ha)

* Assuming 1600 seedlings/ha at 20.5c per seedling.
** Assuming 2667 seedlings/ha at 27c per seedling.

A 30-year rotation is assumed for both plantations. In reality, both tree species can keep producing timber and oil indefinitely (although pines need replanting after clearfell). However, for modelling purposes, the time frame needs to be specified. Thirty years is often cited as a practical rotation length for maritime pines (e.g. Shea et al. 1998; CALM 1999). When modelling an enterprise where most benefits and costs are accrued in the future in a single-period model such as MIDAS, certain assumptions need to be applied. Firstly, it is assumed that the annuity is received at the end of the year, so that no interest on these benefits is received throughout the year. Secondly, future benefits need to be discounted by the real interest rate a farmer would face. The discount rate is equated to the real interest rate. The maintenance costs of both tree crops presented in Table 5 are in nominal terms. Like the establishment costs, oil mallee maintenance costs are greater than that of the maritime pines. This is due to a greater cost of replanting if there is failure of establishment (which is simply a function of a greater seeding density) and greater weed control requirements. Maritime pines require fertilisation after each harvest, which is not required for oil mallees (Abbott 1989; Specht 1996; Weiss 1997).

Table 5: Nominal maintenance costs ($/ha)

* Five percent of establishment cost.

The timber harvested from the maritime pines and the price received for the timber is presented in Table 6. In years 12, 18 and 24, the plantation is thinned to get harvest yields outlined. In year 30 the maritime pine forest is clear-felled³. A total of 360 m³/ha is harvested over the 30-year rotation that equates to approximately 375 tonnes/ha of timber⁴.

Table 6: Timber harvested for maritime pine (m³/ha)

Source: (CALM 1999; Ritson 2000)

As mentioned previously, the biology and silviculture of oil mallees is quite different to that of maritime pines. The typical proposed rotation is to harvest first in year six and then every three years subsequently. Hence, in the 30-year life of the plantation, the crop will be harvested nine times. The average above ground biomass is 25kg/ha (Herbert 2000). Each subsequent triennial coppice is the same as the first harvest. The mallees are harvested by cutting them off a few centimetres above the ground and they re-sprout from underground lignotubers. Assuming a seeding rate of 2667 trees/ha, 9 harvests in the 30-year rotation and 25kg/tree, total biomass harvested is 600 tonnes/ha. The gross price received for the oil mallees is $30/t over the 30 years. However, transport, harvest and harvest coordination costs are approximately $15/t, hence the net price received is $15/t (Herbert 2000). This $15/t is received for all above ground biomass.

Financial returns of the two tree species are presented in Table 7. As was discussed earlier, the costs (establishment and maintenance) for oil mallees are greater than those for the maritime pines. However, the total returns for the oil mallees are also greater than those of the maritime pines, so much so that the net present value (with a 6% discount rate) of the oil mallees is over double that of the pines. This is largely due to the discounting process (or rather, to interest rates). In the absence of discounting, the present value of the oil mallees is $6299/ha and of the maritime pines is $7007/ha. Benefits accrued from thinnings of maritime pines are not received until year 12 and the great bulk of benefits from clear-felling the plantation are not received for 30 years. Hence, these benefits are discounted heavily. On the other-hand, benefits from oil mallees accrue regularly starting in year 6 and hence, are not as heavily discounted.

Table 7: Financial returns

*The annuity can be equated with the gross margin of each enterprise.

The annuity presented in Table 7 is the present value of the benefits divided by the length of the rotation, and represents the average benefit received in each year. This is regarded as the annual gross margin of the tree enterprises. The gross margin of the sheep enterprise is typically between $70/ha and $200/ha depending on the class of sheep, wool and meat prices, grazing rotation and stocking rate. The main class of sheep is ewes sold at five years of age, which have a gross margin of approximately $110/ha. Hence, the annuity of the oil mallees is larger than the gross margin of the majority of the sheep enterprises, the main production activity of the GSM, and will be selected in preference to the sheep activities in the optimal (profit maximising) solution. On the other hand, the maritime pine annuity is smaller than the gross margin of both the sheep and oil mallee enterprises and, as a consequence, will not be selected in the optimal solution.

The last of the financial return statistics, the internal rate of return (IRR), is the interest rate received for an investment. Hence, the interest rate received for money invested in maritime pines is 3.2 percent and oil mallees is 7.6 percent. It is assumed that no carbon is emitted at harvest in this analysis and that treatment of harvested timber is not the responsibility of the grower. All subsequent results should be viewed in the light of this.

Carbon sequestration rates for maritime pines are presented in Table 8 (Ritson 2000). Average gross carbon sequestration rates are presented for three components of the plantation:

1. Trees, the component of the plantation left standing after a thinning (although this is removed at clear-fell);

2. Residues, the component of the forest which falls from the tree or is removed during the thinnings (i.e. small branches, leaf litter); and

3. Timber, biomass that is removed from the plantation for commercial purposes.

Table 8: Carbon sequestration rates for maritime pines up to clearfelling at 30 years (t C/ha/year)

Because research into carbon sequestration by plantations is still in its infancy, no data are available at present for oil mallees. Hence, sequestration rates were estimated through use of the following assumptions:

1. Aboveground biomass production is a sigmoid function resulting in 25kg/tree at first harvest (year 6);

2. Aboveground biomass in the first year after is 6kg/tree, the second year after coppice is 15 kg/tree, and the third year after coppice is 25kg/tree prior to the next coppice (McCarthy 2000);

3. Dry weight of the aboveground biomass is 50 percent of the wet weight (Shea 1999);

4. Carbon weight of the biomass is 50 percent of the dry weight (Hassall and Assoc 1996); and

5. Belowground biomass is 20 percent of aboveground biomass (Hassall and Assoc 1996).

Given these assumptions, and converting carbon weight to carbon dioxide weight⁵, the carbon dioxide sequestration rates for maritime pines and oil mallees (including above and below ground biomass) are presented in Table 9. These rates are low compared with those quoted in the literature⁶. However, it is considered safer to err on the side of caution given the uncertainty still surrounding the properties of tree crops as carbon sinks.

Table 9: Carbon dioxide sequestration rates for maritime pines and oil mallees

An interesting trade-off arises between the benefits of oil mallees versus maritime pines with the introduction of these sequestration rates. In the absence of accreditation of plantations as a sink, oil mallees have the greater investment value (see Table 7). However, maritime pines have a six-fold larger capacity to sequester carbon dioxide. In the following results section it will become apparent whether this sequestration capacity adds enough value to the maritime pines to make them the more profitable tree crop in the presence of greenhouse gas abatement requirements.

2.3 Recharge flows from the system

Recharge values for each rotation are presented in Table 10. Values were obtained using the AgET Water Balance Calculator⁷, using rainfall data for Kojonup for the years 1954 to 1993. The first author, in consultation with Paul Raper⁸, specifically created data files that matched the biological and physical characteristics of the GSM LMUs. The recharge values depend on the type, number and order of crop and pasture phases in each rotation. Recharge is assumed to be negligible under the tree crop rotations given that studies have demonstrated that recharge under native vegetation is less than one mm per year (Allison et al. 1990; Kennett-Smith et al. 1992; Salama et al. 1993)⁹.

3. Results and Discussion

This section comprises three parts. Firstly the economic performance of oil mallees and maritime pines in the Great Southern region is presented (Section 3.1). Secondly, the economics of commercial tree crops for greenhouse gas abatement is discussed (Section 3.2). Thirdly, the recharge abatement value of greenhouse gas abatement policies is demonstrated (Section 3.3).

Table 10: Recharge levels for each rotation in GSM (mm/year)*

* Note that the Great Southern aquifers have a storage coefficient of between 0.1 and 0.2, so that 10mm of recharge equates to 50 to 100mm groundwater rise.

3.1 Economic performance of the tree crops in the Great Southern

Consider first the economics of oil mallees in the Great Southern. Figure 1 presents farm profit for increasing areas of oil mallees under different wool prices. Firstly, farm profit, especially in the absence of oil mallees, is highly dependent on wool price. This is not surprising given the dependence of the system on sheep production (85 percent of the farm is typically allocated to sheep production). Secondly, the optimal area of land planted to oil mallees increases with decreasing wool price¹⁰. This occurs as the farm substitutes oil mallees for sheep production as the profitability of the sheep enterprise decreases. Note that the optimal area of oil mallees depends on the farmer’s attitude towards risk. A formal risk analysis is not included here, but it follows that a risk-averse farmer would assume a high wool price when deciding on the optimal area of land to plant to oil mallees due to the irreversibility of the planting decision¹¹. Further discussion of decision-making under risk and uncertainty could be the topic of another paper.

Figure 1: The impact of the introduction of oil mallees on farm profit for varying wool prices (wool price is in c/kg greasy) in the absence of any greenhouse penalties

It is clear that it is profitable to plant part of the farm to oil mallees even at a high wool price. At the highest wool price (450c/kg greasy) it is optimal to plant all of LMU3 (deep sands) to the mallees. LMU3 is less suited to crop and pasture production than LMUs 4 and 5, hence the opportunity cost of growing mallees on LMU3 is the smallest. With decreasing wool price, it is optimal to plant more mallees on LMU 5 and then LMU 4, as the opportunity cost of planting the mallees on LMU5 is less than that of LMU 4. Note, also, that for all soil types and wool prices, mallees generally replace both crop and pasture land simultaneously, although the rate of replacement is faster on cropped land than pasture as cropping is the less profitable of the two enterprises. The medium term forecast of wool price is currently 400c/kg greasy.(This was the forecast at the time the paper was written.) All subsequent results will be presented assuming this forecast price.

3.1.1 Sensitivity analysis

Results presented so far have assumed the financial details presented in Table 7. Now consider a sensitivity analysis on these financial assumptions. Figure 2 is a presentation of the impact of the annuity on farm profit with different oil mallee areas. Farm profitability is highly sensitive to this annuity, especially with relatively large areas of land allocated to the mallees. It is important to note that while farm profit is increased substantially with increased mallee annuity, the mallee enterprise becomes unprofitable on all soil types if the annuity is decreased. It must be remembered for all subsequent results that mallee production is profitable under present or more optimistic financial assumptions but is unprofitable under less optimistic financial assumptions.

Figure 2: Sensitivity on oil mallee annuity

The economics of maritime pines is not so positive. The gross margins of the crop and pasture enterprises are higher than that of the maritime pines, hence the pines are not included in the optimal solution of GSM. Even with a 30 percent increase in the current annuity estimates, maritime pines are still not an economic alternative for the Great Southern. (Maritime pines can only be grown on five percent of the farm area, hence increases in areas of maritime pines has little effect on farm profit.)

Results presented so far are dependent on the relative areas of each LMU. A sensitivity analysis on these relative proportions is not included here; however, it is clear that a farm with greater proportions of the productive LMUs (i.e. 4 and 5) would be more profitable than farms with greater proportions of the more marginal LMUs (i.e. 1, 2 and 3). Additionally, the commercial potential of these tree crops would be larger for a more marginal farm (at least one with a higher proportion of LMU3) due to the smaller opportunity cost of crop or livestock production associated with marginal land compared with productive land.

3.2 Greenhouse abatement options in the presence of commercial tree plantations

Petersen et al. (2001) found that in the absence of on-farm greenhouse gas sinks, any farm-level policy for greenhouse gas abatement would have dramatic negative consequences on the farm enterprise, and in the absence of technological change, would cause the current farming systems to fail and be replaced by alternative land-uses. However, Petersen et al. (2001) found an emissions restriction policy (one where the farmer is legally required to restrict emissions) to be the more effective and efficient than a policy of taxing emissions. With this policy it was found that farmers were able to remain profitable while abating up to 48 percent (850 t CO₂-e) of their emissions levels by substituting pasture for crop on the most productive soil types (LMU 3, 4 and 5). Note that the farmers are not financially compensated for meeting these restrictions. A policy of restricting greenhouse gas emissions is the tool considered in this section.

The impact of varying levels of emissions restrictions on profit with and without the inclusion of tree crops is presented in Figure 3. As was found by Petersen et al. (2001), in the absence of tree crops, the farm falls to zero profits at a level of 48 percent abatement. In the presence of trees as a carbon sink, farm profitability increases with higher levels of abatement. First note that the model was allowed to select oil mallees, maritime pines or a combination of species. However, even at extreme abatement levels the model only selects oil mallees due to their high return to investment relative to maritime pines. Even though the pines have six-fold greater carbon sequestration efficiency, the sequestration efficiency of the oil mallees is more than enough to meet the farm’s sequestration needs. At 100 percent abatement, the area of oil mallees needed for the farm to be emissions-neutral is only 37 hectares (sequestering 1745 t CO₂-e). Farm profit increases for increasing levels of abatement as the profit-maximising area of mallees is 93 hectares, where sequestration exceeds greenhouse gas production on the farm (2571t CO₂-e). This is about 10% of the farm area with the given LMU distribution.

Figure 3: The impact of greenhouse gas abatement on farm profit

Consider now the impact of emissions restrictions on land use (Figure 4). In the absence of trees, pasture area decreases with increasing abatement as the farm substitutes out of pasture production into crop production, a relatively more efficient enterprise in terms of greenhouse gas abatement than pasture production. At an approximately 90 percent abatement level the system substitutes back into pasture production but removes livestock from the system (crop production emits more gases than pasture production in the absence of livestock), hence the upswing at the end of the curve in Figure 4. However, the farm is made unprofitable at an abatement level of 48 percent at which the system fails and, in the absence of technological advancement, would be replaced by alternative land-uses. In the presence of the accreditation of tree crops as carbon sinks, land-use changes only marginally, even at 100 percent abatement levels. As mentioned earlier, 37 hectares of trees are planted, mostly at the expense of pasture.

Figure 4: Land use with different levels of GHG abatement

A sensitivity analysis on sequestration efficiencies is presented in Figure 5. Results of the analysis show that earlier findings are very robust. Even at an extremely low efficiency, for example 20 percent of the estimated efficiency, the area of oil mallees required for 100 percent abatement is only approximately 95 hectares (or 9.5 percent of farm area). (Note that the assumed sequestration rates are low compared with those estimated in the literature. Hence, optimal areas of oil mallees here are over-estimated if anything.)

Figure 5: Area of oil mallees for different proportions of net carbon sequestration. (Numbers in the legend represent the proportion of the estimated sequestration efficiency of the oil mallees i.e. 1 can be equated to the bold line in Figure 4.)

So far it has been assumed that farmers are not compensated for restricting emissions. Petersen et al. (2001) found that it would cost the regulator approximately A$3.5 million a year in subsidies to achieve approximately 50 percent abatement in the Great Southern alone. If commercial plantations were credited as a source of sequestration under the Kyoto Protocol, implementing such a policy would virtually be costless to the regulator as optimal areas of these plantations exceed areas required for 50 percent (or even 100 percent abatement) with current or more optimistic financial assumptions. Furthermore, the value of the tree crops is increased in the presence of this policy and would be especially so if tradeable emission permits were permitted. Such a mechanism is likely to encourage faster growth of the tree crop industry in the Great Southern of Western Australia.

Results of this section have indicated that tree crops are very efficient carbon sinks. Only very small plantation areas are required for the farm to be emission-neutral. In addition, it was found that oil mallees have commercial value for farmers in the Great Southern of Western Australia although this value is sensitive to profit levels. They also substitute for fossil fuels in generating energy for other uses. On the other hand, maritime pines are not profitable, but have extremely high sequestration capabilities. In the presence of greenhouse gas abatement policies the value of both these tree crops is substantially advanced and is likely to lead to the greater production of these trees than would be economic otherwise. While the profitability of maritime pines is not great enough to encourage substantial plantings in the Great Southern under current financial conditions, the introduction of an emissions trading system may increase their value such that they are a viable option.

A Protocol that does not allow for accreditation for trees as carbon sinks would require expensive adaptation of current global practices to meet the short-term targets set at Kyoto. An example of the costs is the likely failure of the current Great Southern farming system (assuming the Australian government chooses to enforce restrictions in the agricultural sector). Such a Protocol would stifle business, discourage innovation and risk failure. A Protocol that includes carbon sinks in trees would ‘buy time’ and encourage innovation of greener technology. Furthermore, expansion in tree crop areas has other external benefits, one of which is salinity abatement (the topic of the next section). However, it is acknowledged that a conservative definition of sinks is required before a more accurate understanding of the ability of trees to sequester carbon is gained.

3.3 Abatement of recharge¹²

It is not suggested here that a policy should be introduced with the aim of decreasing greenhouse gas emissions and soil water recharge simultaneously. This would be breaking the cardinal policy sin of trying to achieve to objectives with the one policy instrument (Tinbergen 1952). Neither objective would be achieved efficiently. What is being investigated is the impact of a greenhouse gas abatement policy on recharge abatement.

First consider the impact of recharge abatement on farm profit (Figure 6). In the absence of tree crop production, it is only possible to abate 35 percent of recharge as no enterprise uses all the rainfall. Farm profit is relatively insensitive to low levels of recharge abatement. At approximately 20 percent abatement, profit decreases relatively quickly as the system substitutes into more crop intensive enterprises due to their relative recharge abatement efficiency (see Table 10). When tree crop production is introduced farm profit is maximised at a level of approximately 15 percent recharge abatement where 93 hectares of oil mallees are planted. Greater levels of abatement come at a cost. Note that it is still not possible to abate 100 percent of recharge as only 75 percent of land is suitable for commercial oil mallee or maritime pine production. A maximum of 88 percent of recharge can be utilised at a cost to the farmer of approximately A$15,000 per year (25 percent of optimal farm profit). Note that 88 percent recharge abatement in the presence of a commercial plantation comes at a similar cost as 35 percent recharge abatement in the absence of a commercial plantation.

Figure 6: The effect of recharge abatement on farm profit

The effect of recharge abatement on land use is illustrated in Figure 7. First consider the impact of recharge abatement on pasture area in the absence of a tree crop. To achieve 35 percent abatement, pasture area decreases by approximately 200 hectares (or 20 percent), and is replaced by crop production. In the presence of a tree crop, a similar decrease in pasture production is required to achieve 35 percent recharge abatement, however, this abatement comes at a profit rather than a cost as pasture is replaced by the more profitable oil mallee enterprise. As stated earlier, a high proportion of recharge is possible in the presence of a tree crop, although this requires the majority of the farm to be set aside for tree production.

Figure 7: The effect of recharge abatement on land use

4. Conclusions

This study focuses on the predominantly grazing systems of the Great Southern region of Western Australia. The two tree crops suitable for the region were analysed. Firstly, maritime pines were not found to be a profitable option for the region. On the other hand, oil mallees, a native to the region, were found to be a profitable enterprise for the region. The lower the wool price, the larger the profit-maximising area planted to oil mallees. This is understandable given the reliance of the system on sheep production. Farm profit was found to be sensitive to the oil mallee annuity. At standard and increased annuity, farm profit increased. However, decreases in the mallee annuity caused decreases in farm profit highlighting the importance of the financial assumptions for all subsequent results.

With the inclusion of an emissions restriction policy, the system was found to fall to zero profits at 48 percent abatement in the absence of tree crops. In the presence of tree crops, the optimal area of oil mallees was 93 hectares (less than ten percent of the farm area) leading to annual carbon dioxide sequestration volume of 2571 tonnes. This sequestration level is more than the volume emitted by the rest of the system achieving more than 100 percent abatement in the absence of any abatement policy. The initial assumptions of the model relating to sequestration rates were low compared with rates cited in the literature. However, a sensitivity analysis of these rates found that, even with an 80 percent reduction in the already conservative sequestration rates, a maximum of only ten percent of farm area is required to achieve 100 percent abatement.

From this analysis, it is seen that tree crops lead to a net sequestration of farm carbon. Their accreditation as carbon sinks (especially in the presence of emission trading) is likely to increase their profitability and hence, lead to expansion of area allocated to commercial plantations in Western Australia, in turn further increasing greenhouse gas sequestration. Furthermore, this study shows that the recharge abatement value of tree crops is substantial.

A Kyoto Protocol that does not allow for accreditation of commercial plantations as carbon sinks imposes heavy costs on the economies of some of the ratifying countries. The failure of the current Great Southern farming systems is one possible cost for Australia (assuming the Australian government applies restrictions on the agricultural sector), along with the sacrifice of environmental benefits of plantations such as salinity abatement. Given the political sensitivity of these issues, a Protocol that excludes accreditation risks failure. Accreditation of tree crops as sinks will buy time and is likely to create more business incentives for innovation of greener technologies than a Protocol that requires onerous short-term targets of expensive adaptation.

This paper has focused on the role of commercial tree crops for greenhouse gas abatement in south-western Australia. An interesting extension to this work would be to analyse the economics of non-commercial tree plantations in these farming systems. O’Connell (1999) found that the cost associated with establishing enough non-commercial trees to significantly impact on recharge in the eastern wheatbelt of Western Australia is likely to lead to negative profits. Although Hassall and Assoc (1999) found that carbon farming is only profitable when associated with commercial trees, the Protocol may increase their value enough for more widespread plantings of trees than is happening now for environmental purposes.

Petersen et al. (2001) and this analysis have focused on greenhouse gas abatement in the predominantly grazing systems of south-western Australia. Further analysis should focus on similar studies of the predominantly cropping systems of south-western Australia. Predominantly cropping systems are not so dependent on ruminant livestock and may have other options for cost-effective greenhouse gas abatement, such as stubble retention, lowering or changing fertiliser inputs, lowering fuel use and minimum tillage. Howden et al. (1994) found these options just mentioned to be cost-effective and efficient at reducing emissions for the Wimmera region of Victoria. Furthermore, commercial or non-commercial plantations are not so prevalent in the predominantly crop-based farming systems of Western Australia east of the Great Southern Region studied here, due to rainfall constraints. Accreditation of trees as carbon sinks may add sufficient value to the trees to encourage expansion of tree plantings in these areas.

References

Abbott, P. S. (1989). Commercial Eucalyptus oil production. Eucalyptus oil production. Gnowangerup, WA: 29-48.

ABS (1997). Agstats: Small Area Agricultural Commodity Data. Perth, Australian Bureau of Statistics.

AGO (1999). National Emissions Trading: Issuing the Permits. Canberra, Australian Greenhouse Office.

Allison, G. B., P. G. Cook, et al. (1990). "Land clearance and river salinisation in the Western Murray basin, Australia." Journal of Hydrology 119: 1-20.

Barton, A. F. M. and A. R. Knight (1997). "High cineole eucalypts in degreasing applications." Chemistry in Australia 64(1): 4-6.

Boland, D. J. (1991). Brief history of the eucalyptus oil industry and essential oil research in Australia. Eucalyptus leaf oils: use chemistry, distillation and manufacturing. D. J. Boland, J. J. Brophy and A. P. N. House. Canberra, Inkata Press: 3-10.

CALM (1999). CALM's maritime Pine Project. Perth, Department of Conservation and Land Management.

Canadell, J. and L. Lopez-Soria (1998). "Lignotuber reserves support regrowth following clipping of two mediterranean shrubs." Functional Ecology 12: 31-38.

Campbell, N., R. George, et al. (2000). Using Natural Resource Inventory Data to Improve the Management of Dryland Salinity in the Great Southern, Western Australia. National Land and Water Resources Audit, Implementation Project 2, Theme 2 - Dryland Salinity.

Chegwidden, A., D. Harrison, et al. (2000). "Mallee eucalypts for electricity, activated carbon and oil, a new regional industry". Proceedings Bioenergy 2000: Making a Business from Renewable Energy for Society and the Environment, Gold Coast, Queensland (in press).

Cooper, D. (1999). An economic analysis of oil mallee industries in the wheatbelt of Western Australia. Agricultural and Resource Economics. Perth, University of Western Australia.

Grzyb, H., P. Mcleod, et al. (2000). Great Southern Regional Wood Processing Industry Capability Study, Report prepared for Timber 2002.

Hassall and Assoc (1996). Sequestration of Atmospheric Carbon Dioxide in Trees. p14, Report prepared for the Australian Nature Conservation Agency.

Hassall and Assoc (1999). Greenhouse, Carbon trading and Land Management. Canberra, Land and Water Resources Research and Development Corporation.

Herbert, A. (2000). Economics of oil mallees. South Perth, Agriculture Western Australia.

Howden, S. M., D. H. White, et al. (1994). "Methods for exploring management options to reduce greenhouse gas emissions from tropical grazing systems." Climatic Change 30: 1-19.

IPCC (2001). Climate Change 2001: The Scientific Basis - Summary for policy makers. Report prepared by Working Group 1 of the Intergovernmental Panel on Climate Change, Shanghai, January 2001.

Kennett-Smith, A. K., G. J. O'Leary, et al. (1992). Estimation of recharge from soil water profiles under dryland agriculture, Eucalyptus plantation, an natural woodland in the Wimmera district of Victoria. Canberra, ACT, Australia, CSIRO Division of Water Resources.

McCarthy, A. (2000). , Department of Conservation and Land Management (CALM).

Morrison, D. A. and J. Young (1991). "Profitability of increasing lambing percentage in the Western Australian wheatbelt." Australian Journal of Agricultural Research 42: 227-241.

NGGI (1998a). National Greenhouse Gas Inventory Fact Sheet 4: Agriculture. Canberra, Australian Greenhouse Office.

NGGI (1998b). Workbook for Livestock. Canberra, Australian Greenhouse Office: National Greenhouse Inventory Committee.

NGGI (1998c). Workbook for Non-Carbon Dioxide Gases from the Biosphere. Canberra, Australian Greenhouse Office: National Greenhouse Gas Inventory Committee.

NGGI (1998d). Workbook for Transport (Mobile Sources). Canberra, Australian Greenhouse Office: National Greenhouse Inventory Committee.

O'Connell, M. (1999). The Impact of Non-Commercial Trees on Groundwater Recharge and Farm Profits in the Eastern Wheatbelt. South Perth, Agriculture Western Australia.

Pannell, D. J. (1996). "Lessons from a decade of whole-farm modelling in Western Australia." Review of Agricultural Economics 18: 373-383.

Pannell, D. J. (1999). "On the estimation of on-farm benefits of agricultural research." Agricultural Systems 62(2): 123-134.

Petersen, E.H., S. Schilizzi and D. Bennett (2001). "An assessment of policy options for greenhouse gas abatement on mixed cropping enterprises in the Great Southern of Western Australia". Contributed paper presented at the 45th Annual Conference of the Australian Agricultural and Resource Economics Society, 22-25 January 2001, Adelaide, South Australia (Submitted for publication in Agricultural Systems).

RIRDC (1999). Integrated tree processing of mallee eucalypts, Rural Industries Research and Development Corporation.

Ritson, P. (2000). Department of Conservation and Land Management.

Salama, R. B., P. Farrington, et al. (1993). "Salinity trends in the wheatbelt of Western Australia: results of water and salt balance studies from Cuballing catchment." Journal of Hydrology 145: 41-63.

Schmidt, C. and D. J. Pannell (1996). "Economic issues in management of herbicide resistant weeds." Review of Marketing and Agricultural Economics 64: 307-308.

Shea, S. (1999). Farming Carbon. Landscope Tree Crops for Farms: p36.

Shea, S., G. Butcher, et al. (1998). The potential for tree crops and vegetation rehabilitation to sequester carbon in Western Australia. Paper presented at the Carbon Sequestration Conference, 19-21 October, 1998, Le Meridien at Rialto, Melbourne, Department of Conservation and Land Management.

Specht, R. L. (1996). The influence of soils on the evolution of the eucalypts. Nutrition of Eucalypts. P. M. Attiwill and M. A. Adams. Collingwood, Australia, CSIRO Australia: 31-60.

Tinbergen, J. (1952). On the Theory of Economic Policy. North Holland, Elsevier.

UNFCCC (1997). Kyoto Protocol to the United Nations framework Convention on Climate Change. Available from the UNFCCC website: http://www.unfccc.de/.

WA Government (1998). Environment Western Australia 1998 State of the environment report. Perth, WA, Australia, Department of Environmental Protection.

Weiss, E. A. (1997). Essential oil crops. Wallingford, UK, CAB International.

Wildy, D. T. (2000). "Sapling and coppice biomass production by alley-farmed 'oil mallee' Eucalyptus species in the Western Australian wheatbelt." Australian Forestry 63(2): 147-157.

Wood, W. E. (1925). "Increase of salt in soil and streams following the destruction of native vegetation." Journal of the Royal Society of Western Australia X: 35-47.

Young, J. (1995). MIDAS: Model of an Integrated Dryland Agricultural System: Manual and Documentation for the Great Southern Model Version GS92-3. Nedlands, Centre for Legumes in Mediterranean Agriculture.

Endnotes:

1. Although emissions from livestock decreased by 5.1 percent from 1990 to 1998 , mostly due to the drop in sheep numbers associated with a drop in wool price.

2. It is assumed that the opportunity cost (and marginal cost) of labour is zero as the farmer does the planting when no other work is required on the farm.

3. In the case of oil mallees, the remaining trees are harvested in year 30. Note that if the land is required for other uses after the final clear-fell, then uprooting and de-stumping would be required at an extra cost.

4. The ratio of tonnes per cubic metre changes throughout the life of the tree. A ratio of 1.95, 1.32, 1.06 and 0.88 tonnes per cubic metre are assumed at age 12, 18, 24 and 30 respectively (Ritson 2000).

5. This is done by multiplying the carbon weight by the atomic weight of carbon dioxide (44) and dividing by the atomic weight of carbon (12).

6. Shea, Butcher et al. (1998) quote average annual sequestration rates of maritime pines and oil mallees as 378t/ha and 114t/ha respectively, but do not offer an explanation for how they obtained these numbers.

7. AgET was developed by Agriculture Western Australia and The University of Melbourne. It can be obtained from the Agriculture Western Australia website: http://www.agric.wa.gov.au/

8. Paul Raper, Research Hydrologist, Agriculture Western Australia, Bunbury

9. Note that a large number of hydrogeologists and agronomists estimate that trees will lower high water tables and hence recover saline and waterlogged land.

10. Optimal mallee area for each wool price is as follows: 50ha (450c/kg greasy), 93ha (400 c/kg greasy), 250ha (350 c/kg greasy) and 594ha (300 c/kg greasy).

11. This should also be equated with the riskiness of the tree crop enterprise including price, production, flood and fire risks.

12. In this section we have not included the changes in area of LMU2 (waterlogged), as this would require MIDAS to contain a spatial analysis of recharge effects. This is beyond the ability of the model in its current configuration.

Citation: Petersen, E., Schilizzi, S. and Bennett, D. (2001). An economic assessment of the role of commercial tree crops to achieve greenhouse gas neutrality in predominantly grazing systems of south-western Australia. SEA Working Paper 01/03, Agricultural and Resource Economics, University of Western Australia. http://www.general.uwa.edu.au/u/dpannell/dpap0103.htm

SEA News issue #9

The SEA News index is at http://welcome.to/seanews

Copyright © 2001 Elizabeth Petersen, Steven Schilizzi and David Bennett
Last revised: May 21, 2003.