Feral Camel Control Strategies

Transcription

1 54 Economics of camel control in the central region of the Northern Territory AG Drucker Report

2 Economics of camel control in the central region of the Northern Territory AG Drucker 2008

3 Contributing author information Enquiries should be addressed to: Adam G Drucker: School for Environmental Research, Charles Darwin University, Darwin, NT, 0909, Australia. Desert Knowledge CRC Report Number 52 Information contained in this publication may be copied or reproduced for study, research, information or educational purposes, subject to inclusion of an acknowledgement of the source. ISBN: (Online copy) ISSN: Citation Drucker AG , DKCRC Research Report 52. Desert Knowledge CRC, Alice Springs. Available at au/publications/contractresearch.html The Desert Knowledge Cooperative Research Centre is an unincorporated joint venture with 28 partners whose mission is to develop and disseminate an understanding of sustainable living in remote desert environments, deliver enduring regional economies and livelihoods based on Desert Knowledge, and create the networks to market this knowledge in other desert lands. For additional information please contact Desert Knowledge CRC Publications Officer PO Box 3971 Alice Springs NT 0871 Australia Telephone Fax Desert Knowledge CRC 2008 The project was funded by Australian Government. The views expressed herein do not necessarily represent the views of Desert Knowledge CRC or its participants. II Desert Knowledge CRC

4 Contents Tables...IV Figures...IV List of shortened forms...iv Acknowledgements... V Executive summary...vi Key findings and recommendations...vii Recommendation 1:...VII Recommendation 2:...VII 1. Introduction Conceptual background and modelling approach Economics of feral animal control Model description Cost-Benefit analysis Population modelling Feral camel control costs Population Control method Alternative control strategies and target densities Benefits of control Direct economic impacts: cattle production impacts and damage to infrastructure Indirect economic impacts Total benefits Discussion Sensitivity analysis Population growth and carrying capacity cap Aerial shooting costs Degree of competition between feral camels and cattle Methane emissions Infrastructure constraints Discount rate and time horizon Conclusions References Desert Knowledge CRC III

5 Tables Table 1: Feral camel population and control costs under an annual removal strategy (1)...7 Table 2: Present benefits under Strategy 1 with a 12-year time horizon and a 5% discount rate...11 Table 3: Net present value of alternative control strategies...13 Table 4: Present control costs and benefits under alternative feral camel population growth models and assumptions...16 Table 5: Total present costs of Strategy 1 under alternative discount rates and time horizons ($m)...19 Table 6: Net present value of control (Strategy 1) under alternative discount rates and time horizons ($m)...20 Figures Figure 1: Map of central NT...2 List of shortened forms AGDCC INRM IPCC NPV NRETAS NRM NTG PB PC SER/CDU Australian Government Department of Climate Change Integrated Natural Resource Management Intergovernmental Panel on Climate Change net present value Natural Resources, Environment, The Arts and Sport (NT Government Department) Natural Resources Management Northern Territory Government present benefits present costs School for Environmental Research, Charles Darwin University IV Desert Knowledge CRC

6 Acknowledgements I should like to thank NRETAS Biodiversity Conservation staff Glenn Edwards, Keith Saalfeld, and Benxiang Zeng for their assistance in providing the data upon which the model development and analysis is based. Thanks also to Clive McMahon and Stephen Garnett (SER/CDU) regarding advice about population modelling. The model development also benefited from previous work carried out under the Australian Government funded NRETAS project Review of threats to biodiversity in the Northern Territory. The work reported in this publication is supported by funding from the Australian Government through the Desert Knowledge CRC; the views expressed herein do not necessarily represent the views of Desert Knowledge CRC or its participants. Desert Knowledge CRC V

7 Executive summary A cost-benefit analysis based on a bio-economic model was carried out to evaluate specific feral camel control strategies in the central region of the Northern Territory (NT). Based on expert opinion obtained through a series of workshops and meetings, and with a view to achieving the NT Integrated Natural Resources Management (INRM) Plan goal by 2020, specific control strategies for feral camels in the central region of the NT were identified. Two different aerial control strategies were modelled. Strategy 1 involved annual removals, while Strategy 2 involved periodic removals only when a specific feral camel density was reached. The direct economic benefits for the pastoral industry of feral camel control were also modelled in terms of reduced grazing competition together with infrastructure damage. A single environmental service related to reduced methane emissions was further considered. Although cultural values and other environmental services are also likely to be important, their modelling was beyond the scope of this study. Consequently, the analysis carried out in this report does not account for these values. The total present value of costs of the feral camel control strategies ranged from $5.39m (Strategy 2) to $6.00m (Strategy 1) over a 12-year time horizon (at a 5% discount rate), equivalent to an annualised present cost of $ $ , respectively. Depending on how such a control program were implemented, these costs could be both public and private in their incidence (i.e. incurred by government and/or landholders). Of the $6.00m Strategy 1 costs, $3.74m (62.3% of total) would be spent in year 1; $ (15.2% of total) in year 2; and $ in each year thereafter. It is therefore apparent that the vast majority of the control costs are spent in the first two years of the control program, making the costeffectiveness of a go-stop policy low (Strategy 2). Although control costs are large, they are far outweighed by the direct economic benefits to the livestock industry from reduced competition between livestock and feral camels ($50.68m under Strategy 1 or 57.9% of total present benefits). The value of reduced methane emissions is also large ($35.24m or 40.3% of total present benefits), while reduced infrastructure damages make a relatively small contribution to total present benefits ($1.62m or 1.8%). Total present benefits under Strategy 1 are thus $87.54m over 12 years or $9.88m per annum and were larger than those found under Strategy 2 ($83.98m). The difference between the economic benefits under the different strategies suggests that a control strategy based on annual removals is almost always likely to be preferred. We can therefore conclude that the magnitude of the benefits arising from a given control strategy should play a key role in control strategy choice. We also note that approximately 60% of the benefits (i.e. from reduced grazing competition and infrastructure damage) will accrue privately to pastoralists, while the remaining 40% (methane emissions avoided) will accrue publicly. The net present value of control (i.e. total present benefits minus total present costs) is $81.54m under Strategy 1. Delays in implementation of a control program could, however, reduce this value significantly. For example, a one-year delay could reduce this value by $7.7m, largely because of benefits forgone during the delay. Given the large positive net present value of control and the robustness of the overall findings, there would appear to be a very strong argument for considering the implementation of a full-scale, long-term feral camel control program in the near future. VI Desert Knowledge CRC

8 Key findings and recommendations Finding 1: The total present value of costs of NT INRM Plan compatible feral camel control strategies to 2020 in the central region range from $5.39m (Strategy 2 periodic removals) to $6.00m (Strategy 1 annual removals), equivalent to an annualised present cost of $ $ respectively. Finding 2: The vast majority ( %) of the total present costs of control are spent during the first two years. As such, annualised figures tend to significantly underestimate the control agency s funding requirement in the first years of a control program. Finding 3: Although control costs are large, they are far outweighed by the economic benefits to the livestock industry from reduced competition between livestock and feral camels ($50.68m), as well as to society as a whole through reduced methane emissions ($35.24m). Including reduced infrastructure damage, the net present value of control is $81.54m under Strategy 1 and $78.59m under Strategy 2. Recommendation 1: The difference between the present value of the economic benefits under the different strategies suggests that a control strategy based on annual removals should be preferred over a strategy of periodic removals. Finding 4: Although cultural values and other environmental services are also likely to be important, their quantification in economic terms is not required to justify a decision regarding whether to undertake a control program or not. Finding 5: The findings were found to be robust under a series of alternative assumptions. The uncertainty regarding feral camel population estimates needs to be addressed in future years through improved monitoring and data collection. Finding 6: Delays in implementation of a control program can reduce the present value of the benefits gained from a control program significantly. Recommendation 2: Given the large positive net present value of control and the robustness of the overall findings, there would appear to be a very strong argument for considering the immediate implementation of a fullscale, long-term feral camel control program. Desert Knowledge CRC VII

9 VIII Desert Knowledge CRC

10 1. Introduction Like other regions of the world, the natural resources of the Northern Territory (NT) face a range of threats, many of which are costing government, business, and individuals a great deal of money and effort to counter. The threats include fire regimes that are changing vegetation patterns, introduced animals that compete with or kill native wildlife, weeds that spread and replace native vegetation, and land uses that destroy or degrade vegetation (Price et al. 2007). Exotic pest animals have major economic, environmental, and social impacts across Australia (Australian Pest Animal Strategy 2007). In a major review of the most significant threats to biodiversity in the NT (covering fire, feral animals, pastoralism, weeds, and land clearing), one of the highest ranked threats across all regions was related to the presence of large feral herbivores (Price et al. 2007). There are 19 species of exotic vertebrate pests in the NT. Arabian camel, donkey, horse, cane toad, pig, water buffalo, fox, and cat are considered major pests because they have a high level of overall impact at current densities and distributions. Other species such as the European rabbit, wild dog (excluding dingoes), and goat are considered to be moderate pests because they have lesser impacts to biodiversity, at current levels. Other species such as the house sparrow, rock pigeon, turtle dove, samba deer, black rat, brown rat, and banteng are considered minor pests as their overall impact is relatively minor (Price et al. 2007). Despite recognising the existence of these threats, their extent and severity is not well understood. Furthermore, it is unclear whether the funding made available for the control of feral animals (by the NT and Australian Governments) is either adequate or being effectively targeted. In principle, funding should be directed to those species and regions where either the most cost-effective outcomes can be achieved (in terms of reducing population numbers) or the highest net benefits can be obtained (accounting for a reduction in production and environmental losses). This paper attempts to support such an analysis by assessing the relative costs and benefits of two feral camel control strategies. In order to carry out such an assessment, a cost-benefit analysis is carried out with regard to the central region of the NT (see Figure 1) feral camel (Camelus dromedarius) control activities. Based on expert opinion obtained through a series of workshops and meetings, population data were obtained and, with a view to achieving the NT Integrated Natural Resource Management (INRM) Plan goal by 2020, specific aerial control strategies were identified and modelled. The direct economic benefit to the pastoral industry of feral camel control was also considered, together with the indirect economic benefits associated with reduced infrastructure damage and reduced methane emissions. Model findings are subjected to a sensitivity analysis in order to assess their robustness. The remainder of this report is organised as follows: Section 2 details the conceptual background and modelling approach, Section 3 details the costs of control, Section 4 details the benefits of control, Section 5 discusses the findings, Section 6 subjects the model findings to a sensitivity analysis and Section 7 presents conclusions. Desert Knowledge CRC 1

11 Figure 1: Map of central NT 2 Desert Knowledge CRC

12 2. Conceptual background and modelling approach 2.1 Economics of feral animal control The management of pests involves making choices that determine how much pest control will cost and what benefit it will deliver. In order to make informed choices, the effect that alternative courses of action have on how the costs and benefits of pest control accrue should, ideally, be understood. To understand how benefits and costs vary between different pest management strategies, the biological and management components of a pest/resource system must be linked so that its economic inputs and outputs can be estimated and compared (Choquenot & Hone 2000). Bioeconomics, that is, the economic analysis of biological systems (Clark 1990), provides a potentially powerful approach to the analysis of pest management by furnishing a quantitative framework for considering the benefits and costs of alternative pest control strategies. Bioeconomic models can have varying levels of complexity, linking economic inputs, such as the costs of feral animal control, with consequent economic outputs, such as the benefits associated with reduced pest numbers. Economic inputs and outputs are analysed to identify feral animal control strategies that produce optimal and/or cost-effective outcomes (Choquenot & Hone 2000). Many feral animals are pests when they impact on grazing capacity and hence the income of pastoralists. McLeod (2004) notes that, consequently, economic outputs have often been related to improved agricultural or livestock productivity associated with specific pest species control. In order to estimate this impact, the distribution of the pest is estimated, the value of agricultural production within the range of the pest identified, and an assessment of the reduced value of production as a result of the pest is calculated (for examples see Hone 2006 and Barlow 1987). Any research or management/ monitoring costs associated with the pest may also be accounted for. 2.2 Model description Following this conceptual approach, we apply a simple bioeconomic model in which economic inputs (costs of control) are integrated into a feral camel population model, leading to economic outputs related to the present costs and benefits of control. In our bioeconomic model, in addition to the initial starting population, two factors influence the feral camel population stock: the number of animals removed each year (subjected to different control strategies) and the population growth rate. The number of feral animals removed each year is one of the economic components and, together with their density, determines the control costs of a strategy. The other economic components include the direct economic benefits of increased cattle production that result from reduced competition with the feral camels that are removed from the grazing system, as well as the indirect economic benefits associated with reduced infrastructure damage and reduced methane emissions. Although we recognise that feral camels may also have important economic impacts on other environmental services, such as biodiversity, as well as socio-cultural impacts (e.g. damage to sacred sites, reduced bush-tucker abundance, etc.), we are unable to quantify such impacts given the current data available. We nonetheless note that where the cattle production, infrastructure, and methane emission impacts are of sufficient importance to justify a feral camel control program, valuing other environmental and socio-cultural impacts may not be crucial to the choice that needs to be made regarding the degree of pest control necessary. Desert Knowledge CRC 3

13 The model developed is in the form of a spreadsheet decision-support tool (the model can be found at that feral camel managers can use to inform future decision making. To this end, the report directs the reader to specific cells in the spreadsheet model where the data in question can be found Cost-Benefit analysis A comparative analysis of the present costs and benefits of control is carried out. The net present value (NPV) of control is a measure of the financial resources required to reach a target population density of feral camels relative to the current density. The discounted stream of future costs and benefits associated with specific control strategies over a given time horizon is calculated as follows: (1) where PB and PC are, respectively, the present benefits and present costs of control. The use of a discount rate r over i years (time horizon) of a control program is used so that future costs and benefits can be expressed in present value terms. For simplicity, it is assumed that all costs and benefits of a control program occur at the end of the year in which they are undertaken. 2.4 Population modelling In the absence of alternative data, a simple growth rate model was used which included a carrying capacity cap (k). If the feral camel population is below the carrying capacity cap, the population P for the following year is calculated by: for all P < k (2) where i signifies the year and r the growth rate. The expected feral camel population size in the central region of the NT for the year 2009 (year 1 of our analysis) was determined by extrapolating from the 2001 aerial survey estimates of Edwards et al. (2004) using a growth rate r of 10% p.a. 1 Cell references are presented as superscripts in the following format: where the reader is directed to a cell on the 1. Summary-Input Entry & Results page of the spreadsheet, only the cell reference is used e.g. B4 ; where the reader is directed to one of the other spreadsheet pages, the cell reference is preceded by the sheet number e.g. 2B4 for cell B4 on the 2. Camels CBA Model page, 3B4 for the equivalent cell on the 3. Initial Population Data page, etc. For example, the text in section 3.1 below: a natural growth rate of 10% C33 per year refers the reader to cell C33 in the spreadsheet, which gives the source of the population growth rate of 10%. 4 Desert Knowledge CRC

14 3. Feral camel control costs 3.1 Population Camels were first introduced into Australia in the 1840s to assist in the exploration of inland Australia. It is estimated that in the period between and camels were imported from India, with an estimated 50 65% landed in South Australia (McKnight 1969). It is not known when the first feral population established, but some escaped during the Burke and Wills expedition in The feral animal population increased substantially after the 1920s when trucks became a widespread form of transport. Australia now has the largest wild population of camels in the world. In 2001 the Australian feral camel population was estimated to be in the order of (Edwards et al. 2004). Feral camels are widely distributed across 2.8 million km 2, or 37% of the Australian mainland, including the rangelands of Western Australia (WA), South Australia (SA) and the NT (Short et al. 1988). Feral camels range in pastoral land in arid and semi-arid Australia, with pastoral areas dominated by Acacia trees and shrubs particularly well suited to feral camel grazing (Short et al. 1988). In the NT, feral camels are mainly confined to the southern third of the land area. A 2001 aerial survey indicated that there was a minimum of feral camels in the NT and that the population was doubling every eight years (Edwards et al. 2004). Based on the results of the 2001 census and an assumed actual feral camel population of at that time (i.e. the 2001 NT population estimate of was a minimum) and a natural growth rate of 10% C33 per year, by 2009 (year 1 of our analysis) the estimated feral camel population in the NT portion of central Australia ( C36 km 2 ) will be approximately M9. In the absence of control the population is assumed to increase to C35 before stabilising as the maximum capacity of the resource base to support feral camels is reached. Such a hypothetical equilibrium would be reached by the beginning of year 6 2I4 (2014). 3.2 Control method The preferred method of controlling feral camels is shooting by helicopter. Shooting from an aerial platform (helicopter) herein referred to as aerial shooting involves the use of a helicopter flying at low altitudes and low speed to position a marksman relative to the target animals so as to have a clear and unimpeded shot to obtain a humane kill. Both the helicopter pilot and marksman have to have undertaken specific training and received recognised accreditation before engaging in aerial shooting operations (Anon 1991). Aerial shooting has long been recognised as the only practical method of controlling a number of large vertebrate feral animals, including camels, across large-scale regions, in inaccessible areas, or to achieve rapid density reductions (Anon 1991, Dobbie et al. 1993, Edwards et al. 2004, Norris & Low 2005). Norris and Low (2005) identify aerial shooting from helicopters as probably one of the best control techniques for large feral herbivores in the Rangelands. In some instances control can be assisted through trapping and/or mustering for the purpose of commercial sale (Dobbie et al. 1993). However, the extent to which trapping and mustering can be used depends on market demand and the accessibility of the animals under management. In the NT, all three techniques are used to manage horses and donkeys. Based on Bayliss and Yeomans (1989), as well as NT Department of Natural Resources, Environment, The Arts and Sport (NRETAS) data collected from actual control activities, it is recognised that there is an increasing marginal cost associated with aerial culling as feral animal densities decline. This is because the labour and helicopter time required to shoot individual animals increases as fewer target animals can be identified. Thus, at densities of equal to or above 0.25 B9 feral camels/km 2, total aerial Desert Knowledge CRC 5

15 control costs are estimated to be $20 C9 /feral camel. At densities down to 0.15 B10 feral camels/km 2, these costs are $40 C10 /feral camel, increasing to $60 C11 /feral camel for densities down to 0.1 B11 animals/km 2. Below this density, costs increase to $110 C12 /animal (K Saalfeld 2008, NRETAS, pers. comm.). In addition to aerial control costs, costs for management and administration of control methods were estimated based on feral camel density. At densities equal to or above 0.25 feral camels/km 2 management costs are assumed to be $ C22 per year, falling to $5000 C23 at densities lower than this. In addition, status monitoring costs are fixed at $30,000 C21 per year regardless of density. The total present value of the control costs is thus closely related to initial and target feral camel densities. 3.3 Alternative control strategies and target densities. Two alternative control strategies are modelled. Both strategies aim to achieve a feral camel density of 0.25 B16/B17 animals/km 2 in the first year and then 0.1 C16/C17 animals/km 2 thereafter. The latter target density is chosen on the basis that, in the view of feral animal control experts, this is the density that is compatible with the NT s INRM Plan goal of no deterioration in the extent, condition and functionality of the native Territory environments in which the feral camels are found. Strategy 1 aims to attain this level as quickly as possible and then maintain that level through annual removals. By contrast, Strategy 2, while also aiming to attain the 0.1 animals/km 2 goal as quickly as possible, permits densities to rise to 0.25 E17 animals/km 2 before triggering a further round of removals to reduce feral animal densities down to 0.1/km 2. A priori, it is expected that the latter strategy may be cheaper as there are increasing marginal costs of removing feral animals as densities decline. Based on the above population figures, where control Strategy 1 is used (Table 1), the initial density of 0.98 I39 animals/km 2 means that approximately D53 feral camels need to be removed (almost 75% of the current population). Taking into account the natural growth rate of feral camels (10% per year), the population at the beginning of the second year will have reached approximately E44 and a further E53 feral camels need to be removed to reach the target density of 0.1 animals/km 2. Total population will then be E57 feral camels and the 10% natural increase of F53 will need to be removed each year thereafter. Table 1 also presents the control costs that occur in each year under an annual feral camel removal strategy. Given that marginal control costs increase as densities decline, it is assumed for simplicity s sake that the cost level to be applied is that which is related to the density at the beginning of each year. Although costs per animal removed increase with declining density, the absolute number of animals removed declines rapidly, leading to a fall in total control costs from $3.92m 2D64 in year 1 to $1.0m 2E64 in year 2 and $ F64 thereafter. Given that the control costs presented in Table 1 occur in different years, in order to determine their total present value it is necessary to discount them to a present value. Using a typical real discount rate of 5% C58 (based on an approximate capital opportunity cost of 9% and an inflation rate of 4%) and assuming a 12 C56 -year time horizon (i.e. covering the period ), it can be seen that a feral camel control program would cost approximately $6.00m D69, which is equivalent to an annual present cost of approximately $ D70. A similar analysis under a density sensitive control strategy (2) reveals total present costs of approximately $5.39m D104 over 12 years, which is equivalent to an annualised present cost of approximately $ D105. Strategy 2 control costs are lower as removals are only carried out in years 1, 2, and 12 when feral camel densities are above the trigger density of 0.25 animals/km 2. Feral 6 Desert Knowledge CRC

16 camel numbers removed in years 1 and 2 are identical to those under Strategy 1. In subsequent removal years (years 12 and every 10 years thereafter 2 ) approximately O88 feral camels would have to be removed, compared with 2620 per year under Strategy 1. Table 1: Feral camel population and control costs under an annual removal strategy (1) Parameter Year 1 Year 2 Year 3 and thereafter Total Feral camel population including natural growth Total area (km 2 ) Feral camel density/ km Target density/km Feral camels to be removed to achieve target (number) Total animals remaining at end of year after removals Control costs (Strategy 1: annual removals) Aerial shooting cost ($/animal) $20 $20 $60 Total cost of aerial shooting $ $ $ Status Monitoring $ $ $ Management of control program $ $ $5 000 Total control cost in each year (Aus$) $ $ $ Total present costs of control (over 12 5% discount rate) Annualised present costs of control (over 12-year time 5% discount rate) $ $ Note that the latter do not play a role in the current 12-year analysis. Desert Knowledge CRC

17 8 Desert Knowledge CRC

18 4. Benefits of control The benefits of feral animal control can be understood in terms of costs avoided as a result of the implementation of a given control strategy. Costs avoided can be categorised as environmental, cultural, and economic. Environmental In general, feral camels are known to have negative impacts on sensitive and threatened plants and plant communities through grazing and trampling in areas where they occur. They contribute to soil erosion, damage vegetation, and foul waterholes (Dörges & Heucke 2003, P Latz, Ecological consultant, pers. comm.). Feral camels are also considered to have a direct impact on sensitive and threatened animals through habitat modification and competition for food and other resources where they occur at moderate to high densities. Grazing and trampling also lead to a direct impact on landscape function. Camels are responsible for methane emissions as a by-product of their digestive process. Cultural Feral camels have negative impacts on Aboriginal cultural values. Impacts occur through habitat modification, damage to culturally important sites including waterholes, damage to cultural resources such as bush foods and trees used for artefact production, and the loss of totemic animal species (refer to Edwards et al for more details). Economic In terms of production, there is a direct impact through competition for food and habitat modification in areas where feral camels overlap with pastoralism, agriculture, and bush food production. In times of scarce forage, and particularly in arid areas, feral camels are likely to compete for herbage with sheep and cattle. This competition inflicts a direct cost on Australia s grazing industries (McLeod 2004) while indirect negative impacts (e.g. through damage to infrastructure and the spread of weeds) also occur on production. The magnitude of all the above impacts is assumed to increase with the feral camel density. While the economic impact of feral camels in all three of these categories may be significant, as noted above, due to data availability constraints we are only able to focus on the direct economic costs in terms of the forgone (private) income from cattle production and the costs of infrastructure damage. A single indirect cost associated with an environmental service methane emissions avoidance is also considered. Assessing the economic values of the other environmental and cultural values is beyond the scope of this study Direct economic impacts: cattle production impacts and damage to infrastructure Pastoralism is the predominant land use in the NT in terms of area, with approximately 55% of the land under some form of pastoral management (pastoral leases plus pastoral operations on some Aboriginal land tenures) (NTG 2008). Pastoral land use spread through most suitable areas of the NT during the 1870s 1890s. The industry is now primarily based on breeding and turning off cattle for live export or fattening elsewhere in Australia. Grazing is generally based on native pastures, although introduced Desert Knowledge CRC 9

19 species are used in some areas, and property and paddock sizes are generally very large (Oxley et al. 2005). Currently, increasing demand and rising costs, as well as high land values, are placing pressure on pastoralists to increase productivity, leading to further intensification of pastoral use through infrastructure development, increased stocking rates, and greater use of exotic pastures (Ash et al. 2006). We assume a negative relationship between yields from cattle production and feral camel density. Reducing the numbers of feral camels hence has potential benefits for the pastoralist industry. We conservatively estimate the net income forgone per annum to the pastoralist from each head of cattle that is replaced by feral camels to be $100 C42. The magnitude of the production loss avoided depends on a number of factors. These include: The current feral camel population. This is extrapolated to the present from the most recent census data that is available and varies in future years according to natural growth rates and removal efforts. The proportion of the total feral camel population that is found on pastoral stations. This is estimated as 20%. C45 The proportion of pastoral properties that provide good grazing and where competition with feral camels actually takes place. This is estimated as 62% C44. Combined with the previous assumption this means that 12.4% (0.62 x 0.2) of feral camels are directly competing with cattle. The degree to which feral camels are considered to compete with cattle for scarce grazing resources. The degree of competition is expressed as a proportion of the feral camel feed requirements to cattle. This is assumed to be 1.5 C43 times higher, that is, 1 feral camel = 1.5 cows. Each feral camel can therefore be considered to cause an annual financial loss to cattle producers through increased grazing competition of $18.60 (0.124 x 1.5 x $100), equivalent to $4.78m for a herd of feral camels. According to a pastoral property survey data (refer to Zeng & Edwards 2008 for more details) covering infrastructure damage during the previous two years, damages to fences, yards, and water points occurred on two-thirds of pastoral properties. Such damage was estimated by the landholders as totalling $ per year; assuming that the amount of damage is directly proportional to feral camel densities, we extrapolate this to $ C49 per year at 2009 predicted densities. We also recognise that this figure is at best a lower-bound estimate of infrastructure damage as the survey excluded a number of other types of infrastructure, as well as some Aboriginal managed lands and conservation areas. 4.2 Indirect economic impacts Methane is a greenhouse gas 21 times more powerful than carbon dioxide. Methane is produced in herbivores as a by-product of enteric fermentation, a digestive process by which carbohydrates are broken down by micro-organisms into simple molecules for absorption into the bloodstream. Both ruminant animals (including cattle and sheep) and some non-ruminant animals produce methane, although ruminants are the largest source since they are able to digest cellulose, a type of carbohydrate, due to the presence of specific microorganisms in their digestive tracts. The amount of methane that is released depends on the type, age, and weight of the animal, the quality and quantity of the feed, and the energy expenditure of the animal (IPCC 1997). According to the IPCC (1997, p 4.10), camels enteric fermentation emissions are 46 kgs/animal, which is equivalent to 0.97 t/animal 4A7 of CO 2 e per year. 10 Desert Knowledge CRC

20 The existence of approximately feral camels in the central region of the NT thus generates approximately tons of CO 2 e per year, equivalent to 1.6% of total NT emissions (Garnett et al. 2008). Removing feral camels can therefore make a contribution to NT emission-reduction strategies insofar as only a small proportion of removed feral camels are replaced by cattle, which have slightly higher emissions (1.31 t/animal 4A22 per year for rangeland beef cattle: AGDCC 2008). Helicopter CO 2 e emissions resulting from removal activities are taken into account and are found to be relatively low at t 4A16 CO 2 e per feral camel (Garnett et al. 2008). Following estimates by Hatfield-Dodds et al. (2007, p.8) 4, which focus on the prospects for rural Australians becoming valued service providers in Australia s low carbon future, we assign a conservative value of $15 C52 per ton of CO 2 e emitted. At the 2009 estimated feral camel population level, the value of CO 2 e emissions is thus approximately $14.50/animal per year, equivalent to $3.73m 4A Total benefits Given the above, the total present value of feral camel impacts (i.e. as a result of grazing competition, infrastructure damage, and methane emissions) in the absence of any control program is $100.79m I9 over 12 years. This is equivalent to $11.37m per year. Under control strategies 1 and 2 these impacts would be reduced to $13.25m I13 and $16.81m I24, respectively, over 12 years. Consequently, as can be seen in Table 2, the total present benefits of control strategy 1 are $87.54m I17 ($100.79m - $13.25m) over 12 years at a 5% discount rate (equivalent to an annualised value of $9.88m I18 ), of which $50.68m (57.9%) is related to reduced grazing competition, $1.62m (1.8%) is related to reduced infrastructure damage, and $35.24m (40.3%) 5 is related to the value of reduced methane emissions. As can be seen, the values of reduced grazing competition and methane emissions are both large, and individually several times greater than the $6.00m costs of control. Table 2: Present benefits under Strategy 1 with a 12-year time horizon and a 5% discount rate Type of benefit Present benefits ($) Proportion of total present benefits (%) Present benefits from reduced grazing competition % Present benefits from reduced infrastructure damage % Present benefits from reduced methane emissions % Total present benefits % Annualised present benefits A similar analysis for Strategy 2 reveals $83.98m I28 ($100.79m - $16.81m) of economic benefits, equivalent to an annualised value of $9.48m I29. 3 For simplicity, it is assumed that the camel population is made up largely of adults emitting methane at the levels stated above. 4 Hatfield-Dodds et al. (2007, p.8) present a range of estimates of the Australian carbon price associated with steady action to achieve significant reductions in emissions from 1990 or 2000 emission levels, along with a mid-range estimate of international carbon prices associated with feasible global action to avoid dangerous levels of climate change. While these different estimates reflect different levels of annual and cumulative emissions, they suggest a likely price range of $15 $65 in 2020 and $20 $75 in 2025, and an effective mid-term price floor of $15 $20 even with a very modest long-term emissions target or with offset sales targeting only overseas markets. 5 To generate these numbers, set two of the three values in C42, C49, and C52 to zero and then look up the result in I17. Desert Knowledge CRC 11

21 12 Desert Knowledge CRC

22 5. Discussion As can be seen in the summary presented in Table 3, the costs associated with the density sensitive Strategy 2 ($5.39m) are lower than those of Strategy 1 ($6.00m). The slightly higher cost effectiveness of Strategy 2 is related to the fact that control costs increase exponentially and, therefore, it is cheaper to only remove animals at higher densities (i.e. when their population reaches a density equal to, or greater than, 0.25 animals/km 2 ). This is in contrast to removing animals when they are always close to 0.1 animals/km 2, as is the case under Strategy 1. Nevertheless, the difference between the two strategies over 12 years is small. By contrast, the difference in the benefits of control between the two strategies is much larger. Total present benefits are $87.54m for Strategy 1 and $83.98m for Strategy 2. Benefits are lower under Strategy 2 as higher feral camel populations are tolerated between removal years. Under both strategies it is clear that the total present benefits far outweigh the total present costs of control. While infrastructure damage avoided represents a relatively small proportion of the benefits (1.8%), reduced grazing competition and reduced methane emissions are individually several times greater than the control costs. Hence, we observe a high net present value for both control strategies ($81.54m I20 for Strategy 1 and $78.95m I31 for Strategy 2, over 12 years). This is equivalent to an annualised value of $9.20m and $8.87m respectively. The benefit-cost ratio is greater than 14 in both cases. Table 3: Net present value of alternative control strategies Strategy 1 $ Benefit-cost Total present costs Total present cenefits Net present value Annualised net present value Benefit-cost ratio 14.6 Strategy 2 Total present costs Total present benefits Net present value Annualised net present value Benefit-cost ratio 15.6 Despite the fact that Strategy 2 was slightly more cost effective, given that the net present value 6 of Strategy 1 is larger than that of Strategy 2, it is clear that the former would be the preferred control strategy. As such, the remainder of the analysis in this report focuses only on Strategy 1. We now subject the above findings to a sensitivity analysis. 6 Although the benefit-cost ratio of Strategy 2 is larger than that of Strategy 1, the actual choice criterion should be based on the net present value figures. Desert Knowledge CRC 13

23 14 Desert Knowledge CRC

24 6. Sensitivity analysis The control costs and benefit estimates derived above are entirely dependent on the data provided by the relevant experts. Insofar as the results obtained can provide useful ball park figures upon which policy recommendations and future research priorities can be defined, it is useful to assess their robustness by exploring the degree to which the model results are driven by and sensitive to particular assumptions. Sensitivity analyses are thus carried out covering a range of factors, including feral camel population growth rates, aerial shooting costs, the degree to which feral animals compete with livestock for grazing resources, methane emission values, infrastructure constraints, discount rates, and time horizons. 6.1 Population growth and carrying capacity cap The rate at which the natural population growth of feral camels takes place is an important factor in determining total present control costs in the model. The annual natural population growth rate applied to the 2001 aerial survey population was 10% and a maximum central region carrying capacity of feral camels was assumed. We note that the model findings appear to be robust even under much higher population growth rates. For example, as can be seen in Table 4 (Model 1b), a 50% increase in the growth rate (i.e. to 15% per year 7 ) would lead to an increase in Strategy 1 total present costs of only $2.7m (from $6.00m in Model 1a to $8.70m). In addition to the fact that modelling such a 50% increase might lead to projected species growth occurring beyond maximum population growths observed in practice, we also note that such an increase in control costs of $2.7m would still be small relative to the estimated direct economic benefits of control. This finding holds true even at very low rates of feral camel population growth, and with both high and low carrying capacity caps. In addition, the baseline model (Model 1a) has been adapted to incorporate some of the main population data parameters and models analysed by McLeod and Pople (2008). Model 1c uses the parameters provided by McLeod and Pople s exponential model, while Model 1d uses the parameters provided by their logistic model. Models 2, 3, and 4 incorporate McLeod and Pople s exponential, logistic, and theta-logistic models, respectively, in place of the simple population model used in Model 1a. The population parameters used in the McLeod and Pople models are also used. As can be seen in Table 4, the alternative assumptions and underlying population models do have a large influence on present control costs and the net present value of control. Strategy 1 present costs of control vary from $3.67m (Model 4) to $4.95m (Model 2), while the respective net present values of control decline to $35.66m and $75.22m. 8 Relative to the Model 1a baseline, these reductions occur principally through the large changes in the estimated 2009 initial feral camel population (under in Model 4) and the estimated maximum carrying capacity ( under Model 3). Despite the sensitivity of the findings to the underlying assumptions made regarding feral camel population growth, and the clear need for improved monitoring and population data collection (Clive McMahon 2008, Research Fellow, Charles Darwin University, pers. comm.), it is apparent that the finding that the net present value of control is large (benefit-cost ratio larger than 10, even under Model 4) is robust. 7 To generate the numbers below, adjust C33. 8 To generate these numbers use the alternative spreadsheets provided and adjust C32, C33, and C35. Desert Knowledge CRC 15

25 Table 4: Present control costs and benefits under alternative feral camel population growth models and assumptions Model 1a (Baseline) Baseline (simple population model) Model 1b Model 1c Model 2 Model 1d Model 3 Model 4 Baseline variant (simple population model higher growth) Baseline variant (using exponential model data parameters) Exponential Population Model Baseline variant (using logistic model data parameters) logistic population model Theta-logistic population model Implied initial feral camel population in 2009 Implied initial feral camel density (animals/km 2 ) Growth rate per year (r or r_max) Maximum carrying capacity (k) Not binding ( ) Not binding ( ) Theta (θ) 0.79 No control 1. Total present value of feral camel impacts $ $ $ $ $ $ $ Total present cost of control $0 $0 $0 $0 $0 $0 $0 Control strategy 1 3.Total present value of feral camel impacts $ $ $ $ $ $ $ Total present cost of control $ $ $ $ $ $ $ Annualised present costs of control $ $ $ $ $ $ $ Total present benefits (1-3) $ $ $ $ $ $ $ Annualised present benefits $ $ $ $ $ $ $ Net present value of control ( ) $ $ $ $ $ $ $ Benefit cost ratio Control strategy 2 10.Total present value of feral camel impacts $ $ $ $ $ $ $ Total present cost of control $ $ $ $ $ $ $ Annualised present costs of control $ $ $ $ $ $ $ Total present benefits (1-10) $ $ $ $ $ $ $ Annualised present benefits $ $ $ $ $ $ $ Net present value of control ( )) $ $ $ $ $ $ $ Benefit cost ratio Desert Knowledge CRC

26 6.2 Aerial shooting costs Aerial shooting costs were assumed to increase in inverse proportion to the density of feral camels being removed. The cost levels used in this report are based on those for actual control programs conducted in different environments and at different target animal densities (refer to Saalfeld & Zeng 2008). These were estimated as $20/animal at densities of equal to or above 0.25 feral camels/km 2, $40/ animal at densities down to 0.15 feral camels/km 2, $60/feral camel for densities down to 0.1 animals/ km 2, and $110/feral camel below 0.1 animals/km 2. We note that in our model, aerial shooting cost levels are determined by the feral camel density at the beginning of each year. Given the non-continuous nature of the cost function used, this means that under Strategy 1 control costs in the first two years are at the $20 level and at the $60 level thereafter. This is because although in year 1 densities are brought down from 0.98 to 0.25 animals/km 2, by the beginning of year 2 natural growth rates have lifted the density back to 0.28 animals/km 2. Similarly, years 3 and onwards always start off with a density of 0.11 animals/km 2. As the target density is 0.1 animals/km 2 (achieved at the end of each year, year 3 onwards), no removals ever take place at the $110 level. It is therefore worth exploring the degree to which varying control costs and minimising the discontinuous nature of our cost curve might affect our overall findings. In the case of buffalo in Arnhem Land, Baylis and Yeomans (1989) argued that control costs increase exponentially according to the following formula: (3) where C is the cost per kill and D is the density/km 2. On the basis of this model, and taking into account that helicopter flying costs per hour appear to have increased substantially 9, aerial shooting costs would thus now be in the region of $92 $148 per animal killed at a density of 0.12 animals/km 2. This would be equivalent to a cost times greater than we have currently modelled. Accounting for an across-the-board increase in aerial shooting costs of 2.5 times 10 increases the total present cost of Strategy 1 to $14.37m (2.4 times that of our baseline). However, given that total present benefits are $87.54m, it is apparent that even much larger increases in aerial shooting costs will not have a significant effect on the large and positive net present value of control. Adjusting the target density downwards has a similar effect to increasing control cost levels. At levels that do not allow natural population growth to result in densities at a different cost level at the beginning of the following year, we can also smooth the discontinuous nature of our cost curve. At our baseline cost levels (i.e. $20, $40, $60, and $110), reducing the first year target density to animals/km 2 and subsequent year targets to animals/km 2 would increase Strategy 1 total present control costs to $7.00m, which is only $1.00m more than under our baseline assumptions. This result occurs as the vast majority (77.6%) of feral camel removal still takes place in year 1 at $20/feral camel. While removals in year 2 do take place at $60/feral camel and in year 3 at $110, relatively few camels are being removed. Total present costs of control are thus particularly sensitive to the cost levels at the initial very high 9 Bayliss estimated that helicopter costs in 1989 were approximately $220 per hour, which is equivalent to $500 in 2007 dollars. Given that helicopter costs are currently approximately $800, it appears that such costs have increased much faster than suggested by the ABS consumer price index for transport in general. 10 To generate the numbers below, adjust C9 C Adjust B Adjust C16 and D16. Desert Knowledge CRC 17