Water resource assessment for the Roper catchment Australia’s National Science Agency A report from the CSIRO Roper River Water Resource Assessment for the National Water Grid Editors: Ian Watson, Cuan Petheram, Caroline Bruce and Chris Chilcott ISBN 978-1-4863-1905-3 (print) ISBN 978-1-4863-1906-0 (online) Citation Watson I, Petheram C, Bruce C and Chilcott C (eds) (2023) Water resource assessment for the Roper catchment. A report from the CSIRO Roper River Water Resource Assessment for the National Water Grid. CSIRO, Australia. Chapters should be cited in the format of the following example: Petheram C, Bruce C and Watson I (2023) Chapter 1: Preamble: The Roper River Water Resource Assessment. In: Watson I, Petheram C, Bruce C and Chilcott C (eds) (2023) Water resource assessment for the Roper catchment. A report from the CSIRO Roper River Water Resource Assessment for the National Water Grid. CSIRO, Australia. Copyright © Commonwealth Scientific and Industrial Research Organisation 2023. To the extent permitted by law, all rights are reserved and no part of this publication covered by copyright may be reproduced or copied in any form or by any means except with the written permission of CSIRO. Important disclaimer CSIRO advises that the information contained in this publication comprises general statements based on scientific research. The reader is advised and needs to be aware that such information may be incomplete or unable to be used in any specific situation. No reliance or actions must therefore be made on that information without seeking prior expert professional, scientific and technical advice. To the extent permitted by law, CSIRO (including its employees and consultants) excludes all liability to any person for any consequences, including but not limited to all losses, damages, costs, expenses and any other compensation, arising directly or indirectly from using this publication (in part or in whole) and any information or material contained in it. CSIRO is committed to providing web accessible content wherever possible. If you are having difficulties with accessing this document, please contact Email CSIRO Enquiries . CSIRO Roper River Water Resource Assessment acknowledgements This report was funded through the National Water Grid’s Science Program, which sits within the Australian Government’s Department of Climate Change, Energy, the Environment and Water. Aspects of the Assessment have been undertaken in conjunction with the Northern Territory Government. The Assessment was guided by two committees: i. The Assessment’s Governance Committee: CRC for Northern Australia/James Cook University; CSIRO; National Water Grid (Department of Climate Change, Energy, the Environment and Water); NT Department of Environment, Parks and Water Security; NT Department of Industry, Tourism and Trade; Office of Northern Australia; Qld Department of Agriculture and Fisheries; Qld Department of Regional Development, Manufacturing and Water ii. The Assessment’s joint Roper and Victoria River catchments Steering Committee: Amateur Fishermen’s Association of the NT; Austrade; Centrefarm; CSIRO, National Water Grid (Department of Climate Change, Energy, the Environment and Water); Northern Land Council; NT Cattlemen’s Association; NT Department of Environment, Parks Australia; Parks and Water Security; NT Department of Industry, Tourism and Trade; Regional Development Australia; NT Farmers; NT Seafood Council; Office of Northern Australia; Roper Gulf Regional Council Shire Responsibility for the Assessment’s content lies with CSIRO. The Assessment’s committees did not have an opportunity to review the Assessment results or outputs prior to its release. This report was reviewed by Kevin Devlin (Independent consultant). For further acknowledgements, see page xxii. Acknowledgement of Country CSIRO acknowledges the Traditional Owners of the lands, seas and waters of the area that we live and work on across Australia. We acknowledge their continuing connection to their culture and pay our respects to their Elders past and present. Photo Looking along the Roper River at Red Rock, Northern Territory. Source: CSIRO – Nathan Dyer Part III Opportunities for water resource development Chapters 4 and 5 provide information on opportunities for agriculture and aquaculture in the Roper catchment. This information covers: • opportunities for irrigated agriculture and aquaculture (Chapter 4) • opportunities to extract and/or store water for use (Chapter 5). Mangoes in the Mataranka area Photo: CSIRO – Nathan Dyer 4 Opportunities for agriculture in the Roper catchment Authors: Chris Stokes, Ian Watson, Seonaid Philip, Tony Webster, Peter Wilson Chapter 4 presents information about the opportunities for irrigated agriculture and aquaculture in the catchment of the Roper River, describing: • land suitability for a range of crop group × season × irrigation type combinations and for aquaculture, including key soil-related management considerations • cropping and other agricultural opportunities, including crop yields and water use • gross margins at the farm scale • prospects for integration of forages and crops into existing beef enterprises • aquaculture opportunities. The key components and concepts of Chapter 4 are shown in Figure 4-1. Figure 4-1 Schematic diagram of agriculture and aquaculture enterprises as well as crop and/or forage integration with existing beef enterprises to be considered in the establishment of a greenfield irrigation development For more information on this figure please contact CSIRO on enquiries@csiro.au 4.1 Summary This chapter provides information on land suitability and the potential for agriculture and aquaculture in the Roper catchment. The approach used to generate the results presented in this chapter involves a mixture of field surveys and desktop analysis. For example, the land suitability results draw on extensive field visits (to describe, collect and analyse soils) that are integrated with state-of-the-art digital soil mapping. Many of the results are expressed in terms of potential. The area of land suitable for cropping or aquaculture for example, is estimated by considering the set of relevant soil and landscape biophysical attributes at each location and determining the most limiting attribute among them. It does not include water availability, cyclone or flood risk, or legislative, regulatory or tenure considerations, ecological, social or economic drivers that will inevitably constrain the actual area of land that is developed. Crops, forages and cropping systems results are based on data analysis and simulation models and assume good agronomic practices producing optimum yields given the soil and climate attributes in the catchment. Likewise, aquaculture is assessed in terms of potential, using a combination of land suitability and the productive capacity of a range of aquaculture species. Information is presented in a manner to enable the comparison of a variety of agricultural and aquaculture options. The results from individual components (land suitability, agriculture, aquaculture) are integrated to provide a sense of what is potentially viable in the catchment. This includes providing specific information on a wide range of crop types for agronomy, water use and land suitability for different irrigation types; analyses of economic performance, such as crop gross margins (GMs); how more intensive mixed cropping systems might be feasible with irrigation; and analyses of what is required for different aquaculture development options to be financially viable. 4.1.1 Key findings Major questions for any agricultural resource assessment are how much soil is suitable for a particular land use and where that soil is located. Based on a sample of 14 individual crop group × season of use × irrigation type combinations, the amount of land classified as ‘moderately suitable with considerable limitations’, or better, ranges from 106,000 ha (Crop Group 19, wet-season furrow) to 3.9 million ha (Crop Group 14, perennial species, spray) before constraints such as water availability, environmental and other legislation and regulations, and a range of biophysical risks are considered. In contrast with other catchments assessed in northern Australia, the Roper catchment has a relatively large percentage of soils classed as ‘Suitable, with minor limitations’, principally the red loamy soils of the Sturt Plateau. However, the Sturt Plateau has uncoordinated drainage patterns and where coordinated drainage features do exist, they do not flow reliability (Section 2.5.5). The plateau is, however, largely underlain by the regional-scale Cambrian Limestone Aquifer (Section 2.5.2). Dryland cropping Despite the theoretical possibility that dryland crops could be produced using the considerable rainfall that arrives during the wet season, in practice there are significant agronomic and market- related challenges to dryland crop production that have prevented its expansion. Loamy Kandosols have low water-holding capacity and are hard-setting, which makes consistently achieving viable yields difficult. Small areas of heavier clay soils along the Roper River and its major tributaries store more plant available water (PAW) that could support higher potential crop yields, particularly if cropped opportunistically in wetter years. However, frequent inundation and waterlogging of clay soils means that optimal farming operations would be disrupted, decreasing the chance that opportunistic high potential yields could be achieved in reality. Despite these challenges, higher value crops such as pulses or cotton show possible potential, especially if used to augment production from irrigated farming. Irrigated cropping Irrigation not only reduces crop water stress but also provides greater control over scheduling of crop operations to optimise production, including the option of growing through the cooler months of the dry season. Analyses of the performance of 19 potential irrigated cropping options in the Roper catchment indicate that achievable annual GMs could be up to about $4,000/ha for broadacre crops, $7,000 to $9,000 per ha for annual row crop horticulture, $11,000/ha for perennial fruit tree horticulture, and $3,000/ha for silviculture (plantation trees). While GMs are a key partial metric of farm performance, they should not be treated as fixed constants determined by the cropping system alone. They are a product of the farming and business management decisions, input costs and market opportunities. As such there are often niche opportunities to improve farm GMs and profitability, but these usually come at the expense of scalability. Farm financial metrics like GMs greatly amplify any fluctuations in commodity prices and input costs, such that the mean GM does not accurately reflect the often substantial cashflow challenges in managing years of losses between those of windfall profits (particularly for horticulture). Crop yields and GMs presented in this chapter are indicative of what might be attained for each cropping option once they have achieved their sustainable agronomic potential. It is unrealistic to assume that these levels of performance would be achieved in the early years of newly established farms, and allowance should be made for an initial period of learning (see Chapter 6). Potential crop species that could be grown as a single crop per year were rated and ranked for their performance in the Roper catchment. Wet-season crops (planted December to early March) that are rated the most likely to be viable are cotton (Gossypium spp.), forages and peanuts (Arachis hypogaea). Dry-season crops (planted late March to August) that are rated the most likely to be viable are annual horticulture, cotton and mungbean (Vigna radiata). Financial viability is determined not just by crop options with the highest GMs, but also depends on associated capital and fixed costs, which are higher in more intensive farming like horticulture. The farm-scale measures of crop performance presented in this chapter are intended to be used in conjunction with the scheme-scale analyses of financial viability in Chapter 6 (as part of an integrated multi- scale approach). Sequential cropping systems involve planting more than one crop in the same year in the same field. These systems have the potential to significantly increase farm GMs. Annual broadacre and horticultural crops have been grown sequentially for many decades in tropical northern Australia. A wide range of sequential cropping options are potentially viable in the Roper catchment. Most suitable crop sequences include wet-season cotton, dry-season cotton, dry-season annual horticulture, or a dry-season forage. Scheduling back-to-back crops could be operationally tight in the Roper catchment, particularly on clay-rich soils with poor drainage. Crop selection is market driven in northern Australian regions like the Roper catchment, so rotations and crop sequences are dynamic, as growers develop an understanding of the benefits, trade-offs and management needs of different crop mixes, and adapt to changing opportunities. Aquaculture There are considerable opportunities for aquaculture development in northern Australia given its natural advantages of a climate suited to farming valuable tropical species, the large areas identified as suitable for aquaculture, political stability and proximity to large global markets. The main challenges to developing and operating modern and sustainable aquaculture enterprises are regulatory barriers, global cost competitiveness, and the remoteness of much of the suitable land area. The three species with the most aquaculture potential in the Roper catchment are black tiger prawns (Penaeus monodon), barramundi (Lates calcarifer), and red claw (Cherax quadricarinatus). Land suitability for lined ponds for freshwater species is widespread throughout the catchment due to the extensive distribution of favourable soil and land characteristics (flat land, non-rocky, deep), whereas for freshwater species in earthen ponds, options are restricted to the impermeable alluvial clays to allow retention of water. Marine aquaculture’s range is restricted to the tidal zones of the catchment on the coastal plain where access to marine water is within 2000 m. High annual operating costs (which can exceed the initial capital costs of development) mean that managing cash flow in the establishment years is challenging, especially for products that require multi-year grow-out periods. Input costs scale with increasing productivity, so improving production efficiency (such as feed conversion rate or labour-efficient operations) is much more important than increasing yields for aquaculture to be viable in the Roper catchment. It would be essential for any new aquaculture development to refine the production system and achieve the required levels of operational efficiency (input costs per kg of produce) using just a few ponds before scaling the enterprise (to a larger number of ponds). 4.1.2 Introduction Aspirations to expand agricultural development in the Roper catchments are not new and across northern Australia there have been a number of initiatives to put in place large-scale agricultural developments since the Second World War (Ash, 2014; Ash and Watson, 2018). Ash and Watson (2018) assessed 11 such agricultural developments, 4 of which continue to operate at a regionally relevant scale. Their assessment included both irrigated and dryland developments and considered natural capital, human capital, physical capital, financial capital and social capital. Key points to emerge from these analyses include the following: • The natural environment (climate, soils, pests and diseases) makes agriculture in northern Australia challenging, but these inherent environmental factors are not generally the primary reason for a lack of success. • The speed with which many of the developments occurred did not allow for a ‘learning by doing’ approach, leading at times to costly mistakes. • Physical capital, in the form of on-farm infrastructure, supply chain infrastructure and crop varieties, was a significant and ongoing impediment to success. For broadacre commodities that require processing facilities, these facilities need to be within a reasonable distance of production sites and at a scale to make them viable in the long term. • Financial plans tended to over-estimate early production and returns on capital and assumed overly optimistic expectations of the ability to scale up rapidly. This led to financial pressure on investors and a premature end to some developments. Furthermore, the need to have well- connected and well-paying markets was often not fully appreciated. In more remote regions, higher value products such as fruit, vegetables and niche crops proved more successful, although high supply chain costs to both domestic and export markets remain as impediments to expansion. • Most of the developments began in areas with no history of agricultural development and there was little in the way of a community of practitioners who could share experiences. • Management, planning and finances were the most important factors in determining the ongoing viability of agricultural developments. For developments to be successful, all factors relating to climate, soils, agronomy, pests, farm operations, management, planning, supply chains and markets need to be thought through in a comprehensive systems design. Particular attention needs to be paid to scaling up at a considered pace and being prepared for reasonable lags before positive returns on investment are achieved. This chapter seeks to address the following questions for the Roper catchment: ‘How much land is suitable for cropping and in which suitability class?’, ‘Is irrigated cropping economically viable?’, ‘Which crop options perform best and how can they be implemented in viable mixed farming systems?’, ‘Can crops and forages be economically integrated with beef enterprises?’ and ‘What aquaculture production systems might be possible?’. The chapter is structured as follows: • Section 4.2 describes how the land suitability classes are derived from the attributes provided in Chapter 2, with results given for a set of 14 combinations of individual crop group × season × irrigation type. Versatile agricultural land is described and a qualitative evaluation of cropping is provided for a set of specific locations within the catchment. • Section 4.3 provides detailed information on crop and forage opportunities, including irrigated yields, water use and GMs. Agronomic principles, such as selection of sowing time, are provided including a cropping calendar for scheduling farm operations. The information is synthesised in an analysis of the cropping systems that could best take advantage of opportunities in Roper catchment environments while dealing with farming challenges. • Section 4.4 provides synopses for 11 crop and forage groups, including a focus on specific example species. • Section 4.5 discusses the candidate species and likely production systems for aquaculture enterprises, including the prospects for integrating aquaculture with agriculture. 4.2 Land suitability assessment 4.2.1 Introduction The overall suitability for a particular land use is determined by a number of environmental and soil attributes. These include, but are not limited to, climate at a given location, slope, drainage, permeability, plant available water capacity (PAWC), pH, soil depth, surface condition and texture. Examples of some of these attributes are provided in Section 2.3. From these attributes a set of limitations to suitability are derived, which are then considered against each potential land use. Note that the use of the term ‘suitability’ in the Assessment refers to the potential of the land for a specific land use such as furrow-irrigated cotton, while the term ‘capability’ (not used in the Assessment) refers to the potential of the land for broadly defined land uses, such as cropping or pastoral (DSITI and DNRM, 2015). 4.2.2 Land suitability classes The overall suitability for a particular land use is calculated by considering the set of relevant attributes at each location and determining the most limiting attribute among them. This most limiting attribute then determines the overall land suitability classification. The classification is on a scale of 1 to 5 from ‘Suitable with negligible limitations’ (Class 1) to ‘Unsuitable with extreme limitations’ (Class 5) as shown in Table 4-1 (FAO, 1976; 1985). The companion technical report on digital soil mapping and land suitability (Thomas et al., 2022) provides a complete description of the land suitability assessment method and the material presented below is taken from that report. Note that for the land suitability maps and figures presented in this section there is no consideration of flooding, risk of secondary salinisation or availability of water as discussed by Thomas et al. (2022). Consideration of these risks and others, along with further detailed soil physical, chemical and nutrient analyses would be required to plan development at scheme, enterprise or property scale. Caution should therefore be employed when using these data and maps at fine scales. Table 4-1 Land suitability classes based on FAO (1976, 1985) as used in the Assessment For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 4.2.3 Land suitability for crops, versatile agricultural land and evaluation of specific areas of interest The suitability framework used in this Assessment aggregates individual crops into a set of 21 crop groups (Table 4-2). The groups are based on the framework used by the Northern Territory Government (Andrews and Burgess, 2021), with some additions considered prospective based on previous CSIRO work in northern Australia (e.g. Thomas et al., 2018). From this set of crop groups, land suitability has been determined for 58 land use combinations of crop group × season × irrigation type (including rainfed) (Thomas et al., 2022). Table 4-2 Crop groups (1 to 21) and individual land uses evaluated for irrigation (and rainfed) potential Land uses are based on those used by Andrews and Burgess (2021) with amendment for the Roper catchment with the addition of crop groups 18–21, based on CSIRO’s previous work in northern Australia, including those used in the Northern Australia Water Resource Assessment (Thomas et al., 2018) which are in boldface. MAJOR CROP GROUP CROP GROUP INDIVIDUAL CROPS ASSESSED Tree crops/horticulture (fruit) 1 Monsoonal tropical tree crops (0.5 m root zone) – mango, coconut, dragon fruit, Kakadu plum, bamboo, lychee 2 Tropical citrus – lime, lemon, mandarin, pomelo, lemonade, grapefruit Intensive horticulture (vegetables, row crops) 3 Cucurbits – watermelon, honeydew melon, rockmelon, pumpkin, cucumber, Asian melons, zucchini, squash 4 Fruiting vegetable crops – Solanaceae (capsicum, chilli, eggplant, tomato), okra, snake bean, drumstick tree 5 Leafy vegetables and herbs – kangkong, amaranth, Chinese cabbage, bok choy, pak choy, choy sum, basil, coriander, dill, mint, spearmint, chives, oregano, lemon grass, asparagus Root crops 6 Carrot, onion, sweet potato, shallots, ginger, turmeric, galangal, yam bean, taro, peanut, cassava Grain and fibre crops 7 Cotton, grains – sorghum (grain), maize, millet (forage) 8 Rice (lowland and upland) Small-seeded crops 9 Hemp, chia, quinoa, medicinal poppy Pulse crops (food legumes) 10 Mungbean, soybean, chickpea, navy bean, lentil, guar Industrial 11 Sugarcane Hay and forage (annual) 12 Annual grass hay/forages – sorghum (forage), maize (silage) 13 Legume hay/forages – blue pea, burgundy bean, cowpea, lablab, Cavalcade, forage soybean Hay and forage (perennial) 14 Perennial grass hay/forage – Rhodes grass, panics Silviculture/forestry (plantation) 15 Indian sandalwood 16 African mahogany, Eucalyptus spp., Acacia spp. 17 Teak Intensive horticulture (vegetables, row crops) 18 Sweetcorn MAJOR CROP GROUP CROP GROUP INDIVIDUAL CROPS ASSESSED Oilseeds 19 Sunflower, sesame Tree crops/horticulture 20 Banana, coffee 21 Cashew, macadamia, papaya A sample of 14 of these individual land use combinations is shown in Figure 4-2. Depending on land use, the amount of land classified as Class 3 or better for these sample land uses ranges from almost 106,000 ha (Crop Group 19 under wet-season furrow irrigation) to closer to 4 million ha (Crop Group 14 under spray irrigation). Much of this land is rated as Class 3, and so has considerable limitations, although there are nearly 1.7 million ha of Class 2 land available for Crop Group 14 crops under spray irrigation and between about 450,000 ha and about 600,000 ha of Class 2 land for the other crop groups under spray or trickle irrigation. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 4-2 Area (ha) of the Roper catchment mapped in each of the land suitability classes for 14 selected land use options A description of the five land suitability classes is provided in Table 4-1 and more detail on the crop groups is found in Table 4-2. In order to provide an aggregated summary of the land suitability products, an index of agricultural versatility was derived for the Roper catchment (Figure 4-3). Versatile agricultural land was calculated by identifying where thehighest number of the 14 selected land use optionspresented in Figure4-2were mapped asbeing suitable (i.e. suitability classes 1 to 3). Qualitative observationson each ofthe areas mapped as ‘A’ to ‘E’ inFigure4-3are provided in Table4-3. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure4-3Agriculturalversatilityindex map for theRopercatchment High index values denote land that is likely to besuitable for more of the 14 selected land use options. The map alsoshowsspecific areas of interest(A to E) from a land suitability perspective, discussed in Table4-3. Note that this mapdoes not take into consideration flooding, risk of secondarysalinisation or availability of water. 194|Water resource assessment forthe Ropercatchment Table 4-3 Qualitative land evaluation observations for locations in the Roper catchment shown in Figure 4-3 Further information on each soil generic group (SGG) and a map showing spatial distribution can be found in Section 2.3. For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Land suitability and its implications for crop management are discussed in more detail for a selection of crops in Section 4.4, where land use suitability of a given crop and irrigation combination are mapped, along with information critical to the consideration of the crop in an irrigated farm enterprise. Land suitability maps for all 58 land use combinations are presented in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). 4.3 Crop and forage opportunities in the Roper catchment 4.3.1 Introduction This section presents results on the farm ‘performance’ of individual crop options, where ‘performance’ is quantified specifically as crop yields, the amount of applied irrigation water (including on-farm losses), and GMs. This is presented together with information on agronomic principles and farming practices to help interpret the viability of new (greenfield) farming opportunities in the Roper catchment. The individual crop options are grouped into dryland broadacre, irrigated broadacre, horticulture, and silviculture (sections 4.3.3 to 4.3.7). The viability of these options is then discussed in a section on cropping systems (Section 4.3.8), that considers the mix of farming opportunities and practices that have most potential to be profitably and sustainably integrated within Roper catchment environments, for both single and sequential cropping systems. The final section evaluates the viability of integrating irrigated forages into existing beef production (Section 4.3.9). These farm-scale analyses are intended to be used in conjunction with the scheme-scale analyses of viability in Chapter 6 (as part of an integrated multi- scale analysis). Nineteen irrigated crop options were selected to evaluate their potential performance in the Roper catchment (Table 4-4). The crops were selected to be compatible with the land suitability crop groups (Table 4-2), provided that they had the potential to be viable in the Roper catchment (based on knowledge of how well these crops grow in other parts of Australia), were of commercial interest for possible development in the region, and there was sufficient information on their agronomy and farming costs/prices for quantitative analysis. The analyses used a combination of Agricultural Production Systems sIMulator (APSIM) crop modelling and climatically informed extrapolation to estimate potential yield and water use for each crop, and those values were then used in a farm GM tool specifically designed for greenfield farming developments (like the Roper catchment, where there are very few existing commercial farms or farm financial models). In particular, extrapolations made use of close similarities in climate and soils between possible cropping locations in the Roper catchment and established irrigated cropping regions at similar latitudes near Katherine (NT) and the Ord River Irrigation Area (WA) (Figure 4-4). Full details of the approach are described in the companion technical report on agricultural viability and socio-economics (Stokes et al., 2023). Section 4.4 provides further details on opportunities and constraints in the Roper catchment for example crops in each of the agronomic ‘crop types’ listed in Table 4-4. Table 4-4 Crop options where performance was evaluated in terms of water use, yields and gross margins The methods used for estimating crop yield and irrigation water requirements are coded as: A = APSIM; E = climatically informed extrapolation. Where two letters are used, the first is the primary method, and the second is used for sensibility testing (A, E) or applying adjustments (E, A; with adjustment multipliers shown in parentheses where the APSIM median was more than 10% outside the range of sensibility testing estimates). Mango (KP) is Kensington Pride, and Mango (PVR) is an indicative new high-yielding variety, likely to have plant variety rights (e.g. Calypso). Note that crops that are agronomically similar, in terms of the commodities they produce (as categorised below), may differ in how they respond to soil constraints. The ‘crop type’ categories below are therefore necessarily different to those used in the land suitability section (that grouped crops according to shared soil requirements and constraints: Table 4-2). CROP TYPE CROP IRRIGATION WATER ESTIMATE METHOD YIELD ESTIMATE METHOD Broadacre crops Cereal Sorghum (grain) A, E A, E Pulse Mungbean E, A (1.30) A, E Chickpea E, A (1.19) E, A (1.14) Soybean A, E A, E Oilseed Sesame E E Peanut A, E A, E Industrial Cotton (dry season) E, A (1.69) A, E Cotton (wet season) E, A (1.34) E, A (1.24) Hemp E E Forage Rhodes grass A, E A, E Horticulture (row) Rockmelon E E Watermelon E E Onion E E Capsicum E E Horticulture (tree) Mango (PVR) E E Mango (KP) E E Lime E E Plantation tree African mahogany E E Sandalwood E E (a) Mean monthly rainfall (b) Mean daily maximum temperature (c) Mean daily solar radiation (d) Mean daily minimum temperature For more information on this figure please contact CSIRO on enquiries@csiro.au 050100150200250SepOctNovDecJanFebMarAprMayJunJulAugmm For more information on this figure please contact CSIRO on enquiries@csiro.au 010203040JanFebMarAprMayJunJulAugSepOctNovDec°C For more information on this figure please contact CSIRO on enquiries@csiro.au 0510152025JanFebMarAprMayJunJulAugSepOctNovDecMJ/m2/day For more information on this figure please contact CSIRO on enquiries@csiro.au 010203040JanFebMarAprMayJunJulAugSepOctNovDec°C KatherineOrdBulmanMatarankaNgukurrLarrimahDaly Waters Figure 4-4 Climate comparisons of Roper sites versus established irrigation areas at Katherine (NT) and Ord River (WA) Roper catchment locations are Bulman, Mataranka, Ngukurr, Larrimah and Daly Waters. Three locations were selected for the APSIM simulations to represent some of the best potential farming conditions across the varied environments available in the Roper catchment: • a sandy Kandosol, locally called Blains, (SGGs 4.1 and 4.2, marked A in Figure 4-3, 79 mm PAWC for grain sorghum (Sorghum bicolor), as an indicator crop) with a Mataranka climate (14.92 °S, 133.07 °E, mean annual rainfall of about 950 mm) • a Vertosol (SGGs 2 and 9, marked B in Figure 4-3, 212 mm PAWC for grain sorghum) with a Ngukurr climate (14.73 °S, 134.73 °E, mean annual rainfall of about 850 mm) • a Dermosol (SGG 2, marked C and D in Figure 4-3, 156 mm PAWC for grain sorghum) with a Bulman climate (13.66 °S, 134.33 °E, mean annual rainfall of about 1000 mm). To assist with interpreting the later results, some information is first provided on agronomic principles related to the scheduling of critical farm operations such as sowing and irrigation in relation to Roper catchment environments. 4.3.2 Cropping calendar and time of sowing Time of sowing can have a significant effect on achieving economical crop and forage yields, and on the availability and amount of water for irrigation required to meet crop demand. Cropping calendars identify optimum sowing times of different crops and are essential tools for scheduling farm operations (Figure 4-5) so that crops can be reliably and profitably grown. Prior to the Assessment, no cropping calendar existed for the Roper catchment. Sowing windows vary in both timing and length among crops and regions and consider the likely suitability and constraints of weather conditions (e.g. heat and cold stress, radiation, and conditions for flowering, pollination and fruit development) during each subsequent growth stage of the crop. Limited field experience currently exists in the Roper catchment for the majority of crops and forages evaluated. This cropping calendar (Figure 4-5) is therefore extrapolated from knowledge of crops derived from past and current agricultural experience in the Ord River Irrigation Area (WA), Katherine and Douglas–Daly regions (NT), and the Burdekin region (Queensland). Some annual crops have both wet season (WS) and dry season (DS) cropping options. Perennial crops are grown throughout the year, so growing seasons and planting windows are less well defined. Generally, perennial tree crops are transplanted as small plants, and in northern Australia this is usually timed towards the beginning of the wet season to take advantage of wet-season rainfall. The cropping calendar presented here considers only the optimal climate conditions for the crop growth and is intended to be used to together with considerations of other local-specific operational constraints. Such constraints would include wet-season difficulties in access and trafficability and limitations on the number of hectares per trafficable day that available farm equipment can sow/plant. For example, clay-rich alluvial Vertosols, such as those found along the Roper River and its major tributaries, are likely to present severe trafficability constraints throughout much of the wet season in the Roper catchment, while sandier Kandosols would present far fewer trafficability restriction in scheduling farming operations (Figure 4-6). Many suitable annual crops can be grown at any time of the year with irrigation in the Roper catchments. Optimising crop yield alone is not the only consideration. Ultimately, sowing date selection must balance the need for the best growing environment (optimising solar radiation and temperature) with water availability, pest avoidance, trafficability during the season and at harvest, crop rotation, supply chain requirements, infrastructure development costs, market access considerations, and potential commodity price. Many summer crops from temperate regions are suited to the tropical dry season (winter) because temperatures are closer to their optima and/or there is more consistent solar radiation (e.g. maize (Zea mays), chickpea (Cicer arietinum) and rice (Oryza sativa)). For sequential cropping systems (that grow more than a single crop in a year in the same field), growing at least one crop partially outside its optimal growing season can be justified if total farm profit per year is increased and there are no adverse biophysical consequences (e.g. pest build-up). For more information on this figure please contact CSIRO on enquiries@csiro.au CROP TYPECROPDECJANFEBMARAPRMAYJUNJULAUGSEPOCTNOVCROP DURATION(days) Cereal cropsSorghum (WS)ssssssgggg110—140Sorghum (DS)gssssssssssggg110—140Maize (WS)ssssssssgggg110—140Maize (DS)ssssssgggg110—140Rice (WS)ssssgggg120—160+ Rice (DS)ssssgggg90—135 Pulse crops (food legumes)Mungbean (WS)ssssssggg70—85Mungbean (DS)ssssggg70—85Chickpeassssgggg100—120OilseedsSoybean (WS)ssssssgggg110—130Sesamessssgggg110—130Root cropsPeanut (WS)ssssssggggg100—140Peanut (DS)gssssgggg100—140Cassavassssssssssssssggggg180—210Industrial cropsCotton (WS)ssssssgggg100—120Cotton (DS)ssssgggg100—120Hemp (fibre)ssssssssgggg110—150Forage, hay, silageRhodes grassggspspspgggspspspspPerennial (regrows) Forage sorghumssssssssgggssssssgg60—80 (regrows) Forage milletssssssssgggssssssgg60—80 (regrows) Forage maizegssssssgggssssssgg75—90Forage legumesCavalcadessggggggssss150—180Lablabssssssssssggggg130—160Horticulture (row crops)Melonsssssssgggg70—110Oniongssssssssssgggg130—160Capsicum, chilli, tomatossssggggg70—90 from transplantPineapplespspspgggggggPerennialHorticulture (vine)Table grapesspspspgggggggggPerennialHorticulture (tree crops)MangospspspgggggggggPerennialAvocadospspspgggggggggPerennialBananaspspspspggggggggPerennialLimespspspgggggggggPerennialLemonspspspgggggggggPerennialOrangespspspgggggggggPerennialCashewspspspgggggggggPerennialMacadamiaspspspgggggggggPerennialPlantation trees (silviculture)Africian mahoganyspspspgggggggggPerennialIndian sandalwoodspspspgggggggggPerennialSowing window forannual cropsGrowingperiodFallowSowing window for perennial cropsLikely sowing period Figure 4-5 Annual cropping calendar for irrigated agricultural options in the Roper catchment WS = wet season; DS = dry season. For more information on this figure please contact CSIRO on enquiries@csiro.au 0% 20% 40% 60% 80% 100% 1–Jan1–Feb1–Mar1–Apr1–May1–Jun1–Jul1–Aug1–Sep1–Oct1–Nov1–Dec% of years PAW is below thresholdKandosol 80% thresholdKandosol 70% thresholdVertosol 80% thresholdVertosol 70% threshold Figure 4-6 Soil wetness indices that indicate when seasonal trafficability constraints are likely to occur on Kandosols (sandy) and Vertosols (high clay) with a Bulman climate The indices show the proportion of years (for dates at weekly intervals) when plant available water (PAW) in the top 30 cm of the soil is below two threshold proportions (70% and 80%) of the maximum PAW value. Lower values indicate there would be fewer days at that time of year when fields would be accessible and trafficable. Estimates are from 100-year APSIM simulations without a crop: in actual farming situations, once a crop canopy is established later in the season, crop water extraction from the soil would assist in alleviating these constraints. Growers also manage time of sowing to optimally use stored soil water and in-season rainfall, and to avoid rain damage at maturity. Access to irrigation provides flexibility in sowing date and in the choice and timing of crop or forage systems in response to seasonal climate conditions. Depending on the rooting depth of a particular species and the length of growing season, crops established at the end of the wet season may access a full profile of soil water (e.g. 200+ mm PAWC for some Vertosols). While timing of sowing to maximise available water can reduce the overall irrigation requirement, it may expose crops to periods of lower solar radiation and extreme temperatures during plant development and flowering. It may also prevent the implementation of a sequential cropping system. 4.3.3 Dryland cropping Dryland cropping (crops grown without irrigation, relying only on rain) has been attempted by farmers in the NT for almost 100 years, yet only small areas of dryland crop production currently occur each year. This indicates that despite the theoretical possibility that dryland crops could be produced using the significant rainfall that occurs during the wet season in the Roper catchment, in practice there are significant agronomic and market-related challenges to dryland crop production that have prevented its expansion to date. Without the certainty provided by irrigation, dryland cropping is opportunistic in nature, relying on favourable conditions in which to establish, grow and harvest a crop. The annual cropping calendar in Figure 4-5 shows that, for many crops, the sowing window includes the month of February. For relatively short-season crops, such as sorghum and mungbean, this coincides with both the sowing time that provides close to maximum crop yield and the time at which the season’s water supply can be most reliably assessed with a high degree of confidence. Table 4-5 shows how plant available soil water content at sowing and subsequent rainfall in the 90 days after each sowing date varies over three different sowing dates for a Vertosol in the Roper catchment at Bulman. As sowing is delayed from February to April, the amount of stored soil water increases. However, there is a significant decrease in rainfall in the subsequent 3 months after sowing. Combining the median PAW in the soil profile at sowing, and the median rainfall received in the 90 days following sowing, provides totals of 581, 464 and 311 mm for the February, March and April sowing dates, respectively. For ‘drier than average years’ (80% probability of exceedance), the soil water stored at sowing and the expected rainfall in the ensuing 90 days (<460 mm) would result in water stress and comparatively reduced crop yields. In ‘wetter than average years’ (20% probability of exceedance) the amount of soil water at the end of February combined with the rainfall in the following 90 days (764 mm) is sufficient to grow a good short-season crop (noting that the timing of rainfall is also important since some rain is ‘lost’ to runoff, evaporation and deep drainage between rainfall events). Opportunistic dryland cropping would target those wetter years where PAW at the time of sowing indicated a higher chance of harvesting a profitable crop. Table 4-5 Soil water content at sowing, and rainfall for the 90-day period following sowing for three sowing dates, based on a Bulman climate on Vertosol PAW = plant available water stored in soil profile. The 80%, 50% (median) and 20% probability of exceedance values are reported, for the 100 years between 1920 and 2020. The lower-bound values (80% exceedance) occur in most years, while the upper-bound values only occur in the most exceptional upper 20% of years. SOWING DATE PAW AT SOWING DATE (mm) RAINFALL IN 90 DAYS FOLLOWING SOWING DATE (mm) TOTAL STORED SOIL WATER + RAINFALL IN SUBSEQUENT 90 DAYS (mm) 80% 50% 20% 80% 50% 20% 80% 50% 20% 1 February 80 151 212 299 424 614 457 581 764 1 March 143 228 305 146 250 405 354 464 617 1 April 193 274 299 16 53 128 269 311 393 Figure 4-7 highlights the impact on dryland crop yields of the diminishing water availability and increasing evapotranspiration as the season progresses. This constraint is much more severe for sandier soils that have less capacity to store plant available water (like Kandosols in the Roper catchment: Figure 4-7a), than finer textured soils (like the alluvial Vertosols in the Roper catchment: Figure 4-7b). However, the frequent inundation and waterlogging of clay soils (Figure 4-6) means that crops cannot always be sown at optimum times, fertiliser can be lost due to runoff, drainage and denitrification, and in-crop management (e.g. for weed, disease and insect control) cannot be undertaken cost-effectively with ground-based equipment in a timely manner, a critical requirement for dryland crop production to succeed. Those disruptions decrease the chance that high potential yields in the top 20% of the seasons could be achieved in practice. (a) Bulman Kandosol (sandy, PAWC 79 mm) (b) Bulman Vertosol (high clay, PAWC 212 mm) For more information on this figure please contact CSIRO on enquiries@csiro.au 0123456701-Jan15-Jan01-Feb15-Feb01-Mar15-Mar01-Apr15-AprYield (t/ha) Sow dateRangeMedian For more information on this figure please contact CSIRO on enquiries@csiro.au 0123456701-Jan15-Jan01-Feb15-Feb01-Mar15-Mar01-Apr15-AprYield (t/ha) Sow dateRangeMedian Figure 4-7 Influence of planting date on dryland grain sorghum yield at Bulman for (a) a Kandosol and (b) a Vertosol Estimates are from APSIM simulations with planting dates on the 1st and 15th of each month. PAWC values give the plant available water capacity that each soil profile can store (for sorghum). The shaded band around the median line indicates the 80% to 20% exceedance probability range in year-to-year variation. Soil is rarely uniform within a single paddock, let alone across entire districts. Without the homogenising input of irrigation to alleviate water limitations (and associated high inputs of fertilisers to alleviate nutrient limitations), yields from low-input dryland cropping are typically much more variable (both across years and locations) than yields from irrigated agriculture. Furthermore, the capacity of the soil to supply stored water varies not only with soil type, but also depends on crop type and variety because each crop’s root system has a differing ability to access water, particularly deep in the profile. This makes it harder to make generalisations about the viability of dryland cropping in the Roper catchment as farm performance (e.g. yields and GMs) is much more sensitive to slight variations in local conditions. Rigorous estimates of dryland crop performance would require detailed localised soil mapping and crop trials before investment decisions could be confidently made. Despite the challenges described above, recent efforts in the NT have identified potential opportunities for dryland farming using higher value crops, such as pulses or cotton. A preliminary APSIM assessment of the potential for dryland cotton in the region suggested that mean lint yields of 2.5 to 3.5 bales/ha may be possible at a range of locations in the vicinity of the Roper catchment (Yeates and Poulton, 2019). However, there was very high variability in median yields between farms (1 to 5 bales/ha), depending on management and soil type. 4.3.4 Irrigated crop response and performance metrics Crops that are fully irrigated can yield substantially more than dryland crops. Figure 4-8 shows how yields for grain sorghum grown on a Kandosol in the Roper catchment increase as more water becomes available to alleviate water limitations and meet increasing proportions of crop demand. With sufficient irrigation, yields are highest for (wet-season sown) crops grown over the dry season when radiation tends to be less limiting (plateau of Figure 4-8a versus b). For wet-season sowing, unirrigated yields can approach fully irrigated yields in good years (yields exceeded in the top 20% of years, marked by the upper shaded range in Figure 4-8a). However, irrigation allows greater flexibility in sowing dates, allows sowing in the dry season too (for crops that would then grow through the wet season), and generates more reliable (and higher median) yields. The simulations did not seek to ‘optimise’ supplemental irrigation strategies in years where available water was insufficient to maximise crop yields: irrigators would need to make those decisions in years where available water was insufficient to fully meet crop demand. A key advantage of irrigated dry-season cropping in northern Australia is that the availability of water in the soil profile and surface water storages for growing the crop is largely known at the time of planting (near the start of the wet season: Table 4-5). This means irrigators have good advance knowledge for planning how much area to plant, which crops to grow and what irrigation strategies to use, particularly in years where they have insufficient water to fully irrigate all fields. A mix of irrigation approaches could be used, such as expanding the scale of a core irrigated cropping area with other less-intensively farmed areas, opportunistic dryland cropping, opportunistic supplemental irrigation, opportunistic sequential cropping, and/or adjusting the area of fully irrigated crops grown to match available water supplies that year. (a) 1 February sowing (wet season) (b) 1 August sowing (dry season) For more information on this figure please contact CSIRO on enquiries@csiro.au 01234567012345Yield (t/ha) Available irrigation water (ML/ha) RangeMedian For more information on this figure please contact CSIRO on enquiries@csiro.au 01234567012345Yield (t/ha) Available irrigation water (ML/ha) RangeMedian Figure 4-8 Influence of available irrigation water on grain sorghum yields for planting dates (a) on 1st February and (b) 1st August, for a Kandosol with a Bulman climate Estimates are from 100-year APSIM simulations. The shaded band around the median line indicates the 80% to 20% exceedance probability range in year-to-year variation. Dryland production is indicated by the zero point where no allocation is available for irrigating. Measures of farm performance (in terms of yields, water use and GMs) are presented for the 19 cropping options that were evaluated (Table 4-4). Given the limited commercial irrigated farming that currently occurs in the Roper catchment to provide real world data, estimates of crop water use and yields should be considered as indicative, and to have at least a 20% margin of error at the catchment scale (with further variation expected between farms and fields). The measures of performance should be considered as an upper bound of what could be achieved under best- practice management after learning and adapting to location conditions. GMs are a key partial metric of farm performance but should not be treated as fixed constants determined by the cropping system alone. They are a product of the farming and business management decisions made by individual farmers, input prices, commodity prices and market opportunities. As such, the GMs presented below should be treated as indicative of what might be attained for each cropping option once their sustainable agronomic potential has been achieved. Any divergence from assumptions about yields and costs would flow through to GM values, as would the consequences of any underperformance or overperformance in farm management. It is unrealistic to assume that the levels of performance in the results below would be achieved in the early years of newly established farms, and allowance should be made for an initial period of learning when yields and GMs are below their potential (see Chapter 6). Collectively however, the GMs and other performance metrics presented here provide an objective and consistent comparison across a suite of likely cropping options for the Roper catchment and an indicative maximum performance that could be achievable for greenfield irrigated development for each of the groupings of crops below. 4.3.5 Irrigated broadacre crops Table 4-6 shows the farm performance (yields, water use and GMs) for the ten broadacre cropping options that were evaluated. For crops that were simulated with APSIM, estimates are provided for locations with three different soil types associated with climates in the Roper catchment (Kandosol at Mataranka, Vertosol at Ngukurr, and Dermosol at Bulman) and include measures of variability (expressed in terms of years with yield exceedance probabilities of 80%, 50% (median) and 20%). For other crops, yield and water use estimates (and resulting GMs) were estimated based on expert experience and climatically informed extrapolation from the most similar analogue locations in northern Australia where commercial production currently occurs. The broadacre cropping options with the best GMs (>$1500/ha) were cotton (both wet- and dry- season cropping), forages (Rhodes grass (Chloris gayana)), and peanuts. These suggest GMs of $4000 to $5000 might be achievable for broadacre cropping in the Roper catchment, although not necessarily at scale. Mungbean, chickpea and industrial hemp (Cannabis sativa ssp. sativa) had intermediate GMs (about $1000/ha). The GMs for the sorghum, soybean (Glycine max) and sesame (Sesamum indicum) were low (<$800/ha in most cases). Simulated yields (and consequent GMs) were generally lowest on the Kandosol and highest on the Vertosol because of the increased buffering capacity that a high PAWC clay soil provides against hot weather that triggers water stress even in irrigated crops. The Dermosol yields and GMs were slightly lower than the Vertosol due to its lower PAWC. It was not possible to model cotton on the Vertosol in APSIM because of the difficulty in replicating the nuances of managing waterlogging on inter-furrow mounds on these heavy clay soils, and the sensitivity of cotton roots to waterlogging. Estimates of cotton yield (used in place of cotton simulations) for Vertosols assume that this waterlogging could be managed if fields were carefully sited and furrows were skilfully managed. However, as illustrated before, some Vertosols in the Roper catchment present particularly severe drainage challenges (Figure 4-6) that could limit the suitable area for farming, and may require more careful management than Vertosols that are currently used for cotton farming in other parts of Australia. A breakdown of the variable costs for growing broadacre crops showed that the largest two costs are the costs of inputs (31%) and farm operations (35%). Both of these cost categories would have only moderately higher dollar values when growing the same crop in southern parts of Australia, but the cost category that puts northern growers at greatest disadvantage is the higher market costs (23%: freight and other costs involved in selling the crop). Total variable costs consume 58% of the gross revenue generated, which leaves sufficient margin for profitable farms to be able to temporarily absorb small declines in commodity prices or yields without creating severe cashflow problems. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 4-9 A melon crop growing in the Mataranka area of the Sturt Plateau Photo: CSIRO - Nathan Dyer Table 4-6 Performance metrics for broadacre cropping options in the Roper catchment: applied irrigation water, crop yield and gross margin (GM) for three environments Performance metrics are an indication of the upper bound that could be achieved after best management practices for Roper catchment environments had been identified and implemented. All options are for dry season (DS) irrigated crops sown between mid-March and the end of April (end of the wet season), except for the wet season (WS) cotton, sown in early February. Variance in yield estimates from APSIM simulations is indicated by providing 80%, 50% (median) and 20% probability of exceedance values (Y80%, Y50% and Y20%, respectively), together with associated applied irrigation water (including on-farm losses) and GMs in those years. The lower-range yields (Y80% exceedance) occur in most years, while the upper-range Y20% yields only occur in the most exceptional upper 20% of years. Note that applied irrigation water is not always higher in years with higher yields (Y20%). ‘na’ indicates 20% and 80% exceedance estimates that were not applicable because APSIM outputs were not available and expert estimates of just the median yield and water use were used instead. Peanut is omitted for the Vertosol location because of the practical constraints of harvesting root crops on clay soils. Freight costs assume processing near Katherine for cotton and peanut, and that hay is sold locally. No crop model was available for sesame or hemp, so indicative estimates for the catchment were used. Cotton yields and prices are for lint bales (227 kg after ginning), not tonnes (t). PAWC = plant available water capacity. CROP APPLIED IRRIGATION WATER CROP YIELD YIELD UNIT PRICE VARIABLE COSTS TOTAL REVENUE GROSS MARGIN (ML/ha/y) (Yield units) ($/unit) ($/ha/y) ($/ha/y) ($/ha/y) Y80% Y50% Y20% Y80% Y50% Y20% Y80% Y50% Y20% Dermosol (156 mm PAWC), Bulman climate (~1000 mm annual rainfall) Cotton WS 6.9 6.1 3.8 10.4 11.2 12.0 bales/ha 580 3604 7448 3439 3844 4366 Cotton DS 8.0 7.2 8.5 8.4 9.1 9.8 bales/ha 580 3291 6073 2400 2782 3110 Sorghum (grain) 4.0 5.6 4.9 7.5 7.9 8.2 t/ha 310 1734 2449 674 715 801 Mungbean 3.2 3.2 3.2 1.7 1.9 2.0 t/ha 1100 940 1919 797 979 1068 Chickpea 4.8 4.3 5.4 2.5 2.7 3.0 t/ha 750 1119 2052 772 933 1053 Soybean 7.6 6.6 7.3 3.3 3.6 3.8 t/ha 570 1342 2052 540 710 784 Peanut 4.5 5.3 5.5 4.5 4.8 5.1 t/ha 1000 3126 4800 1508 1674 1888 Rhodes grass (hay) 13.3 10.8 12.4 34.2 35.1 36.1 t/ha 42 4189 8775 4266 4586 4671 Kandosol (79 mm PAWC), Mataranka climate (~950 mm annual rainfall) Cotton WS 3.6 5.8 7.1 6.9 10.9 12.5 bales/ha 580 3544 7283 1760 3738 4510 Cotton DS 8.9 8.7 10.5 8.3 9.1 10.0 bales/ha 580 3530 6073 2091 2543 2890 Sorghum (grain) 6.8 6.5 5.9 6.1 6.4 6.8 t/ha 310 1730 1984 142 254 416 Mungbean 4.1 4.6 5.0 1.6 1.7 2.0 t/ha 1100 1156 1717 507 561 809 CROP APPLIED IRRIGATION WATER CROP YIELD YIELD UNIT PRICE VARIABLE COSTS TOTAL REVENUE GROSS MARGIN (ML/ha/y) (Yield units) ($/unit) ($/ha/y) ($/ha/y) ($/ha/y) Y80% Y50% Y20% Y80% Y50% Y20% Y80% Y50% Y20% Chickpea 6.2 5.3 4.5 2.3 2.4 2.6 t/ha 750 1340 1796 302 455 692 Soybean 7.1 7.4 7.3 2.3 2.5 2.6 t/ha 570 1637 1425 –283 –212 –147 Peanut 4.6 6.1 6.5 3.6 3.8 3.9 t/ha 1000 2850 3800 936 950 1001 Rhodes grass (hay) 19.3 19.9 18.6 32.0 32.9 33.4 t/ha 42 4694 8225 3407 3531 3709 Vertosol (212 mm PAWC), Ngukurr climate (~850 mm) Cotton WS na 6.0 na na 11.0 na bales/ha 580 3807 7341 na 3535 na Cotton DS na 8.0 na na 9.5 na bales/ha 580 3599 6340 na 2741 na Sorghum (grain) 4.6 5.7 5.7 7.6 8.1 8.3 t/ha 310 1714 2511 715 797 839 Mungbean 2.6 5.2 3.9 2.1 2.3 2.4 t/ha 1100 1000 2323 1192 1323 1438 Chickpea 5.4 5.4 5.4 2.7 3.0 3.2 t/ha 750 1137 2223 939 1086 1234 Soybean 9.0 8.0 9.1 3.8 4.1 4.4 t/ha 570 1372 2337 803 965 1083 Rhodes grass (hay) 13.2 12.1 15.1 35.6 36.8 38.5 t/ha 42 4377 9200 4538 4823 4947 General estimate for Roper catchment (not soil specific) Sesame na 6.2 na na 0.9 na t/ha 1300 1737 1170 na –567 na Hemp (grain seed) na 5.9 na na 1.1 na t/ha 3150 2149 3465 na 1316 na Narrative risk analyses were conducted for the two broadacre crops with the highest GMs: cotton and forages. The cotton analysis explored the sensitivity of GMs to opportunities and challenges created by changes in cotton lint prices, crop yields and distance to the nearest gin (Table 4-7). Results show that high recent cotton prices (about $750/bale) have created a unique opportunity for those looking to establish new cotton farms in NT locations like the Roper catchment, since growers could transport cotton to distant gins or produce suboptimal yields and still generate GMs above $3000/ha. At lower cotton lint prices, a local gin becomes more important for farms to remain viable. Recent high cotton prices have reduced some of the risk involved in learning to grow cotton to its full sustainable potential in the region and while awaiting the commissioning of the new gin 30 km north of Katherine (due in 2023). At high yields and prices, the returns per ML of irrigation water may favour growing a single cotton crop per year, instead of committing limited water supplies to sequential cropping with a dry-season crop (that would likely provide lower returns per ML and be operationally difficult/risky to sequence). Table 4-7 Sensitivity of cotton crop gross margins (GMs) to variation in yield, lint prices and distance to gin The base case is the Ngukurr Vertosol (Table 4-7) and is highlighted for comparison. The gin locations considered are a local gin near a new cotton farming region in the Roper catchment, the new gin in Katherine, and two other potential gins in the NT (Adelaide River) and northwest Queensland (Richmond). Cotton lint prices are for the average over the past decade ($580/bale), recent high prices ($750/bale), and lower prices from about 10 years ago ($450/bale). Effects of a lower yield are also tested (the 9.5 bales/ha estimated as the dry-season yield for this location versus the base case of 11 bales/ha for wet-season cropping). FREIGHT COST $/t (DISTANCE TO GIN) LINT PRICE = $450/bale LINT PRICE = $580/bale LINT PRICE = $750/bale YIELD YIELD YIELD 9.5 bales/ha 11 bales/ha 9.5 bales/ha 11 bales/ha 9.5 bales/ha 11 bales/ha $13 (50 km to local gin) 1881 2517 3116 3947 4731 5817 $79 (300 km to Katherine gin) 1526 2105 2761 3535 4376 5405 $113 (500 km to Adelaide River gin) 1342 1892 2577 3322 4192 5192 $317 (1700 km to Richmond gin) 242 619 1477 2049 3092 3919 The narrative risk analysis for irrigated forages also looked at the sensitivity of farm GMs to variations in hay price and distance to markets, but here focuses on the issues of local supply and demand (Table 4-8). Forages, such as Rhodes grass, are a forgiving first crop to grow on greenfield farms as new farmers gain experience of local cropping conditions and ameliorate virgin soils. While there are limited supplies of hay in the region, growers may be able to sell hay at a reasonable price, given the large amount of beef production in the region and challenges of maintaining livestock condition through the dry season when the quality of native pastures is low. This would particularly be the case in dry years when the quantity and quality of native pasture is low and demand for livestock dietary supplements increases. But the scale of unmet local demand for hay limits opportunities to scale hay production without depressing local prices and/or having to sell hay further away, both of which lead to rapid declines in GMs (to below zero in many cases, Table 4-8). Another opportunity for hay is for feeding to cattle during live export which could be integrated into an existing beef enterprise to supply their own live export livestock: this would require the hay to be pelleted. Section 4.3.9 considers how forages could be integrated into local beef productions systems for direct consumption by livestock within the same enterprise. Table 4-8 Sensitivity of forage (Rhodes grass) crop gross margins (GMs) to variation in yield and hay price The base case is the Ngukurr Vertosol (Table 4-7) and is highlighted for comparison. Transporting the hay further distances would increase opportunities for finding counter-seasonal markets paying higher prices, but this would be rapidly offset by higher freight costs. FREIGHT COST (DISTANCE TO DELIVER) HAY PRICE $150/t $250/t $350/t $20 (Local) 1142 4823 8502 $79 (300 km to Katherine) –1028 2651 6331 $317 (1700 km to Richmond) –9787 –6107 –2427 4.3.6 Irrigated horticultural crops Table 4-9 shows estimates of potential performance for a range of horticultural crop options in the Roper catchment. Upper potential GMs for annual horticulture (about $9,000 per ha per year) were less than upper potential GMs for farming perennial fruit trees (about $11,000 per ha per year). Capital costs of farm establishment and operating costs increase as the intensify of farming increases, so ultimate farm financial viability is not necessarily better for horticulture compared to broadacre crops with lower GMs (see Chapter 6). Note also that perennial horticulture crops typically requires more water than annual crops because irrigation occurs for a longer period each year (mean of 9.0 versus 4.8 ML per ha per year, respectively in Table 4-9): this also, indirectly, affects capital costs of development since perennial crops require a larger investment in water infrastructure compared to annual crops to support the same cropped area. Table 4-9 Performance metrics for horticulture options in the Roper catchment: annual applied irrigation water, crop yield and gross margin (GM) Applied irrigation water includes losses of water during application. Horticulture is most likely to occur on well-drained Kandosols. KP = Kensington Pride mangoes and PVR = new high-yielding mangoes varieties with plant variety rights (e.g. Calypso). Product unit prices listed are for the dominant top grade of produce, but total yield was apportioned among lower graded/priced categories of produce too in calculating total income. Transport costs assume sales of total produce are a split among southern capital size markets in proportion to their size. Applied irrigation water accounts for application losses assuming efficient pressurised micro irrigation systems. CROP APPLIED IRRIGATION WATER CROP YIELD PRICE PRICING UNIT VARIABLE COSTS TOTAL REVENUE GROSS MARGIN (ML/ha/y) (t/ha/y) ($/unit) (unit) ($/ha/y) ($/ha/y) ($/ha/y) Row crop fruit and vegetables, annual horticulture (less capital intensive) Rockmelon 5.3 25.0 28 15 kg tray 40,819 44,000 3,181 Watermelon 6.0 47.0 450 500 kg box 42,756 42,300 –456 Capsicum 3.2 32.0 19 8 kg carton 66,757 76,000 9,243 Onion 4.7 30.0 15 10 kg bag 35,661 41,850 6,189 Fruit trees, perennial horticulture (more capital intensive) Mango (KP) 7.8 9.3 24 7 kg tray 20,751 28,398 7,648 Mango (PVR) 7.8 17.5 21 7 kg tray 40,386 47,250 6,864 Lime 11.4 28.5 18 5 kg carton 89,451 100,890 11,439 Crop yields and GMs can vary substantially amongst varieties, as is demonstrated here for mangoes (Mangifera indica). Mango production is well-established in multiple regions of northern Australia, including in the Darwin, Douglas–Daly and Katherine regions of the NT, with a smaller area of orchards at Mataranka in the Roper catchment. For example, the well-established Kensington Pride (KP) mangoes typically produce 5 to 10 t/ha while newer varieties can produce 15 to 20 t/ha. These new varieties (such as Calypso) are likely to be released with plant variety rights (PVR) accreditation. Selection of varieties also needs to consider consumer preferences and timing of harvest relative to seasonal gaps in market supply that can offer premium prices. Prices paid for fresh fruit and vegetables can be extremely volatile (Figure 4-10) because produce is perishable and expensive to store, and regional weather patterns can disrupt target timing of supply that can result in unintended overlaps or gaps in combined supply between regions. This creates regular fluctuations between oversupply and undersupply, against inelastic consumer demand, to the extent that prices can fall so low at times that it would cost more to pick, pack and transport produce than farms receive in payment. Amongst this volatility are some counter- seasonal windows in southern markets (where prices are typically higher) that northern Australian growers can target. Figure 4-10 Fluctuations in seedless watermelon prices at Melbourne wholesale markets from April 2020 to February 2023 Source: ABARES (2023) Horticultural enterprises typically run on very narrow margins, where about 90% of gross revenue would be required just to cover variable costs of growing and marketing a crop grown in the Roper catchment. This makes crop GMs extremely sensitive to fluctuations in variable costs, yield and produce prices, amplifying the effect of already volatile prices for fresh fruit and vegetables. The majority of the variable costs of horticultural production occur from harvest onwards, mainly in freight. This affords the opportunity to mitigate losses if market conditions are unfavourable at the time of harvest, since most costs can be avoided (at the expense of forgone revenue) by not picking the crop. For more information on this figure please contact CSIRO on enquiries@csiro.au A narrative risk analysis for horticulture used the crop with the lowest GM (watermelons (Citrullus lanatus): Table 4-7) to illustrate how opportunities for reducing freight costs and targeting periods of higher produce prices could improve GMs to find niches for profitable farms (Table 4-10). Reducing freight costs by finding backloading opportunities or concentrating on just the smaller closest southern capital city market of Adelaide would substantially improve GMs (to $6547 and $4039 per ha per year, respectively). The base case already assumed that growers in the Roper catchment would target the predictable seasonal component of watermelon price fluctuations (Figure 4-10), but any further opportunity to attain premiums in pricing could help convert an unprofitable baseline case into a profitable one. This example also highlights the issue that while there may be niche opportunities that allow an otherwise unprofitable enterprise to be viable, the scale of those niche opportunities also then limits the scale to which the industry in that location could expand, for example, there is a limit to the volume of backloading capacity at cheaper rates; only supplying produce to the closest market excludes the largest markets (e.g. accessing the larger Sydney and Melbourne markets remains nonviable except when prices are high, Table 4-10); and chasing price premiums restricts the seasonal windows into which produce is sold or restricts markets to smaller niches that target specialised product specifications. Niche opportunities are seldom scalable, particularly in horticulture, which is a contributing factor to why horticulture in any region usually involves a range of different crops (often on the same farm). Table 4-10 Sensitivity of watermelon crop GMs to variation in melon prices and freight costs The base case (Table 4-9) is highlighted for comparison. FREIGHT COST MELON PRICE (PERCENTAGE DIFFERENCE FROM BASE PRICE) (MARKET LOCATION) $225 (–50%) $337 (–25%) $450 (BASE PRICE) $675 (+50%) $900 (+100%) $210/T $210/t (backloading to Adelaide) –11,642 –2,588 6,547 24,736 42,925 $263/t (close market: Adelaide) –14,150 –5,056 4,039 22,228 40,417 $359/t (all capital cities) –18,662 –9,568 –456 17,716 35,905 $387/t (Sydney) –19,978 –10,884 –1,789 16,400 34,589 $391/t (Melbourne) –20,166 –11,072 –1,977 16,212 34,401 The risk analysis also illustrates just how much farm financial metrics like GMs amplify fluctuations to input costs and commodity prices to which they are exposed. For horticulture, far more than broadacre agriculture, it is very misleading to look just at a single ‘median’ GM for the crop, because that is a poor reflection of what is going on within an enterprise. For example, the –50% to +100% variation in watermelon prices shown in Figure 4-10 would result in theoretical annual GMs fluctuating between –$18,662/ha and $35,905/ha (Table 4-10). While, in practice, potentially negative GMs could be greatly mitigated (by not harvesting the crop), this still creates cashflow challenges in managing years of negative returns between years of windfall profits. This amplified volatility is another contributor to horticulture farms often growing a mix of produce (as a means of spreading risk). For row crop production, like melons, another common way of mitigating risk is using staggered planting through the season, so that subsequent harvesting and marketing are spread out over a longer target window to smooth out some of the price volatility. 4.3.7 Plantation tree crops Estimates of annual performance for African mahogany (Khaya ivorensis) and sandalwood (Santalum album) are provided in Table 4-11. The best available estimates were used in the analyses, but information on plantation tree production in northern Australia is often commercially sensitive and/or not independently verified. The measures of performance presented therefore have a low degree of confidence and should be treated as broadly indicative noting that actual commercial performance could be either lower or higher. Table 4-11 Performance metrics for plantation tree crop options in the Roper catchment: annual applied irrigation water, crop yield and gross margin (GM) Yields are values at final harvest and for sandalwood are just for the heartwood component. Other values are annual averages assuming a 20-year life cycle of the crop (representing the idealised ultimate steady state of an operating farm that was set up with staggered plantings for a steady stream of harvests). No discounting is applied to account for the substantial timing offset between when costs are incurred and income is received: any investment decision would need to take that into account. African mahogany performance is for unirrigated production. CROP CROP LIFE CYCLE APPLIED IRRIGATION WATER CROP YIELD AT HARVEST PRICE PRICING UNIT VARIABLE COSTS TOTAL REVENUE GROSS MARGIN (y) (ML/ha/y) (t/ha) ($/unit) ($/ha/y) ($/ha/y) ($/ha/y) African mahogany 20 unirrigated 160 4,000 t 682 4,000 3,318 Sandalwood 20 4.7 4 8,800 t heartwood 901 1,760 859 Plantation forestry has long life cycles with low-intensity management during most of the growth cycle, so variable costs typically consume less of the gross revenue (27%) than broadacre or horticultural farming. However, long life cycle production systems have additional risks over annual cropping in that there is a much longer period between planting and harvest for adverse events to affect the yield quantity and/or quality, prices of inputs and harvested products could change substantially over that period, and market access and arrangements with buyers could also change. The long lags from planting to harvest also mean that potential investors need to consider other similar competing pipeline developments (that may not be obvious because they are not yet selling product) and long-term future projections of supply and demand (for when their own plantation will start to be harvested and enter supply chains). The cashflow challenges are also significant, given the long-term outlay of capital and operating costs before any revenue is generated. Carbon credits might be able to assist with some early cash flow (if the ‘average’ state of the plantation, from planting to harvest, stores more carbon than the vegetation it replaced). 4.3.8 Cropping systems This section evaluates the types of cropping systems (crop species x growing season x resource availability x management options) that are most likely to be profitable in the Roper catchment based on the analyses of farm performance above, information from companion technical reports in this Assessment, and cropping knowledge from climatically analogous regions (relative to local biophysical conditions). Cropping system choices could include growing a single crop during a 12- month period, or growing more than one, commonly referred to as sequential, double or rotational cropping. Since many of the issues for single cropping options were already covered above, this section focuses more on sequential cropping systems and the mix of cropping options that might make up a new farming area in the Roper catchment. Cropping system considerations In addition to the challenges of choosing an individual crop to farm in the Roper catchment, selecting two or more crops to grow in sequence brings additional complexity. The rewards from successfully growing crops in sequence (versus single cropping) can be substantial if additional net annual revenue can be generated from the same initial capital investment (to establish the farm). Markets Whether growing a single crop or sequential cropping, the choice of crop(s) to grow is market driven. As the price received for different crops fluctuates, so too will the crops grown. In the Roper catchment freight costs, determined by the distance to selected markets, will also need to be considered. A critical scale of production may be needed for a new market opportunity or supply chain to be viable, (e.g. exporting grains from Darwin would require sufficient economies of scale for the required supporting port infrastructure and shipping routes to be viable). Crops such as cotton, peanut and sugarcane (Saccharum officinarum) require a nearby processing facility. A consistent and critical scale of production is required for processing facilities to be viable. From 2023 cotton will have the advantage of local processing when a gin will be operational 30 km north of Katherine. Transport of raw cotton from the Roper catchment to this gin would go a long way to improving the viability of cotton production (Table 4-7). Most horticultural production from the Roper catchment would be sent to capital city markets, often using refrigerated transport. Roper catchment horticultural production would have to accept a high freight cost relative to producers in southern parts of Australia. The competitive advantage of horticultural production in the Roper catchment is that higher market prices can be achieved from ‘out of season’ production compared to large horticultural production areas in southern Australia. Annual horticultural row crops, such as melons, would be grown sequentially, for example, with fortnightly planting over a 3-to-4-month period, to reduce risk of exposure to low market prices and to make it more likely that very high market prices would be achieved for at least some of the produce. Operations Sequential cropping can require a trade-off in sowing times to allow crops to be grown within a back-to-back schedule. This trade-off could lead to slightly lower yields from planting at suboptimal times. For annual horticulture crops there would be an additional limitation on the seasonal window over which produce can be sent to market (reducing opportunities to target peak prices and/or mitigate risks from price fluctuations). Growing crops sequentially depends on timely transitions between the crops and selecting crops with growing seasons that will reliably fit into the available cropping windows. In the Roper catchment’s variable and often intense wet season, rainfall increases operational risk via reduced trafficability and the subsequent limited ability to conduct timely operations. A large machinery investment (either multiple or larger machines) could increase the area that could be planted per day when fields are trafficable within a planting window. With sequential cropping, additional farm machinery and equipment may be required where there are crop-specific machinery requirements, or to help complete operations on time where there is tight scheduling between crops. Any additional capital expenditure on farm equipment would need to be balanced against the extra net farm revenue generated. Sequential cropping can also lead to a range of cumulative issues that need careful management, for example, build-up of pests, diseases and weeds; pesticide resistance, often exacerbated by sequential cropping; increased watertable depth; and soil chemical and structural decline. Many of these challenges can be anticipated prior to commencement of sequential cropping. Integrated pest, weed and disease management would be essential when multiple crop species are grown in close proximity (adjacent fields or farms). Many of these pests and controls are common to several crop species where pests move between fields (e.g. aphids). Such situations are exacerbated when the growing seasons of nearby crops partially overlap or when sequential crops are grown, because both scenarios create ‘green bridges’ facilitating the continuation of pest life cycles. When herbicides are required, it is critical to avoid products that could damage a susceptible crop the following season or sequentially. Water Sequential cropping leads to a higher annual crop water demand because: the combined period of cropping is longer (versus single cropping); it includes growing during the Roper catchment dry season; and PAW at planting will have been depleted by the previous crop. Typically, an additional 1 ML/ha on well-drained soils, and 1.5 ML/ha on clays, is required for sequential cropping relative to the combined water requirements of growing each of those crops individually (with the same sowing times). This additional water demand needs consideration during development where on- farm water storage is required, or dry-season water extraction is necessary. Irrigating using surface water in the Roper catchment would face issues with the reliability and the timing of water supplies. River flows are unlikely to be sufficient to trigger pumping into on-farm storages for irrigation (i.e. to meet environmental flow and river height requirements) before mid to late wet season (mid-February to March) in the mid Roper catchment (see companion technical report on river modelling by Hughes et al. (2023)). The timing of water availability is therefore not well suited to crops that would need to be reliably sown by March (e.g. wet-season grain sorghum, soybean and sesame) and would push cotton planting to the later part of the wet-season window (Figure 4-5). Late availability of water for extraction each wet season reduces the options for sequencing a second crop. Soils The largest arable areas in the Roper catchment are loamy Kandosols of the Sturt Plateau (SGGs 4.1 and 4.2, marked A in Figure 4-3) and the cracking clay Vertosols on the alluvial plains of the major rivers (SGGs 2 and 9, marked B in Figure 4-3). There are good analogues of these Roper catchment environments in successful irrigated farming areas in other parts of northern Australia: Katherine is indicative of farming systems and potential crops grown on well-drained loamy soils irrigated by pressurised systems, and the Ord River Irrigation Area is indicative of furrow irrigation on heavy clay soils. The good wet-season trafficability of the well-drained loamy Kandosols permits timely cropping operations and would enhance the implementation of sequential cropping systems. However, Kandosols also present some constraints for farming. Kandosols are inherently low in organic carbon, nitrogen (N), phosphorus (P), sulfur (S), zinc (Zn) and potassium (K) with other micronutrients often requiring supplementation (molybdenum (Mo), boron (B), and copper (Cu)). Very high fertiliser inputs are therefore required when first cultivated. Due to the high risk of leaching of soluble nutrients (e.g. N and S) during the wet season, in-crop application (multiple times) of the majority of crop requirement for these nutrients is necessary. In addition, high soil temperatures and surface crusting combined with rapid drying of the soil at seed depth reduce crop establishment and seedling vigour for many broadacre species sown during the wet season and early dry season (e.g. maize, soybean, cotton). In contrast, the cracking clay Vertosols have poor trafficability following rainfall (Figure 4-6) or irrigation, disrupting cropping operations. Farm design is a major factor on cracking clay soils to minimise flooding of fields from nearby waterways, ensure prompt runoff from fields after irrigation or rain events, and maintain trafficability of farm roads. Timely in-field bed preparation can reduce delays in planting. Clay soils also have some advantages, particularly in costs of farm development by allowing lower cost surface irrigation (versus pressurised systems) and on-farm storages (where expensive dam lining can be avoided if soils contain sufficient clay). Clay soils also typically have greater inherent fertility than Kandosols (but initial sorption by clay means that phosphorus requirements can be high for virgin soils in the first 2 years of farming). Potentially suitable cropping systems Potential crop species that could be grown as a single crop per year were identified and rated for the Roper catchment (Table 4-12) based on indicators of farm performance presented above (yields, water use and GMs), together with considerations of growing season, experiences at climate-analogous locations, past research, and known market and resource limitations and opportunities. Annual horticulture, cotton, peanut and forages are the most likely to generate returns that could exceed farm development and growing costs (Table 4-12). Table 4-12 Likely annual irrigated crop planting windows, suitability, and viability in the Roper catchment Crops are rated as to how likely they are to be financially viable: *** = likely at low-enough development costs; ** = less likely for single cropping (at current produce prices); * S = marginal but possible in a sequential cropping system. Rating qualifiers are codes as L development limitation, M market constraint, P depends on sufficient scale and distance to local processor, and B depends on distance to and type of beef (livestock production) activity it is supporting. Farm viability is dependent on the cost at which land and water can be developed and supplied (Chapter 6). na = not applicable. CROP RATING CROP RATING Wet season (planted December to early March) Dry season (planted late March to August) Cotton *** P Annual horticulture *** M Forages *** B Cotton *** P Sugarcane *** LP Niche grains (e.g. chia, quinoa) *** SM Peanut (not on clay) *** LMP na na Mungbean ** Mungbean ** Maize ** na na Chickpea ** na na Rice ** L na na Sorghum * S Sorghum * S Soybean * S Soybean * S Sesame * S Sesame * S Due to good wet-season trafficability on loamy soils, there are many possible sequential cropping options for the Roper catchment Kandosols (Table 4-13). Due to the predominance of broadleaf and legume species in many of the sequences (Table 4-13), a grass species is desirable as an early wet-season cover crop. Although annual horticulture and cotton could individually be profitable (Table 4-12), an annual sequence of the two would be very tight operationally. Cotton would be best grown from late January with the need to pick the crop by early August, then destroy cotton stubble, prepare land and remove volunteer cottons seedlings. That scheduling would make it challenging to fit in a late-season melon crop that would need to be sown by late August to early September. Similar challenges would occur with cotton followed by mungbean or grain sorghum. Table 4-13 Sequential cropping options for Kandosols E = early in month; L = late in month; M = middle of month. SPECIES GROWING SEASON SPECIES GROWING SEASON Wet season (planted December to early March) Dry season (planted late March to August) Mungbean E-February to L-April Annual horticulture From M-May to L-October Grain sorghum January to April Peanut (not on clay) January to April or February to May Cotton L-January to E-August Mungbean M-August to L-October Grain sorghum M-August to M-November Forage/silage to E-November; cut then retained as wet-season cover crop Mungbean E-February to L-April Cotton E-May to E-November Mungbean Peanut Sesame Soybean E-February to L-April E-January to L-April E-January to L-April E-January to L-April Maize May to October Sesame or Grain sorghum (grain) January to L-April Chickpea May to August Mungbean Sesame Soybean E-February to L-April January to L-April January to L-April Grass forage/silage May to E-November; cut then retained as wet-season cover crop Fully irrigated sequential cropping on the Roper catchment Vertosols would likely be opportunistic and favour combinations of short-duration crops that can be grown when irrigation water reliability is greatest (March to October), for example, annual horticulture (melons), mungbean, chickpea, and grass forages (2 to 4 months growing season length). Following a rain-grown wet- season grain crop with a dry-season irrigated crop is also possible. However, seasonally dependent soil wetting and drying would limit timely planting and the area planted, which means that farm yields between years would be very variable. Grain sorghum, mungbean and sesame are the species most adapted to dryland cropping due to favourable growing season length, and their tolerance to water stress and higher soil and air temperatures. 4.3.9 Integrating forage and hay crops into existing beef cattle enterprises A commonly held view within the northern cattle industry is that the development of water resources would allow irrigated forages and hay to be integrated into existing beef cattle enterprises, thereby improving their production and potentially, their profitability. Currently, cattle graze on native pastures, which rely solely on rainfall and any consequent overland flow. The quality of these pastures is typically low, and it declines throughout the dry season, so that cattle either gain little weight, or even lose weight, during this period. Theoretically, the use of on-farm irrigated forage and hay production would allow graziers greater options for marketing cattle: meeting market live weight specifications for cattle at a younger age; meeting the specifications required for different markets than those typically targeted by cattle enterprises in the Roper catchment; and providing cattle which meet market specification at a different time of the year. Forages and hay may also allow graziers to implement management strategies, such as early weaning or weaner feeding, which should lead to flow-on benefits throughout the herd. Some of these strategies are already practiced within the Roper catchment but are reliant on hay or other supplements purchased on the open market. By growing hay on- farm, the scale of these management interventions might be increased, at reduced net cost. Furthermore, the addition of irrigated feeds may also allow graziers to increase the total number of cattle which can be sustainably carried on the property. The use of irrigated hay or forage production to feed cattle on-farm in the Roper catchment is very little used, if at all (Cowley, 2014). In fact, there are very few cattle enterprises in northern Australia which are set up to integrate on-farm irrigation, notwithstanding the theoretical benefits. Despite its apparent simplicity, fundamentally altering an existing cattle enterprise in this way brings in considerable complexity, with a range of unknowns about how best to increase productivity and profitability. There is still much to be learned about the most appropriate forage and hay species to grow, how best to manage the forages and hay to ensure high-quality feed, which cohort(s) of cattle to feed, how the feeding should be managed and which market specifications should be targeted to obtain maximum return. Because there are so few on-ground examples, modelling has been used in a number of studies to consider the integration of forages and hay into cattle enterprises (Watson et al., 2021). The most comprehensive guide to what might be possible to achieve by integrating forages into cattle enterprises can be found in Moore et al. (2021), who used a combination of industry knowledge, new research and modelling to consider the costs, returns and benefits. Bio-economic modelling was used in the Assessment to consider the impact of growing irrigated forages and hay on a representative beef cattle enterprise on the red earths of the Sturt Plateau, using Larrimah as the rainfall record (see the companion technical report on agricultural viability and socio-economics (Stokes et al., 2023) for more detail). The enterprise was based on a self- replacing cow-calf operation, focused on selling into the live export market. Broadly speaking, these enterprise characteristics can be thought of as a typical cattle enterprise within the Roper catchment with a size of about 100,000 ha and an Owner-Manager. The modelling considered a number of scenarios: (i) a baseline; (ii) baseline plus buying-in hay to feed weaners; growing forage sorghum, an annual forage grass species, and feeding either as (iii) stand and graze or (iv) as hay; (v) growing lablab (Lablab purpureus), an annual legume, and feeding as stand and graze; and (vi) growing Rhodes grass, a perennial tropical grass, and fed as hay. Ideally, production would increase by allowing male animals to reach minimum selling weight at a younger age and allowing for greater weight gain during the dry season when animals on native pasture alone either lose weight or gain very little weight. The addition of forages and hay also allows more cattle to be carried, while still maintaining a utilisation rate of native pastures at around 15%. A GM per adult equivalent (AE) was calculated as the total revenue from cattle sales minus total variable costs (Table 4-14). A profit metric, earnings before interest, taxes, depreciation and amortisation (EBITDA), was also calculated as income minus variable and overhead costs, which allows performance to be compared independently of financing and ownership structure (McLean and Holmes, 2015) and is used in the analysis of net present value (NPV). Three sets of beef prices were considered: • LOW beef price. Beef prices were set to 275 c/kg for males between 12 months and 24 months old, declining across age and sex classes to 134 c/kg for cows older than 108 months. • MED beef price. Beef prices were set to 350 c/kg for males between 12 months and 24 months old, declining across age and sex classes to 170 c/kg for cows older than 108 months. • HIGH beef price. Beef prices were set to 425 c/kg for males between 12 months and 24 months old, declining across age and sex classes to 206 c/kg for cows older than 108 months. At all three beef prices, total income was highest for the four irrigated forage or hay scenarios compared to the two baseline scenarios but the higher costs for the irrigated scenarios led to similar or lower GMs. Table 4-14 Production and financial outcomes from the different irrigated forage and beef production scenarios for a representative property on the Sturt Plateau Details for LOW, MED and HIGH beef prices are found in the text in Section 4.3.9. Scenario descriptions are found in the companion technical report on agricultural viability and socio-economics (Stokes et al., 2023: Section 5.4). AE = adult equivalent; EBITDA = earnings before interest, taxes, depreciation and amortisation. BASELINE BASELINE PLUS HAY FORAGE SORGHUM – STAND AND GRAZE FORAGE SORGHUM – HAY LABLAB – STAND AND GRAZE RHODES GRASS – HAY Forage/hay None Bought hay Forage sorghum Forage sorghum Lablab Rhodes grass Maximum number of breeders 2030 2070 2400 2300 2250 2290 Herd size (AE) averaged across calendar year 2752 2760 3316 3215 3167 3170 Pasture utilisation (%) 15.1 15.0 15.2 15.0 15.1 15.0 Weaning rate (%) 64 63 63 64 64 63 Mortality rate (%) 6.9 6.9 6.4 6.4 6.5 6.5 Average weight of all castrate males sold in May (kg/animal) 343 331 355 356 356 357 BASELINE BASELINE PLUS HAY FORAGE SORGHUM – STAND AND GRAZE FORAGE SORGHUM – HAY LABLAB – STAND AND GRAZE RHODES GRASS – HAY Average weight of 18 month old (i.e. end- November-born) castrate males sold in May (kg/animal) 307 311 354 352 350 352 Average weight of 30 month old (i.e. end- November-born) castrate males sold in May (kg/animal) 378 387 n/a n/a n/a n/a Average age of castrate males sold in May (months) 24 20 18 18 18 18 Percentage of castrate male cohort aged 15 months to 19 months (compared with 27 to 31 month cohort) sold in May (%) 51 87 100 100 100 100 Beef produced per year (kg) 380,119 390,161 478,419 465,534 455,166 460,597 Gross margin ($/AE) (LOW BEEF PRICE) 142 133 95 100 113 136 Profit (EBITDA) ($) (LOW BEEF PRICE) 128,073 103,770 52,157 58,223 93,104 166,263 Gross margin ($/AE) (MED BEEF PRICE) 226 220 181 188 200 224 Profit (EBITDA) ($) (MED BEEF PRICE) 359,466 342,142 337,246 339,901 369,890 445,448 Gross margin ($/AE) (HIGH BEEF PRICE) 310 306 267 275 288 312 Profit (EBITDA) ($) (HIGH BEEF PRICE) 590,860 580,513 622,335 621,580 646,676 724,633 At MED beef prices, EBITDA was highest for Rhodes grass hay ($445,448/year). The EBITDA for all other scenarios was between $337,246/year (forage sorghum stand and graze) and $369,890/year (lablab stand and graze). While production (measured as beef sold per financial year) is clearly boosted by the introduction of irrigated forages or hay, the profitability is highly sensitive to the cost of the irrigated scenarios. NPV analysis showed that only one scenario had a positive NPV, that of Rhodes grass hay at HIGH beef price and the lower of two development costs per ha ($15,000/ha as opposed to $25,000/ha). All other scenarios gave a negative NPV and even the one positive NPV was low ($312,793), suggesting that a decision to irrigate would need to assume beef prices well above their 10-year average in order to be viable. The EBITDAs would need to increase by about $2,000 per year per irrigated ha at the $15,000/ha development cost in order to meet the costs of development or about $3,000 per year per irrigated ha at the $25,000/ha development cost. Much of the animal production and EBITDA increases due to the irrigated forage scenarios came from the increased number of breeders which could be carried, while still keeping the utilisation rate of native pastures at about 15%. The two irrigated hay scenarios allowed the highest number of breeders to be carried, an average of 2295, compared with 2030 and 2070 for the two baseline scenarios. This flowed through to the total number of AE carried being about 15% to 20% higher than the two baseline scenarios averaged across all years. The amount of beef produced each year was about 20% to 24% higher, using the same scenario comparison. The average sale weight and average age of all castrate males sold at the May sales requires some explanation and is due to the age at which cattle are sold (Table 4-14). The average weight for cattle in the baseline and baseline plus hay scenarios (343 kg and 331 kg) is similar to those in the four forage or hay scenarios (between 355 kg and 357 kg). The reason for this is that only 51% (baseline) and 87% (baseline plus hay) of animals were sold in their second May (15 to 19 months old) with the remainder sold in their third May (27 to 31 months old). By contrast, 100% of the animals in all four forage and hay scenarios were sold in their second May (15 to 19 months old) and their average weight at the May sale reflects this. For 30 month old steers in the two baseline scenarios, the average sale weights were 378 kg and 387 kg. While there are advantages to some form of irrigated forage or hay production, the introduction of irrigation to an existing cattle enterprise is not for the faint-hearted. The scenarios here range from an area which would require 1.5 pivots of 40 ha each to an area which would require 5 pivots of 40 ha each. A water allocation of about 0.8 to 1.2 GL would be required to provide sufficient irrigation water. The capital cost of development would range between $900,000 for 60 ha of Rhodes grass hay at a development cost of $15,000/ha to $5,000,000 for 200 ha of lablab at a development cost of $25,000/ha. In addition, the grazing enterprise would need to develop the expertise and knowledge required to run a successful irrigation enterprise of that scale, which is quite a different enterprise to one of grazing only. This is a constraint recognised by graziers elsewhere in northern Australia (McKellar et al., 2015) and almost certainly contributes to the lack of uptake of irrigation in the Roper catchment. 4.4 Crop synopses 4.4.1 Introduction Note that the estimates for land suitability in these synopses represent the total areas of the catchment unconstrained by factors such as water availability, landscape complexity, land tenure, environmental and other legislation and regulations, and a range of biophysical risks such as cyclones, flooding and secondary salinisation. These are addressed elsewhere by the Assessment. The land suitability maps are designed to be used predominantly at the regional scale. Farm-scale planning would require finer-scale, more localised assessment. 4.4.2 Cereal crops Cereal production is well-established in Australia. The area of land devoted to production of grass grains (wheat, barley (Hordeum vulgare), grain sorghum, maize, oats (Avena sativa), triticale (× Triticosecale) etc.) each year has stayed relatively consistent at about 20 million ha over the decade from 2012–13 to 2021–22, yielding over 55 Mt with a value of $19 billion in 2021–22 (ABARES, 2022). Production of cereals greatly exceeds domestic demand, and the majority (82% by value) was exported in 2021–22 (ABARES, 2022). Significant export markets exist for wheat, barley and grain sorghum, with combined exports valued at $15 billion in 2021–22. There are additional niche export markets for grains such as maize and oats. Among the cereals, sorghum (grain) is promising for the Roper catchment. Sorghum is grown over the summer period, coinciding with the Roper wet season. Sorghum can be grown opportunistically using dryland production, although the years in which this could be successfully done will be limited. Cereal crop production is higher and more consistent when irrigation is used. From a land suitability perspective, cereal crops are included in Crop Group 7 (Table 4-2; Figure 4-11). The loamy soils of the Sturt Plateau, the Wilton River Plateau and scattered elsewhere make up about 43% of the catchment. Much of this area is suitable (with moderate or minor limitations) for spray irrigation in the dry season but inadequate drainage in the wet season substantially reduces the area suitable for wet-season spray irrigation. Clays (cracking, non-cracking and clay loams) in the Gulf Fall region and the Sturt Plateau make up about 19% of the catchment. Inadequate drainage and deep gilgais (Sturt Plateau especially) reduce the prospects for furrow irrigation, particularly in the wet season. Shallow and/or rocky soils make up 35% of the catchment, and by definition they are unsuitable. Assuming unconstrained development, approximately 3.2 million ha of the Roper catchment is considered to be suitable with moderate limitations (Class 3; Table 4-1) or better (Class 2 or Class 1) for irrigated cereal cropping (Crop Group 7; Table 4-2) using spray irrigation in the dry season. For spray irrigation in the wet season, nearly 2.0 million ha is suitable with moderate limitations (Class 3) or better. Land considered suitable with moderate limitations for furrow irrigation is limited to about 290,000 ha in the dry season and only 110,000 ha in the wet season, due to inadequate soil drainage in clay soils (and/or gilgais are too deep) and because the loamy soils are too permeable. There is potential for dryland cereal production in the wet season over an area of about 440,000 ha. Note that from a land suitability perspective, Crop Group 7 contains both cereal crops and cotton, which is considered under industrial (cotton) in these crop synopses (Section 4.4.6). The ‘winter cereals’ such as wheat and barley are not well-adapted to the climate of the Roper catchment. If grown during winter, they would require full irrigation. To grow cereal crops, farmers will require access to tillage, fertilising, planting, spraying and harvesting equipment. Harvesting is often a contract operation, and in larger growing regions other activities can also be performed under contract. Because of the low relative value of cereals, good returns are made through production at a large scale. This requires machinery to be large so that operations can be completed in a timely way. Table 4-15 provides summary information relevant to the cultivation of cereals, using sorghum (grain) (Figure 4-12) as an example. The companion technical report on agricultural viability and socio-economics (Stokes et al., 2023) provides greater detail for a wider range of crops. Figure 4-11 Modelled land suitability for Crop Group 7 (e.g. sorghum (grain) or maize) using furrow irrigation in (a) the wet season and (b) dry season Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). Figure 4-12 Sorghum (grain) Photo: CSIRO For more information on this figure please contact CSIRO on enquiries@csiro.au Table 4-15 Sorghum (grain) For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 4.4.3 Pulse crops (food legume) Pulse production is well-established in Australia. The area of land devoted to production of pulses (mainly chickpea, lupin (Lupinus spp.) and field pea (Pisum sativum)) each year has varied from 1.1 to 2.0 million ha over the decade from 2012–13 to 2021–22, yielding over 3.8 Mt with a value of $2.5 billion in 2021–22 (ABARES, 2022). The vast majority of pulses (93% by value) were exported in 2021–22 (ABARES, 2022). Pulses produced in the Roper catchment would most likely be exported, although there is presently no cleaning or bulk-handling facility. Pulses often have a short growing season, and are suited to opportunistic dryland production over a rainy season or more continuous irrigated production, often in rotation with cereals. Not all pulse crops are likely to be suited to the Roper catchment. Those that are ‘tender’ such as field peas and beans may not be well-suited to the highly desiccating environment and periodically high temperatures. Direct field experimentation in the catchment is required to confirm this, for these and other species. In the Roper catchment, mungbean and chickpea are likely to be well suited. From a land suitability perspective, pulse crops are included in Crop Group 10 (Table 4-2; Figure 4-13). The loamy soils of the Sturt Plateau, the Wilton River Plateau and scattered elsewhere make up about 43% of the catchment. Much of this area is suitable (with moderate or minor limitations) for spray irrigation in the dry season. Clays (cracking, non-cracking and clay loams) in the Gulf Fall region and the Sturt Plateau make up about 19% of the catchment. Inadequate drainage and deep gilgais (Sturt Plateau especially) reduce the prospects for furrow irrigation. Shallow and/or rocky soils make up 35% of the catchment, and by definition they are unsuitable. Assuming unconstrained development, approximately 3.1 million ha of the Roper catchment is considered to be suitable with moderate limitations (Class 3; Table 4-1) or better (Class 2 or Class 1) for irrigated pulse cropping (Crop Group 10; Table 4-2) using spray irrigation in the dry season. Land considered suitable with moderate limitations for furrow irrigation is limited to about 210,000 ha in the dry season, due to inadequate soil drainage in clay soils (and/or gilgais are too deep) and because the loamy soils are too permeable. There is potential for dryland pulse production in the wet season over an area of about 350,000 ha. Note that from a land suitability perspective, Crop Group 10 contains the pulse crops mungbean and chickpea, while soybean, is considered under oilseeds in these crop synopses (Section 4.4.4). Pulses are often advantageous in rotation with other crops because they provide a disease break and, being legumes, can provide nitrogen for subsequent crops. Even where this is not the case, their ability to meet their own nitrogen needs can be beneficial in reducing costs of fertiliser and associated freight. Pulses such as mungbean and chickpea can also be of high value (historical prices have reached >$1000/t) and so the freight costs as a percentage of the value of the crop are lower compared with cereal grains. To grow pulse crops, farmers will require access to tillage, fertilising, planting, spraying and harvesting equipment. Harvesting is generally a contract operation, and in larger growing regions other activities can also be performed under contract. The equipment required for pulse crops is the same as is required for cereal crops, so farmers intending on a pulse and cereal rotation would not need to purchase extra ‘pulse-specific’ equipment. Table 4-16 provides summary information relevant to the cultivation of many pulses, using mungbean (Figure 4-14) as an example. The companion technical report on agricultural viability and socio-economics (Stokes et al., 2023) provides greater detail for a wider range of crops. Figure 4-13 Modelled land suitability for mungbean (Crop Group 10) in the dry season using (a) furrow irrigation and (b) spray irrigation Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). Figure 4-14 Mungbean Photo: CSIRO For more information on this figure please contact CSIRO on enquiries@csiro.au Table 4-16 Mungbean For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 4.4.4 Oilseed crops The area of land devoted to production of oilseeds (predominantly canola (Brassica napus)) each year has varied between 2.1 and 3.4 million ha over the decade from 2012–13 to 2021–22, yielding over 8.4 Mt with a value of $6.1 billion in 2021–22 (ABARES, 2022). The majority of oilseed production (98% by value) was exported in 2021–22 (ABARES, 2022). Canola dominates Australian oilseed production accounting for 98% of the gross value of oilseeds in 2021–22, while soybeans, sunflower (Helianthus annus) and other oilseeds (including peanuts) each accounted for less than 1%. Soybean, canola and sunflowers are oilseed crops used to produce vegetable oils, biodiesel and as high protein meals for intensive animal production. Soybean is also used in processed foods such as tofu; it can provide both green manure and soil benefits in crop rotations, with symbiotic nitrogen fixation adding to soil fertility and sustainability in an overall cropping system. Soybean is used commonly as a rotation crop with sugarcane in northern Queensland. Summer oilseed crops such as soybean and sunflower are more suited to tropical environments than winter-grown oilseed crops such as canola. Cottonseed is also classified as an oilseed and is used for animal production. Soybean is sensitive to photoperiod (day length) and requires careful consideration in selection of the appropriate variety for a particular sowing window. From a land suitability perspective, soybean is included in Crop Group 10 (Table 4-2; Figure 4-15). The loamy soils of the Sturt Plateau, the Wilton River Plateau and scattered elsewhere make up about 43% of the catchment. Much of this area is suitable (with moderate or minor limitations) for spray irrigation in the dry season. Clays (cracking, non-cracking and clay loams) in the Gulf Fall region and the Sturt Plateau make up about 19% of the catchment. Inadequate drainage and deep gilgais (Sturt Plateau especially) reduce the prospects for furrow irrigation. Shallow and/or rocky soils make up 35% of the catchment, and by definition they are unsuitable. Assuming unconstrained development, approximately 3.1 million ha of the Roper catchment is considered to be suitable with moderate limitations (Class 3; Table 4-1) or better (Class 2 or Class 1) for irrigated soybean cropping (Crop Group 10; Table 4-2) using spray irrigation in the dry season. Land considered suitable with moderate limitations for furrow irrigation is limited to about 210,000 ha in the dry season, due to inadequate soil drainage in clay soils (and/or gilgais are too deep) and because the loamy soils are too permeable. There is potential for dryland soybean production in the wet season over an area of about 350,000 ha. Note that from a land suitability perspective, Crop Group 10 contains both soybean and pulse crops such as mungbean and chickpea, which are considered in Section 4.4.3. To grow oilseed crops, farmers will require access to tillage, fertilising, planting, spraying and harvesting equipment. Harvesting is generally a contract operation, and in larger growing regions other activities can also be performed under contract. The equipment required for oilseed crops is the same as is required for cereal crops, so farmers intending on an oilseed and cereal rotation would not need to purchase extra ‘oilseed-specific’ equipment. Table 4-17 provides summary information relevant to the cultivation of oilseed crops, using soybean (Figure 4-16) as an example. The companion technical report on agricultural viability and socio-economics (Stokes et al., 2023) provides greater detail for a wider range of crops. Figure 4-15 Modelled land suitability for soybean (Crop Group 10) in the dry season using (a) furrow irrigation and (b) spray irrigation Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). Figure 4-16 Soybean Photo: CSIRO For more information on this figure please contact CSIRO on enquiries@csiro.au Table 4-17 Soybean For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 4.4.5 Root crops, including peanut Root crops including peanut, sweet potatoes (Ipomoea batatas) and cassava (Manihot esculenta), are potentially well-suited to the lighter soils found across much of the Roper catchment. Root crops such as these are not suited to growing on heavier clay soils because they need to be pulled from the ground for harvest, and the heavy clay soils, such as cracking clays, are not conducive to mechanical pulling. While peanut is technically an oilseed crop, it has been included in the root crop category due to its similar land suitability and management requirements (i.e. the need for it to be ‘pulled’ from the ground as part of the harvest operation). The most widely grown root crop in Australia, peanut is a legume crop that requires little or no nitrogen fertiliser and is very well-suited to growing in rotation with cereal crops, as it is frequently able to fix atmospheric nitrogen in the soil for following crops. The Australian peanut industry currently produces approximately 15,000 to 20,000 t/year from around 11,000 ha, which is too small an industry to be reported separately in Australian Bureau of Agricultural and Resource Economics and Sciences statistics (ABARES, 2022). The Australian peanut industry is concentrated in Queensland. In northern Australia a production area is present on the Atherton Tablelands, and peanuts could likely be grown in the Roper catchment. The Peanut Company of Australia established a peanut-growing operation at Katherine in 2007 and examined the potential of both wet- and dry-season peanut crops, mostly in rotation with maize. Due to changing priorities within the company, coupled with some agronomic challenges (Jakku et al., 2016), the company sold its land holdings in Katherine in 2012 (and Bega bought the rest of the company in 2018). For peanuts to be successful, considerable planning would be needed in determining the best season for production and practical options for crop rotations. The nearest peanut processing facilities to the Roper catchment are Tolga on the Atherton Tablelands or Kingaroy in southern Queensland. From a land suitability perspective, peanut is included in Crop Group 6 (Table 4-2; Figure 4-17). The loamy soils of the Sturt Plateau, the Wilton River Plateau and scattered elsewhere make up about 43% of the catchment. Much of this area is suitable (with moderate or minor limitations) for spray irrigation in the dry season but inadequate drainage in the wet season substantially reduces the area suitable for wet-season spray irrigation. Clays (cracking, non-cracking and clay loams) in the Gulf Fall region and the Sturt Plateau make up about 19% of the catchment and these heavier textured soils are generally unsuited to root crops. Shallow and/or rocky soils make up 35% of the catchment, and by definition they are unsuitable. Assuming unconstrained development, approximately 2.9 million ha of the Roper catchment is considered to be suitable with moderate limitations (Class 3; Table 4-1) or better (Class 2 or Class 1) for irrigated root crops (Crop Group 6; Table 4-2) using spray irrigation in the dry season. For spray irrigation in the wet season, nearly 1.4 million ha is suitable with moderate limitations (Class 3) or better. Furrow irrigation is not suited to either season with wetness on the heavier textured soils being the limitation and the lighter textured soils being too permeable and therefore furrow irrigation was not considered in the land suitability analysis. To grow root crops, farmers will require access to tillage, fertilising, planting, spraying and harvesting equipment. The harvesting operation requires specialised equipment to ‘pull’ the crop from the ground, and then to pick it up after a drying period. Peanuts are usually dried soon after harvest in industrial driers. Table 4-18 provides summary information relevant to the cultivation of root crops, using peanut (Figure 4-18) as an example. The companion technical report on agricultural viability and socio- economics (Stokes et al., 2023) provides greater detail for a wider range of crops. Figure 4-17 Modelled land suitability for peanut (Crop Group 6) using spray irrigation in (a) the wet season and (b) the dry season Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). Figure 4-18 Peanuts Photo: Shutterstock For more information on this figure please contact CSIRO on enquiries@csiro.au Table 4-18 Peanut For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 4.4.6 Industrial (cotton) Dryland and irrigated cotton production are well-established in Australia. The area of land devoted to cotton production varies widely from year to year, largely in response to availability of water, varying from 70,000 to 600,000 ha between 2012–13 and 2021–22, with an average of 400,000 ha grown per year for the decade (ABARES, 2022). Likewise, the gross value of cotton lint production varied greatly over the past decade, from $0.3 billion in 2019–20 to $5.2 billion in 2021–22. Genetically modified cotton varieties were introduced in 1996 and now account for almost all cotton produced in Australia (over 99%). Australia was the fourth largest exporter of cotton in 2022, behind the United States, India and Brazil. Cottonseed is a by-product of cotton processing and is a valuable cattle feed. Mean lint production in Australia in 2015–16 was 2.0 t/ha (ABARES, 2022). Commercial cotton has had a long but discontinuous history of production in northern Australia, including in Broome, the Fitzroy River and the Ord River Irrigation Area in WA; in Katherine and Douglas–Daly in the NT; and near Richmond and Bowen in northern Queensland. An extensive study undertaken by the Australian Cotton Cooperative Research Centre in 2001 (Yeates, 2001) noted that past ventures suffered from: • a lack of capital investment • too rapid movement to commercial production • a failure to adopt a systems approach to development • climate variability. Mistakes in pest control were also a major issue in early projects. Since the introduction of genetically modified cotton in 1996, yields and incomes from cotton crops have increased in most regions of Australia. The key benefits of genetically modified cotton (compared to conventional cotton) are savings in insecticide and herbicide use, improved tillage management and human health benefits associated with reduced handling of farm chemicals. In addition, farmers are now able to forward-sell their crop as part of a risk management strategy. Growers of genetically modified cotton are required to comply with the approved practices for growing the genetically modified varieties, including preventative resistance management. Research and commercial test farming have demonstrated that the biophysical challenges are manageable if the growing of cotton is tailored to the climate and biotic conditions of northern Australia (Yeates et al., 2013). In recent years irrigated cotton crops achieving 10 bales/ha have been grown successfully in the Burdekin irrigation region and experimentally in the Gilbert catchment of north Queensland. New genetically modified cotton using CSIRO varieties that are both pest and herbicide resistant are an important component of these northern cotton production systems. Climate constraints will continue to limit production potential of northern cotton crops when compared to cotton grown in more favourable climate regions of NSW and Queensland. On the other hand, the low risk of rainfall occurring during late crop development favours production in northern Australia, as it minimises the likelihood of late-season rainfall that can downgrade fibre quality and price. Demand for Australian cotton exhibiting long and fine attributes is expected to increase by 10 to 20% of the market during the next decade and presents local producers with an opportunity in targeting production of high-quality fibre. From a land suitability perspective, cotton is included in Crop Group 7 (Table 4-2; Figure 4-19). The loamy soils of the Sturt Plateau, the Wilton River Plateau and scattered elsewhere make up about 43% of the catchment. Much of this area is suitable (with moderate or minor limitations) for spray irrigation in the dry season but inadequate drainage in the wet season substantially reduces the area suitable for wet-season spray irrigation. Clays (cracking, non-cracking and clay loams) in the Gulf Fall region and the Sturt Plateau make up about 19% of the catchment. Inadequate drainage and deep gilgais (Sturt Plateau especially) reduce the prospects for furrow irrigation, particularly in the wet season. Shallow and/or rocky soils make up 35% of the catchment, and by definition they are unsuitable. Assuming unconstrained development, approximately 3.2 million ha of the Roper catchment is considered to be suitable with moderate limitations (Class 3; Table 4-1) or better (Class 2 or Class 1) for irrigated cotton (Crop Group 7; Table 4-2) using spray irrigation in the dry season. For spray irrigation in the wet season, nearly 2.0 million ha is suitable with moderate limitations (Class 3) or better. Land considered suitable with moderate limitations for furrow irrigation is limited to about 290,000 ha in the dry season and only 110,000 ha in the wet season, due to inadequate soil drainage in clay soils (and/or gilgais are too deep) and because the loamy soils are too permeable. There is potential for dryland cotton production in the wet season over an area of about 440,000 ha. Note that from a land suitability perspective, Crop Group 7 contains both cotton and cereal crops, which are considered elsewhere in these crop synopses (Section 4.4.2). In addition to a normal row planter and spray rig equipment used in cereal production, cotton requires access to suitable picking and module or baling equipment, as well as transport to processing facilities. Initial development costs and scale of establishing cotton production in the catchments would need to consider sourcing of external contractors and could provide an opportunity to develop local contract services to support a growing industry. Cotton production is also highly dependent on access to processing plants (cotton gins). The first cotton gin in the NT is set to open in mid-2023 near Katherine and would be the processing facility for cotton grown in the Roper catchment. Niche industrial crops, such as guar (Cyamopsis tetragonoloba) and chia (Salvia hispanica), may be feasible for the Roper catchment, but there is only limited verified agronomic and market data on these crops. Past research on guar has been conducted in the NT and current trials are underway. Hemp is a photoperiod-sensitive summer annual with a growing season between 70 and 120 days, depending on variety and temperature. Hemp is well suited to growing in rotation with legumes as hemp can use the nitrogen fixed by the legume crop. Industrial hemp can be harvested for grain with modifications to conventional headers, otherwise all other farming machinery for ground preparation, fertilising and spraying can be used. There are legislative restrictions to growing hemp in Australia, and jurisdictions including the NT are implementing industrial hemp legislation to license growing of industrial hemp to facilitate development of the industry. The companion technical report on agricultural viability and socio-economics (Stokes et al., 2023) provides greater detail for a wider range of industrial crops. Table 4-19 describes some key considerations relating to cotton production (Figure 4-20). Figure 4-19 Modelled land suitability for cotton (Crop Group 7) using furrow irrigation in (a) the wet season and (b) the dry season Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). Figure 4-20 Cotton Photo: CSIRO For more information on this figure please contact CSIRO on enquiries@csiro.au Table 4-19 Cotton For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 4.4.7 Forages Forage, hay and silage are crops that are grown specifically for consumption by animals. Forage is consumed in the paddock in which it is grown, which is often referred to as ‘stand and graze’. Hay is cut, dried, baled and stored before being fed to animals at a time when natural pasture production is low (generally towards the end of the dry season). Silage use resembles that for hay, but crops are stored wet, in anaerobic conditions where fermentation occurs to preserve the feed’s nutritional value. Dryland and irrigated production of forage crops is well-established throughout Australia, with over 20,000 producers, most of whom are not specialist forage crop producers. Approximately 85% of forage production is consumed domestically, with the rest primarily used on live export ships often in a pelleted form. The largest consumers are the horse, dairy and beef feedlot industries. Forage crops are also widely used in horticulture for mulches and for erosion control. While there is currently already consumption use of forages by the northern beef industry, forage costs comprise less than 5% of beef production costs (Gleeson et al., 2012), so there is likely room for further expansion of forage production. Non-leguminous forage, hay and silage The Roper catchment is suited to dryland or irrigated production of non-leguminous forage, hay and silage. A significant amount of dryland hay production occurs in the Douglas–Daly region, south of Darwin. Most of the hay produced in the NT is for feeding cattle locally destined for live export or used as part of a feed pellet used on boats carrying live export cattle. Forage crops, both annual and perennial, include sorghum (Sorghum spp.), Rhodes grass, maize and Jarra grass (Digitaria milanjiana ‘Jarra’), with particular cultivars specific for forage. These grass forages require considerable amounts of water and nitrogen as they can be high-yielding (20 to 40 t dry matter/ha). Given their rapid growth, crude protein levels can drop very quickly, reducing their value as a feed for livestock. To maintain high nutritive value, high levels of nitrogen need to be applied and in the case of hay, the crop needs to be cut every 45 to 60 days. After cutting, the crop grows back without the need for re-sowing. The rapid growth of forage during the late spring and summer months can make it challenging to match animals to forage growth so that it is kept leafy and nutritious and does not become rank and of low quality. Dryland hay production from perennials gives producers the option of irrigation when required or, if water becomes limiting, allowing the pasture to remain dormant before water again becomes available. Silage can be made from a number of crops, such as grasses, maize and forage sorghum. From a land suitability perspective, Rhodes grass is included in Crop Group 14 (Table 4-2; Figure 4-21). The loamy soils of the Sturt Plateau, the Wilton River Plateau and scattered elsewhere make up about 43% of the catchment. Much of this area is suitable (with moderate or minor limitations) for spray irrigation in the dry season but inadequate drainage in the wet season substantially reduces the area suitable for wet-season spray irrigation. Clays (cracking, non-cracking and clay loams) in the Gulf Fall region and the Sturt Plateau make up about 19% of the catchment. Inadequate drainage and deep gilgais (Sturt Plateau especially) reduce the prospects for furrow irrigation, particularly in the wet season. Shallow and/or rocky soils make up 35% of the catchment, and by definition they are unsuitable in all but a few instances. Assuming unconstrained development, approximately 3.2 million ha of the Roper catchment is considered to be suitable with moderate limitations (Class 3; Table 4-1) or better (Class 2 or Class 1) for irrigated cropping of annual forages (Crop Group 12; Table 4-2) using spray irrigation in the dry season. For spray irrigation in the wet season, nearly 2.0 million ha is suitable with moderate limitations (Class 3) or better. Land considered suitable with moderate limitations for furrow irrigation of annual forages is limited to about 290,000 ha in the dry season and only 110,000 ha in the wet season, due to inadequate soil drainage in clay soils (and/or gilgais are too deep) and because the loamy soils are too permeable. There is potential for dryland production of annual forages in the wet season over an area of about 410,000 ha. For the perennial Rhodes grass, nearly 4.0 million ha are suitable with moderate limitations under spray irrigation and about 330,000 ha under furrow irrigation. Apart from irrigation infrastructure, the equipment needed for forage production is machinery for planting and fertilising. Spraying equipment is also desirable but not necessary. Cutting crops for hay or silage requires more specialised harvesting, cutting, baling and storage equipment. Table 4-20 describes Rhodes grass production (Figure 4-22) for hay over a one year of 6-year cycle. Information similar to that in Table 4-20 for grazed forage crops is presented in the companion technical report on agricultural viability and socio-economics (Stokes et al., 2023). For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 4-21 Modelled land suitability for Rhodes grass (Crop Group 14) using (a) spray irrigation and (b) furrow irrigation Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). Figure 4-22 Rhodes grass Photo: CSIRO Table 4-20 Rhodes grass For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Forage legume The use of forage legumes is similar to that of forage grasses. They are generally grazed by animals but can also be cut for silage or hay. Some forage legumes are well-suited to the Roper catchment, and would be considered among the more promising opportunities for irrigated agriculture (Figure 4-23). Forage legumes are desirable because of their high protein content and their ability to fix atmospheric nitrogen. The nitrogen fixed during a forage legume phase is often in excess of that crop’s requirements, which leaves the soil with additional nitrogen. Forage legumes are being used by the northern cattle industry, and farmers primarily engaged in extensive cattle production could use irrigated forage legumes to increase the capacity of their enterprise, turning out more cattle from the same area. Cavalcade (Centrosema pascuorum ‘Cavalcade’) and lablab are currently grown in northern Australia and would be well-suited to the Roper catchment. Cavalcade is already grown in the catchments and used for grazing and for hay. Hay crops are commonly used as a component of forage pellets that are used to feed live export cattle in holding yards and on boats during transport. From a land suitability perspective, forage legumes such as Cavalcade and lablab are included in Crop Group 13 (Table 4-2; Figure 4-23). The loamy soils of the Sturt Plateau, the Wilton River Plateau and scattered elsewhere make up about 43% of the catchment. Much of this area is suitable (with moderate or minor limitations) for spray irrigation in the dry season but inadequate drainage in the wet season substantially reduces the area suitable for wet-season spray irrigation. Clays (cracking, non-cracking and clay loams) in the Gulf Fall region and the Sturt Plateau make up about 19% of the catchment. Inadequate drainage and deep gilgais (Sturt Plateau especially) reduce the prospects for furrow irrigation, particularly in the wet season. Shallow and/or rocky soils make up 35% of the catchment, and by definition they are unsuitable. Assuming unconstrained development, approximately 3.3 million ha of the Roper catchment is considered to be suitable with moderate limitations (Class 3; Table 4-1) or better (Class 2 or Class 1) for irrigated forage legumes (Crop Group 13; Table 4-2) using spray irrigation in the dry season. For spray irrigation in the wet season, nearly 1.7 million ha is suitable with moderate limitations (Class 3) or better. Land considered suitable with moderate limitations for furrow irrigation is limited to about 300,000 ha in the dry season and only 20,000 ha in the wet season, due to inadequate soil drainage in clay soils (and/or gilgais are too deep) and because the loamy soils are too permeable. There is potential for dryland forage legume production in the wet season over an area of about 280,000 ha. The equipment needed for grazed forage legume production is similar to that for forage grasses, that is, a planting method, with fertilising and spraying equipment being desirable but not essential. Cutting crops for hay or silage requires more specialised harvesting, cutting, baling and storage equipment. Table 4-21 describes Cavalcade production over a 1-year cycle. The comments could be applied equally to lablab production (Figure 4-24). For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 4-23 Modelled land suitability for Cavalcade (Crop Group 13) in the wet season using (a) spray irrigation and (b) furrow irrigation Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). Figure 4-24 Lablab Photo: CSIRO Table 4-21 Cavalcade For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 4.4.8 Horticulture Horticulture is an important and widespread Australian industry, occurring in every state. Horticulture production encompasses a very wide range of intensive cultivated food and ornamental crops, including the vast range of fruit and vegetable crops. Horticultural production varied between 2.9 and 3.3 Mt per year between 2012–13 and 2021–22, of which 65 to 70% was vegetables (ABARES, 2022). Unlike broadacre crops, most horticultural production in Australia is consumed domestically. The total gross value of horticultural production was $13.2 billion in 2021–22, up from 9.3 billion in 2012–13, of which 24% was from exports (ABARES, 2022). Horticulture is also an important source of jobs, employing approximately one-third of all people employed in agriculture. Production of horticultural crops is highly seasonal and may involve individual farms growing a range of crops, or growing the same crop with sequential planting dates. The importance of freshness in many horticultural products means seasonality of supply is important in the market. The value of horticulture crops can vary widely, with price changes occurring over very short periods of time (weeks). Part of the attraction to growing horticulture crops in the Roper catchment is to supply southern markets when southern growing regions are unable to produce due to climate restrictions. Transport of horticulture produce can involve significant costs, so achieving a price premium for ‘out of season’ production will be required for successful production in the Roper catchment. This requires a heightened understanding of risks, markets, transport and supply chain issues. Horticultural production systems are generally more intensive than broadacre farming, requiring higher capital investment in establishing farm infrastructure, and requiring higher ongoing inputs for production. Picking and packing operations involve significant labour. Attracting sufficient seasonal workers to the Roper catchment for harvesting season would need consideration. Horticulture (row crops) Horticulture row crops are generally short-lived, annual crops, grown in the ground such as watermelon and rockmelon (Cucumis melo var. cantalupensis). Almost all produce is shipped to major markets (cities) where central markets are located. Row crops such as watermelon and rockmelon use staggered plantings over a season (for example every 2 to 3 weeks) so that the period over which harvested produce is sold can be extended. This strategy allows better use of labour and allows better management for risks of price fluctuations. Often only a short period of time with very high prices is enough to make melon production a profitable enterprise. Horticultural row crops are well-established throughout the NT. The NT melon industry, consisting of watermelon (seedless), rockmelon and honeydew (Cucumis melo (Inodorus Group) 'Honey Dew'), produces approximately 25% of Australia’s melons. Melon production is well suited across many parts of the NT and would be well suited to the Roper catchment. From a land suitability perspective, intensive horticulture row crops such as rockmelons are included in Crop Group 3 (Table 4-2). The loamy soils of the Sturt Plateau, the Wilton River Plateau and scattered elsewhere make up about 43% of the catchment. Much of this area is suitable (with moderate or minor limitations) for spray irrigation in the dry season but inadequate drainage in the wet season substantially reduces the area suitable for wet-season spray irrigation. In addition, disease risk is very high for horticulture row crops in the wet season. Clays (cracking, non-cracking and clay loams) in the Gulf Fall region and the Sturt Plateau make up about 19% of the catchment. Inadequate drainage and deep gilgais (Sturt Plateau especially) reduce the prospects for furrow irrigation, particularly in the wet season. Shallow and/or rocky soils make up 35% of the catchment, and by definition they are unsuitable. A wide range of horticultural row crops are considered in the land suitability analysis (crop groups 3, 4, 5 and 18; Table 4-2; Figure 4-25). Assuming unconstrained development, between about 3.1 million ha and 3.4 million ha of the Roper catchment is considered to be suitable with moderate limitations (Class 3; Table 4-1) or better (Class 2 or Class 1) using spray or trickle irrigation in the dry season. Land considered suitable with moderate limitations for furrow irrigation of sweet corn (Crop Group 18) is limited to about 330,000 ha in the dry season and only 110,000 ha in the wet season, due to inadequate soil drainage in clay soils (and/or gilgais are too deep) and because the loamy soils are too permeable. Horticulture typically requires specialised equipment and a large labour force. Therefore, a system for attracting, managing and retaining sufficient staff is also required. Harvesting is often by hand, but packing equipment is highly specialised. Irrigation is mostly with micro equipment, but overhead spray is also feasible. Leaf fungal diseases need to be more carefully managed with spray irrigation. Micro spray equipment has the advantage of also being a nutrient delivery (fertigation) mechanism, as fertiliser can be delivered via the irrigation water. Table 4-22 describes some key considerations relating to row crop horticulture production, with rockmelon (Figure 4-26) as an exemplar of those relating to row crop horticultural production more broadly. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 4-25 Modelled land suitability for (a) cucurbits (e.g. rockmelon) (Crop Group 3) using trickle irrigation in the dry season and (b) root crops such as onion (Crop Group 6) using spray irrigation in the wet season Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 4-26 Melon crop in Mataranka area Photo: CSIRO Table 4-22 Rockmelon For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au Horticulture (tree crops) Some fruit and tree crops – such as mangoes and citrus (Citrus spp.) – are well-suited to the climate of the Roper catchment and mangoes are already grown within the region. Other species such as avocado (Persea americana) and lychee (Litchi chinensis) are not likely to be as well- adapted to the climate and soils. Tree crops are generally not well-suited to cracking clays, which make up some of the suitable soils for irrigated agriculture in the Roper catchment. Fruit production shares many of the marketing and risk features of horticultural row crops, such as a short season of supply and highly volatile prices as a result of highly inelastic supply and demand. Managing these issues requires a heightened understanding of risks, markets, transport and supply chain issues. The added disadvantage of fruit tree production is the time lag between planting and production, meaning decisions to plant need to be made with a long time frame for production and return in mind. Mango production in the NT is buffered somewhat against large- scale competition as its crop matures earlier than the main production areas in Queensland and it can achieve high returns. Mango production in the NT had a gross value of $129 million in 2020, accounting for 38% of the $341 million total value of horticultural production in the NT, and half of mangoes produced in Australia (Sangha et al., 2022). The perennial nature of tree crops makes a reliable year-round supply of water essential. However, some species, such as mango and cashew (Anacardium occidentale), can survive well under mild water stress until flowering (generally August to October for most fruit trees). It is critical for optimum fruit and nut production that trees are not water stressed from flowering through to harvest. This is the period approximately from August up to November through to February, depending on the species. This is a period in the Roper catchment when very little rain falls, and farmers would need to have a system in place to access irrigation water during this time. From a land suitability perspective, intensive horticultural tree crops such as mango are included in Crop Group 1, the monsoonal tropical tree crops (Table 4-2). The loamy soils of the Sturt Plateau, the Wilton River Plateau and scattered elsewhere make up about 43% of the catchment. Much of this area is suitable (with moderate or minor limitations) for spray irrigation. Inadequate drainage in the wet season constrains a larger area. Clays (cracking, non-cracking and clay loams) in the Gulf Fall region and the Sturt Plateau make up about 19% of the catchment. Inadequate drainage and deep gilgais (Sturt Plateau especially) reduce the prospects for horticultural tree crops. Shallow and/or rocky soils make up 35% of the catchment, and by definition they are unsuitable. A wide range of horticultural tree crops are considered in the land suitability analysis (crop groups 1, 2, 20 and 21; Table 4-2; Figure 4-27). Assuming unconstrained development, between about 2.1 million ha (citrus) and 3.5 million ha (mango) of the Roper catchment is considered to be suitable with moderate limitations (Class 3; Table 4-1) or better (Class 2 or Class 1) using spray or trickle irrigation. Furrow irrigation was not considered for horticultural tree crops. Specialised equipment for fruit and nut tree production is required. The requirement for a timely and significant labour force necessitates a system for attracting, managing and retaining sufficient staff. Tree pruning and packing equipment is highly specialised for the fruit industry. Optimum irrigation is usually via micro spray. This equipment is also able to deliver fertiliser directly to the trees through fertigation. Table 4-23 describes some key considerations relating to mango production (Figure 4-28) in the Roper catchment, as an exemplar of those relating to tree crop production more broadly. Similar information for other fruit tree crops is described in the companion technical report on agricultural viability and socio-economics (Stokes et al., 2023). For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 4-27 Modelled land suitability for (a) mango (Crop Group 1) and (b) lime (Crop Group 2), both grown using trickle irrigation Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). Figure 4-28 Mangoes Photo: Shutterstock Table 4-23 Mango For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 4.4.9 Plantation tree crops (silviculture) Of the potential tree crops that could be grown in the Roper catchment, Indian sandalwood and African mahogany are the only two that would be considered economically feasible. Many other plantation species could be grown; however, returns are much lower than for these two crops. African mahogany is well-established in commercial plantations near Katherine and Indian sandalwood is also grown in Katherine, the Ord valley Western Australia, and in north Queensland. Plantation timber species require over 15 years to grow, but once established can tolerate prolonged dry periods. Irrigation water is critical in the establishment and first 2 years of a plantation. In the case of Indian sandalwood, the provision of water is not just for the trees themselves but also for the leguminous host plant associated with Indian sandalwood, as it is a semi-parasite. From a land suitability perspective, plantation tree crops such as Indian sandalwood, African mahogany and teak (Tectona grandis) are included in crop groups 15, 16 and 17 (Table 4-2). The loamy soils of the Sturt Plateau, the Wilton River Plateau and scattered elsewhere make up about 43% of the catchment. Much of this area is suitable (with moderate or minor limitations) for trickle irrigation but inadequate drainage in the wet season substantially reduces the area suitable for teak. Clays (cracking, non-cracking and clay loams) in the Gulf Fall region and the Sturt Plateau make up about 19% of the catchment. Inadequate drainage and deep gilgais (Sturt Plateau especially) reduce the prospects for Indian sandalwood. Shallow and/or rocky soils make up 35% of the catchment, and by definition they are unsuitable. Depending on the specific tree species being planted and their tolerance to poorly drained soils and waterlogging, the suitable areas vary considerably. A range of silviculture trees were considered in the land suitability analysis (crop groups 15, 16 and 17; Table 4-2). Assuming unconstrained development, between about 2.6 million ha (teak) and 3.7 million ha (African mahogany) of the Roper catchment is considered to be suitable with moderate limitations (Class 3; Table 4-1) or better (Class 2 or Class 1) using trickle irrigation (Figure 4-29). Furrow irrigation was considered for Indian sandalwood only and 170,000 ha was assessed as suitable with moderate limitations. Table 4-24 describes Indian sandalwood production (Figure 4-30). For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 4-29 Modelled land suitability for Indian sandalwood (Crop Group 15) grown using (a) trickle or (b) furrow irrigation Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). Figure 4-30 Indian sandalwood and host plants Photo: CSIRO Table 4-24 Indian sandalwood For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au For more information on this figure, table or equation please contact CSIRO on enquiries@csiro.au 4.4.10 Niche crops Niche crops such as guar, chia, quinoa (Chenopodium quinoa), bush foods, and others may be feasible in the Roper catchment, but there is limited verified agronomic and market data available for these crops. Niche crops are niche due to the limited demand for their products. As a result, small-scale production can lead to very attractive prices, but only a small increase in productive area can flood the market, leading to greatly reduced prices and making production unsustainable. There is growing interest in bush foods, but insufficient publicly available information for inclusion with the analyses of irrigated crops options in this Assessment. Bush food production systems could take many forms, from culturally appropriate wild harvesting targeting Indigenous consumers to modern mechanised farming and processing, like macadamia (Macadamia integrifolia) farming. The choice of production system would have implications for the extent of Indigenous involvement throughout the supply chain (farming, processing, marketing and/or consumption), the scale of the markets that could be accessed (in turn affecting the scale of the industry for that bushfood), the price premiums that produce may be able to attract, and the viability of those industries. The current publicly available information on bush foods is mainly focused on eliciting Indigenous aspirations, on biochemical analysis (for safety, nutrition and efficacy of potential health benefits), and on considerations of safeguarding Indigenous intellectual property. Analysing bush foods in a comparable way to other crop options in this report would first require these issues to be resolved, for communities to agree on the preferred type of production systems (and pathways for development), and for agronomic information on yields, production practices and costs to be publicly available. Past research on guar has been conducted in the NT and current trials are underway in north Queensland, which could prove future feasibility. There is increasing interest in non-leguminous, small-seeded crops such as chia and quinoa, which have high nutritive value. The market size for these niche crops is quite small compared with cereals and pulses and so the scale of production is likely to be small in the short-to-medium term. There is a small, established chia industry in the Ord River region of WA, but its production and marketing statistics are largely commercial in confidence. Nearly all Australian production of chia is contracted to The Chia Company of Australia or is exported to China. In Australia, The Chia Company produces whole chia seeds, chia bran, ground chia seed and chia oil for wholesale and retail sale and exports these products to 36 countries. The growing popularity of quinoa in recent years is attached to its marketing as a super food. It is genetically diverse and has not been the subject of long-term breeding programs. This diversity means it is well-suited to a range of environments, including northern Australia, where its greatest opportunity is as a short-season crop in the dry season under irrigation. It is a high-value crop with farm gate prices of about $1000/tonne. Trials of quinoa production have been conducted at the Katherine Research Station (approximately 50 km from the western edge of the Roper catchment), with reasonable yields being returned. More testing is required in the northern environments of the Roper catchment, before quinoa could be recommended for commercial production. For more information on this figure please contact CSIRO on enquiries@csiro.au Figure 4-31 Quinoa crop Source: Jonas Ingold, LID 4.5 Aquaculture 4.5.1 Introduction There are considerable opportunities for aquaculture development in northern Australia given its natural advantages of a climate suited to farming valuable tropical species, large areas identified as suitable for aquaculture, political stability and proximity to large global markets. The main challenges to developing and operating modern and sustainable aquaculture enterprises are regulatory issues, global cost competitiveness and the remoteness of much of the suitable land area. This section draws on a recent assessment of the opportunities for aquaculture in northern Australia (Irvin et al., 2018) summarising: the three most likely candidate species (Section 4.5.2); an overview of production systems (Section 4.5.3); land suitability for aquaculture within the Roper catchment (Section 4.5.4); and the financial viability of different options for aquaculture development (Section 4.5.5). 4.5.2 Candidate species The three species with the most aquaculture potential in the Roper catchment are black tiger prawns, barramundi and red claw. The first two species are suited to many marine and brackish water environments of northern Australia and have established land-based culture practices and well-established markets for harvested products. Prawns could potentially be cultured in either extensive (low density, low input) or intensive (higher density, higher inputs) pond-based systems in northern Australia, whereas land-based culture of barramundi would likely be intensive. Red claw is a freshwater crayfish that is currently cultured by a much smaller industry than the previous two species. Black tiger prawns Black tiger prawns (Figure 4-32) are found naturally at low abundances across the waters of the western Indo-Pacific region, with wild Australian populations making up the southernmost extent of the species. Within Australia, the species is most common in the tropical north, but does occur at lower latitudes. Figure 4-32 Black tiger prawns Photo: CSIRO Barramundi Barramundi (Figure 4-33) is the most highly produced and valuable tropical fish species in Australian aquaculture. Barramundi inhabit the tropical north of Australia from the Exmouth Gulf in WA through to the Noosa River on Queensland’s east coast. It is also commonly known as the ‘Asian sea bass’ or ‘giant sea perch’ throughout its natural areas of distribution in the Persian Gulf, the western Indo-Pacific region and southern China (Schipp et al., 2007). The attributes that make barramundi an excellent aquaculture candidate are: fast growth (1 kg or more in 12 months); year- round fingerling availability; well-established production methods; and hardiness (i.e. they have a tolerance to low oxygen levels, high stocking densities and handling, as well as a wide range of temperatures) (Schipp et al., 2007). In addition, barramundi are euryhaline (able to thrive and be cultured in fresh and marine water) but freshwater barramundi can have an earthy flavour. Figure 4-33 Barramundi Photo: CSIRO Red claw Red claw is a warm water crayfish species that inhabits still or slow-moving water bodies. The natural distribution of red claw ranges from the tropical catchments of Queensland and the NT to southern New Guinea. The name ‘red claw’ is derived from the distinctive red markings present on the claws of the male crayfish. The traits of red claw that make them attractive for aquaculture production are: a simple life cycle, which is beneficial in that complex hatchery technology is not required (Jones et al., 1998); they can tolerate low oxygen levels (<2 mg/L), which is beneficial in terms of handling, grading and transport (Masser and Rouse, 1997); they have a broad thermal tolerance, with optimal growth achievable between 23 and 31 °C; and they can remain out of the water for extended periods. 4.5.3 Production systems Overview Aquaculture production systems can be broadly classified into extensive, semi-intensive and intensive systems. Intensive systems require high inputs, with expected high outputs: they require high capital outlay; high running costs; specially formulated feed; specialised breeding, water quality and biosecurity processes; and have high production per hectare (in the order of 5,000 to 20,000 kg per ha per crop). Semi-intensive systems involve stocking seed from a hatchery, routine provision of a feed, and monitoring and management of water quality. Production is typically 1000 to 5000 kg per ha per crop. Extensive systems are characterised by low inputs and low outputs: they require less-sophisticated management and often require no supplementary feed because the farmed species live on naturally produced feed in open-air ponds. Extensive systems produce about half the volume of global aquaculture production (but there are few commercial operations in Australia). Water salinity and temperature are the key parameters that determine species selection and production potential for any given location. Suboptimal water temperature (even within tolerable limits) will prolong the production season (slow growth) and increase the risk of disease, reducing profitability. The primary culture units for land-based farming are purpose-built ponds. Pond structures typically include an intake channel, production pond, discharge channel and a bioremediation pond (Figure 4-34). The function of the pond is to be a containment structure, an impermeable layer between the pond water and the local surface water and groundwater. Optimal sites for farms are flat and have sufficient elevation to enable ponds to be completely drained between seasons. It is critical that all ponds and channels can be fully drained during the off (dry-out) season to enable machinery access to sterilise and undertake pond maintenance. Figure 4-34 Schematic of marine aquaculture farm Most production ponds in Australia are earthen. Soils for earthen ponds should have low permeability and high structural stability. Ponds should be lined if the soils are permeable. Synthetic liners have a higher capital cost but are often used in more intensive operations, which require high levels of aeration; conditions that would lead to significant erosion in earthen ponds. Farms use aerators (typically electric paddlewheels and aspirators) to help maintain optimal water quality in the pond, provide oxygen, and create a current that consolidates waste into a central sludge pile (while keeping the rest of the pond floor clear). A medium-sized 50-ha prawn farm in Australia uses around 4 GWh annually, accounting for most of an enterprise’s energy use (Paterson and Miller, 2013). Back-up power capacity sufficient to run all the aerators on the farm, usually via a diesel generator, is essential to be able to cope with power failures. Extensive production systems do not require aeration in most cases. Black tiger prawns For black tiger prawns, a typical pond in the Australian industry would be rectangular in shape, about 1 ha in area and about 1.5 m in depth. The ponds are either wholly earthen, lined on the banks with black plastic and earthen bottoms, or (rarely in Australia) fully lined. Pond grow-out of black tiger prawns typically operates at stocking densities of 25–50 individuals per square metre (termed ‘intensive’ in this report). These pond systems are fitted with multiple aeration units (that could double from 8 to 16 units as the biomass of the prawn crop increases) (Mann, 2012). At the start of each prawn crop, pond bottoms are dried and unwanted sludge from the previous crop is removed, and if needed, additional substrate is added. Prior to filling the ponds, lime is often added to buffer pH, particularly in areas with acid-sulfate soils. The ponds are then filled with filtered seawater and left for about 1 week prior to postlarval stocking. Algal blooms in the water are encouraged through addition of organic fertiliser to provide shading for prawns, discourage benthic algal growth, and stimulate growth of plankton as a source of nutrition (QDPIF, 2006). Postlarvae are purchased from hatcheries and grow rapidly into small prawns in the first month after stocking, relying mainly on the natural productivity (zooplankton, copepods and algae) supported by the algal bloom for their nutrition. Approximately 1 month after the prawns are stocked, pellet feed becomes the primary nutrition source. Feed is a major cost of prawn production; around 1.5 kg of feed is required to produce 1 kg of prawns. Prawns typically reach optimal marketable size (30 g) within 6 months. After harvest, prawns are usually processed immediately, with larger farms having their own production facilities that enable grading, cooking, packaging and freezing. Effective prawn farm management involves maintaining optimal water quality conditions, which becomes progressively complex as prawn biomass and the quantity of feed added to the system increases. As prawn biomass increases, so too does the biological oxygen demand required by the microbial population within the pond in breaking down organic materials. This requires increases in mechanical aeration and water exchanges (either fresh or recycled from a bioremediation pond). In most cases water salinity is not managed, except through seawater exchange, and will increase naturally with evaporation and decrease with rainfall and flooding. Strict regulation of the quality and volume of water that can be discharged means efficient use of water is standard industry practice. Most Australian prawn farms allocate up to 30% of their productive land for water treatment by pre-release containment in settlement systems. Barramundi The main factors that determine productivity of barramundi farms are the provision of optimal water temperature, dissolved oxygen, effective waste removal, expertise of farm staff, and the overall health of the stock. Barramundi are susceptible to a variety of bacterial, fungal and parasitic organisms, and are at highest risk of disease when exposed to suboptimal water quality conditions (e.g. low oxygen or temperature extremes). Due to the cost and infrastructure required, many producers elect to purchase barramundi fingerlings from independent hatcheries, moving fish straight into their nursery cycle. Regular size grading is essential during the nursery stage due to aggressive and cannibalistic behaviour. Size grading helps to prevent mortalities and damage from predation on smaller fish and assists with consistent growth. Ponds are typically stocked to a biomass of about 3 kg per 1000 L. Under optimal conditions barramundi can grow to over 1 kg in 12 months and to 3 kg within 2 years (Schipp et al., 2007). A pellet feed is produced by the two largest Australian aquafeed manufacturers (located in Brisbane and Hobart), providing a specific diet promoting efficient growth and feed conversion. The industry is heavily reliant on these mills to provide a regular supply of high-quality feed. Cost of feed transport would be a major cost to barramundi production in the Roper catchment. As a carnivorous species, high dietary protein levels, with fishmeal as a primary ingredient, is required for optimal growth. Barramundi typically require between 1.2 and 1.5 kg of pelleted feed for each kilogram of body weight produced. Warm water temperatures in northern Australia enable fish to be stocked in ponds year round. Depending on the intended market, harvested product is processed whole or as fillets and delivered fresh (refrigerated, ice slurry) or frozen. Smaller niche markets for live barramundi are available for Asian restaurants in some capital cities. Red claw Water temperature and feed availability are the variables that most affect crayfish growth. Red claw are a robust species but are most susceptible to disease (including viruses, fungi, protozoa, bacteria) when conditions in the production pond are suboptimal (Jones, 1995). In tropical regions, mature females can be egg-bearing year round. Red claw breed freely in production ponds, so complex hatchery technology (or buying juvenile stock) is not required. However, low fecundity, and the associated inability to source high numbers of quality selected broodstock, is an impediment to intensive expansion of the industry. Production ponds are earthen lined, rectangular in design and average 1 ha, and are sloping in depth from 1.2 m to 1.8 m. Sheeting is used on the pond edge to keep the red claw in the pond (migration tendency) and netting surrounds the pond to protect stock from predators (Jones et al., 2000). At the start of each crop, ponds are prepared (as for black tiger prawns above) then filled with fresh water and left for about 2 weeks prior to stocking. During this period, algal blooms in the water are encouraged through addition of organic fertiliser. Ponds are then stocked with about 250 females and 100 males that have reached sexual maturity. Natural mating results in the production of around 20,000 advanced juveniles. Red claw are omnivorous, foraging on natural productivity such as microbial biomass associated with decaying plants and animals. Early-stage crayfish rely almost solely on natural pond productivity (phytoplankton and zooplankton) for nutrition. As the crayfish progress through the juvenile stages, the greater part of the diet changes to organic particulates (detritus) on the bottom of the pond. Very small quantities of a commercial feed are also added on a daily basis to assist with the weaning process and provide an energy source for the pond bloom. The provision of adequate shelters (net bundles) is essential at this stage to improve survival (Jones, 2007). Approximately 4 months after stocking, the juveniles are harvested and graded by size and sex for stocking in production ponds. Juveniles are stocked in production ponds at 5 to 10 per square metre. Shelters are important during the grow-out stage, with 250/ha recommended. During the grow-out phase pellet feed becomes an important nutrition source, along with the natural productivity provided by the pond. Current commercial feeds are low cost and provide a nutrition source for natural pond productivity as much as the crayfish. Most Australian farmers use diets consisting of 25 to 30% protein. Effective farm management involves maintaining water quality conditions within ranges optimal for crayfish growth and survival as pond biomass increases. As with barramundi, management involves increasing aeration and water exchanges, while strictly managing effluent discharges. Red claw are harvested within 6 months of stocking to avoid reproduction in the production pond. At this stage the crayfish will range between 30 to 80 g. Stock are graded by size and sex into groups for market, breeding or further grow-out (Jones, 2007). Estimated water use An average crop of prawns farmed in intensive pond systems (8 t/ha over 150 days) is estimated to require 127 ML of marine water, which equates to 15.9 ML of marine water for each tonne of harvested product (Irvin et al., 2018). For pond culture of barramundi (30 t/ha over 2 years), 562 ML of marine water, or fresh water, is required per crop, equating to 18.7 ML of water for each tonne of harvested fish. For extensive red claw culture (3 t/ha over 300 days), 240 ML of fresh water is required per pond crop, equating to 16 ML of water for each harvested tonne of crayfish (Irvin et al., 2018). 4.5.4 Aquaculture land suitability The suitability of areas for aquaculture development were also assessed from the perspective of soil and land characteristics using the set of five land suitability classes in Table 4-1. The limitations considered included clay content, surface pH, soil thickness and rockiness, and mainly relate to geotechnical considerations (e.g. construction and stability of impoundments). Other limitations, including slope, and the likely presence of gilgai microrelief and acid-sulfate soils, infer more difficult, expensive and therefore less suitable development environments, and a greater degree of land preparation effort. More detail can be found in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). Suitability was assessed for lined and earthen-impounded ponds, with earthen ponds requiring soil properties that prevent pond leakage. Soil acidity (pH) was also considered for earthen ponds as some aquaculture species can be affected by unfavourable pH values exchanged into the water column (i.e. biological limitation). Representative aquaculture species were selected to represent environmental needs of marine species (represented by prawns), and freshwater species (red claw). Additionally, barramundi and other euryhaline species, which can tolerate a range of salinity conditions, may be suited to either marine or fresh water, depending on management choices. Except for marine species’ aquaculture, which for practical purposes are restricted by proximity to sea water, no consideration was given in the analysis to proximity to suitable water for fresh and euryhaline species aquaculture. It was not possible to include proximity to fresh water due to the large number of potential locations that water could be captured and stored within the catchment. Note also that the estimates for land suitability presented below represent the total areas of the catchment unconstrained by factors such as water availability, land tenure, environmental and other legislation and regulations, and a range of biophysical risks such as cyclones and flooding. These are addressed elsewhere by the Assessment. The land suitability maps are designed to be used predominantly at the regional scale. Planning at the enterprise scale would demand more localised assessment. The suitability for marine aquaculture has been restricted to a distance of 2 km to a marine water source. Aquaculture land suitability in lined ponds is shown in Figure 4-35a and shows suitability restricted to the areas under tidal influence and the river margins where cracking clay and seasonally or permanently wet soils dominate. These soils show the desired land surface characteristics such as no rockiness, suitable slope and sufficient soil thickness. However, these soils have the risk of acid-sulfate soils and so need to be managed accordingly. Approximately 4,500 ha (0.06% of the catchment) is highly suited (Class 1) to marine aquaculture in lined ponds, 7,700 ha (0.1%) as Class 2 (see Table 4-1), and 48,000 ha (0.62%) as Class 3. The land suitability patterns for marine species in earthen ponds (Figure 4-35b) closely mirror those of the marine in lined ponds, although areas are restricted to slowly permeable cracking clay soils. Approximately 43,000 ha (0.56% of the catchment) is mapped as suitability Class 3. Aquaculture marine map \\FS1-CBR.nexus.csiro.au\{lw-rowra}\work\3_Land_suitability\3_Roper\1_GIS\1_Map_docs\LL-R-526_Suit_aquaculture-Marine-LINED_aquaculture-Marine-EARTHEN_20211103.mxd For more information on this figure or equation please contact CSIRO on enquiries@csiro.au Figure 4-35 Land suitability in the Roper catchment for marine species aquaculture; (a) lined ponds and (b) earthen ponds Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). The aquaculture land suitabilities for freshwater species are shown in Figure 4-36. This shows that a significant proportion of the catchment is suitable for freshwater lined aquaculture (Figure 4-36a). Much of the Sturt Plateau is highly suitable (Class 1) or suitable with minor limitations (Class 2). This is because low slope gradient and low surface rockiness correspond in the red loam soils, meaning that the need for intensive land preparation would be limited. Other areas of Class 1 and Class 2 are found in the alluvial areas of the Wilton River Plateau, and alluvial areas throughout the Gulf Fall country on friable non-cracking clay or clay loam, cracking clay and red loam soils, which are generally Class 2. In the Gulf Fall country there are significant instances of Class 1 soils associated with the alluvial soils along the major watercourses where friable non- cracking clay or clay loam and cracking clay soils dominate, including in the headwaters of the Hodgson River. Near to the river mouth in the coastal plain, seasonally or permanently wet soils coincide with Class 1 suitability and in the same area, cracking clay and sand or loam over sodic clay subsoils contribute to Class 2 and Class 3 suitabilities. Approximately 2,476,000 ha (32% of the catchment) is highly suited (Class 1) for freshwater lined aquaculture, 1,695,500 ha (22%) is mapped as Class 2, and 162,500 ha is mapped as Class 3. In comparison, opportunities for freshwater species in earthen ponds in the Assessment area are fewer (Figure 4-36b). There are minor areas of Class 2 associated with cracking clay soils on the Sturt Plateau. The moderately to highly permeable soils are unsuited to earthen water impoundments. Areas of Class 3 suitability on slowly permeable clays are found on the Wilton River Plateau and in the Gulf Fall country. There are also significant areas of the coastal plain near the river mouth of Class 3 suitability on slowly permeable seasonally or permanently wet soils, sodic soils, sand or loam over sodic clay subsoils and cracking clay soils. These coastal plains have potential acid-sulfate soils that would require appropriate management. Freshwater species using earthen ponds shows a very small proportion of Class 2 suitability totalling 8,500 ha (0.11% of the catchment) and 537,000 ha (7%) as Class 3. Aquaculture freshwater map \\FS1-CBR.nexus.csiro.au\{lw-rowra}\work\3_Land_suitability\3_Roper\1_GIS\1_Map_docs\LL-R-527_Suit_aquaculture-Fresh-LINED_aquaculture-Fresh-EARTHEN_20211103.mxd For more information on this figure or equation please contact CSIRO on enquiries@csiro.au Figure 4-36 Land suitability in the Roper catchment for freshwater species aquaculture; (a) lined ponds and (b) earthen ponds Note that this land suitability map does not take into consideration flooding, risk of secondary salinisation or availability of water. The methods used to derive the reliability data in the inset map are outlined in the companion technical report on digital soil mapping and land suitability (Thomas et al., 2022). 4.5.5 Aquaculture viability This section provides a brief, generic analysis of what would be required for new aquaculture developments in the Roper catchment to be financially viable. First, indicative costs are provided for a range of four possible aquaculture enterprises that differ in species farmed, scale and intensity of production. The cost structure of the enterprises was based on established tools available from the Queensland Government for assessing the performance of existing or proposed aquaculture businesses (https://publications.qld.gov.au/dataset/agbiz-tools-fisheries- aquaculture). Based on the ranges of these indicative capital and operating costs, gross revenue targets are then calculated that a business would need to attain to be commercially viable. Enterprise-level costs for aquaculture development Costs of establishing and running a new aquaculture business are divided here into the initial capital costs of development and ongoing operating costs. The four enterprise types analysed were chosen to portray some of the variation in cost structures between potential development options, not as a like-for-like comparison between different types of aquaculture (Table 4-25). Capital costs include all land development costs, construction, and plant and equipment, accounted for in the year production commences. The types of capital development costs are largely similar across the aquaculture options with costs of constructing ponds and buildings dominating the total initial capital investment. Indicative costs were derived from Guy et al. (2014), and consultation with experts familiar with the different types of aquaculture, including updating to 2021 dollar values (Table 4-25). Operating costs cover both overheads, which do not change with output, and variable costs that increase as the yield of produce increases. Fixed overhead costs in aquaculture are a relatively small component of the total costs of production. Overheads consist of costs relating to licensing, approvals and other administration (Table 4-25). The remaining operating costs are variable (Table 4-25). Feed, labour and electricity typically dominate the variable costs. Aquaculture requires large volumes of feed inputs, and the efficiency with which this feed is converted to marketed produce is a key metric of business performance. Labour costs consist of salaries of permanent staff and casual staff who are employed to cover intensive harvesting and processing activities. Aerators require large amounts of energy, increasing as the biomass of produce in the ponds increase, which accounts for the large costs of electricity. Transport, although a smaller proportional cost, is important because this puts remote locations at a relative disadvantage to aquaculture businesses that are closer to feed suppliers and markets. In addition, transport costs may be higher at times if roads are cut (requiring much more expensive air freight or alternate, longer road routes) or if the closest markets become oversupplied. Packing is the smallest component of variable costs in the breakdown categories used here. Revenue for aquaculture produce typically ranges between $10 and $20 per kg (on a harvested mass basis), but prices vary depending on the quality and size classes of harvested animals and how they are processed (e.g. live, fresh, frozen or filleted) and farms are likely to deliver a mix of products targeted to the specifications of the markets they supply. Note that the mass of sold product may be substantially lower than the harvested product (e.g. fish fillets are about half the mass of harvested fish), so prices of sold product may not be directly comparable to the costs of production below (which are on a harvest mass basis) (Table 4-25). Table 4-25 Indicative capital and operating costs for a range of generic aquaculture development options Costs are provided both per ha of grow-out pond and per kg of harvested produce, although capital costs scale mostly with the area developed and operating costs scale mainly with yield at harvest. Capital costs have been converted to an equivalent annualised cost assuming a 10% discount rate and that a quarter of the developed infrastructure was for 15-year life span assets and the remainder for 40-year life span assets. Indicative breakdowns of cost components are provided on a proportional basis. PARAMETER UNITS PRAWN (EXTENSIVE) PRAWN (INTENSIVE) BARRAMUNDI RED CLAW (SMALL SCALE) Scale of development Grow-out pond area ha 20 100 30 4 Total farm area ha 25 150 100 10 Yield at harvest t/y 30 800 600 32 Yield at harvest per pond area t/ha/y 1.5 8.0 20.0 3.0 Capital costs of development (scale with area of grow-out ponds developed) Land and buildings % 56% 26% 23% 30% Vehicles % 5% 2% 2% 11% Pond-related assets % 27% 67% 70% 41% Other infrastructure and equipment % 11% 6% 5% 17% Total capital cost (year 0) $/ha 65,000 125,000 129,000 143,000 Labour costs % 47% 13% 12% 57% Electricity costs % 16% 24% 30% 9% Packing costs % 2% 4% 3% 2% Transport costs % 6% 16% 16% 11% Overhead costs (fixed) % 17% 8% 1% 12% Total annual operating costs $/kg 16.88 10.90 10.89 15.56 $/ha/y 25,321 87,227 217,854 46,683 Total costs of production Total annual cost $/kg 21.63 12.62 11.60 20.78 $/ha/y 32,400 100,900 232,000 62,400 PARAMETER UNITS PRAWN (EXTENSIVE) PRAWN (INTENSIVE) BARRAMUNDI RED CLAW (SMALL SCALE) Equivalent annualised cost $/kg 4.75 1.71 0.71 5.22 $/ha/y 7,122 13,695 14,134 15,668 Operating costs (vary with yield at harvest, except overheads) Nursery/juvenile costs % 12% 9% 7% 1% Feed costs % 0% 26% 30% 8% Commercial viability of new aquaculture developments Capital and operating costs differ between different types of aquaculture enterprises (Table 4-26), but these costs may differ even more between location (depending on case-specific factors such as remoteness, soil properties, distance to water source and type of power supply). Furthermore, there can be considerable uncertainty in some costs, and prices paid for produce can fluctuate substantially over time. Given this variation among possible aquaculture developments in the Roper catchment, a generic approach was taken to determine what would be required for new aquaculture enterprises to become commercially viable. The approach used here was to calculate the gross revenue that an enterprise would have to generate each year to achieve a target internal rate of return (IRR) for given operating costs and development costs (both expressed per hectare of grow-out ponds). Capital costs were converted to annualised equivalents on the assumption that developed assets equated to a mix of 25% 15-year assets and 75% assets with a 40-year life span (using a discount rate matching the target IRR). The target gross revenue is the sum of the annual operating costs and the equivalent annualised cost of the infrastructure development (Table 4-26). Table4-26Gross revenue targets required to achieve target internal rates of return(IRR)for aquaculturedevelopments with differentcombinations of capital costs and operating costs All values are expressed per hectare of grow-out ponds in the development.Gross revenue is the yield per ha of pond multiplied by the price received for produce (averaged across products and on aharvest mass basis). Capital costswere converted to an equivalent annualised cost assuminga quarter of the developed infrastructure wasfor15-yearlifespan assets and the remainderfor40-year lifespanassets. Targets would be higher after taking into account riskssuch as initial learning and market fluctuations. OPERATING($/ha/y) GROSS REVENUE REQUIRED TO ACHIEVE TARGET IRR($/ha/y) Capital costs of development ($/ha) 60,000 70,000 80,000 90,000 100,000 110,000 125,000 150,000 175,000 7% target IRR 20,000 25,022 25,859 26,696 27,533 28,371 29,208 30,463 32,556 34,648 50,000 55,022 55,859 56,696 57,533 58,371 59,208 60,463 62,556 64,648 100,000 105,022 105,859 106,696 107,533 108,371 109,208 110,463 112,556 114,648 150,000 155,022 155,859 156,696 157,533 158,371 159,208 160,463 162,556 164,648 100,000 105,022 105,859 106,696 107,533 108,371 109,208 110,463 112,556 114,648 200,000 205,022 205,859 206,696 207,533 208,371 209,208 210,463 212,556 214,648 250,000 255,022 255,859 256,696 257,533 258,371 259,208 260,463 262,556 264,648 10% target IRR 20,000 26,574 27,669 28,765 29,861 30,956 32,052 33,695 36,434 39,174 50,000 56,574 57,669 58,765 59,861 60,956 62,052 63,695 66,434 69,174 100,000 106,574 107,669 108,765 109,861 110,956 112,052 113,695 116,434 119,174 150,000 156,574 157,669 158,765 159,861 160,956 162,052 163,695 166,434 169,174 100,000 106,574 107,669 108,765 109,861 110,956 112,052 113,695 116,434 119,174 200,000 206,574 207,669 208,765 209,861 210,956 212,052 213,695 216,434 219,174 250,000 256,574 257,669 258,765 259,861 260,956 262,052 263,695 266,434 269,174 14% target IRR 20,000 28,776 30,238 31,701 33,163 34,626 36,089 38,283 41,939 45,596 50,000 58,776 60,238 61,701 63,163 64,626 66,089 68,283 71,939 75,596 100,000 108,776 110,238 111,701 113,163 114,626 116,089 118,283 121,939 125,596 150,000 158,776 160,238 161,701 163,163 164,626 166,089 168,283 171,939 175,596 100,000 108,776 110,238 111,701 113,163 114,626 116,089 118,283 121,939 125,596 200,000 208,776 210,238 211,701 213,163 214,626 216,089 218,283 221,939 225,596 250,000 258,776 260,238 261,701 263,163 264,626 266,089 268,283 271,939 275,596 In order for anenterprise to be commercially viable,the volume ofproduce grown each yearmultiplied bythe sales price of that produce would need to match or exceed the target valuesprovided above. For example, a proposed development with capital costsof $125,000/ha and operating costs of$200,000/ha/yearwould need to generate gross revenue of $213,695/ha/yearto achieve a target IRR of10% (Table4-26). If theenterprise received $12/kgfor produce(averaged acrossproduct types, on a harvest mass basis), then it would needto sustain average 268|Water resource assessment forthe Ropercatchment long-term yields of 18 t/ha (= $213,695/ha/y ÷ $12/kg × 1t/1000kg) from the first harvest. However, if prices were $20/kg, average long-term yields would require 11 t/ha (= 213,695/ha/y ÷ $20/kg × 1t/1000kg) for the same $125,000 capital costs per hectare, or only 8 t/ha prices if the capital costs were lowered to $100,000 per hectare. Target revenue would be higher after taking into account risks, such as learning and adapting to the particular challenges of a new location and periodic setbacks that could arise from disease, climate variability, changes in market conditions, or new legislation. Key messages From this analysis, a number of key points are apparent about achieving commercial viability in new aquaculture enterprises: • Operating costs are very high and the amount spent each year on inputs can exceed the upfront (year zero) capital cost of development (and the value of the farm assets). This means that the cost of development is a much smaller consideration for achieving profitability than ongoing operations and costs of inputs. • High operating costs also mean that substantial capital reserves are required, beyond the capital costs of development, as there will be large cash outflows for inputs in the start-up years before revenue from harvested product starts to be generated. This is particularly the case for larger size classes of product that require multi-year grow-out periods before harvest. Managing cashflows would therefore be an important consideration at establishment and as yields are subsequently scaled up. • Variable costs dominate the total costs of aquaculture production so most costs will increase as yield increases. This means that increases in production, by itself, would contribute little to achieving profitability in a new enterprise. What is much more important is increasing production efficiency, such as feed conversion rate or labour-efficient operations, so that inputs per unit of produce are reduced (and profit margins per kg are increased). • Small changes in quantities and prices of inputs and produce would have a relatively large impact on net profit margins. These values could differ substantially between different locations (e.g. remoteness, available markets, soils and climate), and depend on the experience of managers. Even small differences from the indicative values provided above could render an enterprise unprofitable. • Enterprise viability would therefore be very dependent on the specifics of each particular case and how the learning, scaling up, and cash flow were managed during the initial establishment years of the enterprise. It would be essential for any new aquaculture development in the Roper catchment to refine the production system and achieve the required levels of operational efficiency (input costs per kg of produce) using just a few ponds before scaling any enterprise. 4.6 References ABARES (2022) Agricultural commodities: September quarter 2022. Australian Bureau of Agricultural and Resource Economics and Sciences, Canberra. September CC BY 4.0. https://doi.org/10.25814/zs85-g927. ABARES (2023) Australian horticulture prices. Australian Bureau of Agricultural and Resource Economics and Sciences, Canberra. Viewed 10 March 2023, https://www.agriculture.gov.au/abares/data/weekly-commodity-price-update/australian- horticulture-prices#daff-page-main. Andrews K and Burgess J (2021) Soil and land assessment of the southern part of Flying Fox Station for irrigated agriculture. Part B: Digital soil mapping and crop specific land suitability. Department of Environment, Parks and Water Security, Northern Territory Government, Darwin. Ash AJ (2014) Factors driving the viability of major cropping investments in Northern Australia – a historical analysis. CSIRO, Australia. Ash A and Watson I (2018) Developing the north: learning from the past to guide future plans and policies. The Rangeland Journal 40, 301–314. Cowley T (2014) The pastoral industry survey – Katherine region. Northern Territory Government, Australia. Gentry J (2010) Mungbean management guide, 2nd edition. Department of Employment, Economic Development and Innovation, Queensland. Viewed 19 October 2017, https://era.daf.qld.gov.au/id/eprint/7070/1/mung-manual2010-LR.pdf. DSITI and DNRM (2015) Guidelines for agricultural land evaluation in Queensland. Queensland Government (Department of Science, Information Technology and Innovation and Department of Natural Resources and Mines), Brisbane. FAO (1976) A framework for land evaluation. Food and Agriculture Organization of the United Nations, Rome. FAO (1985) Guidelines: land evaluation for irrigated agriculture. Food and Agriculture Organization of the United Nations, Rome. Gleeson T, Martin P and Mifsud C (2012) Northern Australian beef industry: assessment of risks and opportunities. ABARES report to client prepared for the Northern Australia Ministerial Forum, Canberra. Guy JA, McIlgorm A and Waterman P (2014) Aquaculture in regional Australia: responding to trade externalities. A northern NSW case study. Journal of Economic & Social Policy 16(1), 115. Hughes J, Yang A, Marvanek S, Wang B, Petheram C and Philip S (2023) River model calibration and scenario analysis for the Roper catchment. A technical report from the CSIRO Roper River Water Resource Assessment for the National Water Grid. CSIRO, Australia. Irvin S, Coman G, Musson D and Doshi A (2018) Aquaculture viability. A technical report to the Australian Government from the CSIRO Northern Australia Water Resource Assessment, part of the National Water Infrastructure Development Fund: Water Resource Assessments. CSIRO, Australia. Jakku E, Thorburn PJ, Marshall NA, Dowd AM, Howden SM, Mendham E, Moon K and Brandon C (2016) Learning the hard way: a case study of an attempt at agricultural transformation in response to climate change. Climatic Change 137, 557–574. Jones CM (1995) Production of juvenile redclaw crayfish, Cherax quadricarinatus (von Martens) (Decapoda, Parastacidae) III. Managed pond production trials. Aquaculture 138(1), 247–255. DOI: https://doi.org/10.1016/0044-8486(95)00067-4. Jones C (2007) Redclaw package 2007. Introduction to redclaw aquaculture. Queensland Department of Primary Industries and Fisheries, Brisbane. Jones C, Grady J-A and Queensland Department of Primary Industries (2000) Redclaw from harvest to market: a manual of handling procedures. Queensland Department of Primary Industries, Brisbane. Jones C, Mcphee C and Ruscoe I (1998) Breeding redclaw: management and selection of broodstock. QI98016. Queensland Department of Primary Industries, Brisbane. Mann D (2012) Impact of aerator biofouling on farm management, production costs and aerator performance. Mid project report to farmers. Seafood CRC Project. Masser M and Rouse B (1997) Australian red claw crayfish. The Alabama Cooperative Extension Service, USA. McKellar L, Bark RH and Watson I (2015) Agricultural transition and land-use change: considerations in the development of irrigated enterprises in the rangelands of northern Australia. The Rangeland Journal 37, 445–457. McLean I and Holmes P (2015) Improving the performance of northern beef enterprises, 2nd edition. Meat and Livestock Australia. Moore G, Revell C, Schelfhout C, Ham C and Crouch S (2021) Mosaic agriculture. A guide to irrigated crop and forage production in northern WA. Bulletin 4915. Western Australia Department of Regional Industries and Regional Development, Perth. Paterson B and Miller S (2013) Energy use in shrimp farming, study in Australia keys on aeration and pumping demands. Global Aquaculture Advocate. QDPIF (2006) Australian prawn farming manual: health management for profit. Queensland Department of Primary Industries and Fisheries, Brisbane. Sangha KK, Ahammad R, Mazahar MS, Hall M, Owens G, Kruss L, Verrall G, Moro J and Dickinson G (2022) An integrated assessment of the horticulture sector in northern Australia to inform future development. Sustainability (Switzerland) 14(18), 1–18. DOI:10.3390/su141811647. Schipp G, Humphrey JD, Bosmans J and Northern Territory Department of Primary Industry, Fisheries and Mines (2007) Northern Territory barramundi farming handbook. Northern Territory Department of Primary Industry, Fisheries and Mines, Darwin. Stokes C, Jarvis D, Webster A, Watson I, Jalilov S, Oliver Y, Peake A, Peachey A, Yeates S, Bruce C, Philip S, Prestwidge D, Liedloff A, Poulton P, Price B and McFallan S (2023) Financial and socio-economic viability of irrigated agricultural development in the Roper catchment. A technical report from the CSIRO Roper River Water Resource Assessment for the National Water Grid. CSIRO, Australia. Thomas M, Gregory L, Harms B, Hill JV, Holmes K, Morrison D, Philip S, Searle R, Smolinski H, Van Gool D, Watson I, Wilson PL and Wilson PR (2018) Land Suitability Analysis A technical report from the CSIRO Northern Australia Water Resource Assessment to the Government of Australia. CSIRO, Canberra. Thomas M, Philip S, Stockman U, Wilson PR, Searle, R, Hill J, Bui E, Gregory, L, Watson, I, Wilson PL and Gallant G (2022) Soils and land suitability for the Roper catchment, Northern Territory. A technical report from the CSIRO Roper River Water Resource Assessment for the National Water Grid. CSIRO, Australia. Watson I, Austin J and Ibrahimi T (2021) Chapter 8 Other potential users of water. In: Petheram C, Read A, Hughes J, Marvanek S, Stokes C, Kim S, Philip S, Peake A, Podger G, Devlin K, Hayward J, Bartley R, Vanderbyl T, Wilson P, Pena Arancibia J, Stratford D, Watson I, Austin J, Yang A, Barber M, Ibrahimi T, Rogers L, Kuhnert P, Wang B, Potter N, Baynes F, Ng S, Cousins A, Jarvis D and Chilcott C (2021) An assessment of contemporary variations of the Bradfield Scheme. A technical report to the National Water Grid Authority from the Bradfield Scheme Assessment. CSIRO, Australia. Yeates SJ (2001) Cotton research and development issues in northern Australia: a review and scoping study. Australian Cotton Cooperative Research Centre, Darwin. Yeates SJ, Strickland GR and Grundy PR (2013) Can sustainable cotton production systems be developed for tropical northern Australia? Crop and Pasture Science 64, 1127–1140. Yeates SJ and Poulton PL (2019) Determining dryland cotton yield potential in the NT: Preliminary climate assessment and yield simulation. Report to NT Farmers, Queensland Cotton and the Cotton Research and Development Corporation. CSIRO, Canberra.