The process of selecting new crops in which to invest research and development effort and funds has, in the past, been largely subjective or based on a limited number of criteria.
Improved screening and selection procedures will assist in choosing new crops worthy of future development by industry and will reduce the number of expensively funded failures.
The RIRDC/GRDC funded project, UQ33-A, at the University of Queensland Gatton College seeks to develop a comprehensive and objective procedure for selecting new crops suitable for northern Australia. In this paper, we will review some recent developments in new crop selection and present the approaches that are being evaluated. We will also report progress on constructing information systems and a selection procedure and their ultimate extension into a decision support system.
A crop is a group of agricultural or horticultural plants cultivated for a purpose, usually with some form of harvest in mind. The most important crops used by man are listed in Table 1, ranked in terms of production. Each of these crops has a combined annual value of more than $US1 billion (Harlan, 1975).
The most important tree crops are highlighted in Table 1 and include citrus (ranked eleventh in production), banana (fifteenth), palm oil (twentieth), coffee (twenty fourth), rubber (twenty sixth) and cocoa (twenty seventh).
It is surprising that man has domesticated and developed so few crops for extensive use. There are at least 3000 potential food plant species available for development and as many as 20,000 plants available for domestication (Jain, 1983).
New crops include those plants which have not yet been domesticated, new species, plants adapted to new climates, cropping systems or areas, and plants giving rise to new products (Figure 1).
Figure 1. The production of a new product from a new crop is the ultimate challenge: the production of new products from established crops or established products from new crops is the more likely scenario in Australia.
| Cropping system (Production) |
| ||
| |||
Table 1. The 28 most important crops in the world, ranked in terms of estimated production (Harlan, 1975)
| Crop | Production (M metric tons) | Crop | Production (M metric tons) | ||
| 1 | Wheat | 343 | 15 | Bananas | 28 |
| 2 | Rice | 308 | 16 | Tomatoes | 28 |
| 3 | Maize | 308 | 17 | Millets | 22 |
| 4 | Potatoes | 306 | 18 | Cottonseed | 22 |
| 5 | Barley | 152 | 19 | Sesame | 21 |
| 6 | Manioc (cassava) | 92 | 20 | Palm oil | 20 |
| 7 | Oats | 54 | 21 | Peanuts | 18 |
| 8 | Sorghum | 49 | 22 | Sweet potatoes & yam | 15 |
| 9 | Soybean | 49 | 23 | Cotton (fibre) | 11 |
| 10 | Cane sugar | 41 | 24 | Coffee | 4.9 |
| 11 | Citrus | 37 | 25 | Tobacco | 4.5 |
| 12 | Beet sugar | 31 | 26 | Rubber | 3.5 |
| 13 | Beans,peas,chickpeas | 31 | 27 | Cocoa | 1.5 |
| 14 | Rye | 31 | 28 | Tea | 1.3 |
The Australian Institute of Agricultural Science National Conference in 1993, entitled 'Making It Happen' was told, by representatives from their Queensland Zone, that 'there has been a preference in Australia to focus on established crops. There needs to be much greater emphasis on new crops and new products, which will help to reduce the impact of global crop wars...' (AIAS, 1993).
Fletcher (1993) produced a listing of over 4500 potential new crops for Australia, from descriptions of plant species which have been found to be useful somewhere in the world. Among this listing, there were over 1600 tree crops, including 1451 fruit crops, 160 nut crops and 44 windbreak crops (Table 2). The fruit and nut crops from this listing are included in Appendix 1.
This listing demonstrated the scale of the problem of selecting potential new crops, such as new tree and nut crops, in which to invest research and development effort. The aim of this paper is to address the most appropriate method(s) for tackling this problem.
Table 2. Summary of the different crop types included in the Listing of Potential New Crops (Fletcher, 1993)
| Number | |||||
| 1 | Beverage | 15 | Oil | ||
| 2 | Cereal | 16 | Pesticide | ||
| 3 | Drug | 17 | Pseudocereal | ||
| 4 | Dye | 18 | Resin | ||
| 5 | Elastomer | 19 | Root | ||
| 6 | Forage grass | 20 | Spice | ||
| 7 | Fibre | 21 | Starch | ||
| 8 | Forage legume | 22 | Soil stabiliser | ||
| 9 | Fruit | 23 | Sugar | ||
| 10 | Green material | 24 | Sweetener | ||
| 11 | Gum | 25 | Tannin | ||
| 12 | Legume | 26 | Vegetable | ||
| 13 | Medicinal | 27 | Wax | ||
| 14 | Nut | 28 | Windbreak | ||
| TOTAL(some crops were classified as having multiple uses) 5081 | |||||
Recent history of new crop development
Wood, Chudleigh and Bond (1994) investigated the factors determining success or failure for 35 new Australian crop industries established since 1950 (Table 3). Among their general conclusions, they found that:
Table 3. New crop industries established in Australia since 1950, grouped according to their estimated annual value, in $AUD million (Wood, Chudleigh and Bond, 1994)
| High | $M | Medium | $M | Low | $M | Zero | $M | |
| Cotton | 806 | Sunflower | 48 | Oilseed poppy | 10.0 | Rice (WA) | 0 | |
| Lupin | 160 | Broccoli | 37 | Pyrethrum | 8.0 | Arrowroot | 0 | |
| Mushroom | 99 | Soybean | 35 | Safflower | 6.4 | Cassava | 0 | |
| Melon | 35 | Rice (Qld) | 5.6 | |||||
| Canola | 27 | Blueberry | 4.5 | |||||
| Triticale | 27 | Ginger | 4.5 | |||||
| Avocado | 23 | Mungbean | 4.4 | |||||
| Macadamia | 22 | Navy bean | 4.0 | |||||
| Chickpea | 21 | Duboisia | 3.5 | |||||
| Mango | 16 | Pecan | 2.2 | |||||
| Kiwifruit | 15 | Persimmon | 2.0 | |||||
| Almond | 15 | Eucalyptus oil | 2.0 | |||||
| Garlic | 1.5 | |||||||
| Tea | 1.3 | |||||||
| Alfalfa sprout | 1.0 | |||||||
| Aloe vera | 0.4 | |||||||
| Jojoba | 0.1 | |||||||
Figure 2. Representation of the change in gross value of a new crop industry (derived from information included in Wood, Chudleigh and Bond, 1994 and other sources).
**scan**
The factors having the greatest impact on the development of new crop industries in Australia were the availability of research and development funds in Australia, and the marketability of the crop product, especially in terms of satisfactory quality (Table 4).
The impact of government support in new crop industries was significant, through the provision of research and development funding, especially in the area of plant improvement.
The growth characteristics of many of the new crop industries, in terms of gross value, could be approximated by the relationship in Figure 2. Growth was rapid at first (stage A), after which a plateau was reached (B). After this time, the performance of the new crop industry was likely to be determined by the research and development conducted (Table 4); the outcome could have been positive (D) or negative (C). At stage E, the crop industry could probably be referred to as being established.
The growth characteristics of the macadamia, mango and almond industries indicated that these industries had progressed beyond the plateau stage (to stage D), whereas the avocado industry appeared to be at the plateau stage (Wood, Chudleigh and Bond, 1994).
Table 4. Summary of the more significant success factors identified among the 35 new crop industries developed in Australia since 1950 (Wood, Chudleigh and Bond, 1994).
| Success factors: | |||
| Positive Production Factors: | |||
| Research and development conducted in Australia Plant improvement (20); Cultural practices (15); Pests and diseases (6); Mechanisation(5)] | |||
| Availability of overseas technology | |||
| Investment incentives | |||
| Dissatisfaction with alternatives | |||
| Negative Production Factors: | |||
| Pests and diseases (often solved) | |||
| Reduced productivity | |||
| Scarce labour | |||
| Marketing Factors: | |||
| Demand | (positive) | ||
| (negative) | |||
| Market development | (positive) | ||
| (negative) | |||
| Quality | (positive) | ||
| (negative) | |||
| Competition | (negative) | ||
| Processing Factors: | |||
| Facilities established | |||
| Overseas technology available | |||
| Government Factors (apart from research funding) | |||
| Regulation of this or some other industry | |||
| Assistance | |||
The current situation in new crop research and development in Australia
A survey was conducted among the 500 Australian readers of the Australian New Crops Newsletter in July 1994. Among the 400 respondents, 70 had an interest in new fruit crops, 56 were interested in new nut crops, and 45 in windbreak crops (Fletcher, 1995a).
Specific crops identified by these readers as the subjects of current or past investigations are listed in Table 5.
Table 5. Tree and nut crop species being researched and species indicated as having been the subject of past experiences by the readers of the Australian New Crops Newsletter (Fletcher, 1995a)
| Anacardium spp. (Cashew) | Macadamia spp. |
| Annona cherimola (cherimoya) | Mangifera indica (mango) |
| Athertonia diversifolia | Musa spp. (banana) |
| Averrhoa carambola (starfruit) | Myrciaria cauliflora (jaboticaba) |
| Carica spp. (Papaya) | Nephelium lappaceum (rambutan) |
| Ceratonia siliqua (carob) | Olea europaea (olive) |
| Cocos nucifera (coconut) | Prunus mume (Japanese apricot) |
| Diospyros spp. (Persimmon) | Psidium guajava (guava) |
| Euphoria longana (longan) | Santalum acuminatum (sweet quandong) |
| Ficus spp. (Fig) | Santalum spp. (sandalwood) |
| Flacourtia inermis | Sclerocarya caffra |
| Garcinia mangostana (mangosteen) | Syzygium spp. |
| Inocarpus tugiter | Terminalia karnbachii |
| Litchi chinensis (lychee) | |
| Albizia spp. | Indigofera spp. |
| Castanospermum australe (blackbean) | Luffa spp. |
| Eriobotrya japonica (loquat) | Pyrus spp. (nashi) |
| Flindersia grayliana (Queensland maple) | Ziziphus jujuba (common jujube) |
Techniques and approaches in new crop selection
The development of new crops does not only involve new species producing new products (Figure 1). Such a combination is the ultimate challenge. A new crop may have some advantage over established crops in producing a product already well established in the marketplace. Conversely, an old crop producing a new product or being introduced into a new geographical area can be considered as a new crop, having the difficulties associated with new crop development.
In developing an approach to selecting new crops for research and development, four decisions need to be made:
Each of these matters will be discussed briefly to indicate the reasoning behing the decisions which have been reached.
The total information approach is an attempt to identify all criteria likely to be relevant to the selection of new crops and attempts to provide information for each of the criteria. Table 6 describes a potential information structure that may support this process, with the most likely sources for the relevant information.
The disadvantages of the total information approach are that the numbers of relevant criteria will be substantial and reliable information for new crops is not often available. Reliability may have significant implications in a process such as new crop selection: small differences in the initial conditions of some variables may provoke large differences in the final outcomes (Hall, 1992; Sigmund, 1993). The advantage of the total information approach is that the generic set of criteria created requires that many of the potentially important factors will need to be considered before any research and development commences and are thus not likely to be overlooked.
The major factor approach has the advantage that it will be easier to assemble and carry out. Single factors can sometimes have a major influence on the success of new crops and may warrant analysis as a predictor of future potential among new crops. It is difficult, however, to make such an approach generic. Should the major factor change in its importance, the data and the problem will need to be reassessed.
As an example of the major factor approach, would monitoring of the shelves of US supermarkets give a sufficient indication, by itself, of the future trends in the marketability of new food innovations? Although such an approach may have some usefulness in certain situations, it does not appear to warrant general application.
Table 6. List of criteria for the selection of new crops and the likely sources of the relevant information for use in making these decisions.
| Information category (criterion) | Information source / format |
| 1 (Ethno)botanical/geographical | |
| 1.1 Image of plant or parts | |
| 1.2.1 Family | SEPASAL,Index Kewensis,NICU,Grays |
| 1.2.2 Genus | SEPASAL,Index Kewensis,NICU,Grays |
| 1.2.3 Species | SEPASAL,Index Kewensis,NICU,Grays |
| 1.2.4 Synonyms | SEPASAL,Index Kewensis,NICU,Grays |
| 1.2.5 Common names | SEPASAL,Index Kewensis,NICU,Grays |
| 1.3 Life form, habit | SEPASAL, Literature |
| 1.4 Ontogeny | SEPASAL, NICU |
| 1.4.1 Physiology/genetics of ontogeny | Literature |
| 1.5 Geographical origin | SEPASAL, Literature |
| 1.5.1 Centres of origin | SEPASAL, Literature |
| 1.6 Current distribution | SEPASAL, Literature |
| 1.7 History of domestication | SEPASAL, TDWG |
| 1.8 Uses | SEPASAL |
| 1.8.1 Useful plant parts | SEPASAL |
| 1.8.2 Image of useful plant parts | |
| 2 Agronomic | |
| 2.1 Climate, | PLANTGRO, CLIMPROD, ECOCROP,CLIMEX |
| 2.1.1 Matched climate in Australia | PLANTGRO, CLIMEX |
| 2.2 Landscape linked to regions | SEPASAL |
| 2.2.1 Matched landscape in Australia | SEPASAL |
| 2.3 Soil types | FAO, SEPASAL |
| 2.3.1 Matched soil type in Australia | FAO |
| 2.4 Areas of cultivation (if any) | Literature |
| 2.4.1 World map | GIS |
| 2.4.2 World past areas of cultivation Literature, | GIS |
| 2.4.3 World current areas of cultivation Literature, | GIS |
| 2.4.4 World potential areas of cultivation | GIS |
| 2.4.5 Australian map | GIS |
| 2.4.6 Australian past areas of cultivation | GIS |
| 2.4.7 Australian current areas of cultivation | GIS |
| 2.4.8 Australian potential areas of cultivation | |
| 2.5 Breeding system | GRIN, Literature |
| 2.6 Genetic diversity available | |
| 2.6.1 World accessions held | IPGRI |
| 2.6.2 Australian accessions held | Genetic Resource Centres |
| 2.7 Breeding method to be used | |
| 2.7.1 Type of commercial cultivar to be used | |
| 2.7.2 Selection or crossing | |
| 2.7.3 Objectives comprising ideal cultivar | |
| 2.7.3.1 Critical objectives | |
| 3 Production | |
| 3.1 Quantity of information/support available Worldwide Literature | |
| 3.1.1 Related information available | |
| 3.1.2 Australia | Literature, experts |
| 3.1.2.1 Related information available | |
| 3.2 Propagation | Literature, Horticulture Section |
| 3.3 Planting | Literature, AgMech |
| 3.3.1 Preparation of planting site | |
| 3.4 Degree of mechanisation possible AgMech | |
| 3.4.1 Planting | |
| 3.4.2 Maintenance during growth | |
| 3.4.3 Harvesting | |
| 3.4.4 Post-harvest treatment | |
| 3.5 Current/potential yields | FAO, ABS, PLANTGRO |
| 3.5.1 Generic yield models available | e.g. CERES |
| 3.6 Harvesting | AgMech |
| 3.6.1 Specialisation of labour required FAO | |
| 3.6.2 Seasonality of harvesting | FAO |
| 3.7 Nutrition | Horticultural/Agronomy Sections |
| 3.8 Major pests/diseases | Plant Protection Section |
| 3.8.1 Availability of control measures | |
| 3.8.2 Comparative damage | |
| 3.9 Pesticide registrations | PESKEM |
| 3.10 Rotations available | Literature |
| 3.11 Landcare considerations | Literature, Land Resources Section |
| 4 Production economics | |
| 4.1 Competing alternative crops | |
| 4.2 Gross margins analysis | Greg Ferguson, QDPI |
| 4.3 Crop budget | QDPI software |
| 4.3.1 Inputs | |
| 4.3.1.1 Variable | |
| 4.3.1.2 Fixed | |
| 4.3.1.3 Random | |
| 4.3.2 Prices | |
| 4.3.3 Land/site preparation | |
| 4.3.4 Cultivation | |
| 4.3.5 Harvest rate | |
| 4.4 Tax incentives | ATO, DPIE |
| 5 Industry economics | |
| 5.1 Critical mass for infrastructure | |
| 5.2 Specialised processing | |
| 6 Domestic production | |
| 6.1 Areas of production | GIS, ABS |
| 6.1.1 Dispersal/concentration | |
| 6.2 Volumes | GIS, ABS |
| 6.2.1 By areas | GIS |
| 6.2.2 By years | |
| 6.3 Characteristics of producers | |
| 7 World trade | |
| 7.1 FAO statistics | FAO, USDA |
| 7.2 Trade volume | FAO, USDA |
| 7.3 Price | FAO, USDA |
| 7.4 Stability | Derived |
| 7.5 Dominant players/markets | FAO, USDA |
| 7.5.1 Timing of production | |
| 8 Australian trade | |
| 8.1 Trade statistics | ABARE, ABS, AQIS |
| 8.1.1 Customs import/export codes | SITC |
| 8.2 Trade volume | ABARE, ABS, AQIS |
| 8.3 Price | ABARE, ABS, AQIS |
| 8.4 Stability | Derived |
| 8.5 Tariffs and import duties | ACS |
| 8.6 Quarantine considerations | AQIS |
| 8.7 Dominant players/trading system Derived | |
| 9 Marketing | |
| 9.1 Product category | ACS, ABS |
| 9.2 Potential substitutes | ABS, experts |
| 9.3 Market characteristics | Chai McConnell |
| 9.4 Quality considerations | AQIS, wholesalers, manufacturers |
| 9.5 Public awareness | Experts |
| 9.6 Marketing bodies | Chai McConnell |
| 9.7 Registration/recognition barriers | AQIS, Customs, AMA |
| 10 Handling | |
| 10.1 Technology transfer | AgMech, FST |
| 10.2 Transport system | AgMech, FST |
| 10.3 Packaging standards | FST |
| 10.3.1 Image of packaging standards | |
| 10.4 Shelf life | Horticulture Section |
| 10.4.1 Wholesale, retail | Commercial resellers |
| 10.4.2 Maturity indices, profiling | Horticulture Section |
| 10.5 Postharvest technology | Horticulture Section |
| 11 Processing | |
| 11.1 Potential products/uses | Literature |
| 11.1.1 Image of products | |
| 11.2 Product standards | Relevant Industry bodies |
| 11.3 Specific product characters | FST |
| 11.4 Processing technology | FST |
| 11.4.1 Image of machinery | |
| 11.5 Processing infrastructure | FST |
| 11.6 Health/hygiene considerations | FST |
| 12 Funding of domestic research | |
| 12.1 Current research | Survey via Newsletter |
| 12.2 Potential funding bodies | Experts, literature |
| 13 General | |
| 13.1 Industry champion | |
| 13.2 Seminal references | CDROM abstracts |
| 13.3 World authorities/references | FAO, Literature |
| 13.4 Australian authorities/references | DPIE, CSIRO, survey |
| 13.5 Potential interest groups | Survey |
| 14 Glossary | ANBG |
Abbreviations:
ABARE: Australian Bureau of Agricultural Resources and Economics
ABS: Australian Bureau of Statistics
ACS: Australian Customs Service
AgMech: Agricultural Mechanisation Centre, Department of Plant Production, The University of Queensland Gatton College, Gatton.
ANBG: Australian National Botanical Garden
AQIS: Australian Quarantine Inspection Service
AMA: Australian Medical Association
ATO: Australian Taxation Office
CDROM: Compact Disc Read-only Memory
CERES: Software package
CLIMEX: Software package
CLIMPROD: Software package
DPIE: Commonwealth Department of Primary Industries and Energy
CSIRO: Commonwealth Scientific and Industrial Research Organisation
ECOCROP: Database
FAO: Food and Agricultural Organisation of the United Nations
FST: Food science and Technology Department, The University of Queensland Gatton College
GIS: Geographic Information System
Grays: Database
GRIN: Genetic Resource Information Network, National Resources Program, United States Department of Agriculture
IPGRI: International Plant Genetics Resources Institute
NICU: Names in Current Use
PESKEM: Database
QDPI: Queensland Department of Primary Industries
SEPASAL: Survey of Economic Plants of Arid and Semi-arid Lands
SITC: System of International Trade Classifications
TDWG: Taxonomic Databases Working Group
USDA: United States Department of Agriculture
Figure 3. Representation of the need for more attention in the choosing of which potential new crops are investigated.
Identifying a plant as useful indicates that its product has been defined. The Listing of Potential New Crops for Australia (Fletcher, 1993), adopted this approach.
The advantage of the usefulness approach was demonstrated in Figure 1. To consider species that are useful, avoids the problems of establishing a new crop product from a new crop. A defacto product has already been identified in these cases.
Many systems to comprehensively evaluate 'promising' new crops already exist (step C in Figure 3). Such an approach has been used by such organisations as the USDA and the New Zealand and Tasmanian Departments of Primary Industry and Fisheries.
The nature of the process in step B that identified the promising crops to be investigated is not clear. The current project aims to carry out both steps B and C systematically and in one operation.
Due to limitations on research and development resources, selecting one new crop would be advantageous, yet risky. The new crop that is chosen will have to be commercially successful to warrant its choice.
From an economic standpoint, a basket of new crops has many advantages.
An economic rationalist would select new crops worthy of financial support by benefit/cost analysis. Such analyses, as conducted on established crops, would need to be modified before being used for new crops.
An investment that produces a marginal improvement in an established industry, in the short term, will usually be more attractive than the same investment in a new crop industry. This would occur because new crop industries have extra costs associated with establishment of the industry. With a basket of new crops, these costs can be shared. As well, the benefits from new crops will be longer in coming.
Most of the benefits from established crop research flow to the consumer. Increased productivity results in lower product prices in the marketplace and for exported products, these benefits are shared globally.
For new crops, there is more opportunity for the producer, as well as the consumer, to benefit from research and development. With a basket of new crops, these benefits would be spread more widely. There would also be greater potential to retain, within the national economy, a larger share of the research benefits from new crops research.
There is also significant risk associated with funding of any new crop research. This risk is likely to be higher, and probably more difficult to estimate, than that associated with established crop research. The risk with new crops has also been publicised through several ill-advised ventures that have given rise to new crop failures in the past.
The higher risk involved with new crop industries is quantified in benefit/cost analyses through discounts applied to benefits. The selection of a basket of new crops would avoid the discounts associated with these higher risks.
Portfolio selection in stock market trading has techniques available to measure and select groups of shares which provide a particular expected return, with a certain level of associated risk. Such methods can be used to select portfolios of new crops on which research and development funds can be spent, with a predetermined level of risk. The method of genetic algorithms has potential for this purpose.
The choice of a selection procedure
Several methods have been investigated in a preliminary fashion to determine which may be most useful in comparing potential new crops and selecting the most suitable for research and development. Currently, genetic algorithms has the most promise and will be investigated further. An example of the use of genetic algorithms to choose new crops worthy of research and development is included below.
Development of information systems
There is a lack of high quality, structured and reliable information available on new crops. This is especially true of commercially relevant information and is a factor preventing the adoption of new cropping options and hence the diversification of Australian agriculture.
A survey of fifteen extension personnel responsible for advising primary producers on new crop choices (Fletcher, 1995b) has identified that the unavailability of useful information is their major difficulty in providing interested producers with advice on alternative cropping options (Table 7).
Industry is understandably reluctant to commit research and development funds to any project for which the outcomes are uncertain. Thus, in any investigation of new crops for commercial adoption, lack of information becomes a de-facto selection criteria itself, prompting the discarding of those potential crops about which little information is readily available. This situation is not desirable.
As long ago as 1957, R.D.Lewis, as the chairman of the US New and Special Crops Task Group, in reporting to the President's Appointed Bipartisan Commission of Increased Industrial Use of Agricultural Products, recognised this. His task group reported that collating and structuring the existing literature was the first requirement for expanded new crops research and was the foundation for any intelligent program of plant evaluation. He advocated the total information approach, but, in the intervening forty years, this has not been undertaken (Jolliff, 1995).
There is, therefore, a need for some form of information system on new crops.
There are two distinct and possibly conflicting requirements for information systems on new crops.
Table 7. The sources for the information on new crops supplied by extension personnel to primary producers (Fletcher, 1995b).
| Personally-produced information: |
| Verbal discussion based on previous experience |
| Personally-conducted trial work (more limited these days as time available becomes short) |
| Personal experience generally with crops |
| Information generated from within the industry-based group or project team |
| Personally-conducted market research |
| Use of files collected over time |
| Information sourced from an expert: |
| Referral to an industry expert on that crop (if available) |
| Discussions with other producers who have grown the new crop |
| Contacts established with leading farmers with experience from other areas |
| Personal and direct contact with importers and local buyers |
| Personal and direct contact with overseas grower associations and networks |
| Contracts for market research |
| Contracts for whole farm economic analysis |
| Information sourced from those with more generic knowledge: |
| Contacts established with those in related industries, innovative producers, others with specific information, researchers from various agencies |
| Discussion with relevant officers of state departments of agriculture, company representatives, traditional primary producers |
| Word of mouth |
| Information from publications: |
| Specific interest magazines |
| Relevant state government department of agriculture published research |
| Locally produced information sheets |
| State department of agriculture information centres |
| Articles from the local press and rural newspapers |
| Local publications; Interstate publications; Books |
| Scientific literature; journals |
| General magazines (which can be incomplete in their information) |
Construction of a Decision Support System
A decision support system is any resource that we use to assist in making better informed decisions. Computer-based decision support systems provide a means by which we may process large quantities of information and make comparisons between that information, based upon pre-defined rules.
Human experts typically work on a select subset of any given information. The particular subset depends on their area of expertise. As the level of expertise increases, the ability to articulate knowledge and thought processes (that is, to elucidate evaluation techniques) decreases. In other words, several experts can often provide conflicting opinions on a subject and yet be unable to provide justification for their decision beyond invoking "experience".
The main advantage of applying computer technologies to new crop selection is that data may be treated objectively and comprehensively, even if they may vary in quality or completeness.
Improved methodologies and skilful identification of the information will decrease the complexity of the task. Good design is needed to ensure that all vital information is included.
An example of the use of genetic algorithms
The software 'Evolver' (Axcelis Inc., 4668 Eastern Avenue North, Seattle, Washington, USA 98103-6932) has been used for this demonstration. A model comprising a number of useful plant species, a number of criteria and ratings was firstly set up in an Excel database (Table 8).
An estimate for the cost of fixing all limitations for each plant species was derived by multiplying the ratings by the cost and summing (Table 9). The basket of the five most worthy plant species for research and development included species 15, 7, 12, 18 and 4 (Table 10). A range for each cost was then used with the 'Evolver'software to find a basket of five species with the lowest total cost and the basket included the same five species (Table 10).
There are many possible developments for improving this model, including such elements as:
Table 8. The model for choosing new crops worthy of research and development: twenty useful plant species, fifteen criteria and ratings for the limitations of each new crop in terms of each criterion (this is an example including random numbers for the ratings).
Table 9. Derivation of the total cost (in arbitrary units of value) of rectifying the limitations for each useful plant species in the process of choosing new crops worthy of research and development.
Table 10. The total costs of fixing all limitations for each of twenty useful plant species and the results of using genetic algorithms to choose the five species producing the minimum cost of fixing all limitations, once a range of costs has been estimated for each criterion.
Conclusions
From the considerations of the approaches to the selection of potential new crops it is concluded that:
The technique of genetic algorithms offers an opportunity to select such new crops as soon as the information is compiled.
There are great opportunities looming in the field of new crop development. However we need to heed the lessons from the past and cooperate throughout the industry to ensure that those crop industries which are targeted for research and development are the ones most likely to succeed with the limited resources available.
AIAS (1993). Making it happen. A national strategy for professional support for Australian agriculture to the year 2020. Proceedings of the Australian Institute of Agricultural Science National Conference pp. 14-20. Australian Institute of Agricultural Science Canberra ACT.
Fletcher R.J. (1993). Listing of Potential New Crops. Department of Plant Production The University of Queensland Gatton College Gatton Queensland.
Fletcher R.J. (1995a). Survey of the Australian readership of the Australian New Crops Newsletter. Summary of responses: January 1995. The Australian New Crops Newsletter 3 2-5.
Fletcher R.J. (1995b). Survey of Extension Activity in New Crops in Australia. The Australian New Crops Newsletter 4 12.
Hall N. (ed.) (1992). The New Scientist guide to chaos. Penguin Books London England.
Harlan J.R. (1975). Crops & man. American Society of Agronomy Madison Wisconsin U.S.A.
Jain S.K. (1983). Domestication and breeding of new crop plants. In Crop breeding (ed. D.R.Wood K.M.Rawal and M.N.Wood) pp. 1-20. American Society of Agronomy and Crop Science Society of America Madison Wisconsin U.S.A.
Jolliff G. (1995). Personal communication.
Lewis R.D. (1957). Report of President's Appointed Bipartisan Commission on Increased Industrial Use of Agricultural Products by the Task Group on New and Special Crops. Texas Agricultural Experiment Station College Station Texas.
Sigmund K. (1993). Games of life. Explorations in ecology evolution and behaviour. Oxford University Press Oxford England.
Wood I. Chudleigh P. and Bond K. (1994). Developing new agricultural industries. Lessons from the past. RIRDC Research Paper series 94/1. Rural Industries Research and Development Corporation Barton ACT.