University of Western Sydney

Project commencement: The project commenced with the first face-to-face project workshop held in Port Moresby on 30 June 2006. Discussions were held between the PNG and Australian collaborators and personnel from ACIAR on which districts (and which sites within these districts) in PNG we would use to carry out experiments, which crops to study and which PNG agencies and personnel were available to participate in this project.
Project workshops and training: In addition to the PNG/Australia workshop mentioned above there have been several more conducted around the four districts in PNG at which various aspects of the project will be studied. NARI staff from Port Moresby travelled to the experimental sites/agencies in the four districts to demonstrate how to collect fruit, rear fruit flies and identify fruit flies for the project.
Experimental design: On a second trip to PNG Andrew Jessup and Andrew Beattie travelled to the four districts with district collaborators to view potential farms and orchards on which to carry out experiments. Experimental designs were worked out to suit each site/crop/district/fruit fly species combination with a view to each design being as similar to each other between districts as possible to allow for direct comparisons.
Farmer survey: Anna Kawi set up a Farmer Questionnaire on the knowledge of fruit flies and their impact on horticultural production. The survey has gone out to officers in all districts. The questionnaire is designed to be carried out at the beginning of the project (already achieved), during the middle and towards the end to gauge changes in grower knowledge/perception of fruit flies.
Fruit survey: Staff in each district are collecting fruit from identified farms within their district and rearing out flies which are counted and identified. We will soon have comprehensive lists of fly species, crops attacked, the severity of infestation, time of year crops are attacked as well as the economic impact of infestations on the grower.
Trap survey: Staff from each district have set up and placed out fruit fly traps capable of trapping the majority of pest fruit fly species. Flies are periodically collected, counted and identified to give a list of fruit fly species, numbers trapped, time of year and district.
PNG experiments: Experiments have commenced on bagging (a physical fruit fly exclusion method) individual guava fruit on tree, netting entire capsicum plants (another physical fruit fly barrier) and fruit fly baits in comparison with insecticides. More experiments are planned once experimental material is sourced in Australia and shipped to PNG.
Aust experiments: Experiments have commenced in Australia to design a physical system for the disposal of fallen/damaged/infested fruit in-field whilst allowing the survival and dispersal of fruit fly parasitoids (a form of biological control of fruit flies). Also tests on new female fruit fly lures and traps have commenced.
Analysis: Data from previous ACIAR projects on PNG are being analysed by John Allwright Fellows in the Queensland University of Technology.

An impact assessment survey has been conducted in eight locations over three provinces in Papua New Guinea (PNG). The survey has shown that farmers recognise fruit flies as major pests, but don't generally use control strategies because they don't fully understand the ecology of the insect. These results indicate that more information should be provided to farmers to help them understand how to better manage fruit flies. It is planned that this survey will be conducted at least once more during the course of the project to gauge the impact that the project has on farmer's awareness and management strategies for fruit flies.

A bagging experiment on guava fruit conducted earlier in the project demonstrated that bags made from old newspapers provide complete protection from fruit fly infestation when they are applied to fruit prior to maturity. Replicated trials with capsicum (3 locations) and banana (1 location) are currently being conducted in PNG. These trials compare the efficacy of different management methods, including protein bait spraying and male annihilation technique (MAT). The results from these trials conducted in PNG will form the basis of fruit fly management strategies to be developed for each of the different regions being investigated in the project. The results from these trials will also be used to educate farmers on the management options that may be available to them and to produce extension materials. In Australia, work has continued on the development of improved female fruit fly lures and traps.

The replicated trials are visited weekly by technical staff and this provides an opportunity for ongoing communication with local farmers. Farmer training has been conducted by
NARI and FDPA staff in locations where the experiments, fruit fly trapping and fruit sampling have been taking place. The training covers topics including the ecology of fruit flies, different fly species, and different management options. It is planned that demonstration trials will be set up later in 2008 to help continue this training. Written extension materials will also be prepared. Students at PAU and technical staff from the collaborating agencies have also been involved in a number of training workshops.
A draft manuscript for submission to a peer-reviewed journal has been prepared that reports the relationship between the distribution and abundance of populations of five economically important fruit flies (data collected in previous project CS2/96/225) and environmental factors, such as altitude and annual rainfall. It is anticipated that this paper will be published in the near future.

Research trials were conducted in four provinces in Papua New Guinea (PNG) to investigate the efficacy of various fruit fly management strategies. These have been completed and the results are now being collated for analysis. The results of these trials will provide useful information regarding the effectiveness of management strategies such as protein baiting, male annihilation technique, fruit bagging and local pesticide use in each location and crop. An economic analysis based on a single farm case study should help to determine if the level of control achieved from the management strategies provides benefits that outweigh the cost of the technologies themselves. If this is found to be the case, a recommendation will be made to further develop the supply-chain for these technologies, which are currently not commercially available in PNG. These research trials have also provided an opportunity to train regional research and extension staff. The knowledge that they have gained through this project will enable them to train new staff joining their organisations, as well as lead farmers in their respective communities.
Extension activities including the development of printed materials, radio and farmer talks have been ongoing throughout the project. The aim of these activities has been to create awareness about the project and to provide basic information about what fruit flies are, how they affect crops and what can be done to reduce their damage. These activities will be continued. Following on from the results achieved in the experimental trials, it is anticipated that an extension strategy will be implemented to allow some of the producers with larger farms to trial the technologies for themselves. If these producers are satisfied with the results of their own trials it may help to generate a demand for the technologies, which will aid in developing a supply-chain for the products. This exercise will also provide a participatory learning environment for local extension staff.
Research trials in Australia investigating mineral oil products and changes to existing protein baiting programmes are ongoing. The results from these trials will be used to help make regulatory decisions regarding fruit fly eradication methods and to assist horticultural industries make informed decisions regarding the various fruit fly control options available to them.
With the finalisation of all research trials and the development of extension materials and strategies, this project will be completed during the next nine months. It is anticipated that the outputs from this project will form the basis for the development of a supply-chain for the technologies trialled, which are currently not available in PNG.

Vietnam is one of many countries in South East Asia which experience serious preharvest fruit and vegetable losses to fruit flies ranging from 70 - 100%. As a result, fruit flies are seen as a major contributor to the ongoing problems of hunger, poor nutrition and poverty, especially in the rural communities, which comprise over 75% of Vietnam's population. The purpose of the project is to develop and introduce simple yet practical, in-field solutions to the fruit fly problem that will result in a direct and positive influence on household and food security for Vietnam.

The key collaborating institutions in Vietnam are the Ministry of Agriculture and Rural Development (MARD) through the National Institute of Plant Protection (NIPP) in Hanoi, the Southern Fruits Research Institute (SOFRI) at Long Dinh, the Provincial Plant Protection Departments (PPPD), Foster's Asia (Tien Giang Brewery) and AVENTIS Crop Science (Vietnam office). Collaboration within Australia is with Foster's Brewery, Melbourne, AVENTIS Crop Sciences and the Crawford Fund.

The various components of the project are progressing well and have yielded good results as follows:
Surveys of fruit flies in Vietnam are being conducted through trapping and host fruit collecting. These surveys have revealed nine species of fruit flies as being of major economic importance, along with another 20 non-pest species in Vietnam.
Studies to define the pest species of fruit flies, their geographic and host ranges, and levels of damage to major food crops have shown that crops such as gourds and water apple experience losses in excess of 95% in unprotected situations.
The construction of a protein manufacturing plant at Foster's Brewery, Tien Giang, although initially delayed, has now been completed and the necessary equipment (yeast waste evaporator, digestor and drum dryer) has been installed. Production of protein from the brewery yeast waste is expected to commence in the final quarter of 2002.
As an interim measure, Vietnamese collaborators were taught the procedures and have begun processing Foster's Brewery waste into protein bait in the laboratory at SOFRI. This laboratory prepared bait is being used for the scheduled laboratory and field tests to determine the correct formulations and application rates for the new protein product for efficient use in fruit fly control programs.

The first of three training workshops planned for the project was conducted in SOFRI from 3 - 7 June 2002. The purpose of the workshop was to train a core group of Vietnamese trainers who will then conduct further training for other Vietnamese staff, especially Plant Protection Sub-Department staff. A comprehensive training manual covering morphology, taxonomy and management of fruit flies in Vietnam was prepared and distributed to all participants. A total of 27 participants from 12 different agencies in Vietnam attended the training workshop. A major outcome of the workshop was the design of a brochure in the Vietnamese language by workshop participants for local farmers on fruit fly identification, biology and management. This brochure is to be produced by NIPP and SOFRI and widely distributed to Vietnamese farmers. The circulation of this farmer brochure will be made in conjunction with the release of the new protein bait from Foster's Brewery at Tien Giang.

Extensive collections of adult fruit flies collected from male lure traps and reared from host fruits (cultivated and wild) over 23 provinces covering north, central and south Vietnam have revealed that there are nine species of fruit flies are of economic importance to horticultural production and export trade in Vietnam. These are Bactrocera dorsalis, B. carambolae, B. correcta, B. cucurbitae, B. diversa, B. latifrons, B. pyrifoliae, B. zonata and B. tau. The species causing the greatest damage in north Vietnam are B. dorsalis, B. pyrifoliae and B. cucurbitae, whereas in south Vietnam the species causing greatest damage are B. dorsalis, B. correcta and B. cucurbitae. Crop losses ranging from 40 to 100% are being recorded in a wide range of fruits and vegetables when no control measures are applied.

To assist with field control studies, laboratory colonies (rearing on artificial diet) of B. cucurbitae have been established at the National Institute of Plant Protection (NIPP). Colonies of two other major pest species-B. dorsalis reared from litchi and B. pyrifoliae reared from peach are also being set up at NIPP. At the Southern Fruits Research Institute (SOFRI) colonies of pest species B. correcta and B. dorsalis have been successfully established. These laboratory colonies currently provide adult flies for laboratory and field testing of the new protein bait produced at Foster's Tien Giang brewery.

The production plant at Foster's Brewery at Tien Giang to process brewery waste into a fruit fly bait has been fully commissioned. Batches of protein bait produced at the plant are currently being evaluated by NIPP for field control of fruit flies infesting peach and bitter luffa, and by SOFRI on water apple and guava.

An extensive training program for on the biology and control of fruit flies for Provincial Plant Protection Department (PPPD) staff, as well as for farmers has been successfully implemented in various provinces around Vietnam. Project staff from Brisbane initially ran a week-long workshop in June 2002 to train a core group of trainers from NIPP, SOFRI, PPD and selected universities in Vietnam. These staff members from NIPP and SOFRI have subsequently completed training a total of 177 PPPD staff as well as 1600 farmers from 16 provinces across Vietnam. The training was accompanied by the distribution of over 3000 illustrated brochures in the Vietnamese language on the biology and management of fruit flies, incorporating the new bait spray technology.

Continued collections of adult fruit flies collected from male lure traps and reared from host fruits (cultivated and wild) over various provinces covering north, central and south Vietnam have not revealed any new pest species beyond the nine that have so far in the project been identified as being of economic importance to horticultural production and export trade in Vietnam. These are Bactrocera dorsalis, B. carambolae, B. correcta, B. cucurbitae, B. diversa, B. latifrons, B. pyrifoliae, B. zonata and B. tau. In the current sampling period, however, B. correcta has also been recorded in North Vietnam where it was not recorded before. Crop losses ranging from 40 - 100% are still being recorded in a wide range of fruits and vegetables when no control measures are applied.

Laboratory colonies (rearing on artificial diet) of B. cucurbitae at the National Institute of Plant Protection (NIPP) and B. correcta and B. dorsalis at the Southern Fruits Research Institute (SOFRI) continue to be maintained. These laboratory colonies currently provide adult flies for laboratory and field testing of the new protein bait produced at Foster's Tien Giang brewery.

The protein bait production plant at Foster's Tien Giang has been fully commissioned and was officially launched together with the bait called SOFRI PROTEIN on 16 April 2004 by the Australian Ambassador to Vietnam, the Hon Joe Thwaites. The high profile ceremony was attended by senior officials from the Tien Giang provincial government, Ministry of Agriculture and Rural Development (MARD) Vietnam, Fosters Brewing International, Griffith University, Queensland Government, AusAID and ACIAR.

The one-day training program for farmers on the biology and control of fruit flies for Provincial Plant Protection Department (PPPD) staff, as well as for farmers has been successfully continued. In South Vietnam, training in this reporting period has focussed on Barbados Cherry farmers with whom an large area fruit fly control program is being organised. Over 4,000 copies of farmer extension leaflets have been printed and distributed by NIPP and SOFRI.

To ensure more effective control of fruit flies, two large-area fruit fly control programs have been implemented, one on peach grown by minority hill tribe people in Moc Chau, North Vietnam, and the second in Barbados cherry grown largely in Tien Giang province in South Vietnam. These trials combine the use of male lures with SOFRI PROTEIN bait sprays for more effective fruit fly population suppression. Farmers are organised in large groups of 25 or more and apply these treatments in a coordinated manner. Initial results show excellent suppression of pest fruit fly populations and very low levels of fruit damage to the benefit of a large number of farmers in an area.

SUMMARY FROM FINAL REPORT:
Extensive collections of adult fruit flies obtained from male lure traps and reared from host fruits (cultivated and wild) over 23 provinces covering north, central and south Vietnam have revealed that there are eight species of fruit flies are of economic importance to horticultural production and export trade in Vietnam. These are Bactrocera dorsalis, B. carambolae, B. correcta, B. cucurbitae, B. latifrons, B. pyrifoliae, B. zonata and B. tau. The species causing the greatest damage in north Vietnam are B. dorsalis, B. pyrifoliae and B. cucurbitae, whereas in south Vietnam the species causing greatest damage are B. dorsalis, B. correcta and B. cucurbitae. Crop losses ranging from 40 - 100% are being recorded in a wide range of fruits and vegetables when no control measures are applied.

To assist with field control studies and other ACIAR funded projects to improve postharvest fruit quality, laboratory colonies (rearing on artifical diet) of B. cucurbitae B. dorsalis and B. pyrifoliae were established at the National Institute of Plant Protection (NIPP). At the Southern Fruits Research Institute (SOFRI) colonies of pest species B. correcta and B. dorsalis were successfully established. These laboratory colonies provide adult flies for laboratory and field attractancy testing for on-going quality control of the new protein bait produced at Foster's Tien Giang brewery.

The protein bait production plant at Foster's Tien Giang was officially launched on 16 April 2004 by the Australian Ambassador to Vietnam, the Hon Joe Thwaites. The ceremony was attended by senior officials from the Tien Giang provincial government, Ministry of Agriculture and Rural Development (MARD) Vietnam, Fosters Brewing International, Griffith University, Queensland Government, AusAID and ACIAR, as well as over 200 local farmers from the surrounding provinces. The protein product has been officially registered in Vietnam under the trade name SOFRI Protein 10DDD (yeast protein + fipronil insecticide) and is being made available for sale to farmers through a local pesticide distribution company that has numerous outlets in the Mekong delta. The price has been kept very low and is controlled by a steering committee comprising Griffith University, ACIAR Vietnam, SOFRI, Foster's Tien Giang and Cantho Pesticide Company. The protein production plant at Foster's Tien Giang has an annual production capacity of about 50,000 litres of protein.

The use of SOFRI Protein 10DD both in small farms and over large areas has provided excellent control of pest fruit flies across a range of fruit and fruiting vegetable crops in Vietnam, resulting in major increase in incomes for farmers. For example, H'Mong minority hill tribe people in the northern province of Son La, Moc Chau district cultivate peach as their main crop but have always had to harvest the fruits hard green because ripe fruits are 100 per cent infested by fruit flies. Using the spot spray protein bait technology, fruit fly damage was reduced to less than 5 per cent, resulting in farmers harvesting ripe fruits from which they obtained higher yields and better prices. Overall this resulted in a 4-fold increase in incomes in one season for the peach farmers. Barbados cherry farmers in Go Cong province in the Mekong Delta have also experienced a similar benefit with a 2-fold increase in incomes resulting from higher yields in orchards using SOFRI Protein 10DD. Control trials in other crops like guava, jujube, luffa and bitter gourd have reduced fruit fly damage from over 70 per cent to less than 5 per cent. The protein bait spot spray technique also provides major health and environmental benefits by utilizing extremely low amounts of pesticide and spray volumes in comparison with cover sprays of insecticides. The Ministry of Agriculture and Rural Development now plans to set up more protein production plants in other breweries across the country and extend the benefits of the protein spot spray bait technology to the wider farming community in Vietnam.

An extensive training program on the biology and control of fruit flies for Provincial Plant Protection Department (PPPD) staff, as well as for farmers has been successfully implemented in various provinces around Vietnam. Project staff from Brisbane initially ran a week long workshop in June 2002 at SOFRI to train a core group of trainers from NIPP, SOFRI, PPD and selected Universities in Vietnam. These staff from NIPP and SOFRI have subsequently completed training a total of 290 PPPD and sub-PPPD staff as well as 4445 farmers from 16 provinces across Vietnam. The training was accompanied by the distribution of over 5000 illustrated brochures in the Vietnamese language on the biology and management of fruit flies, incorporating the new protein bait spray technology introduced by the project.

Year 5 report not sent due to a gap between final report and new variation.

Year 6 progress summary is for the period 1 May 2006 - 30 April 2007.
1. Pairs of male lure traps (methyl eugenol and cue-lure) were set up in provinces that were not surveyed before. These are Kien Giang, Thua Thien Hue, Ca Mau, Tra Vinh, An Giang and Binh Phouc in South Vietnam, and Son La, Thai Nguyen, Tuyen Quang and Ha Giang in north Vietnam. No new species of flies have been recovered so far and the species composition remains the same as for the provinces surveyed in the earlier project. Additional samples of host fruit have been collected to verify uncertain host records.

2. Trials on peach and plum integrating fruit fly control using spot sprays of protein baits with improved crop management techniques and post-harvest handing introduced through ACIAR project CP/2002/086 - Improving postharvest quality of temperate fruits in Vietnam and Australia were planned for various locations in Moc Chau province. However, these trails could not be implemented this year because of high variability in fruiting and fruit load within some of the experimental sites. It is thus proposed to conduct these trials during the next fruiting season in March - June 2008.

In the Mekong Delta, trials were planned to integrate fruit fly control using spot sprays of protein baits with improved control of pests other than fruit flies using white mineral oils introduced through ACIAR Project CP/2000/043: Huanglongbing management for Indonesia, Vietnam and Australia. One trial on Barbados cherry has been established to evaluate the effectiveness of integrating SOFRI Protein and methyl eugenol trapping for fruit fly control, and the mineral oil SK Enspray 99 for other such as aphids, mites, scales and mealy bugs. The trial is in progress.

A new protein bait plant has been constructed at An Thinh Brewery in Hanoi, and the facility was officially launched on 9 May 2007.

3. The farmer training programme is continuing with another 2950 farmers having received training in fruitfly biology and field pest management in the provinces of Hue, Quang Nai, Quang Nam, Lam Dong, Contum, An Giang, Kien Giang, Ca Mau, Soc Trang and Bac Lieu.

The list of fruit fly species occurring in Vietnam has been completed and this will be of great value to Vietnam in meeting with the SPS requirements, and to quarantine authorities and in trade negotiations.

A major development in the project has been the construction of a protein bait plant at An Thinh Brewery near Hanoi that has a capacity to produce 100,000 litres of bait a year. This plant was constructed by the Vietnamese using their own funds. The bait is being marketed under the trade name Ento Pro 10DD and is available to farmers in north Vietnam.

Farmer training continues to extend the protein bait spot spray technology to a number of new crops like guava, sapota, dragon fruit, mango and pomelo, all of which are in the national priority list for export development.

The first annual planning and review meeting was held in Hanoi, in February 2003.. Experimental protocols were planned for most project activities after relevant presentations from participants. Orchards were also visited at Cao Phong in Hao Binh province. Protocols were circulated for comment in May 2003 but progress during the year was hindered by travel restrictions related to regional uncertainty.

In September/October 2003 Dr Andrew Beattie (the project leader) and Dr Paul Holford visited the Indonesian Botanic Gardens in Bogor and Gadjah Mada University in Yogyakarta. During this visit potential sites for field experiments were inspected and plans for constant temperature facilities required for other experiments were reviewed. Dr Paul De Barro (CSIRO) suggested modifications to the facilities when he visited Indonesia in December 2003 to plan experiments and visit field sites. Construction of the facilities has commenced and completion was scheduled for March 2004. In late 2003 a decision was made to appoint an entomologist (Dr Zamir Hossain) from January 2004, to undertake project activities in Indonesia, and to help coordinate other activities.

Dr Holford and Dr Beattie visited Vietnam from 5 to 10 October 2003 to review project protocols and continue planning of experiments. Dr Beattie also discussed these plans with project personnel during other non-project visits. Dr Mabberley helped to prepare a list of plants (citrus and citrus allies) to test as hosts for huanglongbing and the Asiatic citrus psyllid (Diaphorina citri), and to resolve uncertainty about relationships between these plants-an issue vitally important for the assessments.

During 2004, significant progress was made in overcoming setbacks stemming from the impact of international events in 2003. Field sites were established for a major objective in Vietnam in which management strategies (including use of white mineral oils and imidacloprid, a systemic insecticide) for control of huanglongbing and its vector, the Asiatic citrus psyllid Diaphorina citri, are being compared. Screenhouse studies on the effectiveness of imidacloprid were also undertaken in northern Vietnam. Progress was also made towards resolving crucially important issues related to the identity of true species of Citrus and Citrus relatives in Vietnam. These issues arose in 2003 when it became clear that published descriptions of these plants are in many instances of questionable accuracy and not supported by 'voucher' specimens in herbariums. These issues are important for management of the disease and conservation of unique plant material.

In Indonesia, controlled environment and screenhouse facilities required for research and propagation of plants were completed in July and became operational in September-October. Personnel required for each of the project objectives were chosen in April and they commenced work in July. Sites for major field experiments were selected at three altitudes(50, 650 and 1300 m above sea level). These sites will be used to assess management practices and the impact of climate on levels of infestations of the vector and the rate of spread of the disease. Initial controlled environment studies on the impact oil white mineral oil deposits on the feeding behaviour of the vector indicated that deposits of 0.25% to 2% sprays significantly reduce feeding. Confirmation of these results will boost our confidence in the use of oil sprays to reduce transmission of disease through control of the psyllid (based on ACIAR CS2/1993/005 in China, ACIAR CS2/1996/176 research in Sarawak (East Malaysia), and subsequent research in Sarawak) and reduced feeding by adult psyllids. This should apply particularly to situations where it will be possible to maintain populations of the psyllid at low or negligible levels, such as in large orchards or smaller orchards in areas/localities with common management strategies.

The annual project planning and review workshop was held from 22-26 November in Yogyakarta. It was attended by more than 10 participants from Indonesia, four from Vietnam, three from China, one from Japan and three from Australia. Highlights of the workshop were presentations by UGM personnel, a valuable review of research on the disease and the vector in China, and preliminary evaluation of a simple field microscope for detecting disease-damaged tissue in leaves. This technique should prove very practical for field use, and as an alternative to expensive molecular techniques that can only be used in laboratories. It was originally used in Indonesia 20-30 years ago and its usefulness appears to have been overlooked with recent focus by scientists on the use of molecular techniques. The disease leads to degeneration, and ultimately premature death, of cells (known as phloem) that transport sugars and other chemicals in plants. In Indonesia, the disease is known as 'citrus vein phloem degeneration' disease.

Dr Zamir Hossian was appointed as an entomologist at the University of Western Sydney (UWS) in early January 2004 to help manage project activities, particularly in Indonesia. He spent 16 weeks in Indonesia during the year, helping with the selection of personnel to work on project activities at Gadjah Mada University (UGM) and assisting senior Australian (Prof Beattie and A/Prof Holford at UWS, and Dr De Barro from CSIRO) and Indonesian (Prof Susampto, Dean, Faculty of Agriculture, UGM, and Drs Siti Subandiyah and Andi Trisyono) with planning and supervision.

During 2005, significant progress was made in overcoming setbacks stemming from the impact of international events in 2003 and ongoing impacts of similar events on travel. The project was reviewed from 6-12 November by Prof Myron Zalucki (University of Queensland), Dr Nguyen Cong Thuat (Ministry of Agriculture and Rural Development, Hanoi, Vietnam) and Dr Baharuddin (Hasanuddin University, Makassar, Indonesia).

In southern Viet Nam, studies at the project field site at Cai Be in Tien Giang in the Mekong Delta in which application of mineral oils, pesticides and other management strategies are being compared continued. Populations of the Asiatic citrus psyllid and the incidence of huanglongbing were low, much lower than anticipated, an outcome seemingly linked to the presence of guava trees interplanted between citrus trees at the site. The implications for control of the vector and the disease will have major repercussions for citrus production throughout Southeast Asia should ongoing sampling and future experiments confirm the observations.

In northern Vietnam, growth of trees at the project field site at Cao Phong, where a similar experiment to the one at Cai Be, but without guava interplants, is being conducted, was slower than in the tropical south. Populations of the psyllid were very low throughout the year (3 were recorded) and seemingly related to low incidence of psyllid in the local area due to increased use, and more thorough applications, of pesticides in recent years. Infestations of the defoliating gold dust beetle (Hypomeces squamosus) were of concern in all treatments, and raised the possibility that high incidence of the weevil could be due to the widespread absence of the once common predatory ant Oecophylla smaragdina (golden weaver ant) in northern Vietnam.

In northern Vietnam, limited surveys were undertaken to locate and collect material from endemic Citrus spp. and Citrus relatives. This work was done in conjunction with placement of an Australian Business Volunteer (Mr Harry Jones) at the National Institute for Plant Protection (NIPP). Prof David Mabberley (University of Washington) visited Hanoi shortly after Mr Jones returned to Australia. During this visit he examined plant material collected by Mr Jones, visited Cuc Phuong National Park 80 km to the west of Hanoi and the Science & Technology Research Institute, and met with Dr Dzuong Duc Huyen (Head of Botany and the local co-ordinator for PROSEA - a largely Dutch initiative linked to the University of Wageningen) with whom he discussed herbarium records in Vietnam.

Experiments on the impact of deposits of 0%, 0.25%, 0.5%, 1% and 2% of aqueous mineral oil emulsions of two mineral oils on oviposition by Diaphorina citri were also undertaken in screen-houses at NIPP. Deposits of emulsion with  0.5% oil reduced oviposition on buds shorter than 6 mm and longer than 10 mm by more than 75%, with effects more evident for the nC24 agricultural mineral oil tested than for the nC21 horticultural mineral oil tested. Mortalities of eggs laid on deposits, and of first instar nymphs that hatched from eggs laid on the deposits, were also significant: total mortality for the two life-cycle stages in treatments with  0.5% oil exceeded 55%.

In Indonesia, preparation and establishment of all field sites was completed in November. These sites are being used for research on: (a) the impact of altitude (40 m, 670 m and 1300 m asl) on the vector and the disease; (b) the susceptibility of Citrus and Citrus relatives to the disease (c) studies on natural enemies of the vector; and (d) the impact pest management strategies, including the use of mineral oils and insecticides. Experiments in screen-house and constant environment facilities at Universitas Gadjah Mada focused on feeding preferences psyllid adults on citrus relatives, and the impact of aqueous mineral oil deposits on feeding and oviposition by adults on mandarin plants in choice and non-choice assays. In the host-preference studies feeding on Swinglea glutinosa was significantly higher than on the citrus relatives tested. The next most preferred plants were Aegle marmelos and Murraya paniculata, then Limonia acidissima, Triphasia trifolia and Glycosmis pentaphylla. Males showed least interest in feeding on Murraya exotica, but no species was rejected for feeding.

In the choice tests with aqueous oil emulsions applied to mandarin plants, the proportion of adults landing on the plants treated with 0%, 0.25%, 0.5%, 1% and 2% oil emulsions was 49%, 18%, 17%, 9.4% and 6.4% respectively. Avoidance was more noticeable in the no-choice test. Mortality of adults 1 d after they were caged with plants sprayed with 0.25%, 0.5%, 1% and 2% oil emulsions was 22.5%, 37.5%, 40%, and 65% respectively: 100% adults died 2 d after they were caged with plants treated with 2% oil. Mortality appeared to be related to reduced feeding on plants sprayed with oil.

In southern Viet Nam, studies at the project field site at Cai Be in Tien Giang in the Mekong Delta in which application of mineral oils, pesticides and other management strategies are being compared continued. Guava interplants within the site continued to limit ingress of huanglongbing and its vector, the Asiatic citrus psyllid, into the site. The level of infection three years after citrus trees were planted reached about 20%. This compares well with about 24% two years after planting in French and Vietnamese studies at Cai Be in which the systemic insecticide imidacloprid was applied initially as soil drenches and then directly to the trunks of trees. The impact of guava on spread of the disease has the potential to radically alter the economics of citrus production in the Mekong Delta, and in similar environments, through additional income from sale of guava fruit, dramatic reductions in pesticide use, and, through increased longevity of citrus orchards, increased income from the sale of citrus fruit. The impact of guava interplants on spread of the disease has gained worldwide attention. In northern Vietnam, growth of trees at the project field site at Cao Phong, where a similar experiment to the one at Cai Be, but without guava interplants, is being conducted, continued to be slower than in the tropical south. No psyllids were observed within the site during 2006 and symptoms of the disease were not observed. Infestations of the defoliating gold dust beetle (Hypomeces squamosus) were, as in 2005, of concern in all treatments. Mineral oil sprays gave good control of citrus leafminer (Phyllocnistis citrella).

In Indonesia, plants in all field sites established in late 2004 grew slowly until the end of the extended dry season, and then quite rapidly. No psyllids were observed in these sites at mid (670 m asl) and high (1,300 m asl) altitudes. No psyllids were observed in mature orchards at 1,300 m asl, or in D-vac samples from these orchards. The major impediments to citrus production at this altitude appear to propagation on non-disease free plants, and inadequate, but relatively easy, control of heavy citrus powdery mildew (Oidium tingitaninum) infections and citrus leafminer (Phyllocnistis citrella) infestations. The diversity and abundance of predatory coccinellid species varied consistently with altitude. The incidence of the predators was highest at the lowest altitude and lowest at the highest altitude. Black citrus aphid (Toxoptera citricida) was the main host of the coccinellids, and heavy infestations were observed at all altitudes. Both primary parasitoids of the psyllid were recorded, with the endoparasitoid Diaphorencyrtus aligarhensis being more common than the ectoparasitoid Tamarixia radiata. The favoured hosts of the psyllid among 20 plant species and cultivars being compared at Purworejo (40 m asl) were, from October 2005 to July 2006, Citrus - junos, then Swinglea glutinosa and Murraya exotica. During the same interval nymphs were most common on Citrus hystrix, M. exotica, S. glutinosa and Bergera koenigii. More than 240 adults and 420 nymphs were recorded over the 9 months. At this point, M. paniculata, as a host of adults, ranks thirteenth among the 20 plant species and varieties in the trial, and no nymphs were recorded on any of the 16 M. paniculata plants in the trial. These results were unexpected, as M. paniculata is considered to be the favoured host of D. citri, and we are not certain if plants named as M. paniculata represent a single highly variable species, a hybrid (M. exotica), or two species (M. paniculata and M. exotica), as the validity of M. exotica as a species is uncertain. However, in our studies, significant differences in morphology and molecular biology were detected between the two species as named and M. exotica is growing and flowering more prolifically at Purworejo than M. paniculata. These findings are important as it now clear that M. exotica is susceptible to huanglongbing whereas M. paniculata is possibly not or is asymptomatic and plants grown as M. paniculata in China, Florida and Brazil resemble M. exotica.

No psyllids were recorded in the Purworejo where six treatments are being compared for control of the psyllid. However, some 20 sprays have been applied in the oil treatments since late 2004. No oil-induced phytotoxicity as been observed but some minor, not readily noticeable, oil soaking has occasionally been observed along the midveins of mature leaves. Trees in the imidacloprid treatment did not appear to be any better than trees in other treatments. Citrus greasy spot (Mycosphaerella citri) was less common on oil sprayed trees. Experiments in controlled environment facilities at Universitas Gadjah Mada demonstrated that psyllid females can detect mineral oil deposits on treated surfaces before landing, most probably though detection by antennal receptors of volatile oil molecules.

Studies on huanglongbing and its vector, the Asiatic citrus psyllid (Diaphorina citri) continued in Viet Nam and Indonesia. In southern Viet Nam, where applications of mineral oils, pesticides and other management strategies are being compared for control of the psyllid in an orchard at Cai Be in Tien Giang in the Mekong Delta, no differences have been detected between treatments. This outcome is due to the presence of guava trees inter-planted with king orange trees in the orchard. The level of disease in the orchard reached 24% in March 2008, increasing slightly from 20% in early 2007. This level of disease is dramatically lower than in orchards in the absence of guava. The psyllid has not been observed in the orchard, or in similar orchards, by University of Western Sydney personnel, scientists and citrus growers from Florida (April 2007), and scientists from China (May 2008), whereas, psyllids are easily found in orchards were guava trees are not planted. As a result, evaluation of guava interplants has commenced in northern Viet Nam, Indonesia, Florida and China; work in China has shown that guava volatiles repel adult psyllids. Other sources of volatiles are being considered and use of guava and other plants as groundcovers in inter-row spaces has been contemplated.
At Cao Phong in Hoa Binh near Ha Noi in northern Viet Nam, no psyllids were observed in the project orchard where applications of mineral oils, pesticides and other management strategies are being compared. The absence of the psyllid was most probably due to high pesticide use in adjacent orchards and to climate (Cao Phong is cooler than Cai Be, where the potential for psyllid populations to increase is higher).
In Indonesia, application of foliar fertilisers was shown to reduce disease severity in mandarin trees. Other studies have demonstrated that the psyllid populations and incidence of the disease are related to altitude: higher at low altitudes (10 to 60 m asl), uncommon at medium altitudes (600 m asl) and absent at high altitudes (1,300). Species abundance and diversity of natural enemies also varied with altitude. Studies on the influence of mineral oils on the behaviour D. citri continued to show that oil deposits reduce feeding and egg-laying by adult psyllids. Psyllid populations, though increasing, remained too low (due to slow ingress from surrounding orchards) during 2007 and early 2008 to determine differences between treatments in the field experiment where applications of mineral oils, pesticides and other management strategies are being compared.
Ongoing reviews of literature continued as part of the project and for preparation of an incursion management plan, for both the disease and its vectors, for the Australian citrus industry. These reviews indicate that Citrus evolved in Australasia, not Asia, as widely accepted, and that huanglongbing did not, as widely accepted, originate in China. Plans for extension activities from 2009 were considered at an annual workshop in Indonesia in January 2008. The workshop was attended by extension personnel and farmers, and inter-planting of guava in citrus orchards in Java, and more widely in Indonesia, commenced in preparation for these extension activities. Supplies of a mineral oil were also arranged for these activities.
Interest in the project led to the chief investigator (Dr Beattie) being invited to the annual Citrus Expo in Florida in August 2007, where he addressed some 700 people, mostly growers. Interest in the project and Australian expertise also led to Dr Beattie being invited by the United States National Science Academy to participate in a meeting in April 2008 to determine priorities for research on huanglongbing and D. citri in Florida.

Studies on huanglongbing and its vector, the Asiatic citrus psyllid (Diaphorina citri) continued in Viet Nam and Indonesia. As in previous years, populations of D. citri and the incidence of huanglongbing remained low in an 0.5 ha guava-interplanted king orange orchard in which mineral oils, pesticides and other management strategies are being compared for control of the psyllid in southern Viet Nam. The orchard, at Cai Be in Tien Giang, was planted in 2004. Differences among treatments remained undetectable and no psyllids were observed. Other studies and observations in the Mekong Delta, and research in Indonesia and China, continued to indicate that low incidence of the psyllid and the disease is due to the impact of guava volatiles on psyllid behaviour. Differences among the same treatments in similar trails at Cao Phong in Hoa Binh near Ha Noi in northern Viet Nam and at Purworejo, Central Java, Indonesia, in the absence guava interplants are still not apparent. At Cao Phong no psyllids were observed. As in previous years, absence of the psyllid was most probably due to high pesticide use in adjacent orchards and colder winters than in southern Viet Nam. However, huanglongbing is present in the orchard at low levels and assessments will continue to in order to determine differences between treatments. Pysllids and huanglongbing have been recorded at Purworejo and the trail will also continue there for another two years in order to determine differences between treatments. No mineral oil-induced phytotoxicity was observed at Cai Be, Cao Phong or Purworejo.
In studies on the impact of altitude on incidence of the psyllid and the disease in Central Java, both continued to be more prevalent at 40 to 60 m asl than at 600 m asl, and absent at 1300 m asl. Parasitism of the psyllid by its two primary parasitoids, Tamarixia radiata Diaphorencyrtus aligarhensis also varied with altitude, and was lower at 600 m asl than at 40 to 60 m asl. These studies also showed that powdery mildew is the most serious impediment to citrus production in orchards above 700 m in Central Java. Trials are being undertaken to evaluate sulfur and mineral oils for the control of the disease.
Studies in Indonesia suggest that higher than normal levels of catechol may occur in huanglongbing infected trees than in disease-free trees. Secondary symptoms of huanglongbing resemble those of zinc deficiency to which abnormally high levels of catechol and phloem degeneration were linked in the 1940s. The nature of phloem degeneration (hyperplasia and hypertrophy of phloem parenchyma leading to crushed sieve tubes) in huanglongbing-affected and zinc deficient, but otherwise healthy, plants appears to be similar. Studies are continuing to determine if excessive levels of catechol in huanglongbing affected plants are related to the response of plants to the pathogens (liberibacters) that cause the disease.
Results from a field site at Purworejo in Central Java where the susceptibility Citrus and Citrus relatives to D. citri and huanglongbing is being assessed were reported at an international conference in Florida in December 2008. The site was established in November 2005. In 2008, the highest populations were recorded on M. paniculata var. exotica then M. paniculata, Bergera koenigii and the other species and varieties that included C. - aurantium L. (sour orange, natsudaidai and Japanese citron), C. hystrix, C. maxima, C. reticulata, Aegle marmelos, Feroniella lucida, Limonia acidissima, Triphasia trifolia from the Aurantieae, and Cl. harmandiana, Cl. lansium, and Glycosmis pentaphylla from the Clauseneae. In late 2008, chlorosis was visible on the foliage of some, but huanglongbing-positive PCR results were only obtained for natsudaidai, pomelo (C. maxima) and C. reticulata var. Grabag.
A 'Pest-Specific Contingency Plan for the Citrus and Nursery and Garden Industries' for huanglongbing and its vectors was submitted to Horticulture Australia and Citrus Australia and reviewed at meeting in Melbourne in February 2009. Preparation of this document would not have been possible without ACIAR's support for research on huanglongbing, D. citri, and other citrus pests and diseases in Asia.
The chief investigator (Dr Beattie) gave an invited keynote address on the origins of citrus, huanglongbing, Diaphorina citri and Trioza erytreae at an international research conference on huanglongbing in Orlando, Florida, in December 2008.

The project is making satisfactory progress with several major objectives met in oil palm and cocoa.
Oil Palm
In oil palm, the mobile card payment scheme which guarantees payment of family and hired labour for work on smallholder oil palm blocks has been introduced at Bialla. The payment initiative, which was developed and trialled under ASEM2002/014, will be introduced to Hoskins smallholders.
In addition, a draft Clan Land Usage Agreement (CLUA) has been developed after extensive consultation with OPIC, customary landowners and migrant smallholders growing oil palm on customary land in Bialla, Hoskins and Popondetta. OPIC supports introducing the new draft CLUA on future smallholder blocks involving land transactions between "outsiders" and customary landowners. The project is in the process of seeking agreement with OPIC to conduct a trial of the new agreement to late 2009 or early 2010.
Cocoa
Cocoa Pod Borer (CPB) is now having a devastating impact on the livelihoods of smallholder cocoa growers in ENBP with almost 90% loss of production in areas with high CPB infestation rates. A social and economic impact assessment of CPB was completed in May 2009, which included a strategy for addressing the problem. Managing CPB requires farmers to raise inputs of labour to undertake CPB management techniques such as weekly harvesting (every mature pod), centralised pod breaking and pod burial. High levels of block management (pruning, shade control and weed control) must also be maintained for effective control of CPB. PNG cocoa farmers conventionally use a low labour input system of production (the foraging strategy) which, with CPB present, can lead to over 90% crop losses. The project is working with commercial sector partners to deliver new forms of extension that aim to mobilise labour for effective control of CPB. While still early days, there are promising results emerging for the monitoring programme with some farmers switching to a high input system and effectively controlling CPB.

The project has two broad objectives: 1) to improve extension delivery through greater commercial sector engagement with smallholders; and 2) to develop effective land use agreements between the commercial sector and customary landowners. The project is making good progress towards these objectives.
In oil palm, the mobile card payment scheme which was designed to mobilise labour in smallholder production by guaranteeing payment of family and hired labour for work on family oil palm blocks has been introduced at Bialla. The payment initiative, which was developed and trialled under ASEM2002/014 is now in the process of being bedded down at Bialla.
In addition, a template for a Clan Land Usage Agreement (CLUA) which was developed last year in consultation with OPIC, customary landowners and migrant smallholders growing oil palm on customary land in Bialla, Hoskins and Popondetta is now available at all OPIC project sites. The template is also being modified for in-fill plantings by the Smallholder Agricultural Development Project, funded by the World-Bank. This new CLUA gives greater recognition to the underlying land tenure rights of the customary landowning group while providing greater tenure security for famers with secondary rights in the land.
The project is also evaluating the range of land use agreements between commercial sector organisations and customary landowners in the cocoa, coffee and oil palm sectors.
In cocoa, the Cocoa Pod Borer (CPB) is having a devastating impact on the livelihoods of smallholder growers in ENBP with almost 90% loss of production in areas with high CPB infestation rates. Provincial production has fallen by around 50% to 10,000 tonnes in 2009, and CPB has now spread to Bougainville where it is anticipated it will decimate the cocoa crop. The pest is causing enormous hardship by undermining people's capacity to earn a living, to meet their needs in education and health and to maintain their general quality of life. However, it is possible for farmers to manage CPB but this requires shifting from the low labour input system of smallholder production that typifies most export cash crop cropping in PNG to a high labour input system involving weekly harvesting, regular pruning and shade control and other new practices to control CPB.
Over the past year project staff have been working with an exporting company, NGIP-Agmark, on new extension strategies to assist farmers make the transition to high input farming. This extension strategy is proving successful with many farmers now adopting the high input system of production. Monitoring of the strategy reveals that those farmers able to make the switch to high input farming are earning good returns on their labour. High input production means much reduced losses from pests and diseases like Black Pod and a yield response to pruning and improved maintenance of cocoa stands. This is leading to much higher yields than those achieved pre-CPB under the low input system of production.
The switch to high input farming involves a lifestyle change for smallholder families. To make the transition easier for families, centralised processing leading to higher quality and higher value dry beans is being trialled by the company. This reduces some of the labour demands on smallholder families by centralising processing, thereby providing more time for growers to undertake maintenance and CPB control practices on their family cocoa blocks. However, not all families are willing or able to make the transition to high input farming. For these farmers the project has been investigating options for diversification into alternative crops that do not depend on the high labour inputs now required for cocoa production in a CPB environment.

The two broad objectives are to: 1) improve extension delivery through greater commercial sector engagement with smallholders; and 2) develop effective land use agreements between the commercial sector and customary landowners.
The mobile card payment mechanism to mobilise labour in smallholder oil palm production by guaranteeing payment of family and hired labour is operating at Bialla. Growers are finding new ways to adapt the payment mechanism to meet their needs. Elderly growers reluctant to hand over block management to their sons because of potential loss of income have adopted the card to guarantee an income in their retirement, thereby lessening the generational conflict and disruption of production often associated with the transfer of block management from fathers to sons. Growers call this the "papa levy".
The new template for a Clan Land Usage Agreement (CLUA) developed in consultation with customary landowners, settlers and other stakeholders is now available at all OPIC project sites, and is being modified for in-fill plantings by the Smallholder Agricultural Development Project, funded by the World-Bank. This new CLUA gives greater recognition to the underlying land tenure rights of the customary landowning group while providing greater tenure security for famers with secondary rights in the land.
The project is also evaluating the range of land use agreements between commercial sector organisations and customary landowners in the cocoa, coffee and oil palm sectors.
The Cocoa Pod Borer (CPB) has spread to eight provinces. By the end of 2010, production in ENBP - the worst affected province - fell by 80% to 5000 t. The pest is causing severe financial hardship to smallholder families with many families reverting to subsistence production as a survival strategy. Smallholders can adapt to living with CPB, but they must adopt much more intensive cocoa management practices including weekly harvesting, regular pruning and shade control and other labour-intensive CPB control techniques.
Working with the exporting company, NGIP-Agmark, the team has been monitoring the delivery, uptake and effectiveness of new extension strategies (Training by Association model) and other services provided by the company. New services for growers include: purchase of cocoa at the roadside edge of the cocoa block, seedling support program including credit and training in cloning techniques and nursery establishment, general extension and advisory services including CPB training and farmer group support, a savings scheme and fermentary credit scheme for farmers, and farm diversification support for growers affected by CPB.
For the 200 farmers participating in NGIP-Agmark's Training by Association model, the results have been remarkable. A relatively high proportion of farmers has successfully made the transition to high input farming, and most of them are achieving yields well above levels of pre-CPB times. General improvements in block maintenance and sanitation mean that losses due to pod rot and Black Pod (around 30% of pods in pre-CPB times) are negligible, and losses to CPB are on average less than one pod per tree.
The switch to high input farming involves a lifestyle change for most smallholder families. However, time allocation surveys reveal that smallholder families that have successfully switched to high input farming have more leisure time and spend more time on community and church activities than families who have not made the transition. This is because the latter group have to spend more time on food garden production for home consumption and sale at local markets. Not all families are willing or able to make the transition to high input farming, and for these farmers the project has been investigating options for diversification into alternative crops that do not require high labour inputs.

During 2008-2009 progress has occurred on several fronts. A field research program has commenced on priority research topics identified by project scientists and local extension officers and farmers. Policy and institutional settings have been further studied through two detailed field trips. An approach for analysing household livelihoods will be based on previous ACIAR funded research in Indonesia. Details are outlined below.
A workshop held in Qingyang in April 2009 identified research topics for on-farm research during the current growing season. Farmers, local extension officers, Chinese and Australian project staff participated in research priority identification. Detailed research protocols were developed and implemented in the following areas:
Evaluation of alternative forage crop species in each of three research locations.
Identification of lucerne management and harvest practices for improved quantity and quality of forage.
Exploration of winter wheat as a dual purpose grain and graze forage option.
Identification of improved ration mixes for sheep and goats utilising available feed resources.
Two extensive field trips in the study region have characterised current policy and intuitional settings relevant to local livestock enterprises. A greatly improved understanding of the complex and sometimes conflicting policies that relate to livestock production has emerged. An important finding relevant to both researchers and policy makers is the challenge of connecting local household livestock producers to livestock markets.
A workshop held in St. Lucia, Australia in August 2008 evaluated alternative analytical approaches to understanding livestock production from integrated crop - livestock farming systems. The key characteristics of the local Gansu system include a diverse range of potential forage sources and crop residues, pen feeding of sheep and goats, severely constrained land resources, constrained labour resources, climate variability and market uncertainty. The Integrated Analysis Tool (IAT), developed in previous ACIAR projects, was selected as the main framework for analysing household livelihoods in this project. The IAT will require significant development, for example inclusion of sheep and goats and local forage and crop options, before it can be applied to local farming system issues.

During 2009-2010 significant progress has been made under Objective 1, implementing a participatory research program into forage production and livestock growth; Objective 2, analysis of institutional and policy settings; and Objective 3, developing capacity for analysing crop-livestock farming systems. Work on Objective 4, evaluation of productivity and resource-use efficiency under farmer-managed conditions will commence in 2010-2011.
A series of forage agronomy experiments were conducted during the 2009 and 2010 seasons under Objective 1. The experiments included; lucerne harvest time, lucerne cutting height, dual purpose (grain and graze) winter wheat, and a comparison of annual forage crops. All experiments were conducted as part of a Masters or PhD postgraduate research project enrolled with Lanzhou University. Highlights from the 2009 season include a significant opportunity for increasing harvestable lucerne forage by shifting to a harvest schedule based on appearance of first-flower, a large grain yield penalty from grazing winter wheat, and higher yield potential from summer forage crops compared to winter forages. Research is proceeding in 2010.
In addition, a cashmere goat feeding trial evaluated the influence of different proportions of lucerne hay in the diet on liveweight gain of weaned kids. The inclusion of lucerne in the diet increased the rate of liveweight gain compared to a diet comprised only of maize straw and feed concentrate. Once lucerne was included in the diet, liveweight gains increased at a decreasing rate with any additional lucerne. This suggests the possibility of conserving lucerne and feeding out gradually over the year rather than the current practice of feeding lucerne when it is green and growing. Further feeding trials are planned in 2010, focussing on the role of maize silage.
Research on Objective 2 of the project - to identify and analyse institutional and policy settings relevant to crop livestock systems in Qingyang City - has progressed essentially as planned. The significant amount of fieldwork and the detailed policy review required to fulfil this project objective has been conducted, and information has been collated and ordered. An early analysis of this information is reported in Brown et al. (2009), while several other papers that link Objective 2 with other objectives of the project are under development. In doing so, further information gaps have become apparent, especially on how policies and institutions on crop-livestock systems in Qingyang relate to those at higher, especially province and central, levels. Interviews in Beijing in June 2010 aim to fill these information gaps.
At the same time, fieldwork allowed for the identification of other issues that are beyond the scope of the initial policy and institutional analysis but which will impact significantly on the broader project objectives. In particular, as a result of several major external shocks (the global financial crisis and volatile food prices), the period 2007-2009 has been time of major change in agricultural and rural markets and industries in China. Data on broader industry developments and market prices was collated in form suitable for input into the economic modelling component of the project.
As envisioned in the project design, the policy, institutional, market / industry information collated so far is important in defining the broader environment in which the farm-levels analysis of the project is conducted. While these settings have been communicated between project collaborators, and policy-makers have been consulted throughout the course of the project, the more formal policy-science linkage workshops planned for the project have been delayed to commencement in 2011, when more developed project findings can be presented to policy makers, thus forming a two-way exchange that is a key component of the project.
The Integrated Analysis Tool (IAT) has been adapted to local farming systems as part of research activity under Objective 3. Key changes included parameterisation for local breeds of sheep (Tan and Han) and goats (Cashmere, Longdong), and development of forage supply quantities and forage quality for key local forage sources including; lucerne, maize silage, wheat straw and feed concentrates. Initial analyses have focussed on farm land allocation to grain or forage crops, lucerne harvest practices, and comparing different feeding strategies to utilise available feed resources. A complementary analysis will identify feed gaps under traditional farmer practice and strategies for addressing the feed gap through a combination of improving the feed resource base and improving the utilisation of available feed.
The 2010-2011 reporting period will see a gradual shift in research focus towards integrated analysis of whole farm productivity, profitability and resource use efficiency at the household scale. Preliminary IAT land allocation results have been obtained in one of the three sites, and it is expected that by June 2011 modelling results will available for three scenarios in the three project locations. The three scenarios are all linked to small ruminant production and include: different land allocations, higher lucerne yields obtained through optimal harvesting and feed year planning centred on strategic feeding of lucerne. The main interest is whether significant gains in household income can be achieved for only small losses in food supply when alternative lucerne harvesting and feeding strategies are used.

In the past reporting period progress has been made towards better understanding the biophysical and economic processes operating in the project's location. This has been achieved through a) conducting crop and livestock experimental trials b) developing location specific whole-farm models to assess net income income-grain supply tradeoffs between different enterprise mixes and government policies and c) monitoring on-farm practices. This research has been an integrated effort between Lanzhou University students, Lanzhou University staff and Australian scientists. The modelling and interpreting of research results is currently progressing. In the upcoming year there will be an increased interaction between extension agents, farmers and project scientists.
Progress related to the different components of the project is as follows:
A series of forage agronomy experiments were conducted in 2009 and 2010. These experiments highlight how the management of current crops and the adoption of alternative crops can influence livestock feed availability, and hence household livelihoods.
Two cashmere goat feeding trials have now been conducted. One trial evaluated how different proportions of lucerne hay in a diet can influence liveweight gains of weaned kids. The second trial focused on the effect of lucerne quality on the liveweight gains of weaned kids. In 2011-2012 a third experiment will focus on how feeding maize silage influences the liveweight gains of weaned kids.
The Integrated Assessment Tool (IAT) was used to assess changes in net income and food security when households altered their crop-livestock mix to meet government targets. The IAT used three farm types. These farm types were determined using Principal Component Analysis and Cluster Analysis.
A village-level model was developed to help scale up the farm-level analyses. Lucerne-growing subsidies and more integrated livestock feed markets are required if households are to meet livestock development goals.
A feed-demand analysis was conducted to assess if current farming systems can support increased livestock numbers. Results from the IAT and feed-gap analysis are consistent with each other, and highlight the need to increase livestock forage resources.
In relation to the policy-related work, preliminarily results related to the above research were presented to the Huanxian Animal Husbandry Bureau in June 2010. This interaction session created an opportunity for project staff and local officials to interact and exchange ideas in an informal setting. The interaction session had approximately 20 participants. The topics covered included the timing and frequency of feed shortages in current farming systems, potential for implementation of alternative feeding strategies and optimal carrying capacity of local farming systems. There were two main outcomes of the session: a) local extension officers gained a better appreciation of how lucerne can be fed in conjunction with other feed sources to reduce feed deficits b) project staff gained a better appreciation of extension officers' ideas for feed management, including a desire to increase the use of maize silage in local systems. To improve the dissemination of results and increase linkages with policy makers, a more formal policy-science linkage workshop is planned for 2012.
A key component of the past year has been the development of strong linkages between Lanzhou University students, Lanzhou University staff and Australian scientists. These linkages are expected to have a lasting impact on the capabilities of new agricultural researchers in China and to lead to peer-reviewed publications. In addition, the capacity of Chinese counterparts has been enhanced by the Australian Youth Ambassador for Development (Joshua Philip), who has spent approximately one year in Lanzhou. Adam Komarek is spending up to one year in Lanzhou to continue research related to this ACIAR project.

In previous ACIAR studies it has been demonstrated that activities which have the highest adverse impact on agricultural sustainability (and therefore long term economic sustainability) in upland farming on Bohol, include: up and down cultivation on sloping lands, continuous use of nutrient-depleting crops such as corn and cassava, and extensive cultivation of steep upland soils. These non-sustainable farming systems have been identified as those with the greatest potential for improvement to reduce soil loss and poverty. In addressing these issues, a key objective of this project is to quantify, demonstrate and provide examples to farmers of the environmental and farm-level economic benefits that can be realised by implementing selected best management practices for soil, water and crop management in affected areas.
The Project team has successfully established 3 improved practice and 3 conventional practice sites in the upper Inabanga watershed, with highly erodible soils on sloping uplands and which include a corn-cassava rotation. The sites are located in the municipalities of Carmen, Sierra Bullones and San Carlos/Pilar. Farmer-cooperators at the improved practice sites have been guided in implementing improved practices. Additionally, sites have been established in the upper Abatan watershed on steeply sloping land where erosion control measures have been implanted for at least 5 years and also on conventionally farmed land. The sites are in the Cadungao municipality, and had received earlier guidance in improved practices through previous ICRAF Landcare activities.
All the sites have been topographically surveyed, geo-referenced and characterised to determine site soil and water components, including physico-chemical properties, soluble and exchangeable cations, pH and nutrients. On-site weather station and automatic water flow and sampling equipment commissioning is being completed. Cropping maps and 5-year cropping schedules have been completed for the sites, with ongoing labour, planting material and fertiliser inputs recorded for all cropping activities. Additionally, crop transport and crop return revenues are recorded. Training workshops on record keeping using farm journals and farm budgeting have been conducted for farmer-cooperators, local government and Municipal Agriculture Officer (MAO) representatives. FMD teams have developed 5-year cropping plans and have been guided in appropriate land management, following initial characterisation of the farm sites.

The socio-economic team has carried out baseline and end-of-cropping-season surveys to quantify productivity of crops, input costs, marketable yields, market prices and overall economic returns on all farms within the improved and unimproved subwatershed sites. Case study data has been collected by the farmer cooperators and the Project team and is being used to guide the team in upcoming cropping periods.

A key objective of this project is to quantify, demonstrate and provide examples to farmers of the environmental and farm-level economic benefits that can be realised by implementing selected best management practices for soil, water and crop management in affected areas.
In addressing this objective, the Project team has established three improved practice and three conventional practice farmer managed demonstration (FMD) sites in the upper Inabanga watershed, with highly erodible soils on sloping uplands and which include a corn-cassava rotation. The sites are located in the municipalities of Carmen, Sierra Bullones and San Carlos/Pilar. Farmer-cooperators at the improved practice sites have been guided in implementing improved practices. Additionally, two FMD sites have been established in the upper Abatan watershed on steeply sloping land. On one site, an improved site, erosion control measures have been implemented for at least 5 years and on a conventionally farmed, unimproved site, no improved land management practices have been used. The sites are in the Masonoy municipality, and farmers for the improved site received earlier guidance in improved practices through previous ICRAF Landcare activities.
All the sites have been topographically surveyed, geo-referenced and characterised to determine site soil and water components, including physico-chemical properties, soluble and exchangeable cations, pH and nutrients. Eight gauging stations with on-site weather station and automatic water flow and sampling equipment are operational and regularly monitored. Initially developed cropping maps and 5-year cropping schedules have been reviewed and updated for sequential cropping periods. Detail of ongoing labour, planting material and fertiliser inputs, crop transport and crop return revenues are recorded and used in proactive planning for subsequent cropping and land management planning.

In addition to soil/ water management and agro-economic/socio-economic issues, the project has actively introduced improved farming practices that have the potential to increase farmers' production. All of the Project FMD sites have improved soil and water management practices implemented (for the improved sites), however there is a need for improved production technology techniques and strategies to complement established Landcare practices. Farmer cooperator groups are being introduced to a range of different and appropriate crop production technologies. They may then evaluate which among the practices can increase their production and income. These practices may then be promoted to other farmers.

Training-of-trainers workshops and farmer field schools which focus on improved crop production strategies have been established in a cooperative initiative involving the BSWM, ICRAF, East-West Seed Co., Department of Agrarian Reform (DAR), DENR, LGUs and Australian project staff.

Significant progress has been achieved towards meeting key objectives of this project, to quantify, demonstrate and provide examples to farmers of environmental and farm-level economic benefits, which can be realised by implementing selected best management practices for soil, water and crop management in affected areas.
In the establishment and ongoing maintenance of farmer managed demonstration (FMD) sites, local municipalities in the San Isidro and Inabanga watersheds have provided consumables such as fertilizer, seed and cuttings. ACIAR Project partners provide instrumentation, training and soil management/cropping expertise. Additional fertiliser and seed supplies are provided by Project partners to enable recommended cropping practices to be implemented in the FMD sites. The establishment and ongoing community management of the FMDs has assisted in the process of improving adoption by provincial farmers.
It has been estimated by Project training staff and local municipal agriculture officers that approximately 50% of the farmers in the upper Inabanga watershed are now adopting improved farming practices as a consequence of project training and dissemination of project results, cropping advice and supply of planting materials. The improved practices include use of contour cultivation, natural vegetated strips (NVS and NVS with crops, eg. pineapple), minimum tillage planting, cover crop plantings, alternative cropping including vegetables, natural residue and/or plastic mulching, accelerated composting and Vermiculture techniques.
Agro-economic data analysis and soil/water monitoring budgets are being developed to identify which farming practices and which crops were most successful in improving economic returns in relation to those that were most successful in improving soil and water losses.

This project provides a foundation for improved food security and increased cash income for some of the poorest farmers in India, through improved understanding of small watershed hydrology with respect to water harvesting, developing easy-to-use principles for design and implementation of water harvesting technology, evaluation of downstream hydrologic impacts, and developing crop options and farming systems that effectively use land and water resources. The project commenced on 1 April, 2006. Activities within each of 5 project objectives include:

1. Develop, validate and promote water harvesting principles
The basis of WSD, as practiced by the NGO partner, PRADAN, is social development through self-help groups, a process which has commenced and will proceed throughout the project.

The first step in developing a watershed development plan is to understand the resources available and any constraints to development. PRADAN is undertaking socioeconomic 'mapping', and other team members are focused on the biophysical assessment. Soil and landscape assessment within the case study watershed (Pogro) has focused on: a) soil descriptions and classification, with 12 soil pits completed, representing the major landscape units and variation within them; b) surface soil sampling and analysis for fertility assessment within each of the main landscape units (~100 sites in Pogro watershed, and in Amagara where agronomic work will be located initially), with the partially completed analyses revealing higher pH than expected for this region (lowland mean pH 7.2, medium lowland 6.0 and uplands 5.5) but consistent with observations of annual Medicago spp., and indicating organic carbon is low at all locations (0.6-0.7%) as with exchangeable potassium; c) water resource assessment using EM38 and EM31 electromagnetic induction, with watershed maps of conductivity having been produced and instrument calibration proceeding to convert this to estimates of profile water. PRADAN have produced a 'resource map' of the Pogro watershed. All data will be georeferenced and mapped onto cadastral maps (scanned into ArcView) or a topographic map which has been created in ArcView from our dGPS.

Work has just commenced on developing databases of climate and hydrology for subsequent modelling purposes. In Pogro, baseline hydrologic conditions are being established through gauging surface flow at the outlet by placing gauges on the wall of two culverts and training a local farmer to take readings as required. Thirteen piezometers have been installed for monitoring shallow groundwater, and water level in 7 wells is also being monitored, as well as water level in 4 'ponds' in which staff gauges have been located.

A conceptual model for hydrology of small watersheds has been developed that could form the basis for modelling from which criteria ('thumb rules') can be developed to either assess the suitability of small watersheds for development or to plan, design and install water harvesting measures. Fill rate tests have been conducted on bore holes as a first step to develop design principles for 'seepage tanks'.

2. Make a preliminary assessment of the applicability and sustainability of water harvesting across the East India Plateau.
The first step has been to review hydrologic and geo-hydrologic models.

3. Develop cropping system options and improved agronomy to effectively use harvested water.
Participatory workshops and other activities led to a set of issues where the objectives of the project and farmers coincided. Three action-learning fertiliser workshops (75 farmers) followed, which included subsequent on-farm experiences. Planning for this 'experience', together with earlier workshops, led to agreed experiments evaluating maize, black gram, upland rice and finger millet as alternative crops for upland areas and poorer upland paddy, all with fertiliser treatments, and to evaluation of fertiliser responses on kharif rice. In Amagara, rice of 3 different durations was planted as the commencement of a cropping systems experiment that evaluates the opportunities and costs associated with shorter duration rice varieties, which has emerged as a central theme for the agronomy/farming systems work. Other experiments explored the growing of early season vegetables, but the focus here is on the change process amongst farmers. In all, more than 90 farmers have been involved in the design, implementation and management of these trials, all of which are replicated across farmer's fields. Harvest has not been completed, but already it is clear that there are very large responses to P in all crops, to K in legumes, and to N in non-leguminous crops. Native rhizobia appear to be effective on the legumes examined.

4. Evaluate biophysical and socioeconomic impacts
A workshop was held in Toowoomba involving most team members (8 from India) at which a participatory evaluation process and framework was developed and a set of indicators.

5. Enhance the capacity of PRADAN (and other NGO's) to undertake watershed development.
PRADAN has been exposed to new ways of conducting on-farm investigations which should involve no more resources but provide much better information. Rigorous evaluation of all activities has been implemented, to introduce greater discipline into the (continuous) learning process. New crop options have been collaboratively developed for evaluation. PRADAN have been exposed to GIS, which may be introduced into broader activities.

Watershed development (WSD) is a major strategy for livelihood improvement. Often the strategy is not underpinned by sound hydrology; agronomic improvement is not integrated with 'water harvesting'; and questions have been raised about the costs and benefits, especially in lower-rainfall areas. This project uses participatory action research to develop principles for WSD for the high rainfall East India Plateau. It will:

1. Develop, validate and promote water harvesting principles

2. Make preliminary assessment of the applicability and sustainability of water harvesting

3. Develop cropping systems and improved agronomy to effectively use harvested water,

4. Evaluate biophysical and socioeconomic impacts, and

5. Enhance the capacity of PRADAN and other NGO's to undertake WSD.

Implementation of the research follows 5 concurrent 'themes' in WSD: (i) Resource assessment, (ii) Developing social capacity, (iii) Improving agronomic skill (iv) Developing crop options and farming systems and (v) Integrating water resource management.

Pre-and post-intervention monitoring of socioeconomic and biophysical indicators is being undertaken in a case-study catchment (Pogro). In the pre-intervention period (2006-'08), intensive resource assessment (i) and capacity building (ii and iii) are being undertaken and evaluated, along with preliminary evaluation of new crop options and their management for the kharif in uplands (iv). Novel farming systems (iv) and integration of rainfed and irrigated cropping (v) are being developed in Amagara. Options developed in Amagara will be integrated and evaluated following intervention in Pogro (v).

Key Findings

Process
Villagers have participated fully and enthusiastically in the action learning cycle of "planning (the research activity) - doing (planting /managing fields) - observing (gathering data, making visual observations) and reflecting (workshops for making sense, understanding)". Following a review of all results in a crop season ('reflection'), farmers and the research team rate each activity for its contribution to learning about important issues and assess whether more work on a topic or new topics is needed, followed by further planning etc. In the kharif seasons of 2006 and 2007, experiments were placed in 111 and 257 farmer's fields, respectively, whilst a further 34 fields were involved in the 2006/07 rabi season (Amagara only). We have witnessed increasing willingness and capacity of farmers to take responsibility for improving their situation. Involving women's self-help groups (SHG's) to organise on-farm activities has released the research team from the need to encourage and sometimes cajole farmers to do what they had agreed to do, resulting in a much higher compliance with previously agreed actions. Farmers at Amagara show evidence of a 'transformation' in attitude from 'beneficiary of aid' to 'manager of resources'.

Resource assessment - Pogro
Profile soil descriptions and other data were used to map soil types. Surface soil fertility was assessed in conjunction with a survey of crop yields (see below).

Soils are generally less acid (pH 5.5 -7) than expected, except for isolated upland surface soils (pH<5). Near-neutrality is indicated by the annual legume Medicago sp., a valuable animal feed and medicinal herb. Soil nitrogen (N) and organic carbon (C) are low, as expected, and 'plant available' phosphorus (P) is almost universally very low. Deficient P and ubiquitous Medicago suggest the catchment ecosystem might be primarily P-constrained, not N-constrained. Infiltration rates were moderately high in uplands (10-30 mm/hr) but low in drained rice fields, both at the surface and below the puddled zone.

The cadastral map has been scanned and ortho-rectified. A digital terrain model has been developed from dGPS data. Hydrologic features such as drainage lines, wells, ponds have been geo-referenced and water levels monitored and recorded, in 15 piezometers (not all active), 7 wells and 4 ponds, and in 2 culverts at the catchment outlet.

Electro-magnetic induction (EM31/38) Surveys of Pogro and Amagara have been completed in wet and dry periods. The signal responds to profile water content, but not as strongly as expected, presumably because of low electrical conductivity of these highly leached profiles. Evaluation of EMI for water resource assessment is continuing.

PRADAN have prepared resource maps and intervention plans based on their conventional approach for later evaluation using project outputs.

Developing social capacity and improving agronomy (human capacity)
The participatory process following the PRADAN model has been extended to include systematic investigation of soil and water resources with the intention of further increasing social and human capacity to undertake the complex task of watershed development. An important learning is that the role of women's SHG's can be expanded from micro-finance to include oversight of technical issues. Workshops on fertilisers included field experiments by farmers, that excited much interest and enhanced basic agronomic skills. In Amagara, the team responded to a request for work on early-season vegetables by developing the activity as an enquiry into the extension process. There was evidence that this contributed to the transformation in attitude referred to above. Soil/crop surveys also excited interest, with villagers calling for them to be repeated and expanded in 2007.

Developing crop options and farming systems
Experiments in Pogro in the kharif 2006 and '07 explored constraints to productivity and new crop options for uplands or 'poor' rice fields. Experiments in Amagara in the '06/'07 rabi explored cropping on residual water under rainfed and irrigated conditions, aimed at efficient irrigation strategies combined with improved crop nutrition. Short-duration rice is being examined as a means of reducing climate risk whilst increasing the opportunity to take a second crop (with or without supplementary irrigation). Experiments are statistically designed for scientific rigour yet compatible with farmer-participation.

Key findings from the 2006 kharif were:
o In the rice crop survey at Pogro:
Yield was equal in 'good' and 'poor' rice land in this favourable year, showing that differences in perception of land quality relate to reliability for rice production, not intrinsic agronomic potential.
Yields were positively (significantly) related to available P and mineralisable N.
Unquantified management factors were also important.
o In comparisons of matched rice fields, increasing P by 10 kg/ha over the farmer's rate (often zero) increased yield by 20%, with further increase with additional N.
o P doubled yields in black gram at Pogro, and residual response to P is evident in natural Medicago.
o Responses to P were observed in maize, finger millet and direct-seeded rice.
o In the rabi experiments at Amagara:
Wheat and mustard failed without P, but good yields were obtained with P up to 50kg (P)/ha.
One or two irrigations (each ~ 40 mm) gave yields approaching full irrigation (4-5 irrigations), apparently by using residual water in the soil after rice.

Water resource management
Conceptual models have been developed to simulate the water balance at the scales of individual intervention, small catchment (1-2 km2) or large catchment (2-300 km2). These will be used to develop water harvesting design principles and ultimately 'thumb rules' for use by agencies involved in WSD. They will also allow evaluation of cumulative out-of-catchment impacts as WSD is scaled up.

The farming systems model APSIM is being parameterised using local data to explore long-term viability of alternative crop options, the costs and benefits of short-duration rice, and alternative irrigation strategies.

This project uses participatory action research to develop and promote principles for Watershed Development (WSD) for the high rainfall East India Plateau (EIP).

Key Findings 2007-2008
Process
Villagers continued to participate fully and with increasing enthusiasm in the action learning cycle (Plan, Do, Observe, Reflect). It is apparent that development intervention needs to go beyond engaging self-help groups (SHG's), to 'family-wise' planning that ensures all members understand their role and agree on what can be done and who will do it. This is to avoid further increasing the workload of women, to ensure equitable sharing of work, and to ensure thorough understanding of technical requirements.

Developing social capacity and improving agronomy (human capacity)
A Village Core Committee (VCC) was formed from the SHG's in Amagara in 2007 to improve project implementation. This shifted the ownership and responsibility for trial site selection and management from the team to the farmers. The high standard of management at trial sites and the reduced level of team support needed, point to a great success. Further development and wider evaluation of this process is planned, as it may support implementation of any complex intervention such as WSD. A mixed-gender group of 48 adults met in Amagara in May, 2008 to finalise plans for the kharif. It comprised the VCC and other interested farmers. The size of this voluntary group and the high level of participation indicate the growing importance attached by villagers to the planning process. This large group of poorly-educated subsistence farmers engaged in a business-like way in planning a complex research project, indicating development in their capacity to deal with complex issues and integrate them into village life. Similar observations have been made at Pogro. Involving the VCC has meant that training in agronomic skills provided by the team is now being taken up more carefully and more widely. This training is being supported by research findings on crop nutrition and irrigation and the continuing crop and fertility survey work. Improved crop management practices have been introduced to more than 100 farmers, and we believe substantial upskilling in agronomy has occurred.

Developing crop options and farming systems
- Soil and crop management survey (Pogro and Amagara).
Again there was no significant difference in rice yield between the land classes despite a trend towards lower yields in baid. In Pogro, rice yields were positively related to available P and mineralisable N (R2 = 0.6, P<0.05). There was no significant association between yield and exchangeable K at either location, despite generally low soil-K concentrations (<100 mg/kg). Management factors were also important.

- Split-field nutrition studies
Phosphorus - At Pogro, there was a 23% yield response to 30 kg/ha P across all land classes (P<0.01). In Amagara, the response to P was significant only in Baid land, but it was detectable despite lack of significant response overall in the fertility survey.
Potassium - At Pogro and Amagara there were 27% (P<0.01) and 7% (P<0.1) yield increases with added K, despite there being no apparent association between soil-K and yield in the soil/crop survey. More work is needed.
Zinc - There was no response to Zn at either location.
Nitrogen - Many of the high K and High P plots appeared to be N deficient.

The survey work is well received by farmers and raising awareness about sources of yield variation. With sufficient fields and measurements, relationships between soil attributes (eg P) and yield can be established, but imprecisely and at significant cost. Paired-plots are more efficient for identifying nutrient deficiencies, but they don't give the dose-response function. We are working on a composite approach to test in 2008. The alternative of a replicated P-response experiment at a single site is largely irrelevant in this highly variable landscape.

- Black gram
Three-fold yield increases were obtained with 15-20 kg P/ha, with a possible further small increase with added K. Nodulation was good.

- Rabi experiments at Amagara
Wheat and mustard failed without added P, but yields were good with up to 50 kg P/ha. One or two irrigations (each c. 40 mm) gave yields approaching full irrigation (4-5 irrigations), apparently by promoting use of residual water in the soil after rice.

Water resource management
Conceptual models have been developed to simulate the water balance at the scales of individual intervention, small catchment (1-2 km2) or large catchment (2-300 km2). These are being used to develop water harvesting design principles and 'thumb rules' for use by agencies involved in WSD, and will be used to evaluate cumulative out-of-catchment impacts as WSD is scaled up. The farming systems model APSIM is being locally parameterised to explore long-term viability of alternative crop options, the costs and benefits of short-duration rice, and alternative irrigation strategies.

This project uses participatory action research to develop principles and improved practices for Watershed Development (WSD) for the high rainfall East India Plateau (EIP). Villagers participate in an action learning cycle (Plan, Do, Observe, Reflect) which guides the overall project as well as most activities.
Key Findings
Process - It has emerged that villager's perceptions about themselves and their resources change, and activities succeed, when villagers genuinely control the implementation of projects, whether the resources are supplied by the project or by the villagers. This degree of control can be problematic for implementing agencies, but is nevertheless essential.
Developing social capacity and improving agronomic practise (human capacity)
The idea was extended from Amagara to Pogro of using a Village Core Committee (VCC) comprised of Self Help Group (SHG) representatives to improve project implementation and build social capacity, by shifting ownership, responsibility and control from the team to the villagers. In Pogro, the VCC oversaw (with project support) the initial implementation of the watershed development plan during the dry season of 2007/08 and its recent completion. Initial implementation focused on part of the watershed, including a 'learning cluster' of 6 families. The cluster is a focal point for the whole village and surrounding communities to 'action-learn' about the effective use of water resources.
Learning activities have been designed to develop agronomic skills (capacity) and transform farmer's self perceptions and perceptions about natural resources. As an outcome, in Amagara, a linear study of land use over time has revealed significantly increased cropping intensity and crop diversity. Case studies reveal major improvement in the livelihoods of individual families. In a turnaround in perceptions, agriculture is seen by the community in Amagara as a viable livelihood, with 80 of 146 Amagara families reporting income improved substantially, by at least Rup 15,000.
Gender studies have shown that family-wise action which engages with women as farmers, not merely as members of an SHG, builds individual self-confidence more strongly and exposes women to a wider set of ideas, apparently without alienating men.
Developing crop options and farming systems
Kharif soil and paddy management survey - 54 sites/fields over two watersheds were surveyed in 2008. Differences in fertility within and between land classes mirrors previous years. Mean yield was slightly less than previously because of lower yields in medium-upland, much of which succumbed to dry conditions late in the monsoon. Many drought-affected crops at Pogro had ample soil water to 'finish' the crop, so the possibility of low subsoil pH/Al toxicity is being examined, which may inhibit root growth and subsoil water-use. High rainfall early in the monsoon made N management more difficult, accounting for some of the variation in lowland yields.
Omission trial - nutrition studies - The nutrition-response work adopted an omission trial approach, in which the nutrients examined were P, K and Zn, with P at two rates. Unlike previous years, there was no response to either P or K at either watershed. Participating farmers remain convinced about the responses to P and K overall, explaining the lack of response in 2008 in terms of site selection/management and drought. Lack of nutrient responses made it impossible to evaluate the omission trial approach against earlier approaches tried in the project. A pot experiment confirmed that large responses to superphosphate are to the P, not S.
Alternatives to paddy rice
Upland, direct-seed rice succeeded at Amagara, despite a short monsoon. Mean yield was 2.4 t/ha. Yield variation (1.3-4.6 t/ha) was apparently related to soil water availability. We can now formulate recommendations that should allow for modest yields with low risk of failure in upland sites, with soil fertility maintained. Direct-seeded rice again performed poorly at Pogro, possibly because growth slowed with the onset of dry conditions, again suggesting the possibility of Al toxicity.
Black gram nutrition and weed management experiments failed because of late planting. This failure emphasises the need for farmers to perceive black gram as a premium crop and treat it accordingly once line-sowing and improved nutrition have been adopted. It has traditionally been treated as a low-yielding subsistence crop, which is broadcast.
Early-season vegetable production in Amagara has continued to expand although it is no longer a focus for the project. Most villagers now crop vegetables in the late monsoon, and into the rabi.
In the 2008/09 rabi at Amagara, mustard work focused on helping selected farmers to 'experience' further improved agronomy. Yields of participating farmers averaged an excellent 1.9 t/ha. At Pogro in the 'learning cluster', the focus was on improving extension processes, working with a group of families to evaluate options for using water 'harvested' from their new water harvesting structures. Credible yields of 4 t/ha from wheat and 1.6 t/ha from mustard were achieved. The learning process was documented and case studies documented of participating families.
A linear study of land-use and crop diversity at Amagara shows clearly that farmers have integrated information on new crop options, including their irrigation and fertiliser requirements, to develop much more intensive and diverse systems. The APSIM cropping systems model has been parameterised to support further exploration of farming systems options.
Water resource management
Observations of the climate, coupled with observations of water fluxes and storages (discharge from study site, pond water levels, as well as shallow and deep groundwater table) have been used to develop models simulating the hydrologic behaviour of selected interventions. This includes structures to hold water in upland areas as well as structures that can tap into shallow groundwater systems in the lower parts of the catchment. Based on these observations and models, guidelines to help with the design of watershed development work have been developed to assist the planner with determining what interventions would be suitable in specific areas, and estimating the volume of potentially 'harvestable' water. These guidelines have been applied to the Pogro catchment to develop a plan for intervention.

Participatory action research is being used to develop principles and improved practices for Watershed Development for the East India Plateau (EIP). Villagers participate in a learning cycle (Plan, Do, Observe, Reflect) which guides the overall project and most activities.
Community engagement processes for upscaling
PRADAN (Purulia) implemented a program to reach 5,000 families in 2010, starting them on the road to improved livelihoods through better use of land and water resources. This program applies project findings that (i) implementation improves when women are engaged equally with men (and continually re-engaged) in learning and deciding about improving livelihoods through agricultural innovation, (ii) substantial improvements in livelihoods demands radical changes in farming systems to better use natural resources and (iii) an adult-learning approach rather than demonstrations of technological packages facilitates complex change, fosters independence, and minimises NGO-dependency. Resource-poor families are being helped to envisage improved livelihoods through agriculture, plan new farming systems including a range of crop options, and develop locally-relevant practices that make better use their resources. This program takes radically new ideas arising from the project to large numbers of people using the resources normally available to an NGO.
Understanding the resources and developing crop options and farming systems
Rainfall was low across India in 2009. Good opening kharif rains from 30th June in Pogro (a project site) allowed cultivation and establishment of rice nurseries, but poor rainfall delayed transplanting until August. The monsoon finished early (late September). Most rice yielded poorly, except for earlier sown lowlands. Research focussed on consolidating learning and developing training and decision support material.
Soil fertility An experiment with potassium in 2009 across 24 farmer's fields confirmed earlier responses in rice at Pogro (16% yield increase with 30 kg K/ha), but only in the kanali and bohal (lowlands) that have long been used for rice and are lowest in exchangeable K.
Previously we reported unexpectedly severe P-deficiency in crops, including rice, along with high between-field variability. A decision support 'tool' has been developed for rice and several non-flooded kharif and rabi crops. It can be used with a soil test, but is best suited for use by farmers conducting fertiliser test strips. This is preferred because high field variability means each field needs to be tested, which is impracticable for poor farmers with several fields. A generic P response curve relates soil P to crop yield for different crops. Users determine where a field lies on the curve by measuring soil P or crop response to added P in a fertiliser test strip. The curve is divided into three zones, unresponsive (no need for P-fertiliser - but periodically conduct a test-strip), highly responsive (add 2x expected P removal by a crop, except to flooded rice - P removal is provided in a look-up table - because highly P-deficient soils 'fix' about half the P applied into unavailable forms, and marginally responsive (replace expected P removal by the crop). This approach puts the farmer in control of finding out about fertilisers.
Crop options Previously we reported on good results with directed-seeded upland rice (DSR) in Amagara. The potential for DSR to combat climate variability was revealed in a case study in 2009. This farmer planted DSR on early rain, whilst the interrupted monsoon delayed or prevented transplanting of lowland rice. The upland variety used with DSR was unaffected by the dry period following the early monsoon rain, and being quick maturing it was harvested by late September to return a good yield of ~4 t/ha. There was enough rain in early October to plant a second crop which matured on residual moisture. So the farmer grew two good crops on land that was once regarded as risky for rice. Other villagers harvested little or no transplanted rice on similar or better lands. However, DSR has not performed well at Pogro, and we suspect Al toxicity. Research into this possibility in 2009 failed to provide conclusive evidence.
Black gram is a potentially valuable crop on the EIP, but yields are low as a subsistence crop. In 2009, Pogro farmers participated in experiments to improve the agronomy, but on their initiative they treated non-experimental fields conventionally. All fields were affected by unusually severe virus disease, but the important message was that farmers are yet to perceive black gram as a cash crop. When this was discussed with farmers, they saw the value of planting small areas and managing them well, especially since market price is now >Rp 40 /kg. We are working further with farmers to change perceptions of black gram as well as to improve practices.
Vegetable production in Amagara continues to expand, although no longer a focus for the project. A linear study of land-use and crop diversity shows many villagers now grow vegetables from pre-monsoon through to the rabi, with some cropping in 2010 through to early May. Cropping into summer is risky biophysically and economically if there is insufficient water, and there is social risk associated with competition for community water resources. These risks need attention.
Farmer-learning at Pogro Here, farmers are at an earlier stage in development than Amagara. Work continued with farmers to learn how to use "harvested" water effectively for rabi crops - yields of mustard and wheat varied widely between fields, and much of this variation seemed to be associated with the effects of degraded soil structure and poor infiltration following rice. This limits the amount of water that can be applied in any irrigation without water ponding and risking poor germination and later crop damage. Following previous observations like this in Amagara, work commenced in 2009 on green manuring to address the soil problems following rice which are a barrier to adopting potentially more reliable and profitable non-flooded kharif or rabi crops on medium uplands.
A workshop to help farmers learn about their water resources was developed and tested in Pogro and Amagara.
Water resource management
Based on observations and modelled results, guidelines to help with the design of watershed development have been developed to assist the planner with determining what interventions would be suitable in specific areas, and estimating the volume of potentially 'harvestable' water. These guidelines were trialled with a group of PRADAN professionals. A new phase of the work commenced in 2010 with implementation of the intervention plan in Pogro, and the commencement of post-intervention hydrologic monitoring. New instrumentation has been installed to improve flow gauging at two catchment outlets at Pogro.

Participatory research is being used to develop principles and improved practices for Watershed Development for the East India Plateau (EIP). Villagers participate in a learning cycle (Plan, Do, Observe, Reflect) which guides the overall project and most activities. The main foci in 2010/11 were out-scaling project learning about agronomy and community engagement processes, through PRADAN (Purulia team), and finalising hydrologic work, especially rating control structures at the catchment outlets in Pogro watershed. A soil survey was conducted in Jharkhand to support out-scaling.
Rainfall was again low in East India, totalling only 553 mm at Pogro for the June-September monsoon period (average 1,200 mm), by far the driest year of the project. Little rice was transplanted across the region, and distressed migration started in September when all hope of a kharif crop had been abandoned. Despite the critical lack of water for transplanted rice, a soil water model developed in the project showed that available water for rainfed crops exceeded 50 mm from 10th July-29th October (Pogro weather data). This is sufficient to grow a crop without water stress, as evidenced by black gram, vegetables and direct-seeded (un-puddled) rice (DSR) grown in several project villages. This reinforces the finding that East India experiences 'drought' because of the susceptibility of transplanted rice to rainfall deficits and confirms that there are safe, productive alternatives, including DSR. This is the central theme of out-scaling.
Out-scaling PRADAN (Purulia team) aimed to reach 5,000 families in 200 villages in 2010. The out-scaling applies project findings that: (i) implementation improves when women are engaged equally with men in learning and deciding about improving livelihoods through agricultural innovation, (ii) significantly improved livelihoods demands radical change in farming systems and (iii) an adult-learning approach facilitates complex change, fosters independence, and minimises NGO-dependency. Resource-poor families were helped to plan new farming systems and develop locally-relevant practices to better use their resources. This applies project learning at scale, using resources normally available to an NGO, and allows for evaluation and improvement of the process.
Six PRADAN professionals were trained and community service providers (CSPs) were recruited from amongst the villagers and trained to assist with procuring inputs and providing day-to-day technical support to families. Representatives from women's self-help groups (SHGs) were formed into Agriculture Management Committees (AMCs) and trained to support the process of change. Participating families paid an Rs25 fee to the Federation of SHGs, partly covering costs of participation.
Planning meetings facilitated by PRADAN professionals were held with ~4,000 families, of whom 2,000 'signed up' and paid their subscription fee. Families developed their own plans from a range of options explained to them. As the kharif progressed, each Professional looked after 8-10 villages, meeting every 1-2 weeks with CSP's and AMC's to make a field visit and monitor implementation and assist where needed.
Initial enrolment in the program was encouraging, but it was difficult to maintain enthusiasm with so little rain and few opportunities to implement plans. However, a survey of 537 participating families (one-quarter of participants) found 45% of families had net earnings greater than Rs7,500 from the interventions, in a year when cropping largely failed across the region. Only 20% of families made less than Rs2,500 and 20% earned more than 15,000, exceeding our target of Rs8,000 /family.
Important learning for PRADAN was that cropping plans need to be more flexible to allow families to respond to the weather (and markets), and the process needs to be streamlined to demand less professional time and be amenable to Professionals with limited agricultural experience. The role of women needs nurturing. A revised process supported by newly-developed learning tools will be used in 2011.
Understanding the resources and developing crop options and farming systems As a basis for future out-scaling, a soil fertility survey in 8 watersheds (stratified by toposequence) sampled the range of soils and rainfall in Jharkhand. Results confirmed widespread P deficiency but with high variability between fields, even in the same toposequence position in the same watershed. Thus fertilizer prescriptions must lead to inefficient fertilizer use. Site-specific nutrient management is needed, but as soil testing is not feasible other tools such as omission trials are needed. K was commonly deficient but more work is needed to assess fertilizer response.
Water resource management A new phase of the work commenced in 2010 with implementation of the watershed plan in Pogro, and commencement of post-intervention hydrologic monitoring. New instrumentation was installed to improve flow gauging at two catchment outlets at Pogro. However, with the very low rainfall there was only a single small runoff event at Pogro, and on that occasion the instruments failed.

During 2005 we carried out research work and training at the Highlands Aquaculture Development Centre (HAQDEC). This facility is located at Aiyura and run by Officers of Eastern Highlands Province. It is nationally significant because it is the main fish hatchery in PNG. Also, research was carried out at laboratories in the University of Western Sydney, Hawkesbury campus. In summary, the activities that were undertaken relate to 4 of the 5 objectives i.e. Objective 2 (Develop improved on-farm feeds and fertilisers for Smallholder farmers), Objective 3 (Substantially increasing the fingerling supply from HAQDEC), Objective 4 (Trial aquatic bacteria in feed to improve fish nutrition) and Objective 5 (Extend the fish husbandry package and findings to farmers). Activities relating to Objective 1 (Develop a fish husbandry package for smallholder farmers) will commence in 2006. The progress made in 2005 is summarised as follows.

For Objective 2 (Develop improved on-farm feeds and fertilisers), at UWS a comparative study was undertaken on the impact of six (6) types of fertiliser on the bacteriology of water and sediment in extensive, freshwater aquaculture. The two key factors that were investigated were the a) effect of the fertilising regime on bacterial succession in fish ponds, and b) the nutritional value for fish of the benthic and planktonic biota for each fertiliser. The fertilisers were cow manure, chicken manure, mushroom compost, dried lucerne, dried oat hay, and soluble fertiliser (urea and superphosphate). Experiments were carried out in triplicate mesocosms (50 litre volumes) and repeated. Unfertilised mesocosms were used as a control. Two variations were studied in which results for a) high vs low level of each fertiliser, and b) new vs aged sediment were compared. Funding for this study was provided by an UWS internal competitive grant (50%) and ACIAR project funds (50%). The final collection and analysis of samples will be completed in February 2006. The data will then be fully analysed and a decision will be made on publication and the next direction in the investigation. It is hoped that the work will provide evidence for determining the most appropriate fertilising regime for extensive farming in new and aged fish ponds in PNG (i.e. for ponds without powered aeration and with low rates of water exchange). Data on types and numbers of bacteria, extracellular microbial enzymes, nutrient concentrations, biophysical measurements and survival & growth rate of fish is currently stored on Excel data bases. Draft reports are in preparation. Aquatic bacteria that have been isolated during the study will be used in Objective 5 in 2006.

For Objective 3 (increasing fingerling supply from HAQDEC), progress was made in all of the six activities listed in the Project document. The results will a) reduce bottlenecks at the hatchery stage of fingerling production (ie juvenile fish) and b) substantial increase fingerling supply to smallholder farmers throughout PNG. The enthusiastic officers at HAQDEC were assisted through visits by international experts - Dr Mohammed Akterazzuman from Bangladesh, Mr Satya Nandlal from Fiji, and Dr Geoff Allan, Mr Peter Graham and Mr Brett Herbert from Australia. Also, improvements were made in aspects of hatchery operations as a result of experiments performed by Project Officers (Ms Hopa Simon and Mr Wally Solato) and staff at HAQDEC. A summary of the activities and the results are as follows.

3.1 A major workshop, "The Hatchery Review of Operations at HAQDEC" was held from 7th to 11th November 2005 at Aiyura - the program is attached (Appendix 1). More than 20 farmers and officers from 6 Provinces of PNG participated. Two experts in breeding and culture of GIFT fish were invited from Bangladesh (Dr Md Akteruzzaman) and SPC in Fiji (Mr Satya Nandlal). They generously shared their knowledge with participants and provided instructive training and extension. A full day of panel sessions provided the Workshop with excellent notes that will form the basis of the Hatchery Manual that we are currently drafting. A method based on Hazard Analysis and Critical Control Points (HACCP) was used to identify key problems and solutions. The Hatchery Manual should be completed by mid-2006 and it will be tested and modified by staff at HAQDEC under operating condtions. It will also form the basis of a Smallholders Hatchery Manual that will be produced by the project and distributed by the Project throughout PNG at the workshops and through Co-operator farms.

3.2 Ms Hopa Simon, one of the project's Technical Officers, performed the experiments to determine the cause of mortalities of GIFT fingerlings during transport and to determine the appropriate protocols for ensuring high survivals during transport. Up until now, the mortalities for GIFT fingerlings have been averaging around 70-90% for trips of 3 to 4 hours, even though bags were oxygenated. It has been almost impossible to transport fingerlings to farms that are more than 12 hours from Aiyura. The research was designed to measure survival of fingerlings while testing a range of sedatives, chemicals, packing densities and water quality. A key finding was that the use of rainwater from storage tanks at Aiyura was a major cause of mortalities. By using low concentrations of salt in the rainwater, or using reservoir water, it was possible to reduce mortalities significantly. The further addition of low concentrations of sedative and the use of battery-operated aerators during harvest and transport was also very important. Ms Simon gave an excellent presentation of the results at the Hatchery Review and she is currently drafting a report on the study. We are planning to submit a manuscript to the journal of "Asian Fisheries Science" for publication.

3.3 Mr Wally Solato, our other Technical Officer, has been installing a system for managing separate broodlines for GIFT. Three families of fish have been collected from the original stocks that were distributed to regions of PNG. Although only a handful of the original GIFT survived in quarantine at Aiyura in 1998-2002, the offspring that are now spread throughout many farms in the regions of PNG, have fine characteristics. There is a good argument for introducing replacement families of GIFT, however the specimens that we have observed in PNG do not suffer from deformities or other characteristics of weak, inbred fish. So we will continue with our families of broodlines and gain training as we go.

3.4 Techniques for constructing a new water bore at HAQDEC have been applied to a low lying section of the Fisheries Station. Initial tests were performed in October and a second attempt was made in November 2005. Once the necessary pipe fittings are purchased, the bore should be pumped in. Hopefully this will be completed in early 2006. Assistance has been provided by the mechanic that we employed, Mr Steven.

3.5 The first training and extension workshops at HAQDEC for small-scale hatchery operators was held in November 2005 at the Hatchery Review, as outlined in 3.1 (above). At the Hatchery Review smallholder hatchery operators attended from 6 provinces and the manual that we are preparing will extend the skills and training at the future Workshops.

3.6 We have recorded the data on fingerling production and distribution by HAQDEC since 1980 for common carp and GIFT as well as for Erap since 2003 for GIFT. We have not yet made accurate records of the number of fingerlings distributed by smallholder hatcheries, however this will be attempted in 2006. The data has been entered into an Excel spreadsheet and was transmitted to NFA in August 2005 for their databases. The database will be updated every 6 months.

For Objective 4 (Trial aquatic bacteria in feed to improve fish nutrition) one activity has been undertaken. Aquatic microbes have been isolated and selected for testing on feeds. So far the bacteria have been cultured with a broth based on oaten straw. The growth and survival rates of atremia have been remarkable. The assessment of the bacteria's capability to improve the nutritional status of this low-value feed is being assessed. This work is on-going and testing will expand in 2006. Funding was provided by UWS (50%) and ACIAR (50%) in 2005.

For Objective 5 (Extend the fish husbandry package and findings to farmers), some skills and knowledge has been provided to the industry at the Hatchery Review (as outlined above in 3.1). At this stage only a small number of smallholder farmers benefited, however the accumulation of extension material has begun.

This is a project with a national focus and in the last 12 months training and extension of skills have been given to smallholder farmers, NGOs and government officers from a range of Lowland Provinces at two project Workshops at Erap's Aquaculture Centre.

The major events that have occurred during the last 12 months are as follows.
At Erap (Lowland Aquaculture Facilites of NDAL):
1. The TOs, Billy Kerowa and Bangan John were new recruits and they started in February 2006 working under Johnney Soranzie, the OIC at Erap Aquaculture Centre. Unfortunately Bangan John left soon after starting and has not been replaced. Billy Kerowa has been learning the skills for operating fish ponds at Erap and assisting Johnney in the construction of a GIFT hatchery at Erap. Billy travelled to Aiyura for further training - particularly in using pH, Dissolved Oxygen, salinity meters and loggers.
2. The Co-operator farms. Johnney Soranzie selected 10 co-operator farms and has supplied them with tools (shovels, etc). He held the first introductory workshop at Erap in February 2006. This went well. He needs a reliable vehicle in order to maintain visits to farms and this is under discussion.
3. Farm-based workshop. Johnney also organised the first farm-based workshop in August 2006. At the workshop the farmers carried out activities that are part of the "Halpim long pis farma" (i.e. fish husbandry package).
4. Feed work. Johnney visited Project Support Services (PSS) to obtain information on feed machines for Geoff Allan's ACIAR project. Johnney and other team members attended ACIAR's feed workshop at Aiyura in November 2006 and received training on making pellet fish feed.
5. Siot long wokim (Shirt for work). The Lae soccer team participated in the program - the team carried out activities at Erap station and after the work, the team received the ACIAR sponsored football shirts.

At Aiyura (Highland Aquaculture Development Centre of EHP Government):
1. The TOs, Hopa Simon and Wally Solato, worked on feed trials and broodstock respectively. Hopa received training at a Feed workshop in Thailand and Wally received training at SPC in Fiji on broodstock. Wally prepared 5,000 fingerlings for the SPC-ACIAR project in Yonki Reservoir. Unfortunately, Hopa left the project in about September-October 2006 after suffering from Typhoid and a number of other problems. Hopa has not been replaced yet and it is still not clear whether she will be returning to the project..
2. Yonki cage culture. Awepstar Seka, the officer in charge returned to Goroka and EHP administration worked out a plan for meeting its own needs and the commitments to ACIAR for Aiyura and Yonki, especially the cage culture project. Wally Solato assisted Peter Minimulu and Jacob Wani with Cathy Hair's ACIAR-SPC cage culture project.
3. Management and Infrastructure. Kine Mufuape, the OIC at HAQDEC was suspended by EHP Government in June 2006 and replaced with Mr Silas. Prior to this, Kine contracted a carpenter to build a smoke house for drying feed and fish. Also he contracted an electrician to complete the supply of power to all houses, including the new meter boxes for the field workers and their families. The water tank was picked up from Lae after the Canter truck was re-registered and repaired. The plumber has connected the water supply to houses however the water supply is still insufficient. The poor water supply at the HAQDEC aquaculture station as well as the whole of the Aiyura valley is very serious. The Technical Advisor, Mr Nephion Tarapi was successful in getting funds from EHP Government to improve the water supply to HAQDEC and the AVI, Mr Kevin Lange has been assisting with implementing the water plan. The new clutch for the Landcruiser was purchased and installed. The new computer was faulty and returned to Port Moresby for repair. It was returned to Aiyura but there is no email connection in operation. The phone/fax machine at HAQDEC was out of service for most of the year. Jacob Wani, Peter Minimulu, Mr Silas and the team at HAQDEC assisted with the Feed Workshop that was conducted under ACIAR's Feed initiative.
4. Feed work. Hopa Simon carried out excellent work in 2005 to identify the cause of fingerling mortalities during transport. Also she started investigating the usefulness of seeds from Elephant grass as a feed for GIFT fish in a 100-day experiment using triplicates for a range of treatments. She ran the experiments in the 12 experimental ponds - triplicates for 3 treatments and a control. The experiment was not completed at the time Hopa left her position. Unfortunately while she was ill with Typhoid, the grass that she was using was cut down and the experiment needs to be repeated. Further work on a whole range of "nil cost" feeds are planned to be carried out during the project once the team can meet and make a plan. The team received training for preparing pellet feeds using the feed machinery from Geoff Allan's project. Funds are needed to purchase ingredients and maintain this important initiative.
5. Farm-based workshop. The first farm-based workshop for the highlands was organised by Kine Mufuape and Hopa Simon for July 2006 in Simbu Province, however this was cancelled because of the re-structure of administration of HAQDEC. Efforts are being made to organise a new workshop in the Highlands. The Project Leader travelled to the highlands in October 2006 and selected 3 Co-operator farms in the Goroka area in an effort to encourage the team to select the remaining co-operator farms. The remaining Co-operator farms still remain to be selected by the team.
6. Siot long wokim (shirt for work). There are two football teams in Aiyura wanting to participate in the program - a boys and girls' team. Previously Kine Mufuape distributed a selection of second-hand shirts brought up by the Project Leader in the March trip as a trial. There were a few problems, but now that we have sorted that out, the next time will be with proper ACIAR shirts. Shirts have been purchased and emblems have been added to the shirts. The shirts are to be taken to the highlands for the next workshop.
7. Kevin Lange, the AVI from Tasmania, arrived in March 2006 and has settled into his accommodation at Aiyura High School. He is assisting the team with water installation, repair of vehicles, general back-up and technical advice.

In April, 2007 the project team met and reviewed the project. As a result, the distribution of remaining funds and objectives were revised. Since then, the team has made substantial efforts to get back on track. We welcome the return of Ms Hopa Simon to the position of Technical Officer at Aiyura. Mr Wally Solato successfully moved from Aiyura to Goroka and is now working with NDAL. Mr Billy Kerowa and Mr Douglas Kawa both continue at Erap. The Australian Volunteer, Mr Kevin Lange, who was stationed at HAQDEC in Aiyura for two years (2006-8), completed his stay in PNG in March 2008.

All Technical Officers have prepared workplans for their station and off-station work. These workplans vary from officer to officer - topics include studies with formulated fishmeal-based feeds, GIFT broodstock production trials, fingerling trials, polyculture and trials with sex-reversal of GIFT fish. The feed-manufacturing machines arrived at all stations (Goroka, Erap, Port Moresby, Banz and Aiyura) and the Officers are working to get them installed and, in some cases, they are now successfully producing pelleted fishmeal-based feed.

The publication by ACIAR of our book: "Aquaculture in PNG: Status of freshwater fish farming" was printed and distributed by ACIAR. Farmers who participated in the study, NGOs, Officers and all interested parties received copies. The monograph is available on the ACIAR website.

A strategic review of the operations of the Highland Aquaculture Development Centre (HAQDEC) at Aiyura, was carried out by the team in July 2007. This review gave the staff at Aiyura a chance to express themselves to the project. Judging by follow up trips in October 2007 and May 2008, the staff at Aiyura is working well towards substantial improvements. Major improvements include: repair of the inlet canal to the water reservoir and replacement of timbers of monk - now the reservoir is full for most of the year.

Two good second-hand vehicles were purchased for the project and are operating in Morobe and Eastern Highlands Provinces. The vehicles were paid for from the project's funds. Fuel, registration and maintenance are provided by the project and at the conclusion of the project in 2009, the two vehicles will become the property of NDAL.

The project has extended its "Fish Husbandry Package" at four workshops at demonstration farms have trained 250 new farmers in the last 12 months. The workshops are based on getting the farmers to carry out activities over three days. The team brings GIFT fingerlings (and common carp) so that the farmers can stock ponds once they have constructed them (usually 3 to 6 months later). Workshops were held as follows:
February 2007 - Workshop Avini Village, Koruo Fish Farm (80 fish farmers attended).
November 2007 - workshop at Asai village, Madang Province (75 attended)
November 2007 - Bihute workshop, EHP (35 prisoners attended)
May 2007 - Bihute Training workshop EHP (40 prisoners & officers attended)
November 2007 - Banz workshop, WHP (52 attended) (5 days workshop)
February 2008 - Kundiawa workshop, Simbu 75 attended
November 23 - 26th 2007 - Faiyantina Workshop (33 participants)
February 2007 - Workshop Avini Village, Koruo Fish Farm (80 fish farmers attended).

The farmers are very keen to learn at these workshops - they always ask for a return visit in 12 months to reinforce and renew their training so that they get more skills and confidence. The farmers in greatest need are in remote regional areas away from the big towns. Some farmers walk for 2 days to attend. We have found that the workshops have a high impact and within 12 months, the farmers generally have a surplus of GIFT fingerlings.

The project has started working with 20 co-operator farms - 10 in the highlands and 10 in the lowlands. Preparations are being made for off-station work with trial feeds and other aspects of the "Fish Husbandry Package". While the project's activities at Co-operator farms are yet to reach our planned goals, there were three visits by officers to initiate the work at the farms, as follows:

June 2007 - visited 10 lowland co-operator farms
July 2007 - visited 10 highland co-operator farms
September 2007 - visited 10 highland co-operator farms

At the University of Western Sydney (UWS), Australia, an experiment on polyculture (carp and redclaw) was carried out - it showed that this could be a useful culture system for reducing the over-breeding of prolific breeders (such as redclaw crayfish and GIFT). This may have application in PNG - Billy Kerowa started work to apply the results to polyculture with common carp and GIFT fish. Also at UWS, several strains of bacteria were isolated which have interesting enzyme profiles which may be useful for aquaculture.

ASSESSMENT OF THE RISK OF INTRODUCTION OF FMD IN INDONESIA AND IMPROVEMENT OF EARLY DETECTION
Questionnaires have already been administered by Dr Pebi Suseno to Quarantine officers and members of the Animal Health Division to identify potential routes of introducing FMD into Indonesia. This includes both identification of geographical regions and evaluation of the likelihood of infection from legal and illegal sources of animal and animal product movement. This information together with livestock numbers has been used to identify the 10 provinces with the highest risk of having an introduction of FMD. Five villages from each province will be selected for strategic (targeted) sampling for FMD. Two workshops with Quarantine Officers and members of the DGLS were held in 2006/07 to discuss risks of FMD, the consequences of introduction and to identify deficiencies in knowledge and awareness of the disease by field staff and farmers. A presentation to the annual conference of the Indonesian Veterinary Association was made on the risks of FMD. A deficiency in current material suitable for farmers on the dangers and signs of FMD was identified and existing material will be updated during 2007/08.
SURVEILLANCE PROJECT- ALOR ISLAND
Important background information collected:
To date the Australian research team has made three trips to Alor and Kupang. There was a delay in commencing this component of the project because of workloads by staff in Kupang and Alor. During the last 10 months essential background information has been collected in relation to the role and function of the Department of Agriculture on Alor and the importance of livestock to farmers. It has become evident that although there are strong procedures in place to provide animal health services- the lack of resources, and interestingly farmer perceptions about the importance of livestock health, make providing adequate services difficult. A two-day workshop was organised on CSF and diseases of pigs in Alor and animal health workers were involved in a training session on collection of blood samples from chickens and pigs.
It is apparent that there are difficulties not only associated with the ability of remote villages to effectively contact the Agriculture Department but also with the ability of the Department to respond promptly and provide the necessary resources for a disease problem. There is also reluctance for many farmers to notify the Department of instances of disease and death in their livestock.
There has been a vaccination programme for CSF in place on Alor since 2002. The aim of this program is to vaccinate every pig on Alor yearly, in the hope of achieving eradication in the future. Unfortunately, some farmers continue to refuse the vaccination of their pigs.
Although there is no denying the importance of animals such as pigs in the culture of the people of Alor - it would seem that the farmer's main concern is the mere existence of the pig regardless of body condition and/or health. The pig functions as an asset and is traded for money to pay for such things as a child's education. It is also slaughtered for religious ceremonies, special events and parties. An issue highlighted by one of the volunteers from a farmer's NGO group we visited during one of our visits was that there is no market system in place in Alor. Farmers therefore are not motivated to produce the largest, healthiest pig for market. Animals are thus mainly only kept for private consumption or as a long term asset. Crops on the other hand provide daily returns and are probably the mainstay of the farmer's income, and as a consequence receive the most time and effort. This could have implications for the effectiveness of syndromic surveillance given that the farmers seem to pay little attention to the health of their animals and spend the majority of their day tending to their gardens and crops.
Survey
Thirty seven villages have been selected and demographic data collected. Questionnaires and instructions for the cross-sectional, cohort and socioeconomic studies have been designed. The questionnaire for the socioeconomic study has been pre-trialled and refined and the questionnaire for the cross-sectional study is currently being trialled. It is hoped that the socioeconomic and cross-sectional study will get underway shortly. The logistics associated with the cohort study and vaccine trial is currently being worked on; this aspect of the project will commence subsequent to the onset of the cross-sectional study.

In light of the background information collected, incorporated into the questionnaires, are targeted questions pertaining to identifying reasons for why farmers often fail to contact the Agricultural Department in the evident of disease or death in their livestock, and also why some often object to the routine vaccination of their pigs against CSF. Once this information has been collected it is hoped that we will be able to develop incentives aimed at improving this situation.

A framework for a national surveillance system for foot and mouth disease (FMD) in Indonesia was established at a workshop organised as part of this project and attended by staff from Government, University and Research institutions. Technical guidelines have been drafted for this surveillance. As part of the improved surveillance system a field investigation of a suspect case of FMD was investigated and samples collected and tested by the national FMD laboratory (PUSVETMA). All samples were test-negative and the results highlight the success of having a national integrated surveillance system.
Another important outcome of the work has been in relation to incentives. Villagers were asked what type of incentives would encourage them to report animal diseases and a high percentage stated that more information about diseases was necessary. They also indicated that free vaccinations and free vitamin injections would also encourage them to report to the government officials about diseases. An outcome of this research was the need to explore the decision making processes of villagers. An understanding of how decisions are made about animal disease identification and reporting will be important towards the development of a surveillance program on Alor Island and will form the basis of the upcoming years work.
A pre-trial of the questionnaires for the cross-sectional study, prospective cohort study and vaccination trial for classical swine fever (CSF) was conducted in July 2007. Active surveillance (sampling) for CSF has subsequently commenced. The cross sectional study is complete and in total 690 samples were collected. The cohort study is on-going and to date 529 samples have been collected. A field based vaccination trial of 4 commercially available CSF vaccines has also been instigated and to date 1526 samples have been collected. A valuable component of the cohort and vaccination studies has been collecting data on changes in animal numbers and animal movements over time. Knowledge of movement of animals is essential to understand disease dispersal and to design potential methods of disease control. Questionnaires have been administered to farmers involved in these studies and the results of these will aid in further understanding farmers actions and help in designing effective surveillance systems for remote areas.

The AYAD student, Michael Bragg, was located on Alor Island for 5 months in 2007 and during this time, studied 16 villages across Alor Island. This allowed a better understanding of the importance of livestock in villages on Alor and enabled Michael to successfully complete his Honours degree at the University of Western Sydney with second class honours. This analysis provided valuable insights into the importance of livestock to farmers and the day to day functioning of villages in rural Alor. Although pigs are culturally important to most villagers, it was found that they had little concern for disease in their pigs and were more concerned about their family and crops. Pigs were less important for day to day consumption in comparison to crops, as pigs were mainly eaten at cultural events.

Guidelines for a surveillance program for FMD has been established and incorporated into a contingency plan in the event FMD is detected in Indonesia. A quantitative risk assessment was conducted identifying high risk areas for targeted sampling in Indonesia and the INDOVET PLAN for FMD was updated and results presented to the 29th World Veterinary Congress in Vancouver, Canada in July 2008.
The cohort study for Classical Swine Fever (CSF) in Alor is on-going and is due to be completed in October 2009. Of the 300 pigs initially involved in the study, 103 remain and 954 blood samples have been collected over five sampling periods.
The vaccine trial for CSF is due for completion in September 2009. Of the 300 pigs initially involved and vaccinated with one of four vaccines or a control, 101 remain and 1758 blood samples have been collected from eight sampling periods.
Our research has determined that syndromic surveillance cannot be effectively implemented on Alor given the current situation. For syndromic surveillance to be effective, farmers must first recognise signs of disease and then report this to an animal health authority. Our research identified a multitude of factors associated with farmers failing to report disease including a lack of: awareness of disease signs, incentives to report and access to animal health officials. On Alor farmers spend most of their time tending to their crops, which are located up to five kilometres from the households where livestock are kept. As a result, farmers spend very little time feeding, observing or caring for their pigs, even though they have significant cultural importance. The low priority placed on pigs together with a severe lack of education relating to pig health, husbandry and nutrition, culminates in the farmer's inability to recognise signs of disease. If farmers do recognise disease it is often difficult for them to access an official to report to, and there is no incentive to do so. Only two villages have an animal health worker (AHW) in permanent residence; farmers in other villages may have to travel a day or more to reach one. Because travel is costly and diverts time from income-earning activities, farmers are unwilling to spend time travelling to report diseases, especially because they receive no perceivable benefit from doing so.
There is no conventional livestock market system in Alor nor is there an abattoir for pigs. Pigs are sold and traded between friends, family and neighbours and consequently the movements of animals go unrecorded and uncontrolled. This combined with the failure of farmers to report livestock diseases, means that disease surveillance currently relies on active monitoring by authorities. However, due to monetary and personnel constraints this system is ineffective and expensive. Consequently, for surveillance to be improved it must be acknowledged that the farmer is an important source of information and an effort must be made to improve their knowledge and reporting levels.
By addressing the key factors affecting disease reporting, the situation may be improved to a point where syndromic surveillance is possible on Alor. However we are of the opinion that syndromic surveillance could work in other regions with more intensive livestock systems. In Alor we identified two main incentives that could be used to encourage farmers to report diseases. Firstly, farmers have a strong desire to improve their knowledge on livestock issues. By offering training to farmers on livestock health, husbandry and nutrition the farmers' ability to detect and prevent disease/poor performance will be improved and they may be more motivated to report disease. We are conducting research to determine information delivery techniques appropriate for Alor. Secondly, farmers have a strong desire to have healthy pigs. Most farmers interviewed stated that they liked their pigs receiving injections because it made them healthy. Hence, offering timely and effective treatments to sick pigs could also motivate farmers to report disease. However, access to treatments and vets/animal health workers remains an issue. In one village with a permanent AHW all nine farmers interviewed told the AHW when they had a sick pig. In another village, five of seven farmers travelled five kilometres to a neighbouring village to report livestock diseases to the AHW. Providing access to an AHW in every village or cluster of villages could greatly improve disease reporting rates, however this is expensive and alternatives are being investigated.
The improved plan for the surveillance of FMD in Indonesia is of significant benefit to Australia and helps maintain confidence in Indonesia's status of freedom from FMD. If all of Australia's immediate northern neighbours (Timor Leste, Indonesia and Papua New Guinea) are free from FMD it reduces the risk of disease incursion into the northern region of Australia. Knowledge of the Indonesian situation for CSF, which has been exotic to Australia for over 40 years, is also of benefit in developing appropriate surveillance measures in the northern part of Australia.