Year 3 (01/01/2003-31/12/2003)
Indonesian project members have endeavoured to continue the establishment and maintenance of the field trials at Soppeng and Ladongi. Progress has been slow, especially at Soppeng, because of difficulties in getting the farmers on whose trees the test clones have been grafted to follow advice about pruning of the mother trees to allow full establishment of the grafts.

The July 2003 review highlighted the progress made in the project in (i) forming linkages between ICCRI, BRIEC, BPTP Kendari, PT Effem and the SUCCESS Project in Indonesia, and between Indonesian scientists and their counterparts in PNG, (ii) training and giving experience to two younger scientists from BPTP and a young cocoa breeder from ICCRI, (iii) establishing collections of 30 local and 30 international cocoa clones ready for screening for resistance and quality, and (iv) developing methods for on-farm selection and screening of improved cocoa types. Reviewers noted the difficulty encountered in developing fully replicated and established screening trials on farms.

Progress on increasing the number of replicate grafts in the Soppeng trial has been slow (only 95 new grafts have been made over the last 6 months); 30 out of the total of 48 clones in the collection at this site still need further replicate grafts to reach the required number of 10 replicates. Farmers involved in these trials are reluctant to conduct vigorous pruning of the mother tree to allow full establishment of the grafts. As long as the grafts get established as extra branches on the mother trees, they may still produce enough pods to allow some of the assessments that we have planned. In December 2003, a farmer cooperator conducted severe pruning of 50 trees he has side grafted on his farm. It is hoped that this may be a good demonstration to the other farmers that side grafting and pruning will result in a productive new tree, thus pushing them to allow the full establishment of the genotypes on their farms. VSD was found to be quite common at Soppeng and the scientist took some infected twigs back to BRIEC to isolate the fungus.

Progress in the trial at Ladongi has been more positive. During the last 6 months of 2003, 160 successful grafts were made, to bring to 28 the number of clones with at least 10 replicates (leaving 20 clones still needing further grafting). Local partners are now skilled at doing the grafting, although the follow-up careful removal of the bags and nursing of the grafts is still a problem because they visit the trials only monthly. One farmer cooperator has now been shown how to nurse the grafts and hopefully this will facilitate the complete establishment of that trial. Also, local Dinas Perkebunan staff will help with the follow-up work. The two participating farmers at Ladongi are prepared to do the heavy required pruning of the mother trees.

The third trunk injections for the phosphonate trials in Soppeng and Ladongi were done in July 2003 and results of harvests have been collated.

1. Collection and resistance screening of material and assessments of quality characteristics of cocoa lines
Field evaluation procedures were finalised and data sheets for field evaluations prepared. Evaluations have been conducted every fortnight since mid-2004, by BPTP staff at Ladongi and Dinas Perkebunan staff (recruited to conduct fortnightly evaluations) at Soppeng. So far only 24 clones at Ladongi, and 15 at Soppeng, have produced a sufficient number of pods (8 or more in 17 evaluations) to enable an assessment of their comparative resistance to CPB and PPR. Data obtained on CPB-incidence for ripe (harvestable) pods in Ladongi and the proportion of infested pods that are heavily or moderately-heavily infested suggest that the 5 most resistant clones (out of 24 assessed at Ladongi) are RCC 73, Haris 1, Aryadi 2 and 3, and Darwis 3. Other promising clones are Catongo, VSD 1 Ladongi, BAL 209 and KKM 22. These evaluations are continuing on a fortnightly basis. One of the most important characteristics for cocoa growers and processors is the pod value (no. pods per kg dry beans). The pod values for most of the 24 clones assessed in the Ladongi trial indicate a substandard yield of beans per pod. The lowest pod values were obtained from the international or Jember clones: BAL 209, DRC 16, KKM 22 and RCC 73. It is worth noting that RCC 73 has also had a consistently low CPB incidence and a light infestation so is potentially a CPB-resistant clone that also has a relatively low pod value. Clones such as Catongo and Haris 1 show promise of partial resistance to CPB but also have very high pod values. The extremely high pod value of 'Catongo' reflects its very small pod size and suggests that this clone has been misidentified. Its pod value is more characteristic of Scavina clones than the authentic Catongo clone, which has a good yield. This example indicates the need for molecular characterisation of Indonesian collections (see below). Over the same evaluation period, PPR incidence was 0% for DRC 16 (already recognised as PPR-resistant), ICS 13, VSD 1 Ladongi, and Haris 1. Very low levels of incidence (less that 8%) were also found for Darwis 2, VSD 3 Ladongi, BAL 209, RCC 73, Haris 2, and Aryadi 3. In this evaluation period, the most susceptible clones to PPR (with PPR incidence rates up to 30%) were Amiruddin, Baharuddin, Pp Ladongi, Aryadi 2, PBK 1 Ladongi and VSD 2 Ladongi. VSD assessments have been conducted every four months. The data collected indicates a particularly low severity of VSD infection, therefore high resistance to VSD, for DRC15, KA2106 (both already recognised as VSD-resistant) and Catongo. However, the high pod value of Catongo suggests its resistance characteristics will be of the most use when incorporated by breeding into genotypes with lower pod values. In the Soppeng trial, pod production has lagged behind Ladongi. As reported last year, project team members tried to persuade farmers at the Soppeng trials to adequately prune the mother trees in order to allow the side grafts to fully establish. Compensations were paid to the farmers for any loss of production due to pruning. Farmers have responded by pruning at the trial site but only to a light level. Therefore, inadequate sunlight is reaching foliage and this has inhibited pod production on the grafted cocoa clones. The project team is continuing to talk to farmers about this problem. Quality analyses were conducted by PT Effem for the clones that had produced a sufficient quantity of beans. The results of these analyses indicate that these clones have substandard quality characteristics. This would suggest that, should clones prove to have resistance to CPB, PPR or VSD, they would need to be backcrossed with clones of good quality (ie with high fat content etc).
Links with the SUCCESS Alliance project (a USDA funded project administered by ACDI/VOCA*) were strengthened during the year to foster wider evaluation of promising screening material and recommendations for cultural management. A MOU was signed between the SUCCESS Alliance project and La Trobe University. New joint genotype-testing trials were established in Southeast Sulawesi at three ACDI-VOCA training sites in different parts of Kolaka District. Blocked trials with 3 replicates have been established at each site to test 15 clones (side-grafted onto mature trees in 6-10 tree plots). Very dry conditions in the latter half of 2004 caused a high mortality of side-grafts. Consequently, it has been necessary to allow time for more grafting budwood to be produced at the ACIAR trial sites so that grafting can continue to complete at least 6 trees in each plot.
2. Uptake of cultural practices to minimise disease and pest risks. A two-day training workshop was conducted in Kolaka District, Southeast Sulawesi for farmers, extension officers (including ACDI-VOCA officers). The workshop covered pest and disease management for smallholders. It included opening speeches, presentations and field demonstrations. Participants enthusiastically welcomed the opportunity to hear about approaches to pest and disease management from the trainers including project staff from ICCRI and Masterfoods. Topics covered by the workshop were management strategies for CPB, PPR, canker and VSD, and rehabilitation of farms by grafting, especially chupon grafting which is now routinely practiced in Brazil. Over 80 participants attended on each day including provincial Dinas Perkebunan officers, ACDI-VOCA* field technicians and trainer farmers, smallholders, BPTP staff and Kendari town council staff. New trials established jointly with SUCCESS Alliance have also encouraged the uptake of cultural farm practices. Farmers participating in the SUCCESS Alliance training programs have been doing the grafting and the assessments of graft survival. This program aims to encourage farmer-centred research and the linking on-farm research with extension and training objectives. As well as having the research aims of retesting clones selected from the ACIAR trials, the new trials aim to introduce methodologies of farm improvement and pest/disease management to smallholders.
3. Enhancing the infrastructure and capability within Indonesia for the collection, maintenance and screening for pest and disease resistance of cocoa genotypes During 2004, project staff from La Trobe University visited Indonesia in March, August-October and November 2004. Our project team member from Masterfoods Australia also made regular visits to Sulawesi. These visits enabled the project staff to address issues such as pruning at the ACIAR trials sites, the procedures for evaluation of the clones (some of which are now producing pods) and also to run a field workshop for training of extension staff and farmers in Southeast Sulawesi (see above). BPTP SULTRA staff were trained in evaluation of clonal performance by the Australian project staff and PT Effem. Since mid-2004, they have been actively commencing the evaluation of the trials. ICCRI staff have developed procedures building on their experience for assessing clones for CPB and PPR resistance and characterisation of the clones. Some of the clones included in the ACIAR trials are probably mislabelled, a common problem in cocoa research around the world - there is a possibility that USDA could assist in analysis of samples sent from ICCRI for genotypic identification providing appropriate material transfer agreements are in place.
(* ACDI/VOCA resulted from the 1997 merger of US NGOs - Agricultural Cooperative Development International and Volunteers in Overseas Cooperative Assistance. The merger blended ACDI's systemic, long-term approach to development and VOCA's people-to-people volunteer activities.)

Year 3 (01/01/2003-31/12/2003)
Indonesian project members have endeavoured to continue the establishment and maintenance of the field trials at Soppeng and Ladongi. Progress has been slow, especially at Soppeng, because of difficulties in getting the farmers on whose trees the test clones have been grafted to follow advice about pruning of the mother trees to allow full establishment of the grafts.

The July 2003 review highlighted the progress made in the project in (i) forming linkages between ICCRI, BRIEC, BPTP Kendari, PT Effem and the SUCCESS Project in Indonesia, and between Indonesian scientists and their counterparts in PNG, (ii) training and giving experience to two younger scientists from BPTP and a young cocoa breeder from ICCRI, (iii) establishing collections of 30 local and 30 international cocoa clones ready for screening for resistance and quality, and (iv) developing methods for on-farm selection and screening of improved cocoa types. Reviewers noted the difficulty encountered in developing fully replicated and established screening trials on farms.

Progress on increasing the number of replicate grafts in the Soppeng trial has been slow (only 95 new grafts have been made over the last 6 months); 30 out of the total of 48 clones in the collection at this site still need further replicate grafts to reach the required number of 10 replicates. Farmers involved in these trials are reluctant to conduct vigorous pruning of the mother tree to allow full establishment of the grafts. As long as the grafts get established as extra branches on the mother trees, they may still produce enough pods to allow some of the assessments that we have planned. In December 2003, a farmer cooperator conducted severe pruning of 50 trees he has side grafted on his farm. It is hoped that this may be a good demonstration to the other farmers that side grafting and pruning will result in a productive new tree, thus pushing them to allow the full establishment of the genotypes on their farms. VSD was found to be quite common at Soppeng and the scientist took some infected twigs back to BRIEC to isolate the fungus.

Progress in the trial at Ladongi has been more positive. During the last 6 months of 2003, 160 successful grafts were made, to bring to 28 the number of clones with at least 10 replicates (leaving 20 clones still needing further grafting). Local partners are now skilled at doing the grafting, although the follow-up careful removal of the bags and nursing of the grafts is still a problem because they visit the trials only monthly. One farmer cooperator has now been shown how to nurse the grafts and hopefully this will facilitate the complete establishment of that trial. Also, local Dinas Perkebunan staff will help with the follow-up work. The two participating farmers at Ladongi are prepared to do the heavy required pruning of the mother trees.

The third trunk injections for the phosphonate trials in Soppeng and Ladongi were done in July 2003 and results of harvests have been collated.

1. Collection and resistance screening of material and assessments of quality characteristics of cocoa lines
Field evaluation procedures were finalised and data sheets for field evaluations prepared. Evaluations have been conducted every fortnight since mid-2004, by BPTP staff at Ladongi and Dinas Perkebunan staff (recruited to conduct fortnightly evaluations) at Soppeng. So far only 24 clones at Ladongi, and 15 at Soppeng, have produced a sufficient number of pods (8 or more in 17 evaluations) to enable an assessment of their comparative resistance to CPB and PPR. Data obtained on CPB-incidence for ripe (harvestable) pods in Ladongi and the proportion of infested pods that are heavily or moderately-heavily infested suggest that the 5 most resistant clones (out of 24 assessed at Ladongi) are RCC 73, Haris 1, Aryadi 2 and 3, and Darwis 3. Other promising clones are Catongo, VSD 1 Ladongi, BAL 209 and KKM 22. These evaluations are continuing on a fortnightly basis. One of the most important characteristics for cocoa growers and processors is the pod value (no. pods per kg dry beans). The pod values for most of the 24 clones assessed in the Ladongi trial indicate a substandard yield of beans per pod. The lowest pod values were obtained from the international or Jember clones: BAL 209, DRC 16, KKM 22 and RCC 73. It is worth noting that RCC 73 has also had a consistently low CPB incidence and a light infestation so is potentially a CPB-resistant clone that also has a relatively low pod value. Clones such as Catongo and Haris 1 show promise of partial resistance to CPB but also have very high pod values. The extremely high pod value of 'Catongo' reflects its very small pod size and suggests that this clone has been misidentified. Its pod value is more characteristic of Scavina clones than the authentic Catongo clone, which has a good yield. This example indicates the need for molecular characterisation of Indonesian collections (see below). Over the same evaluation period, PPR incidence was 0% for DRC 16 (already recognised as PPR-resistant), ICS 13, VSD 1 Ladongi, and Haris 1. Very low levels of incidence (less that 8%) were also found for Darwis 2, VSD 3 Ladongi, BAL 209, RCC 73, Haris 2, and Aryadi 3. In this evaluation period, the most susceptible clones to PPR (with PPR incidence rates up to 30%) were Amiruddin, Baharuddin, Pp Ladongi, Aryadi 2, PBK 1 Ladongi and VSD 2 Ladongi. VSD assessments have been conducted every four months. The data collected indicates a particularly low severity of VSD infection, therefore high resistance to VSD, for DRC15, KA2106 (both already recognised as VSD-resistant) and Catongo. However, the high pod value of Catongo suggests its resistance characteristics will be of the most use when incorporated by breeding into genotypes with lower pod values. In the Soppeng trial, pod production has lagged behind Ladongi. As reported last year, project team members tried to persuade farmers at the Soppeng trials to adequately prune the mother trees in order to allow the side grafts to fully establish. Compensations were paid to the farmers for any loss of production due to pruning. Farmers have responded by pruning at the trial site but only to a light level. Therefore, inadequate sunlight is reaching foliage and this has inhibited pod production on the grafted cocoa clones. The project team is continuing to talk to farmers about this problem. Quality analyses were conducted by PT Effem for the clones that had produced a sufficient quantity of beans. The results of these analyses indicate that these clones have substandard quality characteristics. This would suggest that, should clones prove to have resistance to CPB, PPR or VSD, they would need to be backcrossed with clones of good quality (ie with high fat content etc).
Links with the SUCCESS Alliance project (a USDA funded project administered by ACDI/VOCA*) were strengthened during the year to foster wider evaluation of promising screening material and recommendations for cultural management. A MOU was signed between the SUCCESS Alliance project and La Trobe University. New joint genotype-testing trials were established in Southeast Sulawesi at three ACDI-VOCA training sites in different parts of Kolaka District. Blocked trials with 3 replicates have been established at each site to test 15 clones (side-grafted onto mature trees in 6-10 tree plots). Very dry conditions in the latter half of 2004 caused a high mortality of side-grafts. Consequently, it has been necessary to allow time for more grafting budwood to be produced at the ACIAR trial sites so that grafting can continue to complete at least 6 trees in each plot.
2. Uptake of cultural practices to minimise disease and pest risks. A two-day training workshop was conducted in Kolaka District, Southeast Sulawesi for farmers, extension officers (including ACDI-VOCA officers). The workshop covered pest and disease management for smallholders. It included opening speeches, presentations and field demonstrations. Participants enthusiastically welcomed the opportunity to hear about approaches to pest and disease management from the trainers including project staff from ICCRI and Masterfoods. Topics covered by the workshop were management strategies for CPB, PPR, canker and VSD, and rehabilitation of farms by grafting, especially chupon grafting which is now routinely practiced in Brazil. Over 80 participants attended on each day including provincial Dinas Perkebunan officers, ACDI-VOCA* field technicians and trainer farmers, smallholders, BPTP staff and Kendari town council staff. New trials established jointly with SUCCESS Alliance have also encouraged the uptake of cultural farm practices. Farmers participating in the SUCCESS Alliance training programs have been doing the grafting and the assessments of graft survival. This program aims to encourage farmer-centred research and the linking on-farm research with extension and training objectives. As well as having the research aims of retesting clones selected from the ACIAR trials, the new trials aim to introduce methodologies of farm improvement and pest/disease management to smallholders.
3. Enhancing the infrastructure and capability within Indonesia for the collection, maintenance and screening for pest and disease resistance of cocoa genotypes During 2004, project staff from La Trobe University visited Indonesia in March, August-October and November 2004. Our project team member from Masterfoods Australia also made regular visits to Sulawesi. These visits enabled the project staff to address issues such as pruning at the ACIAR trials sites, the procedures for evaluation of the clones (some of which are now producing pods) and also to run a field workshop for training of extension staff and farmers in Southeast Sulawesi (see above). BPTP SULTRA staff were trained in evaluation of clonal performance by the Australian project staff and PT Effem. Since mid-2004, they have been actively commencing the evaluation of the trials. ICCRI staff have developed procedures building on their experience for assessing clones for CPB and PPR resistance and characterisation of the clones. Some of the clones included in the ACIAR trials are probably mislabelled, a common problem in cocoa research around the world - there is a possibility that USDA could assist in analysis of samples sent from ICCRI for genotypic identification providing appropriate material transfer agreements are in place.
(* ACDI/VOCA resulted from the 1997 merger of US NGOs - Agricultural Cooperative Development International and Volunteers in Overseas Cooperative Assistance. The merger blended ACDI's systemic, long-term approach to development and VOCA's people-to-people volunteer activities.)

1. Collection and resistance screening of material and assessments of quality characteristics of cocoa lines
Field evaluation of pod-bearing clones continued on a fortnightly basis in Ladongi and Soppeng . However, pod production in Soppeng remains very low (see below) because of the reluctance of farmers to adequately prune the mother trees. The trial at Ladongi has been well pruned and the grafts have grown prolifically and are now producing sufficient pods for the assessment of cocoa pod borer (CPB) and Phytophthora pod rot (PPR) and quality characters. At Ladongi, twenty clones (fourteen local and six international selections) produced a sufficient number of pods per replicate in 2005 to enable statistical analyses. CPB incidence ranged from 41% (Aryadi 2) to 95% (a susceptible control) and the incidence of severely infested pods (in which 50% or more beans were damaged) ranged from 0% to 63% (see Table 2, attached). However, the pods of 17 of the 20 clones had rates of severe infestation of 20% or less. Mann-Whitney tests indicated that Haris 1, Aryadi 2, KKM 22 and Darwis 3 had significantly lower rates of severe infestation than 10 of the other twenty clones. Of these four clones only KKM 22 is not a local selection, indicating the potential of using selection and propagation of genotypes currently being grown on farms to minimise CPB damage. Among the twenty clones examined, PPR incidence ranged from 4% to 31% but significant differences were not discerned due to the variability of the data (see Table 3, attached). The susceptible control, PpLadS, had the third highest PPR incidence (23.5%). Six clones had a PPR incidence of less than 10%; four of these were local selections, indicating again the promise of employing on-farm selection and propagation of genotypes as part of an overall pest/disease management strategy. A number of local selections and Indonesian clones from the Indonesian Coffee and Cocoa Research Institute (ICCRI) collection had relatively low pod values (no. pods per kg dry beans): two Sumatra clones, a Papua New Guinea clone and two Sulawesi selections had pod values <30 (see Table 1, attached). Three other selections from Southeast Sulawesi had pod values in the range 30-35. However, some local selections that demonstrated possible resistance to CPB, PPR and vascular-streak dieback (VSD) (Aryadi2, Aryadi3 and VSD2LadR) had higher pod values (42.6, 38.9 and 79.4, respectively). Quality analyses of some of the Ladongi clones are currently under way at PT Effem. This year as the SUCCESS Alliance project finished in January 2005, the genotype testing trials established jointly with the SUCCESS Alliance project (see 2004 report) were taken over by the ACIAR project. Further work was conducted on these trials, especially re-grafting to replace unsuccessful grafts (caused by dry conditions, see 2004 report). However, the shortage of grafting material available was a constraining factor and, therefore, greater focus has been placed on one of the trials (at Tirawuta, Rate-Rate). This will ensure that a sufficient number of replicates are propagated in at least one of the trials. As with the initial larger trials, the benefit of having more than one trial in different locations is evident.

2. Uptake of cultural practices to minimise disease and pest risks.
Farmer groups such as the SUCCESS Alliance groups involved in on-farm research with ACIAR have also been educated on cultural practices through the joint research with the ACIAR project and their close involvement with Dinas Perkebunan, Kendari. From training received from sessions with extension organisations and others such as the 2004 ACIAR training workshop (see 2004 report), farmers involved in these groups take back a greater understanding of cultural practices that can be applied to their own farms to manage pests and diseases. Other farmers in the area also learn about these methods by word-of-mouth. In April, meetings were held between the project coordinator and BPTP SULTRA and Dinas Perkebunan Kendari and, in South Sulawesi, with DInas Perkebunan and BPTP Makassar to discuss approaches to encourage the research initiated by the ACIAR project and to foster uptake of methodologies and techniques developed by the project. The general consensus was that it was important to apply methods developed in the project and to test other methods (e.g. pod sleeving) in a number of locations with a greater focus on extension (for example, by using demonstration plots). This is an indication of the strong interest among some Indonesian institutions in conducting work with joint research and extension goals. Demonstration plots would provide a good educational tool for disseminating cultural practices as well as testing technologies. A visit was made by the project coordinator to ICCRI in Kaliwining, Jember in May 2005. Discussions were held there to identify priorities for continuing the ACIAR project research in Sulawesi. The incorporation of breeding clones with promising resistance, yield or quality characteristics in future work was felt to be important (see below). The ideas gained from discussions in Jember and in South and Southeast Sulawesi contributed to the preparation by the project leader of a pipeline proposal for a new project building on the current project. This proposal was submitted on 23 September 2005.

3. Enhancing the infrastructure and capability within Indonesia for the collection, maintenance and screening for pest and disease resistance of cocoa genotypes
A project staff member from La Trobe University visited Indonesia in April-May 2005 and in September 2005 to present the work of the project to the International Conference for Crop Security, Malang, East Java. Our project team member from Masterfoods Australia also made regular visits to Sulawesi. These two project staff members visited the Soppeng trial site. This trial is evaluated on a fortnightly basis but still has a generally low pod production. The Ladongi site, by contrast, has been well-managed and the root-stock trees have been pruned sufficiently to allow good development of the grafts. Visits were made to the Ladongi trial site and to two of the more recent trials established with SUCCESS Alliance (see 2004 report). During the visits availability of grafting material was assessed in the Ladongi trial and as much material was collected as possible to provide to the new trials SUCCESS trials to replace failed grafts. Farmers in two locations (in North Kolaka and Tirawuta) were affected severely by PPR in the 2005 wet season and this was the greatest current concern of the cocoa producers in the area. PPR had caused 50% losses in the Rate-Rate area and even higher losses in North Kolaka. Graft mortality at the trial sites was high, possibly due in part to the particularly wet conditions. Scions collected from the Ladongi trial were distributed to the three new trials. In the future it is expected that the greatest emphasis will be placed on the trial at Tirawuta (see above). An important aspect of developing resistant clones is to cross genotypes with good resistance characters with other genotypes having desirable yield and bean quality characters. At ICCRI, fourteen hybrid crosses are being tested on the Kaliwining cocoa research plantation against VSD. The possibility was raised of retesting some of these hybrids in Sulawesi. For example, a cross between ICS13 and Sca 12 produced a VSD-resistant hybrid that was more productive than the resistant Sca 12 parent. An hypothesis that could be investigated regarding VSD resistance relates to leaf size. The most VSD-resistant clones at ICCRI tend to have small leaf size. These include KA2 106, Sca 12, DRC 15 and KEE 2. This might be related to a lower transpiration rate in these clones compared to other clones with larger leaves. An investigation into this could facilitate selecting genotypes for VSD resistance, particularly in drier areas which appear to be more prone to VSD epidemics.