The Project seeks to increase the efficiency and sustainability of yam production in the South Pacific through improved crop nutrition. Yams (predominantly Dioscorea alata and D. esculenta, but recently also D. rotundata) are important staple foods in the Pacific, and hold an esteemed place in cultural heritage. Among the tropical root crops, they are the most demanding of high soil fertility, and this has contributed to the decline in yam production as pressure on land and labour has increased. However, a healthy yam crop can be highly productive and give very high economic returns. This project aims to determine the extent to which South Pacific yam yields are limited by nutritional stresses, and whether yields can be economically elevated using appropriate soil fertility management.

The following activities have been achieved in the year to June 2002:

(i) At the University of Queensland, solution culture experiments with D. alata cv. Mahoa'a were conducted varying rates of Ca, Mg, S, K and NaCl (salinity). Critical concentrations have been estimated for these nutrients. The tentative critical concentrations for leaves sampled at the 5th and 6th nodes are: 1.1%, 0.15%, 0.12%, 3.3%, for deficiencies of Ca, Mg, S and K respectively. A nutrient omission trial was conducted in solution culture to compare deficiency symptoms on two genotypes of D. alata and two of D. rotundata. Yam germplasm multiplication has continued, and there is now sufficient material of several cultivars of Dioscorea rotundata and D. alata to undertake comparitive experiments in the coming growing season. Leaves were also analysed from the field-grown germplasm collections, and revealed relatively little variability among genotypes within species but some significant differences between the two species. Furthermore, the samples from field-grown plants showed little variation in nutrient concentration with leaf age, which contrasts with observations on plants grown in solution culture.
(ii) In Tonga, 16 field trials were conducted, 13 testing the P response of yam on various soils throughout the Kingdom, and three testing the ability of legume green manure crops to increase P availability to the subsequent yam crop. Damage by pigs or cyclone affected five trials, but seven P rate trials gave significant responses, with yield increased by approximately 30-50% by optimum P rates between 50 and 200 kg/ha. The green manure trials found that mucuna was able to improve P nutrition of the yam crop equivalent to 50 kg/ha P. Dolacus lablab was also tested on one site and had a positive effect about half that of mucuna.
(iii) In PNG, nutrient omission pot trials were completed on two soils in the new greenhouse. Poor rainwater collection and storage facilities were identified as a constraint, and are currently being replaced with new plastic fittings. The field trial at Bubia comparing different fallow vegetation gave expected responses to prior fallow type, but no response to fertilizer. A survey of yam production and nutritional status in Kiriwina and Alotau areas in Milne Bay Province was completed in February. Preparations have been made to begin field activities on Kiriwina in July. Two sites in the Markham valley have also been established with improved fallow species, to be planted with yam in November. The field trial at Dugumor village in Madang Province was planted to yam in October and has been visited several times for maintenance. It will be harvested in August 2002.
(iv) In Vanuatu, pot experiments have continued to characterise a further three soils, on which it is intended to establish field trials. A second attempt was made to measure P response by yams in the germplasm collection at VARTC, and a small positive response was detectable despite high yield variability among plants. A PxK rate experiment was established at Montmartre using Dioscorea esculenta, and is yet to be harvested. Marie Melteras began studies at the University of Queensland in preparation for her Masters program, and worked with UQ staff to develop an extensive field program to begin in July.
For detailed reports on research activity in each country program, see individual program reports appended.

Year 4 (01/07/2002-30/06/2003)
The following activities were achieved in the year to June 2003:
At the University of Queensland, solution culture experiments were conducted inducing deficiencies of iron, boron, manganese and zinc, and toxicities of boron, manganese, zinc and copper. Improvements in nutrient management resulted in much better growth of the yams in solution culture than in previous seasons. Planned experiments on copper and molybdenum deficiencies were not conducted due to failure of sprouting of the planting material.
Some 3000 leaf samples were processed for analysis, including samples from the overseas programs and carry-over from last year. However, temporary decommissioning of the laboratory due to building refurbishment has prevented completion of this season's analyses. A critical concentration for manganese deficiency has been determined at 10 mg/kg in leaves of the 5th and 6th nodes from the tip. Yam germplasm maintenance continued, with the whole collection being planted at Redland Bay Farm for the first time.
In PNG, the project has installed a new 9000 litre plastic water tank and plumbing on the laboratory building at Bubia, to supply high quality rainwater for distillation purposes. This has allowed pot trials to be conducted. A sand culture experiment to induce nutrient deficiency in D. esculenta yam has been completed. Symptoms of most deficiencies were successfully induced and photographed. Leaf nutrient levels are awaiting analysis at UQ. On-station active fallow and live-staking demonstration plots with D. rotundata and D. esculenta completed their first crop of yam. The first season of on-farm yam trials at Dugumor, Madang Province were harvested in August 2002 and the plots replanted in October 2002. A fertiliser response trial was also planted in October 2002. On-farm trials were established on Kiriwina Island, Milne Bay Province, including four Gliricidia live stake plots (to be planted with yam in the 2003-04 season), and two fertilizer response trials in yam gardens. Demonstration plots of active fallow and live-staking with gliricidia were further developed at Tanam and Mutzing sites in Markham Valley, but only the Tanam plot could be planted this season, and yams grew poorly apparently due to zinc deficiency. Experiments to address the zinc deficiency are planned for the coming season. The socio-economic surveys in East Sepik, Madang and Milne Bay Province (conducted between May 2000 and February 2002) have been collated and a report drafted.
In Tonga, experiments continued to focus on the phosphorus response of yam, and on the contribution of a Mucuna cochinchinensis (velvet bean) cover crop to the nitrogen (N) and phosphorus nutrition of subsequent crops. In Vanuatu, 10 field trials were conducted during the year. Eight evaluated fertilizer responses, based on deficiencies identified through pot trials. One experiment examined effect of Gliricidia leaf mulch, and another (still in progress) is comparing planting sets of high and low nutrient status.
The year reported was to be the final year of the project. Research efforts in all programs were expanded, in an effort to make up for delays in the first two years. Good progress was made, with most project objectives achieved by the end of the year. However, some prescribed items remained to be completed, and in other areas an extension would allow promising initiatives to come to fruition. An extension of 18 months was granted to the end of 2004, allowing one additional cropping season and time to collate results.

The Project seeks to increase the efficiency and sustainability of yam production in the South Pacific through improved crop nutrition. Yams (predominantly Dioscorea alata and D. esculenta, but recently also D. rotundata) are important staple foods in the Pacific, and hold an esteemed place in cultural heritage. Among the tropical root crops, they are the most demanding of high soil fertility, and this has contributed to the decline in yam production as pressure on land and labour has increased. However, a healthy yam crop can be highly productive and give very high economic returns. This project aims to determine the extent to which South Pacific yam yields are limited by nutritional stresses, and whether yields can be economically elevated using appropriate soil fertility management.

The following activities have been achieved in the year to June 2002:

(i) At the University of Queensland, solution culture experiments with D. alata cv. Mahoa'a were conducted varying rates of Ca, Mg, S, K and NaCl (salinity). Critical concentrations have been estimated for these nutrients. The tentative critical concentrations for leaves sampled at the 5th and 6th nodes are: 1.1%, 0.15%, 0.12%, 3.3%, for deficiencies of Ca, Mg, S and K respectively. A nutrient omission trial was conducted in solution culture to compare deficiency symptoms on two genotypes of D. alata and two of D. rotundata. Yam germplasm multiplication has continued, and there is now sufficient material of several cultivars of Dioscorea rotundata and D. alata to undertake comparitive experiments in the coming growing season. Leaves were also analysed from the field-grown germplasm collections, and revealed relatively little variability among genotypes within species but some significant differences between the two species. Furthermore, the samples from field-grown plants showed little variation in nutrient concentration with leaf age, which contrasts with observations on plants grown in solution culture.
(ii) In Tonga, 16 field trials were conducted, 13 testing the P response of yam on various soils throughout the Kingdom, and three testing the ability of legume green manure crops to increase P availability to the subsequent yam crop. Damage by pigs or cyclone affected five trials, but seven P rate trials gave significant responses, with yield increased by approximately 30-50% by optimum P rates between 50 and 200 kg/ha. The green manure trials found that mucuna was able to improve P nutrition of the yam crop equivalent to 50 kg/ha P. Dolacus lablab was also tested on one site and had a positive effect about half that of mucuna.
(iii) In PNG, nutrient omission pot trials were completed on two soils in the new greenhouse. Poor rainwater collection and storage facilities were identified as a constraint, and are currently being replaced with new plastic fittings. The field trial at Bubia comparing different fallow vegetation gave expected responses to prior fallow type, but no response to fertilizer. A survey of yam production and nutritional status in Kiriwina and Alotau areas in Milne Bay Province was completed in February. Preparations have been made to begin field activities on Kiriwina in July. Two sites in the Markham valley have also been established with improved fallow species, to be planted with yam in November. The field trial at Dugumor village in Madang Province was planted to yam in October and has been visited several times for maintenance. It will be harvested in August 2002.
(iv) In Vanuatu, pot experiments have continued to characterise a further three soils, on which it is intended to establish field trials. A second attempt was made to measure P response by yams in the germplasm collection at VARTC, and a small positive response was detectable despite high yield variability among plants. A PxK rate experiment was established at Montmartre using Dioscorea esculenta, and is yet to be harvested. Marie Melteras began studies at the University of Queensland in preparation for her Masters program, and worked with UQ staff to develop an extensive field program to begin in July.
For detailed reports on research activity in each country program, see individual program reports appended.

Year 4 (01/07/2002-30/06/2003)
The following activities were achieved in the year to June 2003:
At the University of Queensland, solution culture experiments were conducted inducing deficiencies of iron, boron, manganese and zinc, and toxicities of boron, manganese, zinc and copper. Improvements in nutrient management resulted in much better growth of the yams in solution culture than in previous seasons. Planned experiments on copper and molybdenum deficiencies were not conducted due to failure of sprouting of the planting material.
Some 3000 leaf samples were processed for analysis, including samples from the overseas programs and carry-over from last year. However, temporary decommissioning of the laboratory due to building refurbishment has prevented completion of this season's analyses. A critical concentration for manganese deficiency has been determined at 10 mg/kg in leaves of the 5th and 6th nodes from the tip. Yam germplasm maintenance continued, with the whole collection being planted at Redland Bay Farm for the first time.
In PNG, the project has installed a new 9000 litre plastic water tank and plumbing on the laboratory building at Bubia, to supply high quality rainwater for distillation purposes. This has allowed pot trials to be conducted. A sand culture experiment to induce nutrient deficiency in D. esculenta yam has been completed. Symptoms of most deficiencies were successfully induced and photographed. Leaf nutrient levels are awaiting analysis at UQ. On-station active fallow and live-staking demonstration plots with D. rotundata and D. esculenta completed their first crop of yam. The first season of on-farm yam trials at Dugumor, Madang Province were harvested in August 2002 and the plots replanted in October 2002. A fertiliser response trial was also planted in October 2002. On-farm trials were established on Kiriwina Island, Milne Bay Province, including four Gliricidia live stake plots (to be planted with yam in the 2003-04 season), and two fertilizer response trials in yam gardens. Demonstration plots of active fallow and live-staking with gliricidia were further developed at Tanam and Mutzing sites in Markham Valley, but only the Tanam plot could be planted this season, and yams grew poorly apparently due to zinc deficiency. Experiments to address the zinc deficiency are planned for the coming season. The socio-economic surveys in East Sepik, Madang and Milne Bay Province (conducted between May 2000 and February 2002) have been collated and a report drafted.
In Tonga, experiments continued to focus on the phosphorus response of yam, and on the contribution of a Mucuna cochinchinensis (velvet bean) cover crop to the nitrogen (N) and phosphorus nutrition of subsequent crops. In Vanuatu, 10 field trials were conducted during the year. Eight evaluated fertilizer responses, based on deficiencies identified through pot trials. One experiment examined effect of Gliricidia leaf mulch, and another (still in progress) is comparing planting sets of high and low nutrient status.
The year reported was to be the final year of the project. Research efforts in all programs were expanded, in an effort to make up for delays in the first two years. Good progress was made, with most project objectives achieved by the end of the year. However, some prescribed items remained to be completed, and in other areas an extension would allow promising initiatives to come to fruition. An extension of 18 months was granted to the end of 2004, allowing one additional cropping season and time to collate results.

The past 12 months have encompassed the fifth and final experimental season for the project. Research programs were implemented as outlined in the project extension proposal. Many trials have only recently been harvested, and a few are still to be harvested in the July-September quarter (mainly in PNG). Analysis and collation of trial results will be completed over the final six months of the project, and full reports on the below-mentioned trials will be appended to the project's completion report.

The following activities have been achieved in the year to June 2004:

(i) At the University of Queensland, solution culture experiments were completed for deficiencies of nitrogen, boron, iron, copper and molybdenum. In addition, toxicities of boron, zinc and copper were developed to greater intensity than achieved in the trial last year. This season, pH buffering using Mes buffer was introduced, with 0.5 mM Mes achieving reasonably good pH stability (+/- 0.4 units) between weekly pH adjustments. This improved pH stability was considered important for stable nutrient buffering by chelating resins. The copper deficiency experiment successfully used Amberlite IRC-718 cation-chelating resin to induce a range of deficiency levels, with well-developed symptoms. The same resin was used at pH 5.0 to pre-clean nutrient salts and water used in the Mo experiment, but was not successful in eliminating contamination. The borate-binding anion-exchange resin Amberlite IRA-743 was trialled for the second time, again with poor results suggesting phytotoxicity to yam. Over 1000 leaf samples were generated by the solution culture experiments, and some 350 samples received from overseas programs. These have been processed in the laboratory, but analysis results are not yet collated for reporting. Due to the sale of the University's research farm at Redland Bay, the rootcrops germplasm collection was moved to the DPI Redlands Research Farm at Cleveland. All but one of the yam genotypes received from IITA have been destroyed on account of virus infestation. From the remaining four D. alata genotypes, leaves were sampled monthly to track changes in leaf nutrient concentrations in leaves at various positions on the vine.

(ii) In Tonga, the project established 5 sites comparing D. alata response to N and P rates on soils planted 4 to 5 months with Mucuna cochinchinensis (velvet bean) and Panicum maximum fallow. The results did not show any significant responses as last year. The project staff in Tonga reported that severe drought during tuberization was the likely cause of no significant responses. Mucuna roots were examined microscopically and found to be heavily populated with vesicular-arbuscular mycorrhizae (VAM). In the greenhouse, yams were successfully grown in pot experiments demonstrating P responses on two soils, both taken from sites continuously cropped for 5 years prior to sampling.

(iii) In Papua New Guinea, Gliricidia live-staking plots have been planted with yam for the third year in Dugumor, Madang Province, for a second year at Bubia Research Station and for the first season at Mutzing in the Markham Valley and in three villages on Kiriwina Island in the Trobriand group, Milne Bay Province. In addition, fertilizer rate trials were planted at Dugumor and on Kiriwina. At Tanam in the Markham Valley, an experiment is investigating zinc deficiency and possible remedies with both yam and maize. Mid season observations show that zinc deficiency was exacerbated by NPK fertilization, and ameliorated by either zinc application (soil-applied or foliar sprays) or chicken manure. Harvests are underway at the time of reporting.
In Vanuatu, the phosphorus rate trial at Malo was harvested in August, and replanted for a second year, this time with potassium applied to two replicates. A small P response was measured, but K response was not evident. At Tagabe Research Station, parallel experiments compared the response of Dioscorea alata, D. rotundata and sweetpotato to N, P and K applications, but no responses were evident. At VARTC, the growth and development of D. esculenta was studied, with particular attention to root morphology. Destructive harvests each month involved careful excavation of the roots and recording of root length and depth. Roots were observed to remain very shallow (less than 10 cm) while extending out from the plant for over 2 m, but ultimately growing down to about 30 cm depth. Feeder roots were not concentrated in or under the mound, except for those growing from developing tubers. These results call to question the efficacy of fertilizer placement in the mound, and the adequacy of 1 m guard rows between fertilizer treatments.

The period of this report encompassed the final six months of the project's first extension of 18 months. Research activities involved the final stages of some 2003-04 season field trials in Vanuatu and PNG, laboratory analyses of leaf, tuber and soil samples as well as a small field trial at the University of Queensland, and data analysis and write-up.

In September 2004, four project participants (Dr. O'Sullivan, Dr. Halavatau, Ms. Melteras and Mr. Ernest) contributed to the 4th International Plant Nutrition Colloquium in Brisbane. Two posters were presented, with full papers published in the proceedings (see publication list). Following the congress, James Ernest remained in Brisbane for a further week, to process PNG samples in the laboratory and work with Dr. O'Sullivan on statistical analysis and report structure for the PNG program. Ms. Melteras was based in Brisbane from July 2004 to July 2005, to process samples from the Vanuatu program and write up her MPhil thesis encompassing the research of the Vanuatu program. She was supported by a John Allwright Fellowship during this year. There has been no activity in the Tonga program in the reporting period, as funding for this program ceased in June 2004. In the University of Queensland program, samples from the 2003-04 experimental season were analysed and the collation of data begun. However, this work remains incomplete, due to Dr. O'Sullivan's time being taken up with ACIAR Project SMCN/2000/060 and in supporting the finalization of the Vanuatu and PNG programs.

One highlight of the 2003-04 experimental season was the experiment in Vanuatu following the development of D. esculenta plants. The experiment continued to October 2004, documenting a full annual cycle, with monthly excavation of roots showing that roots extend over considerable distance (up to 4 m) from the planting position. This raised questions about the most effective placement of fertilizer, and the distance required to separate fertilizer treatments in the field. To address these questions, a small trial was conducted by Dr. O'Sullivan this year in Brisbane, using the D. alata germplasm maintenance planting. Strontium was used as a tracer of root activity. This is a non-toxic element of naturally low abundance, which is taken up in a similar manner to calcium. By "spiking" specific locations in the soil with Sr, the ability of the roots to reach that location can be detected by measuring elevated Sr levels in the leaves. This experiment was very successful, and established that yam roots may reach horizontally for at least 5.5 m, and to depths exceeding 40 cm. It also showed that fertilizer placed in the planting hole under the sett is accessed rapidly by the plant, but that uptake was greater when placed in a ring around the set, within the mound.

Progress over the past year:
Marie Melteras (Vanuatu project research officer and John Allwright Fellow) submitted her Masters of Philosophy thesis in July 2005 before returning to Vanuatu. The title of the thesis is "Soil fertility as a factor affecting the production of yams (Dioscorea spp.) in Vanuatu". The thesis examiners' reports were received in November, recommending a number of minor changes. After an extension of time due to Ms Melteras' current work responsibilities, these were completed and the revised thesis submitted in July 2006.

Dr. O'Sullivan completed work as contracted on the PNG project reports in July 2005. In September, she attended the 13th International Plant Nutrition Colloquium in Beijing, with support from the Plant Nutrition Trust (Alf Anderson Award). She presented a poster entitled "Interpretation of tissue analysis for potassium nutrition in tropical root crops dependent on sodium concentration" describing work resulting from the Masters program of Sharryl Ivahupa, a John Allwright Fellow under ACIAR LWR2/1991/01. A more detailed account of this work was published in the Journal of Plant Nutrition in June 2006 (Volume 29(6), 1095-1108).

September also saw the completion and publication of the internet and CD resource "Sweetpotato DiagNotes", the product of ACIAR SMCN/2000/060. The product has been widely acclaimed. The internet version is available at
http://www.lucidcentral.org/keys/sweetpotato.

A paper on nutritional deficiencies of Dioscorea alata was accepted for publication by the Journal of Plant Nutrition, and will include two colour plates containing 23 images of deficiency symptoms. The cost of publishing the colour plates was covered by ACIAR in the current extension budget.

Manuscripts of three papers resulting from the project were drafted in October-December 2005 and circulated for co-author comment. These have since been submitted for publication.

Of the work identified to be completed in the 2006 extension period,
- the slide collection of symptoms of nutritional disorders in yams (1600 slides) has been scanned at high resolution to a digital database. This is yet to be annotated.
- the leaf analysis samples identified as requiring reanalysis have been rerun, completing the primary data from the project.
Work remaining includes completion of data analysis and compilation of results.