Overview Objectives

The aim of the project is to characterise the virus diseases of taro, a Pacific Island staple crop, and to develop sensitive specific tests for each virus.

Project Background and Objectives

Taro is widely grown in Papua New Guinea (PNG) and many other Pacific Island countries. It also plays an important cultural role. The roots are a source of carbohydrate, and the foliage is also eaten. It is cultivated mainly in gardens for local use, but there is also a domestic and export market. Over the last 20 years there has been a gradual decline in the production of taro because of the effects of pests and diseases. Taro leaf blight, caused by a fungus, is the most serious and widespread disease of the plant in Pacific countries. It has long been present in Micronesia, Papua New Guinea and the Solomon Islands, but in 1993 it spread to American Samoa and Samoa with devastating consequences. Many growers have since abandoned taro cultivation in these countries, causing major social and economic problems. Export earnings in Samoa fell from 9.5 million Tala to 158,000 Tala in just one year after the arrival of the blight.

The genetic diversity of taro is poorly known, but some described varieties are resistant to the fungus. In 1993, a breeding program started up in PNG to develop these varieties but the taro germplasm cannot be moved between countries because of the presence of a lethal virus disease known as alomae. It is now important to characterise this disease (which seems to be associated with the presence of two viruses together) and develop reliable tests for the presence of both viruses within taro germplasm. This should then allow free movement of germplasm and thereby help in combating leaf blight and in developing other features of the plant.

Increased knowledge of alomae will be helpful of itself because this disease is now the main constraint on taro production in PNG and the Solomons. Elsewhere it seems that the two viruses do not occur together; when only one virus is present, disease symptoms are much milder. Characterising the virus diseases of taro, a Pacific Island staple crop, is underway as the first step to developing sensitive specific tests for each virus.

Progress Reports (Year 1, 2, 3 etc)

There are several viruses known to infect taro. Dasheen mosaic potyvirus (DMV) is well characterised and has a worldwide distribution. Taro small bacilliform virus (TSBV), a putative badnavirus, is also thought to have a widespread distribution. There also appears to be 2 putative rhabdoviruses associated with taro in the Pacific. One is thought to occur in Vanuatu and Fiji (known as taro veinal chlorosis virus), while the other, known as taro large bacilliform virus (TLBV), is thought to be restricted to Papua New Guinea and the Solomon Islands. Being a vegetatively propagated crop, there is a strong possibility there will be other viruses infecting taro.

Achievements to date

In summary, the work that has been achieved to date is as follows:
A microsatellite enriched genomic library for taro has been created.
A core set of taro germplasm from the Pacific Island countries have been screened with eight different primer pairs for microsatellite loci.
75% of microsatellite pairs revealed polymorphism.
ISSR fingerprinting has been optimised for use with taro germplasm.
The ISSR data set revealed unique fingerprints for the majority of Pacific Island germplasm screened.
A data module that will manage the molecular marker data generated has been developed within ICIS (the International Crop information System).
Preparations finalised for a 12 week visit by 2 PNG scientists to the University of Queensland for technology transfer.

2.1. Development and application of molecular markers to taro germplasm

Two molecular marker techniques have been developed for taro germplasm; microsatellites (SSRs: Simple Sequence Repeats) and ISSRs (Inter Simple Sequence Repeats).

2.1.1. Microsatellites

Microsatellites exploit hypervariable regions in the genome, which are comprised of simple sequences repeated in tandem, e.g. ‘AAAAAA’, ‘CACACA’. Hence, microsatellites are also known as Simple Sequence Repeats (SSRs), and also, more generally, VNTRs (variable number of tandem repeats). These repeats vary in number and length, and are known as microsatellites when the basic repeat unit is around 2 to 8 base pairs in length.

SSRs are very powerful markers, in that they are single locus, codominant and multiallelic. They are extremely robust and easily exchanged between labs and don’t require radioactivity for detection. Multiplex reactions can be run also, to speed up the process where the products have non-overlapping size-ranges. Traditionally, the use of microsatellites has been limited by the initial cost in finding and sequencing loci. The standard approach to obtain microsatellites is by screening a size-selected genomic library with oligo probes containing different repeat motifs, and with this methodology the proportion of recombinants containing microsatellites usually varied between 0.5 and 2% (Hammond et al, 1998). However, in recent years, different approaches have been developed with the aim to enrich for microsatellites and so circumvent the limitations associated with their use. The SSR enrichment protocol as described by Edwards et al. (1996) has been applied to taro microsatellite capture. This methodology completely eliminates the need for construction of a genomic library, through the hybridisation capture of microsatellites directly from genomic DNA.

The following synthetic oligonucleotide fragments have been used to enrich for microsatellite containing DNA fragments: [CT]15, [CA]20, [GA]15, [ACC]10, [GT]15, [GAC]10, [CAT]10, [AGC]10, [GATA]8, [GAG]10, [GAA]10, [AGA]12, [ACT]10, [TAA]10.
Following optimisation of the protocol described by Edwards et al. (1996), the average enrichment rate of the proportion of recombinants containing microsatellites is 25% (Mace & Godwin, 2000). The majority of the repeat motifs were either dinucleotide or trinucleotide repeats, of which 37.7% were compound perfect repeats and the remaining 62.2% were compound imperfect repeats. One heptamer was also recorded with 2 repeats. The dinucleotide repeats had, on average, higher numbers of repeats (9) than the trinucleotide repeats (7). The most common motif found in taro was GT/CA.

To date, 30 microsatellites have been used for primer design as they contained either dinucleotide repeats with more than seven repeat motifs and trinucleotide with more than four repeat motifs, or a mononucleotide string of more than 10, in addition to having sufficient flanking sequence to permit the design of a PCR primer pair.

The primer pairs were then applied to a core set of Colocasia esculenta var. esculenta (dasheen type) accessions selected based on country of origin, passport data and, where available, isozyme electrophoretic patterns (TANSAO, 1999), to be representative of the genetic diversity in the Pacific Island region. An accession of C. esculenta var. antiquorum (eddoe type) and Xanthosoma species, both from Fiji, were also included for comparative purposes.

Several difficulties have been encountered in this reporting period;

The budget for travel and accommodation/sustenance (calculated by a collaborator on the complementary Tarogen (AusAID-funded project) was under-estimated which has resulted in reduced travel to Pacific Island countries.

Communication with PNG (Unitech) has been a problem due to staff absences.

The timetable and funding of the virus indexing of taro plants for the Tarogen project is causing some concern. It is unlikely that virus diagnostics will be developed prior to the end of the project and this has a significant impact on virus indexing of taro plants for the Tarogen project, since there will be no personnel or budget to fund the work. To achieve the desired outcomes of both the ACIAR and Tarogen projects, it would be highly desirable to extend the project for 1 year with a budget to include a research associate and consumables - this will enable the taro to be indexed by an experienced researcher using the recommended schedule (growing the taro for 1 crop cycle).

Funding for the project finished in December, 2001. Dr Mace left the project to take up a new position at ICRISAT in India, commencing in October, 2001.

The project was reviewed for ACIAR in February, 2002. As a result of the reviewers’ recommendations, the DNA fingerprinting component was granted a 12 month extension, with the major objective to DNA fingerprint the Solomon Islands collection. The extension will include provision for salary for a technician, consumables, and a small travel budget. This extension will commence in July, 2002.

2.1 Development and application of molecular markers to taro germplasm

As reported previously, two molecular marker techniques have been developed for taro germplasm: microsatellites (SSRs: Simple Sequence Repeats); and ISSRs (Inter Simple Sequence Repeats).

The most polymorphic SSR markers have been applied to the collections, or sample collections from all Pacific Island countries involved in the study, with the exception of the Solomon Islands. These markers have been used in combination with morphological descriptors to develop country core collections, encompassing approximately 10% of the original collection.

The Solomon Islands collection is currently held as tissue cultures at SPC. The first 20 accessions have been received and DNA extractions are in progress. The remaining accessions will be hand-carried to UQ by Ian Godwin after his attendance at the TGRC meeting in Suva in May, 2002.

2.1.2 ISSRs
ISSR fingerprinting differs from the SSR marker approach in that no prior sequence knowledge is required. The technique can therefore be applied immediately to the germplasm, with only a small amount of optimisation necessary for different taxa (Godwin et al., 1997). However, the ISSR approach does have limitations, principally in that they are dominant markers, and hence homozygotes can not be distinguished from heterozygotes.

The technique has been applied to many different plant taxa for genetic diversity analysis to date, e.g. finger millet (Salimath et al., 1995) and sorghum (Yang et al., 1996). Primers based on a repeat sequence, such as (CA)n, are designed with a 3’-anchor, such as (CA)8RG or (AGC)6TY (where R=purines: G or A; Y=pyrimidines: C or T). The resultant PCR reaction amplifies the sequence between two SSRs. 33 primers were initially screened, and four were selected to continue with, based on the clarity of the bands, the presence of polymorphism and repeatability of the banding patterns. These 4 primers were based on (GA)n repeats and (ACC)n repeats. The PCR products were separated by both agarose gel and polyacrylamide gel electrophoresis. With the agarose gels, the bands were visualised by staining with ethidium bromide, and with the polyacrylamide gels, the bands were visualised through silver staining. The majority of the ISSR work was carried out by a Masters student, Ms Nur Zuhairawaty. These markers will be useful in distinguishing among accessions which have an identical fingerprint, and may be instrumental in identifying true duplicates within the collection.

2.2 Database Development

One of the other major objectives of the DNA-fingerprinting component is to develop a data module that will handle DNA fingerprint data. The database management system ICIS (the International Crop Information System; http://www.cgiar.org/icis/homepage.htm) has been investigated for its potential to handle this type of information. This is a database system for the management and integration of global information on genetic resources and crop improvement for any crop. Currently, it is being used for crops such as maize, wheat, barley, rice, potato. ICIS was initially developed for the unique identification of germplasm, the management of nomenclature and the management of pedigree information. However, the scope of ICIS has now been expanded to also handle molecular marker data.

A new module, GEMS (the Gene Management System), has recently been developed and is currently being alpha-tested by users for its suitability to handle various types of molecular marker data. It is envisaged that it will form the basis of a new database for taro - MIRACLE (the Molecular Informatics Resource for the Analysis of Colocasia esculenta). This work has ceased since the departure of the postdoctoral research fellow, Dr Emma Mace, in September, 2001. It is not envisaged that any further progress will be made in the provision of this database, as the level of expertise and support is not sufficient for this to occur.

2.3 Technology Transfer

Dr Mohammed Wagih from Unitech in Lae, PNG spent 1 week in Ian Godwin’s laboratory at UQ to gain some hands-on experience with the techniques used in taro DNA fingerprinting. During this short time, he was able to perform DNA extractions, radiolabelled PCR to amplify SSR markers, agarose and polyacrylamide gel electrophoresis, and visualisation and scoring of gels. He was also able to hold discussions on further collaborative opportunities with UQ.

2.4 Publication of results

A book chapter has been published:
Godwin ID, Mace ES and Nurzuhairawaity, 2001. Genotyping Pacific Island Taro (Colocasia esculenta (L.) Schott) Germplasm. In: Plant Genotyping - the DNA fingerprinting of plants (Henry, R. ed). CAB International, Wallingford, England pp 109-128.

A journal article has been accepted for publication:
Mace, ES and Godwin, ID 2002. Development and characterisation of polymorphic microsatellite markers in taro, Colocasia esculenta (L.) Schott. Genome 45: (in press)

Please refer to year 3 summary.

This summary is taken from the final report.
Objective 1:

(i) Characterisation of all taro viruses and development of diagnostics

(a) Taro reovirus (TaRV): This newly discovered virus appears to have a genome comprising 10 segments of dsRNA. We have obtained the complete genomic sequences of segments 3 and 4 (S3 and S4), while partial sequences have been obtained for S1, S2 and S10. Based on comparisons with other reoviruses, degenerate primers were designed to conserved regions of S4 to amplify a 1.7 kbp fragment. We investigated the sequence variability in the genome of isolates from PNG, Solomon Islands, New Caledonia and Vanuatu TaRV in order to develop a more specific PCR test. Based on this data, a specific PCR-based diagnostic assay for this virus has now been developed.

(b)Taro vein chlorosis virus (TaVCV): We have cloned and sequenced the entire genome of a Fijian TaVCV isolate. The sequence comprises approximately 12,000 nt and shows most similarity to Sonchus yellow net virus (SYNV), a nuclearhabdovirus. Sequence variability studies were done on the L and capsid genes. A diagnostic PCR test has now been developed for this virus based on a 220 nt region of the L-gene.

(c) Colocasia bobone disease virus (CBDV): We have cloned and sequenced approximately half (~5000 nt) of a PNG isolate of CBDV. Analysis of this sequence indicates that CBDV is a distinct rhabdovirus from TaVCV and appears to be a cytorhabdovirus. A PCR-based diagnostic has been developed based on a region of “gene 3”.

(d) TaBV-like sequence: Preliminary characterisation and analysis studies have indicated that the TaBV-like sequences present in taro are integrated sequences. Further work is needed to determine whether these sequences can be activated and cause disease.

KEY IMPACTS AND PRACTICAL IMPLICATIONS: The characterisation of TaRV and the two rhabdoviruses has enabled the subsequent development of sensitive diagnostic tests for these viruses. As such, diagnostics tests have now been developed for all known viruses infecting taro. The availability of a suite of taro virus diagnostics will now enable taro germplasm to be virus-indexed, thus facilitating safe international movement of taro germplasm.

(ii) Virus Survey

Surveys were conducted in Vanuatu, Samoa, American Samoa, Fiji, PNG, Solomon Islands and New Caledonia. Samples were also provided from Micronesia and the Cook Islands. These samples have been indexed for all known viruses using the newly developed molecular-based diagnostic tests.

KEY IMPACTS AND PRACTICAL IMPLICATIONS: The virus surveys, conducted in countries wishing to share germplasm under the TaroGen project, provided updated information on virus distribution. This data, combined with the results from virus-indexing from the TaroGen germplasm collection, allowed countries to make informed decisions on the importation of taro germplasm.

(iii) Virus indexing of TaroGen taro germplasm collection held at SPC, Fiji

Approximately 450 tissue-cultured taro lines held in the germplasm collection at SPC have been sent to Brisbane for growing in AQIS-greenhouses and indexing. Of these, 159 have been indexed for each of the taro viruses according to an internationally-recommended schedule.

KEY IMPACTS AND PRACTICAL IMPLICATIONS: Safe international transfer of indexed taro germplasm will now be possible, allowing countries access to a diverse pool of germplasm with disease resistance and other agronomic qualities.

Objective 2:

(i) DNA fingerprinting of national taro collections

Taro collections from 9 PICs were DNA fingerprinted using radiolabelled SSR (Simple Sequence Repeat) markers. These markers were developed as a resource for taro research both within the region and internationally. Taro collections were made in most cases as part of the TaroGen germplasm collection. Entire collections from Fiji, Samoa, Tonga, Niue, Palau, Cook Islands were fingerprinted. A 20% sample of the country collection was fingerprinted from Papua New Guinea, Solomon Islands, Vanuatu and New Caledonia, as these collections were too large to fingerprint all accessions. There were some delays in receiving collections from some countries. Civil unrest, disease and cyclones prevented the Solomons collection from being received until 2002. Even then, this was not a representative country collection, rather it was separate samples from 3 provinces: Choiseul, Malaita and Temotu. The Guadalcanal collection was lost to virus and other diseases before samples were taken. We received 2 collections representing Samoa, one from the Botanic Gardens of the University of Hawaii, and one from MAFF in Samoa. In the case of the Solomon Islands, Fiji and the Polynesian countries, collections were received from the Regional Germplasm Collection at SPC as tissue cultures. In other cases, we received leaf samples directly from the country.

From the overall collection of 2206 accessions, 527 were DNA fingerprinted with SSR markers. The 7 SSR primer pairs used generated 38 alleles across the collection. SSR marker information was used to assess within and between country taro genetic diversity, and to identify a DNA fingerprint of accessions.

It was evident that most (if not all) of the genetic diversity within South Pacific taros could be sampled from PNG and the Solomons. Interestingly, 2 SSR alleles were found only in the Solomons. This was somewhat unexpected, and is a tantalizing hint at a localized adaptation or separate introduction within taros in the Solomons. It is worth noting that none of the PNG taros fingerprinted were from Bougainville, which is geographically closer to the Solomons than it is to other islands of PNG. There wer some quite diverse taros in the small sample of accessions from Palau, probably representing some more Asian taro types. Evidence from our study and that of the EU-TANSAO study suggest that there are 2 Centres of Diversity, one in PNG/Solomons and the other in Indonesia/Malaysia.

KEY IMPACTS AND PRACTICAL IMPLICATIONS: Countries have information regarding the diversity of the germplasm, and have DNA fingerprint information on important accessions. From a regional viewpoint, the major finding is that PNG and the Solomons hold most or all of the genetic diversity of the region. These should be seen as major sources of future genetic diversity for genetic improvement programs.

(ii) Rationalisation of taro germplasm to form a core collection

Based on DNA fingerprints, we selected a core collection for each country. The aim of the core collection is to reduce the size of the collection to about 10% of the total accessions while attempting to maintain at least 85% of the genetic diversity available. The core collection can then be conserved more easily and utilized more effectively, as these accessions can be more extensively characterized (for example for pest and disease resistances, corm attributes and response to abiotic stresses). Once core collections were selected for each country, the combined all country dataset was analysed to improve the level of overall genetic variation for the region. In total, 211 accessions make up the core collection, as described below:

Country/Region Number of Accessions

Papua New Guinea 83
Solomon Islands 48
Vanuatu 44
New Caledonia 8
Fiji 8
Palau 4
Niue 6
Samoa 4
Tonga 3
Cook Islands 3
TOTAL 211

The core collection is stored as in vitro tissue cultures, primarily at the Regional Germplasm Collection at SPC in Suva, Fiji. Duplicate collections are kept at USP Alafua Campus, Samoa, with plans to maintain a sample ate the International Potato Centre (CIP) in Peru, and negotiations are underway to maintain a duplicate collection in PNG.

KEY IMPACTS AND PRACTICAL IMPLICATIONS: The core collection is kept centrally at SPC and can be protected with DNA fingerprint information. Accessions of interest can be provided to interested parties subject to virus indexing. Requests for germplasm have already been received from Samoa, Marshall Islands, New Caledonia and Hawaii, and tissue cultures have been provided. In addition SPC has provided accessions to Australian researchers and grower in central and north Queensland. It is hoped that the RGC receives continued support for the maintenance of taro germplasm (both from TaroGen and TANSAO - which includes Asian taro types), as well as the other species in the collection (including sweet potato, yam, banana, kava and tannia).

Objective 3:

QUT Component
(i) Mr Macquin Maino from University of Technology, Lae, PNG, completed his Master of Science Degree within the Centre for Molecular Biotechnology, QUT. This project involved the development of diagnostic tests for Dasheen mosaic virus.
(ii) Mr Apaitia Macanawai from USP (Samoa) completed his Master of Agriculture Degree. Mr Macanawai was investigating the epidemiology of taro small bacilliform virus and examined alternative hosts, vectors and seed transmission.
(iii) The taro virus surveys were made with staff from research, extension and quarantine divisions. During the surveys, informal training was provided in virus symptom recognition, general pathology and control.
(iv) A “Taro Virus Diagnostics” Workshop was held at USP, Suva, from Sept. 1-4, 2003. This workshop was attended by ~20 participants from Fiji, Papua New Guinea and French Polynesia.

UQ Component
(i) Mr Tom Okpul from NARI, Bubia spent 3 months in 2000 at UQ learning the techniques of DNA fingerprinting. He applied these techniques to a diverse set of germplasm from the NARI breeding program. He has since moved to Vudal University in Rabaul, where he has instigated taro breeding.
(ii) Mr Robert Plak Pawilnga from Unitech, Lae visited for 2 week in 2001 and learnt techniques involved in database maintenance and analysis of DNA fingerprint data. He returned in 2003 where he is currently completing his Masters of Agricultural Studies. His research is focussed on tissue culture and genetic transformation of taro.
(iii) A DNA Fingerprinting Workshop was held at USP, Suva from 26-30 May, 2003. This involved theory, and hands-on lab experience in DNA extraction, gel electrophoresis, PCR and other marker techniques, scoring of fingerprint gels, analysis of molecular data and problem solving. The workshop was attended by 20 participants from Fiji, PNG, Solomon Islands and Samoa.
(iv) Basic DNA fingerprinting equipment is available at the Institute of Applied Science and School of Biology at USP, Suva. Scientists were trained in the use of this equipment, and a group of USP/SPC scientists have undertaken some DNA fingerprinting of the in vitro sweet potato germplasm collection. In addition, others from Fiji are using these techniques for coral identification and Tongan scientists want to perform varietal identification of vanilla accessions.

KEY IMPACTS AND PRACTICAL IMPLICATIONS: Mr Maino has returned to UniTech, Lae, where it is envisaged that he will play a key role in the continued development of the Agricultural Biotechnology Centre, and training of new scientists. The virus diagnostics workshop held at USP provided both theoretical and hand-on training in plant virus characterisation and diagnosis, using taro viruses as an example. SPC has been proactive in the development of an in-house virus indexing capacity in Fiji - this workshop provided both technology transfer and an opportunity to discuss issues such infrastructure, equipment and logistics. With the availability of DNA fingerprinting equipment, expertise and interest in Fiji, there are currently projects ongoing to fingerprint more taros, with a student from USP undertaking collections in Kiribati. As already stated, there is an ongoing effort to fingerprint sweet potato at the RGC and scientists form the regions are interested in applying the techniques to coconut, vanilla and kava.

Improved knowledge of virus epidemiology:

An awareness seminar was held at Unitech to discuss the aims and objectives of the project with selected farmers from several different regions around Lae. Farmers were enthusiastic and willing to participate - three sites have now been established to conduct on-farm trials. It was decided to delay the production of the flyer (showing virus symptoms and possible control strategies) until the later stages of the project when more information was available.

The vector transmission studies to determine the vector of TaVCV and to determine the role of viruses in Alomae disease complex have commenced. Cages have been constructed and insects have been collected from infected taro and placed on healthy taro. These plants are currently being observed for symptoms. A research technician and postgraduate student have been appointed to the project. The success of this aspect of the project relies on the implementation of taro virus diagnostic tests at Unitech. Most of the consumables for these tests have now been either received or are in the process of ordering. It is envisaged that known positive control samples will be tested for viruses in parallel in the Unitech and QUT labs to ensure quality control - this will commence in the next month.

Safe distribution of taro germplasm:

A research technician has been appointed to assist with the project. Again, this objective relies on the implementation of virus diagnostics at Unitech, which are in the process of being established.

Improved knowledge of virus epidemiology:
The majority of information dissemination activities are being deferred until the end of the project when more information has been obtained. Discussions and planning for these activities are currently in progress.

Known healthy, but disease-susceptible taro plants are being multiplied in tissue culture for use in transmission experiments. Potential virus vectors are being collected and placed on these healthy plants which are then being examined for virus symptoms.

Known TaVCV-infected taro are being multiplied in order to have sufficient numbers to conduct transmission experiments with Tarophagus.

Potential vectors of alomae disease and known virus-infected plants are being multiplied to enable tests to determine the etiology of alomae disease.

Safe distribution of taro germplasm:

This objective relies on the use of PCR-based virus diagnostics at Unitech. New primers and other reagents have been sent to Unitech from QUT. Despite initial difficulties with the diagnostics, it now appears that the problems have largely been overcome. Once the technique has been proven reliable, germplasm can be tested for viruses prior to release.

Project Outcomes

Characterisation of all taro viruses and development of diagnostics
(a) Taro reovirus (TaRV): Investigation of the sequence variability in the genome of isolates from PNG, Solomon Islands, New Caledonia and Vanuatu TaRV enabled researchers to gather sufficient data to develop a specific PCR-based diagnostic assay for this virus.
(b)Taro vein chlorosis virus (TaVCV): The researchers cloned and sequenced the entire genome of a Fijian TaVCV isolate. A diagnostic PCR test has now been developed for this virus based on a 220 nt region of the L-gene.
(c) Colocasia bobone disease virus (CBDV): The researchers cloned and sequenced approximately half (~5000 nt) of a PNG isolate of CBDV. A PCR-based diagnostic has been developed based on a region of ‘gene 3’.
(d) TaBV-like sequence: Preliminary characterisation and analysis studies indicated that the TaBV-like sequences present in taro are integrated sequences. Further work is needed to determine whether these sequences can be activated and cause disease.

Thus the characterisation of TaRV and the two rhabdoviruses has enabled the subsequent development of sensitive diagnostic tests for these viruses. As such, diagnostics tests have now been developed for all known viruses infecting taro. The availability of a suite of taro virus diagnostics will now enable taro germplasm to be virus-indexed, thus facilitating safe international movement of taro germplasm.

Virus survey
Surveys were conducted in Vanuatu, Samoa, American Samoa, Fiji, PNG, Solomon Islands and New Caledonia. Samples were also provided from Micronesia and the Cook Islands. These samples have been indexed for all known viruses using the newly developed molecular-based diagnostic tests. The virus surveys, conducted in countries wishing to share germplasm under the TaroGen project, provided updated information on virus distribution. These data, combined with the results from virus-indexing from the TaroGen germplasm collection, allowed countries to make informed decisions on the importation of taro germplasm.

Virus indexing of TaroGen taro germplasm collection held at SPC, Fiji Approximately 450 tissue-cultured taro lines held in the germplasm collection at SPC have been sent to Brisbane for growing in AQIS-greenhouses and indexing. Of these, 159 have been indexed for each of the taro viruses according to an internationally-recommended schedule. Safe international transfer of indexed taro germplasm will therefore now be possible, allowing countries access to a diverse pool of germplasm with disease resistance and other agronomic qualities.

DNA fingerprinting of national taro collections
Taro collections from nine Pacific Island Countries were DNA fingerprinted using radiolabelled SSR (Simple Sequence Repeat) markers. These markers were developed as a resource for taro research both within the region and internationally. Taro collections were made in most cases as part of the TaroGen germplasm collection. Entire collections from Fiji, Samoa, Tonga, Niue, Palau, Cook Islands were fingerprinted. A 20% sample of the country collection was fingerprinted from Papua New Guinea, Solomon Islands, Vanuatu and New Caledonia, as these collections were too large to fingerprint all accessions. There were some delays in receiving collections from some countries. Civil unrest, disease and cyclones prevented the Solomons’ collection from being received until 2002. Even then, this was not a representative country collection, rather it was separate samples from three provinces: Choiseul, Malaita and Temotu. The Guadalcanal collection was lost to virus and other diseases before samples were taken. Researchers received two collections representing Samoa, one from the Botanic Gardens of the University of Hawaii, and one from MAFF in Samoa. In the case of Solomon Islands, Fiji and the Polynesian countries, collections were received from the Regional Germplasm Collection at SPC as tissue cultures. In other cases, they received leaf samples directly from the country.

From the overall collection of 2206 accessions, 527 were DNA fingerprinted with SSR markers, which were used to assess within- and between-country taro genetic diversity, and to identify a DNA fingerprint of accessions. It was evident that most (if not all) of the genetic diversity within South Pacific taros could be sampled from PNG and Solomon Islands. Interestingly, two SSR alleles were found only in the Solomons. This was somewhat unexpected, and is a tantalising hint at a localised adaptation or separate introduction within taros in the Solomons. It is worth noting that none of the PNG taros fingerprinted were from Bougainville, which is geographically closer to the Solomons than it is to other islands of PNG. There were some quite diverse taros in the small sample of accessions from Palau, probably representing some more Asian taro types. Evidence from our study and that of the EU-TANSAO study suggest that there are two ‘Centres of Diversity’, one in PNG/Solomons and the other in Indonesia/Malaysia.

Rationalisation of taro germplasm to form a core collection
Based on DNA fingerprints, researchers selected a core collection for each country. The aim of the core collection is to reduce the size of the collection to about 10% of the total accessions while attempting to maintain at least 85% of the genetic diversity available. The core collection can then be conserved more easily and utilised more effectively, as these accessions can be more extensively characterised (for example for pest and disease resistances, corm attributes and response to abiotic stresses). Once core collections were selected for each country, the combined all-country data set was analysed to improve the level of overall genetic variation for the region. The core collection is stored as in vitro tissue cultures, primarily at the Regional Germplasm Collection (RGC) at SPC in Suva, Fiji. Duplicate collections are kept at USP Alafua Campus, Samoa, with plans to maintain a sample at the International Potato Centre (CIP) in Peru, and negotiations are under way to maintain a duplicate collection in PNG.

Regional expertise
This has been boosted with the training at Queensland University of Technology of Mr Macquin Maino from University of Technology, Lae, PNG, who completed his Master of Science Degree within the Centre for Molecular Biotechnology. He has returned to UniTech, Lae, to play a key role in the continued development of the Agricultural Biotechnology Centre, and training of new scientists. Also at QUT, and Mr Apaitia Macanawai from USP (Samoa) who completed his Master of Agriculture Degree. Mr Macanawai investigated the epidemiology of taro small bacilliform virus and examined alternative hosts, vectors and seed transmission. At University of Queensland Mr Tom Okpul from NARI, Bubia spent 3 months in 2000 learning the techniques of DNA fingerprinting. He applied these techniques to a diverse set of germplasm from the NARI breeding program. He has since moved to Vudal University in Rabaul, where he has instigated taro breeding. Also at UQ Mr Robert Plak Pawilnga from Unitech, Lae visited for 2 weeks in 2001 and learnt techniques involved in database maintenance and analysis of DNA fingerprint data. He returned in 2003 where he is currently completing his Masters of Agricultural Studies. His research is focused on tissue culture and genetic transformation of taro.

A virus diagnostics workshop held at USP provided both theoretical and hand-on training in plant virus characterisation and diagnosis, using taro viruses as an example. SPC has been proactive in the development of an in-house virus indexing capacity in Fiji - this workshop provided both technology transfer and an opportunity to discuss issues such infrastructure, equipment and logistics.

Project ID
CP/1994/043
Project Country
Inactive project countries
Commissioned Organisation
Queensland University of Technology, Australia
Project Leader
Associate Professor Rob Harding
Email
r.harding@qut.edu.au
Phone
07 3864 1379
Fax
07 3864 1534
Collaborating Institutions
Secretariat of the Pacific Community, Fiji
Ministry of Agriculture, Forests, Fisheries and Meteorology, Samoa
University of Queensland, Australia
University of Technology, Papua New Guinea
University of the South Pacific, Samoa
National Agricultural Research Institute, Papua New Guinea
Project Budget
$1,576,307.00
Start Date
01/07/1998
Finish Date
30/06/2001
Extension Start Date
01/07/2006
Extension Finish Date
30/06/2007
ACIAR Research Program Manager
Dr T K Lim