1 Genetic resources of taro (Colocasia esculenta (L.) Schott) in Papua New Guinea: A review Okpul, T.1, D. Singh1,3, M.E. Wagih2 and D. Hunter3 Summary The status of taro, Colocasia esculenta (L.) Schott, genetic resources and the efforts of the National Agricultural Research Institute, together with several regional networks, in the acquisition, conservation, evaluation and utilization of the germplasm are reviewed. Wild and cultivated genotypes show enormous genetic variation suggesting Papua New Guinea (PNG) is a center of diversity. However, traditional varieties are being displaced due to biotic and abiotic factors. A representative sample of the genetic diversity of taro in PNG has been collected and is maintained ex situ. The need to develop alternative conservation strategies for efficient management and utilization is considered. Keywords: Taro, Colocasia esculenta, genetic resources, Papua New Guinea. Introduction Taro, Colocasia esculenta (L.) Schott, is an edible aroid belonging to the Araceae family. It is an antique staple root crop in Papua New Guinea (PNG). Taro cultivation at Kuk in the Western Highlands Province, for example, has been dated to be 9, 000 B.P. (Golson 1977). It has been suggested that it was the principal crop in the highlands prior to the introduction of sweet potato (Ipomoea batatas) 300-500 years ago (Clarke 1977; Bayliss-Smith 1982, 1985). With a long history of association with human society, taro has emerged as an extrinsic symbol in traditional beliefs and rituals among many ethnic groups (Barrau 1965; Panoff 1972; Barth 1975; Morren and Hyndman 1987). Taro is cultivated as an important staple up to an altitude of 2, 200 m (Bourke et al. 1998). It is grown as a minor food crop as high as 2, 740 m (Bourke 1982). It is primarily grown for its edible corm. Leaves, inflorescence and tender inner layers of petiole sheaths are also used, to a lesser extent, as vegetables. It is now being cultivated as a semi-commercial crop with surplus produce sold at about 4.00 PNG-Kina (US$1.20) per kg (Anonymous 2001). Annual production of taro is estimated at 170, 000 tons from an area of 31, 000 ha (FAO 2001). 1National Agricultural Research Institute, WLMP-Bubia, P. O. Box 1639, Lae, Papua New Guinea; 2Unitech Biotechnology Center, Papua New Guinea University Of Technology, PMB, Lae, Papua New Guinea; 3Secretariat of the Pacific Community, Suva, Fiji 2 Taro production, however, has been declining in the recent past as reported by Wagih et al. (1994). Taro leaf blight (TLB), caused by Phytophthora colocasiae Racib., taro beetles (Papuana spp.), the Alomae – Bobone virus complex (ABVC) and declining soil fertility impact negatively on yield. Their impacts have contributed, in part, to the dwindling production (Sar et al. 1998). Changing dietary habits and preferences for exotic foods, and also the introduction of other crop species with better comparative advantages such as Chinese taro (Xanthosoma sagittifolium) and sweet potato have also had negative influences on taro production (Waddell 1972; Bourke 1982; Joughin and Kalit 1986). The impact of these factors on production has unfortunately led to loss of traditional cultivars in some parts of PNG (Wagih et al. 1994). This paper reviews research efforts to preserve the existing taro genetic diversity by collecting and conserving farmer’s varieties and evaluation and utilization of the genetic resources to improve production in PNG. Past efforts, current status and future prospects are discussed, highlighting collaboration with regional networks. Taro diversity Taro is postulated to have originated in the Indo-Malayan region, either between Eastern India and Bangladesh (Plucknett 1983; Purseglove 1988) or in Southern China (Cable 1984). The origins of wild taros and their domesticated relatives and the directions of dispersal remain debatable (Matthews and Terauchi, 1994 cited by Yen, 1995). Three hypotheses have been proposed for PNG taro germplasm. Firstly, it may be an eastward extension of the Indo- Malayan flora suggesting a single center of origin in South Asia (Plucknett et al. 1970; Leon 1977; Yen 1982; Coates et al. 1988). Secondly, it may have been dispersed by early migrants and other seed vectors (Okada and Hambali 1989; Pawley and Green 1973). And thirdly, it may have resulted from an independent origin in PNG suggesting multiple origins of the crop (Matthews 1990, 1991; Lebot 1999). Nevertheless, PNG is considered as a centre of diversity (Leon 1977). Three botanical varieties of C. esculenta are prevalent in PNG. Variety aquatilis is the commonly occurring wild type (Matthews 1991, 1987), whereas variety esculenta (dasheen type) and variety antiquorum (eddoe type) are cultivated in gardens, either in swiddens or irrigated plots. A third cultivated form expresses intermediate traits between the dasheen and eddoe types and is believed to be a hybrid (Lebot et al. 2000). The distribution of the cultivated forms depends to some extent on their status as a component of the cropping system practised, where taro is cultivated either as a dominant or co-dominant staple (Figure 1). The dasheen types are commonly grown throughout the country. The eddoe and the intermediate types, on the other hand, are rarely grown in higher altitude areas and in locations characterized by distinct dry seasons but where taro is an important staple such as in parts of Central, East Sepik, Madang, Milne Bay, Morobe, Northern, West Sepik and Western Provinces. 3 Numerous varieties exist in gardens (Panoff 1972; Rangai 1982; Morren and Hyndman 1987). The variability may be attributed to sexual recombination, and perhaps somatic mutation, associated with continuous vegetative propagation and the subsequent selection by farmers based on adaptability and culinary qualities from exotic and novel varieties. Taro flowers and sets seed naturally. Protogyny and self-incompatibility systems in the inflorescence facilitate cross-fertilisation, which usually results in variable progenies (Shaw 1975; Okpul and Ivancic 1996; Johnston and Gendua 1998). Farmers have developed practices to identify, evaluate and select new varieties from natural crosses. Such practices have been observed in Pomio in West New Britain Province (Panoff 1972) and in Manus, Morobe and the Western Highlands Provinces. On Manus Island, after the rainy season farmers search for seedlings of new varieties in ant nests on small trees and in bamboo patches, and in soils shifted by landslides. Similarly, in Morobe and the Western Highlands Provinces, seedlings germinating in drains or ditches and alongside creeks and streams are nursed and culinary qualities are assessed at maturity (T. Okpul, personal observation). Continuous vegetative propagation has been observed to complicate colour or pigment patterns resulting in the creation of numerous morphotypes (T. Okpul, personal observation). Exchange of such variants between communities may have resulted in new identities. A general indication of variability within the genepool can be traced from vernacular names. Novel varieties are usually named after the founder or are given names relating to culinary qualities such as texture and aroma, or even names that are analogous to human anatomy relating to the shape or pigmentation of plant parts, while introduced varieties are given the names of their place of acquisition. Clonal variants (or even unidentified but morphologically similar varieties), on the other hand, are given additional names that describe the phenotypic variation. For example, in the Morobe Province, a popular variety, Numkowec, has three variants: Numkowec-koko, Numkowec-sisip, and Numkowec-yangyang, which are distinguished by their red, white and green basal ring colors, respectively. The variability has been indicated in diversity studies based on quantitative traits (Ivancic et al. 1995), isozyme variations (Lebot and Aradhya 1991), variations in ribosomal and mitochondrial DNA restriction sites (Matthews 1990), Randomly Amplified Polymorphic DNA (Irwin et al. 1998), Amplified Fragment Length Polymorphisms (AFLP) (van Eck et al. 1998) and microsatellites (Mace and Godwin 2000; Godwin et al. In press). These studies have shown that a higher level of genetic diversity exists in PNG as compared to neighbouring Pacific Islands. Nevertheless, and as exemplified by Waddell (1972), the displacement of taro is a growing concern (Wagih et al. 1994; Kesavan and Aburu 1982). During 1998 the present authors observed villagers in Lababia (population of >1000), Morobe Province abandon taro cultivation in favor of sweet potato as a result of high infestation by taro beetle and epidemics of TLB. Consequently, all traditional varieties (estimated to be over 40) selected over the years under a taro-based farming system are now displaced. Such a situation raises the need to explore, collect and safeguard the existing genetic diversity for potential use. 4 Acquisition and conservation Collection of germplasm representing the genetic diversity is a prerequisite for its effective study, conservation and utilization for crop improvement. Past collections of cultivated and wild taros were conducted on an ad hoc basis by various researchers in the Department of Agriculture and Livestock (whose research functions are now under the National Agricultural Research Institute (NARI)), and PNG University of Technology (UniTech). A total of 461 accessions were collected and maintained at the Highlands Agricultural Experiment Station (48), the Lowlands Agricultural Experiment Station, LAES (120), Laloki Research Station (135), Saramandi Research Station (28), Kuk Agricultural Research Station (10), and UniTech (120) (Aburu 1980; Jackson 1994; Levett et al. 1985). Amalgamation of the remnants of these collections at Bubia Agricultural Research Center, now Wet Lowlands-Mainland Program (WLMP) of NARI, formed the basis for the national taro germplasm collection (Arura 1985; Kambuou 1995). Other institutions were involved on a number of occasions. The International Plant Genetic Resources Institute (IPGRI), formerly the International Board for Plant Genetic Resources (IBPGR), funded several collection missions in the 1980s. Additional explorations resulted in the collection accruing 602 entries by 1989 (Akus et al. 1989), which were maintained in an ex situ collection. A lot of accessions were lost before being characterized and evaluated, the losses mainly attributed to inadequate level of husbandry in terms of weed control, and insect pest and disease management. Natural calamities such as floods and prolonged dry periods also had impacts on the field genebank, which was reduced to 437 accessions by 1995. Frequent staff turnover resulted in the loss of the passport data and field plans, which eventually invalidated the remaining entries and impeded efforts to re-collect (Kambuou 1998). Such errors have proved expensive and unsustainable (Godden 1999). From these experiences, the concept of a core collection and the use of complementary conservation strategies, especially in vitro techniques, for efficient conservation and utilization of taro germplasm were necessitated. Recently, local cultivars were re-collected throughout the country under the auspices of several regional networks: the Pacific Agricultural Regional Program (PRAP) of the Secretariat of the Pacific Community (SPC), the European Commission–funded Taro Network for Southeast Asia and Oceania (TANSAO) and the Taro Genetic Resources: Conservation and Utilization project (TaroGen) funded by the Australian Agency for International Development (AusAID). A total of 859 accessions from 16 provinces (Figure 1) have been collected, maintained ex situ in an augmented block design, and characterized. A core collection has been developed and a duplicate is being conserved in vitro in the Regional Germplasm Center (RGC), Fiji (TaroGen, 2001c). International germplasm exchange can play an important role in broadening genetic diversity. This is an approach being emphasized by the TaroGen and TANSAO networks. The current germplasm collection also holds several exotic accessions from the Asia-Pacific region (Table 5 1). Recently, 134 accessions included in the TANSAO regional core collection were acquired. This set of lines is currently being maintained as tissue cultures under quarantine and will be virus-indexed and cleaned before adding to the field collection (Gunua et al. 2000). Variety identification and evaluation Agro-morphological characterization Taro exhibits a wide array of agro-morphological polymorphism. Numerous variable but clonally stable traits are being used as markers for varietal identification and assessment of genetic diversity. The first descriptor list for taro, developed by IBPGR (IBPGR 1980), was used in earlier attempts to characterize the germplasm (Levett et al. 1985; Akus et al. 1989). An edited version, developed by IPGRI in collaboration with TaroGen (IPGRI 1999), is currently being used. Additionally, the TANSAO network developed a descriptor list using major agro-morphological markers for use in all of the partner countries for selection of national core samples to form a regional core collection (TANSAO 1998). A PNG core sample comprised of thirty-one varieties (Table 2) was selected based on the principal component score strategy from an initial collection of 279 accessions. All accessions collected and maintained ex situ have now been characterized using the IPGRI descriptor list under TaroGen. The selected core sample from the TANSAO project has also been included. The data matrix was subjected to diversity analysis for the selection of a tentative core collection comprising 20% of the most diverse varieties. This work was conducted with assistance from IPGRI. Molecular characterization Lebot and Aradhya (1991) evaluated a total of 452 PNG accessions using isozyme markers. However, the information was not used to rationalize the collection due to loss of passport data and the field map. Although numerous diversity studies included PNG accessions (Matthews 1990; Irwin et al. 1998; Mace and Godwin 2000; Godwin et al. In press), the germplasm was not systematically characterized for varietal identification to rationalize the collection and to enable thorough assessment of the extent of genetic diversity. The national core sample developed under TANSAO has been fingerprinted using AFLP at Wageningen Agricultural University and the information was used to compare genetic diversity among the accessions of the region (TANSAO, 1999, 2000). While the core collection developed by TaroGen was fingerprinted at the University of Queensland using ISSR and SSR markers under the Taro virus indexing and DNA fingerprinting project funded by the Australian Center for International Agriculture Research (ACIAR). A finalized 10% of the whole collection comprising 84 varieties of the most diverse accessions was selected from the tentative core 6 collection to form the national core collection (Mace et al. 2001; TaroGen 2001a, 2001b, 2001c). Agronomic evaluation Limited evaluation work has been conducted on yield and screening for disease resistance. Evaluations of varietal yields were, in most cases, preliminary and were sometimes inconclusive with respect to making recommendations of promising varieties for farmer adoption (Levett et al. 1985; Akus et al. 1989). In exploiting the germplasm for varietal resistance, Hicks (1967) screened for resistance against P. colocasiae Racib. and noted varying levels of resistance among tested accessions. However, no further work was pursued until recently, when Kokoa and Darie (1994) screened for and identified three resistant genotypes, namely Ph-15, Ph-17, and Ph-21, as possible donors for resistance against P. colocasiae. The collection is currently being systematically evaluated under the breeding program for other agronomic traits. Initial evaluations are based on the augmented block design in which performances of accessions within the collection are compared prior to formal testing and release for use (Figure 2). The selected TANSAO core samples are also being evaluated for agronomic traits and quality in formal trials (Gunua et al. 2001). Use of conserved germplasm Germplasm can be utilized directly in the form of better performing landraces or farmer varieties or indirectly as donors of useful genes for the development of novel varieties. The germplasm has not been systematically evaluated for its direct use. Genetic improvement, on the other hand, was attempted over the years. Earlier taro breeding work was conducted at LAES. The Department of Agriculture (Anonymous 1941) reported on single plant selections that developed two high yielding varieties, ‘Utility’ and ‘King’, yielding up to 17.6 t/ha. No further improvement work was conducted until the 1980s when breeding against TLB was first attempted. A resistant Thai wild type identified as ‘Bangkok’ was used in crosses involving promising cultivars. This program was, in fact, an extension of Dr. G. V. H. Jackson’s work in the Solomon Islands. However, no improved lines were released from LAES (Gunawardhana, 1984). 7 Recently, three PNG resistant genotypes, Ph-15, Ph-17 and Ph-21, together with ‘Bangkok’ were used as donor parents in the base population of a modified recurrent selection procedure adopted for taro improvement at WLMP. Several cultivars (both indigenous and exotic) were used as recurrent parents in selection for yield and culinary qualities (Ivancic and Okpul 1996, 1997). The breeding work is currently in its fourth cycle (or generation). In the first cycle, selection was directed towards TLB resistance because of the strong influence of wild germplasm on yield and culinary quality (Okpul et al. 1997). A set of 12 lines was finally selected for intermediate trials, which resulted in the recommendation of seven lines (Table 3) for evaluation in national multi-location trial (Singh and Okpul 2000). These lines were evaluated in seven different sites throughout PNG (TaroGen 2001). Further, multi-character selection in the third cycle population has resulted in the selection of 49 lines, which are being bulked for further testing in intermediate trials. The fourth cycle population is currently being evaluated on a single plant basis. The genebank has also provided a means for interested farmers to recover their lost varieties. Many farmers have also used the opportunity to diversify the number of varieties used in their gardens. International students and researchers have also accessed the germplasm (Table 4). The regional approach to germplasm conservation being undertaken will ensure easier access by scientists and other bonafide users. Future prospects Despite waning production, taro is still a staple of high market value in PNG. This highlights its potential as a commodity on which to base economic development in rural areas. However, taro is a difficult crop to grow because of its high requirements for soil nutrients, moisture and labour. It also has numerous pests and diseases, which can have a drastic impact on yield. These factors have, cumulatively, caused farmers to lose interest in taro farming. As observed, the abandonment of taro cultivation is a serious threat to traditional varieties. In order to renew farmers’ interest and prevent genetic erosion, more research needs to be conducted in developing options to circumvent production constraints. Major factors limiting production are TLB, taro beetle, declining soil fertility and ABVC together with its vectors, particularly Tarophagus spp. and Aphis gossypii. Other pests and diseases of some concern to growers include the dasheen mosaic virus, leaf defoliators (especially, Hippotion celeria L. and Spodoptera litura) and the nematode (Hirschmaniella miticausa). Several leaf blight resistant lines have been developed and are in the process to be released as recommended varieties to farmers. Nevertheless, ABVC and taro beetle still remain as major constraints to production. Jackson and Gollifer (1975) have reported on varieties expressing varying levels of resistance to ABVC. The study identified varieties of the ‘male types’ (no relation to sexuality but generally larger plant forms) to be susceptible while the smaller ‘female types’ express some level of resistance. These observations need to be further investigated, although the situation is complicated by the fact that several viral particles are involved whose combinations are usually manifested in varying symptoms (Rodoni et al. 1994). 8 Active search for satisfactory control measures for taro beetle has been concentrated in the last decade and is still continuing. Several biological control agents and chemicals have been identified in laboratory studies but their practical relevance is yet to be established (Thistleton et al. In press). Although the beetle has a wide host range, and varietal resistance does not seem to exist in taro as it attacks all varieties, it does not attack taros cultivated in flooded fields. This indicates the potential for using permanent flooding to circumvent beetle damage. Flooded or paddy culture may be an effective control measure, but it is not feasible in most areas of the ruggedly undulating terrain of the country. Besides, cultivars in PNG are mostly adapted to rain- fed cropping systems and usually do not tolerate moisture levels above saturation point. Hence, genotypes tolerant to flooded conditions need to be identified from the germplasm collection. Taro breeding, as with other crop species, is strongly dependent on the availability of genetic resources. Collection and conservation of germplasm is therefore vital and a fair representation of the diversity has been collected from most provinces. A core collection of maximal diversity has been selected. In essence, the core collection was developed to ease management and access to conserved genetic diversity, but the germplasm, however, is conserved ex situ, where it is continuously exposed to biotic and abiotic stresses. Hence, complementary conservation strategies need to be developed to avoid costly errors of the past. Medium to long-term storage in vitro should be considered. Although a duplicate of the core collection will be maintained in vitro in the RGC (Fiji), it is imperative to adopt such a complementary strategy nationally for the ease of domestic access. Additionally, the feasibility of other options such as in situ, on- farm, conservation needs to be investigated. Development of efficient screening methods and evaluation of the germplasm for varietal resistance against the main pests and diseases, and for adaptability, particularly to permanent flooding, is of high priority. Long-term strategies have to be developed for pragmatic exploitation of the germplasm. Acknowledgement The support from the Taro Genetic Resources: Conservation and Utilisation project (TaroGen), funded by AusAID; the Taro Network for Southeast Asia and Oceania (TANSAO), funded by the INCO – DC program of the European Commission; and the Secretariat of the Pacific Community – Pacific Regional Agricultural Program, is gratefully acknowledged. The authors wish to thank Messrs J. Maro, M. Tina, Dr. G. Wiles and the Australian Contribution to National Agricultural Research Support Project for their technical assistance. References 9 Aburu, K. 1980. Germplasm at Lowlands Agricultural Experimental Station, Keravat, Papua New Guinea. International Board on Plant Genetic Resources South East Asia Newsletter 4(3): 6-9. Akus, W. L., T. Pisea, M. Arura, E. Ososo, and J. Bokosou 1989. Taro germplasm collection and maintenance. In Bubia Agricultural Research Centre, Annual Report. Department of Agriculture and Livestock, Lae. Anonymous 1941. Department of Agriculture Annual Report. New Guinea Agriculture Gazette 7(2): 76-116. Anonymous 2001. Market Prices. Fresh Produce News 150: 14-29. Arura, M. 1985. Taro germplasm collection. In Bubia Agricultural Research Centre, Annual Report 1984. Department of Agriculture and Livestock, Lae. Barrau, J. 1965. L’humide et le sec. Journal of Polynesian Society 74: 329-346. Barth, F. 1975. Rituals and knowledge among the Baktaman of New Guinea. Yale University Press, New Haven. Bayliss-Smith, T. P. 1982. Taro (Colocasia esculenta) in the New Guinea highlands. Pp. 134- 147, in Proceedings of the Second Papua New Guinea Food Crops Conference, Part One, R. M. Bourke and V. Kesavan (eds). Department of Primary Industry, Port Moresby. Bayliss-Smith, T. P. 1985. Pre-ipomoean agriculture in the New Guinea highland above 2000 metres: some experimental data on taro cultivation. Pp. 285-320, in Prehistoric Subsistence Agriculture in the Tropics, E. S. Farrington (ed.). BAR International Series 232, Oxford. Bourke, R. M. 1982. Root crops in Papua New Guinea. Pp: 51-63 in Proceedings of the Second Papua New Guinea Food Crops Conference, Part One, R. M. Bourke and V. Kesavan (eds). Department of Primary Industry, Port Moresby. Bourke, R. M., B. J. Allen, P. Hobsbawn and J. Conway 1998. Agricultural Systems of Papua New Guinea. Working paper No. 1. Australian National University, Canberra. Bryne, P. N. 1984. Department of Primary Industry, Crop Research Report (July 1969 to December 1982). World Bank/ Papua New Guinea Agricultural Support Services Project, Port Moresby. 10 Cable, W. J. 1984. Spread of taro (Colocasia sp.) in the Pacific. Pp: 28-31 in Edible Aroids. S. Chandra (ed.). Oxford University Press, Oxford. Clarke, W. C. 1977. A change of subsistence staple in pre-historic Papua New Guinea. Pp. 159-163, in Proceedings of the Third International Symposium on Tropical Root Crops, C. L. A. Leaky (ed.). International Society for Tropical Root Crops and International Institute for Tropical Agriculture, Ibadan. Coates, D. J., D. E. Yen and P. M. Gaffy 1988. Chromosome variation in taro, Colocasiaesculenta: implications for origin in the Pacific. Cytologia 53: 551-560. FAO 2001. FAO Statistical (FAOSTAT) database. Http://apps1.fao.org. Godden, D. 1999. Economic rationale for cooperation in germplasm conservation. Paper presented at ACIAR/NARI Workshop on Regional Co-operation in Conservation and Management of Plant Genetic Resources in the Pacific (30-31 March 1999). Lae. Godwin, I. D., E. S. Mace and Nurzuhairawaity In press. Genotyping Pacific Island taro (Colocasia esculenta (L.) Schott) germplasm. In Plant Genotyping – the DNA Fingerprinting of Plants, R. Henry (ed.). CAB International. Golson, J. 1977. The making of the New Guinea Highlands. Pp: 45-56 in The Melanesian Environment, J. H. Winslow (ed.). The Australian National University Press, Canberra. Gunawardhana, S. A. 1984. A review of taro breeding, pests and diseases. Unnumbered, in Food Crops Advisory Meeting (2/3rd October 1984) Report. J. George and D. Moles (eds). Bubia Agricultural Research Center, Lae. Gunua, T. G., T. Okpul and D. Singh 2001. Activities of TANSAO in Papua New Guinea. 2000 Annual Report. WLMP, National Agricultural Research Institute, Lae. Hicks, P.G. 1967. Resistance of Colocasia esculenta to leaf blight caused by Phytophthora colocasiae. Papua New Guinea Agriculture Journal 19(1): 1-4. IBPGR 1980. Descriptors for taro (Colocasia esculenta). International Plant Genetic Resources Institute, Rome. IPGRI 1999. Descriptors for taro (Colocasia esculenta). International Plant Genetic Resources Institute, Rome. Irwin, S. V., P. Kaufusi, K. Banks, R. de la Pena and J. Cho 1998. Molecular characterization of taro (Colocasia esculenta) using RAPD markers. Euphytica 99: 183-189.
Description: