Comparative study of indigenous Vigna vexillata (L.) A. Rich. accessions from different latitudes in Indonesia and Australia

Introduction

Vigna vexillata (L.) A. Rich. is a tropical herbaceous legume, widely distributed throughout southern and eastern Africa, the Indian subcontinent, south-east Asia, Indonesia, Papua New Guinea and Australia (Verdcourt 1970; Vanderborght 1989). The species is considered to have two centres of diversity, namely Africa, and south-east Asia including northern Australia (Wong 1997; Garba and Pasquet 1998). James and Lawn (1991) reported that accessions from Africa and Australia were inter-fertile and described the inheritance of a range of traits.

The species is highly variable morphologically, but usually takes the form of a vigorous twining or scrambling vine with large, showy purple or purplish-yellow flowers, hirsute pods and foliage, and fleshy tuberous roots from which the plants perenniate. In parts of its range of occurrence, V. vexillata is collected from the wild, and in some locations is cultivated as a food crop, primarily for its tuberous roots, but occasionally for its pods and seeds. For example, it is used as a tuber and pulse in north-east India (Bhattacharyya et al. 1984) and in the foothills of Himalayan regions (NRC 1979).

Recently, Karuniawan et al. (2006) reported collecting cultivated forms of V. vexillata in several localities in Bali and West Timor in Indonesia, where it is reputed to be more adapted to drought compared with other traditional, non-legume, root crops. In Bali, the local people know the crop as Jempirangan, while in West Timor, it is recognized as Kamberiti. In the islands of Sumba and Flores, where the plant is called Oehala and Fanuatufui, the tubers are collected from plants in the wild. Wild forms of V. vexillata have also been used as a ‘bush tucker’ plant by Aborigines in Australia (Lawn and Cottrell 1988). It has also been used as a forage legume in Africa and Australia.

V. vexillata is of interest as a useful species in its own right, and because it is considered by some to be a ‘linking’ species with affinities to both the African Vignas (subgenus Plectrotropis) and Asiatic Vignas (subgenus Ceratotropis). Garba and Pasquet (1998) suggested it may be closely related to cowpea (V. unguiculata). Using in vitro embryo rescue techniques, Gomathinayagam et al. (1998) reported the successful recovery of V. vexillata × V. unguiculata hybrids. Evans (1975) reported partial development of pods from crosses between V. vexillata and mung bean (V. radiata). V. vexillata may therefore be considered part of the tertiary gene pool of several cultivated Vigna spp.

The African gene pool of V. vexillata has generally received more research attention than that from south-east Asia. The objective of the present study was to undertake a comparative study of the growth and development of wild accessions of V. vexillata from a wide geographical range of locations from eastern Indonesia to south-eastern Australia. In anticipation that accessions collected at different latitudes may show differential adaptation with respect to photoperiod, based on preliminary observations reported by Grant et al. (2003), the plants were exposed to a range of extended photoperiods during their growth. Many wild and cultivated Vigna species are quantitative short-day plants (Lawn 1979; Ellis et al. 1994; Rebetzke and Lawn 2006). 

Materials and methods

Germplasm
Twelve accessions of V. vexillata were chosen for study: two from Indonesia and 10 from Australia (Table 1). The accessions sampled naturally occurring populations at regular latitudinal intervals across a wide geographical range, from the low-latitude tropics in West Timor, Indonesia, to the subtropics in eastern Australia.

Cultural details
The plants were grown in cylindrical pots (330 mm diameter, 300 mm depth) filled with a high quality commercial potting mix, located in a field at James Cook University, Townsville, Queensland Australia (Lat. 19°20’ S; Long 146°46’ E). Prior to sowing, the seeds were scarified to overcome hardseededness, by removing a small piece of the testa with a scalpel. Several seeds were sown directly into each pot at a depth of 1.5 cm on 16 August (late winter) and excess seedlings later thinned to leave two plants per pot. Where initial germination was inadequate, seedlings were transplanted between replicate pots or re-sowing took place, to ensure two healthy plants per pot.

Each plant was provided with a 1.2 m bamboo stake for support, around which they were hand-trained during growth to prevent intertwining. Applications of Osmocote Plus® controlled-release fertilizer to each pot provided sufficient nutrients for vigorous growth. Plants nodulated freely without inoculation. Pots were watered daily by hand and via a water-filled saucer under each pot. Minimum night temperatures were initially cool (<12°C) and germination was slow (average 11 days). By the end of August, minimum night temperatures were >18°C and thereafter plant growth was vigorous.

Experimental treatments and measurements
In order to test the hypothesis of Grant et al. (2003) that accessions of V. vexillata from different latitudes are adapted to different day lengths, the plants were exposed to a range of artificial photoperiods using 100 w incandescent lights suspended above the plants, as described by Summerfield et al. (1993).

The photoperiod extension treatments were commenced on August 27, when seedlings of all accessions had emerged. Local day length (sunrise to sunset) at that time was 11 hr 38 min. Automatic timers were used to switch the lights on, to extend the natural day length pre-dawn, for 0 (control), 0.5, 1.0, 1.5 and 2.0 hours. The timers were adjusted twice weekly to compensate for the natural increase in day length during the experiment. The statistical design was a split plot with two replications.

Several attributes of plant growth and development were recorded on a per-pot basis during growth, including the dates of emergence, first flowering and maturity of the first flush of pods; the length and breadth of the uppermost fully expanded terminal leaflet at 30, 40 and 50 days after sowing (DAS); the number of trifoliolate leaves on the main stem at 30 DAS; the number of primary branches; and plant habit (1 = bushy, 5 = twining). Pods were collected as they ripened daily to avoid seed losses from dehiscence, and the numbers of pods per plant, seeds per pod, seed yield (g per plant) and 1000-seed weight were calculated from the cumulative harvests. The evaluation was terminated when the first flush of pods in each pot had matured, at which time the plants were destructively harvested and the oven-dry weights of the aboveground biomass and tuberous roots were recorded.

Results and discussion

Effects of day length extension
The plants flowered rapidly, within seven weeks of sowing in all treatments (Table 2), with no apparent effects of day length extension nor any apparent photoperiod × accession interaction effects.

However, there were apparent photoperiod effects on several attributes of growth and partitioning (Table 2). Firstly, there was a tendency for terminal leaflet size, as measured by length and breadth, to be larger with longer photoperiods. Terminal leaflet shape, as reflected in the length:breadth ratio, was unaffected.

There was also a trend for the longer days to stimulate vegetative growth but not reproductive growth (Table 2). The number of main stem leaves at 30 DAS and the aboveground vegetative biomass at maturity of the first flush of pods was highest in the longer day lengths, whereas the proportion of total biomass allocated to seeds, tubers and regenerative tissues in total (i.e. seeds + tubers) declined with longer day lengths. 

There were small but statistically significant interaction effects between accession and photoperiod extension treatment on several of the attributes shown in Table 2. However, there were no consistent patterns relating interaction effects to latitude of provenance.

The trend for longer days to stimulate vegetative development, relative to reproductive growth, is consistent with short-day photoperiodic response, and is characteristic of many summer-growing legume crops (e.g. Lawn 1989). The absence of any apparent photoperiodic effects on time to flowering was thus surprising.

The fact that the photoperiod treatments were effective in eliciting responses in terms of growth and partitioning, but not in flowering, may indicate that day length was not extended sufficiently beyond the prevailing short, late-winter – early-spring days to delay flowering. Alternatively, it is possible that floral induction had already occurred by the time the day-length treatments were started. In some plants, as few as one or two short days at the seedling stage are sufficient to irreversibly induce flowering (Galston and Davies 1970).

In their natural environments, most of the accessions would normally not germinate or re-sprout from tubers until after the early- to mid-summer rains that precede the summer wet season in late November to late December, by which time natural day lengths are longest. The plants then flower and set pods in the shortening days of late summer – autumn. There is evidence in crop legumes that post-flowering development is sensitive to shorter day lengths than is time to flowering (Lawn 1989).

Variation among accessions
Averaged over treatments, there was highly significant variation among the different accessions for a wide range of phenological and morphological traits (Table 3). While significant, the genotypic differences in time to flowering and to maturity of the first flush of pods were small relative to the related wild species V. radiata subsp. sublobata (Rebetzke and Lawn 2006). Flowering and podding also occurred rapidly in all accessions and treatments, relative to cultivated V. vexillata (Karuniawan et al. 2006). The rapid progress toward flowering, combined with the small variation among accessions, suggested that floral induction occurred rapidly and early in seedling development.

Plant habit ranged from sub-erect bush types, like ACC332 from Timor, to twining vines like ACC 336 from coastal Queensland (Table 3). Among accessions, there was a two-fold range in vegetative biomass and in seed yield, and a four-fold range in tuber dry weight per plant. In relative terms, the variation among accessions for pod and seed traits was not as great as that in seed yield and biomass.

Variation with provenance
There were similarities in some traits, and also some statistically significant differences, between the Australian accessions as a whole and the Indonesian accessions (Table 4), although interpretation of the differences needs to be tempered by the small number of Indonesian accessions sampled.

Average vegetative biomass production was similar between the two provenances (Table 4), whereas seed yield, tuber yield, pods per plant, seed size and the proportion of regenerative biomass (i.e. seeds + tubers) were all smaller in the Indonesian accessions. The Indonesian accessions were also more bush like, while more of the Australian accessions were twining vines.

Relations between traits and with latitude
Accessions that flowered earliest tended to flower for longer (r = 0.77**) so that the dispersion among accessions in days to maturity of the first flush of pods was smaller than that in the time to flower (Table 3). An analogous relation has been observed in V. radiata subsp. sublobata (Rebetzke and Lawn 2006).

Among accessions, those that produced more aboveground vegetative biomass (e.g. ACC 336, ACC 317 and ACC 383 – Table 3) tended to produce smaller seed yields (r = -0.76**). However, accessions that produced more aboveground vegetative biomass tended to produce more tuber dry weight (r = 0.63*).

Among accessions, seed yield per plant was positively related (r = 0.81**) to the number of pods per plant, but there were no consistent relations between seed yield per plant and either seed size or seeds per pod. Among pod and seed traits, there was a negative correlation (r = -0.63*) between the number of pods per plant and the number of seeds per pod. Pod length was positively related to the number of seeds per pod (r = 0.57*) while pod width was positively related to 1000-seed weight (r = 0.66*).

There was no apparent relation between growth habit score and other attributes reported in Table 3, not even the number of primary branches.

There were no consistent relations between phenology (Table 3) and latitude of provenance (Table 1). Nor were there any apparent relations between latitude and biomass, or latitude and seed yield. However, both the absolute tuber dry weight (r = 0.79**) and the tuber harvest index (r = 0.82**) tended to be greater in accessions from subtropical latitudes. This is again consistent with quantitative short-day photoperiodic response, with those accessions adapted to higher latitudes more readily forming tubers at given day length than accessions from the topics.

Acknowledgements

The provision to AK of an ATSE Crawford Fund post-doctoral training award at James Cook University (JCU) is acknowledged with appreciation. Special thanks also to Mr Chris Gardiner, International Postgraduate Liaison Officer, JCU School of Marine and Tropical Biology, for his helpful advice in obtaining access to resources and facilities.

References

Bhattacharyya PK, Ghosh AK, Sanyal B, Deb Ray G. 1984. Grow Vigna vexillata for protein-rich tuber-cum-pulse crop in North-eastern Hill Region. Seeds and Farms (India) 10:33–36.

Ellis RH, Lawn RJ, Summerfield RJ, Qi A, Roberts EH, Chay PM, Brouwer JB, Rose JL, Yeates SJ. 1994. Towards the reliable prediction of time to flowering in six annual crops. III. Cowpea (Vigna unguiculata). Experimental Agriculture 30:17–29.

Evans AM. 1975. Species hybridisation in the genus Vigna. In: RA Luse and KO Rachie (editors). Proceedings of the IITA Collaborators Meeting, Grain Legume Improvement. International Institute of Tropical Agriculture, Ibadan, Nigeria.  pp. 31–34.

Galston AW, Davies PJ. 1970. Control Mechanisms in Plant Development. Prentice-Hall Inc., New Jersey, USA.

Garba M, Pasquet RS. 1998. Isozyme diversity in Vigna vexillata (L.) A. Rich (Fabaceae) complex. South African Journal of Botany 64:163–175.

Gomathinayagam P, Ram GS, Rathnaswamy R, Ramaswamy NM. 1998. Interspecific hybridization between Vigna unguiculata (L.) Walp. and V. vexillata (L.) A. Rich through in vitro embryo culture. Euphytica 102:203–209.

Grant T, Lawn RJ, Bielig LM. 2003. Variation among Australian accessions of Vigna vexillata for traits of agronomic, adaptive or taxonomic interest. Australian Journal of Agricultural Research 54:243–250.

James AT, Lawn RJ. 1991. Inheritance of selected traits in accessions of Vigna vexillata (L.) A. Rich of Australian and African origin. Australian Journal of Botany 39:415–429.

Karuniawan A, Iswandi A, Kale PR, Heinzemann J, Grüneberg WJ. 2006. Notes on neglected and underutilized crops: Vigna vexillata (L.)A. Rich. cultivated as a root crop in Bali and Timor. Genetic Resources and Crop Evolution 53:213–217.

Lawn RJ. 1979. Agronomic studies on Vigna spp. in south‑eastern Queensland I. Phenological response of cultivars to sowing date. Australian Journal of Agricultural Research 30:855–870.

Lawn RJ. 1989. Agronomic and physiological constraints to productivity of tropical grain legumes and prospects for improvement. Experimental Agriculture 25:509–528.

Lawn RJ, Cottrell A. 1988. Wild mungbean and its relatives in Australia. Biologist 35:267‑273.

NRC [National Research Council]. 1979. Tropical Legumes: Resources for the Future. National Academy of Sciences, Washington DC, USA.

Rebetzke GJ, Lawn RJ. 2006. Adaptive responses of wild mungbean (Vigna radiata ssp. sublobata) to photo-thermal environment I. Phenology. Australian Journal of Agricultural Research 57(8):917–928.

Summerfield RJ, Lawn RJ, Qi A, Ellis RH, Roberts EH, Chay PM, Brouwer JB, Rose JL, Shanmugasundaram S, Yeates SJ, Sandover S. 1993. Towards the reliable prediction of time to flowering in six annual crops. II. Soybean (Glycine max). Experimental Agriculture 29:253–289.

Vanderborght T. 1989. Some observations on seedlings of Vigna vexillata (L.) A. Richard (Fabaceae). Bulletin du Jardin Botanique National de Belgique 59:179–187.

Verdcourt B. 1970. Studies in the Leguminosae – Papilionoideae for the ‘Flora of Tropical East Africa. IV.’ Kew Bulletin 24:504–555.

Wong KC. 1997. Vigna vexillata (L.) A. Richard. In: PROSEA, Plant Resources of South-East Asia No.11: Auxiliary plants. Prosea Foundation, Bogor, Indonesia. pp. 261–263.

Contact PGRN-manuscripts@cgiar.org about this page