Keywords

Genetic Resources

The importance of germplasm as a basic tool for crop improvement is well recognised. They provide the basic material for selection and improvement through breeding to ensure food security needs of the world’s rapidly rising population. Wild relatives and progenitors of cultivated plants together with semi-domesticates represent a strategic part of germplasm collections. Genetic variation is fast eroding as natural habitats of wild relatives of cultivated species are being destroyed. As the genetic base of modern varieties is narrow and variability fast eroding, introgression of genes from wild species can substantially influence the breeding progress. Generally, the Indian genetic resources of Momordica are not threatened.

Gene Pool Classification

Gene pool serves as a tool for conceptualising the ability of plant population to cross with the conspecific population and with those of other species (Harlan and de Wet 1971). Based on the available literature on inter-specific hybridization and evaluation of their progeny, the following gene pool classification of the cultivated/semi-cultivated species is proposed (Bharathi et al. 2012a). The nature of these relations is important for the application of appropriate technologies to transfer desirable genes from wild Momordica species to the cultivated/semi-cultivated species.

Gene pools of M. charantia. The primary gene pool of M. charantia is again divided into two subclasses I and II. Gene pool I consists solely of its various commercial cultivars and its interfertile variety, the wild-type var. muricata. There is evidence for introgression of M. charantia var. charantia genes into wild-type in Taiwan (Liao et al. 2012). There are a great many commercial cultivars with particular characteristics that together with local land races and populations of wild varieties constitute extraordinary genetic resources. The next level of compatibility involves the M. balsamina (wild species). Despite the high degree of morphological/cytological similarity between these two species, they are reproductively isolated from each other in terms of barriers to hybridisation and is very difficult to obtain hybrid seeds (that too only in one direction—M. charantia as seed parent) and therefore M. balsamina is placed in primary gene pool II. The dioecious species represent the tertiary gene pool.

Gene pools of M. dioica. None of the dioecious species is reproductively isolated from the other completely. The primary gene pool is represented by its land races/varieties and M. sahyadrica. The anthesis of M. dioica occurs in the evening while that of M. sahyadrica and their hybrid progeny in the morning, which can be explored to get greater pollinator choice. Its secondary gene pool includes M. cochinchinensis and M. subangulata subsp. renigera while all the monoecious species are included in tertiary gene pool.

Gene pools of M. subangulata subsp. renigera. As it is a tetraploid species and the rest of the dioecious species is diploid, the hybrid progeny are triploid and sterile. Therefore, primary gene pool includes only infra-specific types and secondary gene pool includes the rest of the dioecious species and tertiary gene pool is formed by monoecious species.

Genetic Erosion and Threat Status

The only reference to the threatened status of Momordica is found (Anonymous 1997) in ICUN Red Data Book where M. subangulata Blume. from Wyanad (Kerala) and south Canara (Karnataka) is accorded threatened-indeterminate status (taxa known to be extinct, endangered, vulnerable or rare but where there is not enough information to say which of the four categories is appropriate). The material referred to as M. subangulata from Kerala and Karnataka is actually M. sahyadrica and true M. subangulata is of restricted distribution in north–east India. Jha and Ujawane (2002) consider M. balsamina as nearing extinction in Saurashtra, Gujarat and M. cochinchinensis as endemic to Assam forests. However, M. cochinchinensis is not endemic to Assam as the authors have spotted the species in abundance in the north, south and middle Andamans and also there are reports of its distribution in a vast region in South–East Asia. Zuberi and Biswas (1998) reports M. dioica in Bangladesh in the endangered category. Dwivedi (1999) considers M. dioica as endangered in Madhya Pradesh. However, most of these reports are based on certain assumptions and do not have the support of authentic fieldwork. Recent studies revealed a grave threat for M. dioica in its entire range and M. sahyadrica in the Western Ghats of Kerala. Overall, M. charantia var. muricata faces a medium level of threat across its geographic range. Habitat loss and fragmentation brought about by population pressure and developmental activities, poor distribution and low population density of Momordica species coupled with inadequate in situ conservation efforts, and acculturation of the forest dwelling communities are the major factors attributed to their heightened threat status affecting their long-term survival in the wild (Joseph and Antony 2007).

Present Status of Germplasm Holdings

Bettencourt and Konopka (1990) have given a compilation of ex situ holdings of Momordica germplasm worldwide. A large collection of M. charantia is maintained in the national gene bank and by different organisations in India and in other countries (Table 7.1). Species representation of the genus Momordica in various herbaria/gene banks around the world is presented in Table 7.2. It seems that wild Momordica are underrepresented in gene banks. Recently, descriptors for dioecious Momordica spp. have been published (Joseph and Antony 2011). Evaluations of genetic resources for traits of horticultural interests are regularly conducted for yield and fruit quality or for pest and disease resistance.

Table 7.1 Present status of germplasm holdings in Momordica species
Full size table
Table 7.2 Species representation in various herbaria/gene banks around the world
Full size table

Plant Descriptors

Habit

All dioecious species [M. subangulata (subsp. renigera, subsp. subangulata), M. dioica, M. sahyadrica, M. cochinchinensis, M. rumphii, M. clarkeana, M. denticulata, M. denudata] and a monoecious species (M. cymbalaria) are perennial climbers with tuberous roots. Monoecious species, viz. M. charantia and M. balsamina are annuals with fibrous roots. Perennial species undergo dormancy during winter/summer months and new shoots are produced upon favourable conditions. However, in M. cochinchinensis, the aerial stem does not wither or dry up completely upon cessation of favourable growth season.

Seedlings

All species have distinct seedling morphology. Annual species have epigeal germination (Fig. 7.1a), whereas perennial species have hypogeal germination (Fig. 7.1b). Polyembryony was observed rarely in M. dioica, M. subangulata subsp. renigera and M. sahyadrica (Fig. 7.2). Robustness and size of the cotyledon was greater in M. charantia var. charantia and progressively reduced to M. charantia var. muricata and M. balsamina was most fragile. In the dioecious group, M. cochinchinensis is most robust and fast in emergence and has triangular non-cordate leaves. M. dioica and M. sahyadrica differ in lobing of first few leaves, M. dioica being more deeply lobed and very fragile.

Fig. 7.1
figure 1

Germination behaviour. a M. charantia showing epigeal germination, b M. subangulata showing hypogeal germination

Full size image
Fig. 7.2
figure 2

A polyembryonic seedling of M. dioica

Full size image

Roots

The annual species produce fibrous roots, which die at senescence along with the aerial parts. However, the perennial taxa produce storage roots with which they perennate during the unfavourable growth period. In M. sahyadrica and M. dioica, the seedling tap root gets thickened with the accumulation of food and secondary thickening and side roots which are formed from the base of the bulged part are fibrous and non tuberous. In the case of M. subangulata subsp. renigera tap root gets branched slightly below the caudex, gets swollen at intermittent places and undergoes repeated branching (Fig. 7.3a). The number of swollen tubers in the case of M. dioica and M. sahyadrica were one each (taproot sometimes forked), whereas in the case of M. subangulata subsp. renigera, it varied from 5 to 15. In M. cochinchinensis, the tap root and its primary branches becomes woody (Fig. 7.3b) and areal stem remains alive to a considerable height during unfavourable season. After a period of active growth, with the advent of unfavourable season for growth, the plants of these dioecious species show symptoms of senescence, leaves become yellow and dry up, vine also withers and the plant perennates with the help of storage roots underneath. In case of M. dioica the stem portion consisting of basal 2–3 nodes remains alive while in M. cochinchinensis all the nodes of the main vine remain alive with only reduction in new foliar growth, which upon favourable conditions, put forth branched sprouts. In case of M. sahyadrica, sprouts emerge from the root–shoot transition zone (caudex). In case of M. subangulata subsp. renigera, there is no polarity and specification; sprouts emerge from any part of the tuber surface, even from wiry roots.

Fig. 7.3
figure 3

Nature of roots. a Teasel gourd showing adventitious root tubers. b Woody roots of M. cochinchinensis

Full size image

Sexual Reproductive System

Most of the species (M. subangulata (subsp. renigera and subsp. subangulata), M. dioica, M. sahyadrica, M. cochinchinensis, M. denudata, M. denticulata, M. rumphii, M. clarkeana) are dioecious and only three (M. charantia, M. balsamina and M. balsamina) are monoecious. Occasionally, hermaphrodite flowers in M. subangulata subsp. renigera are observed in nature (unpublished).

Tendrils

Tendrils are simple and unbranched. However, in some wild varieties of M. charantia bifid tendrils (Fig. 7.4) are also observed. In M. cochinchinensis, tendrils are robust.

Fig. 7.4
figure 4

Bifid tendril of M. charantia

Full size image

Leaves

In Momordica the leaves are simple, with the blade either entirely or variously (deeply) lobed or (sub) pedately 3–5 foliate. The lobing may be variable within a species. The leaves of M. subangulata subsp. renigera are entire or angled while the leaves of other species are much dissected. However, in M. dioica mixed occurrence of entire as well as lobed leaves in the same plant has also been noticed. Umbilical glands in petiole and lamina base (which is present in M. cochinchinensis) act as a good taxonomic trait (Fig. 7.5).

Fig. 7.5
figure 5

Leaf of M. cochinchinensis with umbilical glands

Full size image

Inflorescences/Flowers

Male flowers are solitary or in short loose pseudo-racemes, each flower stalk with a persistent hooded bract (Fig. 7.6). Female flowers are solitary, in axils also with a conspicuous or rudimentary bract. Male flower pedicels are short or long; the receptacle is short, cupular (M. charantia, M. balsamina, M. dioica, M. sahyadrica, M. rumphii, M. clarkeana) or saucer shaped (e.g. M. cochinchinensis, M. subangulata subsp. renigera, M. denticulata) or obconical (M. cymbalaria) calyx lobe entire or scarious, adnate at base. Petals 5, free, entire; stamens 5, anthers 3, 1—one thecous, 2—two thecous, filaments very short, free inserted at mouth of the receptacle tube; thecae usually coherent, connective sometimes swollen, pistil lode absent. Female flowers calyx as in the male or distinct, petals as in the male, ovary oblong-fusiform, warty or soft papillose, ovules mostly many, horizontal, stigma 3-lobed; staminode absent.

Fig. 7.6
figure 6

Male flower of M. cochinchinensis with hooded bract

Full size image

Fruits

The fruit is fleshy, various in size and shape, pyriform, globose, ovoid or ellipsoid. Fruits are spiny (M. dioica, M. subangulata subsp. renigera, M. sahyadrica, M. cochinchinensis) or warty (M. balsamina) or tuberculate (M. charantia) or ribbed (M. cymbalaria, M. subangulata subsp. subangulata). The nature of epicarp is delicate in all the species except M. cochinchinensis which is shell like and leathery. The soft pulp inside the mesocarp cavity contains the seeds and has a scarlet red colour and slimy aril (Fig. 7.7) characteristic of the genus.

Fig. 7.7
figure 7

M. subangulata subsp. renigera fruit exposing the seeds with red coloured arils

Full size image

Seeds

Many enclosed in orange red sarcotesta or creamish yellow (aril), small or large, flattened or turgid on faces, smooth or sculptured margins often undulate and dentate. M. balsamina and M. cymbalaria stand out in its seed shape and sculpturing. M. charantia var. muricata has close resemblance to cultivated bitter gourd and is difficult to distinguish except for the small size. The dioecious group has a general resemblance, all being basically black and cog wheel shaped. Fresh seeds of M. dioica have golden striation on testa which fades away on drying. M. subangulata subsp. renigera had short rectangular seeds with six projections. M. cochinchinensis has the biggest seed with deep sculpturing and irregular projections on the sides in a broadly stellate fashion. The surface is flat without any sculpturing. Seeds of M. cymbalaria are ovoid-subglobose and obscurely sculptured and are different from the seeds of other Asiatic Momordica spp.

Descriptors and Descriptor States

Internationally recognised descriptor lists are published by Biodiversity International for major crops. However, there is no published descriptor for bitter gourd, teasel gourd, sweet gourd or spine gourd or any Momordica species by Biodiversity International. Only few references to Momordica descriptor are available (Srivastava et al. 2001; Rasul et al. 2004; Joseph 2005; Joseph and Antony 2011). As a preliminary step, sets of most significant descriptors (minimal descriptor list) for Momordica (Srivastava et al. 2001) were prepared under the National Agricultural Technology Project in which bitter gourd, sweet gourd and spine gourd were treated together. They, being evolutionarily divergent groups (bitter gourd on the one hand, sweet gourd and spine gourd on the other hand), should be treated separately as they vary by more than 75 % characters by virtue of their breeding behaviour and growth forms (Joseph 2005).

Rasul et al. (2004) proposed a descriptor with 29 morphological and physiological characters for M. dioica. Descriptor lists for monoecious species (Joseph 2005) and dioecious species (Joseph and Antony 2011) have been developed (Tables 7.3, 7.4, 7.5, 7.6 adapted from Joseph 2005; Joseph and Antony 2011) based on the observed variability in national collections (observable from herbarium sheets), published descriptions of these taxa in various flora together with ex situ study of germplasm collections comprising M. dioica, M. sahyadrica, M. subangulata subsp. renigera and M. cochinchinensis. Exploitation of some characters for inter- and infra-specific categorisation is based on the current state of the author’s knowledge of both levels of variation. Further collection and study of variability across the country will lead to spotting of more diverse types and accordingly the descriptor states need elaboration and modification. Present treatment of some traits such as leaf shape, fruit shape, etc., are not exhaustive as numerous types are difficult to describe in technical terms, but easy to depict through illustrations found in the existing collection itself.

Table 7.3 Descriptors and descriptor states for characterization of balsam pear and balsam apple
Full size table
Table 7.4 Descriptors and descriptor states for evaluation of balsam pear and balsam apple
Full size table
Table 7.5 Descriptors and descriptor states for characterisation of sweet gourds (to be used in continuation of passport data)
Full size table
Table 7.6 Descriptors and descriptor states for sweet gourds evaluation
Full size table

Collection, Regeneration and Maintenance

Lack of information about a taxon’s precise distribution in different ecosystems is a major constraint to biodiversity conservation (Arora 1998). The findings of the ecogeographical analysis give a clear-cut picture of areas of distribution, hotspots, infra-specific variability and phenology. The distribution maps give a holistic picture of the distribution of component taxa, areas of overlapping distribution and higher concentrations that need to be targeted for maximum assemblage of genetic diversity, using which a prospective collector can have access to the exact site.

Area-wise gaps in germplasm collection can be ascertained by comparing the gene bank passport data with the distribution maps. Analysis of species distribution maps based on herbarium survey and locality data of collections reveal the need for more intensive exploration in species hotspots. A good representation of diversity in M. charantia has been assembled in India through various explorations conducted by various organisations like NBPGR, New Delhi; Indian Institute of Vegetable Research (IIVR), Varanasi, Uttar Pradesh; Indian Agricultural Research Institute (IARI), New Delhi; Indian Institute of Horticultural Research (IIHR), Bengaluru, Karnataka; Vivekananda Parvathiya Krishi Anusandhan Shala (VPKAS), Almora, Uttar Pradesh; Tamil Nadu Agricultural University (TNAU), Coimbatore, Tamil Nadu; Kerala Agricultural University (KAU), Thrissur, Kerala, University of Agricultural Science, Dharwad, Karnataka; Acharya NG Ranga Agricultural University (ANGRAU), Hyderabad, Andhra Pradesh; Mahatma Phule Krishi Vidyapeeth (MPKV), Rahuri, Maharashtra; Govind Ballabh Pant University of Agriculture and Technology (GBPUAT), Pantnagar, Uttaranchal; Konkan Krishi Vidyapeeth, Dapoli, Maharashtra, etc., though there are still a few grey areas to be explored more intensely. However, other species and areas need extensive coverage. Similarly, more than 70 samples of teasel gourd were collected from north-eastern India by the Central Horticultural Experiment Station, Bhubaneswar, in collaboration with NBPGR, New Delhi, which represent good morphological variability.

Experience of germplasm collection of Momordica species across the Western Ghats revealed certain general factors affecting wild species survival. Momordica species were found to be subjected to varied types of threats such as changes in agricultural practices affecting species dependent on prevailing agricultural systems and other factors such as forestry plantations, monoculture practices, continuous weeding preventing reproductive maturity, pressure from introduced plants (smothering by Mekania micrantha, competition from Mimosa incisa, Lantana camera, etc.) and collecting for horticultural purpose thus leading to critically low population level with subsequent danger of breeding collapse (Rajashekaran et al. 2011).

Characterisation and Evaluation

The nature and magnitude of genetic diversity in any crop determines and often limits its utilisation in breeding programmes. Genetic diversity was studied by various authors using various tools and materials. Characterisation studies employing solely morphological methods typically focused on revealing valuable horticultural traits. Indian researchers gave considerable attention to the evaluation of local M. charantia germplasm with the goal of identifying valuable accessions for breeding which resulted in the development of many varieties across India. However, other Momordica species of Indian occurrence are not given due attention in collection and characterisation of germplasm and reports are scanty.

Morphological characters have been widely used to characterise the collections while lately, DNA markers are popular in these studies. Wide morphological variations have been reported in M. charantia accessions collected from six countries namely India, China, Japan, Taiwan, Thailand and USA (Kole et al. 2010), Asia (Marr et al. 2004; Dalamu et al. 2012); India (Sirohi and Choudhury 1983; Behera 2004; Yadav et al. 2008; Joseph and Antony 2009; Paul et al. 2010); Bangladesh (Islam et al. 2010); Thailand (Promote et al. 2011) and Romania (Botau et al. 2010). The accessions of M. dioica collected from eastern and northern India (Ram et al. 2001; Bharathi et al. 2005, 2010) and Bangladesh (Rasul and Okubo 2002; Rasul et al. 2004) also showed a considerable range of diversity in qualitative and quantitative traits. The above studies provide broad phenotypic species variation in morphological (qualitative and quantitative) characters like sex expression, growth habit, maturity, fruit shape, fruit size, fruit length, fruit colour, surface texture, number of fruits per plant, yield per plant, etc. The entities collected represented a wide range of variability from almost near wild-types, semi-domesticated to cultivated types.

Based on evaluation the accessions of bitter gourd viz. IC-44428B, IC-85604A, IC-85608BC, IC-85611, IC-85636, EC-110596 have been identified as high yielders (Ghosh and Kalloo 2000). At the Central Horticultural Experiment Station (CHES, IIHR), Bhubaneswar, India, 60 accessions each of M. dioica, and M. subangulata subsp. renigera and 8 accessions of M. cochinchinensis have been studied for morphological variability (Fig. 7.8) which lead to identification of two high yielding clones, viz., Arka Neelachal Sree and Arka Neelachal Gaurav in M. dioica and M. subangulata subsp. renigera, respectively, for commercial cultivation (Vishalnath et al. 2008a, b).

Fig. 7.8
figure 8

Variability for fruit shape in dioecious Momordica species. a M. dioica, b. M. subangulata subsp. renigera, c. M. cochinchinensis

Full size image

Isozyme variation in M. charantia germplasm supported a single domestication event but did not clarify the place of domestication (Marr et al. 2004). In domesticated M. charantia, the absence of multiple alleles at allozyme loci and fixation for the same alleles across a great geographical distance indicate that gene flow from wild M. charantia into the domesticate is rare. This suggests that the morphological variation is due to conscious or unconscious selection on a local scale, rather than to introgression with the wild form. In a genetic diversity study involving seven genera of the family Cucurbitaceae, isozymes could not distinguish between Momordica and Luffa (Sikdar et al. 2010).

Many molecular markers have been used to characterise Momordica germplasm including both plastid and nuclear markers. The random molecular markers like random amplified polymorphic DNA (RAPD) and inter simple sequence repeat (ISSR) are particularly useful for studying polymorphism and genetic diversity pattern in plant species where no genomic information is available. Species-specific fragments detected by RAPD and ISSR have potential applications in introgression breeding of Momordica and these markers can be utilised for inter-specific hybridisation followed by marker-assisted monitoring of introgression. A wider range of molecular diversity detected in various studies (Table 7.7) by both RAPD and ISSR markers reflected the presence of high level of genetic variation among the species. High level of polymorphism was detected in dioecious species than monoecious species (Bharathi et al. 2012b). Genetic affinities among the cultigens were defined by their geographic origin, suggesting that opportunities exist for broadening the existing Indian germplasm collection (Behera et al. 2008a). RAPD and ISSRs to describe patterns of genetic variation among seven species of Momordica gave similar results for each marker type (Bharathi et al. 2012b); however, ISSR was more effective than RAPD analysis at intra-specific variation studies in M. charantia (Behera et al. 2008b).

Table 7.7 Molecular characterization of Momordica accessions
Full size table

Apart from RAPD, ISSR other molecular markers such as simple sequence repeats (SSRs), sequence characterised amplified region (SCAR), amplified fragment length polymorphism (AFLP) have also been used to assess the genetic diversity of different Momordica species in India and Bangladesh (Table 7.7). Simple sequence repeats due to the advantage of high variability, co-dominance and ubiquity in eukaryotic genomes, have become a useful molecular marker in population genetic analysis (Walter and Epperson 2001). AFLP analysis was discriminatory and allowed for a more complete dissection of unique differences within and between collection sites (Behera et al. 2008a) while RAPD and ISSR were not able to uniquely discriminate (Dey et al. 2006; Singh et al. 2007). Recently, Wang et al. (2010) developed polymorphic microsatellite markers which will be useful to study the genetic diversity and population structure within and between M. charantia and other related species. Among the Asiatic Momordica species only M. charantia germplasm have been characterised for SSR variation (Kole et al. 2009; Wang et al. 2010).

Plastid markers are typically conserved, making them especially valuable for revealing phylogenetic relations at or above species level (Chung and Staub 2004). The first report investigating mt, cp and n DNA sequence analysis was presented by Schaefer and Renner (2010). They studied 122 accessions of 58 Momordica species including the Asiatic species and suggested that the genus consists of 11 well-supported clades and monoecy evolved from dioecy seven times independently.

Conservation Strategies

In agro biodiversity conservation, wild plant genetic resources have received relatively lesser emphasis and attempts to conserve them face considerable constraints basically due to lack of information about the biology and ecology of the species and their precise distribution in different ecosystems (Arora 1998). The presence of genetic resistance to pathogens in wild populations is a reason frequently cited for the importance of conserving the genetic diversity present in the wild ancestors of domesticated species. Genetic erosion is very high due to habitat destruction, spread of alien weeds and anthropogenic factors. In this context, conservation of genetic diversity in the genus Momordica assumes significance by virtue of being a wild relative of bitter gourd which is an important vegetable and has manifold medicinal uses, e.g. to treat cancer, diabetes, psoriasis and many infectious diseases. References to conservation of Momordica species are scanty. Neglect of genetic resources of wild species and semi-domesticates in ex situ gene banks has been a universal feature (Heywood 1998). Momordica species assume significance for conservation as crop relative, source of economic product of aesthetic and ecological interest, of potential horticultural value and as future raw material for the medicinal and pharmaceutical industry.

Tissue culture, pollen storage and in situ conservation can be a valuable conservation tool especially in species which are amenable to vegetative propagation, viz., spine gourd, teasel gourd, sweet gourd and M. sahyadrica. In vitro conservation was attempted in dioecious Momordica species, viz., M. subangulata ssp. renigera, M. sahyadrica and M. dioica (Rajashekaran et al. 2011). The species were established in vitro (MS medium supplemented with growth regulators) and the cultures could be maintained in vitro (standard culture conditions) for 6 months without any subculture. Application of cryogenic techniques for conserving nuclear genetic diversity of rare, endangered and threatened plant species sourced from wild habitats would enable extended use of the male gametophyte for providing access to the conserved nuclear genetic variability, biotechnology research besides genetic enhancement of derived crops. The pollen of M. dioica can be stored at 0° C for 45 days (Islam and Khan 1998) but pollens showed little tolerance under long-term freezing conditions (−5 °C). However, the pollen viability was determined based on acetocarmine staining which is not a vital stain (Lebeda et al. 2006). Cryopreserved pollen (−196 °C) of M. dioica and M. sahyadrica showed 67–74 % germination after 48 h (Rajashekaran et al. 2010).

By establishing a few genetic reserves in selected protected areas in the Western Ghats, North–East and Andaman Islands Momordica species can be afforded in situ protection. Good populations of M. balsamina thrive in Machia safari park, Jodhpur, Rajasthan, India. Artificial seeding and in situ protection in sacred groves, especially for M. dioica needs consideration in the light of its endangerment especially in coastal lowlands in Kerala. Several tribal families across India were found to grow various species of wild Momordica in their homesteads in a simulated in situ condition. Often in the case of M. dioica and M. sahyadrica, the planting material, i.e. tuber is collected from the forest. M. charantia var. muricata being exclusively seed propagated, domestication attempts have progressed further. Hence, the conservation of semi-domesticates and pre-domesticates in home gardens is a viable option.

Momordica species including balsam pear, balsam apple, spine gourd and sweet gourd are treated as ornamentals in Europe and America, where it is grown in glasshouses since Victorian times for its beautiful foliage, pendant orange ripe fruits embedded in green foliage and star-like configuration of bursting fruits (Walters and Decker-Walters 1988; Robinson and Decker-Walters 1997). Miniature fruited M. charantia var. muricata and M. balsamina have beautiful foliage and orange red fruits. M. dioica has musky scented flowers and M. sahyadrica has large showy yellow flowers in profusion, besides both have ivy-like beautiful foliage and pendant fruits turning orange and bursting in star-like configuration. All this offers scope for adoption by urban gardeners, thus giving another dimension to on-farm conservation.

It has been observed that in primitive societies, gathering of wild vegetables is usually done by women. Often they do this while collecting firewood or fodder, which is a regular work, carried out by tribal women. On-farm conservation is carried out by them intentionally or unknowingly. As it is always the women who cook food, it is she who disburses mature or ripe seeds, some of which germinate and develop as new plants.

A careful breeding strategy involving extensive field survey in the fruiting season followed by rescue collection and seed multiplication in on-farm sites and a subsequent ex situ approach is needed for the conservation of variability in semi-domesticate landraces of M. charantia var. muricata. Artificial seeding and rehabilitation in sacred grooves may be attempted for M. dioica in coastal Kerala. Establishment of genetic reserves inside protected areas must be attempted for conserving diversity in M. dioica and M. sahyadrica in the Western Ghats. Ex situ conservation in home gardens and on-farm conservation in tribal homesteads in forest pockets is a viable option for conservation of Momordica gene pool as the taxa are still wild or semi-domesticate with high dependence on biotic agents for pollination and seed dispersal. Popularisation as ornamental plants and kitchen garden vegetables will enhance survival of the taxa and establishment of farms for tuber production will reduce pressure on wild population.

The study of genetic diversity, population ecology and conservation of Momordica species is inadequate and limited. As all wild Momordica species are potential vegetables besides genetic resources of bitter gourd, sweet gourd and spine gourd, IPGRI through AVRDC should initiate a collection and ex situ conservation programme for all the Asiatic wild Momordica species. A good representation of diversity in M. charantia has been assembled from India though there are still a few grey areas to be explored more intensely. However, other species and areas need extensive coverage. In the absence of any earlier attempt to collect and conserve this diversity, immediate steps need to be taken in this direction. This perhaps also serves as introspection to the poor state of wild Momordica gene pool collection and conservation in the National Agricultural Research System (NARS).

There is a need to conserve the highly heterozygous germplasm of dioecious species by establishing field gene banks. Further, under MTA, it should be made available to gene banks across South and SE Asia for domestication and utilisation. M. clarkeana, M. denticulata, M. rumphii (all SE Asia-Malesia) and M. denudata (Sri Lanka) need special attention. AVRDC/National agricultural research agencies should develop a strategy for (a) an update of conservation (ex situ) status of wild Momordica genetic resources, (b) ex situ regeneration protocol for rare endemics, (c) regulated supply of genuine planting materials to researchers across nationality borders under MTA, (d) clear and concise distribution maps for individual species based on field and herbarium survey and (e) a database on ethno-botanical uses of various species by aboriginal people.