Bitter Melon in Australia

A report for the Rural Industries Research and Development Corporation

by Dr Wendy Morgan and Professor David Midmore Plant Sciences Group

Central Queensland University Rockhampton Qld 4702

November 2002

RIRDC Publication No 02/134 RIRDC Project No UCQ-10A

ii

© 2002 Rural Industries Research and Development Corporation. All rights reserved. ISBN 0642 58533 4 ISSN 1440-6845 Bitter Melon in Australia Publication No. 02/134 Project No. UCQ-10a The views expressed and the conclusions reached in this publication are those of the author and not necessarily those of persons consulted. RIRDC shall not be responsible in any way whatsoever to any person who relies in whole or in part on the contents of this report. This publication is copyright. However, RIRDC encourages wide dissemination of its research, providing the Corporation is clearly acknowledged. For any other enquiries concerning reproduction, contact the Publications Manager on phone 02 6272 3186. Researcher Contact Details Professor David Midmore Plant Sciences Group Central Queensland University Rockhampton QLD 4702 Phone: 07 49 309770 Fax:07 49 309255 Email:d.midmore@cqu.edu.au

Internet: http://science.cqu.edu.au/psg/ In submitting this report, the researcher has agreed to RIRDC publishing this material in its edited form. RIRDC Contact Details Rural Industries Research and Development Corporation Level 1, AMA House 42 Macquarie Street BARTON ACT 2600 PO Box 4776 KINGSTON ACT 2604 Phone: 02 6272 4539 Fax: 02 6272 5877 Email: rirdc@rirdc.gov.au. Website: http://www.rirdc.gov.au Published in November 2002 Printed on environmentally friendly paper by Union Offset

iii

Foreword While Australian rural industries and research institutions are typically based more on a competitive spirit than on active cooperation and coordination, there are a number of instances where a particular industry or section of an industry has adopted a successful collaborative approach to the development of new export or domestic industries Collaborative research amongst groups of specialists located across the length and breadth of Australia has the potential to be an effective catalyst to the development of cooperative arrangements for continuous supply of perishable vegetables to satisfy domestic and export markets. This publication, one of a number arising from the RIRDC project UCQ-10a, presents data from cooperative trials on bitter melon run in the Northern Territory, Queensland and New South Wales, and illustrates how various institutes can cooperate in such ventures. This project was funded from RIRDC Core Funds which are provided by the Federal Government, with varying levels of co-funding from the cooperating institutes. This report, a new addition to RIRDC’s diverse range of over 800 research publications, forms part of our Asian Foods R&D program, which aims to support industry in its drive to develop new products and markets and to gain competitive advantage through improving productivity in, and achieving price premiums for, Australian production. Most of our publications are available for viewing, downloading or purchasing online through our website: · downloads at www.rirdc.gov.au/reports/Index.htm

· purchases at www.rirdc.gov.au/eshop

Simon Hearn Managing Director Rural Industries Research and Development Corporation

iv

Acknowledgements The sub-project team members: CQU, Rockhampton, Qld. Rachel Brimblecombe Tom Osborne D. BIRD, Darwin, NT Mark Traynor Peter Hopkinson Emma Maroulis Greg Owens NORADA, NSW David Hicks Craig Harman Dr David Gallagher, for most of the Literature Review (current knowledge) and Peter McLaughlin for his practical information generously contributed to the review. Mr Tony Byrne, RIRDC Asian Food Program Manager for continued interest and support of the project and its aims. Dr Nguyen Quoc Vong, for practical advice to NORADA and DBIRD, Darwin. Participating institutions in UCQ-10a Contact person: Agriculture WA, Bunbury Ms Vynka McVeigh Central Queensland University, Rockhampton Professor David Midmore Northern Rivers Agricultural Development Association, Murwillumbah Mr Peter McLaughlin Northern Territory Department of Business Industry and Resource Development, Darwin Ms Melinda Gosbee NRE, Institute for Horticultural Development, Knoxfield Dr Wendy Morgan NSW Agriculture, Gosford Dr Nguyen Quoc Vong Queensland Department of Primary Industries, Mareeba Lester Loader Tasmanian Institute for Agricultural Research, Dr Frank Hay Burnie University of Queensland, Gatton Assoc. Prof. Alan Wearing

v

Contents Foreword ............................................................................................................................................... iii

Acknowledgements.............................................................................................................................. iv

Contents ................................................................................................................................................. v

Tables and Figures............................................................................................................................... vi

Executive Summary ............................................................................................................................ vii

1. Introduction........................................................................................................................................ 1

1.2 Current Knowledge ...................................................................................................................... 2

2. Materials and methods ................................................................................................................... 12

2.1 Replicated field trials................................................................................................................... 12

2.2 Observation trials ........................................................................................................................ 13

3. Results.............................................................................................................................................. 15

3.1 Replicated field trials................................................................................................................... 15

3.2 Observation field trials ............................................................................................................... 20

4. Discussion ...................................................................................................................................... 24

5. References ....................................................................................................................................... 27

6. Appendix 1 ....................................................................................................................................... 29

vi

Tables and Figures Table 1. Bitter melon market preferences. Table 2. Seasonal margins for bitter melon planting in Australia. Table 3. Respiration and ethylene production of selected Asian vegetables. Table 4. Northern Territory production and value of selected Asian vegetables. Table 5. Bitter melon cultivars: average fruit weight, size and Brix 0 over a five week harvest,

Northern Territory, 1999. Table 6. Bitter melon cultivars: fruit appearance over a five week harvest period, Northern

Territory, 1999. Table 7. Bitter melon cultivars: fruit appearance rated with BMPDL, Northern Territory,

1999. Table 8. Bitter melon cultivars: yield parameters, Northern Territory, 2000. Table 9. Bitter melon cultivars: yield and quality characteristics, Burringah NSW, 1998/99. Table 10. Fruit characteristics and water use of bitter melon cultivars grown hydroponically on

two trellis types, Rockhampton, Qld., 2000. Figure 1. Price of bitter melon at Flemington Markets, Sydney, January 1996- June 1999. Figure 2. Weekly harvest yields, Northern Territory, 2000. Figure 2. Cumulative harvest yield, Northern Territory, 2000. Figure 4. Correlation of fruit number with maximum and minimum temperature for trellis type

(a) and cultivar (b).

vii

Executive Summary Bitter melon (Momordica charantia L.) is frequently found growing in tropical and sub-tropical climates, and is consumed in most Asian countries. It is also produced in relatively smaller volumes in Australia, with winter production in the Northern Territory (NT), Queensland and northern Western Australia and summer production steadily southwards to include New South-Wales (NSW) and Victoria (Vic). Varieties used in Australia are generally open-pollinated (OP) and selected by growers, even though hybrid varieties are available from overseas. In order to strengthen the industry in Australia, from 1999 to 2001 a series of trials were established to identify national production capability, and supply and productivity issues for supply chain development of domestic and future export markets. This report presents the outcomes of those trials, and a brief review of relevant literature and market situations. Bitter melon is harvested at a physiologically immature stage, before the true onset of ripening. It is important, if fruit is to be transported overland (taking from 1-5 days from origin to market place), that fruit be selected that have not started the physiological process of ripening, for they will produce ethylene that hastens ripening of adjacent stored fruit. Yields of around 30 t/ha in the NT, and 25-80 t/ha in northern NSW are in line with reported yields overseas. Hybrid varieties did not out-yield OP varieties, but tended to have more fruit set and lower individual fruit weight. This inverse relationship was evident in Queensland too. The height at which fruit set in NT may have been responsible for the recorded individual fruit weights of hybrids not reaching their claimed size, but effects of planting density may also have been involved in this anomaly. A period of about three weeks from fruit set to first harvest was evident in Queensland (and reported for the Philippines), and earliest harvest in NT was 56 days after sowing pre-germinated seed, in line with the 60 days after sowing in Rockhampton (Qld). Premature maturity (ie before fruit reaches a marketable size generally considered to be 18-22 cm long) occurred more frequently in OP than hybrid varieties in NT. The main harvest peak in NT was 11-12 weeks after planting, which coincided with the flush of flowering and fruit set on primary lateral branches. Trellising to spread the canopy and to contain growth to primary and secondary branches is important for bitter melon production, and a small trial at Rockhampton showed that overhead trellising was superior to vertical (fence) trellising. The former made for easier harvest (fruit were more visible), less fruit blemish and better fruit set and aeration for diminished disease outbreak. Based on the trials reported here, the recommended cultivars for NT are Moonlight, Baizin, Long Type and the white-fruit cultivar Moon Beauty (although market acceptance of this cultivar is uncertain), and for northern NSW Baizin and Moon Beauty.

1. Introduction Bitter melon is a member of the cucurbit family commonly grown in tropical and sub tropical countries. There are a number of other commonly called melons such as winter melon, hairy melon, long melon. In this publication we are concerned only with bitter melon, Momordica charantia. In Australia bitter melon is grown in the Northern Territory, Queensland and New South Wales and a small amount in Western Australia. Small amounts are also grown in polyhouses in Victoria and South Australia to meet local market demand. Most is grown by Vietnamese growers and hence the most common varieties are those favoured by Vietnamese consumers which are pearly green and have a very rough or bumpy skin. The bumps are usually small to medium size and are irregular in size and shape. It is particularly difficult to determine when to harvest bitter melon. Consumers eat bitter melon when the fruit is physiologically immature or unripe and therefore when discussing bitter melon, maturity refers to maturity for eating quality. Maturity is very hard to judge from the outward appearance of the fruit but external fruit colour of the whole fruit can be used. This needs to be complemented with some assessment of seed coat colour which is a better indicator, but obviously has disadvantages as a tool in decision making. The seed coat must be creamy or pale green-brown with over maturity indicated by any shade of pink seed coat. Fruit continues to mature after harvest. Therefore fruit for local markets should be harvested when mature (still physiologically immature) for eating and fruit for long distance transport (interstate or export) when immature. Sections 4.4 and 4.5 in the Bitter Melon Quality Description Language Manual, BMQDL, (Vujovic et al. 2000) have colour photos which show these colour maturity indicators. Most bitter melon grown in Australia is open pollinated with seed usually selected by individual growers to suit their customers’ requirements and their production methods. There are hybrid varieties available in Australia and from overseas which have the potential to increase yield and hence profitability. It is unknown whether hybrids have better yields and meet market requirements or whether they have the same maturity issues. This report draws together research undertaken at regional, state and national level between 1999 and 2001. The aim of the research was to identify national production capability and supply and productivity issues for supply chain development of domestic and future export markets. Recommendations for the highest yielding cultivars which meet market, particularly Flemington Market and Melbourne Markets, requirements are made.

2

1.2 Current Knowledge Bitter melon, Momordica charantia L. is a member of the Cucurbitaceae family. It has also been called:

Citrillus lanatus (Thunb.) Matsum. & Nakai Momordica chinensis Sprengel Momordica elegans Salisb. Momordica indica L. Momordica muricata

Bitter melon is known by many common names both between and within countries: African cucumber English alligator pear English amargoso Filipino ampalaya Filipino assorossie French balsam pear English balsamapfel German balsamina Spanish balsamini longa Portugese balsamini lunghi Italian bitter cucumber English bitter gourd English bitter melon English bitterer Balsamkürbis German bittergourd English calabaza africana Spanish caranza Italian carilla gourd English Chinese melon English cundeamor Spanish daun peria (leaves) Malaysian daun peria katak (leaves) Malaysian estropajo Spanish foo gwa yip (leaves) Chinese foo gwa Chinese fu kua Chinese fu kwa Chinese haix Laotian jin li zhi Chinese karela Danish, Indian karella English kaveli Indian khô qua Vietnamese kor-kuey Chinese ku gua Chinese kugua Chinese la khô qua (leaves) Vietnamese la kwa Vietnamese lao pu tao Chinese maha Thai maiden apple West Indian

3

maiden's blush West Indian mara Thai margase French melão de São Caetano Portugese momordique a feuilles de vigne French mreah Khmer muop dang Vietnamese niga-uri Japanese palia Filipino pare pahit Indonesian pare Indonesian paria Filipino, Indonesian paroka French paya-aki Thai peria katak (short variety) Malaysian peria laut Malaysian peria Indonesian, Malaysian periok Malaysian phakha Thai pomme de merveille French pomo balsamo Italian reishi Japanese s'aix Laotian springkomkommer Dutch tsuru reishi Japanese vite indiana Italian wunder-Balsampfel German

Bitter melon is thought to have been domesticated in east India and southern China (Yang and Walters 1992) and is a commonly grown in South East Asia, India, South America, East Africa and the Caribbean. It is used as a food and a medicine. The immature fruit is valued for its bitter flavour and is considered to have an enormous range of health benefits. Cooking styles vary between regions, and produce can be sold fresh, pickled, canned in brine or dehydrated. Young shoots, leaves and flowers are also eaten, for their flavour and for medicinal purposes.

The fruit is rich in vitamins A, B and C and minerals such as calcium, phosphorous, potassium and iron (according to the Philippines Food and Nutrition Research Institute (1980), in Rasco and Castillo (1990)).

People of different nationalities have their own distinctly favoured shape and size of bitter melon, requiring different varieties (Table 1).

Table 1: Bitter melon market preferences (Peter McLaughlin 1998, pers. comm.).

Country Size (length. x diam., cm) Description Japan 35-45 x 5-7 Long, skinny and very bitter. Vietnam 16 x 4-5 Pearl coloured round, rough skin China 20-26 x 5-7 Dark green, smooth skin

4

Cultivars Most of the bitter melon produced in Australia is open pollinated (usually the Vietnamese type) with many growers selecting their own seed of these open pollinated varieties to produce their customers’ requirements. More recently, as a result of this project, growers are adopting hybrid cultivars which have greater yields and can meet customer expectations. Considerable improvement in quality, consistency and yield can be achieved with hybrid seed. Both hybrid and open pollinated seed is available through Australian seed companies, though the range is limited and hybrid seed is usually obtained from overseas. A white cultivar is also available.

Hybrid cultivar (seed catalogue) descriptions For seed supply contacts see Appendix 1. Kiew Yoke 59 Productive cultivar with long, smooth light green fruit. Fruit weight 500-600 g. Kiew Yoke 68 Vigorous, disease tolerant cultivar with large, broad shouldered, glossy fruit. Fruit weight 500-600 g. good shelf life and recommended for raining season production. Known You Green A Taiwanese cultivar with a smooth, shiny, beautiful green skin. Plants are early, vigorous and prolific. Fruit has ribbed stripes and weighs 400-700 g. The flesh is green and mildly bitter. Fruit is good for salad and frying. Verdure An early, vigorous, productive, high yielding cultivar. Fruits are short, beautifully shaped with a maximum weight of 500 g. Its green skin and light green flesh are suitable for stewing. Moonrise Plants are early, vigorous and prolific. Fruit are long shaped, with light green skin and flesh and weigh up to 700 g. It has an excellent crispy and tender taste and is ideal as a fresh vegetable or for stir frying. Moonlight An early, prolific cultivar which produces medium-long fruit that seldom crack. Fruit weigh up to 650 g. and have light green skin and flesh. This cultivar is suitable for use as a fresh vegetable and for stir frying.

Moon Beauty Plants are early, vigorous and high yielding. Fruit are oblong shaped and have shiny white skin with a wart like surface. Moon Beauty fruit are 30 cm long and 9 cm width and weigh 700 g. It has thick and crispy flesh with great taste. New South Wales Local OP selection The variety called Local OP, used as a “control” in trials in NSW and NT is the F8 selection of a cross between a locally grown Vietnamese type (of unknown origin) and Okinawa (Kurname Vegetable Seed Co., Japan). It is not commercially available.

5

Production Weed Wild varieties exist in Australia, which have much smaller fruit than the commercial varieties. Weedy forms of bitter melon have escaped ornamental cultivation and become a problem in citrus groves of Florida (Robinson and Decker-Walters 1996).

Climate / season Bitter melon is grown commercially along the Australian east coast from Melbourne to Cairns, and in the Northern Territory (Lee 1995, Table 2). It is tolerant to a range of environments (Lim 1998) and can be grown in tropical and subtropical climates (Reyes et al. 1994). Two crops are produced annually in Florida (Lamberts 1992).

Table 2: Seasonal margins for bitter melon planting in Australia (E: early, L: late, +: all).

Planting date Location J F M A M J J A S O N D Victoria + . . . . . . . . + + + Gosford . . . . . . . . . + + + Murwillumbah . . . . . . . . E + + + Darwin . . E + + + + L. . . . . Kununurra + + + +

In Murwillumbah bitter melon is harvested from early January through to May, in Darwin from May to October and in Kunannura, June to September. Temperature A minimum temperature of 18 C during early growth is preferred (Larkcom 1991), with 24-27 C being optimum (Desai and Musmade 1998). This range has been confirmed at Murwillumbah, with prolific growth occurring at day / night temperatures of 28-35 / 20-25 C and severe reduction in growth with night temperatures of 16 C or less (Peter McLaughlin 1998, pers. comm.). It requires more heat than the other gourds to produce maximum yield (Larkcom 1991), but is also more resistant to lower temperatures (Desai and Musmade 1998).

pH Optimum is 6.0-6.7 (Desai and Musmade 1998).

Soil type Bitter melon grows best in a well drained sandy loam, rich in organic matter, but will tolerate many soils (Cantwell et al. 1996, Reyes et al. 1994).

Soil preparation It is preferable to grow the crop on raised beds or ridges (Lim 1998). Chicken manure may be dug in and watered 2-4 weeks before transplanting or sowing but can be associated with food safety issues. Base fertiliser should be applied to the top of the bed (Traynor et al 2001. In press).

Germination Seeds rapidly lose viability, so pre-germinating is recommended for all but very fresh seed to ensure that only viable seed is planted. Soak for 24 hours and wrap in damp paper towelling, then keep in a plastic bag at 26-29 C for germination within a couple of days (Larkcom 1991). Germination ceases outside the range of 10-50 C (Singh 1991). Germination percentage can be increased by soaking in 1% KNO3 (Devi and Selvaraj 1994). Sow germinated seeds into pots, or directly in the field if the temperature is warm (Larkcom 1991). Seedlings emerge 5-7 days after sowing (Reyes et al. 1994). They require protection from wind, chill and excessive sun, and water should be prevented from collecting nearby (Desai and Musmade 1998).

6

Seed weight Weight varies considerably, with reports from 6,000 seeds/kg (Desai and Musmade 1998) to 17,000 seeds/kg (Reyes et al. 1994).

Plant density Plant spacing of 50 cm to 1 m between plants and 2-3 m between rows are used in Australia (Traynor et al. 2001 In press). Optimum plant density varies with cultivar, from 6 500 to 11 000 plants/ha (Reyes et al. 1994) or 20 000 plants/ha (Huyskens et al. 1992).

Transplanting It may be necessary to grow the first plants of the season in a hot bed to ensure enough warmth for immediate germination, and then transplant them at the 4-6 leaf stage (Peter McLaughlin 1999, pers. comm.).

Trellising Bitter melon is grown on 2 m high fences or overhead trellising in Darwin. Higher yields are obtained with 2 m than 1 m high trellises (Abusaleha and Dutta 1994) and the crop is more accessible. Overhead or T-trellising may increase the proportion of marketable fruit (Huyskens et al. 1992). See section on trellises.

Remove lateral branches below the first production wire of the trellis, and remove the tip of the main runner when it reaches the top wire, to induce early cropping. Removal of lateral branches in the first ten nodes has a positive effect on total yield.

Without pruning, most of the female flowers occur between the 10th and 40th nodes, or at 0.5 - 2 m height (Rasco and Castillo 1990).

Grafting Luffa provides an excellent root stock for bitter melon, and grafting can increase yields substantially in Taiwan, due to Fusarium wilt control (Lin et al. 1998). Luffa rootstock is also more resistant to flooding (Liao and Lin 1996).

Nutrition If soil is highly fertile and prepared with enough organic matter, further feeding may not be necessary (Larkcom 1991). Add 2 L/ha Nitrofoska monthly after planting until the plant reaches trellis height, then potassium nitrate (at 50 kg/ha ) up to flowering. Reduce nitrogen application during fruit set as nitrate is known to suppress flowering in many species. The addition of calcium nitrate at 50 kg/ha until flowering improves shelf life (Peter McLaughlin 1998, pers. comm.). N:P:K basal applications of around 100:50:50 kg/ha respectively, are recommended by Robinson and Decker-Walters (1996).

Water Irrigate at least weekly, beginning from the day of sowing (Desai and Musmade 1998). Tensiometers are used and watering is applied to maintain 10-20 centibar tension in the root zone (Traynor 2000 pers. comm.). The plant is intolerant to flooding (Reyes et al. 1994), with 4 days of flooding producing significant changes in morphology (Liao and Lin 1994).

7

Flowering Bitter melon is monoecious with male and female flowers borne separately on the same plant. Yield is determined by the number of successfully pollinated female flowers. Bitter melon will produce its first flower at 45-55 (Reyes et al. 1994) or 35 days (Rasco and Castillo 1990) after sowing if conditions are optimal, and continue throughout a season of usually 6 months (Reyes et al. 1994). Pollination can be achieved with bees or, in prolonged overcast weather, by blowing with an empty mister. Pollen looses viability as the day advances and may be fully unviable by midday (Desai and Musmade 1998). The stigma is usually receptive for one day either side of flower opening, after which it turns brown and dries (Rasco and Castillo 1990). Flowering behaviour varies with cultivar and climatic conditions (Deshpande et al. 1979). Male/female flower ratio Average ratio of male to female flowers throughout the flowering period of two cultivars was 50 males for every female flower (Rasco and Castillo 1990). Long days cause the male flowers to bloom up to 2 weeks before the female flowers, while short days have the reverse effect (Huyskens et al. 1992 and references within). Nearly ninety percent of female flowers develop on the first forty nodes with the majority at nodes 21 to 30 (Rasco and Castillo 1990). Pruning the lower laterals increases the total number of flowers per plant by increasing the number of flowers on higher laterals (Rasco and Castillo 1990). Allowing the vine to grow on overhead trellis produced more fruiting nodes (Traynor 2001 pers comm.).

Maturity It is difficult to determine when to harvest bitter melon. Indications include a slight change in fruit colour and the fullness of ridges and bumps. Both can be difficult to observe. Consumers eat bitter melon when the fruit is physiologically immature or unripe and therefore when discussing bitter melon maturity for harvest, this refers to eating quality (and is really immaturity). A physiologically mature fruit is unsuitable for marketing. Commercial maturity is very hard to judge from the outward appearance of the fruit but external fruit colour of the whole fruit can be used. This needs to be complimented with seed coat colour which is a better indicator but obviously has disadvantages as a tool in decision making. The fruit colour must be green which it is whilst commercially immature and mature and seed coat colour is needed as a further indication of maturity. The seed coat must be creamy or pale green-brown with over maturity indicated by any shade of pink. Since fruit continues to mature after harvest, fruit for immediate sale in local markets should be harvested mature (still physiologically immature) and fruit for long distance transport (interstate or export) when immature. Sections 4.4 and 4.5 in the Bitter melon Quality Description Language Manual, BMQDL, (Vujovic et al. 2000) have colour photos which show these colour maturity indicators. Harvest It typically takes about 15-20 days after fruit set (90 days from planting) to reach marketable age (Reyes et al. 1994), but bitter melon can be harvested at any stage before, depending on the market. Fruit should be light green, thick and juicy (Lim 1998) and the seeds should be soft and white (Huyskens et al. 1992) creamy to pale green-brown (Vujovic et al. 2000). Harvest every 2-3 days as the fruit ripens quickly (Desai and Musmade 1998). Fruits increase in bitterness with time due to a build up of the alkaloid momordicine, and then lose the bitterness during ripening (Cantwell et al. 1996).

Harvest should occur before ripening (Traynor, 2000 pers. comm.) and the petiole (stem on which the fruit is found) should be cut cleanly with a sharp knife (Reyes et al. 1994).

Yield Yields of 20-30 t/ha are common (Reyes et al. 1994), and over 50 t/ ha have been reported (Huyskens et al. 1992, Kuang et al. 1997). Yields of 61-108 t/ha were achieved in Taiwan with plastic housing and grafting (Liao and Lin 1996).

8

Postharvest Bitter melon is a climacteric fruit that continues to ripen towards physiological maturity after harvest (David Hicks 2002, pers. comm.). Hence reducing fruit temperature is important to slow maturity. As it ripens bitter melon produces ethylene which can cause other bitter melon in close proximity eg. in a carton, to over ripen. Pick early morning and remove field heat immediately before storing at the correct temperature. Bitter melon should not be stored or transported with ethylene producing fruit such as banana, tomato, mango, papaya and guava. Storage temperature Optimal storage temperature for fruit is 7-10ºC (Gosbee and Lim, 2000). Fruit can be stored at temperatures down to 4ºC for short periods but prolonged exposure to low temperatures can cause chilling injury. Fruit stored above 10ºC continue ripening. The optimum temperature for setting refrigeration may be lower, and is different for each set of storage conditions. Transport refrigeration and cool room settings need to be calibrated. Temperature settings for transport of 5-7ºC. have been quoted by suppliers. If air circulation is low, heat will build up in the cartons causing fruit temperature to be higher than air temperature (Melinda Gosbee 2001, pers. comm.).

Lower temperatures cause chilling injury, evident as pitting, decay and discolouration and as higher ethylene production (faster ripening) when temperature is subsequently later raised to 15 C (Zong et al. 1995).

Higher temperatures cause ripening. More mature fruit (slightly soft but still green) will rapidly begin to ripen if the temperature is too high, risking the contents of the entire carton/container, but maturity has less effect on postharvest life when fruit is stored at the correct temperature (Zong et al. 1995).

The effect of temperature on respiration and ethylene production is given in Table 3.

Table 3: Respiration and ethylene production of selected Asian vegetables (Zong et al. 1992, Zong et al. 1995).

0°C 5°C 10°C 15°C Respiration (µl CO2 /g/h) Bitter melon 7.8 15.0 26.6 Hairy melon 8.7 11.4 21.1 Luffa 13.7 19.0 34.4 Yard long bean 20.0 23.0 46.0 Ethylene production (µl/g/h) Bitter melon <0.1 <0.2 <0.3 Hairy melon <0.1 <0.1 <0.2 Luffa <0.1 <0.1 <0.2

Post harvest change in pigmentation has been studied (Tan et al. 1999). Storage humidity Stored at 85 - 90 % humidity bitter melon fruit is quite resistant to water loss (Cantwell et al. 1996). Humidity can be achieved by lining boxes with plastic which reduces movement of all gases, including water vapour. However, plastic causes problems with bitter melon in Australia because more mature fruit produce ethylene, risking the entire box when ventilation is low. Paper lining is thus preferable at the moment to allow ventilation (Gosbee and Lim 2000). It is not necessary to use plastic bags.

Shelf life A life of 2-3 weeks is possible with good varieties stored under optimum conditions. Field heat, harvest maturity and mode of transportation are all critical in determining shelf life. Temperature has a greater effect on bitter melon shelf life and quality than packaging (Gosbee and Lim, 2000). Low

9

temperature reduces ripening and hence quality loss. Optimum shelf life is achieved at 5 C. However chilling injury (dark watery pits on skin) was observed after 3 weeks at 5 C.

Packaging Pack 10 kg into a 13 kg banana box with wet paper between each layer, so that the box is not quite full, and ventilate to remove ethylene (Peter McLaughlin 1998, pers. comm.).

Fruit wrapped individually in low density polyethylene film and stored at 5-7 C were still marketable after 21 days. The film delayed appearance of chilling injury as well as other symptoms of quality loss (Mohammed and Wickham 1993). However, Gosbee and Lim (2000) reported that packing 10 kg fruit in newspaper (in a carton with double newspaper thickness placed at top and bottom), compared with low density polyethylene plastic, increased ventilation and reduced ethylene triggering of fruit ripening. They also stress that cooling produce to less than 10 C before packing is essential.

Quality assessment

Bitter melon fruit should have a fresh appearance, uniform colour and be free from visual defects. They must be firm without excessive seed development. Common post harvest defects include over-maturity, seed development, softening, yellowing and ripening with internal or external colour changes, scuffing of the peel and chilling injury (Zong et al. 1992, Cantwell et al. 1996). Pictures of quality fruit are presented in the Bitter Melon Quality Description Language (Vujovic et al., 2000).

Pests and diseases Foliage pests and diseases were considered not to be a problem, due to toxic compounds present in the plant (Robinson and Decker-Walters 1996). Control of ants is important as they farm some insect larvae. Thrips, which can devastate a crop, can be controlled chemically. Fungal infections can occur during prolonged wet periods. Potential viruses are listed at Plant Viruses Online (Brunt et al. 1996 and onwards). Pests and diseases present in the trial sites are presented in the results section of this report. In the Northern Territory thrips, cucumber moth, heliothis, whitefly and nematodes are common and powdery mildew and virus (especially with hybrids) can also be a problem.

Fruit fly (Strumetia cucurbitae) is a problem in the US, combated by covering fruit with protective paper bags (Cantwell et al. 1996). Fruit fly does not appear to be a problem in Queensland. Double layer paper bags are used against melon fly in Taiwan, and are applied when fruit measure 2-3 cm in length (Fang and Chang 1987).

No pesticides are specifically registered for use on bitter melon in Australia. However, off-label use permits exist for a systemic fungicide for control of powdery mildew on bitter melon and a miticide/insecticide for control of Helicoverpa spp, Two Spotted Mite, Plague Thrips, Greenhouse whitefly and Whitefly (Bremia spp.) on melons, for all states except Victoria. As well, there is an off label use permit for a number of biological insectides for control of Helicoverpa spp. and cucumber moth (Phakellura indoa) larvae on cucurbits and melons in Northern Territory. These permits come with strict conditions for use and are issued with expiry dates. Check with your own state department for details about the use of off-label pesticides.

10

“Soft”chemical control

Neem potassium soap destroys the skin of whitefly, aphids, thrips and mites and in the Northern Territory is recommended as part of a spray program for whitefly control.

Biocontrol Use of Trichoderma sp. will suppress growth of Sclerotium rolfsii on fruit during storage and transport (Mukherjee and Raghu 1997). Domestic market Australian consumers (largely Vietnamese, but also other Asians) appear to prefer dark green shiny fruit but this may be because that is what they are sold. Development prospects were rated as high for both fresh and processed bitter melon (Vinning 1995). There is also potential for fresh cut product.

Pickled bitter melon is imported from India, Malaysia and the Philippines. As these are low cost producers, it is unlikely that Australia will be able to compete on this market (Vinning 1995).

New South Wales Prices are lowest during the summer when southern states are producing (Figure 1), and peak in May/June, prior to the peak season in northern Australia (July-November).

Figure 1: Price of bitter melon at Flemington Markets from January 1996 to June 1999 (Flemington Market Reporting Service, NSW Agriculture).

$0

$10

$20

$30

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Pric

e ($

/10

kg b

ox)

1996 1997 1998 1999

11

Northern Territory The price paid to producers in Darwin averaged $1.84/kg in 1994/5, and $2.00, $2.00 and $2.24/ kg in 1998, 1999 and 2000, respectively (Northern Territory Department of Primary Industries and Fisheries, 1996, 1999 & 2000) and production increased substantially (Table 4). Northern Territory producers have received up to $3.50/kg (Lim 1998).

Table 4: Northern Territory production of selected Asian vegetables

GrossValue Volume ($’000) (tonne) 1995 1996 1998 1999 2000 1995 1996 1998 1999 2000 Yard long beans 403 484 959 1,092 1,226 175 125 376 436 455 Bitter melon 389 981 1,028 1,028 1,370 189 337 514 514 611 Luffa 111 171 470 488 284 55 69 188 195 142 Winter melon 68 81 92 165 254 38 32 61 110 127 PrimeStats Horticulture Industry of the NT 2000, Office of Resource Development, NT DPIF (now DBIRD). Victoria Like Sydney, Melbourne prices have fallen in recent years. Bitter melon leaves are available during summer for $1.50 to $3.00/bunch retail, sourced from Victoria (Chew and Morgan 1996). Average retail prices vary seasonally between $5 to $1/kg, with highest prices between May to August and in January (Vujovic et al. 2001).

Export market Bitter melon will remain an important vegetable crop in South East Asia and may gain popularity with the release of lower bitterness cultivars (Reyes et al. 1994). Major markets include the Philippines, Malaysia, Hong Kong and Taiwan, but trade data are scanty (Vinning 1995).

Philippines Production rose to 18 000 tons in 1992 (Lim 1998). Retail prices rose markedly during the 1980's (Vinning 1995). It is one of the ten most widely grown vegetables (Rasco and Castillo 1990).

Malaysia Production rose to 19 000 tons in 1994 (Lim 1998).

Taiwan Production climbed to 35 000 tons in 1993 (Lim 1998). Taipei wholesale prices are falling (NT$28/kg in 1994) but volumes are increasing (8000 tons in 1994). Prices peak in March and April (Vinning 1995).

Thailand Production was 17 749 tons in 1994 (Lim 1998).

Sri Lanka Production was 19 266 tons in 1987 (Lim 1998).

China Chinese use it for a range of medicinal purposes (Yang and Walters 1992).

Indonesia Imported 2 t from Taiwan in 1992, then none from 1993-5 (Taiwanese official trade statistics)

India Preferences are for a small, dark green, very bitter type (Lamberts 1992).

12

2. Materials and methods 2.1 Replicated field trials Northern Territory Replicated field trials were carried out by NT DPI&F (now Department of Business, Industry and Resource Development, DBIRD) between 1998 and 2001 at the Coastal Plains Horticultural Research Farm, 60 km south east of Darwin (12 36'S and 131 18'E). Soil was a sandy red massive earth type. A randomised complete block design was used with two plant rows comprising one replicate (1999) or each plant row one replicate (2000), with varieties randomised within the row. Inter and intra row spacing varied between trials. Plants were trained to grow up a vertical string and lateral branches were removed except where three horizontal wires 30-40 cm apart were used to hold the lateral branches for fruit production. The plants were pinched off when they reached the third or top wire. Irrigation through surface laid T tape and measurement of soil moisture with tensiometers at 20 and 40 cm were employed to maintain soil moisture at 20 centibars in the root zone and to apply nutrients. Nutrients were injected into the irrigation system at rates of 25 kg N, 5 kg P and 18 kg K /ha/ week for early growth and 12 kg N, 5 kg P, 18 kg K and 5 kg Ca /ha/ week after fruit set or from about 6 weeks onwards. Lime and basal fertiliser were applied and trace elements, (zinc sulphate (30 kg/ha), manganese sulphate (10 kg/ha), and molybdenum (1 kg/ha)) and Solubor (2 kg/ha) were injected during early crop growth. Pest and disease control was performed as required. 1999 Four plants of 14 cultivars were grown in each of two replicate plots, with inter-row spacing of 3 m and intra row spacing 1 m. There was a 3 m buffer between plots in each row. Seed was soaked for 24 hours prior to planting on 23/6/1999. Fruit was harvested over five weeks (it was felt the harvest could have been longer) commencing 25/8/1999, nine weeks from sowing. Twelve fruit per plant were assessed for fruit weight, length and width, flesh thickness and Brix . Total marketable yield was not recorded. Appearance is critical for market acceptance. Six relatively experienced growers, with 5-6 years experience in growing bitter melon, were invited to closely inspect the cultivars and make comments on appearance and suitability of the cultivars for their markets. Appearance was also rated using the Bitter Melon Quality Description Language (Vujovic et al. 2000). 2000 Four plants of three cultivars (which performed well in 1999) were grown in each of four replicate plots, with inter-row spacing of 3 m and intra-row spacing of 2 m with a 4 m buffer between plots in a row.

Harvest commenced on 16/6/2000, eight weeks after transplanting and continued until week 17. Fruit were harvested as they reached "maturity". Fruit number and weight were recorded. After the final

13

harvest the number of fruit left to mature on the vine was recorded. Marketable yield was calculated for 1666.7 plants per hectare (spacing 2 m x 3 m). 2.2 Observation trials New South Wales 1998/99 The first “national project” trials were carried out by the Northern Rivers Agricultural Development Association (NORADA) at Burringar on the north coast of NSW (28.43 S, 153.46 E). The site had not recently been used for bitter melon production. Any cover crop or fallow plant material was treated with a general herbicide and ploughed in. Beds were formed and a pre-emergent herbicide used to suppress the growth of any emerging weed seedlings. The trial used three rows within a 0.5 ha commercial bitter melon crop with north/south orientation. Seeds of eight cultivars were sown in seedling trays on 20/8/1998 and transplanted on 8/10/1998 at the third true leaf stage. Transplants were planted in rows of 2.4 m width and 1.65 m between rows. The 2.1 m high overhead type trellis consisted of coppers logs anchors, posts, and cross-members; high tensile fruiting wire connected posts, and pig wire was loosely secured across the high tensile wire by Jerrard rings. Each plant was trained up nylon baling twine secured by polypipe at ground level and the high tensile fruiting wire at trellis canopy level. The apical meristem was excised at trellis canopy height and the two growing shoots were encouraged in east/west growth to the trellis edge, where direction changed either north/south toward the adjacent planting pair. Lateral shoots were pruned up to trellis canopy height only. Soil pH was adjusted from 6.0 to 6.5 with 1 t/ha of lime and 20 t/ha chicken manure was incorporated prior to planting. Six fortnightly applications of KNO3 at 50 kg/ha were imposed up to flowering when applications of Ca(NO3)2 at 50 kg ha-1 were applied fortnightly until crop senescence. Three 2 L/ha applications of Foliar Nitrophoska (N:12, P:2.1, K:7.0) containing minor and trace nutrients were applied monthly after planting. Irrigation was supplied twice weekly via 13 mm polypipe with a single, 360o micro irrigation spray jet situated between each planted pair. Harvest commenced on 20/01/1999, approximately 15 weeks after transplanting and continued through to mid-May. Percentage of seedling emergence, time period between sowing and fruit set, mean marketable fruit weight (over 2 week picking period), diameter of fruit size range, colour, and marketable yield per plant were recorded. Brix at 20 C was measured and an approximation of marketable yield per hectare was calculated. An assessment of marketability & viability was made. Queensland 2000 An observation trial on the effect of cultivar and trellis type in a hydroponic growing situation on plant yield and water requirement was conducted in the experimental grounds of the Plant Sciences Group, Central Queensland University, Rockhampton (23.37 S, 150.52 E). Eight simple non-circulating hydroponic boxes (similar to those used at AVRDC; Midmore, 1994), measuring 0.575 m x 0.373 m x 0.26 m were lined with black plastic and filled with 35 litres of Manutec hydroponic nutrient providing 215 ppm N, 37 ppm P and 218 ppm K, plus micro nutrients. Pre-germinated seed of two varieties, Long Shape and Baizan, were sown on 13/3/2000 at one plant per box (0.575 m within row spacing). Two types of trellis were constructed over a four-box row; a vertical fence and an overhead gourd trellis (See Figure 1). For the vertical fence horizontal wires were tied at approximately 0.3 m intervals from the top of the box (virtual ground level) to a height

14

of 2 m. While for the overhead trellis, vertical wires were dropped from a 2 m high overhead wire grid (0.2 m x 0.2 m) to guide each plant. Two plants of each cultivar were grown on each trellis type. The first lateral branches of all plants were trailed along the first horizontal wire of both trellis systems, to achieve early canopy growth and light interception. For the vertical fence trellis, lateral branches were either trailed or removed depending on their proximity to the horizontal wires. For the overhead trellis, all subsequent branches were removed until the growing point reached the trellis roof, then all branches were left and the plant allowed to reach full canopy. The water level in the boxes was checked regularly and 10 litres were added when the depth went below 10 cm from the top. Harvest commenced on 12/5/2000, 60 days after sowing and fruit number, weight, length and width were recorded for each plant. Water use was recorded during growth. Northern Territory 2001 A demonstration trial to quantify the effects of plastic mulching and entraining plants on either fence-like or overhead trellises commenced on 23/5/2001, but was terminated prematurely 12 weeks after planting due to fire. Yields were taken for three weeks before the fire but as they are inconclusive only observations are reported. Statistical Analysis Northern Territory For the 2000 field trial, ANOVA on the average and total fruit weight and total fruit number of three cultivars was conducted.

15

3. Results 3.1 Replicated field trials Northern Territory 1999 The first fruits of all cultivars were harvested 10 weeks after sowing. Average (of 12 fruit) length of fruit varied with cultivar from 14.5 to 32.0 cm, width from 5.0 to 7.0 cm and weight from 110 to 395 g (Table 5). Yield records were not collected. It was considered that the selection of types suited to market requirements was more important at this stage of investigation. Table 5. Bitter melon cultivars average fruit weight, size and Brix , over a five weeks harvest, Northern

Territory 1999 *

Cultivar

Length (cm)

Width (cm)

Weight (g)

Flesh Thickness

(mm)

Brix (mid fruit)

1. Baizin Green (hyb.)1 23.5 6.0 360 9 2.8 2. Kiew Yoke 59 (hyb.) 32.0 5.0 380 7 3.5 3. Kiew Yoke 68 (hyb.) 28.0 5.5 365 8 3.0 4. Bitter Spindel (hyb.) 14.5 5.0 110 7 3.3 5. Polo (hyb.) 23.5 5.0 245 8 2.8 6. Known You Green (hyb.) 26.0 6.0 318 9 2.7 7. Verdure (hyb.) 17.5 7.0 311 10 2.8 8. Moon Rise (hyb.) 28.0 5.5 332 8 2.8 9. Moon Light (hyb.) 21.5 6.0 306 8 3.1 10. ‘Long Type’ (OP)2 22.0 6.0 308 8 2.7 11. ‘One End Large’ (OP) 17.0 6.0 247 8 2.8 12. ‘Big Dooms’ (OP) 15.0 7.0 276 9 3.2 13. NSW local OP 24.0 7.0 395 10 2.9 14. Moon Beauty (hyb.) 20.5 7.0 367 9 2.9

* average of twelve fruit.

1Hybrid 2Open pollinated Generally, most of the hybrids were uniform over the 5-week harvest period. Hybrid Baizin Green was a particularly good type with good uniformity of shape and size while the three open pollinated cultivars Long Type, One End Large and Big Dooms were quite variable. The open pollinated, NSW local selection is a particularly good type and similar to local NT selections (rough bumpy type). Brix varied from 2.7 - 3.5 with all open pollinated cultivars, except Big Dooms, less than or equal to 2.9. The hybrids had values ranging from 2.7 – 3.5. Detailed field assessments on the plots showed the cultivars with most potential were the hybrids Baizin Green, Polo, Known You Green and Moon Light and the open pollinated line NSW local selection. The hybrids Baizin Green and Known You Green are very similar in appearance and performance.

16

Kiew Yoke cultivars 2 and 3 had poor plant vigour, sparse plants which produced long thin fruit and poor leaf cover. Hybrid Bitter Spindel had very vigorous plants but overall was a very poor performer with small fruit size. Fruit appearance recorded as shape, skin colour and brightness and presence and arrangement of bumps are presented in Table 6. Fruit assessment using Product Descriptor Language (Vujovic et al. 2000) is shown in Table 7. All the growers said that NSW local selection (OP) was a good type (rough irregular lumps). What they meant was that it was most similar to the local selection they are growing now. One grower considered Moon Light to be a good smooth type, very shiny and bright with a good shape. The smoother hybrids have fewer bumps and would have less damage in transport, are easier to pack and he had tried them before.

Table 6. Bitter melon cultivar fruit characteristics during a five week harvest period, Northern Territory 1999.

Cultivar

Shape

Bumps

Skin Brightness

Skin Colour

Baizin Green

(hyb.)1 Good uniform even shape

Large Smooth uniform ridges

Fair Light/Mid

Kiew Yoke 59 (hyb.)

Fairly uniform, some bent Long & thin

Uniform, medium size Smooth ridges

Fair Light/Mid

Kiew Yoke 68 (hyb.)

Fairly uniform, some bent Tapered.

Medium size Fairly smooth ridges

Poor/fair Light

Bitter Spindel (hyb.)

Uniform small even shapes Too small

Small, Irregular, rough bumps

Poor Light/Mid/Dark

Polo (hyb.) Mostly uniform, some bulges Mostly large Irregular smooth bumps

Good Mid

Known You Green (hyb.)

Good uniform even shapes Medium size, Smooth uniform ridges

Fair Mid

Verdure (hyb.)

Uniform short blocky shape Tapered

Medium size, Fairly smooth uniform bumps

Fair Mid/Dark

Moon Rise (hyb.)

Fairly uniform, some bent Long & thin

Fairly uniform, Smooth bumps. Some ridges

Fair Mid

Moon Light (hyb.)

Mostly uniform even shapes Small/medium, Irregular, smooth bumps

Very good Mid/Dark

‘Long Type’ (OP)2

Variable shape Some pointed, some blunt

Irregular. Some large ridges, some rough

Poor Light/Mid

‘One End Large’ (OP)

Very variable Short and long shapes

Irregular medium bumps, Mostly smooth

Fair/good Dark

‘Big Dooms’ (OP)

Variable Short block or top shapes

Fairly smooth irregular bumps Fair Mid/Dark

NSW local selection (OP)

Mostly uniform, Some bulges, some pointed

Small/medium irregular rough bumps

Fair Mid

Moon Beauty (hyb.)

Fairly uniform. Some block shapes

Very irregular small/large smooth bumps

Good Cream /White

1Hybrid 2 Open pollinated

17

Table 7. Bitter melon cultivars fruit appearance (rated using BMQDL, Vujovic et al 2000), Northern Territory, 1999.

Cultivar

Type Size

Length/

Diameter

Colour

Shape

Bumps

Baizin Green (hyb.)1 2 * 23/6 2 2 3 Kiew Yoke 59 (hyb.) 2 * 32/5 3 1 3 Kiew Yoke 68 (hyb.) 2 * 28/5.5 2 1 3 Bitter Spindel (hyb.) 1,3 & 4 1 14.5/5 3 &W 3 1 Polo (hyb.) 1 * 23.5/5 2 1 2 Known You Green(hyb.) 2 * 26/6 2 1 3 Verdure (hyb.) 2 * 17.5/7 4 2 2 Moon Rise (hyb.) 2 * 28/5.5 3 1 2 Moon Light (hyb.) 1 2 21/6 3 2 2 ‘Long Type’ (OP)2 2 2 22/6 2 1 3 ‘One End Large’ (OP) 2 1 17/6 4 2 2 ‘Big Dooms’ (OP) 1 * 15/7 4 2 2 NSW local selection OP) 1 * 24/7 2 2 1 Moon Beauty (hyb.) 4 2 20.5/7 W 2 1

Quality Descriptor Page (BMQDL 2000)

2.1

3.1

Cm

3.2

3.4

3.5

Ideal” Quality Descriptor

1

2

22/6

2

2

1

* These cultivars did not fit the size description category 1Hybrid 2 Open pollinated Only one cultivar, NSW local selection, had four bitter melon quality descriptors that were ideal, for type, colour, shape and bumps. Three cultivars, Long type, Moon Beauty and Moon Light had three descriptors considered ideal. 2000 Harvest of all varieties commenced at 8 weeks from planting with a peak at 11-12 weeks which was more distinct with the hybrid lines (Figure 2). This peak was probably a reflection of the initial fruit set on primary lateral branches. Both hybrids had similar cumulative yields with the local selection showing lower yield 11 weeks after planting but by week 17, total cumulative yield was not significantly different from the hybrids (Figure 3).

18

Figure 2. – Weekly harvest yields, Northern Territory, 2000.

Figure 3. Cumulative harvest yield, Northern Territory, 2000

Total fruit weight was the same for the three cultivars but total fruit number of the two hybrids was greater than the local selection (Table 8). The average fruit weight was greatest for the local selection with both hybrids, especially Moon Light, having significantly lighter fruit weight. Although the hybrid fruit size was lighter, it was still within the acceptable market size range. Fruit percentage that ripened on the plant before reaching a marketable size was greater with the NT local selection (Table 8).

Progressive Yield - Rep Average

0

4

8

12

16

20

8 9 10 11 12 13 14 15 16 17

Weeks from planting

Yiel

d (k

g)Known You Green Moon Light Local Selection

Cumulative yield - Rep average

0

20

40

60

80

8 9 10 11 12 13 14 15 16 17

Weeks from planting

Yiel

d (k

g)

Known You Green Moon Light Local Selection

19

Table 8. Bitter melon cultivars yield parameters, Northern Territory, 2000.

Cultivar Total Fruit Wt per

plot (kg)

Total Fruit Number

per plot

Ave. Fruit Wt.

(g.)

Marketable yield (t/ha)

Matured fruit on vine

(%)

Total fruit no. per plant

Known You Green

71.6 a*

218.5 b

328 b

29.8 a

12.1

61 b

Moon Light

73.9 a

256.5 b

288 a

30.8 a

10.8

71 b

NT OP1 Local Selection

66.7 a

177.5 a

376 c

27.8 b

18.4

53 a

* Values followed by different letters within columns are significantly different (P<0.05). 1 Open pollinated Observations made during replicated trials in Northern Territory Bitter melon are harvested at the ‘mature green’ development stage before the onset of ripening. This is an essential market requirement but difficult to determine at harvest. Even fruit showing no external colour change may have commenced ripening within the seed cavity. These fruit will continue to ripen in transit with ethylene production causing ripening of the whole box. Regular internal checks of harvested fruit showed this problem was evident in all varieties. Yield pattern was really important – it was discovered from this and another trial (data not presented) that bitter melon produce their first fruit at about 8 weeks from planting, regardless of the size of the vine. This is similar to crops like zucchini. It has real importance in vine management, as vigorous early growth will increase yields. There were obvious differences in fruit uniformity between the hybrids and the open pollinated line. Both hybrids produced very uniform fruit with size, shape and colour remaining constant over the harvest period. While the local selection fruit weight was fairly consistent, the fruit shape and colour were quite variable. Cultivars showed differences in vine vigour. Whereas both hybrids produced adequate vegetative growth to support and protect the fruit, the OP Local Selection was more vigorous with many fruit totally hidden by vine growth. The local selection seemed well adapted to the local environment with less signs of leaf wilting during periods of heat stress than the hybrids. Pest and disease observations In the Northern Territory thrips, cucumber moth, Heliothis spp, whitefly and nematodes are common and powdery mildew and virus (especially on hybrids) can be a problem.

20

3.2 Observation field trials New South Wales 1998/99 Seedling emergence of the hybrids Moon Beauty and Spindle, 90 and 92% respectively, was greater than other cultivars (Table 9). The cultivars Baizen and Moon Beauty had the best yield per plant and good fruit weight. They also took less time from establishment to fruit set than all other varieties except Spindle (Table 10). The varieties 1 end large and 1 end small were observed to perform poorly in terms of growth, fruit set and fruit quality. The NSW local selection OP (grown commercially in the region) was observed as the most consistent and prolific, though no data was recorded.

Table 9. Bitter melon cultivar yield and quality characteristics, NSW, 1998/99.

Cultivar

Emergence (%)

Fruit Diameter (cm.)

Average Fruit Weight * (g)

Fruit Colour

Fruit Set (days after emergence

Yield per plant (kg)

Marketable Yield per hectare (t)

Brix

NSW local selection (OP)

ND

Spindle Fairbanks

90

2.5-3.75

221

Light pale green

97

13.6

22.67

2.80

Long Type Minara*

70

4-6.5 5.0

330

Pale to medium green

111

14.7

24.51

2.80

Verdue Known-you

60

5-8.5

335

Medium green

124

29.4

49.01

3.45

Baizin Fairbanks

60

5-8.5

294

Pale to medium green

83

48.8

81.35

2.73

Moon Beauty Known-you

92

6.0-9.0 7.0

305

Pale green

97

43.5

72.51

2.73

1 end small Minara

ND

1 end large Minara

ND

* average fruit weight after two weeks picking ND-data not supplied The addition of calcium nitrate after flowering and during fruit set appears to improve shelf life by slowing onset of ripening. Growth of bitter melon appears to be night temperature dependent, with prolific growth occurring at day / night temperatures of 28-35 / 20-25 C and severe reduction in growth with night temperatures of less than 16 C.

21

Fruit set is reduced during overcast days with reduced sunshine hours and luminous intensity. The poor fruit set during these periods was attributed to flowers not opening, a phenomena characteristic of many species in the curcurbit family, and the reduced activities of pollinators. Physiologically mature fruit, if left on the vine, will influence the ripening and senescence of fruits within its vicinity, regardless of size. Waste of harvested fruit is high due to the presence of ripening fruit that influence the ripening of otherwise physiologically immature fruit. To minimise respiration and maintain air circulation to remove build up of ethylene, 10 kg of fruit is packed into a 13 kg banana box with wet paper between each layer to maintain humidity. Do not over fill the box and ventilate to remove ethylene. Pest & disease observations The major pests encountered include Heliothis sp. caterpillars and whitefly. Heliothis caterpillars were readily managed within an acceptable economic threshold. Caterpillars were observed to attack foliage, younger instars producing a ‘window pane’ effect and older caterpillars completely skeletonising leaves. Large caterpillars were also observed boring into fruits, ruining the marketability of the individual fruit. The onset of ripening around the wound entrance rapidly increased maturity rate of the whole fruit and influenced that of the surrounding fruit. The whitefly incidence was not deemed important, though it was recognised it had the potential for greater crop damage if the environmental conditions favoured its population growth or it were the vector for viral transmission. Disease incidence was restricted to powdery mildew and virus. The powdery mildew was prevalent in areas of low air circulation, poor light under the canopy or where vines had been left unmanaged over the trellis in parts of the commercial crop. Control was achieved by physical removal of diseased leaves. Similarly, the virus, which appeared to reduce growth vigour and affect leaf colour, was of minimal incidence. Queensland 2000 Growth Fruit set started approximately six weeks (42 days) after sowing in March and continued up to June. Cultivar Long Type was more vigorous in early growth resulting in a greater proportion of vegetation on the trellis and greater fruit set. Water use was greater for Long Type than Baizen, and with overhead trellis compared with fence trellis. Yield Both cultivars had harvestable fruit 60 days after sowing. Number of fruit was greater for cultivar Long Type than Baizan and when plants were grown on overhead compared with fence trellis (Table 10). Fruits of both cultivars were similar in length, with Baizen fruit wider and heavier at harvest.

Further fruit set after early June was low and corresponded with a temperature drop (minimum 2.4 C on June 3rd). Once fruit had set it took approximately 20 days to reach harvestable size, so the low

22

temperatures in early June, causing a reduction in fruit set, would have resulted in the low fruit numbers at the end of June/early July (Figure 4).

Figure 4. Relationship between fruit number with maximum and minimum temperature and days after sowing (das). Fruit numbers split by trellis type, (a), and variety, (b). In graph (a) grey bars denote overhead trellis and black bars

vertical fence. In graph (b) black bar denotes Baizan variety and grey bars Long Shape.

(a)

das

40 50 60 70 80 90 100 110 120

fruit

num

ber/t

empe

ratu

re C

0

5

10

15

20

25

30

35

(b)

das

40 50 60 70 80 90 100 110 120

friut

num

ber/t

empe

ratu

re C

0

5

10

15

20

25

30

35

Trellis effects Fruit number and weight were greater with overhead trellis which had twice as many fruit as the fence trellis (Table 10). Water use was greater with overhead trellis. Although fruit width was unaffected by trellis type, fruit length was considerably less with overhead trellis.

Table 10. Fruit characteristics and water use of bitter melon cultivars grown hydroponically on two trellis types,

Rockhampton, 2000.

Fruit Weight Length Width Water use Number (g) (mm) (mm) (l) Variety Baizan 5 421.20 201.25 72.75 101.1 (mean) Long Type 29 166.38 173.63 46.63 164.3 Trellis Overhead 22 281.03 147.67 58.07 175.8 (mean) Fence 12 221.63 218.00 52.60 108.3

23

Pest and disease observations Powdery mildew appeared at the beginning of May (six weeks after sowing). The polyhouse environment contributed to disease spread by being hotter than the outside environment and reducing light levels causing the plants to produce large leaves and, hence, a dense canopy. Disease incidence was greater on the vertical fence than the overhead trellis and did not vary with cultivar. Northern Territory 2001 Yield results are not available as a bush fire destroyed the trial four weeks after harvesting commenced. Observations made before fire: Harvest commenced eight weeks after transplanting confirming results of the 1999 and 2000 trials (8 weeks in 2000 and 10 weeks in 1999). There was a very high percentage of rejects in the first four weeks of harvest due to premature ripening, misshaped fruit or skin blemishes caused by wind rub. Rejected fruit appeared to have poor seed development but this could have been due to harvest timing. Misshapen and early ripening fruit were thought to be due to pollination issues. Plants grown using overhead trellis had less fruit with wind rub damage after four weeks harvest than those grown on vertical trellis. Plastic mulch resulted in more fruit and higher fruit weight after four weeks harvest than no plastic mulch, particularly with plants grown on vertical trellis.

24

4. Discussion Bitter melon is grown successfully in most states of Australia. Marketable yields have been shown to be around 30 t/ha in Northern Territory and in NSW between 23 – 81 t/ha depending upon the cultivar. Similarly, yields of 20-30 t/ha are reported as common (Reyes et al. 1994), and over 50 t/ha has been widely reported (Huyskens et al. 1992, Kuang et al. 1997, Liao and Lin, 1996). Yield per plant was greater with the hybrids Baizen and Moon Beauty than the open pollinated cultivar Long Type in New South Wales. In Queensland when grown hydroponically yield per plant was greater with Long Type than Baizen. In the Northern Territory the hybrids were not significantly different in weight per plot to the open pollinated local selection. This would be expected as the local selection has been selected for site-specific performance over a number of years. Whilst the yield (total fruit weight per plot) was the same with three cultivars in the Northern Territory 2000 trial, the fruit weight and number varied with cultivar. Those with lower average fruit weight, Moon Light and Known You Green had higher fruit number than the open pollinated local selection. That is, there was an inverse relationship between fruit number and average fruit weight. The two hybrids also had a greater number of fruit per plant than the OP local selection. Similarly, in Queensland the two cultivars varied in fruit number and inversely in fruit weight. Long Type had greater fruit number per plant with lower fruit weight and size and Baizen much fewer fruit of greater weight and size. This is similar to results obtained with other members of the cucurbit family such as kabocha, Cucurbita maxima and Japanese pumpkin C. moschata, and inter-specific crosses between them them (Morgan and Gallagher 2002, in press). Fruit weight of the two hybrid cultivars in Northern Territory 2000 and NSW 98/99 trials were considerably less than claimed as possible in seed catalogue descriptions. This may be because they were grown using same practices as used for the traditionally grown open-pollinated varieties. Further work on agronomic practice for hybrid varieties may improve yield and fruit size. Long Type grown hydroponically used 64% more water, and therefore more nutrients than Baizen, probably because of its greater total yield. However its large number of fruit were smaller with an average length just below the domestic market requirements of 18-22 cm and its yield characteristics and water use efficiency may have been enhanced with early lateral pruning. It had the more vigorous growth of the two cultivars suggesting a lot of lateral development. Long Type grown in Northern Territory had the desired market length at harvest but would not have had the extent of vegetative growth as plants grown hydroponically. Management of vegetative growth and optimum yield needs more refining. Yield is dependent on the number of female flowers successfully pollinated (Rasco and Castillo 1990) and growers in the Philippines monitor female flower opening to predict yield and harvest time (approximately 19 days after female flower opening) and when to terminate harvesting. As well, removal of lateral shoots in the first ten nodes enhanced flowering and fruiting of the two bitter melon cultivars investigated and can be used to increase yield (Rasco and Castillo 1990). Some work is required in Australia to define more closely the time periods between flowering and harvest and extent of lateral shoot removal to suit our conditions and cultivars. A practical indicator of main bulk harvest dates in the Philippines is 16-19 days after each peak female flowering (Rasco and Castillo 1990) which varies with cultivar. In Rockhampton it was reported that fruit set at 6 weeks and harvest started at 60 days after sowing which is approximately 18 days later, supporting the harvest indicator as a potentially useful tool for growers. Knowing the number of flowers set weekly can also be used to predict percentage of expected total yield at any

25

point. This could then be used in management decisions such as labour requirement and forward marketing. Data on flowering time were not available from these trials but this is worth following up by growers. First fruit matured earlier at Darwin in 2000 and 2001, 56 days after sowing pre germinated seed compared with 1999 when first fruit were harvested 70 days after transplanting. The later time to harvest in 1999 was probably due to inexperience in deciding on maturity in the first year of trials. In Rockhampton fruit grown hydroponically first matured for commercial harvest 60 days after sowing. Knowing that fruit matures approximately 56-60 days after sowing will assist growers when assessing maturity for harvest. Maturity or readiness for harvest is a very difficult decision to make for bitter melon, with no simple outward signs of maturity. Indications include a slight change in fruit colour and the fullness of ridges and bumps, both of which are difficult to observe. A percentage of fruit mature on the vine before reaching a marketable size. This can be caused by poor pollination or early injury to fruit, but is also noted as a varietal characteristic. In Darwin 2000 the local selection had more fruit maturing on the vine than the two hybrid cultivars (Table 4). This is a common problem for growers who collect their own seed, and open-pollinated seed quality can vary between farms. When fruit size is small, the fruit will ripen on the plant before it has reached a marketable size. Hence, the growers place a great deal of importance on selecting for fruit size (as well as other important market requirements) to help overcome this problem. In Darwin 2000, harvest peaked 11-12 weeks after planting especially with the hybrid cultivars. This peak was probably a reflection of the initial fruit set on primary lateral branches. Further smaller peaks in harvest would come from the flowering and fruit set on secondary lateral branches. This trend in flowering and fruiting was also evident in yield figures from an unpublished observation trial in 2000. Fruit set was less on secondary and later laterals so it is important to get early vigorous growth and highest yield as possible at first peak. This would also reduce disease problems and help define harvest termination date. The trellis type effect at Rockhampton was confirmation of prevalent knowledge. Vertical trellis makes fruit sighting and picking harder, can cause blemishes in fruit colour and/or lighter fruit colour due to heavy shading effects and in Queensland created a more humid environment suitable for disease spread and reduction in pollination, with resultant yield reduction (Table 9). The overhead trellis created a more open growing environment with more aeration, less disease and provided they are a height suitable for comfortable harvest are recommended. It has been recommended that grape trellises could also provide the agronomic advantage of overhead trellis and be economical to install (Osborne et al. 2000). Initial comments by growers during inspection of the Darwin 2000 trial indicated that they preferred the local selection with its rough irregular bumps because their agents wanted it. However, recently some agents have indicated that they also want the smooth skinned Moon Light and growers are now planting it more commonly. Cost of hybrid seed was calculated at approximately 4 cents per seed and growers consider this to be worth it. These trials have provided a basis for growers to make more informed management decisions when growing bitter melon. However there is a need for more agronomic work on nutrition and water management particularly of hybrid cultivars and further understanding of the relationship between flowering and peak harvest would allow for scheduling for more consistent market supply.

26

Recommended cultivars for Darwin are Moon Light, Baizen, Long Type and the white fruit cultivar Moon Beauty. As yet market acceptance of Moon Beauty is uncertain. For northern NSW Baizen and Moon Beauty had greatest yields. It is too soon for recommendations to be made for central Queensland.

27

5. References Abusaleha and O. P. Dutta (1994b). Performance of bittergourd (Momordica charantia) under

different training systems. Indian Journal of Agricultural Sciences 64(7): 479-481.

Cantwell, M., X. Nie, R. J. Zong and M. Yamaguchi (1996). Asian vegetables: Selected fruit and leafy types. Progress in new crops. Ed.: Janick, J. Arlington, VA, ASHS Press: 488-495.

Chew, M. and Morgan, W. (1996). Melbourne retail Asian vegetable survey. Melbourne, Agriculture Victoria 143 pp.

Desai, U. T. and Musmade, A. M. (1998). Pumpkins, squashes and gourds. In: Handbook of vegetable science and technology: production, composition, storage and processing. (Ed; Salunkhe, D. K. and Kadam, S. S.). New York, Marcel Dekker 273-298.

Deshpande, A. A., Venkatasubbaiah K., Bankapur V. M. and Nalawadi U. G. (1979). Studies on floral biology of bitter gourd (Momordica charantia L.). Mysore Journal of Agricultural Sciences 13(2): 156-159.

Devi, J. R. and J. A. Selvaraj (1994). Effect of pre-sowing treatment on germination and vigour in bitter gourd (Momordica charantia L.) cv. Co.1. Seed Research (22: 64-65.

Gosbee MJ and Lim TK (2000). Postharvest Handling of 'Asian' vegetables in the Northern Territory. ACIAR Proceedings 100 “Quality Assuance in Agricultural Produce. Proceedings of 19th ASEAN/1st APEC Seminar on Postharvest Technology. Ho Chi Minh, Vietnam, 9-12 November 1999.pp 456-460.

Huyskens, S., Mendlinger, S., Benzioni, A. and Ventura, M. (1992). Optimization of agrotechniques for cultivating Momordica charantia (karela). Journal of Horticultural Science 67(2): 259-264.

Lamberts, M. (1992). Production trends for specialty Asian vegetables in Dade County, Florida. Acta Horticulturae 318: 79-85.

Larkcom, J. (1991). Oriental vegetables: the complete guide for garden and kitchen. London, John Murray 232 pp.

Lee, B. (1995). Audit of the Australian Asian vegetables industry. RIRDC Research Paper No. 95/13. Canberra, Rural Industries Research and Development Corporation. 97 pp.

Liao, C. T. and C. H. Lin (1994). Effect of flooding stress on photosynthetic activities of Momordica charantia. Plant Physiology & Biochemistry 32(4): 479-485.

Liao, C. T. and C. H. Lin (1996). Photosynthetic responses of grafted bitter melon seedlings to flood stress. Environmental & Experimental Botany 36(2): 167-172.

Lim, T. K. (1998). Loofahs, gourds, melons and snake beans. The New Rural Industries. Ed.: K. W. Hyde. Canberra, Rural Industries Research and Development Corporation: 212-218.

Lin, Y. S., Hwang, C. H. and Soong, S. C. (1998). Resistance of bitter gourd-loofah grafts to Fusarium oxysporum f.sp. momordicae and their yield. [Chinese]. Chih Wu Pao Hu Hsueh Hui Hui K'An 40(2): 121-132.

Midmore, D.J., (1994) Simple hydroponics for food security. ILEIA Newsletter, 10:11 – 12.

Mohammed, M. and L. D. Wickham (1993). Extension of bitter gourd (Momordica charantia L.) storage life through the use of reduced temperature and polyethylene wraps. Journal of Food Quality 16(5): 371-382.

Morgan, W.C. and Gallagher, D. (2002) Kabocha and Japanese pumpkin in Australia. RIRDC Publication (in press).

Mukherjee, P. K. and K. Raghu (1997). Trichoderma sp. as a microbial suppressive agent of Sclerotium rolfsii on vegetables. World Journal of Microbiology & Biotechnology 13(5): 497-499.

28

Osborne, T., Gallacher, D. and Midmore, D.J. (2000) Hydroponic bitter melon variety trial in Central Queensland. Access to Asian Vegetables Issue 33.

Rasco, A.O. and Castillo, P. S. (1990). Flowering patterns and vine pruning effects in bittergourd (Momordica charantia L.) varieties 'Sta. Rita' and 'Makiling'. Philippine Agriculturist 73: 3-4.

Reyes, M. E. C., Gildemacher, B. H. and Jansen, G. J. (1994). Momordica L. In: Plant Resources of South-East Asia: Vegetables. (Ed.: Siemonsma, J. S. and K. Piluek). Wageningen, The Netherlands, Pudoc Scientific Publishers 206-210.

Robinson, R. W. and Decker-Walters, D. S. (1996). Cucurbits. New York, CAB International 226 pp.

Rubatzky, V. E. and M. Yamaguchi (1997). World vegetables; principles, production, and nutritive values. New York, Chapman and Hall 843 pp.

Singh, D. K. (1991). Effect of temperature on seed germinability of Momordica charantia L. cultivars. New Agriculturist 2(1): 23-26.

Tan, X. P., Ueda, Y., Imahori, Y. and Chachin, K. (1999). Changes in pigments in pulp and aril of Balsam pear (Momordica charantia L.) fruit during development and storage [Japanese]. Journal of the Japanese Society for Horticultural Science 68(3): 683-688.

Traynor, M., Owens, O., and Gosbee, M. (in press) Growing bitter melons and other Asian vegetables

Vinning, G. (1995). Market Compendium of Asian Vegetables. RIRDC Research Paper No. 95/12. Canberra, Rural Industries Research and Development Corporation 386 pp.

Vujovic,S., Chew,M. and Morgan,W. (2001). Melbourne retail and wholesale Asian vegetable survey January 1997- June 2001. Melbourne, Department of Natural Resources and Environment,62pp.

Vujovic, S., Gosbee, M., Marte, S., Thomson, G., Chew, M. and Morgan, W. (2000). Bitter Melon Quality Description Language. Melbourne, Department of Natural Resources and Environment. 25pp.

Waters, C. T., Morgan, W. C. and McGeary, D. J. (1992). How to identify, grow and use oriental vegetables. Melbourne, Agmedia 128 pp.

Yang, S. L. and T. W. Walters (1992). Ethnobotany and the economic role of the Cucurbitaceae of China. Economic Botany 46: 349-367.

Zong, R. J., Cantwell, M., Morris, L. and Rubatzky, V. (1992). Postharvest studies on four fruit-type Chinese vegetables. Acta Horticulturae 318: 345-354.

Zong, R. J., Morris, L. and M. Cantwell (1995). Postharvest physiology and quality of bitter melon (Momordica charantia L). Postharvest Biology & Technology 6(1-2): 65-72.

29

6. Appendix 1 Seed companies that supply bitter melon Australia

Fairbank’s Selected Seed Co. pty. Ltd.

542 Footscray Road Footscray Vic. 3011 Phone: 03 9869 4500 Fax 03 9687 7089

Mangus Kahl Seeds PO Box 125 Metung Vic 3904 Phone 03 5156 2370 Fax 03 5156 2522

Minara seeds Pty. Ltd. 25 Centurian Street Bridgeman Downs Qld.

4090 Phone 07 3263 4090 Fax 07 3263 3182

International East West Seed Company 50/1 Moo 2 Bang Buoi Thong- sai noi Road Amphur Sai noi Nonthaburi 1150 Thailand Phone 662 597 1225 Fax 662 597 1229 Email Info @eastwestseed.com

Known-You Seed Co., Ltd. 26 Chung Chen 2nd Road Kaohsiung Taiwan 802 R.O.C. Phone 886 224 1106 Fax 886 222 7299