HAND POLLINATION OF CUSTARD APPLES
A DPI Technical Feature

SCIENTIFIC NAME: Annona reticulata
FAMILY: Annonaceae

The custard apple flower is hermaphroditic (male and female parts in the same flower) and exhibits protogynous dichogamy (stigmas = female parts are receptive before the pollen is shed by the anthers).

The separation of the female and the male stages appears to be the the main factor limiting the level of self pollination. Fruit set can occur as a result of self pollination, insect-assisted pollination or hand pollination. The level of wind pollination is low. Studies in Australia and in Israel have shown that the main insect pollen-carrying vectors are nitidulid beetles. These beetles are capable of transferring pollen from the stamens to the stigmas in the same flower or to other flowers at the time of anther dehiscence. The population of these beetles in orchards at flowering time may vary and consequently fruit set may be low.

When do custard apples flower?
Custard apples flower between October to February inclusive. Flowering is associated with the vegetative flushing pattern. Flowering intensity is dependent on tree vigour, tree age, lateral size and variety. Some varieties such as Pink's Mammoth, exhibit strong apical dominance, producing few flowers until 5-7 years of age. Other varieties, such as African Pride, produce an abundance of flowers within 2-3 years after planting.

Hand pollination
The advantages of hand pollination are:

1. Increased fruit set of poorer yielding varieties;

2. Manipulation of the harvest period;

3. Better quality (size and shape) fruit.

Currently hand pollination is mainly used to increase fruit set of poorer yielding varieties. Considerable opportunity also exists to manipulate flowering and fruit set periods to produce high priced early or late set fruit. Fruit quality is also improved as a result of hand pollination, making fruit more suitable for export markets. A premium price is often paid for hand pollinated fruit.

Pollen collection and storage
Flowers from which pollen is to be collected should be picked from trees during mid to late afternoon. The petals should be nearly fully opened and the pollen sacs should have turned creamy-grey in colour and less tightly held together. Immature pollen sacs are white and tightly packed. On hot, low humidity days, the flowers will release their pollen sacs readily. These can be collected by shaking the flowers over a piece of paper. The pollen sacs should be placed in a small container and stored at room temperature ( <20°C) overnight for use the following morning. Alternatively if the pollen sacs do not readily separate from the flowers, the flowers should be spread thinly over shallow layered trays and left overnight. Pollen sacs should be then separated on the following morning. Flowers should not be placed in a closed container as a build up of ethylene and moisture will cause the pollen sacs to turn brown and pollen germination will be severely reduced. Preliminary studies suggest that custard apple pollen can only be stored for very short periods of time (2-3 days).

Frequency of pollination
Trees can be hand pollinated every week from December to February. Records on the number of flowers pollinated on each tree and set should be kept so as to prevent overcropping and consequently reduced fruit size.

In South-East Queensland, the fruit development period for African Pride is approximately 26 weeks, while the period for Hillary White is 19 weeks. It is possible to manipulate components of the fruit development cycle by such things as defoliation, hand pollination and summer tip pruning and consequently the time of fruit maturity.

Environmental factors affecting fruit set
The success of hand pollination depends on environmental conditions during fruit set. Conditions conducive to high fruit set are moderate temperatures, high relative humidity, lack of tree water stress and no wind or continuously heavy rain. If relative humidity falls below 70 percent, fruit set and fruit shape are adversely affected. High relative humidities in the orchard can be maintained by establishing windbreaks. Light applications of irrigation in the late afternoon using wide-throw minisprinklers also will help opening flowers maintain their stigma receptivity until the following morning.

The most favourable period to hand pollinate flowers is in the early morning prior to 11 am. Fruit set is reduced if flowers are pollinated during the hottest period of the day. Hand pollination should not be attempted on days when light rain is falling as fruit set on these days is reduced.

Tree physiological factors affecting fruit set
Fruit set is reduced during periods of strong vegetative flushing, due to competition between the growing points and the flowers. Larger flowers set more heavily than smaller ones. Flowers produced on the basal nodes (node positions 1-5) set more heavily than those closer to the apices. Trees should be maintained in a state of good health. Foliar applications of boron during the flowering period will also improve fruit set and fruit quality.

Economics
Flowers can be hand pollinated using the brush technique at the rate of 150-200 flowers per hour with success rates varying between 80-100 per cent. Using pollination guns this rate can be increased to 400 flowers per hour. Using the brush technique and assuming a minimum success rate of 80 per cent, 160-250 fruit can be set for every hour of hand pollinating. This is enough to fill about ten cartons. Using 1995 prices of $12 per carton, a gross return of $120 could be obtained for every hour of hand pollinating. Even greater returns could be achieved if a more effective hand pollination gun could be developed.

Pollen sources for hand pollination in custard apples

Hand pollination is required to produce a commercial crop of Pinks Mammoth and Hillary White custard apples. Pollen from these varieties is not always available at the right time in large enough quantities to set a viable commercial crop. Other varieties, such as African Pride (which produces large quantities of flowers), can be a good source of pollen in these times of pollen shortfall.

Trialling alternative pollen sources
Trials were conducted to assess the use of alternative pollen sources. Flowers from African Pride, Pink's Mammoth and Hillary White varieties were harvested mid-afternoon and left overnight. The following morning, the pollen sacs were extracted from the flowers of each of the varieties. Flowers from both Pink's Mammoth and Hillary White were hand pollinated with pollen from Pink's Mammoth, Hillary White and African Pride. African Pride was only pollinated with its own collected pollen, to confirm its pollen viability. African Pride was not pollinated with the other two pollen sources as it usually does not require hand pollination.

All pollinated flowers were tagged and the pollen source recorded. Fruit, resulting from the pollination, was assessed at harvest.

African Pride pollen is best

African Pride shows a high level of self-compatibility when using its own pollen for hand pollination.

In contrast, varieties such as Pink's Mammoth show a high level of self-incompatibility which results in reduced fruit set and poorer fruit shape.

African Pride is therefore the preferred pollen source for most commercial varieties of custard apples.

Methods of pollen application in custard apples

A variety of methods are available for applying pollen to flowers when hand pollinating custard apples. These methods include pollination guns, puffers and the use of a camel hair brush. All of these techniques are effective when pollen is not strongly diluted with a carrier such as lycopodium (a tan powder made from the spores of clubmosses. It is highly flammable) or PVC dust.

Photo of a pollen gun, a puffer and a camel hair brush.

Which application technique is best?
Flowers from African Pride were harvested mid-afternoon and left overnight. The following morning. pollen sacs were extracted and diluted with various amounts of PVC dust. The pollen mixtures were used to pollinate Hillary White flowers. Two methods of application were used - a camel hair brush and a pollination gun (based on a prototype of the persimmon pollination gun). All flowers were tagged and pollen dilution and application methods were recorded. Fruit was assessed at harvest.

Discussion
The brush application technique is a more effective method of increasing fruit set and symmetry. The quality of the fruit from this method is only affected when the dilution of pollen falls below 30 per cent.

The brush technique produces more reliable results due to a greater amount of pollen applied to the stigmas.

Pollination guns tested need to be modified so as to apply a more accurate and uniform dose of pollen to the flower.

The pollen storage and delivery system of these guns need to be pressurised so that the gun can be used at any angle.

Flower size affects fruit size in custard apples

Choosing the right flowers to hand pollinate can improve custard apple fruit quality and size. Hand pollination can be time consuming so it is important to ensure that the right flowers are pollinated to give the best results.

Trialing the effect of flower size
A trial was conducted to see if flowers size affected fruit quality and size.

African Pride flowers were harvested mid-afternoon and left overnight. The following morning, pollen sacs were extracted and used to pollinate Pink's Mammoth flowers of three specific size grades.

These flowers were tagged and the subsequent fruit assessed at harvest.

Larger flowers produce better fruit
It is evident that the size of the flower affects the size of the fruit produced and the level of fruit set.

Flowers that were 40-50mm long had a 98 per cent level of fruit set with excellent fruit symmetry.

This is in contrast to flowers that were 20-30mm in length which produced a low level of fruit set (48 percent) with reduced, but acceptable, fruit symmetry.

This finding is similar to findings in other temperate fruit tree crops which have shown that tree size control and canopy and nitrogen management can all influence flower size.

It is recommended that leaf nitrogen levels be maintained in the optimum range and for greater productivity, larger flowers should be targeted in hand pollinating operations.

Provided these problems could be overcome, the pollination gun would be a much faster method of application (nearly twice as fast).

In California, the Japanese persimmon pollination gun has proven to be highly effective in pollinating cherimoya, unfortunately these guns are no longer manufactured.

At low dilution rates, the pollination gun produced comparable fruit set and symmetry results to that of the brush application technique.

The slight reduction in fruit set results would be compensated by the increased pollination rate that is achievable (x5 the brush application technique) with this device.

This would make the use of a hand pollination gun a much more profitable practice.

Pollen dilution for hand pollination in custard apples
Custard apple growers use hand pollination to increase and control fruit set.

Collecting sufficient pollen to pollinate an economical number of flowers is time consuming.

Pollen can be extended by using dilutants such as PVC dust, lycopodium and talc.

Pollen sources and dilution rates
Trials were conducted to investigate the use of dilutants to extend pollen. African Pride flowers were harvested mid-afternoon and left overnight. The following morning, the pollen sacs were extracted and some was diluted with either PVC dust or lycopodium. The dilution rates were 50 per cent (Pollen): 50 per cent (Dilutant), 33.3 per cent (Pollen): 66.6 per cent (Dilutant) and 100 per cent (Pollen). The prepared dilutions were then used to pollinate Pink's Mammoth flowers. These flowers were tagged and the subsequent fruit assessed at harvest.

What's recommended
At low levels of dilution (less than one third - 33 per cent), there appears to be less than a 15 per cent drop in the efficacy of lycopodium or PVC dust when compared to undiluted pollen. At higher rates of dilution, PVC dust performed better as a carrier than lycopodium. There was approximately a 30 per cent reduction in the level of fruit set using PVC dust, whereas, an 85 per cent reduction occurred using lycopodium.

Fruit symmetry appeared to be affected minimally by the use of PVC dust, whilst the use of lycopodium produced a drastic reduction in fruit symmetry.

The results achieved through the use of dilutants are still lower than those achieved from undiluted pollen.

The benefit of using PVC dust as a dilutant is the increased number of flowers that can be pollinated per gram of pollen harvested.

It appears that the pollen grains adhere much more easily to the PVC dust than they do to lycopodium.

PVC dust is also less expensive than lycopodium.

A.P. George, M.J. Thomas, J.A. Campbell and R.J. Nissen
Queensland Fruit & Vegetable News, Nov. 16, 1995

DATE: March 1996

* * * * * * * * * * * * * * *