3. PLANT DEVELOPMENT AND WEATHER AND SOIL RELATIONSHIPS

Overview

Lychee trees go through several phases of plant development during the year. These include leaf expansion, flowering, anthesis and fruit development. There is generally considerable variation in the timing of these different growth stages amongst trees within an individual orchard.

High-yielding trees typically have one or two flushes after harvest, followed by another one in winter. The first flushes are usually vegetative, and the one in winter, floral. This is provided cool weather occurs during early bud development. Inflorescence development continues uninterrupted and leads to anthesis six to twelve weeks after panicle emergence. Fruit set in spring normally lasts two to six weeks for an individual cultivar in an orchard. Fruit mature after 12 to 16 weeks, depending on temperatures during fruit development.

Commercial activity in Asia and the Pacific is mainly found in sub-tropical environments from 17 to 30 degrees latitude. There are also some plantings at elevation in the tropics. Most of the sub-tropical areas have cool or cold winters and warm to hot summers, while rainfall is highest in summer and least in winter and spring. Temperatures below 20°C induce flowering, whereas drought is not essential. Extremes of temperature influence productivity by affecting male and female flowers, pollination and fruit set. There can also be problems if trees are droughted during fruit development. A high proportion of the fruit can brown, split or abscise before harvest in some locations. Average yields are low compared with many tropical fruit such as avocado and mango, usually less than 1 to 5 tonnes per ha, although yields of 10 tonnes per ha or more have been recorded in some areas, with close spacings and irrigation.

A model showing the relationship between potential flowering with latitude along eastern Australia can be used to estimate the reliability of cropping at different elevations in more tropical areas.

Orchards can be established on many different soils, provided they are well drained to at least a metre. Clay loams of medium to high fertility are preferred. Light sandy soils may dry out during hot weather, while there can be problems with micronutrients at extremes of soil pH. Many soils in the Region are acid or have been acidified and applications of lime or dolomite are required. In contrast, many of the soils in India are calcareous with a pH above 7.0 and trees are also susceptible to moderate levels of salinity. Mature trees may have many roots below a metre and are thus able to extract soil water to a considerable depth. Mycorrhizal fungi have been isolated from roots, but whether they are required for commercial lychee production is not known.

3.1 Plant development

Lychee trees go through several phases of plant development during the year. These include leaf expansion, flowering, anthesis and fruit development. There is generally considerable variation in the timing of these different growth stages amongst trees within an individual orchard. There can also be differences between different branches on an individual tree.

High-yielding trees typically have one or two leaf flushes after harvest, followed by a floral one in winter, if cool weather occurs during early bud development. In southern Queensland, the main period of leaf growth occurs from January to March for early cultivars such as “Tai So” and from March to May for late cultivars such as “Wai Chee”. This would be equivalent to July to September, and September to November in Guangdong. Inflorescences develop without a dormant period, with the flowers opening after six to twelve weeks. In southern Queensland, panicles normally emerge in April to May in early cultivars and from June to July in late cultivars. These periods are equivalent to October to January in southern China. Inflorescences can have several leaves especially when buds develop during warm weather. There may also be vegetative outgrowth of lateral buds below the inflorescences.

Anthesis in spring normally lasts two to four weeks for an individual cultivar in an orchard, with fruit mature 12 to 16 weeks later. The duration of each stage varies from orchard to orchard and year to year. Anthesis occurs from August to October in southern Queensland, and from February to April in Guangdong. Maximum growth occurs during the last six weeks of fruit development. Fruit are harvested from December (“Tai So”) to February (“Wai Chee”) in southern Queensland, and from May to August in much of Asia.

3.1.1 Leaf development

Shoot elongation is by repeated flushes during which several leaves and internodes expand. At the end of leaf expansion, the leaves thicken and change from light to dark green. The minimum interval between successive vegetative flushes (or between vegetative and floral shoots) is approximately six weeks. The interval can be much longer, depending on the weather and the physiological state of the plant. Low temperatures, low light, drought and nutrient deficiencies increase the interval between successive flushes. Pruning can be used to alter the pattern of shoot elongation, and if carried out at the correct time can induce flowering in some locations (see Figure 4).

Figure 4. Flush development in lychee cultivar Kwai May Pink in northern New South Wales.

(Shoot elongation shown for two trees pruned initially in September (broken line) and October (solid line). Both trees flowered the following spring. Pruning for tree size control is normally carried out after harvest from January to March.)

3.1.2 Panicle development

The inflorescence is composed of several multiple-branched panicles initiated on the present season's wood. The panicles are normally produced terminally in clusters of ten or more, although in some trees, a high proportion of axillaries may be produced. Inflorescences are generally mixed, with the lowest buds producing leaves only, the middle buds producing floral buds in the axils of the leaves and the topmost buds producing only floral branches and sometimes very small leaves which do not persist. This pattern of development is related to differences in temperature experienced by different buds during early shoot development.

Growth of the inflorescences is usually complete in six to twelve weeks, with considerable variation in the growth of individual branches within a tree. It is possible to determine shoot development by studying the external appearance of the buds as they emerge. Shoots with the terminal and axillary buds dormant tend to remain vegetative. Shoots with the terminal bud dormant, but axillary buds stalked tend to form small panicles, while those with both buds active give rise to regular panicles.

Differences exist between cultivars with respect to the type of panicle initiated. Early cultivars such as “Tai So” in Australia produce large multi-branched panicles with large numbers of mostly male flowers, while late cultivars such as “Wai Chee” produce small panicles with few branches and mostly female flowers. Flower development after initiation is normally earlier in warm weather and is delayed or prevented by frosts. When the terminal buds are frosted, a large number of axillaries may be produced. Some growth regulators can elicit a similar response.

The proportion of female flowers varies with orchard, year and cultivar. Only the female flowers can set fruit. Cultivars with a high number of female flowers have the potential to be high yielding. Inflorescences that develop early in autumn during warm weather in southern Queensland often have predominantly male flowers. This also affects lychee in Asia. Cultivars vary in the number of male and female flowers: “Wai Chee” produces more female flowers than “Kwai May Pink” in southern Queensland.

3.1.3 Flowering

Flowers normally open for 20 to 45 days within an individual orchard and cultivar, depending on seasonal conditions. Flowering is more compact when it occurs late in spring in warm weather. There is no pollination unless the male and female stages overlap. This can be a problem in some seasons when the weather warms up very quickly. These conditions increase the rate of development of the male flowers. Consequently, the male flowers have finished when the female flowers open. Generally, the last stage of male flowering provides most of the pollen for the female flowers.

Flower opening occurs during both the day and night, with peak opening in the early morning, provided temperatures are above 15°C. Flower opening in Queensland normally corresponds with the dry season. Under very dry or warm conditions, the young flowers may wither and fail to develop. In contrast, it is often wet during fruit set in southern China. Male flowers shed pollen for three days after opening, however, not all the anthers shed pollen at the same time. Pollen is short-lived, lasting no more than a day after shedding. Individual female flowers also have a limited life. If the female flower is not pollinated within three days of opening, it will not set. Once again, extended hot or dry weather can dry out the stigmas.

The flowers posses nectaries and attract many insects, including native and European honey-bees. Some authors have shown that bagging inflorescences or screening trees more or less completely prevented pollination. However, others have achieved satisfactory fruit set without insects. This area requires further research. Cool overcast weather, strong winds and some insecticides reduce foraging by the bees. Hives are installed during flowering in some orchards within the Region. Usually two or three hives per hectare of orchard are sufficient. Application of carbaryl and some other insecticides should also be avoided at this time.

Far more female flowers are produced than develop into fruit. This can be due to premature flower shedding, excessive numbers of male flowers or occasionally poor pollination. In some seasons, insects such as flower-eating caterpillars, thrips, flower-eating beetles and erinose mites can damage the flowers and reduce yields. From 1 to 10 percent of the female flowers carry a fruit to harvest, with some cultivars more productive than others. This is a relatively high rate of set compared with other tropicals such as avocado and mango, which may set less than 0.1 percent of the female flowers.

3.1.4 Fruit growth

Only one of the two ovaries of the female flower normally develops into a fruit. Very rarely, two lobes develop, with the mature fruit superficially resembling two fruit adhering to each other at their bases, each containing a seed. Depending on the season and cultivar, fruit take about 12 to 16 weeks to mature. Fruit growth is normally faster when it occurs late in the season during warmer weather.

Not all parts of the fruit develop at the same time. During the first seven to eight weeks after fertilization, the fruit skin, the embryo and the seed skin are formed. At the end of this stage, the aril or flesh is only a negligible portion of the fruit. During the next two to three weeks, the cotyledons (or seed leaves) that comprise most of the seed are formed, and the development of the aril begins. At the end of this stage, the aril is about a third of fruit fresh weight. The final period of fruit growth is dominated by rapid growth of the aril (seed development also continues). At fruit maturity, the aril is about 65 to 75 percent of fruit weight (Figure 5).

Maximum fruit weight occurs about two to three weeks before the fruit mature. In most cultivars, the colour changes from green to yellow-red to red with advancing maturity. This change is associated with a flattening of the skin segments and protuberances, and an increase in sugar/acid ratio and eating quality.

Fruit weight is related to weather and tree culture, and ranges from 15 to 35 g for different cultivars. Cultivars that have a high proportion of chicken tongue seed normally produce smaller fruit. Some of these cultivars may also produce nearly seedless fruit. These normally weigh only 8 or 10 g.

High leaf nitrogen and potassium concentrations and regular irrigation are essential for good fruit yields. Temperature can also affect the plants. High temperatures often accelerate fruit development at the expense of fruit weight. However, at very low temperatures, photosynthesis is reduced. The largest fruit are generally produced at intermediate temperatures.

3.1.5 Fruit abscission

Far more fruit are set than harvested. Typically, premature fruit abscission commences soon after anthesis and continues to fruit maturity, with most fruit abscising in the first two to six weeks (Figure 5). This varies greatly with locality, year, cultivar, weather and culture, and in some cases all of the fruit are shed. The initial abscission is thought to be due to failure of fertilization. Fruit can also fall after embryo abortion.

Later abscission is thought to be due to competition for assimilates. Girdling at this stage often reduces fruit drop, while drought, shade and leaf removal increase it. Fruit thinning at this time also increases the retention of the remaining crop. Surprisingly, the major fruit drop period occurs before the peak demand in carbohydrates by the developing crop. The young green fruit can photosynthesise, however, most of the carbohydrates for the fruit come from current assimilation in the leaves behind the fruit clusters. Reserves in the branches can also be used. Young leaves do not induce fruit abscission unless they develop directly behind the fruit cluster. This generally only occurs when fruit set is poor.

Nutritional and hormonal imbalances have been implicated in premature fruit abscission. Experiments by Israeli scientists have shown that fruit retention can be improved by applying auxins when the fruit weigh about 1 to 2 g. Earlier or later applications are ineffective. Some of these growth regulators can also increase fruit size.

Fruit-sucking bugs and fruit borers induce fruit abscission in many orchards within the Region. In some areas, they can account for more than 90 percent of green fruit drop.

3.1.6 Fruit disorders

Sunburning and skin-cracking (splitting) occur throughout Asia and the Pacific, and are often associated with hot, dry weather, drought and low calcium concentrations. However, the relationship between these disorders and tree management is not clear.

Skin cracking often occurs when trees are droughted soon after fruit set. If the drought is severe enough, fruit development will be affected, particularly the development of the fruit skin. Cell division is reduced and the fruit skin becomes inelastic, and often splits when the aril grows rapidly before harvest. This can occur after irrigation or heavy rain, or just an increase in relative humidity.

Insects, hail, and the sun can damage the skin during cell expansion and induce cracking towards harvest. These damaged areas cannot expand with the rest of the fruit, creating a weakness in the skin that splits.

Figure 5. Pattern of fruit drop and fruit growth for lychee cultivar Tai So. (Trees droughted from panicle emergence until after harvest)

3.2 Relationship between plant development and weather

Although lychee has a long history in Asia, there have been few critical studies on the response of the plant to weather. Many of the earlier studies were conducted under field conditions where sunshine, temperature and water supply are often correlated. It was not until the late 1960s, that the first glasshouse experiments on flower initiation were initiated. Later studies examined the relationship between flower development, pollination, fruit growth and environment.

3.2.1 Weather in different areas

The main commercial plantings in Asia and Australia are found at low elevation in the sub-tropics from 17 to 30 degrees latitude. A few small industries are also based at 300 to 600 m in tropical locations in the Central Plains of Thailand near Bangkok and in a few selected areas of the Philippines and Indonesia. Most of the sub-tropical areas have cool to cold winters and warm to hot summers (Table 3). Rainfall is highest in summer and least in winter or spring. Lychee is found in a narrow range of climates, whereas many other tropical fruit such as citrus, mango and banana are cultivated from the cool sub-tropics to the warm equatorial tropics.

Most of the commercial areas have winter minima below 20°C and usually below 15°C (Table 3). Winters are dry, with rainfall of less than 50 mm. Maxima during fruit set are usually between 20° to 30°C. Rainfall is usually light, with less than 50 mm in spring, although some areas such as Fuzhou have more than 100 mm. Summers are warm to hot, with maxima of 28° to 33°C. Average summer rainfall is at least 150 mm and usually more. In near equatorial areas such as Ho Chi Minh City (latitude 11°N; elevation 9 m), minima do not fall below 20°C during the year and yields are very unreliable, even though there is a distinct dry season.

3.2.2 Effect of solar radiation on plant development

Lychee originated as one of the dominant species in sub-tropical rainforests of Asia. However, as with many crops, the original environment may not be ideal for commercial production. Both flowering and fruiting are reduced once adjacent trees start to crowd each other in an orchard, and thinning becomes necessary. The decline in yield in crowded orchards begins when sections of the canopy are shaded for most of the day.

Don Batten analysed data collected by Xu in Fujian (latitude 24° to 25°N) and showed that yield (3 to 9 tonnes per ha) was correlated (r² = 64%) with March sunshine hours (20 to 220 h) over ten years. No similar relationship could be established for Alstonville in Australia (latitude 29°S), which has more sunshine hours than Fujian in September (equivalent to March): 244 h compared with 106 h. This work suggests that light may limit flower development in Fujian, although high sunshine hours would be expected to be correlated with higher temperatures and therefore earlier anthesis as proposed by Batten.

Table 3. Climates of different growing areas.

(Data presented for winter, spring and summer. Temperatures are means for the three months and rain is total for the three months.)

The reported reduction in fruit set during cloudy weather in Fujian could be due to lack of assimilates for flower development, but is more likely to be related to a direct effect of rain on the anthers or stigmas. Overcast weather may have also reduced bee activity, although their role in pollination is yet to be resolved.

Weather data in Zhang Zhou, Fujian over 22 years showed that in the first ten days of April, the average temperature was 18.4°C and rainfall 49 mm compared with 20.5°C and 43 mm for the middle 10 days of the month and 21.9°C and 39 mm for the last ten days. It was suggested that the early flowering failed because of cool, overcast weather during fruit set.

The effects of light (average irradiance of 4, 7, 9.5 or 13.5 MJ per m² per day (from 280 to 2,800 nm) on the growth and flowering of “Wai Chee” were studied over two seasons in Brisbane, Australia (latitude 28°S). Plants were shaded from June to September in year one, and from February to September in year two. Inflorescences emerged from August to September. More than 75 percent of terminal branches flowered, even if the plants were shaded several months before flowering. Average seasonal changes in light would not be expected to strongly influence flowering, unless overcast weather persists for several weeks.

Heavy shade for one week increased fruit drop in cultivar H1224 in Guangzhou (latitude 23°N). Branches were covered with shade cloth to reduce light levels to 10 percent of full sun. With shading at full bloom, the number of fruit per panicle after three weeks was 0.2 compared with 8.5 in the control. When shading began three weeks after full bloom, the number of fruit retained per panicle three weeks later was 0.8 and 2.2. Overcast weather is common in southern China, although most commercial areas in Asia have clear, dry weather during anthesis.

3.2.3 Effect of temperature on plant development

High temperatures increase the rate of shoot elongation. In contrast, a few weeks of cool weather in winter favour flowering. Extended periods of temperatures above 30°C during anthesis and fruit development can also reduce fruit set and possibly fruit quality.

The effects of temperature on vegetative growth were initially studied in Australia using seedlings. High day/night of 30°/25°, 25°/20° and 20°/15°C compared with 15°/10°C increased shoot growth in six selections, with a mean daily base temperature of 11°C. In a later study with marcots, trees flushed twice at 30°/25°C and once at 25°/20°C over 18 weeks. High temperatures reduced both the duration of flushing and the interval between flushes.

The time of floral initiation in “Calcuttia” and “Rose-Scented” was studied at Kanpur, India. Longitudinal sections of apical buds were sampled every one to two weeks from mid-September (year one) or mid-November (year two). The first signs of floral differentiation occurred about three to four weeks after the minima fell below 10°C, although sampling in the first year missed the actual start of floral initiation. Daily maxima at the start of these observations were as high as 30°C. These studies highlight the difficulty of relating productivity of fruit trees to weather.

Nakata and Watanabe from Hawaii provided the first direct evidence that low temperatures promote flowering. Marcots were placed outdoors or in a glasshouse, with some of the plants moved to a growth room at night. Average daily minima of 13.9°C in the growth room compared with 22.2° to 22.7°C outdoors and in the glasshouse increased flowering. The greatest number of inflorescences per branch occurred if the low temperatures were maintained until anthesis, although flowers were slower to develop compared to those on trees moved outdoors after floral induction. No plants flowered in a growth room at 23.9°C. Flowering only occurred when the night temperature was maintained at 15.6°C for two months. In Australia, all cultivars flowered at 15°/10°C and remained vegetative at 25°/20°C or higher.

Temperature also affects the rate of reproductive development, with panicles emerging earlier at 15°/10°C than at 20°/15°C, but taking longer to reach anthesis. This is consistent with the behaviour of cultivars in Australia. In cooler sub-tropical areas such as Nambour (latitude 27°S), panicles emerge from “Tai So” in May and fruit are harvested in December. However, in warmer tropical areas such as Cairns (latitude 17°S), fruit are harvested in November, although panicles do not appear until July.

In Australia, higher numbers of female flowers were associated with an average maximum during early flower development of 18°C, with lower numbers at 23°C. In contrast, the rate of flower opening was related to the number of flowers per panicle. It was concluded that areas with winter maxima above 25°C were not well suited for lychee culture.

The relationship between fruit set and weather is not well understood. There was no correlation between the proportion of female flowers setting fruit (19 to 26 percent) and daily maximums from 25° to 35°C or maximum vapour pressure from 1.5 to 3.5 kPa in northern New South Wales. However, continuous hot, dry conditions may reduce yields, since fruit set failed at a constant 33°C in a glasshouse. Bagging can improve fruit quality, possibly due to cooler temperatures and higher humidities.

Temperature has been shown to have strong effects on pollination, but these responses do not necessarily translate into better fruit production. The relationship between pollination and temperature was studied by using glasshouses maintained at 15° to 33°C. The normal time for fertilization to occur was estimated by counting pollen tubes in the ovaries. Maximum fertilization occurred when the number of pollen tubes per ovary did not increase with time after fertilization. Pollination was optimum at 19° to 22°C, with maximum fertilization obtained after seven days. At 15°C, pollen tube elongation was strongly inhibited. However, from 15° to 27°C, at least 10 percent of ovules contained pollen tubes indicating that they were fertilized. Such a level of fertilization appears sufficient for most cultivars to produce a high yield, although at 33°C, all female flowers abscised, suggesting a limitation for good yields when days are above 30°C for long periods.

In southern Queensland, the proportion of female flowers that set was greater with later flowering when the maximum was 30°C than with earlier flowering when the maximum was 24°C. In contrast, fruit set or yield in northern New South Wales could not be attributed to differences in average or maximum temperatures during anthesis. It was proposed that fruit set failed because the male flowers failed to produce pollen. The other possibility was that the early female flowers were sterile.

The average number of days from full bloom to harvest in “Shahi” in India was 68 days, equivalent to an average of 813 degree-days above 15°C. These authors choose the base temperature from data of Batten and Lahav that were based on stem growth not fruit development, although other workers reported that shoot growth still occurred with days of 15°C. Ray et al. showed a strong correlation (r² = 99%) between the number of days from full bloom to harvest and the number of degree-days above 15°C, although there were two years out of five with the same number of days to harvest, but with different numbers of degree-days. This agrees with the more rapid fruit development in tropical areas.

3.2.4 Effect of drought on plant development

Drought can assist flower initiation, but is not essential. In contrast, drought during fruit development generally reduces production.

Nakata and Suehisa studied the effects of irrigation in eight year old “Tai So” trees in Hawaii, where it is generally dry between April and November. The 'wet' treatment maintained y_S (soil water potential) at 45 cm depth above -0.03 MPa from June to February. Panicles emerged in December. The 'dry' treatment had an average y_S of about -0.5 MPa from June to August and then a y_S of -1.5 MPa from September to December. Heavy rain occurred in December and y_S rose to -0.03 MPa. In the 'covered' treatment, y_S declined from -0.03 MPa in October to -0.8 to -0.9 MPa during December and January, and then increased to -0.03 MPa in March after irrigation. Only 50 percent of tagged branches flowered in the 'wet' plot compared with 80 and 85 percent in the 'covered' and 'dry' plots, respectively. Average yields were 50, 71 and 84 kg per tree.

A similar trial was conducted in Israel with six year old trees of “Mauritius” (“Tai So”) and “Floridian” (“Brewster”?). It is generally dry from April to October. A week after water was withdrawn from a set of trees, y_S (30 to 90 cm depth) declined to -0.07 MPa. Irrigation was withheld for a further two weeks until the mature leaves started browning (equivalent to a noon y_L or leaf water potential of -3.2 MPa compared with -1.5 MPa in control trees). 'Dry' trees were then given limited irrigation of 1 mm per day for another week that would hardly balance evapotranspiration. Full irrigation at this time of the year was 3 mm per day. The severe drought in October inhibited leaf growth in November and increased flowering and yield. Flowering occurred after the trees were re-watered. These results demonstrate that drought can induce flowering, but the response is probably related to a shift in the timing of shoot growth. Several glasshouse experiments in Australia showed that drought had no direct effect on flowering.

Shoot growth is very sensitive to changes in tree water status. Menzel et al. examined the vegetative flushing of “Kwai May Pink” under different irrigation regimes in a glasshouse. Growth decreased as the level and duration of drought increased, but none of the trees flowered at high temperatures. A period of drought before flower induction may assist flowering by delaying early shoot growth until winter. This can be used in areas such as northern Thailand that have dry winters.

Once flower panicles are initiated, best fruit set is achieved when plants are well watered. A cyclic drought (predawn y_L of -2.0 MPa) achieved by watering the plants every four to seven days to field capacity reduced panicle growth and the numbers of flowers compared with plants watered daily (y_L above -0.7 MPa). Most of the flowers abscised prematurely in droughted plants and the few flowers that reached anthesis were male. These results indicate that trees should be irrigated from panicle emergence to prevent water deficits reducing fruit set, although they do not indicate a threshold y_L below which production is affected. Experiments in small pots may not necessarily predict the response of mature trees in the field, with a deep root system and slower development of drought.

There is very little information on the response to irrigation during fruiting. The results on hand indicate that there may be different effects on fruit production depending on the level and timing of the water shortage.

Batten et al. compared a set of unirrigated trees and trees irrigated weekly to replace 85 percent of potential evapotranspiration at Alstonville in Australia (latitude 29°S). Potential evaporation is the water use of a well-watered grass sward. This was not mentioned in the text. For a Class A pan with a wire bird cover surrounded by grass, potential evapotranspiration of the grass is about 85 percent of the evaporation from the pan. Consequently, the irrigated trees were watered to replace 72 percent of the pan evaporation (pan factor of 0.85 and a crop factor of 0.85). The eight to ten year old “Bengal” trees were growing in a deep, well drained clay soil and were droughted from flowering until harvest.

Predawn and noon y_L declined to -0.9 and -2.4 MPa in unirrigated trees, while minimum y_L in the controls were -0.4 and -2.0 MPa. It took six weeks before any significant difference in y_L between the two groups was noted. Fruit were 10 percent smaller in the unirrigated trees than in control trees, but the number of fruit was more than double in the dry treatment (26 fruit per panicle compared with 12 fruit per panicle in the controls). Greater fruit retention was attributed to less competition between leaf flushes and fruit, although no shoot growth data were presented.

The effects of irrigation on “Tai So” were studied in South Africa. A 'wet' group of trees was irrigated weekly to replace evapotranspiration, while a 'dry' set was allowed to dry out gradually over six months from panicle emergence. Minimum y_L declined to -2.8 MPa in the early afternoon in the 'dry' treatment compared with -2.2 MPa in the 'wet' treatment. Minimum y_L on the shaded side of the trees at 0900 h were -2.6 and -1.5 MPa in the 'dry' and 'wet' treatments, respectively. It took about six weeks before there were appreciable differences in tree water status between the two groups of plants. Drought reduced the number of fruit per tree, average fruit weight, flesh recovery and yield. The main reason for the lower yield in the 'dry' treatment was increased rate of fruit splitting just before harvest compared with control trees. The differences in the results in Australia and South Africa need to be resolved.

Skin cracking is a serious problem in many countries such as India where up to 50 percent or more of the crop may be lost. Temperatures are above 38°C and relative humidity below 60 percent during much of fruit development. However, it is not a major problem in Viet Nam, where the weather is less extreme.

The role of hot, dry conditions on fruit drop is not known. There have been no experiments in which humidity and temperature conditions have been controlled or the pattern of fruit drop has been correlated with daily weather data. Fruit drop in sub-tropical Australia was not related to rainfall after fruit set in irrigated orchards, although higher rainfall would be expected to increase relative humidity. Spotting bugs (Amblypelta nitida and A. lutescens) are more important factors in some areas, accounting for 25 to 99 percent of green fruit drop in several locations.

3.2.5 Predicting areas suitable for lychee production

The key factors to consider when assessing the potential of different areas for lychee are temperatures in winter that affect flower initiation, temperatures and light levels in spring which affect fruit set, and reliability of rainfall which affects fruit development. Normally temperatures below 20°C induce flowers, while flowering is irregular at higher temperatures, with the exception of a few tropical ecotypes in Thailand.

A short drought in winter may assist flowering, especially in the more tropical cultivars, but is not essential. Annual rainfall of 1,200 to 1,500 mm is probably required in the absence of irrigation. Long dry periods during fruit development will invariable reduce returns. This will limit production to the wetter areas in Asia.

The other critical part of the crop cycle is fruit set that is reduced when temperatures fall below 20°C for extended periods during flowering. Persistent cloud cover at this time can also be a problem. This could be a concern at higher elevation in some areas in southern Australia and elsewhere.

Olesen developed a model showing the relationship between potential flowering with latitude along eastern Australia (Figure 6). This was related to the number of days per year with mean temperatures below 20°C. At lower latitudes or more tropical sites, there were few days suitable for flowering, while at higher latitudes or more sub-tropical sites, there were several weeks of suitable temperatures. This model is supported by the relative performance of mature trees in the different areas. The data can be used to show the changes in mean temperature in July with latitude as well (Figure 7). You can then predict flowering in other environments if you have access to temperature data (Figure 8), with a plot of likely flowering versus mean temperatures for the coldest month.

The model can be used to estimate the reliability of flowering at different elevations, instead of latitude in Asia. These can be derived by estimating the change in mean temperature with elevation, using a base temperature for a site that is close to sea level (Figure 9). McAlpine et al. used a similar model to derive changes in temperature with elevation in Papua New Guinea. Other models are available, but they are generally similar, with temperature falling by about 0.6°C for each 100 metre rise in elevation. Once mean temperatures for the coldest month are determined, estimates can be made of flowering at different elevations for a more tropical location, say at a latitude of 12° (Figure 10). This analysis is dependent on the actual temperature at elevation being close to that predicted by the model. Previous work using data from five sites indicate a difference of ± 1.0°C between the predicted and actual temperatures. The reliability of the model was confirmed.

Figure 6. Relationship between number of days per year suitable for lychee flowering and latitude in Australia. (Latitude varies along 2,000 km of coastline. Data from Menzel et al. 2000).

Figure 7. Relationship between average temperatures in July and latitude along eastern Australia. (Latitude varies along 2,000 km of coastline. Data from Menzel et al. 2000).

Figure 8. Relationship between number of days per year with means below 20°C, and mean daily temperature in July in eastern Australia. (Data from Menzel et al. 2000. Mean temperatures in July have been calculated from Figure 7).

Figure 9. Relation between temperature in January and elevation at a tropical location (latitude 12°). (Analyses from McAlpine et al. 1983).

Figure 10. Relationship between number of days per year with means below 20°C, and elevation at 12° latitude.

3.3 Relationship between cropping and soil type

3.3.1 Soil type

Lychees can be found growing on a range of soils, including alluvial sands, loams, heavy clays, and soils with a high content of organic matter, lime or rocks. Trees perform best on well drained clay loams of medium to high fertility with a minimum one metre of well drained topsoil. They may die on heavy clay soils that become waterlogged. There can also be problems on very sandy soils that dry out during hot weather, and on calcareous soils with potential iron, zinc or manganese deficiencies. These soils need to be carefully managed. Slopes greater than 15 percent are also best avoided as they do not allow the safe use of machinery, and may erode.

In Guangdong, many of the newer plantings have been established on heavy clays. Traditionally, the best trees were found close to the rivers, on alluvial sands with good drainage and access to the water table. There were also many orchards planted in terraces 1.5 to 2.0 m wide in gravelly sandy loams and in swampy areas bisected by canals where the soil was built up in levees about 0.5 to 1.0 m above high tide.

In Guangxi, most of the trees are found on heavy red clays on slopes, although sometimes they are grown on sandy loams of alluvial origin along the rivers. The bulk of the red clays are of low to medium fertility, with only average concentrations of organic matter, phosphorus and potassium. The soils are generally acid, and need regular applications of lime or dolomite. Some of the trees are grown in mounds (less preferred) or on mounded rows to improve drainage during extended wet weather, however, these mounds may dry out in hot weather.

The soils in Fujian are high in clay, poorly drained and acid. When trees are grown in terraces, the planting site is generally filled with quality loam and organic matter to improve soil structure and fertility.

In Viet Nam, trees are grown in many different soils, from silty loams to clay loams with a wide range in colours from red, brown, yellow and grey. Physically, most of these soils are suitable for cropping, provided organic materials are added, and are well drained to at least one metre. However, chemical and pH levels vary, and these need to be managed carefully.

In India, well drained alluvial soils with access to the water table are considered ideal. Production is generally much lower in the poorly drained, heavy clays. In northern Bihar, there are many calcareous soils, with a pH of 7.5 to 8.0. Nutrition has to be carefully managed on these soils to avoid deficiencies of micronutrients such as iron.

3.3.2 Water relations and root growth

Lychees can withstand up to 14 days of immersion, provided the water does not become stagnant, but will die after prolonged waterlogging. Trees subjected to continued flooding in China are smaller than those on better drained soils. Poor drainage in heavy clays can increase the incidence of collar rots and root diseases. In southern Queensland, hilling of the soil along the rows to give ridges 0.5 m high is recommended in wet sites. The addition of drainage pipes can also assist growth in wet soils.

Nel observed a tremendous network of roots in “Tai So” growing down to one metre in sandy soils in South Africa, while trees growing in clay soils had a shallow root system. Most of the roots of an eight year old “Tai So” tree growing in a sandy clay loam overlying a heavy clay in Queensland were in the top 30 to 40 cm. Other experiments showed that soil type influenced total root density and feeder root distribution (depth of the soil where 80 percent of roots are located). There were more roots in a sand than in a clay, but a smaller proportion was found at depth (feeder root distributions of 0 to 20 cm and to 0 to 60 cm, respectively). About 90 percent of the roots were less than 2 mm in diameter, with no effect of soil type or depth.

Howard showed that although some roots were found below 300 cm in a deep calcareous sandy loam in India, most roots were located in the top 45 cm. The deep roots were, however, capable of absorbing enough water during the dry season to support a large crop.

3.3.3 Soil pH

Trees are capable of growing on either acid or alkaline soils, although there is little critical information on the optimum pH. Most growers aim for a pH between 5.5 and 6.0, although lower pH is probably acceptable. Nutrition management, especially the application of micronutrients needs to be modified at extremes of soil pH.

The pH in China is usually about 5.5, with the soils naturally acid or acidified by liquid fertilizers or organic mulches such as straw. In contrast, in India, many soils are alkaline with up to 30 percent free lime.

Table 4 shows the suggested rates of lime application for soils with different pH in Queensland. No more than 5 tonnes of lime per ha should be applied in a single application on sandy soils. Where more lime is required, a second amount should be applied three months later. Dolomite can be used instead of the lime, if soil magnesium concentrations are low.

Table 4. Lime requirement (tonnes per ha at 10 cm depth) for soils with different pH.

(Only apply lime when the soil pH (water) is lower than the target pH. Data from Phil Moody and Bob Aitken, Queensland Department of Natural Resources and Mines).

3.3.4 Salinity

There is little information on the response to excess salts. Lychee appears to be less sensitive than avocado or macadamia, but is still in the low tolerance class of plants. It is recommended that trees should not be irrigated with water having an electrical conductivity greater than 0.5 dS per m or about 500 mg soluble salts per litre. Damage sometimes occurs during dry weather, especially when young trees are over-fertilized. The tips and margins of the old leaves die.

Australian Scientists grew marcots in sand culture irrigated with 6 or 12 mM NaCl. At both concentrations, older leaves were shed with each new flush of growth. “Tai So” was more sensitive than “Bengal”, and this was reflected in greater uptake of salts. The concentrations of Na in the leaves of “Tai So” after 13 months in the control and 12 mM NaCl treatments were 240 and 22,000 ppm, respectively. Similarly, leaf Cl concentrations were 0.3 and 2.6 percent.

3.3.5 Mycorrhiza

Coville was the first to detect mycorrhiza in lychee. Fungi were isolated from root tubercles of seedlings grown in peat and sand, whereas no such tubercles were found on plants grown in the standard mix of loam, sand and manure. Seedlings with the tubercles were larger and had more roots than plants without the fungi. Kadman and Slor showed that “Tai So” seedlings were larger when grown in peat plus mycorrhizal soil compared with peat plus regular soil.

Pandey and Misra described the taxonomy, morphology and habit of the mycorrhiza. Rhizophagus litchi belongs to the vesicular-arbuscular group of phytomycetous endophytes. The endophyte could not be cultured on artificial media, the presence of living roots being necessary for its survival. Mycorrhiza were only found on short-lived sublateral roots. The fungi penetrated the roots through the epidermal cells into the cortex, whereas the root hairs, endodermis and vascular tissue were free of infection.

Since the earlier work of Coville, many authors have suggested that lychee requires mycorrhiza to grow satisfactorily, although healthy plants have been examined which were completely devoid of tubercles. In China and India, it is suggested that new plants be grown in soil taken from the vicinity of old trees to introduce the mycorrhiza. Further experiments are required to establish the role of these organisms in commercial production.

Bibliography

Batten, D. J. 1986. Towards an understanding of reproductive failure in lychee (Litchi chinensis Sonn.). Acta Horticulturae 175, 79-83.

Batten, D. J., McConchie, C. A. and Lloyd, J. 1994. Effects of soil water deficit on gas exchange characteristics and water relations of orchard lychee (Litchi chinensis Sonn.) trees. Tree Physiology 14, 1177-89.

Batten, D. J. and McConchie, C. A. 1995. Floral induction in growing buds of lychee (Litchi chinensis) and mango (Mangifera indica). Australian Journal of Plant Physiology 22, 783-91.

Chaikiattiyos, S., Menzel, C. M. and Rasmussen, T. S. 1994. Floral induction in tropical fruit trees: effects of temperature and water supply. Journal of Horticultural Science 69, 397-415.

Huang, H. B. 2001. Towards a better insight into the development of the arillate fruit of litchi and longan. Acta Horticulturae 558, 185-92.

Huang, X. M., Li, J. G., Wang, H. C., Huang, H. B. and Gao, F. F. 2001. The relationship between fruit cracking and calcium in litchi pericarp. Acta Horticulturae 558, 209-15.

Li, J. G., Huang, H. B., Gao, F. F., Huang, X. M. and Wang, H. C. 2001. An overview of litchi fruit cracking. Acta Horticulturae 558, 205-7.

McAlpine, J. R., Keig, G. and Falls, R. 1983. Climate of Papua New Guinea. Commonwealth Scientific Industrial Research Organization, Canberra pp. 89-101.

Menzel, C. M. 1983. The control of floral initiation in lychee: a review. Scientia Horticulturae 21, 201-15.

Menzel, C. M. 1984. The pattern and control of reproductive development in lychee: a review. Scientia Horticulturae 22, 333-45.

Menzel, C. M. and Simpson, D. R. 1987. Lychee nutrition: a review. Scientia Horticulturae 31, 195-224.

Menzel, C. M. and Simpson, D. R. 1988. Effect of temperature on growth and flowering of litchi (Litchi chinensis Sonn.) cultivars. Journal of Horticultural Science 63, 347-58.

Menzel, C. M., Rasmussen, T. S. and Simpson, D. R. 1989. Effects of temperature and leaf water stress on growth and flowering of litchi (Litchi chinensis Sonn.). Journal of Horticultural Science 64, 739-52.

Menzel, C. M. 1991. Litchi. In Plant Resources of South-East Asia Vol. 2. Edible Fruit and Nuts (E. W. M. Verheij and R. E. Coronel, Editors). Pudoc, Wageningen, The Netherlands pp. 191-5.

Menzel, C. M. and Simpson, D. R. 1992. Flowering and fruit set in lychee (Litchi chinensis Sonn.) in sub-tropical Queensland. Australian Journal of Experimental Agriculture 32, 105-11.

Menzel, C. M. and Simpson, D. R. 1994. Lychee. In The Handbook of Environmental Physiology of Fruit Crops Vol. II. Sub-tropical and Tropical. (B. Schaffer and P. C. Andersen, Editors). CRC Press, Boca Raton, Florida USA pp. 123-41.

Menzel, C. M., Oosthuizen, J. H., Roe, D. J. and Doogan, V. J. 1995. Water deficits at anthesis reduce CO₂ assimilation and yield of lychee (Litchi chinensis Sonn.) trees. Tree Physiology 15, 611-7.

Menzel, C. M. and Simpson, D. R. 1995. Temperatures above 20°C reduce flowering in lychee (Litchi chinensis Sonn.). Journal of Horticultural Science 70, 981-7.

Menzel, C. M. Olesen, T. and McConchie, C. A. 2000. Lychee, Longan and Rambutan. Optimising Canopy Management. Final Report to the Rural Industries Research and Development Corporation, Canberra 92 pp.

Roe, D. J., Menzel, C. M., Oosthuizen, J. H. and Doogan, V. J. 1997. Effects of current CO₂ assimilation and stored reserves on lychee fruit growth. Journal of Horticultural Science & Biotechnology 72, 397-405.

Shukla, R. K. and Bajpai, P. N. 1974. Blossom bud differentiation and ontogeny in litchi (Litchi chinensis Sonn.). Indian Journal of Horticulture 31, 224-8.

Stern, R., Adato, I. and Gazit, S. 1990. Autumn water stress as a means of increasing flowering and improving fertility of young litchi trees. Alon Hanotea 44, 391-4.

Stern, R.A. and Gazit, S. 1997. Effect of 3, 5, 6-trichloro-2-pyridyl-oxyacetic acid on fruitlet abscission and yield of 'Mauritius' litchi (Litchi chinensis Sonn.). Journal of Horticultural Science & Biotechnology 72, 659-63.

Yuan, R. C. and Huang, H. B. 1988. Litchi fruit abscission: its pattern, effect of shading and relation to endogenous abscisic acid. Scientia Horticulturae 36, 281-92.