VARIATION IN SEEDLING CHARACTERS OF THREE DIFFERENT COCONUT CULTIVARS AND THEIR USE IN IDENTIFICATION IN THE NURSERY

There is no standard guide for the identification of seedlings of different cultivars at the nursery. Diagnostic traits are necessary to detect illegitimates amongst hybrid coconut seedlings and for sorting out the types, in the case of accidental mixing during transport. Requests fora guide to the identification of planting material have been made by overseas buyers. Several metric traits of 3eedlings of three different Coconut (Cocos nucifera L) cultivars were studied in order to identify those of value in distinguishing between these types in the nursery. Seedlings of Dwarf, Tall and the hybrid, Dwarf x Tall were used.


INTRODUCTION
Distinguishing illegitimate Dwarf seedlings from the hybrids (Dwarf x Tall) is important in the production of hybrid seedlings. Such an identification of Dwarf types in the nursery is easier when Dwarf yellow or Dwarf red has been used as the female parent due to the presence of a colour marker. For Dwarf green there is no such marker so that distinguishing between the two cultivars is difficult. Accidental mixing of seednuts may take place during transport, especially while exporting and requests for a guide to identification have been made by buyers. The hybrid seednuts of Dwarf green x Tall can be differentiated from tall seednuts because of the small nut size and its characteristic features of the Dwarf nut but the Dwarf yellow x Tall seednut cannot be identified likewise. Hence the need to evolve a practical guide for the identification ofDwarf, Tall and hybrid seedlings of coconut is essential. Considerable variation in germination time between different cultivars of coconut has been reported (Foale 1968;Wuidart 1979) and Whitehead (1965 b) indica' ^d the rate of germination, as a character of taxonomic significance. Seedling characters were measured on Tall, Dwarf and hybrid seedlings and variation was studied in order to select which character or characters would be useful discriminators to distinguish among the cultivars.

MATERIAL AND METHODS
Tall x Tall and Dwarf green x Tall seednuts produced at the Isolated Seed Garden at Ambakelle by directed natural pollination were used. The. Dwarf green seendnuts were produced by self pollination by bagging the inflorescences. Tall  The final germination percentage achieved by the seednuts laid during the dry season (June 85) was much lower for all cultivars, with Dwarf and hybrid seednuts being more sensitive to the relatively dry conditions. In dwarfs the time taken for the sprouts to emerge was much longer in the seednuts laid in June compared to seednuts laid in November and the percentage germination was even lower than the hybrids. However, the differences in the pattern of germination among the three cultivars remained similar in both seasons.

Seedling characters at 6 months from laying
Multivariate discriminant analysis indicated highly significant differences between cultivars in height, girth, number of leaves, number of split leaves, length of petiole of first leaf and width of 2nd leaf (Table 1). However the two characters mentioned last, although statistically significant, were not very clearly different visually and hence were not considered to be of much practical Dwarf seedlings appeared to be the most vigorous at the early stages with as many as four leaves at 6 months from laying by which time splitting of leaves had already begun. These seedlings were about 85 cm tall with a mean girth of 9 cm. At 6 months, the Tall seedlings were relatively small with only two unsplit leaves; the mean girth at collar was 7 cm and height 60 cm.  Visual observations were also found to be useful, For instance at about 2 to 3 months from laying, Dwarfs had about 2 to 3 leaves which were pale green in colour, curled and smooth in texture. At a later stage Dwarf leaf blades appeared narrower than Tall leaves and the length from the tip of the leaf up to the point of dischotomy was greater. Consequently Dwarf leaves appeared more elongated than Tall leaves. In Dwarfs the canopy width was lesser as the leaves were pointing upwards whilst the Tall had a good canopy spread. Table 2 indicates the percentages of seedlings correctly classified into each treatment and the percentage misclassified, using the most significant variables in the analysis at 6 month and 9 month stages. The percentages correctly classified under each treatment were more than 75% and this confirms that the variables used are suitable enough to distinguish among the cultivars. It was seen that 20% of Tall seedlings had been misclassified as hybrids at 6 months but only 7% of hybrids were incorrectly classified as talis. The chance of Dwarfs being misclassified as Tall or hybrid was lesser. At 9 months the misclassification percentage of hybrids and Tall were more than at 6 month stage and therefore it would be much easier to distinguish the cultivars at 6 month stage. These observations were confirmed by the generalized squared distances (D 2 ) values obtained according to Mahalanobis (1936) ( Table 3). The shortest distance is between hybrid and Tall and the longest is between Tall and Dwarf. ..

Table 2
Percentages classified into treatments from the multivariate discrimi nant analysis using the significant characters (Table 1

CONCLUSION
Almost all characters measured showed continuous variation and therefore was difficult to categorize into discrete groups. One feature alone may not usually serve to distinguish between cultivars but several characters which show small but significant and detectable differences between cultivars may be used in combination to identify these cultivars. Certain detectable characters are visible at various stages and identification has to be carried out during this whole period for higher accuracy. Recommendations can be given to reject any seedling that sprouts before the 6th week in discarding the illegitimates from hybrid seedlings. Tails have a distinctly different period for sprouting from Dwarfs; so these can be identified at the sprouting stage itself. However, for distinguishing tall from hybrids other characters indicated may be used. 28