Production and distribution of dry matter in trees of Coffea arabica L. in Kenya as affected by seasonal climatic differences and the presence of fruits

SUMMARY Standard growth analysis procedures were used to study the production and distribution of dry matter in 3 1/2 to 5-year-old coffee trees through their first and into their second commercial fruiting year. The trees were growing in the field and were treated according to normal commercial practice. Up to fifteen fruiting and deblossomed trees were harvested on each of seven occasions, at intervals of 63–90 days. The dry weights of four aerial fractions, five root fractions and all fallen, picked and pruned material were recorded. The net assimilation rate (E) of deblossomed trees was as large as that recorded in East Africa for coffee seedlings (0·13 g dm-2wk-1). Fruiting trees increased in dry weight faster than deblossomed trees, even when their leaf area was 30% smaller. Their E was up to 0·19 g dm-2wk-1when expressed on a leaf area basis and up to 0·16 g dm-2wk-1even when the total green fruit surface area was included in the calculation. In the hot, dry season, January–February 1968, all parts of the trees increased in dry weight relatively slowly, except the thin roots (< 3 mm diameter) which, in this season, took about 10% of the dry weight increment. The trees had been pruned in December and E of deblossomed trees was only 0·09 g dm-2wk-1. Flushes of shoot growth occurred at the beginning of the Long Rains 1967 and 1968 (February–March). During the 1967 shoot growth flush, 61% of the dry weight increment was used in the production of new, large leaves with a mean specific leaf area of over 140 cm2g-1. The thin roots took about 10 % of the increment, but the thicker roots increased in weight very slowly. The branches and trunk extended rapidly, but their radial growth was relatively slow. Following the shoot growth flushes, the leaf area ratio of the trees was large (0·38 dm2g-1in 1967) and during the Long Rains (April–June 1967, 1968), when conditions were favourable for photosynthesis (E 0·13 g dm-2wk-1), all parts increased in dry weight relatively rapidly, although in 1968 the thin roots took as little as 3 % of the total increment. Leaf area ratio decreased during the Long Rains owing to a large decrease in specific leaf area (in 1967: 118 to 95 cm2g-1). Pruning was carried out in June and new leaf production during the cool, dry season, July to mid-September, was very slow. Consequently, the total dry weight increase of the trees during this season was relatively small, although the E of deblossomed trees was 0·13 g dm-2wk-1. The leaves, which took 33 % of the increment, decreased further in specific leaf area (to 83 cm2g-1) and most root fractions increased in dry weight rapidly, the thin ones taking 17 % of the increment. Light fruiting in 1967 did not affect the seasonal periodicity in growth described. Both light and heavy fruiting tended, eventually, to lessen the dry weight increase of leaves and thin roots proportionately more than that of the trunk and thick roots. In 1968 fruiting trees retained over 8000 fruits per tree, which took over 70 % of the dry weight increment during the 1968 Long Rains, and became 36 % of the trees' dry weight. Some rootlets on these trees decreased in weight. Aspects of the productivity, growth periodicity and fruiting of coffee are discussed, and some management implications are noted.