Estimation of leaf emergence rates of bananas

Abstract

Leaf emergence rates (LER, leaves/day) of 17 AAA genome banana cultivars (Musa sp., subgroups Gros Michel, Cavendish, Red/Green-red) were estimated using a model based on mean monthly air temperature (13.3-27.9¦C), daylength (10.2-13.8 h day), age of planting (48-1361 days), plant density (1111-2222 mats/ha) and cultivar stature (0.88-4.15 m). The LER model involved the five factors in a multiplicative form and was developed initially for one set of 612 observations on all 17 cultivars at Alstonville, N.S.W. but at one density only. The model indicated that LER increased with increasing temperature up to an optimum of c. 28.5¦C. LER increased during plant establishment to a maximum at 64 days, then decreased to a stable value with increasing age. LER increased with increasing daylength, but decreased with increasing stature. The model predicted LER without bias for 16 of the cultivars, but slightly over-predicted LER in winter for the only representative of the Gros Michel subgroup (Highgate).Examination of independent data from South Africa revealed that LER decreased with increasing plant density. After the LER model had been adjusted for this effect, it gave an excellent prediction of LER for 229 observations in 20 additional data subsets from Alstonville, Innisfail (Queensland), Burgershall (South Africa) and Azaguie (Ivory Coast), with LERobs = (1.03 ¦ 0.04) LERpred. - (0.005 ¦ 0.004) and coefficient of determination of 0.71. After pooling all 841 observations into one data set, the coefficients of all five factors of the LER model were re-estimated. There were only minor revisions to the coefficients previously obtained, but LER was predicted with slightly greater precision than in the validation test above.