APS_October 2018

J ournal of the A merican P omological S ociety

270

Figure 5. Dendogram of 52 centennial carob trees constructed on the base of morphological traits recorded, using rescaled squared Euclidean distance. Figure 5. Dendogram of 52 centennial carob trees constructed on the base of morphological traits recorded, using rescaled squared Euclidean distance.

represented by only the Baasir MLC1 tree, with the longest and heaviest pods with high seed number per pod. Cluster II was divid- ed into two main sub-clusters based on pod length and width, whereas sub-cluster II.1 contained 22 trees with small to medium pod size of which seven were from Mount Lebanon, eight from the South, six from the North and one from Beirut. Sub-cluster II.2 contained 10 trees with big pods of which six were from Mount Lebanon, two from the North, one from the South and one from Bei- rut. At 2.5% dissimilarity and below, trees of the South (which have small and light pods and seeds) and the ones of Mount Lebanon (with heavy and long pods and seeds) are clearly differentiated, while the trees of Bei- rut and the North are distributed in various subgroups. Discussion  In Lebanon, carob has always been an im- portant component of the natural landscape and the traditional agroforestry system, but

morphological variability of the species at country level has been rarely assessed (Tal- houk et al ., 2005). The present work is the first phenotypic assessment of the Lebanese carob germplasm, using a set of morphologi- cal traits related to tree, pod, seed and leaf. Morphological descriptors have been widely used to assess variability of carob germplasm and to differentiate cultivars and promising trees (Batlle and Tous, 1997; Sidina et al ., 2009; Naghmouchi et al ., 2009; Sidina et al ., 2009; Tous et al. , 2013).  The most variable traits found in this study included number of aborted seeds per pod, seed yield, pulp weight, pod weight and pod volume with respective coefficients of varia- tion of 93.9, 44.6, 42.9, 39.3, and 36.1. These traits are mostly complex and controlled by polygenetic features, and are strongly influ- enced by the environmental factors (Sidina et al ., 2009). On the other hand, seed traits var- ied less and seem less influenced by environ- mental factors, similar to the results reported previously for the Spanish cultivars, and the 31