Leaf Structure of Jamaican Upper Montane Rain-Forest Trees

Abstract

Leaves of 50 tree species from five montane rain-forest types at c. 1500 m in Jamaica were examined microscopically. The following characters were measured (mean values for all species in parentheses): blade thickness (237iLm), thickness of the outer walls of the upper (5.4[um) and lower (4.3[Lm) epidermis (includirg cuticles), the non-palisade mesophyll to palisade mesophyll ratio (2.1:1) the incidence of: a hypodermis (28%); crystals (76%); sclerenchyma along vascular bundles (92 %); and transcurrent sclerenchyma (26%); the density of stomata (299 mm -2) and mean guard cell length ( 17.6i,m). There were no statistically significant differences bqtween four forest types studied in detail for any leaf character, though the most physiognomically impoverished forest had the most xeromorphic leaves. Correlations between leaf characteristics were examined. Lamina thickness is positively correlated with both upper and lower outer epidermal wall thickness. There is no correlation between lamina thickness and specific leaf area, indicating that increased leaf thickness is due only to increased water content or air space rather than increased dry weight. The Jamaican montane forest leaves are compared with those from other tropical rain forests. Consistent differences between montane and lowland rain forests are smaller leaves with thicker outer epidermal walls and lower non-palisade to palisade ratio in montane rain-forest leaves. All other measured characters do not show consistent differences. IT HAS LONG BEEN KNOWN that leaves of wet trolpical montane forest trees have different characteristics from those of the lowland rain forest (Schimper 1903), but only recently have examples of detailed studies on their structure been published; by Howard (1969) for Puerto Rico, and Grubb (1974) for New Guinea. This paper presents the results of a study of tree leaves collected from montane forests at c. 1550 m in the Blue Mountains in Jamaica, prior to an investigation of their physiology. Grubb and Tanner (1976) have described 10 forest types from this area. The leaves for this study were collected from the Mor Ridge, Mull Ridge, Wet Slope, Dry Slope, and Gap (Gully) forests. Of these, the Mor Ridge forest is most poorly developed, having a low canopy (57 m), relatively few tree species (16), and low foliar mineral concentrations (mean nitrogen 1.05 % dry weioht and phosphorus 0.05%, Tanner 1977). Mull Ridge, Wet Slope, and Dry Slope forests are better developed with a higher canopy (8-13 mi), more tree species (c. 35 each), and higher foliar mineral levels (means for Mull Ridge forest: nitrogen 1.61%, phosphorus 0.07%). Gap forest is the best developed, having the highest canopy (12-16 m), slightly fewer tree species (26), and the highest foliar mineral levels (means: nitrogen 1.76%, phosphorus 0.10%). The study was designed with the following questions in mind: are the differences in structural and physiognomic development of the Jamaican forest types reflected in differences in the internal structure of their leaves? Are certain leaf characteristics associated and, if so, how might this be related to selection pressures? Are leaf characteristics similar in geographically separated but structurally similar forests, and do they differ between structurally different forests? In the long term we seek answers to such questions as: which environmental factors are most important in determining leaf characteristics? And, are differences in forest development, e.g., lowland rain forests vs. uppar montane rain forest (seizsu Grubb and Tanner 1976), predictably reflected in the structural characteristics of their leaves?