WITHIN-YEAR AND AMONG-YEAR VARIATION IN THE LEVELS OF HERBIVORY ON THE FOLIAGE OF TREES FROM A MEXICAN TROPICAL DECIDUOUS FOREST

Discrete (i.e., one-time) measurements of the levels of herbivore dama ge on the foliage of trees (16 species) were made for three consecutiv e years in the Chamela dry forest. There was a marked similarity in th e overall levels of damage despite the fact that there were considerab le climatic differences among years and that some of the species showe d at least one significant among-year difference. In one of the three years of study, long-term measurements of damage (i,e., on individuall y marked leaves) were also made to assess within-year variation (by ca lculating the races of herbivory early and late in the wet season) and to assess the reliability of che discrete measurements. With this typ e of measurement the mean leaf area loss of the studied species was 17 percent. This value is higher than the values commonly reported for t ropical forests. An outstanding example of defoliation was that of Jat ropha standleyi, with 73 percent leaf area eaten. There was a very mar ked within-season variation in the rates of herbivory: during the earl y parr of the rainy season the mean race (0.352% leaf area eaten per d ay) was considerably higher chan lace in the season (0.097). Analyses of leaf nitrogen and water content early and late in the season indica ted chat these components of the nutritional quality of the foliage de crease as the season progresses. The discrete measurements of herbivor y underestimated the true (long-term) values of leaf damage. A discrep ancy ratio long-term/discrete measurement varied considerably among sp ecies with an overall mean of 1.99. Thus, in general, discrete measure ments underestimated herbivory by half. It appears that che major caus e of che underestimates of the discrete measurements is that the leave s of some species are eaten completely and thus missed in a one-rime m easurement. Another disadvantage of the discrete measurements is chat information on che timing of damage is lost. This study showed that fo livory is nor only concentrated to the season of leaf availability, bu t chat it occurs within an even more restricted period.