Oxygen isotope ratios were determined in leaf cellulose from two plant species at Barro Colorado (Republic of Panama) in 4 different plots, two of which were undergoing an irrigation treatment during the dry season. There is a gradient in δO values of leaf cellulose from the understory to canopy leaves, reflecting the differences in relative humidity between these two levels of the forest. This gradient is most pronounced in irrigated plots. For irrigated plots there was a highly significant correlation between δO and δC values, which was not observed in control plots. This relationship can be explained by humidity controlling stomatal conductance. Low humidity affects δO values of leaf water during photosynthesis, which isotopically labels cellulose during its synthesis. Low humidity also decreases stomatal conductance, which affects discrimination against carbon-13 by photosynthetic reactions, thus affecting the δC values of photosynthates. WUE values calculated by using plant carbon and oxygen isotope ratios were similar to those observed with gas exchange measurements in other tropical and temperate area. Thus the concurrent analysis of carbon and oxygen isotope ratios of leaf material can potentially be useful for long term estimation of assimilation and evapotranspiration regimes of plants.

In a Neotropical moist forest at Barro Colorado Island, Panama, °13C values of CO2 in air and °13 values of leaf tissue exhibit parallel patterns of variation between the forest floor and the canopy. During the daytime, °13C values of CO2 from air sampled at 1 m and 0.5 m were significantly less than that at 25 m. Based on mass balance equations, up to 18% of the CO2 in air at 0.5 m above the forest floor is from respiration. Respired CO2 is responsible for 31 and 37% of the variation in isotope composition in leaves of two species of herbaceous bamboo grown in a well—ventilated sun treatment and in the forest understory. Respired CO2 accounts for 45—70% of the difference in °13C values between understory and canopy leaves for three species growing in large—scale irrigation and control treatments. Understory leaves of these species show °13 values consistent with higher ratios of intercellular to ambient CO2 in irrigated relative to control treatments. Estimates of water—use efficiency from leaf carbon isotope content should be corrected for the contribution of the carbon isotope composition of respired CO2 in closed—canopy forests.

No abstract