RESUMEN
The water relations of Pentaclethra macroloba (Willd.) Kuntze, a dominant, shade-tolerant, tree species in the Atlantic lowlands of Costa Rica, were examined within the forest canopy. Pressure-volume curves and diurnal courses of stomatal conductance and leaf water potential were measured in order to assess differences in water relations between understory, mid-canopy and canopy leaves. Leaves in the canopy had the smallest pinnules but the largest stomatal frequencies and stomatal conductances of the three forest levels. Osmotic potentials at full turgidity decreased with height in the forest; in the canopy and midcanopy they were reduced relative to those in the understory just enough to balance the gravitational component of water potential. Consequently, maximum turgor pressures were similar for leaves from all three canopy levels. Bulk tissue elastic modulus increased with height in the canopy. Leaf water potentials were lowest in the canopy and highest in the understory, even when the gravitational component was added to mid-canopy and canopy values. As a result, minimum turgor pressures were also lowest in the canopy compared to those at lesser heights, and approached zero in full sunlight on clear days.Osmotic potentials at each canopy level were similar for both wet and dry season samples dates suggesting that seasonal osmotic adjustment does not occur. Despite lowered predawn water potentials during the dry season, turgor was maintained in the understory by reduced stomatal conductances.
RESUMEN
Whole ecosystem CO2 flux under ambient (340 µl/l) and elevated (680 µl/l) CO2 was measured in situ in Eriophorum tussock tundra on the North Slope of Alaska. Elevated CO2 resulted in greater carbon acquisition than control treatments and there was a net loss of CO2 under ambient conditions at this upland tundra site. These measurements indicate a current loss of carbon from upland tundra, possibly the result of recent climatic changes. Elevated CO2 for the duration of one growing season appeared to delay the onset of dormancy and resulted in approximately 10 additional days of positive ecosystem flux. Homeostatic adjustment of ecosystem CO2 flux (sum of species' response) was apparent by the third week of exposure to elevated CO2. Ecosystem dark respiration rates were not significantly higher at elevated CO2 levels. Rapid homeostatic adjustment to elevated CO2 may limit carbon uptake in upland tundra. Abiotic factors were evaluated as predictors of ecosystem CO2 flux. For chambers exposed to ambient and elevated CO2 levels for the duration of the growing season, seasonality (Julian day) was the best predictor of ecosystem CO2 flux at both ambient and elevated CO2 levels. Light (PAR), soil temperature, and air temperature were also predictive of seasonal ecosystem flux, but only at elevated CO2 levels. At any combination of physical conditions, flux of the elevated CO2 treatment was greater than that at ambient. In short-term manipulations of CO2, tundra exposed to elevated CO2 had threefold greater carbon gain, and had one half the ecosystem level, light compensation point when compared to ambient CO2 treatments. Elevated CO2-acclimated tundra had twofold greater carbon gain compared to ambient treatments, but there was no difference in ecosystem level, light compensation point between elevated and ambient CO2 treatments. The predicted future increases in cloudiness could substantially decrease the effect of elevated atmospheric CO2 on net ecosystem carbon budget. These analyses suggest little if any long-term stimulation of ecosystem carbon acquisition by increases in atmospheric CO2.