干旱区生态环境知识资源中心

In the arid southwest of North America, winter precipitation penetrates to deep soil layers, whereas summer "monsoon" precipitation generally wets only surface layers. Use of these spatially separated water sources was determined for three dominant tree species of the pinyon-juniper ecosystem at six sites along a gradient of increasing summer precipitation in Utah and Arizona. Mean summer precipitation ranged from 79 to 286 mm, or from 18% to 60% of the annual total across the gradient. We predicted that, along this summer rainfall gradient, populations of dominant tree species would exhibit a clinal off-on response for use of water from upper soil layers, responding at particular threshold levels of summer precipitation input. This prediction was largely supported by our observations of tree water source use over a two-year period and from irrigation experiments.

Hydrogen and oxygen stable isotope ratios (deltaD and delta O-18) of tree xylem water were compared to that of precipitation, groundwater, and deep and shallow soil water to distinguish among possible tree water sources. deltaD-delta O-18 relationships and seasonal xylem water potential changes revealed that trees of this ecosystem used a mixture of soil water and recent precipitation, but not groundwater. During the monsoon period, a large proportion of xylem water in Pinus edulis and Juniperus osteosperma was from monsoon precipitation, but use of this precipitation declined sharply with decreasing summer rain input at sites near the regional monsoon boundary in Utah. Quercus gambelii at most sites along the gradient used only deep soil water even following substantial inputs of summer rain. Populations of Quercus at sites with the highest average summer precipitation input, however, predominantly used water in upper soil layers from recent summer rain events. Soil temperature correlated with patterns of summer precipitation use across the gradient; high soil temperatures north of the monsoon boundary may have inhibited surface root activity for some or all of the three tree species.

Irrigation experiments with deuterium-labeled water revealed that Quercus gambelii in northern Arizona and southern Utah did not use water from surface layers. In contrast, Juniperus osteosperma at these sites responded significantly to the irrigations: between 37% and 41% of xylem water originated from irrigations that wetted only the top 30 cm of soil. Responses by Pinus edulis to these irrigations were variable; uptake of labeled water by this species was greater in September at the end of the summer than during the hot midsummer period. Inactivity of Pinus roots in midsummer supports the hypothesis that root activity in this species is sensitive to soil temperature.

Seasonal patterns of leaf gas exchange and plant water potential corresponded to the seasonality of rainfall at different sites. However, no correlation between a species’ ability to use summer rainfall and its tolerance to water deficits at the leaf level was found. Midday stomatal conductance (g(s)) for Pinus needles approached zero at predawn water potentials near -2 MPa, whereas g(s) in Quercus and Juniperus declined to zero at -2.8 and -3.7 MPa, respectively. The relationship between photosynthesis (A) and g(s) was similar among the three species, although Quercus maintained higher overall rates of gas exchange and tended to operate higher on the A/g(s) curve than the two conifers. At sites in eastern Arizona where Quercus fully used moisture from summer rains, leaf gas exchange characteristics were similar to those of Pinus and Juniperus.