B23E-01 INVITED
Linking roots and rhizospheres to hydrological processes
There is ample evidence that shows how plants can exert very significant and often dominant –controls" over the manner and magnitude by which water and other soil-borne resources cycle through diverse ecosystems on Earth. The use and redistribution of soil water resources by root systems has been a particularly important addition to our understanding of how the movement of soil water resources can impact hydrological processes at a range of scales. When soil and plant water relations data are coupled with land-use and climatic change data and predictive models for seasonally-dry ecosystems they have revealed new insights into how the water cycle is also changing and the role that plant root functions plays in shaping fundamental aspects of the hydrological cycle. I will highlight the ways my research group as well as the work of others have used a range of methods to explore the links between roots and rhizospheres and hydrological processes. The detailed analyses of the stable isotope composition of plant and soil water and precipitation and the temporal and spatial patterns of water use by diverse trees in temperate and tropical biomes when coupled with ongoing modeling research has revealed new insights into how belowground and aboveground plant water use behaviors can impact the manned and magnitude of water cycling on local and regional scales. Further, new results clearly show the impacts that plant water uptake and use have on ecosystem carbon fixation and both temperature and precipitation patterns over vast regions like the Amazon as well as other parts of the globe covered by trees and deeply rooted woody vegetation. The combination of empirical and theoretical research results shows that plants can help sustain water recycling, can significantly impact carbon and nutrient cycles, and impact regional climate, drought and its seasonality thereby establishing a direct link between plant functioning, resource movement and the climate system across the globe.
B23E-02
Hydraulic Redistribution of Soil Water in a Drained Loblolly Pine Plantation: Quantifying Patterns and Controls over Soil-to-Root and Canopy-to-Atmosphere Interactions
The conversion of wetlands to intensively managed forest lands in eastern North Carolina is widespread and the consequences on water and carbon balances are not well studied. Quantification of evapotranspiration (ET), tree transpiration and their biophysical regulation are needed for assessing forest water management options. We characterized vertical variation in the diurnal and seasonal soil volumetric water content at 10 cm intervals to evaluate changes in water availability for root uptake and monitored eddy covariance ET and tree transpiration (sap flux) in a drained Loblolly pine (Pinus taeda L.) plantation. We also quantified the magnitude of hydraulic redistribution (HR), the passive movement of soil water from deep to shallow roots, to identify factors affecting the seasonal dynamics of root water uptake, root and plant water potentials and stomatal conductance. Soil water content varied with soil depth and total water use from the upper 1m peaked between 4 and 6.5 mm/day during the growing season and was strongly correlated and similar to ET (ET represented 90-95% of total water depletion). After periods of more than 10 days without rain, water extraction shifted to the deeper layers, and recharge from HR approached 0.5 mm/day in the upper 60 cm. However, the upper 30cm accounted for 40% of total water depletion from the upper 1m at peak water uptake (>4 mm/day), and increased to 65% during days of low water uptake (<2 mm/day), illustrating the contribution of deeper roots to water uptake during days of high evaporative demand. This result was supported by the fact that deep roots (from 30-50cm) accounted for 65% of the total water redistributed. Because of stomatal regulation to prevent water potentials from reaching critical values that would cause significant loss of tree hydraulic conductivity, maximum tree transpiration during high evaporative demand remained constant at around 3 mm/day. Tree transpiration represented on average 60% of ET. However, it represented only 50% of ET on days following rain events and up to 80% of ET after prolonged periods without rain. We propose that HR prevented predawn water potentials from decreasing during periods of increasing soil water deficit, therefore maintaining a constant driving force for water uptake of around 1.7 MPa. It was thought that HR was an important mechanism for maintaining shallow root function during drought and preventing total stomatal closure but our study shows that even in wet conditions with soil water potentials never dropping below -0.6 MPa, HR may play a role in wetland hydrological balance. This first approximation of the extent of HR in this ecosystem suggests that it is likely to be an important process in wet forests of North Carolina.
B23E-03
Decoupling structural and environmental determinants of sap velocity
Characterization of transpiration based on the water use of individual tress has the advantage of preserving vital information on the plant-environment functional links and flux partitioning between species and landscape areas. Whole-tree transpiration has been estimated by means of sap velocity probes, which offer the dual advantages of practicality and repeatability. However, the assumptions underlying the technique require careful verification in order to determine total sap flow from point-based estimates of sap velocity. Our work presents a novel theoretical framework for the study of individual tree sap flow that incorporates both spatial and temporal variability in sap velocities. The instantaneous sap velocity at any point in the radial profile of xylem tissue is defined as the product of two components: (1) a time-invariant sap velocity distribution linked to the species- specific anatomical and structural properties of the conducting xylem, and (2) a time-varying term linked to the dynamics of the atmospheric water demand and available soil moisture. The separation of structural and temporal variation in sap velocity observations provides a direct mechanism for investigating how sap flow is governed by variation in environmental conditions as well as a means for comparing characteristic rates of plant water use among individuals of varying size. Most critically, this approach allows for a consistent and physically meaningful method for extrapolating point observations of sap velocity across the entire depth of conducting xylem. Experimental evidence supports our theoretical framework in the case of a population of sugar maples in a mixed deciduous forest, where observations were taken from a wide range of tree sizes, under varying soil water availability and atmospheric transpiration demand. We have also applied our approach to a small homogeneous sample of dwarf apple trees in a managed orchard, with favorable results. While these results require further confirmation in order to be generalized, they nevertheless offer the basis to improve both the specific sampling strategies used to estimate whole-tree transpiration using sap velocity probes as well as methods employed to upscale water use of individual trees to larger scales for evaluation of landscape water balance.
B23E-04
Effects of Post-fire Succession and Edaphic Conditions on Tree Transpiration in a Boreal Black Spruce Forest
Boreal forest ecosystems play an integral role in global climate change because of their large land area and ability to store large quantities of carbon. Quantifying and explaining tree water use in both well- and poorly- drained soils and across successional development is critical in understanding the influence of physiological processes on carbon, water, and energy cycling. Four black spruce stands burned in 1850, 1930, 1964, and 1989 were chosen for this research because they had been shown in previous studies to represent critical stages of forest development that capture the successional impacts of both leaf area and species composition change. We hypothesized that tree transpiration will differ between well- and poorly-drained areas and with age due to 1) tree size and age and edaphic-related hydraulic adjustments and 2) tree size will be explained by species specific growth differences from edaphic conditions. Sap flux, leaf water potential (\PsiL), site specific allometric relationships between sapwood area and leaf area and soil properties such as texture and organic matter depth in each of the four burn ages were utilized to test these hypotheses. Results show that sap flux for Picea mariana at the 1964 burn age differed between well- and poorly-drained soils when scaled per unit xylem area with trees located on poorly-drained soils experiencing higher sap flux rates than trees in well- drained areas (101.79 & 83.02 g cm-2 day-1 respectively). However, when scaled to transpiration on a per tree basis, taking tree size into account, trees on well-drained soils had higher rates than those in poorly- drained locations (366.96 & 216.82 g tree-1 day-1 respectively). The presence of Pinus banksiana and Populus tremuloides in the well-drained areas increased stand transpiration rates for these areas considerably as compared to the poorly-drained areas. Midday \PsiL for all four burns show no significant difference between well- and poorly-drained (average midday \PsiL = -1.23 & -1.29 MPa respectively) sites for Picea mariana (t-value = -0.591, df = 6, p-value = 0.576). This indicates that tree size, which is constrained by growth and anaerobic conditions, drives differences in tree transpiration for well- and poorly-drained soils.
B23E-05
Drivers of Variability in Water Use of Native and Non-native Urban Trees in the Greater Los Angeles Area
A number of cities, including Los Angeles, have started programs aimed at increasing urban tree cover, due to growing recognition that urban trees can mitigate urban heat island effects, storm water runoff, atmospheric CO2 emissions and pollution, and energy expenditures. There is considerable interest in trying to quantify the magnitude of these tree-induced benefits, yet water use and water relations of urban trees have rarely been studied. Urban trees are exposed and presumably adapted to a different set of growing conditions than natural trees, including growth-enhancing environmental alterations such as irrigation and fertilization, but also detrimental conditions like increased ozone and restricted rooting area. In order to study the factors which control whole tree water use of common species in the Los Angeles Basin urban forest, four sites in Los Angeles and Orange County have been instrumented with sap flow and meteorological sensors. These sites allow comparisons of the water use of the same species under different environmental conditions: urban vs. non- urban, coastal vs. inland climates, and comparisons of native vs. non-native species. We found that over a similar vapor pressure deficit (daytime average 0.5-3.5 kPa; VPD) range, California native sycamores in the inland urban site exhibited far higher (up to 3 times) rates of daily sap flow (g cm-2 d-1) than sycamores in an inland non-urban (natural riparian, non-irrigated) site. However daily sap flow rates of sycamores at an urban coastal site, where VPD is generally lower (daytime average <1.2 kPa), were much lower than in the inland urban site even at the same VPD. In contrast, Canary Island pines showed lower daily sap flow rates over the same VPD range at an inland, urbanized site in comparison to a coastal, irrigated urban site. These differing behaviors in contrasting environments could result from differences in a number of allometric or hydraulic properties. In order to determine what drives the observed differences we are also measuring leaf area, plant water sources, vulnerability to cavitation, and leaf nutrient and isotopic composition. Our observations thus far reinforce the idea that measurements of natural forests cannot accurately predict the physiological responses and water relations of urban trees.
B23E-06
Dynamic Responses and Controlling Factors of Evapotranspiration in a Sagebrush-Steppe Ecosystem in North America
Current and future changes in precipitation in North America will alter key ecosystem such as water cycling and surface energy balance. Although sagebrush-steppe is the largest ecosystem in North America, observations of ET at ecosystem-level have been limited. We used eddy covariance and associated above- and belowground micrometrolgy techniques to determine the dynamics and controlling factors of ET. Measuring occurred in a sagebrush-steppe ecosystem in south-central Wyoming, USA for two growing seasons (2004 and 2005). 2004 had a dry spring while 2005 had an unusually high amount of precipitation in late spring. During the measurement years, the highest rates of daily ET (1.7 mm day-1) occurred in the early summer season (June, 2005) driven by high soil moisture and contribution from developing grasses and forbs. The general pattern of daily ET was more strongly correlated with the deep soil moisture (15 – 45 cm) than the shallow soil moisture (4 cm). Net radiation was a major driver of ET regardless of soil moisture availability while vapor pressure deficit (D; atmospheric drought) was a major driver of ET when the ecosystem was not limited by soil moisture reflecting stomatal closure in response to atmospheric drought. The influence of D was also reflected in the decoupling factor (Ømega ranged from 0.2 to 0.7), showing the lowest values (near 0.2) when D was high. Penman-Monteith model ET was calculated using surface conductance estimated from a Jarvis-type model which included functions of D, light, and soil moisture. Model results indicate that interactions of soil moisture, light and D components are critical predictive understanding of ET in the sagebrush-steppe ecosystem.
B23E-07
What do the towers see at night? An exploration of nocturnal eddy covariance evapotranspiration fluxes from three adjacent ecosystems in the Southeastern U.S.
While it is commonly assumed that stomatal closure prevents nocturnal transpiration in C3 plants, recent evidence from a number of studies suggests that significant transpiration occurs at night across a wide range of species. We explore the magnitude of nocturnal evaportranspiration (ET) fluxes measured by multiple years of eddy-covariance data from a successional gradient (grassland, planted pine forest, and hardwood forest) in the Southeastern U.S. all experiencing similar climatic and edaphic conditions. After removing unreliable data points collected during periods of low turbulence, nocturnal ET fluxes averaged 27.1, 19.1 and 5.5 percent of the magnitude of mean daytime ET fluxes in the grassland, pine forest, and hardwood forest, respectively. Because little attention has been paid to deriving appropriate methods to gapfill missing eddy covariance ET fluxes, we employed several gapfilling methodologies on these datasets, with a focus on assessing appropriate methods to gapfill nocturnal fluxes. These gapfilling procedues include multiple linear regression between nocturnal ET and meteorological variables (namely air temperature, vapor pressure deficit, and mean wind speed), relationships between nocturnal and daytime ET and conductance measurements, and a multiple imputation Monte Carlo technique. The utility of the gapfilling procedures is then assessed by comparing simulated fluxes to reliable measured fluxes using randomly generated gaps, and by assessing the difference between the annual sums of ET across these three sites as generated by the different gapfilling techniques. The seasonality of these nocturnal ET fluxes and their relationship to leaf area is also explored.