Maintaining high rates of water loss during times of high resource availability could allow establishing woody desert perennials to grow quickly by allowing them to take advantage of the fleeting but abundant monsoonal moisture typical of warm deserts like the Chihuahuan. However, a plant cannot endlessly increase water loss in order to grow faster --there are hydraulic constraints on rates of water loss. The hydraulic properties of each particular plant xylem and soil microsite, as well as the AR:AL absorbing root area to transpiring leaf area ratio) interact to set limits on rates of water loss. If transpiration rates become too high, cavitation may limit the ability of the xylem to supply water to the leaves. The main objective of this study was to test two hypotheses on a population of Larrea tridentata at the Sevilleta LTER in central New Mexico (1) do small plants grow faster and use water less conservatively than large, and (2) are there differences in the hydraulic constraints on small and large plants. Measurements were made every six weeks in the spring, summer and fall from April 2002 - August 2003. Field measurements of shoot growth, gas exchange and plant and soil water potentials were made to determine growth rates and water use. Measurements of leaf specific conductance determined the ability of the xylem to supply water to the leaves. Excavation findings were used to estimate (AR:AL). Xylem vulnerability curves and soil texture analysis were used to determine the hydraulic properties of the plant xylem and soil. A model determined where the limiting conductance occurred in the plant-soil continuum.
For gas exchange measurements a LiCor 6400 portable gas exchange system was used. Measurements took place in June, August and September 2002 and May, June, and August 2003. Two measurements were made on each plant at approximately 7-9:30AM and 10-12:30PM. One branch tip was chosen and marked on each 10 small and 10 large plants. The same branch tip was used for measurement throughout the day unless it broke, at which time another branch tip was chosen and marked. Stomatal Ratio was set to one because stomates are present on both sides of the leaf in this species. Because of the small size of the leaves, an energy balance approach was used to calculate the leaf temperature in the chamber.
Chamber temperature and humidity were controlled at ambient and reference CO2 was set to 400ppm. Using natural light plants were clamped into the chamber, oriented in their original direction, chamber conditions were allowed to stabilize. Leaf area was set to one during measurement. Because of the small leaves of the species, a branch tip had to be measured. Measured branch tips were cut and returned to the lab where their leaf area was measured.
A Vista Scan Scanner was used to create an image of the leaves. The bitmap image was then analyzed for number of pixels using Scion Image. A regression equation was developed which converted pixel number into leaf area in cm2.The gas exchange data was then recalculated to adjust for leaf area.
For plant water potential a Scholander Pressure bomb was used to measure branched stem tips consisting of 15-20 leaves and a woody base. Predawn water potential samples were collected between 4AM and 5AM. Midday water potentials were collected between 11AM and 1PM. Samples were placed into a plastic baggie with a moist paper towel during transport. Samples were collected in May, June, August, September and December 2002 as well as January, March, May, June and August 2003.
For whole plant hydraulic conductance measurements large sections of the xylem were measured using a vacuum canister to generate known vacuum pressures. The plant was attached to a water filled container on a balance via tiagon tubing. Changes in weight on the balance, and thus flow rate (mg/s) through the plant were measured by a computer using a program written in Turbo Pascal. Each sample was measured at four or five pressures and the change in flow rate with pressure was calculated as the total hydraulic conductance of all tissues contained in the sample. All samples were immediately placed into a plastic bag with a wet paper towel and transported to the lab where they were measured within 48 hours of collection.
For large plants an entire stem of the plant was cut at the base. The stems were cut under water at the lab, to about 30cm. For small plants the entire plant was excavated, and any roots larger than about 2mm in diameter were kept intact. Back at the lab, most of the root system was cut off under water, leaving the root collar and the initial un-branched portion of the main root which obviously supplied the entire plant. For both sizes, all green material was removed from the tips of the branches, leaving only woody stems.
The green portion of the plant included all leaves and sometimes large (up to 150mm) branched sections. For root hydraulic conductance measurements root segments +AD4-50cm were cut in the field and transported to the lab wrapped tightly in 3 plastic bags containing wet paper towel. Segments were re-cut underwater and the ends shaved off with a razor blade. They were then placed on a manifold and flow through the segment was measured. Segments were then flushed for 15 minutes with distilled water at 100kPa. Flow was then re-measured. Percent loss of conductivity is calculated as the difference between pre and post flush flow divided by post flush flow and multiplied by 100. Due to time constraints only one flush was performed on each sample. For soil water potential monitoring soil thermocouple psychrometers were placed under 4 large and 4 small plants at 30cm and 45cm below the soil surface. Measurements were made at or around 2AM when temperature gradients were at a minimum. A Campbell datalogger reported millivolt output which was converted to MPa using calibrations determined in the lab. Calibration involved regression of millivolt output against solutions of known salt concentration for each psychrometer.
Additional Information on the personnel associated with the Data Collection / Data Processing Joy Francis, a post-doc with Jim Gosz, was instrumental in setting up this study.