Laserfiche WebLink
58 R.M. Gazal et a1.lAgricultural and Forest Meteorology 137 (2006) 56-67 <br />Table 1 <br />Structural characteristics of cottonwood clusters at Lewis Spring (n = 9) and Boquillas (n = 10) sites along the San Pedro River in southeastern <br />Arizona <br />Site <br />Diameter (cm) <br />Height (m) <br />Canopy area' (m2) <br />SWAb (cm2) <br />Leaf Area/SWA*" (m2 cm-2) <br />Boquillas <br />Lewis Springs <br />61 <br />49 <br />27 <br />25 <br />1037 <br />421 <br />1223 f 231 <br />797 f 243 <br />0.21 f 0.02 <br />0.31 t 0.04 <br />a Canopy area refers to the planar area of the canopy as seen from aerial photographs (Schaeffer et al., 2000). <br />b Sapwood area. <br />`** AL:As (±S.E.) significant at P= 0.10. <br />most of the San Pedro generally grow within the active <br />floodplain and upon highly permeable alluvium, <br />consisting primarily of sand and gravel (Pool and <br />Coes, 1999) <br />2.2. Basal sap flow measurements <br />Sap flow of four cottonwood trees, a diffused porous <br />species, within each cluster was measured using <br />constant heat flow Granier -type probes (TDP -30 and <br />TDP -80, Dynamax, Inc., Houston, TX). Thermocouple <br />needles 1.2 and 1.65 mm in diameter (TDP -30 and <br />TDP -80, respectively) were installed as a vertically <br />aligned pair 4 cm apart within the sapwood of each tree. <br />The system measures the temperature difference <br />between two probes inserted radially in the xylem, <br />one constantly heated and the other unheated. Sets of <br />these probes were implanted on the north and south side <br />of each tree at 1.75 m above the ground. <br />Sap flow was measured continuously from April 5 <br />to November 9, 2003 using a datalogger (CR10x <br />datalogger, Campbell Scientific Inc., Logan, UT). <br />Plastic putty was installed around the needles for <br />water protection and foam quarter- spheres were <br />tightly secured on both sides of the needles to protect <br />the wire from bending stress and to provide thermal <br />insulation to the needles. Reflective bubble wrap was <br />installed around the tree bole for additional insula- <br />tion. <br />The Granier (1987) method was used to calculate the <br />dimensionless parameter (K) as: <br />K _ aTn, - aT (1) <br />aT <br />where aT is the measured difference in temperature <br />between the heated needle, referenced to the lower non - <br />heated needle and aTm is the value of aTwhen there is no <br />or minimum sap flow. Average sap flow velocity [V, <br />cm s -1] was calculated as ( Granier, 1987): <br />V = 0.0119K1231 (2) <br />Sap flow velocity was then converted to sap flow rate <br />[J5, cm h -1 or g cm -2 h-11 from: <br />J, = SWA x V x 3600) (3) <br />where SWA is the sapwood area (cm 2) and 3600 is in <br />s h-'. <br />2.3. Scaling <br />Cottonwood stand transpiration, E (mm day-'), was <br />calculated based on individual tree sap flow, total <br />sapwood area and crown area of the cluster (Wulls- <br />chleger et al., 1998). At the end of the season, sapwood <br />area (SWA) was determined from increment cores taken <br />as close as possible to the probe insertions at each side <br />of the tree. Sapwood was identified from heartwood by <br />color change from lightly colored to darkly colored and <br />water saturated heartwood. Total canopy area and <br />diameter of all the trees in the cluster were measured at <br />both sites (Table 1). The sapwood area -to- diameter <br />relationship was used to estimate the total sapwood area <br />of all the trees in the cluster (Schaeffer et al., 2000). <br />For trees instrumented with TDP -30, sap flow was <br />assumed to be constant over the radial sapwood profile <br />since radial profile information was lacking for our study <br />site. However, using the TDP -80, sap flow rate per tree <br />(Js) was scaled based on the sapwood area that covers the <br />position of the two thermocouples per probe. Scaling of <br />sap flow from those trees instrumented with probes to the <br />whole patch of trees involves several sources of <br />uncertainty such as estimates of sapwood area and <br />within -tree and between -tree variation in sap flow. <br />J, from the north and south side of each thermo- <br />couple position was averaged and multiplied by the <br />sapwood area corresponding to the depth of the <br />thermocouple in the sapwood (Js). JS from the two <br />thermocouple positions was then added and the sum <br />was divided by the total sapwood area of the tree (Js). J, <br />from the instrumented trees in the cluster was averaged <br />and multiplied by the total sapwood area of the entire <br />cluster to get the total water use of the stand (kg d -'). <br />