Stem flow and porometer measurements of transpiration from honey mesquite (Prosopis glandulosa)

Abstract

The objective of this study was to compare stem flow and porometer methods of measuring transpiration of honey mesquite (Prosopis glandulosa) trees on a semiarid site. Stem flow was measured using heat balance stem flow gauges. Porometer measurements of leaf stomatal conductance (g_s) were made within foliage layers of each stem and scaled to transpiration values for the entire stem (E_stem) using stem leaf area. Simultaneous measurements using both methods were made diurnally and under artificially imposed stem shading or defoliation in June and October 1990. Stem flow and E_stem had similar diumal patterns except on 2 d in June when E_stem increased during the afternoon while stem flow declined relative to midday values. During October, E_stem was greater than stem flow throughout the day. This was attributed to sampling error in which only undamaged leaves were used for porometer measurements yet, by this time in the growing season, many leaves on each stem were damaged from insects or wind and likely had lower transpiration rates. A regression coefficient between E_stem and stem flow of 0.79 in June and 0.91 in October suggested the two methods were comparable, but there was considerable variation between methods during peak transpiration rates. Both techniques detected that artificial shading or defoliation caused similar relative declines in transpiration. Results imply that estimates of stem transpiration can be obtained by scaling porometer measurements of leaves but accuracy declines at higher transpiration rates.