Sap flow is Underestimated by Thermal Dissipation Sensors due to Alterations of Wood Anatomy

Abstract:

The thermal dissipation technique (TD) is one of the most commonly adopted methods for sap flow measurements. However, underestimations of up to 60% of the tree transpiration have been reported with this technique, although the causes are not certainly known. The insertion of TD sensors within the stems causes damage of the wood tissue and subsequent healing reactions, changing wood anatomy and likely the sap flow path. However, the anatomical changes in response to the insertion of sap flow sensors and the effects on the measured flow have not been assessed yet.

In this study, we investigate the alteration of vessel anatomy on wounds formed around TD sensors. Our main objectives were to elucidate the anatomical causes of sap flow underestimation for ring-porous and diffuse-porous species, and relate these changes to sap flow underestimations. Successive sets of TD probes were installed in early, mid and end of the growing season in Fagus sylvatica (diffuse-porous) and Quercus petraea (ring-porous) trees. They were logged after the growing season and additional sets of sensors were installed in the logged stems with presumably no healing reaction. The wood tissue surrounding each sensor was then excised and analysed by X-ray computed microtomography (X-ray micro CT). This technique allowed the quantification of vessel anatomical characteristics and the reconstruction of the 3-D internal microstructure of the xylem vessels so that extension and shape of the altered area could be determined.

Gels and tyloses clogged the conductive vessels around the sensors in both beech and oak. The extension of the affected area was larger for beech although these anatomical changes led to similar sap flow underestimations in both species. The higher vessel size in oak may explain this result and, therefore, larger sap flow underestimation per area of affected conductive tissue. The wound healing reaction likely occurred within the first weeks after sensor installation, which impeded the observation of differences in the extension or density of tyloses among 5, 11 and 22 weeks old wounds. This was also consistent with sap flow measured with the corresponding sensors. These results will help to improve the accuracy of sap flow measurements with thermal dissipation TD probes.