Thermal dissipation probe (TDP) method (Granier, 1985) is widely used to estimate tree transpiration (i.e., the water evaporated from the leaves) because it is simple to build, easy to install, and relatively inexpensive. However, the universality of the original calibration has been questioned and, in many cases, proved to be inaccurate. Thus, when the TDP is used in a new species, specific tests should be carried out. Our aim was to propose a new method for improving the accuracy of TDP on trees in the field. Small hazelnut trees (diameter at breast height 5 cm) were used for the experiment. The response of TDP sensors was compared with a reference water uptake measured with an electronic potometer system provided with a high precision liquid flow meter. We equipped three stems where we measured the sap flow density, the sapwood area (by using fuchsine), the total tree water uptake (reference), and the main meteorological parameters during summer 2018. Results confirmed that the original Granier’s calibration underestimated the effective tree transpiration (relative error about −60%). We proposed a new equation for improving the measurement accuracy within an error of about 4%. The system proposed appeared an easier solution compared to potted trees and particularly suitable for orchards, thus contributing to improve the irrigation management worldwide.

Climate change and the global economy impose new challenges in the management of food-producing trees and require studying how to model plant physiological responses, namely growth dynamics and phenology. Hazelnut (Corylus avellana L.) is a multi-stemmed forest species domesticated for nut production and now widely spread across different continents. However, information on stem growth and its synchronization with leaf and reproductive phenology is extremely limited. This study aimed at (i) defining the sequencing of radial growth phases in hazelnut (onset, maximum growth and cessation) and the specific temperature triggering stem growth; and (ii) combining the stem growth phases with leaf and fruit phenology. Point dendrometers were installed on 20 hazelnut trees across eight orchards distributed in the Northern and Southern hemisphere during a period of three growing seasons between 2015 and 2018. The radial growth variations and climatic parameters were averaged and recorded every 15 min. Leaf and reproductive phenology were collected weekly at each site. Results showed that stem radial growth started from day of year 84 to 134 in relation to site and year but within a relatively narrow range of temperature (from 13 to 16.5 °C). However, we observed a temperature-related acclimation in the cultivar Tonda di Giffoni. Maximum growth always occurred well before the summer solstice (on average 35 days) and before the maximum annual air temperatures. Xylogenesis developed rapidly since the time interval between onset and maximum growth rate was about 3 weeks. Importantly, the species showed an evident delay of stem growth onset with respect to leaf emergence (on average 4–6 weeks) rarely observed in tree species. These findings represent the first global analysis of radial growth dynamics in hazelnut, which is an essential step for developing models on orchard functioning and management on different continents.