Evaluation of different non-invasive methods for root research in paved sites

The availability of reliable, non-invasive methods the detection of the root system of urban trees would increase the knowledge about root growth in paved sites, with beneficial consequences for tree preservation during redevelopment. Ground Penetrating Radar (GPR) and sonic tomography have been proposed as non-invasive technologies for root studies in paved sites This experiment has the aims of: 1) evaluate the capacity of GPR and sonic tomography to locate roots and predict root density of urban trees in paved sites; 2) evaluate the effects of pavement typology on the root system characteristics on urban trees. The experiment was carried out at Fondazione Minoprio, using 16 trees (Celtis australis and Fraxinus ornus) planted in March 2012 in either an unpaved soil or in soil covered by pavements. In 2020, the Tree-Radar-Unit GPR (900 MHz antenna) and the Arboradix tomograph were used to measure root attributes of the trees. Concentric scans and tomographies were obtained using a 20 cm pace from the root flare to a maximum radius of 2.60 m. Results of these non-invasive methods were validated against the root number and the root density data obtained by direct measurement and Lidar scanning of the root system. Direct measurements were performed after removing the pavements and pulling away the soil using AirspadeTM and a suction excavator. Arboradix could provide an reliable estimation of root number in Fraxinus (P<0.000; R2 = 0,654), but not in Celtis (P=0,391). The presence of pavements can affect the speed of the soundwave, thus the output of the measurement. GPR could provide an estimation of root density for both species (P<0,000), despite the wellness of fit was higher in Celtis (R2=0,790) than in Fraxinus (R2=0,500). Because of higher replicability, GPR and airspading were then used to evaluate the effects of impermeable, permeable, and porous pavements on root growth. Results showed that root density declined under every type of pavements, compared to the unpaved control, with the larger reduction observed with impermeable pavements. Average root length was slightly affected by pavements, but root slenderness decreased significantly in impermeable and permeable pavements, compared to porous ones and control. Also, the amount of fine roots per unit coarse root biomass decreased when plants were grown in impermeable pavements. Results of this research showed that GPR can be applicable for rough estimates of rooting area and root density of individual, largely-spaced trees, but inaccuracies can arise when multiple root system overlap. Sonic tomography can be applicable in trees with a strictly radial root distribution, but inaccuracies can arise when circling roots are present. Pavements affect root system morphology mainly as an effect of altered gas exchange between the soil and the atmosphere. This research has been funded by TREE FUND – Research Fellowship Award 2016-2021.