Young trees share soil water with wheat in an alley-cropping system in a wet crop year: Evidence from 2H2O artificial labeling.

• Deuterium labeling was used to monitor tree and wheat water uptakes in agroforestry. • Young trees and wheat took up water mainly from the upper soil layer. • Wheat in alley-cropping took up more water at 30 cm depth than in sole-crop plot. • Intercropped wild cherry and alder may exploit deep soil water during dry events. Despite numerous studies investigating competition and/or facilitation for soil water resources in alley-cropping systems (AC), the share of water at the early stage of tree establishment in AC has been poorly examined. This work aimed to explore the water share between crops and trees after four years of tree establishment in AC using deuterium labeling tracer. The investigations were carried out at the Ramecourt AC experimental site, in northern France, according to a randomized block design with three replications. In mid-spring, we injected 300 mL of 10 % deuterated water at 30, 50 and 100 cm soil depths at a 1.5 m distance from a reference tree (Alnus glutinosa or alder, Carpinus betulus or hornbeam, Prunus avium or wild cherry) in an AC, in a pure-forest control plot with ryegrass (FC) and in a randomly chosen area in monocrop wheat control (CC) plot. The tracer uptake was monitored by collecting tree leaves and wheat and ryegrass (RG) whole-plant samples every two weeks in labeled and unlabeled areas. For deuterium natural abundance analyses, the global mean δ2H was significantly lower for wheat (-44 ± 4 ‰) than for RG (-27 ± 6 ‰) and trees (-20 ± 3 ‰). The plant species, the block and the plant water content (PWC) accounted for 38 % of the δ2H variability (p < 0.01). This result indicated that the most active sites of plant water uptake were different between species. The mean wheat δ2H was 2481 ± 523 ‰, 715 ± 218 ‰, and 133 ± 68 ‰ at the 30, 50 and 100 cm labeling depths, respectively. The value was significantly higher in the AC (2883 ± 585 ‰) than in the CC (1131 ± 274 ‰) only at the 30 cm labeling depth. For trees, the δ2H of labeled samples remained negative, unlike wheat. Particularly in the AC, alder and wild cherry presented significantly a higher δ2H at 8 and 28 days after labeling, respectively, from 50 and 100 cm labeling depths, compared to the 2H natural abundance. We concluded that trees and wheat take up their water in the upper soil layer, but in the AC, trees favor wheat water uptake in the topsoil and can flexibly shift their water source to the deep layer in the case of low water availability in the upper soil layer. [ABSTRACT FROM AUTHOR]