Data from: Testing the divergent adaptation of two congeneric tree species on a rainfall gradient using eco-physio-morphological traits

In tropical Africa, evidence of widely distributed genera transcending biomes or habitat boundaries has been reported. The evolutionary processes that allowed these lineages to disperse and adapt into new environments are far from being resolved. To better understand these processes, we propose an integrated approach, based on the eco-physio-morphological traits of two sister species with adjacent distributions along a rainfall gradient. We used wood anatomical traits, plant hydraulics (vulnerability to cavitation, wood volumetric water content and hydraulic capacitance) and growth data from the natural habitat, in a common garden, to compare species with known phylogeny, very similar morphologically, but occupying contrasting habitats: Erythrophleum ivorense (wet forest) and Erythrophleum suaveolens (moist forest and forest gallery). We identified some slight differences in wood anatomical traits between the two species associated with strong differences in hydraulics, growth, and overall species distribution. The moist forest species, E. suaveolens had narrower vessels and intervessel pits, and higher vessel cell-wall reinforcement than E. ivorense. These traits allow a high resistance to cavitation and a continuous internal water supply of the xylem during water shortage, allowing a higher fitness during drought periods, but limiting growth. Our results confirm a trade-off between drought tolerance and growth, controlled by subtle adaptations in wood traits, as a key mechanism leading to the niche partitioning between the two Erythrophleum species. The generality of this trade-off and its importance in the diversification of the African tree flora remains to be tested. Our integrated eco-physio-morpho approach could be the way forward. AE_WP_ErythrophleumCollected AE signals and water potential used to construct an acoustic vulnerability curves for E. ivorense ansd E. suaveolens. The AE signals were cumulated over the measurement period, and averaged over 5 minutes. The endpoint of the VCAE was based on the local maximum in the third derivative, following the strongest decrease in the first derivative. Cumulative acoustic emissions were then rescaled between zero and the defined endpoint to obtain a relative percentage of cavitation-related AE (%). The continuous water potential x-axis of the VCAE was obtained from a segmented linear regression relation between point measurements of xylem water potential, and xylem shrinkage monitored by dendrometers.wood_traits_ErythrophleumTangential vessel diameter, vessel density, vessel grouping index (ratio of total number of vessels to total number of vessel groupings), proportion of vessel lumen, fibre, axial and radial parenchyma measured for the two species, E. ivorense and E. suaveolens, in their natural habitat (branch and stem samples) and in a common garden experiement (only branch samples). On branch sections, these measurements were done on 0.16 mm² quadrats systematically distributed from the pith to the bark (from [0;400] to [2800;3200]), in order to include radial variation in wood traits. For stem sections, measurements were done on two quadrats of 9.40 mm². Intervessel wall thickness and fiber wall thickness (measured as the double wall between two adjacent fibers) were also measured at higher magnification. Conduit wall reinforcement was calculated by determining the wall thickness (t) to conduit wall span (b) ratio . On tangential sections, we measured the horizontal pit membrane diameter at its widest point but only for branch samples from natural habitat.growth_common_gardenGrowth in diameter in the common garden experiement for E. ivorense and E. suaveolens.growth_natural_habitatGrowth in diameter and diameter classes for the two species E. suaveolens and E. ivorense in their respective natural habitat.growthnatural_habitat.csv