Physical constraints and environmental factors shape phloem anatomical traits in woody angiosperm species

Abstract

Xylem trait studies have enhanced our understanding of how plants strategically adapt their morphological and anatomical features to diverse climates. Despite the importance of the phloem in plant functioning, similar studies of phloem traits are lacking. To tackle this knowledge gap, we analyzed phloem anatomical traits of woody angiosperm species in relation to climate and the distance of samples to the stem tip. We collected main stem or branch cross-sections of 188 angiosperm woody species, which represent a wide range of climates and diverse families. Measurements of xylem vessel and phloem sieve element diameter, density, and lumen fraction were used in phylogenetic structural equation models to disentangle internal and climatic constraints on their morphological and anatomical features. Our results showed that distance-to-tip mainly affects sieve element and vessel diameter and density, while climate more strongly influenced conduit lumen fraction. Vessel size was positively correlated with temperature after correcting for the distance-to-tip, while sieve element diameter was correlated with water availability. Our results highlight the need to account for distance-to-tip when accessing anatomical variations linked to the environment, and show that sieve element traits respond to other climatic drivers than vessel traits rather than simply mirroring them.

Keywords: adaptation; allometry; phloem sieve element; tip‐to‐base conduit widening; xylem vessel.

Fig. 1

Anatomical feature. (a) Sieve element lumen area (in green), density, and fraction measurement. (b) Xylem vessel lumen area (in yellow), density, and fraction, using Fagus sylvatica L. as an example. (c) Detrended value. The black solid line represents linear regression between log‐transformed trait and log‐transformed distance‐to‐tip. For a given species A (red dot in the figure), the detrended trait value is the residual (dashed line d), which means the distance between raw value and the predicted value.

Fig. 2

Structural equation model scenario. Arrows represent the influence of climate factors (PCs) or distance‐to‐tip on trait values. (a) A null model supposes that there is no influence of either climate (PCs) or the distance to the tip. (b) An indirect model supposes the influence of the climate is on distance‐to‐tip and then subsequently influences the trait value. (c) A direct model assumes that the climate has impacts on trait values; there is no correlation between the distance‐to‐tip and trait value. (d) A full model allows both direct and indirect influences.

Fig. 3

Scaling factors for (a) median diameter in the main stem sample, (b) median diameter in branch sample, (c) vessel and sieve element conduit density, and (d) vessel and sieve element conduit lumen fraction with distance‐to‐tip. The scaling factors b are shown in each graph. Solid line: slope b is significantly different from zero (P < 0.05). Dashed line: slope b is not significantly different from zero (P ≥ 0.05).

Fig. 4

Pearson product–moment correlation between sieve element (SE) and vessel anatomical traits in main stem samples. The left triangle shows correlation between traits' raw value; the right triangle shows correlation between traits' detrended value, corrected after distance‐to‐tip. The color and the shape reflect the value of Pearson correlation coefficients. Significant code: ***, P < 0.001; **, P < 0.001; *, P < 0.5; ns, not significant, P ≥ 0.05.

Fig. 5

The best‐fit phylogenetic structural equation model for sieve element (a) median diameter, (c) density, and (e) fraction; and vessel (b) median diameter, (d) density, and (f) fraction in main stem samples. Arrows represent the influence of climate factors (PCs) or distance‐to‐tip on trait values. Red arrows show positive correlation, and blue arrows show negative correlation. The thickness of the arrow indicates the weight. Solid lines indicate 95% confidence intervals (CIs) do not overlap zero, and dashed lines indicate 95% CIs overlap zero.