3.1. Responses of aerial and underground parts of riparian plants to hydrogeomorphological in situ constraints
When individuals of P. nigra establish on alluvial bars in dynamic rivers, they can strongly affect fluvial geomorphology by trapping fine sediments on alluvial bars during floods, and thus build new biogeomorphological landforms (Gurnell, 2014; Hortobágyi et al., 2018a). The engineer effect of P. nigra on alluvial bars depends on its architecture, biomass, and exposure to mechanical stress. The species has a strong phenotypic plasticity that allows to adjust its physiological, morphological and biomechanical traits according to local hydrogeomorphological conditions. The quantification of the morphological and biomechanical variation of response traits withinP. nigra populations which depend on exposure to mechanical constraints related to flow is essential to better understand why, to what extent, and where the black poplar is able to modulate geomorphology and thus gain in its survival, growth, and reproduction.
In an empirical study conducted on the Allier River, Hortobágyi et al. (2017) quantified the variation in response traits of P. nigrawithin a population that was two years old, based on three levels of exposure to mechanical constraints. In highly exposed areas, at the head of an alluvial bar, saplings developed response traits that allow for greater resistance, including a reduced size (avoidance strategy; see Bornette and Puijalon, 2011; Puijalon et al., 2011), a flexible and inclined stem, and also a more robust root system (tolerance strategy; see Bornette and Puijalon, 2011; Puijalon et al., 2011). In a highly exposed context, avoidance morphological and biomechanical responses led to a decrease in drag during floods. These responses that increase the resistance of young individuals nevertheless limit the engineer effect, i.e., their ability to trap fine sediment and nutrients in relation with increased surface biomass and roughness. This limitation of the engineer effect of black poplar in the most exposed areas, at the head of the bar, is therefore translated by a limitation of sediment trapping and niche construction at the very early stage of establishment.
This observation suggests the existence of a functional compromise in the expression of traits between resistance to floods (i.e., improvement of survival) and niche construction (i.e., improvement of habitat conditions under biotic control and therefore improvement of resource acquisition and growth). In less exposed areas of alluvial bars – in the middle part and downstream of the bars as well as within near secondary channels – young poplars developed a wider, longer, less flexible, and less inclined stem, as well as larger leaves. This expresses a prioritization of resource acquisition and competition abilities and at the same time the possibility for niche construction during ordinary floods. The total length (underground part + aerial part) of the plants is maximal at the tail of the bar with a notable effect on the trapping of fine sediments.
Complementary observations of individual roots system on the same alluvial bars showed that plant burial by fine sediment systematically stimulates the production of adventitious roots and aerial biomass, which in turn increases sediment trapping (Ding, 2014). This was interpreted as a positive biogeomorphological feedback of sedimentary accretion and plant growth. The construction of fluvial landforms by black poplar cohorts seems to correspond to a positive niche construction strategy leading to the progressive reduction of mechanical constraints and to a modification of the habitat that positively influences the survival, growth, and ultimately the sexual reproduction of poplars (Corenblit et al., 2016b, 2018, 2020a,b). Biogeomorphological processes occurring at a small scale caused by different morphological and biomechanical plant traits therefore may have an influence on the processes of fluvial landform construction at a larger spatial scale (Merritt, 2013; Corenblit et al., 2015; O’Hare et al., 2016; Diehl et al., 2017).