The roles of species' relatedness and climate of origin in determining optical leaf traits over a large set of taxa growing at high elevation and high latitude

Abstract

Climate change is driving many mountain plant species to higher elevations and northern plant species to higher latitudes. However, various biotic or abiotic constraints may restrict any range shift, and one relevant factor for migration to higher elevations could be species' ability to tolerate high UV-doses. Flavonoids are engaged in photoprotection, but also serve multiple ecological roles. We compared plant optical leaf trait responses of a large set of taxa growing in two botanical gardens (French Alps and southern Finland), considering potential constraints imposed by the relatedness of taxa and the legacy of climatic conditions at plants' original collection sites. The segregation of optically measured leaf traits along the phylogeny was studied using a published mega-tree GBOTB.extended.tre for vascular plants as a backbone. For a subset of taxa, we investigated the relationship between climatic conditions (namely solar radiation, temperature and precipitation at a coarse scale) at the plants' original collection site and current trait values. Upon testing the phylogenetic signal (Pagel's λ), we found a significant difference but intermediate lambda values overall for flavonol or flavone index (I_flav) and anthocyanin index (I_ant), indicating that phylogenetic relatedness alone failed to explain the changes in trait values under a Brownian motion model of trait evolution. The local analysis (local indicator of phylogenetic association) indicated mostly positive autocorrelations for I_flav i.e. similarities in optically measured leaf traits, often among species from the same genus. We found significant relationships between climatic variables and leaf chlorophyll index (I_chl), but not I_flav, particularly for annual solar radiation. Changes in plants' I_flav across microhabitats differing in UV irradiance and predominately high F _v /F _m indicated that most plants studied had sufficient flexibility in photoprotection, conferred by I_flav, to acclimate to contemporary UV irradiances in their environment. While not explaining the mechanisms behind observed trait values, our findings do suggest that some high-elevation taxa display similar leaf flavonoid accumulation responses. These may be phylogenetically constrained and hence moderate plants' capacity to adjust to new combinations of environmental conditions resulting from climate change.

Keywords: UV radiation; botanical gardens; flavonoids; leaf traits; mountain species; photoprotection.

Figure 1

Relative differences in leaf traits from 672 taxa in the alpine botanical garden (Col du Lautaret, France) shown in a hexagonally arranged Kohonen self-organising map (SOM) with 25 nodes. Plants were sampled for optically measured leaf traits; indices of adaxial flavonol/flavone (I_flav, dark blue, above right wedge) and anthocyanin (I_ant, turquoise, above left wedge), and leaf chlorophyll (I_chl, blue, above middle wedge) during the summer of 2014. Panel (A) shows the relative differences (radius of each wedge) in three scaled mean leaf traits in the upper half circle. The lower half circle contains relative differences in categorised light condition of the plants, categorised phenological advancement, and categorised leaf inclination angle. Panel (B) shows loadings of the nodes i.e., number of taxa grouped in the node as points, and neighbour distance (sum of distances to all immediate neighbours) as colour scale with red for the greatest distance. These analyses allowed for distinction of different developmental stages of those species where such an approach was deemed necessary. Range in mean optically measured leaf traits among taxa are given lower right.

Figure 2

Used phylogeny for the studied 622 taxa from the alpine botanical garden (Col du Lautaret, French Alps) based on previously published mega-tree (GBOTB.extended.tre) and methodology (scenario 1) from Jin and Qian (2019), is plotted with mean leaf trait values (I_flav =A & B, I_chl = C and I_ant = D) as colour scale by using R package phytools (Revell, 2012 and Revell, 2013). Colour bar length is a scale for branch lengths (million years, Myr). The plotting method estimates ancestral trait values but estimate of uncertainty is not shown, and ancestral values are mainly shown for visual clarity. The (B) panel shows mean I_flav of taxa with no plants from shaded microhabitats (categorised light condition maximum of minor shade over day) with differing scale of values. Infraspecific taxa (i.e. subspecies and variety) were included in the phylogeny by combining them with their parental species. High-quality pdf -versions with tip labels may be found at the end of the Supplementary A1 .