Trait-Mediated Competition for Light Underpins Plant Diversity Loss Under Eutrophication

Abstract

Eutrophication is a major driver of plant diversity loss, yet the underlying mechanisms remain poorly understood. In particular, the role of eutrophication-induced light limitation in regulating plant diversity in natural communities has rarely been examined directly. Here we show that experimental light addition to the understory of a natural alpine grassland consistently restored lost diversity under different nutrient enrichment regimes. Under nitrogen enrichment, light addition recovered diversity primarily by promoting species gains, whereas under phosphorus enrichment, it primarily reduced species losses. When both nitrogen and phosphorus were enriched, light addition simultaneously increased species gains and reduced losses. These effects were primarily driven by shifts in the colonization and extinction of species with resource-acquisitive strategies (i.e., those with high specific leaf area and low leaf dry matter content), emphasizing the critical role of trait-mediated competition for light in biodiversity loss. Our findings point to light competition as a key driver of eutrophication-induced plant diversity loss, suggesting that managing light availability could help mitigate these losses in natural ecosystems.

Keywords: functional trait; light competition; nutrient enrichment; plant economics spectrum; species diversity; species gain and loss.

FIGURE 1

Experimental setup. (a) Landscape view of the light competition experiment in a natural alpine grassland. (b) Photographs illustrating the experiment: The upper panel shows the layout of the experimental plots; the middle and lower panels depict light addition to the understory via LED strips at night and in the daytime, respectively. (c) Overview of the experimental design, highlighting treatment groups and expected treatment effects. Details on the light addition device are provided in Figure S1.

FIGURE 2

Effects of experimental treatments on plant species richness, gains and losses. The effects of light addition (L), N addition (N) and P addition (P) on plant species richness at the end of the 3‐year experiment (i.e., in 2023) (a) and the number of species gains and losses between 2023 and 2020 (b). Data are presented as mean + s.e.m. Different letters denote significant differences between treatment means (p < 0.05), conducted separately for species gain and loss.

FIGURE 3

Traits of species lost under eutrophication. Principal component analysis of plant traits from control plots, highlighting traits of species that were lost after N (a) and NP (b) addition. Species losses were primarily observed among resource‐acquisitive species. Each circle represents a species that was lost, with the circle size proportional to the number of extinctions observed. Axis labels show each principal component (PC1, 44.51% of total trait variation explained; PC2, 22.7% of total trait variation explained). Leaf graphics by T. Saxby and L. Heydon (lntegration and Application Network, University of Maryland Center for Environmental Science, http://ian.umces.edu/imagelibrary/).

FIGURE 4

Traits of species gained or rescued from loss after light addition. Principal component analysis of plant traits from control plots, highlighting traits of species gained only after light addition in N‐amended (a) and NP‐amended plots (b), as well as traits of species rescued from loss by light addition in P‐amended (c) and to the NP‐amended plots (d). Both gaines and rescues were primarily observed among resource‐acquisitive species. Each circle represents a species gained or rescued, with the circle size proportional to the number of extinctions observed. Axis labels show each principal component (PC1, 44.51% of total trait variation explained; PC2, 22.7% of total trait variation explained). Leaf graphics by T. Saxby and L. Heydon (lntegration and Application Network, University of Maryland Center for Environmental Science, http://ian.umces.edu/imagelibrary/).

FIGURE 5

The results of the random forest model for predictors of species richness. Light availability emerged as the best predictor of species richness, surpassing microclimate, soil nitrogen, soil pH, the number of added resources and soil metal concentration. Significant levels: *p < 0.05 and **p < 0.01.