Opposing shifts in distributions of chlorophyll concentration and composition in grassland under warming

The CV represents the discrete degree of trait values, that is, the size of the trait space (Fig. 6b; Supplementary Note S1); S and K are generally used to describe the shape of trait distribution (Fig. 6c,d, Supplementary Note S1). Environment filtering can force a trait to deviate from the original distribution, with characteristics of smaller CV and larger S and K values^16,17. Partly consistent with our hypothesis, MAT significantly exerted positive effects on the total concentration of CV, S, and K, but weaker negative effects on the three values of Chl a/b (Fig. 6a). That is, the distributions of Chl concentration and composition shifted in opposite directions under global warming: Chl concentration was distributed in a broader but more differential way (Fig. 7a), while Chl a/b was distributed in a narrower but more uniform way (Fig. 7b).

The trait distributional shift under warming is possibly caused by the relative role of species turnover and intraspecific variation (due to plasticity and/or heritable differentiation)²⁵. For Chl concentration and composition, very weak phylogenetic signals were found in three plateaus (Supplementary Table S2), indicating the phenotypic plasticity of Chl, which environments have influenced during the long-term evolution. However, plasticity and intraspecific variation are not the focus of the discussion. Because the species compositions were significantly different among the three plateaus: with only a few species overlapping (Supplementary Fig. S3), and the dominant species and co-existing species gradually varied along the 30 sites (Supplementary Table S3). Shifts in Chl distributions under warming may be interpreted mainly by the alternation of species composition.

For Chl concentration, a broader trait space (higher CV) and a more skewed distribution (higher S and K) under warming conditions indicate several new species that differ in functions (here refer to rare species with higher Chl concentration) appeared or increased. This contributed to the long tail of the curve and raised the average Chl concentration. At the same time, most of the other species converged at lower Chl concentrations; that is, Chl concentration undergoes more substantial differentiation and functional contrasting species co-exist under warming. The concentration of Chl is representative of plant growth rate and production ability. Its distribution shift may imply a possible trend of polarisation in functions: both acquisitive and conservative species occur simultaneously. This alteration in species composition indicates changing biotic interactions²⁶. The co-existence of functional contrasting species allows individuals to avoid competition and enhance the exploitation of resources and niche^27,28, which is of great importance in optimising community functions^28,29. In desert and alpine regions, functional contrasting species with large inter-specific trait variations improve community multi-functionality and enable better resistance to climate change^17,30.

However, despite the shift in species composition, the distribution of Chl a/b only changed slightly compared to the Chl concentration under warming. The ratio of Chl a to Chl b represents the plant allocation to RC and LHC in PS and the efficiency trade-offs between light capture and light conversion^6,7. This ratio is characteristic of conservatism which is mainly manifested in the following aspects: (1) Chl a/b is independent of Chl concentration (orthogonal relationship of the two; Supplementary Fig. S2); (2) Chl a/b distributed more converged with higher K and lower CV (Supplementary Table S1); (3) relative fixed allometric relationships were found between Chl a and Chl b (beside TP; Fig. 8). Plants may adjust their RC and LHC allocation to a common ratio of 3:1 despite large variations in light availability or Chl concentration, which has also been confirmed by a study from forests¹⁴. Considering that RC is costlier than LHC, plants tend to sustain the Chl a/b as low as possible unless there is a functional imbalance caused by environmental stress such as warming^9,31.

Consequently, the shift in species composition may not be reflected by the conservative distribution of Chl composition. However, there is still a significant opposing trend for Chl a/b, indicating a highly uniform rather than a differential distribution of Chl composition under warming. The effect size is relatively weak, indicating that the shift in Chl composition may be slower or slighter than the Chl concentration under the same warming conditions and may not be suitable for prediction.

The effect of warming on community composition is a persistent concern, which will have severe consequences on community and ecosystem functions and processes. The significant reported effects of warming on grassland community structure include alterations in species coverage and important value³², reduction of dominant species and species richness³³, conversion among functional groups³⁴, and distribution shifts of phylogeny and function from divergence to randomness³⁵. The conclusions vary depending on the regions, biomes, and succession stages of the community. For example, evidence from the tundra showed that warming typically supports many vascular species, shrubs, and fewer cryptogams³⁶. A study in the northern Mongolian steppe found that locally restricted species are more sensitive to warming than more widespread species, based on their narrower niche breadth³⁷. Fertile or early successional grasslands, composed of fast-growing and short-lived species, respond rapidly to climate warming³⁸. Globally, warming has resulted in the migration of alpine vegetation toward higher latitudes and elevations, leading to the prevalence of thermophilic species^2,39,40.

It is difficult to obtain a unified explanation for how the species respond to warming because species respond differently depending on their strategies, such as resource allocation, which will be influenced by warming through limitations on other functions or traits. For example, nitrogen (N) is a key component of Chl, photosynthetic enzymes, and other functional proteins. How the species allocate its leaf N to the photosynthetic apparatus under warming will greatly impact its growth, interaction with other species, and community structure. We examined the relationship between leaf N and Chl concentrations in the three plateaus. We found that a significantly higher proportion was allocated to Chl for the species in LP than in TP when leaf N increased by the same amount (Supplementary Figure S4). In TP, low temperatures may limit the metabolic and growth rates of species, and strong direct radiation may cause photodamage. In this case, the most important thing for a species is to survive rather than produce. The strategies adopted by species to enhance their persistence and tolerance are associated with increased resistance and protective proteins⁴¹, leading to competition for leaf N. In warmer places, the synthesis of leaf N will not be restricted, and species tend to allocate more leaf N to Chl to support production. Consequently, Chl concentration shifts under warming are possibly caused by changes in other traits. This can be checked through trait-trait relationships, indicating species-specific trade-offs between different functions.

Space-for-time substitutions use extant spatial patterns between biota and environmental factors to reveal the warm impacts on plants. However, this method was thought to overestimate the effect of warming because there are unavoidable covariations in geographical locations, soil resources, and climate factors, making it difficult to separate the warming effect independently^36,42. In particular, species sensitivity to warming can be influenced by these covariation factors^36,37,43. However, this is still a preferred method for predicting the long-term response of plants under warming because there are several ways to avoid the shortcomings of this method or modify the effect magnitude as much as possible^42,44.

First, the study sites could be carefully selected with replications to avoid or minimise the coupling between temperature and potential confounding factors⁴². In this study, 30 sites were distributed mainly along the MAT gradient, which was almost independent of precipitation, soil resources, and aridity index (the orthogonal relationships showed between MAT and a suite of soil and climate factors; Supplementary Figure S5). This situation may be explained by the fact that the coupling of environmental factors was loosened at high elevation, so increasing temperature may not induce the additional effects of other environmental factors⁴⁵. In addition, the 10 replicated sites within each plateau were distributed along the precipitation and soil nutrient gradients, with the availability of soil resources gradually decreasing from east to west (Supplementary Figure S5). Thus, additional effects of covariant factors may be offset in the entire 30 sites of grassland, and thus differences in geographical location are thought to translate into a temperature gradient. Although a weak negative correlation was observed between MAT and PAR, it was due to elevation change. However, despite the variability in elevation-climate relationships, it may be that at lower elevations, conditions tend to favour acquisitive species that can take advantage of high resource levels since higher temperatures stimulate microbial activity and increase resource availability and vice versa⁴⁴. Consequently, species response and composition directions, shifting the Chl due to warming, should be reflected in this spatial transect.

Second, manipulative experiments or historical monitoring should be combined with space-for-time substitutions to draw a comprehensive conclusion⁴⁶. We summarised recent artificial warming experiments conducted in these three plateaus to understand the warming effects on species composition and possible interactions with other factors. For example, an increase in temperature changed the important values and coverage of different species and functional groups in Inner Mongolia and Tibet grasslands. Still, a concurrent decrease in soil moisture hardly affected the community composition because the plants can morphologically spread their roots deeper to maintain water use^47,48. Variations in precipitation induced by only warming influenced species phenology rather than composition^49,50. However, some argued that the combined effects of increased temperature and precipitation exhibited strong spatial heterogeneity⁵¹, with more significant sensitivity of vegetation at higher elevations⁵². Concomitant N deposition may exert the same effect as warming on community structure but through different mechanisms, which affect asynchronous population dynamics³³ or migration and availability of soil elements⁵³. In addition, from a 6-year experiment, nitrogen addition did not affect the abundance of major species and functional groups while exerting a more direct effect on ecosystem functioning⁵⁴. Although some conclusions support weak coupling between the roles of temperature and that of other factors, we should be aware of the defects of a large spatial span. In the plateau regions, the interaction between temperature and radiation should be considered. Informative results from manipulative experiments should be integrated into space-for-time substitutions to modify the effect size and consider the response time lag to warming in the near future.

In conclusion, warming may cause the Chl concentration to shift toward a broader but more differential distribution (functional heterogeneity), while Chla/b shifted toward a narrower but more uniform distribution (functional homogeneity). The shifts in the Chl distribution curve may be caused by phenotypic plasticity, alternation in species composition, and leaf N allocation patterns of species. The Chl composition is more conserved and insensitive to environmental change, which may not be suitable for use as a tool to predict grassland dynamics under climate change. At the regional level, Chl concentration and composition depend on local climate and limiting factors. Space-for-time substitutions are an informative method for estimating the long-term response of plants under warming. Still, the study sites should be carefully selected to minimise the additional effects of covariation factors, and conclusions should be drawn at the base of artificial warming experiments and historical monitors.

Source: Ecology - nature.com