Are litter decomposition and fire linked through plant species traits?

Contents 653 I. 654 II. 657 III. 659 IV. 661 V. 662 VI. 663 VII. 665 665 References 665 SUMMARY: Biological decomposition and wildfire are connected carbon release pathways for dead plant material: slower litter decomposition leads to fuel accumulation. Are decomposition and surface fires also connected through plant community composition, via the species' traits? Our central concept involves two axes of trait variation related to decomposition and fire. The 'plant economics spectrum' (PES) links biochemistry traits to the litter decomposability of different fine organs. The 'size and shape spectrum' (SSS) includes litter particle size and shape and their consequent effect on fuel bed structure, ventilation and flammability. Our literature synthesis revealed that PES-driven decomposability is largely decoupled from predominantly SSS-driven surface litter flammability across species; this finding needs empirical testing in various environmental settings. Under certain conditions, carbon release will be dominated by decomposition, while under other conditions litter fuel will accumulate and fire may dominate carbon release. Ecosystem-level feedbacks between decomposition and fire, for example via litter amounts, litter decomposition stage, community-level biotic interactions and altered environment, will influence the trait-driven effects on decomposition and fire. Yet, our conceptual framework, explicitly comparing the effects of two plant trait spectra on litter decomposition vs fire, provides a promising new research direction for better understanding and predicting Earth surface carbon dynamics.

Variation in trait trade-offs allows differentiation among predefined plant functional types: implications for predictive ecology.

Plant functional types (PFTs) aggregate the variety of plant species into a small number of functionally different classes. We examined to what extent plant traits, which reflect species' functional adaptations, can capture functional differences between predefined PFTs and which traits optimally describe these differences. We applied Gaussian kernel density estimation to determine probability density functions for individual PFTs in an n-dimensional trait space and compared predicted PFTs with observed PFTs. All possible combinations of 1-6 traits from a database with 18 different traits (total of 18 287 species) were tested. A variety of trait sets had approximately similar performance, and 4-5 traits were sufficient to classify up to 85% of the species into PFTs correctly, whereas this was 80% for a bioclimatically defined tree PFT classification. Well-performing trait sets included combinations of correlated traits that are considered functionally redundant within a single plant strategy. This analysis quantitatively demonstrates how structural differences between PFTs are reflected in functional differences described by particular traits. Differentiation between PFTs is possible despite large overlap in plant strategies and traits, showing that PFTs are differently positioned in multidimensional trait space. This study therefore provides the foundation for important applications for predictive ecology.

Inclusion of ecologically based trait variation in plant functional types reduces the projected land carbon sink in an earth system model.

Earth system models demonstrate large uncertainty in projected changes in terrestrial carbon budgets. The lack of inclusion of adaptive responses of vegetation communities to the environment has been suggested to hamper the ability of modeled vegetation to adequately respond to environmental change. In this study, variation in functional responses of vegetation has been added to an earth system model (ESM) based on ecological principles. The restriction of viable mean trait values of vegetation communities by the environment, called 'habitat filtering', is an important ecological assembly rule and allows for determination of global scale trait-environment relationships. These relationships were applied to model trait variation for different plant functional types (PFTs). For three leaf traits (specific leaf area, maximum carboxylation rate at 25 °C, and maximum electron transport rate at 25 °C), relationships with multiple environmental drivers, such as precipitation, temperature, radiation, and CO2 , were determined for the PFTs within the Max Planck Institute ESM. With these relationships, spatiotemporal variation in these formerly fixed traits in PFTs was modeled in global change projections (IPCC RCP8.5 scenario). Inclusion of this environment-driven trait variation resulted in a strong reduction of the global carbon sink by at least 33% (2.1 Pg C yr(-1) ) from the 2nd quarter of the 21st century onward compared to the default model with fixed traits. In addition, the mid- and high latitudes became a stronger carbon sink and the tropics a stronger carbon source, caused by trait-induced differences in productivity and relative respirational costs. These results point toward a reduction of the global carbon sink when including a more realistic representation of functional vegetation responses, implying more carbon will stay airborne, which could fuel further climate change.

A fully traits-based approach to modeling global vegetation distribution.

Dynamic Global Vegetation Models (DGVMs) are indispensable for our understanding of climate change impacts. The application of traits in DGVMs is increasingly refined. However, a comprehensive analysis of the direct impacts of trait variation on global vegetation distribution does not yet exist. Here, we present such analysis as proof of principle. We run regressions of trait observations for leaf mass per area, stem-specific density, and seed mass from a global database against multiple environmental drivers, making use of findings of global trait convergence. This analysis explained up to 52% of the global variation of traits. Global trait maps, generated by coupling the regression equations to gridded soil and climate maps, showed up to orders of magnitude variation in trait values. Subsequently, nine vegetation types were characterized by the trait combinations that they possess using Gaussian mixture density functions. The trait maps were input to these functions to determine global occurrence probabilities for each vegetation type. We prepared vegetation maps, assuming that the most probable (and thus, most suited) vegetation type at each location will be realized. This fully traits-based vegetation map predicted 42% of the observed vegetation distribution correctly. Our results indicate that a major proportion of the predictive ability of DGVMs with respect to vegetation distribution can be attained by three traits alone if traits like stem-specific density and seed mass are included. We envision that our traits-based approach, our observation-driven trait maps, and our vegetation maps may inspire a new generation of powerful traits-based DGVMs.