Leaf biomechanical traits predict litter decomposability

Leaf biomechanical strength is important not only in plant defense strategies but also in "after-life” effects—determining leaf-litter decomposability. It is a composite metric that can be evaluated by fracturing a leaf using multiple methods. However, such after-life effects have not been systematically evaluated. We assessed 40 leaf functional traits, including 12 biomechanical traits measured through three standard tests (i.e., punch, tensile, and shearing tests) and categorized as fracture length-, fracture area-, or mass-based traits, to predict leaf-litter decomposition dynamics among 186 species from diverse functional groups. Categorized as fracture length-based traits, they outcompeted fracture area- and mass-based traits in predicting decomposition rates, with “force to punch” emerging as the best predictor, followed by “work to shear”. After incorporating all studied traits into a multidimensional trait space, the first principal component axis accounted for 44.3% of the total variation in decomposition rates, whereas excluding biomechanical traits reduced the variation explained to 31.6%. The leaf’s inherent resistance properties independently influenced litter decomposability beyond tissue density and lamina thickness, rendering leaf mass per area an incomplete proxy for biomechanical traits. Additionally, using the tensile force for leaves with parallel veins would underestimate leaf-litter decomposition rates. In contrast, focusing on punch force and shearing work as principal biomechanical traits offers a promising research avenue for an improved understanding of how leaf decomposability is determined.

Synthesis. Our results provide the first evaluation of leaf biomechanical traits from three standard physical resistance tests as predictors of leaf-litter decomposability. These biomechanical traits complement chemical, structural, and morphological traits and should be more effectively integrated into existing models to enhance our comprehension of the leaf-litter decomposition process.

Methods
We selected 186 species across 69 families as our target plants for collecting plant materials and the related ecology data needed in this study. These species span broad taxonomic groups or clades, including 96 woody deciduous, 68 woody evergreen, 13 graminoid, and 9 forb plants. We measured 40 leaf functional traits specifically related to leaf-litter decomposition research, including 12 biomechanical, 20 chemical, and 8 structural/morphological traits, across these species. The litterfall from these 186 species was also collected and used in the decomposition experiments. The mass remaining percent at intervals of 6, 12, 18, and 24 months since the beginning of the experiments were described via the four most-commonly employed empirical models: (1) the single- exponential model, (2) the double- exponential model, (3) the asymptotic model, and (4) the Weibull model. Thereafter, the decomposition rates obtained from Weibull model (the best fitting model) were linked with leaf functional traits.