Climate and traits drive bark decomposition patterns at global scale

Tree bark represents a large global carbon stock, comprising 2-20 % of woody biomass, and plays a distinct role in carbon and nutrient cycling. It is poorly understood how different abiotic and biotic drivers contribute to bark decomposition globally, or how these relationships play out in a changing climate. Here, we compile a global dataset to estimate the spatial variability of current bark decomposition rates (k values) and employ three machine learning approaches to project the response of k in future climate scenarios. We find that, globally, bark decomposes 2.9 times faster than the rate previously estimated for tree trunks (including bark). Mean k values increase latitudinally from 0.093 year^-1 in the boreal region to 1.339 year^-1 in the tropical region. Angiosperm bark decomposes faster (k = 0.450 year^-1) than gymnosperm bark (0.198 year^-1). Climate, especially mean annual temperature and mean annual precipitation, and bark phosphorus concentration, are the predominant factors driving bark decomposition. Four leading future climate scenarios predict bark decomposition to slow down in warm and wet regions and accelerate in cold and dry regions. Our results highlight the importance of considering bark decomposition in carbon cycling models.