Search for: All records

Purpose A better knowledge of how deadwood decomposes is critical for accurately characterizing carbon and nutrient cycling in forests. Fungi dominate this decomposition process, but we still have limited understanding of fungal community structuring that ultimately controls the fate of wood decomposition. This is particularly true in tropical ecosystems. To address this knowledge gap, our study capitalized on an extreme storm event that caused a large and synchronized input of deadwood to the forest floor. Methods Here we report data for the first year of wood decomposition of trees in a Puerto Rican dry forest for nine tree species that were snapped by Hurricane Maria in 2017. We measured wood properties and the associated fungal communities after 12 months of decomposition and compared them with initial wood properties and stem-inhabiting fungal communities to identify the best predictors of wood decomposition rates and chemical changes. Results Changes in wood chemistry were primarily explained by rapid xylan losses, the main hemicellulose component for the studied tree species. Fungal communities were dominated by saprotrophic and plant pathogenic fungi and showed moderate changes over time. The initial relative abundances and ratios of different fungal functional guilds were significant predictors of both xylan and glucan losses, with plant pathogenic fungi accelerating cellulose and hemicellulose decomposition rates compared to saprotrophs. Conclusion Our results confirm that fungi present at the time of treefall are strong drivers of wood decomposition and suggest that plant pathogenic fungi might act as efficient early decomposers of hemicellulose in dry tropical forests. 

Abstract. Interactions between wind and trees control energy exchanges between theatmosphere and forest canopies. This energy exchange can lead to thewidespread damage of trees, and wind is a key disturbance agent in many ofthe world's forests. However, most research on this topic has focused onconifer plantations, where risk management is economically important, ratherthan broadleaf forests, which dominate the forest carbon cycle. This studybrings together tree motion time-series data to systematically evaluate thefactors influencing tree responses to wind loading, including data from bothbroadleaf and coniferous trees in forests and open environments. We found that the two most descriptive features of tree motion were (a) the fundamental frequency, which is a measure of the speed at which a treesways and is strongly related to tree height, and (b) the slope of the powerspectrum, which is related to the efficiency of energy transfer from wind totrees. Intriguingly, the slope of the power spectrum was found to remainconstant from medium to high wind speeds for all trees in this study. Thissuggests that, contrary to some predictions, damping or amplificationmechanisms do not change dramatically at high wind speeds, and therefore winddamage risk is related, relatively simply, to wind speed. Conifers from forests were distinct from broadleaves in terms of theirresponse to wind loading. Specifically, the fundamental frequency of forestconifers was related to their size according to the cantilever beam model(i.e. vertically distributed mass), whereas broadleaves were betterapproximated by the simple pendulum model (i.e. dominated by the crown).Forest conifers also had a steeper slope of the power spectrum. We interpretthese finding as being strongly related to tree architecture; i.e. conifersgenerally have a simple shape due to their apical dominance, whereasbroadleaves exhibit a much wider range of architectures with more dominantcrowns.