Dead wood diversity promotes fungal diversity
Cite this dataset
Yang, Shanshan (2021). Dead wood diversity promotes fungal diversity [Dataset]. Dryad. https://doi.org/10.5061/dryad.4b8gthtdk
Dead wood is a source of life as it provides habitat and substrate for a wide range of fungal species. A growing number of studies show an important role of wood quality for fungal diversity, but in most cases for a limited number of wood traits or tree species. In this study, we evaluate how abiotic and biotic factors affect the fungal diversity and composition during dead wood decomposition. For 10 common European tree species, fresh similar-sized logs were incubated simultaneously in two Dutch forests. Annual surveys of fungal fruiting bodies were made for an 8-year period. For each tree species, 20 fresh stem traits were measured that are important for chemical and physical defence and for nutritional quality. Throughout 8 years, 4,644 fruiting bodies belonging to 255 species and 90 genera were recorded on the logs of 10 tree species. Fungal frequency and richness were higher for Angiosperms than for Gymnosperms, both for individual tree species and as a group, and higher for tree species with more acquisitive stem trait strategies (i.e., high nutritional value and low physical defence). Differences in fungal communities were strongly driven by phylogenetic group (Gymnosperms vs. Angiosperms), stem traits, decay time and forest sites, together explaining 23% of the variation. Fungal communities in sandy site diverged early in the decay process but converged later because of substrate homogenization. Of the 128 fungal species included in the analyses, 41% showed a preference for specific tree species and 34% for a specific successional year. In conclusion, dead wood quality, determined by tree species and decay stage, is an important driver of fungal diversity. For forest management, promoting a wide array of dead tree species (especially angiosperm species), a range of stem trait values and decay stages will increase fungal and, thereby, forest biodiversity.
We surveyed visible sporocarps as they are easy to detect and monitor, and are a conspicuous component of the fungal community and appreciated by naturalists and conservationist. Ecologically, sporocarps represent spore production and dispersal, which is a key event in the life cycle of fungi, critical to the distribution and long-term success of the species. Mycological surveys were conducted annually in the peak fructification time (October/November) from 2012 to 2019. No survey was made in 2018 due to the extremely dry conditions and lack of fungal fruiting bodies. In total, fungal fruiting body data were collected 1680 times (12 tree “species”×2 sites×2 individuals×2 logs×7 surveys).
During each visit, all fruiting bodies that were visible by eye were identified to species, but in a few cases only to genus, following the nomenclature from Arnolds and van den Berg (2013). When macroscopic identification was uncertain, samples were collected from the log and brought to the lab, then the species were microscopically identified. Fungal abundance is difficult to quantify as it is impossible to tease apart whether fruiting bodies belong to the same mycelium or not, or to count different fruiting bodies, which are difficult to distinguish for clumps, and tend to be highly dynamic over time. We quantified therefore fungal frequency instead. The frequency of a fungus was calculated as the number of logs on which a specific fungal species was present.
We quantified the fungal community composition based on fungal frequency. Before analysis, rare species with only one or two records in all survey years were omitted (only for analysis of fungal composition, for fungal frequency and richness, all observed fungi were considered) (cf. Buée et al. 2011, Peter et al. 2001, Pouska et al. 2011), because they might introduce high levels of noise and have a strong effect on the results. If a taxon could be identified to genus level only, then the taxon was either omitted when other taxa of this genus could be identified to species level (to avoid inflating the species number per genus), or treated as a species (if no other members of this genus were found).
Wood density and heartwood proportion in the stem cross-section area were measured on five individual trees for each tree species. Wood density was calculated as dry mass of oven-dried wood samples per fresh volume (g cm-3), which was measured by the water displacement method. The heartwood proportion was calculated using heartwood area divided by total disk area (excluding bark); for non-heartwood forming species, the heartwood proportion was set to zero.
Wood anatomical traits (ray fraction, conduit density and conduit diameter) were measured for three individual trees per tree species. Micro-thin sections were cut from the cross-section using a sledge microtome and stained with a mixture of Astrablue and Safranin for at least 5 minutes; then samples were washed with demi-water and dehydrated in ethanol series (50%, 96% and 100%); finally, thin sections were dewaxed using Roticlear® (Carl Roth, Karlsruhe, Germany) and permanently embedded with Roti®-Mount (Carl Roth, Germany). High-resolution digital images of anatomical sections were made using a camera mounted on an optical microscope. Gymnosperm sections with narrow tracheids were measured using a 10× magnification, while angiosperm sections with wider vessels were measured using a 5× magnification. Digital images were calibrated with a slide-mounted micrometer and then analysed using Fiji/ImageJ (Schindelin et al. 2012); manual adjustments were added if necessary.
Chemical traits of wood and bark were determined from stem disks of five individual trees per tree species (n =5). Carbon and nitrogen concentrations were determined by dry combustion using a Flash EA 1112 elemental analyser (Thermo Scientific, Rodana, Italy). Phosphorus was determined from sawdust samples, those samples were first digested with HNO3/HCl (1:4 mixture of 37% HCl and 65% HNO3), and then phosphorus concentration was determined with spectrophotometry using the ammonium molybdate method (at a wavelength of 880 nm) (Murphy and Riley, 1962). Lignin concentration was determined following Poorter and Villar (1997). Samples were extracted with water, methanol and chloroform to remove the soluble sugars, soluble phenols and lipids. Then starch, fructan, pectin and a part of the hemi-cellulose were removed during acid hydrolysis. Finally, after correction for ash concentration (including silicates) and remaining proteins, the lignin concentrations were calculated based on C concentration. Phenol concentrations were measured using the Folin-Ciocalteu method. A 50% methanol solution was used to extract the phenolic hydroxyl groups. Then the phenols were coloured with Folin-Ciocalteu reagent and measured at 760 nm on a spectrophotometer. Finally, total phenol concentration was calculated according to a tannic acid-based calibration curve. Finally, the pHH2O was measured following Cornelissen et al. (2006). A 0.15 ml sawdust sample and 1.2 ml demi-water were added to an Eppendorf tube and shaken for 1 h at 250 rotations per minute. Then the tubes were centrifuged for 5 min at 13,000 rpm and the supernatant was used to measure pHH2O with a WTW SenTix Mic electrode.
Nora Croin Michielsen Fonds, Award: 2018
China Scholarship Council, Award: 201706910085