Signature wood modifications reveal decomposer community history

Abstract

Correlating plant litter decay rates with initial tissue traits (e.g. C, N contents) is common practice, but in woody litter, predictive relationships are often weak. Variability in predicting wood decomposition is partially due to territorial competition among fungal decomposers that, in turn, have a range of nutritional strategies (rot types) and consequences on residues. Given this biotic influence, researchers are increasingly using culture-independent tools in an attempt to link variability more directly to decomposer groups. Our goal was to complement these tools by using certain wood modifications as 'signatures' that provide more functional information about decomposer dominance than density loss. Specifically, we used dilute alkali solubility (DAS; higher for brown rot) and lignin:density loss (L:D; higher for white rot) to infer rot type (binary) and fungal nutritional mode (gradient), respectively. We first determined strength of pattern among 29 fungi of known rot type by correlating DAS and L:D with mass loss in birch and pine. Having shown robust relationships for both techniques above a density loss threshold, we then demonstrated and resolved two issues relevant to species consortia and field trials, 1) spatial patchiness creating gravimetric bias (density bias), and 2) brown rot imprints prior or subsequent to white rot replacement (legacy effects). Finally, we field-tested our methods in a New Zealand Pinus radiata plantation in a paired-plot comparison. Overall, results validate these low-cost techniques that measure the collective histories of decomposer dominance in wood. The L:D measure also showed clear potential in classifying 'rot type' along a spectrum rather than as a traditional binary type (brown versus white rot), as it places the nutritional strategies of wood-degrading fungi on a scale (L:D=0-5, in this case). These information-rich measures of consequence can provide insight into their biological causes, strengthening the links between traits, structure, and function during wood decomposition.

Fig 1. Multi-species survey.

(A) Birch (Betula papyrifera) and (B) southern pine (eg, Pinus taeda) dilute alkali solubility (DAS) as a function of mass loss after decay by 25 white rot (WR; n = 11) and brown rot (BR; n = 14) wood-degrading fungi listed in Table 1. Blocks were harvested at weeks 3, 6, and 12, and brown rot line fits are polynomial and white rot fits were linear. Bole decay classes per tree species are shown using species-specific density ranges from Harmon et al. [42]. Four soft rot species were tested but lacked sufficient mass loss to plot. Lines (C) from this study (2014) are shown overlapping those of Worrall et al. (1997) to show the broad similarity within rot type.

Fig 2. Ratio of lignin to density loss (L:D) in birch samples from multi-species fungal survey.

(A) Summary of L:D ratios by species, delineated by decay type. (B) L:D after decay by brown rot, soft rot, and white rot as a function of total mass loss. A distinction is made between known selective and simultaneous white rot species.

Fig 3. Multi-species isolate survey dilute alkali solubility (DAS) for (A) Birch (Betula papyrifera) and (B) southern pine (eg, Pinus taeda) as a function of mass loss.

Contour lines are overlaid representing predicted mass loss-dependent probabilities (90%, 50%, and 10%) of observing brown rot at the corresponding DAS. DAS observations above the lines have a greater probability of being brown rot. Probabilities across the DAS/mass loss variable space were generated using a binary logistic regression model ($logit \left(P\left(brown rot\right)\right) = B0 + B1 * mass loss + B2 * DAS + B3 * mass loss * DAS$) developed from experimental values for each substrate independently.

Fig 4. Density bias.

A single disc from a standing birch tree (~15cm dia) was cut, air-dried, and sanded (photo). The brown rot zone was obvious, located not only by color but by having a Piptoporus betulinus sporophore emerging (inset) on the adjacent disc. Similar was done in the white rot zone using a Fomes fomentarius sporophore. Triplicate cores show the density disparity along with DAS. Milling and homogenizing the whole disc, the actual DAS was higher than predicted based on area%, due to the biasing effects of higher brown rot density.

Fig 5. DAS sample size correlations.

The DAS (wt%) values from wood collected with a 3/16ths Imperial drill bit from Radiata pine in New Zealand revealed obvious value inflation with smaller sample weights, a pattern shown here among all replicate samples including natural variability and some brown rot. This supports conservatively using more than 100 mg of material for DAS, as suggested by Shortle et al. 2012 [40].