Forest Ecology and Management 391 (2017) 86–95

Contents lists available at ScienceDirect

Forest Ecology and Management

journal homepage: www.elsevier.com/locate/foreco

Wood decay rates of 13 temperate tree species in relation to wood

properties, enzyme activities and organismic diversities
Tiemo Kahl a,b,⇑,1, Tobias Arnstadt c, Kristin Baber a,d, Claus Bässler e, Jürgen Bauhus a, Werner Borken f,
François Buscot g,h, Andreas Floren i, Christoph Heibl e, Dominik Hessenmöller j, Martin Hofrichter c,
Björn Hoppe g,k, Harald Kellner c, Dirk Krüger g, Karl Eduard Linsenmair i, Egbert Matzner f, Peter Otto l,
Witoon Purahong g, Claudia Seilwinder m, Ernst-Detlef Schulze j, Beate Wende i, Wolfgang W. Weisser m,
Martin M. Gossner m,n,⇑,1
a
Chair of Silviculture, University of Freiburg, Tennenbacherstr. 4, 79106 Freiburg, Germany
b
UNESCO Biosphere Reserve Thuringian Forest, Brunnenstr. 1, 98711 Schmiedefeld am Rennsteig, Germany
c
Department of Bio- and Environmental Sciences, International Institute Zittau, Technische Universität Dresden, Markt 23, D-02763 Zittau, Germany
d
Department of Systematic Botany and Functional Biodiversity, Institute of Biology, University of Leipzig, Johannisallee 21-23, 04103 Leipzig, Germany
e
National Park Bavarian Forest, Freyunger Str. 2, 94481 Grafenau, Germany
f
Soil Ecology, University of Bayreuth, Bayreuth Center of Ecology and Environmental Research (BayCEER), Dr.-Hans-Frisch-Str. 1-3, 95448 Bayreuth, Germany
g
UFZ – Helmholtz Centre for Environmental Research, Department of Soil Ecology, Th.-Lieser- Str. 4, D-06120 Halle (Saale), Germany
h
German Centre for Integrative Biodiversity Research (iDiv), Halle-Jena-Leipzig, Deutscher Platz 5e, D-04103 Leipzig, Germany
i
Department of Animal Ecology and Tropical Biology, Biocenter, University of Würzburg, Am Hubland, D-97074 Würzburg, Germany
j
Max-Planck-Institute for Biogeochemistry, Hans-Knöll-Strasse 10, 07745 Jena, Germany
k
Julius Kühn-Institute – Federal Research Centre for Cultivated Plants, Institute for National and International Plant Health, Messeweg 11/12, D-38104 Braunschweig, Germany
l
Department of Molecular Evolution and Plant Systematics, Institute of Biology, University of Leipzig, Johannisallee 21-23, Leipzig 04103, Germany
m
Terrestrial Ecology Research Group, Department of Ecology and Ecosystem Management, Center of School of Life and Food Sciences Weihenstephan, Technische Universität München,
Hans-Carl-von-Carlowitz-Platz 2, D-85354 Freising-Weihenstephan, Germany
n
Swiss Federal Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, Switzerland

a r t i c l e i n f o a b s t r a c t

Article history: Deadwood decay is an important ecosystem process in forest ecosystems, but the relative contribution of
Received 5 October 2016 specific wood properties of tree species, activities of wood-degrading enzymes, and decomposer commu-
Received in revised form 12 January 2017 nities such as fungi and insects is unclear. We ask whether wood properties, in particular differences
Accepted 6 February 2017
between angiosperms and gymnosperms, and organismic diversity of colonizers contribute to wood
decomposition. To test this, we exposed deadwood logs of 13 tree species, covering four gymnosperms
and nine angiosperm species, in 30 plots under different forest management in three regions in
Keywords:
Germany. After a decomposition time of 6.5 years Carpinus betulus and Fagus sylvatica showed the highest
Wood decomposition
Ecosystem function
decay rates. We found a positive correlation of decay rate with enzyme activities, chemical wood prop-
Saproxylic beetles erties (S, K concentration) and organismic diversity, while, heartwood character, lignin content, extrac-
Biodiversity Exploratories tive concentration and phenol content were negatively correlated with decay rate across all 13 tree
Deadwood experiment species. By applying a multi-model inference approach we found that the activity of the wood-
degrading enzymes laccase and endocellulase, beetle diversity, heartwood presence, wood ray height
and fungal diversity were the most important predictor variables for wood decay. Although we were
not able to identify direct cause and effect relations by our approach, we conclude that enzyme activity
and organismic diversity are the main drivers of wood decay rate, which greatly differed among tree spe-
cies. Maintaining high tree species diversity will therefore result in high structural deadwood diversity in
terms of decay rate and decay stage.
Ó 2017 Elsevier B.V. All rights reserved.

⇑ Corresponding authors at: UNESCO Biosphere Reserve Thuringian Forest,

Brunnenstr. 1, 98711 Schmiedefeld am Rennsteig, Germany (T. Kahl). Swiss Federal 1. Introduction
Research Institute WSL, Zürcherstrasse 111, CH-8903 Birmensdorf, Switzerland
(M.M. Gossner).
Deadwood is known to be an important structural component
E-mail addresses: kahltiemo@web.de (T. Kahl), martin.gossner@wsl.ch
(M.M. Gossner). of forests (Bauhus et al., 2009) that influences a large number of
1
These authors contributed equally to the manuscript. ecosystem functions (Harmon et al., 1986; Cornwell et al., 2009)

http://dx.doi.org/10.1016/j.foreco.2017.02.012
0378-1127/Ó 2017 Elsevier B.V. All rights reserved.
 T. Kahl et al. / Forest Ecology and Management 391 (2017) 86–95 87

and is a critical resource for wood-dwelling organisms (Lonsdale Furthermore differences in the presence of organisms and their
et al., 2008; Rondeux and Sanchez, 2010; Stokland et al., 2012). activity might affect wood decay. A high diversity of saproxylic
The rate of deadwood decay, as a combination of biological res- fungi and beetle species in logs of different tree species have been
piration, leaching and fragmentation (Harmon et al., 1986), is shown in previous studies (Baber et al., 2016; Gossner et al., 2016).
mainly the result of organismic activity (e.g. fungi, insects), deter- In wood decomposition mediated by fungi, the main secreted
mined by intrinsic (e.g. tree species properties) and environmental enzymes are oxidoreductases modifying lignin via radical forma-
factors (e.g. temperature, moisture). In the early phase of decay, tion with the help of dioxygen (laccase) or hydrogen peroxide (sev-
biological respiration (i.e. mineralization of wood components to eral peroxidases), and diverse glycoside hydrolases, which
CO2) is the main process, while leaching only accounts for about hydrolytically cleave the major polysaccharides (cellulose, hemi-
1–2% of mass loss (Kahl, 2008; Bantle et al., 2014). Fragmentation celluloses) to sugars (Hatakka and Hammel, 2010; Kellner et al.,
can also cause a substantial mass loss of deadwood, e.g. 18–30% of 2014; Arnstadt et al., 2016a; Noll et al., 2016). While substantial
decomposition rates over the first 18 years in some European tree lignin decomposition is exclusively mediated by white-rot basid-
species (Herrmann et al., 2015), and it may become even more iomycetes possessing high-redox potential peroxidases (i.e. class
important in the late phase of decomposition (up to 40%; II manganese and lignin peroxidases, Floudas et al., 2012), cellulose
Harmon et al., 1986). and hemicelluloses can be disintegrated by all wood-decomposing
Decay rates for common European tree species have been well fungi (white-, brown- and soft-rot fungi), as well as by specialized
documented for different climatic regions (Pietsch et al., 2014). bacteria. The resulting sugars, e.g. cellobiose, glucose, xylose, man-
First estimates were reported for Quercus spp., Fraxinus excelsior, nose, and arabinose, serve as carbon and energy source (Riley et al.,
Betula spp. and Corylus avellana (Swift et al., 1976; Christensen, 2014). Against this background, the activities of laccase, man-
1977) already in the 1970s. However, decay rates for less abundant ganese and general peroxidases can be used as proxy for lignin
tree species such as Carpinus betulus, Prunus avium or Tilia spp. attack. The same is true for endo-1,4-b-glucanase and endo-
have not been quantified. 1,4-b-xylanase, which mediate the first step in the breakdown of
Estimates on decay rates are often based on a chronosequence the polysaccharide backbone by forming oligo-, di- and monosac-
approach, where in-situ deadwood is sampled of different times charides (Kögel-Knabner, 2002) representing the enzymatic
since death and compared. This approach has a sampling bias, polysaccharide disintegration. Despite the importance of these
due to the fact that deadwood with low decay rates is more likely enzymes for wood decomposition, their relation to decay rates of
to be sampled than deadwood with high decay rates, which may deadwood has never been studied. Only few publications have so
have already decayed completely at the time of observation far analyzed enzymes in deadwood (Valaskova et al., 2009;
(Kruys et al., 2002). This may result in an underestimation of dead- Větrovský et al., 2011; van der Wal et al., 2015; Arnstadt et al.,
wood decay rates. To quantify true decay rates, these should be 2016a; Baldrian et al., 2016; Noll et al., 2016; Purahong et al.,
measured during the entire time course of wood decay. An unbi- 2016).
ased estimation of decay rates can either be achieved by resam- Differences in fungal species identity and fungal communities
pling of sites that include all recently killed trees, or by can cause distinct differences in decay rates (Boddy, 2001; Kahl
experiments that place deadwood logs in forests. Such site-based et al., 2015). However, to date no consistent relationship between
multi-species studies on decay rates including angiosperms and deadwood decomposition and fungal decomposer diversity has
gymnosperms are, however, still largely lacking (Weedon et al., been demonstrated (van der Wal et al., 2013). It has been shown
2009). that decay rates can be either positively (Setälä and McLean,
Decay rates of different tree species could be largely affected by 2004; Tiunov and Scheu, 2005; Toljander et al., 2006; Valentín
functional group or species-specific wood traits. A meta-analysis of et al., 2014; Kahl et al., 2015) or negatively (Fukami et al., 2010)
Weedon et al. (2009) showed that gymnosperms generally tend to correlated with fungal species richness. Based on the quantification
have lower decay rates than angiosperms. In gymnosperms, sol- of fungal operational taxonomic units (OTU), Hoppe et al. (2015)
vent extractable organic compounds (extractives) comprise mainly found no correlation between total richness (using molecular
terpenes and their derivatives, which slow down fungal growth NGS 454 data) and wood decomposition rate in both coniferous
and degradation activity (Dix and Webster, 1995; Pearce, 1996). (Picea abies) and deciduous (Fagus sylvatica) deadwood. Admit-
In contrast, in angiosperms the chemical composition of extrac- tedly, there is an undeniable relationship between fungal commu-
tives is more diverse, reaching from resin-like materials of carbo- nity composition and wood decomposition. The fungal assembly
hydrate gum in cherry (Pearce, 1996), which may not affect history (Fukami et al., 2010), the presence of species with extraor-
decay, to phenolic compounds (e.g. quercitin) and tannins in oak, dinary decomposition activities (Lindner et al., 2011; Kubartová
which inhibit fungal growth (Hart and Hillis, 1972; Aloui et al., et al., 2012) and the environmental conditions (Renvall, 1995;
2004). Furthermore, the lignin of gymnosperms is more difficult Høiland and Bendiksen, 1996) can determine the strength of this
to oxidize compared to that of angiosperms, since it consists to relationship. Van der Wal et al. (2015) showed that the most
more than 90% of guaiacyl moieties, compared to 0–33% in angios- important factors explaining variation in wood decay rates can
perm lignin. The lignin of angiosperms contains a higher propor- change over time and the strength of competitive interactions
tion of sinapyl moieties that degrades more easily (Brunow, between fungi may level off with increased wood decay. This study
2001; Higuchi, 2006). Guaiacyl units make the lignin of gym- furthermore showed that wood moisture content contributed
nosperms particularly compact, which impedes enzymatic attack especially to explain sapwood decay in early decay stages, whereas
in general (Hatakka and Hammel, 2010). fungal community composition and species richness were the best
Within angiosperms, wood traits such as phosphorus and nitro- predictors for mass loss in the later stages (van der Wal et al.,
gen content, and low C:N ratio are positively correlated with decay 2015).
rates (Weedon et al., 2009). It is also well known (EN350-2, 1994) Insects also play an important role in the wood decomposition
that tree species with distinct heartwood (e.g. Quercus spp.) tend to process. Based on a meta-analysis, Ulyshen (2016) estimates the
have higher resistance against wood decaying fungi. The content of contributions of invertebrates to wood decomposition to account
extractives, such as tannins and terpenoids/resins, in heartwood is for 10–20%. He suggests four main underlying mechanisms:
often higher than in sapwood and the extracts from heartwood are (1) enzymatic digestion by endogenous enzymes (Watanabe and
usually more toxic (Scheffer, 1966; Hillis, 1999; Gierlinger et al., Tokuda, 2001, 2010) as well as by enzymes produced by endo-
2004). (gut microbes; e.g. Suh et al., 2005) and ectosymbionts (e.g. fungi
88 T. Kahl et al. / Forest Ecology and Management 391 (2017) 86–95

cultivating ambrosia beetles; e.g. Hofstetter et al., 2015a), (2) sub- (n = 1170 logs). Tree species include nine Angiosperms: Acer spp.
strate alteration by tunnelling and fragmentation and therefore (A. pseudoplatanus, A. platanoides, A. campestre), Betula pendula,
building entrance ports for other species, e.g. fungi or later succes- Carpinus betulus, Fagus sylvatica, Fraxinus excelsior, Populus spp. (P.
sional species, (3) biotic interactions with microbes (e.g. alteration tremula and hybrids), Prunus avium, and Quercus spp. (Q. robur,
of microbial communities) and other invertebrates (e.g. by preda- Q. petraea), Tilia spp. (T. cordata, T. platyphyllos) and four Gym-
tion, competitive exclusion) and (4) nitrogen fertilization which nosperms: Larix decidua, Picea abies, Pinus sylvestris, and Pseudot-
promotes nitrogen fixation by endosymbiotic and free-living suga menziesii). All logs are approximately 4 m long and have a
bacteria. In Central Europe, where termites are absent, large mean diameter of 31 ± 5.9 cm (SD). All logs were cut in winter
wood-boring beetles are assumed to be most influential and their 2008/2009 in the federal state of Thuringia (Germany) and then
presence more important than species richness itself, but detailed transported to each plot. Within a subplot, the 13 logs were placed
studies on the importance of insect species richness for wood in random order beside each other with a distance of ca. 1 m
decomposition of different tree species are still lacking (Ulyshen, between logs. We carefully followed a protocol to minimize the
2016). potential bias by colonization of logs prior to translocation to the
While our knowledge on wood decay rates in relation to wood experimental sites. Owing to limited supplies, P. avium logs are
properties and organismic diversity has increased in the past dec- missing on 27, Acer spp. on 2 and Fagus sylvatica on 1 of the 90 sub-
ades, there are still large uncertainties how the different factors plots (Kahl et al., 2015). One subplot with 12 logs in the explora-
interact and differ between tree species. Unbiased estimates from tory Schorfheide was unlawfully removed in 2012, reducing the
field deadwood experiments are scarce. Here, we report on the total number of logs to 1128.
results of a large-scale field experiment where logs of 13 tree spe-
cies were placed into 30 forests stands in Germany. We studied 2.3. Wood decay rate
decay rates, wood properties and organismic diversity.
We asked the following questions: Deadwood decay rate was calculated based on density loss. Ini-
tial wood density was determined using discs that were cut in
1. Do gymnosperms show consistently lower rates of decay than 2009 immediately after exposure from all logs. The discs were then
angiosperms under common environmental conditions? dried at 65 °C and stored in a warehouse. After removing the bark
2. Are anatomical, chemical or structural wood traits good predic- the discs were drilled with a 1 cm diameter auger and a cordless
tors of decay rates? power drill in radial direction to the pith. The second sampling
3. Are biotic factors such as enzymatic activities of wood- from the exposed logs took place in June 2015. Owing to the
degrading enzymes and species richness of fungi and insects advanced decay in 2015, logs were, after removing the bark, drilled
good predictors of decay rates? with a 4 cm diameter Forstner bit in 60 cm distance from one end
of the log. The Forstner bit allows drilling clean cylindrical holes
2. Methods even in decayed wood material. In both 2009 and 2015, wood
shavings were carefully collected and dried at 65 °C until no fur-
2.1. Study area ther decrease in weight was detected. Dry wood density (g cm3)
in 2009 and 2015 was calculated as ratio of dry wood weight (g)
The study was conducted within the ‘‘Biodiversity Explorato- and drill hole volume (cm3). Decay rate constant (k) using density
ries”, a large-scale and long-term project to investigate the rela- loss for each log was calculated based on an exponential decay
tionship between land-use intensity, biodiversity and ecosystem function as proposed by Olson (1963). Decay time between 2009
processes in three different regions (exploratories) in Germany and 2015 was set to 6.5 years. Negative decay rates occurred in
(for details, see Fischer et al., 2010); the UNESCO Biosphere Reserve individual logs of 12 out of 13 tree species (except Carpinus betulus;
Schorfheide-Chorin in North-Eastern Germany (52°470 2500 –53° Supplement Table A3-1), which seems to be a specific problem of
130 2600 N/13°230 2700 –14°080 5300 E, 3–140 m a.s.l.), the National Park our applied methods. Since negative decay rates, which would
Hainich and surrounding Hainich-Dün area in Central Germany mean an increase in density during decay, are rather unlikely, we
(50°560 1400 –51°220 4300 N/10°100 2400 –10°460 4500 E, 285–550 m a.s.l.), believe that our methods for the determination of wood density
and the UNESCO Biosphere Reserve Swabian Alb in South- were not as precise as they should be. The following reasons could
Western Germany (48°200 2800 –48°320 0200 N/9°100 4900 –09°350 5400 E, lead to the observed discrepancies: First, initial wood density was
460–860 m a.s.l.). With 700–1000 mm, annual precipitation is based on discs cut from each log in 2009. These discs represent
highest in the Swabian Alb, followed by Hainich-Dün (500– only about 2% of the log, while samples in 2015 were taken at least
800 mm) and Schorfheide-Chorin (520–580 mm). The mean 50 cm away from where the discs were cut. We do not know how
annual temperature increases from Swabian Alb (6.0–7.0 °C) to large the density variability is within each log. Second, the discs
Hainich-Dün (6.5–8.0 °C) to Schorfheide-Chorin (8.0–8.5 °C) were dried in 2009 before sampling while the logs in 2015 were
(Fischer et al., 2010). moist. The shrinkage of the wood after drying in 2009 should have
caused an even higher density. The average volume shrinkage is
2.2. BELongDead experiment about 0.42% per percent water content (Wagenführ, 2006). Third,
differences in the used drilling devices (d = 1 cm auger vs.
The BELongDead (Biodiversity Exploratories Long term Dead- d = 4 cm forstner bit) in dry vs. moist wood. It could be that our
wood) Experiment was established on a total of 30 plots (9 in method of volume determination was not precise enough. To this
Schorfheide-Chorin and Swabian Alb, and 12 in Hainich-Dün) in point we cannot say why we found these negative decay rates in
2009. The 30 plots cover a range of forest types and management some of the logs. Although this is a major concern in our study
intensities including extensively managed or unmanaged forests we believe that all tree species were affected in the same way
dominated by European beech (9), managed even- and uneven- and that the relative relationship in decay rates between tree spe-
aged forests dominated by European beech (12), and intensively cies remained largely unaffected (Supplement Fig. A3-1 & A3-2). In
managed even-aged coniferous (Norway spruce and Scots pine) cases where negative decay rates were measured for individual
plantations (9). logs between inventory periods, these rates were set to 0. These
On each plot deadwood logs of 13 tree genera (henceforth tree results were only used for the illustration of decay rate distribution
species for simplicity) were exposed in three replicates (subplots) within and among tree species.
 T. Kahl et al. / Forest Ecology and Management 391 (2017) 86–95 89

For correlations with predictors of wood decay we used decay were summed up to total peroxidase (POX) activity. Hydrolytic
rates based on single exponential models of wood density loss of enzyme activities related to polysaccharide disintegration, i.e.
all logs for each tree species. Average wood density of all logs endocellulase (EC 3.2.1.4), and (endo)xylanase (EC 3.2.1.8), were
per tree species in 2009 (undecayed) and in 2015 (after six years measured spectrophotometrically applying AZO-cellulose and
of decay) were used as input variables for the calculation of decay AZO-xylan as substrates, respectively, following the protocol of
rates. Megazymes (Bray, Ireland) (Vetrovsky et al., 2011).

2.4. Predictors of decay rates 2.4.3. Fungal species richness based on fruit bodies
Fungal fruit body data were gathered on 379 out of 1128 logs (1
We measured a number of anatomical, mechanical, physical subplot per plot) in autumn (September to October) 2012 during
and chemical wood properties, enzyme activities, and organismic the peak fruiting season (Halme and Kotiaho, 2012) parallel to
species richness for all 13 tree species, to analyze how these corre- the wood sampling for molecular, wood chemical and enzyme
late with wood decay rates. Due to methodological and time limi- activity analysis. We collected macrofungi with a fruit body, fruit
tations some of the traits were extracted from literature or body aggregation or stroma size of at least 5 mm diameter. The
assessed only on a subset of logs. Hence, the number of replicates fungal inventory covered Basidiomycota (mainly corticioid and
for measured variables varies for each method (Table 1). We used poroid fungi) and macrofungal groups from the phylum Ascomy-
mean values per tree species for the analysis (Supplement cota (especially Pyrenomycetes and Discomycetes). We investi-
Table A3-2). gated exclusively the visible part of the log and did not remove
moss or bark. The fungal species were identified in situ or, if nec-
2.4.1. Wood properties essary collected for later microscopical identification. Collected
2.4.1.1. Anatomical, mechanical and physical wood proper- specimens were dried and deposited at the Herbarium Universi-
ties. Species-specific bark proportion, occurrence of distinct heart- tatis Lipsiensis LZ (University Leipzig, Germany). We used species
wood, ratio of longitudinal parenchyma in wood to total wood richness for each tree species averaged per log as well as rarefied
volume, wood ray height and wood ray thickness were taken from to lowest sample size (20 logs) as predictor variables in our
the literature (Wagenführ, 2006). Wood density in 2009 was mea- models.
sured as described above.
2.4.4. Fungal molecular operational taxonomic units
2.4.1.2. Chemical wood properties. Initial chemical properties of sap- In addition to the identification of macrofungi, we also used a
wood were determined from discs of three logs per tree species molecular method to quantify fungal richness. Fungal automated
(n = 3), taken immediately after exposure of logs on the forest floor, ribosomal intergenic spacer analysis (F-ARISA) was performed as
then dried and cut in wedge-like pieces, representing the propor- described in Purahong et al. (2014a, 2016). Briefly, DNA was
tion of all tree rings. Carbon and nitrogen concentrations of milled extracted from 100 mg of each homogenized wood sample (see
sapwood were analyzed using a CN analyzer. Following pressure method for enzyme activities above) using the ZR Soil Microbe
digestion, phosphorous, sulphur, potassium, calcium, magnesium DNA MiniPrep kit (Zymo Research, Irvine, CA, USA), according to
and manganese were determined by inductively coupled plasma the manufacturer’s instructions. Fungal ITS was amplified using
with optical emission spectrometry (ICP-OES). Sapwood shavings FAM-labelled primer ITS1-F (50 -CTTGGTCATTTAGAGGAAGTAA-30 ,
were extracted with deionized water (1:15) for spectroscopic anal- Gardes and Bruns, 1993) and unlabelled ITS4 (50 -TCCTCCGCTTATT
ysis of water soluble phenols (Box, 1983), calibrated with phenol as GATATGC-30 , White et al., 1990). The amplified PCR fragments
standards. were discriminated by capillary electrophoresis and raw profiles
were analyzed using the Gene Mapper Software 4.0 (Applied
2.4.1.3. Organic extractives, pH-value and lignin. Organic extractives Biosystems). All peaks of the fragments between 390 and
(e.g. tannins and terpenoids/resins) were measured gravimetrically 1000 bp that appeared in two technical PCR replicates were used
after extraction of the dried and milled wood samples (same sam- for further analyses. Operational taxonomic unit (OTU) binning
ples as for chemical wood properties) with acetone in an acceler- was carried out using an interactive custom binning script
ated solvent extraction device (Dionex ASE 200). Afterwards, the (Ramette, 2009) in R version 2.14.1 (R Core Team, 2011). Fungal
Klason lignin content of the extracted wood samples was deter- OTUs were defined as described previously by Green et al. (2004)
mined gravimetrically after further grinding and sequential for the F-ARISA approach and may differ from those OTU defini-
hydrolysis with sulphuric acid. The part of acid soluble lignin tions relying on sequences (Purahong et al., 2014b). They could,
was measured spectrophotometrically in the hydrolysate and alternatively, be called ribotypes (Gleeson et al., 2005). We used
summed up with Klason lignin to the total lignin content fungal OTU richness for each tree species averaged per log as well
(Arnstadt et al., 2016a). For pH measurement, the milled wood as total OTU richness of 27 logs (equal number of replicates per
samples were extracted with deionized water and the pH detected tree species) as predictor variables in our models.
with an electrode in the aqueous supernatant (Arnstadt et al.,
2016a). 2.4.5. Saproxylic Coleoptera
Saproxylic Coleoptera emerging from the logs were sampled by
2.4.2. Enzymatic activities closed emergence eclectors for three years, 2010–2012 (for details,
In 2012, wood samples were taken using a cordless drill (Makita see Müller et al., 2015; Gossner et al., 2016). These traps sample
BDF 451) equipped with a wood auger (20 mm  450 mm). Activ- insects emerging from the logs over a full season. Traps were
ities of degradative enzyme were determined in 379 wood samples installed in March 2010 (first subplot) and March 2011 (second
(1 subplot per plot) in 2012. In the aqueous extracts of the milled subplot). Sampling vials were emptied monthly until the end of
samples, lignin-related oxidoreductases (laccase EC 1.10.3.2, man- October and samples were stored in 70% ethanol. In autumn, traps
ganese peroxidase EC 1.11.1.13 and general peroxidase activity were dismantled for the winter. The traps were moved 35 cm
EC 1.11.1.7/14/16; Kellner et al., 2014) were assayed spectrophoto- down the log every year. Specimens were sorted to order by stu-
metrically with ABTS (2,20 -azino-bis(3-ethylbenzothia-zoline-6-sul dent helpers and all Coleoptera were identified to species level
fonic acid) as substrate according to Arnstadt et al. (2016a, 2016b). by taxonomic specialists. Species were classified as saproxylic
The activities of manganese peroxidase and general peroxidase according to Seibold et al. (2015). We calculated beetle species
90 T. Kahl et al. / Forest Ecology and Management 391 (2017) 86–95

Table 1
Number of replicates (logs) per tree species and years of sampling for decay rate and each predictor category.

Tree Decay rate Wood density Wood chemistry Enzymes Fungi fruit Fungi OTU Coleoptera
species 2009–2015 2009 2009 2012 body 2012 2012 2010–2012
Angiosperms
Acer 85 85 3 29 29 27 59
Betula 87 87 3 30 30 27 60
Carpinus 88 88 3 30 30 27 60
Fagus 85 85 3 30 30 27 59
Fraxinus 87 87 3 30 30 27 60
Populus 86 86 3 30 30 27 60
Prunus 62 62 3 20 20 27 46
Quercus 87 87 3 30 30 27 60
Tilia 88 89 3 30 30 27 60
Gymnosperms
Larix 89 89 3 29 29 27 60
Picea 87 88 3 30 30 27 60
Pinus 85 86 3 30 30 27 60
Pseudotsuga 87 87 3 31 31 27 60

richness for each tree species by averaging over logs and year, rar- axes explaining at least 70% of the variance were selected. We
efied to lowest sample size of all tree species (104 log-years). selected the variable that showed the strongest correlation with
the corresponding PCA axis. The correlations among selected vari-
ables was Pearson’s r < 0.70 in all cases. According to Tabchnick
2.5. Tree phylogeny
and Fidell (1996) the independent variables with a bivariate corre-
lation stronger than 0.70 should not be included in multiple regres-
Because tree species may not be regarded as independent sam-
sion analysis. The selected set of variables were standardized
ples, we considered the phylogenetic relatedness between tree
(scaled to zero mean and unit variance) and then used in either lin-
species. Therefore we used a bioinformatics pipeline implemented
ear (function lm) or phylogenetic linear multiple regression models
in the R package megaptera (Heibl, 2014), to derive a set of align-
(phylolm function in the R package phylolm; Ho and Ane, 2014)
ments on which our phylogenetic hypothesis is based. We
with decay rate as dependent variable. For model selection, we
screened all protein-encoding and rRNA genes of chloroplast and
applied the multimodel inference approach as proposed by
mitochondrion as well as the nuclear rRNA genes, the histone H3
Grueber et al. (2011). We first formulated a global model including
gene and the phytochrome C gene. For each tree genus and genetic
all predictor variables und conducted automated model selection
locus, we downloaded all available accessions from GenBank and
with subsets of the global model by the dredge function based
BOLDSYSTEMS. Under the assumption of monophyletic genera,
on AICc (small-sample-size corrected version of AIC) in the R pack-
we created majority-rule consensus sequences of all congeneric
age MuMIn (Bartoń, 2016). We subsequently selected the best
sequences that passed the quality checks. 11 loci provided the best
models based on delta AIC < 2 (substantial support) and AIC < 7
taxonomic coverage and were concatenated into a final superma-
(some support) as suggested by Burnham et al. (2011) using the
trix: atp1, atp6, atp9, ndhF, rbcL, cox1, cox3, matR, nd4L, nd6, 16S
get.models function. The relative importance of each independent
rRNA. Tree topology and ultrametric branch lengths were modelled
variable was assessed by calculating the cumulative Akaike
in a maximum likelihood framework using RAXML version 8.2.4
weights of models containing a particular predictor.
(Stamatakis, 2014) and the function chronos in the R package
We calculated the phylogenetic signal (Pagel’s k and Blomberg’s k)
ape (Paradis et al., 2004). Confidence in the topology of our phylo-
for each variable by using the phylosig function in the R package
genetic hypothesis was assessed with non-parametric bootstrap
phytools (Revell, 2012).
using stopping criteria described in Pattengale et al. (2010). All
branches received bootstrap support of 83% or more and the tree
topology is in accordance with current knowledge (Supplement 3. Results
Fig. A1-2).
3.1. Decay rates of different tree species
2.6. Statistical analysis
Average decay rates (k; incl. negative values) of gymnosperms
(0.014 ± 0.016 y1, N = 4) were significantly lower than those of
All analyses were conducted in R version 3.2.2 (RCoreTeam,
angiosperms (0.046 ± 0.021 y1, N = 9; LM: t = 3.178, p = 0.009,
2015). We calculated Spearman’s rank correlation coefficients to
R2adj=0.431). Highest decay rates in angiosperms were found for
show the correlations between the dependent variable decay rate
Carpinus betulus and Fagus sylvatica, while lowest were found for
and all independent variables.
Fraxinus excelsior and Quercus spp. The overall lowest decay rates
For analyzing the importance of different predictor variables for
were recorded for Larix decidua and Pseudotsuga menziesii (Fig. 1).
wood decay, we used two approaches; in one we considered the
phylogenetic relatedness between tree species, in the other not.
To reduce the number of explanatory variables, we first classified 3.2. Effect of predictor variables on decay rates
them in three groups: (i) anatomical, mechanical and physical
wood traits, (ii) chemical wood traits and (iii) enzymes and organ- The strongest correlations with decay rate were found for pres-
ismic diversity. We then used either uncorrected (function ence of heartwood and organic extractives, both of which were
prcomp) or phylogenetically corrected PCAs (function phyl.pca in negative (Fig. 2 & Supplement Fig. A1-1). All enzymatic activities,
the R package phytools; Revell, 2012) based on correlation matri- fungal species richness and S and K concentrations showed strong
ces. Variables were log or square root transformed when necessary positive correlations (correlation coefficient >0.70; Fig. 2 & Supple-
to meet assumptions of normality. For each PCA, the first three PCA ment Fig. A1-1). Among independent variables, parenchyma ratio,
 T. Kahl et al. / Forest Ecology and Management 391 (2017) 86–95 91

K, Lignin and mean fungi fruit body richness showed a strong phy-
logenetic signal (Pagel’s lambda and Blomberg’s k < 0.05; Supple-
ment Table A1-1).

3.3. Correlations among predictive variables

In uncorrected PCA, ray height, proportion of bark and the

occurrence of distinct heartwood were selected among physical
wood traits (Supplement Fig. A2-1a). For chemical wood traits
(Supplement Fig. A2-1b), the concentrations of phosphorus, phe-
nols and lignin were chosen. For enzyme activities and organismic
diversity, laccase activity, rarefied coleopteran species richness and
mean fungal fruit body species richness were selected (Supplement
Fig. A2-1c). The phylogenetically corrected PCA resulted in the
same variables except for laccase which was replaced by endocel-
lulase (Supplement Fig. A2-2). For details see Tables A2-1 & A2-2 in
the Supplement.

3.4. Relative importance of predictive variables

Model selection revealed that the variables best predicting

wood decay were laccase activity, rarified Coleoptera species rich-
ness, the occurrence of distinct heartwood and wood ray height
(within the best models; DAIC < 7). When a phylogenetic correc-
( )
tion was performed, the variables were the same except for laccase
Fig. 1. Decay rates of 13 tree species using density loss between 2009 and 2015 activity, which was replaced by endocellulase activity. In the
based on exponential decay. Proposed phylogeny of tree genera is shown for uncorrected model additionally mean fungal fruit body species
illustrative purpose only. A tree including bootstrap values is provided in the richness was selected. When using a DAIC of <2, only the models
supplement (Fig. A1-2). Median values, 25%/75% percentiles (Box) and Min-Max
with rarefied Coleoptera species richness and laccase activity was
values (Whiskers) excluding outliers are shown.
selected as best model based on the uncorrected approach and
the models with rarefied Coleoptera species richness and endocel-
lulase activity and either with or without heartwood were selected

Fig. 2. Correlation between the dependent variable decay rate and the independent variables for all tree species. The Spearman’s rank correlation coefficient for each variable
is given. For phylogenetic signal in traits, see Supplement Table A1-1.
92 T. Kahl et al. / Forest Ecology and Management 391 (2017) 86–95

Table 2
Best models (DAICc < 2) for linear multiple regression models with decay rate as
dependent variable. Model selection was conducted based on AICc (small-sample-size
corrected version of AIC). Please note that in the uncorrected model laccase and in the
phylogenetically corrected model endocellulase was used as predictor variable. NA:
This predictor is not included in a particular model.

Uncorrected Corrected
Rank 1 Rank 1 Rank 2
AICc 41.928 43.113 41.307
DAICc 0.000 0.000 1.806
Weight 0.238 0.448 0.182
Adj. R2 0.769 NA NA
df 4 5 6
Intercept 0.172 0.169 0.187
Bark ratio NA NA NA
Phenol NA NA NA
P NA NA NA
Rarefied beetle richness 0.002 0.019 0.021
Laccase/Endocellulase 0.022 0.041 0.033
Heartwood NA NA 0.033
Lignin NA NA NA
Mean fungal fruit body richness NA NA NA Fig. 3. Importance values (assessed by the cumulative Akaike weights of models
Wood ray height NA NA NA containing a particular predictor) of predictor variables in explaining decay rates of
the 13 different tree species based on model averaging with DAIC < 7. Values for
uncorrected and phylogenetically corrected models are shown. Signs indicate
whether the effect on wood decay is positive or negative. For details see Tables S1
as best models based on the phylogenetically corrected approach
and S2 in the Supplement.
(Table 2, Supplement Table A3-1 & A3-2). This shows that laccase
and endocellulase activity, rarified Coleoptera species richness
and the occurrence of distinct heartwood were of the selected ones The available data on wood density in 2009 and 2015 do not
the most important predictors of wood decay of the 13 studied tree allow us to say whether the exponential decay is the best way to
species (Fig. 3). Mean fungal species richness and lignin content model decay rates for each of the 13 tree species. Nevertheless,
had a relatively higher importance when using uncorrected models we hypothesize that the relative differences between tree species
compared to phylogenetically corrected models, which can be will stay largely constant. The experimental design did allow us
explained by the strong phylogenetic signal in these variables to study the decay rate and its influencing factors independently
(see Supplement Table A1-1). of the autochthonous tree species, of the cause of tree death and
of climatic and edaphic conditions. Since all logs were cut in the
4. Discussion same area, we assume that all logs have a similar history of initial
properties such as colonization with endophytic fungi and insects.
4.1. Decay rate Whether or not this reduced the variability of the observed decay
rates cannot be deduced from our data.
This is the first study that presents decay rates of 13 tree species
in a deadwood experiment in temperate European forests, includ- 4.2. Effect of wood properties and diversity on wood decay rate
ing tree species for which decay rates were so far unknown, i.e.
Carpinus betulus, Prunus avium and Tilia spp. Interestingly, the 4.2.1. Physical, anatomical and chemical wood properties
respective data for decay rates of Carpinus betulus were found to With respect to physical and anatomical wood properties, we
be the highest compared to all other tree species in this study found negative correlations for bark ratio and presence of distinct
(Fig. 1). The decay rates of 10 tree species studied in common with heartwood with decay rate. The latter was an important predictor
Pietsch et al. (2014) correlated well between both studies in our multiple regression analysis. Thicker bark with its high
(r2 = 0.49, p = 0.02). This is additionally supported by a strong lin- amount of hydrophobic suberin can decrease the decay rate by
ear correlation (r2 = 0.59, p = 0.002) of our data with CO2 emission protecting wood from the colonization by fungi and insects
rates in 2012 (Kahl et al., 2015), which can be seen as a proxy for (Freschet et al., 2012; Graça, 2015). Heartwood is known to contain
actual decay rates. The observed lower decay rates for gym- higher amounts of secondary metabolites (phenols, tannins, stilbe-
nosperms than angiosperms has already been described by nes) than sapwood. These metabolites hamper the growth of fungi
Weedon et al. (2009) and answers question one. resulting in lower decay rates (Dix and Webster, 1995). Wood ray
The presented decay rates are based on 6 years of observation in width, height and parenchyma ratio were all positively correlated
an experimental setup. This time frame is rather short in terms of with decay rate. All of these wood properties are expressions of
deadwood decay in temperate European forests where the resi- parenchyma cells in wood. These cells, on the one hand, allow
dence time (t0.95) of Fagus sylvatica is estimated to be between for radial transport of small molecules including crucial dioxygen
30 (Müller-Using and Bartsch, 2009) and 50 years (Kahl et al., (O2), low molecular compounds and nutrients in wood and could
2012). Using our decay rates, we would estimate the residence therefore be responsible for increased concentrations of these
time (t0.95 = 3/k) to range between 39 years for Carpinus betulus compounds inside a tree, which in turn could increase decay rate.
and a rather unrealistic 1901 years for Pseudotsuga menziessii On the other hand, they are essential for the process of thylosis
(Fagus sylvatica: 46 years). It is thus likely that decay rates will that is important for heartwood formation and may finally
change in the coming years. For some tree species the decomposi- decrease decay rates. Furthermore, a high ratio of ray cells will
tion process is probably still in an initial stage, or in the stage of a facilitate the colonization of wood by hyphae and improve their
lag phase, as proposed by Harmon et al. (1986) and Freschet et al. radial distribution within the xylem (Schwarze, 2007). Wood den-
(2012), in particular in Pseudotsuga menziesii. Probably some key sity was positively correlated with decay rates in our experiment,
players of the initial phase such as bark beetles are less successful for all tree species (Fig. 2), but also for the angiosperms only. The
in this non-native tree species (Roques et al., 2006). opposite was shown for 26 neotropical species in Amazonia
 T. Kahl et al. / Forest Ecology and Management 391 (2017) 86–95 93

(Hérault et al., 2010). A positive correlation of wood density with Organisms such as fungi and Coleoptera are key agents of wood
decay rate in all 13 tree species could be explained by higher resin decay. Fungal and coleopteran diversity (alpha diversity, i.e. mean
content and increased proportion of recalcitrant guaiacyl units in species richness per log and gamma diversity, i.e. rarified species
lignin in gymnosperms resulting in lower decay rates. The reasons richness across all logs) were both positively correlated with decay
for the positive relation in angiosperms remain unclear. rates. Both were selected as important variables in a multiple
With respect to chemical wood properties, all element concen- regression analysis. Saproxylic fungi and beetles have evolved
trations were positively correlated with the decay rate, except for many mutualistic commensal relationships (e.g. Hofstetter et al.,
the C content (though the interspecific variation in the C concen- 2015b), and thus their unidirectional effects are not surprising.
tration was rather small). Surprisingly, sulphur showed the stron- CO2 emission rates of these logs also support a positive correlation
gest effect on decay rate. It has been shown that sulphur addition with fungal species richness across tree species, and also indepen-
increases fungal growth in wood (Schmalenberger et al., 2011). dent of tree species (Kahl et al., 2015). Easily degradable wood
Besides N and P (Sinsabaugh et al., 1993), S, which is essential might be inhabited by a large or a small number of species, but still
for two amino acids and various biochemical cofactors, seems to may have comparable fungal and/or beetle biomass, which could
be one of the limiting macronutrients for fungal growth and hence result in similar decay rates. But obviously this is not the case,
for wood decay. We attribute the negative correlation between C and easily degradable wood tends to be inhabited rather by a larger
content and wood decay to the higher relative carbon content of number of species than by a few species. Our results are therefore
lignin as compared to cellulose or hemicelluloses. This is due to in line with studies demonstrating a positive relationship between
the high degree of lignin methoxylation and the fact that carbon species richness and decay rate which supports the view of syner-
in the aromatic rings of lignin binds less hydrogen atoms or func- gistic or facilitative interactions among species (Tiunov and Scheu,
tional groups than carbon in the pyranose rings of polysaccharides. 2005).
In other words, the variation in the carbon content coincides in Looking at the whole residence time of logs, the positive corre-
part with the variation in lignin concentration. lation between diversity and decay rate could also be an observa-
The concentrations of phenols, lignin and extractives were all tion bias due to organismic succession. Comparing two tree
negatively correlated with decay rate. Phenols and organic extrac- species, one with a low and one with a high decay rate, the species
tives such as resins are known to generally inhibit fungal growth having a low decay rate will always have a longer residence time in
(Scheffer, 1966; Hillis, 1999; Gierlinger et al., 2004). Lignin as a the forest providing substrate for organisms. At the same point in
molecular glue represents a barrier to wood decomposition (Dix time, the two tree species will be in different successional phases.
and Webster, 1995; Stokland et al., 2012) by hindering the pene- In a finite substrate such as deadwood, three successional phases
tration of enzyme molecules to the lignocellulose complex occur: initial, optimal, final with low, high and low species diver-
(Hatakka and Hammel, 2010), and as such a higher amount will sity, respectively (Stokland et al., 2012). Comparing a tree species
slow down the decay. Freschet et al. (2012) demonstrated previ- such as Carpinus betulus, which is after 6 years in the optimal phase
ously that the initial lignin content of trees is negatively correlated of succession, with Fraxinus excelsior, which is in the initial phase,
with decay rate. Moreover, we found for pH a weak positive corre- seems incorrect. A correct solution would be a comparison after
lation with decay rate, similar to Freschet et al. (2012), which the logs are fully decomposed or by comparing them at the same
might be attributed to non-structural secondary metabolites of point of relative mass loss (Arnstadt et al., 2016b). By continuing
heartwood, like organic acids, which contribute to decay resistance our experiment over the next decade we will be able to better esti-
(Harmon et al., 1986; Pearce, 1996). mate this potential bias.

4.2.2. Biotic factors 4.3. Management implications

Activities of all observed enzymes (laccase, total peroxidase,
endocellulase, xylanase) were positively correlated with decay Our results indicate that fungi, beetles and enzymes play an
rates, and were revealed to be the most important predictors of important role for wood decomposition. The detailed pathways
wood decay in our multiple regression analysis. Enzyme activities of the relationships could not be identified by the present study
can be seen as a proxy for fungal activity in deadwood. Especially and needs further investigation. Nevertheless, our results suggest
endocellulase activity, which can actually be a conglomeration of that not only particular key species, but also overall organismic
genetically diverse enzymes from different organisms including diversity is important for wood decay. An expected negative influ-
fungi, insects and their microbial symbionts (Baldrian and Valášk ence of intensified forest management on saproxylic species diver-
ová, 2008; Watanabe and Tokuda, 2010; Lombard et al., 2014), is sity therefore most likely also negatively affects the process of
a major driver of microbial decay. Given the fact that saprotrophic wood decomposition. Due to different successional pathways of
microorganisms have to gain their metabolic energy from sugars, different tree species, showing great differences in decay rates
and cellulose (b-1,4-linked glucose homopolymer) that constitutes and therefore residence times, a mix of different tree species in
up to 50% of wood, it seems rather logical to find this strong deadwood enrichment strategies might be most effective for
correlation and it as best model predictor. (Endo)xylanase, which promoting species diversity.
hydrolyzes the hemicellulose backbone (b-1,4-linked xylose
heteropolymer) contributes also substantially to the carbon Acknowledgments
(pentoses) and energy supply of microorganisms including fungi
(Polizeli et al., 2005). Further enzymes, like laccase and peroxidases We thank Angela Günther, Gerald G. Hirsch and Frank Dämm-
mediate the rather slow process of lignin decomposition and rich for the identification of selected corticoid and poroid fungi,
also contribute to the detoxification of plant ingredients Constanze Stark and Sabrina Leonhardt for help during wood
(e.g. aromatics); lignin degradation is mainly mediated by basid- sampling, Iris Gallenberger, Nadja Simons, Petra Freynhagen,
iomycetous fungi causing white-rot (Baldrian, 2006; Hatakka and Marco Lutz and student helpers for supporting the Coleoptera sam-
Hammel, 2010). However, in particular the activities of lignin- pling in the field and insect sorting in the lab and Boris Büche,
modifying enzymes were shown to be highly variable in larger Frank Köhler, Torben Kölkebeck and Thomas Wagner for identifica-
wood sampling campaigns (Arnstadt et al., 2016a; Baldrian et al., tion of Coleoptera species. We thank the managers of the three
2016; Noll et al., 2016), and thus it is remarkable to see here these Exploratories, Kirsten Reichel-Jung, Swen Renner, Katrin Hartwich,
strong and also meaningful correlations. Sonja Gockel, Kerstin Wiesner, and Martin Gorke for their work in
94 T. Kahl et al. / Forest Ecology and Management 391 (2017) 86–95

maintaining the plot and project infrastructure; Christiane Fischer Implementing large-scale and long-term functional biodiversity research: the
Biodiversity Exploratories. Basic Appl. Ecol. 11, 473–485.
and Simone Pfeiffer for giving support through the central office,
Floudas, D., Binder, M., Riley, R., Barry, K., Blanchette, R.A., Henrissat, B., Martínez, A.
Michael Owonibi for managing the central data base, and Markus T., Otillar, R., Spatafora, J.W., Yadav, J.S., Aerts, A., Benoit, I., Boyd, A., Carlson, A.,
Fischer, Eduard Linsenmair, Dominik Hessenmöller, Jens Copeland, A., Coutinho, P.M., de Vries, R.P., Ferreira, P., Findley, K., Foster, B.,
Nieschulze, Daniel Prati, Ingo Schöning, François Buscot and the Gaskell, J., Glotzer, D., Górecki, P., Heitman, J., Hesse, C., Hori, C., Igarashi, K.,
Jurgens, J.A., Kallen, N., Kersten, P., Kohler, A., Kües, U., Kumar, T.K.A., Kuo, A.,
late Elisabeth Kalko for their role in setting up the Biodiversity LaButti, K., Larrondo, L.F., Lindquist, E., Ling, A., Lombard, V., Lucas, S., Lundell, T.,
Exploratories project. The work has been funded by the DFG Prior- Martin, R., McLaughlin, D.J., Morgenstern, I., Morin, E., Murat, C., Nagy, L.G.,
ity Program 1374 ‘‘Infrastructure-Biodiversity-Exploratories” Nolan, M., Ohm, R.A., Patyshakuliyeva, A., Rokas, A., Ruiz-Dueñas, F.J., Sabat, G.,
Salamov, A., Samejima, M., Schmutz, J., Slot, J.C., St John, F., Stenlid, J., Sun, H.,
(DFG-BA 2821/9-3, DFG-WE 3081/21-1). Field work permits were Sun, S., Syed, K., Tsang, A., Wiebenga, A., Young, D., Pisabarro, A., Eastwood, D.C.,
issued by the responsible state environmental offices of Baden- Martin, F., Cullen, D., Grigoriev, I.V., Hibbett, D.S., 2012. The Paleozoic origin of
Württemberg, Thüringen, and Brandenburg (according to § 72 enzymatic lignin decomposition reconstructed from 31 fungal genomes.
Science 336, 1715–1719.
BbgNatSchG). Freschet, G.T., Weedon, J.T., Aerts, R., van Hal, J.R., Cornelissen, J.H.C., 2012.
Interspecific differences in wood decay rates: insights from a new short-term
method to study long-term wood decomposition. J. Ecol. 100, 161–170.
Appendix A. Supplementary material Fukami, T., Dickie, I.A., Paula Wilkie, J., Paulus, B.C., Park, D., Roberts, A., Buchanan, P.
K., Allen, R.B., 2010. Assembly history dictates ecosystem functioning: evidence
from wood decomposer communities. Ecol. Lett. 13, 675–684.
Supplementary data associated with this article can be found, in Gardes, M., Bruns, T.D., 1993. ITS primers with enhanced specificity for
the online version, at http://dx.doi.org/10.1016/j.foreco.2017.02. basidiomycetes – application to the identification of mycorrhizae and rusts.
Mol. Ecol. 2, 113–118.
012. Gierlinger, N., Jacques, D., Schwanninger, M., Wimmer, R., Pâques, L.E., 2004.
Heartwood extractives and lignin content of different larch species (Larix sp.)
and relationships to brown-rot decay-resistance. Trees 18, 230–236.
References Gleeson, D.B., Clipson, N., Melville, K., Gadd, G.M., McDermott, F.P., 2005.
Characterization of fungal community structure on a weathered pegmatitic
granite. Microb. Ecol. 50, 360–368.
Aloui, F., Ayadi, N., Charrier, F., Charrier, B., 2004. Durability of European oak
Gossner, M.M., Wende, B., Levick, S., Schall, P., Floren, A., Linsenmair, K.E., Steffan-
(Quercus petraea and Quercus robur) against white rot fungi (Coriolus
Dewenter, I., Schulze, E.-D., Weisser, W.W., 2016. Deadwood enrichment in
versicolor): relations with phenol extractives. Holz als Roh- und Werkstoff
European forests – which tree species should be used to promote saproxylic
62, 286–290.
beetle diversity? Biol. Conserv. 201, 92–102.
Arnstadt, T., Hoppe, B., Kahl, T., Kellner, H., Krüger, D., Bässler, C., Bauhus, J.,
Graça, J., 2015. Suberin: the biopolyester at the frontier of plants. Front. Chem. 3.
Hofrichter, M., 2016a. Patterns of laccase and peroxidases in coarse woody
Green, J.L., Holmes, A.J., Westoby, M., Oliver, I., Briscoe, D., Dangerfield, M., Gillings,
debris of Fagus sylvatica, Picea abies and Pinus sylvestris and their relation to
M., Beattie, A.J., 2004. Spatial scaling of microbial eukaryote diversity. Nature
different wood parameters. Eur. J. For. Res. 135, 109–124.
432, 747–750.
Arnstadt, T., Hoppe, B., Kahl, T., Kellner, H., Krüger, D., Bauhus, J., Hofrichter, M.,
Grueber, C.E., Nakagawa, S., Laws, R.J., Jamieson, I.G., 2011. Multimodel inference in
2016b. Dynamics of fungal community composition, decomposition and
ecology and evolution: challenges and solutions. J. Evol. Biol. 24, 699–711.
resulting deadwood properties in logs of Fagus sylvatica, Picea abies and Pinus
Halme, P., Kotiaho, J.S., 2012. The importance of timing and number of surveys in
sylvestris. For. Ecol. Manage. 382, 129–142.
fungal biodiversity research. Biodivers. Conserv. 21, 205–219.
Baber, K., Otto, P., Kahl, T., Gossner, M.M., Wirth, C., Gminder, A., Bässler, C., 2016.
Harmon, M., Franklin, J., Swanson, F., Sollins, P., Gregory, V., Lattin, J., Anderson, N.,
Disentangling the effects of forest-stand type and dead-wood origin of the early
Cline, S., Aumen, N., Lienkaemper, G., Cromack, K., Cummins, K., 1986. Ecology
successional stage on the diversity of wood-inhabiting fungi. For. Ecol. Manage.
of coarse woody debris in temperate ecosystems. Adv. Ecol. Res. 15, 133–302.
377, 161–169.
Hart, J.H., Hillis, W.E., 1972. Inhibition of wood-rotting fungi by Ellagitannins in the
Baldrian, P., 2006. Fungal laccases-occurrence and properties. FEMS Microbiol. Rev.
heartwood of Quercus alba. Phytopathology 62, 620–626.
30, 215–242.
Hatakka, A., Hammel, K.E., 2010. Fungal biodegradation of lignocelluloses. In:
Baldrian, P., Valášková, V., 2008. Degradation of cellulose by basidiomycetous fungi.
Osiewacz, H.D. (Ed.), The Mycota: A Comprehensive Treatise on Fungi as
FEMS Microbiol. Rev. 32, 501–521.
Experimental Systems for Basic and Applied Research. Springer, pp. 319–340.
Baldrian, P., Zrůstová, P., Tláskal, V., Davidová, A., Merhautová, V., Vrška, T., 2016.
Heibl, C., 2014. megaptera: MEGAPhylogeny Techniques in R. R Package Version
Fungi associated with decomposing deadwood in a natural beech-dominated
1.0-0. <https://CRAN.R-project.org/package=megaptera>.
forest. Fungal Ecol. 23, 109–122.
Hérault, B., Beauchêne, J., Muller, F., Wagner, F., Baraloto, C., Blanc, L., Martin, J.-M.,
Bantle, A., Borken, W., Ellerbrock, R.H., Schulze, E.D., Weisser, W.W., Matzner, E.,
2010. Modeling decay rates of dead wood in a neotropical forest. Oecologia 164,
2014. Quantity and quality of dissolved organic carbon released from coarse
243–251.
woody debris of different tree species in the early phase of decomposition. For.
Herrmann, S., Kahl, T., Bauhus, J., 2015. Decomposition dynamics of coarse woody
Ecol. Manage. 329, 287–294.
debris of three important central European tree species. For. Ecosyst. 2, 1–14.
Bartoń, K., 2016. Package ‘MuMIn’: Multi-Model Inference. R-Package 1.15.6
Higuchi, T., 2006. Look back over the studies of lignin biochemistry. J. Wood Sci. 52,
<https://CRAN.R-project.org/package=MuMIn>.
2–8.
Bauhus, J., Puettmann, K., Messier, C., 2009. Silviculture for old-growth attributes.
Hillis, W.E., 1999. The formation of heartwood and its extractives. In: Romeo, J.T.
For. Ecol. Manage. 258, 525–537.
(Ed.), Phytochemicals in Human Health Protection, Nutrition, and Plant Defense.
Boddy, L., 2001. Fungal community ecology and wood decomposition processes in
Springer, Boston, MA, US, pp. 215–253.
angiosperms: from standing tree to complete decay of coarse woody debris.
Ho, L.S.T., Ane, C., 2014. A linear-time algorithm for Gaussian and non-Gaussian trait
Ecol. Bull., 43–56
evolution models. Syst. Biol. 63, 397–408.
Box, J.D., 1983. Investigation of the Folin-Ciocalteau phenol reagent for the
Hofstetter, R.W., Dinkins-Bookwalter, J., Davies, T.S., Klepzig, K.D., 2015a.
determination of polyphenolic substances in natural waters. Water Res. 17,
Symbiontic associations of bark beetles. In: Vega, F.E., Hofstetter, R.W. (Eds.),
511–525.
Bark Beetles. Academic Press, San Diego, pp. 209–245 (Chapter 6).
Brunow, G., 2001. Methods to reveal the structure of lignin. Lignin, humic
Hofstetter, R.W., Dinkins-Bookwalter, J., Davis, T.S., Klepzig, K.D., 2015b. Symbiotic
substances and coal. In: Hofrichter, M., Steinbüchel, A. (Eds.), Biopolymers,
associations of bark beetles. In: Vega, F.E., Hofstetter, R.W. (Eds.), Bark Beetles.
Lignin, Humic Substances and Coal, vol. 1. Wiley-VCH Verlag GmbH & Co.,
Academic Press, San Diego, pp. 209–245 (Chapter 6).
Weinheim, pp. 89–99.
Høiland, K., Bendiksen, E., 1996. Biodiversity of wood-inhabiting fungi in a boreal
Burnham, K.P., Anderson, D.R., Huyvaert, K.P., 2011. AIC model selection and
coniferous forest in Ser-Trendelag County, Central Norway. Nord. J. Bot. 16,
multimodel inference in behavioral ecology: some background, observations,
643–659.
and comparisons. Behav. Ecol. Sociobiol. 65, 23–35.
Hoppe, B., Purahong, W., Wubet, T., Kahl, T., Bauhus, J., Arnstadt, T., Hofrichter, M.,
Christensen, O., 1977. Estimation of standing crop and turnover of dead wood in a
Buscot, F., Krüger, D., 2015. Linking molecular deadwood-inhabiting fungal
Danish oak forest. Oikos 28, 177–186.
diversity and community dynamics to ecosystem functions and processes in
Cornwell, W.K., Cornelissen, J.H.C., Allison, S.D., Bauhus, J., Eggleton, P., Preston, C.
Central European forests. Fungal Divers., 1–13
M., Scarff, F., Weedon, J.T., Wirth, C., Zanne, A.E., 2009. Plant traits and wood
Kahl, T., 2008. Kohlenstofftransport aus dem Totholz in den Boden. In, Fakultät für
fates across the globe: rotted, burned, or consumed? Glob. Change Biol. 15,
Forst- und Umweltwissenschaften. Albert-Ludwigs-Universität Freiburg im
2431–2449.
Breisgau, Freiburg im Breisgau, p. 108.
Dix, N.J., Webster, J., 1995. Fungal Ecology. Springer, Wallington, Surrey.
Kahl, T., Baber, K., Otto, P., Wirth, C., Bauhus, J., 2015. Drivers of CO2 emission rates
EN350-2, 1994. Durability of Wood and Wood-based Products – Natural Durability
from dead wood logs of 13 tree species in the initial decomposition phase.
of Solid Wood. Guide to Natural Durability and Treatability of Selected Wood
Forests 6, 2484.
Species of Importance in Europe.
Kahl, T., Mund, M., Bauhus, J., Schulze, E., 2012. Dissolved organic carbon from
Fischer, M., Bossdorf, O., Gockel, S., Hansel, F., Hemp, A., Hessenmoller, D., Korte, G.,
European beech logs: patterns of input to and retention by surface soil.
Nieschulze, J., Pfeiffer, S., Prati, D., Renner, S., Schoning, I., Schumacher, U., Wells,
Ecoscience 19, 1–10.
K., Buscot, F., Kalko, E.K.V., Linsenmair, K.E., Schulze, E.D., Weisser, W.W., 2010.
 T. Kahl et al. / Forest Ecology and Management 391 (2017) 86–95 95

Kellner, H., Luis, P., Pecyna, M.J., Barbi, F., Kapturska, D., Krüger, D., Zak, D.R., Blanchette, Henrissat, B., Martin, F., Cullen, D., Hibbett, D.S., Grigoriev, I.V., 2014.
Marmeisse, R., Vandenbol, M., Hofrichter, M., 2014. Widespread occurrence of Extensive sampling of basidiomycete genomes demonstrates inadequacy of the
expressed fungal secretory peroxidases in forest soils. PLoS ONE 9, e95557. white-rot/brown-rot paradigm for wood decay fungi. Proc. Natl. Acad. Sci. 111,
Kögel-Knabner, I., 2002. The macromolecular organic composition of plant and 9923–9928.
microbial residues as inputs to soil organic matter. Soil Biol. Biochem. 34, 139– Rondeux, J., Sanchez, C., 2010. Review of indicators and field methods for
162. monitoring biodiversity within national forest inventories. Core variable:
Kruys, N., Jonnsson, B., Stahl, G., 2002. A stage-based matrix model for decay-class Deadwood. Environ. Monit. Assess. 164, 617–630.
dynamics of woody debris. Ecol. Appl. 12, 773–781. Roques, A., Auger-Rozenberg, M.-A., Boivin, S., 2006. A lack of native congeners may
Kubartová, A., Ottosson, E., Dahlberg, A., Stenlid, J., 2012. Patterns of fungal limit colonization of introduced conifers by indigenous insects in Europe. Can. J.
communities among and within decaying logs, revealed by 454 sequencing. For. Res. 36, 299–313.
Mol. Ecol. 21, 4514–4532. Scheffer, T.C., 1966. Natural resistance of wood to microbial deterioration. Annu.
Lindner, D.L., Vasaitis, R., Kubartová, A., Allmér, J., Johannesson, H., Banik, M.T., Rev. Phytopathol. 4, 147–168.
Stenlid, J., 2011. Initial fungal colonizer affects mass loss and fungal community Schmalenberger, A., Pritzkow, W., Ojeda, J.J., Noll, M., 2011. Characterization of main
development in Picea abies logs 6 yr after inoculation. Fungal Ecol. 4, 449–460. sulfur source of wood-degrading basidiomycetes by S K-edge X-ray absorption
Lombard, V., Golaconda, R.H., Drula, E., Coutinho, P.M., Henrissat, B., 2014. The near edge spectroscopy (XANES). Int. Biodeterior. Biodegrad. 65, 1215–1223.
carbohydrate-active enzymes database (CAZy) in 2013. Nucleic Acids Res. 42, Schwarze, F.W.M.R., 2007. Wood decay under the microscope. Fungal Biol. Rev. 21,
D490–D495. 133–170.
Lonsdale, D., Pautasso, M., Holdenrieder, O., 2008. Wood-decaying fungi in the Seibold, S., Brandl, R., Buse, J., Hothorn, T., Schmidl, J., Thorn, S., Müller, J., 2015.
forest: conservation needs and management options. Eur. J. For. Res. 127, 1–22. Association of extinction risk of saproxylic beetles with ecological degradation
Müller-Using, S., Bartsch, N., 2009. Decay dynamic of coarse and fine woody debris of forests in Europe. Conserv. Biol. 29, 382–390.
of a beech (Fagus sylvatica L.) forest in Central Germany. Eur. J. For. Res. 128, Setälä, H., McLean, M.A., 2004. Decomposition rate of organic substrates in relation
287–296. to the species diversity of soil saprophytic fungi. Oecologia 139, 98–107.
Müller, J., Wende, B., Strobl, C., Eugster, M., Gallenberger, I., Floren, A., Steffan- Sinsabaugh, R.L., Antibus, R.K., Linkins, A.E., McClaugherty, C.A., Rayburn, L., Repert,
Dewenter, I., Linsenmair, K.E., Weisser, W.W., Gossner, M.M., 2015. Forest D., Weiland, T., 1993. Wood decomposition: nitrogen and phosphorus dynamics
management and regional tree composition drive the host preference of in relation to extracellular enzyme activity. Ecology 74, 1586–1593.
saproxylic beetle communities. J. Appl. Ecol. 52, 753–762. Stamatakis, A., 2014. RAxML version 8: a tool for phylogenetic analysis and post-
Noll, L., Leonhardt, S., Arnstadt, T., Hoppe, B., Poll, C., Matzner, E., Hofrichter, M., analysis of large phylogenies. Bioinformatics 30, 1312–1313.
Kellner, H., 2016. Fungal biomass and extracellular enzyme activities in coarse Stokland, J.N., Siitonen, J., Jonsson, B.G., 2012. Biodiversity in Dead Wood.
woody debris of 13 tree species in the early phase of decomposition. For. Ecol. Cambridge University Press, Cambridge.
Manage. 378, 181–192. Suh, S.-O., McHugh, J.V., Pollock, D.D., Blackwell, M., 2005. The beetle gut: a
Olson, J., 1963. Energy storage and the balance of producers and decomposers in hyperdiverse source of novel yeasts. Mycol. Res. 109, 261–265.
ecological systems. Ecology 44, 322–331. Swift, M.J., Healey, I.N., Hibberd, J.K., Sykes, J.M., Bampoe, V., Nesbitt, M.E., 1976.
Paradis, E., Claude, J., Strimmer, K., 2004. APE: analyses of phylogenetics and Decomposition of branch-wood in canopy and floor of a mixed deciduous
evolution in R language. Bioinformatics 20, 289–290. woodland. Oecologia 26, 139–149.
Pattengale, N.D., Alipour, M., Bininda-Emonds, O.R.P., Moret, B.M.E., Stamatakis, A., Tabchnick, B.G., Fidell, L.S., 1996. Using Multivariate Statistics. Harper Collins, New
2010. How many bootstrap replicates are necessary? J. Comput. Biol. 17, 337– York.
354. Tiunov, A.V., Scheu, S., 2005. Facilitative interactions rather than resource
Pearce, R.B., 1996. Antimicrobial defences in the wood of living trees. New Phytol. partitioning drive diversity-functioning relationships in laboratory fungal
132, 203–233. communities. Ecol. Lett. 8, 618–625.
Pietsch, K.A., Ogle, K., Cornelissen, J.H.C., Cornwell, W.K., Bönisch, G., Craine, J.M., Toljander, J.F., Eberhardt, U., Toljander, Y.K., Paul, L.R., Taylor, A.F.S., 2006. Species
Jackson, B.G., Kattge, J., Peltzer, D.A., Penuelas, J., Reich, P.B., Wardle, D.A., composition of an ectomycorrhizal fungal community along a local nutrient
Weedon, J.T., Wright, I.J., Zanne, A.E., Wirth, C., 2014. Global relationship of gradient in a boreal forest. New Phytol. 170, 873–884.
wood and leaf litter decomposability: the role of functional traits within and Ulyshen, M.D., 2016. Wood decomposition as influenced by invertebrates. Biol. Rev.
across plant organs. Glob. Ecol. Biogeogr. 23, 1046–1057. 91, 70–85.
Polizeli, M.L.T.M., Rizzatti, A.C.S., Monti, R., Terenzi, H.F., Jorge, J.A., Amorim, D.S., Valaskova, V., de Boer, W., Klein Gunnewiek, P.J.A., Pospisek, M., Baldrian, P., 2009.
2005. Xylanases from fungi: properties and industrial applications. Appl. Phylogenetic composition and properties of bacteria coexisting with the fungus
Microbiol. Biotechnol. 67, 577–591. Hypholoma fasciculare in decaying wood. ISME J. 3, 1218–1221.
Purahong, W., Arnstadt, T., Kahl, T., Bauhus, J., Kellner, H., Hofrichter, M., Krüger, D., Valentín, L., Rajala, T., Peltoniemi, M., Heinonsalo, J., Pennanen, T., Mäkipää, R., 2014.
Buscot, F., Hoppe, B., 2016. Are correlations between deadwood fungal Loss of diversity in wood-inhabiting fungal communities affects decomposition
community structure, wood physico-chemical properties and lignin- activity in Norway spruce wood. Front. Microbiol. 5, 1–11.
modifying enzymes stable across different geographical regions? Fungal Ecol. van der Wal, A., Geydan, T.D., Kuyper, T.W., de Boer, W., 2013. A thready affair:
22, 98–105. linking fungal diversity and community dynamics to terrestrial decomposition
Purahong, W., Hoppe, B., Kahl, T., Schloter, M., Schulze, E.-D., Bauhus, J., Buscot, F., processes. FEMS Microbiol. Rev. 37, 477–494.
Kruger, D., 2014a. Changes within a single land-use category alter microbial van der Wal, A., Ottosson, E., de Boer, W., 2015. Neglected role of fungal community
diversity and community structure: molecular evidence from wood-inhabiting composition in explaining variation in wood decay rates. Ecology 96, 124–133.
fungi in forest ecosystems. J. Environ. Manage. 139, 109–119. Vetrovsky, T., Voriskova, J., Snajdr, J., Gabriel, J., Baldrian, P., 2011. Ecology of coarse
Purahong, W., Kahl, T., Schloter, M., Bauhus, J., Buscot, F., Krüger, D., 2014b. wood decomposition by the saprotrophic fungus Fomes fomentarius.
Comparing fungal richness and community composition in coarse woody debris Biodegradation 22, 709–718.
in Central European beech forests under three types of management. Mycol. Větrovský, T., Voříšková, J., Šnajdr, J., Gabriel, J., Baldrian, P., 2011. Ecology of coarse
Prog., 1–6 wood decomposition by the saprotrophic fungus Fomes fomentarius.
R Core Team, 2011. R: A Language and Environment for Statistical Computing. R Biodegradation 22, 709–718.
Foundation for Statistical Computing, Vienna, Austria. Wagenführ, R., 2006. Holzatlas. Fachbuchverlag Leipzig im Carl Hanser Verlag,
Ramette, A., 2009. Quantitative community fingerprinting methods for estimating Leipzig.
the abundance of operational taxonomic units in natural microbial Watanabe, H., Tokuda, G., 2001. Animal cellulases. Cell. Mol. Life Sci.: CMLS 58,
communities. Appl. Environ. Microbiol. 75, 2495–2505. 1167–1178.
RCoreTeam, 2015. R: A Language and Environment for Statistical Computing. R Watanabe, H., Tokuda, G., 2010. Cellulolytic systems in insects. Annu. Rev. Entomol.
Foundation for Statistical Computing, Vienna, Austria. 55, 609–632.
Renvall, P., 1995. Community structure and dynamics of wood-rotting Weedon, J.T., Cornwell, W.K., Cornelissen, J.H.C., Zanne, A.E., Wirth, C., Coomes, D.A.,
Basidiomycetes on decomposing conifer trunks in northern Finland. Karstenia 2009. Global meta-analysis of wood decomposition rates: a role for trait
35, 1–51. variation among tree species? Ecol. Lett. 12, 45–56.
Revell, L.J., 2012. Phytools: an R package for phylogenetic comparative biology (and White, T.J., Bruns, T.D., Lee, S., Taylor, J., 1990. Amplification and direct sequencing
other things). Methods Ecol. Evol. 3, 217–223. of fungal ribosomal DNA genes for phylogenies. In: Innis, M.A., Gelfand, D.H.,
Riley, R., Salamov, A.A., Brown, D.W., Nagy, L.G., Floudas, D., Held, B.W., Levasseur, Sninsky, J.J., White, T.J. (Eds.), PCR Protocols: A Guide to Methods and
A., Lombard, V., Morin, E., Otillar, R., Lindquist, E.A., Sun, H., LaButti, K.M., Applications. Academic Press, San Diego, pp. 315–322.
Schmutz, J., Jabbour, D., Luo, H., Baker, S.E., Pisabarro, A.G., Walton, J.D.,