Oecologia

, Volume 167, Issue 4, pp 1177–1184

Simulated nitrogen deposition affects wood decomposition by cord-forming fungi

Authors

    • Earthwatch Institute
  • Sarah C. Watkinson
    • Department of Plant SciencesUniversity of Oxford
  • Lynne Boddy
    • Cardiff School of BiosciencesCardiff University
  • Peter R. Darrah
    • Department of Plant SciencesUniversity of Oxford
Global change ecology - Original Paper

DOI: 10.1007/s00442-011-2057-2

Cite this article as:
Bebber, D.P., Watkinson, S.C., Boddy, L. et al. Oecologia (2011) 167: 1177. doi:10.1007/s00442-011-2057-2

Abstract

Anthropogenic nitrogen (N) deposition affects many natural processes, including forest litter decomposition. Saprotrophic fungi are the only organisms capable of completely decomposing lignocellulosic (woody) litter in temperate ecosystems, and therefore the responses of fungi to N deposition are critical in understanding the effects of global change on the forest carbon cycle. Plant litter decomposition under elevated N has been intensively studied, with varying results. The complexity of forest floor biota and variability in litter quality have obscured N-elevation effects on decomposers. Field experiments often utilize standardized substrates and N-levels, but few studies have controlled the decay organisms. Decomposition of beech (Fagus sylvatica) blocks inoculated with two cord-forming basidiomycete fungi, Hypholoma fasciculare and Phanerochaete velutina, was compared experimentally under realistic levels of simulated N deposition at Wytham Wood, Oxfordshire, UK. Mass loss was greater with P. velutina than with H. fasciculare, and with N treatment than in the control. Decomposition was accompanied by growth of the fungal mycelium and increasing N concentration in the remaining wood. We attribute the N effect on wood decay to the response of cord-forming wood decay fungi to N availability. Previous studies demonstrated the capacity of these fungi to scavenge and import N to decaying wood via a translocating network of mycelium. This study shows that small increases in N availability can increase wood decomposition by these organisms. Dead wood is an important carbon store and habitat. The responses of wood decomposers to anthropogenic N deposition should be considered in models of forest carbon dynamics.

Keywords

Basidiomycete fungiCarbon cycleForest litter decompositionWood decayNitrogen cycle

Introduction

Anthropogenic nitrogen (N) deposition now rivals natural N inputs, and this has had widespread effects on terrestrial and aquatic ecosystems (Vitousek et al. 1997; Fenn et al. 1998; Langham 1999). One of the most intensively-studied aspects is the effect of atmospheric N deposition on litter decay in forests (Berg and Laskowski 2005), as alterations in decomposition rates will effect forest carbon sequestration. It has long been recognized that the carbon/nitrogen (C/N) ratio of forest litter is far greater than microbial biomass, and decomposition rates are generally greater in plant materials with low C/N ratios (Swift et al. 1979; Melillo et al. 1982). Leaf litter decomposition often increases along N-deposition gradients in terrestrial (Fenn and Dunn 1989; Kuperman 1999) and aquatic environments (Gulis et al. 2004; Ferreira et al. 2006). These observations imply that litter decomposition is N-limited. However, many other studies contradict the theory of N-limitation in decay micro-organisms. Contrary to the expectation of the N-limitation theory, the vast majority of results show either no effect, or a negative effect, of N on decomposition rate, particularly in litters with high C/N ratio (Fog 1988; Knorr et al. 2005; Berg and Laskowski 2005).

These contradictory observations may be due to a number of factors, including the type of litter, the source of N in the system, and the taxa of micro-organisms involved in decomposition. The two major types of forest litter are leaf litter and woody litter (Table 1). These litter types differ greatly in C/N ratio (Table 1), and different communities of fungi specialize on these two resources. The decay of woody, lignin-rich litter is largely due to basidiomycete fungi known as ‘white rots’, which are highly efficient in utilizing and recycling N, and are adapted to low N environments (Watkinson et al. 2006). White rots synthesize lignolytic enzymes during N starvation, breaking down the lignin matrix to allow access to cellulosic material and proteins within the wood (Kirk and Farrell 1987; Rayner and Boddy 1988; Eriksson et al. 1990). The suppression of woody litter decay by N appears to be due to the enzymatic responses of white rot fungi that degrade the lignin matrix: if sufficient N is present, then access to further resources in the wood is not required (Kirk and Farrell 1987; Carreiro et al. 2000). Another possibility is reduced competitiveness of white rots compared with other microbes (Carreiro et al. 2000; Waldrop et al. 2004).
Table 1

Summary of leaf and coarse woody litter composition and dynamics in temperate forest ecosystems

Measure

Tree type

Leaf littera

n

Woody litterb

n

Nitrogen (mg g−1)

Gymnosperm

5.5 (4.3–7.8)a

8

0.7 (NA)b

11

Angiosperm

11.0 (5.1–30.1)a

7

1.2 (NA)b

11

Lignin (mg g−1)

Gymnosperm

288 (249–381)a

5

300 (260–340)b

16

Angiosperm

296 (264–330)a

3

240 (160–320)b

10

Input rate (Mg ha−1 year−1)

Gymnosperm

2.9 (1.2–4.4)a

26

2.0 (0.3–30)b

18

Angiosperm

3.4 (2.4–5.2)a

20

0.9 (0.0–14.5)b

14

Mass (Mg ha−1)

Gymnosperm

27.0 (2.9–188.4)c

91

115.0 (3–490)b

41

Angiosperm

19.0 (2.1–102.9)c

55

22.0 (7.5–49.3)b

20

Half life (year)

Gymnosperm

NA

 

69.7 (3.5–172)b

15

Angiosperm

NA

 

9.1 (2.3–24)b

7

Mean residence time (year)

Gymnosperm

10.0 (1.0–67.0)

90

NA

 

Angiosperm

3.7 (0.3–33.0)

55

NA

 

Data are medians with ranges in parentheses

n The number of data points in the source literaturea,b

aData from Berg and Laskowski (2005)

bData from Harmon et al. (1987)

cData from Vogt et al. (1987). Includes leaf litter and fine woody litter <1 cm diameter

While there are numerous studies on the effects of N on decay of leaf litter and carbon compounds in the field and laboratory (Fog 1988; Knorr et al. 2005; Berg and Laskowski 2005), there is remarkably little information on the role of N deposition in the decay of wood in forests. Experimental studies on wood splints and chips have produced variable results (Downs et al. 1996; Micks et al. 2004; Hobbie 2005). A meta-analysis has shown that wood decomposition rate increases with wood N content (Weedon et al. 2009), but the effect of atmospheric N deposition is essentially unknown.

Given the known differences in the composition, dynamics, and microbial communities between leaf litter and dead wood, and the lack of experimental data on wood decay under N-deposition in the field, we conducted an experiment to test whether N-deposition affects: wood decomposition, as measured by mass loss; the size of the fungal mycelium, as measured by mycelial mass and extension from wood resource; and N content of the mycelium, wood, and underlying soil.

Materials and methods

Fungal cultures

Hypholoma fasciculare (Huds.) Quél. and Phanerochaete velutina (DC.) Parmasto were maintained on 2% malt extract (Oxoid, Basingstoke, UK) agar (Agar No. 3; Oxoid) in a dark room at 19°C. Both H. fasciculare and P. velutina are white rots, with the ability to decompose both lignin and cellulosic material, and have been used in previous studies of wood decomposition in forests (Dowson et al. 1988). The species both form corded mycelia, but differ in their development, with H. fasciculare producing a denser mycelium than P. velutina (Bolton and Boddy 1993; Dowson et al. 1989). Beech wood was obtained from freshly cut trees (Bagley Wood, Oxfordshire, UK) and cut into 2 × 2 × 2 cm (mean dry density 0.66 ± 0.004 g cm−3, n = 270) blocks. Blocks were sterilized via γ-irradiation (Isotron, Swindon, UK).

Study site

Field work was conducted at Wytham Wood, Oxfordshire, UK (51°47′N, 1°20′W) in 2005. Wytham Woods is a mixed woodland of 400 ha. The wood is situated on a small hill rising to 165 m a.s.l. from the surrounding plain at 60 m a.s.l. The soil is a thin band of sandstone, with patches of sandy Frilford series soil. The top of the hill is composed of coral rag limestone covered by extremely thin, well-drained soils of Sherborne and Morton series. Mean annual precipitation between March 1992 and March 2006 was 706 mm, mean monthly relative humidity 82%, mean daily temperature minima ranged between ~0 and 14°C in winter and summer, respectively, and mean daily temperature maxima ranged between ~5 and 23°C in winter and summer, respectively [Natural Environment Research Council Environmental Change Network (NERC ECN) data]. Mean N deposition between May 1993 and April 2005 was 2.9 kg ha−1 year−1 as NH4+, and 0.7 kg ha−1 year−1 as NO3 (ECN data). Mean shallow soil (10 cm depth) NO3 concentration is 4.0 mg l−1, NH4+ is 0.038 mg l−1, pH (ECN data), Ca+ 22.3 mg l−1, CaCO3 20.9 22.3 mg l−1, and pH 7.2. Humus is mull form.

There are three vegetation types of roughly equal area: ancient semi-natural (NVC classification W8; Rodwell 1991) Fraxinus excelsiorAcer campestre—Mercurialis perennis); secondary (also NVC W8); and plantations of beech (Fagus sylvatica L.), the majority of which were planted 40–50 years ago. Field experiments were conducted in these beech plantations.

Wood decomposition

Sterilized beech wood blocks were inoculated with either H. fasciculare or P. velutina by placing onto fungal mycelium growing on 2% malt agar and leaving for 2 weeks. Four beech plantation sites were selected in Wytham Wood, separated from one another by a minimum distance of 200 m. The experiment was a randomized block design, with ten replicates per fungal species per N treatment per site. In February 2005, a grid with 4-m spacing between sample points were laid out in each block. At each grid location, an inoculated wood block was placed beneath the leaf litter on the soil surface, and pressed into the soil 1–2 mm. The N treatment was the addition of 50 mg NaNO3 in 1.0 l deionized water in a 0.5-m2 area around blocks, once per month for 10 months. This was the equivalent to an additional N load of 2.8 kg N ha−1 year−1. The control was the addition of 1.0 l of deionized water over the equivalent area. The position of the blocks was marked with a plastic flag.

After 10 months, the wood blocks were carefully exposed by removing overlying litter. Any mycelial network was carefully exposed, and the extension of mycelial cords in each of four quadrants around the wood block was measured (recording zero if no cords were present). The cords and wood blocks were then removed, cleaned of attached soil and organic matter, freeze-dried, weighed, ground to fine powder, and analyzed for total N (NERC Stable Isotopes Facility, Centre for Ecology and Hydrology, Lancaster, UK). Only blocks that had been inoculated with P. velutina were analyzed, because only these had developed sufficient mycelium.

Four soil samples were taken at each grid point at which P. velutina blocks had been placed. Surface litter was removed, and 5-cm-deep cores, 2.5 cm in diameter, were taken, located 0, 5, 10 and 20 cm from where the wood block had been placed. The samples were then dried at 105°C, sieved to 2-mm particle diameter, bulked to give one sample per wood block, and analyzed for total N (SOYL, Berkshire, UK).

Results

Wood decomposition

The experiment was analyzed as a randomized block design, i.e. site was included as a factor in all models. Effects of N treatment, and differences between the two species of fungi, were detected (Fig. 1). Mass loss was greater for P. velutina (F1,154 = 13.2, p = 0.0004) and for N treatment plots (F1,154 = 8.9, p = 0.003), but did not differ significantly among sites (F3,154 = 1.0, p = 0.4). More wood blocks decomposed completely (i.e., none of the wood block remained) under the N-treatment than in controls (Z = 3.28, p = 0.0012), but there was no difference between species (Z = 0.39, p = 0.70).
https://static-content.springer.com/image/art%3A10.1007%2Fs00442-011-2057-2/MediaObjects/442_2011_2057_Fig1_HTML.gif
Fig. 1

Mean wood mass loss (%) by species and treatment. Points show fitted values from linear model for Site 1, bars show 95% confidence intervals. H is Hypholoma fasciculare, P is Phanerochaete velutina, C control, T treatment

Mycelial growth

The mycelium of H. fasciculare formed fine mats interspersed within decaying leaves, and could not be successfully extracted from this matrix. P. velutina mycelial cords grew between the soil-leaf litter interface (Online Resource 1), and could be harvested successfully except at the outer margins where it became thinner, more diffuse, and fragile. Because the mycelium of H. fasciculare was impossible to harvest, further analyses were conducted using only those wood blocks inoculated with P. velutina.

Cord system extension in P. velutina was recorded for 50 sample points, and cord system mass for 48 sample points, the other sample points having no mycelium or cord systems present that could be confidently associated with the experiment. Cord systems of P. velutina extended to a maximum radius of 50.1 ± 2.2 cm from wood blocks, though mean extension derived from four quadrants was lower (Table 2). P. velutina cord system mass ranged between 31 and 222 mg (Table 2), and was correlated with mean extension (r = 0.44, df = 42, p = 0.003). Greater cord system mass and extension were correlated with wood block mass loss (Fig. 2). There was no significant effect of N treatment on mycelium mass (F1,43 = 0.11, p = 0.74).
Table 2

Summary statistics for wood blocks and mycelium

 

Control

 

Nitrogen

 
 

Mean ± SE

n

Mean ± SE

n

Wood mass loss, H. fasciculare (%)

67 ± 3

40

73 ± 4

40

Wood mass loss, P. velutina (%)

79 ± 2

40

87 ± 2

40

Complete decomposition, H. fasciculare (%)

12.5

40

32.5

40

Complete decomposition, P. velutina (%)

12.5

40

37.5

40

Wood block nitrogen (% dry mass)

0.72 ± 0.06

35

0.77 ± 0.07

24

Wood block nitrogen (mg)

7.0 ± 0.4

34

6.8 ± 1.2

23

Mycelium nitrogen (% dry mass)

1.5 ± 0.1

25

1.7 ± 0.1

26

Mycelium nitrogen (mg)

2.0 ± 0.2

23

2.5 ± 0.3

25

Mycelium mass (mg)

132 ± 7

23

140 ± 10

25

Mycelium extent, excl. zero, mean (cm)

25.8 ± 2.7

25

28.3 ± 2.7

25

Mycelium extent, incl. zero, mean (cm)

16.1 ± 2.6

40

17.7 ± 2.8

40

Soil nitrogen (% dry mass)

0.27 ± 0.02

25

0.31 ± 0.06

23

See text for statistical analyses of differences

https://static-content.springer.com/image/art%3A10.1007%2Fs00442-011-2057-2/MediaObjects/442_2011_2057_Fig2_HTML.gif
Fig. 2

Mycelium growth versus wood decay. a Mycelium mass versus wood mass loss for P. velutina (Pearson correlation coefficient r = 0.29, df = 46, p = 0.04). Control blocks are denoted by circles, Treatment blocks by triangles. b Mean mycelium extent versus wood mass loss (r = 0.26, df = 42, p = 0.06)

Nitrogen

Mycelial N concentration ranged between 0.45 and 2.7% dry weight, while the total N content of recovered mycelium ranged between 0.5 and 5.0 mg (Table 2). Wood block N concentration ranged between 0.24 and 1.6%, wood block N content between 1.5 and 32 mg, and soil N between 0.14 and 0.48% (Table 2). Mycelium N concentration was positively correlated with wood block N concentration (r = 0.45, df = 40, p = 0.003), but mycelial N content and wood block N content were negatively correlated (r = −0.42, df = 40, p = 0.006). The N content of the system changed as wood decomposition proceeded (Fig. 3). Wood block N concentration increased with wood decomposition (r = 0.61, df = 57, p < 0.0001), but the total mass of N in the wood blocks decreased (r = −0.76, df = 54, p < 0.0001). Mycelium N concentration and N mass increased with wood decomposition (r = 0.42 and 0.53, respectively).
https://static-content.springer.com/image/art%3A10.1007%2Fs00442-011-2057-2/MediaObjects/442_2011_2057_Fig3_HTML.gif
Fig. 3

Nitrogen content versus wood decay. a Wood N concentration versus wood mass loss for P. velutina (r = 0.61, df = 57, p = 0.000). Control blocks are denoted by circles, Treatment blocks by triangles. b Wood N mass versus wood mass loss (r = −0.76, df = 54, p = 0.000). c Mycelial N concentration versus wood mass loss (r = 0.32, df = 49, p = 0.019). d Mycelial N mass versus wood mass loss (r = 0.40, df = 46, p = 0.004)

N treatment did not significantly affect N concentration in the mycelium (F1,46 = 3.06, p = 0.087), the wood block (F1,52 = 0.005, p = 0.94), or the soil (F1,52 = 0.005, p = 0.94). There was significant variability in soil N concentration among sites, from a mean of 0.21% in Site 4 to 0.30% in Site 2.

Discussion

Using realistic N levels and a single decomposer species representative of the wood decay fungi present in the habitat, we showed enhancement of decay induced by N elevation. This differs from results obtained by previous leaf decomposition experiments, and is consistent with observations on N deposition gradients (Fenn and Dunn 1989; Kuperman 1999) and wood N concentration (Weedon et al. 2009).

One explanation for enhanced decomposition with N treatment may lie in the quantity of additional N added to the system. Early studies indicated that the responses of wood-decay fungi to nitrate can be subtle and non-linear, with decomposition increasing with low rates of N addition, but decreasing with high N availability (Findlay 1934; Schmitz and Kaufert 1936). Most recent experimental studies have added very large quantities of N, up to more than 1,000 times ambient deposition, with none less than twice the ambient rate (Knorr et al. 2005). Such large, acute additions may be uninformative of the response of decomposers to anthropogenic N deposition. Large additions are also likely to alter the decomposer community (Carreiro et al. 2000; Waldrop et al. 2004; Sinsabaugh et al. 2005; Moorhead and Sinsabaugh 2006), to an extent that bears little relationship to the effects of anthropogenic N deposition. In the present study, the additional nitrate was approximately equal to the wet N deposition for the site. Although the total nitrate addition was low compared with other studies, the concentration of nitrate in solution was approximately an order of magnitude greater than site soil nitrate concentration. The results of the present study support the idea that low rates of nitrate addition can have the opposite effect to high rates of addition. By controlling the species of wood decay fungi and inoculating standardized woodblocks, we minimized variability due to decomposer community composition and substrate properties.

It is also possible that too much emphasis has been placed on lignin metabolism. Lignins usually comprise less than a third of the biomass of plant litter (Table 1), and there is considerable variability among white rot species in the response of their lignolytic systems to N (Kirk and Farrell 1987; Eriksson et al. 1990). In contrast to lignin, cellulose decomposition increases with N availability in laboratory studies (Fog 1988), and cellulase activity is usually enhanced by N addition in the field (Carreiro et al. 2000; Saiya-Cork et al. 2002; Sinsabaugh et al. 2005).

Another reason for the unexpected effect of N fertilization may relate to background rates of N deposition at Wytham Wood. Compared with the modeled UK mean (8.6 kg N m−2 year−1; EMEP data), the rate at Wytham is low (2.9 kg N m−2 year−1). It is well established that stimulation of decomposition via N fertilization is more likely to occur in sites with low background N deposition rates (Hobbie 2005; Knorr et al. 2005). In studies in which the decomposer organisms colonize the litter from the soil, this effect could be due either to N-limitation or to shifts in the decomposer community under high N deposition. In the present study, the decomposer organisms were experimentally controlled, suggesting N-limitation of wood decay rates.

Wood decomposition rates were comparable with other studies on the same species. Dowson et al. (1988) inoculated 8-cm3 beech wood blocks with H. fasciculare and P. velutina (among other species) and measured mass after 2 years. Almost half (44 ± 16%, mean of five sites) of blocks inoculated with H. fasciculare had disappeared by the end of the experiment, while 70 ± 10% of blocks inoculated with P. velutina had completely decayed. There was considerable variability in decay rate across the experiment, and it is possible that in some locations other factors, most probably water availability, were limiting (Rayner and Boddy 1988).

Previous studies have reported variable effects of N addition on fungal biomass (Entry and Backman 1995; Sinsabaugh et al. 2002; Frey et al. 2004; Gallo et al. 2004), but N-deposition rates in these experiments were extremely high compared with anthropogenic levels, so it is difficult to interpret these findings. In contrast, laboratory studies generally support increased biomass production with N-availability (Boyle 1998; Kachlishvili et al. 2005). It is also known that the fungal mycelium can sequester ‘luxury’ N (Watkinson et al. 2006), and are therefore possible candidates for the missing sink of anthropogenic N in forest soils (Currie 1999). Although the data suggested that N treatment increased the N-concentration of mycelium, wood blocks, and soil, there was no statistical support for these hypotheses.

Responses of soil N measures to N addition can be weak (Aber et al. 2003), and take several years to become evident even when additions are high (Dise and Gundersen 2004). Both observational (e.g., Dise et al. 1998; Falkengren-Grerup et al. 1998) and experimental (Magill et al. 2004) studies have failed to detect an effect of N addition on forest soil N, therefore the lack of response in the present study is not surprising. Similarly, although some studies have demonstrated positive relationships between soil N and decomposition rates (e.g., Kuperman 1999), the variability of soil N is likely to have obscured any relationship between total soil N and decomposition in the present study, because soils were sampled only once. Saprotrophic fungi build highly adaptive foraging mycelia (Boddy 1999), and there is perhaps little reason to expect that total N in soil cores would approximate N availability as detected by the fungus.

An increase in litter N concentration through time is well documented for both leaf (Berg and Laskowski 2005) and woody litter (Laiho and Prescott 2004), as is the release of N in the later stages of decomposition (Laiho and Prescott 2004). In the present study, the mass of N in wood blocks decreased, while that in the mycelium increased, suggesting that the growing mycelium was extracting N from the wood following initial import (Watkinson et al. 2006). However, the overall decrease in N in the wood–mycelium system with decomposition indicates that this process is inefficient, and that leaching, or perhaps mycophagy (Tordoff et al. 2008), released N into the ecosystem.

The data suggest that results from previous leaf litter experiments that utilize greatly elevated N deposition levels, and from laboratory studies on lignin decomposition in isolation, cannot be readily extrapolated to changes in wood decay rates under realistic N deposition levels. A non-linear response to N deposition is supported, with low deposition rates enhancing decomposition, and high deposition rates reducing decomposition. Further work on this process is required, since dead wood is an ecologically important resource, particularly as a carbon pool and habitat (Harmon et al. 1987; Freedman et al. 1996; Grove and Meggs 2003; Laiho and Prescott 2004; Jonsson et al. 2005). Increasing dead wood inputs are expected due to global forest dieback (Bréda et al. 2006; Jurskis 2005), and through alterations of forest management practices to more closely mimic natural disturbances (Bebber et al. 2004; Jonsson et al. 2005; Lindenmayer et al. 2006).

Acknowledgments

The authors thank Mike Morecroft, Michele Taylor, George Tordoff, and Juliet Hynes. The study was funded by NERC grant NER/A/S/2002/882.

Supplementary material

442_2011_2057_MOESM1_ESM.doc (826 kb)
Supplementary material 1 (DOC 826 kb)

Copyright information

© Springer-Verlag 2011