Plant and Soil

, Volume 333, Issue 1, pp 275–285

Does light exposure make plant litter more degradable?

Authors

    • Institute of Biological and Environmental SciencesUniversity of Aberdeen
    • Department of Crop and Soil SciencesCornell University
  • Jessica Bellarby
    • Institute of Biological and Environmental SciencesUniversity of Aberdeen
  • Wolfram Meier-Augenstein
    • Scottish Crop Research Institute, Invergowrie
    • Centre for Anatomy & Human IdentificationUniversity of Dundee
  • Helen Kemp
    • Scottish Crop Research Institute, Invergowrie
Regular Article

DOI: 10.1007/s11104-010-0342-1

Cite this article as:
Foereid, B., Bellarby, J., Meier-Augenstein, W. et al. Plant Soil (2010) 333: 275. doi:10.1007/s11104-010-0342-1

Abstract

Many field experiments have indicated that litter decomposition in semi-arid areas may be partly or fully controlled by photodegradation. We devised a study to test our hypothesis that light exposure makes plant litter more degradable. Dry, senescent, aboveground plant litter from Miscanthus x giganteus was exposed to light including ultraviolet (UV) radiation for various lengths of time from 0 to 289 days. Weight loss was measured after exposure and appeared to be modest and did not increase with time of exposure. The litter of the longest and shortest exposure time as well as controls were then incubated with soil and moisture for 35 days and CO2 and N2O production were measured. The longest exposed litter degraded much faster than any other treatment during incubation with moisture, about twice as fast as the unexposed control. The shortest exposed however, degraded only slightly faster than the unexposed control. This suggests that increasing litter degradability is a more important mechanism for photodegradation than direct light-induced mass loss. N2O production from decomposition of the exposed litter was high in the beginning, suggesting that nitrogen may be released quickly. The mechanism is probably that light exposure leaves the nitrogen in plant litter easily available to microbial utilisation upon wetting. Such a mechanism might play an important role for nutrient cycling in semi-arid areas.

Keywords

PhotodegradationDecompositionNitrous oxide13C

Introduction

Understanding the global carbon cycle is necessary to predict future biotic feedbacks to global change (Cox et al. 2000). Carbon turnover in soils is the largest uncertainty in current models of such feedbacks (Friedlingstein et al. 2006). Generally soil and litter decomposition are assumed to depend on moisture and temperature. This has been found to be broadly correct for a large number of environments (Cadish and Giller 1997; Couteaux et al. 1995; Parton et al. 2007). However, carbon cycling in arid and semi-arid areas is imperfectly understood. Decomposition is faster than would be expected from moisture availability alone (Parton et al. 2007; Tian et al. 2007). As drylands cover more than 40% of the terrestrial land surface (Reynolds 2001), it is important to understand what drives biogeochemical cycles in these areas.

The degradation of materials like plastics by light, particularly in the UV range, has been well-studied (Moorhead and Callaghan 1994). The role of irradiance in leacning and breaking down dissolved and particulte organic matter to inorganic carbon in aquatic environments are well known (Mans et al. 1998; Anesio et al. 1999; Bertilsson and Tranvik 2000). Photooxidation may make material both more and less available for microbial degradation (Tranvik and Bertilsson 2001). There may be a complicated interaction between microbial degradation and photodissolution, where products from microbial decay can be made more available for further microbial degradation upon irradiation (Mayer et al. 2009).

It has more recently become clear that photodegradation may also play a role in litter decomposition in terrestrial environments (Austin and Vivanco 2006; Day et al. 2007; Pancotto et al. 2005; Parton et al. 2007; Vanderbilt et al. 2008). It is mainly in seasonally dry environments that photodegradation appears to be important. In less dry environments shading may be more common all year round, and other factors well known to be drivers of decomposition (climate, soil factor) may play larger roles (Brandt et al. 2007). Most studies of photodegradation have been done by putting out litter bags or boxes on the soil surface. In most cases theses studies have found that light, particularly in the UV range, has an effect, sometimes a large effect on mass loss rate degrading litter (Austin and Vivanco 2006; Day et al. 2007; Gallo et al. 2009; Pancotto et al. 2005; Vanderbilt et al. 2008).

To find the mechanism for photodegradation, controlled experiments need to be carried out. Gallo et al. (2006) and Brandt et al. (2009) are the only studies to date looking at photodegradation of litter under controlled conditions. Brandt et al. (2009) conducted a series of experiments testing different hypotheses for the mechanism for terrestrial photodegradation, but they used relatively low radiation levels and short exposure times. They did not find any significant effect of irradiation on subsequent leaching or microbial degradability. They concluded that photodegradation was a direct result of CO2 production upon irradiation, and that surface area was the most important factor determining the amount of degradation. However, Gallo et al. (2006) found partly contradictory results in much longer exposure treatments. The treatments were dark-wet, light-dry and light-wet, but there was no dark-dry control. No significant difference in mass loss between the treatments was found, but enzyme activity and chemical composition of dissolvable organic matter was different between treatments. In a subsequent field-based study Gallo et al. (2009) investigated the effect of photo-exposure on degradation rate and microbial parameters. They found exposed litter degraded faster regardless of litter type. They also concluded that the fungal community found in this arid environment was different from that in mesic environment in that it was dominated by early colonisers throughout the degradation process. This could indicate that the exposure to light may change the litter to make it more easily degradable. They also found that dissolved organic matter extracted from photo-exposed litter was more aromatic than that from shaded litter, suggesting that breaking up and dissolving recalistrant structures may be one way photo-exposure makes the litter more easily accessible for microbial degradation.

In the semi-arid zone, where photodegradation is likely to be most important, photodegradation is likely to mainly take place in the dry season, whilst microbial degradation is likely to mainly take place in the rainy season. There may also be good reason to assume that microbial and photodegradation are separated in time because light in the UV range may be harmful for microbes (Hughes et al. 2003). The chemical changes in the plant material during photodegradation may make it more accessible to microbial degradation, particularly as light in the UV range is known to preferentially degrade lignin (Davidson 1996; Gallo et al. 2006).

We hypothesise that light exposure makes plant litter more degradable whereby the dry season acts as a “pre-treatment” of litter for degradation during the wet season. In this paper we report the results of a study comprising a series of experiments or treatments where light exposure and moisture for microbial degradation were separated in time to test the above hypothesis. We first irradiated the dry litter for different lengths of time and then incubated it with soil and moisture. We measured both the direct effect of irradiation on mass loss and the subsequent effect on degradability, to be able to assess the relative significance of each. To the best of our knowledge this is the first time photodegradation has been studied under controlled conditions with a complete set of non-exposed controls and radiation amount and wavelength distribution approaching field conditions.

Material and methods

Plant material

Plant material was from Miscanthus x giganteus planted in 2003 on clay soil in Wales, UK (52°26′N, 4°1′W). Miscanthus is a C4 grass grown as feedstock for energy production and harvested annually (Lewandowski et al. 2000). The standing dead, senesced, aboveground plant material was harvested in April 2006 and air-dried and stored until December 2007 when the experiment started. Some simple chemical parameters are given in Table 1. More information about the chemical composition of Miscanthus, can be found in (Ververis et al. 2004) (α-cellulose 41.5%, Clason lignin 27.6%). The material was cut into smaller, more even pieces using a blender.
Table 1

Characteristics of the plant and soil material used in the incubation. Both exposed and unexposed plant material were analysed, but as there was no difference between them, only an average is shown. pH was measured in water. Water holding capacity (WHC) was measured by weighing the soil after it had drained excess water for 24 h under cover and then weighing it again after drying (110°C). Other measurements are described in the text. Values are means and standard error in brackets (n = 4 for soil, n = 16 for plant material)

 

Carbon (%)

Nitrogen (%)

δ13C (‰)

pH

WHC (%)

Soil

2.00 (±0.06)

0.12 (±0.01)

−28.95 (±0.13)

3.38 (±0.05)

53.0 (±3.0)

Plant material

44.47 (±0.09)

0.18 (±0.01)

−11.98 (±0.03)

Conditions during light exposure

The growth chamber light was supplemented with 4 Arcadia 3D reptile lamp, 90 cm, which were changed after 5 months. Total light energy was 33 Wm−2 and the temperature was kept at 30°C. The wavelength distribution in the chamber was measured with SR9910-V7 Spectroradiometer (Macam., UK) and is given in Fig. 1. Dry plant material (5–6 g) was weighed and placed in foil trays. Before each weighing the trays were dried (40°C) overnight to minimize possible effects of varying moisture content. The trays were moved around in the chamber every two weeks to avoid potential effects of differences in light in different parts of the chamber (although these differences were rather small). Exposure times were: 43, 83, 94, 147, 206 and 289 days.
https://static-content.springer.com/image/art%3A10.1007%2Fs11104-010-0342-1/MediaObjects/11104_2010_342_Fig1_HTML.gif
Fig. 1

Wavelength distribution of the radiation in the growth chamber. Radiation was constant, so that time in exposure is proportional to total cumulative radiation

Incubation with soil

Plant material from each tray was incubated with soil in 100 mL flasks in the dark. The soil used was from Kapil Sepilok forest, Malaysia (Dent et al. 2006), a mixture of the soil types in that experiment. This soil was from a tropical area, but from a forest rather than grassland in order to be able to differentiate the 13C signature of the soil from that of the C4 plant litter (Table 1). Details of soil and plant material can be found in Table 1. Fifty g soil and 4 g of plant material was weighed and put in each flask, the litter on top of the soil to simulate the condition in the field where photo-exposed litter would be found on top of the soil. Material was wetted with deionised water to near field capacity and incubated at 30°C and 80% humidity. More deionised water was added as needed during the incubation (about 5 mL per week).

Treatments

The incubated treatments were: the longest exposed (289 days), the shortest exposed (43 days), not exposed (only kept in coldroom, here referred to as unexposed) and kept dark in the growth chamber for 289 days (here referred to as dark). A soil only control was also included, as well as empty bottle controls. 4 replicates were used. The dark controls were shaded with non-transparent plastic sheeting. The plant material was stored in a coldroom at 4°C and weighed before and after. All plant material was stored in coldroom when not exposed.

Gas analysis

Flasks were closed for between 1 and 4 h before each sampling using subaseal and samples were taken out with a syringe. 1 mL samples were taken for nitrous oxide analysis and 5 mL were taken for carbon dioxide measurements. CO2 in the gas samples was measured on a GC-FID C with a methanizer and a Porapak Q column (Chrompack, Middelburg, The Netherlands). The instrument was calibrated using standards of 350, 1,000, 5,000 and 10,000 ppm CO2 before the start of each measurement series. N2O was measured on a GC-ECD 68Ni (Perkin Elmer, USA) with a Hayesep Q column (Chrompack, Middelburg, The Netherlands). The instrument was calibrated using standards of 0.5, 1 and 5 ppm N2O before each measurement series.

For measurement of 13C abundance of CO2 in gas samples, gas samples were collected in 12 mL Exetainer tubes (Labco, High Wycombe, UK) and stored until measurement. Gas samples were analysed on an AP2003 gas isotope ratio mass spectrometer [IRMS] (Analytical Precision, Norwich, UK), an instrument set-up specifically designed for the isotope analysis of gaseous CO2 (Carre et al. 2005; El-ghali et al. 2006). This set-up comprises an autosampler, an in-line water trap, an automated gas sampling loop, a GC column and a reference gas module. Each sample was analysed in triplicate and for each analysis a subsample of 200 μL was taken from a 12 mL Exetainer by means of an automated gas sampling loop. From the sample loop the gas sample is transported in a helium carrier gas stream through a GC column for separation of CO2 from any other permanent atmospheric gases that may be present in the gas sample. Once the CO2 peak has been admitted to and detected by the IRMS a reference gas pulse from a cylinder of CO2 of known isotopic composition was admitted into the IRMS for isotopic calibration of the measured sample CO2 isotope ratio.

Plant and soil analysis

Plant and soil material was powderised on a ball mill. Samples were weighed into tin capsules (SerCon Ltd, Crewe, UK) and crimped for total carbon and nitrogen as well as 13C analysis. Carbon and nitrogen content were measured on a Fisons NA 1500 NCS micro-elemental analyser (Italy). Bulk 13C isotope analysis of soil and plant material samples was carried out using an automated nitrogen−carbon analyser (ANCA) coupled to a 20/20 isotope ratio mass spectrometer (SerCon Ltd, Crewe, UK) (Hopkins et al. 2009). Measured 13C/12C isotope ratios were processed using proprietary software Calisto (SerCon Ltd, Crewe, UK) and anchored to the VDPB (Vienna Peedee Belemnite) scale by contemporaneous analysis of international reference materials IAEA-600 (δ13C = −27.77‰; IAEA, Vienna, Austria) and IAEA-CH6 (δ13C = −10.45‰. IAEA, Vienna, Austria). Mineral nitrogen (ammonium and nitrate) in the soil was measured at the start and the end of the incubation. The soil was sieved (2 mm) and extracted in 1 M KCl (Allen 1989). Nitrate and ammonium in the extracts were measured on a Foss Tecator Flow Injection Analyzer (Sweden).

Calculations and statistics

As the ratio 13C /12C ratio is very small in most materials, it is customary to express it as δ13C in units of ‰ compared to the international reference material VPDB carbonate (Craig 1957) with δ13C being defined as:
$$ {\delta^{13}}{\hbox{C}} = \left( {{{\left( {{{\hbox{R}}_{\rm{sample}}} - {{\hbox{R}}_{\rm{reference}}}} \right)} \mathord{\left/{\vphantom {{\left( {{{\hbox{R}}_{\rm{sample}}} - {{\hbox{R}}_{\rm{reference}}}} \right)} {{{\hbox{R}}_{\rm{reference}}}}}} \right.} {{{\hbox{R}}_{\rm{reference}}}}}} \right) \times 1000 $$
where R = 13C/12C for sample and reference. As plant material contains less 13C than the VDPB standard, corresponding δ13C-values will be negative.
Measured 13C/12C isotope ratios of gas sample CO2 analysed on the AP2003 were processed using proprietary software AP 2.0 (Analytical Precision, Northwich, UK) to yield gross δ13C values, i.e. not corrected for background, while measured 13C/12C isotope ratios of soil samples analysed on the 20/20 IRMS were processed using proprietary software Calisto (SerCon Ltd, Crewe, UK). Isotope abundances were used to calculate how much of the CO2 that came from the soil and how much came from the litter. Since no δ13C value for litter only CO2 could be obtained, this was calculated assuming that the ratio between soil CO2 and soil carbon was the same as the ratio between litter CO2 and litter carbon (fractionation factor). A similar procedure was suggested by Bernoux et al. (1998). The percentage of CO2 derived from the litter can then be calculated as:
$$ \% \;{\hbox{litter}}\;{\hbox{derived}}\;{\hbox{C}}\;{\hbox{in}}\;{\hbox{sample}} = {{\left( {{\delta^{13}}{{\hbox{C}}_{{\rm{CO2}}\;{\rm{sample}}}} - {\delta^{13}}{{\hbox{C}}_{{\rm{soil}}\;{\rm{CO2}}}}} \right)} \mathord{\left/{\vphantom {{\left( {{\delta^{13}}{{\hbox{C}}_{{\rm{CO2}}\;{\rm{sample}}}} - {\delta^{13}}{{\hbox{C}}_{{\rm{soil}}\;{\rm{CO2}}}}} \right)} {\left( {{\delta^{13}}{{\hbox{C}}_{\rm{litter}}} \times {\hbox{fractionation}}\;{\hbox{factor - }}{\delta^{13}}{{\hbox{C}}_{{\rm{soil}}\;{\rm{CO2}}}}} \right) \times 100}}} \right.} {\left( {{\delta^{13}}{{\hbox{C}}_{\rm{litter}}} \times {\hbox{fractionation}}\;{\hbox{factor - }}{\delta^{13}}{{\hbox{C}}_{{\rm{soil}}\;{\rm{CO2}}}}} \right) \times 100}} $$

CO2 concentration and corresponding δ13C values for empty flasks were assumed to be the situation when the flasks were closed. CO2 production rate was calculated by subtracting the amount of CO2 in the empty flasks from that in each of the sample flasks and dividing by the time they were closed. The fraction of this coming from litter and soil respectively was calculated using the formula above.

Total loss of carbon from the litter during the experiment was calculated assuming linear change in CO2 production rate between measurement days. The percentage of total carbon lost at each time point was calculated assuming the carbon content of the plant material was as in Table 1. Cumulative loss over the whole period as well as mass remaining at each time point were calculated. An exponential decay function was fitted to mass remaining:
$$ {\hbox{C}}\left( {\hbox{t}} \right) = 100{{\hbox{e}}^{{\rm{ - kt}}}} $$

C(t) is the percentage of carbon left at time t and k is a constant.

Statistical analysis was carried out using R (R Development Core Team 2008). Paired t-test was used to test if weights before and after exposure to radiation were different. It was also tested if changes in weights were different for each length of exposure using ANOVA. Each gas sampling date was analysed separately using ANOVA to determine how the treatments affected each time point. If ANOVA showed significant difference, the treatments were subsequently compared using pairwise t-test with significance level adjusted for multiple comparisons.

Results

Weight loss was measured after light exposure while CO2 evolution was measured during incubation. In this way it was assessed how much material was degraded during photoexposure and microbial degradation, respectively.

Light exposure led to a small but significant decrease in weight for most lengths of exposure (Fig. 2). The control, kept dark in the exposure chamber showed a non-significant increase in weight. However, there was no significant difference between different times and no indication that the weight loss increased with time under exposure. There were no significant differences in carbon and nitrogen content or 13C signature between exposed and unexposed litter (Table 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs11104-010-0342-1/MediaObjects/11104_2010_342_Fig2_HTML.gif
Fig. 2

Weight loss during exposure to radiation as a function of length of exposure. Error bars are standard error (n = 12). One star indicates significantly different from start value at 0.05 level, two stars significant at 0.01 level

CO2 production from litter during incubation was initially not much different between treatments, but the longest exposed treatment reached a higher CO2 production rate from 5–7 days onwards (Fig. 3, Table 2). This was sustained during the whole incubation period, except the last sampling (Table 2). The shortest exposed treatment was not much different from the unexposed treatments, but sometimes had slightly higher CO2 production rate (Table 2). The CO2 production pattern from the treatment kept dark in the growth chamber had different CO2 production pattern from the unexposed kept in the coldroom. The unexposed (kept cold) showed higher CO2 evolution in the beginning, but declining more at the end than the ones kept dark (Fig. 3, Table 2). The total amount of carbon lost during the incubation was almost twice as large in the longest exposed as in the unexposed treatment (Table 3). The carbon remaining fitted an exponential decay function well (Table 3).
https://static-content.springer.com/image/art%3A10.1007%2Fs11104-010-0342-1/MediaObjects/11104_2010_342_Fig3_HTML.gif
Fig. 3

Respiration rate from soil and plant litter as a function time since start of the incubation (moisture addition). Fraction of total respiration from soil and plant is calculated using 13C signatures of the CO2, except for the soil only treatment. Unexposured means samples that have not been in the growth chamber, and dark means kept shaded in the growth chamber for the maximum length of time. Error bars are standard error (n = 4)

Table 2

Results from statistical tests for CO2 production rate for the soil (top) and litter (below) parts. Results show p-values from multiple comparisons with significance level adjusted for each day number in incubation (horizontal) and each treatment combination (vertical)

Soil

Day

2

5

7

11

15

18

22

28

35

Soil – dark

>0.01

0.04

>0.01

>0.01

>0.01

>0.01

>0.01

0.05

0.67

Soil – unexposed

1.00

>0.01

>0.01

>0.01

>0.01

>0.01

>0.01

0.03

0.72

Soil – 43 days

0.06

0.01

>0.01

>0.01

>0.01

>0.01

>0.01

0.02

0.72

Soil – 289 days

0.94

0.06

>0.01

>0.01

>0.01

>0.01

>0.01

>0.01

0.04

Dark – unexposed

0.02

0.85

0.20

1.00

0.01

0.02

0.15

1.00

1.00

Dark – 43 days

0.94

1.00

0.21

1.00

0.39

0.39

0.15

1.00

1.00

Dark – 289 days

0.06

1.00

0.43

0.36

0.52

0.39

0.05

1.00

0.72

Unexposed – 43 days

0.15

1.00

>0.01

1.00

0.17

0.39

1.00

1.00

1.00

Unexposed – 289 days

1.00

0.59

0.41

1.00

0.09

0.39

1.00

1.00

0.63

43 days – 289 days

0.39

0.06

0.07

0.42

0.68

0.97

1.00

1.00

0.60

Litter

Day

2

5

7

11

15

18

22

28

35

Dark – unexposed

0.12

0.96

>0.01

0.03

0.20

0.22

0.12

0.45

1.00

Dark – 43 days

0.41

0.96

>0.01

0.02

0.20

0.32

0.51

0.76

1.00

Dark – 289 days

0.12

0.70

>0.01

>0.01

>0.01

>0.01

>0.01

0.03

0.18

Unexposed – 43 days

0.01

0.49

>0.01

>0.01

0.02

0.05

0.25

0.76

1.00

Unexposed – 289 days

>0.01

0.18

>0.01

>0.01

>0.01

>0.01

>0.01

>0.01

0.06

43 days – 289 days

0.41

0.96

>0.01

>0.01

>0.01

>0.01

>0.01

0.01

0.07

Table 3

Parameters calculated from litter degradation data. Total carbon loss was calculated from CO2 production rate and total amount of material and carbon content in the beginning (n = 4). Turnover rate was calculated assuming exponential decay, r2-value for that is also shown

 

Dark

Exposed 0 days

Exposed 43 days

Exposed 289 days

Total carbon loss during incubation (% of start)

4.54

3.56

4.49

7.36

Turnover time (days)

722

860

692

476

r2

0.99

0.94

0.97

0.98

The addition of litter material stimulated CO2 production in the soil, but all the litter types stimulated it equally (Fig. 3). CO2 production rate from soil was significantly different in the soil only treatment from all or some of the other treatments on most of the sampling days (Table 2).

Nitrous oxide production was very high in the beginning of the incubation in the longest exposed material (Fig. 4). However, in all the treatments nitrous oxide emission quickly declined to close to zero. Nitrous oxide emission from the longest exposed treatment was highly significantly different from all the other treatments at the first sampling (day 3). At the fourth sampling (day 14) the longest exposed treatment was also weakly significantly different from all the others except the soil only treatment. There were no other significant differences in nitrous oxide production rates between any of the treatments.
https://static-content.springer.com/image/art%3A10.1007%2Fs11104-010-0342-1/MediaObjects/11104_2010_342_Fig4_HTML.gif
Fig. 4

Nitrous oxide emission rate as a function of time since start of the incubation (moisture addition). Unexposured means samples that have not been in the growth chamber, and dark means kept shaded in the growth chamber for the maximum length of time. Error bars are standard error (n = 4)

There was hardly any nitrate in the soil, and ammonium in the soil in the end of the experiment showed no difference between treatments except that it was significantly higher in the soil only than in all the other treatments (data not shown). Soil ammonium concentration in the other treatments was not significantly different from that in the beginning.

Discussion

The experiment indicates that an important mechanism for photodegradation is that it increases litter degradability. Although there was a small direct effect of radiation on mass loss, the main effect was to make the litter more degradable. This is in contrast to the results of Austin and Vivanco (2006) who found photodegradation to be sustained, even in the absence of any effect of microbial degradation. The reason may be that in nature, photodegradation may interact both with microbial degradation and physical disturbance (wind, temperature changes). Throop and Archer (2007) found that soil and litter mixing enhanced decomposition in a semidesert grassland, and they speculated that physical disturbance may be the mechanism. The absence of mechanical disturbance during exposure in our experiment could explain why weight loss did not increase with time under exposure in this experiment. The litter in this experiment also received the radiation from the same angle most of the time, whilst in nature the angle will change. Brandt et al. (2009) found that litter with higher surface area degraded faster, consistent with the hypothesis that a chaning angle may increase photodegradation.

Brandt et al. (2009) did not find any effect of pre-treatment of litter with UV light on subsequent microbial degradation, apparently contradicting our results. The reason for that is probably that their pre-treatment was relatively short (3 weeks) and their radiation low (14 Wm−2). Exposure time and total radiation during exposure in their experiment were both well below those even the shortest exposure treatment in our experiment, although the total amount of UV received during the exposure were probably comparable. We also did not observe much effect of the pre-treatment in the shortest exposed treatment. This all indicates that relatively large radiation doses are required have a significant effect. The pre-treatment effect of photo-exposure will therefore be important in environments were the litter is exposed to high radiation for extended periods of time, such as tropical environments with an extended dry season. Brandt et al. (2009) observed a loss of CO2 as a direct result of photo-exposure. This is consistent with our observation of a small weight loss during light exposure, and most of the weight was lost shortly after the start of exposure.

In the beginning of the incubation most of the CO2 came from the soil, and there was therefore little difference between the treatments. However, during this period the nitrous oxide emission was high in the longest exposed treatment (see below), indicating that at least in that treatment, the litter was already contributing to microbial activity. The CO2 production rate increased until about day 5 and then declined in all the treatments except the longest exposed, where it remained almost constant. The reason might be that in the photoexposed litter, supply of easily available carbon remained higher than in the unexposed litter due to photodegradation of recalistrant carbon. Gallo et al. (2009) found that the fungal community in photoexposed litter remained dominated by early colonizer for the whole decomposition process, supporting this theory.

The unexposed (kept cold) treatment released more CO2 in the beginning and slightly less later in the incubation than the other treatments. A possible explanation for that is that some of the most easily degradable material may have been lost or changed when it was in the warm growth chamber. This must have happened to some extent also in the absence of light and could be an effect of temperature.

The incubation shows that the longer photoexposure did change litter degradability. This is consistent with Gallo et al. (2006) who found that chemical composition and enzyme activity changes after photoexposure. The total radiation received during this experiment is modest compared to what it would receive during a dry season in the semi-arid zone. As the change in degradability was shown to depend on total length of exposure, even larger changes in degradability could be expected in the natural environment. The enhanced degradability is therefore important for the carbon cycle, particularly in areas with a prolonged dry season. A delayed effect of photo-exposure has been observed in the field by Henry et al. (2008) who found that exposure to radiation during the dry season increased litter degradation during the next wet season, which in conjunction with the observations presented here further supports our hypothesis of dry seasons “pre-treating” litter for degradation during the wet season.

Addition of fresh carbon to the soil has been found to stimulate soil respiration in a number of studies (Foereid et al. 2004; Fontaine et al. 2003, 2004, 2007; Fu and Cheng 2002). This is called a priming effect (Kuzyakov et al. 2000). Priming was also observed in this experiment. However, the magnitude of the effect did not increase with the degradation rate of the litter. It is possible that there is a threshold for how much priming there can be, and this was exceeded even at lowest rate of carbon supply in this experiment.

The high nitrous oxide emission early in the experiment in the exposed litter indicates a very quick release of nitrogen upon wetting. A rapid nitrogen mineralisation upon re-wetting of dry soil in semi-arid climates is known as the “birch effect” (Rey et al. 2005). A delayed response to photo-exposure might be part of the mechanism for this effect, as we found much more nitrous oxide emission from the longest exposed treatment than the other treatments. Parton et al. (2007) found that aboveground litter in seasonally dry grassland does not immobilise nitrogen. That may indicate that the litter exposed to radiation releases its nitrogen quickly, which is consistent with the findings here. The period of quick release was very short probably because the microbes quickly started to use the excess nitrogen, so that in the end of the incubation there was no difference in mineral nitrogen content. The low soil pH could also limit the amount of nitrogen emitted as nitrous oxide (Sitaula et al. 1995). We did not study the relevant period intensively enough to really conclude about nitrogen dynamics, but this is an area that deserves further study. A possible mechanism could be that exposure to radiation breaks up the cell walls in plant litter and so allows nitrogen to leak out when wetted.

We conclude that facilitation of microbial degradation is an important mechanism for photodegradation. This facilitation was more important than the direct effect of light on mass loss in our experiment. Photo-exposure probably also makes the nitrogen in the litter much more easily available. Further work should focus on the role of photo-degradation in the nitrogen cycle as well as the carbon cycle.

Acknowledgements

The authors wish to thank Dr. John Clifton-Brown at IGER, UK for supplying the Miscanthus plant material, and Dr. David Burslem at Aberdeen University for supplying soils from a tropical environment. Professor David Robinson at Aberdeen University is acknowledged for helping with light spectrum measurements. This study was funded by the University of Aberdeen. Professor Pete Smith at Aberdeen University is acknowledged for helping to obtain funding.

Copyright information

© Springer Science+Business Media B.V. 2010