Preliminary characterization of microbial functional diversity using soleC-source utilization profiles in Tremoal do Pedrido mire ( Galicia , NW Spain )

Peatlands are a global carbon sink. It has been estimated that Northern peatlands store 257-342 Gt of carbon. Most carbon is in the form of peat organic matter whose transformation is largely dependent on microbial activity, and this in turn has a major influence on carbon cycling. The presence of two main physico-chemical environments in the peat deposits -the upper, aerobic, acrotelm and the lower, anaerobic catotelmsuggests that there could be differences in both the microbial activity and communities’ composition with depth. In this preliminary study we analyzed the depth variation in microbial community functional diversity using sole–carbon–source utilization profiles, in a raised bog from Galicia (NW Spain). Substrate utilization was quantified by measuring the absorbance at 590 nm of the colour developed by the reduction of the tetrazolium dye contained in each carbon source. Substrates consumption was followed for 10 days and expressed as average well colour development (AWCD) and microbial diversity (Shannon-Wiener index, H). The highest activity (AWCD 1.6-2.0) and microbial diversity (H 3.3) were found at the surface (upper 2 cm) of the peatland. No substrate was used below a depth of 52 cm. Principal components analysis showed three main depth records of degradation: i) substrates (N-compounds, carbohydrates, carboxylic acids) only used at the surface of the peatland, ii) substrates (mostly carboxylic acids) used at the surface and at 46-48 cm, and iii) substrates (N-compounds and polymers) used from the surface to a depth of 48-52 cm. The overall kinetics of substrate utilization showed four patterns: i) asymptotic, ii) exponential (although of low activity), iii) linear, and iv) no reaction over the whole incubation period. Some differences were observed, both in intensity of substrate degradation and total time of reaction, when substrates were grouped according to the results of principal component analysis. These findings suggest a “stratification” of the microbial communities that may be controlled by the varying geochemical conditions (humidity, temperature, acidity, nutrientand oxygen availability) with depth, and that the acrotelm/catotelm boundary is an effective barrier for oxidative degradation of the organic matter in peatlands. AUTHORS


Introduction
According to the definition of the Natura 2000 framework, peatlands are wetlands formed by the accumulation of peat and have current peat-forming vegetation.Peat accumulates because the net production of organic matter exceeds its decomposition.This fact makes peatlands a global carbon sink.They contain one third of the global soil carbon and 10% of the global freshwater (Bartalev et al. 2004a, b).Estimations of the total carbon reservoir in northern peatlands by different authors vary between 257-342 Gt (Tarnocai and Stolbovoy 2006).Although dominant in boreal areas, peatlands are also present in temperate and tropical latitudes, where they also play an important role in the water and carbon cycles.In Galicia, for example, it has been estimated that mountain peatlands have accumulated 10-16 Mt of carbon (Pontevedra Pombal et al. 2004), representing 4.7-7.5% of the carbon stored in forests biomass in Spain and 24-38% of the forests biomass in Galicia.
Carbon is fixed by photosynthesis and is stored initially in the acrotelm, i.e, the relatively young, well-aerated surface layer of the peat deposit (Ingram 1978).After that, carbon will be retained at depth in the anaerobic catotelm.However, not all the carbon that enters the acrotelm is later stored in the catotelm because of the loss by aerobic decomposition.About of 90-97% of the carbon fixed by living plants is typically lost by decomposition (Francez and Vasander 1995), a process that is carried out by microorganisms (Gilbert and Mitchell 2006).Despite their role as an atmospheric C sink, peatlands can also become a source of C to the atmosphere if the balance between decomposition and accumulation is modified, as it has been suggested to occur due to climate change (Tarnocai and Stolbovoy 2006).Predictions made using global circulation models indicate that the northern areas (where peatlands are extensive) will be the most affected by climate warming (Tarnocai and Stolbovoy 2006).Climate change could not only alter the carbon cycle in peatlands, but also enhance the export of stored contaminants in peat (i.e.Hg, Pb, organohalogens) to the aquatic systems or to the atmosphere (Martínez Cortizas et al. 2007).
Taking into account the huge amount of carbon accumulated in peat deposits and the risk of their change from sink to source, the importance of understanding all the aspects related to decomposition mechanisms is justified.
Peatlands are widely studied systems in different fields.As methods and techniques have evolved in recent years, there has been a large increase of knowledge regarding the diversity and composition of their microbiota (Andersen et al. 2013).Relatively few studies deal with peatland microbial communities as a whole, as most research was done on particular groups (e.g.methanotroph) (e.g.Dedysh 2009) and mostly for taxonomic purposes.Within this category there are studies related to the use of microbial indicators, such as testate amoebae, for the reconstruction of past environmental conditions in paleoecological studies (Charman et al. 1999;Charman 2001;Mitchell et al. 2001).
In the functional approach, on the other hand, the focus is on the role of microorganisms in the cycling of nutrients (carbon and nitrogen) and organic matter decomposition, targeting isolated enzymes and is almost exclusively based on cultivation-based studies (Artz 2009).Despite the limitations of both approaches, the importance of peatlands as carbon reservoirs, sinks, and sources in the global carbon cycle (Andersen et al. 2013), and their relevance to the cycles of greenhouse gases (CO 2 and CH 4 in particular) justifies the development of a functional approach.
One possible approach to the characterization of microorganism activity in soils is to analyze the functional diversity of the microbial community using sole-carbon-source utilization profiles; and for this the Ecoplates from BIOLOG have been proposed as a simple and efficient method.The Biolog plates method was used, for example, to compare the metabolic activity of heterotrophic microbial communities from different habitatswater, soil and wheat rhizospheres (Garland andMills 1991). Garland (1996)  rotation on the diversity and community structure of soil bacteria.Pietikäinen et al. (2000) studied the microbial substrate utilization pattern (using Biolog Ecoplates) to determine the effect of burning in the microbial community structure in humus layers.In the field of peat research, Fisk et al. (2003) used Ecoplates to compare patterns of soil microbial activity in northern peatlands which differed in vegetation communities.
In this paper we describe the preliminary results of the application of the Ecoplates method to a peat core, with the aim to study the microbial activity patterns with depth.As already mentioned, the upper part of the peat deposits contains two main geochemical environments that largely differ in water content, oxygen and nutrient availability, and thus also likely differ in their microbial activity.The peatland we sampled is located in NW Spain, a temperate area, and its functionality may serve as an analogue for northern peatlands under a scenario of climate warming.
With the previous considerations, the objectives of this work were: i) to characterize the microbial activity patterns at different depths in the upper part (100 cm) of the peat deposit where the most intense biogeochemical changes occur; ii) to characterize the kinetics of the decomposition of the different carbon substrates; and iii) to study the relationship between the diversity of microbial availability and the kinetics with other physico-chemical properties of the peat.

Location and sampling
The peat core (OBX) used for this study was sampled in June 2013 in Tremoal do Pedrido (29T 0619124 4812082 UTM), an ombrotrophic raised bog located in the Xistral Mountains (Northwestern Spain) at an elevation of 695 m a.s.l., and 29 km south of the northern coast of Galicia (Figure 1).The raised bog originated from an earlier minerogenic mire which started to developed on the fluvial terrace of the Pedrido River ~12,500 years ago (based on radiocarbon dating of the base of the core; data not shown).
Sampling was done with a Russian peat corer to a depth of 100 cm.To avoid disturbance of the peat core, two parallel hemi-cores were sampled at short distance (20-30 cm apart).The first core section included the upper 50 cm and the second one from 50 to 100 cm.Immediately after sampling of each section the temperature was measured every 2 cm using a Mini Temp Ray tec MT4.Then the sections were wrapped in plastic film, protected in PVC hemi-tubes and taken to the laboratory.Once in the laboratory both sections were cut into 2 cm slices and placed in polyethylene bags and preserved at 4 ºC in a freezer until analysis.

Characterization of the peat core
To do a basic physico-chemical characterization of the OBX core, the fresh samples were analyzed for water content, peat density, ash content and pH in water and KCl.We used the protocols described in (Pontevedra Pombal 2002).These results were compared with those obtained in a previous study of another core (TPD, sampled and analyzed in 2000) from the same bog and 20 m apart, in order to provide a general characterization of the vertical variations in the peat deposit.For the physico-chemical characterization of TPD, the above-mentioned protocols were also used.Carbon and N were determined on dry milled and homogenized samples using an elemental analyzer LECO TruSpec CHN.

Microbial metabolic fingerprinting
To study the microbial activity in the core we used Ecoplates from Biolog.Ecoplates are designed to obtain a community metabolic fingerprint which indicates differences in community composition (Campbell et al. 1997).Of the wide range of Biolog plates, we selected Ecoplates because they contain more ecologically relevant (i.e.root exudates) substrates.Each microplate contains 96 wells that are triplicates of 31 organic substrates, which can be grouped into six chemical guilds (Dobranic and Zak 1999) (Table 1) and a blank.The organic compounds were selected because they are highly discriminating between soil microbial communities (Hitzl et al. 1997), while the blanks were represented by 3 wells with water to serve as a control (Lagomarsino et al. 2007).When a substrate is used, a color is developed because microbial respiration reduces a tetrazolium dye that is included with the carbon source.Preston-Mafham et al. (2002), in their review of the application of sole-carbon-source utilisation profiles to analyse microbial functional diversity, gave five main recommendations: 1) to reduce the time between sampling and inoculation; 2) use equivalent sample sizes; 3) inoculate the plates with a soil suspension; 4) take multiple point readings to allow full kinetic analysis; and 5) take into account inoculum density.Of these we followed the first four recommendations, as described below.Although it may pose some limitations to the interpretation of our results, for this preliminary study of the application of Ecoplates to peat research we did not make an assessment of the inoculum density.We believe that this approach better reflects the conditions of the natural environment of the peat deposit.
With the aid of the temperature record we selected 11 peat sections (0-2 cm, 10-12 cm, 20-22 cm, 26-28 cm, 32-34 cm, 38-40 cm, 46-48 cm, 52-54 cm, 62-64 cm, 78-80 cm and 96-98 cm) as representative of the vertical variations.Three of the sections were done in duplicate (0-2 cm, 20-22 cm and 96-98 cm).Immediately after sampling -the same day-, fixed volumes of peat were taken from the center of each selected peat slice using a syringe (1.6 cm of diameter and 2 cm in thickness = 4.02 cm 3 of peat).Each plug was transferred to Falcon tubes and 30 ml of sterile MiliQ water were added.Samples were allowed to stir for 13 h.Then they were filtered through sterile quantitative analysis filters ALBET 140 (pore diameter 15-20 μm) and 100 μL of the water extracts were inoculated into each well of the Ecoplates.All used materials were sterilized using an autoclave at 121 ºC during 1 hour.
Monitorization of the changes in the consumption of the substrates was done measuring the absorbance at 590 nm with the aid of a microplate reader (Model 680, Bio-Rad Laboratories, Hercules, CA).The first measurement (time zero, T0) was taken immediately after inoculation.Then, the microplates were incubated at constant temperature (26 ºC) during 10 days.A preliminary test performed on peat that had been stored in a refrigerator showed no evidence of substrate consumption (i.e.development of color) in the first three days, thus the second measurement was set at 65 h (time 1, T1), and then readings were done twice every day.
Considering that microplates are only able to compare microbial functional diversity in different soils or states and not to obtain direct information on microbial community function (Preston-Mafham et al. 2002), we used the first peat section as a reference to compare those taken at the other depths because it showed the highest response (i.e.absorbance) and the largest number of substrates reacted.

Statistical methods and calculations
A principal components analysis (PCA) was performed on the average absorbance of each carbon substrate (after subtraction of the blank) for the readings obtained at 185 h of incubation, using a varimax rotation to maximize the loadings of the variables and provide the best separation among the components (Tabachnick and Fidell 1989).In some cases (5% of the whole data set), one of the three replicates of each microplate was found to be a gross outlier (Baxter 1999) and was not included to calculate the average.Only in a few occasions (1% of the data set) two of the replicates did not react at all but one showed color development.In these cases we used the absorbance of the reacting well, assuming that the presence of reaction indicates that for the given peat layer there must be microorganisms capable of substrate consumption.The PCA was done using SPSS 20.0 software.
To express the overall absorbance of the Ecoplates we calculated the AWCD (Preston-Mafham et al. 2002;Fisk et al. 2003;Stefanowicz 2006;Weber and Legge 2010) for each of the 12 reading times to get an approximation to the kinetics of the consumption of the substrates.Taking the advantage of the PCA results, we also calculated AWCD values for the most representative substrates (those with loadings greater than 0.7) of each component, as it will be described later.
As a measure of the number of substrates utilized (substrate richness) and diversity of the extent of utilization of particular substrates (substrate evenness) we applied the Shannon-Wiener index (H) (Zak et al. 1994;Stefanowicz 2006) and the transformation proposed by Jost (2006), also known as "effective number of species" (MacArthur 1965), to obtain true diversity measures.

Physico-chemical characterization
Peat temperature showed a continuous decline with the depth (Figure 2) from a maximum of 14.8 ºC at the surface to a minimum of 6.8 ºC at 94-98 cm of depth, with a small increase at 40-52 cm (up to ~11 ºC).From 70 to 100 cm, the temperature was almost constant.
The depth variation of the volumetric water content (WC) (Figure 2) shows two distinct sections.In the upper 55 cm of the core WC values were almost constant with an average value of 54.8±6.7%.Below 55 cm average WC increased to 72.2±11.4% but showed a larger variability.The minimum value was found at the surface (44%) and the maximum at 98-100 cm (94%).
Figure 2 shows the pH in water (pHw) and in KCl (pHk) values with depth.As expected, pHw is higher than pHk, but in both cases they showed little variation with depth.pHw varies between 3.98 and 4.98, while pHk is between 1.5 (in the upper 10 cm) and 1.2 units lower.
Peat bulk density (BD) ranged between 0.08 and 0.23 g cm -3 , the higher values were found in the top 25 cm of the core (Figure 2).Ash content is typically higher in the upper 18 cm (6-12%) and below it is fairly stable (around 2%).
The depth records of the physico-chemical properties measured in both cores (OBX and TPD) can be found in the supporting material.The quite similar BD and ash content records (r 0.87 and 0.79 respectively) indicate that there is a good stratigraphical correlation between the OBX and the TPD cores, and thus other properties analyzed in the later can also be of help to characterize the changes with depth in the OBX core; for example organic carbon (C), and nitrogen (N) contents and C/N ratios (supporting material).Organic carbon content ranges between 45% and 52.9%, and N between 1.32% and 2.64%.Both are inversely correlated (r = 0.92).Organic carbon content increases and N content decreases rapidly from the surface to 40 cm of depth.Below 40 cm contents show little variation.The C/N ratios record is almost identical to that of C (r 0.93), with values varying between 17.5 at the surface and around 30-40 below 40 cm (supporting material).

Microbial functional diversity: Ecoplates
The results of absorbance at the end of the incubation period for each substrate (average of the three replicates) and depths analyzed are shown in the supporting material.Substrates codes are according to Table 1.In the upper peat sample all compounds reacted except for γ-hydroxybutyric acid (E3); while low absorbance was found for D,L-a-Glycerol phosphate (H2).Also, almost all substrates had the highest reactivity (i.e.absorbance) in the upper peat sample.The other depths analyzed showed a large variability in the degradation of the substrates.For example, the peat section at 10-12 cm showed low activity (i.e.average absorbance) compared to the upper ones.Samples corresponding to depths 20-22 cm and 26-28 cm presented high to medium absorbance values for most polymers and almost half of the carbohydrates (see supporting material).Samples of depths 32-34 cm and 38-40 cm showed almost no substrate utilization.Moreover at 46-48 cm and 52-54 cm there was a relatively high activity in half of the substrates (see supporting material).
Eleven substrates (mostly amino acids, polymers and carbohydrates) reacted at greater depth (until 48 cm) (see supporting material).Of these, only Tween 40 (C1) and N-acetyl-D-glucosamine (E2) showed a significant reaction until 54 cm.Peat samples below 54 cm showed no activity in the whole incubation period, except for glycogen (F1).
The Shannon-Wiener index (H) of entropy at the end of the incubation period showed a similar pattern to the AWCD (Table 2).The upper sample (0-2 cm) had a value of 3.3; intermediate depths (10-12 cm, 20-22 cm, 26-28 and 46-48 cm) had values between 1.2 and 1.8.The remaining samples had values lower than 0.8.The AWCD and H are highly correlated (r = 0.98).In both cases, duplicate samples showed the same variation among them that was observed for the absorbance values.
The difference between the maximum and minimum D values is greater than those of H and AWCD, resulting in a larger differentiation between samples.The upper sample (0-2 cm) had an average value of 27.6; intermediate depths (10-12 cm, 20-22 cm, 26-28 and 46-48 cm) had values between 2.2 and 5.9.The remaining samples had values lower than 1.5 (Table 2).
Considering the number of substrates that have reacted, AWCD, H and D values, the duplicates of surface (0-2 cm) and deeper samples (96-98 cm) presented quite similar results (Table 2).On the other hand, the duplicate of the sample 20-22 cm showed quite different values (almost a 2-fold difference).

PCA analysis of functional diversity
A certain pattern of substrate degradation is intuited from Table 2, but to reduce the dimensionality of the data set and to investigate into the causes of depth variability, we performed a principal components analysis (PCA) on the average absorbance of the substrates for all peat depths at the end of the incubation period.Substrate E3 (γ-hydroxybutyric acid) was not considered because of its lack of reaction.Four components explained 91.2% of the total   4).
To provide a visual representation of the changes of substrate utilization with depth at the end of the incubation period, a coloured map of relative absorbance for each compound at each depth is given in Figure 3. Values of relative absorbance were calculated with respect to the uppermost peat sample (0-2 cm) that, as already mentioned, showed the highest values for all substrates (except for A4, E2 and F2).
The substrates were grouped according to the principal component to which they contribute.Substrates characteristic of Cp1 reacted only in the superficial peat sample; those in Cp2 also reacted in the sample at 46-48 cm; while those in Cp3 are most of the substrates that reacted at greater depths.The only substrate in Cp4 (F1, Glycogen) only showed a significant reaction at the upper peat sample (0-2 cm) and was included in Figure 3 among the substrates of Cp1.

Kinetic of the substrates
AWCD results for all depths and readings presented four main kinetic patterns.The surface sample (0-2 cm) showed an asymptotic (Table 5) evolution leveling off after 100 hours, to reach a maximum absorbance of 1.7 at 200 hours (Figure 4); samples of depths 10-12, 20-22, and 26-28 cm showed an exponential increase (Table 5) but with very low end values of AWCD (0.3-0.5); absorbance in sample at 46-48 cm increased linearly with time (Table 5) to a maximum value of 0.62; while the other samples did not show any reaction or pattern, and AWCD values were very low.
Given that the PCA enabled us to identify three main depth distributions, we also calculated the AWCD for selected substrates representative of each component (those with loadings greater than 0.7; see Table 3. Substrates of Cp1 showed two kinetic patterns (Figure 4): that of the surface sample (0-2 cm) can be approximated to a sigmoidal (Table 5) model; while in the other samples the substrates did not show any activity.Substrates of Cp2 and Cp3 (Figure 4) presented the same kinetic models as those described for the AWCD calculated with all substrates, with slightly higher maximum AWCD values in Cp3 than in Cp2, except for the sample at 46-48 cm of depth for which it was similar (Figure 4).
As mentioned before, AWCD and H index are highly correlated and therefore the latter fits exactly to the same models, providing any additional information on the kinetics of microbial functional diversity in our study case.4. Discussion

Physico-chemical properties
The main objective of the properties analyzed was to characterize the chemical environment in the upper meter of the sampled peatland and help determine its possible link to the observed microbial functional diversity.Regarding temperature, we did not find any published data, neither reference values nor records of vertical variation, to compare with ours.Although the measurements were done in only one core and after coring, and thus their interpretation should be taken with care, the data seem to be appropriate because the values show a reasonable depth trend of overall exponential (r = 0.96) decreasing temperature.
For In the core studied by us and in published ones, the depth variation in WC, C and N contents, and C/N ratios, suggests a change in the geochemical environment of the peat deposit that could correspond to the boundary between the acrotelm -the aerobic, acidic, upper section of the peat deposit (with higher susceptibility to oxidative microbial activity)-, and the catotelm -the anoxic, less acidic and thicker section of the peat deposit.The comparison with published records (Pontevedra Pombal et al. 2006) suggests that this boundary is located at different depths in each peat deposit, between 25-60 cm, and the OBX core falls within this range.Nevertheless, some of the observed differences (water content, bulk density and ash content), are probably due to the fact that all records (ours and published ones) represent single measurements in the time domain.Thus, the actual local and temporal variability of these properties has not been assessed and it is likely that variation within the mire is greater (or at least comparable) than between mires.

Microbial functional diversity
Regarding microbial activity, the maximum values of absorbance we obtained are similar or higher to those given for different experiments in the literature (between 1.6-2.7)(Campbell et al. 1997;Preston-Mafham et al. 2002), although most publications do not provide the raw results (e.g.Fisk et al. 2003).AWCD values, integration of the overall microplate absorbance at the end of the incubation period, are comparable to those found in the literature: 0.4-1.1 at 150h (Campbell et al. 1997) (2006), transformed index usage is not common in Ecoplates methodology.
The relative variation between duplicates (e.g. between 0-2 cm and 0-2 cm r, see Table 2) analyses found in our experiment may be due to i) spatial heterogeneity, as suggested by Barkovskii et al. (2009) who found that there are differences in organic-matter-driven diagenesis depending on the bacterial fraction, or ii) it may be because the inoculation was not equally successful.
The PCA results showed that there were four patterns of substrate degradation.As a synthesis it can be said that i) some of the most simple C-sources (those grouped in Cp1) showed maximum reaction at the surface of the mire and no reaction at the other depths; ii) four of the carboxylic acids-substrates (Cp2) were preferentially degraded at the surface of the peatland and right above the limit of the acrotelm/catotelm boundary; iii) the more complex C-sources (polymer-substrates in Cp3) and some amino-substrates showed reactivity to a greater depth; and iv) no reaction was found below 52-54 cm.No comparable experiments were found in the literature, and thus, to our knowledge, these are the first results on vertical variations of microbial functional diversity in peatlands using sole-C-source utilization profiles.Fisk et al. (2003) is the only publication we were able to find dealing with an experiment of microbial functional diversity in peatlands, but they compared reactions on superficial samples in different locations and mires and not with depth, and their results are not directly comparable because they used a different type of Biolog microplate.
The degradation of N-containing substrates at greater depth is also consistent with findings by different researchers (Bridgham et al. 1998;Aerts et al. 1999), who showed that N cycling in peatlands is quite active due to its nature as limiting nutrient.A similar reasoning can be applied to the significant consumption of the more complex substrates (polymers Tween 40 and 80) in the upper half meter of the peat deposit.Research on the depth changes in C-groups and molecular composition of the peat (Pontevedra Pombal 2002;Buurman et al. 2006;Kaal et al. 2007;Barkovskii et al. 2009) showed an exponential decrease of polysaccharides with depth; thus, it is likely that the most labile compounds (including polysaccharides) are preferentially degraded in the upper part of the peat deposit while the more complex and recalcitrant ones are still degraded at greater depth.As for the preferential degradation of some carboxylic acids at the acrotelm/catotelm boundary, we do not have an explanation and it can only be speculated that oxygen availability and temperature (which showed a secondary maximum at this depth) may have played a role.
Taken together, the patterns of microbial functional diversity and the records of the physico-chemical properties and elemental composition suggest that the acrotelm/ catotelm boundary is an effective one for oxidative degradation of organic substrates.Almost all substrates were degraded in the upper sample, where higher temperature and, likely, highest oxygen availability may have promoted a maximum in microbial activity in the peat column.But the change of geochemical environment (oxygen availability, temperature, nutrients, organic matter quality, etc.) with depth results in increasing limitations to bacterial growth.On the other hand, studies using more sophisticated techniques (like PLFAs, T-RFLP, FISH) also found that microbial communities and associated decomposition processes of all peatlands types are vertically stratified, as redox conditions and carbon quality change (Williams and Crawford 1983;Sundh et al. 1997;Morales et al. 2006;Dedysh et al. 2006).This all supports the suggestion of Andersen et al (2013) that the vertical stratification of microbial communities in peatlands arises primarily from energy constraints.
Thus, our results are consistent with previous knowledge on physico-chemical and geochemical properties and the functioning of peatlands.Nevertheless, as we commented in the methods section, no assessment of the density of the inoculum was done.This may have implications on the results that can be obtained using Ecoplates, as differences in carbon consumption at different depths with different, standardized, microbial biomass are expected.On the other hand, our aim was to more closely reflect the conditions of the natural environment.As found by Barkovskii et al. (2009), there can be large differences in bacterial fractions and composition at short distances.

Kinetics of microbial functional diversity
AWCD overall results showed that substrate degradation followed four kinetic patterns: asymptotic, exponential, linear and no reaction with time.In the literature, the general kinetic pattern is sigmoidal (Campbell et al. 1997;Preston-Mafham et al. 2002;Stefanowicz 2006) and therefore we believe our results should also agree with a sigmoidal evolution.The lack of coincidence is possibly due, on one hand, to the absence of measurements in the first 50 hours.
The kinetics of substrates grouped based on the results of the PCA showed certain differences respect the general model, including two main features: intensity of substrate utilization and total time of reaction.The results for the Cp1-substrates in the upper sample fit to the expected sigmoidal model, thus indicating that the asymptotic model of the superficial sample obtained with the pooled substrates (overall AWCD) is in fact dominated by the kinetics of Cp2-and Cp3-substrates, and not by Cp1substrates.The results also suggest that full substrate reaction may have only been reached in the uppermost peat sample, with larger values for Cp1-than Cp3-and Cp2-substrates both in intensity (i.e higher overall AWCD, 2.7 vs 1.4-1.7)and utilization time (asymptotic behavior at ~200 hours and ~100 hours respectively).Campbell et al. (1997) also studied the kinetic of the reaction depending on the chemical composition of the substrates, but their results are hardly comparable since, as we have found in our study, substrates of the same chemical group may show different behaviour and therefore integrated results may not be representative.Previous PCA analysis seems to be a proper statistical strategy to overcome this problem.

Conclusions
Our analysis of microbial functional diversity using sole C-source utilization profiles applied to a peat core (OBX) sampled in the Tremoal do Pedrido raised bog showed that there are significant variations with depth in the degradation of Ecoplates carbon sources.These variations were reflected both by the AWCD and the Shannon-Wiener index (H) of diversity (which were highly correlated) and more accentuated on the "effective number of species" (D).
The degradation patterns we observed indicate that some of the most accessible C-sources may be rapidly used at the surface of the mire while N-compounds and the most complex substrates are still degraded at great depth.Some differences were observed, both in intensity of substrate degradation and total time of reaction, when substrates were grouped according to the results of the principal components analysis.
The results suggest that there can be a "stratification" of the microbial communities, so that microorganisms which preferentially utilize simple sources may concentrate in the surface of the mire while those able to degrade more complex compounds (which are left) are still active in the deeper sections of the acrotelm.
It also seems that oxygen availability and temperature at specific depths may influence the biodegradation of carboxylic acids.We did not find substrate utilization below the acrotelm/catotelm boundary, suggesting that this boundary is an effective one for oxidative degradation of the peat organic matter.Although this is consistent with previous research on peat organic chemistry we do not actually know if the position of the boundary is a seasonal feature (as it may be expected).
Interpretation of results must be cautious because the use of Ecoplates implies selecting a very specific part of the microbial population (metabolically active and aerobic bacteria capable of growing at specific lab conditions).Also, adding more uncertainty to the results is the fact that Ecoplates allow the characterization and not only the comparison of microbial communities, because of the concentration and type of carbon substrates (as also discussed by Bossio andScow 1995, andPreston-Mafham et al. 2002).Therefore a study using a selection of peat representative organic compounds, such as those characterized by pyrolysis-GC/MS in previous studies (Buurman et al. 2006), as carbon substrates and at different lab conditions could allow better information to be obtained on the microbial activity in peatlands.Further research is needed to determine the influence of others factors such as inoculum density, incubation time, temperature and spatial and seasonal variability within the mire.Otherwise it could be interesting to combine Ecoplates with other techniques, to provide information on the relative contribution of different microorganisms to the dynamics of organic matter in peatlands.• Fontúrbel MT, Barreiro A, Vega JA, Martín A, Jiménez E, Carballas T, Fernández C, Díaz-Raviña M. 2012.Effects of an experimental fire and post-fire stabilization treatments on soil microbial communities.Geoderma 191:51-60.

Figure 1 .
Figure 1.Location map of the Xistral Mountains and the Tremoal do Pedrido mire.

Figure 3 .
Figure 3. Colour-coded map of relative absorbance (the upper sample taken as reference).The substrates are grouped according to the principal components (compounds with loadings greater than 0.7).

Figure 1 .
Figure 1.Supporting Material Depth records of: ash content (Ash); bulk density (BD); carbon (C) and nitrogen (N) contents; carbon/nitrogen ratio (C/N).Circles OBX core, squares TPD core.The shadows indicate the depths at which the samples were taken for Ecoplates.

[
PÉREZ-RODRÍGUEZ M. & MARTÍNEZ CORTIZAS A. ] used BIOLOG plates to characterize the patterns produced by different microbiological samples from root exudates while Lupwayi et al. (1998) used them to investigate the effects of tillage and crop [ PÉREZ-RODRÍGUEZ M. & MARTÍNEZ CORTIZAS A. ]

Table 1 .
Carbon sources used as substrate in Ecoplates.Substrates are classified in chemical guilds according toDobranic and Zack (1999)

Table 2 .
Diversity index at the end of the incubation period for each sample Average well colour development (AWCD); Shannon-Wiener index (H); Effective number of species (D); Duplicates (r).

Table 3 .
Substrates grouped according to PCA analysis performed with AWCD values at the end of the incubation period Component loadings (Cp1, Cp2, Cp3 and Cp4) > 0.7 are in bold.γ-hydroxybutyric acid and water are not included.Substrates ecoplate code (Cd); Chemical guild (CG).

Table 5 .
Statistical summary of the equations of the kinetic patterns Equations (Y corresponds to the AWCD of the absorbance at 590 nm, except for the last equation in which it corresponds to the scores of the first principal component; t is time in hours).

Table .
Supporting material 1.Average absorbance values at the end of the incubation period for each sample Substrates code (A1, A2 ... H4) according to (Table 1); Duplicates (r).[ PRELIMINARY CHARACTERIZATION OF MICROBIAL FUNCTIONAL DIVERSITY USING SOLE-C-SOURCE UTILIZATION PROFILES IN TREMOAL DO PEDRIDO MIRE (GALICIA, NW SPAIN) ]