To our knowledge this is the first study to compare variation in microbial, biochemical and micromorphological indicators in a cut-over peatland where the surface is in different stages of regeneration. Both biological and biochemical indicators were consistent in showing clear differences between the profiles from the unexploited part of the peatland and those from its regenerating parts.
testate amoebae and bacteria indicators
Nevertheless, to date testate amoebae have only rarely been included in studies of cut-over peatland regeneration (Buttler et al. 1996; Davis & Wilkinson 2004). In a study of naturally regenerating cut-over bogs in the Jura Mountains, Buttler et al. (1996) observed that testate amoebae responded rapidly to changes occurring at the surface. This study also provided data on recent stages of succession not included in this study. The authors observed a convergence of vegetation and testate amoeba community structure in the advanced regeneration stage, regardless of the initial conditions.
In this study, we observed clear differences in species richness, diversity, density, biomass and average species size. Species richness and diversity increased but density declined from the recent to the advanced regeneration stages and the unexploited site. This result is in agreement with patterns of community assembly of various groups of organisms during primary and secondary succession (Odum 1971). However, recent studies of testate amoeba diversity and community structure in chronosequences, and relationships between plant and testate amoeba diversity, have shown contrasting responses, suggesting that testate amoebae might not respond in the same way as larger organisms (Ledeganck, Nijs & Beyens 2003; Wanner & Xylander 2005). In the cut-over peatland secondary succession sequence, we found shifts in community composition, rather than simply an addition of new taxa as observed by Wanner & Xylander (2005) in sand dunes. However, peatlands are quite different from sand dunes and mineral soils in their evolution. It can be assumed that the changes in ecological conditions (e.g. moisture and pH) associated with the development of a new, actively growing peat layer acts as a strong ecological filter that causes early colonizers to disappear from the community.
Biomass and the average size of species declined over the regeneration sequence but were higher in the unexploited site of the peatland. These changes also agree well with the changes in ecological conditions over the regeneration sequence. Testate amoebae are aquatic organisms and respond to soil moisture content in a size-specific way: larger species are more stimulated by wet conditions than smaller ones (Lousier 1974). However, larger species are also less numerous and heavier than smaller ones and therefore they are less likely to be transported over long distances. In a broader context, it has been suggested that testate amoebae larger than 100–150 µm may not be cosmopolitan (Wilkinson 2001). The processes responsible for such biogeographical patterns probably affects the recolonization of secondary habitats at a finer scale. Therefore we would not expect large species to colonize favourable habitats very quickly. This hypothesis is supported by the available data on primary succession. In a series of primary colonization experiments in mineral soils, Wanner and co-workers have observed that all successional stages are dominated by small Euglyphid testate amoebae, while larger taxa characteristic of forest humus only occur in late succession stages (Wanner & Dunger 2001; Wanner & Xylander 2005). Such differences in the size and quality of the testate population may serve as predictors of the rate and directions of change as the regenerating peat community becomes more established. Beyond the quantitative inference of key ecological variables, the structure of the testate amoeba community might provide information on the degree of ‘naturalness’ of a site.
Bacterial biomass is a relatively crude measure of microbial activity in ecosystems including secondary succession in cut-over peatlands. Nevertheless clear changes were observed. Beyond biomass, changes in bacterial community structure and associated processes can be expected. A community approach for bacteria was beyond this study. However, low densities of bacteria were recorded during the apparently drier phases (see below) and in the more advanced regeneration stages (albeit not significant). Similarly, Gilbert (1998) observed lower chemoheterotrophic assimilation (mainly bacterial) during the dry period of mid-summer in a Sphagnum-dominated peatland. This apparent negative effect of dry conditions on bacteria density and production parallels the pattern of testate amoeba density where low numbers were found in the more advanced, drier secondary sites and in the unexploited site. Testate amoebae feed on a broad range of micro-organisms, thus the lower density of testate amoebae in the drier sites matches the density patterns of at least some of their prey (the bacteria) and microbial secondary production (Yeates & Foissner 1995; Gilbert et al. 2000). These results could also suggest that the larger species of testate amoebae that are characteristic of the unexploited site may be less directly dependant on the abundance of bacteria and instead feed more (or perhaps exclusively) on fungi. Such an assumption is commonly made, although there is still little reliable data on the feeding habits of most testate amoeba species.
organic matter and biochemical indicators
The high preservation of organic material in peat that results from low pH and anoxia make the peat archives particularly useful for palaeoenvironmental reconstructions. Nevertheless, to date the biochemical composition of peat OM has rarely been used as an indicator for past environmental conditions, particularly in ombrotrophic peatlands (Morita & Montgomery 1980; Nott et al. 2000; Pancost et al. 2002; Nichols et al. 2006), and none of these studies concern formerly cut-over sites. Recently, a study of a regenerating cut-over bog in the Jura Mountains allowed Comont, Laggoun-Défarge & Disnar (2006) to obtain insights into changes in the OM sources and dynamics of inherited biopolymers along the regeneration sequence.
Field observations did not suggest that the ‘old’ peat beneath the regenerated peat was different between the recent and advanced regeneration stages. However, from detailed microscopic observations the proportion of Sphagnum (either well-preserved or not) was higher in the catotelm peat of the advanced stage than of the recent stage. In the recent stage, in contrast, the peat contained a higher proportion of unidentifiable remains. However, the biochemical signature revealed that, in both stage, the new peat was of comparable origin. We interpret this as an indication of a similar original composition but differential preservation because of the conditions of the sites at the onset of regeneration
At the unexploited site, the irregular but overall progressive decrease of total sugars with increasing depth depicts typical diagenetic evolution. Nevertheless, the high and nearly constant C/N ratio values (i.e. 60–80) recorded along the peat profile, and the abundance of well-preserved tissues mainly derived from Sphagnum mosses, are typical of rather well-preserved inherited plant material. In contrast, the two sections taken between 20–25 and 48–62 cm depth that display much lower C/N ratios, lower total sugar yields and OM dominated by decomposed plant tissues, suggest an increasing degradation of OM. At the top of these two sections well-preserved Cyperaceae-derived tissues replaced the Sphagnum-derived tissues. These features suggest a change in vegetation and environmental conditions that might have been provoked by drier phases in the history of the bog. Such a dry event would have shifted the competition between Sphagnum and Eriophorum in favour of the latter and increased peat mineralization. The causes of these two dry phases are uncertain but drainage from peat cutting is most likely to be responsible for changes recorded in the upper peat. Taken together, these results illustrate well the fact that, although this part of the bog has not been exploited for peat, drainage related to peat harvesting affected the vegetation and therefore the botanical composition of the peat. These changes were well recorded in the existing peat.
In the regenerating sites (sites 1 and 4), vertical patterns of OM composition revealed a limit between the upper ‘new’ peat and the lower ‘old’ catotelm peat. The latter, especially at site 1, was characterized by a pronounced OM degradation, as attested by relative low C/N ratios and sugar contents and a predominance of amorphous OM and mucilage. In contrast, the new regenerated peat was dominated by moss-derived tissues. This was confirmed by distributions of individual hemicellulose sugars displaying high proportions of mannose and, to a lesser extent, galactose compounds typical for mosses (Comont, Laggoun-Défarge & Disnar 2006). The new peat also exhibited much higher yields of total hemicellulosics in comparison with total cellulosic sugars. Such a discrepancy might reflect a higher contribution of moss to the peat, these plants being richer in hemicellulosic sugars than sedges (Comont, Laggoun-Défarge & Disnar 2006). However, a relative enrichment of the hemicellulosic carbohydrate pool as a result of cellulose destruction cannot be excluded. Nevertheless, the amount of total sugars recorded in this peat layer, which were in the same range as in living plants, are indicative of a good OM preservation. Surprisingly, although the vegetation cover is currently dominated by mosses and sedges, no evidence of any Cyperaceae-derived material, and/or related sugar biomarkers, was identified by analyses. In fact, sugar markers of Cyperaceae, i.e. xylose and arabinose (Bourdon et al. 2000), were not present in the new peat but in the upper levels of the old catotelm peat (Fig. 5). This overall lack of Eriophorum record in the new peat can be attributed to its higher decomposability compared with Sphagnum mosses (Coulson & Butterfield 1978; Clymo & Hayward 1982; Chague-Goff & Fyfe 1996).
Over the secondary succession, a close examination of organic composition of the new peat revealed changes from the recent to the advanced stages. At site 1 (the recent stage) the peat was dominated by Sphagnum remains, while at site 4 (the advanced stage) it had a more heterogeneous botanical composition, with better carbohydrate preservation (c. 337 mg g−1 vs. c. 243 mg g−1 at site 1). In addition to the original botanical composition, such contrasting composition might also be related to abiotic factors, i.e. trophic conditions inducing differences in biodegradation processes between the two sites. The surface vegetation suggests that the environmental conditions of the recent regenerating stage (site 1) are probably more minerotrophic and consequently more favourable to microbial activity than the more advanced regenerating stage (site 4) (E. Samaritani, A. Siegenthaler, Y. M. li-Petäys, A. Buttler, P.-A. Christin and E. A. D. Mitchell, unpublished data). This explanation was supported by (i) the bacterial biomass, which was about twice as high in the new peat of the recent regenerating stage as in the advanced stage and in the unexploited site, and (ii) a shift in testate amoebae from a fen community towards a more acidic, drier bog community. In the same way, when considering the whole profiles (new and old peat), it appeared that the highest bacterial biomass was recorded in the Sphagnum-dominated peat layers and the lowest in the highly decomposed and deeper peat layers. This was interpreted, at least at site 1, as a consequence of drainage phases during peat extraction.
conclusions and implications for management
Our aims were to assess how testate amoebae, bacteria and peat OM were correlated and to identify specific indicators of changes in the structure or functioning of the ecosystem. While bulk chemical OM characterization allowed the newly regenerated peat to be differentiated from old peat, OM indicators (carbohydrates and botanical composition of the peat) combined with heterotrophic bacteria biomass and testate amoebae diversity revealed contrasting signatures between the recent and advanced stages of regeneration. At the natural unexploited site, specific OM indicators provided information on past changes in vegetation and related environmental conditions, well recorded in the accumulated peat.
This study illustrates how biochemical markers and testate amoebae could provide additional information on the functioning of the ecosystem as well as the observation of the present vegetation, which is commonly used to assess the state of the ecosystem. Testate amoebae appear to be particularly useful because (i) they provide information on the soil biota, (ii) they are preserved in the peat deposits, thus allowing palaeoenvironmental reconstruction, and (iii) their analysis does not require expensive equipment or consumables. Indicators from OM also appear useful because they allow the botanical and biochemical composition of even quite decomposed peat to be determined. Unlike accepted ideas regarding the rapid consumption of carbohydrates in modern environments, their good preservation in peats provides additional information on past anthropogenic perturbations in bogs and their consequence in terms of OM recycling and storage.
Understanding ecosystem dynamics in secondary ecosystems is challenging because we rarely have accurate information on the nature of the ecosystem prior to disturbance and a detailed account of human impact. It is valuable to compare independent lines of evidence to determine such characteristics of the site. A multidisciplinary assessment approach may therefore prove useful for the management of abandoned cut-over peatlands.