For vascular plants nomenclature follows Mossberg et al. (1992) (subarctic and cool-temperate regions), Van der Meijden (1996) and Van der Meijden et al. (1983) (warm-temperate region); for Sphagnum mosses nomenclature follows Bouman & Van der Pluijm (2002).
Are growth forms consistent predictors of leaf litter quality and decomposability across peatlands along a latitudinal gradient?
Article first published online: 29 JUN 2005
Journal of Ecology
Volume 93, Issue 4, pages 817–828, August 2005
How to Cite
DORREPAAL, E., CORNELISSEN, J. H.C., AERTS, R., WALLÉN, B. and VAN LOGTESTIJN, R. S.P. (2005), Are growth forms consistent predictors of leaf litter quality and decomposability across peatlands along a latitudinal gradient?. Journal of Ecology, 93: 817–828. doi: 10.1111/j.1365-2745.2005.01024.x
- Issue published online: 29 JUN 2005
- Article first published online: 29 JUN 2005
- Received 15 December 2004 Revision accepted 16 March 2005 Handling Editor: Richard D. Bardget
- climate change;
- environmental gradient;
- high latitude;
- leaf litter;
- litter chemistry;
- nitrogen deposition;
- plant functional types;
- Top of page
- Supplementary material
- 1Plant growth forms are widely used to predict the effects of environmental changes, such as climate warming and increased nitrogen deposition, on plant communities, and the consequences of species shifts for carbon and nutrient cycling. We investigated whether the relationship between growth forms and patterns in litter quality and decomposition are independent of environmental conditions and whether growth forms are as good as litter chemistry at predicting decomposability.
- 2We used a natural, latitudinal gradient in NW Europe as a spatial analogue for future increases in temperature and nitrogen availability. Our screening of 70 species typical of Sphagnum-dominated peatlands showed that leaf litters of Sphagnum mosses, evergreen and deciduous shrubs, graminoids and forbs differed significantly in litter chemistry and that the ranking of the growth forms was independent of the region for all litter chemistry variables. Differences among growth forms were usually larger than differences related to the environmental gradient.
- 3After 8 and 20 months incubation in outdoor, Sphagnum-based decomposition beds, growth forms generally differed in decomposability, but these patterns varied with latitude. Sphagnum litters decomposed slower than other litters in all regions, again explaining its high representation in organic deposits of peatlands. Forb litters generally decomposed fastest, while the differences among the other growth forms were small, particularly at higher latitudes.
- 4Multiple regression analyses showed that growth forms were better at predicting leaf litter decomposition than chemical variables in warm-temperate peatlands with a high N-load, but less so in the subarctic, low-N region.
- 5Our results indicate that environmental changes may be less important in determining ecosystem leaf litter chemistry directly than are their indirect effects through changes in the relative abundance of growth forms. However, climatic and nutritional constraints in high-latitude peatlands promote convergence towards nutrient-efficient plant traits, resulting in similar decomposition rates of vascular growth forms despite differences in litter chemistry. The usefulness of the growth-form concept in predicting plant community controls on ecosystem functioning is therefore somewhat limited.
- Top of page
- Supplementary material
Changes in climate due to increasing concentrations of glasshouse gases are expected to be most pronounced at northern high latitudes (Houghton et al. 2001). Subarctic and boreal peatlands contain about one-third of the total world soil carbon pool and have acted as a long-term carbon sink, currently estimated to take up 0.07 Pg carbon year−1 (Gorham 1991; Clymo et al. 1998). Increases in temperature, and the subsequent increases in nutrient availability (Rustad et al. 2001), or changes in precipitation may alter the balance between carbon uptake and release through changes in biomass production and decay, which may have important consequences for the global atmospheric carbon balance. Decomposition of dead organic matter regulates the release of nutrients for plant growth as well as the release of carbon into the atmosphere. It is therefore a key process in the responses and feedback of northern peatlands to climate change.
The rates of decomposition processes are directly related to temperature and moisture or nutrient availability (Robinson et al. 1995; Hobbie 1996; Aerts 1997; Ågren et al. 2001; Liski et al. 2003), but are also indirectly influenced through their effects on litter quality (Fig. 1, relationships 1 and 2; Coulson & Butterfield 1978; Robinson et al. 1995; Aerts 1997; Shaw & Harte 2001). Both experimental warming and increased nutrient availability have been shown to affect litter N and P concentrations, C : N and C : P ratios, and carbon, lignin and phenolic concentrations (Graglia et al. 2001; Shaw & Harte 2001; Dormann & Woodin 2002; Van Heerwaarden et al. 2003), although the responses of litter chemistry to environmental changes may vary among individual species (Aerts & De Caluwe 1997). At the ecosystem level, however, changes in climate and nutrient availability may also affect decomposition indirectly through changes in community composition (Fig. 1, relationship 3). Experimental warming or fertilization of northern peatlands and heaths has been shown to induce strong shifts in species abundance (Chapin et al. 1995; Press et al. 1998; Weltzin et al. 2003). Plant species in peatlands clearly differ in their decomposition rates (Coulson & Butterfield 1978), and the consequences of species shifts for decomposition at an ecosystem level may be as strong as direct temperature effects (Hobbie 1996; Shaw & Harte 2001).
In order to compare and predict plant community responses to changes in environmental conditions, it has become customary to classify species into functional groups (Gitay & Noble 1997; Arft et al. 1999; Dormann & Woodin 2002). In cool and cold biomes a classification based on plant growth forms (Chapin et al. 1996) is often used and this distinguishes trees, deciduous shrubs, evergreen shrubs, sedges, grasses, forbs, Sphagnum mosses, non-Sphagnum mosses and lichens. Differences in maximum height, responses to temperature, access to moisture and nutrients, the length of their photosynthetically active period and their effects on albedo, carbon gain, decomposition and methane emission are therefore reflected. Cluster analysis of 37 tundra species, based on traits that were expected to influence ecosystem processes, including responses to climate, resource acquisition rates and factors affecting nutrient use, resulted in a similar classification (Chapin et al. 1996). For functional groups based on growth form to be a useful tool in predicting the responses of ecosystems to changes in climate or nutrient availability, however, growth forms must differ consistently in their effects on ecosystem properties, such as litter chemistry and decomposition, and these differences should be independent of environmental conditions, such as climate and nutrient availability.
Screenings of large numbers of species from a range of ecosystem types showed that growth forms indeed account for part of the variability in leaf litter decomposability (Cornelissen 1996; Cornelissen et al. 1999; Pérez-Harguindeguy et al. 2000; Quested et al. 2003). However, all species within a single ecosystem type are essentially constrained by the same environmental conditions, and this may reduce both the diversity and the number of species and growth forms. Plant growth in many high-latitude and peatland ecosystems, for example, is strongly nutrient limited, and mature leaf N and P concentrations of all growth forms are lower in peatlands than in other terrestrial ecosystems (Aerts et al. 1999). This low nutrient availability may decrease, or even obscure, differences in litter chemistry and decomposition among the growth forms. It therefore remains to be tested whether consistent relations between growth forms, litter quality and decomposability can be observed within single ecosystem types (Fig. 1, relationships A and B). Furthermore, individual species within growth forms have been shown to respond differently to experimental warming or fertilization, both in litter chemistry (Shaw & Harte 2001; Van Heerwaarden et al. 2003) and decomposability (Robinson et al. 1995; Hobbie 1996). It is therefore not clear whether patterns such as the relative ranking of growth forms according to a chemical variable or decomposition rate are independent of environmental conditions (Fig. 1, interactions C and D).
We therefore investigated the consistency of patterns in leaf litter chemistry and decomposability among growth forms in three regions, using a latitudinal gradient in NW Europe (52°−68° N) as a spatial analogue for environmental change. We confined ourselves to peatland ecosystems and selected 70 species of the growth forms that are most commonly encountered in those systems, viz. evergreen shrubs, deciduous shrubs, forbs, graminoids and Sphagnum mosses. We therefore included both direct and indirect effects of changes in climate and nutrient availability, through differences in growing conditions (temperature, precipitation, nitrogen deposition), changes in species composition (within the growth forms) and differences in decomposition conditions (climate, decomposer community). We hypothesized (i) that growth forms differ in their litter chemistry and decomposability within peatlands, although the differences may be small, and (ii) that the patterns among the groups are not affected by differences in climate or nutrient availability. For several decades, leaf litter decomposability has been related quite successfully to various variables describing initial ‘bulk’ litter chemistry (e.g. total N, P, C, phenol and lignin concentrations and various ratios: Coulson & Butterfield 1978; Swift et al. 1979; Melillo et al. 1982; Hobbie 1996; Aerts & Chapin 2000; Pérez-Harguindeguy et al. 2000; Aerts et al. 2003). As growth forms differ in both nutrient and carbon uptake and allocation strategies, we further hypothesized (iii) that growth forms would be as good at predicting decomposability as litter chemistry.
- Top of page
- Supplementary material
study areas and species
Our study regions ranged from 68° N in the north to 52° N in the south of NW Europe (Fig. 2), hereafter also referred to as subarctic (north Sweden), cool-temperate (south Sweden) and warm-temperate (the Netherlands and Belgium). The gradient is characterized by strong differences in temperature and precipitation, as well as in nitrogen deposition (Table 1).
|Mean annual temperature (°C)||−0.5*||7.1‡||9.8§|
|Mean summer temperature (°C)||6.4*||13.0‡||14.4§|
|Mean winter temperature (°C)||−7.4*||1.1‡||5.2§|
|Annual precipitation (mm)||323*||716‡||780§|
|N-deposition 2000 (kg ha−1 year−1)||3–5†||10–15†||34¶|
In each region we collected leaf litter from several Sphagnum-dominated peatlands, comprising a broadly similar representation of minerotrophic fens and bog margins as well as strictly ombrotrophic bogs. In north Sweden we collected leaf material at the Stordalen mire and several other peatlands near Abisko (68°21′ N, 18°49′ E): this litter was incubated in Abisko. In south Sweden we collected leaves on Store Mosse, Dala Mosse, Kopparåsmyren, Åkhultmyren and Björnekulla in the province of Småland (57°07′ N, 14°30′ E) and incubated it, for practical reasons, in Lund (55°42′ N, 13°12′ E). In the Netherlands and Belgium we included material from Het Guisveld, De Amstelveense Poel and De Westbroekse Zodden near Amsterdam (52°21′ N, 4°55′ E), from De Grote Peel in the south of the Netherlands and from Fagne Wallone (50°50′ N, 6°08′ E) in Belgium. All this warm-temperate material was incubated in Amsterdam. In each region, we selected litter of five species for each growth form from species typical of, and generally abundant in, its Sphagnum-dominated peatlands. The range of species and families included was as broad as possible, each species was included in only one climatic region, and one species from a particular vascular plant genus in the same region, unless there was no alternative. The monocots Iris pseudacorus and Narthecium ossifragum were classified as forbs (i.e. broad-leaf herbs) based on their broad (non-graminoid) leaves. All species are listed in Appendix S1 in Supplementary Material.
Recently senesced leaves were collected in September 2000 around Abisko, in October 2000 in Småland and in November 2000 in the Netherlands and Belgium, following the criteria described by Cornelissen (1996). Laminae and stalks of compound leaves were both included provided they fell off as a unit (only laminae were therefore used for Filipendula ulmaria and Sorbus aucuparia). Sphagnum sods were cut in the field and stored at 4 °C until further use, when the upper, green and living part and the lowest part, which was dark brown and showed clear signs of disintegration, were removed. The middle part of the shoots, which was usually light brown, and had a fresh and coherent structure, was used as litter, after it had been frozen briefly in liquid nitrogen to ensure its death before incubation. Possible freezing artefacts were tested by incubating frozen litter of one species of each vascular growth form and unfrozen Sphagnum fuscum litter.
Where sufficient material was available, samples of 1.0 ± 0.05 g air-dried litter were weighed to the nearest mg and sealed into tubular bags of 0.9-mm mesh polyester net, otherwise 0.5 ± 0.03 g was used. Bags for very small leaves, needles and fine graminoid leaves were made from 0.3-mm mesh. Leaf material from 13 species was incubated as both 1.0 g and 0.5 g samples and leaf material of 12 species (both selections representing all growth forms and regions) was incubated in both 0.9-mm and 0.3-mm mesh bags, to test for procedural artefacts. In order to standardize the degree of contact among litter pieces, the size of the bags varied between approximately 4 × 4 cm and 5 × 10 cm, depending on the volume of leaf material. Leaves of graminoids were cut into 7 cm long pieces and other large and fragile leaves were cut into halves before weighing. For most species and control treatments we prepared three litterbags per harvest (six in total). For Betula pubescens, Carex rostrata and Empetrum nigrum, however, there were six litterbags per harvest, because these species were used simultaneously in a more detailed study of fewer species, the results of which we will present elsewhere. For each species, a small litter subsample was weighed and oven-dried (70 °C, 48 hours) to determine the relationship between air-dry mass and oven-dry mass, and thus to calculate the initial oven-dry mass in each litterbag. These subsamples were subsequently used for analysis of the initial chemistry.
Our experiment was designed to compare patterns in decomposition for different growth forms in relation to regional variation in climate and other environmental conditions. In order to minimize the effects of microclimatic and microtopographic variability in natural peatlands, we incubated litter under more controlled, semi-natural conditions, using outdoor, Sphagnum-based decomposition beds (Cornelissen 1996; Quested et al. 2003), with a flat microtopography, comparable moisture conditions and limited fluctuations in water-table depth. Each bed consisted of plastic trays (30 × 40 cm × 15 cm height), containing live, 11-cm thick Sphagnum sods. The sods in Abisko consisted primarily of Sphagnum fuscum, those in Lund were cut in Småland and were composed of S. magellanicum and those in Amsterdam contained primarily S. palustre and S. recurvum. Large holes in the side of the trays, 1.5 cm below the Sphagnum surface, allowed for the drainage of excess rainwater, but the bottom was watertight. During the summers, we regularly replenished the soil water by adding de-mineralized water below the Sphagnum surface, in order to prevent strong drought effects. The experimental design consisted of 12 (Lund, Amsterdam) or 18 (Abisko) trays, which were randomly assigned to three blocks. Within each block half of the trays were assigned to the first harvest and the remaining ones to the second harvest (four or six trays per block). The trays in each block were placed adjacent to each other in experimental gardens in Abisko, Lund and Amsterdam in autumn 2000 and dug 7 cm into the ground to smooth temperature fluctuations.
In early 2001 (9 February in Lund, 11 February in Abisko and 21 February in Amsterdam), one litterbag of each species (two for B. pubescens, C. rostrata and E. nigrum) was placed on top of the Sphagnum layer for each block of trays and each harvest date. The litterbags were re-moistened with de-mineralized water, except in Abisko, where a 15-cm thick snow layer had to be removed and watering might have caused frost damage. The litterbags were then covered by a 3-cm thick layer of mixed leaf-mould from locally abundant peatland species (Rubus chamaemorus, B. pubescens and Salix spp. in Abisko, Narthecium ossifragum and B. pubescens in Lund, Phragmites australis, Rubus fruticosus and Schoenoplectus lacustris in Amsterdam) in order to prevent differential dehydration of litterbags. The litter layer was secured by a double layer of 3-cm mesh soft nylon net, and the removed snow in Abisko was replaced.
We harvested the litterbags after 8 months (11 and 20 October and 12 November 2001 in Abisko, Lund and Amsterdam, respectively), and after 20 months (8 and 17 October and 7 November 2002), when the summer season was over in all regions. We removed adhering litter from the outside of the harvested litterbags and removed roots from the litter inside. Upon opening the Sphagnum litterbags that were harvested after 20 months, it became clear that not only had roots grown into the bags, but also that partly decomposed fragments of the covering litter layer had entered and been caught in the dense, sponge-like structure of the Sphagnum. Such problems in using litterbags to determine Sphagnum decomposition over periods longer than 1 year are well recognized (Clymo 1965; Johnson & Damman 1993) and all Sphagnum litterbags from the second harvest were discarded. We collected the remaining litter from all other litterbags and determined its oven-dry mass (70 °C, 48 hours).
Because of the large number of species included, only the most frequently used, ‘bulk’ chemical variables were measured, but these have been related to decomposability in many studies (Coulson & Butterfield 1978; Swift et al. 1979; Melillo et al. 1982; Aerts & Chapin 2000; Pérez-Harguindeguy et al. 2000). Initial total nitrogen and carbon concentrations of the litters were determined by dry combustion on a Perkin Elmer 2400 CHNS analyser, and total phosphorus concentration by colourimetry using the ammonium molybdate method (Murphy & Riley 1962), after digestion in 37% HCl : 65% HNO3 (1 : 4, v/v). After extraction in 50% MeOH, total concentration of soluble phenolics was determined by means of the Folin-Ciocalteu method, with tannic acid as a standard (Waterman & Mole 1994). Although this method may extract a variety of phenolics, negative relations with nitrogen mineralization have been found consistently (Palm & Rowland 1997) and this measure may therefore give a crude but meaningful indication of the effects of bulk soluble phenolics on decomposition for a range of species. It is, moreover, robust against chemical changes during preservation (Waterman & Mole 1994). Initial lignin concentration was determined as described in Poorter & Villar (1997). In brief, after ground, oven-dry plant material has undergone several (polar, non-polar and acid) extraction steps, the mass of the residue, corrected for ash content, and its C and N concentrations are used to calculate the lignin concentration based on the difference in carbon content between cellulose and lignin, after correction for remaining proteins. Although the so-called ‘lignin’ fraction determined in this and other frequently applied methods (acid-insoluble carbon, Klason lignin) may contain other recalcitrant C-fractions besides true lignin, it has often been successfully used as a litter quality index (Hobbie 1996; Preston et al. 1997, and references therein), and this fraction is therefore referred to hereafter as lignin. The initial chemistry values of all litters are given in Appendix S1.
The percentage litter mass loss (relative to initial dry mass) in all litterbags for each species per harvest was averaged before analysis (see Appendix S1). We first tested whether the differences in mesh size, initial litter mass per bag and freezing with liquid nitrogen affected mass loss at either harvest. No correction of the mass loss data was needed for starting mass (1.0 vs. 0.5 g; paired samples t-test (d.f. = 12) t = 1.41, P = 0.19 (8 months); t = 0.18, P = 0.86 (20 months)) or for freezing (t = 1.70 (d.f. = 3), P = 0.19 (8 months)). The mass loss in the fine-mesh litterbags was, however, lower than in the coarse litterbags, but only after 8 months (t = 2.39 (d.f. = 11), P = 0.04; t = 1.43 (d.f. = 11), P = 0.18 after 20 months). We therefore corrected mass loss after 8 months by adding 5.5% for all species that had been incubated in fine-mesh litterbags.
Differences in initial litter chemistry and mass loss after 8 months among the five growth forms and among the regions, and any interactions, were analysed in separate two-way anovas. Differences in decomposability after 8 and 20 months (vascular growth forms only) were analysed with a repeated-measures anova (Pillai's trace), with ‘incubation time’ as within-subject factor and ‘growth form’ and ‘region’ as between-subject factors. The repeated-measures anova was followed by separate two-way anovas for each harvest because of a significant interaction of region with incubation time. To ascertain which factors can be used to predict decomposability, we used multiple regression analyses to compare the individual and combined abilities of growth form and litter chemistry to explain the variation in litter decomposition among species. In order to be able to use categorical variables in the regression analyses, we constructed four binary ‘dummy variables’ (Zar 1999) to represent growth form and two to represent region. Litter mass loss of all species was then regressed on region alone, or with either or both growth form and chemistry. Comparison of the predictive power of different combinations of the measured variables or ratios that were not strongly correlated with each other (r < 0.50), showed that total N, P, C and phenol concentrations were the best chemical variables for regression analysis of mass loss after 8 months, while C was replaced by lignin for mass loss after 20 months.
Data were tested for homogeneity of variances with Levene's test. Ln-transformation considerably improved the normality and homogeneity of variances of most of the chemical variables, but not of the C and lignin data or the mass loss data. Mass loss data (fractions) were arcsine-square-root transformed, except when Sphagnum mosses were included in the analyses, because of slightly negative mass loss values for three of the Sphagnum species in Amsterdam after 8 months. Because analyses of variance and regression analyses are robust to considerable heterogeneity of variances as long as the sample sizes are nearly equal (Zar 1999), we proceeded with the analyses even when homoscedasticity assumptions were not fully met. Post-hoc multiple comparisons were included if growth form or region was a significant factor (Tukey HSD or Games-Howell), and these tests were performed for separate regions if the interaction between the main factors was significant. All analyses were performed with SPSS for Windows 10.1.
- Top of page
- Supplementary material
Leaf litters of the five peatland growth forms differed significantly in initial concentrations of N, P, soluble phenolics, C and lignin, as well as in their most commonly used ratios, phenol : N, phenol : P, C : N, C : P, lignin : N and lignin : P (Table 2, Fig. 3). Total N concentrations were lowest in Sphagnum and graminoid litters and highest in forb litters. Soluble phenolic concentrations were highest in evergreen and deciduous shrubs and lowest in Sphagnum litters. Total C and lignin concentrations were highest in evergreen shrubs, followed by deciduous shrubs, and lignin, in particular, was much lower in forbs, graminoids and Sphagnum litters. The patterns among growth forms of the C : N and C : P ratios were the opposite of those for total N and P, while the patterns of the phenol : N, phenol : P, lignin : N and lignin : P ratios resembled those of the phenolic and lignin concentrations, respectively (results not shown).
|GF × region||0.56||0.81|
|GF × region||0.84||0.58|
|GF × region||1.21||0.31|
|GF × region||1.34||0.24|
|GF × region||1.46||0.19|
|Phenol : N ratio||GF||12.37||< 0.001|
|GF × region||1.14||0.35|
|Phenol : P ratio||GF||9.69||< 0.001|
|GF × region||0.97||0.47|
|C : N ratio||GF||9.74||< 0.001|
|GF × region||0.66||0.73|
|C : P ratio||GF||3.63||0.010|
|GF × region||0.86||0.55|
|Lignin : N ratio||GF||22.97||< 0.001|
|GF × region||1.68||0.12|
|Lignin : P ratio||GF˙||4.04||0.006|
|GF × region||0.66||0.73|
Total N concentrations were significantly lower in leaf litters from the subarctic than from either temperate region (Tukey HSD P < 0.05), while total P concentrations were lower in the cool-temperate region than in the other regions (Games-Howell P < 0.05). Leaf litters in the cool-temperate region also contained significantly less soluble phenolics than subarctic litters (Games-Howell P < 0.05), while total C and lignin concentrations did not differ among the regions (Table 2). None of the chemical variables showed significant interactions between growth form and region, indicating that the differences in litter chemistry among growth forms are robust against changes in climate or nutrient availability, and in species composition within growth form (Table 2).
Litter mass loss after 8 months showed clear differences among growth forms (Fig. 4), but the pattern was not consistent across regions (significant GF × region interaction, Table 3). Sphagnum litter decomposed slowest, while forbs generally decomposed fastest. The differences in average mass loss among the four vascular growth forms were relatively small, except in the warm-temperate region. The repeated measures analysis of the 8- and 20-month harvests showed that patterns in mass loss in different vascular growth forms also varied across regions (Table 3, growth form × region P = 0.008), although the interaction was only marginally significant for the separate analysis of the 20-month data (Table 3, growth form × region P = 0.08). The interaction is probably related to the relatively low mass loss of the forbs and deciduous shrubs in the subarctic and cool-temperate regions, although these groups decomposed fastest in the warm-temperate region. Moreover, evergreen shrubs decomposed slowly compared with the other vascular groups, particularly graminoids, in the subarctic region, while there were no significant differences in the cool-temperate and warm-temperate regions (Fig. 4).
|Variable||Source||All growth forms||Vascular growth forms|
|Mass loss 8 and 20 months||Incubation time||48.80||< 0.001|
|Incubation time × GF||0.57||0.64|
|Incubation time × region||24.43||< 0.001|
|Incubation time × GF × region||1.63||0.16|
|GF × region||3.34||0.008|
|Mass loss 8 months||GF||34.42||< 0.001||9.86||< 0.001|
|Region||35.42||< 0.001||45.42||< 0.001|
|GF × region||7.17||< 0.001||4.54||0.001|
|Mass loss 20 months||GF||5.07||0.004|
|GF × region||2.00||0.08|
growth forms vs. litter chemistry
To test whether growth forms were as good at predicting short-term leaf litter decomposability as the most commonly used litter chemistry variables, we added the two factors, separately and in combination, to multiple regression models including data from all regions, or from separate regions. In the model including the 8-month harvest data from all regions, the addition of growth form increased the explained variance by 44% from that with region alone, resulting in a fairly good prediction of leaf litter decomposability (total R2 = 0.65, Table 4). In contrast, the set of initial litter chemistry variables (total N, P, soluble phenolics and C concentrations) added only 26% explained variance to the model (total R2 = 0.47), compared with a model using region only. The combination of both growth form and litter chemistry in the model increased the total R2 to 0.74. This indicates that only 9% of the variance was explained by chemical aspects that were unrelated to growth form and region, whereas information uniquely related to growth form accounted for 27% of the total variance in litter decomposition.
|Variables included||All regions||Subarctic||Cool-temperate||Warm-temperate|
|8 months (all growth forms)|
|Region + chemistry||0.47***b||0.43*a||0.22a||0.48**a|
|Region + GF||0.65***b||0.49*a||0.59**b||0.84***b|
|Region + chemistry + GF||0.74***c||0.62*a||0.75**b||0.88***b|
|8 months (vascular growth forms)|
|Region + chemistry||0.69***c||0.32a||0.39a||0.58**a|
|Region + GF||0.62***b||0.16a||0.22a||0.65**a|
|Region + chemistry + GF||0.73***c||0.46a||0.53a||0.72*a|
|20 months (vascular growth forms)|
|Region + chemistry||0.38***b||0.54*b||0.50*a||0.47*a|
|Region + GF||0.31**b||0.23a||0.07a||0.48*b|
|Region + chemistry + GF||0.62***c||0.65*b||0.79**b||0.74**b|
When Sphagnum mosses were excluded, the chemical variables increased the R2 of a model including the mass loss data after 8 months and region as sole variable by 0.23 (total R2 = 0.69), but growth form alone was a weaker predictor, only adding 16% explained variance (total R2 = 0.62). The combination of growth form and litter chemistry increased the total R2 to 0.73, indicating that vascular growth forms added unique information to the model explaining only 4% of the variance, compared with 11% for information uniquely related to litter chemistry (Table 4). After 20 months, differences in mass loss among the vascular leaf litters were poorly explained by region, but again the set of chemical variables (including lignin instead of C) improved the total explained variance of the model more than growth forms, and also contributed more unique information (Table 4).
A comparison of the regression models among the three regions showed that, when all growth forms were included, litter chemistry variables generally explained less variation in litter decomposition than growth forms (Table 4). Furthermore, the percentage of variance explained by growth forms alone increased with decreasing latitude (R2 = 0.49 in the subarctic vs. R2 = 0.84 in the warm-temperate region). However, exclusion of Sphagnum decreased the percentage of variance explained by growth forms in the subarctic and cool-temperate regions, both after 8 and 20 months (Table 4).
- Top of page
- Supplementary material
This study has provided the first large-scale test of (i) whether growth forms differ in leaf litter chemistry and potential decomposition in a single ecosystem type, viz. peatlands, (ii) whether the patterns among growth forms are independent of environmental conditions such as climate or nutrient availability, and (iii) whether classification of species into growth forms allows as accurate predictions of litter decomposability as a set of the most widely used aspects of litter chemistry. Although our screening of a broad range of species along a latitudinal gradient shows several consistent patterns, interactions occur with regional climate or nutrient availability, reducing the usefulness of the growth form concept.
environmental and growth-form controls over litter quality
The five growth forms most common in peatlands showed distinct differences in leaf litter chemistry, and the patterns were consistent across a range of environmental conditions for all measured variables. As in other ecosystems (Pérez-Harguindeguy et al. 2000; Shaw & Harte 2001), the total N concentrations of forb litters were higher and C : N ratios and, to some extent, C : P ratios much lower than those of any of the other growth forms and deciduous shrubs had relatively high N concentrations and low C : N ratios. Although forbs tend to dominate vegetation on nutrient-rich sites (Chapin et al. 1996) they tend to be relatively unimportant, in terms of cover and biomass production, in nutrient-poor peatlands, especially bogs (Press et al. 1998; Weltzin et al. 2000; Berendse et al. 2001). Although we tried to get equal representation of material from minerotrophic fens and ombrotrophic bogs for all growth forms, most forb species had to be collected on fens or relatively minerotrophic margins of large bogs. However, climate warming and subsequent increases in nitrogen availability may increase the biomass of forbs and deciduous shrubs (Arft et al. 1999; Jonasson et al. 1999), which might result in higher ecosystem litter N concentrations, lower C : N ratios and, probably, lower C : P ratios. Total concentrations of soluble phenolics were 13 times higher in leaf litters of deciduous and evergreen shrubs than in Sphagnum litters and four times higher than in graminoid litters. Lignin and C concentrations were also highest in the evergreen and deciduous shrubs, indicating that the woody nature of these growth forms is not restricted to their stems. Even in a strongly nutrient-limited ecosystem type such as peatlands, growth forms thus differ substantially in bulk litter chemistry (Fig. 1, relationship A), and climate or nutrient availability does not interfere with these patterns (Fig. 1, interaction C).
Almost all litter chemistry variables and ratios varied across regions and, as consistent patterns were observed among the growth forms, are more likely to be due to regional differences in environmental conditions than in species composition. N concentrations were lower and C : N ratios higher in the subarctic than in both temperate regions, while P concentrations were lower and C : P ratios higher in the cool-temperate region. This is probably the result of a shift at lower latitudes from nitrogen- to phosphorus-limited growth, due, at least in part, to anthropogenic nitrogen deposition (Aerts et al. 1992). Future climate warming might lead to greater nitrogen mineralization and thus induce such a shift at higher latitudes as well (Rustad et al. 2001). The twofold greater phenolic concentration of leaf litters in the subarctic compared with either temperate regions may reflect a stronger nutrient limitation on growth in the subarctic area. Phenols have been shown to serve as an alternative carbon sink when growth is limited by factors other than C-assimilation (Jones & Hartley 1999; Graglia et al. 2001). Although our results suggest that litter chemistry shows consistent responses to combined changes in temperature, precipitation and nutrient availability (Fig. 1, relationship 2), its response to experimental fertilization and, particularly, to warming varies among species, growth forms and studies (Robinson et al. 1995; Arft et al. 1999; Shaw & Harte 2001; Dormann & Woodin 2002; Van Heerwaarden et al. 2003). Moreover, the variation in litter chemistry among the three regions in our study was usually smaller than the variation among growth forms (Table 2). Changes in the relative abundance of the growth forms may therefore be more important in determining leaf litter chemistry at an ecosystem level than large, long-term changes in temperature, precipitation and nitrogen availability, or shifts in species composition within growth forms.
environmental, chemical and growth-form controls over decomposition
Although differences between growth forms in nutrient- and carbon-use strategies were reflected in litter decomposability, even in nutrient-limited peatlands, the significant interaction of growth form and region was counter to our hypothesis that environmental conditions along a large latitudinal gradient would not interfere with decomposition patterns among growth forms. Forb litters usually decomposed fastest and Sphagnum litters slowest, but differences among the other vascular growth forms were generally small, especially at higher latitudes, and patterns were not consistent across regions. These findings correspond to previous screenings of decomposability in peatlands and subarctic ecosystems (Verhoeven & Toth 1995; Aerts et al. 1999; Quested et al. 2003), although larger differences among vascular growth forms have been reported for a range of other ecosystems (Cornelissen 1996; Pérez-Harguindeguy et al. 2000). The interaction between growth form and environmental region was also apparent when we compared litter chemistry and growth form as predictive variables for decomposition. Growth form appeared to be the better predictor when all growth forms were included in the analysis, but the strength of the relationship between decomposition and growth form decreased considerably with increasing latitude. Differences in decomposability among growth forms have been reported to be larger under favourable conditions (Cornelissen et al. 1999), and our results may indicate that climatic and nutritional constraints for plant growth in peatlands at higher latitudes result in an overall convergence towards stress-tolerant, nutrient-efficient plant traits for all growth forms. Although large differences in specific litter quality variables are found, the combined effects of several chemical variables probably constrain leaf litter decomposability in all growth forms. For instance, decomposition of evergreen and deciduous shrub litters in the subarctic may be primarily hampered by its high phenolic, C and lignin concentrations, while graminoid and Sphagnum litter decomposition may be restricted by low N concentrations (Fig. 3). Differences in climatic conditions or nutrient availability thus seem to have different effects on the decomposition of different growth forms, which interferes with the patterns among the groups (Fig. 1, relationships A and B, interaction D).
Both chemistry and growth form explained significant proportions of the total variation in decomposition across regions. Growth form, however, contained a relatively large portion of unique information that was not covered by the selected litter chemistry variables, especially when all growth forms were included (27% for all regions, up to 40–53% in the temperate regions). We may thus have omitted to measure some chemical or physical characteristics of the leaf litters that are important determinants of decomposition. Partly because of practical constraints, we selected the best combination from those chemical variables and ratios that are usually well-correlated with decomposition (Coulson & Butterfield 1978; Swift et al. 1979; Melillo et al. 1982; Hobbie 1996; Aerts & Chapin 2000; Pérez-Harguindeguy et al. 2000; Aerts et al. 2003), and obtained comparable R2 values, particularly in the warm-temperate region or for vascular growth forms only. Lignin : N ratio was not among the selected variables (initial N, P, phenolics and C (8-month harvest) or lignin (20-month harvest) concentrations), although it is often considered to be an important indicator of the decomposability of deciduous and evergreen forest leaf litters (Melillo et al. 1982; Aerts 1997). For wider ranges of growth forms or habitats, however, initial concentrations of lignin, C, N or phenolics, or the C : N or C : P ratios alone may explain as much variation as lignin : N (Coulson & Butterfield 1978; Aerts & Chapin 2000; Pérez-Harguindeguy et al. 2000; Aerts et al. 2003), which may indicate that the factors controlling decomposition differ between woody and non-woody growth forms. A more likely explanation is related to chemical compounds in Sphagnum, with very specific decay-inhibiting effects, such as sphagnan and sphagnum acid (Johnson & Damman 1993; Verhoeven & Toth 1995) or specific phenolic compounds (Swift et al. 1979). Another alternative is leaf tensile strength, which has also been shown to explain variation in leaf litter decomposability, and is usually high in graminoids and low in forbs (Cornelissen et al. 1999; Pérez-Harguindeguy et al. 2000).
growth forms and environmental change in northern peatlands
Our latitudinal gradient can be regarded as a spatial analogue for future changes in climate and nutrient availability at higher latitudes. Mean annual temperatures increased by 10.3 °C, while the annual precipitation more than doubled and nitrogen deposition showed a tenfold increase from north to south. Over the coming century, arctic and subarctic mean summer and winter temperatures are expected to increase by 1.0–8.2 °C and 2.0–14.0 °C, respectively, while precipitation is predicted to increase by 5–80% (Maxwell 1992; Houghton et al. 2001; McCarthy et al. 2001). Experimental increases in temperature have been shown to stimulate nitrogen mineralization (Chapin et al. 1995; Hobbie 1996; Robinson 2002), and increased nitrogen availability may therefore be a realistic indirect consequence of climate warming. However, although tenfold increases in N-mineralization have been observed upon experimental warming (Nadelhoffer et al. 1991), the average increase in a meta-analysis of warming experiments was only 46% (Rustad et al. 2001). Moreover, the higher temperatures in the more southern regions may also result in higher mineralization rates, thus increasing the range in N availability along our gradient even further. The changes in temperature, precipitation and N availability along the latitudinal gradient were therefore in the same direction, but probably larger than can be expected at high latitudes according to realistic global change scenarios.
The regions also differed in the plant species collected, and possibly in the composition and activity of the decomposer community in our decomposition beds, partly because different Sphagnum species constituted the incubation medium. The composition of the decomposer community may vary with both resource quality and climate (Swift et al. 1979), but along climatic gradients this does not seem to affect the ranking of plant species according to their litter decomposability (Cornelissen et al. 1999; González & Seastedt 2001). Changes in climate and nutrient availability have been shown to induce shifts in plant community composition (Chapin et al. 1995; Press et al. 1998; Weltzin et al. 2003). Indeed, it was hard to find five relatively abundant forbs on the peatlands in the subarctic, or five woody evergreen species in the warm-temperate region (E. Dorrepaal, personal observation). Although the actual shifts in the vegetation due to changes in climate or nutrient availability may not be as broad as the range of species included in this study, we deliberately chose a wide range of species, and equal numbers of species within each growth form. We were therefore able to test the true effect of growth form, independent of the identity of the species concerned, as a predictor of litter properties. The highly consistent patterns in leaf litter chemistry indicate that the main differences in litter quality, decomposability and incubation conditions among the three regions were most likely related to direct and indirect effects of differences in temperature, precipitation and nitrogen availability, i.e. the key factors in global change scenarios.
Experimental warming of arctic and peatland ecosystems tends to enhance shrub abundance, but responses of forbs, graminoids, Sphagnum and other mosses vary considerably (Arft et al. 1999; Weltzin et al. 2001; Dormann & Woodin 2002; Dorrepaal et al. 2003; Van Wijk et al. 2003; Weltzin et al. 2003). Responses to increased nutrient availability seem to be more straightforward, with usually positive effects on graminoids, positive or no effects on forbs and deciduous and evergreen shrubs, and negative effects on Sphagnum mosses (Press et al. 1998; Jonasson et al. 1999; Dormann & Woodin 2002; Van Wijk et al. 2003). Our data suggest that warming and increased nutrient availability may thus increase the ecosystem-level concentrations of phenolic compounds, C and lignin because of shifts towards shrubs, but may also increase litter N concentrations and reduce the C : N ratios because of reduced production of Sphagnum, possibly increased production of forb litter, and direct effects on all growth forms. Reduced Sphagnum production would strongly increase the overall peatland decomposition rate. However, because of ecological convergence and inconsistent patterns, the use of changes in vascular growth form abundance and litter chemistry to predict leaf litter decomposability in peatlands appears to be limited, particularly at higher latitudes.
- Top of page
- Supplementary material
We thank Miranda de Beus, Dorien van Biezen, Ivanka Bijlemeer and Jeroen Cornelissen for invaluable field and laboratory assistance, Hans Kruijer and Cris Hesse for their help in identifying the Sphagnum species and Nico Schaefers for preparing Fig. 2. We also thank the Abisko Scientific Research Station and its staff for assistance and hospitality. This study was financially supported by USF grant 98/24 to R.A. The fieldwork at the Abisko Scientific Research Station was financially supported by grants of the Royal Swedish Academy of Sciences to E.D.
- Top of page
- Supplementary material
The following supplementary material is available for this article from http://www.Blackwell-Synergy.com.
Appendix S1 Scientific species names, initial chemistry and mass loss after 8 and 20 months incubation in Sphagnum-based decomposition beds for all included leaf litters.
- Top of page
- Supplementary material
- 1997) Climate, leaf litter chemistry and leaf litter decomposition in terrestrial ecosystems: a triangular relationship. Oikos, 79, 439–449. (
- 2000) The mineral nutrition of wild plants revisited: a re-evaluation of processes and patterns. Advances in Ecological Research, 30, 1–67. & (
- 1997) Nutritional and plant-mediated controls on leaf litter decomposition of Carex species. Ecology, 78, 244–260. & (
- 2003) Plant community mediated vs. nutritional controls on litter decomposition rates in grasslands. Ecology, 84, 3198–3208. , & (
- 1999) Plant-mediated controls on nutrient cycling in temperate fens and bogs. Ecology, 80, 2170–2181. , & (
- 1992) Growth-limiting nutrients in Sphagnum-dominated bogs subject to low and high atmospheric nitrogen supply. Journal of Ecology, 80, 131–140. , & (
- 2001) Combining theory and experiment to understand effects of inorganic nitrogen on litter decomposition. Oecologia, 128, 94–98. , & (
- 1999) Responses of tundra plants to experimental warming: meta-analysis of the international tundra experiment. Ecological Monographs, 69, 491–511. , , , , , et al. (
- 2001) Raised atmospheric CO2 levels and increased N deposition cause shifts in plant species composition and production in Sphagnum bogs. Global Change Biology, 7, 591–598. , , , , , et al. (
- 2002) De Nederlandse veenmossen: flora en verspreidingsatlas van de Nederlandse Sphagnopsida. KNNV Uitgeverij, Utrecht, The Netherlands. & (
- 1996) Plant functional types as predictors of transient responses of arctic vegetation to global change. Journal of Vegetation Science, 7, 347–358. , , & (
- 1995) Responses of arctic tundra to experimental and observed changes in climate. Ecology, 76, 694–711. , , , & (
- 1965) Experiments on breakdown of Sphagnum in two bogs. Journal of Ecology, 53, 747–758. (
- 1998) Carbon accumulation in peatland. Oikos, 81, 368–388. , & (
- 1996) An experimental comparison of leaf decomposition rates in a wide range of temperate plant species and types. Journal of Ecology, 84, 573–582. (
- 1999) Leaf structure and defence control litter decomposition rate across species and life forms in regional floras on two continents. New Phytologist, 143, 191–200. , , , , , et al. (
- 1978) An investigation of the biotic factors determining the rates of plant decomposition on blanket bog. Journal of Ecology, 66, 631–650. & (
- 2002) Climate change in the Arctic: using plant functional types in a meta-analysis of field experiments. Functional Ecology, 16, 4–17. & (
- 2003) Summer warming and increased winter snow cover affect Sphagnum fuscum growth, structure and production in a sub-arctic bog. Global Change Biology, 10, 93–104. , , , & (
- 1997) What are plant functional types and how should we seek them? Plant Functional Types: their Relevance to Ecosystem Properties and Global Change (eds T.M.Smith, H. H.Shugart & F.I.Woodward), pp. 3–19. Cambridge University Press, Cambridge. & (
- 2001) Soil fauna and plant litter decomposition in tropical and subalpine forests. Ecology, 82, 955–964. & (
- 1991) Northern peatlands: role in the carbon cycle and probable responses to climatic warming. Ecological Applications, 1, 182–195. (
- 2001) Environmental control and intersite variations of phenolics in Betula nana in tundra ecosystems. New Phytologist, 151, 227–236. , , , , & (
- 1996) Temperature and plant species control over litter decomposition in Alaskan tundra. Ecological Monographs, 66, 503–522. (
- 2001) Climate Change 2001: the Scientific Basis – Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. , , , , , et al. (
- 1993) Decay and its regulation in Sphagnum peatlands. Advances in Bryology, 5, 249–296. & (
- 1999) Responses in microbes and plants to changed temperature, nutrient, and light regimes in the arctic. Ecology, 80, 1828–1843. , , & (
- 1999) A protein competition model of phenolic allocation. Oikos, 86, 27–44. & (
- KNMI (2002) Klimaatatlas van Nederland: De normaalperiode 1971–2000. Uitgeverij Elmar, Rijswijk, The Netherlands.
- 2003) Climatic effects on litter decomposition from arctic tundra to tropical rainforest. Global Change Biology, 9, 575–584. , , & (
- 1992) Arctic climate: potential for change under global warming. Arctic Ecosystems in a Changing Climate: an Ecophysiological Perspective (eds F.S.Chapin, R.L.Jefferies, J.F.Reynolds, G.R.Shaver & J.Svoboda), pp. 11–34. Academic Press, San Diego, California. (
- 2001) Climate Change 2001: Impacts, Adaptation, and Vulnerability – Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge. , , , & (
- 1982) Nitrogen and lignin control of hardwood leaf litter decomposition dynamics. Ecology, 63, 621–626. , & (
- 1992) Den Nordiska Floran. Wahlström & Widstrand, Stockholm, Sweden. , & (
- 1962) A modified single solution method for the determination of phosphate in natural waters. Analytical Chimica Acta, 27, 31–36. & (
- 1991) Effects of temperature and substrate quality on element mineralization in six arctic soils. Ecology, 72, 242–253. , , & (
- 1997) A minimum dataset for characterization of plant quality for decomposition. Driven by Nature: Plant Litter Quality and Decomposition (eds G.Cadisch & K.E.Giller), pp. 379–392. CAB International, Wallingford. & (
- 2000) Chemistry and toughness predict leaf litter decomposition rates over a wide spectrum of functional types and taxa in central Argentina. Plant and Soil, 218, 21–30. , , , , & (
- 1997) The fate of acquired carbon in plants: chemical composition and construction costs. Plant Resource Allocation (eds F.A.Bazzaz & J.Grace), pp. 39–72. Academic Press, San Diego, California. & (
- 1998) Responses of a subarctic dwarf shrub heath community to simulated environmental change. Journal of Ecology, 86, 315–327. , , , & (
- 1997) 13C nuclear magnetic resonance spectroscopy with cross-polarization and magic-angle spinning investigation of the proximate-analysis fractions used to assess litter quality in decomposition studies. Canadian Journal of Botany, 75, 1601–1613. , , & (
- 2003) Decomposition of sub-arctic plants with differing nitrogen economies: a functional role for hemiparasites. Ecology, 84, 3209–3221. , , , , , et al. (
- RIVM (2002) Luchtkwaliteit, Jaaroverzicht 2001. RIVM, Bilthoven, The Netherlands.
- 2002) Controls on decomposition and soil nitrogen availability at high latitudes. Plant and Soil, 242, 65–81. (
- 1995) Responses of plant litter decomposition and nitrogen mineralisation to simulated environmental change in a high arctic polar semi-desert and a subarctic dwarf shrub heath. Oikos, 74, 503–512. , , , , , et al. (
- 2001) A meta-analysis of the response of soil respiration, net nitrogen mineralization, and aboveground plant growth to experimental ecosystem warming. Oecologia, 126, 543–562. , , , , , et al. (
- 2001) Control of litter decomposition in a subalpine meadow-sagebrush steppe ecotone under climate change. Ecological Applications, 11, 1206–1223. & (
- SMHI (2002) Väder och Vatten. SMHI, Stockholm, Sweden.
- 1979) Decomposition in Terrestrial Ecosystems. Blackwell Scientific, Oxford. , & (
- 2001) Nordklim Data Set 1.0 – Description and Illustrations. Klima Report no. 08/01. DNMI, Oslo, Norway. , , , , , et al. (
- 1996) Heukels’ Flora van Nederland, 22nd edn. Wolters-Noordhoff, Groningen, The Netherlands. (
- 1983) Heukels/Van der Meijden – Flora van Nederland, 20th edn. Wolters-Noordhoff, Groningen, The Netherlands. , , & (
- 2003) Nitrogen and phosphorus resorption efficiency and proficiency in six sub-arctic bog species after 4 years of nitrogen fertilization. Journal of Ecology, 91, 1060–1070. , & (
- 2003) Long-term ecosystem level experiments at Toolik Lake, Alaska, and at Abisko, Northern Sweden: generalizations and differences in ecosystem and plant type responses to global change. Global Change Biology, 10, 105–123. , , , , , et al. (
- 1995) Decomposition of Carex and Sphagnum litter in fens: effect of litter quality and inhibition by living tissue homogenates. Soil Biology and Biochemistry, 27, 271–275. & (
- 1994) Analysis of Phenolic Plant Metabolites. Blackwell Scientific, London. & (
- 2003) Potential effects of warming and drying on peatland plant community composition. Global Change Biology, 9, 141–151. , , , & (
- 2001) Production and microtopography of bog bryophytes: response to warming and water-table manipulations. Oecologia, 128, 557–565. , , , & (
- 2000) Response of bog and fen plant communities to warming and water-table manipulations. Ecology, 81, 3464–3478. , , , , & (
- 1999) Biostatistical Analysis, 4th edn. Prentice Hall, Englewood Cliffs, New Jersey. (