• global warming;
  • internal nutrient cycling;
  • leaf litter;
  • nutrient-limited growth


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Plant growth at high-latitude sites is usually strongly nutrient-limited. The increased nutrient availability predicted in response to global warming may affect internal plant nutrient cycling, including nutrient resorption from senescing leaves.
  • 2
    The effect of increased N supply (10 g N m−2 year−1) on nitrogen and phosphorus resorption efficiency and proficiency in six sub-arctic bog species, belonging to four different growth-forms, was studied in northern Sweden.
  • 3
    We hypothesized that while increased N supply would not affect N or P resorption efficiency, it would lead to lower N resorption proficiency (higher N concentrations in leaf litter) and higher P resorption proficiency (lower P concentrations in leaf litter). We also investigated whether the basis on which resorption was expressed (leaf mass, leaf area or unit leaf) influenced the patterns observed.
  • 4
    Contrasting with our hypothesis, a general trend of decreased N resorption efficiency occurred in response to increased N supply, but the expected decrease in N resorption proficiency was seen in all species except Betula nana.
  • 5
    P resorption efficiency did not change in four species (B. nana, Empetrum hermaphroditum, Eriophorum vaginatum and Rubus chamaemorus) but it decreased in Andromeda polifolia, and increased in Vaccinium uliginosum. P resorption proficiency showed the expected increase in only two species (B. nana and V. uliginosum).
  • 6
    Apart from P resorption efficiency, the different calculation methods generally produced similar responses of resorption efficiency and proficiency to N supply.
  • 7
    Increased N supply at high-latitude sites clearly leads to more N being returned to the soil through leaf litter production. However, decomposition of such litter will probably become P-limited.
  • 8
    Considerable interspecific differences in nutrient resorption proficiency were found, indicating that long-term changes in vegetation composition need to be considered when evaluating plant-mediated effects on ecosystem nutrient cycling in response to increased nutrient supply.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Resorption of nutrients from senescing leaves is of great adaptive significance, because such nutrients are directly available for further use, making a species less dependent on current nutrient uptake (Aerts & Chapin 2000). It has often been suggested that species from low-nutrient habitats have higher nutrient resorption efficiencies (percentage of a nutrient withdrawn from mature leaves before leaf abscission). However, the evidence available so far does not support this contention: high nutrient resorption efficiency is characteristic of all perennial growth-forms and appears not to be very responsive to changes in nutrient supply (Aerts & Chapin 2000). This implies that although nutrient resorption is an important nutrient conservation mechanism, it does not explain the distribution of growth-forms over habitats differing in soil fertility.

Nutrient resorption can also be quantified by resorption proficiency, a parameter describing the level to which a nutrient is reduced during senescence (Killingbeck 1996), with higher proficiencies corresponding to lower final nutrient concentrations. Resorption proficiency seems to be more responsive than resorption efficiency to nutrient availability. Several studies reported that nitrogen (N) fertilization resulted in higher N concentrations in the litter of most species (Shaver & Mellilo 1984; Kemp et al. 1994; Vitousek 1998), suggesting that N fertilization leads to lower N resorption proficiencies. This relationship between N resorption proficiency and N availability is also found along natural fertility gradients (Pugnaire & Chapin 1993; Eckstein et al. 1999). In contrast to resorption efficiency, there are clear differences in resorption proficiency among growth-forms (Aerts 1996). On average, evergreen species have higher N proficiencies than woody deciduous species, which in turn have higher proficiencies than herbaceous species. This has profound consequences for soil nutrient cycling as the nutrient content of herbaceous and deciduous leaf litter that is returned to the soil is higher than that of evergreens. This is an important aspect of the positive feedback between growth-form dominance and nutrient availability in ecosystems (Chapin 1993; Aerts 1999), which, although plant-driven, may be affected by external nutrient inputs.

In many ecosystems N availability has increased during the past decades because many industrialized countries are subject to enhanced atmospheric N deposition (Bobbink et al. 1998) and also because of climate change. Glasshouse gas emissions are predicted to raise mean global temperatures by 1.0–3.5 °C in the next 50–100 years, with above-average increases at high-latitude and high-altitude sites (Houghton 2001). This temperature increase is expected to have a significant impact not only on the abundance and phenology of organisms (Wookey et al. 1993; Wookey et al. 1994; Chapin et al. 1995; Harte & Shaw 1995), but also on element cycling in those regions where plant growth tends to be severely constrained by nutrient availability (Gorham 1991). A global meta-analysis of 32 warming experiments, including several at high-latitude and high-altitude sites, has shown that, across all sites and years, 2–9 years of experimental warming in the range of 0.3–6.0 °C significantly increased soil net N mineralization rates by 46% (with a 95% confidence interval of 30–64%) (Rustad et al. 2001). There are no clear effects of warming on soil phosphorus (P) availability, because the breakdown of detrital N is a biological process, whereas breakdown of detrital P is a chemical process with a much lower Q10 value (Aerts & Chapin 2000). These observations raise the question of how increased inorganic N availability affects N and P resorption from senescing leaves of dominant species and thereby the amount of N and P returned to the soil through leaf litter production. A secondary effect of increased N availability in N-limited ecosystems is a shift towards P-limitation (Bobbink et al. 1998), which seems to lead to lower P concentrations in litter, due to growth dilution of the P in the leaves (Shaver & Mellilo 1984; Vitousek 1984; Kemp et al. 1994).

We hypothesized that, in N-limited ecosystems, increased inorganic N supply would not affect leaf resorption efficiency for either N or P, but would lead to lower leaf resorption proficiency for N and higher values for P. Leaf litter concentrations are thus expected to increase for N, and to decrease for P. We investigated these hypotheses by studying the effects of 4 years of N fertilization on leaf nutrient resorption of six dominant plant species in a subartic bog in northern Sweden. We also investigated, as a secondary aim, whether the observed patterns in resorption efficiency and proficiency were dependent on the basis on which resorption was expressed (per unit of leaf mass, leaf area or per unit leaf).

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

field site and experimental set-up

Leaves were collected from a long-term N fertilization experiment that started in 1998 in a peat bog at Stordalen, near the Abisko Scientific Research Station, northern Sweden (68°21′ N, 18°49′ E). This bog, which is a former tundra site of the International Biosphere Programme (IBP), has been described in detail by Sonesson (1980). Annual precipitation amounts to 320 mm per year, and mean summer and winter temperatures are 7 °C and −6 °C, respectively. The length of the growing season is 130 days. The vegetation of the bog is characterized by great heterogeneity both at small and large scales, reflecting the topographical heterogeneity of the site. We conducted our study in the elevated ombrotrophic part of the bog in the Empetrum hermaphroditum–Vaccinium microcarpum association.

In May 1998, 16 plots measuring 1.5 by 1.5 m were selected on drier hummocks. The plots were separated by buffer zones of at least 0.5 m. Since 1998, eight replicate plots have been N fertilized each year (10 g N m−2 year−1 as NH4Cl). Nitrogen was added in solution in three equal portions in mid-June, mid-July and mid-August. The other eight replicate plots served as controls (C) and received only an equal amount of water to that used to dissolve the NH4Cl.

Six dominant perennial plant species that occurred within each plot were selected; namely: Empetrum hermaphroditum Hagerup and Andromeda polifolia L., both woody evergreen; Vaccinium uliginosum L. and Betula nana L., both woody deciduous; Eriophorum vaginatum L., graminoid; and Rubus chamaemorus L., herb. From their mycorrhizal infection and low leaf ∂15N values (cf. the data of Michelsen et al. 1996, 1998 for this area), the four woody species are most probably capable of taking up organic N sources. No such data are available for E. vaginatum and R. chamaemorus. However, given the widespread capability of northern non-mycorrhizal species to take up organic N (Näsholm et al. 1998), these species are probably also capable of such assimilation. It must be emphasized, however, that the present study focuses on the effects of increased inorganic N supply on nutrient resorption from senescing leaves.

leaf collection

In July 2001, preceding the N fertilization that month, mature green leaves of the selected species were collected at random from each plot. In autumn (late September to early October) recently senesced, brown, but still attached leaves were collected in the same plots. For each plot and each species, sufficient leaves were collected to enable chemical analyses. For E. hermaphroditum this meant randomly sampling 200 leaves from whole branches, as different cohorts were not clearly distinguishable. Approximately 50, 60 and 35 leaves were collected from B. nana, V. uliginosum and A. polifolia, respectively, and in autumn, recently senesced leaves were collected from the plants as well as from the ground. For R. chamaemorus about 12 leaves were collected in summer, but 25 were collected in autumn. About 30 leaves were collected from E. vaginatum, but as they had not yet totally senesced down to the base in the control plots only the senesced parts (> 75%) were collected in autumn. The leaves were transported to the laboratory, oven dried at 30 °C to a constant weight and stored until further analysis.

laboratory methods

For both mature and senesced leaves, the number of collected leaves of each species in a plot was counted, and their dry mass measured. Total projected leaf area per sample was determined for each species using a scanner and Photoshop 5.0. However, this was not possible for the evergreen species A. polifolia, because the tough and bent leaves of this species did not allow accurate measurement of leaf area. For E. vaginatum, which has long, narrow and partly folded leaves, a similar problem occurred, and for this species the total leaf length per plot was determined as an alternative. Therefore, only four species provided data on all three different measurement bases (mass, area, unit leaf) for comparison. Leaf material of each species was bulked per plot and ground. Leaf P concentration was determined after digesting ground leaf material in 37% HCl : 65% HNO3 (1 : 4, v/v) (cf. Sneller et al. 1999). Phosphorus was measured colorimetrically at 880 nm after reaction with molybdenum blue (Murphy & Riley 1962). Total N concentration was measured using a Perkin Elmer 2400 series II CHNOS/O analyser.

nitrogen and phosphorus resorption

Resorption of N and P was calculated for each species and plot individually. Resorption efficiency (RE) was calculated as:

  • RE = 100% × (1 − [nutrient]dead / [nutrient]green

where [nutrient] is the N or P concentration of the dead or the green leaves expressed as the nutrient mass per leaf dry mass. Because of the sampling of the leaves in summer and autumn, alterations in leaf structure during senescence (due to mass resorption or area shrinkage) can bias the parameter resorption efficiency to a large extent (Van Heerwaarden et al. 2003). Therefore, resorption efficiency was also calculated on a leaf unit basis (REunit), or calculated per unit leaf area (REarea). Different ways to express nutrient levels (mass based (concentration), leaf area based (content), and leaf unit based) were used as measurement of resorption proficiency (Killingbeck 1996) for the same reason.


Due to unequal variances, data of N and P concentrations were log(x) transformed, and resorption efficiencies were log(1/1−x) transformed prior to statistical testing (SPSS for Windows 10.1.0, SPSS Inc., Chicago, USA). Data were tested for homogeneity of variance with Levene's test. When transformed data showed unequal variances, data were tested non-parametrically (ranked), but only if this increased Levene's P-value.

In a two-way anova, with species and N treatment as factors, we tested whether treatment effects were general or species specific (shown by a significant treatment–species interaction term). The results of the tests on the differently based data (mass, area or unit leaf) were compared to see whether the measurement basis influences the outcome of the test. As we were only able to collect mass- and area-based, as well as leaf unit-based, data for four out of six species, we repeated the statistical analysis including only data from these four species. The statistics of the two-way anovas are therefore presented both for the complete set of species, and for only the four species for which we had complete data. To test whether the N treatment influenced resorption of N or P for each species, the resorption efficiencies and proficiencies were also separately tested in a one-way anova with N treatment as a fixed factor with two levels.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

nitrogen fertilization effects on green leaf n and p concentrations

In the control plots, the N concentration of mature leaves clearly differed among species (Fig. 1a–c, Table 1). The leaves of the evergreen shrubs E. hermaphroditum and A. polifolia had the lowest N concentrations, and the herb R. chamaemorus the highest, while the other species exhibited intermediate values (Fig. 1a,b). For P the patterns were different, with the highest P concentrations occurring in the two deciduous species B. nana and V. uliginosum (Fig. 1d,e, Table 1), although the large leaves of R. chamaemorus logically had the highest nutrient contents when expressed on a unit basis (Fig. 1f).


Figure 1. Effects of 4 years of N fertilization (10 g N m−2 year−1) on green leaf N and P concentrations (mean ± SE, n= 8) of six sub-arctic bog species expressed on dry mass basis (a, d), area basis (b, e), and unit leaf basis (c, f, note log-scale). White bars, unfertilized controls; grey bars, + N. For Eriophorum vaginatum, N and P are expressed on leaf length (µg N or P mm−1 leaf) rather than area basis. Species: Empetrum hermaphroditum (Eh), Andromeda polifolia (Ap), Vaccinium uliginosum (Vu), Betula nana (Bn), Eriophorum vaginatum (Ev), and Rubus chamaemorus (Rc). m.v. = missing value; *indicates significant fertilization effect (P < 0.05); NS = non-significant.

Download figure to PowerPoint

Table 1.  Results of two-way anova on the effects of species and N treatment and their interaction on N and P concentrations in green leaves of six sub-arctic bog species, growing on eight replicate plots. The left-hand set of results included all species available for the measurement basis concerned: dry mass (DW)-based data, six species; unit-based data, five; area-based data, four. The right-hand set analyses only the four species that had all three sets of data; these were used to compare results between measurement bases
BasisSource4–6 species 4 species
 DWSpecies5 1930.0013 2790.001
N1 3570.0011 2400.001
Species × N5  920.0013 1260.001
N1 2030.0011 1520.001
Species × N4  560.0013  780.001
 AreaSpecies3  700.0013  700.001
N1 2250.0011 2250.001
Species × N3 1400.0013 1400.001
 DWSpecies5  860.0013 1330.001
N1  610.0011  480.001
Species × N5   3.20.0113   1.90.136
N1   9.50.0031  130.001
Species × N4   2.40.0593   1.80.151
 AreaSpecies3  140.0013  140.001
N1  540.0011  540.001
Species × N3   5.60.0023   5.60.002

In general, N treatment led to higher N concentrations in green leaves, but the size of this effect was species-specific, as shown by the significant species–N treatment interaction (Table 1, Fig. 1a–c). The N fertilization effect was not significant in B. nana and R. chamaemorus. In E. hermaphroditum a more than threefold increase in N concentration occurred, whereas the other species increased their green leaf N concentration by up to 75%. Only in R. chamaemorus did the measurement basis (dry mass, area or leaf unit) influence the significance of these responses.

Phosphorus responded differently, with N fertilization generally resulting in a decreased P concentration in green leaves, although not in all species as indicated by the significant species–N interaction (Table 1). Green leaf P concentration decreased by almost 30% in some species (V. uliginosum and E. vaginatum), while in others (A. polifolia, R. chamaemorus, and unit-based E. hermaphroditum) differences were not significant (Fig. 1d–f). The different measurement bases generally yielded the same patterns of significance in all the statistical tests, except for E. hermaphroditum individually and the interaction term of the overall test. This irregularity in the overall test could be due to the different number of species used. However, when only the four species for which dry mass-, unit- and area-based data were available were analysed, rather than all species, only the interaction term of dry mass-based data changed significance (Table 1).

N fertilization changed the N : P ratios, clearly reflecting the contrasting effects of N fertilization on N and P concentrations in green leaves. In the control plots, N : P ratios of green leaves were lower than 14 in most species, which points to N-limited plant growth (Fig. 2). In the N-treated plots, N : P-ratios were significantly higher in all species (except in B. nana) and exceeded the critical value of 16 that may indicate P-limited plant growth.


Figure 2. Effects of 4 years of N fertilization (10 g N m−2 year−1) on green leaf N : P ratios (mass basis) (mean ± SE, n= 8) of six sub-arctic bog species. White bars, unfertilized controls; grey bars, + N; *indicates significant fertilization effect (P < 0.05). Horizontal lines in the graph represent the boundary N : P values correlating with N limitation (N : P < 14) and P limitation (N : P > 16) as proposed by Koerselman & Meuleman (1996). See Figure 1 for abbreviation of species.

Download figure to PowerPoint

nitrogen fertilization effects on n and p resorption efficiency

In the unfertilized controls, N resorption efficiency also clearly differed between species, with values between 40 and 80% (Fig. 3a). No clear patterns emerged in relation to growth forms. In general, N treatment led to lower N resorption efficiencies but, as indicated by the significant species–N interaction, the response was species-specific (Fig. 3a, Table 2). At the individual species level, N resorption efficiency significantly decreased only in A. polifolia, R. chamaemorus (only on DW basis) and V. uliginosum. The largest decrease in N resorption efficiency after N treatment occurred in A. polifolia, down to a value that was not even significantly different from zero (P = 0.513, DW-based data). The type of measurement basis did not influence the significance of these responses, except for R. chamaemorus and V. uliginosum.


Figure 3. Effects of 4 years of N fertilization (10 g N m−2 year−1) on resorption efficiency (mean ± SE, n= 8) of nitrogen (a) and phosphorus (b) of six sub-arctic bog species expressed on dry mass basis (white bars), area basis (light grey bars), unit leaf basis (dark bars), and leaf length basis (striped bar, only for Eriophorum vaginatum). C, unfertilized control; + N, N fertilized; *indicates significant fertilization effect (P < 0.05) within each measurement basis. See Figure 1 for abbreviation of species.

Download figure to PowerPoint

Table 2.  Results of two-way anova on the effects of species and N treatment and their interaction on N and P resorption efficiency of six sub-arctic bog species. See Table 1 for further explanation
BasisSource4–6 species4 species
Species × N5110.0013110.001
Species × N4130.0013 4.90.004
Species × N3120.0013110.001
N1110.0021 0.150.704
Species × N5140.0013 3.50.021
 UnitSpecies4 60.0013 7.20.001
N1 2.250.1381 3.00.087
Species × N415.00.0013 2.50.073
N1 0.900.3471 0.900.347
Species × N3 1.90.1433 1.90.143

Resorption efficiency of P also differed significantly among species in the control plots (values between 30 and 90%), but again no clear pattern emerged when considering growth-form differences (Fig. 3b). Nitrogen fertilization did not have a clear general effect on P resorption efficiency, as responses were species-specific, dependent on the measurement basis (Fig. 3b, Table 2) and for most species insignificant. P resorption efficiency only changed significantly in A. polifolia, decreasing (DW basis) from 71% to 20%, a value not significantly different from zero (P = 0.115, cf the effect on N resorption efficiency). It was also less clear whether P resorption efficiency of V. uliginosum and R. chamaemorus responded to N fertilization, as differences were only significant when expressed on unit or DW basis, respectively. Moreover, in the overall statistics, only the DW-based data yielded a significant interaction term. An explanation for this dependency on measurement basis is that a different number of species had been included in each test (Table 2), although differences and responses in mass resorption may also have played a role. In contrast with N, the measurement basis significantly influenced the overall statistical significance pattern for P when either all species or the four with complete data sets were compared (Table 2 and Fig. 3b).

nitrogen fertilization effects on n and p resorption proficiency

The patterns in N and P concentrations in senescent leaves in the unfertilized control plots were noticeably similar to the N and P concentration patterns in green leaves (compare Fig. 4 with Fig. 1). Thus, there is high N proficiency (low litter N concentration) in both evergreen shrubs and low N proficiency in the herb R. chamaemorus (Fig. 4a–c, Table 3). The patterns were also comparable in fertilized plots, although N and P resorption proficiencies clearly responded differently to N fertilization in A. polifolia. In this species, differences between N concentrations in the control and fertilized plots were considerably larger in litter compared with green leaves; for P, there was no significant N fertilization effect in green leaves, but litter P concentration increased threefold.


Figure 4. Effects of 4 years of N fertilization (10 g N m−2 year−1) on dead leaf N and P concentrations (mean ± SE, n= 8) of six sub-arctic bog species expressed on various bases. For Eriophorum vaginatum (Ev), N and P are expressed on leaf length (µg N or P mm−1 leaf) rather than area basis. White bars, unfertilized controls; grey bars, + N; *indicates significant fertilization effect (P < 0.05). See Figure 1 for further explanation.

Download figure to PowerPoint

Table 3.  Results of two-way anova on the effects of species and N treatment and their interaction on N and P resorption proficiency of six sub-arctic bog species. See Table 1 for further explanation
BasisSource4–6 species 4 species
 DWSpecies5 840.0013  750.001
N12890.0011 1600.001
Species × N5 300.0013  330.001
N1 440.0011 1060.001
Species × N4  30.0013  250.001
 AreaSpecies3 600.0013  600.001
N11470.0011 1470.001
Species × N3 340.0013  340.001
 DWSpecies51000.0013  680.001
N1  0.1730.6791  180.001
Species × N5 120.0013   4.50.007
 UnitSpecies4 410.0013 7570.001
N13400.5621  160.001
Species × N4  20.0013   4.90.004
 AreaSpecies3 240.0013  240.001
N1 250.0011  260.001
Species × N3  5.10.0043   5.10.004

In general, N treatment led to lower N resorption proficiencies, but the effect sizes were species-specific as indicated by the significant species–N treatment interaction (Table 3). The N treatment effect was significant except in B. nana (Fig. 4a–c), and the largest response occurred in A. polifolia, which had a fourfold higher N litter concentration when expressed on DW basis, and almost fivefold higher when expressed on unit basis. Empetrum hermaphroditum and V. uliginosum responded with more than doubling of litter N concentration, whereas other species had smaller increases.

Nitrogen fertilization also affected P litter concentrations, but responses were species-specific (right-hand side of Table 3, Fig. 4d–f). When more than the four ‘complete’ species were tested the overall N fertilization effect was not significant (left-hand side of Table 3).

Considering individual species, responses of P resorption proficiency to N fertilization were more variable than green leaf P responses (which decreased in all species, Fig. 1). Both B. nana and V. uliginosum had almost 50% lower P concentrations in their litter in response to the N treatment, meaning that they had achieved higher P resorption proficiency, whereas no change occurred in other species, apart from A. polifolia where it decreased dramatically. This clearly confirms that the response of P resorption proficiency to N fertilization is species-specific. Different measurement bases gave similar patterns for both N and P resorption proficiency in overall and species-level statistics.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

methodological considerations

It is well known that values of resorption efficiency can be biased by the measurement basis that is used. In a theoretical paper we have shown that considerable under-estimations of resorption efficiency can occur when mass-based data are used instead of area-based data (van Heerwaarden et al. 2003). This is due to the fact that, during leaf senescence, substantial resorption of non-structural compounds can occur, although leaf shrinkage can also lead to bias when using area-based data. In this study we also found considerable mass loss (up to 25%) during senescence in some species (data not shown). In comparative studies such as this one, the species-specific biases have to be taken into account. Moreover, when applying a treatment there is also the possibility of treatment-induced change in mass loss or leaf area during senescence that can lead to bias in the treatment effect. This can partly be overcome by expressing the data on various measurement bases, and checking if the significances of the treatment effect are similar among measurement bases. Our data showed that the treatment and species patterns were hardly influenced by the different measurement bases except for P resorption efficiency. This could be caused by the inhomogeneous group-variances in the P resorption efficiency data, in combination with a species-specific treatment-induced change in mass loss and leaf area shrinkage during senescence, as was the case in R. chamaemorus (area-based mass loss in control plots, 20% ± 3.6 (SD); in fertilized plots, 26% ± 2.1 (SD); P = 0.002) and V. uliginosum (unit-based area loss in control plots, 11% ± 13 (SD); in fertilized plots, 17%± 8 (SD); P = 0.001).

Another general cause for bias in results is that nutrient resorption efficiency has to be determined from two different cohorts of green and dead leaves, as nutrient concentration is measured destructively. This is a problem inherent to all resorption studies. These problems can partly be minimized if the leaves are pre-selected and followed (e.g. photographically) throughout the growing season. This might be feasible for studies of species with large leaves that can easily be tagged and followed. However, most species in this sub-arctic bog habitat are extremely small (height less than 5 cm, i.e. considerably smaller than similar or related species in more fertile and southern habitats) and their leaves are tiny (millimetres, for some of the species). Thus, for a species such as E. hermaphroditum it is not possible to differentiate between leaf samples with an average leaf length of 4 mm or of 5 mm (i.e. a 20% difference) when the leaves are collected in the field. In addition, it is of course possible to choose which mature leaves are sampled, but for the dead leaves of the evergreen species one has to sample ‘what is there’. Moreover, tagging of leaves of these species is practically impossible due to their small size. Thus, in the present study this problem cannot be solved.

There is a risk that strongly different leaf cohorts are compared, especially when investigating evergreen species with long leaf life span. Theoretically, evergreen dead leaves could have been subjected to pre-experimental conditions, whereas green leaves developed during the treatment. This could be the case in E. hermaphroditum, as this species keeps some leaves for about 4 years (Jonasson 1989). Given that this is just as long as the duration of the experiment this type of error is probably small.

For some species we found that leaf areas of green and dead leaves were significantly different. In the control plots, the leaf area of dead Empetrum leaves was 11% larger than that of the green leaves (14% for B. nana). This might be caused by sampling young, not yet maximally expanded, green leaves, or by a change in leaf shape during senescence. In contrast, the V. uliginosum leaf cohort collected in autumn from N-treated plots had 17% smaller leaves than the cohort collected in summer, whereas in the control plots no significant differences were observed. These findings could be explained by accelerated senescence due to N treatment as, if larger leaves had already dropped due to accelerated senescence and their larger weight, this would have led to selectively finding the smaller leaves still attached to the plants in the N-treated plots. Such size-related leaf shedding was also found in another study, where the oldest leaf cohort of the evergreen E. hermaphroditum consisted of few and relatively smaller leaves (Jonasson 1989). Accelerated senescence was also observed in this study in the graminoid E. vaginatum during autumn collection, as leaves from N-treated plots had totally senesced all the way to the leaf base, whereas they had not in the control plots. However, other studies mainly reported that N fertilization delayed senescence (Larcher 1995). Thus, it seems that the response of senescence timing to N fertilization is species-specific.

leaf n and p resorption efficiency and proficiency in sub-arctic peat bog species

The plant species in our study site showed high N and P resorption efficiencies. The four investigated sub-arctic bog species with area-based data available withdrew between 54 and 82% of the N from their senescing leaves. The graminoid E. vaginatum even withdrew 86% (length basis) of its N, and the evergreen A. polifolia 69% (unit basis). These resorption efficiency values are rather high compared with the average N resorption efficiency (50%) found in a large literature survey (Aerts 1996), but are comparable with those found in another study in the same region (Quested et al. 2003). This was also the case for P resorption efficiency: it varied between 39 and 72% in the four species with area-based data, but was even 92% (length basis) in the grass E. vaginatum, and also very high (78%, unit basis) in the evergreen A. polifolia. The lowest P resorption efficiency values were those of the two deciduous species. Our values are very similar to the values reported by Berendse & Jonasson (1992), who made a compilation of resorption data for arctic and sub-arctic plant species, including four species that were also studied by us (B. nana, V. uliginosum, E. hermaphroditum and E. vaginatum). These data suggest that nutrient resorption efficiency in high-latitude species is at the high end of the values as compiled by Aerts (1996) and Killingbeck (1996) for a wide range of ecosystems.

The average litter N concentrations in the control plots varied between 3.7 and 13 mg N g−1 DW among the different species and, for the species with-area based data available, between 33 and 125 µg N cm−2. Phosphorus concentrations in dead leaves grown in the control plots varied between 0.12 and 1.25 mg P g−1 DW among the different species, or (for the species with area-based data available) between 4.1 and 10.9 µg P cm−2. When comparing these resorption proficiency values with the values of Killingbeck (1996) that are indicative for the completeness of resorption, the data for the woody perennial species suggest that resorption of N is complete in some species and intermediate in others, whereas P resorption is classified as less complete (intermediate to incomplete). This pattern correlates well with the observation that plant growth in sub-arctic peat bogs is N-limited, as was found in an earlier study (Aerts et al. 1992), because plants under N limitation are more likely to reach complete resorption of this nutrient compared with the unlimiting nutrients.

effects of increased n supply on n and p resorption efficiency and proficiency

We hypothesized that N resorption efficiency would not be affected by increased N supply, but we found a general decrease in N resorption efficiency, although decreases were significant in only two species individually. In accordance with our hypothesis, N resorption proficiency decreased in all species except B. nana upon N treatment. This implies that the litter N content was higher, and that the amount of N returned to the soil by leaf litter production probably increased.

Our hypothesis that P resorption efficiency would not change after N fertilization was supported by four out of six species. The other two species showed a large decrease (A. polifolia) and a moderate increase (V. uliginosum) in P resorption efficiency after N fertilization. In contrast to our hypothesis, P resorption proficiency only increased in two species (B. nana and V. uliginosum), showing no response to N fertilization in three species, and a decrease in A. polifolia. This pattern might also be explained with the concept of ‘complete resorption’. Using the reference values of Killingbeck (1996), the evergreens A. polifolia and E. hermaphroditum had complete resorption of P in the control plots, whereas B. nana and V. uliginosum had incomplete resorption. As the demand for P probably increased after N fertilization, the species with incomplete resorption could hypothetically increase the resorption proficiency to use the potentially available P, leading to lower litter P concentrations after N fertilization. The species that already had complete resorption under normal conditions were unable to acquire extra P from their senescing leaves owing to physiological constraints and did not show a lower P concentration in their senesced leaves. The herbaceous species did not show a significantly changed P resorption proficiency, which could mean that these species also had complete resorption proficiency. In this case, P proficiency could not increase even though the need for P became higher, as indicated by the increased green leaf N : P ratio in all species.

In this study A. polifolia showed anomalous behaviour compared with the other species, with the results suggesting prematurely arrested resorption of N and P after N fertilization. In the fertilized plots the mature leaves appeared to be red instead of green, probably due to anthocyanins (personal observation), and indicating a plant stress response (Chalker-Scott 1999). Such a response is not a general observation as in other studies A. polifolia showed a significant increase in net primary production (Thormann & Bayley 1997) and increased density (Redbo-Torstensson 1994) after N fertilization. We suspect that under the specific conditions of our experiment this species suffered from strong P limitation, as anthocyanin accumulation is often a sign of severely P-limited plant growth (Marschner 1995).

consequences for nutrient cycling

Given the expected increase in nutrient availability at high-latitude sites due to global warming, our fertilization study provides insight into possible changes in the amount of nutrients returned to the soil through leaf shedding. Mass-based nutrient resorption proficiency is directly related to decomposition characteristics (in the form of litter nutrient concentration) (cf. Quested et al. 2003) and, together with litter quantity, determines the amount of nutrients returned to the soil. Our data show that in all the studied species (except B. nana) the concentration of N in the shed leaves increased strongly (e.g. in A. polifolia fourfold) upon increased N availability. It is most likely that litter quantity also increased upon fertilization, thus adding to the amount of nutrients returned to the soil. The secondary effect of increased N availability on P resorption proficiency is also important: it increased in B. nana and V. uliginosum upon fertilization. Combining the patterns in N and P proficiency suggests that, in many species, the N : P ratio of leaf litter will strongly increase in response to increased N availability, thereby possibly leading to a strong P control over the decomposition process. Such a change from N-controlled to P-controlled litter decay has already been observed in southern Sweden and north-west Europe as a result of increased atmospheric N deposition during the past few decades (Aerts & De Caluwe 1997; Bobbink et al. 1998). It should be noticed, however, that we added high amounts of N (10 g N m−2 year−1) to our plots, almost certainly overestimating future effects. Nevertheless, it is obvious from our data that the potential changes are substantial.

Earlier work (Killingbeck 1996; Aerts & Chapin 2000) and also this study have shown that there are considerable species-specific and growth-form differences in nutrient resorption proficiency, and that growth-form dominance is thus an important component of the plant-mediated feedback on soil nutrient cycling (Aerts 1999; Chapin 1993). It has also been shown that increased N availability can have direct effects on decomposition rates, but also indirect effects through shifts in the species composition of the vegetation (Wookey et al. 1993; 1994; Chapin et al. 1995) and subsequent effects on the chemical composition of the bulk litter (Shaver & Chapin 1991; Hobbie 1996; Cornelissen 1996; Shaw & Harte 2001). We did not determine quantitative changes in species composition of our plots, but cover estimates carried out after 4 years of N fertilization suggested that the cover of both A. polifolia and R. chamaemorus had decreased, whereas the cover of E. vaginatum had increased (data not shown). Increased dominance of a graminoid species following long-term N fertilization was also found in a dwarf shrub heath community close to our site (Press et al. 1998). This emphasizes that, for long-term assessment of plant-mediated effects on soil nutrient cycling, including the partitioning of nutrients between the resorption and the decomposition pathway, we need to know not only how given species respond to global changes, but also how the species composition of the vegetation changes.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

We thank Staffan Karlsson, Richard van Logtestijn, Ellen Dorrepaal and Ivanka Bijlemeer for field assistance and the staff of the Abisko Scientific Research Station in Sweden for facilitating this study. The critical comments of Hans Cornelissen are gratefully acknowledged. This study was supported by USF grant 98.24 of the Vrije Universiteit Amsterdam to RA.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Aerts, R. (1996) Nutrient resorption from senescing leaves of perennials: are there general patterns? Journal of Ecology, 84, 597608.
  • Aerts, R. (1999) Interspecific competition in natural plant communities: mechanisms, trade-offs and plant-soil feedbacks. Journal of Experimental Botany, 50, 2937.
  • Aerts, R. & Chapin, F.S. III (2000) The mineral nutrition of wild plants revisited: a re-evaluation of processes and patterns. Advances in Ecological Research, 30, 167.
  • Aerts, R. & De Caluwe, H. (1997) Nutritional and plant-mediated controls on leaf litter decomposition of Carex species. Ecology, 78, 244260.
  • Aerts, R., Wallén, B. & Malmer, N. (1992) Growth-limiting nutrients in Sphagnum-dominated bogs subject to low and high atmospheric nitrogen supply. Journal of Ecology, 80, 131140.
  • Berendse, F. & Jonasson, S. (1992) Nutrient use and nutrient cycling in northern ecosystems. Arctic Ecosystems in a Changing Climate: an Ecophysiological Perspective (ed. F.S.ChapinIII). Academic Press, San Diego.
  • Bobbink, R., Hornung, M. & Roelofs, J.G.M. (1998) The effects of air-borne nitrogen pollutants on species diversity in natural and semi-natural European vegetation. Journal of Ecology, 86, 717738.
  • Chalker-Scott, L. (1999) Environmental significance of anthocyanins in plant stress responses. Photochemistry and Photobiology, 70, 19.
  • Chapin, F.S. III (1993) The evolutionary basis of biogeochemical soil development. Geoderma, 57, 223227.
  • Chapin, F.S. III, Shaver, G.R., Giblin, A.E., Nadelhoffer, K.J. & Laundre, J.A. (1995) Responses of arctic tundra to experimental and observed changes in climate. Ecology, 76, 694711.
  • Cornelissen, J.H.C. (1996) An experimental comparison of leaf decomposition rates in a wide range of temperate plant species and types. Journal of Ecology, 84, 573582.
  • Eckstein, R.L., Karlsson, P.S. & Weih, M. (1999) Leaf life span and nutrient resorption as determinants of plant nutrient conservation in temperate-arctic regions. New Phytologist, 143, 177189.
  • Gorham, E. (1991) Northern peatlands – role in the carbon-cycle and probable responses to climatic warming. Ecological Applications, 1, 182195.
  • Harte, J. & Shaw, R. (1995) Shifting dominance within a montane vegetation community: results of a climate-warming experiment. Science, 267, 876880.
  • Hobbie, S.E. (1996) Temperature and plant species control over litter decomposition in Alaskan tundra. Ecological Monographs, 66, 503522.
  • Houghton, J.T. (2001) Climate Change 2001: the Scientific Basis. Cambridge University Press, Cambridge.
  • Jonasson, S. (1989) Implications of leaf longevity, leaf nutrient re-absorption and translocation for the resource economy of five evergreen plant species. Oikos, 56, 121131.
  • Kemp, P.R., Waldecker, D.G., Owensby, C.E., Reynolds, J.F. & Virginia, R.A. (1994) Effects of elevated CO2 and nitrogen fertilization pretreatments on decomposition on tallgrass prairie leaf litter. Plant and Soil, 165, 115127.
  • Killingbeck, K.T. (1996) Nutrients in senesced leaves: keys to the search for potential resorption and resorption proficiency. Ecology, 77, 17161727.
  • Koerselman, W. & Meuleman, A.F.M. (1996) The vegetation N : P ratio: a new tool to detect the nature of nutrient limitation. Journal of Applied Ecology, 33, 14411450.
  • Larcher, W. (1995) Physiological Plant Ecology. Springer-Verlag, New York.
  • Marschner, H. (1995) Mineral Nutrition of Higher Plants, 2nd edn. Academic Press, London.
  • Michelsen, A.S., Quarmby, C., Sleep, D. & Jonasson, S. (1998) Vascular plant 15N natural abundance in heath and forest tundra ecosystems is closely correlated with presence and type of mycorrhizal fungi in roots. Oecologia, 115, 406418.
  • Michelsen, A.S., Schmidt, I.K., Jonasson, S., Quarmby, C. & Sleep, D. (1996) Leaf 15N abundance of subarctic plants provides field evidence that ericoid, ectomycorrhizal and non- and arbuscular mycorrhizal species access different sources of soil nitrogen. Oecologia, 105, 5363.
  • Murphy, J. & Riley, J.P. (1962) A modified single solution method for the determination of phosphate in natural waters. Analytica Chimica Acta, 27, 3136.
  • Näsholm, T., Ekblad, A., Nordin, A., Gieslr, R., Högberg, M. & Högberg, P. (1998) Boreal forest plants take up organic nitrogen. Nature, 392, 914916.
  • Press, M.C., Potter, J.A., Burke, M.J.W., Callaghan, T.V. & Lee, J.A. (1998) Responses of a subarctic dwarf shrub heath community to simulated environmental change. Journal of Ecology, 86, 315327.
  • Pugnaire, F.I. & Chapin, F.S. III (1993) Controls over nutrient resorption from leaves of evergreen Mediterranean species. Ecology, 74, 124129.
  • Quested, H.M., Cornelissen, J.H.C., Press, M.C., Callaghan, T.V., Aerts, R., Trosien, F. et al. (2003) Litter decomposition of sub-arctic plant species with differing nitrogen economies: a potential functional role for hemiparasites. Ecology, 84 (12), in press .
  • Redbo-Torstensson, P. (1994) The demographic consequences of nitrogen fertilization of a population of sundew, Drosera rotundifolia. Acta Botanica Neerlandica, 43, 175188.
  • Rustad, L.E., Campbell, J.L., Marion, G.M., Norby, R.J., Mitchell, M.J., Hartley, A.E. 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, 543562.
  • Shaver, G.R. & Chapin, F.S. III (1991) Production: biomass relationships and element cycling in contrasting arctic vegetation types. Ecological Monographs, 61, 131.
  • Shaver, G.R. & Mellilo, J. (1984) Nutrient budgets of marsh plants: efficiency concepts and relation to availability. Ecology, 65, 14911510.
  • Shaw, M.R. & Harte, J. (2001) Control of litter decomposition in a subalpine meadow-sagebrush steppe ecotone under climate change. Ecological Applications, 11, 12061223.
  • Sneller, F.E.C., Van Heerwaarden, L.M., Kraaijeveld-Smit, F.J.L., Ten Bookum, W.M., Koevoets, P.L.M., Schat, H. et al. (1999) Toxicity of arsenate in Silene vulgaris, accumulation and degradation of arsenate-induced phytochelatins. New Phytologist, 144, 223232.
  • Sonesson, M. (1980) Ecology of a Subarctic Mire. LiberTryck, Stockholm.
  • Thormann, M.N. & Bayley, S.E. (1997) Response of aboveground net primary plant production to nitrogen and phosphorus fertilization in peatlands in southern boreal Alberta, Canada. Wetlands, 17, 502512.
  • Van Heerwaarden, L.M., Toet, S. & Aerts, R. (2003) Current measures of nutrient resorption efficiency lead to a substantial underestimation of real resorption efficiency: facts and solutions. Oikos, 101, 664669.
  • Vitousek, P.M. (1984) Litterfall, nutrient cycling, and nutrient limitation in tropical forests. Ecology, 65, 285298.
  • Vitousek, P.M. (1998) Foliar and litter nutrients, nutrient resorption, and decomposition in Hawaiian Metrosideros polymorpha. Ecosystems, 1, 401407.
  • Wookey, P.A., Parsons, A.N., Welker, J.M., Potter, J.A., Callaghan, T.V., Lee, J.A. et al. (1993) Comparative responses of phenology and reproductive development to simulated environmental change in sub-arctic and high arctic plants. Oikos, 67, 490502.
  • Wookey, P.A., Welker, J.M., Parsons, A.N., Press, M.C., Callaghan, T.V. & Lee, J.A. (1994) Differential growth, allocation and photosynthetic responses of Polygonum viviparum to simulated environmental change at a high arctic polar semi-desert. Oikos, 70, 131139.