SEARCH

SEARCH BY CITATION

Keywords:

  • grassland;
  • leaf dry-matter content;
  • leaf nitrogen content;
  • primary productivity;
  • specific leaf area

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • 1
    By comparing plant species under the same experimental field conditions, the direct effects of plant traits on ecosystem processes can be studied. We have analyzed the role of leaf traits (leaf lamina dry matter content, LDMC; leaf lamina N content, LNC and specific leaf lamina area, SLA) for the annual above-ground primary productivity (ANPP) and quality (pepsin-cellulase digestibility, crude protein content) for herbivores of 13 perennial C3 pasture grass species.
  • 2
    These relationships were investigated over 2 years with monocultures grown in a fully factorial block design crossing the plant species, the cutting frequency and the N supply factors.
  • 3
    The within species variation in leaf traits, ANPP, digestibility and protein content was less than between species variation. Species ranks for leaf traits were conserved among N supply and cutting frequency levels. Highly significant (P < 0·001) between species allometric relationships were found for LNC × SLA and SLA × LDMC, with common slopes but differences in intercept and shifts among factor levels.
  • 4
    The between species variation in ANPP was strongly (P < 0·001) and negatively correlated with the fresh-matter based leaf N content (i.e. LDMC × LNC) and was not affected by SLA, apparently because of a trade-off between SLA and leaf lamina fraction. Digestibility increased with SLA and declined with LDMC. Protein content increased with both fresh and dry-matter based LNC.
  • 5
    N supply increased LNC and SLA but reduced LDMC. Cutting frequency increased LDMC and reduced LNC. In response to cutting frequency, changes in digestibility and in fresh-matter based LNC were positively correlated.
  • 6
    We conclude that the between species variation in the annual production of digestible energy and of proteins by pasture grasses is controlled in an additive way by two leaf traits: LNC and LDMC.

Introduction

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

Understanding the role of species for ecosystem functioning (e.g. net primary productivity, litter decomposition rate, herbivory) is one of the major objectives of studies in ecology (Chapin et al. 2000). Various attempts to generalize plant species effects on ecosystem processes have focused on single traits, called ‘functional effect traits’ (Díaz & Cabido 2001; Lavorel & Garnier 2002). Individual traits should, however, not be considered in isolation, because some pairs of traits are sufficiently closely coordinated (Wright et al. 2004) to be thought of as forming a single dimension of strategy variation compounded from several traits (Westoby & Wright 2006).

By comparing plant species under the same experimental field conditions, the direct effects of plant traits on productivity can be studied (Craine et al. 2002). This experimental design allows studying the within species variation in trait values in response to stress and disturbance factors without interferences between species. There are, however, few published studies (e.g. Craine et al. 2002; Fargione & Tilman 2006) analyzing relationships between productivity and plant traits under experimental field conditions.

Grazing, which is one of the most globally widespread land uses (Díaz et al. 2006), has a major importance for primary productivity and chemical composition of pasture species (e.g. Briske, Fuhlendorf & Smeins 2005). Several studies have shown that leaf traits of dominant species vary along gradients of disturbance by grazing and cutting (e.g. Díaz, Noy-Meir & Cabido 2001; Cingolani, Posse & Collantes 2005; Louault et al. 2005; Díaz et al. 2006). Grazing avoidance traits (such as high leaf dry matter content, LDMC) are usually associated with low palatability (Wardle et al. 1998; Cornelissen et al. 1999; Díaz et al. 2001). Grazing tolerance would be favoured by a high specific leaf area (SLA) which increases shoot regrowth ability (Westoby 1999) and by a high leaf N content (LNC) which increases leaf quality and selectivity by herbivores (Pérez-Harguindeguy et al. 2003; Cingolani et al. 2005).

With 13 co-occurring perennial C3 pasture grass species grown in monocultures, we have investigated the role of three leaf traits (LDMC, LNC and SLA) for the above-ground net primary productivity (ANPP) and for the nutritive value (pepsin-cellulase digestibility and crude protein content) for herbivores. To assess species plasticity in response to nitrogen and disturbance, the grass species were compared in a fully factorial block design crossing the N supply and the cutting frequency factors.

Materials and methods

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

PLANT MATERIAL

Thirteen perennial C3 grasses that co-occur in semi-natural mesic grasslands were studied: Alopecurus pratensis, Anthoxanthum odoratum, Arrhenatherum elatius, Dactylis glomerata, Elytrigia repens, Festuca arundinacea, F. rubra, Holcus lanatus, Lolium perenne, Phleum pratense, Poa pratensis, Poa trivialis and Trisetum flavescens. These species are all among the 20 most widely distributed Poaceae species in the French Massif Central (Antonetti et al. 2006). Seeds of these grass species were collected in their native habitat, within 20 km of the study site, in moderate to high fertility semi-natural mountain grasslands managed by grazing and cutting in the French Massif Central.

EXPERIMENTAL DESIGN

The experiment was established in an upland area of central France (Theix, 45°43′N, 03°01′E, 870 m a.s.l.) on a granitic brown soil (Cambic soil, FAO) (43% sand, 36% silt, 21% clay, pH (H2O) 6·2, 5·2% OM). The local climate is semi-continental, with a mean annual temperature of 9 °C ranging from 1 °C in January to 20 °C in August and an average annual rainfall of 760 mm.

A factorial complete block design crossing three factors (plant species, cutting frequency and N supply) with three replicates per treatment combination was used. In May 2001, the grass monocultures were sown in rows (eight rows 18 cm apart) in each plot (2·8 × 1·5 m). The cutting frequency and N fertilizer treatments were started in spring 2002. Two cutting frequencies (bimonthly, three cuts per year, and monthly, six cuts per year, C– and C+, respectively) and two levels of mineral N (NH4NO3) supply (12 and 36 g N m−2 year−1 at N– and N+, respectively, supplied in split applications after each cut) were compared. The N supply levels were established to obtain limiting and non-limiting N nutrition at N– and N+, respectively (Pontes 2006). Cutting frequencies were selected to simulate defoliation frequencies found in hay meadows (C–) and in grazed pastures (C+). In a given cutting treatment, all plots were cut simultaneously at 6 cm height with a mower (Haldrup, Logstor, Denmark). One cut out of two (cuts number 2, 4 and 6) was in common to the two cutting treatments. Phosphorus (8 g P2O5 m−2 year−1) and potassium (24 g K2O m−2 year−1) were supplied at non-limiting rates for growth. Soil volumetric water content was followed fortnightly in the twelve D. glomerata plots using time domain reflectometry (TDR, Trime-FM, Medfield, USA). When soil water content was below 10%, all plots were irrigated. A total of 100 mm of water was supplied in five applications during summers of 2003 (between July 7 and August 15) and 2004 (between June 28 and August 5).

ABOVE-GROUND NET PRIMARY PRODUCTIVITY (ANPP)

At each cutting date, the fresh harvested biomass of each individual plot was automatically collected by the mower and weighed. A subsample was immediately taken, weighed and dried at 60 °C for 48 h to determine the dry matter (DM) content of the harvested biomass and calculate the ANPP of each plot (g DM m−2). The annual ANPP (g DM m−2 year−1) was calculated as the sum of the six (C+) and three (C–) cuts performed each year.

DRY MATTER DIGESTIBILITY AND CRUDE PROTEIN CONTENT

The subsamples used to determine DM content were ground with a 1 mm screen through a Cyclotec sample mill (Model 1093 FOSS TECATOR Inc., Höganäs, Sweden). Each forage sample was analyzed via near-infrared reflectance spectroscopy (NIRS) for crude protein (CP) content and enzymatic pepsin-cellulase dry matter digestibility (DMD). NIR spectra were collected with a Foss-NIRSystems 6500 monochromator (FOSS-NIRSystems, Silver Spring, MD, USA) which scans the spectral range of 400–2500 nm. Modified Partial Least Square (MPLS) calibration equations were developed using 144 samples which were selected from among all the spectra collected (n = 1512). The calibration set was analyzed for CP (using Kjeldahl N × 6·25) and DMD (Aufrère & Demarquilly 1989). All spectra and reference data were recorded and managed with the software winisi Version 1·5 (Infrasoft International, Port Matilda, PA, USA). For CP and DMD, respectively, the results of calibration statistics obtained were: minimum and maximum range (101–288, 492–870 g kg−1), standard error of cross-validation (6·6, 28 g kg−1) and r2 of cross-validation (0·92 for both).

LEAF TRAITS

Leaf traits were measured in June and in September 2003 and 2004, 3 weeks after a cut, which was common to the C– and C+ cutting treatments. Ten tillers were collected at random in each plot, avoiding 20 cm edges, cut with scalpels at ground level and kept in a cold box. In the laboratory, at plant level, the sheath length was first measured (SL). The tiller base was cut in de-ionized water and was then placed at 4 °C in the dark for at least 6 h to allow for full rehydration (Garnier et al. 2001a). After rehydration, the lamina of the youngest fully expanded leaf of each of the ten individuals was measured (LL), weighed and their area was measured with an electronic planimeter (LI 3100, Li-cor, Lincoln, NE, USA). The leaves were then oven dried at 60 °C for 48 h and weighed. Leaf dry matter content (LDMC, leaf lamina dry mass/leaf lamina fresh mass) and specific leaf area (SLA, leaf lamina area/leaf lamina dry mass) were calculated. The leaf lamina N content (LNC) was determined with an elemental analyser (Carlo Erba Instruments, CNS NA 1500, ThermoFinnigan, Milan, Italy) for each sample. The fresh matter based leaf N content (LNCF, g N g−1 FM) was calculated as LNC × LDMC.

DATA ANALYSIS

Means of traits are means of two measurement dates (June and September) and of 2 years (2003 and 2004). Pontes (2006) has shown that species ranks are conserved between seasons and years. For each plot, annual means of DMD and CP were calculated as a weighted average, based on the harvested dry matter (ANPP) value at each cut. This procedure allows the calculation of a mean annual quality for the total herbage harvested during the 2 years. Analyses of variance (anova) were performed using the statistical analysis package SAS (sas Institute 2000, ver. 8, Cary, NC, USA) with the species, cutting regime, N supply and block factors. Plant tissue composition data (LDMC, LNC, CP and DMD) were transformed prior to anova using the Arcsin (square root) function and ANPP data were transformed by square root to normalize the data.

Relationships among leaf traits were studied after log-transformation of values using standardized major axis (SMA) regression (Wright et al. 2004). SMA slopes were calculated as the linear regression slope divided by the correlation coefficient (r) (Sokal & Rohlf 1995). Statistical comparisons among treatments for SMA slopes and intercepts were performed according to Warton et al. (2006) in R (Ihaka & Gentleman 1996).

Relationships between productivity, nutritive value and leaf traits, were analyzed by ancova (Statgraphics Plus, Manugistics, Rockville, MA, USA) using species means within each block (fixed factor) and within each N supply and cutting frequency level (n = 39). ancova between responses to treatment factors were performed for ANPP and nutritive value and leaf traits. Spearman's rank coefficients were used to compare species ranks for traits among treatments.

Results

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

CUTTING FREQUENCY AND N SUPPLY EFFECTS

The species and N supply factors were significant for all variables analyzed (Table 1). The species factor was the single largest source of variation (44%–91% of the total variance explained) for all variables but CP. The cutting frequency factor was significant for all variables but SLA. Despite a significant effect for leaf traits, the block factor accounted only for a small share of the total variance (Table 1). For LNC, the species × N supply and the N supply × cutting frequency interactions were significant. For ANPP, CP and DMD the first order interactions were significant (except the species × N interaction for CP and DMD, and the cut × N interaction for DMD).

Table 1.  Statistical significance of F ratios in anova's for leaf traits, productivity and nutritive value of the grass populations
 dfLDMCLNCSLAANPPCPDMD
VEFPVEFPVEFPVEFPVEFPVEFP
  1. Means of 2 years were analyzed for each individual plot (n = 156). The factors included in the model are: species, cutting frequency, N supply, block and their interactions (except interactions with the block). The second order interaction between species, cutting frequency and N supply was never significant. df, degree of freedom; VE, variance explained (%); ns, not significant; *P < 0·05; **P < 0·01; ***P < 0·001. LDMC, leaf dry matter content; LNC, leaf lamina N content; SLA, specific leaf area; ANPP, above-ground net primary productivity; CP, crude protein; DMD, pepsin-cellulase dry matter digestibility.

Species127454***4481***91167***7284***1835***5571***
Cut 12·320***6·3135***< 11·6ns4·664***40921***25390***
N 15·447***36796***1·534***8·1112***33766***7·8120***
Block 22·812***6·470***1·315***< 11·4ns< 10·5ns< 12·6ns
Species × Cut12nsnsns3·84·4***2·03·6***3·44·4***
Species × N12ns1·52·9**ns2·32·7**nsns
Cut × N 1ns< 117***ns< 16·8*< 115***ns

Relative responses to cutting frequency and N supply were plotted as box plots indicating the variability observed among species (Fig. 1). Data (means per species within each N supply and cutting frequency levels) used to calculate these relative responses are shown in Supplementary Material (Supplementary Tables S1 and S2).

image

Figure 1.  Box and Whisker plot of the responses to N supply (a, b) and cutting frequency (c, d) of leaf traits (a, c) and of above-ground DM productivity, DM digestibility and crude protein content (b, d). N–, 12 g N m−2 year−1; N+, 36 g N m−2 year−1; C–, three cuts per year; C+, six cuts per year. LDMC, leaf dry matter content; LNC, leaf lamina N content; LNCF, fresh matter based leaf N content; SLA, specific leaf area; ANPP, above-ground net primary productivity; CP, crude protein; DMD, pepsin-cellulase dry matter digestibility. Values are means per species (n = 13). The centre lines within each box show the location of the sample medians. The lower whisker is drawn from the lower quartile to the smallest point within 1·5 interquartile ranges from the lower quartile. The other whisker is drawn from the upper quartile. Circles indicate species outside this range.

Download figure to PowerPoint

For all species, LNC was increased by N supply and reduced by cutting frequency (Fig. 1a,c). In the same way, for most species, SLA (Fig. 1a,c) was increased by N supply (except for D. glomerata) and was reduced by cutting frequency (except for D. glomerata, Ph. pratense and P. trivialis). Only A. pratensis displayed a positive response of LNCF (i.e. the leaf N content per unit fresh matter, mg N g−1 FM) to cutting frequency (Fig. 1c). Except for F. arundinacea and P. trivialis, the relative changes of LDMC (Fig. 1a,c) were negative in response to N supply and positive in response to cutting frequency.

The ANPP was usually increased by the N supply and reduced by the cutting frequency factors and displayed large but highly variable relative responses among species (Fig. 1b,d). Only two species (L. perenne and P. trivialis) had their production increased at a high compared to a low cutting frequency. There was a considerable variability of the grass species in the relative response to N supply (+4·8% to +52% increase between N– and N+). In contrast, nutritive value variables (DMD and CP) were stimulated by cutting frequency and N supply with relatively small variation among species (Fig. 1b,d).

For all traits, as well as for productivity and nutritive value variables, the species ranking was highly correlated (P < 0·001) between the cutting frequency and N supply (Table 2) treatments.

Table 2.  Comparisons of species ranking (Spearman's rank correlation) for leaf traits, productivity and nutritive value variables among cutting frequency and N supply treatments. Values are means per species (n = 13). Same abbreviations as in Table 1
 Cutting frequencyN supply
  • *

    P < 0·001.

LDMC0·95*0·98*
LNC0·98*0·94*
SLA0·98*0·98*
ANPP0·94*0·93*
CP0·81*0·96*
DMD0·89*0·97*

RELATIONSHIPS AMONG LEAF TRAITS

Allometric relationships among leaf traits were studied using SMA regressions (Table 3). The SLA × LDMC regression (P < 0·001) displayed a slope between –3·35 and –3·05. Hence, a 10-fold increase in SLA coincided with 2-fold decline in LDMC. LNC was positively correlated (P < 0·001) with SLA (SMA slope between 0·39 and 0·45). At C– and at N+, a negative correlation (P < 0·05) was observed between LNC and LDMC. Individual slopes per factor level did not differ significantly from the common slope for all pairwise regressions (Table 3).

Table 3.  Allometric relationships between leaf traits
(a) Relationships for each cutting frequency level
Y (log)X (log)C–C+Cut effect
r2Sloper2SlopeP slopeP interceptP shift†
LNCLDMC0·18**–1·26 (–1·70, –0·94)0·08 ns–1·34 (–1·84, –0·97)nsns**
SLALDMC0·48***–3·05 (–3·87, –2·40)0·39***–3·33 (–4·31, –2·58)nsnsns
LNCSLA0·49*** 0·41 (0·33, 0·52)0·40*** 0·40 (0·31, 0·52)ns**ns
(b) Relationships for each N supply level
Y (log)X (log)N–N+N effect
r2Sloper2SlopeP slopeP interceptP shift
  1. Standardized major axis (SMA) regressions were calculated after log-transformation of each variable. Coefficients of determination (r2), SMA slopes (95% confidence intervals in parentheses) and intercepts of slopes were calculated according to Warton et al. (2006) with means per species and block (n = 39) within each factor level. For abbreviations, see Table 1. C–, three cuts per year; C+, six cuts per year; N–, 12 g N m−2 year−1; N+, 36 g N m−2year−1. ns, not significant; *P < 0·05; **P < 0·01; ***P < 0·001.

  2. †Testing for significance between treatment differences in shift along the x axis.

LNCLDMC0·10 ns–1·39 (–1·90, –1·02)0·16*–1·29 (–1·74, –0·95)ns******
SLALDMC0·47***–3·13 (–3·98, –2·46)0·39***–3·35 (–4·33, –2·59)nsnsns
LNCSLA0·38*** 0·45 (0·34, 0·58)0·47*** 0·39 (0·30, 0·49)ns******

Significant intercept differences between cutting frequencies occurred only for LNC × SLA (Table 3, P < 0·01). An increase in cutting frequency increased the intercept of the LNC × SLA regression (Fig. 2a). An increase in N supply increased intercepts of both LNC × SLA and LNC × LDMC regressions (Table 3, P < 0·001) (Fig. 2b). Shifts in SLA and LDMC were also found in response to N supply and cutting frequency (Table 3).

image

Figure 2.  Standardized major axis (SMA) relationships between LDMC (mg g−1), LNC (mg g−1 DM) and SLA (m2 kg−1 DM), within each cutting frequency (a) and N supply (b) level. Solid lines indicate SMA slopes for C– (three cuts per year) and N– (12 g N m−2 year−1) and dashed lines indicate slopes for C+ (six cuts by year) and N+ (36 g N m−2 year−1). Slope and r2 values are given in Table 3. Alopecurus pratensis (Ap), Anthoxanthum odoratum (Ao), Arrhenatherum elatius (Ae), Dactylis glomerata (Dg), Elytrigia repens (Er), Festuca arundinacea (Fa), Festuca rubra (Fr), Holcus lanatus (Hl), Lolium perenne (Lp), Phleum pratense (Php), Poa pratensis (Pp), Poa trivialis (Pt) and Trisetum flavescens (Tf). ‘+’ for C+ and N+; ‘–’ for C– and N–.

Download figure to PowerPoint

SLA (P < 0·001) and LNC (P < 0·05) were both positively correlated with the ratio of sheath to leaf lamina length (SL/LL) (Fig. 3).

image

Figure 3. Relationships between the ratio of sheath to leaf lamina length (SL : LL) and specific leaf area (SLA) (a) and leaf lamina N content (LNC) (b). (a) n = 13, r2 = 0·81, P < 0·001; (b) n = 13, r2 = 0·62, P < 0·05. Values are means of 2 years and of four treatments per species. Same abbreviations as in Fig. 2.

Download figure to PowerPoint

PRODUCTIVITY NUTRITIVE VALUE AND LEAF TRAITS

First order interactions (species × N and species × cut) were significant but not the second order interaction (species × N × cut) (Table 1). Thus, the role of leaf traits was analyzed by ancova at each N supply level and at each cutting frequency level (Table 4). ANPP was negatively correlated with LNC (P < 0·05) and with LNCF (P < 0·001) for all factor levels (Table 4). DMD was always negatively correlated to LDMC (P < 0·01) and positively correlated to SLA (P < 0·001). CP was positively correlated (P < 0·01) to LNC and LNCF. Finally, DMD was also significantly correlated to LNC for all factor levels but C–. Individual slopes did not differ significantly between C– and C+, and between N– and N+ (Table 4), except for ANPP × LNCF (Table 4a). The intercept of this regression was also significantly affected (P < 0·05) by the cutting frequency (Table 4a).

Table 4.  Analyses of covariance between leaf traits, productivity and forage quality variables were performed with means per species and block (n = 39, block factor was not significant) within each factor level (± SE)
(a) Relationships for each cutting frequency level
 C–C+Cut effect
r2SlopeInterceptr2SlopeInterceptP slopeP intercept
ANPP × LDMCns–4·6 ± 2·691939 ± 600·00·13*–3·4 ± 1·571544 ± 361·4nsns
ANPP × LNC0·20**–33 ± 11·52259 ± 473·20·12*–16 ± 7·61389 ± 289·9nsns
ANPP × LNCF0·45***–226 ± 42·52976 ± 388·20·30***–104 ± 27·31677 ± 237·6***
CP × LNC0·28**1·5 ± 0·4292 ± 17·50·51***2·3 ± 0·38104 ± 14·7nsns
CP × LNCF0·22**5·9 ± 1·98100 ± 18·10·34***7·4 ± 1·79126 ± 15·6nsns
DMD × LDMC0·27**–1·2 ± 0·34918 ± 76·30·51***–1·4 ± 0·231024 ± 52·2nsns
DMD × LNCns3·4 ± 1·71513 ± 70·70·24**4·6 ± 1·37534 ± 52·2nsns
DMD × SLA0·39***5·1 ± 1·10528 ± 27·40·62***5·4 ± 0·71579 ± 17·6nsns
(b) Relationships for each N supply level
 N–N+N effect
r2SlopeInterceptr2SlopeInterceptP slopeP intercept
  1. Same abbreviations as in Table 1. C–, three cuts by year; C+, six cuts by year; N–, 12 g N m−2 year−1; N+, 36 g N m−2 year−1. LNCF is the fresh matter based leaf N content (calculated as LNC × LDMC). Only significant correlations for at least one factor level are presented. r2, coefficients of determination. ns, not significant; *P < 0·05; **P < 0·01; ***P < 0·001.

ANPP × LDMC0·11*–3·7 ± 1·801622 ± 416·7ns–4·2 ± 2·531864 ± 560·0nsns
ANPP × LNC0·18*–23 ± 8·41574 ± 300·70·13*–24 ± 10·81956 ± 469·7nsns
ANPP × LNCF0·39***–141 ± 29·71916 ± 246·00·32***–170 ± 42·32552 ± 403·6nsns
CP × LNC0·38***1·9 ± 0·4286 ± 15·00·40***1·8 ± 0·37112 ± 16·0nsns
CP × LNCF0·25**6·5 ± 1·90102 ± 15·70·30***6·7 ± 1·76124 ± 16·8nsns
DMD × LDMC0·43***–1·4 ± 0·28988 ± 63·80·42**–1·4 ± 0·27995 ± 60·0nsns
DMD × LNC0·20**4·5 ± 1·58503 ± 56·80·14*3·5 ± 1·44546 ± 62·7nsns
DMD × SLA0·51***5·9 ± 0·98527 ± 23·60·51***4·8 ± 0·79576 ± 20·2nsns

DMD response to cutting frequency was significantly (P < 0·01) and positively correlated to LDMC (r2 = 0·31) and to LNCF (r2 = 0·27) responses. ANPP and CP responses to N supply and cutting frequency were not significantly correlated to leaf trait responses (data not shown). The annual production of DMD and of proteins (see Supplementary Table S3) was significantly and negatively correlated (P < 0·05) to LNCF for all factor levels (Fig. 4a,b).

image

Figure 4. Relationship between the annual digestible dry matter (ANPP × DMD, g digestible DM m−2 year−1) and crude protein (ANPP × CP, g proteins m−2 year−1) production and the leaf N content per unit of fresh matter (LNCF, mg g−1), within each cutting frequency level (a) and N supply level (b). Solid lines indicate slopes for C– and N– and dashed lines indicate slopes for C+ and N+. Slopes and intercepts were significantly different among cutting frequency levels for digestible DM × LNCF regression (P < 0·05). Same abbreviations as in Fig. 2.

Download figure to PowerPoint

Discussion

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

BETWEEN AND WITHIN SPECIES VARIATION IN LEAF TRAITS

The perennial C3 grass species that originated from a common habitat (mesic semi-natural grasslands) had contrasting leaf trait values. On average 70% of the total variance in LDMC, LNC and SLA could be ascribed to the species factor (Table 1). The within species variation of leaf traits in response to N supply and cutting frequency was small (average relative change in trait values below 14%, Fig. 1a,c), except for the 23% increase in LNC in response to N supply. Therefore, in agreement with other findings at the plant community scale (Garnier et al. 2004), the within species variation in leaf traits was less than the between species variation. Moreover, the species ranks of all leaf traits were conserved for two contrasting levels of cutting frequency and of N supply (Table 2). This is in good agreement with previous findings showing that species ranks for LDMC and SLA are conserved both between different years and sites (Garnier et al. 2001b) and in response to resource gradients (Wilson, Thompson & Hodgson 1999; Al Haj Khaled et al. 2005).

CORRELATIONS AMONG TRAITS

Our results confirm the correlations already found between SLA, LNC and LDMC (Wilson et al. 1999; Wright et al. 2004; Shipley et al. 2005). The main new contribution is the planned comparison between resource (N) and disturbance (cutting frequency) levels allowing us to compare SMA slopes and intercepts. For two leaf trait combinations (LNC × SLA and SLA × LDMC), common SMA slopes were found across N supplies and cutting frequencies (Table 3). Therefore, the co-ordinated relationship along resource and disturbance gradients, already reported for LNC × SLA (Reich, Ellsworth & Walters 1998; Wright, Reich & Westoby 2001; Wright et al. 2004), also applies to SLA × LDMC in perennial grasses. In contrast, the LNC × LDMC regression had a lower r2 (between 0·08 and 0·18) and was not always significant at P < 0·05.

The SMA slope range reported here for LNC × SLA (0·30–0·59) is lower than the range (0·76–0·81) reported by Wright et al. (2004). This result was expected due to the narrower range of growth forms and geographic regions in our study, compared to that by Wright et al. (2004), which covers a large array of plant species from contrasting biomes.

Differences in intercepts in response to N supply were found in the co-ordinated relationships among leaf traits (Fig. 2b). This appears to constitute physiological plasticity in response to nutrient availability (Wilson et al. 1999; Wright et al. 2005), leading to increased SLA and LNC, and reduced LDMC at high compared to low N (Fig. 1, Table 1), which could favour resource acquisition.

In response to cutting frequency, a difference in intercept (Table 3) was also found for LNC × SLA, indicating a decline in LNC in response to increased cutting frequency (Fig. 1, Table 1). This decline may lead to reduced N losses during defoliation (Gutschick 1999) and could also result from the reduction in N reserves observed in frequently defoliated plants (Thornton & Millard 1997).

Cutting frequency induced a decline in LNC and an increase in LDMC and, hence, reduced the nutritive value at leaf lamina level. Nevertheless, at the plant level, the nutritive value (DMD and CP) of shoots was increased (Fig. 1d) because of a reduced development of stems and spikes (data not shown, Pontes 2006) at frequent compared to infrequent cuts. Indeed, at the plant level, changes in cutting frequency affect herbage digestibility mostly through alterations in the ratio between vegetative and reproductive shoots (Santis et al. 2004; Pontes 2006).

Grazing avoidance results from slow growing shoot tissues rich in structural defences (Briske 1996). It has been associated with changes in leaf morphology that reduce both the digestibility and the prehension of leaves by grazers (Westoby 1999; Díaz et al. 2001; Cingolani et al. 2005). Defoliation by cutting would trigger the same type of avoidance response, as we have observed a reduced palatability of leaves mediated by a rise in LDMC and a decline in LNC at high compared to low defoliation frequency. In contrast, we find no evidence for a tolerance response to cutting that would increase the rate of leaf recovery (Caldwell et al. 1981; Westoby 1999), since neither SLA nor LNC increased at high compared to low defoliation frequency.

PRODUCTIVITY AND LEAF TRAITS

Our results do not confirm previous findings by Poorter & De Jong (1999) who reported a positive relationship between ANPP and SLA. According to Reich, Walters & Ellsworth (1992), different patterns of biomass allocation may cause the lack of relationship between ANPP and SLA. Species with low SLA have been shown to allocate a relatively large fraction of their total biomass to foliage. Conversely, species with high SLA have a relatively large fraction of the above-ground biomass allocated to stems, due to a higher efficiency of their foliage biomass (Reich et al. 1992; Wright et al. 2001). Under our conditions, the SLA was positively correlated with the sheath to leaf lamina length ratio (Fig. 3a). Thus, in good agreement with the hypothesis of Reich et al. (1992), a grass species with high SLA would allocate relatively less growth to leaf laminae than to sheath. Because of this trade-off with the leaf lamina fraction, the ANPP would not increase with SLA.

Among grass species, ANPP was negatively correlated to dry-matter and especially fresh-matter based leaf N content (LNCF, P < 0·001, Table 4). Craine et al. (2002) have reported for N limited plants a negative relationship between ANPP and plant N content. Grass RGR was also negatively correlated with LNC (Soussana et al. 2005) because of a reduction in N productivity, which is the ratio of biomass production per unit time to the amount of N in the plant. A high LNCF would reduce ANPP through a decline in N productivity, since for the same leaf volume more dry-matter and more N per unit dry-matter are required at high compared to low LNCF.

In grasses the leaf growth zone is the main site of shoot growth where anatomical and chemical characteristics of leaves originate. The leaf elongation rate and the length of the leaf growth zone (region where cell division and expansion occurs) exhibit negative correlations to LDMC (Arredondo & Schnyder 2003). LNC is closely associated with N deposited during cell production (Gastal & Nelson 1994). Therefore, we suggest that N productivity declines with N deposition per unit cell volume in the growth zone. Accordingly, a large N dilution would result in a fast regrowth rate through increased N productivity.

In plants that are not severely disturbed, a long leaf life span favours N conservation (Berendse & Aerts 1987; Aerts & Chapin 2000). We have found a positive relationship between LNC and the ratio of sheath to leaf lamina length (Fig. 3b). Grasses with a high LNC (low N productivity) would therefore invest relatively more in sheaths compared to laminae, which would increase the mean residence time of N (Berendse & Elberse 1989) and reduce N losses by defoliation.

Taken together, our results suggest a trade-off between the two components (N productivity and mean residence time of N) of nitrogen-use efficiency (Aerts & Chapin 2000): grasses with high LNC and high sheath fraction would have a low N productivity and a high mean residence time of N. Conversely, grasses with low LNC and low sheath fraction would have a high N productivity and a low mean residence time of N. Defoliation removes three to six times more N than leaf senescence in grass monocultures cut each month or every 2 months (Soussana et al. 2005). Therefore, cuts strongly constrain the mean residence time of N, which explains why ANPP is more affected by N productivity than by mean residence time of N under our experimental conditions.

NUTRITIVE VALUE AND LEAF TRAITS

At the plant community scale, the shoot digestibility was negatively correlated with the community-aggregated LDMC in a long-term grazing experiment (Louault et al. 2005). At the leaf lamina scale, digestibility was correlated negatively with LDMC and positively with SLA (Al Haj Khaled et al. 2006).

Our results show that the between species variation in the digestibility of grass shoots is highly correlated (P < 0·01) with LDMC and with SLA. Moreover, these relationships are conserved among N supply and cutting frequency levels (Table 4). A low investment in cell wall material (lignin, hemicellulose and cellulose) and relatively more cytoplasmic elements such as proteins (Van Arendonk & Poorter 1994) would explain the higher digestibility of leaves from grass species with low LDMC and high SLA values. The between species variation in shoot crude protein content could be predicted from the fresh and dry-matter based leaf N content (LNCF and LNC).

In summary, only two leaf traits (LDMC and LNC) were required to predict the above-ground productivity and quality (digestibility and crude protein content) under our experimental conditions. An integrated measurement of the energy and protein supply to herbivores is the annual production of digestible dry matter (i.e. ANPP × DMD) and of proteins (ANPP × CP) (Fig. 4). Both variables were highly and negatively correlated (P < 0·001) with a single trait, the leaf N content per unit fresh matter (LNCF), which has been calculated in our study as LNC × LDMC. Hence, this single trait is likely to be a useful indicator of the provision of energy and proteins to herbivores by pasture grass monocultures. However, our findings with monocultures cannot be extrapolated without care to grassland communities.

In conclusion, this study shows that between species variation in above-ground productivity and quality of temperate pasture grasses is partly controlled by leaf traits and especially by the leaf lamina N content per unit fresh matter.

Acknowledgements

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

The authors thank B. Pons and S. Toillon for help in field and laboratory measurements and Drs K. Klumpp, D. Vile and S. Fontaine for valuable comments on previous versions of the manuscript. The financial support of the French ECCO PNBC ‘GEOTRAITS’ research project and of INRA for a Doctoral grant to L. da S. Pontes are gratefully acknowledged. This work contributes to the GDR ‘UTILITERRES’.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Aerts, R. & Chapin, F.S. (2000) The mineral nutrition of wild plants revisited: a re-evaluation of process and patterns. Advances in Ecological Research, 30, 167.
  • Al Haj Khaled, R., Duru, M., Decruyenaere, V., Jouany, C. & Cruz, P. (2006) Using leaf traits to rank native grasses according to their nutritive value. Rangeland Ecology Management, 59, 648654.
  • Al Haj Khaled, R., Duru, M., Theau, J.P., Plantureux, S. & Cruz, P. (2005) Variations in leaf traits through seasons and N-availability levels and its consequences for ranking grassland species. Journal of Vegetation Science, 16, 391398.
  • Antonetti, P., Brugel, E., Kessler, F., Barbe, J.P. & Tort, M. (2006) ‘Atlas de la Flore d’Auvergne’, p. 981. Conservatoire botanique national du Massif Central, Le Bourg. ISBN 2952472203.
  • Arredondo, J.T. & Schnyder, H. (2003) Components of leaf elongation rate and their relationship to specific leaf area in contrasting grasses. New Phytologist, 158, 305314.
  • Aufrère, J. & Demarquilly, C. (1989) Predicting organic matter digestibility of forage by two pepsin-cellulase methods. Proceedings of the 16th International Grassland Congress, Nice, France, 2, 877878.
  • Berendse, F. & Aerts, R. (1987) Nitrogen-use efficiency: a biologically meaningful definition? Functional Ecology, 1, 293296.
  • Berendse, F. & Elberse, W.T. (1989) Competition and nutrient losses from the plant. Causes and Consequences of Variation in Growth Rate and Productivity of Higher Plants (eds H.Lambers, H.Konings, M.L.Cambridge & T.L.Pons), pp. 269284. SPB Academic Publishing, The Hague, Netherlands.
  • Briske, D. (1996) Strategies of plant survival in grazed systems: a functional interpretation. The Ecology and Management of Grazing Systems (eds J.Hodgson & A.W.Illius), pp. 3767. CAB International, Wallingford.
  • Briske, D.D., Fuhlendorf, S.D. & Smeins, E.E. (2005) State- and transition models, thresholds, and rangeland health: a synthesis of ecological concepts and perspectives. Rangeland Ecology and Management, 58, 110.
  • Caldwell, M.M., Richards, J.H., Johnson, D.A., Nowak, R.S. & Dzurec, R.S. (1981) Coping with herbivory: photosynthetic capacity and resource allocation in two semiarid Agropyron bunchgrasses. Oecologia, 50, 1424.
  • Chapin, F.S. III, Zavaleta, E.S., Eviner, V.T., Naylor, R.L., Vitousek, P.M., Reynolds, H.L., Hooper, D.U., Lavorel, S., Sala, O.E., Hobbie, S.E., Mack, M.C. & Díaz, S. (2000) Consequences of changing biodiversity. Nature, 405, 234242.
  • Cingolani, A.M., Posse, G. & Collantes, M.B. (2005) Plant functional traits, herbivore selectivity and response to sheep grazing in Patagonian steppe grasslands. Journal of Applied Ecology, 42, 5059.
  • Cornelissen, J.H.C., Perez-Harguindeguy, N., Díaz, S., Grime, J., Marzano, B., Cabido, M., Vendramini, F. & Cerabolini, B. (1999) Leaf structure and defence control litter decomposition rate across species and life forms in regional floras on two continents. New Phytologist, 143, 191200.
  • Craine, J.M., Tilman, D., Wedin, D., Reich, P., Tjoelker, M. & Knops, J. (2002) Functional traits, productivity and effects on nitrogen cycling of 33 grassland species. Functional Ecology, 16, 563574.
  • Díaz, S. & Cabido, M. (2001) Vive la difference: plant functional diversity matters to ecosystem processes. Trends in Ecology and Evolution, 16, 646655.
  • Díaz, S., Lavorel, S., McIntyre, S., Falczuk, V., Casanoves, F., Milchunas, D.G., Skarpe, C., Rusch, G., Sternberg, M., Noy-Meir, I., Landsberg, J., Zhang, W., Clark, H. & Campbell, B.D. (2006) Plant trait responses to grazing – a global synthesis. Global Change Biology, 12, 129.
  • Díaz, S., Noy-Meir, I. & Cabido, M. (2001) Can grazing response of herbaceous plants be predicted from simple vegetative traits? Journal of Applied Ecology, 38, 497508.
  • Fargione, J. & Tilman, D. (2006) Plant species traits and capacity for resource reduction predict yield and abundance under competition in nitrogen-limited grassland. Functional Ecology, 20, 533540.
  • Garnier, E., Cortez, J., Billes, G., Navas, M.L., Roumet, C., Debussche, M., Laurent, G., Blanchard, A., Aubry, D., Bellmann, A., Neill, C. & Toussaint, J.P. (2004) Plant functional markers capture ecosystem properties during secondary succession. Ecology, 85, 26302637.
  • Garnier, E., Laurent, G., Bellmann, A., Debain, S., Berthelier, P., Ducout, B., Roumet, C. & Navas, M.-L. (2001b) Consistency of species ranking based on functional leaf traits. New Phytologist, 152, 6983.
  • Garnier, E., Shipley, B., Roumet, C. & Laurent, G. (2001a) A standardized protocol for the determination of specific leaf area and leaf dry matter content. Functional Ecology, 15, 688695.
  • Gastal, F. & Nelson, C.J. (1994) Nitrogen use within the growing leaf blade of tall fescue. Plant Physiology, 105, 191197.
  • Gutschick, V.P. (1999) Biotic and abiotic consequences of differences in leaf structure. New Phytologist, 143, 318.
  • Ihaka, R. & Gentleman, R. (1996) R: a language for data analysis and graphics. Journal of Computational and Graphical Statistics, 5, 299314.
  • Lavorel, S. & Garnier, E. (2002) Predicting changes in community composition and ecosystem functioning from plant traits: revisiting the Holy Grail. Functional Ecology, 16, 545556.
  • Louault, F., Pillar, V.D., Aufrère, J., Garnier, E. & Soussana, J.F. (2005) Plant traits functional types in response to reduced disturbance in a semi-natural grassland. Journal of Vegetation Science, 16, 151160.
  • Pérez-Harguindeguy, N., Díaz, S., Vendramini, F., Cornelissen, J.H.C., Gurvich, D.E. & Cabido, M. (2003) Leaf traits and herbivore selection in the field and in cafeteria experiments. Australian Ecology, 28, 642650.
  • Pontes, L. da S. (2006) Diversité Fonctionnelle des Graminées Prairiales: Conséquences Pour la Productivité et Pour la Valeur Nutritive. Thesis, Université Blaise Pascal, Clermont-Ferrand, France.
  • Poorter, H. & De Jong, R. (1999) A comparison of specific leaf area, chemical composition and leaf construction costs of field plants from 15 habitats differing in productivity. New Phytologist, 143, 163176.
  • Reich, P.B., Ellsworth, D.S. & Walters, M.B. (1998) Leaf structure (specific leaf area) modulates photosynthesis–nitrogen relations: evidence from within and across species and functional groups. Functional Ecology, 12, 948958.
  • Reich, P.B., Walters, M.B. & Ellsworth, D.S. (1992) Leaf life-span in relation to leaf, plant, and stand characteristics among diverse ecosystems. Ecological Monographs, 62, 365392.
  • Santis, G.D., Ianucci, A., Dantone, D. & Chiaravalle, E. (2004) Changes during growth in the nutritive value of components of berseen clover (Trifolium alexandrinum L.) under different cutting treatments in a Mediterranean region. Grass and Forage Science, 59, 378388.
  • Shipley, B., Vile, D., Garnier, E., Wright, I. J. & Poorter, H. (2005) Functional linkages between leaf traits and net photosynthetic rate: reconciling empirical and mechanistic models. Functional Ecology, 19, 602615.
  • Sokal, R.R. & Rohlf, F.J. (1995) Biometry: The Principles and Practice of Statistics in Biological Research. W.H. Freeman, New York.
  • Soussana, J.F., Teyssonneyre, F., Picon-Cochard, C. & Dawson, L. (2005) A trade-off between nitrogen uptake and use increases responsiveness to elevated CO2 in infrequently cut mixed C3 grasses. New Phytologist, 166, 217230.
  • Thornton, B. & Milland, P. (1997) Increased defoliation frequency depletes remobilization of nitrogen for leaf growth in grasses. Annals of Botany, 80, 8995.
  • Van Arendonk, J.J.C.M. & Poorter, H. (1994) The chemical composition and anatomical structure of leaves of grass species differing in relative growth rate. Plant, Cell and Environment, 17, 963970.
  • Wardle, D.A., Barker, G.M., Bonner, K.I. & Nicholson, K.S. (1998) Can comparative approaches based on plant ecophysiological traits predict the nature of biotic interactions and individual plant species effects in ecosystems? Journal of Ecology, 86, 405420.
  • Warton, D.I., Wright, I.J., Falster, D.S. & Westoby, M. (2006) Bivariate line-fitting methods for allometry. Biological Reviews, 81, 259291.
  • Westoby, M. (1999) A LHS strategy scheme in relation to grazing and fire. Proceedings of the VIth International Rangeland Congress (eds D.Eldridge & D.Freudenberger), pp. 893896. Australian Rangeland Society, Queensland, Australia.
  • Westoby, M. & Wright, I. (2006) Land-plant ecology on the basis of functional traits. Trends in Ecology and Evolution, 21, 261268.
  • Wilson, P.J., Thompson, K. & Hodgson, J.G. (1999) Specific leaf area and leaf dry matter content as alternative predictors of plant strategies. New Phytologist, 143, 155162.
  • Wright, I.J., Reich, P.B., Cornelissen, J.H.C., Falster, D.S., Garnier, E., Hikosaka, K., Lamont, B.B., Lee, W., Oleksyn, J., Osada, N., Poorter, H., Villar, R., Warton, D.I. & Westoby, M. (2005) Assessing the generality of global leaf traits relationships. New Phytologist, 166, 485496.
  • Wright, I.J., Reich, P.B. & Westoby, M. (2001) Strategy shifts in leaf physiology, structure and nutrient content between species of high- and low-rainfall and high- and low-nutrient habitats. Functional Ecology, 15, 423434.
  • Wright, I.J., Reich, P.B., Westoby, M., Ackerly, D.D., Baruch, Z., Bongers, F., Cavender-Bares, J., Chapin, T., Cornelissen, J.H.C., Diemer, M., Flexas, J., Garnier, E., Groom, P.K., Gulias, J., Hikosaka, K., Lamont, B.B., Lee, T., Lee, W., Lusk, C., Midgley, J.J., Navas, M.-L., Niinemets, U., Oleksyn, J., Osada, N., Poorter, H., Poot, P., Prior, L., Pyankov, V.I., Roumet, C., Thomas, S.C., Tjoelker, M.G., Veneklaas, E.J. & Villar, R. (2004) The worldwide leaf economics spectrum. Nature, 428, 821827.

Supporting Information

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

Table S2. Mean annual above-ground net primary productivity (ANPP, g DM m-2 yr-1), crude protein content (CP, g kg-1 DM) and pepsin-cellulase dry matter digestibility (DMD, g kg-1 DM) per species. Results are means (? s.e.) of four treatments and of two years (n = 6). C-, bimonthly cuts; C+, monthly cuts; N-, low N supply (12 g N m-2 year-1); N+, high N supply (36 g N m-2 year-1).

Table S3. Analyses of covariance between leaf traits, digestible dry matter (DDMY) and crude protein (CPY) yield were performed with means per species and block (n = 39, block factor was not significant) within each factor level (? s.e.). Same abbreviations as in Table 1. C-, three cuts by year; C+, six cuts by year; N-, 12 g N g-2 year-1; N+, 36 g N m-2 year-1.

FilenameFormatSizeDescription
FEC1316_TableS1.doc75KSupporting info item
FEC1316_TableS2.doc78KSupporting info item
FEC1316_TableS3.doc56KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.