• grassland restoration;
  • plant diversity;
  • plant-soil feedback;
  • soil fertility;
  • soil invertebrates;
  • succession


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Plant community development depends to a great extent on the availability of soil nutrients, but recent studies underline the role of symbiotic, herbivorous and pathogenic soil biota. We tested for interactions between these biotic and abiotic factors by studying the effects of additional nutrients and the removal of soil biota on the replacement of grassland plant species typical of different successional stages.
  • 2
    Species representing each of the early, mid and target phases of secondary succession in a grassland community (four per phase) were grown in mid-successional grassland soil. The mixed plant communities were grown in sterilized and non-sterilized soil, at three nutrient supply levels. The distribution of shoot biomass over the different plant species, and the total root biomass, were determined, as well as the composition of nematode and microarthropod communities and soil decomposition rates.
  • 3
    The effect of nutrient supply on plant community composition depended on soil sterilization. In sterilized soil, the plant community was initially dominated by grasses that increased in dominance even without fertilization. In non-sterilized soil, the plant community was more diverse and grass dominance decreased over time, except with high fertilization. Fertilization enhanced the productivity of the plant community in sterilized and, to a greater extent, in non-sterilized soil.
  • 4
    The abundance of root-feeding nematodes was positively related to increased root biomass, pointing to a strong bottom-up control. Increased levels of nutrient supply were associated with reduced abundance of omnivorous nematodes, the cause of this reduction being less clear. Increased soil fertility altered the functional diversity of the soil nematode community, which might, in the longer term, also affect their feedback effects on the plant community.
  • 5
    Increased nutrient supply reduced soil decomposition activity in the non-sterilized soil, but not in the sterilized soil.
  • 6
    Our results imply that soil biota may reduce the effects of nutrient supply on plant dominance. Incorporating the effects of soil biota on plant species interactions into studies on succession, plant species diversity and restoration may therefore considerably increase our understanding of the observed plant community patterns.


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

Plant species dominance is, to a great extent, determined by abiotic factors, such as light and nutrient competition (Tilman 1982). Recent work, however, shows that soil biota affect plant biomass production, plant species diversity and plant succession. Soil biota can enhance plant diversity and plant succession by a specific reduction of dominant plant species (Van der Putten et al. 1993; De Deyn et al. 2003) or by promotion of subordinate plant species (Van der Heijden et al. 1998). The effects of these direct interactions between plants and soil biota on the plant community are, however, context dependent, because arbuscular mycorrhizal fungi can also promote dominant plant species (Hartnett & Wilson 1999) and soil pathogens may reduce rare plants more than dominants (Klironomos 2002), resulting in reduced plant species diversity.

Positive plant-soil microbe feedback has been suggested to play a central role in early successional communities on nutrient poor soils, while negative feedback is expected to contribute to species replacement and diversification in later successional communities at more fertile sites (Reynolds et al. 2003). In general, positive feedback involves mutualistic associations with symbionts or decomposers, while negative feedback involves pathogens, parasites or herbivores. It is suggested that the impact of plant-soil biota feedback on plant community dynamics interacts with soil nutrient availability (Van der Putten & Peters 1997; Wardle 2002; Reynolds et al. 2003). However, the consequences of such interactions for plant community development have hardly been studied experimentally (Bardgett & Wardle 2003).

In order to investigate the interaction between soil fertility and soil biota on plant community development we performed a microcosm experiment in which grassland plants representative of different stages of secondary succession were grown together in sterilized or non-sterilized soil of intermediate productivity, with three levels of nutrient supply. The development of the plant community was evaluated over 1 year. The shoot biomass distribution over the different plant species was assessed three times and the total root biomass was determined at the end of the experiment. The effect of nutrient supply on different trophic groups of soil fauna was studied via the response of the nematode and microarthropod communities of the non-sterilized soils in relation to nutrient supply level and root biomass, as well as via the soil decomposer activity. We hypothesized that increased soil fertility would stimulate plant dominance, but the effect would be less marked when a natural soil biota community was present. We expected that the fast growing plant species that dominate at high nutrient availability (i.e. early successional grasses) would be relatively most reduced by the presence of soil biota.

Materials and methods

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

experimental design

The effects of soil sterilization, nutrient supply and their interaction on plant community development were studied in a glasshouse microcosm experiment. Six treatment combinations were used, three nutrient supply levels (none, low, high) in sterilized and non-sterilized soil, with eight replicates of each. All replicates of every nutrient treatment were randomly distributed over three trolleys, but separate trolleys were used for sterilized and non-sterilized soil in order to minimize the risk of contamination. The trolleys were re-randomized every second week in order to avoid effects of microclimate differences within the glasshouse.

soil characteristics and soil sterilization

The soil was collected from the upper 10 cm of a perennial mid-secondary succession grassland, ‘de Born’ in Wageningen (51°59′ N, 5°40′ E). Until 1973, the site was used as production grassland grazed by cattle, but fertilization was then stopped and the site has been managed by mowing for hay collection twice every year. The productivity of the grassland was 6.3 ton dw ha−1 and 85.3% of the ground cover consisted of grasses. The soil was of a sandy loam type. The soil was sieved using a 0.5-cm mesh in order to remove roots and stones and kept in sealed polypropylene bags stored outdoors at ±18 °C. Half the soil was sterilized by γ-irradiation (25 kGy). Microcosms of (L × D × H) 17 × 17 × 22 cm3 were filled with 4420 g of either sterilized or non-sterilized soil, with the soil moisture set at 20% (w:w). In the non-sterilized soil the total P content was 680 mg kg−1, the total N content 2987 mg kg−1, and the percentage organic matter was 7.57%. The nutrient availability in the sterilized and non-sterilized soil at the start of the experiment is presented in Table 1.

Table 1.  Nutrient availability in the sterilized and non-sterilized soil from ‘de Born’ at start (mean ± 1 SE in mg kg−1, n = 8) and in the supplied nutrient solution. The P value indicates the significance of the difference between non-sterilized and sterilized soil
NutrientsNon-sterilized soilSterilized soilPLow (¼H 100 mL−1)High (½H per 100 mL−1)
NO3-N11.85 ± 0.06 0.02 ± 0.02< 0.00110.3320.67
NH4-N  1.9 ± 0.2 85.6 ± 1.2< 0.001  
(NO3 + NH4)-N 13.8 ± 0.2 85.6 ± 1.2< 0.001  
P 0.18 ± 0.01 0.74 ± 0.01< 0.001 0.34 0.69
K  3.2 ± 0.3 33.0 ± 0.4< 0.001 2.61 5.21
Mg 32.7 ± 0.4 56.8 ± 0.7< 0.001 0.54 1.08
Na  6.5 ± 0.2  7.0 ± 0.5= 0.83  
pH4.318 ± 0.0024.510 ± 0.002< 0.001  


The plant community in the microcosms was composed of a mixture of plant species typical of either recently abandoned production grassland (early successional), grassland that had been under restoration for 20 years (mid-successional) or species-rich natural grassland, the target community in biodiversity restoration and conservation projects on similar soil types. Each microcosm was planted with two individuals of Poa trivialis (L.), Lolium perenne (L.), Stellaria media (L.), Rumex obtusifolius (L.) (early secondary succession species), Agrostis capillaris (L.), Festuca rubra (L.), Plantago lanceolata (L.), Prunella vulgaris (L.) (mid species), Anthoxanthum odoratum (L.), Festuca ovina (L.) and Campanula rotundifolia (L.) (target species). In addition, due to poor germination of the fourth selected target species (Succisa pratensis (L.)), half of the replicates received two Centaurea jacea (L.) seedlings, while the other half received one Centaurea jacea (L.) and one Succisa pratensis (L.) seedling. The seeds had been obtained from the field and were provided by a small, specialized supplier (‘Cruydt-hoeck’, Groningen). Plant seeds were sown on glass beads, moistened with demineralized water and placed in a germination cabinet for 1 week at 16/8 L/D photo regime, 18/22 °C, and transplanted 1 week after germination in a grid of 6 × 4 positions. Each of the eight replicates had a different plant configuration in order to avoid positioning effects. One week after transplanting, dead seedlings were recorded and replaced. Plants that died after the first week were not replaced. There were minimally 16 hours of light (240 W m−2) and the L/D temperature regime was 21/16 °C. The plants were watered daily and the initial soil moisture level was re-set weekly to 20% by weighing.

nutrient treatment

Nutrient addition treatment was started 6 weeks after planting the seedlings. Every 2 weeks each experimental unit received 100 mL of water or either 1/4 (low) or 1/2 (high) strength Hoagland nutrient solution (Hewitt 1966). After 6 months the nutrient solution was added weekly. The amount of nutrients added to each 100 mL of solution is shown in Table 1.


Soil and plant nutrient analysis

Soil concentrations of NO3-N, NH4-N, P, K, Na and Mg, as well as the pH in non-sterilized and sterilized soil were assessed after CaCl2-extraction (Houba et al. 1994). Nutrient availability in the soil was checked twice (at the time of seedling planting and at the final harvest following shoot collection) on a soil subsample of four soil cores per microcosm. At the end of the experiment the N, P and K concentration in Plantago lanceolata and Anthoxanthum odoratum shoots was determined as for the soil nutrient availability, after drying and grinding the shoots.

Plant biomass

After 4 and 6 months shoots were clipped at 4 cm above the soil surface and dry weights of all individual plant species, as well as the total dead plant parts per pot, were determined after drying for 2 days at 75 °C. After 12 months the shoots were all clipped at soil surface level, sorted out to plant species, the dead parts were separated and the total root biomass was collected, dried and weighed.


In four replicates, six soil cores of 22 cm deep and 2.5 cm diameter were collected from each microcosm and nematodes were extracted from 100 cm3 of this soil by Oostenbrink elutriators (Oostenbrink 1960) and from the roots using a mist chamber. The total number of nematodes was counted in 10% of the total extracted soil volume. Nematodes were determined up to genus or family level according to Bongers (1988), using an inverted light microscope (10 × 20), allocated to functional groups according to Yeates et al. (1993), and distributed in C-P classes according to Bongers (1990). The most abundant nematode genera were Helicotylenchus (Steiner, 1945), Pratylenchus (Filipjev, 1936) and Paratylenchus (Micoletzky, 1922).


In the remaining four replicates, microarthropods were isolated from soil cylinders of 10 cm diameter and 5 cm deep, using Tullgren extraction funnels (Van Straalen & Rijninks 1982). Collembolans and mites were counted using a binocular and light microscope and the collembolans were identified to species level (10 × 5 and 10 × 40).

Cellulose decomposition

The decomposition of cellulose was assessed by determining the weight loss of a pre-weighed dry strip of cellulose filter paper incorporated in Petri dishes with soil collected from each microcosm and moistened to 55% water content. The sealed Petri dishes were incubated for 6 weeks at 20 °C in the dark and the remains of the cellulose strip were then removed, dried for 24 h at 75 °C, and weighed.

data analysis

The data were analysed with the statistical program Statistica version 10.0. All plant data were analysed by repeated measures anova (analysis of variance), or via a multivariate GLM (general linear model) with soil sterilization, nutrient supply level and their interaction as factors. The assumption of normality was checked with Kolmogorov Smirnov (Lilliefors) procedures and that of homogeneity of variances (homoscedasticity) with Levene. The statistical difference between groups was determined by Tukey's multiple range test, P < 0.05. Data were transformed where necessary to achieve normality, but if this was not possible, a non-parametrical test was conducted. The final root biomass and the shoot biomass at all three harvests were square-root transformed prior to analysis. Shoot biomass as a proportion of total plant biomass and forb biomass as a proportion of total shoot biomass were arcsine-square-root transformed.

Plant community diversity was calculated as Shannon-Wiener Evenness index J′ = −∑(pi × log2pi) × 1/log2S, where pi represents the proportional contribution of the ith species to the plant community biomass and S the number of species present. The effect of sterilization, nutrient supply and their interaction on the diversity index was statistically tested by a multivariate GLM. Treatment effects on the plant community were investigated by principal component ordination analysis of proportional biomasses in canoco version 4.5 and subsequent Kruskal–Wallis anova by ranks of the ordination score values on the primary axis. The effect of the treatment factors (nutrient supply and soil sterilization) on the plant species community composition was investigated by redundancy analysis (RDA), using the proportional biomasses. The significance of the canonical axes was investigated by partial Monte Carlo permutation tests, using nutrient supply as factor and sterilization as cofactor and vice versa. Treatment effects (nutrient supply, soil sterilization and their interaction) on the proportional biomass of individual plant species were analysed by a multivariate GLM.

The nematode abundances were log10 (x + 1) transformed in order to meet the assumptions for general linear model testing; original values were used when treatment effects were tested via a Mann–Whitney U non-parametric test. In order to check the effect of root biomass and nutrient supply level separately on the abundance of Helicotylenchus and of all plant-feeding nematodes, a GLM with type I sequential decomposition of the variance with root biomass as first and nutrient supply as second factor was used. The effect of soil sterilization and nutrient supply level on the nutrient availability in the soil and nutrient concentration in the plants was tested with Mann–Whitney U and Kruskal–Wallis non-parametric tests.


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

soil nutrient availability and nutrient concentration in plants

Soil pH, concentrations of phosphorus, potassium, magnesium and ammonium, and the total amount of available nitrogen (ammonium plus nitrate) were initially significantly higher in sterilized soil, but nitrate availability was higher in non-sterilized soil (Table 1). Soil nutrient availability had generally decreased by the end of the experiment, although concentrations of ammonium in unfertilized non-sterilized soil and nitrate availability in sterilized soil increased, and potassium levels in non-sterilized soil with nutrient supply remained constant. Soil pH decreased over time in all treatments. At the end of the experiment, concentrations of ammonium and total inorganic nitrogen were lower in sterilized than in non-sterilized soil, while phosphate availability showed the reverse pattern. Nutrient supply affected availability of all nutrients measured, except for nitrate and phosphorus. Ammonium and total nitrogen were lower with than without nutrient supply, while potassium, magnesium and sodium availability, as well as soil pH, were higher with than without nutrient supply (Table 2). Nutrient supply affected the nitrogen and potassium concentration in P. lanceolata and A. odoratum, as well as the phosphorus concentration in A. odoratum. Soil sterilization only affected the phosphorus concentration in P. lanceolata (Table 2). The N : P and N : K ratios were strongly affected by nutrient supply in both plant species (P < 0.001), while only in P. lanceolata soil sterilization significantly affected the N : P (P < 0.01) and N : K (P < 0.05) ratios. The N : P ratios in the unfertilized plants ranged from 27 ± 1 and 25 ± 2 in non-sterilized, to 12 ± 1 and 20 ± 3 in sterilized soil, for A. odoratum and P. lanceolata, respectively. The N : K ratios in unfertilized plant communities ranged from 22 ± 2 and 21 ± 1, in non-sterilized, to 11 ± 2 and 13 ± 1 in sterilized soil, for A. odoratum and P. lanceolata, respectively. Their N : K ratios decreased drastically with nutrient supply to 2.9 ± 0.1 and 4.0 ± 0.3, in non-sterilized, and 3.1 ± 0.1 and 5.9 ± 0.9 in sterilized soil (low supply level).

Table 2.  Soil nutrient availability and nutrient concentration in foliage after 12 months in the sterilized and non-sterilized soil with different nutrient supply levels (mean ± 1 SE in mg kg−1, n = 8 for soil data and n = 4 for plant data). The P value indicates the significance of the difference between non-sterilized and sterilized soil (S) and the nutrient supply levels (F)
 NutrientsNon-sterilized soilSterilized soilP value
SoilNO3-N  0.4 ± 0.2 0.9 ± 0.2 0.5 ± 0.1  2.1 ± 0.4  0.4 ± 0.2  0.6 ± 0.3   0.38   0.28
NH4-N 24.8 ± 5.7 1.3 + 0.4 0.4 ± 0.2  4.0 ± 2.1 0.06 ± 0.04 0.04 ± 0.03< 0.001< 0.001
(NO3 + NH4)-N 25.2 ± 5.7 2.2 + 0.6 0.9 ± 0.2  6.1 ± 2.3  0.5 ± 0.2  0.6 ± 0.3   0.024< 0.001
P 0.12 ± 0.020.07 ± 0.010.08 ± 0.01 0.16 ± 0.03 0.17 ± 0.04 0.16 ± 0.02< 0.001   0.41
K  1.4 ± 0.2 3.7 ± 0.6 7.2 ± 0.9  1.4 ± 0.4  3.3 ± 0.7  5.3 ± 0.8   0.24< 0.001
Mg 12.3 ± 0.615.6 ± 0.717.0 ± 1.0  9.8 ± 0.9 12.5 ± 1.3 18.7 ± 1.6   0.19< 0.001
Na  2.2 ± 0.3 2.9 ± 0.4 3.5 ± 0.4  1.9 ± 0.3  2.4 ± 0.4  4.1 ± 0.9   0.40< 0.01
pH 3.95 ± 0.034.10 ± 0.024.11 ± 0.02 4.04 ± 0.01 4.05 ± 0.01 4.06 ± 0.01   0.15< 0.01
P. lanceolataN2321 ± 127 809 ± 38 895 ± 501234 ± 1731051 ± 1071246 ± 179   0.15< 0.05
P  95 ± 4 108 ± 8  87 ± 8    66 ± 7    72 ± 10  99 ± 21< 0.01   0.31
K 110 ± 11 207 ± 15 320 ± 17    97 ± 12 197 ± 30 277 ± 49   0.10< 0.001
A. odoratumN2028 ± 124 692 ± 40 607 ± 301269 ± 188 646 ± 31 539 ± 19   0.43< 0.001
P  75 ± 4  88 ± 4  74 ± 5   102 ± 7  65 ± 4  52 ± 3   0.29< 0.01
K  98 ± 15 237 ± 12 297 ± 11   119 ± 11 211 ± 16 248 ± 21   0.24< 0.001

total shoot and root biomass

Total shoot and root biomass produced increased with increasing levels of nutrient supply both in the sterilized and in the non-sterilized soil. Total shoot and root biomass produced was also enhanced by soil sterilization, but there was no interaction between sterilization and nutrient supply (Table 3, Fig. 1). The proportional root biomass (i.e. as a percentage of shoot + root biomass) increased with nutrient supply, indicating a relatively higher investment in root than in shoot biomass, and was higher in sterilized than in non-sterilized soil (Table 3). Proportional root biomass was lowest for non-sterilized unfertilized soil (35 ± 6%), increasing to 46 ± 6% for the low level, and 63 ± 4% for the high level of fertilization; in sterilized soil values increased from 61 ± 6% (without nutrient supply), to 81 ± 1% and 82 ± 2% for the low and high nutrient supply level, respectively.

Table 3. F values derived from analysis of variance for measured response variables in response to soil sterilization (S) and nutrient supply (F) at three consecutive samplings (T1, T2, T3 = 4, 6, 12 months). Asterisks denote significant differences at *P < 0.05, **P < 0.01 and ***P < 0.001, and NS = non-significant. Degrees of freedom between parentheses, error 28, 42 or 84 for analysis per nutrient level, on final biomasses and repeated measures, respectively
Dependent variableS (1)F (2)S × F (2)R (time) (2)S × R (2)F × R (4)S × F × R (4)
Total shoot biomass (T1 + T2 + T3)250***108***1.06    
Total root biomass123*** 34***2.22    
Shoot/total biomass 53.17*** 14.57***1.22    
Shoot biomass (T1, T2, T3)344.08***161.04***9.56***206.89***75.84***19.31***0.68
J′ (Shannon-Wiener evenness)
% forbs (analysed per nutrient supply level) 41.14***  0.173.72*    
No nutrient supply 20.72***    5.34*22.68***  
Low supply level 79.64***    2.0742.46***  
High supply level 49.46***    4.01 8.42**  

Figure 1. Effects of soil sterilization and nutrient supply level on total shoot and root biomass (n = 8). Different letters denote significant differences between means at P < 0.05.

Download figure to PowerPoint

shoot biomass production pattern over time

At 4, 6 and 12 months, shoot biomass was enhanced by nutrient supply and by soil sterilization, and there was a significant interaction between the two factors (Table 3, Fig. 2).


Figure 2. Effects of soil sterilization and nutrient supply level on the shoot biomass over time (n = 8). Different letters denote significant differences between means at P < 0.05.

Download figure to PowerPoint

After 4 months shoot biomass in the sterilized soil was higher than in the non-sterilized soil, independent of nutrient supply level. By 6 months (i.e. 2 months after the first clipping), shoot biomass was greatest in the sterilized soil with the highest level of nutrient supply. However, at the lower level of fertilization shoot biomass was not significantly different from that in unfertilized plant communities in sterilized soil, or those growing in non-sterilized soil at the higher nutrient level. After 12 months shoot biomass remained significantly higher in sterilized than in non-sterilized soil only in unfertilized plant communities.

dominance of grasses over time affected by nutrient supply and soil sterilization

The proportional forb biomass was affected by soil sterilization at all nutrient supply levels and its pattern over time interacted significantly with the initial soil condition (sterilized or not) at all levels of nutrient supply (Table 3, Fig. 3).


Figure 3. Effects of soil sterilization and nutrient supply level on the proportion of forbs in the total biomass (n = 8, back transformed mean ± 1 SE).

Download figure to PowerPoint

In sterilized soil the proportional forb biomass decreased over time, to the benefit of grass proportional biomass, but in non-sterilized soil proportional forb biomass increased over time, except at the highest nutrient level.

plant community diversity and composition

The effect of nutrient supply on plant community diversity (Shannon-Wiener evenness J′) after 12 months depended on whether the soil was sterilized (Table 3). In the non-sterilized soil J’ declined from 0.71 ± 0.04 to 0.62 ± 0.05 from zero to high nutrient addition, but increased from 0.43 ± 0.03 to 0.54 ± 0.04 in sterilized soil. Kruskal–Wallis anova on the ranks of the ordination score values determined by principal component analysis (PCA) of the plant communities revealed significant treatment effects (P < 0.01). Treatments ranked as no added nutrients in sterilized soil < no addition in non-sterilized soil < low additions in non-sterilized soil < high additions in sterilized soil < low in sterilized-soil < high in non-sterilized soil. Thus the order in which the nutrient supply levels ranked differed between sterilized and non-sterilized soil. Treatment effects on the plant community and individual plant species were determined by redundancy analysis (RDA, Fig. 4). Both canonical axes, representing sterilization and nutrient supply, were significant (partial Monte Carlo permutation tests, F = 17.06 for sterilization and F = 8.90 for nutrient supply, P < 0.01 for each factor). The first canonical axis separates the communities in sterilized vs. those in non-sterilized soil, while the second canonical axis separates the nutrient supply treatments. The forbs C. jacea, C. rotundifolia, S. pratensis (species of target communities), P. vulgaris and P. lanceolata (mid), performed better in non-sterilized soil. In contrast, the grasses L. perenne, P. trivialis (early) and A. capillaris (mid) performed better in sterilized soil. At 4 months similar results were found for C. rotundifolia and for the grasses, but not for the other forb species (Table 4). The positioning along the second axis shows that the grass species L. perenne and A. odoratum performed better with increased nutrient supply, while F. rubra did better without nutrient supply, in line with its response after 4 months. A. capillaris (mid) and C. rotundifolia (target) performed better without, while P. trivialis (early) and P. lanceolata did better with nutrient supply, but these responses appeared to interact with soil sterilization (interaction sterilization × nutrient supply in GLM analysis, P < 0.05). Despite significant treatment effects after 4 months (Table 4), after 12 months F. ovina and R. obtusifolius showed no significant response to either sterilization or nutrient supply, while S. media had disappeared from all treatments.


Figure 4. Plant community and plant species response after 12 months to nutrient supply level (no, low, high) and soil sterilization (s = sterilized, ns = non-sterilized). Significance of factors on plant community tested by partial Monte Carlo permutation test, per plant species by multivariate GLM (letters between brackets denote significant effects of S = sterilization, F = nutrient supply and S × F = interaction; n = 48 and P < 0.05).

Download figure to PowerPoint

Table 4.  Proportional plant species composition after 4 months, in response to nutrient supply (F) and soil sterilization (S) (mean proportion of total shoot biomass ± 1 SE, n = 8). Asterisks denote significant differences at *P < 0.05, **P < 0.01 and ***P < 0.001, and NS = non-significant
 Nutrient supplyNon-sterilized soilSterilized soilEffects
NoneLowHighNoneLowHighSFS × F
ForbsStellaria media 1.0 ± 0.4 3.8 ± 0.9 7.8 ± 1.8 1.0 ± 0.2 1.0 ± 0.5 0.3 ± 0.1*********
Rumex obtusifolius0.06 ± 0.050.19 ± 0.111.24 ± 0.560.29 ± 0.040.89 ± 0.230.50 ± 0.16NS**
Prunella vulgaris1.16 ± 1.160.01 ± 0.010.38 ± 0.200.22 ± 0.100.23 ± 0.050.31 ± 0.10NSNSNS
Plantago lanceolata 6.6 ± 2.414.7 ± 3.910.9 ± 3.4 8.6 ± 1.311.7 ± 2.1813.1 ± 1.7NSNSNS
Campanula rotundifolia0.46 ± 0.140.74 ± 0.181.39 ± 0.410.09 ± 0.040.11 ± 0.030.30 ± 0.12****NS
Centaurea jacea 1.6 ± 0.6 1.6 ± 0.4 1.6 ± 0.9 1.4 ± 0.4 1.0 ± 0.3 0.9 ± 0.3NSNSNS
Succisa pratensis000.13 ± 0.0800.01 ± 0.010NSNSNS
GrassesLolium perenne 7.9 ± 2.911.4 ± 2.5 9.5 ± 3.117.8 ± 2.217.1 ± 2.816.1 ± 1.7***NSNS
Poa trivialis 1.9 ± 0.8 3.5 ± 1.0 4.6 ± 1.3 8.1 ± 1.512.8 ± 2.7 9.8 ± 1.8***NSNS
Festuca rubra 5.4 ± 0.6 2.1 ± 1.0 2.3 ± 0.6 4.1 ± 0.6 3.9 ± 0.6 2.0 ± 0.4NS**NS
Agrostis capillaris13.5 ± 2.514.9 ± 1.718.7 ± 4.335.0 ± 2.330.4 ± 3.925.4 ± 3.0***NSNS
Festuca ovina 1.6 ± 0.7 4.0 ± 0.6 5.9 ± 1.2 4.0 ± 0.6 3.0 ± 0.4 3.2 ± 0.7NSNS**
Anthoxanthum odoratum58.9 ± 4.043.2 ± 6.735.5 ± 6.419.5 ± 3.018.4 ± 3.028.1 ± 2.5***NS**

soil biota response to soil nutrient supply

Nematode abundance

At the end of the experiment, the sterilized soil contained nematodes, but all were bacterial feeders. In the non-sterilized soil, plant-feeding and plant-associated nematodes were the most abundant feeding groups. The major plant-feeding taxa were Helicotylenchus, Pratylenchus and Paratylenchus. Their averages per nutrient supply treatment ranged from 162 to 940 individuals per 100 g of soil for Helicotylenchus, 242–437 for Pratylenchus and 49–115 for Paratylenchus. The abundance of Helicotylenchus significantly increased in the fertilized communities (F2,9 = 5.78, P < 0.05). At the highest nutrient supply level 940 ± 382 individuals 100 g of soil−1 were found, compared with 162 ± 57 in the unfertilized soil. Sequential decomposition of the variance, with root biomass and nutrient supply as explaining variables, showed that the number of Helicotylenchus was only significantly affected by root biomass (root biomass, F1,8 = 10.46, P < 0.05; nutrient supply, F2,8 = 0.98, P = 0.41). The strong positive correlation between the most dominant plant-feeding nematode taxon and root biomass was also reflected in the response of the total plant-feeding nematode abundance, pointing to a significant bottom-up response (root mass, F1,8 = 5.75, P < 0.05; nutrient supply, F2,8 = 0.29, P = 0.76; Pearson correlation coefficient r = 0.66, P < 0.05). The fungal-feeding nematode abundance also correlated positively with the root biomass, but root biomass had no significant effect in the GLM analysis when nutrient supply was also included as a factor (Pearson correlation coefficient r = 0.66, P < 0.05; root mass, F1,8 = 2.20, P = 0.18; nutrient supply, F2,8 = 0.14, P = 0.87). Although the abundances of bacterial and fungal feeders did not differ significantly between nutrient supply levels, the proportion of fungal feeders was significantly higher at higher levels of nutrient supply (P < 0.05). Nutrient supply reduced the numbers of omnivorous nematodes (F2,9 = 15.42, P < 0.01; Fig. 5a), as well as the nematode maturity index, a measure of soil disturbance (F2,9 = 7.78, P < 0.05; Fig. 5b).


Figure 5. Effects of nutrient supply level on (a) the abundance of omnivorous nematodes, and (b) the nematode community maturity index in non-sterilized soil (n = 4). Different letters denote significant differences between means at P < 0.05.

Download figure to PowerPoint

Micro-arthropod abundance

In the non-sterilized soil, the abundances of mites and collembolans were not significantly affected by nutrient supply (F2,9 = 0.41, P = 0.67 for the mites and F2,9 = 0.95, P = 0.42 for the collembolans). Mite abundance ranged from 111 ± 29 per 100 g dry soil with high nutrient addition to 172 ± 49 with low addition and collembolan abundance from 22 ± 13 per 100 g dry soil in the control to 52 ± 19 with the highest nutrient supply. For Collembola, the number of taxa was marginally significantly increased by adding nutrients, from 2.3 ± 0.6 to 2.5 ± 0.3 and 4.0 ± 0.4 (F2,9 = 4.16, P = 0.052).

Decomposition of cellulose

In non-sterilized soil, the decomposition of cellulose decreased at the higher level of nutrient addition (F2,21 = 10.10, P < 0.001). Decomposition rate was faster in sterilized than in non-sterilized soil (F1,46 = 19.71, P < 0.001), but was no longer affected by nutrient supply (F2,21 = 0.12, P = 0.89; Fig. 6).


Figure 6. Decomposition of cellulose in relation to soil sterilization and soil nutrient supply level (n = 8). Different letters denote significant differences between means at P < 0.05.

Download figure to PowerPoint


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

As soil sterilization changes both (micro)biological and chemical soil properties, the differential response of the synthesized plant communities to the nutrient supply treatments in sterilized and non-sterilized soil results from the net effect of the absence of root pathogens, root herbivores and root symbionts (Bever et al. 1997), as well as from the initial nutrient flush and the temporary absence of nutrient immobilization into soil microbial biomass in the sterilized soil (Troelstra et al. 2001). The original soil was nutrient poor and, assuming the critical N : P ratio to be 16, A. odoratum and P. lanceolata appeared to be P limited (Koerselman & Meuleman 1996). However, according to the triaxial diagram of Venterink (2000), the percentages of N/(N + 10P + K), K/(N + 10P + K) and 10P/N + 10P + K in the foliage of unfertilized A. odoratum and P. lanceolata plants pointed to limitation by K or by both K and N.

As expected, total plant biomass increased with increased nutrient supply. Root biomass was relatively more affected than shoot biomass, suggesting significant below-ground plant community changes. The shoot biomass distribution over the different plant species depended on nutrient supply and on soil sterilization. Soil sterilization favoured fast growing mid- and early successional plant species (L. perenne, P. trivialis and A. capillaris), whilst in non-sterilized soil mid- and late successional forbs (P. lanceolata, C. rotundifolia, C. jacea and S. pratensis) established well, resulting in higher plant community diversity. A low level of nutrient addition to non-sterilized soil increased the biomass of all plant species proportionally and, in contrast to Rajaniemi et al. (2003), plant community diversity was not reduced. Higher nutrient supply, however, favoured A. odoratum disproportionately, increasing its dominance and consequently decreasing plant community diversity. Nutrient supply had the opposite effect in sterilized soil, where there was a trend of increasing Shannon-Weiner evenness with increasing nutrients.

Differences between grasses and later successional forbs in growth rate and in tolerance to herbivory might explain our results. Slow growing plants of nutrient-poor (or late secondary successional) habitats often have longer lived and better defended leaves and roots than faster growing plants from more fertile, or earlier successional, habitats (Grime 1977; Bazzaz 1979; Chapin 1980; Coley 1988; Herms & Mattson 1992; Van der Krift & Berendse 2002). Generalist herbivores tend to prefer nitrogen-rich, poorly defended plant species (Mattson 1980; White 1984; Buckland & Grime 2000), while nutrient supply can enhance the tolerance of plants to and regrowth after herbivory (Steinger & Müller-Schärer 1992). The observed increase in evenness in communities in sterilized soil as nutrient supply increased could be explained by the dominant species needing a smaller proportion of the available nutrients and therefore more being left for the subordinate plants.

Total shoot and root biomass produced in the non-sterilized soil with the highest nutrient supply equalled that in sterilized soil without added nutrients. Although, according to theory, plant diversity is expected to be similar at similar productivity levels (Al-Mufti et al. 1977; Grime 1979; Marrs 1993), plant diversity (Shannon-Weiner evenness) at the end of the experiment differed between these two treatments. Patterns of shoot biomass production during the experiment also differed, suggesting that the production process, over time, may be as important as single observations.

Several plant species showed significantly different responses to nutrient supply in non-sterilized and sterilized soil. Forbs only benefited from (low) nutrient addition in non-sterilized soil, while for grasses the outcome of the interaction between soil sterilization and nutrient supply differed between plant species. The proportional biomass of P. trivialis increased with nutrient supply in sterilized soil, while A. capillaris increased proportionally with nutrient supply in non-sterilized but decreased with nutrient supply in sterilized soil. The decrease in sterilized soil was probably due to interspecific competition with A. odoratum, a plant species that also benefited strongly from increased nutrient supply in the non-sterilized soil. This might be due to its superficial rooting system (Berendse 1983), in combination with nutrient supply at the soil surface. However, as deep-rooting plants (P. lanceolata) were also favoured by nutrient supply in the non-sterilized soil, a higher herbivore tolerance in A. odoratum than in A. capillaris is not unlikely. The observations that P. lanceolata is a poor host for plant-parasitic nematodes (Wardle et al. 2003; De Deyn et al. 2004), while A. capillaris appears to be sensitive to root herbivory (De Deyn et al. 2003) also support the possible involvement of differences in tolerance to herbivory between A. capillaris and A. odoratum.

In diverse plant communities shoot herbivores reduce plants with high relative growth rate more at increased soil fertility, indirectly enhancing slower growing, stress-tolerant plant species (Fraser & Grime 1999) and promoting plant species diversity (Proulx & Mazumder 1998). In our experiment, fast growing early secondary successional plant species were most reduced by soil biota in unfertilized soil and least affected at high nutrient level, probably due to compensatory growth. In natural communities, above-ground herbivory should be more important in suppressing dominant plant species at high soil fertility, while the relative importance of soil biota in suppressing dominant plant species will increase with decreasing soil fertility. Verschoor et al. (2002a) did not find an interaction between root herbivory and nutrient availability on shoot biomass production in Holcus lanatusA. odoratum mixtures, although grasses and forbs in species mixtures might respond very differently. Regrowth depends on resource availability such that at high resource availability herbivory enhances regrowth of grasses while forbs are favoured at lower nutrient levels (Hawkes & Sullivan 2001).

In contrast to Verschoor et al. (2002b), plant-feeding nematode abundance increased in response to nutrient supply and increased root biomass. However, not only did root quantity increase but also it is likely that root quality, in terms of nitrogen content, and species composition did as well. The plant-feeding nematodes recovered are all known to have a wide host range, consuming roots of most, if not all, of the plant species present. Nematode preference for certain plant species and differences in plant tolerance to nematode feeding are, however, likely (Yeates 1987; Trudgill 1991; De Deyn et al. 2004). The decline of omnivorous nematode abundance with increased nutrient supply could be due to increased competition with specialist plant-feeding nematodes, but their higher sensitivity to nutrient addition, in combination with their slow recovery rate, seems a more likely explanation (Bongers 1990; Bongers & Bongers 1998).

The activity of the decomposer soil biota results in nutrient cycling via decomposition. In contrast to our expectation, cellulose decomposed faster in sterilized than in non-sterilized soil. Air-borne microorganisms such as saprophytic fungi and bacteria may have re-colonized the sterilized soil and could easily have spread and reproduced due to reduced fungistasis and the absence of fungivores (De Boer et al. 2003). The reduced decomposition rate in non-sterilized soil at higher nutrient supply might be due to reduced nitrogen availability. Decomposers of cellulose require nitrogen (McClaugherty et al. 1985), the availability of which was lower in the nutrient-supplied plant communities. Moreover, in the non-sterilized soil the fungivorous nematodes and microarthropods were more abundant at higher nutrient levels, which might have caused top-down control of the fungi (Wardle 2002), hence reducing their ability to decompose cellulose.

In conclusion, our results show that the effect of increased nutrient availability on plant community development and diversity depends on the interaction with soil biota. In the absence of root-feeding invertebrates and high initial nutrient availability, fast growing grass species increase in dominance, independent of subsequent nutrient levels, while other plant species are suppressed. Lower initial nutrient availability and presence of soil biota reduced the growth rate of fast growing plant species, resulting in higher plant community diversity. Grass dominance declined over time, except at high nutrient availability, when increased growth was probably able to compensate for the negative effects of soil biota.

Nutrient supply increased root production, which correlated well with root-feeding nematode density. When nutrient supply is ceased a negative feedback to the most nutrient-responsive plant species can be expected, because nutrients for compensatory growth will no longer be available. Our data suggest that the effect of higher nutrient availability on the interspecific competition of successional plants is not only determined by their nutrient acquisition and growth rates but also by the degree to which they benefit from the interaction with soil biota.


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

We thank Wiecher Smant for practical assistance, Peter C. de Ruiter and Herman A. Verhoef for critical reading of a previous version of the manuscript and the Dutch NWO-ALW stimulation programme Biodiversity for funding. Publication 3379N100-KNAW Netherlands Institute of Ecology.


  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Al-Mufti, M.M., Sydes, C.L., Furness, S.B., Grime, J.P. & Band, S.R. (1977) A quantitative analysis of shoot phenology and dominance in herbaceous vegetation. Journal of Ecology, 65, 759791.
  • Bardgett, R.D. & Wardle, D.A. (2003) Herbivore-mediated linkages between aboveground and belowground communities. Ecology, 84, 22582268.
  • Bazzaz, F.A. (1979) The physiological ecology of plant succession. Annual Review of Ecology and Systematics, 10, 351371.
  • Berendse, F. (1983) Interspecific competition and niche differentiation between Plantago lanceolata and Anthoxanthum odoratum in a natural hayfield. Journal of Ecology, 71, 379390.
  • Bever, J.D., Westover, K.M. & Antonovics, J. (1997) Incorporating the soil community into plant population dynamics: the utility of the feedback approach. Journal of Ecology, 85, 561573.
  • Bongers, T. (1988) De Nematoden van Nederland. KNNV, Utrecht.
  • Bongers, T. (1990) The maturity index: an ecological measure of environmental disturbance based on nematode species composition. Oecologia, 83, 1419.
  • Bongers, T. & Bongers, M. (1998) Functional diversity of nematodes. Applied Soil Ecology, 10, 239251.
  • Buckland, S.M. & Grime, J.P. (2000) The effects of trophic structure and soil fertility on the assembly of plant communities: a microcosm experiment. Oikos, 91, 336352.
  • Chapin, F.S. (1980) The mineral nutrition in wild plants. Annual Review of Ecology and Systematics, 11, 233260.
  • Coley, P.D. (1988) Effects of plant growth rate and leaf lifetime on the amount and type of antiherbivore defense. Oecologia, 74, 531536.
  • De Boer, W., Verheggen, P., Klein Gunnewiek, P.J.A., Kowalchuk, G.A. & Van Veen, J.A. (2003) Microbial community composition affects soil fungistasis. Applied and Environmental Microbiology, 69, 835844.
  • De Deyn, G.B., Raaijmakers, C.E., Van Ruijven, J., Berendse, F. & Van Der Putten, W.H., (2004) Plant species identity and diversity effects on different trophic levels of nematodes in the soil food web. Oikos, 106, 576586.
  • De Deyn, G.B., Raaijmakers, C.E., Zoomer, H.R., Berg, M.P., De Ruiter, P.C., Verhoef, H.A. et al. (2003) Soil invertebrate fauna enhances grassland succession and diversity. Nature, 422, 711713.
  • Fraser, L.H. & Grime, J.P. (1999) Interacting effects of herbivory and fertility on a synthesized plant community. Journal of Ecology, 87, 514525.
  • Grime, J.P. (1977) Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. American Naturalist, 111, 11691194.
  • Grime, J.P. (1979) Plant Strategies and Vegetation Processes. Wiley, Chichester.
  • Hartnett, D.C. & Wilson, G.W.T. (1999) Mycorrhizae influence plant community structure and diversity in tallgrass prairie. Ecology, 80, 11871195.
  • Hawkes, C.V. & Sullivan, J.J. (2001) The impact of herbivory on plants in different resource conditions: a meta-analysis. Ecology, 82, 20452058.
  • Herms, D.A. & Mattson, W.J. (1992) The dilemma of plants: to grow or to defend. Quarterly Review of Biology, 67, 283335.
  • Hewitt, E.J. (1966) Sand and Water Culture Methods Used in the Study of Plant Nutrition. CMI Bureau, Bucks.
  • Houba, V.J.G., Novozamsky, I. & Temminghoff, E. (1994) Soil analysis procedures − extraction with 0.01 M CaCl2. Soil Science and Plant Nutrition (ed. V.J.G.Houba), pp. 943 . Wageningen University, Wageningen.
  • Klironomos, J.N. (2002) Feedback with soil biota contributes to plant rarity and invasiveness in communities. Nature, 417, 6770.
  • 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.
  • Marrs, R.H. (1993) Soil fertility and nature conservation in Europe: theoretical considerations and practical management solutions. Advances in Ecological Research, 24, 241300.
  • Mattson, W.J. (1980) Herbivory in relation to plant nitrogen content. Annual Review of Ecology and Systematics, 11, 119161.
  • McClaugherty, C.A., Pastor, J. & Aber, J.D. (1985) Forest litter decomposition in relation to soil nitrogen dynamics and litter quality. Ecology, 66, 266275.
  • Oostenbrink, M. (1960) Estimating nematode populations by some selected methods. Nematology (eds J.N.Sasser & W.R.Jenkins), pp. 85102. University of North Carolina Press, Chapel Hill.
  • Proulx, M. & Mazumder, A. (1998) Reversal of grazing impact on plant species richness in nutrient-poor vs. nutrient-rich ecosystems. Ecology, 79, 25812592.
  • Rajaniemi, T.K., Allison, V. & Goldberg, D.E. (2003) Root competition can cause a decline in diversity with increased productivity. Journal of Ecology, 91, 407416.
  • Reynolds, H.L., Packer, A., Bever, J.D. & Clay, K. (2003) Grassroots ecology: plant–microbe–soil interactions as drivers of plant community structure and dynamics. Ecology, 84, 22812291.
  • Steinger, T. & Müller-Schärer, H. (1992) Physiological and growth-responses of Centaurea maculosa (Asteraceae) to root herbivory under varying levels of interspecific plant competition and soil-nitrogen availability. Oecologia, 91, 141149.
  • Tilman, D. (1982) Resource Competition and Community Structure. Princeton University Press, Princeton.
  • Troelstra, S.R., Wagenaar, R., Smant, W. & Peters, B.A.M. (2001) Interpretation of bioassays in the study of interactions between soil organisms and plants: involvement of nutrient factors. New Phytologist, 150, 697706.
  • Trudgill, D.L. (1991) Resistance to and tolerance of plant parasitic nematodes in plants. Annual Review of Phytopathology, 29, 167192.
  • Van der Heijden, M.G.A., Klironomos, J.N., Ursic, M., Moutoglis, P., Streitwolf-Engel, R., Boller, T. et al. (1998) Mycorrhizal fungal diversity determines plant biodiversity, ecosystem variability and productivity. Nature, 396, 6972.
  • Van der Krift, T.A.J. & Berendse, F. (2002) Root life spans of four grass species from habitats differing in nutrient availability. Functional Ecology, 16, 198203.
  • Van der Putten, W.H. & Peters, B.A.M. (1997) How soil-borne pathogens may affect plant competition. Ecology, 78, 17851795.
  • Van der Putten, W.H., Van Dijk, C. & Peters, B.A.M. (1993) Plant specific soil-born diseases contribute to succession in foredune vegetation. Nature, 362, 5356.
  • Van Straalen, N.M. & Rijninks, P.C. (1982) The efficiency of Tullgren apparatus with respect to interpreting seasonal changes in age structure of soil arthropod populations. Pedobiologia, 24, 197209.
  • Venterink, H.O. (2000) Nitrogen, phosphorus and potassium flows controlling plant productivity and species richness. Dissertation, Utrecht University, Utrecht, The Netherlands.
  • Verschoor, B.C., De Goede, R.G.M. & Brussaard, L. (2002b) Do plant parasitic nematodes have differential effects on the productivity of a fast- and slow-growing grass species? Plant and Soil, 243, 8190.
  • Verschoor, B.C., Pronk, T.E., De Goede, R.G.M. & Brussaard, L. (2002a) Could plant-feeding nematodes affect the competition between grass species during succession in grasslands under restoration management? Journal of Ecology, 90, 753761.
  • Wardle, D.A. (2002) Belowground consequences of aboveground food web interactions. Communities and Ecosystems (eds S.A.Levin & H.S.Horn), pp. 105137. Princeton University Press, Princeton, New Jersey.
  • Wardle, D.A., Yeates, G.W., Williamson, W. & Bonner, K.I. (2003) The response of a three trophic level soil food web to the identity and diversity of plant species and functional groups. Oikos, 102, 4556.
  • White, T.C.R. (1984) The abundance of invertebrate herbivores in relation to availability of nitrogen in stressed food plants. Oecologia, 63, 90105.
  • Yeates, G.W. (1987) How plants affect nematodes. Advances in Ecological Research, 17, 63113.
  • Yeates, G.W., Bongers, T., De Goede, R.G.M., Freckman, D.W. & Georgieva, S.S. (1993) Feeding-habits in soil nematode families and genera – an outline for soil ecologists. Journal of Nematology, 25, 315331.