SEARCH

SEARCH BY CITATION

Keywords:

  • assembly rules;
  • functional groups;
  • legumes;
  • resistance to invasion;
  • seed-addition experiment

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Although experimental studies usually reveal that resistance to invasion increases with species diversity, observational studies sometimes show the opposite trend. The higher resistance of diverse plots to invasion may be partly due to the increased probability of a plot containing a species with similar resource requirements to the invader.
  • 2
    We conducted a study of the invasibility of monocultures belonging to three different functional groups by seven sown species of legume. By only using experimentally established monocultures, rather than manipulating the abundance of particular functional groups, we removed both species diversity and differences in underlying abiotic conditions as potentially confounding variables.
  • 3
    We found that legume monocultures were more resistant than monocultures of grasses or non-leguminous forbs to invasion by sown legumes but not to invasion by other unsown species. The functional group effect remained after controlling for differences in total biomass and the average height of the above-ground biomass.
  • 4
    The relative success of legume species and types also varied with monoculture characteristics. The proportional biomass of climbing legumes increased strongly with biomass height in non-leguminous forb monocultures, while it declined with biomass height in grass monocultures. Trifolium pratense was the most successful invader in grass monocultures, while Vicia cracca was the most successful in non-leguminous forb monocultures.
  • 5
    Our results suggest that non-random assembly rules operate in grassland communities both between and within functional groups. Legume invaders found it much more difficult to invade legume plots, while grass and non-leguminous forb plots favoured non-climbing and climbing legumes, respectively. If plots mimic monospecific patches, the effect of these assembly rules in diverse communities might depend upon the patch structure of diverse communities. This dependency on patch structure may contribute to differences in results of research from experimental vs. natural communities.

Introduction

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

Competition from established vegetation normally leads to some degree of invasion resistance (Elton 1958; Turnbull et al. 2000; Levine et al. 2004). For example, experimental enhancement of resident species diversity (Knops et al. 1999; Joshi et al. 2000; Hector et al. 2001), the presence of particular resident species (Crawley et al. 1999; van Ruijven & Berendse 2003; Meiners et al. 2004), or a combination of the two (van Ruijven & Berendse 2003), can reduce overall invasion success. Many of these experimental studies have considered invasion by non-resident but non-exotic species as a way to deepen our understanding of the rules underlying community structure. Understanding these rules is an important first step in understanding why some communities are particularly susceptible to invasion by exotics (Leishman & Thomson 2005; Von Holle 2005) and why some exotic species are particularly successful invaders (Crawley et al. 1996; Grotkopp et al. 2002). The increased resistance of diverse communities or communities containing a strong competitor is likely to occur because residents deny invaders access to limiting resources (Tilman 2004). For example, a set of complementary species might achieve more complete resource depletion above or below ground (Tilman 1999; Hector et al. 2005; Spehn et al. 2005), or the presence of a particularly good competitor might lead to unusually low resource availability in communities containing that species (Wedin & Tilman 1993). These results imply that assembly rules are resource-based. If so, species are more likely to invade if they are able to use resources that the current residents cannot access and the match between the invader and the residents should be a key determinant of invasion success (Tilman 2004).

There has been rather little experimental work linking invasion success to the resource-use characteristics of both resident communities and invading plants. However, several recent studies have examined whether invasion is easier for species belonging to functional groups or guilds that are either absent or rare in the resident community (Symstad 2000; Fargione et al. 2003; Von Holle & Simberloff 2004; Xu et al. 2004). This should occur if species belonging to different functional groups have reduced resource-use overlap. The results of such experiments have been mixed. Using artificially assembled communities, Fargione et al. (2003) showed that each resident functional group was most effective at inhibiting invaders from the same functional group, although all invaders found plots containing C4 grasses the most difficult to invade. In contrast, Von Holle & Simberloff (2004) manipulated functional group diversity in natural communities. Removal or reduction of a particular functional group in this case did not make it easier for species from the removed group to invade.

Invasion experiments can also help to reveal competitive dynamics within guilds. For example, if strong competitive hierarchies exist within guilds (Tilman 1994; Turnbull et al. 2004) then the likelihood of successful invasion by one species into a community that already contains a second from that guild would depend on their relative positions within this hierarchy: a poor competitor from a particular functional group may be an unsuccessful invader while a good competitor may invade easily (Wedin & Tilman 1993). However, if species within functional groups are competitively equivalent, with dynamics governed by drift (Hubbell 2001), all species from a particular functional group would be equally capable of invading a community containing any other species from the same functional group given equivalent propagule pressure. Finally, competitive outcomes within guilds might be determined by environmental conditions (Tilman 1982) and the relative success of particular invaders would therefore depend on the conditions prevalent in the resident community, for example on the availability of soil nitrogen (McKane et al. 2002) or water (Silvertown et al. 1999).

Conditions in the resident community are partly determined by the resident plants; for example, the canopies of different tree species transmit different amounts of light (Canham et al. 1994). This in turn helps to maintain diversity by allowing the persistence of both shade-tolerating and pioneer species that, respectively, specialize on low and high light patches (Pacala et al. 1996). Monospecific or monodominant stands within resident communities could therefore potentially generate a mosaic of patches favouring the establishment of different species. Establishing experimental monocultures on uniform soil is one way of ensuring that the initial abiotic conditions are the same and that differences in conditions have been created by the species themselves. If the relative success of invading species changes with monoculture characteristics we can conclude that biotically generated heterogeneity can contribute to the creation and maintenance of diversity.

Here we focus on species from one particular functional group, legumes, and determine their ability to invade monocultures belonging to a range of species and functional groups. We test whether legumes are less able to invade legume monocultures than those belonging to two other functional groups (grasses and non-leguminous forbs) as predicted if resource-use patterns are more disparate between than within functional groups. We used the same seven native European legume species as invaders and as monocultures to obtain a fully reciprocal design for the within-functional group invasion test. By adding seeds of the seven invaders mixed together we ensured there would be competition between invaders. We then examined whether the relative success of different legume species and types (climbers and non-climbers) changed across monocultures with different characteristics, as predicted if resident species offer a range of opportunities to invaders. Finally, we related the effectiveness of each legume species at invading other legumes to its ability to resist invasion itself. Such a correlation is expected if there are strong dominance hierarchies within the legume guild.

Methods

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

the established monocultures

Monocultures of 36 common species from European grassland were established from seed in May 2002 as part of a biodiversity experiment. The design was not fully randomized; instead species were divided into groups of six and monocultures (50 × 50 cm) of each species in the group were always grown together in a spatial unit (called here a sub-block). Each sub-block was then replicated three times within a blocked design, although the monocultures of only 27 species were used in the experiment described here (81 plots in total). During the first growing season plots received nitrogen fertiliser for the purposes of the biodiversity experiment but these treatments were discontinued during 2003 (the year of the seed-addition experiment described here). All plots were weeded during 2002 and once prior to addition of legume seeds in April 2003, but not thereafter. Plots were cut to 5 cm twice during 2002 and in June 2003 to mimic the typical grassland management regime.

the seed-addition experiment

We sowed a mixture containing all seven legume species into all 81 monoculture plots. Each selected monoculture plot was divided into two halves, and one half randomly assigned to sowing while the other half remained as an unsown control. Seed mixtures contained 100 seeds of each legume species (700 seeds in total; a density of 5600 seeds m−2). Seeds were obtained from commercial suppliers of wild flower seed. The seven legume species can be classified as either climbers (Lathyrus pratensis L., Vicia cracca L. and Vicia sepium L.), which have modified leaflets or tendrils allowing them to use other species for physical support, or non-climbers (Lotus corniculatus L., Medicago lupulina L., Trifolium pratense L. and Trifolium repens L.), which lack this adaptation. In addition, climbers have much heavier seeds than non-climbers (13.7 ± 3.13 vs. 1.50 ± 0.33 g).

In September 2003, prior to the destruction of the plots, the central 20 × 20 cm area of both sown and unsown halves was cut to ground level and sorted into the seven sown legume species, other unsown invaders and the original monoculture species. The unsown invaders were not sorted further. Samples were then oven-dried for 48 hours at 80 °C and weighed. None of the invading legumes were flowering although plants were generally robust and well beyond the seedling stage. Because of the clonal nature of their growth, data on individual plant biomass are not available. Negligible amounts of sown legume species were recovered from the unsown subplots and analysis was therefore conducted on the uncorrected legume biomass from the sown subplots. The invading biomass of a legume species invading its own monoculture had to be assumed to be zero. This might lead to a lower total invading legume biomass in legume monocultures simply because there is always a species missing. To assess the magnitude of this effect, we deleted the biomass of one legume species from each grass and non-leguminous forb monoculture plot (selecting a new species at random from each plot) and then re-fitted the model. We repeated this procedure 50 times and assessed the significance of the functional group effect each time.

We calculated two covariates that might explain differences in invasibility of plots with different functional groups and species: the average total biomass of each monoculture species and a measure of the height of the above-ground biomass (here referred to as the average biomass height). Average total biomass was estimated from the three replicate unsown subplots. Our measure of average biomass height combined information on both canopy height and plant architecture and was obtained from the previous year's layered harvest. For this harvest the plots were cut into six vertical layers, 0–5 cm, 5–15 cm, 15–25 cm, 25–35 cm, 35–45 cm and > 45 cm. The measure is then calculated (following Spehn et al. 2000) using:

  • image(eqn 1 )

where mi is the biomass of the ith layer, hi is its width in cm, and zi is the mean height of the ith layer measured from the ground. For a species whose biomass is distributed evenly through all the canopy layers the value of H is simply half the maximum canopy height, while species with proportionally more biomass in higher layers would have a higher value of H and vice versa.

statistical analysis

We analysed total legume biomass, the relative abundance of climbers and the identity of the dominant legume species (the invading sown legume species with the highest biomass) to examine the effect of monoculture functional group, species and covariates on the success of the whole legume guild, and the relative success of climbers using the R statistical package ( R Development Core Team 2003). Invading legume biomass required log-transformation to meet the assumptions of anova. To avoid problems with heteroscedasticity we excluded plots where the total biomass of sown legume species was zero (Schmid et al. 1994). There were 11 such plots and they were distributed equally among monoculture functional groups: Fisher's exact tests revealed no significant differences between the proportion of grass (2/25), non-leguminous forb (3/29) and legume (6/21) plots in which there were no invading legumes. We used one-degree-of-freedom contrasts to break down functional group effects into legume vs. non-legume, and grass vs. non-leguminous forb components. We analysed the biomass of invaders belonging to other species in unsown subplots to assess whether these species (which were a mixture of unidentified grass, non-leguminous forb and unsown legume species) responded in the same way as sown legumes to the measured variables. Functional group effects, covariates and their interactions were always tested using monoculture species as the error term. In addition, there were two spatial components to the analysis: block and sub-block. Sub-block formed the error term for block.

Results

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

total legume biomass

The mean biomass of resident species, sown legume invaders and spontaneously arriving invaders in monocultures belonging to each functional group is shown in Fig. 1. In the analysis of plots with non-zero invading legume biomass, the only significant term was functional group (F2,20 = 12.76, P = 0.0002: block, sub-block and species identity were not significant). Contrasts revealed that this effect was almost entirely due to the difference between legume and non-legume plots (F1,20 = 22.24, P = 0.0001) while the difference between grass and non-leguminous forb plots was marginal (F1,20 = 3.27, P = 0.086). The direction of the effect shows that legume species are particularly poor at establishing in legume monocultures. The randomization test yielded 50 anovas in which these effects were never qualitatively different: the legume contrast was significant in every case (range of F1,20 values = 15.98–22.58, range of P-values = 0.0007–0.0001), showing that this is not an artefact of effectively having six invading species in legume plots but seven in grass and non-leguminous forb monocultures.

image

Figure 1. The mean (± 1 SE) resident biomass (resident), sown legume biomass (legume) and non-sown invading species biomass (unsown) in monoculture plots belonging to the three functional groups.

Download figure to PowerPoint

Average monoculture biomass had a less negative effect on invader biomass than average biomass height (Tables 1a and 2a), and was no longer significant if fitted after biomass height (F1,19 = 1.60, P = 0.27). In the model with average biomass height as covariate, the functional group effect is still significant but there is no significant interaction between the covariate and functional group, indicating that invader responses to biomass height within functional groups are parallel (Fig. 2a–c, Table 1a). The relatively poor ability of legumes to invade other legumes cannot therefore be attributed to differences in biomass or biomass height between legume species and other monocultures.

Table 1.  (A) anova of total sown legume biomass (> 0) with the covariate biomass height. The appropriate error term is given in each case. (B) Analysis of deviance of relative abundance of climbers with the covariate biomass height using approximate F-tests (to allow the use of correct error terms)
(A)
TermError termd.f.Sum of squaresMean squareFP(< F)
1 Block2 316.279 5.426 1.518 0.264
2 Sub-block71139.318 3.574 1.564 0.161
3 Biomass height6 151.55851.55812.99 0.00219
4 Functional group6 256.24928.125 7.087 0.00577
5 Biomass height: functional group6 2 0.203 0.102 0.0256 0.975
6 Monoculture species71767.462 3.968 1.737 0.0907
7 Residual 3068.557 2.285  
(B)
TermError termd.f.DevianceResidual d.f.Residual devianceFP(< F)
1 Block2 3 1.0796355.831 0.3810.769
2 Sub-block71110.3935245.438 1.7980.0989
3 Biomass height6 1 0.4025145.036 0.7530.398
4 Functional group6 210.8914934.14510.200.0012
5 Biomass height: functional group6 212.2664721.87911.480.0007
6 Monoculture species717 9.0793012.799 1.0170.468
7 Residual 30     
Table 2.  (A) anova of total sown legume biomass (> 0) with the covariate average biomass. The appropriate error term is given in each case. (B) Analysis of deviance of relative abundance of climbers with the covariate biomass using approximate F-tests (to allow the use of correct error terms)
(A)
TermError termd.f.Sum of squaresMean squareFP(< F)
1 Block2 316.279 5.426 1.5180.2644
2 Sub-block71139.318 3.574 1.5640.1608
3 Average biomass6 111.26811.268 2.8190.111
4 Functional group6 287.30043.65010.920.00089
5 Biomass: functional group6 2 8.967 4.483 1.1220.349
6 Monoculture species71767.938 3.996 1.7490.088
7 Residual 3068.557 2.285  
(B)
TermError termd.f.DevianceResidual d.f.Residual devianceFP(< F)
1 Block2 3 1.0796355.8310.3810.7689
2 Sub-block71110.3935245.4381.7980.0989
3 Average biomass6 1 0.0025145.4360.0020.96
4 Functional group6 214.7664930.6708.4470.00283
5 Biomass: functional group6 2 3.0124727.6571.7230.208
6 Monoculture species71714.8583012.7991.6650.108
7 Residual 30     
image

Figure 2. Mean (± 1 SE) sown legume biomass (a–c) and the relative abundance of climbers (d–f) recovered from monocultures differing in average biomass height belonging to three functional groups. For total biomass data are natural log-transformed, fitted lines are from analysis of covariance with the same slopes (the interaction was not significant). For relative abundance fitted lines are from logistic regression. Here, the interaction between the covariate and functional group is significant. Where no error bar is given the other replicates had zero biomass and were therefore excluded from this analysis.

Download figure to PowerPoint

In contrast to legume invaders, the total biomass of unsown invaders was not significantly affected by monoculture functional group (Table 3), and was lowest in non-leguminous forb rather than legume plots. It does not therefore appear that legume plots are generally harder to invade than those belonging to other functional groups. Biomass of unsown invaders declined with both monoculture biomass and biomass height, with the two covariates having similar explanatory power (Table 3).

Table 3. anova of total unsown invader biomass (> 0) with the covariates average biomass (A) and average biomass height (B). The appropriate error term is given in each case
(A)
TermError termd.f.Sum of squaresMean squareFP(< F)
1 Block2 3 2.595 0.865 0.4170.742
2 Sub-block71150.309 4.574 2.2060.0510
3 Average biomass6 116.51216.51210.500.0048
4 Functional group6 2 3.617 1.808 1.1500.340
5 Biomass: functional group6 2 2.896 1.448 0.9210.177
6 Monoculture species71926.738 1.573 0.7590.718
7 Residual 2449.751 2.073  
(B)
TermError termd.f.Sum of squaresMean squareFP(< F)
1 Block2 3 2.595 0.865 0.4170.742
2 Sub-block71150.309 4.574 2.2060.0510
3 Average height6 114.80914.80910.190.00534
4 Functional group6 2 8.183 4.091 2.8150.0879
5 Height: functional group6 2 2.062 1.031 0.7090.506
6 Monoculture species71924.709 1.453 1.4530.701
7 Residual 2449.751 2.073  

relative abundance of climbers

The relative abundance of climbers was analysed using logistic regression. However, quasi-F-tests were performed to allow the use of correct error terms (McCullagh & Nelder 1989). Both monoculture functional group (F2,20 = 5.28, P = 0.014) and species identity (F20,30 = 2.03, P = 0.038) significantly affected the relative abundance of climbers. In this case the main functional group effect is not due to the legume vs. non-legume contrast (F1,20 = 2.95, P = 0.1) but to the difference between grass and non-leguminous forb plots (F1,20 = 7.61, P = 0.012), with climbers performing relatively better in non-leguminous forb plots than in grass plots. Again, average biomass height had greater explanatory power than average biomass, although the main effect was non-significant in both cases (Tables 1b and 2b). However, the interaction between average biomass height and functional group was significant (Table 1b). This interaction can again be broken down using contrasts: in this case the difference between legume and non-legume plots is marginal (F1,17 = 2.99, P = 0.1), while the grass vs. non-leguminous forb component is highly significant (F1,17 = 19.98, P = 0.00034). The significant interaction occurs because the relative abundance of climbers declines with average biomass height in both grass and legume plots but increases strongly in non-leguminous forb plots (Fig. 2d–f). With average biomass height as covariate, the effect of functional group remains significant while the effect of species identity does not (Table 2b). The differential success of climbing and non-climbing legumes in grass vs. non-leguminous forb plots indicates that invasion success is not just a matter of general competitive ability but depends on the specific environment provided by the resident community.

dominance

Dominance (the invading sown legume species with the highest biomass) of climbers and non-climbers was non-randomly distributed across monoculture functional groups (χ2 = 7.456, d.f. = 2, P = 0.024). Contrasts reveal that this was due to the difference between grass and non-leguminous forb plots (χ2 = 7.431, d.f. = 1, P = 0.0064) and not to the difference between legume and non-legume plots (χ2 = 0.0256, d.f. = 1, P = 0.87). In this case, a species of non-climber (usually T. pratense) was most often dominant in grass plots, while a species of climber (usually V. cracca) was most often dominant in non-leguminous forb plots (Fig. 3).

image

Figure 3. The number of plots in which each invading legume species was dominant in monocultures belonging to three functional groups: (No = no successful legume invasion, T.p. = Trifolium pratense, T.r. = Trifolium repens, L.c. = Lotus corniculatus, M.l. = Medicago lupulina, V.c. = Vicia cracca, V.s. = Vicia sepium and L.p. = Lathyrus pratensis). Non-climbing legumes (filled bars) and climbing legumes (unfilled bars) have been grouped together.

Download figure to PowerPoint

competitive hierarchies

We analysed the ability of each legume species to resist invasion by other legumes (measured as the amount of invader legume biomass found in each plot). There were several zero-biomass plots (where residents totally resisted invasion) so we analysed log (invader biomass + 0.0005 g), as the lowest non-zero biomass was 0.005 g. This analysis revealed a significant effect of resident legume species identity on invasion resistance by legumes (F6,14 = 5.08, P = 0.0058), indicating that some legume species are more resistant to invasion by legumes than others. We then calculated the total biomass of each legume species found as an invader in other legume monocultures (invasion success) to see if it explained some of the variation between species in invader resistance. The covariate was significant (F1,14 = 11.60, P = 0.0043), although species identity remained significant (F5,14 = 3.78, P = 0.022). Species that are good at invading other legume monocultures are therefore more resistant to invasion themselves (Fig. 4), implying that a competitive dominance hierarchy exists within the legume guild.

image

Figure 4. The resistance of each of the legume species to invasion by other legumes declines with their ability to invade the legume plots. All data are natural log-transformed; fitted line is from linear regression. The intercept is low because the zero values pull the line down.

Download figure to PowerPoint

Discussion

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

The resistance of a plant community to invasion is rarely absolute (Levine et al. 2004). In our study, the seven legume species were capable of invading the majority of plots (85%) to some degree, although their success varied greatly. Invasion should be more difficult when the niches (or patterns of resource use) of invading species closely overlap those of established species (Tilman 2004). Invading legume biomass was lowest in legume monocultures and, as this functional group effect remained after fitting covariates that accounted for differences in total biomass or biomass height of resident species, it cannot be attributed to variation in either of these factors. In addition, legume plots were no less susceptible than other functional types to spontaneous invasion by other unsown species: similar total biomass of these species (mainly grasses and non-leguminous forbs) in all plots indicated that it is the match between the functional group of invaders and resident species that is critical to successful invasion. The relative success of climbing and non-climbing legumes was also affected by the functional group and characteristics of resident species: climbers dominated in tall non-leguminous forb plots, while non-climbers dominated in grass plots, regardless of height. This confirms that all legume species are not functionally identical and that conditions generated by resident species can shift the balance of competition between invaders from the same guild. This was confirmed by analysis of the identity of the dominant invader: T. pratense tended to be the dominant invader in grass plots, while V. cracca tended to dominate non-leguminous forb plots.

Total legume biomass was one to two orders of magnitude lower in legume plots than in either grass or non-leguminous forb plots of the same height. It is known from biodiversity experiments that plots containing at least one legume species have higher availability of soil nitrogen (Spehn et al. 2002; Scherer-Lorenzen et al. 2003), which might inhibit the germination of invading legumes. Alternatively, it is known that legume monocultures are notoriously difficult to sustain because of a rapid build-up of fungal pathogens (Spehn et al. 2002). It is possible that some of these pathogens or predators are not species-specific but can attack other species of legumes and therefore reduce their establishment. Irrespective of mechanism, the reduced ability of legumes to invade established patches of other legume species amounts to a priority effect (D’Antonio et al. 2001; Munday 2004). This can be highly stabilizing over a system of patches, as once established, each species effectively has a refuge from within-guild competition.

The experiment also shed light on competitive interactions within the legume guild. The invasion of legume monocultures by other legumes provided evidence for a strict competitive dominance hierarchy, although there were competitive reversals across non-legume monocultures. Legume species which themselves were most resistant to invasion (e.g. T. pratense) were also best at invading other legumes. However, T. pratense was not always the best invader of non-legume plots: V. cracca was usually the most successful invader in non-leguminous forb plots, and climbers in general made up a greater proportion of invader biomass as height increased in these plots, but not in grass plots. Competitive interactions between legume species can therefore be mediated by the presence of other species. While we might expect climbers to always have an advantage in a tall canopy, they also require structural support. Non-leguminous forbs tend to have a more robust growth form than grasses, as well as woodier flowering stems, which may provide a better scaffold for climbers and tip the balance of competition in their favour. The higher seed weight of climbers has probably also evolved to provide critical resources during early growth until higher canopy levels are reached (Westoby et al. 1996).

To date, most studies of community invasibility have focused on large pools of invaders belonging to several functional groups or on a particular species of interest (Knops et al. 1999; Joshi et al. 2000; Naeem et al. 2000; Hector et al. 2001; van Ruijven & Berendse 2003; Xu et al. 2004). Here we investigated the resistance of monocultures to invasion by several species belonging to one particular functional group. While we found that legume monocultures were harder to invade, this was only true for legume invaders and not for unsown invaders belonging to a range of functional groups. This supports the previous work of Fargione et al. (2003) and shows that the match between invaders and resident species is critical to invasion success. It also supports the notion that non-random assembly rules exist in grassland communities; something that is very difficult to demonstrate using observational data because of the statistical difficulty in distinguishing neutral from non-neutral models (Fridley et al. 2004).

Previous invasion studies using experimental diversity gradients usually show that diverse mixtures have greater resistance to invasion (Levine & D’Antonio 1999), while studies from natural communities sometimes show the opposite result (Robinson et al. 1995; Von Holle 2005). Our results suggest that established species could create heterogeneity, so that, for example, a two-species grass/non-leguminous forb mixture might offer opportunities for both climbing and non-climbing legumes. This would, however, depend on whether a mixture behaves as an average of its components or maintains a distinctive patch structure. Biodiversity experiments establish higher diversity plots from seed mixtures and may therefore lack the patch structure of natural communities where recruitment limitation and vegetative growth over long periods of time might create larger, monospecific or, indeed, monofunctional patches. This could explain some of the discrepancy between observational and experimental studies that link invasibility to resident diversity (Levine & D’Antonio 1999). Clearly, distinguishing between different functional types and species of invaders rather than treating all species as equal should help to shed light on why some invaders succeed where others fail, and under which circumstances, and for which species, biotic resistance is expected to be effective.

Acknowledgements

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

We thank the staff at Zurich-Reckenholz for all their help in maintaining this experiment. We also thank the reviewers for their helpful and constructive comments. The research was supported by the Swiss National Science Foundation (nr. 31–65224.01).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • Canham, C.D., Finzi, A.C., Pacala, S.W. & Burbank, D.H. (1994) Causes and consequences of resource heterogeneity in forests – interspecific variation in light transmission by canopy trees. Canadian Journal of Forest Research, 24, 337349.
  • Crawley, M.J., Brown, S.L., Heard, M.S. & Edwards, G.R. (1999) Invasion-resistance in experimental grassland communities: species richness or species identity? Ecology Letters, 2, 140148.
  • Crawley, M.J., Harvey, P.H. & Purvis, A. (1996) Comparative ecology of the native and alien floras of the British Isles. Philosophical Transactions of the Royal Society of London Series B-Biology Sciences, 351, 12511259.
  • D’Antonio, C.M., Hughes, R.F. & Vitousek, P.M. (2001) Factors influencing dynamics of two invasive C-4 grasses in seasonally dry Hawaiian woodlands. Ecology, 82, 89104.
  • Elton, C.S. (1958) The Ecology of Invasions by Animals and Plants. Methuen, London.
  • Fargione, J., Brown, C.S. & Tilman, D. (2003) Community assembly and invasion: an experimental test of neutral versus niche processes. Proceedings of the National Academy of Sciences of the United States of America, 100, 89168920.
  • Fridley, J.D., Brown, R.L. & Bruno, J.E. (2004) Null models of exotic invasion and scale-dependent patterns of native and exotic species richness. Ecology, 85, 32153222.
  • Grotkopp, E., Rejmanek, M. & Rost, T.L. (2002) Toward a causal explanation of plant invasiveness: seedling growth and life-history strategies of 29 pine (Pinus) species. American Naturalist, 159, 396419.
  • Hector, A., Dobson, K., Minns, A., Bazeley-White, E. & Lawton, J.H. (2001) Community diversity and invasion resistance: an experimental test in a grassland ecosystem and a review of comparable studies. Ecological Research, 16, 819831.
  • Hector, A., Spehn, E.M., Joshi, J., Scherer-Lorenzen, M., Schmid, B., Bazeley-White, E. et al. (2005) Ecosystem effects of the manipulation of plant diversity in European grasslands: data from the BIODEPTH project. Ecological Applications, M075–001, S1.
  • Hubbell, S.P. (2001) The Unified Neutral Theory of Biodiversity and Biogeography. Princeton University Press, Princeton.
  • Joshi, J., Matthies, D. & Schmid, B. (2000) Root hemiparasites and plant diversity in experimental grassland communities. Journal of Ecology, 88, 634644.
  • Knops, J.M.H., Tilman, D., Haddad, N.M., Naeem, S., Mitchell, C.E., Haarstad, J. et al. (1999) Effects of plant species richness on invasion dynamics, disease outbreaks, insect abundances and diversity. Ecology Letters, 2, 286293.
  • Leishman, M.R. & Thomson, V.P. (2005) Experimental evidence for the effects of additional water, nutrients and physical disturbance on invasive plants in low fertility Hawkesbury Sandstone soils, Sydney, Australia. Journal of Ecology, 93, 3849.
  • Levine, J.M., Adler, P.B. & Yelenik, S.G. (2004) A meta-analysis of biotic resistance to exotic plant invasions. Ecology Letters, 7, 975989.
  • Levine, J.M. & D’Antonio, C.M. (1999) Elton revisited: a review of evidence linking diversity and invasibility. Oikos, 87, 1526.
  • McCullagh, P. & Nelder, J.A. (1989) Generalized Linear Models. Chapman & Hall, London.
  • McKane, R.B., Johnson, L.C., Shaver, G.R., Nadelhoffer, K.J., Rastetter, E.B., Fry, B. et al. (2002) Resource-based niches provide a basis for plant species diversity and dominance in arctic tundra. Nature, 415, 6871.
  • Meiners, S.J., Cadenasso, M.L. & Pickett, S.T.A. (2004) Beyond biodiversity: individualistic controls of invasion in a self-assembled community. Ecology Letters, 7, 121126.
  • Munday, P.L. (2004) Competitive coexistence of coral-dwelling fishes: the lottery hypothesis revisited. Ecology, 85, 623628.
  • Naeem, S., Knops, J.M.H., Tilman, D., Howe, K.M., Kennedy, T. & Gale, S. (2000) Plant diversity increases resistance to invasion in the absence of covarying extrinsic factors. Oikos, 91, 97108.
  • Pacala, S.W., Canham, C.D., Saponara, J., Silander, J.A., Kobe, R.K. & Ribbens, E. (1996) Forest models defined by field measurements: estimation, error analysis and dynamics. Ecological Monographs, 66, 143.
  • R Development Core Team (2003) R: a Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna.
  • Robinson, G.R., Quinn, J.F. & Stanton, M.L. (1995) Invasibility of experimental habitat islands in a California winter annual grassland. Ecology, 76, 786794.
  • van Ruijven, J. & Berendse, F. (2003) Positive effects of plant species diversity on productivity in the absence of legumes. Ecology Letters, 6, 170175.
  • Scherer-Lorenzen, M., Palmborg, C., Prinz, A. & Schulze, E.D. (2003) The role of plant diversity and composition for nitrate leaching in grasslands. Ecology, 84, 15391552.
  • Schmid, B., Polasek, W., Weiner, J., Krause, A. & Stoll, P. (1994) Modeling of discontinuous relationships in biology with censored regression. American Naturalist, 143, 494507.
  • Silvertown, J., Dodd, M.E., Gowing, G.D.J. & Mountford, J.O. (1999) Hydrologically defined niches reveal a basis for species richness in plant communities. Nature, 400, 61.
  • Spehn, E.M., Hector, A., Joshi, J., Scherer-Lorenzen, M., Schmid, B., Bazeley-White, E. et al. (2005) Ecosystem effects of biodiversity manipulations in European grasslands. Ecological Monographs, 75, 3763.
  • Spehn, E.M., Joshi, J., Schmid, B., Diemer, M. & Korner, C. (2000) Above-ground resource use increases with plant species richness in experimental grassland ecosystems. Functional Ecology, 14, 326337.
  • Spehn, E.M., Scherer-Lorenzen, M., Schmid, B., Hector, A., Caldeira, M.C., Dimitrakopoulos, P.G. et al. (2002) The role of legumes as a component of biodiversity in a cross-European study of grassland biomass nitrogen. Oikos, 98, 205218.
  • Symstad, A.J. (2000) A test of the effects of functional group richness and composition on grassland invasibility. Ecology, 81, 99109.
  • Tilman, D. (1982) Resource Competition and Community Structure. Princeton University Press, Princeton, New Jersey.
  • Tilman, D. (1994) Competition and biodiversity in spatially structured habitats. Ecology, 75, 216.
  • Tilman, D. (1999) Ecological consequences of biodiversity: a search for general principles. Ecology, 80, 14551474.
  • Tilman, D. (2004) Niche tradeoffs, neutrality, and community structure: a stochastic theory of resource competition, invasion, and community assembly. Proceedings of the National Academy of Sciences of the United States of America, 101, 1085410861.
  • Turnbull, L.A., Coomes, D.A., Hector, A. & Rees, M. (2004) Seed mass and the competition/colonization trade-off: competitive interactions and spatial patterns in a guild of annual plants. Journal of Ecology, 92, 97109.
  • Turnbull, L.A., Crawley, M.J. & Rees, M. (2000) Are plant populations seed-limited? A review of seed sowing experiments. Oikos, 88, 225238.
  • Von Holle, B. (2005) Biotic resistance to invader establishment of a southern Appalachian plant community is determined by environmental conditions. Journal of Ecology, 93, 1626.
  • Von Holle, B. & Simberloff, D. (2004) Testing Fox's assembly rule: does plant invasion depend on recipient community structure? Oikos, 105, 551563.
  • Wedin, D. & Tilman, D. (1993) Competition among grasses along a nitrogen gradient – initial conditions and mechanisms of competition. Ecological Monographs, 63, 199229.
  • Westoby, M., Leishman, M.R. & Lord, J. (1996) Comparative ecology of seed size and dispersal. Philosophical Transactions of the Royal Society of London B, 351, 13091318.
  • Xu, K.Y., Ye, W.H., Cao, H.L., Deng, X., Yang, Q.H. & Zhang, Y. (2004) The role of diversity and functional traits of species in community invasibility. Botanical Bulletin of Academia Sinica, 45, 149157.