Survey
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project
Species Diversity, Invasion, and Alternative Community States in Sequentially Assembled Communities. Author(s): Lin Jiang, Lauren Brady, Jiaqi Tan Reviewed work(s): Source: The American Naturalist, Vol. 178, No. 3 (September 2011), pp. 411-418 Published by: The University of Chicago Press for The American Society of Naturalists Stable URL: http://www.jstor.org/stable/10.1086/661242 . Accessed: 22/02/2012 15:27 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. The University of Chicago Press and The American Society of Naturalists are collaborating with JSTOR to digitize, preserve and extend access to The American Naturalist. http://www.jstor.org vol. 178, no. 3 the american naturalist september 2011 Species Diversity, Invasion, and Alternative Community States in Sequentially Assembled Communities Lin Jiang,1,* Lauren Brady,2 and Jiaqi Tan1 1. School of Biology, Georgia Institute of Technology, Atlanta, Georgia 30332; 2. Department of Biology, Kenyon College, Gambier, Ohio 43022 Submitted April 12, 2011; Accepted May 13, 2011; Electronically published July 19, 2011 Online enhancement: appendix figure. Dryad data: http://dx.doi.org/10.5061/dryad.469kf. abstract: The relationship between resident species diversity and invasion is generally negative in experimental studies but takes various forms in observational studies of natural communities. We hypothesized that stochastic species colonization, which applies to natural communities but not to experimental communities generally assembled through simultaneous species introduction, may lead to nonnegative diversity-invasion relationships via incurring priority effects. To test this hypothesis, we manipulated both resident species diversity and colonization history in sequentially assembled communities of bacterivorous protist species. We found that, despite a significant effect of assembly history on invader abundance, invader abundance decreased with diversity. This result was largely driven by positive selection effects associated with the dominant influence of an invasion-resistant species, which shared the most similar resource use pattern with the invader, and by the overall weak priority effects observed for the resident communities. Increasing species diversity, however, significantly strengthened priority effects, providing the first experimental support for the idea that larger species pools promote alternative community states. We suggest that elucidating mechanisms regulating the strength of priority effects may help in understanding variation in diversity-invasion relationships among natural communities. Keywords: alternative stable states, biological invasions, community assembly, community invasibility, priority effects, species diversity. Introduction Many of the Earth’s ecosystems are undergoing major changes in biodiversity, including both the loss of native species and the addition of exotic species (Sax and Gaines 2003). One active area of research, linking native species extinction with exotic species introduction, asks whether species diversity of native (resident) communities influences their susceptibility to the invasion of exotic (nonresident) species. In agreement with the early proposition * Corresponding author; e-mail: [email protected]. Am. Nat. 2011. Vol. 178, pp. 411–418. 䉷 2011 by The University of Chicago. 0003-0147/2011/17803-52971$15.00. All rights reserved. DOI: 10.1086/661242 of Elton (1958), the majority of experiments that directly constructed resident communities found that increasing diversity tended to reduce invasion (Levine et al. 2002). This pattern can often be attributed to one or both of the following two processes—niche complementarity among resident species, which allows more-diverse communities to cover more of the niche space potentially utilized by invaders, and the positive selection effect (also known as the sampling effect), in which more-diverse resident communities are more likely to contain invasion-resistant species (e.g., Fargione and Tilman 2005). By contrast, studies of invasions in natural communities have produced widely different results, with positive, negative, and null relationships between resident species diversity and the success of invading species all being common (Herben et al. 2004; Fridley et al. 2007). Hypotheses explaining this discrepancy have focused on associating positive diversity-invasion relationships with broadscale processes, particularly the role of spatial heterogeneity in abiotic and biotic conditions (summarized in Fridley et al. 2007). However, although these hypotheses can potentially account for positive diversity-invasion patterns that dominate observational studies at large spatial scales, they do not provide an explanation of the frequent nonnegative patterns found in observational studies conducted at scales comparable to those of experimental studies. One factor that may potentially influence the diversityinvasion relationship but that remains unexplored in this context is the history of community assembly. Resident communities in experimental studies of diversity and invasion are generally assembled by introducing all member species at the same time. However, this assumption of simultaneous species arrival is unlikely to hold for natural communities, which are typically colonized by species from the regional pool sequentially (Fukami 2010). Frequently, in situations in which sequential community assembly leads to priority effects (e.g., Almany 2004), the stochastic nature of species colonization can result in al- 412 The American Naturalist ternative community states characterized by the dominance of different early-colonizing species in otherwise similar habitats. Under this scenario, one may expect both the positive selection effect and niche complementarity to be less important. Priority effects reduce the importance of the positive selection effect when invasion-resistant species, as late colonizers, attain small population sizes or even fail to establish viable populations. Priority effects also tend to reduce the importance of niche complementarity, because it results in communities with more-uneven species abundance, in which the magnitude of complementary interactions is constrained by the limited contribution from the low-abundance organisms (Jiang et al. 2009). We hypothesize that the diminished role of both processes, which are associated with stochastic assembly history, could potentially bring about nonnegative diversity-invasion relationships in natural communities. To examine this hypothesis, we assembled aquatic bacterivorous protist communities that differed in species richness and in the order of sequential species colonization, and we subjected these communities to the invasion of a nonresident species. Protist microcosms are particularly suitable for answering community assembly questions (e.g., Fukami and Morin 2003; Jiang and Patel 2008), mainly due to their amenability to experimental manipulations and the short generation times of protists that permit communities to quickly attain steady states. Because invasion resistance is often considered as a type of ecosystem function (Levine et al. 2002), our work is among the few to explore potential functional consequences of community assembly (also see Zhang and Zhang 2007; Fukami et al. 2010), but is the first, to our knowledge, to directly manipulate both species diversity and assembly history. By doing so, we explored whether the assembly history of a community affects its susceptibility to invasion and whether communities assembled from different histories still give rise to negative diversity-invasion relationships, as seen in experiments with simultaneous species introduction. Our experiment also allowed us to address the question of how the size of the species pool affects the likelihood of alternative community states. Theory suggests that differences in assembly history are more likely to result in alternative community states as the species pool size increases (Law and Morton 1993; Chase 2003; Fukami 2004b), but an experimental test of this prediction is lacking. Material and Methods Study Organisms We assembled resident protist communities from a pool of five bacterivorous ciliate species, including Colpidium kleini, Loxocephalus sp., Paramecium aurelia, Spirostomum teres, and Tetrahymena pyriformis, and subjected these communities to the invasion of a sixth bacterivorous ciliate, Paramecium multimicronucleatum. Prior to the experiments, each of these species was grown separately in its stock culture on a multispecies bacterial assemblage, including Bacillus cereus, Bacillus subtilis, Serratia marcescens, and a number of unidentified species, which were introduced to experimental microcosms along with their protist consumers (see Microcosms). Microcosms Each microcosm was a 250-mL Pyrex glass bottle filled with 100 mL of protozoan pellet medium (concentration, 0.55 g of protozoan pellet [Caroline Biological Supply, Burlington, NC] per 1 L of deionized water). Bottles and medium were autoclave-sterilized, and the medium was inoculated with bacteria from the protist stock cultures. To prepare for bacterial inoculation, we mixed samples from the stock culture of each protist species and removed the protists from the mixed sample by running it through a 1.0-mm sterile filter. The filtrate, containing only bacteria, was used to inoculate the medium. Each microcosm also received two autoclave-sterilized wheat seeds as an additional carbon source. Invasion Experiment Resident protist communities differed in three aspects: species richness (one, two, or five species), composition, and colonization sequence. Single- and two-species communities each had five different compositions, with each resident species represented once in single-species communities and twice in two-species communities, whereas five-species communities included all five resident species. We varied species colonization history for two- and fivespecies communities, allowing each species to be the first colonizer once within each composition; this resulted in two assembly sequences for each two-species composition and five assembly sequences for the five-species composition (see fig. 1B for the list of treatments, in which resident species were represented by the first letter of their genus names; e.g., CP stands for the two-species cultures where Colpidium colonized before P. aurelia). We replicated each assembly sequence (20 total) three times, resulting in 60 microcosms. During resident community assembly, resident species were added at 4-day intervals. Resident communities were allowed to equilibrate for another 2 weeks after the introduction of the last colonizers before each was challenged with P. multimicronucleatum. The inoculum of each protist species (including the invader) consisted of approximately 200 individuals, with Community Assembly and Invasion 413 day-old cultures for inoculation. Microcosms were colonized by the first protist species 24 h after bacterial inoculation and were run as semicontinuous cultures for the duration of the experiment (52 days), with 7% content of each microcosm replaced with fresh medium each week. All microcosms were maintained in an unlighted incubator at 22⬚C. We sampled microcosms to estimate population density of each protist species three times a week, starting from 2 days after the introduction of the first species (the first sampling day designated as day 0). This procedure consisted of three steps. First, the microcosm was swirled to ensure an even distribution of protist individuals in the medium, and a small sample (0.3–0.4 mL) was withdrawn using a sterile Pasteur pipette. Second, the sample was weighed to determine its volume and distributed onto a petri dish in small drops. Third, the petri dish was placed under a stereoscopic microscope, and the number of individuals of each species in the sample was counted. We diluted samples for high-abundance species before counting. All protist density data were recorded as the number of individuals per milliliter. Bacterial Experiment Figure 1: Final population density of the invader at each resident species richness level (A) and in each assembly history treatment (B). In A, the Poisson generalized estimating equation line is shown along with the data. In B, values are arithmetic means ⫹ SE. Horizontal bars in B indicate nonsignificant differences in the leastsquares means between two assembly sequences within the same species composition treatment (e.g., PT vs. TP). C, Colpidium species; L, Loxocephalus species; P, Paramecium species; S, Spirostomum species; T, Tetrahymena species. the exception of S. teres, which consisted of 20 individuals because of its lower stock culture abundance. To minimize variation in physiological conditions of initial protist populations introduced at different times, we set up fresh stock cultures of each species periodically and always used 14- To gain a more mechanistic understanding of resourcebased competitive interactions between protists (see Fox 2002), we performed an experiment in which we estimated the impacts of individual protist populations on bacterial resources. This experiment had seven treatments, including a control with bacteria only and six protist monoculture treatments corresponding to each of the six protist species used in the invasion experiment. Each treatment was replicated five times, for a total of 35 microcosms. This experiment proceeded as in the invasion experiment, except that only a single protist species was introduced into the microcosms. We ran the experiment for 3 weeks, at the end of which period we sampled each microcosm for bacterial abundance. Samples were serially diluted and spread on nutrient agar plates, and the number of colonyforming units (CFUs) for each bacterial morphospecies was counted after 10 days’ incubation at room temperature. Bacterial density data were recorded as the number of CFUs per milliliter. Data Analysis Because protist communities appeared to approach their steady states toward the end of the invasion experiment (fig. A1 in the online edition of the American Naturalist), data from the last sampling day were used to quantify treatment effects. We used Poisson generalized estimating equations (GEEs) to model invasion success, measured as 414 The American Naturalist final invader population density, as a function of species richness, composition, and colonization history. GEEs were chosen over ordinary Poisson regression models because they can account for possible nonindependence among data and tend to provide more-robust estimates of standard errors of model parameters. For our data, GEE yielded essentially the same parameter values as the Poisson regression model, but generated larger confidence intervals for the same parameters, compared with the latter models; results based on GEEs are thus more conservative. We ran three separate GEEs: the first to test for the general relationship between resident species richness and invader density, the second for the effect of resident species composition (nested within richness) on invader density, and the third for the effect of assembly history (nested within composition) on invader density. Given the overall significant effect of history (see “Results”), we also used planned contrasts to assess the difference in invader density among communities experiencing different histories within each species composition. We used the overyielding index Dmax (Loreau 1998), calculated for each multispecies culture, to assess the relative importance of complementarity and selection effects on community invasion resistance. The additive partitioning model of Loreau and Hector (2001) was not used because it was not possible to determine the contribution of each resident species in multispecies cultures to invasion resistance. The original Dmax was developed for biomass production and attains positive values when the performance (biomass) of a polyculture exceeds the monoculture performance of its most productive member species when grown alone. In the context of invasion resistance, however, better performance corresponds to lower invader abundance. To facilitate comparison with previous work, we calculated Dmax on invader abundance according to its original formula but reversed its sign; positive Dmax thus corresponds to invader abundance of a polyculture being lower than that of the monoculture of its most invasionresistant member species when grown alone. A one-sample t-test tested whether Dmax value in each polyculture significantly differed from 0. Significant positive Dmax values, indicative of the ability of multispecies assemblages to lower invader abundance beyond that of the most invasion-resistant species, would suggest niche complementarity, whereas Dmax not different from 0, which is indicative of the dominant role of the most invasion-resistant species, would suggest the positive selection effect. We quantified the degree of similarity in the final structure of protist communities that shared the same species composition, but with different assembly histories, using the Morisita-Horn similarity index. This index ranges from 0 (no species shared by the two communities being compared) to 1 (two communities sharing the same species and the same relative abundance of these species) and is known to be robust to changes in species richness (Wolda 1981), which facilitates the comparison between species richness levels. For comparison, we also calculated the widely used Bray-Curtis similarity index. Results based on the two similarity measures were similar, and we report only those based on the Morisita-Horn measure here. To test the idea that species pool size may influence the likelihood of alternative community states, we compared similarity values in the two- and five-species cultures using a permutation test that randomly shuffled data between the two richness levels 1,000 times. Analyses included the invader as a component of the protist community, but results were qualitatively similar when considering resident protist species only. We performed principal component analysis (PCA) of bacterial density data from the bacterial experiment to discern the difference in bacterial community structure among protist monocultures. Principal components were calculated based on the correlation matrix of bacterial densities, which were log transformed (log10[x ⫹ 1]) prior to the analysis. Results Invasion success, measured as invader population density at the end of the experiment, decreased significantly as the number of resident species increased (fig. 1A; GEE: x 2 p 8.58, df p 1, P p .0034 for richness). The species richness effect remained significant (x 2 p 9.18, df p 2, P p .0101) in a GEE model that treated species composition as a nesting factor within species richness, which also revealed a significant effect of species composition (x 2 p 20.12, df p 8, P p .0099). A third GEE model that treated species assembly history as a factor nested within species composition indicated significant effects of both factors on invasion success (composition: x 2 p 26.81, df p 10, P p .0028; history: x 2 p 18.09, df p 9, P p .0342). The effect of assembly history, however, was not uniform across species composition treatments. Planned contrasts revealed significant differences in the invader density of communities sharing the same composition but with different assembly histories in three of five cases for two-species communities and three of 10 cases for fivespecies communities (fig. 1B). Negative values were attained for Dmax in most multispecies cultures (fig. 2); only in one treatment (LS) was it significantly different from 0 (t 2 p 4.38, P p .0484). None remained significant after Bonferroni adjustment. The dominance of early-colonizing species over later ones was observed only in the five-species communities, not two-species communities (fig. A1). For example, reversing the order of the introduction of the two most Community Assembly and Invasion 415 Figure 2: Dmax value (mean ⫹ SE) for each assembly history treatment. In all but one treatment (LS), Dmax is not significantly different from 0 in one-sample t-tests. None of the values are significantly different from 0 after Bonferroni adjustment. C, Colpidium species; L, Loxocephalus species; P, Paramecium species; S, Spirostomum species; T, Tetrahymena species. ident communities, such that species dominance patterns were unaffected by history, the abundance of the subordinate species did change with history in these communities (fig. A1). In particular, Spirostomum attained larger populations as a subordinate species when it was the first colonizer rather than when it was the second colonizer, and this increased abundance of Spirostomum corresponded to the reduced invader abundance in the twospecies communities in which Spirostomum was present (fig. A1). Additional evidence for the linkage between Spirostomum and invasion success came from the five-species communities, where stronger priority effects led to changes in the identity of the dominant species as well as to changes in Spirostomum abundance. A GEE model relating invader abundance to all resident species in these communities identified Spirostomum abundance as the only significant predictor of invader abundance (x 2 p 4.93, P p .0264). Note that Spirostomum was the least abundant resident species in these communities (fig. A1), and some of the other resident species, when alone, posed greater resistance to invasion (fig. 1B). Therefore, the negative association between Spirostomum and invasion may not be solely attributable to the inhibitive effect of Spirostomum on the invader via the reduction of their shared bacterial re- abundant species (Colpidium and Loxocephalus) in the five-species communities caused the reversal of their relative abundance (fig. A1). As a result, similarity in the final structure of communities that shared identical species composition but underwent different assembly histories, measured as the Morisita-Horn index, was significantly greater for two-species communities than for five-species communities (fig. 3; randomization test: P ! .001). PCA revealed both similarities and differences in the structure of bacterial communities among protist monocultures. In particular, bacterial communities in the invader (Paramecium multimicronucleatum) cultures were most similar to those in the cultures of Paramecium aurelia, which shares the same genus as the invader; Spirostomum cultures possessed distinct bacterial communities from other protist cultures (fig. 4). Discussion Our experiment clearly showed that the assembly history of a community can affect its susceptibility to invasion. This result was closely linked to changes in the structure of resident communities with different assembly histories. Although priority effects were weak in the two-species res- Figure 3: Morisita-Horn similarity between communities with the same species composition but different assembly sequences, averaged across all compositions for each species richness level. A permutation test that randomly assigned values to the two richness treatments showed that the similarity among the five-species communities was significantly smaller than that among the two-species communities (P ! .001). Values are mean ⫹ SE. 416 The American Naturalist Figure 4: Plot of the first and second principal component scores (PC1 and PC2, respectively) from the principal component analysis of bacterial community structure in the bacterial experiment. PC1 and PC2 accounted for 19.5% and 12.5% of variation in the data, respectively. Each treatment had five replicates. B, bacteria only; C, Colpidium monocultures; I, invader (Paramecium multimicronucleatum) monocultures; L, Loxocephalus monocultures; P, Paramecium aurelia monocultures; S, Spirostomum monocultures; T, Tetrahymena monocultures. sources, independent of other species in resident communities. More likely, it may have reflected its complementary use of bacterial resources with other resident species. Consistent with this idea, we found that Spirostomum imposed different influences on bacterial communities than did other protist consumers (fig. 4), which suggests its distinct niche. Overall, our findings agree with those of several recent studies that found significant effects of species colonization history on ecosystem functioning, including biomass production in micro-algal communities (Zhang and Zhang 2007) and grassland plant communities (Korner et al. 2008) and wood decomposition in fungal communities (Fukami et al. 2010). Together, these results suggest that history-dependent alternative community states that differ in species composition and/or abundance may exhibit different levels of ecosystem function, which underscores the need for future assembly studies to consider the functional consequences of assembly history. As in many experimental studies that directly manipulate resident species diversity (Levine et al. 2002), a negative diversity-invasion relationship emerged in our experiment. This negative effect of resident species diversity on invasion can be primarily attributed to the positive selection effect, as indicated by the Dmax values, which rarely differed from 0. The positive selection effect was largely driven by the fact that more-diverse communities have an increased likelihood of including Paramecium aurelia, whose monoculture showed the greatest resistance to invasion; resident communities containing this species exhibited a substantially lower level of invasion than did those communities that lacked this species (fig. 1B; planned contrast: P ! .0001). Note that the invader (Paramecium multimicronucleatum) belongs to the same genus as P. aurelia but does not belong to the same genus as other resident species. Darwin (1859) hypothesized the difficulty of a nonnative species becoming naturalized in habitats where its congeners are present, reasoning that the substantial ecological similarity of congeners results in strong competition between them. The lower abundance of P. multimicronucleatum in the presence of P. aurelia, coupled with the similarity in their use of bacterial resources, provided strong support for this hypothesis (see Jiang et al. 2010 for a formal experimental test of the hypothesis). Notably, changing assembly history did not eliminate this positive selection effect associated with interactions of closely related species. In both two-species communities with weak priority effects and five-species communities with stronger priority effects, P. aurelia, as the most invasion-resistant species, reached appreciable abundance, which rendered the positive selection effect robust to variation in assembly history. The positive selection effect also dwarfed any complementarity effects among resident species (undetected using Dmax) and any possible change in the strength of complementary interactions as a result of history-dependent species abundances. In all, the strong positive selection effect led to a decrease in invasion success with increasing species diversity, despite the influence of assembly history on invasion. As species diversity increased, the structure of communities that shared the same species composition but underwent different assembly histories became less similar. One could argue that the lower similarity in the five-species communities may simply be a transient phenomenon that may go away if the experiment runs longer, because these communities experienced a longer initial assembly period (with the last resident species introduced on day 20) than did the two-species communities (in which the last resident species introduced on day 8). Our data, however, indicate that most communities attained relatively steady states by the end of the experiment (fig. A1). Moreover, even if the five-species communities were not at their steady states (see Fukami 2004a for an example of community assembly with long transient dynamics), they demonstrated longer-lasting priority effects than did the two- Community Assembly and Invasion 417 species communities. For instance, P. aurelia was more abundant than was Tetrahymena for only 5 days in the PT treatment after both species were introduced, before yielding its dominance to the latter; Tetrahymena dominated P. aurelia in the TP treatment throughout the experiment (data not shown). By contrast, when introduced earlier in the five-species communities, each of the two species held its dominance over the other for the duration of the experiment (fig. A1); this is especially evident when comparing treatments CSLPT and LTPCS, in which the two species were introduced 4 days apart, as in the two-species communities. Considering the short generation times (less than 2 days) of these organisms, these results suggest that species pool size can influence priority effects over ecologically meaningful timescales. Indeed, these results are consistent with the theoretical prediction that increasing species pool size tends to increase the likelihood of alternative community states (Law and Morton 1993; Chase 2003; Fukami 2004b). It has been suggested that communities with larger species pools tend to contain more species with similar traits, the interaction of which promotes priority effects (Chase 2003). This is certainly the case for the invader, for which the priority effects associated with its resident congener P. aurelia, which was more frequently present in more-diverse communities, were primarily responsible for its decrease with an increase in resident species diversity (i.e., the positive selection effect). However, this hypothesis cannot account for the fact that priority effects involving the same pair of resident species (e.g., P. aurelia and Tetrahymena) were stronger with increasing diversity in our experiment, which instead may be better explained by complementary resource use (fig. 4) among earlier-colonizing resident species leaving fewer resources for later colonizers in more-diverse communities (i.e., niche complementarity). Future studies should further explore the idea that niche complementarity and positive selection effects may both contribute to the increase of priority effects with diversity. We note that both mechanisms are often found to play important roles in causing the increase of ecosystem biomass production with diversity (Cardinale et al. 2007). Two aspects of our invasion experiment warrant some clarification. First, for convenience, we confined the term “community assembly” to resident communities and the term “invasion” to the nonresident species, which gives the impression that community assembly and invasion are two independent concepts. These two concepts, however, are in fact closely linked, because community assembly involves the invasion and subsequent interactions of species, regardless of whether the species are resident (native) or nonresident (nonnative). Both terms have often been used together in the same context (e.g., Fargione et al. 2003; Tilman 2004). As such, our experiment (and many other invasion experiments) may also be considered as an assembly experiment with the nonresident species introduced last. Second, our experiment had only three replicates. We chose to replicate treatments three times because our previous assembly experiments (Jiang and Patel 2008; Jiang et al. 2010, 2011) showed that this level of replication was adequate at detecting treatment differences, which is also the case for the present experiment. There was indeed some unexpectedly appreciable variation among replicates in a few treatments, which nevertheless did not prevent us from detecting significant diversity and assembly history effects. One may also ask whether increasing replication may increase the chance of finding stronger priority effects in some of the treatments, based on the premise that replicates may randomly diverge along different trajectories (e.g., Jiang et al. 2011). However, such random divergence has not been found for communities of bacterivorous protists (Jiang et al. 2011), probably because stable bacteriaprotist interactions minimize the effects of demographic stochasticity that trigger divergence. Consistent with this, there was little sign of random divergence in our experimental communities of bacterivorous protists. Nevertheless, our experiment could certainly benefit from more replication, which would have increased statistical power. Our study provides the first experimental evidence for the increased likelihood of alternative community states in more-diverse communities and demonstrates the importance of resident community assembly history for invasion. However, we still found a negative diversityinvasion relationship in our sequentially assembled communities, largely driven by an invasion-resistant resident species that remained fairly abundant with changes in assembly history. Nevertheless, we caution against jumping to the conclusion that assembly history contributes little to the nonnegative diversity-invasion patterns that are observed in many natural communities. Priority effects involving all resident community members (including the invasion-resistant species) in our experiment were generally weak, which limited the effect of assembly history on invasion. In contrast, priority effects of various strength, including strong effects that result in the extinction or severe reductions in the density of later colonizing species (e.g., Zhang and Zhang 2007), have been reported (Chase 2003). Our understanding of the role of assembly history for community resistance to biological invasions and other ecosystem functions will certainly benefit from a better knowledge of the mechanisms regulating the strength of priority effects (e.g., Chase 2003, 2007, 2010). Acknowledgments We thank Z. Pu, W. Ryberg, and two anonymous reviewers for constructive comments that significantly improved this 418 The American Naturalist manuscript. This project was supported by the National Science Foundation (grants OCE-0851606 and DEB064041), and L. Brady was supported by OCE-0851606 as a National Science Foundation Research Experience for Undergraduates student. Literature Cited Almany, G. R. 2004. Priority effects in coral reef fish communities of the Great Barrier Reef. Ecology 85:2872–2880. Cardinale, B. J., J. P. Wright, M. W. Cadotte, I. T. Carroll, A. Hector, D. S. Srivastava, M. Loreau, et al. 2007. Impacts of plant diversity on biomass production increase through time because of species complementarity. Proceedings of the National Academy of Sciences of the USA 104:18123–18128. Chase, J. M. 2003. Community assembly: when should history matter? Oecologia (Berlin) 136:489–498. ———. 2007. Drought mediates the importance of stochastic community assembly. Proceedings of the National Academy of Sciences of the USA 104:17430–17434. ———. 2010. Stochastic community assembly causes higher biodiversity in more productive environments. Science 328:1388– 1391. Darwin, C. 1859. The origin of species. J. Murray, London. Elton, C. S. 1958. The ecology of invasions by animals and plants. Methuen, London. Fargione, J., C. S. Brown, and D. Tilman. 2003. Community assembly and invasion: an experimental test of neutral versus niche processes. Proceedings of the National Academy of Sciences of the USA 100:8916–8920. Fargione, J. E., and D. Tilman. 2005. Diversity decreases invasion via both sampling and complementarity effects. Ecology Letters 8:604– 611. Fox, J. W. 2002. Testing a simple rule for dominance in resource competition. American Naturalist 159:305–319. Fridley, J. D., J. J. Stachowicz, S. Naeem, D. F. Sax, E. W. Seabloom, M. D. Smith, T. J. Stohlgren, et al. 2007. The invasion paradox: reconciling pattern and process in species invasions. Ecology 88: 3–17. Fukami, T. 2004a. Assembly history interacts with ecosystem size to influence species diversity. Ecology 85:3234–3242. ———. 2004b. Community assembly along a species pool gradient: implications for multiple-scale patterns of species diversity. Population Ecology 46:137–147. ———. 2010. Community assembly dynamics in space. Pages 45– 54 in H. A. Verhoef and P. J. Morin, eds. Community ecology: processes, models, and applications. Oxford University Press, Oxford. Fukami, T., and P. J. Morin. 2003. Productivity-biodiversity rela- tionships depend on the history of community assembly. Nature 424:423–426. Fukami, T., I. A. Dickie, J. P. Wilkie, B. C. Paulus, D. Park, A. Roberts, P. K. Buchanan, et al. 2010. Assembly history dictates ecosystem functioning: evidence from wood decomposer communities. Ecology Letters 13:675–684. Herben, T., B. Mandak, K. Bimova, and Z. Munzbergova. 2004. Invasibility and species richness of a community: a neutral model and a survey of published data. Ecology 85:3223–3233. Jiang, L., and S. N. Patel. 2008. Community assembly in the presence of disturbance: a microcosm experiment. Ecology 89:1931–1940. Jiang, L., S. Q. Wan, and L. H. Li. 2009. Species diversity and productivity: why do results of diversity-manipulation experiments differ from natural patterns? Journal of Ecology 97:603–608. Jiang, L., J. Q. Tan, and Z. C. Pu. 2010. An experimental test of Darwin’s naturalization hypothesis. American Naturalist 175:415– 423. Jiang, L., H. Joshi, S. K. Flakes, and Y. Jung. 2011. Alternative community compositional and dynamical states: the dual consequences of assembly history. Journal of Animal Ecology 80:577–585. Korner, C., J. Stocklin, L. Reuther-Thiebaud, and S. Pelaez-Riedl. 2008. Small differences in arrival time influence composition and productivity of plant communities. New Phytologist 177:698–705. Law, R., and R. D. Morton. 1993. Alternative permanent states of ecological communities. Ecology 74:1347–1361. Levine, J. M., T. Kennedy, and S. Naeem. 2002. Neighbourhood scale effects of species diversity on biological invasions and their relationship to community patterns. Pages 114–124 in M. Loreau, S. Naeem, and P. Inchausti, eds. Biodiversity and ecosystem functioning: synthesis and perspectives. Oxford University Press, New York. Loreau, M. 1998. Separating sampling and other effects in biodiversity experiments. Oikos 82:600–602. Loreau, M., and A. Hector. 2001. Partitioning selection and complementarity in biodiversity experiments. Nature 412:72–76. Sax, D. F., and S. D. Gaines. 2003. Species diversity: from global decreases to local increases. Trends in Ecology & Evolution 18: 561–566. 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 USA 101:10854–10861. Wolda, H. 1981. Similarity indexes, sample-size and diversity. Oecologia (Berlin) 50:296–302. Zhang, Q. G., and D. Y. Zhang. 2007. Colonization sequence influences selection and complementarity effects on biomass production in experimental algal microcosms. Oikos 116:1748–1758. Associate Editor: Oswald J. Schmitz Editor: Judith L. Bronstein