Aim The invasion of natural communities by alien species represents a serious threat, but creates opportunities to learn about community functions. Neutral theory proposes that the niche concept may not be needed to explain the assemblage and diversity of natural communities, challenging the classical view of community ecology and generating a lasting debate. Biological invasions, when considered as natural experiments, can be used to contrast some of the predictions of neutral and classic niche theories.
Methods We use data from biological invasions as natural experiments to contrast some of the fundamental predictions of neutral theory.
Results Some emerging patterns did not differ from neutral model expectations (e.g. the relationship between native and exotic species richness, invasibility of resource-rich habitats, and the relationship between propagule release and invasion success). Nevertheless, other patterns (e.g. experimental evidence of the relationship between diversity and susceptibility to invasion, the invasion of communities with a low resource availability, invasiveness related to species traits) contrasted with the predictions that can be inferred from neutral theory.
Main conclusions Neutral theory correctly highlights the need to include randomness in models of community structure. Biological invasion patterns show that neutral forces are important in structuring natural communities, but the patterns differ from those inferred from a complete neutral model. For biodiversity-conservation purposes, the implications of accepting or not accepting neutral theory as a process-based theory are very important.
The invasion of exotic species into natural communities is a world-wide phenomenon with significant concomitant impacts on community structure and dynamics (e.g. Carlton & Geller, 1993; Mack et al., 2000). To set priorities for the control of invasive species, the recovery of impacted communities, or the prevention of future invasions, it is important to understand how community properties affect vulnerability to invasions (e.g. Kolar & Lodge, 2001; Shea & Chesson, 2002). Identifying the causes of invasion will not only provide tools for conservation purposes, but will also provide insight into processes structuring natural communities, thus improving our understanding of nature (see Elton, 1958; Bruno et al., 2005; Catford et al., 2009).
Ecological theory has been built on the idea that species differ in their needs from the environment and in how their activities affect the environment (e.g. Hutchinson, 1959; MacArthur & Levins, 1967; Caswell, 1978). Indeed, most ecologists would agree that ecological differences among species are fundamental in maintaining the diversity observed in natural communities (see Hutchinson, 1959). However, most ecologists would also agree that stochastic forces are important in shaping species diversity, and that a theory without randomness is as realistic as a theory without species interactions (see Tilman, 2004; Chase, 2005). Despite this apparent agreement, a recent series of publications has shown that patterns of relative species abundance (RSA) could be explained by a theory in which all interspecific interactions are turned off (e.g. McGill, 2003; Volkov et al., 2003, 2005, 2007). This idea (postulated by Hubbell, 2001) states that the niche concept may not be needed to explain the observed assemblage of natural communities and developed an alternative approach based on sampling theory. This theory, called the unified neutral theory of biodiversity and biogeography (Hubbell, 2001), has become one of the pillars of macroecology (e.g. Chave, 2004; Hu et al., 2006) and also has become a controversial issue (see Chase, 2005).
The unified neutral theory of biodiversity and biogeography states that species in the same functional guild are ecologically the same, and their abundances simply fluctuate randomly over time. At the metacommunity level, species dynamics depend on the probabilities of mortality, speciation and reproduction. The neutrality assumption implies that these probabilities are the same for all species. At local community levels, where speciation is not expected, dynamics are maintained by births, deaths and migration from the metacommunity.
Since its inception, neutral theory has faced strong scepticism concerning its fundamental assumption of the equivalence among individuals of different species. Resistance to this theory comes from community ecologists who find difficult to accept that the differences between species that they recognize in their work have no effect on the structure of natural communities (e.g. Bell, 2001; Chave, 2004). Although some evidence supports species equivalence (Alonso et al., 2006) and others reject it (see Chave, 2004), the discussion is not whether individuals are ‘really the same’, but whether the differences between individuals of different species are important or not in shaping the community structure (see Hubbell, 2005).
Despite its success in generating patterns like those in nature, the only evidence supporting neutral theory relies on comparisons between observed and theoretically predicted patterns of RSA from species-rich communities. The similarity in ‘curve fitting’ achieved by both neutral and niche theories (McGill, 2003; Tilman, 2004; Pueyo et al., 2007) makes those approaches inadequate for discriminating the relative contribution of niche and neutral processes to the structure of natural communities (Nee & Stone, 2003). Given that it is impossible to conduct experiments at such a large scale, the debate arrives to a dead end (Adler et al., 2007). However, the invasion of natural communities by alien species, when considered as a large-scale experiment, can be used to contrast some predictions of neutral theory. Some data have been interpreted as evidence of incompatibilities between neutral theory and invasion ecology (Tilman, 2004; Chase, 2005), but the same data were also used to support the theory (Davis, 2003; Herben et al., 2004). Therefore, we explicitly developed neutral predictions and contrasted them with widespread ecological patterns shown in studies on biological invasions.
Neutral theory and invasion processes
In recent years, ecologists have made efforts to develop predictive theories to determine priorities for the control of invasive species and the prevention of future invasions (e.g. Kolar & Lodge, 2001; Shea & Chesson, 2002). Although generalizations are hard to develop, there is enough information to answer questions such as ‘are all communities equally susceptible to be invaded?’, ‘are all species equally invasive?’, or ‘what makes some species more invasive than others?’ (see Richardson & Pyšek, 2006; for a recent review on this subject). Those answers may also provide insight into the processes structuring natural communities (see Elton, 1958; Bruno et al., 2005) and can be used to contrast some fundamental predictions of neutral theory (see Table 1).
Table 1. Taking invasion ecology as a large biogeographical experiment, some derivations of neutral theory can be contrasted. The table provides a summary of expected results from neutral theory and published evidence as answers for key questions regarding invasion ecology.
|Is invasion affected by species diversity?||Success of invaders is not affected by species diversity in a community||Poorly diverse communities are usually more easily invaded than highly diverse ones (because of resource-use complementarity or decreased chances for the presence of highly suppressive species)||Tilman (1997), Knops et al. (1999), Naeem et al. (2000), Wardle (2001), Kennedy et al. (2002), van Ruijven et al. (2003) and Tilman (2004)|
|Does resource availability or disturbance frequency affect invasion?||Invasion is expected to increase as resources or disturbance increase||Even though invasion is more frequent in resource-rich or frequently disturbed communities, it also occurs in resource-poor environments||Daehler (2003), Tilman (2004) and Funk & Vitousek (2007)|
|Are all species equally invasive?||All species in the same functional guild are equivalent, and thus there should be no differences in their invasiveness||Invasive species differ from non-invasive species in diverse ecological traits (e.g. form of reproduction, leaf morphology, shade tolerance, or differential interactions with soil biota)||Rejmánek & Richardson (1996), Kolar & Lodge (2001), Grotkopp et al. (2002), Wolfe & Klironomos (2005), Wu et al. (2005), Duyck et al. (2007), Funk & Vitousek (2007) and Burns (2008)|
|Do species similarities affect invasion resistance?||Coexistence relies on species similarities, hence there should not be increasing inhibitory effects with increasing similarity||Resident species more strongly inhibit invading species that are functionally similar||Fargione et al. (2003) and Porter & Sagivnano (1990)|
|Do some invasive species affect RSA?||Impacts of different species are similar and stochastic||Some invasive species that create or destroy niches can dramatically affect diversity||Crooks (2002)|
Are high- and low-diversity communities equally susceptible to invasion? Species equivalence means that there is no functional difference between a monoculture and a very diverse community for systems of the same size (i.e. the same number of individuals; see Herben, 2005). Thus, neutrality would infer that both high- and low-diversity communities are equally susceptible (see Table 1). The relationship between species diversity and resistance to invasion is probably one of the most studied patterns in invasion ecology. The diversity resistance hypothesis argues that diverse communities are highly competitive and can resist invasion (see Elton, 1958; Kennedy et al., 2002). Two proposed mechanisms attempt to explain biodiversity as a barrier to invasion. One is resource-use complementarity, in which nearly all niches are saturated as diversity increases (Knops et al., 1999; Naeem et al., 2000). The other is sampling effect, in which increasing diversity increases the possibility that a competitively superior species will occur (Wardle, 2001; van Ruijven et al., 2003).
Field data at local spatial scales have shown a negative correlation between native and invasive species richness (e.g. Tilman, 1997; Naeem et al., 2000; Kennedy et al., 2002). Conversely, large-scale observational studies usually show a positive correlation between the richness of native and introduced species (Shea & Chesson, 2002; Stohlgren et al., 2004; Davies et al., 2005; Melbourne et al., 2007). This discrepancy is known as the invasion paradox. The negative relationship between native biodiversity and invasion at local spatial scales apparently show that communities with high species richness are less susceptible to invasion and, thus, already invoked as important evidence against the neutral theory (see Tilman, 2004; Chase, 2005). Using null models, however, Herben et al. (2004) and Fridley et al. (2004) showed that the observed negative correlation is mathematically inevitable for communities with a fixed and small number of species (as species are either ‘exotics’ or ‘natives’). The null expectation for randomly assembled communities is, thus, the observed negative relationship between native and exotic species richness (Fridley et al., 2004). At a larger scale, and solving the paradox, null models produce a positive correlation between these variables (see Fridley et al., 2004; Stark et al., 2006) and provide a strong support to neutral theory predictions (Stohlgren et al., 2003; Herben et al., 2004 but see Rejmánek, 2003).
The discrepancy between small and large spatial scale patterns can also be explained by co-varying external factors. Recent analyses indicate that positive correlations arise because both native and invader richness are correlated with spatial heterogeneity of abiotic conditions (Davies et al., 2005; Fridley et al., 2007; see also Richardson & Pyšek, 2006). The broad-scale positive relationship may be the outcome of combining data from a series of negative relationships coming from communities where species richness decreases susceptibility to invasion (Shea & Chesson, 2002). Another support to the diversity resistance hypothesis is that evidence of a negative relationship between local diversity and invasibility not only comes from field sampling data, but also from manipulative experiments. In these experiments, the explanation of an ‘inevitable’ negative relationship does not apply since there is not a constraint in the number of species but diversity was manipulated and invasibility was then measured (e.g. Kennedy et al., 2002; van Ruijven et al., 2003; Maron & Marler, 2008).
In conclusion, relationships between native and exotic species richness based on descriptive field survey data do not differ from neutral model expectations since null models generate the observed patterns. Experimental results, however, do seem to contrast with its predictions.
Is there any relationship between resource availability and invasion? The zero-sum assumption of neutral theory postulates a constant community size because of the saturation of resources. Thus, increases in resource levels would drive increases in the number of individuals that fit in the community. As regional diversity depends directly on the number of individuals (see Hubbell, 2001), increases in resource levels increase regional diversity. Consequently, for neutral theory, increases in resource levels can lead to the invasion of a new, previously absent, species (see Table 1). The niche counterpart is the fluctuating resource availability hypothesis (Davis et al., 2000), which proposes that increases in resource availability can relax competition intensity between resident species and potential invaders, which in turn leads to increased vulnerability to invasion.
As a generalization, and in concordance with both neutral and niche explanations, invaders are more commonly found in resource-rich habitats (Daehler, 2003), and experimental increases in resource availability can cause increased invasion (Davis & Pelsor, 2001 but see Maron & Marler, 2008). As disturbance creates resource opportunities, a similar rationale can be used to explain invasion in commonly disturbed or undisturbed communities (e.g. Burke & Grime, 1996; Davis, 2005) and null models can successfully explain the invasion of frequently disturbed habitats (Herben, in press).
However, invasions also (and commonly) occur in resource-poor environments and non-disturbed habitats (Funk & Vitousek, 2007; Martin et al., 2008). In these cases, invaders are thought to be better adapted to handle limiting factors, showing higher resource-use efficiency than native species (Funk & Vitousek, 2007; Martin et al., 2008). Hence, invaders are capable of invading high-resource and low-resource environments. Neutral theory can successfully explain the invasion of communities with high-resource availability, but fail to explain the invasion of low-resource communities that are better explained by invoking competitive advantage.
What makes some species more invasive than others? The neutral theory assumes that species belonging to the same functional guild are equivalent (see Table 1). Thus, the only factors that can affect species invasiveness would be the frequency of introductions and the number of individuals introduced to new communities (i.e. factors that affect the migration rate of a particular species). Those factors are, indeed, recognized as having strong effects on species invasiveness (see Levine, 2000; Kolar & Lodge, 2001; Colauti et al., 2006). The positive correlation between the number of propagules released into a new zone and invasion success (Duncan, 1997) highlight the importance of demography in invading processes. There is, nevertheless, overwhelming evidence that other traits, such as reproductive rates, forms of reproduction, adult and seed/propagule size, leaf morphology, shade tolerance, toxicity, resource-use efficiency (see Kolar & Lodge, 2001; Sutherland, 2004; Richardson & Pyšek, 2006; Funk & Vitousek, 2007; for recent reviews), or differential interactions with soil biota (Wolfe & Klironomos, 2005) strongly differentiate invading from non-invading species, even in species of the same genus. Although it is impossible to find a set of traits associated with invasiveness that applies to all species, it is possible to find traits at finer taxonomic scales or particular environments (Richardson & Pyšek, 2006). Short juvenile periods and short intervals between reproductive events, for example, can successfully discriminate invasive from non-invasive tree species of the genus Pinus (Rejmánek & Richardson, 1996; Richardson & Rejmánek, 2004).
It can be argued that patterns and processes at the species level, like the identities of the species, are beyond the scope of neutral theory. It can also be said that it is obvious that identities and traits of species involved are highly relevant, because ‘super species’ will, by definition, successfully invade communities that were neutral before the introduction. If we think of invasion processes in the context of neutral models, it becomes obvious that main neutral predictions can be contrasted. For neutral theory, in a local community with n species, the relative abundances of each species are the result of demographic processes and stochastic forces (see Hubbell, 2001). The introduction of a new species into the local community simply changes the number of species from n to n + 1. The neutral communities are thus expected to be invaded by both ‘super species’ and ‘normal species’ (i.e. demographic forces may be sufficient to achieve invasion, and thus ‘normal species’ can also be invasive) but empirical evidence shows that the success or failure of an invasion depends, at least in part, on the capacity of invading species to take over a niche (i.e. demography is important but not enough).
Some particular species can have tremendous effects on the diversity and RSA of the communities they invade (Crooks, 2002; Richardson & Pyšek, 2006; White et al., 2006). Evidence suggests that introduced species that increase habitat heterogeneity (i.e. create niches) can increase the abundance and species richness of the communities, while species that decrease heterogeneity (i.e. destroy niches) usually have the reverse effect (Crooks, 2002). This highlights the importance of the availability of niches before and after the introduction.
Do residents and invasive species actually compete? Competition is a mechanism that acts at the population level and can also be seen beyond neutral theory, but neutral theory postulates that coexistence does not rely on species differences, but rather on species similarities (i.e. diversity is the result of random sampling of a regional species pool and there is no difference in competitive abilities between species). Under neutral theory, although species are not ‘the same’, the differences between individuals of different species are not important in shaping community structure (see Hubbell, 2005). Indeed, this is the essential ‘neutral’ assumption of the theory (see Hubbell, 2001). The vision of a community as an assemblage where there are ecological roles and differences between species that are important in allowing them to coexist (Hutchinson, 1959) is the opposite. In this context, the negative effect of local communities on invader performance (i.e. resistance to invasion) usually depends on the existence of local species that are ecologically equivalent to the invader (i.e. communities that resist invasion already have a species very similar to the invader; see Table 1). Experimental studies on prairie grassland communities, for example, demonstrated that resident species strongly inhibit functionally similar invading species, whereas the simple removal of those local species drives invasion (Fargione et al., 2003). Moreover, the negative effects of invading species on native species are, in some cases, positively correlated with their similarity. For example, in Texas (USA), the fire ant (Solenopsis invicta) reduced native ant diversity by 70%, while overall non-ant arthropod diversity was reduced by only 30% (Porter & Sagivnano, 1990). Those patterns suggest that natural communities behave, at least in part, as assemblages where each species has some niche and species with similar niches are partially incompatible.
On the other hand, there are only a few examples where the introduction of a new species drives the extinction of a local species. Furthermore, none of these few examples is attributed to the introduction of a competitor, but all are the result of the introduction of predators or pathogens (see Sax et al., 2002; Davis, 2003). This can be interpreted as evidence that interspecific competition is unimportant in structuring natural communities, thus supporting neutral theory (see Davis, 2003).
Neutral theory explains the relative abundance of species, species-area relationships, and spatial or temporal turnover in species community composition by assuming that all species are the same. Of course, all naturalists know that species differ, and that neutral theory cannot be falsified simply by showing differences between species. The fundamental issue in the ongoing debate about neutrality in community ecology is regarding whether such differences are important in determining broad patterns of distribution and abundance in communities (see Hubbell, 2005). Some expectations of the neutral theory are indeed consistent with evidence from biological invasions. Other evidence of widespread patterns of invasion in a wide variety of environments clearly contrasts with the predictions of neutral theory apparently showing that differences between species are important. One possible caveat in favour of the neutral theory is that all empirical evidence of invasion ecology may come from non-equilibrium conditions and this theory only applies to ‘steady state’ communities (see Volkov et al., 2007; but see also Azaele et al., 2006). This restriction, however, also blocks any attempt to falsify the theory with other methodologies except for the curve fitting of descriptive data, and restarts the debate regarding whether natural communities are really in ‘steady states’ or whether they are the result of dynamic endless processes (see Levin & Paine, 1974; Wiens, 1984). Moreover, this argument can be used to refute any descriptive data that do not fit the model because the community may not be in a ‘steady state’.
Demonstrating that differences between species are important does not mean that species differences are the only important matter. The inclusion of randomness in models that attempt to explain the structure of natural communities is, without doubt, a realistic necessity to achieve a truly unified theory (see Chase, 2005). There are already some sound approximations (Tilman, 2004; Pueyo et al., 2007; Alonso et al., 2008) that by combining sampling effect with species differences (as differences in growth rates) can successfully explain patterns of invasion (see Tilman, 2004). Neutral theory is a simple theory that clearly highlights the importance of stochastic process in natural communities. However, it is far from being a valid mechanistic model because it cannot explain widespread evidence outside the few data sets that have been used to test its predictions (where nearly all models can perform as well as it does; see Pueyo et al., 2007). For environmental decision making, especially on biodiversity conservation, the implications of accepting or rejecting neutral theory as a process-based theory are very important. One direct and fundamental implication of neutral theory is that, as natural communities are far from being complex assemblages, the only requirement for achieving the successful restoration of any community that is lost, or may be lost in the future, is saving a significant amount of seeds or propagules. This implies that we should not be worried about losing (or adding) a particular species (see Lepš, 2004); increasing disturbance; decreasing habitat heterogeneity; or changing the climate, because all of these factors are just anecdotic in the simple lottery of nature. Of course, the evidence shows that this conclusion is unlikely to be true.
This project was supported by grants from the Universidad Nacional de Mar del Plata, the Fundación Antorchas and CONICET, ANPCyT, PNUD/GEF/Patagonia (B-B-01) to O.I., P.D. and J.A. were supported by Doctoral scholarship from CONICET (Argentina).