Does competitive neutrality occur?

One perspective on the structure of communities that rose to prominence about 20 years ago is the idea that many communities consist of ecologically equivalent competitors that persist for long time periods only undergoing random changes in population size.

This idea was not accepted in the last millennium; Armstrong and McGehee's influential 1980 article on coexistence actually defined such competitive neutrality as exclusion under the assumption that the ultimate outcome of a single-species system would be reached relatively quickly. However, Hubbell (2001) promoted the proposition that most rainforest trees (and probably other sets of species) are competitively equivalent. He argued that species-abundance distributions provided strong evidence of this, and that speciation could balance the relatively high rate of extinction that such a system would likely exhibit. However, the number of species on a given trophic level is influenced by all of their interspecific interactions, as well as the available species pool, and temporal fluctuations in abundance. It is therefore difficult to assign a single mechanism underlying a distribution of abundances. As with most generalized statistics in large biological communities, we know very little about the factors influencing the spectrum of abundances expected in a complete community or in segments thereof. The statistical distribution of species abundances is likely influenced by many factors other than competition (Abrams 2001b).

Leibold and McPeek (2006) and Velland (2010, 2016; Velland et al. 2014), among others, have supported aspects of Hubbell's argument in a more qualified manner. However, there is still no conclusive evidence for neutral coexistence in any pair of species, let alone large groups of species. On the other hand, there are many reasons for rejecting the idea. First, the accumulation of deleterious mutations occurs more rapidly in species with smaller population sizes (Lynch et al.

1995a, b). This provides an automatic advantage to the most common in a set of otherwise equivalent competing species, and would convert neutrality to a case of exclusion with a priority effect.

Secondly, species-specific diseases, particularly sexually transmitted ones, provide a stabilizing force that by itself would convert neutrality to coexistence. Third, sexual behaviours themselves can cause density-dependent mortality or density-dependent reduction in fertility (Gomez-Llano et al. 2021), again converting neutrality to stable coexistence. Fourth, adaptive evolution occurs at different rates depending on population size, and, even when it produces convergence, it in general is not expected to equalize competitive abilities (Pasztor et al. 2020). While some of these forces act in opposition, there is no reason for them to exactly cancel out at exactly the same population size in two or more species. A variety of differences in spatial distribution and differences in associations with mutualists provide other possible mechanisms producing coexistence among species having no differences in ‘resource use', when the latter is defined narrowly. Although evolution may favour convergence of resource use in some systems, it is not expected to give rise to exact equality of competitive abilities.

McPeek and Gomulkiewicz (2005) and Leibold and McPeek (2006) cite a relatively small number of sets of species having similar resource use in their sympathetic reviews of the idea of neutral coexistence. However, for none of these is the full life cycle studied, nor are spatial differences in species distribution considered thoroughly. Even Pasztor et al. (2020), who criticize the idea of neutrality, seem to agree with McPeek and Gomulkiewicz (2005) in their assertion that, ‘The vast number of co-occurring, seemingly identical, ecologically equivalent sister and cryptic species provide empirical grounds for challenging conventional ecological wisdom’. (They meant the wisdom that neutrality was unlikely.) If the number of such sets of species were really ‘vast’, and if other mechanisms affecting competition (e.g., sexually transmitted diseases, sexual selection and/or conflict, and hidden spatial resources) had been studied and ruled out, the cases cited might indicate high levels of competition.

Neither of these circumstances applies to our current knowledge.

It is important to note that McPeek and Siepielski (2019) and Ousterhout et al. (2019) have significantly revised the interpretation of coexistence in Enallagma damselflies, one of the primary examples of possible neutral coexistence in McPeek’s earlier work. The more recent studies argue that interspecific effects tended to be weak and/or asymmetrical, and that at least some species exhibit much stronger intraspecific than interspecific competition. Pasztor et al. (2020) also end up concluding that neutrality is unlikely. Leibold et al. (2019) have recently suggested that the appearance of neutrality at large spatial scales actually represents the outcome of adaptive evolution of niche-partitioning at local scales.

In other recent work, Gomez-Llano et al. (2021), and McPeekand Siepielski (2019) argued that reproductive interactions could result in intraspecific density dependence, and that this might explain cases of coexistence with high overlap in consumed resources. Reproductive competition is one of a large number of potential ‘limiting factors’ whose importance in coexistence was stressed by Levin (1970), and was something I regarded as common knowledge when I noted their potential role in character displacement 35 years ago (Abrams 1986a, p.111). Gomez-Llano et al. (2021) are correct in arguing that the details of several of the various mechanisms of sex-related density dependence have received relatively little study and deserve more attention. However, they do not support an argument for neutrality.

Neutrality would greatly simplify competitive relationships if it occurred. This may be in part responsible for the popularity of the idea. Another simplifying assumption in most competition theory (including many examples in this book) is symmetry (roughly that the effect of species 1 on species 2 is similar to the reverse effect of 2 on 1.) In many cases, symmetry of competitive interactions in simple models is required to obtain an analytical solution for the equilibrium.

In other cases, it is required so that the analytical solution is sufficiently simple to understand. In any case, the reality is that there exists no theoretical reason to expect symmetry, and even the earliest reviews of empirical work revealed that a large proportion of relationships are asymmetrical. More than 80% of the studies reviewed by Schoener (1983) that both found competition and examined the issue of symmetry found that interactions were asymmetric.

Asymmetry has many potential effects beyond this simple inequality of the effects between a pair of species. Abrams and Cortez (2015a) examined asymmetrical cases of the widely used 2-consumer-2-resource MacArthur model with the addition of asymmetric Lotka-Volterra competition between the two resources. Asymmetrical resource competition, as expected, caused asymmetrical interactions between the consumers. However, it also frequently changed the sign of some or all of the commonly used measures of inter-consumer effects, and frequently produced cyclic dynamics in the two-consumer system in cases where the single-consumer subsystems are both stable and the analogous symmetric case is stable.

7.5

<< | >>

↑

Source: Abrams Peter A.. Competition Theory in Ecology. Oxford University Press,2022. — 336 p.. 2022

Does competitive neutrality occur?

More on the topic Does competitive neutrality occur?: