Modes of speciation and the neutral theory of biodiversity

Abstract

Hubbell's neutral theory of biodiversity has generated much debate over the need for niches to explain biodiversity patterns. Discussion of the theory has focused on its neutrality assumption, i.e. the functional equivalence of species in competition and dispersal. Almost no attention has been paid to another critical aspect of the theory, the assumptions on the nature of the speciation process. In the standard version of the neutral theory each individual has a fixed probability to speciate. Hence, the speciation rate of a species is directly proportional to its abundance in the metacommunity. We argue that this assumption is not realistic for most speciation modes because speciation is an emergent property of complex processes at larger spatial and temporal scales and, consequently, speciation rate can either increase or decrease with abundance. Accordingly, the assumption that speciation rate is independent of abundance (each species has a fixed probability to speciate) is a more natural starting point in a neutral theory of biodiversity. Here we present a neutral model based on this assumption and we confront this new model to 20 large data sets of tree communities, expecting the new model to fit the data better than Hubbell's original model. We find, however, that the data sets are much better fitted by Hubbell's original model. This implies that species abundance data can discriminate between different modes of speciation, or, stated otherwise, that the mode of speciation has a large impact on the species abundance distribution. Our model analysis points out new ways to study how biodiversity patterns are shaped by the interplay between evolutionary processes (speciation, extinction) and ecological processes (competition, dispersal). S n (n 0 ; J)0 X

Speciation-rate dependence in species–area relationships

“…In particular, the assumptions of a point speciation mode (i.e. the fact that each individual has a fixed probability to speciate) is known to be crucial and the results may change dramatically when considering "fission modes", corresponding to allopatric speciation (Hubbell, 2001(Hubbell, , 2003Ricklefs, 2003;Etienne et al, 2007a). Indeed, it seems like the most important assumption of neutral models is that new species enter the system with a population of a single individual (Zillio and Condit, 2007).…”

The Unified Neutral Theory of Biodiversity and Biogeography at Age Ten

“…Etienne and Haegeman [34] recently noted that this produces identical results to MacArthur's broken stick model [35], which was motivated by random apportioning of limiting resources to niches that would then dictate relative species abundances. Surprisingly, point mutation speciation fits empirical data much better than do random fission and a variety of other alternatives [34,36,37]. This suggests that actual species may arise as fairly small populations [38].…”

“…The likelihood of each generated data set is then compared with the likelihood of the real data; if the likelihood of this empirical data is around the median of the artificially generated likelihoods, the fit is good. Similar procedures have since been developed for more complex versions of the model, such as when multiple local communities are linked to the same metacommunity [53,55] or when other speciation mechanisms are considered [34,36,37,98].…”

“…For example, Tomasovych and Kidwell recently used the neutral theory framework, calibrated with data from living taxa, to investigate the effects of time averaging on fossil data [75]. Alternative modes of speciation can be investigated where speciation rates depend on species abundance rather than on species richness [36]. Furthermore, spatially explicit coalescence methods have the potential to investigate the effects of variation in sample size over time and space at large scales, including the possibility of accounting for preservation bias.…”

Why abundant tropical tree species are phylogenetically old

“…In others, species randomly split into two new species with a probability proportional to population size (1,5,6). NTB models are remarkably successful in predicting distributions of species abundances, particularly with the assumption of mutation speciation (4,7,8). Although the assumption of ecological neutrality has been continuously challenged (9)(10)(11)(12)(13), the predictions of NTB have been shown to be robust to the existence of niches if species diversity is sufficiently high (2,14,15), because random drift can still occur between the relative abundances of species within the same niche or between species that share very similar niches (16,17).…”