Protracted speciation revitalizes the neutral theory of biodiversity

Protracted speciation revitalizes the neutral

theory of biodiversity

James Rosindell,1* Stephen J.

Cornell,1 Stephen P. Hubbell 2 and

Rampal S. Etienne3

1 Institute of Integrative and Comparative Biology, University of Leeds, Leeds LS2 9JT, UK

2 Department of Ecology and Evolutionary Biology, University of California, Los Angeles, CA 900954, USA

3 Community and Conservation Ecology Group, Centre for Ecological and Evolutionary Studies, University of Groningen, Box 14, 9750 AA Haren, The Netherlands

*Correspondence: E-mail:

james@rosindell.org

Ecology letters 2010 v.13 no.6 pp. 716-727

Abstract    

  Understanding the maintenance and origin of biodiversity is a formidable task, yet many  ubiquitous ecological patterns are predicted by a surprisingly simple and widely studied  neutral model that ignores functional differences between species. However, this model  assumes that new species arise instantaneously as singletons and consequently makes  unrealistic predictions about species lifetimes, speciation rates and number of rare  species. Here, we resolve these anomalies – without compromising any of the original  model’s existing achievements and retaining computational and analytical tractability – by  modelling speciation as a gradual, protracted, process rather than an instantaneous event.  Our model also makes new predictions about the diversity of incipient species and rare  species in the metacommunity. We show that it is both necessary and straightforward to  incorporate protracted speciation in future studies of neutral models, and argue that nonneutral  models should also model speciation as a gradual process rather than an  instantaneous one.  

Keywords:  Incipient species, log-normal, log-series, neutral model, neutral theory, speciation, species abundance, species longevity.

Figure 3 Species abundance distributions (SADs) for the log-series and difference log-series (DLS). Left panel: histogram of numbers of species in each abundance class for JM¼109 and l¼0.000001 predicted by the log-series distribution (red line) and the DLS with a variety of different transition times s¼0, 1, 10, 100, 1000, 10 000 (bars with darker colours corresponding to larger s and s¼10 000 in green). The jth abundance class is defined as containing species with j£ abundance <2j. As s increases, there are dramatically fewer rare species, but the more abundant species follow the log-series distribution very closely. The remaining panels show the same data for the log-series (red line) and DLS with s¼10 000 (green line) as a plot of log abundance vs. rank in abundance. The whole metacommunity is shown in the centre, where the DLS is clearly distinct from the log-series. A random sample of 0.01% of individuals is plotted on the right showing that after sampling it is nearly impossible to observe the difference between the DLS and log-series.

Figure 4 Top left panel: a fit to coral reef metacommunity data (shown in blue) using the standard point mutation model with h¼19.85 (red) and protracted speciation model with h¼20.13 and s ¢ ¼2.23·10)6 (green). These parameters represent the global optimum in parameter space calculated with eqn 4. The two black lines represent 1 SD on either side of the mean prediction under protracted speciation, produced by analysing 4 000 000 simulations of the model (using algorithms described in Appendix S1). It can clearly be seen that both models fit the abundance data equally well: we increased the thickness of the red line otherwise it would be completely hidden behind the green line. The top centre and top right panels show the species lifetimes (in generations) and speciation rates (per species per generation), respectively, and were plotted using the same parameter values as those used for the top left panel. The lower set of panels show the same results for a reef fish metacommunity sample for which the best fitting h was 15.82 under point mutation (red) and h¼16.21, s ¢ ¼3.77·10)5 under protracted speciation (green). 

Figure 5 Local community species abundances compared with the regional species abundances and empirical data. Top left panel: a metacommunity abundance distribution with JM¼1010, h¼100 under different modes of speciation. Point mutation speciation (red bars), protracted speciation with s¼100 (green line) protracted speciation with s¼10 000 (black line). Top centre and right panels show dispersal limited local community samples m¼0.1 (top centre) and m¼0.7 (top right) for all three metacommunities. It is clear that in these local community samples we no longer see the very significant differences in the metacommunities from which they are drawn. A fit to locally sampled species local community abundance data for trees on BCI (Condit 1998; Hubbell et al. 1999, 2005) is shown in the bottom left panel. The parameters used were h¼47.67 and m¼0.093 for point mutation (red) and h¼47.96, s¢¼8.15·10)8 and m ¼ 0.093 for protracted speciation (green). Blue bars show empirical data and the two black lines represent the 1 SD on either side of the mean of the protracted speciation result, calculated by simulation as in Fig. 4. The bottom centre and bottom right panels show speciation rate and species lifetime for the same optimal parameters as used to fit the species abundances in the bottom left panel.

Full Text

Download     Click here   Or    Click here

Read and download                Click here