Neutral theory of biodiversity (NTB) revolutionized the way we think about species co-existence. Even with its seemingly oversimplified assumptions, NTB is capable of predicting realistic species-abundance distributions, species-area curves, as well as many other macroecological patterns. Whilst most of the current research focused on these predictions and their comparison with empirical data there has been one aspect of NTB that has remained almost completely ignored – the population dynamics. Population dynamics of a species within a neutral model has most frequently been described as neutral drift which is supposed to be a random walk of the abundances of species caused by demographic stochasticity in birth, death and dispersal. In our study we ask if the neutral drift is random walk sensu stricto, or whether it has some non-random properties that are typical for natural populations. We simulated population time series in simple, spatially implicit neutral model and measured rate of density dependence, ‘redness’ of spectra and slopes of Taylor’s power law. We also explored the influence of NTB parameters (size of local communities, number of local communities, mortality rate, migration and mutation rates) on these properties.

Results/Conclusions

We showed that, in our neutral model, population time series can produce realistic slopes of the Taylor’s power law, can have spectral properties similar to natural populations (pink spectra) and seem to be more frequently density-dependent (stable) than expected in strict random walk. We detected higher rate of non-randomness in local communities than in meta-communities. We attribute the observed non-random behavior to: (1) the presence of implicit upper limit of number of individuals in a community, (2) the presence of static constraints of migration rates, (3) high mortality rates eliminating rare species end enabling dominant species to reach the community ceiling and (4) low mortality rates enabling rare species to maintain constant population sizes.