Deadwood stochasticity

Stochasticity is one of the principal processes of community assembly. However, experimental studies rarely target stochasticity in natural communities, and hence experimental validation of stochasticity estimates in observational studies is lacking. Here we combine experimental and observational data to unravel the role of stochasticity in the assembly of wood-inhabiting fungi. We carried out a replicated field experiment where natural colonization of a focal fungal species was simulated through inoculation, and local fungal communities were monitored through DNA barcoding before and after inoculation. The results revealed that the amount of stochasticity in fungal colonization is less pronounced than expected from the amount of unpredictability in observational data, suggesting that stochasticity may play a smaller role in fungal occurrence than anticipated earlier, or it may be more related to dispersal and establishment phases than colonization per se. Our results confirmed the expectation that stochasticity is more prominent in the initial phase of community succession, with the earliest successional stage one year after inoculation involving a higher level of stochasticity than the later successional stage after two years. We conclude that experimentally measuring the amount of stochasticity in community assembly processes is feasible even for species-rich communities under natural conditions and highlight the importance of experimentally testing the accuracy of stochasticity estimates based on observational data.