Solving the puzzle of yeast survival in ephemeral nectar systems: exponential growth is not enough

We here examine how sufficiently high cell densities of nectar yeast can develop in a flower. In laboratory experiments, we determined the remaining fraction of nectar yeast cells after nectar removal, and used honeybees to determine the number of transmitted yeast cells from one flower to the next. The results of these experiments directly fed into a simulation model providing an insight into movement and colonization ecology of nectar yeasts. To understand the effect of many consecutive pollination events on population size and dispersal potential we developed a stochastic simulation model (NetLogo 5.3.1; Wilensky 1999) of nectar yeasts in one single flower. The model calculates the population size and the amount of dispersed cells of a single nectar yeast population over time, dependent on: pollination time and chance, inoculated cells during first pollination event, transmitted cells to the next flower, cells that remain in the flower during pollination, nectar production rate and growth rate of yeast cells with lag phase. Modelling was done with NetLogo 5.3.1 (Wilensky 1999). The model works stochastically, exclusively with global variables without individuals or space. One time step is one hour.