Disturbance events shape plant communities depending on the disturbance regime as well as on the properties of the species constituting the community. We studied a lakeshore community at Lake Constance, a pre-alpine lake in Central Europe, where disturbance by flooding has a strong impact as the water level is almost not regulated by man. The lakeshore habitats are shaped by annual summer flooding during the vegetation period of plants, with seasonal course and magnitude of water-level fluctuations varying considerably between years. By a combination of field work and modelling we examined (1) whether flooding tolerance and interactions between individuals can explain zonation and species coexistence at lakeshore, (2) how strongly invasive species are affecting habitat specific species, and (3) whether changing flooding regimes due to potential climate changes will affect species composition.

The study was based on a long-term monitoring record (23 years; 1988–2005) of presence/absence of six species along two transects in a species-poor community at Lake Constance and on detailed records of annual flooding. We succeeded in parameterizing a spatially explicit individual-based life-cycle model of community dynamics on a pattern-oriented basis. This gave results on species demography, dispersal, and interactions. The parameterized model could be verified along additional data recorded in 2011.

The results of model parameterization showed that the habitat specialists (characteristic species for the studied lakeshore community) will be able to coexist with stable zonation over the coming decades, mainly due to a contrast between flood tolerance and strength of competitive interactions forming a variant of the competition–colonization trade-off model of coexistence. As a consequence, long-term shortening of the average flooding period, as predicted for the future, should negatively affect flood-tolerant but weakly competing habitat specialists. Here, an avoiding strategy, a shift of flood-tolerant plant species to lower parts of the lakeshore beyond transect boundary and hence beyond present model range, is to be expected.

The modelling results revealed that two native but invasive species will have an increasing impact on the community, and they are predicted to endanger habitat specialists on a long run. This was verified already by the re-examination in 2011. Following the parameter values, invasive species threatening the lakeshore community are both flood-tolerant and competitive. Their expected importance for the fate of the community even exceeds that of possible changes in flood duration.

The study demonstrates that an individual-based model can be developed on the basis of nonstationary, temporally and spatially changing local abundance data. Such a model goes beyond conventional matrix modelling by the inclusion of nonlinear features as a consequence of individual interactions, and such features turn out to be decisive factor for dealing with species coexistence and displacement in a plant community.