Fragment dispersal and plant-induced dieback explain irregular ring-shaped pattern formation in a clonal submerged macrophyte

•We formulated a deterministic mathematical model of macrophyte meadow expansion due to both clonal propagation and fragment dispersal.•The model accounts for fragment motion preferentially in the direction of the wind, and a negative feedback on macrophyte dieback due to macrophyte-induced stratification.•Fragment dispersal was predicted to be the primary mechanism for macrophyte meadow expansion.•Model predictions of central dieback, due to the stratification feedback, were consistent with field observations.•This study highlights the significance of fragment dispersal to invasive macrophyte spreading.

Submerged macrophytes can colonize shallow lakes via several reproductive mechanisms, and can in turn substantially alter these environments by modifying the thermal structure and dissolved oxygen levels within these lakes. Although multiple mechanisms of submerged macrophyte expansion have been described, the relative contribution of each of these in shallow lake environments has been largely overlooked. In this study we analyzed the spatial spread and patterning during seasonal growth of a globally invasive submerged macrophyte, Potamogeton crispus, in a shallow urban lake (Lake Monger, Western Australia). We used underwater and aerial imagery to estimate the spatial pattern of the P. crispus bed. By comparing the spatial extent of the bed at different times during the growing season, we found linear expansion rates two orders of magnitude higher than those previously estimated through rhizome elongation. We formulated a deterministic mathematical model that accounted for the ability of P. crispus to spread through rhizomes and fragments broken off by the feeding activities of aquatic birds, to assess the contribution of fragment dispersal to the emergent patterns of the submerged macrophyte bed. In addition to accounting for dispersal from fragments, the model also accounted for a hypothesized feedback between macrophyte-induced thermal stratification and central dieback. Comparison of our model results against field data indicated that the model accurately represented the spatial spread of the macrophyte bed when fragment dispersal was included. When fragment dispersal was not included in the model, the spatial spread of the bed was largely underestimated, suggesting that fragment dispersal may well account for the fast seasonal spread of this species. The model also captured the formation of a ring-shaped pattern in spatial macrophyte distribution suggesting that both fragment dispersal and the feedback between stratification and dieback are necessary to reproduce the spatial structure of the macrophyte bed. Our results highlight the potential important role of fragment dispersal in facilitating colonization and submerged macrophyte invasion in shallow lakes.