LAUSR

It remains an open question whether or not artificial macrophytes are good alternatives to natural macrophytes in studies of periphyton abundance and composition in lakes. Here, a mesocosm experiment was conducted in winter (when plant growth is low) to compare simultaneously the periphyton community on three submerged macrophytes (Potamogeton lucens, Vallisneria sp. and Cabomba caroliniana) with contrasting leaf structural complexities (leaf fractal dimension = 1.12, 1.17 and 1.37, respectively) and on three types of artificial macrophytes with similar morphologies as the natural plants. We also compared intertreatment differences in phytoplankton sampled from mesocosms. Both for natural and artificial macrophytes, the periphyton chlorophyll a (Chl-a) was positively associated with leaf fractal dimension. Although the morphological structure of natural and artificial plants and the physicochemical characteristics of the water were similar, the periphyton community differed between natural and artificial macrophytes, with the difference being dependent on the leaf structural complexity of the macrophytes. For leaves with a simple structural complexity, the abundance and composition of periphyton on natural and artificial plants were not statistically different. In addition, periphyton Chl-a, density and biovolume were higher on the adaxial side than on the abaxial side of natural P. lucens leaves, but no differences were found between sides of the artificial leaves. For leaves with a medium structural complexity, the abundance of periphyton was lower on the natural than artificial plants, and the proportion of diatoms to the total community differed. For leaves with a high structural complexity, periphyton Chl-a of the artificial plants was notably higher than on the natural plants, while no significant differences were found for periphyton density, biovolume, and the proportion of diatoms and green algae. Permutational multivariate analysis of periphyton genus composition confirmed that periphyton composition on the artificial plants (medium and high leaf structural complexities) was different overall from that on the natural plants. Phytoplankton Chl-a, density, biovolume, and diversity did not show any pronounced differences among treatments. Our results suggest that artificial macrophytes cannot fully substitute for natural plants even when they are morphologically similar. Artificial macrophytes should therefore be used with caution when investigating the periphyton community on macrophytes.