Abstract
Growth patterns of aquatic macrophytes have been shown to vary in response to hydrological properties; however, such properties are typically characterized by water level fluctuation, flow velocity, flooding season, and sedimentation, but not by water exchange rate (WER). Herein, we experimentally investigated how WER (three levels: exchange 0%, 20%, and 40% of total water per day) affects water and sediment properties, and the consequences that these variations have on the individual responses of two submerged macrophytes, Hydrilla verticillata and Myriophyllum aquaticum which were planted in two different sediment types (sand and clay). In the experiment without ramets, it was found that turbidity, pH value, and dissolved carbon dioxide concentration of the system water were statistically unaffected by WER, while water dissolved oxygen (DO) concentration and sediment oxidation-reduction potential (ORP, in both sediments) consistently increased with increasing WER, regardless of experimental time. In the experiment containing ramets, biomass accumulation and relative growth rate (RGR) of both species gradually increased with increasing WER regardless of sediment type. The mechanisms were related to (a) increased oxygen availability, as indicated by gradually increased water DO concentration and sediment ORP; and (b) enhanced phosphorus (P) and nitrogen (N) absorbing abilities associated with stimulated root growth, reflected in increased mean root length, specific root length, and the root/above-ground biomass ratio, with increasing WER. Additionally, in the experiments containing ramets, significant linear relationships were consistently detected between sediment ORP and root parameters, root parameters and plant nutrients (N and P), and plant nutrients and plant growth conditions (biomass accumulation and RGR). These results demonstrate that WER plays an important role in determining oxygen availability and thus impacts the growth of submerged macrophytes by altering the ability of roots to absorb nutrients, indicating that ecosystem functions are more sensitive to WER than previously recognized.