Macrophytes play a key role in stabilizing clear-water conditions in shallow freshwater ecosystems. Their populations are maintained by a balance between plant grazing and plant growth. As a freshwater snail commonly found in shallow lakes, Radix swinhoei can affect the growth of submerged macrophytes by removing epiphyton from the surface of aquatic plants and by grazing directly on macrophyte organs. Thus, we conducted a long-term (11-month) experiment to explore the effects of snail density on macrophytes with distinctive structures in an out-door clear-water mesocosm system (with relatively low total nitrogen (TN, 0.66 ± 0.27 mg L^-1) and total phosphorus (TP, 36 ± 20 μg L^-1) and a phytoplankton chlorophyll a (Chla) range of 14.8 ± 4.9 μg L^-1) based on two different snail densities (low and high) and four macrophyte species treatments (Myriophyllum spicatum, Potamogeton wrightii, P. crispus, and P. oxyphyllus). In the high-density treatment, snail biomass and abundance (36.5 ± 16.5 g m^-2 and 169 ± 92 ind m^-2, respectively) were approximately twice that observed in the low-density treatment, resulting in lower total and aboveground biomass and ramet number in the macrophytes. In addition, plant height and plant volume inhabited (PVI) showed species-specific responses to snail densities, i.e., the height of P. oxyphyllus and PVI of M. spicatum were both higher under low-density treatment. Thus, compared to low-density treatment, the inhibitory effects of long-term high snail density on macrophytes by direct feeding may be greater than the positive effects resulting from epiphyton clearance when under clear-water conditions with low epiphyton biomass. Thus, under clear-water conditions, the growth and community composition of submerged macrophytes could be potentially modified by the manual addition of invertebrates (i.e., snails) to lakes if the inhibitory effects from predatory fish are minor.

The experiment lasted from May 2016 to April 2017. After one-month pre-cultivation, water samples were collected to determine initial total nitrogen (TN), total phosphorus (TP), and phytoplankton chlorophyll a (Chla) content before SN treatment. During the experiment, sampling was conducted in June, August, September, November, December, February, and April. During each sampling event, dissolved oxygen (DO) and light attenuation coefficients (Kd) were first determined. DO was measured using a ProODO Optical Dissolved Oxygen Instrument (YSI, USA) at a depth of 30 cm and Kd was calculated by light attenuation at a depth of 30 cm using the formula in Kirk (1977) based on a Li-192 underwater quantum sensor and data-logger Li-1400 (LI-COR, USA). The outlet taps were then opened, and water samples were collected in 250-ml and 1-L plastic bottles. The 1-L water samples were used for phytoplankton Chla content, which was determined by ethanol extraction after filtering through Whatman GF/C filters (Huang et al., 1999). The 250-ml water samples were taken back to the laboratory for determination of TN, TP, pH, and alkalinity. Both TN and TP content were ascertained using spectrophotometry after digestion with K₂S₂O₈ solution (Huang et al., 1999), and alkalinity was determined by titration using 0.1 mmol L^-1 HCl. Finally, plant traits, including plant height, plant volume inhabited (PVI), ramet number, and flower number (if the plants flowered), were measured and recorded. PVI was calculated by: (average coverage) × (community plant height) / water depth.

At the end of the experiment, macrophyte and epiphyton samples were collected. The second mature leaves (for Pc and Pw) and new 10-cm long shoots (for Po and Ms) were harvested and stored in plastic bags at 4 °C for determination of epiphyton biomass (Cao et al., 2015). The remaining macrophytes were harvested and cleaned in tap water, together with the plants collected for epiphyton, and then dried at 80 °C for 48 h to determine above- and below-ground biomass. The ratio of belowground and aboveground biomass (BG/AG) was calculated. The water in the plastic tanks was then discharged, and snails in the mesocosm were collected by a 500-μm invertebrate net. Snail biomass and abundance were recorded.