The cross-ecosystem impact of deer on an endangered submerged macrophyte, Ranunculus nipponicus var. submersus

Macrophytes are a critical component of freshwater ecosystems and are often eaten by cervids. However, the impact of cervids on macrophytes is not well known. In this study, we investigated the effect of sika deer (Cervus nippon yesoensis) on the endangered macrophyte Ranunculus nipponicus var. submersus in a spring stream in southwestern Hokkaido, Japan. We monitored the frequency of stream habitat use by deer by using sensor cameras in photography mode for four seasons. We also monitored deer feeding behavior on R. nipponicus var. submersus using sensor cameras in movie mode. To quantitatively evaluate the impact of deer on R. nipponicus var. submersus, we conducted a field experiment in which deer were excluded from part of the stream. We selected 10 pairs of adjacent patches of R. nipponicus var. submersus and set up exclosures covering one patch in each pair. We assessed the frequency of deer feeding and trampling on the control patches using the sensor cameras in photography mode and measured the mean macrophyte stem length in the exclosure and control patches every month for four seasons. To compare abiotic conditions between the exclosure and control patches, we investigated canopy openness, water depth, water temperature, electrical conductivity, pH, current velocity, and water quality at each patch during the growing season. The frequency of deer in the stream habitat was higher from spring to summer than in other seasons. Direct evidence of deer feeding behavior on R. nipponicus var. submersus was recorded using the sensor cameras. Deer often fed on and trampled on the control patches, particularly from spring to summer. The R. nipponicus var. submersus stem length was longer in the exclosure patches than in control patches (PP 0.189). Stem growth of R. nipponicus var. submersus differed among seasons (PP <0.001), and was low from winter to spring. In addition, exclosure and seasonality significantly affected stem length (PP <0.001), and the difference between the exclosure and control patches was largest in the spring. These results indicated that deer feeding and trampling negatively affected R. nipponicus var. submersus, and this effect exhibited seasonality.