OBJECTIVETo study the impact of water level changes on the distribution of snails around lakes, and to study the possibility of forecasting the distribution of snails according to changes of water level reacted by remote sensing images.

METHODSAll information including water level information of Xingzi Hydrometric Station from 1997 to 2007, twelve TM images of Poyang Lake between 1998 and 2006 were collected. The water areas of Poyang Lake was extracted by MNDWI threshold approach, and the relationships between water areas from TM images and water level from Hydrometric Station was analyzed; Majiawan and Menqiantan marshland of Zhuxi village, Xingzi county was chosen as study pilot to set frames and investigate snails 2991 frames in 2007, and the information of snails of this pilot from 1998 to 2006 was collected to analyze the relationships between snails and water level.

RESULTSThe water areas of Poyang Lake extracted from Remote Sensing images from 1998 to 2006 was 1658.73 - 5873.48 km(2), and the corresponding daily water level of Xingzi Hydrometric Station was 9.96 - 21.72 m, according to correlation analysis, they were positive correlation relationship, the correlation coefficient was 0.96 (P < 0.01), and build a model: S = 5189.96 x ln(h) - 10 284.97; the percent of frames with living snail of zhuxi pilot from 1998 - 2007 was 28.33% - 77.46%, the corresponding monthly average water level of last August was 13.19 - 21.96 m, the corresponding highest water level of last year was 15.97 - 21.96 m, according to correlation analysis. The percent of frames with living snail of zhuxi pilot was negatively correlated with the corresponding monthly average water level of last August and the corresponding highest water level of last year, and the correlation coefficient was -0.69 (P < 0.05), and -0.77 (P < 0.01) respectively.

CONCLUSIONWater areas of Poyang Lake extracted from remote sensing images could well express the actual changes of water level, and the highest water level ahead of the year investigating snails influence the distribution of snails, thus we can estimate the distribution of snails around lakes by remote sensing images.