[Impact of the changing water level on the variance of Oncomelania hupensis populations in Lake Area with general additive model]
To explore the impact and regularity of snail population after changing of water level and to develop effective control and prediction programs. A typical snail habitat closed to Changjiang county in Yueyang city, Hunan province was selected as the survey field. Data on water level and the changing trend of snail population during 2001 - 2009 including the average water level, maximum and minimum water levels, snail and infected snail densities, proportion of sampling frames with living and infected snails etc. were collected. The distribution of water level and snail indexes were described and a general additive model (GAM) for the relationships between these indexes were also fitted. Impacts of the average water level and the watered-out days in current year in the surveyed field on the snail density were statistically significant (P<0.01), showing a quadratic curve association. The snail density remained at the lowest level when the average water level was at 25.0 m and the surveyed field watered-out days was 120. The average water level in current year and the field watered-out days in the past year showed statistically significant impacts on the infected snail density (P<0.001), the proportion of sampling frames with infected snails (P<0.05) and living snails (P value neared 0.05), presented a quadratic curve, a cubic curve or even a more complicated piecewise curve association. Both the infected snail density and the proportion of sampling frames with infected snails remained at the lowest level when the average water level was at 25.0 m and the watered-out days in the past year was 110, while the proportion of sampling frames with living snails kept the lowest level when the average water level was at 25.0 m and the watered-out days in the past year was 160. The water level and the field watered-out days affected the development of snail population directly. The changing water level had an impact on snail population change, which mainly presented as nonlinear smooth function relation. Impact of the field watered out days on the infected snail density and the proportion of sampling frames with living snails and infected snails showed a hysteresis effect. The snail density was predicted to be retaining a high level when the water level was 24.0 m and the field watered-out days was 3 months. It had obvious advantages to fit the relationship of the changing water level and the snail indexes with a GAM which could get closer to the reality as well as easier to find and explain the potential associations and regulations.