Cumulative Effect Of Land Use On Fish Community Structure In Different Scales And Impact Prediction Research

One of the main threats to biodiversity in freshwater ecosystems is the conversion of land use from natural forest land to non-natural types, which is mainly affect the water environmental elements in different temporal and spatial scales by influencing nutrients, water temperature, organic content, dissolved oxygen, hydrological characteristics and other factors. The impact mechanism of individual types of land use change on aquatic ecosystems is determined more clearly than the cross-impact of many types. What’s more, as lack of basis, the measurement of land-use planning negative effect on the aquatic environment only a rule of thumb. Therefore, Based on the land use data of the point upstream catchment and upstream for 10 km with 1km wide buffer, in Taizi River Basin, the key environmental factors were screened out by many tools, such as principal component analysis, canonical correspondence analysis, and so on. After that, the classification and regression tree model was attempted to analyze the scale effect and cumulative effects of land use on fish communities, while the indicator species threshold analysis was used to determine the development threshold of land use in protection of fish sensitive species and communities. Finally, the effect of land use was predicted in Bayesian network model under different scenarios. The conclusions are as follows:(1) In buffer scale, with the higher fish diversity, the cross-impact of grass land(<2.65%), forest(>15.15%) and farmland(>24.82%) is more intensity than only grass(<2.65%) and forest(<15.15%). But the minimum of fish integrity scores was appeared with the cross-impact of forest(<34.07%) and urban land(>15.45%). In catchment scale, the highest fish diversity was appeared with the cross-impact of farmland(>13.35%) and urban land(1.07%-2.54%), while the cross-impact of forest(>80.1%) and urban land(<1.28%) resulted in higher fish integrity scores.(2) In buffer scale, P.lagowskii, N.nudus, C.granoei and O.obscura negative response(threshold range of 38.86%-45.42%) to farmland interference, and C.auratus and A.chankaensis response(threshold range of 44.41%-47.32%) to farmland positively. In addition to the four kinds of fish species who negative response to farmland interference, A.liaoningensis response(threshold range of 3.43%-13.51%) to the urban land disturbance negatively, while the C.auratus, H.leucisculus, G.rivuloides and A.chankaensis showed a positively response(threshold range of 6.02%-8.43%). Meanwhile, the five species who response to urban land negatively all showed a positively response(threshold range of 22.68%-41.54%) to forest. In catchment scale, C.auratus, A.chankaensis and A.rivularis response to farmland interference positively(threshold range of 13.35%-20.82%), while response negatively to forest(threshold range of 79.03%-86.67%). P.lagowskii, N.nudus, C.granoei and O.obscura response to farmland interference negatively(threshold range at 7.69%-22.81%), while response negatively to urban land(threshold range of 1.28%-2.62%) and response positively to forest(threshold range of 73.71%-88.77%). Meanwhile, C.auratus, H.leucisculus and A.chankaensis showed a positive response(threshold range of 2.07%-3.997%) to urban land.(3) In buffer scale, as the fish community show a negative response, the development threshold of farmland, forest and urban land is 45.42%, 28.64% and 8.43%. In contrast, the farmland(threshold of 46.15%), forest(threshold of 34.08%) and urban land(threshold of 12.83%) also should be considered for fish community responding positively. In catchment scale, in consideration of the negative response of fish communities, the development threshold of farmland, forest and urban land is 14.04%, 81.37% and 2.62%. In contrast, the development threshold of farmland, forest and urban land is 13.94%, 80.48% and 2.07% in view of fish communities’ positive response.(4) The results showed that when the forest increased, but urban land and farmland decreased, the number of fish individuals, the diversity index and integrity of fish species will all increased. When the forest and urban land increased, but farmland decreased, the number of fish individuals and fish integrity will decreased. When the forest and farmland increased, but the urban land decreased, the number of fish individuals, the diversity index and fish integrity will increased. When lowering the proportion of forest, but the urban land and farmland are increased, the number of fish individuals will be declined more obviously than fish diversity index and integrity. When the proportion forest and farmland decreased, while the urban land increased, the number of fish individuals and integrity will be reduced. When the proportion of forest land and urban land decreased, while the farmland increased, only the diversity of fish individuals will be increased, while the integrity of fish decreased. As a whole, if we want to improve the survive condition of fish communities, the proportion of forest should be increased and considered in land use planning.