[Vegetation landscape health assessment in Changshan Archipelago, North Yellow Sea].

Island vegetation is an important component of island ecosystem. Multi-targets of island ecosystem health integrated with landscape ecology theory were employed to construct the index system for island vegetation health assessment in terms of landscape vigor, landscape stressing intensity and landscape stability. The Changshan Archipelago in the North Yellow Sea was chosen as a case to apply the island vegetation health assessment index system. The results showed that the overall vegetation health status in Changshan Archipelago was good and had a big island variation. The vegetation health index for Haiyang Island and Zhangzi Island was above 0.80, belonging to first eco-health level area, whereas that for Dachangshan Island, Xiaochangshan Island and Dawangjia Island ranged from 0.70 to 0.80, which could be categorized as the second eco-health level area. Guanglu Island and Shichen Island could be termed as the third eco-health level area with the vegetation health index below 0.70. The distance of island to mainland, area of island together with industrial structure were the main driving forces for the variation of vegetation landscape heath between different islands.

[Research on remote sensing inversion biomass method based on the Suaeda salsa's measured spectrum].

Suaeda salsa is one of characteristic vegetation of wetlands in Northern China. By measuring the spectral data and leaf area index (LAI) of the Suaeda salsa in the ShuangTai Estuary of Liaodong Bay by the use of portable spectrometer and vegetation canopy analyzer, collecting the biomass of the Suaeda salsa samples, setting up the spectral reflectance curve of the Suaeda salsa, probing into the relationship between the vegetation index and the leaf area index of the Suaeda salsa, carrying out regression analysis of LAI and biomass and constructing the function equation, some conclusion were drawn: (1) By the end of September the spectral characteristics of Suaeda salsa show that at the red band 630 nm there is a clear reflection of the peak with a reflection rate of 12%-15%; there is a clear "red valley" configuration between 680 and 700 nm, and there is a clear "red edge" reflection rate of 25%-30% about 760 nm. (2) It was found that there is best correlation between vegetation index (SAVI and MSAVI) and LAI compared to other vegetation index in the regression analysis of the LAI and vegetation index. The correlation coefficient R2 is 0.711. By comparison of vegetation index linear regression equations, the correlation coefficient (SAVI and LADI) R2 is 0.696; the value of R2 (LAI and MSAV) is 0.695; the value of R2 (RVI) is 0.664; the value of R2(NDVI) is 0.649 and the value of R2 (PVI) is 0.466. (3) The value of correlation coefficient is low between the biomass and the vegetation indexes (RVI and NDVI) and the value of linear regression equation's R2 is 0.342 and 0.316, and the Logarithmic regression equation's R2 is 0.319 and 0.21, and the quadratic equation's R2 is 0.589 and 0.568, the value of correlation coefficient is high between the biomass and the vegetation indexes (PVI, SAVI and MSAVI), the value of linear regression equation is 0.626, 0.698 and 0.679, that of logarithmic regression equation is 0.592, 0.706 and 0.683 and that of the quadratic equation is 0.688, 0.711 and 0.683.

Ecosystem health assessment of the Jinghe River Watershed on the Huangtu Plateau.

An improved Costanza model was developed to assess the health of the Jinhe River Watershed ecosystem. The watershed is located at the center of the Huangtu Plateau in China and has suffered a severe disturbance in the last few decades. Three indicators including vigor, organization, and resilience were calculated respectively by merging ground-based observations with remotely sensed data on a watershed scale. Health indices of 12 topographic sub-watersheds were calculated using a modified Costanza formula. Health evaluated results indicated that sub-watersheds in the Huangtu mountain region were relatively healthy ecosystems with scores over 0.673. The sub-watersheds in the loess mountain and the loess gully regions, e.g., Jinghe, Heihe, and Honghe regions, scored moderately; their evaluated value ranged from 0.505 to 0.606. The two sub-watersheds in the loess gully region and all sub-watersheds in the loess hilly region scored the lowest, less than 0.50 and were considered unhealthy ecosystems. It can be argued that the loess hilly region and the loess gully regions should be in primary consideration for ecological protection and rehabilitation. This study provided a possible quantitative model for ecological planning and landscape management with respect to topographic conditions in this area.

[Empirical study on non-point sources pollution based on landscape pattern & ecological processes theory: a case of soil water loss on the Loess Plateau in China].

Non-point sources pollution is one of main pollution modes which pollutes the earth surface environment. Aimed at soil water loss (a typical non-point sources pollution problem) on the Losses Plateau in China, the paper applied a landscape patternevaluation method to twelve watersheds of Jinghe River Basin on the Loess Plateau by means of location-weighted landscape contrast index(LCI) and landscape slope index(LSI). The result showed that LSI of farm land, low density grass land, forest land and LCI responded significantly to soil erosion modulus and responded to depth of runoff, while the relationship between these landscape index and runoff variation index and erosion variation index were not statistically significant. This tell us LSI and LWLCI are good indicators of soil water loss and thus have big potential in non-point source pollution risk evaluation.

[Relationships between soil and water loss and landscape pattern on Loess Plateau].

Based on the theories of detrended canonical correspondence analysis (DCCA), 12 soil and water loss indices ("species") and 7 landscape indices ("environmental factors") were adopted to quantitatively analyze the effects of landscape pattern on the soil and water loss in Jinghe River basin on Loess Plateau. The results showed that the first four DCCA axes were significantly correlated to agricultural land ratio, landscape diversity index, forest land ratio, and landscape contagion index. The characteristics of soil and water loss changed obviously with landscape pattern gradient, e. g., Sanshuihe watershed with a forest cover larger than 65% and low landscape diversity had a big and stable runoff, but small sediment and low river sand content. The runoff generation coefficient and runoff depth as well as the sediment transport and river sand content increased with the increasing agricultural land ratio and landscape diversity and the decreasing forest land ratio. As for Honghe watershed with a very low forest cover but high agricultural land ratio (53.41%), and with a relatively high landscape diversity and complex landscape pattern, it had the highest river sand content and the biggest sediment transport, and its monthly runoff and sediment transport varied markedly. The soil and water loss in the Upper-Huan, Down-Huan and Dongchuan watersheds with smaller forest and agricultural land ratio and relatively simple landscape pattern was smaller than that of the watersheds with dominated agricultural land.