[Effects of land use and landscape pattern on nitrogen and phosphorus exports in Lanlingxi Watershed of the Three Gorges Reservoir Area, China].

The temporal and spatial characteristics of N, P exports and effects of land use and landscape pattern on N, P exports were analyzed in the Lanlingxi Watershed of the Three Gorges Reservoir Area. The results showed that the TN, TP and NO3(-) -N were mainly generated by non-wood forest, the N, P exports in flood period (June to September) were significantly higher than the non-flood period (January to May). The NH4(+) -N export was derived from the residential area in the non-flood period, while from non-wood forest in the flood period. In addition, the performance of samples N, P exports with forest distributed were lower in both two periods. Also, the proportion of forest significantly negatively correlated with NO3(-) -N, TP in the non-flood period and TN, TP in the flood period. The residential area proportion notably positively correlated with NO3(-) -N, TN in non-flood period and NO3(-) -N, TN, TP in the flood period. The non-wood forest proportion also significantly positively correlated with NH4(+) -N, TN in the flood period. Moreover, PD closely positively correlated with N exports in non-flood period, with NO3(-) -N, NH4(+) -N in flood period. The CONT index strongly negatively correlated with N exports in flood period and TP in non-flood period. However, the proportions of farmland, unused land and the indices of ED were relatively weakened with N, P exports in both periods, while SHMN and water proportion did not show any positive or negative correlation. Moreover, the regression fitting degree of NH4(+)-N was superior to NO3(-) -N, TN and TP with the adjust R2 of 0.885 and 0.969 in two periods, while the regression relation was better than that of non-flood period. The result of redundancy analysis further demonstrated that the landscape fragmentation caused by patches types of different land uses could better explain impacts on the exports of nitrogen and phosphorus. The two canonical axes accumulated explained the 90% proportion of the variables and the highest contribution was PD, which was an important indicator for watershed water quality assessment and prediction.