| In order to carry out the health evaluation of grassland ecosystems on a large scale and achieve real-time and rapid monitoring targets,in this study,the ground-measured combined with remote sensing data under different grazing intensities of Yak in the alpine meadows of the Qinghai-Tibet Plateau(Maqian County and Maqu County)from 2016-2018 were used to conduct a health evaluation of the study area,based on the previous studies.The results are as follows:(1)From 2016-2018,there was a significant negative correlation between grazing intensity and aboveground biomass(P<0.01),while Normalized Difference Vegetation Index(NDVI)positively correlated with aboveground biomass(P<0.01),the fitting between NDVI and aboveground biomass exhibited the same tendency during 3 years.With the increase of NDVI,the grazing intensity increased first and then decreased,and its peak appeared at the middle level of NDVI.(2)The Shannon index was decreased with increasing of the grazing intensity(P<0.05).The varied the Pielou index of Maqin area was not significant under different grazing intensity from 2017-2018(P>0.05),but the area of Maqu was significant(P<0.05)from 2016-2017.The Simpson index of Maqu area increased first and then decreased with the increase of grazing intensity in 2017,and there was significantly higher than the grazing intensity of 0.67 heads/hm~2 and 7.4 heads/hm~2compare to the control,but no significant in 2016(P>0.05).From 2017-2018,the Simpson index of the Maqin region increased with the increase of grazing intensity(P<0.05),and the variation in responses of dominance degree of vegetation community to grazing intensity in different years and study areas.(3)From 2016-2018,there was significant difference in fat content(except for Maqu in 2017)(P<0.05)and the fat content increased first and then decreased when grazing intensity increased.The content of crude protein(CP)was significantly increased with increase of grazing intensity(P<0.05),while the neutral detergent fiber(NDF)and acid detergent fiber(ADF)significantly decreased(except for Maqu 2016)(P<0.05)and the ADF was significantly higher than the others compare to the control in the Maqu(2017)area and Maqin(2018)area.(4)From 2016-2017,the surface soil water content,organic carbon(SOC)and total nitrogen content were significantly decreased with increase of grazing intensity at 0-10 cm depth(P<0.05)in the Maqu area,while in Maqin area(2017-2018)showed a trend of increased first and then decreased with the increase of grazing intensity(P<0.05)and grazing intensity significantly increased the soil bulk density(P<0.05).(5)The health level of the sample plots was varied between healthy and alert.From 2016 to 2017,the health level of Maqu area(herds 2 and 3)changed from healthy to alert and the health status deteriorated.The health level of Maqin(S1 and M2)changed from alert to healthy,the health status improved and other health grades did not change in 2017 and 2018.Comparing the overall level of the comprehensive health evaluation index(CVOR)of the two study areas,the overall health status of Maqu area is better than that of Maqin area.In addition,there was a significant difference between CVOR and grazing intensities of study areas in 2016-2018(P<0.05).(6)Five estimation models based on NDVI were established to evaluate CVOR of the grasslands,the results demonstrated that NDVI has a high correlation with CVOR,the determination coefficients of the five models were higher than 0.54 and the estimation accuracy was more than 90%,which indicating that the remote sensing estimation of the comprehensive index of grassland health evaluation is feasible and the estimated results of each model are not much different. |
