Studies On The Relationship Between Macro Ecological Changes Of Marsh And Climate In The Yellow River Source Region

The response of the regional ecosystem to global climate change has been a main theme of global change research.Alpine wetland ecosystem has become one of the research hotspots,probably due to its sensitivity to climate change.In the northeast of the Tibetan Plateau and the marginal area of East Asian monsoon,the influence of the climate change(especially the warming)on alpine wetland of the Yellow River source region has attracted the attention of the academic community.Previous studies focused on the wetland area,landscape pattern and its relationship with climate.However,there was little concern about the changes in marsh ecosystem,one of key wetland types in the region,and its relationship with climate.Also,many researchers failed to recognize the climate differentiation,and its possible influence on the response of the alpine marsh to climate change.Marsh of the Yellow River source region served as the study object,and the characteristics of spatial-temporal variation,such as area,vegetation growing season characteristics and NDVI,were analyzed at both regional(source region)and local scale(three representative climatic units are Marduo,Dari and Ruoergai from the northwest to the southeast).The Mixed Generalized Additive models(GAMMs)was applied to simulate relationships between area and climatic factors,NDVI of growth season and climatic factors to assess spatial influence of climatic factors.NDVI anomaly of marsh vegetation growing season under the influence of gley environment,its correlation with climatic factors and their differences in different climatic space units were analyzed with the grassland NDVI as baseline NDVI under the same climatic conditions.The main conclusions are as follows:(1)The annual average temperature has markedly ascended at both regional and local scale except for Henan station.The annual precipitation has increased but not significant at the regional scale.At the local scale,the annual precipitation of Marduo has increased significantly,while the other stations have increased or decreased but not significantly.However,the climate changed from the relatively cold period to relatively warm period at local scale were inconsistent,Marduo,Xinghai and Qiabuqia was 1986,Hongyuan,Jiuzi,Ruoergai and Dari was 1997.The changes of precipitation in each station were mainly fluctuant.(2)The area of marsh in the Yellow River source region showed a trend of growth from 1990 to 2010.At the local scale,the area changes of marsh were different,Maduo and Ruoergai showed a decrease,while Dari showed an increase.The relationship between area of marsh and annual precipitation,average annual temperature and annual relative humidity was non-linear.The three climatic factors had a significant impact on the area of marsh(P<0.01),but only partly explained the changes in the area of marsh(adjusted R~2=0.172).All three climatic factors had no obvious influence on their respective low value range,obvious influence on their median value range(positive correlation)and high value range(negative correlation).(3)The multi-year averaged growing season of marsh vegetation in the Yellow River source region(2000-2013)was from late April to mid-October.The beginning of growing season was delayed half a month respectively from Ruoergai to Dari and then to Maduo,but the end of the growing season was the same.The average NDVI in each year and years(2000-2013)were as follows:Ruoergai,Dari and Maduo,which were consistent with the climate spatial differentiation of the source region.The relationship between the NDVI of the marsh vegetation growing season and the precipitation,average temperature during the same period was non-linear.The influence of two climatic factors on NDVI of the growing season was very significant(P<0.01),which represented the main influencing factors of NDVI in the growing season(adjusted R~2=0.716).The precipitation of the growing season was positively correlated with NDVI,and was obviously influence(positive correlation)on the200-500mm range.The influence of mean temperature was more complex,which was obviously influence on its low and high value ranges(negatively correlated)and 7-11?range(positively correlated),especially 7-9.5?range.(4)Under the effect of gleization,NDVI anomaly of marsh vegetation showed a growth trend at the beginning and later periods of the growth season,a decrease trend in the middle of the growth season at the regional scale.Mardo NDVI anomaly showed a decrease trend form the early of the growth season to the peak of grass vegetation NDVI,an increase trend after that.Dari NDVI anomaly showed an increase trend at the beginning of the growth season,a decrease trend from the medium period to the peak of growth season,an increase trend from the medium and later periods after the peak of NDVI.The changes of Ruoergai NDVI anomaly were not obviously.The climatic factors had obviously influence on the NDVI anomaly at the region scale,and the influence was obviously different in the different climatic units.The growing season NDVI anomaly was significantly or highly significant negative correlated with the mean temperature during the same period,which was more obviously in the northwest alpine region.The correlation between NDVI anomaly and growing season precipitation was significantly lower than that of NDVI anomaly and growing season mean temperature.(5)The regional average pattern can not be simply adopted because of the obvious difference in the beginning of the growing season,when analyzing the response of local vegetation to climate change for the Yellow River Yellow River source region and similar macroscopic region.The growing season of the local scale is calculated according to the climatic space differentiation for the same vegetation type.NDVI is calculated according to the corresponding growing season to analyze its dynamic change and response to climate change.The GAMMs model indicated that NDVI of marsh vegetation growing season in the Yellow River source region had obviously spatial differentiation in response to climatic factors.Therefore,it must be considered space differentiation under the effects of the same amplitude climate change.The response analysis should be carried out according to different climatic units,rather than simply reasoning that the climate unit will change in the same way as the regional scale does.