||Studies on changes of vegetation cover of pasturelands in the eastern Alai Valley, Kyrgyzstan [an abstract of entire text]
Kyrgyzstan became an independent country after the collapse of the Soviet Union in 1991. This resulted in great socio-economic changes occurred in the country. The Alai Valley, located in the southern margin of Kyrgyzstan, is characterized by a long history of pastoral practice. However, no detailed research about the socio-political influences on the grazing activities and vegetation cover of pasturelands was conducted in the valley. Moreover, understanding future ecosystem response to climate change is important. This study examined the influence of socio-political and climate factors on vegetation cover of the pasturelands in the eastern Alai Valley after the independence of Kyrgyzstan in 1991, which is the period with a drastic socio-political change. First, seasonal pasture use (seasonality) of the pasturelands in the eastern Alai Valley were surveyed by interviews. As a result, three types of the pasture-use seasonality was identified: (1) spring and autumn pastures (SA-type pastures), (2) summer pastures (S- type pastures), and (3) all-season pastures (A-type pastures). Normalized Difference Vegetation Index (NDVI) values of Landsat imageries (acquired on June 2, 1990 and June 25, 2013) were calculated and were represented as the situation of vegetation cover of pasturelands at the end of the Soviet era and two decades after the independence. According to the threshold method of NDVI values, the land use/land cover was classified into three types: (1) bare and sparsely vegetated area (NDVI≤0.2), (2) moderately vegetated area (0.5≥NDVI>0.2), and (3) densely vegetated area (NDVI>0.5). The result on the changes of the pasture resource from 1990 to 2013 shows that densely vegetated area increased by 20.8%, while bare and sparsely vegetated area and moderately vegetated area decreased by 5.0% and 15.8%, respectively. It is indicated that vegetation cover has generally recovered since 1991 in the eastern Alai Valley due to the great decrease in the livestock number. Moderate Resolution Imaging Spectroradiometer Enhanced Vegetation Index (MODIS EVI) time series data were used to identify the dynamics of vegetation cover of the pasturelands from 2000 to 2013. A seasonal-trend decomposition procedure (STL) was applied to decompose the time series data of precipitation, air temperature and EVI into seasonal component, trend component and remainder component. A relationship between the trend component of climate variables and the trend component of EVI indicates that variation of the vegetation cover was more related to precipitation than air temperature. The 8-month lagged effect of the trend component of precipitation on the trend component of EVI suggests that snowfall in winter has an influence on the vegetation cover in summer. The lag effect of precipitation also implies that the growth of the vegetation does not respond to the precipitation instantly but to the history of soil moisture. The MODIS EVI time series analysis shows that the general trend of the vegetation cover increased from 2000 to 2013. However, in details, the trend of the vegetation cover decreased from 2005 to 2008 due to the decrease in the trend of precipitation. The grazing model (Howard and Higgins 1987) was used to assess the grazing intensity. Summer pastures (S-type pastures) own the highest grazing intensity due to the highest percentage of the overgrazed slopes (53.8%) and the lowest percentage of the slopes without grazing steps (34.6%). A-type pastures have the lowest grazing intensity, and are characterized by the lowest percentage of the overgrazed slopes (29.5%) and the highest percentage of the slopes without grazing steps (60.0%). Vegetation cover mainly decreased on S-type pastures (2.1%) and near the main river across the entire area. This result may be due to higher grazing intensity in these areas. The analysis of the changes of vegetation cover and grazing intensity suggests that summer grazing (S-type pasture) exerts a more negative influence on the sustainability of the vegetation cover on the south-facing slopes. Although the vegetation cover of the pasturelands has recovered after the country’s independence, some species were identified as unpalatable. Moreover, the species that are regarded as the indicator of overgrazing exist in the study area (Artemisia sp., Trisetum sp. and Polygonum viviparum). The results of the prediction of a future scenario of the dynamics of vegetation cover in 2043 based on the Markov Chain analysis and CA-Markov analysis imply that the decrease in precipitation in the future may lead to the degradation of the vegetation cover. Local people are suggested to graze their livestock on the flat, unused area south to the study area to mitigate the grazing pressure of summer grazing activities on the south-facing slopes.
Hokkaido University（北海道大学）. 博士(環境科学)