Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Allemagne → 2018 → Effects of land use and vegetation changes on soil erosion of alpine grazing lands - Fergana Range, Southern Kyrgyzstan

Universität Hamburg (2018)

Effects of land use and vegetation changes on soil erosion of alpine grazing lands - Fergana Range, Southern Kyrgyzstan

Kulikov, Maksim

Titre : Effects of land use and vegetation changes on soil erosion of alpine grazing lands - Fergana Range, Southern Kyrgyzstan

Auswirkungen von Landnutzung und Vegetationsänderungen auf die Bodenerosion alpiner Weideflächen - Ferghanagebirge, Südkirgisistan

Auteur : Kulikov, Maksim

Université de soutenance : Universität Hamburg

Grade : Doktorgrades der Naturwissenschaften 2018

Résumé partiel
Human civilization depends greatly on ecosystems and the services they provide. These include soil, rangelands, precipitation and temperature regimes and many others. All these factors are in constant interaction and human impact can affect the balance in ecosystems. Thus, it is important to understand how and to what extent the natural resources can be sustainably used without severe consequences. The aim of this research is to assess the interactions between soil, vegetation and climatic factors and quantify them for better prediction in different utilization and climate change scenarios. We attempt to investigate the impact of existing grazing practices on rangelands, its soil and vegetation resources, as well as vegetation dependence of climatic factors and its capacity to prevent soil erosion. The study area of this Ph.D. thesis included the Fergana range in the south of Kyrgyzstan. The research focused on mountain rangelands, used as summer pastures by local population. The study area represents subhumid-semiarid mountain slopes with grazed subalpine-alpine mat vegetation. The spring season has maximum precipitation. The terrain is rugged with different slope gradients and altitudes between 2000 and 3000 m a.s.l. It was necessary to cover human, soil, vegetation and climatic factors, so the research included several aspects. Soil samples and vegetation information were collected during field trips, together with human impact data. The soil samples were analyzed in the laboratory of the Department of Physical Geography of University of Hamburg. The remotely sensed data, representing vegetation, soil and climatic factors were collected from open sources, including Landsat, SRTM, ASTER, DWD and MODIS. To understand the interactions, we applied statistical analysis of field data and remotely sensed data, modelling and development of digital maps, illustrating them. The risks of soil erosion were assessed using RUSLE (Revised Universal Soil Loss Equation) approach by calculating and assessing soil erodibility and vegetation protection factor (K-factor and C-factor) and their relations with climatic factors. For K-factor estimation soil samples were collected, representing different terrain features, they were analyzed in the laboratory for the grain size distribution and organic content, which were used for calculation of K-factor. Then K-factor map was created with universal kriging using spatially explicit terrain factors as auxiliary data. Soil features indicated their relations with terrain, which allowed to develop a scheme for prediction of soil erodibility in Fergana mountains. Vegetation cover protection or soil loss ratio (C-factor) was calculated from vegetation physical condition data, collected in the field. Then the map of C-factor was developed with universal kriging using NDVI (Normalized Difference Vegetation Index) for different months, based on annual phenology and monthly NDVI images. Climatic data, such as temperature and precipitation were collected from local weather station. Grazing pressure was assessed with interviews of shepherds in the field. C-factor indicated temporal correlation with climatic factors with time lag and spatial correlation with grazing pressure and terrain features. Modelling of vegetation and climate interrelations were done on a larger scale – for the entire country. NDVI was correlated spatially and temporarily with climatic factors as temperature and precipitation. The correlation pattern and strength of NDVI predictability with climate varied between different parts of the country. Based on this variation five clusters were developed, indicating the areas of different vegetation-climate interrelations.


Version intégrale (14,2 Mb)

Page publiée le 18 janvier 2019