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

Titre : Analysis of spatial variability of soil and plant diversity in semi-arid lands

Yarı kurak meralarda bitki ve toprak özelliklerinin uzaysal değişkenliğinin analizi

Auteur : DİKMEN Ülkü D

Université de soutenance : Çankırı Karatekin Üniversitesi

Grade : Master 2013

Résumé
Semi-arid areas comprise a considerable portion of Turkey and the vegetation cover is very sensitive in these areas. This study was conducted in a typical semi-arid area on Çankırı Karatekin University Campus for the purpose of analyzing spatial relationship between soil properties and number of plant species (NPS) per square meter (plant diversity), which is an important indicator for desertification studies. Two transect (one normal and other parallel to the slope) was taken on north-east and another two transect were taken on south-west aspect of a small hill in the study area. Soil samples were taken from 5-m apart on each transect and NPS per square meter was determined at each sampling point. The soil samples were analyzed for physical (sand, silt, clay, bulk density, field capacity, wilting point, plant available water content, aggregation index) and chemical (pH, electrical conductivity, organic matter and CaCO3 contents). The properties depth of topsoil was also measured at the sampling. First descriptive statistics were calculated for the soil variables and NPS. The results showed that the NPS was similar in mean and coefficient of variation on all the four transects. Results of the correlation analysis among variables and NPS showed that the correlation coefficient ranged from 0 to 0.82. The correlagoram analysis of soil variables and cross-correlagrom analysis between soil variables and NPS were performed. The results of geostatistical analysis showed that topsoil depth, sand content, bulk density, plant available water content, silt content and EC spatially related to NPS and that, in general, different soil variables related to NPS on different transects. The results further showed that correlogram analysis was superior to semivarience analysis in that further could yield more precise and clear results on the distance and strength of spatial dependence and of spatial relations between soil variables and NPS. The results have a potential for use in studies of grassland development and assessment of desertification

Présentation

Version intégrale