Soil quality cannot be improved after thirty years of land use change from forest to rangeland.

Soil quality can be assessed by measuring its physical, chemical and biological properties. In terrestrial ecosystems, the knowledge of the status of soil quality under different land use/cover can increase our understanding of processes related to soil functioning and help to properly managing ecosystems and increase their services. Conversion of the forest to rangelands is one of the most common forms of land use change having a significant effect on soil quality indicators. Here, we addressed the following objectives: (ii) to study the current status of soil physical, chemical and biological characteristics after more than thirty years of land use change from forest (dominated by Carpinus betulus and Parrotia persica) to rangeland, and (ii) to provide an overview of the spatial distributions of soil properties in forest and rangeland covers using a geostatistical method. For this, two sites (i.e., forest and rangeland) were selected in northern Iran. Within each site, 50 soil samples were collected at 0-10 cm depth along two sampling lines (250 m length) with a total of 100 soil samples for each site. Results showed that following the change of land use from forest to rangeland soil porosity, aggregate stability, pH, electrical conductivity and nutrient (i.e., total N and available P, K, Ca and Mg) contents increased, whereas soil bulk density and C/N ratio decreased. In addition, the population of soil biota (i.e., earthworms, acarina, collembola, nematode, protozoa, bacteria and fungi), microbial and enzyme activities decreased after more than thirty years of land use change from forest to rangeland. Principal component analysis confirmed that forest site had a more fertile soil and a higher biological activity than rangeland cover. Based on heat plots of soil properties, forest ecosystems created hot spots of soil quality indicators in the study area. Based on the geostatistical approach, most of the soil variables in the rangeland site followed a linear model, while in the forest site, most models were exponential and spherical. The fractal dimension values of the soil properties in the forest (1.62-1.99) had larger variations than in the rangeland (1.75-1.99) site. As a general conclusion, soil quality was not improved after more than thirty years of land use change from forest to rangeland, suggesting that degraded forest habitats should be restored by native tree species rather than converted to other land uses.