the city, accounting for 40.7%, 21.1%, and 28.5% of the total land, respectively. Grassland only accounted for 6.9%. Forestland was mainly distributed in the western and eastern mountains; farmland was mainly in the southern plains, with some farmland in the northern plains; construction land was mainly in the central plains; and grassland was mainly in the northeastern and northwestern mountains.
Soil sampling and chemical analysis
A total of 139 topsoil samples (to a depth of 20 cm) were collected from the agricultural lands (farmland, forestland, and grassland) in Taiyuan from April to May 2013 (Fig.2). The sampling design was based on the distribution of the local agricultural land uses, and the sampling intensity was roughly proportional to the proportion of the total area covered by each land use. iSampling sites were randomly chosen, but were located at least 200 m from roads and industrial or re
sidential areas. Near the center of each site, 5 individual sub-samples were collected from a 10 m × 10 m plot using a wood spade. The sub-samples were then bulked to obtain a single composite sample for each site. The coordinates of each sampling point were recorded using a global positioning system receiver. Information on the vegetation, soil type, and agricultural land use were also recorded. Samples from the forestland, grassland, and farmland sites accounted for 39.6%, 7.9%, and 52.5% of the samples, respectively. Approximately 1.0 kg (fresh weigh) of each soil sample was taken and mixed manually. The samples were then oven-dried at 45 ◦C for 3 d, sieved through a 2-mm plastic sieve to remove large debris (including gravel-sized materials, plant roots, and waste materials), and stored in sealed plastic bags until analysis. Soil was digested with a mixture of 5:2:3 (volume/volume/volume) HNO3-HClO4-HF (Li et al., 2012). The digested solutions were analyzed by