Overgrazing and overtreading have caused a decrease in the diversity of species and community structure and a degradation of the soil in many places, especially near human habitation. For example, vegetation within 10 km of the Baiyinxile State Farm has severely deteriorated. With increasing grazing intensity, the average height, cover, and density of the dominant species Stipa grandis and Aneurolepidium chinense decrease sharply, whereas undesirable plants of the Artemisia frigida and Potentilla species, increase. Five distinctive stages in grazing succession have been identified: slightly grazed, moderately grazed, slightly overgrazed, overgrazed, and severely overgrazed. General trends in community dynamics caused by increased grazing have also been described. First, community productivity declines, with a decrease in plant height, cover, and density. Second, changes in species composition, including a decline in the number of palatable grasses and an increase of unpalatable species, also occur. Although changes in water content, structure, and organic matter of soil, as well as selective browsing, are believed responsible for changes in vegetation, the mechanisms behind this phenomenon remain poorly understood.
The major types of mown grasslands in Xilingele are Aneurolepidium chinense steppe and Bromus inermis meadow. Increased mowing produces effects similar to those caused by overgrazing: decrease in plant height, undesirable changes in species composition, and decline in productivity (Li et al., 1988). Both empirical and experimental data on mowing succession have accumulated in the last several years (e.g., Zhong et al., 1987, 1988; Zhong and Piao 1988). Yong (1984) and Li et al. (1988) found that left untended, abandoned agricultural lands may return to their precultivation state, although research on this phenomenon has been limited. In general, fallow-land succession passes through four stages: from fallow land to tall annual and biennial forbs (1–2 years after abandonment), to rhizome grass (2–3 years), to rhizome bunchgrass (5–10 years), and finally to bunchgrass (15–20 years). Detailed information on changes in community structure and function during old-field succession has been reported by Li et al. (1988).
Biomass Production Since the 1950s, several studies have been done on community production in the Xilin River Basin. During a survey of natural vegetation in the 1960s, the productivity of different vegetation types was measured, and based on these measurements, estimates were made of standing crops and animal carrying capacities (Li et al., 1988). The productivity of major vegetation types found on the Xilingele Station has been monitored, and some general characteristics of biomass production have begun to emerge.
Most work on community productivity has been done by researchers from the Institute of Botany and Inner Mongolia University and published in Research on Grassland Ecosystem [Caoyuan shengtai xitong yanjiu ], which is edited by the Inner Mongolia Grassland Ecosystem Research Station. Research on biomass production has included studies of qualitative and quantitative rela-