This study was carried out at 4137 m altitude in Gansu Yanchiwan National Nature Reserve where the typical cushion vegetation is dominated by Thylacospermum caespitosum (Camb.). First, we measured the cushion plant cover at the study site to control the estimation of community attributes in engineered situations where there were differences in habitat availability. A ratio of 3:17 was used to set the sample plots, i.e. 60 cushion patches and 340 open areas of the same size were chosen because the cushion cover was 15%. If open areas are considered as a habitat that is not affected by cushions, and open area and cushion patches are considered as a habitat that is affected by cushions, then the species composition, diversity, and density of each habitat can be subsequently determined. Additionally, the capacity of T. caespitosum (Camb.) to influence and maintain the plant community was quantitatively assessed by determination of the relative habitat richness (RHR) at a patch level, the landscape richness enhancement (LRE) at a landscape level, the habitat rescue potential (HRP) at a patch level, and the landscape insurance potential (LIP) at a landscape level. The results showed that cushion plants lead to the creation and maintenance of new habitat patches in landscapes that have different physicochemical conditions compared to the surrounding unmodified habitat matrix. Such a change in habitat conditions may directly affect the distribution of other species, and improve the species diversity of the community. In this study, the cushion plants resulted in the addition of new species to the landscape, changed the species composition, enhanced the species diversity within the community, and altered the abundance of some species. Among the 20 species identified at the study site, four were found exclusively in cushion patches, three in open areas, and 13 in both cushion and open areas, which indicated that the cushion plant T. caespitosum (Camb.) led to the addition of four new species to the habitat. Among the 13 species distributed in both cushion and open areas, four were present at a higher density in habitats that contained by cushion plants. When the species density, diversity, and evenness of two habitats that were or were not influenced by cushions were compared, species density and diversity were both found to be enhanced in habitats that were influenced by cushions, although species evenness was reduced. This shows that in the investigated community, some plant species were more dependent on the environment that was modified by T. caespitosum (Camb.). Therefore, increasing the number of cushion patches will lead to an increase in species richness at a landscape level. This indicates that T. caespitosum (Camb.) has the ability to improve community species richness by adding new plant species. Furthermore, T. caespitosum (Camb.) consolidated the survival of other plant species and reduced their risk of extinction. At a landscape level, the habitat diversity created by T. caespitosum (Camb.) guaranteed the maintenance of species richness and consequently reduced the risk of species loss, and thus, showing that the cushion plant T. caespitosum (Camb.) can potentially maintain species diversity.