4 Discussion

The growth and distribution of species and habitat condition, including soil, climate and topography et al., are inseparable (Piao et al., 2012; Liu et al., 2014; Liao et al., 2020). In this study, there are no significant differences for the soil properties between different topographies (Table 2). However, there are significant differences for the individual growth between different topographies. The experimental plots were all located on the same alluvial fan with the same climate overall. While local modification of the climate by topography and vegetation produces microclimates at the land–air interface (Zellweger et al., 2019). Therefore, the micro-topographic condition was main constraints that affected the changes of microclimates. And hence the spatial complexity may be important for adaptations to climate change (Kooijman et al., 2019). The warming temperature and the change in precipitation have provided the positive effects of vegetation dynamics in recent decades in this region (Li et al., 2015; 2016). Thus, the unreasonable human activity, such as overgrazing, extensive reclamation and excessive collecting live plants for firewood, remain the dominant factor caused aeolian sandy land development severely in 1990s in this region (Shen et al., 2012; Li et al., 2016). Ecological restoration programs, including active restoration (i.e. planting the artificial forest and shrubs, conversion of cropland to grassland, shrub or forest, and establishing the sand-protecting barriers) and natural restoration (i.e. forbidding grazing or collecting firewood and controlling the number of livestock), have been launched and conducted successfully since 2000. In this study, the coverage for S. moorcroftiana population shows well after decades of natural restoration (Fig. 1). Population structure is the synthetic action of biological characteristics, environmental factors, and intraspecific interactions which can be used to reflect the succession dynamics and development trends of a population (Liu and Zhang, 2018). The relationship between CPA and height was fitted well by a linear model in each topography from the field survey (p < 0.01) in this study. Therefore, the plant height of S. moorcroftianapopulation was regarded as the measure of age grade. This study showed that the height structure of S. moorcroftiana population approximated a Gaussian distribution model (Fig. 3). The population structure was dominant for the seedings in 2011, however the number of individuals was dominant for the adults of 30-60 cm in 2017. Compared with the early study in 2011, the age structure of S. moorcroftiana population increased significantly to 2017. Aggregated spatial distribution patterns are commonly observed in many restored ecosystem and small individuals showed higher degree of aggregation than large individuals (Zhao et al., 2012). The present study also found that the height class of less than 90 cm (class 1, 2, 3) had higher degree of aggregation than those of more than 90 cm (class 4, 5) for each topography (Fig. 3), and this phenomenon of aggregation was observed more clearly with increasing elevation and slope.
It is commonly agreed upon in the academic community that climate change can cause alterations to the structure of communities and also cause changes in the spatial distribution of terrestrial populations (Gelviz-Gelvez et al., 2015). Therefore, the micro-topography was also an important driver of individual distribution. Revealing the spatial distribution of S. moorcroftiana population contribute to illustrate the effects of various micro-topography on the dispersal and growth of S. moorcroftiana . Our results showed that the correlations between the individual height of S. moorcroftiana population and micro-topography (elevation and slope) reached the 99% confidence level (p < 0.01), however there were obviously differences for the individual distribution on different topographies. Although the distribution of S. moorcroftiana population decreased along increasing elevation, there was a peak value at 3624 m (T2) (Fig. 6). There was similar to the overall trend of spatial distribution forS. moorcroftiana population along increasing slope, with some different changes minimally happened after approximately 20°. While theS. moorcroftiana population was mainly distributed on the southwest in T1 and T2 and on the northeast in T3. Diamond et al. (2019) reported that the spatial distribution of relative water levels is changed by the wetland micro-topography and affected individual distribution and growth. However, for sandy ecosystem, the changes of solar radiation with micro-topography also must be taken into account (Li et al., 2019). In general, the west would receive excessive solar radiation and evaporation in the afternoon, and this is not conducive to survival of plants. Therefore, there was little distribution of individuals on the west in all of micro-topography in this study. Generally speaking, the short time of solar radiation and fine condition of water have been found at the shade slope, north slope of mountain, and it was beneficial for vegetation growth. Therefore, the moving sandy land was mainly distributed on the north of Yarlung Zangbo River, while the fixed sandy land was mostly located in the south of Yarlung Zangbo River, which was the north slope of mountain. However, there were minimum density of vegetation at the north slope from the micro-topography. One conjecture, the effect of low temperature in alpine valley, was submitted for the difference between overall and micro-topography scale. Less solar radiation with lower temperature at the shade slope blocked the survival of plants. Comprehensively analyzing the structure, distribution and growth of population, we concluded that the habitat conditions on the southwest slope of approximately 20°-25° with altitudes of 3593-3643 m most favor to the natural restoration of S. moorcroftiana population on aeolian sandy land.