|其他摘要||Aim: The pattern of species richness along an environmental gradient and its potential mechanism are the study focus of biologists, especially in the recent years, the references of gamma diversity at large-scale space have increased by geometric series. Environmental variables along an elevational gradient drastically vary within a narrow spatial range, so elevational pattern of species richness receives more attention. The elevational gradient in the Hengduan Mountains (Mount Hengduan) is one of the longest bioclimatic elevational gradients in the world, from a subtropical zone to a zone of permanent frost, and there distributes rich species, so it is ideal for the analysis of elevational pattern of species richness. This paper primarily aims to (1) explore the division of the southern and northern Mount Hengduan; (2) describe the elevational patterns of seed plant richness in the Mount Hengduan; (3) describe the elevational patterns of species richness for different life forms and some selected taxa; (4) analyze the variation of species endemism along this elevational gradient; (5) analyze the relationships between species richness and area, mid-domain effect (MDE) and climatic factors; (6) estimate the species richness in special eco-environment of dry valley or elevations above timber line and its variation along this elevational gradient.
Methods: Data of species richness and distribution at elevational space came from publications, 85,482 specimens and field investigation. The Mount Hengduan region was divided into 9 latitudinal belts using 1 degree of latitude to define the latitudinal belt. The distribution of species in each latitudinal belt was recorded, and Jaccard similar index of species between adjacent latitudinal belts was calculated by EstimateS 7.51 software. The elevations from 800 to 5500 m were divided into 47 100-m elevational belts, or 16 300-m elevational belts, and species richness, area and climatic data in each 100-m or 300-m elevational belt were calculated. Species richness was estimated by interpolation and rarefaction. A digital elevational model (DEM) was used to calculate the area in each 100-m elevational belt in the region using ESRI’s Arcview 3.1. Climatic factors were obtained from 90 climate stations with a record length of 30 yr (1951-1980). The linear regression between temperature and elevation was significant and strong, so it was used to calculate temperate data for these 100-m elevational belts where no distribution of climate stations. The Kriging algorithm was used to realize spatial interpolation of rainfall using ESRI’s Arcview 3.1. RangeModel software was used to generate a null distribution predicted by the mid-domain effect. The ordinary least squares (OLS) and conditional autoregressive models (CAR) were used to correlate the relationship between species richness and explainable variables.
Results: (1) Flora of seed plants in the Mount Hengduan. There distributed 8590 seed plants, of which 2783 species were endemic to the region of the Mount Hengduan, holding 32.4% of total seed plants. There were 6070 herb species, 1743 shrub species and 777 tree species. The herb species had the highest endemism (35.5%), followed by shrub species (29.0%), and tree species had the lowest endemism (15.7%).
(2) Division of the southern and northern Mount Hengduan. Species similarity between adjacent latitudinal belts across 29ºN latitude showed a visibly low value, and latitudinal patterns of species richness, shown by the c-value of species-area power function and species / area ratio, showed a sharp decrease at the 1 degree latitudinal belt of 29.0º-29.9ºN. So, the 29ºN latitudinal curve can be as the division of the southern and northern Mount Hengduan. According to this division, the southern subregion occupied 40% of total area, but contained more than 80% of species, almost twice as much as in the northern subregion; 50% of total species, 60% of the endemic or tree species were only distributed in the southern subregion.
(3) Elevational patterns of species richness. Elevational patterns of species richness estimated by interpolation and rarefaction both presented to be unimodal, and the highest richness almost located at the same elevations. Total species had the highest richness at the elevations of 2000-3200 m in the southern subregion, and at the elevations of 2500-3800 m in the northern subregion. The elevations having the highest richness of endemic species were located at the same elevational range (3000-4000 m) between the southern and northern subregions. Furthermore, elevational patterns of species richness adjusted by area also presented to be unimodal, but the elevations with the highest richness were far lower. Among the different life forms, their elevational patterns of species richness also presented to be unimodal, and the elevations with the highest richness of herb species were higher than that of shrub and tree species.
(4) Variation of endemism along the elevational gradient. Endemism increased with ascending elevation below 4300 m, and then decreased with elevation gain. There were three patterns of endemism at generic level with increasing elevation: namely it (i) increased, (ii) decreased, or (iii) peaked at middle elevations.
(5) Relationships between elevational patterns of species richness and explainable variables. The mid-domain effect (MDE) had the strongest relationship with elevational patterns of species richness, and it can explain more than 60% of variation of species richness for most species groups, and MDE had a stronger relationship with broad-ranged species than narrow-ranged species. Area had the highest explainable ability for species richness in the southern subregion, followed by the entire region, and it had the lowest ability to explain the variation of species richness in the northern subregion. Area had a stronger relationship with endemic richness than that of non-endemic richness. Among the different life forms, the relationship between area and herb richness was significant, but it was not significant between area and shrub and tree richness of most groups. In the selected climatic factors, moisture index showed a significant and strong relationship with species richness, but mean annual rainfall or potential evaporation showed a weak relationship with species richness. The relationship between climatic factors and species richness was the highest for tree richness or non-endemic richness, followed by shrub richness or total richness, and it was the lowest for herb richness or endemic richness. The multi-relationship between explainable variables and species richness showed that all explainable variables had a powerful ability to explain the elevational patterns of species richness, and explained more than 96% of variation of non-endemic richness, and 93% of variation of total richness, and 73% of variation of endemic richness.
(6) Species richness in the special eco-environments. There distributed 1487 seed plants in the dry valley of the southern subregion, of which 124 species were endemic to the Mount Hengduan region, holding 8.3% to total species. There were 2927 seed plants in the dry valley of northern subregion, of which 568 species were endemic to the Mount Hengduan region, holding 19.4% to total species. Area and climatic factors showed a significant and strong relationship with species richness in dry valley. There distributed 1820 species in the elevations above timber line, of which 655 species were endemic to the Mount Hengduan region, holding 35.5% to total species. Area, potential evaporation and mean annual rainfall all showed a significant and strong relationship with species richness above timber line, and they together explained more than 97.6% of variation of species richness.
Main conclusions: (1) The 29ºN latitudinal curve is an important division of the southern and northern Mount Hengduan, and the southern Mount Hengduan is the core of this biodiversity hotspot. (2) The unimodal patterns of seed plant richness along the Mount Hengduan elevational gradient have been approved, and the elevational range of 3000-4000 m. has the highest richness of endemic species, and species in this elevational range have experienced the most drastic differentiation and evolution, so this elevational range is important for the biodiversity conservation. (3) At generic level, multi-patterns of endemism along the same elevational gradient show that endemism is strongly related with evolution and dispersal of species in the elevational space. (4) MDE, area and climate are the important factors to influence the elevational pattern of species richness, and they together have a strong power to explain the variation of species richness along the elevational gradient. (5) Elevational pattern of species richness in higher elevations can not be well described for the insufficiency of data, especially the intensity of plant diversity investigation and plant specimen collection are low in the elevations above timber line in the Mount Hengduan, and more investigations and collections are urgently needed.|