RESUMO
Uninhabited islands are important for providing isolated habitats for unique biological resources, and revealing the spatial pattern of plant diversity is of great significance for the island biodiversity conservation. A total of 15 uninhabited islands in Miaodao Archipelago, a group of typical uninhabited islands in North China, were selected as the study area. The multiple gradients at island and site scales were identified and quantified based on field investigation and remote sensing methods, and seven "from macro to micro" aspects, including morphology, proximity, landscape, terrain, atmosphere, soil, and vegetation, were selected to cover all aspects of factors influencing the plant diversity. Then, the single and comprehensive effects of the multiple gradients on the spatial pattern of plant diversity at the dual scales were analyzed using methods of regression analysis and canonical correspondence analysis ordination. Results indicated that 130 plant species were recorded. The species accumulation curves proved the sufficiency of the numbers of sampling sites and islands to represent the overall characteristics of plant diversity. The species composition on the uninhabited islands possessed common characteristics with the neighboring inhabited islands and mainland, meanwhile, showed unique features on the dominant species. The α diversity showed distinct spatial heterogeneities at the dual scales; the ß diversity indicated the great difference of species composition within an island and among different islands. At island scale, island area, vegetation condition, and terrain complexity contributed the most to the spatial pattern of plant diversity. At site scale, biodiversity indices changed irregularly along the multiple gradient factors, yet all aspects of gradients showed significant effects on the species composition and distribution. The island area played a fundamental role in determining the α diversity at island scale and generating the ß diversity within an island, however, was not significantly correlated with the diversity at site scale.
Assuntos
Biodiversidade , Ecossistema , Monitoramento Ambiental , Plantas , China , IlhasRESUMO
Based on the 1975-2010 multi-temporal remotely sensed TM and ETM images and meteorological data, in combining with wavelet analysis, trend surface simulation, and interpolation method, this paper analyzed the meteorological elements' spatial distribution and change characteristics in the middle reaches of Heihe River, and elucidated the process of wetland landscape fragmentation in the study area by using the landscape indices patch density (PD), mean patch size (MPS), and patch shape fragment index (FS). The relationships between the wetland landscape fragmentation and climate change were also approached through correlation analysis and multiple stepwise regression analysis. In 1975-2010, the overall distribution patterns of precipitation and temperature in the study area were low precipitation in high temperature regions and high precipitation in low temperature regions, and the main characteristics of climate change were the conversion from dry to wet and from cold to warm. In the whole study period, the wetland landscape fragmentation was mainly manifested in the decrease of MPS, with a decrement of 48.95 hm2, and the increase of PD, with an increment of 0.006 ind x hm(-2).