| Pinus taiwanensis is the main constructive species of mountain vegetation in the coastal margin of southeast china. The 6400 hm2 original nature P. taiwanensis community in Daiyun Mountain National Nature Reserve is typical representative of warm coniferous forest of mid-subtropics. The objective of this study is aimed to consummate the Chinese Forest Biodiversity Monitoring Network and to provide reference for the near-natural operation and protection of P. taiwanensis plantation and biodiversity conservation. Based on 1 hm2 plot data of P. taiwanensis forest, we investigate its main population spatial patterns and associations, community species abundance patterns, spatial structure characteristics, community phylogenetic structure, spatial heterogeneity of vegetation parameter, spatial variability and patterns of environmental factors, and relationship of vegetation and environment, and main results are as follows:Distribution patterns and spatial associations of main populations at different growth stages and different storeys are studied using 0-ring statistical. Results show that the class structure of diameter at breast height (DBH) of the whole communities, populations of P. taiwanensis, Clethra delavayi and Rhododendron latoucheae all exhibit an invert J shape, and their population all show a stable growth type. P. taiwanensis, C. delavayi and R. latoucheae show aggregated distribution pattern at the scales of 0 to 16 m, and with the scale increases, their distribution pattern tend to be random or uniform. As age (seedling, sapling, medium tree to large tree) and tree height (forest canopy layer T, sub-forest layer S, shrub layer U) increase, the aggregation degree of P. taiwanensis, C. delavayi and R. latoucheae are reduced. Spatial distribution pattern types of P. taiwanensis are greatly influenced by DBH classes, and this influence is random. At most scales, spatial associations of P. taiwanensis and C. delavayi, C. delavayi and Bredia sinensis, C. delavayi and Rhododendron mucronatum, C. delavayi and Eurya groffii, and R. latoucheae and R. mucronatum are negative; P. taiwanensis and E.groffii, P. taiwanensis and B. sinensis, E.groffii and B. sinensis, E.groffii and R. mucronatum, and R. mucronatum and B. sinensis show positive spatial associations. The difference of spatial association among in P. taiwanensis, C. delavayi and R. latoucheae with different growth stages and different forest layers are obviously different. To some extent, habitat heterogeneity, density dependence and species isolation can explain the spatial pattern of P. taiwanensis forest community.The species abundance at different scales of P. taiwanensis community is investigated using AIC test and X2 test. Result show that the optimal model for quadrates with radius of 2.5 m,10 m and 20 m are preemption mode, while optimal model for quadrates with radius of 5 m,15 m and 25 m are Zipf-Mandelbrot mode. Except for log-normal mode, the other five modes pass the χ2 test while quadrates with radius of 2.5 m and 5 m; the χ2 test refuse modes except for neutral model, preemption mode, brokenstick mode, and Zipf mode while quadrates with radius of 10 m; χ2 test results show that quadrates with radius of 15 m,20 m and 25 m can not be fitted by log-normal mode、Zipf-Mandelbrot mode, and preemption mode, respectively.To solve the problem of biased estimation originated from select four neighbor tree (n= 4), we quantify the stand spatial structure base on Voronoi map. Result show that mingling degree, neighborhood comparison, uniform angle index, open degree, and competition indexes of P. taiwanensis forest are 0.3796, 0.5304,0.5102,0.3921 and 1.4851, respectively. Each stand spatial structure indexes has an obviously character as the DBH classes increase, and the rules vary among in different stand spatial structure indexes. It is worth mentioning that that a new competition indexes (V_ u_a_CI,) base on voronoi map and intersection angle is proposed as we study the spatial competition among in woods.Phylogenetic structure of P. taiwanensis community at different scale plots (5 x 5 m,10 x 10 m,20 x 20 m,30 x 30 m, and 40 x 40 m) and different tree size class (small diameter:1 cm≤DBH<5 cm, medium diameter:5 cm≤DBH<10 cm, large diameter:DBH≥10 cm) are analyzed using different zero model. Results show that NRI and NTI both decrease as the scale increase under the non constrained zero model, while under the zero model constraint conditions, NRI spectrum at all scales on the structure were gathered, lineage divergence aggregation/strength changing with the scale showed no regularity; NTI show a divergent phylogenetic structure at 5×5 m scales, while it manifest aggregation phylogenetic structure in the other five scales. The NRI increase as the DBH increase at each scales, the phylogenetic exhibit divergence structure at small diameter and medium diameter, and the large diameter emerge aggregation phylogenetic structure, while the variation of NTI is contrary to NRI. In the plot, habitat filtering and density dependence may be an important species diversity maintenance mechanism of P. taiwanensis community.Spatial heterogeneity of vegetation factors and environmental factors in P. taiwanensis forest is investigated using geostatistics combined with GIS. Result show that the Pielou evenness index exhibit strong variability, soil pH and soil temperature exhibit weak variability, and other indicators all exhibit medium variability. The vegetation parameters, soil pH, hydrolyzable nitrogen, soil phosphorus and soil temperature have a moderate spatial autocorrelation pattern with nugget/sill between 25% and 75%. Soil moisture physical properties, total nitrogen, nitrate nitrogen, ammonium nitrogen, soil potassium, soil organic carbon density, and photosynthesis active radiation have a strong spatial autocorrelation pattern, which indicate that their spatial heterogeneity is dominated by structural factors. The spatial autocorrelation range of vegetation parameters, soil moisture physical properties, soil pH, soil nutrient, soil organic carbon density, soil temperature, and photosynthesis active radiation are 23.5～401.1 m,14.9～19.4 m,53 m, 11.06～244.2 m,14.55 m,93.3～115.7 m, and 22.17 m, respectively. All the indexes exhibit anisotropy, and the rules of their spatial heterogeneity are consistent with the fractal characteristics. Moran I coefficient of each index shows obvious spatial positive (or negative) correlation as the spatial scales increase. All the indexes emerge obvious mosaic distribution with zonal and patchy, the difference of area of each patchy are significant, and proportion of patchy area with maximum and minimum are small. These results can provide theoretical basis for the sampling design of soil properties and the regeneration pattern and process of seedlings in P. taiwanensis forest.Based on GIS overlay analysis, we find that the law of arbor basal area and shrub basal area response to environmental factor are similar. The arbor and shrub exhibit preference in the soil moisture, soil bulk density, capillary porosity, soil pH, soil total nitrogen, hydrolyzable nitrogen, nitrate nitrogen, ammonium nitrogen, total phosphorus, available phosphate, total potassium, available potassium and soil organic carbon density are 27.57～41.49 %,0.7773～1.1564 g/cm3,0.5914～0.6085,4.27～4.90,0.1419～0.2409 g/kg,131.14～240.85 mg/kg,1.17～2.29 g/kg,16.73～47.13 g/kg,0.0611～0.1121 g/kg,4.82～5.94 mg/kg and 6.14～8.69 mg/kg,16.18～18.84 g/kg and 19.65～24.43 g/kg,69.59～92.10 mg/kg, and 6.78～ 13.19 kg/m2. The arbor and shrub also prefer the environment in soil temperature with 17.7～18.8 (April), 23.4～26.4 (August), and in photosynthesis active radiation with 17～67 umol/(s·m2). Result of principal component analysis show that the top nine principal components can explain contribution rate of variance is 80.63%, which means that soil moisture physical properties, hydrolyzable nitrogen (AN), and vegetation parameters are more important in the P. taiwanensis forest ecosystem. Stepwise regression analysis suggests that tree height (H) is more closed to elevation (ELV), and diameter at breast height (DBH) is more closed to AN. Regression models of H and DBH are:H=-23.203+0.018 ELV; DBH= 34.003-0.02 AN, respectively. Results of RDA ordination analysis show that environmental variables can explain 45.54% species abundance variation, which include 25.97%,5.11%, and 14.46% originated from soil variables, topography variables, and both of them, and there are still 80.68% species abundance variation can not explain by environmental variables. The species diversity maintaining mechanism of P. taiwanensis community in Daiyun mountain high altitude area is niche differentiation. |
PDF Full Text Request