[ad_1]
MaxEnt model parameter optimization results
As detailed in Table 1, after optimizing the modulation multiplicity (RM) and feature combination (FC) of the MaxEnt model using the Kuenm package in R 4.2.0 software, the average value of the akaike information criterion (ΔAICc) was 0. The 10% omission rate (10% OR) consistently remained below 5%, and the AUC ratios (AUCratio) were all greater than 1.7, meeting the evaluation criteria for an optimal model. When using the optimized MaxEnt model to simulate suitability zones for rare Michelia species, the mean AUC for each species exceeded 0.9, with standard errors ranging from 0.007 to 0.0129. Michelia guangxiensis had the smallest mean AUC at 0.9497, while Michelia pubinervia had the largest mean AUC at 0.9986. These results indicate that the optimized MaxEnt model significantly outperforms the default parameter model in predictive accuracy and transferability, providing a more precise prediction of the distribution of rare Michelia species under both baseline (1970–2000) and future (2061–2080) climate scenarios.
Species richness fitting and centroid migration in suitability zones
As shown in Fig. 3, during the baseline period (1970–2000), the species richness of rare Michelia species was broadly distributed, encompassing approximately 25.35% of the land area of China. In northern China, only eastern Shandong Province, northern Gansu Province, and southern Shanxi Province exhibit limited distribution, while the majority of the remaining zones are predominantly located south of the Qinling Mountains-Huaihe River, a region characterized by a typical monsoon climate. Low richness is primarily distributed in southern-central Tibet, eastern-central Sichuan, southern Gansu, southern Shanxi, eastern-central Hubei, northern Hunan, northern Jiangxi, southern Anhui, eastern Shandong, southern-central Jiangsu, Shanghai, and northern Zhejiang Provinces, covering an area of approximately 110.65 × 104 km2, which accounts for about 11.53% of the land area of China. the species richness of rare Michelia species is primarily distributed in southeastern Tibet, northern Yunnan, southern-central Sichuan Province, Chongqing Province, southern-central Guizhou, western Hubei Province, western and southern Hunan Province, Jiangxi Province, southern Zhejiang Province, Fujian Province, Taiwan Province, Guangdong Province, Guangxi Province, and Hainan Province, covering an area of approximately 105.13 × 104 km2, which accounts for about 10.96% of the land area of China. Higher richness is primarily distributed in southern-central Yunnan Province, central Guangxi Province, central Guangdong Province, central Fujian Province, central Taiwan Province, and southern-central Hainan Province, covering an area of approximately 26.66 × 104 km2, which accounts for about 2.78% of the land area of China. High richness is primarily distributed in southern-central Yunnan Province, covering an area of approximately 0.82 × 104 km2, which accounts for about 0.08% of the land area of China.
Dynamics of spatial variation in species richness. Dynamics of spatial variation in species richness. Map generated using ArcGIS 10.8.1 software (https://desktop.arcgis.com/). S is the area of the species richness of the rare Michelia in China.
The significance and explanatory analysis of the spatial distribution of species richness for rare Michelia species were conducted using the Geographic Detector optimized in R software to identify the driving factors. The effects of natural ecological factors on the spatial distribution of species richness are ranked in the following order of magnitude: Minimum temperature of coldest month (q = 0.8205), Annual precipitation (q = 0.7432), Precipitation of warmest quarter (q = 0.7424), Temperature seasonality (q = 0.6031), Precipitation of driest month (q = 0.5847), Base saturation the clay in the topsoil (q = 0.5523), Percentage of the clay in the topsoil (q = 0.4344), Isothermality (q = 0.1648), Distance drom river systems (q = 0.1443), Slope (q = 0.0991), Altitude(q = 0.0918). According to the definition of q-value, among the 11 natural environmental factors, the minimum temperature of the coldest month, annual precipitation, and precipitation of the warmest quarter were the main driving factors influencing the spatial distribution of species richness.
In the future (2061–2080), under the two climate scenarios, low richness is projected to spread to the central Tibet Autonomous Region, northwestern Sichuan Province, southern Gansu Province, central Shanxi Province, Hubei Province, Henan Province, southwestern Shanxi Province, and Shandong Province, with richness expected to be maintained at 11.47–16.84% of the land area of China. Medium richness is expected to spread to zones such as eastern Sichuan Province, northern Guizhou Province, northern Hunan Province, northern Jiangxi Province, and southern Anhui Province, with richness expanding to 1.17–2.73% of the land area of China. Higher richness is projected to gradually disappear in the northern Guangxi Zhuang Autonomous Region, Fujian Province, and Taiwan Province. High richness will be maintained in southern Yunnan Province; however, it is anticipated to decrease by 0.02% under the SSP585 climate scenario.
Figures 4 and 5 show that under future climate scenarios, the potential suitability distribution zones for most rare Michelia species will migrate northwest and southwest within China, while a few will shift to the northeast and southeast. Furthermore, with the increase in CO2 emission concentrations, the potential suitability zones are projected to migrate even farther. Among them, the migration distances of species such as M. gioi, M. guangdongensis, M. velutina, M. guangxiensis, M. iteophylla, M. martinii, M. odora, M. fujianensis, M. wilsonii, M. coriacea, M. masticata, and M. szechuanica are projected to exceed 150 km under the SSP585 climate scenario.
Centroid migration of potentially suitable zones for critically endangered and endangered species of Michelia.
Centroid migration of potential suitability zones for susceptible and near-threatened species of Michelia in China.
Human interference factor calculation
As shown in Fig. 6, the Analytic Hierarchy Process (AHP) and the Entropy Weight Method (EWM) were used, respectively, to calculate the weights of the various socioeconomic factors. An optimization decision matrix is then constructed using the Oriented Distance Function (ODF) to calculate the combined weights for each factor. Subsequently, a raster calculator for superposition analysis is employed to obtain the human interference factor raster, taking into account various socioeconomic factors based on their combined weights.
Weighting values for each human disturbance factor. AHP is Analytic hierarchy process, EWM is Entropy weight method, ODF is Distance function-based combination weighting.
As shown in Fig. 7, the intensity of anthropogenic interference trends gradually increases from west to east. Zones with higher levels of anthropogenic interference are primarily located east of the Heihe-Tengchong line in China, with significant interference activities concentrated around major urban agglomerations, such as the Beijing-Tianjin-Hebei, Yangtze River Delta, Central Plains, and Chengdu-Chongqing urban agglomerations. Overall, the zoning of anthropogenic disturbance intensity in China is more consistent with the anthropogenic disturbance factor data generated from each economic factor. The conservation cost, determined by allocating costs to each sub-basin planning unit, was derived from the data on anthropogenic disturbance factors. This conservation cost was then used as the input file for the Marxan modeling procedure.
Combined human disturbance factors. Combined human disturbance factors. Map generated using ArcGIS 10.8.1 software (https://desktop.arcgis.com/).
Marxan model parameter optimization results
To conduct a sensitivity analysis, BLM values were varied within a range of 0 to 100 using a specific formula, and the planning results exhibited substantial variation across the 14 distinct BLM values employed. As illustrated in Fig. 8, BLM values ranging from 0 to 0.064 show a sharp decrease in boundary length as conservation costs increase. However, as conservation costs continue to rise, the rate of change in boundary length progressively diminishes. A distinct inflection point is observed at a BLM value of 0.064, where both the conservation cost and boundary length are relatively low. This BLM value is considered relatively optimal for the study. The optimized BLM values were used as a reference, and subsequently, the SPF values were adjusted based on the sensitivity analysis. These adjusted SPF values predominantly fell within the range of 0.1 to 0.3, all of which are lower than the default SPF values used in the Marxan model.
Variation of cost and boundary length for different BLM values.
Analysis of priority conservation zones
After conducting 100 iterative runs of the Marxan model, planning units with irreplaceability values ranging from 80 to 100 were identified, totaling 539 units. This resulted in a prioritized conservation zones of 8.27 × 104 km2, primarily concentrated in the southeastern Tibet Autonomous Region, south-central Yunnan Province, central Sichuan Province, western Chongqing Province, southern Guizhou Province, northern Guangxi Zhuang Autonomous Region, and southern Hunan Province. Additional zones include northern Guangdong Province, eastern and southern Jiangxi Province, northwestern Fujian Province, southern Zhejiang Province, central Taiwan Province, and southwestern Hainan Province. This conserved zone constitutes only 0.86% of of the land area of China. Overlay analysis of the priority conservation areas identified by the Marxan model with the existing protection system found that the priority conservation areas cover an area of 1.67 × 104 km2 within nature reserves and national parks, and that 79.8% of the priority conservation areas are still not protected by nature reserves and national parks.
As shown in Fig. 9, the Priority conservation zones for rare Michelia species covers an area of approximately 0.56 × 104 km2 within the nature reserves. Among them, the zones conserved by national nature reserves primarily include the Yarlung Tsangpo Grand Canyon Nature Reserve in the Tibet Autonomous Region, the Xishuangbanna National Nature Reserve in Yunnan Province, the Maolan National Nature Reserve in Guizhou Province, the Wuyi Mountain National Nature Reserve in Jiangxi Province, the Longqishan National Nature Reserve in Fujian Province, the Matou Mountain National Nature Reserve in Jiangxi Province, and the Shaoguan Danxia Mountain National Nature Reserve in Guangdong Province, among others. Those zones primarily conserved by provincial and municipal nature reserves include the Guilin Ocean Mountain Autonomous Nature Reserve in the Guangxi Autonomous Region, the Qiandongnan Moon Mountain Nature Reserve in Guizhou Province, the Hezhou Xiling Mountain Nature Reserve in Guangxi Autonomous Region, the Jiangjin Sifang Mountain Nature Reserve in Chongqing Municipality, the Tongbiguan Nature Reserve in Yunnan Province, and the Qiandongnan Libo Jialiang Sanlian Cave Nature Reserve in Guizhou Province, among others.
Priority conservation zones within nature reserves. Priority conservation zones within nature reserves. Map generated using ArcGIS 10.8.1 software (https://desktop.arcgis.com/).
As shown in Fig. 10, Priority Conservation zones for rare Michelia species covers an area of approximately 1.11 × 104 km2 within the parks. Among them, the zones conserved by national parks primarily include the Tropical Rainforest National Park in Hainan Province, the Danda Wildlife Important Habitat in Taiwan Province, the Wuyi Mountain National Park in Taiwan Province, the Yushan National Park in Taiwan Province, the Dongjiang Lake National Wetland Park in Hunan Province, and the Giant Panda National Park in Sichuan Province. The largest zones conserved by provincial and municipal parks primarily include the Heyuan Wanlvhu Forest Nature Park in Guangdong Province, the Mengla Yiwu Forest Nature Park in Yunnan Province, the Shaoguan Nanxiong Zhugui Meiguan Forest Nature Park in Guangdong Province, the Qingyuan Tianhu Forest Nature Park in Guangdong Province, the Shaoguan Renhua Forest Nature Park in Guangdong Province, and the Ningde Mindong Grand Canyon Forest Nature Park in Fujian Province.
Priority conservation zones within parks. Priority conservation zones within parks. Map generated using ArcGIS 10.8.1 software (https://desktop.arcgis.com/).
[ad_2]
Source link
