海洋生物分类与系统演化实验室知识库

熊本牡蛎Crassostrea sikamea（Amemiya, 1928）隶属牡蛎科（Ostreidae）、巨蛎属（Crassostrea），为我国南方的原生物种。熊本牡蛎与长牡蛎Crassostrea gigas（Thunberg, 1793）等经济品种生殖周期不同，且在繁殖期肉质依旧可口细腻，因此能够填补因繁殖所导致的夏季牡蛎市场空缺，具有很大的养殖潜力。受全球气候变化的影响，目前熊本牡蛎的分布范围北缘已成功跨越了位于30-31°N的长江口生物地理屏障，逐步扩展到位于33-34°N的苏北生物地理屏障附近。物种分布的迅速变化将会不可避免地影响到物种所处新地区物种群落的组成结构及种间关系，进而影响生物多样性和生态系统稳定性。为探究未来气候变化下熊本牡蛎于我国近海的整体分布格局和迁移趋势，本研究基于课题组近20年在我国近海大量的熊本牡蛎采样数据，采用biomod2平台构建了9种单一物种分布模型和集合模型，评估比较了不同模型的精度差异，在最优模型基础上预测当前和四种气候情景（RCP2.6、RCP4.5、RCP6.0和 RCP8.5）下未来2050s和2100s熊本牡蛎在我国近海的潜在分布情况，分析了影响其地理分布的重要环境变量，量化了其未来的气候生态位动态变化。本研究结果对我国熊本牡蛎种质资源保护和利用具有潜在的指导意义，还可以为研究全球气候变化对潮间带其他物种的适应性机制提供参考。

本研究的主要结论如下：

（1）采用AUC、TSS和Kappa三种模型评估指标对9种物种分布模型进行评价后，选取了三种模型评估指标皆为“极好”（AUC≥0.95、TSS≥0.90且Kappa≥0.80）的单一模型（包括ANN、CTA、FDA、GBM、GLM和MARS）构建了集合模型。集合模型构建完成后，得到AUC值为0.99，TSS值为0.95，Kappa值为0.80，三种模型评估指标皆为“极好”，集合模型精度较单一模型显著提高。通过构建最优集合模型能够提高物种分布模型的预测精度，使得预测结果更加准确可靠。

（2）熊本牡蛎当前在我国沿海岸线连续广泛分布，南起海南省和北部湾沿岸、北至江苏省连云港市沿岸。预计未来我国沿岸熊本牡蛎在大致保持其当前气候情景下的适生区的同时，将随温室气体排放水平的升高不断向北扩张适生区，覆盖山东省至辽宁省沿岸的大片区域，北部适生区内的适生情况会逐渐变好。适生区中部和南部将保持相对稳定。适生区南端不会有明显的纬度迁移，大致稳定在海南省和广东省沿岸，但南部适生区内的适生情况会随着温室气体排放水平的升高逐渐下降。未来一个世纪内，我国近海的熊本牡蛎分布范围极有可能会越过苏北生物地理屏障，并不断迁入北方沿海地区。熊本牡蛎作为潮间带物种中向北迁移的先驱物种，其分布范围的显著变化很可能预示着未来将会发生一场巨大的南方物种向北迁徙活动。

（3）离岸距离是熊本牡蛎建模过程中重要性最大的环境变量，这说明离岸距离这一变量在限定熊本牡蛎适生区为潮间带范围时具有特殊性和重要性。影响熊本牡蛎地理分布的环境变量主要是温度，当前情境下熊本牡蛎的平均温度生态位约在23-26℃，未来熊本牡蛎会向更低温的环境（约16-20℃）扩张生态位，同时原先23-26℃的生态位将略向更高温的环境（约26-28℃）漂移。盐度和酸碱度的重要性大致相同，且未来熊本牡蛎预计保持与当前大体上一致的最大盐度生态位和酸碱度生态位。熊本牡蛎在适应未来气候变化的过程中，能够保持与当前相对稳定的生态位特征，但是可能会占据与如今显著不同的生态位空间。

Kumamoto oyster Crassostrea sikamea (Amemiya, 1928), belonging to the family Ostreidae and the genus Crassostrea, is a natural species in southern China. The Crassostrea sikamea has a different reproductive cycle from that of economic species such as Crassostrea gigas (Thunberg, 1793), and has great potential for cultivation because it can fill the gap in the summer oyster market due to reproduction, as the meat quality is still tasty and delicate during the reproductive period. Due to global climate change, the northern edge of its distribution range has successfully crossed the Yangtze (Changjiang) River Estuary Biogeographical Barrier at 30-31°N, and has gradually expanded to the vicinity of the Subei biogeographical barrier at 33-34°N. The rapid changes in species distribution will inevitably affect the composition and interspecific relationships of species communities in the new areas where the species are located, which in turn will affect biodiversity and ecosystem stability. In order to investigate the overall distribution pattern and migration trend of Crassostrea sikamea in the offshore of China under the future climate change, the present study, based on the large amount of Crassostrea sikamea sampling data in the offshore of China in the past 20 years, constructed nine single-species distribution models and ensemble model by using the biomod2 platform, and evaluated and compared the differences in the accuracy of the different models. Based on the optimal model, we predicted the potential distribution of Crassostrea sikamea in the offshore of China in the 2050s and 2100s under the current and four climate scenarios (RCP2.6, RCP4.5, RCP6.0, and RCP8.5), and analyzed the important environmental variables affecting its geographic distribution and quantified the dynamic changes of its future climatic ecological niche. The results of this study have potential guiding significance for the conservation and utilization of Crassostrea sikamea germplasm resources in China, and can also provide a reference for the study of the adaptive mechanism of global climate change on other species in the intertidal zone.

The main conclusions of this study are as follows:

(1) After evaluating nine species distribution models using three model assessment indicators, namely AUC, TSS and Kappa, single models (including ANN, CTA, FDA, GBM, GLM, and MARS) with "excellent" (AUC ≥ 0.95, TSS ≥ 0.90, and Kappa ≥ 0.80) on all three model assessment indicators were selected to construct an ensemble model. After the construction of the ensemble model, the AUC value is 0.99, the TSS value is 0.95, and the Kappa value is 0.80, which significantly improves the accuracy of the model compared with the single model. The optimal ensemble model can improve the prediction accuracy of the species distribution model to a certain extent, and the prediction results have higher accuracy and better reliability.

(2) Crassostrea sikamea is widely distributed along the coastline of China, from the coast of Hainan Province and Beibu Gulf in the south to the coast of Lianyungang City in Jiangsu Province in the north. It is anticipated that in the future, the suitable habitat of Crassostrea sikamea along the coast of China will generally remain within their current climatic conditions. However, with the increasing levels of greenhouse gas emissions, the suitable habitat will continuously expand northward, covering a large area from Shandong Province to Liaoning Province along the coast, and the suitability within the northern habitat will gradually improve. The central and southern parts of the zone will remain relatively stable. There will be no significant latitudinal shift at the southern end of the zone, which will be roughly stable along the coasts of Hainan Province and Guangdong Province, but the suitability within the southern zone will gradually decline with the increasing levels of greenhouse gas emissions. It is highly likely that the distribution range of Crassostrea sikamea offshore China will cross the Subei biogeographical barrier and keep migrating into the northern coastal areas in the coming century. As a pioneer intertidal species that migrates northward, the significant change in the distribution range of Crassostrea sikamea may indicate that a huge migration of southern species to the north will occur in the future.

(3) Offshore distance is the most important environmental variable in the modeling process of Crassostrea sikamea, which indicates that the variable of offshore distance is special and important in limiting the suitable area of Crassostrea sikamea to the intertidal zone. The main environmental variable affecting the geographic distribution of Crassostrea sikamea is temperature. The average temperature ecological niche of Crassostrea sikamea in the current situation is about 23-26℃, and in the future Crassostrea sikamea will expand its ecological niche to cooler environments (about 16-20℃), while the ecological niche of the former 23-26℃ will drift slightly to hotter environments (about 26-28℃). The importance of Salinity and pH is roughly equal, and in the future Crassostrea sikamea is expected to maintain a maximum salinity ecological niche and a pH ecological niche that are roughly consistent with the current ones. Crassostrea sikamea will be able to maintain relatively stable ecological niche characteristics as compared to the current ones as it adapts to future climate change, but it may occupy significantly different niche space under future climatic conditions than it does today.

第1章 绪论     1

1.1 研究背景及意义      1

1.2 国内外研究进展      3

1.2.1 气候变化对物种分布的影响    3

1.2.2 物种分布模型       4

1.2.3 物种分布区中的生态位动态量化   6

1.3 研究内容及技术路线     7

1.3.1 研究内容 7

1.3.2 技术路线 8

第2章 数据获取及预处理  10

2.1 物种分布数据的获取和预处理   10

2.1.1 物种分布数据的获取  10

2.1.2 物种分布数据的预处理     10

2.2 环境变量数据的获取和和筛选   11

2.2.1 环境变量数据的获取  11

2.2.2 环境变量数据的筛选  13

2.3 本章小结    17

第3章 模型构建及评估      18

3.1 biomod2程序包 18

3.2 模型构建    20

3.2.1 单一模型构建       20

3.2.2 集合模型构建       21

3.3 模型评估    21

3.3.1 模型评估指标       21

3.3.2 模型性能评估       22

3.4 本章小结    23

第4章 气候变化下熊本牡蛎的适生区分布预测 24

4.1 用于构建集合模型的单一模型适生区分布预测   24

4.1.1 基于ANN模型的适生区分布预测 24

4.1.2 基于CTA模型的适生区分布预测 26

4.1.3 基于FDA模型的适生区分布预测 28

4.1.4 基于GBM模型的适生区分布预测       29

4.1.5 基于GLM模型的适生区分布预测 30

4.1.6 基于MARS模型的适生区分布预测     32

4.2 集合模型适生区分布预测    34

4.3 环境变量重要性分析     37

4.4 本章小结    38

第5章 气候变化下熊本牡蛎的生态位动态量化 40

5.1 气候生态位的动态量化分析       40

5.2 气候变化下熊本牡蛎的气候生态位动态量化特征      41

5.3 影响熊本牡蛎地理分布的重要环境变量的生态位变化     44

5.4 本章小结    46

第6章 讨论与结论       47

6.1 讨论     47

6.1.1 影响熊本牡蛎地理分布的重要因素      47

6.1.2 各模型对熊本牡蛎适生区分布预测结果的异同       48

6.1.3 气候变化下熊本牡蛎在我国近海整体分布格局和迁移趋势 49

6.1.4 气候变化下熊本牡蛎在我国近海的生态位动态变化      50

6.1.5 熊本牡蛎迁入北方带来的养殖机遇和生态挑战       51

6.2 结论     52

6.3 不足与展望       53

参考文献    55

附录     73

致谢     79

作者简历及攻读学位期间发表的学术论文与其他相关学术成果   81