黑水河中华纹胸鮡栖息地适宜性指数算法筛选及评估  

ALGORITHM SCREENING AND EVALUATION OF HABITAT SUITABILITY INDEX FOR GLYPTOTHORAX SINENSE IN THE HEISHUI RIVER

作　　者：谢伟 邓华堂[1] 蒲艳[1] 倪达富 成必新 唐锡良 陈大庆 段辛斌[1] 田辉伍[1] XIE Wei;DENG Hua-Tang;PU Yan;NI Da-Fu;CHENG Bi-Xin;TANG Xi-Liang;CHEN Da-Qing;DUAN Xin-Bin;TIAN Hui-Wu(National Agricultural Science Observing and Experimental Station of Chongqing,Yangtze River Fisheries Research Institute,Chinese Academy of Fishery Science,Wuhan 430223,China;College of Aquaculture and Life Science,Shanghai Ocean University,Shanghai 201306,China;College of Aquaculture,Southwest University,Chongqing 400715,China;Shanghai Investigation,Design and Research Institute Co.Ltd,Shanghai 200434,China;China Three Gorges Corporation,Wuhan 430010,China)

机构地区：[1]中国水产科学研究院长江水产研究所,国家农业科学重庆观测实验站,武汉430223 [2]上海海洋大学水产与生命学院,上海201306 [3]西南大学水产学院,重庆400715 [4]上海勘测设计研究院有限公司,上海200434 [5]中国长江三峡集团有限公司,武汉430010

出　　处：《水生生物学报》2025年第4期142-153,共12页Acta Hydrobiologica Sinica

基　　金：国家重点研发计划(2022YFC3202001);中国三峡建设管理有限公司项目(JG/18056B和JG/18057B);国家自然科学基金(51909271和51509262);中国水产科学研究院中央级公益性科研院所基本科研业务费专项资金(2023TD09)资助。

摘　　要：为研究中华纹胸鮡(Glyptothorax sinensis)栖息地适宜性指数(Habitat suitability index,HSI)模型最优算法,科学评估其适宜栖息地分布,基于2018—2019年黑水河中华纹胸鮡渔获物数据及同步采集的13个环境因子,采用一元非线性函数拟合构建单个环境因子SI曲线,并结合最大值法(Maximum,MAX)、最小值法(Minimum,MIN)、算术平均法(Arithmetic mean model,AMM)、几何平均法(Geometric mean model,GMM)、加权平均法(Weighted moving average,WMA)分别计算中华纹胸鮡HSI值。计算结果表明:在各模型算法中,算术平均法和加权平均法两种方法的预测结果误差最小,最大值法与最小值法的预测结果与中华纹胸鮡实际分布偏差较大,在进行算法选择时要慎重考虑。黑水河中华纹胸鮡栖息地适宜性指数总体呈现上游至下游纵向上升趋势,HSI值大于0.7的点位为下游自然河段S3和S4。水温、海拔等物理环境是驱动中华纹胸鮡栖息地空间分布差异的主要因素。算术平均法及加权平均法为黑水河中华纹胸鮡栖息地适宜性指数模型预测最优算法。研究结果可为黑水河鱼类栖息地评估提供参考资料,促进鱼类栖息地保护。Habitat assessment is a prerequisite for in-situ conservation of fish,and appropriate model algorithms is an important basis for improving the prediction accuracy of habitat assessment models.This study aims to identify the optimal algorithm for the habitat suitability index(HSI)model of Glyptothorax sinensis and to scientifically evaluate its suitable habitat distribution.We utilized fishery catch data and 13 environmental factors collected synchronously from the Heishui River from 2018 to 2019.A one-dimensional nonlinear function was employed to fit a single environmental factor curve,and HSI values of G.sinensis were calculated using the maximum value method,minimum value method,arithmetic mean model,geometric mean model,and weighted moving average.The results indicate that among the various model algorithms,the arithmetic mean model and weighted moving average model exhibited the smallest prediction error,while the maximum and minimum value method showed significant prediction error and deviated notably from the actual distribution of G.sinensis.Therefore,caution should be taken when choosing model algorithms.Overall,the HSI index of G.sinensis in the Heishui River shows an upward trend from the upstream to the downstream,with sections S3 and S4 exhibiting HSI values greater than 0.7.Water temperature and elevation are the main driving factors for the spatial distribution differences in G.sinensis habitat.The arithmetic mean model and weighted moving average model are identified as the optimal algorithm for predicting the habitat suitability index of G.sinensis in the Heishui River.The research results can provide reference for the assessment of fish habitats in the Heishui River and promote the protection of fish habitats.

关 键 词：栖息地适宜性指数 最优算法 黑水河 中华纹胸鮡 

分 类 号：Q178.1[生物学—水生生物学]