产漂流性卵小型鱼类的生态位建模及分析

发布时间：2019-05-29 06:39

【摘要】：针对近年来美国产漂流性卵小型鱼类濒临灭绝的问题,本文以具有代表性的阿肯色河流域鲤鱼(Arkansas River Shiner, Notropis girardi,以下简称ARS)为研究对象,开展了产漂流性卵小型鱼类的生态位建模及分析研究,旨在为河流生态系统的保护和濒危鱼类生境恢复提供参考依据。本文在美国地质调查局(United States Geological Survey)的"Biotic and Abiotic Factors Related to Declining Distribution of a Federally Threatened Cyprinid"课题资助下完成,具体研究内容如下：(1)提高生态位模型转移能力的方法研究以生态位模型MAXENT为例,基于投影概念和Parolo等提出的计算模型转移前后预测结果的相关性理论,定义了模型的转移能力系数：“预测”能力系数和“追溯”能力系数。通过分析四种常规方法：基于环境变量间Spearman's rho相关性系数的变量筛选方法、基于环境变量方差膨胀因子VIF的变量筛选方法、基于主成分分析PCA的环境变量重组方法和调节模型自适应参数"Regularization Multiplier"方法,提出了提高生态位模型转移能力的最佳方法。研究结果表明：四种常规方法均能有效提高模型的转移能力,其中,增加"Regularization Multiplier"值提高效果最明显,“预测”能力系数提高了67%,“追溯”能力系数提高了74%,但"Regularization Multiplier"值过高会影响模型的模拟效果(自适应性)。而利用调节"Regularization Multiplier"与变量相关性分析方法相结合的方式,可以在保证生态位模型自适应性的基础上,达到模型转移能力提高的最佳效果。(2)ARS栖息地适宜性分析方法研究基于地理信息系统(GIS)和遥感技术(RS),研究收集自1950年至2010年可能影响阿肯色河流域ARS潜在分布的自然环境数据,包括生物气候数据、地形地貌数据、坡度高程数据、河流流态数据和外来入侵物种数据等；计算人为活动诸如主要河流周边的点源、地下水开采井及拦水大坝的分布数据。基于MAXENT生态位模型,针对不同时期ARS的分布情况进行建模,旨在分析自然环境和人为活动对ARS潜在分布的影响,确定濒危鱼类对象适宜的生存环境。研究结果表明：河流流态(年平均流量和河流等级)及人为活动(地下水开采因子和点源因子)对不同时期MAXENT生态位模型贡献率较大(5%),即对ARS的潜在栖息地分布的影响较大。不同时期ARS适生曲线与各因子的反应曲线形状不同,主要原因在于两个时期ARS的分布数量与范围相差较大。建立大尺度流域环境空间和大跨度时间范围内的栖息地适应性模型,从宏观角度上确定了特定鱼类对象适宜的生存环境。(3)ARS繁殖期关键生态水文因子识别方法研究研究选取分布于整个流域40个水文站点自1950年至2010年的日径流数据,分析产漂流性卵小型鱼类繁殖期特有的水文条件需求,给出了3个生态水文因子：繁殖期高流量、繁殖期高流量历时与繁殖期断流历时。将生态水文因子分别与其他环境因子相结合,构建了不同时期的MAXENT生态位模型,发现大型水利工程建设时期之前和之后的流域自然流态变化较大。研究结果表明：繁殖期高流量对生态位模型的贡献率最高(历史时期65.8%,当前时期55.7%),进一步证明了繁殖期高流量对于ARS产卵期的影响。此外,研究得到的历史时期ARS对高流量的适应范围,有助于量化形成河流生态目标,指导实施河流生态系统的重建或调控操作。(4)气候变化情形下ARS的潜在生境动态分析方法研究研究利用IPCC AR5发布的最新气候情景数据,分析生物气候因子的生态学意义及其相关性,优选5个气候因子,模拟CCSM4.0共同体气候系统模式下2030年、2050年和2070年阿肯色河流域ARS潜在适宜生境的变化规律。针对阿肯色河流域东西区域的气候分布不均匀的情况,结合Palmer干旱强度指数(PDSI),提出按照最小外包络矩形分为东西两个子区域建模分析方法,通过对比子区域模型与整体区域模型预测结果的增加率(Gain)和丢失率(Loss),可验证模型的预测精度。研究结果表明：增加率比(Gain Rate)和丢失率比(Loss Rate)范围均在1±0.3内,表明子区域模型和整体模型预测结果具有一致性,模型预测的结果较为准确。通过绘制2030年、2050年和2070年三个时期ARS潜在适宜生境的变化趋势,发现ARS的地理分布格局受温度变化方差和最热月份最高温度变化的影响,将整体向西部的加拿大河上游河段偏移,东部区域除锡马龙河和阿肯色河的交汇处有增加趋势外,其他河段的分布概率将越来越小。研究方法有利于降低未来气候变化对濒危物种的影响。
[Abstract]:In order to solve the problem of the extinction of the small-scale fish in the USA in recent years, the ecological niche modeling and analysis of the small fish with rafting eggs are carried out in the paper by Arkanas River Siner, Norois girardi (hereinafter referred to as" ARS "). The aim of this paper is to provide a reference for the conservation of the river ecosystem and the restoration of the endangered fish habitat. In this paper, under the aid of the "Biotic and Abiotic Factors Related to Declining Distribution of a Federally Threatened Cyprinid" of the United States Geological Survey, the research contents are as follows: (1) The method of improving the transfer ability of the niche model is studied with the ecological niche model MAXENT as an example. Based on the concept of projection and the correlation theory of the prediction results before and after transfer, the transfer capability coefficient of the model is defined: the "forecast" capacity coefficient and the "retroactively" capacity coefficient. The invention provides a variable screening method based on a Spearman's rho correlation coefficient between environment variables, a variable screening method based on an environment variable variance expansion factor VIF, an environment variable recombination method based on the principal component analysis PCA, and an adaptive parameter "Regularization Multiplier" method for adjusting the model, The best way to improve the transfer ability of niche model is put forward. The results show that four conventional methods can effectively improve the transfer ability of the model, in which, the increase of the "Regularization Multiplier" value is the most obvious, the "forecast" capacity coefficient is improved by 67%, the "retroactively" capacity coefficient is improved by 74%, But the high "Regularization Multiplier" value can affect the simulation effect (self-adaptability) of the model. By means of the combination of the adjusting "Regularization Multiplier" and the variable correlation analysis method, the best effect of improving the model transfer capability can be achieved on the basis of ensuring the adaptability of the ecological niche model. (2) The research on the suitability analysis method of the ARS habitat is based on the geographic information system (GIS) and the remote sensing technology (RS), and the data collected from 1950 to 2010 may affect the potential distribution of ARS in the river basin, including the biological and climatic data and the topographic and geomorphological data. Slope elevation data, river flow pattern data and alien invasive species data, etc.; the distribution data of point sources, groundwater exploitation wells and water retaining dams, such as major river perimeters, are calculated. Based on the MAXENT niche model, the distribution of ARS in different periods is modeled. The purpose of this paper is to analyze the influence of natural environment and man-made activities on the potential distribution of ARS, and to determine the appropriate living environment of the endangered fish. The results show that the flow regime of the river (annual average flow and river level) and man-made activities (groundwater exploitation factors and point source factors) have a great contribution to the model of the MAXENT niche model in different periods (5%), that is, the influence of the potential habitat distribution of ARS is large. The response curve of ARS in different period is different from that of each factor. The main reason is that the distribution quantity of ARS is different from the range. The environment space and the habitat adaptability model of the large-scale basin are set up, and the appropriate living environment of the specific fish object is determined from the macroscopic angle. (3) The research and study on the key ecological hydrological factors of the ARS propagation period selected the daily runoff data from 1950 to 2010 in the 40 hydrological stations of the whole basin, and analyzed the special hydrological condition demand of the small-scale fish breeding period, and gave three ecological hydrological factors: The high-flow rate, the high-flow duration of the breeding period and the duration of the breeding period. The ecological hydrological factors are combined with other environmental factors to construct the MAXENT niche model in different periods, and the natural flow pattern of the basin before and after the construction of the large-scale water conservancy project is found to be great. The results showed that the contribution rate of high-flow rate to niche model was the highest in the breeding period (65.8% in the historical period and 55.7% in the current period), and the effect of high-flow rate on the egg-producing period of ARS was further proved. In addition, the high-flow adaptation range of ARS is helpful to quantify the formation of river ecological targets and to guide the implementation of the reconstruction or control operation of the river ecosystem. (4) The research and research on the potential habitat dynamic analysis of ARS in the case of climate change, using the latest climate scene data published by the IPCC AR5, and analyzing the ecological significance and the correlation of the bioclimate factors, preferably 5 climate factors, The changes of the potential suitable habitat for ARS in the Meihe River Basin,2030,2050 and 2070, are simulated by the CCSM4.0 community climate system model. In view of the non-uniform distribution of the climate in the east and west of the river basin, combined with the Palmer's drought intensity index (PDSI), the method of modeling and analyzing two sub-regions according to the minimum outer envelope rectangle is proposed. The prediction accuracy of the model can be verified by comparing the increase rate (Gain) and the loss rate (Loss) of the sub-region model and the overall region model prediction result. The results show that the increase rate ratio (Gain Rate) and the loss rate ratio (LosRate) range are within 1-0.3, indicating that the sub-region model and the overall model prediction result are consistent, and the result of the model prediction is more accurate. By plotting the change trend of the potential suitable habitat for ARS in the three periods of 2030,2050 and 2070, the distribution pattern of ARS is found to be affected by the variation of temperature variation and the highest temperature in the hottest month, and the whole is shifted to the upstream reach of the Canadian river in the west. In the eastern region, the distribution probability of other reaches will be smaller and smaller in addition to the increasing trend of the junction of the Maolong River and the East River. The research approach is beneficial to reducing the impact of future climate change on endangered species.
【学位授予单位】：中国农业大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：S931.1

【相似文献】