Study on Water Environment and Fish Behavior in Agricultural Canals

对于农业水渠生态系统的保护和恢复，不仅要量化水质和水力条件等有利生境特征，而且要量化鱼类或野生动物的环境偏好。本研究采用室内实验和实地考察相结合的方法，对农业渠道中日本鳉鱼（Japanese Medaka Fish，鳉鱼，Oryzias latipes）的环境偏好和水环境进行了调查。之所以选择日本水豚作为本次研究的河鱼，是因为日本水豚现在被认为是日本乡村生态系统恢复的象征鱼类。通过对近年来在农业渠道内新建的人工筑鱼窝块内及周围的水力测量，选择水深、流速和覆盖3个环境因子作为影响农业渠道鱼类栖息地可得性和质量的主要因素。基于鱼的行为本质上是不确定的，以及个体和环境对环境变化的行为反应差异很大的背景，采用模糊逻辑模型和实验室明渠实验对三种环境因素的偏好强度进行了量化。引入了一种简单的遗传算法来搜索偏好强度的最优函数表示。通过室内水箱实验和现场检验，对模糊遗传偏好强度模型进行了验证。在考虑流速和盖度两个因子时，预测分布与观测分布的一致性较好，表明这两个因子对日本稻螺的环境偏好有明显的影响。为了利用基于个体的模型方法对鱼类栖息地的水环境进行评价，本文分析了日本米卡鱼的鱼群行为，以确定个体之间的相互作用：个体之间的间距、个体的空间分布和交流过程。引入了行为模式模型和人工网络模型来描述鱼群的群集行为。模型的数值模拟结果与实验结果在极化、最近邻距离、扩散和分形维数等方面基本一致。通过计算游泳轨迹、游泳速度和游泳角度的回归图、分形维数、相关维数和小波变换，量化个体间游泳行为的差异。利用计算结果，验证了特定环境下鱼类张力或休息程度评价和水环境中化学污染物鱼类生物监测的可行性。少

For the preservation and restoration of ecosystems in agricultural canals, it is important to quantify not only the characteristics of preferable habitat such as water quality and hydraulic condition but also environmental preference of fish or wildlife. In this research, environmental preference of Japanese Medaka Fish (Japanese killifish, Oryzias latipes) and water environments in agricultural canals were investigated by laboratory experiments and on-the-spot examinations. Japanese Medaka Fish was selected as a river fish subject to this research because Japanese Medaka is now regarded as the symbolic fish for the restoration of countryside ecosystems in Japan.Judging from the hydraulic measurements in and around artificial blocks that are recently constructed for providing fish nest in agricultural canals, the three environmental factors of water depth, current velocity and cover were selected as the principal factors affecting the availability and quality of the fish habitat in agr … More icultural canals. Based on the background that fish behavior is essentially noncommittal and that the behavioral responses to an environmental change vary widely among both individuals and circumstances, the preference intensities of Japanese Medaka to the three environmental factors were quantified by fuzzy logic modeling and laboratory open-channel experiments. A simple genetic algorithm was introduced to search for the optimal functional representation of preference intensity. The fuzzy-GA preference intensity model was then successfully verified by laboratory water-tank experiments and on-the-spot examinations. The good agreement between the predicted and observed distributions in considering the two factors of current velocity and cover indicated that the two factors obviously affected the environmental preferences of Japanese Medaka.For the purpose of evaluating the water environment of fish habitat through the individual-based model approach, the schooling behavior of Japanese Medaka was analyzed to determine the interactions between individuals: the spacing between individuals, the spatial distribution of individuals and the communication processes. The behavioral pattern model and the artificial network model were introduced to describe the behavior of fish schooling. The results of the numerical simulations using the models agreed with the experimental data in many regards, including the polarization, the nearest neighbor distance, the expanse and the fractal dimension. The differences of swimming behavior among individuals were also quantified by calculating return map, fractal dimension, correlation dimension and wavelet transform for swimming trajectory, swimming velocity and turning angle respectively. The feasibility of the assessment for the degree of tension or rest of fish in specific environment and the fish-based biomonitoring for chemical contaminants in water environment was examined using the calculated results. Less