Manila clam Venerupis philippinarum as a biomonitor to metal pollution     

吴惠丰  吉成龙  王清  刘小莉  赵建民  冯江华 《中国海洋湖沼学报》2013,31(1):65-74

The Manila clam Venerupis philippinarum is a good biomonitor/bioindicator to marine metal pollution and is frequently used in aquatic toxicology. Two dominant pedigrees (white and zebra) of clam are distributed in the Bohai Sea; however, little attention has been paid to potential biological differences between these two pedigrees. In this study, we tested the sensitivity of both pedigrees to marine metal (cadmium and zinc) pollution biomonitoring and marine environmental toxicology. Results demonstrate significant biological differences in gills of white and zebra clams based on metabolic profiles and antioxidant enzyme activities. In addition, we found that hypotaurine, malonate and homarine were relatively high in white clam gills, while alanine, arginine, glutamate, succinate, 4-aminobutyrate, taurine and betaine were high in zebra clam gills. Zebra clam gills were also more sensitive to a mixture of Cd and Zn, as shown by antioxidant enzyme activities and metabolic profiles, but white clam gills could accumulate more Zn. Therefore, we suggest that the white pedigree can be used as a biomonitor to marine Zn pollution, whereas the zebra pedigree can be used for toxicology studies on Cd and Zn mixed pollution.  相似文献   

内华达试验基地雷尼尔台地裂隙凝灰岩中水流...     

Russ.  CE 赵建民 《国外地质(成都)》1991,(4):40-46,24

  相似文献   

基于深度卷积神经网络的海洋牧场岩礁性生物图像分类     

孙东洋  刘辉  张纪红  孙利元  王清  赵建民 《海洋与湖沼》2021,52(5):1160-1169

水下视频可直观记录和反映海洋牧场生物资源的现状和变动,目前亟待开展基于图像的海洋牧场生物识别分类方法研究,以充分发挥图像处理技术在海洋牧场生物群落监测领域的应用潜力。利用采集自我国北方烟威地区包含鱼礁、藻床和泥沙三种图像背景的水下视频,开展了图像增强、图像分类数据集的建立和3种分类模型的应用。对比了基于绿通道的色彩补偿和限制对比度的自适应直方图均衡等方法在海洋牧场水下图像增强上的效果。建立了北方海洋牧场常见岩礁生物图像分类数据集,包括花鲈(Lateolabraxjaponicus)、(Lizahaematocheilus)、许氏平鲉(Sebastes schlegelii)等鱼类11种、棘皮类3种和蟹类1种,共23 211张图像。基于飞桨深度学习框架和PaddleX全流程开发工具,选择AlexNet、MobileNetV3和ResNet50三种图像分类卷积神经网络进行迁移学习,并分别验证了其在含噪音水下图像上的鲁棒性。结果表明,三种模型在测试集的类准确率分别达到96.64%、94.75%和99.23%,其中ResNet50模型在含有高斯噪音的图像集验证具有更好的鲁棒性。综之,基于深度学习的计算机视觉技术在我国海洋牧场生物群落监测中具有较大应用潜力,可为我国海洋牧场监测和管理提供新的思路和方法。  相似文献   

带鱼(Trichiurus haumela)鱼卵DNA的提取及其18S rDNA初步分析   总被引：2，自引：0，他引：2  

张继民  刘霜  赵建民  马兆党  张洪亮 《海洋通报》2009,28(6)

目前应用显微镜技术很难进行鱼卵种类的准确鉴定,本文尝试采用DNA技术开展鱼卵的鉴定工作。以带鱼(Trichiurus haumela)鱼卵为研究对象,采用苯酚-氯仿法提取了其DNA基因组,并采用PCR及克隆测序的方法,对其18S rDNA序列进行了测定与初步分析。结果表明,提取的带鱼(T.haumela)鱼卵DNA基因组片段长度约23kb,PCR扩增片段长度约1.8kb;带鱼(T.haumela)鱼卵18S rDNA与其它已知鱼类的18S rDNA序列具有一定的差异性,可以用其进行带鱼(T.haumela)鱼卵的初步鉴定。  相似文献