山东近海三种底层鱼类的摄食生态与食物竞争

研究根据2016—2017年在山东近海进行的四个季节的渔业资源底拖网调查和胃含物分析数据, 应用分类树和生态位重叠指数法等, 研究了该海域高眼鲽(Cleisthenes herzensteini)、黄鮟鱇(Lophius litulon)和小眼绿鳍鱼(Chelidonichthys spinosus)3种主要底层鱼类的食物组成及其营养-空间二维生态位的重叠情况, 并分析环境因素对其摄食习性的影响, 旨在为该海域渔业资源的可持续利用和科学管理提供基础资料。研究发现, 这3种底层鱼类均主要以虾类和鱼类为食, 属于底栖动物食性。3种鱼类的摄食习性存在一定的差异, 可以通过构建分类树将其进行区分。通过CCA分析发现, 这3种鱼类的摄食习性受不同的环境因素影响, 其中高眼鲽的摄食习性主要与其体长有关, 随着体长的增加, 高眼鲽会摄食更多的鱼类和虾类; 小眼绿鳍鱼的摄食主要与季节有关, 不同季节小眼绿鳍鱼摄食鱼类和虾类的比例不同; 而黄鮟鱇的摄食主要受空间分布的影响, 高纬度的黄鮟鱇摄食更多的虾类, 而随着经度的增加, 黄鮟鱇会摄食更多的鱼类。黄鮟鱇的营养生态位和空间生态位宽度均最高, 而其余两种鱼的营养和空间生态位宽度相对较小。山东近海这3种鱼类的营养生态位重叠指数都较高, 但是空间生态位重叠指数较小, 在一定程度上可以缓和它们之间的种间竞争。     

鱼类摄食代谢和运动代谢研究进展

摄食和运动不仅是动物最主要的生理活动,同时也是机体代谢能量消耗的主要过程。相关研究表明鱼类摄食代谢主要由营养物质同化过程的耗能组成,其食物蛋白质同化耗能远低于陆生脊椎动物,而摄入营养物质不平衡可能导致摄食代谢耗能增加;鱼类摄食代谢和运动代谢上可能存在能量消耗与性能维持之间的权衡,且都可能受最大代谢能力限制。鱼类不仅在摄食和运动代谢的相对大小及其他特征上存在差异,而且在摄食和运动代谢竞争上存在不同的模式。从功率分配的角度,研究鱼类摄食和运动代谢特征及其与物种生态习性的关系将成为鱼类能量学研究的重要方向之一。     

位相分辨偏振灵敏光学相干层析术:组织双折射特性的成像与定量分析

我们研制了一种基于光纤的位相分辨偏振灵敏光学相干层析成像系统。该系统中的偏振状态控制设量在参考臂而非光源臂上,因而使得光抵达样品的传输效率大大提高。鉴于光源的部分偏振性,入射于样品上的光含有任意偏振状态的分量,通过对参考光偏振状态的调制,就可相干地提取对应于入射光四种正交偏振状态并经样品后向散射的光信号。基于斯托克斯矢量夹角在无损光纤系统传输的变换不变性,我们能利用测量臂中光信号的斯托克斯参数来确定双折射样品深度分辨的位相延迟信息。利用所研制的偏振灵敏光学相干层析成像系统,不仅确认了韧带和软骨的双折射性质,而且定量分析了不同条件下韧带的双折射变化．研究结果表明：韧带松弛可使其双折射特性明显减弱,而韧带经拉伸后,其双折射特性的变化却不明显。     

基于基因表达谱的肿瘤特异基因表达模式研究

基于肿瘤基因表达谱, 利用生物信息学的方法, 从肿瘤与正常组织的样本分类入手就肿瘤特异表达基因的发现及其表达模式问题进行了分析和研究, 进而探讨了肿瘤在基因表达上的特点. 首先, 在分析肿瘤基因表达谱特点的基础上, 提出了基于Relief算法的样本分类特征基因选取策略; 然后, 以支持向量机为分类工具进行样本类型的识别, 以分类错误率为标准选取样本分类特征基因, 并对其中反映肿瘤与正常样本组织构成特点的组织特异表达基因进行排除以突出肿瘤样本真实的类别特征; 最后结合统计学方法, 从信息学的角度论证了分类特征基因在肿瘤组织中特异表达的确实性和普遍性, 并对这些基因在肿瘤组织中呈现出的特异的表达模式进行了分析.     

基于信号稀疏表示的高光谱散射图像苹果粉质化分类

苹果的粉质化是指苹果果肉发软、汁液减少等一系列物理和生理变化现象,采用高光谱散射图像技术结合信号稀疏表示分类算法(SRSA)研究了苹果的粉质化分类问题。首先利用平均反射算法(MEAN)提取了600～1000 nm的高光谱散射图像特征;引入遗传算法(GA)解决分类样本的不均衡问题,在此基础上,把苹果的粉质化分类问题,转化为一个求解待识别样本对于整体训练样本的稀疏表示问题。仿真结果表明,基于信号稀疏表示分类算法的苹果粉质化分类精度为79.8%,高于偏最小二乘判别分析(PLSDA)的74.8%,为苹果的粉质化分类提供了一种新的有效的方法。     

鱼类通过鱼道内水流速度障碍能力的评估方法

鱼类通过鱼道内水流速度障碍能力的量化对鱼道设计有重要理论和实际价值,其基础是鱼类游泳能力的测定.首先对鱼类游泳能力的研究方法进行了概述总结,指出了鱼类游泳能力经典测试方法存在测定流场与自然情况相差较大的不足;分析了关键要素如鱼类行为特征、生理耗能规律及水力特性对鱼类通过水流速度障碍能力的影响;提出了分析鱼类游泳行为和能力与特征流场的关系,探讨鱼类通过水流障碍行为规律和生理疲劳恢复特征,通过研究仿自然流态下的鱼类自由游泳行为、水力计算及生理耗能的关系,构建多因素鱼类游泳能力关系式,定量评价鱼类通过鱼道内水流速度障碍的发展方向.     

基于机器学习技术的非染色细胞凋亡检测分类新方法

细胞凋亡检测和分类在生物医学研究中具有重要意义。该研究目的是在我们多年研究的基础上建立了一种基于偏振衍射成像流式细胞检测系统和机器学习技术的凋亡检测新方法,具有更高的时间效率和更好的应用前景。化学诱导K562和HL60细胞凋亡。通过荧光激活细胞分选仪结合荧光双染技术将细胞分为三个亚群:健康细胞、早期凋亡细胞、晚期凋亡/坏死细胞。应用偏振衍射成像流式细胞检测技术采集每个亚群细胞的衍射图像。基于局部二值模式算法提取图像纹理特征生成训练和测试数据集,研究了不同的分类算法,建立凋亡分类模型。对模型性能和时间效率做出对比分析,筛选出具有更高时间效率的模型。新方法可实现90%的分类准确度,并在时间效率上具有优势。该研究成功开发了一种快速的无染色细胞凋亡检测新方法,使检测后的细胞可直接用于后续实验。     

基于深度学习的蝴蝶科级标本图像自动识别

【目的】本研究旨在探讨深度学习模型在蝴蝶科级标本图像自动识别中的可行性和泛化能力。【方法】为了提高识别模型的鲁棒性和泛化能力,将锤角亚目中6个科1 117种蝴蝶标本图像通过水平翻转、增加图像对比度与亮度以及添加噪声的方式增强图像数据集。在Caffe框架下,利用迁移学习方法,首先使用Image Net数据集中的图像训练Caffe Net模型,迭代31万次后得到初始化的网络权值;然后利用蝴蝶图像训练已预训练好的Caffe Net模型,通过参数微调,获得一个蝴蝶科级标本图像自动识别的卷积神经网络模型。为了比较深度学习和传统模式识别两种方法建立的模型的泛化能力,对相同训练样本提取全局特征和局部特征,训练支持向量机(support vector machine,SVM)分类器。所有的模型在与训练样本图像来源一致和不一致的两个测试样本集上进行测试。【结果】当测试样本与训练样本来源一致,均为蝴蝶标本图像时,基于Caffe Net的蝴蝶识别模型对6个科的蝴蝶识别准确率平均达到95.8%,基于Gabor的SVM分类器也获得了94.8%的识别率。当测试样本与训练样本来源不一致,为自然环境下拍摄的蝴蝶图像时,两种方法获得的识别率均下降,但Caffe Net模型对蝴蝶自然图像的平均识别率仍能达到65.6%,而基于Gabor的SVM分类器的识别率仅为38.9%。【结论】利用Caffe Net模型进行蝴蝶科级标本图像识别是可行的,相比较传统模式识别方法,基于深度学习的蝴蝶识别模型具有更好的泛化能力。     

基于设计模板的BRD-like折叠类型综合分类方法

蛋白质折叠规律研究是生命科学重大前沿课题,折叠类型分类是蛋白质折叠研究的基础。构建BRD-like折叠类型模板数据库,建立了基于多模板的综合分类方法,并用于该折叠类型的分类。对实验集的12 117个样本进行检验,结果的敏感性、特异性分别为0.923和0.997,MCC值为0.72;对独立检验集2 260个样本的检验,结果发现:敏感性、特异性分别为0.941和0.998,MCC值为0.86.结果表明:基于多模板的综合分类方法可用于蛋白质折叠类型分类。     

东太平洋公海茎柔鱼饵料中鱼类和头足类组成初步分析

茎柔鱼(Dosidicus gigas)广泛分布于东太平洋海域,在海洋生态系统中扮演着重要角色。本研究根据2019年6—12月我国鱿钓生产船在东太平洋公海采集的62尾茎柔鱼样本,对其胃含物残留的4131个耳石和75个角质颚样本进行鉴定分析。结果表明: 茎柔鱼主要摄食鱼类10种、头足类4种。从摄食饵料的出现频率和数量百分比来看,主要摄食的鱼类饵料为荧串光鱼、朗明灯鱼和墨西哥尾灯鱼,荧串光鱼在不同胴长组的茎柔鱼胃内均有出现,是茎柔鱼最重要的饵料;摄食的头足类主要有茎柔鱼、鸢乌贼和爪乌贼。随着胴长的增大,茎柔鱼摄食的饵料种类不断增加,头足类和灯笼鱼饵料比例增加,且摄食的饵料尺寸增大,摄食营养结构发生改变,摄食营养水平随之提高。研究结果可为评估不同饵料生物在茎柔鱼食性转变过程中的食物贡献提供基础资料。     

White Blood Cells Image Classification Using Deep Learning with Canonical Correlation Analysis

White Blood Cells play an important role in observing the health condition of an individual. The opinion related to blood disease involves the identification and characterization of a patient's blood sample. Recent approaches employ Convolutional Neural Network (CNN), Recurrent Neural Network (RNN) and merging of CNN and RNN models to enrich the understanding of image content. From beginning to end, training of big data in medical image analysis has encouraged us to discover prominent features from sample images. A single cell patch extraction from blood sample techniques for blood cell classification has resulted in the good performance rate. However, these approaches are unable to address the issues of multiple cells overlap. To address this problem, the Canonical Correlation Analysis (CCA) method is used in this paper. CCA method views the effects of overlapping nuclei where multiple nuclei patches are extracted, learned and trained at a time. Due to overlapping of blood cell images, the classification time is reduced, the dimension of input images gets compressed and the network converges faster with more accurate weight parameters. Experimental results evaluated using publicly available database show that the proposed CNN and RNN merging model with canonical correlation analysis determines higher accuracy compared to other state-of-the-art blood cell classification techniques.     

Fish image recognition method based on multi-layer feature fusion convolutional network

Solving the problem of fish image classification is important to conserve fish diversity. This conundrum can be addressed by developing a new fish image classification method based on deep learning by training data with complex backgrounds. To this end, this paper proposes a fusion model, referred to as Tripmix-Net. The backbone network of the proposed model primarily consists of multiscale parallel and improved residual networks that are connected in an alternate manner, and network fusion is used to integrate the information that is extracted from shallow and deep layers. Experiments conducted on the 15-category WildFish fish image dataset verified the efficacy of the proposed Tripmix-Net for classifying same-genus fish images with complex backgrounds. The model achieved an accuracy of 95.31%, which is a significant improvement over traditional methods. The proposed approach serves as a new concept for the fine-grained image classification of fish against complex backgrounds.     

Dietary self-selection in fish: a new approach to studying fish nutrition and feeding behavior

The principles of modern aquaculture encourage the development of fish feeds containing low fish meal content and several types of plant ingredients plus nutrients to avoid depleting global fish stocks and to reduce costs. However, food constituents can affect animal nutrition and feeding behavior, so the effect of different diets on fish behavior and growth needs to be understood to optimize the use of nutrients and to improve fish welfare. The development of multiple-choice self-feeding systems led to a new perspective for investigating these issues in aquaculture species. Our purpose with this review is to summarize the information that has been published to date on this topic and to identify gaps in knowledge where research is needed. Key subjects are assessed under the following major headings: How do we study dietary selection in fish? What food signals do fish use to choose the right diet? and How do fish respond to food challenges? The present review will provide a picture of the main results obtained to date in these studies in aquaculture fish species, as well as perspectives for future research in the field.     

Status and impact of rural aquaculture practice on Lake Victoria basin wetlands

Rural aquaculture in Lake Victoria basin is a fast increasing nontraditional farming activity which if not appropriately practised will lead to degradation of the wetlands. As part of a study to develop appropriate guidelines and model systems for wetlands‐based rural aquaculture in the basin a survey was conducted to assess the status and the ecological and socioeconomic impact of rural aquaculture on wetlands and wetlands communities. Aquaculture practice was found to be common but not as a major activity. Aquaculture in the wetlands can be described as a low input‐low output production activity and subsistence based on ponds under 400 m² using free seed from public agencies with hardly any supplementary feeding. Men owned most of the ponds and women only contributed to the management of the fishponds by feeding the fish. Poorly constructed ponds and loss during harvesting have led to the escape of cultured species into the wild. Introduction of nonnative species in the basin has already led to wide ranging ecological, environmental and socioeconomic changes whose impact and usefulness are still very much contentious. Repeat of such scenarios can be avoided if appropriate and science‐based models for rural aquaculture farming are developed, tested and disseminated to the communities.     

Low-Dose Micro-CT Imaging for Vascular Segmentation and Analysis Using Sparse-View Acquisitions

The aim of this study is to investigate whether reliable and accurate 3D geometrical models of the murine aortic arch can be constructed from sparse-view data in vivo micro-CT acquisitions. This would considerably reduce acquisition time and X-ray dose. In vivo contrast-enhanced micro-CT datasets were reconstructed using a conventional filtered back projection algorithm (FDK), the image space reconstruction algorithm (ISRA) and total variation regularized ISRA (ISRA-TV). The reconstructed images were then semi-automatically segmented. Segmentations of high- and low-dose protocols were compared and evaluated based on voxel classification, 3D model diameters and centerline differences. FDK reconstruction does not lead to accurate segmentation in the case of low-view acquisitions. ISRA manages accurate segmentation with 1024 or more projection views. ISRA-TV needs a minimum of 256 views. These results indicate that accurate vascular models can be obtained from micro-CT scans with 8 times less X-ray dose and acquisition time, as long as regularized iterative reconstruction is used.     

Host impact of monogenean Lepidotrema bidyana infection and intensity estimates for onsite monitoring

We developed a rapid effective method for accurate estimation of intensity for the monogenean Lepidotrema bidyana, a gill parasite of silver perch Bidyanus bidyanus. This parasite requires monitoring because high-intensity infections reduce host growth and can lead to secondary bacterial and fungal infections. The most accurate method for counting L. bidyana was visual examination of fresh gills. There was a significant relationship between fish size and parasite intensity; however, there was no significant relationship between fish condition and parasite intensity. Parasite intensity estimates were generated by using the mean intensity of worms on the posterior hemibranch on the first left gill arch, compared to the total mean intensity of worms on all hemibranchs. Estimates were validated by predicting L. bidyana intensity from a random sample of silver perch obtained from aquaculture ponds. Parasite intensity estimates correlated strongly to real counts, and this method can be used to accurately predict parasite intensity on an individual host, and thus represents an improvement over previous methods.     

Facilitation and Dominance in a Schooling Predator: Foraging Behavior of Florida Pompano,Trachinotus carolinus

Presumably an individual’s risk of predation is reduced by group membership and this ‘safety in numbers’ concept has been readily applied to investigations of schooling prey; however, foraging in groups may also be beneficial. We tested the hypothesis that, when feeding in groups, foraging of a coastal fish (Florida Pompano, Trachinotus carolinus) on a benthic prey source would be facilitated (i.e. fish feeding in groups will consume more prey items). Although this question has been addressed for other fish species, it has not been previously addressed for Florida Pompano, a fish known to exhibit schooling behavior and that is used for aquaculture, where understanding the feeding ecology is important for healthy and efficient grow-out. In this experiment, juvenile Florida Pompano were offered a fixed number of coquina clams (Donax spp.) for one hour either in a group or as individuals. The following day they were tested in the opposite configuration. Fish in groups achieved greater consumption (average of 26 clams consumed by the entire group) than the individuals comprising the group (average of 14 clams consumed [sum of clams consumed by all individuals of the group]). Fish in groups also had fewer unsuccessful foraging attempts (2.75 compared to 4.75 hr^-1) and tended to have a shorter latency until the first feeding activity. Our results suggest fish in groups were more comfortable feeding and more successful in their feeding attempts. Interestingly, the consumption benefit of group foraging was not shared by all – not all fish within a group consumed equal numbers of clams. Taken together, the results support our hypothesis that foraging in a group provides facilitation, but the short-term benefits are not equally shared by all individuals.     

Structured modeling of fish physiology.

The use of models in simulation and state estimation has proved useful in diverse applications, especially in industrial process control. The project presented here looked into the modeling of fish physiology for applications in fish physiology research and aquaculture. The models deal with gastric evacuation, metabolism, kidneys, gills, the cardiovascular system, and feeding behavior and are based on data from the literature. Model responses are mostly in accord with real responses in principle, but, as with most models of complex biological systems, the numerical accuracy is low in several cases. However, these structured models enable researchers to test hypotheses by altering the submodels and parameters.     

Development of ethologo-physiological differentiation between parrs and smolts of the Atlantic salmon <Emphasis Type="Italic">Salmo salar</Emphasis>

Dynamics of the level of monoamines and their metabolites in the brain of juveniles of the Atlantic salmon Salmo salar is investigated in the initial period of differentiation of the progeny into parrs and smolts. The first differences in the level of metabolism of noradrenaline (NA) and 5-HT arise in the optical tectum of the brain before the beginning of smolting. Appearance of the first signs of smolting in juveniles is accompanied by the increase of differences in the level of activity of HA-and 5-HT-ergic systems in the forebrain and hypothalamus. The first differences in the activity of the DAergic system between parrs and smolts were observed in the period of visible differences in fish coloration. In the same period, considerable differences in the level of adrenaline and noradrenaline in the fish blood were recorded. In experiments on a prolonged exposure of fish to a current of different intensity, it was shown that behavior and physiological response of the fish organism are different: in parrs, the stressogenous response is developed, they stop feeding, weight loss takes place, and they cannot withstand the water flow; smolts adapt to the current using heterogeneity of water movements.     

Orthogonal-polarization-gating optical coherence tomography for human sweat ducts in vivo

We propose an orthogonal-polarization-gating optical coherence tomography (OPG-OCT) for human sweat ducts in vivo. OPG-OCT is composed of the orthogonal linearly polarized light of a sample arm individually interfering with orthogonal linearly polarized lights of the reference arms, where OPG-OCT induces two images, one reflecting the projection intensity and the other the horizontal linear diattenuation (HLD). The results demonstrate that OPG-OCT projection intensity could improve the image quality of sweat ducts. HLD also clearly illustrates the spiral shape of the sweat ducts. Finally, sweat ducts in intensity image are segmented by employing convolutional neural networks (CNN). The proportions of left-handed and right-handed ducts are extracted to characterize the sweat ducts based on HLD. Therefore, the OPG-OCT technique employing CNN for the human sweat glands has the potential to automatically identify the human sweat ducts in vivo.