Using enriched stable isotopes of barium and magnesium to batch mark otoliths of larval golden perch (Macquaria ambigua,Richardson)

Woodcock SH, Gillanders BM, Munro AR, McGovern F, Crook DA, Sanger AC. Using enriched stable isotopes of barium and magnesium to batch mark otoliths of larval golden perch (Macquaria ambigua, Richardson). Ecology of Freshwater Fish 2011: 20: 157–165. © 2010 John Wiley & Sons A/S Abstract – Enriched stable isotope immersion techniques were used to mark the otoliths of larval golden perch (Macquaria ambigua) immediately post‐hatch. Two experiments were undertaken: the first involved rearing larvae in water enriched with three concentrations of ¹³⁷Ba for 1–5 days. Marks were produced in as little as 1 day; however, otolith isotope ratios reached equilibrium with the water in 5 days at 90 μg·l^?1. The second experiment involved rearing larvae in isotope enriched water with combinations of stable isotopes of Ba and Mg for 4 days after hatching. Seven significantly different isotopic signatures were produced using three Ba isotopes, which were reflective of the water. Only slight differences were found in otoliths of larvae that were reared in combinations of Mg isotopes, which did not reflect the water chemistry. The length of golden perch at 3 weeks of age showed that isotope immersion did not negatively affect early growth.     

农业机械合作组织在农业现代化中应用意义浅探

通过分析基层农业机械化合作组织发展演变过程,认为新生的基层农业机械化合作组织既可以增强家庭联产承包制度活力,又可以提高集约化农业生产效率.     

基于双正则化的图像超分辨率盲重建

图像超分辨率重建是利用数字信号处理技术由一系列低分辨率观测图像得到高分辨率图像。大多数重建算法假设成像系统的模糊特性也即点扩散函数(PSF)已知,然而实际的应用环境下PSF事先不知道或部分知道。为此,将未知PSF模型化,提出基于双正则化的图像超分辨率盲重建算法,并且正则化作用的强度随重建图像局部光滑程度的变化而自适应地改变,以便能保护图像细节同时抑制平滑区域的噪声。求解过程中采用交替最小化方法估计PSF参数和高分辨率图像,并随着迭代次数的增加逐步提高每次寻优的精度以节省计算开销。实验结果表明,该算法能够比较准确地估计出PSF参数并取得较好的图像重建效果。     

亚麻优异种质创新利用及发展

亚麻种质创新工作采用外源总体 DNA导入 ,Co6 0— γ射线处理种子、植株以及利用雄性核不育材料进入转育等方法创造新类型、新材料 ,提供给育种家 ,培育出一批优异亚麻新品种 ,促进了亚麻生产的飞速发展。随着新的生物技术和信息技术的突破 ,对今后种质创新工作进行展望     

Growth and viability of hatchery-reared Argyrosomus japonicus released into open and semi-closed systems

Abstract  Hatchery-reared Argyrosomus japonicus (Temminck & Schlegel) were released into estuaries of varying habitat between 1996 and 2004, and growth and recruitment to the fishery evaluated. Fish stocked earlier in summer had significantly faster growth rates ( P  < 0.01), and post-stocking mortality ranged from 1.1% to 11.7% within 7 day post-release. Argyrosomus japonicus stocked in the Georges River in 2003 and 2004 yielded fishery independent recapture rates up to 0.2%. For A. japonicus stocked in Smith's Lake in 1997 and 2004, growth rates, timing of recaptures and increases in commercial catches suggested recruitment to the fishery by 18 months. The 1997 stocking led to a 30-fold increase in A. japonicus catch, but stocking was not financially viable, with a cost:benefit ratio of 2.1. Small, shallow coastal lagoons may represent experimental units for refining stocking strategies, but are not optimal for A. japonicus . The potential for density-dependent effects, and complex relationships between growth, habitat and mortality, highlighted the need for a thorough understanding of species – system interactions. Pilot releases can contribute to this understanding, particularly assessment of habitat requirements, and season, site and size-of-release.     

基于信用提升来解决食品安全问题的探讨

基于食品安全频发的现状,深入分析食品安全问题产生的原因,主要是一线人员意识薄弱漠视生命安全、食品生产经营者过度追逐利益、行政监督监管缺位滞后和缺乏诚信公心的社会文化4个方面的原因,并针对这4个方面,从改善信用状况角度,提出了相应的对策建议。     

浙江海域曼氏无针乌贼资源演变及修复路径探讨

就我国浙江海域曼氏无针乌贼种群分布、生命周期、洄游、近海生长等资源特征以及从20世纪50-60年代的盛产,经历了80-90年代的衰竭,到本世纪初有恢复迹象这段时间内的各种变化及其产生原因展开综述,并探讨了曼氏无针乌贼资源修复路径——加大曼氏无针乌贼增殖放流工作力度;产卵场的修复和保护;洄游路径上禁止捕捞作业;严格控制年捕捞量,以及拟定了资源保护和修复方面相关渔政措施及建议。     

双峰MWD聚α-烯烃油品减阻剂的研制

以络合型TiC13催化剂、N型Ziegler-Natta催化剂和[O,O,N,N]茂金属为主催化剂,采用本体聚合方法,制备了单峰和双峰聚α-烯烃油品减阻剂,考察了催化剂的种类与用量对聚α-烯烃分子质量、分子质量分布及减阻剂性能的影响.结果表明:3种催化剂中,[O,O,N,N]茂金属催化剂的催化效率最高,制备的聚α-烯烃分子质量、分子质量分布及减阻性能最优;使用双金属络合型TiC13和[O,O,N,N]及N型Z-N和[O,O,N,N]制备的双峰聚α-烯烃减阻剂的起效速度和减阻效果均优于单峰聚α-烯烃减阻剂.     

基于近红外光谱和深度学习数据增强的大米品种检测

近红外光谱和深度学习结合的思路是大米品种检测的重要研究方向，其准确检测模型的建立依赖大规模的样本数据，然而采集和预处理样本耗时巨大，对准确性的提升造成限制。为解决上述不足，便于深入探究近红外光谱结合深度学习方法在大米品种检测领域应用的可行性，该研究提出基于近红外光谱结合改进型深度卷积生成式对抗神经网络（deep convolutional generative adversarial network，DCGAN）数据增强的大米品种检测方法。首先，在相同环境下采集4种大米品种的近红外光谱并对原始光谱数据进行预处理，使用去趋势校正（detrend correction，DC）和无信息变量消除算法（uninformative variable elimination，UVE）消除无用光谱特征点。然后，建立改进型DCGAN模型对预处理后的光谱数据进行数据增强，对比试验结果表明，改进型DCGAN相比与传统数据增强方法，改进型DCGAN生成数据的结构相似度指标更优。最后，研究不同数据增强方法结合不同分类方法建立大米品种分类模型的性能，对比试验结果表明，改进型DCGAN数据增强结合一维卷积神经网络（one-dimensional convolution neural network，1D-CNN）分类算法所建模型面向测试集的准确率最高，为98.21%，为简便准确的大米品种检测方案提供了新思路。