An examination of the use of angling data to estimate total returns of Atlantic salmon, Salmo salar, to two rivers in Newfoundland, Canada

Conservation spawning requirements (limits) for Atlantic salmon, Salmo salar L., have been developed for a number of rivers in Newfoundland. Status of stocks is evaluated against these limits and scientific advice and recommendations on levels of recreational catch provided to resource managers. A critical factor in this process is the determination of total returns. This has been accomplished for 30–40% of the rivers in question over the years through the use of fish counting fences and traps installed in fishways. In Newfoundland, it is not logistically or financially possible to operate fish counting facilities in all of the approximately 200 Atlantic salmon rivers. Thus angling catch data have been used as an alternative means of estimating total returns to rivers without counting facilities. Estimates of total returns were three to four times higher than the actual values for one of the two rivers studied and deviated from the actual by as much as 60% for the other. Management implications of the approach are discussed.     

Plant regeneration via protoplast electrofusion in cassava

Protoplast electrofusion between callus protoplasts of cultivar TMS60444 and mesophyll protoplasts of cultivar SC8 was performed as an approach for the genetic improvement of cassava. The fusion products were subsequently cultured in protoplast culture medium(TM2 G) with gradual dilution for approximately 1–2 months. Then the protoplast-derived compact calli were transferred to suspension culture medium(SH) for suspension culture. The cultured products developed successively into embryos, mature embryos, and shoots on somatic embryo emerging medium(MSN), embryo maturation medium(CMM), and shoot elongation medium(CEM), respectively. And the shoots were then rooted on Murashige and Skoog(1962) medium(MS). Sixty-six cell lines were obtained and 12 of them developed into plantlets. Based on assessment of ploidy level and chromosome counting, four of these plantlets were tetraploid and the remaining eight were diploid. Based on assessment of ploidy level and simple sequence repeat(SSR) analysis, nine tetraploid cell lines, one diploid variant plant line and nine variant cell lines were obtained. The diploid variant plant line and the nine variant cell lines all showed partial loss of genetic material compared to that of the parent TMS60444, based on SSR patterns. These results showed that some new germplasm of cassava were created. In this study, a protocol for protoplast electrofusion was developed and validated. Another important conclusion from this work is the confirmation of a viable protocol for the regeneration of plants from cassava protoplasts. Going forward, we hope to provide technical guidance for cassava tissue culture, and also provide some useful inspiration and reference for further genetic improvement of cassava.     

水稻胚乳细胞计数方法研究

在改进Ｓｉｎｇｈ小麦胚乳细胞计数方法的基础上，研究提出水稻单个籽粒胚乳细胞的计数方法，进行了细胞计数过程中各环节的误差分析，并对计数方法鉴别样品差异的能力做出统计推断。证明是一种可用于较大样本量、简便易行的水稻胚乳细胞计数方法。     

氚水乳化测定方法研究

比较了几种种氚的液闪测定方法，系统地分析了乳化法，结果表明：乳化法具有低的本底、淬灭及化学发光；计数效率高且稳定；制样简单、价格便宜、适应性广，可作常规分析。     

利用libsys3.0汇文系统典藏模块进行馆藏清点工作的实践探讨

结合工作实践，详细介绍了如何利用libsys3．0汇文系统典藏模块进行书刊清点工作。     

薄面光折射式小麦种子流多通道并行检测装置设计与试验

针对小麦高速播种作业过程中高频排种种子流精准检测困难的问题,该研究设计了一套薄面光折射式小麦种子流多通道并行检测装置。基于将高通量变为低通量多通道并行同步检测的思路,设计了种子流分流结构。根据小麦种子物理特性,在已有传感原理的基础上,提出了一种\     

基于深度残差网络的麦穗回归计数方法

单位面积的穗数是估算小麦产量的重要指标,针对传统麦穗计数方法效率低、主观性高等问题,将基于深度残差网络的密度回归模型引入麦穗的计数领域,建立原始图片与密度图的对应关系,以密度图像素值总和确定图像中麦穗数量.对ResNet34网络进行改进,提出了ResNet-16模型,实现端对端的麦穗计数.针对ResNet34网络复杂度...     

融合YOLOv7和BYTE多目标跟踪的多类别海珍品计数方法

针对目前养殖过程中海珍品计数方法成本高、效率低、计数精度难以保障等问题,该研究以真实底播养殖环境下的海珍品为研究对象,以水下拍摄的海珍品视频为数据源,提出一种基于视频多目标跟踪的多类别海珍品计数方法。首先,采用性能优异的YOLOv7算法实现海珍品目标检测器,为多目标跟踪提供输入;然后,结合真实养殖环境下同类别海珍品外观相似性高、不清晰等特点,借鉴BYTE算法的多目标跟踪思想,设计多类别轨迹生成策略和基于轨迹ID号的计数策略,提出一种多类别海珍品跟踪与计数方法。并提出一套更适用于基于轨迹ID号计数方法的评估指标。试验结果表明,改进平均计数精度、改进平均绝对误差、改进均方根误差及帧率分别为91.62%、5.75、6.38和32帧/s,各项指标多优于YOLOv5+DeepSORT、YOLOv7+DeepSORT、YOLOv5+BYTE、YOLOv7+BYTE等算法,尤其改进平均计数精度、帧率指标比YOLOv5+DeepSORT高了29.51个百分点和8帧/s,且在改进平均绝对误差、改进均方根误差指标上分别降低19.50和12.08。该研究方法可有效帮助水产养殖企业掌握水下海珍品数量,为现代化渔业的测产研究提供技术参考,为水产养殖的智慧管理提供科学决策依据。     

基于自纠正NMS-ByteTrack的套袋葡萄估产方法

针对套袋后的葡萄体积增加和葡萄叶片表面积大容易出现重叠遮挡,及人工拍摄视频的速度不稳定可能导致套袋葡萄目标丢失的问题,该研究提出一种基于自纠正NMS(non-maximum suppression)-ByteTrack的套袋葡萄估产方法。该方法首先通过目标检测方法YOLOv5s检测视频中的套袋葡萄,将检测阶段的NMS操作后置到追踪阶段,保留因遮挡而被过滤的果实检测框;其次在ByteTrack的基础上加入相机运动补偿和改进的卡尔曼滤波算法,以自动纠正果实预测框的位置并进行追踪;最后提出一种划线计数策略对套袋葡萄自动计数。试验结果表明,该方法的多目标追踪准确率、多目标追踪精度和ID调和平均数分别为64.6%、82.4%和80.8%,相比ByteTrack分别提高了1.7个百分点、1.0个百分点和4.1个百分点,平均计数精度达到82.8%。因此,基于自纠正NMS-ByteTrack的估产方法能有效解决套袋葡萄的追踪计数问题,实现对套袋葡萄更精确地估产。     

水稻土和湿地土壤有机碳测定的CNS元素分析仪法与湿消化容量法之比较

土壤固碳研究中需要精确的(有机)C计量,而常规的湿氧化法与CNS元素分析仪法测定结果的吻合性足C计量中的问题.国外对旱地土壤的研究表明,这两种方法的结果基本可以对比,但是否同样对于湿地土壤也适用还不清楚.采用CNS元素分析仪(仪器法)和重铬酸钾外加热湿氧化法(容量法)对20个水稻土样品,26个淡水湿地土壤样品和20个沿海湿地土壤样品进行了总有机C(TOC)的对比测定.结果表明,无论是土壤的表土样品还是剖面样品,仪器法测定结果约低于容量法10%以下.两种方法对淡水湿地土壤和水稻土的有机C含量的测定可以对比,但是重铬酸钾氧化容量法对于含氯化物较高的沿海湿地土壤有机C的测定可能不精确.CNS元素分析仪测定的精密度和准确度较高,可以用于各种湿地土壤的C计量.