银杏种子后熟生理与内源激素变化的研究

研究银杏种子后熟期间的生理变化以及内源激素对种子休眠与萌发过程的调控作用.结果表明:银杏种胚分化不完全、发育不充分是种子休眠的主要原因,银杏种子的后熟过程实质上是种胚不断分化生长的过程.在此过程中,物质代谢、酶活性的变化均有利于种胚发育和种子发芽,种子内部经历了一系列生理生化变化去调整内源激素平衡、完成种胚发育.研究发现,内源激素在银杏种子休眠与萌发过程中起着重要的作用:GA1 3是银杏种胚后熟的关键物质;ZRs对银杏种子生长起重要的促进作用;GA1 3/ABA及ZRs/ABA相对比例的变化对种子发育、胚的后熟及种子萌发均有重要的影响.     

NaHCO3前处理对生物炭性质及其磷吸附能力的影响

为探究NaHCO₃前处理对不同种类生物炭性质及其磷吸附能力的影响,借助元素分析、光电子能谱、孔径分析、扫描电镜等手段对比处理前后秸秆、壳核及其他3类生物炭表面特性和孔结构的差异,基于吸附等温线和Freundlich与Dubinin-Radushkevich模型拟合,探讨生物炭性质控制磷吸附的机理。结果表明：NaHCO₃前处理总体提高了各类生物炭的比表面积和孔体积,增幅分别为2.70%～110.84%和1.42%～123.80%,提高其芳香性(C=C),H/C增幅为5.56%～29.41%,同时降低了极性官能团(C—O和C=O)含量,极性指数[(O+N)/C]降幅为13.18%～46.34%。原始生物炭的磷释放量范围为78.33～568.33 mg·kg^-1,NaHCO₃前处理显著增加各类生物炭对磷的吸附,使其表现出近似的磷吸附能力(Freundlich吸附系数K_F范围为119～254 mg^1-n·Lⁿ·kg^-1)...     

基于YOLOv5s改进模型的小白菜虫害识别方法

小白菜是中国种植面积较广、深受大众喜爱的蔬菜,但真实菜地环境中虫害往往出现在叶片的特定区域,且受环境因素如光照和背景干扰较大,影响对其的智能检测。为提高小白菜虫害的检测效率和准确率,该研究提出一种基于YOLOv5s网络框架改进的YOLOPC小白菜虫害识别模型。首先,引入CBAM（Convolutional Block Attention Module）注意力机制,将其放在CBS（卷积层Convolution+归一化层Batch normalization+激活函数层SILU）的输入端构成CBAM-CBS的结构,动态调整特征图中各个通道和空间位置的权重;使用上采样和1×1卷积操作来调整特征图的尺寸和通道数,实现不同层次特征的融合,增强模型的特征表示能力。同时,改进损失函数,使其更适合边界框回归的准确性需求;利用空洞卷积的优势提高网络的感受野范围,使模型能够更好地理解图像的上下文信息。试验结果表明,与改进前的YOLOv5s模型相比,YOLOPC模型对小白菜小菜蛾和潜叶蝇虫害检测的平均精度（mean Average Precision, mAP）达到91.40%,提高了12.9个百分点;每秒传输帧数（Frame Per Second, FPS）为58.82帧/s,增加了11.2帧/s,增加幅度达23.53个百分点;参数量仅为14.4 MB,降低了25.78个百分点。与经典的目标检测算法SSD、Faster R-CNN、YOLOv3、YOLOv7和YOLOv8相比,YOLOPC模型的平均精度分别高出20.1、24.6、14、13.4和13.3个百分点,此外,其准确率、召回率、帧速率和参数量均展现出显著优势。该模型可为复杂背景下小白菜虫害的快速准确检测提供技术支持。     

基于遥感信息和产量形成过程的小麦估产模型

为提高小麦遥感估产的精确性和机理性,在广域环境条件下进行了多品种小麦的种植试验,并基于遥感信息获取的瞬时性与广域性,结合小麦产量形成的生理生态过程及其与气候环境的相互关系,建立了较为简化的小麦估产模型.通过组件化的设计方法实现了遥感信息和估产模型的耦合,即利用抽穗期遥感影像反演的LAI和生物量及时替换小麦估产模型对应参数变量,进而实现对小麦产量的估测.通过试验验证,小麦产量的预测值与实测值较为一致,预测小麦产量的RMSE为354.18 kg·ha-1,利用小麦估产模型可以对不同年份、不同区域的小麦产量形成情况进行监测预报.     

进水流速对圆形循环水养殖池流场特性影响的数值模拟

通过对工厂化循环水养殖进水流速的智能调控,可降低饵料残留,避免水质恶化。为此,本研究采用数值模拟方法探究了进水流速对工厂化循环水养殖池流场特性的影响,并基于该研究设计出一套确定进水流速调控的实验方法。首先,通过对比Standard k-ε、RNG k-ε和Realizable k-ε 3种湍流模型及多种壁面函数的仿真效果,确定RNG k-ε模型和标准壁面函数作为仿真配置。同时,针对多相流模型,对欧拉多相流模型和DPM离散相模型进行对比,为提高计算准确性选用DPM离散相模型,并基于上述模型进行网格无关性验证、制定网格划分方案。其次,以大菱鲆(Scophthalmus maximus)养殖为例,模拟不同进水流速下养殖池流场、排污和水温调节的效果。最后,针对仿真结果提出进水流速调控方案。结果显示,日常采用1.0 m/s的进水流速,可有效提高适宜流速区面积并控制水处理成本;投饵前,采用0.2 m/s的进水流速可以解决循环水养殖中存在的饵料浪费问题;进食结束后,采用1.2 m/s的进水流速可快速排出残饵避免水质恶化;水温异常时,采用15 ℃的水、以1.2 m/s的进水流速注水230 s,可使20 ℃的水下降到正常水平,精准化控制水温。采用本研究提出的方法,可针对不同养殖生物和养殖环境设计进水流速智能调控策略,可用于解决循环水养殖过程中饵料浪费、水质变差和水温异常等问题。     

复方抗菌剂对鲫鱼细胞色素P4503A活性的影响

试验研究了五倍子、复方抗菌剂对鲫鱼CYP3A酶活性的影响。设计复方抗菌剂浓度分别为30mg/kg、10 mg/kg和五倍子浓度为30 mg/kg,采用灌肠给药对鲫鱼进行处理;在试验的第3、7、10天取样测定其肝和肾组织中CYP3A的酶活性变化。结果表明:复方抗菌剂30 mg/kg处理组和五倍子30 mg/kg处理组的肝CYP3A活性较对照组显著降低(P<0.05),且在处理第10天,肝CYP3A酶活性较前2次的降低极显著(P<0.01);而复方抗菌剂10 mg/kg组的肝CYP3A酶活性无显著变化。对于肾组织,各药物组的CYP3A酶活性均无明显变化。     

绿色经济体系下饲料企业经济模式转变研究

近年来,绿色经济成为席卷全球的潮流.绿色经济带给经济社会的改变将是全面而深刻的,饲料行业也不会例外.绿色经济既会给饲料行业带来落后产能淘汰、产业结构升级的机遇,也会带来利润下降、承包上升的短期阵痛.饲料企业想要抓住机遇、迎接挑战,就必须尽快转变当前的经济模式,建立绿色企业文化,加大研发投入,调整生产模式,加强上下游产业...     

拟南芥U-box蛋白的研究进展

泛素-26s蛋白酶体途径（ubiquitin-26s proteosome pathway,UPP）是真核生物体内一种高度选择性的蛋白降解途径。U—box蛋白是该途径中决定底物特异性的一种泛素连接酶E3。U—box结构域大约由70个氨基酸残基构成,在酵母、植物和动物等真核生物中高度保守。植物U-box蛋白（plant U—box protein,PUB）的数目远多于其他生物,在植物的生长发育过程中起着重要作用。目前在拟南芥中已发现60多个PUB蛋白（Arabidopsis plant U-box protein,AtPUB）,成为近年来受到较多关注的一类蛋白。本文根据蛋白的结构特征和系统进化树比较,将61种AtPUB蛋白分为7大类。同时,对已报道AtPUB蛋白的功能进行了综述并提出了研究展望。     

Dihydroartemisinin enhances radiotherapy efficiency in H22 cell tumor-bearing mice by regulating PI3K/AKT signaling pathway

AIM To study the effect of dihydroartemisinin （DHA） on the radiotherapy efficiency in hepatocellular carcinoma H22 cell tumor-bearing mice and the role of phosphatidylinositol 3-kinase （PI3K）/protein kinase B （AKT） signaling pathway in this process. METHODS A model of H22 cell tumor-bearing mice was established. The mice was divided into model group, single radiotherapy group, 5-fluorouracil （5-FU） group, and low-, medium- and high-dose DHA groups. The body weight and tumor volume in each group were measured every other day. At the end of administration, blood was collected from the tail of the mice and the animals were killed by neck removal immediately. The synergistic effect of DHA on radiotherapy was determined, and tumor growth inhibitory rate was calculated. The degree of lymphocyte transformation and natural killer （NK） cell activity were measured by MTT, the serum levels of interleukin-2 （IL-2） and IL-4 were measured by ELISA, and the protein levels of PI3K, AKT and p-AKT were determined by Western blot. RESULTS The H22 cell tumor-bearing mouse model was successfully constructed. Compared with model group, the TGT₃ （tumor growth time to reach 3 times of volume） of single radiotherapy group was remarkably increased （P<0.05）, while tumor weight, lymphocyte transformation degree, NK cell activity, IL-2 and IL-4 levels, PI3K protein level and AKT phosphorylation level were remarkably decreased （P<0.05）. Compared with single radiotherapy group, TGT₃, EF （enhancement factor）, tumor inhibitory rate, lymphocyte transformation degree, NK cell activity, IL-2 level and IL-4 level were increased with the increase in DHA dose （P<0.05）, and the PI3K protein level and AKT phosphorylation level were decreased （P<0.05）. CONCLUSION DHA may enhance the immunity of tumor-bearing mice by inhibiting the activity of PI3K/AKT signaling pathway, thereby enhancing the efficacy of radiotherapy.     

A GIS-based plant prediction model for wetland ecosystems

An existing non-spatial model for the prediction of response of wetland plant species on ecological factors has been transformed into a GIS-based prediction model which produces spatial output at the landscape scale. The input, spatial patterns of the ecological factors, were constructed with geostatistical spatial interpolation (kriging). With this GIS-based model the spatial patterns of presence and absence of 78 wetland plant species are predicted for an area with wetlands in the Netherlands of approximately 500 square kilometers. The GIS-based model has been validated, and the estimated uncertainty of the input has been propagated through the model. At the species level the output shows spatially coherent and non-random patterns. The validation is affected by the propagation of input errors through the model. The number of valid predictions declines approximately 10–20% when 95% confidence intervals are used in the validation. This study shows that it is feasible to use a geostatistical interpolation method to construct spatial patterns of ecological factors on a landscape scale and to use these patterns as input for a GIS-based prediction model. The added uncertainty on the input values however, affects the number of valid predictions of the model.