基于DXNet模型的富士苹果外部品质分级方法研究

针对传统计算机视觉技术在苹果外部品质分级中准确率较低、鲁棒性较差等问题,提出了基于深度学习的苹果外观分级方法（多卷积神经网络融合DXNet模型）。首先,在延安市超市、果园等场所实地拍摄不同外观等级的苹果图像15000幅,并进行人工标记,建立了外部品质信息覆盖度广、样本量大的苹果图像数据库;然后,在对比分析经典卷积网络模型的基础上,采用模型融合的方式对经典模型进行优化改进,抽取经典模型卷积部分进行融合,作为特征提取器,共享全连接层用作分类器,并采用批归一化和正则化技术防止模型过拟合。试验评估采用15000幅图像进行训练、4500幅图像进行测试,结果表明,DXNet模型的分级准确率高于经典模型,分级准确率达到97.84%,验证了本文方法用于苹果外部品质分级的有效性。     

‘国光’苹果及其红色芽变花青苷合成与相关酶活性的研究

 以‘国光’苹果及其红色芽变为试材, 测定了果实发育期间的花青苷含量及其相关酶活性,并研究了套袋对芽变花青苷合成的影响。结果显示: ①在果实发育期间, 芽变果皮的花青苷含量明显高于‘国光’, 尤其是成熟期芽变果皮花青苷含量为132170 U·g^-1FW, 而‘国光’仅为49140 U·g^-1FW; ②在果实发育期间, 两个品种间PAL 和UFGT的酶活性无明显差异, 但芽变的CHI和DFR酶活性明显高于‘国光’, 表明芽变花青苷合成能力的提高与CHI和DFR酶活性高有关; ③套袋抑制芽变果皮花青苷的合成, 但解袋后花青苷的含量极显著升高, 解袋后4种酶的变化趋势差异较大, CHI和UFGT活性均迅速升高, 明显高于对照, 这与解袋后花青苷迅速合成相吻合。综上结果, 芽变与原有品种在着色机理上的关键指标是果皮花青苷含量和CHI酶活性。     

苹果园春季土施尿素的利用及其在土壤中的累积

 以7年生嘎拉苹果/平邑甜茶为试材, 利用¹⁵N示踪技术, 研究了苹果生产体系中氮素年周期 的利用、残留、损失及在土壤中的迁移动态。结果表明, 对春季土施尿素的利用率盛花期较低, 为11.38% , 至采收后达到最高。土壤氮素残留率盛花期最高, 为57.10% , 果实采后残留率最低。年周期中各土层氮素残留量不同, 盛花期20～40 cm土层残留量最高, 采收后最低, 0～20 cm土层在新梢旺长期残留量最高, 果实成熟期最低, 而40～60 cm土层在果实膨大期残留量达到最高, 在采收后0～20 cm土层残留有所增加, 其余各土层残留量均达到最低值。氮素损失与土壤残留呈相反的变化趋势, 氮素损失率在盛花期最低为31.53% , 随物候期的推迟逐渐升高, 在果实采后高达58.40%     

MdMYB1转录因子调控红星苹果果实着色的计算机模拟分析及表达验证

 从‘红星’苹果(Malus domestica‘Delicious’) 果实中克隆获得了MdMYB1基因, 通过生物信息学分析( in silico analysis) 和活体细胞融合蛋白荧光观察, 确定其编码蛋白定位于细胞核。同时, 克隆了MdDFR和MdUFGT基因的启动子, 模拟分析表明2个基因的启动子序列中含有MYB结合的顺式作用元件。为了探讨MdMYB1转录因子是否调控MdDFR和MdUFGT基因的表达, 对不同光照条件下果皮花青素含量和基因表达进行了检测, 结果表明光照能够诱导花青素积累, 并迅速启动3个基因表达, 且3个基因表现出高度相似的表达模式, 表明MdMYB1转录因子可能直接调控MdDFR和MdUFGT基因的表达。     

苹果可溶性固形物含量的检测方法——基于可见光近红外光谱技术

为了快速检测苹果的可溶性固形物(SSC)含量,采用可见光近红外光谱技术,结合主成分分析(PCA)和BP神经网络技术,来建立苹果SSC的预测模型.获取苹果样本在345～1039 nm波段的漫反射光谱,采用DPS数据处理系统对其进行主成分分析,并提取出累计可信度大于95 %的5个新主成分.建立一个3层的BP神经网络模型,并将这5个新的主成分作为BP神经网络模型的输入量,其结果是98 %以上预测样本的预测相对误差在5 %以下.该研究表明,采用近红外光谱技术来建立苹果可溶性固形物的预测模型是可行的.     

苹果果实袋成型装置的对比研究

苹果果实套袋技术是提高果品质量的重要措施之一,果实袋的类型及性能是影响苹果质量的主要因素.为此,对果实袋成型工艺及装置的结构进行对比研究,提出了改进和提高果实袋性能的措施,满足了生产高质量果实的要求.     

黄土高原丘陵沟壑区苹果园水肥耦合试验研究

在西北黄土高原丘陵沟壑区选择试验点,开展了苹果园水肥耦合试验,研究内容包括苹果园土壤农化性状测试分析;不同水肥耦合试验组合对土壤养分含量的影响;不同水肥耦合试验组合对苹果树的干周、产量、质量及效益的影响等。试验研究结果证明,实行平衡施肥后,土壤中的养分含量磷素提高到2~3级;氮素保持5级,钾素保持2级,有机质含量均有提高。提出了在苹果树不同树龄段,最适的水肥耦合量比例。按照该水肥耦合比例,能有效提高肥料利用率和土壤养分含量,基本达到了各树龄段的优质果园标准,经济效益显著。     

采后照光处理对富士苹果增色及贮藏的影响

采后人工照光使富士苹果增色的试验结果表明：白色荧光灯照射后果实花青素形成量最多;８℃和 ２０℃下果实花青素形成最快,２℃下最慢,但形成量最多;８０ Ｗ照射果实花青素形成量大于４０ Ｗ照射;２℃和 ８℃下采用白色荧光灯分别照射 １２０、１９２ｈ,可使果实照光面 １００％着色。对果实随后的低温贮藏无显著不良影响。     

Investigation and genetic mapping of a Glomerella leaf spot resistance locus in apple

Apple Glomerella leaf spot (GLS) is a severe fungal disease that damages apple leaves during the summer in China. Breeding new apple varieties that are resistant to the disease is considered the best way of controlling GLS. Fine mapping and tightly linked marker are critically essential for the preselection of resistant seedlings. In this study, a population of 207 F₁ individuals derived from a cross between ‘Golden Delicious’ and ‘Fuji’ was used to construct a fine simple sequence repeat (SSR)‐based genetic linkage map. The position of R_gls, a locus responsible for resistance to GLS, was identified on apple linkage group (LG) 15 using SSR markers CH05g05 and CH01d08, which was adapted from a published set of 300 SSR markers that were developed using the bulked segregant analysis (BSA) method. These two SSR markers flanked the gene, and its recombination rate was 8.7% and 23.2%, respectively. A total of 276 newly developed SSR markers around the target region and designed from the genome apple assembly contig of LG15 were screened. Only nine of these were determined to be linked to the R_gls locus. Thus, a total of 11 SSR markers were in linkage with R_gls, and mapped at distances ranging from 0.5 to 33.8 cM. The closest marker to the R_gls locus was S0405127, which showed a genetic distance of approximately 0.5 cM. The first mapping of the gene R_gls was constructed, and the locations of the 11 effective primers in the ‘Golden Delicious’ apple genome sequence were anchored. This result facilitates better understanding of the molecular mechanisms underlying the trait of resistance to GLS and could be used in improving the breeding efficiency of GLS‐resistant apple varieties.     

Preliminary study of patients with chronic mountain sickness by GC-TOF-MS based serum metabolomics analysis

AIM: To evaluate specific metabolomics profiles in the serum of patients with chronic mountain sickness (CMS) and to explore the potential metabolic biomarkers in the native Tibetans living on the Qinghai-Tibet Pla-teau.METHODS: A gas chromatography time-of-flight mass spectrometry (GC-TOF-MS) approach as a metabolomics technique was used to evaluate metabolic differences. The native Tibetan CMS patients (n=10) and healthy Tibetan controls (n=10) were enrolled from YuShu in Qinghai province in this study. The serum samples were collected and analyzed by GC-TOF-MS coupled with a series of multivariate statistical analyses such as principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA) and orthogonal partial least squares discriminant analysis (OPLS-DA).RESULTS: The intergroup differences between CMS patients and control subjects have been observed. A list of differential metabolites and several top altered metabolic pathways have been identified. The levels of fumaric acid, an intermediate in the tricarboxylic acid (TCA) cycle, and inosine were highly upregulated in the CMS patients, suggesting a greater effort to hypoxic adaptation in high elevation area. Other differential metabolites, such as methyl phosphate, 2-ketoadipate, lyxose and phytanic acid were also identified. Importantly, the differential metabolites possessed higher area under the ROC curve (AUC) values, indicating an excellent clinical ability for the prediction of CMS. Increased levels of amino acids (isoleucine, glycine, serine, L-cysteine, citrulline and trimethyllysine) were detected in CMS group, yet significantly decreased levels of sulfuric acid, oxamic acid, lyxose and glutamine were also detected in CMS group than those in control group. At the same time, the levels of ribose and glucose-1-phosphate were markedly elevated in CMS group (P<0.05).CONCLUSION: The metabolic activities are significantly altered in the serum of CMS patients. High altitude hypoxia may act on the disturbed glucose metabolism and amino acid metabolism in part of the Tibetan triggered by CMS.