阿尔蚜茧蜂对不同龄期豌豆蚜的寄生及后代适合度研究

为了明确阿尔蚜茧蜂(Aphidius ervi Haliday)对寄主豌豆蚜[Acyrthosiphon pisum(Harris)]龄期的寄生偏好及其后代适合度特征,在24℃下进行了阿尔蚜茧蜂对不同龄期豌豆蚜的寄生作用研究。结果表明:豌豆蚜被寄生时的龄期大小对阿尔蚜茧蜂的寄生和后代适合度有显著影响(P0.05)。阿尔蚜茧蜂能寄生各发育期豌豆蚜,但对2龄若蚜的寄生率最高(56.67%),其次为1龄若蚜(30.67%),而对4龄若蚜和成蚜的寄生率最低;寄生低龄蚜虫的后代蜂羽化率显著高于寄生高龄若蚜和成蚜。阿尔蚜茧蜂寄生1龄若蚜后,僵蚜出现历期最长(9.47 d),后代蜂羽化历期也最长(5.51 d),随着被寄生时寄主龄期的增大,僵蚜出现历期和后代蜂羽化历期都逐渐缩短;后代蜂发育历期在寄生1龄若蚜时最长(14.85 d),而寄生成蚜时最短(7.87 d)。后代蜂雌性比例在寄生1龄若蚜时最低(24.52%),而寄生成蚜时最高(57.33%),且后代蜂体型随寄主龄期(体型)增大而增大。寄主龄期(体型)与阿尔蚜茧蜂后代适合度呈正相关。因此,从后代适合度的角度,阿尔蚜茧蜂偏好选择龄期(体型)大的寄主寄生,但取得较高的后代适合度的代价是降低其寄生率。说明寄主龄期(体型)大小对阿尔蚜茧蜂的寄生选择及其后代适合度具有重要影响。     

Comprehensive Analysis of Plant Gene Expression in Soybean Root Nodules at Different Growth Stages

Gene expression profiles in early (12 d after sowing, DAS), mature (35 DAS) and senescent (77 DAS) soybean ( Glycine max (L.) Merr. cv. Enrei) root nodules were analyzed using Lotus japonicus cDNA macroarray. The cDNA macroarray membranes contained 18,144 independent cDNAs of L. japonicus isolated from various organs. The membranes were hybridized to ³³P-labeled single-strand cDNAs transcribed from mRNA extracted from the soybean root nodules at three different growth stages. The results showed that over 75% of the cDNA clones gave a relative signal intensity (RSI) lower than 300, irrespective of the soybean growth stage. cDNA with an RSI value over 5,000 accounted for 2.1, 1.2 and 0.5% of the total cDNA clones at 12, 35 and 77 DAS, respectively. The highest RSI value was detected in a clone hybridized to putative pectinesterase (MWM097c10_r) at 12 DAS, while the RSI value of lycopene β-cyclase (MWL025d06_r) was the highest at 35 and 77 DAS. At 77 DAS, the percentage of transferase gene cDNA clones accounted for 20.5% of the total cDNA clones showing an RSI value over 500, a value clearly lower than that at 12 and 35 DAS. On the other hand, the percentage of hydrolase gene cDNA clones accounted for 45.2% at 77 DAS, a value around 1.5 times higher than that at 12 and 35 DAS. The chronological RSI profiles of individual cDNA clones were classified into nine patterns, and the profiles of the cDNA clones belonging to 15 homologous gene groups encoding enzymes of nitrogen and carbon metabolism, nodule specific proteins, enzymes of RNA and protein hydrolysis, and synthesis of jasmonic acid precursor were compared within and between groups. As a result, significant variations both within- and between-groups were revealed.     

海南尖峰岭热带山地雨林生理生态功能群特征

在海南尖峰岭热带山地雨林,野外测定获得3个与植物-大气相互作用直接相关的生理生态特性指标(最大净光合速率、最大气孔导度和比叶面积),应用等级聚类分析和NMS排序方法对热带山地雨林的主要植物进行功能群划分,分析热带山地雨林4个演替阶段(初期、早期、中期和后期)生理生态功能群组成及其动态规律。结果表明:海南尖峰岭热带山地雨林的87个主要植物种类可划分为8个功能群,检验结果显示功能群间存在显著差异,而功能群内同质性较高;功能群的丰富度及相对多度在不同演替阶段差异显著;功能群的丰富度在演替初期最低,演替早、中期最高,演替后期略有降低;演替初期的优势功能群光合能力强,水分利用效率低,演替早、中期的优势功能群光合能力、水分利用效率均为中等;演替后期的优势功能群光合能力偏低,水分利用效率高。     

松材线虫虫体特异表达cDNA文库的构建与分析

松材线虫取食和致病相关基因的鉴定是揭示松材线虫病致病机制的重要研究内容。以松材线虫混合虫体为测试方,线虫卵为驱动方,构建二者差减cDNA文库。文库测序结果经序列拼接、功能注释,共获得松材线虫虫体特异表达的有效EST序列354个,拼接后获得34个重叠群,其中15个重叠群获得明确的功能注释,其中MixC12、MixC25、MixC33分别与过氧化物酶、β-1,4内切葡聚糖酶、锌指家族蛋白有较高的同源性,这些基因被认为与植物线虫的取食和致病性密切相关。RT-PCR分析验证了所获得的差减cDNA的代表性。     

油茶不同苗龄造林后的生长节律研究

用1年生与2年生油茶苗木造林对比试验结果表明：油茶在相同的生长时间内（含苗龄时间）,苗高生长、地径生长、叶片性状、生物量都有明显的差异。通过对比试验,掌握了油茶不同苗龄造林当年生长发育的特点,为油茶幼林的科学管理提供理论依据。     

贡嘎山海螺沟冰川退缩区土壤序列矿物组成变化

阐明土壤中矿物随时间变化的机制是理解矿物风化和土壤发育的基础。利用X射线衍射法对贡嘎山海螺沟冰川退缩区土壤矿物组成随成土作用时间变化进行了定量分析。结果表明,冰川退缩区成土母质的矿物组成同质性较高,以硅酸盐矿物为主(约90%),包括:斜长石(28.5%)、石英(24.5%)、黑云母、钾长石、普通辉石、角闪石、绿泥石、蛭石;并有少量碳酸盐矿物,如方解石(8%)、白云石(2.3%);以及磷酸盐矿物磷灰石(2.1%)。退缩区土壤的矿物组成总体呈新发育土壤特征,随着成土年龄的增加,方解石逐渐被风化成为草酸钙石,角闪石、黑云母、磷灰石和绿泥石含量逐渐降低,长英质矿物的相对含量有所增加。成土作用中矿物组成的变化受植被原生演替和土壤p H的影响,快速发育的植被导致土壤p H迅速降低,风化程度增强。     

不同抗旱性小麦叶片膜脂过氧化的氮素调控机制

在田间条件下研究了施氮对不同抗旱性冬小麦叶片全生育期黄嘌呤氧化酶(XOD)、超氧化物歧化酶(SOD)、过氧化氢酶(CAT)活性以及过氧化氢(H2O2)和丙二醛(MDA)含量的影响。结果表明,施氮提高了叶片CAT和SOD活性,降低了XOD活性和MDA含量,以N180处理效果最明显,而且产量最高,表明N180处理对小麦膜脂抗过氧化能力和产量形成最为适宜;不施氮(N0)处理条件下,旱地品种较水浇地品种具有较高的保护酶系活性和较低的XOD活性、MDA含量。而适量氮素能够提高叶片保护酶系活性和降低XOD活性、MDA含量,所以氮素能够通过提高叶片膜质抗过氧化能力来增强小麦对干旱的适应。由于降低了超氧阴离子的生成量,使XOD活性降低和CAT活性提高,H2O2和MDA含量维持在较低水平,最终提高小麦产量。     

Time-course of nitrogen fixation in two bambara groundnut (Vigna subterranea L. Verdc.) cultivars

TheA-value method, involving the application of a higher¹⁵N rate to a reference non-N₂-fixing plant, was used to assess the magnitude of N₂ fixation in two bambara groundnut cultivars at four growth stages [vegetative, 0–47 days after planting (DAP); early pod-filling, 47–99 DAP; mid-pod-filling, 99–120 DAP; physiological maturity, 120–148 DAP). The cultivars were Ex-Ada, a bunchy type, and CS-88-11, a slightly spreading type. They were grown on a loamy sand. Uninoculated Ex-Ada and CS-88-11 were used as reference plants to measure the N₂ fixed in the inoculated bambara groundnuts. In this greenhouse study, soil was the major source of N in bambara groundnuts during vegetative growth, and during this period it accounted for over 80% of the N accumulaed in the plants. However, N₂ fixation became the major source of plant N during reproductive growth. There were significant differences between the two cultivars in the ability to fix N₂, and at physiological maturity, almost 75% of the N in CS-88-11 was derived from the atmosphere compared to 55% in Ex-Ada. Also, the total N fixed in CS-88-11 at physiological maturity was almost double that in Ex-Ada. Our data indicate that the higher N₂ fixation in CS-88-11 was due to two factors, a higher intensity of N₂ fixation and a longer active period of N₂ fixation. The results also suggest that bambara groundnut genotypes could be selected for higher N₂ fixation in farining systems.     

缓/控释氮素肥料玉米苗期养分释放特点

采用盆栽试验．模拟田间生态环境．研究施用不同种缓／控释氮素肥料玉米苗期土壤尿素态氮、硝态氮、铵态氮和速效态氮含量，分析比较其在玉米苗期氮素养分释放特点。研究表明，在玉米苗期，施用醋酸酯淀粉包膜脲酶抑制剂nBPT涂层尿素肥料。土壤中尿素态氮和铵态氮的积累量最多．分别为21．72mg／kg和48．31mg／kg；醋酸酯淀粉包膜尿素肥料，硝态氮和速效氮含量最多．分别为102．90mg／kg和135．25mg／kg；丙烯酸树脂包膜脲酶抑制剂nBPT涂层尿素肥料，尿素自膜内迁移到土壤中的量较少，硝态氮和速效态氮含量最少。分别为53．74mg／kg和93．70mg／kg。包膜与脲酶抑制剂nBPT相结合的缓／控释肥料，对减少土壤石硝态氮的生成效果最为明显．明显优于其他缓／控释肥料．丙烯酸树脂包膜nBPT涂层尿素肥料控释效果最好。     

基于RAdam卷积神经网络的水稻生育期图像识别

为了解决现阶段水稻发育期信息的获取主要依靠人工观测的效率低、主观性强等问题,该研究提出一种基于Rectified Adam(RAdam)优化器的ResNet50卷积神经网络图像识别方法,开展水稻关键生育期的自动识别。连续2a对12块试验田的水稻物候特征进行持续自动拍摄,对采集的水稻图像进行预处理,得到水稻各发育期分类图像数据集;采用ExG因子和大津法(Otsu)算法相结合的方法对水稻图像分割,减小稻田背景干扰;对比分析了VGG16、VGG19、ResNet50和Inception v3四种模型下水稻生育期图像分级识别的性能,选取性能较优网络模型并进行了网络参数调优;对比试验了不同优化器下模型准确率和损失值的变化,选取了RAdam优化器。结果表明,采取基于RAdam优化器卷积神经网络构建的模型,在真实场景下分类识别准确率达到97.33%,网络稳定性高、收敛速度快,为水稻生育期自动化观测提供了有效方法。