Tamura模型下DNA序列点突变熵性质的研究

【目的】研究在较复杂的核苷酸替换模型,即Tamura模型下DNA序列中4种核苷酸频率随世代变化的规律,以得到序列熵的性质,为了解生物进化机理奠定基础。【方法】在Shannon信息熵理论基础上,利用La-grange乘数法进行理论证明,并用Mathematica软件模拟了Tamura模型下DNA序列中4种核苷酸组成的变化规律和序列熵的变化过程。【结果】DNA序列在Tamura模型下发生突变,各核苷酸频率随世代变化逐渐趋于平衡,且序列熵趋于最大值。【结论】核苷酸突变具有保熵性,验证了前人生物是朝着多样性增大的方向发展进化的结论。     

Heterogeneous Nature of a ‘new’ Pathotype of Melampsora Rust on Salix Revealed by AFLP

An outbreak of rust on Salix × mollissima (S. triandra × S. viminalis) Q83, an important biomass willow, was first observed at several locations in the UK in 1992. Rust collections obtained from Q83 in 1992 at Long Ashton (south west England), Markington (Northern England) and Loughgall (Northern Ireland), were tested for pathogenicity and examined using amplified fragment length polymorphism (AFLP). All collections showed the same pathogenicity patterns on the eight willow differentials and were assigned to f.sp. larici-epitea typica of Melampsora epitea. A total of 304 AFLP markers was scored for 54 rust isolates, 20 from Long Ashton, 20 from Markington and 14 from Loughgall. Cluster analysis placed the isolates into three distinct groups according to the collection sites. Within each site, Markington isolates were least variable, Nei & Li's similarity coefficients averaging 0.996. Average similarities within isolates from Long Ashton and Loughgall were 0.899 and 0.883, respectively. Average per-locus diversity within site (H_j), calculated using Shannon information index, was 0.014 in Markington, 0.24 in Long Ashton and 0.23 in Loughgall population. Most diversity (69.1%) was partitioned between populations. An analysis of molecular variance (AMOVA) attributed 85.8% of variance to between populations and 14.2% to the individuals within populations. The results suggest that, in 1992, this previously unknown pathotype was not spread from a common source but from separate sources. The AFLP analysis and early records on the host range of M. epitea indicate that the rust virulent to S. × mollissima may have existed in nature before 1992.     

Soybean [<Emphasis Type="Italic">Glycine max</Emphasis> (L.) Merrill] rhizobial diversity in Brazilian oxisols under various soil,cropping, and inoculation managements

In this study, soybean nodules were collected from 12 sites in the State of Mato Grosso, in the Brazilian Cerrados, where both exotic soybean [Glycine max (L.) Merrill] and bradyrhizobial strains have been introduced from 1 to 18 years before. All soils were originally devoid of rhizobia capable of effectively nodulating soybean and varied in terms of chemical and physical properties, inoculation procedures, and cropping systems. Rhizobial genetic diversity was assessed on 240 isolates by rep-PCR fingerprinting with BOX primer, and indices of diversity (abundance-based coverage estimator and traditional and modified Shannon indices) were applied to the profiles obtained. The genetic diversity was much greater than expected, as after the introduction of a maximum of four strains, up to 13 profiles were identified, some sharing many similar bands with the inoculant strains, but others quite distinct from the putative parental genotypes. The increase in the number of rep-PCR profiles could be attributed to genetic variability due to the stressful tropical environmental conditions, but also might indicate that indigenous rhizobia become capable of nodulating the host legume. After the third year of cropping with the host legume, inoculation did not affect rhizobial diversity. A high content of clay decreased diversity in comparison with that seen in a sandy soil, probably due to reduced aeration. Diversity was higher under the no-tillage system when compared to the conventional tillage management, highlighting the importance of maintaining crop residues in tropical soils. Understanding the ecology of exotic rhizobia after being introduced into new cropping areas represents a first step towards the establishment of better strategies of inoculation, which in turn may result in sustainability and higher plant yields.     

土地利用分类粒子群优化概率神经网络半监督算法

针对以往土地利用监测大都采用监督分类算法,成本较高、错分漏分严重且受人为因素影响较大的问题,提出了一种粒子群优化概率神经网络的半监督分类算法。该算法通过粒子群优化算法优化分类器的参数,提高分类器的精度,运用香农熵选择高置信度的样本扩展初始训练样本集,将大量未标记的样本扩展到训练样本集中,减少了初始标签样本的数量,节约了成本,并与随机森林法、最大似然法、概率神经网络算法进行对比分析,总体精度较其他算法提高了1.25～6.57个百分点,Kappa系数达到0.8以上。对新乡市1996年、2004年、2013年、2020年的遥感影像进行土地分类,结果表明1996—2020年间新乡市的建设用地以中部地区新乡县为中心不断扩张,耕地面积也在不断增加,其他用地面积不断减少,沿黄河绿地面积不断增加;土地流转方面耕地转建设用地最为明显,本研究为新乡市进一步合理开发土地资源提供了理论依据。     

沈阳新城区土地利用结构及生态足迹分析

依据用地面积分布数据和开发建设规划,利用信息熵和土地利用均衡度指标,分析了沈北新区土地利用结构的变化趋势.结果表明:沈北新区土地利用信息熵和均匀度总体呈下降趋势,土地利用系统向相对有序状态发展,土地利用的均质性减小,土地利用方式趋向多样化;研究区人均生态足迹为1.22 hm2/人,生态赤字达0.93 hm2/人,处于不可持续发展状态.     

Are beech (Fagus sylvatica) roots territorial?

Root competition between conspecific neighbours has been notoriously difficult to determine in intact ecosystems. Here we applied microsatellite markers to genotype beech trees (Fagus sylvatica, L.) and roots in subplots in an undisturbed, old-growth monospecific forest (Hainich, Germany). Roots were quantified and assigned to individuals. As a measure for territoriality, we determined root segregation. The hypothesis was tested that intraspecific root diversity was higher in the geometric centre of a triangle formed by beech trees than at positions closer to the stem of a tree. Space exploration measured as presence of an individual was correlated with stem diameter but not soil occupation. Fine root biomass, intraspecific Shannon diversity, effective numbers of genotypes and Evenness were stable and unrelated to the distance of the target tree. We conclude that beech shows no indication for territoriality or asymmetric competition in the presence of conspecific neighbours.