Search

The capacity of the bread wheat (Triticum aestivum) genome to tolerate introgression from related genomes can be exploited for wheat improvement. A resistance to powdery mildew expressed by a derivative of the cross bread wheat cv. Tähti ⨯ T. militinae (Tm) is known to be due to the incorporation of a Tm segment into the long arm of chromosome 4A. Here, a newly developed in silico method termed RICh (rearrangement identification and characterization) has been applied to characterize the introgression. A virtual gene order, assembled using the GenomeZipper approach, was obtained for the native copy of chromosome 4A; it incorporated 570 4A DArTseq markers to produce a zipper comprising 2,132 loci. A comparison between the native and introgressed forms of the 4AL chromosome arm showed that the introgressed region is located at the distal part of the arm. The Tm segment, derived from chromosome 7G, harbors 131 homoeologs out of the 357 genes present on the corresponding region of Chinese Spring 4AL. The estimated number of Tm genes transferred along with the disease resistance gene was 169. Characterizing the introgression's position, gene content and internal gene order should facilitate not only gene isolation, but may also be informative with respect to chromatin structure and behavior studies. Funding agencies: National Program of Sustainability I [LO1204]; Czech Science Foundation [14-07164S]; Estonian Ministry of Agriculture

Käesolevas töö eesmärgiks oli anda puistute okaspuude ja lehtpuude suhtelise koosseisu hinnang RMK Pikknurme metsandikus spektraalse segu lahutamise meetodiga ja hägusate hulkadena klassifitseerimise meetodiga. Pikslile mitmest klassikuuluvust lubavate klassifitseerimise meetodite tulemusi võrreldi pikslile vaid ühest klassikuuluvust lubavate lähima kauguse ja suurima tõepära meetoditega. Analüüsiks kasutati satelliidi Landsat 7 skanneri Enhanced Thematic Mapper Plus varasuvist satelliidipilti kuupäevaga 10. juuni aastast 2000. Näidiste valikuks kasutati RMK Pikknurme metsandiku takseerandmete andmebaasi. Sateliidipiltide töötlus on tehtud geoinfosüsteemi (GIS) ja pilditöötluspaketi Idrisi Andes vahenditega. Vektorandmestike andmetöötluseks ja päringute tegemiseks kasutati GISi paketti Mapinfo Professional 10.0. Töö tulemustest selgus, et spektraalse segu lahutamise meetod (üldine täpsushinnang 54%) sobib puistutele koosseisu hinnangu andmiseks paremini kui hägusate hulkadena klassifitseerimise meetod (üldine täpsushinnang 45%). Piksleile mitmest klassikuuluvust lubavate klassifitseerimise meetodite tulemusi võrreldi jäikade klassifitseerimise meetodide tulemustega ning suurima tõepära meetod (üldine täpsushinnang 79%) andis parema tulemuse kui lähima kauguse meetod (üldine täpsushinnang 53%). Suurima tõepära meetod andis üldse kõige paremad tulemused kõikidest töös kasutatavatest meetoditest. Spektraalse segu lahutamise meetodil ja hägusate hulkade klassifitseerimise meetodil andsid paremaid tulemusi puhtpuistud ja suure lehtpuude või okaspuude osakaaluga puistud võrreldes ligilähedaselt võrdsete kaaludega segapuistutega. Okaspuude osakaaludel oli RMK andmetes esindatud osakaaludega parem kooskõla kui see oli lehtpuudel. Kõige suuremad vead klassifitseerimisel tekkisid nooremates puistutes, mille tõttu tuleks neid puistuid vanematest puistutest eraldi vaadelda. The aim of this study was to estimate forest stand composition in Pikknurme State Forest District with linear spectral mixing and fuzzy set classification methods. The results obtained with these classification methods were compared with minimum distance and maximum likelihood classification results. The satellite image used in the study was Landsat 7 scanner Enhanced Thematic Mapper Plus early summer image from year 2000, 10th of June. The satellite image processing was made with geographic information system (GIS) and image processing software Idrisi Andes. Vector data processing and queries were done using GIS software Mapinfo Professional 10.0. The results of the analysis showed that linear spectral mixing method (overall accuracy 54%) was better than fuzzy set classification method (overall accuracy 45%). Maximum likelihood method (overall accuracy 79%) was better than minimum distance method (overall accuracy 53%). Overall, maximum likelihood method gave better results than all other methods. Soft classification methods gave better results in pure stands and in almost pure stands, then in mixed stands. Coniferous forests performed better results if compared to the data in the forestry database than did deciduous forests. The most significant misclassifications were detected in case of young stands. It is recommended that in subsequent analysis these young stands should be treated separately.