Hyperspectral data for classification of vegetation of alpine treeless in the Krkonoše Mts. ABSTRACT The Master Thesis is a part of the HyMountEcos project, which deals with a complex evaluation of mountain's ecosystems in the Giant Mountains National Park using the hyperspectral data. The area of interest is alpine treeless in the Giant Mountains National Park. The main goal of this thesis was to create detailed methodology for classification of vegetation cover using hyperspectral data from AISA DUAL and APEX sensors, to find a classification method, which would improve the accuracy of the results compared to those found in the literature, and to compare the accuracy reached with these two types of the data. Many different classification algorithms (Spectral Angle Mapper, Linear Spectral Unmixing, Support Vector Machine, MESMA a Neural Net) were applied and the classification results were statistically evaluated and compared in the next part of the work. The classification method Neural Net was found as the most accurate one, as it gives the most accurate results for APEX data (the overall accuracy 96 %, Kappa coefficient 0,95) as well as for AISA DUAL data (the overall accuracy 90 %, Kappa coefficient 0,88). The resulting accuracy of the classification (the overall one and also for some classes) reached...

<p>This work investigates the assignment of unknown sequences to their species of origin. In particular, I examine four questions: Is existing (GenBank) data reliable for accurate species identification? Does a segregating sites algorithm make accurate species identifications and how does it compare to another Bayesian method? Does broad sampling of reference species improve the information content of reference data? And does an extended model (of the theory of segregating sites) describe the genetic variation in a set of sequences (of a species or population) better? Though we did not find unusually similar between-species sequences in GenBank, there was evidence of unusually divergent within-species sequences, suggesting that caution and a firm understanding of GenBank species should be exercised before utilizing GenBank data. To address challenging identifications resulting from an overlap between within- and between species variation, we introduced a Bayesian treeless statistical assignment method that makes use of segregating sites. Assignments with simulated and <em>Drosophila</em> (fruit fly) sequences show that this method can provide fast, high probability assignments for recently diverged species. To address reference sequences with low information content, the addition of even one broadly sampled reference sequence can increase the number of correct assignments. Finally, an extended theory of segregating sites generates more realistic probability estimates of the genetic variability of a set of sequences. Species are dynamic entities and this work will highlight ideas and methods to address dynamic genetic patterns in species.</p> / Doctor of Philosophy (PhD)

Abstract Boreal and arctic plant communities are responding to anthropogenic climate change that has been exceptionally rapid during the recent decades. General responses include increased productivity, range expansions and biodiversity changes, all of which affect ecosystem functions. Vegetation dynamics are however controlled by multiple drivers, and the outcomes under the changing climate are not yet fully clear. As high latitude areas often lack long-term monitoring of vegetation, alternative methods are required to observe and understand vegetation changes and dynamics. Recently, resurveying historical vegetation data has become a valuable method of studying vegetation changes over the past few decades. In this thesis, I studied multidecadal (23–60 years) vegetation changes in forest and treeless heath and tundra plant communities along a latitudinal gradient in northern Fennoscandia using both vegetation resurveys and long-term experimental data. In addition to examining climate-driven vegetation changes, I related changes in plant communities to key local drivers of each context including mesotopography, grazing, soil moisture and soil fertility. General trends among the resurveyed treeless heath sites were the pronounced increase of the dwarf shrub Empetrum nigrum ssp. hermaphroditum in snow-protected habitats and the decrease of lichens throughout. Southernmost heath communities showed strong responses to multidriver effects and had shifted towards new community states. The long-term experiment in the tundra confirmed that depending on driver combinations, tundra communities evolve towards divergent alternative states, highlighting the importance of local drivers in modifying tundra vegetation over time. Communities in fertile forest sites experienced greater temporal turnover compared to infertile forest sites, suggesting that the soil fertility level is a key predictor of vegetation changes under climate change. This particularly important finding previously relied mainly on experimental evidence. Despite these generalities, changes in diversity, plant groups and species varied under a rather uniform climatic warming trend and were often habitat- or region-specific. Thus, the results of my thesis highly motivate continued monitoring and resurveying of vegetation under rapid environmental change and also form baseline time-series data for future studies. / Tiivistelmä Poikkeuksellisen nopea ilmastonmuutos on johtanut viime vuosikymmenten aikana muutoksiin boreaalisissa ja arktisissa kasviyhteisöissä. Muutoksiin lukeutuvat tuottavuuden lisääntyminen, levinneisyysrajojen siirtyminen sekä muutokset biodiversiteetissä, mitkä kaikki muuttavat ekosysteemien toimintaa. Kasvillisuuden dynamiikkaa säätelevät kuitenkin useat paikallistason tekijät, minkä seurauksena ei ole täysin selvää, miten kasvillisuus on eri alueilla ja habitaateissa muuttunut. Koska kasvillisuuden jatkuva monitorointi on harvinaista pohjoisilla alueilla, vanhojen kasvillisuusaineistojen uudelleenkartoituksista on tullut tärkeä menetelmä muutosten havaitsemiseksi. Tutkin väitöskirjassani vuosikymmenten kuluessa tapahtuneita (23–60 vuotta) kasvillisuusmuutoksia Pohjois-Fennoskandian metsissä, puuttomilla kankailla ja tundralla uudelleenkartoitusten ja kokeellisen tutkimuksen avulla, ja kytkin ne ilmastonmuutokseen sekä tärkeimpiin paikallisiin tekijöihin. Yleisiä trendejä uudelleenkartoitetuilla puuttomilla kankailla olivat variksenmarjan (Empetrum nigrum ssp. hermaphroditum) voimakas lisääntyminen lumensuojaisissa habitaateissa sekä jäkälien väheneminen kaikkialla. Yhteisöjen kokonaismuutos oli voimakkainta eteläisillä puuttomilla kankailla, jossa se korreloi yhtä aikaa lisääntyneiden lämpötilojen ja laidunpaineen kanssa. Kokeellinen tutkimus tundralla osoitti, että kasviyhteisöt kehittyvät hyvin erilaisiksi paikallisten tekijöiden voimakkuussuhteista riippuen, jotka voivat joko hidastaa tai nopeuttaa ympäristömuutoksista johtuvia kasvillisuusmuutoksia. Metsien uudelleenkartoitus osoitti yhteisöjen kokonaismuutoksen olevan pitkällä aikavälillä suurempaa tuottavilla maaperillä lehtometsissä verrattuna karumpiin kangasmetsiin. Tutkimuksen mukaan maaperän tuottavuus on avaintekijä, joka ennustaa kasvillisuusmuutosten voimakkuutta ilmastonmuutoksen aikana. Tästä tärkeästä löydöstä oli aiemmin pääasiassa vain kokeellista tutkimustietoa. Yleisistä trendeistä huolimatta, muutokset diversiteetissä, kasviryhmissä ja yksittäisissä lajeissa olivat kuitenkin vaihtelevia ja usein habitaatti- tai aluesidonnaisia. Väitöskirjani tulokset, jotka muodostavat myös aikasarjan tuleville tutkimuksille, osoittavat kasvillisuuden monitoroinnin ja uudelleenkartoitusten olevan ensisijaisen tärkeitä, jotta kasvillisuuden dynamiikkaa voidaan ymmärtää paremmin nopeasti muuttuvissa olosuhteissa.