Natural and anthropogenic environmental variables pressures exert fitness pressures on our flora and vegetation and necessitate an understanding of the mechanisms of plant responses to those pressures. Overgrazing and desertification are examples of the stressors that affect plant community composition, on both large and small scales. In order to investigate the plant community response to these stressors it has been shown that it is more practical and informative to group plant species according to their functional characteristics into functional types than attempting to predict plant response solely based on previous range distribution data. The present study involved a survey of the dunefield vegetation of several privately owned farms as well as a communal farm in the southern Kalahari. The farms were subjectively chosen to represent a gradient from underutilized to overutilized. Plant community surveys were conducted on the farms and the dominant plant species were sampled and researched for selected functional traits to explore possible relationships between traits and grazing pressure. Ordination techniques were used to identify plant functional grazing response types. The identification of plant functional types through easy measurable traits will be a great aid in modelling and longterm environmental monitoring approaches. Of the three dunefield habitats, the dune street habitat was identified as the most vulnerable to overgrazing. A reduction in cumulative cover and an increase in bare surface cover indicated increased grazing pressure in the dune crest habitat. In the dune streets shrub cover and annual grass cover increased when veld was overgrazed. Functional analysis of 112 Kalahari plant species for a set of 47 plant traits showed that species clustering were strongly related to vegetative characteristics. The emerging clusters were strongly associated with major life forms, which comprised phanerophytes of different size classes, geophytes and herbaceous perennials, as well as herbaceous annuals. These clusters were identifiable even with the inclusion of reproductive traits. Selected vegetative traits are good indicators of plant response to land-use intensity since these traits lend competitive advantages to plant persistence under utilisation pressure. Plant height, spinescence, specific leaf area, leaf dry matter content, leaf sclerophylly and stem density were found to be examples of such traits. Phanerophyte and chamaephyte dominated functional groups were mostly associated with lower range condition and hemicryptophyte dominated groups with higher range condition, whereas herbaceous therophyte dominated groups did not show any specific association. / Dissertation (MSc)--University of Pretoria, 2011. / Plant Science / unrestricted

Throughout the past century, vegetation scientists have been studying plant communities to develop classification standards for global mapping purposes. In Turkey, although there are several vegetation classification schemes in use by botanists, foresters or geographers, none is widely accepted by all for vegetation mapping. In this study, a hierarchical, ecologically meaningful, physiognomic-floristic classification system was adapted and developed for terrestrial vegetation in Turkey. The system has eight hierarchical levels, with the alliance and the association as its lowest two floristic levels, and the classes are designed to be easily detected in the field or indirectly through remote sensing. To test of its utility, a supervised vegetation classification of the whole Mediterranean Region of Turkey was carried out using Landsat ETM images. The accuracy of the classification ranged between 55% and 69% depending on the level of hierarchy. A further test for its ecological utility was carried out by comparing identified vegetation classes with breeding bird communities derived from data gathered through field observations at 193 ecological communities at 83 sites. The proposed classification scheme has proven to be reasonably accurate when widely available satellite imagery is used and ecologically meaningful as shown by a high concordance with observed bird community patterns. It is suggested that this new system can be safely applied to other regions of Turkey for purposes of vegetation mapping, species habitat modeling, and nature conservation if proper image set and ancillary data is used.

Thesis (Ph. D.)--University of Washington, 2000. / Vita. Includes bibliographical references (leaves 107-124).

A four year experimental study of initial stages of succession on different substrata (sand, peat and gray clay overburden from brown coal mining) was carried out. The substrata were transplanted into two different landscape types in the Czech Republic. Control plots were filled with autochthonous topsoil treated with high temperature vapor. Plots with un-treated autochthonous topsoil were also established. Plant species cover and several soil-biological parameters (soil microbial biomass, basal soil respiration, exchangeable pH(KCl), organic-C extractable by potassium sulfate solution, and nitrate-N content) were measured once a year. Obtained data were mainly processed by unimodal (plant cover data) and liner (soil-biological data) ordination methods. Biological species traits (life span, seed dispersal vector, seed bank type, ecological strategy type following Grime, life form following Raunkiaer) were used to characterize early stages of plant succession. Vegetation and measured soil-biological parameters, except pH(KCl), were significantly changing during four growing seasons. Variance partitioning analysis based on plant species cover data indicated the landscape type to be the main governing factor and physico-chemical characteristics of substrate to be less important during the studied initial stages of succession. The same analysis based on soil-biological parameters showed the reverse result. Great variability of plant species ecological traits was noticed during four growing seasons. The earliest plant colonizers were mainly ruderal, anemochorous terophytes with persistent seed bank. The number of species with C-strategy type increased during the study. The relationship between the average plant species cover and measured soil-biological parameters was also examined during the experimental study of early succession. Direct and indirect ordination revealed the average sum of plant species cover to be weakly positively correlated with soil microbial biomass, basal soil respiration, nitrate-N, potassium sulfate extractable organic-C, metabolic quocient (qCO2) and negatively correlated to exchangeable pH(KCl). The relationship was significant only in the case of microbial biomass and potassium sulfate extractable organic-C. More detailed studies of the relationship between aboveground plant species biomass and soil microbial communities in initial stages of succession are required. Comparability of presented study is limited because not many experimental investigations of similar kind have been presented until now.