Abstract In this thesis, I studied the patterns in the assemblage composition as well as the biogeography and ecology of spring macroinvertebrates and bryophytes in Finland. My main objectives were to assess the importance environmental variables to macroinvertebrate and bryophyte assemblage composition in springs at the level of multiple spatial scales. In addition, I assessed the importance of springs in the boreal mire landscape, and sought the ecological and environmental determinants of a key species in boreal springs. In a large-scale study, I also examined the concordance between macroinvertebrates and bryophytes across boreal ecoregions, and assessed how macroinvertebrate assemblage variation corresponds to terrestrially-based ecoregions. Locally, spring macroinvertebrate assemblage structure displays high variation between different kinds of mesohabitats within springs, highlighting the importance of careful sampling of all habitat types in spring surveys. Helocrenes and other aquatic-terrestrial ecotone habitats harbour the highest species diversity and most spring-dependent species among spring habitat types. Further, spring-influenced mire patches were shown to have distinct cranefly assemblages in the mire landscape and to harbour higher cranefly diversity than mire types with lower trophic status, emphasising the importance of springs for mire biodiversity. Regionally, a red-listed spring-dependent caddisfly species appeared to be a surrogate for a high spring conservation value, indicating high overall species diversity and the occurrence of additional red-listed species. On a large geographical scale, intersecting the boreal ecoregions, a pattern of gradual change of benthic macroinvertebrate assemblage composition from south to north was detected, largely corresponding to terrestrially-derived ecoregions. However, the physical attributes of springs also need to be taken into account in bioassessment studies. Macroinvertebrate assemblage variation also correlated with physical habitat-scale variables, but not with changes in water chemistry. In contrast, spring bryophyte assemblages showed a distinct response to variation in water chemistry, but not to variation in physical habitat characteristics. Bryophytes and insect assemblages were concordant with each other on the large geographical scale, although the concordance was rather weak. Because of their different kind of responses to the physical and chemical variables, insects and bryophytes of springs are poor surrogates for each other in boreal springs.

Protected area site selection is generally carried out using biodiversity data as surrogates. However, reliable and complete biodiversity data is rarely available due to limited resources, time and equipment. Instead of drawing on inadequate biodiversity data, an alternative is to use environmental diversity (ED) as a surrogate in conservation planning. However, there are few studies that use environmental diversity for site selection or that evaluates its efficiency / unfortunately, no such example exists for Turkey, where biodiversity is high but our knowledge about it is unsatisfactory. Hence, this study was carried out to investigate the efficiency of environmental surrogates and the utility of different biological taxa in conservation planning. The objective was to find out the most efficient surrogates, either environmental or biological, for conservation planning, so that limited resources can be used more efficiently to establish an effective protected areas network. The study was carried out in northeastern Turkey, within the Lesser Caucasus ecoregion. The taxonomic groups considered include large mammals, breeding birds, globally threatened reptiles and amphibians, butterflies, highly threatened plants, and ecological communities. The distribution data was taken from a previous study, while climate and topographical data were obtained from various sources and produced through spatio-statistical techniques. Complementarity-based site selection was carried out with Marxan software, where the planning unit was the 100 sq.km. UTM grid square. Various statistical methods, including geographically weighted regression, principal components analysis, and p-median algorithm, were used to determine ED across the units. Performance of different approaches and different sets of surrogates were tested by comparing them to a random null model as well as representation success. Results indicate that endemic or non-endemic highly threatened plant species, butterfly species and ecological communities represent biodiversity better than other taxa in the study area. As such, they can be used on their own as efficient biodiversity surrogates in conservation area planning. Another finding is that highly threatened plant species are required to be used in the site selection process if they need to be represented well / in other words, they are their own surrogates. It was demonstrated that while ED alone can be used as a surrogate to represent biodiversity of an area, they are not as good as biodiversity surrogates themselves. It is also suggested that using species taxa with smaller distributional ranges or taxa that complement each other due to ecological differences as surrogates provide better results. On the other hand, ED might be a more suitable surrogate if resources are very limited or field work is impossible. In such cases, using ED in conjunction with one of the better biodiversity surrogates is probably the best solution.