Local and Regional Scale Determinants of Biodiversity Patterns in Boreal Agricultural Landscapes

This thesis examines the local and regional scale determinants of biodiversity patterns using existing species and environmental data. The research focuses on agricultural environments that have experienced rapid declines of biodiversity during past decades. Existing digital databases provide vast opportunities for habitat mapping, predictive mapping of species occurrences and richness and understanding the speciesenvironment relationships. The applicability of these databases depends on the required accuracy and quality of the data needed to answer the landscape ecological and biogeographical questions in hand. Patterns of biodiversity arise from confounded effects of different factors, such as climate, land cover and geographical location. Complementary statistical approaches that can show the relative effects of different factors are needed in biodiversity analyses in addition to classical multivariate models. Better understanding of the key factors underlying the variation in diversity requires the analyses of multiple taxonomic groups from different perspectives, such as richness, occurrence, threat status and population trends. The geographical coincidence of species richness of different taxonomic groups can be rather limited. This implies that multiple geographical regions should be taken into account in order to preserve various groups of species. Boreal agricultural biodiversity and in particular, distribution and richness of threatened species is strongly associated with various grasslands. Further, heterogeneous agricultural landscapes characterized by moderate field size, forest patches and non-crop agricultural habitats enhance the biodiversity of rural environments. From the landscape ecological perspective, the major threats to Finnish agricultural biodiversity are the decline of connected grassland habitat networks, and general homogenization of landscape structure resulting from both intensification and marginalization of agriculture. The maintenance of key habitats, such as meadows and pastures is an essential task in conservation of agricultural biodiversity. Furthermore, a larger landscape context should be incorporated in conservation planning and decision making processes in order to respond to the needs of different species and to maintain heterogeneous rural landscapes and viable agricultural diversity in the future.

Advanced chemometric methods: Applicability on industrial data

Recent years have produced great advances in the instrumentation technology. The amount of available data has been increasing due to the simplicity, speed and accuracy of current spectroscopic instruments. Most of these data are, however, meaningless without a proper analysis. This has been one of the reasons for the overgrowing success of multivariate handling of such data. Industrial data is commonly not designed data; in other words, there is no exact experimental design, but rather the data have been collected as a routine procedure during an industrial process. This makes certain demands on the multivariate modeling, as the selection of samples and variables can have an enormous effect. Common approaches in the modeling of industrial data are PCA (principal component analysis) and PLS (projection to latent structures or partial least squares) but there are also other methods that should be considered. The more advanced methods include multi block modeling and nonlinear modeling. In this thesis it is shown that the results of data analysis vary according to the modeling approach used, thus making the selection of the modeling approach dependent on the purpose of the model. If the model is intended to provide accurate predictions, the approach should be different than in the case where the purpose of modeling is mostly to obtain information about the variables and the process. For industrial applicability it is essential that the methods are robust and sufficiently simple to apply. In this way the methods and the results can be compared and an approach selected that is suitable for the intended purpose. Differences in data analysis methods are compared with data from different fields of industry in this thesis. In the first two papers, the multi block method is considered for data originating from the oil and fertilizer industries. The results are compared to those from PLS and priority PLS. The third paper considers applicability of multivariate models to process control for a reactive crystallization process. In the fourth paper, nonlinear modeling is examined with a data set from the oil industry. The response has a nonlinear relation to the descriptor matrix, and the results are compared between linear modeling, polynomial PLS and nonlinear modeling using nonlinear score vectors.

Outcomes of cervical human papillomavirus (HPV) infections among mothers in the Finnish Family HPV Study

To understand the natural history of cervical human papillomavirus (HPV)-infections, more information is needed on their genotype-specific prevalence, acquisition, clearance, persistence and progression. This thesis is part of the prospective Finnish Family HPV study. 329 pregnant women (mean age 25.5 years) were recruited during the third trimester of pregnancy and were followed up for 6 years. The outcomes of cervical HPV infections were evaluated among all the mothers participating in the study. Generalized estimating equation (GEE)-models and Poisson regression were used to estimate the risk factors of type-specific acquisition, clearance, persistence and progression of Species 7 and 9 HPV-genotypes. Independent protective factors against incident infections were higher number of life-time sexual partners, initiation of oral contraceptive use after age 20 years and becoming pregnant during FU. Older age and negative oral HR-HPV DNA status at baseline were associated with increased clearance, whereas higher number of current sexual partners decreased the probability of clearance. Early onset of smoking, practicing oral sex and older age increased the risk of type-specific persistence, while key predictors of CIN/SIL were persistent HR-HPV, abnormal Pap smear and new sexual partners. HPV16, together with multiple-type infections were the most frequent incident genotypes, most likely to remain persistent and least likely to clear. Collectively, LR-HPV types showed shorter incidence and clearance times than HR-HPV types. In multivariate models, different predictors were associated with these main viral outcomes, and there is some tentative evidence to suggest that oral mucosa might play a role in controlling some of these outcomes.