prof. Miska Luoto " Modern Geo-statistical Methods in Landscape Ecology"

04.02.2008 18051 views public lecture Medical science


Miska Luoto Department of Geography University of Oulu, Finland Which factors explain the distribution and the abundance of a species has continued to be a central question in landscape ecological and biogeographical research since the pioneering work "The distribution and abundance of animals" of Andrewartha & Birch in 1954. Nowadays, human activities increasingly affect the terrestrial biosphere, resulting in habitat loss and degradation which ultimately impair ecosystem function and services. Due to the rapidity of this process, there is utmost need to detect and predict changes in the natural and human environments and assess the spatial distribution of valuable sites and habitats. Current European environmental regulations often require assessments of the ecological effects of land use planning. Not only detailed knowledge of the present patterns of species richness and distributions but also accurate spatial predictions for more poorly known regions are needed. The challenge in land use planning is that biodiversity data, as accurate as possible, is required within a limited amount of time and over considerably large areas. Unfortunately, such data are often sparse and expensive to acquire by traditional field inventories. One potential means to complement the insufficient information concerning the distribution of species and suitable habitats for them is provided by species distribution modelling. Recently, species and habitat modelling has become one of the key issues in landscape ecology. This development is based on two trends: growth in the availability of remotely sensed (RS) and geographic information (GI) data and development of GIS techniques, and the progress of novel statistical techniques. In landscape ecology, statistical techniques in conjunction with RS-GI data and GIS techniques can provide effective means to identify the main environmental factors underlying the distribution patterns of species. Remote sensing generates a remarkable array of ecologically valuable measurements as well as the capacity to detect natural and human-induced land cover changes. Used in an integrative mode, RS-GI data can provide information about both historical and current habitat and land cover factors affecting biodiversity patterns. Hitherto, the benefits of remote sensing, GI data sets and spatial modelling have not been fully utilized in landscape ecological studies. Moreover, there are several critical issues in applying RS-GI based spatial models which have been insufficiently explored. These include the applicability of different RS data sets in biodiversity assessments, and the effect of scale, statistical techniques and model complexity on RS-GI based landscape ecological modelling. In this J.G. Granö lecture, I present several novel spatial approaches and techniques in landscape ecological research. I wish that the issues discussed in this lecture can have relevance in several fields of application of spatial data in geography and related environmental sciences.


Lisainfo veebis: http://www.finst.ee/grano/