Biodiversity ecological analysis at regional level and spatial

Biodiversity ecological analysis at regional level and spatial modeling of species distribution Prof. Dr. Tatiana Rogova Kazan State University General Ecology Department

Biosphere Ecological problems 1. p 2. p 3. p 4. p 5. p Water resources deficit Water and air pollution Biodiversity decreasing Deforestation and desert increasing Landscape fragmentation Every country on the planet have one or an other problem, or all of them at the same time

Species Biodiversity p Necessity of conservation for all species: -There is no useless species, every one takes specific place in ecosystem food nets. Loosing one species may lead to extinction of another p -Mankind have not neglect any one species, because its useful effect may be evident only in future p

Necessary prerequisites for successful biodiversity conservation 1. p 2. p 3. p 4. p Social & Political Economical Legislative Scientific (Last but not the least)

Biodiversity conservation World Biodiversity Conservation Strategy p European Biodiversity Conservation Strategy p Biodiversity Conservation Strategy of Russian Federation p Regional Biodiversity Conservation Strategy p

SCIENTIFIC BACKGROUND AND FUNDAMENTAL KNOWLEDGE p INFORMATION THEORY p & COMPUTER SCIENCE p p p BIOSPHERE THEORY GEOGRAPHY PEDOLOGY ENVIRONMENTAL SCIENCE CARTOGRAPHY p p p GENERAL SYSTEM THEORY & MODELING BIOLOGY VEGETATION SCIENCE GENERAL ECOLOGY ANIMAL & PLANT ECOLOGY BIOGEOGRAPHY

Spatial biodiversity heterogeneity Realized species niche in environmental superspace (из Hirzel et al. , 2002) Spatial biodiversity heterogeneity Anisotropy of habitat space Geographical space Time space Biotic space Ecological factors space Anthropogenic impact

Biodiversity analysis at regional level (Tatarstan)

Ecological Biodiversity Ecological regions Scale 1: 200 000 Vegetation map (Tatarstan Republic)

Ecosystem diversity

Spatial-ecological Analyses & Biodiversity Modeling Necessary data and methodic approaches: p Computer information system on regional biodiversity p Accessibility and completeness of environment data p Appropriate mathematical methods

Data and information sources. Data and information used in flora biodiversity research come from different kind of sources. Among them are: 1. field data with species inventory; 2. topographic and soil maps; meteorological data, etc.

Forming and keeping information data base “Flora” in regional GIS p By using information system «Flora» cadastre of vascular plants worked out as a component of regional GIS. Cadastre includes reference information: plant taxa reference contains more than 1600 species, vegetation formations - 19, landscape regions - 11, Information on areal, ecological and coenotic elements of flora, plant life forms. p Supporting flora cadastre in regional GIS system allows to fulfill exact coordinate attachment of geobotanical plots and species locations. p The cadastre information can be used for cartographic modeling of species spatial distribution for the purpose of evaluating tendencies of flora chorology, landscapes dynamics, monitoring endangered species and flora diversity.

Environmental information at regional level (Tatarstan) Absolute altitude Earth surface temperature (may) Hydrotermic coefficient Plant species are associated in groups by environment ecotope conditions, so environmental variables are informative features for ecological analysis of biodiversity. NDVI (may)

Data and information sources. To represent physical characteristics of habitat as patch types, we need: first to combine all the various spatial resolution and uncertainty level information using data fusion technique then apply some kind of habitat characteristics classification and ordination using predefined rules or exploratory data analysis to receive patch types description related to the biodiversity and species spatial distribution Among with traditional (statistical) methods, modern techniques, such as artificial neural nets, could be used to solve the task.

Species Composition Probabilistic Model of Vegetation Cover Probabilistic approach based on: p Stable state of flora composition according to the species pool concept (Abbott, 1977; Abbott, Black, 1980; Van der Maarel, 1997; Ewald, 2002) p Species ecological individuality hypothesis (Ramenskiy, 1938; Witteker, 1980) p Stochastic species distribution on vegetation continuum as a reaction on environmental gradients

Flora and vegetation gradient analyses p p p Direct and indirect vascular plant species ordination Probabilistic model application SOM method application for vegetation ordination

Indirect gradient analyses Field flora species list Probabilistic species list

Direct gradient analyses Field flora species list Probabilistic species list

Species communities ordination А B • Humid. Clay. Sand. Dry C • • • D Ordination of ecologocoenotical species groups at local level (micropatches) 1 -15 numbers of plots; Isoclinical lines show density of micropatches; line thickness show % of ecological group cover in the patches. В pine wood C nemoral, D boreal complexes

Classification & Ordination Classes hierarchy Data structure exposure on similarity indices Scaling Initial data Classes Ordination

Classification & Ordination Kohonen neural nets (SOM) and Generative Topographic Mapping (GTM) provide a new means to combine species communities classification and ordination Probabilistic species composition model developed give means for each site actual species pool estimation for usage in communities classification and ordination taking in account vegetation composition continuum. Species communities classes, mapped on the ordination plane, reflect continuous change in species community composition. This non-linear, axis-free ordination simultaneously show all the gradient changes: from nemoral to boreal communities; from arid to wetland communities; from ruderal and disturbed communities to intact communities.

Species communities ordination received by the means of SOM and GTM could be used also to discover species, communities and environmental variables interrelation Mercurialis perennis L. Hydrotermic coefficient

Spatial modeling of floristic complexes : 1 - expert complement of failing data (raster model); 2 - mathematical interpolation on Delone triangulation method; 3 – spatial autocorrelation Kriging –method.

Modeling results: nemoral (left), boreal (right) complexes

Spatial modeling of species distribution Species presence probability forecast (Vaccinium myrtilus)

Species distribution dynamics – Picea x fennica; – Ephedra distachya. : – Aster alpinus; – Cicerbita uralensis.

Map modeling of potential habitats for rare plant species (left – boreal species, right – south species at the north border of areal space)

Forest fragmentation North, RT (Tulyachinskiy district) West, RT (Kaybitckiy district)

Forest fragmentation(Bavlinskiy district, Tatarstan) 1953 г. 1975 г.

Vegetation dynamics in anthropogenic landscapes (left – vegetation map, M. Markov, 1944; right – actual vegetation map) Vegetation map (M. Markov 1944) Actual vegetation map

Forest dynamics in anthropogenic landscapes Forest map (F. Boyko, 1986)

Modeling prognoses of retrospective and potential regional biodiversity Methodological approach: expert system forming based on fuzzy logic Expert system – complex of data bases: 1. information about investigated object (biological, environmental, etc. ); 2. list of formal rules for making system decision; 3. software programs for checking the results and truthfulness determination The expert systems based on the GIS databases open a possibility in ecosystem dynamics modeling and paleolandscape reconstruction. Expert system formed for vegetation dynamics reconstruction of foreststeppe region is based on the long settling history of Tatarstan region.

Modeling prognoses of retrospective and potential regional biodiversity Anthropogenic loads (influence) distribution (X-XIII c. ) based on relief model and archeological monuments map.

Modeling prognoses of retrospective and potential regional biodiversity Forest vegetation IX-X c. (reliability scale: 1 - presence; 0 – absence) Forest vegetation XIV c. (reliability scale: 1 presence; 0 – absence)

Modeling prognoses of retrospective and potential regional biodiversity Forest vegetation ( XVII–XVIII) (reliability scale: 1 - presence; 0 - absence) Forest vegetation at the end of XX c. Actual data

Thanks for attention