Research team:
Coordinator
Researcher´s:
Petra Šímová, Tomáš Klouček, Adam Chrumko, Ondřej Lagner, Jiří Prošek, Petr Klápště, Jaroslav Nauč, Šárka Krčílková
Evaluation of fragmentation and biotope connectivity in the Czech Republic in connection with species with different ability to spread (ACTIVITY 2)
The aim of ACTIVITY 2 is to prepare map and other foundation analyzing actual fragmentation and connectivity of ecosystems in the Czech Republic considering species with different ability to spread. Another target is to assess the time trend of connectivity development, both in detailed and generalized measuring scale.
The evaluation will provide foundation for strategic planning within the Czech Republic. Detailed studies will provide information for targeted species protection and biotope management in local and regional level.
Research state:
Fragmentation of ecosystems is closely related to the loss of landscape (biotope) connectivity. Connectivity indicates the degree to which a landscape allows the propagation of species (individuals, genes) between source areas. Functional connectivity of a landscape takes into account explicitly the different behaviour of an organism in various landscape elements (Tischendorf and Fahrig, 2000), and determines what part of the landscape is available for a specific type of organism from a given place in the landscape (Saura et al., 2011). The impacts of fragmentation of ecosystems and the loss of biotopes and their connectivity are thus species-specific. A lack of landscape connectivity and subsequent isolation of biotopes may adversely affect the dissemination of seeds, the gene flow, migration of animals and other ecological processes (Saura et al., 2011). Generally speaking, the fragmentation threatens more species with a limited propagation ability species with a narrow ecological valence.
The aim of this part of the project was to (i) prepare maps and other data analysing the fragmentation and connectivity of ecosystems in the Czech Republic in terms of species with different abilities of propagation, and (ii) to evaluate temporal trends in the fragmentation and connectivity. The activities leading to the fulfilment of these objectives focused on the scale of the Czech Republic and the possibilities for in-depth research.
Result: A tool to automate the calculation of connectivity
Indices for species-specific quantification of a certain type of connectivity of environments based on the graph theory can be calculated using the freeware program Conefor. A limiting factor, however, is that Conefor allows calculations for the one location. If we want to express a connectivity space so that the entire territory is divided into dozens or hundreds of mapping units (mapping project in blocks and districts throughout the Czech Republic), we are faced with a task that is not time-consuming without automation. Appropriate automation of this calculation is one of the results of the project. The process is facilitated by our specialised ArcGIS Toolbox. The tool is programmed in Python and is executable in ArcGIS 10, which communicates with Conefor.
Result: Map portal
To allow other researchers and bodies active in the decision-making process in the protection of nature and landscape to the spatial and temporal changes in connectivity of biotopes we created in the project a web map portal, www.fzp.czu.cz/mapyfragmentace, mediating the intermediate results obtained through modelling of the connectivity of biotopes (main types of landscape cover) on the scale of the Czech Republic. The map portal is another important result of the project. The portal content is divided into two parts, one of which takes into account the main Conefor metrics, expressing the connectivity, including the ability of a species to propagate in the form defined by the dispersion distance. The second part provides maps of the key metrics of fragmentation and heterogeneity, such as the density of the edges, the size of patches and indices according to Shannon and Simpson. The basis for all the calculations were CORINE Land Cover data from 1990, 2000, 2006 and 2012. Connectivity and fragmentation is expressed at the level of districts in the Czech Republic and at the level of squares, which are used in species distribution atlases. The individual maps show both the values of the metrics for those periods and percentages of the changes between 1990 and 2012. A more detailed examination of the changes is allowed by intuitive tools for overlapping of the single layers.
The Conefor metrics, including the ability of a species to spread (distance distribution), are calculated for each of the eight main types of environments based on the CORINE categories for each type of environment a separate mapping application is created. All the applications are identical in structure and operation, allowing users easy navigation and use of the data portal without a deeper understanding of GIS. In principle, some knowledge of current mapping applications like Google Maps suffices. Due to the large number of the connectivity maps (20 dispersion distance x 8 types of biotopes x 2 spacing units x 2 indices x 4 timeframes = 2,560 maps) for each combination of habitat, index, spatial unit and time period in the application includes maps for a distribution distance of 500 m and for a predetermined threshold distance. The value of the dispersion distance indicates that for species with a greater propagation capacity the differences in connectivity shown in the maps were not significantly different (significantly different in a statistical sense). The specific values of these 'disruptive' dispersion distances for that type of habitat, index and space unit can be read directly from the names of each connectivity layer. They range from 2,500 to 7,000 metres.
Result: Spatially detailed identification of biotopes in a fragmented landscape - Analysis of satellite images in very high resolution
One of the problems arising at the research habitat requirements of species is the use of a sufficiently accurate environmental data to calculate predictors usable in species-habitat and species distribution models (SDM). The knowledge gained through these models may provide valuable input for decision support and direction of management measures in a fragmented landscape. We must always consider, however, the level of detail that can be captured in mapping environmental characteristics. Generally speaking, for SDM the degree of spatial detail, in the sense of resolution, the environmental data must be in line with the positional accuracy of the species data.
To explore the detailed relationships in a fragmented landscape we use data with a relatively precise location (using a tourist GNSS receiver with a positional error between 5 and 10 metres.
The only existing adequate spatial data set concerning the biotopes of the Czech Republic is Biotope Mapping Layer (BML) generated within the system NATURA2000. Its usefulness for research on animal species in fragmented landscape components is however limited, on the one hand, by the species being distinguished, which does not necessarily reflect the parameters relative to the animal’s environment, and due to the difficulty of nationwide terrestrial mapping it does not as a rule match the time of acquisition of the habitat data and the acquisition of species data. Problematic areas also more frequent repeats in order to capture during the season, capturing the development of successional areas.
For these reasons, the project tested the possibility of using satellite data in very high resolution (VHR) for mapping biotopes in a landscape, where there are particular types of biotopes (in terms of terminology, remote exploration of the Earth for each classified class) scattered in small fragmented patches, where the environment creates a fine-grained spatial mosaic.
Based on the analysis of the two different anthropogenically affected areas with a fine-grained mosaic of the biotopes (the Doupov mountains affected by long-term military operations and post-emission landscape plateaux of the Ore Mountains), we verified in the project it was possible to classify the VHR images to produce thematic maps of the biotopes with the standard accuracy for these types of landscapes. In the standard landscape of the Czech Republic similar or better results can be expected. We found suitable products for evaluation of the fragmented biotopes satellite images WorldView-2, which provides and supports since 2009 direct customers’ tasking of images with a spatial resolution of 1.84 m in 8 multispectral bands. We consider the created methodological framework for the classification of these data is considered to be one of the project outputs.
Result: The trends of fragmentation and connectivity in a generalised view
The connectivity maps created in the first part of the project for the map portal correspond to the possibilities of propagation of species that are not significantly limited by linear structures when moving across a landscape. In other words, the analysis and resulting maps show changes in connectivity that are not directly caused by linear structures or other terrestrial obstacles.
Based on the results of the project, it must be stressed that even without this the connectivity of some biotopes decreases. The loss of connectivity threatens some types of non-forest biotopes. Management measures to mitigate the consequences of landscape fragmentation should be on this scale directed primarily at increased connectivity of scrubland biotopes, nearby meadows and agricultural areas with nature-friendly features. For all the rated types of biotopes it is also true that activities should be geographically oriented to the places mentioned in the final project report and others derivable from the map portal for decision support, www.fzp.czu.cz/mapyfragmentace. Those recommendations can be applied to animals with different propagation abilities according to the dispersion distances on the order of hundreds of metres or several kilometres.
Result: The trends of fragmentation and heterogeneity in spatial detail
The analysis of time trends on the detailed spatial scale was focused on changes in the structure of the agricultural landscape since the 1950s to the present, focusing on a potential turning point in the evolution of landscape structures and representations of close to nature elements in the agricultural landscape after the change of the socio-economic system in the Czech Republic after 1989.
The input for the detailed analysis of thematic layers biotopes for 357 sample localities in a square 1 x 1 km in size. The locations are distributed randomly in the landscape with agricultural matrix
For all the localities we created detailed thematic layers of biotopes based on vectorisation and visual interpretation of aerial photographs. The layers covering three time horizons: the 1950s showing the fine-grained structure of agricultural land before collectivisation, the turn of the 1990s showing the status of agricultural land in the period around 1989, and the year 2013 representing the current state. The images were vectorised on a scale 1: 1500 following a procedure that minimises the emergence of false changes, i.e. vectorisation of the current state according to the current detailed colour images and the subsequent adjustment of the layers according to historical photographs.
In the detailed evaluation of the scale trends in connectivity and fragmentation in the agricultural landscape it can be said that the problem of fragmentation biotopes of the type of herbal patches and belts, brushwood, parkways, and the like in an agricultural landscape does not consist in their overall losses or their increasing fragmentation into patches of insufficient size. The problem is the fragmentation of such biotopes with extensive expanses of arable land. Managementovou priority should therefore be not a simple increase in areas of stabilising elements in the agricultural landscape, but their targeted placement within large blocks of arable land, increasing the connectivity of these elements, and hence increasing the permeability of farmland for many species, particularly species with small dispersion distances. These findings are in principle consistent with the outputs from Activity 1 of this project, and landscaping is the anticipated significant systemic tool for the implementation of improvement measures.
Summary of benefits from project Activity 2
This part of the project brought a number of research and practical results. In addition to the proposed management priorities in a global perspective in the Czech Republic and the detailed spatial scale, aimed at the small stabilising elements in the agricultural landscape the prime benefit of this project is the methodological evaluation of the fragmentation and connectivity tools using GIS and remote exploration of the Earth. Part of the activities for the development of knowledge in the field of biotopes and fragmentation and options for mitigation of its consequences can be followed thanks to the project and disseminated by specialists in GIS and exploration of the Earth. Those project outputs provide recommendations for the creation of digital thematic maps of biotopes based on satellite images in very high resolution and a software tool for calculating graph-oriented connectivity metrics based on any thematic map of biotopes for any territory and any mapping unit. Experts active in decision-making in landscape management and the protection of nature and landscape, but also the general professional and lay public, can make use of the freely accessible web portal www.fzp.czu.cz/mapyfragmentace maps. This part of the project was, ultimately, a significant educational contribution. Two student members of the team contributed to the overall solution with their final-year theses (Adam Chrumko, Jaroslav Nauč, supervisor Petra Šímová). One of the theses was received a prize from the Rector of the Czech University of Life Sciences in Prague and one a prize from the Minister of the Environment of the Czech Republic.
Photogallery
