07/2/14 6:31 PM

Belgium invigorates the taxonomic backbone

A seal in the Flanders coast
A seal in the Flanders coast

Flanders Marine Institute (VLIZ) and the Research Institute for Nature and Forest (INBO) are contributing to the central European LifeWatch components with a taxonomic backbone, and to the regional components with marine, freshwater and terrestrial observatories, several biodiversity data systems, web services and models.

The central taxonomic backbone facilitates the standardization of species information and the integration of the distributed biodiversity facilities. The taxonomic backbone aims to align several existing taxonomic databases (WoRMS, PESI, CoL, etc.) to make them more compatible, and aims to build access services to bring this taxonomic data to the LifeWatch infrastructure. A central data management taskforce is providing technical, logistic and financial support for upgrading and completing the component databases (WoRMS, FADA, SCAR-MarBIN, etc.). LifeWatch is also providing logistic and financial support for workshops and meetings with and between the taxonomic societies (Smebd, TDWG, PESI, EDIT, Marbef+).

For the marine observatory, monthly and seasonal monitoring campaigns are organized with the research vessel Simon Stevin, during which water and sediment samples are taking and several environmental parameters are measured. In addition to these sea-going monitoring campaigns, VLIZ has access to real time environmental parameter readings from continuously measuring buoys in the Ostend harbor. This sensor network is being expanded to the entire Belgian part of the North Sea.

For the freshwater-terrestrial observatory, the INBO started four main projects in 2012-2013: (1) GPS tracking network for large birds (large gulls and the western marsh harrier), (2) UAS monitoring of Natura 2000 habitats, (3) Acoustic telemetry for eel management plan, and (4) Remote groundwater monitoring.Two pilot projects were selected as well: (1) Monitoring seabird densities, and (2) Fish monitoring. All real time data generated by the observatories and sensor networks will be available through the www.lifewatch.beportal.

VLIZ and the INBO manage several biodiversity databases and data systems: WoRMS, EurOBIS, Marine Regions, Broedvogel, Florabank, VIS, Vlinder, Watina, Watervogels, Wildbeheer, etc. These databases are being updated on a regular basis and are integrated into the central LifeWatch infrastructure. During the construction phase of LifeWatch a lot of effort is made to digitize and thus rescue old, "lost" datasets, through so-called data archaeology activities.

One of the general objectives of LifeWatch is to identify, analyze and design onlinedata services, models and applications which could be valuable contributions to the data available from the supporting data systems, the taxonomic backbone, and the data generated by the observatories and sensor networks. Several web services are already available through the portal: taxonomic, geographic, tidal and data format validation services.

Earth and Life Institute (UCL)



During the preparation phase of the project, a users' requirements survey and a literature review have been carried out which allowed identification of three products. Images from spatial sensors are the basis of the project, but existing ancillary data are also integrated. In the near future, the launch of the Sentinel satellites will increase the images availability and increase the monitoring potential at the European level.

The first product is a European habitats database that integrates a large set of biotic and abiotic factors. The high spatial resolution images (RapidEye) are segmented into habitats and each of them are classified automatically. Habitats are the backbone of the LW-WB project. In addition, this backbone is enriched by a ready-to-use synthesis of existing ancillary data such as topographic and climatic variables. The minimum mapping unit is 0.5 hectare.

In order to reflect the complexity of the biotopes, an Object Oriented Data Model (OODM) has been selected for the habitats mapping. The main rationale is to propose an alternative to Corine Land Cover (hierarchical classification) and Copernicus products (continuous variables), and to take into account the strengths and weaknesses of remote sensing with respect to the objectives of the study. The main strength of remote sensing is the repeatability. The main weakness is the limited discrimination potential when it comes to the fine characterization of species, particularly compared with the field observation. An object-oriented database with a large component from quantitative variables has therefore been preferred to a direct classification.

The second product is an incremental layer to quickly identify disturbance or restoration of habitats. Monitoring change is particularly useful for the biodiversity community and this need has been highlight during the users' requirements survey. Remote sensing methods provide a good tool to identify the changes; it is a cheap method that can be used in large areas. Beside the categorical changes of land cover and/or land use, quantitative changes are also included in the changes layer.

The third product is an annual summary of ecosystems dynamic. Five temporal variables have been identified as the main needs: phenology of vegetation, sunshine, snow, floods and fire. The average profile of these time series will be calculated on the basis of satellite images of medium resolution (MODIS and MERIS), except for the sunshine that primarily relies on a GIS analysis.

EU Research & Innovation
EU-Brazil OpenBio