Sitemap Contact Events New Home

Database of research projects FEDRA


Research actions



Research and applications > Projectendatabanken > Database of research projects FEDRA

Linking dispersal, connectivity, and landscape structure to produce habitat evaluation/restoration guidelines

Research project EV/16 (Research action EV)

Persons :

Description :

Habitat loss and fragmentation, associated with land conversion for human activities, are recognised as a major driving force in this dramatic reduction of Earth’s biodiversity.

Traditional, small-scale conservation efforts - either species or area oriented in their approach - have recently been recognised as being largely insufficient to preserve major parts of the biodiversity in a more sustainable way. Therefore, conservation biology has shifted towards landscape-level approaches, including ecological networks.

There is an evident need for sound, scientifically-based procedure to evaluate different network scenario’s on their conservation efficiency for a series of (target)-species and their habitats.

Dispersal was formerly largely considered from the point of view of the source population, i.e. as a form of mortality. It is now recognised as a major process enabling fragmented populations to overcome the local extinction events, whose probability increases with decreasing habitat patch size. Modelling metapopulation dynamics is a major scientific challenge, with the aim to estimate its viability under particular conditions. Such models must integrate characteristics of the species, and characteristics of the landscape: they are called "spatially explicit". Current metapopulation models treat the landscape as a map where each piece of land ("pixel" of the map) is either habitable (habitat patches) or not (matrix surrounding the habitat patches; it is supposed that dispersal may occur through the matrix - connectivity is not nil). Also, local population dynamics is usually ignored, so that the metapopulation dynamics only depends upon extinction and re-colonization rates.

The aim of this project is apply a spatially realistic modelling approach and to introduce local population dynamics into the metapopulation models.

This means on the one hand that the landscape structure underlying the model will be that of a real target area; habitat patches will be distributed and a connectivity indice will be calculated and mapped with the help of a Geographic Information System, basing on maps of land use and other pertinent environmental data. The connectivity indice developped by the UA team will be applied, and the G.I.S. expertise of the UCL team will allow the construction of pertinent landscape maps.

On the other hand, long term studies by both teams on butterflies and on rodents have already provided very important data allowing a precise modelling of the demography of these species. UCL team is specialized in the analysis of capture-mark-recapture data.

Finally, an important ingredient of metapopulation models is dispersal rate. It is extremely difficult to evaluate, even when radio-tagging methods are used - because dispersal is a rare event. The UA team is expert in molecular genetic techniques that will allow us to assess dispersal in the studied populations.

The expected output of this project is an estimation of the time to metapopulation extinction, of the probability of survival over long periods (100 years) and of the risk of genetic loss under various scenarios of habitat restoration, in order to provide guidelines and criteria to evaluate the "quality" of a landscape in conserving biodiversity.


1. Population dynamics: Standard capture-mark-recapture methods are applied, with technical adaptations specific to each target species.

2. Description of population dynamics: Demographic parameters (survival rates and recruitment) of populations inhabiting local habitat patches are estimated using «high-tech» methods (constrained linear models), based on probabilistic models describing the outcomes of capture-mark-recapture experiments.

3. Estimation of dispersal rates: For some species (butterflies) direct observation of movements between habitat patches is possible. For mammals, radio-tracking may occasionally reveal such movements; but we mainly rely upon molecular methods to identify immigrants by comparing their «genetic signature» to the profile of genotypes in the local population.

4. Evaluation of connectivity and of habitat quality: Physical and vegetal characteristics of the target regions are mapped. Using GIS tools and models of habitat patch occupancy and of «cost distance», maps of habitat quality and of resistance (against movement between habitat patches) are produced for the focal species and used in spatially explicit models of the dynamics of these species over the target regions.

Interactions between the different partners

This project is clearly interdisciplinary. UIA and UCL teams have recognised expertise in complementary fields, which they share in the frame of this project (including mutual training sessions). Specific UIA expertise is in molecular methods and in cost-distance modelling. UCL expertise is in the estimation of demographic parameters and in population viability analysis. Both teams have long-term data on several species (butterflies, rodents) in different regions, which are used in the present project.

Expected results/output

The expected output of this project is a tool to estimate the time to metapopulation extinction, the probability of survival over long periods (100 years) and the risk of genetic loss under various scenarios of habitat restoration, in order to provide guidelines and criteria to evaluate the "quality" of a landscape in conserving biodiversity.
The proposed methodology does not provide a global solution to the problem of biodiversity conservation at the scale of a landscape. A landscape must support a sustainable ecosystem, and not merely be viable for some emblematic species. The exercise must be performed for an array of species, hopefully encompassing all functional compartments of the ecological web. The best scenario for one species may be a bad scenario for another species. The globally optimal scenario necessarily will be a compromise. A perspective towards such a compromise may be to try to apply the concept of desirability functions used in the statistical problem of finding experimental conditions that optimise a multivariate response. Our proposal is the first necessary step towards this direction.



Eric Le Boulengé: We aim at understanding and modelling processes underlying (meta)population dynamics. We specialise upon demographic parameter estimation, experimental planning, analysis methods, and field techniques.

Michel Baguette: Since 1994 we initiated several projects on population ecology and genetics of different taxa in their natural habitats. The general aim is to investigate mechanisms responsible for the adaptation of populations to selective pressures of a changing, heterogeneous environment.

Erik Matthysen: We investigate the functioning of animal populations, focusing on interaction between individual and population processes. Fundamental questions concern adaptation in behavioural and ecological traits, and micro-evolutionary processes. Applied ones concern the impact of environmental changes on natural populations.


Eric Le Boulengé
Biodiversity Research Centre,
Environmetry and Geomatics Unit,
Place Croix du Sud, 2, B.P. 16,
B-1348 Louvain-la-Neuve

Michel Baguette
Biodiversity Research Centre
Ecology and Biogeography Unit
Place Croix du Sud, 4
B-1348 Louvain-la-Neuve

Erik Matthysen
Universiteit Antwerpen
Department of Biology
Laboratory of Animal Ecology
B-2610 Antwerp

Users Committee
Prof. dr. Paul Opdam
ALTERRA, Research Institute voor de Groene Ruimte
Postbus 47, NL-6700 AA Wageningen (The Netherlands)

Dr Geert De Blust
Institute of Nature Conservation
Kliniekstraat, 25
B-1070 Brussels

Dr Marc Dufrêne
Research Centre on Nature, Forests and Wood
Avenue Maréchal Juin, 23
B-5030 Gembloux

About this website


© 2017 Science Policy PPS