Monitoring and predicting elusive species colonisation –
Application to the otter in the Cévennes National Park (France)
The movements of some wildlife species are secretive and the process
of change in their population distribution is often poorly documented.
The objective of this research is to develop an integrative approach
to monitor and predict the colonisation of elusive species. The study
focuses on the Eurasian otter (Lutra lutra L.) in the mid-mountainous
landscape of the Cévennes National Park (France). Statistical
analyses of 15-years field survey, GIS-modelling and non-invasive genetics
are combined to (1) optimise the species detection; (2) define a relevant
framework for data analysis; (3) retrace the settlement chronology; (4)
get genetic information about colonisers; and (5) predict the spatial
process of colonisation.
Findings were both methodological and specific to otter ecology. Looking
for spraints, i.e. otter faeces, along riverbanks, we showed that one
long (> 5.5 km) or multiple short surveys (e.g. 13 x 50 m) were necessary
to predict otter presence with 95% efficiency in a basin at a colonisation
front. The marking density and frequency were relevant indicators of
the settlement process. An original modelling approach validated the
use of functional distances from detection places to predict otter occurrence
where it was not detected, which may contribute to address the issue
of ‘false-absences’. By this method, we retraced the time
sequence of colonisation at the basin scale showing a mean rate of about
10 km / year. Extraction and analysis of DNA from spraints were used
as the only efficient technique to better understand the population distribution
dynamics (73% PCR amplification success; 98% genotyping reliability;
0.6% probability to obtain an identical genotype among relatives). 70
individuals were genotyped describing the population along 1650 km of
rivers. Sex ratio was 57% males. Individual movements were observed up
to 60 km. Two genetic clusters were identified (Fst: 0.102) partially
separated by watershed lines. Assignment tests demonstrated that rivers
and marshy passes between valleys were colonisation paths and that steep
ridges (> 8° on average) were effective barriers.
Combining different advanced tools in a multidisciplinary approach proved
to be efficient in inferring animal movements from indirect signs only.
Such approach should be used with population dynamics models in order
to predict space and time status changes of elusive and vulnerable species.
Please contact me (firstname.lastname@example.org)
if interested for the full text - PDF (included publications are still “submitted” or “in
PhD from Université catholique de Louvain
Unit of Environmetry and geomatics