The abandonment of military areas leads to
succession processes affecting valuable open-land habitats
and is considered to be a major threat for European butterflies.
We assessed the ability of hyper spectral remote
sensing data to spatially predict the occurrence of one of
the most endangered butterfly species (
Hipparchia statilinus
)
in Brandenburg (Germany) on the basis of habitat
characteristics at a former military training area. Presence–
absence data were sampled on a total area of 36 km
2, and
N = 65 adult individuals of
Hipparchia statilinus could be
detected. The floristic composition within the study area
was modeled in a three-dimensional ordination space.
Occurrence probabilities for the target species were predicted
as niches between ordinated floristic gradients by
using Regression Kriging of Indicators. Habitat variance
could be explained by up to 81 % with spectral variables at
a spatial resolution of 2 × 2 m by transferring PLSR
models to imagery. Ordinated ecological niche of
Hipparchia
statilinus was tested against environmental
predictor variables. N = 6 variables could be detected to
be significantly correlated with habitat preferences of
Hipparchia statilinus. They show that
Hipparchia statilinus
can serve as a valuable indicator for the evaluation of
the conservation status of Natura 2000 habitat type 2330
(inland dunes with open
Corynephorus
and
Agrostis
grasslands) protected by the Habitat Directive (Council
Directive 92/43/EEC). The authors of this approach, conducted
in August 2013 at Döberitzer Heide Germany, aim
to increase the value of remote sensing as an important tool
for questions of biodiversity research and conservation.
