International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
According to Turner et al. (2003) the direct remote sensing of
individual organisms, species assemblages, or ecological
communities from airborne or space borne becomes more and
more important. Remote sensing data provide increased
opportunities to develop quantitative models on the relationship
between species diversity and the diversity of land cover
elements (Noss 1990, Nagendra and Gadgil 1999). Providing
consistent and reproducible information on land cover at
different scales proves to be the main advantage of remote
sensing as a tool for both ecological analyses and biodiversity
assessment studies. Particularly regression analyses have been
broadly applied for the modelling of the spatial distribution of
species and communities up to date (Guisan et al. 2002). Thus,
in combination with regression analyses high resolution remote
sensing data may considerably help to assess biodiversity of a
region. Estimates of species richness of a region can then be
used to focus on targets in inventories so that appropriate levels
of sampling can be reached in these areas. Calculation of
potential biodiversity hotspots might be helpful for conservation
efforts in a region, e.g. for an assessment of the landscape itself
and for future protection planning.
The present study is focused on an assessment of lichen species
richness for six test sites in the Swiss Pre-Alps following a
gradient of land use intensity combining remote sensing
techniques and regression analyses. This study ties in with the
European Union Project BioAssess which aimed at quantifying
patterns in biodiversity and developing “Biodiversity
Assessment Tools” that can be used to rapidly assess
biodiversity. For the BioAssess project seven biological
indicators (soil macrofauna, collembola, ground beetles, plants,
butterflies, birds and lichens) as well as remote sensing based
indicators (non-biological) for a biodiversity assessment were
collected in the test sites for eight participating countries.
Lichens are mutualistic symbiotic organisms. Many species
have evolved a requirement for substrates that are themselves
by-products of advanced succession in more dominant
ecosystems. Lichens are affected by various forms of
anthropogenetic disturbance such as agricultural and forest
management (Scheidegger and Goward 2002), atmospheric
pollution and climate change (Nimis et al. 2002). These
disturbances can be detected using remote sensing data and
ecological modeling. Some studies show the combination of
lichens with remote sensing methods: e.g. in Nordberg and
Allard (2002) lichens have particularly been used as an
indicator of ecosystem disturbance or serve as indicators of
forest age (Nilsson 2004).
The objective of the present study is to correlate ecological
meaningful variables derived from airborne remote sensing data
with field sampled lichen species richness and to develop
regression models to predict lichen diversity on the investigated
test sites.
2. MATERIAL AND METHODS
2.1 Study-area
The study area is located in the northern Pre-Alps of the central
part of Switzerland in the region of Entlebuch which has been
accredited as an UNESCO Biosphere Reserve since September
2001. The region is characterized by a complex topography
with impenetrable gorges, rocky slopes, karst areas and
fluviatile deposits. The region covers an area of 395 square
kilometres and ranges from the montane (700 m) to alpine zone
(2300 m). It is mainly dominated by fragments of forest, rich
and poor pastures and natural grassland, mires as well as rocks
and small settlements. The study area consists of six landscape
types also called land use units (LUU) with an extent of 1 km x
1 km along the BioAssess gradient of land use intensity (see fig.
1). LUU1 contains more then 50 % old-grown forest and
represents extensive land use. LUU6 on the other end of the
gradient contains more then 50 % grassland and represents
intensive land use. The other LUUs are distributed according to
management intensity, which is defined after the percentage of
different land use classes inside the test areas.
"T. m, à 2m p y. f
ap = ; dh $f | A
pes,
RY Yen » i
N Fürstein i A
wdyrstemn s
AS St 4
V Sese ore.
Y, D4
184 pe
: A EN Schaónbü
s N EC S
cations of LUU1-6 with 6x16 sampli
p Af 4d
Lo
Figure 1.
2.2 Training and reference data sets
2.2.1 Field data — lichen relevés: A training data set is
required to calibrate the models whereas reference data are
required to validate the quality of the calibrated models. In our
case we used training data of the lichen surveys. A total of 96
sampling plots (6 x 16) were collected that form a grid of 200 m
mesh size (fig. 2). All 96 lichen sampling plots were set up by
differential GPS measurements with an accuracy of +/- 0.5 m.
. Lichen surveys were carried out in the years 2001-02 on the 96
sampling plots (16 per LUU) on a circular area of 1 ha (56.41 m
radius). Within each sampling plot 12 collecting sites were
selected randomly (fig. 3).
0 75 150 200 Meter $WSL. 2003
Figure 2. BioAssess sampling design for LUU6 with the 16
corresponding sampling plots (circles of 56 m radius)
846
Inte
2.2.
prec
ecol
this
info
Six
as t
appi
groi
of 1
gree
(75(
info
calc
botl
Furt
reso
Lan
(DS
sets
To
mea
betv
info
absc
grou
imag
orth
