et al. 2005). Some halophilic ground beetle species
were more abundant in grazed than in abandoned
sites and vice versa. In general, no effect of manage-
ment on species richness was found for ground
beetles. Generally, spiders seem to be more depen-
dent on vegetation and litter structure than ground
beetles (Pe ́tillonet al. 2008).
The aforementioned experiment in the artificial
marsh of the Ley Bucht aimed to study the effect of
various stocking rates on the invertebrate fauna
(Andresenet al. 1990). For invertebrates, it may
not only be the plant species composition that is
important. Non-floweringAsterswere found only
at the higher salt marsh within the highest stocking
rate. The canopy height of the understorey was
higher in the abandoned site than in the grazed
sites (Andresenet al. 1990). In the third year of
cessation of grazing, positive effects for several in-
vertebrate groups were recorded for Collembola,
Aranea, Amphipoda, Coleoptera and Diptera. This
was attributed to the accumulation of litter, increase
of flowering plants and hence availability of pollen
and nectar and therefore higher aboveground bio-
mass for leaf- and stem-dwelling species (Irmler
and Heydemann 1986).Erigone longipalpis, a halo-
philic species, is the most important spider species
in thePuccinellia maritimacommunity. Other spe-
cies occur mainly in theFestuca rubracommunity
and cannot be considered halophilic, namelyOe-
dothorax retusus, Pardosa agrestis andPachygnata
clerki. WhereasErigone longipalpisstill occurred in
high abundance in theFestuca rubracommunity at
the start of the experiment in 1980, it has since 1982
moved to the lowerPuccinellia maritimaandSalicor-
niaspp. communities in the abandoned site. The
other species spread into the lower salt marsh at
different rates (Andresenet al. 1990). A distinct
zonation of the invertebrate communities was ob-
served in the first three years of the experiment. The
community diversity was highest in the abandoned
site, since communities of the higher marsh spread
into the lower marsh. In 1988, however, the com-
munity of the higher marsh had spread over the
entire elevation gradient, and completely replaced
the communities of the lower marsh in the aban-
doned site. Hence, eventually the community di-
versity was lowest at the abandoned site.
However, the number of species became highest at
the abandoned site, partly as a result of immigra-
tion from adjacent grassland. But the main reason
was that many species are damaged by grazing
(Irmler and Heydemann 1986). The authors espe-
cially stress the damaging effects of grazing on
many arthropod communities.
Two and three years after the start of the afore-
mentioned grazing experiment in Friedrichskoog
and So ̈nke-Nissen-Koog, invertebrates were moni-
tored. Mainly herbivorous and flower visitors were
positively affected by cessation of sheep grazing
and the resulting flowering ofAster tripoliumand
Plantago maritima. A minor part of the herbivorous
fauna profits from enhanced plant growth in mod-
erately grazed sites. Typical soil dwellers benefit
from grazing owing to greater amounts of bare
soil (Meyeret al. 1995).
In general, the community structure changes
from a dominance of detritivores to a dominance
of herbivores after the cessation of sheep grazing,
and after the cessation of cattle grazing. The num-
ber of species and individuals increases shortly
after the cessation of grazing, but after a longer
period of cessation of grazing typical halophilic
species may decrease. For the time being it is not
possible to discuss top-down or bottom-up con-
cepts with respect to the interaction between vege-
tation and invertebrates. Food web studies could
help in this discussion and are currently being car-
ried out.10.10 Exclusion of large herbivores: effects on birds
10.10.1 Migrating birds
In order to evaluate the importance of livestock
grazing for habitat use by geese in the Wadden
Sea, a large-scale inventory was made. Sixty-three
transects were established, subdivided over 38
sites. Only those sites with a stable and clearly
defined management regime for at least six preced-
ing years were included. Management was subdi-
vided into ‘long-term ungrazed’(>10 years), ‘short-
term ungrazed’ (6–10 years), ‘lightly grazed’ (low
stocking rates, i.e.4.5 sheep/ha or1 cow/ha),
and ‘intensively grazed’ (i.e. with high stocking
rate). Only marshes with sufficiently large surface144 APPLICATIONS