The Structure and Dynamics of Agricultural Landscapes as Drivers of Biodiversity - Landscape Ecology for Sustainable Environment and Culture - page 286

moment of bee sampling and the proportion of two types of crop rotations. For the
last variable, a crop rotation map summarising land-use history over a period of
five years was realized. As cereals and grassland are the dominant land uses in the
LTER site we distinguished two classes of crop fields: the fields that were sown
only with cereals (wheat and maize) during the last five years and fields where the
crop rotation included from one to four years of grassland (referred to as ''mixed
fields''). Solitary bees were trapped on 50 field margins, 15 of which were along
oilseed rape fields and the 35 others were randomly located along other fields.
Landscape composition was quantified in square windows centered on sampling
points. Three window sizes were chosen, covering the range of relevant scales for
flight and foraging distances of solitary bees (400, 800 and 1,200 m in width).
We found contrasted effects of non-flowering crops according to the type of
margin and the spatial scale (Fig. 14.4 ). Solitary bee abundance in margins of
oilseed rape fields deeply increased with the proportion of non-flowering crops at
the moment of bee sampling at the finest spatial scale while it remained unchanged
in margins of non-oilseed rape fields (Fig. 14.4 a). This result shows that the
attractiveness of mass-flowering crops depends on the quality of the surrounding
landscape: the use of oilseed rape by solitary bees is higher when the surrounding
area provides few floral resources.
Long-term grasslands and crop rotation influence local richness and abundance
of bees at large spatial scales. Probably due to a masking effect of mass-flowering
crops, these influences are only detected in margins of non-oilseed rape fields.
Solitary bee abundance increased with the increasing proportion of long-term
grasslands (Fig. 14.4 b). Moreover solitary bee abundance and species richness
increased with the increasing proportion of ''mixed fields'' (at least one year of
grassland in the past five years) (Fig. 14.4 c), whereas the proportion of fields only
sown with cereals during the last five years had the opposite effect (Fig. 14.4 d). The
positive effect of long-term grassland is already known in landscape scale studies
on wild bees (e.g. Steffan-Dewenter et al. 2002 ; Morandin et al. 2007 ). This type of
fields is typically likely to provide wild flowers and suitable nesting sites. The
originality in our results is to show that introducing temporary grasslands in cereal
rotations is beneficial to bees. As they are generally sown with Poaceae species only
and fertilized, the suitability of temporary grasslands for bees remains to be sup-
ported by further data. Nevertheless, the introduction of this cover type in cereal
rotations could imply a less intensive farming system, potentially beneficial to
solitary bees, thanks to (1) reduced pesticides and fertilizer inputs over the whole
rotation cycle (2) greater floral resources in properly managed temporary grasslands
(3) less disturbed soils better suited for ground-nesting bees. Our result is consistent
with Steffan-Dewenter ( 2001 ) and Kuussaari et al. ( 2011 ) who showed the positive
effect of the introduction of set-asides in cereal rotations on pollinator insects.
To sum up our findings, the composition of the landscape at the time of sam-
pling had a direct impact on the spatial distribution of solitary bees only at the
finest scale (400 m). On the contrary, when considering the landscape structure
over several years (crop rotations and semi-natural elements like long-term
grasslands), the effects occurred at the larger scales (800 and 1200 m). Therefore,