Community Ecology Processes, Models, and Applications

may for economic reasons specialize on a single or a
few crops. This is especially a risk as contracts with
large retailer chains such as Wal-Mart or Sainsbury
become more common.
The expected positive effects of organic farming
on biodiversity are one of the reasons it has been
included in agri-environmental schemes. An im-
portant applied question is then whether organic
farming really delivers a higher biodiversity. For
more basic questions, the advantage of using organ-
ic farming is that it is afarm-scalemanipulation of
landscape structure. Organic farming mainly af-
fects how fields are managed, and, for organisms
in natural habitat patches, it may mainly increase
the quality of the matrix (Vandermeer and Perfecto
2007). This is likely to have effects on metacommu-
nity dynamics and trophic interactions.
Our studies were conducted in a mosaic agricul-
tural landscape in central Sweden, where fields are
interspersed with areas of glacial till and bedrock
which are not used for crops but often contain semi-
natural pastures or small forests. We did not find
any positive effects of organic farming on diversity
among butterflies, carabid beetles, plants in field

margins or spiders, although landscape heteroge-

neity at the farm scale influenced these groups pos-

itively (Weibullet al. 2000, 2003, O ̈berget al. 2007).

In addition, local habitat type was more important

for diversity than landscape factors (Table 9.2).

The importance of local habitat and landscape for

species composition appeared to vary between the

three organism groups (plants, carabids and butter-

flies) in relation to mobility (Table 9.2). The propor-

tion of explained variation accounted for by

landscape measurements increased from almost

zero for plants to almost 50% in butterflies, which

in general are the most mobile of the three groups.

Other studies have also found that responses to the

landscape may vary with differences in mobility

(Steffan-Dewenteret al. 2002; Tscharntkeet al. 2005

for an overview). This is in accordance with meta-

community theory. The importance of the land-

scape, either other patches or the matrix, should

increase with organism mobility and dispersal

rates (second expectation, above; Fig. 9.2a).

Such patterns may also be found for the same

group in different landscapes, as suggested by a

study of plants in field buffer zones in Finland

Table 9.2Organism mobility in relation to contributions of landscape, farming system and

local habitat type to (A) species richness (diversity) and (B) variation in species composition on

16 farms in mosaic agricultural landscapes in east central Sweden

Mobility: PlantsÆ CarabidsÆ Butterflies

A. Diversity

Landscape (þ)r¼0.47 (þ)r¼0.58 (þ)r¼0.52

Farming system ns (þ)P¼0.02 ns

Habitat þþþ

B. Species composition

Landscape 0% 5% 10%

Habitat 10% 5% >10%

Modelr^2 15% 15% 20%

Mobility is assumed to increase from plants to carabids to butterflies, based on data on

adult movements. Habitat types were pasture, semi-permanent ley and field margin. The

table summarizes results in Weibullet al.(2003) and Weibull and O ̈stman (2003).

For diversity (A), only habitat was significant in the full model, but effects of landscape

(highest correlations with landscape variables shown) and farming system (carabid diver-

sity lower on organic farms) were found in other analyses (see Weibullet al.2003).þ,P<

0.05; ns, not significant.

In (B) the figures refer to the proportion of the variation in species composition explained

by canonical correspondence analysis (CCA) axes related to landscape or habitat (Weibull

and O ̈stman 2003) and interpreted by this author. Modelr^2 is the total variation explained

by the CCA model.

DIVERSITY AND ECOSYSTEM SERVICES 125