(Berg and Matzner 1997). The net effect is a relative
increase in recalcitrant substrates as decomposition
proceeds, while the amounts of un-decomposable
material accumulate with depth in the organic pro-
file. As nitrogen is usually the limiting nutrient in
the initial phase of litter decomposition, the C:N
ratio of organic matter is a general index of the
quality of litter (Gadisch and Giller 1997; Berg and
Laskowski 2006). However, to understand the me-
chanisms that regulate the processes of decomposi-
tion, the type of carbon in organic matter, the
concentration of other nutrients than nitrogen and,
especially, the composition of various secondary
plant compounds are also important. For instance,
litters with high lignin content, and that are rich in
polyphenols, have a low degradation rate. Lignin is
difficult to decompose by microbial enzymes (Shah
and Nerud 2002) and can mask cell wall polysac-
charides from degradation (Chesson 1997). Poly-
phenols, condensed tannins, terpenes and surface
waxes strongly reduce the activity of microbes and
the palatability of plant residues for detritivores
(Ha ̈ttenschwiler and Vitousek 2000).
During degradation, the morphology, biochemis-
try and physical properties of organic matter are
continuously modified. These physicochemical
transformations result in a decline in substrate
quality for microbes and detritivores and are ac-
companied by succession of soil biota (Ponge 1991;
Dilly and Irmler 1998). Over time, this older organic
matter is covered by newly formed detritus with a
higher quality, and this stratification in substrate
quality results in shifts in the vertical distribution
of soil biota (Faber 1991; Ponge 1991; Berget al.
1998a). In the absence of earthworms detritus is
not mixed through the soil and a stratified organic
horizon is formed, with a subsequent litter, frag-
mented litter and humus layer on top of parent
material. Although in the presence of earthworms
organic matter is mixed through the soil, there still
exists a vertical stratification in organic matter qual-
ity and quantity, but on a larger vertical spatial
scale.
These examples imply that the distribution of
soil organic matter and soil organisms is often
strongly interlinked. To evaluate the importance
of organic matter for community composition
and to assess whether variability in community
composition is linked to variability in soil organic
matter quality, I present the results from an exper-
iment that simultaneously describes variability in
organic matter degradation and community struc-
ture over time and across horizontal and vertical
space.Thisstudywasperformedinafirst-genera-
tion, 40-year-old Scots pine,Pinus sylvestris,forest,
in The Netherlands, planted on a heather field on
an inland dune of a former sand-drift area (Berg
et al. 2001). First, I shall describe the variability in
community composition over time, then introduce
some factors that can explain temporal variability
in community composition and finally indicate
possible consequences for food web studies.6.4 Variability in time in soil communities
Although it has been repeatedly pointed out that
temporal and spatial heterogeneity in soil is crucial
for understanding the distribution of soil organ-
isms and how biodiversity affects key ecosystem
processes (Bengtsson 1994; Ettema and Wardle
2002; Wardle 2002), surprisingly little is known
about temporal and spatial variability in soil com-
munities and detrital food webs (Bengtsson and
Berg 2005). Variability in community composition
has most often been studied in aquatic ecosystems.
These studies found a considerable temporal and
spatial variability in the composition of the food
web of a freshwater pond (Warren 1989), an inter-
mittent stream (Closs and Lake 1994) and a tidal
freshwater river (Findlayet al. 1996). Schoenly and
Cohen (1991) analysed a set of aquatic and terres-
trial food webs for temporal variability in commu-
nity composition. Very few species in the collection
of food webs occurred on each sampling occasion,
and most of the species were found only once. To
examine the temporal and spatial variability in the
structure of a soil community I performed a strati-
fied litterbag experiment (for details of the design,
see Fig. 6.2). Compositional variability, expressed
using the Bray–Curtis (BC) similarity index, was
measured over 2.5 years, for three organic horizons,
namely litter, fragmented litter and humus (for de-
tails, see Fig. 6.3). From litter to humus the quality
of organic matter for soil biota strongly declines. As
far as I know, this is the first time that temporal and72 SPACE AND TIME