Agriculture Reference
In-Depth Information
allocated at random; control and input plots within a 'block' were located
40 m apart. Earthworm inoculations were carried out during the second
year following soil replacement and aimed to create a population typical of
undisturbed pasture over a short period of time. Inputs were equivalent
to almost 70 individuals m
−2
(for details, see Scullion, 1994) and were
representative of the full range of species in local pastures. Field drainage
and general management (grazing of clover-grass swards by sheep, with an
annual topdressing of 8 t fresh weight of poultry manure ha
−1
year
−1
) were
considered (Scullion, 1994) to favour earthworms.
In the fourth year after soil replacement, populations on input plots
were similar to those on adjacent, undisturbed pasture (Marashi, 1995)
and included a high proportion of the deep burrowing species
Lumbricus
terrestris
(L.) and
Aporrectodea longa
(Ude). Prior to the final earthworm
survey in year 6, populations on control plots were markedly lower than
those of input plots and were dominated by the surface-dwelling, early
colonizing species
Lumbricus rubellus
Hoff. and
Allolobophora chlorotica
Sav.
In the final study, differences in soil microbial organic C indices under
grass and woodland were assessed on the three land types described in the
first study. Woodland soils were sampled in coarse grass under
Alnus
glutinosa
, the most productive species in the woodland mix (Scullion,
1994). Trees were planted in the first year after restoration on former
opencast land and at the same time as the R21 site on undisturbed land.
During the first five or more years after planting, when tree growth was
slow, organic returns would have been markedly higher under grassland.
Carbon cycling was also favoured under grassland by the regular return of
faeces during grazing. These effects led to markedly larger earthworm
populations on grassland compared with woodland sites (Scullion, 1994).
Woodland areas adjacent to the high input grassland sites used in the first
study were selected for this comparison. These grassland and woodland
sites were sampled concurrently, with sampling restricted to the surface
10 cm where soil changes were expected to be more pronounced.
Between six and ten replicate samples were taken from each area in the
above investigations. Mean values of each measured parameter for different
soil groups were compared by one-way analysis of variance, using the
Statgraphics version 7 (Statistical Graphics Corporation, 1993) statistical
package.
Soil Analyses
Organic contents (LOI) were measured by loss on ignition (Ball, 1964) at
400
C. Total carbohydrates (CARB) were extracted using strong acid
hydrolysis followed by refluxing at lower acid concentrations (Cheshire
et al
., 1983) and measured by the anthrone method (Brink
et al
., 1960).
°
Search WWH ::
Custom Search