2
A. M A C F A D Y E N
It is essential to appreciate, that an experimental design should be
conceived as a whole, that the most elegant statistical methods cannot
compensate for biased sampling nor for inappropriate application and
that the most complete extraction is invalidated by faulty experimental
design. Nevertheless it is convenient to discuss the subject under two
headings, the experimental design of the survey and the field and laboratory methods to be used, because the first has a wider application and so
can be treated mainly by reference to the literature, whilst the latter are
more confined in their range of interest to the present topic and therefore demand fuller treatment. In order to relate the methods discussed
to practical needs the following section (Section 11) suggests a number of
types of soil arthropod survey and the appropriate sampling requirements whilst in the final section (Section V) an attempt at synthesis is
made by considering some specific projects.
11. CHARACTERISTIC SAMPLING REQUIREMENTS OF SOME
DIFFERENT TYPES OF SOIL ARTHROPOD SURVEY
Since at least fifty percent of the world’s arthropod fauna remaim to
be described there is still a great need for geographically widespread
taxonomic collections to be made without regard to quantitative considerations, especially from regions outside Western Europe. However,
our knowledge of the taxonomy of European soil arthropods has improved so much in the last twenty-five years that serious study of ecological problems is now possible because the ecologist can see beyond
taxonomic difficulties which formerly occupied his full attention or deterred him from studying the soil fauna entirely. Thus while exploratory
surveys can and should be conducted with simple, robust equipment,
used on collections obtained in a haphazard manner, ecological work
now demands that greater attention be devoted to sampling methods.
Ecological problems are mainly of two kinds; firstly there are those
involving comparisons between species lists (sometimes of limited taxonomic range) which have been obtained from different habitats, parts of
habitats or at different times, including distributional, seasonal and successional studies. Secondly we have studies that are concerned with
trophic structure of communities including food chains, biomass
measurements, biological interrelationships and population-size
measurements within a single community. The former, which can be
called community studies, tend to make greater demands on taxonomic
precision than the latter or “trophic” type because a greater range of
related species is encountered and distinctions between them are often
critically important in detecting changes in community compositlon. In
a trophic study confined to a single habitat, fewer species are encountered and general conclusions will hardly be invalidated if an isolated
A. M A C F A D Y E N
It is essential to appreciate, that an experimental design should be
conceived as a whole, that the most elegant statistical methods cannot
compensate for biased sampling nor for inappropriate application and
that the most complete extraction is invalidated by faulty experimental
design. Nevertheless it is convenient to discuss the subject under two
headings, the experimental design of the survey and the field and laboratory methods to be used, because the first has a wider application and so
can be treated mainly by reference to the literature, whilst the latter are
more confined in their range of interest to the present topic and therefore demand fuller treatment. In order to relate the methods discussed
to practical needs the following section (Section 11) suggests a number of
types of soil arthropod survey and the appropriate sampling requirements whilst in the final section (Section V) an attempt at synthesis is
made by considering some specific projects.
11. CHARACTERISTIC SAMPLING REQUIREMENTS OF SOME
DIFFERENT TYPES OF SOIL ARTHROPOD SURVEY
Since at least fifty percent of the world’s arthropod fauna remaim to
be described there is still a great need for geographically widespread
taxonomic collections to be made without regard to quantitative considerations, especially from regions outside Western Europe. However,
our knowledge of the taxonomy of European soil arthropods has improved so much in the last twenty-five years that serious study of ecological problems is now possible because the ecologist can see beyond
taxonomic difficulties which formerly occupied his full attention or deterred him from studying the soil fauna entirely. Thus while exploratory
surveys can and should be conducted with simple, robust equipment,
used on collections obtained in a haphazard manner, ecological work
now demands that greater attention be devoted to sampling methods.
Ecological problems are mainly of two kinds; firstly there are those
involving comparisons between species lists (sometimes of limited taxonomic range) which have been obtained from different habitats, parts of
habitats or at different times, including distributional, seasonal and successional studies. Secondly we have studies that are concerned with
trophic structure of communities including food chains, biomass
measurements, biological interrelationships and population-size
measurements within a single community. The former, which can be
called community studies, tend to make greater demands on taxonomic
precision than the latter or “trophic” type because a greater range of
related species is encountered and distinctions between them are often
critically important in detecting changes in community compositlon. In
a trophic study confined to a single habitat, fewer species are encountered and general conclusions will hardly be invalidated if an isolated