Free Download - HCCREMS

CHaPTeR 3 Method of data analysis and collection
Teams located sites as closely as possible to the supplied coordinates using hand-held GPS units.
The actual coordinates and the GPS accuracy were recorded on the survey proforma. Most GPS
units currently are able locate a point to an accuracy of 10 m or better, but this is only under ideal
circumstances. Where satellite reception is limited, such as under dense canopy, actual margins
of error can be closer to 50 m. This level of accuracy may have implications for distribution
modelling, where many of the input layers may have resolutions of 25 m or better.
All sites were marked, where appropriate, with a marker at either the centre point or at one
of the corners of the survey quadrat. This potentially will allow sites of particular interest to
be used for longer term research purposes. In previous studies sites have often been marked
using star pickets, however, this option is cumbersome, particularly in steep country, and may
present a safety hazard in some circumstances. For these reasons sites were marked using a
type of engineering survey marker which is a metal peg hammered into the ground and to
which a durable plastic makers is attached to the top and which sits flush with the ground.
Field botanists entered the majority of data into an empty copy of the YETI database; this was
then returned, checked and imported into a central version of the database (see below section
3.3.4). Some data collected could not be entered into this database, either due to the way the
information was collected or simply the lack of a relevant field of information in the YETI
database. For entry and storage of such data a separate database was developed and empty
copies also distributed to field botanists. This database can be linked easily to the YETI database
for analysis of both datasets.
3.3.4 Quality assurance checks of the surveys
All data collected during the project were given quality assurance checks for both errors of
omission and commission. This included checks on whether data had been collected and entered
in the correct procedure and that all sections had been filled out (the quality assurance proforma
is included in Appendix 2). In addition, a random check of 10% of all returned data against the
original datasheets was undertaken to assess the accuracy of the data entry process.
3.4 Vegetation community analysis
3.4.1 Introduction
The main purpose of classification is to simplify complex sets of information by reducing a
large number of objects under consideration to a smaller set of groupings based on similarities
between objects. For vegetation community classification this means reducing the complexity of
a large number of floristic survey sites by grouping them together into ‘communities’ of species
which tend to occur together in the landscape. This in turn allows decisions to be made, or
analyses undertaken, on the basis of this reduced number of communities without having to take
into consideration the full diversity of the individual sites which make up those communities.
Such simplification is necessary to understand and describe complex environmental processes
and relationships. It should be noted, however, that, as stated by Beadle and Costin (1952) ‘any
attempt to classify a continuously varying system into several categories must necessarily be
somewhat arbitrary’.
In deciding on an approach to vegetation community classification consideration must be
given to the nature of the available dataset. Classification is often (particularly in the case of
29