Interpretation of plant distributions an OverView
A principle focus of Kew’s research in Madagascar is to study key plant
families in depth, to provide a substantial body of sound systematic information
on which to develop advice for biodiversity conservation. The Madagascar
programme started with three major groups which have a complementary distribution
on the island; the palms,
legumes and orchids.
It is now being extended to the Rubiaceae
and Euphorbiaceae. From these studies large databases have now been established
with records of plant distributions taken from herbaria in Antananarivo,
Paris and the Missouri
Botanic Garden, as well as from our own collections and records at
Kew. For the Papilionoideae alone we now have over 13,000 specimen records.
As these records are from different eras and are of varying accuracy, a
suite of programs have been developed to interpret this data in a way that
has never been possible before and to an extent that Kew may now be considered
to be a leader in this field.
Chart of the World’ is used as a standard base map giving global coverage
to which other more detailed regional maps can be attached, and also to
provide the base for species distribution maps (Figure
3). These maps can be of any area, at any scale and with a wide range
of geographical information selectively superimposed as required (e.g.
lakes, rivers, contours, roads and towns). User-friendly programs have
been developed to allow these species point distribution maps to be overlaid
on and compared to the maps of geology, altitude and vegetation type (Figure
2). The accuracy of herbarium specimen records can be weighted to allow
for the spectrum of difference between a modern GPS recording and vague
record from the last century based on a large town or a generalised physical
feature. Histograms can then be produced to show the range of substrates,
vegetation types or altitudes which the species prefers, and the statistical
relevance of these observed preferences can be examined (below).
The demonstrated preferences of individual species have been outstanding
and tally extremely well with field observations that have been made on
selected species. There is clearly a great deal more objective data that
can be obtained from dot maps than has ever been envisaged before, and
these techniques provide information which is directly relevant to floristic
work and taxonomic revisions. GIS analyses are ideally suited to the exploitation
of the geographical nature of the label data on herbarium specimens.
Analysis of the distribution of dicraeopetalum mahafaliense
in relation to geological substrate, showing a very strong preference for
Tertiary limestones (click on image for an enlargement).
Other applications of GIS include the use of Gap Analysis and related
techniques to allow an objective prediction of the full distribution of
a species from incomplete point distribution maps, based on its ecological
preferences. It can also be applied to the assignation of the new IUCN
red list criteria for rare and endangered species. The measures for extent
of occurrence (EOO) and area of occupancy (AOO) of a species have now been
automated within ArcView, producing
accurate measures, in km2, of the area within which a given
species occurs (see below).
This data has already been passed to WCMC
for woody species of Papilionoid Legumes.
Analysis of point localities within ArcView for Cordyla madgascariensis
In the longer term we hope to explore the use weather satellite imagery
to provide a measure of seasonality, and other climatic factors which are
major factors in determining the distributions of plants and very obviously
so in Madagascar, where dramatic changes occur over short distances. It
is difficult to get images totally free of cloud cover, but strategies
are being devised to sample at the same time over several years and make
limited extrapolations to provide a reliable guide. A map of bioclimates
has been digitised by the Missouri
Botanical Garden and data from the two sources will be compared.
Extent of occurrence (EOO)
Further analysis using Rapoport's mean propinquity method
It is envisaged that the combination of physical, geological and meteorological
maps will provide a very powerful tool for the further analysis of the
data from Madagascar. This could be extended to other regions of the world,
eventually providing a standardised suite of techniques and maps which
will allow comparative analyses to be made between any regions, and a generic
baseline for species distribution and conservation applications.
Back to top