endangered species, density of endangered
languages, energy budgets for countries,
roads ands connectivity, and many more.
Armed with this list, we searched for global
data to incorporate in the GIS. We quickly
narrowed the list down to those data
sources that were available and could be
used for the GIS. A common problem in
using different data sets is that the resolu-
tion, scale and map projections differ, mak-
ing it crucial to accurately standardise the
datasets within the GIS. Early on we decid-
ed to rely as little as possible on datasets
that were based on country boundaries in
favor of a more accurate spatial distribution
that did not take into account political
boundaries. This would enable us to
demonstrate the distribution of biocultural
diversity geographically rather than political-
ly.
The problem of mapping biological diversity
is a considerable one, and a discussion of
these difficulties is beyond the scope of this
paper. Suffice to say that many researchers
have explored this issue with varying
degrees of success. In our opinion, the
team led by Wilhelm Barthlott at the
University of Bonn has produced one of the
more sophisticated analyses of spatial distri-
bution of biological diversity, specifically
vascular plant diversity^2. Rather than rely on
distribution in political units (e.g.flora of
countries) or geographical characteristics
(e.g.flora of the Amazon) they calculate
vascular plant diversity based on standard-
ised units of area (10,000 sq. km). This
allows for comparable diversity categories
on a global scale. They use ten categories
of diversity based on number of vascular
plant species which they call diversity
zones. The ten diversity zones (DZ) with
species number per 10,000 s. km are DZ 1
<100 spp; DZ 2 100-200 spp.; DZ 3 200-
500 spp.; DZ 4 500-1000 spp.; DZ 5 1000-
1500 spp.; DZ 6 1500-2000 spp.; DZ 7
2000-3000 spp.; DZ 8 3000-4000 spp.; DZ 9
4000-5000 spp.; DZ 10 >5000 spp. We con-
tacted Barthlott and his associates and they
generously provided a base map of global
plant diversity based on the above figures
that we incorporated into the GIS.
The other crucial dataset required was the
global distribution of languages. For these
data we relied on the Ethnologue database
produced by the Summer
Institute of Linguistics^3.
Ethnologue is widely
regarded as the most
comprehensive data
source of current lan-
guages spoken world-
wide. Its major limitation
is that it does not indi-
cate spatial extent of the
speakers for a given language, only a singu-
lar point denoting the most central location
of the population. However, on a global
scale this is not really problematic, and by
using a singular point it allows for the
graphical depiction of all of the world’s lan-
guages on a single map.
Global distribution of biocultural
diversity
With these layers we were able to produce
an accurate depiction of the distribution of
linguistic and biological diversity worldwide
(please see Figure 1, placed in the back of
the back page of the journal, as the figure
needs color). For the sake of graphical sim-
plicity we aggregated the ten diversity
zones into five categories with Very Low
<200 spp.; Low 200-1000 spp.; Medium
1000-2000 spp.; High 2000-4000 spp.; Very
High >4000 spp. This map clearly demon-
strates the co-occurrence of high linguistic
diversity with high biological diversity.
Several regions in particular stand out in
this regard: Mesoamerica, Andes, West
Africa, Himalayas, and South Asia/Pacific.
The general trend towards increasing lin-
guistic diversity in areas of increasing plant
diversity (or vice versa) is noted in Table 1
and Figure 2, with the diversity zones disag-
gregated back to the original ten classes. A
regression analysis shows that r^2 = 0.9873,
noting a strong correlation between increas-
History, cculture aand cconservation
...one iimmediate ttrend
that wwe nnoticed wwas tthe
relationship bbetween
low ppopulation ddensity
and, iin ssome aareas,
high bbiocultural ddiver-
sity....
