Geoscience Reference
In-Depth Information
not yet available for most taxonomic groups, even those that are well studied. For example, the genus
Acacia
is one of the most speciose and well studied of the Australian vascular plant groups, with
1029 known species in the genus. Despite being a very well-studied group, the available phylogeny
covers less than 20% of known species (see Miller et al., 2011). The level of detail contained in a
phylogeny is also a function of what the taxa are and the extent to which they have been studied. For
example, where available, mammal species typically have phylogenies at the species level, insects
at the genus level and bacteria at no more detail than the family level or higher, except where they
are related to human diseases. Many organisms have yet to even be described by science. These
are typically smaller organisms such as insects, but such organisms represent the great majority of
species on the Earth.
Given the difficulties associated with the identification of bacterial and related organisms, one
alternative approach being explored is to develop phylogenies of the genetic diversity of what is
present at a location rather than attempting to identify individual species (Venter et al., 2004). Such
an approach avoids the sometimes contentious issue of defining species units and instead operates at
the level of the individual organism. Perhaps the example that relates most to readers is the assess-
ment of genetic variation among a sample of humans.
Genetics is also a field that moves at an extremely rapid rate, with advanced technologies from
10 years ago already being redundant. The current state-of-the-art approach is in next-generation
sequencing (e.g. Egan et al., 2012), a set of methods with the potential to generate extremely large
amounts of complex geospatial data in very short periods of time.
A further limitation of phylogenies is the availability of data over temporal periods relevant to
evolution. Genetic samples are comparatively easy to obtain for extant (living) organisms. However,
fossil and sub-fossil data are at the mercy of preservation, assuming they can be found in the first
place. This means that one needs to use morphological traits that are preserved in the fossil record
such as dentition (teeth) or leaf structures.
There also always looms the possible effect of phylogenetic revisions as new data and methods
become available. However, change in the underlying data is not an abnormal occurrence in GC
or indeed in any field of research. The main effect in the case of phylogenetic analyses is that any
such modifications to the structure of the phylogeny will propagate their way through a subsequent
spatial analysis operation, to deliver changes in the resultant spatial patterns that are directly pro-
portional to changes in the original phylogeny.
6.2.2 g
eneS
and
o
ther
M
olecular
d
ata
Phylogenies are generated primarily from genetic data. This is done using a variety of methods,
but fundamentally it is based on a comparison of shared DNA base pairs (the nucleotides A, C, G
and T). These base pairs are the rungs on the double helix ladder that is DNA, with the full set of
base pairs being the genome. The more pairs that are common to two organisms, the more shared
DNA they have.
From these, genetic data can be derived matrices of genetic relatedness. These matrices can be
used to derive phylogenies or analysed directly (Bickford et al., 2004) by linking them to geolocated
objects at whichever level is appropriate (e.g. individual, group, population, taxon).
These genetic matrices represent one of the challenges of spatially analysing biological data.
GIS software has not been developed with such data structures in mind. Small matrices can be
stored as attribute tables, but new data structures analogous to spatial weight matrices found in
specialist spatial analysis tools are needed to analyse them spatially. This then requires new and/
or more complex algorithms to access the data to use in an analysis of a location or neighbourhood
of locations.
While genetic data are comparatively common, other molecular data such as metabolites are also
the subject of research (Kulheim et al., 2011) and represent an additional form of biological data to
which GC methods could be applied.
