Geoscience Reference
In-Depth Information
alternate search heuristics (Ciarleglio et al., 2009). Second, one can include alternate weightings of
the taxa considered, for example, weighting by the inverse of a taxon's range such that endemic taxa
contribute more to the complementarity than widespread species (this is directly analogous to the
WE metric of Equation 6.2). One can also include phylogenetic weightings. The reason these have
not been explored is simply due to the fact that the majority of commonly used software packages do
not support extensions. Of course, this should not stop a good GC researcher developing their own
solution, and developing such solutions has been one of the drivers behind GC research.
6.3.4 d
iSeaSe
S
Pread
The spread of disease through populations, particularly those involving uncontrolled animal species
such as feral animals and unfenced livestock, is something that must often be understood from a
paucity of available data. It is in these circumstances that spatio-temporal simulation modelling, an
important research area of GC, is often the only recourse to assess the potential impact of a disease
or virus incursion and therefore the development of response plans and policy Ward et al. (2007).
Disease outbreaks are spatio-temporal phenomena. Artificial life models, such as geographic
automata, therefore represent an important modelling approach. Geographic automata models
are simply extensions of cellular automata (see also Batty and Longley, 2014) that work with non-
symmetric geographic units. Fundamentally they treat space and time as discrete units and allow
interactions to occur between local neighbours using a set of simple rules (Doran and Laffan, 2005;
Ward et al., 2007; Laffan et al., 2011). Models developed for epidemiology typically also use a state-
based approach, where an individual or group can be in one of four states at any one time step -
susceptible, latent (infected), infectious and recovered/removed. The complex behaviour of epidemics
and their evolution over time (Figure 6.5) is simulated by the repetitive application of the rules con-
trolling transmission of the disease and temporal transition between sequential individual states.
Such automata approaches are extremely flexible and have the potential to be applied to any sys-
tem where the interactions are predominantly local. Where interactions are essentially non-local,
such as long-distance human travel networks, they become less applicable and other epidemiologi-
cal approaches are needed.
Beyond simulation modelling, one can visualise the spatio-temporal spread of diseases as they
evolve. A good example of this is the tracing of outbreaks of disease as the virus or bacterium
mutates. For example, Perez et al. (2010) traced the development of an outbreak of vesicular stoma-
titis in the United States over a 2-year period using scan statistics and space-time visualisation of
the phylogenetic change in this disease as the outbreak spread.
6.3.5 M
oVeMent
a
nalySeS
One part of biology that represents a broad and interesting arena for GC research is animal move-
ment. Developments in GPS, radar and other tracking technologies are making it possible to col-
lect vast amounts of data about the movements of animals as small as insects (Riley et al., 1996),
with sensors mounted on orbital platforms such as the International Space Station being planned
(Pennisi, 2011). Indeed, as more technology is developed, not only will we get more data but the
data sets that are available for analysis will also undoubtedly become more complex. The analysis
of such data can be linked to the environments that the animals prefer to utilise (Zhang et al., 2011b)
or, for detailed data sets, to which other animals they are interacting with. Many applications use
simple kernel density analyses (Taylor et al., 2006), but it is clear that research developments in
space-time geography (Miller, 1991; Demšar and Virrantaus, 2010; Laube and Purves, 2011) and
agent-based modelling (Benenson, 2014) are directly relevant.
Of particular promise are radar analyses, although in terms of animal movements, the scientific
application of this technology has largely so far been restricted to tracking insects (Riley et al.,
1996). However, more recent developments in the use of PTT and RFID tagging have also enabled
