Animal Movement - Mobility Data - page 285

example of fallow deer, the possibility of observing deer behavior while record-
ing spatio-temporal positions allowed us to understand whether or not foraging
behavior was optimal and to evaluate trade-offs between foraging efficiency and
protection, a factor that dominates the life of these animals.
The analyses of animal movement can be done through two different
approaches: mechanistic and statistical. A mechanistic model needs data col-
lected at a spatial and temporal scale compatible with the (behavioral) scale at
which the organism takes its own decisions or interacts with its environment,
for example, selection of food items or habitat, predator avoidance, interaction
with conspecifics, and so forth. The chance of collecting such data depends
on the context (e.g., field or experimental settings, studied species) and on the
development of appropriate technologies. For instance, the recent release of
video camera collars will allow researchers to collect data on foraging with
an unprecedented resolution on large mammals. However, the step of analysis
allowed by actual technology is often coarser than the typical behavioral scale
and the use of mechanical statistical models is appropriate to perform an analysis
of movement patterns, which can give us relevant insights in the processes of
spatial distribution of animal population, such as dispersal and habitat exploita-
tion. On the other hand, the choice of appropriate models to analyze movement
data is dependent on the sampling design used by the researcher, which, in its
turn, depends on trade-offs among costs of tags, costs of capture, weight of
devices, and so forth.
We expect that important developments may arise by the application of meth-
ods from statistical mechanics to animal movement as suggested in the recent
topic by Viswanathan et al. (2011). There are many instances of transfer of
methodologies from statistical mechanics to ecology; since Okubo's topic, sev-
eral types of randomwalks have been first developed in physics and then applied
to organismal movement. Even results relative to diffusion are shared in the two
disciplines. This has not been a swift process, as methods and ideas that are
well established in statistical mechanics have demonstrated to be unsuitable or
problematic for applications in ecology. The study of animal movement is a chal-
lenging task: animal behavior is dictated by drives that have evolved over a long
time. On the other side, the natural environment, influencing animal movements,
is highly heterogeneous in time and space. Interdisciplinary research between
engineers, physicists, ecologists, and ethologists is more than a rhetoric plea: it
is the key to relevant breakthroughs in the future.
13.4 Bibliographic Notes
The literature on animal movement is immense. We have kept citations to a bare
minimum. An useful review of old literature can be found in Fraenkel and Gunn
( 1961 ). Okubo ( 1980 ) represents the passage of a number of models and concepts