Animal Movement - Mobility Data - page 284

animals in groups performed a Brownian motion, indicating that living in group
reduces foraging efficiency (Figure 13.5 b). This effect has indeed demonstrated
that foraging rates decrease with group size. The mechanisms seem linked to a
variation in the foraging efficiency of animals moving inside the group, which
forage less than deer moving on the borders of the group. Because with large
groups there is a larger proportion of animals inside the group, the average
foraging efficiency is lower in large groups. This is the price to pay to enjoy the
better protection afforded by large groups. Besides that both LW and CCRW
may simulate the process of area-restricted search, the two models are basically
different: Levy walks are scale-free, while the two-level scale-specific CCRWs
are a mixture of two movements characterized by specific scale (typically, inter-
and intrapatch movement).
13.3 Conclusions
The last years have seen impressive development in the study of animal move-
ment. Important breakthroughs originated from the technological development
of biologging devices, which have allowed researchers to collect huge amounts
of movement data from a large number of animals. All possible scales of analysis
(frommigration to food search) were investigated. In parallel with technological
development we have witnessed substantial improvements in data storage and
data mining and statistical models became more and more flexible. An applica-
tion of multiscale movement analyses to the understanding of animal ecology
was provided in relation to post-reintroduction displacement of elk ( Cervus ela-
phus ). This analysis documents behavioral shifts at different spatial and temporal
scales that permitted to these animals to survive in a difficult environment.
Despite these improvements, the analysis of animal movements remains chal-
lenging and requires important progress.
The first, apparently banal, observation is that biological samples are often
biased: to fit animals with a biologger, the animals have to be captured, thus
incurring behavioral biases, for example, hihger ability to escape or a higher
attitude to use baited traps than the average individual. Further, the device itself
maymodify the behavior of an animal because of its weight and shape. Moreover,
the sample size might be inadequate to express the variability of the population.
Last but not least, the devices determining animal positions are subject to errors
as are all instruments. Finally, a movement is a continuous process but we are
forced to sample it in a discrete manner. Careful experimental design is strongly
advertised.
As also noted in Chapters 1 and 2 , it is quite important to use semantically
rich trajectories. Of course the use of biologically meaningful attributes can
enhance our understanding of animal behavior and data enrichment is an impor-
tant future task to be developed with the use of innovative technology. In the