Around the world there are concerns over the impacts of land use change and the developments (such as wind farms). These concerns have led to the implementation of tracking studies to better understand movement patterns of animals. Such studies have provided a wealth of high-resolution data and opportunities to explore sophisticated statistical methods for analysis of animal behaviour.
We use accelerometer data from aerial (Verreaux’s eagle in South Africa) and marine (blacktip reef shark in Hawai’i) systems to demonstrate the use of hidden Markov models (HMMs) in providing quantitative measures of behaviour. HMMs work really well for analysing animal accelerometer data because they account for serial autocorrelation in data. They allow for inferences to be made about relative activity and behaviour when animals cannot be directly observed too, which is very important.
In addition to this, HMMs provide data-driven estimates of the underlying distributions of the acceleration metrics – and the probability of switching between states – possibly as a function of covariates. The framework that we provide in ‘Analysis of animal accelerometer data using hidden Markov models‘ can be applied to a wide range of activity data. It opens up exciting opportunities for understanding drivers of individual animal behaviour.
The following images provide an inside view into the ecosystems in which the Verreaux’s eagle and blacktip reef shark reside.
Soaring with Veraux’s Eagles
Megan Murgatroyd equipping an eagle with a UvA-BiTS tag used to collect the data presented in our paper. These tags collect high-resolution GPS locations, altitude and three-axis acceleration data. They are able to collect large volumes of data because information is downloaded to a ground station rather than being transmitted via satellite. (© Gareth Tate)
Verreaux’s eagles intersperse their flights with periods of perching, which can be detected by low values of acceleration. Eagles like to perch on rocky outcrops and are usually found together. Both perching and flying behaviours persist for periods of time, e.g. birds can perch for 10min between flights of 45min, and this is explicitly taken into account by virtue of the structure of HMMs. (© Megan Murgatroyd)
Large stick nests are built on cliffs or rocky outcrops and are used for successive years. Especially during nest building and incubation, birds can spend quite a bit of time walking around on the nest, this is the type of behaviour that is likely captured by the low activity state described in the paper. (© Megan Murgatroyd)
A Verreaux’s eagle seen here soaring in its typical habitat. Updrafts on cliff faces likely aid soaring flight. Away from cliffs eagles rely on thermal lift to achieve low cost flights. We found that birds were more likely to remain in an active state at higher wind speeds, which are known to aid soaring flight especially over ridges. (© Megan Murgatroyd)
Verreaux’s eagles lay one to two eggs but only raise one chick. These are normally hatch during mid-winter and stay on the nest for around 90 days. The parents provision the chick during this time, delivering prey to the nest. Rock hyrax is a frequent item on the menu, as seen here. Having concurred data from accelerometers and nest cameras for the same bird could allow for activity level to be linked to successful prey acquisition or prey type. (Photo taken by a nest camera installed as part of the Black Eagle Project, Megan Murgatroyd)
Swimming with Blacktip Reef Sharks
How many sharks can we spot in this photo? The waves breaking in the background while the sharks cruise around the reef. Due to bans on fishing, large numbers of predators (sharks, snapper, and trevally) are found at the atoll, which makes it a perfect site to study blacktip reef sharks. (© Yannis P Papastamatiou)
A blacktip reef shark swims around ourresearch boat at Palmyra atoll. In total, we received accelerometer data from four blacktip reef sharks. We calculated overall dynamic body acceleration and used this metric as a proxy for general activity levels. The hidden Markov model was able to distinguish between high level and low level activity, allowing us to understand when the shark was more active. (© Yannis P Papastamatiou)
A blacktip reef shark cruising solo at the border between the backreef and forereef, where we see the waves breaking in the background. A few sharks had cameras attached, as well as the accelerometers. However, even with cameras it’s hard to know everything the shark is doing! Acceleration data provides a picture of what the shark is doing without having to spend all day underwater swimming after it. (© Yannis P Papastamatiou)
A view from the top of a blacktip reef shark. Applying a hidden Markov model to the observed overall dynamic body acceleration data, we discovered that the sharks use deeper water during the day when they were least active. (© Yuuki Y Watanabe)
Here comes the blacktip reef shark! Photographed during the day, when these sharks tend to be less active. They ‘rest’ during these hours in low energy environments, though, naturally, they never stop swimming! (© Yannis P Papastamatiou)
To find out more, read our Methods in Ecology and Evolution article ‘Analysis of animal accelerometer data using hidden Markov models’.
Researchers are increasingly interested in how social behaviour influences a range of biological processes. Social data have the interesting mathematical property that the number of potential connections among individuals is typically much larger than the number of individuals (because individuals can interact with every other member of their group). This introduces a huge challenge when it comes to collecting data on social interactions—not only does the amount of data needed increase exponentially with group size, the data can also be more difficult to record.
