In our latest post, Korbinian Pacher takes us out to hunt as he share insight behind the paper: “Evidence for a by-product mutualism in a group hunter depends on prey movement state”. Korbinian explains how striped marlin groups hunt schools of sardine and opens our eyes to the endless possibilities of studying behavioural ecology in the open ocean, using aerial vehicles no less!
About the paper
When we hear the term “group-hunting”, we almost immediately think of large terrestrial carnivores such as wolves or lions, working together to first subdue their prey and afterwards defend the kill from scavengers. But if we take a step back from these traditional group hunting systems, we suddenly see that the phenomenon of group hunting (hunting together with conspecifics) is also prevalent in marine systems, especially the open ocean. What makes these systems so interesting is that the dynamics of open ocean group hunting are fundamentally different to terrestrial! Prey organisms are usually much smaller than their predators and have to rely on grouping as their most effective defence. On the other hand, predators are often present in opportunistic non-kin and highly dynamic groups. This means that individuals cannot monopolize the resource and have to attack the prey group repeatedly to maximise their success.
Most interestingly, while these group hunts are highly dynamic and competitive, inter-predator aggression is basically absent! Through recent advances in technology (especially in the field of unmanned aerial vehicles), these intriguing systems have become more and more accessible to scientific observation and that’s exactly what we did in this paper: We filmed group hunting striped marlin; large teleosts from a family widely known as “billfish”, preying upon schools of sardines in the open ocean from underwater and with drones. With this, we explored how the movement dynamics of the prey schools influenced the performance of these predators. With our results we hope to shed light on the mechanisms and functions of group hunting in the open ocean and to understand the factors that have led to the evolution of this type of predator-prey interactions in the open ocean. We compared two fundamentally different prey motion states. The first one we called “stationary”. Stationary is characterized by predominantly circular and localized movement of the prey school that is closely surrounded by the predator group. The second one was called “the mobile state” which characterized the prey moving in rather linear ways while being chased by the predators (often over the extent of several kilometres!). In the mobile state lies the big novelty of our study. Only by use of drone were we able to capture a very interesting dynamic: The capture frequency of marlin that attacked the prey school was roughly the same between stationary and mobile schools, however when the prey schools were mobile many prey fish got isolated.
An isolated prey fish in this system loses every protection by the group and gets captured almost immediately by one of the predators. Yet most interestingly, the predators that captured these isolates were mostly non-active participants of the predator group and almost never the individual that is attacking the prey school. This meant that in the act of attacking the prey group, the disruption of the prey school created isolated prey items for opportunistic feeding. This represents a classic example for a by-product mutualism. These isolation events are highly prevalent only during mobile high-speed chases and drive the group level prey intake during this stage of the hunt. We believe this might partly explain the large predator-groups during these predation events that bother to follow the prey. Although only one individual can get access at a time, they may just stay around and wait for isolated prey. We detail these interesting findings and more in our paper, but that’s what I would choose for the take home message of our work!
About the research
The field work for this study was carried out on the open pacific in Baja California, Mexico, which is the only place to have striped marlin be reliably observed during hunting. Living on a small boat in the middle of the pacific for several days and spending every minute of daylight hunting for predation events and flying drones is extremely exciting but can also be exhausting, especially on quiet days where you cruise the ocean for hours without any success.
Back in the lab we analysed the data, which of course meant watching and sorting hours of video footage. For the analysis of predator performance, I chose a classical approach of timestamping certain behaviours, like attacks, captures and isolated capture, which was quite straight forward. However, striped marlin are among the fastest fish in the ocean so that meant having to get the complete picture from the drone videos be screened in 0.25x speed several times per video. This can be a tedious task at times!
A whole different story, however, was the development of the computer vision methodology to track the prey schools. As everything is filmed in the open ocean there are no reference points in the drone-recording only predators and prey in the deep blue of the pacific. In addition, both, the animals and the drone are constantly moving. The challenge here was to find an approach to track a constantly moving object (the prey school) in the open ocean filmed by another constantly moving object (the drone). As behavioural ecologists this was out of our league, but we are very lucky that Berlin offers a fruitful research environment with great and diverse collaborators in the same city. So, we teamed up with computer vision specialists from the Excellence Cluster Science of Intelligence. They tackled this problem and developed a particle filter algorithm that not only tracks the prey school in the video frame but integrates data from the built in flight log of the drone, to get prey trajectories in real world coordinates.
The next step in the project is relatively straight forward: as the trajectories for the prey are now accessible, the predators would be next! Having full trajectories of predators and prey in this system will open many interesting questions for scientists such as, how prey schools change from stationary to mobile and whether or not this dramatic shift in the motion dynamic is driven by the attack behaviour, spacing of the predators or if they are purely reacting to the collective movement decisions of the prey.
About the author
I got involved in classical ecology during my Bachelors, with a thesis on soil food webs, however my focus has shifted since then and nowadays some would purely call me a behavioural ecologist. The work on the striped marlin was what got me excited about this field of research. I started to work on the topic as a student assistant with Jens Krause as my supervisor and Matthew Hansen, who is the PI of the striped marlin project. Our paper detailing this work came out of the main findings from my master´s thesis. The striped marlin are an incredibly picturesque and fascinating study system, however, what really captivated me while I was working on the system, was that I learned what it means to work in a team of researchers: the everyday reality of thinking about biological questions, solving problems and discussing ideas with colleagues and collaborators to develop new questions and repeat the whole process again and again, was what really got me involved into behavioural ecology in particular and also research and academia on a broader sense. Now I am a PhD student in the same lab but switched my focus to a different study system.
In my PhD, I study the predator prey interactions between extremophile live-bearing fish and various piscivorous birds in southern Mexico. And while this means I had to move away from the open ocean, the systems have a lot in common. I think what generally fascinates me are how prey can use collective behaviour as an effective defence against predation and how predators on the other hand can adjust their behavioural strategies to overcome these defences. A scientific obsession I don´t want to leave unmentioned here is the functional morphology of the billfish rostrum (the name giving bill). My favourite fact about this structure is that it is covered in hundreds of thousands small micro-teeth, that have all diagnostic features of typical vertebrate teeth but are not even a millimetre long. I think everyone who wants to understand this enigmatic fish family should take a close look at the rostrum at least once!
Like the blog post? Read the research here.
