CE 394 Fall 2002

Introduction:

How predators search for, locate, stalk, and subdue prey has long been of interest to researchers. This is especially true when referring to mammalian species. Much of the research to date has focused on terrestrial species because of the increased ease in observation as compared to underwater species. There has therefore been little research conducted on the foraging habits of marine mammals. Previous research techniques for observing these highly mobile marine creatures (i.e. fixed location cameras, manned submersibles, divers, etc.) have provided only short glimpses into their diving behaviors. Recent advances in technology are now enabling researchers to equip animals with not only data recorders but video systems as well (1). The combination of video images and recorded data allows researchers to study marine mammals three-dimensionally. This is of obvious importance since marine mammals locate, pursue, and capture prey in three spatial dimensions. Fig 1 (left) Seal with attached video system and data recorder.

Fig 2 (right) Weddell Seal pup.

Searching for and capturing prey can often be a challenging activity (3). This is even more so the case when referring to marine mammals. Most marine mammals have developed a wide variety of methods for dealing with these challenges, and the Weddell Seal (Leptonychotes weddellii) is a perfect example of this. These large, marine predators are highly adapted for hunting in the shore fast and pack ice habitats of Antarctica (1). In order to survive, these seals must hunt through dark, cold waters while also holding their breath. The ability of these creatures to survive in such difficult hunting conditions has raised a number of questions. One particular group of researchers from several universities is in the process of examining the specific behavioral and energetic adaptations used by the Weddell seal during foraging.

Fig 3 (left) Video recorder and data logger

Fig 4 (right) Internal components video/data system

During the past five years, these researchers have made a number of trips to Antarctica to visit the Weddell seals. Using the new animal borne video recorders and data loggers, they were able to gather a great deal of data regarding the seals foraging behaviors. The video system was responsible for recording images of the seal’s head and the environment immediately in front of the animal. The data logger was responsible for recording time, depth, water speed, compass bearing, water temperature, and stomach temperature once every second (1). Because of the three-dimensional nature of the dive paths, a great deal of the data analysis has centered on the reconstruction of these dive paths in some three-dimensional manner. The dive paths for some of the recorded dives have already been constructed but a great deal of data still remains to be examined. The dive paths developed have already been quite helpful in understanding the predatory behaviors of Weddell seals, and further analysis of the data will only improve our understanding of the their behavior.

Fig 6. Weddell World

Fig 5. Location of research site known as Weddell World. Fig 7. Aerial photo of research site

Objective:

One of the many outstanding features of ArcGIS is its ability to display data three-dimensionally in ArcScene. ArcScene is often used to display land surface terrain using elevation values, but it has other three-dimensional applications as well. The goal of my project was to use this component of GIS to reconstruct several three-dimensional dive paths of one Weddell Seal (Seal 16), and apply this is to the analysis of seal behaviors while foraging.

Methods:

The data for this project was supplied directly from researchers involved in the experiment. Since each seal is deployed a number of different times, and within each deployment they can dive multiple times, tremendous amounts of data are collected for each seal. For this project I only used data from one seal (Seal 16). I focused primarily on the fifth deployment since this was when the seal appeared to have the most interaction with prey, therefore providing me with the most insight into the seal’s foraging behaviors. All the information obtained from the data logger is placed in Excel spreadsheets. This raw data can then be used to calculate the x, y, and z values of the seal for every second along its dive, enabling us determine the exact location of the seal the entire time it is underwater.

Fig 8. Excel spreadsheet used to create .dbf file.

The first step in generating a 3-D dive path was to create .dbf files from the excel spreadsheets I had been given. I chose to not only include the necessary x, y, and z values, but the speed values as well. This would allow me to easily display how quickly the seal was moving along its dive. I then loaded the .dbf file into ArcMap. After the file had been added I was able to display the x, y data. This gave me a nice two-dimensional picture of the seal’s dive path. Once a 2-D dive path had been created in ArcMap I was able to transfer the data to ArcScene. The first step within ArcScene was to select the layer and use the 3-D Analyst button. Within the 3-D Analyst extension, there is an option called Convert Features to 3-D. When this is selected, you are then allowed to choose exactly which layer you would like your z values to be added from, as well as where you would like them to be displayed. The final step is to choose the actual source of the z values. In my case, they were listed as attributes in the attribute table, so I chose the option “Input feature attribute.” After hitting OK, I then had the three dimensional dive path of a Weddell Seal.

Results:

I was very excited with my results, but I quickly noticed a problem. The seals appeared to be diving up into the air, rather than down into the water. Being the astute biologist that I was, I did not believe there was any way this could have actually happened. After some small experimentation with the data (and a little assistance from Venkatesh), I realized I had to convert my z attributes to negative values in order for the seal to dive in the proper direction. I now had a true 3-D dive path of a Weddell seal.

Fig 9. Oops! The seal appears to be diving into the air. Fig 10. Dive s16d05b. No prey interaction. Fig 11. Dive s16d05a. No prey interaction.

While both dive paths with and without prey interactions offer interesting insights into the life of a seal, it is the dives with prey interactions that are of most importance to my project. After watching the video images from each of the dives, researchers were able to establish the exact point at which the seal had an interaction with a prey species. They then noted these points within the excel spreadsheets. With these points I was able to create layers of prey encounters using the same methods I had with the actual dive paths. By adding these layers, I was able to see at which exact point the seal made Fig 12. P. borchgrevinki are one

contact with the prey. When put in the context of the entire dive, this also allows source of food for the Weddell seals.

researchers to know exactly what the seal was doing before and after its encounter with

the prey, as well as how fast it was traveling. Fig 13. Images recorded from the animal-borne video camera. The seal is flushing P. borchgrevinki from the loose ice below the surface (1).

Fig 14. Dive s16d05f with prey interactions. Click to see animation. Fig 15. Dive s16d05k with prey interactions. Click to see animation.

Fig 16. Dive s16d05n with prey interactions. Click to see animation. Fig 17. Enlarged portion of dive s16d05n. Click to see animation.

Conclusions:

After examining the dive paths, there are several important things to notice. As the seals are descending through the water, they appear to maintain a fairly slow and steady speed (slow speeds are represented by the green colors). This is representative of a technique referred to as gliding. By using changes in their buoyancy the seals are able to descend without consuming a great deal of their available energy. This is particularly important in dives of greater depth since they are more energy costly (1). There are also some important observations to be made about the seals foraging behaviors. As Weddell seals search for their prey, they appear to use a technique known as straight walk (Fig 10. is a good example). This means that as the seal descends it will travel along a straight path in one direction and then ascend along a roughly parallel path. Each dive will normally extend in a different direction, allowing the seal to achieve the best coverage of the territory (2). If the seal is fortunate enough to spot prey, it will then conduct what is known as area-concentrated search (Fig 14-16). It is easy to see from the dive paths in which there was prey interaction that the seal performs a much different searching technique. This is not surprising since seals do the majority of their feeding on fish that frequently display schooling behavior (2). The last important observation to examine is how the seals attack their prey. Frequently, the seals will approach their prey from below (Fig 17). This is believed to be due to the combination of the seal’s reliance on vision for hunting and the dark diving conditions in which the seals must hunt. By approaching the fish from a greater depth, the seals can benefit from the backlighting provided by the ice. Lighting is not as much of an issue at shallower depths, but this technique can be rather important as the seals descend (1).

ArcGIS had never previously been used by the researchers to reconstruct the dive paths of Weddell seals. Before beginning this project, I had no idea as to whether or not GIS would be an effective way to create 3-D dive paths. Perhaps the most important conclusion I was able to draw from this whole experience was that GIS is in fact very effective at building dive paths in three-dimensions.

Future Endeavors:

After having such successful results, I hope to see the researchers on this project attempting to use GIS more in the interpretation of their data. I believe that if the 3-D dive paths are used in combination with the video images, we will be better able to tell exactly what the seal is doing along each dive path. Venkatesh Merwade is currently working on creating a 3-D time series animation using speed for each dive. This will truly allow us to watch the seal as it goes throughout its dive path.

Within the next few months the same research techniques (i.e. animal-borne video recorder and data logger) are going to be used on Stellar Sea Lions in Alaska. Hopefully this technique will make it easier to analyze the foraging behaviors of these similar creatures.

Works Cited:

1. Davis, R.W., L.A. Fuiman, T.M. Williams, “Hunting Behavior of a Marine Mammal Beneath the Anarctic Fast Ice,” Science 283 (1999).

2. Davis, R.W. Project Proposal, “Project Title: Foraging Ecology and Hunting Behavior of Adult and Juvenile Stellar Sea Lions,” (2001),

3. Williams, T.M., R.W. Davis, L.A. Fuiman, “Sink or Swim: Strategies for Cost-Efficient Diving by Marine Mammals,” Science 288 (2000).