A modeler is the person, who dares to point out the places where certain type of seaweed grows even though the site had never been visited before. A modeler is the person who says where the field staff will be sent to find a rare species of beetle that lives in the sea. A modeler is also the one who takes the heat when exhausted divers return without ever encountering this mythical creature.
A map on my screen shows that we have mapped most of the broad areas in our coastal seas. A closer look; however, reveals that only a small portion of the seafloor has been actually visited. Mapping every inch of the seafloor on site is not possible, so we have to find alternative means to understand what is in the areas between the visited locations.
Identifying the type of surrounding environment based on few sample points can be difficult. The type of environment might be totally different in a location 100 m away from a diving transect or a video point. You could imagine doing mapping on land by filming your feet and trying to guess what kind of environment is surrounding you. Under your feet might be a road, but next to it could be a field, forest or someone’s yard.
The job of a modeler is to map areas, where certain species or species groups are most likely to occur. Modeling is done on the basis of known traits of the environment and what kind of locations the species in question prefers. If we try to model suitability for a delicate plant that will break in strong wave activity, a modeler will limit the possible areas to sheltered locations. If we know that the plant is a marine species that cannot handle low salinities, we can leave those areas out. We know that plants need sunlight, so we can select the areas of the seafloor that are well enough illuminated. Sounds easy right?
Even though the basic concept is simple, there are a couple of major obstacles. One is lacking data on the constantly changing environment. There are very few stable traits in the sea, so we must work with statistical means, minimums and maximums a lot. For e.g. in the Baltic Sea, salinity is dependent on rainfall, flow of the rivers and pulses of more saline water from the Atlantic Ocean. More saline water is heavier, so we might have layers of water with different salinities. A storm the next day might mix these layers and again the salinity changes.
Another major obstacle is our information about what kind of environments certain species or species groups prefer. We most often look at our field inventory data to see where we have found certain species before. The size of our boats can affect the depth distribution of our inventories, and hard to reach places, such as outer reefs or densely vegetated inlets, are less often visited. The number of our observations is affected by the places most visited according to our mapping efforts. The locations where we have found most often the species to be modeled might not be in surroundings they prefer to grow in.
We can use statistical calculations to correct the observations with the inventory effort, but there is also trouble with communicating the results. When people see the model and try to evaluate how good it is, they tend to visually or statistically inspect, how well the observations are aligning with the model. This will not tell the whole truth. A modeler has at this point considered, where the species has been found, but also suggests where it prefers to live and should be looked at more closely when working in the field in the future.
Written by Matti Sahla
Parks & Wildlife Finland