Journal Biomagnification of Mercury

At the beginning of the third week we were ready to add to our simple model.  We decided to work separately so that we could each get practice in using the modeling tools in STELLA.

Kathy worked on biomagnification from the aqueous system through the large fish.
Leslie worked on adding pH and temperature as variables in the aquatic environment and creating stocks of natural and anthropogenic sources of mercury.  The work had become highly mathematical for both of us at this point.

Here's what we individually did and learned:
 
 

Kathy see Model 2

Leslie see Model 3

 

I set out to make our model "run," that is to adjust the initial numbers and relationships so that our STELLA model would approximate the way in which mercury biomagnification actually occurs. I reviewed our research and chose a few crucial numbers: (1) 10 nanograms per liter (ng/L) is the average concentration of mercury in ecosystems for which atmoshperic deposition is the dominant source, i.e. no spills of mercury have occurred; (2) concentrations of  methyl- mercury [CH3Hg]  in plankton increase about 10,000 fold over water concentrations; and (3) at each successive trophic level, mercury accumulates at a factor of 10x  Using these numbers, I adjusted initial values for stocks, flows and converters, trying to avoid the premature depletion of  populations and to maintain a continual mercury flow in order to see the biomagnification effect.  Because worldwide consumption advisories for mercury-contaminated fish are expressed in parts per million (ppm), I converted all values to these units. I decided to treat water, plankton and fish as similar "containers" for CH3Hg. To streamline the model somewhat, I also chose to ignore other factors (metabolic losses, plankton or fish deaths other than by "consumption," evasion or other decreases of mercury in the water, rates of preferential retention, etc.)  Finally, my section of the model did, in fact, run, producing the type of graphs I expected, based on published research, and generating the range of numerical values that fit our mental model of biomagnification.

I had read that maximum methylation of mercury in the sediments by bacteria occurs at slightly acidic conditions.  In addition, at very acidic or very basic conditions it is likely that no methylation will occur, since the organism which provides this service cannot live in severe pH environments.  To use pH as a variable in STELLA, I used a tool called a converter. In this instance the converter was used to designate a pH range within which the model would run and to define optimal methlyation parameters. This was pretty simple work once I became familiar with the software.  I repeated this logic with the temperature converter.  It makes sense that there would be no methylation at certain temperature ranges, and that there would be an optimal range.  In adding anthropogenic and natural sources I used the ratios that I found in our data to adjust the stocks and flows.  Most of this work was "seat of the pants," because we couldn't dig up enough data quickly enough for it to be useful.  At any rate, we had decided that we were using this process more to understand the system and the modelling process than to solve a problem which would require very accurate data.  As a teacher I want my students to learn that science is driven by questions. An important question to ask is how well a given model reflects reality.  I learned a lot about the details of the system and I had to ask a lot of questions in order to understand the details.  We are leaving our work with questions and a depth of knowledge about this system and systems in general.