The hypothesis was correct: much of the E. coli placed into the artificial wetland models would be absorbed through the plants' roots. Both containers showed substantial decreases in E. coli content within the first 24 hours. Of the containers with plants, the rate of decrease in E. coli was greater in Model A, which contained the Giant Bur-reed plants.

However, Model C, the control, which did not contain any plants, decreased at the greatest rate. The E. coli may have attached itself to the soil, since there were no plant roots. Unfortunately, it is quite difficult to test the soil for this because of the different properties bacteriophages adopt when growing on a solid surface. Another sample was taken to test this theory, but the results received were disputable, so they were not included in the Observation and Analysis section of the Scientific Method. Instead, they were listed under the heading Extension after the Scientific Method.

The results of this final sample did support the theory that the bacteria had attached itself to the soil. The petri dishes in which a sample from Model C had been plated showed a much larger amount of colonies than the previous sample, while the amount in Container B remained the same. The amount of bacteria in Container A decreased to none, since no colonies appeared in the plates overnight, making it the most successful samples. Assuming that the bacteria remained in the container, there are several possibilities. First, that the bacteria did stick to the soil, and either multiplied until it was forced to spread back into the water, or simply detached itself from the soil. Second, that the bacteria remained on the soil, and the colonies that appeared on the plates were another type of bacteria that had grown in the water.

If the first theory is correct, then the E. coli in the other two containers would also have attached itself to the surface of the soil. However, the final test results did not show any increase at all in the bacteria content of Sample A, and Sample B only showed a small increase. Most likely, this was related to the amount of surface area in the containers of Bur-reed and Arrowhead that was covered by plant roots. The amount of exposed roots appears to be related to the amount of E. coli found in each container in the final sample. The Giant Bur-reed roots in Container A covered almost the entire surface of the soil. No colonies appeared on this plated sample. In Container B, the Arrowhead roots covered about half of the soil surface. An average of 8.5 colonies were visible in its plates. Container C did not contain any plants. When plated, 16 colonies appeared in the agar.

The second theory (the colonies that appeared on the plates were another type of bacteria) is also possible, however there are no results to prove that it occurred. To conclude, it is quite possible that the E. coli attached itself to the soil and multiplied, although with the given data, this can only be speculated. Therefore, the amount of E. coli in the soil of each of the containers is difficult to determine.

The results can be applied in many situations involving water pollution. The rate and amount of contaminant absorbed by the plants depends on the number of roots it is exposed to. If, for example, the pollutant is being removed from a river to prevent its transportation to a larger water body, a wetland would be constructed between the river and larger water body. It would have to be much wider and more shallow than the river itself. It would also have to be exposed to the maximum amount of roots possible.

Extension

After graphing these results, I was surprised that the bacteria levels decreased at a similar rate in each of the containers, in particular Model C, which did not contain any plants. Originally I had decided to stop taking samples after 72 hours, but after reviewing the graphed data, I decided the results merited further investigation.

First I reviewed the original theory. I hypothesised that both Models A and B, the containers with plants, would show large decreases in bacteria content within several days, because the E. coli would be absorbed through the plants' roots. I made the assumption that the content of E. coli in Model C would remain the same. I thought that without contact with plant roots, there would be nowhere for the bacteria to go. I was surprised when the levels of bacteria decreased at an even greater rate than those of the plant models.

I assumed that the E. coli in Model C must be sticking to the soil, explaining the low content of bacteria in the water. Several weeks passed before I decided to take a final water sample. I thought the bacteria in Model C may have multiplied and therefore increased the water's bacteria content. The increase in colonies provided evidence for this idea, although this could only be proven by taking soil samples from each of the containers to determine the amount of bacteria sticking to its surface.

However, I discovered that bacteria stuck to a solid surface behaves differently and has different properties than bacteria colonies formed in water. (Sorger 2004) In addition to this, depending on the properties of the soil, it is possible that the original bacteria died off, and the increase seen on the plates was another type of bacteria which the water had been exposed to.

552 hour (23 day)results

Model C: increase in
Bacteria