First, let's identify the independent variable. This should be something we have control over and already know. In this case, what plants are in the soil. The three plants we'll want to have are smooth cordgrass, dwarf glasswort, and barley cordgrass. We should also have a control, which in this case could be no plant.The control will help account for non-plant related things that could remove salt from soil, such as water.The dependent variable, then, will be the salt content of the soil. We should also keep some constants, or what stays the same for each condition. This could be length of time, type of soil, time of year, amount of water, location, etc.
Now let's come up with a hypothesis. This does not have to be correct! But it should be based on some criteria. Dwarf glasswort is the succulent of this list, so my hypothesis is that dwarf glasswort will remove the most salt from the soil.
Now we can put all that together for an experiment.
What we could do is go to the same plot of land and randomly select some patches of soil, let's say 40. Then we'll randomly select 10 of these patches for each condition (no plant, smooth cordgrass, dwarf glasswort, and barley cordgrass. We're looking for the change in salt content, so let's measure the salt content of the soil before putting the plants (or no plant) in.
Now we'll put in the plants (or no plant) and wait for 6 months to give time for the plants to grow and establish. During this time, we'll give each plot the same amount of water, light, etc.
Now after 6 months, measure the salt contents of each soil patch. Identify the average change in soil content for each condition, and find out whether any of these changes are statistically significant from each other. Then find the condition with the greatest statistically significant decrease in salt content from the control (if any. The plants might not impact salt content at all or could even increase it)
