|MadSci Network: Agricultural Sciences|
Removing the microbes from soil is not an option. What you need to do is kill them (called "sterilization"), hopefully without changing anything else in the soil very much.
The best way to kill microbes and not do much to the rest of the soil would be to apply about 50 kilograys of gamma rays or x-rays, but most of us don't have the facilities to do this. The main alternatives, then, are wet or dry heat. If your family has a pressure cooker, small quantities of soil could be sterilized in glass jars with steam under 15 pound-per-square-inch pressure (the temperature will be about 250 degrees Fahrenheit) for at least 15 minutes. Dry heat sterilization calls for about 340 degrees Fahrenheit for 2 hours; most regular kitchen ovens can do this. Either method will kill the microbes (bacteria, fungi, actinomycetes, etc.) in soil, but the cooked soil won't be exactly like the original soil, minus the microbes. Consider how different the Thanksgiving turkey and stuffing looked, coming out of the oven, from when they went in. The minerals in soil will probably not change much, but "organic matter" (the part of the soil that used to be part of a plant or animal) is affected by the high temperatures needed to kill microbes. This may have some effect on how plants grow in the soil, but it can't be helped.
OK, let's assume that your heat treatment killed all the microbes, and now you want to run your experiment. Those of us who do microbiology for a living learn early that "sterile" is a temporary condition. If I scoop the sterile soil out of the sterilization container into a flowerpot, there will be bacteria, etc., on the scoop and on the inside of the flowerpot that will get into the soil at that stage, unless the scoop and the flowerpot were sterilized the same way that the soil was. And I had better not touch the soil with my fingers, even after I've washed them, if I don't want skin bacteria on it. The transplanted plants or seeds will also have bacteria on them, and perhaps other microbes; you'll have to water the soil to make the plants grow, and water will also have microbes in it, even if it's chlorinated or boiled (although it can be sterilized in a pressure cooker).
So, does that mean you can't do your experiment? Not necessarily. The real question an experiment like yours should probably be asking is how the SOIL microbes affect plant growth. There are a wide variety of microbes that live in soil, growing actively some of the time, and just hanging out the rest of the time, depending on temperature and available moisture. If you kill all of these, the bacteria from your fingers or from boiled or tap water probably won't replace them in soil. If you grow your plants from seeds that have not been in contact with the soil, and either sterilize your flowerpots or buy new ones that have not had soil in them, you can probably determine how soil microbes affect plant growth. Be sure to include a "control" pot or pots with soil that was not sterilized.
If this were a university research project, you might go a step further and try to learn how much baking the soil had messed it up. This would call for a third pot or set of pots. The soil in these would be sterilized, but re-inoculated with a little bit of real soil microbes (which should grow in warm, moist soil, just like the plants). The way to get these would be — when watering a pot that unsterilized soil in it, gather some of the water on the soil surface with an eyedropper (or even a spoon) and put this water on top of the sterilized soil in another pot. With time (days, at least, but your plants don't grow all that fast, either), these "inoculated" soil microbes should be carried down into the soil with successive waterings and take up residence. Plants growing in this or these pots will have real soil microbes around them, but the soil will have been previously heated. If anything important in the way of organic matter got taken out by heating, the microbes probably won't be able to put it back, so you'll be able to tell whether any differences between your sterile-soil pots and your control pots was really due to lack of microbes.
When doing experiments like this, it's good to change the positions of the flowerpots occasionally, just in case one location gets more light or a different temperature than another. And of course, even dust from the air carries microbes, so you want to be sure that not too much stuff from the air lands in your pots.
Try the links in the MadSci Library for more information on Agricultural Sciences.