First, a disclaimer: The Kirkwood Gaps are not visible gaps like the Cassini Division in Saturn's rings. There are pleanty of asteroids in the Kirkwood Gaps. However, most of them are asteroids with semi-major axes that do not equal the gap distance, but with enough eccentricity that they can cross the gaps. There are very few of them that possess semi-major axes equal to those of certain resonances. So the Kirkwood Gaps are only gaps in a histogram of semi-major axes like the image Mal posted.
Here's the simulations I promised you:
Asteroids in Jupiter's 3:1 interior resonance go around the sun three times for every time one time Jupiter goes around the sun. So their periods should be 1/3 Jupiter's period. Jupiter's orbital period is 4331.572 days. So their orbital periods should be 4331.572 /3 = 1443.85733333333 days. Using the formula

, and converting everything into proper units (just use the calculator here:
http://orbitsimulator.com/gravity/articles/formula55.html , enter 1443.85733333333 for P, choose "d" for units, enter 1 for mass and choose "solar masses" for unit, and choose AU for output units), asteroids in Jupiter's 3:1 interior resonance should have semi-major axes of 2.50 AU.
The simulation
http://orbitsimulator.com/gravity/simulations/j3_1.gsim begins with an accurate representation of the solar system's 8 planets, plus Earth's moon and Pluto. Then 90 massless objects are added near Jupiter's 3:1 interior resonance. There are 10 each at:
2.46 AU
2.47 AU
2.48 AU
2.49 AU
2.50 AU
2.51 AU
2.52 AU
2.53 AU
2.54 AU
The objects are color-coded and are given names that reflect their initial semi-major axis. For example, "2p47 28" has an initial semi-major axis of 2.47, and it is the 28th asteroid in the simulation.
Let the simulation run at a time step of 16384 seconds, and in 3 million years, you'll find that most of the objects that started at semi-major axes of 2.50 and 2.51 (2p50 nn & 2p51 nn) are no longer in circular orbits, and many have very high inclinations. All other asteroids are virtually unchanged.
If you don't want to wait the simulated 3 million years, the simulation
http://orbitsimulator.com/gravity/simulations/j3_1_3millionYears.gsim shows this system 3 million years later. This simulation begins in "ecliptic view" so you can see the high inclinations. Drag the scroll bar to the top to return to a top-down view.