Cavitation occurs primarily by two mechanisms. First, your pump can ingest air either by ingesting frothy water, or if the pump intake rises above the water, such as when you are hopping over a wave. Your pump sucks air, rather than liquid water, and your thrust goes to approximately zero. Certainly this is bad if you are climbing a rapid and need thrust to avoid calamity.
Second, cavitation can occur when liquid water turn into steam. This is similar to boiling water, where the liquid water turns into steam bubbles. At sea level, boiling occurs at 212 F. At higher elevations, the atmospheric pressure is lower, so boiling occurs at a lower temperature. For exampe, at 2,000 ft above sea level, boiling occurs at 208 F, rather than 212 F. As pressure further decreases, the boiling temperature also decreases. If you walk up hill, the pressure decreases, and so does the boiling temperature. For example, if you keep walking up hill to about 60,000 feet (more than twice Mt. Everest), then the boiling temperature is only 100 F, and your body will begin to boil.
Interestingly, when water is pulled into your jet pump, the water pressure decreases. If the pressure decreases enough, then the water will boil. In addition, gases such as oxygen that fish breathe, leave the liquid and and become bubbles. In this situation, the liquid water becomes filled with cavities of gas or steam (water vapor). Consequently, your pump thrust decreases significantly. In some situation this can be a simple irritation. In other cases, such as when climbing a rapid, it can be disastrous.
You can observe cavitation around rocks that are completely submerged in fast-moving water. On the downstream side, you can sometimes see bubbles appear "by magic." On the downstream side of the rock, the pressure can be very low, so the water boils and dissolved gases escape in bubbles. The same thing can happen if your intake grate is restrictive, or if rocks are stuck in your grate, or if you are simply pulling a lot of water through your pump.
When cavitation occurs, the engine's work deceases, so RPMs will increase quickly (possibly above red line). You may feel extra vibration, and you may also hear a high-pitched noise from the bubbles. The effect is reduced thrust, possibly when you need it most. Therefore, cavitation is something to avoid and prevent.
In my personal experience, a stock Berkeley cavitates somewhat in white water. It is actually quite effective, except in extremely frothy water. Some folks report horrible performance; however, my guess is some other extenuating circumstance is present, such as poor hull design. Even folks with Hamilton 212 (or similar) sometimes report excessive cavitation, so the pump isn't always to blame.
Relative to stock Berkeley performance, if you add an inducer impeller, then it won't cavitate at all. The inducer is a small impeller that sits in front of the main impeller, and effectively "pre-pressurizes" the water prior to the main impeller. The primary difficulty with an inducer is installation: the pump shaft must be pulled out forward. To make room for this, either the entire pump must be removed, or the engine must be removed. If you're not particularly mechanical, then you'll have to hire somebody compentent. If you are mechanical but lack an acceptable engine hoist, then you'll have to pull the pump. On some boats, pulling the pump is easier, and on others the engine is easier. Additionally, the inducer adds extra resistance to the pump, so it will decrease you max RPM by roughly 150-200.
Another down-side rumor is that the Berkeley is inefficient and lacks thrust. There is some truth to this, when compared to a Hamilton 212 (for example). Extensive side-by-side testing of two very similar boats showed almost identical performance of a Ford 460/Berekely and a Chevy 350/212. The Berkeley had an A2 impeller and inducer, and the 212 had a 2.4 impeller. Both setups had about the same max RPM of roughly 4200. All else being the same, the bigger engine should outrun the smaller engine. But it didn't. Top speed was nearly the same. It's also known that the 460 will turn a 3.4 impeller (in a Hamilton 212) to at least 4000 RPM, which implies substantially more thrust than a Hamilton/2.4. The conclusion is that the Berkeley is less efficient than the Hamilton 212. However, it is still plenty capable.