Imagine a prairie which normally suffers a fire every two to three years, but has not had one for 100. One would imagine that this lack of disturbance would lead to a flourish in species diversity in this community. This, however, is not the case.
Let us take some examples. In forests, especially those with a closed canopy (like tropical rainforests, deciduous forests, coniferous forests, etc.), not a lot of light can penetrate to the forest floor. This means that small plants and new trees will never have the opportunity to grow, thereby limiting species diversity. In the event of a fire, or the falling of a tree, however, the canopy will be broken up which creates new ecological niches, and new trees will have an chance to grow. In a redwood forest without any disturbance ever, there will never be anything but redwoods.
To much disturbance poses different problems. In forest with frequent fires, or the constant felling of trees, species will be destroyed because of the destruction of their niches. The population of a monkey whose preferred habitat is the upper canopy of a rainforest will not be destroyed if a tree falls. But if humans come and begin felling too many trees, the monkey could go extinct in that region.
A prairie which does not suffer any disturbance does not open up any new ecological niches, and limits species diversity. Perhaps the population of large grazers like buffalo has exploded because nothing threatens their primary food source, grass. This could have a negative impact on other species. An occasional fire would destroy part of the grass, and decrease the population of buffalo, giving other species an opportunity to recover. This theory that says that occasional disturbance is necessary for the diversity of a community is called the intermediate disturbance hypothesis.