When you are comparing species richness between two types of land management, how do you determine which is better? Can you measure the mean and...

It really depends on what you're trying to determine.

The t-test would make sense if you have some species richness estimates for plan A and plan B which have means and standard errors attached, and this is the question you're trying to ask: "Will plan A most likely result in more mean species richness than plan B?"

But maybe you're not as worried about mean species richness; maybe you are risk-averse, and one plan is riskier than others. In that case you'd want to maximize expected utility where you have some utility function defined over species richness (maybe log of richness or something like that). That might mean taking the plan that has a lower mean but also a lower standard error.

Also, the t-test only works if you're only comparing two options. To compare 3 or more options, use ANOVA.