This graph is from The Paleo Answer (Loren Cordain). It shows the proportion of plant food versus animal food consumed in isolated populations (on a total of 228 populations).
Frequency is the number of "tribes" examined.
Dependence by class interval (%) is the percentage of plant versus animal food (in calories).
For example, there are 45 populations relying on 26-35% on plant food, the rest of the food is from animal origin. There are no populations relying entirely (86-100%) on plant food. This means that human diet contained animal food, there are no vegetarians in the populations examined. The highest number of these populations relied on roughly 1/3 of plant food and 2/3 of food of animal origin. In India, people became vegetarians for purely religious, and not for health reasons. The irony is that the hindus (vegetarians) suffered more from diabetes and cancer than other meat-eater populations, for instance the Christians and Muslims, or the Inuit, who where decidedly "carnivorous".
Our adaptation to a certain proportion of animal food in our diet suggests that we also evolved to absorb nutrients from these foods. In fact, vitamin B12 can only be absorbed from animal food. You can read more about this in Chris Kresser's article.
I often get emails from people saying that in certain studies the conclusion is that a vegetarian diet is healthier than an eat-what-you-want diet. For this, I refer you to Denise Minger's article on the mistakes of these studies.