Jared Diamond repeatedly credits the availability, in different wild environments, of plant species that are edible and suitable for domestication as a major factor in determining the technological viability of different cultures. For instance, he suggests that agriculture in what is now the northern United States might have begun several thousand years later than in the fertile crescent and suggests that the primary cause for this is that the Americas had fewer plant species that were available for domestication.
He makes a similar argument about the land that is now considered New Guinea to suggest why the indigenous people there became colonized rather than becoming colonizers themselves. Diamond discusses how the food crops there which could be domesticated could not be grown at high elevations and did not provide significant sources of protein. He argues that this—combined with the fact that they did not domesticate a cereal grain because there were no large-grained grasses native to the area and that they had no large animals to domesticate as livestock—meant that the people of New Guinea had an extremely protein-limited diet. This all changed when the sweet potato arrived from South America and began to be cultivated. The new crop, which tolerated higher elevations, gave higher yields, and provided more protein allowed for a population explosion.
Diamond tracks a similar population explosion as this in the northern United States which he associates with the arrival of the Mexican "trinity" of crops: corn, beans, and squash. These crops, when grown together, nourish each other and the soil they are grown in and, when eaten together, provide a substantial amount of plant protein.
Diamond then extends this argument even further to suggest that places on the earth where agriculture was never developed as a technology were specifically restricted by the limited availability of native plants that were suited to cultivation or agricultural approaches.