In the human past, is the perennial consumption of starch a deep or shallow phenomenon?

Recent research has argued that digestible carbohydrates, rather than just meat, were necessary to accommodate the metabolic demands of the evolving human brain. This argument is supported by genetic and biological evidence that Homo sapiens was especially adapted to a starch diet within the Palaeolithic, making starch consumption a deep phenomenon. Yet there is no deep time archaeological evidence of starch consumption by early humans. A significant body of evidence attests to the increasing starch diet of humans from the Upper Palaeolithic, through the advent of agriculture into the present day where cereals represent 70% of worlds energy intake. Without archaeological evidence of regular early human starch consumption, this appears to be a shallow phenomenon on the timescale of human evolution. Based on botanical evidence from Klasies River and Blombos Caves, among the earliest human Middle Stone Age (MSA) occupation sites of South Africa, this thesis makes three key arguments: firstly that cooked roots and tubers provided a high quality starch diet for MSA hunter-gatherers from at least 120 thousand years ago in the Cape region; secondly, that pounded, cooked tuber remains and mixed plant food 'recipes' from two hearths (85 kya and 81 kya) at Blombos Cave represent a consistent subsistence strategy through climatic change, one hearth dating to a warm interglacial cycle and the other to a very cold glacial cycle; thirdly, that cooking and processing roots and tubers represent a dietary mainstay for MSA hunter-gatherers in this region for at least the 55 ky span of the hearths sampled. The collection and analysis of these data involved an innovative extension of method to overcome difficult preservation conditions. The analysis also benefited from the creation of a regional parenchyma modern reference collection and co-operation with geoarchaeologists to understand the micro-context of each hearth, especially at Klasies River. Combined these method extensions have provided better quality data analysis. This research offers evidence of the earliest yet known evidence of early human consumption of cooked and processed starchy plants from the MSA of the Cape coast of South Africa. The findings conclude with the hypothesis that perennial starch consumption is a deep phenomenon in the evolution of Homo sapiens. The further question must be: How deep is the phenomenon in the ancestors of Homo sapiens?