[This post is copied from the archaeobotanist blog] Neumann et al. in a new Quaternary International article "First farmers in the Central African rainforest: A view from southern Cameroon", report a combination of archaeobotanical, apynological and historical linguistic evidence for the nature of early Bantu economies on the northwestern rainforest along margins of central Africa in the First Millennium BC. This includes updated and important discussions of pearl millet, tree nut use (like Canarium and oil palm). These societies brought savanna millet agriculture with them and took advantage of drier conditions to cultivate millet in marginal forest environments, while utilizing (and managing?) forest tree resources as well. Of relevance to those who have argued that bananas were fundamental to early Bantu economies in the rainforest zone (e.g. Blench 2009), however, is the lack of evidence for bananas in these newer excavations. The article includes a short paragraph on bananas, with some quite critical comments on the issue of early African banans ("act two" in thehistory summarized below on the archaeo botanist blog). They note that study in their samples "several thousand phytoliths already counted, no evidence for Musa could be detected. This sheds further doubt on the banana phytoliths from the contemporary third millennium BP site Nkang" and also they argue that, "There are also ecological arguments against cultivation of banana during this period. As is shown in the following, the climate was much more seasonal in the second half of the third millennium BP and thus unfavourable for plantains which require a humid climate without any major oscillations".
So the debate is out in the open. I don't think there is question of whether the reported phytoliths of Nkang are from Musa, but the worry is surely whether these phytoliths are actually of Iron Age date. They are not directly dated, and the possibility of intrusive or contaminating material from later, when bananas are such a prominent part of the present landscape, contamination is what we need to worry about. On the other hand Nkang is not in exactly the same area as the site studied by Neumann et al, so supporters of the early banana hypothesis might point to diverse and varied economies in the Iron Age. All the more reason to chase more archaeobotanical sampling in the region: we are still reliant on a just a few sites.
Nkang is at present the main data point in African archaeology that supports the notion of a prehistoric, first millennium BC diffusion of tropical crops from Asia to Africa ('the trio'= bananas, taro, and Asian yam), as summarized in the sealinks Antiquity article last year.