The "millets triangle" was introduced in 2022 as a way to bring all cultivated millets together in one conceptual schematic, which could be used to convey certain kinds of information about them. It was revised in that same year, and recently I revised it again (this 2024 version is attached).

The subject of millets can be a bit daunting. There are many of them, their common names can be confusing, they come from different regions, and some are widely grown & consumed, white others are very local. And behind all of that, the grains we call millets are botanically somewhat diverse.

The placement of millets in the diagram is organized with three sets of relationships in mind:
1. Level of production, considered ona global level. So the millets are listed on four levels according to some rough criteria. 
1a. Sorghum, which in terms of area planted and quantity of grain harvested is in a class by itself - totalling perhaps twice the volume of all other millets combined. This is a global crop with total annual production in the neighborhood of 60 million metric tons.
1b. The 4 most produced grains surnamed "millet," namely pearl millet, foxtail millet, proso millet, and finger millet. These have production probably between 2-15 million MT/yr, with pearl millet on the upper end of that range, and the others in the single digits. (Pearl millet by itself is said to have as much production as all the rest of the millets.) Each of these is a significant crop in more than one world region, and is traded to some degree internationally.
1c. Several other millets that are significant grain crops mainly in one world region and have annual production well under 1 million MT. Additionally there is some wider trade in these grains or products made with them.
1d. Several more millets that tend to be more local, with production probably under 0.1 million MT/yr,sometimes well below that. Trade in these is generally more local or perhaps regional.  Included on this level are three genera - Panicum, Setaria, and Echinochloa - in which some number of species are or have been grown or managed for food in certain limited areas

2. Geographic origin. With only very minor exceptions (important in their own right, but rarely discussed on a global level), all the millets we are talking about come from two continents: Asia and Africa. So, in this diagram, millets of Asian origin are mainly to the left and those from Africa to the right. 

3. Botanic relationships. Millet species of the same sub-family (of Poaceae), "tribe," and genus are placed in proximity (adjacent cells), to the extent possible, given the above criteria.

Changes from the previous version:
* Adlay / Job's tears is moved up a level, and browntop millet down. The former appears to have a larger production (on the order of 0.2 million MT/yr) than the latter, which is produced mainly in India (one estimate of production there was ~0.02 million MT/yr).
* The open "others" cell is now "Other Echinochloa species." This is based on the fact that species in this genus other than E. frumentacea and E. esculentus are discussed in the literature as being in active cultivation. These include E. colona, and some others in southern China. These "barnyard millets" are for me at least the most confusing part of the millets space.
* Cells shifted on the 3rd and 4th levels to retain proximity of geographic origin and of botanic relationships.
* Edits to nomenclature for the two African Digitaria millets, known collectively as fonio. For most purposes, including international marketing, "fonio" means D. exilis, the more widely cultivated of the two. The other one, D. iburua, which is local in parts of Nigeria and into Benin, Togo, and perhaps Niger, has been known as "black fonio," due to the dark appearance of its grain hulls. My impression is that this has led to a back formation of "white fonio" for D. exilis. Currently, one also sees the Hausa loanword for fonio - "acha" - used to distinguish D. exilis from D. iburua, which is by the same toke referred to as "iburu" (the local name in central Nigeria that is incorporated into its scientific name). So the current rendition of names for these two in the millets triangle attempts to cover all bases.
* The cells for raishan and Guinea millet were rotated horizontally in order to accommodate their respective geographical origins and botanic relationships (i.e., keeping those in adjacent/contiguous cells per the above criteria).

How to use the "millets triangle"?

The triangle can be used as such to facilitate an introduction to millets. It is possible to shade or color selected cells in order to situate discussion of certain millets in the larger context of the group as a whole. In the attached image, there are two examples.

What's next?

This is an ongoing project, but I personally do not anticipate any major revisions at this time - unless anyone spots some serious error(s). A new presentation of it and potential uses is envisioned. It would be nice to have a more polished image than what I was able to produce with my current skills and available software, if anyone is interested in helping on that. A dynamic and hyperlinked version could be

Since it is open source, others may modify or adapt it as they wish, but please note the source (per the CC license). 

One could arrange the millets in some other fashion. There is certainly nothing sacrosanct about a triangle or pyramid form for this  subject - that is just what occurred to me when thinking about the relative production levels of these crops. The rest of the thinking about botanic relationships and geographical origin then seemed to layer well (if not always easily) onthat base.

Anyway,feedback and suggestions are invited.
 
Don

Don Osborn, PhD
(East Lansing, MI, US)
North American Millets Alliance