Part of a batch of Sidra coffee seed in Ecuador grew into plants genetically closer to Kenya's Batian

Coffee takes three to four years to reach significant production. Over that period the grower pays for seed, planting and crop management, expecting a harvest with specific characteristics for a specific market ^[11]. On this farm, the check came at the low end of that window, around the time the trees would have started producing a crop ^[17].

The label matters because a variety's identity is tied to genetic traits that shape growth, productivity and cup quality ^[15]. The team used two lines of evidence. They commissioned a molecular-marker analysis from a coffee genetics specialist, comparing short regions of DNA across plants ^[7]. They also scored traits visible in the field, including plant architecture and the shape of leaves, branches and fruit ^[8]. Half of the four morphotypes did not carry the identity printed on the seed purchase ^[16].

The two mismatches differ in kind. Batian is not part of the genetic group of Ethiopian landraces that Sidra is related to, so that morphotype comes from a different lineage than the one ordered ^[3]. The Gesha plants did sit inside the expected group. They were an Ethiopian landrace, and the discrepancy is that they were not identical to one reference accession held by CATIE in Costa Rica ^[4].

How cheap these checks can be is less clear from this work. Verification methods can be costly or hard for producers to access, according to the phys.org account of the study ^[12]. The cheaper tool the study points to is visual: taken together, the physical traits assigned a high proportion of plants to the right variety on this farm ^[10]. The account gives neither a price for the marker analysis nor the exact share of plants the visible traits classified correctly.

The authors are plain about the denominator. "The study does not allow us to determine how common this type of discrepancy is in Ecuador, as it was carried out on a single farm with a limited number of plants. Nevertheless, this case highlights the need for identification tools and certification systems that are accessible to producers," the researchers said ^[13].

I think the case supports checking planting material before it goes into the ground, with the condition the authors set. It shows that a mislabel can happen and can stay hidden for about three years ^[6]. It does not show how often that happens ^[13]. It also leaves open whether a DNA test or a field screen of visible traits is the cheaper first check, because the trait screen has so far been tried only on the varieties growing on one farm ^[10].