Pretty off-topic but just a little video about ‘unrelated identical twins’ as an example relating to the traditional landrace phenomenon of having members in a population that have the same phenotype but different genotype:
This is kinda same thing that annoys me in some facebook groups when people ask id for let’s say tomato and people who have grown 3 tomato varieties come to identify it as something they have grown because it’s small and red. Many times it’s only the fruit that looks the same and there atleast subtle differences in growth. Sometimes in peppers the pepper is clearly different species. Even though domesticated species have lost lot of their diversity there is still a lot of unique compinations. One of those things where the more you know, the less you actually know to know.
Not to mention invisible traits, like disease resistance and heat or cold tolerance. Or, for that matter, how much fertilizer a plant expects to be grown with, how drought tolerant it is, what kind of soil it wants, and what insects it’s good at protecting itself from. A lot of the most important traits are invisible at a glance.
It can be a very positive thing. Having close enough phenotypic similarities can be good in terms of marketing and food consumption, but with genetic differences that make the population more resilient. While ‘modern landraces’ may have much more phenotype difference, it seems traditional landraces are quite consistent in phenotype and for SC species have a lower rate of crossing than the modern landrace method, but that would seem to be because the modern method is focusing on the beginning stage of creating a new landrace quickly, or at least on the beginning stages, whereas the traditional landrace, so far as I can tell, is generally something that has been well settled for centuries or even millennia, and so only needs a low rate of outcrossing, to handle local fluctuations in conditions.
I seem to remember interesting info on this regarding potato populations in … probably the Andes, for example.
So in such cases, things like taste, how much fertiliser needed and so on, should come under the phenotypic homogeneity, since they’d be grown all in the same field for so many generations, and eaten by the same people in the same dishes.
But I also like the modern idea of limiting how many phenotypes we try to stabilise. Like Joseph’s squashes. Choosing the few phenotypes that are important to stabilise, like size (for cooking time and cooking method), and taste, but allowing for different shapes for example. And, with the aim of sharing seeds around the country/world, the high diversity of the modern method is eminently useful for enabling sharing with a wider array of conditions.