12 December 2012

Some Carrot (and other crop) Breeding Concerns

Carrots need to have a wide gene pool in the population or after a number of generations, inbreeding depression may occur resulting in poor, unproductive plants.  So it’s not a good idea to collect seed from only one or two plants - just like I have done with my self-seeded Baby carrots. So how many should you use? Best advice is a couple of hundred plants, but a minimum of around 50 is about as low as you should go.
I need to make a slight diversion. Hybrids, crosses between different varieties of the same vegetable, are known to be desirable. They often have better vigour than either parent, so modern agriculture has really focussed on the development of hybrid varieties. For inbreeders like tomatoes, this isn’t too hard- remove the pollen bearing bits from a flower before it has properly opened, get some pollen from another variety, transfer it to the first plant, and wait for fruits and seeds to develop. Each fruit produces numerous seeds, and each of these child plants will produce lots of fruit for market. So it’s commercially viable to produce F1 hybrid tomatoes year after year by crossing the original parents each year to produce new F1 seed.

But some vegetables, like carrots and parsnips are problematic. You need to have a big population of plants to collect seed from, the individual flowers are tiny, and they are bunched together in big bundles called umbels (thus the scientific name for this family of plants, the Umbellifereae.) So ensuring that the plant doesn’t pollinate itself is next to impossible. You could isolate individual flowers in the flowerhead, remove the pollen, and transfer pollen from another variety, but that would only give you one seed for each crossed flower, producing only one hybrid carrot for market – not a viable concern, really, and not enough to plant out a whole field of carrots or parsnips.

So how come some seed sellers advertise F1 hybrid carrots and parsnips? How do they ensure that every seed is the cross between two parent varieties?

It just so happens that there is a mutant form of some vegetables that don’t produce pollen. All the rest of the flower apparatus is in working order, they just don’t produce pollen. This is known as cytoplasmic male sterility ( I'll do a full post on this later). If this form is one of the parents, then the seed collected from these must have been pollinated by another plant. So grow one row of pollen free parsnips next to a row of a pollen producing variety, and only collect seed from the pollen free ones. Easy! But because of the nature of cytoplasmic male sterility, none of the F1 children will be able to produce pollen – this is not a self sustaining population. Unless there is some pollen producing plants around, the crop will die out.

So this is not a desirable characteristic to introduce into a breeding population if you want to develop a new variety to share.
Unfortunately two of the coloured carrot varieties I was planning on to supply coloured genes for my project are hybrids – so I can’t incorporate them into my breeding mix. But I do have a couple of dozen white, a few yellow yellow, and a couple of  purple carrots to provide some color diversity - hope they work!


  1. Just found your blog! I've saved carrot and parsnip seed and always try to give them descent population but hadn't thought about this conundrum in regards to commercial hybrids.

  2. Thanks O G. Joseph over at Homegrown Goodness put me onto the whole male sterile thing, and I've just noticed it in my flowering carrots. I've ot some pics so will make another post.