Return to Andigena Accessions Index  

Characterizing USDA Andean Potato Accessions

This page has attracted a lot of traffic recently and generated some questions.  The most common has been whether we sell the varieties pictured.  We don’t.  I make selections from the genebank accessions that I use for breeding, but keep very few overall.  I save seeds from the best of them and create true potato seed mixes that you can use to select your own varieties.  For casual growers, I recommend that you start with one of those, for a variety of reasons.  I have already picked the accessions with the greatest potential, the seed is fresher and should be easier to germinate and more vigorous, and I offer more seeds per packet.  For the average home grower, growing from our mixes vs. genebank accessions probably saves you at least two years of work.

The USDA National Plant Germplasm System holds a large collection of Andean potato germplasm and makes it available to breeders, researchers, and educators. They maintain just about 1,000 available accessions of Andean domesticated potato species. Unfortunately, 93% of them are true seed accessions with little to no morphological description, so it takes up to two years just to figure out whether or not an accession has traits that you want to work with. The collection is really fascinating, perhaps as much for that incompleteness as for its huge number of accessions.  These accessions were collected all over the Andes, from Venezuela to Argentina, beginning in the 1930s and mostly ending in the 1990s.  In some cases, there is a snippet of description about the collection, such as whether the plants had berries or what the tubers looked like.  More often than not, though, there is no description at all.  Collections were made at markets, farms, in home gardens, in roadside ditches, anywhere a collector spotted a plant.  Some are misidentified.  Many are probably hybrids with wild species, with traits like long stolons and high glycoalkaloids.  This is both fascinating and frustrating.

I started out by following the trail of breadcrumbs in the GRIN database and requesting accessions based on traits that might be useful in my work, like high pollen production, high tuber production (a likely proxy for long day tuberization), and disease resistance.  I tried to match these traits with varieties that had original names that were descriptive of morphological traits that I was interested in.  Some of the accessions have Spanish, Quechua, or Aymara names that mention shape or color.  This was probably better than blindly selecting varieties from the database, but not by much.  Many of the traits noted in the database are inconsistent from one study to the next and also inconsistent with my results.  For example, I initially avoided accessions that were marked as no or minimal tuberization, but many of those accessions have turned out to have high tuberization under my conditions.  A lot of this can undoubtedly be attributed to the great genetic diversity of the potato.  These are mostly true seed accessions and growing a small number of seedlings won’t always give a good representation of the trait distribution.  Having decided that the descriptions weren’t worth much, I decided to request the majority of diploid andigena accessions over several years and evaluate them.  That was a lot of work, but it was invaluable, both as a method of identifying valuable breeding lines and learning about Andean potatoes.  There aren’t that many ways to learn about these potatoes short of rolling up your sleeves and growing them.  Growing the diploids taught me a lot and convinced me that there was more to learn, so I decided to start in on the much larger number of tetraploids and that continues.

When I began this project, I just wanted to identify some materials to use for further breeding, but the experience of working with the accessions also unexpectedly woke up some exploratory drive in me.  I learn so much growing them and see so many things that I never expected to see that I increasingly just want to discover what comes from that nameless potato that was collected in a ditch in Bolivia 60 years ago, whether it is immediately useful or not.  As I worked through these accessions, it occurred to me many times that I can’t possibly be the first person to grow these.  In fact, I know other potato breeders who have grown quite a few of them.  It really seems like a shame that people have had to rediscover this information again and again.  If you search on most of these accessions in Google Scholar, you will rarely find more than two or three matches and those will almost always be wide surveys of pest or disease resistance.  If you want to know something as simple as what the tubers look like, or whether the accession leans toward domesticated or wild traits, you’re out of luck.  So, I decided that the least I can do is grossly characterize these accessions and take some pictures.  Even that small amount of information will probably help people to avoid wasting their time growing accessions that have four foot long stolons or that are obligate short day tuberizers.  That would have been a huge help to me and I hope that this will serve as a helpful supplement to others who are trying to use the collection.  Also, this is a function of our government that most people are not aware of and will never experience in any direct way.  With pictures, you at least have the opportunity to see and appreciate some of the agricultural diversity that our government holds in trust for all mankind.

I don’t know if I will ultimately grow out all 1,000 accessions.  I have about 300 done or in progress and that has yielded more genetic diversity than I could ever conceivably use.  On the other hand, I have learned so much from growing these accessions that I can’t help but think that I would be a shame not to finish.

Thanks to the United States Potato Genebank for maintaining and providing these accessions.

The Process

I aim to grow 16 to 32 plants of each accession, but usually fall short of that.  We grow in raised beds and the number of plants is dictated by the size of those beds.  Each bed accommodates two rows of 16 plants.  Most accessions are supplied as true seed in packets of 50 seeds.  Germination rates vary widely.  On average, I get 9 seedlings suitable for planting out from each packet of 50 seeds, so I often have less than 16 plants.  In that case, I cross whatever I did get to bulk up the seed and then try again the next year.  In many cases, I get a good yield the first year.  In some cases, the seedlings are late, slow growing, or have puny yields, in which case I grow them again from tubers the next year for another look.  If I end up with three or fewer plants to evaluate, then I request the accession again for a second try.  If I don’t do any better on the second try, I usually give up.

My germination rates are generally worse than those reported by the USDA, sometimes much worse.  I’m not sure why.  I start potato seeds under controlled conditions, 65F during the day and 55F at night.  I grow them in 4 inch pots with fresh potting soil.  I have no trouble with these conditions with my own seeds.  I have experimented with treating seeds with gibberellic acid as recommended by the Potato Introduction Station, but I have found that this results in worse germination almost as often as it does improved germination.  I also think that the vigor of seedlings from the gene bank is usually inferior to those from fresh seed, probably as a result of age.

So far, I have kept clones for further breeding from about 1 in 4 accessions.


They say that a picture is worth a thousand words, which is great because I don’t have time to write a thousand words about each accession.  Pictures can also be misleading though.  Because it takes me a long time to harvest and photograph thousands of plants, I pull most of them long before maturity.  That means that you shouldn’t draw any conclusions about size or yield from the pictures.  Pictures are taken at different times of day under natural light and I don’t have time to go back and take another picture of one doesn’t turn out well.  The pictures are primarily useful for showing tuber shape and color.  If an accession has noteworthy flesh color, I usually also take a picture of some cut tubers.  The pictures can also give you a rough idea of how well a variety tuberizes under long days.  Pictures that include many different phenotypes typically indicate at a glance that the accession is mostly long day.  Pictures that show only one or two phenotypes probably indicate that the accession is mostly short day (or that it simply had poor survival, which I note in the description).  Almost everything is harvested before the autumn equinox, so the pictures show only long day phenotypes.