How I started apple breeding

Karim with new apples from his seedling 73

When I joined Keepers Nursery in 2010, as someone planning to spend his career growing and propagating (grafting) apple trees using material from a huge range of well-known and established varieties, the concept of apple breeding (growing from seed) to create new varieties wasn’t on my radar. Surely the world has enough varieties already? It is well documented that it takes in the region of 10,000 specifically crossed seeds to create a new commercial variety (Google it right now). A typical commercial breeding program will take about 20 years to release new varieties which consists of 5-10 years for the seedling to fruit and then a period of testing and trialling the best prospects. As a Biology graduate from Oxford University I had some idea of the amount of work which would go into these big, long-term projects, not something a small, family-run nursery could attempt… 

But then again who doesn’t want their own unique apple variety?

I was also curious to know what the fruit of a seedling apple tree was actually like. Since apple seedlings don’t produce fruits the same as their parents and each tree produces a new and unique variety of apple, would the fruit be a mixed up version of the parents? My father, an Oxbridge scientist himself and nurseryman for more than 20 years at the time, gave me the stark stats. I recall he suggested that most seedlings would be sour and green “spitters”, as they are known in the States. Surely they can’t all be bad? Surely most of the best known apples weren’t part of an epic 20 year breeding program? I had come across wild seedlings, trees “accidentally” grown from seed in hedges for example, their fruit often lacked good attributes. I had also read stories of wonderful varieties such as ’Cox’ being a discovered accidental seedling , surely the discovery of it wasn’t the 10,000th time someone tried an apple from a wild seedling tree? 

So, in the autumn of 2015, stubborn and full of youthful determination I decided to have a go. I selected three varieties - Kidd’s Orange Red, Rajka and Rubinola - and set about germinating about 150 collected seeds from information I had found on wikihow. This was just one of many spinning plates I had going at the time, and my enthusiasm for the project waned over the next few years - it takes years for anything exciting to happen in apple breeding!

The seedlings were grown in pots, most of which survived until March 2017, at which point I planted the seedlings in a long row in the ground. Left alone, a seedling tree will have a long juvenile stage of up to 10 years, during which it will not fruit. I had read that to reduce the juvenile stage one should graft wood from the young seedlings onto a precocious M9 rootstock so I did this in August 2017. 

The newly grafted seedling trees grew well over the following few seasons and in late April 2020, seven seedlings flowered for the first time. By that September, I was actually eating new apple varieties of my own creation! They all looked different, ripened at different times and most importantly all tasted different. One early apple was exceptionally sweet, unlike anything we already had in our collection of about 700 apple varieties. Too sweet, perhaps but certainly not a sour “spitter”. All but one of the remaining six were unremarkable. Seedling 73 was very good, easily as good as the varieties I already knew so well, I thought I must have been exceptionally lucky to have success so soon. How could it have been so easy to create a high quality apple variety? It had taken only 5 years, I had planted only 150 seedlings, 5% had fruited so far, I didn’t even know the pollen parent of this seedling-it had been created by nature. Was this a fluke?

The following spring, 2021, 25% of the grafted seedlings flowered. That summer, I grafted copies of seedling 73, whilst eagerly watching all the other new fruits develop. In Autumn 2021, 45 seedlings had fruited, many for the first time. I decided to record everything in detail; ripening times, descriptions and a score between 1-5. I decided to be very critical.

• 1 was a score for a bad apple of no worth.

• 2 was an apple with at least 1 major flaw which I might finish but wouldn’t eat a second time.

• 3 was a good apple with minor flaws, possibly worth keeping.

• 4 was a great apple with no flaws and is as good as the best known cultivars of the same season, these would be named and propagated.

• 5 was the unachievably high standard of ‘exceptional apple’, which I never expected to breed from such a small sample; an apple this good would probably outlive me.

That season, most apples were 2s or 3s, a classic bell curve, and a very small number of 1s and 4s. These were similar results to the 2020 findings. Without jumping to conclusions, I could see a few trends. Openly pollinated seedlings (from within my huge nursery gene pool) are vastly different. The odds of creating new a high quality “garden” variety are small but it does not take 1000s of seedlings, and the odds that a seedling will be very unpleasant are also just as small. In autumn 2021, seedling 77 and 73 (again) were the pick of the pack and I was now very excited about the remaining 100 or so seedlings left to fruit.

In autumn 2021, high on my recent successes and keen to continue, I decided to collect more seeds. This second batch became known as the “2021 seedlings”. I germinated over 600 seeds that winter and these seedlings were eventually grafted in August 2023. I expect 5% to fruit in the summer of 2026 if they follow the same time-frame as the “2015 seedlings”. 

The process of raising 600 seedlings was very time consuming and in an attempt to cut corners I learnt the hard way that potted plants need watering every day and that rodents love apple seeds. I actually lost my entire original batch of 2021 seedlings to hungry mice in December 2021 and had to replace them with whatever I could find still on the trees. Statistically, I suspect several excellent varieties were lost that day.

In the 2022 season, 50% of the “2015 seedlings” fruited and my recording job became somewhat harder. I found that figuring out when an apple hits its peak between August and December requires dedication. I also found that some seedlings are inconsistent, much like many well respected cultivars, some 3s and some 4s became 2s and on some rare occasions some 2s and 3s became 4s; a lot depended on the season and my timing. I could now appreciate why trials can take up to 10 seasons. 

However I also started to contemplate how large scale commercial trials might miss things; they obviously have objectives and likely reject upon appearance alone. From my findings I’m certain the top 1% of a commercial trial could be very high quality apples that never get released, which is a shame. As fate would have it, I didn’t miss a very special apple.