Question about Ukrainian persimmon breeding

The Ukrainian program of persimmon breeding has produced some very noteworthy hybrids, all of which are some combination of Japanese kaki and American virginiana. But details seem lacking about the specific varieties used in the crosses. Does anybody know anything that they can report?

I had assumed that the kakis were all PCA because the hybrids are all astringent. The main point of the Ukrainian effort seemed to be to enhance the cold-hardiness of kaki-type fruits, not to reduce astringency. But a forum friend recently noted that Gora Roman Kosh reportedly loses astringency while still firm, which lead me to wonder why. Specifically, could GRK have enough PCNA ancestry to be genetically NA? If not, how do we explain the observation?


Currently, we don’t explain it with genetics. The diagnostic tools and understanding of functional biology simply aren’t that advanced for persimmon. Instead, we take comfort in the phrase “breeding happens”.


When species with such high ploidy levels as persimmons are crossed, all bets are off. One pair of chromosomes can have one genetic information, and another pair something completely different, and so on. You never know how this genetic mess will be expressed in the phenotype under various conditions.


You gotta have heat (warm summers) to remove astringency is my thoughts after what Stan said. Again, I don’t even know what ploidy means so I cannot process what you guys say when you use science words.

I think if trees aren’t stressed (watered when needed, fed; sprayed when attacked, et-) if the climate is good for the Genus/species, then it’s going to be pretty ‘solid’ in how it performs whether California zone 9 and hot summers or whether Virginia zone 7a hot summers or me zone 5b Illinois hot summers.

I wouldn’t put it past Ukrainian breeders to want PCNA or PVNA hybrid kaki characteristics. Since we dont’ know what they used but we do know that Nikitaskaya Bordovaya is about the #1/main cultivar crossed back to kaki male flowers to produce the hybrid-kaki type fruits in most all occasions. I already read that on the group. Arthus76 member wrote it.


@Stan – I think that Japanese and Chinese scientists have worked out the genetics of two routes to PCNA. Japanese PCNAs are homozygous for (6 copies of) a single recessive gene; Chinese PCNAs are heterozygous for (1 copy of) a single dominant gene. Two different loci. These two gene patterns appear to produce NA pretty reliably across conditions, though some warmth seems required for Japanese PCNAs to lose astringency totally.

What I’m wondering about Gora Roman Kosh is whether we have (1) some new genetic route that is highly reliable across conditions, or (2) a lucky genetic combination that just happens to express a NA phenotype in some conditions.

Why this matters is breeding. If #1, then maybe offspring of GRK will be NA too – GRK is the founder of a potential NA hybrid line. If #2, then maybe not – GRK is a fluke.


As part of my present research, I have been studying the work of the Chinese and Japanese investigators of persimmon, but in relation to a different fruit. Their statements in recent publications concerning total gene counts, correlations to specific RNA sequences in individual chromosomes, etc. are incorrect. I suspect this is due to (a) use of freeware floating around in academic circles, (b) use of faulty PCR methods, and (c) an eagerness to provide genomic evidence for the morphology they observe in their breeding programs.

Please note that I hold the investigators harmless in their activities. Much of the academic freeware has been used in published results and unfortunately is considered valid in the horticultural sciences academic community. The use of PCR to determine the presence of a sequence is certainly a valid practice but does not tell you where (which chromosome, what location) it occurs. But the use of PCR counts (the number of inflorescences) is a highly questionable practice that does not hold up when an examination of a full high-resolution sequencing of the genome is made.

Investigation of current, state-of-the-art sequences is no minor feat either. They are not entirely accurate and there are multiple types of inaccuracies. A simple search for occurrences of a subsequence is an erroneous approach. Generalized discrete 1D correlation is necessary. The computation time per chromosome strand using 8-way parallelism on a 10-core 20-thread 4GHz platform is over 24 hours. The software must be designed for fault-tolerance, e.g. in the event of a power failure the computation can be restarted near where it left off. The results from processing an entire genome produce a 2 x number of strands fold increase in data storage requirements, spread across a few 10’s of thousands of output files. All of this is within the realm of a computational scientist but it is rare to see one listed among the authors of horticultural science papers.


@Richard – I have zero training in anything other than very basic genetics, so I defer to your expertise. Being untrained, I can’t even begin to determine the implications of your criticism. So I will ask you to connect some dots. Specifically,

  1. Do you think the evidence supports a conclusion that NA in Japanese Kakis is determined by a single recessive gene? Your answer could take the form of a likelihood estimate, e.g. “It’s roughly 65% likely that . . . .”

  2. Do you think the evidence supports a conclusion that NA in Chinese Kakis is determined by a single dominant gene? Ditto.

  3. Do you think the evidence (growers’ reports not DNA analysis) supports a conclusion that NA in the hybrid Gora Roman Kush is heritable? We don’t understand the genetics but if “breeding happens” then maybe understanding is beside the point.

I ask these questions hoping the answers will help inform amateur breeders, such as members of this forum, as to where to devote their energies. A cross of a Chinese NA with an astringent DV seems beyond the capacity of us amateurs. A cross of Gora Roman Kush with any other persimmon might not produce NA offspring, depending on the heritability of the observed NA in the parent. I’ve been pushing a back-cross of JT-02 x Taishu because it seems feasible but that’s assuming our understanding of non-astringency in Japanese PCNAs is correct.


To be honest, the majority of my education beyond basic genetics has only occurred in the past 3.5 years and it is primarily focused on structural genetics, diagnostic (e.g. sequencing) techniques, and computational methods – all of which in relation to genetic ID and genomic determination of ancestry. However, I believe this enough to answer your questions.

  1. They (nor anyone else) do not have diagnostic equipment to determine basic parameters of gene measurements in the persimmon genome. In particular, I have recently demonstrated that the very same detection technique they are using returns values that have no correspondence what is actually in a genome. To make matters worse, there is no complete list of genes in the persimmon genome. One estimate is that we know 10% of them.

  2. Ditto.

  3. Time will tell.



OK, while I digest Richard’s response, I;d like to refocus the thread on the main question posed in the OP: Does anyone know anything about the specific varieties used in the Ukrainian breeding program?

1 Like

Since I read both Ukrainian and Russian, I tried to do some research online. A lot of info to digest, and I did not have enough time. A few quick observations:

  • It seems like Ukrainian persimmon breeders relied a lot on open pollination, so while you can follow (in some cases) the seed parentage line, the pollen line is mostly unknown.
  • According to some sources, ‘Gora Roman Kosh’ is an open-pollinated seedling of ‘Nikitskaya Bordovaya’.
  • In rare occasions, new varieties are sports. One example is ‘Pamyat Pasenkova’ (Memory of Pasenkov) – a sport of ‘Rosseyanka’, characterized by better resistance to late-winter and early-spring frosts.
  • Astringency is very rarely discussed in Ukrainian sources, it’s almost never mentioned in brief descriptions of varieties (in contrast to winter hardiness, fruit size, ripening time, pollination requirements and sweetness). I think, there are two reasons for this. First, PCNA kaki would not survive the winter unprotected anywhere in Ukraine except the southern coast of Crimea, and therefore they are not considered relevant. Second, obsession with texture (firmness and crunchiness) of fruit is a pure American phenomenon, while most people outside of the US do not care much about this. They only care about the flavor and hence have no problem waiting for fruit to soften and lose astringency, so this point is mute.

Great comments, Stan. Thanks.

So I infer from your comments that GRK could not be an accidentally non-astringent offspring of PCNAs. If the observed nonastringency is heritable then there must be a different genetic pathway.

One follow-up question: The V x K cross that produced JT-02 was accomplished with great effort. Apparently subsequent crosses of the hybrid JT-02 are easier and might be accomplished by open pollination. Did the Ukrainian breeders get the initial Virginiana x Kaki crosses by open pollination or by embryo rescue, etc?

This statement appears to imply that NA can only be achieved in nature by heredity, which is not true.

1 Like

@Richard – I don’t think so. It implies merely that heritable astringency can only be achieved by genes. I didn’t say “If non-astringency, then genes.” I said “If heredity, then genes.”

I get that non-astringency might also occur as an interaction of genes and environment. For example, hot weather might accelerate the loss of astringency in some varieties. That interaction would be worth documenting. But what we really want is non-astringency that is robust across environments.

The pragmatic issue is whether Gora Roman Kosh can be one parent of a family of non-astringent hybrids. I figure that if GRK is non-astringent because of robustly non-astringent genes, then the answer could be yes. If GRK is non-astringent only under restricted environmental conditions, then the answer would seem to be no.

1 Like


It could also occur without interaction with the environment. Chromosomes and biologic entities in general are dynamic - very different from static mechanical devices.

1 Like

Ok, I’m willing to learn. Can you give me some examples. I get mutations, when the genes change. I also get modified expression of genes via epigenetics. What else?


The first V x K cross, ‘Rosseyanka’, was obtained via some embryo manipulation. After that, AFAIK, all crosses have been done via cross pollination. ‘Nikitskaya Bordovaya’ is an open-pollinated seedling of ‘Rosseyanka’, GRK is an open-pollinated seedling of ‘Nikitskaya Bordovaya’, etc. ‘Nikitskaya Bordovaya’ was selected at the Nikitsky Botanical Garden in Crimea, where they have a large collection of Kaki, including PCNA. One source I read says that Jiro has been used quite a lot in the breeding program, but does not specify its off-springs. So, it’s not impossible for GRK to have some PCNA heritage via its seed parent, ‘Nikitskaya Bordovaya’. We don’t know the pollen parent of GRK but it could be a seedling of ‘Rosseyanka’ or another seedling of ‘Nikitskaya Bordovaya’, or some later-generation cross. My understanding is that breeders in Ukraine did a lot of seed swaps, and some seeds could come from the Nikitsky Botanical Garden where they have a collection of Kaki or from some other breeders that grew Kaki in a protected environment. Hence, we cannot exclude some degree of PCNA parentage in GRK.


Chromosomes in plants are not just a blueprint, nor are all segments in a chromosome independent from each other. Under an electron microscope, a single diploid chromosome appears as a rope with many coils - so from one end you see a short section of “rope” entering a “coil”, then exiting and entering another coil, etc. These coils are the subject of what is termed “the protein folding problem”. We know the coils provide compact storage of the otherwise meter(s) long “rope”. Some have been determined to also act as factories and/or regulators. They might all do this - we don’t know. But within these coils, genes and other functional segments have the opportunity to interact. Note that these coils are relatively loose - enzymes and proteins essential to cell operations (e.g. RNA) are free to exit and enter. Also worth noting: these coils are not rigid. Although they retain their basic configuration - they are known to undulate on very short time scales. There is some evidence that chromosomes reconfigure a few coils, changing the total number. At present it is not known why or how often this occurs. There is of course abundant speculation. Too many biologists have a tendency to put the cart before the horse.

Yet another fact to keep in mind is that chromosomes produce inexact copies of RNA (and themselves) on occasion. For example, a functional segment of 100 base pairs will sometimes be reproduced at a length of (e.g.) 112 base pairs, with some extra non-functional spacers inserted. And likewise, truncation or deletion of an interior section will occur to produce a length of (e.g.) 96 base pairs. In addition, the reproduction might be the same length but not contain exactly the same bases, e.g. 4 base pairs are different. All of this could be swept under the rug of “mutation” but it actually occurs often enough to be considered normal and usually has no known effect on plant functionality or morphology – until bam! – an orange tree sprouts a branch with seedless fruit, or a cross between two known apricot cultivars produces a genetic dwarf with less than 1" spacing between nodes (without use of “catalysts” and without any ancestral history of dwarfing).

Now take the complexity I’ve described above and consider persimmon genetics, and additionally some members of the mulberry family etc. with 2n = 4, 8, … 306. Some of these are not at all simple with 3 bundles of 2 chromosomes wrapped on each other. We don’t know if or how the coils in adjacent chromosomes interact. So in my opinion it is too early to speculate on what “gene” or functional unit in one or more chromosomes is responsible for an observed morphologic trait.


<< The first V x K cross, ‘Rosseyanka’, was obtained via some embryo manipulation. After that, AFAIK, all crosses have been done via cross pollination. >>

Stan – Thanks. That makes sense. It would parallel the experience with JT-02.

1 Like

Yeah, like I said, I get mutation. The fact that it’s common and normal and complicated doesn’t mean it’s not genetic or heritable. Like, you described one example as a “genetic dwarf.” The dwarfing is inherited.

It was not. It occurred during procreation.

I have changed modulate to undulate, and put catalysts in quotes for clarity

True. However: