The Tropical Fruit Forum

Used ChatGPT to look into cultivation of a citrus variety that is obscure outside of Japan. Told it to search Japanese language web pages. Yielded much more informative results than my own searches in English.

Propagation Methods of 黄金柑 (Ōgonkan) – A Review of Japanese Sources

黄金柑 (Ōgonkan, also called Golden Orange or Ki-mikan) is a Japanese citrus variety known for its bright yellow rind and sweet, aromatic flesh. Japanese-language sources – including academic insights, industry publications, and government reports – consistently emphasize that citrus like Ōgonkan are primarily propagated by grafting onto hardy rootstocks, rather than by seed. Other propagation methods such as cuttings and air layering are possible but less common. Below is a summary of key findings on Ōgonkan propagation methods from Japanese sources, with a focus on seeds vs. grafting vs. other vegetative techniques.

Seed Propagation (実生繁殖)
   •   Seeds are present but not the preferred method: Ōgonkan fruits contain a few seeds (typically 3–4 per fruit). While these seeds can germinate and grow into new trees, commercial growers in Japan rarely use seed propagation for citrus. The Ehime Prefectural citrus Q&A explains that raising citrus from seed takes a long time to reach fruiting and runs the risk of yielding a plant with different traits than the parent . In other words, seedlings have a long juvenile period and may not come “true to type”, so they are generally avoided in favor of faster, more reliable methods.
   •   Polyembryony and nucellar seedlings: Like many citrus, Ōgonkan is polyembryonic, meaning a single seed can produce multiple embryos. Often, one embryo is sexual (zygotic) and the rest are nucellar embryos – clonal seedlings of the mother plant . Japanese sources note that polyembryonic citrus seeds tend to sprout many nucellar seedlings, which are genetically identical to the seed parent . This phenomenon has practical implications: even if grown from seed, Ōgonkan seedlings might be clones of the mother tree rather than new hybrids. For example, Kanagawa’s development of the hybrid Shōnan Gold suggests that one of the seedlings from an Ōgonkan × unshū mandarin cross was actually a nucellar (clone) Ōgonkan seedling. While breeders sometimes exploit nucellar seedlings for selecting variations, farmers typically do not propagate orchard trees from seeds due to the long wait for maturity and the unpredictability of getting a true hybrid or a clone .
   •   Use in breeding vs. commercial planting: In summary, seed propagation of Ōgonkan is mainly of interest for breeding or hobby growing, not for commercial orchard establishment. A hobbyist report notes that planting Ōgonkan seeds is feasible (with germination in a few weeks), but it took about 7 years for a loquat grown from seed to fruit, implying a similar wait for citrus. Commercial growers avoid this delay.

Grafting (接ぎ木)
   •   Dominant propagation method: All reviewed sources agree that grafting is the standard propagation method for Ōgonkan and other citrus in Japan . Grafting involves taking a bud or scion (枝接ぎ穂) from an existing Ōgonkan tree and uniting it with a separate rootstock plant. The rootstock is usually trifoliate orange (Poncirus trifoliata, カラタチ) or related hardy citrus, which confers vigor and disease resistance . The Ehime Prefecture’s guidance explicitly states that citrus propagation “usually uses grafting” and that seed propagation (実生) is “normally not used” due to the issues mentioned above .
   •   Advantages of grafting: Grafted trees have several benefits. The rootstock’s established root system can more efficiently uptake water and nutrients, leading to faster growth and earlier fruiting compared to seedlings . Grafting onto dwarfing rootstocks like trifoliate also keeps the tree smaller and more manageable, and ensures the fruit quality remains true to the scion variety. Japanese experts note that if citrus were grown on their own roots (as in cuttings or seedling trees), they often become too vigorous and take longer to bear fruit. By using grafting, growers can have Ōgonkan trees that flower and fruit in just 2–3 years instead of 7+ years, with predictable varietal traits.
   •   Industry practice: Nursery catalogs and industry publications consistently list Ōgonkan as available only as grafted nursery stock (接木苗). For instance, one nursery’s online catalog labels its Ōgonkan saplings as “1-year grafted seedlings” and notes explicitly, “苗木は接木で増やします” – nursery trees are propagated by grafting. Another fruit-tree nursery explains that virtually all citrus for sale are grafted onto rootstocks, often delivered as budded or grafted one- or two-year-old trees . This is the de facto standard in the industry.
   •   Grafting techniques: In Japan, common grafting methods for citrus include whip grafting and budding (芽接ぎ) in spring. Some prefectural agricultural centers have published guides and even videos on proper citrus grafting techniques . Grafting is typically done in late spring (around April, when sap flow is active) to ensure good cambium contact and healing . The success rate is high when done properly, and once the scion bud takes, the young Ōgonkan tree will essentially behave like a mature tree branch on a young root system, blooming and setting fruit much sooner than a seedling.
   •   Summary: All evidence indicates grafting is the preferred and most widely used propagation method for Ōgonkan. This method guarantees genetic fidelity to the original variety and significantly reduces the time to fruiting . Japanese government and academic sources emphasize that grafting is almost indispensable for citrus cultivation at commercial scale.

Cuttings (挿し木)
   •   General feasibility: Propagating Ōgonkan by cuttings (rooting a piece of stem) is possible, but it is not commonly practiced. Citrus cuttings are known to be tricky – woody citrus cuttings often do not root easily under standard conditions. A gardening Q&A response on a Japanese site confirms this: “Lemon cuttings will root about half the time even if just stuck in soil and left alone, but other citrus hardly root at all – they usually just wither”. This reflects the experience that lemons are relatively easy to root, whereas mandarins (mikan) are much more difficult.
   •   Specific to Ōgonkan: Interestingly, one respondent on a Japanese forum speculated that “Ōgonkan has a high possibility of success” with cuttings. This anecdotal claim suggests that Ōgonkan might root better than Satsuma mandarins (温州みかん), perhaps due to varietal differences. However, this is not a formally documented result – it’s an opinion from an experienced grower, not a controlled study. So while a few growers have reported or hypothesized better rooting with Ōgonkan, it remains anecdotal evidence.
   •   Challenges and professional advice: A more detailed expert explanation (chiming in on another Q&A) provides insight into why cuttings are generally avoided for citrus in Japan. The expert notes that under controlled conditions (proper temperature, humidity, bottom heat, etc.), citrus cuttings can be made to root. However, the key reasons cuttings are not the norm are: (1) Rootstock advantages – using a strong rootstock via grafting helps dwarf the tree and induce early fruiting, which a cutting on its own roots would lack; and (2) Vigor control – a tree grown from a cutting will express the variety’s full vigor (which in many citrus can lead to a large, unwieldy tree that is slow to fruit). In short, “Yes, you can do cuttings… but normally one grafts onto rootstock”, because it’s much more practical.
   •   Success factors for cuttings: For those who do attempt citrus cuttings (often hobbyists or researchers), timing and technique are important. Some Japanese horticultural guides suggest that late spring (March–April) or late summer (August) are suitable times to take citrus cuttings when conditions are warm and humid enough to encourage rooting. Using rooting hormone (発根剤) and maintaining high humidity (for example, by covering the cutting or using mist) can improve success . Even so, the expected rooting period can be several weeks to months, and success rates are variable. There is little specific published data on Ōgonkan cutting success rates, so this remains a marginal propagation method.
   •   Bottom line: Cuttings are not a standard propagation method for Ōgonkan in Japan, though they are biologically possible. Commercial nurseries do not propagate this way because grafting is more efficient. Cuttings might be used on a small scale by enthusiasts or for experimental purposes, but Japanese industry literature overwhelmingly favors grafting over cuttings for citrus.

Other Vegetative Methods (Layering, Tissue Culture)
   •   Air layering (取り木): Another vegetative technique is air layering, where a branch is induced to form roots while still attached to the mother tree, and then severed to become a new plant. In general, air layering is considered difficult for citrus, but it can be done with careful technique. A Q&A on a Japanese site details a gardener’s attempt to air-layer a mikan branch: after a few months it had formed a mass of callus but no roots, indicating partial success . The expert advice was to peel a wider ring of bark and use sphagnum moss without disturbance, to encourage root formation . This suggests that with proper method (ring-barking the branch, applying rooting hormone, keeping the moss moist), citrus branches can be induced to root. The Tokyo Metropolitan Agriculture center has even published a video demonstrating citrus air layering techniques . However, similar to cuttings, air layering is rarely used in commercial propagation of Ōgonkan. It’s more labor-intensive and not guaranteed to succeed, so it remains a niche method for specialty propagation or salvaging branches.
   •   Tissue culture and micrografting: In academic contexts, plant tissue culture has been explored for many crops, including citrus, to produce virus-free clones or mass-produce rootstock seedlings. In Japan, citrus micropropagation is not widely reported in public sources for this specific variety, but research stations have used in-vitro techniques and nucellar embryo culture for breeding disease-free lines. One example is the use of micro-grafting (grafting a tiny bud onto a seedling in vitro) to eliminate pathogens while propagating citrus clones – a technique used by some Japanese citrus research labs (though specific published reports on Ōgonkan were not found in this review). For the scope of practical propagation, tissue culture remains a scientific tool rather than a routine method.
   •   Budding: It’s worth noting that “grafting” in nursery practice often means bud grafting (芽接ぎ) – this is essentially the same outcome (joining Ōgonkan wood to rootstock) but done by inserting a single bud rather than a scion stick. Japanese nurseries frequently bud-graft citrus in summer, then grow the plant for a year. Budding is encompassed under grafting methods and is the technique by which most Ōgonkan trees are produced.

Conclusion

Japanese sources unanimously highlight grafting as the primary propagation method for Ōgonkan, due to its efficiency and reliability. Government agricultural extensions explicitly advise using grafting for citrus and caution against seed propagation except for breeding . Industry publications (nurseries and growers) reinforce that all commercial Ōgonkan saplings are grafted. Academic knowledge of citrus physiology (e.g. polyembryonic seeds) explains why seedling populations are avoided in cultivation despite the interesting trait of nucellar embryony that can produce clones. Cuttings and air layering are acknowledged but infrequently used – they are possible routes that some enthusiasts may try, yet they play a negligible role in mainstream Ōgonkan production due to lower success and agronomic drawbacks.

In summary, Ōgonkan is propagated almost exclusively by grafting in Japan, with seeds used only in special cases (such as breeding programs) and vegetative methods like cuttings or layering remaining experimental. This ensures that orchards of Ōgonkan maintain true-to-type fruit quality and bear fruit as early as possible, following the long-established practices in Japanese citrus cultivation .

Sources:
   •   愛媛県庁 かんきつ類Q＆A (栽培編) – Explanation of citrus propagation via grafting vs. seed  .
   •   花ひろばオンライン苗木部 – Ōgonkan product page (nursery stock description).
   •   富久寿園 柑橘辞典 – Ōgonkan entry (notes on Shōnan Gold nucellar seedling).
   •   Yahoo知恵袋 (Japan) – Citrus propagation questions (expert answers on cuttings and layering) .
   •   果物ナビ – General information on Ōgonkan fruit characteristics (for context on seeds).

Tried this at Frankie's today. It's incredibly hard to describe.

The texture is like an *extremely* overripe mango. The flavor and aroma are both very similar to durian, but significantly milder. Durian is too intense for me but I've enjoyed things like durian-flavored macarons; it was like that. The difference between durian and this fruit is like the difference between biting into a lemon and drinking lemonade, or the difference between biting a garlic clove and eating something flavored with garlic. This is not to say cempedak tastes like lemons or garlic, rather just an analogy regarding intensity.

In addition to the durian flavor, there is almost a nuttiness. At times I tasted something that reminded me of Thai basil. And there is also a sort of acidity reminiscent of a slightly underripe grocery store mango. All these background complexities are hard to describe. But overall if you say 'extremely overripe mango texture, with the milder durian aroma and flavor you'd get in a durian flavored dessert, quite sweet, moderately acidic (*just* acidic enough to balance out the sweetness)' you are about 80% of the way there in terms of explaining it.

It wasn't my favorite fruit ever, but it was without a doubt the most fascinating fruit I've ever tried, because of how challenging it is to describe.

20137 Pioneer Blvd
Lakewood, CA 90715

I remember when these first hit the market they were enormous and always very high quality. Now they are the size of regular oranges (if not smaller) and very hit-or-miss in terms of sugar and acid balance.

Not asking to complain; rather, it seems like home growers could learn from this (i.e. mimic the growing practices that made the original fruit so good, and avoid whatever changes that have caused the quality to go down)

I visited a fun place today in Oceanside (south of Los Angeles, north of San Diego). They have a bunch of guava trees and let you pick your own, then they weigh it out and you pay. There's many trees and a lot of variation between them (two of the varieties, "Tropic Pink" and a strawberry scented Psidium guajava from Cal Poly, seem to be seed-grown).

Would really recommend anyone here who likes tropical fruit to check it out. Not associated with this place at all, just think it's a hidden gem.

Some photos attached from the visit and of the guavas I picked.














Here's their website: https://www.osidesullacollina.com/

There is something very entertaining called the "Dwarf Tomato Project." It has produced quality tomato varieties that grow less than one foot tall.

Some info on it (from this source: https://www.craiglehoullier.com/dwarf-tomato-breeding-project)

Back in 2005, Craig LeHoullier of Raleigh, NC, long time heirloom tomato enthusiast and SSE member, discussed an idea with Patrina Nuske Small of Australia on the popular garden site Garden Web.  Craig and his wife, Susan, have been selling tomato seedlings at the local farmer’s market for years, and a frequent request was for delicious, unusual tomato varieties that would do well when grown on decks or patios, in reasonably sized containers.  Though his customers loved the big fruited, differently colored heirlooms, the tall vines proved to be quite a challenge for many of them. 

Craig, through his seed collecting, was aware of a very few so-called dwarf growing varieties, distinct from the determinate types (which were also relatively compact) in that the flavor seemed in general to be better in the dwarfs.  These dwarf types are very distinctive in having a thick central stem, stout compact growth, and dark green, crinkly looking (so-called rugose) foliage. 

Craig also collects old seed catalogs to satisfy his interest in the historical development of tomatoes in the United States, and noted in his collection a 1915 catalog from the Isbell Seed Company a listing for New Big Dwarf.  What caught Craig’s eye was that the creation of New Big Dwarf was clearly described.  This involved a cross of a known medium sized pink fruited dwarf, Dwarf Champion (well known from the late 1800s), with the largest tomato known at that time, Ponderosa.  Following the cross and a few generations of selection, a new variety was born – New Big Dwarf – which had the compact growth of Dwarf Champion, but the large fruit size of Ponderosa.

Realizing that through their Garden Web chats Patrina (also an avid heirloom tomato grower) was also skillful at carrying out crosses, Craig explained the Isbell Seed Company listing to Patrina.  It was clear that the approach Isbell used was successful, yet strangely never further explored to create a range of new dwarf growing varieties with large fruit, in different colors.  So Craig and Patrina described a project in which some strategic crosses would be made, the resulting new hybrids grown out, then seeds shared for work on selection of promising new varieties, followed by several generations to reach a stable new non-hybrid variety with the desired characteristics.

Around that same time, a new tomato chat website emerged, Tomatoville, which provided both a structure for tracking the project, as well as a magnet for the kind of extreme tomato enthusiasts who could make up a perfect volunteer team to take this new project forward.  Craig set about to gather volunteers in the US and Canada, and Patrina in Australia, Tasmania and New Zealand.  By working the project across hemispheres, two growing seasons could be achieved in one calendar year, thus reducing the new tomato variety development time by half.

So, armed with more than 100 willing volunteers, the project set off with the first crosses being grown out in 2006.  As of this time, over 40 interesting crosses have been made by Patrina and others, resulting in dozens and dozens of possible color, size and shape combinations.  Craig and Patrina are quite sure that this is the very first all volunteer world-wide tomato breeding project in documented gardening history.  None involved are botanists or horticulturists – just avid gardeners with a keen interest in learning about tomato genetics or discovering interesting new tomatoes. 

Craig selected four seed companies (Tomato Growers Supply Company, Southern Exposure Seed Exchange, Sandhill Preservation, and Victory Seeds) to introduce small quantities of a few of the new dwarf varieties each. Over the years, other companies have joined, including Gleckers, The Sample Seed Shop, Casey's Heirlooms, Tatiana's TOMATObase, Victory Seeds and Heritage Tomato Seeds.

None of those involved with the project receive any payment for the development of these new varieties – it is just the fulfillment of a passion to create what will hopefully end up being considered heirloom tomato varieties many years from now – a donation to the tomato gene pool, and to those space-challenged gardeners the world over.  Full descriptions of the varieties used and volunteers involved in the development will be given for each new variety. 

There are at least two dwarf Psidium guajava varieties in the United States. I know because I have both.

This one (tiny plant, tiny leaves): https://tom-piergrossi.squarespace.com/plants-m-r/psidium-guajava-compact-form found in the wild in Hawaii, but similar to another clone being passed around in Asia

This one (tiny plant, normal leaves): https://www.logees.com/dwarf-guava-psidium-guajava-nana.html that was found being cultivated in Hawaii but has an original origin unknown to me

I think it could be very interesting to take pollen from these and use them to fertilize regular, high-quality guava varieties. Especially because a lot of people who are into tropical plants live in temperate climates and like to grow stuff in pots.

By using the dwarf variety as the pollen parent, and the full-sized variety as the seed parent, you would know pretty quickly if you were successful (versus having accidentally selfed the seed parent) by looking at the growth habit of the resulting seedlings.

Right now the only Psidium guajava I have are the two dwarf ones, but in March I'm going to try to get a really good variety, 'Ramani,' from Frankie's in Hawaii. If I'm successful (i.e. if Frankie's has this variety in stock, if I clean its roots well enough to pass inspection, and if it survives the trip home), in time I will start pollinating some of its flowers with pollen from the dwarf guavas and distribute any resulting seeds to interested forum members who have space to grow them out. This particular variety, 'Ramani' from Frankie's, is basically seedless (soft seeds), so even just the production of seeds should be a good hint that the cross worked. It will then just be a matter of seeing if the dwarfing aspect of the pollen parent, and the fruit quality of the seed parent, was passed on to the offspring.

Unlike the 'Dwarf Tomato Project' we wouldn't need to stabilize the resulting hybrids. We would just need one good one and then it could be distributed via air layering and grafting.

It will be a while until this actually happens, if it even does, but I thought it'd be fun to make the first post now.

In a parking lot in East Los Angeles. Fruits are large (my hands are big which makes it look a bit small in the picture). Tree is tall (12-15 feet). Tried a fruit that had fallen to the ground (the dirt at the base of the tree; the ones that had fallen to the pavement were smashed). No insect or animal damage to it. Had hard seeds (doesn't bother me), was very soft, was sweet and creamy. It's the middle of winter; notable to me that it ripened sweet in cold weather. Maybe being surrounded by concrete has something to do with it.

I feel like if you grew this one you could just set up a net under the canopy for the ripe fruit to fall into

Was over there the other day. Big Jim loquat, Tropic Pink guava, seed-grown Manila mango roughly 5-6 feet tall (good for top working), some interested grafted mango varieties (one that caught my eye was "ST Maui"), starfruit, pitangatuba, etc. Notably a lot of their plants have fruit on them which I thought was neat. Worth checking out.

There's also two "Mimosa" nurseries in Los Angeles County but I think they're under different ownership.

Saw this on Facebook

Found this, useful https://www.ars.usda.gov/pacific-west-area/hilo-hi/daniel-k-inouye-us-pacific-basin-agricultural-research-center/tropical-plant-genetic-resources-and-disease-research/docs/guava-collection/

You can click on the name to see photos

The website also has similar collections for other fruits

Hi everyone!

I'm looking to make a multi-variety tree. It would be in-ground. I think I would ultimately want it to have Meyer lemon, Valencia orange, New Zealand Lemonade, and a number of Asian citrus varieties (yuzu, kishu, etc).

I've been looking at published reports about rootstock performance but would also really appreciate any insight from members here about which rootstock might be a good choice.

I live in Los Angeles County (summer days can sometimes hit 100+ F, freezes are unheard of). Having a long-lived tree is important to me (for example, I don't want to have to inarch yuzu or something like Japanese growers have to once their plants start to decline on trifoliate).

Does anyone have any suggestions? Or are there any rootstocks I should avoid?

Link (captions available in English):

https://www.youtube.com/watch?v=JvvtFmVEn2M

Description (translated to English):

José Aurélio Vilela can be considered the precursor to the "Jabuticabas Collectors".

Despite being widely cultivated and used in food by indigenous peoples since before the arrival of the first Europeans, and having soon attracted attention and recognized as the best native fruit of Brazil by the Portuguese, jabuticaba still has its immense genetic variability unknown to the majority of the population.

In March 1995, so almost 30 years ago, a 2-page report with colour photos in Globo Rural Magazine revealed to the country the work started by Vilela in the late 1960s, of selecting and gathering more than 40 varieties of jabuticaba in an orchard in your farm in the southern state of Mines Gerais.

The strong impact caused by the matter and mainly by the photos showing some of these varieties has inspired nearly a dozen new "Jabuticabas Collectors," including the author of this post.

Today the J name tho Aurelio Vilela is still almost unknown to the general public, and even among the most fans any reference to the "Jabuticabeiras Collection" still involves a deep aura of mystery.

The "E-GARDEN ON TOUR" program, which PREMIERES TODAY on YouTube, kicks off a series of episodes telling all about the Vilela Collection and showing in detail some of its best Jabuticabas!

They aren't shipping them; you have to buy them in-person. They're having people sign a waiver agreeing that they won't grow them commercially. Hopefully this means the variety starts getting passed around.

More information on their Instagram: https://www.instagram.com/frankiesnursery

To those who don't know: the Meli Kalima variety was bred by Frankie's Nursery in Hawaii. They were sold at local markets and somewhat legendary among those who managed to try them. Although the fruit is very small, it is intensely sweet. Frankie's recently ended their production of the variety and it seemed they were looking for people to buy the patent.