Seed Production in Petrocosmea

As I continue to harvest that last seedpods from this winter's Petrocosmea hybridizing adventures, I find myself, one again, feeling excited and grateful that I've had success in producing seeds from my favorite genus of gesneriads. For years, around ten to be exact, I consistently failed to produce even a single seed..or a seedpod either for that matter. But, at least I was consistent! One of the most requested tidbits of information I receive is for advice on how to produce seeds in Petrocosmea. I thought I would outline my process here on the blog.

First, let me say that this is what works for ME, in MY environment, with MY plants. Other growers, in other conditions, may have very different results. I don't suggest that I am an expert, only that I've been lucky for a short time now, and this is what worked for me. While I have shared this with other accomplished growers in the recent past, I have heard from only one person who was successful in producing seeds. There are others, who've had success, of course. Other well before I had luck. Mr. Nakayama, in Japan, in fact, produced the first Petrocosmea hybrid around a decade ago... P. 'Momo'...the result of a cross between P. nervosa and P. flaccida. He has since remade the same cross and produced two selections..P. 'Asa Blue' and P. 'Imperial Butterflies'. Mrs. Joyce Stork, here in the US, produced introduced two hybrids a couple of years ago, P. 'Fluffer Nutter' and P. 'Short'nin' Bread' both from a cross between P. forrestii and P. flaccida. And Mr. Jeff Foerderer of the Netherlands, had produced a number of crosses recently. He registered two of them with the Gesneriad Society....making these the first registered Petrocosmea crosses. Those were P. 'Lexi' and P. 'Milan'. So we are beginning to see some success with hybridizing Petrocosmea. To date, I have been successful in producing seeds from a total of 22 seperate hybrid crosses over the past three years. Most of those, 14 of them, produced during this past winter.

So what has worked for me? Well, the key factors for me, which I have tested and proven during the past two winters, have been a combination of high humidity during flower production and seedpod maturity, and cool to cold temperatures. If either of these factors was changed or missing, I have had total failure in the crosses. Furthermore, I must say that those same two factors have been proven critical to successful seed germination as well. But first, lets get the seeds and then we can discuss how to germinate them.

So here are the steps I follow to produce seeds:

1. The seedpod parent must be flowering in high humidity...100% if possible. As plants come into bud, I enclose the whole plant inside humidity domes over the nursery trays on my lightstand shelves. Also the area should be cool...around 40 degrees F. A bit cooler is fine...I've had successful pod formation in 38 degrees F and as warm as 50 degrees, but no warmer.

2. Select flowers for pollination that are at least four days old. Older is fine as long as the pistil and stigma are still in good condition. Around the time the flower is about to drop seems the best, but the stigma is usually not receptive until around four days.

3. Select pollen from the freshest flowers you can find...freshly opened up to three or four days seems optimal. Pollen should look "fresh and fluffy" dry dusty pollen does not work for me. Frozen or refrigerated pollen has also failed to work for me too. I use a simple process of collecting the pollen on my thumbnail from the anthers so that I can see it easily. "Pack" as much pollen on each stigma as you can get it to accept.


This photo, taken by my friend and master hybridizer Dale Martens, shows the reproductive structures of a Petrocosmea flower. Note that the stigma at the tip of the pistil is purple. The filaments supporting the anthers have furry white hairs.

4. Label each flower pollinated. I use a small square of paper with a hole in the center and a small slit on one side to allow it to be slipped onto each flower peduncle. I also record the parentage and date of the cross in a manual which I keep all of my hybridizing records in. This becomes a most valuable resource later on.

This photo shows seedpods forming on P. forrestii. Note that after successful pollination, the tip of the pistil dries to just below the stigma. the ovary swells quickly and pollination will be evident in around 8-10 days after making the cross if it has been successful. These seedpods are about one month out from pollination.

5. Wait. This is the hardest part. Ripening for me has taken 60-80 days with around 64 days being the most common. I have had successful germination from seeds taken from a pod that was only 49 days old, when the mother plant began to deteriorate and I was afraid of loosing the cross. Generally, though, I wait until the seedpod browns and is just beginning to split open. Seedpods will split from the base upwards along two sides of the pod. The seedpods appear to have two chambers.

Another excellent photo taken and shared with me by my friend Dale Martens, shows Petrocosmea seeds. Viable seeds appear plump while non-viable seeds appear more linear and smaller.

6. Once the seeds are ripe and harvested, I sow them immediately. I only sow a small amount of the seeds and freeze the rest. I sow seeds on moist, fine vermiculite or moist peat moss pellets. I enclose the seeds once planted and keep them moist with high humidity. In the same temperatures used for seed production...around 40 degrees F, I get germination in 14-28 days from fresh seeds. Once seeds are a bit older, I find germination takes longer, and it has taken as long as 3 months on older seed. Seeds sown in warmer temperatures have consistently failed to germinate for me and several other growers with whom I've shared seeds for testing.

Once seeds germinate, I begin to fertilize them with a weak solution of water soluble fertilizer at a rate of 1/8 - 1/16 teaspoon per gallon of water. I keep them under lights in the same conditions the mature plants grow in.

Once the seedlings are large enough, I transplant them into community pots and treat them just like the mature Petrocosmeas in every way. Of course, labeling and careful record keeping are maintained at every step of the way. The photo above shows seedlings at about three months from germination.

Again, cool temperatures and high humidity are what have made the critical differences for me. I hear often from growers who have tried repeatedly to cross Petrocosmea and failed. When I ask about the conditions in which they attempted the crosses, one or both of these factors always seems to be missing.

This year, I tested this several times by attempting crosses in lower humidity and in warmer conditions. I had NO success. Only when both factors were present did the crosses take. Even then, with the two critical factors in place, my success rate with total attempts remains under 50%. I make sure to pollinate around ten to twelve flowers each time I make a cross. Sometimes I get as many as 7 seedpods, other times, I may get only one. And of course, sometimes I get none. But, the challenges are what makes this so much fun and so rewarding once those seedlings finally do bloom. This is part of what fuels my passion for these magnificent creations of nature! One thing that alway occurs 100% of the time though, is that I learn something every time I try a cross. And I have fun 100% of the time too!

A Silky Haired Trio - The Petrocosmea sericea Group - III

Should our trio become a quartet? In the last two posts, I have been discussing the three plants that I place with a group I call the "Sericea Group". This is my own placement and terminology around the three plants labeled P. sericea, P. sp. 'HT-2' and a newly collected plant labeled P. sp. ?'JR2008-1'. But, should this trio include one more plant? Study the photo below in comparison to these photos of the flowers on these three plants from my last post.

Very close isn't it? I think so. Morphologically, the flowers and foliage of the plant I acquired last year labeled P. cavaleriei are so very close. I have sent plant material to have it's DNA studied in relation to the DNA of the other three, but do not yet have results. Hopefully those will come soon. But, independent of the DNA data, I feel this plant belongs to the same group as the other three. It does not totally match the published description of P. cavaleriei, but then neither does P. sericea.....or most other Petrocosmeas which currently bear a name. Again, the identity crisis for this genus.
The flowers of P. cavaleriei closely match those of P. sericea and the others in the "Sericea Group". It clearly fits within the Section Anisochilus, as do the other three plants. Other than form of the petals, it is identical.

Petrocosmea cavaleriei pictured above, looks very much like P. sericea, and P. sp. 'HT-2'...especially the latter. It differs only in it's small stature. This plant stays small, after two years, this plant still remains in a three inch pot. Cymes are single flowered. I'm anxiously awaiting the DNA results to see how closely it is related to the others in this group.

I thought I would conclude by showing a bit of what P. sericea is capable of as a parent in hybridizing. My very first hybrid, pictured above, is the result of a cross using P. sericea as the pollen parent with P. rosettifolia #3. The hybrid was named for a dear friend, P. 'Rosemary Platz'. It inherited a heavier leaf substance, the leaf shape, and size from P. sericea. Flower size and coloration on this hybrid also show the influence of P. sericea. The serrated leaf edges, the glossy green leaf color with lighter yellowish veins clearly came from the P. rosettifolia parent.

A close up of the flower of P. 'Rosemary Platz' shows a nice lavendar color from P. sericea, while the absence of the dark purple blotch at the base of the tube, along with the two yellow stipes are from the P. rosettifolia parent. Flower size is similar to P. sericea.

A trio of first flowers from three seedlings resulting in my second P. sericea cross. This time P. sericea was the seed parent with pollen coming from P. minor veined leaf form. The variation in the purple color of the flowers shows the lightest being the color of P. sericea and the darkest coming from the deep purplle of P. minor. Two flowers have the white throat of P. minor, while the one on the left has no white in the throat...which is most like P. serice. All three inherited the deep purple blotch at the base of the corolla, which both parent have. Genetically, P. minor and P. sericea are closely related to each other. None of these seedlings have been named as yet, but the three pictured are held for another years growth to evaluate their perfomance as they mature. While P. sericea can be slow to mature, these three all bloomed as tiny plants in one ounce pots.... each seedling flowered with fewer than ten leaves.

A Silky Haired Trio - The Petrocosmea sericea Group - II

In the last installment, I discussed the foliar characteristics of the trio of plants I call the "Sericea Group" due to the fact that they all share very similar foliar and floral characteristics. In addition, recent DNA data shows all three plants to be closely related to each other, so close, in fact, that they may be simply three unique individuals within the same species.

In this post, I would like to focus on the characteristics of the floral morphology. I'll start with the first plant to come into culitivation... the plant we currently grow as P. sericea.


The flower of P. sericea shows characteristics which clearly place the plant within Section Anisochilus, having shorter upper petals, one half the length of the lower petals. The flowers typically have five petals, with the shorter upper two petals being fused into a two lobed "hood" which bend forward over the pistil. The anthers cup together to form a "beak" with the pollen ejecting from the pointed tip of the "beak" when pressure is applied to the wider base of the anthers. P. sericea's flowers are usually born singly on the cyme, with a nice bluish lavendar color. At the base of the tube inside, is a darker purple spot. The petals on P. sericea are a bit longer and narrower in the lower portion of the corolla.

The cymes of P. sp. 'HT-2' are often multi-flowered...having from 1 to five flowers...most commonly two or three flowers. The petals on the flowers of this form are shorter and rounder at the lower portion of the corolla, giving a "fuller, rounder" look to the flowers when compared to P. sericea's flowers. Coloration is identical on the two forms.

The lower three petals of the flowers of P. sp. 'HT-2' are rounder and shorter than those of P. sericea, giving the flowers a fuller appearance. In this photo, you can barely see the darker purple spot at the base of the corolla around the ovary of the pistil.

While my own plant of P. sp. ? 'JR2008-1' has not yet flowered, this photo shows the plant from which my plant was propogated. The calyx lobes on this form appear hairier and more silvery, when compared to P. sericea and P. sp. 'HT-2', but the flowers otherwise appear nearly identical. Like P. sp.' HT-2' this plant appears to have multiple flowers on each cyme. This photo also shows the more open form of the rosette on this form and also that the leaves are softer adn have a less succulent nature than the other two forms in this group. The leaves are more pliable and softer to the touch.

A close up shot of the cyme and flower of P. sp. ? 'JR2008-1' shows the longer, more silvery calyx and cymes on this form...an attractive feature in my opinion. I look forward to the flowering of my plant, which so far, has been slow to flower.

While the flowers are very similar, there are key differences which allow each form to be identified as unique in it's own right. Still, I feel these are all representative of the same species. Time will tell.

There is one last installment in this series....... Should Petrocosmea cavaleriei be included in the "Sericea Group"?

A Silky Haired Trio - The Petrocosmea sericea Group - I

In my continuing quest to know the genus Petrocosmea better, I've been focusing lately on a trio of plants that I call the "Sericea Group". Why do I call them this? Well, because these three plants are yet more of our wonderful Pet species which are currently victims of the identity crisis that plagues this genus. Only one currently has a species name attached and even that one may be incorrectly attached to this plant. That species, Petrocosmea sericea (pictured below) is the first of the group to come into cultivation in the USA, and this is the name applied to it from that time until today. So, I use that one,being the first and the only one with a species name, as the species around which I "lump" the other two plants...those being Petrocosmea sp. 'HT-2' and Petrocosmea sp. ? 'JR 2008-1'. These three plants are the focus of this short series.

So, what do I know about these three beautiful species? Well, the do all appear to be very closely related to each other....perhaps even three forms of the same species. The early DNA analysis does place them very closely related to each other, but three clear distinct individuals. In other word, they are not the same individuals...the DNA is a bit different, but they may be the same species. My personal opinion is that they are. Morphologically, they are very close also.

Secondly, they do appear to fit within the Section Anisochilus of the genus. Morpholically, the flower's anthers and corolla easily fit the defining characteristics to fit within that section. This section also includes the plants we currently grow as P. rosettifolia, and the P. minor (s). If you grow all of these, think about the flower structure and you can easily see that aside from size and color, the flowers are shaped almost identically in may respect. These all have the two upper petals that are shortened and bend forward, fused into a "hood" at the top of the flower. The flowers of the plants in the "Sericea group" all are nearly identical, with the only differences being that sp. 'HT-2' has a branced cyme with 1-5 flowers per cyme, while P. serciea most often has a single flower per cyme. I have not personally flowered the P. sp. ? 'JR 2008-1', but have seen photos of the flowers, which appear almost identical to the flowers of the other two species. All three plants have silvery hairs on the calyx lobes and cyme, making it appear silvery, frosted and silky.

Third, the leaves of all three species are similar, with the following differences. P. sericea has the most succulent leaves of the three plants, with the margins curled up forming a "spoon", which to my eye, is very attractive. The margins of the leaves of P. sp. 'HT-2' curl downward
in the opposite manner of the leaves of P. sericea, and are a bit smaller than the leaves of P. sericea, but are otherwise identical. The leaves of P. sp. ? 'JR2008-1' are the "thinnest" and softest of the three, and are clearly the largest of the group. They are softer to touch and more flexible. The "silkiest" if you will. This form also forms the most "open" rosette of the three plants. All three plants offset rarely for me, and tend to grow slowyly. The newest of the three.. sp. ? 'JR2008-1' grows the slowest of all. The plant pictured below is now nearly two years old and still is a small plant.

Petrocosmea sericea - the first of the group to come into cultivation around 2003. It is "close" to the published description of this species, although the published description mentions the hairs being golden or yellowish... the hairs on this plant are clearly silvery and "white". Otherwise the description is pretty close. I have to wonder if the description might have been written using dried or preserved plant material to describe, and perhaps the hairs turned more yellow or golden with age and drying? I have collected some leaves to press and dry in order to see if the colors of the hairs changes with age and drying.

The leaves of this plant cup upward and are the heaviest substance of the three...being quite succulent. This plant grows slowly, compared to all other species I grow, and makes a spectacular rosette as it ages. Two years, minimum, seems required to get a "mature" plant from a leaf propagation. The plant only gets better and flowers more heavily with age.
It also tolerates and seems to appreciate a bit of drying between watering and the leaves spoon more and look more silvery with higher light.


A comparison of two mature plants of P. sericea and P. sp. 'HT-2'...both of these plants are in 5" pan pots and are three years old when this photo was taken. They were grown side by side with identical culture to test the differences in the two species. Note that P. sericea, (left) is more silvery in appearance and the leaf margins cup upward forming a spoon. The leaves of P. sp. 'HT-2' (right) are greener in appearance and the margins are flat or curl downward. Leaf size is close, with the leaves of P. sp. 'HT-2' being a bit smaller on the average.

The newest species in the trio is one labeled P. sp. ? JR2008-1'. It came to me from a friend who got it from the collector in China in late 2008. This plant grows slowly. The plant in the photo above is two years old, and is still in a 3.5" pot. It has not yet flowered for me, but I have photos of the flowers which I will post in the next post on this group. The plant in all respects is very close to the other two morphologically. Leaves are much larger than the other two plants, even on this young plant. I currently have six smaller plants that were propped from this plant as "insurance" and to share with other Pet friends so that we can begin to get this form into broader cultivation. I am also sending a plant to a botanical garden as part of the collection there. It is always exciting to get new genetic material and new forms of Pet species in cultivation.

This photo shows the leaves of the three forms discussed in this post, for comparison. They are (Left to right): P. sericea (left), P. sp. 'HT-2' (middle) and P. sp. ? JR2008-1' (right).

These three plants add beautiful and interesting diversity to the family Petrocosmea in any collection. There is no other Pet that has the characteristics of these three. While P. nervosa and P. flaccida also have the silvery, silky hairs covering the leaves, the flowers of these are quite different. P. flaccida and P. nervosa have floral characteristics that place them within Section Petrocosmea of the family, while these plants clearly fit into Section Anisochilus. Additionally, the DNA data suggests that these three plants are very closely related to the group of P. minor(s). The"minors" are their closest relatives, according to the DNA. I have been successful in crossing P. sericea with P. minor veined leaf, when P. minor would not cross with other species. This willingness to hybridize with the minors suggests a close relationship also.

In the second installment on this topic, I will focus on the flowers of these three plants.