Hand pollination of Asclepiadaceae

Friederike Hübner

Asclepiads are interesting and sometimes bizarre plants, and to me they are also beautiful. Compared with my husband’s cacti, however, they have one disadvantage: it is not possible simply to use a small brush to obtain deliberately pure seed – ideally even from plants of the same locality. One aim of enthusiasts should therefore be to conserve species in collections by propagating them from cuttings and from seed produced in cultivation. Unfortunately, the appropriate pollinators are absent in our climate.

In spring 2002 I therefore began, with the aid of a stereomicroscope at 16–32× magnification, to study the structure of different asclepiad flowers, the “harvesting” of pollinaria and the technique of hand pollination. My first attempts using forceps and various loops failed. Forceps are usually too large for the tiny guide rails, while loops and clamps require more patience than I possess – even a good red wine did nothing to improve that. The remaining option was to glue the translator to a needle. Contrary to expectation this works well, provided that the adhesive does not touch the pollinia. In most cases a glued pollinarium can even be used twice: after insertion and pulling through the guide rail, a correctly seated pollinium breaks off at a predetermined weak point directly at the pollinium body.

An important point in pollination trials is that asclepiad genera often produce few or no seeds with their own pollen. They are self-incompatible. All my attempts at self-pollination failed except one in Stapelia longipedicellata. Three pollinations produced one pair of follicles and seed could be harvested. So far, however, the seeds sown successively have not germinated. They may have only very limited viability.

Table 1. Summary of the pollination trials carried out to date (total period: one year).

Table 1 summarises the trials and their success rates, from the rather discouraging beginnings to the present. In total, 26 species from eight genera were pollinated.

The genera from the perspective of pollination

Huernia has the simplest flowers. Their structure is very clear, and both the guide rails and the pollinaria are relatively large. In most cases, however, about half of the corolla has to be removed, most conveniently with a razor blade.

Stapelia has relatively large flowers, but the guide rail is hidden within the funnel-shaped structures of the corona. In addition, nectar or a nectar-like substance is often present at the bottom of the funnel and makes insertion more difficult.

The same applies to Caralluma. I was able to practise a great deal on this genus, but only rarely had two clones of the same species – and they never flowered at the same time. I am currently raising Caralluma socotrana deliberately and hope that two grafted seedlings will flower together this summer.

Pollinating Hoodia requires a steady hand and a good stereomicroscope. The effort is certainly worthwhile because both seed and plants are not commonly offered. So far I have successfully pollinated Hoodia officinalis and H. pedicellata.

Larryleachia is very small and correspondingly difficult to handle. The decisive factor is the angle at which the pollinium must meet the guide rail. If the angle is correct, insertion is quick. Unfortunately, several pollinated flowers died after four to eight weeks, and one plant carrying two developing fruits became infected and rotted within two days. This reduced the success rate considerably.

Orbea and Piaranthus again have comparatively clear floral structures, and pollination worked well in both genera.

Ceropegia and Brachystelma are the most problematic genera because all the species have very small gynostegia and even smaller pollinaria. The pollinia are almost spherical. A minute triangular part has to be inserted into the guide rail, but this is extremely difficult because the spherical pollinia obscure the view.

Special aspects of hand pollination in asclepiads

Each genus requires a different angle between pollinium and guide rail. This applies both to the angle seen vertically from above and to the angle seen from the side.

Simply inserting a pollinium does not mean that pollination has succeeded. Only when the translator breaks away is the pollinium correctly seated in the rail.

Even correct insertion is no guarantee of visible success. When growth ceases, plants tend to abort incompletely developed “fruits” – in fact slightly swollen flower stalks that remain persistently attached to the plant. This was particularly noticeable in pollinations carried out from August onwards. The few successful spring pollinations resumed growth rapidly and developed fruits; examples include Stapelia divaricata, Stapelia longipedicellata and Hoodia pedicellata.

Avoiding self-pollination by using two different clones produces distinctly higher success rates. Whenever possible, material from the same geographical locality should be used.

The longer the interval between pollination and the next growth phase, the more developing fruits are lost.

Nevertheless, during the last two weeks Orbea variegata, Huernia laevis, Huernia keniensis var. keniensis, Stapelia gettleffii and Larryleachia cactiforme have pushed their follicles out from the swollen flower stalks. Only Piaranthus framesii, Larryleachia marlothii, Hoodia pedicellata and Hoodia officinalis are still pending. These plants, however, also show no clear signs of renewed growth and have not yet ended their dormant period.

Spontaneous fruit formation

Last autumn again brought spontaneous fruit set that had not resulted from my hand pollination. Fruits developed on Ceropegia africana, Ceropegia woodii, Ceropegia stapeliiformis subsp. stapeliiformis and Ceropegia stapeliiformis subsp. serpentina – which may have pollinated one another – as well as on Stapelia gigantea. Examination of Ceropegia flowers, however, showed a much higher proportion of inserted pollinia than the number of fruits produced. Self-incompatibility may therefore also play a major role here.

← Back to article overview