Skip to main content
You are here
Facebook icon
Pinterest icon
Twitter icon

Orchid seeds – Nature’s tiny treasures

Wolfgang Stuppy
28 November 2013
Blog team: 
Orchids have the smallest seeds in the world and they produce millions of them, but why? Kew's seed morphologist Wolfgang Stuppy explains the clever survival plan that lies behind this seemingly wasteful strategy.


A selection of images of beautiful orchids
Breathtakingly beautiful and vast in number - with more than 26,000 species, orchids are the largest family of flowering plants, not to mention the more than 100,000 hybrids humans have created.

The biggest seed in the world

Seeds come in all shapes and sizes. Famed for both its volume and suggestive shape, the seed (actually a single-seeded stone) of the Seychelles nut or double coconut (Lodoicea maldivica, Arecaceae) holds the unbeaten record for the world’s largest seed. It can weigh up to 18 kg and resembles something that, while bobbing in the waves of the Indian Ocean, gave sailors in the Middle Ages all kinds of, well, “seedy” ideas.


Keith Manger, Lab Manager at the Millennium Seed Bank, proudly demonstrating the physical attributes of a Seychelles nut
Keith Manger, Lab Manager at the Millennium Seed Bank, proudly demonstrating the physical attributes of a Seychelles nut

Seeds like dust

At the other extreme of the spectrum we find the seeds of orchids. Famed for their beautiful and fascinating flowers, with over 26,000 species worldwide, orchids are the largest of all flowering plant families. What’s more, they also hold the world record for having the smallest seeds of all flowering plants. A typical orchid seed is merely the size of a speck of dust. 

Flower of Stanhopea Assidensis [= S. tigrina x S. wardii] and seeds of Stanhopea tigrina (0.66 mm long); below: flower and seeds (0.6-0.8 mm long) of the common spotted orchid (Dactylorhiza fuchsii) [Images from SEEDS – Time Capsules of Life by Rob Kesseler & Wolfgang Stuppy; Copyright Papadakis Publisher, Newbury, UK] 


Top: seeds of the Wild Coco (Eulophia alta), on a British one penny coin. Below: two seeds of the same species in the SEM (the scale bar shows half a millimetre)

Tiny ones and not so tiny ones

To give an impression of the dimensions involved: a single capsule of the tropical American orchid Cycnoches chlorochilon produces almost four million seeds, and one gram of seeds of the southeast Asian species Aerides odorata contains 3.4 million seeds. At around 0.2 mm in length, Aerides odorata has the smallest seeds I have ever come across at Kew’s Millennium Seed Bank. According to the literature [Arditti, J. & Abdul Karim Abdul Ghani (2000) Numerical and physical properties of orchid seeds and their biological implications (Tansley Review No. 110). New Phytologist 145: 367-421], there are orchids with even smaller seeds. Those of the New Caledonian species Anoectochilus imitans are said to be the smallest of all, measuring just 0.05 mm in length. At a ‘gigantic’ 6 mm, the seeds of the lopsided star orchid (Epidendrum secundum) are allegedly the longest of any orchid. 


Top: two seeds of Acanthephippium splendidum measuring c. 3 mm in length. Below left: three seeds of Aerides odorata measuring c. 0.2 mm in length (the scale bar shows half a millimetre)

Left: flower of bee orchid (Ophrys apifera). Right: the seeds of early spider orchid (Ophrys sphegodes) measure a bit more than half a millimetre (c. 0.6 mm) [Images from The Bizarre and Incredible World of Plants by Wolfgang Stuppy, Rob Kesseler & Madeline Harley; Copyright Papadakis Publisher, Newbury, UK] 

Balloon travellers

The reduction in seed size and weight is mainly achieved at the expense of embryo and endosperm, the latter failing to develop in orchids. At the time of dispersal, orchid seeds consist of a spindle-shaped, wafer-thin seed coat that encloses an extremely small and simplified embryo in the shape of a spherical cluster of cells. Just one single cell layer thick, the seed coat (also called testa) forms a balloon around the embryo, a clear adaptation to wind dispersal.


A seed of Clowesia russelliana, showing the thin, transparent seed coat enclosing a tiny spherical embryo
A seed of Clowesia russelliana, an epiphytic orchid from tropical America, showing the thin, transparent seed coat enclosing a tiny spherical embryo.

With a little help from their friends

Because orchid seeds lack a food reserve in the form of an endosperm or a large embryo, most of them, especially terrestrial ones, are generally unable to germinate on their own. They first have to engage in a mycorrhizal relationship with a fungus that helps to feed the emerging seedling. Some orchids are able to join up with many different species of fungi whilst others only accept a very specific fungus to enter their lives (or rather roots). Few orchids don’t need any fungus at all for their germination, such as certain species of Disa from South Africa, a remarkable exception among terrestrial orchids. 

Seedlings of the neotropical orchid Encyclia chloroleuca growing in a Petri dish. Placed on a nutrient medium under sterile conditions most epiphytic orchids can germinate without their fungal partner. (Photos: Suzie Woods)

Why so many?

Their dependence on certain fungal partners is most probably the reason why orchids produce vast numbers of tiny seeds. With their small size, low weight and balloon-testa, orchid seeds are perfectly adapted to wind-dispersal. However, their strategy is not to travel long distances. Scattering large numbers of seeds with the wind merely heightens the chances that at least some end up in a place where they are lucky enough to meet their specific fungal partner without which they cannot germinate.

Long-distance dispersal would mean that the same amount of seed is distributed over a larger area which could actually lower the odds of encountering a compatible host in a suitable location. The fact that many orchid species are endemics with very limited distributions supports this theory. This does not mean, however, that their seeds are not able to cover long distances. Orchids managed to reach isolated islands far away from the mainland. As famously documented, they were among the first pioneers to resettle on the islets of Krakatoa after the catastrophic volcanic eruption of 27 August 1883. 


An open fruit of the Asian tiger orchid (Grammatophyllum speciosum) showing the placenta with some remaining seeds and a close-up of the seeds
An open fruit of the Asian tiger orchid (Grammatophyllum speciosum) showing the placenta with some remaining seeds. Top right: close-up of the seeds (scale bar shows 1 mm; photos)

Why so small?

Shedding millions of seeds most of which go to waste, seems very wasteful. However, evolution shows no mercy with wasters and given the orchids’ success, their seed dispersal strategy must pay off. In fact, producing lots of very small seeds with literally no food reserve (apart from some oil droplets and starch grains in the embryo) is energetically inexpensive and doesn’t take up that much of a plant’s energy at all.

The survival benefits of producing millions of tiny seeds clearly outweigh the costs of producing them. Not only orchids prove this point. Other families, like the Orobanchaceae (broomrape family), pursue the same strategy. As parasites, they have a similar problem to orchids: they need to get their seeds to meet the right host partner in order to grow into a new plant. 


Left: Ivy broomrape (Orobanche hederae, Orobanchaceae) growing outside the School of Horticulture at Kew Gardens. Right: just under 0.4 mm long, the tiny seeds of this parasite look similar to those of certain orchids but they lack the balloon-like seed coat.

Vanilla ice cream and seed morphology

Since we are talking orchids here and most of us love ice cream, here’s a seed morphological nugget for you. Next time you treat yourself to some good quality vanilla ice cream you can discover that the tiny black spots in it are actually real vanilla seeds (in cheap ice cream they might be fake!). Vanilla is made from the fermented fruits (‘pods’) of the vanilla orchid (Vanilla planifolia). That’s how all those seeds end up in your ice cream. Sadly, though, the seeds of vanilla are nowhere near as exciting as those of other orchids. They are just very simple, unexciting looking, tiny black discs. Lacking the transparent balloon-like seed coat so typical of other orchids, their seeds are obviously not wind-dispersed.

In fact, the seed dispersal strategies of vanilla orchids are still enigmatic. The fruits of many Vanilla species, including the ones of V. planifolia, open when ripe to expose their tiny seeds covered in an extremely sticky layer of oil. The oil might serve as an adhesive to attach the seeds to visiting animals, which could either be insects or vertebrates. For example, it has been observed that euglossine bees are attracted by the fragrance of vanilla fruits and act as seed collectors and potential dispersers. 


Delicious vanilla ice cream image above and below image of the vanilla seeds
Top: Vanilla ice cream (exciting!). Below: scanning electron micrograph of a vanilla seed (not very exciting!)

Orchid seed research at the Millennium Seed Bank

At this point I asked my colleague, Tim Marks, to tell us something about the research into orchid seeds he is involved in at the Millennium Seed Bank and he writes:

'Being wind-dispersed, orchid seeds are naturally dry at release and appear to be desiccation tolerant. The latter is essential for us to be able to preserve them under very dry and very cold (freezing!) conditions, as we do with other seeds in the Millennium Seed Bank.

'Unfortunately, orchid seeds have the reputation to be short-lived under seed banking conditions. Our research is engaged in finding out why this is and how we can extend their survival.

'A basic concept in understanding their specific requirements for storage is to test the relationship between temperature and moisture content upon viability and germination. By running long-term storage experiments with temperatures between -196°C (liquid nitrogen) and +20° (ambient), and a variety of moisture contents, it is possible to identify species-specific requirements.

'Some orchid species prove tolerant to a range of conditions, while others store better in liquid nitrogen. However, to prevent repeating this on all species, we are looking at a number of seed characteristics that could affect this response. One of these is lipid content of the seed, the physical properties of which could affect seed physiology as they go through the freeze and thaw cycles that stored seeds are subjected to. It is possible to produce thermal fingerprints describing the phase transitions between liquid and solid states that these go through, with the intention of developing a predictive model that will describe the observed responses to storage and during germination.' 


Our cryo-storage facility at the Millennium Seed Bank where we keep orchid seeds at -196°C
Our cryo-storage facility at the Millennium Seed Bank where we keep orchid seeds at -196°C


Related Links


20 February 2014
Like other orchid seeds, the seeds of Disa are endosperm-less. However, mature seeds of Disa do contain substantial reserves of lipid and protein (no starch!) in the embryo which enable the seeds to germinate on their own.
Abuse Reported. Comment will be reviewed and removed if necessary.
13 February 2014
If Disa orchids do not use a mycorrhizal fungus, does this mean that, unlike most orchid seed, their seeds contain an endosperm or other food store?
Abuse Reported. Comment will be reviewed and removed if necessary.
10 December 2013
Thank you for the information. Would wish to add here that orchid seeds are diverse even in temperature specificity upon storage. I think for better conservation of such fragile plants, specific genebanks for orchids could be a future choice, may be with specific conservation conditions on the level of subfamily or even genus of Orchidaceae.
Abuse Reported. Comment will be reviewed and removed if necessary.

Browse by blog team