Derek the Hybrid Detective

DD1211 Is Billbergia brasiliensis a hybrid?
by Derek Butcher and Eric Gouda, December 2011

Summary – Ever since this taxon was described in 1943 there have been difficulties in linking live material with this name. All the past references have been revisited and others not previously quoted, investigated and it has been concluded that this name should continue as a recognised species within the subgenus Helicodea with slight amendments to the description. B. kuhlmannii, B. mohammadii and B. velascana are treated as synonyms.

Let us look at how problems of identity can arise where you have no collection data and you live in Australia!

Billbergia brasiliensis L.B. Smith. (by Derek Butcher, Sth Australia in Bromeletter 30(3): 5. 1992)

Have you ever been bothered by a problem that keeps coming back and just nags and nags? Well, here is one which has been partly laid to rest or fully laid to rest if you change your labels.
It all started in the 1970's when we South Australians got a plant from Victoria called Billbergia kuhlmannii and in those days we believed everything the Victorians said. It was a pretty plant with few leaves, broad white banded and forming a tight tube. It had blue flowers which mostly coiled up like a watchspring. It easily set seed and the seed pod was always 3 sided (trigonous) with many irregular longitudinal grooves (sulcate).
A similar looking plant arrived from Queensland under the name of Billbergia venezuelana and another from N.S.W. as B. brasiliensis. We were swamped with names! In 1980 in the Brom Soc Inc seed bank, there appeared B. exotica. Was this a hybrid with such a fancy name? It was listed in the 1979 International Checklist of Bromeliad Hybrids with no mention of parentage. In the latest "Beadle" listing, we have B. ‘Exotica’ as a name given by Kent to a possible B. vittata cultivar. ( this is confirmed in the Bromeliad Cultivar Registry 1998 but the description is so vague it could well be describing B. brasiliensis). All the plants that grew up under our tender care seemed to be exactly the same as our blue flowered B. kuhlmannii. In March, 1984 the Brom Soc. Inc seed bank offered B. alphonsi-johannis which four years later under our tender care produced a blue flowered - yes , you've guessed it !
By this time I was becoming desperate and just had to confide in Bill Morris – an Australian who is one of the Bromeliad Society International Trustees. The only plant that tallied according to my interpretation was yet another name - B. velascana. This was the only one with blue flowers and a grooved trigonous seed pod. Bill did calm me down (or did it just to shut me up) by asking me if the ovaries on page 2033 of Flora Neotropica were from flowers just opened, post floral, or full of seed. How was I to know? Secondly what were the chances of South Australia having a rare B. velascana from Bolivia against the more common B. brasiliensis which comes, we think, from near Rio de Janeiro, which is where many of the first Bromeliad imports to Australia came from. This didn't seem very scientific to me especially when Smith's writings had suggested the likelihood that B. brasiliensis was a hybrid. If he meant an F1 hybrid then seedlings from this should bear same relationship to either of the original parents (viz. the grandparents). However, our experience showed that self pollination produced regularity in progeny. So I forgot about the problem!
At a South Australian Society meeting in November 1991, a member came up with plant in hand and wanted to know its name. Yes, it was our blue flowered Billbergia and I quietly went to a corner to bang my head there.
'WRITE TO LUTHER' kept drumming in my head. So I sent a slide, but even from that meagre picking Harry was fairly sure we had B. brasiliensis. I now know that the ovary becomes sulcate as it matures, and that the lower ovaries can be trigonous before and at flowering. Neither of these facts are on page 2007 of Flora Neotropica.
By the way, B. velascana has gone by the wayside and B.kuhlmannii has a tubular ovary and is so rare that it is not even in the Marie Selby collection. So if you have a Billbergia with few, white-banded stiff leaves, a blue flower which invariably winds up like a watchspring, and a three sided sulcate ovary which becomes almost globular when ripe, then you have Billbergia brasiliensis.

In Europe in 2006 Eric Gouda was having similar identification problems and we decided to do something about it!

We know that Lyman Smith tried to link plants from horticulture with plants collected in the wild even though they appeared to be of man-made hybrid origin. Vriesea Xmorreniana is one example where the alleged plants found in the wild have now been identified as species in their own right as Vriesea flava.
When Lyman Smith coined the name Billbergia brasiliensis in Arq. Bot. D Paulo II. 1: 105. 1943 he commented that the taxon could well be a hybrid because of the inconsistency of the coiling of the petals.(fig 1) This naming was necessary because there were two taxa with the name Billbergia leopoldi. He could not follow Linden and Houllet because they created a naming conflict with their transfer of Helicodea leopoldi in 1869 to Billbergia leopoldi which had already been used by Koch in 1857.

Let us look at a bit of history that prompted Lyman Smith to treat this as a possible natural hybrid. No further collections are reported in Flora Neotropica (1979) and we would suggest that the inconsistent petal coil would not have been evident in the herbarium specimen Glaziou 16421 held at Kew. - In 1856 Koch wrote about a Billbergia leopoldi which was eventually treated as a synonym of B. vittata. This takes this taxon out of the discussion. In any event the herbarium material held in Berlin was destroyed by fire in WWII.
- In 1862 icon 67 at Kew was painted by M Severyns for Morren
- In 1864 Lemaire wrote about Helicodea leopoldi Hort. Versch. that had been collected by De Vos in Santa Catarina in 1847.There is also an illustration which shows the fruit of this taxon. It is triangular in cross section (Fig 2). Note that Reitz (1983) reports this plant has not been sighted in this area.
- In 1866 icon 289 at Kew was painted in 1866 for Morren and is annotated B. brasiliensis by Dave Philcox who was collaborating with Lyman Smith at the time.
- In 1869 Houllet discussed the growing of B. leopoldi in apartments and was clearly referring to the plant we now know as B. vittata. Smith’s reference is in error.
- In 1871 Belgique Horticole 21: 1-6. (Fig 3) Morren again had a painting done.
Paul Wilkin from Kew advises that all three paintings are different but petals are rolled back to the same degree in all 3! We know that horticulturally ‘Helicodea’ are not always consistent with the extent of the coiling especially as the inflorescence ages. E. Morren described the following “Ovario cylindracea-trigono, costato, farinoso. Baccis 9-12 costatis, subrotundis, parum carnosis, trilocularibus, rimis longitudinalibus irregularibus loculicido dehiscentibus.” We intend to emend the description in Flora Neotropica to include this data.

If we now refer to Flora Neotropica page 2035 we read that Witte in Cat. Bromel Lugd. Bat.11. 1894 treats B.leopoldii as a hybrid and quotes the parentage of decora x saundersii. However, on referring to the actual publication in 1894 we find the following:
52. Leopoldii Ed. Morr. (Brasil)
Belg. Hort. 1871, pag.1, tab. 1-4
Helicodea Leopoldii Lem.
Billbergia ianthina Hort.
Billbergia nuptialis Hort. Mak.

Smith was in error in quoting this as a hybrid. Also, where did Lyman Smith get the alleged parentage? The notation that this is not the B. leopoldi of Morren, 1871 is also confusing because this painting is actually referred to by Witte. We have already proved that Morren 1871 and Morren 1866 paintings have similar traits which makes it hard to believe Smith did not accept these paintings as representing the same species. Is there something that Smith knew 85 years later that Witte was not aware of? We find it hard to believe that there was yet another plant only this time called ‘x Leopoldi’.

For many years now, plants have been identified as B. brasiliensis in the USA and Australia, and B. kuhlmannii in Brazil and Europe, and all breed true from self set seed. All 3 petals coil as expected with sub-genus Helicodea. This indicates that instead of being a hybrid they are a species. However, they have an ovary that is triangular in cross section which differs from Smith’s description which states subterete.

In investigating the possibility of this being a man made hybrid there is little material to refer to, other than the works of Duval. It seems that active hybridising did not start in this genus until the 1880’s and we are talking about a plant described some 20 years before. The chances of it being a man-made hybrid are remote.

Let us now look at the protologue of B. kuhlmannii. Smith did his differential diagnosis against B. rubicunda another species with peculiar petal coiling habits and also said to be of hybrid origin. He said “The species seems similar to Billbergia rubicunda Mez, but very different by the dense scaled inflorescence, the short epigynous tube, and blue petals.” Interestingly B. brasiliensis has a dense scaled inflorescence, a short epigynous tube and blue petals! Could B. kuhlmannii also have a tendency to irregular petal coiling habits? In the Billbergia key in Flora Neotropica pp. 1977-8 we see B. brasiliensis distinguished from B. kuhlmannii by having a non-sulcate wholly white farinose ovary which does not agree with the illustrations.

Recent collections of a similar plant in Brazil are being identified as B. kuhlmannii which is supposed to have ellipsoid ovaries but are triangular in cross section in the live collections! (Fig 4) Add to this the fact that Billbergia mohammadii Vasquez & Ibisch, Die Brom. 1: 11-13. 2000 was considered synonymous with B. kuhlmannii in the Bromeliad Binomial List by Harry Luther in April 2002. Admittedly, no reason was given for this action. B. mohammadii does have an ovate ovary that is triangular in cross section.

When describing the new species B. mohammadii Vasquez & Ibisch compared their plant with the rarely encountered B. velascana Cardenas (Brom Soc. Bull. vii. 35) which had been collected in an adjacent area in Bolivia in 1955. B. velascana is an enigma because only one collection was made and no further collections have been reported. In fact, when Vasquez & Ibisch described their new species, comparison was made with the protologue data of B. velascana with no comment about having seen live specimens of this species which is what we always expect with their indepth reporting. Allowing for variables you would expect from a description based on more than one collection, the only thing that seems to have some significance is the length of the epigynous tube which is quoted at 12mm long. This seems an odd amount because the ovary is also quoted as 12mm. This figure was changed without comment to ‘short’ in Flora Neotropica. This is within the “short but distinct’ quoted for B. brasiliensis by Lyman Smith. It is clear, at this moment, that B. velascana cannot be distinguished from B. brasiliensis and should be treated as a synonym too.

Something must be done with B. brasiliensis, B. kuhlmannii, B. mohammadii, and B. velascana and a decision made as to a correct name. We maintain this is B. brasiliensis (Fig 5) because it predates all others. If the description of B. brasiliensis is amended to delete reference to the contorted petals and to add the fact that the ovary is triangular in cross-section and that the fruit is subrotund with 9-12 ribs, all these species will fit comfortably.

Finally, we did notice that Smith in Flora Neotropica Monograph 14, Part 3, Bromelioideae, 1979 refers to B. leopoldi Lem. as the 7th synonym of B. zebrina on page 2026. This is incorrect because in 1864, although reference was made to both plants there was no suggestion that they were the same taxon.

1864 Helicodea leopoldi Verschaffelt Hortus ex Lemaire, Ill. Hort. 11: sub pl. 421.
1869 Billbergia leopoldi Linden ex Houllet, Revue Hort. 41: 87, fig. 21.
1871 Morren E. in Belgique Horticole 21: 1-6.
1894 Witte in Cat. Bromel Lugd. –Bat.11.
1943 Smith L. B. in Arq. Bot. D Paulo II. 1: 105.
1957 Cardenas M, B. velascana in Brom. Soc. Bull. 7(3): 35.
1979 L B Smith & R J Downs in Flora Neotropica Monograph 14, Part 3, Bromelioideae
1983 Reitz R. in Bromeliaceas E A Malaria – Bromelia Endemica, p486
1990 Leon Duval in The Bromeliads, edited by Rothenberg & Read, Big Bridge Press, Pacifica, Cal., USA
2000 Vasquez & Ibisch, in Die Brom. 1: 11-13.
2002 Luther H. Bromeliad Binomials p12. 2002
We would like to thank Jason Grant of Neuchatel University in Switzerland, and Paul Wilkin of Kew for providing us with historical data and advice.

Billbergia brasiliensis

Inconsistency of petal coiling.
Fruit triangular in cross section.
Painting in Belgique Horticole 21:1-6.1871.
Billbergia brasiliensis. Derek Butcher 11/11.
B. kuhlmannii in Brazil.

Updated 02/12/11