An interpretive nightmare:
The Aechmea distichantha / wittmackiana Complex

By: Derek Butcher and Peter Franklin

We have unearthed a number of photographs of plants in reputable bromeliad books which do not appear to link with the descriptions or key given by Lyman B Smith in Flora Neotropica Bromelioideae (1977). We have also investigated plants being grown in Australia which correspond to these photographs. It has been suggested that we take herbarium specimens of the plants in Australia but we decided against this because it would cloud the situation even more and, of course, we do not have habitat information.

It has also been suggested that our solution would not be upheld because we had not critically examined the herbarium specimens or the original descriptions. This would have been difficult from here in Australia and the fact that we are amateurs, access to the various herbaria would have been very difficult. In any event we feel some trust should be made in the descriptions by Lyman Smith even though interpretation may be in error.

So we leave it to posterity where some professional botanist may take up the challenge and is welcome to use what material we have unearthed. In the meantime we will add our views to cyber-space rather than publish in some printed form according to ICBN rules.

The following will show you how we have addressed the problem and offer possible solutions.

There appears to be quite a diversity of these plants in Australia and in the United States with some misnaming involved. Let us see how Lyman Smith sees the key differences.

Scape stout, erect, inflorescence amply compound, spikes many flowered, placentae apical, leaf blades not channelled = Aechmea distichantha

Scape slender, decurved, inflorescence depauperately compound, spikes 1 - 3 flowered, placentae median, leaf blades channelled. = Aechmea wittmackiana

When you check the description of A. wittmackiana you find that placentae is sub-apical to sub-central, which means that the characters used in the key don't match the description.

Now let us look at the differences within A. distichantha

1.  Inflorescence lax or sub-lax, usually broadly pyramidal, spikes more or less spreading, many flowered, leaves usually acute or acuminate. var. distichantha
 2. Petals purple or blue var distichantha forma distichantha
 2. Petals white var. distichantha forma albiflora
 1. Inflorescence dense, spikes erect, few flowered, leaves usually attenuate or rounded and apiculate  
 3. Inflorescence elongate, slenderly cylindric or fusiform, plants large, leaves usually attenuate var. schlumbergeri
     3. Inflorescence short, ovoid, plants small, leaves usually rounded and apiculate var. glaziovii

Just after the Smith and Downs Monograph was published (1979), Pereira named Aechmea distichantha var. vernicosa (Bradea 2 : 308. 1979). The interesting thing here is that this plant was described simultaneously with Billbergia pyramidalis var. vernicosa. It has since been found that the Billbergia specimen had had insecticide sprayed on it giving it that 'varnished' look! The Billbergia naming has been abandoned, but what about Aechmea distichantha var. vernicosa? We'll exclude it, assuming that Aechmea distichantha var. vernicosa is a normal Aechmea distichantha that was also sprayed with the insecticide.

Aechmea 'Glutinosa'
Aechmea 'Glutinosa'
We have been looking at this group from different angles. Remember Peter Franklin's "Exactly what is Aechmea 'Glutinosa'" in Bromeletter 1999 #1? Remember too that this Aechmea 'Glutinosa' had been proudly brought to Australia as a "NEW" Aechmea species from California in 1982 when Uncle Derek was just Derek! In addition Derek has received further living material from Ruby Ryde (the inflorescence was a bit past it but it was still alive!). Aechmea 'Glutinosa'
Aechmea 'Glutinosa'

Let us digress. You may recall that we stirred up a problem with Foster's Aechmea distichantha var. canaliculata which was collected somewhere in the State of Sao Paulo. Here we had problems with Foster's collection number 3043. A note in Smith and Downs page 2078 indicated that this collection had been split, suggesting that one part was in A. wittmackiana and the other in a new variety of A. distichantha. Jason Grant has checked the 3043 at the Smithsonian and ALL of this collection went to A. wittmackiana and had not been split! Harry Luther was brought into the discussion and he agreed with Jason's findings. As is usual, I got photographs of herbarium specimens! These showed an inflorescence that was cylindrical at the top but with spreading distichous spikes at the bottom. The voucher for Foster's 3043 says simple top half, branched below, and the Selby Gardens voucher indicated that their specimen was probably a clone of Foster's 3043.

Back to Ruby's living material. Derek had always wondered why Foster used the word "canaliculata" which means grooved or channelled. But now he knows because of the pieces that Ruby sent. The leaves have two keels forming a box gutter which is very distinctive. We cannot find anything in Stearn's Botanical Latin to describe this structure but perhaps sub-bicarinate might do.

Aechmea wittmackiana - Ruby Ryde plantAechmea wittmackiana leaf - Ruby Ryde plant
A. wittmackiana
Ruby Ryde plant
Ruby had sent me two specimens, one from the Buchanans at Wardell in New South Wales, and the other she and Keith had collected. The Buchanans' one had allegedly been sourced from Marie Selby Gardens in Florida whereas Ruby's had been collected near the coast at the edge of rainforest between Ubatuba and Caraguatuba in Sao Paulo State, Brazil. The only differences Derek could see were attenuate leaves with a little barring on the back for the Buchanans' plant compared to rounded, mucronate and very faint barring on the other. Aechmea wittmackiana - Buchanan plantAechmea wittmackiana leaf - Buchanan plant
A. wittmackiana
Buchanan plant

The best way to look at this problem is via a chart so that you can see how the various cultivated forms line up against certain parts of descriptions of the allied species.

Anyone investigating this group will be aware of the sharp teeth on the leaf edge in the larger forms but we decided against using this facit because it is difficult to judge in a photograph AND Lyman Smith shows us that for Aechmea distichantha leaves are finely serrulate or armed with stout spines 4 mm long. We were unable to check on the original descriptions but Baker reveals Aechmea glaziovii as having small teeth.

Scape stout and erect X X   X X X X X X
slender/decurved     X            
Amply compound X           X X X
Depauperate compound   X X X X X      
Dense   X X X X X      
Open X           X X X
Spikes many flowered X           X X X
1 - 3 flowered   X X X X X      
Leaf not channelled X X              
channelled     X X X X X X X
25 - 80 mm wide X X   X X X      
to 25mm wide     X       X X X

D = distichantha
S = schlumbergeri
W = wittmackiana
OQ = Olwen's Query
G = 'Glutinosa'
MS = McNamara's schlumbergeri
BW = Buchanans wittmackiana
RW = Ruby's wittmackiana
GW = Golinski's wittmackiana
Please note that Golinski's wittmackiana started off as a Quesnelia but was then renamed as an Aechmea.

Aechmea wittmackiana Golinski
A. wittmackiana
You will see from this chart that there are a few problems if we follow Lyman Smith's interpretation. The Buchanan's , Golinski's and Ruby's plants agree with each other, but do not correspond to the description because of the inflorescence structure. Peter's A. 'Glutinosa' Marjory's var. schlumbergeri and "Olwen's Query" which we cannot check for habitat information, agree with each other and link very closely with A. distichantha var. schlumbergeri. Olwen's Query
Olwen's Query

While on the subject of A. distichantha, Peter and I have been discussing whether true A. distichantha var. glaziovii is in cultivation. The plant in Australia looks like the plant in Baensch's Blooming Bromeliads (1994: 63) but differs from the description by having an amply compound inflorescence which is open not dense, and the spikes are many flowered not 1 - 3. Reference to Baker's "Bromeliaceae" (1889) seems to confirm that we should be looking for a smaller version of var. schlumbergeri rather than a smaller version of var. distichantha. A black and white photograph in Rauh's Bromeliads for Home, Garden, and Greenhouse (1979: ill. 185) shows what we are looking for.

Identification of Aechmea distichantha var. schlumbergeri is to our mind also in doubt . If you refer to Baensch's Blooming Bromeliads (1994: 63 ) you will see a spindle shaped, sub-lax inflorescence with many-flowered side-spikes. Only the spindle shape agrees with the description! A closer match to the described var. schlumbergeri is a plant that Peter obtained from Marj McNamara which has, you've guessed it, sub-bicarinate channelled leaves, providing a link between schlumbergeri and wittmackiana.
A. distichantha
v. schlumbergeri

If you really want more problems then read Baker's "Bromeliaceae" (1889) for descriptions of several species now treated by Lyman Smith as synonyms. All are clearly A. distichantha with some that seem to be describing var. schlumbergeri but all are treated by Lyman Smith as var. distichantha! Interestingly, no one mentions the sub-bicarinate leaves and this includes Aechmea jucunda which was placed in synonymy under Aechmea wittmackiana. We wonder too whether sub-bicarinate leaves can be easiily detected in herbarium specimens. One can only assume that Lyman Smith saw ALL the herbarium specimens involved and decided differently. However, this has not been recorded. Regrettably, too, we have no access to herbarium specimens and can only rely on photographs in reputable books where others have done the identification, namely, Rauh, Reitz, Luther and Read, Leme, and Steyermark.

Aechmea distichantha
A. distichantha
A. distichantha v. glaziovii
A. distichantha
v. glaziovii
A. distichantha v. glaziovii
A. distichantha
v. glaziovii
BSIJ 1977 p168
Aechmea jucunda painting
Aechmea jucunda
BSIJ 1986 p8

The differences between the forms we have looked at seem to hinge upon:
1. Compound inflorescence, branches spreading versus depauperate compound, branches erect, simple at top.
2. Plant big versus plant small.
3. Leaves channelled versus non-channelled.
4. Scape stout versus slender.

Which of these characteristics is significant enough to split the distichantha/wittmackiana complex into species or varieties? Lyman Smith seemed to prefer channelling and then compound inflorescence and then size, which we will follow. So using those assumptions, we have the following key which quite successfully describes our investigations and allows the placement of all the the material that we have seen.

1. Leaf blades sub-bicarinate, channelled   = Aechmea wittmackiana
2. Inflorescence depauperate compound (simple at top)  
    3. Leaves narrow to 25mm   = A. wittmackiana var. wittmackiana
Includes Fosters 3043, Selby Gardens herbarium specimen, Marnier Lapostolle picture BSB 1967 p75, and A. jucunda BSIJ 1986 p8
3. Leaves wide, 25-80mm   = A. wittmackiana var. schlumbergeri
Includes A. 'Glutinosa', Olwen Ferris's Query, Marj McNamara's schlumbergeri
2. Inflorescence compound with open spreading branches  = A. wittmackiana var. ramosa
Includes Ruby's plant, Buchanans' plant, and Golinski's plant
1. Leaf blades not channelled  = Aechmea distichantha
4. Inflorescence compound with open spreading branches  
5. Plant large, leaves 25-80mm wide  = A. distichantha var. distichantha
6. Petals blue  = A. distichantha var. distichantha f. distichantha
Includes most forms in cultivation and photograph in Bromeliaceas of Venezuela (1987: 52) photograph page 63 in Baensch's book as var. schlumbergeri, painting and description in Reitz 1983.
6. Petals white  = A. distichantha var. distichantha f. albiflora
No material seen but allegedly in Australia.
5. Plant small, leaves to 25mm wide  = A. distichantha var. minor
Includes plant generally grown as var. glaziovii, Photo in Baensch's Blooming Bromeliads p 63 as var. glaziovii, and Photo BSIJ 1977 p 168 as var. glaziovii.
4. Inflorescence depauperate compound (simple at top), leaves to 25mm wide  = A. distichantha var. glaziovii
Includes photo in Rauh's Bromeliads for Home, Garden and Glasshouse.(1979: ill. 185)
Not seen in cultivation.

To accommodate these changes we propose the following names:

1. Aechmea wittmackiana (Regel) Mez var. ramosa Butcher and Franklin var. nov.
Differt a forma typica sed inflorescencia ramosa laxior

2. Aechmea wittmackiana (Regel) Mez var. schlumbergeri (E, Morren) Butcher and Franklin comb. nov.

3. We considered the possibility of having another form of Aechmea distichantha v. distichantha to cover those plants that have been misidentified as var. glaziovii. However to accommodate the existing described forms based on petal colour and following the logic of Lyman Smith in separating var. schlumbergeri from var. glaziovii on size at varietal level, we decided that the same differentiation be used for:-

Aechmea distichantha Smith var. minor Butcher and Franklin var. nov.
Differt a forma varietata sed inflorescencia et folii brevior.

Identification of A. wittmackiana and the varieties of A. distichantha seem to hinge on how lax is lax, how channelled is channelled, and how few is few. An intriguing puzzle which you can easily solve by allocating numbers to the various clones just like Peter does to remove the influence that is undoubtedly present because of preconceived notions like "This plant must be correctly named because I got it from so and so!"

The A. wittmackiana (Buchanan's) from Marie Selby Botanic Gardens, presumably identified by Harry Luther suggests that a depauperate compound inflorescence or slender scape are not significant. Now, if the same logic were applied to the allied species Aechmea distichantha it would have no formal varieties! We can find no written information on the subject. We wonder whether this is yet another case of too many forms and not enough accepted varieties, and so these forms have been given an arbitrary but not necessarily correct name. If this is the case then it adds weight to Peter's new Key and solution to the problem.

If we refer to Reitz's Bromeliaceas (1983: 413 - 417) we will see Aechmea distichantha described but only using the synonyms used by Lyman Smith for A. distichantha var. distichantha f. distichantha. However, the description does refer to erect to sub-erect branches of 2 to 7 flowers. The low number of flowers adds problems to Lyman Smith's key in differentiating A. wittmackiana from A. distichantha on the one hand and the few flowered branched varieties of A. distichantha on the other.

Secondly, Leme seemed loth to use A. distichantha var. schlumbergeri on the photo on page 104 in Bromeliads in the Brazilian Wilderness 1993 even though the inflorescence pictured is far from open, amply compound and spreading. This has the spindle shaped inflorescence but the branches seem few flowered and the side branches are held tight to the main rhachis. It may well be that this plant has erect rather than spreading branches because of the climatic conditions at the time the photo was taken.

The BSI made a giant step forward by having an Identification Centre with Harry Luther running it at Marie Selby Botanical Gardens. Thus, the Alphabetical List of Bromeliad Binomials plays an important monitoring role with a steadying effect particularly now that the De Rebus series in Selbyana provides a formal back-up. Here we find that Smith's interpretation is still upheld for Aechmea distichantha and Aechmea wittmackiana. A curious note is that a plant called Aechmea wittmackiana purported to have come from Marie Selby Gardens seems to contradict Smith's findings and seems to be different to the Selby herbarium specimen.

All of these differences, anomalies, omissions and discrepancies between live plants, descriptions and botanists' interpretations show that something needs to be done to clarify the issues. Our suggestion to have two new names and a new combination of existing names is intended to test the water as to whether the Smith approach should prevail. We have deliberately taken a "splitter" approach, however the problem can be just as successfully resolved by a "lumper" approach. We await some "steadying" effect from Harry Luther where we may see all with strong teeth being Aechmea distichantha and those serrulate being Aechmea wittmackiana.

Peter is already thinking along the lines that the described var. glaziovii fits better in wittmackiana than it does in distichantha but the theory is impossible to check without finding channelling in the leaf blades of the herbarium specimens of var. glaziovii. This will have to remain a theory until further evidence is available.

Please check the photographs linked to this article and to those referred to in other books and advise us if you see any flaws in our argument.



Baker J. G. 1889 Handbook of the Bromeliaceae ; Plant Monograph Reprints J. Cramer and H. K wann 1972

Leme E. M. C. & Marigo L. C. 1993 Bromeliads in the Brazilian Wilderness

Oliva-Esteva F. & Steyermark J. A. 1987 Bromeliaceas of Venezuela

Rauh W. 1979 Bromeliads for Home, Garden and Greenhouse (English version)

Reitz R. 1983 Bromeliaceas E A Malaria-Bromelia Endemica

Smith L. B. & Downs R. J. 1979 Bromelioideae (Bromeliaceae) In Flora Neotropica, Monograph No. 14 part 3