Identifying GENUS DATURA L.

Difficulties in connection with proper identification of the species in the Genus Datura L.  A suggestion of including sub-groups into the sections.


The Datura Collection International

The Datura Collection International was started in 1990 in Sorø, Denmark. The idealistic purpose was to carry out morphological examinations of the herbaceous species found within the Genus Datura L. and to run a series of test experiments to determine, how to preserve such seed collection in the best possible manner. 

From 1995 Datura seeds were bred on a scale of 50-1.000 grams each year in order to be distributed for scientific purposes and for mutual exchange between botanical gardens and non-botanists seriously interested in this Genus. 

Also from 1995, a limited number of seed accessions were released on the seed market, and in some cases – in order to enlarge the seed collection by including cultivated strains of the wild-growing Datura species – the same accessions were exchanged for seeds coming from the trade.

A selection program for Datura specimens especially suited as garden ornamentals was established in 1997. The species involved in the program is D. inoxia Mill., D. meteloides Dun. In D.C. and D. wrightii Regel. 

Several F1-hybrids and selections have been made, but none of these have not yet been released on the market. 

The selection criteria’s was cold-tolerance; rapid growth; eye-catching flowers; and stable flower production, when grown out in hardiness zone 5-7.

The necessity of genetic diversity

Quite from the beginning, Datura seeds were collected with the purpose of personal studies of this inviting Genus. The extensive amount of literature available, concerning investigations in this field, were also an object of study. As a direct consequence of these studies and the first series of test experiments carried out, it became evident, that the reliability of the test results obtained was indeed intimately related to the “quality” of the seed collection at hand. A large collection containing several different samples of D. meteloides Dun. in D.C. and of D. wrightii Regel may convince the serious student, that both are distinct species, for instance by a close examination of the leaves. Much of the misapplication of species names could have been avoided in this manner. The past nine years of my own morphological examinations proves this fact clearly.

Also in the literature, it is evidently clear, how limited access to certain seeds or certain branches of the literature can be very damaging to an investigation, as it has nearly always lead to inadequate conclusions when such investigation is based on insufficient plant materials.

That is the exact reason, why it is considered to be of greatest importance to preserve the diversity of genetic resources found within the Genus Datura. This, for instance, implies the preservation of propagation materials collected as well in native and adventive habitats, as well as propagation materials originating from selected medicinal crops and garden ornamentals. It is by no means easy to build up a seed collection containing samples from that many sources, but again I must stress, just how vital this point is to the reliability of the scientific results obtained by the investigators in the field of botany, biochemistry, pharmacy, genetics, etc. It is vital to any investigator to gain free access to varied and well-documented propagation materials representing the species serving as objects for his particular research.

The value of proper identification

In connection with the establishment of a Datura seed collection in Sorø, several taxonomical and nomenclatural problems related to the issue of proper identification of plant materials were encountered. Proper identification is of great importance serving as documentation of the propagation materials. This implies a radical reexamination of the taxonomy and nomenclature of the Genus. This also implies growing out the entire seed collection (currently DATU-01 to DATU-127) in order to re-identify these accession numbers and also to continue to study the range of morphological expressions in each individual under different cultural conditions. Whenever it is possible, the original species descriptions will be re-assumed. The sections established by Bernhardi (Bernhardi, 1833) are maintained, but also separated further into series, which is based on distinct morphological expressions of the species.

A serious problem was encountered by maintaining D. stramonium L. and D. tatula L. as two distinct species. Both species include a variety carrying spineless fruits (DATU-16 – DATU-27). There is only one variety name available to both selected taxa: inermis Jacq. To avoid further confusion of these two taxa, I find it necessary – at least in these papers and as a temporary solution to this problem – to re-assuming D. laevis L. for the white-flowered variety, but in a new combination: D. stramonium L. var. laevis (L.) -. 

In connection with D. metel L., Danert (Danert, 1954) has established the hitherto most detailed and best-structured classification of this taxon. However, for some reason, Danert did not include the yellow, double or triple-flowered form very much resembling his D. metel L. var. muricata (Bernh.) Danert.

Nor has he included the single-flowered forms similar to his descriptions of D. metel L. var. fastuosa (Bernh.) Danert f. fastuosa Danert and D. metel L. var. fastuosa (Bernh.) Danert f. malabarica Danert.

The woody Datura species of sect. Brugmansia is sometimes regarded as a separate Genus (Lockwood, 1973). In 1999 it became possible to grow out several species in Sorø and some examinations of the tree-Datura morphology have been carried out, but several have yet to be made in order to establish a reliable thesis on this section of South-American Datura species and in this connection, it also necessarily to enlarge the seed collection in regard of seeds stemming from sect. Brugmansia. However, in the following, I have written down some of my current impressions based on detailed examinations of nine specimens of that particular section.

The application of sectional sub-groups

The Genus Datura L. is according to Bernhardi (Bernhardi, 1833) separated into four sections: sect. Stramonium Bernh., sect. Dutra Bernh., sect. Ceratocaulis Bernh. and sect. Brugmansia Bernh. The distinction between the sections rests on several morphological characters. The most frequent object of discussion is the direction of the flower and the character of the fruit.

The species of sect. Stramonium forms an erect capsule covered with rigid spines or unarmed capsules, splitting Flowers erect, small. The Corolla limb is mostly rounded or pentagonal with five primary teeth.

The fruits of sect. Dutra is deflexed, berry-like (succulent), and armed capsules consisting of fragile and spongy valves, which at first splits open irregularly and next fall apart. Flowers are erect, either very large or very small. Corolla limb rounded or pentagonal – rarely genuine decagonal – with five primary teeth separated by an equal number of secondary teeth making the limb appear to be decagonal. Calyx tube rounded, apex splitting regularly into 5 teeth, circumscissile near the base.

The fruit of sect. Ceratocaulis are distinct from the fruits of sect. Dutra by being unarmed, more a berry, than a capsule, and these are usually referred to as berries. Flowers erect, large. Corolla limb decagonal with five primary teeth separated by five secondary teeth. Calyx tube rounded, spate-like not circumscissile, falling off entirely.

The fruits belonging to sect. Brugmansia is unarmed, glabrous, or pubescent berries. Some fruits split open irregularly. Other fruits never open, but fall to the ground, where they decompose and thereby release the seeds. Calyx tube rounded or pentagonal, apex splitting regularly or irregularly into 5 teeth or like-like. In some species, the calyx is persistent and encloses the fruit, in other species it fall off entirely. Flowers pendant. Corolla is more or less pendant limb rounded or pentagonal with five primary teeth separated either by an equal number of sinuses or by five secondary teeth.

I find, however, that there are several serious discrepancies connected to such distinction. These discrepancies were not only found to persist in the much-discussed sect. Dutra, but also in sect. Stramonium, even if it is rarely – or not at all – mentioned in the literature. I do not stress these facts in any way to underestimate the work of Bernhardi. The work of Bernhardi stands as evidence of mental acuteness and thanks to his genius he created new angles to the problems of Genus Datura and enabled many serious workers after him to continue their attempts to solve some of these enigmas.

Commonly D. ferox (DATU-33 – DATU-34 – DATU-35) D. quercifolia (DATU-36) and D. villosa ( – ) is distinguished from D. stramonium (DATU-01 to DATU-13), D. tatula (DATU-18, DATU-20), and D. bernhardtii (DATU-19 – DATU-91) by their fruits having fewer, but more stout spines. Another distinction can be based on the inner of the fruit walls, which are more fleshy and tends to dry out over a longer period of time compared to the three former species. In contrary to the slow drying out of the inner walls, the outer epidermis tends to decompose slightly prior to the breaking up of the capsule valves, a characteristic often observed in the species of sect. Dutra.

In the older literature, D. discolor was often considering the fruit, which is a deflexed, berry-like capsule, and the spines are somewhat fleshy and the entire fruit is covered by short, close pubescence. The inner of the fruit is succulent and – at least in some fruits – breaks up irregularly. On the other hand, there is the fact, that a varying number of fruits in the same specimen breaks up regularly in four equal – or nearly so – valves (DATU-51 – DATU-52) and that the color and structure of the testa suggest a close relationship to the species of sect. Stramonium. Another morphological character pointing at sect. Stramonium is that the calyx apparently is pentagonal.

The ideal determination of D. discolor would properly be to state, that it belong neither to sect. Stramonium or to sect. Dutra, but that it is both ways related.

A similar argument can be used about a Mexican Datura, D. leichhardtii. These small species form tiny, deflexed fruits armed with short, soft conical spines. Like it can be observed in the fruits of D. discolor, the cells forming the succulent inner walls in D. leichhardtii (DATU-37 – DATU- 110) tend to burst before the breaking of the fruit, which causes the iron content of the cell fluid to discolor due to oxidation. This without any doubt points at, that this species by all good rules can be included in sect. Dutra. However, in all specimens grown out in 1996-1999 50-75% of all fruits broke up regularly into four valves. Also, the morphological expression of leaves and the stem give rise to some consideration concerning the relationship of D. leichhardtii to sect. Stramonium.

Observations of D. metel varieties in the test field give raise to considerations similar to those given account for in the cases of D. discolor and D. leichhardtii. The morphological habit of its stem, branches, leaves, and of corolla limb strongly suggest, that it should be placed in sect. Stramonium. Only in two out of six samples of D. metel var. muricata grown out in 1992-1997 deflexed fruits could be observed. In both single- and double-flowered forms of D. metel var. fastuosa f. fastuosa and – f. malabarica grown out in the same period no deflexed fruits were observed. In all the forms mentioned erect fruits were found to be predominant, except for a few examples of semi-erect fruits. These exceptions were not due to the curving of the fruit stalks. The only morphological expression found in these varieties justifying their placement in sect. Dutra is the fact, that the inner of the fruit walls is succulent and that the fruits break up irregularly.

A recently new seed sample of D. metel L. var. metel reveals valuable new information on several Datura species and their relation to the four sections. The sample was received in March 2001 from New Mexico, USA, and contains seeds, which develop into plants showing all the characteristics of a truly wild plant. The interesting thing about this plant is, that it apparently unites typical morphological features from species usually considered remotely related and also belonging to two different sections. The stem and leaves are of exactly the same type as common for D. stramonium and D. tatula and their smell is also identical to these species. The fruits are almost erect, but bear great resemblance in surface and spine structure to D. leichhardtii and D. pruinosa.

The doubts surrounding D. ceratocaula is not as much it’s rightfully belonging to sect. Ceratocaulis, which is true beyond any doubts and nor the morphological expression of its calyx or its fruits. Rather the fact, that in the literature it is stated about this species, that it differs from the species of the other sections by being more tolerant to water. It is however true, that D. ceratocaula is characterized by having hollow stems and branches, but this is easily proven to be a feature, that this species shares with any herbaceous Datura species of a little age.

Concerning the fact, that D. ceratocaula (DATU-65 – DATU-66 – DATU-71) also is thought of as a species having a weak root system, this could not be observed in any specimen grown out in Sorø from 1996-1999. The pot-grown specimens were grown out in 6-inch pots until they reached the growth stage, where the stem started to rise above the earthborn leaf rosettes. In reaching this stage the specimens quickly outgrew our 15-inch pots – within 3-4 weeks – and large specimens would not be happy until repotted into 30-inch pots. This indicates, that D. ceratocaula in a cultural environment is in possession of a well-developed root system.

The fact, that D. ceratocaula sometimes has been observed growing in shallow water does not seem to turn it into a water-loving plant. The natural law of flow and ebb is also applying to desert lakes and also the fact, that a period of heavy showers increases the water level in the lakes in question. No specimen of D. ceratocaula exanimated could stand more than 72 hours with the lower part of the root system covered by water. Blakeslee (Avery et Al., 1959) seemed to be a little surprised by the fact, that D. ceratocaula did pretty well as a pot plant treated as any other Datura species.

All Datura species are dependent on a certain level of water in the soil. I do not mean to generalize and state, that all Datura species are equally dependent on a high content of soil moisture during their entire life circle. Actually, there are quite conspicuous differences between the species in this regard. Some Datura species own the ability to store water and nutrition in the lower stem and are fully able to recover after a period of severe drought.

D. inoxia (DATU-38 – DATU-39 – DATU-41 – DATU-87), D. meteloides (DATU-44 – DATU-R-45 – DATU-46-B – DATU-72) and D. wrightii (DATU-82 – DATU-107 – DATU-125 – DATU-100 – DATU-101 – DATU-102 – DATU-138

Finally, there is the discussion of the woody Datura species belonging to sect. Brugmansia. As mentioned earlier, these have sometimes been regarded as a separate Genus, Brugmansia. The following considerations of the differences and the similarities between the herbaceous and the arborescent Datura species is, as mentioned in the previous page, based on detailed examinations of a limited number of species and hybrids of sect. Brugmansia, which are  D. arborea L., D. aurea (Lagerh.) Safford, D. candida (D. aurea x D. versicolor) (Pers.) Safford, D. suaveolens Humb. Et Bonpl.

A part of the morphological expression found in some of the arborescent species are also found in a single herbaceous Datura species, D. ceratocaula (

DATU-65 – DATU-66 – DATU-71). In “Über die arten der Gattung Datura“, Bernhardi (Bernhardi, 1833) points out, that more morphological features, especially some characters found in the fruits of D. ceratocaula, are also present within the arborescent species. Other aspects of the morphology of D. ceratocaula, for instance, the habit of the flower, are found in all of the herbaceous Datura species. These facts are used as a link or as a bridge between the arborescent and the herbaceous Datura species of sect. Dutra.. 

The examination of the arborescent Datura species carried out in Sorø in 1999, when holding together with several years of experiences with the herbaceous species, seems to support Bernhardi in assuming, that parts of the morphological expression of D. ceratocaula can be found as well in the arborescent, as in the herbaceous species.

It also became evident, that some of the morphological characters are typical for sect. Brugmansia could be produced as pheno-typical expressions in certain members of sect. Dutra, when these were exposed to certain ecological pressures.

This may indicate, that besides the fact, that D. ceratocaula serves as a link between these species, there is also more a direct link between the species of sect. Brugmansia and sect. Dutra.

These experiences gave rise to more open-ended questions than answers in regard to a clear distinction between the herbaceous and the arborescent Datura species.

A proposal for a new, integrated structure of the four sections is:

Sect. Stramonium

Sub-group 1: Datura stramonium  L. – Datura tatula  L. – Datura bernhardtii  Lundstr. – 

Sub-group2: D. ferox  L. and D. quercifolia H.B.K.

Sect. Dutra:

Sub-group 1: Datura metel L. – Datura leichhardtii F. Muell. ex Benth. – Datura pruinosa Greenm. – 

Sub-group 2: Datura discolor Bernh. – Datura inoxia Mill. – Datura meteloides Dun. in D.C. – Datura wrightii Regel ex hort.

Sect. Ceratocaulis:

Sub-group 1: Datura ceratocaula Ort.

Sect. Brugmansia:

Sub-group 1: Datura arborea L. – Datura sanguinea Ruiz. et Pav. – Datura vulcanicola Barcl. ex Bye

Sub-group 2: Datura aurea (Lagerh.) Saff. – Datura candida Saff.  – Datura suaveolens H.B.K. – Datura versicolor L.

The distinction between the arborescent and the herbaceous Datura species

Another important question that arises in connection with the distinction between the herbaceous and the arborescent Datura species is the question of exact, what premises and definitions such distinction rest. Apart from a purely genotypical point of view, it would also be refreshing to examine a wide range of pheno-typical expressions of every species of this Genus. Both in the literature and the work in the field it has been like an unspoken and maybe even unconscious assumption, that the “genuine” morphological expression of a given Datura species is indisputable equivalent to a given expression observed in a given habitat in a given year. But this may show to be a stylized and highly idealized picture of the ever-moving Mother Nature herself. I will give an example:

Try to imagine, that You and Your friend are looking at a video movie with Marlon Brando. At a certain point in the movie, when Marlon Brando is smiling, You take a snapshot. Later on, in the movie, Your friend takes another snapshot and that, at a point, where Marlon Brando is about to get into a fight. Now, when the movie is over You start to compare the two snapshots. At this point, it is, that You starts to argue, which of You are holding the snapshot of the “genuine” morphological expression of Marlon Brando. In order to stretch this example – Who of You are in fact holding the “genuine” snapshot of Marlon Brando, The smiling or the fighting Marlon Brando? The disappointing answer is: None of You because Marlon Brando expressed himself through the entire movie, and he continued to do so in the breaks, when the movie was taken and the same phenomenon continued, when he came home to his own place as a private person, etc., etc. The snapshots are taken expressed only maybe 1/1.000 of a second of the man’s entire life. 

This is exactly, what happens if we are too eager to hold on to a stylized snapshot of the living nature. The morphological expression observed in a plant population will vary from year to year in accordance with changes in the climate. Therefore, in order to create a strong, dynamic picture of the Datura species, it is important to consider that each plant population, even though that these specimens from year to year grow from seeds of the same genetically constitution as the parent plants, have the possibilities to have a wide range of pheno-typical expressions.

The morphological differences between the herbaceous and the arborescent Datura species are not as genetically different, as they are differences due to several thousand years of adaptation to a particular ecological environment. Some of these morphological differences, as for instance the arborescent habit of the species within sect. Brugmansia, are fixed, which means that they will remain unchanged, when the plant is transferred to a different environment, while other characters, for instance, the calyx remains in certain species of sect. Dutra, will undergo remarkable changes.

Normally the species of sect. Brugmansia is characterized as arborescent, whereas the species of the other sections are determined as being herbaceous.

This was also found to be true in the specimens examined in Sorø, but truth with certain modifications. In the specimens of sect. Brugmansia only the older stems were considered to be woody, whereas most of the young root suckers in the examined species, except for D. sanguinea, remained herbaceous and at the same time carrying lots of flowers.

When the surface of the older stems was lightly scratched in with a nail, it was observed, how the thin, corky layer of the stem, which gives these species their Woody appearance, was easily rubbed of and revealed an interior herbaceous layer similar to that of the herbaceous species. The woodiness of these species is only remotely related to the woodiness one would expect from for instance oak, but very close indeed to be called herbaceous.

The same was true about some members of Sect. Dutra. The stems found in specimens belonging to the taxa D. inoxia (

DATU-38 – DATU-40 – DATU-43 – DATU-75 – DATU-87 – DATU-112), D. meteloides (DATU-44 – DATU-45 – DATU-R-45 – DATU-46-A – DATU-48 – DATU-49 – DATU-50 – DATU-76 ) and D. wrightii (DATU-82 – DATU-107 – DATU-125 ) also develops the woodiness similar to that of the arborescent species. In five-month-old specimens (DATU-38 – DATU-40 – DATU-44 – DATU-48 – DATU-82 – DATU-107 – DATU-112 – DATU-125 ) the woody layers covered the lower part of the stem. 

In the upper zone of the lower stem, this layer could easily be rubbed off as it was possible to do it with the arborescent species, but the layers near the stem basis have become thick and hard and more arborescent than observed in the older parts of the stem in a four-year-old specimen of D. aurea.

It was also observed, that the morphological habit of the stem and the branches showed a high degree of resemblance. 

Specimens of both sections developed a straight stem, which divides into two branches, and each of these further divided into two new, but longer branches, which developed several short flowering shootings. 

The morphological behavior of the root suckers, as we know it from the arborescent species could also be observed in D. metalloids (ALL ACCESSION NUMBERS ) and in D. wrightii (ALL ACCESSION NUMBERS ). In some instances, these suckers grew larger, than the parent plants, before the development of the floral branches. At the end of the 

season, the roots of some of the parent plants were lifted out of the soil and examined. The suckers had developed an entire root system of their own and in some specimens, the root system had become independent of the roots of the parent plant.

This was considered as an important discovery because one of the differences between the arborescent species and the herbaceous species have been, that the species of sect. Brugmansia could be propagated from both seeds and large cuttings, whereas the herbaceous species could only be propagated from seeds. It has been attempted several times to root cuttings taken from the lower part of the stem of D. meteloides (DATU-44 – DATU-R-45 – DATU-50 ) and of D. wrightii (DATU-82 – DATU-107 ). All attempts, however, sooner or later failed. But still, it can now be recognized, that there is an alternative to see the propagation of two of the herbaceous Datura species and a possibility too, by the aid of root suckers, produce new plants, which are genetically identical to the parents.

In the flowering specimens of sect. Brugmansia it was observed, that short, flowering shootings were developed in the main stem. The same morphological behavior could be observed in some herbaceous species, however not the species of sect. Dutra, except for D. discolor, but largely in all species of sect. Stramonium. D. tatula (DATU-18 – DATU-20 – DATU-21 – DATU-60 – DATU-63) ) and D. quercifolia (DATU-36 ) was the most outspoken species in this regard.

Two characters have to be discussed in connection with the differences between the herbaceous and arborescent Datura species in regard to the calyx.

D. ceratocaula is, as previously mentioned, considered to be a link connection sect. Brugmansia to the species of the other sections, because of its like-like calyx, the berry-like fruit, and because there is found no remains of the calyx in the fruits.

The berry-like calyx apparently connect D. ceratocaula to D. arborea, D. versicolor, D. vulcanicola, D. candida (D. aurea x D. versicolor. D x flava (D. arborea x D. sanguinea) and D. suaveolens x versicolor.

The absence of calyx remains in the fruits of D. ceratocaula links this species to D. arborea, D. aurea, D. suaveolens, D. versicolor, D. vulcanicola, D. candida (D. aurea x D. versicolor), and D. aurea x D. suaveolens, which have a similar morphological expression.

In regard to the fruit, there are certain similarities, but also some differences. D. ceratocaula, as well as the arborescent species have fruits, which are more like berries, than capsules. The fruits are glabrous and without any spines. But the fruit in D. ceratocaula breaks up irregularly in large fragments, which curls up and falls apart. In the fruits of the arborescent species, this mechanism is entirely missing.

In regard to the calyx certain species of sect. Dutra also seems to form a link to some of the species of sect. Brugmansia. Some specimens of D. meteloides (DATU-44 – DATU-45 – DATU-46-A – DATU-46-B – DATU-48 – DATU-58 ) also form like-like calyxes, which links these forms to the arborescent species mentioned the connection with D. ceratocaula.

The observations of like-like calyxes in D. meteloides was at first assumed to belong to the morphological behavior provoked by environmental stress, which caused the calyx teeth to separate incompletely, but these specimens continued to form these characteristic calyxes, when exposed to different growth environments.

In some of the arborescent Datura species, the fruit is partly or entirely enclosed by the calyx. A such behavior could also be observed in specimens of D. meteloides (ALL ACCESSION NUMBERS ) and D. wrightii (ALL ACCESSION NUMBERS ). This phenomenon, which is commonly associated with the arborescent species only, serves to connect the two taxa to D. sanguinea, D. insignis (D. suaveolens x D. versicolor x D. suaveolens), D. suaveolens x D. versicolor and D. candida (D. aurea x D. versicolor) var. Culebra (also described as a separate Genus,  Methysticodendron amesianum).

The corollas of sect. Brugmansia seems to be morphological identical to the flowers of the other sections, in some species, however, these are quite much larger.

The corollas in the herbaceous species are erect to nodding (D. meteloides and D. metel), these rarely exceed a length of 25 cm (DATU-50 – DATU-82 – DATU-107 ), whereas the flowers in the arborescent species are nodding to pendulous and in some species, the flower can reach a length of + 55 cm (D. versicolor).

A significant difference was observed in the duration of the flowers. In the herbaceous species, the corolla open early in the evening and closes again the following morning. In most of the examined specimens of sect. Brugmansia, the corollas remained open in 4-7 days.

The shortest duration was in D. sanguinea, where the corollas generally remained open and intact in 3-4 days and the longest duration was observed in a specimen of D. aurea, where a duration period of 7-8 days were not uncommon.

Scroll til toppen