Post #25 Pelargonium aff. dasycaule Haw.

The section Otidia encompasses ca. 16 species, of which P. dasycaule is arguably the least known. It has been debated for a long time whether it should be recognised as a taxon at species level at all, as it wasn’t clear how it differed from P. ceratophyllum. This was compounded by the fact that the last observation of what appeared to be this species was in 1977. Furthermore, the location of this observation remains unknown and no herbarium records seem to exist. It is not surprising therefore that in 1990, Vorster concluded that the taxon should be treated as synonymous with P. ceratophyllum [1].

On the other hand, Becker, in his Thesis [2], provides DNA evidence, on the basis of specimens collected in 1977, which indicated that the taxon is separate from two other most closely related taxa, P. ceratophyllum and P. albersii. Despite this, it has remained impossible to resolve the status of this taxon without a natural population that would match the literature. This contribution attempts to rectify the issue.

Historical overview

Pelargonium dasycaule Haw. was published in 1812 [3]. The early history of the taxon is riddled with uncertainties, as is so often the case with the genus, however, we are helped by Clifton’s meticulous research [4], which much of this paragraph is based on. Haworth’s description is unfortunately accompanied by neither a herbarium specimen nor a drawing, and even if a herbarium specimen once existed (Haworth was known to habitually press his plants), it did not survive. Since the description cannot be linked to a type, Vorster [1] and later Becker [2] used Sims’s name P. dasycaulon published in 1819 [5], which was accompanied by the drawing reproduced in Fig. 1. The fact that Sims’s description of the taxon was largely based on Haworth’s is used by Clifton to argue that Sims’s illustration should be used as the iconotype for Haworth’s name; however, it is unclear whether the plant illustrated by Sims was obtained from Haworth and we may never know. Where these specimens originated (there could have been several, even from several collectors), also appears to be unclear as the two collectors who operated in the Cape in the late 18^th century, Hove and Masson [4], supplied plants to Kew and diverse nurseries. Given the uncertainties, and the absence of a neotype, it would seem proper to use the original name P. dasycaule Haw. for the time being [6].

Vorster, in his 1990 review of the section, established that Sims’s illustration “strongly suggests that P. dasycaulon is identical to P. ceratophyllum”. As we established in the last Newsletter [7], Vorster’s definition of P. ceratophyllum in 1990 was as illustrated in Pelargoniums of Southern Africa vol. 2 [8], which is what we now call P. albersii. No justification for his statement is provided either, and a comparison of Sims’s (Fig. 1) and Ward-Hilhorst’s [8] illustrations clearly reveal very different plants, one with thicker vertically growing stems and an upright, dichotomously branched inflorescence, and the other one with thin decumbent stems and a sparsely branched inflorescence [8].

Becker’s “P. dasycaulon”

Becker based his description on Sims’s illustration and a plant from the Stellenbosch Botanical Gardens (STEU 1404), which could be traced to the last known presumed collection of P. dasycaule, by Pieter Drijfhout (2134). Both can be compared in Fig. 1.

Unfortunately, we do not have the collection data for this specimen, although Clifton [4] very helpfully lists Drijfhout’s collections made just prior and just after his specimen 2134: (i) 2094 was collected on 19 March 1977 opposite Kolmanskop near Lüderitz; and (ii) 2170 was gathered on 23 March 1977, also at Lüderitz. It is thus highly likely that Drijfhout’s collection 2134 originated at Lüderitz or its vicinity and was collected between 19 and 23 March 1977.

However, the immediate vicinity of Lüderitz [7] is populated by P. ceratophyllum and it would be unusual for two very closely related taxa, flowering at the same time, to share the same habitat, as the available pollinators are sparse and would be likely shared. It is thus plausible that Drijfhout collected 2134 slightly further away from Lüderitz but not very far, given that he collected in the vicinity in the days prior and in the days after 2134 was collected.

Fig. 1: Left: Sims’s illustration of P. dasycaulon published in 1818 [5]. Right: Drijfhout 2134 (STEU 1404) in cultivation, reproduced from [2] (scale unclear).

In any case, Becker assumed that STEU 1404 was the originally collected plant, and not its potentially hybridized offspring. Sadly, recent communication with the Garden curator revealed that this plant has since been lost [9].

In Fig. 1, the depicted specimens compare reasonably well, which is how Becker justified that Drijfhout 2134 is identical to Sims’s P. dasycaulon. Becker compares the characters of what he calls P. dasycaulon with those of ceratophyllum, as most previous authors (including Vorster [1]), treated these as synonymous. Clifton seemed to confuse P. ceratophyllum, with its upright growth exhibited in cultivation, with P. dasycaule [4]. Therefore, it would have been helpful if in his publication, Becker compared albersii with both ceratophyllum and his understanding of dasycaulon as prior to his description of P. albersii [10], the characters of ceratophyllum and albersii were conflated by most authors.

Fig. 2: DNA evidence [2] showing P. dasycaule as part of a trio of species, well separated from “further taxa of section Otidia” (German: “weitere Taxa der Sekt. Otidia”).

Becker points out the small, succulent stem, tripinnate laminas with cylindrical pinnulets and narrow, pointed petals. However, it is the inflorescence that is characteristic for his understanding of P. dasycaule: it is vertical and dichotomously branched, with pseudo-umbels with often more than 4 flowers, as opposed to the short and barely branched inflorescence of ceratophyllum, with 3-5 (-8 in albersii) pseudo-umbels, with up to 4 flowers. Although Becker was aware of the importance of root morphology in the section, he does not mention the presence or absence of tubers in his P. dasycaule, unfortunately.

Further evidence for the separation of P. dasycaule from P. ceratophyllum and P. albersii is presented through DNA analysis (for details, see [2]), Fig. 2. P. dasycaule clearly stands apart from both closely related species.

Becker’s hypothesis about the plausible habitat of P. dasycaule is interesting. Based on the morphological differences, he assumes that in contrast to albersii, it is likely that dasycaule occupies rocky outcrops.

Taxonomic status of P. dasycaule Haw.

On the basis of the data reviewed above, it is thus impossible to make a clear link between P. dasycaule Haw. and any live material for the following reasons:

• Haworth’s description offers no unique character that could not be used to describe also P. ceratophyllum or P. albersii,
• Sims’s description, while based on and referring to Haworth’s, cannot be unequivocally tracked to the plants Haworth used to describe dasycaule, hence his illustration may be of a different taxon altogether,
• Sims’s illustration and the Drijfhout 2134 (STEU 1404) specimen appear to be similar but there are also important differences: the latter has shorter petals and more flowers per inflorescence than the former. It is unclear if either has tubers and their natural growth habit is unknown.

Even if Drijfhout 2134 is indeed a separate taxon as indicated by the DNA data presented in Fig. 2, and even if populations are located close to Lüderitz that turn out to resemble Drijfhout 2134, it would appear impossible to use the name P. dasycaule Haw. for these.

Given the historical uncertainties, the latter name would best be consigned to history as an insufficiently known species. I will therefore consistently use P. aff. dasycaule Haw. in the text to follow.

P. aff. dasycaule near Lüderitz

In 2019, when exploring the mountains inland from Lüderitz, I came across an unusual population of otidias at Haalenberg, part of a group of inselbergs that harbour many locally endemic taxa, e.g. Eriospermum haalenbergense, Conophytum haalenbergense etc. The vast expanses of Southern Namib around these mountains are largely devoid of vegetation, however, the steep slopes seem to be able to catch some condensation from the fogs rolling in from the Atlantic and provide some shelter from the hot sun and strong winds.

Fig. 3: P. aff. dasycaule in habitat, 2019. Left: plant habit, the remains of last year’s inflorescences are clearly visible, right: main stem showing points of re-growth and branching possibly due to grazing, as well as clearly visible round tubers.

Fig. 4: P. ceratophyllum (GER485) in cultivation, clearly developing a tuberous root.

A small population of otidias took shelter in situations with firm ground (i.e. not sand, as preferred by albersii), and sheltered by small rocks (Fig. 3). The plants are small, no taller than 10 cm when not in flower, much branched and although one would be tempted to dismiss them as albersii, the remains of previous-year’s peduncles and bi- to tri-pinnate leaves piqued my interest. 2019 was a dry year, at the end of an extended drought that lasted for a decade, and the leaves looked compact and silvery, the dense hairs evidently providing some shielding from the constant scorching sun.

On carefully digging up a plant (later re-planted), an unexpected character was revealed: the plants had small, round tubers (Fig. 3). While this is not a character usually associated with P. ceratophyllum, the latter can also develop tuberous roots, as evident in Fig. 4 (new observation since [7]). The three taxa can thus not be separated on the basis of absence or presence of tubers. Other Namibian otidias, also characterised by tubers, are P. parviflorum ssp. tuberculum, however, this is an altogether different taxon, with yellowish petals shorter than sepals.

A similar Namibian taxon, P. parviflorum ssp. karasbergense has a long thick taproot, while P. dasyphyllum, a Northern Namaqualand species, and P. alternans, with a large distribution from Namaqualand to Klein Karoo, of which some populations have also been observed to have round tubers [11], do not have bi- to tri-pinnate succulent leaves.

Comparison of characters of P. aff. dasycaule, P. albersii and P. ceratophyllum

Nothing is known about P. aff. dasycaule in cultivation, and we only have a single cultivated Drijfhout 2134 for comparison (Fig. 1), which exhibited vertical growth and stems about as wide as the flowers, i.e. ca. 10-14 mm. Drijfhout 2134 was unbranched, whereas habitat plants are branched at base, Fig. 3 (right) showing possible signs of grazing and regrowth. It is likely that habitat plants remain low-growing, and exhibit upright growth only in cultivation, similarly to P. ceratophyllum [7].

The leaves of the three species (Fig. 5) are remarkably similar and are also similar to P. crithmifolium and P. paniculatum in that they are succulent to the point of almost being cylindrical, and have a central groove. The rest of the species in the section have less succulent leaves, although those of P. parviflorum and P. brevipetalum can develop the same character if drought-stressed, when the pinnae margins roll in. However, this is reversed when the growing conditions normalise.

The leaf indumentum of the three species is also the same, consisting of adpressed non-glandular hairs: (0.06-)0.07-0.13(-0.16) mm interspersed with extremely short glandular hairs. This is at odds with Becker’s claim that P. albersii only has short glandular hairs [10]. Similarly, the epidermises show similar irregular concave cells in all three species, and that of P. albersii is no different.

Fig. 5: Leaves of A – P. albersii [GER245], B – P. ceratophyllum [GER485], and C – P. aff. dasycaule [GER795]. Above: the laminas of the latter are markedly more divided, however, the leaves of P. albersii can also be tripinnatisect. The leaves of P. ceratophyllum are significantly more succulent than the other two. Below: the leaf epidermises and indumentum.

 

The pedicel of aff. dasycaule, (6.0-)6.5-8.5(-10.0) mm, is typically slightly longer than the hypanthium, (5.50‑)5.75-7.25(-8.00) mm (Fig. 6), whereas the pedicel of ceratophyllum and albersii can be much longer, up to 20 mm [2]. The ratio of hypanthium to pedicel length is about 1, typically 0.75-1.20, whereas for P. albersii this is 0.5-0.9 (GER170, GER245).

Fig. 6: Hypanthium and pedicel lengths, and their ratio (H/P) for P. aff. dasycaule (N = 8) and for P. albersii (N = 7). (The box min and max represent 25% and 75% of the measurements, the whiskers represent 1.5x of the interquartile range and the black points represent the outliers. The horizontal line within a box represents the median and the hollow square the mean.)

 

Becker’s photograph [2] (Fig. 7) shows off-white, linear petals with weak markings, slightly shorter than tepals. The petals of GER795 are yellowish and are slightly longer than the tepals. The posterior petals are (8.5-)9.0-10.0(-10.5) x (2.3-)2.6-3.0(-3.5) mm and have weak feather-like markings with densely hairy ears (Fig. 8), while the anterior petals are (7.6-)8.0-9.0(-9.5) x (1.8-)2.0-2.8(-3.2) mm and are unmarked. The tepals are reflexed, 7.0-8.0 x 2.5-3.5 mm. Fertile stamens are 5, the staminodes are 7-9 mm long. The pollen is orange. The flowers have a perceptible odour.

Fig. 7: Pseudo-umbels of GER795 (left) and Drijfhout 2134 (i.e. STEU1404, right). Scale: 2 cm.

Fig. 8: Hairs on the auricle of a posterior petal of P. aff. dasycaule.

There are populations of two Otidia taxa at Haalenberg: (i) P. aff. anauris, and (ii) P. aff. dasycaule. These grow about 100 m apart, and it is unclear if hybridisation occurs given their respective flowering times. Plants of what appears to be a viable population of P. aff. dasycaule (ca. 50 plants observed) exhibit characters that set them somewhat apart from P. ceratophyllum and P. albersii (Tab. 1).

Fig. 9: The upright growing habit of P. aff. dasycaule (GER795) with terminally borne, repeatedly branched inflorescences with up to 6 flowers per pseudo-umbel.

Tab. 1: Characters of P. aff. dasycaule, P. ceratophyllum and P. albersii, for the latter two adapted after [6], based on observations of GER245 and GER485, respectively.

	P. aff. dasycaule (GER795)	P. ceratophyllum	P. albersii
Habit	Much-branched upright shrublet up to 20(‑30) cm tall.	Much-branched shrublet to 20(-30) cm tall.	Sparsely branched decumbent shrublet to 10(-20) cm tall.
Roots	Tubers present.
Stems	6-8(-10) mm, with persistent remains of petioles and semi-persistent remains of peduncles.	(in cultivation: 10-) 20-30 mm, with persistent remains of petioles.	(in cultivation: 3-)5-10 mm, with persistent remains of petioles.
Stipules	Triangular.
Leaves	Tri-pinnatisect, succulent. Pinnulets cylindrical, terminal pinnulets often incised at the end. Petiole, rachis and pinnulets with a central groove. Lamina (15-)17-45(‑65) x (20-)33-75(-90) mm, petiole (10‑)15-30(-50) mm. Indumentum: short glandular and recurved non-glandular hairs, (0.06-)0.07-0.13(-0.16) mm.	Bi- or tripinnatisect, succulent. Pinnules cylindrical, terminal pinnules often incised at the end. Petiole, rachis and pinnules with a central groove. Lamina (10-)20-40(-60) x 10-20(-40) mm, petiole up to 40 mm. Indumentum: short glandular and recurved non-glandular hairs, (0.06-)0.07-0.13(-0.16) mm.
Inflorescence	Borne terminally on branches, central rachis red, vertical, 2-3.5(‑4) mm at base, repeatedly dichotomously branched with pseudo-umbels with (2-)3-5(-6) flowers. Peduncle 30-50(-60) mm.	Borne terminally on branches, peduncle and central rachis red, thin (<2 mm), sparsely branched with up to 8 pseudo-umbels with up to 4 flowers.
Pedicel and hypanthium	Similar in length, pedicel 6-8(‑9) mm, hypanthium (5.5-)6-8(‑10) mm.	Similar in length, 4-15(-20) mm.
Petals	5, linear, ovate, auricled, white to yellowish. Ears hairy. As long as or longer than tepals. Posterior petals with feather-like wine-red markings.	5, linear, ovate, auricled, white in P. ceratophyllum or white to yellowish in P. albersii. Ears hairy. As long as, but typically longer than tepals. Posterior petals with feather-like wine-red markings.
Scent	Perceptible.	Imperceptible.	Perceptible.

P. aff. dasycaule has an upright habit, consistently tripinnatisect leaves, a thicker, semi-persistent, repeatedly dichotomously branched inflorescence rachis, pseudo-umbels of up to 6 flowers with shorter pedicels. Further populations of what appear to be the same taxon were reported from the Sperrgebiet in Southern Namibia in September 2023 [12].

If GER795 is indeed the same as Drijfhout 2134 (STEU1404), for which we currently lack evidence [given the differences in the habits of GER795 (Fig. 9) and Drijfhout 2134 (Fig. 1)], there appears to be sufficient DNA data (Fig. 2) to separate this taxon from P. ceratophyllum and P. albersii. On the other hand, morphologically, P. aff. dasycaule could potentially be seen as an extreme form of P. albersii.

Conclusion

After 42 years, a viable population of what appears to be P. dasycaule has been found close to Lüderitz and an herbarium specimen deposited at the NBG. The plants exhibit hitherto unknown characters, such as tuberous roots as well as vertical, dichotomously branched, semi-persistent inflorescence, which separates them from both P. ceratophyllum and P. albersii. These three species with succulent, bi- and tri-pinnate leaves and cylindrical pinnules (pinnulets) with a central groove, seem to be genetically closely related, although distinct.

However, since there is no clear link between dasycaule Haw., dasycaulon (Haw.) Sims., Drijfhout 2134 (STEU 1404) and the population discussed here, the latter can be only tentatively identified as aff. dasycaule Haw.

This note has brought P. dasycaule Haw. a bit closer, however otidias at the two extremes of the geographical distribution of the section remain a unclear. We will have a closer look at P. parviflorum ssp. karasbergense, broadly distributed in small, scattered populations in the summer-rainfall mountains of Southern Namibia in one of the future Newsletters. P. anauris is another otidia from the winter rainfall area of Namibia that is less well researched.

In addition, P. parviflorum-type populations in the summer-rainfall areas of the South-eastern Karoo also remain unclear and present some interesting combinations of characters that may need a closer examination in the future.

By Matija Strlič, Ljubljana, Slovenia

Specimens examined

Tab. 2: Live Geraniaceae Reference Collection (GER) material examined.

Taxon	Grid ref.	Location	GER code
P. aff. dasycaule	2516CB	Haalenberg (Namibia)	GER795
P. albersii	2816DA	Beauvallon (RSA)	GER170, GER245
P. ceratophyllum	2615CA	Lüderitz (Namibia)	GER485

 

Literature

1. Vorster P. J. (1990), Taxonomy of the genus Pelargonium: Review of the section Otidia. Proceedings of the International Geraniaceae Symposium, University of Stellenbosch, 24-26 September 1990, reprinted by the Geraniaceae Group, 1994, pp. 279-294.
2. Becker (2006), Revision der Pelargonium – Sektion Otidia (Geraniaceae) aus dem Winterregengebiet des südlichen Afrikas und Bewertung evolutiver Strategien der Pelargonien aus der Capensis. PhD Thesis, University of Münster.
3. Haworth A.H. (1812), Pelargonium dasycaule. Synopsis plantarum succulentarum, London, p. 309.
4. Clifton R. (2014), Pelargonium section Otidia, Typification and Verification Notes for ceratophyllum, P. albersii, P. dasycaule, P. dasycaulon and P. hospitans. Geraniaceae Group Associated Notes: no. 75.
5. Sims J. (1818), Pelargonium dasycaulon. Curtis’s Botanical Magazine, London, vol. 46, t. 2029.
6. Manning J. (2023), personal communication, 10.10.
7. Strlič M. (2023), P. ceratophyllum and P. albersii – Two Lesser-Known Sister Otidia Species. Geraniaceae Group News #170, pp. 16-28.
8. Van der Walt J. J. A., Vorster P. J. (1981), Pelargonium crithmifolium. Pelargoniums of Southern Africa, Juta, Cape Town, vol. 2, pp. 27-29.
9. Kirkwood D. (2023), personal communication, 27.9.
10. Becker M. (2008), Pelargonium albersii, a new species of Pelargonium section Otidia (Geraniaceae). Schumannia 5, 157-168.
11. Strlič M. (2023), Observations of tuberous P. alternans. Geraniaceae Group News #169 (2023), pp 28-31.
12. McCleland W. (2023), iNaturalist observations: https://www.inaturalist.org/observations/194648226 and https://www.inaturalist.org/observations/194089386 (accessed 26/12/2023).

Citation and Copyright

© The Author. This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

ISSN 2464-014X.

This article was first published in Geraniaceae Group News #171. Cite as: M. Strlič: Pelargonium aff. dasycaule Haw. Geraniaceae Group News #171 (2023), pp 17-32.