R E I N W A R D T I A

Published by Herbarium Bogoriense, Bogor, Indonesia
Vol. 8, Part 2, pp. 319 — 321 (1972)

THE TAXONOMIC POSITION OF PAPUODENDRON
C.T. WHITE AS ELUCIDATED BY ANATOMICAL CHARACTERS

W. A. VAN HEEL

Rijksherbarium, Leyden, Netherlands ,

When, in 1946, White described Papuodendron lepidotum from
New Guinea, he apparently hesitated to incorporate this species in the
Bombacaceae. He considered that whereas on the one hand the congestion
of the stamens on top of the stamen tube pointed to Bombacaceae, on the
other hand the absence, of tile cells in the rays indicated the Hibisceae.
Van Steenis (1947), in an enumeration, listed Papuodendron under the
Bombacaceae. However, Kostermans (1960) decided that Papuodendron
was Malvaceous, that it most probably belonged in Hibiscus, with affinity
to a small group of New Guinean species, namely Hibiscus carrii,
H. womersleyanus and H. pulvinulifer, which were all described by van
Borssum Waalkes (1956). It was the opinion of Kostermans (1960) that
the congestion of the stamens on top of the tube was insufficient for
separating Papuodendron from Hibiscus, in which genus the stamens are
spread along the distal part of the tube. Accordingly Kostermans reduced
Papuodendron to Hibiscus, and renamed Papuodendron lepidotum
White as Hibiscus papuodendron Kosterm. In addition Kostermans
described a second 'Papuodendron' species, Hibiscus hooglandianus.
The discussion of van Borssum Waalkes (1966) on the problem appears
contradictory. Whereas on the one hand he claimed that Papuodendron
is at any rate a link between the Hibisceae and the Durioneae, he stated
on the other hand that Papuodendron resembles very much Hibiscus
pidvinulifer. He rejected the names of Kostermans, reestablishing
Papuodendron, and did not treat the genus in his revision of Malesian
Malvaceae.

In the following I shall bring forward a number of anatomical
arguments, in three stages, in favour of the view of Kostermans in this
matter.

— 319 —



320 R E I N W A R D T I A [VOL. 8

1. Papuodendron is Malvaceous

As I explained in 1966, the stamens in Malvaceae and many
Bombacaceae are arranged along the free distal parts of five lobes that
are fused with their bases into a tube. These staminate lobes consist of
a median and two lateral parts, each part with its own vascular bundle
supply. However, whereas in Malvaceae the, median part and its vascular
supply is reduced or almost so, in Bombacaceae this part is either
staminiferous or staminodial, and shows a corresponding vascular bundle
supply. In this respect Papuodendron follows the Malvaceous structure.

Another argument is derived from the structure of the pollen.
As noticed by several authors the pollen of Papuodendron are echinate,
as is the rule in Malvaceae. However, it is important to add that the pollen
are periporate, meaning that many simple pores are spread all over the
surface of the pollen grain. Whereas this particular structure is prevalent
in Malvaceae, it is absent in Bombacaceae. It is conceivable that together
these two pollen characters could greatly serve to delimit Malvaceae and
Bombacaceae.

2. Papuodendron belongs to the Hibisceae

Accepted as Malvaceous, the genus finds its place in the Hibisceae.
This is evident from general habit, as well as from the capsular fruits,
the free capitate stigmata, and such special features as the stamens being
divided serially, not collaterally. For these reasons Papuodendron must
be placed in the Hibisceae, preferably like redefined by Fryxell (1968).

3. Papuodendron is a Hibiscus species

In Hibiscus the, monothecous anthers show a characteristic vascular
bundle supply, consisting of a dichotomy of two sharply anatropous
branches that leave some residual tissue in the branch angle. This precise
pattern neither occurs in Bombacaceae nor in other genera of Malvaceae,
except for the dissimilar eastern-European Kitaibelia vitifolia.
Papuodendron does show this pattern.

The second argument is the resemblance of Papuodendron with
a small group of New Guinean Hibiscus species, as mentioned above.
This concerns general appearance (indumentum, inflorescence) as well
as two noteworthy special characters. Firstly all the Hibiscus spp. under
consideration belong to a part of the section Azanza that shows five
radial false septs in the ovary cells, creating the appearance of ten cells.



1972] VAN H E E L : The -position of Papuodendron 321

Papuodendron also shows these false septs. Secondly Hibiscus pulvinulifer
possesses large, profusely vascularised, glands where the petals are
attached to the stamen tube. These glands are, obscured by long surrounding
hairs. In the whole of Bombacaceae and Malvaceae these antepetalous
glands are restricted to Hibiscus pulvinulifer and Hibiscus sciadolepidus
(Hochr.)Borss., a species equally belongs to the Azanza section. However,
also Papuodendron shows this peculiar feature.

One single character remains against reducing Papuodendron to
Hibiscus, namely the position of the stamens together on the summit
of the stamen tube, instead of spread along the distal part of the tube.
Moreover the sterile extension of the tube, as present in Hibiscus,
is lacking in Papuodendron. Kostermans thought that a slight difference
in length of the filament could be of no value. To that may be added that
in young flowers of Hibiscus hooglandianus — the second 'Papuodendron'
species — I noticed sterile lobes at the end of the tube. However, I am not
certain that these young lobes will be visible in mature flowers. Also,
in these young flowers, the stamens were slightly spread along the distal
part of the tube. However that may be, more relevant is that also in
Hibiscus pulvinulifer a sterile tube ending is absent, and that the stamens
are rather closely packed along the distal part of the tube.

Thus a close affinity of Papuodendron to this group of New Guinean
Hibiscus species in the Azanza section is firmly established, especially to
Hibiscus pulvinulifer. For some common special features in the
receptacular vascular bundle pattern I may refer the reader to my paper
of 1966; these features indicate a similar way of development of that
region.

The conclusion is that the reduction of Papuodendron to Hibiscus,
as proposed by Kostermans, is fully corroborated by anatomical study.

I dedicate this paper to Dr. A. J. G. H. Kostermans, on the occasion of
his 65th anniversary. What else could the anatomist add but some security
to what is grasped at once by the able tropical botanist?

REFERENCES

BORSSUM WAALKES, J. VAN, 1956, in Reinwardtia 4, 1.
_ , 1966, in Blumea 14, 1.

FBYXELL, P. A., 1968, in Bot. Gaz. 129, 4.

HEEL, W. A. VAN, 1966, in Blumea 13, 2.
KOSTERMANS, A. J. G. H., 1960, in Reinwardtia 5, 3.
STEENIS, C. G. G. J. VAN, 1947, in J. Arnold, Arb. 28.
WHITE, C. T., 1946, in J. Arnold Arb. 27.


	hal 319-328_Page_1
	hal 319-328_Page_2
	hal 319-328_Page_3