The history and identity of
Boesenbergia longiflora
(Zingiberaceae) and descriptions of five related new taxa
John D. Mood
1
, Linda M. Prince
2
, J.F. Veldkamp
3
and Santanu Dey
4
1
Lyon Arboretum, University of Hawaii, 3860 Manoa Road,
Honolulu, HI 96822, USA
zingiber@warmlava.com
2
The Field Museum, Department of Botany,
1400 S Lake Shore Dr., Chicago, IL 60605, USA
3
National Herbarium of The Netherlands, Naturalis Biodiversity Center,
P.O. Box 9514, 2300 RA Leiden, The Netherlands
4
Department of Botany, Nagaland University, Lumami 798627,
Zunheboto, Nagaland, India
ABSTRACT. The history of
Boesenbergia longiflora (Wall.) Kuntze (Zingiberaceae) is
reviewed, its identity is discussed and a lectotype designated. Five new, related taxa are
described and illustrated: B. collinsii Mood & L.M.Prince, B. hamiltonii Mood, S.Dey &
L.M.Prince, B. kerrii Mood, L.M.Prince & Triboun, B. kingii Mood & L.M.Prince, and B.
maxwellii Mood, L.M.Prince & Triboun. A phylogenetic analysis of plastid
trnK intron
(including matK) and nuclear ITS DNA sequence data indicate these Boesenbergia species
form a clade within Boesenbergia. Results of the molecular data analyses in concert with
several diagnostic characters, support the recognition of the new taxa. Historical illustrations,
colour plates, a field key to the species, a comparative table, a listing of the botanical history of
B. longiflora, and a geographical distributional map are provided.
Keywords. Burma, Curcumorpha, Gastrochilus, India, Thailand
Introduction
Boesenbergia longiflora (Wall.) Kuntze was described in 1829 as Gastrochilus
longiflora by Nathaniel Wallich based on his collection from Rangoon, Burma [Wallich
6589A (K-W)]. Boesenbergia was proposed in 1891 when Otto Kuntze observed that
Gastrochilus Wall. is a homonym of Gastrochilus D.Don (1825; Orchidaceae). In
1974 it was regarded as belonging to the monotypic genus Curcumorpha A.S.Rao &
D.M.Verma. This change was countered by Larsen (1997) but is still in use by some
(Kress et al. 2002, Gao et al. 2004).
During a botanical survey in the Lampang Province, Thailand in 1998, numerous
plants which at first were identified as B. longiflora were observed growing on the
margins of secondary forest. These were easily recognised by their robust stature and
large, white flowers. Later, another Boesenbergia, reminiscent of B. longiflora, but with
yellow flowers, was encountered on a limestone outcrop [Mood & Pedersen 1455 (C)].
47
Gardens’ Bulletin Singapore 65(1): 47–95. 2013
Its specific identity remained unclear. Several years later, yet another
Boesenbergia
with yellow flowers was collected by Mark Collins (1963-2011) from a limestone
area in Loei Province, Thailand. Living plants of this field collection (T-3068) were
later determined to be similar to the earlier Lampang collection. Identification of the
two collections began with a review of Wallich’s description of
G. longiflora which
was described as having small, yellow flowers with red markings on the labellum
(Fig. 1). This information was very surprising, as it was inconsistent with the current
perception of
B. longiflora being a ginger with large, white flowers with pink markings
on the labellum (Wu & Larsen 2000; Kress et al. 2002; Gao et al. 2004, 2005; Larsen
& Larsen 2006). This significant discrepancy precipitated a complete review of the
history of
B. longiflora (Appendix A).
The historical findings indicated that the true B. longiflora (Wall.) Kuntze was
very similar to the two aforementioned Thailand collections. The current perception of
a white-flowered ginger appears to have been fostered by a watercolour published by
Hooker in 1843 (Fig. 2). Based on a reference by Baker (1890), the plant Hooker used
for this illustration was sent to Kew from the Calcutta Botanical Gardens (Wallich
1840), which was identified with the unpublished name, G. jenkinsii. Although Baker
(1890) mentioned this error, he described white flowers, as did Schumann (1904).
Holttum (1950) also noted the discrepancy between the Hooker illustration and
Wallich’s, but did not pursue the matter. Rao and Verma (1974) described their C.
longiflora as white-flowered, but in this case, it was based on 20 specimens from
Assam and Meghalaya, India, the area where Wallich found G. jenkinsii (Voigt 1845).
Although the historical information improved the taxonomic understanding of
B. longiflora, several taxa remained unconfirmed or unidentified: Wallich’s G. jenkinsii
from Assam (Fig. 2); the large white-flowered Boesenbergia common to N. Thailand;
the specimens cited by Rao and Verma; and finally, the two yellow-flowered Thailand
collections. In order to properly identify these and other specimens, the present
research was initiated in 2008 using an integrated taxonomic/phylogenetic approach.
Materials and methods
Taxonomy. Historical references, both text and illustrations (Appendix A) were used to
provide an understanding of past taxonomic perceptions. Also, over 200 digital images
of
B. aff. longiflora obtained from various sources were reviewed and categorised by
flower morphology and geographic location, providing early clues on diversity and
phytogeography. Herbarium specimens (75) labelled as
B. longiflora or C. longiflora
were examined and photographed at ASSAM, BK, BKF and CAL in 2009. Another
75 specimens were viewed via digital images from AAU, C, E, K, L, P, SING and
US. This exercise provided an important finding—the B. longiflora complex was
extremely difficult to sort out and distinguish one taxon from the other in the herbarium.
Vegetative characters often used to determine the identity of dried specimens in
other Zingiberaceae were not particularly useful here. Additionally, flowers did not
preserve well, changed colour upon pressing and became indistinguishable by shape.
48
Gard. Bull. Singapore 65(1) 2013