The specific epithet is derived from Kemumu, a village
near the locality where this plant was collected in Northern
Bengkulu, Sumatra, Indonesia.
Susatya A, Hidayati SN, Riki S. 2017. Rafflesia kemumu (Rafflesiaceae), a new species from Northern Bengkulu, Sumatra, Indonesia. Phytotaxa 326:211-220.
These authors have named what they consider a new species of Rafflesia from the Palak Siring area of Northern Bengkulu. Two other populations, from the Kuro Tidur area (40 km away) and the Ipuh Production Forest (150 km away) are also considered this taxon, however, they admit that the latter site may contain R. gadutensis instead (as identified by Willem Meijer 1997). Having reviewed the manuscript for Phytotaxa and read the final published manuscript, it is my opinion that the taxon being described is synonymous with R. gadutensis. Although Susatya et al. (2017) provide what they perceive to be species-level differences between R. kemumu and R. gadutensis, they have either purposely or erroneously misrepresented the range of morphological features seen in R. gadutensis. Moreover, they also misrepresent some morphological features in their “new” species where the photograph disagrees with the text description. Indeed, they have described, sometimes in excruciating detail (e.g. the ramenta) what they have observed on flowers from these populations. But what seems missing here is the concept of phenotypic plasticity and polymorphism that is known to occur in Rafflesia. Seven "key differences" between these species are given in Table 1 of Susatya et al. (2017). The following is a critique of the information in that Table.
• The color of the upper perigone lobe surface for R. gadutensis is given as "maroon red" whereas that of R. kemumu is "orange to dark orange". It is incorrect to characterize the R. gadutensis perigone as only red-maroon. Various flowers can be seen on PhytoImages HERE and one can see that the flower color is often more orangish, essentially identical to Figs. 1B and 1C of R. kemumu. It is highly likely that flower color is polymorphic among various populations.
• Upper surface of the diaphragm. For R. gadutensis it is given as "pinkish red, warts present" and for R. kemumu as "light orange, wart absent, ring of circular depressions present". I consider these differences very minor and much less than variants seen in, for example, R. lobata that can have an entire diaphragm or a deeply lobed diaphragm. Although the authors indicate that warts are not present in R. kemumu, close examination of their Fig. 2B shows that indeed there are warts present, albeit not raised as much as in typical R. gadutensis.
• Windows. Rafflesia gadutensis has a window of five concentric series, two series closest to the diaphragm opening, very often consisting of merged or long rectangular blots. For R. kemumu, four series, two series closest to the opening, consisting of circular blots distant to each other. The windows, otherwise known as white blots in Meijer (1984), that are present near the diaphragm opening in both taxa are in two series. The main difference appears to be the degree to which the windows merge with each other. I consider this a feature that could be variable across individuals of one species.
• Ramenta. Susatya et al. (2017) claim that the "toadstool" type is absent in the upper part of the perigone tube in R. gadutensis but present in R. kemumu. Quoting from the description of R. gadutensis in Meijer (1997, p. 24, Flora Malesiana): “Between the ramenta on the inner side of the perigone tube some scattered toadstool-like protuberances close to the inside of the diaphragm, but with only 1-2 mm long stalks.” The part of the perigone tube near the diaphragm is the upper part, so this contradicts Susatya et al. (2017).
• Processes. Susatya et al. (2017) have attempted to make the number of
processes between the two taxa an important taxonomic character. For R.
gadutensis this is reported as "14-15 or 10 in outer ring, 4-5 or 7
in inner ring". For R. kemumu they report "15 in outer, 7 in
inner". This method of counting is confusing and misleading. Meijer (1984)
reports 17-30 for R. gadutensis, which clearly encompasses the
range in both taxa. Furthermore, Susatya et al. (2017) claim that the apex
of each process in R. gadutensis is flat and truncate whereas in
R. kemumu it is conical and rounded. But looking at this photo of
processes from R. gadutensis HERE,
one can see that
both flat and conical processes are present in the same flower, as are
truncate and rounded.
• Number of anthers. Meijer (1984) reports "about 30" for R. gadutensis whereas Susatya et al. (2017) give "26", which in my mind is "about 30".
Taken together, it can be seen that many of the "key" features that supposedly distinguish R. kemumu from R. gadutensis do not stand up to scrutiny. They are either encompassed by the variation already reported for R. gadutensis or they misrepresent the range of characters seen in that species. Minor variation in flowers of "R. kemumu" that have not been reported in R. gadutensis only serve to extend the range of variation in the latter species. This is comparable to the Philippine taxon "R. banoana" that is simply a large-flowered form of R. leonardi as explained in Barcelona et al. (2011). It should be borne in mind that R. gadutensis is an extremely rare species that may be nearing extinction owing to forest destruction and trampling (Jeremy Holden, personal comm. January 2018). Because of this rarity, gene flow may be restricted among populations on the island of Sumatra, thereby magnifying local populational differences.
Literature cited
Barcelona,
J. F., E. S. Fernando D. L. Nickrent, D. S. Balete, and P. B.
Pelser. 2011. An amended description of Rafflesia
leonardi and
a revised key to Philippine Rafflesia(Rafflesiaceae).
Phytotaxa 24: 11-18.
Meijer
W. 1984. New species of Rafflesia (Rafflesiaceae). Blumea 30: 209-215.
Meijer W. 1997. Rafflesiaceae. Pages 1-42 in Kalkman C, Kirkup DW, Nooteboom HP, Stevens PF, de Wilde WJJO, eds. Flora Malesiana, vol. 13. Leiden: Rijksherbarium.
