﻿A new combination and a new synonym of Gesneriaceae in China

﻿Abstract Didymocarpussubpalmatinervis W.T.Wang was shown to be conspecific with Petrocodonlithophilus Y.M.Shui, W.H.Chen & Mich.Möller, by checking original literature, examining specimens, tracing specimen collecting history, and conducting field surveys. The results show morphological characteristics and geographical distribution overlaps between these two species. The transfer of Didymocarpussubpalmatinervis to Petrocodon as a new combination with Petrocodonsubpalmatinervis (W.T.Wang) F.Wen & Z.L.Li is proposed here, and P.lithophilus is synonymized with P.subpalmatinervis. Color photographs and essential information are also provided, including a detailed comparison of description, distribution, habitat, and the proposed IUCN conservation status.


Introduction
Didymocarpus Wall.has a complex taxonomic history, saturated with doubtful taxa (Weber and Burtt 1998).Originally, in the taxonomic system of Burtt and Wiehler (1995), it was the largest genus in the tribe Didymocarpeae Endl., subfam.Cyrtandroideae, with more than 180 species.Weber and Burtt (1998) excluded the section Heteroboea, which was considered as part of Didymocarpus s. l (Hilliard and Burtt 1995;Weber and Burtt 1998).In addition, they considered that 1) D. hancei Hemsl.(endemic to China) and D. bonii Pellegr.(distributed in Vietnam and East Thailand) should be included in Calcareoboea C.Y.Wu ex H.W.Li, as well as D. mollifolius W.T,Wang and D. niveolanosus D.Fang & W.T.Wang; 2) D. demissus Hance should be assigned to the previous genus, Chirita Buch.-Ham.ex D. Don (Wei et al. 2010;Wang et al. 2011;Weber et al. 2011).Their notion of separating D. hancei from Didymocarpus Wall.was supported by Li (2007) based on morphological and molecular evidence.Weber et al. (2011) Wang).D. subpalmatinervis W.T.Wang, a species with no collection record after 1905, also has morphological characteristics placing it in section Heteroboea.The taxonomic status of this species remained unchanged because of the difficulty in obtaining type materials.Its problematic taxonomic status has previously been recognized by (Li et al. 2015;Möller et al. 2016;Hong et al. 2018).
The French Catholic priest Francois Ducloux (1864Ducloux ( -1945)), the head of the Kunming Church from 1889 to 1945, hired people to collect plant specimens extensively in central and northern Yunnan (Qu 2014).In 1905, Ducloux collected a Gesneriaceae-like taxon, perennial and acaulescent in Y-dje, near Loulan, Yunnan, China (collection number Fr. Ducloux 3711).He sent three sheets back to the Museum of Natural History in Paris.Since then, those specimens were neglected in P for over 90 years, until in 1996, Wen-Tsai Wang identified Ducloux's specimens as a new species belonging to Didymocarpus sect.Heteroboea and published it as Didymocarpus subpalmatinervis subsequently a year later (Weitzman et al. 1997).
Petrocodon lithophilus was described at the morphological and molecular level from Naigu Stone Forest, Yunnan Province, where Michael Möller and Yu-Min Shui first discovered it in August 2012.It is placed in Petrocodon by some obvious characters of this genus, viz. a rhizomatous rosette with leaves usually rounded or elliptic, mainly in an alternate arrangement, filaments straight, stigma discoid, and fruits dehiscing loculicidally into two valves (Chen et al. 2014).
When sorting out the specimens of Didymocarpus sect.Heteroboea of, the type specimen of D. subpalmatinervis (Fr.Ducloux 3711), morphologically, was found to be extremely similar to Petrocodon lithophilus.According to protologue and label information on the type specimen, the type locality of Didymocarpus subpalmatinervis is near Lunan Stone Forest, only 13 kilometers away from Naigu Stone Forest.During field investigations near the type locality of Petrocodon lithophilus, several populations of Didymocarpus aff.subpalmatinervis were also found on the hills.The authors concluded that D. subpalmatinervis is conspecific with Petrocodon lithophilus.Accordingly, it is reasonable to make a new combination P. subpalmatinervis, and reduce P. lithophilus to a synonym.

Materials and methods
A thorough comparison of the type material of Petrocodon lithophilus and Didymocarpus subpalmatinervis was made.Their protologues and relevant records were studied intensively.The geographical distribution of the two species was outlined by careful field surveys of the type locality areas.Classical plant taxonomic methods were involved.Major online herbarium databases, including P (https://science.mnhn.fr/institution/mnhn/search),E (http://www.rbge.org.uk/),K (https://www.kew.org/),A (https://huh.harvard.edu/),PE (https:// pe.ibcas.ac.cn/index.html)and Chinese Virtual Herbarium (https://www.cvh.ac.cn/), et al., were checked.The only recorded specimens of D. subpalmatinervis stored at P were affirmed and checked, utilizing high-resolution digital images of the type specimen.In addition, the authors observed and recorded morphological characters of Petrocodon lithophilus in the field and examined its type specimen at KUN.

Comparison and discussion of morphological characteristics of sect. Heteroboea
There are numerous distinct morphological characteristics distinguishing Didymocarpus subpalmatinervis from other species within sect.Heteroboea, such as leaves' margins obtusely or doubly dentate (vs.margin irregularly triangular denticulate), triangular lobes (vs.rounded or oblong lobes), straight filaments (vs.curving), separate anthers (vs.confluent anthers).Detailed morphological comparisons with sect.Heteroboea are provided in Fig. 1.Based on the aforementioned, it becomes evident that D. subpalmatinervis is not a species within sect.Heteroboea.Additionally, owing to its stemless habit, it does not align with sect.Didymocarpus either.Consequently, the taxonomic classification of this species falls outside the scope of the Didymocarpus genus.Given its overall vegetative traits, it should be reclassified under the genus Petrocodon.

Petrocodon lithophilus
Comparing the description of Didymocarpus subpalmatinervis and Petrocodon lithophilus in their protologues, we found consistency in habit, leaves, calyx, corolla, disc, and ovary (indicated by "•" in Table 1).Because Wen-Tsai Wang described D. subpalmatinervis (Weitzman et al. 1997) according to the over 90-year-old type specimen, there may be some distortion compared with wild-living plants in morphology.Moreover, terminology may alter to a certain degree among different taxonomists due to personal preference in the use of words.Taxonomic descriptions have evolved over the years, as have identification tools.For example, Wang selected 'puberulous' and Shui picked up 'pubescent' to describe the hair of Didymocarpus subpalmatinervis and Petrocodon lithophilus, respectively.Such equivalents used by them in the description are marked with "★" in Table 1.The only significant difference appeared in the statement of staminode's number: 2 in Didymocarpus subpalmatinervis and 3 in Petrocodon lithophilus (indicated by "▲" in Table 1).This feature has already been questioned by Wen et al. (2020).Previous researchers often made mistakes in describing the number of staminodes because 1) they followed generic diagnoses of Didymocarpus and paid no attention to the actual stamineal condition; 2) staminodes are inconspicuous in many species unless they are carefully observed, but the correct number of staminodes is 3. Detailed morphological comparisons with P. lithophilus are provided in Fig. 2.
Chinese, and Y-dje corresponds to 维则 (Wéizé).The scope of Lou-lan is not indicated on the type specimens of D. subpalmatinervis.China in 1905 was still in the era of the Qing Dynasty.Lunan then refers to Lunanzhou, which is now Shilin County.Naigu Stone Forest is included in Lunan County (now Shilin County).Lunan refers to Lunan Stone Forest.Lunan Stone Forest and Naigu Stone Forest are both in the Stone Forest scenic spot today.Our field survey shows only one species of Gesneriaceae with similar morphological characteristics in these two regions.The distribution of the two species thus overlaps no matter which geographical entity Lunan refers to in the past or present.
In summary, based on literature research, geographical distribution, and morphological analysis, we find no discontinuities and recognize Didymocarpus subpalmatinervis and Petrocodon lithophilus as conspecific.Therefore, a new combination, P. subpalmatinervis (W.T.Wang) F. Wen & Z.L.Li, comb. nov. is proposed here.

Table 1 .
The description and comparison of the D. subpalmatinervis and P. lithophilus.