﻿Rosafuningensis (Rosaceae), a new species from Yunnan, China

﻿Abstract A new species Rosafuningensis and its variant R.funingensisf.rosea, both collected from Yunnan Province, China, are, for the first time, documented and illustrated in this study. Morphological analysis in comparison with two related species in the wild, R.gigantea and R.rubus, presents distinguishable features through leaf surfaces, inflorescences and the shape of styles. R.funingensis leaf surfaces are abaxially villous, purple-red, pale green when mature, adaxially glabrous, dark green; inflorescences solitary or 2–5(7) in corymbose cyme; and styles connate into a column or not, exserted.


Introduction
There are about 150-200 species of roses around the world, widely distributed throughout the Northern Hemisphere, with Central and Southwest Asia being the centres of distribution of the genus (Rehder 1951;Ku and Robertson 2003;Quest-Ritson and Quest-Ritson 2003). China has 95 species of the genus Rosa, of which 65 species are endemic (Ku and Robertson 2003); there are also 34 varieties, totalling 129 taxa of roses (Liu and Lian 2014). The number of wild rose species in China are found to decrease gradually from the southeast to the southwest and northwest of the country (Yu and Lu 1985). Yunnan is one of the main distribution centres and differentiation centres of Rosa (Xu 2001), with 41 species and 17 varieties of wild Rosa (Chen and Li 2006).
On 5 April 2018, a unique species of Rosa was discovered during an investigation of wild rose resources in Funing County, Wenshan Zhuang and Miao Autonomous Prefecture, Yunnan Province, China. It shared certain morphological characteristics with R. gigantea and R. rubus, while being distinguished in terms of leaf, inflorescence and shape of styles. Subsequently, this species was introduced to the Kunming South Tropical Garden (Kunming Nanguo Shanhua) Horticulture Technology Co. Ltd., Yunnan Province for fur-ther observation and study. After a thorough examination over a period of three years, it was determined that the specific morphological characteristics of this species and its variant were stable, indicating that they were, indeed, new to the Rosa genus. In 2021, phylogenetic relationships were analysed after collecting the specimens to confirm their status as a new species within the genus Rosa. As a result of this research, the new species was described and named as Rosa funingensis L. Luo & Y. Y. Yang. Additionally, a form of this new species was identified and documented as Rosa funingensis L. Luo & Y. Y. Yang f. rosea L. Luo & Y. Y. Yang, characterised by its light salmon-pink flowers that fade to white.

Field observations, comparative morphology
We studied living plants of the new species in their natural habitats and documented their known distribution ranges. Morphological descriptions and illustrations were based on mature foliage, fresh flowering material and mature fruit of living plants and dried specimens of R. funingensis and R. funingensis f. rosea.

Phylogenetic analysis
Sixteen taxa of the genus Rosa, including Rosa funingensis and two outgroups (Fragaria vesca and Potentilla tanacetifolia) were used to reconstruct a phylogenetic tree. Sequences of R. Chinensis 'Old Blush' (sequence number: SRR6175515), Fragaria vesca (sequence number: SRR12536045) and Potentilla tanacetifolia (sequence number: SRR8208352) were downloaded from Gen-Bank. The other 14 taxa were selected from six sections. Their complete genomic DNA was extracted from silica-gel-dried leaves using the CTAB method (Porebski et al. 1997) and sequenced using Illumina NovaSeq.
The sequenced data were quality-controlled to obtain clean data. Genome alignment was performed using MINIMAP2 v.2.21. PCR de-duplication and SNP detection were performed using gatk v.4.2.0.0. SNP annotation was performed using ANNOVAR v.2020-06-07. The processed SNP data were analysed in VCFTOOLS v.0.1.17 for Fst analysis (fst-window-size 100000 fstwindow-step 10000), thetaΠ analysis (window-pi 100000 window-pi-step 10000) and Tajima's D analysis (TajimaD 100000). The resulting data were analysed in MEGA11 (Tamura et al. 2021). The evolutionary history was inferred by using the Maximum Likelihood method and the Kimura 2-parameter model (Kimura 1980). The tree with the highest log likelihood (-9601.46) is shown. The percentage of trees in which the associated taxa clustered together is shown next to the branches. Initial tree(s) for the heuristic search were obtained automatically by applying Neighbour-Joining and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach and then selecting the topology with superior log likelihood value. There were a total of 3560 positions in the final dataset.

Phylogenetic analysis
The ML phylogenetic tree ( Fig. 1) showed that Rosa glomerata, R. soulieana (sect. synstylae) and R. kweichowensis (sect. microphyllae) and all the species of sect. chinenses formed a well-supported clade (purple), with the sister group of R. lasiosepala, R. luciae and R. rubus of sect. synstylae (orange). The putative new species, R. funingensis, is placed into a well-supported clade with R. gigantea (sect. chinenses) and formed a larger clade with sect. chinenses.

Discussion
Rosa funingensis is often in association with R. rubus and R. gigantea in the wild where there are no other members of the genus Rosa present. The overlap of the flowering period of R. rubus (late March to late April) and R. gigantea (March) generates the possibility of natural hybridisation to produce offspring.
Morphologically, Rosa funingensis shares similarities with both R. rubus and R. gigantea, but it is not exactly the same as either one, which provides further evidence that R. funingensis may be a natural hybrid. The molecular evolutionary tree also supports this inference.
The discovery of this new species enriches the resources of Rosa and provides new materials for interspecific hybridisation. Hybridisation between sections has been a great challenge in rose breeding, making wild Rosa resources not sufficiently exploited (Zhao et al. 2015). The hybridised experiment during 2020-2022 between R. funingensis and the wild species of sect. chinenses shows that it is, indeed, a good breeding material. The inclusion of R. funingensis as a new germplasm resource for breeding between sect. chinenses and sect. synstylae has the potential to enhance the genetic diversity and improve the breeding outcomes of the genus Rosa.
Additionally, during our field research, we also found plants that are similar to R. funingensis, but with smaller leaflets (7-9); stipule margin covered with sparsely glandular hairs; flowers showing light salmon-pink at the beginning and turning white at the later stage; hip obovoid. We speculate that these plants may be a form of R. funingensis, with an increased number of leaflets and this is currently under observation. Diagnosis. Rosa funingensis is mostly similar to R. gigantea. However, Rosa funingensis differs significantly from R. gigantea by having leaves abaxially villous, purple-red, pale green when mature, adaxially glabrous, dark green (vs. both surfaces glabrous), rachis and petiole shortly prickly, glandular hairs and villous (vs. sparsely shortly prickly and glandular pubescent), inflorescences solitary or 3-5(7) in corymbose cyme (vs. solitary or 2 or 3 and fasciculate) and styles connate into a column or not (vs. free) ( Table 1).
Phenology. Flowering in early April, fruiting from July to October. Etymology. The species epithet refers to Funing County, where the new species was first discovered. The variant with pink flower colour is proposed to be named "Rosa funingensis f. rosea".
Distribution and habitat. New species are currently known from Funing County, Wenshan Zhuang and Miao Autonomous Prefecture in eastern Yunnan, at elevations between 400 m and 1400 m. They grow on hillsides, roadsides and riversides.  Conservation status. Based on currently available data, the newly-described Rosa funingensis species and its variants should be assigned to the 'Data Deficient' (DD) category of IUCN (2022). The precise conservation status of the population(s) has not been determined. Further explorations are needed to assess its distribution and conservation status. The known distribution of this species is limited. The type locality of this new species is an unprotected mountainous area. Increasing human activities and habitat destruction may cause a threat to the existence of this rare species.