Alseodaphnopsis: A new genus of Lauraceae based on molecular and morphological evidence

An investigation of a questionable species of the genus Alseodaphne led to the discovery of a new genus Alseodaphnopsis H. W. Li & J. Li, gen. nov., separated from Alseodaphne Nees, and a new species Alseodaphnopsis ximengensis H. W. Li & J. Li, sp. nov., endemic to Yunnan province, China. This new species is characterized by having big, axillary, paniculate inflorescences, as well as large, subglobose fruits. Based on DNA sequence data from two gene regions (nuclear ribosomal ITS and LEAFY intron II), we investigate its phylogenetic position within the Persea group. Phylogenies using maximum parsimony (MP) and Bayesian inference (BI) support the recognition of Alseodaphnopsis as a distinct genus but do not resolve well its relationship within the Persea group. The new genus is circumscribed, eight new combinations for its species are made, and a description and illustration of the new species are provided.

The delimitation of Alseodaphne Nees historically has been difficult and circumscriptions have been variable since the genus was first described by Nees [8], who incorporated four species, of which A. semecarpifolia Nees only stands.The three other species belong to Phoebe and Litsea Lam.(Lauraceae), and Castanopsis (D. Don) Spach (Fagaceae).Meisn.[9] followed Nees and moved Phoebe excelsa Nees to Alseodaphne.Bentham & Hooker [10] treated Alseodaphne and Nothaphoebe as sections within Persea.Gamble [11] recognized Alseodaphne and Persea as two distinct genera, while Hooker [12] reduced Nothaphoebe to Alseodaphne, and Boerlage [13] moved all Malesian Nothaphoebe to Alseodaphne.Ridley [14], on the other hand, kept the two genera Nothaphoebe and Alseodaphne separate in his The Flora of the Malay Peninsula.Kostermans [15] included Alseodaphne and Nothaphoebe in Persea in his overview of all Lauraceae, but changed his opinion and recognized Alseodaphne and Nothaphoebe as independent genera later [16].He also thought that these two genera possibly might be fused again in the future, because they are extremely close to each other.van der Werff [17] came to a similar conclusion.He found no significant difference between the two genera and included Nothaphoebe in Alseodaphne, estimating the number of species to about 90.He also expressed doubts about the delimitation between Alseodaphne and Dehaasia.Julia et al. [18], however, recognized Alseodaphne, Dehaasia and Nothaphoebe as distinct genera based on a combination of numerous morphological characters.In the light of the DNA phylogenetic results, Rohwer and Rudolph [5] found that Alseodaphne perakensis (Gamble) Kosterm.and Dehaasia cuneata (Bl.)Bl. formed a strongly supported clade.Rohwer et al. [1] found that the relationship of Nothaphoebe umbelliflora (Bl.)Bl. and some species of Alseodaphne was very close, but Alseodaphne did not appear monophyletic in their study, although with insufficient support.The work of Li et al. [2] showed that Alseodaphne was clearly a polyphyletic group.Its species were placed in two distinct clades, one of which included also Dehaasia (5 spp.examined) and Nothaphoebe umbelliflora.The species of this clade are mainly distributed in tropical Asia.The second clade (Clade III in Li et al., 2011) was poorly resolved at the base and included also the species of Phoebe as well as a few Neotropical species currently placed in Persea.The Alseodaphne species belonging to this clade are mainly distributed in the northern margin of tropics in southwestern China.In our study, we won't discuss in depth the relationships of species within Alseodaphne, Nothaphoebe, and Dehaasia, as our samples from tropical SE Asia are limited, but aim at the species of Alseodaphne with their main distribution in southwestern China.
We conducted an exploratory trip to Yunnan, China, and collected a questionable plant possibly belonging to Alseodaphne, and this was the reason to study the species from SW China in more detail.In this study, the phylogenetic relationships of the genus Alseodaphne distributed in southwestern China are assessed by using DNA sequence data in addition to morphological data.

Ethics statement
Collection of these species was conducted in compliance with existing regulations for plants defined as non-commercial, as determined by local government offices.In addition, these sample collections were performed in China with the written approval from the National Forest Bureau and relevant local governments, complying with Chinese and international regulations for the collection of native plant samples.

Morphological observations
Material of the questionable taxon was collected in November 2007, January 2008, August 2015 and November 2015 from the county of Ximeng (22˚41 0 28.36@N, 99˚38 0 59.68@E), Yunnan, China.The morphological description is based on fresh and pressed specimens.Details of the flowers were examined and photographed under a stereomicroscope (ZEISS Discovery V12.0).The morphological comparison with other closely related species is based on study of living plants in the field as well as herbarium specimens, supplemented by information gathered in the relevant literature [6][7]16].The specimens examined have been deposited in the herbarium of HITBC.Inflorescence size data and fruit size data of the new taxon and known Alseodaphne species were collected from pressed specimens or literature searches, and compared using the independent samples t test.Significance was assumed at p<0.05.All statistical analyses were performed using the SPSS statistical software package 16.0 [19].

Taxon sampling
In the present study, the ingroup sampling included 66 samples (including the new taxon) representing most of the genera within the Persea group (Machilus, Phoebe, Dehaasia, Nothaphoebe, and Alseodaphne).As in the work of Li et al. [2], nine species from four closely related genera (Actiondaphne, Lindera, Litsea, Neolitsea) were selected as the outgroups.Voucher information and GenBank accession numbers are listed in S1 Table.

Nomenclature
The electronic version of this article in Portable Document Format (PDF) in a work with an ISSN or ISBN will represent a published work according to the International Code of Nomenclature for algae, fungi, and plants, and hence the new names contained in the electronic publication of a PLOS ONE article are effectively published under that Code from the electronic edition alone, so there is no longer any need to provide printed copies.
In addition, new names contained in this work have been submitted to IPNI, from where they will be made available to the Global Names Index.The IPNI LSIDs can be resolved and the associated information viewed through any standard web browser by appending the LSID contained in this publication to the prefix http://ipni.org/.The online version of this work is archived and available from the following digital repositories: PubMed Central, LOCKSS.

DNA extraction, PCR amplification and sequencing
Total genomic DNA was extracted from silica-gel dried leaf specimens using the Plant Genomic DNAKit (Tiangen Biotech, Beijing, China).The analyses presented here used the sequence data from two DNA regions, the internal transcribed spacer (ITS) region (ITS1-5.8S-ITS2) of the nuclear ribosomal DNA and LEAFY intron II.These regions have been shown to be valuable in phylogenetic studies within the Persea group [2].The ITS and LEAFY intron II regions were amplified and sequenced following the work of Li et al. [2].The amplified products of LEAFY intron II were purified using the EZNA Cycle-Pure Kit (Omega Bio-Tek, Georgia, USA) before cloning.Cloning was performed using the pEASY-T3 Cloning Kit (TransGen Biotech, Beijing, China).At least 6 positive clones from each individual sample were sequenced and up to 12 positive clones were sequenced for some samples.The sequence chromatogram output files were assembled and edited using Sequencer 4.5 (GeneCodes, Ann Arbor, Michigan, USA).

Sequence alignment and phylogenetic analyses
DNA sequences were aligned by the program Clustal X 1.81 [20] and edited manually using BioEdit 7.0.9.0 [21].A single representative sequence was chosen randomly from multiple clones of each individual sample, as all clones from the same individual sample invariably formed a single clade in a preliminary analysis.Individual and combined datasets for the two markers were assembled as ITS, LEAFY intron II, and ITS + LEAFY intron II.Phylogenetic relationships based on the individual and combined datasets were inferred using unweighted maximum parsimony (MP) by the program PAUP Ã 4.0b10 [22], and Bayesian inference (BI) analyses by the program MrBayes 3.1.2[23][24].
In the MP analyses, a heuristic search was performed with 100 random addition sequence replicates, tree-bisection-reconnection (TBR) branch swapping, collapse of zero length branches, Multrees on and character state changes unordered and equally weighted.Each random addition sequence replicate was allowed to save up to 1000 trees.Bootstrap support values (BS) of the internal nodes were obtained with 100 bootstrap replicates, using the same options as described above.
In the BI analyses, the best-fit model of evolution was chosen for each dataset (ITS and LEAFY intron II) by the program Modeltest 3.7 [25][26] based on the Akaike information criterion (AIC).The Markov chain Monte Carlo (MCMC) algorithm was performed for 2,000,000 generations with one cold and three heated chains, starting from random trees and saving one tree each 100 generations.The first 5000 trees (25%) were discarded as burn-in after checking for stability on the log-likelihood curves, and the remaining 15,000 trees were used to construct the consensus tree.The branch support was determined as Bayesian Posterior Probabilities (BPP).

Comparison of morphological characters
A comparison of morphological characters among the different species of Alseodaphne revealed two groups, consisting of species mainly distributed in tropical Asia (group 1) and species mainly distributed in the northern marginal zone of the tropics in southwestern China (group 2), respectively.Group 1 includes the type species, A. semecarpifolia, whereas the new taxon examined here is placed in group 2. The morphological differences between group 1 species and group 2 species are listed in Table 1.Most of these characters are quantitative and cannot distinguish these two group if considered separately, but in combination, they are effectively separating them.Among the more reliable differences are deciduous perianth lobes in young fruit, not perulate terminal buds as well as whitish twigs contrasting with blackish petioles in dried specimens in group 1 vs. persistent perianth lobes in young fruit, perulate terminal buds, not obviously whitish twigs in group 2. In addition, the error bar charts of the inflorescence size and fruit size between group 1 species and group 2 species showed that both the them were significant (p = 0 and p = 0.016 respectively) (Fig 1).The data of inflorescence size and fruit size are listed in S2 and S3 Tables, respectively.

Phylogenetic analyses
The two DNA loci, ITS and LEAFY intron II, included 600 and 769 aligned position respectively.Modeltest suggested that their evolution was best explained by the TVM + I + G and  The Bayesian consensus tree obtained from the ITS+LEAFY intron II dataset (Fig 2) is largely congruent with phylogenies inferred from separate datasets but more resolved and better supported internally.Thus, the ITS + LEAFY intron II tree is used for the following discussion and its topology is described below.
Within the ITS + LEAFY intron II tree, all species (so far investigated) from the Persea group form a well-defined monophyletic clade (92% BS and 100% PPS).We  subclade is Alseodaphne cladeⅠ, which consists of four Alseodaphne species, Dehaasia hainanensis, and Nothaphoebe umbelliflora.The Alseodaphne clade I receives 99% BS and 100% PPS, and shows very good internal resolution.Clade II consists of Alseodaphne species distributed mainly in southwestern China, viz., A. andersonii, A. hainanensis, A. petiolaris, A. rugosa, A. sichourensis and the new species described below, plus two unidentified samples (Alseodaphne sp.NP, Alseodaphne sp.W14264) that also may represent new species.It receives 86% BS, 100% PPS and has two principal subclades, identical with the Alseodaphne clades II a and II b retrieved from the LEAFY intron II analysis.Clade III consist of all Phoebe species in this study and received 100% PPS but no significant BS.Its two principal subclades, however, are strongly supported in the MP as well as in the BI analyses.

Discussion
Machilus clade and Phoebe clade-These two clades have been retrieved in almost identical composition and topology in an earlier analysis [2], so that there is no need to discuss them here again.
Alseodaphne clade I and Alseodaphne clade II-Just as in the work of Li et al. [2], Alseodaphne appears polyphyletic within the Persea group, with at least two different origins.Of all the Alseodaphne species investigated, four species (A.gracilis, A. huanglianshanensis, A. semecarpifolia and A. gigaphylla) and one unidentified sample (Alseodaphne sp.W17084) fall into Alseodaphne clade I, whereas six species (A.andersonii, A. petiolaris, A. sichourensis, A. rugosa, A. hainanensis, and the new taxon described below) and two unidentified samples (Alseodaphne sp.NP, Alseodaphne sp.W14264) are found in an independent clade (Alseodaphne clade II).The Alseodaphne clade I includes the type species, A. semecarpifolia, which is most dry-resistant species in Alseodaphne [16], so that the name Alseodaphne will stay with this clade, which represents the traditionally recognized typical Alseodaphne species distributed mainly in tropical Asia.
The origin of the Alseodaphne clade II species is apparently different.Most of them are from southwestern China, and Alseodaphne sp.W14264 was collected in northern Vietnam, not far from the Chinese border.The earliest fossil of Alseodaphne was found in Changchang Basin of Hainan Island (China), and the extinct Alseodaphne changchangensis is closest to the living species A. hainanensis [27].Among the extant species, A. hainanensis and A. rugosa occur in Hainan, and both are members of Alseodaphne clade II.Alseodaphne hainanensis also has been reported from northern Vietnam [7].Hainan Island belongs to the same phytogeographical region as tropical southern China, and it may have been connected to northern Vietnam and Guangxi at least in the Eocene [28].We also find that the morphological characters within the independent Alseodaphne clade II are mainly consistent with the molecular results.This clade is divided into two subclades receiving 92% BS, 100% PPS and 98% BS, 100% PPS respectively (Fig 2).In most of the species of Alseodaphne clade II b the lower surfaces of the leaves are distinctly glaucous (except Alseodaphne sp.NP), while all of the species of Alseodaphne clade II a have completely green lower leaf surfaces.
The fact that this independent Alseodaphne clade II differs also morphologically from the traditionally recognized Alseodaphne species has already been discussed by Rohwer et al. [1].In this study, we also find some differences between these two clades (Table 1).We therefore think that the independent Alseodaphne clade II should be recognized as a new genus, which we call Alseodaphnopsis.A formal description of the new genus is provided below.The vegetative characters may be insufficient to distinguish the two genera independently, as most of them are quantitative characters, but in combination, they can be used to segregate the two genera.The principal characters to distinguish the two genera include: 1) twigs thick, 4-11 mm in diameter, not obviously whitish in color vs. thin, 2.5-4.5 mm in diameter, and obviously whitish in color; 2) terminal buds perulate vs. not perulate; 3) perianth lobes persistent at least in young fruit vs. early deciduous; 4) inflorescences relatively large, 8.5-35 cm long, generally many-flowered, with 3-4 order of branching vs. 3-20 cm long, few-flowered, with 1-2 orders of branching; and 5) mature fruit relatively large, 3-5 cm or < 2.5cm in diameter.We found that twigs are not obviously whitish in color and the terminal buds are perulate in almost all species of the new genus Alseodaphnopsis, whereas the twigs are obviously whitish in color and terminal buds are not perulate in most of the traditional Alseodaphne species.This may be an adaptation to the more seasonal climate in the area of distribution of Alseodaphnopsis, with colder winters than in the area of distribution of most Alseodaphne species.In addition, the perianth lobes are persistent in young fruit period in at least some species of Alseodaphnopsis.We have observed this character in A. andersonii, A. petiolaris and A. hainanensis, but not in the type species A. semecarpifolia, possibly also caused by the difference of climate.Although the size of panicles and fruits look as quantitative ones, perulate terminal buds as well as persistent perianth lobes look as more or less qualitative ones, these qualitative characters are better to be called as the key characters to define Alseodaphne clade II.Alseodaphne gracilis and A. huanglianshanensis, which are distributed in northern tropical Asia, are nested in Alseodaphne clade I, not in Alseodaphne clade II, as most of the other species from tropical China.Also morphologically, they are more similar to A. semecarpifolia, which is likewise located in Alseodaphne clade I, than to most other Chinese species.We have observed that A. gracilis has no perianth lobes in its young fruit, which matches with its placement in traditional Alseodaphne.In addition, A. gracilis, A. huanglianshanensis and A. semecarpifolia possess thinly leathery leaves and thin twigs with conspicuously whitish bark, whereas the Alseodaphnopsis species have leathery leaves and thick twigs that are green or brown, but not whitish in fresh material.
Based on both morphological and molecular evidence, we therefore propose a new genus Alseodaphnopsis, separated from the traditional genus Alseodaphne, to accommodate the independent Alseodaphne clade II including most of the species distributed in the northern margin of tropics in southwestern China (including the new species, Alseodaphnopsis ximengensis).In Yunnan (China), at least two further Alseodaphne species (A.hokouensis H. W. Li, A. marlipoensis (H.W. Li) H. W. Li) belongs to this clade, judged from morphology, so that we can expect this group to become larger with increasing taxon sampling.
Description: Trees evergreen.Terminal buds perulate.Twigs robust, 4-10 mm in diam., not whitish.Leaves alternate, pinninerved, always clustered at the ends of the branchlets, abaxial side glaucous or not.Inflorescence axillary, paniculate, bracts and bracteoles deciduous.Flowers bisexual, trimerous.Receptacle short; perianth lobes 6, subequal or extremely unequal, slightly dilated after anthesis and ± persistent at least in young fruit.Fertile stamens 9, in 3 HKY + G evolutionary models, respectively.The topologies of the consensus trees obtained from the MP and Bayesian analyses, based on different datasets (ITS, LEAFY intron II and ITS + LEAFY intron II), were mostly congruent.Most of their major clades were identical, and only minor variation in the composition and relationships of a few terminal nodes were detected, possibly caused by insufficient phylogenetic signal in the data and insufficient sample size.Moreover, these inconsistencies received only very weak support.Here, only Bayesian consensus trees with bootstrap support (BS) values and posterior probability support (PPS) values are presented for demonstration.The Bayesian consensus tree obtained from the ITS dataset (S1 Fig) is largely congruent with the work of Li et al. [2].The two principal clades within the Persea group are (1) the Machilus clade, and (2) a clade including species of Alseodaphne, Dehaasia, Nothaphoebe, and Phoebe.Some major clades in the ITS tree, which also appear in the ITS + LEAFY intron II tree (Fig 2), are labeled for comparison.The Bayesian consensus tree obtained from the LEAFY intron II dataset (S2 Fig) is also largely compatible with the work of Li et al. [2].The three principal clades within the Persea group are (1) Alseodaphne clade II a, (2) Alseodaphne clade II b+ Phoebe clade, and (3) the Machilus clade + a clade including Alseodaphne, Dehaasia, Nothaphoebe.As in S1 Fig, some major clades are labeled for comparison with the ITS + LEAFY intron II tree.
defined three principal clades (clade I, II, and III) within the Persea group.Clade I consists of all Machilus and Dehaasia species included in the present study, and four Alseodaphne species distributed in tropical Asia (A. semecarpifolia and A. gigaphylla) and southwestern China (A. huanglianshanensis, A. gracilis), as well as Nothaphoebe umbelliflora.It receives 56% BS and 100% PPS, and has two principal subclades.The major component of the first principal subclade is the Machilus Clade, which comprises all representatives of Machilus included in the present study.It receives 88% BS and 99% PPS, and shows very little internal resolution.The second principal