A Phylogenetic Morphometric Investigation of Interspecific Relationships of Lyponia s. str. (Coleoptera, Lycidae) Based on Male Genitalia Shapes

Simple Summary Resolving phylogenetic relationships among animals remains one of the most challenging issues in systematics. Currently, molecular phylogeny is the standard for inferring evolutionary relationships, but morphological analysis still cannot be replaced or neglected. Male genitalia have been proven to be valuable in phylogenetic analyses usually in the higher taxonomic grades but are rarely studied at the lower level. In the present study, we performed a taxonomic review (with two new species described) and further investigated the interspecific relationship of Lyponia s. str. based on the morphometric data of phallus shapes using geometric morphometric (GM) and phylogenetic morphometric (PM) methods. As a result, the produced topologies (of the unweighted pair group method using arithmetic averages (UPGMA), neighbor-joining (NJ) and maximum parsimony (MP) analyses) provide a general framework of the morphological evolution of this subgenus. To be exact, these species are divided into two clades that represent two shapes of the phallus. The results provide better understanding of the species diversity and evolution of Lyponia s. str. and shed new light on investigations of the phylogenetic relationships of insects based on male genitalia shapes, which is particularly useful when molecular data are unavailable. Abstract The nominate subgenus Lyponia Waterhouse, 1878 from China is reviewed, with two new species described and named L. (s. str.) ruficeps sp. n. (China, Yunnan) and L. (s. str.) zayuana sp. n. (China, Xizang). A distribution map and a key to all species of Lyponia s. str. are provided. Moreover, the phenotypic relationships among the species of Lyponia s. str. are investigated based on phallus shapes using geometric morphometric and phylogenetic morphometric analyses. The topologies demonstrate that the species are divided into two clades. One clade is composed of six species (L. ruficeps sp. n., L. zayuana sp. n., L. kuatunensis, L. shaanxiensis, L. hainanensis, and L. tamdaoensis) and is supported by a stout phallus (less than 3.6 times longer than wide). The other clade includes the remaining species (L. nepalensis, L. debilis, L. cangshanica, L. delicatula, and L. oswai) and is supported by a slender phallus (at least 4.1 times longer than wide). These results provide better understanding of the species diversity and evolution of Lyponia s. str. Nonetheless, more samples and loci are required in the future to verify the present results.


Introduction
Insecta is the largest organism group on the earth, with a high diversity of morphological variations [1].Male genitalia of insects are one of the most evolutionarily variable morphological characters, and their apparently rapid rate of morphological changes is assumed to result from sexual selection [2].They are helpful for allowing systematists to identify diagnostic features at various taxonomic levels [3].As noted in other insects, male genitalia are the most important and useful diagnostic characteristics of species of net-winged beetles, including Lyponia [4].They have also been demonstrated to be useful in phylogenetic analyses [5] and are usually applied in higher-level classifications and phylogenetic frameworks [6][7][8].Although there are few studies assessing the phylogenetic value of male genitalia in the lower grades due to the uniformity of their structures, it has become possible to elucidate the phenotypic relationships among the species with the advent of geometric morphometric (GM) [9] and phylogenetic morphometric (PM) methods [10].Following our preceding publication [11], herein we construct the phenotypic relationships of Lyponia s. str.based on male genitalia shapes.Although Bocak [12] conducted a cladistics analysis of Lyponia, he noted that the analysis was only for the species groups not the species given that the high morphological similarity in some species preventing coding the dataset.More recently, Li et al. [13] proposed a phylogenetic hypothesis for Lyponiini based on molecular data, in which only a few species of Lyponia s. str.were included in the analysis because of limited availability of the material.Therefore, it is necessary for us to comprehensively assess the phylogenetic relationships among the species of Lyponia s. str.herein.
The genus Lyponia was proposed by Waterhouse for L. debilis Waterhouse, 1878 [14].It was always mixed with Ponyalis Fairmaire, 1899 [15] due to their similarity in general appearances [15][16][17][18][19].For this reason, Nakane [20] created a broad concept of Lyponia s. l. that included both Ponyalis and Lyponia s. str.Thirty years later, Lyponia s. l. was systematically revised by Bocak [12], in which Ponyalis was treated as a subgenus of Lyponia, and another subgenus Weiyangia was proposed.Soon after that, Kazantsev reinstated Ponyalis as a separate genus and subdivided Lyponia into five subgenera: Lyponia s. str.; Weiyangia Bocak, 1999; Poniella Kazantsev, 2002 (the L. nigroscutellaris group by Bocak, 1999); Mimoditonecia Kazantsev, 2002;and Sundolyponia Kazantsev, 2002 [21].Regardless of the classification system, Lyponia is most closely related to Ponyalis, and they belong to the subtribe Lyponiinina [7] or the tribe Lyponiini [8,[22][23][24] based on either morphological or molecular evidence.The members of Lyponia are distinguished from those of Ponyalis based on the following characters: free basal part of the coxite, while basally fused in Ponyalis; antennomere I of both sexes progressively widened and not compressed anteriorly, while abruptly widened near the base and more or less compressed at the anterior margin in Ponyalis; primary elytral costa III not going beyond the apical fifth, while nearly reaching the apex of the elytra in Ponyalis; aedeagus absent with numerous minute thorns in the preapical portion of the median lobe, while present in Ponyalis [21].
At present, a total of 27 species of Lyponia sensu Kazantsev are hitherto known [4,12,[14][15][16][17][18][19][20][21][25][26][27][28][29][30][31], and they are widely distributed in the eastern Palaearctic and Oriental regions [4,12,21].Among the subgenera, Lyponia s. str. is easily distinguished from others based on the larger eyes and male antennomere III equipped with slender lamella [21].Prior to this study, this subgenus contained nine species [4,12,14,21,25,29,30], of which five species are present in China.Recently, we obtain some specimens of Lyponia s. str.from China.After our examination and identification, we discover two new species and a species newly recorded in China that are reported in the present study.Meanwhile, the previously known species are illustrated to make comparisons with the new species.Moreover, the phylogeny of Lyponia s. str. is constructed to assess the phonetic relationships among the species.The results of this study will allow us to better understand the species diversity and evolution of Lyponia s. str.

Material
The studied material is preserved in the following collections: IZAS (Institute of Zoology, Chinese Academy of Sciences, Beijing, China) and MHBU (Museum of Hebei University, Baoding, China).

Morphological Techniques
We softened the dry specimens with water, then dissected the aedeagus, and cleared it in 10% NaOH solution.After that, we examined and took photos of the aedeagus, which was immersed in glycerol, and then glued it on a paper card for permanent preservation.We obtained images of the habitus with a Canon EOS 80D digital camera (Canon (China) Co., Ltd., Zhuhai, China) and aedeagi using a Leica M205A stereomicroscope (Leica Microsystems Inc., Heerbrugg, Switzerland).Then, we stacked images of different layers using Helicon Focus 7. In the end, we edited the final permutation in Adobe Photoshop CS3.10.0.1.
We performed the following measurements using Image J 1.50i (NIH, Bethesda, MD, USA): body length (from the anterior margin of head to the elytral apex), body width (the width across the humeri of elytra), pronotal length (from middle of anterior margin to middle of posterior margin of pronotum), pronotal width (the width across the widest part of pronotum), eye diameter (the maximal width of an eye), and interocular distance (the minimal distance between eyes).

Map Preparation
We collected the distribution information from the original publications [2][3][4]6] and the material in this study.Based on the retrieved data, we prepared the distribution map using ArcMap 10.Then, we input the morphometric data into the tps-UTIL 1.43 software [33] and used TpsSmall (ver.1.20, F. Rohlf, see http://life2.bio.sunysb.edu/morph, assessed on 10 September 2023) to test whether the observed variation in shape was small enough that the distribution of points in the tangent space can be used as a good approximation of the distribution in shape space.We analyzed the coordinates using TpsRelw 1.46 software [34] to calculate the Eigen values for each principal warp.Finally, we showed the shape changes of different species as thin-plate splines based on variation along the first two relative warp axes.

Data Analyses
We analyzed the aligned landmarks data (which was converted from semi-landmarks in the tps.file manually) using MorphoJ 1.06d software [35] to examine the shape variation.We employed principal component analysis (PCA) to test how well the species can be separated based on the shape of the phallus.The observed variation patterns among the species frequently corresponded to the characters with high loading value in PCAs.We used canonical variates analysis (CVA) to analyze the relative similarity and discrimination of the test groups.The shape values that maximize group means relative to variation within groups were determined using CVA, in which the covariate matrices are assumed to be identical [36].We computed the Procrustes distances and Mahalanobis distances (the square root of the sum of squared differences between corresponding points) between each of the species and produced the matrices using Mor-phoJ software [35].

Phylogenetic Morphometric (PM) Analyses
We investigated phylogenetic relationships among the species of Lyponia s. str.based on the morphometric data of male genitalia using unweighted pair group method using arithmetic averages (UPGMA), neighbor-joining (NJ) and maximum parsimony (MP) analyses [37,38].
We conducted UPGMA and cluster analyses based on the Procrustes and Mahalanobis distances matrices (Supplementary material: Table S1).Then, we loaded the Procrustes and Mahalanobis distance score matrices in PAST 2.17 [39], separately, to assess the phonetic relationships among the species.
In addition, we constructed NJ trees [40] using PAST 2.17 with 1000 bootstrap replicates to display the Mahalanobis and Procrustes distances between populations, separately.
Moreover, we produced the tps files in tps-DIG (Supplementary material: Table S2) to perform MP analysis using TNT 1.5 [38].We followed a heuristic (traditional search) search strategy.Here, random addition sequences were utilized, and tree bisection reconnection (TBR) was used as a branch swapping algorithm.In each replicate, a separate

Data Analyses
We analyzed the aligned landmarks data (which was converted from semi-landmarks in the tps.file manually) using MorphoJ 1.06d software [35] to examine the shape variation.We employed principal component analysis (PCA) to test how well the species can be separated based on the shape of the phallus.The observed variation patterns among the species frequently corresponded to the characters with high loading value in PCAs.We used canonical variates analysis (CVA) to analyze the relative similarity and discrimination of the test groups.The shape values that maximize group means relative to variation within groups were determined using CVA, in which the covariate matrices are assumed to be identical [36].We computed the Procrustes distances and Mahalanobis distances (the square root of the sum of squared differences between corresponding points) between each of the species and produced the matrices using MorphoJ software [35].

Phylogenetic Morphometric (PM) Analyses
We investigated phylogenetic relationships among the species of Lyponia s. str.based on the morphometric data of male genitalia using unweighted pair group method using arithmetic averages (UPGMA), neighbor-joining (NJ) and maximum parsimony (MP) analyses [37,38].
We conducted UPGMA and cluster analyses based on the Procrustes and Mahalanobis distances matrices (Supplementary material: Table S1).Then, we loaded the Procrustes and Mahalanobis distance score matrices in PAST 2.17 [39], separately, to assess the phonetic relationships among the species.
In addition, we constructed NJ trees [40] using PAST 2.17 with 1000 bootstrap replicates to display the Mahalanobis and Procrustes distances between populations, separately.
Moreover, we produced the tps files in tps-DIG (Supplementary material: Table S2) to perform MP analysis using TNT 1.5 [38].We followed a heuristic (traditional search) search strategy.Here, random addition sequences were utilized, and tree bisection reconnection (TBR) was used as a branch swapping algorithm.In each replicate, a separate tree was obtained, and the runs were repeated 1000 times (mult = ras tbr hold 1 rep 1000) [10].

Key to the subgenera of
Distribution (Figure 2).China (Anhui, Fujian, Hubei, Hainan, Jiangxi, Guangdong, Guangxi, Guizhou, Yunnan, Zhejiang, Xizang), Japan, Nepal, and Vietnam.Distribution (Figure 2).China: Yunnan.Remarks.In the original publication [4], the male habitus was well illustrated, but the aedeagus was only described with illustrations in the ventral view.Here, we provided macrophotographs of the aedeagus in ventral, dorsal and lateral views for this species to make it better understood.In addition, we described the male antennae in more detail to make comparisons with other species, since this character is vital for specific identification.3D-F): phallus stout and 2.6 times as long as wide, widened near middle part and tapered at apex in dorsal and ventral views (Figure 3D,E), nearly straight in the lateral view (Figure 3F), internal sac robust and obviously protruding over the apex of the phallus (Figure 3D-F).
Remarks.Although the male of this species has been mentioned several times [4,12,21], its aedeagus was illustrated in the ventral view only by Bocak [12].In the present study, the habitus of the male and female for this species are provided for the first time, and the aedeagus is illustrated in ventral, dorsal and lateral views to provide more information.Descriptive notes.Male (Figure 4C).Antennae flabellate, overlapping the midlength of elytra when inclined.Antennomere II transverse; III-XI lamellate, and lamella of IV ca.1.2 times longer than the joint itself, and lamella of IX 1.6 times as long as the joint itself.

Lyponia (s. str.) kuatunensis
Aedeagus (Figure 3G-I): phallus moderately slender and about 3.5 times as long as wide, feebly widened apically and tapered at apex in dorsal and ventral views (Figure 3G,H), feebly bent dorsally in lateral view (Figure 3I), internal sac robust, but never protruding over the apex of the phallus (Figure 3G-I).
Female (Figure 4D).Similar to male, but body stouter, antennae serrate.Distribution (Figure 2).China (Fujian, Guangxi, Hunan, Guizhou).Remarks.Bocak [12] provided the illustrations of male habitus and aedeagus in ventral view for L. kuatunensis.In this study, the habitus of both sexes, and the aedeagus in ventral, dorsal and lateral views are illustrated to provide more morphological details.Aedeagus (Figure 3J-L): phallus slender and approximately 4.1 times as long as wide, widened apically and tapered at apex in dorsal and ventral views (Figure 3J,K), feebly bent dorsally in the lateral view (Figure 3L), internal sac well developed and protruding over the apex of the phallus (Figure 3J-L).

Lyponia
Distribution (Figure 2).China (Hainan, Guangxi).Remarks.In the original description provided, the photos of the male habitus and aedeagus were in the ventral view [12].In this study, we illustrate the aedeagus in ventral, dorsal and lateral views to make easier to compare with others.Descriptive notes.Male (Figure 5A).Antennae flabellate, overlapping basal two-thirds length of elytra when inclined.Antennomere II transverse; III-XI lamellate, lamella of IV ca.1.9 times as long as the joint itself, and lamella of IX 3.0 times as long as the joint itself.6A-C): phallus slender and about 3.6 times as long as wide, widened apically and rounded at apex in ventral and dorsal views (Figure 6A,B), "S"-shaped in lateral view (Figure 6C), internal sac slender and slightly protruding over the apex of the phallus (Figure 6A-C).
Aedeagus (Figure 6A-C): phallus slender and about 3.6 times as long as wide, widened apically and rounded at apex in ventral and dorsal views (Figure 6A,B), "S"-shaped in lateral view (Figure 6C), internal sac slender and slightly protruding over the apex of the phallus (Figure 6A-C).
Female (Figure 5B).Similar to male, but more slender than male, and antennae serrate.
Distribution (Figure 2).China (Henan, Shaanxi).Remarks.In the present study, the female is firstly reported, and the habitus of both sexes are provided for the first time.In addition, we discover that the body coloration of this species is somewhat variable, and the individuals from Shaanxi always have a black patch on the pronotum and black scutellum (Figure 5B), while those from Henan have a dark-brown patch on the pronotum and light red scutellum (Figure 5A).Remarks.In the present study, the female is firstly reported, and the habitus of both sexes are provided for the first time.In addition, we discover that the body coloration of this species is somewhat variable, and the individuals from Shaanxi always have a black patch on the pronotum and black scutellum (Figure 5B), while those from Henan have a dark-brown patch on the pronotum and light red scutellum (Figure 5A).
Descriptive notes.Male (Figure 5C).Antennae flabellate, overlapping basal two-thirds length of elytra when inclined.Antennomere II transverse; III-XI lamellate, lamella of IV ca.1.5 times as long as the joint itself, and lamella of IX twice as long as the joint itself.
Aedeagus (Figure 6D-F): phallus stout and 3.0 times as long as wide, widened to one side and widely rounded at the apex in dorsal and ventral views (Figure 6D,E), nearly straight in the lateral view (Figure 6F), internal sac reduced and minimally visible (Figure 6D-F).
Remarks.The species is recorded to China for the first time.Differential diagnosis.It is most similar to L. nepalensis on basis of the general appearance but can be easily distinguished from the latter based on the following characters: lamella of antennomere IX slender in male, about 10.0 times longer than wide (Figure 5D), while robust in L. nepalensis, about 6.0 times longer than wide (Figure 5C); pronotum with posterior margin arched (Figure 5D), while nearly straight in L. nepalensis (Figure 5C); phallus nearly parallel-sided with a robust internal sac (Figure 6G-I), while widened to one side with feebly visible internal sac in L. nepalensis (Figure 6D-F).
Head dorsally flat, antennal tubercles and transversal depression present; antennae almost reaching the basal two-thirds length of the elytra, antennomere II transverse; III-XI with short and thin lamellae, lamella of IV 1.4 times as long as the joint itself and 1.7 times as long as wide, lamella of IX 1.8 times longer than the joint itself and 10.0 times as long as wide.
Pronotum almost trapezoidal, with rounded anterior angles and acute posterior angles, anterior margin widely rounded and slightly producing anteriorly and posterior margin strongly arched posteriorly.Scutellum slightly narrowed posteriorly and straight at the apex.
Elytra slightly widened posteriorly, only primary costae II and VI stouter than secondary ones, secondary costae III shortened to the apical fifth length, cells irregular at the basal part of the elytra.
Aedeagus (Figure 6G-I): phallus moderately slender and 3.5 times as long as wide, nearly parallel-sided, narrowly rounded at apex in dorsal and ventral views (Figure 6G,H), feebly bent dorsally in the lateral view (Figure 6I), internal sac moderately developed and robust, but never protruding over the apex of the phallus (Figure 6G-I).
Etymology.The name of the species is derived from the name of the type locality "Zayu", Xizang Autonomous Region, China.Differential diagnosis.It looks similar to L. kuatunensis in general appearance, but can be differentiated by the uniformly red pronotum (Figure 5E,F), while a dark-brown patch is always present in L. kuatunensis (Figure 4C,D); lamella of antennomere IX slender in male, about 15.0 times longer than wide (Figure 5E), while much shorter in L. kuatunensis, about 7.0 times longer than wide (Figure 4C); phallus subparallel-sided in ventral view (Figure 6J), while progressively widened apically in L. kuatunensis (Figure 3G).
Head dorsally flat, antennal tubercles and transversal depression present; antennae nearly reaching middle length of the elytra, antennomere II transverse; III-XI with short and thin lamellae, lamella of IV 1.4 times as long as the joint itself and 2.0 times as long as wide, lamella of IX 1.8 times longer than the joint itself and 15.0 times as long as wide.
Pronotum almost trapezoidal, with rounded anterior angles and rectangular posterior angles, anterior margin widely rounded and projecting anteriad, lateral margin sinuates and posterior margin bisinuate.Scutellum feebly narrowed posteriorly and obviously emarginate at the apex.
Elytra parallel-sided, all primary costae stouter than secondary ones, cell irregular, primary costa I and secondary costa I merged at apical part of disc, primary costa II stouter than all other costae.
Aedeagus (Figure 6J-L): phallus slender and about 4.5 times as long as wide, nearly parallel-sided and tapered at apex in dorsal and ventral views (Figure 6J,K), feebly bent dorsally in lateral view (Figure 6L), internal sac moderately developed and stout, but never protruding over apex of phallus (Figure 6J-L).
Female (Figure 5F).Stouter than male, antennal serrate, reaching nearly the midlength of elytra, pronotum with anterior angles confluent with anterior margin, scutellum feebly emarginate at the apex.

GM Analyses of Male Genitalia Shapes
The correlation (uncentered) between the tangent space (Y) regressed onto the Procrustes distance (geodesic distances in radians) was 0.999997.There was little doubt on the basis of the result from tps-SMALL, which supported that the data are acceptable by geometric morphometric methods, since the results from the statistical test performed using tps-SMALL proved the acceptability of the data set for further statistical analysis [41].
The first three principal components of the shape of phallus explain 94.433% of the micromesh variation and were 68.041%, 21.877% and 4.525%, respectively (see Supplementary material: Table S3).They were plotted to indicate variation along the first two relative warp axes (Figure 7).The shape changes of all species of Lyponia s. str.were shown as deformations of the least squares reference using thin-plate splines (Figure 7A-L).
Comparison of the tps configurations indicated that the average shape of the phallus of Lyponia s. str. is asymmetrical and feebly bent to one side, almost even in width except feebly narrowed at the apical one-seventh, and rounded at apex in dorsal view (Figure 7A).
The first principal component (x-axis) dominated the uniformity of the phallus shape, and the species on the left seemed narrowed apically (e.g., Figure 7B,C,I-L).In contrast, those on the right appeared widened near apex (e.g., Figure 7D-G).Meanwhile, the second principal component (y-axis) decided the width of the phallus.The species below were generally stout (e.g., Figure 7H-L), while those above seemed slender (e.g., Figure 7B-H).
Moreover, the CVA and PCA scatter plots of shape differences of the phallus shapes in dorsal views (see Supplementary material: Figures S1 and S2) showed that each species of Lyponia s. str.independently occupied an area and were separated from one another.Comparison of the tps configurations indicated that the average shape of the phallus of Lyponia s. str. is asymmetrical and feebly bent to one side, almost even in width except feebly narrowed at the apical one-seventh, and rounded at apex in dorsal view (Figure 7A).
The first principal component (x-axis) dominated the uniformity of the phallus shape, and the species on the left seemed narrowed apically (e.g., Figure 7B,C,I-L).In contrast, those on the right appeared widened near apex (e.g., Figure 7D-G).Meanwhile, the second principal component (y-axis) decided the width of the phallus.The species below were generally stout (e.g., Figure 7H-L), while those above seemed slender (e.g., Figure 7B-H).
Moreover, the CVA and PCA scatter plots of shape differences of the phallus shapes in dorsal views (see Supplementary material: Figures S1 and S2) showed that each species of Lyponia s. str.independently occupied an area and were separated from one another.
In the NJ tree of Procrustes distances (Figure 8C), clade I was recovered, but the interrelationships among the species were different from the above (Figure 8A).Although the species of clade II were not clustered together, the two-component small branches (L.nepalensis + L. oswai) + L. cangshanica (Figure 8e) and L. delicatula + L. debilis (Figure 8g) were recovered, separately.In comparison, both (L.hainanensis + L. tamdaoensis) of clade I and ((L.nepalensis + L. cangshanica) + L. oswai) of clade II were only partially recovered in the NJ tree of Mahalanobis distance (Figure 8C).
The MP tree of two landmark configurations (Figure 8E) produced a significantly different topology from the above phenograms.Only the sister relationship of L. delicatula + L. debilis of clade II was recovered.Although the coupled clade of L. nepalensis + L. oswai (Figure 8f) was different from that shown in clade I (Figure 8A,B,D,d), it was similar to that of the NJ tree of Mahalanobis distance (Figure 8C).

Separate Status of the New Species
Like other net-winged beetles, reliable identification of the species is regularly based on the shape of the male genitalia [4].Within the subgenus Lyponia s. str., the morphological interspecific divergence of male genitalia always exhibit quantitative variations [4,12,21].Geometric morphometrics is a rigorous method [42,43] to identify and analyze shape variations between species [44].
In the present study, the phallus shapes of Lyponia s. str.were analyzed using GM analyses.The statistical test performed by TpsSmall suggested that our obtained data on male genitalia are acceptable for geometric morphometric analysis.Further, CVA analysis suggested that all species of Lyponia s. str.can be distinguished from one another based on male genitalia shapes.Moreover, the tps configurations effectively display the shape of the phallus for each species (Figure 7A-L), so it could be included in the comparison between sibling species and complied in the key, thereby facilitating and stabilizing the taxonomy of Lyponia.Also, a logical conclusion drawn from the observations is that male genitalia have always been useful as a taxonomic character for the species, which suggests that they must obtain a new form in every new species [45].This hypothesis was validated in our study, and the two new species discovered here, L. zayuana sp.n. and L. ruficeps sp.n., have phallus shapes that differ from the previously known species.
In addition, the shape of male antennae is rarely used as a conspicuous variable within species [12,21], so it is also applied in the identification key for the species.
Key to the species of Lyponia s. str.

1.
Lamella of male antennomere IX extremely long, at least 2. Aedeagus: phallus slender and at least 4.1 times as long as wide in the dorsal view (e.g., Male antennomere IV nearly as long as the joint itself; elytra with primary costae III disappearing at basal third part; aedeagus: phallus ca.4.6 times as long as wide in the dorsal view (Figure 7E and [4]: Figure 18

Phenotypic Relationships Based on Aedeagi Shapes
In the present study, we analyzed the phallus shapes of all species of Lyponia s. str.(a total of 11 species) using GM analyses and further constructed their interspecific relationships using different methods (Figure 8).Given that any method has its own defects and limitations, we tried different methods to obtain more objective and dependable results.As expected, there is some discrepancy in the produced topologies, which likely resulted from different analysis data and optimal criteria of the methods [11].However, most produced topologies (Figure 8A-C) demonstrate that closer relationships existed among the species of clade I (including L. hainanensis, L. tamdaoensis, L. shaanxiensis, L. ruficeps sp.n., L. kuatunensis, and L. zayuana sp.n.) and those of clade II (including L. nepalensis, L. cangshanica, L. oswai, L. debilis, and L. delicatula), which are supported by synapomorphy (Figure 8a,b).This is consistent with the results of the comparative morphology of phallus shape, which is either stout (Figure 7B-G) or slender (Figure 7H-L).Within clade I, a sister relationship of L. hainanensis + L. tamdaoensis is frequently recovered (Figure 8A-D) and supported by synapomorphy (Figure 8c).This finding is congruent with the GM analysis, and the two species are in closer positions in the first two relative warp axes (Figure 7).However, within clade II, L. nepalensis + L. cangshanica + L. oswai (Figure 8A-D) and L. debilis + L. delicatula (Figure 8A,C,E) are recovered as sister groups, separately, but they do not seem congruent with their relative shape variations of the phallus (Figure 7).Procrustes and Mahalanobis distances are used to capture shape variation, which is considered the best method for measuring shape differences among taxa [41,[46][47][48][49][50].So, they can effectively indicate phonetic relationships, summarizing overall patterns of similarity [41,51].
In constructing morphological phylogeny, male genitalia have been widely used in many arthropod groups by systematists [52], owing to its effective phylogenetic value in higher-level classifications [53][54][55][56][57][58][59].Despite the opinion that genital evolution occurs too fast at the specific level to observe its phylogenetic inertia in the structures [60,61], this notion is contrary to the findings of Song and Bucheli [52] and most recently those of Liu et al. [11].
Unexpectedly, a gap still exists between our obtained results and the previous molecular phylogenic analysis [13].In the study by Li et al. [13], five species of Lyponia s. str.were included in the analysis, and the phylogenetic relationship was recovered as (L.delicatula + L. kuatunensis) + (L.oswai + L. shaanxiensis + L. debilis).However, none of our produced topologies based on male genitalia shape are congruent with this result.Losos [62] argued that if rates of character evolution to speciation are relatively high, there may not be a relationship between the degree of the phylogenetic relationship and phenotypic similarity.If so, it is presumed that some non-genital characters of male [52] or female genitalia [63] are also involved in the speciation of Lyponia.Despite this, we should note that although molecular phylogenetic analysis has become the standard for inferring evolutionary relationships at present [64], the lack of samples of the representative species [e.g., 18] inevitably influenced the ability to elucidate the relationships among the species [65].Nevertheless, more samples or loci are required in the future to reconstruct the phylogeny of Lyponia s. str.

Conclusions
In the present study, we review the lycid subgenus Lyponia s. str.and describe two new species from China, including L. ruficeps sp.n. and L. zayuana sp.n.Then, we investigate the phenotypic relationships among the species of Lyponia s. str.based on the phallus shapes with GM and PM analyses (using UPGMA, NJ and MP analyses).The produced topologies demonstrate that the species are usually divided into two clades (clade I and clade II) using UPGMA and NJ analyses based on the Procrustes and or Mahalanobis distances.Clade I is composed of six species (L.hainanensis, L. tamdaoensis, L. ruficeps sp.n., L. shaanxiensis, L. kuatunensis, and L. zayuana sp.n.) and is supported by a stout phallus (less than 3.6 times longer than wide).Clade II includes the remaining species (L.nepalensis, L. cangshanica, L. oswai, L. debilis, and L. delicatula) and is supported by a slender phallus (at least 4.1 times longer than wide).The sister groups, L. hainanensis + L. tamdaoensis (of clade I), L. nepalensis + L. cangshanica + L. oswai, and L. debilis + L. delicatula (of clade II) are frequently recovered.These results provide better understanding of the species diversity and evolution of Lyponia s. str.Nonetheless, more samples or other types of data are required in the future to verify the present results.

Supplementary Materials:
The following supporting information can be downloaded at https: //www.mdpi.com/article/10.3390/insects15010011/s1,Table S1: Difference in the phallus shapes among the groups: Mahalanobis distances p-values (above) from permutation tests (10,000 permutation rounds); Procrustes distances p-values (below) from permutation tests (10,000 permutation rounds); Table S2: TPS data used to perform MP analysis.Table S3: Eigen values and contributions of the principal components analysis of the phallus shapes in dorsal views; Figure S1: CVA scatter plots of differences in phallus shapes in dorsal views; Figure S2: PCA scatter plots of differences in phallus shapes in dorsal views.

21 Figure 1 .
Figure 1.The curve used in the GM analysis as shown for Lyponia (s.str.) ruficeps sp.n.

Figure 1 .
Figure 1.The curve used in the GM analysis as shown for Lyponia (s.str.) ruficeps sp.n.

Figure 7 .
Figure 7. Shape differences in the phallus among the species of Lyponia s. str.: shape changes among species are implied by variations along the first two relative warp axes; (A-L) splines indicate deformation of the landmarks in comparison with the reference configuration.

Figure 7 .
Figure 7. Shape differences in the phallus among the species of Lyponia s. str.: shape changes among species are implied by variations along the first two relative warp axes; (A-L) splines indicate deformation of the landmarks in comparison with the reference configuration.

Figure 8 .
Figure 8. Comparisons of phenotypic relationships of Lyponia s. str.: (A,B) phylogenetic hypothesis based on Mahalanobis (A) and Procrustes (B) distances using UPGMA; (C,D) neighbor-joining tree for the species of Lyponia s. str.based on Mahalanobis (C) and Procrustes (D) distances with 1000 bootstrap replicates; (E) phylogenetic hypothesis based on two landmark configurations using MP analysis.The branches in different colors or dashed boxes represent the same clade recovered in different phenograms.The average shape of each branch is displayed in (a-g), respectively.

Figure 8 .
Figure 8. Comparisons of phenotypic relationships of Lyponia s. str.: (A,B) phylogenetic hypothesis based on Mahalanobis (A) and Procrustes (B) distances using UPGMA; (C,D) neighbor-joining tree for the species of Lyponia s. str.based on Mahalanobis (C) and Procrustes (D) distances with 1000 bootstrap replicates; (E) phylogenetic hypothesis based on two landmark configurations using MP analysis.The branches in different colors or dashed boxes represent the same clade recovered in different phenograms.The average shape of each branch is displayed in (a-g), respectively.

Author
Contributions: Conceptualization, C.F., Y.Y., X.Y. and H.L.; methodology: C.F. and Y.Y.; formal analysis, C.F. and Y.Y.; investigation: C.F., Y.Y. and H.L.; writing-original draft preparation: C.F. and Y.Y.; writing-review and editing: C.F., Y.Y., X.Y. and H.L. All authors have read and agreed to the published version of the manuscript.Funding: This study is financially supported by the National Natural Science Foundation of China (No. 32270491, 31772507), the Natural Science Foundation of Hebei Province (No. C2022201005), the Excellent Youth Scientific Research and Innovation Team of Hebei University (No. 605020521005), and the Interdisciplinary Research Program of Natural Science of Hebei University (No. DXK202103).