https://orcid.org/0000-0002-2706-8166
Figures
Abstract
Continental shelf islands in Southeast Asia remain poorly studied with respect to insect diversity and evolutionary history. Here, we provide an integrative assessment of species diversity, DNA barcode variation, diversification dynamics, and historical biogeography of leaf beetles (Coleoptera: Chrysomelidae) on Cat Ba Island, northern Vietnam. Based on field surveys, 36 morphospecies operational taxonomic units (OTUs) belonging to 30 genera and five subfamilies were documented, with Galerucinae representing the most species-rich lineage. DNA barcoding of the mitochondrial COI gene generated 31 Barcode Index Numbers (BINs), most of which represent new records in the Barcode of Life Data System, highlighting substantial undocumented genetic diversity. COI-based analyses were further used to explore broad temporal and evolutionary patterns. Within this exploratory framework, divergence-time estimation under a relaxed molecular clock suggests that major lineages may have originated during the early–middle Miocene, predating the formation of the present-day island. Diversification analyses support relatively constant rates through time with low inferred extinction, consistent with expectations for continental shelf island systems shaped by repeated connectivity and isolation. Model-based biogeographic analyses indicate predominantly localized ancestral ranges, with Cat Ba Island and adjacent mainland regions playing recurrent roles in the assembly of the fauna. Together, these results provide baseline taxonomic and genetic data for a poorly known insular insect assemblage while offering a preliminary evolutionary context that should be interpreted with caution and that can serve as a foundation for future biodiversity monitoring and comparative studies in dynamic island–mainland systems.
Citation: Nguyen DT, Ho LT (2026) Leaf beetle diversity on a Southeast Asian continental island: Taxonomy, DNA barcoding, and preliminary evolutionary insights from Cat Ba Island, Vietnam. PLoS One 21(6): e0351706. https://doi.org/10.1371/journal.pone.0351706
Editor: Michael Schubert, Laboratoire de Biologie du Développement de Villefranche-sur-Mer, FRANCE
Received: January 29, 2026; Accepted: May 29, 2026; Published: June 18, 2026
Copyright: © 2026 Nguyen, Ho. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The DNA barcode sequences generated in this study are publicly available in the Barcode of Life Data Systems (BOLD) under project ID VAST. The corresponding sequences are publicly available in GenBank under accession numbers PZ276027–PZ276050, PZ270350–PZ270351, PZ275966–PZ275968, PX170664–PX170666, PX170669, and PX170674. All accession numbers are provided in the Methods section and Tables 2 and 3. All R scripts and analytical results are available in Zenodo (https://doi.org/10.5281/zenodo.19647930).
Funding: This research was supported by the Vietnam Academy of Science and Technology under grant number VAST04.08/25-26. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Islands have long been recognised as natural laboratories for investigating the processes that generate and maintain biodiversity, including colonisation, isolation, diversification and extinction [1,2]. Classical island biogeography theory emphasised the roles of island area and isolation in shaping species richness [3], whereas more recent frameworks incorporate geological history, temporal dynamics and evolutionary processes to explain patterns of insular diversity [4,5]. From a conservation perspective, islands are of particular importance because they often harbour unique species assemblages while being especially vulnerable to habitat disturbance and environmental change.
Continental shelf islands differ fundamentally from oceanic islands in both origin and evolutionary dynamics [6]. Rather than forming through volcanic processes, continental islands represent fragments of former mainland landscapes that have undergone episodic isolation driven primarily by sea‐level fluctuations [7]. As a consequence, their biotas are often assembled through repeated cycles of dispersal, persistence and limited in situ diversification, producing composite faunas with heterogeneous evolutionary histories [8,9]. Despite their relevance for understanding biodiversity assembly and persistence, the island systems have received comparatively little attention in insect conservation studies, particularly in Southeast Asia.
Southeast Asia is characterised by a complex geological and climatic history, marked by tectonic reorganisation and pronounced Quaternary sea‐level oscillations that repeatedly connected and isolated landmasses across the Sunda Shelf [7,10]. These processes are expected to have had profound effects on the distribution and diversity of terrestrial organisms. However, integrative studies that combine biodiversity surveys with molecular and evolutionary approaches remain scarce for insects, even though insects constitute the majority of terrestrial biodiversity and are increasingly recognised as being at risk from habitat loss and environmental change.
Leaf beetles (Coleoptera: Chrysomelidae) provide an excellent model for addressing these knowledge gaps. As one of the most speciose beetle families, Chrysomelidae comprise more than 35,000 described species worldwide [11–14]. Most species are phytophagous and often exhibit host‐plant specialisation, traits that can restrict dispersal and promote population differentiation under geographic isolation [15,16]. These characteristics make leaf beetles particularly suitable for assessing how island systems contribute to regional insect diversity and for generating baseline data relevant to conservation planning.
Cat Ba Island, located in the Gulf of Tonkin in northern Vietnam, forms part of the Ha Long–Cat Ba limestone archipelago and represents a continental shelf island system. Geological evidence indicates that this region has experienced repeated phases of subaerial exposure and marine inundation since the late Neogene, driven by glacial–interglacial sea‐level fluctuations [9,17]. Cat Ba Island is recognised for its high conservation value and is protected as both a national park and a UNESCO Biosphere Reserve. Nevertheless, the diversity and evolutionary composition of its insect fauna remain poorly understood, and no previous study has integrated species diversity data with molecular and evolutionary analyses for leaf beetles on the island.
In this study, we provide the first integrative assessment of leaf beetle diversity on Cat Ba Island by combining field surveys, morphological taxonomy and DNA barcoding with divergence‐time estimation and historical biogeography. Specifically, we aim to: (1) document patterns of species diversity and community composition; (2) generate DNA barcode reference data to support species identification and future biodiversity monitoring; and (3) place the island fauna in a broader evolutionary and spatial context. By explicitly linking biodiversity patterns with evolutionary history, this study provides essential baseline information for insect conservation on continental shelf islands in Southeast Asia.
Rather than aiming to resolve deep phylogenetic relationships or to test explicit macroevolutionary hypotheses, this study is designed to provide a robust baseline of taxonomic, genetic and evolutionary information for a poorly documented insular insect fauna. The analytical framework adopted here prioritises methodological transparency and data generation, with evolutionary inferences interpreted cautiously and at a coarse-grained level, consistent with the objectives of broad-scope biodiversity and data-driven studies. To our knowledge, this represents the first integrative study of insect diversity on a continental shelf island in Vietnam, providing a novel baseline for linking biodiversity patterns with evolutionary history in this region.
Materials and Methods
Study area and sampling
Cat Ba National Park is situated within the mainland–island complex of northern Vietnam and constitutes the core area of the Ha Long–Cat Ba Biosphere Reserve. The park is characterised by extensive limestone karst formations, evergreen tropical forests, and a high degree of habitat heterogeneity. Owing to these features, it harbours exceptionally rich biodiversity, including numerous endemic and threatened taxa, and plays a crucial role in regional conservation initiatives [18,19].
Field surveys were conducted at multiple forested sites within Cat Ba National Park and its surrounding areas during May and July 2025. Specimens were collected using three complementary methods: (i) direct hand collection without specialised tools; (ii) random sweeping of trees and shrubs along forest roads using an entomological sweep net; and (iii) beating of low branches and understory vegetation up to arm’s reach, with beetles dislodged onto a beating tray and subsequently collected using forceps. The geographic coordinates of all sampling sites are provided in Table 1 and illustrated in Fig 1.
(a) Regional map showing the location of Cat Ba Island within northern Vietnam (scale 1:9371119).(b) Detailed map of Cat Ba Island indicating sampling sites (scale 1:106670).
Maps were created using data from Natural Earth (public domain, http://www.naturalearthdata.com/) and visualized in QGIS
To minimise potential impacts on local populations and to comply with conservation guidelines, sampling at each site was restricted to a maximum of three individuals per morphospecies. All specimens collected from each locality were preserved separately in vials containing 90% ethanol and were labelled with the date of collection, locality, and collector’s name.
Field sampling was conducted in accordance with Vietnamese regulations. No specific permits were required for insect collection because the study did not involve protected or endangered species, and sampling was carried out in areas without restricted access. Permission to access sampling sites within Cat Ba National Park and surrounding areas was granted by local authorities.
Morphological identification and integrative taxonomic framework
Specimens were initially sorted into morphospecies based on diagnostic external morphological characters. Species-level identifications were achieved for morphospecies where feasible through detailed comparison with published taxonomic keys, original descriptions, and authoritative taxonomic revisions [20–38].
Morphological species hypotheses were subsequently assessed using DNA barcoding data where such data were available. For morphospecies represented by successfully sequenced specimens, mitochondrial COI sequences were compared with Barcode Index Numbers (BINs) in the Barcode of Life Data System (BOLD) to evaluate concordance between morphological delimitation and molecular clustering. In contrast, for morphospecies lacking DNA barcode data—either due to unsuccessful amplification or the absence of reference sequences in BOLD—species delimitation relied exclusively on morphological evidence.
Accordingly, molecular data were used as an auxiliary line of evidence rather than a universal criterion for species identification. This combined but flexible framework allowed integrative taxonomic inference where possible, while maintaining a morphology-based approach for taxa without available genetic information.
Taxonomic nomenclature at the family and subfamily levels follows [39], whereas generic and species-level classifications are adopted from [40].
Representative specimens were photographed using a Nikon DS-Fi3 digital camera mounted on a Nikon SMZ800N stereomicroscope. Image stacks obtained from multiple focal planes were processed in NIS-Elements and combined using Helicon Focus 7 to generate fully focused composite images suitable for detailed morphological examination.
DNA extraction, amplification and sequencing
Genomic DNA was extracted from whole specimens using the QIAamp® DNA Investigator Kit (QIAGEN) in accordance with the manufacturer’s instructions. Molecular analyses focused on the mitochondrial cytochrome c oxidase subunit I (COI) gene.
Mitochondrial COI was selected as the primary genetic marker because it represents the most widely used standard for insect DNA barcoding and allows direct comparability with existing reference data in BOLD and GenBank. Although reliance on a single mitochondrial locus may limit phylogenetic resolution, the present study employs COI as a pragmatic and widely accepted marker for species delimitation, baseline genetic characterisation and exploratory evolutionary analyses in a data-deficient system.
An approximately 658 bp fragment of COI was amplified using the universal primer pair LCO1490 and HCO2198 [41]. Polymerase chain reactions (PCRs) were conducted in a total volume of 25 µl, containing 2.5 µl of 10 × reaction buffer (Evrogen, Russia), 0.5 µl of 10 mM dNTP mix, 0.5 µl each of forward and reverse primers (10 µM), 1 µl of 25 mM Mg² ⁺ , 2 µl of template DNA, 0.2 µl of thermostable Taq DNA polymerase (Evrogen, Russia), and nuclease-free deionised water to volume.
Thermal cycling conditions comprised an initial denaturation step at 94 °C for 3 min, followed by 35 cycles of denaturation at 94 °C for 30 s, annealing at 42 °C for 40 s, and extension at 72 °C for 60 s, with a final extension at 72 °C for 5 min. Amplification success was assessed by electrophoresis on 1.5% agarose gels. Successfully amplified products were sequenced bidirectionally using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Foster City, CA, USA) with the same primers employed for PCR.
Following DNA extraction, voucher specimens were air-dried, mounted, and labelled with unique voucher numbers. All voucher material was deposited in the Institute of Biology (IOB), Vietnam Academy of Science and Technology (VAST), to ensure traceability and long-term preservation.
Forward and reverse Sanger sequence reads were assembled, edited, and aligned using Geneious Prime 2025.2 (Biomatters Ltd.). The resulting sequences were deposited in GenBank (NCBI) under accession numbers PZ276027–PZ276050, PZ270350, PZ270351, PZ275966, PZ275967, PZ275968, PX170664, PX170665, PX170666, PX170669, and PX170674 and in the Barcode of Life Data System (BOLD v4) [42]. Automatic assignment of sequences to Barcode Index Numbers (BINs) was performed through the BOLD platform [43]. All sequence data are publicly available.
Divergence time estimation
Twenty-five species of Chrysomelidae collected from Cat Ba Island were selected for divergence-time analyses based on the availability of high-quality, sufficiently long DNA sequences. Phylogenetic relationships and divergence times were inferred within a Bayesian framework using BEAST v10.5 [44]. Sequence evolution was modelled under the GTR + I nucleotide substitution model, with all parameters estimated from the data. Rate heterogeneity among lineages was accommodated using an uncorrelated lognormal relaxed molecular clock (UCLD), which allows substitution rates to vary independently across branches. Lineage diversification was modelled using a Birth–Death process.
Because no reliable fossil calibrations are currently available for the focal taxa, divergence-time estimation relied exclusively on molecular rate calibration. A prior on the mitochondrial substitution rate was applied to the COI gene, with a mean rate of 0.0115 substitutions per site per million years per lineage. This rate has been widely employed in beetle and insect phylogenetic studies and is commonly used for estimating interspecific divergence times [45,46]. Importantly, this value was implemented as a prior distribution on the clock model rather than as a fixed rate, thereby explicitly incorporating uncertainty associated with molecular rate variation.
Markov chain Monte Carlo (MCMC) analyses were run for a sufficient number of generations to ensure adequate sampling of the posterior distribution, with parameters sampled at regular intervals. Convergence diagnostics and mixing performance were examined in Tracer v1.7.2 [47], and all key parameters exhibited effective sample size (ESS) values > 200. Posterior distributions of trees were summarised as maximum clade credibility (MCC) trees after discarding the initial 25% of samples as burn-in using TreeAnnotator v10.5.0 [44].
The resulting time-calibrated phylogeny forms the basis for the divergence-time estimates presented in the Results section, with node ages and their associated 95% highest posterior density (HPD) intervals visualised and interpreted in an evolutionary and biogeographical context using FigTree v1.4.4 [48].
These analyses were conducted to place local species diversity within a broad temporal framework, rather than to test detailed macroevolutionary hypotheses or provide precise estimates of speciation and extinction dynamics. Given the moderate taxon sampling and reliance on a single mitochondrial marker, the results should be interpreted cautiously and are not intended as definitive tests of alternative macroevolutionary scenarios.
Inference of Diversification Dynamics of Chrysomelidae on Cat Ba Island
Diversification dynamics of Chrysomelidae on Cat Ba Island were investigated using complementary maximum-likelihood (ML) and Bayesian approaches to ensure robustness of inference. All analyses were based on a time-calibrated (ultrametric) phylogenetic tree comprising 25 extant taxa, inferred from Bayesian divergence-time analyses. The tree was imported into R v. 4.4.2 [49] and verified to be ultrametric prior to diversification analyses. Branch lengths represent divergence times (in millions of years) and were assumed to adequately reflect the evolutionary history of the focal clade.
Maximum-likelihood diversification analyses
ML analyses were conducted in R v. 4.4.2 using the package diversitree [50], which implements likelihood-based methods for estimating diversification parameters from time-calibrated phylogenies. Analyses were performed on the ultrametric tree, with likelihoods conditioned on survival of the clade to the present.
Two constant-rate diversification models were fitted. First, a Yule (pure-birth) model was applied, assuming a constant speciation rate (λ) through time and no extinction (μ = 0). Second, a constant-rate birth–death (BD) model was fitted, allowing both speciation (λ) and extinction (μ) to occur at constant rates. The Yule model included one free parameter (λ), whereas the BD model included two free parameters (λ and μ).
Model parameters were estimated by maximising the likelihood function using numerical optimisation. To ensure stable parameter estimation and avoid convergence on local optima, each model was fitted using multiple starting values for λ (and μ in the BD model). All optimisation runs converged on the same maximum-likelihood estimates.
Model fit was assessed using the Akaike Information Criterion (AIC), calculated as
, where k is the number of free parameters and log L is the maximum log-likelihood. Relative support for competing models was evaluated using differences in AIC (ΔAIC), with lower AIC values indicating better-supported models.
Bayesian diversification analyses
To further explore diversification dynamics and to test for potential temporal heterogeneity in diversification rates, a Bayesian framework was implemented using the R v. 4.4.2 package TESS v. 3.0 [51]. Two alternative models were analysed: (1) a constant-rate birth–death model, assuming time-homogeneous speciation (λ) and extinction (μ) rates, and (2) a rate-shift birth–death model, allowing diversification rates to vary through time.
Analyses were conducted on the reconstructed phylogeny while explicitly accounting for incomplete taxon sampling via a sampling probability parameter. Posterior distributions of diversification parameters were estimated using Markov chain Monte Carlo (MCMC) sampling. Model fit was evaluated using posterior log-likelihoods and AIC values, with ΔAIC > 10 interpreted as strong support for the model with the lower AIC.
These analyses are intended to provide a coarse-grained characterisation of diversification patterns rather than precise estimates of speciation and extinction rates.
Ancestral area reconstruction
Historical biogeographic analyses were conducted to infer ancestral range evolution and dispersal processes in Chrysomelidae using the R v. 4.4.2 package BioGeoBEARS [52]. Species distributions were coded into four discrete geographic areas: (1) Cat Ba Island, (2) mainland northern Vietnam, (3) adjacent regions of southern China, and (4) India. These areas were defined to reflect the major biogeographic units relevant to the evolutionary history of the studied lineages, capturing both continental source regions and insular systems while maintaining a tractable number of areas for likelihood-based inference.
Analyses were based on a mitochondrial COI dataset comprising 658 base pairs. A total of 75 Chrysomelidae species were included, with representatives from Cat Ba Island (25 species), mainland Vietnam (17 species), India (12 species), and mainland China (21 species). COI sequences for taxa from China, India, and nine species from mainland Vietnam were obtained from GenBank, whereas the remaining sequences were newly generated in this study (Table 2).
Ancestral range estimation was performed using the likelihood-based Dispersal–Extinction–Cladogenesis (DEC) model and its extension incorporating founder-event speciation (DEC + J). Both models were implemented on a time-calibrated ultrametric maximum clade credibility (MCC) tree inferred in BEAST, with branch lengths proportional to divergence time. Species distributions were coded as presence–absence across the predefined areas, and the maximum range size was constrained to three areas per lineage to reflect biologically realistic distributions.
All analyses assumed a fixed dispersal matrix with constant dispersal and extinction rates through time. Model parameters were estimated using maximum likelihood, and model fit was evaluated by comparing log-likelihood values and the Akaike Information Criterion corrected for small sample size (AICc). The comparison between DEC and DEC + J was explicitly designed to assess whether allowing cladogenetic founder-event dispersal improved model fit and altered inferred ancestral ranges, particularly in the context of continental–island biogeographic systems.
Rather than conducting an exhaustive comparison across all available BioGeoBEARS models, our analytical framework focused on evaluating the relative support for gradual range evolution (DEC) versus scenarios that additionally permit jump dispersal at speciation (DEC + J). This targeted approach was chosen to directly test hypotheses regarding the role of founder-event speciation in shaping present-day distribution patterns, while acknowledging ongoing discussions regarding the interpretation of the J parameter. Accordingly, results from DEC + J are interpreted cautiously and used primarily to identify qualitative spatial patterns rather than to infer explicit dispersal mechanisms.
The inclusion of the DEC + J model is intended to improve descriptive fit to the observed distribution patterns rather than to imply explicit founder-event dispersal processes. In line with recent discussions on the interpretation of the J parameter, results from DEC + J are treated as phenomenological and are used primarily to highlight qualitative spatial structuring of lineages rather than to infer specific dispersal mechanisms.
Results
Species diversity and taxonomic composition of Chrysomelidae on Cat Ba Island
All specimens collected during field surveys were assigned to morphospecies operational taxonomic units (OTUs). In total, 36 morphospecies OTUs were recognised, of which 16 were identified to species level, while the remaining OTUs were identified to genus level. These OTUs represent 30 genera belonging to five subfamilies of Chrysomelidae recorded from Cat Ba Island (Table 3, Figs 2–8).
(a, b) P. septemvittata; (c, d) G. tredecimaculata; (e, f) L. cyaneicollis; (g, h) L. egena; (i, j) L. impressa; (k, l) L. (L.) rondoniana.
(a, b) Melixanthus sp.; (c, d) Demotina sp. 1; (e, f) Pseudometaxis sp.; (g, h) H. granulosa.
(a, b) Colaspoides sp.; (c, d) Colasposoma sp.; (e, f) Basilepta sp.; (g, h) Cleoporus sp.; (i, j) Pagria sp.
(a, b) Altica sp.; (c, d) Hemipyxis sp.; (e, f) N. chrysomeloides; (g, h) Paradibolia sp.; (i) P. dalmani.
(a, b) Clitenella sp.; (c) Dercetina sp. 1; (d) Dercetina sp. 2; (e, f) Dercetina sp. 3; (g, h) Morphosphaera sp.
(a, b) Sermyloides sp.; (c, d) S. nebulosa; (e, f) Cassena sp.; (g, h) H. unicolor; (i, j) K. unicolor.
(a, b) Mimastra sp.; (c, d) Monolepta sp. 1; (e, f) Monolepta sp. 2; (g, h) T. (T.) bowringii; (i, j) T. (P.) nigribasalis; (k, l) O. decempunctata.
Species richness differed markedly among subfamilies. Galerucinae was the most diverse subfamily, comprising 21 species distributed across 17 genera. Eumolpinae was the second most diverse group, with eight species in eight genera, followed by Criocerinae, with four species in two genera, and Chrysomelinae, with two species in two genera. Cryptocephalinae exhibited the lowest diversity, being represented by a single species in one genus.
Most morphospecies showed restricted spatial distributions across sampling sites. H. unicolor was the most widely distributed species, occurring at 9 of the 21 sampling sites, followed by S. nebulosa (8/21 sites) and K. unicolor (5/21 sites). Hemipyxis sp. was recorded from three sampling sites, whereas Colaspoides sp., Colasposoma sp., and O. decempunctata were each collected from two sites. The remaining 29 morphospecies OTUs were each recorded from a single sampling locality.
DNA barcoding and genetic diversity
All OTUs were subjected to DNA extraction, amplification, and sequencing of the 658-bp fragment of the mitochondrial cytochrome c oxidase subunit I (COI) gene. However, successful amplification and sequencing were achieved for only 31 OTUs.
These sequences were submitted to the Barcode of Life Data System (BOLD) and were assigned to 31 Barcode Index Numbers (BINs). Among them, 25 were singletons and constitute newly recorded BINs in BOLD generated by this study (Table 3).
Analysis of sequence composition in BOLD showed that the mean nucleotide frequencies were 16.17% for G, 16.46% for C, 29.66% for A, and 35.71% for T. The mean GC content across sequences was 34.64%, with a marked GC bias at the first codon position (mean = 44.12%).
Divergence time estimates
The relaxed clock model was strongly supported, with a high coefficient of variation (CV ≈ 0.55), rejecting a strict molecular clock. Likelihood values were stable across analyses, and nucleotide substitution parameters were highly consistent, indicating robust phylogenetic signal. Effective sample size (ESS) values for all key parameters exceeded 200, indicating adequate convergence of the Markov chain Monte Carlo analyses.
Divergence time estimates obtained from BEAST analyses are summarised in the maximum clade credibility tree (Fig 9). The most recent common ancestor of the study group was dated to approximately 19.75 Ma (95% HPD: 14.86–26.47 Ma), corresponding to the early–middle Miocene. Major lineages diverged shortly after the origin of the group, followed by successive diversification toward the present. Deeper nodes exhibited broader uncertainty, whereas more recent splits showed narrower credibility intervals.
Node ages are shown as mean estimates. The estimated age of the root is 19.75 Ma (95% highest posterior density [HPD]: 14.86–26.47 Ma). Horizontal bars at nodes represent 95% HPD intervals of divergence times. Numbers at nodes indicate posterior probabilities. Node ages are colour-coded according to the scale shown on the left, and the time scale (Ma) is shown along the bottom axis.
Diversification dynamics
Maximum-likelihood constant-rate models.
ML analyses revealed broadly similar patterns of lineage accumulation under both constant-rate models. Under the Yule (pure-birth) model, the maximum log-likelihood was −26.54, with an estimated speciation rate of λ ≈ 0.079. The constant-rate birth–death model yielded a slightly lower maximum log-likelihood (−26.78), with an estimated speciation rate of λ ≈ 0.086 and an extinction rate μ close to zero.
Model comparison based on AIC slightly favoured the Yule model (ΔAIC = 2.48), providing moderate support for the simpler pure-birth model (Table 4).
Bayesian analyses using TESS.
Bayesian analyses implemented in TESS further supported a constant-rate birth–death model over a rate-shift model. The constant-rate model achieved a higher posterior log-likelihood and a substantially lower AIC, whereas the rate-shift model was associated with broader HPD intervals and increased parameter uncertainty.
Under the constant-rate model, the posterior mean speciation rate was λ = 0.104 (95% HPD: 0.0645–0.140), while extinction rates were comparatively low (μ = 0.0177; 95% HPD: 0.00022–0.0502), yielding a net diversification rate of 0.086 (95% HPD: 0.0453–0.1282). In contrast, the rate-shift model showed broad HPD intervals for all parameters, including negative values of net diversification (r), indicating substantial uncertainty and poor parameter identifiability. (Table 4).
Both analytical frameworks (ML and Bayesian analyses using TESS) broadly support the conclusion that diversification in the study lineage is primarily driven by speciation, with extinction being weakly supported and poorly constrained.
Inferred biogeographic history of Chrysomelidae on Cat Ba Island.
Model comparison based on likelihood and information criteria favoured the DEC + J model over the DEC model (Table 5). The DEC + J model achieved a higher log-likelihood and a lower AICc value despite including one additional free parameter.
Under both models, ancestral range reconstructions inferred predominantly single-area ancestral ranges, with Cat Ba Island and adjacent mainland regions frequently reconstructed at deep and intermediate nodes. The DEC + J model inferred more geographically restricted ancestral ranges at several internal nodes compared with the DEC model. Taken together, the inferred patterns are consistent with spatially structured range inheritance and repeated dispersal into island regions, as captured by the DEC + J framework.
Ancestral state reconstructions under the DEC + J model are illustrated in the BioGeoBEARS plot (Fig 10). Cat Ba Island (A) and China (C) were most frequently inferred as ancestral areas at deeper nodes, whereas mainland Vietnam (B) was mainly reconstructed at shallow nodes. India (D) was inferred sporadically and did not represent a dominant ancestral region. Most nodes were reconstructed as single-area states rather than widespread multi-area distributions.
Colors represent geographic areas: A = China, B = mainland Vietnam, C = Cat Ba island, and D = India. Pie charts at nodes indicate relative probabilities of ancestral ranges. The reconstruction highlights the predominance of Cat Ba Island and China as ancestral areas and supports patterns consistent with founder-event-like processes.
Discussion
The present study provides an integrative assessment of species diversity, genetic structure, evolutionary history and biogeographic patterns of Chrysomelidae on Cat Ba Island, a continental shelf island in northern Vietnam. By combining morphological surveys, DNA barcoding, divergence-time estimation and model-based biogeographic analyses, the results offer a coherent framework for understanding how geological history, dispersal processes and ecological heterogeneity have shaped an insular leaf beetle assemblage.
The present study documents a diverse assemblage of Chrysomelidae on Cat Ba Island, comprising 36 morphospecies OTUs assigned to 30 genera and five subfamilies. This level of diversity is comparable to that reported from other island faunas in Vietnam, such as Phu Quoc Island with 44 OTUs [58], as well as islands in other tropical regions, including a west coast island of Sabah, Malaysia (68 OTUs [59]), and the island of St Vincent (47 species [60]).
Galerucinae was the most species-rich subfamily, accounting for more than half of the recorded OTUs. This dominance is consistent with both regional and global patterns, as Galerucinae is widely recognised as the most diverse and ecologically versatile lineage within Chrysomelidae, particularly in forested habitats [58,59,61]. The high generic diversity observed within Galerucinae on Cat Ba Island further reflects the broad host-plant associations and ecological plasticity characteristic of this group.
Eumolpinae represented the second most diverse subfamily, with eight species distributed across eight genera. Members of this subfamily are predominantly associated with woody vegetation and forest understorey plants, habitats that are well developed within the limestone and evergreen forests of Cat Ba Island. Comparable levels of Eumolpinae diversity have been reported from other tropical forest systems in the region [14]. In contrast, Chrysomelinae and Cryptocephalinae were represented by relatively few taxa. The low diversity of Cryptocephalinae, represented by a single species, may reflect both genuine ecological rarity and limitations of sampling, as many species are canopy-dwelling or exhibit seasonal activity patterns that reduce detectability [62,63].
At the spatial scale of the island, most morphospecies exhibited restricted distributions and were recorded from a single sampling locality. Such patterns are commonly observed in tropical beetle assemblages and are often attributed to habitat heterogeneity, patchy host-plant distribution and limited dispersal ability [64,65]. Nevertheless, several species showed relatively broad distributions across Cat Ba Island. H. unicolor, S. nebulosa and K. unicolor were recorded from multiple sampling sites, suggesting broader ecological tolerances and the ability to exploit widely distributed host plants. Similar contrasts between widespread and narrowly distributed species have been reported in leaf beetle communities from other Southeast Asian islands and mainland regions [58,59].
Species such as Hemipyxis sp., Colaspoides sp., Colasposoma sp. and O. decempunctata displayed intermediate distribution ranges, occurring at two or three sampling sites. These taxa may be more closely associated with specific habitat conditions or host plants, although further ecological data are required to clarify these relationships. Overall, the combination of a small number of widespread species and a large proportion of narrowly distributed taxa suggests that the chrysomelid fauna of Cat Ba Island is shaped by both dispersal ability and habitat specialisation, a pattern characteristic of insular ecosystems where geographic isolation and environmental heterogeneity promote high beta diversity [3,4].
To complement morphological diversity patterns, DNA barcoding of the COI gene generated sequence data for 31 species, resulting in the assignment of 31 Barcode Index Numbers (BINs), of which 25 were singleton BINs newly recorded in the Barcode of Life Data Systems (BOLD). The high proportion of singleton and novel BINs clearly reflects a pronounced deficiency of publicly available DNA barcode data for Vietnamese leaf beetles. Such gaps in reference databases constrain accurate species delimitation, hinder comparative biogeographic analyses, and reduce the effectiveness of biodiversity assessments in the region. From a conservation perspective, the incomplete documentation of Chrysomelidae diversity is particularly concerning for island ecosystems such as Cat Ba, where habitat fragmentation and anthropogenic pressures may disproportionately affect narrowly distributed or cryptic taxa. By contributing new barcode records to BOLD, this study represents an important step towards closing these data gaps and provides an initial genetic reference framework to support taxonomic research, biodiversity monitoring, and conservation planning for the Chrysomelidae fauna of Vietnam [42,66,67]. The observed nucleotide composition and GC bias at the first codon position are consistent with patterns commonly reported for insect mitochondrial genomes and COI barcode datasets [58,59] supporting the reliability of the generated barcode data and reinforcing the utility of DNA barcoding as a tool for species delimitation and future taxonomic research [67].
These analyses were conducted to place the observed diversity within a broad temporal and evolutionary context, rather than to test detailed macroevolutionary hypotheses. Within this exploratory framework, COI-based temporal analyses suggest that the major lineages of Chrysomelidae represented on Cat Ba Island may have originated during the early–middle Miocene. Divergence-time estimation under a Birth–Death model with an uncorrelated lognormal relaxed clock suggests rejection of a strict molecular clock, reflecting substantial among-lineage rate heterogeneity typical of insect mitochondrial datasets [46,68]. The inferred age of the most recent common ancestor at approximately 19.75 Ma is consistent with a period of major tectonic and climatic reorganisation in Southeast Asia, including intensified uplift and the establishment of monsoonal climates [9,69]. Although Cat Ba Island itself emerged much later, these results are consistent with the possibility that diversification was initiated in mainland prior to the formation of the present-day island landscape.
Exploratory analyses of diversification dynamics further suggest that the assembly of the Cat Ba chrysomelid fauna may not have involved pronounced bursts of diversification. Both maximum-likelihood and Bayesian approaches are broadly consistent with relatively constant diversification rates through time, with low inferred extinction. Under such conditions, pure-birth or constant-rate birth–death models provide parsimonious and statistically robust descriptions of lineage accumulation, particularly for phylogenies of moderate size [70,71]. The lack of strong support for rate-shift models may indicate that diversification did not proceed via episodic radiations linked to discrete environmental perturbations or key innovations [72,73]. Instead, the positive net diversification rate is consistent with sustained lineage accumulation without strong evidence for diversity-dependent slowdown.
This pattern is broadly consistent with the geological history of continental shelf islands such as Cat Ba, where cyclical sea-level fluctuations during the Quaternary repeatedly connected and fragmented island and mainland habitats [7,10]. These dynamics likely facilitated dispersal and recolonisation, whereas interglacial isolation may have promoted divergence without leading to extensive in situ radiation. These dynamics are expected to produce relatively homogeneous diversification trajectories over time rather than pronounced rate heterogeneity.
Biogeographic reconstructions provide preliminary insight into the spatial processes underlying this evolutionary history. Model comparison favoured the DEC + J framework in a statistical sense over the standard DEC model, indicating that spatially discrete range inheritance patterns captured by founder-event processes provide a better statistical fit to the data [74]. Nevertheless, as highlighted by earlier methodological assessments, the biological meaning of the j parameter requires careful interpretation, especially in cases where dispersal and extinction estimates lie near their boundary conditions [75]. In this context, DEC + J is best viewed as a phenomenological model describing patterns of range inheritance rather than a literal representation of dispersal mechanisms.
Ancestral area reconstructions suggest predominantly single-area ancestral ranges, with Cat Ba Island and China most frequently reconstructed at deep and intermediate nodes. Mainland Vietnam was mainly inferred at shallow nodes, suggesting more recent colonisation events, whereas India appeared sporadically and did not constitute a dominant ancestral region. These patterns are consistent with a spatially structured evolutionary history characterised by recurrent colonisation and limited range expansion, rather than widespread ancestral distributions. The recurrent inference of Cat Ba Island and China as ancestral areas suggests that these regions may have played central roles in the assembly of the fauna through multiple, independent colonisation events rather than a single directional dispersal pathway [4].
Taken together, the temporal framework, diversification analyses and biogeographic reconstructions are broadly consistent with a scenario in which the chrysomelid fauna of Cat Ba Island originated from mainland lineages during the Miocene and was subsequently assembled through repeated dispersal events linked to changes in continental shelf configuration. Diversification appears to have proceeded gradually at broadly constant rates, while founder-event-like range inheritance may have contributed to the spatial structuring of lineages [74]. In such systems, alternating phases of connection and fragmentation may promote lineage persistence and incremental diversification rather than rapid adaptive radiations [4,76].
For phytophagous beetles, the pronounced ecological heterogeneity of Cat Ba Island, including limestone-dominated forests and a diverse flora of potential host plants, may have provided additional opportunities for ecological differentiation at local scales. These conditions are consistent with a model of gradual diversification driven by the interaction between geological dynamics and fine-scale ecological variation, rather than by single pulses of ecological opportunity.
Despite its integrative approach, this study is subject to limitations. Moderate taxon sampling and the absence of fossil calibrations may constrain the detection of subtle shifts in diversification dynamics [70,71]. Broader geographic sampling across Cat Ba Island and adjacent mainland regions would improve inference within a dynamic shelf-island context [7,10]. In addition, reliance on mitochondrial COI data alone may be sensitive to lineage-specific rate variation, introgression, incomplete lineage sorting, and purifying selection, all of which may obscure true species boundaries and evolutionary relationships [46,66,77,78]. Furthermore, phylogenetic patterns inferred from mitochondrial data may differ from those based on nuclear markers, a phenomenon known as mitonuclear discordance [79]. Recent studies have also shown that closely related species may exhibit little or no mtDNA divergence despite being clearly distinct at nuclear loci [80], highlighting the need for caution when interpreting COI-based results. Future analyses incorporating nuclear markers are therefore expected to substantially strengthen phylogenetic inference. Finally, integrating ecological data, particularly host-plant associations, would allow more explicit tests of how ecological differentiation interacts with dispersal and isolation to shape biogeographic patterns [4,15].
Conclusions
This study presents the first integrative synthesis of species diversity, genetic structure and evolutionary history of Chrysomelidae on Cat Ba Island. The available evidence suggests that the present-day fauna is composed of lineages that originated on the mainland during the Miocene and were subsequently assembled through repeated episodes of dispersal and isolation associated with continental shelf dynamics.
Diversification patterns inferred from mitochondrial COI data are broadly consistent with relatively constant rates through time, while spatially structured range inheritance appears to have contributed to present-day distribution patterns. Although these inferences are necessarily coarse-grained, they provide a useful evolutionary context for interpreting current biodiversity patterns.
Overall, the baseline taxonomic, genetic and biogeographic data generated in this study constitute a reference framework for future biodiversity monitoring, conservation assessments and comparative studies of insular insect faunas in Southeast Asia.
Acknowledgments
We express our gratitude to the authorities responsible for Cat Ba National Park for their support in facilitating this research.
Permission to reproduce material from other sources
All figures and tables are original. Fig 1 was created using public domain data from Natural Earth and QGIS software; therefore, no permission is required.
References
- 1.
Darwin C. On the origin of species. London: John Murray. 1859.
- 2.
Wallace AR. Island life: or, the phenomena and causes of insular faunas and floras, including a revision and attempted solution of the problem of geological climates. London: Macmillan. 1880.
- 3.
MacArthur RH, Wilson EO. The theory of island biogeography. Princeton: Princeton University Press. 1967.
- 4. Whittaker RJ, Triantis KA, Ladle RJ. A general dynamic theory of oceanic island biogeography. J Biogeogr. 2008;35(6):977–94.
- 5. Gillespie RG, Roderick GK. Evolution: Geology and climate drive diversification. Nature. 2014;509(7500):297–8. pmid:24828187
- 6. Ali JR. Islands as biological substrates: continental. J Biogeogr. 2018;45:1003–18.
- 7. Woodruff DS. Biogeography and conservation in Southeast Asia: how 2.7 million years of repeated environmental fluctuations affect today’s patterns and the future of the remaining refugial-phase biodiversity. Biodivers Conserv. 2010;19(4):919–41.
- 8. Lohman DJ, de Bruyn M, Page T, von Rintelen K, Hall R, Ng PKL, et al. Biogeography of the Indo-Australian Archipelago. Annual Review of Ecology, Evolution, and Systematics. 2011;42:205–26.
- 9.
Hall R. Sundaland and Wallacea: geology and palaeogeography. In: Gower DJ, ed. Biotic evolution and environmental change in Southeast Asia. Cambridge: Cambridge University Press. 2012. 32–78.
- 10. Voris HK. Maps of Pleistocene sea levels in Southeast Asia: shorelines, river systems and time durations. Journal of Biogeography. 2000;27(5):1153–67.
- 11.
Schmitt M. Phylogenetic system of chrysomelidae. In: Jolivet P, Cox ML, ed. Chrysomelidae biology. Amsterdam: SPB Academic Publishing. 1996. 57–96.
- 12.
Futuyma DJ. Preface. In: Jolivet P, Santiago-Blay JA, Schmitt M, ed. New developments in the biology of Chrysomelidae. The Hague: SPB Academic Publishing. 2004. 19–20.
- 13. Ślipiński SA, Leschen RAB, Lawrence JF. Order Coleoptera Linnaeus, 1758. In: Zhang ZQ, ed. Animal biodiversity: an outline of higher-level classification and survey of taxonomic richness. 2011. p. 203–8.
- 14. Jolivet P. Together with 30 years of Symposia on Chrysomelidae! Memories and personal reflections on what we know more about leaf beetles. Zookeys. 2015;(547):35–61. pmid:26798313
- 15. Farrell BD, Sequeira AS. Evolutionary rates in the adaptive radiation of beetles on plants. Evolution. 2004;58(9):1984–2001. pmid:15521456
- 16. Gómez-Zurita J, Hunt T, Kopliku F, Vogler AP. Recalibrated tree of leaf beetles indicates independent diversification. PLoS One. 2007;2:e360.
- 17.
Hutchison CS. Geology of Southeast Asia. Kuala Lumpur: Geological Society of Malaysia. 2007.
- 18.
UNESCO. Cat Ba Biosphere Reserve, Vietnam. 2004. https://www.unesco.org/en/mab/cat-ba
- 19.
Sodhi NS, Brook BW. Southeast Asian biodiversity in crisis. Cambridge: Cambridge University Press. 2006.
- 20.
Beenen R. Galerucinae. In: Löbl I, Smetana A, ed. Catalogue of Palaearctic Coleoptera. Stenstrup: Apollo Books. 2010. 443–91.
- 21. Bezděk J. Revisional study on the genus Mimastra (Coleoptera: Chrysomelidae: Galerucinae). Acta Entomol Mus Natl Pragae. 2009;49:819–40.
- 22. Bezděk J. Taxonomic and faunistic notes on Oriental and Palaearctic Galerucinae and Cryptocephalinae. Genus. 2012;23:375–418.
- 23. Kimoto S, Gressitt JL. Chrysomelidae of Thailand, Cambodia, Laos and Vietnam I. Pac Insects. 1979;20:191–256.
- 24. Kimoto S, Gressitt JL. Chrysomelidae of Thailand, Cambodia, Laos and Vietnam II. Pac Insects. 1981;23:286–391.
- 25. Kimoto S, Gressitt JL. Chrysomelidae of Thailand, Cambodia, Laos and Vietnam III. Esakia. 1982;18:1–141.
- 26. Kimoto S. Chrysomelidae of Thailand, Cambodia, Laos and Vietnam IV. Esakia. 1989;27:1–241.
- 27. Kimoto S. Chrysomelidae of Thailand, Laos and Vietnam V. Bull Inst Comp Stud Int Cult Soc. 1998;21:1–88.
- 28. Tishechkin AK, Konstantinov AS, Bista S, Pemberton RW, Center TD. Review of the continental Oriental species of Lilioceris Reitter (Coleoptera, Chrysomelidae, Criocerinae) closely related to Lilioceris impressa (F.). Zookeys. 2011;(103):63–83. pmid:21747681
- 29. Medvedev LN. New Chrysomelidae from Southeast Asia. Annales Historico-Naturales Musei Nationalis Hungarici. 1998;90:163–74.
- 30. Medvedev LN. Chrysomelidae from the Nepal Himalayas. Stuttgarter Beiträge zur Naturkunde A. 2000;616:1–32.
- 31. Medvedev LN. Revision of the genus Colaspoides. Russian Entomological Journal. 2003;12:257–97.
- 32.
Medvedev LN. Alticinae of Indochina. Moscow: KMK Scientific Press. 2009.
- 33. Medvedev LN. Chrysomelidae of high mountain regions of NW Vietnam. Russian Entomological Journal. 2009;18:201–8.
- 34. Medvedev LN. New and poorly known Oriental Chrysomelidae. Vernate. 2015;34:319–35.
- 35. Moseyko AG. Notes on Asiatic Eumolpinae (Coleoptera, Chrysomelidae). Entomological Review. 2020;100:843–62.
- 36. Moseyko AG, Medvedev LN. A review of the chrysomelid genus Hyperaxis Gemm. et Har., 1874 (Coleoptera, Chrysomelidae: Eumolpinae) from Vietnam. Entomological Review. 2017;97:624–42.
- 37. Romantsov PV. New species of the subfamily Galerucinae (Coleoptera: Chrysomelidae) from South-East Asia. Cauc Entomol Bull. 2018;14(1):47–53.
- 38. Romantsov PV. Two new species and new records of Galerucinae (Coleoptera: Chrysomelidae) from Vietnam. Cauc Entomol Bull. 2020;16(1):79–84.
- 39. Bouchard P, Bousquet Y, Davies AE, Alonso-Zarazaga MA, Lawrence JF, Lyal CHC, et al. Family-group names in Coleoptera (Insecta). Zookeys. 2011;(88):1–972. pmid:21594053
- 40. Seeno TN, Wilcox JA. Leaf beetle genera (Coleoptera: Chrysomelidae). Entomography. 1982;1:1–222.
- 41. Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R. DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I. Mol Mar Biol Biotechnol. 1994;3:294–9.
- 42. Ratnasingham S, Hebert PDN. BOLD: the Barcode of life data system. Mol Ecol Notes. 2007;7(3):355–64.
- 43. Ratnasingham S, Hebert PDN. A DNA-based registry for all animal species: the barcode index number (BIN) system. PLoS One. 2013;8(7):e66213. pmid:23861743
- 44. Baele G, Ji X, Hassler GW, McCrone JT, Shao Y, Zhang Z, et al. BEAST X for Bayesian phylogenetic, phylogeographic and phylodynamic inference. Nat Methods. 2025;22(8):1653–6. pmid:40624354
- 45. Brower AVZ. Rapid morphological radiation and convergence among races of the butterfly Heliconius erato. Proc Natl Acad Sci USA. 1994;91(14):6491–5.
- 46. Papadopoulou A, Anastasiou I, Vogler AP. Revisiting the insect mitochondrial molecular clock: the mid-Aegean trench calibration. Mol Biol Evol. 2010;27(7):1659–72. pmid:20167609
- 47. Rambaut A, Drummond AJ, Xie D, Baele G, Suchard MA. Posterior summarization in Bayesian Phylogenetics Using Tracer 1.7. Syst Biol. 2018;67(5):901–4. pmid:29718447
- 48. Rambaut A. FigTree v1.4.4. http://tree.bio.ed.ac.uk/software/figtree
- 49.
R Core Team. R: a language and environment for statistical computing. Vienna: R Foundation for Statistical Computing. 2025.
- 50. FitzJohn RG. Diversitree: comparative phylogenetic analyses of diversification in R. Methods Ecol Evol. 2012;3(6):1084–92.
- 51. Höhna S, May MR, Moore BR. TESS: Bayesian inference of lineage diversification rates. Bioinformatics. 2016;32:789–91.
- 52. Matzke NJ. Probabilistic historical biogeography: new models for founder-event speciation, imperfect detection, and fossils allow improved accuracy and model-testing. Frontiers of Biogeography. 2013;5(4).
- 53. Center TD, Rayamajhi M, Dray FA, Madeira PM, Witkus G, Rohrig E, et al. Host range validation, molecular identification and release and establishment of a Chinese biotype of the Asian leaf beetleLilioceris cheni(Coleoptera: Chrysomelidae: Criocerinae) for control ofDioscorea bulbiferaL. in the southern United States. Biocontrol Science and Technology. 2013;23(7):735–55.
- 54. Nie RE, Wei J, Zhang SK, Vogler AP, Wu L, Konstantinov AS, et al. Diversification of mitogenomes in three sympatric Altica flea beetles (Insecta, Chrysomelidae). Zool Scr. 2019;48(6):743–59.
- 55. Nie R-E, Breeschoten T, Timmermans MJTN, Nadein K, Xue H-J, Bai M, et al. The phylogeny of Galerucinae (Coleoptera: Chrysomelidae) and the performance of mitochondrial genomes in phylogenetic inference compared to nuclear rRNA genes. Cladistics. 2018;34(2):113–30. pmid:34645082
- 56. Lalrinfeli R, Behere GT, Firake DM, Sharma B, Banerjee A, Rajesh T. Development of DNA barcodes for major insect pests of cole crops in mid-hills of Meghalaya. Int J Curr Microbiol Appl Sci. 2019;8(2):789–99.
- 57. Nguyen DT. The DNA barcoding data and genetic distance of leaf beetles (Coleoptera, Chrysomelidae) in Vietnam. Acad J Biol. 2020;42(4):25–31.
- 58. Nguyen DT. Exploring the diversity of leaf beetles (coleoptera: chrysomelidae) on the islands of Vietnam: a survey of Phu Quoc Island, South of Vietnam. Int J Trop Insect Sci. 2022;42(5):3525–39.
- 59. Yeong K-C, Takizawa H, Liew T-S. Investigating leaf beetles (Coleoptera, Chrysomelidae) on the west coast islands of Sabah via checklist-taking and DNA barcoding. PeerJ. 2018;6:e5811. pmid:30386703
- 60. Peck SB. The beetles of the island of St. Vincent, Lesser Antilles (Insecta: Coleoptera): diversity and distributions. Insecta Mundi. 2010;144:1–77.
- 61. Nguyen DT, Gómez-Zurita J. Subtle ecological gradient triggers high species turnover. PLoS One. 2016;11:e0156840.
- 62.
Reid CAM. Cladistic analysis of Chrysomelidae. In: Pakaluk J, Ślipiński SA, ed. Biology, phylogeny and classification of Coleoptera. Warszawa: PAN. 1995. 559–631.
- 63.
Jolivet P, Verma KK. Biology of leaf beetles. Andover: Intercept Publishers. 2002.
- 64.
Ødegaard F. Host specificity, alpha- and beta-diversity of phytophagous beetles in two tropical forests in Panama. In: Hawksworth DL, Bull AT, ed. Arthropod diversity and conservation. Dordrecht: Springer. 2006. 69–91.
- 65. Novotny V, Miller SE, Hrcek J, Baje L, Basset Y, Lewis OT, et al. Insects on plants: explaining the paradox of low diversity within specialist herbivore guilds. Am Nat. 2012;179(3):351–62. pmid:22322223
- 66. Hebert PDN, Cywinska A, Ball SL, deWaard JR. Biological identifications through DNA barcodes. Proc Biol Sci. 2003;270(1512):313–21. pmid:12614582
- 67. Pentinsaari M, Hebert PDN, Mutanen M. Barcoding beetles: a regional survey of 1872 species reveals high identification success and unusually deep interspecific divergences. PLoS One. 2014;9(9):e108651. pmid:25255319
- 68. Drummond AJ, Ho SYW, Phillips MJ, Rambaut A. Relaxed phylogenetics and dating with confidence. PLoS Biol. 2006;4(5):e88. pmid:16683862
- 69. Morley RJ. Assembly and division of the South and South-East Asian flora in relation to tectonics and climate change. J Trop Ecol. 2018;34(4):209–34.
- 70. Nee S, Holmes EC, May RM, Harvey PH. Extinction rates can be estimated from molecular phylogenies. Philos Trans R Soc Lond B Biol Sci. 1994;344(1307):77–82. pmid:8878259
- 71. Rabosky DL. Extinction rates should not be estimated from molecular phylogenies. Evolution. 2010;64(6):1816–24. pmid:20030708
- 72. Rabosky DL, Lovette IJ. Explosive evolutionary radiations: decreasing speciation or increasing extinction through time?. Evolution. 2008;62(8):1866–75. pmid:18452577
- 73. Stadler T. Mammalian phylogeny reveals recent diversification rate shifts. Proc Natl Acad Sci U S A. 2011;108(15):6187–92. pmid:21444816
- 74. Matzke NJ. Model selection in historical biogeography reveals that founder-event speciation is a crucial process in Island Clades. Syst Biol. 2014;63(6):951–70. pmid:25123369
- 75. Ree RH, Sanmartín I. Conceptual and statistical problems with the DEC J model of founder-event speciation and its comparison with DEC via model selection. J Biogeogr. 2018;45:741–9.
- 76. Ali JR. Islands as biological substrates: classification of the biological assemblage components and the physical island types. J Biogeogr. 2017;44:984–94.
- 77. Ho SYW, Phillips MJ. Accounting for calibration uncertainty in phylogenetic estimation of evolutionary divergence times. Syst Biol. 2009;58(3):367–80. pmid:20525591
- 78. Funk DJ, Omland KE. Species-level paraphyly and polyphyly: frequency, causes, and consequences, with insights from animal mitochondrial DNA. Annu Rev Ecol Evol Syst. 2003;34:397–423.
- 79. Bocek M, Motyka M, Kusy D, Bocak L. Genomic and Mitochondrial Data Identify Different Species Boundaries in Aposematically Polymorphic Eniclases Net-Winged Beetles (Coleoptera: Lycidae). Insects. 2019;10(9):295. pmid:31514482
- 80. Duran DP, Laroche RA, Roman SJ, Godwin W, Herrmann DP, Bull E, et al. Species delimitation, discovery and conservation in a tiger beetle species complex despite discordant genetic data. Sci Rep. 2024;14(1):6617. pmid:38503840