Scarabaeoidea constitute one of the most speciose radiations of Coleoptera, comprising approximately 35,000 described species distributed across 13 families (Browne and Scholtz 2002; Ahrens et al. 2014; Scholtz and Grebennikov 2016). Despite extensive taxonomic work on adults, the larval stages of many lineages remain insufficiently documented. This gap is striking given the ecological importance of scarabaeoid larvae, which occupy a wide range of trophic niches—including saproxylic (Huang 2018; Ulyshen 2018), saprophagous (Micó and Galante 2003; Rodrigues et al. 2018), coprophagous (Cabrero-Sañudo and Zardoya 2004; Scholtz et al. 2009), and phytophagous (Zhang 1984; Ansari et al. 2006; Harrison and Wingfield 2016; Long et al. 2024). These diverse dietary modes contribute substantially to ecosystem functioning, such as organic matter turnover (Nichols et al. 2008; Doube 2018), but also include species of major agricultural concern due to their feeding on subterranean plant organs (Cave and Ratcliffe 2008; Chen et al. 2025).

Mouthparts, which mediate food acquisition and processing, represent one of the most functionally and phylogenetically informative components of larval morphology (Blanke et al. 2015; Krenn 2019). Although most insect larvae possess generalized chewing-type mouthparts (Stehr 1987; Angelini and Smith 2019), substantial variation in structure is evident among scarabaeoid lineages. Historically, these characters have played important roles in larval taxonomy and identification (Hayes 1928, 1929; Böving 1942; Medvedev 1952; Ritcher 1966; Zhang 1984; Sawada 1991), and contemporary studies continue to underscore their relevance for understanding lineage differentiation (Šípek 2010; Sousa et al. 2018). Yet fewer than 2% of insect larvae have been formally described (Lawrence et al. 2011; Šípek and Král, 2012), leaving the breadth of larval mouthpart diversity, particularly in Scarabaeoidea, far from comprehensively documented.

A persistent challenge in interpreting mouthpart morphology lies in disentangling homology from homoplasy. Recent studies have revealed that morphologically similar structures may evolve independently in unrelated scarabaeoid groups (Jia et al. 2021; Zhang et al. 2024), while considerable structural divergence can occur within closely related clades (Fang et al. 2018; Sun et al. 2024). Such patterns imply that convergent evolution may be widespread, likely driven by dietary transitions or shifts in ecological niche. Without a phylogenetic context, it remains difficult to ascertain whether a character state reflects shared ancestry (synapomorphy) or is the result of convergent or parallel evolution (homoplasy). Consequently, systematic and phylogenetically informed comparative analyses are essential for clarifying the evolutionary origins of larval mouthpart traits across Scarabaeoidea.

To address these gaps, the present study provides detailed examinations of third-instar larval mouthparts for ten species representing major lineages of Scarabaeoidea, supplemented by twelve additional species for which both morphological descriptions and mitochondrial genomes are available. In total, seventeen mouthpart characters were analyzed within a phylogenetic framework to evaluate their evolutionary polarity and homology. Specifically, we aimed to identify mouthpart characters that may represent candidate synapomorphies, clarify the distribution of homoplastic traits across lineages, and assess how specific morphological features relate to larval feeding habits. These analyses yield new insights into the evolutionary diversification of scarabaeoid larval mouthparts and provide a comparative framework for studies of insect functional morphology and character evolution.

This study provides the first comparative SEM-based assessment of larval mouthparts across ten scarabaeoid lineages and evaluates seventeen characters within an explicit molecular phylogenetic framework. Three of these characters were recovered as putative synapomorphies within the sampled framework, being potentially informative for Scarabaeidae, for the clade Cetoniinae + (Rutelinae + Dynastinae), and for the coprophagous Aphodiinae + Scarabaeinae, respectively. Several additional traits show repeated evolutionary origins and exhibit clear associations with feeding mode, illustrating how larval mouthpart morphology tracks functional diversification across major scarabaeoid lineages.

Extensive phylogenetic studies of Scarabaeoidea have relied on diverse datasets, including Sanger loci and multilocus DNA sequences (Caterino et al. 2005; Smith et al. 2006; Ahrens et al. 2014; McKenna et al. 2015; Song and Zhang 2018; Ahrens et al. 2020; Beza-Beza et al. 2020; Cai et al. 2022; Dietz et al. 2023) as well as morphology-based character matrices (Browne and Scholtz 2002; Grebennikov and Scholtz 2004; Lawrence et al. 2011). The topology obtained here is broadly congruent with previous mitogenomic studies (Song and Zhang 2018; Ayivi et al. 2021; Guo et al. 2022), and with multi-locus and genomic reconstructions (McKenna et al. 2019; Cai et al. 2022; Dietz et al. 2023), in recovering a major split within Scarabaeidae between the coprophagous clade Aphodiinae + Scarabaeinae and the phytophagous scarabs, and in placing Sericinae as the earliest-diverging lineage of the phytophagous assemblage, with Cetoniinae + (Dynastinae + Rutelinae) forming the more derived part of that clade. However, our results differ from some previous studies in several respects. Most notably, the monophyly of Melolonthinae recovered here is inconsistent with the result of Gunter et al. (2016). Likewise, the sister-group relationship between the coprophagous and phytophagous clades in our topology was not supported by Ahrens et al. (2014), Bocak et al. (2014), or Grebennikov and Scholtz (2004). Such differences may be attributable, at least in part, to the use of different datasets and analytical frameworks, but they may also more directly reflect the limited taxon sampling in the present study.

The epipharynx is one of the most variable structures among scarabaeoid larvae, with pronounced differences in spines, sensilla, setae, heli, phobae, and sclerotized processes across lineages and feeding guilds (Hayes 1928, 1929; Browne and Scholtz 2002; Grebennikov and Scholtz 2004; Šípek and Ahrens 2011; Qu et al. 2019). Within Scarabaeidae, two key structures—helus and circularly arranged phobae—are mutually exclusive and have never been documented together in a single species (Ritcher 1966; Zhang 1984). Circularly arranged phobae or subcircular phobal fields have been mostly clearly documented in coprophagous Aphodiinae and Scarabaeinae (Verdú and Galante 2000; Martínez and Lumaret 2005; Frolov et al. 2021), whereas helus has mainly been reported from phytophagous Sericinae, Melolonthinae, and several Rutelinae (Anomalini, Adoretini) (Morón and Paucar-Cabrera 2003; Šípek and Ahrens 2011; Cuate-Mozo et al. 2014; Lugo-García et al. 2018; Begha and Santos 2020; Zhang et al. 2024). Notably, although larvae of Glaphyridae, a basal lineage of Scarabaeoidea, are also soil-dwelling and phytophagous, available descriptions suggest that they lack helus (Ritcher 1966), implying that similar feeding habits do not necessarily involve the same epipharyngeal modification across Scarabaeoidea. Our ancestral-state reconstruction supports circular phobae as a putative synapomorphy of the sampled coprophagous Aphodiinae + Scarabaeinae, whereas helus represents a homoplastic acquisition among phytophagous white grubs (Jia et al. 2023; Cao et al. 2024; Fuhrmann et al. 2024), suggesting that helus may have evolved convergently within Scarabaeidae as one recurrent morphological response to subterranean phytophagy rather than as a universal scarabaeoid adaptation.

Mandibular morphology also varies widely across families and feeding modes (Lawrence and Ślipiński 2013; Beutel and Yavorskaya 2019). Mandibles are symmetric only in Passalidae (Grebennikov and Scholtz 2004; Mattos et al. 2015), but become apically curved, dentate, and sharply pointed in coprophagous Aphodiinae and Scarabaeinae (Edmonds and Halffter 1978; Li et al. 2019) or saprophagous Lucanidae, Cetoniinae, and Dynastinae (Sousa et al. 2018; Jang and Kim 2019; Qu et al. 2019; Dong and Jiang 2023). Blade-like incisors characterize phytophagous Sericinae, Melolonthinae, and Rutelinae (Ernesto and Carrillo 2010; Filippini et al. 2017; Neita-Moreno and Morón 2017; Jia et al. 2020; Šípek and Ahrens 2011; Cao et al. 2024). Our analyses corroborate that these blade-like incisors have evolved multiple times in plant-feeding lineages and likely represent functional adaptations to cutting subterranean roots (Ritcher 1966; Jia et al. 2021, 2023; Zhang et al. 2024). Even within Rutelinae, the trait is not phylogenetically stable, occurring in phytophagous species but not in saproxylic taxa (Fang et al. 2018; Sun et al. 2024). This pattern is consistent with the broader ecological diversity of Rutelinae larvae, especially within Rutelini, where larvae are frequently saproxylic and associated with decaying wood or other decomposing organic substrates rather than living roots (Jameson and Morón 2001; Sun et al. 2024), and further supports the view that blade-like incisors track feeding substrate rather than subfamily identity.

Mandibles also play a central role in larval stridulation (Ritcher 1966; Dong and Jiang 2023) producing species-specific signals (Görres and Chesmore 2019) using ventral ridges whose number and arrangement vary widely (Balaguer et al. 2008). Our reconstruction indicates that dense stridulatory ridges may represent a putative synapomorphy of the sampled Cetoniinae, Rutelinae, and Dynastinae (Ritcher 1948; Bedford 1974; Ibarra-Polesel et al. 2020, 2022; Uliana et al. 2023). Other lineages lack these ridges: they are absent in Lucanidae, Passalidae, and Geotrupidae (Grebennikov and Scholtz 2004), replaced by a single sclerotized ridge in some Sericinae (Šípek and Ahrens 2011; Cao et al. 2024), a glabrous surface in some Melolonthinae (Jia et al. 2020, 2023; Zhang et al. 2024), or granular processes in others (Jia et al. 2021; Rana et al. 2022). The striking heterogeneity within Melolonthinae likely reflects the known paraphyly of this subfamily (Ahrens et al. 2014; Guo et al. 2022; Dietz et al. 2023).

Maxillary stridulatory teeth, described across many white grubs (Orozco and Pardo-Locarno 2004; Roggero et al. 2013; Shabalin 2017), typically interact with mandibular ridges in Cetoniinae, Rutelinae, and Dynastinae (Ratcliffe and Skelley 2011; Dong et al. 2021; Sun et al. 2024). Unexpectedly, however, maxillary teeth also occur in Aphodiinae, Scarabaeinae, Sericinae, Melolonthinae, and Euchirinae (Edmonds and Halffter 1978; Šípek et al. 2011; Li et al. 2019; Jia et al. 2023), even though these lineages lack mandibular ridges. The widespread occurrence of maxillary teeth implies that stridulatory structures originated before the evolution of specialized mandibular files, but the acoustical mechanism remains poorly understood. Our analyses support maxillary stridulatory teeth as a potential synapomorphy of Scarabaeidae within the sample taxa (Fig. 11; Clade D).

The limited sampling in the present study did not include Geotrupidae, Bolboceratidae, Trogidae, or Glaphyridae, mainly because specimens for which both larval morphological information and mitogenomic sequences are available remain extremely scarce. Nevertheless, comparison with previous reports on larval morphology and biology (Verdú et al. 2004; Bezborodov and Koshkin 2015; Costa-Silva and Smith 2024), suggests that some of the patterns inferred in our sampled framework may extend beyond the taxa included here. In particular, the circular phoba complex together with the absence of nesium, which in our sampled taxa appears to characterize coprophagous larvae, is also reported in the coprophagous Geotrupidae and Bolboceratidae (Verdú et al. 2004). The circular phoba complex is likewise present in the keratin-feeding larvae of Trogidae (Costa-Silva and Smith 2024), although the absence of nesium does not appear to apply to this group (Ritcher 1966). These comparisons suggest that the circular phoba complex may be more broadly associated with non-phytophagous feeding habits, whereas loss of the nesium may be more restricted. At the same time, such inferences should be treated cautiously, because the present study is limited by incomplete larval sampling across Scarabaeoidea, and broader taxon coverage will be needed to test the generality of these patterns.

Larval mouthparts are multifunctional structures involved not only in feeding but also locomotion and communication (Stehr 1987; Chapman 2013), and thus are shaped by strong selective pressures. Within a phylogenetic framework, we show that three characters emerge as potentially informative synapomorphies within the sampled framework, enhancing their value for morphological systematics (Hayes 1929; Van Emden 1957; Ritcher 1966; Browne and Scholtz 2002; Grebennikov and Scholtz 2004). Fourteen additional characters show homoplastic distributions, yet still retain an apparent ecological signal, particularly with respect to feeding mode (Krenn 2019). The diversity documented here likely represents only a fraction of the morphological variation present across Scarabaeoidea. Broader sampling and detailed morphological-functional analyses will be essential for elucidating larval evolution and for improving pest management or conservation applications.

Conflict of interest. The authors declare that they have no conflicts of interest in relation to this work.

Author contributions. Wenjun Cao (Conceptualization [Equal], Data curation [Lead], Investigation [Lead], Methodology [Lead], Visualization [Lead], Writing–original draft [Lead]), Lu Jiang (Corresponding Author, Conceptualization [Equal], Supervision [Lead], Writing–review and editing [Lead])

Data availability. All new five mtGenomes generated in this study were deposited in GenBank under accession numbers PQ067330-PQ067331 and PQ083081-PQ083083.