Cyclocephalini is the second most species-rich tribe of Dynastinae after Pentodontini, and it contains 14 genera and over 500 species and subspecies (Smith 2006, Moore et al. 2015, 2018b, Ratcliffe and Cave 2017). Historically, the tribe Cyclocephalini was characterized by the absence of characters found in other dynastines. These characters included: 1) lack of horns, tubercles, carinae, or foveae on the head and prothorax; 2) absence of a stridulatory area (pars stridens) on the propygidium; 3) simple mandibles that lack dentition distal to the molar region; 4) metatibial apex truncate and without produced teeth or a crenulated margin; and 5) metatarsus with basal joint simple and not triangular (Ratcliffe and Cave 2017). The sexual dimorphism found in cyclocephalines is not as pronounced as it is in the horned dynastines. However, most cyclocephaline species display sexual dimorphism of the protarsus (enlarged in males; simple in females) and elytral epipleural margin (simple in males; expanded and modified in females of some species). Moore (2012) hypothesized that during mating, there was an interaction between the enlarged male protarsal claw and the female epipleural expansion, making it easier for the male to clasp the female during copulation and for mate guarding. Moreover, as in all dynastines, the apex of the last abdominal sternite is emarginate in males and entire or rounded in females (Figs 1, 2).

Figures 1–2. 

Last sternite in Dynastinae. 1 male with apex emarginate 2 female with apex entire.

Cyclocephalini, while relatively morphologically uniform, is not well defined, and monophyly of the tribe still needs to be evaluated (Ratcliffe and Cave 2017). In that endeavor, some generic-level taxa have been removed from the tribe while others have been transferred into Cyclocephalini. Coscinocephalus Prell was transferred from Cyclocephalini to Pentodontini and is considered to be most similar to Orizabus Fairmaire (Morón and Ratcliffe 1996). The bizarre, monotypic genus Acrobolbia Ohaus was transferred from Rutelinae to Cyclocephalini by Jameson et al. (2002), and those authors compared the genus to Ancognatha Erichson. Peltonotus Burmeister, with 25 species, was transferred from Rutelinae to Dynastinae (Jameson 1998) and specifically to Cyclocephalini (Jameson and Jakl 2010). Additionally, the monophyly of several cyclocephaline genera is in doubt. Ratcliffe (2003) stressed that further research is needed on the genera Cyclocephala Dejean, Mimeoma Casey, Aspidolea Bates, and Ancognatha to ascertain if they should be maintained as valid genera or some should be folded into others. Moore et al. (2015) evaluated the monophyly of Mimeoma and its relationship with Cyclocephala by using a combined molecular and morphological analysis. These data showed that the five species of Mimeoma clustered within an apical clade of other Cyclocephala species, rendering Cyclocephala paraphyletic. As a result, Mimeoma was synonymized with Cyclocephala.

Pentodontini is the largest tribe of dynastines, comprising about 100 genera and over 550 species distributed worldwide (Ratcliffe and Morón 1997, Ratcliffe and Cave 2017). Adult pentodontines are distinguished by: 1) the presence of tubercles, a carina, or a fovea on the head and pronotum; 2) broad mandibles with or without teeth on the scissorial region; 3) propygidium with or without a pars stridens; 4) lateral margin of the protibia usually tridentate; 5) apex of the metatibia usually truncate and margined with short, spine-like setae; 6) protarsus occasionally enlarged in males (Ratcliffe and Morón 1997, Ratcliffe and Cave 2017).

Dimorphism between males and females is slight in most species (Ratcliffe and Morón 1997), although males sometimes have larger protarsi and tubercles on the head and pronotum (López-García et al. 2015), and the pronotal fovea is more pronounced. Pentodontines, along with all dynastines, display sexual dimorphism of the last abdominal sternite, which is emarginate in males and entire or rounded in females (Figs 1, 2).

López-García et al. (2015) reported that for over 100 years, there was no consensus whether Pentodontini should be treated as a family, subfamily, or tribe. Historical workers prioritized different criteria: Mulsant (1842) considered the categories Pentodonaires and Oryctésaires as separate groups; followed by Bates (1888) who designated Pentodontinae as a subfamily (Ratcliffe and Morón 1997, López-García et al. 2015); Casey (1915) established Pentodontini as a tribe and later Leng (1920), Arrow (1937), Blackwelder (1944), and Arnett (1973) did not recognize any of these former designations and included pentodontine genera within Oryctini (Ratcliffe and Morón 1997). Endrődi (1969) re-established the tribes Pentodontini and Oryctini as they are currently used, but he considered that transitional characters blurred the distinction between the tribes. Consequently, the monophyly of Pentodontini is in doubt (Ratcliffe 2003, Gasca-Álvarez et al. 2008, López-García et al. 2015, Sanabria-García et al. 2012).

Morphological descriptions and temporal and distributional data were based on the study of 425 specimens from three sources: (1) field collecting expeditions by the authors and colleagues; (2) data recorded from the literature; and (3) specimens from the following museums and private collections: Canadian Museum of Nature (Ottawa, Canada), Canadian National Collection of Insects (Ottawa, Canada), Florida State Collection of Arthropods (Gainesville, Florida, United States), Museo Ecuatoriano de Ciencias Naturales (Quito, Ecuador), Museo de la Escuela Politécnica Nacional (Quito, Ecuador), National Museum of Natural History (Prague, Czech Republic), Museo QCAZ-Invertebrados de la Pontificia Universidad Católica del Ecuador (Quito, Ecuador), Stephane Le Tirant Collection (Terrabonne, Québec, Canada), University of Nebraska State Museum (Lincoln, Nebraska, United States), and United States National Museum (Washington, DC, currently on long-term loan to University of Nebraska State Museum, Lincoln, Nebraska, United States).

This study was developed as part of the broader project “The Dynastine Scarab Beetles of Ecuador”. For this reason, we provide only generalized, province-level distribution data for Pucaya and Parapucaya species in Ecuador. More detailed distribution data for these genera will be released as part of that forthcoming monograph. Collecting methods utilized were: 1) light traps using mercury vapor and ultraviolet bulbs; 2) foliage gleaning; 3) excavating rotting logs and stumps; and 4) manual collecting around public lights. Ecuadorian collecting, mobilization, and export permits were obtained with the support of QCAZ in Quito, Ecuador.

The species descriptions encompass the range of variation observed in the specimens at hand. They were based on the following characteristics (from Ratcliffe and Cave 2017): 1) length from apex of the clypeus to the apex of the elytra; 2) width across elytral humeri; 3) coloration and markings; 4) interocular width (number of transverse eye diameters across the frons between the eyes); 5) form and sculpturing of the head, pronotum, elytra, and pygidium; 6) form of the prosternal process; and 7) form of the parameres. Punctures were considered simple unless otherwise noted. Minute punctures were generally not visible with 12.5× magnification but were easily seen with 50× magnification. Small punctures were clearly visible with 12.5× magnification and can be seen with the naked eye. Large punctures are easily seen without magnification. Sparse punctures were characterized by greater than 5 puncture diameters between them. Punctures moderate in density had 3–5 puncture diameters between them. Dense punctures had only 2 or fewer puncture diameters between them.

Previous studies by Gunter et al. (2016) and Ahrens et al. (2011, 2014) generated DNA sequence data that served as a phylogenetic scaffold for testing the classification of Pucaya and Parapucaya within Pentodontini (Tab. 1). GenBank was datamined for 28S, CO1, and 16S/ND1 sequences from diverse tribal-level exemplars for higher Scarabaeidae (Tab. 1). Among Dynastinae, there were tribal-level exemplars with at least partial data for all three gene regions from six of the eight commonly recognized tribes (minus Hexodontini and Agaocephalini). 16S and 28S data were generated from exemplar specimens of Pucaya pulchra Arrow and Parapucaya amazonica Prell to incorporate into this phylogenetic framework. Based on shared morphological characters with Pucaya species, Cyclocephala freyi Endrődi exemplars were also targeted for DNA extraction and PCR.

DNA extractions of metafemoral tissue from specimens of C. freyi, P. amazonica, and P. pulchra were performed using guanidinium thiocyanate following the QCAZ Molecular Biology Laboratory protocol (unpublished). 28S sequence data was gathered using the primers Bulbasaur/28SR (Moore et al. 2015, Whiting et al. 1997, Whiting 2001) and thermocycles from Moore et al. (2015). 16S sequence data was generated using the universal primer LRJ-12864 and 16Sar-L (Palumbi et al. 1991) with the following thermocycle: 1) 94 °C for 2 minutes; 2) 94 °C for 40 seconds; 3) 54 °C for 40 seconds; 4) 68 °C for 1 minute and 30 seconds (34 cycles of steps 2–4); and a final extension of 68 °C for 1 minute. Forward and reverse sequence traces were trimmed and assembled into contigs in Geneious 5.6.2 (Kearse et al. 2012).

Based on the results of Gunter et al. (2016), a species of Isonychus Mannerheim (Scarabaeidae: Melolonthinae: Macrodactylini) was used as an outgroup for all phylogenetic analyses. Isonychus sp. was recovered as the most early-diverging member of a melolonthine clade sister to the clade containing all Cetoniinae + Rutelinae + Dynastinae exemplars (Gunter et al. 2016), making this taxon a suitable outgroup for examining relationships among these subfamilies and placing Pucaya and Parapucaya at the tribal level. Sequences were aligned using ClustalW (Larkin et al. 2007), with default settings, as implemented in MEGA7 (Kumar et al. 2016). The resulting concatenated sequence alignment contained 3,537 bp positions (1530 bp 16S; 550 bp ND1; 805 bp CO1; 652 bp 28S).

Maximum likelihood analyses of this matrix were conducted in W-IQ-TREE (Trifinopoulos et al. 2016). The matrix was partitioned by gene (16S, 28S, and ND1) and codon position (CO1). The best-fit model of sequence evolution for each partition (GTR+F+I+G4 for 16S; TPM3u+F+G4 for ND1 and CO1 third position; TIM2e+I+G4 for 28S; HKY+F+I+G4 for CO1 first position; SYM+I+G4 for CO1 second position) was selected by ModelFinder (Kalyaanamoorthy et al. 2017), as implemented in W-IQ-TREE, using the Bayesian information criterion. Bootstrap support values for the most likely tree were calculated using 10,000 ultrafast bootstrap replicates (Hoang et al. 2017). Bayesian phylogenetic analyses were run in MrBayes 3.2.2 (Ronquist and Huelsenbeck 2003). Models of sequence evolution for the MrBayes analysis were determined with PartitionFinder 2.1.1 (GTR+I+G for 16S, ND1, 28S, CO1 second position, and CO1 third position; HKY+I+G for CO1 first position) (Lanfear et al. 2016).

Bayesian analyses comprised four independent runs, each with four chains (one cold and three heated). Partitions had their parameters unlinked and allowed to vary independently. Flat priors were used. Chains were run for 1 million generations, with trees sampled every 1,000 generations. Convergence was evaluated by examining the standard deviation of split frequencies among runs and by plotting the log-likelihood values from each run using Tracer 1.6 (Rambaut and Drummond 2013). Tracer diagnostics indicated that runs converged within 10,000 generations, and trees sampled during this period were discarded as burn-in before obtaining clade posterior probabilities. Parsimony tree searches were performed in MPBoot (Hoang et al. 2018). Heuristic searches were conducted using default parsimony ratchet search options in MPBoot. Bootstrap analyses were performed using the same ratchet search options and included 10,000 bootstrap replicates.

Morphological observations show that Parapucaya shares characters with genera in Cyclocephalini, most notably with some Cyclocephala species. For example, the two Parapucaya species share characters with C. almitana Dechambre, C. macrophylla Erichson, C. melanocephala (Fabricius), and C. pseudomelanocephala Dupuis. These characters include: 1) frontoclypeal suture complete; 2) clypeus weakly emarginate with lateral and apical margins reflexed; 3) clypeal apex broadly truncate; 4) the generally exposed and slender mandibles that lack lateral teeth; 5) mandibular apex acuminate and curved upward; 6) protibia strongly tridentate with the basal tooth removed from other two teeth; 7) protarsus in males enlarged (the larger claw strongly curved and incised at apex), while females have a simple protarsus; 8) inner portion of the apical margin of the 5^th protarsomeres in males eroded, allowing the enlarged protarsal claw to be further articulated; 9) metatarsi reduced, shorter than metatibia, more evident in females (character shared with C. melanocephala and C. almitana); 10) prosternal process moderately long, columnar, with its apex densely setose, flattened, and with a large, raised, round “button” covering half of the apex; 11) hindwing vein RA, proximal to apical hinge, with 2 rows of pegs extending distally nearly to margin of apical hinge; and 12) anterior edge of hindwing distal to apical hinge lacking setae and with a produced, membranous border (Figs 3, 4).

Figures 3–4. 

Hindwing vein RA3, distal to apical hinge. 3 Parapucaya sp. 4 Pucaya sp. Photo credits to Gavin J. Martin.

Like Parapucaya, Pucaya species share many characters with some Cyclocephala species (e.g., C. freyi). Pucaya also shares the character of a medially incomplete frontoclypeal suture with Ancognatha species. In Pucaya individuals, the frontoclypeal suture is visible from the lateral margins along the external side of the frontal horn, where it becomes obsolete medially. Pucaya and some Ancognatha species display weakly developed “armature” of the head and thorax. For example, Ancognatha castanea Erichson has tubercle-like swellings on the frontoclypeal region of the head. Ancognatha jamesoni Murray and A. horrida Endrődi show enlargement of the pronotum in males.

Other shared characters with other cyclocephalines include; 1) clypeus with lateral and apical margins reflexed; 2) clypeal apex broadly truncate, subquadrate; 4) maxillary galea with four teeth on inner margin (shared with C. freyi (Figs 5, 60), 5) slender mandibles that lack lateral teeth; 6) protibia strongly tridentate with the basal tooth removed from other two teeth; 7) protarsus in males enlarged (the larger claw strongly curved and incised at apex), while females have a simple protarsus; 8) inner portion of the apical margin of the 5^th protarsomeres in males eroded, allowing the enlarged protarsal claw to be further articulated; 9) prosternal process short to moderately long, columnar, with its apex densely setose, flattened, and with a large, raised, round “button” covering half of the apex; 11) hindwing vein RA, proximal to apical hinge, with 2 rows of pegs extending distally nearly to margin of apical hinge; and 12) anterior edge of hindwing distal to apical hinge lacking setae and with a produced, membranous border.

Figures 5–6. 

Maxillary galea showing four teeth. 5 Cyclocephala freyi Endrődi 6 Pucaya pulchra Arrow. Photograph credits to Gavin J. Martin.

W-IQ-TREE analyses found the most likely tree with a log likelihood score of -42928.5840. MPBoot heuristic tree searches recovered most parsimonious trees of score 9992 steps. Bayesian posterior probabilities and parsimony bootstrap support values for nodes are reported on the maximum likelihood bootstrap consensus tree topology (Fig. 7). Analyses conducted on the concatenated dataset recovered 27 strongly supported internal nodes (>75 BS and >0.95 PP) from all three tree search strategies. All three analyses strongly supported the monophyly of Cetoniinae and Dynastinae (Fig. 7). Like the analyses of Gunter et al. (2016), the subfamily Rutelinae was recovered as paraphyletic. Parapucaya amazonica, P. pulchra, and C. freyi were recovered together as a clade (94 ML BS, 0.97 PP, 73 Parsimony BS) sister to the other three Cyclocephala exemplars. Together, these six exemplars form a strongly supported cyclocephaline clade (99 ML BS, 1.0 PP, 79 Parsimony BS) within the broader Dynastinae clade (99 ML BS, 1.0 PP, 91 Parsimony BS) (Fig. 7). The remaining 14 Pentodontini species included here did not form a monophyletic group. Six pentodontine species fell out in a clade that includes Cryptodus sp. (Dynastinae: Phileurini) (96 ML BS, 1.0 PP). Eight pentodontine species were recovered in a clade (98 ML BS, 1.0 PP, 86 Parsimony BS) that also included Oryctes nasicornis (Linnaeus) (Dynastinae: Oryctini).

Figure 7. 

Bootstrap consensus tree from W-IQ-TREE analysis. Node support values from left to right are ML bootstrap, Bayesian posterior probability, and parsimony bootstrap. Support values labeled with a “*” have 100% bootstrap support or 1.0 posterior probability. Support values labeled with a “-” have bootstrap supports lower than 50% or posterior probability lower than 0.95. Nodes labeled “--” indicates that node was not recovered by an analysis.

(Modified from Jameson et al. 2002)

Males: Apex of last abdominal sternite emarginate (Fig. 1). Protarsomeres 4–5 and/or anterior claw enlarged in all genera except Stenocrates and Erioscelis.

Females: Apex of last abdominal sternite entire, evenly parabolic (Fig. 2). Protarsomeres 4–5 and anterior claw always simple, not enlarged.

Figures 8–15. 

Form of clypeal apex. 8 Peltonotus silvanus Jameson and Wada (subquadrate, note chitinized labrum) 9 Ruteloryctes morio (Fabricius) (emarginate, note round mandibles) 10 Surutu seabrai d’Andretta and Martínez (narrowly parabolic) 11 S. hesperius Ratcliffe (broadly parabolic) 12 Protibia of Harposceles sp. 13 Acrobolbia sp. (pentagonal) 14 Arriguttia sp. 15 Augoderia sp. (thickened apex) (Figures 9–15 modified from Jameson et al. 2002, by permission).

Figures 16–24. 

Form of clypeal apex. 16 Aspidolea sp. (subquadrate, dirvergent from base to apex) 17 Erioscelis sp. (subquadrate truncate) 18 Erioscelis sp. (subquadrate emarginate) 19 Ancognatha sp. (narrowly parabolic) 20 Ancognatha sp. (parabolic) 21 Cyclocephala sp. (rounded) 22 Cyclocephala sp. (truncate) 23 Cyclocephala sp. (emarginate) 24 Cyclocephala sp. (sharply acuminate) (Figures modified from Jameson et al. 2002 and Ratcliffe 2003, by permission).

Figures 25–30. 

Form of clypeus and pronotum. 25 Parapucaya sp. (clypeus thickened with sharp crest on margins) 26 Cyclocephala melanocephala (Fabricius) (clypeus not thickened, margins weakly raised, rounded). 27 Stenocrates sp. (clypeus trapezoidal, pronotum not produced anteriorly at middle) 28 Dyscinetus sp. (clypeus subtrapezoidal, pronotum with anterior margin produced at middle). Form of clypeus in posterodorsal view Mentum 29 Ancognatha sp. (with furrow) 30 Cyclocephala sp. (without furrow, weakly emarginate). (figures 28–30 from Ratcliffe 2003, by permission).

Figure 31. 

Parapucaya amazonica Prell.

Figure 32. 

Parapucaya amazonica Prell parameres (from Ratcliffe 2003, by permission).

Figure 33. 

Parapucaya nodicollis (Kirsch).

Figure 34. 

Parapucaya nodicollis (Kirsch) parameres.

Figures 35–36. 

Pronota. 35 Pucaya pulchra Arrow 36 Pucaya castanea Ohaus.

Figure 37. 

Pucaya castanea Ohaus.

Figure 38. 

Pucaya castanea Ohaus parameres (from Ratcliffe 2003, by permission).

Figure 39. 

Pucaya pulchra Arrow. Image by B.C. Ratcliffe.

Figure 40. 

Pucaya pulchra Arrow parameres.

Males with protarsomeres enlarged, protarsus with one claw simple and one enlarged. Females with protarsomeres slender, protarsus with both claws simple.