ZOOTAXA ISSN 1175-5326 (print edition) ISSN 1175-5334 (online edition)Copyright © 2017 Magnolia Press Zootaxa 4347 (2): 301–315 http://www.mapress.com/j/zt/ Article https://doi.org/10.11646/zootaxa.4347.2.6 http://zoobank.org/urn:lsid:zoobank.org:pub:24DE9911-A704-47C9-BAC8-21AD27E233EE Troublesome Trimes: Potential cryptic speciation of the Trimeresurus (Popeia) popeiorum complex (Serpentes: Crotalidae) around the Isthmus of Kra (Myanmar and Thailand) DANIEL G. MULCAHY1,5, JUSTIN L. LEE2,3, ARYEH H. MILLER2,4 & GEORGE R. ZUG2 1Global Genome Initiative, National Museum of Natural History, Smithsonian Institution, Washington, DC, 20013 USA 2Department of Vertebrate Zoology, National Museum of Natural History, Smithsonian Institution, Washington, DC, 20013 USA 3College of Computer, Mathematical and Natural Sciences, University of Maryland, College Park, MD, 20742 USA 4Department of Biology, University of North Carolina Asheville, Asheville, NC 28804. 5Corresponding author. E-mail: MulcahyD@si.edu Abstract The taxonomic identity of the Trimeresurus (Popeia) popeiorum complex from the Isthmus of Kra and to the north was investigated. Several studies over the last decade have produced several specimens and associated mtDNA sequence data for a variety of individuals of the T. popeiorum and “T. sabahi” complexes. Here, we combine four mitochondrial genes (12S, 16S, ND4, and CytB) from all available specimens in GenBank with the addition of five new specimens collected from the mainland, Tanintharyi Region of Myanmar. Maximum Likelihood and Bayesian analyses identified that T. po- peiorum sensu lato is paraphyletic with two geographically distinct clades: a northern clade representing populations from northern Myanmar, Laos and northern Thailand and a southern clade representing samples from the Tanintharyi Region and adjacent west Thailand. While the two clades have considerable genetic distance, they appear to be morphologically identical, leading to the hypothesis that the southern clade represents a cryptic, undescribed species. Because they appear to be cryptic species and the limitation of only five specimens from the southern lineage, this does not permit us to for- mally describe the new species. In accordance to past molecular studies, we uncovered paraphyly and lack of genetic sup- port for the validity of taxa within the T. sabahi complex. However, we suggest recognizing these populations as subspecies within T. sabahi. Key words: Cryptic speciation, Myanmar, Southeast Asia, Subspecies, Tanintharyi Region, Thailand, Trimeresurus Introduction Often times new species are described based on few individual specimens available, sometimes only from the type series, which can be problematic for interspecific comparisons. This can be particularly problematic in species with sexual dimorphism and/or from cryptic species complexes. Additionally, sequences in GenBank are often not represented by voucher specimens and sequences can be misidentified, which confounds resolution of relationships. Cryptic species, morphologically indistinguishable but genetically and/or reproductively isolated (Bickford et al. 2006; Jörger et al. 2013), can confound taxonomic matters even further (Funk et al. 2012). Molecular sequencing methods can be extremely useful for determining cryptic species (Hebert et al. 2004), particularly in groups with limited samples for morphological comparisons. Southeast Asian Green Pitvipers (Genus: Trimeresurus) are notoriously difficult to classify. Despite the abundance of specimens in museum collections for some species, morphological conservatism in the genus makes taxonomic studies challenging and some species are only represented by few museum specimens. Such limited samples have resulted in misidentifications within the genus (e.g. Orlov & Helfenberger, 1997) that were subsequently corrected by others (Malhotra & Thorpe 2000; Giannasi et al. 2001; Tillack et al. 2003). Nevertheless, species diversity in Trimeresurus is likely underestimated, as detailed examination of several groups has revealed undescribed or revalidated species (Vogel et al. 2004; David et al. 2006; Grismer et al. 2006; David et Accepted by T. Nguyen: 9 Oct. 2017; published: 13 Nov. 2017 Licensed under a Creative Commons Attribution License http://creativecommons.org/licenses/by/3.0 301 al. 2009; Guo & Wang, 2011; Malhotra et al. 2011; Sumontha et al. 2011; Vogel et al. 2014a; Vogel et al. 2014b). Molecular phylogenetic analyses also have identified cryptic diversity in some species (Guo et al. 2015; Thorpe et al. 2015). Recent (2015–2016) rapid assessment surveys in southern Tanintharyi Region, Myanmar by one of us (DGM) yielded three specimens, which we assigned to the subgenus Popeia (Malhorta & Thorpe, 2004), specifically to Trimeresurus popeiorum Smith, 1937. Recently, several new species in this and the T. sabahi complexes have been described (e.g. Vogel et al. 2004; Grismer et al. 2006; David et al. 2009). Additional studies using mitochondrial DNA (mtDNA) sequence data have attempted to unravel these complexes (Malhorta and Thorpe 2004; Sanders et al. 2006). A recent study claims to resolve most taxonomic issues for this group (Wostl et al. 2016); however, a comprehensive analysis has yet to be conducted, and questions remain as to what species occur in the Tanintharyi. Briefly, the Popeia subgeneric group occurs from northern India to southern China (Guo et al. 2015) southward through Myanmar, Thailand, Indochina, and Malay Peninsula into Borneo and Sumatra. The type species, Trimeresurus popeiorum, was described in 1937 by Malcolm A. Smith without any precise type locality or type specimen. This was subsequently corrected by Taylor & Elbel (1958). Later, Regenass & Kramer (1981) described two new subspecies (T. p. barati and T. p. sabahi) in the complex. In 2004, Malhotra & Thorpe performed a revision of Southeast Asian Trimeresurus. Their study included both morphological (hemipenes, scalation) and genetic (mtDNA) characters. They suggested the subdivision of the Trimeresurus species into several “new” genera. One of these genera was Popeia for T. popeiorum. In the same year, Vogel et al. (2004) revised T. popeiorum and, on the basis of morphological characters, recognized two new species in Peninsular Malaysia: Trimeresurus fucatus Vogel, David & Pauwels, 2004 in southern Thailand and Peninsular Malaysia and Trimeresurus nebularis Vogel, David & Pauwels, 2004 restricted to the Cameron Highlands of Peninsular Malaysia. They also elevated Trimeresurus barati (Regenass and Kramer, 1981) for the Sumatran populations and T. sabahi (Regenass and Kramer, 1981) for the “popeiorum” of Borneo. This new taxonomy resulted in the distribution T. popeiorum being the northern portion of the previous Southeast Asian-wide range but also identified the populations as far south as Myeik, (Tanintharyi, Myanmar), as T. popeiorum. Their distributional concept, however, disagreed with their identification of several specimens, i.e., they considered a female BMNH 1924.5.20.38 from “Taok Plateau, Tenasserim” (now Mt. Pya Taung, Tanintharyi, Myanmar) as T. popeiorum, and two males: BMNH 56.5.6.105 from Myeik, Tanintharyi, Myanmar and BMNH 1940.3.9.43 from Kanmaw Kyun Island (= Kisseraing Island), Tanintharyi, Myanmar as T. fucatus. Subsequently, Pauwels and Chan-ard (2006) identified Trimeresurus (Popeia) from Keang Krachan National Park, Thailand as T. fucatus. Sanders et al. (2006) using an expanded molecular and morphological data from Malhotra and Thorpe (2004), defined two clades within the Popeia complex. The Northern Clade contained all specimens north of the Isthmus of Kra and one specimen (B467) near Phang-nga (south Thailand). The Southern Clade contained all specimens further south in the Malay Peninsula and the islands of Indonesia. Sanders et al. (2006) recommended a more conservative taxonomy, placing all Popeia species from the Sundaland region into T. sabahi, except for T. nebularis and retaining T. popeiorum for specimens north of the Malay Peninsula. Sanders et al. (2006) also examined the morphology of a specimen from Tanintharyi, identified by Vogel et al. (2004) as T. fucatus (BMNH 1940.3.9.43), and it was identified as part of the “Northern Clade”. However, they did not examine any other individuals from Tanintharyi and provided no meristic data from the BMNH specimen. They included uncatalogued specimens from Phetcheburi Province (AMB52; B34) in their statistical sample and identified them as T. popeiorum; these specimens were paraphyletic with respect to other T. popeiorum specimens in their molecular phylogenetic analysis. Another uncatalogued specimen from Phang-nga Province, Thailand (B467) was also included in the northern clade, and this inclusion also rendered T. popeiorum paraphyletic. The taxonomy suggested by Sanders et al. (2006) remains controversial. Subsequent studies focusing on the Popeia subgenus continue to follow the taxonomy suggested by Vogel et al. (2004). Grismer et al. (2006) described T. buniana from Pulau Tioman Island off the coast of Peninsular Malaysia. David et al. (2009) described northern Sumatran specimens as T. toba and recommended that the genus Popeia (Malhotra & Thorpe, 2004) be used as a subgenus in order to preserve the definitive nature of the genus Trimeresursus. David et al. (2011) expanded on this recommendation and identified the nucleospecies (= type species) of the genus Trimeresurus as T. viridis Lacépède, 1804 (= T. insularis Kramer, 1977) and officially recognized all genera proposed by Malhotra & Thorpe (2004) as subgenera with the exception of Ovophis and Protobothrops. Based on morphometrics, Sumontha et al. (2011) described T. phuketensis, a species endemic to Phuket Island, Thailand. It is unique among Popeia in that both MULCAHY ET AL. 302 · Zootaxa 4347 (2) © 2017 Magnolia Press males and females contain a bicolored postocular and ventrolateral stripe. Most recently, Wostl et al. (2016) added molecular data for two previously un-sampled taxa (T. barati and T. toba), but used only two (of four) mtDNA genes used by Sanders et al. (2006). Wostl et al. (2016) also identified all Popeia from the Sundaland biogeographic region as T. sabahi, including T. buniana and T. toba. However, they did not evaluate the taxonomic status of T. phuketensis as there were no genetic sequences available; they also did not include the south (B467) and west (AMB52; B34) Thai samples of T. popeiorum—which rendered the species paraphyletic in Sanders et al. (2006). In spite of the large number of taxonomic studies focused on the Popeia subgenus, the identity of populations in the Tanintharyi region and the Isthmus of Kra remains uncertain. No comprehensive molecular dataset has been used to examine the affinities of the green Trimeresurus from this area. Here, we use recently collected specimens, augmented with additional California Academy of Sciences (CAS) specimens from the Tanintharyi Region, to determine which species occur in this region. We investigate the identity and relationships of these specimens with those in adjacent areas using molecular and morphological data by including all available specimens available in GenBank, all four mtDNA loci, and morphological data provided in previous studies. Material and methods All three specimens from the 2015–2016 survey were deposited in the National Museum of Natural History, Smithsonian Institution (USNM). The first specimen (USNM 587588) was collected in Lenya in May 2015. Two specimens, both adult females, (USNM 587918 and USNM 587919) were collected in Ywahilu in May 2016. One of these (USNM 587918) was found dead on the road and partially skeletonized, hence unavailable for morphological study. However, tissue samples were taken and the specimen is vouchered as a skeleton. Two specimens (a juvenile female USNM 587920, adult male USNM 587921) collected from Kawthaung, Tanintharyi, Myanmar, were included in the morphological analysis. Our molecular analyses also included specimens from the California Academy of Sciences: Dawei Township (a male CAS 245932) and Kawthaung (a female CAS 247754). Tissue samples were taken from the liver and heart and preserved in salt-saturated DMSO/EDTA buffer for genetic analyses. Extractions of genomic DNA were conducted on small pieces of liver or muscle tissue and run on an Auto-Genprep 965 (2011 AutoGen, Inc.), using standard phenol manufacturer protocols. Genomic DNA was eluted in 100 µl of AutoGen R9 re-suspension buffer. We sequenced four mitochondrial genes CytB, ND4, 16S and 12S. Primers used for each gene are identified in Table 1. Cycle-sequence reactions were performed in both directions, using the PCR primers using BigDye Terminator v3.1 Cycle Sequencing Kit’s in 0.25 × 10 µl reactions run on and ABI3730 Sequencer (2011 Life Technologies) using the 950 chemistry. Raw trace files were edited in Geneious 9.1.5 (Biomatters Ltd 2005–2016), complementary strands were aligned, edited, and inspected for translation. All sequences were deposited in GenBank under accession numbers MF476856–MF476874. Outgroups were chosen based on close phylogenetic relationship between taxa (Alencar et al. 2016). Additional genetic material along with our outgroups came from published records in GenBank (see Table 2). We performed maximum-likelihood (ML) analyses on the concatenated mtDNA using RAxML (v8.2.9, Stamatakis, 2014) with the rapid bootstrap inferences (1000 replicates) and subsequent GTRCAT thorough ML search, with each gene as a separate partition. We also conducted Bayesian analyses using MrBayes (v3.2.6; Ronquist et al. 2012). We partitioned our dataset by locus, applied the GTR+I+G model, and unlinked all partitions. We ran our analyses for 10 x 106 generations with four chains, sampling every 1000 generations. Stationarity was assessed by the average standard deviation of split frequencies (ASDSF < 0.01) and visual plots of log-likelihood by generation in Tracer v1.2 (Rambaut and Drummond, 2004); the first 1,000 trees (of 10,000) were discarded as the burn-in. A 50% majority-rule with compatible groups consensus was taken from the remaining trees and posterior probabilities (pp) of 0.95 or above were considered significant. We examined morphological characters considered diagnostic to the Popeia subgenus based on previous studies (Pope & Pope, 1933; Regenass & Kramer, 1981; Vogel et al. 2004). Although the most recent taxonomic treatment of the subgenus Popeia (Wostl et al. 2016) indicated that all Sundaic populations should be recognized as T. sabahi, we only compared the morphology of our specimens to the Thai-Malaysian populations of T. sabahi recognized as T. fucatus by Vogel et al. (2004) (see Table 2). This decision makes it easier for us to compare our specimens on a local basis, as the allopatric populations of T. sabahi defined by Wostl et al. (2016) as well as Zootaxa 4347 (2) © 2017 Magnolia Press · 303CRYPTIC SPECIATION IN TRIMERESURUS Sanders et al. (2006) each contain relatively stable morphologies, sexual dimorphism and ecology. Ventral scale count methodology follows Dowling (1951). Color pattern vocabulary follows Vogel et al. (2004). TABLE 1. List of primers used to amplify each mitochondrial gene in our study. Results We obtained alignments of the mitochondrial genes CytB (826 bp), ND4 (846 bp), 16S (539 bp) and 12S (410 bp) for a total of 2621 bp of aligned sequence data. Our ML analyses placed specimens from northern Myanmar sister to T. nebularis, with poor bootstrap value support (<50%, Fig. 1). An outlying specimen (B467), from Phang-nga Province, south Thailand, initially identified as T. popeiorum was placed at the base of a clade containing the northern Myanmar T. popeiorum + T. nebularis specimens with strong support (91%). The latter two were sister to one another, but by a very short branch length with poor support (<50%). The Tanintharyi Region and western Thailand specimens (AMB52 and B34 of Sanders et al. 2006) were placed sister to this clade (Fig. 1). However, similar to the taxonomy of Wostl et al. (2016) and Sanders et al. (2006), we recovered a single, well-supported clade containing all Sundaic populations of Popeia with strong support (100%). Our Bayesian results were very similar, with the main difference being the south Thai sample (B467) was placed sister to T. nebularis clade (albeit with poor posterior probability support and short branch; <.50 pp), and they were placed sister to the northern T. popeiorum samples, with strong support (pp = 0.99). The T. sabahi clade was resolved with strong support (pp = 0.99), relationships among the lineages in this clade were slightly different from the ML topology, but were also poorly supported (values shown in Fig. 1). The morphology is summarized in Table 3. All specimens are described as followed: TailL/TotalL ratio 20.9% in the male, 15.0–17.0% in females. The dorsal pattern in all specimens was solid green, except for the juvenile specimen (USNM 587920), which had faint irregular vertebral crossbars. It is unclear what color they were in life, but they are dark green in preservative. Postocular striping in females is faint but present in all specimens as a thin white line; in the male (USNM 587921), the postocular streak is bicolored with the thin section (bottom) plain white and the wide section (top) red. Ventrolateral striping in females is extremely faint, less than half a dorsal scale wide, visible as margins on the dorsal scales and is plain white. In the male (USNM 587921), the ventrolateral stripe is bicolored with the bottom being deep red and the top plain white, extending to the tail where it becomes sporadic. The eye color in life (available from photographs of female specimens USNM 587588 and USNM 587919) is red. The tail is mottled in rusty-red in all specimens with no clear distinction between the two colors, but females appear to have a green border laterally. Snout truncated; distinct but no sharp canthus rostralis; rostral visible from above; occipital scales distinctively keeled in the male (USNM 587921), slightly keeled in females; temporals only slightly keeled in all specimens. Loreal pit in contact with second labial; nostril always distinct from first labial; two preoculars in contact with loreal pit; single subocular always long and crescent shaped; one or two rows of scales between subocular and supralabials; first infralabial largest; two chin shields; mental never in contact with chin shields. Dorsal scales keeled and in 21 rows at midbody; Ventral scales in females 165–171, 169 in male specimen; subcaudals 57–65 in females, 72 in male specimen, and all have a single anal plate. Locus Primer Direction Temp. Sequence 5' to 3' Reference 12S 12SI Forward 48 TGCCAGCAGYCGCGGTTA Puillandre et al. 2009 12S 12SIII Reverse 48 AGAGYGRCGGGCGATGTGT Puillandre et al. 2009 16S 16Sar–L Forward 54 CGCCTGTTTATCAAAAACAT Palumbi et al. 1991 16S 16Sbr–H Reverse 54 CCGGTCTGAACTCAGATCACGT Palumbi et al. 1991 CytB Gludge Forward 48 TGACTTGAARAACCAYCGTTG Parkinson et al. 2002; CytB ATRCB3 Reverse 48 TGAGAAGTTTTCYGGGTCRTT Parkinson et al. 2002; ND4 HypLeu2r.1 Forward 48 TACCACTTGGATTTGCACCA Mulcahy 2008 MPE ND4 HypNad4f.1 Reverse 48 TGCCTAGCAGCCTTYATAGCTA Mulcahy 2008 MPEMULCAHY ET AL. 304 · Zootaxa 4347 (2) © 2017 Magnolia Press                                          ac ce ss io n                                         ! "  #    $ %" $ %    "            &     ' " ( &) * + $ &   )      *    +     %    ' " ( ) , -. /+ 0$ )    ,              1 %    2 #  0     ( ' * ) $ '      *     )     3     4 %  ' " (       #           #     -             )     5 6    7 88 9          7 88 :9 ;                                      Tr im er es ur us a lb ol ab ri s " ) " < =>  " &< >< ??  " @ A= ? AB " @ A= >  " @ A= ? 8=  +   C   D  )     5 6    ! 88 ;  Tr im er es ur us in su la ris " ) " < 8B  " @ A= > :> " @ A= ? AA " @ A= > A? " @ A= > BB  0    E  2 #  )     5 6    ! 88 ;  Tr im er es ur us ti be ta nu s ,)  := : < " @ A= > B " @ A= ? <8 " @ A= > <= " @ A= > >:  *    +    /  #  )     5 6    ! 88 ;  Tr im er es ur us st ej ne ge ri " ) " < :8  " &< >< ?B : " @ 8= B= BA " @ 8= B= == " @ 8= B= AB  6    6  %    )     5 6    ! 88 ;  Tr im er es ur us b ar at i ) , -. /+ 0= C /? BB :  C /B AB A  F F 0        2   /  #  3   et a l. ! 8< :;  Tr im er es ur us b ar at i ' 6" -G : <: 8  C /? BB : A C /B AB A < F F 0        2   /  #  3   et a l ! 8< :;  Tr im er es ur us b ar at i ) , -. /+ 0= " @ A> <> :B " @ A> <> =A  0            / #  )     5 6    ! 88 ;  Tr im er es ur us b un ia na " )  = <> >? " @ A> <> =  )     /   6       e t a l ! 88 :;  Tr im er es ur us fu ca tu s  :  " @ 8= B= >8 " @ 8= B= ?B " @ 8= B= =: " @ 8= B= 8  )     &  I +   )     5 6    ! 88 ;  Tr im er es ur us fu ca tu s  >?  " @ A> " @ A> <> ?8 " @ A> <> =8  )     &  I +   )     5 6    ! 88 ;  Tr im er es ur us fu ca tu s " ) "  8A  " @ A> <> B: " @ 8= B= ?? " @ 8= B= =A " @ 8= B= A>  6     *     6      /  #  )     5 6    ! 88 ;  Tr im er es ur us fu ca tu s " ) "  8  " &< >< B8  " @ A> <> >8 " @ A> <> AB  6     6     /  #    e t a l ! 88 :;  Tr im er es ur us fu ca tu s "  : " @ A> <> B  6     6     /  #     e t a l ! 88 :;  Tr im er es ur us fu ca tu s < B " @ A> <> >B F 6     6     /  #  )     5 6    ! 88 ;  Tr im er es ur us n eb ul ar is "  " @ A> <> >A " @ A> <> :  )     %   +         e t a l ! 88 :;  Tr im er es ur us n eb ul ar is  A=  " @ A> <> 8  )     %   +         e t a l ! 88 :;  Tr im er es ur us n eb ul ar is  A:  " @ A> F " @ A> <> >  )     %   +        e t a l ! 88 :;  Tr im er es ur us n eb ul ar is  A>  " @ A> <> ?  )     %   +         e t a l ! 88 :;  Tr im er es ur us n eb ul ar is  A?  " @ A> <> > " @ A> <> A>  )     %   +        e t a l ! 88 :;  Tr im er es ur us n eb ul ar is A =  " @ A> <> >= F )     %   +      )     5 6    ! 88 ;        … … co nt in ue d on th e ne xt p ag e Zootaxa 4347 (2) © 2017 Magnolia Press · 305CRYPTIC SPECIATION IN TRIMERESURUS     !%    ;                                            Tr im er es ur us n eb ul ar is A :  " @ A> < " @ 8= B= B8 " @ 8= B= = " @ 8= B= A?  H   /    /  #     e t a l. ! 88 :;  Tr im er es ur us p op ei or um &) * +  =? B= 8 " @ 8= B= >< " @ 8= B= B8 " @ 8= B= = " @ 8= B= A?  H   /    /  # (/      4   )     5 6    ! 88 ;  Tr im er es ur us p op ei or um %"  8 =? >  " @ A> <> ?A " @ A> <> =<  )       4 # $    @      e t a l ! 88 :;  Tr im er es ur us p op ei or um %"   < B=  " @ A> <> >> " @ A> <> A?  )     )     $C   6       e t a l ! 88 :;  Tr im er es ur us p op ei or um %"  < :: 8B  " @ A> <> >: " @ A> <> A?  )     )     $C    6       e t a l ! 88 :;  Tr im er es ur us p op ei or um %"   =B A  ) & >: ?: = ) & >: ?> 8 ) & >: ?: 8 ) & >: ?= : )     6    4 # ( 4   4   6     Tr im er es ur us p op ei or um %"   >> =  ) & >: ?: : ) & >: ?> < ) & >: ?: < ) & >: ?= > )     6    4 # ( C      4   6     Tr im er es ur us p op ei or um  ' * ) = ?> B? ? ) & >: ?: > ) & >: ?>  ) & >: ?:  ) & >: ?= ? )     6    4 # ( H   6     Tr im er es ur us p op ei or um ' * ) = ?> B< ? ) & >: ?: ? ) & >: ?> A ) & >: ?: A F )     6    4 # ( @     6     Tr im er es ur us p op ei or um  ' * ) = ?> B< B ) & >: ?: B ) & >: ?>  ) & >: ?:  ) & >: ?= B )     6    4 # ( @     6     Tr im er es ur us p op ei or um " 8  " &< >< B8  " @ A> <> ? " @ A> <>   6     %   G  /  #  )     5 6    ! 88 ;  Tr im er es ur us p op ei or um " 8 = " &< >< B8 : " @ A> <> :> " @ A> <> <  6     J*   J    e t a l ! 88 :;  Tr im er es ur us p op ei or um *  :>  " @ A> " @ A> <> ?< " @ A> <>   6     /  -   / #     e t a l ! 88 :;  Tr im er es ur us p op ei or um " )  =  " @ A> <> :? " @ A> <> =  6     /     /  #     e t a l ! 88 :;  Tr im er es ur us p op ei or um A  " @ 8= B= > " @ 8= B= B< " @ 8= B= =? " @ 8= B=   6     /     /  #  )     5 6    ! 88 ;  Tr im er es ur us sa ba hi " )  A   " @ A> <> >< " @ A> <> A:  )           )     5 6    ! 88 ;  Tr im er es ur us sa ba hi " )  A <  " @ A> <> > " @ A> <> A  )       !        ;   e t a l ! 88 :;  Tr im er es ur us to ba  ) , -. /+ 0= A  C /? BB : : C /B AB A  F F 0            '    / #  3   et a l ! 8< :;  Tr im er es ur us to ba  ' 6" -G : <: <  C /? BB : = C /B AB A A F F 0            '    / #  3   et a l. ! 8< :;   K3               T . p hu ke te ns is ! 4     ; MULCAHY ET AL. 306 · Zootaxa 4347 (2) © 2017 Magnolia Press    !                        Tr im er es ur us !P op ei a; p op ei or um   Tr im er es ur us !P op ei a; fu ca tu s !               8 8 ;                     6     4 #   )     %    ' * ) = ?> =? ?  ' * ) = ?> B< B ' * ) = ?> B 8 ' * ) = ?> B < T. (P .) po pe io ru m T. (P .) po pe io ru m T. (P .) fu ca tu s T. (P .) fu ca tu s L M  M  M  M  M  M  M  M   H 8 : =   : B = F F F F 6 H  >?  ?8  >  :  F F F F 6 H D6   H     8 <> 8 8 <= 8 8 <: 8 8 8 B 8 A 8 8 < <: B <= < <= >F <> 8      :8  =>  :=  >  =B F> = =: F:  :B F?  =B F> A       <8 D< < <8 D< < << D< 8 << D< < BF <<  BF <<  BF <  BF <  0