TEXAS TECH UNIVERSITY urn Natural Science Research Laboratory Occasional Papers Museum of Texas Tech University Number 286 8 April 2009 Sportive Lemur Diversity at Mananara-Nord Biosphere Reserve, Madagascar Holland’s Sportive Lemur Lepi lemur hollandorum Masoala Penninsula 40 Km Mananara-Nord Ivotaka-Nord Ivotaka-Sud Verezanantsoro Mananara-Nord Biosphere Reserve Front cover: Description of a new sportive lemur, Holland’s or Mananara-Nord sportive lemur, from Mananara-Nord Biosphere Reserve, Madagascar. Created by Lisa Kimmel. Sportive Lemur Diversity at Mananara-Nord Biosphere Reserve, Madagascar Borome Ramaromilanto, RunhuaLei, Shannon E. Engberg, SteigE. Johnson, Brandon D. Sitzmann, and Edward E. Louis, Jr. Abstract Molecular genetic sequence variation among the sportive lemurs (genus Lepilemur) of Madagascar was investigated utilizing mitochondrial DNA (mtDNA) sequence data ( ca . 3,000 base pairs). We infer Lepilemur phylogenetic relationships based on the topology of mtDNA trees generated with 225 individuals from 24 currently recognized species of the genus from 43 field sites, along with a previously undefined taxon from Mananara-Nord Biosphere Reserve. We formally describe this new species based on genetic analyses and biogeographic information (including the recent Inter-River-Systems model), supplemented with morphological data. Key words: biogeography, Lepilemur , Madagascar, Mananara-Nord Biosphere Reserve, phylogenetics, sportive lemur Introduction Madagascar’s remarkable species diversity and high levels of endemism are persistently under threat from anthropogenic pressures (Mittermeier et al. 2006; Harper et al. 2007). Consequently, the island has been ranked among the world’s most important biodiversity hotspots, underscoring the need for coordinated conser¬ vation efforts (Myers 2000; Groombridge and Jenkins 2002). While most of the Malagasy fauna is susceptible to extinction risk, lemurs are especially vulnerable due to their relatively small and often fragmented geographic ranges (Jemvall and Wright 1998). As a result, lemurs are protected under the Convention on International Trade in Endangered Species (CITES) and many species are red-listed as Critically Endangered, Endangered or Vulnerable by the IUCN SSC (IUCN 2008). Many species are still data deficient (IUCN 2008), and their number is likely to increase due to the recent rapid expansion in the number of recognized species (e.g., Andriaholinirina et al. 2006; Louis et al. 2006b; Rabarivola et al. 2006; Craul et al. 2007,2008; Lei et al. 2008). In addition, the description of new spe¬ cies often reduces the geographic range of traditionally recognized taxa (Louis et al. 2006b). Thus, frequent re-evaluation of the conservation status of all lemur species is necessary, using newly available informa¬ tion on taxonomy, biogeography, ecology, ethology, and current threats. Among the most widely distributed lemur groups, the sportive lemurs (genus Lepilemur ) are medium¬ sized nocturnal lemurs that were originally thought to consist of only two species, L. mustelinus from the eastern rainforests and L. ruficaudatus from the western and southern dry forests of the island (Schwarz 1931; Hill 1953). Recent investigations using molecular ge¬ netic, cytogenetic, and morphological data have greatly expanded the diversity of this genus (Louis et al. 2006b; Rabarivola et al. 2006; Craul et al. 2007, 2008; Lei et al. 2008). These developments in Lepilemur taxonomy are a salient example of the need for verifying unique diversify in order to assess the conservation status of a group. Currently, there is one vulnerable, one endan¬ gered, and one critically endangered Lepilemur species, while 21 of the listed species (most of them recently described) remain data deficient (IUCN 2008). Northeastern Madagascar is one region where sportive lemur taxonomy is currently in revision. L. mustelinus was formerly the only recognized taxon between the Bemarivo and Mangoro Rivers (Petter et 2 Occasional Papers, Museum of Texas Tech University al. 1977; Tattersall 1982). More recently, two previ¬ ously undefined species have been described from this region. Louis et al. (2006b) described L. seali from Anjanaharibe-Sud Special Reserve, while Lei et al. (2008) introduced L. scottorum from the Masoala Peninsula (Fig. 1). Furthermore, Craul et al. (2008) extended the distribution of L. seali south of the Antainambalana River. Louis et al. (2006b) assigned the sportive lemur from Mananara-Nord Biosphere Reserve (MNBR) to L. seali although the available molecular data suggested that this population would eventually be described, pending further field studies. With only a single representative from MNBR avail¬ able, Lei et al. (2008) established it as an undefined species, Lepilemur species nova #2. By incorporating sequence data from Louis et al. (2006b), Craul et al. (2008) subsequently reconfirmed the uniqueness of the population at MNBR. Three biogeographic models have been proposed based on different relative contributions of factors including large rivers (>50 m wide at 20 km inland), retreat dispersion watersheds, and topographical bar¬ riers, i.e. mountains (Martin 1995; Wilme et al. 2006; Craul et al. 2007; Olivieri et al. 2007). In Olivieri et al. (2007) and Craul et al. (2008) the authors presented biogeographic models in which “centers of endemism” were defined based on the isolation effects of paired rivers, or Inter-River-System (IRS; Fig. 1). During the course of several biogeographic revisions of northern and northwestern Madagascar, the number of IRS has increased from four (Martin 1995), to five (Wilme et al. 2006), and to nine (Craul et al. 2008). In Craul et al. (2008), an initial IRS model was presented to delin¬ eate the ranges of the sportive lemurs of northeastern Madagascar. We aim to add to the baseline understand¬ ing of the diversity in the greater Antongil Bay region, which includes Anjanaharibe-Sud Special Reserve, Masoala National Park, Makira Forest, Mananara- Nord Biosphere Reserve, and adjacent habitats. By revising the known biodiversity of sportive lemurs for this region, we present an amended IRS model to one of the largest remaining tracts of intact forest in Madagascar (Fig. 1). Historically, the Biological Species Concept (BSC), emphasizing reproductive isolation, has been the most common approach to define species (Mayr 1942). However, when a putative species is geo¬ graphically isolated from closely related species, this concept is difficult to implement. The Phylogenetic Species Concept (PSC) employs a cladistic perspec¬ tive, incorporating evolutionary patterns of ancestry and descent as it defines species operationally as the smallest diagnosable, distinct cluster of individuals (Cracraft 1983; Wheeler and Platnick 2000; Groves 2001a, b). It is also useful in defining conservation units (Vogler and DeSalle 1994). In this paper, we present a taxonomic analysis of the genus Lepilemur using the PSC, including a formal description of a new species from MNBR. As previously discussed in Andriantompohavana et al. (2006), Louis et al. (2006a, b) and Thalmann and Geissmann (2005), the utilization of whole vouchers as the designated holotype for a new species is not a prerequisite for species descriptions; opportunis¬ tic collections can later supplement morphological, vocalization, and/or molecular data in combination with curated blood and/or tissue samples (Jones et al. 2005). The sportive lemurs are a prime candidate for this methodology because the highly folivorous dietary requirements of this group currently preclude any attempts to curate “live vouchers” (Thalmann and Geissmann 2005). Total genomic DNA for the three paratype specimens is currently curated at the Museum of Texas Tech University (TK125726; TK125727; TK125728). Additionally, an electronic database that contains all Lepilemur field data and photographs, in¬ cluding data for the paratype specimens, is curated at the Museum of Texas Tech University. The database is stored in the Type Specimen Collection in multiple media formats. This collection of field data and pho¬ tographs, as well as additional tables and figures, is also available online at the website of Omaha’s Henry Doorly Zoo. See Appendices I-III for a directory of appropriate website addresses. Ramaromilanto et al.—Mananara-Nord Sportive Lemur 3 IKS U .Ambatoledama Forest Mangabe Rantabc River Fanartfhanjg Fahumtralfl Vlaiiinuor> Rivtr Onibc River i> km 00 km Bemarivo River -—Marojejy National Park Anjanaharibe-Sud UJ Special Reserve 4ukavana River Makira Forest IRS A R [I Forest Masiaposa W Forest Lokaitra ET Forest IRS MAA IRS Rl h Masoala National Park _Antsahabe Q] Forest IRS FM Manana rn River Ivotaka-Nord -Ivotaka-Sud IRS MS -r— j- Mananara-Nord Verezanantsoro [3j Biosphere Reserve novc Rivci IRS SMAS - J | . Amhatovakv UJ Special Reserve __ Marotandrano 121 Special Reserve SiHiilralsio Ri^ er Zaliamena National Park Betampona Special Reserve IRS Mza Figure 1. Map of northeastern Madagascar. This regional map features the potential boundaries and forest tracts for the sportive lemurs (genus Lepilemur) in northeastern Madagascar. Samples collected in the Ivontaka-Sud and Verezanantsoro parcels of Mananara-Nord Biosphere Reserve, Madagascar, were evaluated, along with sequence data from Louis et al. (2006b) and Lei et al. (2008). The Inter-River-System (IRS) model was adapted from Craul et al. (2008). IRS A and IRS RF from Craul et al. (2008) were revised to IRS AA and IRS RFa and RFb, respectively. IRS SMAS, IRS SM, and IRS MZO were inserted to define the potential boundaries for L. species nova #2 (Lei et al. 2008) and L. mustelinus. IRS AAMB was inserted to define the potential northern boundary for L. seali. IRS MAA was inserted to define the potential boundaries for L. scottorum. The boxed numbers found below the horizontal lines indicate what sportive lemur is designated for that site or IRS as follows: 0 specifies L. seali ; 0 specifies L. scottorum ; 0 specifies L. sp. nova #2; 0 specifies L. mustelinus ; and 0 specifies undefined species or data unavailable. 4 Occasional Papers, Museum of Texas Tech University Methods Sampling. —All lemurs investigated in this study were wild-caught, free-ranging individuals immobilized with a C0 2 projection rifle or blowgun with 10 mg/kg of Telazol (Fort Dodge Animal Health; Overland Park, Kansas; Fig. 1; Table 1). Four 2.0 mm biopsies and 1.0 cc per kilogram of whole blood were collected from each sedated animal and immediately stored in room temperature storage buffer (Longmire et al. 1992). A Home Again microchip (Schering-Plough Veterinary Corp.; Kenilworth, New Jersey) was placed subcutaneously between the scapulae of each lemur (Appendix 1(a)). This procedure was used to field- catalog each animal with a unique recognition code in order to re-identify all captured individuals during any future immobilizations. In addition, morphometric measurements were taken. For presentation purposes, we present the weight, head crown, body length, and tail length in this publication following the guidelines of Smith and Jungers (1997; Appendices I(a-b)). Field data, including all measurements and e-voucher pho¬ tographs, are available in Appendix 1(b), Louis et al. (2006b), and Lei et ah (2008). Data Collection. —We recorded the location of all immobilized lemurs using a global positioning system (GPS; Appendix I(a-b)). Genomic DNA was extracted from samples using a phenol-chloroform/iso- amyl extraction (Sambrook et al. 1989). We recorded the location of all immobilized lemurs using a global positioning system (GPS; Appendix I(a-b)). From these samples, the following regions of mtDNA were amplified: the displacement loop or control region (D- loop; Baker et al. 1993; Wyner et al. 1999), a fragment of the cytochrome oxidase subunit III gene (COIII), NADH-dehydrogenase subunits 3, 4L, and 4 (ND3, ND4L, and ND4), as well as the tRNA Gly , tRNA Arg , tRNA Hls , tRNA Ser , and partial tRNA Leu genes (subse¬ quently referred to as the PAST fragment; Louis et al. 2006b). The accessioned sequences of L. seali and L. mittermeieri of Craul et al. (2008) and Rabarivola et al. (2006), respectively, were not compatible with this data set so were not utilized in this study. Although avail¬ able as accession sequences in GenBank, we would have had to truncate and eliminate two-thirds of our generated sequence data to include their L. seali and L. mittermeieri accession fragments in these analyses. Using 50 nanograms of genomic DNA, the D-loop (555 bp) and the PAST (2,378 bp) fragments were amplified using the following conditions: 94°C for 30s, 47°C for 45s, and 72°C for 45s for 34 cycles. Since potential nuclear insertions or mitochondrial pseudogenes within the nuclear genome can be amplified inadvertently, we minimized the likelihood by amplifying both mito¬ chondrial DNA regions as intersecting or overlapping segments and confirming these segments with the degenerate oligonucleotide-primed PCR methodology (Telenius et al. 1992; Zhang and Hewitt 1996; Louis et al. 2006b). The samples were electrophoresed on a 1.2% agarose gel to verify the PCR product and purified with Exonuclease I and Shrimp Alkaline Phosphatase (EXOSAP; Silva et al. 2001). The purified products were cycle-sequenced us¬ ing a BigDye terminator sequencing kit (Applied Bio¬ systems; Foster City, California). The sequences were analyzed by capillary electrophoresis with an Applied Biosystems Prism 3130 genetic analyzer. The PCR and sequencing primer suite from Louis et al. (2006b) and Lei et al. (2008) were used to generate the D-loop and PAST fragment sequences. The sequence frag¬ ments were aligned to generate a consensus sequence using Sequencher 4.8 (Gene Codes Corporation; Ann Arbor, Michigan), and the consensus sequences were aligned using ClustalX 1.83 (Thompson et al. 1997). All aligned sequences are available from the first author upon request. All sequences have been deposited in GenBank, and the sequence data and information are available from the referenced accession numbers (Table 1; Appendix 1(a); Louis et al. 2006b; Lei et al. 2008). Phylogenetic Analysis. —Maximum likelihood (ML) analyses for the D-loop and PAST fragment sequence data were performed under the GTRCAT algorithm implemented in the parallel Message Pass¬ ing Interface (MPI) version of RAXML-VI-HPC (Stamatakis 2006; software available at http://icwww. epfl.ch/~stamatak). The best-scoring ML-trees were searched and saved in PAUP* 4.0bl0 (Swofford 2001). Bayesian inference analyses of the D-loop and PAST fragment sequence data were conducted using MrBayes 3.0b4 (Huelsenbeck and Ronquist 2001; Ronquist and Huelsenbeck 2003). A Markov Chain Monte Carlo (MCMC) run with four simultaneous chains and 1,000,000 generations was performed. Table 1. Samples (225 Lepilemur and 29 outgroups total) used in the present genetic study and taxonomic revision of the genus Lepilemur. TK Number “ is referenced voucher curated at the Museum of Texas Tech University. Mitochondrial DNA sequence data for PAST (Pastorini Fragment), and D-loop (D-loop or control region) for each sample are available from GenBank under the listed accession numbers'**. Ramaromilanto et al.—Mananara-Nord Sportive Lemur H C/3 & a 1/3 t =3 u a, c/3 13 g ‘5b 'L CL> X £ =5 z pq H 53 £ X 3- VO t-~ 00 ov o Cl re 3- ov o Cl cn OV VO VO VO VO s IR¬ ts IR¬ iR- in VO VO vo vo VO On ov On On On ON On ON ov ov On OV ov On OV On On ov On ov OV OV Ov ov ov On OV OV On Cl IS is c- c- s- s s- IS s s t~- ir¬ IR¬ S- 'Cl IS IS is IS s- ts s s t> Is t> is IS IS a P P P P P P P P P P P P P p Q m pq m pq pq pq pq pq pq pq pq pq pq pq Cl Cl m 3- in vo IS 00 On o in VO IS 00 Ov 'n 00 00 00 00 00 00 00 00 On s ts s s IS in On ov On On ov ov On On On On On ov OV On ov On ov On On On ov On On ov On On ov ov On Cl s- IS IS IS I s " s IS S S S IS o s S in ts s- S IS ts s- ts ts IS S s- s- s r- a P p P P p P P P p P p p p P Q pq pq pq pq pq pq pq pq m m pq pq pq pq I g R s; =5 C /3 c3 I O > o Z VO Cl r- H "§ j§ 'o *r k a r & si T3 =5 C SI c3 M 1 O > o Z P4 H g R El Sr si | | g £ £ a s ■S ’S' -i J •§ g § o § o § o i o o VJ R QJ VJ << a S 'S "o "o s: ■s: s i S Ik a S s 1 s § I V3 N C3 cs o s s b £ « S 1 s 1 S) 1 S 1 S 'Vj VJ 1 £ s, si si s § JS § JS g § Co Co Co >5 R R R R R R TS *3 •a hR R3 b b k. •SJ vS vS R R 1 g 1 ‘S § "1 -1 "R s 1 R i i R k. k. 5J k. s i s s i R s s 1 a R 1 l R s 'i) 'VJ U ‘Sh 'S, Sr 1 R. S 55 S) s5 S -S3 ^3 s SI SI S o -a o 4 I r I T3 3 op c3 M 1 O > o Z X 3 op c3 M I O > o Z T3 3 op 3 X f o > o Z T 3 3 C/3 3 o > t3 .3 ,g .3 .3 o'' Z § s c < M H VO IS s is o o o o S PQ m m 3 3 3 3 cn Cl o m in (S 3 in PQ £ 2 H y S > H o N < o N S| o o o o _ca X) X) X B a 3 a X a >N o ta X C /5 X 3 X £ < ¥ X 3 o £ o 3 -B o X) o X) TO o X •B o X £ < o N o N 13 o B S s B X < < C < < G ,P C* pq VO P s Anjiamangirana Daubentonia madagascariensis Daubentonia madagascariensis EU779980 EU779964 6 Occasional Papers, Museum of Texas Tech University The model of evolution was selected for the ML infer¬ ences by using Mrmodeltest 2.2, a modified version of Modeltest 3.6 (Posada and Crandall 1998; Nylander 2004). It was performed with HKY+I+G Model for D-loop and GTR+I+G Model for PAST, for 1,000,000 generations. Every hundredth generation, the tree with the best likelihood score was saved, resulting in 4,000 trees. These were condensed in a majority rule consensus tree using PAUP* 4.0bl0 (Swofford 2001), and clade posterior probabilities (PP) were computed. The pattern of sequence evolution was estimated by conducting a minimum spanning network generated with the program NETWORK version 4.500 (Bandelt et al. 1999) and Arlequin version 2.0 (Schneider et al. 2000). MEGA 3.1 (Kumar et al. 2004) was used to calculate uncorrected pairwise distances (‘p’) and Kimura distance measures (Kimura 1980) for D-loop and PAST fragments. We utilized MacClade 3.01 (Maddison and Maddison 1992) and MEGA version 3.1 (Kumar et al. 2004) in a diagnostic search to designate evolutionary significant units (ESU) using population aggregate analysis (PAA) of the D-loop (550 bp) and PAST (2,378 bp) sequence data for genus Lepilemur (Davis and Nixon 1992; Louis et al. 2006a, b; Lei et al. 2008). With the sequential addition of each individual without an a priori species designation, a PAA distinguishes attributes or apomorphic characters according to the smallest definable unit (Davis and Nixon 1992; Vogler and DeSalle 1994; Groves 2001a, b; Louis et al. 2006b; Lei et al. 2008). Results Mitochondrial DNA sequence data were complet¬ ed for two fragments, D-loop and PAST ( ca . 3,000 bp), for 225 individuals representing 24 recognized species of sportive lemurs collected from 43 field sites (Table 1; Louis et al. 2006b; Lei et al. 2008). Due to different mtDNA fragments utilized in Rabarivola et al. (2006), congruent sequence data were not available and thus, L. mittermeieri was not included in this study. Based on the phylogenetic inferences of the ML and Bayesian analyses of D-loop and PAST sequence alignments, 24 Lepilemur species were represented in 24 distinct and well-supported terminal clades. These terminal clades could be partitioned into four geographic regions (Figs 2-3; Appendices Il(a-h)). In general, Section A consists of sportive lemurs from northern and northwestern Madagascar as follows: L. ankaranensis, L. milanoii, L. septentrionalis, L. tymerlachsoni, L. dorsalis , L. sahamalazensis, and L. ahmansonorum. Section B is associated with northwestern Madagascar: L. otto, L. edwardsi, and L. grewcockorum. Section C cor¬ responds to southern and west central Madagascar as follows: L. hubbardorum , L. ruficaudatus , L. aeeclis , L. randrianasoli, L. leucopus, and L. petteri. With the exception of L. microdon , Section D incorporates the sportive lemurs of eastern Madagascar as follows: L. mustelinus, L. jamesorum , L. betsileo, L. fleuretae, L. wrightae, L. seali, L. scottorum, and L. species nova #2. Furthermore, all phylogenetic methods support the uniqueness of the subpopulation, Lepilemur species nova #2 from MNBR (Figs. 2-3). All methods revealed the same phylogenetic proximity between regions and among sportive lemur species, resulting in distinguishable eastern and western clades with the exception of L. microdon (Figs. 2-3; Appendices Il(a-h)). The phylogenetic association of L. microdon to Sections B or C varied according to which mtDNA sequence fragment was analyzed. Based on either phylogenetic method, L. microdon clusters with the northwestern sportive lemurs (Section B) for the D-loop sequence fragment. However, for the PAST sequence fragment, this east coast sportive lemur’s association shifts to the species located in west central and southern Madagascar (Section C; Figs. 2-3; Appendices (a-h)). The complete uncorrected ‘p’ distance and the Kimura two-parameter distance measures for the genus Lepilemur are presented for D-loop and PAST fragments in Appendices Ill(c-d). Values ranged from 7.0% to 10.9% and from to 4.2% to 11.0% for D-loop and PAST, respectively, for the three sportive lemur species geographically closest to L. species nova #2. The minimum spanning network diagrammatically presents the relative evolutionary associations among 25 sportive lemur species (Fig. 4). Ramaromilanto et al.—Mananara-Nord Sportive Lemur 7 Majority rale -ANK7- ■Microcebus ravehbemis -Microcebus rufiis •Allocebus trichotis 'Mirza zaza •Mirza coquereli •Protemur simus •Lemur catta RANO250 ■ TVY7.23 - ANT5.3 — ZOMB54 — KIAN124 RAN0338 REN31— RAN0351 I_„ . RAN0352 j ~ ~• Hapaiemur aureus RAN061 I ,, , RAN 062 !- Hapaiemur griseus AN AL2.23 | Hapaiemur occidenialis FAN 21 Varecia varlegata variegate r RANQ45- Eukmur ruftfions MER1- Eukmur coronatus Eukmur mongo- •Daubentonia madagascariemis •Cheirogakus medius •Cheirogaleus major 'Propithecus edwadsi Propithecus verreauxi •Indri indri Avahi peyrierasi ■Avahi occidenialis ■Phaner Julci/er _ I O —— bp 3 o 99 0.002 LSI _r- 81 0.009 0.002 1 - 77 0.002 _| 0 002 ' Lepilemur ankaranensis Lepilemur milanoii Lepilemur tymerlachsoni Lepilemur septentrional is Lepilemur dorsalis Lepilemur sahamal azensis Lepilemur ahmansonorum -A Lepilemur otto Lepilemur edwardsi Lepilemur erewcock orum Lepilemur hubbardorum Lepilemur ruficaudatus Lepilemur aeeclis Lepilemur randrianasoli Lepilemur leucopus Lepilemur petteri _ Lepilemur microdon Lepilemur mustelinus Lepilemur jamesorum Lepilemur betsileo Lepilemur fleuretae Lepilemur'seali Lepilemur scottorum Lepilemur species nova #2 KALA4.9 KALA5.12 KALA4.I8 Lepilemur wrightae KALA24 KALA4.I6 -B -C -D Figure 2. Phylogenetic relationships between Lepilemur species inferred from the maximum likelihood and Bayesian approaches for the D-loop sequence data from 73 haplotypes from the 225 Lepilemur individuals with 29 outgroup taxa. Numbers above the branches represent posterior probability support. Numbers below the branches represent ML values. We obtained the maximum likelihood tree (-Ln likelihood=l 2,404.88) from the D-loop alignment using a transition/transversion ratio of 2 (k=3.99). Section A consists of sportive lemurs from northern and northwestern Madagascar. Section B consists of sportive lemurs from northwestern Madagascar. Section C consists of sportive lemurs from west central and southern Madagascar. Section D consists of sportive lemurs except for L. microdon (associated with Section C) from eastern Madagascar. Outgroup taxonomy based on Mittermeier et al. (2008). 8 Occasional Papers, Museum of Texas Tech University r^M n ha ANAL 13 ANAL2.1 ANAL2.4 ANAL2.13 LABES. 1 ANAL2.2 CAR46 CAR6 CAR 15 CARI7 CAR2I CAR33 CAR4S F1A5.6 CAR5S CAR59 MER26 LAMES.6 FIA5.1 FI AS. 11 FI AS. 17 99 95 — 0.009 99 0001 99 87 0.016 Connects to Lepilemur Taxa Sections C and D Lepilemur ankaranemis -A Lepilemur milanoii Lepilemur tymerlachsoni Lepilemur ahmcmsonorum Lepilemur sahamalazensis Lepilemur dorsalis Lepilemur septenlrion alis Lepilemur edwardsi Lepilemur grewcockorum Lepilemur otto Lepilemur microdon -B Figure 3a. Phylogenetic relationships between Lepilemur species inferred from the maximum likelihood and Bayesian approaches of PAST fragment sequence data from 161 haplotypes from the 225 Lepilemur individuals with 29 outgroup taxa. Numbers above the branches represent posterior probability support. Numbers below the branches represent ML values. We obtained the maximum likelihood tree (-Ln likelihood=38,278.20) from the PAST alignment using a transition/transversion ratio of 2 (k=4.23). Section A consists of sportive lemurs from northern and northwestern Madagascar. Section B consists of sportive lemurs from northwestern Madagascar. Ramaromilanto et al.—Mananara-Nord Sportive Lemur 9 Connects to Outgroup < i Connects to Lepilemur Taxa Sections A and B Lepilemur petteri Lepilemur leucopus Lepilemur ruficaudatus Lepilemur hubbardorum Lepilemur randrianasoli Lepilemur aeeclis -c Lepilemur mustelinus Lepilemur betsileo Lepilemur jamesorum Lepilemur fleuretae Lepilemur wrightae Lepilemur seali __ Lepilemur species nova #2 HHH Lepilemur scottorum _ -D Figure 3b. Phylogenetic relationships between Lepilemur species inferred from the maximum likelihood and Bayesian approaches of PAST fragment sequence data from 161 haplotypes from the 225 Lepilemur individuals with 29 outgroup taxa. Numbers above the branches represent posterior probability support. Numbers below the branches represent ML values. Section C consists of sportive lemurs from west central and southern Madagascar. Section D consists of sportive lemurs except for L. microdon (associated with Section B) from eastern Madagascar. Figure 4. Minimum spanning network of Lepilemur haplotypes calculated using Arlequin Version 2 and Network Version 4.11. Identification numbers denote haplotypes corresponding to Table 1 (Appendix 111(e)). The minimum number of mutational steps separating matriarchal lines is indicated. Nucleotide substitutions are indicated by dashes. The number of nucleotide differences in their connecting lines is indicated when there are more than two. Missing intermediates are indicated by gray circles. The size of the circles is representative of the number of individuals with matching haplotypes (circles without any number represent one individual). Ramaromilanto et al.—Mananara-Nord Sportive Lemur 11 Discussion Anthropogenic pressure has resulted in the fragmentation of panmictic populations, which has compelled wildlife and conservation agencies to make management priorities according to the current understanding of taxonomy, historical biogeography and present distribution of these now isolated popula¬ tions (Wilme et al. 2006; Kremen et al. 2008). Many studies have shown that molecular genetics offers a reliable and rapid method of identifying unique and/or cryptic biodiversity (e.g., Louis et al. 2006b; Craul et al. 2007, 2008). Through the analyses of accessioned and novel sample sets, we found that each described sportive lemur clusters in distinct and well-supported terminal clades, along with the undefined individuals from MNBR. With this aim, we present here a revision of the genus Lepilemur , concentrating on the distri¬ bution of the sportive lemur species in northeastern Madagascar. Ganzhorn et al. (2006) described the importance of rivers as corridors referring to cytogenetic recon¬ structions of the genus Lepilemur , specifically the sister relationship of a western sportive lemur, L. edwardsi (Ankarafantsika National Park) and an eastern sportive lemur, L. microdon (Ranomafana National Park; Ishak etal. 1992). The current phylogenetic reconstructions from sequence-based data of the genus Lepilemur have substantiated this finding (Louis et al. 2006b; Lei et al. 2008). However, incongruence between the D-loop and PAST inferences in regard to the phylogenetic re¬ lationships ofL. microdon to the northwestern sportive lemurs (Section B) and the west central and southern sportive lemurs (Section C) presents an alternative system of river corridor dispersal (Figs. 2-3). This finding illustrates the importance of developing and assessing novel nuclear DNA based sequence fragments to examine further the significance or function of river corridors in speciation. The population aggregate analysis (PAA) results from the D-loop and PAST sequence fragments are presented in Tables 2Aand 2B, respectively (Appendix Ill(a-b)). Multiple diagnostic characters distinguish each described sportive lemur, along with the unde¬ fined sportive lemur from Mananara-Nord Biosphere Reserve. Lepilemur species nova #2 had 12 diagnostic sites for D-loop and PAST fragments combined. Ac¬ cording to the Phylogenetic Species Concept ( sensu Wheeler and Platnick 2000), diagnostic characters or attributes define Evolutionary Significant Units (ESUs). Several authors suggest that ESUs are equivalent to species and reflect species barriers (Cracraft 1983; Groves 2001a). Lepilemur species nova #2 had mul¬ tiple molecular diagnostic sites (Tables 2Aand 2B; Ap¬ pendices Ill(a-b)) and, given this criterion, represents a distinct ESU. The continuous addition of sequence from novel samples and sites to the PAA data set will dynamically test the distinction and diagnostic ability of these characters, and, therefore, the ongoing status of this and related species. A summary of the morphometric data for the three sportive lemur species of northeastern Madagascar, along with the proposed species, Lepilemur species nova #2, are presented in Tables 3A and 3B (detailed morphological measurements of the novel sportive lemur are available in Appendix 1(b)). No extensive quantitative analyses were conducted on the morpho¬ metric data at this point (Table 3). Therefore, this morphometric information is provided as supplemental data, complementing the molecular data used to parti¬ tion unique biodiversity. Nevertheless, it can be seen that the new species most resembles L. mustelinus in size, but has a distinctly longer tail and shorter pollex and hallux. Meanwhile, we describe the new species as follows: Lepilemur hollandorum , New Species Lepilemur species nova #2 of Lei et al. (2008) Type Specimen. —NARA4.20 (TK125726; TTU- M 109031), adult female; collected on 8 August 2004, captured at Mananara-Nord Biosphere Reserve. Ma¬ terial: Total genomic DNA (50ng/pl) for NARA4.20 (TK 125726), adult female, stored and curated at the Museum of Texas Tech University, Lubbock, Texas, USA. Two 2.0 mm biopsies from ear pinna, and 0.007 cc of whole blood tissues stored at Henry Doorly Zoo, Omaha, Nebraska, USA. A microchip pit tag was placed subcutaneously between the scapulae and re¬ corded as 4556420944. NARA4.20 was collected by Edward E. Louis, Jr., Richard Randriamampionona, 12 Occasional Papers, Museum of Texas Tech University Table 2A. Summary of Population Aggregate Analysis (PAA) D-loop diagnostic sites for Lepilemur. Refer to Appendix 111(a). The locality o/Lepilemur species nova #2 is Mananara-Nord Biosphere Reserve. *No character or attribute is available for this fragment. Species Fragment Size (bp) PAA base pair location Lepilemur ankaranensis 540 * Lepilemur milanoii 540 130 Lepilemur tymerlachsoni 538 * Lepilemur septentrionalis 536 33,43, 111, 113,249 Lepilemur dorsalis 540 536, 537 Lepilemur sahamalazensis 542 * Lepilemur petteri 534 * Lepilemur leucopus 535 19 Lepilemur ruficaudatus 535 103, 126,310 Lepilemur hubbardorum 535 242, 253, 270, 302 Lepilemur randrianasoli 538 33,272 Lepilemur edwardsi 545-546 127 Lepilemur grewcockorum 544 139, 195, 357 Lepilemur ahmansonorum 542 * Lepilemur aeeclis 537-538 21 Lepilemur mustelinus 552-553 * Lepilemur jamesorum 552 132 Lepilemur betsileo 553 272, 273, 286 Lepilemur fleuretae 550 10, 24, 37, 200, 287, 288, 314, 317, 330 Lepilemur microdon 530 25, 34, 107, 110, 120, 121, 123, 124, 125, 127, 137, 139, 396 Lepilemur wrightae 551 55, 58, 275, 301,476, 493 Lepilemur seali 550 54, 159, 221 Lepilemur species nova #2 550 87, 195,231,327, 475 Lepilemur scottorum 551 24, 30, 140, 187, 266 Lepilemur otto 545-547 160, 162 Ramaromilanto et al.—Mananara-Nord Sportive Lemur 13 Table 2B. Summary of Population Aggregate Analysis (PAA) PASTfragment diagnostic sites for Lepilemur. Refer to Appendix 111(b). The locality o/Lepilemur species nova #2 is Mananara-Nord Biosphere Reserve. Species Fragment Size (bp) PAA base pair location Lepilemur ankaranensis 2359-2360 364, 858, 1315, 1804 Lepilemur milanoii 2359 342, 769, 1896 Lepilemur tymerlachsoni 2359 152, 1309, 1378, 1861, 1898, 1995 Lepilemur septentrionalis 2360 44, 113, 211, 214, 274, 353, 354, 533, 551, 555, 576, 674, 734,1103,1174, 1231, 1347, 1399, 1448, 1492, 1550, 1582, 1603, 1630, 1777, 2144, 2146, 2363 Lepilemur dorsalis 2361 579, 717, 746, 1525, 1780, 2163, 2168, 2177, 2236 Lepilemur sahamalazensis 2360 204, 539, 737, 749, 770, 803, 1358 Lepilemur petteri 2360 337, 578,779, 957, 1615 Lepilemur leucopus 2360-2361 220, 719, 836, 1960 Lepilemur ruficaudatus 2360 94, 127, 235, 365, 776, 1074, 1370, 1783, 1835, 1867, 1921, 2068 Lepilemur hubbardorum 2361 350, 543, 566, 629, 681, 1012, 1015, 1240, 1396, 1559, 1906, 1907, 2111 Lepilemur randrianasoli 2360 191, 699, 849, 923, 982, 1018, 1035, 1053, 1219, 1432, 1444, 1753, 1981, 1988, 2250, 2267 Lepilemur edwardsi 2360 1018, 1474, 1979 Lepilemur grewcockorum 2360 406, 888, 896, 988, 1114, 1226, 1354, 1537 Lepilemur ahmansonorum 2360 46, 304, 350, 1096, 1097, 1818, 2141, 2170 Lepilemur aeeclis 2360 535, 548, 563, 581, 975, 1357, 1368, 1423, 1442, 1990, 2089, 2107 Lepilemur mustelinus 2358-2359 85 Lepilemur jamesorum 2359 2144 Lepilemur betsileo 2359 8, 1057 Lepilemur fleuretae 2359 29, 103, 269, 358, 533, 534, 546, 553, 664, 1124, 1574, 2013, 2023 Lepilemur microdon 2360 146, 510, 581, 596, 826, 829, 1171, 1954, 1991, 2077, 2164 Lepilemur wrightae 2359 55,133,663,691,871,907,942,1105,1117,1120,1837,1856,1936,2041, 2096,2181,2185, 2331 Lepilemur seali 2360 124, 147, 290, 626, 665, 692, 722, 1302, 1313, 1371, 1679, 1875, 1879, 1969 Lepilemur species nova #2 2360 86, 567, 1157, 1337, 1483, 1606, 2165 Lepilemur scottorum 2360 72, 256, 1033, 1112, 1167, 1237, 1336, 1538, 1902 Lepilemur otto 2360 115, 196, 328, 379, 702, 988, 1004, 1955 Table 3 A . Morphometric data collected from the four geographically closest species o/Lepi lemur from northeast Madagascar . ( In ¬ dividual morphological data available at Appendix 1 ( b ).) *Only adult individuals were included (based on body weight and canine length ). 14 Occasional Papers, Museum of Texas Tech University £ © v — / T- 1 o CO P ri p ri p "So -H -H -H HH O cn o 00 © P in’ NO K 00 CN ri ri ri "c 3 H 'B © ON r- &b c ,_J o’ ,_H -H -H 4 t -H © 00 in NO o p in’ t> NO CN ri n n T 3 o PQ s' p ON p >n o +1 o’ -H 1 — U ON in r- >n P IN- NO r-’ T 3 cd © ffi _ bO ,—i ON -ct n M n O o P’ o’ o’ O o’ xi if -H -H -H .00 On in 00 On ’© ON On 00 ON £ o’ o’ o’ o’ Z * 27 in in in Vh ^H Vh £ © =s £ £ © © © > c 3 z ts o p ts o O o i Oh CO 75 o Oh 02 05 Oh 02 J 02 £ eS © 73 O o U £ © CO o CO CN =tfc G § K s O a to 6 ? 02 _© ’© © a o © £ S to u to Oh !z a Sn. 5 h. t /2 § s § l © 50 cj © Oh 1 1 l l CO p o 1 o -f -Si .g n K J a o 73 § § £ O -O > § III [p p c3 ’S ^ e :S e H © o o Pi % S ^ = £ 5 U U ffi p 2 ^ u bo C is.s p 3 £ ^*3 u X £ ^ o § O Ph i-4 £ Z Oh CO on o' -H On o' ON o' -H IT- On’ o -H On ri o -H r- o 3 NO r- o 2 o’ -H CN NO -H NO m o' -H NO ri o -H NO o -H N- co o’ -H CN ON NO CN -H ON O -H CN NO o 4 n o’ -H