Page 113 of 152 Systematics and Biodiversity
1 2 3 4 1 Research Article 5 6 7 2 Unrecognized species diversity and new insights into colour pattern 8 9 10 3 polymorphism within the widespread Malagasy snake Mimophis 11 12 13 4 (Serpentes: Lamprophiidae) 14 15 16 5 17 18 1,2 2 3 3 19 6 SARA RUANE , EDWARD A. MYERS , KAHMUN LO , SARA YUEN , 20 21 7 RACHEL S. WELT 2, MAYA JUMAN 3, INDIA FUTTERMAN 3, RONALD A. 22 23 24 8 NUSSBAUM 4, GREGORY SCHNEIDER 4, FRANK T. BURBRINK 2 & 25 26 2 27 9 CHRISTOPHER J. RAXWORTHY 28 29 10 30 31 1 32 11 Department of Biological Sciences, 206 Boyden Hall, Rutgers University, 195 University Ave, 33 34 12 Newark, NJ 07102, USA 35 36 2 th 37 13 Department of Herpetology, American Museum of Natural History, Central Park West and 79 38 39 14 St., NY, NY 10024, USA 40 41 15 3Science Research Mentoring Program, American Museum of Natural History, Central Park 42 43 th 44 16 West and 79 St., NY, NY 10024, USA 45 46 17 4Division of Reptiles and Amphibians, Museum of Zoology, Research Museums Center, 3600 47 48 18 Varsity Drive, University of Michigan, Ann Arbor, MI 48108, USA 49 50 51 19 52 53 20 Running title : Mimophis systematics 54 55 21 56 57 58 59 60 1
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 114 of 152
1 2 3 22 Although many wide ranging taxa occur in Madagascar, phylogeographic studies for most of 4 5 6 23 these species are still lacking. This is especially the case for snakes, where of more than 100 7 8 24 endemic species, the population structure of only two species has so far been examined. Here, we 9 10 11 25 examine genetic population structure of one of the most common snakes of Madagascar, 12 13 26 Mimophis mahfalensis (Grandidier, 1867). This taxon is the only representative of 14 15 27 Psammophiinae in Madagascar, where the majority of species in this subfamily is distributed 16 17 18 28 throughout mainland Africa. Applying an integrative approach, where both morphological data 19 20 29 and genetic results from coalescent species delimitation analyses are considered, we determine 21 22 30 that Mimophis mahfalensis is composed of two distinct taxa: M. mahfahlensis in the central 23 24 25 31 montane and southern parts of Madagascar, and a second new species restricted to the north and 26 27 32 northwest, which we describe here. We also examine the colour pattern polymorphism exhibited 28 29 33 in Mimophis , which has been previously hypothesized as sexually dimorphic and/or 30 31 32 34 geographically correlated. However, we find all three colour morphs in both sexes, and both 33 34 35 species of Mimophis . Our work highlights the importance of phylogeographic studies that 35 36 37 36 examine wide ranging taxa to detect cryptic species diversity, even among species that are 38 39 37 common, or have been previously considered to be well known. 40 41 38 42 43 44 39 Key words : African snakes, Colubroid, Lamprophiidae, Malagasy biodiversity, Psammophiinae, 45 46 40 species delimitation 47 48 41 49 50 51 42 Correspondence to: Sara Ruane. E mail:[email protected] 52 53 43 54 55 56 57 58 59 60 2
URL: http://mc.manuscriptcentral.com/tsab Page 115 of 152 Systematics and Biodiversity
1 2 3 4 44 Introduction 5 6 45 Madagascar is known for having an abundance of biodiversity and endemism (Myers, 7 8 9 46 Mittermeier, Mittermeier, da Fonseca, & Kent, 2000), especially for squamate reptiles, with 10 11 47 >90% of the ~400 species endemic to the island (Uetz & Hosek, 2015). While recent studies 12 13 48 have shown that there are microendemic species restricted to very small ranges (Raxworthy et 14 15 16 49 al., 2008; Ruane, Burbrink, Randriamahtantsoa, & Raxworthy, 2016; Townsend, Vieites, Glaw, 17 18 50 & Vences, 2009), there are surprisingly few phylogeographic studies for wide ranging taxa 19 20 51 crossing multiple habitats. Of the existing phylogeographic studies on squamates of Madagascar, 21 22 23 52 most are on lizards (Boumans, Vieites, Glaw, & Vences, 2007; Chan, Choi, Raselimanana, 24 25 53 Rakotondravony, & Yoder, 2012; Florio, Ingram, Rakatondravony, Louis, & Raxworthy, 2012; 26 27 54 Florio & Raxworthy, 2016; Ratsoavina et al., 2010; Vences et al., 2004) with only two studies 28 29 30 55 examining snake phylogeography: the Malagasy boas (Orozco Terwengel, Nagy, Vieites, 31 32 56 Vences, & Louis, 2008) and the pseudoxyrhophiine Madagascarophis (Nagy, Glaw, Andreone, 33 34 35 57 Wink, & Vences, 2007). 36 37 58 The majority of Madagascar’s terrestrial snake fauna is represented by the 38 39 59 morphologically and ecologically diverse lamprophiid subfamily Pseudoxyrhophiinae (81 40 41 42 60 species, 82% of Malagasy snakes), as well as sanziniine boids (four species) and several lineages 43 44 61 of typhlopoids (~13 species total). Additionally, there is a second lamprophiid subfamily on 45 46 62 Madagascar represented by a single taxon, the monotypic Mimophis mahfalensis (Grandidier, 47 48 49 63 1867). Several studies have shown that this species is most closely related to mainland African 50 51 64 psammophiines (Bogert, 1940; de Haan, 2003; Kelly, Barker, Villet, Broadley, & Branch, 2008; 52 53 65 Nagy, Joger, Wink, Glaw, & Vences, 2003; Vidal et al., 2008; Vidal & Hedges, 2002). Mimophis 54 55 56 66 represents the most recent natural dispersal of snakes into Madagascar (<13 million years ago; 57 58 59 60 3
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 116 of 152
1 2 3 67 Samonds et al., 2013). The combination of the recent arrival and possible potential competitive 4 5 6 68 exclusion from the diverse pseudoxyrhophiines species may account for the lack of species 7 8 69 diversity within Mimophis . 9 10 11 70 Mimophis mahfalensis has one of the largest range sizes for snakes in Madagascar, being 12 13 71 absent from only the humid eastern coast (Glaw & Vences, 2007; Raxworthy, 2003). Mimophis 14 15 72 mahfalensis is primarily terrestrial and can be found at high densities (Schutt, 2008) across a 16 17 18 73 wide variety of habitats (Durkin, Steer, & Belle, 2011; Glaw & Vences, 2007; Ramanamanjato, 19 20 74 Mcintyre, & Nussbaum, 2002) especially in anthropogenically disturbed areas (D’Cruze, Green, 21 22 75 Robinson, & Gardner, 2006; Gardner & Jasper, 2009; McLellan, 2013; Ramanamanjato & 23 24 25 76 Lehtinen, 2006), and is frequently associated with grasslands from which most other Malagasy 26 27 77 snakes are absent (Bond, Silander, Ranaivonasy, & Ratsirarson, 2008). Mimophis mahfalensis is 28 29 78 opisthoglyphic and its bite is known to cause non lethal reactions in humans (Rosa et al., 2014). 30 31 32 79 The venom of M. mahfalensis is likely used to subdue lizard prey. 33 34 80 (Domergue, 1989). 35 36 37 81 While this snake is typically coloured grey and brown, M. mahfalensis exhibits intriguing 38 39 82 colour pattern polymorphisms. There are three readily identified dorsal patterns: 1) a mostly 40 41 83 patternless, uniform grey or tan, with finer, dark spotting or flecking on the dorsum, 2) a 42 43 44 84 background of grey or tan with a darker zig zag pattern down the mid dorsum, and 3) a dark 45 46 85 brown or black stripe down the mid dorsum, followed by a series of four contrasting lighter and 47 48 86 darker dorso lateral stripes before reaching the ventral scales (Fig. 1). It has been suggested these 49 50 51 87 colour differences are due to sexual dichromatism, with females having the patternless grey/tan 52 53 88 coloration, and males having the zig zag pattern (Domergue, 1969; Glaw & Vences, 1994, 2007) 54 55 89 but this has never been formally tested. Furthermore, whether there is a geographic component to 56 57 58 59 60 4
URL: http://mc.manuscriptcentral.com/tsab Page 117 of 152 Systematics and Biodiversity
1 2 3 90 these color patterns has not been studied in detail, although the stripe pattern is considered 4 5 6 91 typical for the central plateau area and the subspecies, M. mahfalensis madagascariensis (Glaw 7 8 92 & Vences, 1994; Gunther, 1868). Because there has been no detailed phylogeographic study 9 10 11 93 examining these snakes, it is also possible that these color patterns could correspond to 12 13 94 unrecognized species or lineages within Mimophis . 14 15 95 Although Mimophis mahfalensis is currently considered a single species, a previous 16 17 18 96 phylogenetic study, based on five samples, suggested that this taxon may be composed of more 19 20 97 than one undescribed taxon (Kelly et al., 2008). Here, we examine population structure and 21 22 98 morphology for M. mahfalensis across its complete range, and test for the presence of cryptic 23 24 25 99 species and the association with environment. We then examine colour pattern distributions to 26 27 100 determine if this coloration polymorphism is related to either sexual dichromatism or 28 29 101 geographically distinct lineages. 30 31 32 102 33 34 103 Materials and methods 35 36 37 104 Genetic samples 38 39 105 We sampled 39 individuals of Mimophis mahfalensis from across Madagascar (32 localities with 40 41 42 106 1−3 individuals per locality) and the outgroup Lamprophis guttatus for genetic data (sample 43 44 107 details in Appendix 1, see supplemental material online, which is available from the article’s 45 46 108 Taylor & Francis Online page at …). Tissue samples included liver or muscle tissue preserved in 47 48 49 109 EtOH and we extracted DNA following the Qiagen® DNEasy tissue protocol. We sequenced 50 51 110 four loci in the forward and reverse directions: the nuclear protein coding genes CMOS (509 bp), 52 53 111 RAG2 (645 bp), and the nuclear intron Nav 5 (512 bp), and the mitochondrial gene, COI (662 54 55 56 112 bp), using the same primers and conditions for PCR and sequencing conditions reported in 57 58 59 60 5
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 118 of 152
1 2 3 113 Ruane, Raxworthy, Lemmon, Lemmon, & Burbrink (2015). Contigs were assembled in 4 5 6 114 Geneious® using the De Novo Assembly option and sequences subsequently aligned using the 7 8 115 Geneious Alignment; sequences were trimmed so that the majority of individuals were the same 9 10 11 116 length for each locus. The substitution model for each locus was determined using the Bayesian 12 13 117 information criterion in jModelTest (Darriba, Taboada, Doallo, & Posada, 2012). We used the 14 15 118 program PHASE v.2.1 (Stephens, Smith, & Donnelly, 2001; Stephens & Donnelly, 2003) to 16 17 18 119 determine the allele phase of nuclear loci for heterozygous individuals. PHASE was run using 19 20 120 the default parameters for 100 iterations, with a thinning interval of 1, and burnin of 100; this 21 22 121 analysis was run three times on each locus to ensure consistency among runs. The outputs for the 23 24 25 122 most probable pair of resulting alleles for the heterozygous individuals were used in subsequent 26 27 123 analyses. 28 29 124 30 31 32 125 Population structure and species delimitation 33 34 126 We used the Bayesian inference program BEAST v.1.8.3 (Drummond & Rambaut, 2007) to infer 35 36 37 127 an initial concatenated tree of all loci, a nucDNA concatenated tree, and individual gene trees to 38 39 128 examine genetic structure among the Mimophis samples. This unphased dataset was partitioned 40 41 129 by locus and each used the most appropriate substitution model as determined using the Bayesian 42 43 44 130 information criterion in jModelTest. We constrained Lamprophis guttatus as the outgroup taxon, 45 46 131 used a Yule process speciation prior, and ran the analysis for 10 million generations, sampling 47 48 132 every 1000 generations. We used the program Tracer v.1.5 (Rambaut & Drummond, 2009) to 49 50 51 133 assess effective sample sizes of parameters, stationarity, and burnin for the BEAST analysis. 52 53 134 Using the phased dataset for Mimophis samples, we ran the population clustering 54 55 135 program Structure v.2.3.4 (Pritchard, Stephens, & Donnelly, 2000) to determine whether M. 56 57 58 59 60 6
URL: http://mc.manuscriptcentral.com/tsab Page 119 of 152 Systematics and Biodiversity
1 2 3 136 mahfalensis is a single, contiguous species or if there are multiple, discrete populations, possibly 4 5 6 137 representing cryptic taxa. Structure groups individuals into K populations by maximizing Hardy 7 8 138 Weinberg Equilibrium and linkage equilibrium. We generated 20 runs for each K value, ranging 9 10 11 139 from 1 to 5 populations for K, using the admixture model and correlated allele frequencies. Each 12 6 5 13 140 of these replicates were run for 10 generations, with a burnin of 5x10 generations. We 14 15 141 determined the optimal value of K through examination of the Ln P(D) plot of each K value, 16 17 18 142 which is the estimate of the posterior probability of the data for a K value and the K Evanno 19 20 143 method to determine the K value that captures the highest level of genetic structure (Evanno, 21 22 144 Regnaut, & Goudet, 2005) using the web interface CLUMPAK (Kopelman, Mayzel, Jakobsson, 23 24 25 145 Rosenberg, & Mayrose, 2015). 26 27 146 Neither BEAST nor Structure explicitly test for species limits, therefore we used the 28 29 147 results from the BEAST tree and the best estimated number of populations of Mimophis from the 30 31 32 148 Structure analyses (see results) to test for these lineage/populations as independent species using 33 34 149 the Bayesian delimitation program BBP v.3.1 (Yang & Rannala, 2010, 2014). This method uses 35 36 37 150 multiple loci and the multi species coalescent (Rannala & Yang, 2003; Yang, 2002) to determine 38 39 151 if user specified putative species are supported as distinct, giving results in the form of posterior 40 41 152 probability distributions (Pp) for either retaining or collapsing nodes between the putative 42 43 44 153 species. We tested species delimitations using the both the mtDNA and nucDNA datasets as well 45 46 154 as the nucDNA dataset alone, all under the following conditions. We used the reversible jump 47 48 155 Markov chain Monte Carlo (rjMCMC) algorithm 0 and adjusted the fine tuning parameters 49 50 51 156 between 0.30 and 0.70 for each parameter to allow mixing of the rjMCMC among the possible 52 53 157 species delimitation models. Each BPP analysis was run for 100,000 generations, sampled every 54 55 158 2 generations, and with a 20,000 generation burnin. We varied the parameterizations to account 56 57 58 59 60 7
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 120 of 152
1 2 3 159 for both large and small population sizes combined with both deep and shallow divergence 4 5 6 160 times, following the parameterizations of Ruane et al. (2016). Each parameterization was run 7 8 161 three times with different starting seeds to ensure consistency among results. 9 10 11 162 12 13 163 Isolation by distance or environment 14 15 164 To test if geographic distance or variation in climate across Madagascar is important for 16 17 18 165 structuring genetic variation across populations, we used redundancy analyses (RDA), which is 19 20 166 a constrained ordination method (Legendre & Fortin, 2010) performed in the statistical software 21 22 167 program R v.3.4.1 (R Core Team, 2013), using the scripts included in Myers, Hickerson, & 23 24 25 168 Burbrink (2016). Current climatic variables interpolated at 2.5 arc min resolution were obtained 26 27 169 from the WorldClim database, which are 19 average monthly climatic data for minimum, mean, 28 29 170 and maximum temperature and precipitation between the period of 1960–1990 (Hijmans, 30 31 32 171 Cameron, Parra, Jones, & Jarvis, 2005), and extracted from the collection localities of 33 34 172 individuals with genetic data. We exclude Isothermality (BIO3) and Temperature Annual Range 35 36 37 173 (BIO7) because they are derived from other variable already included within the WorldClim 38 39 174 dataset. A genetic distance matrix was generated from all genetic sequence data for both the 40 41 175 concatenated dataset (39 individuals total) and the COI dataset (37 individuals total) using the 42 43 44 176 dist.dna function of the R package ape (Paradis, Claude, & Strimmer, 2004) with the GG95 45 46 177 model of sequence evolution (Galtier & Gouy, 1995 ) and used as a response variable, where 47 48 178 geographic distance between sampled localities (32 unique localities across all individuals) and 49 50 51 179 current climate were the explanatory variables. Geographic distance and climate may be 52 53 180 confounded by one another, therefore analyses incorporating only one of these variables were 54 55 181 conditioned on the other (using the Condition option within the rda function of vegan ; (Oksanen 56 57 58 59 60 8
URL: http://mc.manuscriptcentral.com/tsab Page 121 of 152 Systematics and Biodiversity
1 2 3 182 et al., 2007). We ran three different analyses with these variables, where predictions of genetic 4 5 6 183 structure were tested using distance, climate, and a model that included both distance and 7 8 184 climate. RDA analyses return an r2 value and an ANOVA can be used to assess significance. 9 10 11 185 12 13 186 Morphology and colour patterns 14 15 187 Georeferenced records for a total of 239 Mimophis were used from the herpetology collection at 16 17 18 188 the American Museum of Natural History (AMNH), the University of Michigan, Museum of 19 20 189 Zoology (UMMZ), and the University of Antananarivo (UADBA). Of these specimens, nine did 21 22 190 not have morphological vouchers available for examination but had tissues included in the 23 24 25 191 genetic analyses above, and two of these were also coded for colour pattern based on field 26 27 192 photos. 28 29 193 We took morphological data from 30 Mimophis specimens that had corresponding 30 31 32 194 genetic data and examined an additional 51 specimens distributed across the entire range of 33 34 195 Mimophis for the following morphological characters (to the nearest mm where applicable): total 35 36 37 196 length, snout vent length (SVL), tail length, ventral scale count, and subcaudal scale count. We 38 39 197 also determined the relative tail length of individuals using the ratio of tail length and SVL. 40 41 198 Using sexed adults (n = 57 adult individuals total which included complete 42 43 44 199 morphological data, see below on determining sex), we examined differences in morphology 45 46 200 between resulting Mimophis lineages. First, using R (R Core Team, 2013), we tested if resulting 47 48 201 populations differed significantly for ventral and subcaudal counts using Welch’s t test, which 49 50 51 202 provides a better estimate for unequal sample sizes or variance than does a standard t test 52 53 203 (Ruxton, 2006; Welch, 1947). Subcaudal scale numbers are often different between sexes in 54 55 204 snakes, we therefore partitioned the populations by sex, to determine if sexes could be pooled 56 57 58 59 60 9
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 122 of 152
1 2 3 205 within lineages. To examine relative tail length differences between populations, we first tested 4 5 6 206 for allometric growth for both males and females for each lineage to determine that we are only 7 8 207 comparing adults. We used the R package smatr 3 (Warton, Duursma, Falster, & Taskinen, 9 10 11 208 2012), which specifically tests for allometry using power functions. We used the sma function to 12 13 209 determine if measured samples show isometric growth between tail length (TL) and snout vent 14 15 210 length (SVL) using the power function log = log where a is a proportionality 16 17 18 211 coefficient and b is a scaling exponent, and if significantly different from 1.0, would suggest that 19 20 212 allometric growth is present in our samples; TL and SVL are log transformed prior to the 21 22 213 analysis (see Warton et al., 2012 for details). We then used the lm function in R to run a multiple 23 24 25 214 linear regression to determine the significant predictors of tail length (TL) to specifically 26 27 215 determine if resulting lineages (specified here as species for the analysis) differed in tail lengths. 28 29 30 216 We compared models that included the variables SVL, sex, species, the interaction of sex and 31 32 217 species, the interaction of SVL and species, the interaction of SVL and sex, and the interaction of 33 34 218 SVL, sex, and species as predictors of the response variable TL. Both TL and SVL were log 35 36 37 219 transformed to normalize the residuals of the model and improve the fit. Automated model 38 39 220 selection was conducted using the dredge function of the MuMIn package v. 1.15.6 (Barton, 40 41 221 2016). Models were ranked using the Akaike Information Criterion with correction for small 42 43 44 222 sample sizes (AICc). We used the anova function in R to determine if high ranking models were 45 46 223 significantly different from each other and from the complete model including all predictors and 47 48 224 possible interactions. 49 50 51 225 To score colour patterns, we examined all available specimens or field photos for colour pattern 52 53 226 (n = 232). We scored the three readily identified dorsal patterns (Fig. 1): 1) uniform, 2) zig zag, 54 55 56 227 and 3) striped. Snakes that were unclear with respect to pattern due to poor preservation or 57 58 59 60 10
URL: http://mc.manuscriptcentral.com/tsab Page 123 of 152 Systematics and Biodiversity
1 2 3 228 sloughing/loss of skin while in preservation were excluded from scoring. Using a Chi square test 4 5 6 229 of independence in R we determined whether these colour patterns were distributed equally 7 8 230 within resulting Mimophis lineages (using only snakes that could be confidently assigned a 9 10 11 231 pattern and to a lineage). To determine if any colour pattern was restricted to a single sex, we 12 13 232 sexed a sub sample of specimens assigned to each colour morph (17 uniform, 37 striped, and 50 14 15 233 zig zag individuals), based on dissection and examining gonads or the presence of everted 16 17 18 234 hemipenes. All specimen information is provided in Appendix 1 (see supplemental material 19 20 235 online). 21 22 236 23 24 25 237 Results 26 27 238 Genetic structure and species delimitation 28 29 30 239 Lamprophis guttatus and all of the Mimophis individuals were successfully sequenced for at least 31 32 240 two of the four loci, with most sequenced for all four loci (see Appendix 1; GenBank, see 33 34 35 241 supplemental material online). JModelTest indicated that the the best fit substitution models to 36 37 242 our data were: CMOS=HKY; RAG2=F81; Nav intron 5=HKY; COI=HKY+Gamma. 38 39 243 All trees generated in BEAST had high ESS values for all parameters (>200) and after 40 41 42 244 assessment of stationarity in Tracer, 25% of samples were discarded as burnin. The BEAST 43 44 245 concatenated tree (Fig. 2) indicated two main clades of Mimophis , one with a northern and 45 46 246 northwestern distribution (n = 15) and one with a central and southern distribution (n = 24). We 47 48 49 247 compared the four individual nuclear gene trees to the concatenated tree (additional trees are 50 51 248 included as supplementary materials) and found no gene tree was identical to another but there 52 53 249 was overall concordance among trees. Those trees from RAG2 and CMOS were largely 54 55 56 250 congruent, with individuals falling into the northern or southern clades as they did in the 57 58 59 60 11
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 124 of 152
1 2 3 251 concatenated tree, with only one (CMOS) or two (RAG2) individuals differing in their 4 5 6 252 placements (but not the same individuals). The intron Nav 5 has less similarity in structure to the 7 8 253 concatenated tree, with individuals from the northern and southern lineages mixed, but still 9 10 11 254 showed some of the same overall signal/structure, e.g., a clade composed only of southern 12 13 255 individuals. The COI gene tree found all individuals were in the same two clades (i.e., either 14 15 256 northern or southern) as in the concatenated tree. The concatenated nuclear tree results mirror the 16 17 18 257 concatenated mtDNA+nucDNA tree, adding additional support that the mtDNA is not solely 19 20 258 driving the phylogenetic results (but see discussion on UMMZ 237402 below). Minimum and 21 22 259 maximum uncorrected pairwise genetic divergences between the northern and southern clades 23 24 25 260 were: COI= 4.1−8.0%; CMOS = 0.2−0.6%; RAG2 = 0.0−0.4%; NAV intron 5 = 0.0−0.9%. 26 27 261 Uncorrected genetic distances (pairwise identities) were on average greater between the northern 28 29 262 and southern lineages than within them for all loci. For example, for COI the mean pairwise 30 31 32 263 difference between the northern and southern clades was 6.3%, and the mean pairwise 33 34 264 differences within the northern and southern clades were 0.7% and 3.3%, respectively. The 35 36 37 265 within clade diversity of the southern lineage did have pairwise differences that were higher 38 39 266 within the clade rather than between the clades. These elevated within clade divergences were 40 41 267 seemingly due to individuals with unshared substitutions and the high amounts of substructure 42 43 44 268 within the southern clade (Fig. 2). There may also be even greater (species level) substructure 45 46 269 within the southern lineage that additional sampling would make clearer. 47 48 270 Structure results (see supplemental material online) indicated that the Ln P(D) of populations 49 50 51 271 increased substantially from K=1 to K=2, with diminishing increases combined with high 52 53 272 variance at higher K values; similarly, among the assessment of population changes (∆K), K=2 54 55 273 was the optimal value of K using the Evanno method. An examination of the Structure outputs 56 57 58 59 60 12
URL: http://mc.manuscriptcentral.com/tsab Page 125 of 152 Systematics and Biodiversity
1 2 3 274 for each run and K value shows that individuals assigned to the K=2 Structure populations had 4 5 6 275 the highest assignment probabilities, with 38 individuals assigned at >95%, and one individual 7 8 276 assigned at >70% ; these assignments generally corresponded to the same northern and southern 9 10 11 277 populations of Mimophis inferred by the BEAST tree. As the value of K increased beyond two, 12 13 278 individuals from the southern population were assigned to additional populations with no 14 15 279 obvious geographic pattern to the assignments, while those assigned to the northern population 16 17 18 280 typically remained within the same population for any value of K. However, UMMZ 237402 is 19 20 281 strongly supported as sister to the rest of the northern clade in the concatenated BEAST tree (Fig. 21 22 282 2), and is also suggested independently when examining morphological data (Appendix 1), but is 23 24 25 283 part of the southern population in the Structure results. This individual was successfully 26 27 284 sequenced for only two loci (COI, CMOS); only CMOS provided any information for the 28 29 285 Structure analyses as this individual has a private COI haplotype not shared with any other 30 31 32 286 Mimophis sequenced here. As the inference of the tree in BEAST makes use of the sequence data 33 34 287 from all available loci, combined with the independent morphological evidence that suggests it is 35 36 37 288 part of the northern clade (Appendix 1, see supplemental material online), we considered it part 38 39 289 of the northern lineage here for species delimitation purposes. Based on the two population 40 41 290 assignments from Structure but including UMMZ 237402 as part of the northern clade, BPP 42 43 44 291 analyses for each run, using all datasets and under all parameterizations, resulted in two well 45 46 292 supported species of Mimophis (Pp=100% in all runs for all parameterizations). Results from the 47 48 293 RDA analyses demonstrate that neither geographic distance (COI: adjusted r2 = 0.02, p value = 49 50 51 294 0.49; concatenated: adjusted r2 = 0.02, p value = 0.2), current climate (COI: adjusted r2 = 0.02, 52 53 295 p value = 0.53; concatenated: adjusted r2 = 0.07, p value = 0.15), or a combination of these two 54 55 56 57 58 59 60 13
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 126 of 152
1 2 3 296 variables (COI: adjusted r2 = 0.08, p value = 0.76; concatenated: adjusted r2 = 0.07, p value = 4 5 6 297 0.11) significantly explains genetic variation across the distribution of Mimophis . 7 8 298 9 10 11 299 Morphological differences in Mimophis 12 13 300 Using all available data (genetic, morphological and/or geographical) we were able to categorize 14 15 301 a total 239 Mimophis records as being members of a northern lineage (n = 46), southern lineage 16 17 18 302 (n = 187), or an intermediate between the two (n = 6); (Appendix 1, see supplemental material 19 20 303 online). These assignments were first made from the results of the genetic analyses, giving us an 21 22 304 estimate of the distributions for each putative species. Using individuals that had both genetic 23 24 25 305 and morphological data, we subsequently identified morphological differences (described below) 26 27 306 to diagnose specimens into the same groups as the genetic data. We then used this morphological 28 29 307 diagnosis to assign specimens lacking genetic data into the northern or southern groups (n = 75 30 31 32 308 individuals total). 33 34 309 We found consistent differences in the eye stripe coloration feature between these two 35 36 37 310 groups (see Table 1 and the new species description for details). We also found a significant 38 39 311 difference for ventral counts between the northern and southern populations (Welch’s t test, 40 41 312 North vs. South ventrals, t = 2.63, df = 25.45, P = 0.015), with northern individuals having more 42 43 44 313 ventrals on average compared to southern individuals (mean = 161, range = 152–170 North; 45 46 314 mean = 157, range = 144–167 South). For subcaudal scale counts, we found that the northern and 47 48 315 southern populations were significantly different in subcaudal number (Welch’s t test, North vs. 49 50 51 316 South subcaudals, t = 7.5, df = 26.04, P < 0.001), with the northern lineage having higher 52 53 317 numbers of subcaudal scales (mean = 94, range = 79–104 North; mean = 76, range = 44–88 54 55 318 South). We also tested if subcaudal counts might be sexually dimorphic, and found no 56 57 58 59 60 14
URL: http://mc.manuscriptcentral.com/tsab Page 127 of 152 Systematics and Biodiversity
1 2 3 319 differences between males and females in each population (Welch’s t test, North males vs. 4 5 6 320 females subcaudals, t = 0.44, df = 22.24, P = 0.666; South males vs. females subcaudals, t = 7 8 321 0.79, df = 11.08, P = 0.444). 9 10 11 322 We found no significant allometric effect with respect to tail length for either sex in the 12 13 323 northern or southern lineages ( P = 0.085 0.328 for all sexes from all lineages; outputs included 14 15 324 as supplementary material), confirming that we are comparing only adult samples. For the 16 17 18 325 multiple regression analysis of tail length, the three most likely models (respective Akaike 19 20 326 weights of 0.435, 0.176, 0.149) all included SVL and species (the northern and southern 21 22 327 lineages) as predictors, with the best scoring model including just SVL and species (rank 1). The 23 24 25 328 interaction between SVL and species is included in the second model (rank 2) and sex is also 26 27 329 included in the third model (rank 3). The analysis of the three most likely linear models shows 28 29 330 that 1) log(TL) increases significantly with log(SVL) in all models ( P > 0.0001 in all models), 2) 30 31 32 331 log(TL) is significantly larger in the northern group than in the southern group in all models ( P > 33 34 332 0.0001 in all models), and 3) when included in the model, the interaction between SVL and 35 36 37 333 species is not a significant predictor of tail length ( P = 0.4573) and that sex is also not a 38 39 334 significant predictor of tail length ( P = 0.6212). These results indicate that there is no sexual 40 41 335 dimorphism in relative tail length, but there is a significant difference between the northern and 42 43 44 336 southern groups. Relative tail length is greater in the northern population compared to the 45 46 337 southern (tail length/total length ratio: mean = 0.37, range = 0.33–0.43 North; mean = 0.30, 47 48 338 range = 0.20–0.36, South). 49 50 51 339 For colour pattern on the body, we found that specimens from the putative ranges of both 52 53 340 the northern and southern populations included all three patterns (Fig. 3). We analysed these on 54 55 341 specimens that could only be confidently assigned to the northern or southern Mimophis clade (n 56 57 58 59 60 15
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 128 of 152
1 2 3 342 = 223). The three colour patterns were not distributed equally within the northern and southern 4 5 2 6 343 lineages ( X test of independence = 35.9; P <0.001; df = 2); there were fewer zig zag patterned 7 8 344 snakes within the range of the northern population compared to the southern population, with 9 10 11 345 none observed in the extreme north of Madagascar (Fig. 3). Striped snakes are absent in the 12 13 346 southern and southwestern coastal regions, and the central High Plateau lacked zig zag patterned 14 15 347 snakes (Fig. 3). Mimophis is likely not sexually dichromatic; both sexes were found with all 16 17 18 348 three pattern types (see Table 1 for a summary of morphological data, additional details are 19 20 349 available in Appendix 1, see supplemental material online). 21 22 350 23 24 25 351 Taxonomy of Mimophis 26 27 352 Based on our genetic analyses and the corresponding morphological data, we find strong support 28 29 353 for the presence of two distinct species within Mimophis mahfalensis . The original description of 30 31 32 354 Mimophis mahfalensis is based on specimens collected by Grandidier (1867) from the southwest 33 34 355 coast of Madagascar. Grandidier gives two localities: ‘Machikova’, which likely corresponds to 35 36 37 356 the present day Masikoro Port, now known as Androka at 25.029122 S, 44.074371 E, and 38 39 357 Anhoulabé, which is likely present day Ambohibe, a coastal town at 21.353016 S, 43.510772 E. 40 41 358 Both these localities are located within the range of the southern species found in our analyses. 42 43 44 359 Thus this southern species can be confidently assigned to Mimophis mahfalensis . 45 46 360 The southern species (Mimophis mahfalensis ) also includes other previously described 47 48 361 species and subspecies of Mimophis , collected in the central High Plateau region of Madagascar. 49 50 51 362 The putative subspecies M. mahfalensis madagascariensis Gunther 1868, which was given 52 53 363 subspecific status from junior synonymy of M. mahfalensis by Glaw and Vences (1994), was 54 55 364 described from type specimens collected by the Rev. W. Ellis, a missionary who worked in the 56 57 58 59 60 16
URL: http://mc.manuscriptcentral.com/tsab Page 129 of 152 Systematics and Biodiversity
1 2 3 365 Central High Plateau of Madagascar. Glaw and Vences (1994) recognized this subspecies based 4 5 6 366 on the lined colouration pattern type given in the original description, which they thought was 7 8 367 typical for Mimophis from Central Madagascar. Domergue (1969) also considered that these 9 10 11 368 Central High Plateau Mimophis deserved subspecific status based on their lined coloration, and 12 13 369 even assigned a name: M. mahfalensis lineatus . However, no type specimens were designated, 14 15 370 and thus this taxon is not considered valid (Brygoo, 1982). 16 17 18 371 Jourdran (1903) also reported a new variety of Mimophis with an unusual white head that 19 20 372 he referred to as Mimophis albiceps in the text, and Mimophis mahafalensis albiceps in the 21 22 373 associated illustration. However, no type specimen or locality was designated with this 23 24 25 374 description, and this taxon has subsequently been ignored (Guibé, 1958) or else considered a 26 27 375 variety of Mimophis mahfalensis (Brygoo, 1982). We have never seen white headed Mimophis in 28 29 376 the field or collections, and strongly suspect that this coloration feature is an artifact resulting 30 31 32 377 from fluid preservation. In alcohol, Mimophis specimens sometimes shed epidermal skin on 33 34 378 scales, revealing a much lighter colouration below. Thus, this white headed specimen may have 35 36 37 379 acquired this colouration simply by losing surface skin on its head, after it was preserved. 38 39 380 We therefore conclude that the northern species, as identified by our results, represents an 40 41 381 unnamed taxon, which we formally describe below. 42 43 44 382 45 46 383 Mimophis occultus sp. nov. 47 48 384 Figs. 1, 2, 3, 4; Table 1; Appendix 1(see supplemental material online) 49 50 51 385 52 53 386 54 55 56 57 58 59 60 17
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 130 of 152
1 2 3 387 HOLOTYPE: UMMZ 237408 (RAN 68828 field number), adult male, Madagascar, Mahajanga 4 5 6 388 Province, Tsiambara Forest ( 15.920716 S, 45.9886 E), 41 m elevation, on 20 April 2002, 7 8 389 collected by J. Spannring, J. Rafanomazantsoa, and M. Rakotoarivelo 9 10 11 390 12 13 391 PARATYPES: UMMZ 209622 (RAN 38599 field number), juvenile male, Madagascar, 14 15 392 Antsiranana Province, Montagne des Francais ( 12.3000 S, 49.3167 E), 150 m elevation, on 31 16 17 18 393 December 1991, collected by C. J. Raxworthy, A. Raselimanana, and J.B. Ramanamanjato; 19 20 394 UMMZ 209625 (RAN 38969 field number), juvenile female, Madagascar, Antsiranana Province, 21 22 395 Ankarana Reserve, ( 12.9250 S, 49.1000 E), 119 m elevation, on 3 February 1992, collected by 23 24 25 396 C.J. Raxworthy and A. Raselimanana; AMNH 176427 (RAX 12588 field number), adult male, 26 27 397 from Madagascar, Antsiranana Province, Ankarana Reserve ( 12.9311 S, 49.0558 E) on 7 28 29 398 February 2014, collected by B. Randriamahtantsoa, C.J. Raxworthy, and S. Ruane. 30 31 32 399 33 34 400 NONTYPE MATERIAL EXAMINED: AMNH 140240 from Madagascar, Antsiranana Province 35 36 37 401 ( 12.3736 S, 49.3669 E); UMMZ 209624 (Field Number RAN 38846) from Madagascar, 38 39 402 Antsiranana Province ( 12.3000 S, 49.3167 E); UMMZ 209623 (Field Number RAN 38845) 40 41 403 from Madagascar, Antsiranana Province ( 12.3000 S, 49.3167 E); AMNH 153327 (Field 42 43 44 404 Number RAX 2250) from Madagascar, Antsiranana Province ( 13.7884 S, 48.8 E); AMNH 45 46 405 153328 (Field Number RAX 4797) from Madagascar, Antsiranana Province ( 13.3727 S, 47 48 406 50.0029 E); UMMZ 240452 (Field Number RAN 70118) from Madagascar, Antsiranana 49 50 51 407 Province ( 12.9707 S, 48.753 E); UMMZ 209679 (Field Number RAN 44306) from Madagascar, 52 53 408 Antsiranana Province ( 13.7833 S, 48.3333 E); AMNH 176426 (Field Number RAX 12469) 54 55 409 from Madagascar, Antsiranana Province ( 12.1907 S, 49.2197 E); UMMZ 219470 (Field 56 57 58 59 60 18
URL: http://mc.manuscriptcentral.com/tsab Page 131 of 152 Systematics and Biodiversity
1 2 3 410 Number RAN 54964) from Madagascar, Antsiranana Province ( 12.8333 S, 49.5 E); UMMZ 4 5 6 411 219469 (Field Number RAN 54757) from Madagascar, Antsiranana Province ( 12.8333 S, 49.5 7 8 412 E); UMMZ 218448 (Field Number RAN 54630) from Madagascar, Mahajanga Province ( 9 10 11 413 17.4908 S, 45.7464 E); UMMZ 218449 (Field Number RAN 54633) from Madagascar, 12 13 414 Mahajanga Province ( 17.4360 S, 45.7531 E); UMMZ 237398 (Field Number RAN 67632) from 14 15 415 Madagascar, Mahajanga Province ( 17.8944 S, 44.0168 E); UMMZ 237407 (Field Number RAN 16 17 18 416 68393) from Madagascar, Mahajanga Province ( 15.9325 S, 45.7716 E); UMMZ 237403 (Field 19 20 417 Number RAN 68122) from Madagascar, Mahajanga Province ( 16.0212 S, 45.6029 E); UMMZ 21 22 418 222408 (Field Number RAX 0285) from Madagascar, Mahajanga Province ( 16.0176 S, 45.2718 23 24 25 419 E); UMMZ 222407 (Field Number RAX 0215) from Madagascar, Mahajanga Province ( 26 27 420 16.4698 S, 45.3484 E); UMMZ 237400 (Field Number RAN 67719) from Madagascar, 28 29 421 Mahajanga Province ( 17.8944 S, 44.0168 E); UMMZ 222406 (Field Number RAX 0003) from 30 31 32 422 Madagascar, Mahajanga Province ( 16.4698 S, 45.3484 E); UMMZ 218446 (Field Number RAN 33 34 423 51981) from Madagascar, Mahajanga Province ( 14.8508 S, 48.2264 E); AMNH pending (Field 35 36 37 424 Number RAX 14073) from Madagascar, Mahajanga Province ( 15.4924 S, 46.6958 E); UMMZ 38 39 425 218447 (Field Number RAN 54286) from Madagascar, Mahajanga Province ( 18.7080 S, 40 41 426 44.7164 E); UMMZ 219471 (Field Number RAN 55582) from Madagascar, Mahajanga Province 42 43 44 427 ( 16.0430 S, 45.2597 E); UMMZ 237405 (Field Number RAN 68279) from Madagascar, 45 46 428 Mahajanga Province ( 16.0212 S, 45.6029 E); UMMZ 237406 (Field Number RAN 68372) from 47 48 429 Madagascar, Mahajanga Province ( 15.9325 S, 45.7716 E); UMMZ 237401 (Field Number RAN 49 50 51 430 67844) from Madagascar, Mahajanga Province ( 16.7655 S, 44.3805 E); UMMZ 237404 (Field 52 53 431 Number RAN 68124) from Madagascar, Mahajanga Province ( 16.0205 S, 45.6036 E); UMMZ 54 55 432 237402 (Field Number RAN 68052) from Madagascar, Mahajanga Province ( 16.1085 S, 56 57 58 59 60 19
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 132 of 152
1 2 3 433 45.0458 E); UMMZ 237399 (Field Number RAN 67718) from Madagascar, Mahajanga Province 4 5 6 434 ( 17.8944 S, 44.0168 E); UMMZ 240451 (Field Number RAN 69002) from Madagascar, 7 8 435 Mahajanga Province ( 13.0561 S, 48.9814 E); UMMZ 240450 (Field Number RAN 68934) from 9 10 11 436 Madagascar, Mahajanga Province ( 16.3584 S, 46.8259 E); AMNH 160099 (Field Number RAX 12 13 437 9641) from Madagascar, Mahajanga Province ( 16.6298 S, 47.4161 E); UMMZ 218426 (Field 14 15 438 Number RAN 49464) from Madagascar, Mahajanga Province ( 16.3136 S, 46.8173 E); UMMZ 16 17 18 439 218425 (Field Number RAN 49449) from Madagascar, Mahajanga Province ( 16.3136 S, 19 20 440 46.8173 E); UMMZ 237391 (Field Number RAN 66985) from Madagascar, Mahajanga Province 21 22 441 ( 18.8725 S, 44.2399 E); UADBA pending (Field Number RAX 12471) from Madagascar, 23 24 25 442 Antsirinana Province ( 12.1907 S, 49.2697 E); UADBA pending (Field Number RAX 12408) 26 27 443 from Madagascar, Antsirinana Province ( 12.3342 S, 49.3581 E); UADBA pending (Field 28 29 444 Number RAX 12405) from Madagascar, Antsirinana Province ( 12.3342 S, 49.3581 E); UADBA 30 31 32 445 pending (Field Number RAX 11384) from Madagascar, Mahajanga Province ( 16.0750 S, 33 34 446 46.7312 E); see Appendix 1 (see supplemental material online). 35 36 37 447 38 39 448 ETYMOLOGY: The species epithet occultus is Latin for hidden or concealed, referring to the 40 41 449 cryptic status of this taxon and having been hidden in plain sight. 42 43 44 450 45 46 451 DIAGNOSIS: A species of Mimophis that can be diagnosed from M. mahfalensis by a 47 48 452 combination of the following characteristics (see also Table 1): a narrow dark post ocular eye 49 50 51 453 stripe ≤2 temporal scales in width (vs. in M. mahfalensis post ocular eye stripe >2 temporal 52 53 454 scales in width, or indistinct and fused with the dark dorsal head coloration); 79−104 subcaudal 54 55 56 57 58 59 60 20
URL: http://mc.manuscriptcentral.com/tsab Page 133 of 152 Systematics and Biodiversity
1 2 3 455 scales (vs. in M. mahfalensis 44−88 subcaudal scales); and a generally larger tail length/SVL 4 5 6 456 ratio of 0.33−0.43 (vs. in M. mahfalensis 0.20−0.36). 7 8 457 9 10 11 458 DESCRIPTION OF THE HOLOTYPE: Adult male in good state of preservation, tail complete, 12 13 459 removal of dorso lateral musculature from the posterior right side for DNA tissue sample, 14 15 460 hemipenes not everted. Scales smooth, anal plate divided. Total length 615 mm, tail length 172 16 17 18 461 mm. Ventrals 160, subcaudals 92, divided. Head length (as jaw length) 17.7 mm, head width 19 20 462 (widest point) 9.1 mm. Eyes large, 3.0 mm horizontal diameter, pupil round. On both left and 21 22 463 right, supralabials 8, numbers 4 and 5 in contact with the eye. Infralabials 10, first pair in contact 23 24 25 464 behind mental, infralabials 1 5 in contact with inframaxillaries. Rostral broader than high, 2.6 26 27 465 mm wide/1.5 mm high, visible from above. Nasal entire in contact with 1 st and 2 nd supralabials. 28 29 466 Single loreal, in contact with nasal, internasal, prefrontal, and supralabial 2. Circumoculars 7, 1 30 31 32 467 supraocular, 2 preoculars, 2 suboculars, and 2 postoculars; the preocular is divided so that a 33 34 468 lower triangular fragment is still in contact with the eye, otherwise the circumocular count would 35 36 37 469 be 6, with a single preocular. Temporals 2+3/1+3. Dorsal surface of head includes a pair of 38 39 470 internasals (width 1.5 mm/length of suture 1.5 mm), pair of prefrontals (width 2.2/length of 40 41 471 suture 1.9 mm), a pair of supraoculars (width 2.2 mm/length 4.1 mm), a frontal longer than wide 42 43 44 472 (length 5.1 mm/anterior width 2.2 mm), and a pair of parietals (length of suture 3.6 mm). Two 45 46 473 pairs of inframaxillaries separated by a mental groove (anterior inframaxillary length 4.0 47 48 474 mm/posterior inframaxillary length 4.2 mm), with the posterior pair separated posteriorly by a 49 50 th th 51 475 gular scale. Dorsal scale rows 17 17 13 at 10 ventral from anterior, midbody, and 10 ventral 52 53 476 anterior to the cloaca. 54 55 477 56 57 58 59 60 21
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 134 of 152
1 2 3 478 COLOURATION: The overall coloration faded in preservation, with the loss of the top layer on 4 5 6 479 some scales. However the pattern remains clear and is of the “striped” variety. The dorsum of the 7 8 480 head is a tannish ground color with darker longitudinal markings on the internasals, prefrontals, 9 10 11 481 supraoculars, frontals, and parietals. Dark brown marks on the nasal, loreal, preocular and 12 13 482 postocular scales forms a dark stripe passing through the eye and fading posteriorly on the 14 15 483 temporal scales. This dark post ocular eye stripe is narrow, ≤ 2 temporal scales in width. Dark 16 17 18 484 brown blotches are also present on all labial scales, with some blotches centered with pale brown 19 20 485 spots. The ventral surface of the head is of a light tan/cream color and includes dark brown 21 22 486 speckles or marks on the majority of gular scales and the inframaxillaries. The dorsal pattern 23 24 25 487 consists of brown longitudinal stripes on a yellowish tan ground color with a broad mid dorsal 26 27 488 stripe and then 4 thinner stripes on each side of the body moving ventrally. The dark brown 28 29 489 central stripe starts anteriorly as two blotches on the internasal scales of the head, separated by 30 31 32 490 white on middle scales, and merging into one solid stripe approaching the neck and running 33 34 491 down the dorsum on scale rows 8 (partially), 9, and 10; the stripe becomes broken and less 35 36 37 492 distinct approaching the posterior of the tail. This central stripe includes a white center that is 38 39 493 only on the interior of the middle dorsal scale row (9), beginning as a broken white line and 40 41 494 becoming a solid line approximately 1/6th of the way down the body, and becoming less distinct 42 43 44 495 and fading entirely on the tail. On scale rows 5 (partially), 6, 7 and 8 (partially) and counting 45 46 496 from the ventral scale up towards the dorsum mid body, the color is yellowish tan, with 47 48 497 occasional dark brown pigment scattered on these scales. The single apical pit, encompassed in a 49 50 51 498 dot of dark brown pigment is especially prominent on the scales in these rows. Scale rows 1, 2, 52 53 499 and 3 each has a very thin central dark stripe, lighter on the anterior 1/3 of the body but 54 55 500 becoming prominent on the posterior 2/3 of the body. There is lighter brown pigment on the 56 57 58 59 60 22
URL: http://mc.manuscriptcentral.com/tsab Page 135 of 152 Systematics and Biodiversity
1 2 3 501 upper half of scale row 1 and the lower half of scale row 2 forming a faded light brown line. 4 5 6 502 There is a thin stripe of light brown pigment on scale row 5, broken (or blotchy) anteriorly, but 7 8 503 becoming prominent at mid body where it drops to scale row 4 and forms a distinct stripe 9 10 11 504 posteriorly to the level of the anal scale, before breaking up again on the tail. The venter is cream 12 13 505 colored, especially anteriorly, but obscured posteriorly by profuse stippling of darker pigment, 14 15 506 with pairs of thin dark brown marks on each ventral scale. 16 17 18 507 19 20 508 VARIATION: Paratypes similar to the holotype except as follows: UMMZ 209622, juvenile 21 22 509 male, 477 mm total length, 135 mm tail length, ventrals 157, subcaudals 99, number of 23 24 25 510 precoculars 1, temporals 2+3 on both sides of the head; UMMZ 209625, juvenile female, 551 26 27 511 mm total length, 166 mm tail length, ventrals 162, subcaudals 86 (tail missing terminal tip), nasal 28 29 512 divided below nostril, number of precoculars 1, temporals 2+3 on both sides of the head (lower 30 31 32 513 anterior temporal not in contact with postocular); AMNH 176427, adult male, 705 mm total 33 34 514 length, 180 mm tail length, ventrals 164, subcaudals 99, nasal divided below nostril, number of 35 36 37 515 precoculars 1, temporals 2+3 on both sides of the head, body coloration of the patternless type. 38 39 516 See Table 1 for a summary of variation and comparison with M. mahfalensis . 40 41 517 42 43 44 518 DISTRIBUTION AND NATURAL HISTORY: A species of Mimophis found from the northern 45 46 519 tip of Madagascar, throughout central and western Antsiranana Province, and as far south as the 47 48 520 towns of Maintirano, Ambatomainty, and Tsaratanana in Mahajanga Province. The maximum 49 50 51 521 known elevation recorded for this species is 700 m (UMMZ 218449, RAN 54633). There is a 52 53 522 narrow zone between Maintirano, Belo Tsiribihina and Tsiroanomandidy where several snakes 54 55 523 have intermediate morphological characters (see Discussion), and where M. occultus and M. 56 57 58 59 60 23
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 136 of 152
1 2 3 524 mahfalensis may co occur (e.g., Cap Kimby). Because we do not have genetic data for these 4 5 6 525 morphological intermediates, we cannot yet be sure of their taxonomic status. 7 8 526 Based on our field observations, we conclude that M. occultus occurs in similar habitats 9 10 11 527 and has similar behaviors as compared to M. mahfalensis . However, M. mahfalensis is 12 13 528 distributed from sea level to much higher elevations than M. occultus, reaching as high as 1680 14 15 529 m (Ibity). Mimophis occultus is common in disturbed areas (e.g., roadsides, grazed areas, and 16 17 18 530 near human settlements) and throughout open deciduous forested regions. It is a diurnal and 19 20 531 terrestrial snake, but we also observed individuals at night at Montagne des Français, 21 22 532 Ankarandota, and Kelifely in low branches or on the ground; it is unclear if the snakes were 23 24 25 533 active or preferred the low branches as a roosting site. These snakes prey on squamates and 26 27 534 probably frogs, as does M. mahfalensis . The dwarf chameleon Brookesia stumpfii was found in 28 29 535 the stomach of the paratype AMNH 176427. Several M. occultus specimens (UMMZ 219470, 30 31 32 536 237400, 237403, 237404, 237405, 240451) were heavily infested with worms. These appear to 33 34 537 be sparagnosis infestations of plerocercoid larvae, possibly of Spirometra tapeworms. Similar 35 36 37 538 infestations were also seen in M. mahfalensis (e.g., UMMZ 209638); the source could be from 38 39 539 infected freshwater prey, such as frogs. 40 41 540 42 43 44 541 REMARKS: The large distribution and broad tolerance to human disturbance exhibited M. 45 46 542 occultus makes it unlikely to be of conservation concern. 47 48 543 49 50 51 544 Discussion 52 53 545 Considering the species diversity and large ranges of much of the snake fauna in Madagascar, it 54 55 56 546 is surprising that few studies have investigated for potential cryptic structure in Malagasy snakes. 57 58 59 60 24
URL: http://mc.manuscriptcentral.com/tsab Page 137 of 152 Systematics and Biodiversity
1 2 3 547 Using an integrated molecular and morphological approach, we find that one of the most wide 4 5 6 548 ranging snakes in Madagascar, Mimophis mahfalensis , is composed of two independently 7 8 549 evolving lineages and describe the northern cryptic lineage as the new species Mimophis 9 10 11 550 occultus . 12 13 551 Our results are concordant with the only previously published Mimophis molecular 14 15 552 phylogeny that included five samples from across the island. Specifically, Kelly et al. (2008), in 16 17 18 553 estimating a general phylogeny for the Psammophiinae, included a northern Mimophis sample 19 20 554 (from Montagne des Français) and four southerly samples (from Kirindy, Tulear, and Mount 21 22 555 Ibity). Paralleling our results the deepest genetic break for the genus was between the single 23 24 25 556 northern Mimophis and the four southern individuals. Kelly et al. (2008) concluded there was 26 27 557 likely unrecognized diversity in this genus requiring further investigation. While this prior study 28 29 558 was not extensive in its sampling of Mimophis , the similar results using different individuals 30 31 32 559 further support the findings of our study. 33 34 560 The presence of a cryptic taxon within the widespread Mimophis may not be entirely 35 36 37 561 unexpected, but surprisingly, the three previously noted colour pattern polymorphisms are found 38 39 562 in both species and are represented throughout much of the range of M. mahfalensis and M. 40 41 563 occultus (Fig. 3; Appendix 1). We find no support for the putative High Plateau subspecies M. 42 43 44 564 mahfalensis madagascariensis (see Glaw & Vences, 1994), which was primarily described based 45 46 565 on having a striped colour pattern. Kelly et al. (2008) included the two putative subspecies in 47 48 566 their sampling of Mimophis , and their resulting phylogeny also did not show these subspecies 49 50 51 567 formed monophyletic groups; the northern “M. m. mahfalensis ” fell outside of the clade 52 53 568 including the southern “M. m. mahfalensis ” samples. Rather than being a distinct lineage, the 54 55 569 striped pattern is found within both of the Mimophis species, as are the patternless and zig zag 56 57 58 59 60 25
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 138 of 152
1 2 3 570 types (Fig. 3). These patterns are polymorphic within the genus and interestingly many other taxa 4 5 6 571 within Psammophiinae exhibit similar colour pattern polymorphisms including striped, 7 8 572 unpatterned, and zig zag forms, including: Psammophis crucifer , P. notostictus , P. leightoni , P. 9 10 11 573 sibilans, Psammophylax rhombeatus (Branch, 1998); Hemirhagerrhis nototaenia, 12 13 574 Psammophylax variabilis (Spawls et al., 2001); Psammophis schokari (Kark, Warburg, & 14 15 575 Werner, 1997). 16 17 18 576 It has also been suggested that Mimophis is sexually dimorphic with respect to these 19 20 577 patterns, with males having the zig zag and females the patternless colouration (Glaw & Vences, 21 22 578 2007). However, we find all three patterns are present in both sexes of Mimophis (Appendix 1). 23 24 25 579 The potential adaptive significance or cause for these polymorphisms is unclear, but previous 26 27 580 studies of snakes suggests that intraspecific colour pattern variation may be related to anti 28 29 581 predator behaviours, ecologies, ontogenetic changes, general crypsis, body sizes, or 30 31 32 582 thermoregulation, either singly or in some combination (Forsman, 1995). Although we can rule 33 34 583 out ontogenetic changes and body size (these colour patterns are present in all size and 35 36 37 584 ontogenetic classes), a detailed study isolating the other potential causes is required to 38 39 585 understand what is maintaining these polymorphisms within Mimophis . 40 41 586 Mimophis occultus differs morphologically from M. mahfalensis in two external features: 42 43 44 587 (1) the dark post ocular eye stripe, which is narrower on the temporal scales; and (2) the longer 45 46 588 relative tail length (in both sexes), which usually results in a higher number of subcaudal scales 47 48 589 and a larger tail length/SVL ratio. The eye stripe feature is fully congruent with the molecular 49 50 51 590 results. However, the possible adaptive function of eye stripes in Mimophis (and indeed most 52 53 591 snakes) has not yet been investigated. The role of relative tail length in snakes is has been 54 55 592 previously studied (e.g., Guyer & Donnelly, 1990; Lillywhite & Henderson, 1993), with shorter 56 57 58 59 60 26
URL: http://mc.manuscriptcentral.com/tsab Page 139 of 152 Systematics and Biodiversity
1 2 3 593 tails often found in fossorial snakes and longer tails found in arboreal snakes, and with primarily 4 5 6 594 terrestrial species being somewhere in between (Feldman & Meiri, 2012). The relative tail length 7 8 595 differences between the two species of Mimophis is modest and likely reflects the similar 9 10 11 596 terrestrial lifestyle of both taxa. 12 13 597 Interestingly, in the western region between Maintirano, Belo Tsiribihina and 14 15 598 Tsiroanomandidy where the two species may come in contact, we found six snakes with 16 17 18 599 intermediate morphological characters. These snakes have broad eye stripes on the temporal 19 20 600 scales diagnostic for M. mahfalensis , but also have the longer tails typical of M. occultus. 21 22 601 Because we do not have tissues and molecular data for these specimens, we are unable to 23 24 25 602 determine if these are unusually long tailed M. mahfalensis , or whether they might represent 26 27 603 hybrids. To make further progress on resolving this issue will require making new collections in 28 29 604 this region to acquire additional genetic samples. Including genetic samples from this region in 30 31 32 605 future studies will allow for the examination of the degree of hybridization between the two taxa 33 34 606 in contact zones . 35 36 37 607 Speciation typically occurs via sympatry, parapatry, and allopatry, but while allopatry is 38 39 608 the most frequently referenced mode of speciation, paraptry has been found to be more common 40 41 609 that previously recognized (e.g., Fisher Reid et al., 2013) Isolating geographic features, such 42 43 44 610 mountain ranges or rivers, as well as changes in ecotones/ecological gradients have been 45 46 611 previously found important for speciation among Malagasy reptiles (Florio & Raxworthy, 2016; 47 48 612 Pearson & Raxworthy, 2009; Raxworthy, Ingram, Rabibisoa, & Pearson, 2007). The potential 49 50 51 613 barrier that separates the Mimophis sister taxa is uncertain, however results from the RDA 52 53 614 analyses rule out neutral divergence via geographic distance between populations and the 54 55 615 potential effects of adaptation to current climates in driving speciation. There are several rivers 56 57 58 59 60 27
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 140 of 152
1 2 3 616 that may contribute to divergence in the general region of the genetic break and of morphological 4 5 6 617 intermediacy for Mimophis mahfalensis and M. occultus . Specifically the Tsiribihina River, 7 8 618 which has been found to impede gene flow in lemurs (Pastorini, Thalmann, & Martin, 2003) and 9 10 11 619 has been suggested as a putative driver of speciation for sister taxa of Oplurus iguanas (Chan et 12 13 620 al., 2012) may be an isolating geographic feature for these snakes. Targeted genetic sampling in 14 15 621 this region for Mimophis across these rivers will help determine if this river reduces gene flow 16 17 18 622 between the species. 19 20 623 It is likely that many other species in Madagascar will be revealed as complexes of 21 22 624 distinct species; our research highlights the need for phylogeographic examination of Malagasy 23 24 25 625 snakes, especially those that are broadly distributed. Mimophis mahfalensis is perhaps the most 26 27 626 common diurnal snake encountered in most of Madagascar and our analyses confirm that this 28 29 627 wide spread species is actually two distinct species, each with colour pattern polymorphisms. We 30 31 32 628 recommend further examination for Mimophis and other snakes of Madagascar using 33 34 629 phylogenomic datasets, specifically with sampling across potential barriers, as this will allow for 35 36 37 630 better determination of which geographic features are common drivers of diversification for 38 39 631 Malagasy taxa. 40 41 632 42 43 44 633 Acknowledgements 45 46 634 Field studies in Madagascar were made possible due to the assistance of the Ministère de 47 48 49 635 l'Environnement, de l'Ecologie et des Forêts , Madagascar National Parks, and the Université 50 51 636 d’Antananarivo, Departement de Biologie Animale. Thanks to the American Museum of Natural 52 53 637 History Ambrose Monell Cryo Collection (J. Feinstein) and the Science Research Mentoring 54 55 56 638 Program, M. Valencia Mestre for photos of the holotype, and S. Garnier for statistical discussion. 57 58 59 60 28
URL: http://mc.manuscriptcentral.com/tsab Page 141 of 152 Systematics and Biodiversity
1 2 3 639 4 5 6 640 Disclosure Statement 7 8 9 641 The authors declare no conflicts of interest. 10 11 642 12 13 643 14 Funding 15 16 644 Research support was provided by the National Science Foundation under grants DEB 1257610, 17 18 645 0641023, 0423286, 9984496 (CJR), DEB 9625873, 9322600 (RAN and CJR) and 1257926 19 20 21 646 (FTB), the American Museum of Natural History Gerstner Scholars Program (Gerstner Family 22 23 647 Foundation), the American Museum of Natural History Science Research Mentoring Program, 24 25 648 the Theodore Roosevelt Memorial Fund, the Richard Gilder Graduate School, and the Gerstner 26 27 28 649 Scholar Program. 29 30 650 Research conducted at: American Museum of Natural History and Museum of Zoology, 31 32 33 651 University of Michigan 34 35 652 36 37 38 653 Supplemental data 39 40 654 Supplemental information supporting this research can be found online XXX. 41 42 655 43 44 45 46 656 References 47 48 49 50 657 Bartoń, K. (2016). MuMIn: Multi Model Inference. R package version 1.15.6. Retrieved from 51 52 658 https://CRAN.R project.org/package=MuMIn (accessed 30 August 2017). 53 54 55 56 57 58 59 60 29
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 142 of 152
1 2 3 659 Bogert, C. M. (1940). Herpetological results of the Vernay Angola Expedition: with notes on 4 5 6 660 African reptiles in other collections. Part 1, Snakes, including an arrangement of African 7 8 661 Colubridae. Bulletin of the American Museum of Natural History , 77 . 9 10 11 662 Bond, W. J., Silander Jr, J. A., Ranaivonasy, J., & Ratsirarson, J. (2008). The antiquity of 12 13 663 Madagascar’s grasslands and the rise of C 4 grassy biomes. Journal of Biogeography , 35 , 14 15 664 1743–1758. 16 17 18 665 Boumans, L., Vieites, D. R., Glaw, F., & Vences, M. (2007). Geographical patterns of deep 19 20 666 mitochondrial differentiation in widespread Malagasy reptiles. Molecular Phylogenetics 21 22 667 and Evolution , 45 , 822–39. 23 24 25 668 Brygoo, E. R. (1982). Les ophidiens de Madagascar. Memórias do Instituto de Butantan , 46 , 19– 26 27 669 58. 28 29 670 Chan, L. M., Choi, D., Raselimanana, A. P., Rakotondravony, H. A., & Yoder, A. D. (2012). 30 31 32 671 Defining spatial and temporal patterns of phylogeographic structure in Madagascar’s 33 34 672 iguanid lizards (genus Oplurus ). Molecular Ecology , 21 , 3839–51. 35 36 37 673 D’Cruze, N. C., Green, K. E., Robinson, J. E., & Gardner, C. J. (2006). A rapid assessment of the 38 39 674 amphibians and reptiles of an unprotected area of dry deciduous forest in north Madagascar. 40 41 675 Herpetological Bulletin , 11 , 17–25. 42 43 44 676 Darriba, D., Taboada, G. L., Doallo, R., & Posada, D. (2012). jModelTest 2: more models, new 45 46 677 heuristics and parallel computing. Nature Methods , 9, 772. 47 48 678 De Haan, C. C. (2003). Extrabuccal infralabial secretion outlets in Dromophis , Mimophis and 49 50 51 679 Psammophis species (Serpentes, Colubridae, Psammophiini). A probable substitute for 52 53 680 “self rubbing” and cloacal scent gland functions, and a cue for a taxonomic account. 54 55 681 Comptes Rendus Biologies , 326 , 275–286. 56 57 58 59 60 30
URL: http://mc.manuscriptcentral.com/tsab Page 143 of 152 Systematics and Biodiversity
1 2 3 682 Domergue, C.A. 1969. Cle simplifée pour la determination sur terrain des serpents communs de 4 5 6 683 Madagascar. Bulletin de l'Académie Malgache , 45 , 13–26. 7 8 684 Domergue, C. A. (1989). A venomous snake of Madagascar. 2 case reports of bites by 9 10 11 685 Madagascarophis (Colubrida opisthoglypha). Archives de l’Institut Pasteur de 12 13 686 Madagascar , 56 , 299–311. 14 15 687 Drummond, A. J., & Rambaut, A. (2007). BEAST: Bayesian evolutionary analysis by sampling 16 17 18 688 trees. BioMed Central Evolutionary Biology , 7, 214. 19 20 689 Durkin, L., Steer, M. D., & Belle, E. M. (2011, February 11). Herpetological surveys of forest 21 22 690 fragments between Montagne d’Ambre National Park and Ankarana Special Reserve, 23 24 25 691 Northern Madagascar. Herpetological Conservation and Biology , 6, 114–126. 26 27 692 Evanno, G., Regnaut, S., & Goudet, J. (2005). Detecting the number of clusters of individuals 28 29 693 using the software STRUCTURE: a simulation study. Molecular Ecology , 14 , 2611–2620. 30 31 32 694 Feldman, A., & Meiri, S. (2012). Length–mass allometry in snakes. Biological Journal of the 33 34 695 Linnean Society, 108 , 161–172. 35 36 37 696 Fisher Reid, M. C., Engstrom, T. N., Kuczynski, C. A., Stephens, P. R., & Wiens, J. J. (2013). 38 39 697 Parapatric divergence of sympatric morphs in a salamander: incipient speciation on Long 40 41 698 Island? Molecular Ecology , 22 , 4681–4694. 42 43 44 699 Florio, A. M., Ingram, C. M., Rakatondravony, H. A., Louis, E. E., & Raxworthy, C. J. (2012). 45 46 700 Detecting cryptic speciation in the widespread and morphologically conservative carpet 47 48 701 chameleon ( Furcifer lateralis ) of Madagascar. Journal of Evolutionary Biology , 25 , 1399– 49 50 51 702 1414. 52 53 54 55 56 57 58 59 60 31
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 144 of 152
1 2 3 703 Florio, A. M., & Raxworthy, C. J. (2016). A phylogeographic assessment of the Malagasy giant 4 5 6 704 chameleons ( Furcifer verrucosus and Furcifer oustaleti ). Public Library of Science One , 11 , 7 8 705 e0154144. 9 10 11 706 Forsman, A. (1995). Opposing fitness consequences of colour pattern in male and female snakes. 12 13 707 Journal of Evolutionary Biology , 8, 53–70. 14 15 708 Galtier, N., & Gouy, M. (1995). Inferring phylogenies from DNA sequences of unequal base 16 17 18 709 compositions. Proceedings of the National Academy of Sciences of the United States of 19 20 710 America, 92, 11317–11321 21 22 711 Gardner, C., & Jasper, L. (2009). The urban herpetofauna of Toliara , southwest Madagascar. 23 24 25 712 Herpetology Notes , 2, 239–242. 26 27 713 Glaw, F., & Vences, M. (1994). Amphibians and Reptiles of Madagascar (2nd ed.). Cologne, 28 29 714 Germany: Verlag. 30 31 32 715 Glaw, F., & Vences, M. (2007). Amphibians and Reptiles of Madagascar (3rd ed.). Cologne, 33 34 716 Germany: Verlag. 35 36 37 717 Grandidier, A. (1867). Liste des reptiles nouveaux découverts, en 1866, sur la côte sud ouest de 38 39 718 Madagascar. Revue et Magazine de Zoologie (Paris) , 2: 232–234. 40 41 719 Guibé, J. (1958). Les serpents de Madagascar. Mémoires de l’Institut Scientifique de 42 43 44 720 Madagascar (sér. A, Biologie Animale) , 12 , 189–260. 45 46 721 Gunther, A. (1868). Sixth Account of a New Species of Snakes in the Collection of the British 47 48 722 Museum. Annals and Magazine of Natural History , 4, 121. 49 50 51 723 Guyer, C., & Donnelly, M. A. (1990). Length mass relationships among an assemblage of 52 53 724 tropical snakes in Costa Rica. Journal of Tropical Ecology, 6 , 65 76. 54 55 56 57 58 59 60 32
URL: http://mc.manuscriptcentral.com/tsab Page 145 of 152 Systematics and Biodiversity
1 2 3 725 Jourdran, E. (1903). Les ophidiens de Madagascar [The snakes of Madagascar] (Vol. 459). 4 5 6 726 Paris, France; Imprimerie des Facultés. 7 8 727 Kark, S., Warburg, I., & Werner, Y. L. (1997). Polymorphism in the snake Psammophis schokari 9 10 11 728 on both sides of the desert edge in Israel and Sinai. Journal of Arid Environments , 37 , 513– 12 13 729 527. 14 15 730 Kelly, C. M. R., Barker, N. P., Villet, M. H., Broadley, D. G., & Branch, W. R. (2008). The 16 17 18 731 snake family Psammophiidae (Reptilia: Serpentes): phylogenetics and species delimitation 19 20 732 in the African sand snakes ( Psammophis Boie, 1825) and allied genera. Molecular 21 22 733 Phylogenetics and Evolution , 47 , 1045–60. 23 24 25 734 Kopelman, N. M., Mayzel, J., Jakobsson, M., Rosenberg, N. A., & Mayrose, I. (2015). Clumpak: 26 27 735 a program for identifying clustering modes and packaging population structure inferences 28 29 736 across K. Molecular Ecology Resources , 15 , 1179–91. 30 31 32 737 Legendre, P., & Fortin, M. J. (2010). Comparison of the Mantel test and alternative approaches 33 34 738 for detecting complex multivariate relationships in the spatial analysis of genetic 35 36 37 739 data. Molecular Ecology Resources , 10 , 831. 38 39 740 Lillywhite, H. B., & Henderson, R. W. (1993). Behavioral and functional ecology of arboreal 40 41 741 snakes. In R.A. Seigel & J.T. Collins (Eds.), Snakes: Ecology and Behavior , (pp. 1–48). 42 43 44 742 New York, NY: McGraw Hill. 45 46 743 McLellan, F. (2013). New snake records for the island of Nosy Be, northwest Madagascar: 47 48 744 Mimophis mahfalensis (Grandidier, 1867) and Pseudoxyrhopus quinquelineatus (Günther, 49 50 51 745 1881). Herpetology Notes , 6, 295–297. 52 53 746 Myers, N., Mittermeier, R. A., Mittermeier, C. G., da Fonseca, G. A., & Kent, J. (2000). 54 55 747 Biodiversity hotspots for conservation priorities. Nature , 403 , 853–8. 56 57 58 59 60 33
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 146 of 152
1 2 3 748 Nagy, Z. T., Glaw, F., Andreone, F., Wink, M., & Vences, M. (2007). Species boundaries in 4 5 6 749 Malagasy snakes of the genus Madagascarophis (Serpentes: Colubridae sensu lato) 7 8 750 assessed by nuclear and mitochondrial markers. Organisms Diversity & Evolution , 7, 241– 9 10 11 751 251. 12 13 752 Nagy, Z. T., Joger, U., Wink, M., Glaw, F., & Vences, M. (2003). Multiple colonization of 14 15 753 Madagascar and Socotra by colubrid snakes: evidence from nuclear and mitochondrial gene 16 17 18 754 phylogenies. Proceedings of the Royal Society B: Biological Sciences , 270, 2613–21. 19 20 755 Orozco Terwengel, P., Nagy, Z. T., Vieites, D. R., Vences, M., & Louis JR, E. (2008). 21 22 756 Phylogeography and phylogenetic relationships of Malagasy tree and ground boas. 23 24 25 757 Biological Journal of the Linnean Society , 95 , 640–652. 26 27 758 Paradis E., Claude J., & Strimmer K. 2004. APE: analyses of phylogenetics and evolution in R 28 29 759 language. Bioinformatics, 20 , 289–290. 30 31 32 760 Pastorini, J., Thalmann, U., & Martin, R. D. (2003). A molecular approach to comparative 33 34 761 phylogeography of extant Malagasy lemurs. Proceedings of the National Academy of 35 36 37 762 Sciences of the United States of America , 100 , 5879–84. 38 39 763 Pritchard, J. K., Stephens, M., & Donnelly, P. (2000). Inference of population structure using 40 41 764 multilocus genotype data. Genetics , 155 , 945–959. 42 43 44 765 Pyron, R. A., & Burbrink, F. T. (2010). Hard and soft allopatry: physically and ecologically 45 46 766 mediated modes of geographic speciation. Journal of Biogeography , 37 , 2005–2015 47 48 767 R Core Team (2013). R: A language and environment for statistical computing. R 49 50 51 768 Foundation for Statistical Computing, Vienna, Austria. ISBN 3 900051 07 0, URL 52 53 769 Retrieved from http://www.R project.org/ (accessed 30 August 2017). 54 55 56 57 58 59 60 34
URL: http://mc.manuscriptcentral.com/tsab Page 147 of 152 Systematics and Biodiversity
1 2 3 770 Ramanamanjato, J. B., & Lehtinen, R. (2006). Effects of rainforest fragmentation and correlates 4 5 6 771 of local extinction in a herpetofauna from Madagascar. Applied Herpetology , 3, 95–110. 7 8 772 Ramanamanjato, J. B., Mcintyre, P. B., & Nussbaum, R. A. (2002). Reptile, amphibian, and 9 10 11 773 lemur diversity of the Malahelo Forest, a biogeographical transition zone in southeastern 12 13 774 Madagascar. Biodiversity & Conservation , 11 , 1791–1807. 14 15 775 Rambaut, A., & Drummond, A. J. (2007). Tracer v1.5 . Retrieved from http://beast.bio.ed.ac.uk/ 16 17 18 776 (accessed 1 April 2013). 19 20 777 Rannala, B., & Yang, Z. (2003). Bayes estimation of species divergence times and ancestral 21 22 778 population sizes using DNA sequences from multiple loci. Genetics , 164 , 1645–56. 23 24 25 779 Ratsoavina, F. M., Gehring, P. S., Ranaivoarisoa, F. J., Rafeliarisoa, T. H., Crottini, A., Louis Jr, 26 27 780 E. E., & Vences, M. (2010). Molecular phylogeography of a widespread Malagasy leaf 28 29 781 chameleon species, Brookesia superciliaris . Zootaxa , 2554 , 62–64. 30 31 32 782 Raxworthy, C. J. (2003). Introduction to the reptiles. In S. M. Goodman & J. P. Benstead (Eds.), 33 34 783 The Natural History of Madagascar (pp. 934–949). Chicago, IL: University of Chicago 35 36 37 784 Press. 38 39 785 Raxworthy, C. J., Ingram, C. M., Rabibisoa, N., & Pearson, R. G. (2007). Applications of 40 41 786 ecological niche modeling for species delimitation: a review and empirical evaluation using 42 43 44 787 day geckos ( Phelsuma ) from Madagascar. Systematic Biology , 56 , 907–923. 45 46 788 Raxworthy, C. J., Pearson, R. G., Zimkus, B. M., Reddy, S., Deo, A. J., Nussbaum, R. A., & 47 48 789 Ingram, C. M. (2008). Continental speciation in the tropics: contrasting biogeographic 49 50 51 790 patterns of divergence in the Uroplatus leaf tailed gecko radiation of Madagascar. Journal 52 53 791 of Zoology , 275 , 423–440. 54 55 56 57 58 59 60 35
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 148 of 152
1 2 3 792 Rosa, G. M., Boistel, R., Campantico, E., Gillet, B., Eusebio Bergò, P., & Andreone, F. (2014). 4 5 6 793 Case solved: presence of toxin secreting oral glands in the lamprophiid snake Mimophis 7 8 794 mahfalensis (Grandidier, 1867) from Madagascar. Zoomorphology , 133 , 417–423. 9 10 11 795 Ruane, S., Burbrink, F. T., Randriamahtantsoa, B., & Raxworthy, C. J. (2016). The Cat eyed 12 13 796 Snakes of Madagascar: Phylogeny and description of a new species of Madagascarophis 14 15 797 (Serpentes: Lamprophiidae) from the tsingy of Ankarana. Copeia , 104 :712–721. 16 17 18 798 Ruane, S., Raxworthy, C. J., Lemmon, A. R., Lemmon, E. M., & Burbrink, F. T. (2015). 19 20 799 Comparing species tree estimation with large anchored phylogenomic and small Sanger 21 22 800 sequenced molecular datasets: an empirical study on Malagasy pseudoxyrhophiine snakes. 23 24 25 801 BioMed Central Evolutionary Biology , 15 , 221. 26 27 802 Ruxton, G. D. (2006). The unequal variance t test is an underused alternative to Student's t test 28 29 803 and the Mann–Whitney U test. Behavioral Ecology , 17 , 688 690. 30 31 32 804 Samonds, K. E., Godfrey, L. R., Ali, J. R., Goodman, S. M., Vences, M., Sutherland, M. R., … 33 34 805 Krause, D. W. (2013). Imperfect isolation: factors and filters shaping Madagascar’s extant 35 36 37 806 vertebrate fauna. Public Library of Science One , 8, e62086. 38 39 807 Schutt, P. (2008). Analysis of road kill data from Ankarafantsika National Park, Madagascar . 40 41 808 (Unpublished doctoral dissertation). Durham, NC: Duke University, NC. 42 43 44 809 Stephens, M., & Donnelly, P. (2003). A comparison of bayesian methods for haplotype 45 46 810 reconstruction from population genotype data. The American Journal of Human Genetics , 47 48 811 73 , 1162–1169. 49 50 51 812 Stephens, M., Smith, N. J., & Donnelly, P. (2001). A new statistical method for haplotype 52 53 813 reconstruction from population data. American Journal of Human Genetics , 68 , 978–89. 54 55 56 57 58 59 60 36
URL: http://mc.manuscriptcentral.com/tsab Page 149 of 152 Systematics and Biodiversity
1 2 3 814 Townsend, T. M., Vieites, D. R., Glaw, F., & Vences, M. (2009). Testing species level 4 5 6 815 diversification hypotheses in Madagascar: the case of microendemic Brookesia leaf 7 8 816 chameleons. Systematic Biology , 58 , 641–56. 9 10 11 817 Uetz, P., & Hosek, J. (2015). The Reptile Database . Retrieved from http://www.reptile 12 13 818 database.org (accessed 21 October 2016). 14 15 819 Vences, M., Wanke, S., VieitesS, D. R., Branch, W. R., Glaw, F., & Meyer, A. (2004). Natural 16 17 18 820 colonization or introduction? Phylogeographical relationships and morphological 19 20 821 differentiation of house geckos ( Hemidactylus ) from Madagascar. Biological Journal of the 21 22 822 Linnean Society , 83 , 115–130. 23 24 25 823 Vidal, N., Branch, W. R., Pauwels, O. S. G., Hedges, S. B., Broadley, D. G., Wink, M., … Nagy, 26 27 824 Z. T. (2008). Dissecting the major African snake radiation: a molecular phylogeny of the 28 29 825 Lamprophiidae Fitzinger (Serpentes, Caenophidia). Zootaxa , 1945 , 51–66. 30 31 32 826 Vidal, N., & Hedges, S. B. (2002). Higher level relationships of caenophidian snakes inferred 33 34 827 from four nuclear and mitochondrial genes. Comptes Rendus Biologies , 325 , 987–995. 35 36 37 828 Warton, D. I., Duursma, R. A., Falster, D. S., & Taskinen, S. (2012). smatr 3–an R package for 38 39 829 estimation and inference about allometric lines. Methods in Ecology and Evolution , 3, 257 40 41 830 259. 42 43 44 831 Welch, B. L. (1947). The generalization ofstudent's' problem when several different population 45 46 832 variances are involved. Biometrika , 34 , 28 35. 47 48 833 Yang, Z. (2002). Likelihood and Bayes estimation of ancestral population sizes in hominoids 49 50 51 834 using data from multiple loci. Genetics , 162 , 1811–23. 52 53 54 55 56 57 58 59 60 37
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 150 of 152
1 2 3 835 Yang, Z., & Rannala, B. (2010). Bayesian species delimitation using multilocus sequence data. 4 5 6 836 Proceedings of the National Academy of Sciences of the United States of America , 107 , 7 8 837 9264–9. 9 10 11 838 Yang, Z., & Rannala, B. (2014). Unguided species delimitation using DNA sequence data from 12 13 839 multiple Loci. Molecular Biology and Evolution , 31 , 3125–35. 14 15 840 Associate Editor: Mark Wilkinson 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 38
URL: http://mc.manuscriptcentral.com/tsab Page 151 of 152 Systematics and Biodiversity
1 2 3 841 Table 1. Morphological characteristics and geographic limits of Mimophis species. 4 5 842 Mimophis occultus sp. nov. Mimophis mahfalensis 6 843 Holotype All specimens All specimens 7 844 Character n = 43 n = 32 8 9 845 Maximum SVL 443 mm 620 mm 580 mm 10 846 Maximum tail length 172 mm 219 mm 191 mm 11 847 Ratio Tail / SVL 0.39 0.33−0.43 0.20−0.36 12 848 Midbody dorsal scale rows 17 17 17 13 849 Ventral scales 160 152−170 144−167 14 15 850 Subcaudal scales 92 79−104 44−88 16 851 Uniform color morph n/a yes yes 17 852 Striped color morph yes yes yes 18 853 Zig zag color morph n/a yes yes 19 854 Dark posterior eye stripe width ≤ 2 temporal scales width > 2 temporal scales or 20 855 fused to dark dorsal head 21 22 856 Latitude range 16° 12−19° S 18−26° S 23 857 Elevation range 41 m 0−700 m 0−1680 m 24 858 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 39
URL: http://mc.manuscriptcentral.com/tsab Systematics and Biodiversity Page 152 of 152
1 2 3 859 Fig. 1. The three color pattern polymorphisms found in live Mimophis , patternless (1.1), zig zag 4 5 6 860 (1.2), and striped (1.3). Photos (1.1, 1.2) of Mimophis mahfalensis from Beraketa, southeastern 7 8 861 Madagascar and (1.3) of Mimophis occultus sp. nov. from Ankarana, northwestern Madagascar, 9 10 11 862 by CJR. 12 13 863 14 15 864 Fig. 2. Concatenated BEAST tree of all loci for 39 Mimophis samples showing two main clades 16 17 18 865 of Mimophis found in Madagascar, one with a northern and northwestern distribution ( n = 15) 19 20 866 and one with a central and southern distribution ( n = 24), with Lamprophis guttatus as an 21 22 867 outgroup; * indicate posterior probability support ≥ 95%. 23 24 25 868 26 27 869 Fig. 3. (3.1) Sampling localities for genetic samples and morphological samples (total n = 89) 28 29 870 and (3.2) color pattern polymorphism localities for Mimophis specimens examined in this study 30 31 32 871 that could be classified for color pattern (n = 228) with examples of the three color patterns from 33 34 872 preserved Mimophis vouchers (locality details in Supplementary Appendix 1). 35 36 37 873 38 39 874 Fig. 4. The holotype of Mimophis occultus sp. nov. UMMZ 237408 . 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 40
URL: http://mc.manuscriptcentral.com/tsab PageSystematics 1 of 152 and Biodiversity
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 URL:30 http://mc.manuscriptcentral.com/tsab 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Systematics and Biodiversity Page 2 of 152 Lamprophis guttatus Mimophis mahfalensisn 1 Mimophis mahfalensisn 2 3 Mimophis mahfalensis n 4 Mimophis mahfalensisn 5 Mimophis mahfalensi s 6 Mimophis mahfalensisn 7 Mimophis mahfalensis 8 Mimophis mahfalensisn 9 Mimophis mahfalensisn 10 Mimophis mahfalensis 11 Mimophis mahfalensis 12 Mimophis mahfalensis 13 Mimophis mahfalensis 14 Mimophis mahfalensis 15 16 Mimophis mahfalensisn 17 Mimophis mahfalensisn 18 Mimophis mahfalensisn 19 Mimophis mahfalensis 20 Mimophis mahfalensisn 21 Mimophis mahfalensis 22 Mimophis mahfalensisn 23 Mimophis mahfalensisn 24 Mimophis mahfalensis 25 Mimophis mahfalensis n 26 Mimophis mahfalensisn 27 Mimophis mahfalensisn 28 Mimophis mahfalensis 29 n Mimophis mahfalensisn 30 31 Mimophis mahfalensisn 32 Mimophis mahfalensisn 33 Mimophis mahfalensisn 34 Mimophis mahfalensisn 35 Mimophis mahfalensisn 36 Mimophis mahfalensis 37 Mimophis mahfalensisn 38 Mimophis mahfalensis 39 Mimophis mahfalensis 40 Mimophis mahfalensis 41 URL: http://mc.manuscriptcentral.com/tsabMimophis mahfalensis 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 3 of 152 Systematics and Biodiversity
1 2 3 4 5 6 Mimophis occultus 7 8 9 10 11 12 13 14 15 16 Mimophis mahfalensis 17 18 19 20 21 22 23 24 25 26 27 Mimophis 28 29 30 31 32 URL: http://mc.manuscriptcentral.com/tsab 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Systematics and Biodiversity Page 4 of 152
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47
48 209x296mm (300 x 300 DPI) 49 50 51 52 53 54 55 56 57 58 59 60 URL: http://mc.manuscriptcentral.com/tsab