Calisto obscura subsp. grannus

Hausmann, Axel, 2017, A new species of the hysius species-group of Calisto Hübner (Lepidoptera, Nymphalidae, Satyrinae) and insights into the status of different populations currently attributed to C. grannus Bates, Zootaxa 4317 (1), pp. 1-44 : 21

The C. grannus populations

We analyzed 58 sequences belonging to the eight taxa included in the former C. grannus complex or species group due its genitalic morphology and the possession of two circled ocelli with central pupils at the under surface of HW ( Schwartz & Gali 1984, Gonzalez 1987). Some were originally described as separate species from localities separated a few kilometers at the major mountain range of Hispaniola, the Cordillera Central ( C. grannus grannus M. Bates, 1939 , C. grannus dilemma González, 1987 , C. grannus phoinix González, 1987 , C. grannus amazona González, 1987 , C. grannus micheneri Clench, 1944 ) whereas others were discovered at more distant places of the C. Central ( C. grannus dystacta González, 1987 ), the middle Sierra de Neiba ( C. grannus micrommata Schwartz & Gali, 1984 ) or the southern Sierra de Bahoruco ( C. grannus sommeri Schwartz & Gali, 1984 ) ( Fig. 19 ). Sourakov & Zakharov (2011), based on COI barcode sequences obtained from 28 specimens, lowered the status of most of these taxa from species to subspecies level using mitochondrial DNA as evidence of low divergence level between them. We did not include sequences from the latter study in our analysis as they lacked precise geographical data, with the exception of six sequences representing C. grannus sommeri , which we included in our dataset.

No Bayesian, Maximum Likelihood or Neighbor Joining produced a single cluster formed only by specimens from the same locality or belonging to a single taxon with most clades weakly supported ( Fig. 18 ). K2P pairwise distances values were between 0 and 2.1%, or 0 to 12 differences, with intra-taxon minimum pairwise distances ranging from 0 ( C. grannus micrommata ) to 1.1% ( C. grannus phoinix ) and intertaxa minimum pairwise distances from 0.3 ( C. grannus micrommata – C. grannus sommeri ) to 1.2% ( C. grannus amazona – C. grannus micheneri ).

The species identification methods recovered a single entity (all PTP variants, ABGD, BIN) or two and as much as 14 entities, GMYC single and multiple threshold respectively ( Fig. 18 ). The later method also mixed specimens from different named populations mixed within the delimited entities.

The Median Joining haplotype network shed some light on the relationships of these populations ( Fig. 20 ). The analysis identified 36 haplotypes. Calisto grannus sommeri and C. grannus micrommata were represented by 8, 18 sequences, and 2, 4 sequences, haplotypes respectively. These haplotypes are unique from their respective geographic ranges, Sierra de Bahoruco and Sierra de Neiba. Populations of the northern Cordillera Central included 26 haplotypes, 36 sequences, 19 of them represented by unique sequences. Four haplotypes of these populations are shared by two or three named populations ( Fig. 20 ).

Figures

FIGURE 18. Bayesian COI barcodes gen tree reconstructed by MrBayes showing the relationships among taxa belonging to the hysius species group of Calisto. Numbers above branches represent the posterior probalities / boostrap values of Bayesian Inference / Maximum Likelihood. Bold numbers at species clades represent the support values for each species calculated by the Bayesian Poisson Tree Process (bPTP) species delimitation method. Gray box indicating the position of the species described herein, C. bahoruco. Gray bars at the right side of tree showing the species hypothesis obtained by the different species delimitation methods applied: Generalize Mixed Yule Coalescent single and multiple threshold (GMYCs, GMYCm), Poisson Tree Process, PTP, variants: Bayesian (bPTP), Maximum Likelihood (mPTP), and standard (PTP), Automatic Barcode Gap Discovery (ABGD), and Barcode Index Number (BIN).

FIGURES 19 – 20. 19 — Geographic distribution of C. grannus populations, nomenclature follows Sourakov & Zakharov (2011). 20 — Median Joining Haplotype network obtained in Network v. 5.0; circles size proportional to number of sequences; gray circles are hypothetical haplotypes; black circles represent changes; colors represent subspecies sensu Sourakov & Zakharov (2011).