Lygodactylus petteri Pasteur & Blanc, 1967

Vences, Miguel, Multzsch, Malte, Gippner, Sven, Miralles, Aurélien, Crottini, Angelica, Gehring, Philip-Sebastian, Rakotoarison, Andolalao, Ratsoavina, Fanomezana M., Glaw, Frank & Scherz, Mark D., 2022, Integrative revision of the Lygodactylus madagascariensis group reveals an unexpected diversity of little brown geckos in Madagascar’s rainforest, Zootaxa 5179 (1), pp. 1-61 : 19-20

Lygodactylus petteri Pasteur & Blanc, 1967

Lygodactylus madagascariensis petteri Pasteur & Blanc, 1967

Chresonyms:

Lygodactylus madagascariensis petteri: Kluge (1991) ; Glaw & Vences (1992, 1994, 2007); Puente et al. (2009); Gippner et al. (2021)

Lygodactylus (Domerguella) madagascariensis petteri: Rösler (2000b)

Name-bearing type: holotype MNHN 1990.4 , female.—Type locality: “Montagne d’Ambre, forêt ancienne-Roussettes” according to the original description.—Other types: two paratypes ; MNHN 1990.5 , male; and MNHN 1893.194 .—Etymology: eponym for Jean-Jacques Petter .

Identity and Diagnosis. This nomen was coined for specimens from Montagne d’Ambre that were considered to be a subspecies of L. madagascariensis . According to the original description, this subspecies was purported to differ from typical L. madagascariensis by fewer scales in general (i.e., lower values in various scale counts), suggesting overall larger scales; a larger body size; a different coloration (beige vs. brown); and some other possible differences. Indeed, our measurements and scale counts of the name-bearing type (holotype) and one paratype confirmed these are relatively large-sized (SVL 33.2–35.0, thus at and slightly beyond the upper size limit of L. madagascariensis ) and have lower longitudinal counts of dorsal scales (189 in the holotype; vs. 205–258 in L. madagascariensis ) and ventral scales (102 in the holotype and 103 in one paratype; vs. 106–138 in L. madagascariensis ). This suggests the name petteri should be applied to one of the Domerguella lineages occurring at Montagne d’Ambre.

One of these (called L. sp. 10 by Gippner et al. 2021) appears to reach rather large-sizes (36.9 mm SVL in one specimen) but has relatively high longitudinal counts of dorsal scales (239–240 in two available specimens), thus differing from the types of L. petteri . This lineage (known only from the west slope of Montagne d’Ambre) represents a new species that will be formally named and described below.

Another lineage from Montagne d’Ambre (L. sp. 24) is represented by only one genetic sample, the voucher of which was not available for examination. Unfortunately, no information at all on the coloration or morphology of this lineage is available. We can only hypothesize from its rarity (no further specimens found despite intensive surveys in Montagne d’Ambre) that it is unlikely to correspond to the types of petteri .

The third lineage is the one that we have genetically assigned to L. madagascariensis above, and the one individual from Montagne d’Ambre examined ( Table 1 ) agrees well with topotypical specimens of this species, but not with the petteri types.

However, a fourth lineage from Montagne d’Ambre agrees in all morphological characters very well with the petteri types: it consists of relatively large specimens (SVL in our material 30.3–38.5 mm) with few ventral scales (101–113 vs. 102–103 in the types) and dorsal scales (209–222 vs. 189 in the holotype). We therefore are confident that the specimens belonging to this lineage are conspecific with the types of L. petteri . Since this lineage co-occurs on Montagne d’Ambre with L. madagascariensis with deep genetic differentiation in both mitochondrial and nuclear genes, we conclude that the nomen petteri applies to a full species, Lygodactylus petteri , and we therefore herewith formally elevate it to species level.

It needs to be emphasized that due to a lack of comparative morphological data of the only specimen of L. sp. 24 we cannot fully exclude that this lineage also matches morphologically the holotype of L. petteri and may be conspecific with it. Collection of additional material of L. sp. 24, or alternatively, molecular “archival DNA” data from the holotype of L. petteri , is needed to fully ascertain the identity of these geckos from Montagne d’Ambre. However, independent from these remaining questions, it appears we can conclude with sufficient reliability that L. petteri is not conspecific with L. madagascariensis from which it differs morphologically, and we confirm it is distinct from L. sp. 10, which is described as a new species below.

No clear and consistent differences in color or pattern were found between L. madagascariensis and L. petteri ; both showed a considerable variation in dorsal pattern (near-uniform to heavily patterned in L. madagascariensis vs. asymmetrical series of rather small dorsolateral markings or striped phenotype in L. petteri ). However, the two specimens of L. petteri for which life coloration is known ( Fig. 12 ) do not show the longitudinal rows of large beige patches typical for many L. madagascariensis , and furthermore, the male specimen ZSM 195/2018 has yellow elements dorsally, which we have not seen in any L. madagascariensis . Ventrally the throat is yellowish and ranged from near unspotted to weakly and irregularly spotted in both species.

According to the original description of L. petteri by Pasteur & Blanc (1967), it differs from L. madagascariensis by several characters, which we review here. First of all, L. petteri purportedly has fewer scales (and thus larger ones) in general (characters 9, 12, 13, 17, 31, 32, 33 of Pasteur & Blanc 1967). This agrees with our findings for longitudinal counts of dorsal and ventral scales, while for instance the number of supralabials (character 9 of Pasteur & Blanc 1967) does not clearly differ between the two species according to our data. The authors also reported a larger body size for L. petteri , which is in agreement with our data, as well as differences in coloration and in sexual dimorphism, and a possibly larger size of preanal pores in L. petteri . Once more extensive series of both species become available, it will be worth examining whether these characters may indeed constitute diagnostic differences.

Natural history. A half-digested specimen of L. petteri was regurgitated by a young Compsophis sp. aff. laphystius ( Hutter et al. 2018). Two specimens of L. petteri (ACZC 1407 and ACZC 1427) were found under the bark of Eucalyptus sp. trees at the Gîte d’Étape site on Montagne d’Ambre.

Distribution. L. petteri is only known from its type locality, Montagne d’Ambre.

Figures

FIGURE 12. Photos of specimens of Lygodactylus petteri from Montagne d’Ambre in life. A, Adult male ZSM 195/2018 (MSZC 485); B, C, Female ZSM 194/2018 (MSZC 454) in dorsolateral and ventral views. Not to scale.