NORTH-WESTERN JOURNAL OF ZOOLOGY 7 (2): 329-333 ©NwjZ, Oradea, Romania, 2011 Article No.: 111209 www.herp-or.uv.ro/nwjz

Explosion into the canyon: an insight into the breeding aggregation of gottlebei Busse & Böhme, 1992

Gonçalo M. ROSA1,2*, Vincenzo MERCURIO3, Angelica CROTTINI4,5,6 and Franco ANDREONE7

1. Centro de Biologia Ambiental, Departamento de Biologia , Faculdade de Ciências da Universidade de Lisboa, Bloco C2, Campo Grande, 1749-016 Lisboa, Portugal. 2. Department of Anthropology, Durrell Institute of Conservation and Ecology, University of Kent, Canterbury, Kent, UK. 3. Museum für Naturkunde, Leibniz-Institut für Evolutions-und Biodiversitätsforschung an der Humboldt-Universität zu Berlin, Germany 4. Division of Evolutionary Biology, Zoological Institute, Technical University of Braunschweig, Mendelssohnstraße, 4, 38106, Braunschweig, Germany 5. Universitá degli Studi di Milano, Dipartimento di Biologia, Sezione di Zoologia e Citologia, Via Celoria 26, 20133 Milano, Italy 6. Current address: CIBIO, Centro de Investigação em Biodiversidade e Recursos Genéticos, Campus Agrário de Vairão, R. Padre Armando Quintas, 4485-661 Vairão, Portugal. 7. Museo Regionale di Scienze Naturali, Via G. Giolitti, 36, I-10123, Torino, Italy. *Corresponding author, E-mail: [email protected]

Received: 15. December 2010 / Accepted: 19. July 2011 / Available online: 21.July 2011

Abstract. We describe the explosive breeding aggregation event in the rainbow toad, , an endemic of Isalo Massif, south-central Madagascar. The were observed just after the first seasonal rainfall (early November 2009). Males and females aggregated in pools formed within the narrow canyons where they live. The breeding event lasted for three nights, with males mating with females in inguinal amplexus. Males were much more abundant than females, with a tertiary sex ratio of about 34.5% of females. Significant correlation was observed between the snout-vent length and body mass for both sexes.

Keywords: Anura; reproduction, canyon; Isalo; Madagascar; microhylids, sex ratio.

Nine species are currently ascribed to the micro- defining conservation programs (Andreone et al. hylid genus Scaphiophryne from Madagascar (Glaw 2005a, Mercurio et al. 2008), we here provide the & Vences 2007, Vieites et al. 2009). The rainbow first description of the breeding aggregation in S. frog S. gottlebei, endemic of the sandstone Isalo gottlebei and other aspects of its reproductive biol- Massif (south-central Madagascar), is well-known ogy. for its attractive colour pattern with white, green- The aggregation event of S. gottlebei was re- ish, pinkish-red and black blotches, that makes of corded in 2009 in an area known as Zahavola it one of the most requested Malagasy frogs for the (22°37'17.54"S - 45°21'41.49"E, 850 m a.s.l.), which pet trade (Andreone et al. 2005a, 2005c, 2006). The lies just outside the administrative borders of the species is currently listed as CITES II and assessed Parc National de l’Isalo (PNI). In Isalo, rare but in- as Endangered according to IUCN (2010). tense rainfalls usually take place from late October The adults live inside and around the narrow to February (Projet ZICOMA 1999). The Zahavola canyons that cross the massif. They usually hide in site is a small basin (4x3 m width) within a narrow the sandy substrate, or climb up the canyon wall and cave-like canyon (Andreone et al. 2005b), with hiding in small crevices and cavities caused by sandy substrate on the bottom (Fig. 1). pebble falling (Mercurio & Andreone 2006, Glaw The breeding aggregation was recorded be- & Vences 2007). The psammo-nektonic larvae de- tween 1st-3rd of November 2009 after the first in- velop in pools of lentic water within the canyons, tense rains (max. daily precipitation of 70 mm; re- burying in the sand of the bottom during the day, corded air temperature ranged 16.7-18.7º C). This and swimming in the water column during the was the only rainfall event documented during all night (Mercurio & Andreone 2006). the survey from 1st-14th November, 2009. Snout- Previous papers were focused on the species vent length (SVL) of males and females was meas- (see Andreone et al. 2005b, Mercurio & Andreone ured with a vernier caliper (precision of 0.1 mm), 2006, Crottini et al. 2008), but almost nothing was and body mass with an electronic balance (preci- known of the breeding habits and natural history sion of 0.1 g). The breeding activity, stimulated by requirements; as this is missing information for the rainy precipitation, likely started between the 330 Rosa, G.M. et al.

bei, like other representatives of the genus, lacks nuptial pads (Busse & Böhme 1992), and the am- plexus is inguinal, with the male embracing the female at the groin (Fig. 2.A-E). Most females stand still with one or two males on their back. Single males were calling, either from the water or along the water’s edge on the bottom of the rocky wall emitting a loud thrilling chorus (Fig. 2.F-H). No egg-laying was observed on the night of 3rd November. The next morning (4th November) all paired individuals were still mating and contin- ued with this activity throughout the day. This was the first day without rainfall, but in the eve- ning, most of the individuals were still in the wa- ter when the first egg masses were recorded. Small clutches of dark pigmented eggs were attached to submerged tree branches just below the water sur- face (Fig. 2.I). To safeguard the clutch we did not count the eggs in nature, but we analysed the egg numbers from two preserved individuals captured in 2004, consisting respectively of 196 eggs (FAZC 12866, SVL = 36.10 mm) and 168 eggs (FAZC 12757, SVL = 28.95 mm). On 5th November (at 8 am) several males were still calling but we found only three of them in water (Fig. 2.J). Few and less motivated calls came from the upper part of the canyon, but the great majority of individuals was already burrowed into the sand bank, hidden in crevices and holes in the canyon wall (Fig. 2.K). No breeding activity was recorded on the 6th No- vember. The first egg clutches hatched on those days, so after three days from deposition (Fig. 2.L- M). Statistical analyses were conducted with STA-

TISTICA 9.0 (StatSoft, Tulsa, USA). Morphomet- Figure 1. Temporary water basin inside canyon in Za- rics are provided on Table 1 and Figure 3 and havola, Isalo Massif: A. basin completely dried out, be- comparisons were performed between sexes: sex- fore first heavy rainfall (30th October, 2009); B. over- ual dimorphism in body length (SVL) and body flowing water basin after rainfall (7th November, 2009); C. almost dried out pool, 40 days after the first rains mass was tested using Student’s t test. SVL-body (10th December, 2009) (A-B, photos by V. Mercurio; C, mass relationships were described using analysis photo by G. M. Rosa). of covariance, ANCOVA (Day & Quinn 1989, Zar 2009). Body mass was log10-transformed, reducing night of the 2nd November, but our observations heteroscedasticity which minimises chances of started on the night of 3rd November. After the violating the assumptions of the ANCOVA proce- rainfall the pool filled up, with a maximum depth dure (e.g. Day & Quinn 1989, Ramos-Pallares et al. of around 2 m and held the water for the follow- 2010). ing two months (Fig. 1.B). The correlation between SVL and body mass is Twenty-three adult males and 10 adult fe- significant (p < 0.01) in both males (r = 0.638) and males were found around the basin and into wa- females (r = 0.956) (Fig. 3) with larger individuals ter. When the monitoring began, nearly all the fe- presenting higher weight. Females were also sig- male individuals already amplected, both in wa- nificantly larger and heavier than males (SVL: t = ter, on the ground or even within small holes on 2.86, p < 0.01; body mass: t = 4.23; p < 0.01), with a the canyon’s wall (Fig. 2.A-E). Scaphiophryne gottle- different trend between SVL and body mass (AN- Breeding behaviour of Scaphiophryne gottlebei 331

COVA using sex as factor and SVL as covariate): with the availability of breeding sites and, as we

F(1, 36) = 5.501, p = 0.025. confirmed, as soon as the favourable temporary Females also showed a greater body mass compared to males, a difference that can be ex- Table 1. Biometrics from the Scaphiophryne gottlebei plained by the presence of eggs in females. Simi- population at Isalo (Fianarantsoa Province). Meas- larly to other anurans, the evolution of larger body urements are provided as range, followed by mean, in females may result from fecundity advantage, standard deviation and number of sample units (n). increasing clutch size in larger frogs (Shine 1979, Andersson 1994). Sex SVL (mm) Body mass (g) The pronounced seasonal climate of Isalo is 25.15-33.85 1.90-4.00 Males likely the main responsible cause of the explosive (28.91 ± 1.69, n = 29) (3.46 ± 0.42, n = 29) 26.20-34.50 2.80-7.00 reproductive strategy observed in this species. The Females (30.85 ± 2.30, n = 10) (4.50 ± 1.15, n = 10) reproductive period of S. gottlebei is closely related

Figure 2. Mating behaviour and breeding stages on Scaphiophryne gottlebei: A-E. pairs in amplexus; F-H. unpaired calling males; I. mass of eggs attached to a submerged tree branch; J. non-calling male floating; K. individual hid- den in a hole in the canyon wall; L. larvae hatching; M. three days . All photos by V. Mercurio, except for K, by G. M. Rosa.

332 Rosa, G.M. et al.

aquatic site is established, mating takes place im- males with respect to the females (Guarino et al. mediately. An explosive breeding behaviour is 2010). Nonetheless, the strong male biased sex ra- usually considered a trait of species living in tem- tio has been commonly recorded in breeding porary and seasonal habitats. In Madagascar, this population of explosive breeders (Wells 1977). is the case, for example, of Laliostominae species Data is also available for Boophis tsilomaro where in and some Scaphiophryne (e.g., S. brevis). Anyhow, a a group of about 90 males only one female was similar reproduction was documented for some found (Andreone et al. 2002). other species living in temporary pools within Until now the breeding phenology of S. gottle- rainforests and deciduous forests, such as Boophis bei was quite unknown, as well as its overall dis- albilabris (G. M. Rosa, pers. obs.), B. tsilomaro (An- tribution. Our study work in the last years al- dreone et al. 2002) and in the genus Aglyptodacty- lowed us to unveil some aspects of its habitat lus (Glaw & Vences 2007). preferences and ecology. Our observations at Za- Tertiary sex ratio was highly male biased: only havola showed that the unpredictability and 34.5% (n = 39) of breeding individuals were fe- highly seasonal traits of the Isalo Massif are males. Some hypotheses are available to explain among the reasons of concentrated breeding ac- this biased sex ratio observed in S. gottlebei. tivities in S. gottlebei. The peculiar rocky and can- Among these are: an unbalanced sex ratio at birth; yon habitats represent selected breeding habitats a bias in natural history traits, for example it is for the species, and once again show how the known that individuals live up to two years and Malagasy frog species evolved specialisation traits males reach sexual maturity earlier than females, that only continuing and long-term studies may maybe reflecting the smaller body size of the allow us to describe.

Figure 3. Scatterplots of Log10 body mass (g) against SVL (mm) in males (M, shown by filled triangles) and females (F, shown by open circles) of Scaphiophryne gottlebei popu- lation from Isalo Massif.

Acknowledgements. For help in the field, we are Fund, Gondwana Conservation and Research, EDGE, and indebted to our guide Anicet and the collaborators Nirhy Zoological Society of London. H. C. Rabibisoa and Tokihery J. Razafindrabe. The study was carried out in collaboration with Madagascar National Parks and Parc Botanique et Zoologique de Tsimbazaza. We are grateful to the Malagasy authorities, References in particular the Ministère de l’Environnement et des Eaux et Forêts, for research permits. Fieldwork was Andersson, M. (1994): Sexual selection. Princeton: Princeton funded by Mohamed bin Zayed Species Conservation University Press. Breeding behaviour of Scaphiophryne gottlebei 333

Andreone, F., Cadle, J.E., Glaw, F., Nussbaum R.A., Raxworthy, the rainbow frog from the arid Isalo Massif, southern-central C.J., Vallan, D., Vences, M. (2005a): Species review of Madagascar. Zoology 113(6): 378-384. extinction risks in Madagascar: conclusions from the Global Mercurio, V., Andreone, F. (2006): The of Scaphiophryne Amphibian Assessment. Conservation Biology 19(6): 1790-1802. gottlebei (: ) and Mantella Andreone, F., Mattioli, F., Mercurio, V. (2005b): The call of expectata (Mantellidae: Mantellinae) from Isalo Massif, Scaphiophryne gottlebei, a microhylid frog from the Isalo Massif, southcentral Madagascar. Alytes 23(3-4): 81-95. south-central Madagascar. Current Herpetology 24(1): 33-35. Mercurio, V., Aprea, G., Crottini, A., Mattioli, F., Randrianirina, Andreone, F., Mercurio, V., Mattioli, F., Razafindrabe, T.J. (2005c): J.E., Razafindrabe, T.J., Andreone, F. (2008): The of Good news for three and traded frogs Isalo Massif, southern-central Madagascar: High frog diversity from Madagascar. Froglog 72: 2. in an apparently hostile dry habitat. in: A Conservation Strategy Andreone, F., Mercurio, V., Mattioli, F. (2006): Between for the Amphibians of Madagascar. F. Andreone (ed.), pp 143– environmental degradation and international pet trade: 196. Monografia del Museo Regionale di Scienze Naturali 45, conservation strategies for the threatened amphibians of Torino. Madagascar. Natura 95: 81-96. Projet ZICOMA (1999): Les zones d’importance pour la Andreone, F., Vences, M., Guarino, F.M., Glaw, F., Randrianirina conservation des oiseaux à Madagascar. Projet ZICOMA. J.E. (2002): Natural history and larval morphology of Boophis Ramos-Pallares, E., Serrano-Cardozo, V.H., Ramírez-Pinilla, M.P. occidentalis (Anura: Mantellidae: Boophinae) provide new (2010): Reproduction of Ptychoglossus bicolor (Squamata: insights into the phylogeny and adaptive radiation of endemic Gymnophthalmidae) in an Andean coffee shade plantation in Malagasy frogs. Journal of Zoology 257: 425-438. Colombia. South American Journal of Herpetology 5(2): 143-150. Busse, K., Böhme, W. (1992): Two remarkable frog discoveries of Shine, R. (1979): Sexual selection and sexual dimorphism in the the genera Mantella (Ranidae: Mantellinae) and Scaphiophryne Amphibia. Copeia 1979: 297-306. (Microhylidae: Scaphiophryninae) from the west coast of Vieites, D.R., Wollenberg, K.C., Andreone, F., Köhler, J., Glaw, F., Madagascar. Revue Française d'Aquariologie 19: 57-64. Vences, M. (2009). Vast underestimation of Madagascar's Crottini, A., Chiari, Y., Mercurio, V., Meyer, A., Vences, V., evidenced by an integrative amphibian inventory. Andreone, F. (2008): Into the canyons: The phylogeography of Proceedings of the National Academy of Sciences 106(20): 8267- the Malagasy frogs Mantella expectata and Scaphiophryne gottlebei 8272. in the arid Isalo Massif, and its significance for conservation Zar, J.H. (2009): Biostatistical Analysis (5th ed). Prentice Hall, inc. (Amphibia: Mantellidae and Microhylidae). Organisms New Jersy. Diversity & Evolution 8(5): 368-377. Day, R.W., Quinn, G.P. (1989): Comparison of treatment after the analysis of variance in ecology. Ecological Monographs 59(4): 433-463. Glaw, F., Vences, M. (2007): A Fieldguide to the Amphibians and Reptiles of Madagascar. 3rd edition. Vences & Glaw. Verlag, Cologne. Guarino, F.M., Tessa, G., Mercurio, V., Andreone, F. (2010): Rapid sexual maturity and short life span in the blue-legged frog and