I witnessed this explosive breeding spectacle in two consecutive years (both times mid-December - see http://www.inaturalist.org/observations/36714) at Montagne d'Ambre. There were thousands of bright yellow frogs mating noisily in a very shallow pool (video clip with audio available on request).
What puzzles me is that in all the literature concerning all species of Aglyptodactylus, it says that breeding MALES turn yellow (to a greater or lesser degree depending on species) yet in the population I observed clearly BOTH sexes turn yellow.
A. madagascariensis and A securifer are both known from Montagne d'Ambre. My justification for identifying these specimens as the latter is that Glaw & Vences (2007) say the breeding males are 'bright yellow', whereas in the former species they are only described as 'partly yellowish'. However, they were observed in rainforest habitat, which is not typically the domain of A. securifer. I am wondering if these could represent an undescribed species of Aglyptodactylus.
For further information, to see additional images or to enquire about using my photos, feel free to contact me via www.madagascar-photography.com