Great observation.  For frog vocalization, I am sure you know that the vocal sacs are key to the communication system.  There are three basic kinds of muscular vocal sacs: median sub-gular, paired sub-gular, and paired lateral. The most widespread is the single median sub-gular and you mention that in your question.  Derived and possibly selected for variations on the single median sub-gular are what we see in some other frog groups - but the main drivers seem to be mostly about sound production (louder) and types of environments where the frogs need to communicate.  I would have to say that there is most likely no single easy answer, but a complex interplay of evolutionary ecology, species assemblages, and environmental factors.  To have anything other than the presumed primitive feature (single median sub-gular) would mean that natural selection has modified the sound making apparatus.  The reasons could be how other species are competing for 'sound space', so diversification of frequencies and and call patterns would be produced.  Other reasons could be simply adaptation to environmental features (e.g., temperature, etc.) or physiological constraints (energetics of calling is different depending on how much air you move).  This is probably a complicated answer - so I hope this helps!

The situation in Ranidae is further complicated by the realization that among the Rana boylii group in the western US, submerged calling is found in most or all of this group but calling at breeding time and sites is not exclusively from beneath the water. Red-legged frogs, which spend much of the year out of water, can be heard calling from terrestrial locations adjacent to breeding sites but also actively calling while submerged. Similarly, foothill yellow-legged frogs, have been reported to call both from submerged as well as above-water locations. Oregon spotted frog, the most aquatic member of the group, have been documented to call from completely submerged locations at a depth (up to 2-3m deep) that completely obscures the sound from observers above the water. This calling behavior has been found to occur at locations that are distant from breeding sites and taking place days before breeding aggregations form. However, once breeding aggregations do form, both submerged and surface calling takes place at these sites. Submerged calling, obviously, requires sound generation without inflation of vocal sacs, which would cause the frogs to bob to the surface. Not surprisingly, then, these western ranids that call while submerged, produce relatively weak calls compared to those species with resonant vocal sacs. I am unaware of whether the presence and type of vocal sacs have been investigated among these western ranids. It would appear, however, that there is evolutionary plasticity in calling strategy as suggested by the recent report that Lithobates subaquavocalis, the Ramsay Canyon leopard frog, described by Platz as a separate species in 1993 and reported to call exclusively while submerged, actually should be considered in the species Lithobates chiricahuensis. The Chiricahua leopard frogs in Copper Canyon, Mexico, produce clear and quite loud calls from the surface. Thus, even within this one ranid species, calling strategy and presumably the anatomical features used in calling, appear to be variable. .

Thank you David and Jay for your thoughtful answers! The amount of 'sound space' the frogs take up is an interesting aspect to consider, as intraspecific competition is apparent with some ranids to that regard; I'll have to give more thought to interspecific. And then the mode at which the western species may call is very interesting... I'm so used to frogs calling in shallows near the water's surface that I hadn't considered ones that could call while submerged and how structure of vocal sac may attribute to that ability.

I know some frog calls carry over different distances - some being louder than others - but I also have to wonder if the direction of calls vary greatly among sac types? We see the vocal sacs inflate as they vocalize, and we generally hear the calls when in proximity, but would paired sacs have more coverage than a singular? I would think so, but I'll have to research it more... And maybe for those who do need more coverage, do they also have larger ranges of movement?

I just have so many things that I'm curious about! Thank you for taking the time to consider these things.

Hey! In addition to the great answers given by Jay and David, maybe for a general overview, you'd like to read our review article "The anuran vocal sac: A tool for multimodal signaling". :)

Article The anuran vocal sac: A tool for multimodal signalling

Thank you, Iris! I would love to read that! :)

Yay! :)