It's hard as hell to prove a negative, and this is outside my area of expertise, but I think Wikipedia has it wrong here.

In James Higginbotham's, Piscinae: Artificial Fishponds in Roman Italy, he spends over 300 pages discussing the various fish tanks and artificial ponds the romans use. For example he cites several examples of lead lined wood tanks used on ships and in fishmongers shops to keep the catch alive and fresh. He provides examples and details for all kinds of artificial ponds and pools.

However every mention I can find of these piscinae places them outside the domicile.

The garden pond could be enjoyed and appreciated from different points of vantage, both public and private. The locations of many ponds within Pompeian houses permit the general or remote glimpse suitable for the casual acquaintance or client. Apertures or windows in the atrium and, specifically, in the tablinum allowed a more public audience to see the pond. Many of these same piscinae and others could be enjoyed from much more intimate surroundings. The location of dining areas and reception rooms around the garden often took into account the location of the fishpond. Triclinia, set within rooms or shaded by pergolas afforded selected visitors or guests front-row seats to the splendors of the gardens and their ponds. At Pompeii, over twenty examples of piscinae are situated in close proximity to dining facilities with approximately an equal number located where they can be seen easily from reception rooms and garden exedrae. The connection between ponds and dining areas is not solely aesthetic but has a gastronomic component as well. Fish or eels fresh from the piscina would make an impressive addition to private meals among friends and honored guests.

Placing the fish pond in an easily viewed location seems to have been en vogue. But Higginbotham doesn't reference a single tank found within a house, let alone place them below the visitor's couch or bed.

It should be mentioned though, and Higginbotham goes out of his way to highlight, that he ignores a lot of water tanks.

In Roman Italy there are countless pools and basins that, conceivably, could have held fish. The vast majority of these probably held water for other purposes (i.e., drinking, bathing, swimming, etc) and their designs reflect no special accommodation for fish. Consequently many of these structures are not included in this study.

It is possible that indoor fish tanks were kept, but because of their ambiguous nature, Higginbotham ignored them. However the fact that they would be ambiguous would undermine the details of the wiki article. If an expert on piscinae can't identify them as such, details like the positioning of beds seem unlikely to be true.

Higginbotham also spends a full chapter discussing the design and operation of the piscinae, quoting ancient authors such as Varro and Columella. Roman fish ponds always had flowing water and clean input and outflow. Even the fishmonger tanks were generally attached to some sort of water source. Thus the mention of accommodation for fish. Higginbotham only details those pools that had such accommodations.

He also discusses which fish were kept by the Romans, citing eight different ancient sources. Associating specific fish with their Latin names is difficult. But by far the most common kept fish appears to be varieties of eels, followed by bass, mullet, bream, and flatfish. There doesn't seem to be a mention of sea barbels.

In short the indoor tanks wiki describes don't seem to exist, and the sea barbel doesn't seem to be one of the fish the Romans kept.