Anabantoids: labyrinth fishes

Anabantoids, the Labyrinth Fishes. The widely-distributed Anabantoids include the Southeast Asian Bettas and Paradisefish and all the gouramis, but also the African Ctenopoma and its relatives. The labyrinth fishes that we were for so long accustomed to call "Anabantids", which would suggest a closer grouping than -oids implies,  actually fall into three family groups, according to a 2005 study that for a first time bases a revised phylogeny on mitochondrial and nucleotide molecular divergences, rather than the purely morphological similarities and dissimilarities that had given rise to conflicting interpretations for decades. The study's results also suggested that the fascinating brood care that includes bubble-nest building, underwater plant nest building and mouthbrooding evolved separately three times in the Anabantoidea, from an ancestral free-spawning behavior without brood care. 
So, revising my old view, the largest Anabantoid  family is the Osphronemidae of 108 species, with tiny Spike-Tailed Paradisefish, the Betta species and Macropodus, the Paradisefish. The familiar Trichogaster (now including Colisa) gouramis all belong here too, though their familiar scientific names are currently being shuffled to satisfy historical quibbles, and also the original "Goramy," the edible Giant Gourami that is sometimes mistakenly bought for aquaria. "Kissing" Gouramis (Helostoma temminki) , which are highly specialised as filter-feeders, form a hard-to-place outgroup, currently placed in their own family, Helostomatidae. The third family of Anabantoids,  with 28 species, are the Anabantidae themselves, like Anabas, the Climbing Perch, which is actually another Asian food fish, only rarely kept in aquaria. and its sister group, the African labyrinth fishes, such as Ctenopoma.
A disjunct distribution. The modern biogeography of the labyrinth fishes gives clues to where the group evolved, but conflicting clues to how they arrived at their present widely separated and mutually isolated ("disjunct") distribution in Africa and in tropical Asia, separated by the dry Iranian plateau and Arabia. It's generally agreed in principle that the distribution of primary freshwater fishes, those without distant marine ancestors, provides a clear link between the geological and the biotic evolutionary histories of a region. The reason is simple: the dispersal of fishes, since they're confined to water, depends on direct connections between river basins, and the history of basin interconnections reflects the transformations of regional geology.
Today the center of Anabantoid species diversity lies in South-East Asia, ranging from India to Borneo. The stem-group, the least-specialised group, from which the other families have diverged, appears to be the Anabantids, with Ctenopoma a relict outlying genus, though it is distributed all over Africa, where different species are present in every major African watershed, from the Nile to the Zambezi. But there are no anabantoids in Madagascar, where Cichlids, for example, left some representatives in the course of spreading to India while the subcontinent and Madagascar were still bound to Africa. So, the modern distribution could  suggest that "primitive" Anabantids developed in southern Laurasia, before India came ashore from the south and crumpled up the Himalayas, and that they spread as far as Africa all across Iran and Arabia during much wetter Paleocene and Eocene climates, before before the Red Sea split opened to make a saltwater barrier between southwest Asia and Africa, and long before those areas were uplifted as plateaus and dried out, a process starting about 11 million years ago.
To date just one articulated Anabantoid fossil has been discovered; it's tentatively assigned to Osphronemus, the Giant Gourami, and was found in lake-bed shales of Central Sumatra that are dated to  late Eocene to early OIligocene times, that is 37 to 28.5 million years ago.. Significant differences among the Anabantoids in the rates at which their DNA mutates make it impossible so far to assign absolute dates for the splits in the genealogical history, and so to judge the most credible pattern of the fishes' dispersal.  In his section on Anabantoidea for The Timetree of Life (2009), Lukas Rüber finds that mitochondrial and nuclear DNA data give conflicting dates for the separation of the group into its families, dating the split either Mid Eocene (about 40 Mya) or Late Cretaceous (about 90 Mya). That's quite a stretch of time.
Raising anabantoid young. The minute anabantoid fry are famously delicate. At the age of some weeks they may die in droves. The common explanation is that the labyrinth organ is developing at that growth phase, and that the young are especially susceptible to drafts of cold, dry air. Could be. A more recent alternative explanation has a more credible ring to me. It suggests that the fry are stressed by ammonia/nitrite/nitrate, which is beginning to build up, the result of constant feedings in bare-bottomed tanks that depend on sponge filters for biofiltration.
Could I make a second alternative explanation? Many adult fish carry a light but sustainable load of skin and gill flukes, both in the wild and in captivity. Anabantoid fishes don't rely strongly on their gills, and may not show acute distress from a moderate parasite load. Some flukes release eggs that hatch into ciliate-like larvae. Others release fully-developed trematodes. Either type of fluke can be passed through contact from a care-giving parent to the fry. A few flukes may be of scant concern to a full-grown fish, but it doesn't take many to overwhelm the developing fry.
Conflicting requirements for clean water and minimal current are an issue in raising the minute and delicate fry of anabantoids. Wayne Schmidt rigged up an easily replicated solution that provides  high-volume fry-tank  filtration with minimal disturbance in the water. Water changes are an issue with delicate fry. While I'm in the house I keep topping up to replace water drained off by a constant siphon going through an airline, with a wrapping of dense sponge acting as a filter in the fry tank, so that even the most minute fry won't get sucked down the tube.  An airstonw would do, but when I was using them, once a week or so, the clogged airstone needed cleaning, and I removeed it and poured boiling water over the airstone and airline. This was just too labor-intensive, though I rarely have more than one batch of minute fry raising. Between uses, I keep the water-filled tubing and the sponge in a jar of water, so that the siphon doesn't require re-starting.
General labyrinth fish links. The Anabantoid Association of Britain archives a couple of dozen articles, with balanced coverage of the Anabantoids, or, strictly for domesticated fancy show Bettas, the U.S.-based International Betta Congress also mainstains a website.
The American Labyrinth Fish Association (ALFA), founded in 2012 under the guidance of the importer of rare fishes Mark Denaro is currently maintained as a yahoo group: .
There is a pageful of links to articles on keeping and breeding anabantoids at Aquarticles. And Calgary Aquarium Society maintains several articles on anabantoids in their extensive archive.