According to Pichet (Interfish breeder team - Thailand), the first"dragons" were created by using a "super" red plakat, a red copperplakat and Betta sp. mahachai . With some difficulty, the wild Bettasp. mahachai was crossed to the red copper plakat with the intent ofpreserving the Betta sp. mahachai traits in the offspring. Then a youngmale from this spawn was crossed to a super red plakat female. Thethird step was a mother x son backcross: the super red female wascrossed to one of her offspring in a method commonly referred to as'line breeding'. Although the body shape and finnage of the fry was notvery good and some of them even showed malformations in the body, thecolor was there and the first dragon pair was born! The first "dragons"were developed by Mr. Tea. The "Red dragon V1" were first presented tothe Thai public in the December 2004 edition of a magazine called"Fancy Fish" by the Interfish breeder team .
The shiny iridescent layer of the red "dragon" V1 was not completelycovering the body and it was Mr. Somchat (Interfish breeder team -Thailand) who used these fish to improve the color and finnage into thered "dragon" V2. In 2005, the first dragons were shown at Aquarama bythe Interfish breeder team who also first introduced the red "dragon"to the rest of the world. At this stage the fish was still shortfinnedand although it did not take home any prizes, many Asian breeders wereimpressed by its appearance.
The "Armadillo" strain
Around the time the Interfish breeder team was amazing the world withtheir new "dragon" bettas, a strain called the "Armadillo" was beingproduced in the United States by Victoria Parnell-Stark ofBettysplendens.com. The initial idea for them came from a book by WaltMaurus , in which he included a photograph of a very unique lookingbetta that was the result of crossing Betta splendens with wild bettas,such as Betta imbellis. This beauty was called the "Neon" betta.Determined to reproduce the anomaly, Victoria Parnell-Stark acquiredseveral pairs of wilds from Thailand in a shipment that included Bettaimbellis as well as Betta sp. mahachai. As with the results alreadydescribed above, the Betta sp. mahachai crosses proved not onlydifficult but temperamental, and it was only with some trial and errorthat a successful spawn was recorded between a Betta sp. mahachai and agreen Betta splendens. However, the resulting fry were neverparticularly hardy or prolific.
There was considerably more success with the Betta splendens x Bettaimbellis crossings, and some very interesting, highly iridescent andmasked fish were beginning to appear. The first generation yielded manygreen/red and multicolored bettas, some with highly iridescent scalesand brightly colored gills. Sequential breedings enhanced the effect,so that within a few generations bettas with heavier iridescence andfully masked faces were beginning to appear. The Betta sp. mahachai xBetta splendens hybrids were not faring so well, so before theirinfluence was lost entirely they were absorbed into the line of Bettasplendens x Betta imbellis hybrids. In 2006 the first "Armadillo"bettas were offered for sale, and it was not very long before bettaenthusiasts were noticing similarities between the "Armadillos" and the"dragons". Indeed, their origins followed a very correlative formula,with the differences (particularly in color) probably being attributedto the amount of Imbellis influence in the "Armadillo".The "Neon" betta, a Betta splendens x Betta imbellis hybrid "Armadillo" plakat 
"Dragons" - Back to the roots!
Betta imbellis, Betta smaragdina and Betta sp. mahachai naturally havea higher degree of iridescence on the scales compared to Bettasplendens which probably is an adaption to the murkier water in theirnatural habitat. Both "dragons" and "armadillos" described above arecharacterized by a thick metallic iridescent scaling which almostresembles body armor. When we have a look at the origin of metallicsand "dragons" we see quite a lot of overlap. Outcrossing domesticatedbettas to Betta imbellis gave rise to the metallics which in turn alsohave an important contribution to the development of the "dragon". Thisshared origin mostly likely explains the similarity between metallicsand "dragons" but in case of the latter the additional influence ofBetta sp. mahachai seems to be the key to the thick shiny armor-likemetallic scaling. When we compare wildtype Betta imbellis and Betta sp.mahachai we can already see the iridescence is shinier and thicker inthe latter.Betta splendens (wildtype) Betta imbellis (wildtype) Betta smaragdina (wildtype) Betta sp. mahachai (wildtype) 
When we take these four (sub)species theoretically the following crosses could be made:
- Betta splendens x Betta imbellis
- Betta splendens x Betta smaragdina
- Betta splendens x Betta sp. mahachai
- Betta imbellis x Betta smaragdina
- Betta imbellis x Betta sp. mahachai
- Betta smaragdina x Betta sp. mahachai
With exception of Betta splendens x Betta imbellis crosses, not much isdocumented about whether all these species really can interbreedwithout problems. When we look at the origin of the "dragon" asdescribed above, accoring to Pichet it was not easy to cross the redcopper betta x Betta sp. mahachai. Please note that copper is ametallic phenotype which was created by crossing Betta splendens toBetta imbellis, however how many generations this particular fish isseperated from this initial hybridization is unknown. Strikingly,Victoria Parnell-Stark also reported difficulties with the Bettasplendens x Betta sp. mahachai hybrids. In both cases it seems that amixture of these three species was necessary in order to create andfixate a healthy, viable line. So why where some of these crossesdifficult? Was this is just a coincidence or could it be that certainspecies were evolutionary more separated from each other? Does thesimilarity in appearance between Betta imbellis and Betta sp. mahachaiand the readiness to interbreed mean that they are evolutionary closerrelated to each other than to Betta smaragdina and/or Betta sp.mahachai? Based on their phenotype Betta smaragdina and Betta sp.mahachai also show quite some similarities, does this implicate thatthey are more easily easily intercrossed than to Betta splendens and/orBetta imbellis?
Some other unanswered questions/remarks to keep in mind with respect to this subject are:
(1) Does it matter which sex we use to interbreed species? Are theresults different when we breed a male of species A with a female ofspecies B or the other way around?
(2) Are F1 hybrids of the above mentioned crosses normally fertile, andis this sex dependent? It is known that F1 hybrids of two species whichare further apart in the phylogenetic tree could be infertile. Doesthis account for both sexes or is it in some cases possible to usehybrid males or females to cross back to one of the species used in theinitial cross?
(3) Can hybridism function as a genetical "bridge" between two distinctspecies? The idea of the later hypothesis would be that the F1 hybridoffspring of two different species might function as a genetical"bridge" for the interbreeding of two species which are evolutionaryfurther apart."Dragons" - Genetics?
We know that the first metallic bettas were created around the year2000. It took about 5 years before Dr. Leo Buss provided publishedseveral papers which provided us with more information and insightabout the genetic make-up of this phenotype [15-18]. By now the sametime-span has passed since the first "dragons" were developed but sofar not much is known about genetic make-up and inheritance of thisphenotype and thereby leaving us with quite some unanswered questions:What genes determine the "dragon" phenotype?
We can clearly see that the unique appearance of the "dragon" clearlyaffects the iridescent layer. So far several genes have beencharacterized which affect the iridescent layer, for example theclassical iridescent colors steel blue (blbl), turquoise (BlBl) androyal blue (Blbl), metallic or yellow reflecting iridophore (thewildtype variant of the reduced yellow iridophore locus)  and thespread iridescence gene (Si) which is responsible for the increase indensity and distribution of the iridescent color.
In 2007, Joep van Esch of Bettaterritory.nl made some experimentalcrosses with "dragons" in his fishroom in order to learn more about theheredity behaviour of this trait (see pedigree below). After crossing a"dragon" to a regular metallic (BT171007A
)he observed that all offspring had a copper phenotype. Considering thefact that a copper phenotype (homozygous metallic steel blue, blbl ++)can only be obtained when the offspring obtains one steel blue (bl) andone metallic (+) allele from each parent , this result suggests thatboth traits may play an important role in the unique appearance of the"dragon" phenotype. Interestingly, the metallic scaling of theoffspring seemed more thick and solid than observed in regular metallicfish. This characteristic was also passed on to the offspring when aheterozygous "dragon" was crossed to a heterozygous metallic fish (BT271107
). The offspring resulting from crossing two heterozygous metallic sibblings (BT171007B
)resulted in homozygous, heterozygous and non-metallic offspring withoutthis unique appearance which was observed before. Off course we have tokeep in mind that these findings were purely based on macroscopicobservations (with the naked eye). Nevertheless, these crosses showthat the "dragon" trait is hereditable and suggests that it behavesquite dominant as it already clearly affects the iridescent layer inheterozygous fish. Pedigree "dragon" x metallic experiment by Joep van Esch
In the breeding reports of Pichet and Victoria Parnell-Starkoutcrossing a metallic fish to Betta sp. mahachai seems to have beenthe key to the thick shiny armor-like metallic scaling of "dragons".Dr. Leo Buss has reported that domesticated metallic bettas andwildtype Betta imbellis possess yellow reflecting iridophores  butwhether this also hold true for Betta smaragdina and Betta sp. mahachaito our knowledge is unknown. It is known that it is possible thatseveral variants of a certain gene (polymorphisms) can exist within apopulation of a certain species. As Betta splendens, Betta imbellis,Betta smaragdina and Betta sp. mahachai all share a common ancestor,there is a chance such gene variants also exist among these(sub)species and that they were combined by crossing the different(sub)species. Which gene(s) and/or gene variant(s) exactly areresponsible for the unique appearance of "dragons" is difficult toanswer without any microscopic examination of the chromatophoresthemselves and a determined breeding program.Do blue "dragons" exist?
The characteristic thick, silvery/white metallic layer on the body istypically found red, yellow, orange, white and black "dragons". In caseof red, yellow and orange "dragons" this layer can be observed in bothlight- and darkbodied fish.
Interestingly, so far this characteristic silvery/white metallic layerhas not been observed in darkbodied bettas with a classical iridescentcolor (steel blue, turquoise or royal blue) of the body and finnage.The picture shows an imaginary example of what the authors mean withthis.
In case of a dark bodied "dragons" with a classical iridescent base,the silvery/white appearance seems to be masked although based onmacroscopic observation (with the naked eye) their scaling definitelyseems to differ from a regular metallic fish by being more solid. Therehave been reports of red and black dragons with a blueish shine ontheir scaling but so far no real blue "dragons" with this typicalthick, silvery/white scaling have been developed. It would beinteresting to find out whether it is possible to obtain a the thick,silvery/white metallic layer back after crossing a blue or copper"dragon" as depicted above to a red, yellow, orange or black betta.Imaginary example of a blue "dragon" Blue "dragon" Copper "dragon" Is it possible to breed non-masked "dragons"?
In general the "dragon" phenotype seems to be associated with theiridescent layer covering the head, the so-called mask trait. It isknown that the metallic and mask trait are caused by seperate genes andcan work independently. With the exception of a few marble "dragons",so far to our knowledge no non-masked "dragons" have been developed butoff course this may change in the future. To conclude.......
Although the "dragon" phenotype has been introduced into our hobby forabout 5 years ago, its possibilities still seem unlimited and new colorcombinations and patterns are continuously being developed. But therestill are a lot of questions to be answered. It still is not clear howwe should define the true "dragon". In order to answer this much moreresearch has to be done. A detailed microscopic examination of thechromatophores, as previously done with the yellow reflectingiridophore [15-18], in combination with a determined breeding programmight provide more information with respect to which gene(s) and/orgene variant(s) are responsible for the unique appearance andinheritance of the "dragon" phenotype. Welcome in a new era in the world of "bling-bling" bettas!!Red "dragon" variant Acknowledgements:
We would like to thank Dr. Leo Buss (USA), Stefan Psarakos (Australia)and Ursula Bosnjak (Austria) for their suggestions and critical reviewof this article. References/credits:
1. Parnell-Stark, V., "Copper gold", 2003.
2. Parnell-Stark, V., "Metallics and Masks", 2005.
3. van Esch, J.H.M., "Understanding metallic genetics", 2006 and published in FLARE magazine, Vol. 41, No. 6., 2008.
4. Bred by Tada palakul (Smilebetta) - Thailand.
5. Bred by Sanya Ponpal (Interbettas) - Thailand.
6. Bred by Rung Keereelang (Banleangbettas), owned by Robi Iskandar (Robi2000) - Singapore.
7. Bred by Robi Iskandar (Robi2000) - Singapore.
8. Pichet (Interfish Breeder Team - Thailand), personal communication through email, 2008.
9. Fancy Fish Magazine, Thailand, Vol. 5, No. 50, 2008.
10. Maurus, W., "Bettas a complete introduction", ISBN 0-86622-288-x, 1981.
11. Parnell-Stark, V. (USA) www.bettysplendens.com
12. Panitvong, N., International Betta Congress (IBC) - Species Maintainance Program (SMP).
13. PIBK2005, Malaysia, 2006.
14. Bred by Vu Minh Chuong - Vietnam; www.bettasaigon.com
15. Buss, Leo W., "Structural Color", Bettas&More - FAMA, March 2005.
16. Buss, Leo W., "A 'New' Iridophore Color", Bettas&More - FAMA, September 2005.
17. Buss, Leo W., "Inheritance of Metallic Trait", Bettas&More - FAMA, November 2005.
18. Buss, Leo W., "Naming the new metallic gene", Bettas&More - FAMA, March 2006.
19. Picture adapted by Joep van Esch, based on a non-metallic royal HM PK from BT161209.
20. Bred by Le Anh Tuan - Vietnam; www.bettasaigon.com
21. Bred by Kiat Inthamu (Muangkhonbettas) - Thailand