Skip to content

Strategy for eradication of marble trait on bettas phenotype



Have you seen bettas that have colors and color distribution that change throughout their lives?

This is caused by the marble gene when present in the DNA of these specimens. We will deal with this subject in this article, its problem in fixing lineages in greenhouses of serious breeders and present a proposal aiming at its eradication.

Marble trait: synonymous with instability in specimens of bettas

The genotype of the marble trait, when present in the DNA of Betta, results in the instability of its colors and color distribution, that is, they will randomly change and mischaracterize all the work done by the breeder in order to obtain an already established lineage or even a new lineage.

The following photo shows the presence of this gene in the betta phenotype, see:

The Betta fish marketed today, or simply Betta, kept in aquariums in the most diverse places in the world, is a viable and fertile hybrid of the cross between different pure species belonging to the Splendens species group (among them Betta splendens, Betta imbellis, Betta smaragdina, Betta stiktus and Betta mahachaiensis). From those mating between these species, many different characteristics have emerged in the DNA of these hybrids that we have today.


A quick reflection on what it is like in nature: the absence of the marble trace

In its habitat, the betta does not have marbling – presence of the marble trace. This suggests that individuals with this trait, when present, have a lower chance of survival compared to predators. Therefore, they will have little chance of breeding, which would result in the perpetuation of the marble gene.

Still on their habitat, it is known that these ecosystems have a dark bottom, with tea-colored waters, the result of many leaves and branches of plants on the bottom, or even slightly muddy waters.

The action of the marble gene decreases the betta’s natural camouflage against predators, revealing more easily its position in its habitat.

One speculation that can be created is that, due to natural selection, these bettas cease to exist precisely because of the lack of protection caused by the light colors in their phenotype.

However, in the creation of bettas in captivity, these individuals with the presence of the marble gene became supposedly exotic, exhibiting colors and color arrangements never seen before. Therefore, we are of the opinion that this attractive trait has become commercially popular among breeders, and this gene has been fixed.

Therefore, it is a very common trait today, but unwanted by every breeder who aims to obtain consistent work, where he can guarantee and maintain well-defined phenotypes in subsequent litters.

The need to eliminate the marble trace in the selective creation of bettas

Imagine that you have been improving a line of copper bettas for many years, obtaining specimens without any infiltration of other colors when, suddenly, it is necessary to acquire a specimen from another breeder to avoid the high inbreeding in your bloodline or breeding.

And what is the surprise resulting from the crossing of this new specimen? Individuals marbling all over the phenotype, destroying all that work that was being done by you.

This is precisely what we want to avoid, proposing a method for eliminating this trait in your creation!


Features of the marble trace

1 – the marble line is recessive;

2 – even in genetically marbled bettas, there are cases in which the performance of the marble trait does not seem to occur in the phenotype, as it is associated with a factor related to chronological time, the betta’s lifetime, which is defined in the DNA of each specimen.

For example, there are bettas that are already born with the marbled phenotype (time factor = 0), being totally colorless – the so-called cellophane.

On the other hand, there are bettas that have the marble gene, but their phenotype does not show the trait (time factor = 4 years, for example).

That is, even though the betta is genetically marble, it may not exhibit any marbling in the phenotype. To these exemplary marbles, I call hidden marbles.

3 – in the same litter there is the possibility of bettas marbling in different periods.


Simplification of marbling conditions

Practice has shown that, in most cases, the trait manifests itself in litters up to six months of age. That is, after this period, those who did not marbling, will no longer marbling. And indeed, the occurrence of “hidden marbles” is rarer.

As a direct consequence of this simplification, I recommend that the choice of matrices should run after the individuals are six months old for future matings.


Probabilistic modeling and calculations

If the trait is recessive, then we have:

Once the above condition (> 6 months of age in selection / exchange / acquisition) is respected, we have to:

  • bettas that have not marbled after six months of life will not marbled until the end of their life. In other words, the trait “hidden marble” – as we are naming it – is rare, in addition to not being possible to identify it. Therefore, for practical purposes, we will start from the premise that hidden marble will not occur; it is not detectable (gene mb¹mb¹, so called by us);
  • non-marbled bettas will have one of these genotypes: MbMb or Mbmb;
  • marbled bettas, which will not be selected, will have mbmb genotype for this trait;

An important note: from the point of view of this strategy, as well as due to the conditions mentioned above, we will consider the notation mbmb = mb¹mb¹.

The marble trait eradication strategy in practice

We will have MbMb with the desired genotype, which indicates a homozygous betta free of the marble line, where under no circumstances will it transmit marbling to subsequent generations.

However, only by contemplating the phenotypes, we will not be able to know which individual is a carrier and which is not.

From the point of view of phenotypes, MbMb = Mbmb.

As we are starting a selection work for the eradication of the marble trace of a strain, we do not know which initial genotypes may be present in the matrices, that is, we do not know if the couple is (Mb Mb x Mb Mb), (Mb Mb x Mb mb), or (Mb mb x Mb mb).

Note: we remember again that the possibility of having one of the matrices (or both) with the mbmb genotype was not considered, since the couple will be formed after six months. Bettas with this genotype would marble up to that date, and in this strategy we will not have “hidden marbles” (mb¹ mb¹).

It is concluded that, if the F1 juveniles do not marbled up to the age of six months, according to our initial premises, the couple will have the MbMb x MbMb, MbMb x Mbmb genotypes, and the fry will not produce mbmb, which is the betta with possibility of marbling.

But what would be the probability of obtaining couples with the aforementioned genotypes?

We will start by calculating the probabilities of obtaining the MbMb and Mbmb genotypes.

Each genotype can come from any of the three possible combinations of couples formed initially, namely:

Note that these couples are the ones that individuals do not marble.

From these couples, we obtain:

MbMb = 7/12 = 58,3%

Mbmb = 4/12 = 33,3%

mbmb = 1/12 = 8,3%


Therefore, for us to have a Mb Mb x Mb Mb couple we will have the theoretical probability of 7/12 x 7/12 = 49/144. On the other hand, for a MbMb x Mbmb couple the probability is 7/12 x 4/12 = 28/144.

In other words, the theoretical probability of forming a couple free of the marble line, which is what is desired by those who develop and/or improve lines with well-defined color distribution patterns, is 1.75 times greater than that of forming a couple with one of the breeders carrying the trait.

In practical terms, the real probabilities of occurrences of each genotype will be increasingly closer to the theoretical ones as the size of the litters increases in number of individuals. For small litters, the actual occurrence values for each genotype will be quite different from the theoretical values shown above.

Therefore, we chose an alternative path.


Structuring a practical management proposal for the eradication of the marble trait

In this flowchart, the “matrix” may be MbMb or Mbmb, as it does not have marbling.

And the flowchart shows that, if the “matrix” is Mbmb, when crossed with any marble betta (therefore, mbmb), marbled individuals will appear (mbmb).

Imagine a pair of breeders A and B selected within the lineage we are working on, e.g. copper.

As a result, in F1 we ​​obtained copper bettas, some marbling and others not.

We have that marbled bettas have the genotype (mbmb), and non-marbled bettas (MbMb, or Mbmb, or mb¹mb¹).

But, how to know which ones have the MbMb genotype?

We wait for the F1 litter (of the couple A x B) to develop for up to six months, we separate the F1 matrixes that did not marbled and that pleased us from the point of view of the lineage we are working on. Remember that our main objective is to improve a copper line, and we take these selected matrices (which may be MbMb, or Mbmb, or mb¹mb¹), and cross them with marble bettas. That is:

Those matrices crossed with marbles, and which do not have marbled bettas in their litters, will be either MbMb or mb¹mb¹.

However, as the occurrence of “hidden marbles” is very small, now imagine a copper selected for the improvement of our lineage, being crossed with a marble, and obtaining a litter in which only “hidden marbles” occur.

The practical evidence has shown that the probability of this event occurring (only “hidden marbles”) is so low that it is negligible.

Therefore, from now on we can consider that a matrix, when crossed with any marbled betta, and not producing marbles, will be considered MbMb.

Then we can update the flowchart shown earlier:

We remember that only for the first generations will we pay the price of working with older matrices guaranteed against this problem. The performance of this trait in lines that need to be well defined in terms of colors and color distribution must be eradicated.

In conclusion, I reinforce the importance of using techniques for the correct selection of matrices in your breeding stock.

Observation and modeling of the phenotypes found in the litters is essential.

Good luck with your Bettas!

(c) 2014 - 2022 Betta Project - A DNA Experience. All rights reserved