Polyandric Backcross
An attempt to mitigate the potential loss of recessive contributions inherent in observational breeding while simultaneously assessing step-wise male contributions.
“Don’t let perfect be the enemy of the good.”
INTRODUCTION
The backcross method of breeding is a great process for getting an elite clone-only female into seed form. It’s most widely used form can be seen in the cubing method of repeating backcross generations until the resultant progeny have a ~94% genetic contribution of the target plant. Cinderella 99 is a great example of the cubing method.
The backcross process, as has been practically employed in the example of C99, makes the assumption that the successive genetic contribution passes to subsequent generations monolithically. The assumption then concludes that the repeated selection of a male will pollinate the original target female, resulting in progeny that is converging to her phenotype.
While the assumption of monolithic heredity (
the human application would be to assume you would look like your father on the left side and your mother on the right side) seems ignorant of the myriad dynamics of genetic reality, it has utility as a method of herding the genetic range/pool into a purposeful direction (
the target plant).
This ostensible oversimplification is almost mandatory in observational breeding as the selection of the male is based on phenotype instead of genotype. If I can only select based on what I see, then phenotype necessarily becomes the motivation irrespective of the “unseen” alleles. In a lab setting with unlimited technological resources, marker assisted backcrossing and other highly advanced techniques allow for the attempt to include the more nuanced dynamics of heredity. This is not the normal scenario for cannabis breeders, but does not diminish the classical backcross procedure as a powerful breeding tool.
BACKGROUND
The process of backcrossing as practiced, usually calls for the selection of a singular male from the initial F1 generation (
Target female x Outcross Male) that most resembles the target female. This creates a bias for phenotype. A problem can arise if the constituent traits that make up the target female are double recessive. If you select a male based solely on observable phenotype and choose a male whose corresponding trait is heterozygous, then you will have created a scenario in which the targeted double recessive trait of the mother can be “washed out” by the true breeding dominant male (
one could try to recover in a further generation, but this assumes you know it’s required and can select the necessary genotype for this attempt).
Here is an example of a scenario in which the limitation to a single male selection for the next backcross pollination can impact the resultant frequency of a targeted recessive trait:
EXAMPLE:
Let’s assume we have a target female H.A.OG and a male Black Triangle. We would like to backcross with 2 traits in mind, leaf blades and stigma color. After the initial F1 cross, we see that the progeny all have the leaves of the Black Triangle, so we assume the leaf trait is double recessive in the target female (
H.A.OG).
AA = Black Triangle leaves
Aa = Black Triangle leaves
aa = H.A.OG leaves
Now, let’s add a second trait that cannot be observed in the male, stigma color (
pistils).
BB = Black Triangle red stigma
Bb = Black Triangle red stigma
bb = H.A.OG brown stigma
In the scenario where you have only chosen one male based on the observable phenotype, you choose a male with the genotype: AABb (
Black Tri leaves, Black Tri stigma color)
Our genotypes are:
H.A.OG – aabb
Black Triangle –
AABb
We see that the progeny from this cross results in 2 distinct phenotypes suffering from the same issue we had in the previous generation (
remember we already made the F1 and selected a male), namely a “washing out” of the recessive H.A.OG leaves and have a trait, stigma color, which cannot be observed in the males (pic 1). By limiting our choice to one male from this population for the next backcross, we are essentially flipping a coin on the potential frequencies of desirable traits. Since our selections are practically blind (
because we cannot observe the genotype) we are at the mercy of probability.
Let’s see what happens if you selected a male of the AaBb variety (
red boxes) to backcross with our H.A.OG (aabb):
We end up with 25% of the population with the target female phenotype aabb (pic 2). So when you blindly choose a male from this population, you only have a 1 in 4 shot at choosing the “right” male for the next round of pollination.
Now we will see what happens if I had chosen the other phenotype from the previous generation, Aabb (
green boxes) to backcross to the H.A.OG:
In this case, we see that the target female phenotype is 50% of the population (pic 3). Contrast from the other male selection (pic 2), you now have a 1 in 2 shot at choosing the “right” male for the next generation.
The point being that limiting the male selection for the next backcross to one male, can hinder the success rate and total time required to reach the goal of increasing the desired target female traits.
POLYANDRIC BACKCROSS (polyBX)
I endeavor to attempt a modification to the backcrossing method that seeks to manipulate the probabilities such that the inherent blindness to the genotype does not breed one’s self into a corner in the case of singular male selection. I have dubbed this process
Polyandric Backcrossing. "
Polyandric", refers to a female with multiple male mates (
more specific form of polygamy).
Instead of limiting to one male selection for continued backcrosses, multiple males will be chosen to pollinate the target female on different branches and kept separate (
not mixing the pollen of all males). By choosing multiple males, the probability that the selection will include the necessary genetic ingredients to eventually converge on the target phenotype is augmented. In essence, this hedges against a wrong male selection at some point in the extended backcrossing process. The Polyandric Backcross (polyBX) can be thought of as multiple classical single male backcrosses happening concurrently. Since this polyBX process is to have a practical application to the at-home breeder, I will choose 3 males for each backcross generation.
In an effort to better benchmark the contribution of each chosen male, I will flower his daughters while simultaneously pollinating the target female with his son I choose for the next BX generation (pic 4). This allows me to have flowered examples of his contributions to the target female harvested and observed at the time the seeds for the BX are complete and ready to pop for further male selection.
In a scenario where the flowered daughters have shown stark deviations from the female, this line will be discontinued, and only the converging males will continue for further male selections (pic 5). Meaning, I will select my next multiple males only from those male lines that show a converging tendency. This will ensure a benchmarking or testing of each male selection without adding any extra time to this already involved process.
The other advantage to flowering each backcrossed generations’ daughters, is that the extent to which the polyBX generations must continue can be better assessed based on the results, instead of the classical backcross assumption of raw monolith percentage contribution. Where classical BX may do a full cube, the polyBX may only require a couple generations as evidenced by the performance of the flowered daughters.
While cutting down the required BX generations seems promising, it would require saving veg copies of the males in the event you find you’ve “nailed it” at a certain early BX generation. I will instead hybridize the classical and polyBX procedures to set out to do at least a certain number of BX generations (3-6) and save only the last set of selected males. This will allow me to toss earlier males, but keep the final male that gives me the best result for the final parental ingredients (
Target female and polyBX’d male) to make large populations of the “winning” combo.
The final stage of the polyBX process could include filial breeding the final polyBX generation instead of replicating the final BX generation. I believe this decision will become more persuasive to either approach as information is gathered from the process. I can see pros and cons to this approach, but will reserve the decision after having assessed the efficacy of this Polyandric Backcrossing.