Homozygous Parent Lines in Cannabis – Sib-Mating, Selfing and the Cost of Inbreeding

In an outcrossing species, true-breeding is not a state you find but one you build over generations. Two routes lead there, and they are regularly confused: mating related individuals, as is customary in any dioecious species, and chemically induced selfing. The second is faster. The first is what the breeding literature works with.

What true-breeding actually means

Homozygosity is not visible, it is measured. In the study that first published the full route in this species, a line counted as functionally homozygous once the share of heterozygous loci across a 1,500-marker panel fell below 3 %. The open-pollinated source accessions sat between 17 and 35 %.

The decline per generation varied so widely between lines that the authors explicitly recommend measuring it rather than deriving it from a generation count. A statement like “sixth generation” describes the effort, not the outcome.

Sib-mating: slower, and that is the point

The standard route in dioecious species is mating between relatives – full sibs, half sibs, biparental inbreeding. Fixation runs more slowly than through selfing, raising time and cost, and precisely that difference in pace is what makes selfing methods look attractive.

The slower pace has a flip side that the pace calculation leaves out. A population-genetic model of recessive lethals shows that at equal levels of inbreeding, sib-mating clears the lethal load more effectively than selfing: because homozygosity rises gradually rather than in one jump, loci interfere less with one another during selection. At comparable inbreeding coefficients the equilibrium count of heterozygous lethals was consistently lower under sib-mating – 7.11 against 9.3 at a moderate mutation rate, 46.8 against 49 at a high one. The inbreeding coefficients of the mating types show how far the steps sit apart: selfing and full-sib mating both give 1/2, half-sib mating 1/8.

The model describes natural populations, not a breeding programme; what transfers is the mechanism, not the number. It is enough, though, to refute the common equation of “faster” with “better”.

What selfing fixes along the way

In a dioecious species, selfing requires induced sex reversion: female plants form fertile male flowers after treatment with silver thiosulfate, male plants form fertile female flowers after treatment with ethephon. This makes a trait part of the selection that has nothing to do with the breeding goal.

The study states this outright for the route through dioecious females: there, the capacity for sex reversion has to be stabilised where it segregates in the material – with a raised risk of unwanted spontaneous male flowers. Reaching homozygosity by selfing therefore means selecting along the tendency toward hermaphroditism, in every generation.

The abnormalities did not stay theoretical. From the fourth selfing generation they appeared in every line; in the line whose seven generations were grown side by side, the proportion of abnormal flowers rose from 0.4 % in the source generation to 13.7 % in the sixth, with hermaphroditic flowers, flowers on internodes, and intermediate anther-style organs. From that point, seed set could only be rescued by treating different sections of the same plant simultaneously with the ethylene promoter and the ethylene inhibitor.

Which selfing route fails least often

Anyone taking that route anyway has four starting forms to choose from, and their failure rates differ sharply. Through dioecious female plants, 8 of 22 attempts failed, as the induced male flowers yielded little pollen in later generations. Through XX monoecious plants it was 4 of 27, and chemical treatment was not always required. Through dioecious male plants, 2 of 12 attempts failed; the ethephon-induced female flowers set seed even when malformed.

One side question was settled along the way: selfing male plants across three lines and more than 500 offspring gave a one-to-one sex ratio. Viable YY plants, which would shortcut to masculinised seed, do not exist by this account.

The doubled haploid shortcut that works in other crops is so far unavailable in cannabis.

What the finished line is good for

Not for cultivation. An inbred line running above ten percent abnormal flowers is unfit as a production plant. Its purpose is the cross: five hybrids from four finished lines flowered earlier and more evenly than their parents, all but one were larger than their direct inbred parents, and seed yield ran between 3.9 and 155 % above the parental mean. For cannabinoids the gain stayed limited: only two of the five hybrid lines had every individual within ±20 % across all measured cannabinoids.

Loss of vigour was not universal – two of four lines fell short of their source accession, two did not – but the most vigorous seedling was carried forward in every round, so the figures reflect selection as well.

What is missing here: solid figures on recurrent selection with progeny testing as a programme in cannabis. The methods are described in general breeding theory, no survey exists for this species, and the selfing data above come from fibre and grain material rather than drug-chemotype lines.