Breeding claims are only as useful as the identity, population, environment, selection criteria, and records behind them. Genetics explains potential; phenotype records show how that potential was expressed.

Common interpretation trap: Treating a cultivar name, one exceptional plant, or an F-generation label as proof of uniformity, stability, or measured genetic identity.

• Question: What exactly is the target trait?
• Question: How much variation exists in the evaluated population?
• Question: Was the phenotype repeated across siblings, generations, or environments?
• Record: plant and cross identity
• Record: population size
• Record: trait measurements
• Record: selection decisions
• Record: environment during evaluation

Genotype is the inherited genetic constitution of an organism; phenotype is the observable outcome produced by genotype interacting with environment and development. A trait seen in one plant under one set of conditions does not automatically predict how all related plants will perform elsewhere.

Breeding decisions become stronger when traits are defined in measurable terms and evaluated across enough individuals and environments to distinguish repeatable inheritance from temporary environmental effects.

• Define each target trait before selection.
• Record the environment in which a phenotype was measured.
• Avoid treating one exceptional individual as proof of population stability.

Sexual reproduction reshuffles inherited variation. F1, F2, and later filial labels describe generational relationships, not guaranteed uniformity or quality. Segregation can reveal hidden variation in later generations, while selection changes which alleles and trait combinations are carried forward.

Simple Mendelian models are useful teaching tools for some loci, but many cultivation traits are quantitative, polygenic, environment-sensitive, or influenced by developmental timing. Breeding records should reflect that complexity.

• Use generation labels precisely.
• Separate observed ratios from expected textbook ratios.
• Keep population size and selection intensity in the breeding record.

Cannabis commonly shows separate male and female reproductive forms, but sex expression can be more complex. Genetic background, developmental state, and environmental stress may all affect observed expression. A single appearance should be documented carefully rather than generalized to an entire line.

Controlled pollination requires identity control, timing, isolation, labeling, and records. Pollen movement can occur unintentionally, so breeding work must distinguish intended crosses from uncertain seed set.

• Label parents and pollination events at the time they occur.
• Document unexpected sex expression and surrounding conditions.
• Keep uncertain parentage marked as uncertain rather than filling gaps with assumptions.

Selection works on variation that is present in a population. Selecting for one trait can unintentionally shift correlated traits, and severe selection in a very small population can reduce useful diversity. The appropriate population size depends on the breeding goal, trait architecture, available space, and acceptable uncertainty.

Repeatability matters more than attractive labels. Evaluation across siblings, generations, and repeated environments helps show whether a trait is reliably transmitted.

• Record why each plant was kept or removed.
• Preserve raw measurements in addition to final rankings.
• Re-test important traits before describing them as fixed or stable.

Cultivar names, trade names, breeder records, and community histories can be valuable provenance, but they are not the same thing as genomic verification. A responsible record distinguishes source-reported lineage from directly observed traits and laboratory evidence.

Good documentation protects future breeding decisions. Preserve source, date, generation or clone status, parent identifiers, selection notes, photographs, and uncertainty. This makes the breeding story auditable instead of dependent on memory.

• Label source-reported claims as source-reported.
• Maintain unique plant and cross identifiers.
• Keep original records even when a later interpretation changes.

Some traits can be strongly influenced by one or a few loci and may form distinct classes, while many agronomic traits vary continuously because numerous loci and environmental effects contribute to the phenotype. Height, yield, flowering time, architecture, aroma intensity, and stress response are examples of traits that often need quantitative measurement rather than simple present-or-absent scoring.

Heritability describes how much of the observed variation in a particular population and environment is associated with genetic differences. It is not a permanent percentage attached to a cultivar or trait. Changing the population, environment, measurement method, or range of conditions can change the estimate and the response to selection.

• Define whether a trait is scored categorically or measured quantitatively.
• Keep environment and population identity attached to any heritability statement.
• Use replicated measurements for traits that fluctuate strongly with environment.
• Do not translate a heritability estimate directly into certainty about one individual.

Breeding changes allele frequencies through selection, mating structure, and chance. Repeated mating among close relatives can increase homozygosity and expose recessive alleles, while outcrossing can introduce new variation and alter trait combinations. Neither approach is automatically superior; the useful strategy depends on the breeding objective and the genetic material available.

Small populations are especially vulnerable to genetic drift, where allele frequencies change by chance. A severe bottleneck can permanently remove useful variation even when the surviving plants look acceptable. Population size, the number of selected parents, and the contribution of each parent should therefore be part of the breeding record.

• Record the number of candidate and selected parents each generation.
• Track whether one parent contributes disproportionately to the next population.
• Preserve backup seed or clones before narrowing a valuable population aggressively.
• Distinguish intentional selection from variation lost simply because the population was small.

Reciprocal crosses can differ because the seed develops within maternal tissue and because cytoplasmic genomes, seed provisioning, epigenetic state, and maternal environment can influence early phenotype. These effects do not invalidate Mendelian inheritance; they add biological context that can matter when comparing seed lots or reciprocal breeding designs.

Genomic testing can provide evidence about relatedness, identity, diversity, or specific markers, but the conclusion is limited by the markers, reference population, laboratory method, and question being asked. A commercial lineage name is not equivalent to a genomic identity result, and a genomic similarity result does not replace phenotype evaluation.

• Keep the direction of a cross explicit rather than treating reciprocal crosses as identical.
• Record seed-lot harvest date, parent IDs, storage history, and germination performance.
• State exactly what a genetic test measured before using it to support a lineage claim.
• Keep genomic, pedigree, and phenotype evidence as separate but complementary records.