Talk:Local adaptation

Talk

Discussion for: Local adaptation • View • Links • Related changes

Use this page for verification notes, disputes, and action items. Keep narrative content on the article.

Page scope

Local adaptation covers what local adaptation is, how it is shown by transplants and how far genome scans go, then what has been measured in cannabis. The process of selection, with the balance of selection against migration, belongs to Natural selection; any difference among genotypes in their response to the environment, to Genotype-by-environment interaction; the response of one genotype, to Phenotypic plasticity; the plant's responses to climate as botany, to Climate adaptation in Cannabis and Environmental adaptation.

Data verification

Record source conflicts, figures that differ between studies and sources not used here. Keep narrative content on the article.

Discussion

Sources

Kawecki and Ebert (2004), Ecology Letters 7: 1225–1241: reciprocal transplant and common garden experiments, deme by habitat interaction for fitness.
Leimu and Fischer (2008), PLOS ONE 3: e4010: a meta-analysis of reciprocal transplants in plants; the strict sense; large and small populations; distance.
Hereford (2009), The American Naturalist 173: 579–588: the frequency and size of the native advantage; weak trade-offs.
Ågren and Schemske (2012), New Phytologist 194: 1112–1122: five years of transplants between Sweden and Italy in Arabidopsis thaliana.
Mercer, Martínez-Vásquez and Perales (2008), Evolutionary Applications 1: 489–500: maize landraces in common gardens at two altitudes in Chiapas.
Anderson and colleagues (2013), Molecular Ecology 22: 699–708, and Ågren and colleagues (2013), PNAS 110: 21077–21082: antagonistic pleiotropy, conditional neutrality and fitness loci.
Hoban and colleagues (2016), The American Naturalist 188: 379–397, and Lotterhos and Whitlock (2015), Molecular Ecology 24: 1031–1046: genome scans and their pitfalls.
Zhang and colleagues (2018), Frontiers in Plant Science 9: 1876; Chen and colleagues (2022), BMC Plant Biology 22: 371; Babaei and Torkamaneh (2026), The Plant Genome 19: e70243: origins grown at one site.
Tang and colleagues (2016), Industrial Crops and Products 87: 33–44: cultivars at four European sites.
McPartland and Small (2020), PhytoKeys 144: 81–112: Central and South Asian populations.
Woods and colleagues (2023), G3 13: jkac209; Halpin-McCormick and colleagues (2026), Journal of Heredity 117: 24–38: genome scans and environmental genomic selection.
Giupponi and colleagues (2020), Molecules 25: 1381: clones of one hemp cultivar at two elevations.
Casañas and colleagues (2017), Frontiers in Plant Science 8: 145: the definition of a landrace.

Where the sources disagree

How common local adaptation is. Leimu and Fischer and Hereford both report 0.71 for a comparison at one site. Leimu and Fischer conclude that local adaptation is less common than generally assumed, weighing the 45.3 per cent of pairs that meet the strict sense. Hereford concludes that populations are often locally adapted. Hoban and colleagues cite both for local adaptation being common.
Distance. Leimu and Fischer find no effect of the distance between sites. Hereford finds that the size of local adaptation rises with the difference between home environments. The two measure different things: Leimu and Fischer write that the effect of distance depends on how far it goes with environmental difference.
The count in Leimu and Fischer. The abstract gives 45.3 per cent of "the 1032 compared population pairs". The text uses 1,032 for comparisons of local and foreign plants at one site, and its test of the crossing and non-crossing cases has N = 205.
Trade-offs. Hereford finds weak trade-offs. Ågren and colleagues report a third of fitness loci with trade-offs and call them common. Anderson and colleagues report more conditional neutrality than antagonistic pleiotropy across the genome and describe a bias against detecting trade-offs.
Babaei and colleagues. Babaei and colleagues (2024, BMC Plant Biology 24: 151) write that Iranian populations have "acclimated" to their environments. Babaei and Torkamaneh (2026) write that the plateau "fostered local adaptation" and that seed sourcing ensured "regional adaptation". Both studies grew the plants in one greenhouse at Mashhad.

How far the evidence goes

No cited cannabis study grows populations at one another's home sites and compares fitness.
Zhang and colleagues' title speaks of latitudinal adaptation. The trial is one season at one site, Kunming, and the measures are phenology and form. Their chloroplast pattern fits both climate and distance: Mantel tests give r = 0.379 for geographic distance and r = 0.348 for latitude across the 43 populations, and mean day length is the strongest climatic factor for the lineage frequencies (r² = 0.6024).
Chen and colleagues' statement on moving varieties north or south is a general one, cited to earlier work. It is not a result of their trials.
McPartland and Small's statement on Central Asian plants moved to South Asia carries a citation for the comparison of fungal tolerance. The statement on snapped branches has none, and the paper describes the divergence as what "may be presumed".
Woods and colleagues word their reading as a hypothesis. The positive selection reported at SRFR1 rests on a derived allele at high frequency.
Halpin-McCormick and colleagues' values predict climate from genotype. A preprint by Halpin-McCormick, Aina, Kantar and Ellison (2026, bioRxiv, doi 10.64898/2026.02.05.704066, not peer reviewed) extends the method to 909 samples with feral North American populations, reports associations with temperature and rainfall, and compares predictions with growth: partial overlap with height in one field trial in Wisconsin, and more varied growth among seven feral accessions at 35 °C than at 25 °C over two weeks as seedlings. The authors call their values "partially validated", and say that high prediction accuracies in one data set are likely inflated by the few collection points.
Giupponi and colleagues' clones share one genotype and show a response to site. The design compares no populations from different altitudes.
The maize study is the crop comparison on altitude. Its authors describe the results as preliminary, from one season and two gardens, and cannot rule out maternal effects because seed came from farmers' fields.
Hemp trials measure yield and quality in cultivars bred for many purposes. Seed production at a northern site is a part of fitness in one direction only.

Terms

Local against foreign: the comparison of local and foreign plants at one site, and in the strict sense at both sites of a pair.
Reaction norm, or norm of reaction: the phenotype of one genotype or population as a function of the environment.
Antagonistic pleiotropy and conditional neutrality: alleles that reverse rank between environments, and alleles favoured in one and neutral in the other.
Genomic estimated adaptive value (GEAV): the predicted value of a climate variable for a genotype.
Landrace in the tables of Zhang and of Chen: "domesticated, locally adapted, traditional variety".

Further figures from the sources

Mercer and colleagues (2008): the proportion of plants producing good seed was about 88 per cent for lowland landraces in both gardens, 57 and 68 per cent for midland landraces in the midland and highland gardens, and about 24 per cent for highland landraces in the midland garden. Good seed per plant was about 164 g for highland and midland landraces against 122 g for lowland landraces in the highlands.
Giupponi and colleagues (2020): four pairs of clones at Alagna Valsesia and in the province of Vercelli, grown in the 2018 season.
Babaei and Torkamaneh (2026): nucleotide diversity 8.68 × 10⁻⁴ in the warmer group, 7.66 × 10⁻⁴ in the colder group and 3.98 × 10⁻⁴ in the group found in all five zones.