AdaptFit: will it thrive here?

Pick places and a livestock type or crop. Get a clear verdict, test what changes it, and take a one-page brief to your next meeting. Climate, phenotype and genome tools sit behind it for researchers. Free, no sign-up, no keys. Uploaded files stay in your browser.

Up to 5 places can be compared on the Decide tab.

Default: Addis Ababa. Search or type coordinates, then add the place to your list.

1. What are you deciding about?

Verdict bands (advanced)

2. Which places?

Use the place box above, then press "Add this place to my list". Add up to 5.

Research tool: does the trait drop when it gets hot?

Upload a CSV with one row per record: a date and a trait value, plus optionally latitude and longitude (or a site name) and a group such as breed. The app looks up the daily climate for each site and date, then fits trait against THI.

Research tool: which variants track climate across populations?

Upload a CSV with one row per population: population, lat, lon, then one column per variant containing the allele frequency (0 to 1). The app pulls 2001–2020 climate for each population and ranks variant–climate correlations (Spearman).

Research tool: look up a gene

Starter candidate lists from the literature. They are not exhaustive; use the papers panel to check the evidence in your species.

How it works

  • Decide verdict: for each place, the share of days in your season at or above the stress threshold (livestock: daily mean THI; crops: daily maximum temperature), averaged over the latest 15 years. Good, Watch and Risky use bands you can change. "Change" compares with the first 15 years of the record, and "bad year" is the 80th percentile of yearly values. The sowing scan repeats this for each possible start month. The warming slider adds the same number of degrees to every day, and the tolerance slider raises the threshold. Neither is a climate-model projection.
  • THI (temperature–humidity index, NRC 1971): THI = (1.8·T + 32) − (0.55 − 0.0055·RH)·(1.8·T − 26), T in °C, RH in %. Common cattle categories: below 72 no stress, 72–78 mild, 79–88 moderate, 89–98 severe, 99 and above dangerous. Thresholds differ by breed, species and production level, so the stress threshold is adjustable.
  • Typical thresholds in the presets (cattle THI 72; maize 35 °C, wheat 30 °C, sorghum 38 °C, chickpea 35 °C at sensitive stages) are commonly reported literature values and differ by variety, breed and stage. Set your own for local decisions.
  • Daily and historical THI use daily mean temperature and mean humidity. This smooths out afternoon peaks, so treat day counts as a comparative index, not an exact count of hours above a limit. The forecast uses hourly values.
  • Phenotype × climate: ordinary least squares of trait on THI, overall and per group, with separate slopes below and above the threshold. It does not control for stage of lactation, feeding, management or repeated measures on the same animal. Use it to explore, then confirm with a mixed model.
  • Genome × climate: Spearman correlation between allele frequency and each climate variable, with Benjamini–Hochberg q-values. It ignores population structure, which can create false signals. Treat results as candidates to follow up with methods such as LFMM or BayPass. Fewer than about 10 populations gives unreliable p-values.
  • Candidate gene lists are starting points drawn from published work. They are not a substitute for a literature review.

Privacy

Files you upload are processed in your browser and are never sent to this site. To fetch climate, the app sends coordinates and date ranges to Open-Meteo, and gene symbols to Ensembl, UniProt, STRING and Europe PMC. Nothing else leaves your device.

Data sources and terms

Built by Mulusew Kassa Bitew. Research and education use. Not veterinary, agronomic or investment advice.