The quality-filtered native-record profile for Strelitzia reginae centers on 17.9°C annual mean temperature and 821 mm annual precipitation. Its robust p10–p90 climate envelope is wider than those two medians: the coldest-month minimum spans 6.4–13.9°C, while modeled annual rainfall spans 598–1142 mm.

This is a computed environmental fingerprint built from 79 spatially thinned climate points attached to filtered GBIF records inside the POWO native range of Cape Provinces to KwaZulu-Natal. It describes where retained records occur in long-term gridded climate; it is not an indoor-care prescription or a physiological tolerance test.

SMART CARE Keep every plant on track. Get personalised reminders for water, feeding, and more. Download app

Bird of Paradise Climate DNA at a glance

Signal Computed result
Annual mean temperature median 17.9°C
Coldest-month minimum, p10–p90 6.4–13.9°C
Warmest-month maximum, p10–p90 22.6–28.9°C
Annual precipitation median 821 mm
Precipitation seasonality median 44.6 CV
Evidence 79 climate points; high confidence (83/100)

The source plant is described as a subtropical perennial. That botanical context matters because the same regional rainfall, humidity, or radiation can be experienced very differently by a climber, epiphyte, geophyte, tree, or succulent. The page therefore reports the climate evidence without turning it into a universal potting or watering formula.

Strelitzia reginae Climate DNA percentile ranges Five occurrence-derived climate distributions. Thin lines show p10 to p90, thick green bars show p25 to p75, and dots show medians. Annual mean temperature16.1°Cmedian 17.9°C21.2°CColdest-month minimum6.4°Cmedian 10.5°C13.9°CWarmest-month maximum22.6°Cmedian 25.3°C28.9°CAnnual precipitation598 mmmedian 821 mm1142 mmPrecipitation seasonality22.3 CVmedian 44.6 CV64.5 CV
Thin line: p10–p90. Thick bar: middle 50% (p25–p75). Dot: median. Each row uses its own scale.

What stands out in the computed profile

  • Temperature center: The annual-mean median is 17.9°C, with the middle 50% of sampled points between 16.6 and 19.2°C. In this model that is a mild annual baseline.
  • Cold and heat edges: The robust outer envelope runs from 6.4°C at the p10 coldest-month minimum to 28.9°C at the p90 warmest-month maximum. The 22.5°C span is moderate; it is an occurrence-climate span, not a laboratory tolerance range.
  • Rainfall volume: Median modeled annual precipitation is 821 mm, while the p10–p90 range is 598–1142 mm. The wet end is 1.9 times the dry end, evidence that one annual total would hide real geographic variation.
  • Rainfall timing: The precipitation-seasonality median is 44.6 CV, placing this record set in the fairly even band used by this renderer. That describes variation through the year, not how often to water a pot.
  • Atmosphere and open sky: Relative humidity centers on 60.1%, while solar radiation centers on 16.9 MJ/m²/day. Both are gridded outdoor climate values; neither reports the humidity or light at an indoor leaf.

Together, those observations describe the climate attached to retained records of Strelitzia reginae. They do not establish the species’ physiological limits, and they should not be converted directly into thermostat, humidifier, or watering targets.

Native-record climate envelope

Variable p10 p25 Median p75 p90
Annual mean temperature 16.1°C 16.6°C 17.9°C 19.2°C 21.2°C
Coldest-month minimum 6.4°C 8.0°C 10.5°C 12.6°C 13.9°C
Warmest-month maximum 22.6°C 24.5°C 25.3°C 27.0°C 28.9°C
Annual precipitation 598 mm 656 mm 821 mm 934 mm 1142 mm
Precipitation seasonality 22.3 CV 30.0 CV 44.6 CV 61.0 CV 64.5 CV
Relative humidity 57.1% 58.9% 60.1% 61.4% 62.3%
Solar radiation 16.5 MJ/m²/day 16.7 MJ/m²/day 16.9 MJ/m²/day 18.9 MJ/m²/day 19.3 MJ/m²/day

The outer p10–p90 values are a robust descriptive envelope; the middle p25–p75 band shows where the central half of sampled climate values falls. Neither band is an “ideal range.” Occurrence density, collection effort, native-range boundaries, coordinate quality, and CHELSA grid resolution all shape what appears here.

A recent hourly reference: Canggu, Bali, 2015–2025

The same fixed reference is attached to every profile in this release: 96,432 complete hourly ERA5 rows for temperature, relative humidity, precipitation, and 10-meter wind at -8.65°, 115.13°. Using one place and one period makes cross-plant comparisons consistent. It does not claim that Canggu is native habitat for Strelitzia reginae.

Signal Native p10 / median / p90 Canggu annual p10 / median / p90 Canggu minus native median Canggu median vs native envelope
Annual mean temperature 16.1 / 17.9 / 21.2°C 26.4 / 26.7 / 27.0°C +8.8°C outside
Annual precipitation 598 / 821 / 1142 mm 1236 / 1791 / 2409 mm +970 mm outside
Annual mean relative humidity 57.1 / 60.1 / 62.3% 86.0 / 86.3 / 86.8% +26.2% outside
Strelitzia reginae native-record climate compared with the 2015–2025 Canggu reference Green ranges show the native-record p10 to p90 climate envelope. Orange ranges show the eleven annual Canggu ERA5 values, with a dot at each median. Annual mean temperaturenative median 17.9°CCanggu median 26.7°CAnnual precipitationnative median 821 mmCanggu median 1791 mmAnnual mean humiditynative median 60.1%Canggu median 86.3%
Green: native-record CHELSA p10–p90. Orange: eleven annual Canggu ERA5 values, 2015–2025. Rows use independent scales.

Across the three comparable annual signals, the Canggu median falls inside 0 of 3 native-record p10–p90 envelopes. The Canggu archive’s annual median wind speed is 10.6 km/h; wind is reported as local context rather than matched to the CHELSA profile because the native data contract does not yet include a comparable wind variable.

The eleven annual values preserve year-to-year spread that a single climatological number would hide. Hourly values were aggregated into calendar years before annual percentiles were calculated. This recent reference is useful for comparing an outdoor place with the long-term native-record envelope, but neither layer should be translated directly into an indoor watering schedule.

Occurrence footprint and geographic coverage

Quality-filtered Strelitzia reginae occurrence footprint Equirectangular world grid showing usable spatially thinned GBIF coordinates, deterministically sampled when the set is dense.
Usable native-country GBIF coordinates after spatial thinning. Dense sets are deterministically sampled for legibility; the CHELSA extraction sample is reported separately.

The CHELSA climate-extraction sample covers 45 quarter-degree grid cells across the native-country filter (ZA). Its coordinate bounds run from -34.202° to -26.184° latitude and 18.374° to 32.125° longitude. Bounds show sampled coverage, not the legal or botanical outline of the species’ range.

Filtering stage Records
GBIF records returned before filters 425
Accepted before spatial thinning 132
Usable coordinates after spatial thinning 79
CHELSA climate-extraction points 79

Recorded rejection counts were basis of record: 1, license: 292. The pipeline also constrains records to the POWO native-country list, removes duplicate spatial cells, and retains the source timestamp and record-set hash. A low count can be scientifically honest: it is better to expose weak coverage than to fill the gap with invented precision.

How this profile compares inside its 10-plant cluster

Within this release cluster, Bird of Paradise ranks #8 of 10 for median coldest-month temperature, #8 for annual-mean temperature, #9 for annual precipitation, #6 for precipitation seasonality, and #3 for usable climate-point sample size. Ranks compare only the ten selected profiles and are descriptive, not measures of plant quality or indoor difficulty.

The closest profiles in the same normalized seven-variable comparison are:

The cluster was selected deterministically: the remaining ten profiles after the current 20-plant dataset is split at the median coldest-month temperature. This keeps the collection rule reproducible and prevents an editorial label from overriding the data.

What this can—and cannot—change about care

Native-record climate can improve the questions you ask. For Strelitzia reginae, the moderate rainfall median and fairly even precipitation signal indicate that annual rain volume alone is incomplete. Indoors, substrate structure, pot volume, root density, airflow, temperature, and actual light determine drying. Never convert millimeters of regional rainfall into a weekly watering interval.

The same limit applies to temperature, humidity, and solar radiation. CHELSA describes outdoor long-term grid cells; glass, walls, heating, air conditioning, shade, and distance from a window create a different microclimate. Use this profile to understand the source evidence and to frame observations, then use the Bird of Paradise care guide for practical indoor-care context.

Confidence and limits

Confidence is high (83/100). The score reflects sample size, spatial coverage, and coordinate-precision reporting in this pipeline version. It is not a probability that the numbers are “correct,” and it does not measure how well a particular plant will perform in a particular home.

Important limits remain: GBIF is a record of observations and specimens, not a balanced field survey; a native-country filter is coarser than a verified native-range polygon; historical and cultivated records can be difficult to classify; CHELSA values are modeled long-term normals rather than sensor readings at each plant. The page reports ranges and provenance so those constraints remain visible.

Methodology and sources
  1. Resolve the accepted taxon through the GBIF Species API and retain the taxon key 2763116.
  2. Constrain the occurrence query using the POWO native distribution: Cape Provinces to KwaZulu-Natal.
  3. Filter GBIF occurrence records for usable coordinates, uncertainty, basis of record, establishment signals, duplicate locations, native countries, and compatible CC0/CC-BY licensing.
  4. Spatially thin accepted records at 0.05° before climate extraction so dense collecting areas do not dominate the distributions.
  5. Sample CHELSA-climatologies V2.1 1981-2010 grids using all spatially thinned records and calculate p10, p25, median, p75, and p90 deterministically.
  6. Fetch the fixed Open-Meteo Historical Weather API reference with ERA5 for 2015–2025, keep only complete rows, and aggregate hourly values into monthly and annual summaries. The stored series hash is e37928abf36266ca233a022a395523a2a52a994c6fff201cebbdfa2d296b7ef0.
  7. Persist source URLs, timestamps, rejection counts, record hash 340d35b39e6a4fe8067c64efbd9c5832564f0caddb85c6f06ad24333da689137, and fact hash 26a0ce15b1d59a26be614cd677a635f46ee7dc67197e58b0639db912f5f59956.

Source licenses recorded by the pipeline: occurrences, CC-BY-4.0, CC0-1.0 on retained records; CHELSA climate layers, CC0-1.0; Open-Meteo reference data, CC-BY-4.0. Derived summaries and comparisons are transformations of the cited source data.

Continue from the data

Bring the evidence back to your actual plant

Use KnowYourPlant to identify the species, record changes, and compare this broad climate context with conditions in your home.

Open KnowYourPlant