Melanis smithiae
In 3 days ~20% in a week ~40%
| # | Park | Country | Daily sighting probability |
|---|---|---|---|
| 1 | Manu National Park | 🇵🇪 Peru |
More parks will be added in future updates.
| # | City | Country | Daily sighting probability |
|---|---|---|---|
| 1 | Santa Cruz de la Sierra | 🇧🇴 Bolivia | |
| 2 | Curitiba | 🇧🇷 Brazil | |
| 3 | Porto Alegre | 🇧🇷 Brazil | |
| 4 | Rio de Janeiro | 🇧🇷 Brazil |
With luck
About 10% of wildlife watchers report a Smith's Pixie sighting in Manu National Park on a given day, based on 7 verified sightings from GBIF citizen-science data from 2019 to 2026. It is the 203rd most-seen of the 366 insects we can rate there. Over three days the chance rises to about 20%, and over a week to about 40%, if each day is a fresh chance.
September
In that month about 20% of wildlife watchers report a Smith's Pixie in Manu National Park on a given day, against under 5% in January–March, August and October–November, the lowest months. That is one of Manu National Park's driest months, with about 90 mm of rain against 290 mm in February.
Between 2 and 5 am
Across all the parks we cover, that three-hour window holds the most Smith's Pixie sightings with a recorded time, and about 6 in 10 of them come between 6 pm and 6 am: it is seen both by day and at night.
Only Manu National Park, in our data
Manu National Park is the only park on Wallace's Line with enough records to rate Smith's Pixie sightings: about 10% of wildlife watchers there report one on a given day.
Same family: Lycaenidae
The closest relatives of the Smith's Pixie that you can see in Manu National Park (family Lycaenidae, Gossamer-winged Butterflies) are Separated Stripestreak (about 10%), Thecla celmus (about 10%) and Thecla alda (about 10%), with each one's daily chance of a sighting.
GBIF citizen-science records
It is estimated from how often the Smith's Pixie is recorded against all the insects logged in Manu National Park (7 verified sightings from 2019 to 2026), calibrated against known sighting rates. Figures are rounded to the nearest 10% because citizen-science data supports honest approximations, not decimals, and months with records from fewer than five people show "—". How it works.
