Leptocoma zeylonica
Silhouette: Sheatherius · PhyloPic
In 3 days ~90% in a week >90%
The purple-rumped sunbird (Leptocoma zeylonica) is a sunbird endemic to the Indian Subcontinent. Like other sunbirds, they are small in size, feeding mainly on nectar but sometimes take insects, particularly when feeding young. They can hover for short durations but usually perch to lap nectar from flowers. They build a hanging pouch nest made up of cobwebs, lichens and plant material. Males are contrastingly coloured but females are olive above and yellow to buff below. Males are easily distinguished from the purple sunbird by the light coloured underside while females can be told apart from males by their whitish throats.
Purple-rumped sunbirds are tiny, at less than 10 centimetres (4 in) long. They have medium-length thin down-curved bills and brush-tipped tubular tongues, both adaptations for nectar feeding. Purple-rumped sunbirds are sexually dimorphic. The males have a dark maroon upperside with a blue-green crown that glistens at some angles, bright green shoulder patch and violet/purple rump patch which is generally hidden under the wings. The underparts are whitish with dark throat, maroon breast band and purple/violet patch in the throat which is visible... Wikipedia →
| # | Park | Country | Daily sighting probability |
|---|---|---|---|
| 1 | Yala National Park | 🇱🇰 Sri Lanka | |
| 2 | Udawalawe National Park | 🇱🇰 Sri Lanka | |
| 3 | Minneriya National Park | 🇱🇰 Sri Lanka | |
| 4 | Sinharaja Forest Reserve | 🇱🇰 Sri Lanka | |
| 5 | Bandipur National Park | 🇮🇳 India | |
| 6 | Nagarhole National Park | 🇮🇳 India | |
| 7 | Wilpattu National Park | 🇱🇰 Sri Lanka | |
| 8 | Mudumalai National Park | 🇮🇳 India | |
| 9 | Sundarbans National Park | 🇮🇳 India | |
| 10 | Periyar National Park | 🇮🇳 India | |
| 11 | Satpura National Park | 🇮🇳 India | |
| 12 | Horton Plains National Park | 🇱🇰 Sri Lanka | |
| 13 | Eravikulam National Park | 🇮🇳 India | |
| 14 | Kanha National Park | 🇮🇳 India |
| # | City | Country | Daily sighting probability |
|---|---|---|---|
| 1 | Bengaluru | 🇮🇳 India | |
| 2 | Mysuru | 🇮🇳 India | |
| 3 | Madurai | 🇮🇳 India | |
| 4 | Thrissur | 🇮🇳 India | |
| 5 | Vijayawada | 🇮🇳 India | |
| 6 | Kalyan-Dombivli | 🇮🇳 India | |
| 7 | Thiruvananthapuram | 🇮🇳 India | |
| 8 | Bhubaneshwar | 🇮🇳 India | |
| 9 | Mumbai | 🇮🇳 India | |
| 10 | Colombo [Sri Jayawardenepura Kotte] | 🇱🇰 Sri Lanka |
Likely
About 60% of wildlife watchers report a Purple-rumped Sunbird sighting in Udawalawe National Park on a given day, based on 291 verified sightings from GBIF citizen-science data from 2019 to 2026. It is the 21st most-seen of the 166 birds we can rate there. Over three days the chance rises to about 90%, and over a week to over 90%, if each day is a fresh chance.
July
In that month about 70% of wildlife watchers report a Purple-rumped Sunbird in Udawalawe National Park on a given day, against about 40% in September and October, the lowest months. That is one of Udawalawe National Park's driest months, with about 50 mm of rain against 330 mm in November.
Between 7 and 10 am
Across all the parks we cover, that three-hour window holds the most Purple-rumped Sunbird sightings with a recorded time, and about 9 in 10 of them come between 6 am and 6 pm: it is a daytime animal.
Among the best
Of the 14 parks where we can rate Purple-rumped Sunbird sightings, Udawalawe National Park ranks 2nd at about 60%, level with the leaders once rounded. The best odds are in Yala National Park (Sri Lanka, about 60%). All parks ranked.
Same family: Nectariniidae
The closest relatives of the Purple-rumped Sunbird that you can see in Udawalawe National Park (family Nectariniidae, Sunbirds and Spiderhunters) are Purple Sunbird (about 40%) and Long-billed Sunbird (about 20%), with each one's daily chance of a sighting.
GBIF citizen-science records
It is estimated from how often the Purple-rumped Sunbird is recorded against all the birds logged in Udawalawe National Park (291 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.
