Redonda Ecosystem Reserve — Rat and Goat Eradication (Antigua and Barbuda)

Redonda is a small, uninhabited volcanic islet (~150 hectares) belonging to Antigua and Barbuda. Goats were introduced during the 19th-century phosphate-mining era, and black rats established later, stripping vegetation until the island was reduced to bare rock, a 'moonscape'. This pressure from invasive species threatened endemic reptiles, including the Redonda ground lizard and the Redonda ground dragon (anole).

The programme aimed to remove invasive goats and rats from Redonda to halt vegetation loss and allow the islet's ecosystem, including its endemic reptile species, to recover.

Between November 2016 and April 2017, the Environmental Awareness Group (EAG), Fauna & Flora International, New Zealand technical contractor Wildlife Management International, the Government of Antigua and Barbuda, and funders including Re:wild, the Darwin Initiative, USAID and the National Fish and Wildlife Foundation removed about 60 feral goats. Rat eradication using brodifacoum bait ran from February to April 2017, and the island was declared rat-free in July 2018.

A peer-reviewed study in Integrative Zoology (2021) documented Redonda ground lizard density rising from roughly 112–147 individuals per hectare in 2012 to 935.3 individuals per hectare in 2019, while the endemic Redonda ground dragon increased roughly 13-fold since 2017 and invertebrates rose more than fivefold. Vegetation biomass grew by over 2,000%, and plant species richness rose from 17 to about 90 species; fifteen land bird species recolonized the island. On 18 August 2023, Antigua and Barbuda designated the Redonda Ecosystem Reserve, roughly 30,000 hectares of land and sea including a coral reef of about 180 square kilometres, the largest marine protected area in the Eastern Caribbean.

No programme cost figures have been published, land crab recovery was reported only qualitatively without a percentage figure, and long-term funding for continued monitoring is not fully secured. The peer-reviewed study itself flags possible weather and observer bias in its visual-census methods, so results should be read as a strong but imperfect ecological monitoring record.