Assessing pollen foraging dynamics for honey bees using DNA metabarcoding and species distribution modelling across six ecosystems in the Amazonas Region, northern Peru

Oyola J.E.; Llaja J.C.; Rojas-Briceño N.B.; Silva-López J.O.; García L.; Veneros J.; Bustamante D.E.; Calderon M.S. · Ecological Genetics and Genomics (2026) · DOI: 10.1016/j.egg.2026.100532

Resumen

Pollen resource availability critically affects honey bee (Apis mellifera) foraging and colony sustainability but remains poorly characterized in neotropical landscapes. Here, we combined DNA metabarcoding (ITS2) with species distribution modelling (MaxEnt) to assess the pollen diets of honey bees across the six heterogeneous ecosystems present in the Amazonas region of Peru: i) low-montane non-cloud forests (Basimontane Yunga), ii) mid-elevation cloud-influenced forests (Montane Yunga), iii) high-elevation pluvial forests (Altimontane Yunga), iv) flood-prone palm swamps, v) anthropogenic secondary regrowth, and vi) high-Andean grasslands/herbaceous vegetation. Analysis of 18 pollen samples revealed exceptionally high pollen richness, identifying 142 taxa from 61 families, with no core pollen types shared regionally, indicating highly localized foraging. Each ecosystem exhibited distinct pollen dominance: Asteraceae in the Altimontane Yunga Forest, Melastomataceae in the Palm Swamp and Montane Yunga Forest, and Chloranthaceae (Hedyosmum) in the secondary vegetation. Species distribution models for five key forage species highlighted restricted, fragmented habitats, with more than 86% of the region being unsuitable. These findings demonstrate that honey bee foraging closely reflects local pollen availability, emphasizing the vulnerability of colonies to habitat fragmentation and the need for targeted landscape-scale conservation of critical floral hotspots.

Instituciones: Universidad Nacional Toribio Rodríguez de Mendoza de Amazonas; Universidad Nacional de Moquegua