Responses of bees to habitat loss in fragmented landscapes of Brazilian Atlantic Rainforest

• Published in: Landscape ecology Vol. 30; no. 10; pp. 2067 - 2078
• Main Authors: Ferreira, Patrícia A, Boscolo, Danilo, Carvalheiro, Luísa G, Biesmeijer, Jacobus C, Rocha, Pedro L. B, Viana, Blandina F
• Format: Journal Article
• Language: English
• Published: Dordrecht Springer Netherlands 01.12.2015; Springer Nature B.V
• Subjects: Angiospermae; Angiosperms; Apoidea; Bees; Biomedical and Life Sciences; Brazil; Conservation practices; Ecology; Environmental Management; Forests; habitat destruction; Habitat loss; Habitats; Heterogeneity; Landscape architecture; Landscape Ecology; Landscape/Regional and Urban Planning; landscapes; Life Sciences; Magnoliophyta; Nature Conservation; Nesting; nests; planning; Pollinators; rain forests; Rainforests; Regional analysis; Research Article; Sexual reproduction; social behavior; Sustainable Development; uncertainty; Understory; Brazil; Atlantic Forest; Multiscalar approach; Pollinators; Landscape changes; Brazil; Tropical; Bahia; Forest understory
• ISSN: 0921-2973; 1572-9761
• DOI: 10.1007/s10980-015-0231-3

CONTEXT: Loss of natural habitat can isolate pollinator populations and negatively affect sexual reproduction of animal-pollinated plants. OBJECTIVE: We evaluated how the loss of natural forest affects pollinator diversity in the understory of the Atlantic Rainforest in Northeastern Brazil. METHODS: We focused on bees, the main group of pollinators for angiosperms. We assessed how changes in forest cover at regional (36 km²) and local (0.36 km²) scales affect bee richness and abundance. RESULTS: We sampled 492 bees from 59 species, of which 58 % were above ground nesting species and 73 % exhibited some level of sociality. Our results show that the loss of forest had negative effects on understory bee abundance, which was particularly accentuated for species that nest above ground. However, for social bees the effect of changes in forest cover at a local scale depended on regional forest cover, negative effects being only detected when landscapes had at least 35 % of forest. For bee richness, the null model was among the best models bringing considerable uncertainty about landscape effects on bee richness. CONCLUSIONS: These findings suggest that management strategies and conservation practices must integrate proper actions that consider both local and regional scales. For existing fragmented landscapes, it is important to increase forest availability at the regional scale, while also maintaining high environmental heterogeneity at the local scale. We believe that with proper landscape planning this multiscalar approach can be not only more effective, but also easier to implement.