Invasive species can reach high abundances and dominate native environments. One of the most impressive examples of ecological invasions is the spread of the African sub-species of the honey bee throughout the Americas, starting from its introduction in a single locality in Brazil. The invasive honey bee is expected to more negatively impact bee community abundance and diversity than native dominant species, but this has not been tested previously. We developed a comprehensive and systematic bee sampling scheme, using a protocol deploying 11,520 pan traps across regions and crops for three years in Brazil. We found that invasive honey bees are now the single most dominant bee species. Such dominance has not only negative consequences for abundance and species richness of native bees but also for overall bee abundance (i.e., strong “numerical” effects of honey bees). Contrary to expectations, honey bees did not have stronger negative impacts than other native bees achieving similar levels of dominance (i.e., lack of negative “identity” effects of honey bees). These effects were remarkably consistent across crop species, seasons and years, and were independent from land-use effects. Dominance could be a proxy of bee community degradation and more generally of the severity of ecological invasions.

(1) GENERAL INFORMATION

(1.1) Author Information

    A. Principal Investigator Contact Information
        Name: Lucas Garibaldi
        Institution: Instituto de Investigaciones en Recursos Naturales, Agroecología y Desarrollo Rural (IRNAD)
        Address: Mitre 630, CP 8400, San Carlos de Bariloche, Río Negro, Argentina.
        Email:lgaribaldi@unrn.edu.ar

    B. Associate or Co-investigator Contact Information
        Name: Guaraci Duran Cordeiro
        Institution: University of Salzburg. Department of Biosciences
        Address: Hellbrunnerstr. 34, 5020, Salzburg, Austria
        Email: guaradc@gmail.com

(1.2) Date of data collection: 2010-08-19 - 2014-05-28

(1.3) Geographic location of data collection: Brazilian States - Amazonas, Bahia, Ceará, Goiás, Mato Grosso, Pará, Paraiba, Rio de Janeiro, Rio Grande do Sul  

(1.4) Information about funding sources that supported the collection of the data: This Project was supported by the Global Environment Fund (GEF) and coordinated globally by the United Nations Food and Agriculture Organization (FAO), with support from the United Nations Environment Programme (UNEP). In Brazil it was coordinated by the Ministry of the Environment (MMA), with support of the Brazilian Fund for Biodiversity (FUNBIO). We also appreciate funding from Agencia Nacional de Promoción Científica y Tecnológica (PICT 2015-2333, PICT-2018-00941), Universidad Nacional de Río Negro (PI 40-B-567), Conselho Nacional de Desenvolvimento Científico e Tecnológico (305062/2007-7, 556042/2009-3, 556406/2009-5, 305836/2012-9, 304689/2015-7, 303894/2018-0, 556616/2019-0 and 308358/2019-8), Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (203.321/2017), and the 2017-2018 Belmont Forum and BiodivERsA joint call for research proposals (under the BiodivScen ERA-Net COFUND programme and with the funding organisations AEI, NWO, ECCyT and NSF.

(2) METHODOLOGICAL INFORMATION

(2.1) Description of methods used for collection/generation of data: Bee sampling followed the protocol in LeBuhn et al. (2016). At each site, one plot with 15 pan traps (they were made from 150 ml plastic white bowls of 9 cm diameter painted fluorescent blue, fluorescent yellow or left unpainted [white] and filled with soapy water) was deployed within the crop and another plot with 15 pan traps adjacent in the surrounding area

LeBuhn, G., S. Droege, E. Connor, B. Gemmill-Herren, and N. Azzu. 2016. Protocol to detect and monitor pollinator communities: Guidance for practitioners.

(2.2) Methods for processing the data: The abundance of all bees, the abundance of native bees, and species richness were modeled through a general linear mixed-effects approach in R (version 3.6.3, lme4 package, lmer function, Gaussian error distribution) (Bates et al. 2015, R Core Team 2020) and performed by multi-model inference based on the corrected Akaike’s Information Criterion
(AICc) (Harrison et al. 2018).

Bates, D., M. Maechler, B. Bolker, and S. Walker. 2015. Fitting linear mixed-effects models using lme4. Journal of Statistical Software 67:1–48.
Harrison, X. A., L. Donaldson, M. E. Correa-Cano, J. Evans, D. N. Fisher, C. E. D. Goodwin, B. S. Robinson, D. J. Hodgson, and R. Inger. 2018. A brief introduction to mixed effects modelling and multi-model inference in ecology. PeerJ 2018:1–32.
R Core Team. 2020. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

(2.3) Environmental/experimental conditions: The data were sampled in a large-scale across the highly-heterogeneous environments of Brazil.

(3) DATA-SPECIFIC INFORMATION FOR: [GARIBALDIreadme.txt]

(3.1) Number of variables:   9

(3.2) Number of cases/rows:      768

(3.3). Variable List:
study_system - variable that combined crop species, sampling season and year (categorical)  
total_abundance - total abundance accounted by the most-abundant species (numerical)
native_abundance - abundance of native bees (numerical)   
richness - richness of species (numerical)
dominance - dominance of species (numerical)
honey_bee_dominance - dominance of honey bee; yes: honey bee is most dominant species, no: honey bee is not most dominant species (categorical)
blooming_status - status of blooming during the bee sampling; yes: the crop was flowering, no: the crop was not flowering (categorical)
isolation - distance to natural or semi-natural habitats (numerical)
pan_trap_placement - pantraps were placed inside or outside the crop field (categorical)