Modeling the Potential Global Distribution of Honeybee Pest, Galleria mellonella under Changing Climate

Abstract

Beekeeping is essential for the global food supply, yet honeybee health and hive numbers are increasingly threatened by habitat alteration, climate change, agrochemical overuse, pathogens, diseases, and insect pests. However, pests and diseases that have unknown spatial distribution and influences are blamed for diminishing honeybee colonies over the world. The greater wax moth (GWM), Galleria mellonella, is a pervasive pest of the honeybee, Apis mellifera. It has an international distribution that causes severe loss to the beekeeping industry. The GWM larvae burrow into the edge of unsealed cells that have pollen, bee brood, and honey through to the midrib of the wax comb. Burrowing larvae leave behind masses of webs that cause honey to leak out and entangle emerging bees, resulting in death by starvation, a phenomenon called galleriasis. In this study, the maximum entropy algorithm implemented in (Maxent) model was used to predict the global spatial distribution of GWM throughout the world. Two representative concentration pathways (RCPs) 2.6 and 8.5 of three global climate models (GCMs), were used to forecast the global distribution of GWM in 2050 and 2070. The Maxent models for GWM provided a high value of the Area Under Curve equal to 0.8 ± 0.001, which was a satisfactory result. Furthermore, True Skilled Statistics assured the perfection of the resultant models with a value equal to 0.7. These values indicated a significant correlation between the models and the ecology of the pest species. The models also showed a very high habitat suitability for the GWM in hot-spot honey exporting and importing countries. Furthermore, we extrapolated the economic impact of such pests in both feral and wild honeybee populations and consequently the global market of the honeybee industry.

Advances and knowledge gaps on climate change impacts on honey bees and beekeeping: A systematic review

“…Contrarily, a decrease in precipitations can affect the behavior (e.g., early appearance dates; Gordo et al., 2010), reduce honey production and exportations (Gajardo‐Rojas et al., 2022; Gounari et al., 2022), and reduce plant–pollinator interactions (e.g., frequency visiting; Jaworski et al., 2022; Maluf et al., 2022). Furthermore, changes in precipitation patterns are expected to affect the distribution of different pests (Hosni et al., 2022; Tihelka et al., 2021) and diseases (Giliba et al., 2020), as well as produce changes in the geographic distribution of different lineages of A. mellifera (Canovás et al., 2014).…”

Evaluation of Climate Change Impacts on the Global Distribution of the Calliphorid Fly Chrysomya albiceps Using GIS

“…Two modeling techniques (Maxent and Clim model incorporated in DIVA-GIS) were used to predict the current suitable habitat of C. albiceps depending on the climatological parameters. The two techniques are among the most frequently used GIS methods in studying biological units through several previous studies [39,40]. Both are simple, effective, and easy to use, but Maxent is better and widely used [23].…”

Fungus under a Changing Climate: Modeling the Current and Future Global Distribution of Fusarium oxysporum Using Geographical Information System Data

“…For future predictions, parallel datasets of bioclimate covariates were downloaded for two representative concentration pathways (RCP 2.6 and 8.5) covering the two time periods of 2050 and 2070 ( (accessed on 18 November 2021). These future bioclimatic data layers were also converted to ASCII format via ArcGIS v10.7 and used for future projections in Maxent [ 31 , 32 , 33 ].…”

“…In addition, some simple analyses were performed in DIVA-GIS software V7.5 including the envelope test between Bio 1 and Bio 12 and the histogram of annual mean temperature [ 32 ]. Both used presence-only data to predict the species distributions at pseudo-absence points [ 31 ]. The following settings were used for the Maxent model: output format = logistic, random test percentage = 25, regularization multiplier = 1, maximum iterations = 10,000, convergence threshold = 0.0001, and maximum number of background points (as pseudo-absent points) = 10,000.…”