Answers are grounded in this paper and cite their sources.
Climate-only species distribution models (SDMs) omit the host plants phytophagous pests depend on, yet pairing separately fitted pest and host models brings assumptions of its own about score comparability and threshold transfer. We examined both issues for the mango stem borer, Batocera rufomaculata , using tuned MaxEnt models for the beetle and for two economically important hosts, Mangifera indica and Ficus carica . Of five CMIP6 GCMs screened, the climatically central BCC-CSM2-MR performed well and was used for projections under SSP2-4.5 and SSP5-8.5 in the 2050s and 2070s. Focal-host suitability was the cell-wise maximum of the two host surfaces, combined post hoc with pest suitability by harmonic mean. The pest model showed high spatially structured cross-validated discrimination (mean AUC = 0.883). Niche overlap was lower with F. carica (Schoener’s D = 0.14, Warren’s I = 0.38; P > 0.05), higher with M. indica ( D = 0.28, I = 0.49; P < 0.05), and highest for the pooled-host niche ( D = 0.39, I = 0.62; P < 0.05). Using independently derived 10th-percentile thresholds for the pest and co-suitability surface, current classified area increased from about 37.8 to 48.7 million km 2 (+28.8%) after host integration. However, future change relative to each model’s own current baseline was smaller for the host-integrated surface (+1.6 to +3.1%) than for the climate-only pest model (+3.8 to +5.3%), and centroid shifts were strongly continent-specific (29–483 km). Because independently fitted MaxEnt outputs are not calibration-equivalent, the combined raster should be interpreted as a relative co-suitability index rather than a probability of establishment. The results show that host integration can enlarge classified co-suitable area while dampening projected temporal change, and that the integration rule and threshold must be reported explicitly.
No takes yet. Share an insight, caveat, or question.
Hussain et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: