PREDICTING HISTORIC, CONTEMPORARY, AND FUTURE DISTRIBUTIONS OF CULEX CORONATOR FOLLOWING RAPID RANGE EXPANSION

Technical Summary: Predicting Historic, Contemporary, and Future Distributions of Culex coronator Following Rapid Range Expansion

Problem Statement
Culex coronator, a mosquito vector capable of transmitting pathogens such as West Nile virus and St. Louis encephalitis virus, has undergone a rapid and sustained range expansion within the United States since 2004. Historically restricted to southern Texas, Mexico, and Central/South America, the species has now established populations across the Gulf Coast and into the humid subtropical regions of the eastern U.S., reaching as far north as Virginia. While the expansion is documented, the changes in environmental suitability driving this shift across historical, contemporary, and future climate conditions have not been quantitatively evaluated. Understanding these suitability dynamics is critical for anticipating future distributions to improve vector surveillance and public health risk reduction.

Methodology
The study employed Species Distribution Models (SDMs) to evaluate abiotic suitability for Cx. coronator under three distinct temporal frameworks:

1. Historical Model: Calibrated using occurrence records from 1960–1989 (pre-expansion) and projected onto recent climate conditions (2000–2024) to assess changes in suitability along the Gulf Coast.
2. Contemporary Model: Calibrated using the full dataset of occurrence records from 1960–2024 (including post-expansion data) to predict current suitability and future potential.
3. Future Projections: The contemporary model was projected onto climate conditions for 2041–2060 under the Shared Socioeconomic Pathway 3 (SSP3) mid-range emission scenario.

Data and Calibration:

• Occurrence Data: Georeferenced records were aggregated from the Global Biodiversity Information Facility (GBIF) and scientific literature. Data were cleaned, deduplicated, and spatially thinned to a minimum distance of 0.25 decimal degrees to reduce sampling bias.
• Calibration Region (M): Defined using the Biotic-Abiotic-Movement (BAM) framework. The historical model was constrained to areas accessible prior to 1989 (excluding the Andes and southern South America), while the contemporary model included the expanded range within the humid subtropical zone.
• Environmental Variables: Bioclimatic variables at ~5 km resolution were derived from CRU TS.4.09 and WorldClim v2.1 data. Multicollinearity was assessed using Variance Inflation Factor (VIF), retaining variables with VIF < 5.
• Modeling Algorithm: Models were calibrated using the maxnet algorithm in R (package ENMeval), utilizing linear, quadratic, and product feature classes. Model selection was based on corrected Akaike Information Criterion (AICc), training/testing AUC, and AUC difference. The best-performing models were validated using 20 bootstrap replicates in Maxent.

Key Results

• Historical vs. Recent Suitability: The model calibrated on historical data (1960–1989) predicted only modest increases in suitability along the Gulf Coast when projected to recent conditions (2000–2024). Crucially, this historical model failed to identify large areas of the humid subtropical eastern U.S. that are currently occupied by the species, suggesting that the environmental niche of the expanding population differs from the historical baseline.
• Contemporary Suitability: The contemporary model (1960–2024) predicted widespread suitability across the southern and eastern U.S., including the Gulf Coast, inland Texas, Florida, and extending northward through the Mid-Atlantic to Virginia and Maryland. It also identified high suitability in parts of the Appalachian foothills and inland Kentucky, with suitability declining at higher elevations.
• Future Projections: Under the SSP3 scenario (2041–2060), projections indicated a northward shift in suitability, with increased potential in the northeastern U.S. and higher elevations of the Appalachian Mountains.
• Environmental Constraints: Across all models, suitability remained consistently low in arid and semi-arid regions (e.g., western range limits, parts of the Texas-Louisiana corridor). The contemporary model identified "mean temperature of the driest quarter" as the most significant predictor (43.4% contribution), followed by maximum temperature of the warmest month and precipitation of the driest month. This suggests that moisture availability, rather than temperature alone, is a primary constraint on the species' distribution.

Significance and Claims
The paper claims that relying solely on historical occurrence data to model invasive species can lead to significant underestimation of potential ranges, as the historical model failed to capture the environmental conditions now occupied by the expanding Cx. coronator population. By incorporating updated occurrence records, the contemporary model successfully captured the full breadth of the species' realized niche.

The authors conclude that moisture availability is a key abiotic factor limiting the species, particularly in arid zones, while the ability to overwinter as an adult likely facilitates northward expansion. The study underscores the necessity of updating occurrence data in SDM workflows for invasive mosquito species to strengthen surveillance and control strategies. These models provide a data-driven basis for operational control programs to prepare for potential establishment in new areas and to guide taxonomic identification training for public health personnel. The study does not claim to account for biotic interactions (e.g., competition, predation) or sampling bias entirely, acknowledging these as limitations, but asserts that the updated models offer a robust tool for observing patterns during rapid range expansion.