Field comparison of BG-Pro configurations with CDC light traps and evaluation of the BG-Counter 2 trap for automated mosquito surveillance in urban and rural Thailand

Entomological surveillance programmes need traps that perform reliably across vector taxa and ecological settings, and automated tools that reduce labour without compromising data quality. We evaluated BG-Pro trap configurations and the BG-Counter 2 against the standard CDC light trap (CDC-LT) across contrasting urban and rural ecologies in Thailand.

We compared CDC light traps (CDC-LT) baited with CO₂, BG-Pro traps (BG-Lure + CO₂), and inverted BG-Pro traps (BG-Lure + CO₂) using a 3 × 3 Latin square design over 270 trap-days across two urban sites (Bangkok and Nakhon Pathom; Nov 2024–Mar 2025) and one rural malaria-endemic site (Tha Song Yang, Tak; Jul — Oct 2024). Trap performance was analysed using generalized linear mixed models (GLMMs) to estimate incidence rate ratios (IRRs). The BG-Counter 2 accuracy was assessed over 60 trap-days by comparing automated with manual counts.

A total of 54,054 mosquitoes representing six genera were collected, dominated by Culex (77.6%). Across sites, BG-Pro produced the highest overall catches (urban mean 946.04; rural mean 13.29 mosquitoes/trap-day) and outperformed CDC-LT in both urban (IRR = 6.11) and rural (IRR = 2.64) settings, while the inverted BG-Pro traps captured the fewest mosquitoes. Genus-specific patterns were consistent across ecological settings: BG-Pro achieved the highest mean capture rates for Aedes (urban 8.3; rural 1.4 mosquitoes/trap-day) and Culex (urban 936.3; rural 8.3 mosquitoes/trap-day), indicating stronger performance for culicines relative to CDC-LT. In the rural site, overall captures of Anopheles minimus were very low across all traps, while CDC-LT sampled a broader range of anopheline taxa. The BG-Counter 2 median accuracy was high in urban sites (89.0%) but low in the rural site (20.6%), with the sensor overcounting mosquitoes on the majority of trap-days in both settings.

No single trap optimally samples all vector groups across ecological settings. BG-Pro (BG-Lure + CO₂) showed promising utility for operational culicine surveillance in urban Thailand, whereas CDC-LT remains preferable for anopheline monitoring in rural malaria-risk settings. The BG-Counter 2 can support near-real-time trend monitoring in low-background urban environments but requires site-specific calibration before rural deployment. These findings support a context-specific surveillance strategy that combines complementary traps rather than relying on a single platform.