Fig 1.
The geographical distribution of melioidosis cases (n=782) across townships in Taiwan is illustrated, with colors ranging from light to dark indicating case numbers from low to high. The arrow highlights the hotspot of melioidosis in southern Taiwan (A). The annual incidence rates in the hotspot (red) and Taiwan (blue) are depicted, with shaded areas representing 95% CI. The basemap was derived from open-access shapefiles provided by the Taiwan Government Open Data Platform (https://data.gov.tw, Open Government Data License, version 1.0).
Fig 2.
Melioidosis cases, rainfall, and wind speed from January 2003 to October 2024.
Weekly melioidosis cases (n, red), rainfall (mm, blue), and maximum wind speeds (m/sec, green) are presented for the hotspot (upper panel, n=144) and Taiwan (lower panel, n=782). Significant correlations (p < 0.05) with time lags of 0, 1 and 2 weeks were identified using a zero-inflated Poisson regression model (IRR, wind speed, 1.14 [95% CI: 1.07–1.22] in Taiwan, 1.10 [95% CI: 1.03–1.18] in hotspot, per 1 m/sec; IRR, rainfall, 1.15 [95% CI: 1.10–1.21] in Taiwan, 1.12 [95% CI: 1.09–1.14] in hotspot, per 50 mm).
Fig 3.
Melioidosis case numbers, rainfall, and wind gusts during typhoon events.
From January 2003 to October 2024, a total of 70 invasive typhoons making landfall in Taiwan and 462 melioidosis cases during these typhoon events are shown in chronological order (left). The number of melioidosis cases (n, red dots) and rainfall (mm, blue line) occurred in each typhoon period are illustrated (A). Panel (B) shows the distribution of melioidosis cases (red dots) in relation to rainfall (mm) and wind gusts (m/sec), with red dots representing individual typhoon periods. The sizes of the red dots correspond to case numbers, ranging from 1 (smallest) to 62 (largest). A total of 6 dots are not visible in the figure because no cases occurred in these 6 typhoons (B).
Fig 4.
Association between melioidosis incidence and bank repairs.
The association between melioidosis incidence and the lengths of riverbank repairs (m) (r2 = 0.18, β = 0.43, p < 0.05) (A) and seawall repairs (m) (B) is illustrated based on official reports from 2008 to 2023. The dotted line represents the regression slope.
Fig 5.
Hypothesis and distribution of specific orf2 amplicons, rainfall, and wind speed in the north and south of the hills.
It is hypothesized that B. pseudomallei-contaminated aerosols were generated from farming soils in the north to northwestern region and transported southward to the hotspot during typhoon events, with the hills (green) acting as a barrier that interrupted further dissemination. Panel (A) illustrates the wind direction (black wind barbs indicated) during the typhoon period and the melioidosis cases (n) occurring in the north and south of the hills. The dotted line indicates the highest points of the two hills. The triangles (yellow) indicate the air sampling sites. Panel (B) shows the average detectable concentrations of B. pseudomallei-specific orf2 amplicons (copies/m3) in the north and south of the hills from 2019 to 2020. An asterisk (*) indicates statistical significance (p < 0.05). Panel (C) presents the daily average concentrations of the specific DNA (copies/m3), rainfall (mm), and wind speed (m/sec) are shown (C), with significance correlations (p < 0.05) observed for time lags of 3 and 4 days (IRR, wind speed, 1.04 [95% CI: 1.02–1.06] in north sides, 1.20 [95% CI: 1.11–1.29] in south sides, per 1 m/sec; IRR, rainfall, 1.45 [95% CI: 1.44–1.46] in north sides, 1.08 [95% CI: 0.99–1.19] in south sides, per 10 mm). The basemap was derived from open-access shapefiles provided by the Taiwan Government Open Data Platform (https://data.gov.tw, Open Government Data License, version 1.0). Crop fields and urban green areas were manually labeled based on reference locations identified from government datasets.
Fig 6.
Changes of wind direction, wind speed and specific orf2 amplicons during typhoon events for a range of years (2019–2020, n
=6). The figure illustrated the proportion (%) of wind directions, average wind speed (m/sec), and the concentration of contaminated aerosols (median, Q1–Q3; copies/m³) during three phases: precaution, typhoon landing, and three days after the alarm was cleared. Shades of blue, from dark (6-8 m/sec, largest) to light (0-2 m/sec, smallest), divided into 4 intervals, represent decreasing average wind speeds per hour (A). Comparison of aerosol concentrations (median values) between the north and south sides of the hills is illustrated (B, left), and the comparison of average northwesterly wind speeds per hours (median values) between the same regions, during typhoon landing, is illustrated (B, right). An asterisk (*) indicates statistical significance (p < 0.05).
Table 1.
Seropositivity to melioidosis in the northern and southern regions of the hills.