Volume 27, Number 9—September 2021
Ecologic Determinants of West Nile Virus Seroprevalence among Equids, Brazil
|Model||AIC||ΔAIC||AW||p value†||OR (95% CI)||Maximum OR difference among study sites‡||ρ§||Comment#|
|Distance to bird route, km|
|Forest loss, y/n||250.38||2.37||0.17||0.005||5.106 (1.518–31.796)||5.106||0.09||+|
|Presence of natural or primary forest, y/n||253.39||5.38||0.04||0.029||3.186 (1.111–13.48)||3.186||0.08||+|
|Mean temperature, °C||258.03||10.01||0.00||0.719||0.876 (0.427–0.803)||1.613||−0.04||–|
|Human density, no./km2||255.76||7.74||0.01||0.121||1.000 (1.000–1.001)||3.137||−0.01||+|
|Tree cover, %||256.87||8.86||0.01||0.257||0.981 (0.941–1.012)||2.618||−0.09||–|
|Horse density, no./km2||258.10||10.09||0.00||0.817||0.969 (0.741–1.275)||1.170||−0.03||–|
|Mean precipitation, mm||258.15||10.14||0.00||0.948||1.000 (0.999–1.001)||1.047||−0.01||–|
*Models are sorted by AIC, an estimator of the model’s quality; models with lower AIC values are superior to models with higher AIC values. Horse and human densities were based on 2018 data available from the Brazilian Institute of Geography and Statistics (https://www.ibge.org.br). Information on precipitation and mean temperature was obtained from WorldClim version 2 (https://www.worldclim.org). Information on tree cover was obtained from Copernicus Global Land Cover (https://lcviewer.vito.be/download). Information on natural or primary forest loss was obtained from Global Forest Watch (https://www.globalforestwatch.org). AIC, Akaike information criterion; AW, Akaike weight; OR, odds ratio; ΔAIC, the difference between a given and the best-supported model in AIC. †p values were determined by likelihood ratio tests of the different models. ‡Maximum OR difference among study sites indicates the highest OR difference possible for a given variable for better comparability between binary and nonbinary variables. §ρ, the Spearman correlation coefficient, ranges between −1 for negative correlations and 1 for positive correlations. The closer ρ is to 1 or −1, the greater the correlation between the observed variables. #Clarification that the observed variable is associated with an increase (+) or a decrease (–) of West Nile virus prevalence.
1These first authors contributed equally to this article.