Predicting the potential global distribution of Ixodes pacificus under climate change

warmest season), and bio19 (precipitation of the driest season) may cause spatial illusions dur- ing modeling [13]. Therefore, 52 environmental variables are excluded and remain. Import 52 environmental variables into Maxent 3.4.1 software and load distribution point data. Use Jack- knife method to analyze the importance of environmental factors and obtain their contribu- tion. Remove environmental variables with a contribution rate of less than 1.0% to the model, and then use R software to perform Pearson correlation analysis on environmental variables with a contribution rate of � 1%. Variables with an absolute correlation coefficient of less than 0.8 are retained and incorporated into the model. Variables with an absolute correlation coeffi- cient of � 0.8 are considered highly correlated, and the variable with the highest contribution rate is included in the model. Selection of model parameters When there are few species distribution points, the model using default parameters may not be the optimal model [14]. To avoid overfitting of the model, we need to adjust the parameters of the model. all the collected distribution points of the I. pacificus were imported into the MaxEnt software, set to random seed 75% distribution point modeling and 25% distribution point verification modeling, the output format was “Cloglog”, the output file type was “asc”, the maximum iteration mode was to set select “Bootstrap” and the maximum number of repe- titions was 5000. The number of repeated training rounds was set to 20 to reduce the uncer-

tainty caused by outliers by applying threshold rule select “Minimum training presence”. We selected the best model by setting the feature combination (FC) and regularization mul- tiplier (RM). The RM parameter was set to 8 levels: 0.5, 1, 1.5, 2, 2.5, 3, 3.5 and 4, and there were 5 feature combinations: automatic linear (Linear, L), quadratic (Q), fragmentation (hinge, H), product (P) and threshold (T). We manually set 5 characteristic parameters to obtain 8 features (L, LQ, LQP, QHP, LQH, LQHP, QHPT and LQHPT). To get the best model and control overparameterization, the R package “ENMeval” was used. The ENMeval package was implemented in R 4.0.3 “Checkerboard2” to search for the lowest delta value for Akaike’s information criteria corrected for small sample sizes (AICc) to run the MaxEnt software on the candidate models. Evaluation of the model results The area under the curve (AUC) value of the receiver operating characteristic curve (ROC curve) is a commonly used test method to evaluate the accuracy of species distribution models. The closer the training omission rate is to the theoretical omission rate, the higher the accuracy of the constructed model [15]. The evaluation standard of a ROC curve is: AUC value = 0.5– 0.6, failure; 0.6–0.7, poor; 0.7–0.8, fair; 0.8–0.9, good; 0.9–1.0, excellent. Project the potential suitable area Based on MaxEnt, the cloglog result output is an asc file. Using ArcGIS 10.8, the asc file is con- verted into raster data using the ASC Ⅱ to Raster conversion tool, and the Jenks natural breaks

classification method is selected to reclassify the suitable habitat. The criteria are: 0–0.0539, unsuitable area; 0.0539–0.1941, low suitable area; 0.1941–0.4709, middle suitable area; 0.4709– 1, high suitable area. Results Global distribution points and environmental data of I. pacificus. After processing by ENMtools software, we obtained 567 points of I. pacificus. Using ArcGIS 10.8 to sample the PLOS ONE Prediction of the distribution of Ixodes pacificus PLOS ONE | https://doi.org/10.1371/journal.pone.0309367 August 27, 2024 4 / 18