2.2 Selection and comparison of climate variables
Table 1. 19 Bioclimatic variables in the study.
A total of 19 environmental variables for the current (1971–2000s) and
future (2041–2060s, 2061–2080s) were downloaded from WorldClim
(http://www.worldclim.org/) with a resolution of 2.5’ (Table 1). Data of
BCC - CSM2 - MR from Coupled Model Intercomparison Project phase 6 were
selected for future climate data, including three scenarios in shared
socioeconomic pathways: sustainable development SSP1-2.6, moderate
development SSP2-4.5, and conventional development SSP5-8.5 (Riahi et
al. , 2017). The multilinearity of the 19 environmental factor
variables related to PWD can affect the accuracy of MaxEnt model
prediction results, so Pearson correlation analysis was used to test the
correlation of variables. First, 19 climate variables and 647 pine wood
nematode distribution points were loaded into the MaxEnt model. The
importance of 19 climate factors was determined using the knife-cutting
method and ranked according to the contribution rate of the factors.
Additionally, correlation analysis of 19 climate factors was conducted.
If the correlation coefficient of two related variables is greater than
0.8, the lower contribution rate of the variables should be deleted to
avoid the impact of multicollinearity on the model results (Zhanget al. , 2016). Finally, seven variables were obtained through
screening: isothermality (bio2/bio7) (*100) (bio3), temperature
seasonality of warmest month (bio4), max temperature of warmest month
(bio5), mean temperature of driest quarter (bio9), coefficient of
variation of precipitation seasonality (bio14), precipitation
seasonality (bio15), and precipitation of wettest quarter (bio16).
China’s administrative division vector map was obtained from the
national basic geographic information system (http://mail.nsdi.gov.cn).