Environmental variables
We used 19 bioclimatic variables with 30s resolution (~
1km2 near the equator) that gives current climate data
(Fick and Hijmans, 2017 available inhttp://worldclim.org/version2)
as the environmental variables. Bioclimatic variables are derived from
the monthly temperature and rainfall values to make them biologically
more meaningful; for example, mean annual temperature, maximum
temperature of the warmest month, annual precipitation and precipitation
of the wettest quarter etc. (Fick and Hijmans, 2017). In addition, we
also used elevation (AsterDEM_Version3_drukref03.im obtained from
Watershed Management Division, Ministry of Agriculture and Forest) as an
environmental variable. We used bioclimatic variables as environmental
variables even for the fish as previous studies have found SDMs built
using bioclimatic and hydrological variables did not differ for fish
(McGarvey et al. 2018). We also did not reduce bioclimatic variables
since our study is exploratory in nature, and collinearity among
environmental variables was not an issue in a machine learning
environment like Maxent (Elith et al. 2011, Marco Júnior and Nóbrega
2018) though some literature (e.g. Merow, Smith, and Silander Jr., 2013)
suggests being cautious when interpreting SDMs resulting from the use of
correlated environmental variables.