ucfagls-thanks for the response. I’m tried doing what you suggested but it appears based on the bspline help page (https://rdrr.io/cran/mgcv/man/smooth.construct.bs.smooth.spec.html) that the first value in m (m[1]) determines the spline order. If that’s the case putting a temperature value in the m vector wouldn 't make sense.
For example, I tried to do what you suggested but I don’t think the syntax for m is correct.
testmod=gam(Binomial~s(DO.2)+s(Sal)+s(Temp,m=c(10,14,29,32))+offset(Effort)+s(Year,bs=“re”)+s(Month,bs=“re”),data=dat,family=‘binomial’, method=“ML”)
The response over temperature was extremely curvy when I did this and I think it is because the values in m are large. Any idea why this would be?
Thanks,