YES

TRS:

  zeros() -> cons(0(),n__zeros())
  tail(cons(X,XS)) -> activate(XS)
  zeros() -> n__zeros()
  activate(n__zeros()) -> zeros()
  activate(X) -> X

max/plus interpretations on N:

  zeros_A = 0
  zeros#_A = 0
  cons_A(x1,x2) = max{0, x1, x2}
  cons#_A(x1,x2) = max{0, x1, x2}
  0_A = 0
  0#_A = 0
  n__zeros_A = 0
  n__zeros#_A = 0
  tail_A(x1) = max{0, x1}
  tail#_A(x1) = max{0, x1}
  activate_A(x1) = max{0, x1}
  activate#_A(x1) = max{0, x1}

precedence:

  tail > zeros = activate > cons = 0 = n__zeros