Maybe you can consider both scales from Spielberger: STAI and IDER, one for anxiety and the other for depression. Although its true that is very difficult to distinguish between depression and anxiety, this situation is related with the overlaping between these diagnoses more than the scales. I prefer to use something like this and search what is common between the two entities and what is specificity of them.
If you're going to measure general population, maybe as TRAIT anxiety and depression you can use NEO-FFI neuroticism subscales. Neuroticism in NEO-PI-R consists subscales of anxiety, depression, hostility, impulsivity, shyness and overreacting. This might be satisfactory from more conservative, academic point of view
Though trait depression is not a commonly accepted construct, one can use scales of Rumination to assess tendencies to feel depressed. Ruminative thinking patterns are more persistent and pervasive, and are generally the focus of treatment. Current research is focusing a lot in this area, and this "trait" of rumination can be a viable measure of "trait depression".