The formation of planets within a disc must operate within the time frame of disc dispersal, it is thus crucial to establish what is the dominant process that disperses the gaseous component of discs around young stars. Planet formation itself as wel
l as photoevaporation by energetic radiation from the central young stellar object have been proposed as plausible dispersal mechanisms. [abridged]. In this paper we use the different metallicity dependance of X-ray photoevaporation and planet formation to discriminate between these two processes. We study the effects of metallicity, Z, on the dispersal timescale, t_phot, in the context of a photoevaporation model, by means of detailed thermal calculations of a disc in hydrostatic equilibrium irradiated by EUV and X-ray radiation from the central source. Our models show t_phot propto Z^0.52 for a pure photoevaporation model. By means of analytical estimates we derive instead a much stronger negative power dependance on metallicity of the disc lifetime for a dispersal model based on planet formation. A census of disc fractions in lower metallicity regions should therefore be able to distinguish between the two models. A recent study by Yasui et al. in low metallicity clusters of the extreme outer Galaxy ([O/H] ~- 0.7dex and dust to gas ratio of ~0.001) provides preliminary observational evidence for shorter disc lifetimes at lower metallicities, in agreement with the predictions of a pure photoevaporation model. [abridged] We finally develop an analytical framework to study the effects of metallicity dependent photoevaporation on the formation of gas giants in the core accretion scenario. We show that accounting for this effect strengthens the conclusion that planet formation is favoured at higher metallicity. [abridged]
