Can habitable planets form in clustered environments?

We present observational evidence of environmental effects on the formation and evolution of planetary systems. We combine catalogues of resolved protoplanetary discs (PPDs) and young stellar objects in the solar neighbourhood to analyse the PPD size distribution as a function of ambient stellar density. By running Kolmogorov-Smirnov tests between the PPD radii at different densities, we find empirical evidence, at the >97% confidence level, for a change in the PPD radius distribution at ambient stellar densities Σ ≳ 103.5 pc-2. This coincides with a simple theoretical estimate for the truncation of PPDs or planetary systems by dynamical encounters. If this agreement is causal, the ongoing disruption of PPDs and planetary systems limits the possible existence of planets in the habitable zone, with shorter lifetimes at higher host stellar masses and ambient densities. Therefore, habitable planets are not likely to be present in long-lived stellar clusters, and may have been ejected altogether to form a population of unbound, free-floating planets. We conclude that, while highly suggestive, our results should be verified through other methods. Our simple model shows that truncations should lead to a measurable depletion of the PPD mass function that can be detected with ALMA observations of the densest nearby and young clusters.