Disc fragmentation rarely forms planetary-mass objects

It is now reasonably clear that disc fragmentation can only operate inthe outer parts of protostellar discs (r > 50 au). It is alsoexpected that any object that forms via disc fragmentation will have aninitial mass greater than that of Jupiter. However, whether or not sucha process actually operates, or can play a significant role in theformation of planetary-mass objects, is still unclear. We do have a fewexamples of directly imaged objects that may have formed in this way,but we have yet to constrain how often disc fragmentation may actuallyform such objects. What we want to consider here is whether or not wecan constrain the likely population of planetary-mass objects formed viadisc fragmentation by considering how a population of objects at largeradii (a > 50) au - if they do exist - would evolve underperturbations from more distant stellar companions. We find that thereis a specific region of parameter space to which such objects would bescattered and show that the known exoplanets in that region haveproperties more consistent with that of the bulk exoplanet population,than with having been formed via disc fragmentation at large radii.Along with the scarcity of directly imaged objects at large radii, ourresults provide a similar, but independent, constraint on the frequencyof objects formed via disc fragmentation.