The definition is pretty standard. Aggregations like this may be exclusively habitat related but generally that would lead to randomly distributed individuals within the habitat. Social interactions can also lead to randomness (or quasi-randomness) but clumped or uniform distributions are strongly suggestive of social interaxtions, hence shoaling. In reality, the social interactions usually happen alongside externally mediated behaviour.
Absent appropriately designed, controlled experiments or Bayesion modelling of observational data you can only make an informed guess. This photo appears to me to be fish shoaling. The absence of other species is consistent with that interpretation. It could just be fish hanging around but that density of fish in one place would be pretty unusual if there weren’t social pressures over-riding agonistic behaviours.
Schooling is usually more or less straightforward to identify. Shoaling less so.