In our development of Clark's theorem we assumed that everything of value about that cod fishery we were going to fish to extinction could be summed up in the amount of money we could get for selling the cod. But this implicitly assumes that the people who are buying cod can get the benefit they derive from cod by buying some other product. In other words, we're assuming that once we run out of cod we can substitute for them by using some other product. That assumption is known as the substitutability assumption. In many conventional economic analyses resources are assumed to be substitutable. If we run out of heating oil, we can switch to coal or natural gas. Maybe cod are substitutable, but as far as I know, the planet Earth isn't. If the resource in question isn't substitutable, then a fundamental assumption of Clark's theorem is violated, so it shouldn't be applied.
Notice, however, that the question of whether or not a resource is substitutable is only in part a scientific or technological question. Engineers can tell us whether it is feasible to substitute solar or wind power generating facilities for coal-fired power plants. But it is a matter of individual preferences and values whether we are willing to live in a world without cod.8 We can determine how strong those preferences and values are,9 but there's no way to do an experiment or make an observation to determine whether particular preferences are ``good'' or ``bad.'' Those are ethical or aesthetic judgments, and we'll talk about those next Monday.