The following assumptions form the basis of metaphysical or ontological naturalism,
(1) that there is no such thing as a "supernature" : the world/nature is causally closed and there is no interaction with supernatural realms
(2) that the world really exists, so it does not depend on the human mind.
(3) that all things must take place in a lawful, not chaotic, way

However, assumptions of this kind - unlike scientific statements - are basically "not empirically verifiable". Since they do not meet the falsifiability criterion for scientific statements, they are metaphysical assumptions. This means, however, that metaphysical naturalism is ultimately a belief system.

The German philosopher of science Martin Mahner claims "Ontological naturalism is a necessary prerequisite of the natural sciences". The examination of scientific theories is only possible in a naturalistic context.

Even just a glance at the history of science is sufficient to refute this thesis, because what we today call "science" was invented in the early modern period by theists, not by naturalists (!): Copernicus, Kepler, Boyle, Descartes, Newton... all believing Christians. If Mahner's thesis were correct, then it would be a miracle that these men could "invent" and advance the natural sciences.

Furthermore, it is impossible to determine a priori what NATURE is at all. If I define NATURE as follows: "The totality of all really existing things that do not owe their existence to the human mind" (i.e. everything that is not the product of human imagination or creativity), then it is not clear from the outset how far this area goes. So it is not the case that the natural sciences are based on an already known ontological concept of NATURE, but it is the other way round: the scientific method, i.e. the systematic back and forth between hypothesis and observation, successively determines what actually exists and what is due to our imagination. For this method, however, only two provisional basic assumptions are needed: that there is a world "out there", that is, beyond human mental activity, and that this world is essentially lawful.

Using this method, the area we call NATUR has gradually been discovered. If we assume 'for the sake of argument' that there are such things as the "astral bodies" popular with esotericists, and one could have evidence for their existence, as we do with viruses, neutrinos or gamma rays, then these astral bodies would be part of NATURE.

Or to take another example: A few decades ago, the biologist Rupert Sheldrake postulated the existence of "morphic fields". Were these natural or supernatural? Since there is no evidence for their existence so far, Sheldrake's books are now on the shelf for esotericism, not under natural science. But if in the future such evidence could be presented, then these "morphic fields" would be part of nature, and Sheldrake would be a candidate for the Nobel Prize.

In other words, it was only the application (!) of the scientific method that separated NATURE from FICTION. To put it in a nutshell: first you have to look very closely, then you know what NATURE is at all. And everything that does not belong to nature is by definition fiction. A "super-nature" cannot exist at all, but within nature there can be areas and things that are completely unknown to us (as gamma rays were for the people of the Middle Ages) - and perhaps will remain unknown for all time.

Conclusion: It doesn't matter whether a scientist derives his or her motivation from faith in a god, from belief in an ontological naturalism or solely from the success of science (the latter is called pragmaticism, and I have great sympathy for this philsophy). The core of science is its methodology, and it does not need justification to function. On the contrary, the belief that scientific activity requires something like a metaphysical final justification, a fundamentum inconcussum (as the scholastics or rationalists like Descartes called it), has more to do with religion than with science itself. There can be no ultimate justification for anything - but there is no need for it either.