Given the specificity of your question, Jochen's answer is adequate. However, given the information that he provides in his consultation, I would like to ask him if "non-clinical" refers to a directed process? And if it is not directed, is he thinking of an experimental design to obtain data?
In this regard, the answers to these questions determine in the first instance if it is convenient to carry out a Probabilistic or Non-Probabilistic Sampling. and within the Probabilistic, if it aims to establish behaviors of the current population, or how "certain experiment" impacts the experimental units that could be representative of a specific group of analysis.
I agree with the previous answers. For completeness, there are other tools out there for sample size calculation that might be easier to use than Gpower (e.g., sigdaan.com, Pass, xlstata, etc.)