Marine spatial planning (MSP) is positioning itself as a rational decision-making process regulating uses of marine spaces and resources in order to reduce tensions between exploitation and conservation as well as between ocean stakeholders. As global political agendas identified marine protected areas as a key answer to biodiversity erosion, systematic reserve site selection became a critical component of MSP. Establishing an ocean zoning involves the analysis of large quantities of heterogeneous, multi-sources and spatially explicit data. This often leads to problems too complex to be solved by human intuition only, thus calling for optimisation tools to support the decisions. In that context, our work aims at informing practitioners about stakes, possibilities and limitations of MSP approach through reserve site selection tools. We first clarify the reserve site selection framework, especially the underlying mathematics - the problem formulation and the solving method. Then, we highlight potential pitfalls due to input data feeding the reserve-based planning approach. Finally, and more practically, we show to what extent parameters used in reserve selection tools shape the reserve outcome. These elements are explored and illustrated on a real case study, namely the Fernando de Noronha archipelago in the Brazilian tropical Atlantic. This work provides a brief overview of informational challenges brought by decision support tools in marine spatial planning negotiations.