Ecosystem models are currently one of the most powerful approaches used to predict and analyse the consequences of anthropogenic and climate-driven changes in food web structure and function. The modeling community is however still lacking techniques for taking microbial processes properly into account and assessing flow uncertainty explicitly. A linear inverse model of the Bay of Biscay continental shelf was built using a Monte Carlo method coupled with a Markov Chain (LIMMCMC) to characterize the system's trophic food-web status and its associated structural and functional properties. By taking into account the natural variability of ecosystems (and their associated flows) and the lack of data on these environments, this innovative approach enabled the quantification of uncertainties for both estimated flows and derived food-web indices. This uncertainty assessment constituted a real improvement on the existing Ecopath model for the same area and both models results were compared. Our results suggested a food web characterized by main flows at the basis of the food web and a high contribution of primary producers and detritus to the entire system input flows. The developmental stage of the ecosystem was characterized using estimated Ecological Network Analysis (ENA) indices; the LIM-MCMC produced a higher estimate of flow specialization (than the estimate from Ecopath) owing to better consideration of bacterial processes. The results also pointed to a detritus-based food-web with a web-like structure and an intermediate level of internal flow complexity, confirming the results of previous studies. Other current research on ecosystem model comparability is also presented.