The expansion of biological species in natural environments is usually described as the combined effect of individual spatial dispersal and growth. In the case of aquatic ecosystems flow transport can also be extremely relevant as an extra, advection induced, dispersal factor. There is a lack of reproducible experimental studies on biological fronts of living organisms in controlled streaming habitats. It is thus not clear if, and to which extent, the current theoretical and experimental knowledge on advective-reactive-diffusive fronts for chemical reactions can also apply to the expansion of biological populations. We designed and assembled a dedicated microfluidic device to control and quantify the expansion of populations of $E.coli$ bacteria under both co-flowing and counter-flowing conditions, measuring the front speed at varying intensity of the imposed flow. At variance with respect to the case of autocatalytic reactions, we measure that almost irrespective of the counter-flow velocity, the front speed remains finite at a constant positive value. A simple model incorporating growth, dispersion and drift on finite-size hard beads allows to explain this finding as due to a finite volume effect of the bacteria. This indicates that models based on the Fisher-Kolmogorov-Petrovsky-Piscounov equation (FKPP) that ignore the finite size of organisms may be inaccurate to describe the physics of spatial growth dynamics of bacteria.
