We present an hierarchical approach to segment images of yeast cells based on watershed and space-scale analysis. Yeasts belong to an important fungi class and the performance of bioreactors and other chemical processes are greatly influenced by their morphological character. The method proposed is capable of segmenting yeast cells based on the analysis of survival time, shape and gray-scale features of the Tree of Critical Lakes. We show experimental results for one group of yeast images obtained from the School of Food Engineering at Unicamp, Brazil. Preliminary comparison shows that the proposed method provides cells with area 10% lower than traditional approach, as well as its contours preserved.