When plotting microbial death curves, how might they look different for bactericidal versus bacteriostatic treatments?