The problem is that extremely small, biologically uninteresting effects can be statistically significant, as long as the sample size is sufficiently large. For example, automobile accidents increase during full moons, and this result is statistically significant. Such results attracted media attention because they bring to mind stories of werewolves and vampires wreaking havoc on the human population. But the size of the effect is a trivial 1% increase in the accident rate, far too small to make it worth cautioning you to check the phase of the moon before you go driving. In fact, almost any null hypothesis can be disproved with a large enough sample. There are likely few factors in biology whose effects on humans and other organisms are exactly zero. We should care about the effects only if they are large enough to matter.