What is predictable is that if I subject a population of bacteria to ten different stresses, the genome of the population will change such that the stressful conditions become optimal or normal conditions.
Biologists cannot predict WHAT EXACT mutations will arise - but we know that they will - and that those variations that arise that act to reduce the stress on the organism will predominate in subsequent generations.
Moreover evolutionary science demonstrated is predictive utility rather dramatically recently when the existence of a fish-tetrapod transition was predicted to be at a particular location at a particular time period - and low and behold - they found “Tiktaalik”.
You say “Well DUH!!!! Of course change (in DNA/living organisms) is going to happen”.
Because evolution is defined as a change in the genetic makeup of a population you may as well have said...
Well DUH!!!! Of course evolution is going to happen.
What would possibly stop it? Where is the mechanism to keep DNA exactly the same so that adaptive changes don't arise and predominate?
Trivial counterexample, but nonetheless instructive.
Subject a petri dish of bacteria to several moles of elemental Fluorine.
Let me know how that hopey-changey adaptation works out for you, mmmkay?
You are taking for granted four *key* elements.
1) The efficacy of the change in killing the target population.
2) The amount of time required / allowed to adapt (special case: something which is uniformly fatal, allowing *no* survivors) 3) 2nd special case: something which is not uniformly fatal, but for which there exists no successful evolutionary adaptation within relevant time scales)
4) The presence of multiple stressors -- lungfish may survive drought, but not backhoes digging in the mud.
And the fact that most individuals do not survive infanthood without being snarfed up for lunch by a predator, also influences the efficiency of how much a beneficial mutation actually *spreads* once it occurs.
Cheers!