cowman345 said:
Of course fish pass on traits, they have DNA like all eukaryotes and are therefore capable of passing on genetic info. Getting any fish to breed is probably going to be harder than gerbils, and I'm no expert on freshwater, but I think guppies may be a good starting place if you insist on fish. They have highly distinguishable features, at least. (note: if it was me, i'd go with plants, since your results are practically guaranteed).
-dave-
Actually, according to Darwinian and universally accepted evolutionary theory, no animals pass on acquired physical traits genetically. However, you can breed such animals to accentuate physical traits that already exist in their gene pool.
Example: if you have a fish with normal fins and clip its fins (ouch!) and it mates with another normal fish, the offspring will have normal length fins.
Only if such clipped-fin fish mates with a fish with shorter fins would the offspring possibly have shorter fins than its parent. Then again, it might not. Either way, the fish whose fins were clipped cannot pass on the trait for shorter fins just because they were clipped.
What this person will need for their project is a fish that has a short life span and reproduces prolifically (R selected). Generally, larger fish are more K selected and have large body sizes, small litters, and long life spans. It would take a much longer time to reflect changes in such a gene pool because each generation is so much longer.
If I could give a hint/suggestion to the conductor of this experiment, it would be that they should think *very hard* about the environment which they want to control and how to control it so as to get the desired results.
Generally, a very important part of macroevolution is varying survival rates. You will probably need to impose a quantifiable, tightly controlled variable that will impose a fairly strong (not too strong or you'll kill off all of your subjects) selective pressure that will kill off some of the subjects. Those that do survive the variable would be strongly selected for the traits that allowed them to survive in the first place. It's different than "survival of the fittest", actually, because you're looking for persistant changes in subsequent generations long after the survivors mate. Also, you must not let up on the environmental pressure that started the gene shifts in the first place, or the gene pool could revert.
You may want to start with a simple, easily modulated tank variable like temperature. You would then have to crank the temp either up or down towards the very limit of what the fish tolerate without dying off completely. In a relatively decent sized population of fast-reproducing fish, you should see changes between two relatively equal colonies in different conditions. The problem with an experiment like this is that such permanent genetic changes between populations still could take many generations to occur and the changes could be relatively subtle. Also, there is of course the good chance that such stressed populations can still reproduce under the conditions. Ethically, there is a problem as a fish hobbyist of deliberately stressing or killing fish-- most of us want to make it as easy for them as possible to live! Finally, you are highly unlikely to be able to maintain a very large population on which to experiment. Genetic expression of populations in response to environmental stresses/conditions usually works better and faster in very large populations (thousands and thousands of fish in a colony). Either way, you will need to impose a stiff, harsh environmental variable to get the results you want in a reasonable amount of time. For example, if you impose only a few degrees temp difference between two equal and small fish populations, you will probably never see any environmentally imposed genetic changes in the relative populations. It's a very fine line-- the harsher the environmental change, the quicker the poplulation changes through selective breeding and survival; too harsh of a change and you kill off all your subjects.
One final note: environment in this case could be all sorts of user-controlled variables besides temp, like pH, nitrate levels, light exposure, etc., etc. Just my personal opinion, but I think that a test like this would be very difficult to carry out correctly so as to get the desired results.