However, why would it be selected? If it is just beneficial, I fail to see how NS could select it. I only see NS working when it has no choice. Example; Some plants with short roots can survive, but so can longer roots. Longer roots are beneficial so are selected, yet the short roots survive too? Isn't that just no change in the gene pool?
No. The measure used for gene pools is a frequency distribution. A change in the relative frequency of alleles is still evolution.
Here is a quick very basic evolution 101, defining a few important terms.
gene A gene is a particular bit of DNA at a particular location in the genome, with some functional role (usually coding for a protein). The unit of heredity.
allele A specific sequence of DNA for a gene. There can be a number of alleles for a gene in a population, and each individual has two alleles for every gene (one from each parent). In a population of organisms, there is a frequency distribution for alleles indicating how common each allele is within the population.
mutation A change in DNA sequence during replication. New alleles for a gene can arise through mutation.
genetic drift The frequency of alleles changes over time depending on how many instances happen to carry over into the next generation. Often the different alleles don't make any difference to an organism, but the frequencies will still go up or down just by the happenstance. This is called
drift.
selection The number of surviving offspring for an organism in the next generation depends on all kinds of things. But sometimes having one allele or another can make a small change in the probable numbers of surviving offspring. This is called
selection. Positive selection is when a certain allele increases the average expectation for numbers of surviving offspring; and negative selection is when the average expectation is reduced. This depends also on the environment, of course. The same alleles can be positively selected in one environment and negatively selected in another.
fitness The success of an individual, or of an allele, as measured by its probable frequency in successive generations. Alleles under positive selection are said to have increased fitness.
fixation Fixation is when the frequency for an allele reaches 100%. This can occur by drift, or by selection.
evolution The formal definition of evolution is often given as "change in the distributions of alleles in a population over time". Mutation, drift, and selection are all processes which impact the distributions of alleles; and hence they are mechanisms of evolution.
Now let's return to the length of roots. Typically, root length is affected by many different genes and many different alleles. There is no one gene for root length. However, if having longer roots gives an increased expectation for numbers of surviving offspring in the next generation, then this means a positive selective pressure on alleles which are positively correlated with longer roots; and negative selection for alleles that tend to lead to shorter roots.
Under this circumstance, the probability of fixation for new alleles tending to increase root length is raised; and the probability of fixation for new alleles tending to decrease root length is lowered.
Over time, the average root length of the population will increase. Looking at isolated individuals, you will still be able to find ones with short roots that prosper, but they will tend to be a bit less common. Over many generations, the mean root length will increase, and with strong selection or very long periods of time you will get to a point where you no longer find any plants with roots as short as the average length from an earlier time.
This is really basic. Selection can work on tiny changes in probability; it is simply false to say that selection only works when there is no chance at all for something with short roots to survive.
Mike; this is something of a deal-breaker for you. I've tried to set out the matter as clearly as a can in a single post, and I am happy to answer questions; but only if there seems to be some point in doing so. The basic meaning of these terms is not a point for debate. You don't get to invent new meanings for terms, and the phenomena of selection, drift, mutation, fixation, and so on are all directly observed.
Selection is a fairly simple concept, and it can be measured. There is no need for a selected allele to be essential for survival. You are trivially wrong to insist on this point, and you've been refuted on this in the thread many times. The only question at this point is whether or not you are capable of learning something.
When you solicited feedback in another thread, you got rather angry at my answers. I'm not surprised at that; I spoke strongly. It's not that I'm irritated with you. It's that over time I've come to the conclusion that it is a waste of time trying to explain such things.
You'll show I'm wrong when you can stop this childish nonsense about being "irrefutable" and show some capacity for actually learning something and grasping the points being explained.
Cheers -- Sylas