Dr. Pangloss in Voltaire’s Candide is famous for seeing how everything fit perfectly together. Everything is for the best in the best of all possible worlds! Most of us have more difficulty in seeing the relationships between living individuals and their environment. This is especially true when many biological traits or environmental measures are involved.
To begin to see how we might measure fitness, let’s take the case of a single trait adapting to a single aspect of the environment. Let’s look at the length of hummingbird tongues and the flowers that are available for feeding.
If we take the length of the bird’s bill and add the distance that it can extend its tongue beyond the bill, we get its tongue length. This should also be the depth of flower from which it can feed. This length for an individual bird can be compared with the average flower depth of the flowers in that bird’s home range. This can be done for a sample of hummingbirds in a number of different areas with different flower species. The sample in each location can include different hummingbird species to represent the different animals competing for food. If we did this in five locations where longer floral tubes are successively more common, we would expect the following relationship:
As a rule, we would expect birds with longer tongues to live in areas with deeper floral tubes. Under circumstances where alternative food is available or other factors are more important for survival than floral tube depth, the correlation between the animals and their environment ( the fit between them) would be less noticeable. Where the fit is quite good, as in the graph above, birds that have longer tongues than predicted by the relationship would not fit as well as those that fall right on the line. The same is true for those that have shorter tongues than you would expect from the relationship. If tongue length is the only important trait for survival, then natural selection would predict that birds farther from the line will have poorer survival or reproduction than those that are right on the line. Since hummingbirds almost always lay two eggs, the focus would be on adult survival and their success in raising the chicks.
Of course, no one trait can uniquely predict the fate of an individual or a group of related individuals. Fortunately, there is a way to extend this reasoning to as many traits and environmental characteristics as we can measure. The attached paper explains how to find the individuals that differ the most from the relationships that matter the most for survival. It is these individuals which natural selection, if it is operating vigorously, should weed out. If, on the other hand, these odd individuals are doing as well as those that measurably fit their environment, then we must conclude that selection is not occurring.
I wrote the paper some 20 years ago and was unable to publish the idea. I ask you, my dear reader, to have a look at it and see whether we might not develop a predictive theory of natural selection.