Copyright © University of Cambridge. All rights reserved.
This approach is very simple in its conception, and it seems to assume too many things, which make it slightly problematic.
The first problem is that it assumes that dolphins are uniformly distributed throughout the ocean. Factors like food, temperature and predators affect the location where dolphins choose to live. So, in a huge ocean the variation of the sizes of the pods of dolphins is expected to be very large, making the assumption unrealistic.
The second problem is that it only gives us an estimate for the population of dolphins in a very small area of the ocean. Consider, for example, that a medium-sized ocean such as the Indian Ocean contains water of total volume over 290,000 cubic kilometers. So, even if we believe that studying 1% of an area is enough to make deductions about the dolphin distribution, we can see how "insignificant" a result about the number of dolphins in 1000 cubic kilometers is.
And a third problem, is that we can never be sure that we caught all the dolphins within our investigation area.
This approach is slightly more sophisticated, but it still relies on a few assumptions, to which some people might object.
First of all, it assumes that dolphins in general will not travel very far away from the place where they were originally caught. This, however, cannot be taken for granted.
Moreover, what happens if there are 0 tagged dolphins in the second time? The ocean can't have an infinite number of dolphins, so the method cannot provide us with an answer (and so we would need to repeat the experiment until a tagged dolphin gets caught).
Furthermore, if there are too many dolphins in the ocean, we can't know what would be a reasonable number to catch the first time. For example, if there are 50,000 dolphins in an area of the ocean we are looking at, and we only catch 100 at a time, it is highly unlikely that there will be any tagged dolphins in the second set we catch.