The rationale is that if the ratio of bird densities from one area to the next is less than 5%, then there are more than 20 times more birds in one area than the next. Given that birds don't respect postal code boundaries, this is unlikely.
However, the main reason here is that the 110 looks wrong - it's much higher than the other numbers in the table, and out of step with the exam paper (which says that "The 1985 survey showed that the Blue Bird population density was high in the south of Actuaria but decreased towards the north of the country."). So you need to do something to check the data so you can show that you've picked up the error.
I personally would have just used the quote above to test that the density in each area is lower than the one to the south of it - this would have done the job just as well. Alternatively, as there are only 5 numbers, you could say that you'd looked them over by eye, and noted that the 110 looked wrong. For the 2015 data, you need something automated as there's more of it, so having something that works for both is better.
Hopefully that answers the question?
