There are few details in the abstract but wing length is notoriously difficult to compare across measurers - the standard method of measuring the leading edge o

There are few details in the abstract but wing length is notoriously difficult to compare across measurers - the standard method of measuring the leading edge of the wing by gently straightening the primaries along a stopped ruler can easily generate differences of 2-3mm for different people measuring the same bird. It can be done for small groups of people who work together - wader ringers (shorebird banders if you are in the Americas) where they deliberately practise to standardise their technique. If the sample sizes are good then the error should be consistent across time so the conclusion might still be sound. https://en.wikipedia.org/wiki/Bird_measurement#Wing https://en.wikipedia.org/wiki/Bird_measurement#Wing The good news from this is that nature is splendidly adaptable and able to compensate for changes that occur within a few decades. That says that plants and animals are a lot more robust than we think they are. UPDATE: One confounding factor is that standards in handling birds are always improving so there's no way of telling whether measurements are comparable across time. UPDATE: Thanks to throwaway5752 for finding the full text. The birds were measured by a single observer so you can ignore everything I said. Interesting result.