These regions may be widely separated, particularly within animal distributions that are influenced by prey or host densities.
This distribution makes sampling difficult and invalidates commonly-used parametric statistics.
The zoologists M. P. Hassell and R. M. May noted that predators and parasites, too, might aggregate themselves where prey was abundant, choosing some response curve: they observed for example that redshanks (predatory birds) adopted a sigmoid (s-shaped) response to the density of Corophium (amphipod) prey per square metre of mudflats.
[2] Aggregation is seen in directly transmitted parasites (those not using a vector) from many groups: ectoparasites like lice and mites, marine parasites like copepods and cyamid amphipods, and many kinds of nematode, fungi, protozoa, bacteria, and viruses.
Log-transformation of data before the application of parametric test, or the use of non-parametric statistics is recommended by several authors, but this can give rise to further problems, so quantitative parasitology is based on more advanced biostatistical methods.