You are not currently logged in.
Access your personal account or get JSTOR access through your library or other institution:
If You Use a Screen ReaderThis content is available through Read Online (Free) program, which relies on page scans. Since scans are not currently available to screen readers, please contact JSTOR User Support for access. We'll provide a PDF copy for your screen reader.
Since scans are not currently available to screen readers, please contact JSTOR User Support for access. We'll provide a PDF copy for your screen reader.
Preview not available
The existence of assembly rules for biological communities has been controversial, partly because of disagreement on the tests used. Two randomisation tests are compared here, devised by Wilson and Fox respectively, both seeking assembly rules of constant guild representation. The null models used differ in their assumptions on species frequencies. It is demonstrated that Fox's method can give significant results even when species occurrences are allocated at random. This is because it tests mainly whether the guilds are similar in overall species frequency, rather than testing, as Fox intended, for a "tendency for species to be assembled as specified by [a] rule". Therefore, a significant result is likely to demonstrate only that species frequencies are not equal. Such failure to build well-known features into a null model, resulting in a demonstration of the obvious, is termed the 'Jack Horner' effect. In contrast, Wilson's method gives the theoretically expected probability distribution. When Fox and Brown's 'Nevada' dataset is reanalysed with this null model, i.e. allowing for differences in frequency between species, the assembly rule disappears. Thus, Fox's assembly rule for that dataset turns out to represent only that species differ in frequency. It is suggested that danger of the Narcissus effect must be accepted when the only alternative is an ecologically spurious result caused by the Jack Horner effect.
Oikos © 1995 Nordic Society Oikos