Ecological Archives E095-106-A4

Benjamin A. Turschak, David Bunnell, Sergiusz Czesny, Tomas O. Höök, John Janssen, David Warner, Harvey A. Bootsma. 2014. Nearshore energy subsidies support Lake Michigan fishes and invertebrates following major changes in food web structure. Ecology 95:1243–1252.

Appendix D. Statistical procedure for Dunnett’s test post hoc comparisons.

Following the two-way ANOVA procedure described in the methods, a Dunnett’s post hoc comparison was used to determine whether the C and N isotopic signature of individual species changed over time. The test statistic and standard error were calculated using a modified form of the equations from Zar (1999).

where is the mean δ13CNL or δ15Nc of the species being compared from 2010–2012,  is the mean of that species from 2002–2003,  is the mean δ13CNL or δ15Nc of dreissenid mussels from 2010–2012 and  is the mean of dreissenid mussels from 2002–2003. The standard error term for our post hoc pairwise comparisons is dependent upon the sample sizes of both the dreissenid baseline for two time periods and a focal species for two time periods. Therefore the standard error term incorporated all four sample sizes and assumes homogeneity of variance. The standard error term was calculated as follows:

where MSerror is the mean squares error term from the two-way ANOVA,  is the sample size of the species being compared from 2002–2003,  is the sample size of that species from 2010–2012,  is the sample size of dreissenid mussels from 2002–2003 and  is the sample size of dreissenid mussels from 2010–2012.

Literature Cited

Zar, J. H. 1999. Biostatistical Analysis. Fourth Edition. Prentice-Hall, Upper Saddle River, New Jersey, USA.

[Back to E095-106]