Should I calculate the value of this index for each individual and then compare the results with a kruskal-wallis to test if there are differences between each group (age class), or should I just calculate the value of this index for the whole age classes (and hence compare the single values relatice to each class)?
Dear Riccardo, it's a pleasure respond to you! First thing; i would like to suggest to use the (S.I.A) Stable isotopes analysis for the study of trophic ecology. Second thing; the use of Kruskal-wallis test depend on your data, in fact you must test the normality of your distribution and only after apply a non parametric test like Kruscal-wallis.
As well Riccardo, I recommend you to try normalisation by logarithms. I just used for a primary production and general decomposition of ecosystems (plankton, benthos, periphyton) and very effectively. But use the natural log (ln). This goes better. Can do in Stat 8. For my opinion and experience (if I understand well your question) can use both approaches when compare variables of the same type distributions from dif classes but should prove they are the same type.
Andrey
Hey Riccardo, I agree with Davide: the most effective way to determine diet specialization is using stable isotopes supplanted with some gut-content analysis.
Thank you for the answer, Andrey.
I would like to say that I'm studying the diet of an endangered species of trout: Salmo cettii (Rafinesque, 1810), endemic of Italy. To collect the prey items, I used non-invasive methods such as gastric lavage and, since this salmonid is very rare, the application of SIA will be a waste (cause it requires the death of the caught specimens).
Furthermore the ANOVA is obviously not appliable since the distribution of my data is not normal. Very good English, Davide!
- Badges
- Science topic
- Similar topics
- Biological Science
- Zoology
- Ichthyology
- Vertebrates
- Fish