1.) In my design I have biweekly samples throughout the year that yielded no specimen from any species, especially in the winter months. Should those be excluded in biodiversity calculations?
2.) To calculate niche overlap indices, what should be the lowest abundance for a species to be included? My total sample has about 7000 specimen in 37 species - with most species being quite rare.
Hi tom;
Could you specify which index would you use?
Whatever the index you use, I dont think that excluding zeros from your raw data is a good idea, but without knowing what is the index (and thus the purpose) is difficult to determine to leave of remove them (see Gotelli et al. 2015).
Bests
Jon
Many indices such as the Shannon index require division of the abundance of each species by the abundance of all species in the sample. If the total decision is zero you have a problem. What indicts do not involve such operations? Can you provide the DOI of Gotelli's article?
I had a similar problem when studying diversity of soil oribatids in fields (agroecosystems). In my view, omiting samples without presence of studied group would distort results. Some possible approaches (incl. simple indices) fr this situation are mentioned in my work which is on my profile at RG
It will be okay if you calculate MEAN values because the zeros will be eliminated
Zeros should not be excluded from the calculation. They are important values, and can be used in calculations, including models to give more information on species presence and absence.
I should probably rephrase my question. In my study, bi-weekly samples are temporal habitats, so I try to calculate indices for all bi-weekly samples. If I include samples with zero species (15 traps yielded no target species), many (all?) indices fail because they involve a division by zero. Can I assume that the diversity is zero?
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools