I'm working with a data set that has species abundance for two ecosystem engineers along a transect. Samples consisted of counting individuals within quadrats at specified intervals along each transect.
I'm having trouble determining if there is a density of one engineer that is too high for the other to maintain their destiny or be present. I'm not sure what type of statistical approach I could/should take.
The goal is to figure out if the two engineers have any capacity to co-exist within the same environment.
Hello Wesley. Could you elaborate on your problem a bit more? Especially what are the variables in the dataset? Uploading a subset of your data would be the best way I feel, that would make things clear.
This paper may help:
Article cooccur: Probabilistic Species Co-Occurrence Analysis in R
There is a range of issues to think about here. The simplest approach is to plot the abundances of one engineer against the abundances of the other, at least in the first instance. Often people just correlate the values, though there are potential problems with that, and the question you pose is more subtle, suggesting a threshold value. You could try fitting something non-linear. Things to think about though: are you more interested in quadrats where one engineer is present and the other absent? You might try something non-parametric then. Do the actual densities matter? A common approach for these kind of data is to convert them to percentages, so the value for each quadrat is the percentage of the total occurences of an engineer that occur in that particular quadrat. This allows you to deal with the fact that one engineer may naturally occur in greater densities when it does occur (think barnacles vs limpets, for example). While the question is beguilingly simple, there's a lot you could try. See J. exp. mar. biol. 449 (2013) 261-273 for some thoughts and, potentially, a method.
- Badges
- Science topic
- Similar topics
- Engineering