3
$\begingroup$

We are trying to find the right analysis to perform for our university course Biodiversity and Landscape. We have data from the previous 10 years about species abundance [BB-method] and at which height (gradient from sea to more land inward, this is at a saltmarsh) these species are present. We want to find a shift in species abundance along this height gradient over time, as it has been suggested that the dynamics at the saltmarsh are decreasing.

The problem we encounter is in what class do we define variable type Time. Abundance we can explain as a response variable and height as an explanatory variable. Since Time does not directly influence the abundance, but more so the relation between height and abundance.

Improve this question
$\endgroup$

2 Answers 2

Reset to default
6
$\begingroup$

It sounds like you want an interaction term between time and height (also include time as a main effect). Interaction terms tell us how one variable shows change in the relationship between another explanatory and response variable. You may need some nonlinear terms as well if you are looking for the location where a change occurs.

You could also use the height where the species appears (or reaches a given abundance) as a response variable with time as the explanatory variable.

It is hard to say what is really best with short descriptions online, you may be best served by working with a statistical consultant who can sit down with you and go into more details on your data and questions.

Improve this answer
$\endgroup$
2
$\begingroup$

Analysis wise, it looks like ordination, both constrained and unconstrained, could be an interesting approach, it will not only tell you if the abundances change, but also how community composition shifts, assuming you have collected this data. To run this analysis, you could make the grouping factor for the communities be discrete time intervals. David Zeleny has an excellent site detailing ordination theory and statistical techniques in R. I would suggest you start by doing a DCA, an exploratory analysis akin to PCA. From the DCA, check the supplementary variables, height or time could be used here!, and see how, sites etc cluster. After that, you can formulate hypothesis and test them via constrained analysis (CCA, RDA).

Improve this answer
$\endgroup$
1
  • $\begingroup$ If you don´t have species composition data, then maybe a PCA is enough. $\endgroup$ Commented Oct 14, 2025 at 16:21

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.