Is the cloglog output produced from Maxent a probability of occurrence or habitat suitability? I know this is a pretty fundamental question about Maxent. I have read all of the typically cited papers and there is no concensus about this. Any clarification would be appreciated.
Four output formats for model values are supported by Maxent; the raw, cumulative, logistic and cloglog.
Cloglog is the default output, which is the simplest to understand: it gives a probability of occurrence estimate between 0 and 1. The probability of occurrence is highly dependent on sampling design details such as quadrat size and observation time. Cloglog output estimates probability of occurrence by assuming that typical presence localities have an expected abundance of one individual per quadrat, yielding a probability of occurrence of about 0.632. Cloglog output appears to be most appropriate for estimating probability of occurrence.
You can read through the attached articles to get more clarity on the subject matter.
I hope you find it useful.
Habitat suitability is used in the literature, but seems a misnomer to me in this context. "Habitat" implies "suitable" for the species.
Maxent is also known as species "distribution" model" (SDM). In other words, the algorithm predicts, among others the (spatial/geographic) distribution of a species.
Further, "niche" is more than once used in this context. You could say that the output of the algorithm predicts the unoccupied niche (habitat?) from presence points in the occupied niche.
The label habitat and distribution signal explicitly spatial (2D) extent. Niche doesn't. You could argue that the full response curve output of the maxent algorithm depicts the niche of the species.
In conclusion, the SDM output may be labelled in various ways. As always in science, the issue is to define your concept/terminology upfront with authoritative references and use it consistently throughout the paper.
You may also refer to a more fundamental discussion on these terminologies in an earlier Q&A. At the moment, I have no time to look for the link.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Software
- R