Substituting records

Skip to first unread message


Aug 17, 2021, 10:14:03 PMAug 17
to Maxent
Hi Maxent users,

I am attempting to model the distribution of an endangered marsupial in New South Wales, Australia in a national park. The problem is that there is only one record for the species in the entire park. A species of ground-dwelling bird occupies very similar habitat to the  marsupial , but is more widespread throughout the park (26 records).

Is it an acceptable method to use the records for the ground-dwelling bird as a substitute for the marsupial (combined data set of 27 records) given their very similar habitat? All I can find is papers using Target-Group species for background sampling, rather than substituting the records themselves.

Any help would be greatly appreciated.

Kind regards,

Francisco Rodriguez Sanchez

Aug 23, 2021, 5:18:40 AMAug 23

Hi Lachlan,

If there are presence records of the marsupial species outside the park, I'd definitely use them. Otherwise, i.e. if there is only one single location known for the species, I think it makes sense to use a surrogate species with similar habitat occupancy patterns. This has definitely been done before (e.g. see

In any case, it would be very good if you could do some fieldwork after the modelling to search for the marsupial species in the sites predicted as most suitable. If you found new occurrences they could be added to the model in an iterative fashion to gradually get more refined predictions of species distribution.

Hope this helps. Good luck


You received this message because you are subscribed to the Google Groups "Maxent" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
To view this discussion on the web visit
Dr Francisco Rodríguez-Sánchez


Aug 25, 2021, 2:36:38 AMAug 25
to Maxent
Good afternoon (from Australia at least) Paco,

Unfortunately there are no records from outside the vicinity of the national park, there is one record ~50 km away but it has poor spatial accuracy.

Thank you very much for sending the paper by Qiao et al. (2017) through and your suggestions, it is much appreciated! The model will be used to help guide survey efforts for additional colonies by wildlife managers over the next year or so, so the most suitable areas will be ground-truthed. I will be sure to use any additional records they find to refine the model. I just hope they find some additional occurrences during their fieldwork as the future is not looking too bright for the species in this area due to introduced predators and herbivores. Many thanks again!

Cordiales saludos,
Reply all
Reply to author
0 new messages