gSMR spatial points conundrum

[Ella Bradford]

Dear all,

I am in the early stages of attempting to fit a gSMR model to African wild dog data. However, I have found that all examples of gSMR and SCR that I have come across use fixed spatial points - with methods such as camera traps. 

My data does not have fixed spatial points. Our resight method involves receiving a GPS point and then using VHF telemetry to follow up on the pack location, making the packs much easier to refind. However, with the exception of den sites, most locations are only used once.

Does anyone have thoughts on if it is possible/how to adapt a gSMR or SCR model to allow for varying spatial input points rather than fixed spatial points?

Thanks,

Ella Bradford

[Jose Jimenez Garcia-Herrera]

Dear Ella,

 

I’m sure that every colleague who reads your message will offer a different suggestion—and most of them will probably be equally valid.

My first question, which I think is key to being able to answer you properly, is the following: if you have a group of tagged animals and you plan to carry out gSMR—that is, explicitly incorporating the tagging process—what would the re-sighting process look like?

In particular, how would you expect to observe some of the tagged individuals together with a certain number of untagged ones? Or is your idea instead to rely exclusively on the tagging process and use the movement data to help ‘refine’ the detection parameters?

 

José Jiménez

 

De: hmec...@googlegroups.com <hmec...@googlegroups.com> En nombre de Ella Bradford
Enviado el: martes, 7 de abril de 2026 17:41
Para: hmecology: Hierarchical Modeling in Ecology <hmec...@googlegroups.com>
Asunto: gSMR spatial points conundrum

 

No suele recibir correo electrónico de embr...@gmail.com. Por qué es esto importante

--
*** Three hierarchical modeling email lists ***
(1) unmarked: for questions specific to the R package unmarked
(2) SCR: for design and Bayesian or non-bayesian analysis of spatial capture-recapture
(3) HMecology (this list): for everything else, especially material covered in the books by Royle & Dorazio (2008), Kéry & Schaub (2012), Kéry & Royle (2016, 2021) and Schaub & Kéry (2022)
---
You received this message because you are subscribed to the Google Groups "hmecology: Hierarchical Modeling in Ecology" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hmecology+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/hmecology/7389d70b-284d-4972-95a0-e19a43701b21n%40googlegroups.com.

[Ella Bradford]

Dear José,

Thanks for following up on this - this is the point where I'm stuck. I think I'll need to find a way to convert the telemetry points into a fixed point framework.

I know telemetry-based detections are not a standard data input in this style of model - however due to the group association characteristics of wild dogs I'm interested in seeing if I can adapt the work from Meyer et al 2026 ("A novel generalized spatial mark-resight model that accounts for group associations") for this dataset because we do not have independence of individual detections. 

This is the start of one of my PhD chapters and it's still in the planning stage. My current idea is to break up the landscape into pixels and then any detection within the pixel will be assigned to the pixel's centroid point. That would allow for resights to happen from sightings within the pixel. When a pack is resighted, there can be a mix of marked (individually identifiable) and unmarked (unknown/not identifiable) individuals. This dataset has telemetry data where most dogs are identified and citizen science where most dogs are not identified.  I'm primarily working with territory holding packs, so they do have activity centres. However, I'm not sure how creating the fixed points from telemetry data would impact error and uncertainty in any estimations.

The overall goal is to get estimations for abundance and to potentially integrate the results from this model into a broader IPM (adding in capture histories of known individuals and fecundity data). My fallback option is to do a CR model of known individuals, with detection probability varying by pack, rather than individual, but I think a gSMR with group associations would be a better representation of the ecosystem (and the dataset), so I'm trying to see if I can make it work. 

If anyone has thoughts on how to adapt telemetry-based data into a framework that would allow it to work within a gSMR model please reach out and I would love to chat more about it. 

Thanks,

Ella

[Jose Jimenez Garcia-Herrera]

Dear Ella,

 

If the study area can be divided into pixels that are small relative to the movement scale of wild dogs, detections can be aggregated at the center of the corresponding pixel (https://www.nature.com/articles/s41598-018-20675-9; https://doi.org/10.1111/2041-210X.13030). In this way, the spatial location of each pack is available for each daily observation—including both identified and unidentified individuals—across a sufficient number of pixels. Under this framework, a specialized SCR formulation, such as random thinning SCR (https://doi.org/10.1002/ece3.7091), can be applied, thereby avoiding the problem that some marked or otherwise identifiable individuals may be recorded as unidentifiable on certain sampling occasions.

 

In SMR, it is important to recognize the possible presence of three distinct categories of individuals: marked, unmarked, and individuals with unknown marking status. Failure to explicitly distinguish among these categories may introduce bias into inference.

More generally, whether using SCR or SMR, movement data can be incorporated to inform and refine model parameters.

 

Another important concern in both SCR and SMR is the potential bias arising from the gregarious behavior of wild dogs. Conducting a series of simulations may be an appropriate strategy to assess the magnitude of this bias, explore approaches to mitigate its impact, and evaluate model goodness of fit.

 

Best regards,

José

To view this discussion visit https://groups.google.com/d/msgid/hmecology/dcd56b95-a643-4253-b627-015f97782ee2n%40googlegroups.com.