Can data across the years be clumped for analysis

[Nilanjan Basu]

Dear fellow members of the group,

I am trying to analyse line transect data collected over the last 12 years from a tiger reserve in Southern India. The transects remained the same from the first year to the last. I have enough observations of a few species for which I estimated the yearwise density. However, there are a few species for which I do not have enough observations (n<40) to estimate year-wise density. Is it possible to clump multiple years of data together and then estimate the density of those species? For example, instead of year-wise density, I will estimate density for years or something like that.

I look forward to hearing from you.

Regards

Nilanjan Basu

Research Consultant

Hyderabad Tiger Conservation Society

[Eric Rexstad]

Nilanjan

It is indeed possible to combine data across years to have adequate data to fit a detection function. However, if you use that pooled detection function to produce year-specific estimates, those estimates will be biased if detectability differs between years.

To avoid this bias, combine data across years, then fit a model that included year as a covariate. This will account for the variability in detectability between years.

An approximate example of this (using species rather than years) can be found here

https://examples.distancesampling.org/Distance-spec-covar/species-covariate-distill.html

Covariate modeling with rare species Use of covariate to model detectability in multi-species surveys. examples.distancesampling.org

Hope that is useful.

---

From: distance...@googlegroups.com <distance...@googlegroups.com> on behalf of Nilanjan Basu <nba...@gmail.com>
Sent: 27 September 2024 07:03
To: distance...@googlegroups.com <distance...@googlegroups.com>
Subject: [distance-sampling] Can data across the years be clumped for analysis

 

--
You received this message because you are subscribed to the Google Groups "distance-sampling" group.
To unsubscribe from this group and stop receiving emails from it, send an email to distance-sampl...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/distance-sampling/CALtNRJj7nUibrY_euXmZ5zbQ%3DyJM405ZrTxKv4GkW_TxM9e0AA%40mail.gmail.com.

[Nilanjan Basu]

Thank you for the information. Can it be done using the distance software?

Nilanjan Basu

Research Consultant

Hyderabad Tiger Conservation Society

[Eric Rexstad]

Nilanjan

Yes, this can be done using Distance for Windows software. I don't have a ready-made example, but here are some hints.

Organise your data such that years are strata (or regions). Create models where data are stratified, then detection function at the global level (meaning pooled data) and detection function at the stratum level (meaning year-specific detection probabilities). To incorporate year as a covariate, switch to using the MCDS analysis engine. Specify region as the covariate, and tick both global and stratum for the detection function.

If you want more details, ask me off-list.

---

From: Nilanjan Basu <nba...@gmail.com>
Sent: 27 September 2024 08:37
To: Eric Rexstad <Eric.R...@st-andrews.ac.uk>
Cc: distance...@googlegroups.com <distance...@googlegroups.com>
Subject: Re: [distance-sampling] Can data across the years be clumped for analysis

 

[Nilanjan Basu]

This is great, thank you so much! Dr. Rexstad.

Nilanjan Basu

Research Consultant

Hyderabad Tiger Conservation Society