Help Area and data organisation

[riccardo...@gmail.com]

Hello everyone,

I am doing a study on Ds using thermal imaging. The study is divided into two areas with two transects each (4 transects total). What is the best way to organize the data-frame? Combine the data from both areas or conduct the analysis separately for each? How can I proceed to get correct models and data? I have tried both ways but I am not sure if I got correct data..... Mainly I can't understand why the "Covered Area" changes depending on whether the two areas are analyzed separately or in one data-frame.

I would also like to ask what portion of the territory should be considered for the "Area" column. I do not currently have this column in my data, and I think I understand that my problems may stem from this, but I would not know which area to consider. The area of the meadows where I carry out the transects? The entire area of the protected areas within which the study is conducted?

Sorry for the many doubts and thank you very very very much for your help

[Eric Rexstad]

Ricardo

Your challenges extend beyond the organisation of your data. You will struggle to make strong inference with only two replicate transects in each of your strata. Central to inference using design-based methods is the idea that the places you sampled are representative of the places to which you want to make inference. This is best accomplished by a) employing a sampling design that contains an element of randomness and b) having a number of spatial replicates. With only two replicates chance will play a larger role in your findings that with a larger number of replicates.

The "Area" to be specified in your data is the size of the respective strata. The size of the stratum is defined as the area of the region over which your transects were randomly placed (the area over which you wish to make inference). Were you intending to make inference to the protected area or the meadows within the protected area; your description makes it seem like the latter.

Finally, the answer to your question about what detection function model to use (common for both strata, or unique to each stratum) is a model selection question. Fit both models and select the model most appropriate for the data observed. This exact scenario is discussed in our online materials

https://workshops.distancesampling.org/online-course/syllabus/Chapter5/#stratification

Chapter 5 Stratification, clusters and covariates Teaching distance sampling through video lectures and computer exercises. workshops.distancesampling.org

and the exercise accompanying that lecture.

---

From: distance...@googlegroups.com <distance...@googlegroups.com> on behalf of riccardo...@gmail.com <riccardo...@gmail.com>
Sent: 20 April 2023 23:37
To: distance-sampling <distance...@googlegroups.com>
Subject: [distance-sampling] Help Area and data organisation

 

--
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/b51eebba-b4f1-4b76-9247-31de3b7a9d6bn%40googlegroups.com.

[riccardo...@gmail.com]

Hi Eric,

Thank you very much for your clear and timely response and thank you for the shared material which I will study carefully.

To clarify my first question: the transects are yes two for each of the two protected areas but each of them has been traveled every 10 days in the past 3 months, for a total of about 10 runs for each transect.

So as an area can I express either that of the two protected areas, and thus have the same area for the two pairwise transects, or that of the location where I ran the transects and thus have 4 different areas? I know it obviously depends on where I want to refer the inference to, but for the purposes of calculating and running "Distance" does it change anything?

I have one more question: if in the data-frame I expressed the length of the transects in meters, in what unit should the area be? square meters or hectares?

Thank you very much for your help, but I am sure I will come back with more questions as the analysis progresses.

Have a great day

Translated with www.DeepL.com/Translator (free version)

[Eric Rexstad]

Ricardo

Revisiting your two transects in each stratum does not increase the representativeness of them.  Temporal replication (repeated visits) is not equivalent to spatial replication (more unique transects). The threat of non-representative sampling still exists with this more detailed explanation of your survey design.

As you described, you have only two strata of your protected area. It does not make sense to describe your design as consisting of four area each with a single transect.

You can provide the size of your strata in whatever units you wish.  There is a companion function in the Distance​ package called convert_units​. It takes three arguments, the units of measure for perpendicular distance, units of measure for transects and units of measure for area. By appropriately using the convert_units​ function, you can provide measures that make sense for your study animal. It likely makes more sense to express density of deer in numbers per hectare or numbers per square kilometer, than numbers per square meter.

---

From: distance...@googlegroups.com <distance...@googlegroups.com> on behalf of riccardo...@gmail.com <riccardo...@gmail.com>

Sent: 22 April 2023 21:09
To: distance-sampling <distance...@googlegroups.com>
Subject: Re: [distance-sampling] Help Area and data organisation

 

To view this discussion on the web visit https://groups.google.com/d/msgid/distance-sampling/59b0173a-4b5a-482e-8b7d-ed2072ae0e2en%40googlegroups.com.

[riccardo...@gmail.com]

Hi Eric,

Thanks you again for the detailed answer.

Kind regards anche have a nice day