Point transect survey design

[abrun...@gmail.com]

Hello, 

I am analyzing a completed dataset for a study that used point transect sampling for birds (multi-species) to determine the treatment level density of grassland bird species. 

There are 15 plots and in each plot, a transect with 10 points was surveyed. There are 5 treatments and  3 plots per treatment and plots range in size from 200-500 ha.  Within a plot, consecutive points are 400m apart and are meant to cover the area of each plot (picture attached). This means sometimes the nonconsecutive points within a transect are more than 400 m apart. 

My question in this case is what is considered the sampling unit here? The transect line or the point? I tried reading pages 15-17 in Buckland et al. 2015 (Distance Sampling Methods and Applications) but I can't figure out in this case if I should consider each transect a cluster of points or each point a sampling unit. 

My end goal is treatment level density (5 estimates) and my original plan was to pool all the points within a treatment since I do not suspect detection to change within treatments (using treatment as the region) so I would have 30 points in a region. Or should I just have 3 lines in a region (which I know is not ideal).

I appreciate any advice. 

[Eric Rexstad]

Looks like an interesting study; your figure describes the design quite well. The traces of the transects along which the points were placed is apparent, particularly for the irregularly shaped plots (e.g. 3, 14, 13, 9, 11).

I suspect you should treat the transect as the sampling unit rather than the point; it will hamper your estimate of encounter rate variance, but I think that is the lesser evil compared to an inflated number of sampling units. I think this might be the conservative approach, making the treatment effect more difficult to detect rather than "too easy" to detect.

---

From: distance...@googlegroups.com <distance...@googlegroups.com> on behalf of abrun...@gmail.com <abrun...@gmail.com>
Sent: 12 December 2023 20:29
To: distance-sampling <distance...@googlegroups.com>
Subject: {Suspected Spam} [distance-sampling] Point transect survey design

 

--
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/12e12cfa-a8ff-4726-be8f-55eee8a30027n%40googlegroups.com.

[Stephen Buckland]

Adding to Eric’s comments, it’s a tricky one because some of your plots are well covered by the points, and it would be fine to treat the point as the sample units, while others just have a single line of points through the plot.  If plots are very homogeneous, it might not be unreasonable to treat each point as a unit, but to note this issue when writing up.  I guess you could try bootstrapping, once with plots as the sample units and once with points, and see if the 2 options lead to appreciably different inference.

 

For the plots with just a single line of points, you could regard your inference to be restricted to a strip of width 400m centred on the line, in which case you could arguably justify using points as units.

 

Steve Buckland

 

From: distance...@googlegroups.com <distance...@googlegroups.com> On Behalf Of abrun...@gmail.com
Sent: Tuesday, December 12, 2023 8:29 PM
To: distance-sampling <distance...@googlegroups.com>
Subject: {Suspected Spam} [distance-sampling] Point transect survey design

 

Hello, 

--

[abrun...@gmail.com]

Thank you all for your insight and advice! The plots are pretty homogenous, it's a large tract of grassland broken up into plots.  I will try running it both ways and see what comes out.

Thank you again, 

Andrea 

[Eric Rexstad]

Be careful Andrea about "doing it both ways". The danger there is that you will choose which analysis to use coloured by the answers each option provides.

---

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

Sent: 13 December 2023 15:20
To: distance-sampling <distance...@googlegroups.com>
Subject: {Suspected Spam} Re: {Suspected Spam} [distance-sampling] Point transect survey design

 

To view this discussion on the web visit https://groups.google.com/d/msgid/distance-sampling/dd7ba159-f9df-4695-b74f-38c83d4edce2n%40googlegroups.com.