Line transect design on very fragmented strata

[Khadija Amanda Barciela]

Hi!

Hope this finds you well. 

I'm planning a primate census that will consist on counting nests and mesuring perpendicular distances. The protected area where we work is really degradated (cleared areas, crops and anthropogenized areas) and also large areas consist on grassland, where we know the group of great apes that we are studying is not present. So, we want to stratify our study area in two strata consisting on the two different vegetation types where our apes are present. The result is the one shown in the attached image, a really fragmented polygon. As the expected density of nests is much higher in one stratum than in the other and the area of ​​the stratum with the highest density is in turn much smaller, the sampling effort must be different in both strata. The problem is that when I try to run my design of segmented transects with different efforts (different transect total length) on each stratum the Distance program does not generate any transect even though it does not give any error. I don't know if it's due to the shape of our study area or maybe I 'm missing something. Could somebody help us with that? Thanks a lot in advance 

[Eric Rexstad]

Khadija

Thanks for sending along the screenshot. I can't discern the widths (north/south, same as your current transect orientation) of the habitat fragments. I also can't locate your boundary between strata.

However, my initial suspicion is that your specification of the lengths of the transect segments you have specified cannot fit within the boundaries of the strata if you orient them north/south.  What is the relationship between the fragment widths and the segment lengths you are specifying?

If segments cannot "fit" in a north/south orientation, you might need to choose a different design angle.  That said, the degree of fragmentation might not permit segmented transect design to be implemented.

---

From: distance...@googlegroups.com <distance...@googlegroups.com> on behalf of Khadija Amanda Barciela <amanda....@gmail.com>
Sent: 05 April 2023 22:01
To: distance...@googlegroups.com <distance...@googlegroups.com>
Subject: [distance-sampling] Line transect design on very fragmented strata

 

Hi!

Hope this finds you well. 

I'm planning a primate census that will consist on counting nests and mesuring perpendicular distances. The protected area where we work is really degradated (cleared areas, crops and anthropogenized areas) and also large areas consist on grassland, where we know the group of great apes that we are studying is not present. So, we want to stratify our study area in two strata consisting on the two different vegetation types where our apes are present. The result is the one shown in the attached image, a really fragmented polygon. As the expected density of nests is much higher in one stratum than in the other and the area of ​​the stratum with the highest density is in turn much smaller, the sampling effort must be different in both strata. The problem is that when I try to run my design of segmented transects with different efforts (different transect total length) on each stratum the Distance program does not generate any transect even though it does not give any error. I don't know if it's due to the shape of our study area or maybe I 'm missing something. Could somebody help us with that? Thanks a lot in advance 

--
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/CAL%3DxCG15KarchFeEp8YPGBz68HpcYxH15yiJ9Ma-a2Q5wJay3A%40mail.gmail.com.

[Stephen Buckland]

Khadija, I don’t know the scale of your study area, but you may be better to keep it as a single stratum when generating the design.  After generating it, you could either only cover the line segments in the good habitat and restrict inference to that stratum, or you could cover all lines and estimate average density and total abundance for the whole area.  Or you could cover only a proportion of the transect segments in the low quality habitat, and complete a stratified analysis.

 

Steve Buckland

 

From: distance...@googlegroups.com <distance...@googlegroups.com> On Behalf Of Khadija Amanda Barciela
Sent: Wednesday, April 5, 2023 10:01 PM
To: distance...@googlegroups.com
Subject: [distance-sampling] Line transect design on very fragmented strata

 

Hi!

Hope this finds you well. 

I'm planning a primate census that will consist on counting nests and mesuring perpendicular distances. The protected area where we work is really degradated (cleared areas, crops and anthropogenized areas) and also large areas consist on grassland, where we know the group of great apes that we are studying is not present. So, we want to stratify our study area in two strata consisting on the two different vegetation types where our apes are present. The result is the one shown in the attached image, a really fragmented polygon. As the expected density of nests is much higher in one stratum than in the other and the area of ​​the stratum with the highest density is in turn much smaller, the sampling effort must be different in both strata. The problem is that when I try to run my design of segmented transects with different efforts (different transect total length) on each stratum the Distance program does not generate any transect even though it does not give any error. I don't know if it's due to the shape of our study area or maybe I 'm missing something. Could somebody help us with that? Thanks a lot in advance 

 

 

--

[Khadija Amanda Barciela]

Hi again,

Thank you so much for your answers, they were very helpful. 

After several more attempts I was able to verify that the segmented design was not working due to the fragmentation of my study area. So we have chosen to carry out a systematic random sampling that we can run in our study area. Now our question is that, when making this layout, the transects are continuous inside and outside our study area polygon (see QGIS screenshot below). Could we use these transects by cutting them with QGIS to keep only the part within our study area polygon and then load them in DISTANCE for analysis? Or would this generate some statistical problem?

Could you also help me please understand the differences between ESTIMATED ON EFFORT TRACKLINE LENGTH:  and
REALIZED ON EFFORT TRACKLINE LENGTH:  of the design engine output? Do the REALIZED ON EFFORT TRACKLINE LENGTH correspond to the total length of the lines and ESTIMATED ON EFFORT TRACKLINE LENGTH  to that of the lines within our study area? Thank you very much in advance and best regards.

Khadija

[Eric Rexstad]

Khadija

There are no serious problems caused by using the transects you created and cutting them with QGIS. I suspect you should be able to produce roughly the same design by placing the fragmented shape file into Distance and requesting a systematic parallel design (rather than a segmented design). If that works, you can save the step of using QGIS to cut the transects.

Regarding your questions about definitions used in Distance output:

Estimated vs Realised effort. Based on the spacing value provided as input when describing the systematic design, and estimated amount of trackline is determined. However, the realised length of transects in a given realisation of a design depends upon the random placement of the first transect. Hence the difference between estimated and realised values.

---

From: distance...@googlegroups.com <distance...@googlegroups.com> on behalf of Khadija Amanda Barciela <amanda....@gmail.com>

Sent: 13 April 2023 16:02
To: Stephen Buckland <st...@st-andrews.ac.uk>
Cc: distance...@googlegroups.com <distance...@googlegroups.com>; Eric Rexstad <Eric.R...@st-andrews.ac.uk>
Subject: Re: [distance-sampling] Line transect design on very fragmented strata

 

To view this discussion on the web visit https://groups.google.com/d/msgid/distance-sampling/CAL%3DxCG3uufP9x9ge7Qvbbdw3wBfqkEjvK4VVpPP0zz0%2BLiHmTg%40mail.gmail.com.

[Khadija Amanda Barciela]

Thank you so much Eric for your quick response. 

In reference to the realized and estimated length, I understand what you are explaining but the difference between the two quantities is very large (see screenshot below). I don't know if it can depend solely on the random location of the first transect. 

On the other hand, the image of the previous message corresponds to the transects generated by Distance when doing a systematic random sampling introducing my fragmented shapefile as study area (I think it is the same as a systematic parallel design) and the transects are generated as shown, inside and outside the polygon as single continuous transect lines (hence the question to cut them in QGIS, thanks for the answer, I'll cut them like this later for the field work and import them to Distance for the results), and therefore I thought that the big difference between both distances, estimated and realized, was related to this. I would be very grateful if you could confirm that this big difference does not really imply any problem in my design, since I would not like the results obtained to not be able to be analyzed after the sampling effort in the field.

Thank you so much again. Greetings