Varying transect lengths/varying distance walked; track/hair/cadaver observations at PD 0

[Jessica Tax]

Hi,

I am currently working on my master's thesis at the University of Oxford, and am planning on using the distance programme for line transect analyses. After starting today, I have several questions regarding the analysis that I can't seem to figure out. My apologies if this has already been asked before, but I could not find an existing topic just now answering all three!

I have three questions: 

1. I have a set of 14 line transects of different lengths (ranging from 750m to 2300m). The dataset spans multiple years, considering the transects were walked multiple times with different lengths, is there a way to incorporate this in analysis? (e.g. transect PT with total length 950m was walked 4 times, with each walk crossing 950, 950, 650, 800 metres respectively).

2. Additionally, how do I deal with the fact that all my transects are of different length?

3. At our line transects we also took track/hair/cadaver observations into account. These are set at a perpendicular distance of zero. How can I analyse these as well without the zeroes heavily skewing my results? 

I would really appreciate your help! 

Best, 

Jessica Tax

[Eric Rexstad]

Jessica

I think I can speak to your first two questions, I'm less clear on the third.

1. With line transects, effort for the transect "PT" is the sum of the four trips; would be nice to know the reasons for not walking the full transect each time.  If you started at the same end on each walk, this makes the other end of the transect a bit "undersampled".  But for calculation purposes, sum the distances walked.
2. No problem for transects to be of different lengths—that is more common than transects of the same length. Effort for any given transect is the sum of walk lengths.
3. I'm not clear what to say about #3. If your only detections of tracks/hair/cadavers are all at distance zero, then there is no distance sampling to be done.  I'm not sure I understand.
   

---

From: distance...@googlegroups.com <distance...@googlegroups.com> on behalf of Jessica Tax <jessic...@gmail.com>
Sent: 06 September 2022 15:19
To: distance-sampling <distance...@googlegroups.com>
Subject: [distance-sampling] Varying transect lengths/varying distance walked; track/hair/cadaver observations at PD 0

 

--
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/83a27289-0715-431d-860f-2e406460319an%40googlegroups.com.

[Len Thomas]

Hi Jessica,

On 06-Sep-22 3:19 PM, Jessica Tax wrote:
> Hi,
>
> I am currently working on my master's thesis at the University of Oxford, and am
> planning on using the distance programme for line transect analyses. After
> starting today, I have several questions regarding the analysis that I can't
> seem to figure out. My apologies if this has already been asked before, but I
> could not find an existing topic just now answering all three!
>
> I have three questions:
> 1. I have a set of 14 line transects of different lengths (ranging from 750m to
> 2300m). The dataset spans multiple years, considering the transects were walked
> multiple times with different lengths, is there a way to incorporate this in
> analysis? (e.g. transect PT with total length 950m was walked 4 times, with each
> walk crossing 950, 950, 650, 800 metres respectively).

I assume the length of line walked each time had nothing to do with the density
of animals on the line (for example, it's not the case that you stopped walking
the line becuase the ground was flooded and you're survying a ground-dwelling
terrestrial animal). If survey effort is correlated with density then this will
need to be accounted for in the analysis. But I'll assume this is not an issue.

If you're only interested in an estimate of average density across all years
then you can pool the multiple visits to the same line, and for line length just
enter the total (950+950+650+800).

However, assuming you're interested in estiamtes by year as well as the average,
you can enter your data as if you have a different 14 lines each year, and each
year use the correct line length for each line. This causes a minor issue in
that any correlation in encounter rate on lines between years is not accounted
for, potentially affecting your variance estimates. A solution to this is to
implement a nonparametric bootstrap where you resample lines across years, but
this would need to be done programmatically outside of the Distance software
(e.g., in R).

> 2. Additionally, how do I deal with the fact that all my transects are of
> different length?

This is no problem -- Distance allows for a different lenght for each transect.

> 3. At our line transects we also took track/hair/cadaver observations into
> account. These are set at a perpendicular distance of zero. How can I analyse
> these as well without the zeroes heavily skewing my results?

I'm assuming that your main survey was of live animals, and their distance from
the transect line when observed. In that case, these "cues" that the animals
leave (tracks, hairs, dead animals) are not the same type of object, and should
be treated differently. There are methods for track intersections that I
believe have been applied to tracks in snow - I'm not very familiar with them
I'm afraid but perhaps others on the list are. I'm not sure what you can do
with hair. For cadavers I suppose you could consider them as a "cue" and do a
distance sampling cue-based survey, as people do for example for dung surveys.
However I don't think the analogy is direct as animals only proudce one carcass,
and also your sample size is likely to small for distance sampling. So, I'm
afraid I have no firm ideas on what you can do there. Sorry to have been little
help with your last question - perhaps others on the list have more ideas.

Best wishes, Len Thomas

> I would really appreciate your help!
>
> Best,
> Jessica Tax
>

> --
> 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

> <mailto:distance-sampl...@googlegroups.com>.

> To view this discussion on the web visit
> https://groups.google.com/d/msgid/distance-sampling/83a27289-0715-431d-860f-2e406460319an%40googlegroups.com

> <https://groups.google.com/d/msgid/distance-sampling/83a27289-0715-431d-860f-2e406460319an%40googlegroups.com?utm_medium=email&utm_source=footer>.

--
Len Thomas (he/him) len.t...@st-andrews.ac.uk lenthomas.org @len_thom
Centre for Research into Ecological and Environmental Modelling
and School of Mathematics and Statistics
The Observatory, University of St Andrews, Scotland KY16 9LZ
Office: UK+1334-461801 Admin: UK+1334-461842

The University of St Andrews is a charity
registered in Scotland, No SC013532.

[Tiago Marques]

Hi Jessica,

Just an additional note to say that regarding cadavers you have to account for the disappearance rate (some cadavers are removed before you detect them, say by scavengers). IMO, depending on the number of live animals vs cadavers, iyt might be sensible to ignore cadavers (because also hard to know how to relate live animals to dead animals (what is the death rate say). Folks do that a lot for wind turbines, to quantify impacts on birds and bats, you might search under that literature, but for that application death rate is ignored, since you just want to count dead animals, not live ones, which I presume is your main interest here?

Cheers

T

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/92aa2dba-71de-298d-4db3-25a6173265a4%40st-andrews.ac.uk.

[Jessica Tax]

Thank you all for your input! 

Op dinsdag 6 september 2022 om 18:44:48 UTC+1 schreef tiagoand...@gmail.com: