Independent events and maximum number of individuals per independent event

[Emeline Auda]

Hi all,

I had a question concerning the independent events, as unfortunately during our camera trap survey, we did not have enough independent events for most species to be able to produce population estimates. As we are also taking into account the number of individuals on each images, for example for the wild pig, if we are just looking at the number of independent events with a 30 minute-interval (without taking into account the number of individuals), we recorded a total of 64 independent events, which is unfortunately not an "adequate" sample to fit a detection function (we took the threshold of 75 independent events). But in this case, we did not take into account the maximum number of individuals recorded during one independent event. Taking into account the maximum number of individuals, we will have around 93 independent observations if we can call it like that. I am not sure how the number of individuals affect all that and if it is possible to take that into account when looking as independent observations or if we talk about "adequate" sample of observations, it only takes into account the number of independent events and the maximum number of individuals per independent event should not be included.

Thank you for your answer,

Best regards,

Emeline

[Eric Rexstad]

Emeline

I'm afraid I do not have experience to say anything authoritative regarding the interaction of "independent events" and animals occurring in groups. My uninformed opinion is that an event is an event, regardless of whether the event was triggered by a single individual or a group. Others more experienced may have a more enlightened opinion.

I would not be too rigorous in applying a number of detections threshold. The precision with which the parameters of a detection function can be fitted will erode as the number of detections diminish. A survey with a small number of detections will have less precision that a survey with more detections; consistent with common sense. In CTDS studies with which I am acquainted, variability in the between-station encounter rate plays a much larger role in the uncertainty in the density estimate.

---

From: distance...@googlegroups.com <distance...@googlegroups.com> on behalf of Emeline Auda <emelin...@gmail.com>
Sent: 27 July 2022 19:22
To: distance-sampling <distance...@googlegroups.com>
Subject: [distance-sampling] Independent events and maximum number of individuals per independent event

 

--
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/6cf8213a-146c-4b30-a97a-08b165b6744en%40googlegroups.com.

[Eric Howe]

Good day Emeline and thanks Eric,

Different animals in the same image or video are not independent, so this does not boost your sample size. However I agree with Eric that it's worth trying to model the detection function with your 64 independent observations. If there's a reasonably smooth (and initially gradual) decline in detection probability with distance you could still get useful density estimates.

Eric Howe