6
$\begingroup$

I am a master's student working on mixed-species bird flocks in the eastern Himalayan tropical forests. For some administrative and interpersonal reasons, our lab has ended up with a bunch (~25) of Audiomoth devices with no clear use case. Since our primary research concerns the effect of logging, I had initially thought of the following possibilities of utilising the available devices:

  1. To examine if the denser undergrowth in logged forests (due to a more open canopy) drives birds to call from a higher height than in primary forests with sparse undergrowth.
  2. Since I have data on mixed-species bird flocks, I wanted to test if there are ways to use PAM to get at inter-specific interactions (basically, compare flock observation data to an acoustic dataset of associations).

Upon discussing this with a subject-matter expert, I was told that the Audiomoths are not a great way to perform idea #1. I am not sure of Idea #2 (I haven't discussed this with them). However, I was hoping to get your help in learning about what sort of questions can be addressed with PAM devices except from presence-absence studies (considering that any distance-from-recorder calculations aren't feasible).

I look forward to hearing the community's thoughts! Any references would be highly useful.

Improve this question
$\endgroup$

2 Answers 2

Reset to default
3
$\begingroup$

You're entirely right that AudioMoth-based PAM does not easily lend itself to 3D localization nor to population estimates. At the same time, it can serve to characterize relative spatiotemporal changes.

A typical application of PAM beyond presence/absence is the description of"phenology", or the study of temporal variations in calling behavior. For example, if your species of interest exhibits a dawn chorus pattern, you can ask yourself if that pattern will be readable on the call count chart. Repeat over several weeks and you should see the peak of the dawn chorus move gradually later/earlier in Spring/Fall. If you can afford the time and effort, deploy your AudioMoth's along an ecological gradient and measure how that temporal pattern varies spatially.

Improve this answer
$\endgroup$
2
  • $\begingroup$ Hey! Thanks for your reply. Considering that I can do phenology/time of calling studies, do you think the second idea I had mentioned is worth looking into? Again, I will not be able to differentiate between flocking and non-flocking species. $\endgroup$
    – akshay_birdwatch
    Commented Aug 8, 2022 at 5:53
  • $\begingroup$ Yes, you could look at conditional probabilities P[X(t)|Y(t)] where X(t) is "at least one call detected from species X between t and t+delta" and same for Y. But remember that species classifiers such as BirdNET are susceptible to auditory masking. Calls from Y may mask X and thus lower detection rate ... TL;DR assemblages are harder than phenology $\endgroup$
    – lostanlen
    Commented Aug 8, 2022 at 12:44
2
$\begingroup$

Just in case you don't already know, Burivalova et al (2021) show neat whole-soundscape changes related to logging. It might give you some ideas.

Burivalova et al (2021) "The sound of logging: Tropical forest soundscape before, during, and after selective timber extraction" Biological Conservation, 254, 108812

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.