Thread by @olliewearn

You hear lots of people saying that #cameratraps are great. But just how do they stack up to rival methods (line transects, live traps, detector dogs, eDNA etc)? We had a go at answering this question using available data. 1/n

royalsocietypublishing.org/doi/10.1098/rs…
(📷 @Carolina_ASN)

First step: we read a tonne of papers (systematically, mind). From thousands of studies, we whittled it down to 104 studies that yielded insights on camera traps vs other sampling methods. It’s obvious from the start: we realllly need more studies on this to be done. 2/n

Next we summarised what the authors of each study had concluded (if anything) about the relative effectiveness of camera traps vs other methods. We also like proper data (ouch), so as well as mushy qualitative data we extracted quantitative data (e.g. # species detected). 3/n

What did we find out? Shock horror: camera traps are rather good at their job. More studies recommended camera traps over other methods than did not recommend them (44% vs 16%), but lots weren’t sure (39%). The quantitative data, though, showed they were 39% more effective 🔥 4/n

Given that I spend a lot (too much?) of my time yammering on about camera traps, you can imagine that I was relieved to see this headline result. But there’s nuance. There always is. 5/n

Cameras were only clearly better than live-trapping, & perhaps hair-trapping. For other methods, it was a draw. If you a priori expected cameras to crush it, this might be disappointing performance. But objectively, cameras compete on a par with most methods & beat a minority 6/n

Whether you should use cameras also depends of course on your OBJECTIVES. Cameras smash it when the goal is to make a long species list or get lots of detections. But if you want to find species quickly (i.e. detection latency) or cheaply, then other methods might fare better 7/n

What’s more, there were hints that camera traps might be bested by some other methods, notably detector dogs, but there are too few studies currently to get an overall picture. More studies required! 8/n

If you’re still here, then I’ll tell you about 2 more results...

1) Tech advances have massively increased the effectiveness of camera traps in the past, particularly the shift to digital units with passive infrared sensors & IR flashes.

What might future developments do?! (Hello #ConsTech, #AIforConservation and @WILDLABSNET!) 9/n

2) There was evidence that cameras are less effective in open than closed habitats, relative to other methods. This might be because cameras struggle in dense vegetation or with high temps, or it might just mean that other methods do better in the open (think line transects) 10/n

The take home: So, you’re thinking of using camera traps? Great, they are ON AVERAGE great contraptions. But think about your objectives (always think about your objectives) and the specific challenges you might face using cameras in your study (cf …lpublications.onlinelibrary.wiley.com/doi/full/10.10…) 11/n