I am a PhD student new to this area, and I’m developing a field protocol for measuring lizard dorsal patterns for crypsis- specifically, my study is interested in understanding if lizards along a transect are more cryptic with respect to certain substrates than others. We already have Canon EOS 7D cameras which have been modified for full-spectrum images along with relevant vis spec and UV spec filters.
However, the topic of in situ photos vs. ex situ photos for this subject has been a source of debate. Other literature (Yamamoto and Sota 2020, Marshall et al 2015, a number of others) seems to generally imply that in situ photos are preferred, but given the difficulty of photographing these lizards with a diffusion screen and color standards, ex situ photos would be a bit more tenable from a field perspective. Further, the “Background” section of the “Essential Photo-Taking Checklist” page on this site seems to suggest that comparisons between an individual and its substrate are best done ex situ?
Any help or suggestions would be very welcome. Thanks in advance!
Yeah, what Cedric said… The only addition i would have is that photographing in situ allows you to capture natural backgrounds, lighting etc… at a cost of slightly noisier/less standardised images. So this trade-off is difficult to estimate and depends what kind of effects you’re trying to investigate. e.g. if the lizards are highly polymorphic then adding additional environmental noise might be too much to detect a signal.
I guess there are pros and cons for both approaches, as you say.
A possible compromise, given how hard it might be to capture images of your animals in-situ, is a sequential approach. Given your ability to obtain your animals and photograph them in the lab:
1. Take calibrated images of where you have seen an animal, even if the animal has gone. Make sure you know the light environment, i.e. measure it and get a spectrum that you then use for modelling.
2. Take calibrated images of your animals in the lab. Make sure you know the light environment, i.e. measure it and get a spectrum that you then use for modelling.
3. Use the toolbox to quantify the backgrounds and the animals –> draw your conclusions.
Provided you are diligent with calibrating your images etc. this approach would maybe allow you to be more flexible. Make sure to consider the level of variation associated with animals & backgrounds (i.e. don’t just use one animal if there is notable variation in your populations).
I hope this helps,