You first should choose the fragments you want to amplify and adjust your DNA extraction method. The bigger your fragments, the more DNA and high quality you will need to use. I have extracted DNA using a standard Phenol-Chloroform protocol from mites which are also very small, and had approx. 50ng/uL DNA at the end, which was more than enough to amplify a 950bp CoxI fragment.
I think you will get enough DNA for barcoding out of a leg if you can pull it from your specimens.
The question is whether it makes sense for you to use NDE (non-destructive DNA-extraction). At least in ants you might be happy with what you get (the cleared cuticle, which you can wholemount then.
We always use NDE in Protura, which often are below 1mm of length. I would guess that we never had the Problem of too few tissue for DNA-barcoding up till now.
greetings
nikola
P.S. you can see our protocols, but there is plenty of literature on NDE for sure also with beetles and ants
Thanks so much for your responses, this is really helpful info and I reckon from this that two legs from even the very small specimens should be sufficient for the PCR.
It depends on which protocol you are using. If your specimens are very fresh and their DNA is well preserved (e.g., frozen), you may be able to get away with one or two legs. Alternatively, you may wish to look into non-destructive DNA extraction on glass fiber plates followed by voucher recovery - see link below: