Actually your separate issues with your two loci may be related: were the primers and loci developed specifically for this species or another? Sometimes weird interactions can happen with primers developed for another species, such as amplification of PCR artifacts or other portions of the genome. That could also explain why it didn't work for the other locus.

Random things can also happen, such as contamination during PCR set-up. Did you run negative controls? Not everyone does this, but often I run samples twice to compare the results. That also helps with estimating error rates.

Consider how many individuals are showing this pattern. If it's a small number, it could be contamination. If it is a significant number, consider either not scoring individuals at those loci or not including the loci in the dataset at all.

Besides the ploidy level mentioned in previous answers, which could be an explanation, other possibility could be the integrity of your DNA samples. The question is: doues the three peaks appear consistently among other samples (with some grouping factor? locality, variety, treatment or so...) or, does the third peak (the problematic one) appear "alone" (i.e. homocigote)in some sambles or it only came out with a unique three peaks conformation? . Remember that primers are not "perfect2 and depending o the GC % they could not be such effective.

I recently had this happen (example 3rd one down) - a sort of forked peak. It was related to primer fit, but when I had it, I had it on numerous samples - and I was using a primer that was not tested on the species I was running. I always call the back peak if I think this is 'real' and that would put it in line with the allele at 156.7. Line 2 looks like contamination. Contamination can happen anywhere - in field during collection, in lab during extraction or often during plating. The negative control helps identify if contamination is an issue. You cannot score line 2 b/c you do not know which of the 2 of the 3 peaks is the correct ones. Try re-running that one and see if it looks the same. Otherwise, remove the locus or the data point if you cannot sort it out.

It's difficult to say what's going on from the images. The previous suggestions are reasonable things to look into. One more possibility, if you have two dyes in the same lane/capillary and one is overloaded it can produce apparent peaks in both wavelength bands. You would look for peaks at or very near the same mobility in the two colors where one of the dyes is fluorescing very strongly. Overloading can also produce split peaks if the fluorescence exceeds the capacity of the detector.

Hello all! Thanks for your responses. I have collected some fresh samples to test the primers one by one (not in multiplex), and will include a control. These primers were for a related species in the genus - I just found some primers published for the actual study species so will try these too.