It is hard to say anything precisely without complete knowledge on samples, substrate on which it is deposited, calibration method, spectra (both survey and high resolution), and further processing done on the samples (if any).
Apart from carbides (which are not present in your cases, of course), there is a couple of possible explaination in my opinion:
1. there is a large fraction of Csp2-Csp3 boundary defects forming localized clusters ( see for example Article Origins of sp(3)C peaks in C-1s X-ray Photoelectron Spectra ...
)
2. there has been some problem in peaks assignment due to charging.
It may help knowing how have you aligned the BE axis (to the Fermi level) of each spectrum