Using the COI and ITS rRNA gene sets, from the journals I've read, they are the best for molecular markers. The statistics use Bayesian inference to compare the previous research dataset with the likelihood that was obtained when measuring the current population, statistical parsimony and maximum likelihood. you can use this reference: https://www.researchgate.net/publication/324133190_DNA_barcoding_phylogeny_and_phylogeography_of_the_cyst_nematode_species_of_the_Avenae_group_from_the_genus_Heterodera_Tylenchida_Heteroderidae.

There is a problem with formaldehyde-based solutions and molecular techniques; while formaldehyde is a great option when we are talking of preserving sample morphology, it also inhibits PCR and DNA sequencing!

Therefore, you have to select previously what specimens are for molecular analyses and use ethanol, for example.

I don´t know about PCR in nematodes particularly, but a good option is always begin with universal primers (those that have been tested in a high number of independent taxa) e.g:

https://www.mbari.org/wp-content/uploads/2016/01/Folmer_94MMBB.pdf

These works for COI, which is a marker widely used for metazoa, but I suggest you to check on databases (BOLD systems; Genbank) about disponibility of sequences of your taxa of interest (since basically you will have to compare your own sequences with those already available).

If you think you have a new species in your hands, look in the litetature about close relatives e.g: species from the same genus? family? Sure the morphology already gives you some clues about the place of this species!

DESS is the most commonlhy used preservative for molecular studies. you can use it for smaples directly.