Not quite sure, but you're bound to find all kinds of speculations on the Web. The question is which ones to believe. You might want to look it up in Toponymie générale de la France (link) in a largish library. The problem with words with obscure or unknown etymologies is that they attract unqualified word puzzlers to them. That goes doubly so for topnonyms. In one 19th century book I consulted on Google Books, the author purposefully said he was not going to discuss the Langadocian suffix -argues at all.

I did find a nice map of the distribution of placenames ending in -argues (link) in the French Wikipedia article on Toponomie française (link).
 Quote:
Probablement dérivé de ce suffixe, le suffixe -anicum, au pluriel -anicos, qui sous-entend le terme agros (= champs), se retrouve dans de nombreux toponymes du sud de la France sous la forme -argues : Baillargues, Marsillargues, Olargues, Vauvenargues.

Herre it is suggested that it is a derivation of the common Latin suffix -anicum which usually desingates a 'field'. Though elsewhere, I found an alternate etymology 'petit domaine' (link). Whence the intrusive r? Again, I'm not sure, but it might help to look at the history of Occitan (and Provençal). I do know that the southern French r is more likely to be trilled than the northern more uvular r. I hope this helps, and good luck.

Ceci n'est pas un seing.