On my experience, insect gut microbiota tends to have very low diversity.

Things that you could try (if not too much hassle and cost) would be:

Sequence again your DNA samples but without diluting too much your DNA (For Illumina MiSeq I usually go for ~20ng DNA, but I have done it at up to 125 ng DNA). This because if only one genus is predominant in your samples, by having more DNA you have more chances to "see" other low abundant taxa. It will be something like increasing your sample size of DNA fragments to be sequenced.

Change the 16S rRNA database you are using. Specially if you are using GreenGenes database, which its been long since it was updated, try using SILVA. By changing from GG to SILVA I was able to observe two or three more times the amount of ASVs/OTUs in gut microbiota samples from different hosts, including insects, which some times increased the diversity of species observed.

I have seen both high and low diversity of insect/spider gut diversity in published data (and a little of mine). Since Arsenophonus are possibly inctracellar symbionts your data might indicate a higher sampling of tissues rather than the contents of the intestinal content of the thrips. Can you find a way to only sample the gut contents (feces?)?

Another thought - were the trips stressed before they were killed for analysis? If so, it might have caused increased growth of the Arsenophonus and inhibition of other species.

Arsenophonus, in the Gammaproteobacteria, is a diverse symbiotic lineage. Most of the Arsenophonus strains are facultative. Some strains of Arsenophonus may behave as obligate symbionts. Obligate symbionts originate from two endosymbiont genera, Arsenophonus and Sodalis clades. Also Sodalis bacteria show similar gradient of symbiosis as Arsenophonus. So, it is thought that Arsenophonus bacteria can be replaced by Sodalis bacteria. However, what keeps the symbiont consortia limited to the specific bacterial clade is unknown.