It depends on the isotopic baseline in your food web. Primary producers can have a very wide range of d15N, and if the predators specialise on grazers which in turn specialise on a low d15N source, their d15N can be very low as well.
Maarten's point about the importance of isotopic baselines is demonstrated by the attached paper by Cremona et al 2010. I have also attached a papers by Anderson and Cabada 2009 and Gauthier et al 2016 that you should find of interest. If you already have these papers my apologies.
There are factors other than baselines that can explain the phenomenon of low trophic level inferences for predatory aquatic invertebrates. Ontogeny is one possible avenue to consider - predatory invertebrates may change their diet as they grow. While this is possible, we have also noted strange inferred trophic levels for predatory invertebrates in rivers in South Africa. Close inspection of the 13C values undermines the notion that sympatric invertebrates have the same baseline, but unravelling this is complex. They may have completely different baselines depending on whether they predate scrapers, filterers etc, and these strategies can have substantially different diet isotope values. This mechanism invokes constrained rather than generic predator prey relationships within the invertebrate guild.
Look for the Honours degree thesis by Lomarie Cathleen Janse van Rensburg at the University of Pretoria.