This works fine with species that swallow their prey as a whole and hence do not damage it severely. Examples are most finfish, some carnivorous polychaetes and gastropods, some deposit feeders or even filter feeders.
It becomes more difficult in taxa that chew their food, e.g. regular echinoids. The Antarctic sea urchin Sterechinus spp. feeds on bryozoans by crunching their calcareous sceleton and ingesting the pieces (Terpstra 1996). As there are >300 Antarctic bryozoan species, it is quite a challenge to identify the consumed species from this debris.
Rather hopeless are those species which crush or mince their food so that there is nothing left for proper visual identification, e.g. crabs or grazing snails.

If visual identification of food items in the stomach content is impossible, there are two options:
- Identification by other means, and
- Monitoring of feeding in the field

Identification by other means refers to taxon-specific biochemical markers. Plant material may be characterised by specific pigments (e.g. Abele-Oeschger et al. 1992, Bianchi & Findlay 1990) or fatty acids (e.g. Meziane et al. 1997, Napolitano et al. 1997). Material of autotrophic or heterotrophic origin may be identified by immunological antibody responses (e.g. Gorant et al. 1994, Mayfield et al. 2000) or even by DNA probes. As these approaches are laborious and expensive, they are less suitable for large scale quantitative investigations.

Monitoring of feeding in the field can be done in various ways. Reported are identifying the food item actually chewed on once in a while, e.g. in grazing echinoids (e.g. James 2000) or identifying feeding remains, e.g. empty shells along the path of a drilling snail (e.g. Burrows & Hughes 1990). Depending on the species in question other approaches may work, too, e.g. tracking food composition of a hard bottom grazer by the devoided "aisle" left behind the animal.