The occurrence of three-dimensionally preserved, apparently mud-filled guts in otherwise flattened fossils of the Lower Cambrian Chengjiang biota has been widely interpreted as evidence of deposit feeding (i.e., mud eating). Such interpretation, however, is incompatible with the functional morphology of both the guts and the whole animals, and indeed with the accompanying taphonomy: structures preserved three-dimensionally in a compacted shale were, necessarily, resistant to compaction prior to degradational collapse or significant burial.

Three-dimensional guts also occur in otherwise flattened fossils of the Burgess Shale where they result from early diagenetic phosphatization of serially repeated midgut glands; thin-section examination reveals no included sediment. In modern arthropods, well developed midgut glands are characteristic of a predatory lifestyle, consistent with their expression in Anomalocaris, Sydneyia, Leanchoilia, Odaraia and various trilobites. In other words, three-dimensional guts in the Burgess Shale are associated exclusively with non-deposit feeders. Given the severe weathering of the Chengjiang fossils (to the extent of altering/removing all biominerals, including phosphate), the clay minerals that now constitute their guts are best interpreted as secondary replacements of earlier (phosphate?) permineralizations.

The more conventional case for Cambrian deposit feeding comes from trace fossils. Branching forms such as Phycodes are often seen as exemplars of such a habit; however, alternative interpretations such as surface detritus feeding are equally tenable. Indeed, direct evidence of deposit feeding is lacking for all Cambrian trace fossils. Combined with the delayed (Devonian/Carboniferous) appearance of crustacean-type sediment-bearing microcoprolites, it would appear that the strategy of ingesting huge volumes of organically dilute, largely indigestible material represents a relatively derived condition, not generally exploited until after the Cambrian.