Of course insects could hold their breath for years back then, before genetic entropy due to mans evil nature reduced them to the fragile shadows we see now.
Except that according to the requisite mythology, the so-called "fall" happened a good two thousand years or so before
the fantasy Flud. So if wanktard creationists want to go down that route, they've got that little problem to attend to.
That's before we take note of the fact that insects were around for 300 million years or more before humans, and include huge swathes of taxa that could not possibly have survived for any length of time subject even to shallow burial, let alone burial under several kilometres of briny silt. For that matter, out of a million-plus insect species alive today, there are fewer than 50 that regularly live in a marine environment. There is no indication in the fossil record, that this tiny proportion of insects living in a marine environment was ever significantly greater. For example, whenever larvae of aquatic insects happen to persist in the fossil record, the geological facies are always
indicative of the depositional environment being a freshwater one.
A case in point being provided by this paper:Coxoplectoptera, A New Fossil Order Of Palaeoptera (Arthropoda: Insecta), With Comments On The Phylogeny Of The Stem Group Of Mayflies (Ephemeroptera)]/i] by Arnold H. Staniczek, Günter Bechly & Roman J. Godunko, [i]Insect Systematics & Evolution
101-138 (2011) [Full paper downloadable from here
Staniczek et al, 2011 wrote:AbstractMickoleitia longimanus
gen. et sp.n. is described from the Lower Cretaceous limestone of the Crato Formation in Brazil. It is attributed to a new family Mickoleitiidae and a new fossil insect order Coxoplectoptera within the palaeopterous Ephemerida, based on the presence of an elongated costal brace. This fossil insect exhibits a very peculiar combination of derived characters like specialized forelegs with strongly elongated, free coxae, single-clawed pretarsus, and distinctly skewed pterothorax as in dragonflies. On the other hand, several plesiomorphies are present that exclude this taxon from modern Ephemeroptera, namely large hind wings with widened anal area and numerous cross veins that separate the elongate costal brace from the costal margin. Fossil larvae described by Willmann as larval Cretereismatidae are herein attributed to Mickoleitiidae fam.n., based on the shared presence of broad hind wing buds with distinctly broadened anal area, wing bud venation similar to the adult holotype, and subchelate forelegs with elongate free coxae. These larvae are also highly autapomorphic in the structure of their abdominal gills and laterally flattened body with vertically oval section that is unique within Ephemerida. On the other hand they possess plesiomorphic lateral wing pads with pronounced articulation like Palaeozoic pterygote larvae, while wing pads in modern insects are always secondarily fused to the tergum. A similar fossil larva from the Jurassic of Transbaikals was earlier described as Mesogenesia petersae
and classified within modern mayflies. It is herein attributed to Mickoleitiidae fam.n. Coxoplectoptera are recognized as putative sister group of modern Ephemeroptera based on the shared presence of only 7 pairs of abdominal gills, while Permoplectoptera still have retained 9 pairs of gills. The phylogenetic reclassification of the mayfly stem group by Willmann is critically discussed and modified.
This recently discovered fossil clade exhibits a number of features, both in adult and larval form, that immediately
call attention to themselves when examined by any entomologist or insect taxonomist more familiar with modern taxa. The larvae are particularly striking, and exhibit features that would not lead even an experienced taxonomist to consider them related to mayflies upon first examination. The larvae of Coxoplexctoptera look more like mantis shrimps than insect larvae, complete with raptorial forelegs and sabre like mandibles. They also possess shovel like head projections suggestive of a fossorial lifestyle, but one restricted to digging to shallow depths in river sediments. Their assignment as a sister clade to Ephemeroptera is based upon the possession of 7 pairs of abdominal gills (numerous other aquatic insect taxa have larvae with 9 pairs of abdominal gills, the Ephemeroptera being a notable exception) and 3 caudal filaments with dense coatings of swimming hairs (again a feature exhibited uniquely among modern insects by mayflies).
The adults exhibit features again indicating a relationship to Ephemeroptera, including wing venation that indicates a relationship thereto, but also features that set them apart, such as an oblique backward tilt to the pterothorax (the part of the thoxax bearing the wings), which indicates an earlier split of the Ephemeroptera and sister clades from the Odonata some time previously. Like the Odonata, the Coxoplectoptera have raptorial forelegs that are thrust forward by the re-orientation of the pterothorax, large compound eyes and fully functional mandibles (Ephemeroptera adults have non-functional mouthparts, at least amongst modern taxa). The more even distribution of wing area in Coxoplectoptera again suggests Odonata ancestry, but the venation is significantly different from that of stem group Odonata, and is much more akin to that seen in Ephemeroptera, though the hind wing venation is again different, courtesy of the reduction in wing size in many Ephemeroptera, and concomitant vein loss.