Ocean-atmosphere chemistry on Earth has undergone dramatic evolutionary changes through its long history, with potentially significant ramifications for the emergence and long-term stability of atmospheric biosignatures. Though a great deal of work has centered on refining our understanding of false positives for remote life detection, much less attention has been paid to the possibility of false negatives, that is, cryptical biospheres that are widespread and active on a planet's surface but are ultimately undetectable or difficult to detect in the composition of a planet's atmosphere. Here, we summarize recent developments from geochemical proxy records and Earth system models that provide insight into the long-term evolution of the most readily detectable potential biosignature gases on Earth - oxygen (O2), ozone (O3), and methane (CH4). We suggest that the canonical O2-CH4 disequilibrium biosignature would perhaps have been challenging to detect remotely during Earth's ~4.5 billion year history and that in general atmospheric O2/O3 levels have been a poor proxy for the presence of Earth's biosphere for all but the last ~500 million years. We further suggest that detecting atmospheric CH4 would have been problematic for most of the last ~2.5 billion years of Earth's history. More broadly, we stress that internal oceanic recycling of biosignature gases will often render surface biospheres on ocean-bearing silicate worlds cryptic, with the implication that the planets most conducive to the development and maintenance of a pervasive biosphere will often be challenging to characterize via conventional atmospheric biosignatures.