The concentration of atmospheric methane increased from around 360 ppbv at the last glacial maximum (similar to 20 ka before present) to about 700 ppbv in the pre-industrial era (similar to 200 years before present). The sources and/or sinks of methane must therefore have changed during this period; however, the relative sizes of the source- and sink-driven changes in methane concentration remain uncertain. We take the first "bottom-up" approach to identifying any chemical signals preserved in the ice record that could help us to determine these. Using an atmospheric chemistry-transport model, we explore the effects of source- and sink-driven changes in methane on a wide range of chemical species in the Antarctic boundary layer. Though we identify several potentially useful atmospheric signals, a simple and robust constraint on the sizes of the source- and sink-driven changes cannot be readily identified, owing to their preservation in the ice, limitations to the information they hold, and/or ambiguity surrounding their interpretation. This includes the mass-independent fractionation of oxygen isotopes in sulfates, and the concentration of formaldehyde, in which there has been considerable interest. Our exploration is confined to a domain in which NOx emissions and climate remain constant. However, given the uncertainties associated with the changes in these factors, we would anticipate that their inclusion would make it harder still to identify a robust signal. Finally, though formaldehyde cannot provide this, we propose how it might be used to synchronize the gas- and aqueous-phase Antarctic ice records and thus determine the relative phasing of glacial-interglacial changes in Southern Hemisphere CO2 and temperature.