The availability of atmospheric carbon dioxide is the sine qua non for all plant growth and thence for all marine and terrestrial life forms. The purpose of this paper is to show that proposed reductions in anthropogenic emissions of carbon dioxide (CO2) to below the level of observed annual incremental biospheric absorption of those emissions would reduce the growth of the basic feedstock of all life forms. Agronomists have for long known and demonstrated in controlled experiments both in greenhouses and in field studies the dramatic impact of increases in its level on crop yields. These studies have all been local. The regression analysis here of historic data on global food production shows it may well be more dependent on increases in the availability of atmospheric carbon dioxide (henceforth written as [CO2]) than on changes in fertilizer consumption and global mean temperature (GMT). This implies that if the drastic reductions in total anthropogenic emissions of CO2 to be sought at Copenhagen (December 2009) are adopted and applied, they will, even if they aim at only a 60% reduction on the 2000 global level by 2050, bring emissions to below the incremental volume of their biospheric absorption. That could seriously imperil growth of global food production. We show how in its role as a fertilizer that raises global Net Primary Productivity (NPP), increases in [CO2] have a natural negative feedback mechanism that offsets a large proportion of growing emissions: more [CO2] causes more plant growth, but more plant growth takes up more CO2 thus limiting the further rise. This contrasts with the unproven positive feedback assumed in all models deployed by the IPCC whereby, allegedly, rising [CO2] will result in falling biospheric absorption and ever larger increases in [CO2]. We show there is no sign in the observations since 1958 of "saturation" of the capacity of the planet to continue absorbing more than half of all anthropogenic emissions of CO2, so there is no evidence for the IPCC's positive feedback. Biospheric absorption of increases in anthropogenic CO2 emissions would only have to increase from the average 57% of all anthropogenic emissions from 1958 to 2008 to 60% to achieve the likely Copenhagen 60% emissions reduction target. The rapid growth of absorption of total anthropogenic emissions to over 6% p.a. between 1997 and 2006 relative to total emissions growth at 2.6% p.a. over that period (Le Que ?re ? 2008) confirms this manner of attaining the Copenhagen target is easily attainable-and helps to explain the growth of food production at rates in excess of global population growth. It also limited the growth rate of aggregate [CO2] between 1958 and 2008 to only 0.41% p.a. Our results show that with warming in the absence of growing carbon fertilization, agricultural production could be less by more than 10% by 2080 than at present (2007:Table 5.8). That means starvation for most of a global population likely then to be at least 50% larger than now.