The idea of planting trees to reduce carbon emissions is often discussed. The scope of the problem can be nicely illuminated at the global scale. If all land use changes of the previous two centuries were reversed, then carbon accumulation in the atmosphere would be 80 PgC less than the present, leading to a global cooling of 0.4°C. By contrast, total deforestation could add as much as 400 PgC to the atmosphere, leading to a global warming of 2°C. Although these are rough estimates they indicate quite clearly that global scale management of forests, particularly in terms of increasing afforestation and reducing deforestation, has a part to play in future global climate.The global nature of the problem becomes clear when considering sequestration at the country scale. If the whole of the UK were reforested then this would be equivalent to a sequestration of about 1 PgC. If this directly impacted the atmospheric CO2 concentration then there would be a global cooling of 0.005°C. The approximate annual rate of carbon uptake would be 1 PgC, divided by the time to forest maturity. So if the time to maturity were 25 years, then the annual cooling would be 0.0002°C. Annual UK emissions of fossil fuel carbon to the atmosphere cause a global warming of ∼0.0003°C yr−1, while current global emissions of 7.9 PgC yr−1 may cause an estimated 0.016°C yr−1 of warming, using the above rule of thumb. The small contribution of the UK to the global change indicates that global mitigation of climatic change can only be achieved by internationally concerted action to reduce carbon emissions.The oceanic and terrestrial sinks for carbon currently sequester 60% of anthropogenic emissions, but this fraction is likely to decline through the 21st century. There is limited potential to stimulate this sequestration, such as with iron fertilisation of the ocean. Increasing iron stimulates plankton growth and carbon uptake — initially. Herbivores then increase in abundance leading to little change in plankton density and subsequent additions of iron will cause only a limited stimulation of carbon sequestration. There is some suggestion, however, that restoring fish diversity in the over-fished global oceans can enhance production and the resistance of this production to perturbations. As was the case for afforestation the key to success will be concerted global action in controlling and enhancing species diversity.The natural carbon cycle could also be enhanced by human intervention higher up the carbon cascade (Figure 2Figure 2). For example, terrestrial NPP could be harvested to manufacture ethanol as a replacement, or partial replacement for petrol (gasoline) in vehicles. Brazil, for example, produces ethanol from sugar cane that replaces 20–25% of petrol in fuel. The plant source, such as sugar cane, grows and sequesters carbon which is then released again in the car exhaust gases. The ideal net result of this method is a neutral effect on the atmospheric CO2 concentration for using the ethanol, but there would still be CO2 accumulation from the fossil fuel component. At the current fossil fuel emission of 7.9 PgC yr−1, globally, and assuming 1.6 PgC yr−1 as emissions from vehicle transport, then a 22% use of ethanol, globally, would reduce fossil fuel emissions by 0.35 PgC yr−1, broadly equivalent to a cooling of 0.0007°C yr−1; although small this is more than twice the fossil fuel emissions of the UK. But there are hidden carbon costs in terms of fossil fuel requirements for growing, harvesting and converting sugar cane to ethanol, which means the process is not carbon neutral and, furthermore, ethanol has about 65% of the energy content of petrol.Figure 2The average carbon cascade from gross primary production (GPP), to net primary production (NPP), to net ecosystem production (NEP) to net biome production (NBP), for land and ocean, and in comparison with emissions due to human activities.View Large Image | View Hi-Res Image | Download PowerPoint SlideGrowing crops for biofuels is area intensive while mining for fossil fuels is much less so. Nearly 4% of Brazil is used to grow sugar cane for producing ethanol. In the UK, wheat would be the likely source of ethanol, but it is about as half as productive as sugar cane. There are 26 million registered car owners in the UK who, on average, drive 15000 km yr−1. Petrol with 22% ethanol has an average fuel economy of about 9 km l−1. Wheat produces 0.43 t ethanol ha−1. These data imply that one car owner driving an average annual distance would require annually the ethanol from 0.67 ha of wheat crop. For all drivers in the UK this adds up to nearly 75% of the land area of the UK to grow wheat for ethanol production. This is clearly not a practical approach to reducing carbon emissions, using the current technologies for producing biofuels such as ethanol. Increasing the area of crops for biofuel production would exert unacceptably severe impacts on the area under food crops and on the diversity of species in natural and semi-natural vegetation.There are many alternative opportunities to use primary production in order to mitigate the serious consequences of anthropogenic climatic change. The benefit will always be fractional where the real need is to find methods of significantly reducing fossil fuel emissions of carbon dioxide at source and globally. Time is ticking on and many models indicate that the natural oceanic and terrestrial sinks for carbon will slow through the current century. The reduced primary production of the terrestrial biosphere results in particular from warmer and drier conditions on land, increasing the frequency of droughts. Warming the oceans leads to stratification of the warm low density upper ocean water that increasingly reduces the upward movement of nutrients from the denser and colder water beneath. This stratification will reduce the supply of nutrients to the phytoplankton and the solubility of CO2 in the warmer upper waters, leading to a reduction in primary production.The natural sinks for human emissions of CO2 are therefore most likely to slow through the current century, increasing the fraction of CO2 that remains in the atmosphere and its associated warming potential. Area-based human modifications of the carbon cycle, such as by increasing the area of crops for producing biofuels, will exert some mitigating impact on emissions. Their use to mitigate all emissions is unrealistic. For example, with a business-as-usual rate of CO2 emissions fulfilling the aim of stabilising atmospheric CO2 concentration at 550ppm would require a land area about the size of the whole of South America to grow crops for biofuels.Methods for using or enhancing biological productivity to reduce human CO2 emissions, and their climatic consequences, can only be a small part of the mitigation strategy. The only answer is improved methods of reducing emissions from all CO2 sources plus a move away from an energy system based on carbon, such as to hydrogen.