Predicting the impact of changing CO2 on crop yields: some thoughts on food


Author for correspondence: Lewis H. Ziska Tel: +1 301 504 6639 Fax: +1 301 504 5823 Email:



  • Summary 607

  • I. Rising CO2 and agricultural crop yields 608
  • II. Expanding methodologies 608
  • III. Which methodology gives the ‘truest’ prediction   of future yields? 608
  • IV. Comparing responses among methodologies 610
  • V. Methodology vs future uncertainty 611
  • VI. Abiotic uncertainties 612
  • VII. Biotic uncertainties 613
  • VIII. Uncertainty vs methodology 614
  • IX. Modelers and experimentalists 614
  • X. Final thoughts 615
  • Acknowledgements 615

  • References 615


Recent breakthroughs in CO2 fumigation methods using free-air CO2 enrichment (FACE) technology have prompted comparisons between FACE experiments and ‘enclosure studies’ with respect to quantification of the effects of projected atmospheric CO2 concentrations on crop yields. On the basis of one such comparison, it was argued that model projections of future food supply (some of which are based on older enclosure data) may have significantly overestimated the positive effect of elevated CO2 concentration on crop yields and, by extension, food security. However, in the comparison, no effort was made to differentiate ‘enclosure study’ methodologies with respect to maintaining projected CO2 concentration or to consider other climatic changes (e.g. warming) that could impact crop yields. In this review, we demonstrate that relative yield stimulations in response to future CO2 concentrations obtained using a number of enclosure methodologies are quantitatively consistent with FACE results for three crops of global importance: rice (Oryza sativa), soybean (Glycine max) and wheat (Triticum aestivum). We suggest, that instead of focusing on methodological disparities per se, improved projections of future food supply could be achieved by better characterization of the biotic/abiotic uncertainties associated with projected changes in CO2 and climate and incorporation of these uncertainties into current crop models.