, unloading of the phloem at the sinks, or utilization for growth of sinks. To determine the relative importance of leaf versus whole plant level limitations on carbohydrate utilization at elevated CO 2 , and the possible effects on the regulation of photosynthetic capacity, we constructed a treatment system in which we could expose single, attached, soybean leaflets to CO 2 concentrations different from those experienced by the rest of the plant. The single leaflet treatments had dramatic effects on the carbohydrate contents of the treated leaflets. However, photosynthetic capacity and rubisco content were unaffected by the individual leaflet treatment and instead were related to the whole plant CO 2 environment, despite the fact that the CO 2 environment around the rest of the plant had no significant affect on the total non-structural carbohydrate (TNC) contents of the treated leaflets. These results necessitate a re-evaluation of the response mechanisms to CO 2 as well as some of the methods used to test these responses. We propose mechanisms by which sink strength could influence leaf physiology independently of changes in carbohydrate accumulation.Key-words: Glycine max; Fabaceae; acclimation; carbohydrates; carbon dioxide; photosynthesis; rubisco; soybean.
INTRODUCTIONElevated CO 2 increases photosynthetic rates in the short term (minutes to hours) but over the long term (days to weeks) photosynthetic capacities often decrease in response to elevated CO 2 [for a review, see Griffin & Seemann (1996)]. Photosynthetic acclimation to CO 2 may result from feedback inhibition of photosynthetic enzyme synthesis when carbohydrate utilization does not match demand (Stitt 1991). A carbohydrate feedback signal could be generated in response to a limitation in carbohydrate utilization at many different points; export of triose phosphates from the chloroplast, sucrose synthesis and phloem loading, transport in the phloem, unloading of the phloem at the sinks, or carbohydrate utilization for growth of sinks. Studies of carbohydrate allocation in plants tend to assume that one or another of these factors constitutes the major limitation (Farrar 1996). However, control theory suggests that the most efficient utilization of resources results when a plant adjusts the capacities of each limiting step so that the overall process is equally limited by all factors (Farrar 1996). In this case, we would expect that both whole plant and leaf level limitations to carbohydrate utilization would contribute to carbohydrate feedback in plants grown at elevated CO 2 .Most tests of the carbohydrate feedback hypothesis have applied rather drastic and unnatural treatments to increase the carbohydrate content of leaves, for example feeding sugars to leaves or intact plants (Krapp, Quick & Stitt 1991;Krapp et al. 1993;van Oosten, Wilkins & Besford 1994;Winters et al. 1994;Jones, Lloyd & Raines 1996), cold girdling of petioles (Krapp et al. 1993;Krapp & Stitt 1995), heat girdling of petioles (Goldschmidt & Huber 1992), and transforming plant...