COS 207-2 - Temperature response of nitrogen fixation and its respiratory carbon cost in legume-rhizobia symbioses

Abstract: Biological nitrogen (N) fixation is the main input of nitrogen to terrestrial ecosystems. Symbioses between N-fixing bacteria and certain species of vascular plants, including many legume species, have the capacity to bring large quantities of N into ecosystems, but the symbiotic N fixation (SNF) flux into ecosystems is often low because the symbioses are rare or do not fix large amounts of N. One reason N-fixing symbioses might be rare or not fix much N is that the carbon cost of SNF is high compared to acquiring N from the soil. The primary carbon cost of SNF is respiration in nodules (the root structures that house N-fixing bacteria). This carbon cost likely changes across environmental conditions, such as temperature, but a systematic assessment is lacking. SNF itself responds to temperature (unimodal), and respiration in other tissues such as leaves also responds to temperature (rising), so it stands to reason that the cost of SNF (respired C per N fixed) would respond to temperature. Using a new measurement system that allows continuous measurements of SNF and the CO₂ flux from nodules, we asked: How do SNF, nodule respiration, and the respiratory carbon cost of SNF change with temperature in legume-rhizobia symbioses?



We germinated N-fixing legumes and inoculated them with rhizobia. To measure SNF and its carbon cost, we placed whole plants or excised nodules in a chamber connected to a Picarro G2106 ethylene analyzer and a Licor 840a CO₂ analyzer, replacing 2% of the atmosphere with acetylene (which nitrogenase converts to ethylene). Before measuring temperature responses, we assessed how SNF and respiration change in nodules vs. whole plants at constant temperatures (25, 30, 35, and 40 °C). Over three hours, both SNF and respiration declined over time. Excised nodule rates declined somewhat faster than whole plants, and both declined faster at higher temperatures. We corrected for these declines over time when analyzing data from our two-hour temperature ramps with excised nodules. As we had observed previously, the optimum temperature for SNF was quite high (average ~32 °C). Nodule respiration rose with temperature, as expected, but the relative change in SNF was much stronger, so the carbon cost of SNF reached its minimum at a temperature that was only slightly cooler (average ~31 °C) than the optimum for SNF. In addition to advancing fundamental biology, these results can feed directly into models that assume temperature responses of SNF or its carbon cost.