COS 145-1 - Temperature Response of the Respiratory Carbon Costs of Symbiotic Nitrogen Fixation

Background/Question/Methods

Plants capable of symbiotic N fixation (SNF) can acquire N directly from atmospheric N2. However, SNF is costly due to the need to break the triple-bonded N2. Measurements of these costs in nodules have been close to biochemical predictions. However, these measurements have been carried out at constant temperatures whereas the cost might vary widely across temperature. To measure the cost of SNF across temperatures, we grew Robinia pseudoacacia seedlings under controlled conditions, inoculated the seedlings with symbiotic bacteria cultures, and fertilized the plants with limited levels of N (1.5 g N m−2 yr−1) but ample amounts of all other nutrients to promote SNF. Nodules were excised and measured for SNF and respiration continuously across 1-40°C over the course of 3 hours in a sealed chamber with acetylene, and after accounting for leakage and other factors, the rate at which acetylene was reduced to ethylene (measured with a Picarro G2106 laser) provided a measurement of nitrogenase activity. CO₂ flux in the chamber was synchronously measured by a Licor CO₂/H₂O analyzer to determine the temperature response of nodule respiration. Our analysis yielded temperature response curves for SNF, respiration, and the ratio of the two (the C cost of SNF).

Results/Conclusions

We found that the C cost of N fixation by Robinia pseudoacacia has a similar temperature optimum as SNF (34.3 and 34.8°C, respectively). At this temperature, the C cost of SNF is approximately 7.8 g C g N−1, matching biochemical predictions. However, we found that as temperature deviates from the optimum, this cost increases much more steeply than models predict. Furthermore, according to our analysis, C costs can rise to maxima of 150-300 g C g N−1 between 12-17°C whereas according to biochemical predictions, C costs of N fixation peak at about 12 g C g N−1. The paucity of knowledge on how SNF and its C cost respond to temperature has been a major constraint on global biogeochemistry and climate modeling. Temperature response functions for SNF and its cost are already in use in terrestrial biosphere models, despite few data for the temperature response of SNF itself and zero data for the temperature response of the C cost of SNF. Our results demonstrated a warmer temperature optimum as well as a steeper response than many terrestrial biosphere models currently use (i.e., 25°C), which will affect predicted primary production as atmospheric warming occurs.