Associative nitrogen fixation (ANF) in switchgrass (Panicum virgatum) across a nitrogen input gradient

Abstract

Associative N fixation (ANF), the process by which dinitrogen gas is converted to ammonia by bacteria in casual association with plants, has not been well-studied in temperate ecosystems. We examined the ANF potential of switchgrass (Panicum virgatum L.), a North American prairie grass whose productivity is often unresponsive to N fertilizer addition, via separate short-term 15N2 incubations of rhizosphere soils and excised roots four times during the growing season. Measurements occurred along N fertilization gradients at two sites with contrasting soil fertility (Wisconsin, USA Mollisols and Michigan, USA Alfisols). In general, we found that ANF potentials declined with long-term N addition, corresponding with increased soil N availability. Although we hypothesized that ANF potential would track plant N demand through the growing season, the highest root fixation rates occurred after plants senesced, suggesting that root diazotrophs exploit carbon (C) released during senescence, as C is translocated from aboveground tissues to roots for wintertime storage. Measured ANF potentials, coupled with mass balance calculations, suggest that ANF appears to be an important source of N to unfertilized switchgrass, and, by extension, to temperate grasslands in general.

Fig 1. Potential switchgrass root ANF rates from two sites (MI: Kellogg Biological Station in Michigan, USA and WI: Arlington Agricultural Research Station, Wisconsin, USA) across four time periods.

Roots were sampled from plots with a range of N fertilizer addition: A. unfertilized, B. 56 kg N ha^-1 yr^-1, and C. 196 kg N ha^-1 yr^-1. Note that the senescence period has a different scale to accommodate the much higher rates during that time period.

Fig 2. Soil potential ANF rates from two sites (MI: Kellogg Biological Station in Michigan, USA and WI: Arlington Agricultural Research Station, Wisconsin, USA) across four time periods.

Soils were sampled from switchgrass rhizospheres grown with a range of N fertilizer addition: A. unfertilized, B. 56 kg N ha^-1 yr^-1, and C. 196 kg N ha^-1 yr^-1.

Fig 3. Net N mineralization rates, measured via 28-d aerobic incubations, during four time periods.

Best-fit regression lines are shown where slopes were significantly different from 0. Soils were sampled from switchgrass rhizospheres at two sites (MI: Kellogg Biological Station, Michigan, USA and WI: Arlington Agricultural Research Station, Wisconsin, USA) and 3 fertilizer levels (unfertilized, 56 kg N ha^-1 yr^-1, and 196 kg N ha^-1 yr^-1). MI Post-fertilizer: y = 0.40*(x+1)^0.14, p = 0.04; MI Peak biomass: y = 0.07*log(x+1)+0.25, p = 0.002; MI Senescence: y = 0.04*log(x+1)+0.31, p = 0.03; WI Post-fertilizer: y = 0.002*x+0.299, p<0.02; WI Peak biomass: y = 0.127*(x+1)^0.15, p = 0.03; WI Senescence: y = 0.0012*x-0.015, p = 0.004.