One impediment to accurate national estimation of N2O is the difficulty in upscaling N2O measurements made at discrete points to larger field and regional scales. Our objective was to estimate N2O emissions during snowmelt in 2002 for a township (approximately 92 km2) near Laird, Saskatchewan. Chamber measurements were made at 12 sites in the township: four fields with canola (Brassica napus L.) residues, four with pea (Pisum sativum L.) residues, three with wheat (Triticum aestivum L.) residues, and one field that received cattle manure. Ten sampling chambers were used at each site, and N2O samples were made on 7 d during the snowmelt period (from 2002 Apr. 03 to Apr. 17). Cumulative N2O emissions during the 14 days of the snowmelt period differed between crop residue types: cumulative emissions from sites with wheat residues were 105.6 g N2O-N ha-1 and were significantly higher (P < 0.1) than those from fields with pea and canola residues (79.6 and 75.2 g N2O-N ha-1 respectively). The single manured site assessed had the highest cumulative emissions of 330.7 g N2O-N ha-1. The crop-specific emissions from the chamber-based measurements were multiplied by the area of each crop type in the township to calculate an area-weighted value for emissions. Cumulative emissions were 93.4 g N2O-N ha-1 for the chamber-based measurements. Water-filled pore space and soil temperature were not significantly correlated with cumulative emissions. Cumulative emissions from sites with fall nitrate levels below 8.0 kg ha-1 were consistently lower than those above this threshold. The emissions for the Laird township were well below the emissions calculated for most other studies in the Prairies and in central Canada. The lower emissions were probably due to low soil water contents and soil nitrate levels in the fall of 2001 and below normal snowfall in the winter of 2001â??2002. This reinforces the importance in antecedent moisture conditions and soil N levels for modeling of emissions at snowmelt.
