Two experiments were performed: 1) four V x four N rates x two W; 2) four N rates x 4 S (30, 60, 90 and 180 kg ha-1) x two PA (line, broadcast) x two W. In both experiments N rates were 0, 60, 120, and 240 kg N ha-1, applied as urea (46% N) in two equal splits at planting and early tillering, respectively. Water treatments consisted on two water heights (from panicle initiation): 5 and 15 cm. Evaluations included: viable pollen count at anthesis, SPAD readings at flag leaf, yield and yield components, and percent blanking at harvest. Air (at canopy level) and water temperature were measured during the whole season (November-March) at 15-min interval, using a wireless sensor system. Pollen viability was evaluated by dissecting anthers and staining pollen grains with potassium iodine (KI). Total and viable pollen count was done by imagery analysis.
Pollen viability was significantly affected by variety, N rate, and water height. Neither seed rate nor planting arrangement affected pollen viability. A decrease in percent viable pollen was observed with N rate; this decrease was steeper the lower the water height. Pollen viability positively correlated with grain yield and negatively with blanking. Spikelet sterility was not completely explained by pollen viability indicating that other variables besides it are also responsible for blanking.