Net primary production (NPP) in dryland ecosystems is co-limited by water and nutrient availability. Climate change models project a shift towards a more variable precipitation regime that includes a higher frequency of extreme events punctuated by longer drying periods. Meanwhile, nitrogen (N) deposition is increasing through anthropogenic inputs. Desert grasslands are especially sensitive to changes in water and nutrient availability; however, long-term consequences on NPP are not well understood. Since 2006, we have simulated increased precipitation variability in a northern Chihuahuan Desert grassland by experimentally altering the size and frequency of rain events during the summer monsoon (July-September). Irrigation treatments are applied as either 5 mm weekly events (“small”; n=12) or 20 mm monthly events (“large”; n=3). Ambient precipitation is received by all plots year-around. To simulate increased nitrogen deposition, two quadrats per plot are fertilized with 50 kg N ha-1, applied annually ahead of the growing season. Seasonal aboveground NPP (ANPP) is measured after each growing season in fertilized and unfertilized quadrats. Since 2012, we have also measured annual root production (0-30 cm depth) as a proxy for belowground NPP (BNPP) using root ingrowth donuts. In August 2009, a lightning-caused wildfire burned through the experiment.
Results/Conclusions
During 2006-2014, mean ANPP ranged from 3.4 g m-2 in fertilized irrigation control plots (2011) to 382.0 g m-2 in fertilized large irrigation plots (2007). Unsurprisingly, the lowest mean ANPP corresponded with the lowest monsoon rainfall year (85.8 mm, 2011); however, this correlation was not consistently observed across all years. During 2012-2014, mean BNPP ranged from 37.9 g m-2 in irrigation control plots (2012) to 194.2 g m-2 in small irrigation plots (2013). We found a weak positive correlation between ANPP and BNPP (r = 0.25). Multilevel model analyses found that when compared to controls, the small irrigation treatment significantly affected ANPP only in 2008, while the large irrigation treatment was significant only in 2013, implying other factors may be involved in driving ANPP in our system. BNPP differed significantly between the two irrigation treatments during 2012-2014, with consistently higher rates of root production occurring in the small irrigation plots. N fertilization significantly affected ANPP in pre-fire years 2007 and 2008, and in 2013 when monsoon rainfall was 64% above the 2006-2014 average, suggesting N fertilization positively impacts ANPP when fire disturbance is not a factor.