Preliminary response of primary production and community composition to precipitation variation in a temperate grassland

Background/Question/Methods

Grassland ecosystems are water-limited, and show the highest interannual ANPP variability across biomes. Changes in annual amounts or seasonality of rainfall may interact with soil texture to impact grassland ecosystem functions including net primary productivity (NPP). We manipulated precipitation on an upland silty clay and lowland clay soil to a) replace potential evapotranspiration each week, minimizing intra-annual precipitation variability (ET), or b) altering precipitation seasonality by excluding 50% of the winter or spring precipitation (WD or SPD) in a restored perennial grassland in central Texas, USA. We hypothesized that ANPP variability may be accounted for by precipitation seasonality and/or water limitation to productivity mediated by soil texture. We predicted that water limitation will be higher in the upland than the lowland soil, because of their lower water holding capacity.  We compared changes in soil moisture, ANPP and species cover across the manipulations to an ambient precipitation treatment (C). 

Results/Conclusions

ET replacement increased soil volumetric water content across both soils by 17% compared to C (p=0.003), and reduced the CV of soil water content from 25 to 15%. After one year of precipitation manipulations, across both soils, ET replacement increased total ANPP 30% (p<0.005) compared to C, WD or SPD treatments. Total ANPP of seasonal treatments did not differ from C (p>0.05).  Increased grass ANPP (p<0.05) explained the total ANPP increase with ET replacement. Forb ANPP did not differ between treatments. Grass ANPP was about three times higher in lowland clay compared to upland silty clay soils across precipitation treatments (p=0.007), but forb ANPP did not differ between soil types. ET replacement increased mid-season total plant cover by 15% compared to C (p=0.004). Three forb and four grass species accounted for over 80% of total plant cover. Only the cover of a C₄ midgrass Bouteloua curtipendula increased with soil volumetric water content (R²=0.56, p=0.005), whereas the cover of the dominant grass, Nassella leucotricha did not change across soils and treatments. Preliminarily, annual rainfall amount limited ANPP in this grassland by limiting growth, soil moisture predicted the abundance of a sub-dominant but not a dominant C₄ grass, and 50% spring or winter rainfall exclusion was inadequate to further limit ANPP in the first year, and water limitation was similar on these soils.