Climate change and primary production: Forty years in a bunchgrass prairie

value. Therefore, the ratio of precipitation to temperature tended to increase (Mann-Kendall: z = 1.11, P < 0.13) over the period of main plant growth during our study. This indicates reduced evapotranspiration [59,60], which is counter to the annual, Water Year, recharge period, and spring seasonal drying and general warming, and suggests increased annual plant production. There is a period of summer plant senescence (July–Aug decline in “green” plant biomass: Fig 2), as the vegetation is dominated by “cool-season” grasses (C3: >>95% of bomass). Dur- ing this period, temperature increased (Mann-Kendall: z = 3.37, P < 0.0004) and precipitation decreased (Mann-Kendall: z = -2.24, P < 0.01), which are consistent with the summer seasonal trends. This led to a decline in the precipitation to temperature ratio (Mann-Kendall: z = -2.60, P < 0.005), as expected with annual and summer seasonal drying and warming. This suggests that senescence should have intensified over our study. There is a brief period when plants grow again (late-Sept increase in “green” biomass: Fig 2) after summer senescence and before the onset of frosts (Oct). During this period, tempera- ture increased (Mann-Kendall: z = 2.20, p < 0.02) as expected for the summer season, and pre- cipitation was unchanged (Mann-Kendall: z = -0.86, p < 0.20), which is opposite the summer seasonal decrease and fall seasonal increase. This tended to reduce the precipitation to temper- ature ratio (Mann-Kendall: z = -1.11, p < 0.13), which should have decreased this renewal of plant production during our study.

NBR climate patterns are similar (r = 0.92–0.96) to observations at two nearby weather sta- tions (1989–1999 Round Butte: 18.9 km away, elevation = 926m, 704653E 5268696N, and 1978–1999 St. Ignatius: 13.2 km away. Elevation = 897m, 719607E 5244815N). Nitrogen available to plants. Total nitrogen (NO3- + NH4+) was not different between cool (Oct–May) and warm (June–Sept) seasons (Fig 3A: t = -1.14, P < 0.26). Total nitrogen tended to vary among years (1995–2017, Table 2: P < 0.09), especially for the warm season (Table 2: P < 0.02). Total nitrogen tended to exhibit an increasing trend over our study (Table 2: P < 0.09), especially for the warm season (Table 2: P < 0.0002). Cool season NO3-, NH4+ and total nitrogen exhibited no trends over our study (respective GLMM: P < 0.38, Fig 3. NBR soil nitrogen since 1995 for four sample sites studied longest. a) average cool (Oct–May) and warm (June–Sept) season soil total nitrogen at each site, and b) trends over time for warm season soil NO3-, NH4+, and total nitrogen across sites with GLMM results. https://doi.org/10.1371/journal.pone.0243496.g003 PLOS ONE Bunchgrass prairie and climate change PLOS ONE | https://doi.org/10.1371/journal.pone.0243496 December 23, 2020 7 / 21