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

for the late-May–June period, so the ratio of precipitation to temperature (evapotranspiration) increased, which is consistent with the observed increasing ANPP. As a result, the observed late- May–June ratio of precipitation to temperature was positively correlated with ANPP (Table 3). In addition, ANPP was also affected by factors that did not trend over our study (Table 3). First, ANPP was greater with higher ANPP in the prior year. Second, sampling sites were selected to provide an elevation gradient, and ANPP increased with elevation. Finally, the Julian day at peak “green” biomass was negatively correlated with percent of ANPP comprised of grass and posi- tively correlated with late-May–June ratio of precipitation to temperature (Table 3). % Grass. Grass ANPP relationships are the same as for total ANPP, but forb ANPP responded somewhat differently (Table 3). While forb ANPP responded positively with prior year forb ANPP and elevation, it only weakly increased with the ratio of June precipitation to temperature, and current year grass ANPP had a strong negative effect, indicating potential competition. Therefore, while grass ANPP benefitted by the changing environmental condi- tions, forbs did not and were reduced by the increasing grasses. Senescence. Plant senescence in summer increased with summer temperature and the abundance of grass (Table 3). Summer precipitation, the summer ratio of precipitation to tem- perature, and elevation had no effect on senescence. Fall ANPP. Fall plant production increased as summer temperature decreased, i.e., less

summer senescence (Table 3). Fall production was not related to total ANPP (negative correla- tion, P < 0.13), indicating it is independent of the primary growing season (late-May–June) production. Table 3. Backward stepwise regression results, where variables are included if their P < 0.15 and the AIC with them declines. SIGNS r N P AIC Warm season (June–Sept) N - summer temperature + 0.30 94 0.004 452.78 summer temperature, elevation +,+ 0.36 94 0.003 450.28 ANPP - June precipitation/temperature + 0.32 136 0.0001 1512.24 June prec/temp, prev. ANPP +,+ 0.62 125 0.000001 1348.19 June prec/temp, prev. ANPP, elevation +,+,+ 0.63 125 0.000001 1346.63 Grass ANPP - Previous year grass ANPP + 0.70 125 0.000001 1327.37 Previous year grass ANPP, June prec/temp +,+ 0.73 125 0.000001 1318.86 Forb ANPP - Previous year forb ANPP + 0.44 125 0.000001 1165.12 Previous year forb ANPP, grass ANPP +,- 0.47 125 0.000001 1163.51 Previous year forb ANPP, grass ANPP, elevation +,-,+ 0.51 125 0.000001 1159.17 Previous year forb ANPP, grass ANPP, elevation, June prec/temp +,-,+,+ 0.52 125 0.000001 1158.91 Peak biomass—day % grass - 0.27 112 0.004 977.10 June prec/temp + 0.35 112 0.008 972.94 Senescence - Jul–Aug temperature + 0.51 112 0.000001 -75.76 Jul–Aug temperature, grass ANPP +,+ 0.53 112 0.000001 -76.95 Fall ANPP - Jul—Aug temperature - 0.31 32 0.07 -57.71 https://doi.org/10.1371/journal.pone.0243496.t003 PLOS ONE Bunchgrass prairie and climate change

PLOS ONE | https://doi.org/10.1371/journal.pone.0243496 December 23, 2020 12 / 21