The Benefits of Recent Warming for Maize Production in High Latitude China

The benefits of recent warming for maize production in high latitude China

Supporting figures

Figure S1The locations of Heilongjiang Province, weather stations and the six zones, which were divided by ≥0 °C accumulated temperature:1stzone (>2700°C), 2ndzone (2500-2700°C), 3rdzone(2300-2500°C), 4thzone (2100-2300°C), 5thzone (1900-2100°C), and 6thzone(<1900°C). The dots are the thirty-two meteorological stations in Heilongjiang Province

Figure S2 The simulated and observed dry matter in three field experiments. The solid and dashed line represents the simulated total above-ground biomass and grain biomass, respectively. The black and white circle indicates measured total above-ground biomass and grain biomass, respectively. Field experiments were conducted in 2009 on 850 Farm (132.72° E, 45.73° N), 852 Farm (132.65° E, 46.41° N) and Qixing Farm (132.72° E, 47.29° N) in Heilongjiang Province for rainfed maize. At 850 Farm, Zhendan 37 was sown on 2nd May with density of 64,455 plants per ha. At both 852 and Qixing farms, maize variety of Suiyu No. 7 was used while it was sown on 7th and 1st May, respectively. Maize density was 64,530 plants per ha at 852 Farm and 65,000 plants per ha at Qixing Farm. Above-ground biomass of all plots was measured at three-leaf (V3), six-leaf (V6), silking stage (R1), milk stage (R3) and maturity (R6) on 850 and 852 farms and at V3, V6, ten-leaf (V10), R1 and R6 on Qixing Farm. We used the Hybrid-Maize model with local weather data from nearby meteorological recorded stations to simulate the total and grain biomass in these three field experiments

Figure S3 The sowing date for 1980s and 2000s varieties. The sowing date for 1980s varieties was collected from nine agricultural meteorological experiment stations. The sowing date for 2000s varieties was from twelve agricultural meteorological experiment stations. There was no significant difference for average of sowing date between 1980s and 2000s varieties according student’s t-test at 95% confidence level

Figure S4Trends in the simulated crop stages from sowing to flowering (vDays), from flowering to maturity (rDays) and total (tDays), solar radiation and precipitation at the 1st, 2nd, 3rd and 4th accumulated temperature zones during maize growth season (1980-2009) with 1980s varieties. Straight lines show the liner trends against year. **Significant at P < 0.01; *Significant at P0.05

Figure S5The simulated gross assimilation and total respiration for maize at the 1st, 2nd, 3rdand 4thaccumulated temperature zones with 1980s varieties, respectively. **Significant at P0.01

Figure S6The simulated grain yield with 1980s varieties at 1st accumulated temperature zone. (a) Simulation with 1980-2009 weather data. (b) Scenario A. simulation was performed with 1980-2009 weather data when only the temperature data was assumed no change since 1980. (c) Scenario B. Simulation with 1980-2009 weather data when the solar radiation and precipitation was assumed no change since 1980. *Significant at P < 0.05

Figure S7Crops yield (a), harvested area (b), and total production (c) from 1980 to 2010 in Heilongjiang province

Figure S8Expansion of maize area at some high-latitude areas: Canada, Russian Federation and Denmark. Statistical data in Canada and Russian Federation was from Food and Agriculture Organization of the United Nations ( while it was from Denmark National Statistical Databasein Denmark (

Supportingtables

TableS1 The typical varietiesat the first four accumulated temperature zonesin 1980s, 1990s and 2000s, respectively

TableS2 The annual mean, trend and highest positiveand negative departure of climate factors during 1980-2009 in Heilongjiang Province

TableS3The percent of simulated grain yield change per decade (%) for both 1980s and 2000s maize varieties assuming no warming happened from 1980 to 2009

TableS4The available growing degree days between frosts (GDDavailable) (≥10 °C) for maize growth, and the growing degree-days (GDD) of varieties adopted by local farmers at 5th and 6th accumulated temperature zone in 1980s, 1990s and 2000s, respectively. At the 5th and 6th zones in 1980s and 1990s because of the low GDDavailable and no appropriate varieties, there was no maize production in practice

1

Figure S1The locations of Heilongjiang Province, weather stations and the six zones, which were divided by ≥0 °C accumulated temperature:1stzone (>2700°C), 2ndzone (2500-2700°C), 3rdzone(2300-2500°C), 4thzone (2100-2300°C), 5thzone (1900-2100°C), and 6thzone(<1900°C). The dots are the thirty-two meteorological stations in Heilongjiang Province

Figure S2 The simulated and observed dry matter in three field experiments. The solid and dashed line represents the simulated total above-ground biomass and grain biomass, respectively. The black and white circle indicates measured total above-ground biomass and grain biomass, respectively. Field experiments were conducted in 2009 on 850 Farm (132.72° E, 45.73° N), 852 Farm (132.65° E, 46.41° N) and Qixing Farm (132.72° E, 47.29° N) in Heilongjiang Province for rainfed maize. At 850 Farm, Zhendan 37 was sown on 2nd May with density of 64,455 plants per ha. At both 852 and Qixing farms, maize variety of Suiyu No. 7 was used while it was sown on 7th and 1st May, respectively. Maize density was 64,530 plants per ha at 852 Farm and 65,000 plants per ha at Qixing Farm. Above-ground biomass of all plots was measured at three-leaf (V3), six-leaf (V6), silking stage (R1), milk stage (R3) and maturity (R6) on 850 and 852 farms and at V3, V6, ten-leaf (V10), R1 and R6 on Qixing Farm. We used the Hybrid-Maize model with local weather data from nearby meteorological recorded stations to simulate the total and grain biomass in these three field experiments

Figure S3 The sowing date for 1980s and 2000s varieties. The sowing date for 1980s varieties was collected from nine agricultural meteorological experiment stations. The sowing date for 2000s varieties was from twelve agricultural meteorological experiment stations. There was no significant difference for average of sowing date between 1980s and 2000s varietiesaccording student’st-test at 95% confidence level

1

Figure S4Trends in the simulated crop stages from sowing to flowering (vDays), from flowering to maturity (rDays) and total (tDays), solar radiation and precipitation at the 1st, 2nd, 3rd and 4th accumulated temperature zones during maize growth season (1980-2009) with 1980s varieties. Straight lines show the liner trends against year. **Significant at P < 0.01; *Significant at P0.05

1

Figure S5The simulated gross assimilation and total respiration for maize at the 1st, 2nd, 3rdand 4thaccumulated temperature zones with 1980s varieties, respectively. **Significant at P0.01

1

Figure S6The simulated grain yield with 1980s varieties at 1st accumulated temperature zone. (a) Simulation with 1980-2009 weather data. (b) Scenario A. simulation was performed with 1980-2009 weather data when only the temperature data was assumed no change since 1980. (c) Scenario B. Simulation with 1980-2009 weather data when the solar radiation and precipitation was assumed no change since 1980. *Significant at P < 0.05

Figure S7Crops yield (a), harvested area (b), and total production (c) from 1980 to 2010 in Heilongjiang province

Figure S8Expansion of maize area at some high-latitude areas: Canada, Russian Federation and Denmark. Statistical data in Canada and Russian Federation was from Food and Agriculture Organization of the United Nations ( while it was from Denmark National Statistical Databasein Denmark (

TableS1 The typical varietiesat the first four accumulated temperature zonesin 1980s, 1990s and 2000s, respectively

1

TableS2 The annual mean, trend and highest positiveand negative departure of climate factors during 1980-2009 in Heilongjiang Province

Highest positive departure= mean of the highest-five years from 1980 to 2009 –meanfrom 1980 to 2009.

Highest negative departure= mean of the lowest-five years from 1980 to 2009–mean from 1980 to 2009.

**Significant at P < 0.01.

1

TableS3The percent of simulated grain yield change per decade (%) for both 1980s and 2000s maize varieties assuming no warming happened from 1980 to 2009

TableS4The available growing degree days between frosts (GDDavailable) (≥10 °C) for maize growth, and the growing degree-days (GDD) of varieties adopted by local farmers at 5th and 6th accumulated temperature zone in 1980s, 1990s and 2000s, respectively. At the 5th and 6th zones in 1980s and 1990s because of the low GDDavailable and no appropriate varieties, there was no maize production in practice

1