Abstract

The productivity of wheat is highly vulnerable to climate change. Optimizing the sowing period of a crop may be one of the most important climate resilient strategies to optimize yield. First, the CERES-Wheat model was used to analyze effects of climate change on the optimum sowing window of wheat. Second, it was used to determine the optimum sowing window for different zones within Punjab state, India. The simulation results suggested that climate change has caused a shift in the optimum sowing window of wheat. The current (2006–2015 weather data) optimum sowing window is 22–28 October in north eastern Punjab, 24–30 October in central Punjab, and 21–27 October in south western Punjab. The rate of decrease in productivity with delay in sowing from the optimum sowing window by each day was lowest for north eastern Punjab (36.09 kg ha−1 day−1) and highest for south western Punjab (70.80 kg ha−1 day−1). The methodology followed in this study can be useful in determining the optimum sowing time of various crops.

INTRODUCTION

Wheat (Triticum aestivum L.) is very important from the food security point of view and is the second main source of the world's food energy and nutrition. India is the second largest producer of wheat in the world and produced 13% (95.85 million tonnes) of the world's total wheat out of 14% (30.47 million hectares) of global area under wheat during 2014 (FAO 2017). During the year 2014, Punjab produced 18% of Indian and more than 2% of global wheat from an area of 12% and 1% of total Indian and global wheat area, respectively (derived from Bhatti 2015 and FAO 2017), thus making it an important contributor to Indian and global wheat production.

Wheat being a cool season crop is generally sown from the last week of October to the last week of November in this part of India. Further delay in sowing, up to mid-December exposes the crop to high temperatures during the grain filling period, thereby significantly reducing the grain yield and yield components (Rane et al. 2007; Moshatati et al. 2012). Prasad et al. (2006) observed that increase in seed zone temperature beyond optimum resulted in increased germination rate but further increase resulted in reduced seed germination. Moreover, reduction in the rate of germination, seedling emergence, seedling with poor vigor, abnormal seedlings, altered radicle and plumule growth due to temperature stress have been reported in many plant species (Toh et al. 2008; Essemine et al. 2010; Piramila et al. 2012).

Zheng et al. (2012) highlighted that an appropriate sowing date is crucial to allow wheat to flower during the period with minimum stresses, such as frost and heat. Sowing time is a very important zero cost management strategy responsible for good productivity of crops. In general, the optimum sowing window for a crop in a region is determined by planting the crop at weekly or fortnightly intervals at various locations, and such coordinated experiments are conducted for more than one year. Presently, in Punjab state, the recommended time of sowing for long duration wheat varieties (mostly preferred by farmers) extends from the last week of October to the last week of November (Bhatti 2018).

In Punjab, significant changes in climatic parameters have been reported (Kaur et al. 2016) which may affect the optimum sowing window of the crops. In order to assess the effect of such changes on the sowing period of wheat, there is a need to conduct multi-location experiments for several years. The inherent weather varies from location to location within the state and, consequently, the optimum sowing window may also vary with location. The present study was conducted for three different locations in Punjab, representing different edaphic and climatic conditions (Figure 1). Conducting multi-location and multi-year field experimentation is time-consuming and not resource efficient. Crop growth simulation modeling offers an alternative for such conditions wherein the models can be used to simulate the experimental results by using historical weather data for impact assessment studies related to climate change, multi-location studies, and varying cultivation practices (Rosenzweig & Iglesias 1994; Wheeler et al. 1996; Pathak et al. 2006; Hundal & Kaur 2007). An analysis of historic weather data using a crop simulation model such as APSIM (Keating et al. 2003; Holzworth et al. 2014) can help to identify management strategies to achieve the optimum sowing date (Asseng et al. 2001) as attempted in the present study.

Figure 1

Long-term average of weather parameters (wheat season) of the study sites within Punjab. *Long-term actual sunshine hours data for south western Punjab is not available.

Figure 1

Long-term average of weather parameters (wheat season) of the study sites within Punjab. *Long-term actual sunshine hours data for south western Punjab is not available.

The present study was conducted using the CERES-Wheat crop growth model available within the framework of the DSSAT (Decision Support System for Agrotechnology Transfer). Models are designed to estimate production, resource use, and risks associated with crop production (Tsuji et al. 1994; Jones et al. 1998). In the present study, a 7-day sowing window was identified since farmers can complete wheat sowing within 7 days, as they have small farms (average land holding <4 hectares) and assured irrigation (98% of cultivated area is irrigated). However, farmers may take more/less than 7 days to complete sowing operations depending upon availability of inputs, vacation of field, marketing of previous crop, and family engagements. In such situations, farmers may be guided to adjust their farming operations to enable wheat sowing within this period. The objective of the present study was to determine the optimum sowing window, which, if wheat sowing is completed, will lead to enhanced productivity.

METHODS

Site description

The study was conducted for the three distinct sites in Punjab state, India. The first site was Ballowal Saunkhri (31.10°N; 76.39°E; 359 m above mean sea level) situated in the north eastern part of Punjab. This region is characterized by undulating topography along the foot hills of Shivalik mountains, shallow to deep soils, soil erosion by water, poor soil fertility, and shortage of irrigation water. The major cropping patterns followed in this zone are maize–wheat, rice–wheat, maize–pea–sunflower, etc. The second site was Ludhiana (30.90°N; 75.85°E; 251 m above mean sea level) situated in the central region of Punjab. The region is characterized by alluvial soils, high fertility levels, and good quality underground water for irrigation. Some places have problems with excessive seepage loss of water and nutrition deficiencies. The major cropping pattern followed in this zone is rice–wheat. The third site was Bathinda (30.21°N; 74.95°E; 208 m above mean sea level) situated in the south western region of Punjab. The region is characterized by alluvial soils affected by wind erosion and poor quality underground water in most places. The major cropping patterns followed in this zone are cotton–wheat and rice–wheat. The long-term average of weather parameters during the wheat season, for the three sites, are presented in Figure 1.

Data and model used

The CERES-Wheat model, available in the DSSAT framework was used to simulate the growth, development, and yield of wheat using daily weather data. The model takes into account the effects of weather, genotype, soil and crop management practices on growth of the crop. The input data required for simulation are weather, crop genetics, soil and crop management. The weather inputs include daily data on solar radiation (MJ m−2), maximum and minimum air temperature (°C) and precipitation (mm). Crop variety-specific parameters such as phenology, growth and yield attributes were incorporated into the model in the form of genetic coefficients (Ritchie et al. 1998). The crop and meteorological data required for calibration and validation of the model were recorded at Ludhiana, Punjab. The model was calibrated using the field experimental data generated during the crop season of 2010–2011. The field experiment consisted of wheat variety DBW 17 sown on six dates (28 October, 4, 11, 18, 25 November, and 2 December). The crop was raised using the agronomic practices recommended by Punjab Agricultural University, Ludhiana, Punjab. The genetic coefficients (Table 1) were derived by repeated iterations until a close match between simulated and observed phenology and yield was obtained. The data on soil properties were obtained from Vashisht et al. (2013) and Buttar et al. (2012) for Ballowal Saunkhri (north eastern Punjab) and Bathinda (south western Punjab), respectively, and the data for Ludhiana (Table 2) were obtained by analyzing soil samples. Meteorological data used in the model were obtained from the respective stations. Since the solar radiation data for south western Punjab (Bathinda) were not available, they were generated by the model.

Table 1

Genetic coefficients of wheat variety DBW 17

Parameter Value 
P1 V: Days, optimum vernalizing temperature, required for vernalization 14 
P1D: Photoperiod response (% reduction in rate/10 h drop in pp) 40 
P5: Grain filling (excluding lag) phase duration (°C.d) 660 
G1: Kernel number per unit canopy weight at anthesis (#/g) 16 
G2: Standard kernel size under optimum conditions (mg) 44 
G3: Standard, non-stressed mature tiller weight (including grain) (g dwt) 3.2 
PHINT: Interval between successive leaf tip appearances (°C.d) 110 
Parameter Value 
P1 V: Days, optimum vernalizing temperature, required for vernalization 14 
P1D: Photoperiod response (% reduction in rate/10 h drop in pp) 40 
P5: Grain filling (excluding lag) phase duration (°C.d) 660 
G1: Kernel number per unit canopy weight at anthesis (#/g) 16 
G2: Standard kernel size under optimum conditions (mg) 44 
G3: Standard, non-stressed mature tiller weight (including grain) (g dwt) 3.2 
PHINT: Interval between successive leaf tip appearances (°C.d) 110 
Table 2

Physical and chemical properties of the soil profile of the experiment site at Ludhiana (central Punjab)

Depth (cm) Sand (%) Silt (%) Clay (%) Bulk density (Mg/m3Hydraulic conductivity (cm/hr) pH Organic carbon (%) 
15 66.5 20.4 13.1 1.43 2.59 8.1 1.05 
30 71.4 17.6 11.0 1.43 2.59 8.0 1.04 
60 70.7 18.8 10.5 1.44 2.59 8.1 0.96 
120 70.7 18.8 10.5 1.44 2.59 8.1 0.96 
Depth (cm) Sand (%) Silt (%) Clay (%) Bulk density (Mg/m3Hydraulic conductivity (cm/hr) pH Organic carbon (%) 
15 66.5 20.4 13.1 1.43 2.59 8.1 1.05 
30 71.4 17.6 11.0 1.43 2.59 8.0 1.04 
60 70.7 18.8 10.5 1.44 2.59 8.1 0.96 
120 70.7 18.8 10.5 1.44 2.59 8.1 0.96 

The performance of the model was validated with the experimental data from the two crop seasons (2011–2012 and 2012–2013). The model was validated (Figure 2) by using indices like R2, root mean square error (RMSE), and normalized root mean square error (NRMSE):

Figure 2

Validation results of CERES-Wheat model.

Figure 2

Validation results of CERES-Wheat model.

Root mean square error

The root mean square error (RMSE) provides the weighted variations in errors (residual) between the predicted and observed values. Its value equal to zero shows a perfect fit between the observed and predicted data. It was calculated as follows: 
formula
where, P is value predicted by the model, O is observed value, and n is total number of observations.

Normalized root mean square error

Normalized root mean square error (NRMSE) provides a measure (%) of relative difference of predicted versus observed data. NRMSE was calculated as follows: 
formula
where, P is value predicted by the model, O is observed value, n is total number of observations, and Ō is mean of observed values. Jamieson et al. (1991) reported that the simulation is considered excellent, good, fair, and poor if NRMSE is <10, 10–20, 20–30, and >30%, respectively.

After validation, the CERES-Wheat model was used as a research tool to determine optimum sowing windows for different locations in Punjab (Figure 3). The model was used to simulate the crop growth using weather data recorded at respective locations, while considering sowing the crop on daily basis, starting from 15 October up to 30 November (i.e., 47 dates of sowing). All management practices were followed as recommended for irrigated wheat by Punjab Agricultural University, Ludhiana.

Figure 3

Seven days moving average of the simulated average wheat yield (kg/ha) (SD = standard deviation).

Figure 3

Seven days moving average of the simulated average wheat yield (kg/ha) (SD = standard deviation).

The results were presented after calculating average productivity of ten years and were plotted against the date of sowing (Figure 4). The 7-day moving averages (date-wise) of the simulated crop yields (averaged over years) were worked out to find a 7-day sowing window having maximum productivity (Figure 3). Similar to the present study, 15 days moving average was used to determine the optimum flowering period of wheat by Flohr et al. (2017). The average productivity of the identified 7-day sowing window was compared with the average productivity of a 5-week period (last week of October to last week of November) which is the recommended sowing period for long duration varieties in Punjab. A similar comparison was also done with average productivity during the last week of October (early sowing) and average productivity during 7 days following the optimum sowing window. The comparison helped to work out the difference in wheat productivity due to the change in sowing window from the currently recommended sowing time and due to a 1-week delay from the newly established window. In order to determine the effect of temperature on sowing window the mean temperature during the optimum sowing window was calculated from daily average temperatures. Average daily temperature was used because in Punjab during the sowing period of wheat, maximum and minimum temperatures do not have such extremes that could adversely affect sowing, germination, and emergence of the crop. In central Punjab, at Ludhiana (1970–2015), the wheat season mean of day-time temperature has remained significantly unchanged (y = 0.009x + 24.38; R2 = 0.028). On the other hand, during the same time period, the mean of night-time temperature has increased by almost 2 °C (y = 0.052x + 8.746; R2 = 0.636). Hence, it may be the daily average temperature which affected the germination and emergence of wheat. Therefore, in the present study, the analysis was done on the basis of average daily temperatures.

Figure 4

Simulated average yield (kg/ha) of wheat (SD = standard deviation).

Figure 4

Simulated average yield (kg/ha) of wheat (SD = standard deviation).

The temperature range was worked out from the temperature mean using the formula (temperature mean ± standard deviation) for different time periods (as per the availability of data) for three regions in the state. Remaining weather parameters like sunshine hours, wind speed, wind direction, and relative humidity were not considered in this study as these parameters do not have a direct impact on sowing of the crop. Although rainfall has a direct impact on sowing it was not studied as wheat is cultivated mostly under assured irrigation conditions in the state. However, while using the model, data on all the weather parameters including rainfall were used for simulation of the crop growth.

Regression analysis using R software (Figure 5) of average simulated yield (averaged over ten years) from the last decade was performed to work out the rate of change in productivity with a delay of 1 day in sowing. In regression analysis, the dates of sowing varied from location-specific optimum sowing windows to the end of November.

Figure 5

Change in wheat productivity (kg ha−1 day−1) with per day delay in sowing as per weather data of recent decade in different regions of Punjab.

Figure 5

Change in wheat productivity (kg ha−1 day−1) with per day delay in sowing as per weather data of recent decade in different regions of Punjab.

RESULTS AND DISCUSSION

Validation of the CERES-Wheat model

The validation results (Figure 2) showed that the observed and simulated phenological events matched quite well. On average, there was no difference between observed and simulated days taken to reach anthesis stages. Similarly, the average number of observed and simulated days taken for reaching the physiological maturity stage was also the same. In both cases, R2 was statistically significant at P-values shown in Figure 2. NRMSE for both phenological events was in excellent range (<10%) as proposed by Jamieson et al. (1991). Simulated and observed grain yield matched quite well as is evident from R2, P-value, and observed and simulated means. The NRMSE for grain yield was in good range (10–20%) according to Jamieson et al. (1991). The validation results showed that the model was calibrated quite satisfactorily and may be used for further applications.

Shift in optimum sowing window for wheat

Climate change has certainly influenced the optimum sowing window of wheat crop in Punjab.

In north eastern Punjab, the optimum sowing window of wheat was slightly affected due to climate change (Figure 3). In the decade 1986–1995, the optimum sowing window was from 21 to 27 October and during the next decade it was delayed by 1 day, i.e., 22–28 October, and thereafter it remained the same for the current decade, i.e., 2006–2015 (Table 3). The shift of only 1 day in north eastern Punjab might be due to the fact that weather data available for north eastern Punjab were for the recent 30 years (1986–2015) as compared to 46 years for central and 39 years for south western Punjab. In central Punjab, the maximum (4 days) shift in optimum sowing window was observed during earlier time periods, i.e., between 1970–75 and 1976–85 (Table 3). Similarly, in south western Punjab, maximum (4 days) shift in optimum sowing window was again observed during earlier time periods, i.e., between 1977–1985 and 1996–2005. During the recent two decades, the optimum sowing window has not changed much in north eastern and south western Punjab, thereby indicating that most of the shifts occurred during earlier times, i.e., during the 1970s up to the 1990s. The mean and range of daily average temperature for the optimum sowing window were calculated (Table 3) to ascertain the reason for change in the optimum sowing window. During 1986–1995, the mean temperature and its range in the optimum sowing window was 21.6 °C and 20.2–22.9 °C, respectively, for north eastern Punjab. These values were increased by nearly 1 °C during 1996–2005 and, thereafter, remained almost the same. In 2006–2015, during the week preceding the optimum sowing window the temperature range was higher by 1.1–1.9 °C and was lower by 0.6–1.0 °C during the succeeding week. These deviations in temperatures during the preceding and succeeding weeks from the temperature during the optimum sowing window might be the reason for lower simulated wheat productivity.

Table 3

Temperature mean and range of daily average temperature and identified optimum sowing windows

Weekly Time period
 
1970–1975 1976–1985 1986–1995 1996–2005 2006–2015 
 North eastern Punjab (Ballowal Saunkhri) 
Optimum sowing window 
 Temperature mean – – 21.6 (21–27 Oct)a 22.5 (22–28 Oct)a 22.3 (22–28 Oct)a 
 Temperature range – – 20.2–22.9 21.1–24.0 21.1–23.6 
Week preceding the optimum window 
 Temperature mean – – 22.8 23.2 23.8 
 Temperature range – – 21.0–24.7 22.0–24.4 22.2–25.5 
Week succeeding the optimum window 
 Temperature mean – – 21.0 21.5 21.5 
 Temperature range – – 19.7–22.3 19.8–23.1 20.5–22.6 
 Central Punjab (Ludhiana) 
Optimum sowing window 
 Temperature mean 22.3 (20–26 Oct)a 22.1 (24–30 Oct)a 22.0 (25–31 Oct)a 22.9 (25–31 Oct)a 22.3 (24–30 Oct)a 
 Temperature range 20.6–23.9 20.6–23.7 20.8–23.3 21.1–24.8 20.9–23.6 
Week preceding the optimum window 
 Temperature mean 24.7 23.3 22.3 23.7 24.0 
 Temperature range 23.7–25.7 22.0–24.6 20.8–23.8 22.2–25.2 22.4–25.5 
Week succeeding the optimum window 
 Temperature mean 20.5 20.7 20.8 21.9 21.5 
 Temperature range 18.9–22.1 18.7–22.7 19.5–22.1 20.1–23.7 20.6–22.4 
 South western Punjab (Bathinda) 
Optimum sowing window 
 Temperature mean – 23.0b (25–31 Oct)a 22.9 (24–30 Oct)a 23.8 (21–27 Oct)a 23.0 (21–27 Oct)a 
 Temperature range – 21.3–24.8b 21.4–24.4 21.4–26.1 21.3–24.6 
Week preceding the optimum window 
 Temperature mean – 24.1b 23.4 24.7 25.2 
 Temperature range – 22.4–25.7b 21.5–25.2 22.9–26.5 23.4–27.0 
Week succeeding the optimum window 
 Temperature mean – 21.6b 21.3 23.0 22.1 
 Temperature range – 19.3–23.8b 19.9–22.7 20.6–25.3 21.0–23.3 
Weekly Time period
 
1970–1975 1976–1985 1986–1995 1996–2005 2006–2015 
 North eastern Punjab (Ballowal Saunkhri) 
Optimum sowing window 
 Temperature mean – – 21.6 (21–27 Oct)a 22.5 (22–28 Oct)a 22.3 (22–28 Oct)a 
 Temperature range – – 20.2–22.9 21.1–24.0 21.1–23.6 
Week preceding the optimum window 
 Temperature mean – – 22.8 23.2 23.8 
 Temperature range – – 21.0–24.7 22.0–24.4 22.2–25.5 
Week succeeding the optimum window 
 Temperature mean – – 21.0 21.5 21.5 
 Temperature range – – 19.7–22.3 19.8–23.1 20.5–22.6 
 Central Punjab (Ludhiana) 
Optimum sowing window 
 Temperature mean 22.3 (20–26 Oct)a 22.1 (24–30 Oct)a 22.0 (25–31 Oct)a 22.9 (25–31 Oct)a 22.3 (24–30 Oct)a 
 Temperature range 20.6–23.9 20.6–23.7 20.8–23.3 21.1–24.8 20.9–23.6 
Week preceding the optimum window 
 Temperature mean 24.7 23.3 22.3 23.7 24.0 
 Temperature range 23.7–25.7 22.0–24.6 20.8–23.8 22.2–25.2 22.4–25.5 
Week succeeding the optimum window 
 Temperature mean 20.5 20.7 20.8 21.9 21.5 
 Temperature range 18.9–22.1 18.7–22.7 19.5–22.1 20.1–23.7 20.6–22.4 
 South western Punjab (Bathinda) 
Optimum sowing window 
 Temperature mean – 23.0b (25–31 Oct)a 22.9 (24–30 Oct)a 23.8 (21–27 Oct)a 23.0 (21–27 Oct)a 
 Temperature range – 21.3–24.8b 21.4–24.4 21.4–26.1 21.3–24.6 
Week preceding the optimum window 
 Temperature mean – 24.1b 23.4 24.7 25.2 
 Temperature range – 22.4–25.7b 21.5–25.2 22.9–26.5 23.4–27.0 
Week succeeding the optimum window 
 Temperature mean – 21.6b 21.3 23.0 22.1 
 Temperature range – 19.3–23.8b 19.9–22.7 20.6–25.3 21.0–23.3 

–, Meteorological data not available.

aOptimum sowing window identified by the model.

bTime period is 1977–1985.

In central Punjab, a shift of 4–5 days in the optimum sowing window was observed (Figure 3). During the earlier six years (1970–1975), the optimum sowing window was from 20 to 26 October and it was delayed to 24–30 October during 1976–1985 (Table 3). It got further delayed by 1 day (25–31 October) during 1986–1995 and remained the same during the next decade (1996–2005). During the recent decade (2006–2015), the optimum sowing window was preponed to 24–30 October. During 1970–1975, the mean temperature and its range in the optimum sowing window was 22.3 °C and 20.6–23.9 °C, respectively, for central Punjab. They remained almost the same during the rest of the decades (Table 3). In 2006–2015, during the week preceding the optimum sowing window the temperature range was higher by 1.5–1.9 °C and was lower by 0.3–1.2 °C during the succeeding week. These deviations in temperatures during the preceding and succeeding weeks from the temperature during the optimum sowing window might be the reason for lower simulated wheat productivity. An earlier study by Ortiz-Monasterio et al. (1994) concluded that at Ludhiana for long duration durum wheat variety PBW 34 (150 days) and medium duration wheat varieties PBW 154 (145 days) and PBW 226 (140 days) the optimum sowing time is 5 and 15 November, respectively. They conducted the field study during 1987–1992 which corresponds to the 1986–1995 time period of our study. The results of our study during the corresponding time period (1986–1995) showed 25–31 October as the optimum sowing window for long duration bread wheat variety DBW 17 (155 days duration). The difference of 5 days between the results of the two studies may be due to the difference in the genetic characteristics of the varieties. Ortiz-Monasterio et al. (1994) concluded that the optimum heading date for all three varieties of wheat (PBW 34, PBW 154, and PBW 226) at Ludhiana is the same. Thus, the long duration variety needs to be sown earlier than the short and medium duration varieties so that their respective heading phase corresponds to the optimum heading date. Hence, the long duration bread wheat variety DBW 17 (155 days) used in the present study needs to be sown 5 days earlier so that its heading phase corresponds to the optimum heading date.

In south western Punjab, the shift in optimum sowing window followed an entirely different trend. In north eastern and central Punjab, the optimum sowing window was delayed but in south western Punjab it was preponed (Figure 3). The optimum sowing window during 1977–1985 was 25–31 October which was preponed by 1 day (24–30 October) during 1986–1995 (Table 3). During the recent two decades (1996–2005 and 2006–2015), the optimum sowing window became stabilized at 21–27 October, i.e., preponed by 4 days. During 1977–1985, the mean temperature and its range in the optimum sowing window was 23.0 °C and 21.3–24.8 °C, respectively, for south western Punjab and remained almost the same during the rest of the decades (Table 3). In 2006–2015, during the week preceding the optimum sowing window, the temperature range was higher by 2.1–2.4 °C and was lower by 0.3–1.3 °C during the succeeding week. These deviations in temperatures during the preceding and succeeding weeks from the temperature during optimum sowing window might be the reason for lower simulated wheat productivity. Sharma (2000) reported that the favorable temperature regime for optimum growth and yield of wheat crop at sowing is 20–22 °C. The present study also found a similar temperature range (21–24 °C) at the time of sowing of wheat.

Generally, the average wheat productivity of south western Punjab (Bathinda district) is lower than that of central Punjab (Ludhiana district), but the results given in Figures 35 present a different scenario. The results of the present simulation study reveal a higher productivity at Bathinda than at Ludhiana. These locations have different soil, climate, and cropping patterns. The predominant cropping system in south western Punjab is cotton–wheat and that in central Punjab is rice–wheat. An earlier study by Lobell et al. (2013) reported that south western Punjab has sandier soils which are less favorable for rice production and so the cotton–wheat rotation is widespread in the area. As the cotton crop matures later than the rice crop, so the sowing of the wheat crop in south western Punjab gets delayed. This may be the reason for actual lower productivity of wheat in south western Punjab.

Current optimum sowing window

In north eastern Punjab, CERES-Wheat model simulated the optimum sowing window as 22–28 October for higher wheat productivity (Table 3). The productivity of wheat sown during the optimum window was 24.10% higher than the average productivity of currently recommended sowing time (25 October–30 November) and 1.12% higher than early sowing (25–31 October) recommended to the farmers for long duration wheat varieties (Table 4). The productivity of wheat sown during the optimum sowing window was 5.50% higher as compared to the average productivity of wheat sown during the succeeding week. During the last decade (2006–2015), in eight years, the highest wheat productivity was simulated with sowing done during the identified optimum sowing window and in the remaining two years it was on 21 October (preceding day of the identified sowing window). Hence, there is 80% probability of getting higher wheat productivity if the sowing of wheat is completed within the identified sowing window (22–28 October). During the recent decade, 25 October was the best sowing date within the optimum sowing window (22–28 October) which led to the highest wheat productivity (Figure 4). The trend analysis of wheat productivity showed a linear decrease with delay in sowing (Figure 5). In north eastern Punjab, the decrease in yield was 36.09 kg ha−1 day−1 with delay in sowing by 1 day from the optimum 7-day window.

Table 4

Simulated wheat yield average (kg/ha) and advantages in yield of optimum sowing window

Sowing period Average yield (kg/ha)
 
Yield increase during optimum sowing window (%)
 
1970–1975 1976–1985 1986–1995 1996–2005 2006–2015 1970–1975 1976–1985 1986–1995 1996–2005 2006–2015 
 North eastern Punjab (Ballowal Saunkhri) 
Optimum 7 day window – – 3,086 3,266 3,086 – – – – – 
Recommended window* – – 2,592 2,639 2,486 – – 19.9 23.7 24.1 
Early sown**   3,087 3,229 3,052 – – 0.68 1.14 1.12 
7 days following optimum sowing window – – 3,038 3,111 2,925 – – 2.30 4.98 5.50 
 Central Punjab (Ludhiana) 
Optimum 7 day window 4,958 5,276 5,006 4,667 4,482 – – – – – 
Recommended window* 3,755 4,296 4,257 3,850 3,653 32.01 22.80 17.61 21.20 22.7 
Early sown** 4,875 5,274 5,006 4,667 4,474 1.70 0.03 0.00 0.00 0.16 
7 days following optimum sowing window 4,710 5,030 4,771 4,356 4,139 5.26 4.88 4.94 7.15 8.27 
 South western Punjab (Bathinda) 
Optimum 7 day window – 5,582a 5,150 5,189 5,244 – – – – – 
Recommended window* – 4,648a 4,384 4,159 4,024 – 20.1a 17.5 24.8 30.3 
Early sown** – 5,582a 5,146 5,158 5,157 – 0.00a 0.09 0.59 1.69 
7 days following optimum sowing window – 5,315a 5,003 5,042 4,889 – 5.01a 2.95 2.90 5.82 
Sowing period Average yield (kg/ha)
 
Yield increase during optimum sowing window (%)
 
1970–1975 1976–1985 1986–1995 1996–2005 2006–2015 1970–1975 1976–1985 1986–1995 1996–2005 2006–2015 
 North eastern Punjab (Ballowal Saunkhri) 
Optimum 7 day window – – 3,086 3,266 3,086 – – – – – 
Recommended window* – – 2,592 2,639 2,486 – – 19.9 23.7 24.1 
Early sown**   3,087 3,229 3,052 – – 0.68 1.14 1.12 
7 days following optimum sowing window – – 3,038 3,111 2,925 – – 2.30 4.98 5.50 
 Central Punjab (Ludhiana) 
Optimum 7 day window 4,958 5,276 5,006 4,667 4,482 – – – – – 
Recommended window* 3,755 4,296 4,257 3,850 3,653 32.01 22.80 17.61 21.20 22.7 
Early sown** 4,875 5,274 5,006 4,667 4,474 1.70 0.03 0.00 0.00 0.16 
7 days following optimum sowing window 4,710 5,030 4,771 4,356 4,139 5.26 4.88 4.94 7.15 8.27 
 South western Punjab (Bathinda) 
Optimum 7 day window – 5,582a 5,150 5,189 5,244 – – – – – 
Recommended window* – 4,648a 4,384 4,159 4,024 – 20.1a 17.5 24.8 30.3 
Early sown** – 5,582a 5,146 5,158 5,157 – 0.00a 0.09 0.59 1.69 
7 days following optimum sowing window – 5,315a 5,003 5,042 4,889 – 5.01a 2.95 2.90 5.82 

–, Meteorological data not available.

*25 October to 30 November; ** last week of October.

aTime period is 1977–1985.

In central Punjab, the best sowing window for higher wheat productivity is from 24 to 30 October (Table 3). The productivity of wheat sown during the optimum window was 22.7% higher than the average productivity of currently recommended sowing time (25 October–30 November) for long duration wheat varieties (Table 4). The productivity of wheat sown during the optimum sowing window was 8.27% higher as compared to the average productivity of wheat sown during the succeeding week. During the last decade (2006–2015), in five years, the highest wheat productivity was simulated with sowing done during the identified optimum sowing window (24–30 October) and in the remaining four years it was on 31 October (succeeding day of the identified sowing window). Hence, there is 50% probability of getting higher wheat productivity if the sowing of wheat is completed within the identified sowing window (24–30 October). During the recent decade, 27 October was the best sowing date within the optimum sowing window (24–30 October) which led to highest wheat productivity (Figure 4). Similar to north eastern Punjab, in central Punjab the trend analysis of wheat productivity showed a linear decrease with delay in sowing (Figure 5). The decrease in wheat yield was 52.51 kg ha−1 day−1 with delay in sowing by 1 day from the optimum 7-day window.

In south western Punjab, the best sowing window for higher wheat productivity was from 21 to 27 October (Table 3). The productivity of wheat sown during the optimum window was 30.3% higher than the average productivity of currently recommended sowing time (25 October–30 November) and 1.69% higher than early sowing (25–31 October) recommended to the farmers for long duration wheat varieties (Table 4). The productivity of wheat sown during the optimum sowing window was 5.82% higher as compared to the average productivity of wheat sown during the succeeding week. During the last decade (2006–2015), in nine years the highest wheat productivity was simulated with sowing done during the identified optimum sowing window (21–27 October). Hence, there is 90% probability of getting higher wheat productivity if the sowing of wheat is completed within the identified sowing window (24–30 October). During the recent decade, 24 October was the best sowing date within the optimum sowing window (21–27 October) which led to highest wheat productivity (Figure 4). Similar to the other two regions, in south western Punjab the trend analysis of wheat productivity also showed a linear decrease with delay in sowing (Figure 5). The decrease in wheat yield was 70.8 kg ha−1 day−1 with delay in sowing by 1 day from the optimum 7-day window.

Earlier studies have suggested a shift in the date of sowing of wheat as an adaptation strategy to counter the adverse impacts of anticipated rise in temperature due to climate change (Attri & Rathore 2003; Kalra et al. 2008; Wang et al. 2012; Sandhu et al. 2016). However, none of these studies have analyzed the effects of rise in temperature which has already occurred and might have resulted in a shift in the optimum sowing window of wheat. Most of the previous studies considered two or three dates as possible options, but the present study simulated daily sowing of crop to identify the effects of climate change on the sowing window of wheat. Similar to our approach, Zheng et al. (2012) also performed simulations with daily sowing of wheat with the objective to identify the optimal combinations of sowing date and variety for their study region. Similar to our results, Liang et al. (2015) reported that in Loess Plateau of China during 1981–2009, the observed dates of sowing of winter wheat have been delayed by an average of 1.2 days decade−1. Earlier field studies by Randhawa et al. (1981) and Ortiz-Monasterio et al. (1994) reported that delay in sowing of wheat by 1 day leads to a 0.7–1.2% loss in yield, mainly due to the exposure of the crop to high temperature near the end of the growing season. Jalota et al. (2013) suggested a delay in current planting date of rice and wheat by 15–21 days as a probable mitigation option to counter the impact of a rise in temperature by 5.1 °C by the end of the 21st century in central Punjab. As discussed earlier, if wheat sowing is completed within the identified optimum sowing window, there are 80, 50, and 90% chances of getting higher productivity in north eastern, central, and south western Punjab, respectively. Hence, it may be inferred that by taking an average of these three figures, there is a 73% probability of achieving higher productivity of wheat, if the sowing is completed within the identified sowing windows of the respective regions.

CONCLUSION

The study highlighted climate change induced shift in the optimum sowing window of wheat in Punjab. The current identified optimum sowing windows are 22–28 October in north eastern Punjab, 24–30 October in central Punjab, and 21–27 October in south western Punjab. The aim of identifying the best sowing date and sowing window for different regions was to enhance the wheat productivity by encouraging farmers to sow the maximum area within these sowing windows. Proper time of sowing is a non-monetary input which can help in improving the productivity without any additional financial liability. The methodology followed in this study will be helpful in determining optimum sowing time for other crops and in different regions.

REFERENCES

REFERENCES
Asseng
S.
,
Turner
N. C.
&
Keating
B. A.
2001
Analysis of water- and nitrogen-use efficiency of wheat in a Mediterranean climate
.
Plant Soil
233
,
127
143
.
Attri
S. D.
&
Rathore
L. S.
2003
Simulation of impact of projected climate change on wheat in India
.
Int. J. Climatol.
23
,
693
705
.
Bhatti
D. S.
2015
Package of Practices for Crops of Punjab Rabi 2015–16
.
Punjab Agricultural University
,
Ludhiana, India
, pp.
1
18
.
Bhatti
D. S.
2018
Package of Practices for Crops of Punjab Rabi 2018–19
.
Punjab Agricultural University
,
Ludhiana
,
India
, pp.
1
18
.
Buttar
G. S.
,
Jalota
S. K.
,
Sood
A.
&
Vashisht
B. B.
2012
Yield and water productivity of Bt cotton (Gossypium hirsutum) as influenced by temperature under semi-arid conditions of north-western India: field and simulation study
.
Indian J. Agric. Sci.
82
(
1
),
70
75
.
FAO
2017
Cited From FAOSTAT Crop Data
.
Available from: http://www.fao.org/faostat/en/#data/QC (accessed 28 September 2017)
.
Flohr
B. M.
,
Hunt
J. R.
,
Kirkegaard
J. A.
&
Evans
J. R.
2017
Water and temperature stress define the optimal flowering period for wheat in south-eastern Australia
.
Field Crops Res.
209
,
108
119
.
Holzworth
D. P.
,
Huth
N. I.
,
deVoil
P. G.
,
Zurcher
E. J.
,
Herrmann
N. I.
,
McLean
G.
,
Chenu
K.
,
van Oosterom
E. J.
,
Snow
V.
,
Murphy
C.
,
Moore
A. D.
,
Brown
H.
,
Whish
J. P. M.
,
Verrall
S.
,
Fainges
J.
,
Bell
L. W.
,
Peake
A. S.
,
Poulton
P. L.
,
Hochman
Z.
,
Thorburn
P. J.
,
Gaydon
D. S.
,
Dalgliesh
N. P.
,
Rodriguez
D.
,
Cox
H.
,
Chapman
S.
,
Doherty
A.
,
Teixeira
E.
,
Sharp
J.
,
Cichota
R.
,
Vogeler
I.
,
Li
F. Y.
,
Wang
E.
,
Hammer
G. L.
,
Robertson
M. J.
,
Dimes
J. P.
,
Whitbread
A. M.
,
Hunt
J.
,
van Rees
H.
,
McClelland
T.
,
Carberry
P. S.
,
Hargreaves
J. N. G.
,
MacLeod
N.
,
McDonald
C.
,
Harsdorf
J.
,
Wedgwood
S.
&
Keating
B. A.
2014
APSIM − evolution towards a new generation of agricultural systems simulation
.
Environ. Modell. Softw.
62
,
327
350
.
Hundal
S. S.
&
Kaur
P.
2007
Climatic variability and its impact on cereal productivity in Indian Punjab
.
Curr. Sci.
92
(
4
),
506
512
.
Jamieson
P. D.
,
Porter
J. R.
&
Wilson
D. R.
1991
A test of computer simulation model ARC-WHEAT1 on wheat crops grown in New Zealand
.
Field Crops Res.
27
,
337
350
.
Jones
J. W.
,
Tsuji
G. Y.
,
Hoogenboom
G.
,
Hunt
L. A.
,
Thornton
P. K.
,
Wilkens
P. W.
,
Imamura
D. T.
,
Bowen
W. T.
,
Singh
U.
1998
Decision support system for agrotechnology transfer: DSSAT v3
. In:
Understanding Options for Agricultural Production
(
Tsuji
G. Y.
,
Hoogenboom
G.
&
Thornton
P. K.
, eds).
Kluwer Academic Publishers
,
Dordrecht
,
the Netherlands
, pp.
157
177
.
Kalra
N.
,
Chakraborty
D.
,
Sharma
A.
,
Rai
H. K.
,
Jolly
M.
,
Chander
S.
,
Kumar
P. R.
,
Bhadraray
S.
,
Barman
D.
,
Mittal
R. B.
,
Lal
M.
&
Sehgal
M.
2008
Effect of increasing temperature on yield of some winter crops in northwest India
.
Curr. Sci.
94
,
1
10
.
Kaur
P.
,
Sandhu
S. S.
,
Singh
H.
,
Kaur
N.
,
Singh
S.
&
Singh
A.
2016
Climatic Features and Their Variability in Punjab. Bulletin: AICRPAM, School of Climate Change and Agricultural Meteorology, PAU, Ludhiana
. p.
88
.
doi: 10.13140/RG.2.2.29571.91688
.
Keating
B. A.
,
Carberry
P. S.
,
Hammer
G. L.
,
Probert
M. E.
,
Robertson
M. J.
,
Holzworth
D.
,
Huth
N. I.
,
Hargreaves
J. N. G.
,
Meinke
H.
,
Hochman
Z.
,
McLean
G.
,
Verburg
K.
,
Snow
V.
,
Dimes
J. P.
,
Silburn
M.
,
Wang
E.
,
Brown
S.
,
Bristow
K. L.
,
Asseng
S.
,
Chapman
S.
,
McCown
R. L.
,
Freebairn
D. M.
&
Smith
C. J.
2003
An overview of APSIM, a model designed for farming systems simulation
.
Eur. J. Agron.
18
,
267
288
.
Liang
H.
,
Asseng
S.
,
Zhao
G.
,
Wu
D.
,
Yang
X.
,
Zhuang
W.
,
Jin
N.
&
Yu
Q.
2015
Impacts of recent climate warming, cultivar changes, and crop management on winter wheat phenology across the Loess Plateau of China
.
Agric. Forest. Meteorol.
200
,
135
143
.
Lobell
D. B.
,
Ortiz-Monasterio
I. J.
,
Sibley
A. M.
&
Sohu
V. S.
2013
Satellite detection of earlier wheat sowing in India and implications for yield trends
.
Agric. Systems
115
,
137
143
.
Moshatati
A.
,
Siadat
S. A.
,
Kh
A. S.
,
Bakhshandeh
A. M.
&
Kamali
M. R. J.
2012
Effect of terminal heat stress on yield and yield components of Spring bread wheat cultivars in Ahwaz
.
Iran. Int. J. Agric. Res. Rev.
2
,
844
849
.
Pathak
H.
,
Li
C.
,
Wassmann
R.
&
Ladha
J. K.
2006
Simulation of nitrogen balance in rice–wheat systems of the Indo-Gangetic Plains
.
Soil Sci. Soc. Amm. J.
70
,
1612
1622
.
Piramila
B. H. M.
,
Prabha
A. L.
,
Nandagopalan
V.
&
Stanley
A. L.
2012
Effect of heat treatment on germination, seedling growth and some biochemical parameters of dry seeds of black gram
.
Int. J. Pharm. Phyto. Res.
1
,
194
202
.
Randhawa
A.
,
Dhillon
S.
&
Singh
D.
1981
Productivity of wheat varieties as influenced by the time of sowing
.
J. Res.-Punjab Agricultural University
18
(
1
),
52
59
.
Rane
J.
,
Pannu
R. K.
,
Sohu
V. S.
,
Saini
R. S.
,
Mishra
B.
,
Shoran
J.
,
Crossa
J.
,
Vargas
M.
&
Joshi
K.
2007
Performance of yield and stability of advanced wheat cultivar under heat stress environments of the indo-gangetic plains
.
Crop Sci.
47
,
1561
1572
.
Ritchie
J. T.
,
Singh
U.
,
Godwin
D. C.
,
Bowen
W. T.
1998
Cereal growth development and yield
. In:
Understanding Options for Agricultural Production
(
Tsuji
Y. G.
,
Hoogenboom
G.
&
Thornton
P. K.
, eds).
Kluwer Academic Publishers
,
Dordrecht
,
the Netherlands
, pp.
79
98
.
Rosenzweig
C.
&
Iglesias
A.
1994
Implications of Climate Change for International Agriculture: Crop Modeling Study
.
US Environmental Protection Agency
,
Washington, DC
,
USA
.
Sandhu
S. S.
,
Kaur
P.
,
Tripathi
P.
,
Patel
S. R.
,
Prasad
S. R.
,
Solanki
N. S.
,
Kumar
R.
,
Singh
C. B.
,
Dubey
A. P.
&
Rao
V. U. M.
2016
Effect of rise in intra-seasonal temperature at major wheat growing location in India: a study using CERES-Wheat model
.
J. Agromet.
18
,
222
233
.
Sharma
S. N.
2000
Wheat
. In:
Techniques and Management of Field Crop Production
(
Rathore
P. S.
ed.).
Agrobios
,
India
, pp.
99
123
.
ISBN: 9788177541755.
Toh
S.
,
Imamura
A.
,
Watanabe
A.
,
Nakabayashi
K.
,
Okamot
M.
,
Jikumaru
Y.
,
Hanada
A.
,
Aso
Y.
,
Ishiyama
K.
,
Tamura
N.
,
Iuchi
S.
,
Kobayashi
M.
,
Yamaguchi
S.
,
Kamiya
Y.
,
Nambara
E.
&
Kawakami
N.
2008
High temperature-induced abscisic acid biosynthesis and its role in the inhibition of gibberellin action in arabidopsis seeds
.
Pl. Physiol.
146
,
1368
1385
.
Tsuji
G. Y.
,
Uehara
G.
&
Balas
S.
1994
DSSAT Version 3. International Benchmark Sites Network for Agrotechnology Transfer
.
University of Hawaii
,
Honolulu
.
Vashisht
B. B.
,
Mulla
D. J.
,
Jalota
S. K.
,
Kaur
S.
,
Kaur
H.
&
Singh
S.
2013
Productivity of rainfed wheat as affected by climate change scenario in northeastern Punjab, India
.
Reg. Environ. Change
13
(
5
),
989
998
.
Wheeler
T. R.
,
Batts
G. R.
,
Ellis
R. H.
,
Hadley
P.
&
Morison
J. I. L.
1996
Growth and yield of winter wheat (Triticum aestivum L.) crops in response to CO2 and temperature
.
J. Agric. Sci. Camb.
127
,
37
48
.