1 Introduction

Rice is the most important staple food for about 150 million population of the world. It is a water loving plant requiring a heavy rainfall of 125 cm during its growing period and annual rainfall of 100–200 cm [1]. There should be a monthly rainfall of 200 mm to grow lowland rice and 100 mm to grow upland rice successfully [2]. As rice is mainly grown under flooded conditions, any change in climate that lead to reduction in water availability for agriculture might impact the crop and water productivity of rice to a greater extent. As per the current population growth rate of 2 million per year, the demand for rice will grow to an estimated 2000 million metric tons by 2030 [3, 4]. To supply to this increasing demand, the methods of rice productivity will require significant improvement [5]. Achieving this goal, however, is sure to be a challenge with respect to future climatic changes [6], which will basically be characterized by current global warming trends [7].

Climate change and agriculture are interrelated processes, both of which take place on a global scale. Climate is one of the major determinants of agricultural productivity. There is significant concern about the effects of climate change and its variability on agricultural productivity throughout the world. The effect of climate on agriculture is related to variability in local climates rather than global climate patterns. An increase has been predicted in average temperature of 0.5–1.0 °C by mid century and 3.5–4.5 °C by end century with increased frequency of extremely wet rainy seasons in the Indo-Gangetic plains [8]. The rise in temperature and uncertain rainfall associated with climate change may have serious adverse effects either directly or indirectly on the growth, development and yield of rice crop [9]. A decrease of 10–40% in rice yield has been predicted in the Indo–Gangetic plains with increase in temperature [10]. A 10% decline in rice productivity has been estimated with 1 °C rise in minimum temperature in the dry season [11]. The increased temperature leads to forced maturity and poor harvest index due to limited water supply [12]. The water stress during grain filling period may result in decline of grain yield. However, these assumptions are based in a business as usual scenario, with no new technological development and with either no or limited adaptation by all stakeholders. Despite the technological advances such as improved rice varieties and irrigation systems, weather and climate are important factors which play a significant role in agricultural productivity.

The all India mean of total June-September rainfall during 1961–1998 was about 5% below the mean for the previous 30-year period [13]. This reduction is more than double the overall reduction since the late 1800s [14], thus suggesting that the weakening of the monsoon has accelerated. During 1951–2003, the area of India with monsoon rainfall below the mean expanded by nearly 50% [15]. The Indian Punjab has also received below normal monsoon rainfall in most of the years since 2000. Because of climate variability being experienced in the region during last many years, large year to year variability in rice productivity is observed. A significant increase in minimum temperature (at 0.06 °C year−1) and decrease in sunshine hours (at 0.02 h day−1) in the parts of Indian Punjab has been reported [16]. Rice requires higher sunshine hours and significant decrease in sunshine hours can lead to severe implications on rice productivity in the region. Thus, increasing minimum temperature and decreasing sunshine hours can have significant negative impact on its productivity. Though it is accepted that variation in climate and agricultural management practices influence crop yield [17, 18] but there are very few studies in which the combined effect of climate, fertilizer and irrigation on rice yield have been statistically analysed and little is known about the spatially variable determinants of rice yield and climate variables in Punjab. The specific objectives of this study are (1) to assess the temporal variations in rice yield, rainfall, maximum and minimum temperature in Punjab over a period of 40 years, (2) to evaluate the impact of climate, fertilizers and irrigation on rice yield and, (3) to study the spatial variations of rice yield and growing season climatic variability.

2 Study area, materials and methods

The study was conducted for Punjab covering the north-west region of Indo-Gangetic plains of India. It extends from 29°33′N to 32°31′N and from 73°55′E to 76°55′E covering an area of 50,362 km2, which is 1.54% of country’s total geographical area (Fig. 1). The climate of Punjab is dominantly semi-arid and monsoon type, which is highly influenced by the Himalayas in the north and north-east and the Thar Desert in the south and south-west. The major part of the rainfall is received from monsoon between July and September, which is essential for growing summer (Kharif) crops and the subsequent sowing of winter (rabi) crops. The rainfall is also received from western disturbances during winter season. As there is large climatic variability from north-east to south-west region of the state so the whole state is divided mainly into three regions : North-East, Central and South-west.

Fig. 1
figure 1

Location of the study area

Full size image

Rice yield and three climatic parameters (maximum temperature, minimum temperature and rainfall) during rice growing period (June to September) from 1974 to 2013 were collected from different sources including Statistical Abstracts of Punjab, India Meteorological Department and Punjab Agricultural University, Ludhiana. The data of fertilizer consumption and irrigated area was also collected from Statistical Abstracts of Punjab.

The following statistical techniques were used to find the trend and interpolation of rice yield and climate parameters:

  • Descriptive statistics (minimum, maximum and standard error) was used for studying temporal variation in rice yield, rainfall, minimum and maximum temperature.

  • Non-parametric methods: Two non-parametric methods (Mann-Kendall test and Sen’s slope estimator) were used to detect the meteorological variables’ trends, for details see Meals et al. [19] and Kingra et al. [20].

  • Stepwise regression: It was used to identify the climate factors (rainfall, maximum- and minimum temperature) affecting rice yield and the relationship of residuals with fertilizer and irrigation. The regression analysis was performed by SPSS Statistical Software 17.0

  • Interpolation of rice yield and climate parameters: Spatial distribution of rice yield and climate parameters was carried out using Inverse Distance Weighted (IDW) method in Arc GIS 10.2.

3 Results

3.1 Temporal and spatial variability in rice yield

Trend analysis showed a linear increase in rice yield in all the three regions of Punjab. The rate of increase was 30 kg ha−1 year−1 in North-east, 29 kg ha−1 year−1 in Central and 34 kg ha−1 year−1 in the South-west region (Fig. 2). Spatial interpolation of rice yield showed that it varied from 2800 to 3400 kg ha−1 in South-west region and from 2200 to 2800 kg ha−1 in Central and North-eastern regions during 1974–1983. Compared with the first decade (1974–1983), the rice yield during second decade (1984–1993) was increased by 400–600 kg ha−1 in south-west, and by 400–800 kg ha−1 in the Central and North-eastern districts. The area under 3200–3800 kg ha−1 increased in the three regions in the third decade (1994–2003). However, rice yield was 4000–4600 kg ha−1 in South-west, 3400–4400 kg ha−1 in Central region and 3400–4200 kg ha−1 in North-east region from 2004 to 2013 (Fig. 3).

Fig. 2
figure 2

Variability and trend of rice yield over a period of 40 years in different regions of Punjab (India)

Full size image
Fig. 3
figure 3

Spatio-temporal variability in rice yield during a 1974–1975 to 1983–1984, b 1984–1985 to 1993–1994, c 1994–1995 to 2003–2004 and d 2004–2005 to 2103–2014 in Punjab (India)

Full size image

3.2 Temporal and spatial climate variability during rice growing period

3.2.1 Maximum temperature

There were large fluctuations in maximum temperature during different decades and no significant pattern of variability in maximum temperature was observed (Fig. 4). Average maximum temperature was 34.7 °C during 1974–1993, 33.9 °C during 1994–2003 and 34.3 °C during 2004–2013 (Table 1). The results of Mann-Kendall test and Sen’s slope estimator have also depicted large inter- and intra-seasonal fluctuations in maximum temperature during rice season (Table 2). Spatial interpolation of maximum temperature showed that it remained between 35 and 37 °C in South-west region, between 33 and 37 °C in Central region, and between 33 and 35 °C in north-east region with few pockets of 31 and 33 °C. The spatial distribution did not indicate significant variations in maximum temperature during the first three decades (1974–2003), but the area under 31–35 °C has been increased and 35–37 °C has been decreased increased from 2004 to 2013 (Fig. 5).

Fig. 4
figure 4

Variability and trend of maximum temperature during rice growing season over a period of 40 years in different regions of Punjab (India)

Full size image
Table 1 Decade-wise variability in average yield of rice and climatic parameters during its growing period over a period of 40 years in different regions of Punjab
Full size table
Table 2 Results of the statistical tests for variability in maximum temperature during rice growing season over a period of 40 years in different regions of Punjab (India)
Full size table
Fig. 5
figure 5

Spatio-temporal variability in maximum temperature during rice growing season from a 1974 to 1983, b 1984 to 1993, c 1994 to 2003 and d 2004 to 2013 in Punjab (India)

Full size image

3.2.2 Minimum temperature

Minimum temperature increased from 22.5 to 24.0 °C in the North-east region, from 24.0 to 25.1 °C in the Central region and from 24.7 to 25.8 °C in the South-west region during 1974–2013 (Table 1). Mann-Kendall test and Sen’s slope estimator showed significant increase in minimum temperature during different months (Table 3). During rice growing period, the rate of increase in minimum temperature was highest in September in the North-east region, and July in the Central and South-west regions. The rate of increase in minimum temperature during the rice growing season was 0.03 °C year−1 in North-east region and 0.04 °C year−1 in Central and South-west regions of Punjab. In addition to significant increase over time, large yearly fluctuations have also been observed in minimum temperature in all the regions (Fig. 6).

Table 3 Results of the statistical tests for variability in minimum temperature during rice growing season over a period of 40 years in different regions of Punjab (India)
Full size table
Fig. 6
figure 6

Variability of minimum temperature during rice growing season over a period of 40 years in different regions of Punjab (India)

Full size image

Spatial interpolation showed higher minimum temperature in the South-west and lower in the North-east region of the state. During the first decade (1974–1983), minimum temperature was 24–26 °C in South-western region along with parts of Central regions, whereas it was 22–24 °C in North-east region along with the remaining parts of Central region. During the succeeding two decades (1984–2003), area under 24–26 °C increased but the area under 22–24 °C decreased. However, during 2004–2013, the mean minimum temperature was 24–26 °C in the entire state except small area in the North-east region (Fig. 7).

Fig. 7
figure 7

Spatio-temporal variability in minimum temperature during rice growing season from a 1974 to 1983, b 1984 to 1993, c 1994 to 2003 and d 2004 to 2013 in Punjab (India)

Full size image

3.2.3 Rainfall

There was large temporal variability in rainfall during rice growing season in the state, but no significant pattern of rainfall over the years was observed (Fig. 8 and Table 4). Spatial interpolation of rainfall also indicated highly erratic nature of rainfall in the region. On an average, increase in rainfall was observed from 1974 to 2003 in all the regions, but it decreased during 2004–2013 (Fig. 9).

Fig. 8
figure 8

Variability and trend of rainfall during rice growing season over a period of 40 years in different regions of Punjab (India)

Full size image
Table 4 Results of the statistical tests for variability in rainfall during rice growing season over a period of 40 years in different regions of Punjab (India)
Full size table
Fig. 9
figure 9

Spatio-temporal variability in rainfall during rice growing season from a 1974 to 1983, b 1984 to 1993, c 1994 to 2003 and d 2004 to 2013 in Punjab (India)

Full size image

3.2.4 Relative importance of climate parameters, fertilizer and irrigation

Among three climatic variables, 49% of the variance in rice yield was explained by minimum temperature. The relationship of residual of rice yield with irrigation and fertilizers showed that out of the remaining 51% variance not explained by climate variables (rainfall, minimum and maximum temperature), 33% variation in rice yield was explained by irrigation availability. It shows that minimum temperature is one of the main determinants affecting rice yield in Punjab.

4 Discussion

The results showed that minimum temperature explained the major portion of rice yield. It has been established that temperature is one of the limiting factors for rice productivity [21]. Temperature influences not only duration but also growth pattern of rice. The flowering stage (anthesis and fertilization) is most susceptible stage of development to temperature stress [22]. In addition to this, spikelet fertility (seed-set) is an important component of yield that is sensitive to high temperature [23]. Increase in minimum temperature results in shortening of crop duration and hence reduction in productivity. A possible 10% decline in rice productivity has been estimated with 1 °C rise in minimum temperature in dry season [11]. Based on the historical data analysis, a decline in rice grain yield by 10% has also been reported for each 1 °C increase in growing season minimum temperature [24]. Significant increase in minimum temperature can have detrimental effect on rice productivity in the region and it indicates a threat of warming scenarios to rice yields. Large variability in maximum temperature and rainfall can also have adverse effect on rice productivity. Water deficit during vegetative, flowering and grain filling stages has been reported to reduce mean yield by 21, 50 and 21%, respectively [25]. Water stress before or during tillering reduce the number of tillers and panicles per hill which causes reduction in grain weight [26]. The fluctuation of total rainfall, total number of rainy days and period of monsoon season can have adverse effect on rice productivity and the water resources. Due to erratic rainfall in Punjab, ground water is over-exploited in most of the regions [27] and warming scenarios can further aggravate the problem by increasing crop water requirements in future, which indicates an alarming situation regarding food security and water availability in the years to come.

Over the period of four decades, fertilizer consumption in the state has increased by 431% (41–218 kg ha−1), total cropped area by 32% (from 5950 to 7870 thousand ha) and irrigated area by 26.9% from 1974–1975 to 2013–2014. These parameters positively affected rice yield in Punjab. The combined effects of climate and technology on rice yield suggest the adoption of climate smart and resource conservation technologies in agriculture in the region.

5 Conclusions

In this study, the effect of climatic variables was larger than that of other technology variables and it is an important consideration for future sustainability of rice production in Punjab. Climate parameters and yield varied spatially in the region and it requires region specific policies for sustaining food security. As most parts of the Indian Punjab are already suffering from over-exploitation of ground water resources, large fluctuations in maximum temperature and rainfall can further aggravate the problem by increasing water requirements under warming scenarios and decreasing its availability from highly erratic rainfall pattern. Resource conserving and climate smart practices need to be followed for managing the adverse impact of climate variability in the region. Our results also indicate that weather forewarning at a finer scale for agricultural practices may be important in predicting and managing the effect of climate change on rice yield and water requirements.