International Journal of Biometeorology

, Volume 55, Issue 3, pp 387–401

Effects of precipitation and temperature on crop production variability in northeast Iran

Authors

    • Faculty of AgricultureFerdowsi University of Mashhad
  • Sajad Sadeghi Lotfabadi
    • Faculty of AgricultureFerdowsi University of Mashhad
  • Sarah Sanjani
    • Faculty of AgricultureFerdowsi University of Mashhad
  • Azadeh Mohamadian
    • Climate Research CenterKhorasan
  • Majid Aghaalikhani
    • Tarbiat Modares University
Original Paper

DOI: 10.1007/s00484-010-0348-7

Cite this article as:
Bannayan, M., Sadeghi Lotfabadi, S., Sanjani, S. et al. Int J Biometeorol (2011) 55: 387. doi:10.1007/s00484-010-0348-7
  • 407 Views

Abstract

Climate variability adversely impacts crop production and imposes a major constraint on farming planning, mostly under rainfed conditions, across the world. Considering the recent advances in climate science, many studies are trying to provide a reliable basis for climate, and subsequently agricultural production, forecasts. The El Niño-Southern Oscillation phenomenon (ENSO) is one of the principle sources of interannual climatic variability. In Iran, primarily in the northeast, rainfed cereal yield shows a high annual variability. This study investigated the role played by precipitation, temperature and three climate indices [Arctic Oscillation (AO), North Atlantic Oscillation (NAO) and NINO 3.4] in historically observed rainfed crop yields (1983–2005) of both barley and wheat in the northeast of Iran. The results revealed differences in the association between crop yield and climatic factors at different locations. The south of the study area is a very hot location, and the maximum temperature proved to be the limiting and determining factor for crop yields; temperature variability resulted in crop yield variability. For the north of the study area, NINO 3.4 exhibited a clear association trend with crop yields. In central locations, NAO provided a solid basis for the relationship between crop yields and climate factors.

Keywords

Climate changeClimate variabilityRainfed crop productionWeather and crops

Introduction

Among all human activities, agriculture is the most climate dependent. Concerns about climate change and variability and their effects on agricultural production have stimulated research into the analysis of climate and agricultural crop productivity (IPCC 1996). It is recognized that climate change and variability impose a wide range of direct and indirect impacts on crop production; thus, weather fluctuations play a significant role in crop growth and yield. Any change in local weather conditions, especially during critical developmental stages of crops, may adversely impact growth processes and result in enormous yield reductions. This situation makes climate variability a threat to food production with potentially serious economic implications at local, regional, national and global scales (Alexandrov and Hoogenboom 2000).

Climate variability is more prominent in arid and semi-arid conditions. Iran is located within an arid and semi-arid belt. Rainfall variability is critical for agricultural yields. If the pattern of precipitation from the time of planting onward is unknown, farmers are unable to tune their cropping systems to optimize resources. In both very wet and very dry years, farmers face major economic disasters, which usually results in migration from rural areas to cities. Throughout Khorasan province, in the northeast of Iran, the agricultural economy relies strongly on rainfed crops, in either commercial or subsistence farming. Wheat (Triticum aestivum L.) and barley (Horedum vulgare L.) are the two major crops in this region that provide seasonal income and employment for a large segment of the regional rural population. A large body of historical and on going research is devoted to the study of barley and wheat for optimizing both management and environmental suitability to achieve higher yield as the demand for these crops increases annually. High yield variation among cropping seasons is a fact faced by farmers for both wheat and barley. Although it is speculated that much of this variability is related to rainfall fluctuation (Vanderlip et al. 1996), to date there has been no comprehensive study of agrometeorological yield association. Improved climate forecasting (Bannayan and Hoogenboom 2008a, b) and the fact that dissemination of forecast climate information is becoming routine in many countries (Phillips et al. 1999) provide the opportunity to employ climate information for agricultural planning and decision making (Hansen and Sivakumar 2006).

Global climate change would cause stronger interannual variability with more extreme year-to-year climate variations (Monirul and Mirza 2003). Risks to agricultural production due to climate variations have been reported through various studies (Phillips et al. 1999; Reilly 2002). The El Nino/Southern Oscillation (ENSO) is a recurring pattern of climate variability characterized by anomalies in sea surface temperature (SST). ENSO forecasts are now recognized as an important tool for assessing food security (Mjelde et al. 1998; Broad and Agrawala 2000). Numerous studies have shown that climate indices, both on global and regional scales, play a major role in crop yield variability at seasonal to interannual scales in many parts of the world (Trenberth 1996, 1997). Diaz (1995) explored associations between SST indices in the tropical Pacific ocean (as indicators of ENSO conditions), rainfall anomalies, and yields of maize and soybean in key crop districts in the Pampas. He found extremely low correlation between rainfall and maize yield anomalies in four major districts. Diaz (1995) also found rainfall anomalies to be more closely related with national-level maize yield than with yield anomalies. In contrast, Messina et al. (1996a) showed significant correlation between Pacific SST anomalies in December and maize yields in the central part of the Pampas. Messina et al. (1996b) detected a positive correlation between ENSO-related SST anomalies and wheat yields, but only for the southwestern portion of the Pampas.

To determine regional climates across 28 main cities of Iran, Soltani et al. (2007) found three main climatic groups based on time series models: simple, moderate and complex climates. Their results showed a high variation in the temporal pattern of monthly rainfall over Iran except for the margins of the Caspian Sea and the Persian Gulf. Both wheat and barley yield patterns in the northeast of Iran follow the seasonality of precipitation (Fig. 1). The timing of precipitation occurrence also dictates the failure or success of the growing season crop yield. However, such relationships have not been quantified to explore any possible association that would be useful in terms of food security. Nazemosadat and Cordery (2000) investigated the relationships between Iranian autumn rainfall and ENSO and reported that the associations between the Southern Oscillation Index (SOI) and rainfall, and found that, during El Niño episodes, the amount of rainfall over various parts of the country was several times more than during La Niña periods. Direct statistical correlation between study crops yield and SST patterns may provide a basis for crop yield forecasts. Such studies have been reported for various crops at various locations including maize in Africa and Mexico (Jury et al. 1997), and wheat, sorghum and sugarcane in Australia (Meinke et al. 1996)—countries where rainfall shows significant correlations with ocean-atmosphere anomalies related to the ENSO cycle.
https://static-content.springer.com/image/art%3A10.1007%2Fs00484-010-0348-7/MediaObjects/484_2010_348_Fig1_HTML.gif
Fig. 1

Geographical study locations

Our objectives in this study were to evaluate the relationship between precipitation and temperature variability and the yields of two major crops (using historical data), and also their possible association with three different climate indices in northeast Iran.

Materials and methods

Data

Crop yield records of wheat and barley (1985–2005) were obtained from annual national yield reports of the Ministry of Agriculture, which contain data for each location across Khorasan province (Fig 1). On assessment of the soil water supply based on climatic conditions in Khorasan province of Iran, Rozhkov et al. (2007) delineated different agroclimatic zones. The ultra-arid zone is located mainly in the southwestern part of the province. The central (Mashhad) and southern (Birjand) parts belong largely to the arid zone. The northern part (Boujnord) is in the semi-arid zone with steppe vegetation and forest groves. On average, the amount of precipitation during the growing season in the southwestern part of the province reaches 64 mm. In the southern (Birjand) and central (Mashhad) parts of the province, the average annual precipitation reaches 167 and 240 mm, respectively. In northern regions (Bojnourd), it may be as high as 265 mm. The majority of the precipitation falls in the fall, winter, and spring seasons, with almost no summer rainfall. The annual amplitude of the air temperature (calculated as the difference between the mean monthly temperatures of the warmest and coldest months) in the ultra-arid region is equal to 29–33°C. In the arid, semi-arid, and moderately dry steppe regions, it reaches 31–34°C, 25–29°C, and 19–23°C, respectively. In the whole region, drought conditions depend not only on the amount of precipitation but also on the air temperature and the coefficient of water utilization by herbs. In northeast Iran, wheat and barley require 2,300 and 1,900 degree-days from planting to maturity, respectively. Although only 11% of agricultural production is under a rainfed system, and total yields are quite low, many farmers are tied to this level of yield. Adverse impacts on this production would affect the economics of such farms enormously. For our study objectives, non-climatic influences such as improvements in crop genetics and technical factors were removed by detrending the time series in yield productions. Yield data was detrended by means of double exponential smoothing (Joseph and LaViola 2003), and yield anomalies were obtained as the difference between the yield in each year and the average of the long-term observed yield. Long-term records of wheat and barley yield from 1983 to 2006 showed a coefficient of variation of 94% for barley and 78% for wheat. The minimum observed yield was 0 kg ha−1 for both barley and wheat, with maximum yields of 1,900 kg ha−1 for wheat and 2,185 kg ha−1 for barley. Such annual variation has often exposed poor farmers to food security problems and obligatory migration to cities.

Climate indices

Variations in many climate variables are strongly correlated with large scale features of atmospheric circulation, as well as through interactions involving the land and ocean surfaces (Nicholls et al. 1996). Recent advances in forecasting air–climate interaction indices with a lead time of several months (Berliner et al. 2000) may benefit the agricultural sector by enabling farmers to mitigate or adapt to adverse climate conditions if a robust association between SSTs and land production could be established. To explore the impact of climate indices, the association between climate indices, precipitation, temperature, and the direct association between climate indices on rainfed wheat and barley yield were investigated. The relationship between yield, and study factors was investigated using linear correlation and scatter-plots. Both concurrent and lag correlations (up to 2 years) were computed. For the required analysis, the SST in region 3.4 (NINO 3.4), Arctic Oscillation (AO) and North Atlantic Oscillation (NAO) were used as climate indices. The NINO 3.4 is defined as a 3-month average of SST departures from normal for a critical region of the equatorial Pacific (Nino 3.4 region; 120 W-170 W, 5 N-5 S). The AO is an index of the dominant pattern of non-seasonal sea-level pressure variations north of 20 N latitude, and is characterized by pressure anomalies of one sign in the Arctic with the opposite anomalies centered about 37–45 N. The NAO is a climatic phenomenon in the North Atlantic Ocean of fluctuations in the difference of atmospheric pressure at sea-level between the Icelandic low and the Azores high. Through east-west oscillation motions of the Icelandic low and the Azores high, it controls the strength and direction of westerly winds and storm tracks across the North Atlantic. The three climate indices (monthly NAO, AO and NINO 3.4) were obtained from the NOAA website (http://www.cpc.ncep.noaa.gov/data/indices), which covers data from 1985 to 2005.

Results and discussion

Precipitation

In this study we evaluated the SST indices, historical crop yields, minimum and maximum temperatures and precipitation for the study region to look for any association between them. Rainfed crop yields are tied mainly to available soil water and precipitation. Our analysis showed annual changes in October–June—the growing season of both crops—with wheat (Table 1) and barley (Table 2) yields being highly correlated with annual changes in precipitation in both the north (Bojnourd) and south (Birjand) of Khorasan across all the study years. This correlation was statistically significant (P < 0.05) for wheat in two locations and for barley in one location (Tables 1, 2). In the central part (Mashhad), about 30% of annual variation in the yield of both crops was explained by precipitation. A time series of annual precipitation and both barley and wheat yield across three locations (Fig. 2) indicated that highest yields of both crops were obtained in peak precipitation years, and low yields coincided with well below normal precipitation. Such behavior suggests that once the minimum water requirement of the crops is satisfied, yields would increase unless there are other factors limiting productivity. Furthermore, as barley is more resistant to water shortage, barley is more productive than wheat in years with low precipitation, while in wet years wheat was more successful (Fig. 2). Table 1 shows that total precipitation during the growing season has a higher impact on wheat yields for Bojnourd and Birjnad, and that precipitation in April shows a significant correlation with wheat yield. However, in Mashhad the growing season precipitation did not show (P > 0.05) any clear impact on wheat yield, although a significant correlation between wheat yield and monthly precipitation in May was observed (Table 1), and a significant correlation was found between wheat yield and precipitation anomalies (Fig. 2). Talliee and Bahramy (2003) studied the impact of precipitation on wheat yield in the west part of Iran. They reported that higher precipitation in spring showed more positive impact than increasing precipitation in the early months of crop growth and development.
Table 1

Wheat grain yield association with minimum temperature (Tmin, °C), maximum temperature (Tmax, °C), average temperature (Tmean, °C) and precipitation (mm) on a monthly basis and for the whole growing season across study locations

Wheat

Parameters (unit)

 

Mashhad

Bojnourd

Birjand

 

 

r

P-value

r

P-value

r

P-value

October (1982-2004)

Tmin (°C)

0.3993

0.0591**

0.0394

0.8585

-0.3441

0.1079

Tmax (°C)

0.1907

0.3834

-0.1097

0.6184

-0.4684

0.0242*

Tmean (°C)

0.3112

0.1483

-0.0429

0.8460

-0.4370

0.0370*

Precipitation (mm)

-0.2368

0.2767

0.2442

0.2615

0.0724

0.7428

November (1982-2004)

Tmin (°C)

0.4716

0.0231*

0.3560

0.0955**

0.0150

0.9460

Tmax (°C)

0.2086

0.3395

0.0646

0.7695

0.1596

0.4670

Tmean (°C)

0.3155

0.1425

0.1791

0.4134

0.1320

0.5482

Precipitation (mm)

0.1382

0.5296

0.1335

0.5437

-0.0108

0.9608

December (1982-2004)

Tmin (°C)

0.4295

0.0408*

-0.0617

0.7797

0.0605

0.7839

Tmax (°C)

0.0826

0.7079

-0.1654

0.4506

-0.4712

0.0232*

Tmean (°C)

0.2494

0.2510

-0.1254

0.5685

-0.3421

0.1102

Precipitation (mm)

0.1786

0.4148

-0.0562

0.7989

0.3500

0.1016

January (1983-2005)

Tmin (°C)

0.4097

0.0522

0.2620

0.2271

-0.2107

0.3346

Tmax (°C)

-0.0224

0.9193

0.1499

0.4949

-0.2418

0.2664

Tmean (°C)

0.1888

0.3883

0.2045

0.2492

-0.2972

0.1685

Precipitation (mm)

0.0968

0.6604

-0.0376

0.8649

-0.0877

0.6907

February (1983-2005)

Tmin (°C)

0.4575

0.0281*

0.1582

0.4709

0.1800

0.4111

Tmax (°C)

0.0758

0.7312

-0.0090

0.9675

0.0804

0.7152

Tmean (°C)

0.2470

0.2559

0.0666

0.7628

0.1477

0.5011

Precipitation (mm)

0.1404

0.5228

0.2779

0.2164

0.1865

0.3942

March (1983-2005)

Tmin (°C)

0.2489

0.2522

0.1380

0.5300

-0.3127

0.1443

Tmax (°C)

0.2123

0.3307

0.2466

0.2566

-0.5030

0.0144*

Tmean (°C)

0.2412

0.2676

0.2160

0.3222

-0.4627

0.0262*

Precipitation (mm)

-0.1084

0.6224

0.2097

0.3369

0.2120

0.3314

April (1983-2005)

Tmin (°C)

0.2739

0.2060

0.0260

0.9063

-0.3280

0.1266

Tmax (°C)

-0.2451

0.2596

-0.2654

0.2209

-0.5471

0.0069*

Tmean (°C)

-0.0415

0.8508

-0.1857

0.3962

-0.4918

0.0171*

Precipitation (mm)

0.0779

0.7238

0.5209

0.0108*

0.5064

0.0137*

May (1983-2005)

Tmin (°C)

0.0310

0.8883

-0.3696

0.0826**

-0.2228

0.2257

Tmax (°C)

-0.3970

0.0607**

-0.3837

0.0707**

-0.3614

0.0902**

Tmean (°C)

-0.2573

0.2359

-0.3895

0.0662**

-0.3278

0.1268

Precipitation (mm)

0.4838

0.0193*

0.3460

0.1058

-0.1743

0.4264

June (1983-2005)

Tmin (°C)

0.2841

0.1888

-0.3927

0.0638**

-0.2063

0.3450

Tmax (°C)

-0.3003

0.1638

-0.3759

0.0771**

-0.1979

0.3653

Tmean (°C)

-0.0023

0.9918

-0.4063

0.0544**

-0.2420

0.2660

Precipitation (mm)

0.3745

0.0783**

0.8182

0.5911

0.3682

0.0839**

Growing season (October–June)

Tmin (°C)

0.5614

0.0053*

0.1006

0.6477

-0.4532

0.0299*

Tmax (°C)

0.0209

0.9246

-0.2075

0.3421

-0.5373

0.0082*

Tmean (°C)

0.3266

0.8282

-0.0915

0.6779

-0.4802

0.0204*

Precipitation (mm)

0.2972

0.1685

0.5404

0.0078*

0.4426

0.0345*

*Significant at 5% level, ** significant at 10% level

Table 2

Barley grain yield association with minimum temperature (Tmin, °C), maximum temperature (Tmax, °C), average temperature (Tmean, °C) and precipitation (mm) on monthly basis and the whole growing season across study locations

Barley

Parameters (unit)

Mashhad

Bojnourd

Birjand

r

P-value

r

P-value

r

P-value

October (1982-2004)

Tmin (°C)

0.5043

0.0141*

0.4379

0.0366*

0.0487

0.8252

Tmax (°C)

0.2007

0.3585

0.1549

0.4803

-0.1727

0.4307

Tmean (°C)

0.3724

0.0801**

0.3013

0.1623

-0.0562

0.7991

Precipitation (mm)

-0.2080

0.3409

0.0541

0.8065

0.0930

0.6730

November (1982-2004)

Tmin (°C)

0.3540

0.0974**

0.1538

0.4836

0.0562

0.7989

Tmax (°C)

0.1009

0.6469

-0.1888

0.3882

0.2816

0.1931

Tmean (°C)

0.1938

0.3757

-0.0767

0.7279

0.2517

0.2465

Precipitation (mm)

0.1676

0.4447

0.0241

0.9132

-0.2856

0.1865

December (1982-2004)

Tmin (°C)

0.4820

0.0199*

0.0705

0.7491

0.1909

0.3830

Tmax (°C)

0.2183

0.3169

-0.1160

0.5981

-0.0696

0.7522

Tmean (°C)

0.3531

0.0984**

-0.0313

0.8874

0.0401

0.8559

Precipitation (mm)

-0.0465

0.8332

0.0247

0.8109

0.2826

0.1913

January (1983-2005)

Tmin (°C)

0.3404

0.1119

0.2312

0.2886

0.1873

0.3922

Tmax (°C)

0.0158

0.9429

0.0056

0.9796

-0.2538

0.2421

Tmean (°C)

0.1763

0.4210

0.1011

0.6463

-0.0629

0.7756

Precipitation (mm)

0.1087

0.6215

-0.0665

0.7630

0.1087

0.6215

February (1983-2005)

Tmin (°C)

0.5010

0.0149*

0.2471

0.2556

0.2685

0.2154

Tmax (°C)

0.1755

0.4231

0.0950

0.6665

-0.0031

0.9886

Tmean (°C)

0.3272

0.1275

0.1624

0.4591

0.1418

0.5187

Precipitation (mm)

0.0783

0.7227

0.2422

0.2654

0.3726

0.0799**

March (1983-2005)

Tmin (°C)

0.1846

0.3990

0.2703

0.2123

-0.0332

0.8805

Tmax (°C)

0.2292

0.2929

0.2540

0.2422

-0.0302

0.8914

Tmean (°C)

0.2545

0.3030

0.2783

0.1985

-0.0356

0.8718

Precipitation (mm)

-0.2993

0.1653

0.2741

0.1890

-0.1133

0.6069

Apr (1983-2005)

Tmin (°C)

0.3322

0.1215

0.0763

0.7293

-0.1676

0.4447

Tmax (°C)

-0.0985

0.6548

-0.2202

0.3126

-0.1766

0.4201

Tmean (°C)

0.0776

0.7250

-0.1315

0.5498

-0.1883

0.3896

Precipitation (mm)

-0.0026

0.9912

0.4518

0.0305*

0.0861

0.6968

May (1983-2005)

Tmin (°C)

0.1807

0.4092

-0.1647

0.4527

-0.0992

0.6525

Tmax (°C)

-0.1810

0.4085

-0.3723

0.0802**

-0.1827

0.4041

Tmean (°C)

-0.0508

0.8181

-0.3072

0.1540

-0.1533

0.4849

Precipitation (mm)

0.2260

0.2999

0.4959

0.0161*

0.0234

0.9156

Jun (1983-2005)

Tmin (°C)

0.3266

0.1285

-0.1553

0.4793

0.0200

0.9278

Tmax (°C)

-0.1629

0.4578

-0.3196

0.1372

-0.0567

0.7973

Tmean (°C)

0.1028

0.6406

-0.2764

0.2016

-0.0198

0.9285

Precipitation (mm)

0.2835

0.1898

-0.0105

0.9621

0.0780

.7235

Growing season (October-June)

Tmin (°C)

0.6046

0.0022*

0.4016

0.0575**

0.0521

.8133

Tmax (°C)

0.1858

0.3960*

-0.1789

0.4141

-0.1326

.5464

Tmean (°C)

0.4412

0.0351*

0.0559

0.8001

-0.0237

.9144

Precipitation (mm)

0.0649

0.7687

0.5130

0.0123*

0.1876

.3913

*Significant at 5% level, ** significant at 10% level

https://static-content.springer.com/image/art%3A10.1007%2Fs00484-010-0348-7/MediaObjects/484_2010_348_Fig2_HTML.gif
Fig. 2

Time trend of annual wheat and barley yield along with total annual precipitation in three different sites in the northeast of Iran

Analyzing on a monthly basis during the growing season, the highest correlation of association between wheat yield and precipitation for the central region (Mashhad) was in May (Table 1) when it coincided with the grain setting stage, but there was no significant correlation with barley, either in terms of growing season or at any monthly scale (Table 2). For the north of the province, Bojnourd, which is cooler and drier than other two sites, April for wheat, and April and May for barley, showed significant correlation with precipitation. For the south of the province (Birjand), which is hotter and drier than the other two sites, there was a correlation between yield and precipitation in both April (at 5% level) and June (at 10% level) for wheat but in barley only February showed a significant correlation (at 10% level). It seems that as the climate of the study location becomes drier, fall and winter precipitation has more impact, while in less dry locations, spring and early summer precipitation is more influential.

The effect of precipitation on the yields of both crops indicated that the timing of precipitation occurrence, i.e., precipitation distribution within the growing season, resulted in annual variation. In drier areas precipitation towards the end of the winter months is important, while for less dry and cooler areas precipitation in spring and early summer has more impact. Studying various locations in Iran, Karimi (1999) also reported that precipitation during tillering of both wheat and barley, which occurs in early spring, significantly impacts final grain yield.

The association between precipitation anomalies and climate indices is shown in Fig. 3. There was no relationship between precipitation anomalies and climate indices. The figure shows the relationship between the three study indices and the anomaly of annual precipitation at concurrent lag, lag 1 and lag 2. As changing time lags did not change the direction and extent of the relationship, it was concluded that, even if any such association does exist, the concurrent lag is also able to provide the required prediction basis. Thus, there was no impact of lag on the analysis of precipitation and climate indices across the region. This result also showed that there was no delayed impact of ENSO on yield in this part of the world. Yarahmadi and Nassiri (2006) reported that precipitation amount and climate indices showed mostly a poor correlation across Iran during spring but they found a strong correlation between the SOI and monthly precipitation across Iran, with highest correlation in October; however, different locations showed different effects. They also reported a strong positive relation between NINO 3.4 and both fall and winter precipitation.
https://static-content.springer.com/image/art%3A10.1007%2Fs00484-010-0348-7/MediaObjects/484_2010_348_Fig3r1_HTML.gifhttps://static-content.springer.com/image/art%3A10.1007%2Fs00484-010-0348-7/MediaObjects/484_2010_348_Fig3r2_HTML.gifhttps://static-content.springer.com/image/art%3A10.1007%2Fs00484-010-0348-7/MediaObjects/484_2010_348_Fig3r3_HTML.gif
Fig. 3

Relationships between climate indices and precipitation anomalies at concurrent lag and lags 1 and 2

The existence of an impact of climate indices on crop production may provide an incentive to the agricultural sector to take advantage of climate forecasts. This could give the advantage of providing time to consider more flexible management (Bannayan et al. 2003). Considering the correlation between both crop yields and precipitation variation, and the correlation between climate indices and cereals yield, we carried out a tentative analysis of the direct association between wheat and barley yield and major components of climate variability, such as climate indices. Figure 4 shows the association between each of climate indices and crop yield anomalies of both barley and wheat at each study location.
https://static-content.springer.com/image/art%3A10.1007%2Fs00484-010-0348-7/MediaObjects/484_2010_348_Fig4_HTML.gif
Fig. 4

Relationships between climate indices and yield anomalies of wheat and barley

Bojnourd exhibited a significant correlation only between wheat yield anomaly and Nino3.4. Such a relationship in Birjand for both crops was obtained between yield anomalies and AO. For Mashhad, none of the climate indices showed a significant relationship with any crops yield anomalies. For the central part of the study area, the significant correlation was between wheat and NAO, but only in October, with no significant relationship for barley with any of the climate indices in any month. For all three climate indices, the impact on wheat was higher than on barley. In the north of the study area, the highest correlation was obtained in the winter months for wheat, while for barley it was in early fall months. For Bojnourd, while there was no significant relationship between climate indices and barley, wheat yield showed a significant relationship with NINO 3.4 in July (at 10% level), in August (at 10% level) and in December (at 10% level) (Table 3). In Mashhad, NAO showed a clearer relationship with crop yield anomaly (Fig. 4). The same was observed between NINO 3.4 and crop yield anomaly in Bojnourd. In both locations, at positive values of NAO and NINO 3.4, crop yield increased, and, conversely, at negative values of the indices, yields decreased. In the southern part of the area, Birjand, wheat yield showed a significant correlation with AO in August (at 10% level), while for barley the significant relationship was with AO in November (at 10% level) and in December (at 5% level) (Table 3). In October, the relationship between wheat yield and NAO in Mashhad was negative, while there was a positive relationship between NINO 3.4 and wheat yield in Bojnourd (Table 3). The relationship between precipitation and wheat yield in October is also negative in Mashhad and positive in Bojnourd (Table 1).
Table 3

Wheat and barley grain yield association with climate indices for 3 months before and after cultivation across study locations

Parameters (unit)

 

Mashhad

Bojnourd

Birjand

  

r

P-value

r

P-value

r

P-value

Wheat

July (1982–2004)

AO

0.0781

0.7233

-0.1402

0.5234

0.0414

0.8513

NAO

-0.0880

0.6897

0.0110

0.9603

0.2577

0.2352

NINO 3.4

0.2218

0.3090

0.4117

0.0510

0.2387

0.2728

August (1982–2004)

AO

0.0606

0.7837

0.1572

0.4736

0.3724

0.0802**

NAO

0.0112

0.9597

0.0404

0.8548

0.2978

0.1675

NINO 3.4

0.1474

0.5023

0.4351

0.0380*

0.1562

0.4765

September (1982–2004)

AO

-0.1793

0.4130

-0.0982

0.6558

-0.2105

0.3350

NAO

-0.0866

0.6945

-0.2445

0.2609

0.0871

0.6926

NINO 3.4

0.0917

0.6775

0.3142

0.1443

0.1078

0.6245

October (1983-2005)

AO

-0.3195

0.1372

-0.2218

0.3090

-0.0232

0.9165

NAO

-0.4994

0.0153*

-0.3391

0.1135

-0.0868

0.6938

NINO 3.4

0.1857

0.3963

0.3670

0.0850

0.1464

0.5051

November (1983–2005)

AO

0.3491

0.1026

0.3365

0.1164

0.2630

0.2254

NAO

0.1817

0.4068

0.1550

0.4800

0.1134

0.6064

NINO 3.4

0.2375

0.2752

0.3917

0.0645

0.2287

0.2939

December (1983–2005)

AO

0.1231

0.5758

-0.0287

0.8964

0.1501

0.4944

NAO

-0.0534

0.8088

0.0587

0.7904

0.1446

0.5104

NINO 3.4

0.2624

0.2264

0.3811

0.0728

0.2829

0.1909

Barley

July (1982–2004)

AO

0.0461

0.8347

-0.2302

0.2906

-0.0540

0.8066

NAO

-0.1993

0.3620

0.0215

0.9223

0.2047

0.3488

NINO 3.4

0.0941

0.6693

0.2458

0.2583

0.1082

0.6232

August (1982–2004)

AO

0.0371

0.8664

-0.1620

0.4602

0.1985

0.3639

NAO

-0.0603

0.7847

-0.1975

0.3664

0.0363

0.8695

NINO 3.4

0.0302

0.8911

0.3078

0.1530

0.1001

0.6496

September (1982–2004)

AO

-0.1686

0.4419

0.1349

0.5394

0.1225

0.5776

NAO

-0.1235

0.5745

-0.0823

0.7090

0.0565

0.7980

NINO 3.4

-0.0115

0.9585

0.2596

0.2316

0.0360

0.8706

October (1983–2005)

AO

-0.2771

0.2005

-0.2931

0.1747

-0.0512

0.8166

NAO

-0.3396

0.1128

-0.2155

0.3233

-0.1168

0.5955

NINO 3.4

0.1017

0.6443

0.2597

0.2315

0.0931

0.6725

November (1983–2005)

AO

0.1972

0.3670

-0.2844

0.1885

0.3554

0.0961**

NAO

0.0407

0.8538

-0.1050

0.6336

0.0579

0.7930

NINO 3.4

0.1474

0.5020

0.2527

0.2448

0.1441

0.5120

December (1983–2005)

AO

-0.0987

0.6541

-0.3292

0.1251

0.4830

0.0196*

NAO

-0.1671

0.4459

-0.2832

0.1903

0.3303

0.1238

NINO 3.4

0.1562

0.4767

0.2216

0.3095

0.1890

0.3878

*Significant at 5% level, ** significant at 10% level

The above relational associations are able to provide the required basis for projection of final yield situation of wheat crop before harvest. NAO and precipitation in Mashhad also showed a negative association. It seems that when NAO is negative one may expect a higher wheat yield in Mashhad; however, when NAO is negative there is a possibility that precipitation will also be high, but there was a negative relationship between grain yield and precipitation in October, which is when most wheat is planted in this area. Therefore, other factors may also play a role. Based on historical weather data, minimum temperature is more limiting and controlling than other factors. Over the Iberian Peninsula there is strong evidence that positive (negative) values of winter NAO induce low (high) vegetation activity in the following spring and summer seasons (Vicente-Serrano and López-Moreno 2008; Gouveia et al. 2008). This feature is associated mainly with the impact of NAO on winter precipitation. Gouveia and Trigo (2008) mentioned that such a situation reflects the different responses of vegetation to atmospheric variability, in particular changes induced by temperature and precipitation in the annual cycle of heat and moisture. For instance, in the case of wheat, which is grown in both regions, water is the main limiting factor for growth in Iberian Peninsula (Gouveia and Trigo 2008) whereas it is temperature that limits its growth in Northern Europe. Moradi (2004) reported a positive correlation between NAO and precipitation in many parts of Iran. Rodo et al. (1997) revealed that NAO influenced seasonal rainfall in the Iberian Peninsula, but at different temporal and spatial scales. Figure 3 shows that when AO, NAO and NINO 3.4 were negative, there was no clear trend in yield anomalies in either crop, with yield anomalies showing slight positive values but only at positive values of AO and NAO. These results illustrate the idea that linking SST indices directly to wheat and barley cultivation may provide a basis for monitoring annual variation of these two crops at positive values of these indices. However, such calculated associations were found to be more reliable and robust in the central part of the study area.

Temperature

Temperature is central to how climate influences the growth and yield of crops (Wheeler et al. 2000). The growth and development rates of many crops are controlled by temperature (Bannayan et al. 2004). Both wheat and barley across all three locations responded to temperature (Figs. 5, 6) though the response in Bojnourd was lower than at the other two locations. In Mashhad, as the minimum temperature anomaly turned to positive values, wheat and barley yield decreased, except around the year 2000 when an immense drought hit the whole area (Fig. 5). In contrast, in Bojnourd, at negative anomalies of minimum temperature, the yields of both crops increased. Such a situation was more prominent for wheat than for barley (Fig. 5). For Birjand, a linear trend of annual change in minimum temperature from negative values to positive values is obvious in Fig. 5. However, there is no specific relationship between wheat and barley yield with minimum temperature anomalies (Fig. 5). In response to maximum temperature anomalies, all three locations behaved almost similarly (Fig. 6), as at positive anomaly of maximum temperature, rainfed yield of both barley and wheat decreased while negative anomalies had the opposite effect (Fig. 6). Analysing the results during the growing season on a monthly basis, in Mashhad, wheat yield did not show any relationship with average temperature but barley showed a significant correlation in October, November, December and for the whole growing season (Tmean, Table 1). Wheat yield exhibited a significant relationship with minimum temperature in October, November, December and for the whole growing season (Table 1). Barley exhibited similar results except that there was no relation for November (Table 2). For the north of the province, Bojnourd, wheat yield showed a significant correlation with minimum temperature in November, and with maximum, minimum and mean temperature in May and June (Table 1). Barley showed a significant correlation with minimum temperature in October and for the whole growing season, and with maximum and mean temperature in May (Table 2). For the south of the province, Birjand, maximum temperature showed a significant correlation with wheat yield in October, December, March, April, May and the whole growing season, and October, March and the whole growing season for Tmean. Barley did not show any relation with any of the temperature indices at all (Tables 1, 2). Lobell and Asner (2003) investigated relationships between climate trends and yields of corn and soybean in the United States and found that growing season temperature was the only variable correlated with yield. Moradi (2004) found that there was a negative correlation between NAO and temperature in most parts of Iran.
https://static-content.springer.com/image/art%3A10.1007%2Fs00484-010-0348-7/MediaObjects/484_2010_348_Fig5_HTML.gif
Fig. 5

Time trend of annual wheat and barley yield along with minimum temperature annual anomaly in three different sites in the northeast of Iran

https://static-content.springer.com/image/art%3A10.1007%2Fs00484-010-0348-7/MediaObjects/484_2010_348_Fig6_HTML.gif
Fig. 6

Time trend of annual wheat and barley yield along with maximum temperature annual anomaly in three different sites in the northeast of Iran

Conclusion

In general our results showed that, for Birjand, maximum temperature is the main limiting factor and any prediction basis to follow seasonal yield variability should be based on or consider the maximum temperature. For Bojnourd, yields of both crops showed a linear positive relationship with NINO 3.4 (Fig. 3). At positive values of NINO 3.4, yields increased, and at negative values of NINO 3.4, yields decreased. However, there was no clear association between any of climate indices and temperature anomalies. For Mashhad, NAO showed a clear association with crop yields. At positive values of NAO, yields increased, and at negative values of NAO, yields exhibited a declining trend. However, none of the indices in this location showed any relationship with either precipitation or temperature. The existence or absence of a solid relationship between crop yields and/or climate variable indices, which also varies at different locations, indicates the complexity of the influence of ENSO on climate and crop yields.

Acknowledgment

This study has been supported by the grant approval of the Ferdowsi University of Mashhad, Iran and the authors would like to appreciate it.

Copyright information

© ISB 2010