Heat stress in cereals: Mechanisms and modelling
Introduction
Increased climate variability and higher mean temperatures are expected across many world regions (Weisheimer and Palmer, 2005, Tebaldi et al., 2006, Battisti and Naylor, 2009, Field et al., 2012) and are likely to cause large negative impacts on crop productivity (Porter and Semenov, 2005). Empirical evidence increasingly shows that short episodes of high temperature can have large negative impacts on crop yields (Reidsma et al., 2009, Schlenker and Roberts, 2009, Lobell et al., 2013). At a global scale, wheat yields have been negatively impacted by rising temperatures, as detected by Lobell and Field (2007) between 1961 and 2002. The negative trend of decreasing wheat yields with more frequent high temperature extremes during sensitive reproductive stages is apparent across many regions, as found by Gourdji et al. (2013) for recent decades (1980–2011) across Central and South Asia and South America. Wheat yields in Mexico show a significant negative response to higher night-time temperatures (Lobell et al., 2005). Likewise for maize, an analysis of the past 50-years of historical yields in France revealed that since approximately 2000, daily maximum temperatures explain as much yield variability as precipitation (Hawkins et al., 2013), with the cumulative number of days with a maximum temperature over 32 °C associated with yield reductions. Lobell et al. (2011) determined maize kernel set was reduced by 1% per degree day (and 1.7% per degree day under drought stressed conditions) when daily temperatures were above a threshold 30 °C in Sub-Saharan Africa. A national panel analysis of county level maize yields in the United States detected negative impacts on maize yields when daily temperatures were above 29 °C (Schlenker and Roberts, 2009). Evidence in rice suggests that this crop is also sensitive to increasing nighttime temperatures, expected to increase with climate change (Tebaldi et al., 2006). In an analysis of historical station data across China for the period 1981–2000, rice yields declined with higher nighttime temperatures, decreasing at a rate of 4.6% per 1 °C increase in minimum temperature (Tao et al., 2006). The decline in an indica rice varietals’ yield over a 25-year period in the Philippines was associated with an increase in minimum nighttime temperature but not correlated with the concurrent but smaller increase in daily maximum temperature (Peng et al., 2004). As the majority of cereal production, particularly rice and maize, now occurs at mean temperatures above the optimal (Hatfield et al., 2011) increases in global mean temperature would augment yield reductions (Lobell and Gourdji, 2012).
The term heat stress is increasingly used to describe these negative impacts of high temperature on plant growth, though a definitive definition has yet to emerge in the literature and remains elusive. Heat stress has been used to refer to brief episodes of high temperature lying outside of the range typically experienced (Porter and Gawith, 1999, Luo, 2011, Moriondo et al., 2011). Porter and Semenov (2005) and Wheeler et al. (2000) emphasize that negative yield impacts are greatest when high temperatures are experienced during the reproductive phases centered on flowering. Some authors define a high temperature event as heat stress if it results in large, irreversible yield reductions (Wahid et al., 2007). Attribution of yield losses is frequently explained by a reduction in the number of viable seeds produced (Wheeler et al., 2000, Moriondo et al., 2011) or accelerated leaf senescence that reduces yields by shortening the duration of grain filling (Al-Khatib and Paulsen, 1984, Asseng et al., 2011, Lobell et al., 2012). Finally, other authors have defined heat stress as the departure from the regular linear yield response to rising temperatures that occurs when a threshold is surpassed, apparent in the analysis of large panel datasets (Schlenker and Roberts, 2009, Lobell et al., 2011).
The lack of convergence in definitions may simply reflect the need to illustrate specific aspects or levels of detail in different cases. However, it likely also reflects the limitations of our understanding of the mechanisms of high temperature impacts on yield in field crops. Such impacts are the end result of the integration of many processes that operate at the organelle and lower levels all with differing sensitivities to temperature (Sage and Kubien, 2007) and their interactions with other temperature sensitive processes such as transpiration, assimilation and partitioning (Ferrise et al., 2011). These processes are generally studied in isolation (Wahid et al., 2007, Barnabás et al., 2008) and are difficult to abstract to conditions typical in the field. Secondly, the relatively few field scale experimental trials on heat stress have imposed high temperature at different periods, for differing durations and levels, under varying environmental conditions and using different varieties (Lobell et al., 2012), sometimes leading to what seem to be conflicting conclusions. Further, while at the field and larger scales, heat stress is frequently understood to represent a non-linear temperature response, many of the underlying individual mechanisms may not be deviating from their linear response (e.g. the acceleration of crop development with elevated temperatures that results in shorter duration of grain filling). For the remainder of this paper, we use the term very broadly to mean yield reductions resulting from high temperature whose mechanism and impacts are hypothesized to vary with crop, region and the scale considered.
This complexity suggests an important role for crop models to systematize the effects of many processes under a range of environments. However, despite the evidence of the role of high temperatures in reducing grain number (Porter and Gawith, 1999, Wheeler et al., 2000), a key determinant of final yield in cereals (Cirilo and Andrade, 1994, Otegui, 1995, Ferris et al., 1998, Fischer et al., 1998, Hayashi et al., 2012), crop model simulation efforts to date have focused largely on how high temperature accelerates leaf senescence in wheat (Asseng et al., 2011, Lobell et al., 2012) or changes atmospheric water demand and soil water supply in maize (Lobell et al., 2013) and not the direct impacts on grain number (Carberry et al., 1989, Moriondo et al., 2011). While these studies demonstrate that the impacts of high temperature on water use and accelerated senescence dominate as explanations for yield loss in some regions (Asseng et al., 2011), it is not clear if such modelling approaches are appropriate across regions and scales, and perhaps do not adequately reflect the state of the art in understanding crop response to high temperatures (Ferrise et al., 2011, Rötter et al., 2011, White et al., 2011, Eitzinger et al., 2012).
The aim of this review is to compile the state of the art on plant, canopy and regional scale cereal yield formation in response to high temperature stress to serve as a basis for crop models’ improvements. We focus on wheat, maize and rice, as globally, these represent the three most important cereal crops. In Section 2, the influence of temperature, across optimal and higher values, on key physiological processes affecting crop growth and development is reviewed. Section 3 presents the impacts of high temperatures during the flowering and grain filling phases (in the following referred to as “heat stress”) observed for the main yield determinants across crops. Efforts are made to link this knowledge to an underlying physiological process response. Finally, broad approaches to modelling heat stress are reviewed and related to the main mechanisms of heat stress. We conclude with a statement of the research needs to enable better simulation of heat stress impacts in real production settings.
Section snippets
Crop growth and development processes’ response to temperature
Temperature plays a role in nearly all aspects of crop growth and development (Ferrise et al., 2011), such as photosynthesis (Sage et al., 2011), respiration (Atkin and Tjoelker, 2003), transpiration (Crawford et al., 2012), dry matter partitioning (Zhao et al., 2013), plant development (Wolkovich et al., 2012) and root growth (Kaspar and Bland, 1992). The optimal conditions for growth processes of plants usually occur within a range of temperatures (Criddle et al., 1997), with higher or lower
Observed crop specific impacts of heat stress on crop growth
Section 2 described the response of many individual processes to temperature, both in and above their optimal range. In this section, we attempt to describe the impact of heat stress, occurring particularly around flowering and grain filling, on crop yield components. It is likely that the observed impacts represent the integrated response of crops to the various processes described above—some of which will be operating in their optimal range, and other above it. As such, this section attempts
Methods to study heat stress
Understanding crop response to high temperature stress is key to improving crop models with regards to their ability to simulate crop growth at the field and larger scale in warmer and more extreme climates. It is important for crop modellers to understand the conditions under which crop response to temperature has been studied, such that they can better gauge the applicability of such knowledge as the basis for model improvement. For example, the majority of studies on the impacts of heat
Approaches to modelling heat stress
The preceding review of crop response to high temperatures has emphasized that across cereal crops, the most significant impacts of high temperatures on yield formation appear to be associated with reductions in grain number when heat stress occurs at flowering or reductions in grain weight when high temperatures are experienced during grain filling. A summary of the relative importance of various processes’ responses to high temperature on final yield and yield formation are shown in Fig. 4
Canopy temperature
Most crop models consider ambient air temperature to drive various processes and rates, including heat stress effects. For example in their respective heat stress routines, APSIM (Keating et al., 2003) uses daily maximum temperatures, while modified CropSyst (Moriondo et al., 2011) considers mean temperatures between 08:00 h and 14:00 h and SIMPLACE's heat stress module (Eyshi Rezaei et al., 2013) uses a daily average weighted four times more heavily to the daily maximum than the minimum
For understanding heat stress
Knowledge gaps to overcome regarding the impact of heat stress on cereals near flowering, and its interaction with drought, at the plant and greater scales include: (1) the relationship between infrared canopy temperatures, aerodynamic canopy temperatures and existing knowledge on temperature thresholds determined using air temperature, typically generated in chamber studies; (2) the nature of interactions of CO2 levels, heat and water stresses (Mittler, 2006); (3) the importance of
Conclusions
This overview has highlighted the negative impacts of high temperatures, expected to become more frequent and severe with global warming, on cereal grain yields. The critical role of heat stress in reducing grain number is clear, though the mechanisms and sensitivities appear to vary between crops and within varieties. Understanding the direct impact of high temperature on grain formation is complicated due to the interactions with other temperature sensitive processes, such as development,
Acknowledgements
EE and FE contributions were funded by the German Science Foundation (project EW 119/5-1). HW, TG, JN and FE's contributions were funded by the Federal Ministry of Education and Research (BMBF) through WASCAL (West African Science Service Center on Climate Change and Adapted Land Use). FE also acknowledges support from the FACCE JPI MACSUR project (031A103B) through the German Federal Ministry of Food and Agriculture (2812ERA115).
References (245)
- et al.
Evaluation of drought and heat stressed grain sorghum (Sorghum bicolor) for ethanol production
Ind. Crops Prod.
(2011) - et al.
Simulated wheat growth affected by rising temperature, increased water deficit and elevated atmospheric CO2
Field Crops Res.
(2004) - et al.
Thermal acclimation and the dynamic response of plant respiration to temperature
Trends Plant Sci.
(2003) - et al.
An empirical expression to relate aerodynamic and surface temperatures for use within single-source energy balance models
Agric. For. Meteorol.
(2012) - et al.
An overview of the crop model STICS
Eur. J. Agron.
(2003) - et al.
Testing the CERES–maize simulation model in a semi-arid tropical environment
Field Crops Res.
(1989) - et al.
Simulation of the impact of high temperature stress on annual crop yields
Agric. For. Meteorol.
(2005) - et al.
Design and optimisation of a large-area process-based model for annual crops
Agric. For. Meteorol.
(2004) - et al.
Climate changes and trends in phenology of fruit trees and field crops in Germany, 1961–2000
Agric. For. Meteorol.
(2004) - et al.
Effect of heat shock during grain filling on the gluten protein composition of bread wheat
J. Cereal Sci.
(1996)