Effect of High-Temperature Events When Heading into the Maturity Period on Summer Maize ( Zea mays L.) Yield in the Huang-Huai-Hai Region, China

: The predicted increase in the frequency of extreme climatic events in the future may have a negative e ﬀ ect on cereal production, but our understanding of the historical trends of high-temperature events associated with climate change and their long-term impact on summer maize yield is limited. Based on an analysis of historical climate and summer maize yield data from 1980 to 2016 in the Huang-Huai-Hai (3H) region of China, we calculated two high-temperature event indices, namely, high-temperature hours (HTH) and high-temperature degrees (HTD, the sum of the di ﬀ erences between 35 ◦ C and above), and then investigated the temporal trend of high-temperature events from maize heading to maturity and their impact on the yield of summer maize. Our results indicated that the air temperature showed a signiﬁcant upward trend when heading into the maturity period of summer maize in the 3H region from 1980–2016 and that the increase was greater in the northern Huang-Huai-Hai (N3H) region than in the southern Huang-Huai-Hai (S3H) region. The intensity of high-temperature events when heading into the maturity period increased considerably from 1980 to 2016 in the 3H region, especially in the S3H region. The HTH and HTD increased by 1.30 h and 0.80 ◦ C per decade in the S3H region, respectively. Moreover, a sensitivity analysis of panel data showed that the increases in HTH and HTD when heading into the maturity period had a consistent negative e ﬀ ect on yield in S3H and N3H regions; this e ﬀ ect was more obvious in the S3H region. In the S3H region, a 1 h increase in HTH was found to be associated with a 0.45–1.13% decrease in yield and a 1 ◦ C increase in HTD could result in a yield loss of 1.34–4.29%. High-temperature events were detrimental to summer maize production, and the severity of this e ﬀ ect was projected to increase in the 3H region. In this study, we used two indices (HTH and HTD) to quantify the impact of high-temperature events on summer maize yield during the critical growth phase (heading to maturity) at a small timescale (hours and days). The results of this study can provide a reference for policymakers to use in the formulation of corresponding climate change adaptation strategies.


Introduction
With global climate warming, the frequency and severity of extreme weather events have increased [1,2], which have had a negative impact on the growth and development of crops [3,4]. High temperatures that exceed critical thresholds for plants are usually referred to as heat stress, and crops are more sensitive to high temperatures during their reproductive stages than during their vegetative stages [5]. During this critical growth period, even short-term high-temperature stress can cause drastic reductions in yield [6][7][8][9]. For example, high-temperature stress inhibited the elongation of northern sub-regions (Figure 1). The southern region borders the subtropical zone, while the northern region borders the mid-temperate zone. There are certain differences in climatic conditions and summer maize growth duration between the two sub-regions (Table 1). In addition, we also provide the air temperatures and average phenological dates of the summer maize growing season in the 3H region over the past 37 years (Table 2).  Annual average temperature (T); sunshine duration (SSD); precipitation (Prec); vegetative growth period (VGP); reproductive growth period (RGP); whole growth period (WGP).  Annual average temperature (T); sunshine duration (SSD); precipitation (Prec); vegetative growth period (VGP); reproductive growth period (RGP); whole growth period (WGP).

Data Sources
The data, which mainly include meteorology, phenology, and yield data, were obtained from three databases. A comprehensive analysis of these data was conducted to characterize the occurrence of high-temperature events when heading into the maturity period of summer maize and its relationship with yield in the study area. The first data set came from the National Meteorological Information Center of the Central China Meteorological Administration (http://data.cma.cn/) and contained daily maximum temperature (T max ), average temperature (T avg ), minimum temperature (T min ), and precipitation data. In the study area, 112 meteorological stations with complete records were selected and the high-temperature event indices (high-temperature hours (HTH) and high-temperature degrees (HTD)) were calculated by using the daily T max data observed when heading to maturity in summer maize during 1980-2016. These first databases included 60 and 52 stations in the S3H and N3H regions, respectively. The spatial positional data for each meteorological station were obtained from the National Geomatics Center of China (Figure 1).
The second data set consists of the phenological data (sowing, heading, and maturity dates) for summer maize from 1980-2016. In the study, we defined the vegetative growth period as that from sowing to heading and the reproductive growth period as that from heading to maturity. These data also collected from the National Meteorological Administration to evaluate the mean values of the summer maize heading and maturity dates at each station from 1980-2016, which were regularly recorded phenological dates at these stations.

Region
Month T min T avg T max SSD Prec The third database contained the planting area, yield, and annual yield of summer maize in each province from 1980-2016, and these data were extracted from the Agricultural Information Network of China. According to the ratio of the yield to the planted area of each province, we calculated the average yield for the S3H and N3H regions to estimate the impact of high-temperature events from heading to maturity on grain yield.

High-Temperature Event Indices
We treated 35 • C as the critical threshold of high-temperature stress to evaluate the impact of this stress on summer maize yield. The HTH and HTD were used as the main indices to describe the high-temperature events and then to analyze the occurrence and impact of high-temperature events from heading to maturity. Based on the high-temperature event indices at each station over the past 37 years, the temporal trends of the high-temperature index of the S3H, N3H, and 3H regions were fitted by simple linear regression to reveal the local and 3H-regionwide changes in recent decades. The HTH calculation method refers to Bristow and Abrecht [31], and the HTD calculation method was based on Shi et al. [32]. The calculation methods for HTH and HTD are as follows: where T 1 and T 2 (h) are intermediate variables; D is day length (h); T c is the critical temperature (35 • C); T maxi and T mini are the maximum and minimum temperature on day i; HH i is high-temperature hours above the temperature threshold of 35 • C on day i; d h and d m are heading and maturity dates, respectively; HTH indicates the accumulated HH i from heading to maturity; HD i is the high temperature degrees on day i, and HTD indicates the accumulated HD i from heading to maturity.

The Impact of High-Temperature Events on Summer Maize Yield
The yield of maize is affected by climatic factors, as well as non-climatic factors, such as variety improvement, fertilization, and cultivation practices. To analyze the relationship between climatic factors and yield more accurately, the influence of non-climatic factors must be eliminated [33]. Schlenker and Lobell [34] proposed that the linear and quadratic trends of the years can be considered to exclude the contribution of non-climatic factors to yield. Therefore, in this study, the nonlinear model of Lobell et al. [35] was used to perform a sensitivity analysis of the crop yield time series in detail from 1980 to 2016. A regional panel data analysis tool (Eviews6.0) was used to evaluate the effect of high-temperature events on summer maize yield. The specific model is as follows: where i is the sub-region (southern 3H (S3H) and northern 3H (N3H) region) of the study; t = 1980..., 2016; c is a fixed effect; d is the time trend coefficient; β is the independent variable parameter; X(i,t) are Atmosphere 2020, 11, 1291 6 of 12 independent variable parameters, including climatic factors, such as the average temperature (T avg ), minimum temperature (T min ), maximum temperature (T max ), and precipitation; and ε(i,t) is the error term. Statistical analyses were performed using SPSS software (SPSS 25.0, SPSS, Inc., Chicago, IL, USA). Graphics were created with Microsoft Excel 2016 (Microsoft Corporation, Redmond, WA, USA).

Variation Trends of Temperature, Precipitation, and Yield from Heading to Maturity
The changes in T avg , T max , and T min from heading to maturity in summer maize showed an upward trend in the 3H region during 1980-2016 (Table 3). For the S3H and N3H regions during the study period, T avg exhibited a significant increase and T min had an extremely significant increase, but T max only exhibited a significant increase in the N3H region only. Overall, the temperature increase was higher in the N3H region than in the S3H region, and the trend was that T min increased most dramatically, followed by T avg and T max . There was no significant change in precipitation over the study period; the precipitation increased slightly, by 1.61 mm per 10-yr period, but fluctuated greatly interannually. According to the statistics, the average yield of summer maize in the 3H region was 3335 kg ha −1 in the 1980s and rose to 5648 kg ha −1 in 2016, with an average annual increase of approximately 64.26 kg ha −1 . From 1980 to 2016, yields increased significantly in the 3H region, but the difference in yield between the S3H and N3H regions was not significant. The S3H region had the highest maize grain yield at 4951 kg ha −1 , which was 9.55% higher than the yield of the N3H region (4519 kg ha −1 ), but its coefficient of variation was also higher than that of the N3H region.

Variation Trends of HTH and HTD during the Heading to Maturity Period
The values of HTH and HTD from heading to maturity across the 3H region suggested an increasing trend, particularly in the S3H region, during 1980-2016 ( Figure 2). The HTH values increased markedly, by 1.30 h, 0.79 h, and 1.17 h per decade in the S3H, N3H, and 3H regions, respectively. During the study period, the average values of HTH were 3.94 h and 2.68 h for the S3H and N3H regions, respectively. The increases in HTD per 10-yr period were 0.80 • C, 0.41 • C, and 0.59 • C for the S3H, N3H, and 3H regions, respectively. The average HTDs between heading and maturity during the period of 1980-2016 were 2.55 • C and 1.72 • C for the S3H and N3H regions, respectively.

Variation Trends of HTH and HTD during the Heading to Maturity Period
The values of HTH and HTD from heading to maturity across the 3H region suggested an increasing trend, particularly in the S3H region, during 1980-2016 ( Figure 2

Impacts of HTH and HTD Changes on Summer Maize Yield from Heading to Maturity
Panel data analysis was performed to identify the impacts of HTH and HTD variations on yield (Figure 3). The results suggested that the increases in HTH and HTD had a negative effect on maize production. In the S3H region, an increase in HTH by 1 h corresponded to a decrease of 0.54-1.13% in yield, and a 1 °C increase in HTD corresponded to a 1.34-4.29% decline in yield, whereas in the   Note: *, **: significant at p < 0.05 and p < 0.01, respectively.

Impacts of HTH and HTD Changes on Summer Maize Yield from Heading to Maturity
Panel data analysis was performed to identify the impacts of HTH and HTD variations on yield (Figure 3). The results suggested that the increases in HTH and HTD had a negative effect on maize production. In the S3H region, an increase in HTH by 1 h corresponded to a decrease of 0.54-1.13% in yield, and a 1 • C increase in HTD corresponded to a 1.34-4.29% decline in yield, whereas in the N3H region, a 1 h increase in HTH corresponded to a decrease of 0.17-0.89% in yield, while a 1 • C increase in HTD corresponded to a 0.21-0.71% decline in yield. In summary, the high-temperature events have a greater impact on the summer maize yield in the S3H region than in the N3H region.

Discussion
With climatic warming, the frequency and severity of high-temperature events during the reproductive growth period of crops have increased remarkably, and high-temperature events, as a serious threat to maize production, have been a growing concern [36]. Crops have different optimum and critical temperature thresholds in different development and growth phases; when these thresholds are exceeded, a series of physiological processes associated with yield formation are affected, leading to dramatic yield losses [17]. In fact, many studies have reported on the impact of high temperatures around the flowering period on maize, but information obtained from applying indices of high-temperature events to study their influence on the development and yield of summer maize from heading to maturity in the 3H region is inadequate. We used quantitative methods to estimate the variation characteristics of high-temperature events in recent decades based on phenological and meteorological data from the major summer maize production areas in China. Our results indicated that the intensity of high-temperature events between heading and maturity notably increased during 1980-2016 and that there was a difference in their occurrence frequency between the S3H and N3H regions.
In recent decades, the air temperature across the 3H region during the heading to maturity period of summer maize showed a significant upward trend. Tmin increased by 0.40 °C per decade, that is, the night-time temperature increased rapidly and the variation trend was mostly consistent with the trend of global climate change [37,38]. Bahuguna et al. [39] found that warmer nights promote night respiration rates and increase the consumption of photosynthate, resulting in decreased biomass. Wang et al. [21] suggested that at the 30 °C night temperature, yield losses mainly result from a decrease in kernel number rather than in kernel weight under high-temperature stress; compared with those in a control treatment, the maize yield and grain number under a warming treatment decreased greatly by 23.8% and 25.1%, respectively.
In this study, we assessed the response of summer maize yield to HTH and HTD by analyzing

Discussion
With climatic warming, the frequency and severity of high-temperature events during the reproductive growth period of crops have increased remarkably, and high-temperature events, as a serious threat to maize production, have been a growing concern [36]. Crops have different optimum and critical temperature thresholds in different development and growth phases; when these thresholds are exceeded, a series of physiological processes associated with yield formation are affected, leading to dramatic yield losses [17]. In fact, many studies have reported on the impact of high temperatures around the flowering period on maize, but information obtained from applying indices of high-temperature events to study their influence on the development and yield of summer maize from heading to maturity in the 3H region is inadequate. We used quantitative methods to estimate the variation characteristics of high-temperature events in recent decades based on phenological and meteorological data from the major summer maize production areas in China. Our results indicated that the intensity of high-temperature events between heading and maturity notably increased during 1980-2016 and that there was a difference in their occurrence frequency between the S3H and N3H regions.
In recent decades, the air temperature across the 3H region during the heading to maturity period of summer maize showed a significant upward trend. T min increased by 0.40 • C per decade, that is, the night-time temperature increased rapidly and the variation trend was mostly consistent with the trend of global climate change [37,38]. Bahuguna et al. [39] found that warmer nights promote night respiration rates and increase the consumption of photosynthate, resulting in decreased biomass. Wang et al. [21] suggested that at the 30 • C night temperature, yield losses mainly result from a decrease in kernel number rather than in kernel weight under high-temperature stress; compared with those in a control treatment, the maize yield and grain number under a warming treatment decreased greatly by 23.8% and 25.1%, respectively.
In this study, we assessed the response of summer maize yield to HTH and HTD by analyzing panel data, and a negative effect on yield caused by the increase in high-temperature events was noted across the S3H and N3H regions. High temperatures above 35 • C in the 3H region usually occur in July and August, but, in this period, the pollination stage of summer maize also occurs. High temperatures prolong the anthesis-silking interval, decrease pollen activity, induce pollen sterility, and inhibit pollen tube growth, which induces yield reduction [19,28]. Additionally, Zhang et al. [27] reported that high temperatures and droughts generally occurred between the tasseling and kernel milk stages in the 3H region in recent decades and the occurrence of high temperatures increases potential evaporation and reduces the supply of soil water [40], which could be another reason for the decline in production. The background temperature of the S3H region was higher than that of the N3H region, and the time trend analysis revealed that the intensity of high-temperature events increased faster in the S3H region, with HTD and HTH increasing by 0.80 • C and 1.30 h per decade, respectively. Consequently, a more severe negative effect on maize yield was recorded in the S3H region than in the N3H region. It should be noted that temperature variability and extreme temperature events will become more prevalent in the future due to the impact of climate change.
The sowing calendar and the crop growth rate affect whether the critical phenology phase coincides with the high-temperature period. Adjusting the sowing date is an important adaptation strategy for avoiding the risks of high temperature and drought during the critical period for maize, ensuring yield and adapting to climate change [41][42][43]. However, a winter wheat-summer maize rotation system was implemented across most of the 3H Plain. Adjusting the sowing window of summer maize will have an impact on the winter wheat growing season [44]. Yadav et al. [45] suggested that heat-tolerant maize genotypes have higher photosynthetic capacity under heat stress, can better adapt to high-temperature environments, and the yield is less affected by high temperatures. New climate resilient maize showed better adaptability and yield performance than current major commercial varieties under high temperatures, low input, and drought conditions [46,47]. Therefore, selecting and cultivating heat-tolerant summer maize varieties is an effective measure to reduce the damage of high temperatures to maize.
In this study, we used meteorological and yield data from the heading to maturity period of summer maize in the 3H region from 1980 to 2016 to evaluate the impact of high-temperature events on yield. However, crop varieties, soil properties, irrigation management, and cultivation technology are also important limiting factors that contribute to maize yield and yield stability [48,49]. Therefore, to assess the influence of climatic change on crop production more accurately in future studies, these factors should be emphasized and assessed individually.

Conclusions
From 1980-2016, both the air temperature and the intensity of high-temperature events during the maize heading to maturity period across the 3H region showed significant upward trends. The increases in HTH and HTD had a negative effect on the summer maize yield in the S3H and N3H regions, but this effect was more severe in the S3H region. Given the uncertainties surrounding climate change and the increase in extreme high temperature events, and considering the close connection between the planting times of winter wheat and summer maize, improving the heat tolerance of maize through variety improvement is an effective strategy for addressing this issue.
Author Contributions: Conceptualization, S.W. and C.C.; methodology, S.W. and C.C.; software, S.W. and C.C.; formal analysis, S.W. and C.C.; writing-original draft preparation, S.W. and J.L.; writing-review and editing, J.L., T.L., X.W., and A.P. All authors have read and agreed to the published version of the manuscript.