The locust plagues in the Yangtze River Delta of China during the Ming and Qing Dynasties

The Little Ice Age was concurrent with frequent natural disasters in China during the Ming and Qing Dynasties (1368–1911 AD). Using historical documents created in the Yangtze River Delta during these dynasties, we reconstructed the severity and frequency of locust plagues. We uncovered their temporal and spatial patterns, the relevant influencing factors, and the applied control measures. These locust plagues occurred in the Yangtze River Delta during summer and autumn and were considered moderate-degree disasters. The locust plagues during the Qing Dynasty were less harmful than those during the Ming Dynasty. The locust plagues occurred north of 31°N and were concentrated around lakes and rivers. Regarding long-term changes, three periods of frequent locust plagues were identified: 1520–1540 AD, 1610–1670 AD, and 1850–1870 AD. The locust plague cycles were dominated by long-term changes of 100 and 350 years, alternating with short-term changes of 10–30 years. Locust growth and development typically occur under suitable climate conditions. However, in the Yangtze River Delta, the relationship between locust outbreaks and temperature changes was insignificant, whereas flood disasters provided suitable humidity conditions for locust nymph growth. Locust plague outbreaks had a series of social impacts, one direct result was rising food prices. When facing locust plagues, the Ming and Qing Dynasties formed effective top-down preventative measures corresponding to local conditions. The locust plagues in the Yangtze River Delta during the Ming and Qing Dynasties occurred with regularity, were responsive to local hydrological conditions, and impacted social economies and technologies.


Introduction
As one of the major natural disasters that occurred in ancient China, locust plagues critically impacted agricultural production. Throughout history, especially during the Ming and Qing Dynasties, an abundance of literature depicted locust plague records. From the perspective of disaster geography, analyzing the spatiotemporal characteristics of locust plagues, investigating the significance of climate change and social influences on these plagues, and exploring the formation mechanism of locust plagues have important historical and practical value. Locust plagues occur with strong phase, concentration, mobility and extensiveness characteristics (Zhang 2008). In China, the Oriental migratory locust (Locusta migratoria manilensis Meyen.) is one of the most destructive insects (Zhao et al. 2020). This locust is widely distributed in China, occurring south of 42°N in the north, east of 105°E in the west, in coastal areas in the east, and on Hainan Island in the south (Ma 1965).
Changes in natural conditions directly affect the temporal and spatial evolution characteristics of locust plagues. Many studies have been carried out on the relationships between historical locust plagues and climate change. It has been confirmed that the primary factors that caused plagues of the Oriental migratory locust in the historical period of China (before 1911 AD) were related to temperature, drought, and wet factor (Brázdil et al. 2014;Tian et al. 2011;Fei and Zhou 2016). However, previous studies have ignored the differences in climatic conditions among different regions. In addition, some researchers have investigated various natural disasters that cause locust plagues and have generally concluded that climate change triggers droughts and floods, which in turn trigger locust plague outbreaks (Yan et al. 2021;Yu 2009). A locust plague is a biological disaster, and the natural conditions supporting the occurrence and development of locusts are similar among locust plagues. Studies on modern locust plagues (Propastin 2012;Zhao et al. 2020) show that locust spawning sites include riversides, lakesides, and shallow beaches. In years with dry spring and summer seasons, the low water levels of rivers expose the shorelines; this process is conducive to locust reproduction. However, the relationships between historical locust plagues and changes in the hydrological environment have not been clearly elucidated.
Due to the economic system and religious beliefs before the Tang dynasty (618 AD), ancient farmers were not active in locust control. However, with the development of the economy and technology, human attitudes toward locust plague gradually shifted from passive tolerance to active prevention and control. Examples of active measures include the use of plowing tools, which could greatly reduce the survival rate of eggs (Dağyeli 2020;Trumper et al. 2022). The Ming and Qing Dynasties belonged to the embryonic period of Chinese capitalism. With the accumulation of capital, various new ideas and new technologies emerged, which indirectly affected the defensive measures implemented against locust plagues. There have been abundant studies on the impact of humans on locust plagues; however, more work needs to be done to integrate the existing research results and the distribution law of locust plagues obtained in this paper to understand the impact of human factors on locust plagues.
For the long-term series of locust plague research in a certain area, more detailed analysis can be made and more targeted response measures can be proposed. Existing regional studies in China include the Hunan-Jiangxi-Hubei Province (southern China) (Gao et al. 2021), and the Yellow River Basin (northern China) (Fei and Zhou 2016). These studies have elaborated on the occurrence of locust and the response to the global climate. Nevertheless, eastern China has abundant water and heat resources, and systematic research on the locust plague in the Yangtze River Delta is still insufficient. The Yangtze River Delta includes Zhejiang Province, Jiangsu Province, and Shanghai. This region located in the monsoon zone of eastern China (Fig. 1), which has excellent agricultural conditions and for which rich historical records and documents of natural disasters such as locust plagues have been kept. This area is favorable for conducting geographical research on historical disasters.
Based on the valuable historical records of the Ming and Qing Dynasties (1368-1911 AD), a series of locust plague indexes for the Yangtze River Delta are constructed herein. By applying spatiotemporal dynamic statistical methods, we elaborate on the spatiotemporal change characteristics of locust plagues. Based on historical documents and modern locust plague theory, we examine the causes of locust plague formation, relationships between the occurrence of locust plagues and climate conditions, impacts of these plagues on society, and historically established prevention measures. This article supplements the existing record of ancient locust plagues, enriches historical disaster research cases, and provides references for future disaster prevention.

Data sources
The locust plague data were mainly based on the "A Compendium of Chinese Meteorological Records of the Last 3000 Years" (Zhang 2013) (hereafter referred to as the "Compendium") and local chronicles, combined with the "Chinese Locust Plague History" (Zhang 2008), which supplements the locust plague records. The collected data included the time, location, disaster situation, and source of the locust plagues from 1368 to 1911. Considering the rule of "recording exceptions and not routines" in ancient documents, this study regarded the years without locust plague records to be disaster-free years or years with fewer large-scale and extremely harmful locust plague events.
Auxiliary analysis data included historical grain price data, temperature data, and other disaster data. The grain price data came from the research results of the China Economic Research Center of Peking University (Lu and Peng 2005); the Suzhou rice prices from 1644 to 1910 were extracted from the data using silver taels/stone as the unit. The air temperature data came from the millennium-scale air temperature data constructed by the Chinese Academy of Sciences. This study selected the annual average temperature in eastern China from 1368 to 1911; the unit is Celsius (Shi et al. 2012). The historical drought and flood data were selected from the Compendium and screened, and the corresponding statistics were calculated. This study also considered the impact of rivers and lakes on the dynamic changes in locust plagues; hence, the Chinese historical atlas (Tan 1982) was referenced to attain vectorized information for rivers and lakes based on the maps of 1582 and 1820, respectively.

Processing of locust plague data
The locust plague record contains the majority of information about the disaster. In this study, more details were extracted from the documents, including the month and the direction of migration for locusts. The records of the occurrence date of locust plagues generally take the month as the time scale and the lunar calendar as the record, such as "Xuzhou locusts in autumn and July." Therefore, this article includes statistics on the months when the locust plague occurred; a total of 433 historical documents were obtained. In addition, to clarify the path of locust migration, vocabulary related to direction was extracted from the document records. In ancient Chinese literature, the migration path is described in qualitative terms, e.g., "locusts come from the north." A total of 79 documents included orientation records. To visualize these data, this article converts the directional vocabulary into azimuth, that is, starting from the north, and measuring clockwise from the north. The scope of the azimuth varies from 0° to 360°, north is 0°, east is 90°, south is 180°, and west is 270°.
In past research on historical disasters, descriptions found in ancient documents have been widely used to determine the severity levels of historical disasters (Stige et al. 2007;Zheng et al. 2006). According to the duration of the locusts (such as one generation or successive generations of locusts, one year or multiple years of locusts), range of occurrences (in the region or nationwide), degree of agricultural damage, and impact on humans (production reduction, famine, cannibalism, relief, and refugees), the locust plagues were Table 1 The grade standard of historical locust outbreaks (Huang et al. 2018) Note that in Northeast China, there were usually two outbreaks of locusts per year, one in summer and one in autumn  Table 1.
Any year in which locust plagues are recorded in no fewer than 5 counties (rounded by the mean) is considered a locust plague year. A year in which locust plagues are recorded in no fewer than 12 counties (after rounding the sum of the mean and one standard deviation) is marked as a large locust plague year. A single year and multiple consecutive years with large locust plagues are defined as extreme locust plague events.

Statistical and analytical methods
For the spatial scale, the smallest spatial unit of locust plague data is the county, and the changes in administrative boundaries due to dynasties had little impact on the records of disaster threats at the county level (Zheng et al. 1993). Therefore, the county-level historical data used in this study were reliable. Thus, this article uses "county" as the spatial statistical unit to calculate statistics on the occurrence frequency and cumulative frequency of locust plagues. To construct the map and statistics of the frequency of locust plagues, the latitude and longitude of the current government seat of each county are taken as the coordinates of each county.
To understand the time dynamics of the locust plagues, which spanned more than 500 years in the Ming and Qing Dynasties, an analysis of time series changes in the locust plagues considered high-frequency outbreak periods and periodic assessments. To clarify the high-frequency period of locust plague outbreaks in the Ming and Qing Dynasties, this paper uses scales of 50 years and 10 years for temporal dynamic statistical analysis of the counties where locust plagues occur. The wavelet analysis method is used for the periodicity of the locust plague. For a given scale, the wavelet variance represents the strength (energy size) of the periodic fluctuation for its time series. Wavelet variance based on scale can reflect various time scales (periods) contained in the time series and the variation characteristics of their strengths and weaknesses. The scale at the peak is the main period of the sequence, and the other heights are the subperiods. Based on the wavelet analysis tool in MATLAB, wavelet analysis was performed on the locust plague frequency in the Yangtze River Delta, and its periodic variation characteristics were clarified.
In addition, due to the long time period, the changes in the locust plague pattern and the response measures are difficult to analyze with equal time intervals, so the historical period is divided into two time slices for the Ming Dynasty and the Qing Dynasty. When comparing the time series changes in locust plagues with droughts, floods, and rice prices, to reduce the influence of special values in a certain year and facilitate the comparison of the change trends of each series, the annual data are processed using a 10-year moving average. When comparing locust plagues with temperature, droughts, and floods, in addition to comparing the trends, the correlation coefficient calculations and significance tests are used to analyze the long-term series of locust plagues (frequency) and temperature (°C), drought (frequency), and waterlogging (frequency).

Spatial patterns of locust plagues
During the Ming and Qing Dynasties, there were 1,306 records of locust plagues in the Yangtze River Delta region. Most of these locust plagues occurred in Huai'an city, where a total of 37 locust plagues occurred over 544 years. The counties in which the most locust plagues arose during the Ming Dynasty were Dongtai city and Huai'an city, both of which experienced 24 locust plagues. The county in which the most locust plagues occurred during the Qing Dynasty was Xuyi county, with a total of 20 outbreaks. Dongtai city, Huai'an city, and Xuyi county are all located between 32°N and 33°N in the northern part of the Yangtze River Delta.
The locust plagues that occurred in the Yangtze River Delta during the Ming and Qing Dynasties exhibited significant regional characteristics. Regarding the latitude, 31°N can be used as the dividing line. The region north of 31°N (containing 1,042 records in total) experienced more locust plagues than the area south of 31°N (containing 260 records in total) ( Fig. 2a and b). In the region north of 31°N, the incidences of locust plagues exhibited a significant hydrophilic effect ( Fig. 2c and d), namely, the locust plague outbreaks were concentrated around lakes and rivers. This feature was also consistent with the findings of other studies (Li et al. 2017). For more information regarding the relationship between locust plagues and the hydrological system, please refer to Sect. 4.2.
For locust migration (Fig. 2e), there were 41 records of locusts migrating from the north, 25 records of locusts migrating from the northwest, and 13 records of locusts migrating from the east, northeast, southeast, south, or west. Regarding the regional distributions, most locust migration was recorded within 31.01-32.99°N, accounting for 65.82% of the total records (52 of 79 records). In this zone, locusts from the north and northwest accounted for 80.77% of the total records (42 of 52 records).

Temporal dynamics of locust plagues
For the occurrence times of locust plagues, statistics show that locust plagues mainly occurred from May to August in the lunar calendar, as reflected by 83.37% of the plagueindicating documents, while locust plagues occurred less frequently from September to April of the following year. In ancient China, the four seasons were divided based on the lunar month. That is, spring lasted from January to March in the lunar calendar, summer lasted from April to June, autumn lasted from July to September, and winter lasted from October to December. Thus, according to statistics, locust plagues mostly occurred in the summer and autumn seasons.
The Ming and Qing Dynasties (1368-1911 AD) endured for 544 years. Locust plagues occurred in the Yangtze River Delta for at least 270 years of the dynasties. The locust plague frequency derived herein was 49.63%, with an average rate of one plague every 2.0 years. During the Ming Dynasty, locus plagues were recorded in 125 years, at an average recurrence rate of 2.2 years; during the Qing Dynasty, locust plagues were recorded in 142 years, with an average recurrence frequency of 1.9 years. These findings suggest that locust plagues occurred more often during the Qing Dynasty than during the Ming Dynasty. During this long historical period spanning more than five centuries, three periods of recurrent locust plagues were identified (Table 2), namely, the first half of the sixteenth century 1 3 (1501-1550 AD), the first half of the seventeenth century (1601-1650 AD), and the latter half of the nineteenth century to the beginning of the twentieth century (1851-1911 AD). The disaster classes associated with these locust plagues were primarily above the second level, indicating that these locust plagues not only affected wide spatial ranges but also caused high degrees of harm. Throughout the Ming and Qing Dynasties, the occurrences of locust plagues were at levels 2 and 3 (80.63% of all records), and the average grade was 2.23. Therefore, most locust outbreaks were predominantly moderate disasters. However, during the Qing Dynasty, the frequency of levels 3 and above locust plague outbreaks was lower than the frequencies of levels 1 and levels 2 outbreaks. This finding indicates that the locust plagues that occurred during the Qing Dynasty were less dangerous than those that occurred during the Ming Dynasty. The locust plagues that occurred in the regions south and north of 31°N were dominated by levels 2 and 3, accounting for 83% and 80% of the regional records, respectively; the corresponding average levels were 1.86 and 2.78, respectively. These results show that the damage caused by locust plagues was more profound in the area north of 31°N than in the region south of this line of latitude.
According to the frequency and grade characteristics of locust plagues described above, the periods of locust plague occurrence were established in depth, as shown in Fig. 3. Three high-locust plague incidence periods were identified, namely the periods spanning 1520-1540 AD, 1610-1670 AD, and 1850-1870 AD. During these periods, locust plagues occurred in 610 counties, accounting for 58.60% of the total affected counties, with an average occurrence rate of 47 counties every ten years. From the 1620 s to the 1640 s, a total of 203 counties suffered locust plagues, and 16 locust plague years occurred (with ≥ 5 affected counties). From the 1710s to the 1770s, 398 counties experienced locust plagues. This period encompassed the Wanli locust plague (1615-1617 AD) and the Chongzhen locust plague (1635-1641 AD) of the Ming Dynasty and the Kangxi locust plague (1671-1672 AD) of the Qing Dynasty. In the three disastrous locust plagues, a total of 243 counties were affected, resulting in a wide spatial range and severe damage. From the 1850s to the 1870s, a total of 183 counties recorded locust plagues, and seven locust plague years were recorded (in which ≥ 5 counties were affected); the Xianfeng locust plague of the Qing dynasty occurred from 1856 to 1857 in a total of 89 counties. A total of 76 locust plague years (with ≥ 4 affected counties) occurred in the region north of 31°N, and 24 locust plague years (with ≥ 3 affected counties) occurred in the region south of 31°N. The locust plague outbreaks in the zones north and south of 31°N were synchronized temporally, but the locust plague years lasted longer in the region north of 31°N than in the southern area. The results also showed that the locust plagues that occurred in the area north of 31°N were more serious than those in the area south of 31°N.
Although three peak locust outbreak periods were identified during the Ming and Qing Dynasties, clarifying whether periodic changes could be found in the long-term series was still necessary. According to wavelet analysis of the locust plague frequency in the Yangtze River Delta (Fig. 4), locust plagues occurred at multiple time scales, with variance value peaks at timescales of 350 years, 200 years, 100 years, 30 years, 20 years, and 10 years. The 100-year and 350-year timescales corresponded to significant changes throughout the cycle, with strong oscillations and alternating periods of infrequent and frequent locust    plague occurrences. Therefore, the locust plague outbreak cycle in the Yangtze River Delta was dominated by long-term changes on the scales of 100 and 350 years, with short cycles (10-30 years) alternately appearing.

Comparisons of temporal and spatial locust distributions
Most of the migratory locusts in the Yangtze River Delta occur from May to August, while the peak of locust outbreaks in Hunan, Jiangxi, and Hubei Provinces (southern China) is from June to August (Gao et al. 2021). Variations in climate, topography, and altitude cause slight differences in the time of locust concentration. In general, outbreaks of migratory locusts largely occur in summer and autumn. In this study, the outbreak periods of locusts were led by long-term changes of 100 years and 350 years and alternated with short-term changes of 10-30 years. Although the periodicity of locust occurrence has not yet been determined, an 80-year cycle of climate oscillations has been identified in China (Wang et al. 1981;Shen et al. 2006). Furthermore, Zhang (2009) 1368 1388 1408 1428 1448 1468 1488 1508 1528 1548 1568 1588 1608 1628 1648 1668 1688 1708 1728 1748 1768 1788 1808 1828 1848 1868 1888 1908 All North of 31°N Southof 31°N Fig. 3 The locust plague index map in the Yangtze River Delta during the Ming and Qing Dynasties. This figure is divided into three parts: gray signifies the entire study area, dark blue symbolizes the north of 31°N, and light blue represents the south of 31°N. The circle is the number of counties where the locust plague occurred. A higher frequency means a larger ring. The shade is the locust plague year (the counties where the locust plague occurred are ≥ 5, ≥ 4, and ≥ 3) Fig. 4 Wavelet analysis of the counties that suffered from locust plague in the Yangtze River Delta during the Ming and Qing Dynasties argued that 160-170-year cyclical climate cooling causes locust outbreaks by increasing temperature-related drought and flood events. However, in this study, only the Yangtze River Delta was analyzed, and the obtained periodicities were based on locust records, which were inevitably affected by human defenses. In the Yangtze River Delta, most locusts come from the north and northwest. In the north and northwest of the study area is the Huang-Huai River Basin, which is one of the regions hit the hardest by locust plagues (Stige et al. 2007). Therefore, in the event of a nationwide locust plague, locusts in the Huang-Huai River Basin are likely to migrate to neighboring provinces and counties. Locusts in the Yangtze River Delta have significant "hydrophilic" characteristics, but this is not a feature unique to the Oriental migratory locust. The Amudarya River Delta and Ili River Delta are significant breeding regions for the Asian locust (Locusta migratoria L.) due to seasonal floods and the rise and fall of river and lake levels (Sivanpillai et al. 2006;Löw et al. 2016). In addition, related studies (Min 2004;Jiang and Li 2020) have shown that in both long-term series and disaster events, locust plagues in Jiangsu are more regular and serious than those in Zhejiang. Moreover, 31°N is generally consistent with the boundary of Jiangsu and Zhejiang Provinces, so it is practical to use 31°N as the demarcation line for the spatial distribution of migratory locusts. The conclusions presented in this paper regarding the periodicity, migratory patterns, and spatial distributions of Oriental migratory locust are largely consistent with those of previous studies.

Locust plague dynamics and climate change
In terms of biology, the hatching process of locust eggs is reliant on the soil temperature, soil moisture, and weather conditions (Wang et al. 2020;Zhang et al. 2009). However, temperature and precipitation changes affect the seasonality and number of locusts during only some years and do not alter the nature of any given region in terms of the preferential locust areas (Zhao et al. 2020). Therefore, only optimal, appropriate climatic conditions are suitable for the development and growth of locusts. The temperature range required to commence embryonic development in migratory locusts is 15-18 °C. Within the temperature range of 26-32 °C, the developmental speed of locust eggs is positively correlated with temperature, namely, the higher the temperature is, the faster the development speed is (Ma 1958). Relatively warm temperatures increase locust survival in winter, enhance nymphal aggregation, and encourage flight in adult locusts. In this study, we collected the annual average temperature data throughout the historical period (Fig. 5), but we found no significant correlation (r = 0.052) between locust plagues and temperature. This result was also consistent with the conclusions of other studies (Tian et al. 2011).
Locust egg development is heavily dependent on soil moisture, and locust plagues typically occur during drought years or after floods. During dry periods, as water levels drop, habitats suitable for locust spawning appear on riverbanks and lakefronts, and in the years following floods, flood-affected regions also provide suitable habitats for locusts (Stige et al. 2007). Modern biological research on migratory locusts has shown that liquid water is a necessary condition for the development of locust eggs (Qing et al. 1958). The suitable water contents for the spawning of East Asian locusts are 10%-20% in sandy soils, 15%-18% in loamy soils, and 18%-20% in clayey soils (Chen and Kang 2002). If the soil moisture is below 5% or above 25%, the egg number is greatly reduced (You et al. 1958). Nevertheless, the mortality rate of eggs in wet soils is higher than that 1 3  (Qi et al. 2007). Therefore, if the soil moisture increases, the mortality rate of the eggs within the soil also increases.
The temperatures in eastern China were compared with the frequency of locust plagues in the Yangtze River Delta (Fig. 5). Locust outbreaks occurred in both cold and warm periods, and the relationship between the occurrence of locust plagues in the Yangtze River Delta and short-term temperature changes was not significant. In this study, the nymph-bearing areas were primarily concentrated in the region north of 31°N (85.81%). In addition, 82.6% of the area in which nymphs were regularly found (the upper 30% of the nymph records) also contained high incidences of flooding zones (the upper 30% of the flooding records). Among the locust plague years (frequency ≥ 5), flood years (frequency ≥ 11) and drought years (frequency ≥ 8) accounted for 39.44% and 67.61%, respectively. Figure 5 implies that locust outbreaks and floods did not show notable regularity (r = 0.016) in their time series; however, outbreaks were closely associated (r = 0.45, P < 0.01) with drought periods. This is because locusts do not undergo true diapause. Drought conditions thus force locusts to migrate and spread out from their breeding grounds to maintain population-level reproduction. Therefore, long-term and large-scale droughts reliably lead to locust plagues (Ma 1958;Peng et al. 2020).
In summary, the temperatures in the Yangtze River Delta provide the basic suitable heat conditions for the dispersal of Oriental migratory locusts. The hydrological environment influences the growth of locust populations, as locusts are more likely to breed in flooded areas; however, the probability of locust plague outbreaks is highest under drought conditions.

Social impacts
Oriental migratory locusts mainly eat gramineous plants. Among cultivated crops, locusts like to eat wheat, rice, millet, barley, and sorghum. Therefore, reductions in or failures of these food crops represent direct losses often associated with locust disasters. During the Ming and Qing Dynasties, wheat and rice were the main food crops grown in the Yangtze River Delta, and the planting areas of these crops were extensive (Li 1994). The monthly locust plague statistics considered in this study showed that 85.7% of locust plagues occurred from May to August; that is, the wheat and rice growth and harvest seasons coincided with the active locust periods. Therefore, large-scale locust plague outbreaks often resulted in reduced crop yields or even lost harvests, thus resulting in famine. There are many descriptions of locust plagues and famines and even records of cannibalism in the historical materials consulted for this work.
Changes in food prices can reflect the impacts of disasters on socioeconomic development and sociopolitical stability. Although the price of rice in the Qing Dynasty (Lu and Peng 2005) showed a fluctuating upward trend, the years in which locust plagues occurred generally corresponded to years in which grain prices rose. However, there were also cases of years with a low frequency of locust plagues corresponding to years with rising food prices; these anomalies may have been related to other disaster events. After the beginning of the nineteenth century, food costs showed a trend of high prices and slow growth, but no significant natural disasters, epidemics, wars, or particular ecological shifts occurred during this period. This phenomenon can be attributed to the decreased money supply and the altered ratio of silver to copper (Zhu 2014). It should also be noted that lag effects occur when rice prices rise due to disasters; these lag effects are related to the various directions, rates, and magnitudes of disaster transmission conditions in different years.

Locust plague prevention measures
During the Ming and Qing Dynasties, the governments and populations formed various systems and measures to deal with locust plagues. A series of top-down preventative measures were formed in correspondence with the policy system conditions, popular cognition, and prevention or control techniques.
To set up policies and systems to defend against and manage locusts, the governments of the Ming and Qing Dynasties followed the traditions and laws of previous dynasties to control locusts. The Ming Dynasty court and local governments of the disaster-stricken provinces sent specifically appointed officials to organize and guide locust hunts. During the Qing Dynasty, the rulers paid extraordinary attention to locust hunting, and locust hunting management systems were gradually formed and monitored by the emperor and local governments through a combination of rewards, punishments, related laws and regulations (Zhu 2012). For example, the Law of the Qing Dynasty and Regulations for the Ministry of Revenue listed highly detailed rules for addressing the control of locusts. However, due to the influence of the feudal system and the state of the economy, this locust control system operated effectively only when the country's economic power was strong and the governmental control system was complete.
The transformation of ancient Chinese locust control work from the passive act of praying to the active methods of catching and killing locusts occurred during the Kaiyuan age (716 AD) of the Tang Dynasty (Ni 1998). After hundreds of years of experiencing locust plagues, the emperors, ministers, and people of the Ming Dynasty no longer generally deified locusts; instead, they adopted methods for both killing and blessing the locusts to combat locust plagues. Following an outbreak, the rulers and people first took action by hunting and killing the locusts. However, at the same time, influenced by the philosophical view of "heaven-human induction", the rulers of the Ming Dynasty never gave up behaviors to cultivate morality and regarded prayer activities as a significant aid to locust hunting. For this reason, the ruling class also retained various forms of sacrificial and invocation behaviors; these behaviors were also popular among the commonfolk (Wang 2014). After the Qing Dynasty, although praying for blessings and the elimination of disasters still existed in the ideology of farmers, the ruling class took more active measures to address locust plagues. For example, Emperor Kangxi not only issued orders to prevent locust outbreaks many times but also personally visited locust areas to conduct investigations and research and wrote Catching Locusts, in which he discussed the growth laws of locusts and methods for catching locusts in detail. Emperor Qianlong believed that the worship ceremonies often performed before locust hunting were cumbersome, useless and unnecessary and that governmental energy should be devoted to locust hunting (Zhu 2012). In addition, the government of the Qing Dynasty actively promoted locust control initiatives by printing and distributing various booklets and notices on locust management; thus, the concept of locust control profoundly affected more people.
During the Ming and Qing Dynasties, a variety of locust control technologies and methods were developed and accumulated, including artificial, agricultural, biological, and chemical control methods. Although many effective techniques were proposed during the Ming Dynasty, these processes were enhanced, and more were added during the Qing Dynasty based on the experiences of the Ming Dynasty. For example, Emperor Kangxi improved locust-killing technologies and improved the previous method of "following and hitting" to develop the "head-on-head hitting" method, which has a prominent capacity to kill locusts following disasters. Furthermore, during the Kangxi period, a land reclamation policy was vigorously implemented; e.g., in the abandoned Huaihe River region, the soils that were suitable for locust breeding were destroyed in accordance with this land reclamation process (Ma 2002). Therefore, the Ming and Qing Dynasties developed both similar and diverse locust plague prevention and control measures. The actions taken by members of these two dynasties thus effectively influenced the severity of the locust plagues but exhibited different restraint levels.

Conclusion
In this study, based on the historical records of the Ming and Qing Dynasties, we statistically analyzed the dynamics of historical locust plagues in the Yangtze River Delta and discussed the climatic, environmental, and social impacts of these plagues and the responses to them.
During the Ming and Qing Dynasties, locust plagues in the Yangtze River Delta primarily occurred from May to August in the lunar calendar; that is, locust plague outbreaks were most likely to occur in summer and autumn. During the Ming and Qing Dynasties, locust plagues occurred once every two years on average and were dominated by moderate disaster levels. Although locust plagues broke out more frequently in the Qing Dynasty than in the Ming Dynasty, the Qing Dynasty experienced less dramatic plagues than the Ming Dynasty. Throughout the study period, which lasted more than five centuries, we identified three high-locust plague incidence periods, namely 1520-1540 AD, 1610-1670 AD, and 1850-1870 AD. During these highlighted periods, the Yangtze River Delta also experienced many typical locust events, such as the Wanli and Chongzhen locust plagues during the Ming Dynasty and the Kangxi and Xianfeng locust plagues during the Qing Dynasty. In terms of the outbreak cycle of locust plagues, a cycle was dominated by longterm changes on timescales of 100 and 350 years alternating with short-term cycles of 10-30 years. For the spatial distribution of locust plagues, the plagues were concentrated in the region north of 31°N, especially around lakes and along riverbanks, where the locust plague situations were severe due to environmental conditions. The locusts in the Yangtze River Delta mostly migrate from the north and northwest and move into the area of 31.01-32.99°N.
The suitable climatic conditions in the Yangtze River Delta provide a guaranteed optimal temperature for locust plagues, however, the relationship between locust plagues and short-term temperature conditions was not found to be significant in this study. Moreover, as locusts are affected by the hydrological environment, they have a relatively high potential to cause disasters in zones flooded by rivers or lakes and exhibit noteworthy hydrophilic characteristics. One direct social impact of locust plagues is declining agricultural production, which can even cause famine and economic turmoil through factors such as rising food prices. To address these locust plagues, the Ming and Qing Dynasty governments developed various systems and measures according to local requirements. Nonetheless, due to the influence of their policies, regimes, public awareness levels, and prevention and control technologies, the locust control actions developed during the Qing Dynasty were more influential than those of the Ming Dynasty. Therefore, the above conclusions provide historical evidence for the prevention and control of modern locust plagues in the Yangtze River Delta and can support sustainable agricultural development.