4. Discussion **4. Discussion**

#### *4.1. E*ff*ect of Climate Change on the Hot Spot of War*

In this study, we attempted to quantify the cluster e ffect of war and determine the spatiotemporal pattern of war hot spots in historical China by emphasizing a possible climatological root. On the whole, warm and wet (cold and dry) climate and the accompanying shifts in climatic especially temperature zones, may have brought about the north/west/northwestward (south/east/southeastward) movements of agricultural and pastoral zones. The spatial distributions of war hot spots, particularly the intensifying ones, varied correspondingly. Notably, di fferent from traditional hot spots, those in this research contain temporal information (i.e., time steps/series in each bin). The proportions of time steps of later periods were larger in southern hot spots, which was consistent with the overall decline in temperature throughout the last two millennia (in particular, during the Little Ice Age (LIA), Figure 1a). In comparison, precipitation fluctuated as cycles without a significant trend (Figure 1b), which shows that the N–S disparities of hot spots may have been more temperature-dependent. The cyclic and long-term temperature variations and the resultant patterns of war hot spots in this work are similar to the findings in Zhang et al. [21], who used the Standard Deviational Ellipse as another spatial analytical tool but for the detection of directional characteristics/differences of war during three secular temperature cycles.

Specifically, as shown by the results in Figures 3 and 4, two categories of war—i.e., agri-nomadic conflict and rebellion, which responded to climate change differently, had different hot spot patterns. First, for agri-nomadic conflict, the strength of agricultural empire rose under warm and wet conditions, so that the central governmen<sup>t</sup> could either afford expeditions toward Northwest China—i.e., the hinterland of steppes—and maintain control over the nomads there, or at least keep an equilibrium of military power against nomadic tribes/regimes along the borderline. This is why nearly all hot spots of agri-nomadic conflicts were concentrated in the north during such climatic phases. Conversely, cold and dry climate severely affected the economic bases of agricultural empires and nomadic regimes/tribes by reducing grain yields and herding resources, respectively. The weakened power of the central governmen<sup>t</sup> meant that the military initiative held in warm and/or wet periods gave way to passive frontier defense against nomads [40]. Besides, driven by the cooling- and/or arid-induced environmental degradation, nomads had no choice but to migrate southward, and their conflicts with agriculturalists became increasingly intensive, even breaking through the boundaries and massively invading southward [41,42]. Noticeably, after comparing the patterns between Figures 3d and 4d, we observed many more hot spots in Southeast China during cold intervals than in dry periods, which implied that the effect of temperature on war hot spots may have been stronger than that of precipitation.

Second, with respect to rebellion, similar to the case of agri-nomadic conflict, warm and wet climate typically boosted the economy of agricultural empire, so that the governmen<sup>t</sup> was able to initiate military campaigns toward the frontier. Moreover, the economic prosperity possibly resulted in rapid population growth, and the demands for expanding living space and relieving the ever-increasing population pressure would have needed to be met. As a result, accompanied by the outward conquest, mass migrations of agriculturalists also occurred during warm and wet intervals, which may have exacerbated the competitions for resources (especially for land) between the Han Chinese and ethnic minorities. Therefore, the emergence and spread of revolts with separate clusters in remote areas, such as the northwest of the Sichuan Basin, Taiwan, Guangdong, and Guangxi, could be interpreted. In comparison, the rebellion hot spots during cold and dry stages shifted inward and widely covered the middle–lower reaches of the Yangtze River, Yangtze River Delta, and the border among Guizhou, Hunan, and Guangxi, but still manifested in the Central Plain and part of the North China Plain, even extending westward to eastern Gansu, as the agricultural base of the Chinese Empire was heavily undermined and the frontier control over ethnic minorities loosened. Nonetheless, the Central Plain remained the core warring zone regardless of climatic phase, ye<sup>t</sup> rebellions somewhat intensified during cold periods.

#### *4.2. Contributions of Other Possible Factors*

Our investigation into the linkage between climate and war hot spots does not exclude the possible contributions of other non-climatic factors that mediate the relationship. However, these factors are extremely difficult to spatially quantify from historical documents, in which they are often described ambiguously and fragmented in earlier ages. Yet, such limitations do not interfere with our discussion. Clearly, the hot spots (particularly the intensifying ones) of war overlapped with the developed and populous areas in China, e.g., the Guanzhong Basin, the Central–North China Plain, the Jianghuai region, and the Yangtze River Delta. Hence, this spatial consistency could be interpreted as population-pressure-led social contradiction, an important factor that was observed

before [43–47]. Northern China, where the political center of the Chinese Empire was usually located, greatly benefited from warm and/or wet conditions. The rapid economic development and population growth in such an ecologically fragile agricultural–pastoral transition zone sooner or later deteriorated the environment and caused problems like soil erosion, desertification, and salinization, thereby lowering the land-carrying capacity and triggering resource struggles, and even armed conflict, within the Han Chinese peasants/nomads (e.g., during the Sixteen Kingdoms), or between agriculturalists and pastoralists. Likewise, as northern migrants fled nomadic invasions during cold and/or dry intervals, population increased and the economy flourished in southern China, followed by increasing numbers of conflicts and the emergence of the hot spot zone. Nevertheless, the spatial relationship between war and population/migration in ancient China needs further surveys.

As the agricultural basis in the north was damaged by temperature drops (C5–C7, i.e., the LIA) together with long-term drought (i.e., D3), migration also pointed to the southwest during the late imperial epoch. After the migration from Jiangxi to Huguang in the Ming dynasty, the more massive migration in the Qing dynasty (also known as "Huguang fills Sichuan") boosted the population of the entire southwest, since migrants also flooded into the Yunnan–Guizhou Plateau [48]. Overpopulation on these ecologically fragile mountainous karst landforms exhausted the land-carrying capacity and aggravated the tension between the Han and minorities. Consequently, the hot spots of war were concentrated in these regions. Apart from the population stress resulted from mass migration (sometimes launched by the government), governmental policy may have acted as a catalyst for conflict. For instance, throughout the Ming and Qing dynasties, the managemen<sup>t</sup> and exploitation of southwestern China deepened. One momentous measure, the bureaucratization of native o fficers ("Gaitu Guiliu"), aiming at abolishing the hereditary local chieftain system (or "Tusi System") and directly assigning o fficials by the central government, was implemented. This policy worsened the relationships between the central governmen<sup>t</sup> and ethnic minorities and ignited numerous revolts in that region. Hence, for rebellion (some were waged by the Miao, Dong, Zhuang, and Yao people), the concentrations of hot spots in the west of the Sichuan Basin, Guangxi, and the Guizhou Plateau could be ascribed to the strengthened rule of the government.

Another possible factor is pertinent to the geopolitical situation. The N–S regime confrontation, which was qualitatively elucidated in previous studies [22,23,49–52], is one of the most pronounced characteristics in Chinese history. It included two aspects: united empires that occupied most of the land area to the south of the Great Wall (China Proper) versus nomadic tribes/regimes that stayed in the steppes of Inner Asia during warm and wet phases, as well as two or more Han Chinese or Han versus nomadic regimes within China Proper during disintegrated eras and also under cold and dry conditions. Looking at Table 1, Table S2, Figure 3b, and Figure 4b, the hot spots (especially the intensifying ones) in the southeast can be partly explained by the latter aspect. Under the circumstances, battles were dominantly clustered in the vicinity of regime borders, or distributed along natural boundaries, such as the Yellow River (e.g., Northern Wei versus Liu Song during the Southern and Northern dynasties), the Qinling Mountains–Huai River (e.g., Wei versus Shu and Wu in the Three Kingdoms Period and Jin versus Southern Song), or farther south, the Yangtze River (e.g., Sui versus Chen).
