Regional Patterns of Forestry in Four Maps

Since I started this project nearly a decade ago I’ve been trying to develop a comprehensive picture of the regional orientations of forestry institutions. In essence, as I reviewed the most accessible evidence, it became clear to me that forestry simply worked very differently in North China than in South China. As I delved deeper, it became further apparent that there were at least four distinct regional patterns of state interventions into forestry. Furthermore, these interventions showed astonishing persistence over time. Despite substantial shifts in government at the dynastic level, and some specific shifts at the local or level, the four basic regional patterns of forest interventions remained remarkably consistent for at least six centuries (c. 1000–1600), probably more like a millennium (c. 780–1800 or later). Some regional patterns, especially in the north, showed continuities for more than two millennia.

Here is the schematic picture of forest interventions I drew in summer 2015:

The Northwest I now know was an area that both employed forests defensively and was a major area of extraction. The awkward balance between these activities often meant that military auxiliaries were employed as loggers. The balance probably also shifted from more emphasis on extraction up to the Northern Song to more emphasis on defense by the mid-Ming, although the specifics are far from clear.

The West and Southwest was long an area targeted for the extraction of large timber. Timber officials were posted to Sichuan in the Han, and it remained the main source of timber for imperial construction throughout the Ming and up to about 1720. As Sichuan began to run out of large timber, the frontier increasingly shifted further south into northern Guizhou. In the meantime, Hunan and eastern Guizhou were — since at least the Song — sites of timber markets governed by ethnic division of labor.

But the final question is what was going on in the South and Southeast? As of two years ago, I had a rough sense that this was an area with extensive private forestry (rather than state forestry). Based on early 20th century legal records, I extrapolated backwards a very rough sense that private forests in the region were claimed based on four different but overlapping sources of evidence: formal land registration (especially in Huizhou), private deeds (throughout much of Jiangnan and Fujian), fengshui (especially in the Hakka heartland) and settler land claims (especially in Hunan).

Here is the rough schematic map I drew based on this evidence, again in Summer 2015:

Since I drew these two maps, I have been able to add substantial data on patterns of forest registration to the anecdotes and early 20th century evidence. Thanks to two sessions at the Max Planck Institute for the History of Science in Berlin, and a yearlong postdoc at the Yale Program in Agrarian Studies, I had time and access to materials to flesh out these impressions. Thankfully, these data — drawn largely from mid-to-late Ming cadastral records and gazetteers — largely confirm the rough picture I previously developed.

First, patterns of forest registration were strong and universal in Jiangnan (especially Zhejiang, the portion of the Southern Metro south of the Yangzi, and the northern two thirds of Jiangxi). Forest registration was present but weak or irregular in Fujian and the Hakka region of Fujian-Guandong-Jiangxi. Outside of these two regions, there were essentially no registered private forests:

In the map above, dark green represents areas where shan 山 emerged as the standard category for forest, and was used in all counties in the prefecture. I take this as an indication that registration of privately-owned forests was the norm. As you can see, this clearly highlights Zhejiang and most of Jiangxi, as well as most of the parts of the Southern Metro south of the Yangzi. Although there are some areas with data missing. This was definitively the region where forests emerged as private property the earliest (around the twelfth century) and were registered as private property the most consistently.

Light green represents areas where shan emerged as the standard category for forest, but was only used in some counties (generally less than half of the counties in a prefecture). I take this as an indication of incomplete adoption of the practice of registering private forests with the state. Either forests were not treated as private property, or they were claimed on the basis of other forms of evidence (deeds, settler’s rights, fengshui, etc) without registration. This roughly coincides with the Hakka regions of Fujian, Guangdong and Jiangxi, plus one part of Guangxi. I suspect these were areas where the practice of forest registration spread in the Ming (probably in the sixteenth century).

Blue denotes areas where forests were registered under a profusion of different categories, including shan as well as non-standard categories like yuanlin 園林. Dark blue means these categories were used throughout the prefecture, light blue that they were only used in some counties (although generally more than half of the prefecture). This coincides with northern and coastal Fujian, an area surveyed in the Song but that missed out on the reforms conducted in Jiangnan in the Yuan (to be explored in a future post). This explains why it had strong patterns of registration for private forests, but did not standardize registration categories.

Orange indicates areas where I surveyed gazetteers and found no forests (shan or alternatives) listed in landholding figures. Dark orange indicates no forest landholdings listed at prefectural level, light orange indicates no forest landholdings listed at provincial level (where I did not survey lower-level gazetteers). As this map makes apparent, outside of the three regions listed above there was no institutionalized mechanism for registering forests as private property. In much of the North I suspect that forests were maintained as commons. In the Southwest, there was enough natural growth that there was no impetus to claim forests as private property when wood could be cut for free.

A final map further breaks down the sub-regional patterns of forest ownership within the region of extensive forest registration:

In this map, shades of green indicate the proportion of registered acreage in the category of forest (shan):

Lightest green (Nan’an, Ganzhou, etc.): 0–5%

Light green(Ji’an, Nanchang, etc.): 5–10% ,

Medium-light green (Linjiang, Taizhou): 10–20% ,

Medium-dark green (Raozhou, Guanxin, etc): 20–30%

Dark green (Yuanzhou, Ruizhou, etc) 30–40%

Darkest green (Yanzhou): more than 40%

Orange indicates data is missing from the source.

As these data make clear, even within the region where forests were regularly registered as private property there are three or four distinct patterns.

From Lake Poyang to the coast there is a uniform belt where forests made up at least 20% of registered landholdings. Incidentally, while this region overlaps imperfectly with Ming provinces, it coincides almost exactly with the Song circuits of Zhejiang (Zhexi and Zhedong) and Jiangnan Dong. As I will argue in a later post, this is not an accident. This was the region where land surveys were implemented most effectively in the Southern Song, and reformed in the mid-Yuan. It was the original region where private forest plantation developed as a core part of the economy, and it remained an important tree farming area up to the end of the Ming (and indeed to the present).

The second distinct region is western Jiangxi (Ruizhou, Yuanzhou, and Linjiang, and parts of northern Ji’an that are obfuscated in the prefecture-level statistics). The specific history of its pattern of forest ownership are less clear. I suspect that surveys were conducted here in the Southern Song but did not stick, and the region emerged as an important timber-planting region in the mid-Ming, at which point it adopted land registration norms from points further east.

The third region encompasses much of the rest of Jiangxi. Here, there are patterns of forest registration, but production forests were simply not a major part of the local economies.

Finally in southern Jiangxi and western Fujian forest registration rates are very low. In combination with Map 3, we know that this is because the paradigm of private forest ownership and registration was only incompletely adopted by the late Ming. These were areas where wood was still largely cut from natural growth, and where forests were later claimed largely on the basis of settlers rights and fengshui, not deeds or official land records.

I’m not sure how transparent any of this is to someone who has not spent the last few years buried in cadastral data, but to recapitulate my introductory remarks, I think it demonstrates two very important things.

First, the cadastral data support my impressionistic picture that forest institutions were highly regional, and that the Chinese empires integrated several very different regimes of forestry within their territory. Even within each region, there are multiple sub-regional trends. For example, South China was the only region to develop a clear tradition of private forest ownership. Within this region, we can further identify at least three subregions: Jiangnan, which had the strongest tradition of private forest registration; Fujian, which had a distinct and irregular set of private forest ownership practices; and the Hakka lands, which featured highly erratic forest registration. Even within Jiangnan we can identify a difference between the eastern areas with relatively stronger patterns of forest registration, and the western and southern areas with lower registration rates.

Second I will argue that there is a diachronic picture embedded in these synchronic maps. Private forest ownership and forest registration were strongest in Jiangnan because these institutions emerged earliest in Jiangnan (in the 1140s), and were continually reinforced by local practice over five centuries. They were idiosyncratic in Fujian where forests were surveyed early (also the 1140s) but registration was maintained separate from the imperial administration and only erratically. They were weakest in the Hakka region because they were brought there late (probably in the mid-Ming) and without the backing of a regionally- or imperially-organized campaign. I hope to give more evidence of this in a future post.

Together these two conclusions suggest that we cannot think of imperial land administration as something uniform. Instead Chinese imperial land oversight integrated regional systems in ways that were largely agnostic to lower-level institutions.

Related posts:

To become a digital nomad, you need to expect the unexpected. If the term nomad doesn’t tell you anything, you should know that nothing is certain in the digital world.

String theory is a theoretical framework in physics that attempts to reconcile quantum mechanics and general relativity, two fundamental theories that describe the behavior of the universe at very…

All of your blog posts are undoubtedly amazing. You write, post, and interact based on that one blog article. However, are you actually getting the most out of each blog post that you share with…