Environmental degradation and the collapse of civilizations

Full text

Atkins, P.J., Simmons, I.G. and Roberts, B.K. (1998) People, Land and Time London:
Hodder Arnold ISBN: 0340677147 and 0470236590
http://www.routledge.com/books/details/9780340677148/
CHAPTER 5. ENVIRONMENTAL DEGRADATION AND THE COLLAPSE OF
CIVILIZATIONS
‘[Environmental] devastation, with all its consequences, becomes particularly
intense among civilized peoples’. Brunhes, J. (1952) Human geography
London: Harrap
INTRODUCTION
This Section follows on from the discussion and the end of Chapter 4. There the
environmental degradation of early Mediterranean civilizations was reviewed. Here we will
argue that the human abuse of nature can lead to severe economic imbalance and possibly
even to social and political collapse.
COLLAPSE
Archaeologists, historians and historical geographers are like the general public; they prefer
to dwell upon success rather than failure. Studies are made of the rise and florescence of the
great ancient civilizations, such as Egypt, Rome and Greece, with scholars devoting careers
and research resources to the minute investigation of art, material culture, polity and
economy. This work certainly has been important for understanding the foundations of our
own modern civilizations which have been built upon the achievements of our distant
forebears. It has also assisted in developing general principles of societal evolution, although
these have been powerfully unidirectional in their explanatory structure. As a result, until
relatively recently much Anglo-Saxon writing has been of the liberal and whiggish tendency,
steeped in an optimistic, onward and upward outlook. Notably there has been less
enthusiasm for research into failure, whether it be into the medium-scale societies which did
not quite make it to international significance, or the fall of the major powers. Perhaps we do
not like to be reminded of the fragility of advanced societies because there are implications
for potential disintegration in our own age.
By collapse we mean a rapid loss of a major part of the common elements of civilization:
social differentiation, central political control, and economic specialization. One theory is
that civilizations are organisms which follow a natural life cycle of birth, maturity, old age
and death. This is most clearly articulated in the writings of Toynbee and Spengler, but such
organic analogies have been popular since the age of the Greek philosophers.
There is also a school of romantic writers on the subject of lost civilizations who feed our
fascination with the mysterious and the exotic. We are all interested to know how
civilizations could have prospered in the unpromising settings of the Iraqi desert or the
jungles of central America. Thus an industry has grown to satisfy the demand for tourism
among ruined cities, and for books decoding the true meaning of Stonehenge or the
Pyramids.

A third intellectual drift has been the delight taken by popular writers and some scholars of
collapse in shopping at the supermarket of causalism. Simple, penetrating explanations are
obviously preferable to the inconvenience of grasping complex reality, and therefore we are
tempted by the list of causes which have been advanced to account for the historical puzzles
such as the fall of the Roman empire. This week’s special offer may be the lead poisoning
induced by a piped water supply. Next week it may be an equally bizarre theory which has
its own band of advocates. But explanations based upon single causal factors are both
inadequate and misleading. The systems view, in which change is seen as the result of the
interacting elements of a whole, is far more convincing because it can cope better with
understanding the highly complex relationships which characterise all advanced civilizations.
According to Tainter (1988) it is this very complexity which may lie at the heart of the
explanation of collapse. After reviewing a number of civilizations widely separated in time
and space he concludes that, for pre-industrial economies at least, complexity has
disadvantages beyond a certain point (Figure 5.1). In fact the likelihood of a collapse is
increased when to add further investments of capital and labour to bolster an existing power
structure would bring a poor return. Any political or economic problems may then become
more difficult to solve because cheap and accessible resources are not available by way of a
solution.
The classic example of this is the Roman empire. It expanded by a series of conquests, each
of which brought a windfall for the treasury in terms of confiscations, slaves and taxes. Once
all of the fattest milch cows were within the empire, however, the costs of administration
became a burden which could be maintained only by the extraction of an excessive surplus
from the peasantry. A debased currency and weakened agricultural base eventually took their
toll and barbarian incursions became difficult to repulse.
COLLAPSE AND THE ENVIRONMENT
Having pointed to the danger of relying upon limited causalist reasoning, we are happy to

acknowledge that in pre-industrial societies environmental considerations were exceptionally
important because their agricultural economies were reliant upon the vagaries of the elements
for stable food supplies. Particularly where population pressed upon resources, the gearing
of agricultural technology and human muscle power to climatic, soil and topographical
conditions was finely balanced and any disruption could provoke a crisis of subsistence, with
political consequences. For many among the oppressed working population the eventual
collapse may have actually seen an improvement in their standard of living. It is therefore
not unreasonable to ask ‘a collapse for whom?’.
Some economies survived for hundreds or thousands of years in balance with their
environment. This was due to a number of factors, the most important of which is a keen
appreciation, on the part of apparently primitive people, of the dangers of over-exploiting
their resources.
The examples of societies which have damaged their environments are unfortunately more
numerous. Making a direct connexion between such environmental degradation and the
collapse of civilizations is dangerous because we may stray on to the ground of the
environmental determinists who were convinced that human affairs were subordinate to
grosser physical forces such as climate. Although the environment was crucial to the lives of
pre-industrial peoples, we must be careful not to make it the principal god in the pantheon of
explanation without very careful consideration.

CASE STUDY: THE MAYA
The Maya formed a long-lived and highly developed civilization in central America, from
about 3000 B.C. to A.D. 14th century. They were the only native American civilization to
have a fully literate culture and they knew mathematics and made astronomical observations.
They calculated the year to be 365.2420 days long, which differs from the true figure of
365.2422 days only in the fourth decimal place. Their art and architecture were rich and
elaborate, expressed characteristically through a monument cult. They carved their history
and calendar in hieroglyphs on stelae and built monuments of a size and complexity rivalling
any other New World group. Their mausoleums were of stepped pyramid shape (Figure 5.2),
made of limestone blocks, and were supplemented by temples, ritual ball courts, plazas and
palaces.
There is a certain mystery about the Maya. Until relatively recently many aspects of their
society were unclear. Even their material culture was little known because their abandoned
cities were overgrown and inaccessible. Most mysterious of all was the sudden collapse of
the civilization from the ninth century A.D. which, within 100-150 years saw a reduction in
population by at least 70 per cent in the southern lowlands and the desertion and decay of
many of their cities. Greater clarity has come in the last few decades as archaeologists have
investigated a wide range of sites, by digging, field walking, and using technologies such as
remote sensing to map features on a regional scale.
Figure 5.3 shows the region occupied by the Maya. It covers western Honduras, Belize,
Guatemala and four states (Yucatán, Quintana Roo, Campeche and Chiapas) in Mexico’s
Yucatán peninsula. We can distinguish between the mountainous areas in the south and the
lower relief, mostly under 400 metres, to the north. These lowlands are comprised of two
distinct zones: the flatter areas of the northern Yucatán and the rolling topography of the
Petén. Our discussion will focus on the southern lowlands in Guatemala. It was here that
population density was greatest.
This area has (and as far as we know also had in prehistory) a tropical climate with a lengthy

dry season, ranging from luxurious forest vegetation in the south to xerophytic scrub in the
north. Mean annual rainfall varies from 500 mm in the north west to 3,000 mm in the south,
but there are significant annual variations (Turner 1983). The northern part of the peninsula
exhibits typical karst limestone scenery, with little surface water and occasional solutional
depressions such as large poljes (locally called bajos when they contain seasonal wetlands)
and smaller dolinas (here known as aguadas).
This environmental context is interesting. The lowland region is not especially attractive,
with a lesser diversity of ecological niches than surrounding mountainous and coastal
regions. This would have reduced the probability of local trade. The large majority (88 per
cent) of the soils are of moderate or high fertility, particularly the brown/black mollisols
(rendzinas), a much higher proportion than in most of the tropics, but a third (37 per cent) are
classified as easily erodible. In Fedick and Ford’s (1990) sample areas most (89 per cent) of
Maya structures were found in association with these soils, avoiding the acidic and more
difficult-to-work vertisols.
Table 5.1 Periods of Maya development
Period
Date
Stage of development
Early Preclassic
1800-1000 B.C.
Early agriculture.
Middle Preclassic
1000-300 B.C.
Population growing, uniform
pottery
Late Preclassic
300 B.C.-A.D. 250
Writing and calendar,
chiefdoms.
Early Classic
A.D. 250-600
Monuments, towns.
Hiatus
A.D. 534-593
Construction ceases.
Late Classic
A.D. 600-800
High peak of Maya civilization.
Age of expansion
A.D. 600-700
More sites erect monuments.
Age of interconnexion
A.D. 700-750
Greater contact between sites,
spread of lunar cult.
Acceleration of monument
construction.
Incipient collapse
A.D. 750-790
Breakdown of alliances and
decentralization of power.
Fewer sites erect monuments.
Decay
A.D. 790-830
Cultural decline. Invasion of
Seibal.
Terminal Classic
A.D. 830-925
Population collapse in south,
construction ceases, calendar
abandoned. Puuc sites flourish.
Postclassic
A.D. 930-1200
Maya civilization continues in
north. Toltec invasion.
Source: after Coe (1993)
The most important period from our point of view is the Classic period, from roughly A.D.
250-930 (Table 5.1). The civilization had evolved into a number of independent city states,
each with massive temples, palaces, paved roads, ball courts and public plazas, though often
with regional variations on the cultural theme. Until recently these were thought to be

merely ceremonial centres at which the surrounding population would occasionally assemble,
but we have now found signs of relatively mature urban settlements with permanent residents
and extensive residential suburbs. The economy was sufficiently advanced to support a
politico-religious hierarchy and a small army of non-agricultural specialist workers. There
were flint and obsidian workers, potters, woodworkers, dentists, stoneworkers, monument
carvers, textile weavers, human carriers, feather workers, leather workers, musicians,
manuscript painters, merchants, basket makers, bark cloth makers, and many others.
The largest city, Tikal, at its peak probably had an urban population of over 60,000, with a
further 30,000 rural fringe dwellers. It covered sixteen square kilometres and contained
3,000 structures, from temple pyramids to thatched huts. The majority of the other people
were rurally based, living in hamlets of 5-12 houses, with a minor centre for every 50-100
houses. The total lowland Maya population is difficult to estimate, but 8-10 million people at
the apogee seems possible (Coe 1993).
The Maya food supply had originally been based on shifting agriculture. The bulk of their
carbohydrates came from maize, and they also grew or collected beans, squash, avocado,
cacao, and root crops such as malanga (Xanthosoma violaceum), and tree crops like sapodilla
(Achras zapota), nance (Byrsonima crassifolia) and guava (Psidium guajava). Ramón
(breadnut) trees (Brosimum alicastrum) were also plentiful and potentially an important
source of protein. There is no evidence that they were cultivated, but they were a useful
famine food, producing fruit in the driest of seasons. Altogether 150 economic plants were
either grown or collected, many of them encouraged by the management of vegetation in

what amounted to an artificial rainforest (Figure 5.4).
The analysis of carbon and nitrogen isotopes and mineral content in bones of a known age
has allowed the reconstruction of the Maya diet (White and Schwarcz, 1989). Maize
accounted for about 50 per cent of nutrition in the pre-Classic, falling to about one third in
the Terminal Classic due to a diversification of the diet with other plant foods. This may
have been due to success in broadening the agricultural base or the desperation of a
malnourished population eating famine foods. After the collapse the dietary rôle of maize
rose again to 70 per cent, suggesting a reversion to a narrowly based subsistence agricultural
economy, possibly due to a decline in trade.
As population grew, an intensification of agriculture was necessary, as early as the Late
Preclassic period in the central and northern Petén. Several possible adaptations have been
suggested, all of which would have had landscape implications:
1. Home or kitchen gardens (solares) which could be cultivated intensively close to
settlements because labour was at hand for weeding and mulching. Night soil and household
waste would no doubt have been used as fertilizers. Maya urban structure left sufficient
room for this sort of activity, with 0.16-0.20 ha available in central Tikal, for instance.
These infields produced fruits, nuts and vegetables, with the staple maize crop being
cultivated in outfields, perhaps by less intensive swidden (milpa) cultivation. Ethnohistorical
research shows that gardening has been common among the Maya in recent times although it
is very difficult to know if this is the result of continuity from the prehistoric.
2. Wetland agriculture was an intensive adaptation, in seasonally inundated bajos, on lake
margins, and in lowland river valleys. It was attractive because it offered the possibility of
long-term cultivation and removed the tree-felling effort involved in slash and burn (Pohl,
Bloom & Pope 1990). Flood plain soils would have been fertile and relatively weed-free by
comparison with the uplands. Firstly, raised fields (Figure 5.5) were constructed between a
network of canals or ditches. Their surface was raised by the addition of material excavated
from the water courses. Secondly, channelized fields are formed between long trenches sunk
into the margins of wetlands to enhance drainage.
Turner & Harrison (1983) argue that wetland agriculture is likely to have started after upland
milpa cultivation had proved insufficient as a food source. The effort of digging canals and
moving earth would simply not have been worthwhile unless population pressure elsewhere
encouraged a search for intensive alternatives. They also propose an evolutionary model in
which the margins of wetlands would have first been used with (marceño) cultivation of
rapidly maturing crops as the water receded during the dry season. Only subsequently would
intensification have seemed justified. It is thought that channelized fields probably pre-dated
raised fields.
Wetland agriculture is thought to have become intensive in the Late Pre-Classic. Farmers
were then unwilling to write off the significant labour input and came to depend upon élite
groups for the protection of that investment (Pohl 1990). As little as 0.5 ha was sufficient to
support a family. The canals were used for fish farming and their silt for fertilizer. The high
water table kept the root zone of the plants moist.

Satellite radar remote sensing has shown these areas of raised fields to have been extensive
and there is a correlation between swamps and the large centres of population (Tainter 1988).
Water availability must have been an issue because of the annual season of drought and due
to a karstic landscape lacking surface water in about half of the lowlands. Recent work has
indicated that the Maya were practising sophisticated water harvesting and storage
methodologies. In the immediate vicinity of Tikal, for instance, 75 reservoirs have been
found, including six central precinct tanks with a total capacity of 100-250,000 cubic metres.
The expansion of the cultivated area into the dry interior of the Petén was a risk and much of
the Yucatán peninsula is limestone, with little surface water. The Maya built canals, dams,
underground cisterns waterproofed with a plaster lining (chultúnes), wells, and they modified
naturally occurring limestone swallow holes (cenotes) to provide a perennial supply. The
water was used for irrigation, drinking, and for water transport.
3. Hillside terracing was constructed to prevent the erosion of thin soils, to minimize the
leaching of soil nutrients and to control soil moisture (see Section B3.4). Level planting
platforms for ease of cultivation and irrigation are not common, however, the terraced fields
being typically dry and uneven.
Two types of terraces have been identified from detailed field work in the Río Bec region:
linear sloping, dry field terraces and channel bottom check dams or silt traps (Turner 1983).
The terrace walls are either broadbase or stone-slab embankments. These terraces are found
on slopes varying from 4 to 47.
Turner (1983) has calculated for the Río Bec area that a total of up to 16.5 million days of
construction work would have been necessary to establish the terraces in that region. This
means 400 people more or less full-time over a 400 year period. Clearly a high level of
commitment and planning was required, perhaps in response to population pressure and the
dangers of environmental degradation. Towards the end of the Classic period it is

conceivable that demographic instability and reduced work efficiency due to malnutrition
may have made it difficult to maintain such a labour-intensive system of terraces.
Social differentiation and social control.
In the early and mid Classic period increased socio-political complexity was a solution to
shortages for some, with the emergence of élite groups who had privileged access to
resources. The outward sign of such differentiation was public architecture on a scale so
massive that slave or corvée labour must have been involved. The transition from egalitarian
to stratified societies may have been facilitated in areas of settled, wetland agriculture.
Swidden farmers would have been more difficult to control and exploit.
Before about A.D. 400 urban centres seem to have been evenly spaced, with an approximate
equivalence of status. Tikal seems to have emerged as the dominant central place with a
subservient hierarchy of settlements on perhaps two levels. In the hiatus period of A.D. 534-
593 there was a decentralization of the erection of stelae to the peripheries of the Petén,
perhaps indicating a lessening of central power, and in the subsequent Late Classic a number
of major centres were visible, each with their own secondary centres. Tikal was clearly no
longer dominant.
The collapse.
The degree of intensification was remarkable. At the peak of its powers, in about A.D. 800,
the Mayan civilization supported the densest population of any pre-industrial society in
history. In the Rosario Valley, De Montmollin (1989) estimates that there were 400 people
per square kilometre, and a conservative calculation puts the average for the Petén at 200 per
square kilometre (Culbert 1988). Compare this with the population densities of countries
today: China 120, India 264, United Kingdom 231, United States 27. Actually, the Maya
population had probably ceased to rise sometime earlier on, perhaps about A.D. 650,
although the date would have varied from area to area.
There were limits. Short-fallow milpa could have supported at most 30-60 people per square
kilometre, so the bulk of the population relied upon intensive, settled agriculture and inter-
regional trade. By the Late Preclassic much of the central Petén had been deforested and
population densities were such that competition for resources must have occurred. It cannot
be coincidental that evidence of conflict (fortifications, carvings showing prisoners) is
common at this time. Warfare meant the ebb and flow of territories and alliances, with the
capture of high status individuals for sacrificial purposes as one objective.
These was some trade with the hill regions: pottery, cacao, honey and salt from the lowlands,
in exchange for jade, obsidian, quetzal feathers, and granite (for grinding stones). Scope for
short distance trade in the lowlands was limited, however, because of the lack of topographic
and ecological diversity. In times of subsistence crisis, such as during a prolonged drought,
to raid neighbouring groups may have been the most logical short-term solution.
This militarisation seems to have been entwined with population growth and concentration.
It was in the defensive interests of each state to have its people in nucleated settlements
rather than dispersed and therefore vulnerable. The archaeological evidence reveals several

such periods of aggregation, which presumably must have been times when productive food
resources were not fully exploited. In addition we think that population increase may have
been encouraged to guarantee sufficient numbers of able-bodied soldiers for deterrence and
enough labourers for the construction of public works such as monuments. In turn the
architectural display was probably a means of indicating power and wealth to friend and foe
alike in a seemingly endless competitive spiral.
This polity and economy was essentially non-sustainable. The necessary superstructure of
authoritarian control was inherently fragile. The system never reached a stable state but kept
on intensifying, with a larger and larger population, until eventually the system collapsed
under an unsupportable weight.
The evidence of environmental stress is conclusive. Abrams and Rue (1988) have calculated
that pine trees were fully cleared from the Copan area by A.D. 800. Pollen analysis confirms
the absence of deciduous forest and the decline of montane pine forest, with a recovery of
trees only after A.D. 1250. In turn this must have affected the habitats of game species such
as the white-tailed deer, a valuable source of protein.
Immediately prior to the collapse it seems that social complexity and monumental
architecture grew significantly, yet there were signs that all was not well. Population growth
slowed and skeletal remains from this period often show evidence of disease as a result of
specific nutritional deficiencies, such as shortages of vitamin C, which caused scurvy, and
iron leading to anaemia. The Late Classic males were an average of 7 cm shorter than their
Early Classic forebears, and may therefore have been undernourished when children or at
least were short of protein. A calculation of the age of death at burial shows a reduced life
expectancy in the Late Classic. Parasites were probably common and may have contributed
to a reduction in work efficiency.

The collapse took place around A.D. 800-830. During the Terminal Classic population
declined by two-thirds, at Tikal to as little as 1,000, and shrank even further in the Postclassic
(Figure 5.6). Most of the elements of the civilized society were lost: administrative and
residential structures, erection and refurbishment of temples (Figure 5.7), tomb burials, stela
construction, the manufacture of luxury items, writing and the maintenance of the unique and
elaborate calendar (Tainter 1988). People were living in the remains of the former glory,
throwing rubbish in previously restricted areas. They did not abandon their religious
ceremonies but lost the specialized knowledge of the rules of stela erection.
The last inscribed date at Copan was A.D. 800 and at Tikal A.D. 869. The collapse was
therefore not sudden but it was extensive and fundamental. A few cities, such as Altar de
Sacrificios and Seibal, continued to grow, with dated monuments as late as A.D. 889, and in
the northern Yucatán region of Puuc, Chichén Itzá, Mayapán and other cities flourished for a
few centuries more, as did some sites in Belize with access to marine resources.
The precise reason for the collapse is unknown (Table 5.6). Soil exhaustion may have been a
contributory factor or perhaps soil erosion. There is evidence of the latter in the silting of
lakes in the central Petén. A series of seasons when drought reduced crop yields might have
tipped the balance or the introduction of maize mosaic virus. Alternatively the dominant
priestly élite may have been overthrown in a revolution by their oppressed subjects or an
invasion may have come from the north or the Gulf coast. Whatever the immediate cause,
the collapse was undoubtedly exacerbated by the imbalance between people and the land,
and the inherently slow natural regeneration.

Table 5.6: Single factor explanations of the collapse
1. Resource depletion due to soil exhaustion and erosion.
2. Fields choked by weeds.
3. Long-term climatic change.
4. Catastrophic events such as earthquakes or hurricanes.
5. Disease.
6. Social dysfunction leading to revolution against élite.
7. Intersite warfare, invasion from north or west.
Sources: Culbert (1973), Tainter (1988)
CONCLUSION
We must take care not to ascribe monocausal significance to the environmental problem
facing the Maya. It seems likely that unstable political structures, due to internecine strife
within and between élite groups, were contributory, and De Montmollin (1989) reminds us
that no other Mesoamerican polity lasted for longer than a few hundred years. In other words
they had a predisposition to collapse. Nevertheless this is the most dramatic example in
history of how a civilization that over-exploits its environment for short-term gain will
eventually pay the price.
FURTHER READING AND REFERENCES
The mysterious Maya have attracted many scholars, especially American archaeologists.
Their vast outpouring of data and interpretations are best summarised in Coe (1993). The
collapse is dealt with by Culbert (1973, 1988) and Tainter (1988).
Abrams, E.M. & Rue, D.J. (1988) The causes and consequences of deforestation among the
prehistoric Maya, Human Ecology 16, 377-395
Coe, M.D. (1993) The Maya 5th ed. London: Thames & Hudson
Culbert, T.P. (1973) The Maya downfall at Tikal, pp 63-92 in Culbert, T.P. (Ed.) The classic
Maya collapse Albuquerque: University of New Mexico Press
Culbert, T.P. (1988) The collapse of Classic Maya civilization, pp 69-101 in Yoffee, N. &
Cowgill, G.L. (Eds) The collapse of ancient states and civilizations Tucson:
University of Arizona Press
De Montmollin, O. (1989) The archaeology of political structure: settlement analysis in a
Classic Maya polity Cambridge: Cambridge University Press
Driever, S.L. & Hoy, D.R. (1984) Vegetation productivity and the potential population of the
Classic Maya, Singapore Journal of Tropical Geography 5, 140-153
Fedick, S.L. & Ford, A. (1990) The prehistoric agricultural landscape of the central Maya
lowlands: an examination of local variability in a regional context, World
Archaeology 22, 18-33
Pohl, M.D. (Ed.)(1990) Ancient Maya wetland agriculture: excavations on Albion Island,
northern Belize Boulder: Westview
Pohl, M.D., Bloom, P.R. & Pope, K.O. (1990) Interpretation of wetland farming in northern
Belize: excavations at San Antonio Rio Hondo, pp 187-254 in Pohl, M.D. (Ed.)
Ancient Maya wetland agriculture: excavations on Albion Island, northern Belize

Boulder: Westview
Tainter, J.A. (1988) The collapse of complex societies Cambridge: Cambridge University
Press
Turner, B.L. (1983) Once beneath the forest: prehistoric terracing in the Río Bec region of
the Maya lowlands Boulder: Westview
Turner, B.L. & Harrison, P.D. (1981) Prehistoric raised field agriculture in the Maya
lowlands, Science 213, 399-405
White, C.D. & Schwarcz (1989) Ancient Maya diet: as inferred from isotopic and
elemental analysis of human bone, Journal of Archaeological Science 16, 451-74