Late Preclassic Abandonment in the Maya Area: Scale and Causality
During the Late Preclassic Period, particularly from A.D. 200-250, widespread abandonment and/or hiatus in construction occurs across sites in the Maya area. Regions most affected were Chiapas, El Petén and the Yucatán peninsula. A preliminary discussion of the causal factors behind site abandonment is given. In reviewing existing data, it appears that a combination of drought and agricultural intensification with large-scale deforestation were the primary causes for site abandonment. A significant population decline occurs with migration of peoples to more sustainable areas. The construction of hydraulic and defence systems takes place for the first time during this period. At certain sites a hiatus in elite construction occurs which may be an effect of this increase in stress or perhaps may be linked to instability within the institution of divine kingship. Unfortunately, we know little about socio-political complexity in the Late Preclassic Period to be able to clarify this. A tentative model for Late Preclassic abandonment is given, which can be used as a basis for further work on this issue and perhaps can be applied to other cultural areas that have experienced widespread abandonment.
There is an abundance of evidence for widespread abandonment of sites towards the end of the Late Preclassic and the incipient years of the Early Classic Period. However, little research has been carried out on this phenomenon, interest tends to be limited to the Classic Maya Collapse (Culbert 1977, Yoffee and Cowgill 1988, Demarest 1997, Webster 2002, Andrews, Andrews and Robles Castellanos 2003). Yet the study of this area is crucial for several reasons. Firstly, we can gain a deeper insight into a lesser known, but important period of Maya civilization. Secondly, Preclassic abandonment supports the idea of a dynamic civilization that is constantly rising and declining. Thus the Classic Maya collapse was not an unprecedented event, rather a more intense and devastating episode in a history of mini-collapses, abandonments and hiatuses. Lastly, causes for site abandonment during this time period may aid in the understanding of why there was a widespread collapse of Maya sites during the Terminal Classic.
The abandonment of a site can often be a result of a larger occurrence; the collapse of the society. Collapse is generally defined as a society displaying a rapid, significant loss of an established level of sociopolitical complexity (Tainter 1988:4). However, this is very difficult to identify for the Late Preclassic Period, and the Preclassic in general, as we know little about the socio-political complexity of the few sites that have been excavated. We do know that most of the manifestations of complex society during the Late Preclassic had been established, for example, pyramid building and causeways radiating from large sites to smaller subservient sites (El Mirador: Hansen 2001), hydraulic systems (Edzna: Matheny 1976) and hieroglyphic writing (Pyramide de Las Pinturas, San Bartolo). Yet these are a few examples, we do not have a corpus of material regarding kingship and social stratification to work from. For these reasons I will be focusing on site abandonment only.
Abandonment refers to the cessation of architectural buildings, sculpture and trade activities and departure of the vast majority of the population from the site (Manzanilla 2003: 91). Abandonment is an integral part of long-term social processes, as old centres decline and are abandoned, new powers take their place. Site abandonments are not always permanent; we do see re-occupation in many sites by the Late Classic Period, for example at Kaminaljuyu (Michels 1979), and Dzibilchaltun (Andrews V 1981). The rate of abandonment varies depending on the causal factor; a volcanic eruption or earthquake would result in a rapid abandonment, whereas climate change or soil erosion may result in a more gradual abandonment. In the latter case it is difficult to tell if the process took a number of years or decades.
Site abandonments and hiatuses occur throughout the Maya area during the Late Preclassic Period (Figure 1 and Table 1, see also Hansen for an exhaustive list 1990:216-221), particularly in the Yucatán peninsula, the Petén, and Chiapas. The general abandonment period runs from A.D. 50 to A.D. 250 in the Maya area, the peak being between A.D. 200-250 (Figure 2). Sites are abandoned throughout the Yucatán from A.D. 150-250 (Ball 1978:129) and the Mirador Basin is completely abandoned at the same time (Dahlin 1983). Eastern Chiapas sites collapse in A.D. 100, the region not being noted for Early Classic remains (Dahlin, Quizar and Dahlin 1987:373). In addition, Shook and Proskouriakoff (1956) in their survey of settlement patterns in the South Eastern Highlands note a virtual abandonment of much of the area ca. A.D. 260 +/- 85 until the Late Classic Period.
The general explanations for abandonment are divided into climactic/environmental and human causes. The main climatic and environmental causes are drought, floods, earthquakes, hurricanes and volcanic eruptions. Human causes usually given are intensification of agriculture and exhaustion of soils, large population increase, disease, warfare whether internal revolt or foreign invasion, socio-political factors, and loss of trade and economics (Webster 2002:219). Obviously, the factors mentioned could interplay with each other, for example, a significant increase in population could lead to intensification of agriculture and subsequent disease and warfare.
The Late Preclassic was a tumultuous period where we see climate change, a drastic increase in population, the development of elite ideology and the concept of divine kingship, and the development of cities and trade networks. In reviewing the causal data, the most compelling argument for the disruption that occurred is climate change, namely drought. However human agency played a major role. It was precisely the instabilities within the societies; agricultural intensification, large-scale deforestation and lack of water control that made centres vulnerable to the increased stress, resulting in site abandonment. Each causal factor will be reviewed and evidence will be given for why the trigger of drought combined with the silver bullet of agricultural intensification hits sites hard.
Climatic and Environmental Causes
The impact of the eruption of Ilopango Caldera volcano in central El Salvador was felt throughout El Salvador and the adjoining areas of Guatemala and Honduras. The eruption was initially dated to ca. A.D. 260 +/- 114 years and so was given by some archaeologists as the cause for widespread abandonment of Maya sites (Sheets 1983, Sharer 1994:137, Dahlin 1979:28). However AMS dating supports a date of a century and a half later, ca. A.D. 429 (Dull, Southon and Sheets 2001), and so can be ruled out as a factor in Late Preclassic abandonment.
El Chichón volcano, located in Chiapas, was suggested by Gill to have erupted around the time of Late Preclassic abandonment (2000:378). Tree-ring dates from Sierra Nevada and evidence from historical records do indicate that between A.D. 150–250 there was increased volcanic activity and glacier advance (Scuderi 1990:84). Yet we have no confirmed evidence of a specific eruption during this period in the Maya area.
Our evidence for volcanic eruptions actually comes from outside the Maya area. Excavations at the site of La Joya located 3 km southwest of the central plaza at Matacapan, indicate that two volcanic events affected the site, the first being dated to the Late Formative Period. At Bezuapan, a neighbouring site, ash fall was present dated to the same period (Santley 2003:169). Consequently, there is a total population drop at the beginning of the Early Classic in the Tuxtlas. Settlement shift is reflected both by high levels of abandonment of sites and comparatively equal levels of site foundings (Santley 2003:171). In Central Mexico the Popocatepetl volcano erupted after A.D. 100 devastating the site of Cuicuilco, resulting in population decline and shift to the emerging centre of Teotihuacan (Santley 2003:179). Yet there is no evidence that either of these eruptions affected the Maya area and thus volcanic activity can be ruled out as a cause in Late Preclassic abandonment.
Hurricanes and Earthquakes
Evidence for devastating hurricanes is even more lacking for the Maya area, and in itself lacks credibility. From my own experience of living in New Orleans, hurricanes tend to be short and localized. They were and still are today prevalent along the Yucatán coast and can destroy houses and crops, but do not appear to cause extensive damage over a large area resulting in abandonment and resettlement of peoples. Earthquakes have similar effects, although minor tremors must have occurred, evidence is lacking for major structural damage occurring to Late Preclassic sites.
Floods caused by hurricanes or by a drought related rise in sea levels could ravage an area. Bruce Dahlin (personal communication 2003) suspects that the latter was the case in northern Yucatán. There is evidence for particularly wet events during the Late Preclassic Period in 254 B.C, 98 B.C and A.D. 60 in sediments from Lake Punta Laguna (Curtis, Hodell and Brenner 1996:45). Moreover, A.D. 250 in the Yucatán peninsula was marked by a shift to a relatively drier climate. A resulting rise in sea levels on the coast would produce increased sedimentary deposition and would increase salination in water storage facilities (Dahlin 1983:258). We do have evidence for coastal abandonment, such as the site of Komchen by A.D. 250 (Ringle and Andrews 1990:233), and the Dzibilchaltun area at this time (Andrews 1981:324). Cerros, on the coast of Belize, is also abandoned by A.D. 250 (Freidel, 1979:45). Part of the site is under water, which could be due to sea level rise following its construction, but again, could be also due to shifting beach lines (Dahlin 1983:259). A eustatic rise in sea levels as a symptom of drought is an interesting idea, but further work needs to be carried out before it can be given as a causal factor in the abandonment of coastal Preclassic sites.
Palaeoclimatological records show that A.D. 150-200 was a period of severe cold resulting in an arid environment with reduced rainfall and drought in the Maya Lowlands (Gill 2000:379). Due to the seasonal nature of rainfall in the tropics, any pattern of generally drier conditions would manifest in increasingly frequent and severe mid-summer droughts. A prolonged drought thus could seriously affect even the most organized of settlements causing crop failure, starvation, disease, decrease in population and warfare. Yet Denton and Karlén (1973) produced different results when examining the advance and retreat of tree lines in Alaska and Sweden. Radiocarbon dating trees that were overrun by glacial advances showed that the period from 500 B.C.- A.D. 600 was dry but warm in the Maya region. A colder period followed only from A.D. 600-900. Gunn and Adams note that lowland fluorescence occurred during periods of colder climate, as Maya civilization peaked between A.D. 600-900 (1981:94). Yet fluorescence also occurs in the Mirador Basin during the Middle-Late Preclassic, which is supposedly a warmer period. The key point here is that whether warm or cold, the climate was dry towards the end of the Late Preclassic, which could seriously affect the region.
Fortunately, examination of lake level changes can clear up this issue. The analysis of lake sediment cores from Lake Chichancanab in the Yucatán peninsula has confirmed an arid period between A.D. 125 and 210 (Hodell, Brenner, Curtis and Guilderson 2001). Coastal sites would have been particularly vulnerable as potable water in the northern part of the peninsula was acquired from wells, cenotes and caves. The only means of gaining fresh water was through rainfall, thus a drought of several decades would seriously deplete the freshwater supply causing abandonment of the area (Dahlin 1983:258). Supporting archaeological data shows that by A.D 200 sites along the West Campeche-Yucatan coast were few and far between (only 3 recorded) and before the end of the third century ‘even those few lay abandoned’ (Ball 1978:126). Hansen using Eaton’s (1978) study along the entire north Yucatan coast, pointed out that 25 sites of a total of 55 had Preclassic ceramics from test pits and surface collections. Of these Preclassic sites, 92 percent showed no evidence of Early Classic occupation (Hansen 1990:219). Village sites along Rio Lagartos are the only outliers as they appear to have been abandoned not long after 50 B.C. However, it is likely that the population migrated to El Cuyo, a neighbouring site where a subsequent growth is witnessed, though the centre is abandoned by the end of the Preclassic (Ball 1977:173).
Aridity has also been confirmed by oxygen isotope analysis of sediment cores in Lake Salpeten, Guatemala coinciding with site abandonment in the Mirador Basin (Hodell, Brenner, Curtis and Guilderson 2001). A drop in lake levels at Quexil and Petenxil in the Lake Region between A.D. 200-500 shows a reduction in rainfall at this time (Dahlin 1983:249). Therefore there is substantial proof for drought that coincides with the majority of site abandonments in the Maya Lowlands. Yet reasons for site abandonment are much more complex than a simple comparison between climate and occupation. Human action plays a major role in environmental disturbance.
Intensification of Agriculture and Deforestation
The Maya Lowlands on both a regional and local level were not environmentally homogeneous. There are great differences between the tropical forest and deep soils of the Petén and the drier, thinner soils of the Yucatán peninsula. In the Petén alone, 26 varieties of soil exist with significant differences in fertility (Sanders 1977: 295, see also Dunning, Beach, Farrell and Luzzadder-Beach 1998). In the Southern Highlands and Lowlands of Guatemala during the Late Preclassic Period extensive soil erosion and large-scale deforestation occurs.
The site of Kaminaljuyu practiced extensive agriculture by means of a system of irrigation canals from the Miraflores Lake. However, by A.D. 250 there is presence of houses on what had been the bottom of the lake, showing clear evidence that the lake had dried up and the canal system abandoned. Climactic change with over-exploitation of the waters for irrigation and deforestation is indicated as the cause of this transition (Popenhoe de Hatch 2002:284). The major rivers of Southeastern Pacific Guatemala on which major Late Preclassic centres were located appear to have shifted their course during this time, perhaps due to soil erosion caused by deforestation. Aerial photographs and archaeological evidence show that all major Late Preclassic centres were located on old river channels and were subsequently abandoned and never reoccupied (Estrada Belli 2002:118).
Increased values of Alternanthera-type pollen and Cheno-Am pollen at Edzna reflect that in the Late Preclassic Period we see the peak of cleared and regenerating land (slash and burn system), and increased areas of disturbed vegetation (Forsyth 1983:225). Sediment cores taken from a lake near Dos Hombres in the Three Rivers Region, pointed to a significant environmental disturbance during the Late Preclassic (Dunning 2000:193). Investigations in the Far West Bajo at the mouth of Drainage 3 at La Milpa, revealed a peaty horizon containing pollen from wetland and aquatic plants indicative of perennial wetlands dating to the Late Preclassic Period. Also, clayey sediments were present that buried the topsoil, probably being a result of deforestation and soil erosion (Dunning 2000:195). Thus the regional Preclassic populations practised extensive forms of forest clearance and agriculture, which in turn generated extensive environmental disturbance.
Cores derived from Laguna Tamarindito in Río de la Pasión region, the Petén, confirmed a major episode of forest clearance in the Maya Preclassic, the period being also marked by an increase in sedimentation and a change to more mineral-rich sediments, indicating a significant increase in soil loss. Examination of Aguada Catolina on the Petexbatun Escarpment also indicated severe disturbance of soil cover associated with Preclassic agriculture (Dunning 2000:188). Independent studies of the composition, pollen and influx rates for dated sediment zones in cores from lakes Yaxhá and Sacnab indicate a substantial decrease in percentage of arboreal pollens at the close of the Late Preclassic which indicates large-scale deforestation (Hansen n.d.:23)
The northern Petén is a limestone karst environment with bajos characterizing the drainage of the area (Hansen 2002:277). Hansen, who has worked extensively in this region, particularly in the Mirador Basin, has suggested that the climactic shifts and gradual desiccation of the region can be attributed to environmental degradation, extensive deforestation and over-extending the carrying capacity of the land which played a large part in the abandonment of the area (n.d.:10).
Hansen suggests that significant reduction of the thickness of plaster floors in both El Mirador and Nakbe during the Late Preclassic Period, can be given as evidence that significant deforestation was occurring (Hansen n.d.:26). Large amounts of wood were required in the burning of limestone to produce a small amount of lime for the making of plaster. Since in the Mirador Basin, and in other areas for that matter, lime is used predominantly for making floor and architectural constructions, the decrease over time in thickness and quality of stucco could indicate a decrease in the manufacture of lime as a consequence of deforestation. We see a jump from average floor thickness of 9.04 cm (300-150 B.C.), to 7.29 cm (A.D. 50-150) to an average thickness of 6.48 cm by A.D. 250. Although the change in thickness could be due to mere practical reasons or artistic license, a ‘voracious lime production system’ (Hansen n.d:27) must have been in place as witnessed by the monumental size of the structures at the site. Further support of intensive lime production for the making of plaster is that all the Petén lakes contained a fairly large amount of carbon derived from the dissolution of limestone (Binford 1987:119).
Accessing water was an issue in the Mirador Basin, as no perennial freshwater supplies existed near El Mirador, Nakbe and other smaller sites, the only surface water available during the dry season was found in the highly variable aguadas (Hansen 2002:274). Perennial wet marshlands sustained the population and it was the extensive sedimentation of these marshes in the Late Preclassic Period combined with deforestation that brought catastrophic effects to the peoples of the area (Hansen 2002:287). Between A.D. 150-250 we see the cessation of construction at El Mirador, as well as abandonment of chultuns and the presence of intact Late Preclassic vessels on the floors of structures (Hansen 1990: 216). Therefore, a prolonged drought in conjunction with humans exhausting the environment resulted in the abandonment of the Mirador Basin.
Drought and human agency are also behind the transformation elsewhere of bajos from perennial wetlands and shallow lakes to seasonal swamps between 400 B.C. and A.D. 250 (for example at La Milpa). Many of the earliest centres were built along the edges of bajos, which explains the abandonment of many sites and also the construction of complex storage systems (Dunning et al. 2002). There is little evidence to date for agricultural terracing, which could have significantly slowed soil loss appearing to be more a Late Classic phenomenon (Dunning and Beach 1994:52).
Whether drought affecting an already unstable system in regards to agricultural production caused the abandonment of sites, or the change in environment caused by agricultural intensification led to drought and hence abandonment is not known. Results of climate change and human disturbance on the environment are generally the same, making it difficult to distinguish between the two (Leyden et al. 1996:30). To summarize, we can look towards natural and human factors for the change in climate that transforms the environment in the Late Preclassic, which in turn results in site abandonment. The effects of drought would have varied over different areas, those being most heavily affected would have been areas of low rainfall and higher temperatures and precarious water supply. Small (as in population size), and better watered areas would have had an increased chance of survival.
Water Conservation Techniques
Water conservation activities emerge during the Late Preclassic in the Maya Lowlands. Several centres (such as Tikal, Uaxactun, Calakmul, La Milpa) situated next to bajos continued in occupation through the Late Preclassic Period into the Early Classic by constructing reservoirs and land management features (Dunning et al. 2002). Water conservation features are also seen at El Mirador, but were perhaps constructed too late or were not sufficient to sustain the population.
At Edzna by the Late Preclassic a hydraulic system including 31 canals, including one measuring 12km, and 84 reservoirs were constructed (Matheny 1978). The main canal was part of the original city planning and both the moat and main canal sits in alignment with the geographic and ceremonial centre of the city (Matheny 1978:201). Perhaps it was constructed in response to the drought, as its maximum activity occurred towards the end of the Late Preclassic. The manpower involved in the construction of this hydraulic system points to the fact that water at this time was becoming an increasingly valuable commodity.
Water preservation techniques are evidenced further inland at Dzibilnocac. The inhabitants of this site had only rainwater as its source of water and thus constructed wells and used chultuns for storage (Nelson 1973:33-27). Though it is unconfirmed whether the 19 wells found were built during the Preclassic or Late Classic Period. A complex system of water management involving reservoirs and canals occurred at Cerros in Belize (Freidel and Scarborough 1982). There was also a substantial system of artificial reservoirs at Calakmul constructed in Late Preclassic times. We actually see a move at Calakmul to using bajo edges for cultivation, the floor of El Labertino bajo being occupied after A.D. 200 when the bajo waters were at low levels, probably caused by the drought (Gunn et al, 2002:298).
In the Middle and Late Preclassic we see a major population increase in the Maya area, which by the end of the Late Preclassic leads to competition over resources and a move to more intensive agriculture as experienced in the Mirador Basin. Site abandonment should result in population decline, which is exactly what happens at the end of the Late Preclassic throughout the Lowlands (Figure 3). There is also great linguistic divergence in Southern Mesoamerica during this period, which points to population decrease (Dahlin, Quizar and Dahlin, 1987:278).
In the Mirador Basin there is a sharp decline in population (A.D. 150) after reaching its maximum size and density of both residential and architectural constructions (Hansen 2002:273). In the five Lakes Region (Yaxha, Sacnab, Quexil, Macanche and Salpeten) in the Petén, population decline occurred towards the end of the Late Preclassic (Binford 1987:118). Komchen and Dzibilchaltun by A.D. 250 witnessed a dramatic population decline (Ringle and Andrews 1990:224, Andrews 1981:324). At Santa Rita Corozal we also see a population decline around A.D. 200 (Chase 1990:213).
A drastic reduction in the population of Dzibilnocac occurred at the end of the Late Preclassic. This decline is indicated by the fact that little Early Classic pottery was found and no Early Classic architecture (Nelson 1973:30). There is debate over using the existence of limited Early Classic pottery to determine site occupation. It has been suggested that there was prolonged use of Late Preclassic ceramics into the Early Classic and also an increased differentiation of the use of Early Classic ceramics, whereby the elite forming a sub-complex used these vessels while the remainder of the population used Late Classic pottery (Lincoln 1985, Rice and Rice 1990:126). Thus sites that have been identified as being abandoned due to the absence of Early Classic ceramics might actually be erroneous. However, it is unlikely that there would have been such a prolonged use of Late Preclassic ceramics (Andrews, personal communication 2003) and a wide variety of Early Classic household ceramics are present at a large number of sites (Fry 1990:292), putting the ‘elite sub-complex’ theory into question.
By A.D. 250 at Altar de Sacrificios there was a significant decline in the number of house mounds (Smith 1972:188). A similar pattern also occurs along the middle Usumacinta and the Low Sierras (Rands 1977). The Santa Rosa District, Southeastern Guatemalan coast has a very small sample of Early Classic settlements and little Early Classic pottery (A.D. 200-250). A population decline has been suggested for this (Estrada Belli 2002:117).
Where did the population go? Was the population decline due to mortality caused by disease or warfare or rather a result of migration to larger centres that could sustain people? Population losses in Macanche-Salpeten and Quexil-Pentenxil zones were contemporaneous with population increases in vicinities of sites such as Yaxha and Tikal. This decline might be due to the strong influence of migration to Tikal and Yaxha for reasons of higher agricultural potential, of assured hydrological resources in times of stress, or perhaps effective political persuasion and/or coercion (Rice and Rice 1990:134). Interestingly, the ‘exception’ sites; sites that continued to be occupied through the Late Preclassic into the Early Classic such as Río Azul, Uaxactun and Tikal, are sites with little Late Preclassic occupation. In areas of greater population occupation for example, the Mirador Basin, Northern Belize, Southern Campeche and northern Yucatán, the stresses that occurred towards the end of the Late Preclassic Period were sufficient enough to cause their demise (Hansen, n.d. 30).
Population decline can also be attributed to disease perhaps caused by drought and starvation, although this is difficult to confirm due to limited skeletal preservation in the tropics. Disease could have played a role in site abandonment but as a symptom of the increase in environmental stress on the landscape. The primary factors were drought combined with
agricultural intensification, which to led population decline and abandonment in the Maya Lowlands.
Another cause for population decline and abandonment is warfare, whether internal against the ruler or an external invasion from another site. Archaeological evidence for warfare can be difficult to find, as it may not be reflected in weapons or architectural construction. Chagnon in his study of the Yanomamo stated that although these people are a war-faring society no proof of this could be seen in their material record or settlement (Webster 1977:357 from personal communication with Chagnon). We have no evidence of epigraphy or art depicting warfare for this time period, though admittedly we have a limited sample size. As mentioned above, there are too few osteological collections from this period to enable us to make a credible assessment of this factor.
We do have evidence though of defence fortifications being constructed during the Late Preclassic Period. A dry fortification ditch was built at Becán between A.D. 100-250, a protecting rampart ran along the entire interior of the 2 km ditch except where broken for causeways (Webster 1977:360). At the same time a defensive fortress was built at Edzna enclosed by a moat that is identical in size to the defensive ditch surrounding Becán. The fortress is dated to 250 B.C-A.D.100, based on ceramics recovered from cuts in the moat (Matheny 1976:642-646).
El Mirador is to the east and south bordered by a wall and ditch measuring 1270 m long, 20 m wide and 4-6 m high, with a prominent escarpment dated to the Late Preclassic (Hansen n.d.:7). Fortifications are also evidenced at Tikal; a huge ditch 12½ km long with an embankment that possibly encircles the entire site. The fortifications have been preliminary dated to the Early Classic Period, although of all the ceramic complexes represented in the north survey, only Late Preclassic ceramics were found entirely within the earthworks (Rice and Rice 1981:284 from personal communication with Fry 1980). Further work needs to be carried out to fully ascertain the date of construction. A defensive stone wall construction, 4m high having a perimeter of 1.4 km stands at the site of Muralla de Leon, located on an elevated plateau northeast of Lake Macanche, in the Petén. The site is surrounded by water on three sides and the fourth by a steep arroyo. The walls are tentatively dated to the Late Preclassic (A.D. 1-250) (Rice and Rice 1981:281). Lastly, fortifications are present at Calakmul, a wall in the northwestern part of the city standing 6.15m high and 2 m thick running for 200 m and at Oxpemul, a hilltop fortress, although these constructions are undated (Adams 1981:234). A serious threat of warfare is clearly apparent amongst these sites.
Dahlin points to the abandonment of sites in Northern Yucatán and movement southwards as the reason for the defensive structures present at Edzna and Becán (1983:261). I agree with this theory but for the opposite direction, it is just as likely that the demise of the Mirador Basin and subsequent migration of peoples northwards was perceived as a real threat. Warfare could have occurred by population movement from the Southern Highlands of Guatemala and Salvador exhibited in the Floral Park complex of ceramics, which could have been a cause of conflict, though caution is necessary when making the assertion that a movement of ceramics is tied to a movement of peoples (Rice and Rice 1981:284).
At present there is little evidence to suggest external invasion for the abandonment of sites. There is no evidence for invasion along the coast of Yucatán and it appears that defensive constructions were built after the abandonment of the Mirador Basin indicating that another factor was at play, namely land desiccation. The defensive walls and fortifications erected in the Late Prelassic were to keep out migrants from the Mirador Basin rather than outright war with the capital, El Mirador. If warfare was the causal factor it should produce a marked regionalism in pottery styles, though the Chicanel phase is very much homogeneous in the production of pottery from northern Yucatán to the corner of the Guatemalan Highlands (Webster 1977:375).
Internal revolts against the newly institutionalized form of kinship, is an interesting argument for site abandonment. The depiction of a king, on the Murals de las Pinturas, San Bartolo, holding a ceremonial bar and seated on a throne with a jaguar pelt and the inscribed glyph ahaw confirms that the institution of kingship was in place by A.D. 150. A hiatus in elite construction does occur at a variety of sites in both the Highlands and Lowlands (Table 2), concurrently with site abandonment in other areas, all centering around the A.D. 200-250 period (Figure 4). A ‘slump’ occurs at Tikal at the end of the Preclassic Period (A.D. 150-250), the North acropolis with its 8 temples and shrines was abandoned and its stuccoed buildings torn down (W. Coe 1965:21). In contrast to the rich burials, which occur both before and after this time period, there is only one tomb that has been found in the North Acropolis that dates to the Late Preclassic, and is unusual for its lack of grave goods (M. Coe 1965:13). Similarly at Kaminaljuyu there was a collapse of the centralized socio-political organization and cessation of construction activity, to an extent that by the end of the Late Preclassic Period virtually all of the population had abandoned the valley (Popenoe de Hatch 2002:282). A violent overthrow of the palatial residences in the centre of the ceremonial complex at Chiapas de Corzo occurred, again, at the end of the Late Preclassic (Lee 1969:198).
The institution of kingship, a relatively new concept, might have caused internal instability, but until we know more about the socio-political system within sites during the Late Preclassic Period, this factor can be ruled out. The internal tensions among the elite at these sites could easily be a symptom of stress caused by drought and decline in agricultural production.
Loss of trade either as a producer or consumer has been suggested as a causal factor for site abandonment. Ball proposes that the abandonment along the coast of Yucatán may be due to the disintegration of a coastal trading pattern that prevailed in the Late Preclassic and its replacement by interior trade routes leading directly to the Central and Southern Lowlands (1977:181-186), though this claim is unsubstantiated. According to Freidel, Cerros’ livelihood was based on its coastal trade and so the cause for the site’s abandonment towards the end of the Late Preclassic was the loss of trade networks against increasingly strong and well-managed overland routes, which were controlled by other Maya centres (Freidel 1982:21). Yet if this is true why do we not see an affect on Lamanai at this time, particularly as it was presumed to be controlling Cerros (Andrews, personal communication 2003). In a later article by Schele and Freidel, political reasons, namely the institution of kingship, is mentioned as the primary factor in the site’s decline (Schele and Freidel 1990:128). Yet for reasons explained above we cannot safely point to socio-political factors as the primary cause in site abandonment.
Loss of trade relationships could affect and did affect sites, for example, the Kaminaljuyu-Tikal trade network with the former controlling highland products such as obsidian and jade (Sharer 1994:95). Disruption of interregional communal networks could have an affect on the economies, but there is little evidence to indicate that it would have caused site abandonment during this time period. As one megacenter diminished others were always vying to take its place, such as the case of El Mirador with Tikal and Calakmul being candidates for succession. Thus trade, warfare and population decline are not enough in themselves to cause the widespread abandonment of sites during the Late Preclassic, they were the results, rather than causes of abandonment. These factors took minor positions among much bigger players, namely drought in conjunction with extensive forest clearance and agricultural intensification.
A Model for Late Preclassic Site Abandonment
As demonstrated above, site abandonments and hiatuses occur throughout the Maya area during the Late Preclassic Period particularly in the Yucatán peninsula, the Petén and Chiapas. The general abandonment period runs from A.D. 50 to A.D. 250 in the Maya area, the peak being between A.D. 200-250. Using the data cited in this paper, a tentative model for Late Preclassic site abandonment can be given which may be applied to other cultural areas that have experienced widespread abandonment (Figure 5). During the Late Preclassic we see a significant increase in population throughout the Maya area. Data from lake sediments show that A.D. 125-210 was a period of prolonged drought covering the Yucatán peninsula and the rest of the Maya Lowlands. Rising sea levels, a decrease in rainfall and an increase in salinity, all symptoms of drought would be catastrophic for coastal sites along the Yucatán peninsula. These sites, having thin soils and no natural water systems, experiencing the significant decrease in agricultural production would be abandoned and migration would occur to better watered areas.
The Lowlands where soils were richer, nevertheless, also fell to drought. Substantial evidence from lake sediment cores has shown that intensive agriculture and deforestation occurred, causing soil erosion and reduction in the carrying capacity of the land. This is also evidenced in the highlands area around Kaminaljuyu. The limits on food production, and limitation of lime production for architectural construction led to political instability, combined with factors such as state control over surplus production coming into play. Subsequent disease and starvation led to population decline, which is manifested at the close of the Late Preclassic Period. Population decline occurred mainly by migration to more sustainable sites, causing political realignment. Thus many sites having failed demographically and culturally were abandoned.
Areas with low rainfall and high temperatures, or large populations with maladaptive agricultural techniques or inadequate water supply systems were those first to be abandoned. Some sites were permanently abandoned; others were later reoccupied in the Classic Period while others continued in occupation. The fact that other sites survived and prospered during this period is not paradoxical; it is to be expected. People abandoned sites and migrated to others, resulting in either warfare, hence the fortifications, or an integration of different peoples and perhaps site fluorescence. The centres that continued to be occupied were generally sites with little population that had made the necessary adjustments, in constructing canals and reservoirs, to sustain them through the ‘dark times’. In Flores, on Lake Petén Itza there was extensive occupation from ca. A.D. 150-300, with no signs of stress (Hansen n.d:30). La Milpa, Yaxuna and Bilbao all continued to be occupied through the Late Preclassic into the Early Classic without any hiatus (Dunning 2002:272, Dahlin 1983, Parsons 1969:149). There appears also to be continuity of occupation at Río Azul until A.D. 530 (Adams 1999). However, in an earlier article Adam entertains the idea of a serious archaeological crisis at the end of the Late Preclassic, due to population growth and the use of wood for burning and house construction. At the beginning of the Early Classic some regional centres were abandoned which he suggests was either due to the ecological crisis, or a series of crop disasters brought on by a prolonged drought (Adams 1992:40). Outside of the Maya area there is little evidence for site abandonment during this time. The only data found is from Teotihuacan where in A.D. 250 there was a period of political unrest, destruction of structures and looting (Manzanilla 2003).
Site abandonment in the Maya area is best seen as a process taking effect over a period of time. The graph of Site Disturbances (Figure 4) indicates that drought apparently ended before sites were abandoned. The key point here is that it was the over-use of land for agricultural production and large-scale deforestation that guaranteed sites would be unable to recuperate after the climate had improved.
In trying to fit various different factors into the ‘jigsaw’ of Late Preclassic abandonment, I have found that no one ‘piece’ fits best. The variety within the Maya area in terms of agricultural systems, landscape and political organization, makes it difficult if not impossible to give a blanket explanation of why sites were abandoned by the close of the Late Preclassic. But nor should we. Cultures are complex and do not follow defined rules. There are multi-causal factors for widespread abandonment all interplaying with each other to form a downward spiral and there is no easy leap from abandonment to causality. Ironically, the more one knows about the history of a site and the process of abandonment, causality becomes more distant (Demarest, personal communication 2003).
To gain a broader and more accurate picture of site abandonment, regions such as the Petén and the Yucatán peninsula were examined, this is also where archaeological research has been the most intensive. From this data, drought combined with some other factor, perhaps rising sea levels, appears to be the most compelling argument in the Yucatán peninsula. In the Petén, drought combined with agricultural intensification and deforestation occurred, which is perhaps followed by disease and starvation, population decline, socio-political instability and perhaps loss of trade and warfare, which culminates in site abandonment and migration.
Several important questions emerge from this study that hopefully will be answered in ensuing years. What type of socio-political organization was in play during the Late Preclassic Period? Was Maya society in a rather precarious position and only needed a slight push for it to topple? Can factors behind site abandonment be applied to the Classic Maya collapse? As mentioned, this paper is a preliminary study of Late Preclassic abandonment in the Maya area and is subject to change with additional field data. I hope it can be used though, as the first step in finding out what was exactly happening at the end of the Late Preclassic to make so many people get up and leave.
Adams, Richard E W
1977 Rio Bec Archaeology and the Rise of Maya Civilisation. In The Origins of Maya Civilisation, edited by R. E. W. Adams, pp. 77-99. University of Mexico Press, Albuquerque.
1981 Settlement Patterns of the Central Yucatan and Southern Campeche Regions. In Lowland Maya Settlement Patterns, edited by W. Ashmore, pp. 211-257. University of New Mexico Press, Albuquerque.
1992 Early Classic Maya Civilisation: A View from Río Azul. In The Emergence of Lowlands Maya Civilisation, edited by N. Grube, pp. 36-48. Acta Mesoamericana 8. Anton Saurwein, Verlag.
1999 Río Azul. University of Oklahoma Press, Norman.
Adams, R. E. W. and N. Hammond
1982 Maya Archaeology, 1976-1980: A review of major Publications. Journal of Field Archaeology 9(4):487-512.
Andrews, Anthony P, E Wyllys Andrews V and Fernando Robles Castellanos
2003 The Northern Maya Collapse and its Aftermath. Ancient Mesoamerica 14(151-156).
Andrews, E Wyllys V
1981 Dzibilchaltun. In Supplement to the Handbook of Middle American Indians, edited by V. Bricker and J. Sabloff, pp. 312-344. vol. 1. University of Texas Press, Austin.
Ball, Joseph W
1977 A Hypothetical Outline of Coastal Maya Prehistory: 300 B.C.-A.D. 1200. In Social Process in Maya Prehistory, edited by N. Hammond, pp. 167-196. Academic Press, New York.
1978 Archaeological Pottery of the Yucatan-Campeche Coast. Studies in the Archaeology of Coastal Yucatan and Campeche, Mexico 46. Middle American Research Institute, Tulane University, New Orleans.
1980 The Archaeological Ceramics of Chinkultic, Chiapas, Mexico. New World Archaeological Foundation, Bringham Young University, Provo, Utah.
Binford, Michael W1987 Ecosystems, paleoecology and human disturbance in subtropical and tropical America. Quaternary Science Review 6:115-128.
Chase, Diane Z
1990 The Invisible Maya: Population History and Archaeology at Santa Rita Corozal. In Precolumbian Population History in the Maya Lowlands, edited by T. P. Culbert and D. S. Rice, pp. 199-213. University of New Mexico Press, Albuquerque.
Coe, William R
1965 Tikal: Ten Years of Study of a Maya Ruin in the Lowlands of Guatemala. Expedition 8(1):5-56.
Coe, Michel D
1993 The Maya. Thames and Hudson, London.
Cowgill, George L
1979 Teotihuacan, Internal Militaristic Competition and the Fall of the Classic Maya. In Maya Archaeology and Ethnohistory, edited by N. Hammond and G. R. Willey, pp. 51-71. University of Texas Press, Austin.
Culbert, T. Patrick (editor)
1977 The Classic Maya Collapse. University of New Mexico Press, Albuquerque.
Curtis, Jason H, David A Hodell and Mark Brenner
1996 Climate Variability on the Yucatan Peninsula (Mexico) during the Past 3500 Years and Implications for Maya Cultural Evolution. Quaternary Research 46:37-47.
Dahlin, Bruce H
1979 Cropping Cash in the Protoclassic: A Cultural Impact Statement. In Maya Archaeology and Ethnohistory, edited by N. Hammond and G. R. Wiley, pp. 21-37. University of Texas Press, Austin, Texas.
1983 Climate and Prehistory on the Yucatan Peninsula. Climatic Change 5:245-263.
Dahlin, Bruce H, Robin Quizar and Andrea Dahlin
1987 Linguistic Divergence and the Collpase of Preclassic Civilisation in Southern Mesoamerica. American Antiquity 52(2):367-382.
Denton, George and Wibjorn Karlén
1973 Holocene Climatic Variations – Their Pattern and Possible Cause. Quaternary Research 3(2):155-205.
Dull, Robert A, Southon John R and Sheets Payson
2001 Volcanism, Ecology and Culture: A Reassessment of the Volcán Ilopango TBJ Eruption in the Southern Maya Realm. Latin American Antiquity 12(1):25-44.
Demarest, Arthur A.
1997 The Vanderbilt Petexbatun Regional Archaeological Project 1989-1994: Overview, History, and Major Results of a Multidisciplinary Study of the Classic Maya Collapse. Ancient Mesoamerica 8:209-227.
Dunning, Nicholas and Timothy Beach
1994 Soil Erosion, Slope Management, and Ancient Terracing in the Maya Lowlands. Latin American Antiquity 5(1):51-69.
Dunning, Nicholas, Timothy Beach, Pat Farrell and Sheryl Luzzadder-Beach
1998 Prehispanic Agrosystems and Adaptive Regions in the Maya Lowlands. Culture and Agriculture 20(2/3):87-101.
2000 Stability and Instability in Prehispanic Maya Landscapes. In Landscape Transformations in the Precolumbian Americas, edited by D. Lentz, pp. 179-202. Columbia University Press, New York.
Dunning, Nicholas P, Sheryl Luzzadder-Beach, Timothy Beach, Jonathan G Jones, Vernon Scarborough and T Patrick Culbert
2002 Arising from the Bajos: The Evolution of a Neotropical Landscape and the Rise of Maya Civilisation. Annals of the Association of American Geographers 92(2):267-283.
Eaton, Jack D
1978 Archaeological survey of the Yucatan – Campeche Coast. In Studies in the Archaeology of Coastal Yucatan and Campeche, Mexico, pp. 1-67. 46 ed. Middle American Research Institute Publication, New Orleans.
2002 Putting Santa Rosa on the Map: New Insights on the Cultural Development of the Pacific Coast of Southeastern Guatemala. In Incidents of Archaeology in Central American and Yucatán: Essays in Honor of Edwin M Shook, edited by M. Love, M. Popenhoe de Hatch and H. L. Escobedo, pp. 103-128. University Press of America, Maryland.
Forsyth, Donald W
1983 Investigations at Edzná, Campeche, Mexico: Ceramics. Papers of the New World Archaeological Foundation 46(2). Bringham Young University, Provo.
Freidel, David A
1979 Culture Areas and Interaction Spheres: Contrasting Approaches to the Emergence of Civilisation in the Maya Lowlands. American Antiquity 44:36-54.
1982 The Maya City of Cerros. Archaeology 35(4):12-21.
Freidel, David A and Vernon Scarborough
1982 Subsistence, Trade and Development of the Coastal Maya. In Maya Subsistence: Studies in Memory of Dennis E Puleston, edited by K. V. Flannery, pp. 131-155. Academic Press, New York.
Freidel, David A and Linda Schele
1988 Kingship in the Late Preclassic Maya Lowlands: The Instruments and Places of Ritual Power. American Anthropologist 90(3):547-567.
Fry, Robert E
1990 Disjunctive Growth in the Maya Lowlands. In Precolumbian Population History in the Maya Lowlands, edited by T. P. Culbert and D. S. Rice, pp. 285-300. University of Mexico Press, Albuquerque.
Gill, Richardson B
2000 The Great Maya Droughts: Water, Life, and Death. University of Mexico Press, Albuquerque.
Graham, John A, R F Heizer and E M Shook, Eds.
1978 Abaj Takalik 1976: Exploratory Investigations. Contributions of the University of california Archaeological Research Facility; no. 36. University of California, Department of Anthropology, Berkeley.
Gunn, Joel and Richard E W Adams
1981 Climatic Change, Culture and Civilisation in North America. World Archaeology 13(1):87-100.
Gunn, Joel D, John E Foss, William J Folan, Maria del Rosario Domínquez Carrasco and Betty B Faust
2002 Bajo Sediments and the Hydraulic System of Calakmul, Campeche, Mexico. Ancient Mesoamerica 13:297-315.
1981 Settlement Patterns in Belize. In Lowland Maya Settlement Patterns, edited by W. Ashmore, pp. 157-186. University of New Mexico Press, Albuquerque.
1991 Cuello Considered: Summary and Conclusions. In Cuello: An Early Maya Community in Belize, edited by N. Hammond. Cambridge University Press, Cambridge.
n.d. Early Environmental Impact: Preliminary Assessment of the Ecological Consequences of Incipient Maya Settlement. Unpublished manuscript, Los Angeles.
1990 Excavations in the Tigre Complex El Mirador, Petén, Guatemala. El Mirador Series Part 3. New World Archaeological Foundation, Brigham Young University, Provo, Utah.
2001 The First Cities: The Beginnings of Urbanization and State Formation in the Maya Lowlands. In Maya: Divine Kings of the Rainforest, edited by N. Grube.
Hansen, Richard D, Steven Bozarth, John Jacob, David Wahl and Thomas Schreiner
2002 Climatic and Environmental Variability in the Rise of Maya Civilisation. Ancient Mesoamerica 13:273-295.
Hodell, David A, Mark Brenner, Jason H Curtis and Thomas Guilderson
2001 Solar Forcing of Drought Frequency in the Maya Lowlands. Science 292:1367-1370.
Lee, Thomas A. Jr.
1969 The Artifacts of Chiapas de Corzo, Chiapas, Mexico. Papers of the New World Archaeological Foundation 26. Brigham Young University, Provo.
Leyden, Barbara W, Mark Brenner, Tom Whitmore, Jason H Curtis, Piperno Dolores R and Bruce H Dahlin
1996 A Record of Long-and Short-Term Climactic Variation from Northwest Yucatán: Cenote San José Chulchacá. In The Managed Mosaic: Ancient Agriculture and Resource Use, edited by S. L. Fedick, pp. 30-50. University of Utah Press, Salt Lake City.
Lincoln, Charles E.
1985 Ceramics and Ceramic Chronology. In A Consideration of the Early Classic Period in the Maya Lowlands, edited by G. R. Willey and P. Mathews, pp. 55-94. Institute for Mesoamerican Studies, State University of New York, Albany.
2002 Ceramic Chronology of Preclassic Period Western Pacific Guatemala and its Relationship to Other Regions. In Incidents of Archaeology in Central America and Yucatán, edited by M. Love, M. Popenhoe de Hatch and H. L. Escobedo. University Press of America, Lanham, Maryland.
Lowe, Gareth W
1977 The Mixe-Zoque as Competing Neighbors of the Early Lowland Maya. In The Origins of Maya Civilisation, edited by R. E. W. Adams. vol. 197-248. University of New Mexico Press, Albuquerque.
2003 The Abandonment of Teotihuacan. In The Archaeology of Settlement Abandonment in Middle America, edited by T. Inomata and R. W. Webb, pp. 91-101. University of Utah Press, Salt Lake City.
Martin, Simon and Nikolai Grube
1995 Maya Superstates. Archaeology 48(6):41-46.
Matheny, Ray T.
1976 Maya Lowland Hydraulic Systems. Science 193:639-646.
1978 Northern Maya Lowland Water-Control Systems. In Pre-hispanic Maya Agriculture, edited by P. D. Harrison and B. L. Turner II, pp. 185-210. University of Mexico Press, Albuquerque.
Michels, Joseph W
1979 The Kaminaljuyu Chiefdom. Pennsylvania State University Press, Pittsburgh.
Nelson, Fred W
1973 Archaeological Investigations at Dzibilnocac, Campeche, Mexico. Papers of the New World Archaeological Foundation 33. Bringham Young University, Provo.
Parsons, Lee A
1969 Bilbao, Guatemala: An Archaeological Study of the Pacific Coast Cotzumalhuapa Region. Publications in Anthropology 12 2. Milwaukee Public Museum, Milwaukee, Wisconsin.
Pendergast, David M.
1981 Lamanai, Belize: Summary of Excavation Results, 1974-1980. Journal of Field Archaeology 8(1):29-53.
Popenhoe de Hatch, Marion
2002 New Perspectives on Kaminaljuyú, Guatemala, Regional Interaction during the Preclassic and Classic Periods. In Incidents of Archaeology in Central American and Yucatán: Essays in Honor of Edwin M Shook, edited by M. Love, M. Popenoe de Hatch and H. L. Escobedo pp. 277-296. University Press of America, Maryland.
Rands, Robert L
1977 The Rise of Classic Maya Civilisation in the Northwestern Zone: Isolation and Integration. In The Origins of Maya Civilisation, edited by R. E. W. Adams, pp. 159-180. University of New Mexico Press, Albuquerque.
Rice, Don S and Prudence M Rice
1981 Muralla de Leon: A Lowland Maya Fortification. Journal of Field Archaeology 8(3):271-288.
Rice, Don S and Prudence M Rice
1990 Population Size and Population Change in the Central Peten Lakes Region, Guatemala. In Precolumbian Population History in the Maya Lowlands, edited by T. P. Culbert and D. S. Rice, pp. 123-148. University of New Mexico Press, Albuquerque.
Ringle, William M and E Wyllys Andrews
1990 The Demography of Komchen, An Early Maya Town in Northern Yucatan. In Precolumbian Population History in the Maya Lowlands, edited by T. P. Culbert and D. S. Rice, pp. 215-243. University of New Mexico Press, Albuquerque.
Sanders, William T
1977 Environmental Heterogeneity and the Evolution of Lowland Maya Civilization. In The Origins of Maya Civilization, edited by R. E. W. Adams, pp. 287-297. University of New Mexico Press, Albuquerque.
Santley, Robert S
2003 The Tuxtlas as Voclanic Hazard: Volcanism and its Effects on Site Founding and Abandonment in the Tuxtlas Mountains, Southern Veracruz, Mexico. In The Archaeology of Settlement Abandonment in Middle America, edited by T. Inomata and R. W. Webb, pp. 163-181. University of Utah Press, Salt Lake City.
Schele, Linda and David A Freidel
1990 A Forest of Kings: The Untold Story of the Ancient Maya. William Morrow, New York.
Scuderi, Louis A
1990 Tree-Ring Evidence for Climatically Effective Volcanic Eruptions. Quaternary Research 34:67-85.
Shafer, Harry J and Tom R Hester
1983 Ancient Maya Chert Workshops in Northern Belize, Central America. American Antiquity 48:519-543.
Sharer, Robert J
1994 The Ancient Maya. Stanford University Press, Stanford, California.
Sheets, Payson D (editor)
1983 Archaeology and Volcansim in Central America: The Zapotitlan Valley of El Salvador. University of Texas Press, Austin, Texas.
Shook, Edwin M and Tatiana Proskouriakoff
1956 Settlement Patterns in Mesoamerica and the Sequence in the Guatemalan Highlands. In Prehistoric Settlement Patterns in the New World, edited by G. R. Willey. Viking Fund Publications in Anthroplogy 23, New York.
Smith, Robert E
1971 Ethnology Harvard University. vol. 66. Peabody Museum of Archaeology.
Smith, Ledyard A
1972 Excavations at Altar de Sacrificios: Architecture, Settlement, Burials and Caches. Papers of the Peabody Museum of Archaeology and Ethnology, Harvard University 62(2).
Tainter, Joseph A
1988 The Collapse of Complex Societies. Cambridge University Press, Cambridge.
Yoffee, N. and G. L. Cowgill (editors)
1988 The Collapse of Ancient States and Civilizations. University of Arizona Press, Tuscon.
1977 Warfare and the Evolution of Maya Civilisation. In The Origins of Maya Civilization, edited by R. E. W. Adams, pp. 335-371. University of New Mexico Press, Albuquerque.
1979 Cuca, Chacchob and Dzonot Ake. Pennslyvania State University Press, University Park.
2002 The Fall of the Ancient Maya. Thames and Hudson, London.
Willey, Gordon R
1977 The Rise of Classic Maya Civilization: A Pasión Valley Perspective. In The Origins of Maya Civilization, edited by R. E. W. Adams, pp. 133-157. University of New Mexico Press, Albuquerque.