There's no doubt about drought

In addition to those already mentioned, Hodell et al.’s (1995) research prompted several paleoclimatic investigations into the Mayan region that suggest drought conditions were a feature of that period. Each focuses on different locations throughout the Mesoamerica and several proxies are employed. Instead of reviewing each in detail, a selection has been chosen and a summary of their findings is presented here.   

Curtis et al. (1996) reconstructed 3,500 years of climate for the Yucatan Peninsula using a 6.3 m lake sediment core taken from Lake Punta Laguna, Mexico (Figure 1). The proxies measured were δ¹⁸O in mono-specific ostracods and gastropods. 

Figure 1: Map showing location of Lake Punta Laguna in Quintana Roo, Mexico (Source: Curtis et al. 1996)

 The oxygen isotopic record from Punta Laguna indicated that the terminal Classic and earliest Post-classic period (800 to 1050 AD) was one of the driest intervals of the last 3,500 years. The high resolution record produced from Punta Laguna allows identification of peak drought conditions at 862 AD during the terminal Classic period (Figure 2). 

Figure 2: Comparison of the oxygen isotopic record from the Punta Laguna sediment core with Maya cultural periods (Source: Curtis et al. 1996)

 

Islebe and Sanchez (2002) also found there to be a drier period coincident with the period of Maya demographic decline. They analysed the pollen record of a sediment core taken form the Mexican Caribbean coast, showing the development and changes in the mangrove system (Figure 3). 

Figure 3: Location of the coring area along the Mexican Caribbean
coast (Source: Islebe and Sanchez 2002)

 

They found that the mangrove species Conocarpus erecta dominated during the period approximately 1500-1200 ¹⁴C yr BP while R. mangle almost disappeared and other taxa emerged, suggesting drier climatic conditions and generally more open vegetation.

A study by Carrillo-Bastos et al. (2010) once again found there to be drought conditions coincident with the collapse of the Classic Maya. They studied oxygen isotope measurements on a 2.5 m sediment core recovered from Lake Tzib central Quintana Roo, southeast Mexico (Figure 4). 

Figure 4: Map showing the location of Lake Tzib (Source: Carrillo-Bastos et al. 2010)

 

Their data indicated dry conditions between ~1300 and 1200 cal yr BP. They correlated the dry peak at 1200 cal yr BP with a dry period found using a marine core taken in the Cariaco Basin, north of Venezuela (Figure 5).

Figure 5: Correlation of  isotopes from Lake Tzib with Ti% data from the marine Cariaco Basin core (Source: Carrillo-Bastos et al. 2010)

  

It can be seen that the peaks in drought conditions found do correlate with the collapse of the Classic Maya.

 

With so many studies correlating drought conditions to the time of the Classic Mayan collapse, from such diverse parts of the empire, using such a wide variety of proxies, it seems very likely that drought conditions were experienced by all the Maya and caused population decline. Where there may be slight discrepancy in the chronology and exact dating of these is probably due to the differing dating techniques employed. It remains, however, to understand why this drought occurred and how much it affected different groups of Maya from around Central America.

Drought in the south?

Following early work in the northern parts of the Mayan region, studies have been conducted in the southern areas in order to establish that drought was the cause of collapse for the entire Mayan people. One such study was conducted by Curtis et al. (1998). They produced a climate record for Lake Peten-Itza, Peten, Guatemala, in the southern lowlands of the Mayan empire. They used oxygen isotopic records from a 5.5 m core as a proxy for climate change. However, unlike the studies in the northern regions of the Mayan territory, the core displayed no evidence for a Terminal Classic drought in the southern lowlands (Figure 1). 

Figure 1: Oxygen isotopic composition in snail shells (Cochliopina sp. and Pyrgophorus sp.) and ostracod valves (Cytheridella ilosvayi and Candona sp.) versus radiocarbon age in the Peten-Itza core (Source: Curtis et al. 2008)

 

Curtis et al. (1998) conclude that several factors, or a combination of factors, may account for the apparent climatic disparities between the northern and southern areas. First, it might be possible that the drought was local in extent and affected only the more northerly portion of the Yucatan Peninsula, but did not extend into the Peten lowlands. However, it is stated that this scenario is unlikely given the emerging evidence for dry conditions around 900 AD in southerly areas, such as Costa Rica. Second, Curtis et al. (1998) suggest that the sampling resolution of the Peten-Itza core may have been insufficient to record the drought event. However, this is said to be unlikely because the mean sample spacing in the Peten-Itza core for the period in question is about 15 years which exceeds the resolution at of Hodell et al.’s (1995) core from Lake Chichancanab where the drought was clearly evident and was recorded in many samples. It is said that the most plausible explanation for the lack of a drought signal in the Peten-Itza core is that the lake is simply too large (99 km²) and deep (~60 m) to record climatic changes that persist for less than several centuries. In comparison, Lakes Punta Laguna (0.9 km², ~12 m) and Chichancanab (10 km², ~12.5 m) have sufficiently small volumes that their lakewater. Curtis et al. (1998) state that ¹⁸O responds quickly to changes in E/P and Peten Itza’s large volume, and consequent long residence time, make it relatively insensitive to all but the most dramatic, long-term shifts in E/P.

It seems that in order to establish whether drought affected the southern lowlands, where Classic Mayan collapse was most pronounced, it will be necessary to conduct studies in smaller southern-Mayan lakes with high sedimentation rates and continuous records of preserved carbonate microfossils as proxies.

One such study was performed by Rosenmeier et al. (2002) and suggested that drought conditions were apparent at the time of Mayan collapse. They acquired a 4000-yr, 15 m sediment core record from Lake Salpet´en, Guatemala. It is said that Lake Salpet´en is a small, closed-basin lake that lies 104 m above sea level and has a maximum depth of 32 m. It is located in the southern portion of the Mayan empire.

High levels of δ¹⁸O, drought conditions, were found between 850 and 900 cal yr A.D. (Figure 2) and are concordant with the Classic Maya population decline between 800 and 900 A.D.

Figure 2: Oxygen isotope records from Lake Salpet´en (Source: Rosenmeier et al. 2002)

 

Rosenmeier et al. (2002) state that this period of high δ¹⁸O may well have been caused by greater aridity, documented in other northern Yucatan lakes. However they also state that the proxy used may just be recording the decreased hydrologic input to the lake as a consequence of forest recovery.

The evidence for drought conditions presented by Rosenmeier et al. (2002) goes some way in confirming that such conditions were experienced throughout the Mayan empire. It seems that the contradictory results produced by Curtis et al. (1998), discussed previously, were indeed erroneous. The reasoning Curtis et al. (1998) provided for their unexpected findings also seems to be proven correct by Rosenmeier et al. (2002).  It was said, by Curtis et al. (1998), that the lake they studied was too large to detect decadal climate shifts as it experience long residence times, making it relatively insensitive. In a similar area, Rosenmeier et al. (2002) studied a much smaller lake and such smaller scale climate shifts seem to have been picked up.

Summary poster

In order to reinvigorate your interest, and futher appeal to your senses, I have produced a poster that summarises what has been learnt so far during this accademic escapade. It may also prove useful as a convenient starting point for those who have not read previous posts. (Click to enlarge)

Conclusions so far...

 It is often stated that this debate has two sides, those with the social arguments for downfall and those who see climate as its major cause. However, it seems that the correlation between Mayan collapse and climate drought is not simply causal. It is this assumption that often leads some sociologists to pigeon-hole climate scientists as having over-emphasis on environmental determinism. It is evident that there are much more complex relationships occurring, with droughts potentially varying greatly over space and in intensity, affecting groups of Maya with vastly different tolerances to them. It seems then that in order to fully understand why the Mayan civilisation collapsed, there is a need for a complete climate record for the entire Mayan region as well as a knowledge of how the environmental changes might have impacted social processes of the different groups of Maya. Collaboration between paleo-environmentalists and anthropologists/archaeologists seems vital.

News From the Front – A Japanese Rambo!?

Refreshingly, the debate surrounding the collapse of the Classic Maya has arisen elsewhere in cyberspace this week. There is a section of the New York Times website called ‘Scientist at Work – Notes From the Field’, a blog described as the modern version of a field journal – a place for reports on the daily progress of scientific expeditions. On 27^th February Takeshi Inomata, a professor of anthropology specialising in the ancient Maya, draws the reader’s attention to the on-going field science aiming to put the debate surrounding the Mayan population breakdown to rest once and for all.

Whilst primarily studying the anthropological aspects of the ancient Maya, the writer describes the arrival of a team of Japanese geologists and plant scientists who aim to uncover more about the area’s past climate (Figure 1). The group, including a lake-corer known as ‘the Japanese Rambo’, aim to collect sediment from 3 lakes in Guatemala, comprising of Lakes Petexbatun, Las Pozas and Quexil in the central Guatemalan lowlands. Results will stem from the analysis of geochemistry, isotopes, pollens, diatoms and other remains and appear to be successful already, with a series of four-metre-long cores extracted. However, the rich amounts of data on environmental change contained within each core will not be analysed until they are transported back to Japan.  

Figure 1: The group of Japanese scientists with a core sample taken from Lake Petexbatun (source)

 

Building on what we learnt from the paper by Hodell et al. (1995), it is claimed by the writer that though apparent that there were major droughts in the northern parts of the empire at the time of collapse, archaeological evidence suggests that centres prospered in this region during these supposed dry spells. Furthermore, it is stated that the southern regions witnessed a far greater abandonment, but evidence of droughts in this area still remains unclear.  

It is clear to see the need for a greater depth of study into the spatial extent of the defined droughts and the ability of different Mayan groups to resist such conditions. It is as a result of this need that groups of scientists, such as the Japanese, are working to discover new evidence of drought in the southern regions of the Mayan civilization and why anthropologists and archaeologists, like Takeshi Inomata, are working to find out how different sets of people were affected by climate changes.

"Stalagtites hold tight, stalagmites..."

Webster et al. (2007) also outlined a link between drought and Classic Maya collapse. This was done by presenting a climate record for the Maya region using evidence obtained from a stalagmite once located in the Macal Chasm in the Vacal Plateau, Belize (Figure 1).

Figure 1: Location map showing Macal Chasm and other important Mayan sites (Source: Webster et al. (2007))

Past climate was reconstructed from analysis of the reflectance, colour, luminescence, and carbon and oxygen stable isotope records for the period from 1225 B.C. to the present. The record thus encompassed the Maya Pre-Classic, Classic, and Post-Classic periods. 

The stalagmite record indicates that a series of droughts, which collectively form the most prolonged dry interval in the 3300-year record, lasted from A.D. 700 to 1135 and thus coincided with the collapse of the Maya civilization (Figure 2).  

Figure 2: Droughts in Belize indicated by a reduction in stalagmite luminescence compared with Maya monument production (Source: Webster et al., 2007)

For comparison and validation Webster et al. (2007) referenced the work of Keigwin (1996). Keigwin (1996) produced a record of the Holocene variability of sea surface temperatures (SSTs), salinity and flux of terrigenous material in the Sargasso Sea, within the North Atlantic Ocean. A sea-sediment core was taken in order to accomplish this, due to the high rates of sediment accumulation in the region. 

Keigwin (1996) observed that variation in SSTs in the record showed separate distinct rises in temperatures that correlate with the evidence for Mayan collapse. Though the Sargasso Sea is located a significant distance from the location of the Mayan empire, it is claimed that the record, most probably, corresponds to climate change at a “basin or hemispheric scale”. However, this conclusion is yet to be wholly substantiated and is open to debate.

Webster et al. (2007) compiled their own findings with those of Keigwin (1996) and Hodell et al. (1995) (discussed earlier in the blog), as shown in Figure 3.

Figure 3: Comparison of the Belize stalagmite record with Sargasso Sea and Lake Chichancanab data (Source: Webster et al., 2007)

Webster et al. (2007) state that these correlations demonstrate that the Macal Chasm stalagmite provides a regional record of climate, rather than a local record of land-use change or a record only of changes within the cave itself. More significantly, they demonstrate that the dry periods in the Macal Chasm record were widespread events that would have affected the entire Mayan civilization

It is apparent that all three studies show periods of drought occurring at the same time – the Classic Maya collapse. Webster et al. (2007) conclude that severe dryness affected a broad region of Mesoamerica most likely contributed to the collapse of the Maya civilization during the Late Classic period.

While this evidence is still not definitive for the entire Mayan region and does not explore the social implications of the drought, it can be said that the greater the number and spatial extent of the proof that drought did occur at the time of Mayan collapse, the more readily this factor can be accepted as a playing a key role in the demographic decline.

Humble Beginnings: The Early Evidence

There is no better way to begin the accademic side of this blog than by reviewing an early paper that was instrumental in rethinking the Maya collapse, as featured in the previous post's video. In 1995, Nature published a journal article by Hodell et al. that put climate change firmly onto the table of possible causes behind the drastic population decline of the Maya.

Though climate change had been previously considered as a possible trigger of Mayan ruin, prior to this paper’s publication much of the evidence cited had been particularly ambiguous. For example, the suggested use of vegetation records as a proxy for drought were compromised by the vast anthropogenic influence in the area. The extent of human-induced deforestation in the region has altered vegetation such that it mimics past climate shifts, thus making it difficult to discriminate between the effects of natural climate variation and anthropogenic factors.

Here, Hodell et al. (1995) present evidence from sediment cores taken from Lake Chichancanab in Mexico in order to reconstruct a continuous Holocene climate record for the central Yucatan Peninsula. This is done through the study of temporal changes in Oxygen isotope and sediment composition found in a 4.9 metre core. Importantly, the lake is closed-basin and located in the Yucatan, a populous area in Mayan times (Figure 1).

Figure 1: Map of Central America showing location of Lake Chichancanab in the Yucatan Peninsula (Source: Hodell et al. 1995)

 

Hodell et al. (1995) state that the character of past climate can be ascertained by studying the ratio of evaporation to precipitation (E/P), whereby climates have a higher E/P when drier and a lower E/P when wetter. In order to infer the relationship of evaporation to precipitation here, the hydrologic budget of the lake is studied. This is done through two proxies.

The first proxy for climate change applied is that of Oxygen isotope records. It is proposed that when it is dry and E/P is high, there is a greater evaporative loss of H₂¹⁶O from the lake and therefore greater ¹⁸O enrichment.  Therefore, by studying the ratio of ¹⁸O/¹⁶O past climate can be inferred. In this paper, such changes in the ratio of ¹⁸O/¹⁶O are recorded in the δ¹⁸O of carbonate in ostracod and gastropod shells preserved in lake sediment.  

The second proxy used is that of changes in the ratio of gypsum/calcite that can be recorded in sediment cores. It is said, as gypsum does not precipitate into the lake today, that changes in the relative proportions of these minerals in the sediment indicate past variations in saturation of Lake Chichancanab’s water. This therefore reflects on the state of the lake’s hydrologic budget, the ratio of E/P and the overall climate of the time. Drier climates, then, are reflected by a more saturated lake with a greater proportion of gypsum to calcite in the sediments.

To summarise, it is proposed that a drier climate, with a higher E/P, can be reflected by an increase in δ¹⁸O values and an increased proportion of gypsum to calcite found in lake sediments. The opposite is true for wetter climates.

In order to date the sediment in the core, radiocarbon analysis was performed on both aquatic shells and terrestrial material, such as seeds and charcoal. The radiocarbon ages were then converted to calendar years using the decadal tree ring dataset and a specific conversion program called CALIB. Approximate errors of between 30 and 65 years (±) were certified, which appear acceptable.

Figure 2 portrays the result of the sediment core taken from Lake Chichancanab. It can be seen that there is a general consistency between the Oxygen isotope and the gypsum/calcite analyses. It was calculated that the driest conditions, portrayed by the peaks in both δ¹⁸O values and Sulphur (S) content, the measure of gypsum/calcite, occur between AD 800 and 1000 with a peak at approximately AD 922. The timing of this dry spell corresponds well with the collapse of the Classic Maya civilization between AD 750 and 900. The full continuous Holocene climate record produced by Hodell et al. (1995) also sets this period into context as the driest episode in an 8,000 year period. 

Figure 2: Proxy signals of paleoclimate plotted over calendar years to 1000 BC and against subdivisions of Mayan cultural evolution. S content (%) represents the gypsum/calcite ratio, while δ¹⁸O values are represented separately for ostracods and gastropods. (Source: Hodell et al. 1995)

These results presented the first unambiguous evidence for climatic drying at the time of the Mayan collapse. The findings do indicate a causal link between climate change and the dramatic population decline that occurred. However, as the Mayan empire was so geologically, ecologically and climatically diverse, the climate response may have varied throughout the region, suggesting a less than uniform decline than was experienced. Even if the drying occurred throughout the area, its effects may well have varied spatially due to differing sensitivities of cultural and ecological responses.

It is suggested by Hodell et al. (1995) that further study is needed in order to determine the full magnitude and geographical extent of the dry period that occurred between AD 800 and 1000, and in order to explain the intraregional pattern of collapse of the Classic Mayan civilization.

And so the investigation continues...