The Drought Hypothesis

Until recently, it has been difficult to prove the drought hypothesis since the traditional palynological (i.e. pollen and macrofossil) studies used to reconstruct past climates have made it somewhat difficult to distinguish natural versus anthropogenic causes of changes in the assemblages.

Hodell et al. (1995) was one of the first papers to use other methods; they looked at temporal variations in ¹⁸O/¹⁶O ratios in a 4.9-m continuous sediment record, representing 9000 years, from Lake Chichancanab taken in 1993. It is a closed-basin lake (the largest in the Mexican part of the Yucatan peninsula), and so it loses a lot of its water through evaporation (there is not outflow via rivers into the ocean). Therefore, the oxygen isotopic composition of the lake water generally reflects evaporative conditions during the dry season very well, where past changes are encoded within gastropod and ostracod shells preserved in the lake sediments. 

Location of Lake Chichancanab, Yucatan, Mexico. Source: Hodell et al. (1995)

They proposed that during dry conditions with high E/P (evaporation/precipitation ratio), there would have been higher δ¹⁸O values; during wetter conditions, the opposite would have occurred. Through this, they found that the timing of the Classic Maya collapse was coeval to that of climate drying, providing the first unequivocal evidence of drought between 800 and 1000 AD (1,300-1,100 yr BP).  The record also shows that this was the driest period throughout the past 8,000 years. Gypsum/calcite ratios were also used to determine past changes in E/P, as they reflect changes in the hydrologic budget of the lake.

At present, the lake slightly exceeds saturation for gypsum (sulfur) and so gypsum remains in solution in the open water of the lake, as minerals generally only precipitate if the water is supersaturated.  Precipitation only occurs in the shallow areas around the edge, where evaporation is greater. When evaporation is high or rainfall is low, the lake volume is reduced and the gypsum saturation is exceeded. This leads to gypsum precipitation throughout the entire lake that is deposited on the lake floor and becomes preserved in the sediment record, providing a proxy of past increases in evaporation/precipitation ratios (E/P). SO, a lot of gypsum deposition (found in thick layers in the sediment core) is essentially a sign of past drought. This was evident in the sediment record during the Late Classic abandonment, where there were high gypsum concentrations present, suggesting optimal E/P conditions (see figure below). 

Of course, this one record cannot alone support such a theory; yet there has been a multitude of evidence since this paper from various proxies. The team actually returned to  the site five years later to obtain a more highly resolved sediment core that allowed them to look at variations in greater detail and pinpoint smaller-scale changes. They found that the droughts of the Yucatan increased in strength and frequency prior to 1100 AD, parallel to the observations from the northern Great Plains in North America and other parts of Central America.


However, the study also unveiled the effects of deforestation and soil erosion attributed to humans, and so we cannot fully rely on it to support the climate change hypothesis.

To further investigate whether climate change did impact the Maya, Haug et al. (2003) studied a ‘cleaner’ marine sediment core taken from the anoxic Cariaco Basin - just off the coast of northern Venezuela:

Both the Yucatan and the Cariaco Basin experience the same general climate, with distinct dry and wet seasons, a result of the migration of the ITCZ (an equatorial band of rainfall). During the winter it moves south of both places and in summer it moves north, encompassing both areas (shown by the dark green bands).
Why did they choose a record so far away? Because it is completely surrounded by a shallow continental shelf that prevents deep Cariaco waters from mixing with the open ocean. This deep water is anoxic and therefore does not support deep-sea organisms that tend to churn up sediment layers deposited each season. Hence, the undisturbed sediments  preserves a detailed record of past rainfall. Source: American Scientist

Layers of alternating bands of dark and light deposits were discovered in the undisturbed core, where the light colours were algae and other tiny fossils and the dark bands consisted of titanium. This bulk titanium content is thought to reflect changes in input from river banks and the hydrological cycle over the southern Caribbean region. During the rainy season, titanium would be washed into the sea via the rivers, and so thicker dark bands that indicate higher levels of metal, reflect a lot of rain. During drier periods and subsequently a weaker river flow, there would be less titanium washed in and so bands would be thinner. The seasonally resolved record of titanium shows exactly this; there was an extended period of dry conditions (that appears to have lasted from 760-930) around the same time of the Terminal Classic collapse, with three intense droughts centered around 810, 860, and 910 AD.

The bottom panel shows the titanium content for the past 2,000 years, encompassing the terminal Classic collapse as well as  manifestations of what have become known as the "Medieval Warm Period" and the "Little Ice Age". The middle panel shows the record in more detail for the period of interest (it has been smoothed for clarity). The top panel focuses on the Late Classic period, showing evidence of four multi-year droughts (indicated by low titanium content), separated by 40 to 50 years of more moderate conditions. Source: American Scientist 

Now, is it a coincidence that these dates coincide exactly with Gill’s suggested collapse and abandonment of the major cities and diminished monument building? From this evidence, I think it is safe to say that the Maya had never faced such drastic climatic changes, with the most severe drought of the past 7,000 years occurring at the pinnacle of their 1,500 years of existence.