EIGHT
Ice Returns
THE NEOGLACIATION
We are living in an epoch of renewed but moderate glaciation—a “little ice age” that already has lasted about 4,000 years.
FRANÇOIS MATTHES, Report of Committee on Glaciers
THE WARM, DRY CONDITIONS of the mid-Holocene gradually gave way to a cooler, wetter period known as the “Neoglaciation.” It had no clearly demarcated beginning, settling unevenly over the Northern Hemisphere with considerable local variation. In Europe, notably in southern Norway and in the Swiss and Austrian Alps, the onset of this cooler period occurred 5,000 to 4,000 years ago, based on the dating of glacial moraines and of sediments that were eroded by glaciers and transported by glacial meltwaters. Glacial expansion across Iceland, beginning 5,000 years ago, and the retreat of the Eurasian tree line between 4,000 and 3,000 years ago also signaled cooler temperatures.
In the mountain ranges of the American West, many cooler summers passed without melting the previous winter’s snow accumulation. At a pace of tens of feet per year, glaciers in the high mountain ranges of the West, from Alaska to western Canada and into the Cascades, Sierra Nevada, and Rockies, grew and flowed downslope, and the tree line migrated to lower elevations.
This period of Northern Hemisphere cooling also brought wetter conditions to the American West. Lakes formed in unlikely places, including some of the driest places in the West today, such as the Mojave Desert of southeastern California. When researchers cored into the dry Mojave lakebeds, or playas, they were surprised to find evidence that large ephemeral lakes, including Silver Lake today, had formed there 3,620 years ago. In nearby Searles Basin, playa lakes formed about 3,590 years ago.
FIGURE 20. Ancient levels of Mono Lake over the past 3,800 years, reconstructed by using the deltaic sediments of inflowing streams. Note the high stand 3,700 years ago. (Based on data from Stine 1990.)
In eastern California, more evidence for a wet Neoglacial climate was found in Mono Lake in the Owens Valley, to the east of Yosemite National Park. Mono is a special lake, not just for its unique ecosystems and fascinating tufa towers but also for its great age: the lake has existed for over a million years, making it one of the oldest lakes in North America. The sediments accumulated beneath Mono Lake therefore contain a rich archive of past natural history. The lake is a “closed basin,” meaning water flows into it but has no outlet. During dry periods, when the rate of evaporation from the lake exceeds inflows from direct precipitation and streams, the water level subsides. During wetter times, when more water flows into the lake than evaporates, the lake expands. In this way, Mono Lake acts as a prehistoric rain gauge. Several lines of evidence indicate that the lake swelled to its highest known level (or high stand) some 3,730 years ago (see figure 20).
Owens Lake in the Owens Valley south of Mono Lake, having desiccated during the prolonged mid-Holocene drought, also reached a high stand during the Neoglacial. In the Sierra Nevada range, Lake Tahoe, which had been unusually low for hundreds of years during the mid-Holocene, was restored to its early Holocene high level by about 3,000 years ago. The lake again reached the elevation of its sill, causing lake waters to overflow into what is now known as the Truckee River, its only outlet. These waters flowed eastward into Nevada, eventually reaching Pyramid Lake, which also reached a high stand about 3,000 years ago.
Fossilized pollen and plant remains in lakes and meadows in the Sierra Nevada provide further evidence for cooler and wetter climatic conditions. For instance, pollen from Lake Moran in the Sierra Nevada shows a shift in vegetation to cool-adapted fir and pine trees in the region surrounding the lake, starting around 4,000 years ago. An analysis of charcoal levels in Sierra lake sediments shows that these levels declined during the Neoglacial, indicating reduced fire frequency.
CALIFORNIA’S CENTRAL VALLEY
More snow and rain in the Sierra Nevada would have caused more serious flooding along the rivers draining the range, including the Sacramento and San Joaquin, as floodwaters would overtop the banks of the rivers, spread out over the floodplains of the Central Valley, and fill the natural wetlands that once covered the region. Evidence that major flooding occurred in the Central Valley during the Neoglacial period is found in Tulare Lake, located in the southern San Joaquin Valley. Prior to its recent human-caused desiccation, Tulare was the largest freshwater lake in the American West, receiving inflow from the Kings, Kaweah, Tule, and Kern rivers that drain the central and southern Sierra Nevada. During wet periods, these rivers brought even more water to the lake, expanding Tulare and its surrounding wetlands.
Paleoclimate researchers analyzed pollen extracted from Tulare Lake sediments to assess the extent of marshes in the Tulare Lake Basin by determining the proportion of aquatic plants relative to saltbush, a plant that grew around the lake’s shore. These sediment studies, along with shoreline analyses, show that Tulare Lake experienced a high stand between 4,000 and 2,700 years ago. The increased high water and flooding during this generally wetter period would have created conditions much like those seen in 1861–62, when the Central Valley was transformed into an inland sea. These earlier floods would have destroyed any human settlements in the Central Valley. For this reason, it is likely that the native populations used the Central Valley only seasonally and not as a location of permanent settlement.
MOUND BUILDERS OF THE SAN FRANCISCO BAY
Downstream of the Sierra Nevada and the Central Valley, rivers draining these regions flow into San Francisco Bay before reaching the Pacific Ocean. Clues to the climate of the Bay Area during the Neoglacial are found in sediment accumulation beneath the bay waters as well as in archaeological investigations of the shell mound villages along the coast.
The remains of several hundred mysterious mounds that once served as dwellings for a population of hunter-gatherers along the shores of San Francisco Bay have proven to be valuable—if unlikely—sources of information about Neoglacial climate in the West. Archaeologists discovered these mounds in the early twentieth century, along with their most unusual feature: they were composed almost entirely of shellfish remains.
Unlike mounds found elsewhere in North America, which were primarily specialized burial sites, the Bay Area mounds were large enough to suggest that they functioned as residences. Because they were created through the daily accumulation of household debris, the mounds contain a wealth of materials for reconstructing the environment and the climate, including the remains of fish and shellfish from the bay, the bones of terrestrial animals, and a wide array of foraged remains.
But before the significance of the mounds could be understood, archaeologists in the early twentieth century were forced into a race with urban development. One of the leading archaeologists of the time, Nels Nelson, worked heroically to map over 425 of the mounds, but he noted as early as 1909 that most of the larger mounds had been disturbed. Archaeological fieldwork in the Bay Area became largely a salvage operation, with scientists frantically digging ahead of the bulldozers that were poised to level the sites for new roads, parking lots, and businesses. During these rushed excavations, there was barely time to catalog artifacts before whisking them away for storage in museums. Except for those fragments (and a few shell mounds protected intact within state parks), the archaeological record of the earliest human inhabitants of the San Francisco Bay region has been largely erased.
Despite this enormous loss, however, the achievements of Nelson and his successors have helped us piece together a more complete picture of human life during the Holocene, including the Neoglacial. As the droughts of the mid-Holocene gave way to the wetter Neoglacial, the pace of mound-building increased, as shown by radiocarbon dating of successive shell and soil layers of the West Berkeley shell mound. The so-called “Middle Period,” from about 4,000 to 1,450 years ago, was thought to have been a “golden age” of shell mound activity, with the number of mounds around the Bay Area reaching a peak in terms of population numbers and year-round occupation.
If we visualize San Francisco Bay during that time, the curious mounds scattered around its margins were oblong—up to 500 feet long and 30 feet high—and would have been dominant features of the shoreline. We might see smoke curling up from village fires as the people rested from their foraging and prepared their meals. They had for many centuries made the marshes their home for part of each year after sea levels had stabilized 6,000 years ago, allowing the expansion of mudflats and tidal marshes. These, and the rocky intertidal zone and associated ecosystems, supported an abundance of coastal resources. The people caught fish and harvested mussels, clams, and oysters from the shores of the bay. They used rushes that grew close to the channels and in the fresher parts of the marsh for building their homes and making baskets to carry implements and food. Their drinking water was obtained from local creeks that drained the nearby hills, and they foraged riparian forests surrounding the creeks for berries, roots, and herbs. They hunted deer, rabbits, birds, and other game animals.
There were challenges to permanent habitation, however. The Neoglacial period was marked by more frequent rains and increased river flows that raised the shoreline, and storms threatened coastal settlements that were unprepared for them. The shell mound villages represented a creative response to this challenge: residents began “piling up” rather than “carving out” a living. As kitchen waste from each household, including the remains of cooking fires, was collected and deposited outside the simple huts, it became part of a continually rising ground surface. The ground and the mounds rose steadily, fast enough for the villages to keep up with the rising tides that each year reached a little higher onto the marsh. The recycling habit of the mound builders was not just good housekeeping but also an essential community service and survival tool. These mounds therefore contain a stratified archive of the lifeways, environments, and climates of the region over the past several thousand years, and they suggest that the climate was indeed wetter, with higher rates of river inflow, during the Neoglacial.
Sediments that have accumulated beneath San Francisco Bay also contain evidence for prolonged wetness and occasional extreme events during that period. Variations in freshwater inflow alter the bay’s salinity over periods of decades to centuries and affect the organisms living in the estuary—from single-celled plankton and foraminifera to mollusks (oysters, clams, and mussels). The assemblages of species living in the bay adjust in response to changes in river inflow: during higher river inflow, species that prefer lower salinity predominate, whereas during periods of decreased inflow, such as would occur during drought, species that prefer higher salinity conditions predominate. When these organisms die, their hard shell remains settle to the floor of the bay and are entombed in the sediments brought in by creeks and rivers. These accumulated sediments provide a source of environmental information for paleoclimatologists.
In addition to the identification and analyses of bay fossils, the geochemistry of these shells reflects the environmental conditions of the bay water, particularly the salinity and temperature. Such factors in turn reflect changes in the precipitation regime over the bay’s vast watershed that covers about half of the state of California today (see figure 21). This means that the chemistry of the bay water can inform us about climate over a very large region. As shellfish grow and secrete their calcium carbonate shell, they incorporate elements from the water surrounding them, including oxygen. Not all water is the same. Tiny variations in the mass of oxygen atoms (actually, isotopes) have been exploited by geochemists to provide information about past environments. River water entering San Francisco Bay through the delta is lighter than the Pacific Ocean water that enters the bay through the Golden Gate in that it has a greater proportion of oxygen-16. The differences in the amounts of fresh- and saltwater in the bay determine its salinity. The amount of river water flowing into the bay varies with the seasons, with more river water flowing during the wetter winter months. During years and decades marked by higher rainfall, the average salinity in the bay is lower than it is during years and decades of drought.
During the late 1980s and early 1990s, a research group including this book’s authors, in collaboration with other scientists researching San Francisco Bay at that time, collected and analyzed samples of bay waters during different seasons over several years, measuring the salinity and oxygen isotopes of these waters. The samples provided evidence that both the seasonal and the annual average salinity fluctuations were reflected in the proportions of oxygen-16 and oxygen-18 in the bay water. We then collected modern mussel shells growing in parts of the bay of known salinity and were able to show that the oxygen isotopes of these shells reflected the salinity of the bay.
FIGURE 21. Maps of California and the Great Basin showing many of the locations of paleoclimate records discussed in chapters 8–10. In map A the watershed of San Francisco Bay is shaded. Map B of San Francisco Bay shows locations of cores taken from beneath the estuary and surrounding marshes for paleoclimate studies. (Maps by Frances Malamud-Roam.)
These calibration studies were done to support the analyses of sediment cores taken by our research group from the central bay (see figure 21). The cores provide a record that extends back in time more than 5,000 years. Once collected from the bay, the cores were brought to the laboratory, where fossilized shells were separated in sequence from the sediments every two centimeters (three-quarters of an inch). The shells were processed, including cleaning to avoid contamination from the surrounding sediment and powdering to homogenize the samples. A tiny amount of this powder was then analyzed using a mass spectrometer to measure the proportions of oxygen-18 relative to oxygen-16.
The results of these measurements showed that the San Francisco Bay estuary was much fresher during the Neoglacial period. The decrease in salinity suggests that the average amount of river inflow was twice as high as the modern average inflow. The flows peaked between about 4,300 and 3,800 years ago, and average river flows remained relatively high for the following 2,000 years.
We were able to apply newly developed geochemical methods for assessing river inflows and salinity in San Francisco Bay to the shell mound collections. Our collaborators, archaeologists Kent Lightfoot at the University of California, Berkeley, and Ed Luby of San Francisco State University, have spent much of their careers researching the purpose of these mounds, including when and why they were occupied. They provided access to shell mound materials housed in the Hearst Museum of Anthropology on the UC Berkeley campus. The animal remains in some Bay Area mounds had been investigated to assess the season of occupation by analyzing the proportion of migratory species that were hunted by the native populations. For instance, more than half the bird remains in a mound from southern Marin County were migratory species that were present only during the fall and winter. Two large mounds—the West Berkeley and Emeryville shell mounds on the eastern shore—contained abundant chinook salmon remains, suggesting an occupation during the spring and fall migratory periods.
Our geochemical methods would allow us to assess the season of occupation in other mounds that did not contain these types of faunal remains and thus could potentially be applied to a larger number of mounds throughout the Bay Area. We decided to test our methods on two mounds located along the eastern shore of the central part of San Francisco Bay: Ellis Landing, a large shell mound close to present-day Richmond, and a small mound located on a tiny nearby island (Brooks Island).
A graduate student working in our group, Peter Schweikhardt (now a professor at the College of Alameda), developed a novel way of using fossilized shell fragments to determine the season of harvest. The results showed that shells from the Ellis Landing mound were harvested mainly during the spring and fall seasons. Some appeared to have been harvested during the summer, but none were harvested during the winter. This suggests that these mounds were used seasonally during the spring and fall but were largely abandoned during the summer and winter. The smaller mound from Brooks Island also showed seasonal occupation during the fall to early winter, though this mound appeared to have been used more during the summer as well, starting in July. The native populations apparently followed the water and other resources in the region throughout the year as the weather shifted between the cold, wet winter season and the warm, dry summer season. As we shall see in later chapters, this mobility served them well when the climate again shifted to drier, warmer conditions.
A MORE VARIABLE CLIMATE
The oxygen isotopic measurements from the San Francisco Bay sediment cores also show that the abundance of freshwater was part of an overall pattern of climate fluctuations between wetter and drier conditions occurring about every 1,500 years. This climate cycle has appeared elsewhere in the West. For instance, researchers studying Tulare Lake have determined that the lake experienced seven or eight high stands over the past 11,500 years, spaced approximately 1,500 years apart, and coastal sea surface temperatures reconstructed from oxygen isotopic measurements of foraminifera from Santa Barbara Basin sediment cores show a 1,500-year cycle, as we discussed near the end of chapter 6.
In the bay sediment cores, we also detected fluctuations that were smaller and more frequent, with 200-, 90-, and 55-year periods. We will return to these cycles in later chapters, since they appeared in a number of paleoclimate records in the West, as well as globally, and may be related to sunspot cycles and associated changes in solar radiation.
Relative to the early and mid-Holocene, the climate of the late Holocene period—beginning about 3,500 years ago—exhibited larger and more rapid changes. In other words, the climate became more variable. From the coastal ocean to the inland Great Basin, the evidence all points to a less stable climate. Along the coast, for instance, in the Santa Barbara Basin, sea surface temperatures, upwelling, and the thicknesses of the sedimentary layers show greater variability. Pollen separated from marine sediment cores taken off the Central California coast suggests that vegetation growing in those regions began to show more rapid fluctuation from wetter to drier species starting about 3,400 years ago. Inland at Pyramid Lake, in western Nevada, studies of pollen and plankton (diatoms) from lake sediment cores also exhibit an increase in climate variability after 3,500 years ago. This increased variability may be due to an increase in the frequency of El Niño events that appears to have begun at that time. A more detailed discussion about changes in the ENSO and other climate cycles will be given in chapter 11.
The wetter and cooler climate of the Neoglacial gave way to a period of drying conditions after about 1,800 years ago. Was this change favorable or unfavorable for our intrepid mound builders? As we shall see in the next chapter, there is evidence that these mound dwellings, as well as other Native American settlements in California and elsewhere in the West, were abandoned during this warmer and drier period, perhaps in response to the change in climate.