Abstract

The study and positive identification of volcanic tephra is integral to the application of tephrochronology in archaeological research. Volcanic tephra, whether proximal tephra located close to the eruption source or distal tephra located at great distances from the source, provides a chronostratigraphic marker for the date of the eruption. Where tephra appears in archaeological contexts, it can be used to help date occupation features or other strata in the sequence as well as to provide boundary markers for different archaeological time periods. Volcanic eruptions can be identified through their unique petrographic compositions and geochemical signatures. However, distal tephras frequently undergo significant post-depositional mixing and weathering that can alter these signatures. Laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) offers a method for the precise and accurate measurement of the trace element composition of microscopic volcanic glass shards, even allowing for the sampling of individual shards. This method development study employs LA-ICP-MS on individual volcanic glass shards to develop a geochemical signature for the AD 1280 Quilotoa eruption in Ecuador and to use this chemical signature to confirm or reject the presence of Quilotoa ashfall from archaeological contexts at three sites in the northern Ecuadorian highlands.

Research Objective

Highland Ecuador is one of the most volcanically active locations on earth and its frequent eruptions have shaped the nature of human occupations in the region for millennia (Hall and Mothes 2008). The AD 1280 eruption of Quilotoa volcano with a Volcanic Explosivity Index of 6 (VEI=6) was the largest eruption to occur in South America since the start of the Holocene and distributed ashfall over some 40,000 square kilometers (Mothes and Hall 2008). The eruption had a significant impact on the human populations of northern Ecuador leading to widespread abandonment and social and cultural upheaval including most notably the posited abandonment of wetland raised field (RF) agriculture and the initiation of a period of monumental pyramid construction, markers for the transition from the so called Early Integration Period to the Late Integration Period. The cooccurrence of the Quilotoa eruption and the initiation of the subsequent cultural changes has led to the suggestion that post-eruption population migration and environmental changes helped spur the cultural changes (Knapp and Mothes 1998). Some regional climate proxy records have also suggested that Quilotoa marks the end of the warm and wet Medieval Climate Anomaly (MCA) and the gradual transition into the cold and dry Little Ice Age (LIA) in Ecuador (Ledru et al. 2013, 2022; Riedinger 1993). Today, the Quioltoa ashfall is used by archaeologists as a chronostratigraphic marker for the beginning of the Late Integration Period (Mothes and Hall 1998) and based on the regional climate records, has the potential to be used as a marker for the end of the MCA.

Although petrographic analysis has been the dominant form of tephra identification in Ecuador, this method is subject to error due to post-depositional mixing. LA-ICP-MS offers the opportunity for more finite and accurate identifications. Over thirty trace or rare-earth analytes of interest can be detected through LA-ICP-MS on a single glass shard including Li, Be, B, Si, Ca, Rb, Sr, Y, Zr, Nb, Cs, Ba, La, Hf, Ta, Pb, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, Th, and U in addition to those that are normally analyzed through ICP-MS (Pearce et al. 1999). Elements such as Ca, Fe, or Si can be used as an internal standard for the analysis. Incompatible trace element/element ratios can be used to develop a chemical signature to “fingerprint” the tephra to its source eruption. In a recent paper, Bablon et al. 2022 collected tephra samples from offshore marine cores along the Ecuadorian coast and compared trace element isotopic ratios using HR-ICP-MS with proximal samples collected from near the source volcanoes. The authors developed a chemical signature for several of Ecuador’s largest eruptions over the last 10,000 years using La/Yb to Ba/Nb, Zr/YB to La/ZR, Nb to Rb/Th, Ba/La to Zr/Nb, ²⁰⁷Pb/²⁰⁴Pb to ²⁰⁶Pb/²⁰⁴Pb, ⁸⁷Sr/⁸⁶Sr, and ²⁰⁷Pb/²⁰⁴Pb to ⁸⁷Sr/⁸⁶Sr, isotopic ratios that vary based on the north to south axis of the volcanic arc and the eastern and western cordilleras. They have also published their results in a publicly available online database that was removed from embargo on September 30, 2022. For this project, I will develop a method using LA-ICP-MS on several samples of distal tephra that have been tentatively identified as Quilotoa as well as on a proximal Quilotoa control sample in order to develop a chemical signature for the eruption and test the accuracy of the in-field or in-lab petrographic identifications. I will compare these results to those published in the Bablon et al. (2022) online database for Quilotoa and Cayambe tephras as a secondary control.

The main goals of this proposed method development study are threefold:

1. To develop a geochemical signature for the Quilotoa eruption.
2. To test a method for using LA-ICP-MS on distal tephras in Ecuador.
3. To confirm or reject archaeological interpretations based on the presence of Quilotoa tephra.

Justification

Although Bablon et al. (2022) developed a signature for the AD 1280 Quilotoa using proximal tephras, they were unable to detect distal Quilotoa tephra within their cores. To date, no geochemical signature using LA-ICP-MS has been used to identify distal Quilotoa tephra in Ecuador and the technique has yet to be employed in archaeological investigations. This proposed study will be the first use of LA-ICP-MS to identify tephra in archaeological contexts in Ecuador and will contribute to the development of a geochemical signature for the Quilotoa eruption. As discussed above, the proper identification of Quilotoa ashfall is vital to the interpretation of the Ecuadorian archaeological record at the boundary between the end of the Early Integration Period and the beginning of the Late Integration Period as well as the MCA/LIA transition. While Quilotoa ashfall is found through much of Ecuador, smaller eruptions around the same time with more localized impacts can make it difficult to date with certainty archaeological contexts using tephrochronology or to assess the impact of the Quilotoa eruption in these locations. Particularly, an eruption of Cotopaxi volcano around AD 1231 (VEI=5) and eruptions of Cayambe volcano around AD 1040 (VEI=4), AD 1290 (unknown VEI), and AD 1485 (VEI=4) can cast doubts on the accuracy of tephrochronological interpretations made with conventional visual identification at sites near these volcanos.

Three study sites were chosen due to their archaeological features and association with the Quilotoa eruption (Figure 1). The archaeological sites of Cochasquí, La Vega, and Zuleta have Early Integration and Late Integration components including Early Integration hemispherical burial mounds and RFs in addition to Late Integration quadrangular pyramids. Cochasquí and Zuleta are two of the largest earthen mound sites in Ecuador representing some of the largest earth-moving efforts in pre-Hispanic times. La Vega has some of the most extensive and well-preserved RF agricultural features. Much of the chronology of the archaeological record in the northern highlands of Ecuador is based on research that has been conducted at these sites and the relationship between archaeological features at the sites and the Quilotoa ashfall. Yet, their proximity to Cayambe volcano raises questions about the tephras present and accuracy of past interpretations made with visual tephra identification.

FIGURE 1: Study sites, source volcanos, and ashfall extents.

Literature Review

The methods utilized in this study are based primarily on those developed by Westgate et al. (1994), Pearce et al. (1996; 1999; 2004; 2007), and Westgate and Pearce (2017). Westgate et al. (1994) were the first to employ trace element analysis of volcanic tephra using LA-ICP-MS. Westgate et al. compared the viability of this method with the analysis of bulk samples using INAA and XRF methods that were popular at the time noting that these analyses can’t account for the degree of post-depositional mixing and weathering that occurs. Their early process relied on the extraction of 0.01 to 0.05 grams of volcanic glass shards mounted on a card to form a pile a few millimeters high. Although Westgate et al. were only able to produce accurate results within 15% for most elements and the process depended on the need to determine a trace elemental standard to overcome matrix effects from differences in ablation, they noted that the level of accuracy was still very acceptable in tephra studies and that the process was a first step away from the traditional bulk analyses which are more susceptible to contamination towards an idealized single grain analysis.

Pearce et al. (1996, 1999, 2004, 2007) built on the success of Westgate et al. (1994) using advancements in laser technology from the 1064 nm laser used by Westgate et al. to employ a single grain analysis using a 266 nm laser. Pearce et al. also developed a method for using a minor isotope of a major element as an internal standard rather than needing to calibrate an internal standard using an external standard. Pearce et al. (2007) note that laser ablation of single grains requires low background noise with optimal sensitivity and needs to be set to a sufficiently high signal to ablate the glass shard without destroying it too rapidly. Rhyolitic glass shards tend to be shatter under the laser and need to be moved constantly. However, the count rate for an element appears to vary dramatically when the laser is static compared to when it is in motion due to preferential volatilization of some elements. Rastering the laser across the surface of the shard prevents this change in detection from occurring.

More recently, Pearce (2014) and Westgate and Pearce (2017) have detailed methods utilizing a higher frequency 193 nm Excimer laser that is more suitable for smaller glass shards.

Methods

Tephra samples that will be utilized for testing this method include three control samples and thirteen unknown samples. One control sample (n=1) was collected from the Q-I tephra proximally located at the base of Quilotoa volcano and two control samples (n=2) from Cayambe volcano. All control samples were identified and collected by volcanologists Patricia Mothes and Minard Hall. Unknown tephra samples were collected from Cochasquí (n=2), La Vega (n=1), and Zuleta (n=18) during various archaeological investigations carried out between 2019 and 2022. Mothes and Hall tentatively identified two of the unknown samples from Zuleta as Quilotoa AD 1280 and one unknown sample from Zuleta as Cayambe AD 1040 based on their mineralogical constituents. Two additional unknown samples from Zuleta were tentatively identified by archaeologists as Cayambe AD 1040 and AD 1485 based on their stratigraphic relationship to the AD 1280 Quilotoa ashfall.

Sample preparation and instrumentation for this study will follow that outlined in Westgate and Pearce (2017). Samples will be sieved at 100 μm with the both the large and small fractions being retained as the active samples. Samples will then be placed in an ultrasonic bath for 10 minutes to remove fragile pumiceous glass. Both fractions will be mounted on double sided tape in preparation for laser ablation. Laser ablation will be conducted using a 193 nm Excimer laser and analyzed with an Agilent 7500ce quadrupole ICP-MS. Larger shards will be analyzed using a 20 nm beam and smaller shards using a 10 nm beam. Laser energy will be pulsed at 10 J/cm² at 5 Hz using an acquisition time of 24s. NIST 612 will be used as an external calibration standard in conjunction with ²⁹Si as an internal standard.

Expected Results

Laser ablation allows for the analysis of individual glass shards that are representative of the magmatic source of the eruption. I expect that La/Yb to Ba/Nb, Zr/YB to La/ZR, Nb to Rb/Th, Ba/La to Zr/Nb, ²⁰⁷Pb/²⁰⁴Pb to ²⁰⁶Pb/²⁰⁴Pb, ⁸⁷SR/⁸⁶Sr, and ²⁰⁷Pb/²⁰⁴Pb to ⁸⁷Sr/⁸⁶Sr values for the unknown samples tentatively identified as Quilotoa will match those of the control sample while those of the samples identified as Cayambe will produce a different signature. These values should match the expected values based on the magmatic source of the eruption which vary based on location along the northern volcanic arc. Volcanos along the southern half of the arc and the western cordillera such as Quilotoa have high ²⁰⁶Pb/²⁰⁴Pb and ²⁰⁷Pb/²⁰⁴Pb ratios while volcanos such as Cayambe in the northern half of the arc and the eastern cordillera have the highest ²⁰⁶Pb/²⁰⁴Pb ratios, the lowest ⁸⁷Sr/⁸⁶Sr ratios, and the highest ¹⁴³Nd/¹⁴⁴Nd ratios (Babon et al. 2022). Typical analytical precision of 10% should be appropriate for La-ICP-MS of volcanic glass shards (Pearce 2014).

Plan of Work

Samples will be prepared with the aim of being completed by November 12, 2022. The analysis will take place over the course of three weeks from November 13 to December 1. An estimated four hours of analysis time will be necessary at $73 per hour for a total of $286.