Climatic backdrop for Pueblo cultural development in the southwestern United States

While climatic triggers for collapse and population migrations of ancestral Pueblo communities have been proposed, little is known about the overall climatic backdrop for the entire pre-Hispanic Pueblo period (ca. 1300 to 460 B2K). Here, we report data from stalagmite HC-1, from Hidden Cave, Guadalupe Mountains, New Mexico, covering the past 3400 years, showing an interval of increased frequency of droughts from 1260 to 370 yr B2K that is coeval with the entire pre-Hispanic Pueblo period. Our record suggests that this puebloan Late Holocene climatic interval was the most arid and highly variable climatic period of the last 3400 years. Climatic conditions favoring the introduction of cultivation existed prior to the Pueblo period during more pluvial-like conditions from at least 3400 to 1260 yr B2K. Hence, the change from the Desert Archaic/Basketmaker to Pueblo cultures was associated with a quick transition to increasing aridity into and through the Pueblo period associated with greater urbanization and the establishment of pueblo population centers.


S3. Aragonite as an indicator of dry climate in stalagmite HC-1
Stalagmite HC-1 consists almost entirely of calcite growth, but exhibits numerous aragonite layers and growth hiatuses near the stalagmite top. Polyak and Asmerom (2001) 2 noted aragonite that formed on a tilted billet near a drip site in Hidden Cave (Fig. S3). Where water directly dripped onto billets lying flat under drips, calcite formed; however, on the tilted billet, aragonite formed. This demonstrated that the very thin water films on the tilted billet were prone to evaporative conditions in Hidden Cave. Aragonite layers are commonly observed on the sides of stalagmite HC-1 and other samples from this cave. In addition, a drip study in Carlsbad Cavern showed that Mg concentration in drip water increased during winter months when drip rates slowed down 3 . This suggests that the drier conditions during winter produced drip water with higher cation concentrations. The Mg/Ca ratio and CO 3 concentration in the drip water determine the mole% MgCO 3 in the cave calcite 4 . Hidden Cave speleothem calcite has 2.3 to 4.0 mole% MgCO 3 5 , indicating that drip water in this cave contains abundant Mg, derived from the dolostones in which the cave formed. Lippmann (1973) 6 showed that Mg inhibits nucleation of calcite, which can promote precipitation of aragonite. Hidden Cave stalagmites contain both calcite and aragonite laminae, showing that drip water in this cave has suitably high enough concentrations of Mg to promote precipitation of aragonite, a result that was experimentally confirmed 7 . The tilted billet on which aragonite formed represented an environment where changes to thinner water films were more susceptible to the effects of evaporation that lead to higher Mg concentrations and aragonite rather than calcite precipitation. This is also illustrated on the sides of stalagmites where aragonite tends to precipitate. Accordingly, climatic changes that lead to slightly lower relative humidity in the cave and to slower drip rates can result in deposition of aragonite rather than calcite at the tops of actively growing stalagmites. Prior calcite/aragonite precipitation as stalactite growth would also be expected during periods of lower relative humidity and slower drip rates 8 . These more evaporative cave conditions that promote aragonite precipitation over calcite in Hidden Cave are interpreted to be driven by drier surface climate.

S4. Correction of aragonite δ 13 C and δ 18 O values for fractionation differences
We did not correct the aragonite stable isotope values for fractionation that were used in the timeseries of Figure 3 and Figure S5, because we are not reporting quantitative results from our stable isotope values. Sletten et al. (2013) 9 used a routine that corrected aragonite δ 13 C and δ 18 O values where the measured δ 13 C and δ 18 O of 100% aragonite needed to be corrected by subtracting 1.7 ‰ and 0.8 ‰, respectively, from the measured values. Lachniet (2014) 10 show from the fractionation equations of Kim et al. (2007) 11 that the δ 18 O value correction is 0.38 rather than 0.8‰. In Table S2, we provide a time-series of % aragonite for which the aragonite δ 13 C and δ 18 O values can be corrected using the 1.7 and 0.38 ‰ corrections relative to a linear equation that incorporates the % aragonite. We show that these differences are not substantial in Figure 4A, B.

S5. Climate versus culture
The cultural history of our study area back to 4000 yr B2K using the Pecos classification is best represented by the Late Archaic, Basketmaker, and Pueblo cultures 12 . The Late Archaic and early Basketmaker intervals in the study area are also referred to as the Late Archaic/Early Agricultural period 13,14 . During the Late Archaic to Pueblo transition, societies changed in the SW USA from mostly hunter/gatherers (Archaic) to basket makers that implemented agriculture and pit houses (Basketmaker) to remarkable communities of rock building dwellers who utilized ceramics (Pueblo) 15 .
The Basketmaker and Pueblo cultures are well-defined chronologically as late desert Archaic and Basketmaker II (3200 to 1500 yr B2K) 13 , Basketmaker III (1500 to 1300 yr B2K), Pueblo I (1300 to 1110 yr B2K), Pueblo II (1110 to 855 yr B2K), Pueblo III (855 to 715 yr B2K), and Pueblo IV (715 to 460 yr B2K) [16][17][18] . The definition of Pueblo IV 16 differs up to 200 years in its length (ends at 600 yr B2K rather than 460 or 400 yr B2K) compared to those defined in the earlier literature, and other pre-Hispanic Pueblo period boundaries differ slightly depending on the source. To the east of our study area on the southern high plains, archaic cultures lacked materials to build structures that survived weathering, and therefore age models tied to these cultural traditions are less distinctly interpreted and largely rely on identification of pottery shards, projectile points, and other small lithic artifacts. In the greater Southwest, pueblo structures and pottery types are distinct, fascinating, and well-characterized.

S6. δ 13 C and δ 18 O time-series correlations with other Northern Hemisphere records
Stalagmite HC-1 exhibits only two distinct brief aragonite-defined droughts during the 3300 to 1040 yr B2k pluvial period, an interval that can be compared to other similar high-resolution records [19][20][21] Figure S4. Stalagmite HC-1 grayscale time-series shows the character of growth that is used to suggest three sub-periods, pluvial, puebloan, and return to pluvial, of the Late Holocene period. From the beginning of stalagmite HC-1 growth up to 1260 yr B2K, only two aragonite-defined droughts occur during pluvial Late Holocene sub-period.