Late Quaternary Environments of the Teton Mountains , Wyoming : A Pollen Record from Green Lake

Late Quaternary forest succession in Wyoming’s Rocky Mountains, occurred in random patterns, because it reflects differences between the Glacial vegetation at lower elevations on the east vs. the west of the Rockies, as well as along the mountain crest to the south. Differential melting of mountain glaciers resulted in differences in the timing of recolonization. Significant variations in the composition of plant assemblages occurred due to delays in species’ arrival, and even in the exclusion of species. Holocene climate variability, especially ongoing global warming, added to the complex dynamics of plant assemblages with warm climate species replacing early Holocene, cooler climate species. The pollen record from Green Lake (located in a glacial cirque on the west side of the Teton Mountains in Teton County, Wyoming) addresses the local vegetation response from just before the fall of Mazama ash to the end of the middle Holocene warm period. Although the earlier portion of the pollen sequence records some of the last adjustments as some plant species were still arriving after de-glaciation, by the time Mazama ash fell it was climate variation that determined most of the dynamics observed in the Green Lake record. The results reveal a sequence of wetter and drier periods based upon the presence of diagnostic tree species. A moist late early Holocene was followed by a dry middle Holocene, which ended about 6,400 cal. B.P., and was followed between 5,000 to 2,800 cal. B.P. by a sequence of drier and moister climate episodes.


Introduction
In order to track the changes in the late Quaternary history of high elevation lakes, we were examined the record from Green Lake (43.83 N, 110.90 W, 2707 m), which lies on the west slope of the Teton Mountains in Teton County, Wyoming (Figures 1, 2).Microfossils are the most critical group of fossils to interpret the climate and environmental condition (Asgharian Rostami et al., 2013 and2018;Frontalini et al., 2016;Martin et al., 2017;Rodelli et al., 2018;Amao et al., 2018;Balmaki et al., 2012).Here we are focusing our analyses upon the pollen recovered from the middle Holocene portion of this record, but will eventually investigate the complete pollen record, as well as the history of the algae, and charcoal deposited in its sediment in response to climate change.The record from Green Lake in the Grand Teton National Park spans at least 17,000 years and will provide an excellent record of local as well as region vegetation change as post glacial plant community succession and response to climate change occurred.

Previous Research
The publication of R. G. Baker's (1976) description of the late Quaternary vegetation history of the Yellwstone Lake Basin initiated a deluge of paleoenvironmental studies including but not limited to investigation of late Pleistocene forest dynamics after de-glaciation, forest response to climate variations, reconstruction of fire regimes as they related to both to climate change and forest community structure, and the impact of these processes on lake chemistry, and productivity.Much of the research in the Yellowstone -Teton Mountains area of western Wyoming has been conducted by Cathy Whitlock with her students, and colleagues (Whitlock 1993;Whitlock & Bartlein 1993;Whitlock & Sherriff 1996;Whitlock et al. 2008;Huerta, Whitlock & Yale 2009;Whitlock et al. 2012;Mumma, Whitlock & Pierce 2012;Krause & Whitlock 2013;Krause et al. 2015).However, except for Hedrick Pond, Lily Lake and Fen, Divide Lake, Fallback Lake and Emerald Lake most sites that have been investigated lie within or north of Yellowstone National Park.All of these lakes lie to the east across the valley from the Teton Mountains and below the elevation of Green Lake.Although most of these studies have dealt with forest and fire history Krause et al. 2015 includes a study of the diatom flora of Dailey Lake north of Yellowstone National Park.
Figure 1. Green Lake location in the Teton Mountains of western Wyoming Figure 2. Green Lake viewed from the east (Blake Foreshee photo, 10 September 2015)

Geology
Stretching and thinning of the crust about six to nine million years ago resulted in movement along the Teton fault.The west block rose to form the Teton Range, the youngest of the Rocky Mountain ranges.The oldest rocks in the Teton range began forming about 2.5 billion years ago when sand and volcanic debris settled onto the floor of an ancient ocean.Additional deposition was followed by metamorphism into gneiss.Silica-rich magma forced up through cracks in the gneiss formed granite sills anywhere from a few centimeters to hundreds of meters thick.Dikes of igneous diabase formed in cracks.From the Cambrian into the Cenozoic deep marine sediments were deposited over the metamorphic basement.Erosion and ongoing uplift have exposed the metamorphic and intrusive igneous rocks, and as late as 2.1 million years ago volcanic tuffs were deposited on the northwest slope of the range.The EPA has classified the ecoregions of the high Tetons in categories 17G Mid-Elevation Sedimentary Mountains and 17H Alpine Zone.The major vegetation types described around Green Lake include 1) Limestone Pavement Sparse Vegetation in areas scoured by glaciation and steep slopes, 2) Subalpine Mixed Herbaceous Vegetation at higher elevations above the lake; and 3) Whitebark Pine Forest and 4) Subalpine Fir -Engelmann Spruce Forest either separately, or mixed around the lake depending upon slope, aspect, exposure and soil depth.

Climate
The climate in the Tetons is characterized by mild summers and cold winters.Although most precipitation occurs in the winter, there is also a greater abundance during the months of April, May and June (Figure 3).This pattern occurs in a band across the northern Great Basin into northern California.

Core Recovery
Between the 3 rd and 10 th of September 2015, we and a team of volunteers, recovered five 2-inch diameter sediment cores from Green Lake located on the western side of Mt Moran.Working from a portable 6X6 foot raft, that we had packed into the lake and assembled there and anchored to the shore, 11 drives were taken using a modified Livingston piston coring device.The cores were transported out of the lake with pack horses and transoirted to the cold room of the Paleoecology Laboratory in the Geography Department of the University of Nevada, Reno.

Core Dating and Correlation
The cores were first opened in the Palaeoecology lab at the University of Nevada, Reno, and first sampled for AMS dating to establish a chronology.It was clear that two volcanic ashes in the lower third of the record would also provide additional means of both dating and correlating the cores (Figure 4).However, absence of strata that could be correlated in the upper cores necessitated the submission of additional samples for AMS dating to Beta Analytic in Miami, Florida to establish a reliable chronology.Correlation of the cores and their dating is shown in Figure 5.
The correlations shown in Figure 5 using both volcanic ashes and AMS dates, indicate that we have a potential record to 17,000 cal.B.P. The volcanic ashes were identified by the Peter Hooper GeoAnalytical Laboratory at Washington State University in Pullman, Washington.The upper volcanic ash was identified as the Mazama Plinian volcanic ash (8,400 cal.BP).A recent Bayesian modeled age of the Plinian Mazama eruption as dating between 7743 -7543 cal.B.P. (Egan et al 2015).Based upon the mapped spatial distributions of the Glacier Peak B and G volcanic ashes the Glacier Peak ash in the core is most likely Glacier Peak B volcanic ash.(Kuehn et al 2009) (Figure 6).The probability distribution for the modeled age of Glacier Peak tephra indicates a mean age of between 13,710 to 13,410 cal.B.P. (Kuehn et al. 2009).
Seven AMS dates were submitted to Beta Analytic for analysis.The results indicate that the record from Green Lake spans the period from the end of the last Glacial Maximum through about 2700 cal.B.P. above which sediments were too unconsolidated for recovery (Table 2). .Correlation of the core drives from Green Lake.Three drives have "????" at their tops indicating that we do not know the exact age assignment of these cores.These may contain some of the upper portion of the record from Green Lake that is currently missing

Sampling
A series of 47 samples at 4 cm intervals were removed from the cores for analysis of pollen, charcoal, diatoms, algae, grain size, and geochemistry.Of these 30 samples were selected for analysis (Table 3) (Figure 7).

Pollen Extraction
Processing of samples for pollen, and selected spores began with placing 1 teaspoon (4.93 cc) of each sample into a 400 ml plastic beaker with triple distilled water.For statistical purposes two Lycopodium tracer spore tablets (batch #124961 with 12,542 +/-414 spores per tablet) were added to each of the 30 samples.This equaled 25,084+/-828 tracer spores per sample.
In order to ensure the preservation of diatoms, the extraction procedure was modified to remove those steps that might be more damaging to both diatoms and pollen that might be poorly preserved the pollen during processing.
The modified procedure was as follows.
Samples were rinsed through 100-mesh (100 micron) screen using distilled water and transferred to 40-ml test tubes.The samples were then treated with a solution of sodium polytungstate Na 6 [H 2 W 12 O 40 ] with a specific gravity of 2.1.This floated and concentrated all pollen, acid-resistant algae, insect parts, etc. into the upper portion of the tube.The float was decanted into a new test tube and concentrated using three to four distilled water washes with centrifuges and decants after each.
There followed numerous hot distilled water washes to remove suspended clays, charcoal and colloidal materials that were also floated during the sodium polytungstate procedure.As with the standard extraction procedure, the samples were dried with two treatments of ETOH alcohol (the first a 95% concentration and the second a 100% concentration) after the water washes.Samples taken from the Green Lake cores are marked in yellow.Those extracted and counted for this analysis are marked in blue.They are arranged by depth within each drive, and their calibrated age B.P. is assigned as well based upon the dates that we obtained from the cores.stained with safranin during the first alcohol treatment, and the samples were transferred to vials with Tertiary Butyl alcohol and completed with the addition of a mounting medium (2000 cs silicone oil).Finally, the Tertiary Butyl alcohol was evaporated from the samples in a low temperature oven for a several hours.Pollen samples were mounted on glass slides and at least 41 contiguous rows (all rows on the slide) were counted for each slide (in the case of the first sample only 21 rows were counted).This was equivalent to between 250 and 300 pollen grains per slide.Identification of certain unknown pollen types was assisted by the use of a reference slide collection, several online pollen floras (linked to from a website at the University of Arizona: http://www.geo.arizona.edu/palynology/pol_pix.html), and several pollen photomicrograph publications of Southwestern pollen types that were helpful for types that seem to be intrusive from the Southwest (Martin andDrew 1969 and1970;Solomon et al 1973).
Table 3.Samples taken from the Green Lake cores are marked in yellow.Those extracted and counted for this analysis are marked in blue.They are arranged by depth within each drive, and their calibrated age B.P. is assigned as well based upon the dates that we obtained from the cores

Results
The pollen percentages are plotted based upon two pollen totals (Figure 7).Pollen types in green are plotted as a percentage of total pollen.Pollen types in magenta are plotted as percentage of the sum of those types in the magenta with Abies, Pinus and Artemisia outside the sum.This was done to observe the variation in just those types.Pollen type diversity at Green Lake is low.Pollen from Green Lake was dominated by three types Abies (fir), and Pinus (pine).These types occur in abundance around the lake today.Picea (spruce) was not found in the record.It tends to be a very large pollen grain and not well represented in the record unless it is nearby, and due to its large size, it must be close to the edge of the lake to appear in the record with any abundance.
Pollen of Artemisia (sagebrush) of several species ranked third in abundance.Of the rarer pollen types appearing in the record Chenopodiaceae (saltbush-type) was common.This pollen type occurred regularly in the record.More sporadic, but at times abundant, were Aster-type (sunflower) and oak (Quercus) pollen.These were usually out of phase with each other although occasionally not.Pollen recovery statistics are recorded in Table 4. Abundance per cm 3 is plotted both for total pollen and for the major pollen types.Pollen recovery statistics are recorded in Table 3. Abundance per cm 3 is plotted both for total pollen and for the major pollen types.

Pollen Record
he presence of just a few types in the pollen record is typical of the mixed conifer woodland in which one or two species dominate.This explains the dominance of pine (probably white-bark pine) and fir (probably subalpine fir) pollen.Sagebrush may be a mix of both high and low elevation species.Big sagebrush pollen as well as mountain big sagebrush pollen are probably both present in the pollen counts and represent local sagebrush in the forest understory and the sagebrush steppe at low elevations to the west.Chenopodiaceae (saltbush pollen), and oak pollen is probably blowing up from the Snake River Plains and the hills to the west as well.The low pollen influx estimates into Green Lake reflect the low density of the woodland surrounding Green Lake.When viewed on Google Earth, the woodland is clearly very diffuse (Figure 8).A rough estimate suggests that there are considerably less than 1000 trees in the immediate area of the lake.This could explain the sparse pollen in the record overall.Occasional gaps in the distribution of some of the pollen types during periods when we have a low pollen population estimate suggests a brief episode of more rapid deposition rate.This may reflect a runoff or a rapid snow melt event which carried sediment into the lake from above.Pollen influx would be lower, because more sediment was being deposited than pollen during such episodes.
However, there are also episodes when pine is almost absent, but fir is abundant.These may be brief episodes of much cooler, wetter climate.Remember our samples are roughly 50-year snap shots taken every 200 years.Principal component analysis of the major pollen types (Figure 7) indicates pollen types that generally act in concert in the Green Lake pollen record.Fir clearly appears to respond differently than pine and even sagebrush.Saltbush and sunflower-type pollen usually act in concert, as do pollen from sagebrush, pine and oak.
In describing plant species response, fir may be responding to episodes of colder, though not necessarily wetter climate.Pine, sagebrush and oak all respond to increased winter precipitation, whereas saltbush and sunflower pollen may be responding to warmer drier climates at lower elevations and perhaps increased summer precipitation as well.
When pollen assemblages are plotted by depth clear patterns of warm dry and cool wet climate are revealed (Figure 7, 11).In addition, occasional episodes of much cooler climate also seem apparent.These seem to mirror records from other pollen records that indicate a cool-moist late early Holocene, a warm middle Holocene, and a break to cooler, moister conditions after 6,400 cal B.P. Episodes of drier conditions punctuated by a wetter episode at about 3,600 cal.B.P. reflects other records from further west in the northern and central Great Basin.

Climate
Snow Fall -Ice Patch Archaeology -The question of the relationship of use of high elevations by people and the absence of snow is more difficult to address.Prior to the last 1,800 years there are too few dates in the lower 48 states to relate periods of relatively warmer climate with periods when people seem to have frequented higher elevations (Figure 11).annual mean temperature and mean annual snowfall derived from the Bryson meso-scale model (Bryson and Bryson 1997;Bryson and DeWall 2007), were plotted against cal.B.P. dates on artifacts from snow patches (Reckin 2013; Lee 2012; Sgouros and Stirn 2015) (Figure 11) pollen derived climate zones.There appears to be only slight correspondence to warm episodes and use of high elevations.The problem may lie in the fact that palaeoclimatic records lack the resolution to resolve brief episodes of warmer climate.The pollen record sample interval is about 200 years, even the MCM reconstruction interval is 100 years.This is much to course to resolve the climate human use of high elevations issue.The MCM model clearly demonstrates that there are periods of warmer and colder climate that have high frequency variation.
The problem is that mobile artifact occurrence may not be related to episodes of warmer climate at all, but only to seasonal use.This is quite different from the Alta Toquima village excavated by Dave Thomas (1982), which indicated the presence of resource utilizing villages at high elevation after about 2,000 years ago.

Conclusions
An initial investigation of pollen from Green Lake, Teton County, Wyoming reveal a sequence of wetter and drier episodes, based upon the groupings of specific pollen types (Figure 7).In general, these episodes correlate to other records in the region and the western U.S. A moist late early Holocene was followed by a dry middle Holocene which ended about 6,400 cal.B.P. episodes of drier and moister climate characterize the period between 5,000 and 2,800 cal.B.P.
Although some have suggested clear correspondence between longer-term episodes of warmer climate and artifacts recovered from beneath melting ice patches, the data are still inconclusive.Episodes of higher elevation resource utilization could just as easily be explained by change in the groups utilizing the area, and differences in their life ways.We see replacement of peoples along the central California coast at least twice during the last 4,500 years.

Figure 3 .
Figure 3. Monthly average distribution of precipitation and temperature in the Tetons

Figure 4 .
Figure 4. A: Cores laid out in the laboratory for correlation and sampling for volcanic ash, AMS and pollen.In the open core in the photo on the left sample vials are set next to Mazama ash (at the top) and Glacier G volcanic ash below), B: Mazama ash, C: Glacier G volcanic ash

Figure 6 .
Figure 6.Spatial distribution of the Glacier Peak B and G volcanic ashes with the location of Green Lake (from Kuehn et al 2009)

Figure 8 .
Figure 8.View of Green Lake indicating the scarcity of trees in the vicinity of the lake

Figure 9 .
Figure 9. Principle component analysis showing the associations of the major pollen types, and their climates

Table 1 .
Grand Teton Climate Monthly Means

Table 2
. AMS dates from Green Lake Core and Sample Information Sample #s Radiocarbon Age Age cal B.P.

Table 4 .
Green Lake pollen statistic