DWQ-2024-004891

1

Utah Lake Paleoecology Study
Report to DEQ, April 2024

Brahney, J., Powers, M., King, L., Devey, M., Carter, M., Carling, G., Brothers, S., Provard, A., Young, B.,
West, R.

2

Table of Contents
Introduction ..................................................................................................................................................... 5
Core collection, chronology, and core descriptions............................................................................................. 6
Charge Question 1.1 What do the diatom community and macrophyte community in the paleo record tell us
about the historical trophic state and nutrient regime of the lake? ................................................................... 11
Diatoms ..................................................................................................................................................................... 11
Macrophytes ............................................................................................................................................................. 15
Charge Question 1.2. What were the historic phosphorus, nitrogen, and silicon concentrations as depicted by
sediment cores? (add calcium, iron, and potentially N and P isotopes) .............................................................. 21
Phosphorus ............................................................................................................................................................... 21
Nitrogen and carbon mass and isotopes .................................................................................................................. 31
Elemental Composition ............................................................................................................................................ 32
Charge Question 1.4. What do photopigments and DNA in the paleo record tell us about the historical water
quality, trophic state, and nutrient regime of the lake? ..................................................................................... 35
Reference Condition of Utah Lake (Pre-Disturbance) ......................................................................................... 43
1. Large areas of Utah Lake had substantially more macrophyte cover than present ........................................ 43
2.0 Large areas of Utah Lake were less turbid and allowed for greater light penetration to the sediment
environment ............................................................................................................................................................. 45
3.0 The present-day hypereutrophic condition combined with toxic cyanobacteria blooms is unprecedented in
the history of Utah Lake ........................................................................................................................................... 45

3

Table of Figures

Figure 1 Core Collection sites in 2018, 2019, and 2020. ................................................................................................ 6
Figure 2 Concentrations of 210Pb measured (a) throughout the Bird Island (green) and Provo Bay (purple) cores, and
concentrations of 137Cs measured (b) throughout the Goshen Bay (blue) and Provo Bay cores that were used to
develop the respective age-depth models for each sediment core (c), with the shaded areas indicating the
respective 95% confidence intervals. Reproduced from Devey 2022 and King et al. 2024. .......................................... 9
Figure 3. Loss on ignition (LOI) results for % organic matter as well as macrofossil remains recovered for all four
coring locations. ........................................................................................................................................................... 10
Figure 4. Diatom species from Goshen Bay plotted as relative concentration by depth. Red dashed lines show the
significant shifts based on STARS analyses. .................................................................................................................. 12
Figure 5. Diatom from Goshen Bay con’t. .................................................................................................................... 12
Figure 6. Diatoms from Goshen Bay con’t. ................................................................................................................... 13
Figure 7. Diatoms from Goshen Bay con’t. ................................................................................................................... 13
Figure 8 Diatom species from Bird Island plotted as relative concentration by depth. Red dashed lines show the
significant shifts based on STARS analyses. .................................................................................................................. 14
Figure 9. Diatoms from Bird Island con’t. ..................................................................................................................... 14
Figure 10. Diatoms from Bird Island con’t. ................................................................................................................... 15
Figure 11. Diatoms from Bird Island con’t. ................................................................................................................... 15
Figure 12. . Environmental DNA results for Hardstem bulrush and Cyanobacteria from the Goshen Bay core. ........ 20
Figure 13. Population growth within Utah Lake valley (a; World Population Review, 2022) shown alongside profiles
of organic matter (OM %, cutoff length = 40), carbon (C %), nitrogen (N %), C:N ratios, δ13C (‰), δ15N (‰), and
Hydrogen Index (HI, mg HC g) ...................................................................................................................................... 20
Figure 14. Sequential Phosphors Extractions for Goshen Bay. .................................................................................... 25
Figure 15. Sequential Phosphors Extractions for Bird Island........................................................................................ 25
Figure 16. Sequential Phosphors Extractions for Provo Bay. ....................................................................................... 26
Figure 17. Sequential Phosphors Extractions for the North core. ................................................................................ 26
Figure 18. SEM-EDS false color image of elemental abundance within the surface sediment sample collected from
Goshen Bay, Utah Lake. Note, although carbon was removed from the map summary spectrum for quantification,
it was not possible to remove carbon. ......................................................................................................................... 27
Figure 19. Panel A: Population of Utah Valley (World Population Review, 2022) with the covered wagon indicating
European Settlement in 1849 (Janetski, 1990), the fish indicating Carp introduction in ~1883 (Heckmann et al.
1981), and the spigot indicating the implementation of secondary sewage treatment in ~1950. Panel B: The
concentration of phosphorus in physically and chemically separated authigenic calcite grains, Panel C:
accumulation rates of P bound in authigenic calcite, and Panel D: SEM-EDS determined mean concentrations of P
from individual authigenic CaCO3 grains, error bars represent the standard error. E: Total Phosphorus (TP). Dashed
lines indicate significant (p < 0.1) shifts in the mean. .................................................................................................. 28
Figure 19. Panel A: Population of Utah Valley (World Population Review, 2022) with the covered wagon indicating
European Settlement in 1849 (Janetski, 1990), the fish indicating Carp introduction in ~1883 (Heckmann et al.
1981), and the spigot indicating the implementation of secondary sewage treatment in ~1950. Panel B: The
concentration of phosphorus in physically and chemically separated authigenic calcite grains, Panel C:
accumulation rates of P bound in authigenic calcite, and Panel D: SEM-EDS determined mean concentrations of P
4

from individual authigenic CaCO3 grains, error bars represent the standard error. E: Total Phosphorus (TP). Dashed
lines indicate significant (p < 0.1) shifts in the mean. ..................................................... Error! Bookmark not defined.
Figure 20. Regression analyses of calcite-P concentrations against proxies of nutrient loading (Population, 15N)
and water column productivity (sedimentary Canthaxanthin) using the Modified Sequential Extraction and ICP-MS
approach (left) and the SEM--EDS Analysis on Authigenic CaCO3 approach (right). .................................................... 29
Figure 20. Regression analyses of calcite-P concentrations against proxies of nutrient loading (Population, 15N)
and water column productivity (sedimentary Canthaxanthin) using the Modified Sequential Extraction and ICP-MS
approach (left) and the SEM--EDS Analysis on Authigenic CaCO3 approach (right). ....... Error! Bookmark not defined.
Figure 21. Elemental concentrations form the Goshen Bay core. ............................................................................... 33
Figure 21. Elemental concentrations form the Goshen Bay core. .................................. Error! Bookmark not defined.
Figure 22. Elemental concentrations form the Goshen Bay core. ............................................................................... 34
Figure 22. Elemental concentrations form the Goshen Bay core. .................................. Error! Bookmark not defined.
Figure 23. Concentrations of photosynthetic pigments (nmoles pigment g-1 OM), as well as the degradation ratio
(Chl a:Pheo a), measured throughout the Goshen Bay, Provo Bay core, and Bird Island cores. Multivariate BCP
analysis estimated the posterior probability of a change point throughout the cores. Dashed lines indicate
significant abrupt changes as detected by STARS. Figure reproduced from King et al. 2024. ..................................... 36
Figure 23. Concentrations of photosynthetic pigments (nmoles pigment g-1 OM), as well as the degradation ratio
(Chl a:Pheo a), measured throughout the Goshen Bay, Provo Bay core, and Bird Island cores. Multivariate BCP
analysis estimated the posterior probability of a change point throughout the cores. Dashed lines indicate
significant abrupt changes as detected by STARS. Figure reproduced from King et al. 2024. .......Error! Bookmark not
defined.

5

Introduction
Utah Lake is a large shallow eutrophic lake located in north central Utah. In recent years, the occurrence of
frequent and sometimes toxic algal blooms has brought new urgency in understanding the cause and effects of
eutrophication in Utah Lake. The drainage basin of Utah Lake is dominated by carbonate sedimentary materials
and small pockets of phosphate rock occur. Given the bedrock geology and shallow nature of the lake, there has
been uncertainty as to the natural trophic status and ecology of the lake prior to the settlement of Utah Valley.
Thus, defining historical nutrient concentrations and sources in the water column and how these have changed will
assist DWQ in defining numeric nutrient targets while the reconstruction of diatom, macrophyte, and zooplankton
communities will assist in understanding how the trophic status and ecology of the lake has changed.
There is evidence to suggest that sometime in the recent past; Utah Lake underwent a regime shift transitioning
from a clear water macrophyte dominated system to a turbid algal dominated system. Early Latter Day Saint (LDS)
settlers have described the clarity of the lake in their journals, lamenting the loss of submerged vegetation, while
other reports speak of a turbid outflow to the lake (Bushman, 1980). Further, early paleolimnological analyses
indicate a recent shift to a greater proportion of planktonic taxa as well as the historical presence of epiphytic taxa
(Bolland, 1974). At present, it is unclear whether such a regime shift has taken place, and if it has, whether the shift
was primarily associated with carp introduction, nutrient increases, or both. The current state of Utah Lake is
heavily affected by both non-native fish introductions and significant effluent loads from multiple wastewater
treatment plants. In recent years, toxic algal blooms have limited access to the lake and compromised downstream
uses of water, and only two of the thirteen native fish species that inhabited the lake are currently present, one of
which is critically endangered and endemic to the lake (Heckman et al 1981).
Paleolimnology is a well-respected tool used for over 130 years to reconstruct ecological shifts in lake ecosystems
(Cowen et al. 2003, Smol et al. 2017). Paleo investigations have advantages over modern ecological data when the
goal is to establish ecological reference conditions, or natural “baselines”, as well as detecting drivers of ecological
change. Paleo information, or reconstructions, generate information back-in-time, before significant
anthropogenic impacts to ecological systems. Paleo data have become an invaluable source for capturing the
natural range of variability, often prior to formal monitoring programs (Smol et al. 2017). Here we build upon
previous paleolimnological work in Utah Lake using well-established paleo techniques, as well as novel, state-of-
the-art research approaches. Notably, we collected multiple sediments cores using several methods and generated
new baseline data at a high temporal resolution. In addition, we employed robust statistical methods, that are
quickly becoming the new gold standard in published paleolimnological studies.
This study collected multiple sediment cores from four key regions of Utah Lake to address key Charge Questions
outlined by the Utah Lake Science Panel. The study will provide a clear historical framework for the timing of
environmental shifts as they may relate to natural variability of anthropogenic forcing in the catchment and the
lake basin itself. Core collection, descriptions, archiving and chronology are presented at the outset. The results of
this multi-year study are structured around the Charge Questions followed by a summary of the reference
conditions in Utah Lake. Detailed methods are provided in advance of results within each subsection and
supplementary findings are presented at the end of the document.
6

Core collection, chronology, and core descriptions
Methods
Core collection and sample archiving
Sediment cores were collected from five locations in Utah Lake (Figure. 1) using both a Livingstone/Bolivia
piston coring system equipped with polycarbonate tubes and a freeze-coring system. In 2018, cores were collected
in Goshen Bay (18-GB-01;40.10865°, -111.8751°), near Bird Island (18-BI-01, 18-BI-02; 40.17118° -111.7998°), and
Provo Bay (18-PB-01; 40.18166, -111.7181). In 2019, freeze cores and piston cores were collected from the North
Basin of the lake (19-N-01; 40.3402°, -111.8307°) and North of Provo Bay (19-S-01; 40.22850°, -111.77804°). In
2020, piston cores were collected from Provo Bay (20-PB-01, 20-PB-2, 40.188104°, -111.700215°). Goshen Bay and
Provo Bay were selected due to historical reports of abundance macrophyte presence. The core collected close to
Bird Island was selected as it is the deepest area of the lake. The North of Provo Bay and North location were
selected to represent the northern regions of the lake, following the path of the main inflow from the Provo River
to the outlet at the Jordon River. The Bird Island core would have been consistently submerged since before lake
monitoring began in the 1880s. Goshen Bay may have become dry in the 1930 or was submerged by just a few
inches of water. The Provo Bay site was completely dry in the 1930s drought and again in the mid-1990s. Archive
halves of 18-BI-01, 19-N-01, and 20-PB-01 are stored at the LacCore facility at the University of Minnesota. The
remaining cores are stored in the Environmental Biogeochemistry and Paleolimnology Laboratory at Utah State
University and frozen sediment cores 19-N-01-FC and 19-S-01-FC are stored at the Natural History Museum of
Utah.
Figure 1 Core Collection sites in 2018, 2019, and 2020.

7

Chronology
A combination of radioisotopic techniques were used to date the cores, including 210Pb, 137Cs, and
radiocarbon (14C). Gamma spectroscopy (Chronos Scientific Inc. – University of Ottawa, Provo Bay and Bird Island
cores; Paleoecological Environmental Assessment and Research Laboratory – Queen’s University, Goshen Bay core)
was used to measure 210Pb and 137Cs activities in samples from all three cores. Accurate dating of Utah Lake core
intervals is hampered by two key environmental factors. One, the lake sediments are frequently re-suspended
leading to a loss of signal resolution. Second, the atmospheric fallout of 210Pb is relatively low in desert
environments (Appleby 1988, 2008), resulting in sediment activities that may be too low to develop core
chronologies. When 210Pb profiles were inadequate to develop a chronology (Goshen Bay and Provo Bay), we used
137Cs concentrations deposited from open-air nuclear weapons testing to better establish an approximate
chronology. Note that radioactive fallout still occurs due to contaminated soils in Utah and particularly Nevada
where US weapons testing facilities are present, thus modern-day sediments still contain substantial 137Cs.
Additionally, unidentified plant macrofossils from the Goshen Bay core were radiocarbon dated using the
accelerator mass spectrometry (AMS) method at the Illinois State Geological Survey Radiocarbon Dating Lab
(Champaign, IL) and results are reported as calibrated ages. Age-depth models were developed using a Bayesian
statistics approach that reconstructs accumulation histories through a combination of 210Pb, 137Cs, and 14C dates
(Aquino-López et al. 2020).
Defining unique sedimentary environments (facies)
Facies are defined as unique sedimentary environments as defined by sediment composition and
macrofossil content. To identify abrupt or gradual shifts in the sedimentary environment, we sent select cores to
the LacCore facility at the University of Minnesota for high-resolution multi-sensor scans, including magnetic
susceptibility, X-radiography, and X-Ray Fluorescence. During subsampling at 0.5-cm increments, the location of
vegetation and gastropod remains were noted, and large macrofossils were collected. Loss on ignition (LOI) was
used to determine water, organic, and carbonate contents at 1 cm resolution in all cores (Heiri et al. 2001).
Homogenized subsamples of ~1-3 grams were dried at 100°C for a minimum of 12 hours, followed by a 4-hour
ignition at 550°C to remove organic matter, and a 2-hour ignition at 1000°C to remove carbonates. Collectively this
information was used to determine when major ecological changes took place and laid the foundation for further
sampling and analysis as described within each section.
Statistics
The Sequential Three-Step Analysis of Regime Shifts (STARS) method developed by Rodionov (2004), was
used to determine significant changes in the means of the individual proxy records through time (e.g., Daskalov et
al. 2007). This method was used to determine statistically significant shifts in the mean values of a particular proxy.
The method uses a sequential Student’s t-test analyses (two-tailed, unequal variances) under the assumption of
independence and normality. For each proxy, STARS estimates the size and direction of the shift, mean values of
each regime, and the p-value associated with the shift. In addition, relationships between some historical trends of
environmental variables, fossil pigments, and geochemical proxies from the Goshen Bay core were estimated using
simple Pearson correlation coefficients across all time and post-carp introduction intervals. This analysis did not
correct for time lags, temporal autocorrelation, nor synergistic effects between predictors, and so represents only
a preliminary estimate of the relationship between the anthropogenic and climatic predictors, and response
variables.
8

Results
Core Chronologies
Approximate age-depth models were developed using Bayesian statistics and the constant-Constant Rate
of Supply (CRS) model based on measured 210Pb and 137Cs activity, as well as radiocarbon-dated material. The Bird
Island core was the only core that had sufficient 210Pb activities to develop a chronology, which we modeled using
the CRS model as well as the Constant Initial Concentration (CIC) model, which yielded different chronologies.
Chronologies for the Goshen Bay and Provo Bay cores were instead produced using 137Cs age markers, specifically
the clear onset of weapons testing in 1954 and the peak (ca. 1963), which was clearly evident in the Goshen Bay
core. Additionally, radiocarbon ages from plant macrofossils were determined from 47 cm, 56 cm and 60.5 cm
from the Goshen Bay core. Age-depth relationships were modeled in R (version 4.1.0; R Development Core Team,
2021) using the Bacon package (rbacon package, version 2.5.7; Blaauw et al., 2021). The age models provide similar
sedimentation rates for Goshen Bay and Provo Bay, and much lower sedimentation rates for the deeper Bird Island
site. Given the higher quality data and combined radiometric techniques for the Goshen Bay core, this chronology
appears to be the most reliable. Given the compatible sedimentation rates in Provo Bay, we also have reasonably
high confidence in the chronology of this core. The Bird Island core, however, is more questionable given the 210Pb
ages do not align with the facies (see next section) and produce vastly different sedimentation rates.
9

Figure 2 Concentrations of 210Pb measured (a) throughout the Bird Island (green) and Provo Bay (purple) cores, and
concentrations of 137Cs measured (b) throughout the Goshen Bay (blue) and Provo Bay cores that were used to
develop the respective age-depth models for each sediment core (c), with the shaded areas indicating the respective
95% confidence intervals. Reproduced from Devey 2022 and King et al. 2024.

10

Defining unique sedimentary environments (facies)
Up to four unique sedimentary environments can be defined based on shifts in the sediment properties,
which are depicted from the organic content in Figure 3. Typically, abrupt shifts in the amount of the organic
content depict changes in the lake and sediment environment. Goshen Bay, whose sediment record has the most
reliable chronology, shows distinct shifts around 1800 AD (52 cm) “Phase Transition 1 (PT1)” and 1890 AD (30 cm)
“Phase Transition 2 (PT2)”. Phase Transition 3 (PT3), occurring around 10 cm in Goshen Bay, or 1975 AD is not well
documented in the sedimentary organic content, but is visible in the other proxies (see details throughout this
document). Below 30 cm in the Goshen Bay core, almost every interval contained plant remains, and gastropod
shells were common. Above 30 cm, these macrofossils abruptly disappeared. Both plant remains and goastropod
macrofossils, along with Anodonta sp. Shells, were common below 32 cm in the Provo Bay core. A singular
Anodonta shell was found below 30 cm in the North Core (undated), and below 32 cm in the Bird Island core.
Gastropod shells were also recovered from several locations below 32 cm in the Bird Island core. The similarity in
depth and macrofossil content, suggests a common shift close to 30 cm across all coring locations, however, the
chronological information form Bird Island is less clear. Dating was not performed on the North core.

Figure 3. Loss on ignition (LOI) results for % organic matter as well as macrofossil remains recovered for all four coring
locations.
11

Charge Question 1.1 What do the diatom community and macrophyte community in the
paleo record tell us about the historical trophic state and nutrient regime of the lake?
1.1.i. Can diatom (benthic and planktonic) and/or macrophyte extent or presence be detected in
sediment cores? And if so, what are they?
Diatoms
Methods
Diatoms samples were processed and counted by BSA Environmental. When possible, frustules were
identified to species level and at least 500 frustules or fragments were counter per interval. Diatom taxa are
presented as a percent of total count per interval. Diatoms are often not identifiable to the species level when
frustules are broken or key features obscured. When diatoms were not identified to the species level, the most
common, and often only, other taxa identified in the same intervals were assumed to be taxa represented. The
pollution tolerance index was developed using the Biological Condition Gradient (Davies and Jackson 2006), where
1 indicates a specialist/sensitive species and 5 represents a tolerant species. The index was constructed by
summing the taxa weighted by their Biological Condition Gradient. Note that BCG values are not available for a
large proportion of the taxa and the values developed by Davies and Jackson (2006) were for California, and it is
possible that taxonomic relationships with environmental conditions differ regionally.
Results
Diatoms are presented as relative abundance within each sediment interval. Significant shifts in the mean
as determined by the RSI test are shown along with the p-value. Taxa plotted are limited to those that have a
relative concentration greater than 5% of the community at some point in the historical record, and/or show the
temporal transitions between communities. When available the Biological Condition Gradient (BCG) is referenced
where 1 represents the most sensitive and specialized species and 5 the most pollution tolerant. Note that we
cannot use diatom taxa to identify past phosphorus concentrations of Utah Lake, as a training set of 30 to 40
similar lakes is not available given the unique nature of Utah Lake. Diatom counts were performed on the Goshen
Bay, Provo Bay, and Bird Island cores collected in 2018. However, subsequent analyses revealed poor chronological
representation in the Provo Bay core, thus a new core was collected in Provo Bay in 2020. Diatoms counts were not
performed on the 2020 core and the 2018 data will not be interpreted here.
Goshen Bay
In Goshen Bay, distinct shifts in the community composition are frequently observed between 30 and 40
cm (PT 2) and again between 15 and 10 cm depth (PT3), which correspond roughly to the late 19th century and
between 1960-1975 (Figures4-7)(see comments in Chronology section). In general, during the first transition at ~30
cm we see the loss of some species while others for the first time appear in the record. A slower transition away
from epiphytic diatoms (Epithemia spp.) is observed during the first transition, which is followed by a more recent
reduction in epissamic, or sand dwelling species such Staurosirella martyi, S. construens, S. venter, represented by
the ‘small benthic sum’. Large pollution tolerant species begin to increase around ~30-35 cm and continue to
increase upcore. At around 30 cm an increase occurs in some benthic species such an Amphora copulata (BCG 4)
and Pleurosigma spenceri (BCG 4), Navicula cryptochephala (BCG 4; High pH) at the expense of benthic species that
are more sensitive (e.g Pinnularia microstauron (BCG 3). The pollution tolerance index shows a sharp increase at 30
cm and remains high until 2.5 cm where it falls down to lower levels. Pollution tolerant planktonic species such as
Aulacoseira granulata, Cyclotella meneghiniana, and Stephanodiscus hantzschii begin to appear above 12 cm. At
around 10 cm (PT3) we see a significant increase in the representation of planktonic species vs benthic species.
12

In summary, diatom taxa below ~30 cm are represented by more sensitive benthic species (BCG 2, 3) and
includes the presence of epiphytic species. At 30 cm we see a shift in the taxonomic representation of many
species towards more pollution tolerant benthic species (BCG 5). At ~12 cm there is a sharp increase in pollution
tolerant planktonic species and an increase in planktonic representation. The modern community is representative
of shallow, alkaline, and eutrophic conditions whereas the pre-disturbance community is representative of
shallow, alkaline, mesotrophic conditions with a greater presence of macrophytes.

Figure 4. Diatom species from Goshen Bay plotted as relative concentration by depth. Red dashed lines show
the significant shifts based on STARS analyses.
Figure 5. Diatom from Goshen Bay con’t.
13

Figure 6. Diatoms from Goshen Bay con’t.

Figure 7. Diatoms from Goshen Bay con’t.

Bird Island
At Bird Island, distinct shifts in community composition are observed between 40 to 30 cm, 20 cm and 10
cm. The early section of the core (pre 40 cm) is characterized by large sensitive benthic species such as Pinnularia
microstauron, small benthic epissamic species such as, Staurosira spp, most of which completely disappear after 11
cm. Epithemia spp. decline sharply around 40 cm and continue to decrease upcore. At 30 cm we see an increase in
taxa tolerant of fine sediment disturbance, such as Navicula salinarum and Navicula veneta, as well as more
pollution tolerant benthic species mixed with more sensitive species, such as Nitzschia palea (BCG 5), Tryblionella
spp (BCG 4-5). Gyrosigma acuminaun (BCG 4), Diploneis elliptica (BCG 3), and Staurosira spp. (BCG 4-5). We also
see the emergence of the planktonic Discostella psudostelligera (BCG 4). Above 10 cm, D. stelligera is replaced with
pollution tolerant species such as A. granualta (BCG 5), Cyclostephanos dubius (BCG 4), and Cyclotella
meneghiniana (BCG 5), Stephanodiscus sp. as well as the benthic Pleurosigma spenceri (BCG 4). The planktonic to
benthic ratios rapidly and signficantly increases at 10 cm while the pollution tolerance index increases gradually
through the core, peaking at 4.5 cm (~1990). As with the Goshen Bay core, many of the species identified are

14

common in brackish, alkaline waters. Towards the surface of the core the representation by eutraphentic species
increases.
In summary, the early part of the core is characterized by more sensitive species and includes species that
live on aquatic vegetation. This period is followed by an increase in large benthic pollution and turbid tolerant
species. The upper 10 cm is in contrast characterized by an increase in pollution tolerant planktonic species at the
expense of both benthic species and more sensitive planktonic species. The modern community is representative
of shallow, alkaline, and eutrophic conditions whereas the pre-disturbance community is representative of
shallow, alkaline, mesotrophic conditions with a greater presence of macrophytes.
Additional information on the shifts in diatom abundance through time can be found below in the section on algal
pigment analyses.

Figure 8 Diatom species from Bird Island plotted as relative concentration by depth. Red dashed lines show the
significant shifts based on STARS analyses.

Figure 9. Diatoms from Bird Island con’t.

Figure 9. Diatoms from Bird Island con’t.

15

Figure 10. Diatoms from Bird Island con’t.

Figure 11. Diatoms from Bird Island con’t.
Macrophytes
Methods
The historical presence or absence of macrophytes at coring locations were identified through several
means, 1) the enumeration of physical plant remains, 2) metagenomics, 3) carbon and nitrogen mass ratios and
isotopes, 4) Rock-Eval Pyrolysis, and 5) the presence or absence of proxies such as gastropod shells. Freshwater
gastropods feed on plant debris and benthic algae and thus indicate the presence of vegetation and or light
penetration to the benthos. Carbon to nitrogen ratios and Rock-Eval pyrolysis have previously been successfully
used to distinguish between algal and plant biomass within lake sediments (Lüniger & Schwark, 2002; Meyers,
2003; Talbot & Livingstone, 1989). Similarly, these analyses within Utah Lake sediment cores will be used to
determine, both in time and in space, whether a transition from a clear water macrophyte- to a turbid algal-
dominated system occurred.

16

Physical remains of macrofossils
The presence or absence of vegetation remains embedded within the sediments were recorded and
documented during sediment core logging at USU, UU, and LacCore.
Metagenomics
The development of metabarcoding of environmental DNA ‘eDNA’ has opened up new areas of research
using lake sediment archives with unparalleled detail. eDNA based methods have been used to successfully
reconstruct algal, zooplankton, and fish community shifts that have occurred parallel to anthropogenic activities
within catchments (Capo et al., 2016; Monchamp et al., 2016). However, eDNA is not always well preserved and
has the capacity for regional transport. Nevertheless, we expect eDNA to provide a complimentary data on the
species present in the area, and may provide additional historical biomarkers due to the presence of exotic species
(e.g. dandelions) and the establishment of agriculture in the catchment (e.g. curcurbits). eDNA were conducted at
Jonah Ventures in Boulder, CO, and will provide a regional perspective on the types of aquatic plants present at the
time of sediment deposition.
Sedimentary eDNA (sedDNA) samples were analyzed from the Goshen Bay core only due to the specific
requirements around sample extraction. Subsamples (n = 21) were collected from the center of the core during the
extruding process, placed in 20 mL plastic scintillation vials, and frozen (–20°C) until further use. All equipment was
pre-sterilized with UV light, and sterile gloves and ethanol were used to prevent contamination between samples.
Frozen sedDNA subsamples were shipped to Jonah Ventures (Boulder, CO) for analysis, which included next-
generation sequencing (NGS) of phytoplankton and higher plant assemblages. DNA was extracted using the DNeasy
PowerSoil HTP 96 Kit, and sequenced on an Illumina MiSeq (San Diego, CA) using the v2 500-cycle kit. Primers were
used to amplify DNA sequences preserved in the sediments (23S for phytoplankton and trnL for higher plants), and
then sequences were matched to an internal reference database. Only samples that were identified with ≥ 97%
certainty were used in the analysis.
Although sedDNA has been used successfully to reconstruct changes in plankton communities in response
to anthropogenic activities (Monchamp et al. 2017), it is important to note that it is still an emerging tool and that
not all macrophyte populations are captured using this metric (Giguet-Covex et al. 2019). Further research is
needed to better understand the effects of in situ sedDNA degradation, including whether it is selective for
individual markers, and to better curate reference databases. Regardless, sedDNA appears to be able to record
abrupt changes in lake structure and function, particularly when used in combination with other well-tested fossil
metrics of planktonic and benthic community change.
Rock-Eval pyrolysis
Rock-Eval pyrolysis of Goshen Bay sediments was conducted at Brigham Young University to approximate
the Hydrogen Index (HI; Carrie et al. 2012), a measure of organic matter origin based on the hydrogen content of
bulk organic matter (Talbot and Livingstone 1989). Prior to pyrolysis, sediment subsamples (n = 61) were sieved
through a 40-µm pore mesh and crushed to standardize particle sizes. Pyrolysis was conducted using a Wildcat
Technologies HAWK Workstation by means of two phases of combustion: a reducing pyrolysis phase and an
oxidizing stage. The HI (reported as the standard mg of liberated hydrocarbon per g of total organic carbon; mg HC
g-1 TOC) was determined during the pyrolysis phase and is of primary interest as it differentiates organic matter
derived from algal material (high values) and from vascular plants (low values; Meyers & Teranes 2001). Rock-Eval
pyrolysis of Goshen Bay sediments was conducted at Brigham Young University to approximate the Hydrogen
Index (HI; Carrie et al. 2012), a measure of organic matter origin based on the hydrogen content of bulk organic
matter (Talbot and Livingstone 1989) The HI (reported as the standard mg of liberated hydrocarbon per g of total
organic carbon; mg HC g-1 TOC) was determined during the pyrolysis phase and is of primary interest as it
17

differentiates organic matter derived from algal material (high values) and from vascular plants (low values;
Meyers & Teranes 2001).
Carbon:Nitrogen mass ratios
Carbon and nitrogen concentration in cores can be used to determine historical production with greater
concentrations reflecting greater production, and C:N ratios can effectively be used to distinguish between algal
versus plant derived sources of organic matter because plants contain greater C-rich structural compounds
compared to protein (N)-rich algae (Meyers, 2003). Sediment samples for carbon and nitrogen were analyzed at 5-
cm intervals to evaluate shifts organic matter source. Each sample consisted of an aliquot of 600–700 mg of
sediment and which was weighed into centrifuge tubes and placed in an oven to dry for at least 48 hours at 60°C.
Two subsamples of about 300mg each were transferred to new pre-labeled centrifuge tubes. One of the 300mg
samples destined for elemental organic carbon and nitrogen (and isotope) analysis was fumigated with dilute
hydrochloric acid to remove carbonates so that the analysis constituted only the organic fraction. The other 300mg
sample was left untreated and used for carbonate analysis. The sediment samples destined for organic matter
isotope analyses were further rinsed to remove halides, and then dried and homogenized by crushing in a petri
dish. The homogenized samples were then weighed in compressed tin capsules and analyzed together with
internal laboratory standards of known isotopic and elemental composition. The samples were combusted in a
Costech 4010 Elemental Analyzer at 1650ºC and inlet to a Finnigan® MAT 252 Isotope Ratio Mass Spectrometry
(IRMS) in continuous flow mode. The stable isotope analyses were done in the Geochemistry laboratory directed
by Dr. Cerling at the University of Utah and in the Newell Lab at USU.
Carbon:Nitrogen North Freeze core (19-N-01-FC)
Analysis of bulk organic matter and using C:N ratios show significant changes above and below ~23 cm
depth from the North Core (Figure 13). Above ~23 cm depth, C:N ratios decrease steadily to present. The
variations in 13CBOM values are modest with values decreasing steadily toward present. Changes in %C show a
steady increase of ~8 to 8.5% from ~23 cm toward the surface.
The nutrients (C, N) as well as 13CSOM and 15NSOM profiles from the North Core, provide additional
evidence of changes in land-use around Utah Lake since European settlement. Sediment samples analyzed for
percent organic, carbon, nitrogen C:N ratios and 13C and 15N suggest increased nutrient loading into Utah Lake
during the past ~150 years, beginning around ~23-25 cm depth in the North Core (Figure 14). The basin-wide
signal of lake ecosystem change is corroborated by similar evidence from sediment from Goshen Bay cores. The
impact of nutrients associated with European settlement in Utah valley has ultimately re-baselined the
geochemistry of Utah Lake.

18

Figure 12 North Core geochemistry profiles of organic matter (OM %), carbon (C %), nitrogen (N %), C:N ratios, δ13C (‰), and δ15N (‰)
subsampled at ~5 mm contiguous resolution from the frozen sediment core. Dashed lines indicate significant abrupt changes as detected by
STARS.
Typical 15N values for watersheds with limited human activity range from 3.5 to 5.5‰, while freshwater
systems in areas exceeding >15% urban and agricultural development typically range from 6 to 9‰ (Mayer et al.,
2002). The top (~25 cm) sediment 15N values in the North core average ~8.5‰ and are comparable to watersheds
with intensified agricultural or urban land use (i.e., 15N values ranging from 6‰ to 9‰) (Mayer et al. 2002).
Farming and grazing likely contributed to a significant portion of land use change in the Utah Lake watershed
(Squires and Rushforth 1986) and these geochemical changes are captured in multiple sediment cores throughout
the basin.

Figure 13 Utah Lake organic nitrogen from the North Core (19-N-01-FC) show a significant correlation (R2 = 0.89) between 15N and nitrogen
content (%N). This positive trend suggests that an influx of 15N-enriched nutrients into Utah Lake likely originate from external sources.
To explore potential nitrogen sources, a strong correlation between %N and 15N (Figure 15) suggests
nutrient enrichment likely originate outside of the lake. The ~ 3.0‰ enrichment in 15N values between the
bottom (~5.0‰) and the top (~8.0‰) of the North core (and other cores) suggests that the influx of 15N -enriched
19

nutrients drove increased primary productivity within Utah Lake. Human effluent is a common source of elevated
15N values in populated regions over time. Utah Lake has approximately nine sewage plants contributing effluent
to the system historically, including three in Provo Bay (Bushman, 1980).
Nutrient loading can generate positive feedback that maintains a lakes eutrophic state. Land use change
over the last ~150 years has increased nutrient influx to the lake to a point where nitrogen and phosphorus
loadings persist for extended periods, and potentially require unconventional remedial methods (Elliot and Brush
2006). The transformation of Utah Lake from clear to turbid state is likely attributed to resuspension of calcite in
the water column by exotic fish species, including the German Carp (Cyprinus carpio L.), combined with upstream,
watershed changes (e.g. river impoundments and reservoir construction) which have all impacted lake-levels and
geochemistry.
Results
Physical remains
Physical macrophyte remains were recovered in both the Goshen Bay and the 2020 Provo Bay cores. In the
Goshen Bay core, macrophyte remains were recovered from nearly every interval below 30 cm (PT2). No preserved
remains were found above ~30 cm. In Provo Bay, macrophyte remains were recovered below ~22cm, with more
frequent remains occurring below 32 cm (Figure 3).
sedDNA, Goshen Bay
In the Goshen Bay core, next generation sequencing of sedDNA samples (n = 21) detected 278
phytoplankton and 65 higher plant sequences. Most phytoplankton sequences could only be classified at the
phylum level (e.g., 36 cyanobacteria sequences). Cyanobacteria taxa were detected in only two (of nine) sediment
samples below 30 cm; however, they were detected at high frequencies in more recently deposited samples
(Figure 12.). In addition to phytoplankton, next-generation sequencing for higher plants detected sequences that
were identifiable as aquatic macrophytes (Fig. 5a). Hardstem bulrush (Schoenoplectus acutus), an emergent
macrophyte, was the only native aquatic plants detected in historic sediments. Within the top 6 cm of sediment
samples, submerged macrophytes small pondweed (Potamogeton berchtoldii) and sago pondweed (Stuckenia
pectinata), and floating-leaved macrophytes water smartweed (Persicaria amphibia) and water lilies (Nymphaea)
were all detected. Food crops were only detected in sediment layers above ~30 cm, which supports the Goshen
Bay core chronology. The presence of dandelions was only detected in the 30.5 cm sample highlighting the
potential limitation of sedDNA as a proxy because dandelions have been present in Utah since the late 1800s.
20

Carbon:Nitrogen and Rock-Eval, Goshen Bay
Analysis of bulk organic matter using C:N ratios and Rock-Eval pyrolysis suggested that prior to 30 cm (late
19th century) the coring location exhibited a high abundance of benthic plants, which tend to have high C:N ratios
compared to turbid, planktonic-dominated regimes. Rock-Eval HI shows a marked increase at 30 cm. HI increased
from and average of 28 to 85 mg HC g-1 TOC (RSI = -1.6, p < 0.001, whereas bulk C:N ratio decreased in an up-core
direction, although the change was not considered abrupt.

Figure 14. . Environmental DNA results for Hardstem bulrush and
Cyanobacteria from the Goshen Bay core.
Figure 15. Population growth within Utah Lake valley (a; World Population Review, 2022) shown alongside
profiles of organic matter (OM %, cutoff length = 40), carbon (C %), nitrogen (N %), C:N ratios, δ13C (‰), δ15N
(‰), and Hydrogen Index (HI, mg HC g)
21

Proxy evidence
In addition to the physical remains of plant material, the presence of gastropods is indicative of benthic
production as their habitat generally requires decaying vegetation and/or benthic algae and gastropod
concentrations are known to increase in concert with macrophyte density (Diehl and Kornijow in Scheffer and
Jeeppesen 1998). Gastropod remains were found in Goshen Bay below 30 cm only. In Provo Bay, above 22 cm
gastropod remains were found in less than 40% of samples. In contrast, below 22 cm, gastropod remains were
found in more than 65% of the samples, many containing more than 10 individuals/fragments. Anodonta sp. shells
were found in Bird Island, Provo Bay, and the North Core below 30 cm (Figure 3). As noted above, the presence of
diatoms that are epiphytic are found in much greater abundance historically in both cores analyzed (Bird Island and
Goshen Bay). Finally, the Rock-Eval analyses revealed a step shift at 30 cm from low values representative of
vascular plants to high values representative of algal material (Meyers & Teranes 2001) (Figure 12).
Summary
Charge question 1.1 requested information on the historical trophic status and nutrient regime of the lake
as determined through microfossil and macrofossil remains. Here we used diatom records to reconstruct the
community shifts and make inferences on habitat shifts based on the physical requirements and preferences of the
taxa. Diatom data suggest that prior to Phase Transition 2 (around 30 cm depth and 1890 AD), the benthic
environment contained plants, based on the presence of epiphytic taxa, and the sedimentary environment
received more light, based on the presence of sand-dwelling taxa. Together, this suggests that Utah Lake in parts
had abundant macrophyte beds and the turbidity was reduced compared to present day. Because pollution
tolerant taxa become more prominent upcore and pollution sensitive taxa disappear, we can infer that Utah Lake
became more eutrophic through time.
Charge Question 1.2. What were the historic phosphorus, nitrogen, and silicon
concentrations as depicted by sediment cores? (add calcium, iron, and potentially N and
P isotopes)
*Note that P isotopes are not stable in the natural environment and thus cannot be evaluated.
Methods
Phosphorus
Reconstructing historical water column phosphorus concentrations from lake sediments has been a
longstanding challenge due to the post depositional mobility of P within sediments(Carignan & Flett, 1981;
Golterman, 2001). Because of this process, total P or pore-water P cannot be used to determine historical P loading
(Engstrom, Daniel R, Wright, 1984). Alternative means for reconstructing historical phosphorus concentrations rely
on complex statistical relationships using diatom communities. These generally require sampling dozens of similar
lakes to establish a relationship between diatom communities and ambient phosphorus concentrations before
applying this information to sediment core records (Hall & Smol, 1992). Beyond the significant time commitment
involved in producing regional diatom-phosphorus calibration sets and the unavailability of similar systems to Utah
Lake, these reconstructions are fraught with statistical and practical uncertainties (Juggins et al., 2013). We instead
reconstructed historical P loading using three complementary methods. First, we used an existing method
(Ruttenberg, 1992) to sequentially extract different phosphorus fractions. However, most extraction methods are
not target specific leading to operational (acid-leachable) rather than true sediment fraction categories (e.g.
calcite). For example, an acid leach meant to target calcite will also target humic-P and some Fe-P minerals
(Benzing & Richardson, 2005; De Groot & Golterman, 1990). Sequential leaching can also mix between operational
22

categories where P leached from the Fe-/Mn-leach can attach itself to calcium, artificially elevating the acid
leachable (“calcite”) fraction.
Based on these uncertainties, we used two additional novel techniques to reconstruct ambient water
column P concentrations. First, we used a Scanning Electron Microscope with Energy Dispersive X-Ray (SEM-EX) to
measure phosphate inclusions in calcite precipitated in the water column as a measure of epilimnion phosphorus
concentrations. The method is based on the premise that the precipitation of calcite (CaCO3) in the water column
will incorporate PO43- at concentrations that parallel ambient water column conditions. It is well established that
phosphate binds strongly to calcite and can scavenge phosphate from the water column (Whitehouse, 2010).
Because conditions within lake sediments do not generally promote calcite dissolution particularly in Utah Lake
(Randall et al., 2019) post depositional alteration of this fraction is not likely. Thus, P incorporated into crystalline
calcite may accurately record relative, even quantitative, shifts in Utah Lake water column phosphorus. To test this
hypothesis, we have performed a laboratory experiment comparing initial mass in solution to the mass in a calcite
precipitate. Precipitation of CaCO3 in phosphate-spiked solutions was used to determine the relationship between
calcite scavenged P and initial concentrations across a range of P concentrations, which resulted in a tight
relationship (r2 of 0.996, p << 0.001). We further examined P scavenged by calcite using SEM-EDX and LA-ICP-MS to
relate starting P concentrations to SEM-EDX determined P. This step is crucial to establish the relationship between
water column P and sediment core crystalline calcite-P, which can only be determined via SEM-EDX or LA-ICP-MS
analyses. This experiment was also successful producing an r2 of 1, p << 0.001. The second technique involved
physical isolation of the calcite fraction, followed by a surface cleaning and a final digest, which is detailed below.
Sequential Fractionation
Our sequential fractionation scheme consisted of five increasingly harsh reagents and process, dividing TP
into five operational fractions (table 1). Subsamples were freeze-dried and weighed in plastic centrifuge tubes prior
to extractions. Each individual sample had a volume of extractant applied proportional to the mass of the sample
so that the mass-to-volume ratio remained constant. The exact volume of extractant applied to each sample was
calculated by weighing the liquid and converting it to volume from a known density.
Table 1 summarizes the reagents used, basic process, and expected P forms as well as what type (SRP,
NRP, TP) of P was analyzed. The first step extracts P loosely adsorbed to particle surfaces, exchangeable P, and
porewater P, using 1 M magnesium chloride (MgCl2). Salts such as MgCl2 are used in extraction schemes to extract
P by either the formation of a MgPO4- complex or displacement by the Cl- ion. In the second step, the citrate
bicarbonate-dithionite (BD) solution is meant to extract the fraction of P which is bound to reducible forms of Fe
and Mn. A mixed solution of trisodium citrate, sodium bicarbonate, and sodium dithionite (“BD”) reduces oxidized
species of Fe and Mn, thereby releasing adsorbed P from Fe/Mn oxides and hydroxides. Citrate acts as a chelator,
complexing metal ions. When used early in the sequence, CDB has the benefit of not disturbing P associated with
clays or organic compounds. However, citrate has a strong affinity for Ca and may extract some P loosely bound to
the exterior of calcium minerals (Broberg and Persson 1988).
The third step uses sodium hydroxide (NaOH) to extract P by hydrolysis from a broad fraction generally
termed “biogenic”, which consists of P in microorganisms, polyphosphates, detritus, humic acids, and other
organic compounds (Hupfer et al. 1995). It also extracts P bound to Al oxides (Al-P) and any P exchangeable with
OH-. Biogenic P is distinguished from Al-P in this approach by:

𝑇𝑁=𝑁𝑟𝑡ℎ𝑜−𝑁+𝑁𝑅𝑁

Non-reactive phosphorus (NRP) is calculated as the difference of total P (TP) and orthophosphate -P. NRP
corresponds with biogenic P and orthophosphate P corresponds with Al-bound and OH- exchangeable P.
23

The fourth extraction is dilute hydrochloric acid, targeting the more readily soluble forms of CaCO3-P,
including amorphous Ca minerals, biogenic calcium carbonate, crystalline calcium carbonate precipitates, and
possibly authigenic apatite. A 0.5 M HCl solution will not dissolve any detrital apatite in the sediments. This fraction
is considered not bioavailable.
Finally, the refractory P extraction targets biogenic compounds which are otherwise resistant to extraction,
including some phosphate esters and phosphonates (Ingall et al. 1990; Broberg and Persson 1988). Although this
fraction is, by definition, not bioavailable and is generally considered a permanent P pool. Here we used dry
oxidation at 550° C, followed by extraction with 1 M HCl to evaluate the recalcitrant pool. This decision is based on
evidence that wet persulfate digestions underestimate TP (Zhang 2012). Total Phosphorus is represented by the
sum of all the above extractions.

Table 1. Summary of the sequential fractionation scheme used on all four cores, adapted from a scheme suggested
for CaCO3-rich lake environments by Hupfer et al. (2009). Another study on Utah Lake sediments, Randall et al
(2019), also used a similar approach.
Step Reagent Concentration Time Expected P Forms
1 MgCl2 1.0 M 0.5 h P loosely adsorbed to the surfaces
of particles; immediately available P
2 Sodium Bicarbonate
Sodium Dithionite
Trisodium citrate
1.0 M
16 h Reducible P forms, especially bound
to Fe and Mn 0.14 M
0.3 M
3
NaOH 1.0 M
16 h
Al-bound P (Ortho-P) and P in organic material
(Non Reactive P(NRP)), including P in detritus, P
bound to humic compounds, and P in microbes
4 HCl 0.5 M 16 h Ca-P, especially calcite-P co-precipitates
5 Ignition at 550 °C 16 h Refractory P
HCl 1.0 M

After completion, the MgCl2 and NaOH extracts were analyzed for SRP on a SpectraMax® M2 using the
molybdenum blue method. The bicarbonate-dithionite, NaOH, and both HCl extracts were analyzed for TP using an
Agilent Triple Quadrupole ICP-MS at the ICP-MS Metals laboratory at the University of Utah, Utah, USA, as well as
by the same instrument at the Department of Geosciences at Utah State University.
Energy-dispersive X-ray Spectroscopy
To determine CaCO3-bound phosphorus concentrations, samples were digested in 30% hydrogen peroxide
for 72 hours to remove organic matter and organic-associated P, rinsed with a MgCl2 solution, then rinsed in
ultrapure water to remove any weakly-adsorbed P. The material was then suspended again in ultrapure water and
drop cast onto silicon wafers for analysis. The mounted samples were then carbon coated to improve conductivity
for better results. Because silicon and carbon were introduced into the sample as mounting medium and coating,
respectively, they were removed from the final elemental spectra.
Scanning electron microscopy (SEM) was conducted with a FEI Quanta FEG-650 SEM operating at 10 kV
accelerating voltage. Energy Dispersive X-ray spectroscopy (EDS) was performed with an Oxford X-Max detector,
attached to the FEI Quanta FEG, at 10 kV. Sample solution was drop-casted onto 10 x 10 mm Si-chip substrate to
make a thick film, air dried then 10 nm carbon coated via thermal evaporation using EMS-150 ES metal/carbon
coater, before placing in the SEM chamber.
24

Once under the SEM, we arbitrarily chose an area on the sample mount and mapped the image for
constituent elements, including oxygen, carbon, silicon, calcium, and any other trace elements the instrument
could detect (up to parts per thousand). From these maps it was possible to distinguish grains by approximate
composition, including calcium carbonates, calcium-magnesium carbonate, phosphates, aluminosilicates, and iron
oxides. From these maps, we selected ten grains which, based on relative composition, were most likely calcium
carbonate grains for EDS analysis. We allowed each grain 3×106 counts to detect P at 10 kV energy and 10,000x
magnification. If, after all counts were completed and the software was not able to confirm the presence of P at
the given point and we consider the sample to be < 0.0% P by weight, or a non detect.
We analyzed approximately ten randomly selected CaCO3 grains for each of the intervals sampled. We also
collected spectra for several non- CaCO3 grains during this exploration, particularly where P was concentrated
according to the elemental map. It was relatively easy, using this method, for us to avoid concentrations of iron or
aluminum, where P was often concentrated much more than in CaCO3. We were also able to distinguish apatite
grains by the presence of fluorine and very high concentrations of P.
Size-separate CaCO3-Phosphorus Analysis
We used a novel, improved approach to analyzing CaCO3-P by using a less harsh extracting reagent on a
more targeted size fraction of sediments. Analysis of the grain size distribution of mineral types using the SEM/EDS
indicated that the authigenic CaCO3 in the Goshen Bay core belonged to a size class that was distinctly finer than
most detrital grains. Bulk sediments were ignited at 550 °C in acid-washed crucibles to release organic and
refractory P, then vortexed and sonicated in a MgCl2 solution to break up clumps and remove loosely-adsorbed P.
Following the MgCL2 rinse, samples were triple-rinsed in ultrapure water, then filtered to isolate the 0.45-10-
micron fraction. Using ultrapure water and a vacuum system, the sediment material was separated using a 10 µm
Nitex mesh filter. The suspended <10 µm fraction was collected onto a 0.45 µm cellulose acetate filter and put in a
desiccator to dry. The filters were then placed into polypropylene centrifuge tubes, dried again at 110 °C, and
weighed.
With this modified extraction procedure, we sought to address the issue that the HCl reagent used in
typical sequential extraction schemes risks carry-over from other fractions. A buffer solution (pH=4.5), which could
still dissolve CaCO3 but at a much higher pH, was used to dissolve calcium carbonate from the isolated 0.45-10 µm
fraction captured on the filters. Seven mL of the buffer solution was added to each sample tube and shaken for 48
hours at 20 °C. The mixture was centrifuged and extracted, and the remaining detrital sediment and filter paper
were rinsed, dried, and weighed again to obtain a total mass of dissolved sediments. This dissolved material was
assumed to be primarily carbonate material. The extract was analyzed via quadrupole ICP-MS for phosphorus and
a suite of metals typically associated with carbonate minerals (Ca, Mg, Ba, Sr, Mn, and Fe).
Results
Sequential Phosphorus Fractionation
The Goshen Bay core exhibited a steady increase in MgCl2-extractable P between from about 4 to 13 mg g-1
(Figure 14), with significant increases at 31 and 9 cm. The BD-extractable P varied throughout the profile from as
high as 168 mg g-1 to below reagent blank values, but there were no consistent trends throughout the core. The
NRP fraction of the NaOH-extraction that represents organic-P significantly increased around 30 cm and remained
higher throughout the remainder of the core. The HCL extraction exhibited higher values from 45 to 32 cm,
showing a significant decrease at 32 cm, and then rose again around 9 cm towards the surface of the core. There
were no discernable trends in the Total P record for Goshen Bay.

25

The Bird Island core showed modest increases in MgCl2-extractable P up-core, on the order of 3-13 mg g-1
(Figure 15), with a significant shift around 30 cm. In a similar fashion to the Goshen Bay core, BD-extractable P
showed much variability and no apparent trend, except for peaks around 9 and 21 cm, which corresponded with
peaks or troughs in subsequent fractions. The NaOH-NRP exhibited variability between 5-15 mg g-1, and a slight
increased towards the surface of the core after 9 cm. The 0.5 M HCl-extractable fraction remained at or near
detection limits until increasing at 55 cm to the 100- mg g-1 P range, and increasing again around 30 cm to the 300-
400 mg g-1 range until peaking at 9 cm to over 1200 mg g-1 P. Total Phosphorus peaked at 21 and 9-10 cm.

Phosphorus fractions in the Provo Bay core revealed marked transitions that correspond roughly with LOI
trends (Figure 16. In particular the MgCl2, BD, NaOH-Ortho P increased immediately following 30 cm, while the
NaOH NRP, HCL and Total P were depleted in the interval immediately following 30 cm, but increased around 20

Figure 16. Sequential Phosphors Extractions for Goshen Bay.
Figure 17. Sequential Phosphors Extractions for Bird Island
26

cm. NaOH-NRP increased from ~6 mg g-1 between 30 and 20 cm to as high as 46 in newer deposits. HCl-extractable
P experienced a very similar-shaped profile to NaOH-NRP but increased from about values in the low 300 mg/kg
range to values in the 560 mg g-1 range.

Figure 18. Sequential Phosphors Extractions for Provo Bay.

The North core MgCl2-extractable P showed little change until the topmost interval, where it increased
sharply to over 20 mg g-1 (Figure 17). All other fractions showed gradual increases from the base of the core
towards the surface.

Figure 19. Sequential Phosphors Extractions for the North core.
27

Energy-dispersive X-ray Spectroscopy Analysis
CaCO3 grains were easy to detect based on layered color-image maps (Figure 18) and their distinct, fine
grain size. CaCO3 analyzed using SEM-EDS contained variable amounts of P (below detection to 11 ‰), mean
sediment interval concentrations are shown in Figure 2D. The exact portion of non-detects ranged from 10 % (2 of
21, at depth = 0 cm) to 100% (10 of 10, at core base depth = 59.5 cm). Although we only measured a modest
sample size of sediment core intervals, the average wt. % P increased up-core. Again, the highest P content was
observed in the youngest sediments representing 2018, when the core was collected.

Figure 20. SEM-EDS false color image of elemental abundance within the surface sediment sample collected from Goshen Bay, Utah Lake.
Note, although carbon was removed from the map summary spectrum for quantification, it was not possible to
remove carbon.
Size-separated CaCO3-Phosphorus Analysis
The extraction process dissolved 51.7 to 178.1 mg of material (CaCO3) from each sample. Of the dissolved
cations, 90-93% was Ca, whereas Mg, Mn, Fe, Sr, and Ba, made up minor constituents. Mg was the second most
common cation present (approx. 1.8% to 2.8%), likely as either a Ca substitution in calcite, or as dolomite
(mineralogy of Utah Lake sediments explored in Randall et al. 2019).

28

P concentrations in the dissolved material doubled throughout the sediment core, increasing from a
baseline of around 400 mg kg-1 to values close to 800 mg kg-1 at the sediment surface (corresponding to 2018,
when the core was collected; Figure 19B). This P concentration is of the material dissolved during the sodium
acetate buffer extraction, which is assumed to be mostly CaCO3. This calculation is distinct from the more standard
measurement, where P concentrations are based on the mass of the entire sediment sample, regardless of CaCO3
content. However, we also determined the accumulation rate of CaCO3-P, based on the total sediment mass and
the sedimentation rate (Figure 19C). The increase in the P concentration in CaCO3 begins at 28.5 cm based on the
SRSD, corresponding approximately with the year ~1877 (± 34 years), which is consistent with the introduction of
common carp into the lake system and marked changes in other sediment proxies (King et al. 2024). A second
significant increase was observed in both concentration and the accumulation rate post ~1970.
Unfortunately, we were unable to directly compare the results of both methods because the data were
collected using different sediment intervals due to limited material. Still, we evaluated both methods by comparing
them to indices such as Utah Valley population as a proxy for P loading (Figure 19A). We also compared both
CaCO3-P detection methods and TP against other sediment proxy’s indicative of lake water column production
(Canthaxanthin) and wastewater discharge (15N values) (King et al. 2024) (Figure 19); both CaCO3-P methods
showed strong correlations to these indices (Figure 4), while there were no statistical relationships to TP (Figure
18E).

Figure 21. Panel A: Population of Utah Valley (World Population Review, 2022) with the covered wagon
indicating European Settlement in 1849 (Janetski, 1990), the fish indicating Carp introduction in ~1883
(Heckmann et al. 1981), and the spigot indicating the implementation of secondary sewage treatment in ~1950.
Panel B: The concentration of phosphorus in physically and chemically separated authigenic calcite grains, Panel
C: accumulation rates of P bound in authigenic calcite, and Panel D: SEM-EDS determined mean concentrations
of P from individual authigenic CaCO3 grains, error bars represent the standard error. E: Total Phosphorus (TP).
Dashed lines indicate significant (p < 0.1) shifts in the mean.
29

Figure 22. Regression analyses of calcite-P concentrations against proxies of nutrient loading (Population, 15N)
and water column productivity (sedimentary Canthaxanthin) using the Modified Sequential Extraction and ICP-MS
approach (left) and the SEM--EDS Analysis on Authigenic CaCO3 approach (right).

30

Discussion
The most prominent increases in phosphorus across all cores occurs around 30 cm and 10 cm, or Phase
Transition 2 and 3 respectively. Phase Transition 2 is temporally associated with Carp introduction and a
subsequent reorganization of the lake environment through the loss of macrophyte vegetation. Thus, increases in
water column phosphorus and a reorganization of nutrient cycling can be expected. Phase Transition 3 is
associated with an increased trophic status and cyanobacterial presence as determined through diatoms and
metagenomic analyses (above) and pigment analyses (below)

Most cores reveal an increase in the exchangeable and organic P fractions at one or both PT2 and PT3.
Although it is tempting to interpret increases upcore of both these fractions as an indication of increasing trophic
status of the lake, this interpretation is complicated by the degradation of organic matter and high mobility of P
through the sediment pore water. The exchangeable fraction represents a very mobile fraction of P within the
sediment. Exchangeable P originates from the water column where it can be attached to particles that are
sedimenting, but it may also originate within the sediments from the degradation of organic matter or from the
reduction of Fe and Mn oxides. This fraction is thus expected to increase upcore as this soluble P fraction will move
against a gradient in concentration and attach to mineral or organic exchange sites. Organic-P is representative of
algal and bacterial cells as well as other labile biologic material. This fraction is also subject to post-depositional
change as the organic matter is degraded through time within the sediments, thus this fraction is also expected to
decrease downcore, even in the absence of an ecological shift.

The HCL fraction is meant to target CaCO3-P and is ostensibly one of the most stable fractions and can
integrate a water-column concentration of P, which is why it was selected for novel targeted analyses. However,
the standard sequential extraction method cannot isolate the authigenic fraction and is thus a combination of
recalcitrate organic matter, detrital Ca-P, and authigenic P. This method, however, still revealed interesting trends
that likely reflect the eutrophication of the lake environment, as depicted by the other proxies evaluated
(pigments, diatoms). Specifically, in the Goshen Bay core we see an abrupt increase in the HCL-P around PT3, and
both Bird Island and Provo Bay show an increase around PT2.

Despite the high mobility of P fractions within sediment core records, the absolute concentrations, or P
mass, in space (or across lakes) can be a more reliable metric of trophic variability. Among the four cores, Provo
Bay had the highest surface P concentrations in all fractions, suggesting higher modern P concentrations in this
area, which is consistent with contemporary Utah Lake water quality data (Utah Department of Environmental
Quality, 2021). The North Core location has the lowest overall concentrations of P in all fractions, which is expected
due to its distance from wastewater effluent and agricultural runoff. On average, however, Bird Island had the
highest TP concentrations throughout the record, with an average of 1030 µg g-1, followed by Goshen Bay at 840
µg g-1, Provo Bay at 710 µg g-1, and then North at 590 µg g-1. Though the timeframes for which sediments are
averaged here will differ between each core based on individual chronologies.

The most reliable metric of water column-P concentrations through time is derived from the targeted
analyses of CaCO3-P using the two novel approaches, specifically SEM-EDS, and the physical and chemical
separation of authigenic CaCO3. Both these analyses also revealed increases in P at both Phase Transition 2 and 3,
and more remarkably were correlated with metrics representing phosphors pollution from anthropogenic sources.
Both metrics revealed a doubling of the phosphorus concentrations captured within authigenic carbonates.

31

Nitrogen and carbon mass and isotopes
Methods
Elemental composition nitrogen (N) and carbon (C) and isotopic values of each (Talbot 2001) were
analyzed for the Goshen Bay core at the Utah State University (USU) Stable Isotope Laboratory. Sediment
subsamples (n = 41) were freeze-dried, treated with acid fumigation (HCl) to remove carbonates, packed into tared
tin capsules, and measured by combustion using a Costech 4010 Elemental Analyzer. Isotopic values (13C, 15N)
were reported in units per mil (‰) deviation from international standards (PeeDee Belemnite, atmospheric N2).
Elemental composition was estimated as % dry mass (N% and C%) and used to estimate C:N ratios. Because
different sources of nitrogen have distinct isotopic compositions, relative changes in 15N will be used to track
wastewater effluent (10-15‰) as compared to other sources including, inorganic fertilizers (0‰), atmospheric
deposition (<0‰), and natural terrestrial sources (0-5‰)(Finlay & Kendall, 2007). Carbon isotopes can also be used
to distinguish between different sources and in-lake processing (Meyers, 2003).
Results
Prior to European settlement (~1200-1869), analysis of bulk organic matter using C:N ratios (and Rock-
Eval pyrolysis) suggested that the clear-water state exhibited high abundance of benthic plants, which tend to have
high C:N ratios compared to turbid, planktonic-dominated regimes (Meyers 1994; Liiv et al. 2019). Macrophyte
tissues are characterized by C:N ratios > 20, while values between 10 and 20 typically reflect communities with
abundant phytoplankton or phytobentos (Talbot and Lærdal 2000). Although no abrupt change in C:N ratio was
detected, trends suggested slightly higher C:N in deeper sediments (Figure 13), a pattern indicative of vascular
plants rather than phytoplankton (Meyers and Teranes 2001). Similarly, while elevated δ13C values during Phase 1
are consistent with abundant macrophytes (e.g., Klamt et al. 2019), declines in C isotopes in more recently
deposited sediments may also reflect increased use of respired CO2 during phytoplankton blooms, as seen in other
large shallow lakes undergoing eutrophication (Bunting et al. 2016). Low δ15N values are consistent with other
shallow, eutrophic lakes prior to disturbance (Zan et al. 2012; Bunting et al. 2016).
At approximately 1869 we begin to see substantial shifts in organic C and N proxies. In Goshen Bay, the
significant increase in algal pigments and 15N values shortly after ~1869 (± 36 yrs; Figure 13) may be attributed to
increased input of human waste following exponential human population growth within the watershed (Fig. 3a).
Using population growth as a surrogate for wastewater nutrient delivery reveals that there is a moderate
correlation between societal development and sedimentary  15N (r = 0.31, p = 0.02) consistent with the hypothesis
that continued external nutrient loading may have favored transition to phytoplankton dominance. Nitrate derived
from raw sewage is typically known to be isotopically enriched to 10-25‰ relative to natural soils (2-5‰) and
atmospheric sources (0±3‰) (Kendall 1998). Thus, the shift in  15N from 3‰ to 7‰ is consistent with an
increasing contribution of N from sewage to the lake shortly after European settlement. Similarly, the significant
decrease in  13C is consistent with observations from other semi-arid regions, in which elevated primary
production draws on a pool of 13C-depleted respiratory CO2 to fuel eutrophication (Bunting et al. 2016).
Alternatively, the decline in  13C may reflect the Suess effect (Keeling 1979), in which anthropogenic activities (e.g.,
fossil fuel combustion) have altered the abundance of carbon isotopes in organic matter reservoirs, such as lake
sediment (Reavie et al., 2021).
After ~1945, Continued input of WWTP effluent from urbanization and population growth in the
watershed (Figure 13) has resulted in a gradual ecosystem transition in years following the initial abrupt change.
While many cities around the lake constructed WWTPs by 1954, these measures failed to remove dissolved N or P
and therefore continue to contribute to elevated 15N values. WWTPs continue to be among the major sources of
32

nutrient loading to Utah Lake, with levels sufficiently high to be non-limiting to primary producers (Merritt and
Miller 2016). Finally, consistent with recent monitoring, multiple proxies indicate that eutrophic conditions have
continued unabated during the 21st century. For example, 15N values increased to unprecedented levels in the
surface sediments, consistent with continued N influx from urban wastewater.
Elemental Composition
Sediment geochemistry can provide critical information on both changing sediment sources through time
as well as conditions at the sediment-water interface including lake mixing (Brahney et al., 2008). Given the
shallow nature of Utah Lake and extensive fetch, it is unlikely that the lake would stratify to the point at which
suboxic condition might develop near the sediment water interface for any substantive length of time. Note that
Fe-oxide precipitation occurs at a much faster rate than dissolution, thus extended anoxic in a stable hypolimnion
is required to allow for solubilized phosphate to accumulate and influence internal loading. However, if historically
widespread macrophyte beds were capable of stabilizing sediments and reducing water turbulence, intermittent
suboxic conditions might have occurred in a naturally mesotrophic system. Elemental ratios and redox sensitive
elements can be used to infer changes in oxygen concentrations at the sediment water interface (Brahney et al.
2008). Fe and Mn oxides are soluble in their reduced states and insoluble under oxic conditions (Engstrom and
Wright 1984). The normal sequence of electron acceptors along the redox ladder once O2 is consumed is NO3-,
MnOx, Fe(OH)3, SO42- , and then CO2. In general, an electron acceptor must be nearly used up before moving to the
next in the sequence (Schlesinger 1997). However, variation in the natural environment will exist because of the
patchy distribution of electron acceptors and relative reaction rates, e.g. the precipitation of Fe oxides under oxic
conditions occurs at a rate that is substantively faster than the dissolution under reducing conditions. Thus, the
presence or absence of an element is not as informative as the relative concentration through the core ratios of
the core (Brahney et al. 2008, Engstrom and Wright 1984) Accordingly, because Mn-oxides are preferentially
reduced compared to Fe-oxides, the ratio of these two elements can provide information on shifts between weakly
reducing and oxic conditions. In addition, trace elements are influenced by redox conditions through multiple
mechanisms including co-precipitation with iron oxides, sulfides, or associated with organic matter (Smrzka et al.,
2019). Because the elemental composition of iron oxide extractions will be most informative on shifting redox
states, especially if reducing conditions are mild, this will be the only extraction analyzed for elemental chemistry
from select cores. Though more information could be acquired through the analyses of the organic and mineral
fractions, the budget did not allow for this.
Methods
Sodium-citrate/dithionite [(Na3C6H5O7) (Na2S2O4)] was used as a reducing agent to extract the Fe-and Mn-
oxyhydroxides along with the associated trace elements. This treatment does not dissolve magnetite or silicates
(Ross and Wang 1993). ICP-MS was used to measure the extractions for the following elements, P, Ca, V, Cr, Mn,
Fe, Co, Ni, Cu, Ti Zn, Rb, As, Se, La, We, Mo, and U. Pearson correlation coefficients were determined and those
with coefficients greater than 0.45 were considered ‘correlated’, while those with correlations above 0.80 were
considered ‘strongly correlated’.
33

Results
Goshen
Iron and Manganese concentrations were correlated or strongly correlated with Co, V, Cu, As, and La. Fe
was also correlated with Ti and Mn with Rb. Fe was not correlated with P (r = -0.11). A significant shift occurs at
about 30 cm depth (~1880; PT2) that is suggestive of a shift from mildly reducing conditions to oxic conditions. The
Fe:Mn ratio decreases at 30 cm. A higher ratio is indicative or more reducing conditions because Mn reduction will
occur before Fe reduction on the redox sequence. In addition, at 30 cm many elements that can be scavenged be
Fe and Mn oxides increase, these include V, Co, Ni, Cu, Ti, As, and U. However, U displays a more distinctive patter
and increases more monotonically throughout the core as opposed to an abrupt shift. U can be remobilized from
the sediments if redox conditions fluctuate (Algeo and Maynard 2008), which may account for the relatively low
concentrations earlier in the sediment record. Mo is the only elements we measured that had greater
concentrations prior to the shift at 30 cm but fluctuates through the early period. Mo does not accumulate in
mildly reducing conditions, only strongly reducing conditions and this these fluctuations might reflect spatial
heterogeneity in redox conditions through space and time. Mo can also co-precipitate with Mn oxides, but we see
no relationship between Mn and Mo concentrations.
Bird Island
In the Bird Island sediments, Fe correlates with redox sensitive elements Co, Cu, Ti, Rb, As, U, and V, while
Mn correlates with V, Cu, Rb, La, and U. Fe does not correlate with the extracted P. At approximately 30 cm and
abrupt increase in many redox sensitive elements, including Fe, and Mn, occur, which is followed by a second peak
and increase in concentration that occurs around 15 cm. In contrast to Goshen Bay, there is no abrupt shift in the
Fe:Mn ratio. The relative increase in Fe and Mn as associated trace elements may be due to redox cycling within
the sediments, rather than shifts at the sediment water interface. An abrupt peak in Cu occurs around 60 and again
at 30 cm. Cu is often a good tracer of organic carbon remineralisation (Widerlund 1996 Piper and Perkinds 2004),

Figure 23. Elemental concentrations form the Goshen Bay core.
34

and may reflect a substantive die off vegetation, though this is speculative. As with the Goshen Bay core, elements
that can be scavenged be Fe and Mn oxides increase, these include V, Ti, As, Rb, and U.
Summary
Charge question 1.2 requested information on the historic P, N, Ca, Si, and Fe concentrations through the
sediment record, which we interpret as a request for historical information on the changing chemical and nutrient
requirement of the lake through time. Note the Si concentrations require an additional methodological approach
for which we did not have the funds to acquire.
Historical phosphorus concentrations in the Utah Lake environment were evaluated through several
mechanisms. First, we employed a standard sequential extraction method to all sediment cores evaluated at USU.
All cores revealed increases in sediment phosphorus concentrations through time, with distinct shifts occurring
around 30 and 10 cm in most cores, which align with Phase Transitions 2 and 3, which temporally align with Carp
introduction into the Utah Lake system and the more modern-day hypereutrophic conditions. Provo Bay had the
highest modern sediment P concentrations and the North core the lowest, which agrees with conditions present in
the water column as determined through DEQ monitoring. The most insightful records come from the targeted
CaCO3-P analyses, which revealed strong relationships between water column P and anthropogenic sources of
pollution as deduced from strong correlations to the Provo Bay population, metrics of sewage (15N), and
indicators of eutrophic conditions in the lake (water column cyanobacterial pigments). The latter analyses suggest
a doubling of water-column phosphorus through the sediment record.
Historical C and N concentrations as well as N isotopes suggest in an increase in the trophic status of the
lake and a shift to algal derived organic matter. In addition, 15N isotopes increase to values that indicate
contributions from waste effluent to the nitrogen pool, with the highest values occurring in sediments post 2000
AD. Elemental concentrations suggest that prior to PT2 (Carp introduction), lake sediments may have been more
stable, which allowed for greater Fe and Mn oxide reduction post-burial.
In summary, the above information suggests that prior to PT2, Utah Lake had significantly lower P ambient
water column P concentrations (possible reduced by a factor of 2). At this time, lake sediments appeared less likely

Figure 24. Elemental concentrations form the Goshen Bay core.
35

to be mobilized into the water column, which is likely due to the presence of stabilizing macrophyte vegetation.
Post PT3, several proxies suggest increasing eutrophication that appear to be associated with a growing population
in the Utah Lake catchment and the increase in nutrients delivered to the lake via wastewater effluent.
Charge Question 1.4. What do photopigments and DNA in the paleo record tell us about
the historical water quality, trophic state, and nutrient regime of the lake?
Methods
Photopigments
Photopigments were analyzed on the Goshen Bay and Bird Island cores (King et al. 2024). Once cores were
sub-sectioned (see Methods section under Core Collection), care was taken to minimize exposure to light. All
sediments were stored at 4°C. Fossil pigment concentrations were quantified by high performance liquid
chromatography (HPLC) at the University of Regina (Saskatchewan, Canada), following the methods outlined in
Leavitt and Hodgson (2001). Pigments from freeze-dried sediment subsamples (Goshen Bay n = 115, Provo Bay n =
36, Bird Island n = 37) were extracted, and individual compounds were isolated and quantified using an Agilent
model 1100 HPLC fitted with a photodiode array detector. The HPLC system was calibrated with authentic pigment
standards (DHI Denmark) and an internal standard dye (Sudan II). SedDNA was were measured as above.
eDNA
eDNA methods are provided in Section 1.1.
Results
Algal pigments
Concentrations of ubiquitous phototroph pigments that represent total primary production (e.g.
pheophytin a, β-carotene) increased significantly in all cores. The major abrupt change in the Goshen Bay core was
detected at ~30 cm, corresponding with PT2, while the abrupt change in the Provo Bay and Bird Island was
detected at ~27.5 cm and ~9.5 cm, respectively. Although abrupt changes were not detected in pheophytin a
concentration in Provo Bay, there was a continuous increase in inferred production towards surface sediments.
Concentrations of ubiquitous pigments exhibited a subsequent decrease to lower concentrations at ~20 cm in the
Goshen Bay core, but not to pre-impact conditions. No subsequent declines in concentration were observed within
Bird Island or Provo Bay cores.
Algal pigments associated with specific algal groups showed 5- to 10-fold increases from low, stable values
to higher, more variable values in all sediment cores at the same major depth horizons (Figure 23). In the Provo
Bay core, pheophytin b from chlorophytes exhibited a steady increase up-core, whereas aphanizophyll from
diazotrophic cyanobacteria exhibited a sharp increase near the surface of the core. In Goshen Bay, total
cyanobacteria (as echinenone) displayed two sequential increases in concentration, reaching a mean of 3.7 nmole
g-1 OM in recent years that was ~30-fold greater than baseline concentrations. At a similar depth to the third
stepwise increase in total cyanobacteria (echinenone) at ~15 cm, canthaxanthin from Nostocales cyanobacteria
also exhibited an increase to a mean of 5.3 nmole g-1 OM, six times greater than its baseline concentration, and
coeval with the third peak in echinenone (total cyanobacteria). In contrast, the biomarker for N2-fixing
cyanobacteria (aphanizophyll) was detected in the sediments for the first time starting in ca. 1967 (± 12 yrs) and
averaging 2.28 nmole g-1 in recent sediments. Concomitant with subsequent increases in pigments at ~15 cm,
diatoxanthin, mainly from diatoms, was the only compound to undergo a decrease in concentrations in recently
deposited sediments, suggesting a shift in relative taxa from diatoms towards green algae and cyanobacteria.
Despite this shift, concentrations of diatoxanthin remained over three times higher than in the pre-impact era.
Additional pigments were also detected in the core (Figure 23), but were either highly labile (e.g., Chl a,
fucoxanthin) or were associated with multiple algal groups (e.g., lutein from chlorophytes and cyanobacteria,
36

diadinoxanthin and fucoxanthin from diatoms and other siliceous algae) so were not considered further in this
study. Pigments concentrations were classified into there main phases, detailed below.
Summary
Phase 1: Unit 1 and 2 (Core bottoms to ~1890)
Analysis of phototrophic pigments provided insight into both phytoplankton production, and sedimentary
preservation through time. In general, phytoplankton production was relatively low as indicated by low
concentrations of all algal and cyanobacterial pigments in all cores, particularly chemically stable and ubiquitous
pheophytin a and β-carotene (Figure 25). Elevated ratios of labile Chl a to stable pheophytin a suggest that
pigment preservation was generally higher in Phase 1 at Goshen Bay than in subsequent time intervals, despite the
overall low pigment concentrations. Such elevated levels are often seen when phytobenthic populations are
healthy and pigments are produced nearer the sediment-water interface, as seen in other shallow lakes with clear
waters and extensive macrophyte populations (e.g., McGowan et al. 2005). In contrast, Chl:Pheophytin ratios often
decline following a shift to turbid conditions, as most pigment degradation normally occurs during phytoplankton
sedimentation (Leavitt and Hodgson 2001). Differences in the patterns of Chl:Pheophytin among coring sites may
reflect local variation in water-column depth relative to that of light penetration, such that routinely low pigment

Figure 25. Concentrations of photosynthetic pigments (nmoles pigment g-1 OM), as well as the degradation ratio
(Chl a:Pheo a), measured throughout the Goshen Bay, Provo Bay core, and Bird Island cores. Multivariate BCP
analysis estimated the posterior probability of a change point throughout the cores. Dashed lines indicate
significant abrupt changes as detected by STARS. Figure reproduced from King et al. 2024.
37

ratios at Bird Island are consistent with the greater depth, and presumptive lack of macrophytes suppressing
sediment turbidity at this locale.
Phase 2: Unit 3 (~1890-1945)
Many fossil pigment concentrations, and by inference planktonic primary production, increased abruptly
despite likely declines in pigment preservation (Figure 25). The abrupt decrease in the preservation ratio in Goshen
Bay at PT2 (Chl a: Pheophytin a) indicates a fundamental change in the lake environment consistent with increased
degradation of labile compounds. Although speculative, the change from benthic to planktonic dominance may
have increased pigment losses during sinking and resuspension, as demonstrated in a wide variety of lake types
(Leavitt and Hodgson 2001, and references therein). In most lakes, degradation and transformation of
phytoplankton pigment occurs mainly during deposition and is strongly affected by the path length to the
sediments. Consequently, shifts from benthic to planktonic production are often associated with increased loss of
labile components (e.g., Chl a) particularly during state changes in shallow ecosystems (e.g., McGowan et al. 2005).
In Utah Lake, sediment resuspension likely contributes to increased phytoplankton degradation.
Analysis of fossil pigments and sedDNA indicated an increase in phytoplankton production coeval with loss
of macrophytes. Despite changes in fossil preservation, overall algal pigment concentrations are notably greater,
including diverse stable carotenoids known to be less subject to losses during deposition or incorporation into the
fossil record (Leavitt and Hodgson 2001). Diatoms (as diatoxanthin), chlorophytes (pheophytin b), and
cryptophytes (alloxanthin) all exhibited significant increases during this phase in all cores. The loss of macrophytes
may have led to a substantive change in nutrient cycling that also favored greater algal biomass (e.g., Klamt et al.
2019). Additionally, total cyanobacteria (echinenone) displayed a significant increase during this phase, which is
corroborated by the increasing frequency of cyanobacterial sedDNA. Elevated cyanobacterial abundance was
aligned with a decrease in relative abundance of hardstem bulrush in macrophyte sedDNA, as well as
disappearance of macrophyte and gastropod remains. In Goshen Bay, the significant increase in algal pigments
(and δ15N values) shortly after PT2 may be attributed to increased input of human waste following exponential
human population growth within the watershed.
Phase 3 (~1945-2001)
In Goshen Bay, cyanobacterial pigments (echinenone, canthaxanthin) exhibited a second transition to
increased concentrations in the late 20th century (Figure 25. This increase is strongly correlated with the growing
population in the region (echinenone: r = 0.81, p < 0.001; canthaxanthin: r = 0.74, p < 0.001; Fig. S12), and resulting
increase in nutrient influx at the time. Similarly changes in concentrations of aphanizophyll from N2-fixing
cyanobacteria were correlated with human population growth (r = 0.63, p < 0.001), which aligns well with further
transitions observed in other pigments. Rises in aphanizophyll are often associated with blooms of Aphanizomenon
flos-aquae, the taxon currently responsible for many HABs in Utah Lake (Randall et al. 2019).
In contrast to cyanobacteria, siliceous diatom population change (as diatoxanthin) was correlated
negatively with human population growth (r = -0.57, p < 0.01) and exhibited a decrease in concentration suggesting
that cyanobacteria began outcompeting diatoms. Given that β-carotene and pheophytin a (total primary
production), declined, whereas pigments cyanobacteria (echinenone, canthaxanthin) were relatively stable, we
infer that there was a substantial shift from eukaryotic to prokaryotic phytoplankton. Such a shift is known from
other lakes undergoing pronounced state change and may reflect the replacement of negatively-buoyant diatoms
with floating cyanobacterial colonies (Bunting et al. 2016). Regardless, pigment concentrations never returning to
baseline levels seen in the early 1800s demonstrates that Utah Lake remains impacted by urbanization.
Phase 4: Unit 4 (~2001 to coretop)
Concentrations of stable pigments, including those indicative of all primary producers (β-carotene), total
cyanobacteria (echinenone), chlorophytes (and pheophytin b), and cryptophytes (alloxanthin), also displayed
elevated levels in the past 20 years or so (Figure 25), while the frequency of cyanobacteria in sedDNA samples
reached a maximum after the turn of the century. Although post-depositional degradation of organic biomarkers is
38

expected in most lake sediments, both pigment and sedDNA patterns are also consistent with studies documenting
an increase in the frequency of HABs in Utah Lake during recent years (Randall et al. 2019). Additionally, two
cyanobacteria genera, Synechococcus and Cyanobium, identified by sedDNA are also known to produce toxins that
can have negative effects on animals and humans (Jakubowska and Szeląg-Wasielewska 2015). Similarly, the
detection of small pondweed, sago pondweed, and water knotweed by sedDNA in recent sediments is consistent
with recent increases in pondweeds attributed to successful carp removal efforts (Dillingham 2022).
Supplementary Data Methods and Results
Pollen

To determine if changes in vegetation are linked to ecological changes in Utah Lake, and, to provide
chronological sediment marker horizons, pollen assemblages were. Pollen analysis was used as a proxy for
reconstructing both terrestrial and aquatic vegetation change over time. Sediment samples (1cc each) were
obtained at ~0.2mm intervals for the upper 24 cm and averaged ~1.5 cm intervals from 24-46 cm. Pollen
processing follows the methods of Faegri et al. (1989). One Lycopodium tablet was added to each sample as an
exotic tracer. 300 terrestrial grains were counted for each sample processed, counts converted to percentages of
the total terrestrial grains.
Pollen analysis demonstrates that the vegetation surrounding Utah Lake has been a stable, sagebrush-
steppe over the last ~600 years (Figure 16a). A typical sagebrush-steppe is dominated by sagebrush (Artemisia),
with grasses (Poaceae family), amaranths (Amaranthaceae family), and many herbs within the Aster family
(Asteraceae), of which have comprised ~50% of the pollen record over the last ~600 years. Additionally, Sagebrush-
steppe ecosystems in Utah are generally found at low-to-mid elevations, below 1800 m, adjacent to the pinyon-
juniper woodland ecosystem. Pollen analysis suggests the surrounding pinyon-juniper ecosystem has been
prevalent around Utah Lake over the last ~600 years, with ~20% of the pollen found coming from this ecosystem
type.

39

Figure 26 Terrestrial pollen taxa from Utah Lake plotted as percentages over time (and depth), separated into; introduced and non-native taxa, common tree and shrub pollen taxa, and
herbaceous taxa. Also included are the summaries for each category and micro- charcoal from pollen slides. Shaded area equals 10X exaggeration of specific taxa
40

Aquatic Pollen

Within Utah Lake, pollen from aquatic vegetation such as water milfoil (Myriophyllum), cattail (Typha
latifolia) and sedges (Cyperaceae), along with green algae, especially representatives of the
genera Botryoccocus, Pediastrum and Tetraedron suggest a shallow expanse of open water prior to ~1850 (Wacnik,
2009) (Figure 16b). High abundances of Tetraedron also suggests highly productive waters prior to
1850. Tetraedron is highly efficient in the use of HCO3− (i.e., biocarbonate) as a substrate for photosynthesis in
more productive waters (van Hunnik et al., 2000).

Figure 27 Aquatic pollen taxa from Utah Lake plotted as percentages over time, separated into; Aquatic taxa, including algal and fungal
types. Also included are the summaries for each category and micro-charcoal from pollen slides. Shaded area equals 10X exaggeration of
specific taxa.
41

However, starting in the mid-1800s, major changes in both the terrestrial and aquatic ecosystems are reflected in
the pollen record. Specifically, the introduction of non-native tree species, such as holly (Ilex), walnut (Juglans),
mulberry (Morus), hemlock (Tsuga) and elm (Ulmus) appear. Additionally, the appearance of cultivars,
primarily Cerealia-type grasses and Secale-type grass suggests agricultural activity near the lake.
Lastly, Plantago and Rumex species also appear mid-1800s, which are plants introduced with agriculture (Deza-
Araujo et al., 2020). Within Utah lake, a trophic change was observed starting mid-1800s. Furthermore, the
presence of Pediastrum integrum indicates periods of dystrophy within Utah Lake and the decline
of Tetraedron during the past ~150 years suggests a shift toward eutrophication..

Charcoal
2.3.2.2.1 Charcoal Analysis PI Power

Macroscopic charcoal analysis was used as a proxy for reconstructing local fire activity. Macroscopic charcoal
was sieved through mesh screens and tallied to determine the charcoal influx (particles cm-2 yr-1). Approximately ~1
cm-3 sediment sample was obtained from both North freeze core (19-N-01-FC) and the south freeze core (19-S-01-
FC) at ~2-5 mm sediment depth intervals. The samples were disaggregated with potassium hydroxide, washed
through a 125 µm sieve, identified, and counted with a dissecting microscope at 35-40X magnification. Other
identifiable macroscopic plant or animal remains were recorded. Macroscopic charcoal counts were then
transformed to concentrations (particles cm−3) and then to charcoal accumulation rates (CHAR; the number of
particles cm−2 yr−1). Macroscopic charcoal counts were then broken into a low-frequency component, known as
charcoal background (BCHAR), and a peaks component (CHARpeak) using an interpolated temporal resolution of 4-
years sample-1 using CharAnalysis (Higuera et al., 2009). BCHAR was determined using a robust LOWESS regression
with a moving-window width of 50 years which resulted in a robust signal-to-noise index (SNI) > 3.0 (Kelly et al.,
2011). CHARpeak values were calculated as residuals from BCHAR and were separated from the background noise
with a locally defined threshold using the 99th percentile of a Gaussian mixture model. The identified CHARpeaks
were then screened with a minimum count peak-screening test (Higuera et al., 2010). Microscopic-charcoal was used
to reconstruct extra-regional fire activity and was counted simultaneously with pollen analysis at a magnification of
400× and are presented as raw charcoal counts.
Macroscopic charcoal influx from the North core suggest relatively low fire activity locally prior to ~1850
and then a 3-fold increase in fire activity ~1850 (Figure 27). Fire activity substantially increases locally until the mid-
to-late 20th century, a likely response to active fire suppression (Heyerdahl et al., 2011) and potentially reduced
agricultural burning in the watershed. However, fuel build up resulting from fire suppression has likely resulted in
an overall increase in fire activity in western US landscapes (Kreider et al., 2024) and may partially contribute to
variation in the north core fire history reconstruction. The microscopic, or pollen slide charcoal (Figure 28a, b), also
captures alterations in regional fire activity occurring post ~1850.
42

Figure 28 Charred plant materials are summarized to provide a fire history for the past 600 years. Charcoal particles were tallied for two
cores; the North core and South core (lower panels). North core charcoal particles were converted to influx (particles cm-ryr-1) and analyzed
with CharAnalysis software (Higuera et al. 2009) to separate long term trends, BCHAR (LOESS smoother) from high-frequency “Fire episodes”
or CHARpeak (shown as vertical bars, and plotted as fire frequency or # of episodes per 100 years).
N-15 enrichment and wildfire activity
The link between wildfires and nitrogen in lacustrine sediments offers one of many possible drivers of
nutrient enrichment. The transformation of organic nitrogen into inorganic nitrogen during a wildfire is typically
accompanied by increased 15N values in sediments because of kinetic fractionation associated with nitrogen
volatilization (Grogan et al., 2000). Prior to ~1850 CE, the shifts in 15N values of Utah Lake were likely partially
influenced wildfires occurring near the lake or within tributary streams, including the Provo River watershed.
Elevated charcoal influx and 15N during the 16th century captures prehistoric natural variability in the watershed
(Figure 18). This trend is consistent with other observations suggesting wildfires generate an initial pulse of
available nitrogen that can persist for several months after a fire. Following these events, net nitrogen
mineralization resets the baseline for subsequent increases in available nitrogen after a fire (Monleon et al., 1997).
Wildfires in the watershed promote the pyloric release of nitrogen into surrounding soils primarily as ammonium
(NH4+), but may also produce nitrogen in the forms of nitrate (NO3-), and nitrite (NO2-) (Grogan et al., 2000;
Gimeno-Garcia et al., 2000; Wan et al., 2001). It is likely the elevated nutrients, charcoal, and combusted organic
materials are flushed down tributary streams during spring snowmelt (Piatek et al., 2005) and likely constitute the
a punctuated source of nutrient supply to Utah Lake. The record from the north core shows an overall enrichment
of 15N in Utah Lake following increased fire activity during the last ~150 years.
43

Figure 29 Charcoal influx (particles cm-2 yr-1) as a metric of wildfires are compared to 15N results from the North frozen sediment core.
STARS regime detection (blue dashed line) was applied to 15N data and captures a significant change in both charcoal influx influx and 15N
data during the late 19th century. The previous centuries capture low variability in fires and 15N, with the only pre-settlement notable increase in
both fire and 15N occurring during the 1500s, well before the arrival of Euro-American settlers to Utah Lake. Large “peaks” in charcoal influx in
recent decades (extending beyond the x axis) exceed the historic natural range of variability.

Reference Condition of Utah Lake (Pre-Disturbance)
Paleoecology is a powerful and well-established retrospective method that can be used to describe
baseline conditions of a lake ecosystem, including the physical, biological, and chemical environment. In addition,
relative changes amongst proxies and their temporal associated with specific drivers can give insights into which
key climatic or anthropogenic activities are associated with ecosystem shifts. Here we used a multi-proxy approach
to reconstruct the recent history and baseline condition of Utah Lake when European settlers arrived.
1. Large areas of Utah Lake had substantially more macrophyte cover than
present
There are multiple lines of evidence that point to a greater presence of macrophyte beds in Utah Lake. In
particular, it appears that Goshen Bay and Provo Bay had extensive macrophyte cover. Bird Island and the North
Core show evidence that macrophyte beds were proximal to the coring locations, but did not exist at the coring
locations. The evidence for each core is as follows:
Goshen Bay
• We recovered abundant physical remains of macrophyte vegetation from the core prior to ~1900, which is
dated approximately to the time of Carp introduction, which represents the transition to Unit 2 and we
refer to here as Phase Transition 2– (Figure 3).
44

• We recovered the remains of gastropods in most core sections prior to Phase Transition 2~1900 (Figure 3)
• Metagenomic data revealed the presence of hardstem bulrush prior to Phase Transition 2
• An abrupt shift in the Rock Eval Hydrogen Index pre 1900 marked a transition from plant to algal organic
material in the sediments at Phase Transition 2
• Pigments in the sediment cores showed a marked decrease in preservation around Phase Transition 2,
which suggests that prior to this transition sediment conditions were conducive to preservation which
include greater stability and reduced oxygen. After this transition were more frequently mixed into the
water column and exposed to light and oxygen.
• At Phase Transition 2 there is a decrease in the Fe:Mn ratio and increase in oxyhydroxide scavenged
elements suggestive of a sedimentary environment that shifted from a stable lower oxygen environment
to an oxygenated environment, which is expected to occur at the loss of macrophyte beds.
• The greater presence of epiphytic diatoms prior to Phase Transition 2

Provo Bay
**SedDNA, elemental analyses, and Rock-Eval were only performed in Goshen Bay
**Diatom analyses was not performed in Provo Bay
• The Provo Bay sediment record shows a marked increase in plant material prior to the Phase Transition 2
(Figure 3)
• The abundant presence of gastropod shells found below Phase Transition 2
Bird Island
There is limited evidence for the presence of macrophyte beds at the Bird Island coring location, however
there is evidence that macrophyte beds must have been at least proximal to the location. The evidence is as
follows:
• There was a treater abundance of epiphytic diatoms prior to Phase Transition 2
• The presence of gastropods shells prior to Phase Transition 2
• The increase in Cu is a potential indicator of a mass die-ff of vegetation nearby or atmospheric deposition
from mining activities in the region

North of Provo Bay
North Lake
No evidence currently exists for macrophyte cover near the North core site, noting that only macrofossil
analyses were evaluated during core sectioning and no other indicative proxies were evaluated for this core. A
singular Anodonta shell was recovered from this core just below Phase Transition 2.
Summary
Based on the above, we have a high degree of confidence that Goshen Bay had greater representation of
macrophytes such that a substantial reorganization of the ecosystem was observed post colonization. There is also
strong evidence to support the historical presence of greater macrophyte cover in Provo Bay. The presence of
gastropods in one but not both Bird Island cores suggests some localized areas of vegetation may have been
nearby. Finally, there is no evidence to support macrophyte dominance at the North core site, with the
acknowledgement that the only proxy evaluated for this site was the physical presence or absence of macrofossils.
It remains likely that some areas of the lake, but not the lake in its entirety, has greater representation of
45

macrophyte habitats, while other areas of the lake may have been more shallow and therefore turbid. This view
aligns with the historical records that suggest Goshen and Provo Bay were macrophyte dominated while the outlet
of the lake was recorded to be quite turbid, which may indicate that in general Utah Lake had a greater diversity of
habitat types. In addition, reduced pigment preservation and a shift in redox sensitive elements after the abrupt
change in Goshen Bay may reflect the increased mixing and aeration of the sediment associated with the loss of
stabilizing vegetation in the benthos (Rivera et al. 2013). Furthermore, a significant change to higher HI values (Fig.
3b), associated with increased algal components and decreased plant components within the organic matter,
provides further evidence of macrophyte loss.
2.0 Large areas of Utah Lake were less turbid and allowed for greater light
penetration to the sediment environment
There are several lines of evidence that suggest conditions in the Utah Lake water column allowed for
greater light penetration and were thus not as turbid historically. These include,
• A loss of benthic diatom species in all cores evaluated (Bird Island and Goshen Bay)
• A loss of macrophyte vegetation in Goshen Bay and Provo Bay
• A decline in sediment preservation of fossil pigment in Goshen Bay
Summary
The above evidence suggests that areas of Utah Lake prior to European settlement was less turbid than
modern day Utah Lake. Not only is there evidence of macrophyte cover, which would require light penetration to
location of germination at the sediments-water interface, but there was also a significant loss of benthic diatom
species that either live on vegetation or on sandy sediments. Finally, there are sediment proxies that suggest
historic conditions were favorable to pigment preservation including reduced oxygen and light, which could only
occur if sediments remained stable and were not mixed into the water column.
3.0 The present-day hypereutrophic condition combined with toxic cyanobacteria
blooms is unprecedented in the history of Utah Lake
Several lines of evidence suggest that while the lake was likely historically mesotrophic with abundant
macrophyte cover, the community composition was representative of a more functional lake ecosystem. The
evidence for a transition to a eutrophic system are as follows, while details and explanations are provided through
the document:
• Transition from macrophyte dominance to algal dominance (Goshen Bay, Provo Bay)
• Loss of pollution sensitive species such as Anodonta (Bird Island, Provo Bay, North Core)
• Increase in HCL-P (all cores)
• Increase in water-column P concentrations as determined by CaCO3-P targeted analyses (Goshen
Bay)
• Increase in %C and N in the sediments suggestive on increased production.
• Increased values of 15N suggestive of effluent contributions to the lake environment.
• Increased representation of cyanobacteria in the metagenomic analyses
• Increased representation of cyanobacteria and green algae in the pigment analyses
• Decreased abundance of diatom pigments
• A shift in the diatom community from pollution sensitive species to pollution tolerant species
• A loss of diatom species that live on plants and within the sediments.
• An increase in pigment concentrations suggesting an increase in overall lake production.