Isotopic and Geochemical Tracers of Groundwater Flow in the Shivwits Plateau, Grand Canyon National Park

University of Kentucky UKnowledge

Theses and Dissertations--Earth and Environmental Sciences Earth and Environmental Sciences

2020

ISOTOPIC AND GEOCHEMICAL TRACERS OF GROUNDWATER FLOW IN THE SHIVWITS PLATEAU, GRAND CANYON NATIONAL PARK

Jonathan W. Wilson University of Kentucky, [email protected] Author ORCID Identifier: https://orcid.org/0000-0002-2341-8886 Digital Object Identifier: https://doi.org/10.13023/etd.2020.488

Right click to open a feedback form in a new tab to let us know how this document benefits ou.y

Recommended Citation Wilson, Jonathan W., "ISOTOPIC AND GEOCHEMICAL TRACERS OF GROUNDWATER FLOW IN THE SHIVWITS PLATEAU, GRAND CANYON NATIONAL PARK" (2020). Theses and Dissertations--Earth and Environmental Sciences. 85. https://uknowledge.uky.edu/ees_etds/85

This Master's Thesis is brought to you for free and open access by the Earth and Environmental Sciences at UKnowledge. It has been accepted for inclusion in Theses and Dissertations--Earth and Environmental Sciences by an authorized administrator of UKnowledge. For more information, please contact [email protected]. STUDENT AGREEMENT:

I represent that my thesis or dissertation and abstract are my original work. Proper attribution has been given to all outside sources. I understand that I am solely responsible for obtaining any needed copyright permissions. I have obtained needed written permission statement(s) from the owner(s) of each third-party copyrighted matter to be included in my work, allowing electronic distribution (if such use is not permitted by the fair use doctrine) which will be submitted to UKnowledge as Additional File.

I hereby grant to The University of Kentucky and its agents the irrevocable, non-exclusive, and royalty-free license to archive and make accessible my work in whole or in part in all forms of media, now or hereafter known. I agree that the document mentioned above may be made available immediately for worldwide access unless an embargo applies.

I retain all other ownership rights to the copyright of my work. I also retain the right to use in future works (such as articles or books) all or part of my work. I understand that I am free to register the copyright to my work.

REVIEW, APPROVAL AND ACCEPTANCE

The document mentioned above has been reviewed and accepted by the student’s advisor, on behalf of the advisory committee, and by the Director of Graduate Studies (DGS), on behalf of the program; we verify that this is the final, approved version of the student’s thesis including all changes required by the advisory committee. The undersigned agree to abide by the statements above.

Jonathan W. Wilson, Student

Dr. Andrea M. Erhardt, Major Professor

Dr. Michael M. McGlue, Director of Graduate Studies ISOTOPIC AND GEOCHEMICAL TRACERS OF GROUNDWATER FLOW IN THE SHIVWITS PLATEAU, GRAND CANYON NATIONAL PARK

THESIS

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in the College of Arts and Sciences at the University of Kentucky

By Jonathan W. Wilson Lexington, KY

Co-Directors: Dr. Andrea M. Erhardt, Assistant Professor, Earth and Environmental Sciences Dr. Benjamin Tobin, Hydrogeologist, Kentucky Geological Survey Lexington, KY 2020

ISOTOPIC AND GEOCHEMICAL TRACERS OF GROUNDWATER FLOW IN THE SHIVWITS PLATEAU, GRAND CANYON NATIONAL PARK

As the impacts of global climate change on water resources continue to become more apparent, proper understanding and management of groundwater resources will be needed as supplies become more strained. Traditional methods of characterizing groundwater systems are time-intensive, costly, and can be difficult to complete in remote areas. Using ambient geochemical tracers from discrete sampling could aid in characterizing spring systems through determining flow paths, recharge areas, and carbon cycling. However, using discrete seasonal samples to understand the hydrogeology of complex, mixed-lithology aquifers has not been extensively examined. Here we explore using δ13C of dissolved organic carbon (DOC), δ13C of dissolved inorganic carbon (DIC) and fluorescent dissolved organic matter (fDOM), together with water isotopes, major ions, and geochemical modeling, to characterize springs of the Shivwits Plateau in Grand Canyon National Park. Values of carbon isotopes and fDOM for all springs reflect source values for regional surface vegetation and heterotrophic degradation of terrestrial DOM. Principal component analyses show that springs can be grouped into four groups by geochemical variability: 1) a shallow epikarst system, 2) a flow path through gypsiferous beds of the Toroweap Formation on the eastern side of the plateau, 3) a short, canyon slope runoff-dominated flow path through the Supai Group, and 4) a deeper complex flow system in the Redwall Limestone with characteristics of all other flow systems, which indicates mixing. Results show that the methods used can provide a simple conceptual model of a complex groundwater system, but higher–resolution spatial and temporal data are needed to fully understand changes resulting from changing climate. As appropriations from the Colorado River already exceed its annual streamflow and the regional climate is predicted to become more arid, characterizing groundwater resources for water supply will be paramount for the region as well as in other areas that will experience similar transitions. KEYWORDS: Hydrogeology, stable isotope, geochemistry, Grand Canyon, springs ______Jonathan W. Wilson

______23 November 2020______ISOTOPIC AND GEOCHEMICAL TRACERS OF GROUNDWATER FLOW IN THE SHIVWITS PLATEAU, GRAND CANYON NATIONAL PARK

By Jonathan W. Wilson

______Andrea Erhardt Co-Director of Thesis

______Benjamin Tobin Co-Director of Thesis

______Michael M. McGlue Director of Graduate Studies

______23 November 2020 Date ACKNOWLEDGMENTS

I would like to thank my advisors Dr. Andrea Erhardt and Dr. Benjamin Tobin, and committee member Dr. Alan Fryar for the opportunity to conduct this research and for their assistance, guidance, and support throughout the completion of this thesis. Dr.

Erhardt initiated my interest in stable isotope geochemistry and has provided amazing guidance and support throughout my academic career and helped me grow as a geochemist. Dr. Tobin provided me with the chance to work with him on an amazing project and provided invaluable guidance and assistance in field work and feedback for this thesis and comments on other writings. Finally, Dr. Fryar provided ideas for analyses and exceptional feedback for the thesis.

The faculty and students of the Department of Earth and Environmental Sciences all provided support throughout my time with the department. I would like to thank Dr.

Jordon Munizzi and the whole of the Kentucky Stable Isotope Geochemistry Laboratory for the extremely helpful assistance in geochemical analyses and method development for this thesis. Pete Idstein provided assistance whenever needed for field or laboratory equipment and Adrianne Gilley for all things logistical with departmental management.

Finally, this research would not have been possible without the assistance in field work from Stefan Christie, Nicholas Steele, Natalie Jones, Adam Nolte, Vanessa

Fichtner, Sarah Arpin, Hannah Chambless, Matt and Saj Zappitello, Andres Hernandez,

Dr. Gina Lukoczki, and others who assisted.

iii TABLE OF CONTENTS ACKNOWLEDGMENTS ...... iii LIST OF TABLES ...... vi LIST OF FIGURES ...... vii CHAPTER ONE: INTRODUCTION ...... 1 1.1 Background ...... 1 1.2 Study area ...... 5 1.3 Hydrogeologic setting ...... 7 13 13 1.4 δ CDIC and δ CDOC as tracers ...... 11 CHAPTER TWO: METHODS ...... 14 2.1 Sample collection ...... 14 2.2 Stable isotope analyses ...... 15 2.2.1 δ2H and δ18O ...... 15 2.2.2 δ13C of soil samples ...... 16 2.2.3 δ13C and δ18O of carbonate in rock samples ...... 17 2.2.4 δ13C of dissolved inorganic carbon in spring samples ...... 18 2.2.5 δ13C of dissolved organic carbon in spring samples ...... 19 2.3 DOM absorbance and fluorescence analyses ...... 20 2.4 PARAFAC modeling ...... 21 2.5 Other analyses ...... 22 CHAPTER THREE: RESULTS ...... 23 3.1 Spring physical, chemical, and isotopic parameters ...... 23 3.2 δ13C of rock and soil samples ...... 29 3.3 DOM absorbance and fluorescence analyses ...... 31 3.4 Principal component analysis ...... 34 CHAPTER FOUR: DISCUSSION ...... 36 13 13 4.1 Viability of δ CDIC and δ CDOC as tracers ...... 36 4.2 Utilization of fDOM to characterize DOM ...... 40 4.3 Conceptual flow system of the Shivwits Plateau ...... 42 4.4 Seasonality of springs...... 46 CHAPTER FIVE: CONCLUSIONS ...... 49 APPENDICES ...... 51 APPENDIX A: Photographs of springs ...... 51

iv APPENDIX B: Remaining geochemical data ...... 61 REFERENCES ...... 64 VITA ...... 75

LIST OF TABLES

Table 1: Summary of sampling times, location, discharge, and basic water chemistry .. 24 Table 2: Summary of spring geochemistry and major ions ……………………………. 25 Table 3: Summary of spring major anions and saturation indices ……………………... 26 Table 4: Summary of spring stable isotope values …………………………………….. 27 Table 5: Results and characterization of PARAFAC components …………………….. 32 Table 6: Results of dissolved organic matter analyses ………………………………… 33

LIST OF FIGURES

Figure 1: Map setting of study area and previous data ………………………………….. 5 Figure 2: Geologic map and spring locations …………………………………………… 8 Figure 3: Cross-section of study area geology and conceptual flow tracing methods …. 12 Figure 4: Piper plot of spring results …………………………………………………… 23 Figure 5: Plot of stable water isotope values and meteoric water lines ………………... 28 Figure 6: Plot of isotopic results of rock samples ……………………………………… 29 Figure 7: Map of vegetation cover and soil isotope results ……………………………. 30 Figure 8: Results of PARAFAC analyses ……………………………………………… 31 Figure 9: Plot of principal component analysis ………………………………………... 35 Figure 10: Terrestrial DOM and stable isotope correlations …………………………... 39 Figure 11: Conceptual flow systems of the Shivwits Plateau ……………………….…. 45

13 Figure 12: Seasonality of δ CDOC ……………………………………………….…….. 47

vii CHAPTER ONE: INTRODUCTION 1.1 Background

Characterizing groundwater resources in arid regions is important as they are

often the sole source for water resources. In the southwestern United States, surface-

water supplies are limited and expected to become more stressed with groundwater

supplies declining at 0.3 to 3 m per year as the region transitions to a more arid climate

(Zektser et al. 2005; Seager et al. 2007). The Colorado River is the main water source for

the region but has the most complete allocation of its water resources of any river system

in the world and cannot be relied upon for further use, thus groundwater is a growing

alternative (Christensen et al. 2004). As of 2015, groundwater withdrawals comprised approximately 25% of total water use in the United States, up from 21% in 2005 (Maupin and Barber 2005; Lovelace et al. 2020). Groundwater provides drinking water for more than 50% of the US population, with global use of groundwater tripling between 1970 and 2005 (Zektser and Everett 2000; Zektser et al. 2005). Groundwater depletion as a result of increasing demand due to climate change and population growth has long been

recognized as a problem and examined in water-stressed areas worldwide (Chandrakanth and Romm 1990; Al-Sakkaf et al. 1999; Konikow and Kendy 2005; Siebert et al. 2010;

Famiglietti 2014). As aridification and reliance on groundwater supplies increases regionally, proper understanding and management of groundwater resources will be needed.

One of the methods for characterizing water resources is through analyzing components of the carbon cycle. In general, modern carbon enters the biological system through photosynthetic assimilation of atmospheric CO2, while weathering of carbonate

1 - and aluminosilicate minerals to HCO3 mobilizes abiotic carbon (Cole et al. 2007). This

carbon is transported in aqueous systems as dissolved inorganic carbon (DIC), dissolved organic carbon (DOC), and dissolved organic matter (DOM). Stable carbon isotopes in

DIC and DOC are fractionated, becoming either more depleted or enriched in 13C

throughout carbon cycling allowing them to be used as tracers. The normalized ratio of

13C/12C is given as δ13C, where more positive values describe a higher ratio of 13C/12C,

and more negative values have a lower ratio. Isotope ratios are expressed in per mil (‰) δ

form as δX = (Rsample/Rstandard – 1) where X is the isotope of interest and R is the molar ratio of the heavy isotope to the light isotope.

DOC is composed of organic acids and carbohydrates that have leached from vegetation and soils in the source watershed (Florea 2013). Modern vegetation can use one of three metabolic pathways—C3, C4, or Crassulacian acid metabolism (CAM)— with δ13C ranges of –35 to –23‰, –15 to –10‰, and –20 to –10‰, respectively, due to

discrimination during uptake of CO2 (O’Leary 1988; Kingston et al. 1994). If a watershed

has mixed vegetation types, the lack of in–stream microbial modification in surficial

streams means that the δ13C of DOC may allow for determination of source recharge areas (Voss et al. 2017). DIC enters aqueous systems through dissolution of atmospheric

13 13 CO2 (δ C = –8‰), soil CO2 (δ C = –10 to –35‰, depending on vegetation type), or soil

or rock carbonates which have an average value of 0‰ (Amundson et al. 1998; Doctor et

13 al. 2008; Voss et al. 2017). δ CDIC fractionates along flow paths as CO2

enrichment/degassing, bedrock dissolution, microbial uptake, and other processes occur

(Florea et al. 2019). These endmember values and fractionation processes along flow

paths can elucidate source values for DIC as well as assist in grouping spring systems

that experience similar fractionation mechanisms. Finally, DOM is the fraction of organic matter in natural waters that is composed of humic substances (humic acids, fulvic acid, and humins) derived from plant material and amino acids in proteins and peptides derived from bacteria (Hudson et al. 2007). DOM is composed of ~50% DOC with the remaining material being nitrogen, oxygen, hydrogen, and trace amounts of other elements (Moody and Worrall 2017). DOM can be characterized using fluorescence spectroscopy and parallel factor analysis (PARAFAC) in order to determine whether DOM is predominantly humic-like or protein-like. This type of analysis has been used as a tracer for sources of DOM (Stedmon et al. 2003), estimating biological activity (Parlanti et al.

2000), and for characterizing DOM to determine contaminant mobility (Kulkarni et al.

2017).

Traditional hydrochemical methods of characterizing groundwater systems use ambient and/or introduced tracers to characterize sources of recharge, flow paths, travel times, and reactions during flow. Ambient tracers include various solutes, stable isotopes of water (δ2H, δ18O) and solutes (e.g., δ13C, δ15N, δ34S), radioisotopes (e.g., 3H, 14C, 36Cl),

and noble gases, whereas introduced tracers include salts and fluorescent dyes (Käss

1998). However, use of many of these tracers can be time-intensive, costly, and

13 13 impractical in remote areas. Use of DIC and δ C of DIC (δ CDIC) as a tracer in

groundwater often requires high-resolution temporal sampling (Amiotte-Suchet et al.

1999; Hélie et al. 2002; Han et al. 2010; Yamanaka 2012; Peyraube et al. 2013), collection of possible surficial recharge endmembers (Atekwana and Richardson 2004), or coupled with 14C dating (Fritz et al. 1978; Nowak et al. 2017). Like DIC parameters,

13 14 DOC and δ CDOC are often coupled with C dating and high-resolution sampling, or

analyses of specific fractions of organic solutes (Wassenaar et al. 1990; Schiff et al. 1997;

Raymond and Bauer 2001; Deirmendjian et al. 2018). Finally, fluorescent dissolved

organic matter (fDOM) has been extensively used to characterize OM sources and

cycling in groundwaters, but it is not often coupled with isotopic parameters in multi- tracer studies (Birdwell and Engel 2010; Simon et al. 2010; Pavelescu et al. 2013; Huang et al. 2015; Kulkarni et al. 2017). Multi-parameter tracing with stable water isotopes and

13 δ CDIC has been done successfully with high-resolution temporal sampling (Adinolfi

13 Falcone et al. 2008), but seasonal sampling using DOC and δ CDOC as additional

parameters does not seem to have been conducted.

Here I explore using δ13C of dissolved organic carbon (DOC) and dissolved inorganic carbon (DIC), fDOM, other solutes and stable isotopes, and geochemical modeling to characterize springs of the remote Shivwits Plateau in Grand Canyon

National Park (GRCA). These parameters require small sample volumes (< 1 L total) and can be analyzed relatively simply and at low cost, thereby providing valuable insights into groundwater flow-system dynamics. The goals of this study are to: 1) use semi- annual sampling of the aforementioned geochemical parameters to determine controls on groundwater flow in the Shivwits Plateau, 2) examine spatial and temporal variability of

13 13 δ CDIC and δ CDOC to determine how carbon is cycled within the flow system, and 3)

use fDOM to examine microbial activity within the flow system. I hypothesize that this

multi-tracer approach through semi-annual sampling of springs can adequately

characterize the groundwater system of the Shivwits Plateau.

1.2 Study area

Figure 1 – Springs of GRCA. Green dots represent springs with existing data, black dots represent identified springs with no data (National Park Service spring data retrieved from the Spring Stewardship Institute’s Springs Online data portal). Not all springs at river level represented.

The Shivwits Plateau is located ~160 km west of Grand Canyon Village and ~100 km east of Las Vegas, NV, and marks the northern boundary of the Colorado River in western GRCA. Much of the previous work completed in GRCA has been focused in other areas, including the Kaibab Plateau, Coconino Plateau, and the Little Colorado

River Area (Tobin et al. 2018) (Fig. 1). The Shivwits Plateau has few data outside of remote sensing due to difficulty of completing field work. As anthropogenic effects on environmental parameters should be minimal due to the wilderness designation, characterizing mixed-lithology spring systems by using tracers that rely on such parameters should be effective in the area and could be applied in other similar terrains.

According to the Spring Stewardship Institute, the Shivwits Plateau contains over 50 karst or mixed-lithology springs that have little to no hydrogeological data

(http://springsdata.org/). GRCA largely relies on a single spring, Roaring Springs, to fulfill its water needs for over 6 million visitors and residents per year, while hundreds of smaller springs support wildlife and spring-dependent ecosystems (Tobin et al. 2018).

Further research in this area may be essential in the future as GRCA water needs increase due to increasing visitor volume. Thus, the motivation for choosing the Shivwits Plateau as the study area is due to its viability for using inorganic and organic tracers due to few anthropogenic influences, lack of data and understanding of the Shivwits Plateau groundwater system, and its rainfall mechanism of recharge in a region dominated by snowmelt-driven recharge.

Precipitation on the Shivwits largely follows a bimodal seasonal pattern of summer monsoons (July to September) and winter frontal systems (November to March).

Annual precipitation from 1988 to 2019 averaged 41.5 cm, with 21.1 cm occurring in winter months as a mix of rain and snow and summer monsoons accounting for 13.9 cm, with the remaining falling as infrequent rain at other times of the year (Western Regional

Climate Center, Desert Research Institute values for Yellow John Mountain; http://raws.dri.edu). Regional models for the southwest US show that the Shivwits has very little recharge capability due to evapotranspiration potential, low soil-water capacity, and runoff (Flint and Flint 2007). Soils on the Shivwits are shallow and well drained. The western half of the plateau where the Kaibab Formation outcrops is overlain by the

Winona soil complex, composed of loam and gravels with moderate alkalinity and calcareous throughout, with a depth to bedrock of 15 to 50 cm. The eastern half where

6 basalt flows outcrop is covered by the Cabezon complex, composed of stony loam and clays overlying fractured basalt at a depth to bedrock of 25 to 50 cm (Hendricks 1985).

These shallow soils allow for rapid runoff and/or infiltration during precipitation events

(Ingraham et al. 2001).

1.3 Hydrogeologic setting

Uplift of the Colorado Plateau Province began ca. 80–50 Ma and it has reached its current elevation with very little internal deformation at less than 1% internal upper crustal strain (Flowers et al. 2008). Regional dip is ~0.4% to the northeast, but most springs sampled are located on the southwestern edge of the plateau. The Kaibab Plateau shares similar stratigraphy to the Shivwits Plateau. Sinkhole density on the Kaibab

Plateau increases near faults and fractures, suggesting such structural features influence infiltration to the groundwater system (Huntoon 1974; Jones et al. 2018). Many springs of the Shivwits Plateau emerge near known faults and folds, suggesting structural deformation likely controls groundwater flow within the Shivwits Plateau also (Fig. 2).

The hydrostratigraphic section for this study focuses is bound by the Cambrian

Muav Limestone at the base and the Tertiary Shivwits Basalt that caps some of the plateau. The Muav Ls. is composed of fossiliferous limestone, silty limestone, dolomite, and mudstone with interbeds of sandstone, siltstone, mudstone, and shale (Pool et al.

2011). The Muav Ls. has little primary porosity, with its transmission capability resulting from dissolution of carbonate through faults and fractures (Huntoon 1970). The Rampart

Cave Member of the Muav Ls. has been found to be one of the most productive water- bearing units in the region (Huntoon 1977). Overlying the Muav Ls. is the Mississippian

Redwall Ls., which is composed of crystalline, thinly to thickly bedded limestone with

Figure 2 – Geologic map of the Shivwits Plateau with locations of sampled springs (SN – Snap Sp., TI – Tincanebitts Sp., JO – Joe Sp., BU – Burnt Canyon Sp., TW – Twin Sp., GR – Green Sp., DA – Dansil Sp., MU – Mud Sp., LO – Lost Sp., 202 – 202 Ephemeral Sp., CO – Cottonwood Sp., TC – Twin Creek Canyon Sp., TK – Twin Key Sp., HI – Hidden Sp.; LW and PP are locations of standing pool samples that were not groundwaters).

dolomite and thin beds and lenses of chert (Pool et al. 2011). The Redwall and Muav

limestones form the Redwall-Muav aquifer (R aquifer), which is confined by the Bright

Angel Shale below and the lower Supai Gp. above. The Shivwits Plateau contains

numerous breccia pipes that were formed by collapse into caverns formed by dissolution

in the Redwall Ls. (Wenrich and Sutphin 1989). Fractures associated with these collapse

features have been observed to be possible pathways for movement of shallow waters

deeper into the subsurface (Bills et al. 2010). The Surprise Canyon Fm., a valley-fill unit that is composed of a basal unit of cobble conglomerate and calcareous sandstone and a upper unit of marine carbonate and fine-grained clastics, is sometimes present in collapse features within the upper Redwall Ls (Billingsley and Beus 1985). Like the Muav Ls., the

Redwall Ls. has little primary porosity, with permeability resulting from dissolution of carbonates along faults and fractures (Huntoon 1970).

The Pennsylvanian Supai Group is divided into four formations, from lower to upper: the Watahomigi Fm., the Manakacha Fm., the Wescogame Fm., and the Esplanade

Ss. The Watahomigi Fm. is dominantly composed of carbonates with lesser amounts of mud. The Manakacha Fm. is mostly formed of carbonates, while sand and mud comprise the minor components. The Wescogame Fm. varies laterally but is primarily composed of sand and carbonates with little mud in the study area. Finally, the Esplanade Ss. is

predominantly a fine-grained sandstone with minor amounts of siltstone and mudstone

(McKee 1982). The lower formations in the Supai Gp. are confining units for the

underlying R aquifer, while the upper Supai formations are permeable and comprise the

lower section of the Coconino aquifer (C aquifer), which continues through to the

uppermost strata in the section (Pool et al. 2011). Overlying the Supai Fm., the Permian

Hermit Fm. is composed of red sandy shale and fine-grained friable sandstone and may

act as a local confining unit where large amounts of siltstone and claystone are present

(Metzger 1961; Pool et al. 2011). As the Hermit Fm. has extremely low permeability, groundwater movement predominantly occurs laterally above the unit except where faults

or fractures are present (Huntoon 1970).

The Coconino Sandstone overlying the Hermit Fm. is also Permian in age. It is

composed of well-sorted, fine-grained sandstone with silica cement and contains large

crossbeds (Pool et al. 2011). The Coconino Ss. has high primary porosity and

permeability, with groundwater movement being lateral due to the impermeability of the

underlying Hermit Fm. (Huntoon 1970). The unit thins to 2–3 meters within the study

area as the overlying Toroweap Fm. thickens (National Park Service (NPS) Geologic

Resources Inventory (GRI) program, 2000–2013). The Permian Toroweap Fm. is one of

the most complex units in the region, composed of carbonate sandstone, silty sandstone,

siltstone, limestone, and thin beds of gypsum, which thickens into a massive limestone

within the study area (Huntoon 1970; Sorauf and Billingsley 1991; Pool et al. 2011). The complex lithology of the unit leads some layers to act as impermeable units, while gypsum beds can have permeability similar to or higher than the underlying Coconino Ss. due to solution channels in the gypsum (Huntoon 1970). The Permian Kaibab Fm., which unconformably overlies the Toroweap Fm., is the uppermost Paleozoic unit. The Kaibab

Fm. is composed of cherty, thick-bedded limestone to sandy limestone, dolomite, siltstone, sandstone, and gypsum (Pool et al. 2011). While the primary porosity and permeability of the Kaibab Fm. is low, it is the principal recharge medium of some areas of GRCA due to solution widening of joints and bedding planes evident as sinkholes and

solution channels through karstification (Huntoon 1970). Finally, the eastern portion of

the plateau is capped by Tertiary Shivwits Basalt flows, which are composed of finely

crystalline alkali-olivine basalt (National Park Service (NPS) Geologic Resources

Inventory (GRI) program, 2000–2013). Basalt flows are important water bearing units in some nearby regions such as the Little and Big Chino Basin’s where interbedded with basin fill, but are not known to be major water-bearing units in other areas (Pool et al.

2011).

13 13 1.4 δ CDIC and δ CDOC as tracers

13 13 Coupling measurements of δ CDIC, δ CDOC, and fDOM, analyses of water isotopes and major ions, and geochemical modeling provides a multi–tracer approach to

characterize the mixed lithology groundwater system of the Shivwits Plateau (Fig. 3).

13 δ CDIC was chosen as a potential tracer due to the extensive carbonate lithologies present

in GRCA, which likely provide conduits for many karst springs in the study area as well

13 as providing flow paths to springs discharging from other lithologies. δ CDIC may be able to differentiate the predominant lithology of flow paths based on endmember values. For

13 more clastic formations, δ CDIC as well as DIC concentration may be useful in

determining relative residence time in upper carbonate units, whether flow was

predominantly through carbonate units before discharging from a clastic unit, or if flow

mainly occurred within the clastic units themselves. While the terrestrial fingerprint of

13 δ CDIC may be overprinted by fractionation processes such as degassing of CO2 along the

flow path, springs that share a groundwater system will likely show similar fractionation

13 trends in δ CDIC resulting from the system’s microbiome and preferential dissolution within carbonate strata.

Figure 3 – Conceptual cross-section of the section used in the Shivwits Plateau. Vegetation type and resulting DOC will vary spatially across the plateau and thus springs will show the terrestrial fingerprint of the source watershed, represented by flow lines. Carbonate 13 13 lithologies, and resulting DIC values (A, B, C, D), were analyzed for δ C and used as endmembers to aid in characterizing δ CDIC at the spring interface

13 δ CDOC has potential as a tracer of source watershed vegetation, which can be used to group springs. Vegetation across the Shivwits Plateau varies from primarily C3 vegetation such as mixed conifers and grasses to C4/CAM vegetation such as various species of cacti and succulents. Although the plateau is C3-dominated, localized areas of

13 C4/CAM vegetation may alter the δ CDOC signature enough to see variation across larger

areas. DOC concentrations may be especially helpful in deeper units such as the Supai

Gp. and Redwall-Muav system as DOC will be cycled over longer residence times than

DIC. Thus, if a Supai Gp. or Redwall-Muav spring shows a high DOC concentration, while water isotopes have a signature similar to recent precipitation values, it would suggest that more localized flow has occurred through faults, fractures, or conduits, thereby introducing DOC-laden waters into a deep groundwater system. DOM characterization by fluorescence spectroscopy can aid in determination of the source of

DOC at the spring discharge (i.e., whether it retains the terrestrial signature or is overprinted along the flow path).

CHAPTER TWO: METHODS

2.1 Sample collection

Spatial and temporal variability in geochemistry of the Shivwits groundwater system were characterized by two sampling trips (Table 2). The first sampling trip, in

May 2019, captured flow conditions after winter precipitation. The second sampling occurred in September and October 2019 during flow conditions after summer monsoons

(Fall). Target springs and corresponding rock sample collection focused on maximum spatial extent and characterization of all major stratigraphic horizons. Thirteen springs were sampled in May in all major stratigraphic horizons from the Redwall Ls. to the

Shivwits Basalt. Nine springs were sampled in Fall. Two springs (Pen Pocket and Lower springs) sampled during the May trip showed geochemical traits of being recent precipitation and not groundwaters and were not resampled. Four other springs (Burnt

Canyon, Twin Creek, Cottonwood, and Lost springs) could not be resampled for logistical reasons or insufficient flow, and one additional spring (Snap Spring) was sampled in Fall.

Water samples were collected and prepared using a method modified from Doctor et al. (2008), with long-term non-ideal storage methods confirmed by Wilson et al.

(2020). Samples were collected in quadruplicate and analyzed in duplicate or triplicate.

This method utilized 0.2-µm polyethersulfone (PES) syringe filters to remove microbes and halt microbial activity in the samples. Throughout the May sampling trip, samples for carbon and water isotopes were collected in cleaned 5.9-mL Labco Exetainers® that were precombusted at 400°C to remove any trace organic material and sealed with butyl rubber septum caps. For the fall sampling trip, 12-mL Labco Exetainers were used instead to

13 allow duplicate analysis of δ CDOC from the same vial, which requires 4 mL per

analytical sample.

Temperature, pH, specific conductance (SPC), and oxidation-reduction potential

(ORP) were recorded during sampling using a Hanna multiparameter meter, and

discharge was recorded by volumetric means where possible using a container with

known volume and a stopwatch. Spring discharge measurements were taken at most

locations, but some springs discharged directly into a pool (Snap and Twin springs) or

flowed at rates too low to be measured. Rock samples were collected adjacent to springs

where possible by removing a sample of unweathered bedrock and as needed along

vertical transects. Soil samples were collected across the Shivwits Plateau in areas likely

to provide recharge to springs in the study area. Based on GIS analyses that showed the

dominant soil types within the region, soil samples were collected where those soil types

were present and easily accessible from roads and/or off–trail routes.

2.2 Stable isotope analyses

2.2.1 δ2H and δ18O

Stable water isotopes (δ2H and δ18O) were analyzed using laser absorption off– axis integrated cavity output spectroscopy on a Los Gatos Liquid Water Isotope Analyzer

(Los Gatos Research T–LWIA–45–EP) (LGR) at the Kentucky Stable Isotope

Geochemistry Laboratory (KSIGL) at the University of Kentucky. Samples were pipetted into 2-mL autosampler vials and loaded into the LGR sample tray. During analyses, nine separate 1.2-μL injections were introduced into the analyzer cavity for direct measurement. The first four injections were ignored to account for potential memory

effects, and the last five injections were averaged to produce a raw isotope value. Raw

water-isotope ratios were normalized to the VSMOW scale in Laboratory Information

Management Systems (LIMS) for lasers 2015, using USGS49 (δ2H = –394.7‰, δ18O = –

50.6‰) and USGS50 (δ2H = +32.8‰, δ18O = +4.9‰) for normalization and USGS45

(δ2H = –10.3‰, δ18O = –2.2‰) as a blind standard. The precision and accuracy of all

measurements (2σ in this study), assessed with blind standards, were 0.1‰ or better,

while the average difference between sample/duplicate pairs was 0.2‰ or better.

2.2.2 δ13C of soil samples

Soil samples were treated with 20% HCl to remove inorganic carbonate and

flushed with deionized (DI) water until the HCl was removed and pH returned to near-

neutral levels. Samples were then frozen and freeze-dried to remove any remaining water content. A Retsch CyroMill was then used to homogenize the samples before they were weighed into tin capsules. Samples were loaded into an autosampler connected to a

Costech ECS 4010 Elemental Analyzer and sequentially dropped into an oxidation column heated to 1020°C. With the assistance of the tin capsule and research grade

(99.999%) O2, the sample was ultimately combusted at ~1700°C. Halogens and sulfur-

bearing compounds in the sample were trapped using silvered cobaltous oxide in the

bottom of the oxidation column. Sample gases were swept on a stream of ultra-high

purity (99.999%) He to a reduction column heated to 650°C, where excess oxygen was trapped out and nitrogen oxides were reduced to N2. Waters of combustion were removed

using a trap containing magnesium perchlorate and Sicapent (phosphorous pentoxide).

The remaining N2 and CO2 were separated inside a 3-m packed gas–chromatographic

column, then introduced consecutively into the IRMS via ConFlo IV for measurement

against their respective “monitoring” gases.

Following USGS recommendations (Coplen et al. 2006), raw carbon isotope

values were normalized to the VPDB scale using USGS64 (δ13C = –40.8‰) and USGS41

(δ13C = +37.6‰), and Thermo acetanilide as a linearity standard. Elemental

concentrations were determined from peak areas using either the k-factor or linear

regression methods. Blind standards were analyzed to assess the precision and accuracy

of isotopic and elemental percent data. The precision and accuracy of all measurements,

assessed with blind standards, were 0.2‰ or better.

2.2.3 δ13C and δ18O of carbonate in rock samples

Rock samples were powdered by using a Dremel grinder and homogenized in a

Retsch CyroMill. Samples were then loaded into acid-cleaned Labco Exetainers and dried

overnight at ~60°C to remove adsorbed water. Each Exetainer was sealed with a septum

and cap and flushed with ultra–high purity (99.999% pure) He for 13 minutes to remove

atmospheric N2, CO2, and O2. Then, 0.1 mL of >100% H3PO4 was injected through the

septum of each Exetainer and allowed to react with the sample in a heating block for at

least 24 hours at exactly 25°C to produce CO2 in the headspace. Seven consecutive aliquots of headspace gas from each Exetainer were transferred on a stream of He via autosampler to a Thermo Finnigan GasBench II interfaced to a Thermo Finnigan

DELTAplus XP isotope-ratio mass spectrometer (IRMS) at KSIGL. A 30-m PoraPLOT Q

gas-chromatographic column inside the GasBench separated any trace N2 from CO2

evolved from the sample and two Nafion membranes (upstream and downstream of the

GC column) were used to remove any traces of water.

Following USGS recommendations (Coplen et al. 2006), raw oxygen and carbon

isotope values are normalized to the VPDB scale using NBS18 (δ13C = –5.0‰, δ18O = –

23.2‰) and NBS19 (δ13C = +1.9‰, δ18O = –2.2‰). NIST 915b was analyzed as a blind

standard to assess the precision and accuracy of isotopic and elemental percent data.

Analytical duplicates were analyzed to assess the repeatability of measurements. The

precision and accuracy of all measurements assessed with blind standards, were 0.1‰ or

better, while the average difference between sample/duplicate pairs was 0.1‰ or better.

2.2.4 δ13C of dissolved inorganic carbon in spring samples

13 The method used for determination of δ CDIC uses H3PO4 to convert DIC to

aqueous and gaseous CO2 in a pre-flushed He-filled vial. The resulting headspace gas

(CO2 and He) is then introduced to the GC-IRMS system after an equilibration period

(Assayag et al. 2006). Modifications to the method include the use of 85% H3PO4 instead

of 100% H3PO4, as well as different CO2 equilibration conditions. Vials were placed in a

heating block held at 24°C and allowed to equilibrate for a minimum of 24 hours. Before

analysis, vials were placed in a vortex mixer to force any remaining dissolved gas to

transfer to the headspace, then centrifuged for several minutes to remove any

condensation from the bottom of the septa. Vials were then loaded into the GasBench II interfaced to the DELTAplus XP IRMS. Raw carbon isotope ratios were normalized to the

13 VPDB scale using two aqueous NaHCO3 working standards with contrasting δ C values

13 13 (KSIGL SB-1, δ CVPDB = –19.9‰; KSIGL SB-2, δ CVPDB = –3.0‰). Multiple in- session measurements of a third, blind aqueous NaHCO3 working standard (KSIGL SB–

13 3, δ CVPDB = –7.8‰) were used to assess the precision and accuracy of the data. All

three working standards have been tied to the VPDB scale through repeated

measurements (as powder) against NBS-18 and NBS-19 both at KSIGL and at other laboratories. The precision and accuracy of all measurements assessed with blind standards were 0.1‰ or better, while the average difference between sample/duplicate pairs was 0.2‰ or better.

2.2.5 δ13C of dissolved organic carbon in spring samples

13 The method used for δ CDOC was modified from Lang et al. (2012) and Zhou et

al. (2015) using wet oxidation of DOC to introduce CO2. Modifications include using

85% H3PO4 to react inorganic carbon and the use of Na2S2O8 as a substitute for K2S2O8.

Zhou et al. (2015) found that adding AgNO3 as a catalyst and using Na2S2O8 as an

oxidant improved analytical precision. Testing at KSIGL, however, suggested that the

addition of AgNO3 yielded less precise data and Na2S2O8 alone was therefore used for all

13 subsequent phases of the δ CDOC analyses. Improved precision from the use of Na2S2O8

13 was most likely due to a reduced carbon blank, the largest cause of variation in δ CDOC

values for repeated standard analyses.

Before analysis, vials were agitated with a vortex mixer to remove any remaining

dissolved gas from solution to the headspace, then centrifuged to remove any

condensation from the bottom of the septa. Vials were again loaded into a heating block

set to 24°C, using the GasBench II interfaced to the DELTAplus XP IRMS. Two aqueous

13 working standards mixed from reagent-grade powders (D-serine, δ CVPDB = –31.3‰;

13 oxalic acid, δ CVPDB = –6.8‰) were used to normalize raw carbon isotope values to the

VPDB scale. These working standards have been tied to the VPDB scale via repeated

measurement (as powders) against USGS40 and USGS41 at KSIGL. The accuracy of all

13 δ CDOC measurements, assessed with multiple in-session analyses of a blind, third

13 aqueous working standard (D-aspartic acid, δ CVPDB = –22.0‰) was 0.4‰ or better, while the precision, assessed in the same manner, was 1.6‰ or better.

2.3 DOM absorbance and fluorescence analyses

Fluorescence data were collected using an Agilent Varian Cary Eclipse

Fluorometer at excitation wavelengths of 230–455 nm at 5-nm increments and emission wavelengths of 290–702 nm at 2-nm increments. Samples were loaded into a fused quartz cuvette that was cleaned with a 50% HNO3 soak for 10 minutes and rinsed thoroughly with DI water before use and between samples. Blanks using DI water were analyzed for blank corrections at the beginning, middle, and end of each set of samples, and to check for emission inconsistencies due to acid residue between samples.

Absorbance data was collected using a ThermoFisher GENESYS 10S UV-Vis scanning spectrophotometer to collect absorbance values from 230 nm to 750 nm at 2-nm increments. DI water blanks were used to establish baseline absorbance values before each set of samples. For both fluorescence and absorbance analyses, analytical duplicates were analyzed for May samples and field duplicates were used for Fall samples. Standard deviation between duplicates was ±0.016 a.u. or better for absorbance and duplicates were introduced into the parallel factor analysis (PARAFAC) model to correct for variation within samples. The fluorescence index (FI) was calculated as the ratio of emissions at 450–550 nm at 370 nm excitation, giving relative contribution of terrestrial to microbial sources calibrated to known allochthonous or autochthonous sources

(McKnight et al. 2001). The biological index (BIX) was calculated as the ratio of emissions at 380–430 nm at 310 nm excitation, formulated from two common peaks in surface waters attributed to terrestrial and microbial sources (Huguet et al. 2009). Finally,

the humification index (HIX) was calculated as the ratio of the integrated intensities from emissions 435–480 nm divided by 300–345 nm at excitation of 254 nm, giving relative humification values for DOM content (Birdwell and Engel 2010).

2.4 PARAFAC modeling

Characterization of dissolved organic matter (DOM) was performed using the R statistical computing software package starDom (https://cran.r– project.org/web/packages/staRdom/vignettes/PARAFAC_analysis_of_EEM.html), maintained by WasserCluster Lunz and the University of Natural Resources and Life

Sciences in Vienna, to analyze fluorescence and absorbance data. This package utilizes functions from the R packages eemR, dplyr, and tidyr to correct import raw excitation- emission matrix (EEM) and absorbance data, correct and normalize the dataset, extract fluorescence and absorbance indices, perform PARAFAC to develop and validate models of components, and export and interpret the results. These functions and methods follow

Murphy et al. (2013), including corrections for raw data, blank subtraction, inner filter effect, Raman normalization, and removal of Raman and Rayleigh scattering bands. The

PARAFAC models are then created using an alternating-least-squares method to minimize error. Once a model has been chosen, the stability of the model is determined through a split-half analysis, which randomly splits samples into subgroups and then compares how the model fits the subgroups individually. If the model is adequate for all subgroups, then it is considered to be stable and not leveraged by any given group of samples. This is further quantified by calculating Tucker’s congruency coefficients, to examine the correlation of the same component from each subset. When the model has been validated, the results are then uploaded to openfluor.org to compare the components

calculated from the PARAFAC model to components found in other publications. This

allows direct comparison of components across studies to further help characterize the

composition of DOM in water samples (Murphy et al. 2014).

2.5 Other analyses

Metals were analyzed by inductively coupled plasma–mass spectrometry (EPA

- 6020A), anions other than HCO3 were analyzed by anion chromatography (EPA 9056A),

- and alkalinity was determined by titration with HCO3 calculated from geochemistry of

other results (EPA 310.1) at the Kentucky Geological Survey. All statistical analyses

were performed in R with the RStudio integrated development environment. Saturation

indices were calculated using USGS PHREEQC geochemical software

(https://www.usgs.gov/software/phreeqc-version-3).

Principal component analyses were run in RStudio using the prcomp() function

- 2- - with the following variables: concentrations of Ca, Sr, F , SO4 , Mg, Cl , Na, K, B, Ba,

- HCO3 , Si, NO3, and DOC, as well as the calcite and gypsum saturation indices,

- terrestrial PARAFAC component values (CT + CR) (R.U.), CP (R.U.), HCO3 /(Mg+Ca)

2 18 13 13 molar ratio, pCO2, SPC, pH, temperature, δ HH2O, δ OH2O, δ CDOC, and δ CDIC. Other

measured parameters were not included in the PCA because high correlation with

- components that were included (e.g. alkalinity and HCO3 ).

CHAPTER THREE: RESULTS

3.1 Spring physical, chemical, and isotopic parameters

As shown on the Piper plot (Fig. 4), most springs had water types of Ca/Mg-

HCO3, while Twin Key Spring, Mud Spring, and Dansil Spring (Fall samples) were Ca-

-5 3 SO4 type. Very low flow (<0.001 cfs | <3 × 10 m /s) occurred at all springs except for

Cottonwood Spring and 202 Ephemeral Spring, which had discharges of 0.32 cfs (0.009

m3/s) and 0.003 cfs (8.5 × 10-5 m3/s). The pH of spring water was near neutral to slightly

basic (6.71–8.34) for all springs. Temperature at the discharge point had a wide range,

Figure 4 – Piper plot (https://nevada.usgs.gov/tech/excelforhydrology/) of spring samples symbolized by lithology of spring discharge and sampling season.

Table 1 - Collection date, time, lithology, elevation, and discharge of spring (nm–not measured, flow too low), and basic water chemistry. Site Season Date Time Lithology Elevation Discharge Temp. pH ORP SPC (m) (cfs) (m3/s) (°C) (mV) (μS/cm) 202 Ephemeral May 5/9/2019 15:00 Supai 1253 0.00353 1.0 × 10-4 15.75 7.66 238.1 569 Burnt Spring May 5/14/2019 13:00 Supai 1255 0.00001 2.8 × 10-7 26.21 8.05 99.3 607 Cottonwood Spring May 5/10/2019 9:15 Redwall 974 0.32 9.1 × 10-3 21.97 7.51 168.8 520 Dansil Spring May 5/15/2019 13:15 Toroweap 1412 0.00001 2.8 × 10-7 13.8 7.9 127.3 924 Fall 10/1/2019 18:00 Toroweap 1412 0.00001 2.8 × 10-7 13.11 7.88 99.4 922 Green Spring – May 8 May 5/8/2019 13:20 Basalt 1829 0.00066 1.9 × 10-5 12.5 7.83 232.6 623 Green Spring – May 10 May 5/10/2019 18:00 Basalt 1829 0.00074 2.1 × 10-5 11.77 7.41 294.1 670 Green Spring – May 11 May 5/11/2019 16:30 Basalt 1829 0.00001 2.8 × 10-7 12.01 7.69 165.6 667 Green Spring – May 12 May 5/12/2019 ~ Basalt 1829 0.00001 2.8 × 10-7 13.4 8.34 168.2 668 Fall 9/30/2019 ~ Basalt 1829 0.00001 2.8 × 10-7 12.52 6.84 186.4 760 Hidden Spring May 5/9/2019 11:00 Redwall 923 0.00001 2.8 × 10-7 17.16 7.46 176.4 720

24 Fall 9/29/2019 14:31 Redwall 923 nm nm 19.66 7.8 174.5 761 Joe Spring May 5/14/2019 18:00 Coconino 1724 0.00001 2.8 × 10-7 11.33 8.04 115.1 506 Fall 9/30/2019 13:30 Coconino 1724 0.00001 2.8 × 10-7 15.05 7.79 119.5 464 Lost Spring May 5/14/2019 13:45 Supai 938 0.00012 3.4 × 10-6 21.92 7.95 84.8 834 Lower Spring May 5/15/2019 19:15 Kaibab 1732 0.00001 2.8 × 10-7 15.44 8.62 142.9 148 Mud Spring May 5/15/2019 10:45 Toroweap 1388 0.00039 1.1 × 10-5 21.44 8 128.4 1149 Fall 10/1/2019 14:00 Toroweap 1388 0.00001 2.8 × 10-7 16.94 7.43 121.3 1121 Pen Pocket May 5/10/2019 12:30 Basalt 1853 0.00005 1.4 × 10-6 13.37 7.04 284.4 140 Snap Sp Fall 9/30/2019 ~ Toroweap 1707 nm nm 11.82 7.77 157 401 Tincanebitts Spring May 5/14/2019 13:45 Coconino 1693 0.00001 2.8 × 10-7 12.84 7.5 117.1 349 Fall 9/27/2019 13:30 Coconino 1693 0.00044 1.2 × 10-5 14.46 6.71 66.6 831 Twin Canyon Seep May 5/8/2019 14:00 Redwall 886 0.00006 1.7 × 10-6 20.34 7.46 225.4 634 Twin Key Spring May 5/8/2019 19:00 Redwall 986 0.00059 1.7 × 10-5 17.42 7.2 208.4 948 Fall 9/29/2019 18:30 Redwall 986 nm nm 18.01 7.96 140.4 1297 Twin Spring May 5/15/2019 16:30 Kaibab 1707 nm nm 11.02 7.04 142.7 425 Fall 9/28/2019 13:30 Kaibab 1707 nm nm 14.85 7.68 173.1 421

Table 2 – Alkalinity (Alk.), DOC, and major ions (MDL – below detection limit). Site Season Alk. DOC Ba B Ca Mg P K Si Na Sr S (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) 202 Ephemeral May 250 8.4 0.21 0.11 68.3 26.7 0.01 1.76 5.36 17.3 0.18 9.51 Burnt Spring May 258 7.3 0.13 0.13 81.3 18.9 0.02 4.24 5.87 5.45 0.23 12.7 Cottonwood Spring May 210 1.4 0.08 0.15 67.5 23.2 0.02 1.9 6.96 8.67 0.32 20.4 Dansil Spring May 301 2.8 0.04 0.13 126 65.6 0.01 0.63 13.5 19.7 1.41 95.6 Fall 178 1.9 0.01 MDL 129 58.9 MDL 1.55 12.3 17.1 1.64 125 Green Spring – May 8 May 319 2.3 MDL 0.03 86.7 33.5 0.03 0.97 22.2 10.9 0.35 5.33 Green Spring – May 10 May 318 1.7 0.005 0.09 85.2 33 0.04 1.01 22.1 10.8 0.35 5.05 Green Spring – May 11 May 319 1.4 0.003 0.07 85 32.8 0.03 1.01 22.1 10.7 0.35 5.78 Green Spring – May 12 May 318 2.2 0.003 0.06 85 32.9 0.03 0.89 22.2 10.7 0.35 5.93 Fall 405 3.5 0.02 MDL 108 41.6 0.02 1.76 21.1 13.3 0.32 3.75 Hidden Spring May 216 1.6 0.06 0.12 83.1 28.5 0.01 2.74 6.77 22.9 0.42 34

25 Fall 206 1.9 0.04 MDL 84.5 32.2 0.009 2.45 7.38 17.4 0.49 44

Joe Spring May 247 1.7 0.16 0.08 64.1 20.7 0.01 0.82 6.6 14.4 0.12 4.93 Fall 238 2.7 0.16 MDL 64.9 20.6 0.02 0.84 6.51 13.9 0.12 4.75 Lost Spring May 380 3.4 0.16 0.12 69.6 57.9 0.01 2.44 12.4 25.7 0.24 12.2 Lower Spring May 84 11.3 0.02 0.06 27.5 3.52 0.02 1.38 0.61 1.4 0.04 0.72 Mud Spring May 146 3.9 0.01 0.05 170 63.9 0.01 0.88 7.59 9.38 1.62 199 Fall 128 2.2 0.01 MDL 178 66.6 MDL 1.2 8 9.85 1.7 201 Pen Pocket May 76 8.6 0.04 0.07 19.5 5.45 0.02 1.87 4.62 1.06 0.11 0.55 Snap Sp Fall 200 2.5 0.09 MDL 57.5 19.8 0.03 1.22 5.38 7.32 0.1 3.04 Tincanebitts Spring May 223 3.3 0.06 0.07 57.4 19.7 0.01 MDL 4.47 7.06 0.12 4.14 Fall 286 2.1 0.08 MDL 94.5 20.6 0.04 0.69 6.18 7.61 0.14 4.13 Twin Canyon Seep May 224 2.7 0.04 0.08 81 28.3 0.01 2.18 6.88 13.2 0.45 31.5 Twin Key Spring May 188 2.6 0.03 0.13 118 44.6 0.01 3.32 6.66 21.3 0.75 92.6 Fall 129 4.4 0.03 MDL 106 53.4 MDL 4.66 7.92 26.4 0.83 113 Twin Spring May 221 2.5 0.07 0.09 77.7 7.7 0.04 0.88 5.3 4.57 0.07 2.91 Fall 213 2.5 0.08 MDL 83 8.24 0.05 0.94 6.04 4.88 0.07 2.79

Table 3 –Major anions, pCO2, and saturation indices (SI) for calcite and gypsum (MDL – below detection limit). - - - - 2- Site Season HCO3 Cl F NO3 SO4 pCO2 SICalcite SIGypsum (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (log10) 202 Ephemeral May 305 24.6 0.17 MDL 28.6 -2.25 0.36 -2.15 Burnt Spring May 315 4.91 0.16 MDL 39 -2.58 0.97 -1.96 Cottonwood Spring May 256 11.5 0.18 1.85 65 -2.14 0.22 -1.81 Dansil Spring May 367 32.2 0.39 0.2 277 -2.45 0.79 -1.08 Fall 217 26.6 0.38 3.4 369 -2.66 0.54 -0.94 Green Spring – May 8 May 389 29.9 MDL 8.19 15.6 -2.86 1.16 -2.34 Green Spring – May 10 May 388 30.8 MDL 7.53 15.1 -2.34 0.67 -2.35 Green Spring – May 11 May 389 30 MDL 6.65 15.2 -1.92 0.24 -2.35 Green Spring – May 12 May 388 30.2 MDL 6.62 15.2 -2.2 0.52 -2.37 Fall 494 23.6 MDL 0.3 11.4 -1.25 -0.13 -2.42 Hidden Spring May 263 33.5 0.22 15.1 106 -2.11 0.17 -1.55

26 Fall 251 28.1 0.17 5.73 141 -2.46 0.51 -1.44

Joe Spring May 301 16.5 0.26 4.21 14 -2.66 0.65 -2.45 Fall 290 16.8 0.31 4.27 14.6 -2.41 0.45 -2.43 Lost Spring May 463 37.2 0.26 MDL 37.7 -2.34 0.87 -2.13 Lower Spring May 102 1.96 MDL MDL MDL -3.69 0.53 MDL Mud Spring May 178 21.5 0.56 MDL 517 -2.83 0.77 -0.76 Fall 156 24.6 0.87 0.12 574 -2.33 0.1 -0.69 Pen Pocket May 93 MDL MDL MDL MDL -2.13 -1.23 MDL Snap Sp Fall 244 8.03 0.24 2.73 8.89 -2.47 0.27 -2.66 Tincanebitts Spring May 272 6.77 0.27 MDL 11.4 -2.15 0.06 -2.56 Fall 349 6.91 0.25 4.17 10.8 -1.25 -0.41 -2.44 Twin Canyon Seep May 273 19.2 0.2 7.27 96.5 -2.08 0.22 -1.6 Twin Key Spring May 229 34.2 0.22 5.13 294 -1.92 -0.06 -1.06 Fall 157 50.1 0.41 8.88 418 -2.85 0.45 -0.98 Twin Spring May 269 10.1 0.11 0.63 8.1 -1.7 -0.29 -2.57 Fall 260 10.8 0.13 2.8 9.4 -2.34 0.41 -2.5

Table 4 – Isotopic results for stable water isotopes and dissolved carbons species.

2 18 13 13 Site Season δ H–H2O δ O–H2O δ C–DIC δ C–DOC ~ ~ (‰ – VSMOW) (‰ – VSMOW) (‰ – VPDB) (‰ – VPDB) 202 Ephemeral May -82.4 -10.5 -13 -23.9 Burnt Spring May -73.5 -9.5 -12 -25.1 Cottonwood Spring May -86.6 -12 -10.4 -22.4 Dansil Spring May -84.7 -11.2 -11.9 -25.4 Fall -88.1 -12 -9.8 -26.5 Green Spring – May 8 May -85.5 -11.6 -11.4 -28.1 Green Spring – May 10 May -85.7 -11.6 -11.4 -28.4 Green Spring – May 11 May -85.6 -11.6 -11.4 -26.4 Green Spring – May 12 May -85.3 -11.6 -11.4 -28.6 Fall -81 -10.9 -11.4 -25 Hidden Spring May -78.6 -10.6 -11.1 -28 Fall -84.8 -11.5 -8.3 -27.2 Joe Spring May -86.8 -11.7 -7.6 -25.5 Fall -86 -11.6 -7.6 -24.3 Lost Spring May -83.7 -10.9 -13.7 -27.5 Lower Spring May -58.7 -6 -11.9 -24.4 Mud Spring May -81.9 -11.4 -7.9 -28.4 Fall -81.9 -11.2 -7.6 -25.8 Pen Pocket May -71.2 -8.1 -8.8 -25.9 Snap Sp Fall -88.3 -12.5 -8.8 -24.1 Tincanebitts Spring May -85.3 -11.6 -6.5 -29.6 Fall -88.3 -12.1 -10.2 -23.3 Twin Canyon Seep May -86.5 -11.9 -8.8 -27.2 Twin Key Spring May -82.3 -11.2 -10.3 -29.3 Fall -74.2 -9.6 -8 -21.1 Twin Spring May -84.9 -11.8 -8.8 -28.6 Fall -84 -11.6 -8.4 -22.2

from 11.02°C at Twin Spring to 26.21°C at Burnt Spring, both sampled in May.

Bicarbonate values ranged from 156 mg/L at Mud Spring to 494 mg/L at Green Spring, with both minimum and maximum values occurring in Fall. Values for DOC ranged from

1.4 mg/L at both Cottonwood Spring and Green Spring to 8.4 mg/L at 202 Ephemeral

13 Spring, all occurring in May. δ CDOC values were predominantly in the range of C3 plant

matter, ranging from –29.6‰ at Tincanebitts Spring in May to –21.1‰ at Twin Key

13 Spring in Fall. δ CDOC values increased from May to Fall for all springs sampled in both

13 seasons except for Dansil Spring. δ CDIC values varied from –13.7‰ at Lost Spring to –

6.5‰ at Tincanebitts Spring, both during Fall.

Figure 5 – Spring samples plotted relative to the global meteoric water line (GMWL) (Craig 1961), local meteoric water line for the South Rim (Bills et al. 2007), and regional Arizona springs line (Springer et al. 2017). See Figure 2 for springs abbreviations.

Finally, δ2H ranged from –88.3‰ at Tincanebitts Spring in May to –73.5‰ at

Burnt Spring during Fall, while δ18O ranged from –12.5‰ at Snap Spring in May to –

9.5‰ at Burnt Spring during Fall. Figure 5 shows the water isotopes plotted relative to the global meteoric water line (GMWL) (Craig 1961), local meteoric water line for the

South Rim at Grand Canyon Village (Bills et al. 2007), and Arizona regional springs

trend line (Springer et al. 2017). The spring trend line for this study is δ2H = 5.1δ18O –

25.9, with an r2 of 0.90.

3.2 δ13C of rock and soil samples

Figure 6 – Isotopic composition of rock samples grouped by lithology.

The isotopic composition of carbonate in rock samples is shown in Fig. 6. Two samples

of the Shivwits Basalt at Green Spring were analyzed, one sample from above the spring

and one sample below the spring. The sample above the spring contained negligible

carbonate and did not produce isotopic values while the lower sample gave values for

δ13C of –6.8‰ and δ18O of –9.2‰. Both samples were taken from unweathered surfaces, thus this difference is likely due to heterogeneity in the basalt or deposition of a carbonate precipitate within the lower sample. The Kaibab Fm. had the widest spread in values, with δ13C of 0.0 ± 2.6‰ and δ18O of –10.3 ± 2.2‰. The Toroweap Fm. had the

most positive δ13C at +1.6 ± 1.5‰ and δ18O of –8.6 ± 1.8‰. The Coconino Ss. had the

second-most negative values after the Shivwits Basalt, with δ13C of –3.3 ± 1.5‰ and

δ18O of –10.4 ± 2.2‰. The Supai Gp. has a wide range of values, with δ13C of +1.1 ±

2.5‰ and δ18O of –6.8 ± 3.4‰. Finally, the single sample collected from the Redwall

Fm. gave a δ13C of –0.6‰ and δ18O of –4.7‰. All lithologies except for the Shivwits

Basalt have largely overlapping δ13C values near conventional marine limestone values

with an average δ13C of 0.1 ± 2.5 ‰. δ18O values also overlap for most lithologies except

for the Redwall Limestone, with an average of –8.8 ± 2.6‰.

Soil δ13C ranged from –25.1‰ near the western edge of the plateau to –21.3‰

towards the east-central area of the plateau. δ13C values generally become more positive

closer to the canyon rim and decrease northward away from the rim. This trend is shown

in Fig. 7 through interpolated values and generally matches with the National Land Cover

Figure 7 – Location and value of δ13C for soil samples and classification of land cover from the National Land Cover Database (NLCD 2016).

Database (NLCD) transition from evergreen forest classified cover to shrub/scrub and herbaceous cover. Square patches of shrub/scrub in the NLCD plot are grazing plots for

livestock that have been cleared of forest cover. Occurrences of cacti, succulents, and

other CAM plants with more positive δ13C values generally increased toward the rim.

3.3 DOM absorbance and fluorescence analyses

Figure 8 – results of PARAFAC analyses with Coble Peaks labeled on EEMs and representative samples for high CT and CR content (Dansil - May), high CP content (Dansil - Fall), and a well-mixed sample with all three components present in relatively equal quantities (Hidden - May) Coble (1996) peaks are labeled on component EEMs

Table 5 – Results and characterization of PARAFAC components. Peak maximum

shows primary peak with secondary peak in parentheses. Sources: a: Coble, 1996; 1: Lapierre and Giorgio 2014; 2: Murphy et al. 2011; 3: Yamashita et al. 2010; 4:

Gonçalves-Araujo et al. 2016; 5: Dainard et al. 2014; 6: Chen et al. 2018.

Component Peak maximum Coble Component Previous studies Peaka Description CT Ex: <260 (345) nm A + M Terrestrial 1: C2, 2: G1, 3: C1, 4: C1, Em: 460 nm humic–like 5: C1, 6: C<260(365)/476

CR Ex: <260 (315) nm A + M Microbial 1: C5, 2: G2, 3: C6, 4: C2, Em: 400 nm humic–like, 5: C2, 6: C<260(305)/404 reprocessed OM

CP Ex: <260 nm B + T Protein–like, 1: C6, 2: G7, 3: C7, 4: C3, Em: 324 nm amino acids 5: C4, 6: C275/338

Absorbance and fluorescence of spring water samples coupled with PARAFAC analyses resolved three dominant DOM components across all samples (Fig. 8). The components are described in Table 5. Split-half reliability analyses found that the model was internally consistent with an average correlation between split-components of 0.96 ±

0.07. PARAFAC models with 4 and 5 components were also reliable, or internally consistent, at 0.96 ± 0.11 and 0.93 ± 0.18, respectively, but these models did not result in components consistent with any previous literature. This is likely due to the 3-component model being a complete model and 4- and 5-component models subdividing a complete component into unresolvable components. PARAFAC models were run individually for

May and Fall samples, resulting in a reliable model for May and a highly reliable model for Fall. However, the resolved components for May did not coincide with components reported in the literature. This is likely due to the small number of samples used in the model, as PARAFAC analyses are stronger when more samples can be used in resolving

Table 6 – Results of DOM analyses. CT/R/P – PARAFAC component, a254 – absorbance at 254 nm, SUVA – specific ultraviolet absorbance, SR – spectral ratio. Site Season DOC CT CR CP CT CR CP a254 SUVA SR FI HIX BIX (mg/L) (R.U.) (R.U.) (R.U.) % % % (a.u) (L/mg C*m) 202 Ephemeral May 8.4 1.00 0.91 0.02 52 47 1 0.17 2.01 0.89 1.43 0.97 0.58 Burnt Spring May 7.3 0.91 0.76 0.03 54 45 2 0.14 1.85 0.92 1.48 0.94 0.58 Cottonwood Spring May 1.4 0.02 0.02 0.01 38 49 13 0.01 0.36 1.86 2.04 0.74 0.55 Dansil Spring May 2.8 0.54 0.41 0.00 57 43 0 0.08 2.64 0.87 1.43 0.99 0.53 Fall 1.9 0.07 0.06 0.13 27 23 50 0.02 0.88 0.73 1.51 0.34 0.76 Green Spring – May 8 May 2.3 0.02 0.03 0.01 31 47 22 0.01 0.41 0.00 1.73 0.61 0.70 Green Spring – May 10 May 1.7 0.10 0.10 0.07 38 37 25 0.01 0.49 3.00 1.60 0.75 0.88 Green Spring – May 11 May 1.4 0.02 0.03 0.01 32 49 19 0.01 0.66 0.25 1.82 0.96 1.16 Green Spring – May 12 May 2.2 0.02 0.03 0.00 43 56 0 0.01 0.38 0.58 1.29 0.72 0.93 Fall 3.5 0.02 0.03 0.01 33 54 13 0.02 0.67 0.68 1.58 0.74 0.75 Hidden Spring May 1.6 0.04 0.08 0.07 21 42 38 0.01 0.81 0.01 1.81 0.54 1.02

33 Fall 1.9 0.04 0.04 0.02 44 37 19 0.01 0.41 3.64 1.41 0.78 0.62

Joe Spring May 1.7 0.09 0.12 0.06 33 46 21 0.02 1.19 0.25 1.54 0.75 0.88 Fall 2.7 0.11 0.12 0.04 41 43 16 0.02 0.58 0.07 1.81 0.83 0.81 Lost Spring May 3.4 0.34 0.41 0.04 43 52 5 0.04 1.31 0.70 1.49 0.90 0.61 Lower Spring May 11.3 1.07 0.95 0.16 49 44 7 0.24 2.16 1.06 1.36 0.90 0.60 Mud Spring May 3.9 0.35 0.27 0.12 47 36 16 0.06 1.49 0.91 1.47 0.82 0.53 Fall 2.2 0.19 0.11 0.03 59 32 9 0.03 1.14 1.09 1.29 0.92 0.50 Pen Pocket May 8.6 0.50 0.46 0.11 47 43 10 0.14 1.58 1.09 1.36 0.87 0.60 Snap Spring Fall 2.5 0.07 0.09 0.05 34 44 23 0.02 0.67 1.10 1.60 0.66 0.85 Tincanebitts Spring May 3.3 0.17 0.18 0.09 38 42 21 0.03 0.87 0.92 1.51 0.80 0.77 Fall 2.1 0.05 0.07 0.02 37 51 12 0.01 0.26 4.47 1.87 0.77 0.86 Twin Canyon Seep May 2.7 0.00 0.03 0.01 9 65 26 0.01 0.26 3.85 1.80 0.95 0.71 Twin Key Spring May 2.6 0.03 0.06 0.04 23 48 29 0.01 0.29 0.02 1.62 0.61 0.84 Fall 4.4 0.19 0.20 0.15 35 37 28 0.04 0.85 0.71 1.69 0.76 0.86 Twin Spring May 2.5 0.15 0.13 0.00 54 46 0 0.03 1.10 0.67 1.49 0.99 0.65 Fall 2.5 0.16 0.14 0.07 43 37 20 0.03 1.06 0.86 1.51 0.67 0.61

the individual components. The amounts of each PARAFAC component (CT/R/P) are

given in Table 6, along with normalized percentages of each component, absorbance and

specific ultraviolet absorbance at 254 nm, spectral ratio (SR), FI, HIX, and BIX. While FI and BIX values were calculated and reported for this study, DOM content was low enough that fluorescence intensities at the wavelengths used to calculate these indices

were close to background noise from the fluorometer. Thus, these values may not be

representative of the DOM in the sample and BIX and FI will not be discussed in-depth.

3.4 Principal component analysis

The resultant PCA accounted for 28.9% of total variation in the first principal

component (PC1) and 17.33% in the second component (PC2) (Fig. 9). Samples plotted

within four groups. The pink ellipse with Lost, 202 Ephemeral, and Burnt Canyon springs is characterized by higher DOC and terrestrial component (CT + CR) concentrations, more

positive water-isotope values, and lower ion content. The blue ellipse with Dansil and

Mud springs is characterized by high gypsum saturation indices (average of -0.84

2- compared to -2.1 for other springs), high Ca and SO4 concentrations, higher CP values,

- and lower HCO3 . The red ellipse, which contains all Redwall Ls. springs, is not

controlled by any group of loading factors and thus plots near the middle of the PCA.

Finally, the black ellipse contains all other springs and is characterized by higher pCO2

- values, more negative water-isotope values, and high HCO3 .

Figure 9 – Plot of principal components 1 and 2 with loadings and samples (M – May sample, F – Fall sample). See Figure 2 for springs abbreviations.

CHAPTER FOUR: DISCUSSION

13 13 4.1 Viability of δ CDIC and δ CDOC as tracers

To use stable carbon isotopes as tracers for groundwater sources and flow paths, a

variation in δ13C of endmember sources must be present within the study area. Most of the vegetation on the Shivwits Plateau, which consists of mixed conifer forest and shrubland, utilizes the C3 metabolic pathway (Long and Hutchin 1991). Values for soil

δ13C on the Shivwits reflect this as the range of –25.1 to –21.3‰ (average: –23.7 ± 0.9

‰) overlaps the upper range of δ13C for C3 vegetation (–35 to –23‰) (O’Leary 1988).

Microbial degradation and selective preservation of OM within soils results in soil δ13C

being enriched by +1.5 to +4.3‰ relative to the source vegetation (Lichtfouse et al.

1995). When accounting for this, the average δ13C values for vegetation on the Shivwits are depleted relative to the soil and could range from approximately –29.4 to – 22.8‰.

Another reason for the relatively enriched soil δ13C relative to C3 vegetation is likely the minor occurrence of C4 and CAM type vegetation near the edge of the plateau, such as cacti, succulents, and some hedges. While soil δ13C could be seen as an average of

surrounding vegetation, inter-genus variation of junipers has been found to locally vary

from –30.0 to –22.6‰ and pine from –31.0 to –24.6‰ (Mervenne 2015). The range of

13 13 δ CDOC values (–29.6 to –21.1‰) is similar to that of vegetation and soil δ C. In a

13 coniferous catchment, δ CDOC values have been shown to be similar to soil values with no significant difference (Amiotte-Suchet et al. 2007). As seen in the map of soil δ13C

values (Fig. 7), soils become more depleted northwards away from the rim of the plateau.

While there is a slight spatial trend, it is not strong enough and does not correlate with

any spring parameter enough to be used as a sole tracer for possible recharge zones.

Linked to vegetation type, soil CO2 is a major source for DIC, with the proportion

ranging from ~50% in carbonates to ~100% in silicates due to stochiometric dissolution

of the respective lithologies (Shin et al. 2011). These reactions for weathering of

carbonates and silicates (assuming anorthite as the primary silicate phase) is as follows

(Aravena et al. 1992):

Carbonates: (1)

CaCO + CO + H O Ca + 2HCO +2 − 3 2 2 → 3 Silicates: (2)

CaAl (SiO ) + 2CO + H O Ca + Al O + 2SiO + 2HCO +2 − 2 4 2 2 2 → 2 3 2 3 13 In carbonate terrains, the stoichiometry shows that δ CDIC will be composed of a 1:1

molar ratio of C from CaCO3 and soil CO2. In silicate terrains, all DIC will be sourced

from the original soil CO2. As the upper lithologies in the study area are composed of

carbonates, it is likely that carbonate weathering dominates in the region and is

13 represented in the SIcalcite values (-0.41 to 1.16, average: 0.39). Because the δ C of soil

13 CO2, like the δ C of soil OM, is largely controlled by the dominant vegetation type, a

13 back-calculation can be done to determine the approximate δ C of soil CO2 (Amundson

et al. 1998). Respiration from roots and degradation of vegetation yield δ13C values

13 similar to the vegetation itself, and diffusion of C-depleted CO2 to the atmosphere results in a +4.4‰ fractionation in the remaining soil CO2 (Cerling et al. 1991; Aravena

13 et al. 1992). Using the average soil δ C value for the plateau results in an initial soil CO2

δ 13C value of –19.3‰, with a possible range from –20.7 to –16.9‰. Given the average

13 13 carbonate δ C for all lithologies as +0.1‰ results in an average δ CDIC of –9.7‰, with a

13 possible range from –12.4 to –6.5‰. The average δ CDIC of springs is –9.9‰ (–13.7 to –

6.5‰), which shows that the dominant source of DIC for most springs is the

stochiometric weathering of carbonates by soil CO2.

Figure 10 shows that measured δ13C values plot within theoretical ranges for the

13 dissolution of carbonates by soil CO2 (for δ CDIC) and for soil organic matter given

13 plausible vegetation values (for δ CDOC). From the full plots it can be seen that there is

13 increasing depletion in δ CDIC with an increase in terrestrial DOM, while the trend for

13 2 δ CDOC shows enrichment with r values of 0.39 and 0.83, respectively, for terrestrial

DOM >0.4 R.U.. These trends suggest the dominant sources of DOC within the recharge area. At low terrestrial DOM concentrations, the large range in DOC values likely reflects localized variation in vegetation types in the respective recharge areas. DOC can

be transformed by microbial degradation, but the resulting transformed DOC retains the

source value in heterotrophic metabolism (Finlay and Kendall 2007). As terrestrial DOM

concentrations increase, DOC approaches an average value for vegetation or soil OM and

the percentage of protein-like DOM does not appear to control δ13C of DIC or DOC. The

trendline levels out for higher terrestrial DOM concentrations at approximately –24.0‰,

close to the average soil δ13C of –23.7‰, suggesting that terrestrial DOM increases due

to a larger runoff component over a broader area or well-mixed groundwater. The

13 trendline for δ CDIC similarly suggests that DIC at higher terrestrial DOM concentrations

is more strongly controlled by depleted soil CO2 relative to weathering of carbonate

lithologies.

(b)

(a)

Figure 10 – Plots of δ13C vs. terrestrial DOM component concentration with ranges of source values (a) and insets of low terrestrial components with percentage of protein-like DOM labeled (b, c). Dashed boxes are theoretical values from possible vegetation DOC and possible soil CO2 ranges.

4.2 Utilization of fDOM to characterize DOM

Results of fluorescence analyses revealed that DOM within the Shivwits system is

largely composed of terrestrial humic-like DOM (CT), microbially transformed humic- like DOM (CR), and protein-like DOM (CP). In previous studies, CT has been determined

to be composed of terrestrial precursor material from soils and vegetation and shown to

have higher production during warm months (Ishii and Boyer 2012; Tanaka et al. 2014;

Kulkarni et al. 2017). These high-molecular-weight, aromatic compounds are relatively

resistant to further decomposition and represent fulvic acid-like components (Yamashita

et al. 2010; Gonçalves-Araujo et al. 2016). CR is representative of microbially

reprocessed OM and is present in both freshwater and marine systems (Murphy et al.

2011; Lapierre and del Giorgio 2014). This microbial signature reflects the presence of

quinone-like materials used in redox reactions and are byproducts of microbial activity

but can also be produced by degradation of lignin in plant tissues (Cory and McKnight

2005; Yamashita et al. 2010). Whether the microbial matter reflects recent heterotrophic

activity along groundwater flow paths or is due to microbial activity within soils that are

entrained during infiltration cannot be determined without compound-specific analyses

(Ishii and Boyer 2012). Finally, CP has been shown to be a mix of tryptophan and tyrosine amino acids (Yamashita and Tanoue 2003). Tryptophan, which seems to dominate the protein-like signature of spring samples in this study, is a labile compound that has been examined in wastewaters and in marine waters as a byproduct of phytoplankton and bacteria (Baker 2005; Jørgensen et al. 2011). For terrestrial substrates, tryptophan has been examined to a lesser extent, but it is likely sourced from soil or aquatic heterotrophic microbes or from polyphenols in aged plant matter (Maie et al.

2007; Murphy et al. 2008). Like CR, CP cannot be identified as a marker of heterotrophic

metabolism or a derivative of plant matter without examination of specific compounds. If residence time of spring water is long, then it is plausible that labile plant proteins would be cycled within the flow system, but if residence time is short then plant proteins could remain unmodified.

Fluorescence indices are commonly used to characterize DOM based on ratios of fluorescence intensities at various wavelengths. Fluorescence index (FI) values above

~1.7 represent microbially-derived fulvic acids, whereas values near ~1.4 represent

terrestrially-derived fulvic acids (McKnight et al. 2001). Values from springs in this study range from 1.3 to 2.0, with an average of 1.6. However, because of low DOM, fluorescence intensities were low enough in many samples that the wavelengths used for

FI were near background instrument values. Therefore, FI is likely not an accurate representation of DOM in this study. The biological index (BIX) reveals activity of recent autochthonous productivity, with values < 0.6 representing no autochthonous component, values >1 representing strong biological or aquatic bacterial origin, and intervening values representing contributions from each source (Huguet et al. 2009). BIX values for springs range from 0.5 to 1.1, but BIX does not correlate with any other parameters. Like

FI, the wavelengths required for BIX are low in some samples and thus may not be representative of DOM in this study. The humification index (HIX), which describes the relative humification of DOM, ranges from 0 (fresh, labile, low- molecular-weight compounds) to 1 (the most humified, aromatic, and high-molecular-weight compounds)

(Ohno 2002). HIX is the sole index that corresponded with any parameter in the study, as it was relatively correlated with the terrestrially-sourced components. CT and CR had

2 individual r values of 0.28 and 0.17, respectively, but when CT and CR were combined

the correlation increased to 0.64. This suggests that the terrestrial and microbial humic-

like fDOM sources are derived from similar sources, probably humified soil components.

For statistical analyses, they have been grouped together as surficial terrestrial

components.

4.3 Conceptual flow system of the Shivwits Plateau

The stable water-isotope plot (Fig. 5) revealed a strong evaporative signature for

all springs within the study area. The slope of ~5 for the spring trend line is similar to

evaporative signals seen in other water-isotope studies (Clark and Fritz 1997), but is

lower than the regional-springs trend of Springer et al. (2017), indicating that water is

experiencing greater evaporation on the surface or in the soil zone prior to recharge. The

more positive water-isotope values of the Supai Gp., along with higher DOC and

terrestrial DOM content and lower ion concentrations, caused it to plot as a distinct group

in the PCA (Fig. 9). The more positive water-isotope values indicate that the precipitation

occurred during a warmer period and/or experienced more evaporation after precipitation.

None of the Supai Gp. springs sampled in May were still flowing in Fall, which suggests that those springs are largely ephemeral and sourced by runoff on the canyon slopes. As runoff, the water could acquire the high DOC and DOM content while remaining relatively dilute because of short residence time in the subsurface.

Dansil and Mud springs also plotted as a distinct group on the PCA for both May

2+ 2- - and Fall samples (Fig. 9). They are characterized by high Ca and SO4 , lower HCO3 ,

and higher CP values. This shows that these Toroweap springs on the eastern side of the

plateau are predominantly flowing through gypsum beds rather than carbonates. The high

CP loading is due to the Dansil sample in Fall having the highest CP percentage of all

springs at 50%; Dansil-May and both Mud Sp. samples have less than 16%. These

springs appear to have similar recharge mechanisms to other springs in this study and

only differ in flow-path lithology. Dansil and Mud springs are geochemically similar to

springs that discharge from gypsiferous aquifers in other areas (Cardenal et al. 1994;

Bischoff et al. 1994; Chiesi et al. 2010; D’Angeli et al. 2017).

Most other springs, aside from Redwall Ls. springs, plot together within the black

ellipse on the PCA (Fig. 9). These springs are characterized by higher pCO2 values, more

- negative water-isotope values, and high HCO3 concentrations. The more-depleted water isotopes indicate that recharge occurred during cooler temperatures, likely during winter, but the evaporative trend seen in all springs indicates that these waters experienced

- evaporation before or during infiltration. High HCO3 concentrations are driven by the

high pCO2 values, which increase water aggressiveness. The high pCO2 values indicate

that the water spent an extensive amount of time within the soil zone in vegetated areas,

because pCO2 from vegetation generally exceeds atmospheric pCO2 (Peyraube et al.

2013). The likely storage location for these springs is within epikarst in the Kaibab Fm.

and pseudokarst of the Shivwits Basalt flows. Pseudokarst is a karst-like morphology

produced by solutional and non-solutional processes, such as solution of feldspars in silicates and rheogenic pseudokarst present in lava flows as lava tube piping and conduits

(Grimes 1975; Halliday 2007). Epikarst or pseudokarst storage could hold waters near the surface to allow for an evaporative signal to develop in water isotopes, which has been shown to develop as deep as 10 m in arid soils and can develop within a few days following the last infiltration event (Fontes et al. 1986; Sprenger et al. 2016). In areas

with high epikarst fracture density and interconnection of openings, diffuse percolation and flow below the epikarst zone has been observed (Petrella et al. 2007). Geochemistry

- of this group of springs is also similar to the geochemistry (pCO2, HCO3 , and other major ion concentrations) of epikarst springs in other studies (Li et al. 2008; Li et al.

2010; Yang et al. 2012; Peyraube et al. 2013). Storage can occur in the upper epikarst, with transmission through vertical shafts and fissures, and secondary storage or drainage from a lower phreatic zone to the spring discharge (Jacob et al. 2009).

The red ellipse (Fig. 9) contains all Redwall Ls. springs and plots near the middle of the PCA, meaning they are near averages of the loadings in PC1 and PC2. This suggests that the Redwall springs are a mixture of other water sources from upper epikarst storage, gypsiferous aquifers, and runoff (Fig. 11). Various faults and fractures in the study area (Fig. 2) likely transmit upper epikarst waters down to Redwall springs, along with two breccia pipes that are near the largest spring in the study area,

Cottonwood Sp., likely accounting for its high discharge relative to other springs.

Figure 11 – Conceptual flow systems of the Shivwits Plateau. Color of arrows represent ellipse group colors from the PCA.

4.4 Seasonality of springs

From the water-isotope plot, all springs appear to have been sourced from winter

precipitation with various degrees of evaporation. This agrees with the findings of other

studies in the region that winter precipitation, as rain or snow, is the dominant source of

runoff and recharge, whereas summer precipitation is mostly lost to evapotranspiration

(Bills et al. 2007; Springer et al. 2017; Jones et al. 2018; Tobin et al. 2018; Solder et al.

2020). Most sampled springs increased in δ18O and δ2H from May to Fall, indicating that

temperatures during recharge increased between the sampling periods as well. Springs on

the Kaibab Plateau have been shown to have extremely short residence times, from < 1

month to 3 months (Jones et al. 2018). Conversely, springs on the South Rim have been

found to have groundwater ages from 6 to >10,000 years, likely due to regional flow

from the San Francisco Peak’s (Solder et al. 2020; Solder and Beisner 2020). However,

both the Kaibab and Shivwits Plateau’s are local and regional highs and thus are less

likely to be affected by larger regional flow systems. Therefore, assuming similar

residence times for Shivwits Plateau to the Kaibab Plateau springs due to similar

lithostratigraphy and regional prominence, this increase in water-isotope values likely reflects a transition from early winter recharge for May samples to late winter-early spring recharge for Fall samples. Tincanebitts, Dansil, and Hidden springs do not follow this trend, having more negative δ18O and δ2H in Fall than in May. This offset could be

due to a larger contribution of recent waters for May samples, as more recent precipitation would increase the water-isotope values relative to the epikarst storage. In

Fall, when any recent recharge would have already moved through the system, the return

to more negative water isotopes suggests a return to a larger contribution of waters

coming from epikarst storage.

13 Another signature of seasonality for these springs in an increase in δ CDOC from

May to Fall for all springs except for Dansil Sp. (Fig. 12). As Dansil Sp. was one of the

13 Figure 12 – Variation in δ CDOC for repeated samples between May and Fall.

springs with a possible alternate storage component, its deviation to a more negative

value could be explained by that. It also increased from containing no protein-like DOM

13 in May to 50% protein-like DOM in Fall, but whether this caused the decrease in δ CDOC is unclear, as other springs that increased in protein-like DOM did not follow the same

13 trend. The change in δ CDOC could also be due to a change in microbial or vegetation

activity. Microbial degradation preferentially utilizes 12C relative to 13C, which results in

enriched OM and depleted CO2 (Zhou et al. 2015). Reduced microbial activity in winter

13 due to cooler temperatures would result in more negative δ CDOC values. In Fall samples, increased temperatures of recharge waters would likely be accompanied by an increase in

13 microbial activity, thus resulting in more positive δ CDOC values. However, it is unclear

13 if this is the cause, because δ CDIC values and DOC concentrations do not covary with

13 δ CDOC values, as would be expected from seasonality in microbial degradation.

CHAPTER FIVE: CONCLUSIONS

The groundwater of the Shivwits Plateau is a complex system that was able to be characterized by semi-annual sampling for various isotopic and geochemical parameters.

Groundwater flow on the Shivwits appears to be driven by recharge during winter, with summer monsoonal events likely being lost to evapotranspiration. Flow systems can be separated into four groups by PCA: 1) a shallow soil-epikarst-fracture system, 2) a flow path through gypsiferous beds of the Toroweap Fm. on the eastern side of the plateau, 3) a short, canyon-slope runoff dominated flow path through the Supai Gp., and 4) a deeper, complex flow system in the Redwall Ls., with characteristics of all other flow systems showing a mixing of sources. Most springs on the plateau are driven by epikarst storage and relatively isolated from regional flow systems.

Examining solute and isotopic signatures of each spring coupled with multivariate analyses through PCA revealed that spring geochemistry can be adequately characterized for determining groundwater flow systems while individual parameters may not express

13 consistent variability. δ CDIC values of all springs fall within the range of expected values for dissolution of carbonate sources by carbonic acid with no clear seasonality or

13 spatial trends. δ CDOC values of springs show significant differences between May and

Fall (May: –27.0‰; Fall: –24.4‰; p < 0.005), likely because of changing plant respiration and soil microbial activity between cold and warm seasons. fDOM component analyses reveal contain 9–59% of DOM sourced from terrestrial vegetation, 23–63% from microbial cycling of terrestrial OM by soil microbes or heterotrophic organisms along flow paths, and 0–50% from protein-like material likely composed of terrestrial plant phenols or heterotrophic microbial byproducts.

Future work for the Shivwits Plateau should focus on higher-resolution sampling for springs of interest and compound-specific analyses of fDOM fractions, which could reveal more characteristics of DOM sources within the system. This can be achieved through methods such as 13C-nuclear magnetic resonance spectroscopy, Fourier- transform ion cyclotron mass spectrometry, or high performance liquid chromatography- size exclusion chromatography, which were beyond the scope of this study (Hedges et al.

1992; Kaiser et al. 2003; Her et al. 2003; Nebbioso and Piccolo 2013). Although the FI has been shown to be useful in groundwater at other sites (Yang and Hur 2014), fDOM was too low in many of our samples for FI to be usable. Low DOM and DOC concentrations can result in lower analytical accuracy and precision. Similarly,

13 pronounced seasonal variations were evident from water isotopes and δ CDOC values, but

low-resolution temporal sampling limits an understanding of the causes of variability.

This study reveals that initial characterization of a complex groundwater flow

system can be completed with semi-annual sampling of relatively small volumes. Given that common methods to characterize groundwater systems can be time-intensive, costly,

and challenging in remote areas, this approach could be applied elsewhere for initial

characterization of groundwater resources and determine if a more complex

characterization is required. As groundwater resources and the Colorado River are both

being depleted and overdrawn, characterizing alternate groundwater resources will be

important for the region as well as in other areas that experience shortages in water

resources that may be aggravated by climate change.

APPENDICES APPENDIX A: Photographs of springs

Appendix figure A1 – Burnt Spring

Appendix figure A2 – Cottonwood Spring – close-up

Appendix figure A3 – Cottonwood Spring 2 – distal view

Appendix figure A4 – Twin Creek Canyon Seep

Appendix figure A5 – Twin Key Spring

Appendix figure A6 – Twin Spring

Appendix figure A7 – Hidden Spring

Appendix figure A8 – Tincanebitts Spring

Appendix figure A9 – Mud Spring

Appendix figure A10 – Green Spring

APPENDIX B: Remaining geochemical data

Elements not used in analyses due to low concentrations. MDL – below detection limit.

Site Season Al Sb As Br Cd CO3 Cr ~ ~ (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) 202 Ephemeral May MDL MDL MDL MDL MDL MDL MDL Burnt Spring May MDL MDL MDL MDL MDL MDL MDL Cottonwood Spring May MDL MDL MDL MDL MDL MDL MDL Dansil Spring May MDL MDL MDL MDL MDL MDL MDL Fall MDL MDL MDL MDL MDL MDL MDL Green Spring – May 8 May MDL MDL MDL MDL MDL MDL MDL Green Spring – May 10 May MDL MDL MDL MDL MDL MDL MDL Green Spring – May 11 May MDL MDL MDL MDL MDL MDL MDL Green Spring – May 12 May MDL MDL MDL MDL MDL MDL MDL Green Spring Fall MDL MDL MDL MDL MDL MDL MDL Hidden Spring May MDL MDL MDL MDL MDL MDL MDL Fall 0.07 MDL MDL MDL MDL MDL MDL Joe Spring May MDL MDL MDL MDL MDL MDL MDL Fall 0.08 MDL MDL MDL MDL MDL MDL Lost Spring May MDL MDL MDL MDL MDL MDL MDL Lower Spring May MDL MDL MDL MDL MDL MDL MDL Mud Spring May MDL MDL MDL MDL MDL MDL MDL Fall MDL MDL MDL MDL MDL MDL MDL Pen Pocket May MDL MDL MDL MDL MDL MDL MDL Snap Sp Fall MDL MDL MDL MDL MDL MDL MDL Tincanebitts Spring May MDL MDL MDL MDL MDL MDL MDL Fall MDL MDL MDL MDL MDL MDL MDL Twin Canyon Seep May MDL MDL MDL MDL MDL MDL MDL Twin Key Spring May MDL MDL MDL MDL MDL MDL MDL Fall MDL MDL 0.018 MDL MDL MDL MDL Twin Spring May MDL MDL MDL MDL MDL MDL MDL Fall 0.07 MDL MDL MDL MDL MDL MDL

Site Co Cu Au Fe Pb Li Mn Ni ~ (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) 202 Ephemeral MDL MDL MDL MDL MDL 0.01 0 MDL Burnt Spring MDL MDL MDL 0.02 MDL 0.01 0 MDL Cottonwood Spring MDL MDL MDL 0.01 MDL 0.01 MDL MDL Dansil Spring MDL MDL MDL MDL MDL 0.01 MDL MDL MDL MDL MDL MDL MDL 0.01 MDL MDL Green Spring – May 8 MDL MDL MDL MDL MDL 0.01 MDL MDL Green Spring – May 10 MDL MDL MDL MDL MDL 0.01 0 MDL Green Spring – May 11 MDL MDL MDL 0.01 MDL 0.01 MDL MDL Green Spring – May 12 MDL MDL MDL 0.03 MDL 0 MDL MDL Green Spring MDL MDL MDL MDL MDL MDL 0 MDL Hidden Spring MDL MDL MDL 0.04 MDL 0.01 MDL MDL MDL MDL MDL MDL MDL 0.01 MDL MDL Joe Spring MDL MDL MDL MDL MDL 0.01 MDL MDL MDL MDL MDL 0.02 MDL 0.01 MDL MDL Lost Spring MDL MDL MDL 0.02 MDL 0.01 0.02 MDL Lower Spring MDL MDL MDL MDL MDL 0 0 MDL Mud Spring MDL MDL MDL MDL MDL 0.01 0 MDL MDL MDL MDL MDL MDL 0.01 0 MDL Pen Pocket MDL MDL MDL 0.01 MDL 0 0 MDL Snap Sp MDL MDL MDL MDL MDL 0 0 MDL Tincanebitts Spring MDL MDL MDL MDL MDL 0 MDL MDL MDL MDL MDL 0.02 MDL 0 MDL MDL Twin Canyon Seep MDL MDL MDL MDL MDL 0.01 MDL MDL Twin Key Spring MDL MDL MDL 0.01 MDL 0.01 0 MDL MDL MDL MDL MDL MDL 0.01 MDL MDL Twin Spring MDL MDL MDL 0.02 MDL 0 MDL MDL MDL MDL MDL 0.02 MDL MDL MDL MDL

Site Season Se Ag Tl Sn V Zn ~ ~ (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) (mg/L) 202 Ephemeral May MDL MDL MDL MDL MDL MDL Burnt Spring May MDL MDL MDL MDL MDL MDL Cottonwood Spring May MDL MDL MDL MDL MDL 0.01 Dansil Spring May MDL MDL MDL MDL MDL 0 Fall MDL MDL MDL MDL MDL MDL Green Spring – May 8 May MDL MDL MDL MDL MDL MDL Green Spring – May 10 May MDL MDL MDL MDL MDL 0.01 Green Spring – May 11 May MDL MDL MDL MDL MDL 0 Green Spring – May 12 May MDL MDL MDL MDL MDL 0 Green Spring Fall MDL MDL MDL MDL MDL MDL Hidden Spring May MDL MDL MDL MDL MDL 0.01 Fall MDL MDL MDL MDL MDL MDL Joe Spring May MDL MDL MDL MDL MDL MDL Fall MDL MDL MDL MDL MDL MDL Lost Spring May MDL MDL MDL MDL MDL 0.01 Lower Spring May MDL MDL MDL MDL MDL 0 Mud Spring May MDL MDL MDL MDL MDL 0.01 Fall MDL MDL MDL MDL MDL MDL Pen Pocket May MDL MDL MDL MDL MDL 0 Snap Sp Fall MDL MDL MDL MDL MDL 0.01 Tincanebitts Spring May MDL MDL MDL MDL MDL MDL Fall MDL MDL MDL MDL MDL MDL Twin Canyon Seep May MDL MDL MDL MDL MDL 0.01 Twin Key Spring May MDL MDL MDL MDL MDL 0.01 Fall MDL MDL MDL MDL MDL 0.01 Twin Spring May MDL MDL MDL MDL MDL MDL Fall MDL MDL MDL MDL MDL 0.05

REFERENCES

Adinolfi Falcone R, Falgiani A, Parisse B, Petitta M, Spizzico M, Tallini M (2008) Chemical and isotopic (δ18O‰, δ2H‰, δ13C‰, 222Rn) multi-tracing for groundwater conceptual model of carbonate aquifer (Gran Sasso INFN underground laboratory – central Italy). J Hydrol 357:368–388 . https://doi.org/10.1016/j.jhydrol.2008.05.016

Al-Sakkaf RA, Zhou Y, Hall MJ (1999) A strategy for controlling groundwater depletion in the Sa’dah Plain, Yemen. Int J Water Resour Dev 15:349–365 . https://doi.org/10.1080/07900629948862

Amiotte-Suchet P, Aubert D, Probst JL, Gauthier-Lafaye F, Probst A, Andreux F, Viville D (1999) δ13C pattern of dissolved inorganic carbon in a small granitic catchment: the Strengbach case study (Vosges mountains, France). Chem Geol 159:129–145

Amiotte-Suchet P, Linglois N, Leveque J, Andreux F (2007) 13C composition of dissolved organic carbon in upland forested catchments of the Morvan Mountains (France): Influence of coniferous and deciduous vegetation. J Hydrol 335:354– 363 . https://doi.org/10.1016/j.jhydrol.2006.12.002

Amundson R, Stern L, Baisden T, Wang Y (1998) The isotopic composition of soil and soil-respired CO2. Geoderma 82:83–114 . https://doi.org/10.1016/S0016- 7061(97)00098-0

Aravena R, Schiff SL, Trumbore SE, Dillon PJ, Elgood R (1992) Evaluating dissolved inorganic carbon cycling in a forested lake watershed using carbon isotopes. Radiocarbon 34:636–645 . https://doi.org/10.1017/S003382220006392X

Assayag N, Rivé K, Ader M, Jézéquel D, Agrinier P (2006) Improved method for isotopic and quantitative analysis of dissolved inorganic carbon in natural water samples. Rapid Commun Mass Spectrom 20:2243–2251 . https://doi.org/10.1002/rcm.2585

Atekwana EA, Richardson DS (2004) Geochemical and isotopic evidence of a groundwater source in the Corral Canyon meadow complex, central Nevada, USA. Hydrol Process 18:2801–2815 . https://doi.org/10.1002/hyp.1495

Baker A (2005) Thermal fluorescence quenching properties of dissolved organic matter. Water Res 39:4405–4412 . https://doi.org/10.1016/j.watres.2005.08.023

Billingsley GH, Beus SS (1985) The Surprise Canyon Formation - an Upper Mississippian and Lower Pennsylvanian(?) rock unit in the Grand Canyon, Arizona. US Geol Surv Bull 1605-A:A27–A33

Bills DJ, Flynn ME, Monroe SA (2007) Hydrogeology of the Coconino Plateau and adjacent areas, Coconino and Yavapai Counties, Arizona (ver. 1.1, March 2016). US Geol Surv Sci Investig Rep 2005–5222:1–101 . http://dx.doi.org/10.3133/sir20055222

Bills DJ, Tillman FD, Anning DW, Antweiler RC, Kraemer TF (2010) Historical and 2009 water chemistry of wells, perennial and intermittent streams, and springs in northern Arizona. US Geol Surv Sci Investig Rep 2010–5025:141–282

Birdwell JE, Engel AS (2010) Characterization of dissolved organic matter in cave and spring waters using UV–Vis absorbance and fluorescence spectroscopy. Org Geochem 41:270–280 . https://doi.org/10.1016/j.orggeochem.2009.11.002

Bischoff JL, Juliá R, Shanks WC, Rosenbauer RJ (1994) Karstification without carbonic acid: Bedrock dissolution by gypsum-driven dedolomitization. Geology 22:995– 998 . https://doi.org/10.1130/0091-7613(1994)022<0995:KWCABD>2.3.CO;2

Cardenal J, Benavente J, Cruz-Sanjulián JJ (1994) Chemical evolution of groundwater in Triassic gypsum-bearing carbonate aquifers (Las Alpujarras, southern Spain). J Hydrol 161:3–30 . https://doi.org/10.1016/0022-1694(94)90119-8

Cerling TE, Solomon DK, Quade J, Bowman JR (1991) On the isotopic composition of carbon in soil carbon dioxide. Geochim Cosmochim Acta 55:3403–3405 . https://doi.org/10.1016/0016-7037(91)90498-T

Chandrakanth MG, Romm J (1990) Groundwater depletion in India—institutional management regimes. Nat Resour J 30:485–501

Chen M, Jung J, Lee YK, Hur J (2018) Surface accumulation of low molecular weight dissolved organic matter in surface waters and horizontal off-shelf spreading of nutrients and humic-like fluorescence in the Chukchi Sea of the Arctic Ocean. Sci Total Environ 639:624–632 . https://doi.org/10.1016/j.scitotenv.2018.05.205

Chiesi M, De Waele J, Forti P (2010) Origin and evolution of a salty gypsum/anhydrite karst spring: the case of Poiano (Northern Apennines, Italy). Hydrogeol J 18:1111–1124 . https://doi.org/10.1007/s10040-010-0576-2

Christensen NS, Wood AW, Voisin N, Lettenmaier DP, Palmer RN (2004) The effects of climate change on the hydrology and water resources of the Colorado River Basin. Clim Change 62:337–363 . https://doi.org/10.1023/B:CLIM.0000013684.13621.1f

Clark ID, Fritz P (1997) Environmental Isotopes in Hydrogeology. CRC Press

Coble PG (1996) Characterization of marine and terrestrial DOM in seawater using excitation-emission matrix spectroscopy. Mar Chem 51:325–346

Cole JJ, Prairie YT, Caraco NF, McDowell WH, Tranvik LJ, Striegl RG, Duarte CM, Kortelainen P, Downing JA, Middelburg JJ, Melack J (2007) Plumbing the global carbon cycle: integrating inland waters into the terrestrial carbon budget. Ecosystems 10:172–185 . https://doi.org/10.1007/s10021-006-9013-8

Coplen TB, Brand WA, Gehre M, Gröning M, Meijer HAJ, Toman B, Verkouteren RM (2006) New Guidelines for δ13C Measurements. Anal Chem 78:2439–2441 . https://doi.org/10.1021/ac052027c

Cory RM, McKnight DM (2005) Fluorescence spectroscopy reveals ubiquitous presence of oxidized and reduced quinones in dissolved organic matter. Environ Sci Technol 39:8142–8149 . https://doi.org/10.1021/es0506962

Craig H (1961) Isotopic variations in meteoric waters. Science 133:1702–1703 . https://doi.org/10.1126/science.133.3465.1702

Dainard PG Absorbance and fluorescence characteristics of dissolved organic matter in North Atlantic, Pacific and Arctic Oceans. Thesis, Trent University

D’Angeli IM, Serrazanetti DI, Montanari C, Vannini L, Gardini F, De Waele J (2017) Geochemistry and microbial diversity of cave waters in the gypsum karst aquifers of Emilia Romagna region, Italy. Sci Total Environ 598:538–552 . https://doi.org/10.1016/j.scitotenv.2017.03.270

Deirmendjian L, Loustau D, Augusto L, Lafont S, Chipeaux C, Poirier D, Abril G (2018) Hydro-ecological controls on dissolved carbon dynamics in groundwater and export to streams in a temperate pine forest. Biogeosciences 15:669–691 . https://doi.org/10.5194/bg-15-669-2018

Doctor DH, Kendall C, Sebestyen SD, Shanley JB, Ohte N, Boyer EW (2008) Carbon isotope fractionation of dissolved inorganic carbon (DIC) due to outgassing of carbon dioxide from a headwater stream. Hydrol Process 22:2410–2423 . https://doi.org/10.1002/hyp.6833

Famiglietti JS (2014) The global groundwater crisis. Nat Clim Change 4:945–948 . https://doi.org/10.1038/nclimate2425

Finlay JC, Kendall C (2007) Stable isotope tracing of temporal and spatial variability in organic matter sources to freshwater ecosystems. In: Michener R, Lajtha K (eds) Stable isotopes in ecology and environmental science. Blackwell Publishing Ltd, Oxford, UK, pp 283–333

Flint LE, Flint AL (2007) Regional analysis of ground-water recharge. In: Stonestrom DA, Constantz J, Ferre TPA, Leake SA (eds) Ground-water recharge in the arid and semiarid southwestern United States. U.S. Geological Survey Professional Paper 1703, pp 29–59

Florea LJ (2013) Isotopes of carbon in a karst aquifer of the Cumberland Plateau of Kentucky, USA. Acta Carsologica 42: . https://doi.org/10.3986/ac.v42i2-3.668

Florea LJ, Banks SM, Forray FL (2019) Importance of suspended sediments and dissolved organic carbon to carbon exports in karst – The Vadu Crişului karst basin in the Pădurea Craiului Mountains, Romania. Chem Geol 527:118735 . https://doi.org/10.1016/j.chemgeo.2018.04.015

Flowers RM, Wernicke BP, Farley KA (2008) Unroofing, incision, and uplift history of the southwestern Colorado Plateau from apatite (U-Th)/He thermochronometry. Geol Soc Am Bull 120:571–587 . https://doi.org/10.1130/B26231.1

Fontes JC, Yousfi M, Allison GB (1986) Estimation of long-term, diffuse groundwater discharge in the northern Sahara using stable isotope profiles in soil water. J Hydrol 86:315–327 . https://doi.org/10.1016/0022-1694(86)90170-8

Fritz P, Reardon EJ, Barker J, Brown RM, Cherry JA, Killey RWD, McNaughton D (1978) The carbon isotope geochemistry of a small groundwater system in northeastern Ontario. Water Resour Res 14:1059–1067 . https://doi.org/10.1029/WR014i006p01059

Gonçalves-Araujo R, Granskog MA, Bracher A, Azetsu-Scott K, Dodd PA, Stedmon CA (2016) Using fluorescent dissolved organic matter to trace and distinguish the origin of Arctic surface waters. Sci Rep 6:33978 . https://doi.org/10.1038/srep33978

Grimes KG (1975) Pseudokarst: definition and types. In: Proceedings of the 10th Biennial Conference, Australian Speleological Federation, Brisbane. pp 6–10

Halliday WR (2007) Pseudokarst in the 21st century. J Cave Karst Stud 69:103–113

Han G, Tang Y, Wu Q (2010) Hydrogeochemistry and dissolved inorganic carbon isotopic composition on karst groundwater in Maolan, southwest China. Environ Earth Sci 60:893–899 . https://doi.org/10.1007/s12665-009-0226-y

Hedges JI, Hatcher PG, Ertel JR, Meyers-Schulte KJ (1992) A comparison of dissolved humic substances from seawater with Amazon River counterparts by 13C-NMR spectrometry. Geochim Cosmochim Acta 56:1753–1757 . https://doi.org/10.1016/0016-7037(92)90241-A

Hélie J-F, Hillaire-Marcel C, Rondeau B (2002) Seasonal changes in the sources and fluxes of dissolved inorganic carbon through the St. Lawrence River—isotopic and chemical constraint. Chem Geol 186:117–138 . https://doi.org/10.1016/S0009-2541(01)00417-X

Hendricks DM (1985) Arizona soils. College of Agriculture, University of Arizona

Her N, Amy G, McKnight D, Sohn J, Yoon Y (2003) Characterization of DOM as a function of MW by fluorescence EEM and HPLC-SEC using UVA, DOC, and fluorescence detection. Water Res 37:4295–4303 . https://doi.org/10.1016/S0043- 1354(03)00317-8

Huang S, Wang Y, Ma T, Tong L, Wang Y, Liu C, Zhao L (2015) Linking groundwater dissolved organic matter to sedimentary organic matter from a fluvio-lacustrine aquifer at Jianghan Plain, China by EEM-PARAFAC and hydrochemical analyses. Sci Total Environ 529:131–139 . https://doi.org/10.1016/j.scitotenv.2015.05.051

Hudson N, Baker A, Reynolds D (2007) Fluorescence analysis of dissolved organic matter in natural, waste and polluted waters—a review. River Res Appl 23:631– 649 . https://doi.org/10.1002/rra.1005

Huguet A, Vacher L, Relexans S, Saubusse S, Froidefond JM, Parlanti E (2009) Properties of fluorescent dissolved organic matter in the Gironde Estuary. Org Geochem 40:706–719 . https://doi.org/10.1016/j.orggeochem.2009.03.002

Huntoon PW (1974) The karstic groundwater basins of the Kaibab Plateau, Arizona. Water Resour Res 10:579–590 . https://doi.org/10.1029/WR010i003p00579

Huntoon PW (1970) The hydro-mechanics of the ground water system in the southern portion of the Kaibab Plateau, Arizona. Ph.D. dissertation, University of Arizona

Huntoon PW (1977) Cambrian stratigraphic nomenclature and ground-water prospecting failures on the Hualapai Plateau, Arizona. Groundwater 15:426–433 . https://doi.org/10.1111/j.1745-6584.1977.tb03190.x

Ingraham NL, Zukosky K, Kreamer DK (2001) Application of stable isotopes to identify problems in large-scale water transfer in Grand Canyon National Park. Environ Sci Technol 35:1299–1302 . https://doi.org/10.1021/es0015186

Ishii SKL, Boyer TH (2012) Behavior of reoccurring PARAFAC components in fluorescent dissolved organic matter in natural and engineered systems: a critical review. Environ Sci Technol 46:2006–2017 . https://doi.org/10.1021/es2043504

Jacob T, Chery J, Bayer R, Le Moigne N, Boy J-P, Vernant P, Boudin F (2009) Time- lapse surface to depth gravity measurements on a karst system reveal the dominant role of the epikarst as a water storage entity. Geophys J Int 177:347– 360 . https://doi.org/10.1111/j.1365-246X.2009.04118.x

Jones CJR, Springer AE, Tobin BW, Zappitello SJ, Jones NA (2018) Characterization and hydraulic behaviour of the complex karst of the Kaibab Plateau and Grand Canyon National Park, USA. Geol Soc Lond Spec Publ 466:237–260 . https://doi.org/10.1144/SP466.5

Jørgensen L, Stedmon CA, Kragh T, Markager S, Middelboe M, Søndergaard M (2011) Global trends in the fluorescence characteristics and distribution of marine dissolved organic matter. Mar Chem 126:139–148 . https://doi.org/10.1016/j.marchem.2011.05.002

Kaiser E, Simpson AJ, Dria KJ, Sulzberger B, Hatcher PG (2003) Solid-state and multidimensional solution-state NMR of solid phase extracted and ultrafiltered riverine dissolved organic matter. Environ Sci Technol 37:2929–2935 . https://doi.org/10.1021/es020174b

Käss W (1998) Tracing technique in geohydrology: AA Balkema. Brookfield VT

Kingston JD, Hill A, Marino BD (1994) Isotopic evidence for Neogene hominid paleoenvironments in the Kenya Rift Valley. Science 264:955–959 . https://doi.org/10.1126/science.264.5161.955

Konikow LF, Kendy E (2005) Groundwater depletion: A global problem. Hydrogeol J 13:317–320 . https://doi.org/10.1007/s10040-004-0411-8

Kulkarni HV, Mladenov N, Johannesson KH, Datta S (2017) Contrasting dissolved organic matter quality in groundwater in Holocene and Pleistocene aquifers and implications for influencing arsenic mobility. Appl Geochem 77:194–205 . https://doi.org/10.1016/j.apgeochem.2016.06.002

Lang SQ, Bernasconi SM, Früh-Green GL (2012) Stable isotope analysis of organic carbon in small (µg C) samples and dissolved organic matter using a GasBench preparation device: δ13C of organic compounds by GasBench. Rapid Commun Mass Spectrom 26:9–16 . https://doi.org/10.1002/rcm.5287

Lapierre J-F, del Giorgio PA (2014) Partial coupling and differential regulation of biologically and photochemically labile dissolved organic carbon across boreal aquatic networks. Biogeosciences 11:5969–5985 . https://doi.org/10.5194/bg-11- 5969-2014

Li S-L, Liu C-Q, Lang Y-C, Tao F, Zhao Z, Zhou Z (2008) Stable carbon isotope biogeochemistry and anthropogenic impacts on karst ground water, Zunyi, Southwest China. Aquat Geochem 14:211–221 . https://doi.org/10.1007/s10498- 008-9033-4

Li S-L, Liu C-Q, Li J, Lang Y-C, Ding H, Li L (2010) Geochemistry of dissolved inorganic carbon and carbonate weathering in a small typical karstic catchment of Southwest China: Isotopic and chemical constraints. Chem Geol 277:301–309 . https://doi.org/10.1016/j.chemgeo.2010.08.013

Lichtfouse É, Dou S, Girardin C, Grably M, Balesdent J, Behar F, Vandenbroucke M (1995) Unexpected 13C-enrichment of organic components from wheat crop soils: evidence for the in situ origin of soil organic matter. Org Geochem 23:865–868 . https://doi.org/10.1016/0146-6380(95)80009-G

Long SP, Hutchin PR (1991) Primary production in grasslands and coniferous forests with climate change: an overview. Ecol Appl 1:139–156 . https://doi.org/10.2307/1941807

Lovelace JK, Nielsen MG, Read AL, Murphy CJ, Maupin MA (2020) Estimated groundwater withdrawals from principal aquifers in the United States, 2015. U.S. Geological Survey, Reston, VA

Maie N, Scully NM, Pisani O, Jaffé R (2007) Composition of a protein-like fluorophore of dissolved organic matter in coastal wetland and estuarine ecosystems. Water Res 41:563–570 . https://doi.org/10.1016/j.watres.2006.11.006

Maupin MA, Barber NL (2005) Estimated withdrawals from principal aquifers in the United States, 2000. US Geol Surv Circ 1279:1–46 . https://doi.org/10.3133/cir1279

McKee ED (1982) The Supai Group of Grand Canyon. US Geol Surv Prof Pap 1173:1– 504

McKnight DM, Boyer EW, Westerhoff PK, Doran PT, Kulbe T, Andersen DT (2001) Spectrofluorometric characterization of dissolved organic matter for indication of precursor organic material and aromaticity. Limnol Oceanogr 46:38–48

Mervenne C (2015) Isotope ecology of temperate conifers. Thesis, University of Michigan

Metzger DG (1961) Geology in relation to availability of water along the South Rim Grand Canyon National Park, Arizona. US Geol Surv Water Supply Pap 1475– C:105–137

Moody CS, Worrall F (2017) Modeling rates of DOC degradation using DOM composition and hydroclimatic variables. J Geophys Res Biogeosciences 122:1175–1191 . https://doi.org/10.1002/2016JG003493

Murphy KR, A. Stedmon C, Graeber D, Bro R (2013) Fluorescence spectroscopy and multi-way techniques. PARAFAC. Anal Methods 5:6557–6566 . https://doi.org/10.1039/C3AY41160E

Murphy KR, A. Stedmon C, Wenig P, Bro R (2014) OpenFluor – an online spectral library of auto-fluorescence by organic compounds in the environment. Anal Methods 6:658–661 . https://doi.org/10.1039/C3AY41935E

Murphy KR, Hambly A, Singh S, Henderson RK, Baker A, Stuetz R, Khan SJ (2011) Organic matter fluorescence in municipal water recycling schemes: toward a unified PARAFAC model. Environ Sci Technol 45:2909–2916 . https://doi.org/10.1021/es103015e

Murphy KR, Stedmon CA, Waite TD, Ruiz GM (2008) Distinguishing between terrestrial and autochthonous organic matter sources in marine environments using fluorescence spectroscopy. Mar Chem 108:40–58 . https://doi.org/10.1016/j.marchem.2007.10.003

Nebbioso A, Piccolo A (2013) Molecular characterization of dissolved organic matter (DOM): a critical review. Anal Bioanal Chem 405:109–124 . https://doi.org/10.1007/s00216-012-6363-2

Nowak ME, Schwab VF, Lazar CS, Behrendt T, Kohlhepp B, Totsche KU, Küsel K, Trumbore SE (2017) Carbon isotopes of dissolved inorganic carbon reflect utilization of different carbon sources by microbial communities in two limestone aquifer assemblages. Hydrol Earth Syst Sci 21:4283–4300 . https://doi.org/10.5194/hess-21-4283-2017

Ohno T (2002) Fluorescence inner-filtering correction for determining the humification index of dissolved organic matter. Environ Sci Technol 36:742–746 . https://doi.org/10.1021/es0155276

O’Leary MH (1988) Carbon Isotopes in Photosynthesis. BioScience 38:328–336 . https://doi.org/10.2307/1310735

Parlanti E, Wörz K, Geoffroy L, Lamotte M (2000) Dissolved organic matter fluorescence spectroscopy as a tool to estimate biological activity in a coastal zone submitted to anthropogenic inputs. Org Geochem 31:1765–1781 . https://doi.org/10.1016/S0146-6380(00)00124-8

Pavelescu G, Ghervase L, Ioja C, Dontu S, Spiridon R (2013) Spectral fingerprints of groundwater organic matter in rural areas. Romanian Rep Phys 65:1105–1113

Petrella E, Capuano P, Celico F (2007) Unusual behaviour of epikarst in the Acqua dei Faggi carbonate aquifer (Southern Italy). Terra Nova 19:82–88 . https://doi.org/10.1111/j.1365-3121.2006.00720.x

Peyraube N, Lastennet R, Denis A, Malaurent P (2013) Estimation of epikarst air PCO2 13 13 using measurements of water δ CTDIC, cave air PCO2 and δ CCO2. Geochim Cosmochim Acta 118:1–17 . https://doi.org/10.1016/j.gca.2013.03.046

Pool DR, Blasch KW, Callegary JB, Leake SA, Graser LF (2011) Regional groundwater- flow model of the Redwall-Muav, Coconino, and alluvial basin aquifer systems of northern and central Arizona. US Geol Surv Sci Investig Rep 2010–5180:1–101 . https://doi.org/10.3133/sir20105180

Raymond PA, Bauer JE (2001) Use of 14C and 13C natural abundances for evaluating riverine, estuarine, and coastal DOC and POC sources and cycling: a review and synthesis. Org Geochem 32:469–485 . https://doi.org/10.1016/S0146- 6380(00)00190-X

Schiff SL, Aravena R, Trumbore SE, Hinton MJ, Elgood R, Dillon PJ (1997) Export of DOC from forested catchments on the Precambrian Shield of Central Ontario: Clues from 13C and 14C. Biogeochemistry 36:43–65 . https://doi.org/10.1023/A:1005744131385

Seager R, Ting M, Held I, Kushnir Y, Lu J, Vecchi G, Huang H-P, Harnik N, Leetmaa A, Lau N-C, Li C, Velez J, Naik N (2007) Model projections of an imminent transition to a more arid climate in Southwestern North America. Science 316:1181–1184 . https://doi.org/10.1126/science.1139601

Shin WJ, Chung GS, Lee D, Lee KS (2011) Dissolved inorganic carbon export from carbonate and silicate catchments estimated from carbonate chemistry and 13 δ CDIC. Hydrol Earth Syst Sci Discuss 8:1799–1825 . https://doi.org/10.5194/hessd-8-1799-2011

Siebert S, Burke J, Faures JM, Frenken K, Hoogeveen J, Döll P, Portmann FT (2010) Groundwater use for irrigation – a global inventory. Hydrol Earth Syst Sci 14:1863–1880 . https://doi.org/10.5194/hess-14-1863-2010

Simon KS, Pipan T, Ohno T, Culver DC (2010) Spatial and temporal patterns in abundance and character of dissolved organic matter in two karst aquifers. Fundam Appl Limnol Arch Für Hydrobiol 177:81–92 . https://doi.org/10.1127/1863-9135/2010/0177-0081

Solder JE, Beisner KR (2020) Critical evaluation of stable isotope mixing end-members for estimating groundwater recharge sources: case study from the South Rim of the Grand Canyon, Arizona, USA. Hydrogeol J 28:1575–1591 . https://doi.org/10.1007/s10040-020-02194-y

Solder JE, Beisner KR, Anderson J, Bills DJ (2020) Rethinking groundwater flow on the South Rim of the Grand Canyon, USA: characterizing recharge sources and flow paths with environmental tracers. Hydrogeol J 28:1593–1613 . https://doi.org/10.1007/s10040-020-02193-z

Sorauf JE, Billingsley GH (1991) Members of the Toroweap and Kaibab Formations, Lower Permian, Northern Arizona and Southwestern Utah. Mt Geol 28:9–24

Sprenger M, Leistert H, Gimbel K, Weiler M (2016) Illuminating hydrological processes at the soil-vegetation-atmosphere interface with water stable isotopes. Rev Geophys 54:674–704 . https://doi.org/10.1002/2015RG000515

Springer AE, Boldt EM, Junghans KM (2017) Local vs. regional groundwater flow delineation from stable isotopes at Western North America springs. Groundwater 55:100–109 . https://doi.org/10.1111/gwat.12442

Stedmon CA, Markager S, Bro R (2003) Tracing dissolved organic matter in aquatic environments using a new approach to fluorescence spectroscopy. Mar Chem 82:239–254 . https://doi.org/10.1016/S0304-4203(03)00072-0

Tanaka K, Kuma K, Hamasaki K, Yamashita Y (2014) Accumulation of humic-like fluorescent dissolved organic matter in the Japan Sea. Sci Rep 4:5292 . https://doi.org/10.1038/srep05292

Tobin BW, Springer AE, Kreamer DK, Schenk E (2018) Review: The distribution, flow, and quality of Grand Canyon Springs, Arizona (USA). Hydrogeol J 26:721–732 . https://doi.org/10.1007/s10040-017-1688-8

Voss BM, Wickland KP, Aiken GR, Striegl RG (2017) Biological and land use controls on the isotopic composition of aquatic carbon in the Upper Mississippi River Basin: Upper Mississippi Carbon Isotopes. Glob Biogeochem Cycles 31:1271– 1288 . https://doi.org/10.1002/2017GB005699

Wassenaar L, Aravena R, Fritz P, Barker J (1990) Isotopic composition (13C, 14C, 2H) and geochemistry of aquatic humic substances from groundwater. Org Geochem 15:383–396 . https://doi.org/10.1016/0146-6380(90)90165-V

Wenrich KJ, Sutphin HB (1989) Lithotectonic setting necessary for formation of a uranium rich, solution collapse breccia pipe province, Grand Canyon region, Arizona. US Geol Surv Open-File Rep 4:1–33

Wilson J, Munizzi J, Erhardt AM (2020) Preservation methods for the isotopic composition of dissolved carbon species in non-ideal conditions. Rapid Commun Mass Spectrom 34:e8903 . https://doi.org/10.1002/rcm.8903

Yamanaka M (2012) Contributions of C3/C4 organic materials and carbonate rock to dissolved inorganic carbon in a karst groundwater system on Miyakojima Island, southwestern Japan. J Hydrol 412–413:151–169 . https://doi.org/10.1016/j.jhydrol.2011.07.046

Yamashita Y, Scinto LJ, Maie N, Jaffé R (2010) Dissolved organic matter characteristics across a subtropical wetland’s landscape: application of optical properties in the assessment of environmental dynamics. Ecosystems 13:1006–1019 . https://doi.org/10.1007/s10021-010-9370-1

Yamashita Y, Tanoue E (2003) Chemical characterization of protein-like fluorophores in DOM in relation to aromatic amino acids. Mar Chem 82:255–271 . https://doi.org/10.1016/S0304-4203(03)00073-2

Yang L, Hur J (2014) Critical evaluation of spectroscopic indices for organic matter source tracing via end member mixing analysis based on two contrasting sources. Water Res 59:80–89 . https://doi.org/10.1016/j.watres.2014.04.018

Yang R, Liu Z, Zeng C, Zhao M (2012) Response of epikarst hydrochemical changes to soil CO2 and weather conditions at Chenqi, Puding, SW China. J Hydrol 468– 469:151–158 . https://doi.org/10.1016/j.jhydrol.2012.08.029

Zektser IS, Everett LG (2000) Groundwater and the environment: Applications for the global community. Lewis Publishers, Boca Raton, Florida

Zektser S, Loáiciga HA, Wolf JT (2005) Environmental impacts of groundwater overdraft: selected case studies in the southwestern United States. Environ Geol 47:396–404 . https://doi.org/10.1007/s00254-004-1164-3

Zhou Y, Guo H, Lu H, Mao R, Zheng H, Wang J (2015) Analytical methods and application of stable isotopes in dissolved organic carbon and inorganic carbon in groundwater: Analytical methods for carbon isotopes in DIC and DOC. Rapid Commun Mass Spectrom 29:1827–1835 . https://doi.org/10.1002/rcm.7280

VITA Jonathan Wesley Wilson

Education B.S. Geological Sciences, Summa Cum Laude (2018) University of Kentucky

Experience Graduate Teaching Assistant (2020) Department of Earth and Environmental Sciences University of Kentucky, 40506

Graduate Research Assistant (2019-2020) Kentucky Stable Isotope Geochemistry Laboratory (KSIGL) Department of Earth and Environmental Sciences University of Kentucky, 40506

Graduate Research Assistant (2019-2020) Kentucky Geological Survey University of Kentucky, 40506

Professional Publications Wilson J, Munizzi J, Erhardt AM (2020) Preservation methods for the isotopic composition of dissolved carbon species in non-ideal conditions. Rapid Commun Mass Spectrom 34:e8903 . https://doi.org/10.1002/rcm.8903