THE INFLUENCE OF CLIMATE AND LANDSCAPE ON HYDROLOGICAL PROCESSES,

THE INFLUENCE OF CLIMATE AND LANDSCAPE ON HYDROLOGICAL PROCESSES,

1

THE INFLUENCE OF CLIMATE AND LANDSCAPE ON HYDROLOGICAL PROCESSES,

VEGETATION DYNAMICS, BIOGEOCHEMISTRY AND THE TRANSFER OF

EFFECTIVE ENERGY AND MASS TO THE CRITICAL ZONE by

Xavier Zapata-Ríos

__________________________

A Dissertation Submitted to the Faculty of the

DEPARTMENT OF HYDROLOGY AND WATER RESOURCES

In Partial Fulfillment of the Requirements

For the Degree of

DOCTOR OF PHILOSOPHY

WITH A MAJOR IN HYDROLOGY

In the Graduate College

THE UNIVERSITY OF ARIZONA

2015

THE UNIVERSITY OF ARIZONA

GRADUATE COLLEGE

As members of the Dissertation Committee, we certify that we have read the dissertation prepared by Xavier Zapata-Rios, titled the Influence of Climate and Landscape on Hydrological

Processes, Vegetation Dynamics, Biogeochemistry and the Transfer of Effective Energy and

Mass to the Critical Zone and recommend that it be accepted as fulfilling the dissertation requirement for the Degree of Doctor of Philosophy.

2

_____________________________________________________________

Jennifer McIntosh

Date: 03/13/15

_______________________________________________________________ Date: 03/13/15

Peter A. Troch

________________________________________________________________ Date: 03/13/15

Jon Chorover

________________________________________________________________ Date: 03/13/15

Paul D. Brooks

Final approval and acceptance of this dissertation is contingent upon the candidate’s submission of the final copies of the dissertation to the Graduate College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.

________________________________________________________________ Date: 03/13/15

Dissertation Director: Jennifer McIntosh

________________________________________________________________Date: 03/13/15

Dissertation Director: Peter A. Troch

3

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of the requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that an accurate acknowledgement of the source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interest of scholarship. In all other instances, however permission must be obtained from the author.

SIGNED: Xavier Zapata-Ríos

4

DEDICATION

I want to dedicate this work to people who have provided me with great learning opportunities at some point in my educational path. This work is dedicated to great mentors and friends including

Crnl. Jorge Salvador y Chiriboga, Dr. Robert E. Rhoades, Dr. Rolf Ermshaus, and Dr. René M.

Price.

5

TABLE OF CONTENTS

ABSTRACT .................................................................................................................................................. 9

CHAPTER 1: INTRODUCTION ............................................................................................................... 11

1.1. The Critical Zone ........................................................................................................................ 11

1.2 Climate and landscape controls on hydrology and vegetation .......................................................... 12

1.3 Research Objectives .................................................................................................................... 14

1.4 Dissertation format ...................................................................................................................... 17

1.5 References ......................................................................................................................................... 19

CHAPTER 2: PRESENT STUDY .............................................................................................................. 23

2.1 Summary of Paper 1: Climatic and landscape controls on water transit times and silicate mineral weathering in the critical zone. ............................................................................................................... 23

2.1.1 Study objectives ......................................................................................................................... 23

2.1.2 Major findings ............................................................................................................................ 24

2.2 Summary of Paper 2: Influence of terrain aspect on water partitioning, vegetation structure, and vegetation greening in high elevation catchments in northern New Mexico .......................................... 27

2.2.1 Study objectives ......................................................................................................................... 27

2.2.2 Major findings ............................................................................................................................ 28

2.3 Summary of Paper 3: Influence of climate variability on water partitioning and effective energy and mass transfer (EEMT) in the critical zone .............................................................................................. 30

2.3.1 Study objectives ......................................................................................................................... 30

2.3.2 Major findings ............................................................................................................................ 30

2.4 References ......................................................................................................................................... 32

APPENDIX A: ............................................................................................................................................ 34

CLIMATIC AND LANDSCAPE CONTROLS ON WATER TRANSIT TIMES AND SILICATE

MINERAL WEATHERING IN THE CRITICAL ZONE .......................................................................... 34

ABSTRACT ............................................................................................................................................ 35

1.0 INTRODUCTION ...................................................................................................................... 37

2.0 METHODS ....................................................................................................................................... 39

2.1 STUDY AREA ............................................................................................................................. 39

2.2 FIELD INVESTIGATIONS ......................................................................................................... 41

6

3.0 RESULTS ......................................................................................................................................... 46

3.1 Landscape characteristics .............................................................................................................. 46

3.2 EEMT ............................................................................................................................................ 46

3.3 Water stable isotopes .................................................................................................................... 47

3.4 Water Chemistry ........................................................................................................................... 48

3.5 Water transit times (WTT) ............................................................................................................ 48

3.6 Bedrock composition and mass balance analysis .......................................................................... 49

4.0 DISCUSSION ................................................................................................................................... 50

5.0 SUMMARY ...................................................................................................................................... 54

6.0 ACKNOWLEDGEMENTS .............................................................................................................. 55

7.0 REFERENCES ................................................................................................................................ 56

8.0 FIGURES .......................................................................................................................................... 66

9.0 TABLES ........................................................................................................................................... 75

APPENDIX B: ............................................................................................................................................ 80

INFLUENCE OF TERRAIN ASPECT ON WATER PARTITIONING, VEGETATION STRUCTURE,

AND VEGETATION GREENING IN HIGH ELEVATION CATCHMENTS IN NORTHERN NEW

MEXICO ..................................................................................................................................................... 80

ABSTRACT ............................................................................................................................................ 81

1.0 INTRODUCTION ...................................................................................................................... 83

2.0 METHODOLOGY ........................................................................................................................... 85

2.1 Study area ...................................................................................................................................... 85

2.2 Catchment characteristics ............................................................................................................. 87

2.3 Vegetation characteristics ............................................................................................................. 87

2.4 Water partitioning and vegetation greening .................................................................................. 89

3.0 RESULTS ......................................................................................................................................... 94

3.1 Catchment characteristics ............................................................................................................. 94

3.2 Vegetation classification and structure ......................................................................................... 95

3.3 Water partitioning ......................................................................................................................... 95

3.4 Vegetation greening ...................................................................................................................... 98

3.5 Horton index ................................................................................................................................. 98

4.0 DISCUSSION ................................................................................................................................... 99

4.1 Hydrological partitioning ............................................................................................................ 100

4.2 Vegetation water use ................................................................................................................... 102

7

4.3 Vegetation structure and NDVI response ................................................................................... 103

4.4 Water partitioning and vegetation interactions ........................................................................... 105

4.5 Implications for the critical zone ................................................................................................ 106

4.6 Future studies .............................................................................................................................. 107

5.0 SUMMARY .................................................................................................................................... 107

6.0 ACKNOWLEDGEMENTS ............................................................................................................ 108

7.0 REFERENCES ............................................................................................................................... 109

8.0 FIGURES ........................................................................................................................................ 120

9.0 TABLES ......................................................................................................................................... 127

10.0 SUPLEMENTAL MATERIAL .................................................................................................... 132

APPENDIX C: .......................................................................................................................................... 138

INFLUENCE OF CLIMATE VARIABILITY ON WATER PARTITIONING AND EFFECTIVE

ENERGY AND MASS TRANSFER (EEMT) IN A SEMI-ARID CRITICAL ZONE ........................... 138

ABSTRACT .......................................................................................................................................... 139

1.0 INTRODUCTION .......................................................................................................................... 141

2.0 METHODS ..................................................................................................................................... 144

2.1 Study site ..................................................................................................................................... 144

2.2 Climatological stations ................................................................................................................ 145

2.3 Climate variability................................................................................................................. 145

2.4 EEMT estimation .................................................................................................................. 146

2.5 Water availability, water partitioning and climate controls on water availability ...................... 150

3.0 RESULTS ....................................................................................................................................... 150

3.1 Changes in climate ...................................................................................................................... 150

3.2 Water availability ........................................................................................................................ 152

3.3 EEMT .......................................................................................................................................... 153

EEMT emp

........................................................................................................................................... 153

3.4 Climate controls on discharge ..................................................................................................... 155

4.0 DISCUSSION ................................................................................................................................. 156

4.1 Climate variability....................................................................................................................... 156

4.2 Changes in discharge and evapotranspiration ............................................................................. 157

4.3 EEMT components ..................................................................................................................... 158

5.0 SUMMARY .................................................................................................................................... 162

8

6.0 ACKNOWLEDGEMENTS ............................................................................................................ 163

7.0 REFERENCES ............................................................................................................................... 164

8.0 FIGURES ........................................................................................................................................ 172

9.0 TABLES ......................................................................................................................................... 178

10.0 SUPPLEMENTARY INFORMATION ....................................................................................... 182

APPENDIX D: .......................................................................................................................................... 186

NOBLE GASES AND SF

6

CONCENTRATIONS FROM SPRINGS AROUND REDONDO PEAK,

NEW MEXICO ......................................................................................................................................... 186

1.0 NOBLE GASES .............................................................................................................................. 186

2.0 SF

6

................................................................................................................................................... 188

3.0 REFERENCES ............................................................................................................................... 192

9

ABSTRACT

The Critical Zone (CZ) is the surficial layer of the planet that sustains life on Earth and extends from the base of the weathered bedrock to the top of the vegetation canopy. Its structure influences water fluxes, biogeochemistry and vegetation. In this dissertation, I explore the relationships between climate, water fluxes, vegetation dynamics, biogeochemistry, and effective energy and mass transfer fluxes (EEMT) in a semi-arid critical zone. This research was carried out in the upper Jemez River Basin in northern New Mexico across gradients of climate and elevation. The main research objectives were to (i) quantify relations among inputs of mass and energy (EEMT), hydrological and biogeochemical processes within the CZ, (ii) determine water fluxes and vegetation dynamics in high elevation mountain catchments with different terrain aspect and solar radiation, and (iii) study temporal variability of climate and its influence on the

CZ water availability, forest productivity and energy and mass fluxes. The key findings of this study include (i) significant correlations between EEMT, water transit times (WTT) and mineral weathering products around Redondo Peak. Significant correlations were observed between dissolved weathering products (Na

+

and DIC) and maximum EEMT. Similarly,

3

H concentrations measured at the springs were significantly correlated with maximum EEMT; (ii) terrain aspect strongly controls energy, water distribution, and vegetation productivity in high elevation ecosystems in catchments draining different aspects of Redondo Peak. The predominantly north facing catchment, when compared to the other two eastern catchments, receives less solar radiation, exhibits less forest cover and smaller biomass, has more surface runoff and smaller vegetation water consumption. Furthermore, the north facing catchment showed smaller NDVI values and shorter growing season length as a consequence of energy limitation, and (iii) from 1984 to 2012 a decreasing trend in water availability, increased

vegetation water use, a reduction in both forest productivity and EEMT was observed at the upper Jemez River Basin. These changes point towards a hotter, drier and less productive ecosystem which may alter critical zone processes in high elevation semi-arid systems.

10

11

CHAPTER 1: INTRODUCTION

1.1. The Critical Zone

The Critical Zone (CZ) is the uppermost layer of the planet that supports life on earth and extends from the base of the weathered bedrock to the top of the vegetation canopy. Within the

CZ, energy and water fluxes are essential elements that trigger coupled chemical, physical, biological and geological processes [Brantley et al., 2007]. The CZ can be conceptualized as an open system non-linear reactor that is constantly evolving with respect to energy and water fluxes [Figure 1; Chorover et al., 2011; Rasmussen et al., 2011]. Energy and water create an internal structure organization within this reactor, driving processes such as soil organic stabilization, pedon horizonation, flow path formation, and mineral weathering among others.

Dissipative products are generated as a result of these internal processes and leave the reactor as physical and chemical denudation, sediments, water fluxes, solutes, gases and latent heat, etc.

[Chorover, et al., 2011].

Understanding the formation, evolution and functioning of the CZ has been identified as a research priority for the earth science community in order to predict the CZ response to ongoing changes in climate and land cover [National Research Council, 2001]. Gradients of climate and energy have been considered ideal places to study the variability of water and carbon fluxes and their influence in CZ development e.g. soil formation, biogeochemistry, and ecohydrology [Chorover et al., 2011]. Moreover, observations of processes along spatial gradients can provide useful information and can be used as proxies for variability on time

[Troch et al., 2008]. This space for time approach can provide estimations of landscape behavior under future climate regimes [Wagener et al., 2010; Singh et al., 2011].

12

1.2 Climate and landscape controls on hydrology and vegetation

The long-term coevolution between climate, topography, soils and vegetation has been shown to leave a detectable organization in biophysical properties within the landscape that in turn influences the structure and function of hydrological systems [Sivapalan, 2005; Troch et al., 2008;

Hopp et al., 2009; Ehret et al., 2014; Eagleson, 1986]. Despite the significant progress that has been made in the last few decades to understand hydrological processes at small spatial and short temporal scales, this knowledge has been very difficult to upscale to larger spatial and longer temporal scales due to spatial heterogeneities of the landscape and complex hydrologic, biogeochemical, and ecological feedbacks [Sivapalan, 2005; McDonnell et al., 2007; Troch et al.,

2008; Wagener et al., 2010]. In addition, natural and human forces are altering the hydrological cycle in unprecedented ways adding complexity for hydrologist to understand and predict these changes [Eagleson, 1986; Troch et al., 2008; Wagener et al., 2010; Ehret et al., 2014]. The ability to provide solutions for future water management and decision making across wide range of space and time scales depends on reliable estimations of water distribution, fluxes, water storage and quality [Sivapalan, 2005]. Understanding how climate, topography, soils and vegetation interact and have coevolved might provide key information on how they will interact in the future and may result in a new hydrological theory that allows for better hydrological predictions under climate and land cover changes [Hopp et al., 2009; Sivapalan, 2005; McDonnell et al., 2007; Troch et al.,

2008]. All the drivers of landscape evolution take place at the surface of our planet within the critical zone (CZ).

The National Science Foundation (NSF) has funded 13 Critical Zone Observatories

(CZO) across the continental United States along different gradients in climate and energy. One

13 of these observatories, the Santa Catalina Mountains and Jemez River Basin (JRB-CZO), has been established within the Valles Caldera National Preserve (VCNP) in the upper Jemez River

Basin in northern New Mexico. One of the JRB-CZO sites is a volcanic resurgent dome known as Redondo Peak (3430 m) which is located at the center of the VCNP and it is the second highest elevation peak in the Jemez Mountains. Relief and parent material are similar at this site, however differences in terrain aspect control wind exposure, radiation, micro-climate, vegetation

[Lyon et al., 2008] and the mass and energy transferred to the CZ [Chorover et al, 2011]. The

JRB-CZO sites provide unique research conditions to study microclimate, water partitioning, vegetation water use, water storage, streamflow response and water residence time along an energy gradient during a period characterized by years with extremes in precipitation.

High elevation semi-arid regions are water limited environments where the timing and amount of water availability is a fundamental factor that controls processes such as weathering, organic matter decomposition, soil respiration, nutrient uptake, and biomass production among others [Bales et al., 2006]. Understanding the hydrological processes in high elevation semi-arid regions has been limited by a lack of integrated measurements of governing fluxes and states, variability of landscape properties such as soil and vegetation, and small-scale hydroclimatic variability related to elevation, aspect, and vegetation structure that affect the magnitude and partitioning of water and energy fluxes [Musselman et al., 2008; Veatch et al., 2009; Rinehart et

al., 2008; Molotch et al., 2009; Bales et al., 2006; Gustafson et al., 2010]. This research is particularly pertinent to the southwestern US because the Jemez River Basin in northern New

Mexico is part of the head waters that feed the Rio Grande, the third largest river in the country, which sustains regional water supplies for millions of people [Bales et al., 2006]. The hydrology in this semi-arid mountainous region is snow dominated and spring snowmelt provides

ecosystems with most of their annual influx of water [ Bales et al., 2006]. Recent research has

14 shown a climate-driven trend toward earlier snowmelts and reduced snowpacks in this region

[Stewart et al., 2005; Mote et al., 2005]. There are high confidence predictions that snowpacks will continue to decline in northern New Mexico through the year 2100 and projections of snowpack accumulation for mid-century (2041-2070) show a marked reduction for snow water equivalent (SWE) of about 40% [Cayan et al., 2013]. The possible alterations in hydrologic pathways such as evaporation, runoff, infiltration, storage, water residence time, vegetation response due to these reductions remain poorly understood [Bales et al., 2006].

1.3 Research Objectives

The dissertation examines empirical observations in the critical zone across different spatial and time scales to find fundamental relationships between climate, water fluxes, vegetation dynamics, biogeochemistry, and mass and energy fluxes. Figure 1 illustrates a conceptual model of the CZ, highlighting input/output fluxes and processes that store and create an internal structural organization. In this dissertation, variables were selected as indicators to study these input/output fluxes and processes. The indicator variables selected include water residence time,

Horton index, base cations and Na

+

concentrations, mass of mineral dissolution, NDVI, net primary productivity and dissolved inorganic carbon.

15

Figure 1. Conceptual model of the critical zone (CZ). The CZ extends from the base of the bedrock to the top of the canopy. It receives fluxes of radiation, carbon and water that drive internal processes that creates an internal structural organization. Dissipative products leave the

CZ as physical, chemical denudation, baseflow, solutes, DIC/DOC fluxes. Indicators are variables that were selected specifically in this dissertation to understand input and output fluxes and energy storage.

Water, energy, carbon mass balance equations and mineral weathering reactions illustrated in figure 1 are the following:

1. Water mass balance

𝐿

πœ•πœƒ

πœ•π‘‘

= 𝑃 − 𝐸 − 𝑆 − π‘ˆ [L]

16

π‘Š = 𝐸 + π‘ˆ [L]

P= precipitation, E= evapotranspiration, S= surface run-off, U=baseflow, W=wetting

2. Energy balance

𝐢 𝑝

π‘š

πœ•π‘‡

πœ•π‘‘

= 𝑅 𝑛

− πœ†πΈ − 𝐻 − 𝐺 − π΅π‘–π‘œ [ML

2

T

-2

]

T= temperature, cp = heat capacity, m =mass, Rn= net radiation, λE= latent heat, H= sensible heat, G=subsurface heat flux, Bio= energy uptake by vegetation during photosynthesis.

3. Carbon mass balance

πœ•πΆ

πœ•π‘‘

= 𝐢𝑂

2

− 𝑅𝑒𝑠𝑝 − 𝐷𝐼𝐢 − 𝐷𝑂𝐢 − 𝑃𝑂𝐢 [ML

-2

T

-1

]

Resp=respitation, DIC= dissolved inorganic carbon, DOC= dissolved organic carbon, POC= particulate organic carbon.

4. Mineral weathering

π‘ƒπ‘Ÿπ‘–π‘šπ‘Žπ‘Ÿπ‘¦ π‘€π‘–π‘›π‘’π‘Ÿπ‘Žπ‘™π‘ 

(𝑠)

+ 𝐻

2

𝑂

(𝑙)

+ 𝐻

+

(π‘Žπ‘ž)

⇔π‘†π‘’π‘π‘œπ‘›π‘‘π‘Žπ‘Ÿπ‘¦ π‘€π‘–π‘›π‘’π‘Ÿπ‘Žπ‘™π‘ 

(𝑠)

+ π‘π‘Žπ‘‘π‘–π‘œπ‘›π‘ 

(π‘Žπ‘ž)

+ 𝑆𝑖𝑙𝑖𝑐𝑖𝑐 𝐴𝑐𝑖𝑑

(π‘Žπ‘ž)

The main research questions tackled during this dissertation are:

(1) What are the relations between inputs of mass and energy, hydrological and biogeochemical processes within the CZ? I hypothesized that energy and mass flowing into the CZ control its structure thereby influencing hydrologic and biogeochemical processes. In the first paper of this dissertation, I investigated whether the quantification of energy and mass entering the CZ can predict CZ processes, specifically water residence times and mineral weathering. I posited that around Redondo Peak longer residence times and larger mineral weathering fluxes would be observed in waters

17 draining its north facing slopes where higher inputs of mass and energy entering the CZ have been quantified.

(2) What are the water fluxes and vegetation dynamics in high elevation mountain catchments with different terrain aspect and solar radiation? The second paper in this dissertation analyzes the interactions and feedbacks between water availability and vegetation at the catchment scale. Three high elevation streams with different terrain aspects draining Redondo Peak were studied in this paper. I hypothesized that the differences in aspect and solar radiation among catchments would influence the short term partitioning of water and vegetation dynamics around Redondo Peak.

(3) What are the impacts of climate temporal variability on the CZ water availability, forest productivity and energy and mass fluxes? The study presented in the third paper took place in the upper Jemez River Basin with an area of 1200 km

2

. I hypothesized that changes in climate are affecting the CZ by decreasing water availability, altering forest productivity, and decreasing influxes of mass and energy.

1.4 Dissertation format

The dissertation has been organized in three studies focusing on the influence of climate and landscape position on the CZ water fluxes, vegetation dynamics, water transit times, mineral weathering and mass and energy transfer at different spatial scales. Following this first introductory chapter, chapter 2 presents a summary of three papers prepared to be submitted for peer-review publication. These papers are included individually in the appendices.

APPENDIX A. Climatic and landscape controls on water transit times and silicate mineral weathering in the critical zone.

18

APPENDIX B. Influence of terrain aspect on water partitioning, vegetation structure, and vegetation greening in high elevation catchments in northern New Mexico.

APPENDIX C. Influence of climate variability water partitioning and effective energy and mass transfer (EEMT) in a semi-arid critical zone.

19

1.5 References

Bales, R. C., N. P. Molotch, T. H. Painter, M. D. Dettinger, R. Rice, and J. Dozier (2006),

Mountain hydrology of the western United States, Water Resour. Res., 42, W08432.

Brantley, S. L., M. B. Goldhaber, and K. V. Ragnarsdottir (2007), Crossing disciplines and scales to understand the Critical Zone, Elements, 3, 307-314.

Cayan, D., M. Tyree, K. E. Kunkel, C. Castro, A. Gershunov, J. Barsugli, A. J. Ray, J. Overpeck,

M. Anderson, J. Russell, B. Rajagopalan, I. Rangwala, and P. Duffy. 2013. “Future Climate:

Projected Average.” In Assessment of Climate Change in the Southwest United States: A Report

Prepared for the National Climate Assessment, edited by G. Garfin, A. Jardine, R. Merideth, M.

Black, and S. LeRoy, 101–125. Washington, DC: Island Press.

Chorover, J. et al. (2011), How Water, Carbon, and Energy Drive Critical Zone Evolution: The

Jemez-Santa Catalina Critical Zone Observatory, Vadose Zone Journal, 10, 884-899.

Eagleson, P. (1986), The Emergence of Global-Scale Hydrology, Water Resour. Res., 22, S6-

S14.

Ehret, U. et al. (2014), Advancing catchment hydrology to deal with predictions under change,

Hydrology and Earth System Sciences, 18, 649-671.

Gustafson, J. R., P. D. Brooks, N. P. Molotch, and W. C. Veatch (2010), Estimating snow sublimation using natural chemical and isotopic tracers across a gradient of solar radiation,

Water Resour. Res., 46, W12511.

20

Hopp, L., C. Harman, S. L. E. Desilets, C. B. Graham, J. J. McDonnell, and P. A. Troch (2009),

Hillslope hydrology under glass: confronting fundamental questions of soil-water-biota coevolution at Biosphere 2, Hydrology and Earth System Sciences, 13, 2105-2118.

Lyon, S. W., P. A. Troch, P. D. Broxton, N. P. Molotch, and P. D. Brooks (2008), Monitoring the timing of snowmelt and the initiation of streamflow using a distributed network of temperature/light sensors, Ecohydrology, 1, 215-224.

McDonnell, J. J. et al. (2007), Moving beyond heterogeneity and process complexity: A new vision for watershed hydrology, Water Resour. Res., 43, W07301.

Molotch, N. P., P. D. Brooks, S. P. Burns, M. Litvak, R. K. Monson, J. R. McConnell, and K.

Musselman (2009), Ecohydrological controls on snowmelt partitioning in mixed-conifer subalpine forests, Ecohydrology, 2, 129-142.

Mote, P., A. Hamlet, M. Clark, and D. Lettenmaier (2005), Declining mountain snowpack in western north America, Bull. Am. Meteorol. Soc., 86, 39-+.

Musselman, K. N., N. P. Molotch, and P. D. Brooks (2008), Effects of vegetation on snow accumulation and ablation in a mid-latitude sub-alpine forest, Hydrol. Process., 22, 2767-2776.

National Research Council (2001). Basic research opportunities in earth science. National

Academy Press. Washington, D.C.

Rasmussen, C., P. A. Troch, J. Chorover, P. Brooks, J. Pelletier, and T. E. Huxman (2011), An

21 open system framework for integrating critical zone structure and function, Biogeochemistry,

102, 15-29.

Rinehart, A. J., E. R. Vivoni, and P. D. Brooks (2008), Effects of vegetation, albedo, and solar radiation sheltering on the distribution of snow in the Valles Caldera, New Mexico,

Ecohydrology, 1, 253-270.

Singh, R., T. Wagener, K. van Werkhoven, M. E. Mann, and R. Crane (2011), A trading-spacefor-time approach to probabilistic continuous streamflow predictions in a changing climate - accounting for changing watershed behavior, Hydrology and Earth System Sciences, 15, 3591-

3603.

Sivapalan, M (2005), Pattern, process and function: elements of a new unified hydrologic theory at the catchment scale, in Encyclopedia of Hydrologic Sciences, vol. 1, part 1, edited by M.G.

Anderson, chap.13, pp.193-219, John Wiley, Hoboken, N.J.

Stewart, I., D. Cayan, and M. Dettinger (2005), Changes toward earlier streamflow timing across western North America, J. Clim., 18, 1136-1155.

Troch P.A., G.A. Carrillo, I. Heidbuechel, D. Rajagopal, M. Seitanek, T.H.M. Volkmann, M.

Yaeger (2008) Dealing with landscape heterogeneity in watershed hydrology: a review of recent progress toward new hydrological theory. Geography Compass 2: 10.1111/j.1749-

8198.2008.0086.x

22

Veatch, W., P. D. Brooks, J. R. Gustafson, and N. P. Molotch (2009), 'Quantifying the effects of forest canopy cover on net snow accumulation at a continental, mid-latitude site', Ecohydrology,

2, 115-128.

Wagener, T., M. Sivapalan, P. A. Troch, B. L. McGlynn, C. J. Harman, H. V. Gupta, P. Kumar,

P. S. C. Rao, N. B. Basu, and J. S. Wilson (2010), The future of hydrology: An evolving science for a changing world, Water Resour. Res., 46, W05301.

23

CHAPTER 2: PRESENT STUDY

The interactions between climate and topography and their influence on hydrologic processes, vegetation dynamics, biogeochemistry and the effective mass and energy to the CZ are explored in three studies appended to this dissertation. The following pages summarize the major findings and implications of these studies.

2.1 Summary of Paper 1: Climatic and landscape controls on water transit times and silicate mineral weathering in the critical zone.

2.1.1 Study objectives

The CZ can be conceptualized as an open system reactor that is continually transforming energy and water fluxes [Chorover et al., 2011]. In the present study, we test the predictive power of a postulated controlling factor on Water Transit Times (WTT) and mineral weathering derived from climatic data termed effective energy and mass transfer (EEMT) to the CZ

[Rasmussen et al., 2011]. The objective of the present study is to test whether EEMT and/or landscape characteristics are dominant controlling variables on WTT times and the main silicate weathering mass balance reactions in a terrain characterized by a relatively uniform bedrock geology.

Our study site is located around Redondo Peak, a rhyolitic volcanic resurgent dome, in northern New Mexico. At Redondo Peak springs drain slopes along an energy gradient created by differences in terrain aspect. This investigation uses major solute concentrations, mass

24 balance weathering reactions, and the age tracer tritium and compares them to EEMT and landscape characteristics.

The selected springs drain different aspects of Redondo Peak at elevations between 2816 and 3170 m. Contributing areas of springs vary in size from 56 to 1038 (x10

3 m

2

) and their mean cosine angle covers a wide range of aspects from -0.5 south to 0.96 north. The mean slope of contributing areas range between 10 to 20 degrees, and the median water flow path length ranges from 360 to 600 meters with a median gradient between 0.12 and 0.32.

2.1.2 Major findings

The major findings of this study include:

(1) EEMT values for the Redondo Peak region vary between 22 and 59 MJ m-2 year-1.

Large EEMT values above 30 MJ.m-2.year-1 were found on Redondo Peak and low

EEMT values below 30 MJ m-2 year-1 were found along shaded areas by topography such as Redondo Creek in the southwestern part of Redondo.

(2) The mean aspect of a spring contributing area positively correlates with mean EEMT

(R

2

=0.31; p>0.1), maximum EEMT (R

2

=0.90; p<0.0001), and the range of EEMT

(R

2

=0.82; p=0.0007).

(3) Based on the maximum EEMT versus aspect relationship, north facing slopes receive up to 25% more EEMT than south facing slopes.

(4) Spring waters around Redondo Peak are dominated by Ca

+

, Na

+

, Si and HCO

3

-

and have an isotopic signature that indicates they are predominantly derived from infiltration of snowmelt.

25

(5) Dissolved inorganic carbon (DIC) concentrations strongly correlate with the products of silicate weathering, including the sum of all base cations (R

2

=0.92; p<0.0001) and Na

+ concentrations (R

2

=0.86; p=0.001).

(6) Significant correlations were found between all base cations and Na

+

concentrations versus aspect, maximum and range of EEMT. Na

+

is a cation that is not generally cycle to a great degree biologically and it shows the strongest correlations with aspect and

EEMT.

(7) Water with higher concentrations of solutes is found along north facing slopes (R

2

=0.56;

p<0.05), which are characterized in this study by higher EEMT values (R

2

=0.51;

p<0.05).

(8) Maximum flow path length and contributing area are the only two landscape characteristics that show statistically significant correlations with both base cations and

Na

+

concentrations. Larger concentrations of base cations are observed in springs with longer flow paths (R

2

=0.54; p<0.05) and larger contributing areas (R

2

=0.54; p<0.05)

(9) Mean EEMT and other landscape variables, such as mean elevation, mean slope, median flow path length (L), median flow path slope (G), and L/G are poor predictors of both base cations and Na

+

concentrations.

(10) Tritium concentrations measured at the springs ranged from 5.1 to 7.9 TU and water apparent age in the springs ranged from 0.2 to 8.1 years.

(11) Mineralogical analyses of Redondo Peak bedrock samples indicate predominance of primary phases sanidine, oligoclase, anorthoclase, and quartz, accompanied by minor minerals, including apatite, hematite, zircon, titanite, ilmenite, faujasite.

26

(12) Spring waters are undersaturated with respect to silica, halite, calcite, gypsum, sanidine and albite, but saturated to supersaturated with respect to secondary minerals gibbsite, goethite, hematite and kaolinite

(13) Longer WTT based on tritium analysis of springs draining north-facing terrains.

Mineral dissolution fluxes increase with WTT, likely due to enhanced water-rock reaction, and chemical weathering consumes more atmospheric CO

2 along north facing slopes.

Results from this study demonstrate the close interrelationship between landscape, hydrological and biogeochemical processes. This study provides evidence that fluxes of energy and mass, quantified as EEMT, at the catchment scale can effectively predict short time-scale

(months to years) processes within the CZ structure like WTT and silicate mineral weathering.

These results also suggest that basic climatic data embodied in the EEMT term can be used to scale hydrological and hydrochemical responses in other sites.

2.2 Summary of Paper 2: Influence of terrain aspect on water partitioning, vegetation structure,

27 and vegetation greening in high elevation catchments in northern New Mexico

2.2.1 Study objectives

Vegetation and water availability are primary controls on ecosystems structure (Brooks &

Vivoni, 2008), therefore, it is important to study and understand their role within the critical zone. Understanding the feedbacks between water and vegetation in mountainous semiarid catchments can help improve climate change predictions and associated hydrologic and ecologic shifts (Newman et al., 2006; Molotch et al., 2009; Vivoni et al., 2012). Furthermore, field studies along landscape gradients may provide insight on the relationship among topography, vegetation, and water (Kelly and Goulden, 2008; Newman et al., 2006; Chorover et al., 2011).

The objective of this study is to investigate from 2000 to 2012 how terrain aspect influences vegetation structure, the dynamics of hydrological partitioning, and vegetation greening in three high elevation semi-arid catchments using direct and remote sensing observations. A uniform geology and relief around Redondo Peak, located in the center of the

VCNP, make this site an ideal location to empirically study how topographically controlled microclimate (Lyon et al., 2008; Chorover et al., 2011) influences the integrated vegetation and hydrological response. This investigation focuses on three catchments with perennial streams: La

Jara (LJ), History Grove (HG) and Upper Jaramillo (UJ) (Figure 1c). The three catchments together cover approximately an area of 10 km

2

.

28

Vegetation structure was quantified from 1 m LiDAR data while vegetation greening was quantified using remotely sensed NDVI. Hydrological partitioning at the catchment scale was estimated with a metric of catchment-scale water fluxes and vegetation water use (Horton index;

HI).

2.2.2 Major findings

The major findings of this study are:

(1) The predominantly north facing catchment (UJ), when compared to the other two eastern catchments (LJ and HG), receives less solar radiation, and exhibits less forest cover and smaller biomass, has more surface runoff (~15% of P) as a consequence of a smaller vaporization (85% of P) and smaller vegetation water consumption (HI=0.85).

(2) The north facing catchment (UJ) showed smaller peak NDVI values (5.98) and shorter growing season length (76.3 days) as a consequence of energy limitation.

(3) In contrast, the two eastern catchments (LJ and HG) receive larger solar radiation, have more biomass and forest cover (>76%), and smaller surface runoff (<10% P), higher vaporization (>90%P) and vegetation water consumption (HI=0.95).

(4) The eastern catchments (Lj and HG) had larger vegetation greening (6.28-6.58) and a longer growing season (113-156 days).

(5) Snowpack conditions, such as maximum SWE and duration of the snow on the ground, explain over 95% of water partitioning (HI) that in turn influenced annual vegetation greening (R

2

=0.48 - 0.67; p<0.05).

29

This catchment scale study in perennial streams indicates that terrain aspect at a similar altitude (2700 – 3435m) strongly controls energy, water distribution, and vegetation productivity in high elevation ecosystems. Terrain aspect differences in water partitioning fluxes among catchments during wet years can be larger than the water partitioning fluxes variability induced only by climate variability in a single catchment. This study demonstrates that water and energy limits forest productivity and terrain aspect is a landscape characteristic that can further exacerbate the availability of these resources. Aspect controls on water availability and reduced carbon compounds resulting from primary production influences the inputs of EEMT and processes occurring within the CZ.

2.3 Summary of Paper 3: Influence of climate variability on water partitioning and effective energy and mass transfer (EEMT) in the critical zone

30

2.3.1 Study objectives

The objective of this study was to evaluate climate variability and its influence on the temporal changes of water partitioning and EEMT at the catchment scale in a semi–arid CZ over the last few decades. This investigation took place in the upper part of the Jemez River Basin in northern New Mexico, a basin dominated by a wide forest cover and limited human infrastructure, where the Santa Catalina-Jemez River Critical Zone Observatory has established a research site to study CZ processes [Chorover et al., 2011]. Micro-climate variability was studied based on daily records from two SNOTEL stations using records from 1984 through

2012. Water availability and EEMT were estimated during the same time period based on precipitation and temperature from the precipitation-elevation regressions on independent slopes model (PRISM), empirical daily observations of catchment scale discharge, and satellite derived net primary productivity (MODIS).

2.3.2 Major findings

The major findings of this study include:

(1) Clear increasing trends in temperature and decreasing trends in precipitation and maximum SWE at the two SNOTEL stations indicated. Temperature changes include warmer winters (+1.0-1.3 °C/decade), and generally warmer year round temperatures

(+1.2-1.4 °C/decade). Precipitation changes include, a decreasing trend in

31 precipitation during the winter (-41.6-51.4 mm/decade), during the year (-69.8-73.2 mm/decade) and max SWE (-33.1-34.7 mm/decade).

(2) At the upper Jemez River Basin ,all the water partitioning components showed statistical significant decreasing trends including precipitation (-61.7mm/decade), discharge (-17.6 mm/decade) and vaporization (-45.7 mm/decade). Similarly, Q

50

an indicator of snowmelt timing is occurring -4.3 days/decade earlier.

(3) Basin scale precipitation (R

2

=0.56; p=0.003) and baseflow (R

2

=0.41; p=0.02) were the strongest controls on NPP variability indicating that forest productivity in the upper Jemez River Basin is water limited. An increasing trend in Horton index suggests that water limitation and vegetation water use are increasing in the basin.

(4) This study showed a positive correlation between water availability and EEMT. For every 10 mm of change in baseflow, EEMT varies proportionally in 0.6-0.7 MJ m

-

2 year

-1

.

(5) From 1984-2012 changes in climate, water availability, and NPP have influenced

EEMT in the upper Jemez River Basin. A decreasing trend in EEMT of 1.2 to 1.3 MJ m

-2

decade

-1 was calculated in this same time frame.

As the landscape moves towards a drier and hotter climate, changes in EEMT of this magnitude are likely to influence critical zone processes.

2.4 References

Brooks, P. D. and E. R. Vivoni (2008), Mountain ecohydrology: quantifying the role of vegetation in the water balance of montane catchments, Ecohydrology, 1, 187-192.

Chorover, J. et al. (2011), How Water, Carbon, and Energy Drive Critical Zone Evolution: The

Jemez-Santa Catalina Critical Zone Observatory, Vadose Zone Journal, 10, 884-899.

Kelly, A. E. and M. L. Goulden (2008), Rapid shifts in plant distribution with recent climate change, Proc. Natl. Acad. Sci. U. S. A., 105, 11823-11826.

32

Lyon, S. W., P. A. Troch, P. D. Broxton, N. P. Molotch, and P. D. Brooks (2008), Monitoring the timing of snowmelt and the initiation of streamflow using a distributed network of temperature/light sensors, Ecohydrology, 1, 215-224.

Molotch, N. P., P. D. Brooks, S. P. Burns, M. Litvak, R. K. Monson, J. R. McConnell, and K.

Musselman (2009), Ecohydrological controls on snowmelt partitioning in mixed-conifer subalpine forests, Ecohydrology, 2, 129-142.

Newman, B. D., B. P. Wilcox, S. R. Archer, D. D. Breshears, C. N. Dahm, C. J. Duffy, N. G.

McDowell, F. M. Phillips, B. R. Scanlon, and E. R. Vivoni (2006), Ecohydrology of waterlimited environments: A scientific vision, Water Resour. Res., 42, W06302.

Rasmussen, C., P. A. Troch, J. Chorover, P. Brooks, J. Pelletier, and T. E. Huxman (2011), An open system framework for integrating critical zone structure and function, Biogeochemistry,

102, 15-29.

Vivoni, E. R. (2012), Spatial patterns, processes and predictions in ecohydrology: integrating technologies to meet the challenge, Ecohydrology, 5, 235-241.

33

APPENDIX A:

CLIMATIC AND LANDSCAPE CONTROLS ON WATER TRANSIT TIMES AND

SILICATE MINERAL WEATHERING IN THE CRITICAL ZONE

Manuscript in review by the journal Water Resources Research

Authors:

Xavier Zapata-Rios

1*

, Jennifer McIntosh

1

, Laura Rademacher

2

, Peter A. Troch

1

, Paul D.

Brooks

3,1

, Craig Rasmussen

4

, Jon Chorover

4

34

1. Department of Hydrology and Water Resources, The University of Arizona, Tucson, Arizona,

USA

2. Earth and Environmental Sciences, University of the Pacific, Stockton, California, USA

3. Department of Geology and Geophysics, University of Utah, Salt Lake City, Utah, USA

4. Soil, Water and Environmental Science, The University of Arizona, Tucson, Arizona, USA

* Corresponding author:

1133 E James E Rogers Way

J W Harshbarger Bldg Rm 122, PO Box 210011

The University of Arizona

Tucson AZ 85721-0011

Email: [email protected]

35

ABSTRACT

The critical zone (CZ) can be conceptualized as an open system reactor that is continually transforming energy and water fluxes into an internal structural organization and dissipative products. In this study, we test a controlling factor on water transit times (WTT) and mineral weathering called Effective Energy and Mass Transfer (EEMT). We hypothesize that EEMT, quantified based on local climatic variables, can effectively predict WTT within – and mineral weathering products from – the CZ. This study compares the ability EEMT versus static landscape characteristics in predicting WTT, aqueous phase solutes, and silicate weathering products. Our study site is located around Redondo Peak, a rhyolitic volcanic resurgent dome, in northern New

Mexico. At Redondo Peak springs drain slopes along an energy gradient created by differences in terrain aspect. This investigation uses major solute concentrations, mass balance weathering reactions, and the age tracer tritium and compares them to EEMT and landscape characteristics.

Results indicate significant correlations between EEMT, WTT and mineral weathering products.

Significant correlations were observed between dissolved weathering products (Na

+

and DIC) and maximum EEMT. Similarly,

3

H concentrations measured at the springs were significantly correlated with maximum EEMT. In contrast, landscape characteristics such as contributing area of spring, slope gradient, elevation, and flow path length were not as effective predictive variables of WTT, solute concentrations, and mineral weathering products. These results highlight the interrelationship between landscape, hydrological, and biogeochemical processes and suggest that basic climatic data embodied in EEMT can be used to scale hydrological and hydrochemical responses in other sites.

KEYWORDS

Critical Zone, Effective Energy and Mass Transfer (EEMT), terrain aspect, water transit times, weathering of silicates.

36

37

1.0 INTRODUCTION

The Critical Zone (CZ) is the uppermost land surface layer of the planet that extends from the base of the weathered bedrock to the top of the vegetation canopy. Within the CZ, energy and water fluxes drive coupled chemical, physical, biological and geological processes that support life [Brantley et al., 2007]. Understanding the formation, evolution and functioning of the CZ is fundamental for predicting its response to ongoing changes in climate and land cover [National Research Council, 2001].

Mineral weathering is a fundamental process occurring within the CZ that influences soil development and nutrient release [White and Brandley, 1995], as well as the buffering of acids derived from the atmosphere and biosphere [Velbel and Price, 2007]. Furthermore, weathering of silicate minerals impacts the global carbon cycle through the consumption of atmospheric CO

2

[Berner et al., 1983; Berner, 1995; Maher and Chamberlain, 2014], and the aqueous phase products of chemical weathering are determinants of the chemical composition of natural waters

[Bricker et al., 1986; Drever, 1988]. During weathering, atmospheric CO

2

and silicate minerals are converted to alkalinity and dissolved cations [Maher and Chamberlain, 2014]. Many silicate weathering studies demonstrate that hydrologic factors, and particularly water transit time

(WTT), play an important regulatory role on weathering of silicates [Velbel, 1993; White et al.,

2001; Gabet et al., 2006; Maher 2010, 2011].

Water transit time (WTT) refers to the elapsed time between when water enters and exits the hydrologic system [McGuire and McDonnell, 2006]. WTT is a good indicator of the hydrologic response of a system over a period of time and provides information about flow path heterogeneity, subsurface storage capacity, water input-output fluxes, mineral weathering, and

38 solute transport [McGuire and McDonnell, 2006]. Over the last decade, research efforts have focused on understanding the factors that control WTT at diverse spatial scales and in different geographic regions [McGlynn et al., 2003; McGuire et al., 2005; Rodgers et al., 2005; Soulsby

et al., 2006; Stewart et al., 2010; Mueller et al., 2012; many others]. In these investigations, the role of landscape structure, topography, soils, geology and climate on WTT have been tested at different sites with results that have been difficult to transfer from one specific research site to others [Vitvar et al., 2005; Tetzlaff et al.,2009]

In the present study, we test the predictive power of a postulated controlling factor on

WTT and mineral weathering derived from climatic data termed effective energy and mass transfer (EEMT) to the CZ [Rasmussen et al., 2011]. The CZ can be conceptualized as an open system reactor that is continually transforming energy and water fluxes [Figure 1; Chorover et

al., 2011]. Energy and water fluxes generate internal structural organization within this reactor, driving processes such as soil organic carbon stabilization, pedon horizonation, flow path formation, and mineral weathering, among others. Dissipative products resulting from CZ internal processes leave the reactor via physical and chemical denudation, sediments transport, water solutes, gas fluxes, and latent heat [Chorover et al., 2011; Rasmussen et al., 2011].

Energy based pedogenic models attempt to quantify energy fluxes to the soil system that drive mineral weathering and CZ development. Some initial work and semi-quantitative approaches describing soil formation factors (climate, biota, relief, parent material, and time) were developed by Dokuchaev [1967], Jenny [1941], Runge [1973], and Smeck et al. [1983].

Later, these soil forming factors were quantified in terms of energy [Volobuev, 1964; Phillips,

2009; Rasmussen et al., 2005; Rasmussen and Tabor, 2007]. The EEMT term in particular is

focused on the energy and mass transfer to the subsurface in the form of heat energy associated

39 with effective precipitation (water in excess from evapotranspiration) and chemical energy associated with reduced carbon compounds produced through primary production. Recent research along elevation gradients in the western US documented a significant correlation between EEMT and regolith depth, chemical depletion, and denudation rates and demonstrated weak and no correlations considering climatic variables alone such mean annual air temperature and mean annual precipitation [Rasmussen et al., 2005, 2011; Rasmussen and Tabor 2007]. The relationship between inputs of energy into the CZ at a mountain scale (~40 km

2

) and the influence on WTT and mineral weathering has not been explored previously.

Redondo Peak within the Jemez Mountains in northern New Mexico is a research site within the larger Catalina-Jemez Critical Zone Observatory (CZO) that was established to understand CZ processes [Chorover et al., 2011]. Around Redondo Peak, a volcanic rhyolitic resurgent dome, head water springs and streams discharge along different sides of the mountain.

The differences in terrain aspect around Redondo Peak create a natural gradient of energy inputs to the CZ. Previous research at this site has observed differences in solar radiation, water availability, dissolved organic carbon fluxes, and WTT across different terrain aspects of

Redondo Peak [Lyon et al., 2008; Broxton et al., 2009; Perdrial et al., 2014]. The objective of the present study is to test whether EEMT and/or landscape characteristics are dominant controlling variables on WTT times and the main silicate weathering mass balance reactions in a terrain characterized by a relatively uniform bedrock geology.

2.0 METHODS

2.1 STUDY AREA

40

Redondo Peak lies within the Valles Caldera National Preserve (VCNP) in the Jemez

Mountains of northern New Mexico (35˚50’-36˚00’ N; 106˚24’-106˚37’ W). The VCNP is a 21 km wide caldera formed by the collapse of a magma chamber approximately 1.25 Ma before present [Wolff et al., 2011]. The rim of the caldera extends to elevations above 2800 masl and encloses a basin floor that ranges in elevation between 2500 and 2750 masl. Redondo Peak, with maximum elevation of 3435 masl, is a resurgent dome formed by magma flow through ring fracture faults and is located in the center of the caldera [Lyon et al., 2008]. Redondo Peak is characterized by a gradient in terrain aspect that influences wind exposure, radiation, snowmelt, sublimation, evapotranspiration and groundwater recharge [Lyon et al., 2008].

Several springs and first-order streams drain all sides of Redondo Peak feeding the Jemez

River basin, which is a tributary of the Rio Grande River [Ellis et al., 1993] (Figure 2). The geology of Redondo Peak is dominated by Pleistocene aged, densely to partially welded

Bandelier Tuff and older rhyolitic and andesitic rocks associated with older volcanic events

[Goff et al., 2006; Wolff et al., 2011]. Soils across the dome are generally characterized by welldrained Mollisols, Inceptisols and Alfisols that span coarse sandy loam to clay loam textures and contain an abundance of partially to highly decomposed organic matter in surface soil horizons

[http://websoilsurvey.nrcs.usda.gov/app/HomePage.htm]. Vegetation at higher elevations is dominated by Spruce-fir (Picea pungens), ponderosa pine (Pinus ponderosa), aspen (Populas

tremuloides), and gambel oak (Quercus gambelii); lower elevations and valley bottoms are dominated by forest meadows and grasslands [Muldavin et al., 2006; Coop and Givnish, 2007].

The semi-arid climate in this region is continental in nature and highly variable. The climate is characterized by a bimodal precipitation pattern, with ca. 50% of the total annual precipitation falling as snow and rain during the winter months and ca. 50% falling as rainfall during the

41 summer monsoon season [Sheppard et al., 2002]. According to 31 years of records (1981-2012) at the Quemazon SNOwpack TELemetry (SNOTEL) station (35°55’ N; 106°24’ W; 2896 masl), located 5 km east from Redondo Peak, the average annual precipitation is 711 mm, and the average summer and winter temperatures are 10.7˚C and -1.1 ˚C, respectively.

2.2 FIELD INVESTIGATIONS

2.2.1 Landscape analysis

The springs around Redondo Peak selected for this study are characterized by perennial discharge. The springs drain slopes with different aspects (orientations) at elevations above 2800 masl (Table 1). Springs with low discharge, draining the foothills or other domes in the VCNP were not considered in this analysis. Landscape characteristics of springs investigated in this study include aspect, contributing area, water flow length, and slope gradients. Landscape characteristics were defined using a 722 km2 airborne LiDAR flight coverage acquired July 2010 and processed by the National Center for Airborne Laser Mapping (NCALM). The LiDAR coverage has an average point density of 9.68 points per m2 and vertical and horizontal RMSE resolutions of 0.1 m and 1.0 m, respectively. For this study, a spring’s contributing area is an idealized region where energy and matter are transferred to the CZ. Within these boundaries it is assumed that water flow is routed and discharged into the springs and mineral weathering transformations occur. Contributing areas from springs and topographic analysis within each basin were defined using the catchment delineation procedures from TauDEM version 5.0

( http://hydrology.usu.edu/taudem/taudem5.0/index.html. Accessed May, 2014) and following procedures similar to those of McGuire et al. [2005]. An accumulated area grid was obtained applying a multidirectional flow algorithm to define a flow network. Then, flow path length and

42 gradients were computed along each flow line within a spring’s contributing area [McGuire et al.,

2005]. The mean aspect of contributing areas was calculated following the methodology indicated in Broxton et al. [2009], as an average of the x and y composite vectors for each aspect cell. The average aspect calculated in radians was expressed as the cosine of the angle such that north facing terrains have a value of 1 and south facing terrains have a value of -1.

2.2.2 Effective Energy and Mass Transfer (EEMT)

The EEMT model is based on the hypothesis that soil and the larger CZ tend to self-organize to optimize the transmission of energy flowing through the system [Rasmussen et al., 2005;

Rasmussen and Tabor, 2007]. This self-organization, entropy maximization, and use of energy of natural systems is described by principles of open system thermodynamics and traditional quantitative models of soil development [Rasmussen et al., 2005; Kleidon et al., 2012]. The

EEMT model quantifies external energy inputs to the CZ by integrating into one single variable the climatic and biotic forcings [Rasmussen et al., 2005; Rasmussen and Tabor, 2007]. For a more detailed description and derivation of EEMT see Rasmussen et al. [2005, 2011],

Rasmussen and Tabor [2007], and Chorover et al. [2011]. In the present work, we quantified

EEMT following the algorithm described in Chorover et al. [2011] based on a multiple linear regression model between EEMT and variables that exert first-order controls on photosynthesis and effective precipitation including topographically modified temperature, precipitation, and vapor pressure deficit:

EEMT m

=-3.13+0.00879 (T+273.15) + 0.562PPT + 0.03256 (T-17.65)(PPT-9.0)-

0.00235VPD + 0.00062(PPT-9.0)(VPD-662) (1)

43 where,

EEMT m

= EEMT (MJ.m

-2

) on a monthly basis, T= air temperature (˚C), PPT = precipitation (cm) and VPD = vapor pressure deficit (Pa)

EEMT quantification may be applied across multiple spatial (e.g., pedon, catchment, biome) and temporal (e.g., days, months, years) scales. The energy input to the CZ provided by effective precipitation and net ecosystem production was estimated from climatic data during 10 years from 2000 to 2009 [Chorover et al., 2011].

2.2.3 Mineral weathering

Mineral weathering processes were investigated using spring water aqueous phase chemistry.

This methodology examines major solute concentrations in water as products of mineral weathering reactions [Rademacher et al., 2001; Anderson and Dietrich, 2001; Biron et al., 1999;

Campbell et al., 1995; Goddéris et al., 2006; Godsey et al., 2009, Hooper et al., 1990]. Nine perennial springs were sampled monthly from 2011 to 2013 except during those winter months when sites were inaccessible. All spring samples were analyzed for pH, electrical conductivity, dissolved oxygen and temperature in the field. Water samples for cation analysis were collected in acid-washed 250 ml high density polyethylene (HDPE) bottles and acidified with concentrated optima grade nitric-acid. Water samples for dissolved inorganic carbon (DIC) and anions were collected in 500 ml DI-washed and combusted (475ºC, 4h) amber glass bottles. Before water collection, all the bottles were rinsed three times with sample water, then filled completely to eliminate headspace and stored cool (4°C) until laboratory analysis. Water samples from springs were collected for tritium (

3

H) analysis in duplicate 500 mL HDPE bottles during low flow conditions in the fall 2013.

44

Precipitation samples (snow and rainfall) were collected from 2011 to 2013 during the period of maximum snow accumulation and the monsoon season from July to September. Snow samples were collected during the winters in 2011 and 2013 from snow pits following the methodology described in Gustafson et al. [2010], stored in one gallon Ziploc bags, and maintained frozen until they were allowed to melt overnight at room temperature. Bulk rain water samples were collected weekly in two DI-washed 500 ml HDPE bottles at multiple elevations around Redondo

Peak (See Figure 2 for locations of precipitation sampling). One 500 ml bottle for stable isotope analysis contained a thin layer of mineral oil to prevent water loss by evaporation. All the water samples were filtered promptly with a 0.45 μm nylon filter in the laboratory and splits were sent to the corresponding laboratories for further analysis. In addition, this study incorporates precipitation data collected during previous studies around Redondo Peak [Gustafson et al.,

2010; Broxton et al., 2009]

Major cations were measured by inductively coupled plasma mass spectrometry (ICP-

MS) (ELAN DRC-II, Perkin Elmer, Shelton CT) at the University of Arizona Laboratory for

Emerging Contaminants (ALEC). DIC samples were analyzed with a Shimadzu TOC-VCSH

Carbon analyzer (Shimadzu Scientific Instruments, Columbia) in ALEC. All the water samples were analyzed for δ

18

O with a DLT-100 Laser Spectrometer for liquid water stable isotopes with a reported instrument precision of ±0.12 ‰ VSMOW at the University of Arizona. Tritium samples were analyzed by liquid scintillation counting in a Quantulus 1220 liquid scintillation counter with a 0.7 TU detection limit following an eightfold electrolytic enrichment at the

University of Arizona Environmental Isotope Laboratory. Water ages were calculated using the standard tritium decay equation assuming a constant tritium concentration from precipitation

[Stewart et al., 2010]. Eastoe et al. [2012] observed that tritium concentrations in precipitation

at Albuquerque station (70 km from field area) have remained stable, around 9.0 TU, since the

45 early 1990s. A concentration of 8.0 TU was used for age calculations based on a weighted average of tritium concentrations considering only winter precipitation in the Albuquerque station since 1990. Tritium concentrations in precipitation at Albuquerque have been measured since 1962, and the data are available as part of the Global Network of Isotopes in Precipitation

(GNIP) ( http://www-naweb.iaea.org/napc/ih/IHS_resources_gnip.html, Accessed January, 2014 ).

In addition, a quantitative mineralogical analysis was carried out on unweathered bandelier tuff and rhyolite rock samples. Mineral composition was determined by quantitative xray diffraction on random powder mounts using a PANanalytical X’Pert PRO-MPD x-ray diffraction system at the Center for Environmental Physics and Mineralogy (CEPM) at the

University of Arizona [Vazquez-Ortega et al., 2015].

The chemical data were used in geochemical models MINTEQA2 and NETPATH to investigate saturation indices, chemical reactions during mineral weathering, and the mass of minerals dissolving and precipitating along flow paths. MINTEQA2 is an equilibrium speciation model that calculates the equilibrium mass distribution among dissolved and adsorbed species and multiple solid phases [Allison et al., 1991]. NETPATH is a model for simulating net geochemical reactions and calculating the mass-balance reactions between an initial and final water along a hydrologic flow path [Plummer et al., 1994]. Precipitation chemistry was corrected for evaporative concentration of solutes based on the assumption that increases in Cl

were entirely a result of evapotranspiration [Rademacher et al., 2001; Burns et al., 2003].

Precipitation chemistry, corrected for evapotranspiration, was assumed as the initial preweathering composition and the chemistry from spring discharge was used as the final post-

weathering composition. Information on mineral phases and their stoichiometry was based on rock mineralogical analysis. NETPATH calculates the mass of minerals per kilogram of water that dissolve or precipitate along the flow path in every possible combination of the selected phases that accounts for the observed changes in water chemistry [Plummer et al., 1994]. The

46 chemical evolution and mass balance calculated by NETPATH between an initial and final water composition is assumed to only occur along flow paths [Plummer et al., 1994; Rademacher et

al., 2001].

Finally, the estimations of EEMT and the aspect and landscape characteristic in each contributing area of the springs were compared to base cations concentrations and dissolved inorganic carbon concentrations, WTT, and mineral weathering dissolution measured at the springs.

3.0 RESULTS

3.1 Landscape characteristics

The selected springs drain different aspects of Redondo Peak at elevations between 2816 and 3170 m. Contributing areas of springs vary in size from 56 to 1038 (x10

3 m

2

) and their mean cosine angle covers a wide range of aspects from -0.5 south to 0.96 north. The mean slope of contributing areas range between 10 to 20 degrees, and the median water flow path length ranges from 360 to 600 meters with a median gradient between 0.12 and 0.32 (Table 1).

3.2 EEMT

An annual EEMT map (MJ m

-2

year

-1

) for the entire VCNP is presented in Chorover et

al. [2011; figure 2E]. EEMT values for the Redondo Peak region vary between 22 and 59 MJ m

-

47

2

year

-1

. A visual inspection of the map shows large EEMT values above 30 MJ.m

-2

.year

-1 on

Redondo Peak and low EEMT values below 30 MJ m

-2

year

-1 along shaded areas by topography such as Redondo Creek in the southwestern part of Redondo. Mean, maximum, and the range of

EEMT (maximum – minimum EEMT) within contributing areas of springs vary from 39 to

43MJ m

-2

year

-1

, 40 to 50 MJ m

-2

year

-1

, and 1.8 to 11.4 MJ m

-2

year

-1

, respectively (Table 1).

The mean aspect of a spring contributing area positively correlates with mean EEMT (R

2

=0.31;

p>0.1), maximum EEMT (R

2

=0.90; p<0.0001), and the range of EEMT (R

2

=0.82; p=0.0007)

(Figure 3). Based on the maximum EEMT versus aspect relationship, north facing slopes receive up to 25% more EEMT than south facing slopes (Figure 3).

3.3 Water stable isotopes

The stable isotope composition of precipitation exhibits a seasonal pattern with higher

δ

18

O values during the monsoon season and lower values during the winter. There is an altitudinal isotopic effect observed in winter precipitation (1‰ decrease/increase in δ

18

O per

1000 m elevation change), but no corresponding effect observed for summer precipitation

[Broxton et al., 2009; Adams et al., 1995]. The weighted mean and standard deviation of the

δ

18

O values of the snow and rainfall samples are -15.0±2.5 ‰ and -6.2±2.9 ‰, respectively.

Isotopic variability of the spring waters is considerably more dampened and lies between values from snow and monsoon rainfall (Table 2). δ

18

O values from snow, rainfall and springs indicate that the springs are predominantly recharged by spring snowmelt (Figure 5), which is consistent with findings from previous research demonstrating that the source of subsurface hydrologic recharge in the VCNP is snowmelt dominated [Broxton et al., 2009].

48

3.4 Water Chemistry

Spring waters are dominated by Ca

2+

, Na

+

, Si and HCO

3

-

ions (Table 2). Dissolved inorganic carbon (DIC) concentrations strongly correlate with the products of silicate weathering, including the sum of all base cations (R

2

=0.92; p<0.0001) and Na

+

concentrations

(R

2

=0.86; p=0.001) (Figure 4a). Figures 4b-d indicate significant correlations between all base cations and Na

+

concentrations versus aspect, maximum and range of EEMT. Na

+

is a cation that is not generally cycle to a great degree biologically and it shows the strongest correlations with aspect and EEMT. Water with higher concentrations of solutes is found along north facing slopes (R

2

=0.56; p<0.05), which are characterized in this study by higher EEMT values

(R

2

=0.51; p<0.05). Maximum flow path length and contributing area are the only two landscape characteristics that show statistically significant correlations with both base cations and Na

+ concentrations (Figure 4e-f). Larger concentrations of base cations are observed in springs with longer flow paths (R

2

=0.54; p<0.05) and larger contributing areas (R

2

=0.54; p<0.05). As indicated in Table 3, mean EEMT and other landscape variables, such as mean elevation, mean slope, median flow path length (L), median flow path slope (G), and L/G are poor predictors of both base cations and Na

+

concentrations.

3.5 Water transit times (WTT)

Tritium concentrations measured at the springs ranged from 5.1 to 7.9 TU (Table 2).

Water apparent age in the springs ranged from 0.2 to 8.1 years (Table 2). Table 3 shows a significant relation between terrain aspect versus apparent water age (R

2

=0.74; p<0.01) and maximum EEMT versus apparent water age (R

2

=0.61; p<0.01). Water samples in springs with

49 longer WTT based on tritium apparent ages had correspondingly higher Na

+

concentrations (p <

0.05, Figure 6).

3.6 Bedrock composition and mass balance analysis

Mineralogical analyses of Redondo Peak bedrock samples indicate predominance of primary phases sanidine, oligoclase, anorthoclase, and quartz, accompanied by minor minerals, including apatite, hematite, zircon, titanite, ilmenite, faujasite [Table 4; for detailed mineralogy see Vazquez-Ortega et al., 2015]. Spring waters are undersaturated with respect to silica, halite, calcite, gypsum, sanidine and albite, but saturated to supersaturated with respect to secondary minerals gibbsite, goethite, hematite and kaolinite (Table 5). Saturation index results from

MINTEQA2 were used to select mineral phases for inclusion in NETPATH (sanidine, oligoclase, anorthoclase, quartz, calcite, kaolinite, gypsum, gibbsite, goethite, albite, biotite, albite, hornblende, halite and CO

2

gas). NETPATH then generated 64 models that produced a mass balance between the input waters (evaporated snow chemistry) and spring discharge using the selected phases. Figure 7 shows the mean and standard deviation of the mineral mass balance from the 64 possible combinations of mineral phases that explain the chemistry observed for two springs with contrasting aspect. The results from NETPATH are consistent with the results from

MINTEQA2 and indicate for all the springs the same minerals dissolving and precipitating. The model predicts that gibbsite, goethite, hematite and kaolinite precipitate and all the remaining minerals dissolve along the hydrologic flow paths. Mineral dissolution calculated using

NETPATH is strongly correlated with aspect and EEMT; the highest mineral weathering dissolution amounts were observed in springs discharging north-facing aspects (Table 3; Figure

8).

50

4.0 DISCUSSION

This study provides evidence that energy and mass fluxes quantified using EEMT modeling can predict, at a large scale (~40 km

2

), measured components of CZ hydrologic and aqueous geochemical function, such as WTT and incongruent weathering of silicates. The predictive power of EEMT observed in this study is consistent with previous work that demonstrated in the western US strong relations between EEMT and Si fluxes from granitoid watersheds and soil carbon content [Rasmussen et al., 2005]. Around Redondo Peak, Perdrial et

al. [2014] measured higher soluble organic carbon effluxes from catchments draining northfacing aspects characterized in this study by higher EEMT values. Moreover, along elevation gradients on different parent materials in the western US, strong correlations were observed between EEMT and pedogenic indices (pedon depth, total pedon clay content, free Fe oxide to tal Fe oxide ratio, chemical index of alteration minus potassium of the first subsurface genetic horizon) [Rasmussen et al., 2011]. In the study by Rasmussen et al., [2011] poor and no correlations were observed between mean annual temperature and precipitation alone versus pedogenic indices proving that EEMT is a better predictor than only climatic variables. The methodology to estimate EEMT based on average climate data as presented in this study and initially described in Rasmussen et al., [2005] and Rasmussen and Tabor, [2007], was validated with empirical estimations of EEMT at the catchments scale in 86 catchments in the US, showing a significant linear correlation between modeled and empirical EEMT values

[Rasmussen and Gallo, 2013].

The results from the current study demonstrate that in a terrain underlain by rhyolite rocks, higher EEMT is associated with longer WTT and enhanced weathering. Similarly,

51

Broxton et al. [2009] observed longer WTT of surface water along north facing aspects of

Redondo Peak that have been characterized in this study by larger EEMT. Based on water stable isotopes, the mean and standard deviation of WTT was 123±29 days for streams draining all aspects of Redondo Peak, with a maximum difference between south and north facing slopes of

50 days [Broxton et al., 2009]. However, stable isotope tracers can only be used to understand young water systems not older than 5 years [Stewart et al., 2010], therefore this study complemented previous age dating analysis with tritium. The results of this investigation are also consistent with laboratory and field studies that have demonstrated that weathering of silicates increase with WTT [Velbel, 1993; Gabet et al., 2006; Maher, 2010; Maher, 2011; White

et al., 2001; Burns et al., 2003; Rademacher et al., 2001; Rademacher et al., 2005].

This study indicates metrics of mineral weathering are correlated with the range and maximum values of EEMT distributed over the catchment contributing to spring discharge. The observation that mean EEMT showed less predictive power is consistent with the concept of subcatchment “hot spots” contributing disproportionately to a given catchment discharge. The fact that EEMT is not uniformly distributed across a catchment, and that it exhibits strong topographic variation from effects of aspect and topographic convergence of hydrologic flow, particularly in water limited systems, has led to inclusion of such landscape structural parameters into more recent versions of the EEMT calculation than the one employed in the current study

[Rasmussen et al., 2015]. Such an approach could be employed in future analyses of comparable data sets to better resolve why the range and maximum EEMT values are better predictors that mean values.

52

Base cations and especially Na

+

concentrations, around Redondo Peak show strong correlations with dissolved inorganic carbon and WTT, suggesting they may be used as indirect tracers of water WTT with significant improvements in spatial and temporal resolutions of WTT.

Saturation indices (SI) prove that Na

+

is behaving quite conservatively in this system, therefore the increase in concentrations with increase in WTT can be attributed to primary mineral weathering reactions. Chemical formulae of supersaturated secondary phases also indicate that secondary mineral formation do not provide a sink for Na

+

, which is also supported by a previous geochemical study around Redondo, where Na

+

was identified as a conservative tracer and used in an end-member mixing analysis [Liu et al., 2008b]. Liu et al. [2008a; 2008b], observed that groundwater contributions dominate runoff generation throughout the year around

Redondo Peak, with limited contributions from overland flow and shallow subsurface flow. In addition, base cations concentrations in surface water draining Redondo Peak exhibit chemostatic behavior, which suggests a constant source of water supply to surface water, likely from a well-mixed reservoir [Porter, 2012; Perdrial et al., 2014]. Moreover, dissolved organic carbon (DOC) concentrations measured in the springs and first order springs are significantly lower than DOC concentrations from soil waters during baseflow conditions [Porter, 2012;

Perdrial et al., 2014], which supports the hypothesis that groundwater is the dominant source of water in the Redondo system and base cations in the springs are products of primary mineral weathering reactions.

The correlations among energy and mass inputs to the CZ (EEMT), WTT, and mineral weathering products can provide a methodology for indirect estimations of WTT and mineral weathering products. EEMT estimations based on climatic variables, can be used to scale hydrological and hydrochemical responses in other sites. Base cations concentrations combined

53 with EEMT maps can provide estimations of WTT at large spatial scales, thus enabling parameter estimations in hydrological models.

CZ structure influences water movement and solute transport, and an understanding of

CZ heterogeneous structure and organization can contribute to the predictability of hydrological response, especially in regions with scarce hydrological information [Troch et al., 2008].

Concentrations of solutes in water increase with apparent water age as a result of chemical weathering [Drever, 1988; Herczeg and Edmunds, 2000]. Major solutes in water are derived predominantly from water-rock interactions in the unsaturated and saturated zones. In most hydrological systems solutes concentrations increase as water moves down gradient, therefore higher concentrations of ions are associated with older waters and hence major ions may be used as indirect water dating tools [Herczeg and Edmunds, 2000; Rademacher et al., 2001].

Significant work on reactive tracers as indicators of water WTT times has been primarily conducted in aquifers [Hendry and Schwarts, 1990; Edmunds and Smedley, 2000; Burton et al.,

2002], however, a few studies have successfully demonstrated the evolution of water chemistry in shallow groundwater and surface water by taking advantage of piezometers and springs for the application of age tracers: for example in the riparian groundwater in the Panola Mountain

Research Watershed dominated by granitic rocks strong correlations were found between SiO

2

,

Na

+

and Ca

2+

and water age [Burns et al., 2003]. In the Sagehen basin in California, which is dominated by andesite and granodiorite rocks, studies suggest that water ages can be predicted based on major cation concentrations measured at springs [Rademacher et al., 2001; Rademacher

et al., 2005]. Similarly, in New Zealand in a region comprised of andesitic volcanic ash, good correlations were observed between mean transit time and silica concentrations in surface water at the catchment scale [Morgestern et al., 2010].

54

5.0 SUMMARY

Results from this study demonstrate the close interrelationship between landscape, hydrological and biogeochemical processes. Our study highlights statistically significant relations between effective energy and mass transfer (EEMT) to the subsurface critical zone,

WTT, and mineral weathering in springs draining catchments along a gradient of energy controlled by differences in terrain aspect around Redondo Peak, in northern New Mexico.

Spring waters around Redondo Peak are dominated by Ca

+

, Na

+

, Si and HCO

3

-

and have an isotopic signature that indicates they are predominantly derived from infiltration of snowmelt.

Terrain aspect controls EEMT where north facing slopes receive 25% more EEMT than south facing slopes. Larger concentrations of total base cations and Na

+

were observed in springs with longer flow paths, larger contributing areas, and north facing slopes. This result is consistent with longer WTT based on tritium analysis of springs draining north-facing terrains. Mineral dissolution fluxes increase with WTT, likely due to enhanced water-rock reaction, and chemical weathering consumes more atmospheric CO

2 along north facing slopes. This study provides evidence that fluxes of energy and mass, quantified as EEMT, at the catchment scale can effectively predict short time-scale (months to years) processes within the CZ structure like WTT and silicate mineral weathering. These results also suggest that basic climatic data embodied in the EEMT term can be used to scale hydrological and hydrochemical responses in other sites.

6.0 ACKNOWLEDGEMENTS

We are grateful to collaborators Chief Scientist Bob Parmenter and Research Hydrologist Scott

55

Compton, at the Valles Caldera National Preserve. LiDAR data acquisition and processing were completed by the National Center for Airborne Laser Mapping (NCALM), funded by the

National Science Foundation Award EAR-0922307, and coordinated by Qinghua Guo for the

Jemez River Basin and Santa Catalina Mountains Critical Zone Observatory funded by the

National Science Foundation Award EAR-0724958. Logistical support and/or data were provided by the NSF-supported Jemez River Basin and Santa Catalina Mountains Critical Zone

Observatory EAR-0724958 and EAR-1331408). Data access and data sharing policy are available at the Catalina – Jemez River basin Critical Zone Observatory http://criticalzone.org/catalina-jemez/data

56

7.0 REFERENCES

Adams A. I., F. Goff, D. Counce (1995), Chemical and isotopic variations of precipitation in the

Los Alamos region, New Mexico, Rep. LA-12895-MS, Los Alamos Natl. Lab., Los Alamos,

N.M.

Allison, J.D., Brown, D.S., and Novo-Gradac, K.J. 1991. MINTEQA2/PRODEFA2, A geochemical assessment model for environmental systems: version 3.0, Environmental Research

Laboratory, Office of Research and Development, U.S. Environmental Protection Agency,

Athens, Georgia. EPA/600/3-91/021.

Anderson, S. and W. Dietrich (2001), Chemical weathering and runoff chemistry in a steep headwater catchment, Hydrol. Process., 15, 1791-1815.

Berner, R., A. Lasaga, and R. Garrels (1983), The Carbonate-Silicate Geochemical Cycle and its

Effect on Atmospheric Carbon-Dioxide Over the Past 100 Million Years, Am. J. Sci., 283, 641-

683.

Berner, R., A. (1995), Chemical weathering and its effect on atmospheric CO2 and climate. In:

White, A.F., Brantley, S.B. (Eds), Chemical weathering rates of silicate minerals, vol 31.

Mineralogical Society of America, Washington, DC, pp, 565-583; Ch13.

Biron, P., A. Roy, F. Courschesne, W. Hendershot, B. Cote, and J. Fyles (1999), The effects of antecedent moisture conditions on the relationship of hydrology to hydrochemistry in a small forested watershed, Hydrol. Process., 13, 1541-1555.

Brantley, S. L., M. B. Goldhaber, and K. V. Ragnarsdottir (2007), Crossing disciplines and scales to understand the Critical Zone, Elements, 3, 307-314.

57

Bricker, O. (1986), Geochemical Investigations of Selected Eastern-United-States Watersheds

Affected by Acid Deposition, Journal of the Geological Society, 143, 621-626.

Broxton, P. D., P. A. Troch, and S. W. Lyon (2009), On the role of aspect to quantify water transit times in small mountainous catchments, Water Resour. Res., 45, W08427.

Burns, D. A. et al. (2003), The Geochemical Evolution of Riparian Ground Water in a Forested

Piedmont Catchment, Ground Water, 41, 913-925.

Burton, W., L. Plummer, E. Busenberg, B. Lindsey, and W. Gburek (2002), Influence of fracture anisotropy on ground water ages and chemistry, Valley and Ridge province, Pennsylvania,

Ground Water, 40, 242-257.

Campbell, D., D. Clow, G. Ingersoll, M. Mast, N. Spahr, and J. Turk (1995), Processes

Controlling the Chemistry of 2 Snowmelt-Dominated Streams in the Rocky-Mountains, Water

Resour. Res., 31, 2811-2821.

Chorover, J. et al. (2011), How Water, Carbon, and Energy Drive Critical Zone Evolution: The

Jemez-Santa Catalina Critical Zone Observatory, Vadose Zone Journal, 10, 884-899.

Coop, J. D. and T. J. Givnish (2007), Spatial and temporal patterns of recent forest encroachment in montane grasslands of the Valles Caldera, New Mexico, USA, J. Biogeogr., 34, 914-927.

Dokuchaev, V.V (1967) Selected works of V.V. Dokuchev, Israel Program for Scientific

Translations, available from the US Dept. of Commerce, Clearinghouse for Federal Scientific and Technical Information, Jerusalem, Springfield, Va.

58

Drever J.I (1988), The geochemistry of natural waters. Prentice Hall. Second edition

Eastoe, C. J., C. J. Watts, M. Ploughe, and W. E. Wright (2012), Future Use of Tritium in

Mapping Pre-Bomb Groundwater Volumes, Ground Water, 50, 87-93.

Edmunds, W. and P. Smedley (2000), Residence time indicators in groundwater: the East

Midlands Triassic sandstone aquifer, Appl. Geochem., 15, 737-752.

Ellis, S., G. Levings, L. Carter, S. Richey, and M. Radell (1993), Rio-Grande Valley, Colorado,

New-Mexico, and Texas, Water Resour Bull, 29, 617-646.

Gabet, E. J., R. Edelman, and H. Langner (2006), Hydrological controls on chemical weathering rates at the soil-bedrock interface, Geology, 34, 1065-1068.

Godderis, Y., L. Francois, A. Probst, J. Schott, D. Moncoulon, D. Labat, and D. Viville (2006),

Modelling weathering processes at the catchment scale: The WITCH numerical model, Geochim.

Cosmochim. Acta, 70, 1128-1147.

Godsey, S. E., J. W. Kirchner, and D. W. Clow (2009), Concentration-discharge relationships reflect chemostatic characteristics of US catchments, Hydrol. Process., 23, 1844-1864.

Goff,F., J.N. Gardner, S.L. Reneau, and C.J. Goff (2006), Geologic map of the Redondo Peak quadrangle, Sandoval County, New Mexico: New Mexico Bureau of Geology and Mineral

59

Resources, Open-file Map Series OFGM-111, scale 1:24,000, http://geoinfo.nmt.edu/publications/maps/geologic/ofgm/ , accessed January 15, 2013

Gustafson, J. R., P. D. Brooks, N. P. Molotch, and W. C. Veatch (2010), Estimating snow sublimation using natural chemical and isotopic tracers across a gradient of solar radiation,

Water Resour. Res., 46, W12511.

Hendry, M. and F. Schwartz (1990), The Chemical Evolution of Ground-Water in the Milk River

Aquifer, Canada, Ground Water, 28, 253-261.

Herczeg, A.L, and W. M. Edmunds (2000), Inorganic ions as tracers. In: Environmental tracers

in subsurface hydrology (Cook and Herczeg) (eds)

Hooper, R., N. Christophersen, and N. Peters (1990), Modeling Streamwater Chemistry as a

Mixture of Soilwater End-Members - an Application to the Panola Mountain Catchment,

Georgia, Usa, Journal of Hydrology, 116, 321-343.

Jenny, 1941. Factors of soil formation: a system of quantitative pedology.McGraw-Hill book

Company, 1 st

edition, New York, London, xii, 281pp.

Kleidon A, E. Zehe, H. Lin (2012), Thermodynamic limits of the critical zone and their relevance to hydropedology. In: H. Lin. Ed., Hydropedology. Elsevier B.V.

Liu, F., R. C. Bales, M. H. Conklin, and M. E. Conrad (2008a), Streamflow generation from snowmelt in semi-arid, seasonally snow-covered, forested catchments, Valles Caldera, New

Mexico, Water Resour. Res., 44, W12443.

Liu, F., R. Parmenter, P. D. Brooks, M. H. Conklin, and R. C. Bales (2008b), Seasonal and interannual variation of streamflow pathways and biogeochemical implications in semi-arid, forested catchments in Valles Caldera, New Mexico, Ecohydrology, 1, 239-252.

60

Lyon, S. W., P. A. Troch, P. D. Broxton, N. P. Molotch, and P. D. Brooks (2008), Monitoring the timing of snowmelt and the initiation of streamflow using a distributed network of temperature/light sensors, Ecohydrology, 1, 215-224.

Maher, K. (2010), The dependence of chemical weathering rates on fluid residence time, Earth

Planet. Sci. Lett., 294, 101-110.

Maher, K. (2011), The role of fluid residence time and topographic scales in determining chemical fluxes from landscapes, Earth Planet. Sci. Lett., 312, 48-58.

Maher, K. and C. P. Chamberlain (2014), Hydrologic Regulation of Chemical Weathering and the Geologic Carbon Cycle, Science, 343, 1502-1504.

McGlynn, B., J. McDonnell, M. Stewart, and J. Seibert (2003), On the relationships between catchment scale and streamwater mean residence time, Hydrol. Process., 17, 175-181.

McGuire, K. J. and J. J. McDonnell (2006), A review and evaluation of catchment transit time modeling, Journal of Hydrology, 330, 543-563.

McGuire, K., J. McDonnell, M. Weiler, C. Kendall, B. McGlynn, J. Welker, and J. Seibert

(2005), The role of topography on catchment-scale water residence time, Water Resour. Res., 41,

W05002.

61

Muldavin E, Neville P, Jackson C, Neville T. (2006), A Vegetation Map of Valles Caldera

National Preserve, New Mexico. Natural Heritage: New Mexico; 59.

Mueller MH, Weingartner R, Alewell, C. (2012), Relating stable isotope and geochemical data to conclude on water residence times in four small alpine headwater catchment with differing vegetation cover. Hydrol. Earth Syste, Sci Discuss, 9, 11005-11048

National Research Council (2001). Basic research opportunities in earth science. National

Academy Press. Washington, D.C.

Pelletier, J. D. and C. Rasmussen (2009a), Geomorphically based predictive mapping of soil thickness in upland watersheds, Water Resour. Res., 45, W09417.

Pelletier, J. D. and C. Rasmussen (2009b), Quantifying the climatic and tectonic controls on hillslope steepness and erosion rate, Lithosphere, 1, 73-80.

Perdrial, J. N. et al. (2014), Stream water carbon controls in seasonally snow-covered mountain catchments: impact of inter-annual variability of water fluxes, catchment aspect and seasonal processes, Biogeochemistry, 118, 273-290.

Phillips, J. D. (2009), Biological Energy in Landscape Evolution, Am. J. Sci., 309, 271-289.

Plummer L.N, Prestemon, E.C., Parkhurst D.L.(1994) An interactive code (NETPATH) for modeling net geochemical reaction along a flow path version 2.0

62

Porter, C. (2012), Solute inputs to soil and stream waters in a seasonally snow covered mountain catchment determined using Ge/Si,

87

Sr/

86

Sr and major ion chemistry: Valles Caldera, New

Mexico. MS Thesis. University of Arizona, pp 88.

Rademacher, L., J. Clark, D. Clow, and G. Hudson (2005), Old groundwater influence on stream hydrochemistry and catchment response times in a small Sierra Nevada catchment: Sagehen

Creek, California, Water Resour. Res., 41, W02004.

Rademacher, L., J. Clark, G. Hudson, D. Erman, and N. Erman (2001), Chemical evolution of shallow groundwater as recorded by springs, Sagehen basin; Nevada County, California, Chem.

Geol., 179, 37-51.

Rasmussen, C., R. Southard, and W. Horwath (2005), Modeling energy inputs to predict pedogenic environments using regional environmental databases, Soil Sci. Soc. Am. J., 69, 1266-

1274.

Rasmussen, C. and N. J. Tabor (2007), Applying a quantitative pedogenic energy model across a range of environmental gradients, Soil Sci. Soc. Am. J., 71, 1719-1729.

Rasmussen, C., P. A. Troch, J. Chorover, P. Brooks, J. Pelletier, and T. E. Huxman (2011), An open system framework for integrating critical zone structure and function, Biogeochemistry,

102, 15-29.

Rodgers, P., C. Soulsby, and S. Waldron (2005), Stable isotope tracers as diagnostic tools in upscaling flow path understanding and residence time estimates in a mountainous mesoscale catchment, Hydrol. Process., 19, 2291-2307.

Runge, E. (1973), Soil Development Sequences and Energy Models, Soil Sci., 115, 183-193.

63

Sheppard, P., A. Comrie, G. Packin, K. Angersbach, and M. Hughes (2002), The climate of the

US Southwest, Climate Research, 21, 219-238.

Smeck, N.E., E.C.A. Runge, E.E. Mackintosh (1983), Dynamics and genetic modeling of soil systems, 51-81. In L.P. Wilding et al. (ed) Pedogenesis and soil taxonomy. Elsevier, New York

Soulsby, C., D. Tetzlaff, S. M. Dunn, and S. Waldron (2006), Scaling up and out in runoff process understanding: insights from nested experimental catchment studies, Hydrol. Process.,

20, 2461-2465.

Stewart, M. K., U. Morgenstern, and J. J. McDonnell (2010), Truncation of stream residence time: how the use of stable isotopes has skewed our concept of streamwater age and origin,

Hydrol. Process., 24, 1646-1659.

Tetzlaff, D., J. Seibert, K. J. McGuire, H. Laudon, D. A. Burn, S. M. Dunn, and C. Soulsby

(2009), How does landscape structure influence catchment transit time across different geomorphic provinces?, Hydrol. Process., 23, 945-953.

Troch P.A., G.A. Carrillo, I. Heidbuechel, D. Rajagopal, M. Seitanek, T.H.M. Volkmann, M.

Yaeger. (2008), Dealing with landscape heterogeneity in watershed hydrology: a review of recent progress toward new hydrological theory. Geography Compass 2: 10.1111/j.1749-

8198.2008.0086.x

Vazquez-Ortega, A. et al. (2015), Rare earth elements as reactive tracers of biogeochemical weathering in forested rhyolitic terrain, Chem. Geol., 391, 19-32.

64

Velbel, M. (1993), Constancy of Silicate Mineral Weathering-Rate Ratios between Natural and

Experimental Weathering - Implications for Hydrologic Control of Differences in Absolute

Rates, Chem. Geol., 105, 89-99.

Velbel, M. A. and J. R. Price (2007), Solute geochemical mass-balances and mineral weathering rates in small watersheds: Methodology, recent advances, and future directions, Appl. Geochem.,

22, 1682-1700.

Vitvar T, Aggarwal P, McDonnell JJ (2005), A review of isotope applications in catchment hydrology. In Isotopes in the Water Cycle: Past, Present and Future of a Developing Science,

Aggarwal PK, Gat J, Froehlich K (eds). Springer: Dordrecht; 151-170.

Volobuev, V.R. (1964), Ecology of soils. Academy of sciences of the Azerbaidzan SSR. Institute of Soil Science and Agrochemistry. Israel Program for scientific translations, Jerusalem.

White, A., T. Bullen, M. Schulz, A. Blum, T. Huntington, and N. Peters (2001), Differential rates of feldspar weathering in granitic regoliths, Geochim. Cosmochim. Acta, 65, 847-869.

White and Brandley (1995), Chemical weathering rates of silicate minerals: an overview. In:

White, A.F., Brantley, S.L.(Eds.), Chemical weathering rates of silicate minerals, Reviews in

Mineralogy, vol.31. Mineralogical Society of America, Washington, DC, pp 1-22 (Chapter 1).

Wolff, J. A., K. A. Brunstad, and J. N. Gardner (2011), Reconstruction of the most recent volcanic eruptions from the Valles caldera, New Mexico, J. Volcanol. Geotherm. Res., 199, 53-

68

65

8.0 FIGURES

Figure 1. Conceptualization of dominant critical zone (CZ) processes (after Chorover et al.,

2011 and Rasmussen et al., 2011).

66

67

Figure 2. a) Location of the research site in northern New Mexico, b) Location of the Valles

Caldera National Preserve within the upper Jemez River Basin, c) Redondo peak and springs draining different terrain aspects

50

12

R

2

=0.90****

48

10

8

46

6

44

4

42

R

2

=0.82***

40

-1.0

-0.5

0.0

0.5

1.0

cos (aspect)

2

0

Max EEMT

Range EEMT

Figure 3. Maximum and range (maximum – minimum) of EETM values versus mean aspect within springs’ contributing areas. (*** p ≤ 0.001; **** p≤0.0001). There is not a statistically significant relationship between mean aspect and mean EEMT (R

2

=0.31; p>0.1)

68

a)

600

500

400

300

R

2

=0.92*** b)

600

500

400

300

200

100

0

R

2

=0.56*

200

100

0

R

2

=0.86***

R

2

=0.71**

500

400

300

2 4 6 8

DIC (mg/L)

10 12 d)

-1.0 -0.5 0.0

0.5

1.0

cos (aspect) c)

600

R

2

=0.51*

600

500

400

300

R

2

=0.71**

200

100

R

2

=0.67**

0

40 42 44 46 48 50

Max EEMT (MJ/m

2

year) e)

600

500

R

2

=0.54*

200

100

R

2

=0.73**

0

0 2 4 6 8 10 12

Range EEMT (MJ/m

2

year) f)

600

R

2

=0.54*

500

400 400

300 300

200 200

100

R

2

=0.51*

0

500 1000 1500 2000 2500

100

0

Max flow length (m)

R

2

=0.61**

0 500 1000

Contributing Area (x 1000 m

2

)

Na

Base cations

69

Figure 4. Concentrations of base cations, Na

+

and DIC in spring water versus EEMT and landscape characteristics. Statistical significance (*p≤0.05; **p≤0.01; *** p ≤ 0.001)

70

3300

3200

3100

3000

2900

2800

2700

-18 -16 -14 -12 -10 -8 -6 -4

δ

18

O (‰)

Figure 5. δ

18

O values of precipitation and springs relative to elevation.

Mean snow

Mean rainfall

Springs

1St dev snow

1St dev rain

71

10

8

6

4

2

R

2

=0.56*

R

2

=0.61**

0

40 42 44 46 48 50

0

Max EEMT (MJ/m

2

year )

30

25

20

15

10

5

250

200

150

100

50

0

-50

R

2

=0.43*

0 2 4 6 8 10

Age (years)

Figure 6. Max EEMT versus tritium concentration and apparent age (left). Apparent age versus

Na

+

concentrations (right). Statistical significance (*p≤0.05; **p≤0.01)

72

73

1.0

0.8

0.0

-0.2

-0.4

-0.6

0.6

0.4

0.2

UJ2s

Es

Figure 7. Mineral dissolution (+) and precipitation (-) for a north aspect draining (UJ2s) and south aspect draining (Es) spring. Symbols represent mean mineral dissolution and precipitation and error bars represent 1 standard deviation from 64 mass balance models checked with

NETPATH. The mass balance models determined by NETPATH represent every possible geochemical mass balance reaction between the initial and final water given a set of chemical constrains and phases.

74

0.4

0.4

0.3

R

2

=0.71**

0.3

0.2

R

2

=0.76**

0.2

0.1

0.1

0.0

0.0

-0.1

-1.0

-0.5

0.0

0.5

1.0

cosine aspect

-0.1

40 42 44 46 48 50

Max EEMT (MJ/m

2

year)

Figure 8. Dissolution of Anothoclase estimated using NETPATH (Plummer et al., 1994) versus aspect and max EEMT. Statistical significance (**p≤0.01)

75

Climate controls

Springs Latitude Longitude

Code degrees degrees cos aspect

(-)

Mean

EEMT

MJ/m

2 year

Max

EEMT

MJ/m

2 year

Range

EEMTΗ‚ Elevation

MJ/m

2 year

LJ3s

UR1s

UJ2s

Es

UJ1s

UR2s

HGs

LJ2s

LJ1s

35.872

35.882

35.896

35.892

35.896

35.89

35.889

35.877

-106.529

-106.573

-106.538

-106.511

-106.546

-106.567

-106.552

-106.543

-0.12

0.72

0.96

-0.51

0.81

0.87

-0.47

0.27

35.881 -106.549 0.06

Note:

Η‚Range EEMT: Max-Min EEMT

39.72 41.09

42.18 46.46

41.45 48.19

38.89 40.63

2.01

9.65

9.43

2.63

41.42 47.64 8.82

41.43 49.53 11.43

40.18 41.19 1.77

4.53 42.99 45.37

43.08 44.27 1.93

(m)

2816

2842

2848

2860

2876

2877

2908

3070

3170

§According to McGuire et al., 2005 contributing area

Mean slope

(x 10

3

m

2

)

63

287

1038

97

69

404

232

541

56

Landscape controls

(degrees)

16.08

19.94

17.53

13.95

15.90

17.23

10.82

12.78

13.66

Maximum flow

Length

(m)

896

1862

1482

610

851

2140

992

1922

766

Median flow length (L)§

(m)

462

602

474

361

450

509

399

549

510

Median flow path gradient

(G)§

(-)

0.32

0.17

0.26

0.14

0.28

0.24

0.13

0.12

0.18

9.0 TABLES

Table 1. List of springs draining Redondo Peak arranged from low to high elevation. Climate and landscape characteristics defined in a spring’s contributing area.

L/G§

(m)

1460.6

3489.3

1809.9

2514.6

1595.6

2111.4

3056.9

4498.0

2822.3

76

77

Table 3. Linear regression between climate / landscape variables and solutes, mineral dissolution and water age

Cosine aspect (-)

Max EEMT (MJ/m

2

)

Range EEMT (MJ/m

2

)

0.67** 0.67** 0.51* 0.76***

0.82*** 0.73** 0.71** 0.85***

Mean EEMT (MJ/m

2

)

Mean Elevation (m)

Contributing Area (m

2

)

Mean slope (degrees)

Max flow path length (m)

0.04

0.18

0.38

0.59*

0.57*

Median flow path length (L) (m) 0.24

Median flow path slope (G) (-) 0.17

0.04

0.16

0.61*

0.37

0.51*

0.13

0.19

L/G (m) 0.05 0.06

Notes

Η‚ Same R

2

and p values were found for sanidine and oligoclase

0.03

0.22

0.54*

0.56*

0.54*

0.18

0.08

0.02

0.10

0.10

0.27

0.42

0.69**

0.25

0.11

0.02

§According to McGuire et al., 2005

HCO

3

(umol/L)

Na

+

(umol/L)

Base

Cations

(umol/L)

AnorthoclaseΗ‚

(mmol/L)

--------------------R

2

----------------------

0.71** 0.71** 0.56* 0.71**

Water age

(years)

0.74**

0.78***

0.45*

0.49*

0.06

0.19

0.21

0.38

0.47*

0.15

0.00

Statistical significance

* P ≤ 0.05

**

P ≤ 0.01

***

P ≤ 0.001

78

Table 4. Chemical composition of the most abundant minerals in the Valles

Caldera rhyolite (see Vazquez-Ortega et al., 2015, for details).

Rhyolite

Sanidine

Anorthoclase

Oligoclase

Quartz

Minor minerals

Chemical formula

K

0.36

Na

0.54

Al

1.02

Si

2.99

O

8

Na

0.70

K

0.23

Al

1.05

Si

2.94

O

8

Na

0.69

Ca

0.24

Al

1.23

Si

2.76

O

8

SiO

2

Apatite, Hematite, Zircon, Titanite, Ilmenite, Faujasite

79

Table 5. Mineral saturation index log [IAP/K so

] estimated with MINTEQA2

Spring code

Sample ID

SiO2

GIBBSITE

GOETHITE

HEMATITE

NaCl

CALCITE

GYPSUM

KAOLINITE

SANIDINE

UR1s UR2s UJ1 UJ2s Es LJ1s LJ2s LJ4s HGs

Log KΗ‚ NMS1234 NMS1235 NMS1237 NMS1236 NMS1240 NMS1230 NMS1229 NMS1238 NMS1241

-2.71 -0.55 -0.61 -0.64 -0.52 -0.65 -0.43 -0.62 -0.66 -0.39

8.77

0.50

-4.00

1.98

4.94

12.19

0.44

4.75

11.82

1.79

4.19

10.69

0.35

5.21

12.73

0.71

4.40

11.11

1.33

5.15

12.61

0.27

4.65

11.61

0.66

3.74

9.78

0.35

4.02

10.36

1.58

-8.48

-4.85

5.73

1.06

ALBITE 2.59

Η‚ MINTEQ database

-10.29

-1.89

-3.70

6.66

-128.24

-85.35

-10.34

-2.11

-3.89

3.53

-117.43

-79.77

-10.17

-2.85

-3.92

6.11

-128.38

-84.70

-9.98

-1.56

-3.06

3.54

-131.30

-87.60

-9.89

-2.50

-3.48

4.62

-128.80

-85.64

-10.24

-1.70

-3.07

5.69

-129.65

-86.57

-10.23

-2.12

-3.86

3.18

-133.14

-89.02

-9.97

-2.08

-3.24

3.91

-140.43

-92.52

-10.08

-2.49

-3.35

3.82

-130.94

-86.92

Note: Mineral reactions

SiO2SiO

2

+ 2H

2

O = H

4

SiO

4

Gibbsite

Al(OH)

3

+ 3H

+

= Al

+3

+ 3H

2

O

Goethite

FeOOH + 3H

+

= Fe

+3

+ 2H

2

O

Hematite

Fe

2

O

3

+ 6H+ = 2Fe

+3

+ 3H

2

O

NaCl

NaCl = Na

+

+ Cl

-

Calcite

CaCO

3

= Ca

+2

+ CO

3

-2

Gypsum

CaSO

4

:2H

2

O = Ca

+2

+ SO

4

-2

+ 2H

2

O

Kaolinite

Al

2

Si

2

O

5

(OH)

4

+ 6H

+

= 2Al

+3

+ 2H

4

SiO

4

+ H

2

O

Sanidine

KAlSi

3

O

8

+ 4H

2

O + H

+

= Al

3+

K

+

+ 3H

4

SiO

4

Albite

NaAlSi

3

O

8

+ 4H

+

+ 4H

2

O = Na

+

+ Al

+3

+ 3H

4

SiO

4

80

APPENDIX B:

INFLUENCE OF TERRAIN ASPECT ON WATER PARTITIONING, VEGETATION

STRUCTURE, AND VEGETATION GREENING IN HIGH ELEVATION CATCHMENTS IN

NORTHERN NEW MEXICO

Manuscript to be submitted to the journal of ecohydrology

Authors:

Xavier Zapata-Rios

1*

, Peter A. Troch

1

, Paul D. Brooks

2, 1

, Jennifer McIntosh

1

, Qinghua Guo

3

1. Department of Hydrology and Water Resources, University of Arizona, Tucson, Arizona,

USA

2. Department of Geology and Geophysics, University of Utah, Salt Lake City, Utah, USA

3. Sierra Nevada Research Institute, University of California Merced, Merced, California,

USA

* Corresponding author:

1133 E James E Rogers Way

J W Harshbarger Bldg Rm 122, PO Box 210011

The University of Arizona

Tucson AZ 85721-0011

Email: [email protected]

81

ABSTRACT

We investigated vegetation structure, water partitioning dynamics and vegetation greening from

2000 through 2012 in three catchments draining north and east aspects of Redondo Peak in northern New Mexico. Vegetation structure was quantified from 1 m LiDAR data while vegetation greening was quantified using remotely sensed NDVI. Hydrological partitioning at the catchment scale was estimated with a metric of catchment-scale water fluxes and vegetation water use (Horton index; HI). The predominantly north facing catchment, when compared to the other two eastern catchments, receives less solar radiation, and exhibits less forest cover and smaller biomass, has more surface runoff (~15% of P) as a consequence of a smaller vaporization (85% of P) and smaller vegetation water consumption (HI=0.85). Moreover, the north facing catchment showed smaller peak NDVI values (5.98) and shorter growing season length (76.3 days) as a consequence of energy limitation. In contrast, the two eastern catchments receive larger solar radiation, have more biomass and forest cover (>76%), and smaller surface runoff (<10% P), higher vaporization (>90%P) and vegetation water consumption (HI=0.95).

The eastern catchments had larger vegetation greening (6.28-6.58) and a longer growing season

(113-156 days). Snowpack conditions, such as maximum SWE and duration of the snow on the ground, explain over 95% of water partitioning (HI) that in turn influenced annual vegetation greening (R

2

=0.48 - 0.67; p<0.05). This catchment scale study in perennial streams indicates that terrain aspect at a similar altitude (2700 – 3435m) strongly controls energy, water distribution, and vegetation productivity in high elevation ecosystems.

KEY WORDS:

Aspect, vegetation structure, hydrological partitioning, Horton index, NDVI, mountain catchments, critical zone, New Mexico

82

83

1.0 INTRODUCTION

In mid and high latitude regions, the orientation of the land surface, referred to as terrain aspect, controls the partitioning of energy and precipitation, influencing microclimate, vegetation characteristics and water flow paths [Broxton et al., 2009; Gutierrez-Jurado et al., 2013; Desta et

al., 2004; Newman et al., 2006]. Energy and water availability controlled by aspect differences has been proposed as an important regulator of plant microhabitats that in turn influence growth and production of plants in terrestrial ecosystems, particularly in semi-arid regions [Lieth, 1975;

Webb et al., 1986; Noy-Meir, 1973; Wilkinson and Humphreys, 2006; Hinckley et al., 2012].

Aspect influences hydrologic processes occurring above the land surface, such as snow accumulation, snowmelt, sublimation, soil moisture and evapotranspiration [Musselman et al.,

2008; Veatch et al., 2009; Gustaffson et al., 2010; Molotch et al., 2009; McDowell et al., 2007;

Small and McConnell, 2008; Vivoni et al., 2008; Rinehart et al. 2008; Harpold et al. 2014]. Past research on the role of aspect on subsurface hydrologic processes showed that terrain orientation influences soil moisture content, soil water retention capacity [Geroy, 2011], hydraulic conductivity [Casanova et al., 2000] and water movement in hillslopes [Hinckley et al., 2012].

Similarly, in our study site, at Redondo Peak located in northern New Mexico, aspect influences energy, water availability, biogeochemistry and soil development [Lyon et al., 2008;

Broxton et al., 2009; Perdrial et al., 2014; Zapata-Rios et al., in review]. North facing slopes at this site have delayed snowmelt, more water availability in streams [Lyon et al., 2008], longer water residence times and larger mineral weathering dissolution rates [Broxton et al., 2009;

Zapata-Rios et al., in review], larger dissolved inorganic and dissolved organic carbon concentrations [Perdrial et al., 2014].

84

The investigations relating aspect, vegetation, and hydrological feedbacks have primarily concentrated on plots, slope transects and small scale drainages [Gutierrez-Jurado et al., 2006;

Gutierrez-Jurado et al., 2013; Hinckley et al., 2012; Desta et al., 2004], while the effects of aspect on vegetation and hydrological partitioning in perennial streams at the catchment scale remains less understood [Newman et al., 2006]. Understanding the feedbacks between water and vegetation in mountainous semiarid catchments can help improve climate change predictions and associated hydrologic and ecologic shifts [Newman et al., 2006; Molotch et al., 2009; Vivoni et al., 2012].

Furthermore, field studies along landscape gradients may provide insight on the relationship among topography, vegetation, and water [Kelly and Goulden, 2008; Newman et al., 2006;

Chorover et al., 2011]. This comparison of ecosystems across environmental gradients can inform us how climate shapes ecosystems over long time scales, which in turn may help in predicting future changes in climate and land cover [Anderson-Teixeira et al., 2011; Chorover et al., 2011].

The Santa Catalina Mountains-Jemez River Basin (SCM-JRB) critical zone observatory

(CZO) has established a research site in northern New Mexico within the Valles Caldera National

Preserve (VCNP) to study critical zone development and processes along an energy gradient created by differences in terrain aspect [Chorover et al., 2011] (http://criticalzone.org/catalinajemez/). It has been posited by researchers of this observatory that water availability and reduced carbon compounds resulting from primary production are the dominant energy fluxes that drive critical zone development and function [Chorover et al., 2011; Rasmussen et al., 2011].

Vegetation and water availability are primary controls on ecosystems structure [Brooks and

Vivoni, 2008], therefore, it is important to study and understand their role within the critical zone.

The objective of this study is to investigate how terrain aspect influences vegetation structure, the dynamics of hydrological partitioning, and vegetation greening in three high elevation semi-arid

85 catchments using direct and remote sensing observations. A uniform geology and relief around

Redondo Peak, located in the center of the VCNP, make this site an ideal location to empirically study how topographically controlled microclimate [Lyon et al., 2008; Chorover et al., 2011] influences the integrated vegetation and hydrological response.

2.0 METHODOLOGY

2.1 Study area

The Valles Caldera National Preserve (VCNP) (35°50’-36°00’ N; 106°24’-106°37’W) encompasses 359 km

2

and is located within the upper part of the Jemez River Basin, a tributary of the Rio Grande (Figure 1a; 1b). The VCNP encloses a nearly circular caldera that is approximately

21 km in diameter and was formed by the collapse of a magma chamber, around 1.25 Ma. After the caldera’s formation, there were subsequent episodes of volcanism, resulting in the creation of a number of internal domes, including Redondo Peak (3435 m.a.sl) [Wolff et al., 2011]. The geology of Redondo Peak is dominated by densely welded Bandelier Tuff (Tshrige Member) and associate rhyolite/rhyodacite rocks [Goff et al., 2006; Wolff et al., 2011]. Redondo Peak is located approximately in the center of the caldera and headwater streams flow along the different sides of the mountain (Figure 1c). This natural drainage configuration has created catchments with a similar geology [Goff et al., 2006] and relief, but with different terrain aspect affecting wind exposure, radiation, and snow cover duration [Lyon et al., 2008].

The climate is semi-arid, continental and highly variable [Sheppard et al., 2002]. The area is characterized by warm summers and cold winters. The long-term monitoring SNOwpack

TELemetry (SNOTEL) station Quemazon (35°55’ N; 106°24’ W; 2896 m.a.s.l), which is located

5 km to the northeast from Redondo Peak, has been recording climatic variables since 1989. The

86 mean temperature from 1989 through 2012 was -0.87 °C during winter (October – April) and 11.2

°C during summer months (May-September). The minimum temperature occurs between

December-January and the maximum temperature between June-July, with recorded extreme values of -20°C and 25°C, respectively. The study area has a bimodal precipitation pattern; it accumulates an annual winter snowpack between December and April, while the North American monsoon brings high intensity, short duration rainfall between July and September. The annual precipitation at the Quemazon station from 1989 through 2012 ranged from 322 and 1021 mm, with a mean value of 708 mm. Fifty percent of the total annual precipitation occurs during winter

(October-April) and 80% of the winter annual precipitation falls as snow. Most winter precipitation is caused by frontal activity from the west due to Pacific Ocean storms. The summer monsoon originates by weak southeasterly flow from the Gulf of Mexico and short high intensity surges from the Gulf of California and the eastern Pacific Ocean [Sheppard et al., 2002].

Based on information available at the USDA National Cooperative Soil Survey

( http://websoilsurvey.nrcs.usda.gov/app/HomePage.htm

), the soil around Redondo Peak is mostly derived from tuff as colluvium material. The upper elevations have soils with a coarse sandy loam and cobbly loam texture. In the eastern and lower part of Redondo, loamy soils and cobbly coarse sandy loam textures predominate. The northern lower part of Redondo is characterized by fine textures and slightly decomposed plant material.

The prevalent vegetation types around Redondo Peak are spruce-fir forest, mixed conifer forest, woodland, and forest meadow. Spruce-fir forest extends in elevation typically between 2900 and 3430 m. Engelmann spruce (Picea engelmannii) and corkbark fir (A. lasiocarpa var. arizonica

) occupy the highest elevations and north-facing slopes above 2700m. Mixed conifer forest occurs

87 in elevation between 2540 to 3050 m. Forest meadow land cover elevation ranges from 2560 to

3175 m and it is composed by grasslands. The primary grasslands components are Thurber fescue-

Kentucky bluegrass and Arizona fescue-Kentucky bluegrass [Muldavin and Tonne, 2003]. No forest disturbances by logging and fire took place around Redondo Peak in the last four decades preceding and including this study [Anderson-Teixeira et al., 2011; Balmat, 2004].

2.2 Catchment characteristics

This investigation focuses on three catchments with perennial streams: La Jara (LJ),

History Grove (HG) and Upper Jaramillo (UJ) (Figure 1c). The three catchments together cover approximately an area of 10 km

2

. Topographic characteristics of the catchments were defined using a 722 km

2

airborne LiDAR flight coverage (Figure 1c). LiDAR data acquisition (July

2010) and processing were completed by the National Center for Airborne Laser Mapping

(NCALM). The LiDAR coverage has on average a 9.68 point density per m

2

and 30 cm vertical and 20 cm horizontal RMSE resolution. Catchment and stream network delineation and terrain attributes such as area, slope and predominant aspect were defined using the catchment delineation procedures from the spatial analysis tools in ArcGIS 10.1. The amount of annual solar radiation in the catchments was quantified using the solar radiation analysis tools in

ArcGIS 10.1.

2.3 Vegetation characteristics

Land cover classes within the three catchments were estimated using a vegetation map of the VCNP developed by Muldavin et al. [2006], which is based on an analysis of aerial photography from the year 2000, Landsat satellite images from 1999 and 2001, and ground control points. Indicators of vegetation structure were defined using the LiDAR data with 1.0 m

88 resolution. Mean tree height, diameter at breast height (DBH), fractional canopy cover, and leaf area index (LAI) products were created using a regression analysis [Andersen et al., 2005]. This methodology extracts a subset of LiDAR points that were normalized by subtracting the ground points from a digital elevation model (DEM). A height profile was calculated on the normalized points according to the following groups: minimum, maximum, mean, standard deviations, coefficient of variation and the percentiles (1 st

, 5 th

, 10 th

, 25 th

, 50 th

, 75 th

, 90 th

, 95 th

, 99 th

). The LiDAR derived height predictors were compared against field measurements by stepwise regression

[Andersen et al., 2005]. The best models were selected and applied to the entire Redondo Peak area. This process was accomplished by iterating through each pixel of the product grid, extracting

LiDAR points that fall within that pixel and calculating the pixel value using the relation found in the previously mentioned analysis.

Fractional canopy cover (CC) was determined by analyzing the canopy height model

(CHM) at 1 m spatial resolution. The value of the canopy cover pixel was calculated as the ratio of CHM pixels that have a value above a threshold to the total number of extracted pixels from the

CHM [Lucas et al., 2006]. The difference between trees and shrubs was defined by a height threshold of 1.5 m. Each pixel in the canopy cover grid was iterated and CHM pixel values were extracted when they fall within the canopy cover pixel. Leaf area index (LAI) was determined using the LiDAR vegetation points after being normalized by the DEM. Each pixel in the LAI grid was iterated and LiDAR points that fell within the pixel were extracted. An average scan angle was calculated using the extracted LiDAR points as indicated in the following equation:

ang

ο€½

οƒ₯

i n

ο€½

1

angle i n

(1)

89 where ang is the average scan angle, n is the number of extracted points and π‘Žπ‘›π‘”π‘™π‘’(𝑖) is the scan angle for a single extracted point i. The gap fraction (

GF) was calculated using the following equation:

GF

ο€½

n ground n

(2)

Where n

ground

is the number of extracted points for z values smaller than 1.5 m (equivalent to the height of a hemispherical camera) and n is the total number of extracted points. Finally, the LAI was calculated using the following equation:

LAI

ο€½ ο€­ cos(

ang

)

ο‚΄ ln(

GF

)

k

(3) where π‘˜ is the extinction coefficient and assumed equal to 0.5 [Morsdorf et al. 2006; Richardson

et al., 2009]. The vegetation products were extracted at the catchment scale and mean and standard deviation values were computed. Differences in vegetation structure in the catchments were tested with both a Mann-Whitney U test and pair t test.

2.4 Water partitioning and vegetation greening

The water partitioning and vegetation greening analysis for the three catchments around

Redondo Peak was developed using a three step process. First, the Horton index as a metric of catchment scale water fluxes and vegetation water use was calculated for the water years 2008 to

2012 as discharge data was available for this period only. Second, vegetation greening was analyzed using the available MODIS NDVI data from 2000 to 2012. As water partitioning information was available for the years 2008 onwards only, single and multivariate linear

90 regression analysis were estimated with the objective of finding a statistical model to extend

Horton index records back to 2000. Finally, once the best statistical model was selected, Horton index was extrapolated back to 2000 resulting in a dataset that matches the 13 year records of

NDVI.

2.4.1 Water partitioning

2.4. 1.1 Precipitation

Five automatic weather stations have been in operation since 2004 as part of the VCNP monitoring network ( http://www.wrcc.dri.edu/vallescaldera/ ). The five stations are spread out within the VCNP and are located at elevations between 2598m and 3232m (Table S1). The stations record weather conditions, including atmospheric pressure, temperature, wind speed, relative humidity, precipitation and solar radiation. Each station is controlled by a Campbell

Scientific CR1000 data logger and information from the sensors are summarized in 10 minutes intervals. In addition, an eddy covariance flux tower has been operating within the study area since 2005 and is part of New Mexico Upland Flux Network

( http://biology.unm.edu/Litvak/Upland.html

). The tower was set up by the NSF Science and

Technology Center for the Sustainability of Semi-arid Hydrology (SAHRA). The tower site

(VCM; 35° 53.3’ N;106° 31.9’W; 3003 m.a.s.l) is located in the upper part of Redondo Peak surrounded by Douglas and white fir, blue spruce and southwestern white pine (Figure 1c). The eddy covariance derived surface fluxes of water and energy were measured at 10 Hz using a 3axis sonic anemometer (CSAT-3, Campbell Scientific, Logan, UT, USA) and an open path infrared gas analyzer (Li-7500, LiCor Biosciences; Lincoln, NE, USA). 1998). At this site, precipitation is measured by way of a TE525 Texas electronics 6” tipping bucket rain gauge.

Precipitation was distributed at the catchment scale using an elevation lapse rate during the

winter months [Gustafson et al. 2010] and an average precipitation from all the stations during

91 the summer. The lapse rate was evaluated with the observed annual precipitation at the eddy covariance VCM flux tower with an error not larger than 6% on an annual time scale.

2.4.1.2 Streamflow

From 2008 to 2012, streamflow was monitored at the outlets of each catchment (LJ -

2702 m; HG - 2681 m; and UJ - 2723 m) using pressure measurements recorded every 15 minutes with Onset HOBO U20 and In Situ Level TROLL 500 15 psig pressure transducers in

45° trapezoidal flumes. Since the HOBO U20 instruments are non-vented to the atmosphere, the pressure from these sensors was corrected with atmospheric pressure recorded at the flume sites.

The corrected pressures were then transformed into water heights and streamflow by way of standard equations for 45° trapezoidal flumes. The automatic readings by the pressure transducers were validated by manual measurements taken at the flumes. Daily discharge was then calculated and normalized by basin area.

2.4.1.3 Hydrological partitioning and Horton index

Catchment scale hydrological partitioning was calculated following the Horton index methodology from 2008-2012 [Troch et al., 2009; Brooks et al., 2011]. The Horton index is a dimensionless metric of water partitioning at the catchment scale that reflects the influences of both topography and vegetation and ranges from 0 to 1 [Voepel et al., 2011]. The Horton index is defined as the ratio between V and W, and it represents a measure of catchment-scale vegetation water use [Horton, 1933; Troch et al., 2009] (equation 4).

92

𝑉

𝐻𝐼 =

π‘Š

=

𝑃 − 𝑄

(4)

𝑃 − 𝑆

Water partitioning at the catchment scale was based on the analysis of precipitation (P) and discharge (Q). Precipitation was partitioned between storm runoff (S) and catchment wetting

(W). W represents the fraction of precipitation potentially available for vegetation and was calculated as precipitation minus storm runoff (P-S). S was computed by way of a baseflow separation method. Daily stream flow records were partitioned into base flow (U) and storm flow

(S) using a one-parameter recursive filter [Lyne and Hollick, 1979; Arnold and Allen, 1999;

Eckhardt, 2005]:

π‘ˆ π‘˜

= π‘Žπ‘ˆ π‘˜−1

+

1 − π‘Ž

2

(𝑄 π‘˜

− 𝑄 π‘˜−1

) (5)

π‘ˆ π‘˜

≤ 𝑄 π‘˜

Where a is a filter parameter set to 0.925. This filter was passed twice, backward and forward in time, to improve the partitioning of base flow and storm runoff at the beginning of the time series. Then, daily values of streamflow, baseflow and storm runoff were integrated to an annual time scales. Catchment wetting was further partitioned between vaporization and a base flow component. Vaporization (V) at the catchment scale was calculated as precipitation minus discharge (P-Q). The hydrological partitioning analysis was quantified on an annual scale to minimize changes in catchment water storage. Troch et al. [2009] showed that the methodology to separate baseflow from storm runoff plays little influence in the computation of the Horton index. For further details about the Horton index development see Troch et al. [2009].

2.4.2 Vegetation greening

The normalized difference vegetation index (NDVI), a remote sensing measurement of

93 vegetation greenness [Reed et al., 1994], was used as an indicator of vegetation productivity and for the quantification of the interannual and inter-catchment vegetation response. NDVI is a vegetation index independent of any hydrological partitioning variable considered in this study and is widely used for monitoring vegetation conditions, global and regional climate, and modeling global biogeochemical and hydrological processes [Reed et al., 1994; Van Leeuwen et

al., 2006; Hashimoto et al., 2012]. NDVI was extracted from NASA’s Moderate Resolution

Imaging Spectroradiometer (MODIS) Land Products MOD13Q1 and available through the Land

Processes Distributed Active Archive Center (LP DAAC) ( https://lpdaac.usgs.gov/products/ ).

The global MOD13Q1 NDVI data is provided every 16 days at a 250-meter spatial resolution.

An average NDVI value was calculated for each catchment based on each individual NDVI image. In the NDVI images, pixels with clouds and snow were not considered for the analysis and only pixels cataloged in the quality layer with good reliability were included.

A multi-temporal NDVI profile (catchment scale NDVI versus time) was created for each catchment from 2000 to 2012 [Reed et al., 1994]. Two metrics were selected for the quantification of the vegetation response and length of the growing season [Reed et al., 1994]. A time integrated NDVI response per catchment was selected as an indicator of vegetation productivity [Reed et al., 1994]. The time integrated NDVI response was defined as the area under the NDVI curve between the end of the snowmelt season and end of the growing season, between April and September each year (Figure S1). The mean annual NDVI of the three catchments was close to 0.6 and NDVI values above the mean were observed every year during

94 the growing season between April and September. Therefore 0.6 was selected as a signal of the onset of an increase in photosynthetic activity [Reed et al., 1994]. The number of days when the catchment scale NDVI value exceeds 0.6 was used as indicator of growing season length.

Statistically significant differences between the NDVI time series from the three catchments were examined with the Mann-Whitney U test.

2.4.3 Climate controls on Horton index

Single and multivariate regression were used to investigate the climate controls on

Horton index using data from 2008 to 2012. The climate variables from the Quemazon SNOTEL considered in the regression analyses were total annual precipitation, summer and winter precipitation, maximum SWE, timing of initial snow cover, timing of snow melt, duration of snow accumulation and snow cover, total annual, winter and summer temperatures. Previous research has demonstrated that Horton index values vary predictably in a catchment due to climate and landscape and therefore can be predicted based on these variables [Troch et al.,

2009; Voepel et al., 2011).Therefore using the best fitted regression model with climate variables, Horton index values were extrapolated back to the year 2000. Climate controls using the above mentioned variables were also tested against NDVI data.

3.0 RESULTS

3.1 Catchment characteristics

The three catchments are located above 2680 m in elevation. LJ is the highest elevation catchment with the largest area and steepest slope (Table 1). HG and UJ share similar elevations, catchment size, and slope. The catchments are characterized by differences in terrain

aspect, and 58% of UJ area has a predominant northern aspect. In contrast, in HG and LJ 59%

95 and 70 % of their surface area drain to the east (Figure 1c). Aspect differences are reflected in the total annual incoming solar radiation, where the east facing catchments (LJ and HG) received approximately 30% more solar energy than the north facing catchment (UJ) (Table 1).

3.2 Vegetation classification and structure

According to land use classification by Muldavin et al. [2006] the catchments are predominantly covered by evergreen forest, with an extension of 81% in LJ, 76% in HG and

74% in UJ (Table 2). Mean tree height (m) is larger in LJ (15.0 m), followed by HG (14.7 m) and UJ (12.2 m). Similarly, DBH (cm) is larger in trees located in the eastern catchments (LJ and HG) with a mean value of 31.3 cm versus 27.4 cm for UJ. In the same way, fractional canopy cover is larger in the eastern catchments with values of 0.71 and 0.66 (LJ and HG) and

0.57 in the northern catchment (UJ). Mean LAI is 1.41, 1.26 and 1.03 for LJ, HG and UJ, respectively (Figure 2). The land cover classification and the vegetation structure indicators derived from LiDAR indicate that the north facing catchment (UJ) has less above ground biomass than the southeastern catchments (LJ and HG) (Figure 2).

3.3 Water partitioning

3.3.1 Precipitation

During the study period (2008-2012), the three catchments received varying amounts of annual precipitation. The years 2008, 2009, and 2012 registered catchment scale precipitation values above the 31-years mean annual precipitation at the Quemazon SNOTEL station (708 mm). In contrast, precipitation during 2009 and 2010 was below the mean annual precipitation

96 recorded at the Quemazon SNOTEL station. Precipitation in LJ was the largest, followed by HG and UJ with nearly similar annual precipitation values (Figure 3a). LJ is characterized by a highest terrain elevation and received approximately 50 mm more precipitation than the other catchments during the winter season (Figure 3b). During the summer monsoon season, the three catchments received similar precipitation (Figure 3c).

3.3.2 Streamflow response

The study catchments are snowmelt dominated with hydrographs characterized by a significant increase in runoff during spring, reaching a peak at the end of March and beginning of April, and followed by a steady discharge recession that continues until the following spring snowmelt onset. Sporadic increases in discharge occur between July and September as a result of heavy rainfall events during the monsoon season. For the rest of the year, daily runoff for the three catchments is characterized by low flow conditions, the 75th percentile do not exceed a discharge of 0.23 mm/day in the three catchments (Figure S2). During the study period, total annual discharge was strongly correlated to max SWE and the duration of snow on the ground measured at the Quemazon SNOTEL station (R

2

>0.95; p<0.05) (Table S2). Other snow parameters such as timing of initial snow cover, peak of SWE, timing of melt, average air winter temperature [Harpold et al., 2012] were also tested as controls on discharge, however they did not show strong relationships with catchment scale discharge in the multivariate regression analysis. Indicators of magnitude (peak flow and mean discharge) and timing of discharge (day of the water year = DOWY peak flow occurs) show that discharge measured in the north facing catchment (UJ) was the largest and least variable relative to the discharge in the other sites (LJ and HG) (Table 3). In average mean total annual discharge in UJ is 63% and 39% higher than LJ

97 and HG, respectively (Figure 3d). During the five years, discharge on average in UJ represented

15%, in HG 11% and LJ 8% of total annual precipitation. With the exception of the wettest year

(2010), peak flow in the northern catchment (UJ) occurred later than in the eastern catchments

(LJ and HG) (Table 3). Runoff ratio, computed annually as the ratio between discharge to precipitation, was consistently larger in UJ during the five years of analysis. Runoff ratios ranged between 0.08 for the eastern catchments and 0.14 for the northern catchment and they were strongly related to maximum SWE (Figure 3e).

3.3.3 Base flow and storm runoff

Base flow was the largest and dominant component of discharge representing over 90% of the total discharge in the three catchments. However, the north facing catchment (UJ) showed consistently the largest and least variable base flow during the study period. Storm runoff represented less than 10% of annual discharge in all the catchments and once again the north facing catchment (UJ) consistently had larger and less variable storm runoff compared to the other two catchments. The ratio between storm runoff to base flow for all the catchments was more similar during wet years, while the largest differences in this ratio were observed during the relatively dry years of 2011 and 2012 (Figure 3f). Base flow, storm runoff, and discharge were the water partitioning components with the largest variability in all of the three sites.

3.3.4 Vaporization and wetting

The fraction of precipitation (P) that infiltrates and it is potentially availably for vegetation (W) was very high and above 98% of P in all the catchments. A large portion of W was lost by vaporization (Figure 3g). On average vaporization represented 92% from the total

98 annual precipitation in LJ, 91% in HG and 85% in UJ. Water loss by vaporization was smaller in the north facing catchment (UJ) than in the eastern catchments (LJ and HG). A summary of all the water partitioning components during the study period for the three catchments can be found in table S3.

3.4 Vegetation greening

The catchment’s NDVI time series from 2000 to 2012 were statistically different with p values < 0.02. Mean NDVI at the catchment scale for the 12 years was 0.59, 0.58 and 0.56 for LJ,

HG and UJ, respectively and the standard deviation was similar for the three catchments and equal to 0.1. The time integrated NDVI times series and length of the growing season were statistically significantly different between the catchments with p values <0.02. The north facing catchment

(UJ) had the smallest vegetation response indicated by the time integrated NDVI metric (Ε«=5.98,

σ=0.32, C.V.=0.05) followed by HG (Ε«=6.28, σ=0.35, C.V.=0.06) and LJ (Ε«=6.58, σ=0.35,

C.V.=0.34) (Figure 4). In addition, the indicator of growing season length was statistically significantly different among the three catchments (p < 0.05) (Figure 4). Upper Jaramillo had the shortest and more variable growing season (Ε«=76.3, σ=40.9, C.V.=0.54), while LJ had the longest growing season (Ε«=156.4, σ=30.7, C.V.=0.2). The indicator of growing season length in HG had values in between the other two catchments (Ε«=113.2, σ=36.6, C.V.=0.3)

3.5 Horton index

Climatic variables such as maximum SWE and duration of snow accumulation and snow cover were the strongest predictors of Horton index around Redondo Peak (Table 5). Based on these strong multiple regression models (Table 5), Horton index records were back calculated in

99 time to the year 2000 with a resulting mean value of 0.86 for the north catchment (UJ) and 0.92 for the eastern catchments (LJ and HG). The Horton index was larger in the east facing catchments (LJ and HG), suggesting that vegetation uses more of the available water in these two catchments than in the northern site (UJ). LJ had the largest and less variable Horton index ( π‘₯Μ… =

0.92; 𝜎 = 0.03; 𝐢. 𝑉. =0.03) whereas UJ had the lowest and more variable Horton index (π‘₯Μ… =

0.86; 𝜎 = 0.05; 𝐢. 𝑉. =0.06). The inter-catchment variability of the Horton index due to aspect differences can be larger than the interannual Horton index variability by climate in the individual catchments (e.g WY2005; Table 5). During extreme wet years e.g. 2001 and 2005 the variability in Horton index among the catchments increased indicating larger differences in water partitioning among catchments because of aspect differences. In wet years, aspect drive differences among catchments in water partitioning than are more important to the effects of climate in the individual catchments. In contrast, during dry years e.g. 2002 and 2006 the difference in Horton index among the catchments decreases (Table 5).

Climatic variables recorded at the Quemazon SNOTEL station and Horton index were tested as controls of vegetation greening between 2000 and 2012. Horton index was the strongest predictor of vegetation greening for the three study catchments (R

2

>0.48; p<0.05). Annual P and maximum SWE did also show a statistical linear correlation with vegetation greening. Variables such as air winter and summer temperatures did not show statistical significant linear correlation with NDVI (Figure 5).

4.0 DISCUSSION

A conceptual model and summary of vegetation structure, water partitioning and NDVI response in contrasting aspects around Redondo Peak are presented in Figure 6. The results of

100 this investigation indicate that the catchment with a predominant north facing aspect (UJ) received less solar radiation and contains less biomass than the two eastern catchments (LJ and

HG). Furthermore, UJ had more surface water available as indicated by a larger base flow and discharge due to smaller vaporization fluxes and a smaller Horton index. Vegetation greening and length of growing season were consequently smaller in the northern catchment. In contrast, the eastern catchments had smaller surface water availability indicated by annual base flow and discharge fluxes as they received more solar radiation, have larger vaporization and vegetation water consumption. Similarly NDVI response and length of growing season was longer in the two eastern catchments.

4.1 Hydrological partitioning

During the study period, snowmelt was the dominant process controlling runoff generation around Redondo Peak (Figure S2; table S2). The peak stream discharge recorded at the catchments flumes during the five years is related to the winter snowpack conditions, similar to findings by Gottfried et al. [2002] and DaqingYang et al. [2009]. When compared to the years

2008 through 2010, the drier 2011 winter resulted in a reduced snowmelt discharge (Table 3) similar to findings between wet and dry years at other high elevation regions [Williams et al.,

2002]. An analysis of water stable isotopes also indicated that snowmelt is the major contributor of water to the landscape around Redondo Peak [Broxton et al., 2009; Zapata-Rios et al., in review]. The estimated runoff ratios in this study were similar to runoff ratios of 0.03 and 0.11 in an 870m

2

runoff experiment in a ponderosa pine forest on the Pajarito plateau located in close proximity to the study area [Wilcox et al., 1997]. Base flow was the largest and predominant component of discharge in all the catchments, similar to findings by Liu et al. [2008], who found

101 that overland flow was not significant in streamflow generation in two streams around Redondo

Peak based on a geochemical analysis. Moreover, base cations concentration of first order streams draining Redondo show a chemostatic behavior, suggesting a fairly constant source of water supply from a well-mixed reservoir [Porter, 2012; Perdrial et al., 2014].

Annual vaporization (V) represented over 85% of annual precipitation and the largest water loss in the catchments. Similar estimations of annual V have been published for other semi-arid sites. For instance, Baker [1982] estimated in a ponderosa pine forest in north central

Arizona an annual average ET of 506 mm representing 78% of total annual precipitation. In a nearby ponderosa pine forest close to Redondo Peak, Brandes and Wilcox [2000] estimated an average ET during the growing season of 476 mm during the years 1993 through 1995, representing 95% of annual precipitation. Water loss by vaporization was smaller in the north facing catchment (UJ) than in the eastern catchments (LJ and HG). This lower rate of vaporization in UJ is consistent with what would be expected for an area that receives less solar radiation and has smaller biomass than the other (eastern) catchments included in this study

(Table 1; figure 3f). Data from the VCM flux tower provide further evidence of precipitation partitioning into vaporization and the contribution of winter precipitation to water availability.

Precipitation (P), potential evapotranspiration (PET) and vaporization (V) recorded at this station

(Figure S3) showed significantly higher water availability during the winter due to more precipitation than vaporization demand, whereas during the summer season, the difference between precipitation and vaporization decreased. Summer precipitation did not cover vaporization water demand in 2008, 2009 and 2010. In 2007 and 2011, the contribution of summer precipitation was slightly higher than vaporization.

102

In contrast to the east-facing catchments, the north-facing catchment was characterized by larger surface water availability as a consequence of smaller vaporization and water use by vegetation. More water availability at this site is consistent with observations of large mineral weathering rates and longer water residence times along north facing slopes [Zapata-Rios et al., in review; Broxton et al., 2009]. Likewise, Hinckley et al.[2012] found different hydrologic responses in hillslopes with contrasting aspect, where their results indicate that north facing slopes store water more effectively in the near surface and have longer residence times in contrast to south facing slopes, and more water infiltrates into the ground which may pass quickly to deeper weathered bedrock.

4.2 Vegetation water use

Vegetation in the east-facing catchment uses more of the available water than the northfacing catchments, as indicated by a higher and less variable Horton index. This finding is consistent with Troch et al.[2009] who found that semi-arid catchments have a higher and less variable Horton index than more humid catchments. The Horton index increases as a catchment becomes drier and get closer to 1 in drier regions and during dry years [Troch et al., 2009].

Huxman et al. [2004] found that the average rain-use efficiency (RUE), estimated as the ratio of aboveground net primary production to annual precipitation, decreases as precipitation increases.

In contrast, during dry years REU converges to a common maximum RUE similar to the drier regions, regardless of biome type. It has been demonstrated that vegetation of a region uses the largest proportion of the water available (W) and therefore the Horton index presents little variability from year to year despite substantial variability in precipitation [Horton, 1933; Troch

et al., 2009]. Southern aspects are associated with greater insolation, increased potential

103 evapotranspiration and in general drier conditions [Wilkinson and Humphreys, 2006]. A higher

Horton index along eastern catchments is consistent with the large water availability of plants

(W) and drier landscape conditions in areas with greater insolation. More water stressed trees are more likely found on south facing slopes [Safranyik and Carroll, 2006] and plants are using more of its available water on dry landscape positions according to our study.

4.3 Vegetation structure and NDVI response

Frequently larger biomass has been found in all forest types of the northern hemisphere along north facing slopes [Sharma et al., 2011]. For instance, at lower elevations in northern

New Mexico, vegetation in north-facing slopes consists of more mesic plant species in contrast to more xeric species in the south-facing slopes [Gutierrez-Jurado et al., 2006]. In this study we found smaller biomass and smaller NDVI values along the north facing catchment at high elevations (Table 2, Figure 2). A carbon balance study around Redondo Peak has also quantified smaller above ground biomass in the north facing catchment [Perdrial et al., 2013; Perdrial et al., in prep]. The catchment scale biomass observations in this study are in agreement with results from modeling and empirical studies that have indicated that mountain forest productivity are sensitive to energy limitations [Christensen et al., 2008; Tague et al., 2009; Trujillo et al., 2012;

Anderson-Texeira et al., 2011]. Energy limitations in forest productivity have been observed at colder sites along high elevations in northern New Mexico and Redondo Peak [Anderson-

Texeira et al., 2011]. The north facing catchment where a smaller biomass was quantified receives a reduced solar loading compared to the eastern sites (Table 1). Similarly, sap flux sensors installed in dominant spruce and fir trees species located in slopes with contrasting aspect in Redondo Peak at 3000 m of elevation showed radiation as a main control of sap flux

104 rate during the summer, whereas air and soil temperatures were the dominant controls throughout the rest of the year [Mitra and Papuga, 2012]. Additionally, the eastern catchments showed a longer growing season that increases the opportunity of carbon uptake by trees and accumulation of biomass as indicated in past research [Groffman et al., 2012; Rochefort et al., 1994; White et

al., 1999). Analogous biomass observations were reported in a study across an oak dominated temperate watershed in Pennsylvania, where trees on the south aspect stored more carbon per year than trees on the northern aspect [Smith, 2013]. The observations on vegetation biomass in different slopes around Redondo Peak are in line with previous research and indicate that energy and a short growing season length can limit forest productivity at the local scale in high elevation ecosystems [Rochefort et al., 1994].

In high elevation regions snowmelt controls the timing and magnitude of both runoff and soil moisture, recharging subsurface water pools and providing water for plants during the growing season [Molotch et al., 2009; Vivoni et al., 2008]. Molotch et al. [2009] found in a field plot investigation that the beginning of the growing season and the initiation of snowmelt infiltration lag behind each other by a few days. These processes occur in Redondo every year between the middle of March until the middle of April and their timing depends on the amount of snowpack and air temperature during the winter season [Molotch et al., 2009]. Peak soil moisture occurs within few days of snow disappearance and vegetation responds rapidly to water availability [Molotch et al., 2009]. The NDVI time series at the catchment scale in this study indicate similar response of vegetation with increasing NDVI values between the middle of

March and middle of April, where the timing of peak NDVI occurs before the monsoon season begins (Figure S1).

According to our flux tower data (Figure S3) and the observations by Vivoni et al.

[2008], soil moisture in the upper layers of the soil disappears around June at the beginning of the summer and get replenished weeks later with the onset of the monsoon rains, when vegetation again absorbs a large fraction of rainfall and soil moisture derived from the North

105

American monsoon season. The NDVI profiles showed the same dynamics at the catchment scale where NDVI increases at the end of the snowmelt period followed by a short drop in NDVI during the summer at beginning of June. NDVI increases again during the North American monsoon season, a pattern also observed by Molotch et al. [2009] and Vivoni et al. [2008] at small plot scales. In addition, NDVI time series indicated a longer growing season along the eastern catchments, consistent with findings by Lyon et al. [2008], who found that snowmelt occurred early and more readily along terrains with larger solar radiation. Furthermore, Lyon et

al. [2008] found that aspect exerts a strong control on the timing of snowmelt and water availability. For instance, their observations indicated that snowmelt occurs later and first-order streams get dry slower along north facing slopes.

4.4 Water partitioning and vegetation interactions

Snowpack conditions strongly control Horton index that in turn influences vegetation greening. This study indicates that the partitioning of water in the landscape is critical for the annual vegetation productivity. Similarly, Trujillo et al. [2012] found that snowpack thickness and snow duration can lead to changes in annual productivity and ultimately changes in forest stand structure and long term water use. Vegetation water use and productivity are influenced by snow accumulation and melt, particularly in water limited environments [Tague and Dugger,

2010; Molotch et al., 2009]. Therefore, future ecohydrological responses in high elevation

106 ecosystems may depend on snowpack processes and changes in snow distribution, snowmelt, soil moisture, soil temperature [Molotch et al., 2009]. As previous studies have suggested, this study corroborates that a decrease in snow accumulation and melt water might reduce water availability and forest productivity [Molotch et al., 2009; Trujillo et al., 2012].

4.5 Implications for the critical zone

The critical zone is the heterogeneous surficial layer of the planet that extends from the canopy to the base of the groundwater, and where complex interactions between rocks, soil, water, atmosphere, and biota take place [National Research Council, 2001]. It remains highly uncertain how landscape position and climate variability shape the critical zone structure and function [Chorover et al., 2011]. Both water availability (E ppt

) and carbon compounds derived from net primary productivity (E bio

) have been recognized as fundamental sources of effective energy and mass transfer (EEMT) to the critical zone [Rasmussen and Tabor, 2007; Rasmussen

et al., 2011; Chorover et al., 2011]. Previous research quantified higher EEMT along north facing slopes around Redondo Peak [Chorover et al., 2011; Zapata-Rios et al., in review]. More water availability at the north facing catchment found in this study is consistent with previous studies that found higher EEMT along the north facing slopes in Redondo Peak. A higher EEMT in the north facing slope of Redondo Peak is consistent with previously reported higher carbon storage in soil at cooler and wetter sites [Perdrial et al., in prep]. At these sites, trees have lower respiration rates relative to their carbon uptake. In addition higher dissolved organic carbon concentrations has been measured in stream waters and springs in north facing slopes [Perdrial et

al., 2014] indicating that carbon compounds as an a energy source to the CZ are also larger in north facing catchment in agreement with the larger EEMT estimations.

107

4.6 Future studies

Additional factors beyond water and energy can affect the productivity of high elevation ecosystems such as soil, nutrients, light, and roots distribution that need further ecological investigation. For example, details about roots depth and their distribution in the subsurface are unknown around Redondo Peak and can provide an important source of information about plant water use and water pools that provide moisture during the growing season.

On May 31, 2013, the Thompson ridge fire affected the southwestern part of Valles Caldera and a large area of Redondo Peak. The fire consumed a total of 97 km

2

of forest. The three catchments included in this study (LJ, HG and UJ) had a fire severity ranging from moderate to high covering 65%, 81% and 34% of the total catchment area, respectively. The results from this study documents pre-fire water partitioning - vegetation interactions in high elevation forest in northern New Mexico. Having the baseline data presented in this study, presents a unique opportunity for futures studies to further understand vegetation water use and their influence on water partitioning in high elevation semi-arid areas in a post-fire scenario.

5.0 SUMMARY

In Redondo Peak in northern New Mexico we studied three adjacent first order catchments that share similar physical characteristics, but drain different aspects, allowing for an empirical study of how topographically controlled microclimate influences hydrological and vegetation response. Results from this investigation provide evidence that aspect influences the magnitude of water partitioning fluxes and vegetation response at the catchment scale. For instance, differences in water partitioning fluxes during wet years among catchments can be larger than the water

108 partitioning fluxes variability induced by climate variability only at the individual catchment scale.

In contrast, we found that during dry years the catchments behave more similar with one another.

Moreover, this study suggests that terrain aspect is a landscape characteristic that can exacerbate the availability of plant limiting resources such as water and energy. The main findings of this study include a significant difference in forest cover and biomass between the northern and eastern catchment with the northern catchment showcasing a smaller forest cover and biomass. In the northern catchment, smaller vaporization, wetting and plant water consumption was observed and consequently there was more water available as surface runoff and baseflow. In contrast, aboveground biomass was more abundant in the eastern catchments which received a larger solar loading than the northern catchment. In the eastern site of Redondo, vaporization and vegetation water consumption was larger and there was less water available as surface runoff and baseflow.

Vegetation greening indicates larger and longer growing season along the eastern catchments.

Water partitioning around Redondo Peak is highly related to snowpack conditions such as max

SWE and duration of snow on the ground that in turn determines annual vegetation greening.

6.0 ACKNOWLEDGEMENTS

This research was supported by the national Critical Zone Observatory program via NSF EAR-

0724958, NSF EAR-0632516 and NSF EAR-0922307. Robert Parmenter from the VCNP provided access to the meteorological data from the VCNP network. We appreciate the field assistance of Scott Compton and Mark Losleben.

109

7.0 REFERENCES

Andersen, H., R. McGaughey, and S. Reutebuch (2005), Estimating forest canopy fuel parameters using LIDAR data, Remote Sens. Environ., 94, 441-449.

Anderson-Teixeira, K. J., J. P. Delong, A. M. Fox, D. A. Brese, and M. E. Litvak (2011),

Differential responses of production and respiration to temperature and moisture drive the carbon balance across a climatic gradient in New Mexico, Global Change Biol., 17, 410-424.

Arnold, J.G., and P.M. Allen (1999), Automated methods for estimating baseflow and ground water recharge from stream flow records, J. Am. Water Resources. Assoc., 35(2), 411-424.

Doi:10.1111/j.1752-1688.1999.tb03599.x.

Baker, Jr., M. (1982), Hydrologic regimes of forested areas in the Beaver Creek watershed,

USDA Forest Service General Technical Report RM-90, 8p. Rocky Mountain forest and Range

Experiment Station, Fort Collins, Colorado.

Balmat, J (2004), Assessment of timber resources and logging history of the Valles Caldera

National Preserve. M.S. thesis. University of Arizona. 105 pp

Brandes, D. and B. Wilcox (2000), Evapotranspiration and soil moisture dynamics on a semiarid ponderosa pine hillslope, J. Am. Water Resour. Assoc., 36, 965-974.

Brooks, P. D. and E. R. Vivoni (2008), Mountain ecohydrology: quantifying the role of vegetation in the water balance of montane catchments, Ecohydrology, 1, 187-192.

110

Brooks, P. D., P. A. Troch, M. Durcik, E. Gallo, and M. Schlegel (2011), Quantifying regional scale ecosystem response to changes in precipitation: Not all rain is created equal, Water Resour.

Res., 47, W00J08.

Broxton, P. D., P. A. Troch, and S. W. Lyon (2009), On the role of aspect to quantify water transit times in small mountainous catchments, Water Resour. Res., 45, W08427.

Casanova, M., I. Messing, and A. Joel (2000), Influence of aspect and slope gradient on hydraulic conductivity measured by tension infiltrometer, Hydrol. Process, 14, 155-164.

Chorover, J. et al. (2011), How Water, Carbon, and Energy Drive Critical Zone Evolution: The

Jemez-Santa Catalina Critical Zone Observatory, Vadose Zone Journal, 10, 884-899.

Christensen, L., C. L. Tague, and J. S. Baron (2008), Spatial patterns of simulated transpiration response to climate variability in a snow dominated mountain ecosystem, Hydrol. Process, 22,

3576-3588.

Desta, F., J. Colbert, J. Rentch, and K. Gottschalk (2004), Aspect induced differences in vegetation, soil, and microclimatic characteristics of an Appalachian watershed, Castanea, 69,

92-108.

Eckhardt, K. (2005), How to construct recursive digital filters for baseflow separation, Hydrol.

Process., 19, 507-515.

Geroy, I. J., M. M. Gribb, H. P. Marshall, D. G. Chandler, S. G. Benner, and J. P. McNamara

(2011), Aspect influences on soil water retention and storage, Hydrol. Process., 25, 3836-3842.

Goff,F., J.N. Gardner, S.L. Reneau, and C.J. Goff (2006), Geologic map of the Redondo Peak

111 quadrangle, Sandoval County, New Mexico: New Mexico Bureau of Geology and Mineral

Resources, Open-file Map Series OFGM-111, scale 1:24,000, http://geoinfo.nmt.edu/publications/maps/geologic/ofgm/ , accessed July 15, 2014.

Gottfried, G,J., D.G. Neary, and P.F. Ffolliott.2002. Snowpack- runoff relationships for midelevation snowpacks on the workman creek watershed of Central Arizona, USDA., RMRS-RP-

33.

Groffman, P. M. et al. (2012), Long-Term Integrated Studies Show Complex and Surprising

Effects of Climate Change in the Northern Hardwood Forest, Bioscience, 62, 1056-1066.

Gustafson, J. R., P. D. Brooks, N. P. Molotch, and W. C. Veatch (2010), Estimating snow sublimation using natural chemical and isotopic tracers across a gradient of solar radiation,

Water Resour. Res., 46, W12511.

Gutierrez-Jurado, H. A., E. R. Vivoni, C. Cikoski, J. B. J. Harrison, R. L. Bras, and E.

Istanbulluoglu (2013), On the observed ecohydrologic dynamics of a semiarid basin with aspectdelimited ecosystems, Water Resour. Res., 49, 8263-8284.

Gutierrez-Jurado, H. A., E. R. Vivoni, J. B. J. Harrison, and H. Guan (2006), Ecohydrology of root zone water fluxes and soil development in complex semiarid rangelands, Hydrol. Process.,

20, 3289-3316.

112

Harpold, A. A., J. A. Biederman, K. Condon, M. Merino, Y. Korgaonkar, T. Nan, L. L. Sloat, M.

Ross, and P. D. Brooks (2014), Changes in snow accumulation and ablation following the Las

Conchas Forest Fire, New Mexico, USA, Ecohydrology, 7, 440-452.

Harpold, A., P. Brooks, S. Rajagopal, I. Heidbuchel, A. Jardine, and C. Stielstra (2012), Changes in snowpack accumulation and ablation in the intermountain west, Water Resour. Res., 48,

W11501.

Hashimoto, H., W. Wang, C. Milesi, M. A. White, S. Ganguly, M. Gamo, R. Hirata, R. B.

Myneni, and R. R. Nemani (2012), Exploring Simple Algorithms for Estimating Gross Primary

Production in Forested Areas from Satellite Data, Remote Sensing, 4, 303-326.

Hinckley et al. (2012), Aspect control of water movement on hillslopes near the rain-snow transition of the Colorado Front Range. Hydrological Processes,24,74-85,

DOI: 10.1002/hyp.9549.

Horton, R. (1933), The role of infiltration in the hydrologic cycle, Transactions-American

Geophysical Union, 14, 446-460.

Huxman, T. et al. (2004), Convergence across biomes to a common rain-use efficiency, Nature,

429, 651-654.

Kelly, A. E. and M. L. Goulden (2008), Rapid shifts in plant distribution with recent climate change, Proc. Natl. Acad. Sci. U. S. A., 105, 11823-11826.

Liu, F., R. C. Bales, M. H. Conklin, and M. E. Conrad (2008), Streamflow generation from snowmelt in semi-arid, seasonally snow-covered, forested catchments, Valles Caldera, New

Mexico, Water Resour. Res., 44, W12443.

113

Lucas, R., N. Cronin, M. Moghaddam, A. Lee, J. Armston, P. Bunting, and C. Witte (2006),

Integration of radar and Landsat-derived foliage projected cover for woody regrowth mapping,

Queensland, Australia, Remote Sens. Environ., 100, 388-406.

Lyon, S. W., P. A. Troch, P. D. Broxton, N. P. Molotch, and P. D. Brooks (2008), Monitoring the timing of snowmelt and the initiation of streamflow using a distributed network of temperature/light sensors, Ecohydrology, 1, 215-224.

McDonnell, J. J. et al. (2007), Moving beyond heterogeneity and process complexity: A new vision for watershed hydrology, Water Resour. Res., 43, W07301.

Mitra B., Papuga S.A. (2012): Toward an improved understanding of the role of transpiration in critical zone dynamics. Abstract #H53I-1650 presented at 2012 Fall Meeting, AGU, San

Francisco, CA, 3-7 Dec.

Molotch, N. P., P. D. Brooks, S. P. Burns, M. Litvak, R. K. Monson, J. R. McConnell, and K.

Musselman (2009), Ecohydrological controls on snowmelt partitioning in mixed-conifer subalpine forests, Ecohydrology, 2, 129-142.

Morsdorf, F., Kotz, B., Meier, E., Itten, K.I., Allgower, B.(2006), Estimation of LAI and fractional cover from small footprint airborne laser scanning data based on gap fraction. Remote

Sensing of Environment 104 (1), 50-61.

114

Musselman, K. N., N. P. Molotch, and P. D. Brooks (2008), Effects of vegetation on snow accumulation and ablation in a mid-latitude sub-alpine forest, Hydrol. Process., 22, 2767-2776.

Muldavin, E. and P. Tonne. (2003) A vegetation survey and preliminary ecological assessment of the valles caldera national preserve, New Mexico. Natural Heritage New Mexico Publication

No. 03-GR-118. P. 272.

Muldavin, E., P. Neville., C. Jackson., T. Neville (2006), A vegetation map of Valles Caldera

National preserve, New Mexico. Final Report for Cooperative Agreement No. 01CRAG0014,

University of New Mexico, Albuquerque.

National Research Council (2001), Basic research opportunities in earth science. Natl. Acad.

Press, Washington, DC.

Newman, B. D., B. P. Wilcox, S. R. Archer, D. D. Breshears, C. N. Dahm, C. J. Duffy, N. G.

McDowell, F. M. Phillips, B. R. Scanlon, and E. R. Vivoni (2006), Ecohydrology of waterlimited environments: A scientific vision, Water Resour. Res., 42, W06302.

Noy-Meir I, (1973), Desert ecosystems: environment and producers. Annual Review of Ecology

and Systematics 4: 51-58.

Lieth H. (1975), In Primary Productivity of the Biosphere, Lieth H, Whittaker RE(eds) Springer:

New York.

Lyne, V., and M. Hollick (1979), Stochastic time-variable rainfall-runoff modelling, in Hydrol.

And Water Resources. Syp., publ.79/10, pp.89-92, Inst. Eng. Austr. Natl. Conf., Perth, Australia.

115

Perdrial J.N., Brooks P.D., Swetnam T., Lohse K.A., Rasmussen C., Harpold A.A., Litvak M.E.,

Broxton P.D., Mitra B., Condon K., Huckle D.M., Vazquez A., Lybrand R.A., Holleran M.,

Orem C.A., Meixner T., Chorover J. (2013): Do fire disturbances account for missing C in snow dominated headwater catchments in NM?. Abstract B33E-0531 presented at 2013 Fall Meeting,

AGU, San Francisco, CA, 9-13 Dec.

Perdrial, J. N. et al. (2014), Stream water carbon controls in seasonally snow-covered mountain catchments: impact of inter-annual variability of water fluxes, catchment aspect and seasonal processes, Biogeochemistry, 118, 273-290.

Porter, C. (2012), Solute inputs to soil and stream waters in a seasonally snow covered mountain catchment determined using Ge/Si,

87

Sr/

86

Sr and major ion chemistry: Valles Caldera, New

Mexico. MS Thesis. University of Arizona, pp 88.

Rasmussen, C. and N. J. Tabor (2007), Applying a quantitative pedogenic energy model across a range of environmental gradients, Soil Sci. Soc. Am. J., 71, 1719-1729.

Rasmussen, C., P. A. Troch, J. Chorover, P. Brooks, J. Pelletier, and T. E. Huxman (2011), An open system framework for integrating critical zone structure and function, Biogeochemistry,

102, 15-29.

Reed, B., J. Brown, D. Vanderzee, T. Lovel and, J. Merchant, and D. Ohlen (1994), Measuring

Phenological Variability from Satellite Imagery, Journal of Vegetation Science, 5, 703-714.

Richardson, J. J., L. M. Moskal, and S. Kim (2009), Modeling approaches to estimate effective leaf area index from aerial discrete-return LIDAR, Agric. For. Meteorol., 149, 1152-1160.

116

Rinehart, A. J., E. R. Vivoni, and P. D. Brooks (2008), Effects of vegetation, albedo, and solar radiation sheltering on the distribution of snow in the Valles Caldera, New Mexico,

Ecohydrology, 1, 253-270.

Rochefort R.M., Little T.L., Woodward A., D. L. Peterson (1994) Changes in sub-alpine tree distribution in western North America: a review of climatic and other causal factors. The

Holocene 4,1, pp 89-100.

Safrayk L., A. L. Carroll (2006), The biology and epidemiology of the mountain pine beetle in lodegpole pine forests. Victoria, BC Canada.

Sharma, C. M., S. Gairola, N. P. Baduni, S. K. Ghildiyal, and S. Suyal (2011), Variation in carbon stocks on different slope aspects in seven major forest types of temperate region of

Garhwal Himalaya, India, J. Biosci., 36, 701-708.

Sheppard P.R., Comrie A.C., Packin G.D., Angersbach K., Hughes M.K. 2002. The climate of the US Southwest. Climate Research 21. DOI: 10.3354/cr021219.

Small, E. E. and J. R. McConnell (2008), Comparison of soil moisture and meteorological controls on pine and spruce transpiration, Ecohydrology, 1, 205-214.

Smith L.A. (2013). Aboveground carbon distribution across a temperate watershed. M.S thesis.

The Pennsylvania State University. 72 pp.

Tague, C., K. Heyn, and L. Christensen (2009), Topographic controls on spatial patterns of conifer transpiration and net primary productivity under climate warming in mountain ecosystems, Ecohydrology, 2, 541-554.

117

Tague C. and A.L. Dugger (2010), Ecohydrology and climate change in the mountains of the western USA - a review of research and opportunities. Geography Compass 4/11.

Troch, P. A., G. F. Martinez, V. R. N. Pauwels, M. Durcik, M. Sivapalan, C. Harman, P. D.

Brooks, H. Gupta, and T. Huxman (2009), Climate and vegetation water use efficiency at catchment scales, Hydrol. Process, 23, 2409-2414.

Trujillo, E., N. P. Molotch, M. L. Goulden, A. E. Kelly, and R. C. Bales (2012), Elevationdependent influence of snow accumulation on forest greening, Nature Geoscience, 5, 705-709.

Van Leeuwen, W., B. Orr, S. Marsh, and S. Herrmann (2006), Multi-sensor NDVI data continuity: Uncertainties and implications for vegetation monitoring applications, Remote Sens.

Environ., 100, 67-81.

Veatch, W., P. D. Brooks, J. R. Gustafson, and N. P. Molotch (2009), 'Quantifying the effects of forest canopy cover on net snow accumulation at a continental, mid-latitude site', Ecohydrology,

2, 115-128.

Vivoni, E. R. (2012), Spatial patterns, processes and predictions in ecohydrology: integrating technologies to meet the challenge, Ecohydrology, 5, 235-241.

Vivoni, E. R. et al. (2008), Vegetation controls on soil moisture distribution in the Valles

Caldera, New Mexico, during the North American monsoon, Ecohydrology, 1, 225-238.

Voepel, H., B. Ruddell, R. Schumer, P. A. Troch, P. D. Brooks, A. Neal, M. Durcik, and M.

Sivapalan (2011), Quantifying the role of climate and landscape characteristics on hydrologic partitioning and vegetation response, Water Resour. Res., 47, W00J09.

118

Webb, W., W. Lauenroth, S. Szarek, and R. Kinerson (1983), Primary Production and Abiotic

Controls in Forests, Grasslands, and Desert Ecosystems in the United-States, Ecology, 64, 134-

151.

Wilcox, B., B. Newman, D. Brandes, D. Davenport, and K. Reid (1997), Runoff from a semiarid ponderosa pine hillslope in New Mexico, Water Resour. Res., 33, 2301-2314.

Wilkinson, M. T. and G. S. Humphreys (2006), Slope aspect, slope length and slope inclination controls of shallow soils vegetated by sclerophyllous heath - links to long-term landscape evolution, Geomorphology, 76, 347-362.

Williams, M., M. Losleben, and H. Hamann (2002), Alpine areas in the Colorado Front Range as monitors of climate change and ecosystem response, Geogr. Rev., 92, 180-191.

White MA, Running SW, Thornton PE et al., (1999) Length of growing season variability and consequences for carbon sequestration. International Journal of Biometeorology, 42, 139-145.

119

Wolff, J. A., K. A. Brunstad, and J. N. Gardner (2011), Reconstruction of the most recent volcanic eruptions from the Valles caldera, New Mexico, J. Volcanol. Geotherm. Res., 199, 53-

68.

Yang, D., Y. Zhao, R. Armstrong, and D. Robinson (2009), Yukon River streamflow response to seasonal snow cover changes, Hydrol. Process. 23, 109-121.

Zapata-Rios X., McIntosh J., Rademacher L., Troch P.A., Rasmussen C., Brooks P.D., Chorover

J. (in review) Climatic and landscape controls on water transit times and silicate mineral weathering in the critical zone. Water Resour. Res.

120

8.0 FIGURES

Figure 1. The Jemez River basin is located in (a) northern New Mexico with (b) headwaters in the Valles Caldera National Preserve (VCNP). Our first order study catchments: (c) La Jara,

History Grove, and Upper Jaramillo originate on different aspects of the highest peak (Redondo

Peak).

121

25

20

15

10

5

40

35

30

25

20

LJ HG UJ

LJ HG UJ

1.00

0.75

0.50

0.25

2.5

2.0

1.5

1.0

0.5

0.0

LJ HG UJ LJ HG UJ

Figure 2. Lidar derived indicators of vegetation structure in the three study catchments. Whiskers represent one standard deviation. Mean tree height, mean diameter at the breast height (DBH), fractional canopy cover and leaf area index (LAI) are larger in the LJ and HG, the eastern catchments. UJ, the northern catchment, has smaller biomass than the two eastern catchments.

Statistical differences between the northern and eastern catchments were tested with the Mann-

Whitney U test p values <0.05.

122

850

800

750

700

650

600 a)

2008 2009 2010 2011 2012

200 d)

150

100

50

0

2008 2009 2010 2011 2012

800

750

700

650

600

550

500

450 g)

2008 2009 2010 2011 2012

500

450

400

350

300

250 b)

2008 2009 2010 2011 2012

0.3

e)

0.2

0.1

0.0

2008 2009 2010 2011 2012

850

800

750

700

650

600 h)

2008 2009 2010 2011 2012

500

450

400

350

300

250 c)

2008 2009 2010 2011 2012

0.15

f)

0.10

0.05

0.00

2008 2009 2010 2011 2012

1.00

0.95

0.90

0.85

0.80

0.75

i)

2008 2009 2010 2011 2012

La Jara

History Grove

Upper Jaramillo

Figure 3. Water partitioning in the study’s three experimental catchments. The three study catchments showcase similar variability in (a) annual, (b) winter and (c) summer precipitation.

However, LJ the highest elevation catchment receives the largest amount of precipitation, approximately 50 mm more than HG and UJ. Mean discharge (d), runoff ratio (e) and quick to baseflow ratio (f) are consistently large and less variable in UJ, the north facing catchment.

Vaporization (V), wetting (W) and Horton index (HI), shown in plots (g) through (h), are smaller in UJ. HI is an indicator of climate and landscape controls of available water to vegetation. HI is smaller in UJ and larger in LJ and HG, the east facing catchments. A higher HI in the eastern catchments indicates vegetation uses more of its available water than in UJ. These plots indicate the catchments process precipitation differently. Both, W and V variability are controlled principally by climate but W is further influenced by landscape characteristics while V is secondarily controlled by vegetation. A lower V is expected in a catchment that receives a smaller solar loading and in contrast higher V values are expected in the two catchments facing

east. W depends on catchment landscape characteristics and indicates water availability to vegetation. W is consistently smaller in the north facing catchment.

123

124

7.5

7.0

6.5

6.0

5.5

5.0

250

200

150

100

50

0

LJ HG UJ

LJ HG UJ

Figure 4. NDVI response metrics between the three study catchments from 2000 through 2012.

The north facing catchment (UJ) has the smallest vegetation greening indicated by the integrated

NDVI response and the shortest growing season length. Statistical differences in NDVI response between eastern and northern catchments was tested with both the Mann-Whitney U test and pair t test p values <0.05.

125

R

2

Integrated NDVI (Apr-Sep)

1.0

0.8

0.6

0.4

0.2

0.0

*

*

*

LJ

HG

UJ m=0.002

1.0

0.8

0.6

0.4

0.2

0.0

LJ HG UJ

LJ HG UJ

1.0

0.8

0.6

0.4

0.2

0.0

*

*

*

LJ m=0.003

HG m=0.002

UJ m=0.002

1.0

0.8

0.6

0.4

0.2

0.0

LJ HG UJ LJ HG UJ

1.0

0.8

0.6

0.4

0.2

0.0

*

*

*

LJ m=-10.1

HG m=-5.5

UJ m=-4.7

LJ HG UJ

Figure 5. Precipitation at the long-term Quemazon SNOTEL monitoring station versus integrated vegetation response from 2000 through 2012 in the three study catchments. Annual precipitation, SWE, and Horton index (HI) indicate strong relations with vegetation response.

However Horton index is the strongest predictor of vegetation greening. NDVI decreases as a catchment becomes drier as indicated by an increase in HI. Winter P and Summer P are not good predictors of vegetation greening (m represent the slope of the best linear regression, * statistical significant p<0.05).

126

Figure 6. Conceptual model of vegetation structure, water partitioning and vegetation greening between contrasting slopes (northern and eastern) around Redondo Peak. The eastern slopes receive more solar loading and have larger biomass. Snow represents 40% of annual precipitation and controls water partitioning around Redondo Peak. Eastern slopes have larger vaporization, and vegetation takes advantages of a larger portion of its water availability. Also they have larger NDVI response and longer growing season. In contrast, in the northern slopes there is more water availability as baseflow and discharge, and the annual rates of vaporization and Horton index are smaller.

127

9.0 TABLES

Table 1. Physical characteristics of the three study catchments draining Redondo Peak: La Jara

(LJ), History Grove (HG), and Upper Jaramillo (UJ) presented in decreasing elevation order.

The three study catchments are characterized by differences in terrain aspect which affect inputs of solar radiation, mass, and energy into the critical zone.

Catchment LJ HG UJ

Mean elevation (m)

Elevation range (m)

Area (m

2

)

Catchment slope

3,100.0

2,702-3,429

3,671,870.0

2,948.0

2,681-3,308

2,421,410.0

2,925.0

2,723-3,324

3,055,060.0

Mean (SD†)

0.29 (0.17)

East

0.24 (0.13)

East

0.27 (0.19)

North Predominant terrain aspect

Total Annual incoming solar radiation

(KWh.m

-2

)

Mean (SD†) 73.98(14.46)

74.90 (10.04) 57.21(12.49)

†SD= standard deviation

128

Table 2. Main vegetation land cover classes in the three study catchments according to Muldavin et al. (2006). The eastern catchments of LJ and HG have the largest forested areas within the study site. Forest cover represents the largest land cover in the eastern catchments of LJ and HG.

LJ HG Upper Jaramillo UJ

Land cover classes

Coniferous forest

Aspen Forest

Shrubland

Meadow

Grassland

Area (m2)

2,912,352.0

71,340.0

391,056.0

53,992.0

%

79.3

Area (m2)

1,761,676.0

1.9 80,144.0

14,468.0 0.4 14,848.0

10.7 487,188.0

1.5 37,888.0

% Area (m2)

71.6 2,088,532.0

3.3 168,272.0

0.6 8,472.0

19.8 423,152.0

1.5 55,232.0

Wetland

Others†

2,316.0 0.1 2,132.0

226,328.0 6.2 77,700.0

0.1 4,636.0

3.2 366,740.0

TOTAL 3,671,852.0 2,461,576.0 3,115,036.0

†Sparsely vegetated rock outcrop, Felsenmeer rock field, roads-disturbed ground and open water.

%

67.0

5.4

0.3

13.6

1.8

0.1

11.8

129

Table 3. Magnitude and timing of discharge indicators in the three study catchments from water years 2008 through 2012. During the five years of analysis, UJ was characterized by the largest and less variable discharge, and the timing of the maximum peak discharge occurred later than in

LJ and HG, the eastern catchments. The differences between the three study catchments indicated by the discharge indices decreased in 2010, the wettest year within the analysis period.

Discharge indices Catch.

Magnitude

Peak flow mm/day

Mean discharge mm/day

Timing

LJ

DOWY Peak flow HG

DOWY Η‚

Η‚ DOWY= day of the water year

UJ

LJ

HG

UJ

LJ

HG

UJ

Water Year

2008 2009 2010 2011

0.50

0.59

1.21

0.86

1.89

1.68

1.80

3.02

3.22

0.23

0.27

0.86

0.15

0.18

0.28

0.15

0.24

0.28

0.33

0.31

0.41

0.13

0.08

0.15

208

186

214

211

208

217

200

203

199

216

307

326

2012 Mean

0.74

1.06

1.38

0.83

1.37

1.67

STDEV C.V.

0.60

1.11

0.92

0.72

0.81

0.55

0.22

0.17

0.28

0.19

0.20

0.28

0.09

0.09

0.11

0.49

0.46

0.39

184

178

187

208.8

226.0

239.0

6.70

54.81

58.53

0.03

0.24

0.24

Table 4. Climatic controls on Horton index. Multiple linear regression parameters showing influence of maximum SWE, snow cover duration, and snow accumulation on HI using data from the long-term Quemazon SNOTEL station (2008 - 2012). HI during the 5 years study period is highly correlated with snow variables. Other snow parameters like timing to initial snow cover, peak of SWE, timing of melt, average air winter temperature were not strong predictors of HI.

Catchment

Intercept

Max SWE (mm) snow accumulation

(days) snow cover (days)

R

2

p

LJ

p

1.07 0.001

-0.0001 0.023 -0.0002 0.004 -0.001

-0.001

0.001

0.07

0.99

0.001

HG

p

1.12 0.0001

UJ

p

1.08 0.004

-0.001 0.01 -0.0004

0.98 0.95

0.002 0.05

0.04

130

131

Table 5. Interannual and intercatchment variability of the Horton index (HI). The Horton index variability among the three study catchments within a single year can be as large as the Horton index variability due to climate change during the study period e.g. WY 2010. This suggests strong aspect differences in microclimate around Redondo and in the way the catchments filter precipitation. The intercatchment differences can be larger than the hydrological response due to the natural precipitation variability in a single catchment.

Horton Index (HI) Intercatchment comparison

Water

Years LJ

2000 0.92

HG

0.95

UJ

0.89

Mean

0.92

Stdev

C.V.†

0.03 0.03

2001

2002

2003

2004

0.88

0.97

0.89

0.93

0.88

0.99

0.92

0.93

0.78

0.93

0.84

0.86

0.85

0.96

0.88

0.91

0.06

0.03

0.04

0.04

0.07

0.03

0.04

0.05

2005

2006

2007

2008

2009

2010

2011

2012

Mean interannual comparison

Stdev

C.V.

0.90

0.93

0.90

0.93

0.95

0.87

0.94

0.92

0.92

0.03

0.89

0.97

0.90

0.91

0.89

0.84

0.95

0.92

0.92

0.04

0.79

0.92

0.82

0.87

0.87

0.78

0.92

0.86

0.86

0.05

0.03 0.04 0.06

†C.V. coefficient of variation

0.86

0.94

0.87

0.90

0.90

0.83

0.06 0.07

0.03 0.03

0.05 0.06

0.03 0.03

0.04 0.04

0.04 0.05

0.94

0.90

0.02 0.02

0.03 0.04

132

10.0 SUPLEMENTAL MATERIAL

Table S1. Valles Caldera National Preserve meteorological stations listed according to elevation.

Mean annual precipitation and temperature for water years 2008 through 2012. A water year (WY) is considered from October 1 st

to Sept 30 th

.

Latitude Longitude Elevation Average WY 2008-2012

Station Name

San Antonio

(˚)

35.98

(˚)

-106.57

(m)

2598

Air

Precipitation Temperature

(mm)

(˚C)

514 4.8

Headquarters

(HQ)

Los Posos

Valle Toledo

Redondo

35.86

35.92

35.97

35.88

-106.52

-106.42

-106.46

-106.55

2644

2738

2750

3232

680

601

585

789

4.7

4.7

6.1

3.6

133

1.0

0.8

0.6

0.4

0.2

LJ

HG

UJ

0.0

10/07 4/08 10/08 4/09 10/09 4/10 10/10 4/11 10/11 4/12 10/12

Time (months)

Figure S1. Vegetation response for the three study catchments from 2008 through 2012 base on a time series of mean catchment scale NDVI derived from the NASA’s Moderate Resolution

Imaging Spectroradiometer (MODIS). NDVI is not presented for the month of October through the end of March due to slow vegetation activity from winter season. Vegetation response was quantified with the integrated NDVI response estimated as the area under the NDVI curve between April and September. The length of the growing season was estimated as the number of days when NDVI values are above 0.6. NDVI values above 0.6 are generally observed between

May and September each year.

134

3

2

1

Winter (Oct-Apr)

La Jara

History Grove

Upper Jaramillo

0

7/07 11/07 3/08 7/08 11/08 3/09 7/09 11/09 3/10 7/10 11/10 3/11 7/11 11/11 3/12 7/12 11/12

Date (month/year)

Figure S2. Stream hydrographs for water years 2008 through 2012 recorded at the flumes in

LJ,HG and UJ. Discharge shows large interannual variability driven largely by winter precipitation in all years. However, the relative yield is highest in UJ, intermediate in LJ and lowest in HG. The hydrograph of UJ (north facing catchment) indicates that this catchment responds more readily to rainfall events than LJ and HG (both facing east) suggesting differences in the partitioning of precipitation between the catchments. Grey bars represent winter season between October and April.

135

Table S2. Climatic controls on discharge. Multiple linear regression parameters showing influence of maximum SWE, snow cover duration, and snow accumulation on discharge using the data from the long-term Quemazon SNOTEL station (2008 - 2012). Other snow parameters like timing to initial snow cover, peak of SWE, timing of melt, average air winter temperature were not strong predictors of discharge.

Catchment LJ

p

HG

p

UJ

p

Intercept

Max SWE (mm) snow accumulation

(days) snow cover (days)

R

2

P

-49.53 0.07 -92.47 0.09 -65.35 0.08

0.1 0.07 0.13 0.07 0.21 0.04

0.73

0.01

0.91 0.02

0.95

0.8

0.94

0.04

0.98

0.02 0.05 0.06

136

1000

800

600

400

200

0

P

PET

V

S07 W08 S08 W09 S09 W10 S10 W11 S11 W12 S12

Time (seasons)

Figure S3. Records of precipitation (P), potential evapotranspiration (PET) and vaporization (V) from the VCM eddy covariance flux tower. The winter season shows a marked surplus of water availability due to higher P than V during this period, whereas during the summer season, the difference between P and V decreases. Summer P does not cover water demand by V in 2008,

2009 and 2010. In the 2007 and 2011, the contribution of summer P is slightly higher than V. The high PET during winter and summer indicates that this system is water and not energy limited.

137

Table S3. Water partitioning at the catchment scale in LJ, HG and UJ from 2008 through 2012.

Mean, standard deviation and coefficient of variation (CV) are indicated at the bottom of the table

138

APPENDIX C:

INFLUENCE OF CLIMATE VARIABILITY ON WATER PARTITIONING AND

EFFECTIVE ENERGY AND MASS TRANSFER (EEMT) IN A SEMI-ARID CRITICAL

ZONE

Authors:

Xavier Zapata-Rios

1*

, , Paul D. Brooks

2, 1

, Peter A. Troch

1

, Jennifer McIntosh

1

1. Department of Hydrology and Water Resources, University of Arizona, Tucson, Arizona,

USA

2. Department of Geology and Geophysics, University of Utah, Salt Lake City, Utah, USA

* Corresponding author:

1133 E James E Rogers Way

J W Harshbarger Bldg Rm 122, PO Box 210011

The University of Arizona

Tucson AZ 85721-0011

Email: [email protected]

139

ABSTRACT

The Critical Zone (CZ) is the heterogeneous, near-surface layer of the planet that regulates lifesustaining resources. Previous research has demonstrated that a quantification of the influxes of effective energy and mass (EEMT) to the CZ can predict its structure and function. In this study, we quantify how climate variability in the last three decades (1984-2012) has affected water availability and the temporal trends in EEMT. This study takes place in the 1200 km

2

upper

Jemez River Basin in northern New Mexico. The analysis of climate, water availability, and

EEMT was based on records from two high elevation SNOTEL stations, PRISM data, catchment scale discharge, and satellite derived net primary productivity (MODIS). Records from the two

SNOTEL stations showed clear increasing trends in winter and annual temperatures (+1.0-1.3

°C/decade; +1.2-1.4 °C/decade, respectively), decreasing trends in winter and annual precipitation (-41.6-51.4 mm/decade; -69.8-73.2 mm/decade, respectively) and maximum Snow

Water Equivalent (SWE;-33.1-34.7 mm/decade). The water partitioning fluxes at the basin scale showed statistically significant decreasing trends in precipitation (-61.7mm/decade), discharge (-

17.6 mm/decade) and vaporization (-45.7 mm/decade). Similarly Q

50,

an indicator of snowmelt timing, is occurring 4.3 days/decade earlier. Results from this study indicated a decreasing trend in water availability, a reduction in forest productivity (4 g_C.m

-2

per 10 mm of reduction in

Precipitation) and EEMT (1.2 – 1.3 MJ.m

2

.decade

-1

), and increased vegetation water use efficiency. During the study period, a decreasing trend in EEMT was observed of 1.2 and 1.3

MJ/ m

2

decade. These changes in EEMT point towards a hotter, drier and less productive ecosystem which may alter critical zone processes in high elevation semi-arid systems.

KEY WORDS

EEMT, Jemez River Basin, climate variability, critical zone, Northern New Mexico.

140

141

1.0 INTRODUCTION

The critical zone (CZ) is the surficial layer of the planet that extends from the top of the vegetation canopy to the base of aquifers [Chorover et al., 2011; Brandley et al., 2007]. Within its boundaries complex interactions between air, water, biota, organic matter, soils and rocks take place that are critical for sustaining live on Earth [Brandley et al., 2007]. The CZ has been conceptualized and studied as a weathering engine or reactor where interacting chemical, physical and biological processes drive weathering reactions [Anderson et al., 2007; Chorover et

al., 2011]. Over long time scales, the CZ has evolved in response to climatic and tectonic forces and has been recently influenced by human activities [Steffen et al., 2007]. Understanding how climate and land use changes affect CZ structure and related processes has become a priority for the science community due to the implications it may have on the functioning of life supporting resources. It has been hypothesized by the researchers from the Jemez River Basin (JRB) –

Santa Catalina Mountains (SCM) Critical Zone Observatory (CZO)

[http://criticalzone.org/catalina-jemez/] that a quantification of the inputs of the effective energy and mass transfer (EEMT) to the CZ can provide insight about its structure and function

[Chorover et al., 2011]. CZ areas that receive greater EEMT influxes have been shown to have greater structural organization as well as more dissipative products leaving it [Rasmussen et al.,

2011; Zapata-Rios et al., submitted]. The opposite has been observed in regions with less

EEMT.

EEMT is a variable that quantifies energy and mass transfer to the critical zone

[Rasmussen et al., 2011; Chorover et al., 2011]. EEMT integrates within a single variable the energy and mass associated with water in excess from evapotranspiration, quantified as effective precipitation (E ppt

), and reduced carbon compounds resulting from primary production (E bio

)

[Rasmussen et al., 2011]. It has been demonstrated that other possible energy fluxes to the CZ

142 such as potential energy from transport of sediments, geochemical potential of chemical weathering, external inputs of dust, heat exchange between soil and atmosphere, and other sources of energy coming from anthropogenic sources are orders of magnitude smaller [Phillips,

2009; Rasmussen et al., 2011; Rasmussen, 2012]. Therefore the two dominant terms embodied in EEMT are E ppt and E bio

.

Previous research has shown that EEMT can become a tool to predict regolith depth, rate of soil production and soil properties [Rasmussen et al., 2005; Rasmussen et al, 2011; Pelletier and Rasmussen, 2009a,b; Rasmussen and Tabor, 2007]. For instance, strong correlations were found between EEMT, soil carbon, and clay content in soils on igneous parent materials from

California and Oregon [Rasmussen et al., 2005]. Furthermore, transfer functions were successfully determined between EEMT and pedogenic indices, including pedon depth, clay content, and chemical indices of soil alteration along an environmental gradient on residual igneous parent material [Rasmussen and Tabor, 2007]. EEMT has also been incorporated in geomorphic and pedogenic models on granitic rocks to describe landscape attributes and regolith thickness [Pelletier and Rasmussen, 2009 a,b]. Rasmussen and Tabor [2007] demonstrated that regolith depth on stable low gradient slopes increased exponentially with increasing EEMT.

Similarly, Pelletier et al. [2013] found that high EEMT values are associated with large above ground biomass, deeper soils, and longer distance to the valley bottoms across hillslopes in the

Santa Catalina Mountains in southern Arizona. More recently, EEMT estimations haven been strongly correlated with water transit times, water solutes concentrations and dissolution of silicates on a rhyolitic terrain in northern New Mexico [Zapata et al., in review]. In these studies, the main constituents of EEMT (E ppt

and E bio)

were quantified as an average value based

143 on climate records from long-term regional databases as these variables exert first-order controls on photosynthesis and effective precipitation [Rasmussen et al., 2011; Chorover et al., 2011].

It is still uncertain how climate variability influences CZ structure and function

[Chorover et al., 2011]. Climate variability might directly influence changes in the transfer of mass and energy to the CZ as climate has a direct control on both E ppt

and E bio

. In the mountains of the southwestern United States, a large percentage of annual precipitation falls as snow, which is stored during the winter and released as snowmelt during the spring [Clow, 2010]. The water from the winter snowpack constitutes the main source of regional water supplies and the largest component of runoff [Bales et al., 2006; Nayak et al., 2010]. The regional snowpack has been documented to be declining in the southwestern US [Mote et al., 2005; Clow, 2010] and alterations to the snowpack are likely to produce changes in vegetation, impact water availability

[Bales et al., 2006; Harpold et al., 2012; Trujillo et al., 2012] and influence inputs of EEMT.

For instance, significant increasing trends in air temperature and decreasing trends in winter precipitation in the last decades have been documented in the Upper Rio Grande region in northern New Mexico [Harpold et al., 2012].

The objective of this study was to evaluate climate variability and its influence on the temporal changes of water partitioning and EEMT at the catchment scale in a semi–arid CZ over the last few decades. This investigation took place in the upper part of the Jemez River Basin in northern New Mexico, a basin dominated by a wide forest cover and limited human infrastructure, where the Santa Catalina-Jemez River Critical Zone Observatory has established a research site to study CZ processes [Chorover et al., 2011]. Micro-climate variability was studied based on daily records from two SNOTEL stations using records from 1984 through

144

2012. Water availability and EEMT were estimated during the same time period based on precipitation and temperature from the precipitation-elevation regressions on independent slopes model (PRISM), empirical daily observations of catchment scale discharge, and satellite derived net primary productivity (MODIS).

2.0 METHODS

2.1 Study site

The Jemez River is a tributary of the upper reach of the Rio Grande and is located between Jemez and Sierra Nacimiento Mountains in northern New Mexico (Figure 1a). Its headwaters originate within the 360 Km

2

Valles Caldera National Preserve which contains 30% of the total basin surface (Figure 1b). The upper Jemez River Basin is located at the southern margin of the Rocky Mountains ecoregion between latitudes 35.6˚ and 36.1˚ north and longitudes -106.3˚and -106.9˚west. The basin is characterized by a mean elevation of 2591 m and a gradient in elevation ranging from 1712 to 3435 m. Based on a 10 m digital elevation model, the catchment drains 1218 km

2 above the US Geological Survey (USGS) gauge “Jemez

River near Jemez” (35.66˚ N and 106.74˚ W; USGS 08324000) located at an elevation of 1712m.

The basin has a predominant south aspect and a mean catchment slope of 13.7˚. The geology consists of rocks of volcanic origin with predominant basaltic and rhyolitic compositions that overlie tertiary to Paleozoic sediments along the western margin of the Rio Grande rift

[Shevenell et al., 1987]. Common soil types in the basin are aridisols, alfisols, mollisols and inceptisols [Allen et al., 1991, 2002]. Precipitation has a bimodal pattern with 50% of annual precipitation occurring during the winter months (primarily as snow) from October to April and originates from westerly frontal systems. The remaining 50% of precipitation falls as

145 convectional rainfall during the monsoon season between July and September [Sheppard, 2002].

According to the National Land Cover Database (NLCD), the basin is a forested catchment with

79% under evergreen, deciduous and mixed forest cover and only 0.5% of area covered by development and agriculture [http://www.mrlc.gov/nlcd06_leg.php] (Table 1).

2.2 Climatological stations

There are two Natural Resources Conservation Service snowpack telemetry (SNOTEL) stations within the study area with long-term records since 1980

( http://www.wcc.nrcs.usda.gov/snow/; Figure 1b

). The Quemazon station is located at an elevation of 2896 m (35.92 ̊ N and 106.39 ̊ W) and the Señorita Divide#2 station is located at an elevation of 2622 m (36.00 ̊ N and 106.83 ̊ W). The stations collect real-time precipitation, snow water equivalent (SWE), air temperature, soil moisture and temperature, and wind speed and direction. Air temperature records began at the Señorita Divide#2 in 1988 and at the Quemazon station in 1989. There are no stations with long-term records at the lower part of the basin.

2.3 Climate variability

Climate variability was studied based on 13 variables from the two SNOTEL stations, derived from daily air temperature, precipitation, and maximum SWE, following a similar methodology and data processing procedure as in Harpold et al. [2012]. The variables analyzed were winter, summer and annual air temperature ( ̊C), annual and winter precipitation (mm), maximum SWE (mm), maximum SWE to winter precipitation ratio (-), 1 st

of April SWE (mm), first day snow cover (water year day), last day snow cover (water year day), length of snow on the ground (number of days) and SM50, which is the day of the year in which half of the

146 snowpack melts (number of days). Climate records for data analysis were aggregated by water year (from October 1 st

to September 30 th

). Winter season was considered to be between October and April and summer season between May and September. The analysis of climate was conducted from 1984 as a starting year to avoid the anomalous wet years recorded at the beginning of 1980s that were caused by the Pacific Decadal Oscillation (PDO) and El Niño-

Southern Oscillation (ENSO) [Harpold et al., 2012; and references therein]. The presence of a monotonic increasing or decreasing trend in the 13 climate variables recorded at the two individual stations was evaluated from 1984 through 2012 by applying the nonparametric Mann-

Kendall test with a α=0.10 level of significance and the nonparametric Sen’s slope estimator of a linear trend [Yue et al., 2012; Sen, 1968].

2.4 EEMT estimation

In this investigation EEMT was calculated as the sum of E ppt

and E bio

(equation 1). We applied two different methods to estimate E ppt

and E bio

. Following a similar methodology described in Rasmussen and Gallo [2013], EEMT emp

was empirically estimated at the catchment scale based on baseflow estimations and average basin scale net primary productivity (NPP) derived from MODIS satellite data. On the other hand, EEMT model

was estimated at the catchment scale based on long term climate records from Precipitation elevation Regressions on

Independent Slopes Model (PRISM) developed by the climate group at Oregon State University

( http://www.wcc.nrcs.usda.gov/ftpref/support/climate/prism/ ) and described in Rasmussen et al.

[2005; 2011].

𝐸𝐸𝑀𝑇 = 𝐸 𝑝𝑝𝑑

+ 𝐸 π‘π‘–π‘œ

(𝐽 π‘š

−2 𝑠

−1

) (1)

147

EEMT emp

Upper Jemez River Basin precipitation and air temperature from 1984 through 2012 was obtained using PRISM data at an 800 meters spatial resolution [Daly et al., 2002]. Daily discharge data was available from 1984 through 2012 from the USGS Jemez River near Jemez gauge station ( http://waterdata.usgs.gov/nwis ). The upper Jemez River has not been subjected to flow regulation and almost 60% of the annual discharge occurs during the snowmelt period between March and May. Daily discharge records were normalized by catchment area and mean daily discharge was aggregated into water years.

Catchment scale water partitioning fluxes (1984-2012) were calculated following the

Horton index approach of Troch et al. [2009] and Brooks et al. [2011]. The Horton index is an indicator of water partitioning at the catchment scale and integrates both effects of landscape and vegetation in water partitioning [Voepel et al., 2011]. The Horton index represents a metric of catchment scale vegetation water use and was calculated as follows:

𝑉

𝐻𝐼 =

π‘Š

=

𝑃 − 𝑄

(2)

𝑃 − 𝑆 where, V represents vaporization, W catchment wetting, P catchment scale precipitation, Q discharge and S quickflow.

Precipitation (P) on the land surface was partitioned between quickflow (S) and catchment wetting (W). S represents water that directly contributes to streamflow discharge as a response to precipitation events, thus this amount of water is not transferred to the critical zone.

W is the total amount of water that infiltrates the soil, of which a portion is available for vaporization (V) including vegetation uptake. The remaining portion of W flows though the

critical zone and contributes to baseflow (U). V was estimated at the annual scale as the difference between P and discharge (Q). Q was separated between S and U using a oneparameter low-pass filter [Lyne and Hollick, 1979; Arnold and Allen, 1999; Eckhardt, 2005;

Troch et al., 2009] (equation 3).

148

π‘ˆ π‘˜

= π‘Žπ‘ˆ π‘˜−1

+

1 − π‘Ž

2

(𝑄 π‘˜

− 𝑄 π‘˜−1

) (3)

π‘ˆ π‘˜

≤ 𝑄 π‘˜ where a is a filter parameter set to 0.925. This filter was passed twice, backward and forward in time to improve the partitioning of U and S at the beginning of the time series. After this, daily values of Q, U, and S were integrated to annual time scales. Alterations in snowmelt timing were evaluated with Q

50

, which indicates the day of the water year when 50% of the total annual discharge is recorded at the catchment outlet [Clow, 2010; Stewart et al., 2005].

The term Eppt emp

was calculated as stated in equation (4) based on estimations of U and mean PRISM derived air temperature at the catchment scale [Rasmussen et al., 2011; Rasmussen and Gallo, 2013].

𝐸𝑝𝑝𝑑 = π‘ˆ ∗ 𝐢 𝑀

∗ π›₯𝑇 (𝐽 π‘š

−2 𝑠

−1

) (4)

In equation 4, C w

is the specific heat of water (4187 J kg

-1

K

-1

) and ΔT is the difference in temperature between ambient temperature and 0 °C calculated as T ambient

minus T ref

(273.15 °K).

Net primary productivity

Mean annual NPP at the catchment scale was estimated at a 1 km spatial resolution for the years 2000 through 2012 using data MOD17A3 from MODIS [Zhao and Running, 2010]

149

( http://modis-land.gsfc.nasa.gov/npp.html

). E bio

was calculated as indicated in equation (5) and presented in Rasmussen et al. [2011] and Rasmussen and Gallo [2013].

πΈπ‘π‘–π‘œ = 𝑁𝑃𝑃 ∗ β„Ž π‘π‘–π‘œ

(𝐽 π‘š

−2 𝑠

−1

) (5) where, h

bio

is the specific biomass enthalpy and equivalent to 22 kJ m

-2 s

-1

[Lieth, 1975; Phillips,

2009]. As MODIS data was only available from the year 2000 onwards, single and multivariate linear regression analysis were estimated with the objective of finding a statistical model to extend Ebio emp

records back to 1984. Using a similar approach as Rasmussen and Tabor [2007], linear regressions were explored between Ebio emp

and climate variables from the SNOTEL stations and the entire basin.

EEMT model

Eppt model was calculated based on estimations of effective precipitation (P eff

) which is defined as the amount of water that enters the CZ in excess of evapotranspiration and is available to flow through the CZ [Rasmussen et al., 2005; equation 6]

𝐸𝑝𝑝𝑑 𝑖 π‘šπ‘œπ‘‘π‘’π‘™

= 𝑃𝑒𝑓𝑓 𝑖

∗ 𝐢 𝑀

∗ βˆ†π‘‡ (6) where Peff

i

is monthly effective precipitation calculated as the difference between monthly PRISM precipitation and monthly potential evapotranspiration calculated using the Thornthwaite equation

[Rasmussen et al., 2005; Thornthwaite, 1948]. C

w

and βˆ†T are the same parameters described in equation (4). Eppt

i model

was calculated on a monthly basis only for the months when precipitation is larger than evapotranspiration (Peffi > 0) and these values were integrated in water years.

Ebio model was estimated as indicated in equation 5 and NPP was calculated following an empirical relationship based on air temperature [equation 7; Lieth, 1975].

150

3000

𝑁𝑃𝑃𝑖 =

1 + 𝑒

1.315−0.119 π‘‡π‘Ž

∗ π‘‘π‘Žπ‘¦π‘  𝑖

365 π‘‘π‘Žπ‘¦π‘ /π‘¦π‘’π‘Žπ‘Ÿ

(7)

NPPi is monthly NPP in g.m

-2

.year

-1

and Ta is monthly air temperature. Days

i

over the number of days in a year is an NPP time correction. Similar to equation 5, Ebio model

was calculated for the months where Peff

i

>0 only. For a detailed description of EEMT see Rasmussen et al. [2005;

2011; 2015], Rasmussen and Tabor [2007] and Rasmussen and Gallo [2013].

2.5 Water availability, water partitioning and climate controls on water availability

A trend analysis was conducted using data from 1984 through 2012 on each component of the water partitioning analysis (P, Q, U, S, V, W, Q

50

), Horton index and EEMT using the nonparametric Mann-Kendall test and the Sen’s slope estimator of a linear trend with a α=0.10 level of significance [Yue et al., 2012; Sen, 1968]. Relationships between climate, hydrological variables and EEMT were examined by simple and multiple linear regression analysis with parameters fit through a least square iterative process [Haan, 1997].

3.0 RESULTS

3.1 Changes in climate

Records from the Quemazon SNOTEL station from 1984 to 2012 indicated a mean annual precipitation of 701 mm, of which 50% fell during the winter months with a mean maximum SWE of 242.5 mm. The mean annual and winter temperatures at this site were 3.98 ̊

C and -0.87 ̊ C, respectively. During the same time period, Señorita Divide#2 station had a mean annual precipitation of 686 mm, of which 61% fell during the winter with a mean maximum

SWE recorded of 239.2 mm. The mean annual and winter temperatures at the Señorita Divide#2 site were 4.23 and -0.90 ̊ C, respectively (Table 2).

151

During the three decades of analysis, seven out of the 13 climate variables in both stations showed a statistically significant trend (Table 3). Mean winter, summer and annual air temperatures at the Quemazon station increased significantly by 1.3°C, 1.0 °C and 1.4 °C per decade, respectively. Similarly, the same variables at the Señorita Divide#2 station increased 1.0

°C, 1.0 °C and 1.2 °C per decade, respectively. The rates of increase in winter and annual air temperature were larger in Quemazon, the higher elevation station. Annual precipitation decreased in both stations at similar rates per decade. Quemazon station decreased

69.8mm/decade (p≤0.01) and Señorita Divide#2 decreased 73.2 mm/ decade (p≤0.05). Winter precipitation decreased faster at the Señorita Divide #2, the lower elevation station (59.4 mm/decade; p≤0.05) than at the Quemazon station (41.6 mm/decade; p≤0.1). Maximum SWE decreased in both stations at similar rates, -34.7 mm/decade at Señorita Divide #2 and -33.1 mm/decade at the Quemazon station (p≤0.1). There was no significant trend in the ratio between

SWE to winter precipitation at either station. Observed April 1 st

SWE also decreased -60.5 mm/decade (p≤0.05) and -54.4 mm/decade (p≤0.1) at the Quemazon and Señorita Divide#2 stations, respectively. The day of occurrence of maximum SWE recorded at the Quemazon station showed a significant trend indicating that maximum SWE is occurring 5.7 days earlier every decade (p≤0.05). However, this same trend was not observed at the Señorita Divide#2 station. Variables such as SM50, initiation of snow cover, and snow cover duration did not indicate any trend of change in either station at the 90% confidence level. In contrast, there is a decreasing trend in the last day of snow cover, which is happening about 6 days sooner per decade in the Quemazon station (p<0.05). Last day of snow cover at the Señorita Divide #2 station did not show a significant trend (Table 3).

3.2 Water availability

Mean precipitation in the Jemez River Basin from 1984 to 2012 was 617 mm with observed extreme values of 845 mm in 1985 and 336 mm in 2002. During the analysis period,

152 winter precipitation represented 54% of total annual precipitation. Mean annual precipitation at the catchment scale correlated significantly with the mean annual precipitation recorded at the

Quemazon (R

2

=0.45; p<0.0001) and Señorita Divide#2 stations (R

2

=0.73; p<0.0001). In this same timeframe average, minimum and maximum basin scale temperatures were 6.1, -1.5 and

13.6 ̊ C, respectively. In general, January was the coldest and July the warmest month. Basin scale mean annual and winter temperature indicated a statistically significant increasing trend of

0.5° C and 0.4 ° C per decade (not shown). Mean annual temperature in the Jemez River Basin significantly correlated with the mean annual temperature recorded at the Quemazon (R

2

=0.29;

p<0.006) and Señorita Divide#2 stations (R

2

=0.67; p<0.0001) (not shown).

Mean river basin discharge during the study period was 0.15 mm/day and the maximum and minimum historical streamflow discharges were 2.97 and 0.008 mm/day, respectively. In the

29 years of daily discharge records, 90% of the time discharge surpassed 0.03 mm/day and 10% of the time exceeded 0.38 mm/day. Peak discharge occurred between March and May and 58% of the annual discharge flowed between these months.

From 1984 to 2012, three percent of annual precipitation became quickflow and contributed directly to the streamflow discharge (3% P; standard deviation STDEV=1.2% P). As a result, 97% of the annual precipitation (STDEV=1.2% P) infiltrated and was available for vegetation uptake. This 97% of annual precipitation is further partitioned between vaporization and baseflow. The amount of water vaporized into the atmosphere represented 91% of the annual

precipitation (STDEV=3.4% P). Baseflow corresponded to 6.1% of the annual precipitation

(STDEV=2.2% P) and represented the largest component of discharge (73.2% Q; STDEV =

5.4% Q). Quickflow represented the remaining 26.8% of annual discharge (STDEV=5.4%Q).

During the study period, the mean Horton index was 0.94 (STDEV=0.02).

153

There was a significant decreasing trend in precipitation and all the water partitioning components in the upper Jemez River Basin as quantified by the Mann-Kendall test (MKT)

(Figure 2). Precipitation in the basin decreased at a rate of -61.7 mm per decade (p=0.02) (Figure

2a) while discharge decreased at a rate of -17.6 mm per decade (p=0.001) (Figure 2b). The two components of discharge, baseflow and quickflow decreased at a rate of -12.4 mm (p<0.001) and

-5.1 mm (p=0.005) per decade, respectively (Figure 2c, 2d). Water loss by vaporization decreased -45.7 mm per decade (p=0.04; Figure 2e) and wetting decreased -56.7 mm per decade

(p<0.02; Figure 2f). As a result, an increasing water limitation trend was observed in the upper

Jemez River Basin as indicated by the decrease in all the water partitioning variables and increase of 0.014 per decade in the Horton index (p=0.002) (Figure2g). In addition to the decreasing trend in the amount of basin scale discharge, Q

50

showed that 50% of annual discharge is occurring 4.3 days earlier per decade (p=0.03).

3.3 EEMT

EEMT emp

Using the available 2000 through 2012 remote sensing data, mean MODIS NPP was found to be 450 g_C/m

2

(STDEV=57.1 g_C/m

2

). Using these 13 years of data, no trend in the mean annual NPP for the upper Jemez River Basin was found. However, mean annual NPP was positively correlated with basin scale precipitation (R

2

=0.56; p=0.003) and baseflow (R

2

=0.41;

p=0.02) (Figure 3). These results indicated that forest productivity in the upper Jemez River

154

Basin is primarily limited by water availability since other climate variables recorded at the two

SNOTEL stations were not good predictors of NPP. From 1984 through 2012 mean Eppt emp

was

1.03 MJ m

2

year

-1

(STDEV=0.49 MJ m

2

year

-1

) and mean Ebio emp was 9.89 MJ m

2

year

-1

(STDEV=1.26 MJ m

2

year

-1

). Multivariate regression analysis indicated that precipitation at the

Quemazon station and the upper Jemez River Basin were the best predictors of Ebio emp

(R

2

=0.66; p=0.06). Using this multivariate linear regression model, Ebio emp

data was extrapolated for the years 1984 through 1999. Using the combined dataset from extrapolated and measured Ebio emp

the mean annual Ebio emp

was 10.8 MJ m

2

year

-1

(STDEV=1.37 MJ m

2

year

-1

) for the period from 1984 to 2012. Mean EEMT emp

was 11.83 MJ m

2

year

-1

(STDEV=1.74 MJ m

2 year

-1

) and Ebio emp

represented 92% (STDEV=0.03%) of the total EEMT emp

during the study period.

EEMT model

From 1984 through 2012 mean Eppt model

was 0.1 MJ m

2

year

-1

(STDEV=0.07 MJ m

2 year

-1

) and mean Ebio model

was 6.72 MJ m

2

year

-1

(STDEV=2.33 MJ m

2

year

-1

). During this same period, mean EEMT model

was 6.82 MJ m

2

year

-1

(STDEV=2.38 MJ m

2

year

-1

) and Ebio model represented 99% (STDEV=1.2%) of the total EEMT model

.

EEMT emp

was on average 1.7 times larger than EEMT model

. Both EEMT emp

and

EEMT model

showed a significant linear correlation (R

2

=0.42; p=0.0002) and a similar decreasing trend of 1.2 MJ.m

2

.decade

-1

(p≤0.01) and 1.3 MJ.m

2

.decade

-1

(p≤0.05), respectively (Figure 4).

Detailed estimations of EEMT emp

and EEMT model

and its components can be found in table S1

(supplementary material). Figure 5 highlights changes of EEMT in the upper Jemez River Basin

in relation to water availability from 1984 to 2012. EEMT was positively correlated to annual

155 baseflow, increasing during wet years and decreasing during dry years.

3.4 Climate controls on discharge

When compared to the climate variables from the Quemazon station, data from the

Señorita Divide#2 showed the strongest linear correlations with discharge (Table S2). The five variables with the strongest linear correlations to discharge were winter precipitation (R

2

=0.72;

p=0.00001), maximum SWE (R

2

=0.55; p=0.00001), last day of snow cover (R

2

=0.54;

p=0.00001), annual precipitation (R

2

=0.50; p=0.00001), and annual temperature (R

2

=0.49;

p=0.00010). Variables such as first day of snow cover, SWE to winter P ratio and SM50 did not shown any relation.

Similarly, climate variables from Señorita Divide#2 showed the strongest linear correlations with Horton index (Table S3). The five variables with the strongest linear correlations to Horton index were winter precipitation (R

2

=0.59; p=0.00001), maximum SWE

(R

2

=0.59; p=0.00001), last day of snow cover (R

2

=0.55; p=0.00001), occurrence of 50% max

SWE (R

2

=0.41; p=0.00010), and annual temperature (R

2

=0.40; p=0.00070). SM50 was the only variable that did not show a strong linear correlation with Horton index at the Señorita Divide#2 station.

Based on a multivariate regression analysis, annual temperature, max SWE and the length of snow on the ground were the best predictors of discharge and explain above 80% of discharge variability in the basin (R

2

>0.80; p<0.0001) (Table 3). The predictive power of this model was similar regardless of whether data from the Quemazon or Señorita Divide#2 stations was used.

156

From these three predicting variables, annual temperature and max SWE showed decreasing trends that influence the observed decrease in water availability in the basin. Analysis of residuals of the linear model between climate variables and Jemez River discharge indicated that maximum SWE and the duration of the snow cover are the better predictors of discharge residuals variability. As it is shown in figure S1, Q residuals increased during extreme dry and wet years.

4.0 DISCUSSION

4.1 Climate variability

Global climate is changing and the instrumental records in the southwestern US for the last three decades indicate a decline in precipitation and increasing air temperatures in the region

[Hughes and Diaz, 2008; Folland et al., 2001]. Global climate models further predict drier conditions and a more arid climate for the 21 st

century in this region [Seager et al., 2007]. For instance, global climate models indicate, for the future in the southwestern US according to a low and high emissions scenarios, a substantial increase in air temperature between 0.6 to 2.2 °C and 1.3 to 5.0 °C for the period 2021-2050 and by end of the 21 st

century, respectively [Barnett

et al., 2004; Cayan et al., 2013]. An increase in winter temperature of about 0.6 °C per decade was reported from 1984-2012 at a regional level in the upper Rio Grande Basin [Harpold et al.,

2012]. In line with these other studies, we found that mean annual and winter air temperature in the upper Jemez River Basin have increased 0.5 °C and 0.4 °C per decade, respectively.

Changes in climate have been found to be a predominant influence in snowpack decline as oppose to changes in land use, forest canopy or other factors [Hamlet et al., 2005; Boisvenue and Running, 2006]. There are high confidence predictions that snowpacks will continue to

157 decline in northern New Mexico through the year 2100 and projections of snowpack accumulation for mid-century (2041-2070) show a marked reduction for SWE of about 40%

[Cayan et al., 2013]. Harpold et al., [2012] found a decrease in annual precipitation and maximum SWE for the Upper Rio Grande Basin of -33 and -40 mm per decade, respectively. In this study, a clear decreasing trend in annual, winter precipitation and max SWE was observed in records from 1984-2012 in the two high elevation SNOTEL stations. Records in this study showed approximately twice the rate of decrease in annual precipitation and a smaller decrease in max SWE of about 7 mm per decade compared to the regional results from Harpold et al.

[2012]. Harpold et al. [2012] report that SM50 (-2 days per decade), snow cover length (-4.2 days per decade), day of maximum SWE (-3.31 days per decade), and last day of snow cover (-

3.45 days per decade) for the Rio Grande Basin showed statistically significant trends. However, based on our analysis from the individual SNOTEL stations, these variables did not show any statistically significant trends.

4.2 Changes in discharge and evapotranspiration

Decreasing trends in discharge ranging from 10 to 30% are expected during the 21 st century for the western US [Milly et al., 2005] and maximum peak streamflow is expected to happen one month earlier by 2050 [Barnett et al., 2005]. Furthermore, it has been reported that streamflow in snowmelt dominated river basins are more sensitive to wintertime increases in temperature [Barnett et al., 2005]. In this study, we have found that 50.5 % of annual streamflow occurred between (April) and beginning of the summer (June). This result is congruent with other studies in snowmelt dominated systems in the region (Clow, 2010). Previous research in the southwest has found that the timing of snowmelt is shifting to early times ranging from a few

158 days to weeks [Stewart et al., 2005; Mote et al., 2005; McCabe and Clark, 2005]. For instance,

Clow [2010] reports that in southern Colorado rivers, there is a trend toward earlier snowmelt that varied from 4.0 to 5.9 days per decade and April 1 st

SWE decreased between 51 and 95 mm per decade. In this study, it was found that snowmelt timing in the upper Jemez River Basin occurred 4.3 days earlier per decade and April 1 st

SWE decreased between 54 – 60 mm/decade.

Changes in evapotranspiration are related to changes in precipitation, humidity, air temperature, irradiance and wind speed [Barnet et al., 2005]. However, the magnitude and direction of changes in evapotranspiration are still a source of debate and investigation [Ohmura and Wild, 2002]. Pan evaporation in various countries in the northern Hemisphere show that evaporation has been progressively decreasing over the past 50 years [Barnett et al., 2005]. A reduction in evapotranspiration is expected in snowmelt dominated systems, as early snowmelt provides water to the landscape when potential evapotranspiration is low and reducing soil moisture during months with high evapotranspiration demand [Barnett et al., 2005]. In addition, rising CO

2

concentrations will likely increase plant water use efficiency, enhance stomatal closure and reduce transpiration [Betts et al., 2005]. In this study, we found evidence of a decrease in vaporization of 45.7 mm/decade in the upper Jemez River basin.

4.3 EEMT components

Water partitioning

Troch et al. [2009] demonstrated that the Horton index gets closer to 1 in drier regions and during dry years. Similarly, in a study based on 86 catchments in different biomes and ecosystems across the US, Horton index values increased as catchments became more water

159 limited [Brooks et al., 2011]. Huxman et al. [2004] showed that the average rain-use efficiency

(RUE), estimated as the ratio of aboveground net primary production to annual precipitation, decreases as precipitation increases. In contrast, during dry years RUE converges to common maximum RUE similar to the drier regions, regardless of biome type. The increase in water use efficiency by vegetation as the upper Jemez River basin becomes drier indicated by the Horton index variability (0.94; STDEV=0.02) is consistent with the above mentioned studies.

Forest productivity

Reduced carbon compounds resulting from primary production are a fundamental energy component of EEMT [Rasmussen et al., 2011]. Modeling and empirical studies indicate that mountain forest productivity in the southwest is sensitive to water and energy limitations

[Christensen et al., 2008; Tague et al., 2009; Anderson-Teixeira et al., 2011; Zapata-Rios et al., in prep]. Trujillo et al. [2012] found that NDVI greening increased and decreased proportionally to the changes in snowpack accumulation along a gradient in elevation in the Sierra Nevada, while Zapata et al., [in prep] found similar results across a gradient of energy created by aspect differences at higher elevations in the Jemez Mountains. Furthermore, energy limitations to productivity have been observed in colder sites at high elevations [Trujillo et al., 2012;

Anderson-Teixeira et al., 2011; Zapata et al. in prep]. Since the mid-1980 increases in wildfires and tree mortality rates have been documented in high elevation forests due to an increase in spring and summer temperatures and decrease in water availability [Westerling et al., 2006; Van

Mantgem, P.J et al., 2009]. Results from this study indicated that in the upper Jemez River

Basin, forest productivity was primarily responding to water availability (Figure 3).

EEMT variability

160

All of the above results indicate that the Jemez River Basin is highly susceptible to changes in climate that can affect water availability and ecosystem productivity which impacts

EEMT. Rasmussen et al. [2005] estimated low rates of EEMT model

< 15 MJ.m

-2

.year

-1

for the majority of the continental US and proved that E bio

was the dominant component of EEMT with contributions above 50% of total EEMT in different soil orders. Regions dominated by E bio corresponded to regions facing water limitation and where E bio

accounted for up to 93% of the total energy and carbon flux to the CZ [Rasmussen et al., 2011; Rasmussen and Gallo, 2013]. In semi-arid regions vaporization represents over 90% loss of annual precipitation [Newman et al.,

2006] while groundwater recharge accounts for less than 10% of annual precipitation [Scanlon et

al., 2006]. Under these conditions, little water remains for critical zone processes in semi-arid regions. Other studies have found that the contributions of E bio

can be three to seven orders of magnitude larger than other sources of energy influxes to the CZ [Phillips, 2009; Amundson et

al., 2007]. In this study, we confirmed that for the upper Jemez River Basin, E bio

was the dominant term from the total EEMT and E ppt

contributions were small.

A comparison of EEMT model

and EEMT emp

in 86 catchments across the US characterized by having minimum snow influence indicated that model and empirical values were strongly linearly correlated (R

2

=0.75; p<0.0001) and EEMT model

values were larger than EEMT emp

[Rasmussen and Gallo, 2013]. One limitation of the EEMT model

method is that it calculates energy during the months when air temperature is above zero only and assumes no energy associated with precipitation falling as snow. In a snowmelt dominated systems as the upper

Jemez River Basin where snowmelt is the main source of water availability to ecosystems [Bales

et al., 2006], EEMT estimations based only on climate data will likely underestimate the energy transfer to the CZ. Therefore, using EEMT emp

methodology may be more suitable for snowmelt

dominated systems. In this study we found the expected linear correlation between EEMT model

161 and EEMT emp

(R

2

=0.42; p<0.001) however, EEMT model

values were smaller than EEMT emp values. Although the two methods used in this study to calculate EEMT indicated different absolute values of EEMT, the rates of decrease of EEMT per decade are congruent with each other (EEMT emp

=1.2 MJ.m

2

.decade

-1

; EEMT model =

1.3 MJ.m

2

.decade

-1

) (Figure 5).

The rates of EEMT change between 1.2 to 1.3 MJ.m

2

. per decade found in this study in the upper Jemez River Basin can be significant for critical zone processes. In a study conducted in a similar semi-arid region in the Santa Catalina Mountains located in southern Arizona,

Rasmussen et al. [2015] estimated differences in EEMT of about 25 MJ m

2

year

-1

between the upper elevation (2800 m) covered by mixed conifer forest and low elevation (800 m) covered by a dry semi-arid desert scrub ecosystem. These changes in EEMT along the 2000 m elevation gradient in the Catalina Mountains are equivalent to a difference of 1.25 MJ m

2

year

-1

per 100 meters in elevation change. Furthermore, Rasmussen et al. [2015] determined differences of 3.9

MJ m

2

year

-1

between contrasting north and south facing slopes, and of 0.9 MJ.m

2

.year

-1 according to topographic wetness between water gaining and water losing portions of the landscape.

Although the quantification of EEMT using the methodologies applied in this study are suitable for large spatial scales, it is limited in that it is not taking into account small scale variabilities induced by topography in solar energy, effective precipitation, NPP and redistribution of water by differences in micro-topography. Therefore, EEMT estimations at small scales (pedon to hillslopes) need to follow a different approach as indicated in Rasmussen

et al. [2015].

162

5.0 SUMMARY

We investigated how changes in climate in the southwest affect the trends in water availability, vegetation productivity and the annual influxes of EEMT to the CZ. This investigation took place in the 1200 km

2

upper Jemez River basin a semi-arid basin in northern

New Mexico using records from 1984-2012. Results at the two SNOTEL stations indicated clear increasing trends in temperature and decreasing trends in precipitation and maximum SWE.

Temperature changes include warmer winters (+1.0-1.3 °C/decade), and generally warmer year round temperatures (+1.2-1.4 °C/decade). Precipitation changes include, a decreasing trend in precipitation during the winter (-41.6-51.4 mm/decade), during the year (-69.8-73.2 mm/decade) and max SWE (-33.1-34.7 mm/decade). At the upper Jemez River Basin ,all the water partitioning components showed statistical significant decreasing trends including precipitation

(-61.7mm/decade), discharge (-17.6 mm/decade) and vaporization (-45.7 mm/decade). Similarly,

Q

50

an indicator of snowmelt timing is occurring -4.3 days/decade earlier. Basin scale precipitation (R

2

=0.56; p=0.003) and baseflow (R

2

=0.41; p=0.02) were the strongest controls on

NPP variability indicating that forest productivity in the upper Jemez River Basin is water limited. An increasing trend in Horton index suggests that water limitation and vegetation water use are increasing in the basin. This study showed a positive correlation between water availability and EEMT. For every 10 mm of change in baseflow, EEMT varies proportionally in

0.6-0.7 MJ m

-2 year

-1

. From 1984-2012 changes in climate, water availability, and NPP have influenced EEMT in the upper Jemez River Basin. A decreasing trend in EEMT of 1.2 to 1.3 MJ m

-2

decade

-1 was calculated in this same time frame. As the landscape moves towards a drier and hotter climate, changes in EEMT of this magnitude are likely to influence critical zone processes.

6.0 ACKNOWLEDGEMENTS

We thank the funding provided by the NSF-supported Jemez River Basin and Santa Catalina

Mountains Critical Zone Observatory EAR-0724958 and EAR-1331408).

163

164

7.0 REFERENCES

Allen, C., M. Savage, D. Falk, K. Suckling, T. Swetnam, T. Schulke, P. Stacey, P. Morgan, M.

Hoffman, and J. Klingel (2002), Ecological restoration of Southwestern ponderosa pine ecosystems: A broad perspective, Ecol. Appl., 12, 1418-1433.

Allen, R., R. Peet, and W. Baker (1991), Gradient Analysis of Latitudinal Variation in Southern

Rocky-Mountain Forests, J. Biogeogr., 18, 123-139.

Amundson, R., D. D. Richter, G. S. Humphreys, E. G. Jobbagy, and J. Gaillardet (2007),

Coupling between biota and earth materials in the Critical Zone, Elements, 3, 327-332.

Anderson, S. P., F. von Blanckenburg, and A. F. White (2007), Physical and chemical controls on the Critical Zone, Elements, 3, 315-319.

Anderson-Teixeira, K. J., J. P. Delong, A. M. Fox, D. A. Brese, and M. E. Litvak (2011),

Differential responses of production and respiration to temperature and moisture drive the carbon balance across a climatic gradient in New Mexico, Global Change Biol., 17, 410-424.

Arnold, J. and P. Allen (1999), Automated methods for estimating baseflow and ground water recharge from streamflow records, J. Am. Water Resour. Assoc., 35, 411-424.

Bales, R. C., N. P. Molotch, T. H. Painter, M. D. Dettinger, R. Rice, and J. Dozier (2006),

Mountain hydrology of the western United States, Water Resour. Res., 42, W08432.

Barnett, T., R. Malone, W. Pennell, D. Stammer, B. Semtner, and W. Washington (2004), The effects of climate change on water resources in the west: Introduction and overview, Clim.

Change, 62, 1-11.

Barnett, T., J. Adam, and D. Lettenmaier (2005), Potential impacts of a warming climate on water availability in snow-dominated regions, Nature, 438, 303-309.

165

Betts, R. A. et al. (2007), Projected increase in continental runoff due to plant responses to increasing carbon dioxide, Nature, 448, 1037-U5.

Boisvenue, C. and S. Running (2006), Impacts of climate change on natural forest productivity - evidence since the middle of the 20th century, Global Change Biol., 12, 862-882.

Brantley, S. L., M. B. Goldhaber, and K. V. Ragnarsdottir (2007), Crossing disciplines and scales to understand the Critical Zone, Elements, 3, 307-314.

Brooks, P. D., P. A. Troch, M. Durcik, E. Gallo, and M. Schlegel (2011), Quantifying regional scale ecosystem response to changes in precipitation: Not all rain is created equal, Water Resour.

Res., 47, W00J08.

Cayan, D., M. Tyree, K. E. Kunkel, C. Castro, A. Gershunov, J. Barsugli, A. J. Ray, J. Overpeck,

M. Anderson, J. Russell, B. Rajagopalan, I. Rangwala, and P. Duffy. 2013. “Future Climate:

Projected Average.” In Assessment of Climate Change in the Southwest United States: A Report

Prepared for the National Climate Assessment, edited by G. Garfin, A. Jardine, R. Merideth, M.

Black, and S. LeRoy, 101–125. Washington, DC: Island Press.

Chorover, J. et al. (2011), How Water, Carbon, and Energy Drive Critical Zone Evolution: The

Jemez-Santa Catalina Critical Zone Observatory, Vadose Zone Journal, 10, 884-899.

166

Christensen, L., C. L. Tague, and J. S. Baron (2008), Spatial patterns of simulated transpiration response to climate variability in a snow dominated mountain ecosystem, Hydrol. Process., 22,

3576-3588.

Clow, D. W. (2010), Changes in the Timing of Snowmelt and Streamflow in Colorado: A

Response to Recent Warming, J. Clim., 23, 2293-2306.

Daly, C., W. Gibson, G. Taylor, G. Johnson, and P. Pasteris (2002), A knowledge-based approach to the statistical mapping of climate, Climate Research, 22, 99-113.

Eckhardt, K. (2005), How to construct recursive digital filters for baseflow separation, Hydrol.

Process., 19, 507-515.

Folland C.K., T.R. Karl, J.R. Christy, R.A. Clarke, G.V. Gruza, J. Jouzel, M.E. Mann, J.

Oerlemans, M.J. Salinger and S.W. Wange (2001) Observe climate variability and change, In: climate change 2001: the scientific basis. Contribution of working group I to the third assessment report of the Intergovernmental panel on Climate change, ed. J.T. Houghton et al., Cambridge

Uni. Press, 2001

Haan C.T., 1977, Statistical methods in hydrology. The Iowa State University Press, pp 378.

Hamlet, A., P. Mote, M. Clark, and D. Lettenmaier (2005), Effects of temperature and precipitation variability on snowpack trends in the western United States, J. Clim., 18, 4545-

4561.

167

Harpold, A., P. Brooks, S. Rajagopal, I. Heidbuchel, A. Jardine, and C. Stielstra (2012), Changes in snowpack accumulation and ablation in the intermountain west, Water Resour. Res., 48,

W11501.

Hughes, M. K. and H. F. Diaz (2008), Climate variability and change in the drylands of Western

North America, Global Planet. Change, 64, 111-118.

Huxman, T. et al. (2004), Convergence across biomes to a common rain-use efficiency, Nature,

429, 651-654.

McCabe, G. and M. Clark (2005), Trends and variability in snowmelt runoff in the western

United States, J. Hydrometeorol., 6, 476-482.

Milly, P., K. Dunne, and A. Vecchia (2005), Global pattern of trends in streamflow and water availability in a changing climate, Nature, 438, 347-350.

Mote, P., A. Hamlet, M. Clark, and D. Lettenmaier (2005), Declining mountain snowpack in western north America, Bull. Am. Meteorol. Soc., 86, 39-+.

Nayak, A., D. Marks, D. G. Chandler, and M. Seyfried (2010), Long-term snow, climate, and streamflow trends at the Reynolds Creek Experimental Watershed, Owyhee Mountains, Idaho,

United States, Water Resour. Res., 46, W06519.

Newman, B. D., B. P. Wilcox, S. R. Archer, D. D. Breshears, C. N. Dahm, C. J. Duffy, N. G.

McDowell, F. M. Phillips, B. R. Scanlon, and E. R. Vivoni (2006), Ecohydrology of waterlimited environments: A scientific vision, Water Resour. Res., 42, W06302.

Lieth, H (1975). Modeling the primary productivity of the world. P. 237-263. In H. Lieth and

168

R.H. Whittaker (ed.) Primary productivity of the biosphere, Springer-Verlag, New York.

Lyne, V., and M. Hollick (1979), Stochastic time-variable rainfall-runoff modelling, in Hydrol.

And Water Resources. Syp., publ.79/10, pp.89-92, Inst. Eng. Austr. Natl. Conf., Perth, Australia

Ohmura, A. and M. Wild (2002), Is the hydrological cycle accelerating?, Science, 298, 1345-

1346.

Pelletier, J. D. et al. (2013), Coevolution of nonlinear trends in vegetation, soils, and topography with elevation and slope aspect: A case study in the sky islands of southern Arizona, Journal of

Geophysical Research-Earth Surface, 118, 741-758.

Pelletier, J. D. and C. Rasmussen (2009a), Geomorphically based predictive mapping of soil thickness in upland watersheds, Water Resour. Res., 45, W09417.

Pelletier, J. D. and C. Rasmussen (2009b), Quantifying the climatic and tectonic controls on hillslope steepness and erosion rate, Lithosphere, 1, 73-80.

Phillips, J. D. (2009), Biological Energy in Landscape Evolution, Am. J. Sci., 309, 271-289.

Rasmussen, C. (2012), Thermodynamic constraints on effective energy and mass transfer and catchment function, Hydrology and Earth System Sciences, 16, 725-739.

Rasmussen, C. and E. L. Gallo (2013), Technical Note: A comparison of model and empirical measures of catchment-scale effective energy and mass transfer, Hydrology and Earth System

Sciences, 17, 3389-3395.

169

Rasmussen, C., R. Southard, and W. Horwath (2005), Modeling energy inputs to predict pedogenic environments using regional environmental databases, Soil Sci. Soc. Am. J., 69, 1266-

1274.

Rasmussen, C. and N. J. Tabor (2007), Applying a quantitative pedogenic energy model across a range of environmental gradients, Soil Sci. Soc. Am. J., 71, 1719-1729.

Rasmussen, C., P. A. Troch, J. Chorover, P. Brooks, J. Pelletier, and T. E. Huxman (2011), An open system framework for integrating critical zone structure and function, Biogeochemistry,

102, 15-29.

Rasmussen C., Pelletier J.D., Troch P.A., Swetnam T.L., Chorvoer J (2015) Quantifying topographic and vegetation effects on the transfer of energy and mass to the critical zone. Vadose

Journal.

Scanlon, B. R., K. E. Keese, A. L. Flint, L. E. Flint, C. B. Gaye, W. M. Edmunds, and I.

Simmers (2006), Global synthesis of groundwater recharge in semiarid and arid regions, Hydrol.

Process., 20, 3335-3370.

Seager, R. et al. (2007), Model projections of an imminent transition to a more arid climate in southwestern North America, Science, 316, 1181-1184.

Sen P. K (1968), Estimates of the regression coefficient based on Kendall’s tau, J. Am. Stat.

Assoc., 63, 1379-1389.

Sheppard, P., A. Comrie, G. Packin, K. Angersbach, and M. Hughes (2002), The climate of the

US Southwest, Climate Research, 21, 219-238.

170

Shevenell, L., Goff F., vuataz F., Trujillo P.E., Counce D., Janik C and W. Evans (1987),

Hydrogeochemical data for thermal and nonthermal waters and gases of the Valles Caldera –

Southern Jemez Mountains Region. New Mexico. Technical report. Los Alamos National Lab.

NM.LA-10923-OBES.

Steffen, W., P. J. Crutzen, and J. R. McNeill (2007), The Anthropocene: Are humans now overwhelming the great forces of nature, Ambio, 36, 614-621.

Stewart, I., D. Cayan, and M. Dettinger (2004), Changes in snowmelt runoff timing in western

North America under a 'business as usual' climate change scenario, Clim. Change, 62, 217-232.

Tague, C., K. Heyn, and L. Christensen (2009), Topographic controls on spatial patterns of conifer transpiration and net primary productivity under climate warming in mountain ecosystems, Ecohydrology, 2, 541-554.

Thornthwaite, C. W. (1948), An Approach Toward a Rational Classification of Climate, Geogr.

Rev., 38, 55-94.

Troch, P. A., G. F. Martinez, V. R. N. Pauwels, M. Durcik, M. Sivapalan, C. Harman, P. D.

Brooks, H. Gupta, and T. Huxman (2009), Climate and vegetation water use efficiency at catchment scales, Hydrol. Process., 23, 2409-2414.

Trujillo, E., N. P. Molotch, M. L. Goulden, A. E. Kelly, and R. C. Bales (2012), Elevationdependent influence of snow accumulation on forest greening, Nature Geoscience, 5, 705-709. van Mantgem, P. J. et al. (2009), Widespread Increase of Tree Mortality Rates in the Western

United States, Science, 323, 521-524.

Voepel, H., B. Ruddell, R. Schumer, P. A. Troch, P. D. Brooks, A. Neal, M. Durcik, and M.

Sivapalan (2011), Quantifying the role of climate and landscape characteristics on hydrologic partitioning and vegetation response, Water Resour. Res., 47, W00J09.

171

Westerling, A. L., H. G. Hidalgo, D. R. Cayan, and T. W. Swetnam (2006), Warming and earlier spring increase western US forest wildfire activity, Science, 313, 940-943.

Yue, S., P. Pilon, and G. Cavadias (2002), Power of the Mann-Kendall and Spearman's rho tests for detecting monotonic trends in hydrological series, Journal of Hydrology, 259, 254-271.

Zapata-Rios, X.,Troch P.A., Brooks P.D., McIntosh J, (in prep), Influence of terrain aspect on water partitioning, vegetation structure, and vegetation greening in high elevation catchments in northern New Mexico.

Zapata-Rios, X., McIntosh, J. Rademacher L., Troch P.A., Brooks P.D.,Rasmussen C., Chorover

J (submitted) Climatic and landscape controls on water transit times and silicate mineral weathering in the critical zone. Water Resources Research.

Zhao, M. and S. W. Running (2010), Drought-Induced Reduction in Global Terrestrial Net

Primary Production from 2000 through 2009, Science, 329, 940-943.

172

8.0 FIGURES

Figure 1. a) Relative location of study area within the northwestern state of New Mexico, b) upper Jemez River Basin, ~1200 km

2

, delimited above the USGS gauge station “Jemez River near Jemez” (USGS 08324000) based on a 10 m digital elevation model (DEM).

173 a)

900

800

700

600

500

400

300

90

1980 1985 1990 1995 2000 2005 2010

80 c)

70

60

50

40

30

20

10

1980 1985 1990 1995 2000 2005 2010

800

700

600

500

400

300 e)

200

1980 1985 1990 1995 2000 2005 2010

0.98

g)

0.96

0.94

0.92

0.90

0.88

0.86

1980 1985 1990 1995 2000 2005 2010

Time (years)

30

20

10

0 b)

140

120

100

80

60

40

20

0

1980 1985 1990 1995 2000 2005 2010

50 d)

40

Precipitation m = -6.17

p=0.02 (MKT)

Discharge m = -1.76

p=0.001 (MKT)

Base flow m = -1.24

p<0.001 (MKT)

Quickflow m = -0.51

p=0.005 (MKT)

900

1980 1985 1990 1995 2000 2005 2010 f)

800

700

600

500

400

300

1980 1985 1990 1995 2000 2005 2010

220 h)

210

200

190

180

170

160

1980 1985 1990 1995 2000 2005 2010

Vaporization (V) m = -4.57

p=0.04 (MKT)

Wetting (W) m = -5.67

p=0.02 (MKT)

HI m = 0.0014

p=0.002 (MKT)

Q50 m = -0.43

p=0.03 (MKT)

174

Figure 2. Precipitation and water partitioning at the upper Jemez River catchment scale. There was a significant decreasing trend quantified by the Mann-Kendall test (MKT) in the Jemez

River Basin precipitation and all the components of the water partitioning. For instance, precipitation at the catchment scale decreased during the last three decades at a rate of 6.17 mm per year and discharge at 1.76 mm per year. From 1984-2012, Horton index (HI) increased as the basin dried up, indicating that vegetation used more of its available water with increasing water limitation in the basin. Q

50

indicated that discharge is occurring 4.3 days earlier per decade.

175

0.60

0.55

0.50

0.45

0.40

0.35

0.30

R

2

=0.56

p=0.003

0.25

300 400 500 600 700 800

P Jemez River (mm/year)

0.60

0.55

0.50

0.45

0.40

0.35

R

2

=0.41

p=0.02

0.30

10 20 30 40

Baseflow (mm/year)

50

Figure 3. a) Positive linear correlation between precipitation in the upper Jemez River Basin and annual NPP in the upper Jemez River Basin derived from MODIS; b) Linear correlation between baseflow and annual NPP in the upper Jemez River Basin. Forest productivity is water limited in the upper Jemez River Basin. Other variables such as annual, winter and summer air temperature did not correlate with NPP.

176

18 a)

16

14 m= -0.124

p<0.01 (MKT)

12

10

8

6

4

2 m= -0.129

p<0.05 (MKT)

0

1980 1985 1990 1995 2000 2005 2010

Water years

EEMT emp

EEMT model

12

10

8

6

4

18 b)

16

R

2

=0.42

p=0.0002

14

1:1

4 6 8 10 12 14 16 18

EEMT emp

(MJ.m

-2

.year

-1

)

Figure 4. a) EEMT emp

and EEMT model

showed similar significant decreasing trends from 1984-

2012 of 1.2 and 1.3 MJ m

-2

year

-1

b) EEMT emp

and EEMT model

showed a significant linear correlation.

177

16

14

12

10

8

6

4

2

R

2

=0.61

p<0.0001

m=0.07

R

2

=0.21

p<0.01

m=0.06

20 40 60

Baseflow (mm)

80

EEMT emp

EEMT model

Figure 5. Relationship between water availability and EEMT. Baseflow and EEMT showed a positive linear correlation. As water availability in the Jemez River basin decreases indicated by baseflow, EEMT also decreases.

9.0 TABLES

Table 1. Land use classification of the Jemez River Basin area. 79.7% of the total basin is covered by forest according to the National Land Cover Database (NLCD)

[http://www.mrlc.gov/nlcd06_leg.php]

Land use class

Evergreen forest

Deciduous forest

Mixed forest

Grassland/herbaceous

Shrub/scrub

Pasture/Hay

Barren land (rock, sand, clay)

Developed

Cultivated crops

Wetlands

Open water

Total

Area

(Km2) %

847.7 69.60

92.6 7.61

29.8 2.44

128.0 10.51

85.0

1.8

6.98

0.14

1.3

6.1

0.1

25.2

0.10

0.50

0.01

2.07

0.4 0.03

1218.0 100.00

178

179

Table 2. Site and meteorological information for the SNOTEL Quemazon and Señorita Divide

#2 stations located at high elevations in the upper part of the Jemez River Basin.

Station

Id

Station

Name

708 Quemazon

744

Senorita

Divide #2

Elevation

(m)

2896

Mean Air

Temperature

(˚C)

Mean

Precipitation

(mm)

Latitude

(N)

Longitude

(W)

Active since YearΗ‚ Winter†

35.92° -106.39° 1980 3.98 -0.87

YearΗ‚

700.78

Winter†

347.45

Max

SWE

(mm)

242.53

2622 36.00° -106.83° 1980 4.23

Note:

The analysis of precipitation since WY 1984

Η‚Water Year: Oct 1st to Sep 30th

†Winter: Oct 1st to March 31 st

Temperature data availability since 1989 for the Quemazon and 1988 for the Senorita Divide #2 station

-0.90 685.98 422.87 239.20

180

Table 3. Climatic time series trends for the Quemazon and Señorita Divide #2 SNOTEL stations from 1984-2012. A trend in the precipitation time series was evaluated with the Mann-Kendall test (MKT) and Sen’s slope estimator. Trends were considered statistically significant at p≤0.1.

The results showed an increasing trend in winter, summer and annual temperature in the two stations. Annual and winter precipitation, maximum SWE and 1 st

of April SWE decreased in both stations during the 29 years analyzed. The last day of snow cover decreases significantly only at the Quemazon station. No significant trend was observed for the SWE: winter P ratio, duration of snowmelt SM50 and length of snow on the ground.

Variable

Winter Temp

Summer Temp

Annual temp

Annual Precip(mm)

Winter Precip (mm)

Max SWE (mm)

SWE:winter P ratio

1 April SWE

Max SWE day

SM50 (days)

1st day snow cover

(day) last day snow cover

(day) snow on ground (days)

Quemazon

Q Sen's slope estimator

Sig

0.13 ***

0.10 **

0.14 ***

-6.98 **

-4.16 +

-3.31 +

-0.005

-6.05 *

-0.57 *

-0.02

Señorita Divide #2

Q Sen's slope estimator

Sig

0.10 *

0.10 **

0.12 ***

-7.32 *

-5.94 *

-3.47 +

-0.002

-5.44 +

-0.33

0.12

-0.50

-0.65 *

-0.12

0.17

-0.31

-0.60

†Statistical significance

+ P<0.1

** P < 0.01

*** P < 0.001

181

Table 4. Discharge predictors for the Jemez River basin based on climate variables recorded at

Quemazon and Señorita Divide#2 SNOTEL stations. Annual temperature, max SWE and the length of snow on the ground were the best predictors of discharge in the basin. The predictability power of discharge was similar from climatic variables recorded at the Quemazon and Señorita Divide#2 stations. Annual temperature and max SWE climatic variables had a decreasing trend that influenced the decrease in water availability in the basin.

Intercept

Annual Temp (˚ C)

Max SWE (mm)

Snow on the ground (days)

R

2

p

Quemazon

p values

Señorita Divide#2

p values

-7.57

-7.23

0.14

0.32

0.81

<0.0001

0..071

0.0035

0.0003

0.03

37.75

-3.5

0.21

-0.18

0.80

<0.0001

0.0128

0.07

0.0001

0.05

182

10.0 SUPPLEMENTARY INFORMATION

60

40

20

R

2

=0.35

p=0.0004

0

-20

-40

R2=0.61

p<0.0001

-60

0 100 200 300 400 500 600

Max SWE (mm)

60

40

20

0

-20

R

2

=0.38

p=0.0002

-40

-60

R

2

=0.63

p<0.0001

0 50 100 150 200 250

Snow on ground (days)

Quemazon

Señorita

Figure S1. Plot of residuals between max SWE and snow on the ground from the linear model presented in Figure 2b. Maximum SWE and duration of the snow cover are the better predictors of discharge residuals variability. Q residuals increase during extreme dry and wet years.

Table S1. Empirical and modelled EEMT values estimated for the upper Jemez River basin.

Ebio emp

was estimated by multivariable linear regression from annual Precipitation at the

Quemazon station and Jemez River basin between 1984-1999 (R

2

=0.75; p=0.0009)

Water year

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

1992

1993

1994

1995

1996

1997

1998

1999

1984

1985

1986

1987

1988

1989

1990

1991

EEMT emp

EEMT model

Eppt emp

Ebio emp

EEMT emp

Eppt model

Ebio model

EEMT model

1.28 11.27

2.37 12.43

12.55

14.80

0.05

0.20

5.09

5.47

5.14

5.67

1.42 12.48

1.60 11.15

1.16 11.21

0.87 9.28

13.90

12.75

12.37

10.15

0.19

0.09

0.14

0.05

9.34

8.71

8.52

4.18

9.53

8.80

8.66

4.24

0.80 11.77

1.35 13.61

1.77 11.47

1.49 11.43

0.75 11.96

1.74 11.93

0.33 10.13

1.37 12.12

12.56

14.96

13.24

12.93

12.71

13.67

10.46

13.48

0.14

0.27

0.14

0.07

0.15

0.19

0.02

0.11

5.45

14.22

9.11

8.51

8.79

8.72

4.94

7.83

5.58

14.49

9.26

8.58

8.94

8.91

4.96

7.94

1.04 10.94

1.04 11.47

0.60 8.42

1.09 10.20

0.35

0.62

8.36

9.67

0.77 10.03

1.30 10.98

0.48 11.08

1.00 12.56

0.88 10.45

0.65 9.39

0.73 10.39

0.39 8.43

0.50 8.65

11.98

12.51

9.02

11.30

8.71

10.28

10.81

12.28

11.56

13.57

11.33

10.03

11.13

8.82

9.15

0.04

0.21

0.06

0.08

0.05

0.04

0.18

0.08

0.03

0.06

0.01

0.09

0.08

0.03

0.03

4.98

10.90

5.35

5.73

5.78

5.95

5.89

5.66

5.23

5.74

5.24

6.03

5.20

4.29

4.12

5.26

5.80

5.24

6.12

5.29

4.31

4.16

5.02

11.11

5.42

5.81

5.83

5.99

6.07

5.74

183

184

Table S2. Relationship between climatic variables and discharge in the Jemez River Basin (1984-

2012) based on records from the Quemazon and Señorita Divide#2 SNOTEL stations. The variables are listed in decreasing order according to the linear R

2

value from the Señorita

Divide#2 station.

Variable

Winter P (mm)

1 April SWE

Quemazon station slope R

2

p

Discharge

Señorita Divide#2 station slope R

2

p

0.18 0.24 0.00550 0.19 0.72 0.00001

0.18 0.64 0.00010 0.16 0.56 0.00001

0.23 0.71 0.00010 0.19 0.55 0.00001 Max SWE (mm) last day snow cover

(day)

Annual P (mm)

2.01 0.66 0.00001 1.83 0.54 0.00001

0.10 0.19 0.01480 0.13 0.50 0.00001

-

11.03 0.27 0.00850

-

10.90 0.49 0.00010 Annual temp

50% max SWE day

Summer Temp

Max SWE day

1st day snow cover

(day)

1.92 0.48 0.00010 1.71 0.41 0.00010

-2.04 0.01 0.58740 -4.94 0.32 0.00320

Winter Temp -8.57 0.17 0.04400 -9.58 0.29 0.00530 snow in ground (days) 0.88 0.44 0.00001 0.37 0.25 0.00390

1.15 0.31 0.00120 0.65 0.13 0.04800

-0.51 0.09 0.10290 -0.30 0.12 0.05570

SWE:winter P ratio

SM50 (days)

69.29 0.40 0.00020 52.28 0.09 0.09240

-0.22 0.00 0.73350 0.24 0.01 0.58760

185

Table S3. Relationship between climatic variables and Horton index from the Jemez River basin

(1984-2012) based on records from the Quemazon and Señorita Divide#2 SNOTEL stations.

The variables are listed in decreasing order according to the linear R

2

value from the Señorita

Divide#2 station.

Variable

Winter P (mm)

1 April SWE

Max SWE (mm) last day snow cover

(day)

Annual P (mm)

Annual temp

50% max SWE day

Summer Temp

Winter Temp snow in ground (days)

Max SWE day

1st day snow cover

(day)

SWE:winter P ratio

SM50 (days)

Quemazon station slope R

2

Horton index

Señorita Divide#2

p

station slope R

2

p

-

0.0001 0.22 0.00850

-

0.0001 0.60 0.00001

-

0.0001 0.59 0.00001

-

0.0001 0.55 0.00001

-

-

0.0002 0.67 0.00001

-

0.0017 0.62 0.00001

-

0.0001 0.09 0.09940

0.0002 0.59 0.00001

-

0.0016 0.52 0.00001

-

0.0001 0.29 0.00160

0.0103 0.32 0.00430 0.0086 0.40 0.00070

-

0.0017 0.50 0.00001

-

0.0015 0.41 0.00010

0.0019 0.01 0.57230 0.0038 0.25 0.01160

0.0079 0.19 0.03300 0.0079 0.26 0.00910

-

0.0007 0.34 0.00060

-

0.0010 0.33 0.00070

0.0004 0.31 0.00100

-

0.0006 0.15 0.03140

0.0003 0.04 0.26470 0.0003 0.18 0.01890

-

0.0588 0.39 0.00020

0.0002 0.01 0.64710

0.0620 0.18 0.01700

-

0.0001 0.00 0.74040

APPENDIX D:

NOBLE GASES AND SF

6

CONCENTRATIONS FROM SPRINGS AROUND

REDONDO PEAK, NEW MEXICO

186

Figure 1. Springs draining Redondo Peak

1.0 NOBLE GASES

Recharge elevation, recharge temperature and excess air was calculated based on noble gases concentrations from springs (table 1). These variables were inferred applying inverse modeling and the known concentrations of Ne, Ar, Kr and Xe according to the methodology described in

Aeschbach-Hertig et al. (1999) (Table 2). Water recharge around Redondo Peak (3435 m) occurs in all the springs at high elevations above 2970 meters and above 3200 m in 7 springs. The annual air temperature recorded in Redondo station at 3200 m in elevation was 3.8˚C, and water recharge temperatures were estimated between 4.5 and 9.7˚C. There were three springs including Redondo

Meadow, East spring and Redondo spring 3 with temperatures above 7˚C (Figure 2).

Table 1. Noble gases concentrations from the springs

Sample

I.D.

Spring Name Spring

ID

Ar total

(ccSTP/g)

Ne total

(ccSTP/g)

Kr total

(ccSTP/g)

Xe total

(ccSTP/g)

He4

(ccSTP/g)

187

NMS1129

South spring

NMS1130

Top of la Jara

NMS1134

Upper Redondo pond

NMS1135

Upper Redondo spring 3

NMS1136

Upper Jaramillo wall

NMS1137

Upper Jaramillo spring 3

NMS1138

La Jara seep

NMS1140

East spring

NMS1141

Top of History Grove

LJ2

LJ1

UR1

UR3

UJ1

UJ2

LJ4

Es

HGs

2.91E-04

2.90E-04

2.92E-04

2.74E-04

2.84E-04

2.82E-04

2.97E-04

2.77E-04

2.80E-04

1.56E-07

1.55E-07

1.55E-07

1.49E-07

1.53E-07

1.59E-07

1.58E-07

1.52E-07

1.52E-07

6.92E-08

6.99E-08

6.97E-08

6.49E-08

7.09E-08

7.04E-08

7.39E-08

6.76E-08

7.03E-08

1.03E-08

1.00E-08

9.95E-09

9.40E-09

1.03E-08

1.01E-08

1.10E-08

9.77E-09

1.02E-08

3.70E-08

3.62E-08

3.67E-08

3.80E-08

3.62E-08

3.73E-08

3.67E-08

3.59E-08

3.59E-08

Table 2. Recharge elevation, temperature and excess air based on an analysis of Ne, Ar, Kr and

Xe

Spring

South spring

Top of la Jara

Upper Redondo pond

Upper Redondo spring 3

Upper Jaramillo wall

Upper Jaramillo spring 3

La Jara seep

East spring

Top of History Grove

Spring discharge elevation

(m)

Spring

ID

3069 LJ2

3195 LJ1

2839 UR1

2917 UR3

2848 UJ1

2876 UJ2

2817 LJ4

2860 Es

2908 HGs

Recharge elevation

(m)

3200

3336

3433

3252

3323

3222

3043

2979

2978

Temp

(˚C)

5.93

5.71

5.43

7.92

4.90

6.32

4.53

7.77

6.61

Excess Air

A cc/kg = ml

0.84

0.91

1.03

0.70

0.77

1.03

0.70

0.58

0.43

188

10.00

9.00

8.00

7.00

6.00

5.00

4.00

3.00

2.00

2700 2900 3100 3300

Recharge elevations (m)

3500

Recharge Temp.

Ave. Annual Air Temp

Figure 2. Recharge elevation and recharge temperature in springs inferred with noble gases concentrations

2.0 SF

6

SF

6

concentrations from springs can be grouped in two classes (Figure 4). There is a first group with SF

6

concentrations above 200 fg/kg and a second group with concentrations below 150 fg/kg.

These differences in SF

6

concentrations between the two groups define differences in water ages

(Table 3). SF

6

water ages derived from the group 1 were very similar to the water ages estimated with tritium (reported in Appendix A). Understanding the differences in SF

6

concentrations from the two groups needs further research and spring’s resampling. Moreover, there was an inverse relationship between SF

6

concentrations and excess air (Figure 4). Figure 5 and Figure 6 shows tritium ages versus SF

6

group 1 and group 2. Although these plots show large difference in ages,

SF

6

ages were strongly correlated with cations e.g. Na+ similar to the results based on tritium ages

(Figure 7).

189

Table 3. SF

6 concentrations and Age

SF

6

Group Springs

1 Top of History Grove

1

1

1

2

Upper Redondo Pond

East Spring

Upper Jaramillo Wall (UJ2)

Upper Redondo Spring 3

2

3

3

3

La Jara Seep

South spring

Top of La Jara

Upper Jaramillo Spring 3 (UJ1) fg/kg

313.37

283.51

370.63

303.38

231.76

211.20

141.67

111.00

100.77

Utah Lab values

SF

6

Age

(years)

5.2

8.7

0.7

7.2

11.2

15.2

21.2

24.2

25.7 excess air (cm3

STP/Kg)

0.83

0.88

0.91

0.79

0.78

0.99

1.30

1.34

1.54

400

350

300

250

200

150

100

50

0

0.00

y = -270.69x + 514.91

R² = 0.7369

0.50

1.00

Excess air (cm3 STP/kg)

1.50

2.00

Figure 4. SF6 concentrations versus excess air in the springs. Springs can be grouped into two groups.

190

25

20

15

10 y = 1.6519x - 1.3767

R² = 0.8696

5

0

-5

0.00

2.00

4.00

6.00

8.00

10.00

12.00

Tritium Age (Arizona)

Figure 5. SF6 ages from springs with SF6 concentrations above 200 fg/kg versus Tritium ages from Arizona

45.00

40.00

35.00

30.00

25.00

20.00

15.00

10.00

5.00

0.00

0.00

y = 3.8317x - 0.3425

R² = 0.9081

2.00

4.00

6.00

8.00

10.00

12.00

Tritium Age (Arizona)

Figure 6. SF6 ages from the second group (<150 fg/kg).

250

200

150

100

50 y = 7.269x + 81.847

R² = 0.7189

0

0 5 10 15

SF6 (age)

20 25

Figure 7. Na vs SF

6

age based on the samples from the group (SF

6

>200 fg/kg)

191

192

3.0 REFERENCES

Aeschbach-Hertig, W., F. Peeters, U. Beyerle, and R. Kipfer, 1999. Interpretation of dissolved atmospheric noble gases in natural waters. Water Resour. Res. 35:2779-2792

Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement