Mountain Pine Beetle Disturbance and Climate Impacts on … · University of Colorado, Boulder CU Scholar Ecology & Evolutionary Biology Graduate Theses & Dissertations Ecology &

University of Colorado, BoulderCU ScholarEcology & Evolutionary Biology Graduate Theses &Dissertations Ecology & Evolutionary Biology

Spring 1-1-2012

Mountain Pine Beetle Disturbance and ClimateImpacts on Subalpine Forest Carbon CyclingNicole Anne TrahanUniversity of Colorado at Boulder, [email protected]

Follow this and additional works at: http://scholar.colorado.edu/ebio_gradetds

Part of the Biogeochemistry Commons, Ecology and Evolutionary Biology Commons, and theEnvironmental Indicators and Impact Assessment Commons

This Dissertation is brought to you for free and open access by Ecology & Evolutionary Biology at CU Scholar. It has been accepted for inclusion inEcology & Evolutionary Biology Graduate Theses & Dissertations by an authorized administrator of CU Scholar. For more information, please [email protected].

Recommended CitationTrahan, Nicole Anne, "Mountain Pine Beetle Disturbance and Climate Impacts on Subalpine Forest Carbon Cycling" (2012). Ecology& Evolutionary Biology Graduate Theses & Dissertations. Paper 32.

http://scholar.colorado.edu?utm_source=scholar.colorado.edu%2Febio_gradetds%2F32&utm_medium=PDF&utm_campaign=PDFCoverPages

http://scholar.colorado.edu/ebio_gradetds?utm_source=scholar.colorado.edu%2Febio_gradetds%2F32&utm_medium=PDF&utm_campaign=PDFCoverPages


http://scholar.colorado.edu/ebio?utm_source=scholar.colorado.edu%2Febio_gradetds%2F32&utm_medium=PDF&utm_campaign=PDFCoverPages


http://network.bepress.com/hgg/discipline/154?utm_source=scholar.colorado.edu%2Febio_gradetds%2F32&utm_medium=PDF&utm_campaign=PDFCoverPages



http://scholar.colorado.edu/ebio_gradetds/32?utm_source=scholar.colorado.edu%2Febio_gradetds%2F32&utm_medium=PDF&utm_campaign=PDFCoverPages

mailto:[email protected]

!

MOUNTAIN PINE BEETLE DISTURBANCE AND CLIMATE IMPACTS ON

SUBALPINE FOREST CARBON CYCLING

by

NICOLE ANNE TRAHAN

B.A., The Colorado College, 1999

M.S., University of Northern Colorado, 2005

A thesis submitted to the

Faculty of the Graduate School of the

University of Colorado, Boulder in partial fulfillment

of the requirement for the degree of

Doctor of Philosophy

Department of Ecology and Evolutionary Biology

2012

!

This thesis entitled: Mountain pine beetle disturbance and climate impacts on subalpine forest carbon cycling

written by Nicole Anne Trahan has been approved for the

Department of Ecology and Evolutionary Biology

–––––––––––––––––––––––––––––––––––––––––––––––––– Russell K. Monson

_________________________________________________ Steven K. Schmidt

Date_______________

The final copy of this thesis has been examined by the signatories, and we Find that both the content and the form meet acceptable presentation standards

Of scholarly work in the above mentioned discipline.

!

! """!

Trahan, Nicole Anne (Ph.D., Ecology and Evolutionary Biology)

Mountain pine beetle disturbance and climate impacts on subalpine forest carbon cycling

Thesis directed by Professor Russell K. Monson

ABSTRACT

Lodgepole pine (Pinus contorta) subalpine forests are an important regional carbon sink

dependent on winter precipitation. Understanding how climate variability and disturbance impact

the carbon cycle and biogeochemical pools is critical for predicting future carbon sequestration.

In this thesis, I investigated the impacts of mountain pine beetle (MPB) induced tree mortality

and changes in spring snow depth on forest carbon balance and soil biogeochemical pools. Using

parallel disturbance chronosequences of natural and simulated beetle-kill, as well as 8-year

records of gross primary productivity (GPP) and respiration fluxes, I found that soil respiration

sharply declines with GPP after tree mortality, reflecting the loss of autotrophic respiration and

rhizodeposition. During this time, the forest soil also lost a significant fraction of dissolved

organic carbon (DOC) and dissolved organic nitrogen (DON). Five to six years after tree

mortality, soil respiration rates, DOC, and DON recovered before declining again, co-incident

with an increase in decomposing needle litter. Generally, tree mortality caused an increase in soil

moisture and a decrease in soil temperature, although these effects were confounded by stand

density and time since disturbance. MPB disturbance lead to an increase in extractable soil

ammonium, but not nitrate concentrations, and this accumulation was correlated with increased

soil moisture availability and decreases in microbial biomass C:N. Ultimately, consideration of

the dynamics of labile C supply suggest that beetle-killed forests lose less carbon to the

!

! "#!

atmosphere than previously estimated. Finally, I investigated the effects of spring snow pack

depth on the autotrophic and heterotrophic components of subalpine forest soil respiration over

two summer growing seasons. I found that snow additions can transiently increase growing

season soil respiration rates likely through enhanced rhizodeposition, and that the autotrophic

component of soil respiration may be sensitive to the effect of snow moisture availability on the

photosynthetic water-use efficiency of trees. These results collectively demonstrate the

significant effects of climate variability and disturbance on the mechanistic linkages between

gross primary productivity and ecosystem respiration.

!

!

DEDICATION

Mom and Dad, this one’s for you.

!

! #"!

ACKNOWLEDGEMENTS

I have been fortunate in the support and good humor of many people throughout graduate

school. First and foremost I would like to thank my advisor, Russ Monson for providing me

with the mentorship and wonderful opportunities that helped me develop as a scientist. I would

like to also thank my committee members, Dave Moore, Bill Bowman Steve Schmidt and Alan

Townsend. I have learned so much from each of you, and deeply appreciated the support I have

received. A special thank you Mike Weintraub and Dave Bowling who took me into their labs

and taught me how to get things done. The mentorship and generosity of the post-doctoral

researchers in my lab was tremendous, a big thank you to Laura Scott-Denton, Todd Rosenstiel,

Maggie Prater, Diego Riveros-Iregui, Michael Wilkinson, and Allyson Eller. To all my partners

in crime in the Monson lab for both the science and the laughter, thank you! An especially

heartfelt thanks goes to Amy Trowbridge, Jia Hu, Sean Burns, and Lindsay Young.

It goes without saying that my projects would not have been possible without some truly

amazing field crews. Thanks to Stan Lee, who turned out to be just as stubborn as I was. I owe

you one. Thanks to Emily Dynes, Ben Brayden, Claire Gibson, Jennifer Morse, Mitch Krem,

Nate Monson, Ryan Weaver, Meredith Casciato, Jeff Beauregard, Sierra Love-Stowell, and

Danielle Koffler- your hard work, good spirits, and friendship kept me going! Funding from

CIRES, the UGGS, and the EBIO department was also tremendously helpful. The staff and

infrastructure at the Mountain Research Station went above and beyond.

Thanks to Jane Zelikova, Evan Pugh, Ben Brayden (again), Courtney Meier, Anthony

Darouzett-Nardi for collaborations in science and otherwise! And finally to my family and good

friends who were always there, you are very much appreciated.

!

! #""!

TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION……………………………………………………………….1 CHAPTER 2: CHANGES IN CARBON BALANCE AFTER INSECT

DISTURBANCE IN WESTERN U.S. FORESTS……………………………………..5

2.1 INTRODUCTION…………………………………………………………………….6

2.2 METHODS……………………………………………………………………………8

2.2.1 Site descriptions……………………………………………………………..8 2.2.2 Estimating GPP from satellite observations of EVI, Land surface

temperature, and Eddy Covariance Observations……………………….11

2.2.3 Processing of MODIS data and use of basic GPP model………………….11 2.2.4 Satellite data pre-processing………………………………………………14 2.2.5 Bayesian calibration using Metropolis-Hastings………………………….14 2.2.6 Estimation of relative changes in valley respiration from nighttime

accumulation of CO2 at FEF…………………………………………….17 2.2.7 Establishment of disturbance and experimental chronosequences………..24 2.2.8 Soil respiration measurements and extractable dissolved organic

carbon analyses………………………………………………………….25

2.3 RESULTS…………………………………………………..………………………..26

2.4 DISCUSSION…………………………..……………………………………………31

2.5 CONCLUSION………………………………………………………………………32 CHAPTER 3: CHANGES IN SOIL MICROCLIMATE AND NUTRIENT POOLS

FOLLOWING MECHANICAL GIRDLING AND MOUNTAIN PINE BEETLE DISTURBANCE IN TWO COLORADO SUBALPINE FORESTS………………..34

3.1 INTRODUCTION…………………………………………………………………...35

3.2 METHODS…………………………………………………………………………..41

3.2.1 Study site descriptions……………………………………………………...41

!

! #"""!

3.2.2 Establishment of disturbance chronosequences…………………………...43

3.2.3 Plot microclimate, stand characteristics, litter and soil organic layer depths………………………………………………………………45

3.2.4 Soil sampling, extraction, and chemical analyses........................................47

3.2.5 Ion exchange resin capsule installation, extraction,

and chemical analyses...............................................................................48

3.2.6 Statistical analyses…………………………………………………………49

3.3 RESULTS……………………………………………………………………………50

3.3.1 Fraser beetle-kill chronosequence stand and vegetation characteristics…………………………………………………………...50

3.3.2 Soil microclimate and characteristics across the Fraser beetle-kill

and Niwot girdling tree mortality chronosequences…………………….52

3.3.3 Soil and resin extractable biogeochemical pools across the Fraser beetle- kill and Niwot girdling tree mortality chronosequences…………………55

3.3.4 Microbial biomass carbon and nitrogen pools and ratios………………...58

3.4 DISCUSSION……………………………………………………………..…………61

3.4.1 Disturbance induced changes in soil microclimate, plot characteristics, and their interactions with stand structure…………………………...….61

3.4.2 Disturbance induced changes in soil and microbial biomass

biogeochemical pools……………………………………...……………..64 3.5 CONCLUSION………………………………………………………………………71

CHAPTER 4: THE CONSEQUENCES OF CHANGES IN SPRING SNOW PACK DEPTH FOR SUPALPINE FOREST SOIL RESPIRATION……………………………..73

4.1 INTRODUCTION…………………………………………………………………...74

4.2 METHODS…………………………………………………………………………..78

4.2.1 Site description………………………………………………………...…...78

4.2.2 Girdling and snow manipulation treatments………………………………78

!

! "$!

4.2.3 Soil temperature, moisture, and site precipitation…………………………79

4.2.4 Soil respiration measurements……………………………………………..80

4.2.5 Soil extractable dissolved organic carbon and microbial biomass carbon………………………………………………..80

4.26 Assessing the !13CO2 of soil respiration……………………………………81

4.2.7 Statistical analyses…………………………………………………………82

4.3 RESULTS……………………………………………………………………………82

4.3.1 Manipulation of plot spring snow pack depth and

subsequent melt time…………………………………………………….82

4.3.2 Snow and girdling impacts on soil respiration, soil extractable dissolved organic carbon pools, and microbial biomass carbon pools, 2008……..83

4.3.3 Spring snow pack depth and girdling impacts on soil respiration,

temperature, and moisture, 2009………………………………...………86 4.3.4 Spring snow pack depth and girdling treatment impacts on the

!13C of soil respiration, 2008 and 2009…………………………………91 4.4 DISCUSSION………………………………………………………………………..93

4.5 CONCLUSION………………………………………………………………………96

CONCLUSIONS………………………………………………………………………………..98 REFERENCES CITED……………………………………………………………………….101

!

! $!

TABLES

CHAPTER 2 2.1 Original and optimized parameter values for the TG model………………………………..13 2.2 Annual sums for GPPTG and CO2 accumulation from Niwot Ridge and

Fraser Experimental Forest from 2002 through 2010……………………………………18

2.3 Mean growing season soil efflux (!mol CO2 m-2 s-1) and soil total organic carbon (!gC g-1dry soil) from plots across the NWR gridling and FEF beetle-kill chronosequences ………………………………………………………..27

CHAPTER 3 3.1 Mean 2010 plot stand characteristics across the FEF beetle-kill P. contorta

tree mortality chronosequence…………………………………………………………...51 3.2 Mean growing season (Jun-Sept 2010) organic layer microclimate and

soil properties across the FEF beetle-kill and NWT girdling P. contorta tree mortality chronosequences……………………………………………...54

3.3 Mean growing season 2010 resin extractable nutrients across the FEF beetle-kill

and NWT girdling P. contorta tree mortality chronosequences………………………...59 CHAPTER 4 4.1 Analysis of Variance results for the main effects and interactions of spring

snowpack manipulations, girdling and date on measured plot soil respiration for the 2008 and 2009 growing season………………………………………84

!

! $"!

FIGURES CHAPTER 2 2.1 Mountain pine beetle infestation in the Fraser Experimental Forest………………………..10 2.2 EVI data for MODIS pixels containing the NWR flux tower and

the FEF Head Quarters site……………………………………………………………....15 2.3 LST data for MODIS pixels containing the NWR flux tower and

the FEF Head Quarters site………………………………………………………………15 2.4 Probability density functions from the Metropolis-Hastings estimation for

the parameters required for the TG model……………………………………………….18 2.5 Average diurnal CO2 concentrations measured at FEF during June, July, and August

or December, January, and February…………………………………………………….19 2.6 Nightly average CO2 concentrations at FEF and in stable-CO2 conditions at NWR,

and a third site at Storm Peak Laboratories (SPL) in Steamboat Springs, CO…………..19 2.7 Hourly-average CO2 concentrations, wind-speed, wind direction, soil temperature

and soil moisture for two weeks in July, 2009 at FEF Headquarters tower where CO2 is measured and for a tower at 3100 m on the Fool Creek tributary to St. Louis Creek………………………………………………………………21

2.8 Time series and seasonal histograms of summer nocturnal stability for up valley and down

valley as indicated by night average virtual potential temperature gradient between a tower near valley base (Headquarters, 2760 m) and along the valley wall (Fool Creek, 3100 m)………………………………………………………..23

2.9 Frequency histograms of night-time windspeed during periods of down

valley flow for the Headquarters and Fool Creek meteorological stations from 2006 through 2010…………………………………………………………………23

2.10 Mountain pine beetle infestation, relative annual gross primary productivity

and annual respiration flux, as well as their seasonal patterns at the un-infested Niwot Ridge and the infested Fraser Experimental Forest……………….....28

2.11 Soil CO2 efflux and soil dissolved organic carbon as a function of time

since disturbance in FEF and NWR……………………………………………………...30 CHAPTER 3 3.1 Location of the disturbance chronosequence sites showing mapped

!

! $""!

mountain pine beetle infestation in 2000 and 2007……………………………………...42 3.2 Total percent transmitted light to the forest floor and mean growing season

soil temperature across the FEF chronosequence; Mean growing season soil moisture across both the NWT and FEF tree mortality chronosequences…………..43

3.3 Mean growing season 2010 extractable soil biogeochemical pools across the Fraser beetle-kill and Niwot girdling tree mortality chronosequences…………………..56

3.4 Mean growing season 2010 extractable microbial biomass carbon and nitrogen

pools and soil C:N across the Fraser beetle-kill and Niwot girdling tree mortality chronosequences………………………………………………………….60

3.5 Linear regression of the relationship between mean growing season 2010 soil

extractable ammonium and percent soil gravimetric moisture across the Fraser beetle-kill and Niwot girdling tree mortality chronosequences………………………….69

CHAPTER 4 4.1 Mean soil respiration in girdled and ungirdled plots over a treatment

gradient of spring snowpack depth, growing season 2008………………………………85 4.2 Mean growing season 2008 A) extractable dissolved organic carbon and

B) microbial biomass carbon pools in girdled and ungirdled plots……………………...87 4.3 Mean soil respiration in girdled and ungirdled plots over a treatment gradient

of spring snowpack depth, growing season 2009………………………………………..89

4.4 Mean daily ten minute average of A) plot volumetric soil moisture, B) plot soil temperature at 5 cm depth in girdled and ungirdled plots over a treatment gradient of spring snowpack depth, growing season 2009, C) Growing season site precipitation, June through September 2009………………………………...90

4.5 Mean "13C of soil respiration from June 2008- October 2009 in girdled

and ungirdled plots over a treatment gradient of spring snowpack depth, growing season 2009……………………………………………………………...92

!

! %!

CHAPTER 1

INTRODUCTION

In the last two decades the annual rate at which atmospheric carbon dioxide is rising has

tripled. Although largely driven by the escalating fossil fuel emissions, eighteen percent of this

increase has been attributed to a reduction in the efficiency of the planet’s carbon sinks (Canadell

et al. 2007). Critically, these sinks currently absorb around one-third of anthropogenic emissions,

and a reduction in their efficiency suggests our natural systems are becoming less able to

mitigate future climate change (Schimel 2007). Moreover, many coupled climate-carbon models

ascribe reduced sequestration to a positive feedback response to increased climate variability (Le

Quere et al. 2009). For example, widespread forest mortality has been increasingly linked to

global-change-type drought, where water stress and elevated temperatures precipitate regional

vegetation dieback events (Allen et al., 2010; Breshears et al., 2005; van Mantgem et al., 2009).

Ultimately, understanding how climate variability impacts the carbon cycle on local and regional

scales is critical for predicting and understanding sink behavior.

In addition to climate, ecosystem disturbances fundamentally influence carbon and

nutrient cycling. Initially, disturbance shifts an ecosystem to a carbon source, increasing CO2

emissions to the atmosphere, while recovery from disturbance is associated with greater

ecosystem carbon storage (Baldocchi 2008; Kowalski et al. 2004; Magnani et al. 2007; Odum

1969; Pregitzer and Euskirchen 2004). The effects of recovery from disturbance on carbon

sequestration can be as large or larger in magnitude than the effects of climate. Schimel et al.

!

! &!

estimated that of the total annual U.S. carbon sink of ~ 0.3 Pg of C per year, 0.1 Pg could be

attributed to climate and CO2 fertilization effects, and the rest to intensive forest management

and agricultural abandonment (Schimel et al. 2000). In spite of the long recognized importance

of disturbance and age structure on ecosystem processes, current land surface models are not

able to predict episodic disturbances mechanistically given relevant climatic triggers, or their

carbon feedback consequences to the atmosphere (Running 2008). Predicting the future of

terrestrial carbon sinks and their response to climate change requires a precise and process based

understanding of both the differential effects of local and regional climate on gross primary

productivity and ecosystem respiration, together with the localized but large effects of

disturbance.

Mountain forests above 750 m account for seventy percent of the Western U.S. carbon

sink, due to recovery from historical logging, fire suppression practices, and localized higher

moisture availability in the semi-arid west (Schimel et al., 2002; Schimel & Braswell, 2005).

Subalpine lodgepole pine (Pinus contota Dougl. ex Loud.) forests are an important component of

this sink, forming widespread monospecific stands ranging from Colorado to Northern Alberta in

Western North America. These forests are characterized by long cold winters and cool, but often

dry, short growing seasons (Fahey & Knight, 1986). Productivity in these ecosystems is a result

of precipitation, most of which falls as snow (Hunter et al., 2006; Monson et al., 2002). The

amount of late-winter snow and the timing of spring snowmelt exert the primary control over

annual rates of carbon sequestration (Hu et al., 2010; Monson et al., 2005; Monson et al., 2002;

Sacks et al., 2007). However, the role of soil respiration during this critical carbon uptake period

during and immediately following snowmelt has not been investigated.

!

! '!

Recruitment of lodgepole pines is low in understory canopy conditions, and thus these

forests are largely dependent on disturbance events such as windthrow, bark beetle outbreaks,

and harvesting for regeneration (Fahey & Kinight, 1986). Historically stand-replacing fires with

100-400 year return intervals maintained and shaped the development of lodgepole pine forests

(Romme et al., 1986), influencing the evolution of varying degrees of serotiny (Shoennagel et al.,

2003). While fire volatizes nutrients and organic matter in the short term, it ultimately creates the

canopy openings and stimulates regeneration and productivity (Turner et al., 2007). As the

western U.S. experienced economic and resource development, these forests were extensively

harvested from the 1860’s onward during periods of settlement, mining, and railroad building

(Schimel & Braswell, 2005).

Currently lodgepole pine forests are experiencing a historically unprecedented out break

of Mountain Pine Beetle (Dendroctonos ponderosae Hopkins). The mountain pine beetle is an

endemic, aggressive species of bark beetle (Coleoptera: Curculionidae: Scotlytinae), able to kill

mature trees from a wide range of species (Bentz et al., 2010; Hicke et al., 2012). In 2009, U.S.

aerial and ground surveys detected the largest outbreak spread in more than 30 years, with

mountain pine beetles actively killing trees in about 36,000 km2 of forest. In the same year,

active outbreak occurred in over 4,000 km2 of Colorado forests, as the mountain pine beetle

continued to expand east of the Continental Divide into forests of the Rocky Mountain Front

Range (Man, 2010).

Tree death due to insect disturbance alters important forest ecosystem properties such as

community composition, canopy structure, productivity and biogeochemical cycling (Fettig et al.

2007). The effects of tree mortality induced by insect disturbance on soil biogeochemical pools

are potentially variable, and past studies have concluded that the degree of disturbance is

!

! (!

determined both by forest community composition and by the nature and severity of outbreak

disturbance (Hicke et al., 2012). Most studies of insect forest disturbance and forest nutrient

cycling have focused on Lepidopteran and Homopteran pests (e.g. Frost and Hunter, 2004; Houle

et al., 2009; Stadler et al., 2006); only a handful of published studies have directly investigated

the below ground nutrient pool impacts of bark beetle induced tree mortality (Clow et al., 2011;

Griffin et al. 2011; Huber et al., 2005; Morehouse et al., 2008; Xiong et al. 2011). With the

extent of the current mountain pine beetle outbreak predicted to intensify under future climate

projections (Logan et al., 2003; Bentz et al., 2010; but see Evangelista et al., 2011), it is

important that we continue to narrow the uncertainties associated with influences on soil

biogeochemical cycles, as these influences are likely to be crucial to our management of forest

regrowth, nutrient losses into streams and rivers, and the maintenance of multi-trophic animal

communities.

My dissertation is an exploration of these themes: understanding how changes in

interannual climate variability and disturbance alter the subalpine forest carbon cycle and soil

biogeochemical pools. In chapter two, we use 8-year records of gross primary productivity and

respiration fluxes at valley scale, along with parallel disturbance chronosequences, to investigate

the impact of the mountain pine beetle on carbon fluxes. In chapter three, I focus on

understanding the consequences of Mountain pine beetle induced tree mortality on canopy

structure, soil microclimate, and soil carbon and nitrogen pools. In chapter four, I explore how

changes in spring snowpack depth and the timing of melt influence subalpine forest soil

respiration dynamics over the subsequent growing season.

!

! )!

CHAPTER 2

CHANGES IN CARBON BALANCE AFTER INSECT DISTURBANCE IN

WESTERN U.S. FORESTS

Adapted from: Moore1, D.J.P., Trahan1, N.A., Wilkes, P., Quaife, T., Desai, A.R., Negron, J.F., Stephens, B.B. Elder, K., and R.K. Monson. 2012. Changes in carbon balance after insect disturbance in Western U.S. forests. In submission to Nature Geoscience

1 Shared first authorship

Abstract

Mountain pine beetle outbreaks have infested more than 86 million hectares of forest in

the U.S.A. since 2000. Using 8-year records of gross primary productivity (GPP) and respiration

fluxes, and parallel disturbance chronosequences in two lodgepole pine forests in Colorado, we

show that soil respiration sharply declined with GPP after tree mortality, reflecting the loss of

autotrophic respiration and rhizodeposition. Three years post-disturbance, soil respiration rates

recover before declining again, co-incident with a pulse of increased soil organic carbon, de-

coupling respiration from concurrent GPP. The impacts of beetle outbreaks on the carbon cycle

are likely more complex than previous estimates, and consideration of the dynamics of

autotrophic and heterotropic substrate supply suggests that beetle-killed forests lose less carbon

to the atmosphere than previously estimated.

!

! *!

2.1 Introduction

Forests and woodlands cover about 30% of global land area and are the largest

component of the terrestrial land sink, absorbing approximately one third of the CO2 released by

human activities (Manning & Keeling, 2006). The degree to which forests will continue to act as

a net sink for CO2 under the pressures of global change is uncertain (Canadell et al., 2007).

Robust predictions of future forest carbon sequestration rely on mechanistic understanding of

forest responses to multiple environmental changes across multiple scales.

Forest disturbance is a fundamental driver of terrestrial carbon cycle dynamics (Bonan,

2008). Initially, most disturbances shift an ecosystem to a carbon source, while recovery from

disturbance is commonly associated with greater ecosystem carbon storage (Odum, 1969;

Magnani et al., 2007). The magnitude of the effects of disturbance and recovery on carbon

sequestration can be as large as or larger than the effects of climate; two-thirds of the total annual

U.S. carbon sink can be attributed to intensive forest management and agricultural abandonment

(Schimel et al., 2004). High altitude forests, in recovery from historical logging and managed for

some degree of fire suppression, are responsible for the majority of carbon uptake and storage in

the Western U.S. (Schimel & Braswell, 2005). Starting in the early 1990s, catastrophic levels of

tree mortality have been observed across millions of hectares of forest in North America infested

by bark beetles (Kurz et al., 2008; Raffa et al., 2008). The majority of the damage originated

from one aggressive species, the native Mountain Pine Beetle Dendroctonus ponderosae (Man,

2010). In addition, this outbreak occurs within the context of a multi-decadal general increase in

tree mortality across the Western U.S. (van Mantgem et al., 2009) where temperature, evolving

and increasingly susceptible forest conditions, and hydrological change are likely contributors

(Breshears et al., 2009).

!

! +!

Modelling how the carbon cycle in Western North American will respond to this

widespread infestation remains challenging because empirical results are conflicting and are

available over short temporal spans and small spatial scales (Amiro et al., 2010). Results

indicating no changes in soil respiration in stands after beetle attack (Morehouse et al., 2008)

imply the transition of a regional carbon sink into a substantial carbon source, while apparent

recovery of GPP in remaining unaffected vegetation may lead to carbon neutrality within a

decade of disturbance (Brown et al., 2010). Ecosystem-scale CO2 fluxes are notoriously hard to

monitor in mountain forests. Eddy-covariance is limited in scale and subject to errors associated

with complex flows (Yi et al., 2008; Sun et al., 2010), satellite observations have only been

validated against carbon uptake terms (Heinsch et al., 2006; Sims et al., 2008;), and chamber

studies are labor intensive and difficult to scale.

We present a unique set of complimentary observations to characterize the response of

both photosynthesis and respiration in an 84 km2 forested valley over an 8-year period following

beetle disturbance. These data include locally calibrated satellite estimates of GPP at 1 km2

resolution, atmospheric CO2 monitoring that reflects nightly valley-scale respiration, and

spatially distributed soil chamber flux and soil organic carbon measurements representing

chronosequences of time since natural and manipulative disturbance. Collectively, these

measurements and analysis products show that forest respiration sharply decreases in the first 2

years after beetle-disturbance, rebounds during the next 4 years, then declines again from 6-8

years post-disturbance.

!

! ,!

2.2 Methods

We estimated the temporal change in GPP and respiration in a forest impacted by the

beetle (Fraser Experimental Forest; FEF) and a forest free of the outbreak (Niwot Ridge; NWR).

The sites are located in Colorado on either side of the continental divide and share similar

climate, elevation, species composition, disturbance age and stand structure. We used a locally

parameterized version of the Temperature and Greenness model (TG; Sims et al., 2008) to

estimate GPP from Enhanced Vegetation Index (EVI) and surface temperature in both forests

from 2002 onwards. To estimate respiratory nighttime accumulation of CO2 in the valley

atmosphere in FEF, continuous measurements were made at the base of the St. Louis Creek

valley (2,745 m, 39º 54' 25''N; 105º 52' 58''W) and at two nearby alpine sites using well-

calibrated infra-red gas analyzers since 2005 (Stephens et al., 2011). Net Ecosystem Exchange

(NEE) is monitored at NWR using an eddy covariance system (3050m, 40°1’ 58”N; 105°32’

47”W) operated since 1998 (Monson et al., 2002).

2.2.1 Site descriptions

The Fraser Experimental Forest is a 93 km2 research forest maintained by the USDA

located within the 1,789 km2 Sulphur Ranger District, Arapaho National Forest, Colorado. Our

chronosequence sites are located at 39°54’27”N 105°51’10”W at 2,913 m in the lodgepole pine

(Pinus contorta Dougl. ex Loud.) dominated subalpine forest which comprises about half of the

tree cover in the district. Engelmann spruce (Picea engelmannii [Parry]), and subalpine fir (Abies

lasiocarpa [Hook.] Nutt. var. lasiocarpa) make up an additional quarter of the tree cover,

predominantly at higher elevations, on north slopes, and along streams. The forest occupies the

!

! -!

St. Louis Creek valley and several tributaries, with 850 m of vertical relief between ridge tops

and valley bottom. Atmospheric CO2 measurements were made at valley bottom (39º 54' 25'' N

105º 52' 58'' W, 2745 m). Most of the forest became established after a stand-replacing fire in

1685, and low elevation portions, including our chronosequence sites were logged in the early

1900’s. The climate at the site consists of long, cold winters and short, cool summers and an

average annual temperature at 2,745 m of 0.5 °C. Annual precipitation averages 737 mm with

about two-thirds falling as snow from October to May. Soils are typically derived from gneiss

and schist with angular gravel, stone, very little silt and clay, and a thin organic horizon of ~3-12

cm. Due to their permeability, these soils have a high capacity to store considerable water during

snowmelt. More information can be found at http://www.fs.fed.us/rm/fraser/. Since 2002

approximately 70% of the St Louis Creek Valley study area has been infested with mountain

pine beetle (MPB; Figure 2.1).

The Niwot Ridge AmeriFlux site is located at 3,050 m in a subalpine forest just below the

Continental Divide near Nederland, Colorado (40° 1' 58'' N; 105° 32' 47'' W). The site

encompasses a gradient between lodgepole pine and spruce-fir forest types, where 46% of tree

cover consists of subalpine fir, 28% Engelmann spruce and 26% lodgepole pine. The girdling

chronosequence plots are located within the lodgepole pine dominated portion of the site. The

AmeriFlux forest shares similar stand size and structure with the FEF site, also in recovery from

early twentieth century logging. Annual precipitation for the site averages 800 mm

(approximately 65% falling as snow) and the mean annual temperature is 1.5 ºC. The soils are

sandy inceptisols derived from granitic moraine with a thin organic horizon ranging from ~3 to 9

cm. Eddy covariance fluxes, storage fluxes and estimation of NEE have been recorded for eleven

!

!%.!

Figure 2.1 Mountain pine beetle infestation in the Fraser Experimental Forest. The study area within the St Louis Creek Valley is outlined with a dotted line, while the AIRCOA CO2 sensor close to the Fraser Experimental Forest headquarters is indicated with a star. Cumulative estimates of beetle infestation were downloaded from the United States Forestry Service Forest Insect & Disease Aerial Survey Data Download; Available from http://www.fs.fed.us/r2/resources/fhm/aerialsurvey/download/, are plotted in dark grey from 2002 through 2010 indicating that the largest annual increase in infestation occurred between 2005 and 2006. Since 2002, the number of live lodgepole pines per hectare, basal area, and quadratic mean diameter have been reduced by 42%, 69%, and 34%, respectively at FEF and surrounding areas (Klutsch et al., 2009). Topographic variation is shown by the light grey shading and notional night-time flows are indicated by the blue arrows in the 2002 plot.

!

!%%!

consecutive years at the primary AmeriFlux tower; data can be accessed at

http://urquell.colorado.edu/data_ameriflux/.

2.2.2 Estimating GPP from satellite observations of EVI, Land surface temperature, and Eddy

Covariance Observations

The Temperature and Greenness (TG) model of Sims et al., 2008 uses EVI (the Enhanced

Vegetation Index, a measure of the amount of green vegetation) and LST (Land Surface

Temperature) to calculate estimates of Gross Primary Productivity (GPP), removing the need for

re-analysis or downscaled meteorology that is typically required in other satellite derived

estimates of GPP. Both EVI and LST products are routinely generated using data from the

MODIS instruments in orbit on NASA's Terra and Aqua satellites. All MODIS data used in this

research were accessed directly from the NASA ftp site: ftp://e4ftl01u.ecs.nasa.gov/ . The TG

model is globally parameterized, so as to optimize the model for subalpine forest ecosystems, the

parameters of minimum, optimal and maximum operational temperature for photosynthesis were

constrained using GPP data from the Niwot Ridge Ameriflux site using a Markov Chain Monte

Carlo Metropolis Hastings routine as follows.

2.2.3 Processing of MODIS data and use of basic GPP model

To calculate GPP from remotely sensed observations we use a reformulated version of

the TG. The original formulation of the TG model is:

GPP = EVI' # LST' # m

!

!%&!

Where m is an empirical scale factor, and EVI' is given by:

EVI'=EVI-0.1

Where EVI is the value of the enhanced vegetation index from the MODIS product (MOD13A2).

The control of temperature on photosynthesis is given by:

LST' = min[(LST/30), (2.5—0.05.LST)]

Where LST is the land surface temperature as observed by MODIS.

Our alternative formulation modifies LST' as:

LST' = min[(LST-minLST)/(optLST-minLST), (maxLST-LST)/(maxLST-optLST)]

This provides an identical response to that of the original function but with parameters

with an intuitive meaning which can be estimated for a given location: minLST is the lowest

MODIS observed temperature at which photosynthesis takes place, maxLST the highest, and

optLST the temperature at which photosynthesis is not limited by temperature. Because this

formulation is amenable to direct statistical parameterization by tuning the parameters to

maximize the agreement between GPP estimated from the Niwot Ridge Ameriflux tower and the

estimate of GPP from the TG model we were able to estimate parameters for the model

applicable to high elevation subalpine forests. The nominal parameters for our formulation and

parameters to reproduce the original TG model are given in Table 2.1.

!

!%'!

Table 2.1 Original and optimized parameter values for the TG model.

TG original value Optimized value Uncertainty (±1$)

Minimum LST 0°C -2.2°C 1.17°C

Optimum LST 30°C 22.3°C 0.94°C

Maximum LST 50°C 51.7°C 4.43°C

!

!%(!

2.2.4 Satellite data pre-processing

MODIS EVI (MOD13A2) data were downloaded for the relevant regions. The data were

filtered using the MODIS QA bits and only the highest quality retrievals were retained. In

addition some anomalously low values (EVI<0.12) were removed as these are likely due to data

contamination by residual cloud. The temporal signal of the MODIS data was found to have

unrealistic high frequency temporal variability, probably due to a mixture of residual

atmospheric and bi-directional reflectance effects. Consequently, the approach taken here is to

use the annual mean EVI for each site (Fig 2.2). This is justifiable because 1) we are examining

long-term (inter-annual) effects of the mountain pine beetle and 2) intra-annual variability in EVI

caused by beetle infestation is not likely to be significant. MODIS Land Surface Temperature

(LST, MOD11A2) data were retrieved and filtered to remove any data that were not given the

highest QA level. The removed data were then replaced using a simple gap-filling algorithm that

replaces the missing values with the mean of a 41-day temporal window surrounding the missing

observation (Fig 2.3).

2.2.5 Bayesian calibration using Metropolis-Hastings

The TG model is based on a set of parameters that are optimized using data from a wide

range of flux tower sites. Here we recalibrate the model specifically against data from the Niwot

Ridge flux tower to produce temperature response parameters, which are more appropriate for

high altitude forests. Net Ecosystem Exchange (NEE) was measured using the Eddy Covariance

technique as described previously (Monson et al., 2002). Briefly the 30 minute, time-averaged,

lag-corrected, covariance between the CO2 concentration and turbulent fluctuations for vertical

!

!%)!

Figure 2.2 EVI data for MODIS pixels containing the NWR flux tower (closed circles and solid line) and the FEF Head Quarters site (open circles and dashed line). The symbols represent actual MODIS data and the lines are the annual means.

Figure 2.3 LST data for MODIS pixels containing the NWR flux tower (squares) and the FEF Head Quarters site (circles). Filled symbols represent actual MODIS observation and open symbols are gap-filled using the temporal window.

!

!%*!

wind speed measured at 10 m above the canopy were added to the CO2 stored below the canopy

with common corrections applied to estimate the NEE. The concentration of CO2 was measured

using an infra-red gas analyzer (model LI-6262, LI-COR Inc., Lincoln, NE, USA), and wind

speed was measured using a sonic anemometer (model CSAT-3, Campbell Scientfic Inc., Logan,

UT, USA). Air temperature, relative humidity, and atmospheric saturation vapor pressure deficit

(VPD) were measured at two heights (2 m and 21.5 m) on the main tower

(http://urquell.colorado.edu/data_ameriflux/) and a heated tipping bucket rain/snow gauge with a

datalogger (Campbell Scientific, Logan Utah, model 23X) was used to measure precipitation.

Corrections for coordinate rotation and density fluctuations are described by (Turnipseed et al.,

2003).

To estimate Gross Primary Productivity (GPP) at NWR, the exponential temperature

response to total night time ecosystem respiration was first estimated from night time values of

NEE during non-stable conditions (where u* was greater than 0.2 ms-1). Using this static

temperature relationship, total ecosystem respiration was removed from daytime NEE to estimate

GPP. While there are many different approaches to calculate GPP from flux observations, many

of these provide statistically equivalent estimates (Desai et al., 2008) and we chose this method

because there are few and relatively simple assumptions associated with it.

In particular, we wanted to capture the response of the forest to very low temperatures

that are typical of mountain forests. In the original model photosynthesis stops at zero degrees

Celsius, which may bias the results for a site where the temperature is sub-zero for a significant

proportion of the year. This is done using a Metropolis-Hastings algorithm (Metropolis & Ulam,

1949) with a chain length of 200000 and an initial burn in period of 10000 timesteps. Prior

probabilities for the parameters, required by the MH algorithm, were taken from the original TG

!

!%+!

model (i.e. 0°C, 30°C and 50°C) specifying a normal distribution with a standard deviation of

5°C in each case. Error on tower GPP values was approximated as 0.1 g C/m2/day for the

calculation of the likelihoods for the Metropolis-Hastings algorithm. The converged chain was

thinned by a factor of 20 to produce summary statistics. The parameter values were selected by

finding the closest chain member to the mean of the resulting multivariate distribution. The

results are given in Table 2.1. These parameter values are similar to the original TG model, but

notably photosynthesis has a lower optimal and minimum temperature of operation. Figure 2.4

shows the posterior probability distribution functions of the parameters. These PDFs were

constructed from the thinned MCMC chain using a Gaussian kernel density estimator. Error bars

on the TG model GPP estimates in Figure 2.10 were derived by running the model for 1000

parameter combinations in the posterior PDF. The annual totals for GPP estimated by the TG

model (GPPTG) are normalized to 2006 and displayed in Figure 2.10 and the actual annual totals

are shown in Table 2.2.

2.2.6 Estimation of relative changes in valley respiration from nighttime accumulation of CO2 at

FEF

We measured atmospheric CO2 concentrations at the bottom of the St. Louis Creek valley

within the Fraser Experimental Forest, on the Headquarters meteorological tower (Figure 2.5).

At night, during moderate regional wind forcing, radiative cooling consistently leads to a stable

drainage flow pattern, characterized by down-valley and down-slope flows, with vertical mixing

within the valley supported by axial rotor circulations (Whiteman, 2000). Under these

conditions, CO2 concentrations above the ground quickly rise by as much as 50 ppm and remain

elevated through the night (see Figure 2.5). Seasonal variations in the extent of CO2 buildup (Fig.

!

!%,!

Figure 2.4 Probability density functions from the Metropolis-Hastings estimation for the parameters required for the TG model.

Table 2.2 Annual sums for GPPTG and CO2 accumulation from Niwot Ridge and Fraser Experimental Forest from 2002 through 2010 (graphed as normalized values in Figure 2.10a.)

Gross Primary Productivity

Year NWR gC m-2 y-1 FEF gC m-2 y-1

FEF Valley CO2 accumulation (% ppm y-1)

2002 795.4 587.6 . 2003 770 490.3 . 2004 827.9 551.3 . 2005 807.1 514.8 . 2006 766 526.1 5138 2007 817.4 453.7 5083 2008 838.8 432 4189 2009 858.8 419.7 4379 2010 829.3 488.5 4481

!

!%-!

Figure 2.5 Average diurnal CO2 concentrations measured at FEF during June, July, and August or December, January, and February. A de-seasonalized, smoothed, reference timeseries from Mauna Loa Observatory, Hawaii (NOAA ESRL), has been subtracted from the data to allow the combination of data from multiple years.

Figure 2.6 Nightly (0000-0400 LST) average CO2 concentrations at FEF and in stable-CO2 conditions at NWR, and a third site at Storm Peak Laboratories (SPL) in Steamboat Springs, CO. Monthly mean data from Mauna Loa Observatory, Hawaii (NOAA ESRL), are also shown for reference.

!

!&.!

2.6) are driven primarily by soil-temperature effects on respiration. Night-to-night variations

depend primarily on wind-speed and direction, but the influence of soil moisture effects on

respiration are also noticeable (Fig. 2.7).

For these measurements, we used an NCAR AIRCOA system (National Center for

Atmospheric Research, Automated Inexpensive Robust CO2 Analyzer; Stephens et al., 2006;

Stephens et al., 2011, which measures CO2 concentrations from 3 heights at 2.5 min intervals.

Four calibration gases tied to the WMO CO2 scale are analyzed every 4 hours, while a single

calibration gas is analyzed every 30 minutes. Compatibility with the WMO CO2 scale and

between independent systems is estimated at 0.2 ppm or better (Stephens et al., 2011). We have

identical AIRCOA instruments at the Niwot Ridge T-Van site (NWR, 3523 m, 30 km to the NE)

and at Storm Peak Laboratory (SPL, 3210 m, 95 km to the NW). These sites typically

experience descending clean-air conditions at night and we use them as a measure of the

background concentration entering the St. Louis Creek valley.

To construct the interannual CO2-buildup timeseries shown in Figure 2.10, we first

average CO2 concentrations from the highest inlets at FEF, NWR, and SPL, and data from the

Headquarters and Fool Creek (3100 m) met towers, over the period 0000 to 0400 LST. Next we

subtract the nightly mean CO2 concentration of NWR and SPL, filtered for stable CO2

concentrations, from that at FEF. Finally, we select for stable drainage conditions at FEF by

filtering for nights when the following average conditions are met: 1) wind direction at

Headquarters is between 160 and 270 degrees (91% of nights), 2) wind direction at Fool Creek is

between 170 and 230 degrees (77% of nights), 3) wind speed at Headquarters is less than 0.6 m/s

(89% of nights), 4) wind speed at Fool Creek is less than 2.2 m/s (99% of nights), and 5) the

!

!&%!

Figure 2.7 Hourly-average CO2 concentrations for two weeks in July 2009, along with wind-speed, wind direction, soil temperature and soil moisture. Wind speed and wind direction are shown both for the FEF Headquarters tower where CO2 is measured and for a tower at 3100 m on the Fool Creek tributary to St. Louis Creek.

!

!&&!

relative variability in hourly-mean CO2 build-up is less than 200% (88% of nights). Collectively,

these filters remove 34% of the nights from the timeseries.

We use the resulting stable drainage flow CO2 build-up values as a measure of soil

respiration over the major portion of the valley upwind of the Headquarters site. Pypker et al.

(2007a; 2007b) demonstrated the feasibility of using atmospheric CO2 measurements to monitor

valley-scale respiration fluxes. Due to difficulties in quantitatively relating CO2 concentrations

to surface flux magnitudes without high-resolution atmospheric modelling and more extensive

wind observations, we do not attempt to estimate absolute flux numbers. Rather, we have

quantified the temporal dynamics of the relative magnitude of respiration by directly measuring

the background-corrected night-time build-up of CO2 in the valley.

This interpretation assumes that interannual variations in night-time wind speed and

mixing within the valley are not significant. This assumption is justified through an analysis of

proximate measures of the boundary layer conditions within the valley. If there were a

systematic change in the nocturnal drainage flow, we expect to see trends in the vertical stability

gradient within the valley, as the strength and depth of a stable layer is related to this quantity

(Stull, 1988) or in the windspeeds themselves. However, we find no significant difference of

variation in stability with time over Jun-Aug 2007-2010 (when temperature gradient

observations were available) (Figure 2.8). Also there is no change in median windspeed at the

Fool Creek station and only a very small (0.06 m/s) change in median windspeed at the Head

Quarters station (Fig 2.9), furthermore the year-to-year variations in windspeed do not correlate

with variations in CO2 build-up which supports a biological driver for the CO2 build-up

(compare Fig 2.9 and 2.8c). The sum of nighttime CO2 accumulation is shown in Table 2.2.

!

!&'!

Figure 2.8 a) Time series of summer (Jun-Aug) nocturnal stability for up valley (red stars) and down valley (black crosses) as indicated by night average virtual potential temperature gradient between a tower near valley base (Headquarters, 2760 m) and along the valley wall (Fool Creek, 3100 m) for years when data at both towers were available. b) Seasonal histograms of the data shown in (a) reveals a stationary frequency distribution of this gradient indicating no significant change in nocturnal boundary layer strength, which is in contrast to c) showing significant changes in the histogram of nocturnal background-corrected CO2 build-up observed by the AIRCOA sensor.

Figure 2.9 Frequency histograms of night-time windspeed during periods of down valley flow for the Headquarters (a) and Fool Creek meteorological stations (b) from 2006 through 2010.

!

!&(!

2.2.7 Establishment of disturbance and experimental chronosequences

To further investigate valley scale respiration dynamics we took an experimental

approach and substituted space for time since disturbance, combining a chronosequence of

beetle-induced mortality (2002 – 2009; 18 plots at FEF) with manipulative tree girdling (2002 –

2009; 24 plots NWR). Girdling simulates the effects of bark beetle damage to the phloem tissues

of infected trees by severing the connection between the needles and the roots, and immediately

eliminates the deposition of GPP to the soil; tree mortality followed within 1-3 years on our

experimental plots.

In the Fraser Experimental Forest site we selected stands of mountain pine beetle (MPB)

killed P. contorta. Stands were aged to within approximately 1-2 years since time of beetle

infestation by the degradation status of the crown (Klutsch et al., 2009). Stands of three to five

lodgepoles in similar states of crown decay were binned into three age classes, and plots of

approximately 9-15 m2 were established around the dead pines, excluding live trees at a distance

of at least 2 m. For plots 7–8 years after infestation, all needles had been shed from the lodgepole

crowns, fine branches less than approximately 0.5 cm in diameter were absent, and bole bark was

often absent or in a greater state of decay. For the plots 5-6 years after infestation, needles were

also absent, but fine branches remained as well as greater amounts of bark on the bole. For plots

1-2 years after infestation the pines had entered the “red-phase” with most needles remaining on

the crown. Influences of the degree of crown exposure on degradation were also considered in

determining age classes. We set up a total of eighteen plots, three per infestation age class and

six control plots of equivalent size, each with 3-5 live uninfested lodgepole pines of similar

medium size diameter at breast height (dbh), about 10–20 cm. The control plots were

geographically distributed throughout the site.

!

!&)!

At the Niwot Ridge AmeriFlux site, previous researchers investigating tree

rhizodeposition and soil respiration (Scott-Denton et al., 2006; Weintraub et al., 2007) created

plots of P. contorta that were girdled in 2002, 2003 and 2004 or left alive as controls. We also

girdled plots of P. contorta for a prior experiment in 2008, 2009 and 2010. Plot trees were

girdled using a saw to make two parallel incisions approximately 15 cm apart around the

circumference of the tree at breast height, and we used the sharp edge of a hatchet to scrape away

the bark, phloem and cambial layers between the incisions. Any apparent re-growth of cambial

tissue was removed two weeks later. Girdled plots from 2008 and prior were trenched around the

perimeter from 15-30 cm in depth to help minimize rhizospheric inputs from outside the plot.

This work allowed the creation of an eight-year girdling disturbance chronosequence. Three

forest plots about 20 m2 containing between three and twelve medium dbh class lodgepole were

established or re-established for each year plots were girdled, along with six live tree plots for

controls, for a total of twenty-four plots.

2.2.8 Soil respiration measurements and extractable dissolved organic carbon analyses

After the sites became snow-free in mid-June of 2010, we measured soil respiration, and

collected soil from the organic horizon for measuring extractable dissolved organic carbon. We

conducted six rounds of sampling at the FEF site and seven at the NWR site at approximate two-

week intervals from June 21st to September 29th. On each sampling date soil respiration was

measured in four permanently installed polyvinyl chloride collars per plot using LI-COR LI-

6400s equipped with 6400-09 soil CO2 flux chambers (LI-COR Incorporated, Lincoln, NE,

U.S.A.). Soil samples were collected from random locations in the organic horizon between 2cm

and 15cm depth of each plot (NWR= 24; FEF =18) and stored in a cooler on ice for return to the

!

!&*!

lab. Within twelve hours, we sieved and homogenized soils, removing coarse plant litter, debris,

and larger (>1 mm) roots. The soils were then immediately extracted with 0.5M K2SO4 for

organic carbon and frozen at -20 °C until analysis. We quantified dissolved organic carbon using

the non-purgable-organic-C protocol on a Shimadzu total organic carbon analyzer (TOC 5000,

Shimadzu Scientific Instruments, Inc., Columbia, MD, USA).

The influence of time since disturbance on soil respiration and dissolved organic carbon

pools were evaluated using repeated measures mixed model ANOVA (SAS proc mixed; The

SAS Institute, Cary, NC, U.S.A.), with random variation at the plot level. Variation in litter and

organic horizon depths were evaluated by mixed model ANOVA (SAS proc mixed), with

random variation at the plot level and Tukey-Kramer post-hoc comparisions of time since

disturbance classes (! = 0.05). Seasonal averages and uncertainties for both soil efflux and total

organic carbon are shown in Table 2.3.

2.3 Results

GPP at FEF departs sharply from NWR from 2006 onwards, corresponding to the peak of

beetle infestation at FEF (Figure 2.10a; Fig 1), with the trends for the two sites diverging by 30%.

Apparent respiration at FEF, also declines from 2006 (Figure 2.10a and 2.10b). While a clear

statistically significant decline in apparent respiration is detected through 2008 (~20%),

respiration then increases in 2009 and 2010 to rates approximately 10% lower than 2006. GPP

shows a steady decline through 2009 and increases in 2010 (Figure 2.10a). These different

!

!&+!

Table 2.3 Mean growing season soil efflux (!mol CO2 m-2 s-1) and soil total organic carbon (!g C g-1 dry soil) from plots across the NWR gridling and FEF beetle-kill chronosequences. Standard error of the mean values are show in brackets, n=3 for killed plots and n=6 for live plots. (graphed as normalized values in Figure 2.11)

Site Year of Mortality

Soil efflux (!mol CO2 m-2 s-1)

Soil Total Organic Carbon (!gC g-1 dry soil)

NWT Live trees 3.62 (0.22) 775.8 (109) FEF Live trees 3.29 (0.18) 1004 (77) NWT 2010 2.9 (0.30) 735.3 (109) NWT 2009 2.43 (0.30) 547 (109) FEF 2008-09 1.93 (0.23) 253.5 (133) NWT 2008 2.52 (0.30) 530.4 (109) FEF 2004-05 3.17 (0.21) 633.1 (133) NWT 2004 3.24 (0.30) 785.4 (109) NWT 2003 2.6 (0.30) 674.9 (109) FEF 2002-03 2.47 (0.22) 438.7 (133) NWT 2002 2.74 (0.30) 584.3 (109)

!

!&,!

Figure 2.10 (A) The relative annual gross primary productivity (GPP) at the uninfested Niwot Ridge (filled circles) and the infested Fraser Experimental Forest (FEF; open circles) and the coordinated decline in annual respiration flux at FEF (grey triangles) starting in 2006. The shaded area represents the cumulative proportion of the St Louis Valley classified as infested with MPB by the USFS aerial survey. (B) Seasonal patterns in GPP (8-day average) and nighttime valley CO2 accumulations at FEF used to calculate the relative annual fluxes. GPP is estimated using the temperature greenness model parameterized from eddy covariance observations at NWR using a Metropolis-Hastings algorithm.

2002 2004 2006 2008 2010

Nighttim

e Valley C

O2

Accum

ulation (ppm)

0

20

40

60

80

Gro

ss P

rimar

y P

rodu

ctiv

ity

(gC

m-2

8 da

ys-1

)

0

1

2

3

4

Proportional A

rea Infestated (FEF)0.0

0.2

0.4

0.6

0.8

1.0A

nnua

l Flu

x R

elat

ive

to 2

006

0.0

0.6

0.8

1.0

1.2

Area infestedGPP NWRGPP FEFCO2 accumulation

A

B

!

!&-!

patterns suggest that the photosynthetic and respiratory fluxes are decoupling. Ultimately, GPP

provides all the carbon input for ecosystem metabolism but substrate for heterotrophic

respiration can originate from newly fixed carbon or a variety of stored or decomposing pools of

different ages (Ryan & Law, 2005).

Mean growing season soil respiration changed significantly with time since disturbance

for both the FEF beetle kill (F = 9.00, p < 0.05) and NWR girdling (F = 2.89, p < 0.05)

chronosequences (Figure 2.11a). There was a strong decline in measured soil respiration in killed

plots over the first three years after disturbance as compared to plots with undisturbed lodgepole

pines ( = 33% NWR; 40% FEF). This decline was followed by a near complete recovery of

respiration in plots 5-6 years post-disturbance ( = 89% NWR; 96% FEF) lasting for 1-2 years,

and a subsequent decline in respiration rates in the oldest mortality plots (Fig 2.11a). The partial

recovery of respiration did not exceed respiration rates in plots with live lodgepole pines, and

corresponds with a pulse of fallen needle litter from dead trees. At FEF, litter and soil organic

horizon depths were deepest in plots with beetle-killed trees 5-6 years post disturbance ( = 7.56

± 0.63 cm), significantly deeper (F= 4.25, p < 0.05) than in plots with live trees ( = 4.98 ± 0.51

cm) or recently killed trees ( = 4.71 ± 0.65 cm) that still retained their needles. This agrees

with patterns of litter-fall observed post bark beetle disturbance over regional scales in Colorado,

where the depth of the litter layer was greater 4–7 years after beetle infestation compared to more

recent and undisturbed plots (Klutsch et al., 2009).

Since GPP was excluded, the recovery of soil respiration was caused by the increased

availability of decomposing substrate, consistent with an observed increase in soil extractable

dissolved organic carbon after 5-6 years at both sites (FEF F = 3.47, p < 0.05; NWR F = 3.53, p

!

x

!

x

!

x

!

x

!

x

!

!'.!

Figure 2.11 (A) Soil CO2 efflux (µmol CO2 m-2 s-1) and (B) soil dissolved organic carbon (µg C g-1 dry soil) as a function of time since disturbance in FEF (black filled) and the NWR (open symbols). At FEF the disturbance was caused by tree mortality induced by mountain pine beetle while at NWR trees were killed by manual girdling. Values are normalized to undisturbed plots (grey filled), and time since disturbance is estimated for FEF at ± 1-2 years.

Soil

Efflu

x (N

orm

aliz

ed µ

mol

CO

2 m-2

s-1

)

0.0

0.2

0.4

0.6

0.8

1.0

TSD vs NCO2Efflux: 0


TSD vs NCO2Efflux: 0.5









Years since disturbance0 2 4 6 8 10

Soil

Org

anic

Car

bon

(Nor

mal

ized

µgC

g-1

dry

soi

l)

0.0

0.2

0.4

0.6

0.8

1.0

UninfestedFEFNWR

A

B

!

!'%!

< 0.05; Figure 2.11b). The decoupling of GPP and respiration 3-4 years after the peak infestation

at the valley scale results from the transition from newly fixed carbon to decomposing material

as the primary substrate for respiration (Figure 2.10).

2.4 Discussion

While widespread insect-induced tree mortality in western North American forests has

caused regional declines in gross primary productivity (GPP) (Coops & Wulder, 2010), the effect

on ecosystem respiration is harder to measure and models have relied on untested assumptions

about the response of respiration. The model of Kurz et al., (2008) estimates that the current

mountain pine beetle outbreak in Canada will transition forests from a small carbon sink to a

large carbon source for at least two decades following disturbance, but assumes an increase in

heterotrophic respiration immediately after infestation. A different modelling study in the

Western U.S. predicted that after an initial sharp loss, above ground live carbon stocks

continuously increased after beetle infestation, carbon balance recovering to pre-outbreak

conditions in 25 years or less, but did not account for changes in the type of substrate driving

respiration and decomposition (Pfeifer et al., 2011). Similar to other models, the model used by

Kurz et al. (2008) calculates heterotrophic respiration as a baseline decay rate that is a fixed

fraction of decomposing biomass modulated by a temperature function. As such, as live biomass

becomes dead biomass respiration increases in these models, resulting a substantial emission of

carbon to the atmosphere. Our results show this assumption to be false and indicate the loss of

carbon to the atmosphere after an insect disturbance to be substantially lower than expected.

In most ecosystems, nearly half of the CO2 released in soil respiration derives from living

plant roots, mycorrhizal fungi and root-associated microbes, and is driven by carbon from recent

!

!'&!

photosynthesis and its rhizodeposition (Högberg & Read, 2006). A suite of eddy-covariance

studies across broad ecosystem types showed a tight coupling of GPP and respiration, where

inter-annual changes in carbon sequestration associate with simultaneous increases or decreases

in both component fluxes (Baldocchi, 2008). Insect induced mortality cuts off the supply of

freshly fixed carbon so while less carbon is taken up, less is released. The timing of the

secondary pulse in soil efflux linked to older, less labile carbon pools, is likely controlled by the

fall rate of litter, coarse woody debris, root decomposition, and the relative impact of post-

disturbance water availability and soil temperatures on decomposition processes (Griffin et al.,

2011).

2.5 Conclusion

Despite the significance of large scale insect outbreaks to carbon sequestration, the land

surface components of the current generation of global climate models cannot recreate episodic

biotic disturbances mechanistically (Running, 2008) and many models used to predict ecosystem

respiration continue to use zero-order kinetics despite clear evidence for substrate limitation (see

Ryan & Law, 2005; Richardson et al., 2006). Robust predictions of the response of the carbon

cycle to disturbance in the context of other environmental changes are dependent on the

development and parameterization of process based models (Running, 2008; Seidl et al., 2011).

We have coupled direct experimental manipulations with long-term monitoring to elucidate the

mechanisms driving respiration after insect disturbance. The patterns of respiration observed

here over nearly a decade post disturbance indicate that, because of a decline in new substrate for

!

!''!

respiration the immediate impacts of the current beetle outbreak in North America on carbon

release are likely to be lower in magnitude than previous estimates (Kurz et al., 2008).

!

!'(!

CHAPTER 3

CHANGES IN SOIL MICROCLIMATE AND NUTRIENT POOLS FOLLOWING MECHANICAL GIRDLING AND MOUNTAIN PINE BEETLE DISTURBANCE IN TWO

COLORADO SUBALPINE FORESTS

Abstract

Over the last decade, mountain pine beetle beetle (Dendroctonos ponderosae Hopkins;

MPB) outbreaks have caused widespread tree mortality in more than 86 million hectares of

Western U.S. coniferous forests. In spite of the magnitude of the current outbreak, the effect of

beetle-induced disturbance on soil biogeochemical cycles is still not well understood. We

assessed changes to the soil microclimate and carbon and nitrogen pools in two high elevation

lodgepole pine (Pinus contorta var. latifola) forests over eight-year parallel chronosequences

following tree death due to mechanical girdling (which simulates pine beetle disturbance) and

natural mountain pine beetle (MPB) disturbance. Generally, tree mortality caused an increase in

soil moisture and a decrease in soil temperature, although these effects were confounded by

stand density and time since disturbance. Declines in root exudates in the first two years after

disturbance caused sharp declines in specific C and N soil concentrations; soil dissolved organic

carbon concentration declined 47-67%, microbial biomass carbon concentration 33-53%,

dissolved organic nitrogen concentration 30-60% and inorganic phosphorus concentration 60-

75%. Five to six years after disturbance however, concentrations for these pools recovered to

levels nearly equivalent to those measured in undisturbed plots. This recovery was coincident

with a pulse into the soil of decomposing needle litter, and consistent with this we observed

deeper soil organic layers 5-6 years after disturbance in both chronosequences. During the

recovery, soil ammonium, but not nitrate, increased to 2-3 times the levels measured in

!

!')!

undisturbed plots, and this accumulation was correlated with increased soil moisture availability.

Despite recovery in the bulk soil organic matter C/N concentrations, and an apparent increase in

litter deposition and decomposition, soil C and N pools declined slightly over the next two years.

It is not clear what limited these concentrations from increasing further, though microbial

biomass N increased where ammonium was available, lowering the microbial carbon to nitrogen

ratio. It is possible that microbial mineralization of C is limited by available N, and generally

constrained by the loss of more labile substrate pools. This study demonstrated: (1) that forest

soils loose a significant fraction of their C and N immediately after beetle disturbance, and (2)

that soil pools of C and N recover after several years due to increased litter deposition and

decomposition.

3.1 Introduction

Forests and woodlands provide important economic and ecosystem services, including

carbon and nutrient sequestration (Bonan, 2008). Widespread forest mortality has been linked in

recent decades to global-change-type drought, combined with pandemic insect outbreaks (Allen

et al., 2010; Breshears et al., 2005; van Mantgem et al., 2009). These stresses are related to one

another, and they interact in synergistic ways to weaken and eventually kill forests. Drought

stress not only impacts tree physiology directly through xylem cavitation, carbon starvation, and

reduced cellular metabolism (Adams et al., 2009; McDowell et al. 2008), but also diminishes

forest resilience to insects and disease. Drought significantly reduces the capacity for trees to

produce and deploy resin-based defenses against insect attack (Breshears et al., 2009; Hicke et

al., 2012; McDowell et al., 2011), while at the same time, higher temperatures drive population

!

!'*!

increases of pests and pathogens and expand their available geographic ranges (Bentz et al.,

2010; Hicke et al, 2006; Logan et al. 2003; Raffa et al., 2008). Many of the wood boring,

phloem feeding insects have the potential to girdle trees, thus weakening roots and decreasing a

tree's capacity to maintain nutrient and water uptake (Raffa & Berryman, 1983).

Insects and pathogens are the most pervasive cause of disturbance in North American

forests. These biotic agents affect significantly more area than fire, and impose an economic

impact nearly five times as great (Dale et al., 2001). In the United States, rising forest mortality

from insects and disease is approaching 50,000 km2, primarily caused by a historically

unprecedented outbreak of mountain pine beetle (Dendroctonos ponderosae Hopkins; MPB) in

western coniferous forests (Man, 2010). While the mechanisms contributing to the current large-

scale eruptive outbreak of MPB are complex and operate across different spatial and temporal

scales (see Raffa et al., 2008), they include large areas of dense, susceptible-age, stressed trees

(Axelson et al., 2009; Bentz et al., 2010; Fettig et al., 2007; Klutsch et al., 2009), warmer

temperatures accelerating beetle life-cycles and facilitating their expansion into higher elevation

forests (Littell et al., 2010; Logan et al., 2010; Mitton & Ferrenberg, 2012), and release from

winter minimum temperatures that cause larval mortality (Fauria & Johnson, 2009).

The mountain pine beetle is an endemic species of bark beetle (Coleoptera:

Curculionidae: Scotlytinae), able to kill trees from a wide range of species (Bentz et al., 2010;

Hicke et al., 2012). The current infestation is occurring primarily in lodgepole (Pinus contorta),

whitebark (P. albicaulis), and limber (P. flexilis) pine species and ranges in geographical extent

from northern Canada to southern California and New Mexico (Kurz et al., 2008; Man, 2010;

Raffa et al., 2008). In 2009, U.S. aerial and ground surveys detected the largest outbreak spread

in more than 30 years, with mountain pine beetles actively killing trees in about 36,000 km2 of

!

!'+!

forest. In the same year, active outbreak occurred in over 4,000 km2 of Colorado forests, as the

mountain pine beetle continued to expand east of the Continental Divide into forests of the

Rocky Mountain Front Range (Man, 2010). Although the current outbreak is severe, the beetles

are selective in the trees that they kill. In Colorado subalpine forests, stands may be relatively

unaffected whereas other regions have experienced heavy beetle mortality rates. In the Fraser

Experimental Forest, mortality approaches around 70% for P. contorta, the dominant overstory

tree species (Wilkes, 2009).

Mountain pine beetles attack and reproduce in live trees. Upon reaching maturity, adult

beetles undergo mass emergence from an infected tree, and a female beetle will locate and bore

into a new host, often preferentially targeting large diameter trees with thicker phloem (Amman,

1972). If able to overcome the tree’s physical resin flow as well as induced and constitutive

histological and chemical toxicity defenses, the female will release aggregation pheromones to

attract other beetles and initiate a mass attack (Raffa & Berryman, 1983; Raffa et al., 2005).

During a successful colonization, female beetles lay eggs in the cambium layer of the trees, and

the larvae “mine” and consume phloem tissue as they develop, carving galleries that sever the

phloem connection between needles and roots (Raffa & Berryman, 1983). In addition, the beetles

act as a vector for species of mutualistic blue stain fungi (Six & Bentz, 2007), which serves as a

larval food source. Fungal hyphae proliferate within the vascular tissues of the tree and occlude

xylem elements, effectively reducing sap flow rates and thus contributing to tree mortality (Raffa

& Berryman 1983). An infested tree retains green canopy needles during the initial season of

attack. After the first year the needles will redden, though they will remain attached to the tree

for up to three years, depending on the degree of crown exposure (Klutsch et al., 2009).

!

!',!

Following this 'red phase', the tree enters the ‘grey phase’, where the needles have dropped

completely from the tree.

MPB disturbance can radically change forest carbon balance, as tree mortality initially

reduces gross primary productivity (Coops & Wulder, 2010) and increased decomposition

enhances the flux of carbon to the atmosphere (Brown et al, 2010; Kurz et al., 2008; Moore,

Trahan, et al., in submission; Pfeifer, et al. 2011). In the short term, MPB mortality may reduce

the fast cycling soil carbon pool through loss of labile root exudation (Chapter 1; Xiong et al.,

2011; Högberg & Read, 2006); but soil carbon pools may recover over time with an initial pulse

of litterfall to the forest floor (Morehouse et al., 2008; Klutsch et al., 2009; Moore, Trahan, et al.,

in submission) followed by the slower decomposition of recalcitrant standing dead wood (Bigler

& Veblen, 2011; Fahey, 1983). Ultimately, C pools and fluxes will be determined by changes in

plant community composition and forest regeneration rates (Collins et al., 2010; Vyse et al.,

2009), and the competitive release of surviving vegetation (Romme et al., 1986), in the context

of climate and interactions with other disturbances such as land-use change and fire (Bigler et al.,

2005; Parker et al., 2006).

Soil nitrogen pools after MPB outbreaks may increase as tree mortality limits plant

uptake and increased labile litter inputs relocate foliar nitrogen from the canopy to the soil

(Griffin et al., 2011; Morehouse et al., 2008). Furthermore, changes in the soil

microenvironment, particularly with regard to moisture and temperature, may cause an increase

in mineralization rates (Griffin et al., 2011). Concurrently, enhanced nitrification (Moorehouse

et al., 2009) and leaching of NO3- to streams (Huber, 2005) may decrease soil nitrate pools

(Griffin et al., 2011; Fahey et al., 1985). It has been shown that microbial immobilization of

nitrogen may occur during litter decomposition (Yavitt & Fahey, 1984), and the increased

!

!'-!

availability of labile carbon substrates has been hypothesized, however, neither was observed

after recent MPB disturbance in at least one study in Colorado (Xiong et al., 2011). Fresh litter

deposits from recent beetle-kill are higher in quality (lower C:N) than deposits due to natural

needle turnover, as sudden beetle-induced mortality prevents the nutrient re-absorption and

translocation that normally occurs with programmed senescence and needle turnover (Fahey et

al., 1985; Morehouse et al., 2008). Griffin et al. (2011) did indeed observe increases in newly

fallen needle litter percent nitrogen; however, this effect was not enough to significantly impact

the entire litter pool and had a weak though significant effect of increasing net N mineralization

rates. Previous studies of MPB disturbance have all reported increases in soil inorganic nitrogen

(Clow et al., 2011; Huber 2005; Morehouse et. al, 2008; Xiong et al., 2011; Griffin et al., 2011),

but there have been large differences in the reported magnitudes of NO3- and NH4

+ concentration

increases, changes in net mineralization and nitrification rates, as well as total dissolved organic

N persistence in the soil and loss from the soil pool and entry into streams. We are not aware of

any studies that have examined the impact of the beetle outbreak on soil inorganic phosphorus

pools, but studies of girdling and rhizopshere dynamics suggest that changes in below ground

carbon allocation is known to affect varyingly affects plant and microbial competitive ability for

phosphorus (Marschner et al., 2011) and it is known that pine forest disturbance from harvesting

can increase phosphorus pools and nutrient concentration variability (Guo et al., 2004).

Studies assessing changes in soil microclimate with mountain pine beetle disturbance

have reported increases in soil moisture attributed to reduced transpiration (Clow et al., 2011;

Morehouse et al., 2008). Temperature responses are more variable; Morehouse et al., (2008)

reported higher air and soil temperature in infested plots compared to un-infested plots, which

was attributed to greater radiation reaching the forest floor due to canopy loss. In contrast,

!

!(.!

Griffin et al., 2011 reported temporal variability in temperature effects, where air and soil

temperatures were elevated in stands 30 years after disturbance, but soil temperatures were

cooler in stands of more recently beetle-killed trees. This was attributed to the insulating effects

of litter accumulation in previously infested stands (Griffin et al., 2011). Finally, it is important

to note that beetle disturbance also changes forest energy balance by increasing winter albedo,

which may significantly offset the radiative forcing anticipated from lost carbon sequestration

(O’Halloran et al., 2012). In addition, beetle disturbance changes the hydrological dynamics of

ecosystems, increasing snow accumulation, ablation (Boon, 2009; Pugh & Small, 2012), and

watershed runoff (Potts, 1984) potentially leading to increased soil nutrient losses.

In this study, we compared two eight-year parallel disturbance chronosequences with tree

mortality caused by mechanical girdling or natural beetle kill in two stands of Colorado

subalpine forest. To assess the impacts of tree mortality on soil microclimate and

biogeochemical pools, we assessed changes in light, soil temperature, moisture, pH, dissolved

organic carbon (DOC), total dissolved nitrogen (TDN), microbial biomass C and N, and

inorganic pools of nitrogen and phosphorus. We hypothesized that we would observe an increase

in soil moisture over time due to the loss of live tree transpiration, and a decrease in soil

temperature corresponding to increased insulation from greater litter inputs. We predicted that

this decrease in temperature would exceed the potential warming effects of increased light

reaching the forest floor with the loss of overstory canopy. We hypothesized that soil DOC pools

and microbial biomass carbon would initially decrease with tree death due to the loss of the

rhizodepostion of labile carbon, and ultimately increase over time due to the decomposition of

dead plant biomass. We predicted increases in inorganic nitrogen pools due to tree mortality and

a concomitant decrease in plant uptake, as well as increased mineralization and nitrification due

!

!(%!

to increasing soil moisture availability, inputs of canopy litter and turnover of fine roots and

mycorrhizal fungi. We also predicted that the increases in inorganic pools would drive both

greater plant nitrogen availability, immobilization in microbial biomass pools, an increase in soil

pH and uptake by the surrounding, surviving vegetation.

3.2 Methods

3.2.1 Study site descriptions

The Fraser Experimental Forest is a 93 km2 research forest maintained by the US Department of

Agriculture located within the 1,789 km2 Sulphur Ranger District, Arapaho National Forest,

Colorado (40840N, 106800W; Figure 3.1A). The plots are located at 39°54’27”N 105°51’10”W

at 2,913 m in the lodgepole pine (Pinus contorta Dougl. ex Loud.) dominated subalpine forest

which composes about half of the tree cover in the district. Engelmann spruce (Picea

engelmannii [Parry]), and subalpine fir (Abies lasiocarpa [Hook.] Nutt. var. lasiocarpa) make up

an additional "25% of the tree cover, predominantly at higher elevations, on north slopes, and

along stream channels. Most of the forest was established during secondary succession following

a stand-replacing fire in 1685, and low elevation portions, including our site, were logged in the

early 1900’s. The climate at the site consists of long, cold winters and short, cool summers and

an average annual temperature at 2,745 m of 0.5 °C. Annual precipitation averages 737 mm with

about 60% falling as snow from October to May. Soils are typically derived from gneiss and

schist with angular gravel, stone, very little silt and clay, and a thin organic horizon of ~3-12 cm.

Plots at the Fraser site had a mineral layer texture of sandy loam. Due to their permeability, these

soils have a high capacity to store water during snowmelt. More information on site

!

!(&!

Figure 3.1 Location of the Fraser Experimental Forest (FEF) and the Niwot Ridge (NWR; also NWT ) Ameriflux forest in Colorado, U.S.A. encompassing the tree mortality chronosequences. A) Site locations in relation to the nearest towns. B) Mountain pine beetle infestation in red mapped from U.S.F.S. aerial damage surveys showing pre-outbreak conditions in 2000 and peak outbreak conditions in 2006 (Wilkes, 2009).

!

!('!

characteristics can be found at http://www.fs.fed.us/rm/fraser/. The Fraser Experimental Forest

has been heavily impacted by the mountain pine beetle outbreak beginning in 2002, with tree

mortality approaching 70% throughout the valley (Figure 3.1B; Wilkes, 2009).


Continental Divide near Nederland, Colorado (40° 1' 58'' N; 105° 32' 47'' W; Figure 3.1A). The

site encompasses an elevational ecotone between lodgepole pine and spruce-fir forest types,

where 46% of tree cover consists of subalpine fir, 28% Engelmann spruce and 26% lodgepole

pine. The girdling chronosequence plots are located within the lodgepole pine dominated portion

of the site. The AmeriFlux forest shares similar stand size and structure with the Fraser site, also

in recovery from early twentieth century logging. Annual precipitation for the site averages 800

mm (approximately 65% falling as snow) and the mean annual temperature is 1.5 ºC. Soils at the

site are sandy inceptisols derived from granitic moraine with a mineral texture of sandy loam and

a thin organic horizon ranging from ~3 to 9 cm. Eddy covariance fluxes, storage fluxes and

estimation of NEE have been recorded for eleven consecutive years at the primary AmeriFlux

tower (Monson et al., 2002; Monson et al., 2010). For previous studies involving soil and

girdling, see Scott-Denton et al. (2003), Scott-Denton et al. (2006), and Weintraub et al. (2007).

In contrast to the Fraser Experimental Forest, the Niwot Ridge Ameriflux forest has been largely

as yet unaffected by the mountain pine beetle outbreak (Figure 3.1B; Wilkes, 2009).

3.2.2 Establishment of disturbance chronosequences:

The Fraser Experimental Forest site

!

!((!

In collaboration with Jose Negron, a USFS research entomologist specializing in work on

western bark beetles, we selected stands of MPB killed P. contorta at the site located in the

Fraser Experimental Forest. One hillside slope with an age gradient of beetle infestation was

selected for plot establishment to minimize chronosequence spatial, stand age and parent

substrate variability. The plots spanned an approximate 25-hectare area, with plot elevations

ranging from 2,884 m to 2, 909 m, slopes from 3-13° and the aspects from15-331°. Stands were

aged for time since beetle infestation by the degradation status of the crown (Klutsch et al. 2009).

Stands of three to five lodgepoles in similar states of crown decay were binned into three age

classes, and plots of approximately 9-15m2 established around the dead pines, excluding live

trees from the plot center at a distance of at least 2m. In plots 1-2 years after infection, needles

remained on the trees and were mostly red in color (red phase). For the plots 5-6 years after

infestation, needles were also absent, but fine branches and greater amounts of bark remained on

the bole (early grey phase). For plots 7–8 years after infestation, all needles had been shed from

the lodgepole crowns, fine branches less than approximately 0.5cm in diameter were absent, and

bole bark was often absent or in a greater state of decay (late grey phase). Influences of the

degree of crown exposure on degradation were also considered in determining age classes. We

set up a total of eighteen plots, three per infestation age class and six control plots of equivalent

size with 3-5 live, uninfested lodgepole pines of similar medium-size diameter at breast height

(dbh), about 10–20 cm. The control plots were geographically distributed throughout the site.

The Niwot AmeriFlux site

Previous researchers investigating tree rhizodeposition and soil respiration at the Niwot

site (Scott-Denton et al., 2006; Weintraub et al., 2007) created plots of P. contorta that were

girdled in 2002, 2003 and 2004 or left alive as controls. We had also girdled plots for a prior

!

!()!

experiment in 2008, 2009 and 2010. Plot trees were girdled using a saw to make two parallel

incisions approximately 15 cm apart around the circumference of the tree at breast height, and

we used the sharp edge of a hatchet to scrape away the bark, phloem and cambial layers between

the incisions. Any apparent re-growth of cambial tissue was removed by scraping two weeks

later. Girdled plots were trenched around the perimeter from 15-30 cm in depth to help minimize

rhizospheric inputs from outside the plot. This work allowed the creation of an eight-year

girdling disturbance chronosequence. Three forest plots about 20 m2 containing between three

and twelve medium dbh class lodgepole were established or re-established for each year plots

were girdled, along with six live tree plots for controls, for a total of twenty-four plots. Plots

spanned an elevation of 3016-3028 m, a slope of 4.7-6.7 ° and an aspect of 78-94 ° over an

approximate 5.5 ha area.

3.2.3 Plot microclimate, stand characteristics, litter and soil organic layer depths

Concurrent with soil sampling for chemical analyses, we measured soil temperature at

four plot locations integrated over 0-15 cm depths using a soil temperature probe (Omega

Engineering Inc., Stamford CT, USA). We also determined soil moisture gravimetrically after

collecting 5 g of wet soil from the organic horizon and drying it for a minimum of 48 h at 60 °C

before reweighing the sample and calculating it as (g wet–g dry)/g dry. For the Fraser beetle kill

chronosequence, we identified all live and dead standing trees greater than 1 m with respect to

species and measured each tree for diameter at breast height (DBH). Plot stand density was

determined by summing the basal area of living or dead trees. Percent ground cover of coarse

woody debris, defined as non-beetle killed logs, stumps and snags 10 cm and greater (Harmon et

!

!(*!

al. 1986), as well as understory vegetation including shrubs and seedlings were visually

estimated to the nearest 5%.

Leaf area index is defined as one half the total green leaf area per unit ground surface

area.! Effective leaf area index is defined as the product of the clumping index (assumed 0.5) for

the non-random spatial distribution of conifer foliage, and the leaf area index (Chen and Black,

1992). The plant area index, or LAI’ is defined as one half the total area of all above-ground plant

materials per unit ground surface area. Plant materials include leaves, branches (live or dead),

boles, and attachments to plant parts such as lichen and moss. To calculate stand scale LAI’, 4

hemispherical photos were taken per plot in the Fraser beetle-kill chronosequence, 1 in each

cardinal direction surrounding and approximately 5 m away from the plot center. Images were

taken 1 m above the ground using a Nikon D700 camera and Sigma EX 8 mm Fisheye lens

georeferenced to north and leveled to gravity. Once acquired, images were processed using Gap

Light Analyzer software (Frazer et al., 1999) and LAI" was integrated over the zenith angles 0 to

60 ° (Stenberg et al., 1994). Canopy openness was determined as the percentage of the total sky

area that is found in canopy gaps for each sky region. Canopy density was calculated as 100 %

(full canopy) - % canopy openness. Gap Light Analyzer software was also used to model

transmitted light flux through the canopy from July 1 - July 31, 2011, a time representative of

forest growing season photosynthetic activity. The percent total light transmitted is the ratio of

radiation incident on the plot surface when there is blockage of light from the surrounding

topography and overlying forest canopy. Total transmitted light (%) was calculated using a 2-

minute solar step, a cloudiness index of 0.5 (Hardy et al., 2004), a spectral fraction (0.25–25

mm) of 1.0 MJ m-2 d-1, a beam fraction of 0.5, a clear-sky transmission coefficient of 0.7, and an

assumed solar constant of 1367 W m-2.

!

!(+!

Litter depth and depth of the plot organic horizon was measured at four locations

randomized within the plots. At the Niwot plots, number and species and DBH of the girdled or

control trees were recorded, but not understory vegetation characteristics or litter depth as these

aspects had been disturbed to some degree in prior experiments.

3.2.4 Soil sampling, extraction, and chemical analyses

Plot soils were sampled repeatedly throughout the 2010 growing season for a total of six

times at the Fraser site and seven at the Niwot site at approximately two-week intervals from

June 21st to September 29th. As most P. contorta fine roots are located in the upper 10 cm of the

organic horizon, approximately 100 g of soil from the organic layer was collected from a random

location within the plot and stored on ice until processed. Within twelve hours, we sieved and

homogenized soils, removing coarse plant litter, debris, and larger (>1 mm) roots. Soil samples

were then immediately extracted for DOC, DON, DIN, dissolved inorganic phosphorus

following the procedure of Weintraub et al. 2007. Briefly, 25 mL of 0.5 M potassium sulfate

(K2SO4) was added to 5 g of homogenized sample, agitated on an orbital shaker for one hour at

120 rpm and then vacuum-filtered through glass fiber filters. Concurrently, an additional 5 g of

each sample were fumigated with 2 mL of ethanol-free chloroform in Erlenmeyer flask for 24

hours, vented for one hour and then extracted as above to ultimately determine microbial

biomass C and N pools. All extracts were frozen at -20 °C until analysis.

Dissolved organic carbon and total dissolved nitrogen in the K2SO4 extracts and the

fumigated soil K2SO4 extracts were quantified using the non-purgable-organic-C protocol on a

Shimadzu total organic carbon analyzer (TOC 5000) equipped with a total dissolved nitrogen

!

!(,!

(TDN) module (Shimadzu Scientific Instruments, Inc., Columbia, MD, U.S.A.). Soil microbial

biomass C and N were estimated by subtracting concentrations of DOC and DON in the

unfumigated samples as well as soil free controls from the fumigated samples (Brookes et al.

1985; Vance et al. 1987). Dissolved inorganic N and P in the K2SO4 extracts was analyzed

spectrophotometrically on a Bio-Tek Synergy HT microplate reader (Bio-Tek Inc., Winooski,

VT, U.S.A.) using vanadium chloride/sulfanilamide reduction chemistry and colormetric

determination of nitrite for nitrate (Doane and Horwath 2003), ammonium using a modified

Berlethot reaction following the protocol of Rhine et al. (1998) and dissolved inorganic

phosphorus by the malachite green colorimetric procedure (D'angelo et al. 2001). DON was

calculated as TDN – DIN.

In addition to biogeochemical pools, texture, pH and bulk density were assessed in the

plots in 2011. An Oakton pH Testr 3 digital pH meter (Oakton instruments, Vernon Hills, IL,

U.S.A.) was used to assess plot soil pH on a 1:1 solution of 30 g soil to 30 mL deionized water.

We determined bulk density in g cm-1 by coring plot soil with a standard soil corer with a known

radius (0.7106 cm), which was too small for any rocks of significant size to enter and induce

error. Core length was measured and used to calculate the sample core volume. The samples

were then oven dried at 110 °C for 48 hours and weighed to determine mass. Texture in g L-1

was determined in-house and by the Soil, Water and Plant Testing Laboratory at Colorado State

University (Fort Collins, CO U.S.A.) via hydrometer.

3.2.5 Ion exchange resin capsule installation, extraction, and chemical analyses

!

!(-!

Two pre-fabricated, 10 g, 2 cm diameter ion exchange resin capsules (UNIBEST PST-1;

Unibest International Corporation, Kennewick, WA, USA) with a surface area of 12.56 cm2 were

installed in each plot after snowmelt to directly quantify 2010 growing season time-integrated

pools of inorganic nitrogen and phosphorus. At both sites, resin capsules were installed in

random plot locations at the boundary of the soil organic and mineral horizons, which ranged

from a depth of ~3-9 cm at the Niwot site and ~2-10 cm at the Fraser site. Resins were removed

at the end of September and frozen at -20°C until extraction. Briefly, resin beads were removed

from their capsules, added to 40 mL 2 M potassium chloride (KCl) agitated on an orbital shaker

at 80 rpm for 1 hour and incubated at 4 °C overnight. Resin blanks were also extracted as

experimental controls. KCl extraction solutions were then vacuum filtered and frozen at -20°C

until analyses. Resin extractions were spectrophotometrically analyzed for dissolved inorganic

ammonium, nitrate and phosphorus on a Bio-Tek EL800 microplate reader (Bio-Tek Inc.,

Winooski, VT, USA) using the same chemistry protocols for soil inorganic nitrogen and

phosphorus in the K2SO4 extracts described above.

3.2.6 Statistical analyses

Differences in time since beetle or girdling disturbance in soil biogeochemical pools and

plot characteristics were evaluated for significance by repeated measures mixed model ANOVA

(SAS proc mixed; The SAS Institute, Cary, NC, U.S.A.), with random variation at the plot level.

Litter and organic horizon depths, pH and bulk density were evaluated by mixed model ANOVA

(SAS proc mixed), with random variation at the plot level. Where time since disturbance effects

were significant, Tukey’s HSD (! = 0.05 unless specified) was used to determine differences

!

!).!

among the means. Relationships between chronosequence variables were also examined using

ordinary least squares linear regressions (SAS proc reg; ! = 0.05). Statistical comparisons were

performed on log-transformed variables where necessary due to increasing variance in the

residuals.

3.3 Results

3.3.1 Fraser beetle-kill chronosequence stand and vegetation characteristics

By design, P. contorta composed over 98% of the basal area in the selected plots. Stand

density significantly increased along the FEF beetle-kill chronosequence from a total P. contorta

mean basal area of 22.7 ± 3.7 m2 ha-1 in the 1-2-year post infestation class to 52.2 ± 4.1 m2 ha-1

in the 7-8-year post infestation class (Table 3.1). Undisturbed plots were located throughout the

chronosequence and had a mean intermediate basal area of 40.1 ± 2.9 m2 ha-1. Live and dead P.

contorta mean basal area within 5 m of the plot center changed with infestation class, with the

mean percentage of living trees greatest in the undisturbed plots 66.0 ± 8.0, dropping to 8.1 ± 4.5

1-2 years after disturbance before gradually increasing in abundance with stand density in the

oldest classes (Table 3.1).

Reflecting stand characteristics, effective plant area index (LAI’) and percent canopy

density were lowest in the 1-2 year post disturbance class plots (0.7 ± 0.004 and 58.6 ± 1.5

respectively). LAI’ and canopy density were highest in the undisturbed (1.2 ± 0.1; 70.4 ± 2.8)

and 7-8 year post beetle plots (1.4 ± 0.2; 74.8 ± 2.6; Table 3.1). These variables influenced the

amount of light transmitted through the canopy, with significantly more light reaching the forest

floor in the 1-2 year post beetle plots (62.0 ± 1.8) than in the other disturbance classes (average

!

!

Table 3.1

Mean 2010 plot stand characteristics across the FEF beetle-kill P. contorta tree mortality chronosequence. Letters next to values indicate significant differences among classes (Tukey's HSD, ! = 0.05). Undisturbed plot class n = 6, beetle kill plot class n = 4. Error values are standard error of the mean.

Variable Years since disturbance class

0 1-2 5-6 7-8 F, pa

Total P. contorta basal area (m2 ha-1) 40.1 ± 2.9 a 22.7 ± 3.7 b 35.3 ± 3.6 a 52.2 ± 4.1 a 10.80, 0.0006

Live P. contorta basal area (m2 ha-1) 26.7 ± 3.8 a 1.6 ± 0.8 b 6.7 ± 5.2 bc 18.4 ± 3.0 ac 12.47, 0.0003

Dead P. contorta basal area (m2 ha-1) 13.4 ± 2.7 a 21.1 ± 4.1 ab 28.6 ± 4.6 b 33.7 ± 1.8 b 7.69, 0.0028

Live P. contorta (%) 66.0 ± 8.0 a 8.1 ± 4.5 b 17.1 ± 12.9 b 34.8 ± 3.4 ab 10.84, 0.0006

Dead P. contorta (%) 34.0 ± 8.0 a 91.9 ± 4.5 b 82.9 ± 12.9 b 65.2 ± 3.4 ab 10.84, 0.0006

LAI' 1.2 ± 0.1 a 0.7 ± 0.04 b 1.0 ± 0.1 ab 1.4 ± 0.2 a 7.70, 0.0028

Canopy density (%) 70.4 ± 2.8 a 58.6 ± 1.5 b 66.7 ± 2.3 ab 74.8 ± 2.6 a 7.06, 0.004

Percent canopy density per total basal area 1.8 ± 0.1 a 2.8 ± 0.4 b 1.9 ± 0.1 ab 1.5 ± 0.1 a 7.06, 0.0053

Understory vegetation cover (%) 45.8 ± 4.1 ab 56.9 ± 7.6 a 65.3 ± 6.3 a 27.2 ± 6.4 b 7.12, 0.0039

Litter depth (cm) 1.7 ± 0.1 a 1.8 ± 0.1 a 2.4 ± 0.3 a 1.7 ± 0.2 a 1.02, n.s.

Soil Organic layer depth (cm) 3.2 ± 0.3 a 3.1 ± 0.3 a 5.2 ± 0.7 b 3.9 ± 0.3 ab 6.07, 0.001 a ANOVA main effect of time since disturbance F and p value; n.s. = non significant

"#!

!

!"#!

43.7 ± 2.9; Figure 3.2A). Vegetation ground cover was lower in the oldest infestation class

compared to plots with more recent mortality and was similar to undisturbed plot vegetation

cover (Table 3.1). While litter depth did not vary across the beetle-kill chronosequence, larger

soil organic layers were observed in older plots; with a significantly deeper mean O horizon in

plots 5-6 years post infestation (5.2 ± 0.7 cm; Table 3.1).

3.3.2 Soil microclimate and characteristics across the Fraser beetle-kill and Niwot girdling tree

mortality chronosequences

Mean growing season soil temperature integrated over 0-15 cm changed significantly

across both disturbance chronosequences, and had a significant negative linear relationship with

mean growing season gravimetric soil moisture R2 = 0.4769, P < 0.0001). Beetle killed plots at

the Fraser site that were 5-6 and 7-8 years since mortality had significantly cooler (average 9.2 ±

0.1 °C) and wetter (average 31.8 ± 2.5%) soils than the undisturbed and 1-2 year post beetle class

plots (average 9.95 ± 0.1 °C; 26.7 ± 2.9%; Table 3.2; Figure 3.2B). In the Niwot girdling

chronosequence, mean soil temperature and moisture were more variable, where the two year

post girdling plots were the coolest (8.0 ± 0.1 °C) and wettest (45.7 ± 4.0%), and the oldest 8

year post girdling plots were the warmest (10.0 ± 0.1 °C) and driest (30.8 ± 4.4%; Table 3.2;

Figure 3.2B). While the Niwot plots were generally more acidic than the Fraser plots, soil pH

did not change over either disturbance chronosequence (Table 3.2). Soil bulk density in the

organic layer was low at both sites and did not vary significantly across the chronosequences

(Table 3.2).

!

!"$!

Figure 3.2 Plot microclimate conditions across the FEF beetle-kill and Niwot girdling tree mortality chronosequences. A) Bars represent mean percent total light transmitted to the forest floor F = 10.97, P < 0.001 across the FEF beetle chronosequence, modeled July 1-31st 2011. * indicates a significant difference by disturbance class using Tukey’s HSD test (! = 0.05). Circles are mean growing season soil temperature integrated across a depth of 0-15cm across the FEF chronosequence, F = 8.56, P < 0.001. B) Mean growing season gravimetric soil moisture across the Niwot girdling and Fraser beetle-kill chronosequences, NWT F = 3.29, P < 0.05; FEF F = 4.62, P < 0.05. Errors are standard error of the mean and uncertainty around tree mortality times in the FEF chronosequence.

!

!

"#!

!

!""!

3.3.3 Soil and resin extractable biogeochemical pools across the Fraser beetle-kill and Niwot

girdling tree mortality chronosequences

During the years following tree mortality, mean growing season extractable soil total

organic carbon (TOC) pools followed a consistent pattern of initial decline, followed by recovery

and a second decline in both chronosequences. Total organic carbon pools were highest in

undisturbed plots (775.8 ± 85.2 !g C g dry soil-1 FEF; 1004.0 ± 87.7 !g C g dry soil-1 NWT),

and then declined significantly and sharply one to two years following disturbance, with losses at

Fraser of 67.3% and at Niwot of 47.2% (Figure 3.3A). Five to six years following disturbance,

soil C pools recovered to 81.6% at Fraser and 78.2% at NWT of that observed in undisturbed

plots. In plots from 7-8 years following disturbance, soil pools declined again but to a lesser

degree, with pools at 56.5% FEF and 58.2% NWT of undisturbed plots (Figure 3.3A).

Mean growing season extractable total dissolved nitrogen (TDN) did not significantly

change across the Niwot Ridge mechanical girdling chronosequence, but did follow a similar

trajectory as seen in the TOC pools for the Fraser beetle-kill chronosequence (Figure 3.3B). In

the Fraser plots from 1-2 years following tree mortality, TDN declined 55.4% compared to

undisturbed plots from 52.5 ± 5.3 to 23.4 ± 3.4 !g N g dry soil-1. Five to six years after beetle

kill soil nitrogen increased to 64.1 ± 14.2 !g N g dry soil-1 and declined again to 48.9 ± 9.0 !g N

g dry soil-1, which were similar to levels seen in undisturbed stands (Figure 3.3B).

The Fraser beetle-kill chronosequence pattern in soil TDN pools was largely a product of

mean growing season extractable dissolved organic nitrogen (DON) pools, where in plots 1-2

years after beetle kill DON significantly declined 63.7% before recovering 5-6 years after beetle-

kill to 106.6% of pools in undisturbed plots. Dissolved organic nitrogen again declined in plots

!

!"#!

Figure 3.3 Mean growing season 2010 extractable soil biogeochemical pools across the Fraser beetle-kill and Niwot girdling tree mortality chronosequences. At FEF, undisturbed plot class n = 6, beetle kill plot class n = 4; at NWT undisturbed plot class n = 6, girdled plot class n = 3. F and P values are for the time since disturbance effect by repeated measures ANOVA, (" = 0.05); n.s. = non significant. Errors are standard error of the mean, and uncertainty around tree mortality times in the FEF chronosequence. A) Total organic carbon FEF F = 7.74, P = 0.0027; NWT F = 5.07, P = 0.0038. B) Total dissolved nitrogen FEF F = 7.31, P = 0.0035; NWT F = 1.8, n.s. C) Dissolved organic nitrogen FEF F = 7.72, P = 0.0026; NWT F = 2.19, n.s. D) Ammonium FEF F = 3.31, P = 0.0523; NWT F = 5.05, P = 0.0038. E) Nitrate FEF F = 0.67, n.s.; NWT F = 1.0, n.s. F) Inorganic phosphorus FEF F = 7.29, P = 0.003; NWT F = 4.41, P = 0.0073.

!

!"$!

7-8 years after beetle-kill with pools at 75.4% of that seen in undisturbed stands (Figure 3.3C).

While a similar pattern to soil carbon pools again emerged in the Niwot DON pools across the

girdling chronosequences, these differences were not significant (Figure 3.3C).

In contrast to the organic nitrogen pools, mean growing season extractable soil

ammonium at the Fraser beetle-kill plots did not decline immediately after disturbance but

increased gradually as a function of time since disturbance from 5.5 ± 0.6 !g N g dry soil-1 in

undisturbed plots to an average of 14.4 ± 4.1 !g N g dry soil-1 in plots 5-6 years and 7-8 after

beetle-induced mortality (Figure 3.3D). In the Niwot mechanical girdling chronosequence soil

NH4+ increased immediately after disturbance, peaking in plots girdled two years previously at

22.2 ± 3.0 !g N g dry soil-1, more than double the NH4+ measured in undisturbed plots (9.9 ± 0.8

!g N g dry soil-1). Unlike the Fraser beetle-kill chronosequence, soil NH4+ declined in the 6-8

year post girdling plots to a level similar to NH4+ in undisturbed plots (Figure 3.3D). Significant

changes in mean growing season extractable soil nitrate pools were not seen at either site, with

NO3- concentrations remaining near detection limit across the span of time since disturbance

(Figure 3.3E).

Mean 2010 growing season extractable soil inorganic phosphorus also exhibited the

pattern observed in soil carbon pools at both sites. In plots that were measured 1-2 years

following tree mortality, soil PO43- concentrations declined markedly compared to undisturbed

plots (75. % FEF and 58.4% NWT) before recovering 5-6 years following tree mortality to

98.2% FEF and 71.3% NWT of PO43- levels in undisturbed stands. In the 7-8 year post

disturbance plots, soil phosphorus declined in a manner similar to total organic carbon with

concentrations at 52.8% FEF and 38.6% NWT of undisturbed plots (Figure 3.3F).

!

!"%!

While nutrient pools extracted from resin capsules installed over the 2010 growing

season reflected patterns similar to those observed in the soil extractions, the results were more

variable across sampling dates. Total inorganic nitrogen, NH4+, NO3

- and PO43- did not change

significantly over either chronosequence at both sites (Table 3.3). With the exception of PO43-,

the concentrations of nutrients extracted were relatively low compared to soil extractions.

3.3.4 Microbial biomass carbon and nitrogen pools and ratios

Mean microbial biomass carbon (MBC) for the 2010 growing season decreased with tree

mortality in both the Niwot and Fraser chronosequences. Again, the loss was most severe 1-2

years after disturbance, where MBC in beetle-killed stand soils declined 53.7% compared to

undisturbed stands from 2239 ± 157 to 1036 ± 115 !g C g dry soil-1. In the 5-6 and 7-8 year

beetle killed plots, microbial biomass increased to an average of 1552.6 ± 149 !g C g dry soil-1;

69.4% that of the MBC in undisturbed plots (Figure 3.4A). In the Niwot girdling

chronosequence microbial biomass carbon was lowest in plots 1 year after girdling, declining

32.8% compared to undisturbed plots from 2466 ± 140 to 1658 ± 235 !g C g dry soil-1. MBC

then increased in the older girdled plots, averaging 1919 ± 190 !g C g dry soil-1 across the older

chronosequence plots (Figure 3.4A).

In the Fraser beetle-kill chronosequence mean growing season extractable microbial

biomass nitrogen (MBN) was significantly lower in plot measured 1-2 years after tree mortality

(105.0 ± 12.7 !g N g dry soil-1) compared to undisturbed plots (183.1 ± 14.5 !g N g dry soil-1)

and plots 5-6 and 7-8 years after tree mortality (average 156.7 ± 14.6 !g N g dry soil-1). While

!

!

"#!

!

!"#!

Figure 3.4 Mean growing season 2010 extractable microbial biomass carbon and nitrogen pools and soil C:N across the Fraser beetle-kill and Niwot girdling tree mortality chronosequences. At FEF, undisturbed plot class n = 6, beetle kill plot class n = 4; at NWT undisturbed plot class n = 6, girdled plot class n = 3. F and P values are for the time since disturbance effect by repeated measures ANOVA, (! = 0.05); n.s. = non significant. Errors are standard error of the mean and uncertainty around tree mortality times in the FEF chronosequence. A) Microbial biomass carbon FEF F = 12.08, P = 0.0004; NWT F = 2.77, P = 0.0459. B) Microbial biomass nitrogen FEF F = 7.61, P = 0.0029; NWT F = 2.25, n.s. C) Soil C:N FEF F = 4.29, P = 0.0242; NWT F = 8.06, P = 0.0003. D) Microbial biomass C:N FEF F = 7.98, P = 0.0024; NWT F = 6.81, P = 0.0008.

!

!"$!

MBN generally declined overall at the Niwot site as a function of time since disturbance, these

decreases were not significantly different from undisturbed plots (Figure 3.4B).

Although the mean growing season extractable soil carbon to nitrogen ratio declined 1-2

years after tree mortality, it was only after a continued slower decline along the chronosequence

that the ratio became significantly different from that of undisturbed stands, decreasing from

(14.7 ± 0.6 C:N) to (10.1 ± 0.5 C:N ) in plots 7-8 years following beetle mortality (Figure 3.4C).

In the girdled plots, soil C:N decreased significantly 1-2 years after girdling (average (8.5 ± 0.8

C:N) compared to undisturbed plots (13.1 ± 0.6 C:N) before increasing again in the older plots

along the chronosequence (average 11.9 ± 0.8 C:N; Figure 3.4C). Mean growing season

extractable microbial biomass C:N decreased significantly and immediately following tree

mortality compared to undisturbed plots and remained low over the entire Fraser disturbance

chronosequence (Figure 3.4D). In the girdled plots, the mean growing season extractable

microbial biomass C:N dropped significantly lower than undisturbed plots one year after girdling

before rising again to a ratio comparable to live tree plots (Figure 3.4D).

3.4 Discussion

3.4.1 Disturbance induced changes in soil microclimate, plot characteristics, and their

interactions with stand structure

As beetle-killed lodgepole pine trees shed their needles and fine branches, reductions in

canopy density and effective plant area index occur (Pugh & Gordon, 2012). Death of the

canopy, in turn, increases light transmission to the forest floor, potentially altering soil

microclimate and biogeochemical cycling (Morehouse et al., 2008). However, changes in LAI’

!

!"%!

and canopy density were confounded in our MPB chronosequence by a prior existing tree density

gradient. Over the infestation years at the Fraser site, the mountain pine beetle moved over time

in a wave upslope, leaving the most recently killed plots (1-2 year post beetle) concentrated in an

area with significantly lower total stand basal area (Table 3.1). The effect of this tree density

gradient exceeded any beetle-driven changes in canopy structure, with effective plant area index,

canopy density, and canopy density per total basal area lower in the 1-2 years post beetle plots

than all other classes (Table 3.1). The prior stand structure dynamics ultimately resulted in

higher light transmitted to the forest floor in the recently killed plots (Figure 3.2).

As we predicted, and as is seen in other beetle-kill and mechanical girdling studies

(Morehouse et al., 2008; Clow et al., 2011; Zeller et al., 2008; Knight et al., 1991), mean

growing season soil moisture increased after tree mortality, but with subtly different dynamics

depending on the manner of disturbance. Mechanical girdling does not immediately directly

compromise the function of xylem tissues, but transpiration has been shown to decline over time

due to feedback to stomatal conductance caused by reduced carbohydrate allocation to roots

(Domec & Pruyn, 2008; De Schepper et al., 2010). Eventually transpiration ceases with tree

mortality. Soil moisture was not different in undisturbed plots and plots that were mechanically

girdled in 2010, but increased in the first two years after girdling by 28.7%, corresponding with

the majority of tree mortality (Figure3.2B). Soil moisture levels exhibited greater inconsistency

in the older girdled plots, and likely are a product of microclimate variability and variation in the

onset of increased light transmission that accompanies variation in the time of canopy loss. In

the Fraser plots, although transpiration ceased with tree mortality, we did not see a corresponding

increase in organic layer soil moisture as predicted 1-2 years after disturbance (Figure 3.2B).

Instead, soil moisture increased gradually over the subsequent years following tree death, and the

!

!"&!

comparatively lower levels in the recent beetle-kill plots appear to be a direct consequence of

higher light transmission to the forest floor, which caused elevated soil temperatures and

increased surface evaporation (Figure 3.2A).

Mean growing season soil temperature decreased by approximately one degree as age

since tree mortality increased within the Fraser beetle-kill chronosequence, exhibiting a

significant negative relationship with soil moisture (R2 = 0.4769, P < 0.0001) and positive

relationship with transmitted PPFD (R2 = 0.2278, P < 0.05), but in contrast to our hypotheses,

no significant relationship with litter depth was observed (Figure 3.2A, B). Morehouse et al.,

2008 found higher air and soil temperatures as well as light levels in bark beetle-killed stands of

Pinus ponderosa. Griffin et al., 2011 found higher soil temperatures in stands 30 years after

beetle mortality, but lower soil temperature beneath more recently killed stands, which was

attributed to greater insulation from increased litter loads after beetle infestation. Measured litter

depth at the Fraser site showed a trend of increase in older beetle-killed plots, but contrary to

expectations, did not significantly change over our beetle-kill chronosequence (Table 3.1). This

suggests that evaporative cooling or higher heat capacity if the soil from increased soil moisture

(Dai et al., 1999) is also an important environmental influence in the wetter soils of beetle-killed

plots.

With natural mortality, litter depth underneath dead P. contorta trees has been shown to

double (Bigler & Veblen, 2011). Reports of increased litter depth beneath trees killed from beetle

disturbance have been variable; changes in litter depth but not mass across a 30 year beetle

chronosequence have been reported (Griffin et al., 2011), as well as higher litter fall rates in un-

infested plots compared to infested plots (Morehouse et al., 2008). These results depend strongly

on time since disturbance, as trees entering the red phase will retain their needles between one

!

!"'!

and three years after the year of infection (Klutsch et al., 2009). In an extensive survey of 221

0.02 ha plots in the Arapaho National Forest in Colorado, Klutsch et al., 2009 reported a

significant, and consistent increase in litter depth in stands at least four years since the initial

infection.

Increases in forbs and total understory vegetation percent cover were noted in recently

beetle-killed stands of P. contorta in the Greater Yellowstone Region of Wyoming (USA)

(Griffin et al., 2011). In contrast, no changes in percent vegetation cover were seen in the study

by Klutsch et al. (2009) in beetle-killed plots after 7 years since infestation in Colorado, and no

changes in shrub and seedling density were reported for stands of P. Ponderosa that had been

killed by beetle infection (Morehouse et al., 2009). At FEF, total percent understory vegetation

cover increased in more recently disturbed plots compared to undisturbed plots (Table 3.1).

However, the effect was variable and related to decreasing overall stand density (R2 = 0.33, P <

0.05) and percent total light transmission to the forest floor (R2 = 0.28, P < 0.05) rather than

beetle infestation class (R2 = 0.06, n.s). In contrast to several other studies of bark beetle impacts,

which have reported an increase in soil pH coincident with increases of NH4+, (Morehouse et al.,

2008; Griffin et al., 2011; Xiong et al., 2011) we found no significant changes in pH across plots

in both disturbance chronosequences (Table 3.2).

3.4.2 Disturbance induced changes in soil and microbial biomass biogeochemical pools

Tree mortality caused large changes in mean growing season extractable dissolved

organic carbon, dissolved organic nitrogen, and inorganic phosphorus pools that followed similar

trajectories in plots of trees treated with mechanical girdling or beetle disturbance (Figure 3.3).

!

!"(!

As expected for DOC and DON, crashes in these soil pools occurred immediately after

disturbance with the presumed loss of root exudation. The linkage of decreased root exudation

to losses in soil DOC and DON calls attention to the importance of recently-fixed labile carbon

as a control over the cycling of both soil C and N (Högberg & Read, 2006). The 47.2%

reduction in soil carbon pools in plots two years after girdling was equivalent to maximum

differences measured (~50%) between girdled and un-girdled trees in previous experiments in

the Niwot Ridge AmeriFlux forest (Weintraub et al., 2007; Scott-Denton et al., 2006). Notably,

Xiong et al., 2011 reported a 49 % decline in dissolved organic carbon pools in stands of

lodgepole pine trees recently-killed (1-3 years) by beetle infestation within 20 km of the Niwot

Ridge girdling chronosequence plots, further validating similarities in the effect of mechanical

girdling and natural beetle-kill on soil carbon pools. At the Fraser beetle-killed sites, we saw an

even steeper reduction in soil DOC (67.3%) and DON (64%) 1-2 years after MPB disturbance.

While site specific factors no doubt come into play, it is possible that the greater reductions in

these pools also occurred with the death of the root system within the first year of attack rather

than the delayed (2-3 years) mortality that we often see following mechanical girdling. Overall,

the loss of mean growing season microbial biomass carbon 1-2 years after disturbance was 1202

"g C g dry soil-1 (FEF) and 809 "g C g dry soil-1 (NWT). As no significant increases in stream

water DOC export have been found in forest basins in Colorado heavily impacted by MPB

(Clow et al., 2011), DOC loss may be more determined by the dynamics among exudation,

priming, microbial immobilization and respiratory loss.

DON declines measured 1-2 years after mechanical girdling (~30%) were also of similar

magnitude to girdling studies at the Niwot Ridge site (Weintraub et a. 2007), as well as other

temperate forest girdling studies (Kaiser et al., 2011; Zeller et al., 2008). More than one

!

!""!

mechanism may potentially explain the reduction in DON observed here. Lower DON pools

following tree mortality may reflect the loss of root nitrogen inputs (Farrar et al., 2003), but also

disruption to the seasonal and trophic dynamics of the rhizosphere. Carbon rich root exudates

first stimulate microbial immobilization of N, whereafter seasonal or other induced reductions in

rhizodeposition lead to C starvation, reduced microbial N demand, and increased mineralization

(Kaiser et al., 2011; Weintraub et al., 2007). In beetle-killed plots, Xiong et al. (2011) found a

soil food web shift from fungal to bacterial feeding nematodes, whose lower C:N excretions may

enhance mineralization and potentially accelerate microbial turnover (Moore et al., 2003).

Decreases in overall microbial biomass with disturbance, especially in terms of losses of the

mycorrhizal community (Högberg & Högberg, 2002) may also decrease phosphorus

mineralization (Marschner et al., 2011); inorganic phosphorus declined 58.4% in the years

immediately following mechanical girdling and 75.7% in the years following beetle-induced

mortality.

In both chronosequences, organic carbon pools recovered in plots 5-6 years after

disturbance before declining again (Figure 3.3A). Although we measured the greatest amounts of

litter in plots 5-6 years after beetle kill, there was enough variability in litter depths among the

plots that this trend was not significant (Table 3.1). However, there was a corresponding

significantly deeper organic layer in the disturbed plots, likely reflecting decomposition of recent

litter fall from the beetle-killed trees. This increased organic layer depth corresponded with the

measured increases in TOC, DON, TDN and PO43- at the Fraser site. While we predicted a

decline and then increase in soil nutrient pools over time, we did not expect a near complete

recovery in soil pools followed by a subsequent decline. The transient nature of the increase in

these soil nutrients supports the hypothesis that a pulse of litter fall provides a temporary

!

!")!

increase in available substrate that stimulates heterotrophic decomposition and mineralization,

and mean growing season soil respiration fluxes along this chronosequence, followed by a

pattern of recovery and then secondary decline (Moore, Trahan et al., in submission). The

recovery in nutrient pools occurs prior to significant seedling re-establishment in disturbed plots

and in the absence of significant changes in understory vegetation cover both seen at our site

(Table 3.1) and regionally (Klutsch et al., 2009). While previous experiments involving small

areas of selective litter removal and understory vegetation clipping at Niwot undoubtedly

impacted litter and decomposition dynamics, the mean soil organic layer in plots 6 years after

mechanical girdling was 32-45% deeper than the organic layers in all other post-girdling time

classes (data not shown). The Niwot girdled plots exhibited similar behaviors in pool recovery

over a similar timeline, and this can be used to argue that the direct effects of vegetation removal

on soil microclimate and biogeochemical pools were limited compared to the larger effect of tree

mortality and crown thinning. Over time, as a projected four-fold increase in downed woody

debris occurs in forests heavily impacted by the MPB (Klutsch et al., 2009), changes in nutrient

pools will likely reflect the slow climate-limited decomposition of large roots (Silver & Miya,

2001), coarse woody debris and its associated immobilization of nutrients (Fahey, 1983),

concurrent with changes in plant community composition and forest regeneration (Collins et al.,

2010; Vyse et al., 2009).

Levels of inorganic nitrogen displayed a different response to tree mortality, increasing

steadily at the Niwot site and peaking two years after girdling (Figure 3.3A). This increase is

probably a product of decreasing plant uptake as carbon starved roots increasingly lack the

energetic resources for uptake and assimilation of inorganic nitrogen (Kaiser, 2010). Similar or

larger increases in soil NH4+ (~50 "g N g dry soil-1) were reported in lodgepole pine dominated

!

!"*!

plots girdled in a prior experiment (Weintraub et al., 2007). Although pool sizes do not

necessarily reflect or predict mineralization rates (Schimel & Bennet, 2004), several lines of

reasoning suggest that the increase in ammonium reflects not only a decrease in plant uptake, but

also an increase in short-term nitrogen immobilization and mineralization rates. First, increases

in N mineralization have been reported for MPB chronosequences in the Greater Yellowstone

area, which were more strongly correlated with decreased temperatures than litter quality

(Griffin et al., 2011). Secondly, soil NH4+ pools were significantly correlated with increasing

soil moisture (R2 = 0.4552, P < 0.0001) and soil temperature (R2 = 0.2973, P = 0.0002) along

both of the disturbance chronosequences that we studied (Figure 3.5). In contrast to the girdled

plots, and in spite of predicted reduced plant uptake with tree mortality, soil ammonium pools

did not increase in stands 1-2 years after beetle infection at the Fraser site, although higher levels

were seen in the older chronosequence plots (Figure 3.3 D). This suggests that the relatively

warmer drier soils in the low basal area plots suppressed mineralization of fine roots,

mycorrhizae and other below ground dead organic matter, an effect that has also been noted in

post-fire lodgepole pine forests (Metzger et al., 2008). Thirdly, reduced rhizodeposition may

induce C substrate limitation and reduce microbial demand for N, leading to greater N

mineralization (Kaiser et al., 2011). Levels of soil ammonium following beetle and girdling

disturbance were transiently higher in these chronosequences than those reported in other MPB

studies, and approach levels reported for lodgepole pine forests after stand-replacing fire (Turner

et al., 2007).

In accordance with most other studies of MPB impacts on biogeochemistry in the

Western U.S., increases in soil ammonium were not accompanied by consistent increases in soil

nitrate (Morehouse et al., 2008; Griffin et al., 2011; Xiong et al., 2011). Nor did a previous

!

!"+!

Figure 3.5 Linear regression of the relationship between mean growing season 2010 soil extractable ammonium and percent soil gravimetric moisture across the Fraser beetle-kill and Niwot girdling tree mortality chronosequences. Summary statistic for regression performed on natural log transformed extractable NH4

+ due to increasing variance in the residuals; R2 = 0.4552, P < 0.0001.

!

!)#!

girdling study in the Niwot Ridge Ameriflux forest report an increase in soil NO3- (Weintraub et

al., 2007). While increased rates of nitrification have been measured with bark beetle disturbance

(Morehouse et al., 2008; Griffin et al., 2011), along with notable increases in soil water NO3- and

leaching with European bark beetle mortality (Huber, 2005), in Western U.S. lodgepole pine

ecosystems Knight et al. (1991) showed that levels of disturbance mortality must be severe

before significant watershed export of nitrate occurs. These coniferous forest ecosystems exhibit

tight recycling of nitrogen (Fahey & Knight, 1986; Turner et al., 2007) with strong microbial

immobilization of nitrate (Stark & Hart, 1993). This differs from higher fertility Eastern U.S. and

European forests that experience higher rates of nitrogen deposition and more commonly export

nitrogen with forest disturbance (Likens et al., 1970; Galloway et al., 2008; Huber, 2005). In

spite of extensive high elevation regional forest mortality in Colorado, and reports of elevated

soil NH4+ (Xiong et al., 2011) detectable changes in the stream export of nitrate are very low and

not well predicted by MPB infestation (Clow et al, 2011). Instead, inorganic nitrogen appears to

be immobilized in the microbial biomass pool (see below), and assimilated by surviving trees

(Knight et al, 1991; Brayden, Trahan, et al., in prep) and the understory vegetation (Griffin et al.,

2011).

Microbial biomass carbon pools along the Fraser and Niwot disturbance chronosequences

followed the sequential 'decline-recovery-decline' pattern seen in our past studies of soil

respiration and accompanying soil carbon pools, (Figure 3.4). Microbial biomass carbon crashes

with the loss of root exudation, recovers with decomposing litter inputs (Morehouse et al., 2009;

Griffin, 2011; Figure 3.4A,B). Reductions in microbial biomass carbon have been consistently

reported with tree girdling, with ectomycorrhizal fungi composing the most significant portions

of those losses (Högberg & Högberg, 2002; Weintraub et al., 2007). While microbial biomass

!

!)$!

nitrogen pools in the Fraser beetle-kill chronosequence also conformed to this pattern, microbial

biomass nitrogen pools were maintained and even increased after recent girdling, corresponding

with increased soil NH4+ levels (Figure 3.4B,D). Generally, microbial biomass carbon to

nitrogen ratios decreased after disturbance in both chronosequences, corresponding to the

increased availability of soil NH4+ (R2 = 0.3343, P = 0.0008), rather than changes in DON pools

(R2 = 0.0599, P = n.s.), a pattern also noted by Xiong et al., 2011. Soil carbon to nitrogen ratios

decreased initially over the beetle-kill and mechanical girdling chronosequences and remained

lower throughout the beetle-kill plots, indicating that while the loss of TOC and DON with

recent tree mortality was severe, the increases in DIN pools increased the relative amount of

nitrogen to carbon post beetle disturbance. C:N in soil and microbial biomass pools in the oldest

girdling plots recovered to levels equivalent to undisturbed plots coincident with the loss of the

soil NH4+ pulse (Figure 3.4D). In addition, the decomposition of lodgepole pine litter may act as

a substantial microbial N and P sink (Fahey, 1983), which may also, in part, explain the

continued depression in microbial C:N ratios throughout the Fraser beetle-kill mortality

chronosequence (Figure 3.4D). These results collectively indicate NH4+ immobilization by

microbial biomass after tree mortality may help retain N after disturbance, and that increased soil

ammonia may be to some degree assimilated by the existing microbial biomass before uptake by

nitrifiers (Turner et al., 2007) that may already be in low abundance in this N limited system

(Vitousek et al., 1979).

3.5 Conclusions

!

!)%!

Tree mortality caused both by mountain pine beetles and mechanical girdling leads to

substantial and remarkably consistent changes in soil microclimate and biogeochemical pools. In

the first 1-2 years after disturbance, losses of root exudates lead to a decline in soil nutrient pools.

Dissolved organic carbon, microbial biomass carbon, dissolved organic nitrogen, and inorganic

phosphorus pool sizes declined to between one-third and three-quarters those observed in

undisturbed plots. While these losses are large, these soil nutrient pools recovered substantially

during the period 5-6 years after disturbance, returning to levels nearly equivalent to that of

undisturbed stands. This recovery was coincident with deeper soil organic layers, likely a

product of the decomposition of leaf litter following tree mortality. Soil moisture and ammonium

increased with the loss of plant uptake, while cooler, wetter soils likely contributed to increased

rates of nitrogen mineralization. As selective mortality agents, mountain pine beetle induced

changes in soil microclimate and biogeochemical pools and their subsequent interactions are

dependent on existing stand structure, and the degree of induced tree mortality. This study

demonstrates that the effect of bark beetles on soil microclimate and biogeochemical pools can

be well represented by mechanical girdling, and that temporal dynamics in the first decade after

disturbance are an important determinant of the source and size of substrate pools for microbial

decomposition and mineralization. Future changes in biogeochemical pools will depend on the

slow decay of coarse roots and of increased loads of coarse woody debris, concurrent with

seedling regeneration, and potential changes in forest community composition.

!

!)&!

CHAPTER 4: THE CONSEQUENCES OF CHANGES IN SPRING SNOW PACK DEPTH FOR SUPALPINE FOREST SOIL RESPIRATION

Abstract

The amount of late winter snow and the timing of snow melt control rates of carbon

sequestration in subalpine forests, but the influence of snow on soil respiration during the period

of summer and spring carbon uptake is uncertain. We investigated the effects of spring snow

pack depth on the autotrophic and heterotrophic components of soil respiration over two summer

growing seasons in a high elevation forest in Colorado, USA. Spring snowpack was increased or

decreased relative to ambient snow conditions in plots with girdled and ungirdled trees, and soil

respiration rates, soil carbon pools and the #13C of soil respiration were measured throughout the

subsequent growing seasons. Girdling, used to identify heterotrophic versus autotrophic

respiration, decreased soil respiration rates by 38%, extractable soil dissolved organic carbon by

42 % and microbial biomass carbon by 42% in 2008, demonstrating the importance of the

transport of sugars from the shoot to the root, subsequent root exudation and fine root production

to soil respiration. While snow additions did not affect snow respiration rates in 2008, they

increased soil respiration rates by 13% in plots with ungirdled trees for a 22-day period in 2009.

Snow amendment plots with ungirdled trees displayed an ~1 ‰ depleted #13C signal after snow

manipulations that persisted through the early growing season in both years. These results

suggest that gross primary productivity (GPP) and ecosystem respiration (RE) are linked through

rhizodeposition, and that the autotrophic component of soil respiration may be sensitive to the

effect of snow moisture availability on the photosynthetic water-use efficiency of trees, and a

!

!)'!

more sensitive indicator of respiration responses to dynamics in snow pack depth compared to

the overall respiration flux rates.

4.1 Introduction

Subalpine lodgepole pine (Pinus contota Dougl. ex Loud.) forests in Western North

America range from Colorado to Northern Alberta, and are characterized by long cold winters

and cool, but often dry, short growing seasons (Fahey & Knight, 1986). In spite of these

constraints, lodgepole pine ecosystems accumulate organic matter (Pearson et al., 1987), and

have been increasingly recognized as important for carbon sequestration in the Western US.

Mountain forests above 750 m account for seventy percent of the Western U.S. carbon sink, due

to recovery from historical logging, fire suppression practices, and localized higher moisture

availability in the semi-arid west (Schimel et al., 2002; Schimel & Braswell, 2005). Productivity

in these ecosystems is supported primarily by winter and early spring precipitation, most of

which falls as snow (Hunter et al., 2006; Monson et al., 2002). In Colorado subalpine forests,

long-term eddy covariance data show that the primary control over the annual rates of carbon

sequestration is the amount of late-winter snow and the timing of spring snowmelt (Hu et al.,

2010; Monson et al., 2005; Monson et al., 2002; Sacks et al., 2007).

Longer growing seasons in many ecosystem types are associated with increased net

ecosystem productivity (Valentini et al., 2000; Churkina et al., 2005). In Eastern deciduous and

boreal forests, warmer temperatures and an earlier onset of spring may lead to increased carbon

uptake through the earlier initiation of leaf expansion and photosynthesis (Black et al., 2000;

Griffis et al., 2003). However, an extended warmer autumn may lead to higher rates of

!

!)(!

ecosystem respiration and reduced net ecosystem productivity (Paio et al., 2008). In western

high elevation coniferous forests, earlier spring onset and longer growing seasons, which have

been correlated with warmer winters, lead to less carbon sequestration (Hu et al., 2010). As

moisture availability limits photosynthesis in these forests, a pronounced period of forest carbon

uptake occurs following the initiation of snowmelt (Monson et al., 2002). In addition, earlier

snowmelts often coincide with cool temperatures and high atmospheric vapor deficits, both of

which decrease photosynthesis (Monson et al. 2005; Monson et al. 2002).

The role of respiration during this critical carbon uptake period during and immediately

following snowmelt has not been investigated. Ecosystem respiration rather than photosynthesis

have been shown to determine carbon balance in a number of forest types globally (Valentini et

al., 2000). Soil respiration is the major component of ecosystem respiration (Raich &

Schlessinger, 1992); e.g. 70% of ecosystem respiration in temperate forests can be from soil

(Law et al., 1999; Janssens et al., 2001). Changes in winter snowpack have been found to effect

respiration throughout the growing season in mixed grass North American prairie ecosystems

(Chimner & Welker, 2005) and winter precipitation is the strongest predictor of inter-annual

variability in summer soil respiration in Sierra Nevada mixed coniferous forests (Concilio et al.,

2006). Still other studies report no changes in growing season soil respiration with changes in

winter snow (Morgner et al., 2010; Larsen et al., 2007). While it has been well documented that

soil respiration rates are determined globally by temperature and moisture, the interactions of

these two drivers can be complex. A previous study at the Niwot Ridge AmeriFlux forest

determined that while temperature was the most significant predictor of soil respiration rates

across the growing season, moisture availability was a more important determinant of respiration

inter-annually (Scott-Denton et al., 2003).

!

!)"!

Substrate availability further complicates the interactions of temperature and moisture on

soil respiration, as soil respiration is a composite of heterotrophic and autotrophic fluxes, each

with the potential to respond differently to climate (Högberg & Read, 2006; Scott-Denton et al.,

2006). Autotrophic respiration, which includes plant root respiration and the rhizosphere

heterotrophs that receive labile carbon and nitrogen substrates directly from the roots, may be

more influenced by climate effects on recent photosynthesis, carbon allocation, phenology, and

root storage (reviewed in Ryan & Law, 2005). Heterotrophic respiration is chiefly constrained

by the availability and quality of labile substrate found in recent foliage and root litter, and may

be more directly responsive to immediate environmental drivers (Ryan & Law, 2005). For

example, the insensitivity of boreal forest soil respiration to temperature during times of peak

photosynthetic activity has been attributed to increases in below-ground C allocation and

rhizodeposition (Niinistö et al., 2011), and drought reduced the heterotrophic component of soil

respiration more than the autotrophic in a lodgepole subalpine forest (Scott-Denton et al., 2006).

In addition, the carbon stable isotope signature of ecosystem respiration (!13C) is related

to climate variation through soil moisture and atmospheric humidity (Schaeffer et al., 2008). At

the Niwot AmeriFlux site, previous analysis of eddy covariance data and continuous monitoring

of ecosystem 13C/12C has shown significant correlations between the !13C of ecosystem

respiration and the overall respiration rate, consistent with observed isotopic patterns in needle

sugars, and that this signature is different between different years (Riveros-Iregui et al., 2011).

The !13C of autotrophic respiration is therefore likely to reflect changes in photosynthetic

discrimination associated with recent photosynthate (Bowling et al., 2008), while that of

heterotrophic respiration is not likely to change with current environmental conditions, providing

unique climatic signatures for respiration in the short-term.

!

!))!

The goal of this research was to examine the consequences of changes in spring

snowpack depth and the timing of melt on soil respiration dynamics over the course of the

growing season in a Colorado subalpine lodgepole pine forest. We used a nested factorial design

of snow manipulation and tree girdling in eighteen 5 m diameter plots. We girdled trees and

trenched the soil surrounding half the plots to isolate and compare climate effects on the

heterotrophic and autotrophic components of soil respiration. Tree girdling isolates the effects of

rhizodeposition on soil processes with minimal disturbance to the soil environment

(Bhupinderpal-Singh et al., 2003; Högberg et al., 2001; Scott-Denton et al., 2006; Weintraub et

al., 2007). Girdling does not immediately kill the root system or tree, and still initially allows for

some degree of transpiration, affecting soil moisture less than trenching (Subke et al., 2004;

Scott-Denton et al., 2006). After girdling, we altered spring snowpack depth to between

approximately 15-25% above and below a set of ambient snow depth control plots. We assessed

impacts on soil respiration, soil temperature and moisture, soil extractable dissolved organic

carbon and microbial biomass carbon pools as well as the 13C/12C ratio of soil respiration over

two growing season. We hypothesized that earlier springs and lower snowpack causes the soil to

warm earlier, increasing soil respiration rates, and thus reducing growing season net ecosystem

productivity. In addition, we hypothesized that in plots with low snowpack depths, tree

rhizodeposition increases microbial substrate availability and has an additive effect with warmer

soil temperatures earlier in the growing season, increasing seasonal maximum rates of soil

respiration. Finally, we hypothesized that the autotrophic component of soil respiration would be

more sensitive to current environmental conditions than the heterotrophic component. Our

overall aim was to focus on this period of snow melt and dynamics in snow pack depth on the

potential to influence overall rates of carbon sequestration in this subalpine forest ecosystem.

!

!)*!

4.2 Methods

4.2.1 Site description


Continental Divide near Nederland, Colorado (40° 1' 58'' N; 105° 32' 47'' W). The site

encompasses a forest gradient between approximately 100-year-old lodgepole pine and spruce-fir

forest types, where 46% of tree cover consists of subalpine fir, 28% Engelmann spruce and 26%

lodgepole pine. The snow manipulation and girdling plots are located within the lodgepole pine

dominated portion of the site. Annual precipitation for the site averages 800 mm, with

approximately 65% falling as snow, and the mean annual temperature is 1.5 ºC. Soils are sandy

inceptisols derived from granitic moraine with a thin organic horizon ranging from ~3 to 9 cm.

For a more detailed description of the site see Monson et al., (2002) and Monson et al. (2005).

4.2.2 Girdling and snow manipulation treatments

Over the summer of 2007 at the Niwot Ridge AmeriFlux site, eighteen 5 m diameter plots

containing between 5-18 medium dbh class lodgepole pine trees were established for a two-year

experiment consisting of a nested factorial design of snow treatments and girdling. Half of the

plots were girdled in April 2008 before the onset of spring forest photosynthetic activity. We

girdled trees using a saw to make two parallel incisions approximately 15 cm apart around the

circumference of the tree at breast height, and we used the sharp edge of a hatchet to scrape away

the bark, phloem and cambial layers between the incisions. Any apparent re-growth of cambial

tissue was removed by scraping two weeks later. After snow melt, girdled plots were trenched

around the perimeter from 15-30 cm in depth to help minimize rhizospheric inputs from roots

!

!)+!

that might grow into the plots from trees located outside the plots, and saplings and vegetation

were clipped at ground level. Regrowth of understory vegetation was clipped at the soil surface

at weekly intervals and the subsequent litter produced from the clipping was removed from the

plots.

To simulate climate variation, six plots received snow amendments to reach a spring

snowpack depth approximately 25% greater than the average depth of six chosen control plots in

March of 2008. At the same time on the last six plots, snow was removed to a ~15-25% lower

depth than that of the control plots. For all treatments we attempted to preserve or maintain the

natural structure of the snowpack, adding snow by blowing it onto the amendment plots using a

snowblower (model 924HV, Husqvarna, U.S.A.) and transferring snow from beyond the plot

margins. During snow removal and for any work inside the plots (which was avoided whenever

possible), we distributed body weight by standing on 2 by 3 m thin plywood boards. Snow

treatments were completed by April 1st in 2008 and 2009. In 2009 however, a late and heavy

spring storm redistributed snow pack depth in the forest to the extent that plot snow levels were

again adjusted to treatment depths by April 21st.

4.2.3 Soil temperature, moisture, and site precipitation

Plot soil temperature and moisture at 5 cm depths were monitored at 1 minute intervals

by two calibrated Decagon EC-5 dielectric soil moisture probes (Decagon Devices, Inc., Pullman,

WA, U.S.A.) and two Campbell Scientific T107 soil temperature probes per plot, wired to CR10

and CR10x dataloggers (Campbell Scientific, Inc., Logan, UT, U.S.A.). The daily site

precipitation measurements presented here are from the C1 meteorological station, located

!

!*#!

approximately 500 m northwest of the study site, and operated by the Niwot Ridge LTER

program and the University of Colorado Mountain Research Station. Measurements were taken

with a Fergusson-type weighing rain gauge; additional details can be found at:

http://culter.colorado.edu/exec/.extracttoolA?c-1pdayv.ml.

4.2.4 Soil respiration measurements

Over the summer of 2007, we repeatedly measured soil respiration in four randomly

located permanently installed PVC collars per plot using LI-COR LI-6400 infared gas analyzers

equipped with 6400-09 soil CO2 flux chambers (LI-COR Incorporated, Lincoln, NE, U.S.A.).

Concurrent with flux measurements, we measured soil temperature integrated over a 0-15cm

depth next to the collars (Omega Engineering Inc., Stamford CT, USA), and these measurements

provided a pre-treatment baseline used for subsequent factor covariance analyses. After the

girdling and snow treatments, we repeated these soil respiration and temperature measurements

at twice-weekly intervals from snowmelt at the end of the growing season in late June to early

October in 2008 and 2009.

4.2.5 Soil extractable dissolved organic carbon and microbial biomass carbon

We sampled plot soils eight times over approximately two-week intervals during the

2008 growing season. Approximately 100 g of soil from the organic layer was collected from a

random location within each plot and stored on ice until processed. Within twelve hours, we

sieved and homogenized soils, removing coarse plant litter, debris, and larger (>1 mm) roots.

!

!*$!

Soil samples were then immediately extracted for dissolved organic carbon following the

procedure of Weintraub et al., (2007). Briefly, 25 mL of 0.5 M potassium sulfate (K2SO4) was

added to 5 g of homogenized sample, agitated on an orbital shaker for one hour at 120 rpm and

then vacuum-filtered through glass fiber filters. Concurrently, an additional 5 g of each sample

were fumigated with 2 mL of ethanol-free chloroform in Erlenmeyer flask for 24 hours, vented

for one hour and then extracted as above to ultimately determine microbial biomass C pools. All

extracts were frozen at -20 °C until analysis. Dissolved organic carbon in the K2SO4 extracts and

the fumigated soil K2SO4 extracts were quantified using the non-purgable-organic-C protocol on

a Shimadzu total organic carbon analyzer (TOC 5000, Shimadzu Scientific Instruments, Inc.,

Columbia, MD, U.S.A.). Soil microbial biomass C was estimated by subtracting concentrations

of DOC in the unfumigated samples as well as soil free controls from the fumigated samples

(Vance et al. 1987).

4.2.6 Assessing the !13CO2 of soil respiration

In addition, we measured the stable isotopic !13CO2 of soil respiration every two weeks

throughout both growing seasons. Using syringes, we also collected 15 mL of respired gas

samples from L-shaped $” OD stainless steel wells topped with Teflon coated septa installed at

the juncture of the soil organic and mineral layers. Gas samples were injected into 12-mL,

septum-capped, evacuated vials (Exetainer, Labco, High Wycombe, UK) and analyzed for

13C/12C with a tunable diode laser absorption spectrometer (TDL, model TGA100A, Campbell

Scientific, Logan, UT, USA) following the protocol of Moyes et al. 2010. Over the course of the

winters of 2008-09 and 2009-10, we continued to sample for plot integrated values of !13CO2 by

!

!*%!

drawing beneath snow-pack air samples from the center of the plot surface soil using Dekabon $”

OD diameter tubing and a small hand pump.

4.2.7 Statistical analyses

Effects of date, snow manipulation, girdling treatments and their interactions on soil

respiration were assessed for significance by repeated measures mixed model ANCOVA (SAS

proc mixed; The SAS Institute, Cary, NC, U.S.A.), with random variation at the plot level, and

pre-treatment autocorrelation and soil temperature used as a covariate. Time and treatment

effects on soil extractable dissolved organic carbon and microbial biomass carbon, and their

interactions were evaluated by repeated measures mixed model ANOVA (SAS proc mixed), with

random variation at the plot level. Snow manipulation and girdling effects on soil moisture and

temperature were also evaluated by repeated measures mixed model ANOVA (SAS proc mixed),

with random variation at the plot level . Where treatment disturbance effects were significant,

Tukey’s HSD (" = 0.05 ) was used to determine differences among the means. Differences in

mean Keeling plot intercepts of 13C/12C between amendment ungirdled plots and other treatment

groups were evaluated using one-tailed Student’s t-test (" =0.1).

4.3 Results

4.3.1 Manipulation of plot spring snow pack depth and subsequent melt time

In 2008, spring snow manipulations were finished by April 4th, with a mean final depth of

1.2 ± 0.2 m for snow amendment plots, 0.93 ± 0.02 m for snow control plots, and 0.7 ± 0.0 m for

!

!*&!

snow removal plots. The treatment led to a delay in the melt out of snow amendment plots,

which occurred on average on June 18th (Julian day 170 ± 1.3 days) compared to snow control

plots which melted out on average June 10th (Julian day162 ± 1.4 days) and snow removal plots,

which melted out on average June 12th (Julian day 164 ± 1.5 days). In 2009, mean final snow

depths after treatment on April 24th were 1.43 ± 0.01 m for snow amendment plots, 1.17 ± 0.04

m for control plots and 1.06 ± 0.02 m for snow removal plots. On average, complete melt out on

removal plots finished by June 1st, (Julian day 152 ± 2.6 days), on control plots by June 3rd,

(Julian day 154 ± 1.5 days), and by June 10th (Julian day 161 ± 1.3 days) for snow amendment

plots.

4.3.2 Snow and girdling impacts on soil respiration, soil extractable dissolved organic carbon

pools, and microbial biomass carbon pools, 2008

Spring snow manipulations in 2008 did not significantly affect soil respiration in the

following growing season (Table 4.1). While the interaction of snow pack depth, day of year,

and girdling was significant (F = 3.18, P < 0.001), the overall effect of varying snowpack on

seasonal soil CO2 effluxes was localized and small (Figure 4.1). Mean growing season soil

respiration in the snow amendment plots was 3.67 ± 0.4 µmol CO2 m-2 s-1, in the control plots

!

!*'!

Table 4.1 Analysis of Variance results for the main effects and interactions of spring snowpack manipulations, girdling, and date on measured plot soil respiration for the 2008 and 2009 growing seasons, n=3.

!! Factor F P

Soil respirationb 2008 Snow manipulation 0.09 0.93

Girdling 64.22 < 0.0001

Date 189.26 < 0.0001

Snow manipulation and date 3.43 < 0.0001

Girdling and date 34.05 < 0.0001

Snow manipulation and girdling 0.28 0.7575

Snow manipulation and girdling and date 3.18 < 0.0001

Soil respirationb 2009 Snow manipulation 0.55 0.5799

Girdling 42.14 < 0.0001

Date 82.64 < 0.0001

Snow manipulation and date 3.73 < 0.0001

Girdling and date 24.52 < 0.0001

Snow manipulation and girdling 1.03 0.3616

Snow manipulation and girdling and date 5.49 < 0.0001 aResults significant at the P < 0.05 level are shown in bold

bSoil respiration in µmol m-2 s-1 ! !

!

!*(!

Figure 4.1 Mean soil respiration in girdled and ungirdled plots over a treatment gradient of spring snowpack depth, growing season 2008, n = 3. Errors represent standard errors of the least squares mean.

!

!*"!

3.77 ± 0.4 µmol CO2 m-2 s-1, and in the removal plots was 3.76 ± 0.4 µmol CO2 m-2 s-1. These

means were not statistically different from one another. In contrast, the girdling treatments

strongly affected soil respiration rates (Table 4.1), decreasing growing season mean annual soil

respiration rates from 4.61 ± 0.39 to 2.86 ± 0.41 µmol CO2 m-2 s-1 (Figure 4.1).

Girdling also caused significant decreases in soil extractable dissolved organic carbon

pools (F = 16.38, P < 0.0001) and microbial biomass pools (F = 43.18, P < 0.0001). Plots with

girdled trees had significantly reduced DOC (72.5 ± 44.1 µg C g dry soil-1) compared to plots

with no tree girdling (124.6 ± 44.9 µg C g dry soil-1; Figure 4.2A). Soil DOC pools did not vary

significantly by day of year (F = 1.65, P = 0.1319) or snow treatment (F = 2.73, P = 0.0713) or

any of the interactions between girdling, snow manipulations and day of year. Mean growing

season microbial biomass C was 42% lower in the girdled plots at 188 ± 38 µg C g dry soil-1

compared to 313 ± 38 µg C g dry soil-1 in ungirdled plots. Microbial biomass carbon also

changed significantly over the course of the growing season (F = 48.6, P < 0.0001), although the

interactions of date and snow and girdling treatments were not significant (Figure 4.2B).

Although the pattern of these results is likely correct, and the magnitude of soil and microbial

biomass carbon decline with girdling is consistent with previous Niwot AmeriFlux site girdling

studies (Scott-Denton et al., 2006; Weintraub et al., 2007), the values are much lower than

reported for those studies (Weintraub et al., 2007; Chapter 3), and therefore must be considered

suspect until further investigation.

4.3.3 Spring snow pack depth and girdling impacts on soil respiration, temperature, and

moisture, 2009

!

!*)!

Figure 4.2 Mean growing season 2008 A) extractable dissolved organic carbon and B) microbial biomass carbon pools in girdled and ungirdled plots, n = 9. Errors represent standard errors of the least squares mean.

!

!**!

In 2009, plot spring snow pack depth did not significantly explain soil respiration (F =

0.55, P = 0.5799) but the three-way interaction of day of year, girdling, and plot snow depth was

significant (F = 5.49, P < 0.0001; Table 4.1). In contrast to 2008, however, elevated soil

respiration rates were observed in ungirdled snow amendment plots over a 22 day period directly

after melt out (Figure 4.3). Mean soil CO2 efflux over this time period was 5.14 ± 0.56 µmol

CO2 m-2 s-1 in the ungirdled snow amendment plots compared to 4.49 ± 0.49 µmol CO2 m-2 s-1 in

the ungirdled snow control plots and 4.41 ± 0.53 µmol CO2 m-2 s-1 in the ungirdled snow removal

plots. Over the entire growing season and the girdling treatment soil respiration rates were 3.61

± 0.42 µmol CO2 m-2 s-1 in the snow amendment plots, 3.83 ± 0.40 µmol CO2 m-2 s-1 in the snow

control plots and 3.58 ± 0.41 µmol CO2 m-2 s-1. Once again, these mean rates were not

statistically different from one another. As in 2008, however, girdling had a significant effect on

growing season 2009 soil respiration rates (F = 42.14, P = < 0.0001), where plots with girdled

trees exhibited depressed mean growing season soil respiration rates of 2.94 ± 0.41 µmol CO2 m-

2 s-1compared to plots with ungirdled trees with rates of 4.4 ± 0.41 µmol CO2 m-2 s-1 (Figure 4.3).

The day of year over the course of the growing season had a strong overall effect on soil

respiration rates (F = 82.64, P < 0.0001), and the interaction of day of year and spring snow pack

depth as well as day of year and the girdling treatment were also significant (Table 4.1).

Volumetric soil moisture varied across the 2009 growing season, peaking in the

amendment ungirded plots directly after snow melt, and followed patterns of growing season site

precipitation (Figure 4.4). Overall, however, the effect of snow pack manipulations (F = 2.16, P

= 0.1616), the girdling treatment (F = 0.86, P = 0.4515) and their interaction (F = 1.26, P =

0.3446) did not affect soil volumetric moisture throughout the entire growing season. Soil

temperature at 5 cm depth in the experimental plots also changed throughout the growing season,

!

!*+!

Figure 4.3 Mean soil respiration in girdled and ungirdled plots over a treatment gradient of spring snowpack depth, growing season 2009, n = 3. Errors represent standard errors of the least squares mean.

!

!+#!

Figure 4.4 Mean daily averages of 1minute sampling intervals of A) plot volumetric soil moisture, B) plot soil temperature at 5 cm depth in girdled and ungirdled plots over a treatment gradient of spring snowpack depth, growing season 2009, n = 3. C) Growing season site precipitation, June through September 2009.

!

!+$!

decreasing with large site precipitation events, and displaying a depressed pattern in the

amendment girdled plots for the first half of the growing season (Figure 4.4). Similarly, soil

temperature throughout the entire growing season of 2009 was not significantly predicted by

changing spring snow pack depths (F = 0.34, P = 0.7164), the girdling treatment (F = 0.00, P =

0.9988), or their interaction (F = 0.46, P = 0.7997).

4.3.4 Spring snow pack depth and girdling treatment impacts on the !13C of soil respiration,

2008 and 2009

After the snow manipulations finished in 2008 and 2009, a consistent pattern emerged of

a depleted 13C/12C of soil respiration in the amendment ungirdled plots (Figure 4.5). This pattern

appeared with the beginning of our isotope sampling measurements, after plot melt out on June

13th (Julian day 165), and persisted through most of the growing season until August 26th (Julian

day 239). During this time the #13C of soil respiration was approximately 1‰ depleted, on

average -27.3 ± 0.2 ‰ in the snow amended ungirdled plots compared to -26.3 ± 0.5 ‰ in all

other treatment groups. This pattern was again detected earlier in 2009, appearing with beneath

snowpack sampling directly after snow manipulations finished on April 1st (experimental day

457), and preceding snowmelt (Figure 4.4). The pattern persisted for 103 days, through July 11th

(experimental day 560). For 2009, the #13C of soil respiration again displayed the ~1‰

depletion averaging -27.5 ± 0.4 ‰ in the amendment ungirdled plots, compared to -26.4 ±

0.4 ‰ in all other treatment groups.

!

!+%!

Figure 4.5 Mean #13C of soil respiration from June 2008- October 2009 in girdled and ungirdled plots over a treatment gradient of spring snowpack depth, growing season 2009, n = 3. Errors represent standard errors of the mean, * = significantly different from other treatments by a one-tailed student’s t-test, (! = 0.1).

!

!+&!

4.4 Discussion

The only significant observed effect of the alteration of spring snowpack depth was a

delay in the melt out of the plots with additional snow. Plot melt out in snow amendment plots

was delayed by 7-8 days in both 2008 and 2009, whereas snow removal and snow control

treatment plots melted out within 2 days of each other in both years. While shallower snow packs,

generally correlate with earlier melt times (Hu et al., 2010), this pattern was confounded in 2009

by warm spring air temperatures which resulted in an earlier seasonal melt out of plots compared

to 2008 in spite of a deeper snow pack.

In contrast to our hypothesis, changes in spring snowpack did not affect appreciably soil

temperatures. While a lower temperature was observed in the snow amendment girded plots that

did slightly suppress respiration rates (Figure 4.4B; Figure 4.3), this effect was not observed in

the snow amendment ungirdled plots. Nor were removal plots significantly warmer. It is

possible that in a year with a shallower snowpack, the snow removal treatment would have been

enough to change the insulating properties of the snowpack, however the springs of both 2008

and 2009 were years with higher than average spring precipitation and deeper spring snow

depths (data not shown). Instead, increased spring snow pack depth resulted in a transient

increase in soil moisture in plots with additional snow in both 2008 (data not shown) and 2009

(Figure 4.4). During this time period, volumetric soil moisture was generally higher in

amendment plots than what was seen in snow control and removal plots, but girdled plots

displayed lower soil moisture than ungirdled plots overall. Scott-Denton et al. (2006), in a

previous girdling experiment at our site, also noted this pattern of lower soil moisture in recently

girdled plots. In both years alterations in soil moisture from snowpack distribution roughly

persisted until the first large summer rain events (52 mm on Julian day of year 201 in 2008; data

!

!+'!

not shown), and 36 mm over Julian days 171-176 in 2009 (Figure 4.4). While snow removal

plots had generally lower soil moisture throughout the growing season, this effect was not

statistically significant for the overall growing season.

The transient increase in soil moisture in the ungirdled amendment plots coincided with

elevated rates of soil respiration in 2009 (Figure 4.3). Directly after melt-out over Julian days

167-188, we found significantly elevated rates of soil respiration in these plots (Figure 4). This

stimulation of respiration rates was not observed in any other treatment groups, including the

snow amendment plots with girdled trees. Given that only plots with intact trees exhibited a

response to a deeper snowpack, it seems likely that greater respiration rates were driven by an

increase in tree below-ground carbon allocation and rhizodepostion of labile carbon exudates to

the soil pool. Several other lines of evidence corroborate this conclusion. Scott-Denton et al.

(2006) found that dissolved organic carbon and soil microbial biomass reached an annual

maximum during the snow melt period at the Niwot site, and that this maximum does not occur

in plots with girdled trees. The DOC patterns observed in this study in 2008 were also consistent

with that finding, exhibiting a seasonal maximum after snowmelt in the ungirdled plots (Figure

4.3), though this was not statistically significant. On a larger scale, numerous eddy covariance

studies in temperate coniferous forests find that seasonal patterns of GPP and RE are coupled

through this linkage, and peak together, corresponding with the time of greatest net ecosystem

productivity (NEP) (Baldocchi 2008).

Interestingly, we found that shifts in the 13C/12C of soil respired CO2 provided additional

insights into above-ground below-ground linkages during the period of snow melt. Directly

after the beginning of snow melt in the snow addition treatments and over the periods of highest

forest carbon sequestration, which coincided with snow melt, a pattern of depletion in the !13CO2

!

!+(!

of soil respiration was observed in the snow amendment ungirdled plots (Figure 4.5). Again, this

pattern does not appear in any of the other treatment groups, including the snow amendment

plots where trees were girdled. As stomatal controls on carbon fixation make the !13C of recent

photosynthate a proxy for tree photosynthetic water-use efficiency (Farquhar et al. 1982), and

recently fixed carbon can be moved rapidly into the soil pool (Högberg et al., 2010) it is possible

then that this isotopic signal is indicative of the higher stomatal conductance relative to the rate

of photosynthetic uptake in trees in plots with snow amendments, and that this leads to

additional below ground substrate allocation. In 2009, this pattern of depletion in amendment

plots compared to other plots ended on measurement day 580 (Julian day 214) after a singular

large precipitation event on Julian day 210 (Figure 4.4C) that significantly altered soil moisture

patterns (Figure 4.4A). Our findings demonstrate that changes in snow pack and timing of melt

can be detected by changes in the carbon isotope composition of the autotrophic component of

soil respiration before changes in the overall soil respiration rate.

While we saw a 13.4% stimulation of soil respiration rate with a 25% increase in

snowpack depth in 2009, the long-term eddy covariance records from the Niwot Ridge

AmeriFlux forest indicates that annual the sensitivity of GPP to the availability of snowpack

moisture and growing season length is actually higher (Hu et al., 2010) than our data imply for

respiration. This supports the findings of previous studies that carbon sink capacity of the Niwot

Ridge subalpine forest is mainly determined by the response of GPP to the availability of snow

rather than RE (Monson et al., 2005).

The carbon cycle is sensitive to effects on multiple time scales. Increasingly, recognition

is emerging that sensitivity to climate variability in one year is contingent on responses to the

weather in previous years, and the lagged biotic responses to slower shifting environmental

!

!+"!

variables such as soil water pools (Schimel et al., 2005b; Stoy et al., 2009). In a Swiss subalpine

coniferous forest, precipitation during the previous year’s winter had the highest explanatory

power for variations in inter-annual NEE (Etzhold et al., in submission). Over longer time scales

environmental drivers such as temperature and precipitation become less predictive of the

component fluxes of NEE, and biotic responses more predictive (Richardson et al., 2007). Inter-

annual variability in respiration has been linked to plant productivity through litterfall and root

exudation rather than temperature (Janssens et al., 2001). It is possible, therefore, that our snow

amendment treatments provided greater growing season moisture availability during a critical

plant uptake period in 2008, and stimulated rhizodeposition in 2009, causing a lagged substrate

effect on soil respiration rates. Alternatively, the stimulation of soil respiration may simply

represent a threshold response to the greater snowpack depth in the amendment plots in 2009

compared to 2008, especially as respiration rates were relatively invariant to the lower snowpack

treatment depths. In a study involving snow amendments in the high artic, non-linear responses

to snow depth were observed, where modest increases in snowpack were found to stimulate

gross ecosystem photosynthesis, and deeper snow both gross ecosystem photosynthesis and

respiration (Rogers et al., 2011).

4.5 Conclusion

The general pattern of CO2 uptake by the Niwot Ridge forest reflects seasonal maximum

rates during the initial few weeks of the growing season, with relatively low rates during the

autumn (Sacks et al., 2007). Rather than earlier springs leading to warmer soil temperatures and

higher respiration rates, early spring soil respiration appears coupled to uptake of carbon through

!

!+)!

photosynthesis and tree rhizodeposition and therefore sensitive, albeit indirectly, to available

snowmelt water and its effect on GPP. Temperature driven increases of respiration later in the

summer are enhanced by substrate availability, but still constrained generally by the availability

of soil water. The snow-driven effects of an earlier spring warm-up are likely to subtly alter the

balance of GPP and RE, reducing GPP to a greater extent than RE, but potentially stimulating

RE proportionally more in the fall. In the future, lower winter snow packs, earlier melt-outs, and

increased summer precipitation are predicted for high elevation forests in Western US mountain

ecosystems (Mote et al. 2005). Our results demonstrate that the principal influence of shallower

snow packs that melt earlier are more likely to be on GPP, and its linkage to the autotrophic

component of RE, rather than decomposition and the heterotrophic component of RE. The

overall result is likely to be a reduction in the growing season sink strength in this subalpine

forest.

!

!+*!

CONCLUSIONS

My thesis focused on the impacts of Mountain pine beetle (MPB) disturbance and

changes in spring snowpack on subalpine forest carbon cycling and soil biogeochemical pools.

We used 8-year records of gross primary productivity and respiration fluxes at the valley scale,

along with parallel disturbance chronosequences, to investigate the impact of the mountain pine

beetle on carbon fluxes. I then used the same disturbance chronosequences to elucidate the

consequences of Mountain pine beetle induced tree mortality on canopy structure, soil

microclimate, and soil carbon and nitrogen pools. Finally, I explored how changes in spring

snowpack depth and the timing of melt influence subalpine forest soil respiration dynamics over

the subsequent growing season. The main findings of my thesis were:

1. Soil respiration sharply declines with gross primary productivity (GPP) after mountain

pine beetle (MPB) induced tree mortality, reflecting the loss of autotrophic respiration

and rhizodeposition.

2. Three years post-disturbance, soil respiration rates recover before declining again, co-

incident with a pulse of increased soil organic carbon, de-coupling respiration from

concurrent GPP.

3. The impacts of beetle outbreaks on the carbon cycle are likely more complex than

previous estimates, and consideration of the dynamics of autotrophic and heterotropic

substrate supply suggests that beetle-killed forests lose less carbon to the atmosphere than

previously estimated.

4. Mountain pine beetle induced tree mortality caused an increase in soil moisture and a

decrease in soil temperature, although these effects are dependent on stand density and

time since disturbance.

!

!++!

5. Loss of root exudates in the first two years after disturbance led to sharp declines in

specific C and N soil concentrations, however five to six years after disturbance,

concentrations for these pools recovered to levels nearly equivalent to those measured in

undisturbed plots. This recovery was coincident with a pulse into the soil of decomposing

needle litter.

6. During the recovery, soil ammonium, but not nitrate, increased to 2-3 times the levels

measured in undisturbed plots, and this accumulation was correlated with increased soil

moisture availability.

7. Deeper spring snow pack can impact growing season soil respiration; soil respiration

rates increased by 13% in plots with ungirdled trees for a 22-day period in 2009.

8. Snow amendment plots with ungirdled trees displayed an ~1 ‰ depleted #13C signal after

snow manipulations that persisted through the early growing season in both years.

9. GPP and ecosystem respiration (RE) are linked through rhizodeposition, and that the

autotrophic component of soil respiration may be sensitive to the effect of snow moisture

availability on the photosynthetic water-use efficiency of trees.

These results collectively demonstrate the importance of climate and disturbance on the

carbon cycle and biogeochemical pools in the lodgepole pine dominated subalpine forest, an

important regional carbon sink. Predicting the future of carbon sequestration in response to

climate change in this ecosystem requires a process based understanding of regional climate

variability on gross primary productivity and ecosystem respiration, together with the localized

but large effects of disturbance. Understanding these processes is also crucial to our management

!

!$##!

of forest regrowth, nutrient losses into streams and rivers, and the maintenance of multi-trophic

animal communities.

!

!$#$!

REFERENCES Adams H.D., Guardiola-Claramonte M., Barron-Gafford G.A. Camilo Villegasa J.,

Breshears D.D., Zoug C.B., Troch P.A., and Huxman T.E. (2009) Temperature sensitivity of drought-induced tree mortality portends increased regional die-off under global-change-type drought. Proceedings of the National Academy of Sciences of the United States of America 106:7063–7066.

Allen C.D., Macalady A.K., Chenchouni H., Bachelet D., McDowell N., Vennetier M., Kitzberger T., Rigling A., Breshears D.D., Hogg E.H., Gonzalez P., Fensham R., Zhang Z., Castro J., Demidova N., Jong-Hwan L., Allard G., Running S.W., Semerci A., and Cobb N. (2010) A global overview of drought and heat-induced tree mortaliy reveals emerging climate change risks for forests. Forest Ecology and Management 259:660-684.

Amiro B.D., Barr A.G., Barr J.G., Black T.A., Bracho R., Brown M., Chen J., Clark K. L., Davis

K.J., Desai A.R., Dore S., Engel V., Fuentes J.D., Goldstein A.H., Goulden M.L., Kolb T.E., Lavigne M.B., Law B.E., Margolis H.A., Martin T., McCaughey J.H., Misson L., Montes!Helu M., Noormets A., Randerson J.T., Starr G., Xiao J. (2010) Ecosystem carbon dioxide fluxes after disturbance in forests of North America. Journal of Geophysical Research-Biogeosciences 115, G00K02.

Amman G.D. (1972) Mountain pine beetle brood production in relation to thickness of lodgepole

pine phloem. Journal of Economic Entomology 65:138-140. Angert A., Biraud S., Bonfils C., Henning C.C., Buermann W., Pinzon J., Tucker C.J. & Fung, I.

(2005) Drier summers cancel out the CO2 uptake enhancement induced by warmer springs. Proceedings of the National Academy of Sciences of the United States of America, 102, 10823-10827.

Axelson J.N., Alfaro R.I., & Hawkes B.C. (2009) Influence of fire and mountain pine beetle on

the dynamics of lodgepole pine stands in British Columbia, Canada. Forest Ecology and Management 257, 1874-1882.

Baldocchi D. (2008) Breathing of the terrestrial biosphere: lessons learned from a global network

of carbon dioxide flux measurement systems. Australian Journal of Botany 56:1-26. Bentz B.J.R., J. Fettig, C.J. Hansen, E.M. Hayes, J.L. Hicke, J.A. Kelsey, R.G. Negron, J.F., &

Seybold, S.J. (2010) Climate Change and Bark Beetles of the Western United States and Canada: Direct and Indirect Effects. Bioscience 60:602-613.

Bhupinderpal-Singh, Nordgren A., Ottosson Lofvenius M., Hogberg M.N., Mellander P.E., and

Hogberg P. (2003) Tree root and soil heterotrophic respiration as revealed by girdling of boreal Scots pine forest: extending observations beyond the first year. Plant, Cell and Environment 26:1287-1296.

!

!$#%!

Bigler C., Kulakowski D., and Veblen T.T. (2005) Multiple distrubance interactions and drought influence fire severity in rocky mountain subalpine forests. Ecology 86:3018-3029.

Bigler C. & Veblen T.T. (2011) Changes in litter and dead wood loads following tree death

beneath subalpine conifer species in northern Colorado. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 41:331-340.

Black, T.A. Chen W.J., Barr A.G., Arain M.A., Chen Z., Nesic E.H., Neumann H.H., and Yang

P.C. (2000) Increased carbon sequestration by a boreal deciduous forest in years with a warm spring. Geophysical Research Letters 27:1271-1274.

Bowling, D.R., Pataki D.E., and Randerson J.T. (2008) Carbon isotopes in terrestrial ecosystem

pools and CO2 fluxes. New Phytologist 178:24-40. Bonan G.B. (2008). Forests and climate change: forcings, feedbacks, and the climate benefits of

forests. Science 320, 1444). Boon S. (2009). Snow ablation energy balance in a dead forest stand. Hydrological Processes

23:2600-2610. Breshears D.D. Cobb N.S., Rich P.M., Price K. P., Allen C.D., Balice R.G., Romme W.H,

Kastens J.H., Floyd L., Belnap J., Anderson J.J, Myers O.B., & Meyer C.W (2005) Regional vegetation die-off in response to global-change-type drought. Proceedings of the National Academy of Sciences of the United States of America 102:15144-15148.

Breshears, D.D., Myers O.B., Meyer C.W., Barnes F.J., Zou C.B., Allen C.D., McDowell N.G.,

Pockman W.T. (2009) Tree die-off in response to global change-type drought: Mortality insights from a decade of plant water potential measurements. Frontiers in Ecology and the Environment 7:185-189.

Brookes P.C., Landman A., Prudent G., and Jenkinson D.S. (1985) Chloroform fumigation and

the release of soil nitrogen: A rapid direct extraction method to measure microbial biomass nitrogen in soil. Soil Biology &. Biochemistry 17:837-842.

Brown M., Black T.A., Nesic Z., Foord V.N., Spittlehouse D.L., Fredeen A.L., Grant N.J.,

Burton P.J., Trofymow J.A. (2010) Impact of mountain pine beetle on the net ecosystem production of lodgepole pine stands in British Columbia. Agricultural and Forest Meteorology 150:254-264.

Canadell J.G., Le Quéré C., Raupach M.R., Field C.B., Buitenhuis E.T., Ciais P., Conway T.J.,

Gillett N.P., Houghton R.A., Marland G. (2007) Contributions to accelerating atmospheric CO2 growth from economic activity, carbon intensity, and efficiency of natural sinks. Proceedings of the National Academy of Sciences of the United States of America 104:18866-18870.

!

!$#&!

Canadell J.G., Raupach M.R. (2008) Managing forests for climate change mitigation. Science 320:1456-1457.

Chen, J.M. and Black T.A. (1992) Foliage area and architecture of plant canopies from sunfleck

size distributions. Agricultural and Forest Meteorology 60, 249-266. Chimner R.A., and Welker J.M. (2005) Ecosystem respiration responses to experimental

manipulations of winter and summer precipitation in a Mixedgrass Prairie, WY, USA. Biogeochemistry 73:257-270.

Churkina G., Schimel D., Braswell B.H., and Xiao, X. (2005) Spatial analysis of growing season

length control over net ecosystem exchange. Global Change Biology 11:1777-1787. Clow D.W., Rhoades C., Briggs J., Caldwell M., and Lewis, Jr. W.M. (2011) Responses of soil

and water chemistry to mountain pine beetle induced tree mortality in Grand County, Colorado, USA. Applied geochemistry 26:S174-S178.

Collin B.J., Rhoades C.C., Underhill J., and Hubbard R.M. (2010) Post-harvest seedling

recruitment following mountain pine beetle infestation of Colorado lodgepole pine stands: a comparison using historic survey records. Canadian Journal of Forest Research-Revue Canadienne De Recherche Forestiere 40:2452-2456.

Concilio A., Chen J, Ma S., and North M. (2009) Precipitation drives interannual variation in

summer soil respiration in a Mediterranean-climate, mixed-conifer forest. Climatic Change 92:109-122.

Coops N.C., and Wulder M.A. (2010) Estimating the reduction in gross primary production due

to mountain pine beetle infestation using satellite observations. International Journal of Remote Sensing 31:2129-2138.

Dai, A., and Trenberth K.E. (1999) Effects of clouds, soil moisture, precipitation, and water

vapor on diurnal temperature range. Journal of Climate 12:2451-2473. Dale V.H., Joyce L.A., McNulty S., Neilson R.P., Ayres M.P., Flannigan M.D., Hanson P.J.,

Irland L.C., Lugo A.E., Peterson C.J., Simberloff D., Swanson F.J., Stocks B.J., and Wotton B.M. (2001) Climate change and forest disturbances. BioScience 51(9):723-734.

D'angelo E., Crutchfield J., Vandiviere M. (2001) Rapid, sensitive, microscale determination of

phosphate in water and soil. Journal of Environmental Quality 30:2206-2209. Desai, A.R., Desai A.R., Richardson A.D., Moffat A.M., Kattge J., Hollinger D.Y., Barr A.,

Falge E., Noormets A., Papale D., Reichstein M., Stauch V. J. (2008) Cross site evaluation of eddy covariance GPP and RE decomposition techniques. Agricultural and Forest Meteorology, 148(6-7), 821-838.

!

!$#'!

De Schepper V., Steppe K., Van Labeke M., and Lemeur R. (2010) Detailed analysis of double girdling effects on stem diameter variations and sap flow in young oak trees. Environmental and Experimental Botany 68:149-156.

Doane T.A., Horwath W.R. (2003) Spectrophotometric determination of nitrate with a single

reagent. Analytical Letters 36:2713-2722. Domec J. and Pruyn M.L. (2008) Bole girdling affects metabolic properties and root, trunk and

branch hydraulics of young ponderosa pine trees. Tree Physiology 28:1493-1504. Evangelista P.H., Kumar, S., Stohlgren T. J., and Yound N.E. (2011) Assessing forest

vulnerbility and the potential distribution of pine beetles under current and future climate scenarios in the Interior West of the US. Forest Ecology and Management, 262:307-316.

Fahey T. J., and Knight D.H. (1986) Lodgepole Pine Ecosystems: Biotic processes play a critical

role in regulating material flux in Rocky Mountain lodgepole pine forests. BioScience 36:610-617.

Fahey T.J., Yavitt J.B., Pearson J.A., and Knight D.H. (1985) The nitrogen cycle in lodgepole

pine forests, Southeastern Wyoming. Biogeochemistry 1:257-275. Fahey T.J. (1983) Nutrient dynamics of aboveground detritus in lodgepole pine (Pinus Contorta

SSP. Latifolia) ecosystems, southeastern Wyoming. Ecological Monographs 53:51-72. FAO (2006) Global Forest Resource Assessment 2005. In Food and Agriculture Organization of

the United Nations, Rome. Farquhar G.D., Oleary M.H., Berry J.A. (1982) On the relationship between carbon isotope

discrimination and the inter-cellular carbon-dioxide concentration in leaves. Australian Journal of Plant Physiology 9:121-137.

Farrar J., Hawes M., Jones D. and Lindow S. (2003) How roots control the flux of carbon to the

rhizosphere. Ecology 84:827-837. Fauria M.M., and Johnson E.A. (2009) Large-scale climatic patterns and area affected by

mountain pine beetle in British Columbia, Canada. Journal of Geophysical Research 114:G01012, doi:10.1029/2008JG000760.

Fettig C.J., Klepzig K.D., Billings R.F., Munson A.S., Nebeker T.E., Negron J.F., Nowak J.T.

(2007) The effectiveness of vegetation management practices for prevention and control of bark beetle infestations in coniferous forests of the western and southern United States. Forest Ecology and Management 238:24-53.

Frazer, G.W., Canham, C.D., and Lertzman, K.P. (1999) Gap Light Analyzer (GLA), Version

2.0: Imaging software to extract canopy structure and gap light transmission indices from true-colour fisheye photographs, users manual and program documentation. Copyright ©

!

!$#(!

1999: Simon Fraser University, Burnaby, British Columbia, and the Institute of Ecosystem Studies, Millbrook, New York.

Frost C.J. and Hunter M.D. (2004) Insect canopy herbivory and frass deposition affect soil

nutrient dynamics and export in oak mesocosms. Ecology 85:3335-3347. Galloway J.N., Townsend A.R., Erisman J.W., Bekunda M., Cai A., Freney J. R., Martinelli L.A.,

Seitzinger S.P., and Sutton M.A. (2008) Transformation of the nitrogen cycle: recent trends, questions, and potential solutions. Science 320:889-892.

Goodale C.L., Apps M.J., Birdsey R.A., Field C.B., Heath L.S., Houghton R.A., Jenkins J. C.,

Kohlmaier G.H., Kurz W., Liu S., Nabuurs G.J., Nilsson S., and Shvidenko, A.Z. (2002) Forest carbon sinks in the Northern Hemisphere. Ecological Applications 12:891-899.

Goulden M.L., Munger J.W., Fan S.M., Daube B.C., and Wofsy S.C. (1996) Exchange of carbon

dioxide by a deciduous forest: Response to interannual climate variability. Science 271:1576-1578.

Griffin J.M., Turner, M.G., and Simard, M. (2011) Nitrogen cycling following mountain pine

beetle disturbance in lodgepole pine forests of Greater Yellowstone. Forest Ecology and Management 261:1077-1089.

Griffis T.J., Black T.A., Morgenstern K., Barr A.G., Nesic Z., Drewitt G.B., Gaumont-Guay D.,

and McCaughey J.H. (2003) Ecophysiological controls on the carbon balances of three southern boreal forests. Agricultural and Forest Meteorology 117:53-71.

Guo D., Mou P., Jones R.H., and Mitchell R.J. (2004) Spatio-temporal patterns of soil available

nutrients following experimental distrubance in a pine forest. Oecologia 138:61-621. Hardy J.P., Melloh R., Koenig G., Marks D., Winstral A., Pomeroy J.W. and Link T. (2004)

Solar radiation transmission through conifer canopies. Agricultural and Forest Meteorology 126:257-270.

Harmon M.E., Franklin J.F., Swanson F.J., Sollins P., Gregory S.V., Lattin J.D., Anderson N.H.,

Cline S.P., Aumen N.G., Sedell J.R., Lienkaemper G.W., Cromack J.R., and Cummins K.W. (1986) Ecology of coarse woody debris in temperate ecosystems. Advances in Ecological Research 15:133-276.

Heinemeyer A., Hartley I.P., Evan S.P., Carreira De La Fuentes J.A., and Ineson P. (2007) Forest

soil CO2 flux: uncovering the contribution and environmental responses of ectomycorrhizas. Global Change Biology 13.

Heinsch F.A, Zhao M., Running S.W., Kimball J.S., Nemani R.R., Davis K.J., Bolstad P.V.,

Cook B.D., Desai A.R., Ricciuto D.M., Law B.E., Oechel W.C., Kwon H.J., Luo H., Wofsy S.C., Dunn A.L., Munger J.W., Baldocchi D.D., Xu L., Hollinger D.Y., Richardson A.D., Stoy P.C., Siqueira M.B.S., Monson R.K., Burns S.P., and Flanagan

!

!$#"!

L.B. (2006) Evaluation of remote sensing based terrestrial productivity from MODIS using regional tower eddy flux network observations. IEEE Transactions in Geoscience and Remote Sensing 44(7), 1908-1925.

Hicke J.A., Logan J.A., Powell J., and Ojima D.D. (2006) Changing temperatures influence

suitbiity for modeled moutain pine beetle (Dendroctonus ponderosai) outbreaks in the western United States. Journal of Geophysical Research 111: G02019.

Hicke J.A., Allen C.D., Desai, A.R., Doetze M.C., Hall R.J., Hogg E.H., Kashian D.M., Moore

D., Raffa K.F., Sturrock R.N. and Vogelmann J. (2012) Effects of biotic disturbances on forest carbon cycling in the United States and Canada. Global Change Biology 18:7-34.

Högberg M.N. and Högberg P. (2002) Extramatrical ectomycorrhizal mycelium contributes one-

third of microbial biomass and produces together with associated roots, half the dissolved organic carbon in a forest soil. New Phytologist 154:791-795.

Högberg M.N., Briones M.J.I., Keel S.G., Metcalfe D.B., Campbell C., Midwood A.J., Thornton

B., Hurry V., Linder S., Näsholm T., and Högberg P. (2010) Quantification of the effects of season and nitrogen supply on tree below-ground carbon transfer to ectomycorrhizal fungi and other soil organisms in a boreal pine forest. New Phytologist 187:485-493.

Högberg P., Plamboeck A.H., Taylor A.F.S. and Fransson P.M.A. (1999) Natural 13C abundance

reveals trophic status of fungi and host-origin of carbon in mycorrhizal fungi in mixed forests. Proceedings of the National Academy of Sciences 96:8534-8539.

Högberg P., Nordgren A., Buchmann N., Taylor A.F.S., Ekblad A., Högberg M.N., Nyberg G.,

Ottosson M. L., and Read D. J. (2001) Large-scale forest girdling shows that current photosynthesis drives soil respiration. Nature 411:789-792.

Högberg P., and Read J. (2006) Towards a more plant physiological perspective on soil ecology.

Trends in Ecology & Evolution 21:548-554. Houle D., Duchesne L., and Boutin R. (2009) Effects of a spruce budworm outbreak on element

export below the rooting zone: a case sstudy for a balsam fir forest. Annals of Forest Science 66:707, 1-9.

Hu J., Moore D.J.P., Burns S.P., and Monson R.K. (2010) Longer growing seasons lead to less

carbon sequestration by a subalpine forest. Global Change Biology 16:771-783. Huber C. (2005) Long lasting nitrate leaching after bark beetle attack in the highlands of the

Bavarian Forest National Park. Journal of Environmental Quality 34:1772-1779. Hunter T., Tootle, G., and Piechota, T. (2006) Oceanic-atmospheric variability and western U.S.

snowfall. Geophysical Research Letters 33:L13706, 1-5.

!

!$#)!

Janssens I.A., Lankreijer H., Matteucci G., Kowalski A.S., Buchmann N., Epron D., Pilegaard K., Kutsch W., Longdoz B., Grunwald T., Montagnani I., Dore S., Rebmann C., Moors E.J., Grelle A., Rannik U., Morgenstern K., Oltchev S., Clement R., Gudmundsson J., Minerbi S., Berbigier P., Ibrom A., Moncrieff J., Aubinet M., Bernhoffer C., Jensen N.O., Vesala T., Granier A., Schulze E. D., Lindroth A., Dolman A.J., Jarvi P.G., Ceulemans R., and Valentini R. (2001) Productivity overshadows temperture in determining soil and ecosystem respiration across European forests. Global Change Biology 7:269-278.

Kaiser C., Koranda M., Kitzler B., Fuchslueger L., Schnecker, J., Schwieger P., Rasche F.,

Zechmeister-Boltenstern S., Sessitsch A., and Richter A. (2010) Belowground carbon allocation by trees drives seasonal patterns of extracellular enzyme activities by altering microbial community composition in a beech forest soil. New Phytologist 187:843-858.

Kaiser C., Fuchslueger L., Koranda M., Gorfer M., Stange C.F., Kitzler B., Rasche F., Strauss J.,

Sessitsch A., Zechmeister-Boltenstern S., and Richter A. (2011) Plants control the seasonal dynamics of microbial N cycling in a beech forest soil by belowground C allocation. Ecology 92:1036-1051.

Klutsch J.G., Negron, J.F., Costello S. L,. Rhoades C.C., West D.R,. Popp J., and Caissie R.

(2009) Stand characteristics and downed woody debris accumulations associated with a mountain pine beetle (Dendroctonus ponderosae Hopkins) outbreak in Colorado. Forest Ecology and Management 258:641-649.

Knight D.H., Yavitt J.B., and Joyce G.D. (1991) Water and nitrogen outflow from lodgepole

pine forest after two levels of tree mortality. Forest Ecology and Management 46:215-225.

Kowalski A.S. Kowalski, A. S., Loustau D., Berbigier P., Manca G., Tedeschi V., Borghetti M.,

Valentini R., Kolari P., Berninger F., Rannik U., Hari P., Rayment M., Mencuccini M., Moncrieff J., and Grace J. (2004) Paired comparisons of carbon exchange between undisturbed and regenerating stands in four managed forests in Europe. Global Change Biology 10:1707-1723.

Kurz W.A. Dymond C.C., Stinson G., Rample G.J., Neilson, A.L. Carroll E.T., Ebata T., and

Safranyik L. (2008) Mountain pine beetle and forest carbon feedback to climate change. Nature 452:987-990.

Larsen K.S., Grogan P., Jonasson S., and Michelsen A. (2007) Respiration and microbial

dynamics in two subarctic ecosystems during winter and spring thaw: effects of increased snow depth. Arctic, Antarctic, and Alpine Research 39:268-276.

Law B.E., Ryan M.G. and Anthoni P.M. (1999) Seasonal and annual respiration of a ponderosa

pine ecosystem. Global Change Biology 5:169-182. Le Quere C. et al. (2009) Trends in the sources and sinks of carbon dioxide. Nature Geoscience

2:831-836.

!

!$#*!

Liang N., Hirano T., Zheng Z.M., Tang J., and Fujinuma Y. (2010) Soil CO2 efflux of a larch

forest in northern Japan. Biogeosciences 7:3447-3457. Likens G., Bormann F.H., Johnson N.M., Fisher D.W., and Pierce R.S. (1970) Effects of forest

cutting and herbicide treatment on nutrient budgets in the Hubbard Brook Watershed-Ecosystem. Ecological Monographs 40:23-47.

Littell J.S. et al. (2010) Forest ecosystems, disturbance, and climatic change in Washington State,

USA. Climatic Change 102:129-158. Logan J.A., Regniere J., and Powell J.A. (2003) Assessing the impacts of global warming on

forest pest dynamics. Frontiers in Ecology and the Envirnoment 1:130-137. Logan J.A., MacFarlane W.W. and Willcox L. (2010) Whitebark pine vulnerability to climate-

driven mountain pine beetle disturance in the Greater Yellowstone Ecosystem. Ecological Applications 20:895-902.

Magnani F. Mencuccini M., Borghetti M, Berbigier P., Berninger F., Delzon S., Grelle A., Hari

P., Jarvis P.G., Kolari P., Kowalski A.S., Lankreijer H., Law B.E., Lindroth A., Loustau, D., Manca G., Moncrieff J.B., Rayment M., Tedeschi V., and Valentini, R. (2007) The human footprint in the carbon cycle of temperate and boreal forests. Nature 447:848-850.

Man G. (2010) Major Forest Disease Conditions in the United States: (2009) Update. In. Forest

Health Protection Offices, United States Forest Service, Department of Agriculture. Manning, A.C., and Keeling R. F., (2006) Global oceanic and land biotic carbon sinks from the

Scripps atmospheric oxygen flask sampling network, Tellus-B, 58B, 95-116. van Mantgem P.G., Stephenson N.L., Byrne J.C., Daniels L.D., Franklin J.F., Fulé P.Z., Harmon

M.E., Larson A.J., Smith J.M., Taylor A.H., and Veblen T.T. (2009) Widespread Increase of Tree Mortality Rates in the Western United States. Science 323:521–524.

Marschner P., Crowley D., and Rengel Z. (2011) Rhizosphere interactions between

microorganisms and plants govern iron and phosphorus acquistion along the root axis - model and research methods. Soil Biology and Biochemistry 42:883-894.

McDowell N., Pockman, W.T., Allen C.D., Breshears D.D., Cobb N., Kolb T., Plaut J., Sperry J.,

West A., Williams D.G., and Yepez E.A. (2008) Mechanisms of plant survival and mortality during drought: why do some plants survive while others succumb to drought? New Phytologist 178:719-739.

McDowell N.G., Beerling D.J., Breshears D.D., Fisher, R.A. Raffa K.F., and Stitt M. (2011) The

interdependence of mehanisms underlying climate-driven vegetation mortality. Trends in Ecology and Evolution 26:523-532.

!

!$#+!

Metropolis N., and Ulam S. (1949) The Monte Carlo method. Journal of the American Statistical Association, 44: 335-341.

Metzger K.L., Smithwick E.A.H., Tinker D.B., Romme W.H., Balser T.C., and Turner M.G.

(2008) Influence of coarse wood and pine saplings on nitrogen mineralization and microbial communities in young post-fire Pinus contorta. Forest Ecology and Management 256:59-67.

Mitton J.B., and Ferrenberg S.M. (2012) Mountain pine beetle develops an unprecedented

summer generation in response to climate warming. The American Naturalist 179: DOI: 10.1086/665007.

Monson R. K., Lipson D.L., Burns S.P., Turnipseed A.A., Delany A.C., Williams M.W., and

Schmidt S.K. (2006) Winter forest soil respiration controlled by climate and microbial community composition. Nature 439:711-714.

Monson R.K., Burns S.P., Williams M.W., Delany A.C., Weintraub M., and Lipson D.A. (2006)

The contribution of beneath-snow soil respiration to total ecosystem respiration in a high-elevation, subalpine forest. Global Biogeochemical Cycles 20:GB3030, 1-13.

Monson R.K.,Sparks J.P., Rosenstiel T.N., Scott-Denton L.E., Huxman T.E., Harley P.C.,

Turnipseed A.A., Burns S.P., Backlund B., Hu J. (2005) Climatic influences on net ecosystem CO2 exchange during the transition from wintertime carbon source to springtime carbon sink in a high-elevation, subalpine forest. Oecologia 146:130-147.

Monson R.K., Turnipseed A. A., Sparks J. P., Harley P. C., Scott-Denton L. E., Sparks K., and

Huxman T. E. (2002) Carbon sequestration in a high-elevation, subalpine forest. Global Change Biology 8:459-478.

Moore D.J.P., Hu J., Sacks W.J., Schimel D.S., and Monson R.K. (2008) Estimating

transpiration and the sensitivity of carbon uptake to water availability in a subalpine forest using a simple ecosystem process model informed by measured net CO2 and H2O fluxes. Agricultural and Forest Meteorology 148:1467-1477.

Moore J.C., McCann K., Setala H., and De Ruiter P.C. (2003) Top-down is bottom-up: does

predation in the rhizosphere regulate aboveground dynamics? Ecology 84:846-857. Morehouse K., Johns T., Kaye J., and Kaye A. (2008) Carbon and nitrogen cycling immediately

following bark beetle outbreaks in southwestern ponderosa pine forests. Forest Ecology and Management 255:2698-2708.

Morgner E., Elberling B., Strebel, D., and Cooper E.J. (2010) The importance of winter in annual

ecosystem respiration in the high artic: effects of snow depth in two vegetation types. Polar Research 29:58-74.

!

!$$#!

Mote P.W. A.F. H, Clark M.P., and Lettenmaier D.P. (2005) Declining mountain snowpack in Western North America. Bulletin of the American Meteorological Society 86:39-49.

Moyes A.B., Schauer, A.J., Siegwolf, R.T.W., and Bowling, D.R. (2010) An injection method

for measuring the carbon isotope content of soil carbon dioxide and soil respiration with a tunable diode laser absorption spectrometer. Rapid Communications in Mass Spectrometry 24:894-900.

Niinistö S.M., Kellomake S., and Silvola J. (2011) Seasonality in a boreal forest ecosystem

affects the use of soil temperature and moisture as predictors of soil CO2 efflux. Biogeosciences 8:3169-3186.

Odum E.P. (1969) Strategy of Ecosystem Development. Science 164:262. O'Halloran T.L., Law B.E., Goulden M.L., Wang Z., Barr J.G., Schaaf C., Brown M., Fuentes

J.D., Gockede M., Black A., and Engel V. (2012) Radiative forcing of natural forest disturbances. Global Change Biology 18:555-565.

Ollson P., Linder S., Giesler R., and Högberg P. (2005) Fertilization of boreal forest reduces both

autotrophic and heterotrophic soil respiration. Global Change Biology 11:1745-1753. Oquist M.g., and Laudon H. (2008) Winter soil frost conditions in boreal forests control growing

season soil CO2 concentration and its atmospheric exchange. Global Change Biology 14:2839-2847.

Pacala S.W., Hurtt G.C., Baker D., Peylin P., Houghton R.A., Birdsey R.A., Heath L., Sundquist

E.T., Stallard R.F., Ciais P., Moorcroft P., Caspersen J.P., Shevliakova E., Moore B., Kohlmaier G., Holland E., Gloor M., Harmon M.E., Fan S.M., Sarmiento J.L., Goodale C.L., Schimel D., and Field C.B. (2001) Consistent land- and atmosphere-based US carbon sink estimates. Science 292:2316-2320.

Parker T.J., Clancy K.M., and Mathiasen R.L. (2006) Interactions among fire, insects and

pathogens in coniferous forests of the interior western United States and Canada. Agricultural and Forest Entomology 8:167-189.

Pearson J. A., Knight D.H., and Fahey T.J. (1987) Biomass and nutrient accumulation during

stand development in Wyoming lodgepole pine forests. Ecology 68:1966-1973. Pfeifer E.M., Hicke J.A., and Meddens A.J.H. (2011) Observations and modeling of

aboveground tree carbon stocks and fluxes following a bark beetle outbreak in the western United States. Global Change Biology 17:339-350.

Piao S., Ciais P., Friedling P., Peyin P., Reichstein M., Luyssaert S., Margolis H., Fang, J., Barr

A., Chen A., Grelle A., Hollinger D.Y., Tuoas L., Lindroth A., Richardson A.D., and Vesala T. (2008) Net carbon dioxide losses of northern ecosystems in response to autumn warming. Nature 451:49-53.

!

!$$$!

Potts D.F. (1984) Hydrologic impacts of a large-scale mountain pine beetle (Dendroctonus

Ponderosae Hopkins) epidemic. Water Resources Bulletin 20:373-377. Pregitzer K.S., and Euskirchen E.S. (2004) Carbon cycling and storage in world forests: biome

patterns related to forest age. Global Change Biology 10:2052-2077. Pugh E. and Small E. (2012) The impact of pine beetle infestation on snow accumulation and

melt in the headwaters of the Colorado River. Ecohydrology doi: 10.1002/hyp.9312. Pugh E.T. and Gordon E.S. A conceptual model of water yield impacts from beetle-induced tree

death in snow-dominated lodgepole pine forests. Hydrological Processes. doi: 10.1002/hyp.9312.

Pypker, T.G., Unsworth M.H., Lamb B., Allwine E., Edburg S., Sulzman E., Mixa A.C., and

Bond B.J. (2003) Cold air drainage in a forested valley: Investigating the feasibility of monitoring ecosystem metabolism. Agricultural and Forest Meteorology, 145 3-4, 149-166.

Pypker,T.G., Unsworth M.H., Mix A.C., Rugh W., Ocheltree T., Alstad K., and Bond B.J.

(2007). Using Nocturnal Cold Air Drainage Flow to Monitor Ecosystem Processes in Complex Terrain. Ecological Applications, 17(3), 702-714.

Raffa K.F., Aukema B. H. Bentz, B.J., Carroll A.L., Hicke J.A., Turner M. G., and Romme W.H.

(2008) Cross-scale drivers of natural disturbances prone to anthropogenic amplification: The dynamics of bark beetle eruptions. Bioscience 58:501-517.

Raffa, K.F., Aukema, B.H., Erbilgin, N., Klepzig, K.D., and Wallin, K.F. (2005) Interactions

among conifer terpenoids and bark beetles across multiple levels of scale: An attempt to understand links between population patterns and physiological processes. Vol. 39:79-118 in J.T. Romeo, editor. Recent Advances in Phytochemistry. Elsevier, St. Louis, MO. 318 p.

Raffa, K.F., and Berryman A.A. (1983) The role of host plant resistance in the colonization

behavior and ecology of bark beetles (Coleoptera: Scolytidae). Ecological Monographs 53:27-49.

Raich J.W., and Schlesinger W.H. (1992) The global carbon dioxide flux in soil respiration and

its relationship to vegetation and climate. Tellus 44B:81-99. Rhine E.D., Sims G.K., Mulvaney R.l., and Pratt E.J. (1998) Improving the Berthelot reaction for

determining ammonium in soil extracts and water. Soil Science Society of America Journal 62: 473-480.

Richardson, A.D., Braswell B.H., Hollinger D.Y., Burman P., Davidson E.A., Evans R.S.,

Flanagan L.B., Munger J.W., Savage K., Urbanski S.P., and Wofsy S.C. (2006)

!

!$$%!

Comparing simple respiration models for eddy flux and dynamic chamber data. Agricultural and Forest Meteorology 141:219-234.

Richardson A.D., Hollinger D.Y., Aber J.D., Ollinger S.V., and Braswell B.H. (2007)

Environmental Variation is directly responsible for short- but not long-term variation in forest-atmosphere carbon exchange. Global Change Biology 13:788-803.

Riveros-Iregui D.A., Hu J., Burns S.P., Bowling D.R. and Monson R.K. (2011) An interannual

assessment of the relationship between the stable isotopic composition of ecosystem respiration and climate in a high-elevation subalpine forest. Journal of Geophysical Research 106: G02005 doi:10.1029/2010JG001556.

Rogers M.C., Sulliva P.F., and Welker J.M. (2011) Evidence of nonlinearity in the response of

net ecosystem CO2 exchange to increasing levels of winter snow depth in the high arctic of northwest Greenland. Arctic, Antarctic, and Alpine Research 43:95-106.

Romme W.H., Knight D.H., and Yavitt J.B. (1986) Mountain pine beetle outbreaks in the rocky

mountains: regulators of primary productivity? The American Naturalist 127:484-494. Running S.W. (2008) Climate change - Ecosystem disturbance, carbon, and climate. Science

321:652-653. Ryan M.G., and Law B.E. (2005) Interpreting, measuring, and modeling soil respiration.

Biogeochemistry 73:3-27. Sacks W.J., Schimel D.S., and Monson R.K. (2007) Coupling between carbon cycling and

climate in a high-elevation, subalpine forest: a model-data fusion analysis. Oecologia 151:54-68.

Sampson D.A., Janssens I.A., Curiel Y. J., and Ceulemans R. (2007) Basal rates of soil

respiration are correlated with photosynthesis in a mixed temperate forest. Global Change Biology 13:2008-2017.

Schaeffer S.M., Anderson D.E., Burns S.P., Monson R.K., Sun J., and Bowling D.R. (2008)

Canopy structure and atmospheric flows in relation to the !13C of respired CO2 in a subalpine coniferous forest. Agricultural and Forest Meteorology 148: 592-605.

Schimel D. (2007) Carbon cycle conundrums. Proceedings of the National Academy of Sciences

of the United States of America 104:18353-18354. Schimel D., and Braswell, B.H. (2005) The role of mid-latitude mountains in the carbon cycle:

global perspective and a Western US case study. Springer, Dordrecht, The Netherlands. Schimel D.S., Churkina G., Braswell B.H., and Trembath J. (2005b) Remembrance of weather

past: Ecosystem response to climate variability. In: A History of Atmospheric CO2 and its

!

!$$&!

Effects on Plants, Animals and Ecosystems, (J. Ehleringer; Ed.), Springer-Verlag, pp. 350-368.

Schimel D., Kittel T.G.F., Running S., Monson, R., Turnipseed A., and Anderson D. (2002)

Carbon Sequestration studies in Western U.S. mountains. EOS 83:445-456. Schimel D., Melillo J., Tian H., McGuire A. D., Kicklighter D., Kittel T.G.F., Rosenbloom N. A.,

Running S. W., Thornton P. E., Ojima D. S., Parton W. J., Kelly R., Sykes M., Neilson R.P., and Rizzo B. (2000) Contribution of increasing CO2 and climate to carbon storage by ecosystems in the United States. Science 287:2004-2006.

Schimel J.P., and Bennett J. (2004) Nitrogen mineralization: challenges of a changing paradigm.

Ecology 85:591-602. Schindlbacher A., Zechmeister-Boltenstern S., and Jandl R. (2009) Carbon losses due to soil

warming: do autotrophic and heterotrophic soil respiration respond equally? Global Change Biology 15:901-913.

Schoennagel T., Turner M.G., and Romme W.H. (2003) Influence of fire interval and serotiny on

postfire lodgepole pine density in Yellowstone National Park. Ecology 84:2967-2978. Scott-Denton L.E., Rosenstiel T.N., and Monson R.K. (2006) Differential controls by climate

and substrate over the heterotrophic and rhizospheric components of soil respiration. Global Change Biology 12:205-216.

Scott-Denton L.E., Sparks K.L., and Monson R.K. (2003) Spatial and temporal controls of soil

respiration rate in a high-elevation, subalpine forest. Soil Biology & Biochemistry 35:525-534.

Seidl, R., Fernandes, P.M., Fonseca, T.F., Gillet, F., Jöhnsson, A.M., Merganicová, K., Netherer,

S., Arpaci, A., Bontemps, J.D., Bugmann, H., González-Olabarria, J.R., Lasch, P., Meredieu, C., Moreira, F., Schelhaas, M., and Mohren, G.M.J. (2011) Modelling natural disturbances in forest ecosystems: a review. Ecological Modelling 222 (4): 903 - 924.

Silver W.L., and Miya R.K. (2001) Global patterns in root decompositon: comparisons of

climate and litter quality effects. Oecologia 129:407-419. Sims D. A., Rahman A.F., Cordova V.D., El-Masri B.Z., Baldocchi D.D., Bolstad P.V, Flanagan

L.D., Goldstein A.H., Hollinger D.Y., Misson L., Monson R.K., Oechel W.C,. Schmid H.P., Wofsy S.C., and Xu L. (2007) A new model of gross primary productivity for North American ecosystems based solely on the enhanced vegetation index and land surface temperature from MODIS. Remote Sening of the Environment. 112, 1633-1646.

Six D.l., and Bentz B.J. (2007) Temperature determines symbiont abundance in a multipartite

bark beetle-fungus ectosymbiosis. Microbial Ecology 54:112-118.

!

!$$'!

Stadler B., Muller T., and Orwig D. (2006) The ecology of energy and nutrient fluxes in hemlock forests invaded by hemlock woolly adelgid. Ecology 87:1792-1804.

Stark J.M., and Hart S.C. (1977) High rates of nitrification and nitrate turnover in undisturbed

coniferous forests. Nature 385:61-64. Stenberg P., Linder S., Smolander H., and Flower-Ellis J. (1994) Performance of the LAI-2000

plant canopy analyzer in estimating leaf area index of some Scots pine stands. Tree Physiology 14:981-995.

Stephens B.B., Miles N.L., Richardson S.J., Watt A.S., and Davis K.J. (2011) Atmospheric CO2

monitoring with single-cell NDIR-based analyzers. Atmospheric Measurement Techniques Discussions, 4(4),4325-4355.

Stephens, B., Watt, A., and Maclean, G. (2005) An autonomous inexpensive robust CO2

analyzer (AIRCOA). 13th WMO/IAEA Meeting of Experts on Carbon Dioxide Concentration and Related Tracers Measurement Techniques. WMO TD 1359, 95-99, 2006.

Stoy P.C., Richardson A.D., Baldocchi D.D., Katul G.G., Stanovick J., Mahecha M.D.,

Reichstein M., Detto M., Law B.E., Wohlfahrt G., Arriga N., Campos J., McCaughey J. H., Montagnani L., Paw U K.T., Sevanto S., and Williams M. (2009) Biosphere-atmosphere exchange of CO2 in relation to climate: a cross-biome analysis across multiple time scales. Biogeosciences Discussions 6:4095-4141.

Stull, R.B. (1988) An Introduction to Boundary-Layer Meteorology. Dordrecht, Netherlands:

Kluwer Academic Publishers, 670. Subke J-A., Volker H., Battipaglia G., Linder S., Buchmann N., and Cotrufo M.F. (2004)

Feedback interactions between needle litter decomposition and rhizosphere activity. Oecologia 139:551-559.

Sun, J., Oncley S.P., Burns S.P., Stephens B.B., Lenschow D.H., Campos T., Monson, R.K.,

Schimel D.S., Sacks, W.J. DeWekker S.F., Lai C.-T., Lamb B., Ojima D., Ellsworth P.Z., Sternberg L.S., Zhong S., Clements C., Moore D.J., Anderson D.E. Watt, A.S., Hu J., Tschudi M., Aulenbach S., Allwine E., and Coons T. (2010) A multiscale and multidisciplinary investigation of ecosystem-atmosphere CO2 exchange over the Rocky Mountains of Colorado. Bulletin of the American Meteorological Society 91 (2), 209-230.

Tilman D., Reich P., Phillips H., Menton M., Patel A., Vos E., Peterson D., and Knops J. (2000)

Fire suppression and ecosystem carbon storage. Ecology 81:2680-2685. Trumbore S.E. (2006) Carbon respired by terrestrial ecosystems- recent progress and challenges.

Global Change Biology 12:141-153.

!

!$$(!

Turner M.G., Smithwick E.A.H., Metzger K.L., Tinker D.B., and Romme W.H. (2007) Inorganic nigrogen availability after severe stand-replacing fire in the Greater Yellowstone ecosystem. Proceedings of the National Academy of Sciences 104:4782-4789.

Turnipseed, A. A., Anderson D. E., Blanken P. D., Baugh W. M., and Monson R. K. (2003)

Airflows and turbulent flux measurements in mountainous terrain Part 1. Canopy and local effects. Agricultural and Forest Meteorology 119:1-21.

United States Forestry Service Forest Insect & Disease Aerial Survey Data Download; Available

from http://www.fs.fed.us/r2/resources/fhm/aerialsurvey/download/ [Accessed on the 6th June, 2011].

Valentini R., Matteucci G., Dolman A.J., Schulze E.D., Rebmann C., Moors E.J., Granier A.,

Gross P., Jensen N.O., Pilegaard K., Lindroth A., Grell A., Bernhofer C., Grunwald T., Auginet M., Ceulemans R., Kowalski A.A., Vesala T., Rannik U., Berbigler P., Loustau D., Guomundsson J., Thorgeirsson H., Ibrom A., Morgenstern K., Clement R., Moncrieff J., Montagnani L., Minerbi S., and Jarvis P.G. (2000) Respiration as the main determinant of carbon balance in European forests. Nature 404:861-865.

Vance E.D., Brookes P.C., and Jenkinson D.S. (1987) An extraction method for measuring soil

microbial biomass C. Soil Biology & Biochemistry 19:703-707. Vitousek P.M., Gosz J. R., Grier C.C., Melillo J.M., Reiners W.A., and Todd R.L. (1979) Nitrate

losses from disturbed ecosystems. Science 204:469-474. Weintraub M.N., Scott-Denton L.E., Schmidt S.K., and Monson R.K. (2007) The effects of tree

rhizodeposition on soil exoenzyme activity, dissolved organic carbon, and nutrient availability in a subalpine forest ecosystem. Oecologia 154:327-338.

Whiteman, C. D. (2000) Mountain Meteorology: Fundamentals and Applications. Oxford

University Press, New York, 355pp. Wilkes P. (2009) A new remote sensing model to quantify the influence of the mountain pine

beetle on GPP in subalpine forests. Dissertation submitted as part of a MSc degree in Environmental Monitoring, Modelling and Management at King’s College London, 60 pp.

Yi C.X., Anderson D.E., Turnipseed A.A., Burns S.P., Sparks J.P., Stannard D.I., and Monson

R.K., 2008: The contribution of advective fluxes to net ecosystem CO2 exchange in a high elevation, subalpine forest. Ecological Applications, 18, 1379-1390.

Zeller, B., Liu J., Buchmann N., and Richter A. (2008) Tree girdling increases soil N

mineralisation in two spruce stands. Soil Biology & Biochemistry 40:1155-1166.

Documents

Mountain Pine Beetle Disturbance and Climate Impacts on … · University of Colorado, Boulder CU Scholar Ecology & Evolutionary Biology Graduate Theses & Dissertations Ecology &