UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Measuring agricultural greenhouse gas exchange over a conventionally managed highbush blueberry field Pow, Patrick Ka Chun 2019

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2019_may_pow_patrick.pdf [ 12.15MB ]
Metadata
JSON: 24-1.0378327.json
JSON-LD: 24-1.0378327-ld.json
RDF/XML (Pretty): 24-1.0378327-rdf.xml
RDF/JSON: 24-1.0378327-rdf.json
Turtle: 24-1.0378327-turtle.txt
N-Triples: 24-1.0378327-rdf-ntriples.txt
Original Record: 24-1.0378327-source.json
Full Text
24-1.0378327-fulltext.txt
Citation
24-1.0378327.ris

Full Text

MEASURING AGRICULTURAL GREENHOUSE GAS EXCHANGE OVER A CONVENTIONALLY MANAGED HIGHBUSH BLUEBERRY FIELD  by  Patrick Ka Chun Pow  B.Sc., The University of Waterloo, 2017  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  MASTER OF SCIENCE in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Soil Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  April 2019  © Patrick Ka Chun Pow, 2019 ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:  Measuring agricultural greenhouse gas exchange over a conventionally managed highbush blueberry field  submitted by Patrick Pow in partial fulfillment of the requirements for the degree of Master of Science In Soil Science  Examining Committee: T. Andrew Black, Land and Food Systems Supervisor  Mark Johnson, Earth, Ocean and Atmospheric Sciences Supervisory Committee Member  Sean Smukler, Land and Food Systems Supervisory Committee Member Les Lavkulich, Land and Food Systems Examination Chair Ian McKendry, Geography Additional Examiner  Additional Supervisory Committee Members: Rachhpal Jassal, Land and Food Systems Supervisory Committee Member iii  Abstract Agricultural fields are significant sources of carbon dioxide (CO2), nitrous oxide (N2O) and methane (CH4), which has implications for future climate change. In Canada, studies incorporating all three greenhouse gases (GHGs) in agricultural settings are limited to Ontario, Quebec and the Prairies and generally involve short-duration static-chamber measurements. Results from these studies may not generalize well to British Columbia (BC) on the west coast, which has a milder climate and different cropping systems. This study quantified year-round (January 1, 2018 – December 31, 2018) CO2, N2O and CH4 exchange over a conventionally managed highbush blueberry field on Westham Island in Delta, BC, Canada using the eddy-covariance (EC) method. Continuous measurements using EC allowed for quantification of diurnal courses of both CO2 and N2O exchange, whereas sporadic measurements may not accurately reproduce the complete diurnal cycle of GHG emissions. Sawdust mulching may have contributed to a reduction in evapotranspiration but has implications for increased CO2 and N2O emissions. Field management including fertilization and mowing was associated with substantial changes in GHG exchange, suggesting that management strategies can be targeted for potential GHG mitigation. The field was a net source of all measured GHGs and emitted 838 g CO2e m-2 year-1, with CO2 contributing the largest proportion (76%) followed by N2O (20%) and CH4 (4%). The annual net ecosystem exchange (NEE) was 173 g C m-2 year-1 with the ratio of annual gross primary productivity (GPP) to ecosystem respiration (Re) being 0.88. After accounting for inputs and outputs of carbon (C), the field sequestered a net of 231 g C m-2 year-1. While soil temperature was found to be an important environmental factor controlling GHG emissions, soil moisture was also found to be an important factor, which has implications on future feedback cycles and climate change. iv  Lay Summary Agricultural fields are significant sources of greenhouse gases (GHGs). The development of mitigation strategies to reduce GHG emissions from agricultural fields requires actual estimates of GHG emissions. The purpose of this study was to quantify annual emissions of carbon dioxide (CO2), nitrous oxide (N2O) and methane (CH4) exchange over a conventionally managed highbush blueberry field on Westham Island in Delta, BC, Canada. The field was found to be a source of all three GHGs, which is important as other agricultural systems have been found to be sinks for CO2 and CH4. While N2O and CH4 are considered more potent GHGs, CO2 was found to be the greatest contributor, and has implications on future climate change. v  Preface With the support of the UBC Biometeorology and Soil Physics Group, I was responsible for all data analysis, installation and maintenance of field instrumentation. I conducted initial soil sampling for bulk density and water content measurements; additional soil carbon and nitrogen content sampling was conducted by Dr. Sean Smukler’s Sustainable Agricultural Landscapes Lab at UBC. This research was not previously published wholly or in part. vi  Table of Contents Abstract ......................................................................................................................................... iii Lay Summary ............................................................................................................................... iv Preface .............................................................................................................................................v Table of Contents ......................................................................................................................... vi List of Tables ..................................................................................................................................x List of Figures .............................................................................................................................. xii List of Symbols .............................................................................................................................xx Acknowledgements .................................................................................................................. xxiv Dedication ...................................................................................................................................xxv Chapter 1: Introduction ................................................................................................................1 1.1 Agricultural greenhouse gas emissions ........................................................................... 1 1.2 Blueberry management ................................................................................................... 2 1.3 Eddy covariance technique ............................................................................................. 2 1.4 Carbon dioxide ................................................................................................................ 3 1.5 Nitrous oxide ................................................................................................................... 3 1.6 Methane........................................................................................................................... 4 1.7 Research objectives ......................................................................................................... 5 Chapter 2: Methods .......................................................................................................................6 2.1 Site description................................................................................................................ 6 2.2 Climate and soil measurements ...................................................................................... 7 2.3 Eddy-covariance measurements and calculations ........................................................... 8 2.3.1 Instrumentation and methodology .............................................................................. 8 vii  2.3.2 Flux quality control and assessment ......................................................................... 11 2.3.3 Energy balance and radiation .................................................................................... 11 2.3.4 Evapotranspiration .................................................................................................... 12 2.4 Greenhouse gas fluxes .................................................................................................. 14 2.4.1 Carbon dioxide .......................................................................................................... 14 2.4.2 Nitrous oxide ............................................................................................................. 15 2.4.3 Annual greenhouse gas budget and CO2 equivalents ............................................... 16 Chapter 3: Results and discussion ..............................................................................................18 3.1 Weather measurements ................................................................................................. 18 3.2 Soil measurements ........................................................................................................ 20 3.3 Flux footprint models .................................................................................................... 21 3.4 Energy balance and surface properties ......................................................................... 24 3.4.1 Radiation balance components ................................................................................. 24 3.4.2 Energy balance components ..................................................................................... 26 3.4.3 Energy balance closure ............................................................................................. 28 3.5 Evapotranspiration ........................................................................................................ 31 3.5.1 Surface conductance ................................................................................................. 31 3.5.2 Priestley-Taylor equilibrium evaporation ................................................................. 32 3.5.3 Water use efficiency ................................................................................................. 33 3.5.4 Decoupling coefficient .............................................................................................. 36 3.6 Greenhouse gas exchange ............................................................................................. 37 3.6.1 Temporal variation in turbulent exchanges ............................................................... 37 3.6.1.1 Diurnal trends in CO2 exchange ....................................................................... 39 viii  3.6.1.2 Diurnal trends in N2O exchange ....................................................................... 41 3.6.1.3 Diurnal trends in CH4 exchange ....................................................................... 43 3.6.2 Annual greenhouse gas budget ................................................................................. 45 3.6.3 NEE partitioning ....................................................................................................... 47 3.6.4 Annual carbon budget ............................................................................................... 50 3.6.5 Environmental controls on greenhouse gas emissions.............................................. 50 3.6.5.1 Environmental controls on CO2 emissions ....................................................... 50 3.6.5.2 Environmental controls on N2O emissions ....................................................... 54 3.6.5.3 Environmental controls on CH4 emissions ....................................................... 56 Chapter 4: Conclusions ...............................................................................................................57 4.1 Summary of key findings .............................................................................................. 57 4.2 Implications for future research .................................................................................... 58 Bibliography .................................................................................................................................59 Appendices ....................................................................................................................................70 Appendix A Site photos and schematic .................................................................................... 70 A.1 Site photos and blueberry phenology ........................................................................ 70 A.2 Site instrumentation and schematic .......................................................................... 71 Appendix B Soil pit and soil-water retention curve .................................................................. 73 B.1 Soil profile ................................................................................................................ 73 B.2 Soil-water retention curve ......................................................................................... 74 Appendix C Diurnal and annual wind rose at z = 3.0 m ........................................................... 75 C.1 Daytime and nighttime wind rose at the site ............................................................. 75 C.2 Annual flux footprint climatology ............................................................................ 76 ix  Appendix D Instrument and measurement methodology ......................................................... 77 D.1 Gas analyzer comparison .......................................................................................... 77 D.2 Monin-Obuhkov Similarity Theory .......................................................................... 77 Appendix E Approximation of canopy conductance ................................................................ 81 Appendix F Uncertainty in carbon budget ................................................................................ 82  x  List of Tables Table 1: Baseline surface soil characteristics measured at the site in 2015.................................... 7 Table 2: Comparison of 2018 mean monthly air temperature (𝑇a) and precipitation (𝑃) at the site with the 30-year (1981 – 2010) normal at the Vancouver International Airport (YVR) climate station, approximately 2 km away from the research site (49.19°N, 123.18°W, 4.30 m.a.s.l., WMO ID: 71892) .......................................................................................................................... 20 Table 3: Regression parameters of the ordinary least squares regression between 𝐸 and GPP for estimating water use efficiency (WUE) for each month of 2018. ................................................ 35 Table 4: Monthly greenhouse gas emissions of carbon dioxide (CO2), nitrous oxide (N2O) and methane (CH4) measured in carbon dioxide equivalents CO2e using global warming potentials on a 100-year timescale considering climate-carbon feedbacks in g CO2e m-2 month-1. ........... 47 Table 5: Monthly CO2 flux partitioning into total monthly ecosystem respiration (𝑅e), gross primary production (GPP) and net ecosystem production (NEE) at the site ................................ 49 Table 6: Model parameters for the exponential relationship (Humphreys et al., 2005) between nighttime ecosystem respiration (𝑅e) and the 5-cm soil temperature (𝑇s) when the 5-cm volumetric water content (𝜃) > 0.4 during the study period. ...................................................... 53 Table 7: Model parameters for the exponential-temperature relationship with a hyperbolic soil moisture relationship (Gaumont-Guay et al., 2006) between nighttime ecosystem respiration (𝑅e) and the 5-cm soil temperature (𝑇s) during the study period. ........................................................ 53 Table 8: Model parameters for the logistic relationship (Barr et al., 2004) between nighttime ecosystem respiration (𝑅e) and the 5-cm depth soil temperature (𝑇s) during the study period. .. 54 xi  Table 9: Model parameters for the exponential-temperature and Gaussian-moisture relationships between nitrous oxide flux (𝐹N2O) and the 5-cm depth soil temperature (𝑇s) and volumetric water content (𝜃) at the site during the study year. ...................................................................... 56 Table 10: Soil-water retention curve coefficients and model parameters for the van Genuchten model............................................................................................................................................. 74 Table 11: Model parameters for the Monin-Obuhkov similarity relationship between the normalized standard deviation of the vertical wind velocity and the universal stability parameter for unstable and stable conditions at the site during the study year.............................................. 80 Table 12: Range of sawdust application rates in g C m-2 year-1 determined from low and high application height and bulk density. C content was assumed to be 0.50 kg C (kg dry matter)-1 and fraction of the field covered by sawdust estimated to be 0.45 m2 sawdust (m2 field)-1. ............... 82  xii  List of Figures Figure 1: Climate variables at the site during the study year. Panel a) shows 1-day (24-hour) precipitation (𝑃Daily) and cumulative precipitation (𝑃Cum), panel b) shows 1-day mean downwelling shortwave radiation (𝑆d), panel c) shows 1-day mean air temperature (𝑇a) and panel d) shows 1-day mean daytime vapour pressure deficit (𝐷day). Daytime was determined by 𝑆d or potential 𝑆d > 0. ................................................................................................................. 19 Figure 2: Soil variables shown as 5-day means at the site during the study year. Panel a) shows mean soil temperature (𝑇s) and panel b) shows mean volumetric water content (𝜃5) measured at the 5-cm depth. Panel c) shows mean depth to water table (WT). Panel d) shows soil matric potential 𝜑5 measured at the 5-cm depth. ..................................................................................... 21 Figure 3: Annual 24-h flux footprint climatology during the study period at the site. The x- and y-axis indicate the distance east-west and north-south from the tower, with the origin indicating the tower location. The solid contour lines indicate from 10 to 90% the cumulative probability of source area for the measured turbulent fluxes at the tower. The area enclosed by the polygon indicates the boundaries of the blueberry field. ............................................................................ 22 Figure 4: Daytime flux footprint climatology during the study period at the site. Daytime measurements are determined when downwelling shortwave radiation or potential global irradiance is above 0. The x- and y-axis indicate the distance east-west and north-south from the tower, with the origin indicating the tower location. The solid contour lines indicate from 10 to 90% the cumulative probability of source area for the measured turbulent fluxes at the tower. The area enclosed by the polygon indicates the boundaries of the blueberry field. ..................... 23 xiii  Figure 5: Nighttime flux footprint climatology during the study period at the site. Daytime measurements are determined when downwelling shortwave radiation or potential global irradiance is above 0. The x- and y-axis indicate the distance east-west and north-south from the tower, with the origin indicating the tower location. The solid contour lines indicate from 10 to 90% the cumulative probability of source area for the measured turbulent fluxes at the tower. The area enclosed by the polygon indicates the boundaries of the blueberry field. ..................... 24 Figure 6: Monthly mean radiation components of upwelling longwave (↑ 𝐿), downwelling longwave (↓ 𝐿), upwelling shortwave (↑ 𝑆) and downwelling shortwave (↓ 𝑆) and net radiation (𝑅n) at the site during the study year. ........................................................................................... 25 Figure 7: Monthly mean albedo at the site during the study year. ................................................ 25 Figure 8: Monthly mean energy balance components of net radiation (𝑅n), sensible heat flux (𝐻), latent heat flux (𝜆𝐸) and soil heat flux (𝐺) at the site during the study year. ...................... 27 Figure 9: Monthly mean Bowen ratio (𝛽) calculated as the ratio of mean monthly sensible heat flux to latent heat flux at the site during the study year. ............................................................... 27 Figure 10: Monthly mean diurnal sensible (𝐻) and latent (𝜆𝐸) heat flux at the site during the study year. The solid lines indicate 𝐻 and the dotted lines indicate 𝜆𝐸. ...................................... 28 Figure 11: Energy balance closure (EBC) calculated as the slope of the regression of 𝐻 + 𝜆𝐸 vs. 𝑅n − 𝐺 for a) annual (January 1 to December 31, 2018) and b) growing season (May 1 to August 1, 2018) measurements at the site. Annual and growing season EBC was 0.77 and 0.79. The solid black line is the linear regression and the dashed line indicates the 1:1 line. ............................... 30 xiv  Figure 12: Diurnal EBC in each month calculated from half-hourly measurements at the site during the study period. The closure fraction (𝐶MOR) was calculated as the mean of the ratio 𝐻+𝜆𝐸𝑅n−𝐺 for each half-hour. ................................................................................................................ 30 Figure 13: Diurnal EBC in each month calculated from half-hourly measurements at the site during the study period. The closure fraction 𝐶ROM was calculated as the ratio of 𝐻 + 𝜆𝐸̅̅ ̅̅ ̅̅ ̅̅ ̅ to 𝑅n − 𝐺̅̅ ̅̅ ̅̅ ̅̅ ̅ for each half-hour in each month. .................................................................................... 31 Figure 14: Mean monthly diurnal surface conductance (𝑔s) calculated from inverting the Penman-Monteith equation at the site during the study year. The gray area indicates ± one standard deviation of 𝑔s for each half-hour. ................................................................................. 32 Figure 15:  Daily evapotranspiration and Priestley-Taylor 𝛼 at the site during the study period. Panel (a) shows equilibrium evapotranspiration (𝜆𝐸eq) and measured evapotranspiration (𝜆𝐸) as solid and dotted lines respectively. Panel (b) shows the linear regression of 𝜆𝐸eq vs. 𝜆𝐸 and the ordinary least squares line of best fit as the solid black line. The slope of the regression was 0.37 (R2 = 0.86, RMSE = 17 W m − 2, n = 17520) and represents the Priestley-Taylor 𝛼. The dashed black line is the 1:1 line. ................................................................................................... 33 Figure 16: Monthly water use efficiency (WUE) calculated as the slope of the ordinary least squares regression of half-hourly gross primary production (GPP) vs. evaporation (𝐸). Slopes and regression parameters are in Table 3. The solid line is the regression line and the dashed line is the 1 g C to 1 kg H2O line. ........................................................................................................ 35 Figure 17: Mean monthly and daily intrinsic water use efficiency (iWUE) calculated as the ratio of gross primary production to surface conductance at the site during the study period. The gray lines and black lines indicate mean daily and monthly iWUE respectively.................................. 36 xv  Figure 18: Mean monthly McNaughton and Jarvis decoupling coefficient (Ω) at the site during the study year. The solid line indicates the annual mean Ω of 0.25. ............................................ 37 Figure 19: Mean daily fluxes of sensible heat (𝐻;  dotted line), latent heat (𝜆𝐸; solid line), carbon dioxide (CO2), nitrous oxide (N2O) and methane (CH4) at the study during the study year. Positive (negative) values indicate the field was a source (sink). Vertical dashed lines indicate pruning (PR), N-fertilization via four surface applications of ~25 – 30 kg N ha − 1 of ammonium nitrate (F1 − F4) and interrow mowing (M1 −M5). ............................................... 39 Figure 20: Monthly mean diurnal carbon dioxide flux (𝐹c) averaged for each half-hour at the site during the study year. Positive (negative) values indicate that the field was a source (sink) for carbon. The gray shading represents ± one standard deviation of 𝐹c for each half-hour. ............ 40 Figure 21: Fingerprint plot of ensemble diurnal carbon dioxide flux (𝐹c) for each month at the site during the study year. Negative (positive) values indicate that the field was a sink (source) for CO2. ......................................................................................................................................... 40 Figure 22: Monthly mean diurnal nitrous oxide flux (𝐹N2O) for each month at the site during the study year. The crop was fertilized with ~110 kg N ha − 1 of ammonium nitrate over four equal applications between late-April and early-July. The gray area indicates ± one standard deviation of 𝐹N2O for each half-hour. ........................................................................................................... 42 Figure 23: Fingerprint plot of ensemble diurnal nitrous oxide flux (𝐹N2O) for each month at the site during the study year. Negative (positive) values indicate that the field was a sink (source) for N2O. The crop was fertilized with ~110 kg N ha − 1 of ammonium nitrate over four equal applications between late-April and early-July............................................................................. 43 xvi  Figure 24: Monthly mean diurnal methane flux (𝐹CH4) for each month at the site during the study year. The fluxes were sporadic and did not exhibit a clear diurnal signal. The gray area indicates ± one standard deviation of𝐹CH4 for each half-hour..................................................................... 44 Figure 25: Fingerprint plot of ensemble diurnal methane flux (𝐹CH4) for each month at the site during the study year. Negative (positive) values indicate that the field was a sink (source) for CH4. ............................................................................................................................................... 45 Figure 26: Cumulative greenhouse gas emissions of carbon dioxide (CO2 − C), nitrous oxide (N2O − N) and methane (CH4 − C) at the site during the study year. The dashed line indicates the Intergovernmental Panel on Climate Change Tier 1 default emission factor which predicted an annual emission of 2.4 kg N2O − N ha − 1 y − 1 based on the nitrogen fertilization rate of 110 N ha − 1 y − 1. The dotted line indicates the Canada-specific estimate of annual N2O emissions proposed by Rochette et al. (2018) at 2.6 kg N2O − N ha − 1 y − 1. ......................... 46 Figure 27: CO2 flux partitioning, using standard Fluxnet-Canada protocols, into mean daily ecosystem respiration (𝑅e), gross primary production (GPP) and net ecosystem production (NEP) at the site during the study year. The sign convention is such that GPP − 𝑅e =  NEP, where a positive (negative) NEP indicates a carbon sink (source). .............................................. 49 Figure 28: Relationships between half-hourly observed nighttime ecosystem respiration (𝑅e) and modeled 𝑅e using a) an exponential temperature-respiration relationship 𝑅e = 𝑅10𝑄10[(𝑇s−10)/10] excluding measurements of 𝑅e when 𝜃 < 0.40, b) an exponential temperature-moisture relationship of 𝑅e with an additional term describing the sensitivity of 𝑅e  to 𝜃 as 𝑅e  =(𝑎 + 𝑏𝜃 +𝑐𝜃)𝑅10𝑄10[(𝑇s−10)/10] and c) a logistic temperature relationship 𝑅e =𝑟11+𝑒[𝑟2(𝑟3−𝑇𝑠)]  where 𝑇s is the 5-cm depth soil temperature, 𝜃 is the volumetric water content 𝑅10 and 𝑄10 are xvii  derived from log-transformed temperature-respiration relationships and 𝑎, 𝑏, 𝑐, 𝑟1, 𝑟2 and  𝑟3 are empirical coefficients. Coefficients and model parameters are in Tables 6 to 8. ......................... 52 Figure 29: Relationship between nighttime (downwelling shortwave radiation = 0) half-hourly temperature-normalized (ReN) ecosystem respiration (𝑅e) and 5-cm depth soil water content (𝜃).  ReN is expressed as the ratio of observed nighttime 𝑅e to 𝑅e estimated from an exponential temperature-respiration relationship. ............................................................................................ 53 Figure 30: Exponential relationship between 5-cm depth soil temperature (𝑇s) and nitrous oxide flux (𝐹N2O) at the site during the study year. 𝐹N2O was binned into 1℃ increments, with the error bars indicating the standard deviation of the temperature bin. Model parameters are in Table 9. 55 Figure 31: Gaussian relationship between 5-cm depth volumetric water content (𝜃) and nitrous oxide flux (𝐹N2O) at the site during the study period. 𝐹N2O was binned into 0.05 increments of 𝜃, with the error bars indicating the standard deviation of each 𝜃 bin. Model parameters are in Table 9..................................................................................................................................................... 55 Figure 32: Site photos, management examples and progression of blueberry phenology at the site during 2018. a) Eddy-covariance tower and instrumentation hut in January; dormant bud stage. b) March; bud swell stage. c) Late-April; bud burst stage. d) June; early green fruit stage. e) August post-harvest; early fall stage. f) October post-mowing; late fall stage. g) November post-pruning; dormant bud stage. h) December, dormant bud stage. ................................................... 70 Figure 33: Instrumentation hut including the a) LI-7000 infrared gas analyzer, b) the LGR continuous wave laser spectrometer and c) the pumps for each instrument. ................................ 71 Figure 34: Photo of the instrumentation hut and sonic anemometer. The sampling line was insulated and heated. ..................................................................................................................... 71 xviii  Figure 35: Site schematic illustrating the site setup. Bold and solid lines indicate sampling tubing and dotted or dashed lines indicate communications. Fluxes were calculated at the site from the sonic anemometer (Gill R3-50), infrared gas analyzer (LI-7000), and continuous wave laser spectrometer (LGR). ..................................................................................................................... 72 Figure 36: Soil profile (0 – 60 cm) at the row-interrow boundary and initial installation of soil monitoring sensors including CS616, MPS-1, and thermocouples at the 60-cm depth on October 23, 2017......................................................................................................................................... 73 Figure 37: Soil-water retention curve developed with the van Genuchten model using in situ measurements of matric potential 𝜑 and volumetric water content (𝜃) at the 5-cm depth at the site during the study period. 𝜑 is binned into 0.05 increments of 𝜃 with the error bars representing the standard deviation of each bin. ........................................................................... 74 Figure 38: Half-hourly wind rose for daytime (𝑆d > 0 or global potential radiation > 0) and nighttime (𝑆d = 0)  grouped into 10° bins for varying windspeeds at the site during 2018. ....... 75 Figure 39: Google Earth image of the flux footprint climatology at the site during the study year. The flux footprint was calculated using a parameterization by Kljun et al. (2015). The yellow triangle indicates the tower location. The footprint contour lines increase from 10% to 90%. While the 80% contour line is entirely within the field, the 90% contour line extends over an adjacent road and pasture. ............................................................................................................. 76 Figure 40: Comparison of latent heat flux measurements between the Los Gatos Research  (𝜆𝐸LGR) and LI-COR (𝜆𝐸LI7000) gas analyzers. The slope of the regression was 0.99, indicating high agreement. The regression line is the solid black line and the dashed line is the 1:1 line. ... 77 xix  Figure 41: The normalized standard deviation of the vertical wind velocity 𝜙𝑤 as a function of the stability parameter 𝑧m−𝑑𝐿 during stable conditions at the site during the study year. The solid line indicates the fitted function with parameters described in Table 7, and the dashed line is the function predicted from Monin-Obukhov similarity theory (Kaimal and Finnigan, 1994). ......... 79 Figure 42: The normalized standard deviation of the vertical wind velocity 𝜙𝑤 as a function of the stability parameter 𝑧m−𝑑𝐿  during unstable conditions at the site during the study year. The solid line indicates the fitted function with parameters described in Table 7, and the dashed line is the function predicted from Monin-Obukhov similarity theory (Kaimal and Finnigan, 1994). The x-axis is reversed to indicate that the absolute values calculated by the stability parameter are initially negative. ..................................................................................................................... 79 Figure 43: The difference between the canopy conductance (𝑔c) calculated from inverting the Penman-Monteith equation and the approximated canopy conductance (𝑔c ∗) compared to the ratio of aerodynamic to surface conductance  𝑔a𝑔c.  𝑔c − 𝑔c ∗ approaches 0 when  𝑔a𝑔c> 300. ...... 81  xx  List of Symbols Symbols/Acronyms Units Definition 𝐴 W m−2 Available energy flux BC  British Columbia 𝑐𝑝 kJ kg−1 K−1 Specific heat of air C  Carbon CMOR  Closure fraction as mean of ratios CROM  Closure fraction as ratio of means Csawdust g C m−2 year−1 Carbon in sawdust Csequestered g C m−2 year−1 Carbon sequestered on the field Cyield g C m−2 year−1 Carbon in blueberry yield CH4 μmol mol−1 Methane CO2 μmol mol−1 Carbon dioxide CO2e g m−2 time−1 Carbon dioxide equivalent 𝑑 m Zero displacement height 𝐷 kPa Vapour pressure deficit 𝐸 mm Evapotranspiration EC  Eddy covariance EBC  Energy balance closure EF  Emission factor 𝐹c μmol m−2 time−1 Carbon dioxide flux 𝐹CH4 nmol m−2 time−1 Methane flux 𝐹N2O nmol m−2 time−1 Nitrous oxide flux 𝐹s umol m−2 time−1 Scalar gas flux Fert kg N ha−1 year−1 Fertilizer application rate 𝑔a mm s−1 Aerodynamic conductance 𝑔c mm s−1 Canopy conductance 𝑔c ∗ mm s−1 Approximate canopy conductance xxi  𝑔s mm s−1 Surface conductance 𝐺 W m−2 Soil heat flux density GHG  Greenhouse gas GPP μmol m−2 time−1 Gross primary production  GWP g g−1 Global warming potential 𝐻 W m−2 Sensible heat flux iWUE μmol mol−1 Intrinsic water use efficiency IPCC  Intergovernmental Panel on Climate Change IRGA  Infrared gas analyzer 𝑘  Von Kármán constant ↓ 𝐿 W m−2 Downwelling longwave radiation ↑ 𝐿 W m−2 Upwelling longwave radiation 𝐿 m Obukhov length 𝑚s g mol−1 Molecular mass MDV  Mean diurnal variation 𝑀𝑖𝑛𝑁 kg N ha−1 year−1 Annual nitrogen-fertilizer application rate M-O  Monin-Obukhov N  Nitrogen N2O μmol mol−1 Nitrous oxide N2OTierI kg N ha−1 year−1 Tier I emissions of Nitrous oxide N2OTierII kg N ha−1 year−1 Tier II emissions of Nitrous oxide NEE g C m−2 time−1 Net ecosystem exchange NEP g C m−2 time−1 Net ecosystem productivity NH4NO3  Ammonium nitrate NR  Non-stationarity ratio 𝑝𝐻  Soil pH  𝑃 mm Precipitation xxii  ↓ 𝑄 W m−2 Downwelling photosynthetically active radiation ↑ 𝑄 W m−2 Upwelling photosynthetically active radiation 𝑄10  Exponential increase in reaction rate with a 10 K increase in temperature 𝑅10  Initial reaction rate at 10℃ 𝑅e μmol m−2 time−1 Ecosystem respiration 𝑅n W m−2 Net radiation 𝑅s μmol m−2 time−1 Soil respiration 𝑅eN  Temperature-normalized ecosystem respiration 𝑠 kPa ℃−1 Slope of the saturation vapour pressure curve ↓ 𝑆 W m−2 Downwelling shortwave radiation ↑ 𝑆 W m−2 Upwelling shortwave radiation 𝑆𝐴𝑁𝐷  Sand content 𝑇a ℃ Air temperature 𝑇s ℃ Soil temperature 𝑇sonic ℃ Sonic temperature ?̅? m s−1 Mean horizontal wind velocity 𝑢∗ m s−1 Friction velocity ?̅? m s−1 Mean crosswind wind velocity 𝑣 mmol mol−1 Mixing ratio for water vapour ?̅? m s−1 Mean vertical wind velocity WFPS  Water-filled pore space WT mm Water table depth 𝑧m m Measurement height 𝛼  Priestley-Taylor alpha xxiii  𝛽  Bowen ratio 𝛾 kPa ℃−1 Psychrometric constant ∆𝑆 W m−2 Rate of change in energy storage 𝜃 m3 m−3 Volumetric water content 𝜆 kJ kg−1 Latent heat of evaporation 𝜆𝐸 W m−2 Latent heat flux 𝜆𝐸eq W m−2 Equilibrium evapotranspiration 𝜆𝐸LGR W m−2 Latent heat flux measured from the LGR 𝜆𝐸LI7000 W m−2 Latent heat flux measured from the LI7000 ρa kg m−3 Dry air density 𝜎w m s−1 Standard deviation of the vertical wind velocity 𝜑 kPa Matric potential 𝜑h  Diabatic correction for heat and water vapour 𝜑m  Diabatic correction for momentum 𝜑w  Variance of the vertical wind velocity Ω  Decoupling coefficient    xxiv  Acknowledgements I am grateful for the mentorship of Dr. Andrew Black, whose passion for learning and selflessness motivated me to challenge myself. I would also like to thank Drs. Rachhpal Jassal, Sean Smukler, and Mark Johnson who offered insight and inspired me to push further. These people are giants, and it was a privilege to stand on their shoulders. The Biomet group was extremely supportive of my research and fostered a positive environment. I am indebted to Zoran Nesic, whose unyielding support and technical advice ensured successful installation and operation of all instrumentation. He also entertained my constant interferences, which was surely trying. I would also like to thank Jugoslav Kitanovic and Brian Wang who provided technical support in the field. I am also grateful to Hughie Jones, who occasionally graced me with his fleeting presence.  I would like to thank the Husband family of Emma Lea Farms, who accommodated the instrumentation hut and tower in the middle of their field and put up with the late nights I spent fumbling with the system. It was a pleasure working with such progressive growers and I learned deeply from this partnership. I would also like to thank Doug Worthy and Robert Kessler of the Climate Research Division at Environment and Climate Change Canada, Downsview for providing calibration gas, which allowed me to calibrate my gas analyzers. This research was funded by Agriculture and Agri-food Canada through the Agricultural Greenhouse Gases Program. I was partially supported by the National Sciences and Engineering Research Council (NSERC) through the Canada Graduate Scholarships-Master’s program and an NSERC Discovery Grant (T.A. Black). Additional scholarships and travel grants were provided by the UBC Faculty of Land and Food systems and the UBC Faculty of Graduate and Postdoctoral Studies.  xxv  Dedication To my loving partner and family, whose steadfast support ensured this endeavor was a labour of love. 1  Chapter 1: Introduction 1.1 Agricultural greenhouse gas emissions Agricultural soils are significant sources of carbon dioxide (CO2), nitrous oxide (N2O) and methane (CH4), and are estimated to contribute ~11% of global greenhouse gas (GHG) emissions, which has implications for climate change (Smith et al., 2014). In 2016, the agricultural sector contributed approximately 8.5% to Canadian GHG emissions, with the majority of the agricultural emissions (~40%) resulting from livestock production or manure management as CH4 emissions (Flemming et al., 2016). Reductions in beef cattle population and increased use of synthetic nitrogen (N) fertilizers from 1990 to 2016 (114%) has increased the relative importance of N2O emissions from agricultural soils, which in 2016 accounted for ~22% of Canadian agricultural GHG emissions (Flemming et al., 2016). In Canada, recent studies have measured all three GHG emissions from a variety of agricultural systems, but are mostly limited to Ontario, Quebec and the Prairies (Desjardins et al., 2010; Rochette et al., 2018) and generally relied on short-duration, point-source measurements integrated over time (i.e. static chambers). These measurements may not accurately describe seasonal or annual emissions of N2O and CH4 that are sporadic (Zona et al., 2013), close to detection limits (Nemitz et al., 2018), or have strong diurnal influences (Shurpali et al., 2016). While these studies have informed development of Canadian-specific Tier II emission factors (EFs), the uncertainty of Canadian agricultural N2O emissions was relatively high (±17%) compared to estimates from other industries, and contributed to greater uncertainty for estimations of direct emissions from agricultural soils (Flemming et al., 2016). Few continuous measurements of GHG emissions are available in British Columbia (BC), especially for regionally important, localized cropping systems like highbush blueberries which are not well established in the eastern or prairie provinces. As the 2  area of blueberry production increases in the Fraser River Delta, yields are reportedly stagnant, and as a result some growers are increasing rates of N-fertilization and potentially increasing N2O emissions, which has implications on field management and climate change. 1.2 Blueberry management Canada is the second largest producer of blueberries in the world, and BC produces over 90% of Canadian highbush blueberries (AAFC, 2012). Blueberry production in BC has also continuously increased in terms of both yield and harvested area (AgriService BC, 2018). Conventional N-fertilization rates of a highbush blueberry crop range from ~25 to 100 kg N ha−1 y−1 depending on maturity and the timing of mulch application (Larco et al., 2013), which is relatively lower compared to application rates (~156 to 191 kg N ha−1 y−1) for other agricultural crops (Skinner et al., 2014). Conventional management of blueberry fields involves regular application of sawdust mulch (Strik, 2007) for weed control and water conservation, but high C:N ratios in sawdust can immobilize N during decomposition resulting in higher N-fertilizer application rates (Larco et al., 2013). Higher soil respiration (𝑅s) has also been reported from mulched blueberry rows compared to un-mulched rows (Nemeth et al., 2017). While water conservation may be significant in the context of climate change, the potential for increased GHG emissions must also be considered.  1.3 Eddy covariance technique While many studies use static and dynamic chamber methods to estimate GHG emissions from point locations (Rochette et al., 2018), micrometeorological methods such as eddy covariance (EC), an accurate and defensible GHG measurement methodology (Baldocchi, 2003), can be used to quantify GHG exchange spatially and temporally. While EC has been effectively used to quantify the exchange of CO2, water vapour and energy for decades, recent 3  advancements in gas analyzer technology have enabled reliable, high-frequency measurements of non-CO2 GHGs like N2O and CH4 (Baldocchi, 2003; Nemitz et al., 2018; Rannik et al., 2015). However, measurements of N2O and CH4 fluxes are biased towards areas with expectedly high emissions, are generally sporadic with typical values close to the detection limit of conventional analyzers (Nemitz et al., 2018). N2O and CH4 exchanges are also difficult to describe empirically, and there are no standardized methods to assess data quality and for gap-filling missing data, which may have significant implications on estimates of annual GHG emissions (Mishurov and Kiely, 2011; Nemitz et al., 2018). Making continuous half-hourly measurements using EC can provide robust annual estimates of GHG emissions. 1.4 Carbon dioxide Historically, agricultural activities have been considered sources for CO2 due to land use change and intensive management in terms of repeated tillage (Baker and Griffis, 2005). Agricultural fields are also consistently losing carbon (C) as soil respiration (𝑅s), and depending on the crop, C also leaves the field in harvested products, which further contributes to ecosystem C loss (Baker and Griffis, 2005; Suyker and Verma, 2012). It has been suggested that changes in field management can reduce C loss (Hutchinson et al., 2007) or even reverse loss and sequester C (Robertson et al., 2000). While highbush blueberry fields are perennial systems and do not require tillage or other intensive activities associated with long-term soil C loss, management of blueberry fields includes a continuous import of C in the form of sawdust, which has implications on CO2 emissions in terms of increased decomposition and increased soil temperatures (𝑇s).  1.5 Nitrous oxide N2O is a potent GHG and has a 298-times greater global warming potential relative to CO2 on a 100-year timescale considering climate-carbon feedbacks (Myhre et al., 2013). N2O 4  emissions are closely related to soil moisture which can regulate oxygen availability and microbial activity related to nitrification and denitrification processes, which are responsible for the majority (~70%) of global N2O emissions (Butterbach-Bahl et al., 2013). An intermediate volumetric water content (𝜃) or water-filled pore space (WFPS) is considered favourable for N2O emission as lower 𝜃 would not provide sufficient anaerobic sites for denitrification while higher 𝜃 closer to saturation would promote complete denitrification to N2, with optimal WFPS for N2O emissions found to be between 0.65 and 0.80 (Balaine et al., 2013; Butterbach-Bahl et al., 2013). Country-specific Tier II EFs for N2O emissions have been developed in Canada for reporting to the Intergovernmental Panel on Climate Change (IPCC), but these calculations come from studies primarily from eastern Canada and may not account for differences in management or climate across Canada (Flemming et al., 2016; Rochette et al., 2018). While the proposed Canada-specific Tier II EF considers a wide range of environmental variables and is more complex than the IPCC default Tier I EF, studies have shown that the Tier II EF cannot describe the effect of heavy precipitation events or snowmelt (Desjardins et al., 2010; Rochette et al., 2018). Field scale studies can contribute more information on the range of N2O emissions, and provide additional insights on environmental controls, advancing Tier I or Tier II EFs towards Tier III (modelled and measurement-based) EFs regarding agricultural GHG emissions when reporting to the IPCC, especially for N2O. 1.6 Methane CH4 is a less potent GHG compared to N2O and has a 34-times greater global warming potential relative to CO2 on a 100-year timescale considering climate-carbon feedbacks (Myhre et al., 2013). CH4 exchange is controlled by the microbial activity of methanogens and 5  methanotrophs in the soil profile, which generate and consume CH4 in anaerobic or aerobic zones, respectively (Dutaur and Verchot, 2007). The relative depth of the water table (Moore and Knowles, 1989) as well as the degree of soil saturation (Kaiser et al., 2018; Skiba et al., 1996) has a strong influence on the relative zones of each microbial community and can determine the direction of net CH4 exchange. Well-drained soils of agricultural and forest ecosystems and grasslands are considered net sinks for CH4 as the soils are sufficiently aerated to support methanotrophic activity (Dutaur and Verchot, 2007; Robertson et al., 2000; Skinner et al., 2014). 1.7 Research objectives 1. To make continuous, year-round measurements using the EC method of CO2, N2O and CH4 exchanges from a highbush blueberry field in the Lower Fraser Valley, and evaluate its net global warming potential. 2. To determine the complete C balance of the field, including the impacts of management and harvesting. 3. To develop relationships between GHG exchange and controlling soil and meteorological variables.  6  Chapter 2: Methods 2.1 Site description The research site was a conventionally managed highbush blueberry (Vacinium corymbosum L. cvs. Reka and Duke) field (49.078°N, 123.15°W) located on Westham Island, BC in the Fraser River delta. The blueberry plants are planted in a sawdust mulch strip ~1.5 m wide on equally spaced (~3 m) rows with a perennial grass on interrow that is periodically mowed. The blueberry plants are periodically trimmed and maintained at approximately 1.5 m in height. To compensate for the decomposing sawdust mulch, the height of the sawdust mounds is increased by ~0.05 m every ~3 years resulting with a sawdust application rate of ~75 m3 ha−1 y−1. Sawdust was last applied to the field in 2017. The soil series is Westham or Crescent (rego or ortho humic gleysols) and the textural class is either silty clay loam or silt loam (Table 1), characterized by relatively high soil fertility, poor drainage and saturated conditions for much of the year if adequate drainage is not available (Luttmerding, 1981). The site was unusually well-drained due to sub-surface drain tiles, and the field was sub-irrigated through the drain tiles at the ~45-cm depth with water from the Fraser River until late June when water became limited and overly saline. Westham Island is on the western edge of the Fraser River delta and potentially exposed to saline water when freshwater declines after the spring freshet. The water table depth (WT) fluctuated throughout the year, resulting in distinctive gleying and mottling at depths greater than 30 cm (Appendix B.1). During the study period (January 1 to December 31, 2018), the crop was fertilized with ~110 kg N ha−1 of ammonium nitrate (NH4NO3) via four surface applications of ~25 – 30 kg N ha−1 between April and July 2018. The harvested yield of blueberries during the study period was approximately 11,200 kg ha−1 (fresh mass). In a previous study (Neufeld et al., 2017) , soil sampling was conducted from 7  September to October 2015 to measure mean total soil C and other baseline soil characteristics (Table 1).  Table 1: Mean baseline surface soil characteristics measured at the site in 2015.  Depth (cm) Texture (%) Total C (kg m−2) Total N (kg m−2) Bulk density (kg m−3) pH Clay Silt Sand 0 − 15 24.4 70.3 4.0 2.3 0.20 0.81 4.7 15 − 30 26.7 69.3 5.2 1.8 0.15 N/A N/A 30 − 60 23.1 70.5 7.4 1.0 0.08 N/A N/A 60 − 100 21.8 70.6 9.9 0.9 0.06 N/A N/A  2.2 Climate and soil measurements A temperature and relative humidity probe (HMP-45C, Campbell Scientific Inc. (CSI), Logan, UT, USA) mounted at the 2-m height measured half-hourly average air temperature (𝑇𝑎) and relative humidity (RH). Precipitation (𝑃) was measured half-hourly using a tipping-bucket rain gauge (TR525M, Texas Electronics Inc., Dallas, TX, USA) mounted at the 1.5-m height. The tipping-bucket was not modified to receive snowfall. A four-component net radiometer (CNR1, Kipp & Zonen, Delft, The Netherlands) and two quantum sensors (LI-190, LI-COR Inc., Lincoln, NE, USA) mounted at the 2.5-m height measured net radiation (𝑅n) and downwelling and upwelling photosynthetically active radiation (↓ 𝑄 and ↑ 𝑄). Another upward facing quantum sensor was mounted at the 1.0-m height for comparisons of ↓ 𝑄 and to allow for onsite comparisons by switching the upward and downward facing quantum sensors. Nighttime was differentiated from daytime measurements when downwelling shortwave radiation (↓ 𝑆) was ≤ 0 or when potential ↓ 𝑆 was 0 (Stull, 1988). For soil measurements, sensors were installed into 8  both sides of a soil pit at the interface between a blueberry row and a grass interrow. Two profiles of soil temperature (𝑇𝑠) and 𝜃 were measured half-hourly at depths of 5, 10, 30 and 60 cm with copper-constantan thermocouples (Type-T, Omega Engineering, Norwalk, CT, USA) and water content reflectometers (CS616 and CS615, CSI) respectively. Soil matric potential (𝜑) was measured half-hourly at depths of 5, 10, 30, and 60 cm under the blueberry row with dielectric water potential sensors (MPS-1, METER Group, Inc., Pullman, WA, USA). A soil-water retention curve (Appendix B.2) was developed using the van Genuchten model (1980) from in situ measurements of 𝜃 and 𝜑 made at the 5-cm depth. Ground heat flux (𝐺) was measured half-hourly with 6 heat flux plates (CN3, Middleton Solar, Melbourne, VIC, Australia) installed at the 3-cm depth on a 3-m transect covering both a grass-interrow and a blueberry-row. 𝐺 was not corrected for the rate of heat storage change in the 0 to 3-cm depth. All sensors were measured using a datalogger (CR3000, CSI) and two signal multiplexers (AM25T and AM16/32, CSI) every 2-4 seconds for climate variables, and 10 minutes for soil variables, and downloaded to UBC daily through a mobile broadband modem. Datalogger measurements were integrated at 60 Hz to remove electrical noise, predominately generated by the pumps inside the hut.  2.3  Eddy-covariance measurements and calculations 2.3.1 Instrumentation and methodology The eddy-covariance (EC) method was used to measure turbulent exchange of the three GHGs, CO2 (𝐹c), N2O (𝐹N2O) and CH4 (𝐹CH4), and sensible (𝐻) and latent (𝜆𝐸) heat, at a height of 3.0 m above the ground (𝑧m). The EC system comprised a three-dimensional sonic anemometer (Gill R3-50, Gill Instruments, Lymington, UK) and two parallel closed-path gas analyzers including an infrared gas analyzer (IRGA) (LI-7000, LI-COR Inc., Lincoln, NE, USA) and a continuous-wave laser spectrometer (LGR) (model 913-1054, Los Gatos Research Inc., 9  San Jose, CA, USA). Air was independently sampled by the IRGA and the LGR through parallel 5.5 m long, 6.35 mm diameter Synflex 1300 Metal/Plastic composite tubing, with air flow rates of 15-18 L min−1 and 20 L min−1 respectively. The sampling lines were insulated and heated 5-7 K above ambient 𝑇𝑎 to avoid condensation. The sampling line was replaced every ~4 months to sustain fast analyzer response times. Before entry into each gas analyzer, each sampling line was initially fitted with disposable 1 μm plastic filters but substituted with reusable stainless steel 2 μm filters that were replaced weekly. The gas analyzers could not be run in series (i.e. with one sampling line through both analyzers) and required two separate sampling lines due to modification of the air as it travelled through each system and instrument specific internal pressure controls. The R3-50 was mounted such that the middle of the sonic array was at a height of 3.0 m, approximately double the blueberry canopy height (1.5 m) following preseason pruning. To minimize lateral and longitudinal sensor separation, the distance between the sonic array and the air intakes for both analyzers was fixed at ~10 cm (< 10% 𝑧m), and the longitudinal distance (~5.5 m) was addressed by optimizing cross-correlations of measurements from the gas analyzers and the sonic temperature (𝑇sonic)  and shifting the measurements in time (Moore, 1986). The R3-50 and LI-7000 measured at a frequency of 20 Hz while the LGR measured at varying frequencies between 8 – 12 Hz. Measurements from the LGR were linearly interpolated and resampled at 20 Hz to synchronize with other EC measurements. The IRGA and LGR were calibrated in the field manually every 1 to 2 weeks, and automatically every night between June and December using the same calibration gas for both instruments. The zero offsets were measured using high purity (> 99.998) dinitrogen gas and the instruments were spanned using gas of known CO2, N2O and CH4 concentration supplied by Environment and Climate Change Canada. The R3-50, IRGA and LGR were connected to a site computer using RS232 serial 10  connections, and half-hourly EC fluxes were calculated on-site and downloaded daily to the UBC Biometeorology and Soil Physics lab. High-frequency data was stored on a USB flash drive and transferred to UBC every 1 to 3 weeks, from which recalculations were performed and compared against daily calculated values.  Half-hourly fluxes of 𝐹c, 𝐹N2O and 𝐹CH4 were calculated as: 𝐹𝑠 =  𝜌𝑎′̅̅ ̅̅ ̅ ∙ 𝑤′𝑠′̅̅ ̅̅ ̅̅  (1)  where 𝑤′𝑠′̅̅ ̅̅ ̅  is the covariance of the vertical windspeed (𝑤) and one of the scalar gas constituents (𝑠) measured as a mixing ratio, and 𝜌𝑎 is the dry air density, accounting for the effects of density fluctuations from heat and water vapour on flux calculations (Webb et al., 1980). 𝐻 was calculated from the covariance of  𝑤 and 𝑇𝑎 where 𝑇𝑎 is approximated by 𝑇sonic. 𝜆𝐸 was calculated similarly from the covariance of 𝑤 with 𝑣 where 𝑣 is the mixing ratio of water vapour measured using both the IRGA and the LGR. Measurements of 𝜆𝐸 using both the LGR and the IRGA following calibration showed great agreement and the comparison is shown in Appendix D.1. Linear detrending was not performed on scalar constituents to conserve low frequency components and their contributions to flux measurements (Finnigan et al., 2003; Massman, 2000). For each half-hour, three coordinate rotations were performed on the wind velocity components such that after the first rotation, u is aligned parallel to the mean wind direction, after the second rotation ?̅? = ?̅? = 0, and after the third rotation 𝑣′𝑤′̅̅ ̅̅ ̅̅ = 0 where ?̅? and ?̅? are the mean crosswind and vertical velocities, respectively (Kaimal & Finnigan, 1994; Wilczak et al, 2001). MATLAB® was used for all calculations and data analysis (The MathWorks, Inc.). 11  2.3.2 Flux quality control and assessment Fluxes were assessed half-hourly to remove suspect measurements. Fluxes were filtered automatically for spikes (Humphreys et al., 2003), low friction velocity (𝑢∗ < 0.15 m s−1) and unaccepted wind directions between 350° and 90°  (< 20% of the half-hourly wind directions originating from the direction of a nearby road and grass pasture). Filtering for low 𝑢∗ and wind direction removed 28% and 39% of the fluxes, respectively and 48% of the fluxes overall. A non-stationarity ratio (NR) (Mahrt, 1998) was calculated and fluxes were flagged for NR > 3.5 (Humphreys et al., 2005). Due to potentially episodic events of 𝐹N2O and 𝐹CH4 that would exhibit non-stationarity but are biologically significant, fluxes from the three GHGs were only removed due to non-stationarity when 𝐹c was found to be non-stationary (Nemitz et al., 2018). Less than 1% (168 total, 49 additional after other controls) of half-hourly fluxes were removed due to non-stationarity in 𝐹c. Measurements were also rejected when values exceeded sensible limits. Small gaps in the data less than 2 hours were filled with linear interpolation, and larger gaps were filled with the mean diurnal variation (MDV) over one to two weeks (Falge et al., 2001; Nemitz et al., 2018). Measurements of the standard deviation of the vertical wind velocity (𝜎𝑤) from the sonic anemometer were preliminarily assessed using variance similarity theory (Appendix D.2) to describe the variability in 𝑤 compared to the variability predicted by Monin-Obuhkov similarity theory (Foken, 2006; Kaimal and Finnigan, 1994). 2.3.3 Energy balance and radiation Energy balance closure (EBC) was determined to assess the accuracy of measured turbulent fluxes. The sum of turbulent energy exchange of 𝐻 and 𝜆𝐸, and the rate of storage change of 𝐻 and 𝜆𝐸 in the air column below 𝑧m, ∆𝑆, should be equivalent to the sum of available energy (A) to drive that exchange, measured as 𝑅n − 𝐺 − ∆𝑆, such that 𝐻 + 𝜆𝐸 = 𝑅𝑛 − 𝐺. EBC 12  was determined as the slope of the ordinary least squares (OLS) linear regression of 𝐻 + 𝜆𝐸 vs. 𝐴 (Aubinet et al., 2000). To further assess EBC on different time-scales, the closure fraction (𝐶) was also calculated following two methodologies proposed by Kidston et al. (2010) including the mean of ratios (𝐶MOR) and ratio of means (𝐶ROM) methods. 2.3.4 Evapotranspiration Surface conductance (𝑔s) was calculated by inverting the Penman-Monteith equation to determine the impact of both environmental and biophysical controls on evapotranspiration as (Jones et al., 2017; Monteith, 1965):  𝑔𝑠 =𝑔𝑎𝑠𝐴 + 𝜌𝑎𝑐𝑝𝐷𝑔a𝛾 ∙ 𝜆𝐸 −𝑠𝛾 − 1 (2) where 𝑔a is the aerodynamic conductance, 𝑠 is the slope of the saturation vapour pressure curve with temperature, 𝐴 is the available energy flux from 𝑅n − 𝐺 neglecting air column storage change, 𝜌𝑎 is the dry air density, 𝑐𝑝 is the specific heat of air, 𝐷 is the vapour pressure deficit, and 𝛾 is the psychrometric constant.  𝑔a was calculated as (Gash et al., 1999):  𝑔a =𝑢∗𝑢𝑢∗+2𝑘 +𝜑ℎ − 𝜑𝑚𝑘 (3) where  𝑘 = 0.4 is the von Kármán constant. 𝜑ℎ and 𝜑𝑚 are the diabatic stability correction factors for heat and water vapour and momentum respectively, and were calculated for unstable and stable conditions (Campbell and Norman, 1998), respectively as:  𝜑ℎ = −2 ln [1 + (1 − 16𝑧m − 𝑑𝐿 ) =2] ;     𝜑𝑚 = 0.6 𝜑ℎ               for  𝑧m − 𝑑𝐿≤ 0 (4) 𝜑ℎ = 𝜑𝑚 = 6 ln (1 +𝑧m − 𝑑𝐿)                                                            for  𝑧m − 𝑑𝐿> 0 (5) 13  where 𝑧m = 3.0 m is the measurement height, 𝑑 =23ℎ = 1.0 m  is the zero displacement height (Campbell and Norman, 1998) and 𝐿 is the Obuhkov length (Monteith and Unsworth, 2013). Measurements of  𝜆𝐸 were assessed by calculating equilibrium evapotranspiration (𝜆𝐸eq) as (Priestley and Taylor, 1972):   𝜆𝐸eq =𝑠𝑠 + 𝛾(𝑅n − 𝐺) (6) The site specific 𝛼 was also obtained as the slope of the OLS regression of 𝜆𝐸 vs. 𝜆𝐸eq (Pereira, 2004). The decoupling coefficient (Ω) calculated as (McNaughton and Jarvis, 1983): 𝛺 =𝑠𝛾 + 1𝑠𝛾 +𝑔𝑎𝑔𝑠+ 1(7) was used to determine the relative importance of 𝑔𝑎 and 𝑔𝑠. As 𝑔𝑎𝑔𝑠 approaches zero, 𝛺 approaches 1 and the atmosphere and the surface become decoupled, with the surface having low physiological control and 𝐸 is dominated by radiative forcing or 𝑅n (McNaughton and Jarvis, 1983). Conversely, as 𝑔𝑎𝑔𝑠 becomes large and approaches zero 𝛺 approaches 0, and the atmosphere and the surface become highly coupled, indicating that the surface has high physiological control and 𝐸 is dominated by stomatal control and 𝐷 (McNaughton and Jarvis, 1983).  As C assimilated by photosynthesis is related to the amount of water evaporated from the surface, ecosystem water use efficiency (WUE) was calculated as the slope of the OLS regression of daily GPP vs. 𝜆𝐸 annually and monthly (Law et al., 2002). Intrinsic water use efficiency (iWUE) was calculated as the ratio of half-hourly GPP to 𝑔𝑠 (Beer et al., 2009). 14  2.4 Greenhouse gas fluxes 2.4.1 Carbon dioxide  To assess different controls on CO2 exchange processes, 𝐹c measurements were partitioned into the two components of net ecosystem exchange (NEE) such that: NEE = −NEP = −GPP + 𝑅e (8) where NEE is the sum of the EC-measured half-hourly 𝐹c and changes in CO2 storage in the air column beneath the EC measurement height, NEP is net ecosystem productivity, GPP is gross primary production, and 𝑅e is ecosystem respiration. NEE was partitioned using standard Fluxnet-Canada Research network protocols with a moving window approach (Barr et al., 2004; see Falge et al., 2001 and Stoy et al., 2006). Daytime 𝑅e was estimated using a logistic relationship between nighttime 𝑅e and 𝑇s, and GPP was estimated as the sum of daytime NEP and daytime 𝑅e. GPP was gap-filled using an empirical relationship between ↓ 𝑄 and GPP (Barr et al., 2004). An adjustment factor was not applied to force energy balance closure (EBC).  To assess the controls on 𝑅e, an empirical relationship between 𝑇𝑠 and 𝑅e was described empirically using coefficients from the logarithmic transformation ln𝑅𝑒 = 𝐴 + 𝐵𝑇𝑠 to satisfy assumptions for least squares regressions of the exponential equation (Humphreys et al., 2005):  𝑅e = 𝑅10𝑄10[(𝑇s−10)/10] (9. 1) where 𝑄10 is the exponential increase in 𝑅e with a 10℃ increase in temperature, 𝑅10 is the reference reaction rate at 10℃, and 𝑇s is the soil temperature measured at the 5-cm depth. Only nighttime 𝑅𝑒 with 𝜃 > 0.4 were used to remove limitations in 𝑅e due to low soil moisture. The dependence of respiration on both 𝜃 and 𝑇s was described using a similar method (Equation 9.1) but with an additional term such that (Gaumont-Guay et al., 2006): 15  𝑅e = (𝑎 + 𝑏𝜃 +𝑐𝜃)𝑅10𝑄10[(𝑇s−10)/10] (9.2) where 𝜃 is the volumetric water content at the 5-cm depth, and 𝑎, 𝑏 and 𝑐, are coefficients derived from a hyperbolic relationship with temperature-normalized 𝑅e (𝑅eN). A logistic relationship between 𝑇s and 𝑅e was also studied as (Barr et al., 2004): 𝑅𝑒 =𝑟11 + 𝑒[𝑟2(𝑟3−𝑇s)](9.3) where 𝑇s is the soil temperature measured at the 5-cm depth and 𝑟1, 𝑟2, and 𝑟3 and empirical constants. 2.4.2 Nitrous oxide  The IPCC Tier I default EF for N2O, determined globally by a relationship between N-fertilizer application rate and annual N2O emissions from N-fertilized fields was calculated as (Bouwman, 1996): N2OTierI =  1 + 0.0125 Fert (10) where N2OTierI is the annual N2O emission (kg N ha−1 year−1) and Fert is the fertilizer application rate (kg N ha−1 year−1). The relationship assumes a background emission of 1 kg N2O − N ha−1 y−1 (i.e., the 1 in Eq. 10) and predicts a 1.25% N-fertilizer induced increase in N2O emissions. For synthetic N-fertilized agricultural soils in Canada, Rochette et al. (2018) developed an empirical model for estimating cumulative N2O emissions and Canadian-specific EFs:  N2OTierII  = 𝑒[3.91+0.0022𝑃+0.0069𝑀𝑖𝑛𝑁−0.0032𝑆𝐴𝑁𝐷−0.747𝑝𝐻+0.097𝑇𝑎] (11) where  N2OTierII is the cumulative emission from N application, 𝑃 is the growing season precipitation from May to October, 𝑀𝑖𝑛𝑁 is the annual rate of N-fertilizer application, 𝑆𝐴𝑁𝐷 is 16  the sand content, 𝑝𝐻 is soil pH and 𝑇𝑎 is the mean annual air temperature. Both estimates of annual N2O emissions were compared to measured cumulative 𝐹N2O.  The relationship between soil moisture and 𝐹N2O was described using a Gaussian relationship such that (Balaine et al., 2013): 𝐹N2O = 𝑎1𝑒[(−(𝜃−𝑎2)𝑎3)2] (12) where 𝜃 is the volumetric water content at the 5-cm depth and 𝑎1, 𝑎2 and 𝑎3 are empirical constants. 2.4.3 Annual greenhouse gas budget and CO2 equivalents The relative importance of 𝐹c, 𝐹CH4and 𝐹N2O were compared by multiplying each GHG flux by its the 100-year global warming potential (GWP) for N2O and CH4 in terms of CO2 equivalent (CO2e) mass fluxes (Lee et al., 2017):  CO2e = GWP𝑠𝑚𝑠𝐹𝑠  (13) where GWP𝑠 is the global warming potential for each GHG on a mass basis relative to CO2, 𝑚𝑠 is the molar mass of each GHG and 𝐹𝑠 is the molar flux for each GHG. 100-year GWP values for CH4 and N2O were 34 and 298, respectively (Myhre et al., 2013). Molar masses for CO2, CH4 and N2O were 44.01 g mol−1, 16.04 g mol−1 and 44.01 g mol−1, respectively.  The amount of carbon sequestered (Csequestered) on the field was calculated as:  Csequestered = −NEE + Csawdust − Cyield (14) where NEE is the net ecosystem exchange of CO2, Csawdust is the average amount of C entering the field from sawdust application, and Cyield is the amount of C in blueberry yield exported from the field (all in units of g C m−2y−1). The dry bulk density of sawdust was approximated at 17  130 kg dry mass m−3 (Larco et al., 2013; Nemeth et al., 2017). The C content in dry organic matter was estimated to be 50% (Larco et al., 2013).  18  Chapter 3: Results and discussion 3.1 Weather measurements The climate of Westham Island is characterized by relatively mild and wet winters, followed by a dry growing season (Figure 1). At the site, daily mean 𝑇𝑎 was predominantly above 0℃ throughout winter with mean annual value 10.6 ℃ (Figure 1). Annual 𝑃 was 1026.7 mm but growing season (May – August) 𝑃 was only 42.8 mm (Figure 1). A comparison of 𝑇𝑎 and 𝑃 data from the nearby (~2 km away) Vancouver International Airport climate station (YVR) indicated that mean annual 𝑇𝑎 and 𝑃 at the site were higher by 0.2℃ and lower by 162.3 mm, respectively compared to long-term averages at YVR (Table 2). Mean monthly 𝑇𝑎 in January was higher than normal, but lower than normal in February with several days dropping below 0℃ (Figure 1). Overwinter between December and February, 𝑃 was lower than normal by 35.4 mm, but was higher than normal by 33.8 mm in January alone (Table 2). Growing season 𝑃 was much lower than normal by 148.3 mm (Table 2). The tipping bucket at the site was not modified to receive snowfall, but 25.8 cm of snowfall was measured at YVR in February, with 11.0 cm day−1 recorded as the largest snowfall event on February 23, 2018. Although evaporative loss from the tipping bucket would reduce the precipitation measured, most of the snowfall was likely captured after melting as the depth of the funnel collector is approximately 10 cm and sufficiently deep to collect most of the snowfall. Bud burst occurred during the week of April 15, 2018 (Appendix A), coinciding with decreased 𝑃 and increased 𝑆𝑑 (Figure 1). The highest mean daytime D was 1.55 kPa in August, but D is relatively low for most days due to proximity to the ocean, and typically below 1 kPa during the growing season and less than 0.5 kPa during the rest of the year. The dominant wind direction varied diurnally at the site, with 19  the dominant wind direction from the north-west and south-east during the day and from the north-east during the night (Appendix C.1). The mean half-hourly windspeed during the day and night was 2.12 m s−1 and 1.60 m s−1, respectively, and 1.86 m s−1 for the entire study period. J F M A M J J A S O N D012Dday (kPa)-100102030Ta (C)0100200300400Sd (W m -2)010203040PDaily (mm day-1)03006009001200PCum (mm) Figure 1: Climate variables at the site during the study year. Panel a) shows 1-day (24-hour) precipitation (𝑃Daily) and cumulative precipitation (𝑃Cum), panel b) shows 1-day mean downwelling shortwave radiation (𝑆d), panel c) shows 1-day mean air temperature (𝑇a) and panel d) shows 1-day mean daytime vapour pressure deficit (𝐷day). Daytime was determined by 𝑆d or potential 𝑆d > 0.         20  Table 2: Comparison of 2018 mean monthly air temperature (𝑇a) and precipitation (𝑃) at the site with the 30-year (1981 – 2010) normal at the Vancouver International Airport (YVR) climate station, approximately 2 km away from the research site (49.19°N, 123.18°W, 4.30 m.a.s.l., WMO ID: 71892) Month Site 𝑇a 2018 (℃) YVR 𝑇a  2018 (℃)  30-year 𝑇a  1981 – 2010 (℃) Site 𝑃  2018 (mm) YVR 𝑃 2018 (mm) 30-year 𝑃  1981 – 2010 (mm)  January 5.1 5.4 4.1 202.2 249.4 168.4 February 3.3 3.4 4.9 76.3 105.8 104.6 March 5.9 6.1 6.9 79.4 111.8 113.9 April 9.4 9.2 9.4 128.2 134.8 88.5 May 14.5 14.9 12.8 1.6 1.6 65.0 June 15.5 15.9 15.7 34.3 38.8 53.8 July 18.5 19.1 18.0 4.1 5.4 35.6 August 17.3 18.5 18.0 2.8 16.2 36.7 September 13.9 14.6 14.9 119.6 111.0 50.9 October 9.6 9.8 10.3 84.9 110.0 120.8 November 7.4 7.6 6.3 172.3 203.8 188.9 December 5.0 4.9 3.6 121.0 254.8 161.9 Annual 10.6 10.8 10.4 1026.7 1343.4 1189.0  3.2 Soil measurements  In comparison to mean annual 𝑇a, mean annual 𝑇s at the 5-cm depth was higher at 11.9 ℃ and did not decrease below 0℃ annually due to moderation of temperature fluctuations from the sawdust amendments (Figure 2). 𝜃 and 𝜑 measured at the 5-cm depth remained constant through most of the year at 0.5 m3 m−3 and -10 kPa respectively, declining to 0.25 m3 m−3 and -90 kPa, respectively, at the end of the growing season in early September before fall 𝑃 (Figure 2). Mean depth to the WT also remained relatively constant but increased greatly following the first week of August due to limited sub-irrigation. 21   Figure 2: Soil variables shown as 5-day means at the site during the study year. Panel a) shows mean soil temperature (𝑇s) and panel b) shows mean volumetric water content (𝜃5) measured at the 5-cm depth. Panel c) shows mean depth to water table (WT). Panel d) shows soil matric potential (𝜑5) measured at the 5-cm depth.  3.3 Flux footprint models The flux footprint (FFP) was calculated to determine the source area of fluxes and acceptable wind directions using a parameterization by Kljun et al. (2015). The planetary boundary layer (PBL) height was conservatively estimated to be constant at 800 m, approximated from daytime summer LIDAR measurements (Hayden et al., 1997). While the PBL height would change throughout the day and affect the extent and shape of the footprint, the FFP parameterization is relatively insensitive to these changes for shorter towers as the difference between 𝑧 and the PBL remains consistently large (Heidbach et al., 2017). The 80% contour line, or the source area determined by the model to account for 80% of the measured flux, was entirely within the target field when considering the annual FFP climatology (Figure 3). Annually, the 90% contour line extended beyond the blueberry field onto an adjacent road and J F M A M J J A S O N D-100-5005 (kPa)010002000Depth to WT (mm)00.250.50.7515 (%)0102030Ts (C)22  pasture. The 90% contour line also extended beyond the target field for both daytime (Figure 4) and nighttime (Figure 5) measurements despite contraction of the FFP under unstable conditions (Baldocchi, 1997). While the FFP can be extended under stable stratification (Baldocchi, 1997), the 80% contour line was within the field for both daytime and nighttime measurement periods and sufficient to represent measurements from the field. Half-hourly measurements that contained more than 20% measurements from these directions (350° − 90° from the tower) were removed in data processing.  -300 -200 -100 0 100 200 300-350-300-250-200-150-100-50050100East-west distance from tower (m)North-south distance from tower (m) Figure 3: Annual 24-h flux footprint climatology during the study period at the site. The x- and y-axis indicate the distance east-west and north-south from the tower, with the origin indicating the tower location. The solid contour lines indicate from 10 to 90% the cumulative probability of source area for the measured turbulent fluxes at the tower. The area enclosed by the polygon indicates the boundaries of the blueberry field.   23  -300 -200 -100 0 100 200 300-350-300-250-200-150-100-50050100East-west distance from tower (m)North-south distance from tower (m) Figure 4: Daytime flux footprint climatology during the study period at the site. Daytime measurements are determined when downwelling shortwave radiation or potential global irradiance is above 0. The x- and y-axis indicate the distance east-west and north-south from the tower, with the origin indicating the tower location. The solid contour lines indicate from 10 to 90% the cumulative probability of source area for the measured turbulent fluxes at the tower. The area enclosed by the polygon indicates the boundaries of the blueberry field.    24  -300 -200 -100 0 100 200 300-350-300-250-200-150-100-50050100East-west distance from tower (m)North-south distance from tower (m) Figure 5: Nighttime flux footprint climatology during the study period at the site. Daytime measurements are determined when downwelling shortwave radiation or potential global irradiance is above 0. The x- and y-axis indicate the distance east-west and north-south from the tower, with the origin indicating the tower location. The solid contour lines indicate from 10 to 90% the cumulative probability of source area for the measured turbulent fluxes at the tower. The area enclosed by the polygon indicates the boundaries of the blueberry field.  3.4 Energy balance and surface properties 3.4.1 Radiation balance components Monthly 𝑅n and ↓ 𝑆 increased with regular Northern hemispherical seasonality, peaking in July at 157 W m−2 and 299 W m−2 respectively. Although daylength is increasing throughout June until the summer solstice, mean monthly ↓ 𝑆 in June was 252 W m−2, lower relative to both May and July (Figure 3) due to increased backscattering or absorbance of ↓ 𝑆 from cloud cover in June (Figure 1) (Monteith and Unsworth, 2013). Net longwave radiation (↓ 𝐿−↑ 𝐿) was the least negative in January at −11 W m−2 and the most negative in July at −86 W m−2. 𝑅n was on average negative for only the month of December at −5 W m−2. Monthly mean albedo was similar throughout the year at approximately 0.20 except for January and February when it was 25  0.10 and 0.27, respectively. Low albedo in January was related to greater 𝑃 intensity and increased 𝜃, and higher albedo in February was due to snowfall persisting on the surface for several days (Figure 7). J F M A M J J A S O N D-50050100150200250300350400450 S S L LRnRadiation Components (W m -2) Figure 6: Monthly mean radiation components of upwelling longwave (↑ 𝐿), downwelling longwave (↓ 𝐿), upwelling shortwave (↑ 𝑆) and downwelling shortwave (↓ 𝑆) and net radiation (𝑅n) at the site during the study year.  J F M A M J J A S O N D00.10.20.30.40.50.60.70.80.91Albedo Figure 7: Monthly mean albedo at the site during the study year.   26  3.4.2 Energy balance components Monthly 𝐻 and 𝜆𝐸 increased with similar seasonality to 𝑅n, with 𝐻 generally lower than 𝜆𝐸 (Figure 8). 𝐻 was the highest in May at 62 W m−2 while 𝜆𝐸 peaked later in July at 59 W m−2 (Figure 8). Monthly 𝐻 was on average negative for several winter months, but still rose to positive values for short periods during the daytime (Figure 10). The Bowen ratio (𝛽) was approximately 1 during June and July and less than 1 during the rest of the year, with negative values during winter months reaching a low of−5.8 in December (Figure 9). 𝛽 was highest in May at 1.1 (Figure 9), coinciding with suddenly decreased precipitation and 𝜃 (Figure 1). Mean monthly diurnal fluxes of 𝐻 and 𝜆𝐸 peaked approximately 1-3 hours after noon with daytime 𝐻 higher than 𝜆𝐸 for all months except December, when 𝐻 was on average smaller in magnitude than 𝜆𝐸 during the daytime (Figure 10). Consistently negative fluxes of 𝐻 during the nighttime, used to drive small nighttime 𝜆𝐸 values, resulted in lower monthly 𝐻 relative to 𝜆𝐸, thereby resulting in 𝛽 less than one for several months. 𝐺 was on average positive between March and August and negative for the rest of the year (Figure 8). The sum of the energy balance components (𝐻 + 𝜆𝐸 + 𝐺) was highest in July at 123 W m−2 and in May at 122 W m−2, accounting for 0.78 and 0.86 of 𝑅n respectively (Figure 8).  27  J F M A M J J A S O N D-40-20020406080100120140160RnHEGEnergy Balance Components (W m-2) Figure 8: Monthly mean energy balance components of net radiation (𝑅n), sensible heat flux (𝐻), latent heat flux (𝜆𝐸) and soil heat flux (𝐺) at the site during the study year.    J F M A M J J A S O N D-6-5-4-3-2-1012 Figure 9: Monthly mean Bowen ratio (𝛽) calculated as the ratio of mean monthly sensible heat flux to latent heat flux at the site during the study year.  28  -50050100150200250January February March April May June0 12 24-50050100150200250July0 12 24August0 12 24September0 12 24October0 12 24November0 12 24DecemberH and E (W m-2 s-1)Local time (PST)  Figure 10: Monthly mean diurnal sensible (𝐻) and latent (𝜆𝐸) heat flux at the site during the study year. The solid lines indicate 𝐻 and the dotted lines indicate 𝜆𝐸.  3.4.3 Energy balance closure EBC, calculated as the slope of the OLS regression of 𝐻 + 𝜆𝐸 vs. 𝑅n − 𝐺,  was 0.77 (𝑅2 = 0.89, RMSE = 38 W m−2) for the entire study period (Figure 11a). Compared to annual EBC, growing season EBC (May 1 to August 1, 2018) was slightly higher at 0.79 (𝑅2 =0.94, RMSE = 42 W m−2) (Figure 11b). Values of the slopes and intercepts for the annual and growing season periods were 0.77 and 0.79 and −11 and 3 W m−2, respectively, which suggest slightly higher growing season EBC. Values of mean 24-hour closure calculated as either 𝐶MOR or 𝐶ROM were relatively consistent for all months for daytime measurements, with both increasing throughout the day, generally peaking in the late afternoon at values close to unity, suggesting a diurnal increase in closure (Figures 12 and 13). While values of 𝐶MOR are bound between 0 and 1, mean values of the energy balance components for 𝐶ROM indicated that on average, more turbulent exchange was measured than energy was available during nighttime 29  conditions, which may be related to the accuracy of measuring small fluxes at night (Figure 13). Similar to Kidston et al. (2010), 𝐶MOR and 𝐶ROM values for EBC were the lowest shortly before sunrise, and low shortly after sunset and were related to reduced nighttime turbulence associated with the formation of a stable boundary layer (Stull, 1988) (Figure 12 and 13).  EBC can be used as a general indication of EC measurement quality but lack of closure at many EC sites worldwide is well-known (Aubinet et al., 2000; Wilson et al., 2002). In a study comparing EBC at long-term EC sites (Fluxnet), EBC was typically less than unity and on average ~0.80, suggesting the closure measured in this study was comparable to other studies. Fluxnet data is typically spectrally corrected for high frequency loss, incorporating multiplicative factors of up to ~1.3 for 𝜆𝐸 (Massman, 2000; Wilson et al., 2002). Spectral corrections were not applied to measured fluxes in this study, which may have improved EBC. In a study comparing EBC of Fluxnet sites over different land surfaces, Stoy et al. (2013) found that the average EBC of croplands was 0.78 and was lower in comparison to other vegetated landscapes, and suggested that energy storage terms in the soil, particularly soil water, may be significant in irrigated landscapes. Low frequency loss and horizontal and vertical advection may have also reduced EBC (Foken, 2008) at this site due to complex local meteorology.  30  -400 -200 0 200 400 600 800-400-2000200400600800H + E = 0.77 A - 11.01, R2 = 0.89H + E  (W m-2)Rn - G (W m-2)a) Annual-400 -200 0 200 400 600 800H + E = 0.79 A + 3.26, R2 = 0.94Rn - G (W m-2)b) May 1 - August 1 Figure 11: Energy balance closure (EBC) calculated as the slope of the regression of 𝐻 + 𝜆𝐸 vs. 𝑅n − 𝐺 for a) annual (January 1 to December 31, 2018) and b) growing season (May 1 to August 1, 2018) measurements at the site. Annual and growing season EBC was 0.77 and 0.79. The solid black line is the linear regression and the dashed line indicates the 1:1 line.   00.51January February March April May June0 12 2400.51July0 12 24August0 12 24September0 12 24October0 12 24November0 12 24DecemberCMORLocal time (PST)  Figure 12: Diurnal EBC in each month calculated from half-hourly measurements at the site during the study period. The closure fraction (𝐶MOR) was calculated as the mean of the ratio 𝐻+𝜆𝐸𝑅𝑛−𝐺 for each half-hour.   31  012January February March April May June0 12 24012July0 12 24August0 12 24September0 12 24October0 12 24November0 12 24DecemberCROMLocal time (PST)  Figure 13: Diurnal EBC in each month calculated from half-hourly measurements at the site during the study period. The closure fraction 𝐶ROM was calculated as the ratio of 𝐻 + 𝜆𝐸̅̅ ̅̅ ̅̅ ̅̅ ̅ to 𝑅n − 𝐺̅̅ ̅̅ ̅̅ ̅̅ ̅ for each half-hour in each month.   3.5 Evapotranspiration 3.5.1 Surface conductance Monthly mean nighttime values of 𝑔𝑠 at ~2 mm s−1 throughout the year suggest consistent nighttime evaporation from the soil (Figure 14). Growing season 𝑔𝑠 exhibited a diurnal signal associated with daytime photosynthesis, which was pronounced between March and August, with peak afternoon 𝑔𝑠 increasing from 4.0 mm s−1 during winter months to 7.8 mm s−1 in April (Figure 14). Peak 𝑔𝑠 then decreased until August to 5.0 mm s−1 (Figure 14), coinciding with decreased 𝜆𝐸 (Figure 8) suggesting physiological responses to increased 𝐷 or available energy (Figure 1) or decreased 𝜃 (Figure 2). Similarly, canopy conductance was found to decrease following increased 𝐷 in Douglas-Fir stands (Jassal et al., 2009). 32  051015January February March April May June0 6 12 18 24051015July0 6 12 18 24August0 6 12 18 24September0 6 12 18 24October0 6 12 18 24November0 6 12 18 24Decembergs (mm s-1)Local time (PST)  Figure 14: Mean monthly diurnal surface conductance (𝑔s) calculated from inverting the Penman-Monteith equation at the site during the study year. The gray area indicates ± one standard deviation of 𝑔s for each half-hour.   3.5.2 Priestley-Taylor equilibrium evaporation Annual 𝐸 measured at the site was 348 mm year−1. Forcing EBC would increase the annual E to 428 mm year−1. Since 𝑃 was approximately 1027 mm, we estimated that net drainage (as the field was sub-irrigated) of the field was approximately 679 mm of water, making the reasonable assumption that the soil profile water storage was approximately the same at the beginning and end of the year. Measured 𝜆𝐸 was consistently lower than estimated 𝜆𝐸𝑒𝑞, the former peaking only 2.7 mm day−1 and the latter to 5.5 mm day−1, respectively in July (Figure 15). For some days in December, estimated 𝜆𝐸𝑒𝑞 was negative at about −0.5 mm day−1, indicating that the “imposed 𝜆𝐸” term 𝜌𝑎𝑐𝑝𝑔a𝐷𝛾 in the McNaughton and Jarvis (1983) form of the Penman-Monteith equation was large enough to result in positive 𝜆𝐸 (Figure 15). The Priestley-Taylor 𝛼 was 0.37 (R2 = 0.86, RMSE = 17 W m−2, n = 17520). The low 𝛼 can be partly 33  attributed to the sawdust-mulched surface, decreasing 𝑔𝑠 of the soil-atmosphere interface and reducing 𝜆𝐸 relative to 𝜆𝐸𝑒𝑞. The combination of low 𝛼 and relatively low 𝑔𝑠 has also been related to stomatal limitation to transpiration (Jassal et al., 2009). Estimates of peak season LAI (~1) also suggested that blueberry leaves do not have a large canopy leaf area, further reducing transpiration relative to evaporation from the soil. Using a modified Priestley-Taylor equation including LAI and the relative mulch cover, Ai and Yang (2016) showed that progressively increased mulch cover lowered 𝛼 values.   Figure 15:  Daily evapotranspiration and Priestley-Taylor 𝛼 at the site during the study period. Panel (a) shows equilibrium evapotranspiration (𝜆𝐸eq) and measured evapotranspiration (𝜆𝐸) as solid and dotted lines respectively. Panel (b) shows the linear regression of 𝜆𝐸eq vs. 𝜆𝐸 and the ordinary least squares line of best fit as the solid black line. The slope of the regression was 0.37 (R2 = 0.86, RMSE = 17 W m−2, n = 17520) and represents the Priestley-Taylor 𝛼. The dashed black line is the 1:1 line.    3.5.3 Water use efficiency Annually the WUE, calculated as the slope of the OLS regression of half-hourly GPP to  𝜆𝐸 was 2.8 g C (kg H2O)−1 (R2 = 0.76, RMSE = 0.06 g C (kg H2O)−1, n = 17520). WUE J F M A M J J A S O N D-20246Evaporation (mm day-1)a)-100 0 100 200 300 400 500-1000100200300400500Eeq (W m-2)E (W m-2)y = 0.365x+5.9r2=0.86RMSE = 17.457n = 17520b)34  calculated from only the growing season data was slightly lower at 2.6 g C (kg H2O)−1 (R2 =0.72, RMSE = 0.07 g C (kg H2O)−1, n = 5857), suggesting that WUE was lower during the growing season while GPP and 𝐸 were maximal. Monthly WUE peaks after the growing season in September and October, reaching 3.4 and 3.6 g C (kg H2O)−1 (Figure 16 and Table 3). Unlike WUE, iWUE increased during the growing season and peaked in July at 84 μmol mol−1 when WUE was relatively low (Figure 17 and Table 3). Given that GPP was common among both metrics, relative seasonal variations were due to differences in 𝑔𝑠 and 𝜆𝐸. Monthly WUE also demonstrated non-linearity for values of 𝐸 above approximately 0.1 kg H2O m−2day−1 especially during the growing season, suggesting that using a linear regression may have overestimated WUE (Figure 16). Using a non-linear function to calculate WUE would further decrease growing season WUE compared to non-growing season, which have relatively lower 𝐸. Decreased growing season WUE was likely related to moderate but increased 𝐷 (Figure 1), increasing the evaporative demand on the surface. Compared to a synthesis of WUE data for Fluxnet sites, this agricultural site had relatively lower WUE than both grasslands and croplands, which had reported WUE of 3.4 and 3.1 respectively (Law et al., 2002). As Beer et al. (2009) suggested that iWUE can also be approximated as the inherent water use efficiency, GPP𝐷𝐸, the approximation of 𝑔𝑐 using 𝐸𝐷, which was investigated in Appendix D.  35  00.30.6January February March April May June0 0.1 0.200.30.6July0 0.1 0.2August0 0.1 0.2September0 0.1 0.2October0 0.1 0.2November0 0.1 0.2DecemberE (kg H2O m-2 day-1)GPP g C m-2 day-1 Figure 16: Monthly water use efficiency (WUE) calculated as the slope of the ordinary least squares regression of half-hourly gross primary production (GPP) vs. evaporation (𝐸). Slopes and regression parameters are in Table 3. The solid line is the regression line and the dashed line is the 1 g C to 1 kg H2O line.  Table 3: Regression parameters of the ordinary least squares regression between 𝐸 and GPP for estimating water use efficiency (WUE) for each month of 2018.  Month WUE (g C kg H2O−1) 𝑅2 RMSE (g C kg H2O−1) January 1.9 0.44 0.01 February 2.4 0.63 0.02 March 2.4 0.67 0.03 April 2.7 0.75 0.05 May 2.5 0.78 0.06 June 2.3 0.68 0.08 July 2.7 0.77 0.07 August 2.8 0.69 0.08 September 3.4 0.71 0.06 October 3.6 0.76 0.04 November 3.2 0.63 0.02 December 1.5 0.32 0.02   36  J F M A M J J A S O N D-20020406080100120iWUE (mol mol-1) Figure 17: Mean monthly and daily intrinsic water use efficiency (iWUE) calculated as the ratio of gross primary production to surface conductance at the site during the study period. The gray lines and black lines indicate mean daily and monthly iWUE respectively.   3.5.4 Decoupling coefficient Ω was consistently low suggesting that the surface was highly coupled to the atmosphere. As expected of a surface with significant roughness elements, annual Ω was only 0.25, suggesting high coupling to 𝐷 and that 𝑔a𝑔s was large (Figure 18). Ω was relatively low during the winter months of December to February, suggesting higher coupling to 𝐷 overwinter compared to the rest of the year. Ω was lowest in February at 0.18 and reached maximum decoupling in October at 0.30 (Figure 18). Ω at this agricultural site was comparable to coniferous forest and grassland ecosystems with limited water supply and low 𝑔s, with annual Ω values ranging from 0.15 to 0.25, but was lower than for ombrotrophic bogs with values approximately 0.30 which would be similar to irrigated conditions at the site in terms of ample water availability (Brümmer et al., 2012; Lafleur et al., 2003). 37  J F M A M J J A S O N D00.050.10.150.20.250.30.350.40.450.5 Figure 18: Mean monthly McNaughton and Jarvis decoupling coefficient (Ω) at the site during the study year. The solid line indicates the annual mean Ω of 0.25.   3.6 Greenhouse gas exchange 3.6.1 Temporal variation in turbulent exchanges The conventionally managed field had several management operations of grass interrow mowing, blueberry pruning, and NH4NO3 fertilization. Management of the field did not have a discernible effect on 𝐻 and 𝜆𝐸 or 𝛽 (Figure 19) although mowing of the interrow was expected to increase 𝐻 and 𝛽 and decrease 𝜆𝐸 as the leaf area for transpiration would be reduced, partitioning more available energy to 𝐻. All of these events had a large, positive effect on 𝐹C and generally reduced the sink strength of the field. For several days after mowing the interrow, the field had net positive 𝐹C occurrences (Figure 19). As the field was approximately 60% blueberry row and 40% interrow, mowing the interrow grass reduced a large portion of the photosynthetic capacity of the field, especially early in the season when the blueberry plants had not reached peak growth. Also, the grass cuttings left on the field eventually decomposed and increased 𝐹C. 38  The effect of mowing the field and increased 𝐹𝐶 was less pronounced later in the season, suggesting increased photosynthesis in the blueberry plants was able to offset reduced interrow photosynthesis (Figure 19). Like our study, Jassal et al. (2011) found increased 𝑅s following N-fertilization, which may have further increased 𝐹𝐶. However, as N-fertilization has been found to reduce soil microbial respiration (Ramirez et al., 2010), increased 𝐹C emissions following fertilization were likely due to increased autotrophic and root respiration but cannot be confirmed without root exclusion plots. Increased 𝐹N2O was observed following initial N-fertilization, but successive fertilization events did not result in a similar increase in 𝐹N2O. 𝐹N2O declined over the growing season after the initial fertilization event, likely due to increased plant uptake of N (Figure 19). The effect of increased 𝐹N2O was difficult to discern from increased 𝑇𝑠 as both events occurred relatively simultaneously. 𝐹N2O remained elevated until the end of the growing season and declined in August, returning to pre-fertilization values in early September (Figure 19). Emissions were not extremely sporadic compared to other studies; the months of May and June post-fertilization and September and October following sudden precipitation accounted for approximately 30% and 24% of the annual 𝐹N2O, respectively. In comparison to a study that contained disproportionally large 𝐹N2O in a short time period, Zona et al. (2013) measured over 52% of the entire 𝐹N2O in one week over the one-and-a-half-year study period. The magnitude of the initial 𝐹N2O pulse may have been limited as the N-fertilization was implemented successively over four separate applications as opposed to one large event. Management of the field did not have a discernible or consistent effect on the magnitude of 𝐹CH4 (Figure 19), which was sporadic over the entire year but was consistently a source, suggesting the natural variability in 𝐹CH4 had a larger effect than any observed management practice. 39  J F M A M J J A S O N D0  -10-50510CH4(nmol m-2 s-1)-202N2O(nmol m-2 s-1)-6-3036CO2( mol m-2 s-1)-50050100150H and E(W m-2)PR F1M1F2M2F3F4M3M4 PRM5 PRNet Ecosystem Exchange Figure 19: Mean daily fluxes of sensible heat (𝐻;  dotted line), latent heat (𝜆𝐸; solid line), carbon dioxide (CO2), nitrous oxide (N2O) and methane (CH4) at the study during the study year. Positive (negative) values indicate the field was a source (sink). Vertical dashed lines indicate pruning (PR), N-fertilization via four surface applications of ~25 – 30 kg N ha−1 of ammonium nitrate (F1 − F4) and interrow mowing (M1 −M5).  3.6.1.1 Diurnal trends in CO2 exchange 𝐹𝑐 tends to exhibit a strong diurnal signal with negative (sink) values during daytime due to higher GPP and positive (source) values during in the nighttime due to higher 𝑅e (Nemitz et al., 2018). The field was actively photosynthesizing throughout the year, with a clear diurnal signal in all months despite senescence and dormancy of the blueberry crop in the fall and winter months, suggesting that the perennial grass-interrow was active year-round (Figure 20). The mean sink strength of the field increased due to increased daylength and ↓ 𝑄, with 𝐹c decreasing from −1.4 in January to −12.2 μmol m−2 s−1 in July during the daytime then gradually declined whereas nighttime 𝐹c increased relatively less from 1.4 in January to 6.7 μmol m−2 s−1 in September (Figure 21). The half-hourly variability of 𝐹c was higher during the daytime compared 40  to nighttime values, which was due to plant physiological responses to rapidly changing environmental conditions including ↓ 𝑄 and 𝐷 compared to nighttime.  -15-10-50510January February March April May June0 12 24-15-10-50510July0 12 24August0 12 24September0 12 24October0 12 24November0 12 24DecemberFc (mol m-2 s-1)Local time (PST)  Figure 20: Monthly mean diurnal carbon dioxide flux (𝐹c) averaged for each half-hour at the site during the study year. Positive (negative) values indicate that the field was a source (sink) for carbon. The gray shading represents ± one standard deviation of 𝐹c for each half-hour.  Figure 21: Fingerprint plot of ensemble diurnal carbon dioxide flux (𝐹c) for each month at the site during the study year. Negative (positive) values indicate that the field was a sink (source) for CO2.  J F M A M J J A S O N D6121824   Local Time (PST)Fc (mol m-2 s-1)-15-10-5051041  3.6.1.2 Diurnal trends in N2O exchange The field was a consistent source of N2O year-round, with increased 𝐹N2O following N-fertilization and fall precipitation. Low, continuous 𝐹N2Obetween 0 and 0.5 nmol m−2 s−1was observed during the winter and early spring but was elevated to approximately 1 nmol m−2 s−1 following N-fertilization and fall precipitation. Variability in 𝐹N2O also increased in fall months following precipitation, suggesting that increased 𝐹N2O was relatied to higher 𝜃 (Figure 22). While studies estimating annual 𝐹N2O from agricultural systems are generally limited to eastern Canada, these studies have emphasized the disproportionate impact and importance of spring thaw events, contributing up to 40% of the annual 𝐹N2O, following snowmelt due to increased 𝜃 and increased available N from reduced plant activity (Desjardins et al., 2010; Smith et al., 2004). Mild winter temperatures and continous precipitation through winter and spring (Figure 1) in this study did not support high 𝐹N2O in the spring, but similar processes of suddenly increased 𝜃 can explain high 𝐹N2O in the fall after prolonged drying during growing season (Butterbach-Bahl et al., 2013). Compared to 𝐹c, most months did not exhibit a clear diurnal signal in 𝐹N2O. The months of April to June exhibited a weak diurnal signal in 𝐹N2O, with daytime fluxes on average higher by a factor of ~1.2 to 2.3 compared to fluxes during the night (Figure 23). In a study of diurnal variations of 𝐹N2O over an agricultural field, Shurpali et al. (2016) found that given sufficient 𝜃 and evenly distributed 𝑃, 𝐹N2O exhibited a diurnal signal following N-fertilization, which has implications for studies relying on chamber measurements typically performed during the daytime. Unlike 𝐹c, an annual exponential relationship between 𝐹N2O and 𝑇s could not be derived, and 𝐹N2O was not found to be directly related to temperature annually despite a clear diurnal signal in several months following fertilization in April, 42  suggesting that factors other than 𝑇s for the rest of the year were more important in regulating 𝐹N2O annually.   -1.5-1-0.500.511.5January February March April May June0 12 24-1.5-1-0.500.511.5July0 12 24August0 12 24September0 12 24October0 12 24November0 12 24DecemberFN2O (nmol m-2 s-1)Local time (PST)  Figure 22: Monthly mean diurnal nitrous oxide flux (𝐹N2O) for each month at the site during the study year. The crop was fertilized with ~110 kg N ha−1 of ammonium nitrate over four equal applications between late-April and early-July. The gray area indicates ± one standard deviation of 𝐹N2O for each half-hour. 43  J F M A M J J A S O N D6121824   Local Time (PST)FN2O (nmol m-2 s-1)-3-2-10123 Figure 23: Fingerprint plot of ensemble diurnal nitrous oxide flux (𝐹N2O) for each month at the site during the study year. Negative (positive) values indicate that the field was a sink (source) for N2O. The crop was fertilized with ~110 kg N ha−1 of ammonium nitrate over four equal applications between late-April and early-July.  3.6.1.3 Diurnal trends in CH4 exchange Annually methane fluxes were sporadic but centred around zero with no clear seasonality. There was no clear diurnal signal and the variance was relatively consistent throughout the year (Figure 24). The results show that the field was a consistent source of 𝐹CH4 at approximately the same time of the day for several days, possibly due to filling data gaps with MDV (Figure 25).  As the field included grass-interrows that accounted for approximately 40% of the field area and the field was well-drained, site specific conditions were expected to have promoted at least weak 𝐹CH4 uptake. Methanotrophic bacteria can oxidize CH4 as a sole C and energy source using the enzyme methane monooxygenase (MMO), and generally consume most of the CH4 produced by methanogens deeper in the soil in anaerobic environments (Jassal et al., 2011). As 44  MMO has a lack of substrate specificity, a wide range of reactions can be catalyzed including the oxidation of ammonium (NH4+) to nitrite (NO2−) (Bédard and Knowles, 1989; Hanson and Hanson, 1996). While even obligate methanotrophs can oxidize NH4+, the addition of N-fertilizer in the form of NH4NO3 could suppress the uptake of CH4, as methanotrophs with MMO will oxidize NH4+ detrimentally as the substrate cannot be used as a source of C or energy for methanotrophs and will slow growth (Bédard and Knowles, 1989; Hanson and Hanson, 1996).  -200204060January February March April May June0 12 24-200204060July0 12 24August0 12 24September0 12 24October0 12 24November0 12 24DecemberFCH4 (nmol m-2 s-1)Local time (PST)  Figure 24: Monthly mean diurnal methane flux (𝐹CH4) for each month at the site during the study year. The fluxes were sporadic and did not exhibit a clear diurnal signal. The gray area indicates ± one standard deviation of 𝐹CH4 for each half-hour. 45  J F M A M J J A S O N D6121824   Local Time (PST)FCH4 (nmol m-2 s-1)-600-400-2000200400600 Figure 25: Fingerprint plot of ensemble diurnal methane flux (𝐹CH4) for each month at the site during the study year. Negative (positive) values indicate that the field was a sink (source) for CH4.  3.6.2 Annual greenhouse gas budget The field was a source for all GHGs annually and emitted 173 g CO2 − C m−2 y−1,  3.6 kg N2O − N ha−1 y−1 (0.36 g N2O − N m−2 y−1) and 0.71 g CH4 − C m−2 y−1 (Figure 26). The EF estimated from EC-measurements was 2.4% of applied N compared to 1.25% considered by the IPCC, if background N emissions were ~1.0 kg N ha−1 y−1. While N2OTierI predicted by the IPCC Tier I default EF and N2OTierII proposed by Rochette et al. (2018) were similar at 2.4 and 2.6 kg N2O − N ha−1 y−1, respectively, they were both lower than the measured emission by about 1.0 kg N2O − N ha−1 y−1. Both estimates would have more accurately predicted the annual 𝐹N2O if only the initial growing season pulse were considered, suggesting that accounting for events after the growing season including the distribution of annual 𝑃 or increased 𝜃 after prolonged dry periods may be important in estimating annual 𝐹N2O. Predicted N2O emissions at 46  this site are comparable to chamber measurements of N2O emissions in BC from bare soils following dairy slurry application, although the study recognized that emissions could have been underestimated due to interpolation and the length of the sampling period (Bhandral et al., 2008). When the emissions of CO2, N2O and CH4 were considered in CO2e using GWPs on a 100-year timescale, the field was a net source of all GHGs and emitted 838 g CO2e m−2 year−1 (Table 4). As the field was already a net source of 637 g CO2 m−2 year−1, including the emissions of N2O and CH4 as CO2e further increased the annual emission of the field by 168 and 33 g CO2e m−2 year−1, respectively (Table 4). CO2 contributed the largest proportion of GHG emissions at 76%, while N2O and CH4 contributed 20% and 4%, respectively, suggesting that, while N2O and CH4 are more potent GHGs, potential mitigation efforts at this field should be focused on CO2 emissions. J F M A M J J A S O N D-0.500.51g CH4 – C m-2024kg N2O – N ha-1-1000100200g CO2 – C m-2 Figure 26: Cumulative greenhouse gas emissions of carbon dioxide (CO2 − C), nitrous oxide (N2O − N) and methane (CH4 − C) at the site during the study year. The dashed line indicates the Intergovernmental Panel on Climate Change Tier 1 default emission factor which predicted an annual emission of 2.4 kg N2O − N ha−1 y−1 based on the nitrogen fertilization rate of 110 N ha−1 y−1. The dotted line indicates the Canada-specific estimate of annual N2O emissions proposed by Rochette et al. (2018) at 2.6 kg N2O − N ha−1 y−1. 47  Table 4: Monthly greenhouse gas emissions of carbon dioxide (CO2), nitrous oxide (N2O) and methane (CH4) measured in carbon dioxide equivalents (CO2e) using global warming potentials on a 100-year timescale considering climate-carbon feedbacks in g CO2e m−2 month−1. Month CO2 N2O − CO2e CH4 − CO2e Total CO2e January 138 3 5 146 February 95 5 2 102 March 43 5 1 49 April -35 7 4 -24 May -24 29 4 9 June -47 22 1 -24 July -251 15 0 -236 August -75 11 2 -62 September 153 18 2 173 October 237 23 6 266 November 224 19 1 244 December 179 11 5 195 Annual (g CO2e m−2 year−1) 637 168 33 838  3.6.3 NEE partitioning 𝐹c was partitioned into GPP and 𝑅e to determine the environmental controls on 𝑅e and compare the relative importance of each component. The cumulative GPP, 𝑅e and NEE during the study period were 1287, 1460, and 173 g C m−2 year−1 (Table 5). The slope of the linear regression of partitioned 𝑅e − GPP vs. 𝐹c was 0.98 (R2 = 0.93, n = 17520) suggesting that the partitioning was robust. In comparison, a nearby unmanaged grassland on Westham Island had 𝑅e and GPP values of 1215 and 1438 g C m−2 year−1, respectively, resulting in a C sink of −222 g C m−2 year−1 (Lee et al., 2017).  During blueberry dormancy, the grass-interrows still contributed to GPP values of approximately 1 μmol m−2 s−1, but the field remained a source due to higher 𝑅e (Figure 27). While plant dormancy and limited radiation reduced GPP, mild overwinter temperatures and 48  adequate soil moisture were likely favourable for continuous microbial activity and consistent 𝑅e (Falge et al., 2002). In May, GPP increased at a faster rate than 𝑅e, which resulted in several months of net C sequestration (negative NEE) between April and August. Mean daily 𝑅e reached 6.45 μmol m−2 s−1 in June, declined to a minimum of 4.56 μmol m−2 s−1 in late July, and rose finally to 6.20 μmol m−2 s−1 in late September, while GPP displayed typical annual seasonality and continually increased until July at 8.70 μmol m−2 s−1 and decreased steadily back to overwinter values (Figure 27). The field was the strongest C sink in July, reaching mean daily measured NEP of 4.17 μmol m−2 s−1 when 𝑅e was surpressed and GPP was maximal, which resulted in sequestering 71 g C m−2 month−1 (Figure 27 and Table 5).  While limitations in 𝜃 can reduce microbial activity and reduce 𝑅e (Falge et al., 2002; Gaumont-Guay et al., 2006) minimum 𝑅e in July did not coincide with minimum 𝜃, 𝜑 or WT  (Figure 2) when 𝑅e was increased. Progression towards blueberry dormancy following harvest in the fall may have reduced the water demand, allowing enough soil moisture to sustain high 𝑅e. GPP then decreased steadily while 𝑅e remained relatively high compared to GPP for the rest of the year, with the field losing C in the months of September and October at 65 and 66 g C m−2 month−1 (Table 5). Annually, the ratio of GPP to 𝑅e was only 0.88 (Table 5).   49  J F M A M J J A S O N D-6-4-20246810Fc Components (mol m-2 s-1)ReGPPNEP Figure 27: CO2 flux partitioning, using standard Fluxnet-Canada protocols, into mean daily ecosystem respiration (𝑅e), gross primary production (GPP) and net ecosystem production (NEP) at the site during the study year. The sign convention is such that GPP − Re =  NEP, where a positive (negative) NEP indicates a carbon sink (source).  Table 5: Monthly CO2 flux partitioning into total monthly ecosystem respiration (𝑅𝑒), gross primary production (GPP) and net ecosystem production (NEE) at the site Month GPP  (g C m−2 month−1) Re (g C m−2 month−1) NEE  (g C m−2 month−1) January 15 52 37 February 27 48 21 March 63 78 15 April 120 114 -6 May 183 172 -11 June 195 186 -8 July 235 164 -71 August 194 168 -26 September 131 195 65 October 79 145 66 November 32 85 54 December 14 52 39 Annual (g C m−2 year−1) 1287  1460 173  50  3.6.4 Annual carbon budget  The annual C budget of the field was determined by summing NEE and C imports and exports. The values of input of 𝐶sawdust and export of 𝐶yield were 488 and 84 g C m−2 year−1, respectively, resulting in net carbon sequestration of 231 g C m−2 year−1. As the C budget was found to be sensitive to the input of 𝐶sawdust, see Appendix F for a range of possible estimates for inputs and the implications for the annual C budget. In a study characterizing long-term storage of C in highbush blueberry plants, Nemeth et al. (2017) found that the amount of C exported in fruit yield was greater than 100 g C m−2 year−1, which was higher than the estimated 𝐶yield at this site, suggesting that the estimated C export was relatively conservative and could potentially account for more C loss. Forcing EBC and assuming similarity between measurements of 𝜆𝐸 and 𝐹c would further increase the C lost to the atmosphere by an additional 40 g C m−2 year−1.  3.6.5 Environmental controls on greenhouse gas emissions 3.6.5.1 Environmental controls on CO2 emissions Nighttime 𝑅e and 𝑇s was compared using three different relationships. Modeled 𝑅e values using the exponential temperature-respiration relationship 𝑅e = 𝑅10𝑄10[(𝑇s−10)/10] showed relatively poor agreement (𝑅2 = 0.57, RMSE = 2.12 μmol m−2 s−1, n = 8777) with measured 𝑅e (Figure 28). Modeled 𝑅e showed good agreement with measured 𝑅e at lower values of 𝑅e until modeled 𝑅e values reached approximately 7 μmol m−2 s−1, after which the relationship tended to overestimate modeled 𝑅e, suggesting that the simple temperature-respiration relationship may not accurately characterize 𝑅e with continually increasing 𝑇s and required a constraint (Figure 29). The dependence of 𝑅s on 𝑇s has been shown to decrease when 51  𝜃 is limiting during drought conditions (Gaumont-Guay et al., 2006), and 𝑅eN was shown to be weakly related (𝑅2 = 0.24, RMSE = 0.23, n = 8777) with 𝜃 (Figure 29). Including the additional term (𝑎 + 𝑏𝜃 +𝑐𝜃) increased the agreement between measured and modeled 𝑅e (𝑅2 =0.74, RMSE = 0.95 μmol m−2 s−1, n = 8777), suggesting that the relationship between 𝑇s and 𝑅e can be constrained by both low and high 𝜃 (Figure 28) which can reduce soil microbial activity and decomposition (Gaumont-Guay et al., 2006; Lee et al., 2017). The relationship between 𝑇s and 𝑅e was also described with a logistic relationship (Figure 28), which generally showed similar but slightly higher agreement between measured and modeled 𝑅e (𝑅2 = 0.75, RMSE = 0.93 μmol m−2 s−1, n = 8777). While the logistic model was more successful in predicting 𝑅e and was more parsimonious, the exponential-temperature relationship with additional moisture sensitivity constrained using physical variables as opposed to empirically fit parameters. Model parameters for the three relationships are in Tables 6 to 8.   52  0 5 10 15051015Measured Re (mol m-2 s-1)a)R2 = 0.57RMSE = 2.12 mol m-2 s-1n = 87770 5 10 15Modeled Re (mol m-2 s-1)b)R2 = 0.74RMSE = 0.95 mol m-2 s-1n = 87770 5 10 15c)R2 = 0.75RMSE = 0.93 mol m-2 s-1n = 8777 Figure 28: Relationships between half-hourly observed nighttime ecosystem respiration (𝑅e) and modeled 𝑅e using a) an exponential temperature-respiration relationship 𝑅e = 𝑅10𝑄10[(𝑇s−10)/10] excluding measurements of 𝑅e when 𝜃 < 0.40, b) an exponential temperature-moisture relationship of 𝑅e with an additional term describing the sensitivity of 𝑅e to 𝜃 as 𝑅e =(𝑎 + 𝑏𝜃 +𝑐𝜃)𝑅10𝑄10[(𝑇𝑠−10)/10] and c) a logistic temperature relationship 𝑅e =𝑟11+𝑒[𝑟2(𝑟3−𝑇𝑠)] where 𝑇s is the 5-cm depth soil temperature, 𝜃 is the volumetric water content 𝑅10 and 𝑄10 are derived from log-transformed temperature-respiration relationships and 𝑎, 𝑏, 𝑐, 𝑟1, 𝑟2 and 𝑟3 are empirical coefficients. Coefficients and model parameters are in Tables 6 to 8.    53  0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65-1.5-1-0.500.511.522.5RsNR2 = 0.24RMSE = 0.23n = 8777 Figure 29: Relationship between nighttime (downwelling shortwave radiation = 0) half-hourly temperature-normalized (ReN) ecosystem respiration (𝑅e) and 5-cm depth soil water content (𝜃).  ReN is expressed as the ratio of observed nighttime 𝑅e to 𝑅e estimated from an exponential temperature-respiration relationship.   Table 6: Model parameters for the exponential relationship (Humphreys et al., 2005) between nighttime ecosystem respiration (𝑅e) and the 5-cm soil temperature (𝑇𝑠) when the 5-cm volumetric water content (𝜃) > 0.4 during the study period. Function 𝐴 𝐵 𝑅10 (μmol m−2 s−1) 𝑄10 𝑅2 RMSE  (μmol m−2 s−1) 𝑛 Exponential 0.03 0.12 3.4 3.3 0.57 2.12 8777  Table 7: Model parameters for the exponential-temperature relationship with a hyperbolic soil moisture relationship (Gaumont-Guay et al., 2006) between nighttime ecosystem respiration (𝑅e) and the 5-cm soil temperature (𝑇s) during the study period. 𝑎 𝑏 𝑐 𝐴 𝐵 𝑅10 (μmol m−2 s−1)  𝑄10 𝑅2 RMSE  (μmol m−2 s−1) 𝑛 8.9   −10.0    −1.5 0.28 0.08    3.3 2.3 0.74 0.95 8777  54  Table 8: Model parameters for the logistic relationship (Barr et al., 2004) between nighttime ecosystem respiration (𝑅e) and the 5-cm depth soil temperature (𝑇𝑠) during the study period. 𝑟1 (μmol m−2 s−1) 𝑟2 (℃−1) 𝑟3 (℃) 𝑅2  𝑅𝑀𝑆𝐸 (μmol m−2 s−1) 𝑛 6.12 0.29 8.31 0.75  0.93  8777  3.6.5.2 Environmental controls on N2O emissions An exponential relationship with half-hourly 𝑇s and 𝐹N2O was weak (𝑅2 = 0.09, RMSE =0.23 nmol m−2 s−1, n = 17520), suggesting that 𝐹N2O was relatively insensitive to changes in 𝑇s despite evidence of diurnal trends in 𝐹N2O (Figure 30). Denitrification has been shown to be sensitive to temperature with higher 𝑄10 compared to relationships between temperature and CO2 (Butterbach-Bahl et al., 2013). Field management has also been shown to have a large impact on 𝐹N2O, which could reduce the sensitivity of 𝐹N2O to 𝑇s (Rochette et al., 2018; Wagner-Riddle et al., 2007). The 𝑄10 derived in this study for 𝐹N2O was 0.71 (Table 9), which is much lower than the 𝑄10 values for 𝐹c of 2.3 to 3.3 (Tables 6 and 7). The diurnal trends in 𝐹N2O could be instead explained by increased plant activity and root exudates during the daytime which could provide increased C to denitrifying microbial communities (Bouwman, 1996; Butterbach-Bahl et al., 2013). In comparison to 𝑇s, the Gaussian relationship between half-hourly 𝜃 and 𝐹N2O explained more variability (𝑅2 = 0.26, RMSE = 0.36 nmol m−2 s−1, n = 17520) but was still relatively weak, particularly due to higher variability of 𝐹N2O above 𝜃 of 0.45 (Figure 31). 𝐹N2O was maximal when 𝜃 was 0.38 (Table 9), which would represent a WFPS of approximately 60%, and the results were comparable to other studies (Balaine et al., 2013; Butterbach-Bahl et al., 2013) As 𝜃 was only a proxy for the aerobic status of the soil, the relationship could be improved by 55  considering the actual relative gas diffusivity for oxygen into the soil which would better characterize the oxygen availability (Balaine et al., 2013).  0 5 10 15 20 25-0.500.511.5R2 = 0.08RMSE = 0.33 nmol m-2 s-1n = 17520Ts ( C)FN2O (nmol m-2 s-1) Figure 30: Exponential relationship between 5-cm depth soil temperature (𝑇𝑠) and nitrous oxide flux (𝐹N2O) at the site during the study year. 𝐹N2O was binned into 1℃ increments, with the error bars indicating the standard deviation of the temperature bin. Model parameters are in Table 9. 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0.7-4-3-2-101234FN2O (nmol m-2 s-1)R2 = 0.26RMSE = 0.36 nmol m-2 s-1n = 17520 Figure 31: Gaussian relationship between 5-cm depth volumetric water content (𝜃) and nitrous oxide flux (𝐹N2O) at the site during the study period. 𝐹N2O was binned into 0.05 increments of 𝜃, with the error bars indicating the standard deviation of each 𝜃 bin. Model parameters are in Table 9. 56  Table 9: Model parameters for the exponential-temperature and Gaussian-moisture relationships between nitrous oxide flux (𝐹N2O) and the 5-cm depth soil temperature (𝑇s) and volumetric water content (𝜃) at the site during the study year. 𝑇s 𝑅10 (nmol mol−1s−1)  𝑄10  𝑅2 RMSE (nmol mol−1s−1) n  2.04 0.71  0.09 0.23 17520 𝜃 𝐴 𝐵 𝐶  𝑅2 RMSE  (nmol mol−1s−1) n  0.87 0.38 0.10 0.26 0.36 17520  3.6.5.3 Environmental controls on CH4 emissions In general, half-hourly measurements of 𝐹CH4 could not be described with environmental variables due to 𝐹CH4 being close to zero. An exponential-temperature relationship between 𝐹CH4 and 𝑇s explained almost zero variability in 𝐹CH4 (𝑅2 = 0.001, RMSE = 0.0178 nmol m−2 s−1, n = 17520). While Lee et al. (2017) were able to determine an exponential relationship between 𝑇s and 𝐹CH4, the magnitude of the fluxes at their bog site was 5-80 times higher than 𝐹CH4 at this site. Variables related to soil moisture including 𝜃, WT depth or 𝜑 which are proxies for aeration were not found to be linearly related to 𝐹CH4 and using stepwise multiple linear regression suggested that relationships between 𝐹CH4 and soil moisture and could not be described. 𝐹CH4 has been found to be related to WT depth (Moore and Knowles, 1989), soil moisture (Kaiser et al., 2018) and 𝜑 (Ball, 2013), with generally increasing 𝐹CH4 with increased anaerobic conditions.  57  Chapter 4: Conclusions 4.1 Summary of key findings One of the main objectives of this study was to measure continuous fluxes of CO2, N2O and CH4 using an EC system over a conventionally managed highbush blueberry field in Delta, BC and determine its global warming potential. In 2018, the site was a net source for all three measured GHGs, and overall emitted 838 g CO2e m−2 year−1. Despite year-round photosynthesis, the site was a net source of CO2 and lost 637 g CO2 m−2 year−1 (173 g C m−2 year−1), which made the largest contribution to the overall GHG balance. While N2O and CH4 are more potent GHGs than CO2, low emissions of both GHGs resulted in relatively low contributions to the GHG balance. In addition to the GHG balance, this study determined that: • Management of the field effected GHG emissions but the effects varied seasonally. In general, mowing the grass-interrow decreased the C-sink strength of the field and increased N2O emissions were associated with N-fertilization events.  • In addition to temperature, soil moisture was an important variable affecting both N2O and CO2 emissions. Higher N2O emissions were associated with the sudden onset of precipitation following prolonged drying during the growing season. Exponential relationships between temperature and nighttime CO2 emissions were improved by an additional soil moisture term. This has implications in terms of future climate change as the inter-annual and intra-annual distribution of precipitation will likely be modified.  58  • After accounting for imports and exports, the field sequestered C when measurements were presented on an annual basis but depended largely on a continuous application of sawdust mulch to offset C loss and export.  • The field had comparable WUE and iWUE to other vegetated surfaces but differed in that the surface had much lower 𝜆𝐸 relative to 𝜆𝐸eq.  4.2 Implications for future research While the study was successful in determining annual GHG emissions, the use of more complex gap-filling strategies including artificial neural networks (Knox et al., 2015) or look-up tables (Mishurov and Kiely, 2011) was not considered, which may improve annual estimates of non-CO2 GHGs. Error in measurements of NEE could be characterized by simulating data gaps to better characterize the magnitude of C loss or gain (Hollinger and Richardson, 2005). The blueberry field was also a heterogeneous surface which included patches of bare soil, sawdust mulch and grass, and the use of chambers to measure fluxes on the various surfaces could inform management strategies in terms of the row widths and spacing or mowing intensity to reduce GHG emissions by modifying the proportion of each surface. As the sawdust mulch was found to be a significant portion of the overall C balance, root exclusion plots and chamber measurements over mulched and un-mulched surfaces could inform strategies to mitigate 𝑅e. Lysimeters on bare, mulched, and vegetated surfaces combined with measurements of stomatal conductance would help characterize the water retention properties of the sawdust mulch and help parameterize estimates of 𝑔𝑠. 59  Bibliography Agriculture and Agri-Food Canada, 2012. Statistical Overview of the Canadian Blueberry Industry, 2010. AgriService BC, 2018. Fast Stats 2016: British Columbia’s Agrifood and Seafood Sector. Ai, Z., Yang, Y., 2016. Modification and Validation of Priestley – Taylor Model for Estimating Cotton Evapotranspiration under Plastic Mulch Condition. J. Hydrometeorol. 17, 1281–1293. Aubinet, M., Grelle, A., Ibrom, A., Rannik, Ü., Moncrieff, J., Foken, T., Kowalski, A.S., Martin, P.H., Berbigier, P., Bernhofer, C., Clement, R., Elbers, J., Granier, A., Grünwald, T., Morgenstern, K., Pilegaard, K., Rebmann, C., Snijders, W., Valentini, R., Vesala, T., 2000. Estimates of the Annual Net Carbon and Water Exchange of Forests: The EUROFLUX Methodology. Adv. Ecol. Res. 30, 113–175. Baker, J.M., Griffis, T.J., 2005. Examining strategies to improve the carbon balance of corn/soybean agriculture using eddy covariance and mass balance techniques. Agric. For. Meteorol. 128, 163–177. Balaine, N., Clough, T.J., Beare, M.H., Thomas, S.M., Meenken, E.D., Ross, J.G., 2013. Changes in Relative Gas Diffusivity Explain Soil Nitrous Oxide Flux Dynamics. Soil Sci. Soc. Am. J. 77, 1496. Baldocchi, D., 2003. Assessing the eddy covariance technique for evaluating carbon dioxide exchange rates of ecosystems: past, present and future. Glob. Chang. Biol. 9, 479–492. Baldocchi, D., 1997. Flux Footprints within and over Forest Canopies. Boundary-Layer Meteorol. 85, 273–292. Ball, B.C., 2013. Soil structure and greenhouse gas emissions : a synthesis of 20 years of 60  experimentation. Eur. J. Soil Sci. 64, 357–373. Barr, A.G., Black, T.A., Hogg, E.H., Kljun, N., Morgenstern, K., Nesic, Z., 2004. Inter-annual variability in the leaf area index of a boreal aspen-hazelnut forest in relation to net ecosystem production. Agric. For. Meteorol. 126, 237–255. Bédard, C., Knowles, R., 1989. Physiology, biochemistry, and specific inhibitors of CH4, NH4+, and CO oxidation by methanotrophs and Nitrifiers. Microbiol. Rev. 53, 68–84. Beer, C., Ciais, P., Reichstein, M., Baldocchi, D., Law, B.E., Papale, D., Soussana, J.-F., Ammann, C., Buchmann, N., Frank, D., Gianelle, D., Janssens, I.A., Knohl, A., Kostner, B., Moors, E., Roupsard, O., Verbeeck, H., Vesala, T., Williams, C.A., Wohlfahrt, G., 2009. Temporal and among-site variability of inherent water use efficiency at the ecosystem level. Global Biogeochem. Cycles 23, 1–13. Bhandral, R., Bittman, S., Kowalenko, C.G., Friesen, A., Hunt, D.E., 2008. Emissions of nitrous oxide after application of dairy slurry on bare soil and perennial grass in a maritime climate. Can. J. Soil Sci. 88, 517–527. Bouwman, A.F., 1996. Direct emission of nitrous oxide from agricultural soils. Nutr. Cycl. Agroecosystems 46, 53–70. Brümmer, C., Black, T.A., Jassal, R.S., Grant, N.J., Spittlehouse, D.L., Chen, B., Nesic, Z., Amiro, B.D., Arain, M.A., Barr, A.G., Bourque, C.P., Coursolle, C., Dunn, A.L., Flanagan, L.B., Humphreys, E.R., Lafleur, P.M., Margolis, H.A., Mccaughey, J.H., Wofsy, S.C., 2012. How climate and vegetation type influence evapotranspiration and water use efficiency in Canadian forest , peatland and grassland ecosystems. Agric. For. Meteorol. 153, 14–30. Butterbach-Bahl, K., Baggs, E.M., Dannenmann, M., Kiese, R., Zechmeister-Boltenstern, S., 61  2013. Nitrous oxide emissions from soils: how well do we understand the processes and their controls? Philos. Trans. R. Soc. Lond. B. Biol. Sci. 368, 20130122. Campbell, G.S., Norman, J.M., 1998. An Introduction to Environmental Biophysics, 2nd ed. Springer, New York, USA. Desjardins, R.L., Pattey, E., Smith, W.N., Worth, D., Grant, B., Srinivasan, R., MacPherson, J.I., Mauder, M., 2010. Multiscale estimates of N2O emissions from agricultural lands. Agric. For. Meteorol. 150, 817–824. Dutaur, L., Verchot, L. V, 2007. A global inventory of the soil CH4 sink. Global Biogeochem. Cycles 21, 1–9. Falge, E., Baldocchi, D., Olson, R., Anthoni, P., Aubinet, M., Bernhofer, C., Burba, G., Ceulemans, R., Clement, R., Dolman, H., Granier, A., Gross, P., Grünwald, T., Hollinger, D., Jensen, N., Katul, G., Keronen, P., Kowalski, A., Ta, C., Law, B.E., Meyers, T., Moncrieff, J., Moors, E., Munger, J.W., Pilegaard, K., Rannik, Ü., Rebmann, C., Suyker, A., Tenhunen, J., Tu, K., Verma, S., Vesala, T., Wilson, K., Wofsy, S., 2001. Gap filling strategies for defensible annual sums of net ecosystem exchange. Agric. For. Meteorol. 107, 43–69. Falge, E., Baldocchi, D., Tenhunen, J., Aubinet, M., Bakwin, P., Berbigier, P., Bernhofer, C., Burba, G., Clement, R., Davis, K.J., Elbers, J.A., Goldstein, A.H., Grelle, A., Granier, A., Guomundsson, J., Hollinger, D., Kowalski, A.S., Katul, G., Law, B.E., Malhi, Y., Meyers, T., Monson, R.K., Munger, J.W., Oechel, W., Paw, U., Pilegaard, K., Rannik, Ü., Rebmann, C., Suyker, A., Valentini, R., Wilson, K., Wofsy, S., 2002. Seasonality of ecosystem respiration and gross primary production as derived from FLUXNET measurements. Agric. For. Meteorol. 113, 53–74. 62  Finnigan, J.J., Clement, R., Malhi, Y., Leuning, R., Cleugh, H.A., 2003. Re-evaluation of long-term flux measurement techniques. Part I: Averaging and coordinate rotation. Bound. Layer Meteorol. 107, 1–48. Flemming, C., Liang, C., MacDonald, D., 2016. Agriculture (CRF Sector 3), in: National Inventory Report 1990–2016: Greenhouse Gas Sources and Sinks in Canada. Foken, T., 2008. The Energy Balance Closure Problem : An Overview. Ecol. Appl. 18, 1351–1367. Foken, T., 2006. 50 years of the Monin-Obukhov similarity theory. Boundary-Layer Meteorol. 119, 431–447. Gash, J.H.C., Valente, F., David, J.S., 1999. Estimates and measurements of evaporation from wet, sparse pine forest in Portugal. Agric. For. Meteorol. 94, 149–158. Gaumont-Guay, D., Black, T.A., Griffis, T.J., Barr, A.G., Jassal, R.S., Nesic, Z., 2006. Interpreting the dependence of soil respiration on soil temperature and water content in a boreal aspen stand. Agric. For. Meteorol. 140, 220–235. Hanson, R.S., Hanson, T.E., 1996. Methanotrophic Bacteria. Microbiol. Rev. 60, 439–471. Hayden, K.L., Anlauf, K.G., Hoff, R.M., Strapp, J.W., Bottenheim, J.W., Wiebe, H.A., Froude, F.A., Martin, J.B., Steyn, D.G., McKendry, I.G., 1997. The Vertical Chemical and Meteorological Structure of the Boundary Layer in the Lower Fraser Valley during Pacific ’93. Atmos. Environ. 3, 2089–2105. Heidbach, K., Schmid, H.P., Mauder, M., 2017. Experimental evaluation of flux footprint models. Agric. For. Meteorol. 246, 142–153. Hollinger, D.Y., Richardson, A.D., 2005. Uncertainty in eddy covariance measurements and its application to physiological models. Tree Physiol. 25, 873–885. 63  Humphreys, E.R., Black, T.A., Ethier, G.J., Drewitt, G.B., Spittlehouse, D.L., Jork, E.M., Nesic, Z., Livingston, N.J., 2003. Annual and seasonal variability of sensible and latent heat fluxes above a coastal Douglas-fir forest, British Columbia, Canada. Agric. For. Meteorol. 115, 109–125. Humphreys, E.R., Black, T.A., Morgenstern, K., Li, Z., Nesic, Z., 2005. Net ecosystem production of a Douglas-fir stand for 3 years following clearcut harvesting. Glob. Chang. Biol. 11, 450–464. Hutchinson, J.J., Campbell, C.A., Desjardins, R.L., 2007. Some perspectives on carbon sequestration in agriculture. Agric. For. Meteorol. 142, 288–302. Jassal, R.S., Black, T.A., Roy, R., Ethier, G., 2011. Effect of nitrogen fertilization on soil CH4 and N2O fluxes, and soil and bole respiration. Geoderma 162, 182–186. Jassal, R.S., Black, T.A., Spittlehouse, D.L., Bru, C., 2009. Evapotranspiration and water use efficiency in different-aged Pacific Northwest Douglas-fir stands. Agric. For. Meteorol. 149, 1168–1178. Jones, H., Black, T.A., Jassal, R.S., Nesic, Z., Grant, N., Bhatti, J.S., Sidders, D., 2017. Water balance, surface conductance and water use efficiency of two young hybrid-poplar plantations in Canada’s aspen parkland. Agric. For. Meteorol. 246, 256–271. Kaimal, J.C., Finnigan, J.J., 1994. Atmospheric Boundary Layer Flows. Oxford University Press, New York. Kaiser, K.E., McGlynn, B.L., Dore, J.E., 2018. Landscape analysis of soil methane flux across complex terrain. Biogeosciences 15, 3143–3167. Kidston, J., Brümmer, C., Black, T.A., Morgenstern, K., Nesic, Z., Mccaughey, J.H., Barr, A.G., 2010. Energy Balance Closure Using Eddy Covariance Above Two Different Land Surfaces 64  and Implications for CO2 Flux Measurements. Boundary-Layer Meteorol. 136, 193–218. Kljun, N., Calanca, P., Rotach, M.W., Schmid, H.P., 2015. A simple two-dimensional parameterisation for Flux Footprint Prediction (FFP). Geosci. Model Dev. 8, 3695–3713. Knox, S.H., Sturtevant, C., Matthes, J.H., Koteen, L., Verfaillie, J., Baldocchi, D., 2015. Agricultural peatland restoration: effects of land-use change on greenhouse gas (CO2 and CH4) fluxes in the Sacramento-San Joaquin Delta. Glob. Chang. Biol. 21, 750–765. Lafleur, P.M., Roulet, N.T., Bubier, J.L., Frolking, S., Moore, T.R., 2003. Interannual variability in the peatland-atmosphere carbon dioxide exchange at an ombrotrophic bog. Global Biogeochem. Cycles 17, 1–14. Larco, H., Strik, B.C., Bryla, D.R., Sullivan, D.M., 2013. Mulch and Fertilizer Management Practices for Organic Production of Highbush Blueberry. II. Impact on Plant and Soil Nutrients during Establishment. HortScience 48, 1484–1495. Law, B.E., Falge, E., Gu, L., Baldocchi, D.D., Bakwin, P., Berbigier, P., Davis, K., Dolman, A.J., Falk, M., Fuentes, J.D., Goldstein, A., Granier, A., Grelle, A., Hollinger, D., Janssens, I.A., Jarvis, P., Jensen, N.O., Katul, G., Mahli, Y., Matteucci, G., Meyers, T., Monson, R., Munger, W., Oechel, W., Olson, R., Pilegaard, K., Paw U, K.T., Thorgeirsson, H., Valentini, R., Verma, S., Vesala, T., Wilson, K., Wofsy, S., 2002. Environmental controls over carbon dioxide and water vapor exchange of terrestrial vegetation. Agric. For. Meteorol. 113, 97–120. Lee, S., Christen, A., Black, A.T., Johnson, M.S., Jassal, R.S., Ketler, R., 2017. Annual greenhouse gas budget for a bog ecosystem undergoing restoration by rewetting. Biogeosciences 14, 2799–2814. Luttmerding H.A., 1981. Soils of the Langley-Vancouver Map Area. Ministry of Environment. 65  Mahrt, L., 1998. Flux sampling errors for aircraft and towers. J. Atmos. Ocean. Technol. 15, 416–429. Massman, W.J., 2000. A simple method for estimating frequency response corrections for eddy covariance systems. Agric. For. Meteorol. 104, 185–198. McNaughton, K.G., Jarvis, P.G., 1983. Predicting effects of vegetation changes on transpiration and evaporation, in: Kozlowski, T.T. (Ed.), Water Deficits and Plant Growth. Academic Press, pp. 1–47. Mishurov, M., Kiely, G., 2011. Gap-filling techniques for the annual sums of nitrous oxide fluxes. Agric. For. Meteorol. 151, 1763–1767. Monteith, J.L., 1965. Evaporation and Environment. Symp. Soc. Exp. Biol. 19, 205–224. Monteith, J.L., Unsworth, M.H., 2013. Principles of Environmental Physics, 4th ed. Academic Press, Oxford, UK. Moore, T.R., Knowles, R., 1989. The influence of water table levels on methane and carbon dioxide emissions from peatland soils. Can. J. Soil Sci. 69, 33–38. Myhre, G., Shindell, D., Bréon, F.-M., Collins, W., Fuglestvedt, J., Huang, J., Koch, D., Lamarque, J.-F., Lee, D., Mendoza, B., Nakajima, T., Robock, A., Stephens, G., Takemura, T., Zhang, H., 2013. Anthropogenic and Natural Radiative Forcing, in: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, pp. 659–740. Nemeth, D., Lambrinos, J.G., Strik, B.C., 2017. The effects of long-term management on patterns of carbon storage in a northern highbush blueberry production system. Sci. Total Environ. 579, 1084–1093. 66  Nemitz, E., Mammarella, I., Ibrom, A., Aurela, M., Burba, G.G., Dengel, S., 2018. Standardisation of eddy-covariance flux measurements of methane and nitrous oxide. Int. Agrophysics 32, 517–549. Neufeld, K., Paul, S., Smukler, S.M., 2017. Improving On-Farm Drainage Management to Reduce the Impacts of Climate Change in Delta, BC: Final Report for the Farm Adaptation Innovator Program (FAIP), British Columbia Project. Pereira, A.R., 2004. The Priestley – Taylor parameter and the decoupling factor for estimating reference evapotranspiration. Agric. For. Meteorol. 125, 305–313. Priestley, C.H.B., Taylor, R.J., 1972. On the Assessment of Surface Heat Flux and Evaporation Using Large-Scale Parameters. Mon. Weather Rev. 100, 81–92. Ramirez, K.S., Craine, J.M., Fierer, N., 2010. Nitrogen fertilization inhibits soil microbial respiration regardless of the form of nitrogen applied. Soil Biol. Biochem. 42, 2336–2338. Rannik, Ü., Haapanala, S., Shurpali, N.J., Mammarella, I., Lind, S., Hyvönen, N., Peltola, O., Zahniser, M., Martikainen, P.J., Vesala, T., 2015. Intercomparison of fast response commercial gas analysers for nitrous oxide flux measurements under field conditions. Biogeosciences 12, 415–432. Robertson, G.P., Paul, E.A., Harwood, R.R., 2000. Greenhouse Gases in Intensive Agriculture: Contributions of Individual Gases to the Radiative Forcing of the Atmosphere. Science (80-. ). 289, 1922–1925. Rochette, P., Liang, C., Pelster, D., Bergeron, O., Lemke, R., Kroebel, R., MacDonald, D., Yan, W., Flemming, C., 2018. Soil nitrous oxide emissions from agricultural soils in Canada: Exploring relationships with soil, crop and climatic variables. Agric. Ecosyst. Environ. 254, 69–81. 67  Shao, Y., Hacker, J.., 1990. Local similarity relationships in a horizontally inhomogeneous boundary layer. Boundary-Layer Meteorol. 52, 17–40. Shurpali, N.J., Rannik, Ü., Jokinen, S., Lind, S., Biasi, C., Mammarella, I., Peltola, O., Pihlatie, M., Hyvönen, N., Räty, M., Haapanala, S., Zahniser, M., Virkajärvi, P., Vesala, T., Martikainen, P.J., 2016. Neglecting diurnal variations leads to uncertainties in terrestrial nitrous oxide emissions. Sci. Rep. 6, 1–9. Skiba, U.T.E., Sheppard, L.J., Hargreaves, K.J., Smith, K.A., Fowler, D., 1996. Soil environmental variables affecting the flux of methane from a range of forest, moorland and agricultural Soils 4. Skinner, C., Gattinger, A., Muller, A., Mäder, P., Fliebbach, A., Stolze, M., Ruser, R., Niggli, U., 2014. Greenhouse gas fluxes from agricultural soils under organic and non-organic management — A global meta-analysis. Sci. Total Environ. 468, 553–563. Smith, P., Bustamante, M., Ahammad, H., Clark, H., Dong, H., Elsiddig, E.A., Haberl, H., Harper, R., House, J., Jafari, M., Masera, O., Mbow, C., Ravindranath, N.H., Rice, C.W., Robledo Abad, C., Romanovskaya, A., Sperling, F., Tubiello, F., 2014. Agriculture, Forestry and Other Land Use (AFOLU)., in: Climate Change 2014: Mitigation of Climate Change. Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA. Smith, W.N., Grant, B., Desjardins, R.L., Lemke, R., Li, C., 2004. Estimates of the interannual variations of N2O emissions from agricultural soils in Canada. Nutr. Cycl. Agroecosystems 68, 37–45. Stoy, P.C., Katul, G.G., Siqueira, M.B.S., Juang, J., Novick, K.A., Uebelherr, J.M., Oren, R., 68  2006. An evaluation of models for partitioning eddy covariance-measured net ecosystem exchange into photosynthesis and respiration. Agric. For. Meteorol. 141, 2–18. Stoy, P.C., Mauder, M., Foken, T., Marcolla, B., Boegh, E., Ibrom, A., Arain, M.A., Arneth, A., Aurela, M., Bernhofer, C., Cescatti, A., Dellwik, E., Duce, P., Gianelle, D., van Gorsel, E., Kiely, G., Knohl, A., Margolis, H., Mccaughey, H., Merbold, L., Montagnani, L., Papale, D., Reichstein, M., Saunders, M., Serrano-Ortiz, P., Sottocornola, M., Spano, D., Vaccari, F., Varlagin, A., 2013. A data-driven analysis of energy balance closure across FLUXNET research sites: The role of landscape scale heterogeneity. Agric. For. Meteorol. 171–172, 137–152. Strik, B.C., 2007. Horticultural Practices of Growing Highbush Blueberries in the Ever-Expanding U.S. and Global Scene. J. Am. Pomol. Soc. 61, 148–150. Stull, R.B., 1988. Boundary Layer Meteorology. Kluwer Academic Publishers, The Netherlands. Suyker, A.E., Verma, S.B., 2012. Gross primary production and ecosystem respiration of irrigated and rainfed maize – soybean cropping systems over 8 years. Agric. For. Meteorol. 165, 12–24. The MathWorks Inc., n.d. Matlab, Curve Fitting Toolbox and Statistics Toolbox Release 2011a. van Genuchten, M.T., 1980. A Closed-form Equation for Predicting the Hydraulic Conductivity of Unsaturated Soils. Soil Sci. Soc. Am. J. 44, 892–898. Vásquez, C., Díaz-calderón, P., Enrione, J., Matiacevich, S., 2013. State diagram, sorption isotherm and color of blueberries as a function of water content. Thermochim. Acta 570, 8–15. Wagner-Riddle, C., Furon, A., McLaughlin, N.L., Lee, I., Barbeau, J., Jayasundara, S., Parkin, G., von Bertoldi, P., Warland, J., 2007. Intensive measurement of nitrous oxide emissions 69  from a corn–soybean–wheat rotation under two contrasting management systems over 5 years. Glob. Chang. Biol. 13, 1722–1736. Webb, E.K., Pearman, G.I., Leuning, R., 1980. Correction of Flux Measurements for Density Effects. Q. J. R. Meteorol. Soc. 106, 85–100. Wilczak, J.M., Oncley, S.P., Stage, S.A., 2001. Sonic anemometer tilt correction algorithms. Boundary-Layer Meteorol. 99, 127–150. Wilson, K., Goldstein, A., Falge, E., Aubinet, M., Baldocchi, D., Berbigier, P., Bernhofer, C., Ceulemans, R., Dolman, H., Field, C., Grelle, A., Ibrom, A., Law, B.E., Kowalski, A., Meyers, T., Moncrieff, J., Monson, R., Oechel, W., Tenhunen, J., Valentini, R., Verma, S., 2002. Energy balance closure at FLUXNET sites. Agric. For. Meteorol. 113, 223–243. Zona, D., Janssens, I.A., Aubinet, M., Gioli, B., Vicca, S., Fichot, R., Ceulemans, R., 2013. Fluxes of the greenhouse gases (CO2, CH4 and N2O) above a short-rotation poplar plantation after conversion from agricultural land. Agric. For. Meteorol. 169, 100–110. 70  Appendices Appendix A  Site photos and schematic A.1 Site photos and blueberry phenology       a) e) b) c) d) f) g) h) Figure 32: Site photos, management examples and progression of blueberry phenology at the site during 2018. a) Eddy-covariance tower and instrumentation hut in January; dormant bud stage. b) March; bud swell stage. c) Late-April; bud burst stage. d) June; early green fruit stage. e) August post-harvest; early fall stage. f) October post-mowing; late fall stage. g) November post-pruning; dormant bud stage. h) December, dormant bud stage. 71  A.2 Site instrumentation and schematic  Figure 33: Instrumentation hut including the a) LI-7000 infrared gas analyzer, b) the LGR continuous wave laser spectrometer and c) the pumps for each instrument.     Figure 34: Photo of the instrumentation hut and sonic anemometer. The sampling line was insulated and heated.  a) b)c)72   Figure 35: Site schematic illustrating the site setup. Bold and solid lines indicate sampling tubing and dotted or dashed lines indicate communications. Fluxes were calculated at the site from the sonic anemometer (Gill R3-50), infrared gas analyzer (LI-7000), and continuous wave laser spectrometer (LGR).    73  Appendix B  Soil pit and soil-water retention curve B.1 Soil profile  Figure 36: Soil profile (0 – 60 cm) at the row-interrow boundary and initial installation of soil monitoring sensors including CS616, MPS-1, and thermocouples at the 60-cm depth on October 23, 2017.   74  B.2 Soil-water retention curve 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1100101102103-  (kPa) Figure 37: Soil-water retention curve developed with the van Genuchten model using in situ measurements of matric potential (𝜑) and volumetric water content (𝜃) at the 5-cm depth at the site during the study period. 𝜑 is binned into 0.05 increments of 𝜃 with the error bars representing the standard deviation of each bin.  Table 10: Soil-water retention curve coefficients and model parameters for the van Genuchten model.   𝜃𝑟 𝜃𝑎 𝑎 𝑚 𝑅2 𝑅𝑀𝑆𝐸 𝑛 0.25 0.64 0.14 2.06 0.89 0.03 17520    75  Appendix C  Diurnal and annual wind rose at z = 3.0 m   C.1 Daytime and nighttime wind rose at the site  Figure 38: Half-hourly wind rose for daytime (𝑆d > 0 or global potential radiation > 0) and nighttime (𝑆d = 0)  grouped into 10° bins for varying windspeeds at the site during 2018.    Wind Speeds in m/sWS  108  WS < 106  WS < 84  WS < 62  WS < 40  WS < 2Day 0.6%1.2%1.8%2.4%3%0%  EW NS  Wind Speeds in m/sWS  108  WS < 106  WS < 84  WS < 62  WS < 40  WS < 2Night 0.6%1.2%1.8%2.4%3%0%  EW NS  76  C.2 Annual flux footprint climatology  Figure 39: Google Earth image of the flux footprint climatology at the site during the study year. The flux footprint was calculated using a parameterization by Kljun et al. (2015). The yellow triangle indicates the tower location. The footprint contour lines increase from 10% to 90%. While the 80% contour line is entirely within the field, the 90% contour line extends over an adjacent road and pasture.    77  Appendix D  Instrument and measurement methodology D.1 Gas analyzer comparison The slope of the OLS regression of 𝜆𝐸 measurements between the IRGA and the LGR was 0.99 (𝑅2 = 0.99, RMSE = 0.75 W m−2, n = 16226) indicating high agreement between the water vapour concentration measurements (Figure 37).  -50 0 50 100 150 200 250 300 350-50050100150200250300350ELI7000 (W m-2)ELGR(W m-2)R2 = 0.996RMSE = 0.75 W m-2 n = 16226 Figure 40: Comparison of latent heat flux measurements between the Los Gatos Research  (𝜆𝐸𝐿GR) and LI-COR (𝜆𝐸LI7000) gas analyzers. The slope of the regression was 0.99, indicating high agreement. The regression line is the solid black line and the dashed line is the 1:1 line.  D.2 Monin-Obuhkov Similarity Theory Monin-Obukhov (M-O) similarity theory suggests that atmospheric parameters such as, variances, can be represented as universal functions of stability (𝑧𝐿) when normalized by scaling parameters (Kaimal and Finnigan, 1994). Variability in the vertical wind velocity 𝜙𝑤 is especially important as EC calculations are directly affected by variations in 𝑤. 𝜙𝑤 is calculated as the ratio of the standard deviation of vertical wind velocity normalized by 𝑢∗ (𝜎𝑤/𝑢∗) and 78  according to M-O theory can be represented by the universal functions for unstable  (Kaimal and Finnigan, 1994) and stable (Shao and Hacker, 1990) conditions: 𝜙𝑤 ={    𝑎 (1 + 𝑏 | 𝑧m − 𝑑𝐿 |)1/3 , −2 ≤𝑧m − 𝑑𝐿≤ 0𝑎 (1 + 𝑏𝑧m − 𝑑𝐿)1/3      , 0 ≤   𝑧m − 𝑑𝐿≤ 1(15) where 𝑎 and 𝑏 are local site parameters, 𝑧m is the measurement height, 𝑑 is the displacement height, and 𝐿 is the Obukhov length. The predicted 𝜙𝑤 was compared to measured 𝜎𝑤/𝑢∗ and in general shows good correspondence with M-O similarity and the hypothesized 1/3 power law for stable (Figure 35) and unstable (Figure 36) conditions despite poor 𝑅2 values for stable conditions (Table 6). More variance was explained in unstable conditions due to improved mixing of the surface layer, which is required for M-O theory. Unity was not achieved at neutral conditions when 𝑧m𝐿= 0 as predicted by M-O theory, as 𝑎 was approximately 1.20 and 1.17 for stable and unstable conditions, respectively (Kaimal and Finnigan, 1994; Shao and Hacker, 1990).  79  10-710-610-510-410-310-210-1100101100z-d/Lw   Figure 41: The normalized standard deviation of the vertical wind velocity 𝜙𝑤 as a function of the stability parameter 𝑧m − 𝑑/𝐿 during stable conditions at the site during the study year. The solid line indicates the fitted function with parameters described in Table 7, and the dashed line is the function predicted from Monin-Obukhov similarity theory (Kaimal and Finnigan, 1994). 10-610-510-410-310-210-1100101100|z-d/L|w   Figure 42: The normalized standard deviation of the vertical wind velocity 𝜙𝑤 as a function of the stability parameter |𝑧m − 𝑑/𝐿 | during unstable conditions at the site during the study year. The solid line indicates the fitted function with parameters described in Table 7, and the dashed line is the function predicted from Monin-Obukhov similarity theory (Kaimal and Finnigan, 1994). The x-axis is reversed to indicate that the absolute values calculated by the stability parameter are initially negative.   80  Table 11: Model parameters for the Monin-Obuhkov similarity relationship between the normalized standard deviation of the vertical wind velocity and the universal stability parameter for unstable and stable conditions at the site during the study year.  𝑎 𝑏 𝑅2 RMSE Stable  1.20 0.22 0.03 0.17 Unstable 1.17 1.7 0.21 0.16    81  Appendix E  Approximation of canopy conductance If the surface is highly coupled to the atmosphere (i.e. low Ω), it can be assumed that the ratio, 𝑔𝑎𝑔𝑐, is very large. The Penman-Monteith equation can be rewritten such that:  𝐸 =𝜌𝑐𝑝𝐷𝜆 𝛾𝑔𝑐 (16) where 𝐸 is the evapotranspiration rate, 𝜌 is the dry air density, 𝑐𝑝 is the specific heat of air, 𝐷 is the vapour pressure deficit, 𝑔𝑐 is the canopy conductance, 𝜆 is the latent heat of vaporization and 𝛾 is the psychrometric constant, implying 𝑔𝑐 can be approximated as the ratio between 𝐸𝐷. Although the site had low monthly values of Ω,  𝑔𝑎𝑔𝑐 may not always be large enough to simplify the Penman-Monteith equation to Equation 16. The difference between the 𝑔𝑐 calculated from inverting the Penman-Monteith equation and the approximated conductance, 𝑔𝑐 ∗ was compared to  𝑔𝑎𝑔𝑐. The comparison suggested that 𝑔𝑎𝑔𝑐 must be greater than 300 to estimate 𝑔𝑐 with 𝐸𝐷 (Figure 40).0 100 200 300 400 500 600 700 800 900 1000-0.02-0.015-0.01-0.00500.0050.010.0150.02 Figure 43: The difference between the canopy conductance (𝑔𝑐) calculated from inverting the Penman-Monteith equation and the approximated canopy conductance (𝑔𝑐 ∗) compared to the ratio of aerodynamic to surface conductance  𝑔𝑎𝑔𝑐.  𝑔𝑐 − 𝑔𝑐 ∗ approaches 0 when  𝑔𝑎𝑔𝑐> 300.  82  Appendix F  Uncertainty in carbon budget The total C budget relied on measurements of NEE and estimations of C content in both Csawdust and Cyield, of which Csawdust was the largest term and determined that the system was a sink for C on an annual basis. However, there is potential uncertainty in the estimate Csawdust as the bulk density of the sawdust was estimated from a range of literature values (Larco et al., 2013; Nemeth et al., 2017), and it is unlikely that the field received exactly a 0.05-m height application of sawdust homogeneously. A range of high bulk density (140 kg dry matter m−3) and application height (0.06 m), and low bulk density (120 kg dry matter m−3) and application height (0.04 m) were considered to estimate the range of possible sawdust application rates (Table 12). Modifying the originally reported sawdust application rate of 488 g C m-2 year-1 (application height of  0.05 m and a sawdust bulk density of 130 kg m-3), the Csequestered could range from 103 to 373 g C m-2 year-1 and has implications on the overall C budget which demonstrated high sensitivity to Csawdust.  Table 12: Range of sawdust application rates in g C m-2 year-1 determined from low and high application height and bulk density. C content was assumed to be 0.50 kg C (kg dry matter)-1 and fraction of the field covered by sawdust estimated to be 0.45 m2 sawdust (m2 field)-1.   Low application height (0.04 m) High application height (0.06 m) Low bulk density (120 kg dry matter m−3) 360 540 High bulk density (140 kg dry matter m−3) 420 630 Values are calculated by: Application height (m) × bulk density (kg dry matter m3) ×C content (kg Ckg dry matter) × sawdust coverage (m2 sawdustm2 field) ×1000 g Ckg C÷ 3 years. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0378327/manifest

Comment

Related Items