Automated Measurements of CO2, CH4, and N2O Fluxes

from Tree Stems and Adjacent Soils

Josep Barba, Rafael Poyatos and Rodrigo Vargas

University of Delaware

15 papers reporting CH4 and/or N2O stem emissions under field conditions in upland forests

What we clearly know from these studies?

• Stems can emit CH4 and exchange N2O with the atmosphere

• High spatial variability within trees and between trees,

species and ecosystems

• High temporal variability at seasonal scales

Temporal variability

Drivers

Magnitudes &

uncertainties

But what is still unclear?

Temporal patterns

Drivers

Magnitudes &

Uncertainties

Study aims

Use AUTOMATIC measurements of CO2, CH4 and N2O to…

- quantify magnitudes of emissions

- understand seasonal and diurnal patterns of emissions

- describe drivers

- bring some light on the origin

75cm

150cmUpperStem

LowerStem

Soil

CO2, CH4 and N2O

Every hour for 100 days(April – July 2017)

7200 measurements

• Sap flow

• Stem temperature

• Soil temperature

• Soil water content

• Meteorological variables

Carya cordiformis Upland forested area, St Jones Reserve, DE

Experimental design

Experimental design

Li-8150Multiplexer

Li-8100A

Li-8100IRGA

PicarroG2508

ExponentialLinear

Exp: 0.28 nmols m-2 s-1

Lin: 0.16 nmols m-2 s-1

LI-CORVersion SoilFluxPro™Soil Gas Flux Software

Flux calculation

CO2

CH4 & N2O

• DIURNAL patterns and drivers Wavelet coherence analysis

(hourly data)

• SEASONAL patterns and drivers Mixed-effects models

(daily averaged data)

(interactions and temporal autocorrelation)

Statistical analysis

Results – SEASONAL TRENDS

Seasonal course of sap flow per unit sapwood area (SF) and CO2, CH4 and N2O fluxes associated withUpperStem, LowerStem and Soil chambers. Points are hourly means taken from Day of the Year 102 to 202.

Sap Flow

CO2

CH4

N2O

Results – SEASONAL TRENDS

UpperStem: 0.28 ± 0.02

LowerStem: 0.46 ± 0.03

Soil: -0.66 ± 0.06(nmol CH4 m-2 s-1)

UpperStem: -0.014 ± 0.006

LowerStem: -0.017 ± 0.008

Soil: -0.046 ± 0.011(nmol N2O m-2 s-1)

CH4

N2O

MODEL Variables MODEL Variables MODEL Variables

UpperStem Temperature UpperStem Temperature UpperStem TemperatureadjR2 = 0.93 SWC adjR2 = 0.40 SF adjR2 = 0.10 SWCp < 0.001 SF p < 0.001 p = 0.032 SF

Temp*SFSWC*SF LowerStem Temperature LowerStem

adjR2 = 0.33 SWC p = n.s.LowerStem Temperature p < 0.001

adjR2 = 0.92 SWC Soil Temperaturep < 0.001 SF Soil Temperature adjR2 = 0.22

Temp*SWC adjR2 = 0.92 SWC p = 0.001Temp*SF p < 0.001 Temp*SWC

Soil TemperatureadjR2 = 0.99 SWCp < 0.001 SF

Temp*SWC

CO2 CH4 N2O

Results – SEASONAL PATTERNS – mixed-effects model

(daily data)

MODEL Variables MODEL Variables MODEL Variables

UpperStem Temperature UpperStem Temperature UpperStem TemperatureadjR2 = 0.93 SWC adjR2 = 0.40 SF adjR2 = 0.10 SWCp < 0.001 SF p < 0.001 p = 0.032 SF

Temp*SFSWC*SF LowerStem Temperature LowerStem

adjR2 = 0.33 SWC p = n.s.LowerStem Temperature p < 0.001

adjR2 = 0.92 SWC Soil Temperaturep < 0.001 SF Soil Temperature adjR2 = 0.22

Temp*SWC adjR2 = 0.92 SWC p = 0.001Temp*SF p < 0.001 Temp*SWC

Soil TemperatureadjR2 = 0.99 SWCp < 0.001 SF

Temp*SWC

CO2 CH4 N2O

R2 = 0.26 R2 = 0.08

Soil Temperature

adjR2 = 0.92 SWC p < 0.001 Temp*SWC

adjR2 = 0.92

CH4

Temperature SWC

Results – SEASONAL PATTERNS – mixed-effects model

(daily data)

Results – DIURNAL PATTERNS – Wavelet coherence analysis

(hourly data)

Temperature Sap Flow

CO2

CH4

N2O

23%

10%

22%

25%

14%

24%

Wavelet coherence analyses output and the percentage of days with significant correlations between CO2,CH4 and N2O of the LowerStem with Temperature (left panels) and SF (right panels) using hourly data.Yellow color indicates significant temporal correlations (p<0.05).

Temporal & spatial

variability

Drivers

UPSCALING

PROCESSES

ORIGIN

Magnitudes &

uncertainties

Conclusions

Thanks