Soil respiration of three chronosequences in Chequamegon National Forest

1
Soil respiration of three chronosequences in
Chequamegon National Forest
James M. Le Moine 24 November 2004
2
Overview

• Introduction to soil respiration
• Soil respiration models
• Study objectives
• Study site, respiration measurements
• Statistics
• Results
• Summary of key findings
• Future work

3
Soil Respiration
Respiration
Food
Energy Waste
4
Soil Respiration and the Carbon cycle
Soil 4th largest pool 2nd largest source to
atmosphere
http//earthobservatory.nasa.gov
5
Empirical Relations with Microclimate
SRR (gCO2 m-2hr-1)
Soil Moisture ()
Soil Temperature (oC)

• Type of curve fitted depends on data range

• SRRtSRR0e(qt)
• Q10 change in response for 10 Cº increase in
  temperature

6
Established Trends
GPP
Ra
Flux relative to GPP
Rh
Time Since Disturbance
Redrawn from Barnes et al 1980

• Autotrophic and heterotrophic respiration change
  with maturation.
• Soil respiration is comprised of Rh and Ra.

7
Unanswered Questions
Does soil respiration have a trend with
maturation? Do different ecosystems have
different trends in soil respiration with
maturation? Are soil respirationtemperature
relationships consistent across maturation
classes?
8
Objectives

• Model SRR from temperature in clearcuts and three
  age classes of hardwood, jack pines, and red
  pines
• Compare SRR, and its temperature sensitivity
  across forest types, age classes, and age class
  within forest types.
• Correlate SRR and its temperature sensitivity to
  common, easily obtained, vegetation and soil
  metrics
• Compare the above correlations across forest
  types, age classes, and age classes within forest
  types

9
Study site
Chequamegon National Forest

• Heavily managed forest
• 34200 m of outwash sands and loamy sands
• Growing season
• 120 to 140 days
• Precipitation
• 66 to 70 cm rain and 106150cm snow

Figure adapted from Steve Mather
10
Vegetation Types and Age Classes

• Recent Clear Cuts 1, 2, and 4 years since cut
• Mixed hardwoods
• 12, 14, 17,19 years 21, 22, 26
  years 71, 74, 79 years
• Jack Pine Plantations and naturally regenerated
• 11,11,13 years 19, 19, 21, 29 years 68, 68,
  69 years
• Red Pine Plantations
• 11,12,14,14 years 23, 24, 24, 31, 32 years 71,
  72, 74 years

11
Soil Respiration Measurements

• 8 soil respiration collars
• Soil respiration rate
• Soil temperature (5cm)
• Gravimetric soil moisture
• Measurements every 2 wks
• mid Juneearly September 02
• late Aprillate October 03

12
Statistical Analysis

• Shapiro-Wilk (a0.01)verification of normality
• Nonlinear regressionSRRtSRR0e(qt)
• Analyses of Variance (ANOVAs) on SRR0, Q10, SRR15
• Vegetation Type and Age Class(Vegetation)
• Vegetation Type
• Age Class

13
Nonlinear Regression
14
Model fits
Predicted SRR (g CO2 m-2 h-1)
Predicted 0.9408Actual R2 0.6776
Actual SRR (g CO2 m-2 h-1)
Average model had n15,SRR0 0.14, Q102.91,
R20.94
15
Temperature and Moisture Correlation
Spearman-0.4930, p0.0001
SRR (g CO2 m-2 h-1)
Gravimetric soil moisture ()
Individual plot correlations ranged from -0.59 to
-0.80
16
Temperature Data Range
17
Temperature Data Range

• No data prior to 15 May or post 23 August

18
Effects on Q10

• Given SRRtSRR0e qt
• Q10e10q
• Q10 all data e 100.1224
• 3.40
• Q10 parsed 2.30

19
Similarity of SRR0
20
Q10 by Age within Vegetation Type

• Nested Age not significant
• Young and intermediate hardwoods greater than
  other groups

21
Q10 by Vegetation Type
Q10 by Age Class
a
a
a
a

• Hardwood SRR is more temperature sensitive than
  others
• There is no consistent age effect on Q10

22
SRR15 by Age within Vegetation Type
Vegetation type and age class

• Nested Age not significant
• Young and intermediate SRR15 greater than other
  groups

23
SRR15 by Vegetation Type
SRR15 by Age Class

• Hardwood SRR15 is more temperature sensitive than
  others
• There is no consistent age effect on SRR15

24
Summary

• Temperature alone explains 94 of variation in
  SRR. Temperature and moisture strongly
  negatively correlated.
• All SRR0 similar. Reflects severe temperature
  limitation.
• Mean Q10 similar to global average of 2.4. As is
  the range of values (Raich and Schlesinger, 1992)
• Considerable variation between replicates.
• Statistically only need 1 model for hardwoods and
  1 model for all other ecosystems.

25
Objectives

• Model SRR from temperature in clearcuts and three
  age classes of hardwood, jack pines, and red
  pines
• Compare SRR, and its temperature sensitivity
  across forest types, age classes, and age class
  within forest types.
• Correlate SRR and its temperature sensitivity to
  common, easily obtained, vegetation and soil
  metrics
• Compare the above correlations across forest
  types, age classes, and age classes within forest
  types

26
Vegetation and Soils

• Age (Ewel, 1987 Field Fung, 1999 Law et al
  2001 Pypker Fredeen, 2003 and others)
• Basal Area (BA)
• Foliage Mass (FL)
• Canopy Cover (Cover)
• Down Woody Debris (CWD, IWD, FWD) Mallik Hu 2001
• Litter Depth (LD) Euskirchen 2003
• Depth of Organic layer (OD)

27
Vegetation and Soils
28
Foliage Mass

• HaDb
• FL(cDd)(He)
• Ddiameter at breast height
• HHeight
• FLFoliage mass
• a, b, c, d, e species specific coefficients
• Crow and Erdman, 1983 Perala Alban, 1994
  Ter-Mikaelian and Korzukhin, 1997 and
  Young et al. 1980

29
Linear Correlation

• Three steps
• Overall
• By Vegetation Type
• By Age Class
• Pearson correlation for normally distributed data
• Spearman correlation for non-normal
• Significance 0.05
• Marginal Significance 0.05 to 0.10

30
SRR0 Correlation

• Negatively correlated with Q10at all levels
• No correlation with vegetation or soil variables

31
Q10 Correlation with Vegetation

• Overall BA, FL, Cover
• Vegetation Type
• Hardwood Cover
• Jack Pine none
• Red Pine -FWD

• Age Class
• Young marginal BA
• Intermediate none
• Mature Cover

32
Q10 Correlation with Soils

• Overall Positive correlation with OD
• Vegetation Type
• Hardwood Positive with LD, OD
• Jack Pine none
• Red Pine none
• Age Class
• Young none
• Intermediate OD
• Mature none

33
SRR15 Correlation with Vegetation

• Overall Age, BA, FL, Cover
• Vegetation Type
• Hardwood Cover, marginal CWD
• Jack Pine none
• Red Pine Cover, marginally BA

• Age Class
• Young BA, FL, Cover
• Intermediate none
• Mature none

34
SRR15 Correlation with Soils

• Overall LD, OD
• Vegetation Type
• Hardwood OD
• Jack Pine none
• Red Pine none
• Age Class
• Young LD, OD
• Intermediate none
• Mature OD

35
Correlation Summary

• SRR15 had more correlations with vegetation and
  soil variables than did Q10
• Canopy cover has most correllates of vegetation
  variables. Depth of the organic layer is a
  better correlate than litter depth.
• Hardwoods had more correlations than did pines
• The young age class had more correlations than
  did inermediate or mature

36
Future work

• SRR flux source partitioning
• SRR as related to stand GPP and NEE
• Improve landscape level estimates of CO2 efflux
• SRR as related to soil CN, root abundance, root
  turnover, and differences in microbial community.
  

37
Questions and Comments