The Twenty-fourth IHP Training Course

Forest HydrologyConservation of Forest, Soil, and WaterConservation of Forest, Soil, and Water

Resources23 Nov.~7 Dec., 2014 @ Nagoya, Japan

Forest DynamicsForest Dynamics~basis and modelling~

26 Dec 09:30-12:00, Lecture26 Dec 14:00-16:30, Exercise

Hisashi SATOLecturer

Hisashi SATO(Japan Agency for Marine-Earth Sciences and Technology)

MIROC-ESM: Japan’s ESMBlock Diagram

Grid ResolutionsAGCM: T42(128×64), 80 levsOGCM: Cartesian(256×192)

Block Diagram

OGCM: Cartesian(256×192), 44 levs

Developing team:Universities (Tokyo, Hokkaido, Nagoya, Kyushu), JAMSTEC, NIES

i l tf f O tiFrom Watanabe et al. (2011)

Main Platform for Operation:Earth Simulator (JAMSTEC)

Interactions betweenInteractions between Vegetation and Atmosphereg p

[Example] Climatic effects of Tropical deforestation

Foley et al. (2003)

How vegetation affects interactionsbetween terrestrial surface and atmosphere

Carbon Soil SurfaceTemp.

Soil Surface

Carbon Transpiration/ ShortwaveSensible

Carbon Soil SurfaceTemp.

Carboncycle

Transpiration/Latent heat Precipitation Shortwave

reflectionSensible

heat

[Example] Vegetation impact on surface temperature

Geographical distribution of surface temperature observed at Sendai city during daytime on a mid-summer day

serv

ed

),

air u

sing

a m

omet

er (O

bs

Kond

o (2

000)

mad

e in

the

adi

atio

n th

erm

ARA

). Fr

om K

al. (

2014

)

atio

ns w

ere

mpt

er-b

ased

rad

fum

i SUG

AW

Aed

by

Sato

et

Obs

erva

helic

opby

Hiro

fm

odifi

e

Timescales of land-ecosystem processes

PHYSIOLOGICAL ECOLOGICAL

Photosynthesis

Stomatal Opening

Demography(Changes in growth, mortality, & recruitment rates)

Biome Shift(Changes in Stomatal Opening recruitment rates)

Succession(Changes in canopy structure

Leafx

( gdistribution of biome)

(Changes in canopy structure & composition)Phenology

Hours Months Years Decades CenturiesWeather Inter-annual Anthropogenic

Incorporating ecological Processes is critical to forecast the

Prediction Variabilityp g

Climate Change

Incorporating ecological Processes is critical to forecast the responses of land‐ecosystems to anthropogenic climate change

Original Figure: Prof. Moorcroft, P.R.

The Development of Climate models

2001

(IPC

C)

he W

orki

ng

ate

Cha

nge

2Su

mm

ary

of th

port

From

:Clim

aTe

chni

cal S

Gro

up I

Rep

Basis for plant population Basis for plant population dynamicsdynamics

Topics: • One-sided competition for light• Self-thinning rule & Three-halves lawSelf thinning rule & Three halves law• Succession• Gap dynamics• Gap-dynamics• Wild fire as a major disturbance scheme

One sided competition for lightOne-sided competition for light

A Light attenuationTree ATree A

A Light attenuation formulation based on the Lambert‐Beer law

Tree BTree B

Tree CTree CI = e(k×LAI)

Relative Sun‐Light Intensity

Leaf Area Index above the mentioned la er (m2)

I

LAI

GrassRelative Light Intensity Leaf Density Forest Structure

Grass mentioned layer (m2)

Light Attenuation Factor (Typically around ‐0.5)

K

Local vertical structure prominently controlsLocal vertical structure prominently controls competition for light among trees

The Self-thinning & Three-halves lawI d l t l ti th "S lf thi i " ithIn a dense plant population, the "Self‐thinning" occurs with growth of mean plant size. This process generally results in the "Three‐halves law"the  Three halves law

Theory

t (g) Example (Obs.)

W ∽ L 3 A ∽ L 2

per

plan

t

Figure :Iwatsubo (1996) 森林⽣態学

n bi

omas

s

W ∽ A 3/2 A ∽ D -1 Tree Density (m-2)Mea

n

L  : Length of side of a plantW : Mean biomass per plantA : Area per a plantW ∽ D 3/2 A : Area per a plantD : Plant density

W ∽ D -3/2

Secondary Succession

Succession: Sequential changes of physiognomy, those are mediated by plant induced environmental changes (such as soil

eb. 9

Sep

t.

[Example] Temperate forest in North America

mediated by plant induced environmental changes (such as soil formation and sun light interception)

Onl

ine

for K

ids.

We

9013

0>.

[ p ] p

. Art

. Brit

anni

ca O

om/c

ompt

ons/

art-

gure

:on

dary

suc

cess

ion.

/kid

s.br

itann

ica.

co

Sour

ce o

f thi

s fi

fore

st fi

re:

seco

2014

. <

http

://

Secondary succession is a much faster process y pthan to the primary succession

Primary Succession

If succession begins in barren areas without soil, it is 

d

g ,called Primary Succession

for K

ids.

Ret

rieve

d9

[Example] Temperate forest in North America

Brita

nnic

a O

nlin

e f

mpt

ons/

art-

9012

9

re:

essi

on. [

Art]

. In

Brit

anni

ca.c

om/c

om

Sour

ce o

f thi

s fig

urre

e: p

rimar

y su

ccro

m h

ttp:

//ki

ds.b

r

If it begins in barren area where soil was not formed S tr frIf it begins in barren area where soil was not formed, it is called Primary Succession

Changes in land-atmospheric interactions with successionSchematic diagram for post‐fire changes observed 

( )

interactions with succession

in the Alaska (during summer time)

Albedo

Soil Surface temp.

13

Net Radiation

n(2

008)

JGR

1

Sensible Heat flux

L t t H t fl

u &

Rand

erso

nLatent Heat flux

Figu

re :

Liu

Carbon cycles development of a Single Storied Stand

Foliage respiration rate

Foliage biomass ∝ GPP Phase 1

Stem respiration rate

Foliage respiration rate

Phase 2 F li bi

NPPGPP

Phase 2 Foliage biomass reaches maximum

Total respiration rate

Phase 3熱帯

林の

⽣態

NPP

Phase 3Phase1 Phase2 Phase3

Stem biomass still growth, resulting in higherFig

ure:

Kira

(198

4) 熱

Time

in higher respiration

Sour

ce f

this

F

• Foliage biomass saturate faster than stem biomass• So, NPP gradually declines after foliage saturation

Gap DynamicsGap Dynamics

Small gapSmall gap

Saplings of shade‐tolerant species fill the gapspecies fill the gap

Saplings of sun treeSaplings of sun tree species fill the gap

Large gap

Shade tolerant trees Sun trees

g g

Deaths of a canopy tree can cause gap dynamics, which is a partial decay and recovery (due to secondary succession)partial decay and recovery  (due to secondary succession) of small area in a dense forest

Wild Fire as a major disturbance scheme

Fire frequencyFire frequency(ASTER2)

Wild Fire is a common disturbance scheme among gSemi‐arid area and Boreal forest area

Role of wild fire in semi-arid regions

Sparse forest maintain thick grass layer, and it produces 

h f l l dVegetation Type Fire frequency

much fuel load

h d h f lIn the dry season, this fuel load helps to spread wild fire 

Mortality rate due to wild fireFi tl kill ll Short

Tree type

Fire mostly kills small trees, and such size dependency maintains 

Tree type b

Tree type

Tree type d

Tree type

Talla

p ysavanna

b c d e

Hoffmann & Solbrig (2003) For. Ecol. Manage., 180

Wild Fire maintains Savanna ecosystem

Role of wild fire on Larch Forest in East Siberia

Accumulation of organic layer on soil surface

Reduction of tree growth due to little soil water

Recruitment of Larch Trees

Rising Permafrost layerRising Permafrost layer

~10 years 20~40 Years 100 years~

Stand replacing fire (ca. 200 years interval in Eastern Siberia)

Lowering Permafrost layerLowering Permafrost layerExposed soil enhances heat exchanges between

il d t h0 Years

soil and atmosphere

Wild Fire is a key for regeneration of larch forest

L d S f M d lLand Surface Models (LSMs) those treat plant ( S s) ose ea p apopulation dynamics

About SEIB DGVM

VegetationI/O and Composition of SEIB-DGVM

About SEIB-DGVM

VegetationRepresentation

I/O and Composition of SEIB DGVM

A simulation result for

Individual trees compete f li ht d

A simulation result forTemperate forest

for light and space within a virtual forest

Sub-models that compose SEIB-DGVMProcess Approach Source

Physical process

Radiation Beer's law

Penman-Monteith transpirationEvapotranspiration

p+ interception+ evaporation from soil surface

Monteith & Unsworth(1990)

Soil water process Simple bucket model Manabe (1969)

Physiology Photosynthesis Michaelis-type function

Maintenance respiration

respiration rate is in proportion to nitrate contents for each organ Ryan (1991)

Growth respiration based on chemical composition of each organ Poorter(1994)

Stomatal conductance a semi-empirical model as a function of VPD Leuning et al. (1995)

Phenology a set of semi-empirical models of which parameters were estimated from satellite NDVI data Botta et al. (2000)

Decomposition 2 carbon source of decomposition: labile part of litter andpassive part in mineral soil

Sitch et al. (2003) p p

EcologocalDynamics

Establishment climatically favored PFTs establish as small individuals Sitch et al. (2003)

Mortality function of “annual NPP per leaf area”, “heat stress”, “bioclimitic limit” and “fire” Sitch et al. (2003) y bioclimitic limit , and fire

Disturbance (fire) an empirical function of soil moisture and fuel load Kirsten et al (2001)

Plant species were summarized intoPlant Functional Types (PFTs)Plant Functional Types (PFTs)

Woody PFTs (8 types)Tropical broad-leaved evergreenTropical broad-leaved raingreenTemperate needle-leaved evergreenTemperate needle-leaved evergreenTemperate broad-leaved evergreenTemperate broad-leaved summergreenBoreal needle-leaved evergreenBoreal needle-leaved summergreenBoreal broad-leaved summergreenBoreal broad leaved summergreen

Grass PFTs (2 types)C3 - grassC4 - grass

Grass PFTs (2 types)

Herbaceous species are represented by average biomass per unit areaC4 grass by average biomass per unit area

Photosynthesis condition for woody PFT (Di t di ti )PFTs (Direct radiation)

Midday radiation is calculatedfor each individual treefor each 10cm‐interval crown layer.

Lower crown layer suffers from self‐shadingg

To avoid ‘edge effect’ it is assumedTo avoid  edge effect , it is assumed that virtual forest repeats

Photosynthesis condition for woody PFT (Diff d di ti )

Leaf area Relative

PFTs (Diffused radiation)

Leaf areadensity

Relativeradiation

1.00.00.0

Based on average leaf‐area‐density for each crown layer, mean intensity of diffused radiation was calculated for each crown layery y

Horizontal structure was ignoredHorizontal structure was ignored

Growth procedure for woody PFTs (1)

Daily Allocation f

Daily available computationcontrol factors

l fbiomass G h f

resource

constantroot

leaf

biomassbiomass Growth of

Root

Growth of Stock organ

constantstock

leaf

biomassbiomass

Water supply constraindue to the cross‐section

Growth ofLeaves

Packing constrain

area of sapwood

E ill bgdue to the crown size Excess resource will be 

accrued over to the next day

Growth procedure for woody PFTs (2)Allocation

Spatial constraint Adjustments of

Monthly computationAllocation

control factors

Spatial constraintby proximate trees

Adjustments of Trunk, Tree height, and crown diameter growth

Allometry rules among DBH, tree height,  Reproductionand crown diameter with all remaining carbon at this stage

Annual total

Adjustment

Mean NPP for each crown layer Optimum

Annual computationAdjustment of crown depth

y pCrownDepthMean maintenance

respiration cost for depthrespiration cost for each crown layer

Growth procedure for Grass PFTs

DailyAllocation

/

Daily computation

Weekly

Allocation control factors

massleaf Leaf/Root growthRunningmeans

.__ const

massrootmassleaf

Stock Organ growthUntil it comes to same amount

NPP on top of the grass layer Optimal 

Until it comes to same amount of existing leaf mass

Maintenance respiration

LAI

ReproductionAll remaining resource is used for reproduction (add to litter)

respiration rate of leaves

for reproduction (add to litter)

Carbon and Water cyclesy

Carbon cycles Water cycles

From Sato et al. (2007)

Class Discussion

How Dynamic Vegetation Models can be utilized for Forest Managementutilized for Forest Management

SupplementsppPlant geography andPlant geography and BiomeTopics:• Global distribution of natural vegetation• Biomes (Definition, Distribution, and Boundary)( , , y)• Bioclimatic limits

An example of Biome:Tropical rain forest

Dense and stratified

Tropical rain forest

Common characteristicsforest structure

Epiphytic plantsButtress root

Common characteristics

Drip tipsButtress root

Photos are gathered from the Web

These characteristics are commonly observed in Tropical rainThese characteristics are commonly observed in Tropical rain forests, although species composition much differs among regions

Global Distribution of Natural Vegetation1

Tropical Evergreen ForestTropical Evergreen ForestTropical Deciduous ForestTemperate Broadleaf Evergreen ForestTemperate Needleleaf Evergreen ForestTemperate Deciduous ForestBoreal Evergreen ForestBoreal Deciduous ForestBoreal Deciduous ForestEvergreen/Deciduous Mixed ForestSavannaGrassland/Steppe/ShrublandTundraDesert

Biome:Biome: A major regional ecological community characterized by distinctive life forms and principal plant species2by distinctive life forms and principal plant species .Terrestrial ecosystems are typically classified into 5~20 biomes, those are mostly determined by climate.

1 ISLSCP22 A Dictionary of Ecology, Evolution and Systematics

biomes, those are mostly determined by climate.

Biome distribution in eastern AsiaBiome distribution in eastern Asia

Very short summer, andVery short summer, and very cold winter

Needle leaves are tolerant for frost damage and

Winter is not suitable for 

for frost damage and dehydration

Moderate climate throughout the year

photosynthesis

Warm and Humid throughout the year

g y

Adams (2010) Vegetation-Climate Interaction

Warm throughout the year, but dry season exist

In eastern Asia, alternative band of Evergreen and Deciduous forest exists along latitude

Biome boundaryBiome boundary

[Example] Northern part of Mongolia

North-faced slope:Larch forest

South-faced slope: StSteppe

Photo: Forests of Northern Mongolia - FCA Today (www.fca-today.com/page13.html)

Biome boundaries generally transit gradually, but topographic heterogeneity produces mosaic‐structured transition zone

Biome distribution (Whittaker)

This is just an empirical patternImages are gathered from the Web

Biome distributionBiome distribution(Holdridge life zone)

Images are gathered from the Web

Efforts have been paied to establish more mechanistical criterion by employing a Bio‐criterion by employing a Biotemperature and an Aridity index.

Bioclimatic limits 1: Physiological Requirements†

Biome distribution was actually controlled by Bioclimatic limits for each Plant Functional Types (PFTs)

[Example] High-temperature injury49 ° C : For most plant species64 ° C : For some succulent species

[Example] Frost damage-15 °C < T -40 ° C < T

: Evergreen Broad Leaved Species: Deciduous Broad Leaved Species-40 C < T

No limits‡: Deciduous Broad Leaved Species: Boreal Conifer Species

PFTs: a classification of plant species in terms of their responses to environmental changes such as higher air‐temperature and CO

† : Beerling & Woodesrd (2001) Vegetation and the Terrestrial Carbon Cycle: "植⽣と⼤気の4億年(及川武久 監修)“‡ For the Minimum Air temperature in the nature of the earth surface

to environmental changes such as higher air‐temperature and CO2

Bioclimatic limits 2:Bioclimatic limits 2: Requirements for satisfying Life Cycle

[Example] Temperature requirements for Woody SpeciesKoppen (1936) Ojima (1991)T > -5T < 42×log P ­ 106

T : annual mean temperature (in °C)P : annual precipitation (in mm)

P > 100P > 20.0 × T

[Example] Growth Degree Day (GDD)* requirements for woody PFTs in the LPJ‐DGVMfor woody PFTs in the LPJ DGVM

GDD> 1200 : Temperate broad‐leaved (evergreen/summergreen)GDD>  900 : Temperate needle‐leaved evergreenGDD>  600 : Boreal needle‐leaved evergreenGDD>  350 : Boreal summer greeen (neelde/broad‐leaved)

* Annual sum of daily air temperature above which 5 °C.

How to determine Biome in vegetation modelsHow to determine Biome in vegetation models

Cli d

St ti t ti d l D i t ti d l

Climate data

Static vegetation models Dynamic vegetation models

Bioclimatic envelopeBioclimatic envelope determines PFTs to establishBioclimatic envelope determines PFTs to establish

Competition among PFTs

Combinations of PFTs those existCombinations of PFTs those exist

Biome was determined with some criteriaBiome was determined with some criteria

Biome ShiftG hi l Di t ib ti f Cl d F t d D tGeographical Distribution of Closed-Forest and Desert

@Present Day

tion

nter

actio

n 2n

ded

it

]

e: geta

tion-

Clim

ate

I

@18000 yrs ago

urce

of t

he F

igur

eda

ms

(200

7) V

egSo

uAd

Ranges as indicated by pollen percentages in sediment ofRanges, as indicated by pollen percentages in sediment, of spruce and oak in eastern North America

& S

haw

(200

1)

f f f k h l

Dav

is &

Species of trees, not communities of forests, tracked their climatic niches at their own rates and along their own trajectories

Carbon cycles in vegetationCarbon cycles in vegetation

Topics:• GPP, NPP, NEE, ,• Estimation methods for carbon cycles• General patterns of carbon cyclesp y

Carbon Cycles in Land-Ecosystems

CO2 emission CO2 Uptake

h b

GPP (Total Photosynthesis Production)

ΔY: Growth biomass

ΔG: Grazed biomass

ΔL: Litter fall

NPP

Rr: Respiration

BiomassBiomass

ΔL: Litter fall

Rh: HeterotrophicSoil Organic

Matter (SOM)

GPP † = NPP + Rr

pRespiration

( )

NPP ‡ = ΔY + ΔG + ΔLNEP ¶ = GPP – Rr – Rh = ΔBiomass + ΔSOM

NEP Estimation 1 (Summation Method)

NPP by Stem diameter t

ΔY (Growth Biomass)+

measurements=

+ΔL (Litter Fall)

+ http://www.forestry.ac.nz/+ΔG (Grazed Biomass)

Combined with

by Litter TrapMeasurements

Species-SpecificAllometricEquations

by Estimated Root Turnover rate

+Gill & Jackson (2000) New Phytol 147http://www.crestmonsoon.org/maemoh/

NEP Estimation 2GPP Estimate

Measurements of

Integration for both of 

Measurements of photosynthesis rate

Time and Leaf layershttp://www.licor.com/

Respiration rate EstimateStem Diameter Measurements

AllometryRelationship

Respiration rate Estimate

Mori, S., et al. (2010)

PNAS 107

p

PNAS 107

http://www.forestry.ac.nz/

Adaptation and acclimation of photosynthesis properties of leaves to environmental light p p gintensity

High[Example] Stratified forest

Ph

Max

Main

High0

[ p ]in a tropical rain forest

Compensationpointotosyn

Photosy

tenanc

point

nthetic

ynthetic

e Respi

0This illustration is gathered from the Web

rate

c rate

iration

0

Low0

Leaf properties are “optimized” t th li ht i t it f th Light intensityto the sun light intensity of the given environment

Estimation of NEP (or NEE)With th Edd l ti th dWith the Eddy correlation method

Example: Estimation of Vapor fluxFlux tower

VerticalWind Speed

p p

Wind Speed

Absolutehumidity

Vaporflux

humidity

flux

Time

( )Ultrasonic Wind Sensor

NEP = ‐NEE = ‐(FC+SC)FC: CO2 fluxSC: Changes in CO concentration

CO2/H2O AnalyzerFigure: Terashima (2013) 植物の⽣態Photos: www.weather.co.jp/ex/tomakomai.htm、クリマテック株式会社

SC: Changes in CO2 concentration

Estimated carbon cycles of key biomes (1)

Figure: Eddy van der Maarel (2005) Vegetation Ecology

E ti t d b l f k bi (2)

Ranges of estimated NPP/GPP ratio

Estimated carbon cycles of key biomes (2)

0.6

0.7

Ranges of estimated NPP/GPP ratio

0.3

0.4

0.5

0.1

0.2

0.3

Amthor & Boldocchi (2001): Terrestrial higher plant respiration and

0Boreal Forest

Temperate Forest

Tropical Forest

Grassland Cropland

Amthor & Boldocchi (2001): Terrestrial higher plant respiration and net primary production In: Terrestrial global productivity

• Biomes without stem biomass have lower ratio Biomes without stem biomass have lower ratio• Tropical Forest has the lowest ratio among the key biomes