levels of biological integration orest management and...
TRANSCRIPT
Hybrid ecosystem-level forest models as tools for forest management and research
Bla
nco
J.A
.1,2
, K
imm
ins J.P.1
, See
ly B
.1, W
elham
C.1
, Sco
ullar
K.3
1 D
ep. Fore
st S
cien
ces, F
aculty o
f Fore
stry
, The
Univ
ersity
of British
Colu
mbia
, 3041-2
424 M
ain M
all, V
6T 1
Z4, V
anco
uver
, B.C
.
2Conta
ct: ju
an.b
lanco
@ubc.
ca3Life
Sci
ence
Pro
gra
mm
ing L
td., N
aram
ata,
B.C
.
WHY ECOSYSTEM
WHY ECOSYSTEM-- LEVEL MODELS?
LEVEL MODELS?
Levels of biological
Levels of biological
organization
organization
Levels of biological
Levels of biological
integration
integration
Ecosystem
Ecosystem
Understanding and
Understanding andPrediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Ecosystem
Ecosystem
Community
Community
Understanding
Understanding
Population
Population
Understanding
Understanding
Individual
Individual
Understanding
Understanding
and
andPrediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Individual
Individual
Organ systems
Organ systems
Understanding
Understanding
Organs, tissues
Organs, tissues
Understanding
Understanding
Cell
Cell
Understanding and
Understanding andPrediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Prediction
Cell
Cell
Sub
Sub-- cellular
cellular
Understanding
Understanding
Function of level
Prediction: The need for ecosystem level
Prediction: The need for ecosystem level
Bio
logic
al k
now
ledge
is o
rgan
ized
into
lev
els
of
bio
logic
al o
rgan
izat
ion that
are
indispen
sable
for th
e des
crip
tion a
nd u
nder
stan
din
g o
f ev
ents a
nd c
onditio
ns at
eac
h
of
thes
e le
vel
s. H
ow
ever
, pre
dic
tion o
f fu
ture
even
ts a
nd c
onditio
ns
at a
ny o
f th
ese
level
s ca
n o
nly
be
succ
essf
ul in
the
conte
xt of
the
nex
t le
vel
of
bio
logic
al inte
gra
tion
above
(Row
e 1961). Indiv
idual
lev
els of bio
logic
al o
rgan
izat
ion d
efin
e only
a s
ubse
t of
the
pro
cess
es that
affec
t fu
ture
conditio
ns
and e
ven
ts a
t th
at lev
el. O
nly
the
nex
t true
level
of
inte
gra
tion ab
ove
in th
e hie
rarc
hy of
system
co
mple
xity
def
ines
th
e key
det
erm
inan
ts o
f th
e fu
ture
for th
e le
vel
of in
tere
st.
Thus, the
fate
of an
indiv
idual
org
anism
can
not be
def
ined
sole
ly o
n the
bas
is
of know
ledge
of th
e bio
logy o
f th
at indiv
idual
, or of th
e popula
tion o
r ev
en o
f th
e bio
tic
com
munity in w
hic
h it finds
itse
lf. The
popula
tion lev
el f
ails to iden
tify
all the
bio
tic
fact
ors
influen
cing that
indiv
idual
, an
d n
eith
er the
popula
tion n
or th
e co
mm
unity lev
el
addre
ss the
clim
atic
and e
dap
hic
fac
tors
and the
physica
l nat
ura
l distu
rban
ce e
ven
ts that
pla
y such
a k
ey role
in d
efin
ing the
futu
re for th
at indiv
idual
(K
imm
ins et
al. 2
005).
REFERENCES
Kimmins J.P., MaillyD., Seely B. 1999.M
odel
ling fore
st e
cosy
stem
net
prim
ary p
roduct
ion: th
e hybrid sim
ula
tion a
ppro
ach u
sed in F
OR
EC
AST. Eco
l. M
odel
. 122, 195-2
24.
Kimmins J.P., Welham C., Seely B., Meitner M., RempelR., Sullivan T. 2005.Sci
ence
in fore
stry
: W
hy d
oes
it so
met
imes
dis
appoin
t or ev
en fai
l us?
T. For. C
hro
n. 81, 723-7
34.
Rowe J.S. 1961. The
level
-of-
inte
gra
tion c
once
pt an
d e
colo
gy. Eco
logy 4
2, 420–427.
Seely B., Nelson J., Wells R., Peter B., Meitner M., Anderson A., HarshawH., Sheppard S., BunnellF.L., Kimmins H., Harrison D. 2004. The
applica
tion o
f a
hie
rarc
hic
al,
dec
isio
n-s
upport syst
em
to e
val
uat
e m
ultio
bje
ctiv
e fo
rest
man
agem
ent st
rate
gie
s: a
cas
e st
udy in n
orth-e
aste
rn B
ritish
Colu
mbia
, C
anad
a.
THE FAMILY OF MODELS developed by THE FOREST ECOSYSTEM
THE FAMILY OF MODELS developed by THE FOREST ECOSYSTEM
MANAGEMENT SIMULATION GROUP at UBC
MANAGEMENT SIMULATION GROUP at UBC
FO
RECA
ST u
ser in
terfac
e
FORECAST
FORECAST
Hybrid model:
It use
s histo
rica
l/fiel
d gro
wth
dat
a to
sim
ula
te f
utu
re g
row
th &
yie
ld. This s
imula
ted g
row
th
is m
odifie
d b
y som
e bio
logic
al p
roce
sses
:
-light co
mpet
itio
n
-nutrie
nt av
aila
bility
Non-spatial stand-level model
: It d
oes
not ac
count fo
r
indiv
idual
ste
ms
and tre
e positions
in the
stan
d, but
it
does
hav
e a
tree
list an
d tra
cks in
div
idual
ste
m siz
es
Ecosystem management
model:
It
sim
ula
tes
inte
ract
ions
bet
wee
n
ecosy
stem
co
mponen
t (tre
es,
min
or veg
etat
ion -
shru
bs, h
erbs, b
ryophyte
s -an
d the
soil)
and
the
influen
ce
of
diffe
rent
man
agem
ent
pra
ctic
es a
nd n
atura
l distu
rban
ces on them
.
FO
RECA
ST g
raphic
al o
utp
ut
a. Harvest map with underlying forest types / conditions
b. Regeneration pixel groups
c. Light ecotones
a.
b.
c.
Trees
Ecotone
Open
0100200300400
0
100
200
300
400
m
m
Example of 16
ha block
LLEMS
LLEMS
Hybrid model:It w
ill use
s FO
RECA
ST to e
stim
ate
key
ecosy
stem
pro
cess
es (light &
nutrie
nt co
mpet
itio
n).
Spatial (raster-based) stand-level model
: In
div
idual
10x10 plo
ts (p
ixel
s) ar
e sim
ula
ted by FO
RECA
ST,
incl
udin
g
indiv
idual
tree
list
info
rmat
ion.
Ther
e is
bet
wee
n-p
ixel
inte
ract
ion in ter
ms
of light, litte
rfal
l an
d
seed
disper
sal, a
nd w
indth
row
risk.
Landscape ecosystem-level ecosystem: It w
ill al
low
the
use
r to
explo
re a
lter
nat
ive
VR system
s by p
roje
ctin
g the
spat
ial
and
tem
pora
l dev
elopm
ent
of
com
ple
x
cut
blo
cks
crea
ted by par
tial
har
ves
ting in
ar
eas
rangin
g
from
20 to 2
000 h
aFORCEE
FORCEE
Hybrid model:
It use
s FO
RECA
ST to
es
tim
ate
key
ecosy
stem
pro
cess
es (light &
nutrie
nt co
mpet
itio
n).
Spatially-explicit individual tree simulator: It sim
ula
tes
indiv
idual
tree
s an
d pla
nts,
thei
r sp
atia
l position an
d
effe
cts
on light, fo
rest floor
and nutrie
nt
avai
lability,
and c
reat
es a
lig
ht an
d soil "fo
otp
rint“
for ea
ch tre
e.
Ecosystem-level ecosystem:
It sim
ula
tes
inte
ract
ions
bet
wee
n
diffe
rent
ecosy
stem
co
mponen
ts
as
tree
s,
under
story
, bry
ophyte
s an
d so
il an
th
e in
fluen
ces
of
diffe
rent m
anag
emen
t pra
ctic
es o
n them
.
FO
RCEE u
ser in
terfac
e
Possible
Possible
Forest
Forest
Futures (PFF)
Futures (PFF)
Hybrid model:
It
use
s
FO
RECA
ST
to
estim
ate
ecosy
stem
key
pro
cess
es.
Non-spatial
watershed-level
model
: M
anag
emen
t sc
enar
io
anal
ysis
tool
for
educa
tion,
exte
nsion
and
man
agem
ent
gam
ing.
PFF u
ser in
terfac
e
ForW
aDy
ForW
aDy
Stand-level model
: Sim
ula
tion of
hydro
logic
al fluxes
and
gen
eral
ized
en
ergy
bal
ance
. M
ulti
layer
ed
repre
senta
tion o
f ver
tica
l fluxes
in a
dai
ly tim
e step
.
New
hydro
logic
al
subm
odel
for
FO
RECA
ST,
with
feed
bac
k o
n g
row
th rat
es a
nd d
ecom
position rat
es.
Dev
eloped
using the
Ste
lla®
model
ling fra
mew
ork
WHY HYBRID MODELS?
WHY HYBRID MODELS?
Model output is only valid if:
The future growing conditions are sufficiently similar to those
that existed during the development of the sample stands
on which the models are based
yield
time
Past
Present
Better growing
conditions
Similar growing
conditions
Poorer growing
conditions
harvest
Future?
Models under a changing future
Models under a changing future
‘Histo
rica
l bio
assa
y’
model
s bas
ed o
n e
xper
ience
are
val
id f
or
the
spec
ies
involv
ed a
nd the
par
ticu
lar se
t of bio
tic
and a
bio
tic
gro
wth
conditio
ns
that
per
tain
ed
over
the
per
iod o
f gro
wth
. H
ow
ever
, if c
han
ges
in f
utu
re m
anag
emen
t re
gim
es o
r
hum
an im
pac
ts on th
e bio
physica
l en
vironm
ent
signific
antly al
ter
futu
re gro
wth
conditio
ns, the
pre
dic
tions of th
e bio
assa
y a
re u
nlikel
y to b
e ac
cura
te.
Pro
cess
-bas
ed m
odel
s em
piric
ally
der
ived
from
re
lationsh
ips
bet
wee
n a
series
of in
dep
enden
t var
iable
s an
d tre
e gro
wth
hav
e gre
at h
euristic
val
ue,
but m
ost o
f
them
are
not ec
osy
stem
-lev
el m
odel
s an
d a
re r
arel
y u
sed in p
ract
ical
applica
tions
in
fore
stry
, prim
arily bec
ause
w
e do not
know
en
ough ab
out
the
key
ec
osy
stem
pro
cess
es a
nd thei
r in
tera
ctio
ns to
mak
e ac
cura
te p
redic
tions.
A third, ‘h
ybrid’
appro
ach h
as b
een d
evel
oped
whic
h a
ttem
pts to c
om
bin
e
the
stre
ngth
s of th
e oth
er tw
o a
ppro
aches
and ther
eby c
om
pen
sate
for th
eir in
div
idual
wea
knes
ses. T
hes
e m
odel
s ta
ke
the
yie
ld p
redic
tions from
a h
isto
rica
l bio
assa
y m
odel
or ra
w fie
ld d
ata
and m
odify them
acc
ord
ing to a
sim
ula
tion o
f th
e te
mpora
l var
iation
in a
vai
lable
lig
ht an
d n
utrie
nts (K
imm
ins et
al. 1
999).
Projection
Projection
Interpretation
Interpretation
Forest
Forest --level Timber Supply Model
level Timber Supply Model
(ATLAS)
(ATLAS)
Wildlife Habitat Supply Model
Wildlife Habitat Supply Model
(( SimFor
SimFor ))
Polygon
Polygon-- Based
Based
Raster
Raster --Based
Based
HIERARCHICAL DECISION SUPPORT SYSTEMS
HIERARCHICAL DECISION SUPPORT SYSTEMS
Select
Select
treatment
treatment
Stand
Stand-- Level Model
Level Model
(FORECAST)
(FORECAST)
Stand
Stand-- Level Visualization Model
Level Visualization Model
(SVS)
(SVS)
Landscape
Landscape-- Level Visualization Model
Level Visualization Model
(CALP Forester)
(CALP Forester)
Non
Non-- Spatial
Spatial
Visualization
Visualization
Individual trees represented
Individual trees represented
Links between models
Links between models
The
multi-re
sourc
e nat
ure
of
moder
n fo
restry
dem
ands
that
m
anag
ers
asse
ss the
pote
ntial
im
pac
ts o
f th
eir
dec
isio
ns
on a
bro
ad r
ange
of
fore
st a
ttribute
s
rela
ted to
bio
div
ersity
, tim
ber
pro
duct
ion,
carb
on stora
ge,
re
crea
tion an
d oth
er
val
ues
.
The
hie
rarc
hic
al stru
cture
fa
cilita
tes
pro
ble
m an
alysis
acro
ss diffe
rent
pla
nnin
g lev
els
(i.e
tact
ical
vs. s
trat
egic
) by a
llow
ing for th
e ad
ditio
n o
f co
mple
xity
wher
e w
arra
nte
d an
d nec
essa
ry.
More
over
, th
e m
odula
r ap
pro
ach al
low
s fo
r
incr
ease
d f
lexib
ility w
ithin
a D
SS a
s it f
acilitat
es t
he
use
of
diffe
rent
model
s to
addre
ss spec
ific
pro
ble
ms or ec
osy
stem
types
(See
ly e
t al
2004).
Individual trees or stands represented
Individual trees or stands represented