HomeMy WebLinkAboutWildfire in SO:Burning Issue
Wildfire in SW
Oregon:.
A Burning Issue
April 19-20, 2001
Red Lion
Medford, OR
Sponsored by:
SS?c~ety of American Foresters ....,
Iskiyou Chapter
Oregon State University Extension Service .
.".eo.. n......
UNrvCIIWrTY
.XT8N..... ..,""oe
__,_=~~~,~...-......_...,._,,_".~~__. ..,u~_"",, ..,._.,.~ ,~~_,
WILDFIRE IN SW OREGON
Program Agenda
Thursday, April 19, 2001
8:00 Registration
8: 3 0-9: 15 Historical and ecological perspectives on the wildfire issue
Tom Sensenig, BLM
9: 15-10 :00 Effects of fire exclusion on insects and diseases
Katrina Marshall, USFS
10'00-10:15 BREAK
10: IS-II :00 Effects of fire on riparian and aquatic ecosystems
Robert GressweU, USGS
11:00-11:45 Comparing the effects of wildfire, timber harvesting, prescribedfire, and
roads on watersheds
Walt Megahan, NCASI
11:45-12:15 Q & A session with morning speakers
12: 15-1 :30 LUNCH
1 :30-2: 15 Landscape level strategies for forest fuel management
Phil Weatherspoon, USFS, PSW Research Station
2: 15-3 :00 Living in a fire environment: Fire hazard reduction strategies for the
homeowner and rorallot owner
Dennis Turco, ODF
3 :00-3: 15 BREAK
3: 15-4:00 What to do with all those small trees?
Blair Moody, BLM
400-4:45 The power situation in the Northwest
Monte Mendenhall, Pacific Power & Light
--. t"
WILDFIRE IN SW OREGON
Program Agenda
Friday, April 20
Friday, April 20
FIELD TOURS
All tours depart at 8:30 am from the designated meeting location and will return at
approximately 4:30 p.m. or earlier. Please arrive at least 20 minutes early to allow
adequate time for parking and check-in. A box lunch and refreshments will be provided
on all tours. Tours will travel by vans; some walking will be required. You need to bring
appropriate footwear and clothing. Tours go rain or shine!
Tour #1
Reducing Fire Hazard with Active Management Using Commercial and Non-
Commercial Techniques
Tour meets at: BLM office, 3040 Biddle Road, Medford. Park in rear, area will be signed
and flagged.
Tour #2
Fire hazard reduction strategies for the homeowner and rural lot owner
Tour meets at: Oregon Department of Forestry office, 5286 Table Rock Road, Central
Point.
Tour leaders: Dennis Turco and Jim Wolf, Oregon Department of Forestry
Tour #3
Fire ecology, fire effects, and post-fire rehabilitation
Tour meets at: Siskiyou National Forest, 200 NE Greenfield Road, Grants Pass, Parking
area will be signed and flagged.
Tour leader: Dr. Tom Alzet, Area Ecologist, Siskiyou National Forest
\.
"~............-.---,.-"
(
l_
Summary
Fire History, Development and Current and Past Growth, of Old and Young Growth
Forest Stands in the Cascade, Siskiyou and the Mid-Coast Mountains of Southwestern
Oregon
Thomas Sensenig and John C. Tappeiner II
Introduction
Over the past decade forest management objectives for federally managed forests have evolved from an
emphasis on timber production and reforestation to a more comprehensive approach to forest resource
management. This involves the need to grow and produce forests of greater complexity and biological
diversity, including forests possessing old-growth stand structure. Data on the historic development of
old-growth forests and the potential for younger stands to develop old-growth characteristics is generally
lacking.
Forest managers throughout the region have been tasked with managing forests for a broad range of
values including growing and maintaining large trees and diverse structures. Currently, about 40% of the
federally managed forest stands in southwestern Oregon are <40 years of age and nearly 60%, < 120yrs,
As older forests are reduced in size and quantity by logging, mortality and fire, old-growth habitat related
species may correspondingly diminish, Additional reductions as a consequence of reduced habitats
resulting from ongoing forest fragmentation is also likely. Thus it is important to learn more about
developmental patterns of old-growth forests, in terms of frequency of disturbance, species composition,
tree growth and structural characteristics so that silvicultural systems can be developed to manage young
forests toward replacement of old-growth forest ecosystems and existing old-growth to maintain stand
structure, vigor and resiliency.
Methods
A total of 21, 8ha former old-growth stands (clearcuts) and young stands in three forest types were
chosen for this study; 6 stands in the mid-Coast Douglas-fir/tanoak forests, 6 stands in mixed evergreen
forests in the Siskiyou mountains and 6 mixed-conifer stands in the Cascades. We also sampled two
Shasta fir and one sugar pine stands in the Cascades, Three 0.0 I ha plots were installed in each old-
growth c1earcut site. In addition uncut old-growth stands and young stands (60-120yrs) proximal to each
old-growth site was sampled. We evaluated fire intensity, frequency and forest stand development at each
sites. Because we also had a particular interest in the growth of the largest trees data from "big trees"
were also collected opportunistically from stumps of Douglas-fir and ponderosa pine encountered outside
the randomly located plots. A total of 422 were examined, including 149, 136 and 137 in Cascade,
Siskiyou and mid-Coast strands respectively.
All of the stumps within each 8ha (20 acres) old-growth clearcut plots were examined for fire scars,
Where stumps recorded multiple fires each fire was individually dated and a record of the fire interval in
that tree was recorded. The stump height diameter at the time each tree survived its first fire was
measured in cm on each tree recording multiple tires,
Results
Fire was common during the development of old-growth stands in all forest types, A total of 1,262 fire
scars were found on all sites (Cascades mixed-conifer 441; Siskiyou mixed-evergreen, 443; mid-Coast
tanoak 319, Shasta fir 38, sugar pine 21), When fire scars were detected the date of each fire event was
calculated. Because 612 scars dated the same fires, a total of 650 separate fires were recorded. For all
fires in the Cascade, Siskiyou and mid-Coast stands 52%, 76%, and 67% respectively were dated to the
same year. Apparently there was little firc on these sites after 1900. Of the 650 fires detected on 21
sites, only three were dctected after 1900, including one in the Cascades at Shell Peak in 190 I, one in the
"~.-------r~'--~-' ~_",,__"'M_' ~____,.., .__,.,.__,,_.
SisklYous at Yale Creek in 1909, and one in a mid-Coast in 1903 on Big Windy Creek. The oldest fires
detected in the three forest types occurred in 1432, 1623 and 1604 on Hoxie Creek in the Cascades,
Skunk Gulch in the Siskiyous and Rum Creek in the mid-Coast respectively. There were no significant
differences in the number of fire scars detected in 8ha stands among the three forest types (p=0.61 0),
Additionally, stand elevation did not account for differences among sites in any forest type (Cascade
p=0.440. Siskiyou p=O.267. mid-Coast p=1.000), or between forest types (p=0.710).
Fire frequencies ranged from 7-13yrs in the Cascades (ave. llyrs, SE 0,9), 9-14yrs in the Siskiyous
(IOyrs, SE 0.8) and 9-l6yrs (llyrs, SE 1.2) in mid-Coast stands. The Shasta fir stands ranged from 18-
26yrs (ave. Ilyrs, SEI.2). The sugar pine stand had a grouped probable fire frequency of 14yrs (SE 2,8),
An ANOV A detected no significant differences among composite fire intervals grouped +/-3yrs on all
sites in all forest types (P=0.933).
Others sites in the Siskiyous and mid-Coast also had short fire intervals but considerable variations, For
example Yale Creek had an average fire frequency of 14yrs but ranged from 5-35yrs, and Big Windy
Creek had an average 16yrs and ranged from 4-38yrs.
In stands where fires occurred frequently during stand development the length of the periods without fire
are as important as the frequency of fire. Presumably seeding mortal ity was high during periods of
frequent fire yet during periods without fire seedling establishment was higher increasing potential
recruitment and stand density, Also, during periods of less frequently fire the abundance of less tolerant
species increased, In the Cascades, the longest fire-free period ranged from 9yrs at Soda Meadows, to
30yrs at Hoxie Creek, in the Siskiyous from 12yrs at Glade Creek to 35yrs at Yale Creek, and from 11yrs
at Rum Creek to 51 yrs at Dulog in the mid-Coast. The Shasta fir stands had the greatest range of
variability having the longest fire free periods from 89yrs and 56yrs at Vulture Rock and Griffin Pass
respectively, The longest fire free period the sugar pine stand was 28yrs,
Fire occurrence viewed in terms of proportions of decades with fire suggest that fires were quite common
in all forest types from 1700 -1900. Since 1700, for each forest type the average percent of decades in
which fire was detected, based on the grouped fire occurrence data was 67 (SE 8.50), 68 (SE 7.35), 63
(SE 4.55) in Cascade, Siskiyou and mid-Coast stands respectively.
Because smaller trees are more susceptible to fire damage and consequently mortality from fire, the
diameter of the surviving trees is an indication of fire intensity at that point. The size of point sample
trees at the time of first fire also confirms that small trees often survived fires. At the Siskiyou sites 73%
of these trees were < I Ocm dbh at time of first fire and 42% and 60% at the mid-Coast and Cascade sites
respectively. In the Cascades, of the 51 point sample trees surviving multiple fires, 34 (64%) were less
than IOem in diameter at stump height when encountering the first fire. In the Siskiyous, 29 of the 43
(67%) and mid-Coast 19 of 45 (42%) point sample trees were less than IOem at the time of first fire. The
smallest diameter point sample trees surviving first fire were 5.8cm, 5,8cm and 7,6cm in the Cascades,
Siskiyous and mid-Coast stands respectively, The diameters at the time of first fire in the two Shasta fir
and sugar pine point sample trees were 41 cm, 56cm, 8cm respectively,
For all sites 66% of the decades from 1700-1900 experienced fire during stand development (SE 3,9)
and 56% of the decades recorded recruitment into the stand during the same period, indicating that all
sampled trees became established during a period of frequent fire. Because fire events were generally
more frequent than recruitment events, a direct relationship between individual fire events and
subsequent recruitment cannot be drawn.
Developmental patterns of the old-growth trees developed statistically dissimilar to the actual young
stand trees in all forest types. An observation of particulars concern is that decadal increment growth of
-2-
the old-growth trees, remains well above young stands growth through 8 decades in all forest types
studied.
Even though variation in the amount of ir:dividual tree growth within stands was observed a definitive
pattern of tree development is apparent. As depicted in ordination young stands, although exhibiting
relatively high growth rates in early decades, rapidly declined and trees that grew slowly initially
remained slow growing throughout all nine decades. This was not observed in the older stands, in fact
slower growing trees early in the decade often surpassed initially fast growing trees,
Discussion
All old-growth stands on all sites throughout these study areas displayed evidence of frequent fire during
old-growth stand development. For all sites 62% of the decades between 1700 -1900 experienced fire,
An extremely important finding is that there were no scars that were dated after 1903. Since about 1900
virtually all fires have been suppressed rendering most old-growth stands free from disturbance by fire
for an unprecedented period of time.
The results of the study suggest that fires during the period 1700 in 1900 were not intense fires, During
this period seedlings were established and grew into trees with diameters >30cm, However, during the
period especially from 1700 to 1850 a large proportion of the trees were less than 10 and 25 cm in
diameter, Furthermore, 57% of the trees with multiple fire scars recorded their first fire when they were
less than 10 cm diameter and nearly all trees recorded their first fire when they were less than 25 cm
diameter. Thus the fires on these sites did not result in stand replacement. The fact that fires were so
frequent suggested fuels were probably kept at low level, thus reducing the likelihood of intense fire,
These results also have implications for the understanding of forest stand development and stand
dynamics. These forests were the result of rather continuous tree establishment for 1700 in earlier until
1900, under conditions of chronic disturbance by fire. This is a quite different pattern of development
than one where fire, or other disturbance, eliminates an entire stand and regeneration occurs promptly
thereafter resulting in a relatively uniform age/size distribution of trees,
Prior to fire suppression and in the absence of artificial barriers, such as roads, farmland, and European
settlement fires burned continually across the landscape throughout the entire dry region of southwestern
Oregon, Fire redefines stand characteristics and biological processes with each event by determining:
which trees and species will continue to survive and which ones will die; the distribution and density of
surviving trees; the structure, composition, distribution, abundance and species of vegetation and
ultimately what habitats will result. Therefore, wildlife and other organisms that old-growth forests
sustain throughout southwestern Oregon are a consequence of the historic fire environment during stand
development. This analysis confirms that significant differences in the developmental patterns of young
and old-growth stands exist. t is worth noting, that because sampling was confined to Douglas-fir forests
in the Siskiyou mountains of southern Oregon, the area of inference is rather limited, however, the
implications are particularly important.
It is likely that the currently abundant young forest stands may likely to develop much different stand
characteristics than those of old-growth stands. In many young stands, major problems associated with
diminished tree vigor such as insect infestation and competition mortality are probable. Observations
made here should also lend importance to the potential value of early stand intervention and manipulation
such that developmental trajectories can be altered to more closely pattern historic forest development,
and by so doing, provide structure and habitat for associated species.
-3-
""r-'---' '
....<11
......
cSl
.
N ~
.
i
.
.
.
.
....
1lO
~
. ,.
.
.
.
...
100
I:
.
....$
......
CS>
II>
N ~
.
...
......
lit
lIP
I
.
.
.
.
.
...
\0
I
.
.
,
.
.
eft
...
Of .
:€
i1
~
i'
I
I7t
....
oa .
&
-<
.
ii
i
"'<5>>
......
lIP
CIl
to
....
ClI
~
..
10
Z
...
Of
~
..
'"
.
~
N ..
lIIl.II1N of FIN lQre
eft
...
.. .
.
.
I
. . .
I
.
N ..
.
.
. .
..
.
.
.
.
.
.
.
.
.
.
01
...
.. ~
....IV
......
..
.
~
N ..
.
.
.
. .
.
.
. .
. .
.
~
.
.
.
N ~
.
.
.
. .
I
. .
.
.
LV ~,_,_'.,.'.'">,_,,,_'-__~~__,,"_",~'_""""~""'_"_'~.__,.~. ....._.~,..~"~ . IJIW"I
&
.
.
.~ .
,..
to)
i
.
.
1
1
.
I
....
Clll
I
....
ID
:
.....
-..I.
I
.
...
'"
.
.
I .
I
.
.
. .
eft
...
.. .
&
· C)
. i'
t
i
.
.
,.
.
..
CIl
....
IlII .
!.
~
c
i
C)
j
~~ ~"Z
CD Q.)"m C
(/) c;; en 3
(") 0-
~ ~. -; ~
:E ::r ::r 0
Q) CD_
eng.::l::n
c.-cm
0)~3cn
- 0-
CD "'0 CD (")
a.~."" ~
:E.1:l S'(/)
_...., -
:T6'50
S-~CD3
+-...Jc~
- 0"'0 ..:..
I 0"'0 ~
(J.) - (!) 0
'< ...,0
UJ>roo
~;::->.
ooco
:E-o
en roo
-.0) a.
6.(")ro
_.:::1"_
(") coro
Q.) ..., (")
......0)......
CD "'0 ro
:=:::1" a.
3 -- 0-
ro::l'<
en 9: a.
<:(0)
<;0)......
:::1"-ro
roCDa
..., en ....
CD........J
:Ten
3 CD x'
Q::l(f)
CD c: --
......3cn
:To-C:
~~~
o 0 r:::
::l-
CD
a.>
__C
en 0 0
CO '- C
o (1) '"
Y.CI..c
:QE-;
E ::::J ....
C 0
.~ (1) E
en..c
c ..... (1)
o en Q;
Q)
(1).......c
ro~~
-0=0 en
(1)
E'~ E
-o..c:.=
Q) 0. Q)
_CO.....
u'-CO
Q)Olu
.... .J::. 0-
Q)o"O
"Ococ
o (1) en
O_~
0)00
...-- .t:::: .
.9~<(e
o '- >-
o Q) . ('l')
"- 0.0 I
,- 0.0 -
"'--::1/"-+
E ...-- c::
Q) 0'-
e~....~
~ L.--
en.5 9- ~
'- .... 0.... -0
co Q) 0. Q)
0..0 _
en en co
," E (1) "0
.....::J....
t2ccoen
"0 co
- Q).r:. ::>
o.r:......>
Q) r- o~ ro
.CI 0
Een~en
Q) CO Q)
::J_.u '-
Z '(i) en to::
]
~
a:
..
o .
...
~
1
g
CIl
l
L.--L..
Q .
...
.
.
. .
.
. .
.
.
.
.
.
.
.
.
J
. .
.
.
... 01
g
...
0...
.
.
. . :
.
II
.
.
.
.
.
. .
.
.
... N
8
.
...
o
I
o
...
o
...
J
!
~
.g
t
~
&.
.I::
~
'J
.
.
.
.
.
.
.
.
.
.
.
I
.
.
.
.
.
.
.
.
o
o
...
.....
...
...
.
.
.
.
.
.
.
.
...
.,
o
...
0...
01
.... ~ 1.-aumH
..
..
.
...
8.
III
..
~
.
..
~.~ TW
1
r
"3
Q
~
o
-
f
i
~
III
~
o
...
.
-
.
.
..
.
.
.
.
..
8-
.
...
.
.
.
.
.
!
...
.
~
.
.
...
.
I
. .
!
.
.
o
Ie
.
...
.
.
.
.
.
.
ft
i
if. ·
"'0
..,
o
o
...
"'4
~
.
.
.
-<
I
...
.
8
N
.
.
.
.
.
.
N
.
.
.
.
.
.
.
..
~
.. 0
i
I
.
...
. 0
~
~
r
i
+
~
VI
"0
..,
o
o
...
..,
VI
o
..
. .
8
tunMr 0' RNIIQM
~ .
.
.
.
.
.
. .
.
.
N .
.
.
.
.
.
.
.
.
.
.
.
.
.
...
. 0
III
l-
I
~
. 0
N
.
i
.
.
.
.
f
t
.
.
.
. .
.
...
..
g
.
.
. .
.
.
w~ ·
8
. .
.
&
"'0 N ..
:~: :
I .
. .
.
.
.
. .
.. .
.
.
.
.
.
.
.
. .
.
.
...
.
g
.
-
.
...
. a
-+
...
...
i-
I:,
.
It
i
.
...
. a
I
I
&
.
J:
I
f
:=J(J')enz
CD g'~ c
en en CD 3
(") 0-
~ ~;1~
~=rCDa
~fr~:=J
a.-cm
tl> m 3 CI)
- 0'"
(l) "'0 CD (")
0...,.,0)
< -5 -. en
< ..., ::J
;:::;:--~
:r. 0 CD::I" 0
::I --. 3
+-.,JC--.
- 0"0 -J
I 0"0
w. CD 0
'< ...,0
CiJ>roo
. ~ ~ ~
00(0
~-o
Cl)coo
--0> a.
~ 0 (l)
Q::::r.....
(") CO (t)
ID ..., (")
-ro-
(1)"0 (t)
~:::ra.
3 -. 0"
co~'<
009:0-
<00>
<0>-
::r-CD
(1) CD 0
., en :J
(1)_
'::I"CI)
3 CD x.
Q~()
ctl C 0)
-3(J)
:Tc:ro
ID CD 0)
~-'Cl.
o 0 CD
:J""'"
CD
Q)
'0
co
U
U'J
U
U'J
co
U
c
o
01
Q)
'-
o
c
'-
Q)
Ul
Q)
3
.J::.
"5
o
U'J
c
U'J
'0
c
co
Ul
::;
u
'-
co
Q)
(3
co
.c.
co
..-
N
......
o
01$
c U'J
a.~
E.E
co-
C/) C/)
C~
00
- I
.~ '0
.c. '-
Q) E
~~
a co
UJ :J
_ 0
>.
:J
~~
O:::Cf)
CIl
'0
c::
nl
-
en
.c:
i
o
...
Cl
.
'0
o
c::
Q)
...
LL
Cl
c:
'0
...
o
(J
Q)
e:::
CIl
Q)
Gl
...
I-
<
E
o
...
LL
~
o
-
.!!
:x:
Gl
.=
LL
-
o
~
nl
E
E
~
en
"tl
III
Q.
~
o
...
C>
~ ~
~ 0
o ~
;-0::
III
Q.
>-
I-
0;
a:
a
u..
.. "I
] ~I
u ~
~ 'E'
- 0
o u
-;. "
0::
1?~8
11 -'- "
(J)" -I,{) 0 I,() l,()l/} l,{)
IS "0 5""'" S l.O l.D <.0 M
o ~ ~
.... 0 u.
o
'"
8tl)0'l"ltta<O<D
~ ~ ....- ;:. ;:: ;;;. ;.
E--.----
00""' 0...," (',; O'j
LtC"4-MNNN
1? '"
iL'g
-1;
&0..
~ a:
..J ~
MU.
~~~g~~~
- I I , 6 ' t
E ,tH" <0 _ 0"1'
ej..1-:"h;".....
u. I
I
co
.5 l!
"E :l
O(ll
~.!
· c
e iil
...
..
~ r!
l~
...
E
8
"
9
~
...
'c
8.:0( ~..lt CD..ao:: ~
:Rg:R~p
iluilluo"-u
.J:!~':;~tI~~
~Cl~~dlbj~
-8 rn
..
u
<II
..
c..l
....on
<D
. a:>
~w
:;rn
l,()l,() I,(') OQIl)
I,{),..............,a>r--
.0
"''''
:i'
0::'"
('\j.........,O'l<ofM
......... -. -.N._.
~~~~~~
l,(),n.C'ilti,....:Lr).
.....NNNNM
}:ci
<D-
~~
"
!22COMNr--N
fi')aHr.co~,....
I ' f I I I
11~7~9
N...........""iI')N
,_..-......... NN
!!
>00>
::0
.w
~rn
"
~~~~~~
t I 666~
crcr,rt,
"""'M..,.IOW"J
~...,
,,-
~~
"
~~~~~~
H~++~
~-__""N
",N
~!D""'Q)C"')""~~~
"rn
:it...,
~:;i~~;~g~:
10m
M'"
...,. ,.... M""tci
~~~~SS;;"'t'......
"'>w
"rn
",a:>
,-Ma)
""0)1,O""'Q)CI)o ....
IIlMNMN"Nfw
"en
c::
GI
e
:i<
<II
Cii
0;
B
:lM
."
"w
:;rn
..'"
co'"
c: I
"N
0::_
iN
",'"
.w
~rn
"
!!
t:~
_0
,w
~rn
..
t...,
00
i~
"
I,()\{') Ol.O 100
'"f~r-..l.Oll')tO
N~"ltaJ""'O'l
~....._.........._----
.....Nl.DaJ..nO
NN....__N
;:;~~-~.~.~
m....NcnOI()
MCC\n<q'-.:t~
67/';'17
7~~~)~~
0' I,.... I I
N~~""::~
in~~~NCO
'cb~~~~
<r'71/11
""'''''''''\l)''''''''
:;:ll~:<lO:;i1l
,!. I I I , ,
....cp";";'rcp
.!,........~........
...."l
Nr--M,....:!~~~~
"'"
N...
NMo)t--\n,....~~cD
....M...M("}Ll1...tw
.rn
"'0
cD....<OLnlXJMQ)~ai
....r.nl/')r.o...lX)M~u.r
"rn
....'"
... ij) a) ,.... a) In ;Z ~ M
"NNNMM_fw
"en
...
~
III
.!!!
~
o
C x><
==~~e~~~
O!!uu~:Rw
c:u >o"uu"i
:!8'~e€E~
.;;::e>E~E
- Q.2' il rn '" rn
a lIlrn
y
"tl
:i
1i
.9
"''''
<D .
....
.w
:;rn
~'"
:iM
.w
~rn
"
.'"'
0'"
c: I
~;:
!l.:R
Ii t
0::'"
~'"
~,..,:
.w
~'"
..
~'"
~cci
.w
~rn
"
tN
- .
~~
~rn
"
!'"
_0
.w
!:rn
"
,k
~~
"
t~
"'0
~~
"
0'"
. ....: 0)
~ UJ ~
"en
"''''
""..nco
~UJ~
..rn
....0
c.)cir:-1
~:;;~
"rn
~<O
<f-
ew;;;
"en
i
...
.
f
o
...
f
t=
...-r-.-.---.,...-.....
"
B
;,
~
'"
III
E
...
..
; g ~
- a: "-
C e c
SE~
<II ~ c.:>
..>
-; ~-~
JlIS
4> .. '"
"tl....
B~~
<II
..
c..l
00
"'...
"...
Me'
N "'
iii:R
"'a:>
<ON
~:&
I I
<0 a:>
N _
I I
<0'"
0><0
'j"f'
"'-
N-
I ,
<ON
~o
_....N.O
>w
"rn
....0
;:~;S;::
~rn
:!!o
M It) co . CD
...."'r"JIlDW
~en
No
Q)f.O."':C'i
"N~w
"",
"
Ii
..
!;.o
~'"
.UJ
~cn
~I
~'"
",a:>
c: ,
~::
~
~
~.
~'"
<0"
...-
~~
..
<0
'"
I
7
';:
!!
>00
N .
N'"
.w
~:J)
"
a:>
N
I
...
.h
to
=~
.w
~rn
"
<0
N
I
0>
I
"'~
on....
--
..
'"
0>
~
"0
u
..
><
No>
N~ e
- 0
E
o
o
!
ti
Ill~ 'j
<D <II
~
rn
Stand Fire History
20 decades (100"10)
with fire 1700-1900
Cascade
Soda Meadows Old-growth
Individual Tree Basal Area Growth
Per Decade
Before
f
~
138 trees/ha
33.4 m2/ha
basal area
E
u
0-
III
.&:
~
o
...
Ol
'"
ell
"-
III
iii
III
~
'1
Fire
Cessation
Post Fire
Cohort
After
148 trees/ha + 7%
78,5 m2/ha basal area +
f---- --- -~.
Cascade
Soda Meadows Young Stand
Individual Tree Basal Area Growth Per Decade
III
~ 1000
900
800
0- 700
toO 600
i 500
~ 400
:: 300
~ 200
III
iii 100
::I 18~r
..c
E
u
1900
year
Currently
1850
Fire Cessation
1005 treeslha
50.8 m2/ha basal area
I
l___~__
--
1000
<f)
3- 900
800
700
g- 600
~ 500
g. 400
;1l 300
~
Iii 200
~ 100
.0
150%
1550
1600
1650'
1700
1750
year 1800
1850
Siskiyou
Glade Creek Old-growth
Individual Tree Basal Area Growth Per Dee,
Before
'-
S
272 trees/ha
19.89 m2/ha basal area
Fire Cessation
Post Fire
Cohort
0.1%
of final
basal
area
After
Stand Fire Hisrory
15 decad9s (75%)
with firo 1700 . 1900
313 trees/ha + 15 %
48.3 m2/ha basal area + 142%
Siskiyou
Glade Creek Young Stand
Individual Tree Basal Area Growth Per Decade
f
:... 1000
0
~ 900
E 800
u
c- 700
.. 600
I:
~ 500
e 400
co
<<I 300
i e 200
<<I
Ii 100
III
<<I 1880
.Q
~./'
Fire Cessation
Currently
1170 trees/ha
26,18 m2/ha basal area
I
-~
-'~"-"~"~---r--<-""_~_O___" ......-_........____"
~
~ 1000
900
800
c:T 700
., 600
.c
~ 500
o 400
tiI 300
g: 200
...
., 100
-= lScR> ,r, .
~ 1550
1600
1650
1700
1750
year 18
1850
Mid-Coast
Rum Creek Old-growth
Individual Tree Basal Area Growth Per
Before
E
o
102 trees/ha
26.9 m2/ha basal area
Fire Cessation
After
Stand Fire History
11 decades (55%)
with fire 1700-1900
/
102 trees/ha
43.9 m2/ha basal area + 63%
----------1
I
Mid-Coast
Rum Creek Young Stand
Individual Tree Basal Area Growth Per Decade
~f 1000
~
.... 900
----
E 800
u 700
c:T
to 600
.c 500
~
... 400
CII 300
'"
f 200
'"
t; 100
to
III 186\:1'
..a
Currently
-----.
Fire Cessation
527 trees/ha
23.29 m2/ha basal area
""~_' _"_..__."_~_~.....M~.._._...,.__~" _ -'"__. ...
, Proportions of small trees, < 10 and <2Ocm diameter, by 50 year periods from
1700 through 1900 at six sites each in the Cascades, mid-Coast and Siskiyous.
Siskivous
<10 <25 <10 <25 <10 <25 <10 <25
100
90
8 O.
70
60
50 ~$
40
30
20
10
0
Cascades
t.o t
90
. 80
.
0 70
...
t-
- 60
C
0 50
U
... 40
.
Q. 30 ~
20
10
100 Mid-coast
90
80
70
60
50
40
30 ~
20
10
1700-1750 1750-1800 1800-1850 1850-1900
>'--"-~---r-''''--~~"-'^'''''^'.'.'~''-'''_._.'~-'.'~-'--''
E5
~
i4
o
0,3
C'G
~2
...
Q)
0)1
C'G
...
Q)
~O
Comparison of average radial growth of old-growth stands measured at
stump height and young stands measured at dbh and adjusted to stump
height for. three forest types in southwestern Oregon.
Cascade
-
x~
" ------..
~-~-- -_.a_______
_____ x---.
:x "---...
x ______
x x~
- ------ -
-
-
-
- -~!
i
----j
123 4 5 6 7 8 9
10 11 12 1 3 14 15 16 17 18 19 20
Decade
--- old trees stumo helaht --- vouna trees stumo helaht x vouna trees breast helaht
-s
E
u
-
J:4
'i
o
0,3
iti
~2
...
Q)
0)1
C'G
...
Q)
>
<0
Siskiyou
~ ~ ------.,
x" " "----
/~x
/ "'"
/ "'
- x:
------------
'~.
---
x ----~_
x --__...
x~
________ _J
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Decade
__ - old trees stump height ---- young trees stump height x young trees breast height
E5
(,)
-
i4
o
...
C)3
C'G
~2
a:::
Q)
0)1
C'G
....
Q)
~o
mid-Coast
x ~-----
~;-;----,
x"-_
x ______
x --~____
x x
------------- I
I
I
9
14 15 16 17 18
I
19 20
10 11 12
Decade
13
2 3 4 5 6 7 ~
___ old trees stump height ---- young trees stump height x: young trees breast height
~ I
M
o
(J')
T""'
c
.-
L..
OJ
>
"i:
~
.-
c
'i:
.....
OJ
.!:
...
~
o
OJ
.c
L..
OJ
>
'i:
.!:
...
ca
E
E
ca
~
OJ
.!:
...
c
.-
CJ)
0)
o
...J
.-.>'.'l.~"_~...'._",'''.
. _�.._�..__.___..�.__..��_ .w___ _-�.._._.�.__,,.._______._.___...___..
..
'\~.-
Abstract
Effects of Fire Exclusion on Insects and Diseases
Fires, insects and diseases have had a long inter-relationship in the forest. In many ways they
have similar ecological roles In low and mid-elevation pine and mixed conifer forests, fire was
once the primary agent that controlled stocking levels and species composition, Without fire,
drought and insects are now the primary agents that regulate stocking. This is particularly true of
the pine bark beetles. These beetles are attracted to pines of low vigor; often associated with high
density and drought. Pines, especially large pines, are often desirable for growing stock, structure
and habitat. Pines are also valuable because they are resistant to damage by fire and less
susceptible to root diseases and decay Entomologists in Southwest Oregon have developed local
basal area guidelines for pines to reduce the risk of bark beetle attack.
The impact of diseases has also increased where fires were once more frequent. Diseases
generally act more slowly than insects so their effects are less obvious, though not less significant
over time. Fire is believed to have been the major regulator of dwarf mistletoe distribution and
abu ndance. The increasing complexity of stand structure in the absence of fire is very conducive
to the spread of dwarf mistletoes. Fire exclusion has also led to an increase in the dominance of
shade tolerant species These species are generally more susceptible to root diseases.
The reintroduction of fire will also affect insects and diseases. One effect of concern to managers
is the potential for bark beetles to attack residual trees after prescribed fires, Trees that are under
stress before fires or damaged during fires are the ones most likely to be attacked by bark beetles
afterwards. Pre-burn treatments and burning prescriptions that minimize damage will reduce the
risk of beetle attack. Reintroducing fire into stands that are heavily infected with dwarf mistletoe
may result in more extreme fire behavior and higher levels of tree mortality. Root diseases are
unlikely to be eliminated from sites where they are present because they can survive fires in large
roots, buried wood and large stumps. However, over time, reintroducing fire may reduce the
impact of root diseases by shifti ng sites back to greater proportions of less susceptible species.
Root and bole damage caused by fires and mechanical treatments provides an entry point for
certain root diseases and decay organisms. True firs are so easily damaged and so susceptible to
root diseases, decay organisms and bark beetles that underburning in stands where true fir is a
desired component is not recommended.
From an insect and disease perspective, mechanical treatment followed by burning would be better
than burning alone and better than no treatment at all.
Katv AlarshalI
Southwes/ Oregon Fures/lnsee:! and /)isease ,,,'ervlce Center
April 19, 2001
.
Bibliography
Alexander, Martin E. and FG IIawksworth 1975 Wildland fires and dwarf mistletoes a
literature review of ecology and prescribed burning USDA Forest Service General Technical
Report-RM-14 12 pp.
Campbell, Sally, L. Liege! and M. Brookes. 1996 Disturbance and forest health in Oregon
and Washington. USDA Forest Service General Technical Report PNW-GTR-38I. 105 pp
Cochran, P.B. 1992. Stocking levels and underlying assumptions for uneven-aged ponderosa
pine stands. USDA Forest Service Research Note PNW-RN-509. 10 pp.
Filip, Gregory M and L. Yang-ELVIe. 1997. Effects of prescribed burning on the viability of
Armillaria ostoyae in mixed-coni fer forest soils in the Blue Mountains of Oregon. Northwest
Science 71(2): 137-144.
Flanagan, Pau\. 1996. Survival of fire-injured conifers. USDA Forest Service Fire and
Aviation Management, Fire Management Notes 56(2): 13-16.
Harrington, Michael G. and F.G. Hawksworth. 1990. Interaction of fire and dwarf mistletoe
on mortality of southwestern ponderosa pine. In: Effects of fire management of southwestern
natural resources: Proceedings of the symposium, November 15-17,1988. Tucson, Arizona.
USDA Forest Service General Technical Report-RM-191 : 234-240.
Koonce, Andrea L. and L FRoth 1980 The effects of prescribed burning on dwarf mistletoe
in ponderosa pine. In: Proceedings, 6th conference on fire and forest meteorology, April 22-
24,1980, Seattle, WA. Society of American Foresters, Washington, D,C.: 197-203.
Ryan, Kevin C. 1982. Evaluating potential tree mortality from prescribed burning. In: Site
preparation and fuels management on steep terrain, Proceedings of a symposium, February
15-17, 1982, Spokane, Washington. D. Baumgartner, editor. Washington State University,
Cooperative Extension, Pullman, Washington: 167-178.
Scott, Donald W., J. Szymoniak and V. Rockwell. 1996. Entomological concerns regarding
burn characteristics and fire effects on tree species during prescribed landscape burns: burn
severity guidelines and mitigation measures to minimize fire injuries. USDA Forest Service,
Pacific Northwest Region, Blue Mountains Pest Management Zone. BMZ-97-1. 49 pp.
Walstad, John D., S.R. Radosevich and D V. Sandberg. 1990. Natural and Prescribed Fire in
Pacific Northwest Forests. Oregon State University Press, Corvallis, Oregon. 317 pp.
Zimmerman, G Thomas and RD. Laven. 1987. Effects of forest fuel smoke on dwarf
mistletoe seed germination. Great Basin Naturalist 47 (4): 652-659.
KatJ'Marshal!
Southwest Oregon Forest insect and /)isease Service Center
.IWiIIY, ]()()f
'"
\.
-....- ,.
"
~
\
\.-
.".~
Fire and Aquatic Ecosystems: Broad-scale Implications
Robert E Gresswell
USGS Forest and Rangeland Ecosystem Science Center and Department of Fisheries and Wildlife,
Oregon State University, 3200 SW Jefferson Way, Corvallis, OR 97331
Abstracl.--Synthesis of the literature suggests that physical, chemical, and biological elements ofa
watershed interact with long-term climate to influence fire regime, and these factors, in
concordance with the postfire vegetation mosaic, combine with local-scale weather to govern the
trajectory and magnitude of change following a fire event. Perturbation associated with
hydrological processes is probably the primary factor influencing postfire persistence of fishes,
benthic macro invertebrates, and diatoms in fluvial systems It is apparent that salmonids have
evolved strategies to survive perturbations occurring at the frequency of wildland fires (decades
to centuries), but local populations of a species may be more ephemeral. Habitat alteration
probably has the greatest impact on individual organisms and local populations that are the least
mobile, and reinvasion will be most rapid by aquatic organisms with high mobility. It is becoming
increasingly apparent that during the past century fire suppression has altered fire regimes in some
vegetation types, and consequently, the probability of large stand-replacing fires has increased in
those areas. Current evidence suggests, however, that even in the case of extensive, high-
severity fires, local extirpation of fishes is patchy, and recolonization is rapid. Lasting detrimental
effects on fish populations have been limited to areas where native populations have declined and
become increasingly isolated because of anthropogenic activities. A strategy of protecting robust
aquatic communities trom negative effects of anthropogenic activities and restoring aquatic
habitat structure and life history complexity in degraded areas may be the most effective means for
insuring the persistence of native biota where the probability oflarge-scale fires has increased.
.�.._ _____.... _. _..._.�._._._._.__.,..a...o.,.L.�__.__.�_...,.._.�._
....~'---r-._..."' ..,
\
(.
Abstract
ComparinQ the effects of wildfire, timber harvestinq, prescribed fire and
roads on watersheds
By
Walt Megahan
Prepared for the Workshop on "Wildfire in SW Oreqon: A burninQ Issue
Watershed disturbances including forest fire, timber harvesting and road
construction can all cause changes in site conditions that may influence onsite
hydrologic and geomorphic processes that, in turn, can lead to downstream
watershed responses and impact beneficial uses of streams. Watershed scale
responses of concern include sedimentation, streamflow rate, water chemistry,
and water temperature. Of these, massive increases in sedimentation and flood
flows and their associated effects on channels and aquatic resources are the
primary concern.
Forest fire has, by far, the largest potential for impacting watershed resources.
Impacts depend primarily on fire severity and slope gradient. Severe wildfires
provide more water to the soil surface, remove protective litter and organic soil
layers, destroy surface soil structure and can greatly reduce water infiltration.
The combined effects of overland flow, increased surface and mass erosion, and
increased peak flows can cause catastrophic downstream impacts, Fires of low
and moderate severity usually have minimal downstream impacts. Risks from
forest fire decrease rapidly over time with the sprouting and regrowth of burned
vegetation and are usually minimal after three years.
Forest management activities including prescribed burning, timber harvest and
road construction can also Impact watershed resources. However, effects are
much lower than those from severe wildfire and are manageable with the use of
proper procedures including Best Management Practices.
.__..,-'<..............>-'-~_..., ._.~~....._..~.~-_. _.__.""~.'---~--
.-r
.., ......_~--""....-._.....-.._>--.~---
r
~
,
.t~~
,_ ".,..,..~~........____;_,_.......__",,__,,"''''~'' ""__.._,,..4'_~
h +t f' II VVL)'-;, L~L-. ,'r v / S.' (: f / P ~l/) ~ / lv' <: {, / PIJt/I'll G';I.', Fur
,t ,
c. \ " '-
~1\~I' liEF, f"l.~. f,. ~J. ,.{,~ /(""'('_"(".1 -Y-I!-o<,.i,t-Hl
C. PHILLIP WEATHERSPOON
PaCIfic Southwest RI1Search Station
U.S. For..st Service
Redding. California
CARL N, SKINNER
PaCIfiC Southwest Research Station
U.S. Foro~l Servlce
Redding. California
56
Landscape- Level Strategies
for Forest Fuel Management
ABSTRACT
As a resulllargely of human activities during the past 150 years, fires
in Sierra Nevada forests occur less frequently and cover much less
area than they did historically but are much mure likely to be large
and severe when they do occur. High-severity wildfires are consid-
ered by many to be the greatest single threat to the integrity and
sustainability of Sierra Nevada forests. The cuntlnuing accumulation
of large quantities of forest biomass that fuel wildfires points to a
need to develop landscapa-Ievel strategies for managing fuels to re-
duce the area and average size burned by severe fires. Concurrently.
more of the ecosystem functions of natur31 fire regimes-character-
ized in most areas by frequant low- to moderate-severity fires-need
to be restored to Sierran forests. This chapter reviews past and cur-
rent approaches to managing fuels on a landscape basis and, based
on a synthesis of many of these approaches. proposes an outline for
a potential fuel-management strategy for Sierra Nevada forests.
INTRODUCTION
Prior to concentrated Euro-American settlement in the middle
to late 1800s, low- and middle-elevation forests in the Sierra
Nevada were characterized by rt>latively frequent low- to
moderate-severity fires (Skinner and Chang 19(6). These fre-
quent fires performed important ecological functions (Kilgore
1973). As a result largely of human activities during the past
150 years, including but not limited to fire suppression, fires
now occur less frequently and cover much less area but are
much more likely to be large and severe when they do occur
(Husari and McKelvey 1996; McKelvey and Johnston 1992;
Skinner and Chang 1996; U.S. Forest Service 1995;
Weatherspoon et al. 1992). In aggregate, such high-severity
fires are well outside the natural range of variability for these
ecosystems and are considered by many to be the greatest
single threat to the integrity and sustainability of Sierra Ne-
vada forests. In addition, related human-inJuced changes in
forest structure, composition, and processes (including many
of the functions once performed by frequent fires) are in many
areas so profound that they jeopardize ecosystem diversity
and viability even without reference to severe fire (Skinner
and Chang 1996; U.s. Forest Service 1995).
These concerns are prominent among the issues confront-
ing those interested in the well-being of the Sierra Nevada.
This chapter addresses potential landscape-level strategies
intended to reduce the extent of severe fires in Sierra Nevada
forests and to restore more of the ecosystem functions of fre-
quent low- to moderate-severity fires. As a byproduct, these
strategies offer tools that could contribute significantly to
improving the health, integrity, and sustainability of Sierra
Nevada ecosystems.
To keep the scope of the chapter manageable, we focus on
the low- to middle-elevation coniferous forests of the Sierra
Nevada, on both west and east sides of the crest. Our reasons
include the following:
. These forests rank at or near the top among Sierran veg-
etation zones in terms of overall richness and diversity of
resources and values.
. Twentieth-century fire occurrence in these forests has been
much greater than in higher-elevation forests (McKelvey
and Busse 1996). High-severity wildfires are much less a
concern in the higher-elevation forests.
Sierra ~p.vada ECOf,yst[!ffi Project: FinJl report to C ongn?ss, Jol. II, As.sessments and scientific basis for manaj.;cment options. Davis: University of California, Center~ for
W,lter and Wildland Resources, 1996.
1471
, . ~-r--.""'---"'"'' _o"o,~,,_"".'__.~, ....~-t--... ,,"_.,
1472
VOLUME II, CHAPTER 56
. Based on rE'cords o[ twentieth-century [ire occurrence, the
probability of wildfire in low- to middle-elevation conifer-
ous forests is somewhat less tha..'"1 in the lower elevation
foothill woodland and chaparral vegetation types
(McKelvey and Busse 1996). However, the negative effects
of severe wildfire on the dominant vegetation--and by
extension on numerous other resources--generally are
more profound and more long lasting in the coniferous
forests.
. The composition and structure of the dominant vegetation
in low- to middle-elevation coniferous forests probably
have bee~ affected more adversely by removal of the natu-
ral fire regime (and thus potentially could benefit more
from its partial restoration) than in higher or lower veg'
eta tion types.
We recognize the problems associated with the threat of wild-
fires to lives and property in the urban-wildland intermix
areas in the Sierran foothills. Management of foothill vegeta-
tion is mentioned in our discussion of these intermix areas.
Many of the same general principles and approaches for fuel-
management strategies that we discuss for the coniferous for-
ests apply also to the foothill vegetation lypes.
A CAUTIONARY TALE OF
FOREST BIOMASS
A simplified, qualitative accounti.r.g of production and dis-
position of biomass may help to clarify the problem of fuel
accumulation in manv Sierra Nevada forests. As indicated
earlier, low- and middle-elevation forest types-west-side
pine, west-side pine-mixed conifer, and east-side pine--are
emphasized. It is appropriate here to consider only above-
ground biomass, both for simplicity and relevance to the topic
at hand. While we recognize the impOrLl.l1Ce to today's for-
ests of events in the latter half of the nineteenth century
(McKelvey and Johnston 1992), we focus here on contrasts
between the periods before 1850 and after 1900
Biomass Production
Sierra Nevada forests produce a great deal of biomass. While
considerable variation eXists in terms of the site and climatic
variables that largely determine net primary productivity, in
general terms Sierra Nevada forests are quite productive
(Helms and Tappeiner 1996). For the forest types indicated
earlier, the west-side types are substantially more productive
than east-side pine. The average rate of biomass production
during most of the twentieth century probably has exceeded
that which occurred [rom, say, 1650 La 1850 because this cen-
tury generally has been warmer and wetter than the earlier
period (Graumlich 1993). More complete site occupancy, in
the form of denser forests in many areas (GrueIl1994), also
may have contributed to greater production now than then_
Allocation of total biomass production apparently has differed
considerably between the two periods. A much greater per-
centage of biomass historically was stored in the boles of large
trees and in herbaceous vegetation in relatively open stands,
whereas now much more goes into small trees in dense stands_
Biomass Disposition
The main factors accounting for disposition or removal of
forest biomass are decomposition (oxidation), fire (oxidation),
and herbivores and humans (utilization).
Decomposition
In California's Mediterranean climate, decomposition rates
generally are low, limited by low temperatures in the winter
and inadequate moisture in the summer. In some portions of
the Sierra Nevada mixed conifer forest type, however, suffi-
cient moisture may be retained well into the summer to sup-
port fairly high rates of decay (Harmon et al. 1987).
Decomposition rates in Sierra Nevada forests probably have
been greater during this century than during the period 1650-
1850 because (1) this century has been warmer and wetter
(Graumlich 1993), (2) the generally denser stands during this
century have provided more mesic microclimates that favor
decomposition, and (3) more forest floor biomass has been
available for decomposition because it has not been removed
regularly by fire during the twentieth century, Neither his-
torically nor now, however, has decomposition been the pri-
mary remover of biomass in Sierra Nevada forests.
Presettlement Fire
In presetllement forests most biomass ultimately was oxidized
by frequent low- to moderate-severity fires. High-severity fires
more than a few acres in size were unusual (Kilgore 1973;
Skinner and Chang 1996; Weatherspoon et al. 1992). Across
much of the landscape, dead biomass on the forest floor was
kept at low levels, and most small understory trees were killed
and subsequently consumed by fire. While small areas of high-
severity fire killed patches of large trees (Stephenson et al.
1991), most large trees survived the fires and were consumed
at some point after their death by subsequent fires. The lon-
gevity of large snags and downed logs under presettlement
fire regimes is a subject of debate. It seems likely, however,
that relatively few downed logs reached advanced stages of
decay on xeric sites before being consumed by fire, whereas a
greater proportion could last for longer periods (and also
decay faster) on more mesic sites. Physical removal from the
site was a minor component of total biomass disposition, al-
though harvest of biomass by Native Americans, especially
for firewood, may have been a significant factor locally
(Anderson and Moratto 1996).
1473
Landscape-Level Strategies for Forest Fuel Management
Twentieth-Century Fire
[f we skip now to the twentieth century, the relative roles of
fire and biomass removal have changed drastically. As fire
suppression was initiated and took effect early in the cen-
tury, the proportion of biomass consumed by fire dropped
precipitously, as did annual burned area. During the course
of the twentieth century, however, annual burned area for the
Sierra 'Jevada has shown no overall time trend, even though
it has lluctuated considerably from year to year. Large fires
have composed an increasing proportion of that burned area
as the century has progressed (McKelvey and Busse 1996). [n
recent years, large fires have become less controilable and
more severe, evidently reflecting in part increased fuel load-
ings.
Another possible indicator of changing fuel conditions is a
shift in the distribution of fires between human and light-
ning ignitions over the course of the twentieth century. We
observed this shift as part of an evaluation of twentieth-cen-
tury fi re records for Sierran national forests. \Ve used records
for fires greater than 40 ha (100 acres) within the {:v.'enty-four
core SNEP river basins. Because of the extraordinary extent
of the 1987 and 1990 lightning fires, we present the summa-
ries for two intervals of time so as to exclude and include
these two years: 1910 through 1986 (table 56.1) and 1910
through 1993 (table 56.2). We arbitrarily split each interval
into two time periods for these summaries.
TABLE 56.1
rhese summaries suggest some conspicuous differences
between human-caused fires and lightning fires. Whether the
extraordinary years of 1987 and 1990 occurred simply by chance
we cannot SdY based on these limited data. However, whereas
the fire-suFpression organization does appear to have reduced
tolal area burned by, and number of, large human-caused fires,
it has not been effective in reducing either the area burned by
or the nu mber of large lightning fires.
[n table 56.3 we summarize fire characteristics for each of
the three years of greatest burned area for each time period
All six of these years were quite dry. The summaries show
that lotal area burned was similar in these years, However,.
lightning fires contributed only small proportions of total area
burned for the first four years but very large proportions for
the last two years-1987 and 1990. It is interesting to note
that the total number of fires also differs considerably between
the earlier years and 1987 and 1990. Those two years had fewer
and much larger fires contributing most of the area burned.
The pattern of fire starts and the necessary response of the
fire-suppression organization differ considerably between the
two types of ignition. Human-caused fires generally occur as
a singular event or occasionally a few simultaneous events.
This allows the fire-suppression organization to respond to
individual fires with a relatively large body of fire-suppres-
sion resources. Lightning fires, in contrast, usually occur as
simultaneous multiple ignitions. In unusually dry years, reo
source requirements necessary to deal with simultaneous
Summary of fire characteristics for 191047 compared with 1948-86.
Total Annual Maximum Annual Total Annual
Burned Area (ha) Fire Size (ha) Number of Firea
------
Yean. 1910-47 1948-86 1910-47 1948-86 1910-47 1948-86
All Fires
Greater than 40 ha
MinImum 882 125 283 66 5 2
1 st quartile 3,990 1.257 1.260 559 12 6
MedIan 14,483 4.295 3,324 2,026 19 9
3rd quartile 21,285 11.443 8,421 6,599 35 14
MaXimum 95,126 43.330 21,234 18.100 82 23
Total for entIre period 685,880 319.806 983 395
Human-Cau...d FIr..
Gr..ater than 40 ha
Minlmum 882 125 283 66 5 2
1>, quartile 3,732 1.182 1,022 553 11 5
Meoian 14.202 3.781 3,324 1,333 17 8
3rd qU.J!1tl9 20,108 8.690 8,421 6,S99 33 11
MaXImum 93.588 39.402 21,234 18,100 65 20
Total la, entire penod 651,801 273.526 890 318
Lightning-Caused Fires
Greater than 40 ha
Minimum 0 0 0 0 0 0
1st quartile 12 21 12 21 0 1
Median 324 217 175 197 2 1
3rd quartile 901 1,233 550 708 3 4
MaXimum 9.738 7.356 5.748 7,238 18 6
Total for entire penod 34,079 46.280 93 77
".' ...............-.-.....,..'4,._........~_."'_'_
1474
VOLUME II, CHAPTER 56
TABLE 56,2
Summary of fire characteristics for 1910-51 compared with 1952-93.
Total Annual Maximum Annual
Burned Area (ha) Fire Size (ha)
y.a.... 1910-51 1952-93 1910-51 1952-93
All Fire.
Grelltar than 40 ha
MInimum 828 44 283 44
15t quartIle 3,990 1,178 1,260 526
Median 13,856 4.537 3,654 2,107
3rd quartile 20,110 12,125 8,880 6,144
Maximum 95,17.6 81,887 21.234 53,011
Tolal for entire period 730,131 454,861
Human-Causad Fires
Grutar than 40 ha
Minimum 828 0 283 0
1.t quartile 3,732 1,120 1,022 481
Med;gn 13,585 3.108 3,654 1,099
3rt! quartile 19,585 6,993 8,880 4,434
MaxImum 93,588 39,402 21,234 18,100
Tolallor entire period 692,170 267,879
Lightning-Caused Firas
Greater than 40 ha
Minimum 0 0 0 0
151 quartile 12 44 12 44
Median 324 347 175 272
3rd quartile 902 2,625 550 1.960
Maximum 9,738 80.704 5.748 53,011
Total for entire period 17,960 186,982
Total Annual
Number of Fir..
1910-51 1952-93
5 1
12 6
18 9
34 13
82 23
1039 403
5 0
11 4
17 7
32 9
65 20
934 306
0 0
0 1
2 1
3 4
18 13
105 97
multiple ignitions can quickly exceed those available (e.g.. ]977,
1987,1990), Show and. Kotok (1923) recognized early, on the
ba<;is of the 1917 fire season, that general regional lightning
events have the potential to strain the fire-suppression organi-
zation severely.
The period of record is insufficient to conclude that there
is a definite trend toward. larger severe lightning fires or that
a threshold has been cros<;ed. However, we suggest that the
potential influences of changing fuel mosdics. stand condi-
tions, and landscape patterns on the fire environment logi-
cally would begin to show up first in dry years under lightning
situations.
TABLE 56,3
Utilization
[n contrast to the changed role of fire in removing biomass,
utilization of biomass has increased by orders of magnitude
over the levels that prevailed before Euro-American settle-
ment. The components of biomass removed by logging have
changed dramatically from those that previously were re-
moved by fire. Fire-resistant large trees have been harvested
and replaced by much'more fire-susceptible small trees. Dead
biomass in the form of logging slash and natural (i.e., not pro-
duced by management activities) fuels has built up on the
forest floor because of lack of fire and inadequate or nonex-
Fire characteristics in the three major fire years (years of greatest burned area) during 19-10-51 compared with those during
1952.-93.
191 (I-51 1952-93
Year 1924 1928 1931 1959 1987 1990
F"e SIze (ha)
1st quartile 95 101 119 155 182 120
Median 305 222 249 673 277 606
3rd quartile 1.307 572 1,095 3,268 785 3,405
Maximum 15,054 10.252 17.715 7,710 53,011 38,624
T atal burned area 95,126 57,527 52.540 43,330 81,887 57,099
lightning percentage" 2 17 2 9 99 95
Total number of fires 56 80 40 23 18 11
--
oPercentage of total area burned.
.., -.-......
1475
Landscape-Level Strategies for Forest Fuel Management
istent fuel treatment. Total decomposition probably has accel-
erated, but at a rate not nea rly ~ufficien t to compensate for the
increasing fuel load. Together, surface fuels and dense under-
stories have greatly incre,lsed the risk of crown fires (Kilgore
and Sando 1975; Parsons and DeBenedetli 1 '17'1). Heightened
stress from overlr dense stands, often dominated by shade-
tolerant species no longer kept in check by trequent fires, also
has increased mortality from insects (l:'erreII1996). further add-
ing to dead biomass available as fuel.
Fuel Management
As managers began to see the consequences of increased fuel
loads, they undertook a variety of fuel-management activi-
ties. These activities have included a range of treatments that
mimic or facilitate thp. natural processes of biomass disposi-
tion; (1) burning on site (with or without prior piling or rear-
rangement), (2) accelerating decomposition (and reducing
flammability) by rearranging the fuel bed cleser to the ground,
and (3) physical removal from the sIte. Adequacy of slash treat-
ment following timber harvest or other vegetation manage-
ment activity has varied from quite good to nonexistent.
For the Sierra Nevada as a whole, however, vegetation
management activities have produced considerably more new
fuels than they have eliminated. Furthermore, the increasing
problem of live understory fuels has been addressed inad-
equately in silvicultural or fuel-management activities. Efforts
to treat accumulating amounts of natural fuels, ortP.n with
prescribed fire, also have fallen far behind rates of fuel accre-
tion, due in large part to inadequate funding and various con-
cerns about the use of prescribed fire. Even the active prescribed
burning programs in Sierran national parks ever the past
twenty-five years, utilizing both natural and management ig-
nitions, have restored fire to the forests at rates well below
presettlement levels (Botti and Nichols 1'195; Husari and
McKelvey 1996; Parsol15 1995). Consequently, these bums have
been unable even to keep up with new biomass accumulation,
let alone to consume all the excess biomass generated by de-
cades of fire suppression. The basic problem is the same out-
side the parks: current quantities of flammable
biomass-primarily small trees and surface fuels-in low- to
middle-elevation Sierran forests are unprecedented during the
past several thousand years and are continuing to accumulate
at a much faster rate than they are being removed.
The Fuel Problem and the Need for a Strategy
Civen current federal and state budget climates, increasing
suppression costs, and attrition of skilled firefighters, reduc-
tions in suppression rorces seem more likely than substantial
increases (Husari and McKelvey 1996; US. Department of the
Interior and US Department of Agriculture 1995). Accord-
ing to a growing consensus among fire managers, more sup-
prpssion capability is not the solution anyway. This idea is
reinforced, we believe, by the data presented earlier on distri-
butions of lightning and human ignitions. History tells us that
periodic dry years are inevitable and that regional-scale light-
ning events that limit the effectiveness of suppression forces
are not unusual.
If more suppression is not the answer, and if flammable
biomass continues to accumulate at current rates, and if we
do nothing substantive to arrest that accumulation, simple
physics and common sense dictate that the area burned by
high-severity fires will increase. Losses of life, property, and
resources will escalate accordingly. This conclusion is
strengthened by the fact that recent "drought" years, during
which many large, severe fires burned (McKelvey and Busse
1996), appear to be relatively common when viewed on a time
scale of centuries (Graumlich 1993).
Therein lies the rationale for large-scale fuel management.
Given the massive scope of the problem and budget con-
straints, brute force is likely to be neither feasible nor adequate.
A carefully considered strategy is required. Treatments need
to begin in the most logical. efficient, cost-effective places.
Specific components of biomass-mostly small trees and sur-
face fuels-need to be targeted. We must devise ways to pay
for the needed treatments. At least on public lands, treatments
conducted to reduce the hazard of severe wildfires should be
compatible with overall desired conditions for sustainable
ecosystems. In general. conditions need to be moved away
rrom dense, small-tree-dominated forests toward more open,
large-tree-dominated forests. And the rate of treatment needs
to be carefully planned: in the short term, rates of biomass
removal may well need to exceed rates of production in or-
der to return these forests to a more sustainable, fire-resilient
condition. The remainder of this chapter displays and dis-
cusses various considerations for developing such a land-
scape-level fuel-management strategy.
A REVIEW OF FUEL-
MANAGEMENT STRATEGIES
Our use of the term fud-management strategies here rerers to
methods for prioritizing or locating fuel treatments on a land-
scape scale in such a way as to increase their overall effec-
tiveness for reducing the extent of severe wildfires. Most past
fuel management in the Sierra Nevada has taken place in the
national forests. Most of that has not been characterized by
strategic planning: management emphasis and funding have
directed fuel management primarily toward treatment of ac-
tivity fuels following timber sales, and sales usually were not
located with strategic fuel considerations in mind_ In fact, tim-
ber sales often were dispersed-thereby reducing overall ef-
fectiveness of fuel treatments--intentionally in an attempt to
meet various management objectives, such as minimizing
cumulative watershed impacts of harvest-related activities,
In recent years, however, innovative fire and ruel managers
have begun to think much more strategically and to collabo-
. ..----r---.--r- ----.-- .. .-- --.-
1476
VOLUME II. CHAPTER 56
rate with foresters and sil viculturisL~ to add ress landscape-level
forest health concerns. TIlis change has been stimulated and
supported by the general move toward ecosystem manage-
ment and by new capabilities for spatial. landscape-level plan-
ning provided by geographical infurmation system (GIS)
technology.
Some of these evolving ideas are included in the following
sections, which provide a samplir,g of various types of fuel-
management strategies that have been proposed and, to vary-
ing degrees, implemented. Also incorporated here are some
of the ideas discussed by a group of experts in a Fuels Man-
agement Strategies Workshop spunsored by SNEP in March
1995 (Fleming 1996). Three somewhat distinct but certainly
overlapping approache!: have been used: (1) identifying fuel-
management approaches appropriate within each of several
landscape zones defined by general .:haracteristics, uses, or
emphases; (2) setting priorities based on various combina-
tions of risk, hazard, values at risk, a.ld suppression capabili-
ties; and (3) employing a fuelbreak.-type concept intended to
interrupt fuel continuity on a landscape scale and Lo aid in
limiting the size of fires by providing defensible zones for
suppression forces. A fourth "approach." that has received
explicit emphasis recently, although it is implicit to some de.
gree in the other approaches, is rate or timing of imple-
mentation.
Strategies Based on Zones
Arno and Brown (1989) proposed three landscape zones. In
Zone I, wilderness and natural areas, the emphasis would be
on prescribed natural fire (PNF), augmented by management-
ignited prescribed fires (MIPF) as necessary to restore much
of the natural role of fire to these ecosystems. In Zone II, the
general forest management zone, well-planned and well-
implemented fuel management, both in conjunction with and
in addition to proper timber harvests, would contribute sig-
nificantly to good overall management [n Zone HI, the resi-
dential forest, education of homeowners and local officials
about the realities of fire hazards in the wildland-urban in-
terface would go hand in hand with effective, esthetically
pleasing manipulation of fuels. The authors suggested that
shaded fuelbreaks around homes and developments could
be an effective measure. They recommended concentrating
most efforts in Zone HI and adjacent portions of Zone II.
A somewhat different zone approach provides the basis for
fire-management direction in Sequoia--Kings Canyon National
Parks (Manley 1995) . Zones are defined by estimated prox-
imity of current conditions to the natural range of variability.
In Zone 1. areas essentially unaffected by postseltlement ac-
tivities (mostly higher elevations), natural processes, includ-
ing PNF, are permitted to operate with little restriction. In
Zone 2, areas significantly modified by postseltlement activi-
ties, corrective actions, including conservative use of PNF and
MfPF, are required before permitting resumption of all natu-
ral processes. In Zone 3, built-up areas with highly flammable
fuel types near park boundaries, full suppression is combined
wi th mechanical f ue It realmen ts and conservalive use of MIPE
Greenwood (1995) described a land classification system
based on structure density (presumably closely related to
population density) plus appropriate fire-related buffers.
While his analysis was done for the entire state of California,
the subset of Sierra Nevada data could easily be analyzed
separately, and most of his general conclusions probably
would still apply. He labeled the classes wildland, intermix,
and developed, corresponding to increasing structure densi-
ties, and noted the surprisingly high percentage of land in
the intermix category, even 00 public lands. He emphasized
that the presence of people and their structures constrains
many of the options available for both fuel management and
fire suppression. Approaches suggested ranged from reestab-
lishment of presettlement conditions aud processes in some
wildland areas to reliance on fire-safe regulations, public edu-
cation, aggressive initial attack, and only minimal vegetation
manipulation in more densely settled developed areas,
Strategies Based on Risk, Hazard, Values at
Risk, and Suppression Capabilities
To provide a common frame of understanding for the discus-
sion that follows, definitions of "risk," "hazard," and "val-
ues at risk" (McPherson et a!. 1990) are given here.
FIRE RISK: (1) The chance of fire starting, as affected by the
nature and inddence of causative agents. . . (2) Any caus-
ative agent. (P. 45)
!'IRE HAZARD: A fuel complex, defined by volume, type,
condition, arrangement, and location, that determines the
degree of ease of ignition and of resistance to conlrol. (P,
42) "Resistance to control" is related both to fire behavior
and resistance to line construction.
VALU5-Af-RlSK: Any or all natural resources, improvements,
or other values which may be jeopardized if a fire occurs.
(P. 131)
A number of authors have reported the use of decision analy-
sis to aid'in fuel-management decision making (Anderson et
a!. 1991; Cohan et a!. 1983; Radloff and Yancik 1983). Decision
analysis became the cornerstone of the National Activity Fuel
Appraisal Process (Hirsch et al. 1981; Radloff et al. 1982),
which was intended to provide a consistent means of evalu-
ating the important fuctors affecting fuel-treatment decisions.
The Fuel Appraisall'rocess provided probabilities of various-
sized fires by intensity class, based on information about to-
pography, historical weather, historical fire occurrence (risk),
suppression capability, and hazard (measured or projected
based on alternative fuel treatments).
Biehl (1995) described an "all risk management" strategy
in use on the Stanislaus National Forest. Fuel profiles, ex~
1477
Landscape-Level Strategies for Forest Fuel Management
pected ignitiL1ns, and suppression resources are used in con-
junction with management-defined acceptable resource loss
to determine wheLl-jer, where, and what kind of fuel treatment
is needed. The Stanislaus National Forest is combining the
most active prescribed burning program of all California na-
tional forests-concentrated mainly in natural (i.e.,
nonactivityJ fuels-with considerable biomass thinning.
Fuelbreaks are employed, but only as anchor lines to facili-
tate initiation of areawide fuel treatments using pre-
scri bed fi re.
PE:rkins (1995) has devised a similar fire-analysis system
for use on the Klamath National Forest as part of the forest's
landscape-analysis system. Risk, fire behavior potential (based
on fuel classification, slope class, and ninetieth-percentile
summer wildfire weather conditions). and resource values
(based on forest plan direction) are the primary factors used
to determine fuel-management treatment priorities. Fuels in-
formation is derived from vegetation classification, modified
by management history and large-fire history.
James (1994) developed a simple system for estimating a
"catastrophic fire vulnerability rating," based on a point total
derived from separate qualitative assessments of risk, haz-
ard, value, and suppression capability. The system includes
three sets of "firel fuel treatment standards" corresponding
to fire vulnerability ratings of high, moderate, or low. Finally.
it provides a straightforward feedback mechanism for adjust-
ing the posttreatment vulnerability rating. All vulnerability
factors are weighted equally, but local managers should be
able to modify weightings fairly easily to account for their
assessment of the relative importance of various factors.
Strategies Based on Fuelbreaks or Similar
Landscape-Level Interruptions of Fuel
Continuity
rUr.LBREAK5: Generally wide (50--1,000 feet) strips of land
on which native vegetation has been permanently modi-
fied so that fires burning into them can be more readily
controlled. (McPherson et 411. 1990,56)
Early Experiences with Fuelbreaks
Green (1977) rraced the long history of fuelbreaks and their
predecessors, firebreaks (narrower strips usually cleared to
mint::ral soil), in California. Perhaps surprisingly, a recommen-
dation to tbe State Board of Forestry for blocking out the for-
est with strips of "waste" land wide enough to prevent fire
from crossing was made as early as 1886. The Sierra Nevada
was a part of early firebreak history. S. B. Show, District For-
ester. proposed in 1929 that a firebreak be constructed along
the entire length of the western slope of the Sierra Nevada at
the interface of the chaparral and the pine forest. Depression-
related federal funding, especially for the Civilian Conserva-
tion Corps, permitted work to begin in 1933 on what came to
be known as the "Ponderosa Way and Trucktrail." The intent
of this strip, which when completed was about 1,050 km (650
mil long and generally 4~ m (15G-200 ft) wide (Green 1977),
was 10 help prevent fires from burning from the chap-
arral up into the more valuable Sierran timber (Green and
Schimke 1971).
The transition from firebreaks to fuelbreaks came about as
p3rl of preattack planning in the early 1950s (Green 19771-
Most early fuelbreak construction was in southern California
chaparral. The Duckwall Conflagration Control Project on the
StdIlislaus National Forest, initialed in 1962, extended the
fuelbreak concept into the Sierra Nevada mixed conifer for-
est type (Green and Schimke 1971). Green and Schimke (1971),
Pierovich and colleagues (1975), and Green (1977) provided a
number of guidelines for planning, constructing, and main-
taining fuelbreak systems. Among their recommendations:
The number and location of fuelbreaks, along with the size of
blocks to be separated by the fuelbreak network (1,000 ha
[2,500 ac] for the Duckwall program), should be determined
by fire-control objectives as part of the preattack planning
process. Needs for protecting populated areas or high resource
values should be given high priority in fuelbreak location.
Planned management projects-in range, wildlife, recreation,
timber, watershed, and forest roads and trails--should be re-
viewed to see how they might contribute to the fuelbreak
network. Ridges usually are preferred for locating fuelbreaks,
although other locations can be used. Locating fuelbreak.<;
along existing roads where possible was recommended to
facilitate access by suppression forces. Suggesled fuelbreak
widths varied from about 60 to 120 m (200 to 400 ft). The ne-
cessity of maintaining reduced-fuel conditions on fuelbreak.s,
through a combination of appropriate vegetation (e,g., low
volume and lor low flammability) and periodic treatments,
was emphasized.
A number of anecdotal accounts of the effectiveness of
fuelbreaks (or lack thereof) during wildfire incidents, mostly
during the 1960s and early 1970s, were summarized by
Pierovich and colleagues (1975) and Green (1977). Although
experiences were mixed, fuelbreaks were found to be effec-
tive much of the time in stopping wildfires except under the
most extreme conditions, Success was most likely when
fuelbreaks were properly installed, properly maintained, and
adequat~ly staffed by suppression forces during wildfires.
The same authors (Pierovich et a!. 1975; Green 1977) dis-
cussed existing economic analyses of fuelbreak effectiveness,
which differed L.'"1 their conclusions but for the most part found
that a fuelbreak system could be justified economically as part
of a well-integrated fire-management system. A subsequent
study of fuelbreak investments in southern California, using
a linear programming model, predicted that increasing
fuelbreak widths could substantially reduce area burned and
fire-related damages if initial investments were concentrated
in a specific "damage-potential zone" (Dmi 1979). Although
potential corollary-i.e., nonfire--benefits of fuel breaks ha ve
been recognized (Green 1977), such benefits generally have
not been considered in evaluations of their efficacy or cost
effectiveness. [n a study of three forested fuelbreaks ill the
.... r
.,.,......_~.--~......._-, '"_. ..._,-.~_. ~_.->
1478
VOLUME II. CHAPTER 56
central Sierra Nevada, howevpr, Grah and Long (1971) found
that fuelbreak construction increased timber valut>s within the
fuelbreaks by realloeating site resources to larger, faster grow-
ing, and more valuable trees. A portion of fuelbrt>ak costs,
therefore, was offset by the benefit to the timber resource.
Recent Experiences and Recommendations for Using
Fuelbreaks
Fuelbreak construction and maintenance havt> retained some
emphasis in southern California. Salazar and GonzalC'L-Caban
(1987) found that in a large 1985 wildfire in chaparral on steep
terrain, the fuelbreak system apparently influenced the loca-
tion of the final fire perimeter. Except during the most ex-
treme burning conditions, fuelbreaks functioned as intended.
In contrast, most forested areas in the state have seen little
attention given to fuelbreaks over the past twenty years. Fuel
management in Sierra Nevada national forests l1dS been domi-
nated by support of the timber management program du ring
most of that period. Budgets for other fuel acti vities have been
quite limited. Furthermore, many fire and fuel specialists have
viewed fuelbreaks as being of little value for a variety of rea-
sons, including the following: (1) to be effective for stopping
fires, fuelbreaks need to be staffed by suppression forces,
which otten have been unavailable when needed, frequently
because of demands for protecting structures in urban-wild-
land intermix areas; (2) in general, recommended fuelbreak
widths of 60-120 m (200-400 ft) (Gn~en and Schimke 1971.
Green 1977) have been considered too narrow to be effective
under many conditions, especially with ex!ensi ve spotting
(ignition of new fires outside the perimeter of the main fire
by windborne sparks or embers); (3) fuelbreaks often have
been viewed as standalone measures that competed with more
effective areawide fuel treatments; and (4) fire control has bt>en
viewed as the sole beneficiary of fuelbreaks, with little thought
given to other potential resource benefits.
Over the past ten years or so, a number of large, severe
fires in California and elsewhere in the western United States
have emphasized the seriousness and the enormity of the
wildland fuel problem. Fuelbreaks have begun to receive re-
newed attention as one part of the solution. Arno and Brown
(lq89) suggested their use around homes and developments
in the wildland-urban interface. In the recovery plan for the
northern spotted owl, Agee and Edmonds (1992) recom-
mended the use of fuelbreaks along with underburning to
reduce the probability of catastrophic wildfires in "designated
conservation areas" within the Klamath and East Cascades
subregions. Weatherspoon and colleagues (1992) suggested a
two-stage fuelbreak strategy to help reduce the occurrence of
severe fires in California spotted owl habitat in Sierra Ne-
vada mixed conifer forests. Known owl sites first would be
"isolated" using a broad band of prescribed bums, followed
by a more general program of breaking up fuel continuity on
a landscape scale, Fites (1995) proposed a similar approach
to help protect "areas of late-successional forest emphasis"
and to restore more sustainable, fire-resilient cond itions across
the landscape. Arno and Ottmar (1994. 19) pointed out the
need for" an interconnected network of natural fire barriers
and treated stands as zones of opportunity for controlling
wild fires."
In the draft Environmental Impact Statement (EIS) for man-
aging California spotted owl habitat in Sierra Nevada national
foresls (U.s. Forest Service 1995), Alternatives C and Din-
cluded an upper slope/ridge zone that would be dominated
by large, widely spaced shade-irltolerant trees, These alter-
natives were viewed as creating conditions in this zone closer
to those thought to have existed before Euro-American settle-
ment. In addition, the zone would provide many of the fire-
management benefits of a wide shaded fuel break. Alternative
F incorporated some of the fuelbreak-related concepts of the
Quincy Library Group (QLG) proposal (summarized later)
for the northern Sierra Nevada,
LaBoa and Hermit (1995) presented a number of ideas for
strategic fuel planning and treatment, based on their recent
work as members of the California spotted owl ErS Team (suf-
ficiently recent that these ideas were not included in the draft
EIS). They included the use of fuelbreaks; however, they
stressed the need not to stop with a Cuelbreak network but to
build from it to accomplish large-scale fuel modification on a
landscape level.
The most detailed fuel-management strategies to date have
been proposed for the northern end of the Sierra Nevada-
the Lassen and Plumas National Forests and the Sierraville
Ranger District of the Tahoe National Forest. The two strate-
gies. which were developed semi-independently by the QLG
and the U.S. Forest Service, have much in common and build
on many of the ideas cited earlier. Rapid implementation of a
network of broad fueIbreaks is key to both proposals.
QLG is a community-based group whose members repre-
sent a wide range of interests, including fisheries and envi-
ronmental groups, timber industry, and county government.
The group has made strategic fuel management a central fo-
cus of its land management proposal (Quincy Library Group
1994)_ QLG proposes that an intensive four-year effort be fo-
cused on installing a network of strips approximately 0.4 kID
(0.25 mil in width, mostly along existing roads, that break up
fuel continuity across the landscape and provide defensible
zones for suppression forces. During this period, essentially
all forest management activities, including biomass and other
thinnings, salvage activities, and treatment of surface fuels,
would be focused on implementing this fuelbreak network.
Each year 1/32 of the total forest acreage would be treated,
so that at the end of the four-year period 1/8 of the forest
would be a part of these strips. The strips would have reduc-
tions in stand density, lower canopy ladder fuels, and surface
fuels, and they would have relatively low levels of snags and
large downed woody debris. After the initial period, a longer
term fuel-management strategy would add some strips to iso-
late areas of high value and I or high risk, but the emphasis
generally would shift to areawide treatments.
The Technical Fuels Report, prepared by firel fuel special-
1479
Landscape-Level Strategies for Forest Fuel Management
ists from the Lassen, Plumas, and Tahoe National Forests
(Olson et al. 1995), is similar In several respects to the QLG
proposal. The "defensible fuel profile "-One" (DFrZ), a con-
cept fi rst described by Olson (1993), is centr,,1 to the strategy
outlined in the report. Much like a broad fuelbreak, a OFPZ
is a low-density. low-fuel L.one averaging 0.4 km (0.25 mil in
width, located mostly along roads, and designed to support
suppression activities. Like the strips in the QLG proposal.
DFPZs are intended to be installed over a period of just a few
years. The authors point out that DFPZs are intended not to
take the place of widespread fuel treatment but rather to in-
crease the effectiveness of initial fuel treatment and to facili-
tate subsequent treatment of adjacent areas. Olson et al. (1995)
describe the "community defense zone" (COZ) as another
component of their strategy concerned with urban interface
areas within or near national forest boundaries. Similar in
concept to a DFPZ, a COZ is designed to reduce the threat of
wildfire spreading onto national forest land from private land,
or vice versa. Like DFPZs, CDZs would have a high priority
for completior. within a short period of time. The authors
stress the importance of the involvement dnd cooperation of
local communities in implementation of CDZs. A third type
of zone, the "fuel reduction zone" (FRZ), refers to general area
fuel treatment that would take place mainly after the high-
priority system of orrzs and C:DZs is in place. The Technical
Fuels Report (Olson et al. 199')) emphasizes the importance
of site-specific considerations and local decision making in
setting priorities and implementing th~ details of the broad
fuel-management strategy outlined.
A POTENTIAL FUEL-
MANAGEMENT STRATEGY FOR
SIERRA NEVADA FORESTS
The approaches summarized in the previous section, along
with the discussion at the SN EP Fuels Strategies Workshop
(Fleming 1996), seem to point to some degree of convergence
of thinking "bout the fuel problem and some components of
a strategy to deal with it. In this section we attempt to syn-
thesize many of the previously mentioned approaches into
an outline for a potential fuel-management strategy for Si-
erra Nevada forests.
The ideas presented here are necessarily general in nature.
The Sierra Nevada is enormously complex and diverse. Land-
owne~s and ownership objectives vary wid!'ly While agen-
cies and Jarge landowners may choose to set some priorities
on a regional or subregional scale, any attempt on our part to
recommend or prescribe specific management practices
rangewide would be naive, counterproductive. and contrary
to the SNEP charter. Readers should view this "strategy" as a
set of principles and ideas to consider as they develop their
0'''''11 landscape-specific strategic plans. (Additional ideas can
be found in cited references.) Such plans will be greatly fa-
cilitated and improved by developing and maintaining good
GIS databases. Later in this chapter we discuss the nature and
role of such databases for supporting fire and (uel-manage-
ment decision making in the context of adaptive ecosystem
management (Everett et al. 1994; Walters and Holling 1990).
Although landscape-specific planning is focused on a small
portion of the entire Sierra Nevada, it nevertheless requires
thinking on a much broader scale than often has occurred in
the past. Making significant progress toward these goals will
require long-term vision, commitment, and cooperation across
a broad spectrum of land-management agencies and other
entities. Dealing with fuels on only a local. piecemeal basis
will be inadequate.
Goals of the Fuel-Management Strategy
i'he strategy has three general goals, ranging from short to
long term and from relatively narrow to broad. Each goal can
be viewed as nesting within the following one. The goals are
consistent and complementary, as are the means to work to-
ward their accomplishment. For example, the strategy pro-
vides that short-term approaches to reducing hazard be
compatible with longer-term goals of ecosystem sustainability
(Amo and OHmar 1994).
The first goal-the immediate need from a fire-manage-
ment standpoint-is to reduce substantially the area and av-
erage size burned by large, severe wildfires in the Sierra
Nevada. Ideally this will be a short- to medium-term goal.
whose urgency will lessen as the fuel-management strategy
becomes increasingly effective. A second, longer-term goal
should be to restore more of the ecosystem functions of fre-
quent 10w- to moderate-severity fire. The two goals are closely
linked, They could be met simultaneously by replacing most
of the high-severity acreage with the same, or preferably much
greater, acreage of low- to moderate-severity fire. A third,
overarching goal is to improve the health, integrity, and
sustainability of S ierr a N ev ada ecosystems. Thi s goal certainl y
goes beyond fire considerations. Progress toward achieving
the first two goals, however, is critical to the third.
Management actions to progress toward these three goals
should be occurring concurrently. Often it will be possible
for a single treatment or project to address all three goals Si-
multaneously. In fact, opportunities for such congruence
should be sought. In this chapter. however, we spend the most
time addressing the first goal--not because it is most impor-
tant in the long run but because it is the most urgent in the
short ntn lo reduce losses of lives, properly. and resources,
and to make it possible to work more effectively toward
achieving the second and third goals. Stated in another way,
the fuel-management strategy has joint themes of protection
and restoration of ecosystems, and, in many portions of the
Sierra Nevada, protection is a prerequisite to restoration. In a
longer term context, strategies geared specifically toward re-
ducing losses from large, severe wildfires should gradually
.......--.-...,-..
1480
VOLUME II, CHAPTER 56
become less important; restoration in turn should provide a
more fundamental level of protection along with improved
ecosystem health.
Goal 1: Reduce Substantially the Area and Average
Size Burned by large, High-Severity Wildfires
Large, high-severity fires were unusual historically in most
Sierra Nevada forests. Fire regimes in the Sierra Nevada gen-
erally were characterized by relatively frequent, low- to mod-
erate-severity fires (Skinner and Chang 1996). Changes in low-
and middle-elevation forests and their associated fuel com-
plexes, brought about largely by human activities since Euro-
American settlement (including but not limited to fire
suppression), have made these forests much more prone to
large, severe fires (Chang 1996; Husari and McKelvey 1996;
McKelvey and Johnston 1992; Skinner and Chang 1996; U.s.
Forest Service 1995). Such fires, in aggregate, are well outside
the natural range of variability and thus can be considered
detrimental to Sierra Nevada ecosystems (Manley et al. 1995).
Furthermore, the current prevalence of such fires is unaccept-
able socially. The rapidly increasing population of the Sierra
Nevada increasingly places people's houses at risk of loss to
severe wildfires and makes potential solutions to the prob-
lem much more difficult.
In pu rsu ing goal 1, it is essential for the wildland iire agen-
cies to continue support for suppression and prevention ac-
tivities. These tire-management efforts alone, however, cannot
resolve the problems of fire in the Sierra Nevada, Aggressive,
strategically logical fuel-management programs, compatible
with overall desired conditions for sustainable ecosystems,
are necessary to address the basic problem of excessive fuel
accumulation.
Goal 2: Restore More of the Ecosystem Functions
of Frequent Low- to Moderate-Severity Fire
The frequent low- to moderate-severity fires that occurred
throughout lTluch of the Sierra Nevada until about 150 years
ago performed many important ecological functions (Kilgore
1q73; Chang 1996). Wildfires of this type, however, have been
virtually eliminated from Sierra Nevada ecosystems (as mea-
sured by annual area burned by such fires), because these are
the fires that are suppressed most easily. As a result, the eco-
logical functions historically performed by such fires have
been largely lost, with some known and many unknown con-
sequences. It is highly unlikely that fires will ever burn as
much area as often and with the same distribution of severi-
lies as they once did. Nevertheless, it makes sense to try to
restore fire to a more nearly natu ral role in those parts of the
landscape where it is practical to do so. Where fire alone can-
not be used practically, fire surrogates such as silvicultural
techniques and me{'hanical fuel reduction methods (Helms
and Tappeiner 1996; Weatherspoon 1996) can be employed-
either by themselves or in conjunction with prescribed fire-
as appropriate to mimic some of thE> functions of fire and to
move landscapes toward desired conditions (Manley et al.
1995). Over time, adaptive management (Everett et al. 1994;
Walters and Holling 1990) should help us to determine which
ecosystem functions of fire can be emulated satisfactorily by
surrogates, which mdY be irreplaceable, and the implications
for management.
Goal 3: Improve the Health, Integrity, and Sustainability
of Sierra Nevada Ecosystems
The third goal is consistent with the first two and is central to
overall SNEP goals. rt should be achievable (1) by reducing
the incidence of high-severity fires, which are detrimental to
ecosystem sustainability in natural fire regimes characteris-
tic of most of the Sierra Nevada; and (2) by moving ecosys-
tems closer to pre-European-settlement conditions and
processes, assumed by many to be a useful first approxima-
tion of sustainable ecosystems (e,g., Manley et al. 1995;
Swanson et al. 1994), at least on public lands. We cannot de-
fine those presettlement conditions with any great precision,
but we do know enough to be reasonably confident that this
strategy would move us in the desired direction.
Components of the Strategy
The strategy we discuss here has three basic components: (1)
networks of defensible fuel profile zones (DFPZs) (the term
adopted from Olson 1993 and Olson et al. 1995) created. and
maintained in high-priority locations; (2) enhanced use of fire
for restoring natural processes and meeting other ecosystem
management goals; and (3) expansion of fuel treatments to
other appropriate areas of the landscape, consistent with de-
sired ecosystem conditions. We also discuss possible institu-
tional changes that might increase the effectiveness of the
strategy. This strategy builds upon and draws freely from the
various strategies cited elsewhere in this chapter.
Defensible Fuel Profile Zones
Given the massive scope of the problem that goal 1 is intended
to address, a carefully considered strategy is required for pri-
oritizing fuel treatments. Such a strategy should permit man-
agers to multiply the benefits of treatments in order to make
the most rapid and most efficient progress toward achieving
goal 1. We focus our discussion in this section on DFPZ net-
works. Multiple benefits of DFPZs may include (1) reducing
severity of wildfires within treated areas (as with any fuel-
management treatment), (2) providing broad zones within
which firefighters can conduct suppression operations more
safely and more efficiently, (3) effectively breaking up the
continuity of hazardous fuels across a landscape, (4) provid-
ing "anchor" lines to facilitate subsequent areawide fuel treat-
ments, and (5) providing various nonfire benefits, We are
aware of no other strategy with as great a potential in the
short term to progress reasonably rapidly toward achieving
goal 1.
1481
Landscape-Level Strategies for Forest Fuel Management
Rationale
The basic purposes of fuelbre;Iks were summarized earlier.
These stated purposes generally do not include some of the
potential benefits we envision fOl DFPZs, however. We offer
an expanded rationale here, including the reasons for our
choosing not to use the term fuelbreak ~s part of the strategy
we describe.
fuel-management activities in forested ecosystems nor-
mally involve some combination of (1) removing or modify-
ing surface dead fuels to reduce their flammability; (2)
removing or modifying live fuels to reduce their horizontal
and / or vertical continuity, thereby reducing the probability
of crown fire; and (3) felling excess snags that c:auld be safety
hazards and sources or receptors of firebrands.
The kind of protection afforded by fuel-management treat-
ments depends not only on the localized nature of the treat-
ments but also on their scale and spatial relationships. If you
do a good job of treating fuels on a I-acre (0.4 ha) patch of
forest but do nothing in the surrounding forest, the edge ef-
fects probably will overwhelm the treatment in the event of a
severe fire, and the small patch will be lost as well as every-
thing around it. (There is a lesson here for group selection
cuttings [Helms and Tappeiner 1996; Weatherspoon 1996]: it
makes little sense to do fuel treatments in only the small re-
generation openings and ignore the rest of the forest
[Weatherspoon and Skinner 1995].) If you treat fuels to the
same standard in a square 40-acre (16 ha) stand, edge effects
are relatively much less important. Fire intensity will be much
lower than in the surrcunding (untreated) forest, and under
most conditions the majority of the stand probably will sur
vive. However, that 40-acre stand probably will have only a
limited effect on iire damage in the untreated forest down.
wind. If YOil now treat the fuels on n 40-acre stands scattered
randomly across the landscape, essentially the same result is
expected, times n-i.e" the treated stands probably will not
suifer excessive damage from a fire, but their intensity-reduc-
ing effect will not extend much beyond the treated areas. This
last scenario, incidentally, approximates most of our past fuel
treatments, which were not planned with strategic fuel man-
agement in mind.
If you take that same total treated acreage (40n) and string
it together into a broad zone (DFPZ) that makes sefL~e strate-
gically, you have still protected those treated acres, with even
less edge effect. In addition, however, you now have a rea-
sonable chance of putting suppression forces into that zone
and stopping the fire, thereby protecting areas on the down-
wind side of the DfPZ.
The term fuelbreak or shaded fuelbreak has been used to
describe some of the same ideas. We do not use either term in
descrIbing this strategy, however, because they tend to carry
some undesirable connotations:
. A shaded fuelbreak is often envisioned as a strip of land
too narrow (60-120 m [200-400 il] [Green and Schimke 1971;
Green 1977]) to be effecti ve for stopping a fire under many
conditions In contrast, 0.4 km (0.25 mil has been suggested
as a nominal width for DFPZs (Olson et al. 1995; Quincy
Library Group 1994). Use of the term zone (the Z in DFPl)
suggests a broader treated area than fuelbreak.
. A shaded fuelbreak is usually considered to have a single
purpose-a relatively safe, accessible location in which
suppression forces can initiate suppression actions. A DFI'Z
also serves this suppression function, almost certainly more
effectively (because of its greater width) than a normal
shaded fuelbreak. In addition, however, the DFPZ repre-
sents a substantial portion of the landscape-perhaps 10
to 25 percent for a completed network-within which fire
damage is likely to be much reduced in the event of a wild-
fire. Furthermore, a DFPZ network may represent a num-
ber of potential additional benefits, including improved
forest health, greater landscape diversity, increasl'd avail-
ability of open forest habitat, and probably greater prox-
imity to the historic range of variability and desired
conditions.
. A shaded fuelbreak is often envisioned as "an alternative"-
i.e., a standalone option for dealing with fuels, The DFI'Z
incorporates the notion that landscape treatment of fuels
must start somewhere, so it makes sense to begin in strate-
gically logical locations. The DFPZ is a place to start--a
place from which to build out in treating other appropri-
ate parts of the landscape-not an end in itself.
General Location, Description,
Creation, and Maintenance
For the most part, DFPZs should be placed primarily on ridges
and upper south and west slopes. All else being equal, DFPZs
should be localed along existing roads to simplify construc-
tion and maintenance and to facilitate use by suppression
forces. Where roads do not follow ridges, road locations in
relatively gentle terrain-e.g., along broad valley bottoms-
are usually suitable for DFPZs. Roads that follow side slopes
and canyon bottoms in steep terrain should be avoided except
where they might facilitate stream crossings by DFPZs.
A network of DFPZs that define discrete blocks of land
would require some DFPZ segments to cross drainages. De-
cisions about how best to deal with stream crossings should
be based upon site-specific analyses. [n most cases, however,
we anticipale that the function of a DFPZ network would not
be seriously jeopardized by limiting any treatments within
the riparian zone portion of a DFPZ to those treatments (if
any) deemed acceptable elsewhere in the riparian zone. Pre-
scribed burning might be particularly appropriate as a tre,lt-
ment. Because of their relatively moist environment, untreated
or minimally treated riparian zones normally should not
present an undue risk of serving as a "fuse" to spread fire
across a DFPZ adequately staffed with suppression forces.
A reasonable nominal widlh for DFPZs is probably 0.4 km
(0.2:; mil (Olson el al. 1995; Quincy Library Group 1994) ul'.til
..._,------..-- _...._..
1482
VOLUME II, CHAPTER 56
experience indicates otherwise. It seems logical. however, to
vary the width based on strategic importance, topography,
or other conditions. For example, a broad, major ridge with a
main road might warrant a considerably wider [)H'Z than a
spur ridge with steep l>ide slopes. Using the fire-growth model
FARSITE to model various fuel-treatment alternatives, van
Wagtendonk (1996) found that fires burning under ninety-
fifth-percentile weather conditions spotted across 90-m (300
ft) fuelbreaks under most fuel treatment scenarios but did not
spot across 390-m (slightly less than 0.25 mile) fuelbreaks
under any of the scenarios_
The Quincy Library Group (1994) proposed that DFPZs be
used to break up the land into blocks averaging 4,000-5,000
ha (10,000-12,000 ac). We have no reason to argue with that
as a first approximation, but the appropriate area certainly
will vary among la.ndscapes as a function of topography and
the various factors discul>sed later. In many cases it may be
logical to implement an initial high-priority "low-density"
DFPZ network~,g., along major ridges and main roads and
in the vicinity of forest communities. Subsequent efforts
would be a combination of maintaining existing DFPZs, con-
structing flew ones to break up the landscape into sma!ler
blocks, and broadening existing DFPZs in conjunction with
areawide fuel treatments.
Treatment of DFPZs should result ill a fairly open stand,
dominated mostly by larger trees of fire-tolerant species.
DFPZs need not be uniform, monotonous areas, however, but
may encompass considerable diversity in ages, sizes, and dis-
tributions ofteees. The key feature should be the general open-
ness and discontinuity of crown fuels, both horizontally and
vertically, producing a very low probability of sustained
crown fire. Similarly, edges of DFPZs need nol be abrupt but
can be "feathered" into the adjacent forest. Posttreatment
canopy closure usually should be no mOle L~an 40%, although
adjustments in stand density based on local conditions cer-
tainly are appropriate. In some areas, for example, greater
canopy closure may be desirable to slow encroachment by
highly flammable shrubs or other understory vegetation, so
long as tree crowns are high enough that a sustained crown
fire in the denser canopy is very unlikely
Available treatmenttechnique~ for OFPZs include silvicul.
tural cutting methods, prescribed fire, mechanical fuel-reduc-
tion techniques, and combmations of these. [n most cases,
cuttmgs of various kinds will be the most effective initial treat-
ments to accomplish needed adjustments in stand structure
and composition (Helms and Tappeiner 1996; Weatherspoon
1(96). Thinning from below often will be a desirable tech-
nique to move DFPZs from overly dense, small-tree-domi-
nated stands toward more open, large-treed0minated stands.
Prescribed fire frequently will be the treatment of choice fol-
lowing a cutting. In some areas, prescribed fire alone may be
the preferred approach because existing stand conditions are
near desired conditions or because cuttings are precluded or
otherwise inappropriate. Generally, however, prescribed fire
is not likely to be a suitable standalone technique for bring-
ing about major changes in stand structure on the large scale
necessary for timely implementation of DFPZ networks in
Sierra Nevada coniferous forests. Factors that argue against
massive and rapid increases in standalone prescribed burn-
ing include lack of adequate funding (initial burns in
unthinned stands may be quite expensive), air-quality restric-
tions, competition for trained personnel during active wild-
fire seasons, and risk of escapes. Moreover, needed reductions
in stand density using fire alone could require a number of
successive bums spanning several decades. Failure to utilize
biomass in the process would generate large quantities of
smoke from consumption of excess biomass and would forgo
opportunities to generate income to finance treatments. Op-
portunities for economic and social benefits would be forfeited
as well. Furthermore, effects of initial bums probably would
not closely approximate "natural" fire effects because of fuel
complexes that differ greatly from those of the presettlement
era (Skinner and Chang 1996; Weatherspoon 1996).
To ensure effectiveness of a DFPZ, basic adjustmenL~ in
stand structure must be followed by reduction in surface fu-
els to a low-hazard condition using prescribed fire or mechani-
cal methods, or both. In some cases, adequate mechanical
"treatment" may result from crushing of fuels during har-
vest operations, especially where whole trees are removed
from the stand. Prescribed fire was the best choice among van
Wagtendonk's (1996) modeled scenarios from the standpoint
of reducing surface fuels, and it also can raise the bases of
live crowns (by killing lower branches) to increase vertical
discontinuity of live fuels, Where feasible economically, re-
moval and utilization of cut trees are preferable to treating
them in place as fuels, Densities of snags and downed logs
should be kept relatively low and compensated as appropri-
ate by higher densities outside DFPZs.
From a fire standpoint, ridges and upper southerly slopes
generally should benefit more than average from thinning and
hazard reduction: they tend to dry out faster and without treat-
ment would support severe fires a higher proportion of the
time than other aspects and slope positions. The heavy thin-
ning also would promote faster growth of trees into large size
classes less susceptible to fire damage. Their low-fuel charac-
ter, low density of snags, and resistance to sustained crown
fires should make DFPZs substantially safer for suppression
personnel than most other locations. Furthermore, the effi-
ciency and productivity of suppression forces in building am!
holding firelines and in backfire operations should be signifi-
cantly enhanced in DFPZs, especially in those containing
roads. Aerial retardant drops should be considerably more
effective in DFPZs as well because of the open canopy and
relatj VI' ease of getting retardant to the forest floor.
To retain their effectiveness, DFPZs should be maintained
in low-fuel conditions with periodic retreatments, targeting
especially accumulated surface fuels and new growth of un-
derstory vegetation. Retreatment with prescribed burns
should be relatively easy and inexpensive in the open envi-
ronment of DFPZs. (It should be noted in this regard that
1483
Landscape-Level Strategies for Forest Fuel Management
DFPZs are not unique in their ne"d ior maintenance. fuel
treatments anywhere require maintenance to retain their ef-
fectiveness. A OFPZ should cost less to maintain than an equal
area of comparable fuel treatment Elsewhere, however, be-
Cduse of its contiguity and relative accessibility.) Bums may
be required about once every ten years or more often depend.
ing on rate of encroachment by shrubs and other understory
fuels. OFPZ retreatment may be combined with broadened
area treatment, using the Ofrz as an "anchor line." Appro-
priate vegetative grour.d covers, including perennial grasses
and low-volume shrubs (e.g.. bear clover), can reduce main-
tenance needs (Green 197/).
As main canopy trees grow and increase in crown area,
they will need to be thinned periodically to maintain desired
crown spacing. A few may be left to become snags, but snag
density generally should be lower than elsewhere in the for-
est. [n additIon. long-term m<lintenance of a large-tree-domi-
nated DFPZ will require periodic regeneration of portions of
the zone. Long-rotation. low-density versions of group selec-
tion (Weatherspoon 1996) might be the best silvicultural
method for this purpose, because it providf's for regenera-
tion of shade-intolerant (generally fire-tolerant) species and
permits the maintenance of single canopy layers in any given
location, thereby discouraging crown fires. With long rota-
tions, a DFPZ could have Sustainable age-class st:-uctures and
still be occupied mostly by fire-resistant large trees.
Potential Nonfire Benefits
A range of benefits not directly related to fire would be ex-
pected to accrue from having more open stand conditions
along ridges and upper southerly slopes. In general, such open
conditions probably would be somewhat similar to those that
dominated the same topographic positions in presettlement
forests {Skinner and Chang 1996)-on it verage more open than
other sites because of more xeric conditions and more fre-
quent fires. A probable reduction in total evapotranspiration
could lead to increased water yield from these sites. Prob-
abi lity of ad verse watershed effects from harvesting and other
management activities should be reduced because of greater-
than-average distances from stre;uns (Kattelmann 1996). These
areas should contribute to overall habitat diversity and es-
thetic variety in landscapes that currently tend to be deficient
in open, large-tree-dominated structures (Craber 1996; U.s.
Forest Service 1995). forage cond:tions should be improved
in more open forest area~., especially with prescribed fire
(Mellke et al. 1996). and conceivably could help to reduce live-
stock grazing pressure in riparian areas. From a timber stand-
point, tctal production of woody biomass might be reduced
but would be concentrated in larger, more valuable trees (e.g.,
Grah and Long 197]). Lower stand density should reduce
stress on trees and make them less susceptible to insect at-
tack (Ferrell 1996). It is possible. though unproved, that broad
zones of relatively low susceptibility to insects could reduce
"conldgion" effects of insect activity, thus perhaps slowing
movement of outbreaks (Mason and Wickman 1994). If found
to be true, this idea would provide an interesting parallel to
the effect of a low-hazard DfFZ on fire movemenL
The concept that DfPZs may have multiple nonfjre ben-
efits emphasizes the point that strategic fuel management is
an integral component of overall ecosystem management. It
also argues for focusing a large proportion of overall man-
agement efforts in the short term on planning and implement-
ing a sound DFPZ network.
Factors to Be Considered in
Prioritizing DFPZ Locations
In the next sections we present a number of factors that should
be considered in designing a OFPZ network. We do not at.
tempt to set priorities among these factors-to presume, for
example, that values should be weighted more heavily than
historical fire occurrence or that one value is more important
than another value. Such prioritization is best left to local
managers using local fire planning and other information.
"Biggest Bang for the Buck." This concept says, in essence,
" All else being equal, do the cheapest, easiest areas first."
Some stands already may be in an open, low-fuel condi..
tion because of recent management activities. Other areas.
such as rocky outcrops and relatively bare ridges, may pro"
vide natural barriers to the spread of fire. Where it makes
sense strategically to do so, such areas should be incorpo.-
rated into a DFPZ network.
For areas requiring some degree of treatment to be suit..
able as a DFPZ, we suggest that those areas sometimes con..
sidered "most in need of treatment" -i.e.. dense stands and
heavy fuels-should not necessarily be given high priority.
Their costs per unit area may be quite high. This subject can.,
and should, be debated, Our feeling, however, is that from a
strategic standpoint, it seems advisable to treat first those ar..
eas that currenLly would not function effectively as a OFPZ
bllt that could be brought to acceptable standards most quickly
and inexpensively. Thus a greater total length of effective
OFPZ could become functional for a given cost or in a given
period of time. That larger treated area of DFPZ also would
be more likely itself to survive in the event of a severe fire.
Some afeas may be acceptably open but require surface fuel
treatment. Prescribed burning may be the most desirable and
cost-effective option. More often, some thinning is likely to
be necessary. Except in areas where they are precluded for
various reasons, cuttings (preferably with utilization of cut
trees) generally provide a more efficient route to desired foro'
est structures than prescribed burns. Where thinning is
needed. the "biggest bang for the buck" principie may trans..
late to giving priority to multiproduct sales that are economi-
cally self-sustaining by removing some sawtimber to pay for
the removal of smaller trees.
Other examples of locations or conditions that might be
given priority under this principle include (I) accessible ar..
eas with relatively gentle terrain and (2) areas with a signifi..
cant component ol relatively large pine or Douglas fir trees.
-... r'
..... "f ......-...~..-.~,,'.. O_""_..,......_"..._~."
1484
VOLUME II, CHAPTER 56
.A.n additional benef:t of the "biggest bang for the buck"
principle may be in morc quickly developing demonstration
areas or other examples of successful implementation of
DFl'Zs Such areas may be valuable for building and sustain-
ing trust and support for strategic fuel management.
Historical Fire Occurrence and Risk. A major consideration
in locating DFPZs on the landscape shouid be the broad zones
within the Sierra Nevada that have experienced the highest
occurrence of large fires du ring this centu ry-reflecting a com-
bination of rel"ti':ely high risk and high hazard. McKelvey
and Busse (1996) found a strong elevational trend in the oc-
curnmce of twentieth-century fires in Sierran national forests.
The frequency (percentage of area burned at least once) of
large fires was highest below 1.000 m (3.300 ft) elevation and
dropped fairiy rapidly ~t higher elevations. This elevation
zone corresponds generally with the foothill vegetation types
and lower coniferous forests. II is consistent with observa-
tions by others that the highest twentieth-century fire occur-
rence in Sierra Nevada forests has been in the west-side pine
and pine-mixed conifer types and in the east-side pine type
(LaBoa and Hermit 1995; U.S. Forest Service 1995;
Weatherspoon et al. 1992).
This information suggests a fairly simple guideline for ac-
counting for historical fire occurrence: all else being equal.
and in the absence of more site-specific fire-occurrence infor.
mation. begin establishing d DFPZ network at the lowest el-
evations of ponderosa or Jeffrey pine forests and work upward
mto the mixed conifer type. In the general forest zone-i.e..
away from settlements or other hish-value areas-true fir and
other upper montane types probably have low priority for a
DFPZ network from the standpoint of wildfire control. Cer-
tainly other management objectives, however, may call for
zones of more open forest conditions than those common in
most locations today.
Where managers have good "landscape-specific" data on
fire-occurrence, it of course should be weighed more heavily
than regionwidr, trends. Local fire data also may indicate the
direction of prevailing winds that accompany extreme
weather events and I or large fires; this information should
be used in planning DFPZ locations. Current and projected
information on risk-i.e., ignition sources-should be con-
sidered as well. For example. DFPZs should have a role in
isolating heavily traveled transportation corridors and other
areas where ignitions historicaJly have been high. This cer-
tainly applies to urban-wildland intermix areas, which are
discussed next.
Urban-Wildland Intermix Areas. DFPZs have a potential ben-
efit as protective buffers around high-value locations. l:rban.
wildland intermix aeeas ;up prominent in this regard. A
protective buffer should help reduce the incidence of fires
moving from wildlands into these high-value areas and (from
the risk standpoint) also reduce the movement into wildland
arpas of fires initiating in intermix areas. These reasons, along
with the fact that most populated areas in the Sierra Nevada
lie within the elevation zone most frequently burned during
the twentieth century (Greenwood 1995; McKelvey and Busse
]996), give a high overall priority to strategic fuel manage-
ment in urban-wildland intermix areas.
As compared with DH'Zs elsewhere, in forested intermix
areas it may be desirable to focus more on nonfire silvicul-
tural treatment methods in order to minimize concerns about
smoke and potential escapes. In woodland and chaparral veg-
etation types, however, prescribed burning may be the most
practical treatment approach except for limited areas of me-
chanical treatment. Opportunities may exist for the Califor-
nia Department of Forestry and Fire Protection's Vegetation
Management Program (HuSo:l.ri and McKelvey 1996) to de-
velop DFPZs near urban-wildland intermix areas in conjunc-
tion with some of its prescribed burning in foothill vegetation
types.
The need to deal with fire and fuel issues in intermix areas
is confounded by the considerable complexity of those issues.
The physical problems associated with the juxtaposition of
people, personal property. and wildlands are compounded
by an array of problems linked to political and institutional
conditions, multiple and diverse ownerships. and a wide
range in understanding and attitude.
Any overall fuel strategy for urban-wildland intermix ar-
eas must begin with the use of appropriate fire-safe practices
by individual property owners. Prominent among those prac-
tices are adequate clearance between structures and flam-
mable vegetation and the use of fire-resistant roofing and other
fire-safe construction practices (Davis 1990). Part of the pro-
cess of achieving better compliance with fire-safe regulations
is Simply education of property owners-necessarily an on-
going task. Another part may involve stronger incentives,
including significant fines for noncompliance, revision of in-
surance premiums and insurability requirements (Davis 1990),
and possibly increased tax rates, to reflect more accurately
the risk of fire loss in wildland settings as modified by per-
sonal fire-safe practices.
Cooperative efforts to reduce hazard within and around
communities represent another critical component of fuel
management in intermix areas, Partnerships that include lo-
cal governments, local landowners, community groups,
bioregional councils, and, as appropriate. state and federal
agencies could be effective. Fostering such cooperative efforts
is a high priority for the recently formed California Fire Strat-
egies Committee. Sponsored by the California Resources
Agency, the committee consists of representatives of a wide
array of government and private entities with a common in-
terest in dealing effectively with California's wildfire prob-
lems. Members have adopted an ambitious set of action items
in support of the committee's mission "to reduce the risk of
catastrophic fire for the protection of Californians and the
natural envi ronment."
Fuel-management activities in urban-wildland intermix
areas should be coordinated with similar activities on nearby
. ,..._.~'.......
1485
Landscap,,-Level Strategies (or Forest Fuel Management
national forest or other public land and ",'ith activities of large
private landowners. In a recent strategic assessment of fire
management in the U.s. Forest Service, Bacon and colleagues
(19<J5) proposed that priority for hazard mitigation on national
forests in intermix areas be placed on areas where adjacent
landowners agrei! to participate with the U.S. Forest Service
in fuel management and other fire-safety projects. While de-
signing and implementing an effective DFPZ network in and
around complex intermix areas often will not be easy, it will
be greatly facilitated by effective cross-ownership coopera-
tive efforts.
Concems about intermix areas do not stop with current
conditions. Population in Sierran foothill areas is projected to
continue rapid growth (Duane 1996). An important potential
set of solutions related to fire issues rest; with state and local
officials, including legislators and county planning and zon-
ing commissioners, who should implement appropriate limi-
tations and disincentives for new construction in
high-fire-hazard areas_
Fire-related connectionS between urbanized areas and
nearby wildlands go beyond the potentia! spread of fire from
one area to the other. Increasingly in recent years, federal
wildland fire-control agencies have been put into the posi-
tion of having to assume responsibility for structure protec-
tion during major wildfires (Bacon et a\. 1995; Husari and
McKelvey 19%). This imposes costs on other landowners and
thE: general public in two ways: (1) Taxpayers at large pay for
these fire-protection services, and (2) losses to natural re-
sources on public lands increase when these forces are di-
verted to structure protection (Oa vis 1990). Bacon and
colleagues (1995, 4) proposed a redefinition of responsibili-
ties: "(1) fire protection on State and private lands is the re-
sponsibility of State and local governments, (2) homeowners
have a personal responsibility to practice fire safety, (3) the
role of the Forest Service is stewardship of adjacent National
Forests, cooperative assistance to State and local fire organi-
zations, and cooperative suppression during fire emergen.
cies." They suggested two general approaches for the u.s.
Forest Service in response to these responsibilities: (1) The
us Forest Service would phase out of responsibility for di-
rect initial attack in urbanized areas. Existing protection agree-
ments would be renegotiated to reflect this change.
Cooperative fire-protection programs would be expanded to
facilitate state efforts to take on the additional work. (2) Pro-
tection priorities would be changed from the present order of
life first, properly second, and resources th.ird, to life first,
followed by property and resources valued on a par. These
recommendations are consistent with policy changes for fed-
eral agencies proposed in the Federal Wildland Fire Manage-
ment Policy and Program Review (U.s. Department of the
[nterior and U.s. Department of Agriculture 1995). Bacon and
colleagues (1995) also recommended that opportunities be
sought for land exchanges that would improve the ability to
manage fire in urban-wildland intermix areas.
Other High-Value Areas. A number of other kinds of high-
value areas may warrant buffering with DFPZs--e.g., areas
of late-successional emphasis (Franklin et al. 1996),
biodiversity management areas (Davis et al. 1996), and plan-
tations (Wilson 1977). Such protection may be particularly
useful when fuel reduction within the high-value area itself
is undesirable or infeasible because of the nature of the value
being emphasized and! or high costs of treatment. It might
be desirable to treat a high-value area with prescribed fire,
for example, but appropriated funds might be inadequate,
especially since initial reintroduction of fire without mechani-
cal pretreatment can be rather expensive in some places. In
contrast, a OFPZ outside the high-value area could be self-
financing through removal of a product. It also could aid in
the subsequent reintroduction of fire into the area,
OFPZs need not be placed immediately adjacent to a high-
value area. In most cases it probably is desirable to back off to
a location that makes sense for other reasons, as discussed
earlier--e.g., a ridge or an upper south slope, along a road,
relatively cheap to treat.
Using a DFPZ to provide a buffer between adjacent areas
may also be useful where management emphases or intensi-
ties, rather than values per se, differ, For example, it might be
desirable to provide such a separation between an area
managed primarily for natural values, including use of PNF,
and an adjacent area managed primarily for commodities.
This might or might not be associated with an ownership
boundary.
Fire Hazard. Hazard is another factor that needs to be con-
sidered in locating DFPZs. All else being equal. a landscape
dominated by continuous heavy fuels is in greater need of
zones of fuel discontinuity than One with light fuels. Insofar
as possible, however, actual DFPZ location should favor rela-
tively open, low-fuel sites in order to treat more area with the
a vailable funds, In other words, DFPZs should separate high-
hazard areas but not necessarily be built through them.
~t is reasonable to assume that high-hazard areas may be
relatively more of a concem with respect to the potential for
high-severity wildfires in drier years. In such years, a higher
percentage of the total fuel profile (including live fuels) be-
comes readily available for combustion. Drier fuels and drier
microclimate near the forest floor favor easier ignition and
f aster fire spread. The significance of such changes in dry years
is increased by the preponderance of dry years in the past ten
years and by the fact that such years may be more nearly the
norm when viewed on a time scale of centuries (Graum-
hch 1993).
Professional and Public Support. Many forest-management
activities are controversial, among resource professionals as
well as various segments of the public, We believe that creat-
ing and maintaining DFPZs may offer multiple benefits, in-
cluding reduced wildfire hazard, improved forest health, and
utilization of excess forest biomass, which in most cases
.... .. t"
l'~"---"'''-'''''''''''''''"- """---1--. ,--~-~
1486
VOLUME II, CHAF TER 56
should outweigh potential ecosystem damage. Adequately
explained and understood, therefore, DFPZs should be rea.
sonably well supported. Nevertheless, some Meas proposed
for DFPZs may be controversial. All else being equal, we sug-
gest that, at least initially, creation of DFPZ networks be con.
centrated in areas where professional and public support are
relatively high and disagreement relatively low. In most cases,
more than enough work will need to be done to permit ac-
tivities to be focused in these areas and to defer more contro-
versial work. Well-designed and properly implemented early
DFPZs may generate additional support for further develop-
ment of a strategic fuel-management program.
Rate of Implementation and Practicability
We believe that, in the short term, planning and implement-
ing DFPZ networks should have a high priority for manage-
ment of low- to middle-elevation Sierran forests and
appropriate portions of foothill woodland and chaparral
types. Ideally, these networks should be in place within ten
years, Implementing these networks will require a great deal
of concentrated and cooperative effort. [t also may well reo
quire "departures" from nondeclining even flow of timber
volume under the National Forest Management Act. Poten-
tial benefits could be substantial, however, in terms of strate-
gic reduction of wildfire hazard, improvement in forest
conditiorLs, and increases in economic and social well-being
in fore5t-based communities.
By any measure, implementing a rangewide system of
DFPZs within ten (or even twenty) ye:!rs is a formidable un-
dertaking. Responsible managers must be concerned with the
feasibility and potential value of such a task compared with
alternative management actions. Given the high priority of
fire-protection and restoration issues in Sierran forests and
the multiple benefits (cited earlier) that might be anticipated
from DFPZ networks, a number of managers may judge such
networks to have a high overall priority for management.
To be achievable, implementation of a DFPZ system can-
not be viewed Simply as d fire function or goal. Rather, it
should be considered a multi resource or ecosystem manage-
ment goal, with much of the overall activity of the manage-
ment unit in the short term being integrated with and focused
on planning and implementing a sound DFPZ network. Simi-
larly, multifunction funding would improve the feasibility of
accomplishing this task.
How will we pay for all the silviculture and fuel manage-
ment that will be necessary to implement DFPZ networks.
given the large areas that need to be treated? Considering his-
toricallevels of funding and current di reclions of federal bud-
gets, it seems highly unlikely that federal appropriated
funds-even from multiple functions-will be adequate. And
managers may decide that most of the limited appropriated
funds for fuel treatment are best spent to support prescribed
burning of natural fuels in areas with special emphases on
reestablishing natural processes (see the following section).
Thus, truly significant progress on DFPZs and other large-
scale fuel treatments will have to be the result of economi-
cally self-sustaining activities. Yet much of the needed treat.
ment involves removal of small trees that often have marginal
or negative market value. Part of the solution may come from
multiproduct sales, in which sawtimber and other high-value
products subsidize the removal of lower value material. One
of the challenges for managers will be to locate and design
multiproduct or other sales in ways that make them economi-
cally viable. In addition, however, it probably will be impor-
tant to support the establishment of particleboard or other
plants capable of generating value from small trees. Public
land managers and private entrepreneurs need to discuss
whether and how it may be possible to provide sufficient as-
surances of a continuing supply of biomass from public lands
(e.g., for several decades) to warrant the capital investment
in such plants. Research and development efforts also are
needed to develop more efficient technology for harvesting
and processing small material and new markets for utilizing
it (Lambert 1994).
Most resource professionals would agree that fuel reduc-
tion and thinning of overly-dense stands are high-priority
needs in most pine and mixed-conifer forests of the Sierra
Nevada. These are precisely the kinds of activities envisioned
for DFPZs, with the added proviso that they be placed in stra-
tegically logical locations. It is important to note, therefore,
that the major barriers to DFPZ implementation-e.g., eco-
nomic viability of small trees and maintenance of treated ar-
eas-are not unique to DFPZs: they apply much more widely.
Tl)us, these barriers must be resolved in any case if large-scale
thinning and fuel management are to be implemented. The
contiguous nature and relative accessibility of DFPZs, how-
ever, may help to lessen the severity of these problems in
DFPZs.
Enhanced Use of Fire
Restoring the many functions of fire as an ecosystem process
can be accomplished fully only by using fire, Alternative and
supplementary methods must playa large part in needed res-
toration, but they can substitute only partially for fire
(Weatherspoon 1996). In the context of goal 2, therefore, we
believe that a considerably expanded use of prescribed fire
can and should play an important role in the management of
Sierra Nevada ecosystems (Husari and McKelvey 1996; Mutch
et aL 1993).
In some portions of the Sierra Nevada, especially higher
elevation areas, large high..severity fires are not much of a
concern. Thus neither goal 1 nor D[:PZs are particularly ap-
picabe. Many such areas are located in national parks and
wilderness areas, but substantial additional acreage of red fir
and other high-elevation vegetation types fits in this category.
Our suggestion in these areas would be to extend the use of
prescribed natural fire (PNF) as much as possible (including
appropriate areas outside parks and wildernesses) and to
augment PNF with management-ignited prescribed fires
1487
Landscape-Level Strategies for Forest Fuel Management
(M[PE') as needed to reestablish a near-natural distribution of
fi re f requencies_
M [PF also should become a key part of the management of
other areas in which restoration of natural processes is a ma-
jor management objective. Examples of such areas might in-
clude areas of late. successional emphasis (Franklin et al. 1996),
biodiversity management areas (Davis et al. 1996), and re-
search natural areas.
As indicated earlier, DFPZs require periodic maintenance
to retain their effectiveness, and prescribed fire often will be
the treatment of choice. Since the st.ructure and composition
of DFPZs are intended to be closer to presettlement condi-
tions than most other areas of the landscape, it would seem
logical for fire to assume a dual role there-maintenance of
the low-fuel nature of OfPZs and restoration of natural pro-
cesses.
A number of practical and political considerations constrain
the use of both MIPF and PNF on a large scale. Constraints
include risk of escapes, lack of adequate funding, competi-
tion for trained personnel during active wildfire seasons, and
air quality restrictions (Husari and McKelvey 1996; Parsons
1995). The difficulties of applying prescribed fire on a signifi-
cant scale are illustrated by the inability of the prescribed fire
program at Sequoia and Kings Canyon National Parks-cer-
tainly among the most active in the Sierra Nevada~ven to
begin to approach the presettlement fire frequency for the
giant sequoia groves. A National rnteragency Fire Center
study to be undertaken beginning in 1996 will test the feasi-
bility of and constraints on landscape-scale application of
prescribed fire in the Kaweah River drainage of Sequoia Na-
tional Park.
In addition to prescribed burning, significant benefits re-
lated to goal 2 could be achieved by allowing low- and mod-
erate-intensity wildfires to burn. Potentially, many more
burned acres could be achieved by this means than with pre-
scribed fire. The vast majority of ignitions in the Sierra Ne-
vada are suppressed using fast. aggressive control. The
flexibility already existing in present federal fire-management
policy to use alternative suppression responses is rarely ex-
ercised outside the national parks and d few wilderness areas
in the Sierra Nevada (Husari and McKelvey 1996). Fire man-
agers currently are required to select the most economically
efficient suppression option without considering potential
resource benefits of wildfires. Fires that would produce re-
sults most similar to those that occurred under presettlement
conditions are regularly su ppressed while small, because they
are easy and inexpensive to put out. Proposed new federal
policies (U.S. Department of the Interior and U.S. Department
of Agriculture 1995) would permit wi Idfires to be "managed"
if thpy meet resource objectives.
More flexible use of appropriate su ppression responses,
possIble use of managed wildfires to meet resource objectives,
and expanded use of both MIPF and I'NF jointly offer con-
siderable opportunities for managers to restore more of the
ecosystem functions of fire to the Sierra Nevada. All of these
opportun.ities should be enhanced as forest and fuel condi-
tions are improved over time. It should be recognized that in
those areas from which fire continues to be excluded, for
whatever reasons, some ecosystem components and processes
will depart significantly from their natural range of variabil-
ity, with unknown consequences.
Areawide Fuel Treatments
The development of DFPZs described in this chapter is a logi-
cal place to begin, but it is intended to be only a first step
toward achieving the three goals of the fuel-management
strategy discussed earlier. DFPZs should help to limit the spa-
tial extent of severe fires (van Wagtendonk 1996; Sessions et
al. 1996); however, they will not reduce the susceptibility of
the intervening landscape areas to severe fire effects, nor will
they improve forest health or restore more nearly natural pro-
cesses in those intervening areas. l.andscape mosaics and
vegetative profiles will need to be managed on broader scales,
using mainly silvicultural cuttings and fire, to achieve desired
forest conditions and processes (Mutch et al. 1993).
The implementation of areawide landscape treatments
should be significantly facilitated by using previously estab-
lished DFPZ networks as anchor lines from which to build
out. Factors considered in prioritizing DFPZ locations, dis-
cussed earlier, may also be useful as guides for prioritizing
areawide treatments. From the standpoint of topography, for
example, middle and upper south and west aspects on rela-
tively gentle (machine-operable) slopes may be logicalloca-
tions for early work.
RESEARCH AND ADAPTIVE
MANAGEMENT NEEDS
The Role of Adaptive Management
Ecosystem management is increasingly espoused as a guid-
ing concept for managing public lands (Jensen and Bourgeron
1994; Manley et al. 1995; Salwasser 1994). Managing for eco-
system integrity and sustainability, however, is more diffi-
cult and fraught with more uncertainties than managing for
a set of specific outputs. We have much to learn_ For many
reasons, including the complexity and variability of forested
ecosystems and the broad spatiotemporal scale that provides
the conlext for ecosystem management. traditional research
cannot provide all the answers. Scientists, managers, and in-
terested members of the public must work together as part-
ners in a process of learning by doing-i.e., adaptive
ecosystem management (Everett et al. 1994; Mutch et al. 1993;
Walters and Holling 1990)_
A key concept of adaptive management is that we cannot
wait for perfect information, because we will never have it.
Despite the uncertainties, we must move forward with man-
"'''''.n._u~-----,..~..,_;,,,.,~.,..~,,,,,...,.."~,,,..,..,~.,,.., "'_"" .
1488
VOLUME II, CHAPTER 56
aging for sustainable ecosystems using the best information
we have, knowing that with time we willleam more and be
able to manage more intelligently.
The subject of landscape-level fuel-management strategies
is certainly appropriate to address through adaptive manage-
ment. For example, we can make educated d;.SUmptions about
how a network of DFPZs might help to reduce high-severity
fires and contribute to desired conditions and landscape di-
versity. Only through monitoring, experience, and time, how,
ever, will we know the validity of those assumptions. Only
through adaptive management will we learn what locations,
target conditions, and treatment schedules for implementing
a DFPZ network will work for what kinds of landscapes--or
whether a DFPZ network makes sense in the first place.
Similarly, we know that the ecosystem functions of frequent
low- to moderate-severity fire have been largely lost from Si-
erran forests, Restoring these functions can be accomplished
fully only by using fire. Yet in many areas silvicultural tech-
niques and other fire "surrogates" are needed in addition to
or in lieu of fire to accomplish needed restoration
(Weatherspoon 1996). Theextent to which natural fire regimes
can or should be emulated, and the consequences for long-
term ecosystem viability of alternative approaches to usinr;
fire versus fire surrogates on large scales, will become clear
only through carefully designed research and adaptive man-
agement.
A GIS Database in Support of Fuel-
Management Strategies and Adaptive
Ecosystem Management
Good information is essential to intelligent planning of spe-
cific fuel-management strategies in the short term, and to as-
sessing the effectiveness of those strategies (and adjusting
subsequent management as appropriate) in the mid to long
term. An integrated GIS database can provide a good focus
for this information, The concept is quite simple and logical,
given the increasingly GlS-oriented world in which we oper-
ate. Actually accomplishing the monitoring and other data
collection necessary to make it fully functiom.l may be an-
other matter, From a fire standpoint, it probably makes sense
to use the same general priorities for this data collection as
discussed earlier for locating DFPZs.
In the following sections we indicate some thoughts about
the directions in which we should be moving with GIS data-
bases. We are not suggesting a standalone {ire and fuei GIS.
Rather, the following kinds of data needed to support file and
fuel decision making would be integrated into a larger data-
base to inform overall land management.
Management Direction
Management objectives and guidelines, including those spe-
cific to fire and fuel management, should be indicated by area.
Vegetation and Fuels Data
The need for data on vegetation and fuels is basic and well
recogni:.o:ed. (Much of the living vegetation is fuel, of course,
but to simplify the discussion here we list vegetation and fu-
els separately,) Mapping should utilize the best sampling strat-
egies combining remote sensing imagery (perhaps at several
scales) and ground truthing, The reliability of existing veg-
etation maps should be verified before they are incorporated
into the database. Fire-relevant attributes of vegetation (in-
cluding understory composition and structure, and vertical
and horizontal continuity) need to be characterized ad-
equately. Similarly, surface fuels should be described, utiliz-
ing field-verified vegetation/ fuels correlations to the extent
feasible.
Since vegetation and fuels change over time, the dynamics
occurring naturally through succession and growth must be
dealt with using models combined with periodic field evalu-
ations. Natural and human-caused disturbances also change
vegetation and fuels, from a little to a lot, The database must
be updated as needed to reflect these disturbance-induced
changes. To account for these dynamics adequately, we need
to go beyond traditional spatial GIS to incorporate new con-
cepts in spatiotemporal GIS (peuquet 1994; Skinner et aL 1992).
Management Activities and Other Disturbances
For our land management a<.:tivities (including prescribed fire
and fuel management) that significantly alter vegetation and
fuels, monitoring must be carried out to determine the extent
to which management objectives were met and the effects on
vegetation, fuels, and other key ecosystem components. The
GIS database should be updated to indicate the nature, date,
spatial extent, and costs of the activity and the resulting spa-
tially referenced vegetation and fuels, "Natural" or unplanned
disturbances-especially wildfires-must also be incorpo-
rated into the database. Wildfires should be mapped by se-
verity classes and key fire effects. To the extent allowed by
available data, burning conditions at different times and
places on a fire, along with suppression actions and costs,
also should be entered. After postfire activities are completed,
the new vegetation/fuel complex should become part of the
database. To permit long-term evaluation of fires and man-
agement activities, however, it is important to maintain-not
discard-prefire vegetation and fuel data. A spatiotemporal
GIS would serve this purpose more efficiently than the sys-
tems generally available today (Peuquet and Niu 1995;
Peuquet et aL 1992),
Other Fire-Related Data
Risk (historical fire occurrence and historical and projected
ignition patterns), values at risk (for both populated and wild-
land areas), suppression capabilities, and any other spatially
relevant fire-planning data should be included in the data-
base. It may well be advisable for public and private land-
owners to cooperate in establishing data standards and
~~
1489
Landscape-Level Strategies for Forest Fuel Management
protocols applicable to iire and iuels, thereby permitting data
sharing, cross-ownership analyses. and the likf, when mutu-
ally ,ksirable.
Benefits of the GIS Database
ThIS kmd of database, in ever. a rudimentary form. certainly
wlll permit better planning ior iuel-manage,Tlent strategies
As data are Improved and accumulated over time, moreover,
its value will increase. We will begin to have the data neces-
sary to relate wildfire severity and effects to prior manage-
ment activities (induding fuel treatments), fuel conditions,
and site and stand characteristics (e.g., Weatherspoon and
Skinner 1995). Over time, as more wildfires are documented,
au r abili ty to assess the efficacy and cost-effecti veness of vari-
ous fuel-management strategies in terms of both behavior and
eHects of subsequent wildfires and suppression costs will
grow. We also will be able to evaluate trade-oifs involving
environmental effects of the treatments themselves. We will
be much better able to leam by doing and monitoring-the
essence of adaptive management (Everett et a1. 1994; Mutch
et .,1. 1993; Walters and Holling 1990).
Establishing and maintaining an acc.urate GIS database of
this kind will require considerable effort and commitment on
the part of managers and landowners. It will be a long-term.
ongoing process. Many other resource benefits '...-ill accrue.
however, and in fact it is difficult to see how real ecosystem
management in a fire-proI'.e region such as the Sierra Nevada
will be feasible without such a database.
CONCLUSIONS
Fire has been an important component of most Sierran eco-
systems (or thousands of years (Skinner and Chang 1996).
However, human activities since European settlement, along
with variation in climate, have profoundly altered fire re-
gimes, leading to anomalous vegetation and fuel conditions
throughout much of the range. Two major fire-related "prob-
lems" have developed ill the Sierra Nevada: (1) too much high-
severity fire and the potentia! for much more of the same and
(2) too little low- to moderate-severity fire, along with a vari-
ety of ecological changes attributable at least in part to this
deficiency. Clearly, these are not just "fire problems." They
influence virtually all resources and values in the Sierra Ne-
vada and cut across all of SNEP's subject areas.
Given the realities of our modem civiliza.tion, we must rec-
ogniLe that the changes in ecosystem conditions and in the
role of fire are only partially reversible. We can and should
reduce the exrent of large, severe wildfires. However, such
fires will continue at an appreciable level (almost certainly at
a higher level than in the presettlement period) into the fore-
seeable future. We can and should restore more of the ecosys-
tem functions of low- and moderate-severity fire, utilizing
such fire to the extent feasible. It is inconceivable, however,
that fire in its presettlement extent and frequencies could be
restored fully to the Sierra Nevada.
Nevertheless, a partial solution is far beller than no solu-
tion at all or than a continuing deterioration of Sierran forests
from a fire standpoint. There is much that we as land stew-
ards can and should do. The two fire-related problems cited
earlier can be translated into the three strategic goals that have
been discussed in this chapter. Making significant progress
toward these goals will require long-tenn vision, commitment.
and cooperation across a broad spectrum of land-management
agencies and other entities. The problems were created over
a long period of time, and they certainly cannot be solved
overnight. Progress also will require landscape-scale strate-
gic thinking, planning, and implementation. This chapter has
provided some ideas for managers to consider as they de-
velop their own landscape-specific plans.
We have much to learn as we move more fully into an era
of ecosystem management, including strategic fuel manage-
ment. Adaptive management must be an integral part of ollr
management activities, as discussed earlier. It is important to
note in this regard that we do not have to have all the an-
swers before beginning needed restoration work. We know
enough at this point to recognize that current conditions in
most low- to middle-elevation forests of the Sierra Nevada
are unacceptable in terms of wildfire hazard, diversity, and
sustainability. Regardless of the extent to which presettlement
conditions are used as a guide to desired conditions, most
informed people would agree that these forests generally
should be less dense, have less fuels, and have more large
trees. Even if we have not precisely identified target condi-
tions, we certainly know the direction in which we should
begin moving. That beginning alone will require a large mea-
Sure of commitment and hard work, We can adjust along the
way as we learn more and become better able to define de-
sired conditions for Sierran forests.
ACKNOWLEDGMENTS
We would like to thank Russ Jones of the SNEP GIS staff for
his considerable efforts in providing updated data related to
human-caused and lightning fireson Sierra Nevada national
forests during the twentieth century. We gratefully acknowl-
edge the Cal Owl EfS team for providing the original fire data
used by Russ. We also want to thank the following individu-
als for valuable comments on earlier versions of the manu-
script: r. BrenchJey-Jackson, J, Fites, O. Fullmer, S. Husari, J.
LaBoa, D. Leisz, K. McKelvey, 0, Parsons, R. Powers, J. Reiss,
E. Roberson, L. Salazar, 1- Tappeiner, G. Terhune, 1- Wood, and
two anonymous reviewers, Finally, we sincerely appreciate
the efforts and contributions of all those who participated in
the SNEP Fuels Management Strategies Workshop in March
1995 (Fleming 1996).
.., , -'-"-'--""-''''''''''-'- ..."----t--......._........<.'"
1490
VOLUME II. CHAPTER 56
REFERENCES
Agee, J. K., and R. L. Edll\onds 19Y2. Forest protectIOn gUIdelines fm
the northern spotted owL h, Recovery pl.m for the I'orthem spotted
0wl: Appendix r. Wasnington, DC: Us. Depaltment of the Interior.
Ar.derson, M. K., and M.l. Moratto 1996. t\:ative Ameflcdnland-lIs<'
practices and ecologi,:al impacts. [n Sierra ;\Ievada F.cos)'stem Prolect.
Final report to Congress, voL II, chap 9. Davis: Umversity of
California, Centers tor Water and Wildland Resources.
Anderson, P. J., R. E. Martin, and'. K. Gilless. 1991. DeciSIon analysts
in the evaluation of wildfire hazard reduction by prescribed
burning in the wildland .urban interface. In Proceedings of the 11 th
conference on fire and forest meteorology, 291-98. Washington, DC:
Society of American Foresters.
Amo, S. E, and J. K. Brown. 1989. Managing fire in our forests-
Time for a new initiative. Joumal of Forestry 87:44-46.
Amo, S. E, and R. D. Ottmar. 1994. Reducing hazard for catastrophic
firE:. In Restoration of stressed sites. and processes, compiled by R. L.
Everett, 18-19. Vol. 4 uf Easl,ide foresl ecosystem health asse<smcnt.
Genf'raJ Technical Report PNW-GrR-330. Portland. OR: U.S. Forest
Service. PacHic Northwest Research StatIOn.
Bacon, D.. S, Conard. G. Ferry, P. Frey, S. Heywood, D. Lamb, D.
MacCleery, D. Mangan, S. Pedigo, D. Radloff. J. Saveland. C
Swanson, and H. Tran. 1995. Strategic assessment of fire
management in the U.S. Forest Servicl'. Unpublished report.
Washington, DC: U.s. Forest Service.
Biehl, G. 1995. Untitled presentation at S:'>:EP Fuels Managl'ment
Strategies Workshop, Davis, California, March.
Botti. S. J.. and H. T. Nichols. 1995. Availability ot tire rl'sourcl'S =d
funding for prescribed natural fire programs in the N,Hional Park
Service. In Procee<lings: Symposium on fire in wildemess and park
managl'ment, technical coordination by J K. ~rown, R. W. Mutch.
C. W. Spoon, and R. H. Wakimoto, 94-103. General TechnIcal
Report INT.GTR.320. Ogden, UT: US. Forest SerVIce,
Intermountain Research Station.
Chang, C. 1996. Ecosystem responses to fire and variations in fire
regimes. In Sierra Nevada Ecosystem Project: Final report to Congress,
vol. II, chap. 39. Davis: University of California, Centl'rs for Watl'r
and Wildland Resources.
Cohan. 0., S. Haas, and P. J. Roussopoulos. 1983. Decision analysis
of silvicultural prescriptions and fuel managl'ment practices on
an intensively managed commercial forest. Forest Science 29:
858-70.
Davis. J. B. 1990. The wildland-urban interfarl': Paradise or
battleground? Journal of Forestry 81;;26-31
Davis, F. W., D. M. Stoms. R. L. Church, W. J akin. and K. N. Johnson.
1996. Selecting biodiversity manageml'nt areas. In Sierra Nevada
Ecosystem Project Final report to Congress, vol. II, chap. 58. Davis:
University of California, Centers for Watl::[ and Wildland
Resources.
Duane. T. P. 1996. Human settlem..nt, 1850-2040. L" Sierra Nevada
Ecosystem Project: Final report to Congress, vol. n, chap. 11. Davis:
University of California, Cl'nters for Watl'r and Wildland
ResourceS.
Everett. R., C. Oliver, J. Saveland, P. Hl'ssburg. N. Diaz, and L. Irwin
1994. Adaptivl' ecosystem management. In Ecusystem management:
Principles and applications. edited by M. E. Jensen and P. S.
Bourgeron. 340--54. Vol. 2 of Easls.de forest ecosystem health
/lSsessment. General [echnical Report PNW-GTR-318. Portland, OR
U.s. Forest Service, Pacific Northwest Research Station.
Ferrell. G. T. 1996. The influence of insect pests and pathogens on
Sierra forests. [I'. Sierra Nl'vada Ecosystem Project: Final report to
Congress, vol. ll, chap. 45. Davis: University of California, Centn>
for Water and Wildland Resources.
Fites, J. A. 1995. Untitled prl'sentation at SNEP Fuels Mdnagcmenl
Strategies Workshop, Davis, California, March.
Fll'ming, E. 1996. Compilation of workshops contributing to Si.'rra
Nevada assessments. [n Sierra Nevada Erosystem Project: Final report
to Congress. vol. lll. Davis: University of Cahfomia, Centers for
Water and Wildland Resources.
Franklin, J. E, D. Graber, K. N. Johnson, J. A. Fites-Kaufmann,
K. Ml'nning, D. Parsons. J. Sessions. T. A. Spies, 1. c. Tappeiner,
and D. A. Thornburgh. 1996. Comparison of alternative late
successional conservation strategies. In Sil'rra Nevada Ecosystem
Project: Final report to Congress. Davis: University of California,
Centers for Water and Wildland Resources.
Graber. D. 1996. Status of terrestrial vertebrates. [n Sierra Nevada
Ecosystem Project: Final report to Congress. vol. ll, chap. 25. Davis.
University of California, Centers for Watl'r and Wildl.,nJ
Resources.
Grah, R. F., and A. Long. 1971. California fuelbreaks: Costs and
benefits. Journal of Forestry 69:89-93.
Graumlich, L. J. 1993. A 1,000-year record of temperature and
precipitation in the Sierra Nevada. Quaternary Research 39:249-55.
Green. L. R. 1977. Fuelbreaksandotherfuel modification for wildland fire
control. Agricultural Handbook 499. Washington, DC: U.S. Forl'st
Service.
Green, L. R., and H. E. Schimke. 1971. Guides for fuel-breaks in the Sierra
Nevada mixed-conifer type. Station Report. Berkeley, CA: U.s. Forest
Service, Pacific Southwest Research Station.
Greenwood, G. 1995. How population defines fire-managl'ment
environmeI'ts and firl'-planning options. Presl'ntation at the
conference on fire ecology and fuels managl'ment in California
wildlands, 5-7 April. Sacramento, CA. Sponsored by the Amencan
Fishl'ries Society, California Chapter; Society for Rangl'
Management. California Section; Northern California Society of
American Foresters; and Wildlife Society, Western Section.
Grul'll, G. E. 1994. Perspectives on historical vegetation changl' in
the Sierra Nevada. In Proceedings. 15th annual forest vegetation
management conference, 10-24. Redding. CA: Forest Vegetation
Management Conference.
Harmon, M, E., K. Cromack Jr., and B. Smith. 1987. Coarse woody
debris in mixed-conifer forests, Sequoia National Park. California.
Canadian Joumal of Forest Research 17:1265-72.
Helms, J. A., and J. C. Tappeiner. 1996. Silviculture in the Sierra. In
Sierra Nevada Erosystem Project Final report to Congress, vol. II, chap
15. Davis: University of California. Centers for Water and Wildland
Resources.
Hirsch. S. N., D. L. Radloff, W. C. Schopfer, M. L. Wolfe, and R. F.
Yancik. 1981. The activity fuel appraisal procl'ss: Instructions and
examples. General Technical Report RM-83. Fort Collins, CO: US
Forest Service, Rocky Mountain Research Station.
Husari. S. J.. and K. S. McKelvey. 1996. Fire management policies
and programs. [n Sierra Nevada Ecosystem Project Final report to
Congress. vol. II. chap. 40. Davis: University of California, Centers
for Water and Wildland Resources.
James, T. E. 1994. Catastrophic fire vulnerability rating system: "A
process." Unpublished report, Groveland Ranger District,
Stanislaus National Forest. 30 April.
Jensen, M. E.. and P. S. Bourgeron, l'ds. 1994. Ecosystem management:
1491
Landscape-Level Strategies for Forest Fuel Management
pnnc;ples and applications. Vol. 2 of Eastsidl' fOll'st ecosystem health
assessment. General Techf\lcal Report Pt';W-GTR-318. Portland, OR
U.S. Forest Service, Pacific Northwest Researcn Station.
Kattelmann. R. 1996. Hydrology and water resources. in Sierra Nevada
Ecosystem Project Final report to Congress, vol. [I. chap. 30. Davis:
University of Califurnia, Centers for Water and Wildland
Resources.
Kilgore, A. M. 1973. The ecological role of fire in Sierran conifer forests:
Its application to national park management. Quaternary Research
349&--513.
Kilgore, B. M., and R. W. Sando. 1975. Crown-fire potential in a
sequoia forest after prescribed burning. Forest Science 21:83-87.
LaAoa, ,., and R Hermit. 1995. Strategic fuels planning in the Sierra
t':evada province. White paper prepared for SNEP Fuels
Management Strategies Workshop, Davis, CA, March.
Lambert, M. B. 1994. Establish stable stand structures and inereas"
tr~ growth: New technologies in silviculture. In Restoration of
stressed sites, and processes, compiled by R. L. Everett. 93-96. Vol. 4
ot Eastside forect ecosystem health assessment. General Technical
Report P:"JW-GTR-330. Portland, OR: Us. Forest Service, Pacific
Northwest Research Station.
Ylanley, }. 1995. Untitled presentation at SNEP Fuels Management
Strategies Workshop, Davis, California, March.
Ylanley, P. N., G. E. Brogan, C. Cook, \1. E Flores, D. G. Fullmer, S.
Husari, T. M. Jimerson, L. M. Lux, M. E. McCain, J. A. Rose, G.
Schmitt. J. C. Schuyler. and M. ). Skmner. 1995. Sustaining
ecosystems: A runceptual framework. KS-EM-TP-001. San Francisco,
CA: Us. Forest Service, Pacific Southwest Region.
Mason, R. R., and 8. E. WIckman. 1994. Procedures to reduce
landscape hazard from msect outbreaks. In Restoration of stressed
sites, and processes, compiled by R. L. Everet!. 2G-21. Vol. 40f Eastside
forest ecosystem health a~sessrnent. Ceneral Technical Report PNW-
GTR-330. Portland, OR: U.S. Forest Service, Pacific Northwest
Research Station.
McKelvey, K. S., i\f\.d K. K. Busse. 1996. T'.venneth-century fire patterns
on Forest Service lands. In Sierra Nevada Ecosystem Project Final
report to Congress, vol.lL chap. 41. Davis: University of Califomia,
Centers for Water and Wildland Resources.
McKelvey, K. S., and J. D. Jolmston. 1,}92. Historical perspectives on
forests of the Sierra Nevada and the transverse ranges of Sou L"'ern
California: Forest conditions at the turn of the century. In The
Caliiornia spotted owl: A technical assessment of its current status,
technical coordination by J. Verner, K. 5 McKelvey, 8. R. Noon, R.
I. Gutierrez, G. I. Gould /r., and T W. Beck, 225--46. General
Technical Report GTR-PSW-133. Albany. CA: US. Forest Service.
Paciiic Southwest Research Station.
YlcPherso~c, G. R., D. D. Wade, and C A Phillips. 1990. Glossary of
wildland fire ma.nagement terms used ir. the United States. Washington.
DC: SOCiety of American Foresters
Menke, /., C. Davis, and P. Beesley 1996. Rangeland assessment. In
Sierra Nevada Ecosystem Project: Final report to Congress, vol. HI.
Davis: University of California, Centers for Water and Wildland
Resources.
Mutch, R. W., S. F. Arno, J. K. Arown, C E. Carlson, R. D. Ottmar. and
r. L. Pf'terson. 199:'. Forest health in the Blue 1\10untaiUS: A rnanagement
.;trategy for fire-adapted ecosystems. Ceneral Technical Report PNW-
GTR-31O. Portland, OR: Us. Forest Service, Pacific Northwest
Research Station.
Olson, R. D. 1993. Defensible fuel profile zones. Unpublished report.
I.assen National Forest: US. Forest ServICe.
Olson, R., R. Heinbockel. and S. Abrams. 1995. Technical fuels report.
Unpublished report. Lassen, Plumas, and Tahoe National Forests:
U.5. Forest Service.
Omi, P. N. 1979. Planning future fuelbreak strategies using
mathematical modeling techniques. Environmental Management
3:73-80.
Parsons, D. ). 1995. Restoring fire to giant sequoia groves: What have
we learned in 25 years? !n Proceedings: Symposium on fire in
wilderness a:"l.d park management, technical coordination by J. K.
Brown, R. W. Mutch, C. W. Spoon. and R. H. Wakimoto, 256--58.
General Technical Report INT-GTR-320. Ogden, UT: U.5. Forest
Service. Intermountain Research Station.
Parsons, D. ).. and S. H. DeBenedetti. 1979. Impactoftire suppression
on a mixed-conifer forest. Forest Ecology and Management 2:21-33.
Perkins, J. 1995. Untitled presentation at S:--:EP Fuels Management
Strategies Workshop, Davis, California, March.
Peuquet, D. J. 1994. It's about time: A conceptual framework for the
representation of temporal dynamics in geographic information
systems. An.nals of the Association of American Geographers 84:
441~1.
Peuquet, D. J., R. J. Laacke, and C. N. Skinner. 1'192. Representational
issues involved in the development of a decision support system
for spatial and temporal analysis of forest ecosystems. In Resource
Technology 92, 381-87. VoL 5 of Proceedings of the ASPRS I ACSMI
RB2 conference: Monitoring and mapping global change, Bethesda,
MO: American Society for Photogrammetry and Remote Sensing
and American Congress on Surveying and Mapping.
Peuquet, D. J.. and D. A. Niu. 1995. An event-based spatia-temporal
nata model (ESTDM) for temporal analysis of geographic data.
International Journal of Geographical lnformation Systems 9:7-24.
Pierovich. J. M., E. H. Clarke, S. G. Pickford, and F. R. Ward. 1975.
Forest residues management guidelines for the Pacific Northwest.
General Technical Report PNW-33. Portland, OR: U.S. Forest
Service, Pacific Northwest Research Station.
Quincy Library Group. 1994. Fuels management for fire protec-
tion. Unpublished report. Quincy Library Group position paper,
Quincy, CA.
Radloff. D. 1... and R. F. Yandk. 1983. Decision analysis of prescribed
burning. !n Proceedings of the 7th conference on fire and forest
meteorology, 85-89. Washington, DC: Society of American Foresters_
Radloff, D. 1... R. F. Yancik. and K. G. Walters. 1982. User's guide to the
na lional fucl a ppraisal pmcess. Fort Call ins, CO U.5. Forest Service,
Rocky Mountain Research Station.
Salazar, I.. A., and A. Gonzalez-Caban. 1987. Spatial relationships of
a wildfire, fuelbreaks, and recently burned areas. Western Journal
of Applied Forestry 2:55-58.
Salwasser, l-I. 1994. Ecosystem management: Can it sustain diversity
and productivity. Iou mal of Forestry 92:6-7,9-10.
Sessions, J., K. N. Johnson, D. Sapsis, B. BaMo, and J. T. Gabriel. :996.
Methodology for simulating forest growth, fire effects, timber
harvest, and watershed disturbance under d i!ferent management
regimes. In Sierra Nevada Ecosystem Project Final report to Congress.
Davis: University of California, Centers for Water and Wildland
Resources.
Show, S. S., and E.l. Kotok. 1923. Forest fires in California 1911-1920:
An analytical study. Department Circular 243. Wasnington, DC: US
Department of Agriculture.
Skinner, C. N., and C. Chang. 1996. Fire regimes, past and prese:)t. In
Sierra Nl.'Vada Ecosystem Project Final report to Congress, vol. II. chap.
... _c....----...__._c...__._._
1492
VOLUME II, CHAPTER 56
38. Davis: UniversIty of California. Centers for Water and Wildland
Resources.
Skinner, C. N., R. f. Laacke, and D. J. Peuquet. 1992. Developing a
user-orimted natural resource deo,i"n support system. In Resource
Technology 92, 374-ilO. VoL 5 of Proceedings of the ASPRS/ ACSM/
RT92 conference: Monitorir.g and mapping global change. Bethesda,
MD: American Society for Photogrammetry and Remote Sensing
and American Congress on Surveying and Mapping.
Stephenson, N. L., D. J. Parsons, a['.d T. W. Swetnam. 1991. Restormg
fire to the sequoia-mixed conifer forest: Should intense fire playa
role? Tall Tunbers Fire Ecology Conference 17:321-37.
Swanson, F. J., /. A./ones, D. O. Wallin, and J. H. CisseL 1994. Natural
variability-implications for ecosystem management. In Ecosystem
management Principles and applications, edited by M. E. Jensen and
P. S. Bourgeron, 80-94. Vol. 2 of Eastside forest ecosystem health
assessment. General Technical Report PNW-GTR-318. Portland, OR:
u.s. Forest Service, Pacific Northwest Researen Station.
US Department of the Interior and U.s. Department of Agriculture.
1995. Federal wildland fire management Policy and program review.
Washington, DC: U.S. Department of the Interior and U.S.
Department of Agriculture.
US Forest Service. 1995. Draft cnvironmcntal impact statcment:
Managing California spotted uwl habitat in the Sierra Nevllda furests of
California (an ccosystem approach). San Francisco: U.s. Forest Service,
Pacific Southwest Region.
van Wagtendonk, J. W. 19'16_ Use of a deterministic fire growth model
to test fuel treatments. In Sierra NL'Vllda Ecosystem Projcct: Final rq)(lrt
to Consrcss, voL IL chap. 4.3. Davis: University of California, Centers
for Water and Wildland Resources.
Walters, C. )., and C. S. Holling. 1990. Large-scale management
experiments and learning by doing. Ecology 71:2060-68.
Weatherspoon, C. P. 1996. Fire.silviculture relationships in Sierra
forests. In Sierra Nevada Ecosystem Project Final report to Congress,
vol. n, enap. 44. Davis: University of California, Centers for Water
and Wildland Resources.
Weatherspoon, C. P., S. J. Husari, and J. W. van Wagtendonk. 1992.
Fire and fuels management In relation to owl habitat in forests of
the Sierra Nevada and Southern California. In The California spotted
owl: A technical assessment of its current status, technical coordination
by J. Verner, K. S. McKelvey. B. R.. Noon, R, J. Gutierrez, G. L Gould
Jr., and T. W. Beck, 247-60. General Technical Report PSW-133.
Albany, CA: U.S. Forest Service, Pacific Southwest Researen
Station.
Weatherspoon. C. P., and C. N. Skinner. 1995, An assessment offactors
associated with damage to tree crowns from the 1987 wildfires in
northern California. Forcst Scicna 41:430-51.
Wilson, C. C. 1977. How to protect western conifer plantations against
fire. Paper presented at the annual meeting of the Western Forest
Fire Committee, Western Forestry and Conservation Association,
Seattle.
C. PHILLIP WEATHERSPOON
Pacific Southwest RQ~e.rch Statton
U.S. Foresl ServIce
Redding. California
44
Fire-Silviculture
Relationships ill
Sierra Forests
ABSTRACT
Many of the tools available for managing forested ecosystems lie
within the disciplines of silviculture and fire management. These two
sets of management practices, in fact, are commonly used in con-
cer1. Understanding the relationships between these two disciplines,
therefore. can contribute to more intelligent ecosystem management.
Silvicultural techniques mimic to varying degrees some of the distur-
bance functions-such as facilitating establishment of regeneration
and influencing forast structure and composition--performed natu-
rally by fire. This chapter provides a brief overview of some of these
relationships for a range of stand structures and fire regimes. Effects
of partial cuttings on tire hazard also are discussed. Research is
needed to clarify basic relationships between fire regimes and the
dynamiCS and structures of stands and landscapes. Adaptive man-
agement experiments also should be undertaken to determine the
practicability and long-term ecological consequences of a range of
silvicultural and fire treatments.
INTRODUCTION
Hefore Euro-America.ll settlement, relatively frequent fires
strongly influenced the composition, structure, and dynam-
ics of most forest ecosystems in the Sierra Nevada, in concert
with other disturbance factors (I;errell 1996; Skinner and
Chang 1996). These fires, mostly low to moderate in severity,
cauSt.d changes by damaging or killing plants and setting the
stage for regeneration (including sprouting of top-killed
plants) and vegetation succession. They maintained surface
fuels at fairly low levels, and in most areas kept forest under-
stories relatively free of trees and other vegetation. In addi-
tion, fires influenced many processes in the soil and forest
floor, including the organisms therein, by consuming orgamc
matter, affecting nutrient cycling, and inducing other ther-
mal and chemical changes (Agee 1993; Chang 1996). These
fire effects in turn resulted in a wide array of effects on other
ecosystem components and processes, including wildlife com-
munities and watershed properties.
Human activities since the mid-1800s have greatly changed
the occurrence, nature, and effects of fire in the Sierra Ne-
vada (Husari and McKelvey 1996; McKelvey and Johnston
1992; Skinner and Chang 1996; Weatherspoon et al. 1992).
Organized fire suppression, which began early in the twenti-
eth century, has been extremely effective in limiting the area
burned by wildfires (Husari and McKelvey 1996; McKelvey
and Busse 1996). The resulting virtual exclusion of low- and
moderate-severity fire has profoundly affected the structure
and composition of most Sierra Nevada vegetation, especially
in low- to middle-elevation forests. Conifer stands have be-
come denser, mainly in small and medium size classes of
shade-tolerant and fire-sensitive tree species. Stands have also
become more complex when viewed vertically, but less com-
plex and more homogeneous in terms of areal arrangement
(Weatherspoon et al. 1992). "Selective" cutting of large over-
story trees (McKelvey and Johnston 1992) and the relatively
warm and moist climate that has characterized most of the
twentieth century (Graumlich 1993) have probably reinforced
these trends. Excessively dense stands have led to drought
stress and bark beetle outbreaks, resulting in widespread
mortality of trees in many areas and the potential for exten-
sive additional mortality (Ferrell 1996). One consequence of
Sierra '\Jevada Ecosystem Project: Fmal report 10 ConK"re~. vol. II, As.'ies~ments and scientific basi's for mana~ementoption.c:;. DavLe;; University of Califomia, Centers for
Water and Wildland Resources, 1996.
1167
~-~r---""'-^--'-'-'~-'. ^~- ^ -...
1168
VOLUME II, CHAPTER 44
these changes has been a large increase in th!' amount and
continuity of both live dnd dead forest fuels, resu Iting in d
substantial increase in the probability of large, severe wild-
ti res (Weatherspoon and Skinner 1996). In manv ilfeas, eco-
system diversity and sustainability appear leopnrdized by
these manges, even without the threat l)f seven' fi res.
The necessity oi restoring and sustaining these at-risk eco-
systems is emerging as a major challenge confronting those
responsible for managing Sierra Nevada forests. The means
to accomplish these goals is the subject of some controversy.
Some would advocate a hands-off philosophy of forest man-
agement, one of "letting nature take its course." Such a phi-
losophy may well be appropriate for some upper montane
and subalpine forests that have been affected relatively little
by past management activities and in whIch wilderness val-
ues and / or restoration of natural processes are primary man-
agement emphases, especially if lightning fires are permitted
to resume their natural role. This approach, however, is very
unlike]y to be successful in most lower- and middle-elevation
Sierran forests, whose presettlement disturbance regimes were
dominated by frequent low- to moderate-severity fires (Skin-
ner and Chang 1996). Given the exce~sive quantities of fuels
present in most of these forests, continued fire suppression
(which certainly is not a hands-off approach) at it minimum
will be required to avoid wildfire losses that are completely
unacceptable ecologically and socially Suppression alone,
however, will only exacerbate the growing problems of
overly-dense stands and excessive fuels. [n addition to fire
suppression, therefore, some form of active management,
designed to replace critical missi!1g elements of the largely
defunct historic disturbance regimes, is probably essential to
begin to reverse these problems and to ensure the diversity
and sustainable productivity of these forests into the future.
Many of the tools available for managing forested ecosys-
tems, and thereby mimicking to various degrees the functions
of historic fire regimes (Skinner and Chang 1996) or other dis-
turbance processes, lie within the disciplines of silviculture
and fire management. These [-.vo sets of management prac-
tices are commonly used in concert, and in fact the line be-
tween silviculture and fire management can be quite blu rry.
For example, cuttings can be effective in breaking up the hori-
zontal and vertical continuity of live fuels in lower canopy
layers or in pretreating a stand to facilitate the introduction
of prescribed fire. Alternatively, cuttings can add fuels and
otherwise increase wildfire hazard. Prescribed fire and other
techniques are often used for the dual purpose of reducing
hazardous fuels and preparing a site for successful establish-
ment of tree regeneration. Silvicultural techniques are used
to emulate some of the historic effects of fire on forest stmc.
ture. In fact, prescribed fire itself is considered by some to be
a silvicultural technique.
The many ecosystem functions of frequent low- to
moderate-severity fire can be restored fully only through the
use of fire. Silvicultural cuttings and other fire "surrogates"
can substitute only partially tor fire. As is described in the
following section, silvicultural techniques can mimic to vary-
ing degrees some of the functions performed naturally by fire.
including facilitating the establishment of regeneration and
influencing forest structure and composition. A wide <lrray
of thermal and chemical effects of fire (Agee 1993; Kilgore
1973; Chang ]996), however, are not mimicked bv other nleth-
ods. Fire and fire "surrogates" also differ markedly in terms
of other factors, including potential for soil compaction and
components of biomass removed from a site (i.e., fire tends to
consume greater proportions of smaller size classes of bio-
mass, whereas larger size classes typically are removed by
cuttings). Accordingly, it seems desirable for low- to
moderate-severity fire--both prescribed. fire and "managed
wildfire" (Husari and McKelvey 1996; Weatherspoon and
Skinner 1996)-to assume a considerably expanded role in
the management of Sierra Nevada forests. In those areas from
which such fire continues to be excluded, for whatever rea-
sons, managers should recognize that some ecosystem com-
ponents and processes will depart significantly from their
historical ranges of variability (Manley et al. 1995), with
mostly unknown consequences for long-term ecosystem vi-
ability.
Nevertheless, it is important to understand that the rein-
troduction of fire alone cannot restore millions of acres of
degraded Sierra Nevada forests, Silvicultural techniques are
needed in addition to or in lieu of fire in many areas to move
conditions away from dense forests dominated by small trees
and containing excessive fuels toward more open forests
dominated by large trees. Given the realities of modern civi-
lization, it is inconceivable that fire in its presettlement ex-
tent, frequencies, and severities could be restored full y to the
Sierra Nevada, Even at a reduced scale, a number of factors
constrain the use of both management-ignited prescribed fires
and prescribed natural (lightning) fires (Husari and McKelvey
1996; Weatherspoon and Skinner 1996). Furthermore, like
nonfire methods, prescribed fire cannot fully mimic the eco-
system functions of presettlement fire, at least in the short
term, The effects of newly reintroduced fire are likely to be
quite different from those of presettlement fires because the
forests (including fuels) have changed so greatly. If fire alone
were used, several sequential entries with prescribed fire
would probably be necessary, especially in densely stocked
stands with heavy fuel concentrations, before the desired for-
est conditions would be approached. Early prescribed burns
in such stands would tend to be expensive and have a rela-
tively high risk both of escapes and of undesirable fire ef-
fects, In contrast, where feasible and compatible with
management objectives, appropriate silvicultural cuttings
preceding prescribed bums may significantly speed the move-
ment toward desired forest structure and composition and in
turn could hasten the useof prescribed fire in a way that more
nearly mimics the natural ecosystem functions of frecluent
low- to moderate-severity fire. Of course, cuttings also pro-
vide opportunities to meet human needs for jobs and utiliza-
tion of wood fiber.
..".'"'''.....,.......-...--.' ....' --,""~"-"-'..",,~,~... ...--....-........-.
1169
Fire-Silviculture Relationships in Sierra Forests
Understanding the relationships between silviculture and
fire can contribute to more intelligent ecosystem management.
This chapter provides a brief overview of same of these rela-
tionships, discussed in two general categories: (1) silvicultural
cutling methods as approximdtions of sland and landscape
structural effects of fire in different fire regimes, and the com-
patibility of fuel.managementlechni'lues with these cUlling
methods; and (2) effects of partial cuttings on wildfire haz-
ard. Although even-aged cutting methods are discussed
briefly, this chapter emphasizes methods other than even-aged
ones because (1) they more closely mimic the natural distur-
bance regimes prevailinr; in most Sierra Nevada forests, and
(2) any landscape-level needs for large, even-aged stands are
likely to be met by severe wildfires and subsequent planta-
tion establishment for the foreseeable future.
SILVICULTURE, FIRE REGIMES,
AND FUEL-MANAGEMENT
TECHNIQUES
Silviculture was originally developed to produce timber effi-
ciently and sustainably (Smith 1962), and in fact timber pro-
duction has been the principal focus of the discipline during
most of its existence. In the minds of mdny, silviculture is still
the handmaiden of timber management. Over the years, how-
ever, silviculturists and others have come to recognize that
silviculture employs a powerful and flexible set of techniques
for meeting a wide array of resource management objectives
and desired values (Daniel et a1. 1979; Helms and Tappeiner
1996). These techniques need to be used with intelligence and
discnmination. In prescribing and implementing management
treatments, the silviculturist should consider site capabilities,
species requirements, and key ecological processes, indud-
mg the natural disturbance (mainly fire) regimes that pre-
vailed in the area. The extent to which these factors, especially
fire regimes, have been considered in the past has varied con-
siderably.
One key function that both silviculture and natural distur-
bance have in common is facilitating the establishment of re-
generation. The long-term sustainability of any desired forest
condition in the Sierra Nevada depends in part on adequate
establishment of regeneration at suitable intervals, Silvicul-
tural systems are designed to promote the establishment of
regeneration and in fact are classified by the methods they
use to dchieve this goal a.nd the types of structures they cre-
ate (Ford..Robertson ] 971). In most Sierran forest types, fire
historically was. the primary agent that set the stage for re-
generation of conifers and many other plants. Fires typically
produced at least two conditions that promoted conifer re-
generation: they providerl the mineral soil seedbed favored
by many species for seed germination and seedling survival,
and they created openings ranging from a fraction of an acre
to perhaps hundreds of acres--needed for survival and sub-
sequent growth of shade-intolerant species. Other effects of
fire that often influenced regeneration establishment included
increased nutrient availability, reduced density of potentially
competing vegetation, and reduced populations of soil mi-
croorganisms pathogenic to tree seedlings.
In many cases, regeneration was not established after a fire
of low to moderate intensity burned through the understory.
Such fires, however, influenced stand structure and species
composition in other ways. A disproportionate percentage of
smaller trees were kiltI'd by fire, thereby tending to keep the
understory relatively open. In addition, fire discriminated
against thin-barked or otherwise fire-sensitive species. Silvi-
cultural counterparts exist for these nonregeneration functions
of fire: thinning from below (removing smaller trees and leav-
ing larger trees) and thinning to modify species composition.
In fact, the short- to medium-term need most apparent in
many Sierran forests is not the establishment of new regen-
eration but rather the removal, or thinning, of excessive num-
bers of small understory trees. This is a high priority, both to
reduce the haZard of severe wildfire and to begin to restore
forests to a healthier, more sustainable condition
(Weatherspoon and Skinner 1996).
A Range of Fire Regimes and Their
Associated Stand Structures
Fire as a disturbance event, and the variability in the way fire
functions as reflected in various fire regimes, is largely re-
sponsible for the range of natural stand structures found in
forests of the western United States. Stephenson and col-
leagues (1991, 322-23) defined five "fire types" representing
points along a continuum of increasing dominance by intense
fire (and decreasing survival by main canopy trees), in order
to account for the patchy nature of fires:
(1) uniform low intensity, in which all or most canopy
trees survive; (2) low intensity with patchy high inten-
sity. . . in which groups of canopy trees are killed locally
within a matrix of surviving trees; (3) mixed intensity, in
which roughly equal areas of canopy trees are killed and
survive. with neither obviously predominating; (4) high
intensity with patchy Jow intensity, in which groups of
canopy trees survive within a matrix of killed trees; and
(5) uniform high intensity, in which all or mo!>t canopy
trees are killed.
These fire types provide useful reference points in the sec-
tions that follow.
The natural fjre regime of most Sierra Nevada forests is
generally characterized as one of comparatively frequent fires
of low to moderate severity, with small patches of high sever-
ity (Skinner and Chang 1996). This fire regime, which corre-isponds to fire type 2 (Stephenson et a!. 1991), prevailed
cc_. . ,.
1 ,-,"--",,,,~,".,-,,,-,-,,,--l--" ._-~,.-".~"
1170
VOLUME II, CHAPTER 44
historically in most ponderosa pine and mixed conifer ior-
ests both west and east of the Sierran crest and in portions of
the upper montane forests as well. Greater variability in fiie
regimes occurred in more mesic sitt's within the mixed coni-
fer forest type, especially those dominated by white fir, and
in significant portions of the red fir and other upper montane
types (Skinner and Chang 1996). This greater variability in
fire regimes probably translated to greater variability in fire
types as well, so that significant, albeit probably small, pro-
portions of these cooler dnd / or more mesic types may have
been characterized by fire types 3, 4, or 5 (Stephenson et al.
19'!l).
It is noteworthy that the extensive changes in Sierran for-
ests brought about largely by fire suppression and other hu-
man activities over the past 150 years have included a virtual
reversal of fire types (Stephenson et al. 1991). Fire type 2, his-
torically the dominant fire type in Sierra Nevada forests, has
now been virtually eliminated. Conversely, fire types 4 and 5,
relatively rare historically, now account for a large propor-
tion of wildfire acreage in thl' Sierra Nevada.
As was noted earlier, fire type 2 (Stephenson et al. 1991)
corresponds to the presettlement fire regime that evidently
dominated most Sierra Nevada forests, especially those low-
to middle-elevation forests now in greatest need of restorative
management. The corresponding stand structure type (a mo-
saic of small, even-sized groups) and its silvicultural coun-
terpart (the group selection cutting method) are therefore of
special interest in the discussion that follows. Three additional
basic stand structures are discussed, however, in the interest
oi providing information on a more complete range of silvi-
cultural and fire techniques to accommodate varied current
stand conditions and to help meet management objectives for
achieving structural diversity across the landscape. The ex-
tent to which it is desirable to mimic with management the
kinds of stand and landscape structures associated with
presettlement fire regimes (as best we can reconstruct those
structures) is a subject of debate. At a minimum, however,
we need to recognize and understand those historic struc-
tures as a frame of reference, so we know what we are de-
parting from and can better assess the significance and
sustainability of such departures.
The sections that follow contrast even-aged stands with
three other basic types of stand structures that may be found
in more of our managed iorests in the future. These are sim-
plified representations of stand strJcture; the real world is
more complex. Nevertheless, they should provide useful ref-
erence points for illustrating silvicultural alternatives. One
could probably approximate any realistic stand structure by
varying the arrangement and stocking of particular canopy
levels, using one of these four structures as a starting point. A
desired stand strudure could also be viewed as a point on
the multidimensional continuum connecting the four basic
types of structures. For example, as the structure created by
the retention shdterwood cutting method becomes clumpier,
it begins to approximate the structure created by the group
selection cutting method; as openings created by group se-
lection cuttings become larger, they grade into small cl.>ar-
cuts; as the openings become smaller, the structure
approximates that created by the individual tree selection
cutting method. Stand components other than live trees-such
as snags, downed logs, and nontree vegetation-are also im-
portant parts of stand structure for many purposes, and within
limits they can be manipulated silviculturally. For simplicity,
however, the live tree component is emphasized here.
The discussion that follows, which is adapted in part from
McKelvey and Weatherspoon 1992, deals with generalized
stand structures and associated management practices pri-
marily at the stand level. Just as numerous stand-level varia-
tions in structure are possible, as was indicated earlier, it is
important to emphasize that great flexibility also exists for
distributing variations and combinations of these structures
dcross the landscape and through time. This provides oppor.
tunities to arrange landscape-level vegetation structures to
meet varying management objectives.
The sections that follow are organized around stand struc-
tures associated with different regeneration cutting methods.
For each of these structures, however, nonregeneration, or
intermediate, cutting methods such as thinnings are integral
components of the overall silvicultural system, and, like re-
generation cutting methods, mimic natural disturbance func-
tions to various degrees.
Standard silvicultural terminology is used (Daniel et al.
1979; Ford-Robertson 1971; Smith 1962). As was indicated
earlier, these silvicultural systems and the associated termi-
nology were developed in the context of timber management.
The terms, however, are descriptive of cuttings that result in
a broad range of stand conditions-dearly of interest to many
resource areas-and are widely used and recognized.
A short consideration of fuel-treatment options relevant to
each of the basic stand structures is included. It is assumed
that, to the extent practicable, fuels are removed from the site
to promote utilization as well as to reduce wildfire hazard, In
the case of partial cuttings (cuttings other than clear-cuts),
this includes the removal of small understory trees that form
hazardous fuel ladders. Historically, effective fuel manage-
ment has not always been a strong emphasis, due largely to
short-term economic considerations. However, it is becom-
ing an increasingly important concern in treatments pre-
scribed today.
With all of the cutting methods, the use of tractors or other
ground-based machines for yarding logs or for piling or oth-
erwise manipulating harvest residues is limited to relativelyomoderate slopes. Treatment options are much more limited
on steep slopes.
Even-Aged Stands
In an even-aged stand, the ages of all of the trees in the stand
are similar. Natural even-aged stands originate mostly from
high-severity fires that kill the great majority of trees in the
-~-'="""---'
1171
fire Silvlcullure Relallonships III Sierra forests
stand (fire type 5) (Stephenson el al. ]9lJ1). With natural fire
regimes, such fires in coniferous forests norm.llly are sep<.l-
rated by fairly long intervals (usually more than 100 years)
and lypically occur in forest types found in moist or cold re-
gilms
Even-aged forest stands in the Sierra Nevada were prob-
ably relatively uncommon in the presettlemenl era. Such
stands may have been represented best in portions of the
upper montane forests-for example, in some red fir areas-
and in widely-scattered stands of knobcone pme (Skinner and
Chang 1996). In contrast, fire type 5 characterizes d large pro.
portion of current wildfire acreage in the Sierra Nevada be-
cause of increaSed fuel quantities and C'ontinuity.
Silvicultural regeneration cutting methods that produce
even-aged stands include clear-cutting, seed-tree, and
shelterwood cutting. rn a complete cycle of practices in the
even-aged silvicultural system, such a regeneration cutting
would normally be followed by establishment of a planta-
tion or natural regeneration, removal of seed trees or
shelterwood trees (retained initially to provide seed and for
protection for regeneration) where present, appropriate tend-
ing or the young stand, a series of intermediate cuttings
(precommercial and commercial thi!1nings and possible "im-
provement" cuttings), and, at rotation age, another regenera-
tion cutting to begin the cycle again. Either broadcast burning
or machine piling and burning is commonly used to prepare
the site for regeneration (including reducing competing veg-
etation and physical obstacles to planting) following the re-
generation cutting. Underbuming or other fuel treatments
may l2ke place at subsequent times during the life of the stand,
especially after any intermediate cuttings. Prescribed burn
ing is relatively straightforward in even-aged stands except
when the trees are very young.
Even-aged stands resulting from even-aged silvicuItural
systems and from infrequent severe fires may be similar in
terms of the general structure and arrangement of live trees.
Other stand components, however, including large woody
material such as snags and downed logs, and L~eir ecological
functions in the new stand, can be quite d itrerent in the two
kinds of stands.
Two-Storied Stands
As the name suggests, two-storied stands consist of trees of
two quite different ages and sizes. These stands are, in a sense,
intermediate in structure between pven-aged and
uneven-aged stands. Natural two-storied stands tend to be
associated with a moderate- to high-severity fire regime, in
which only scattered live trees or clumps of trees (generally
the larger trees and those of firp-resistant species) survive a
fire within a matrix of killed tn~I.'S (fire type 4) (Slephenson el
al. 1991). The fire also promotes the establishment of a new
age class of trees in the understory Clim~tes tend to be fairly
'~l'\r~t t'll! '~I"l't'\\ h..\l drit"r lh,ll~ I Ill"' Id' iI'!, l,j..l1_,-,,, \'11:1
H.HLb-I;'~'I._J '. <-.j ltlt: Il:bll1ll~-".
The presettlement occurrence of fire type 4 and two-storied
stands in the Sierra Nevada was probably somewhat more
frequent than fire type 5 and even-aged stands, although di-
rect evidence of this is very limited_ Some upper montane
forests, along with the more mesic mixed conifer sites, such
as those dominated by Douglas fir or while fir, may have ac-
counted for much of this stand structure type (Skinner and
Ch,mg 1996).
The silvicultural technique associated with this kind of
stand structure is retention shelterwood (also sometimes
called irregular shellerwood or shelterwood without reo
moval). rypically beginning with a shelterwood seed cutting,
shelterwood trees (and trees reserved for other reasons) are
left in place after regeneration has become established, in-
stead of being removed. These trees may remain in the stand
through much or all of the following rotation. Some will be-
come snags, and some may be removed at the end of the next
rotation (at which time a new set of overstory shelterwood
trees will be selected for retention).
Other conditions could be used as starting points for creat-
ing a two-storied stand structure, Understocked stands, tra-
ditionally a high priority for clear-cutting, could instead b..
underplanted, leaving most of the overstory in place. This
kind of structure could also be initiated in an older planta-
tion by having a heavy commercial thinning double as a
shelten'iood-type regeneration cutting. The cut could be fol-
lowed by site preparation/ fuel treatment and underplanting
with the desired mix of species, Throughout the "rotation" of
such a stand, thinnings could be conducted as needed to
maintain desired size classes and species. These should be
followed by prescribed burning or other fuel treatments such
as mastication or chipping, Snags could be created as needed.
Once created, the stand would never be devoid of large trees:
each regeneration cutting would be accompanied by the re-
tention of some overstory trees.
Fuel treatments, including prescribed burning, should nol
be particularly difficult for a two-storied stand. Initial site
preparation I fuel treatment before establishment of the un-
derstory would be the same as for a shelterwood cut. Subse-
quent treatments would be comparable to those for an
even-aged plantation. Separation of canopy layers would
normally be sufficient to keep wildfires from torching into
overstory crowns.
Uneven-Aged Stands Consisting of a Mosaic
of Small, Even-Aged or Even-Sized Groups
[n an uneven-aged stand of small, even-aged or even-sized
groups, each of several age or size classes occurs in a number
01 small (mostly from 1/4 acre to about 2 acres in size) groups
or aggregations distributed throughout the stand. For the most
part, age or size classes are separated horizontally rather than
vertically. N.1tural stand strllctures of this tyre orir;inatf' l'ri-
r! 1, I r i! ',' t! I I;' t rl '~;; r lIt ':--: I! 1 " I, i. 1, , It....;, 11\ I rLl fl' I ,I i' ,\.1 "" (r.' !1" ! 1 II
Lul bl~IIt..:[....lII) dl Iv....\, Lv IIH)Jt.:ILlll: St:VC(Jly. i\lv.')l ufl.:d.S dll~
._-,- ,.
....."""'t... '___~O'_< _m_'....._"_",.,__,."..,.,~.____.,,
1172
VOLUME II, CHAPTER 44
underburned, with many small trees being killed but nhlst
large trees surviving. Scattered individudls and groups 01
main canapy trees, however, are killed wheel' the fire 10c.)lly
flares up or bums more severely (or groups of frees previ-
ously killed by other agenLs such as bark beetles are consumed
to varying degrees by the fire), leaving s('attered small open-
ings within a matrix of surviving trees (fire type 2.)
(Stephenson et aI. 19<Jl). The locally intense fire exposes min-
eraI soil (a favorable substrate for seedling establishment) dnd
temporarily reduces competing vegetation (including reserves
of dormant seeds stored in duff and soil). Given good cone
crops and favorable soil moisture and other conditions, tree
seedlings become established. Seedlings in an opening may
be even-aged-originating irom a single cone crop-or they
may become established over a number of years. This fire re-
gime and this stand structure were common ouring the
prf'settIement era in the Sierra Nevada, especially in the pon-
derosa pine and mixed conifer forest type~ (Skinner and
Chang 1996).
Silviculturally this kind of stand structure is approximated
with the group selection cutting method. Group sizes should
be large enough to permit successful regeneration of
shade-intolerant tree species. In a sense, each group can bt'
regarded as a small even-aged stand, which can be carried
through the full cycle of regeneration cutting, r"generation
establishment and tending, intermediate cuttings, and regen-
eration cutting once again. So within a stand that contains
many of these small, even.aged groups, the group (regenera-
tion) cuttings can be accompanied by concurrent intermedi.
ate cuttings in the other groups within the stand (mimicking
small. high-severity burn areas within a matrix of low- to
moderate-severity fire). Keeping track of numerous small
openings and groups for management purposes, long con-
sidered a major obstacle to the use cf group selection, should
be significantly easier with the advent of geographic infor-
mation systems and satellite-based global positioning sys-
tems.
[n groups to be regenerated, all trees could be removed, or,
especially in larger groups, scattered live trees and / or snags
could be retained. To facilitdte fuel treatment and reduce dam
age to the surrounding stand, cut trees should be felled a:i
much as possible into the newly created opening.
Openings could be regener<lted, either naturaily or artifi-
cially and with or without vegetation management (reduc-
tion of competing vegetation). Even with planting and
vegetation management, growth of tree seedlings would be
less in an opening typical of group selection than in a large
opening because of competition for site resou rces from large
trees surrounding the opening (Tht' degree of competition
will depend on the dt'nsity or stocking level of the su rround-
ing stand as well as the distance from the edge of the open-
ing.) Without planting and some control of nonconifer
vegetation, however, the development of conifers could be
delayed for several decades. Under such cond itions, fuel treat -
ment would be complicated as well.
The development of a mosaic of small groups could be ini-
tiated in a wide range of stand conditions-for example, in
an older plantation, an uneven-aged young-mature stand, or
an old stand with patchy, uneven distributions of size classes
or species.
Harvesting and other treatments are more difficult and
expensive in an uneven-aged stand with a mosaic of even-
aged or even-sized groups than in an even-aged stand. Imple-
menting group selection cuttings on steep slopes, however, is
especially problematical. Helicoptersciin be used but are very
expensive. This area is ripe for some good logging engineer-
ing research and development Hopefully, practical and eco-
nomically viable methods will be developed for using skyline
systems to yard group selection cuttings while keeping dam-
age to the residual stand within acceptable limits. This could
also provide opportunities for cable yarding of residues or
for the use ot other means of reducing fuel loads, such as re-
moving tree tops (which contain considerable potential fuel)
together with adjacent merchantable logs.
Fuels should be treated not only in the regeneration open-
ings but also in the rest of the stand. On machine-operable
slopes, the whole range of mechanical fuel-management tech-
niques would be available. These could include tractor piling
and burning of slash in regeneration openings, mastication,
and removal (with or without utilization). Residual stand
damage and soil impacts, however, must be kept within ac-
ceptable levels. Machine size and capabilities and operator
skill are all critical factors.
Prescribed understory burning is an option on steep as well
as moderate slopes. Prescribed burning would be more diffi-
cult than in even-aged or two-storied stands, simply because
a variety of conditions and tree sizes occur within the stand.
However, the tact that these size or age classes are separated
horizontally rather than vertically, if combined with proper
temporal spacing of treatments (McKelvey and Weatherspoon
1992), should alleviate many of the potential problems.
l\vo-stage burning (sequential bums under different condi-
tions) or jackpot burning (burning of residue concentrations
under conditions that impede fire spread into adjacent areas)
may be applicable in some situations. One could broadcast-
burn regeneration cut areas after harvest, and then lInderbu rn
the rest of the stand at the same time or perhaps at a laler
stage, when understory fuels have dried a little more. De-
pending on stand conditions, some preburn treatment may
be necessary prior to the first fire entry to reduce fuel Jadders
and overall flammability to acceptable levels. This could be
expensive and might include biomass harvest. culting and
hand piling, or other methods. If litter from ponderosil pine
is available, prescribed bums can be conducted under moister
conditions and therefore in more difficult situations. Again
depending on stand conditions, a first bum might create sub-
stantial additional fuel by scorching or killing (mostly small)
trees, necessitating a second and possibly a third burn to get
the fire hazard down to an acceptable level.
1173
Fire-Silviculture Relationships in Sierra Forests
Uneven-Aged Stands Consisting of a Fine
Mosaic of Individual Trees
In an uneven-aged stand conlalnllllj d fine mosaic of indi-
vidual trees, thrl'e or more sizes and ages of all tree species
preselLt are distributed more or less uniformly throughout the
stand. Openings are very small, the size of individual large
trees. This occurs in nature (at least in a sustainable mode)
only in forest types composed entirely of shade-tolerant spe-
cies and in fire regimes having very long fire-return inter-
vals. [t develops long after a stand-replacement fire, as the
overstory begins to break up and a full range of understory
canopy layers has had a chance to develop. This stand type is
incompatible with frequent periodIc fires. (Some observers
have considered certain open-growing ponderosa pine stands
with short fire-return intervals to have this kind of stand struc-
ture. In such cases, the distinction between stands of uneven-
aged individual trees and stands of uneven-aged groups of
trees becomes largely one at semantics.)
This stand condition is produced and maintained silvicul-
turally using the individual-tree selection cutting method.
Unless a definition of individual-tree selection is used that
includes openings up to 1/4 acre or so (or involves very open
stands), this method will not allow for adequate regenera-
tion and development of shade-intolerant species on most
sites. If the stand does not already consist of shade-tolerant
conifers, it will move in that direction under this cutting
method as long as such specif>s are present in the area. Reten-
tion of the smallest size classes of trees well distributed
through the stand-.a necessity for sustaining this stand struc-
ture through time--creates dangerous fuel ladders and makes
prescribed understory burning essentially impracticable.
On gentle terrain, various machine treatment methods are
available, at least theoretically, tor accomplishing individual-
tree selection cuttings. Residues remaining after harvesting
could be machine piled, chipped, or masticated. But skillful
operators and tight controls over fuel-treatment activities
would be necessary to avoid unacceptable damage to the re-
sidual stand.
Other alternatives include jackpot burning of slash concen-
trations and the much more costly option of hand piling and
burning-either applied preferably al a lime when surround-
ing fuels are too moist to carry fire. Both of these methods
would also be available on steep slopes. [mplementation of
individual-tr..c selection on steep slopes may be feasible only
with expensive helicopter logging systems,
At higher elevations or other mesic sItes where the prob-
ability of sp.vere wildfire is not great, some combination of
lopping, bucking, and scattering of slash, or no fuel trf>atment
at all, may be acceptable. [f individual-tree selection is to be
used at all, it will be on such mesic sites that it probably makes
the most SP.nse anyway because it is more nearly compatible
with presettlement fire regimes and stand and landscape
structures.
EFFECTS OF PARTIAL CUTTINGS
ON WILDFIRE HAZARD
The effects of partial cuttings on wildfire hazard in the re-
sidual stand result from combinations and interactions of two
general factors: effects on fuels, and effects on microclimate.
Effects of Partial Cuttings on Fuels
Thinnings, insect sanitation and salvage cuts, and other par-
tial cuttings add slash, or acti vity-generated fuels, to the stand
unless all parts of the tree above the stump are removed from
the forest. Small trees damaged by harvest activities but not
removed from the forest often add to the fuel load. To the
extent that it is not treated adequately, this component of the
total fuel complex tends to increase the probability of a more
intense, more damaging, and perhaps more extensive wild-
fire. .
Foliage and small branches of live forest vegetation also
contribute to the total amount of available fuel. The position
and continuity of these fuels are important. Dense understory
trees, for example, can provide both the horizontal and the
vertical continuity of live fuels needed to move a fire from
the surface into the main forest canopy and sustain it as a
crown fire. This kind of stand condition is currently wide-
spread in Sierra Nevada forests. Cutting and removing a large
proportion of such a dense understory, thus interrupting much
of the live fuel continuity, can substantially reduce the prob-
ability of a crown fire.
Partial cuttings also have longer-term, more indirect effects
on fuels. Thinning or not thinrung overly dense stands, for
example, influences overall levels of competition for limiting
resources (water, nutrients, and sunlight) in the stand and
consequent levels of stress-induced mortality (induding but
not limited to that caused by insects). Dead trees obviously
add to the total dead fuel load and may increase both the se-
verity of a future wildfire and its spread rate via spotting.
Thinning also influences the subsequent regeneration and
development of understory vegetation-trees, shrubs, and
herbs-which becomes part of the live fuel component.
Effects of Partial Cuttings on Microclimate
A related but separate kind of COncern has to do with changes
in microclimate brought about by stand opening. Thinning
or otherwise opening a stand allows more solar radiation and
wind to reach the forest floor. The net effect, at least during
periods of significant fire danger, is usually reduced fuel
moisture and increased flammability (Countryman 1955). Thp
greater the stand opening, the more pronounced the change
in microclimate is likely to be.
1174
VOLUME II, CHAPTER 44
Interactions of Changed Fuels and
Microclimate
The ways in which changes in these two sets of factors-fuels
and microclimale--as a result of a managemenl activity in.
leract to affect wildfire hazard can be quite complex. The net
effect, in terms of the direction of change in hazard, may be
obvious in many cases, however. For example, removing most
of the large trees from a sland, leaving most of the under-
Slory in place, and doing little or no slash treatment--a situa-
tion all too familiar in the past-will certainly increase the
overall hazard and expected damage to the stand in the event
of a wildfire. Everything points in the same direction: remov-
ing most of the fire-tolerant large trees; retaining most of the
easily damaged small trees; increasing the loading (quantity)
and depth of the surface fuel bed; ilnd creating a warmer, drier,
windier environment near the forest floor du ring times of sig-
nificant fire danger. In contrast, heavily thinning an over-
stocked stand from below and using whole-tree removal (or
chipping and spreading the limbs and tops), followed by a
prescribed understory burn to reduce natural fuels, will al-
most certainly reduce the wildfire hazard of the stand. Com-
puter simulations of the effects of such treatments on fire
behavior (van Wagtendonk 1996), along with anecdotal re-
ports of how such stands have fared during a wildfire in com-
parison with surrounding untreated stands, provide strong
support for this conclusion. [n this case, the "negative" ef-
fects on microclimate of opening the stap.d are outweighed
by the reduction in live and dead fuel loading and continuity.
Past cuttings in the Sierra Nevada (Helms and Tappeiner 1996)
have spanned the range represented by these two contrast-
ing situations but have tended generally, like the first situa-
tion, to create a net increase in fire hazard.
An example of a more complex relationship was reported
by Weatherspoon and Skinner (1995) as part of a large retro-
spective study of factors-including prior management ac-
tivities-that affected the degree of tree damage resulting from
the extensive 1987 wildfires in northern California. Among
three categories 01 uncut or partial-cut stands, they found that
uncut stands (with no treatment at natural tuels) suffered the
least fire damage, followed by partial-cut stands with some
fuel treatment; p<lftial-cut stands with no treatment had the
most damage. The fact that partial-cut stands with no fuel
treatment experienced more damage than partial-cut stands
with some fuel treatment is no surprise. One might wonder,
however, why the uncut stanos experienced less damage than
the partial-cut and treated stands. The explanation probably
lies in a combination of the following factors:
. The partial cuttings created a warmer, drier microclimate
compared with that of the uncut stands-an inevitable ef-
fect of cuttings, as was explained earlier.
. The partial cuttings were typical of many past cuttings that
removed big trees and left small ones. The more readily
scorched small trees thus constituted it higher percentage
of the residual stand. I;urthermore, the live fuel ladder com-
ponent of fire hazard in the uncut stand was not reduced
in the partial-cut stand.
. ruel treatments may have been only partially effective. Two
types of fuel treatments-lop and scatter ,lfid
underbuming-were combined in the analysis (their sepa-
rate effects on fire damage were indistinguishable). Lop-
and-scatter treatments reduced slash depth (and so
presumably reduced flammability compared with no treat-
ment) but did not change the fact that total downed dead
fuel loading in those partial-cut stands (consisting of natu~
ral plus activity-generated fuels) was greater than downed
dead fuel loading in uncut stands (consisting of natural
fuels only). The underburns were not planned treatments
but rather were bums that were allowed to creep around
between clear-cut units that had been broadcast-burned or
to move away from burned roadside piles. Thus, fuel con-
sumption may have been spotty in these areas. More in-
tensive treatment of surface fuels might well have reduced
fire damage further.
. When only the management compartments containing
fuel-treated stands (a small subset of the total number of
compartments in the study) were analyzed separately, dif-
ferences in fire damage between uncut and partial-cut and
treated stands virtually disappeared. Evidently, lower av-
erage levels of damage in uncut stands in the remaining
compartments changed the relationship in the overall
analysis.
CONCLUSIONS AND RESEARCH
NEEDS
Restoration and maintenance of Sierra Nevada forests in pro-
ductive, sustainable conditions will almost certainly require
i:ombinations of silvicultural and fire-management tech-
niques. Understanding the ecological and operational link-
ages between these two disciplines will facilitate this task.
It is generally recognized that recurring fires historically
pIa yed a key role in influencing the species composition, stand
structure, and landscape mosaic of most forest types in the
Sierra Nevada as well as elsewhere in western North America.
But the basic relationships between fire regimes and stand
and landscape dynamics are poorly understood for many for-
est types, including those in the Sierra Nevada, Clarifying
these relationships through research should help managers
as they seek to define desired forest conditions and processes.
We also have little information about the long-term conse-
quences of various forest c(lnditions on a range of ecosystem
components. The long-term nature of these questions and the
need to find answers on a landscape scale means that the nee-
1175
Fire-Silviculture Relationships in Sierra Forests
essary studies will need to be done in the context of adaptive
man,lgement an organized process of learning by doing
(Everett el al. 1'194; Walters and Holling 1990). Managers and
scientists should cooperate in long-term adaptive manage-
ment experiments to (1) devise silvicultural and fire treatnwnls
that mimic historical or other desired conditions in certain
key respects; (2) define treatments representing reasonable
management alternatives that "bracket" those conditions; and
(3) incorporate these treatments into long-term, interdiscipli-
nary studies of the consequences of alternative management
strategies in terms of eco:;ystem productivity, diversity, and
sustainability. Because of the key role of fire historically and
the broad range of fire effects on forest ecosystems, it is im-
portant that the suite of treatments include comparable stand
structures produced and maintained by prescribed fire alone
(requiring multiple bums), through silvicullural cuttings and
mechanical fuel treatments alone (i.e., without fire), and
through combinations of cuttings, mechanical fuel treatments,
and prescribed fire. Only in this way will it be possible to
determine which ecosystem functions of fire can be emulated
satisfactorily by other means, which may be irreplaceable, and
the implications of these findings for management.
Although the basic theory of silvicultural systems has been
well established, actual application of systems other than
even-aged ones in California is quite limited. Practical meth-
ods for implementing such treatments, especially on steep
ground and in conjunction with a variety of fuel-treatment
methods, will require considerable applit'd research as part
of the adaptive management efforts discussed previously.
At least in the short to medium term, much of the needed
silviculture in Sierran forests will involve thinning of small
trees To make such operations economically sustainable, co-
operative research and development efforts are needed to
develop more efficient technology for harvesting and process-
ing of small material and new markets for utilizing it (Lam-
bert 1994)
While we have mucn to learn, it is important to note that
we do not have to have all the answers before beginning
needed restoration work. We know enough at this point to
recognize that cur[(~nt conditions in most low- to
middle-elevation forests of the Sierra Nevada are unaccept-
able in terms of wildfire hazard, diversity, and sustainability.
Regardless of the extent to which presettlement conditions
are used as a guide to de"ired conditions, most informed
people would agree that these forests generally should be less
dense, have less fuels, and have more large trees. Even if we
have not precisely identified target conditions, we certainly
know the direction in which we should begin movin~ That
beginning alone will require a large measure of commitmt>nt
and hard work. We can adjust along the way as we learn more
and become better able 10 define desired conditions for Sier-
ran forests.
ACKNOWLEDGMENTS
I would like to Ihank the following individuals for valuable
comments on earlier versions of the manuscript: L. Blum,
J. Buckley, J. Fites, D. Fullmer, M. Landram, D. Leisz,
K. McKelvey, D. Parsons, C. Skinner, J. Tappeiner, J. Woods,
and two anonymous reviewers.
REFERENCES
Agee, J. K. 1993. Fi re ecology of Pacific Northwest forests. Washington,
DC: I~land Press.
Chang, C. 1996. Ecosystem responses to fire and variations in fire
regimes. In Sierra Nevada Ecosystem Project Final report toCongn'SS,
vol. II, chap. 39. Davi~: University of Califomia, Centers for Waler
and Wildland Resources.
Countryman, C. C. 1955. Old-growth conversion also converts
fireclimate. In Proceed ings of Society of American Foresters Annua I
Meeting, 158-60. Portland, OR: Socicty of American Foresters.
Daniel. T. W., J. A. Helms, and F. S. Baker. 1979. Principles of silviculture.
2nd cd. New York: McGraw-HilI.
Everett, R., C. Oliver, J. Saveland, J. R. Boeder, and J. E. Means. 1994.
Adaptive ecosystem management. In Ecosystem management:
Principles and applications, edited by M. E. Jensen and P. S.
Bourgeron, 340-54. Vol. 2 of Eastside forest ecosystem health
assessment. General Technical Report PNW-GTR-318. Portland, OR:
U.s. Forest Service, Pacific Northwest Research Station.
Ferrell, G. T. 1996. The influence of insect pests and pathogens on
Sierra forests. In Sierra Nevada Ecosystem Project: Final report to
Congress, vol. II, chap. 45. Davis: University of California, Centers
for Water and Wildland Resources.
Ford-Robertson. F. C, ed. 1971. Terminology of forest science, technology;
practice, and products. Washington, DC: Society of American
Foresters.
Graumlich, L. J. 1993 A 1,000-year record of temperature and
precipitation in the Sierra Nevada. Quatemary Research 39:249-.-55.
Helms, J. A., and J. C Tappeiner. 1996. Silvirulture in the Sierra. In
Sierra Nevada Ecosystem Project Final report to Congress, vol.lL chap.
15. Davis: University of California, Centers for Water and Wildland
Resources.
Husari, S. J.. and K. S. McKelvey. 1996. Fire management policies
and programs. In Sierra Nevada Ecosystem Project Final report to
Congress, vol. II, chap. 40. Davis: University of Califomia, Centers
for Water and Wildland Resources.
Ktlgore, B. M. 1973. The ecological role of fire in Sierran conifer fon>sts:
Its application to national park management. Quaternary Research
3:496-513.
Lambert, M. B. 1994. Establish stable stand structures and increase
tree growth: New technologies in silviculture. In Restoration of
stressed sites, and processes, compiled by R. L. Everett, 93-96. Vol. 4
of Eastside fur~st ecosystem health assessment. General Technical
Report PNW-CTR-330. Portland, OR: U.s. Forest Service, Pacific
Northwest Research Station.
Manley, P. N., G. E. Rrogan, C Cook, M. E. Flores, D. G. Fullmer, S.
Husari, T. M. Jimerson, L. M. Lux, M. E. McCain, J. A. R[)~(', C.
Schmitt, J. C. Schuyler, and M. J. Skinner. 1995. Sustaining
ecosystems: A conceptual framework. R5-EM- TP-OOl. San Francisco:
U.S. Forest Service, Pacific Southwest Region.
1176
VOLUME II, CHAPTER 44
McKelvl'Y. K. S., and K. K. Busse. 1996.Twentieth-c,-,ntury fire patterns
on Forest Service lands. [n Sierra Nevada Ecosystem PrD;ect: Final
report to Conl;ress, vol. ll, chap. 41. Davis: University of California,
~:enters for Water and Wildland Resourc'-'s.
\1eKelvey. K. S., and J. D. Johnston. 1992. Ilistoneal perspectives on
forests of the Sierra Nevada and the Transverse Ranges of Southern
Caltfomia: Forest conditions at the lurn of the century. In The
California spotted owl: A technical assessment of its current slatus,
technical coordination by J. Verner, K. S. McKelvey, B. R. Noon,
R. J. GCltierrez, G. I. Could Jr-, and 1'. W. Beck, 225-46. Gen,'ral
T,'chn ical Report PSW-133. Albany, CA: U.5. Forest Service, Pacific
Southwest Resedrch Station.
McKelvey, K. 5., and C. P. Weatherspoon. 1992. Projected trends in
owl habitat. In The California spotted owl: A technical assessment of
Its current status, technical coordination by J. Verner, K. S.
McK.+..ey, B. R. Noon, R. J. Cutierrez, C. I. Could Jr.. and 1'. Beck.
261-73 General Technical Report PSW-133. Albany, CA: US. Forest
Service. Pacific Southwest Research Station.
Skinner. C. :-.I., and C. Chang. 1996. Fire regimes. past and present. In
Sierra Nevada Ecosystem Project Final report to Congress, vol. n, enap.
38. DavIS: University of California. Centers for Water and Wildland
Resources.
Smith. D. M. 1962. The practice of silviculture. 7th cd. New York: John
Wiley.
Stephenson. N. L., D. J. Parsons. and T. W. Swetnam. 1991. Restoring
natural fire to the sequoia-mixed conifer forest: Should intense
fire playa role' Tall TImbers Fire Ecolugy Conference 17:321-37.
US. ForestSecvlce (USFS). 1995. Draft environmental impact statement.
Manag' ng California spotted owl habitat in tilt' Sierra Nevada natioml
for=ts of California (an ecosystem approad,). San FranCISCO: U.s. Fon.>st
Service, PacifIC Southwest Region.
van Waglendonk.)' W. 1996. Use of a deterministic fire growth model
to test fuel tTeatmenls.!nSierra Nevada Ecosystem Project: Final report
to Congress, vol.!1, chap. 43. Davis: University of California, Centers
for Water and Wildland Resources.
Walters, C. J.. and C. S. Holling. 1990. Large-scale management
experiments and learning by doing_ Ecology 7l:2060~8.
Weatherspoon, C. P.. S. J. Husari. and J. W. van Wagtendonk. 1992.
Fire and fuels management in relation to owl habitat in forests of
the Sierra l',;evada and Southern California. In The California spotted
ow I: A technical assessment of ilS current status. tech.'lical roordination
by). Verner, K. S. McKelvey, B. R. Noon, R. J. Gutierrez, G. I. Gould
Jr.. and 1'. W. Beck, 247~0. General Technical Report PSW-133.
Albany, CA: U.s. Forest Service, Pacific Southwest Research
Station.
Weatherspoon, C. P., and C. N. Skinner. 1995. An assessment of factors
associated with damage to tree crowns from the 1987 wildfires in
northern Cal ifomia. Forest Science 41:430-5l.
-. 1996. Landscape-level strategies for forest fuel management.
In Sierra Nevada Ecosystem Project: Final report to Congress, vol. II,
chap. 56. Davis: University of California, Centers for Water and
Wildland Resources.
(
1,-
OREGON DEPARTMENT OF FORESTRY
GUIDELINES FOR RESIDENTIAL ASSESSMENT OF FIRE
HAZARDS
Last updated 4/16/01
Options - Every home site is different. In assessing your hazard situation,
determine what you like about your surroundings. Consider concerns such as
visual barriers, dust barriers, shade, aesthetics, erosion control, etc. As much as
possible, plan your hazard reduction efforts with your personal interests in mind.
When receiving assistance from OOF, or fire department personnel, let them
know what's important to you.
County Codes and Ordinances - Fire Safety Requirements and Guidelines for
Jackson and Josephine Counties are attached. You may want to reference these
to answer questions. Keep in mind that most of the requirements only apply
to new construction.
ADDRESSING
Situation - Many interface residents enjoy living in seclusion. They value the
privacy of living out-of-view of their neighbors. Quite often, an address on a
mailbox is all emergency service providers have to indicate a home's location.
Unless the mailbox is located at the base of a single dwelling driveway,
addressing is inadequate. Too often the numbers are too small, faded, or only
posted on one side of the mailbox. When homes are visible from the street,
addresses posted on the home are sometimes covered by overgrown shrubbery.
They may even end up being the same color as the house, after a fresh paint job.
The single driveway that forks several times can create confusion when the only
address indictors are a row of mailboxes at a driveway's entrance.
Intent - Addresses should be posted in a manner that will enable emergency
services of any kind to find their way into the residence, at all road forks, day or
night, and in bad weather.
Recommendation - Consider 3 to 4 inch numbers on a contrasting background,
reflective if possible, posted at all intersections with arrows if necessary, and
easily visible from the cab of a vehicle day or night, in bad weather or in thick
smoke.
Jackson Co Reference 280 100 - 3. G) Address signs
Josephine Co. Reference 76.040 - Identification numbers
"". yo
r _.._..",.___..,'__,,___...._~w.,.~__.,
ROAD AND DRIVEWA Y ACCESS
Situation - Many driveways in our interface were not designed with fire engine
access, or residential evacuation in mind. Few have all-weather surfacing. Many
cannot accommodate a fire truck's width, height, weight, and length, and are
overgrown with brush. Some residents prefer a heavy growth of vegetatIon as a
dust or sound barrier, or to encourage slower driving. Some driveways are long
single-lane roads with no turnouts. Once at the residence, there may not be an
adequate space for fire equipment to turn around. Random sampling from the
1998 Fire Safety Home Visit program indicated that 56% of 866 residences
visited by the Oregon Department of Forestry in Jackson and Josephine
Counties had no alternate escape routes. Some driveways are simply too steep
for adequate fire engine access.
Intent - Driveways to the home should provide easy emergency vehicle access
and turn-around, and ensure a safe vegetation-free escape route for residents.
This not only includes condition of the access itself, but surrounding vegetation
that can ignite and block the access.
Recommendation(s) - In many cases with existing driveways, there will be little
in the way of modification options. You may consider an all-weather surface,
replacing a weak bridge with a culvert and fill, installation of turnouts where
possible, development of alternate safe, vegetation-free escape routes where
possible, or additional clearance of flammable vegetation.
Jackson Co. Reference 280.100 - 2. C) Emergency Vehicle Access.
Josephine Co. Reference 76.030 - E., F., G. Development Standards
NOTE: In Jackson County, dwellings in existence as of October 28, 1980 are
exempt from compliance with driveway standards of Chapter 280. 100 Fire Safety
Requirements and Guidelines.
RREBREAK/GREENBELT
Situation - Dry grass, under flammable brush, under larger trees, create an
arrangement firefighters call "ladder fuels". This vegetation arrangement gives
fire a ladder to climb As a wildfire climbs this "ladder", it builds in intensity, often
defying fire control efforts. Ladder fuels are prevalent in our area, and from a
vegetation standpoint, pose the greatest threat to structure sUNivability. Both
native and non-native vegetation types can be highly flammable. Many interface
residences don't have the water supplies needed to maintain a greenbelt, by
frequent watering.
2
Intent - Prevent structural ignition from embers, direct flame contact, or radiant
heat, and provide escape routes and safety zones for firefighters and residents.
Removal of other fuel sources close to the home, such as lumber or firewood
may need to be considered in firebreak development.
Recommendations -
. Break the vertical fire ladder by removing flammable, and dead and downed
vegetation, cutting dry grass to a stubble, thinning and spacing brush, and
removing lower tree limbs that may be ignited from vegetation below.
. Fire spreads both vertically, and horizontally, imagine the vegetation ladder
lying down. It's equally important to "break the ladder" to slow a fire's
horizontal spread as well.
. Move other combustibles such as lumber and firewood at least 30 feet away
from the home.
. Limbs should be trimmed back to a minimum distance of 10 feet from
rooflines. Roofs and gutters should be kept free of leaves, needles, and moss
that are easily ignited by falling or wind-blown embers.
. If flammable vegetation is too close to the home, radiant heat can crack or
break windows allowing embers to reach the home's interior. Create a greater
distance between the vegetation and the home, especially if the home has
single-pane windows.
. Consider replacing flammable vegetation with fire resistant plants. Plant lists
are available from ODF, local nurseries, and the Internet,
Jackson Co. Reference 280.100 - 2. A) Fuelbreaks
Josephine Co. Reference Section 76.030 - L. Development Standards
NOTE: In Jackson County, dwellings in existence as of October 28, 1980 are
exempt from fuelbreak standards of Chapter 280. 100 Fire Safety Requirements
and Guidelines.
BUILDING CHARACTERISTICS
Situation - Studies have shown that most structures lost in wildfires were ignited
by radiant heat from nearby flammable vegetation, embers, or "firebrands"
landing on wood shake roofing, embers entering the structure through
unscreened attic or foundation vents, or through broken windows. Embers can
also become lodged under exposed decks where accumulations of leaves or
other combustibles may be present Direct flame contact, and extreme radiant
heat can melt plastic siding and vents, crack and break windows, and ignite
flammable building materials
Intent - Prevent structural ignition from embers, radiant heat, and direct flame
contact.
03
Recommendations -
. Roofing - Install fire-resistant roofing materials when re-roofing the home,
remove overhangmg branches, and keep all roofs and gutters free of
flammable materials such as leaves, needles, and moss.
. Plastic siding and vent frames - Increase the distance between flammable
vegetation and the home (to eliminate direct impingement of flame, and
minimize melting from radiant heat).
. Vents - Screen vents with 1/8 inch non-corrosive metal mesh screening, or
box-in exposed eaves and install screened soffit vents.
. Decks - Enclose or screen areas underneath exposed decks, remove
combustibles from beneath, and keep deck surfaces free of flammable debris.
. Windows - Replace single pane windows with thermopane, or increase the
distance between flammable vegetation and the home (to eliminate direct
impingement of flame, and minimize exposure to radiant heat)
In Jackson Co. Reference 280.100 - 2. E), 3. A), 5.
In Josephine Co. Reference 76.030 - 0.
UTILITIES
Situation - Main distribution lines, and feeder lines to the home and outbuildings
can become overgrown or intertwined with tree limbs and other vegetation. Many
wiring installations to pump houses and outdoor light fixtures are "home project"
jobs completed by non-professionals, and can create a risk to surrounding
flammable vegetation. Snags (dead standing trees) close to powerlines can fall
on, or break lines. LPG (liquid propane gas) tanks are often located too close to
structures, or have inadequate clearance of flammable vegetation,
Intent - Prevent contact between flammable vegetation, powerlines, "home
wiring/l jobs, and LPG tanks.
Recommendations - Report powerline hazards to Pacific Power, Clear
flammable vegetation away from "home wiring job" areas, and pump houses,
Consider professional removal of nearby snags, and clear al/ flammable
vegetation and other combustibles at least 10 feet away from a Liquid Propane
Gas tank.
4
PROTECTlON RESOURCES
Situation - Many interface homes are built outside structural fire district
boundaries and have no structural fire department protection. Most interface
areas have limited water supplies. Electrical service is often interrupted or lost
during wildfires. In some instances, it may be safe for you to remain at your home
to protect it from an approaching fire. In most cases, it isn't. Very few residents
have been trained or are equipped to OSHA standards, to survive in a wildfire
environment, or in the safe and effective use of water and fire tools. Most have
no personal protective eqUipment. Wildfire behavior is always unpredictable. Ufe
safety is our highest priority and should be yours too. In most instances, .
evacuation IS your only safe option.
Recommendations - Consider having additional resources, such as alternate
water sources, gas-powered generators and pumps handy, before a fire
threatens. These may help firefighters protect your home. Ensure fire truck
access to existing water sources such as pools, ponds, storage tanks, etc.
Jackson Co. Reference 280.100 - 3, C), 0).
Josephine Co. Reference 76.030 - K. 1 & 2.
OTHER CONSIDERA T10NS
Situation - Other considerations may include hazards that aren't listed above
such as wooden fencing connected to a house (a fire path?), or a flammable
hedge that leads to the house, deck, or other combustibles. These arrangements
are referred to by firefighters as "wicking".
Intent - To identify additional hazards.
Recommendations - Contact your local fire department or Oregon Department
of Forestry to help you recognize additional hazards, discuss current regulations,
and other fire prevention topics.
Thank you for your interest in fire prevention and life safety. For more
information, please contact:
Dennis Turco
Fire Prevention Specialist
Oregon Department of Forestry
(541) 664-3328
dturco@odf.state.or.us
5
....,.
l"~.'_n'~_"~'<",_""""__-I-_"'_'____"___'_"
9/99
Fire-resistant plants suitable for the
Southwest Oregon climate
GROUND COVERS
WaalyYarraw (Achillea tomentosa) best in full sun, requires very little water
Dwarf Coyote Brush (Baccharis pilularis) best in full sun, requires monthly watering
African Daisy - Cape Marigold (Oimorphotheca) best in full sun, requires little water
Creeping Rosemary (Rosmarinus officinalis) best in full sun, requires little water once
established
Vinca - Periwinkle, Myrtle (Apoeynaceae) best in shade, requires moderate watering
Madagascar Periwinkle (Catharanthus) best in full sun to part shade, requires little water
Sunrose (Helianthemum) best in full sun, do not over water
leeplant (Mesembryanthemum) best in full sun, water during growth & bloom
English Ivy (Hedera) does well in any exposure, water needs vary, water as needed
Gazania (Asteraceae compositae) best in full sun, water occasionally
SHRUBS
Rackrase (Cistus) best in full sun, requires little water once established
Carmel Creeper (Griseus horizontalis) best in full sun to part shade, water as needed
Italian Buckthorn (Rhamnus ala tern us) exposure needs vary, water as needed
Caffeeberry (Rhamnus califomica) exposure needs vary, water as needed
Cascara Sagrada (Rhamnus purshiana) exposure needs vary, water as needed
Saltbush (Atriplex) best in full sun, requires very little water
Oleander (Nerium Oleander - dwarf variety) best in full sun to part shade, loves heat, requires
little water
TREES (common names)
Quaking Aspen
Cherry
Canyon Live Oak
Maple
Paper Birch
Poplar
It is important ta note that a plant's fire performance can be seriously compromised if not
maintained. Plants that are not properly irrigated or pruned, or that are planted in climate
areas not generally recommended for the plant, will have increased fire risk and will
likely make the mature plant undesirable for landscaping in high fire hazard zones.
- '-"--"'T""'"""--r-- -----.----.....-
INFORMA TION RESOURCES
For fire preven'tion and fire-resistant plants
INTERNET SITES
Ilttp//firewise.Oill
http.//www.prefire.ucfpl.ucop.edu/
http//www.prefire.ucfpl.ucop.edu/veqetati.htm
http//selectree. caw. ca~edu/
tJ.J..J.Q//firesafecouncil. orq/
OREGON DEPARTMENT OF FORESTRY*
5375 Monument Drive
Grants Pass, Oregon 97526
(541) 474-3152
5286 Table Rock Road
Central Point, Oregon 97502
(541) 664-3328
* For information on responsibility of seller notifying buyer of reforestation
requirements: ORS 527.665
COUNTY PLANNING DEPARTMENTS
Jackson County Planning and Development Services
10 S. Oakdale, Room 100
Medford, Oregon 97501
(541) 774-6900
Josephine County Planning Office
510 NW 4th Street
Grants Pass, Oregon 97526
(541) 474-5421
OSU EXTENSION SERVICES
(541) 776-7371
LOCAL FIRE DEPARTMENTS
LOCAL NURSERIES
COLLEGES
{
~
f.
1
PACIFICORP: OREGON ENERGY OUTLOOK ANALYSIS
1. What is the outlook for energy prices and supplies?
prices
Average market prices for electricity in the Pacific Northwest are expected to be higher in
200 I than they were in 2000. PacifiCorp expects power prices to be approximately three
times as high as what is reflected in its current rates.
There arc many reasons for the high and volatile market prices, including increased
demand, insufficient new generation and high gas prices: PacifiCorp expects energy
prices to remain high until these market problems are corrected. The FERC has initiated a
proceeding to address some of the price and supply issues, particularly as they arise in the
California market. Increased market stability and less anxiety over evolving rules under
which the industry operates should lead to the development of new generation sources.
Moreover, new gas supplies will eventually make their way to market. These
developments should help drive prices down to more manageable levels and enable
utilities to better meet demand, which is likely to remain high.
In the meantime, PacifiCorp is focusing on keeping its existing resources as low cost and
as fully operationa I as possible.
In 2001, PacifiCorp expects to generate about 65 percent of the power it will deliver to
customers and purchase 35 percent from other sources, some if it under long term
contracts that are not influenced by the uncertain market. Accordingly, in this period of
high energy prices, PacifiCorp customers are better protected than those of utilities that
do not generate as high a proportion of their own power. Nonetheless, PacifiCorp
purchases enough supplemental power from other sources that its overall costs are
severely impacted by these unprecedentedly high prices.
Supplies
Regionally, supplies are going to be limited in 2001. PacifiCorp's system generates
8,000 MW. In addition, the Western Systems Coordinating Council expects 3,300 to
3,550 MW of new generation to come on line between 4th Quarter 2000 and 4th Quarter
2001. Included in the new generation is the Klamath Cogeneration Project, a partnership
ofPacifiCorp Power Marketing and the City of Klamath Falls. At least one energy
marketing and forecasting company is predicting close to a 2.75 percent growth in energy
demand for WSCC-US for 200 I This amounts to a 2, 100 average MW increase in
demand, or 3,200 MW increase in peak, assuming 65 percent load factor. Accordingly,
new generation in the WSCC will barely keep up with demand growth, especially in
summer.
,...<.......~"......----t'._._~.,.,.-.., ._.".~_......___~....,'_u~.,_____.
2. \Vhat does the outlook mean fo." Oregonians?
P r~es
Oregonians wi I! see their rates change in 2001.
PacifiCorp's residential customers will experience a briefrate decrease of approximately
4 percent in January 200 I attributable to the Scottish Power merger and Centralia gain-
on-sale credits However, the high energy prices and other factors have led PacifiCorp to
request a price increase that, when fully implemented on October 1, 2001, will result in
prices 14 9 percent higher than they are today (this excludes the three percent public
purpose charge and low income charge required by SB 1149). Commercial and industrial
customers will experience price increases ranging from 4 to 15 percent effective August
2001.
PacifiCorp also has requested regulatory deferred-accounting treatment on expected
excess costs for purchased energy anticipated before the new rates go into effect. These
may be passed on to customers at a later date.
Supgly
In general, PacifiCorp expects to be able to meet customer demands in 2001 because the
Company's owned and long-term firm resources cover a significant portion of the
Company's resource requirements
However, during high demand periods and when units are offline, the Company
purchases power, sometimes in significant amounts, to meet resource requirements. In
addition, the Company purchases power throughout the year to optimize the Company's
overall system This is especially true when, due to hydro variability, the Company's
hydro projects do not produce as expected. If supply to the region is limited in 2001, in
certain worst case scenarios such as an extended cold/heat spell and related equipment
failure, it is possible that there could be brief limited outages in PacifiCorp's service area.
3. What can Oregonians do to manage their energy costs?
PacifiCorp has several programs for assisting Oregonians with managing their energy
costs. PacifiCorp's initiatives include various brochures on how to reduce energy costs,
webpage tips on energy savings and the Home Energy Checkup program. An
information packet on the Home Energy Checkup Program is attached, complete with:
. A detailed checklist for Oregonians who want to reduce their energy costs;
. Information about the Oregon Residential Energy Tax Credit; and
. Specific information about "Instant Energy Savers"
2
.-.....-..........--'.
4. Are there Oregonians whose job or welfar'e will be at risk? What actions can
energy suppliers or the State of Oregon take to help Oregonians?
PacifiCorp is very concerned about the impact of high prices and limited supply on its
industrial, commercial and residential customers. PacifiCorp has three major projects in
the planning and/or implementation stages that would help mitigate possible harm in its
multi-state service area:
PacifiCorp is considering programs that would provide customers with
incentives for not using high-energy appliances and other devices during
peak times, such as late afternoon and early evening Doing so would help
reduce demands on the regional system during peak hours. It also would
decrease the frequency with which PacifiCorp is required to purchase
supplemental power at the most expensive times of day.
PacifiCorp is considering beginning to offer programs in selected areas in
2001 and may expand as appropriate later
2 PacifiCorp is discussing the possible development of a demand-side
bidding/buy back program similar to programs offered to industrial
customers by the Bonneville Power Administration and Portland General
Electric. From PacifiCorp's perspective, the useful feature would involve
a posted price for customers who want to voluntarily participate in the
program.
PacifiCorp is also in discussion with some customers regarding
coordinating their energy consumption planning and schedules, such as
maintenance scheduling, with the Company's load requirements.
3. PacifiCorp is teaming up with an energy office in another state to develop
an aggressive energy savings campaign, "Save It." The campaign, which
includes public service announcements, emphasizes the broader benefits
of energy savings in such areas as the environment and economic
development The goal is to change usage patterns away from peak times
by focusing on how doing so benefits the community.
PacifiCorp would welcome a partnership with the Oregon Office of
Energy to develop a similar campaign in Oregon
In addition to the above projects, PacifiCorp committed $73 million to Oregon-based
energy efficiency projects in 2000 and expects to expend at least that much in 2001.
Prepared by Peter Cogswell
PacifiCorp Government Affairs
(503) 813-5275
J
. n .... r"
'-~.".....,....,.~--'
AI \ z .If V 7 l -:JPt" ^H /5)1.1/
IV It: \.Y~ltlv.. v I tl f -
/' U-j
James K. Agee, College of Forest Rcsources. University of Washington, Box 352100. Seattle. Washington 98195
The Landscape Ecology of Western Forest Fire Regimes
Abstract
Fire has had a major role in shaping the forested landscapes of the American West. In recent decades. major efforts to quantify that
role have been made. and characteristics of historic tire regimes have been defined: frequency, magnitude. variability. seasonality,
synergism, and extent Together, these char,lcteristlcs also defined the histOlic landscape effects of fire in low-. moderate-, and
high-severity fire regimes. Coarse-filter conservation strategies typically rely on knowledge of natural disturbance regimes to
define appropriate forest structure goals, ooth at the stand and landscape scale. and these will differ by fire regime. Historic patch
size increased across the low- to high-severity spectrum. but edge was maximized in the moderate-severity fire regime. Fire
exclusion in the 20th century has caused two major types of landscape change: loss of openings in once patchy landscapes, and
imposition of high-severity landscape dynamics in areas where wildfires that escape suppression now bum. Effects of historical
fire regimes may be in some cases eiL.1er difficult to mimic or undesirable.
Introduction
Fire has been a central theme of American forest
management during the 20th century. It was the
impetus for the first legislation allowing Federal
forestland purchase (the Weeks Act of 1911), the
first legislation allowing Federal cost-share pro-
grams to the States (Clarke-McNary Act of 1924),
and the focus on impressive technology devel-
oped for purposes of fire control. The control of
forest fire has been one of the most intensive natural
resource investments made by government, yet
paradoxically its success has resulted in a fire
cont.rol problem that now commonly overwhelms
firefighters (Brown and Arno 1990).
Emerging concerns related to biodiversity have
stimulated efforts to favor more "natural" forms
of management, emulating historical disturbances
within the "natural range of variability" (Morgan
et al. 1994). Biodiversity plans can be classified
into coarse and fine filter approaches, and usu-
ally are a combination of both, Coarse filter ap-
proaches focus on management at the ecosystem
level (Hunter 1990), with the assumption that
naturally functioning ecosystem processes will
create and maintain appropriate forest structures
necessary for biodiversity maintenance (Karr and
Freemark 1985, AttiwilI 1994, Swanson et al.
1997). Where this approach leaves certain spe-
cies or guilds at risk, fine filter approaches that
manage at finer scale arc also implemented (Hunter
1990, Haufler et al. 1996). The coarse filter ap-
proach can be successful only if the landscape
ecology of natural disturbance is known, and an
eventual substitution of a few coarse filter ap-
proaches for a plethora of fine-filter approaches
can be justified only if the coarse filter meets the
needs of the fine filter species,
The role of fire in landscape ecology is con-
founded by a lack of understanding of the rela-
tionships between pattern and process, Pattern,
or the architecture of the forest as described by
species composition and structure, including fuel
amounts, size classes, and arrangement, clearly
affects the manner in which the process, fire, burns,
Yet the behavior of a fire is only partly depen-
dent on pattern, as the fire behavior "triangle"
includes not only fuels and topography but also
weather (Agee 1997), which is marginally influ-
enced by pattern. The objectives of this paper are
to describe what is currently known about land-
scape character of western forest fire regimes and
relate these to pattern and process, including
management implications.
Fire has been the most pervasive natural dis-
turbance factor across Western forest landscapes
(Spurr and Barnes 1980), but it did not work in-
dependently of other disturbances. To avoid con-
tradictions in scale terminology (e,g., Silbernagel
1997), fine scale will refer to minute resolution
(large scale in a cartographic sense) and coarse
scale will refer to broad areas (small scale in a
cartographic sense). Fire has had both fme and
coarse scale effects on the forests of western North
America (Agee 1993), but these cffects differed
considcrably by fire regime.
-"-y--""'-4'" ..,..
The Fire Regime
Natural disturbances range from benign to cata-
strophic, and can be generatee! from within or
outside of thc ecosystem (White 1987). The dis-
turbance effects in eithcr case are due in part to
current pattern or structure and to the nature of
thc disturbance. Disturbance is usually charac-
terized by a combination of factors: type, frequency,
variability, magnitude, extent, seasonality, and
synergism with other disturbances (White and
Pickett 1985). Western forest fires have a wide
range of historic frequencies from less than 10 to
over 500 years that vary considerably by forest
typc. Prcdictability is associated with variability,
and either very short or very long tire return in-
tervals compared to the average interval can have
major ecological effects. Non-sprouting species
killed by one fire can be locally extirpated by a
second closely-spaced firc; when fire intervals are
unusually long, fire-sensitive species may pass
through the critical period of their life history.
Magnitudc is often described as firelinc intensity,
a measure of encrgy output rclated to flame length,
although other lcss predictable factors such as
duration of smoldering can also be important.
Extent describes the scale of the fire, but is gen-
erally poorly related to fire effects without knowl-
edge of magnitude. Seasonality describes when
fires occur in the year, In the Amcrican South-
west, spring months are infcrred to be the most
common season (Swetnam and Betancourt 1990),
while in the Pacific Northwest, mid- to late sum-
mer appears to be the most conunon seaSon (Wright
1996), Synergism, or the interaction of fire with
other disturbances, is poorly understood and gen-
erally unpredictable, Insects, disease, and wind
may follow fire events with more than endemic
background effects, and conversely, accelerated
fire effects may follow other disturbances. Many
secondary effects such as soil mass movement
may follow intense fIres (Swanson 1981),
The fire regimes of western forests are usu-
ally described in terms of historical fIres, and in-
terpreted much the same way as potential veg-
etation (e.g., Daubenmire 1968): what occurred
historically and what the trajectories of change
may be with or without management (Agee 1993).
Fire rcgimes based on fire severity (Agee 1993)
are defined by effects on dominant organisms,
such as trees, and although broadly described in
three classes, can be disaggregated to the forest
type or plant association level if desired. The ap-
proach below is to use these broader classes as
an organizing paradigm within which individual
forest types arc discusscd. The l~ily..firc
regi)IleS wcre those in which t~~. _~.f(~~L9f a fire
was usuaIIy a stand replaccmcll-t event (Figure I).
The low-severity fire regimcs were those in which
thc typical fire was benign to dQminant organ-
isms across much of the area it burned, while the
moderate-severity fire regimes .l1ad.as:onml~x. mix
A
Western
hem..lock:
Douglas-fir
B
~
...
co:
....
Q
C
II
U
...
..
~
..
..
"0
.:
&.
i
E
'-'
~
=
e
x.
E
~
Ponderosa
pine
\ Mixed ..--
\ Conifer ",."
,,--.....
I
I
I
I
High /Do Drygl fir Moderate
I U as-
Pacific ~
silver fu \
\
,
,
Moisture Slress Index
100
Low
lntensit y Fi res
High
lntensity Fires
Moderate
lntensity Fires
50
-'- ./ ./'
/".
/ -
/
/
/
/
......
'-
"-
"-
"-
"-
- -. - "
-.,---- "
o
Low
\toderate
High
Fire Regime Severity
Figure 1. A. Historic fire regimes of tbe Pacific Northwest
can be broadly defined into three categories: low-,
moderate- and high-seventy. Each fire regime has
a number of forest types within it that have slmilar
landscape patterns created by fire. B. Historic fire
intensity and associated effects varied by fire re-
glfne.
Figure 2. Ponderosa pine forest has a classic low-severity fire
regime. Patches on this landscape may be as small
as 0.0 I ha. This pattern is disappearing across the
range of the species by ingrowth of trees Slnce the
fire exclusion period and selective harvest of the
older clumps of trees
of only very small patches. The data shown in
Table I are not a comprehensive list of patch sizes
in western forests but are representative of those
SIzes.
The best example of large fires and small patches
(reg~n:cration) are found in the low-severity flfe.
regimes (Figure 2) Fires burned frequently in these
forests, and by regularly consuming fuels, kill-
ing small trees and pruning the boles of residual
trees, maintained a relatively fire-resistant land-
scape across which overstory mortality from fire
was rare. Forests with a large component of pon-
derosa pine commonly had very small patch sizes
(Tabk 1), although historic fires commonly ranged
over hundreds to thousands of ha (Wright 1996).
The Londition that caused the patch was in many
cases senescence of an old group or trees, bark
heetle attack. and subsequent consumption ot' the
debris by fire, resulting in a range of patch sizes
usually < 0.4 ha. The rest of the forest, for most
ecological purposes, was a fairly uniform mosaic
or mature tree clusters and grassy understories.
Larger patch sizes are typical of moderate-se-
verity fire regimes. Although the lower end or the
size range is within that of the low-severity fire
regimes, much larger patch sizes also occur (Table
I). These patches are defined by the amount of
mortality created by the fire, with low-severity
patches underburning with little mortality, mod-
erate-severity patches having some mortality but
substantial numbers of residual trees in the larger
size classes, and high-severity patches where most
trees have been killed (Figure 3), The low- sever-
ity patches may appear much like the unburned
forest, while the moderate-severity patches will
often develop a multiple-age structure. High-se-
verity patches will develop an even-aged struc-
ture if regeneration is immediate, or may revert
to nonforest vegetation if tree seed sources are
limited (Chappell and Agee 1996). Smaller patches
related to canopy gaps at the scale of 0.0025-.04
ha (Taylor and Halpern 1991) also occur in these
forests, and large stand replacement patches >500
ha have occurred in red fir forests in California
(D. Parsons, Sequoia-Kings Canyon National
Parks, pers. comm.),
High-severity fire regimes often have fire events
that are driven by extreme weather (Bessie and
Johnson 1995, Agee 1997), Although the major-
ity of fires in such areas are very small, most of
the area burned is from the few larger fires, They
occur infrequently and in ecosystems where most
or all of the trees species are not adapted to sur-
vive intense fires, so the result is usually a stand
replacement fire event (Figure 4), Because fires
are infrequent, forest structures are often late-suc-
cessional, with multiple crown layers and high
susceptibility to crown flfe behavior. Wind-driven
events can create patches many thousands of ha
in size (Table 1), Boreal forests (Johnson 1992),
subal pine forests (Agee and Smith 1984), and wet
coa<;tal forests (Heinrichs 1983) are the most wide-
spread examples of this type of fire regime in the
West. As with the high-severity patches in the
moderate-severity fire regime, post-fire tree re-
generation depends on seed source availability
or vegetative regeneration.
"..,., ,..~. __."'_'__"__'_"'~""'W""'~"'_'_^"_~
o~s,:,verityJ~y~ls. Thcse artificial classes obscure
the variation that is bctter captured in Figure 1.
Nevertheless, the landscape effects of historic fire
regimes at the left, middle, and right side of the
spectrum were quite different from one another.
Landscape Metrics
There are myriad metrics that may be generated
for landscapes, and the important ones may dif-
fer depending on the problem (McGarigal and
Marks 1994). Patch size, edge, shape, core area,
neaTest neighbor, and diversity metrics are among
the most corrunon, For historical fire regimes, these
metrics are rarely available in quantifiable form,
and because the pattern of scale is so variable,
metrics are not easily compared across fire re-
gimes. The "grain" size may be much less than I
TABLE I. Patch size character of western forest fire regimes.
ha in historic ponderosa pine (Pinus pondcrosa)
forests (White 1985), but thousands of hectares
in subalpine or boreal forests (Bessie and Johnson
1995). Two landscape metrics are compared be-
low across the spectrum of Western forest fire
regImes.
Patch Size
Patch size as used here refers to openings crcat~d
.2L.tl:re withi~ whi<;h post-fi!"e regeneratio_~j~Jil<eJY
J.Q_Qc..<;.urandpersi~t. This is a subjective defini-
tion but one that helps define an ecologically sig-
nificant and recognizable shift in forest structure.
Patch size di ers from fIre extent in that a fire
mAY'~read widely across a landsc;w ut may
cr~at~_G...ondltions for reeeneratiQQ2.~~yjn ~~l~cteq
loc~ A large fire can be associated with creation
Severity of State/
fire regime Province rorest type
Low AZ Ponderosa pine (Cooper 1960)
Low AZ Ponderosa pine (White 1985)
Low OR Ponderosa pine (West 1969)
Low OR Ponderosa pine (Morrow 1985)
Low CA Mixed conifer
(Bonnicksen and Stone 1981)
Moderate OR Red firl
(Chappell and Agee 1996)2
Moderate OR Red fir
(Chappell and Agee 1996)3
Moderate OR Douglas-ftr'
(Morrison and Swanson (990)
Moderate OR Douglas-firS
(Morrison and Swanson 1990)
High WNOR Western hemlock-Douglas-fIt'
(Agee 1993)
High ID Western hemlock6
(Stickney 1986)
High MTIWY Lodgepole pine-subalpine fir6
(Romme and Despain (989)
High OR Mountain hemloc0
(Dickman and Cook 1989)
High AL White and black spruce1
(Eberhart and Woodard 1987)
Patch size (ha)
Mean Median
Range
0.06-0. I 3
0.02-0.29
.25
0.025-0.35
0.03-0.16
2.67
0.84
0.11- 31.09
1.34
0.39
0.12-10.08
8.46
2.22
0.13-74.71
] 1.03
2.70
o 15-253.23
>10,000
>[0,000
>10,000
>3,200
0.01-17,700
I red fir = Abies nwgnifica. western hemlock = Tsuga helerophylla, lodgepole pine = Pinus contorta, subalpine fir = Abies
Lasiocarpa, mountain hemlock = Tsuga merlensiana, white spruce = Picea glauca, black spruce = Picea mariana
2 Goodbye fire, remeasured from maps in report
] Desert Cone fire, remeasured from maps in report
· l893 Event Cook/Quentin Creek. remeasured from maps in report
1 Fires of 1800-19lX) Cook/Quentin Creek. remeasured from maps in report
o Actual patch size may have ranged from 00 I ha spots to Pluch larger sizes than noted
7 Fires <20 ha were omitted from tile analYSIS but occur in the area
Patch Edge
TIle metric chosen to compare edge across fire
regimes is the edge index used by Eberhart and
Woodard (1987). They calculated the edge index
by measuring the perimeter of all burned areas,
including unburned "islands" within the bum, and
computing a ratio between that total perimeter
and the perimeter of a circle of the same area as
the fire (representing the minimum edge condi-
tion). Larger values of the index represent fires
with higher proportions of edge relative to patch
size. This index was adapted for use here by in-
corporating the edge created between patches of
varying severity in moderate-severity fire regimes,
not just the edge between burned and unburned
areas used in high-severity fires. Any edge index
has inherent limitations based on image interpre-
tation, based on the minimum patch size recog-
nized, how many fire severity classes are defined,
and the range of fire severity included in each
class. In this simplified analysis, only three fire
severity levels were used, based on overs tory
mortality, and specific edge indices are not inter-
preted as absolute values but as relative values
comparable broadly between fire regimes.
This index could not be computed for low-se-
verity fire regimes, because edges are diffuse ex-
cept where small old patches are "decaying:' and
where fIres may be more intense (Agee 1993).
Such patches "wink" in and out as they blend with
older forest patches nearby. Structural differences
between a 150-year-old patch and an adjacent 250-
year-old patch are so slight as to be ecologically
meaningless. An edge index for low-severity fire
regimes would probably be less than 1, as in any
defined fire size, that could be represented as a
circle of the same area, the perimeters of the small
patches where fire would be intense would likely
be less than the perimeter of the fire size circle
(which itself would not count as "fire perimeter"
because it does not necessarily create any edge).
Moderate-severity fire regimes appear to have
considerably more edge than low- or high-sever-
ity fire regimes (Table 2). Because these values
were taken from only a few fires, there is prob-
ably a wider range of edge index values than shown
in the table, but the range from 6 to 20 suggests
that moderate-severity fire regimes create substan-
tial patchiness on the landscape. These fires typi-
cally burned for months (van Wagtendonk 1985),
TABLE 2. Patch edge character of Western forest fire regimes
Severity of
Fire Regime
Forest Type
Edge
Index
Moderate Red fir' 11.M
(Agee and Chappell 1996)
Moderate Red fir 6.19
(Agee and Chappell (996)
Moderate Douglas-firJ 11.16
(Morrison and Swanson 1990)
Moderate Douglas-fir' 21.79
(Morrison and Swanson! 990)
High Western hemlockfDouglas-fir 3.72
(Hoh fire-500 ha-Olympic National Park)
High White spruce-black spruce
(Eberhart and Woodard 1987)
20-40 ha fires 2. [7
41-200 ha fires 3.29
201-400 ha fires 3.48
401-2000 ha fires 5.11
2001-2??oo ha fires 7.47
, Goodbye fire, remeasured from maps in report
2 Desert Cone fire, remeasured from maps in report
J [893 Event Cook/Quentin Creek. remeasllred from maps in
report
· Fires of 1800-1900 Cook/Quentin Creek. rcmeasurcd from
maps in report
and burned under severe and benign fire weather,
across complex topography, during the day and
at night, such that substantial variation in burn-
ing conditions resulted. In addition, fuel varia-
tion caused fires to stop or slow at boundaries of
previously burned areas (van Wagtendonk 1985).
High-severity fire regimes have lower edge than
moderate-severity fire regimes, but there appears
to be overlap, particularly as high-severity fires
become larger (Table 2). Wind-driven fires tend
to be elliptical in shape rather than circular (Ander-
son 1983), so an edge index> I is almost certain
even in uniform terrain and fuels. Larger fires tend
to be those that burn over longer periods and are
associated with more weather variation and a higher
probability of the head moving in more than one
direction. This is likely to create more edge. In
addition, larger fires tend to have larger unburned
islands (Eberhart and Woodard 1987) and this is
associated with increased edge effect. Simulated
fire spread models for boreal forests (Ratz 1995)
have produced edge index values similar to those
shown for high-severity fire regimes in Table 2.
Figure 3. Reel fir forest has a moderate-severity fire regime. Stand replacement patches are mixed with those where thinning of
the overstory dom inants has occurred, and those where light underburns have no effect on the overstory at all. Patches
in the foreground that appear to have a smooth canopy texture are stand replacement patches from a fire many decades
old, while a new stand replacement patch IS visible in the background. The rest of the landscape has burned with lower
severity fire.
Figure 4 Subalpine forests, with a high-severity fire rl"gimc. may remain treeless for a century or more after being burned.
Interior areas of large patches arc slower to recolonize than areas near the edge where seed IS more likely to blow in
ffl)fTl adjacent unburned forest
Patch Characteristics, Fire Severity, and
Implications for Management
The landscape metrics of historical forests of the
American West differed by fi.re regime (Table J,
Figure 5). Fires of large extent were common 1[1
all fire regimes. but thcir effects on the lanuscapc
were quite different. Fine-scale pattern was cre-
ateu and maintained in low-severity fire regimcs,
while coarsc-scale pattern occurred in high-sc-
verity fire rcgimes. Wherc forcst types of uiffer-
ent fLce regimcs wcre closely juxtaposed, the
characters of each intenmngled (Agee ct al. 19<)0).
A small inclusion of cool, moist forest cIassifi.eu
as a high-severity flrc regime, surrounded by a
much larger landscape of dry, warm forest with a
low-severity fire regime, tcnded to have some of
the character of the low-severity type: patchier,
morc frequent fire with smallcr patch size and
more edge than found where the type was widely
distributed. The landscape context of the forest,
including landform effects (Swanson et a!. 1988),
inherent edge (Yahner 1988), and the adjacent
forests with their characteristic fire-induced patch
TABLE 3. Relative landscape characters of Westem forest fire regimes.
Fire Regime
Landscape
Character Low-severity Moderate-severity High-severity
Patch Size' Small (- I ha) Medium (1-300+ ha) Large (1-1 ??oo+ ha)
Edge Low Amount High Amount Moderate Amount
Pre-Post Fire
Similarity2 High Moderatc Low
I Thc averagc patch within which tree regeneration wilt be open-grown.
, Of the total area burned, the propoI1ion resemhltng the pre-firc forest structure.
Low-Severity Fire Regime
Moderate-Severity Fire Regime
High-Severity Fire Regime
.
e
.
.
<!)
e
o
.
D Low-Severity Patch
III Moderate-Severity Patch
II High-Severity Patch
Figure S ^ schcmatlc of landscape pattern of fire rcgimcs. Black dots III low-severity fire regimes are very old patches of large,
<lId trees being killed by illsects Jnd decomposed by fire, Jnd gray dots are emerging pole-size stands that have less-
tlerlned edge. The mo(\nate-severity fire regime IS typICally a complex mosaic of larger patches of the three fire severity
levels, while the high-sc:vl'lIty file rc:glme has large stanJ replacemellt patches.
and edge pattcrns, is important in understanding
thc historic landscapc charactcr of a forest type.
The prcdominance of pa~tcrn vcrsus proccss
as controlling factors of firc and forcst landscape
dynamics will vary by fire regime. Most theo~
retical approaches to disturbance and forest pat~
tern have simplified disturbance and fire to a bi~
nary process: a landscape cell is eithcr disturbcd
or not (e.g., Turner et a1. 1989). These modcls
have utility in high-severity fire regimes (Turncr
and Romme 1994, Ratz 1995), but have less util-
ity for moderate- and low-severity fire regimes,
and present scaling problems when applied to real
landscapes. In the low- and moderate-severity fire
regimes, fires spread widely but had significant
pattern effects on only medium to small portions
of the landscape, and the control of process by
pattern is obvious. Frequent, low-intensity fires
created temporary fire barriers by consuming fu-
els, and resulted in ajigsaw-like shape of subse-
quent fires. This has been shown by Wright (1996)
for ponderosa pine, a low-severity fire regime,
and by van Wagtcndonk (1985) for red fir, a mod-
erate~severity fire regime. Pattcrn, as represented
by age-class distribution and spatial stmcture, was
so tire-tolerant in low-severity fire regimcs that
while the forest was dependent on fire in the long
run to maintain its pattern, it was relatively im-
mune to severe short-ternl effects; the interaction
of pattern and process resulted in a quasi-stable
system. While a true equilibrium system with
balanced age classes has not been shown even
for ponderosa pine (Cooper 1960), the low-se-
verity fire regimes were much marc stable than
high-severity fire regimes.
Under "normal" weather, pattern has also been
shown to control process in high-severity fire re-
gimes. Naturally-occurring fires in older forests
at Yellowstone National Park. between 1972 and
1987, tended to slow or stop at boundaries of young
forest with simpler, less fire-prone structure
(Despain and Sellers 1977, Romme and Despain
1989). However, very large hist(;ric fires appear
to ha ve been the source of much of the widespread,
older forest (Rom me 1982), indicating that pro-
ccss must have overwhelmcd pattemat some dis-
tant time in the past. The Yellowstone fifes of 1988
hurned forests of all ages, indicating that process
overwhelmed pattern (Romme and Despain 1989),
which has also been documented for Canadian
subalpine forests (Bessie and Johnson 1995). While
an individual stand fllay appear stahle OV'~r time
to a human observer, the landscapes of high-se-
verity fire regimes (a specific spatial scale) Over
the likspan of a tree (a specific temporal scale)
are considered non-equilibrium systems (Baker
1989, Turner and Romme 1994) because of the
nature of the disturbance process: infrcquent, large,
severe fires that have a persistent crfect (centu-
rics-Iong) on landscape pattern.
Modern forestry has had significant effects on
landscape pattern, but probably the most perva~
sive effect has been that of fire exclusion. Effects
of fire exclusion are extensive inlow-severity fire
regimes (Weaver 1943), less in moderate-sever-
ity fire regimes, and least in high-severity fire
regimes, because fire has been removed for more
fire-return intervals in 10w- and moderate-sever-
ity fire regimes than in the high-severity tire re-
gimes. By allowing forest patches in low-sever-
ity fire regimes to converge in structure, developing
multi-layered character with increased fuel loads,
the infrequent wildfire that escapes control un-
der severe weather conditions now has much more
severe effects (Agee 1997). Forest openings, once
characteristic of many fire-prone landscapes, have
decreased in size as surrounding forests have be-
come more dense (Skinner 1995). Much of the
induced edge that persisted in a shifting mosaic
through the 19th century is now a subtle edge
between mature and old-growth forest (Morrison
and Swanson 1990), and these landscapes are now
more prone to high-severity fire. A persistent but
more unstable landscape pattern is being created.
not only in patch metrics but in susceptibility to
future severe fire. Higher proportions of post - fire
regeneration in sprouting hardwoods and seroti-
Ilous-coned pines will be more likely to be ki lied
by future fires than the more fire-tolerant mature
pines, larches (Larix occidentalis), and Douglas-
firs (Pseudotsuga menziesii) they replace.
Wide riparian forest buffers are being prescribed
for some western forests. While they have been
defended on the basis of protection for aquatic
organisms, including endangered fish, corridors
for wildlife, and sources of coarse woody debris
for future stream habitat, they have also been
documented, under some conditions, as corridors
for severe wildfire (Segura and Snook 1992, Agee,
pers. obs.). There are few data with which to evalu-
ate thc flammability of riparian zones. Marc com-
plex structures often do occur in riparian zones,
but these may be due to better site quality, allow-
ing faster post-fire succession, or 10 less frcquent
'-...-...._.~.~.--'
or [owcr scvcrity fire in these shelrercd locations
(Romme and Knight 1981). In thc dry eastcrn
Washington Ca..'ieades, high-elcvation forest rcfugia
(arcas less likely to burn) wcre identifled as oc-
curflng abovc 1500 rn clevatlOn on north aspccts,
and often adjacent tu the connucnces of peren-
nial streams (Camp ct a1. (997). [n northcrn Cali-
fornia mixed-evergreen forest, late-succcssional
forest structure is most likely to be found in lowcr
slope positions and on north and east aspects (Tay-
lor and Skinner in review). Conversely. in west-
ern Idaho, riparian zones in some locations have
burned marc severely than associated uplands
(Figure 6). Clearly, complex interactions are oc-
curring betwecn process, pattern, and landscape
position of riparian forests, and need to be evalu-
ated in more depth.
The concept called "natural range of variabil-
ity" has been proposed as an appropriate coarse
filter approach to ecosystem management. Simu-
lating natural disturbance processes and patterns
is one way to maintain broadly-defined ecosys-
tem productivity (Attiwill 1994, Swanson et al.
1997). In most Western forest ccosystems, there
is still enough residual cvidence in live trees and
stumps to allow reconstruction of historic land-
scapes over time (Agee, 1993, Wallin et a1. 1996).
In high-severity fire regimes, Baker (1994) sug-
gested that reintroducing the process of fire might
itself be sufficient to restore the natural pattern.
In low-severity fire regimes the issue has been
debated (Bonnicksen and Stone 1982, Parsons et
al. 1986), but the debate has centered more on
objectives of pattern or process rathcr than whether
some type of reconstruction was desirable. Al-
though these debates were first associated with
natural areas, the trend towards coarse filter con-
servation strategies for many forest lands has ex-
panded the potential implications of these argu-
ments to much of the forested land of the West.
There is no "right answer" in these arguments,
but it is clear that in low-severity fire regimes,
modification of fuel structures by underburning
or thinning will move the system towards a more
natural pattern. Continuing to rcmove largc green
trees in conjunction with salvage logging will
exaccrbate currcnt conditions (Agee 1997). [n
moderate-severity fire regimes, timber harvest-
ing that moves away from traditiol1allarge clcarcuts
to (a) partial cuts, (b) small patch cuts with snag
retention, and (c) a system of rescrves, utilizing
Figure 6. Portions of the riparian zone of Little French Creek.
Payette National Forest, Idaho, burned much more
severely than adjacent uplands. The riparian zone
(A) had substantial dead Engelmann spruce (Picea
engelmannii) and multilayered structure, and
burned with a crown fire. The upland forest (B),
which had burned in ca. 1900 and again in 1933,
was composed of self-pruned, widely spaced lodge-
pole pine with a huckleberry (Vacciniwn scoparium)
understory and little coarse woody debris. and spot
fires here did not spread (U.S. Forest Service photo
by Morgan Beveridge).
prescribed fire even i~ reserves, will create more
natural pattern than either past management or a
pure reserve system with no recognition of pro-
cess. [n high-severity fire regimes, large patch sizes,
although perhaps historically present, will be dif-
ficult to managc for and may be perceived as a
"catastrophe best to be avoided" (Hunter 1993).
Whcn the severc weather event occurs, wc may
not have as much control over nature as we think,
so large patch sizes will probably occur in these
systcms regardlcss of our desircs.
Acknowledgements
This paper was suppotted by USDA Forest Ser-
vice Cooperative Agreement PNW 93-040 I be-
tween the Pacific Northwest Research Station and
Literature Cited
Agee, I.K. 1993. Fire Ecology of Pacific Northwest Forests.
Island Press. Washington, D.e.
. 1997. The severe weather wildfire: too hot to
handle? Northw. Sci. 71: 153-156.
Agee, 1.K., M. Finney, and R. deGouvenain. 1990. Forest fire
history of Desolation Peak, Washington. Can. I. ~or.
Res. 20: 350-356.
Agee, I.K., and L Smith. 1984. Subalpine tree reestablish-
ment after fire in the Olympic Mountains, Washing-
ton. EcoL 65: 810-819.
Anderson, H.E. 1983. Predicting wind-driven wild land flIe
size and shape. USDA For. Servo Res. Pap. INT-305.
Intermountain Research Station. Ogden, Utah. 26 pp.
Attiwill. PM. 1994. The disturbance of forest ecosystems:
The ecological basis for conservation management.
For. EcoL and Manage. 63: 247-300.
Baker W.L. 1989. Effect of scale and spatial heterogeneity
on fire-interval distributions. Can. 1. For. Res. 19: 700-
706.
. 1994. Restoration of landscape Structure altered
by fire suppression. Conserv. BioI. 8: 763-769.
Bessie, w.e. and E.A. Johnson. 1995. The relative impor-
tance of fuels and weather on fire behavior in subal-
pine forests. Ecol 76: 747-762.
Bonnrcksen, TM, and E.e. Stone. 1981. The giant sequoia-
mixed conifer forest community characterized through
pattern analysis ilS a mosaic of aggregations. For. Ecol.
and Manage. 3: 307-328.
-~. 1982. Managing vegetation within US. national
parks: A policy analysis. Environ. Manage. 6: 101-
102, 109-122.
Brown, J.K., and S.F Arno. 1990. The paradox of wildland
fire. Western Wildlands (Spring): 40-46.
Camp, A., e. Oliver, P. Hessburg, and R. Everett. 1997. Prc-
dicting late-successional firc refugia pre-dating Eu-
ropean settlement in the Wenatchee Mountains. For.
&01. and Manage. 95: 63-77.
Chappell, C.8., and J.K. Agee. 1996. Fire severity and tree
seedling establishment in Ahies mugnifica forests,
southern Cascades, Oregon. &01. App1. 6: 628-640.
Cooper. c.F. 1960. Changes in vegetation, structure, and growth
of southwestern pine forests since white settlement.
rA:OI. Monogr. 30: 129-164.
Daubenmire, R.E 1968. Plant Communities: a Textbook of
Plant Synecology. Harper and Row. New York.
Despain, D.G., and R.E. Sellars. 1977. Natural fire in
Yellowstone National Park. Western Wildlands 4: 20-
24.
Dickman, A., and S. Cook. 1989. Fire and fungus in a moun-
tain hemlock forest. Can. J. Bot. (,7: 2005-2016.
the University of Washington. The assistance of
Robert Norheim in analyzing patch characters,
and Donald McKenzie in manuscript reView, is
greatly appreciated.
Eberhart, K.E.. and PM. Woodard. 1987. Distribution of re-
sidual vegetation associated with largc fires in Alberta.
Can. J. For. Res. [7: 1207-[212.
Hauner, 18., CA. Mehl, and GJ. Roloff 1996. Using a coarse-
filter approach with species assessment for ecosys-
tem management. Wild!. Soc. Bull. 24,2: 200-208.
Heinrichs, J. 1983. Tillamook. J. For. 81: 442-446.
Humer, M.L. 1990. Wildlife, Forests. and Forestry: Principles
of Managing Forests for Biodiversity. Prentice-Hall.
Englewood Cliffs, New Jersey.
Hunter. M.L. 1993. Natural fire regimes as spatial models
for managing boreal forests. BioI. Cons. 65: 115- 120.
Johnson, E.A. 1992. Fire and Vegetation Dynamics: Studies
from the North American Boreal Forest. Cambridge
University Press. Cambridge.
Karr, J.R., and K.E. Freemark. 1985. Disturbance and verte-
brates: An integrative perspective. In: Pickett, S.T.A..
and P.S. White (eds) The r-:Cology of Natural Distur-
bance and Patch Dynamics. Academic Press. New York.
McGarigal, K.. and B. Marks. 1994. FRAGSTATS: spatIal
pattern analysis program for quantifying landscape
character. Oregon State University, Corvallis, OR.
Morgan, P, G.H. Aplet, J.B. Haufler, H.e. Humphries, M.M.
Moore, and W.D. Wilson. 1994. Historical range of
variability: a useful tool for evaluating ecosystem
change. J. Sust. For. 2: 87 -112.
Morrison, PH., and F.J. Swanson. 1990. Fire history and pat-
tern in a Cascade Range landscape. USDA For. Servo
Gen. Tech. Rep. PNW-GTR-254. Pacific Northwest
Research Station. Portland, Oregon. 77 pp.
Morrow, R.I. 1985. Age structure and spatial pattern of old-
growth ponderosa pine in Pringle Falls Experimental
Forest, central Oregon. Oregon State University,
Corvallis. M.S. Thesis.
Parsons, OJ., OM. Graber, J.K. Agee, and J.W van
Wagtendonk. 1986. Natural fire management in na-
tional parks. Environ. Manage. 10: 21-24.
Ratz. A. 1995. Long-term spatial patterns created by fire: A
model oriented towards boreal forests. [nt. J. Wild-
land Fire 5: 25-34.
Romme, W.H. 1982. Fire and landscape divcrsity in subal-
pine forests of Yellowstone National Park. Eco!.
Mon06r. 52: 199-221.
Romme, W.H., and D.H. Knight. 1981. Fire frequency and
subalpine forest succcssion along a topographic gra-
dient in Wyoming. F.coL 62: 319-326.
Romme, WH., and D.G. Despain 1989. Historical perspec-
tive on the Yellowstone fires of 1988. Biosci. 39: 695-
699.
Segura, G., and L.e. Snook. 1992. Stand dynamics and re-
generation patterns of a pinyon pine forest in east-
central Mexico For. &01. and Manage. 47: 175-194
Stll1C([1,I~,d. J 1997. Scale pcn:eptloll - trom cartography to
el'ology Bull. Ecol. Sac Amer 7R: 160-169.
Sklnnel. CN 1995. Change in spatIal characteristics of for-
l'S{ openIngs mthe Klamath Mountains of northwest.
em California, USA Landscape Ecol. 10: 219-nS
Spurr. S H , and [3 V. Barnes 19XO Forest E,:ology. llmd
[~J. John Wiley and Sons New York
StIckney. P F 1986. [<irst decade plant succeSSIon follOWing
the Sundance forest fire, northern Idaho USDA For.
Servo Gen. Tech. Rep. INT-197. IntermountaIn Re-
search Station. Ogden, Utah. 26 pp
SWanSl1l1, FJ. 19R 1. Fire and geomorphic processes. In
Mooney, H. et al. (eds) Fire regimes and ecosystem
processes: Proceedings of the conference. USDA For
Serv Gen. Tech. Rep. WO-20. Washington, DC pp
.1I OA21.
Swansoll. FJ.. T.K. Kratz. N. Caine, and R.G. Woodmansee.
1988. Landform effects on ecosystem patterns and
processes. Biosci. 38: 92-98.
Swanson. FJ., J.A. Jones, and G.E. Grant 1997. The physi.
cal environment as a basis for managing ecosystems.
Chap. t5ln: Kohm, K.A. and J.F. Franklin (eds) Cre-
ating a Forestry for the 21st Century. Island Press
Washington. D.C
Swetnam. T.W., and J.L. Betancourt. 1990. Fire-Southern
Oscillation relations in the southwestern United States.
SCience 249: 1017-1020.
Taylor. A.H., and e.B. Halpern. 1991. The structure and dy-
namics of Abies magnifica forests in the southem
Cascade Range, USA l. Veg. Sci. 2: 189-200
Taylor. A.H., and e.N. Skinner (in review). Fire history and
landscape dynamics in a Douglas-fir latc-successional
reserve. Klamath Mountains, California, USA. For
Ecol. Manage. xx: yy-zz.
Turner, M.G., and W.H. Romme. 1994. Landscape dynamics
in crown fire ecosystems. Landscape Ecol. 9: 59-77.
Turner. MG. R.H. Gardner, V.H. Dale. and R.V. O'Nelll.
19S1) Predicting the spread of disturbance across het.
erogeneous landscapes. Oikas 55: 121-129.
van Wagtcndonk, J.W. 1985. Fire suppression effects Oil fu-
els and succession in short-fire-irHerval wilderness
ecosystems In: Lotan, J.E., B.M. Kdgore. W.e.
Fischer. and R.W. Mutch (Tech. CoO{ds.) Proceed-
Ings-Symposium and workshop on wilderness fire
USDA [<or. Servo Gen. Te<:h. Rep. INT-182. [ntennoun-
t;llll Research Station, Ogden, Utah. pp. 119-126
Wallin, D A. Fl. Swanson. B. Marks,J.H. Cissel, and J. Kenis.
I <)96. Comparison of managed and pre-settlement land-
scape dynamics in forests of the Pacific Northwest,
USA. For. Ecol. and Manage. 85: 291-305.
Weaver. H 1943. Fire as an ecological and silvicultural fac-
tor in the ponderosa pine region of the Pacific slope.
1. For. 41: 7-15.
West, NE. 1969. Tree patterns in central Oregon ponderosa
pine forests. Amer. MidI. Nat. 81: 584-590.
White, A.S. 1985. Presettlement regeneration patterns in a
southwestern ponderosa pine stand. Ecol. 66: 589-594.
WhIte, P.S. 1987. Natural disturbance, patch dynamics, and
landscape pattern in natural areas. Nat. Areas J. 7: 14-
22.
White, P S, and S.T. Pickett. 1985. Natural disturbance and
patch dynamics: An introduction. Chapter I In: Pickett,
ST..'\., and 1'.5. White, eds. The Ecology of Natural
Disturbance and Patch Dynamics. Academic Press,
!'\ew York.
Wnght, CS. 1996. Fire history of the Tcanaway River drain-
age, Washington. University of Washington, Seattle.
M.S. thesis.
Yahner, R.H. 1988. Changes in wildlife communities near
edges. Cons. BioI. 2: 333-339.
,.
....
James K. Brown
Chapter 9:
Ecological Principles,
Shifting Fire Regimes and
l\IIanagement Considerations
This chapter presents a broader, more fundamental
view of the ecological principles and shifting fire re-
gimes described in the previous chapters that have
important implications for ecosystem management.
Also included are strategies and approaches for man-
aging fire in an ecosystem management context and
sources of technical knowledge that can assist in this
process. Research needs are also described. The eco-
logical fundamentals that underlie the effects of fire
on flora and fuels can be described under four broad
principles:
1. Fire will occur with irregular pattern depending
on climate
2. Diversity of species and vegetation pattern de-
pends on fire diversity
3. Fire initiates and influences ecological processes
such as regeneration, growth and mortality, de-
composition, nutrient fluxes, hydrology, and wild-
life acti\ity
Humans exert a commanding influence on eco-
systems by 19niting and suppressing fire
JSDA Fores! Service Ger' Tech Rep RMRS.GTR.42.vol.2 2000
Ecological Principles
Fire Recurrence
Fire as a disturbance process on wildlands has
occurred as long as vegetation has been present on
earth. The history of fire can be traced through char-
coal fragments back to the Paleozoic Era, several
hli.ndred million years ago (Agee 1993). Lightningthat
can start fires Occurs at a mind boggling rate. Approxi-
mately 8 million strikes per day occur globally (Pyne
1982). Human ignitions were common historically and
continue to be common today. Wildland fires will
continue to happen; the important questions about
fire occurrence are when, where, and of what severity?
The frequency of historical fire varied widely across
North America depending on climate. Fire return
intervals typically ranged from 2 to 5 years in ecosys-
tems supporting abundant cured or dead fine fuels
such as the Southern pines, Southwestern ponderosa
pine, and oak savanna. They ranged from 5 to 35 years
185
Brown
for dry site conifers, shrublands including California
chaparral, and most grasslands; 35 to 200 years for
mesic site Western and Northern conifers; 200 to 500
years for some Eastern hardwoods and wetter site
conifers; and 500 to 1,000 years for extremely cold or
wet ecosystems such as alpine tundra and Northwest-
ern coastal spruce-hemlock forests.
Our knowledge of fire frequency is largely based on
tree ring analyses and postfire stand ages, which only
allow a glimpse of fire history over the past several
hundred years-a rather short climatic period. None-
theless, it provides a basis for understanding the
recurrence offire that can be useful in planning. Keep
in mind that climate could indeed change and in turn
influence the occurrence of fire and the nature of
vegetation response.
Historically, fires have occuned at irregular inter-
vals, largely determined by climate. Dendroclim-
atological studies in western Canada (Johnson and
Larsen 1991) and the United States (Swetnam 1993)
have shown that climatic cycles within cycles some-
times influence fire frequency. For example, in giant
sequoia forests, precipitation was the most important
influence on fire OCCU1Tence over periods of years such
as the recurrent episodes ofthe climatic phenomena EI
Nino and La Nina (Swetnam and Betancourt 1990).
However, temperature was the most important influ-
ence on fire frequency over periods of decades to
centuries. In both cases fuel moisture content was
probably the important fuel property most influenced
by climatic trends in precipitation and temperature. A
study of presettlement fire frequency regimes of the
United States (Frost 1998) suggests that patterns of
fire recurrence, termed "fire periodicity," can be con-
sidered as regular or irregular. For fire regimes hav-
ing high fire frequencies (average fire-return intervals
of 0 to 10 years), individual fire occurrences were
considered nonrandom because they clustered around
a mean fire frequency. For fire frequencies greater
than 10 years, individual fires occurred irregularly or
in a random pattern.
Biodiversity
Biodiversity is broadly defined as the variety oflife
and associated ecological processes that occur in an
area. This variety is sometimes broken down into
genetic, species, and ecosystem components (Salwasser
1990). In dealing with vegetation, it is convenient to
think of the spectrum of components as being plant,
community, and landscape. The landscape can be
viewed as a mosaic of patches , which are plant commu-
nities typically descri.bed as vegetation types, succes-
sional stages, stands, and age classes.
Fire regime types influence biodiversity in various
ways (Duchesne 1994). In forest ecosystems, under-
story fire regimes have the b'Teatest influence on
, o~~
Chapter ~ EcoloQica.l PrinCiples. Shdtlng Fire Re~lmes ar:d i\1anagernenf CcnS,{~erdtlon5
f
~
f
biodiversity Within plant communities because the
understory vegetation is more affected by fire than the
overstory. Stand-replacement fire regimes substan-
tially influence biodiversity across the landscape by
affecting the size, shape, and distribution of patches.
Mixed fire regimes probably have the most influence
on biodiversity within plant communities, but also
affect patch characteristics or between community
diversity. In grassland ecosystems, fire frequency and
seasonal timing largely determine biodiversity.
Biodiversity can be increased by fire in many ecosys-
tems and reduced by eliminating fire (Keane and
others, in press). V ariabili ty offire regimes in time and
space creates the most diverse complexes of species.
Thus, landscapes having fires with high variability in
timing, intensity, pattern, and frequency tend to have
the greatest diversity in ecosystem components
(Swanson and others 1990). The phrase "pyrodiversity
promotes biodiversity" coined by Martin and Sapsis
(1992) aptly summarizes this concept. However, biodi-
versity can be reduced when fires occur much more
frequently than happened under the historical fire
regime. An understanding of the underlying relation-
ships provides a basis for managing fire to meet
conservation of biodiversity goals.
;.
~'
J
,Ii
i
'i
{
~
j
~
~
~
f
),.
I:;
Plant Response to Fire
t
!
\'
Chapter 2 explained the many adaptive traits that
allow plant species to survive fire. In fact, many
species depend on fire to continue their existence.
Traits such as thick bark, fire resistant foliage, and
adventitious buds allow plants to survive low to mod-
erate intensity fires of relatively short duration. Traits
such as fire stimulated germination, belowground
sprouting parts, and serotinous cones allow plants to
reproduce following high severity fires. For any par-
ticular plant to survive and persist, its adaptive traits
must be compatible with characteristics afthe fire and
the timing of its occurrence. Fires can vary in inten-
sity, duration, severity, seasonal timing, and frequency.
Other factors, especially weather and animal impacts,
can greatly affect whether a species can reproduce and
continue its existence following fire. Grazing by ungu-
lates can influence postfire successional patterns and
flammability of future fires (Smith 2000).
Fire severity and intensity have a large influence on
composition and structure of the initial plant commu-
nity following fire. Fire intensity mostly influences
survival of aboveground vegetation. Fire severity ac-
counts for both upward and downward heat fluxes;
thus, it is a better indicator of initial postfire flora and i
other fire effects. For example, when moisture con- ~
tents of the forest floor fuels are high, a surface fire ~
may burn at high intensity yet not damage sprouting - ~
tissues in the duff layer and mineral soil. Conversely, \
under low forest floor moisture contents, a surface fire
"Ii
~.
J
t
5-
'f
t
~
~
i
y
'1
.~
~
~
1
....
rt
i
f
"
I
~
I
I Ic:n^ r:- \(1''--.' ~(>r-!1f-;\ (':.-Ql
nnc' nrMl~"Tn..17.vr.1 .) 2000
l
T,wf~
\.....-: I ,1 ~. t~: :!
~)':i'~,d l-rlrll.;t)I~::;', ~)j "I", ':1 .
..
may burn at low to moderate Intensity yet consume
the forest floor and damage Illany sprouting tissues
As a gcnf~ral rule, burned areas tend to rdurn to the
same flora that was there before fire (Christensen
1985; Lyon and Stickney 1976) However, fin~s ofhlgh
severity create opportunities for new plants to estab-
lish from offsite seed. Large, high severity burns can
be slow to recover depending on available seed sources.
Fires oflaw severity are followed by a strong sprouting
response except where annuals are the dominant
vegeta tion.
The timing of fire including both seasonality and
frequency is crucial to managing for conservation of
biodiversity. This aspect of fire management can be
easily overlooked because of emphasis on controlling
fire and meeting air quality constraints. Seasonal
timing of fire is important because it largely deter-
mines fire severity and related mortality. It particu-
larly affects reproduction of herbaceous plants and
shrubs. For example, in some ecosystems spring and
summer fire may produce abundant postfire flowering
while late summer and fall fires may produce little.
Perennials in Texas survive spring fire, but annuals
arc harmed if fire occurs before seed is produced
(Chandler and others 1983). Evidence suggests that to
maintain long-term (decades) diversity in a tall grass
ecosystem, fire should be applied at different times of
the year to achieve successful seedling establishment
and productivity for a variety of plants (Bragg 19911
Fire frl'quency is a particularly important consider-
ation in short fire return..interval regime types be-
ca use a period of several years to perhaps a decade can
be critical for survival of some species. Frequent fire
regimes that allow control of shrubs are critical to
malOtaining grassland ecosystems (Wright and Bailey
1982). Many rare and threatened species have de-
clined with reduction of fire frequency (see Greenlee
1997) Some fire dependent species in the Southeast-
ern United States seem to require a 1 to 3 year fire
return-interval (Frost 1995). In contrast, local species
extlOctions can occur with fires that occur too fre-
quently, although it is generally accepted that locally
rare plants have greater chances of surviving on land-
scapes having diverse vegetation communities and
structure created by diverse disturbance histories
(Gill and Bradstock 1995). A problem today is that
plants adapted to short fire return-intervals can be
ham1ed by fires burning with high intensity and
sc\'erity in accumulated fuels that resulted from long
firc-free periods (Sheppard and Farnsworth 1997).
fl,
t
\'
,I
J'
~
~
,I
.>
Community and Landscape Responses to Fire
Species diversity within a vegetation community
:;ueh as ;\ stand or a patch depends on the collection of
Spl'elL'S In the community, their adaptivL' traiLs, the
USDA FOC\!'it Service Gen Toch Rep RMRSGTR.42.vol 2 2000
timing of fire, and the nature of fire as it moves
through the community. The spatial arrangement of
fuels and individual plants can be important to sur.
vival, particularly where fuels are unevenly distrib-
uted Variable fire weather can also inQuence :iIU-
vival. Concentrations oflive or dead fuels can generate
much greater fire intensities and severities on rl:/a-
tively small sites. This could enhance or reduce (l1ver-
sity depending on the community. For example, in a
Douglas-fir forest, localized fuel concentrations may
result in fire-created gaps or holes in the canopy This
would create structural diversity and stimulate un-
derstory vegetation, a typical response to fire ] n a
mixed fire regime (fig. 9-1). However, in a ponderosa
pine forest, excessive mortality to highly valued old
growth trees could be a consequence.
Ecosystems and plant communities are considered
to be fire dependent when their continued existence
depends on recurrent fire. Where fires occur regularly
and frequently, such as in African savannas. open pine
communities, and Mediterranean shrublands, t.hey
may remain stable for millennia (Chandler and others
1983). Repeated fires in fire-dependent commul1lties
maintain a dynamic process that creates diversity
across the landscape, but if fire is excluded, biodiver-
sity would probably diminish (Chang 1996). It has
been argued that fire-dependent communities have
evolved flammable characteristics that help ensure
repeated fires and the cycle of renewal (Mutch 1970).
However, the evolutionary argument remains un-
settled (Chang 1996, Christensen 1993b).
Stand-replacement and to some extent mixed re-
gime fires create patches on the landscape of differing
dominant vegetation and stand structures (fig. 9-2).
Patches can vary greatly in size and shape depending
on the biophysical features of the landscape and fire
behavior. Winds of variable speed and direction can
cause fire behavior to create a variety of fire shapes.
Terrain and landforms, rather than other fire influ-
ences, primarily determine patch dynamics in heavily
dissected landscapes (Keane and others, in press) For
example, fires in the nonmountainous boreal forests
were typically large (often well over 10,000 acres) but
medi urn to large (100 to 10,000 acres) in conifer forests
of western mountains (Heinselman 1981). Even in
large fires in mountainous terrain, fire seventy can
vary considerably within the burn, leaving a patchy
distribution of fire effects (Turner and Romme 1994).
Generally, on landscapes characterized by large stand-
replacement fires, the pattern is naturally coarse
grained. On landscapes supporting smaller stand-
replacement fires, the pattern is finer grained On
landscapes having understory fire regimes, occasional
trees are killed, creating gaps. This leaves a fine
grained pattern in the o\'erstory such that the notion
of patches is not as helpful for describing landscape
187
"'--"~----...-.oor."....-.-,.,.,
,~I'.l~\:l~r 'I !-nC;ll(l(j:1.]! P"ln(~lplt~:" ,~llIfll('l;J Fire Rf:glmes <.111,j \lJ'lJ~';',~,,:
1',.'';''".1
Figure 9-1-A mixed severity fire burned through this Douglas-fir stand in Yellowstone National Park killing
about half of the trees leaving gaps and large openings in the canopy.
Figure 9-2-Stand-replacement fire sustained dUring low wind speeds by burning In heavy accumulallons of
dead surface fuels, Yellowstone National Park
188
tl~nA r:-",'.,<'t C'",
" ("1
>lap:cr:) L~c,joglcal PrincIples. S~H'tlf:g t~irc neglmes Jf\J ~./.1nagement Considerations
diversity. In these tire regimes, considerable struc-
tural diversity can exist within communities.
As time since last lire increases, succession ad-
V;1nces all stands to similar communities gradually
: ucing structural diversity (Keane and others, in
press). Extending fire-free periods also increases the
likelihood oflarger fires, hence larger patches and less
patch diversity (Bonnickson and Stone 1982;
Hcinselman 1981; Swetnam 1993). In whitebark pine
forests, Murray (1996) found t.hat lack of fire created
high elevation landscaIJes with high mean patch size
and low diversity. Romme (1982) found that fire con-
trol policies tended to reduce landscape richness and
patchiness and increase evenness in Yellowstone Na-
tional Park, although in some situations, exclusion of
fire actually increased landscape diversity. Knowl-
edge offire regimes can help managers choose alterna-
tive land practices involving fire that favor landscape
dIversity compatible with flatural ecosystems.
Ecological Processes
Fire is an ecological process that triggers an amaz-
ing network of other processes and associated condi-
tions. To explain this network, it can be helpful to
cat.egorize fire effects into first and second orders.
First order effects are the immediate actions of fire
and include plant mortality, consumption of organic
material, creation of smoke, and changes to the physi-
cal-chemical environment. Second order effects are
many and depend on the nature of first order effects
and the postfire environment, especially soil, weather,
and animal activity. For example, here is a partial list
of second order effects:
1. Change in microclimate
2. Increase in range of soil temperatures
3. Change in soil nutrients and microbial activity
4. Regeneration of vegetation
5. Succession and new vegetation patterns
6. Change in plant growth rates and competitive
in teractions
7. Altered wildlife habitat and activity of inverte-
brates and vertebrates
8. Changed water storage capacity and pattern of
runoff
Plant mortality, regeneration, imd growth are fire
effects of obvious importance to land managers be-
cause they determine the characteristics of flora and
fuel that are readily observable as succession pro-
ceeds. Less apparent but nonetheless important, espe-
cially to the pattern of fuel change, is the decomposi-
tion process that invulves fire, inserts, and pathogens
m vary"ng roles.
-'-. ,- r-' rl t . ~ 'r r: Tn"
Hrown
Successional Pathways
The classical concept of succession was based on the
perception that plant communities evolved over time
toward a final climax state that remained stable in-
definitely. However, modern ecologists have rejected
this concept and now view succession as a dynamic
process that can move in alternative directions under
the influence of periodic disturbance and never reach
a stable end point (Christensen 1988). A useful method
of portraying succession utilizes the multiple path-
ways approach (Connell and Slayter 1977; Kessell and
Fischer 1981) where successional classes or stages are
linked along pathways converging to one or several
somewhat st.able late-successional community t.ypes.
Successional classes are described by vegetat.ion t.ype
and structural stage. The number of succession classes,
pathways, and time steps between classes can vary
depending upon knowledge and the application This
approach allows fire of varying severity and other
disturbances such as grazing and silvicultural cut-
tings to be incorporated in the conceptualization of
successional processes.
Time is a key element in understanding succession
(Wright and Heinselman 1973) and explaining it to
others. Some plant communities such as mesic and
wet site grasslands regain their former composition
and structure within only 1 or 2 years after distur-
bance (fig. 9-3). For other ecosystems, some composi-
tional change may continue to occur long into the
future. Forest and shrubland communities vary greatly
in the time necessary for recovery to a mature condi-
tion. In understory fire regimes, vegetation usually
recovers rapidly. Structural changes are small or fine-
grained and may not be readily apparent. In stand-
replacement fire regimes, a youug forest condition
may appear within 20 or so years. But it could take
several times longer in large severe burns where tree
seed sources are limited.
Decomposition
Fire, insects, and pathogens are responsible for the
decomposition of dead organic matter and the recy-
cling of nutrients (Olson 1963; Stoszek 1988). Fire
directly recycles the carbon ofliving and dead vegeta-
tion. The relative importance of fire and biological
decomposition depends on site and climate (Harvey
1994). In cold or dry environments biological decay is
limited, which allows accumulation of plant debris.
Fire plays a major role in recycling organic matter in
these environments. Without fire in these ecosystems,
nutrients are tied up in dead woody vegetation. In
forests, tree denSIty and understories thicken causing
increased competition alld moisture stress. In turn,
i0(l
,.. , ..-'.._....-,.___.._"._._...__.-1-".._.....
Ijr~','",)
L~:1j;'::~~,' J ~l:I)I,j(JIl~(ll;"-~ ",~::plt~"
;"':,rp f-h~q'fT''7~ .l"C ,\!.Ir'd.-,;,"'nlf;i" _' ,'r':)nJ~'t'.I
Figure9-3~One year after a prescribed fire in a mountain big sagebrush community, this mesic site recovered
to domination by perennial grasses and forbs, Caribou National Forest, Idaho.
this incl'!cc1.ses the likelihood of mortality from insects
and diseases leading to increased dead fuels, higher
intensity fires, and possiLJly volatilization of more
nutrients In f,'Tassland ecosystems where both fire
and grazing ilre excluded, thatch or dead herbaceous
litter accumulates, which depresses herbage yields
and the number of plant species (Wright and Bailey
1982), Fire can help control encroaching shrubs and
trees; increase herbage yield, utilization of coarse
grasses, and availability of forage: and impruve habi-
tat for "Uille wildlife species.
Fire both creates and consumes fueL It increases
available fuel by killing shrubs and trees, which leads
to falldown of dead material into the surface fuel com-
plex Moisture contents of dead fuels averaae much
lower than live fuels. which also increases fL:=-el avail-
ability. Insects and diseases perform similar role:::.
They both kill vegetation. wlllch creates available fuel.
and decolllpo,.;e organic matttT Fin: in somo circum-
stances enhances the opportunity for insect and disease
attack For example, bark beetles lllay overwhelm tire-
injured (tmifers, and wood rotting organisms mav in-
vade fire-'icarred deciduous trees ^ complex int~rac-
tlOn th;lt_ t:-; not well understood exists between insech
and di,ot';ht' organi,;rns, fin-, and the envinHllllent
I [O\\,l;vl'r. we do know that firc. ill'iCCts, and palhogl:n'
evtJ!vl:d t')~cd1<'r as vital COi1ljHIIlCnb of I'CO:--;y:-;tt'llh
'Cln
Fuel Accumulation
Fuel accumulation is a term often used 100se1': to
indicate an increasing potentlal for fire to start, spr~;ad,
and intensify as the time since the last fire increases.
Generally, in ecosystems where annual biomass in\:re-
ment exceeds decay, total vegetative biomass increases
steadily with time because photosynthesis is an ongo-
ing process. Fuels accumulate but not necessarily III a
st.eady fashion (Brown 1985a), On forested si tes much
of tile dnnual biomass increment is tied up in live tree
boles where it is unavailable for combustion. In b'T:l:-;S-
lands and forests having short fire intervals, fuels
increase regularly over time as biomass increases.
However, in medium to long fire interval conifer for-
ests, available fuel. ane! fire potential may decrease as
a postfire stand develops. then increase"as the stand
becomes old and overmature i, Brown and See 1981).
Fuel accumulation and associated fire potential de-
pend on fuel quantity as well il:-; other important !'ud
properties such as compactness and continuity (vuti-
cal and horizontal). To be u:.eful for estimating fire
behavior, fuel quantity nllht be expressed by Slze
classes for live and dl:ad component:;. In a given
vcgdalion type, fuel quantIty, SII.I; distribution. dl'ad-
to-lIVI' ratlo, and continlllLy ;'IT Uw !fllpO['tant prupcr-
til~S tlLlt Ch;lllgC as .'iUCl't'SSI(H] progresse:-i. Gellerall,v.
I Il~rl:\ [" , ,;1 c:..-H', I '(l (""fl 1., .\ :;., ;~UP'~ (~IP 1:1 lid ,J
! ; ) 1 ~"
flll,1 qii,i:,l:t It''; illTUillUI,lll' ((, "Ct'dt<-C It:\'('!,; Oil t.lll'
more pr<)'lul'tl\'(' ,;de,; Irl gTd,,~Ll:1d, "hculJl;\fld, and
f(lrt'st I'Clh\''': em,; ,llrown ,lnd ~"I' I ~l01. \Vnghl and
Badey 10:,::;2) In focest l'coc;y,;tl'Il'S much of thl' dl'dd
fur,1 l'XI~I~ i1S end)""e wood\' dt'l\ct.-;. whll'h includl'"
pli'ce,; Lli'L:t'r than:J inchl''' III dlilllwkr ;Ind S()Il1('t Illl('S
Ltrger t heW I inch diamd('r (I iarll1011 and OdH'LC;
1~}S6) The more prnductivC' Sltt'S gTlJ\\' larger tnil;S.
which l'\entually become coarse \\'lcHly dclJl'lS An
ltl1pOrL:llt (OllSlderation in management u('telllpl:rak
l'COSYStl'lllo' I'; that. coarSE; woody debris be recognized
f(lr thi' m.IllY roles It play::;, It illlltrilJutC's to biodl\'er-
"ity by b"lIlg parlofthi: life cycle ofm;),cruinvertebrates.
,;ud mite:" in,;ecb, reptiles, amphibians, blnb, and
lllalllmal- < :\ldvllIln and Cro,;,;le)' 1996} It I'; a ~,uurCl'
of nLltrj(':lt~, habitat for te:-restnrd a:ld aquatl( Ilk,
and fuel I"lr wildhre (Harmon ancl others 19861, As a
fuel itc' 1ll0~t significant featur'c IS that it bCLllmt'~
rotten \, ood, \\hich prolongs burnuut ancl allows tire to
persl::;t l):\ ~ltc for iong periods, IIt,;toric,dly, brge tire,;
occurred because fire remailled smoldenng in rotten
wood and duff for extended periods until low fuc:i
rnoisture~ combined with high wllld speeds to support
Intense. [',1St spreading fires
Flaml\\abluty increases as dead-to-li vc ratios increase,
.-\S fuel" accumulate through gruwth and mort::llity of
plants. t1arnmability threshulds may be rcached that
allow fi:T51.o mcrease greatly in Inten,;it)', Surface lirt's
become lTOI\n fires in conifer forests, and shrub com-
::1Llnitit':' burn intensely as a 5lngle fuel complex
Fuel l';:ntinuity is important because it panh' con"
trols wne,e a fire can go and how fast it tra\'c!s In
gra';:3larld:, and open shrub13nds, he~1\'ily gTazed areas
and art <is oClow productivity form discontin uou" fuel,;
that ltn\lr 5pread offirc, v,:hichcan be a cnticed obstacle
to use of ilrescribed fire, In forests, eXistence of irtdder
fuels from understory veg'etation allov;s surface fire,;
to reach Into the crown canopy Iftlw canopy is mo,;tly
dosed. Cf'l\ln fire can readily de\'elop und!'r adequate
\llnd cpecd~, Open canopi\'s do IIPt ,;upport Gown
!ires lri(n'a~ed fuelcontinuit,\'can account fur changes
In fire :'i;\enty from understory to ll11xed and from
mixed" "Lind,replacement :-1<lf11' tlptions are avail-
ilblc tl, land managers fur altcnng fLll'! COlltlnuity
~hrougl; manipulation of vegetation
Efrt'(t~ ,)t'tire on fud arIse basically t\l'O ways flrst,
t'l'dUi'lrli-; fud through conoiu::\ption, and s('cond, in-
lTl'a';i1Ih fuel by killing vegetatiun, Bptl1 pru(esses
affl'ct"f'\'eral properties of fuel and fire putcntial,
Inrtl 111.. ell'rie! surface fud loadIng,; are n~dul'\:d, also
IO\',l'J'l1 ;~ I!\l' dcad-to-ltve r,ltl(J [I',;ub,;tantlal amounts
of,dll'l~ '. ,;n;al1 conifers, and limbs rWe! follagl' uflarger
lOIllfl'l, an' killed by fire but [elit CllllSUll\('(l. they will
(()[lll'I:,'Jl\' tLl ,;urface fueL; In the \'l',lrS alll'.ld a,; thev
.ICl'Lilll .:"l\' (In tlH' ground Fin' gn'at Iv l!dlu\'llll''; fue'[
CUIlt IIl\Jlt\ b\' creating \ ertlct! and hOrI/,ll1;l.r1 hap,;
\\lIllil, ,tilt! lwtwl'en ,;url',ll(' !Ut'l,; ,111d l'l'(I\\ll flJf'l.~,
l J ~-~~),', ;
~,_,rvIC~:' (;Cll rt'r:~\ f~;,.: qk1H~-)-(~ rf-~ .:,)
Accumulation in FO"L'sts,-LI\'(' aIld dc:\(i fUl'\:-., ,h
1\ (,11 il> :-1:1<111 and large dlrunc(cr fuels, ,an l,dlol\
drf'f('['('lll p,lt!(:rns of'accullll!!atioIl T,\i!lICally, Ii,'" lli'r
!li\CC'OU:-; ,lIld shrub fup!s increase f'olluwll1g fire dlll-IIlg
(';\rI", ,;LI)"l~S of st.and dpvelupnll'nt. Then as lJ'(~(' i'dIl')-
pi"." clll~~C. lIve herbacl'ow; and shrub fuel qual:tltil',o'
t ('nel t II dCL'J'L'a,;c on mesic sites ( Habeck] 9-;- tj. LYOIl d I:d
StlcklH'Y ]~)i61 Howpver, a decrp;1se in biuI1la,;,; [[\,\\
[lot occur where undc'rstories cOIltaln shadl' toler;ll\t
specics, Fine dead fuels from foliage, bark !1akes, lwig",
a nd cured herbaceous vegetation become incorporated
III the ftlrl:oit noor. Once crown canopies do,;c, the
,UllOUllt of litter fuel remains fairly consta.nt a,; Ile\\\'
Ltllen litter IS otfset by older litter nlllving into the dufl
1r1)'I' 1', Dufl quantities continue to increase for sun\!'
tillle until equilibrium with decay is rcached Thl';
period vanes widely from approximately 5 year,; in
Southe;1sti'rrl united States (I\1cNab and others 19-;-8)
11) wel! 0\(' l' a hundred years in some boreal ecosystem,.;
Dead bra.nches and tree boles accumulate on the
ground In response to natural mortality and factors
cau,;ing downfall (Brown 1975). Mortality factors "uch
as fire, insects, disease, canopy suppression. and wind
and sno\'; damage impact stands in a rather haphaz-
ard manner, Thus, accumulation of downed dead fuel
often occurs in an irregular pattern that is correlated
poorly \\ i t h stand age (Brown and See 1951).
Conifer crown fuels increase regularly; however,
likeliho()d uf crown fire may increase then decrease as
the lowcr canopy level grows further above surface
fuels, E\'entually, crown fire potential increases again
when ,;urface fuels increase and understory conifers
becomc ladder fuels. Shade tolerant species tend to
have more foliar biomass than intolerant species due
to their longer needle retention and higher crown
den:'ltiE':' ,Brown 1978; Keane and others 1999), Be-
cau,;!' of rheir shade tolerance they can ti.1l in lTo\\'n
canop\ g:1P:; and develop into understory ladder fueb
FueLs cntIcal to fire spread differ considerably be-
tween "hurt and long fire interval fire regime types
(Brown 1985a), In short fire interval forests, fine fuelS
such ~i" iCrass, live shrubs, and needies create flam-
mable understory fuels eV"en in forests with \'astly
chffel'l'n, decomposition rates such as in longleaf pint;
and poncierosa pine, The substantial quantity of rl[1{~
fuels cllupled with long periods of suitable burning
ctJIlllltflln:; largely account for the uncler~tory Clfe re-
gime, In long fire interval forests the forest He,or and
accumul.ltPd coarse woody deuris are criticai fuels
They burn \\'ith con"iderable heat release over a rcla,
ti\'(~ly [L1r,g duration resulting in exten:,i\'l' mortalIty to
\)\'erstl1r1 trees, They ignite other ,;urfacl' and aerIal
fUl'I,; <lllc1 ::,('rV'e as excellent receptor~ of :'pottini' cm-
I)!'r,; Ih,it llt'ten allow lire to ll\O\'e in a leap frog bshion
FIre lllcc',"\a!:i and envirLlnmcn!S ciltler cunsuj\-'rablv
bet WClT iong fire intervrd t\pes sLlch as cedar-hem-
IllCf.: f<'1. 'I" l)[! warm 1ll0l::'t SI[(''; and suhalpir1l' and
191
GfOWI'
boreal forests on cold, dry sites. Nevertheless. in both
cases accumulated forest floor and downed woody
fuels support stand-replacement fire particularly dur-
IIlg extended dry periods (Romme and Despain 1989)
Accumulation in Shrublands and Grasslands.....--
On many grasslands, grazing eliminates lllO:-it of the
annual pn>duction so fuel accumulation IS inconse-
quentiaL In the absence of grazing, fuel quantities
depend primarily on annual production, which varies
substantially by site potential and annual precipita-
tion (Wright and Bailey 1982). Fuel loading may
increase for several years after a fire as some slow
responding grassland communities recover. Fre.
quently, however, productivity is increased within 10r
2 years following fire (Wright and Bailey 1982). Her-
baceous litter accumulates in some grassland ecosys-
tems but only marginally in others. Ratios of accumu-
lated litter-to-current production typically range from
0_25 to 0.50 (Reinhardt and others 1997).
In shrub and shrub/grass ecosystems young commu-
nities generally have a low dead-to-live ratio. Flam-
mability depends largely on grass and sedge fuels. As
shrubs become senescent or undergo mortality, dead
stem wood accumulates, which significantly increases
potential flammability. Dead fuel quantities tend to
increase with time since last fire or with age of plant
community as suggested for chapan'al, however, not
in a uniform nor readily predictable fashion (Paysen
and Cohen 1996). Besides age, other factors such as
drought, winter kill, insects, and disease can cause
periodic dieback that creates substantial dead fuel
quantities. As cover and height of shrubs such as
sagebrush increase, fire intensity and rate of spread
potential increase markedly (Brown 1982).
Human Influences
People are part of ecosystems and certainly have
exerted a major, far reaching infl uence on fire across
the landscape_ Indian burning was common through-
out the United States and Canada. Pyne (1982) quotes
Henry Lewis as saying, "To simply note that all Indi-
ans used fire to modify their environments is no more
an ecological generalization than to note that all
farmers used plows." The extent of Indian burning
varied considerably, however, depending on locale and
population movements (Boyd 1999; Pyne 1982). In-
dian burning greatly extended grasslands especially
in the Eastern and Midwestern United States. Most of
the coastal plain from Massachusetts to Florida to
Texas was savanna. Western valleys and foothills
were maintained as grasslands and open forests (Gruel!
1985 )
Considerable debate exists about the relative impor-
tance of Native Americans and lightning in maintain-
ing historical fire regimes ( Barrett and Arno 1982;
192
'=:,fldplC'f :_~ [~(~olt)qlcal Pr:(lupII;S, Stldting Fire Regl111es and Mdf1agernenl C,Jn~,:(jerdII(H S
Frost 199B; Keane and others 1999). The relative
importance of Native American fires was probably
greater in topobrraphically complex areas where fire
compartments were smaller and where lighting igni-
tions were infrequent (Frost 1998). Also debated IS
whether anthropogenic burning should be conSidered
part of the native or natural fire regime (Arno 198[);
Kilgore 1985). Fires set by Indians were often of
different seasonality, frequency, and landscape pat.-
tern than those started by lightning (Frost 1998, Kay
1995). Indian and lightning-caused fire existed for
thousands of years, a short evolutionary period but a
long time for plant communities to adjust to fire
disturbance This long period of fire on the landscape
argues strongly for accepting both sources of ignition
in considerations of Euro-American presettlement fire
history used to guide management of ecosystems.
Efforts to suppress fires were modest at first relying
on wet blankets and buckets around dwellings and
campsites (Pyne 1982). Modern suppression capabili-
ties relying on sophisticated communications, rapid
attack, specialized equipment, and many fire fighters
are a far cry from the early 1900s. Fire protection has
succeeded in reducing the extent offire and increasing
fire intervals. Chandler and others (1983) suggested
that as protection succeeds, fire intervals become
greater and flammability increases. Then, more pro-
tection is needed to keep burned acreage down. A gi ven
protection effort and annual burned area will eventu-
ally reach equilibrium. Since the 1980s, the costs of
protection and greater understanding ofthe role of fire
have led to more hazard reduction and ecosysteUl
maintenance rather than just protection.
For the past 100 years or so, human use of fire--
earlier termed controlled burning and now prescribed
fire and wildland fire use--has met with considerable
controversy politically and within land management
organizations. "Light burning" (understory fire) was
once widely applied in the southern pines and ponde
rosa pine type especially in California. However, the
percei ved threat to effective organized fire control
largely curtailed the program on publicly owned lands
(Pyne 1982). Some benefits of controlled burning were
still recognized, especially hazard reduction and prepa-
ration of seed beds for regeneration. In the West
justification for prescribed fire was fuel reduction.
namely slash burning. This single purpose use of
prescribed fire resulted in short-term successes but
long-term failure to optimize societal objectives for
forests (Agee 1993).
More recently, the concept of ecosystem manage-
ment has led to a much wider understanding of the
ecological role of fire and its importance in the func-
tioning of ecosystems. Concerns over air quality, con-
trol of fire, and costs, however, remain as major con-
straints on the application of prescribed fire and
USOA Forest Service Gefl Tech Rep RMRSG Hl.42 vol 2 2000
Cn~':lp\er 9 !~. h.}~i,~,),1 Pr,ncpics, Slid!Ii";J hk i'l-:>Tr;'e<::.: ~"\I'J J\'1d"I.J';,wI:'le'~t c..:onsldcfJ.:lOI!S
wildland fire use. The responsibility to see that fire is
properly managed as a component of the ecosystem is
now greater than ever because land managers have
the power' to delay and exclude fire as well as an
understanding of fire's important ecological role.
Shifting Fire Regimes
Chapters 3 through 7 clearly show that fire regimes
have shifted from what they were historically across
most of the United States and southern Canada. In a
comprehensive assessment of burning in the contigu-
ous United States, Lcenhouts (1998) estimated that
approximately 10 times more area must be burned
than at present to restore historical fire regimes to
non urban and nonagricultural lands. The greatest
departure from historical fire regimes is in the Rocky
Mountains where only a small fracilOn of the pre-1900
annual average fire acreage is being burned today
(Barrett and others 1997). Kilgore and Heinselman
(1990) estimated that the greatest detrimental effects
of fire exclusion were in short interval fire regimes of
the Rocky Mountains. In contrast, in long fire regimes,
the effects offire protection have not had a significant
influence. In the Canadian and Alaskan boreal forest
limited protection due to remoteness has maintained
fire regimes essentially as they were historically.
Extensive grazing by domestic stock that reduces
fuels, and fragmentation by agriculture and human
developments, have also contributed to shifting fire
regimes Lengthened fire return intervals have re-
sulted ill changes of minor to major consequence to
vegetation and fuels by increasing wildfire severity
and decreasing species and structural diversity. A
comparison of historical and current fire regimes in
the Interior Columbia River Basin of about 200 million
acres showed that tires have become more severe on24
percent of the area (Morgan and others 1998) (see fig.
5-1 in chapter 5 of this volume). Fire severity was
unchanged on 61 percent of the area. Fires were less
frequent on 57 percent of the area, unchanged on 33
percent, and more frequent on 10 percent of the land
area. Fire protection, reduced fine fuels from grazing,
decreased fuel continuity from human development,
and in :;ome cases exotic plants arc the most probable
causes (Chang 1996; Keane and others 1999). Further
analyses of changes in fire regimes and condition
classes of vegetation are cUlTently under way for the
United States (Hardy 1999).
Forests and Woodlands
Changes in forest composition and structure due to
shifting fire regimes have been widely documented.
Generally, shade-intolerant species are being replaced
with shade-tolerant species. Stand densities are
USDA Foresl Service Gen Tech Rep. RMRS-GTR.42-vol 2.2000
increasing with development of multiple layer cano-
pies. Outbreaks of insects and occurrence of root dis-
eases appear to be worsening (Stewart 1988). The
greatest impacts have occurred in the understor:.-' tire
regime types typified by ponderosa pine and long-leaf
pine ecosystems (fig. 9-4). Although these two ecosys-
tems experience widely different climates, they share
the same end results of fire exclusion made worse in
some locations by selective harvesting of old growth
trees. Where fire regimes have shifted, growth and
vigor oftrees is reduced, insect and disease mortality
is increased, and understory fuel loadings and conti-
nuity increased so that wildfires tend to be of high
intensity, killing most or all of the overs tory pine,
Diversity of understory herbs and shrubs is decreased.
The loss or depletion of the pyrophytic herb layer is
considered to be one of the unrecognized ecological
catastrophes of landscape history (Frost 1998). The
extent of the problem is greater in ponderosa pine
where relatively little prescribed fire has been ap-
plied. Although prescribed fire is widely applied in the
South it has largely been used only for rough (accumu-
lated understory fuels) reduction during the dormant
season. Thus, lack of seasonal fire diversity in the
southern pine types has limited plant diversity.
In mixed fire regime types such as coastal and
inland Douglas-fir, whitebark pine, red pine, and
pinyon-juniper, the results of fire exclusion have cre-
ated the same problems as found in understory fire
regimes. Mixed fire regimes are experiencing consid-
erably less nonlethal understory fire than in the past
(Brown and others 1994). The mixed fire regime is
shifting toward a stand-replacement fire regime that
favors more shade tolerant species and less landscape
diversity.
In stand-replacement fire regimes, fire intervals
have generally lengthened; however, the effects of this
vary widely depending largely on pre settlement fire
return intervals and accessibility for fire suppression
efforts. For example, in the lodgepole pine/subalpine
fir type, which dominates the Selway-Bitterroot Wil-
derness, presettlement stand-replacement fire was
1.5 times more prevalent than during the recent pe-
riod (Brown and others 1994). The presettlement fire
return-interval was approximately 100 years. In the
same type in Yellowstone National Park, character-
ized by a fire return-interval of about 300 years, the
area burned probably has not differed between pre-
sett.lement and recent periods (Romme and Despain
1989).
The age distribution of marginally commercial and
noncommercial forests such as those in wilderness
areas and parks is shifting to an abundance of older
stands (Brown and Arno 1991). Succession is increas-
ing the shade tolerant component of stands, making
a major species shift likely if tire continues to be
193
....,..-.-.....-....,....... ',- <..".._...,..~,^.~." ,...-.--+.._,.,. ..-.....,
:-=: ~)c)D!t"( !
1,']1, ,J] ~rlf'C ~.:I~'-:; ~..':H,nC; ;"'-.r9 ~,'='; 'pes J'-,l:; "~-1I~,I>
Figure 9-4-.A stand-replacement fire supported by accumulated dead surface fuels and live ladder fuels from dense
understory trees occurred in this understory fire regime type killing the old growth ponderosa pine, Yosemite National Pa:,<.
excluded. In the case of western aspen more than half
of the type has been lost (Bartos 1998), much of it due
to successional replacement by conifers (Bartos and
others 1983). Fire protection policies have resulted in
the fire cycle in aspen shifting from about 100 years to
11,000 years; thus, if this degree of fire exclusion
continues, the loss in biodiversity will be considerable.
Injack pine forests the more shade tolerant balsam fir
is gradually assuming dominance aided by natura!
deterioration and harvesting of jack pine.
Fuel accumulation patterns vary widely in conifer-
ous stand. replacement fire regime types. Mature for-
ests may support abundant or relatively little avail-
able fuel. How(~ver, as fire intervals are allowed to
incre::ise and stands become Over mature, downed
dead woody fuels and live ladder fuels from shade
tolerant understory corufers can be expected to dra-
matically incredse. Th(; result wiIl still be stand-
replacement fire but at higher intensities, which will
tend to propagate larger fires in spite of suppression
efforts. This trend cOljld lrad to fewer but larger fires
burn 109 d uring seven~ fi re weather years. ca using Ies~'
diver.--;rty in patch si;:e and age (Keane and others
1999)
194
Grasslands and Shrublands
Grassland fire regimes have shifted dramatically
from the presettlement period. Many ecologists con-
sider the reduced frequency and extent of fires on
rangelands due to fire protection to be among the most
pervasive influences in the United States by non-
Native Americans (Pieper 1994). The shift to woody
plant domination has been substantial during the past
hundred years. Grazing and possibly climate cha...'1ges
have acted with reduced fire to give a competitive
advantage to woody plant species. Some woody plants
such as honey mesquite hecome resistant to fire, de-
velop fuet discontin uities, and reduce spread offire. In
time, recovery following fire favors shrubs over peren-
nials (Archer 1994). This can alter the composition of
ecosystems to the point that a return to the grassland
type becomes nearly im possible or impractical (Brown
1995>.
Historically, fires were more frequent in Eastern
than in Western grasslands. High productivity of
biomass was m:-tintained in the tallgrass prairie b)i
frequently occurring fire that recycled accumulZlte
thatch. A diverse composition was probably favored b) .
USDA Forest Seem>] Gcn T.]ch Rpio RMRS.GTR.42.vo 2 ~"){]()
. j~ !'
\;lrl:lbk I"c"qu(~nl'\ :\lll! Se;\~"rldllly "I" Lccs {;\bcalll'
,u;c1 (;J1J~OIl 1 (Hit, Ilragg 1 ~I') I! \\,(.,~tcrll l,:r,ls~ldlld,
;lIlfw;\r III hd\'\' :-:"Ill~rallv (':\pl'ri('I\LCd 1"111' 1I';,s ('I'\'-
qll!'nt!\ICruclll:':i;>; \Vright ;llld ILltll~Y 1~)h:2ilrut :->1 ill
t'rei!IJl'rlt Iy enou"h to hold hack 11\\';\0'1011 lIf \\'ood\
p ! ;1 II t s
hr!' ""gllJll'''' il;l\'l~ shifted to ltJ() much fin' III tll..
drier p"rtlOns oft"c' sagebru,;f1-st!'PIJC' cClIsystem that
OCCUPi('''; 0\'('1' 11lU million ;\nes in \Vesterll l-Illu.d
St;l!l':-> Fire frcqucllcy has increased in many al'l;ac;
due to 11:\'aSlon llfcheatgras:-, and n1l'dusahc;1d, intrll-
dUCl'd ;lllrlUals th;1l cun' ('arlv and n'll1dlll f1allllllabic
durin~ ;\ long fin' seaSOll, Increased fire frequl'nC\'
"X1'rls :-'t!'Ong "elective presO'ure against man.\' n;lti\'(,
plants l\c'anc ;cu,d pthers 19;19), A contrasting situa-
t Ion 1'\ I:-'ts for the more mesic nllluntain big sagcbru:,h
type \\:lI're dl'crcased fire frequency and encroach-
ment IJ\' c'onlfers IS causing a rcductlun in hcrtJ2ceolls
and s h I'U b vegeta t Ion (fig, 9-.'5 i,
Managing Fire ___________
FIn' 1";.1n inteb'T:.11 component ofecosy,'~tem:, tllatcdn
aff'c.ct ~dl ,\Speers of ecosystem management.. Fif(~ rc-
gi 111(':" h :1\'1' sh I fled as a fl'SU I t of human in f1 uences and
::1:.1\ C:ilHinuc to shift with clearly detrirncntal result,;;
iu SO[lH' ecos\'stems Land managers need to know IlO\\'
((\ phIl ;\11(1 carrv out fire mandgl'llll'Ilt ;;rrdtq,;l('~ lh;11
SII(TI'Sshllh' i,1cllrporaU' the (,cllloglcal 1'01(' uf' fin'
(;I1llstrdlnl.-; I1n managing prescrtlJl:d fire dllll :;Illok.,
IILI!.:.(' Ii dif'ficult to achieVl' resource gOdls. 1\111'" l)l'o-
IL'cll(lIl ag:lInst wildland fires allO\\'s de\'clup!lll'llt of'
UI1 dl's 1I';!ide ec"log-Ic;d conspq ucnces ( Brow 11 dnl! ;\1' [\(I
19~)]) thercollling tillS predicament rl'CjUln"; t h;tl
land Illanagers and the public alike recog-nlle tit!' nil,'
Of'fIIT in the functioning ofecosyste111s and in 111"1'(111;.':
vdrwd resource objectives.
Strategies and Approaches
Vegetatiun and fire management objectl\!"s should
be derived from broader ecosystem management goals
to achie\'e desirable fire effects. Determining Objl'C-
tives, and the stratq.,ries and approaches for achieving
them, can be simple to complex depending on bnd
ownership and direction provided by the owners, For
example, ;; smelll woodlot owner may simply \\~lIlt to
reduce iire hazard, in which case fuel reduction objec-
tive::; can be clearly stated and, if appropriate, a pre-
scribed fire conducted to reduce the unwanted fuel.
\Vhcre tbe dIrection is ecosystem management., a goal
recently adopted on many Federal and some St.at.e
lands (Salwasser 1994), a more elaborate procl'ss m;\y-
be reqUIred to determine objectives and strategies,
Figu;e 9-5-W,trlOut lk;lurlJi.ll\ce. thiS s,19cL',,,shgrass commur,'ty t)8lng encroached by Douglas-fir wili
e ,crltually become a close(J canopy brest \,.,11 sparse understory vegetation. Deerlodge National F ores!,
~"/~~';I:ana.
J C),\ i
iefl rvc/-, H,.; Hr"iH:-~ I.) rH-,L~ \/':)! ~ '.
1 ~J~,
".-+."-..,......-.---,.......,_......"..."_.._-...-~....__.
8r\)W~1
Cr-I.1;Jter:) Ecological PnnClples. S~llttlng Fire RegImes and Mdndgemenl Cons,dera:lons
To steer this process, a guiding princi pic or goal for
l'Cosystcm management is to provide for con.scrvation
of biodiversity and :-iustainabiJity of ecosystem compo-
sition, structure, and processes (Kaufmann and others
1994l This involves molding a r:wnagcmcnt plan
based on an understandingofecosystclll processes. An
clement missing or minimally considered flam IIlany
past planning efforts was the landscape of varying
scales For this a perspective is needed that involves
consideration of ecological processes across a hierar-
chy of land units (Hann and others 1993)
The settingofgoals and objectives starts out broadly
with a goal specifying the future condition of the
ecosystem or a particular tract of land This desired
future condition is a vision for the future and not an
objective for management action (Kaufmann and oth-
ers 1994). An assessment of the ecosystem, resource
potentials, and needs of people is a prerequisite for
setting the desired future condition. From this, more
specific objectives can be derived for managing fire.
They should be specified in terms that can be moni-
tored. Different approaches may be appropriate for
doing an assessment and setting the desired future
condition and the ensuing management objectives.
Consider the planning task by three types of land
use zones (Arno and Brown 1989):
. Zone I - wilderness and natural areas objec-
tives call for allowing fire to play its natural role to
the greatest extent possible. Fire objectives may
vary depending on whether it is a wilderness or
natural area intended to preserve a particular
condition or process.
· Zone II - general forest and range manage-
ment, where the need to provide resource values
means a wide range of vegetation and fire objec-
tives will be appropriate.
· Zone III - residential wildlands, where the
natural role of fire will be constrained consider-
ably and fuel management is the primary objective.
Two occasionally troublesome facets of setting goals
a:1d objectives in Zones I and II that rely on knowledge
about the ecological role of fire involve the "historical
range of variability" and the goal orientation of "pro-
cess versus structure."
Historical Range of Variability
The historical range of variability (also called natu-
ral range of variability) in ecosystem components can
be used to help set desired future conditions and fire
management objectives. It can serve as a basis for
designing disturbance prescriptions at varying spa-
tial scales and help establish reference points for
evaluating ecosystem management (Morgan and oth-
ers 1994). Reference points to past functioning of
ecusystems can be mtupreted from various sources
196
such as historical records, palyno]()~,!', natural areas,
archivalliteraturc and photographs, GIS data layers.
and predictive models (Kaufmann and other" 1994,
Morgan and others 1994). Historical fire reglnlCS 01
forest ecosystems are often characterized by deter-
mining age distribution and areal extent of ser,d
classes across a large landscape and dating fi["l~ scar"
to detennine fire return intervals These techniques
provide a snapshot ofecosys:em conditions that cuvers
the past 100 to 400 years. Pollen analysis can extend
this period but with less precision about disturbance
events (Swanson and others 1993). Estimation of h15-
toric fire frequencies in grasslands and shrub lands is
more problematical because of a lack of fire scars and
easily determined age classes. It relies largely on
historical accounts of human activities.
To what extent should knowledge of the historical
range in variability be relied upon to help establish
goals and objectives? This depends largely on sound-
ness of the ecological knowledge and other ecosystem
issues such as human needs and threatened and
endangered species (Myers 1997), A strong argument
can be made that knowledge of historical fire should be
used as a guide for understanding landscape patterns,
conditions, and dynamics, but not necessarily for cre-
ating historical landscapes. Knowledge of histoncal
variability provides a basis for bringing the range of
existing conditions in a landscape within the histori-
cal range (Swanson and others 1993).
A scientifically based rationale underlies the use 0,
historical variability as a guide for managing biodiver-
sity. Native species evolved and adapted to natural
disturbance events over at least the past 10,000 years.
Numerous ~cological studies emphasize the close de-
pendence of species on disturbance regimes (Swanson
and others 1993). Genetic di versity (Frankel and Soule
1981) as well as landscape diversity are maintained
through disturbance regimes. Where fire regimes have
shifted markedly, species and landscape dl versity
have declined.
Concerns and limitations to using historical vari-
ability as a guide to managing ecosystems (Morgan
and others 1994; Swanson and others 1993) are:
1. Difficulty interpreting past variability due to
insufficient data.
2. Degree to which past and future environmental
conditions may fall outside the established range
of historical conditions. For example, the possi-
bility of future climate change due to global
warming is a significant concern.
3. Extent to which the range of ecosystem condi-
tions desired by society differs from historical
variability.
The natural range of variability can be determin!'
and applied with reasonable confidence in higl,
USDA Forest Service Gen. Tech Rep RMRS-GTR.42.vol 2. 2000
i
C~laplE;r 9 ECOr~)gl.:al P~J"clpies.. S~ll~tlng ~I~e Regimes a'1.j Managemenl Cons,~era::l:.ns
frequency fire regimes of forests In understory fire
regimes, considerable dat.a on (ire frequeIlcy oft.t~n can
be obtained by consulting published accounts or con-
ducting studies offire intervals on fire scarred trees.
Variability of fire-ret.urn intervals can be quantified
and compared with recent lire histoiY to determine
whet.her a significant. departure has occurred (Brown
1993). In long interval stand-replacement fire regimes
of some forests and tundra, estimates of the histOlical
range of variability are more difficult to establish with
certainty because of the limited r.urnber of distur-
bance events that can be studied. Perhaps the best
technique for measuring fire regime characteristics in
this situation utilizes satellite a.nd GIS technologies to
map vegetation pattern (Morgan a!1d others 1994), an
approach requiring considerable resources.
A question that often arises in inte;:preting fire
history especially concerning wilderness and other
nat.ural areas is how Indian ignitions should be treated
(see Lotan and others 1985). The prevailing thought
seems to be that because Indian burni!1g occurred over
a long period, ecosystems were adjusted to fire effects
from human and lightning ignitions combined and
this reflects historical fire regimes. Disturbance his-
tory can only be readily and reliably measured for the
past 200 to 400 years. Variability in climate, vegeta-
tion composition, and disturbance patterns has been
substantially great.er over the past several t.housand
,'ears than over just the last 400 years. But land
managers need a consistent basis on which to plan,
and using measurable tire history is a practical ap-
proach. The concept of the historical range ofvariabil-
ity can be valuable in understanding and illustrating
the dynanlic nature of ecosystems and in evaluating
current ecosystem health.
Process Versus Structure Goals
Process and structural goal setting approaches are
important to management of Zone I lands. These
concepts originated with establishment of wildernesses
and natural areas where the goal was to manage for
naturalness. The proper role of fire in wilderness and
natural areas has been characterized in terms of
process-oriented and structure-onented goals (Agee
and Huff 1986). Expressed simply, do we want a
natural fire regime (process) or rather the vegetation
that a natural regiuie would have created (structure)
(Van Wagner 1985)? The answer to this may always
involve some degree of debate because of philosophical
differences over the concept of natural (Kilgore 1985l.
In practice, both approaches or a mixture of the two
may be appropriate depending on circumstances. Prac-
tical aspects such as costs, fire safety considerations,
1nd size and boundaries of the ecosystem will often
Jetermine the most appropriate approach.
USDA Forest Service Gen Teen Rep r~MRS.GTR.42'vol 2 2000
Grown
A strictly process-oriented goal is probably only
appropriate in large wilderness areas. The process
goal approach modified by practical considerations
will usually be necessary.
In understory fire regimes where surface fuels have
accumulated to the point that high iiltensity tlre is
likely, a structure-oriented goal is the best approach to
ultimately achieve natural conditions. After fuels have
been reduced using a prescription for low severity fire
to avoid killing the overs tory, a process goal of allow-
ing natural ignitions can be followed ifit will maintain
the understory fire regime (Bonnickson and Stone
1985). Structural goals will continue to find applica-
tion in understory fire regime types to restore and
even maintain the natural role of fire. The structural
goal approach is probably the best for management of
threatened and endangered species. It may also be
more efficient and esthetically pleasing (Agee and
Huff 1986).
Mixed fire regime types in wilderness areas present
variable, complex landscape patterns that can make
structural goals difficult to achieve. Fire frequencies
in the mixed type typically range from 35 to 100 years.
In some localities fire has been absent long enough
that fuels and stand structures appear to be falling
outside the range of historical variability (Arno and
others 2000). In such cases, where accumulated sur-
face fuels and naturally occurring ignitions would
favor stand-replacement fire, structural goals aimed
at retaining a portion of the overs tory may be appro-
priate to restore the mixed fire regime. If excessive
fuels have not accumulated, process goals seem to be
the most reasonable.
Another consideration in wilderness areas, regard-
less of whether structural or process goals are chosen,
is when and where to use prescribed fire to meet
wilderness objectives. In the contiguous United States
75 percent of Congressionally classified Wilderness
areas, which occupy half of the classified wilderness land
area, are too small to maintain natural fire regimes by
relying strictly on natural ignitions (Bro-w"1l 1993). Con-
straints such as concern over escape fire. lack of light-
ning-caused fires, conflicting wilderness goals, and air
quality regulations will require prescribed fire to restore
fire and mimic natural processes. Decisions to use pre-
scribed fire must be ecologically based, but also with the
realization that exacting solutions to mimicking natural
fire processes are probably not feasible. Neither the
determination of fire history nor applications of pre-
scribed fire are precise undertakings.
For residential and commercially zoned lands (Zones
I and II), structura.l goals are the most appropriate.
Clearly definable and measurable end points are be-
ing sought. For example, specific conditions such as
tree species and size, stand age distribution, patch
size. stimulation of shrubs, increased forage production,
and reduced fuel quantities may be desirable objectives.
197
'"."^..........---......-.oor>..-...,-'~,._,'.,_.._,..._-"~'" ._~_. '-'--~
Brown
Landscape Assessment
Managing biodiversity and for sU~itainability of eco-
system components and pfllcesses requires a land-
scape perspective. Small c:cosyst.ems are found within
larger ecosystems, indi vid u aJs occur wi t.hin comm uni-
ties, and short-term processes are nestE.d within longer
term processes (Kaufmann and others 1994). The
various scales fit into il. hierarchical structure that
determmes patterns of d iVHsity for an area (Bourgeron
and J cnsen 1993). A major challenge to setting vegeta-
tion and fire objectives in the context of ecosystem
management is evaluating and interpreting the eco-
logical sigaificance ofmultlfJle scales. Vegetation scales
range from individual plants. communities, seral
stages, potential vegetation types, to the biome level.
Species and individual plant commwlities are dealt
with using a fine filter approach. Traditionally, as-
sessments of fire effe('ts and other environmental
impacts have been done on a ploject basis using fine
and mid scale evaluations. The coarse scale aspects of
ecosystems have been largely neglected. The coarse
filter approach, which deals with higher scale levels
such as aggregations of communities, can operate with
relatively little information, yet be an efficient way to
meet biodiversity goals ( Bourgeron and Jensen 1993;
Hunter 1990; Kaufmann and others 1994). A single
ecosystem can be too small to hold viable populations
of all its species, especially large predators. Thus, the
coarse filter approach is best used on assemblages of
ecosystems such a:s watersheds and mountain ranges.
Both approaches a~e necessary to eval uate all facets of
an ecosystem and meet the goals of ecosystem man-
agement (Hann and others 1993a).
Assessment of landscape and ecosystem properties
can be undertaken with varying degrees ofsophistica-
tion and effort. Some of these planning efforts, which
are evolving through trial and error, are mentioned as
examples. During the past decade agencies such as the
U.S. Forest Service and Bureau of Land Management
have undertaken landscape analyses on extremely
large areas such as the 200 million acre Upper Colum-
bia River Basin (Keane and others 1996) and smaller
areas such as the Pike and San Isabel National For-
ests and Cimarron and Comanche National Grass-
lands in Colorado (lJ.S. Forest Service 1997) and the
130,000 acre Elkhorn Mountains and 46,000 acre
North Flint Creek Range in Montana (O'Hara and
others 1993). Details of these landscape evaluations
varied but they followed three general steps (Hann
and others 1993b):
1. Characterize the general composition, structure,
and processes of the ecosystems and landscapes
within the designated analysis area.
2 Analyze data to assess changes in structure and
composition and relat~ the changes to previous
management treatments.
1 '10
Chapter 9 Ecological PWlc.ples. Stlllli"l) Fife Reg.."1'es and Managemenl ConSldi"allOns
3. Examine the ecosystem processes important for
the area and their cffects on ecosystem anr'
landscape composition. structure, and rate 0,
change.
Succession Modeling--Simulation of succession
provides a means of predicting the long-term interac-
tion of processes such as fire, insects, disease, and
cutting of vegetation on landscapes of varying scale
Simulation can be helpful to managers and the public
by helping them understand how ecosystems function
and for evaluating different management alterna-
tives. The wider availability of powerful computer
capabilities has led to an increase in succession mod-
eling efforts particularly for landscape applications.
Manager-oriented computer models that simulate
successional processes across large landscapes are
faced with a tradeoff between realistic portrayal of
ecological processes and utility of the model. Some
models are too complicated to use without special
training or assistance. Nonetheless, managers are
increasingly using succession models in their plan-
ning while models are continually evolving and com-
puter capabilities growing.
In choosing a model for a particular application, it is
important for the temporal and spatial scales of the
model to match the intended use (Reinhardt and
others, in press). Models that operate over a period or
decades are useful for scheduling treatments. Fa
example, the Fire and Fuels Extension to Forest Veg-
etation Simulator (Beukema and others 1997) simu-
lates fuel quantities, tree characteristics, and tree
mortality in the event of a fire for single stands.
Managers can use the model to help sched ule thinnings
and fuel treatments when potential fire behavior and
fire effects on an area are deemed unacceptable
(Reinhardt and others, in press).
Models that simulate fire effects over centuries are
useful for providing targets for managers, for estimat-
ing the historic range of conditions, for evaluating
implications of climate change, and for understanding
possible long-term consequences of management ac-
tions. For example, CRBSUM was used to simulate
landscape changes for different management scenarios
in the Columbia River Basin (Keane and others 1996).
Some of the current models that have been applied to
assist land managers are summarized in appendix B.
Restoration of Fire
Restoration of fire is needed to varying extents in
most ecosystems of North America to meet the holistic
goals of ecosystem management. The need for restora-
tion is most evident in high fire frequency regime
such as understory fire regime types and some grass
lands and shrublands where fire has been excluded for
several times longer than the average fire return
interval. Although considerable knowledge supports
! ,C'rI/\ C, ,,~,'t C'"",~" r,nll T"I'h l=.J(,,--, nr-.,1P~,r;Tn-d')-\I'" ~ ?tl('\(l
C~I<-l;Jler 'J ~r:., ,.yll:,j! ~)r~nCIIJie::;, :,)1,::1:1-)' 0:,' -~,"-, '
the need for restoration of fIre inCo wildland ecosys-
tems, constraints and obstac~es confront land manag-
ers (Browll and Arno 1991. Mutch 1994) Llmited
funding, zllr quality restrictions, concerns OV(;f escape
fire, and inadequate public support can pose difficul-
ties. Sam,' breakthroughs in manaf:6ngemissions aIlt!
obtaining support have provided marc latitude for
prescribed fire programs (Mutch and Cook 1996)
Successful restoration involves clearly stated objec
tives, plans based on scientific knowledge of fire's role
in the ecosystem, and adaptive learning from pre-
scribed fire efforts. Adaptive learning is important
because !liescribed burning usually improves with
experience. Prescription conditions and tiring tech-
niques may need to be modified to achieve objectives
such as d given level of fuel reduction or to meet
constraints such as holding overstory mortality to
certain limits. Fire may not spread adequately under
an initial prescription, thus requiring lower fuel mois-
ture contents or hig.her wind speeds to be successful.
Restoration of fire can be undertaken on an entire
ecosystem or on an individual plant community basis.
Ideally, restoration of individual plant communities
would be based on ecolo!:,rical considerations of the
broader ecosystem of which they are a purt. The extent
of ecosystem assessment that is appropriate for plan-
ning restoration will depend largely on land owner-
ship and direction given to management. For large
land ownl,rships, restoration of entire ecosystems or
large landscape areas is the soundest approach to
manage landscape pattern and meet biodiversity goals.
It also al;ows for effective placement offuel treatments
designed to disrupt fuel continuity and reduce threat
of large fire occurrences. The steps undertaken by
Keane and Arno (1996) to restore fire in the whitebark
pine ecosystem may be useful in other situations
including grasslands and shrublands. They recom-
mend first, an inventory oflandscape and stand char-
acteristics at multiple scales; then, writing descrip-
tions of the important processes of the landscape and
stands Landscapes and stands can then he prioritized
for restoration treatment and selected based on in ven-
tory, description, priority, and feasibility Treat:nents
should be designed for each selected stand or land-
scape based on inventory and dGse:riptioll information
and implemented as efficiently as possible. finally,
treatments should be monitored to evaluate restora-
tion success.
Restoration offire in grasslands, shrub steppe, and
savannas requires careful consideration of seasonal
timing and frequency to assure that prescribed fires
will spread at appropriate severities. Once woody
plants have encroached to a point ufdorninating a site,
it becomes djfficult to get fire to spread with sutDcient
heat to kdl aboveg-rouod stems such as oak in savan-
nas (Huffman and Blanchard 1991) and juniper in
USDA focest Ser':;ce lien reel) Rep HtARS GTR42-vol 2 2000
sagebrush/grass communities. Perhaps the h'Teatesl
obstacle to success lies with areas that have succes-
sionally lost the native mix of species and lack suf1i-
cient grass fuel to carry fire. Seeding of native species
fullowing fire may be necessary to restore a resem-
blance of former plant composition. Where conifers
invade grasslands such as pinyvn-juniper and inland
Douglas-fir (Gruell and others 1986), successful spread
of surface fire may require fuel enhancement work
such as cutting numerous trees to create adequate
surface fuels. Otherwise, crown fire may be required,
which will necessitate a more flammable, narrow fire
prescription that can limit burning opportunities.
Prescribed Fire and Silviculture
Prescribed fire and silviculture can go hand in hand
for restoration offorest stands and ecosystems. Some
consider prescribed fire to be a silvicultural technique
even though it goes far beyond the usual goals of
silviculture that are oriented to producing tree prod-
ucts and desirable forest stand structures. One debat-
able point is the extent to which it is desirable to have
management mimic the kinds of stands and landscape
structures that typified presettlement fire regimes.
However, an understanding of similarities between
characteristics of fire regime types and silvicultural
stand structures can be helpful for integrating fire
with silviculture to restore fire as a process and meet
ecosystem management goals. The following descrip-
tion of stand structure and silvicultural practices
based on a discussion by Weatherspoon (1996) applies
to individ ual stands. Stands can be treated differently
to manage landscape-level vegetation.
Even-Aged Stands-These stands originated natu-
rally mostly from high-severity, stand-replacement
fires that killed most of the trees. Silvicultural meth-
ods that produce even-aged stands include clear-cut-
ting, seed tree, and shelterwood cutting. Shelterwood
or seed trees are typically removed after regeneration
is secured. Pile burning or broadcast burning is com-
monly used to reduce fuels and prepare sites for
regeneration. Leaving snags, large downed woody
material, and untreated patches in larger treatment
units is important for meeting biodiversity goals.
'fWD-Storied Stands-These stands were associ-
ated with moderate to high severity fire typical of the
mixed fire regime type. Retention shelterwood (also
called irregular shelterwood or shelterwood without
removal) is the silvicultural method for treating the
stand Prescribed underhurning can often be practiced
to manage fuels and create within-stand diversity.
Once created, the stand would never be devoid oflarge
trees because each regeneration cutting would be
accompanied by retention of some overstory trees.
Snags could be readily created.
199
--
-
8(0\1'.111
ChCipter'3 E->:oItJQ1cai Pnnc~pjes. ShIlling Flrp. Regimes and Managf'inent CcnS:df'(,H)()f~
Uneven-Aged Stands with Even-Aged or Even-
Sized Groups--These were associated with low to
moderate severity fires associated with the under-
story fire regime type and perhaps Lo SO/11e extpnt with
the low severity end of the mixed fire regIme type
Silviculturally this stand structure IS mimicked with
the group selection cutting method. Skillful prescribed
underburning is required to apply the proper severity
for maintaining this structure. Jackpot burning and
two-stage burning under different prescription condi.
tions may be appropriate.
Uneven-Aged Stands with Fine Tree Mosaic-
These stands are characterized by three or more sizes
and ages of all tree species distributed rather uni-
formly throughout the stand. This stand type is though t
to have developed primarily with shade-tolerant coni-
fers over long periods following stand-replacement
fire. It is incompatible with frequent fires. The indi-
vidual tree selection method is used to maintain this
structure. This stand structure could be considered to
represent open stands of ponder os a pine and longleaf
pine. Ecologically, however, they fit better with the
previous category of even-aged groups.
Understory Fire Regime Type
Restoration of t.he Ullderstory fire regime type re-
quires application offrequent, low intensity fire, which
has been excluded for excessive periods of time. Rest.o-
ration approaches can vary considerably depending on
stand and fuel conditions. The objective generally is to
create more open stand structures consist.ent with
historical disturbance regimes. A wide range of stand
densities can be appropriate depending on site poten-
tial and silvicultural objectives. Various even-aged
and uneven-aged stand structures can be utilized.
Favoring the long needle pine component through
regeneration and retention of old growth trees is
frequently a high priority need. Often the major prob-
lem to overcome is excessive understory fuel accumu-
lations particularly live ladder fuels, and buildup of
duff around the base of desirable leave trees. Another
consideration is burning to encourage the historical
understory vegetation diversity. This requires burn-
ing during the growing season, which is a departure
from the traditional application of prescribed fire
during the spring, fall, or winter dormant seasons.
Conducting the first prescribed fire after a pro-
longed period of no fire must be done cautiously to
avoid flare-ups in sapling thickets or rough that might
kill desirable trees. For ponderosa pine, thinning of
dense understories and piling and burning slash be-
fore cond uet ing a prescribed lInderburn may be neces-
sa.ry to reduce flammability and remove competitor
species that might survive most prescribed fires (Fiedler
and others 1996). However, too much caution where
the understory consists of thick patches of flr will
result in inadequate fire. Some fuel augmentatlOn by
cuttlllg small fir can help carry the fire with adequate
intensity to kill the fir. A series of prescribed fires
~.1.l med at gradually reducing the accumulated live and
dead f lIels may be necessary to return stands to where
maintenance underburning is easily manageable
(Sackett and others 1996). The best approach to resto-
rat.ion must be determined on a case by base basis. but
it will usually require a combination of mechanical
treatments and prescribed fire repeated over a period
of years.
Mixed and Stand-Replacement Regimes
The mixed fire regime includes a wide range of stand
structures and landscape patterns that result from
highly variable fire severities. Individual fires may be
of either nonlethal understory or stand-replacement
severity, or a combination of both severities. Thus,
managers have considerable latitude in designing
prescribed fire and silvicultural activities (fig. 9-6).
Although little guidance based on past restoration
efforts exists, the best way to determine restoration
objectives is on a large landscape basis because of the
wide latitude in individual stand structures. The chal-
lenge is to provide a diversity of stand structures with
retention of snags and some coarse woody debris in
forest ecosystems and unburned patches in grass-
lands and shrublands. In wilderness and natural area
management where fires have not been previously
allowed, avoiding excessive stand-replacement due to
accumulated fuels may be important.
Stand-replacement fire severities can be created
from either severe surface fire or crown fire. Wildfires
over prolo~ged burning periods can leave large pro-
portions of both severities as observed in lodgepole
pine (Brown and ethers 1994). High severity surface
fires may be more readily prescribed and achieved
than crown fires due to the higher risk and fewer
burning opportunities for prescribed crown fires. Eco-
logical effects of severe surface fire and crown ~re
differ. Crown fire consumes foliage that otherw~
would fall and protect the soil. It can kill seeds lD
cones redistribute nutrients in ash, and provide more
, Where
chance for regeneration by offsite colonizers. .
silvicultural objectives are being pursued, an Impor-
tant consideration is avoidance of excessive fragmen-
tation caused by intensive small-scal~ cutting and ,~
prescribed fire activities. Provision for snags and coane .~
woody debris is also important.
.,d C'"n-Iro r,r>n
Tech Reo
Grazing and Exotic Plants
Introduced exotic species and grazing are two m~,
. h tTorU-
problems that can seriously interfere Wit e . _~
Well-intention"",
restore fire as an ecosystem process.
"-d;--.':r--:: ,1 ~ CUII_,g '-.J ~'f:i)r;I(J:I-,'S St.rltlilJ Fire GCJ:(Jl':.~::; .]'lCl '.-1JrlJ~)>:,q)ent C~J;is,der.)tIOflS
~;:l)Wll
Figure 9-6-Aspen is being successionally replaced by fir, Bridger- T elon National Forest,
Wyomlflg. Restoration will require a stand-replacement disturbance, which could be facilitated by
cutting some of the conifers.
prescrdwd firt', and silvicultural and rangeland en-
hancement a,'ti\'ltips, can fad drastically unless graz-
iLg and exotic plants are anticip::lted and managed
properly.
Grazing--Excessive grazillgcan be the biggest hin-
drance to successfulllse ofpn,scribed fire where grass
\'egetatlOn is a :najor component, parLcularly in west-
ern gTasslands and shrub/grass vegetation types
(\\'right and Bailey 1982). It is more of a problem for
bunchf;f<lsses than rhizomatous grasses (Mack and
ThompS\lll 1982). Overgrazing in the absence offire as
well as fullowing fire can reduce plant diversity. Graz-
ing too S(lon following fire can ellnJlnate or greatly
red UCl' de.;] rable vegetation. In grassland areas woody
plants arc competitively favored, \\'hich could defeat
the purpo:oe of burning to halt woody plant
encroachrnt'nt.
Depcr:ding un sIte potential and grazing presiiure,
gTazing~houjd be deferred 1 to 2 year,.. following fire in
eco::-:.::-tc'IllO' such as sagebruslL/graoes and sC'midesert
shrub i \\' fig!] t and 3a ilt:\, 1982). Iii forests such as the
aspen L'[Jl'. 1~lten"i\:e graiJrlg or spruutlng plants by
li\"f'stock and \\.dd ungulates. espl'u,i1ly elk, following
preo-crclH'd fin',: can greatl\' retard plant recovery
SelJall prl'"crdwd burn:; are particularl\ vulnerable to
()\lTU t Ii I /:ltlun ')('ca use ofcollcen trd kd gT<1/.i ng ( Bart()~
and utlaT"; 1991/.
uso:.., FJrt~:.;t St:C;ILe Get: Tecll flt-:'j.) nfJ\r~,~; GTF-l-t..2-vol :) 2000
'-----
Grazlllg prior to a prescribed burn can easily reduce
fine fuels to a point where fire will not spread success-
fully nor have sufficient heat to ignite or kill woody
plants. At least 600 lb/acre of herbaceous fuel is needed
for successful prescribed fire in grassland and grass!
shrub vegetation (Wright and Bailey 1982).
Exotic Plants-Fire can create f,,\'orable sit.~s for
nonindigenous plant species to become established
and flourish. If exotic plants alread~' grow in or near
areas that are candidates for prescribed fire, a poten-
tial problem exists. Aggressive exotic species can com-
petitively exclude nati\'e vegetatiun. Severe fires that
expuse large areas of mineral soil are most apt to be
invaded by exotic plants; if exotics are already estab-
lished, their dominance may be accelerated. Lower
severity burns are m()re resistant to proliferation of
exotics because many native species ~ prout and quickly
occupy the site.
Cheatgrass. a nonindigenous annual that domi-
nates millions of acres, is an extreme example of a
species fa\'ored by fire Its invasion of the sagebrush-
steppe \'egetation t:r'pe has led to increased frequency
of wildfire due to abundant. early cunng fine fuels The
result is permanent con\'erSlOn to annual grassland
and dlsruptlUn of the historic tirl' rC;lI1lC (\Vhise:lant
1990) Anotht'r problem With llulundigenous plant,;
can cccur from seeding nOllllatl\'C g-rasses sllch as
21)1
- t.
,.,....._.."'''._..'.ce"^_..',,..__._.__.,c._
tJ(':.\i"
annual ryegl'ass on severely burned sites as part of
wildfire rehabilitation efTorts. This practice, which is
intended to stabilize sods, can delay reestablishment
of natIve species and possibly alter long-term commu-
nity composition (Conard and others 1991).
A far difTerent problem is caused by exotics such as
Chmese tallow, which has invaded coastal marshes of
the Southeast. Its invasion causes a shift from grass-
dOllllIlated communities to a sparse forb-dominated
community that is much less flammable and acts as a
fire break. Consequently, once Chinese tallow gains
dommance on a site, prescribed fire cannot be effec-
ti ve Iy used to control the exotic and encroaching woody
pla.nts. Thus, the grass-dominated marsh communi-
ties are reduced.
Fire Prescriptions
Ecosystem management has brought new challenges
to the application of prescribed fire primarily due to
the increased scale and complexity of some prescribed
burning (Zimmerman and Bunnell 1998). Tradition-
ally, prescribed fire was applied on small, relatively
homogeneous units of a single land ownership. Pre-
scribed fire will continue to be important for small-
scale operations. But to meet some ecosystem goals,
prescribed fire needs to be applied over extensive
areas that contain a variety of vegetation communities
and fuel conditions.
In designing fire prescriptions, a strong, clear connec-
tion is needed between ecosystem goals, resource objec-
tives, and fire objectives. This helps assure that pre-
scribed fire will accomplish the desired effects. It can
also help in choosing proper technical aids fordetermin-
ing the prescription and in assuring fires are cost
effective and safely conducted. Designing prescriptions
through a visible, logical process can also demonstrate
professional competence and promote credibility of those
in charge of the prescribed fire activities.
DetIning fire objectives boils down to specifying first
order fire effects that describe what the burningshould
immediately accomplish (Brown 1985b). Treatment
objectives need to specify: (1) how much ufwhat kind
of organic matter should be consumed, (2) what veg-
etation should be killed, and (3) what the size of
burned and unburned patches should be. Constraints
on achieving the treatment objectives must also be
considered. These can be thought of as the fire effects
that should be avoided. Controlling fire, managing
smoke, and avoiding overstory mortality are the com-
mon constraints. Specifying objectives and constraints
is a matter of declaring what the fire should accom-
plish and avoid. Both are fire objectives of sorts, so why
regard them differently? One reason is thai it helps in
demonstrating an awareness of beneficial and unde-
sirable aspects of fire and in explaining the prescribed
fire plans to ot.hers.
202
'-':::~Jc~er (1 f(',)10~:(J' ;:;'II"':~IP:f'~ Srid:.r'G ,=-:re ='e';lr~es d.",J ,\1.1 J-';t"-"'" ~,~"';,:(1f" j ,
Depending on resource objectives, the ti ft' c bJ eet I \ L':'
may call for a wide or narrow prescription wtrldow. Fo:-
example, t.he resource objective to restore fire a,
process in a nonlethal understory fire regime type m.
only require that prescribed fire be able to spread with
minimal mortality to the overstory, an objective that
could be accomplished with a wide prescription \':Ill-
dow. The specific resource objective of attain 109 natu-
ral regeneration while retaining some large downed
woody material may call for a fire objective that
specifies exposure of 20 to 30 percent mineral soil
without consuming more than half of the large downed
woody material. This would require a narro\V prescrip-
tion window.
Occasionally, conflicts may arise between flJ'e objec-
tives and constraints. A common example is betv.reen
the objective to reduce fuels by burning at low fuel
moistures and the constraint to control smoke produc-
tion. Conflict can arise between different objectives;
for example, to expose a high percentage of mineral
soil and to leave large downed woody material for
other ecosystem benefits. When conflicts anse, com-
promise may prevent the fire from achieving the re-
source objectives. It is important to recogmze those
situations so a potentially unsuccessful prescribed fire
can be avoided.
Many technical aids are available to assist in pre-
paring fire prescriptions. Most involve prediction of
information such as weather probabilities, fuel 10"
ings, fuel consumption, fire behavior, tree IDortalic. .
and plant response. Two technical aids-both with
user guides that can help in writing and explaining
prescribed fire objectives and designing fire prescrip-
tions-are relevant for applications across the United
States and much of Canada. They are the Fire Effects
Information System-FE IS (Fischer and others 1996)
and the First Order Fire Effects Model-FOFEM
(Reinhardt and others 1997).
FEIS-This is an easy to use, computerized knowl-
edge management system that stores and retrieves
current information as text organized in an encyclope-
dic fashion. FEIS provides fire effects and related
biological, ecological, and management information in
three major categories: plant species. wildlife species,
and plant communities. The plant species category
includes for each species, information On taxonomy,
distribution and occurrence, value and use, botanical
and ecological characteristics, fire ecology, fire effects,
and references. A citation retrieval system can be
searched independently by author and keyword.
Although the system was originally developed to meet
prescribed fire needs, it is now recognized as a valu-
able aid for obtaining information about species ecol-
ogy for any application. It can be accessed througb
U.S. Forest Service Web site:
http://wwwfs fed usldatabasc/feis
USDA Forest Service Gen Tech Rep RMRS-GTR.42"101 2 2000
Chapter 9 E::COk'\lIc.a1 PnnClples. Shillll)g File Regimes and Management Conslderalions
FOFEM- This system was developed to predict the
direct consequences of fire, that is, first order fire
effects. FOFEM computes duff and woody fuel con-
sumption. mineral soil exposure, fire-caused tree mor-
tality. and smoke production for many forest and
rangeland ecosystems. An update is scheduled to add
soil heating effects. FOFEM contains a fire effects
calculator to predict effects of fire from the burning
conditions and a prescribed fire planner to compute
the burn conditions necessary to achieve a desired
effect. Users may enter their own fuel data or use
default values derived from fuel models provided for
natural and activity fuels by many forest cover types.
The model is implemented in a computer program
available for use on a PC or Forest Service computer.
To obtain a current version of the FOFEM software,
contact the authors at the Intermountain Fire Sci-
ences Laboratory, (406) 329-4800, or PO Box 8089,
Missoula, MT 59807.
Research Needs
The goals of maintaining sustainability of all ecosys-
tem components and processes and conserving biodi-
versity present new challenges to land management
organizations. Knowledge of how ecosystems function
and what they provide is essential to making informed
environmental decisions. The following broadly stated
research needs indicate the knowledge required for
managing tire effects on flora and fuel that will con-
tribute to maintaining sustainable ecosystems.
Characteristics of Fire Regimes
· What is the historical range of variability in fire
regime characteristics especially fire frequency,
seasonality, and severity for fire dependent eco-
systems? This should be answered for multiple
spatial scales because of the hierarchical struc-
ture of ecosystems.
· What are the limits to ecosystem patterns and
processes that signal ecosystems are beyond the
boundaries of the historical range of variability?
· To what extent has climate influenced fire regime
characteristics in the past? How might antici-
pated climate change alter fire regime character-
istics in the future?
Effects of Fire on Ecosystem Processes
and Biodiversity
· What are the long-term effects of fire of varying
frequencies and severities on nutrient dynamics
and vegetation?
'JSDA Forest Se-vlce Gen TeCh Rep. RMRS.GTR.42-vol. 2 2000
8rown
· How does fire of varying frequency, seasonality,
and severity influence individual plant species
and plant community development? The empha-
sis for research should be on rare species and other
vegetation components where knowledge is lacking.
· What interactions between insects and diseases
and fire characterized historical fire regimes, and
how has this affected landscape patterns? How do
these interactions change when ecosystems ex-
ceed the natural range of variability and when
various management activities are applied?
· What is the interaction of different ecosystem
scales on ecosystem processes and biodiversity?
To what extent can coarse scale analysis account
for ecosystem processes and biodiversity?
· What are the long-term effects oflargelyexcluding
fire from ecosystems that evolved under fire
regimes?
Restoration of Ecosystems
· What-approaches and methods involving wildland
fire use, prescribed fire, silviculture, and grazing
can be used to restore ecosystems to a semblance
of the historical range of vegetation composition
and structure while meeting the resource needs of
society?
· What fuel management activities can provide an
acceptable level offire hazard and remain compat-
ible with ecosystem goals, especially needs for
coarse woody debris?
· How can nonindigenous plant species be managed
in combination with prescribed fire and resource
utilization activities to maintain biodiversity?
Development of Ecosystem Evaluation
Methodologies
· Continue with development of simulation models
and ecosystem evaluation techniques that can
help in understanding and managing ecosystem
dynamics. Succession and landscape models are
needed that account for interaction offire, vegeta-
tion, fuels, and climate.
· Fire effects models at small spatial and temporal
scales are needed for rigorous fire effects hypoth-
esis testing and as building blocks for models with
larger temporal and spatial scales.
· Determine organizational approaches that allow
complex ecosystem models requiring specialized
skills and high speed computer facilities to be
accessible to all land management organizations
and units.
203
"-'..- '"
1'-..--...-..-.......'.--..-....-
Notes
, .,.....,-..... r_ ~
r"'~~.;~~ f""."_'~ '.-,..,h n............ o,,~n~ r'..TO 1'"1 "t ') ?nnn