The URL can be used to link to this page

Home My WebLink AboutWildfire 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