a.

href Ch N

nders

Opinionconcerns that too frequent prescription burning can nega-tively affect biodiversity and ecosystem processes. Howev-er, their paper used this issue inappropriately as aspringboard for speculating on the evolution of plant traits,confusing issues of adaptive traits and adaptations, andmissing the fact that different fire regimes can select forvery different plant traits. No species is fire adapted butrather is adapted to a particular fire regime, which, amongother things, includes fire frequency, fire intensity andpatterns of fuel consumption [2]. Species that exhibit traitsthat are adaptive under a particular fire regime can bethreatened when that regime changes. For example, manyof the species-rich Mediterranean-type climate (MTC)shrublands are resilient to periodic high-intensity crownfires at intervals of several decades or more. However,when the fire frequency increases, species can be rapidlylost [3,4]. Bradshaw et al. [1] incorrectly equate adaptationto fire as adaptation to frequent fires and are not justifiedin inferring that frequent prescription burning is an ad-aptationist view. Here, we discuss those traits with ap-

leading to a crown fire regime; traits change to thin bark,retention of lower dead branches, and serotinous conesthat synchronously release seeds following such stand-replacing fires [5,6]. High-intensity crown fires are notstand replacing in Australian eucalypt forests as most ofthe trees, big and small, resprout epicormically along thelength of the bole, replacing the tree canopy within a yearof such fires. In MTC shrublands subject to crown fires,many species restrict seedling recruitment to the immedi-ate post-fire years, arising from previously dormant soil-stored seed banks or synchronous seed release from seroti-nous fruits [2]. These traits are adaptive in fire-proneenvironments and are key to providing resilience to specificfire regimes.

Adaptive traits as adaptations to fireA key evolutionary question about fire-adaptive traits iswhether they represent fire adaptations, defined as traitsthat originated in response to fire. Fire-adaptive traits thatoriginated in response to some other environmental factor,and then were appropriated for their value in fire-pronelandscapes, are exaptations [7]. Distinguishing betweenCorresponding author: Keeley, J.E. (jon_keeley@usgs.gov).Fire as an evolutionshaping plant traitsJon E. Keeley1,2, Juli G. Pausas3, Philip WWilliam J. Bond4 and Ross A. Bradstock5

1US Geological Survey, Sequoia-Kings Canyon Field Station, T2Department of Ecology and Evolutionary Biology, University o3Centro de Investigaciones sobre Desertificacion of the SpanisMontcada, Valencia, Spain4Botany Department, University of Cape Town, Private Bag, Ro5Centre for Environmental Risk Management of Bushfires, Univ

Traits, such as resprouting, serotiny and germination byheat and smoke, are adaptive in fire-prone environ-ments. However, plants are not adapted to fire per sebut to fire regimes. Species can be threatened whenhumans alter the regime, often by increasing or decreas-ing fire frequency. Fire-adaptive traits are potentially theresult of different evolutionary pathways. Distinguishingbetween traits that are adaptations originating in re-sponse to fire or exaptations originating in responseto other factors might not always be possible. However,fire has been a factor throughout the history of land-plant evolution and is not strictly a Neogene phenome-non. Mesozoic fossils show evidence of fire-adaptivetraits and, in some lineages, these might have persistedto the present as fire adaptations.

Adaptation to fireIn a recent publication [1], S. Don Bradshaw et al. raisedconcerns about the widespread application of prescriptionburning in hotspots of biodiversity, and we share those406 1360-1385/$ see front matter . Published by Elsevier Lry pressure

Rundel2,

e Rivers, CA 93271, USAalifornia, Los Angeles, CA 90095, USAational Research Council (CIDE - CSIC), IVIA campus, 46113

ebosch 7701, South Africaity of Wollongong, Wollongong, NSW 2522, Australia

parent adaptive value in fire-prone environments and theextent to which we can demonstrate that they are adapta-tions to fire. We also address the question of what fire-adaptive traits and fire adaptations can tell us about firemanagement.

Adaptive traitsAdaptive traits are those that provide a fitness advantagein a given environment. There are many plant traits thatare of adaptive value in the face of recurrent fire and thesevary markedly with fire regime. For example, North Amer-ican conifers (Coniferophyta) subject to frequent lightning-ignited fires have thick bark, which functions to protect theliving tissues from heat damage during surface fires; theyalso self-prune lower dead branches, which ensures a gapin fuels between the dead surface litter and live canopy.Thick bark and self-pruning are adaptive in this surfacefire regime. At higher latitudes as productivity declines,lodgepole and jackpine forests lack the potential for growthsufficient to keep the canopy away from surface fuels,td. doi:10.1016/j.tplants.2011.04.002 Trends in Plant Science, August 2011, Vol. 16, No. 8

Box 1. Adaptations and exaptations

An adaptive trait enhances fitness in a defined habitat and adaptationis the process of acquiring adaptive value by the natural selection ofnew variants [28,29]. Exaptation defines an adaptive trait with aparticular contemporary function, but one that was previously shapedby natural selection for another function [7,25]. However, exaptationscan become adaptations when natural selection acts and reshapes thetrait.We distinguish five scenarios of change in a trait state (Figure I).

An adaptive trait might not change through time regardless of theselective environment (scenarios 1 and 2 in Figure 1). Such traitscannot be described as adaptations to the current selectiveenvironment as there is no evidence that natural selection shapedthis trait. Other adaptive traits that were shaped by natural selectionunder a previous evolutionary pressure, but not under the current(fire-prone) environment (scenario 3) would be adaptations toprevious evolutionary pressures and exaptations to the currentenvironment.Fire adaptations are those adaptive traits in which natural selection

is acting under the current fire-prone environment to shape the trait,and it is independent of how long this pressure has been present(scenarios 4 and 5). In some cases (scenario 4b) selective pressurescan cause traits to revert to a former state and, when conditionschange, fire might replace the former selective pressure.It is crucial to note that our interpretation of adaptations and

exaptations depends on current understanding of the entire evolu-tionary scenario. If the knowledge base only extends back to thesecond half of the time period in Figure I, it is not possible todistinguish between different scenarios. This problem is inherent inretracing the history of most traits.Plants can have traits of adaptive value under specific fire regimes,

but their origin can follow different evolutionary scenarios. Forinstance, one might find that resprouting is highly adaptive in sometropical forest trees that are suddenly cast into a fire-prone environ-

Opinionadaptations and exaptations is a high bar that is difficult todemonstrate as there is not a simple dichotomy betweenpaths leading to adaptations and exaptations (Box 1).

There are several flaws in Bradshaw et al.s analysis [1]of fire adaptations. One is the misconception that fire is arecent Neogene [23 million years ago (Ma) to the present]phenomenon; however, fire has been part of Earth systemsince the Silurian origin of land plants 443 Ma [8]. It wasglobally important through much of the Mesozoic (25166Ma) [911] and, by the Late Cretaceous (10066 Ma), sometaxa were already specialized for fire-prone environments[12,13]. In the highly diverse Australian Eucalyptus, epi-cormic resprouting regenerates forests rapidly after high-intensity crown fires and phylogenetic studies support anearly Tertiary origin (62 Ma) [14].

ResproutingAnother misconception in [1] is the belief that vegetativeresprouting cannot be an adaptation to fire because it isfound in non-fire-prone habitats. There are many forms ofresprouting and it is expected that different resproutingtypes might have appeared in different lineages in re-sponse to different evolutionary pressures. No one hasever argued that all instances of resprouting are adapta-tions to fire; however, it is still an open question as towhether post-fire resprouting originated in response to firein some lineages [2]. For example, resprouting is coupledwith a lignotuber in various lineages (Figure 1), and thetight coupling of this ontogenetic trait with fire-proneenvironments [15] suggests that it is a fire adaptation.ment owing to anthropogenic impacts, but this adaptive trait mightnot have arisen in response to fire (scenarios 1, 2 or 3). By contrast, itis plausible that some fern lineages have had a long association withfire and resprouting might follow scenario 5 [2]. MTC shrub lineagesresprouting from lignotubers might best be interpreted in terms ofscenario 4a, an old origin of resprouting in response to diverse factorsbut reshaped later by fire.[()TD$FIG]

Time (evolution)

12

34a

4b

5Tra

it st

ate

TRENDS in Plant Science Figure I. Evolutionary scenarios of change in a trait state (continuous lines can

be interpreted as representing a period in which fire acted as an evolutionary

Trends in Plant Science August 2011, Vol. 16, No. 8SerotinySerotiny as an adaptation to fire has also been questionedby [1] because it probably post-dates onset of Tertiarydrying and might be an evolutionary response to low soilfertility. However, phylogenetic studies in the Australiangenus Banksia conclude that fire-adaptive traits, and se-rotiny in particular, originatedmuch earlier and they placetheir origins in the Paleocene (60Ma) [16].Bradshaw et al.s[1] linking of serotiny to infertile soils follows Stephen D.Hoppers [17] contention that low fertility soils, such asthose widespread in Western Australia and South Africa,are derived from ancient Cretaceous substrates. Hopperargues that this ancient association with infertile soils isthe primary factor driving the evolution of plant traits inthese shrublands. However, plants persist in a multivari-ate environment stressed by not only soil fertility, but alsoclimate and potentially fire, not to mention biotic interac-tions. Indeed, low fertility soils select for traits such assmall sclerophyll leaves and provide well-drained sub-strates, both of which increase the likelihood of fires. Inour view, evolution on these substrates is driven by acombination of geology, climate and fire [2]. Singling outone of these factors as the only determinate of trait evolu-tion handicaps our ability to understand plant evolution.

Serotiny is tied to crown fire regimes and, in the absenceof fire, there is relatively little successful recruitment, mak-ing it clearly of adaptive value in these systems. However,demonstrating that serotiny isanadaptation that evolved inresponse to fire, is another matter. Bradshaw et al. [1]conclude that because serotiny is concentrated in the two

pressure and dashed lines a period with a different selective environment).

407

[()TD$FIG]

Opinionsouthern hemisphereMTC regions characterized by lowsoilfertility, it arose in response to soil infertility.However, theyconflate two separate issues: the reproductive strategy of asingle post-fire pulse of seedling recruitment and aerialversus soil seed storage.

Sorting out the adaptive value of serotiny requiresconsideration of ultimate versus proximal causation. Fourof the five MTC regions have species that restrict seedlingrecruitment to a post-fire pulse; the reason for this is thatfires are a predictable ecosystem process and the removalof the dense shrub canopy by fires provides superiorresources for the recruitment of some species. The proxi-mal mechanism of how species do this varies with theregion; in two of the MTC regions, dormant seeds are

Figure 1. Resprouting after fire from a lignotuber in the California chaparral shrub Aden

kingdom. These basal burls are an ontogenetic trait that forms early in development an

burning observed in many woody plants.

408Trends in Plant Science August 2011, Vol. 16, No. 8stored in the soil and serotiny is rare, whereas in two otherregions, many dominant shrubs store seeds in serotinousfruits in the shrub canopy. We argue that low soil fertilityin South African fynbos and Western Australian heath-lands has selected against soil storage in favor of aerialstorage because high-nutrient seeds are subject to moreintense predation when exposed on the soil surface(Figure 2). Higher fertility soils in northern hemisphereMTC shrublands of California and the MediterraneanBasin allow shrub taxa to capitalize on the advantagesof soil-stored seed banks, which includes a capacity foraccumulating larger seed banks and a greater resilience tounpredictably long fire-free intervals. The one MTC regionnot included here is central Chile and the shrublands there

TRENDS in Plant Science

ostoma fasciculatum. It is one of a wide diversity of resprouting modes in the plant

d, in this respect, differ from swollen basal burls initiated in response to cutting or

lack both, a post-fire pulse of seedlings and serotiny, whichis interpreted as a result of the diminution of natural firessince the late Miocene, owing to the rise of the Andes thatblock summer lightning storms [2].

Daniel I. Axelrod [18] also expressed skepticism aboutthe role of fire in the evolution of serotiny in northernhemisphere pines, in part because he understood fire to bean anthropogenic factor and unknown throughout landplant evolution. He proposed that drought was a moreprobable factor selecting for serotiny. We argue that ifdrought were a stronger driving force than fire, it wouldbe more likely to select for a bet-hedging strategy in whichseed dispersal was spread out over multiple years, ratherthan all seeds being dumped immediately after fire, whenthere is a significant probability that it could be a dry year.

[()TD$FIG]

Infertile soils

Large &high nutrient

seedsIntense seed predation

Canopy seedstorage

(+) Reduces seed predation(-) Limits seed bank size to amount that can be stored on plant(-) Increases carbon costs for seed protection(-) Not resilient to fire cycles longer than adult lifespan

TRENDS in Plant Science

Figure 2. Model of factors selecting for serotiny and costs and benefits of serotiny.

Low soil fertility is hypothesized to select for larger seeds with higher nutrient

stores to enhance seedling survival. Predation pressure is heightened owing to the

lower nutrient stores in most plant tissues in the community and this selects for

canopy seed storage. Modified, with permission, from [2].

OpinionBradshaw et al. [1] suggest rainfall predictability is amore probable selective force behind the evolution of se-rotiny than is fire. They contend that winter rainfall is lesspredictable in the two northern hemisphere MTC regionsand, thus, soil seed storage, which could spread germina-tion over multiple years, would be preferable to serotiny.However, northern hemisphere MTC shrublands with soilstored seed banks do not exhibit seed carryover after fireand, just like their southern hemisphere serotinous cou-sins, they too have a single pulse of seedling recruitmentafter fire. Another weakness of this theory is the likelihoodthat subtle differences in rainfall predictability betweenthe northern and southern hemisphere MTCs have fluctu-ated greatly since the Tertiary origin of serotiny [2].

Heat-shock triggered germinationWe agree with Bradshaw et al. [1] that the physical dor-mancy of seeds has had multiple origins; however, becauseof this, Bradshaw et al. cannot use the presence of physicaldormancy in non-fire-prone landscapes as evidence that itis always an exaptation in fire-prone environments. Thepossibility remains that, in some plant lineages, physicaldormancy originated in response to fire and, in otherlineages, it was appropriated by species suddenly cast intoa fire-prone setting. Ultimately, however, sorting out ad-aptation from exaptation might be extremely challengingin some lineages (Box 1). For example, the Fabaceae arewell known for being hard seeded, meaning that theyproduce outer integuments of densely packed palisade cellsand waxy cuticles that make the seed impermeable towater until scarified by heat or other abrasive agent.However, not all Fabaceae seeds are water impermeable.In California grasslands, there are species of Lupinus,Astragalus and Trifolium that germinate every springwithout any apparent physical scarification, but many ofthese same species also occur in adjacent chaparral andhave deeply dormant seeds that seldom germinate untilexposed to heat shock from a fire. Although hard seeded-ness in the Fabaceae is widespread, it appears to wax andwane with local conditions and selection for or againstphysical dormancy depends on local conditions (scenario4b in Figure I, Box 1). When derived from grasslandecotypes, deeply dormant chaparral seeds clearly reflectselection for dormancy, despite the fact that the integu-ment structures are a feature of parallel evolution in theFabaceae. We hypothesize that, although the seed coatstructures per semight have originated in response to anynumber of factors, their precise organization into a heat-shock-dependent seed is an adaptation to fire in somesituations.

Germination triggered by combustion chemicalsSmoke-induced germination is a trait that, on the surface,would seem to be a clear adaptation to fire, but it is morecomplicated than appears at first glance. In most MTCecosystems, there are species with dormant seed banks thatonly germinate in the first growing seasonafter fire, and canbe shown experimentally to only germinate in response tosmoke or charred wood extracts. In these cases, smoke-stimulated germination is clearly an adaptive trait. Wheth-er the origin of this adaptive trait represents an adaptationevolved in response to fire remains an open question.

The relatively recent discovery of a butanolide com-pound (karrikinolide) in smoke that will trigger germina-tion of a vast array of species, many that lack any ecologicalconnection with fire, has raised questions about the selec-tive role of fire inMTC ecosystems [1]. In an apparent questfor the unifying theory of seed germination, several inves-tigators [19,20] have focused intensively on this compoundand ignored reports of other compounds in smoke thattrigger the germination of MTC species with fire-depen-dent seedling recruitment [21] and other negative evidencefor butanolide being a universal signal [22]. All of the workto date on karrikinolides in germination involves lab stud-ies and essentially nothing is known about its ecologicalrole in fire-prone environments. That this compound trig-gers germination at levels of parts per trillion [19] raisesimportant ecological questions. If it is the universal germi-nation trigger, then how do MTC soil-stored seed banksremain dormant for decades in the face of such extraordi-nary sensitivity?

Bradshaw et al. [1] do provide a refreshing new theoryabout the Mesozoic origin of smoke-stimulated germina-

Trends in Plant Science August 2011, Vol. 16, No. 8tion initially in response to organicmatter decay. However,this theory is subject to all of the same criticisms that they

409

level against fire adaptations. The only potential evidencethey present is the observation that karrikinolide-stimu-lated germination is widespread throughout all majorclades of angiosperms and, thus, is a feature of basic seedmetabolism that possibly originated during the early evo-lution of angiosperms. In their view, only organic matterdecay and not fire has been present since early angiospermevolution, but this is not supported by the vast literatureon Paleozoic and Mesozoic fires [813].

However, assuming for arguments sake that karriki-nolides evolved in response to some other environmentalfactor and are the only compounds in smoke that play a rolein germination of species with strict post-fire recruitment,the fact that karrikinolides might have played a role in theevolution of seed metabolism of all angiosperms does notpreclude an adaptive role for species that currently use it tocue germination to post-fire conditions. Given that mostspecies seem to be sensitive to karrikinolides, yet only asmall subset of MTC species have the capacity to remaindormant and then cue germination to the immediate post-fire environment, it seems feasible that these MTC speciesmight have evolved some level of gene regulation control-ling the functioning of karrikinolides in germination. Thiswould constitute a fire adaptation.

Flammability as a fire adaptationEvolution of flammability implies selection for traits thatincrease the frequency or intensity of fires. The idea islargely dismissed by Bradshaw et al. [1] because of theircontention that it is based only on theoretical models.However, studies of trait origins are not readily amenableto experimentation and evolutionary comparativemethodsare among out best available tools. We believe it is a viablehypothesis that species restricting reproduction to a singlepost-fire pulse of recruitment might be under selectivepressure to affect fire activity in ways that enhance theirreproductive success [23]. In crown fire ecosystems, traitsthat could enhance flammability include small leaves,volatile compounds and retention of dead leaves andbranches, to name just a few. There are empirical datashowing that trait distribution in some genera and com-munities is consistent with this hypothesis [6,24,25]. Alsoconsistent with this hypothesis is the observation that,within fire-prone communities, species with post-fire seed-ling recruitment have the most flammable canopies [2,24].This could be selected owing either to greater inhibition ofneighbors [23] or to enhanced seedbank survival by limit-ing deposition of dead fuels on the soil surface [26]. Weacknowledge that there is limited evidence that character-istics contributing to enhanced flammability actually en-hance fitness (i.e. are adaptive) in fire-prone environments,let alone that they represent an adaptation evolved inresponse to fire. However, evidence is likewise lackingthat, in species from crown fire regimes, traits such asretention of dead branches serve another function, whichevolved in response to a non-fire-related pressure.

ConclusionBradshaw et al. [1] raise some legitimate questions about

Opiniontraits that are interpreted to be adaptive in fire-proneenvironments. However, pointing out gaps in the ability

410to trace the origin of many traits to a fire origin is notequivalent to demonstrating these traits arose in responseto other environmental factors. Contrary to their assertion,Bradshaw et al. have not demonstrated that any fire-adaptive trait has a more complex origin or that anysingle trait arose in response to some other factor. Inseveral cases, they have rejected a possible adaptive rolefor traits by defining them in very broad terms so that,across the plant kingdom, the traits appear in many dif-ferent selective environments. More importantly, traitevolution is shaped by many forces that act throughoutthe history of the trait, and it is naive to think that eachtrait should be related to a single evolutionary pressure.

One of the unfortunate aspects of their paper [1] is that ituses evolutionary arguments to draw conclusions aboutappropriate fire management decisions. The assertion ismade that plants with traits interpreted as exaptationsperform worse in the face of recurrent fire than do thosethat represent adaptations originating in response to fire.However, the concept of adaptation versus exaptations onlyrefers to the origin of the trait (Box 1), not to the current role.Thus, there is no reason to believe that plants with exaptedfire traits perform worse than do plants with fire adapta-tions and the authors do not provide any support for such aview. Resource managers and other conservationists areinterested in how plants performunder current fire regimesandmostly regard the evolutionary origin of traits as irrele-vant tomanagement decisions. It is possible formanagers topredict the behavior and dynamics of plants under differentalternativefire regimes on thebasis of their functional traits[27], without regard to their origins.

AcknowledgmentsJGP acknowledges VIRRA (CGL2009-12048/BOS) support from theSpanish Government. Any use of trade, product, or firm names in thispublication is for descriptive purposes only and does not implyendorsement by the US Government.

References1 Bradshaw, S. et al. (2011) Little evidence for fire-adapted plant traits inMediterranean climate regions. Trends Plant Sci. 16, 6976

2 Keeley, J.E. et al. (2012) Fire in Mediterranean Climate Ecosystems:Ecology, Evolution and Management, Cambridge University Press

3 Bradstock, R.A. (2008) Effects of large fires on biodiversity in south-eastern Australia: disaster or template for diversity? Int. J. WildlandFire 17, 809822

4 Keeley, J.E. et al. (2009) Ecological Foundations for Fire Managementin North American Forest and Shrubland Ecosystems, USDA ForestService, Pacific Northwest Research Station

5 Keeley, J.E. and Zedler, P.H. (1998) Evolution of life histories in Pinus.In Ecology and Biogeography of Pinus (Richardson, D.M., ed.), pp. 219250, Cambridge University Press

6 Schwilk, D.W. and Ackerly, D.D. (2001) Flammability and serotiny asstrategies: correlated evolution in pines. Oikos 94, 326336

7 Gould, S.J. and Vrba, E.S. (1982) Exaptation a missing term in thescience of form. Paleobiology 8, 415

8 Pausas, J.G. and Keeley, J.E. (2009) A burning story: the role of fire inthe history of life. Bioscience 59, 593601

9 Scott, A.C. (2000) Pre-Quaternary history of fire. Palaeogeogr.Palaeoclimatol. Palaeoecol. 164, 297345

10 Bond, W.J. and Scott, A.C. (2010) Fire and the spread of floweringplants in the Cretaceous. New Phytol. 188, 11371150

11 Falcon-Lang, H.J. (2000) Fire ecology in the Carboniferous tropicalzone. Palaeogeogr. Palaeoclimatol. Palaeoecol. 164, 355371

Trends in Plant Science August 2011, Vol. 16, No. 812 Watson, J. and Alvin, K.L. (1996) An English Wealden floral list, withcomments onpossible environmental indicators.CretaceousRes.17, 526

13 Collinson, M.E. et al. (1999) Charcoal-rich plant debris accumulationsin the Lower Cretaceous of the Isle of Wight. Engl. Acta Palaeobot.(Suppl. 2), 93105

14 Crisp, M.D. et al. (2011) Flammable biomes dominated by eucalyptsoriginated at the CretaceousPalaeogene boundary. Nat. Commun. 2,18

15 Canadell, J. and Zedler, P.H. (1995) Underground structures of woodyplants in mediterranean ecosystems of Australia, California and Chile.In Ecology and Biogeography of Mediterranean Ecosystems in Chile,California, and Australia (Arroyo, M.T.K. et al., eds), pp. 177210,Springer

16 He, T. et al. (2011) Banksia born to burn. New Phytol. 191, 184196

17 Hopper, S.D. (2009) OCBIL theory: towards an integratedunderstanding of the evolution, ecology and conservation ofbiodiversity on old, climatically buffered, infertile landscapes. PlantSoil 322, 4986

18 Axelrod, D.I. (1980) History of the maritime closed-cone pines Alta andBaja California. Uni. Calif. Publ. Geo. Sci. 120, 1143

19 Flematti, G.R. et al. (2004) A compound from smoke that promotes seedgermination. Science 305, 977

20 Van Staden, J. et al. (2004) Isolation of the major germination cue fromplant-derived smoke. S. Afr. J. Bot. 70, 654659

21 Keeley, J.E. and Fotheringham, C.J. (1997) Trace gas emissions andsmoke-induced seed germination. Science 276, 12481250

22 Downes, K.S. et al. (2010) The fire ephemeral Tersonia cyathiflora(Gyrostemonaceae) germinates in response to smoke but not thebutenolide 3-methyl-2H-furol[2,3-c]pyran-2-one.Ann. Bot. 106, 381384

23 Bond, W.J. and Midgley, J.J. (1995) Kill thy neighbour: anindividualistic argument for the evolution of flammability. Oikos 73,7985

24 Saura-Mas, S. et al. (2010) Fuel loading and flammability in theMediterranean Basin woody species with different post-fireregenerative strategies. Int. J. Wildland Fire 19, 783794

25 Cowan, P. and Ackerly, D. (2010) Post-fire regeneration strategies andflammability traits of California chaparral shrubs. Int. J. WildlandFire 19, 984989

26 Gagnon, P.R. et al. (2010) Does pyrogenicity protect burning plants?Ecology 91, 34813486

27 Noble, I.R. and Slatyer, R.O. (1980) The use of vital attributes topredict successional changes in plant communities subject torecurrent disturbances. Vegetatio 43, 521

28 Dobzhansky, T. (1968) On some fundamental concepts of Darwinianbiology. Evol. Biol. 2, 134

29 Endler, J.A. (1986)Natural Selection in the Wild, Princeton UniversityPress

n

logusolpb.

nseguThel/UK

y Sguer,i/cg

froSen,

Opinion Trends in Plant Science August 2011, Vol. 16, No. 8Plant Science Co

Plant Bio610 AuMinneap

http://www.as

14th Symposium on I1317 Au

Wageningen,http://www.ent.wur.n

Scandinavian Plant Physiolog2125 AuStavang

http://www.spps.f

Plant Organellar Signaling31 August 03

Primoste

http://www.plant-organellarferences in 2011

gy 2011t, 2011is, USAorg/meetings

ct-Plant Interactionsst, 2011Netherlands/SIP+Meeting+2011/

ociety (SPPS) Meeting 2011st, 2011Norwayi-bin/Meetings.pl

m Algae to Higher Plantsptember, 2011Croatia

-signaling.eu/index.html

411

Fire as an evolutionary pressure shaping plant traitsAdaptation to fireAdaptive traitsAdaptive traits as adaptations to fireResproutingSerotinyHeat-shock triggered germinationGermination triggered by combustion chemicalsFlammability as a fire adaptation

ConclusionAcknowledgmentsReferences