You are on page 1of 6

International Association for Ecology

Seed Size and Germination Response: A Relationship for Fire-following Plant Species Exposed to Thermal Shock Author(s): M. E. Hanley, J. E. Unna and B. Darvill Source: Oecologia, Vol. 134, No. 1 (Jan., 2003), pp. 18-22 Published by: Springer in cooperation with International Association for Ecology Stable URL: http://www.jstor.org/stable/4223469 . Accessed: 06/06/2013 00:52
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Springer and International Association for Ecology are collaborating with JSTOR to digitize, preserve and extend access to Oecologia.

http://www.jstor.org

This content downloaded from 137.219.5.13 on Thu, 6 Jun 2013 00:52:26 AM All use subject to JSTOR Terms and Conditions

Oecologia (2003) 134:18-22 DOI 10.1007/s00442-002-1094-2 __0a M S

M. E. Hanley *J. E. Unna * B. Darvill

Seed size and germinationresponse: a relationshipfor fire-following plant species exposed to thermalshock
Received: 22 December 2001 / Accepted: 26 September2002 / Published online: 19 October 2002 ? Springer-Verlag2002

Abstract Thermal shock is well known to be an important stimulus for the germination of soil-stored seeds in fire-prone plant communities. Nevertheless, while the overall germination response of different species is known to vary, the interaction between seed size and germination to a range of thermal-shock temperatures is poorly understood. This interaction may be important in regulating post-fire plant community establishment, since larger seeds are able to emerge from deeper within the soil profile than smaller seeds, and are therefore likely to be insulated against high above-ground temperaturesby a deeper soil covering. In this experiment we examined how germination of eight co-occurring Western Australian fire-followers was influenced by thermal shock, and whether germination was significantly correlated with seed size. We found that small-seeded species not only showed enhanced germination at higher temperatures, but that their ability to germinate at higher temperatures was also greater than that displayed by larger-seeded species. These findings suggest that while seed size may be a useful general predictor of post-fire recruitment success, under different fire regimes the interaction between seed size, maximum seedling emergence depth, and the ability to withstand different thermal-shock temperatures is complex and may confound recent predictive models.
. .

germination arises in response to a series of cues, including fire-stimulated seed release (Lamont et al. 1991), changes in light regime (Bell et al. 1999), and the release of nitrogen compounds (Keeley and Fotheringham 1998). However, by far the most widely studied phenomenon is that of the effect of thermal shock on the germination of soil-stored seeds. A host of experiments have shown that the high temperatures experienced by seeds during fire are important in order to trigger germination (Keeley 1991; Herranz et al. 1998; Tieu et al. 2001). Seed dormancy in firefollowers may be enforced by the possession of an impermeable seed coat (Cavanagh 1987), or by modifications to the structure of the seed coat (McKee et al. 1977). Germination often occurs only when the thermal shock associated with fire scarifies the seed coat allowing water and gas exchange (Thanos and Georghiou 1988). These adaptations not only prevent germination during the unfavourable period between fires, they result in the formation of long-lived seed banks able to take advantage of the immediate post-fire environment. However, the germination response of soil-stored species to different thermal-shock temperatures varies from species to species. For instance, in a study of cooccurring Western Australian species, Hanley and Lamont (2000) found that maximum germination occurred at Keywords Emergence Fabaceae Fire ecology - Heat temperatures ranging between 60?C and 120?C, depending upon the species examined. Not only does this shock - Post-fire regeneration variability in germination response have an important relationship with spatial heterogeneity in fire intensity on post-fire species regeneration (Herranzet al. 1998), it may Introduction also interact with the ability of seeds of different sizes to The role of fire in stimulating germination and seedling emerge from within the soil profile. It is well known that seedlings of large seeded species regeneration in plant species from Mediterranean-climate vegetation worldwide is well known. Increased post-fire can emerge from greater soil depths than those of smallerseeded species (Gulmon 1992; Jurado and Westoby 1992; Jurik et al. 1994). Yet despite a large literature devoted to M.E. Hanley (S) - J.E. Unna * B. Darvill Biodiversity and Ecology Division, School of Biological Sciences, the effects of thermal shock on germination, it has only University of Southampton,Bassett Crescent East, Southampton, recently been suggested that seed size may be correlated S016 7PX, UK with fire intensity (thermal-shock temperature) and depth e-mail: M.E.Hanley@soton.ac.uk of emergence for fire-followers. Bond et al. (1999) argue
Fax: +44-2380-594269

This content downloaded from 137.219.5.13 on Thu, 6 Jun 2013 00:52:26 AM All use subject to JSTOR Terms and Conditions

19

thatsince soil temperature duringfire decreasesmarkedly dishes were covered with lids and randomly arrangedin a dark with depth, so small seeds are likely to be killed at the incubator set at 15?C. These are considered to be optimum germinationconditions for most Western Australianpost-fire colosurface, leaving larger seeds to germinatefrom deeper nizers (Bell 1994). Dishes were examineddaily for 40 days and any within the soil profile. Although this prediction was seeds that had germinatedwere counted and removed. Germination supportedby their own observationson South African was deemed to have occurredwhen the radiclewas visible. Total percentage germination for each species across temperfynbos species, Bond et al. (1999) highlighted several aturetreatmentswas comparedby one-way ANOVA following arcfactorsthat may act to complicatethis relationship. One sine transformation. Differences between germination rates in possibilitythey did not consider,however, was that the unheated controls and heat-treated seeds were located by using germinationresponse of fire-followers might itself be Tukey tests. Having fitted response curves (Ridgman 1975) to the influenced by their ability to withstand different soil percentage germination data for heat-treatedseeds, we estimated the temperatureat which maximum germinationoccurs (Top,). The temperatures duringfire. equation describing the most closely fitting response curve was Because of the inverse relationship between seed size differentiatedand set equal to zero - the value of the gradientat a and the ability of seedlings to emerge from increasing turningpoint. Resulting equations were solved, using the Newtondepth (Juriket al. 1994), there is in fact likely to be a Raphson iterative method (Finney and Thomas 1994). We then values obtainedin this way against mean seed mass strong selection pressure in favour of small-seeded plotted the Topt to examine the relationship between seed size and germination species being able to withstandhighertemperatures than response to thermal-shock temperature. We also estimated the larger-seeded species. There is some circumstantial highest temperature at which 50% of maximum germination evidence to support this hypothesis. In Hanley and occurred(T+50%) by substitutingy=(T,pt)/2into the response curve Lamont's (2000) study, maximum germinationof the equation. These values were also plotted against mean seed mass.
small-seeded species Trymalium ledifolium occurred

at 120?C,while the germination following heat treatment


of larger-seeded species (Acacia pulchella and Daviesia

Results

cordata)peakedat much lower temperatures (100?Cand 800Crespectively).Howeverthereare no accountsin the When compared with unheated controls, thermal shock literaturethat detail the relationshipbetween seed size had a significant stimulatory effect on the germination and germination ratesover a largerangeof temperatures. response of seven species (Table 2). The precise pattern The aim of this study was to examine the relationship of germination response, however, varied markedly betweenseed size, and optimumthermal-shock germina- between individual species. The small-seeded species tion temperature for co-occurring WesternAustralian fire- (Chorizema ilicifolium, Gompholobium knightianum and Sphaerolobium vimineum) germinated well following following plantspecies.

Materialsand methods
Seeds of eight Fabaceae species (supplied by Nindethana Seeds, Albany, WA), native to the Darling Range 20 km east of Perth (31057'S, 115?51'E), were collected from this area during autumn 2000. Like many areas of Western Australia,the vegetation of the Darling Range is prone to frequentepisodes of naturallyoccurring fires. The species were selected on the basis that they co-occurred and covered a range of seed sizes (Table 1). Mean seed mass was determinedby weighing 100 seeds of each species. Seeds were dry heated in an oven for 5 min at 600, 700, 800, 900, 1000, 1100 and 120?C and subsequentgermnination comparedwith those of seeds stored at ambient temperature (ca. 20?C). To avoid problemsassociatedwith pseudoreplication of heat-shocktreatments we heated seeds in single replicate batches (Morrisonand Morris 2000). After heat treatment,the seeds were germinatedin 15-cmdiameterPetri dishes containingtwo layers of Whatmanno. 1 filter paper soaked with 10 ml of distilled water. Each species/treatment groupconsistedof six Petridishes,each dish containing15 seeds. The

treatment at higher temperatures, with G. knightianum and S. vimineum in particular showing strong germination within the highest temperaturegroup. The germination of larger-seeded species by contrast, responded poorly to temperatures in excess of 100?C. Two of these (Hovea and Kennedia) were the only species that failed to germinate at the highest temperature treatment. These general patterns are also shown by the response curves fitted to the percentage germination data for each species (Fig. 1). We obtained very close fits (P<0.01) for the third- and fourth-orderresponse curves fitted for each species, facilitating reliable estimation of Topt. When Topt was plotted against (loglo) mean seed mass (Fig. 2) we found a significant inverse relationship. A similar result was obtained when we plotted the temperature at which maximum germination occurred (Tmax) against mean seed mass (Fig. 2). Both results show that in comparison with larger-seeded species, the germination of small-seeded

Table 1 Details of the eight


Western Australian Fabaceae

Species Bossiaea aquifolium Chorizemailicifolium Daviesia horrida Gastrolobiumspinosum Gompholobiumknightianum Hovea chorizemifolia Kennedia coccinea
Sphaerolobium vimineum

Fire response syndrome Seeder Resprouter Resprouter Seeder Seeder Resprouter Seeder
Resprouter

Mean seed mass (mg) 18.5 1.8 23.7 10.3 2.0 29.3 29.5
1.1

species used to determinethe relationshipbetween mean seed size and the thermal-shock temperature calculated to facilitate optimum germination. Nomenclaturefollows Marchant et al. (1987)

This content downloaded from 137.219.5.13 on Thu, 6 Jun 2013 00:52:26 AM All use subject to JSTOR Terms and Conditions

20 Bossiaea
y = 6E-05x4 - 0.02x3 + 2.66x2 - 150x + 3105 r =0.95 y = - 0.002x3 + 0.539x2 - 44.Ox + 1152 r 0.98
IT .1 1 I 3 2

100 80 60

Gompholobium
y = - 12.2x + 102, r2= 0.51
100

0.047 F = 6.2216 P =

40
20 0 100 80 60

90

80
T opt

Chorizema

Hovea
y = 3E-05x4 - 0.01x3 + 1.38x2 - 74.7x + 1518 r =0.97

120 0)

a)110
100

y=- 11.8x+ 102, r2=0.56 F = 7.4716 P = 0.034

0 co

40

20
0 100 I

3 0.001x +0\249x 2 + -18.4x 486 r209 I I I

E
? C
c

Daviesia
y

Kennedia
y=
+

E a) F0.002x3 r209

90 .

a)

80 60 40

~~~~-

6E-05x4 - 0.02x2 + 3.09x2 176 +3681

0.646x2 59.8x - 1709

80 T + 50%
130 A
\

r2 =0.99

y=-15.0x+124,
F=

r20.80

20*
0

120
110

P = 0.003 ~~~~~~23.916

~A
A

100 80 60 60 40 20 0

Gastrolobium
y = - 0.0004x3 + 0.095x2 -6.62x + 142 9 2 r2 0_99 oggy

Sphaerolobium
100
_.

A 0.5 1.0 1.5

0.0006x3

+JQx + 0.107x

-4.79x + 84.0 r2 0.93

Mean seed mass (mg log10)

60

70

80

90

100 110 120

60

70

80

90

100 110 120

Temperature(OC) Fig. 1 The effects of thermal shock on mean (?SE) total percentage germination of eight Western Australian Fabaceae species. Response curves show significant (P<0.01) relationshipsbetween treatment(5 min heating at 60, 70, 80, 90, 100, 110 or temperature 120?C) and mean germinationrate

Fig. 2 The relationship between pre-emergence thermal-shock temperatureand mean seed mass on germinationof eight Western AustralianFabaceae species. Graphsshow the relationshipsfor the thermal-shocktreatmentpromoting maximum germination(Tmax), the estimated temperatureat which optimum germinationoccurs occurs germination(T+50%)
(Topt), and the estimated temperature at which 50% of total

Table 2 The results of one-way ANOVA (all species P<0.001) to Fabaceae species. Multiple-range (Tukey) tests show significant determine the effect of heat shock (5 min at temperaturesshown) (P<O.05) increases in the germination of treatmentgroups when on total percentage germination of eight Western Australian comparedwith unheatedcontrols, 'Idenotes significant increase Species Ogne-way ANOVA
(F7, 40)

Tukey tests 600C 700C 800C 900C 100?C I I0?C 1200C

Bossiaea aquifolium Chorizemailicifolium Daviesia horrida Gastrolobiumspinosum Gompholobiumknightianum Hovea chorizemifoliaa Kennedia coccineaa Sphaerolobiumvimineum
a

10.50 19.28 11.29 6.74 30.63 10.02 22.57 19.14

'1

Germinationsignificantly reduced when heated to 1200C

species is promoted by relatively high thermal-shock Discussion temperatures. A significant inverse relationship was also was plotted against mean seed mass The inverse relationship between seed size and germinaapparent when T+50% (Fig. 2). This finding suggests that small-seeded species tion response reported here indicates that the prediction of are also more resistant to higher thermal-shock temper- Bond et al. (1999) that "more seedlings of large-seeded species should emerge from high fire intensities because atures than larger-seeded species.

This content downloaded from 137.219.5.13 on Thu, 6 Jun 2013 00:52:26 AM All use subject to JSTOR Terms and Conditions

21

they can emerge from greaterdepths than small-seeded seeds within the soil profile, variations in fire intensity species" may be complicatedby species-specificdiffer- may result in a highly varied and unpredictable pattern of ences in response to thermal-shocktemperatures. It is post-fire seedling emergence. Despite our caveat to the predictive model of Bond et clear fromour resultsthatseed mortality of larger-seeded species is greaterat high temperatures (i.e. both Hovea al. (1999), we would like to emphasise the fact that it does in the 120?Ctreatment), represent an important step towards our understanding of andKennedia failedto germinate and that these species are likely to benefit from their plant community responses to changing fire regimes. abilityto emergefrom deeperwithinthe soil profilethan Notwithstanding the clear interaction between speciessmall-seededspecies. However, it is also apparentthat specific responses to fire and changing fire regime, it is smaller-seededspecies are more tolerantof the higher only recently that ecologists have become more aware of temperaturesfrequently experienced nearer the soil the importance of linking the results of laboratory-based surfaceduringfire. Therefore,while the germination of germination studies with observations of post-fire seedlarge-seeded species at depthmay indeedrespondwell to ling emergence in the field (Ne'eman et al. 1992; Herranz relativelyintense(hot) fires (Bondet al. 1999),ourresults et al. 1998; Bond et al. 1999; Tieu et al. 2001). Since suggestthatsmaller-seeded species nearerthe soil surface many fire-prone ecosystems are now intensively mando not necessarilyalways suffer a reductionin germina- aged, accurate prediction of individual plant species tion. responses to prescribed fires is all the more pressing. This As Bond et al. (1999) point out, the exact relationship study shows that seed size may be linked to speciesbetweenseed size andemergencedepthfor fire-followers specific variation in germination response to fire, a is likely to depend on several factors. It is also well relationship that should be taken into account if ecologists understoodthat germinationand establishmentsuccess are to manage successfully species diversity in fire-prone for fire-followersrepresentsa fine balance between the plant communities. survivalof high temperatures and actually experiencing temperatures high enough to break dormancy(Whelan Acknowledgements We thank Dr. Michael Fenner, and two 1995). Because soil acts as an effective insulator,this anonymous referees for their helpful comments and suggestions. balanceis affectedgreatlyby the depthat which a seed is buried (Bond et al 1999). Few studies, however, have explicitly examined the relationshipbetween soil depth References and temperature duringfire (Whelan1995). Furthermore, factors such as soil moisture content and soil type, Bell DT (1994) Interactionof fire, temperatureand light in the germination response of 16 species from the Eucalyptus togetherwith fire intensity,make the accurateprediction marginata forest of southwestern Western Australia. Aust of soil temperature profiles extremelydifficult (Whelan J Bot 42:501-509 1995;Bond et al. 1999). Nevertheless,it does appearthat Bell DT, King LA, PlummerJA, Taylor SK (1999) Ecophysiological effects of light quality and nitrate on seed germinationin even at relativelyshallowdepths,drop peaktemperatures, species from Western Australia.Aust J Ecol 24:2-10 quickly from those experienced at the soil surface Bond WJ, Honig M, Maze KE (1999) Seed size and seedling et al. 1992; Odion and Davies 1999). (Bradstock emergence: an allometric relationship and some ecological In very high intensityfires where soil surfacetemperimplications. Oecologia 120:120-132 aturesmay exceed 500?C (Odion and Davies 1999), soil Bradstock RA, Auld TD, Ellis ME, Cohn JS (1992) Soil temperatures during bushfire in semi-arid, mallee shrublands. Aust temperatures may be hot enough to kill all but the most J Ecol 17:433-440 deeply buried seeds. In these situations,the seed size/ Cavanagh T (1987) Germination of hard seeded species (Order emergencedepth relationship predictionsof Bond et al. Fabales). In: LangkampP (ed) Germinationof Australianplant (1999) probably hold true. Fires of lower intensity, seed. Inkata,Melbourne,pp 58-70 however, may create a soil temperatureprofile that Finney RL, Thomas GB (1994) Calculus, 2nd edn. AddisonWesley, Reading, Mass. stimulatesthe germinationof small seeds near the soil Gulmon SL (1992) Patterns of seed germination in Californian surface, as well as that of larger-seededspecies lower serpentinegrasslandspecies. Oecologia 89:27-31 down. Why though should small-seeded species defer Hanley ME, LamontBB (2000). Heat shock and the germinationof Western Australianplants species: effects on seeds of soil- and untilrelativelyhigh soil-surface germination temperatures canopy-storedspecies. Acta Oecol 21:315-321 arereached? The answermay lie with the fact thatsurface JM, FerrandisP, Martinez-SanchezJJ (1998) Influence of temperaturesbetween fires can often exceed 70?C Herranz heat on seed germinationof seven Mediterranean Leguminosae (Keeley and Fotheringham2000). If germination of species. Plant Ecol 136:95-103 small-seeded species occurred under these conditions, JuradoE, Westoby M (1992) Seedling growth in relation to seed size among species of arid Australia.J Ecol 80:407-416 competitionfrom neighbouring maturevegetationwould Jurik TW, Wang SC, VanderwalkAG (1994) Effects of sediment make successful seedling establishmentvery unlikely load on seeding emergence from wetland seed banks. Wetlands (Whelan1995).Even in the hottestfires,however,thereis 14:159-165 usuallya largedegreeof heterogeneity in the distribution Keeley JE (1991) Seed germinationand life history syndromes in the Californianchaparral.Bot Rev 57:81-1 16 of fuel, and therefore patchiness in soil temperatures (Ne'emanet al. 1992;Seguraet al. 1998).Combined with Keeley JE, FotheringhamCJ (1998) Mechanismsof smoke-induced seed germinationin a post-fire chaparralannual.J Ecol 86:27the differential germination responses of co-existing 36 species to heat-shock temperatures, and the depth of

This content downloaded from 137.219.5.13 on Thu, 6 Jun 2013 00:52:26 AM All use subject to JSTOR Terms and Conditions

22 Keeley JE, Fotheringham.CJ (2000) Role of fire in regeneration from seed. In: FennerM (ed) Seeds: the ecology of regeneration in plant communities. CABI, Wallingford,pp 311-330 Lamont BB, Le Maitre DC, Cowling RM, Enright NJ (1991) Canopy seed storage in woody plants. Bot Rev 57:277-317 Marchant NG, Wheeler JR, Rye BL, Bennett EM, Lander NS, MacFarlane TD (1987) Flora of the Perth region. Western AustralianHerbarium,Western Australia McKee GW, Pfeiffer RA, Mohsenin NN (1977) Seed coat structure in Coronilla varia and its relationto hardseed. Agron J 69:5358 MorrisonDA, MorrisEC (2000) Pseudoreplication in experimental designs for the manipulationof seed germination treatments. Aust Ecol 25:292-296 Ne'eman G, Lahav H, Izhaki I (1992) Spatial patternof seedlings 1 year after fire in a Mediterraneanpine forest. Oecologia 91:365-370 Odion DC, Davies FW (2000) Fire, soil heating and the formation of vegetation patternsin chaparral. Ecol Monogr 70:149-169 in biology: an introduction to RidgmanWJ (1975) Experimentation design and analysis. Blackie, London Segura AM, Holmgren M, Anabal6n JJ, Fuentes ER (1998) The significance of fire intensity in creating local patchinessin the Chilean matorral.Plant Ecol 139:259-264 Thanos CA, Georghiou K (1988) Ecophysiology of fire stimulated
seed germination in Cistus incanus ssp. creticus (L.) and C. salvifolius (L.). Plant Cell Environ 11:841-849

K (2001) The Tieu A, Dixon KW, Meney KA, Sivasithamparam interactionof heat and smoke in the release of seed dormancy WesternAustralia.Ann Bot in seven species from southwestern 88:259-265 Whelan RJ (1995) The ecology of fire. Cambridge University Press, Cambridge

This content downloaded from 137.219.5.13 on Thu, 6 Jun 2013 00:52:26 AM All use subject to JSTOR Terms and Conditions

You might also like