Early post-fire plant succession in Peninsula Sandstone Fynbos: … · 2012-02-24 · Early...

www.elsevier.com/

Available online at www.sciencedirect.com

South African Journal of Botany xx (2011) xxx–xxx

SAJB-00643; No of Pages 10

www.elsevier.com/locate/sajb

Early post-fire plant succession in Peninsula Sandstone Fynbos: The firstthree years after disturbance

M.C. Rutherford a,b,⁎, L.W. Powrie a, L.B. Husted a, R.C. Turner c

a Applied Biodiversity Research Division, Kirstenbosch Research Centre, South African National Biodiversity Institute, Private Bag X7,Claremont 7735, South Africa

b Department of Botany and Zoology, Stellenbosch University, Private Bag X1, Matieland 7602, South Africac School of Botany and Zoology, University of KwaZulu-Natal, Private Bag X01, Scottsville, Pietermaritzburg 3209, South Africa

Received 16 July 2010; accepted 10 February 2011

Abstract

The early post-fire plant succession in fynbos vegetation in the Mediterranean-type climate area of South Africa was studied. Relatively littlehas been published on this early stage of plant succession in fynbos. Annual sampling over the first three post-fire years confirmed a steady, butrelatively slow increase in plant canopy cover of shrubs and graminoids (mainly Restionaceae), whereas cover of geophytes and other herbspeaked in the first year and declined significantly, thereafter. Cover of annual plants increased each year, which may relate to the persistence of arelatively open vegetation cover by the third year. The responses of reseeder and resprouter species of the Restionaceae to the post-fireenvironment appeared to be habitat dependent. Cover of the reseeders increased rapidly in seep areas, but their recovery was distinctly delayed indryland areas outside the seeps. Re-establishment of the many reseeder Erica species appeared to be delayed until the second post-fire year. Seedbanks of these species were possibly negatively impacted by the fire, and required dispersal of seed from unburnt areas for recruitment. In contrastto some current generalisations, species richness appeared to increase after the fire; less certainly from the first to the second year, but morecertainly from the second to the third year. Therefore, this study does not support a short-term monotonic decline in species richness after fire infynbos.© 2011 SAAB. Published by Elsevier B.V. All rights reserved.

Keywords: Burn; Canopy cover; Diversity; Ephemeral; Mediterranean-type; Reseeder; Resprouter; South Africa

1. Introduction

One of the many fire-prone vegetation types of the world(Bond and Keeley, 2005) is the fynbos of the Cape FloristicRegion — CFR (Bond and Van Wilgen, 1996). The region ischaracterized by exceptional levels of plant diversity (Van derNiet and Johnson, 2009) and endemism, that is unparalleledglobally (Cowling et al., 2009). Response of fynbos to fire iscomplex, and unique compared to other biomes of southernAfrica in that fynbos composition can be drastically altered aftera single fire (Bond, 1997). Fire is thus a major disturbance

⁎ Corresponding author at: Applied Biodiversity Research Division, KirstenboschClaremont 7735, South Africa. Tel.: +27 217998702; fax: +27 217976903.

E-mail address: [email protected] (M.C. Rutherford).

0254-6299/$ - see front matter © 2011 SAAB. Published by Elsevier B.V. All righdoi:10.1016/j.sajb.2011.02.002

Please cite this article as: Rutherford, M.C., et al., Early post-fire plant successionJ. Bot. (2011), doi:10.1016/j.sajb.2011.02.002

factor that shapes the structure and composition of fynbos.Consequently, much of the research effort in fynbos has focusedon the effects of fire and on how fire drives succession (Cowlinget al., 1997). Studies include those on rates of recovery ofvegetation cover, temporary importance of fire ephemerals,differing sensitivity of resprouter and obligate reseeder plants tofire return intervals, and trends in plant diversity (Kruger, 1979;Kruger and Bigalke, 1984; Bond and Van Wilgen, 1996;Cowling et al., 1997).

Post-fire studies in fynbos have generally found a decline inplant diversity in the long-term, typically over periods of longer

Research Centre, South African National Biodiversity Institute, Private Bag X7,

ts reserved.

in Peninsula Sandstone Fynbos: The first three years after disturbance, S. Afr.

mailto:[email protected]

http://dx.doi.org/10.1016/j.sajb.2011.02.002





Fig. 1. Location of (1) study area in relation to other sandstone or sand fynbosstudy sites in the vicinity that have been monitored annually for at least threegrowth seasons since a fire. These are (2) western Table Mountain (Adamson,1935), (3) Jakkalsrivier, Grabouw (Kruger and Bigalke, 1984), (4) Swartboskloof,Stellenbosch (Van Wilgen and Forsyth, 1992), and (5) Pella Nature Reserve,Atlantis (Hoffman et al., 1987).

2 M.C. Rutherford et al. / South African Journal of Botany xx (2011) xxx–xxx

than a decade (Cowling and Pierce, 1988; Cowling and Gxaba,1990; Privett et al., 2001). This long-term decline in plantdiversity after fire appears to also apply in some Mediterranean-type shrublands on other continents (Kruger, 1983; Bond andVan Wilgen, 1996). However, changes in plant diversity infynbos in the early post-fire years are uncertain. In particular,studies that determine changes in species richness over the firstthree post-fire years in fynbos on sandstone-derived or sandsubstrates are few and contradictory. Hoffman et al. (1987)found a significant increase in species richness in lowland sandfynbos from the first to the third year after fire. Similarly,Adamson (1935) reported almost a doubling of the number ofspecies in fynbos vegetation of the lower western TableMountain over the same post-fire period. By contrast, Cowlingand Pierce (1988) indicated a decline in species richness,effectively from the second to third year after fire in coastal dunefynbos in the Eastern Cape Province. Despite these uncertaintiesregarding the early post-fire years in fynbos, generalizationspersist that there is a monotonic decline in the number of plantspecies with increasing stand age after fire (Bond and VanWilgen, 1996), or after an undefined ‘immediate post-fire phase’(Cowling et al., 1992).

This study investigated initial, short-term succession dynamicsof plants for a period of three years after fire in sandstone fynbos.Apart from characterizing short-term responses of species andspecies groups to fire, we also tested the claim that there is amonotonic decline in species richness after fire in fynbos.

2. Study area

The study was conducted after a wildfire at Olifantsbos in theCape of Good Hope section of the Table Mountain NationalPark (Fig. 1). The site is part of a sample area of the BIOTAmonitoring programme (Haarmeyer et al., 2010). The area fallswithin Peninsula Sandstone Fynbos (Rebelo et al., 2006) on theexceptionally flora-rich Cape Peninsula (Simmons and Cowling,1996). The landscape of the study site takes the form of a brokenundulating plain with a number of large boulders. Altitude isbetween 60 and 80 m a.m.s.l. Mean annual precipitation fallsmainly inwinter and amounts to 573 mm, based on the rain stationof Klaasjagersberg that is 3 km from the site. Winter and summertemperatures are ameliorated by the nearby Atlantic Ocean. Thearea is extremelywindywith a high frequency of gale forcewinds.The very infertile soils are derived from sandstone of the TableMountain Group. Soil depth varies, but shallow lithosols andsurface rock are common. In the final year of study, vegetationstructure comprised a dominant lower stratum of 0.2 to 0.4 m tallRestionaceae (e.g. Hypodiscus aristatus) and low shrubs (e.g.Diastella divaricata). Scattered 2 to 3 m tall shrubs (Mimetesfimbriifolius and Leucospermum conocarpodendron) were con-spicuous but not common. A few seep areas, which have standingwater inwinter but dry surface soil in summer, also occurred in thestudy area. These seeps were dominated by Restionaceae plants,mainly Elegia cuspidata, that were 0.8 to 0.9 m tall.

The vegetation of the area studied was burned by the wildfireof April 2007, and had been burned 17 years previously, inJanuary 1991. Before this, fire histories differed. About half the


area had burned five years earlier in February 1986 whereas theother half had not been burned for at least the 16 years for whichreliable fire records had been kept for the area (G.G. Forsyth,personal communication). The average age of vegetation in theTable Mountain National Park is 13.5 years (Forsyth and VanWilgen, 2008) and 10 to 13 years for most of the larger areaselsewhere in fynbos (Van Wilgen et al., 2010). The 2007 burnwas a headfire with flame lengths between two and three metres.South-east winds were in excess of 40 km per hour (C. Dilgee,personal communication). The fire was considered to be a cleanburn that killed all above ground plant parts except for mostof the large individuals of thick-barked M. fimbriifolius andL. conocarpodendron.

3. Methods

Data were collected from sixteen 10×10 m sample plots laidout randomly within the BIOTA sampling grid of 1 km2

(Haarmeyer et al., 2010). Thirteen plots fell in dryland areas andthree plots in seeps. The 13 dryland plots could be clearlydivided according to prior fire history: seven plots with a five-year prior fire return period (short fire interval) and six plotswith a period of more than 16 years (long fire interval) (G.G.Forsyth, personal communication). Data recordings for the firstand second post-fire years were carried out in January 2008 and



a

b

Fig. 2. Post-fire changes in mean canopy cover of a) all plants and b) growthforms. Bars indicate standard error.

3M.C. Rutherford et al. / South African Journal of Botany xx (2011) xxx–xxx

January 2009 respectively. Recordings for the third post-fireyear were done in December 2009 and January 2010. Thisapproach of monitoring the same plots differs from thatcommonly used to analyse post-fire vegetation changes infynbos, namely, comparing independent samples of differentlyaged stands (e.g. Privett et al., 2001). Although the latterapproach is probably effective in longer term studies, this maynot apply in the shorter term where spatial heterogeneity mayconfound detection of changes.

Canopy cover of each plant species was estimated. In addition,a count of individual plants of each species was made, except forgraminoid forms (hereafter applied to include Restionaceae,Poaceae and Cyperaceae) and a few other species in which plantindividuals were difficult to define. Most geophytic specieswere not counted. Counts of plants in dense populations, withsometimes over 1000 plants per 100 m2, were approximate.Mature, surviving plants and seedlings ofM. fimbriifolius and L.conocarpodendron were recorded separately. Plant species wereidentified at the Compton and Bolus Herbaria. Fynbos speciesidentification in the field over the first three post-fire years waschallenging, particularly in the first year. For example, in groupssuch as Restionaceae, seedlings appear very different from adultplants in their first year and only in the third year do plants reachthe typical adult morphology, i.e. without finely-branched culms(Haaksma and Linder, 2000). Also, a number of resprouterspecies could not be separately identified in the first post-fire year.

Twoobservers were used: one for the first post-fire year and theother for the remaining two years. Since plant cover estimates canvary according to observer (Milberg et al., 2008; Bergstedt et al.,2009) due to different approaches or definitions (Fehmi, 2010),the two observers, independently of each other, were asked toestimate, in the same way each had done before, the plant canopycover of a range of species and growth forms on the study site.Close correspondence between the two observers was found incover estimates for graminoid species. For other species, estimatesin the lower range of cover values diverged; consequently, a linearregression was applied to calibrate the cover estimates of the oneobserver against those of the other observer.

Data were not normally distributed and, therefore, the non-parametric, two-tailed Wilcoxon Signed-Ranks test for repeatedmeasurements was used (XLSTAT Version 2010.4.01 byAddinsoft) for each species and species group. Differencesbetween areas, according to prior fire intervals, were analysedusing the non-parametric, two-tailed Mann–Whitney U-test forindependent samples.

4. Results

4.1. Plant growth forms and groups

The post-fire recovery of mean plant canopy cover wasrelatively rapid, yet less than 50% cover was attained by the thirdyear (Fig. 2a). A non-significant decline in cover during the secondyear was followed by a significant increase in the third year(pb0.01). Cover regeneration patterns differed according to growthform. Changes in cover of shrubs and graminoids closely paralleledthat of total plant cover, also with significant increases in the third


year (pb0.05 for shrubs, pb0.01 for graminoids) (Fig. 2b). Bycontrast, cover of herbs and geophytes peaked in the first year.Subsequent decline in their cover was significant (pb0.05 for herbsin the second year, pb0.01 for geophytes in second and thirdyears). Cover of annual plants increased significantly from 0.008%in the first year to 0.25% in the third (pb0.01). Annual plants wererare, and comprised only 3% of the species recorded.

Patterns of post-fire development of canopy cover of mainfynbos families were distinctly different from one another. Changein cover of Restionaceae closely followed that of graminoids(Fig. 3) with a significant decline in the third year (pb0.01).However, cover of Proteaceae peaked in the first year, declinedrapidly and significantly (pb0.01) in the second year, and appearedto stabilise in the third year. By contrast, cover of the species ofEricaceae showed a slow, but significant increase (pb0.001 in thethird year), and reached a similar level to that of Proteaceae (Fig. 3).

The responses of reseeder and resprouter species ofRestionaceae to the post-fire environment appeared habitatdependent. Canopy cover of reseeders increased rapidly in theseep areas in the first year (Fig. 4b), but their recovery wasdelayed in the dryland areas (Fig. 4a). Cover of resproutingspecies showed no such delay in either habitat.

4.2. Plant species

A total of 206 plant species were recorded in the study. Meanspecies richness was observed to increase steadily after fire(Fig. 5). Mean richness increased significantly between the



Fig. 3. Post-fire changes in mean canopy cover for three typical families offynbos vegetation. Bars indicate standard error.

Fig. 5. Post-fire changes in mean species richness. Bars indicate standard error.


second and third years (pb0.001). It is known that richness forthe first year was an underestimate (see Discussion), whichtherefore renders statistical significance of the change inrichness from the first to second years questionable. Ericaspecies constituted 2% of mean species richness in the first year,and rose to 10% by the third year. Only in two seep plots didspecies richness start to decline in the third year (each by lessthan 15%).

The more frequent plant species (those that occurred in morethan two thirds of the sample plots — Table 1) are listed belowaccording to their patterns of response to the post-fireconditions. Species that started to establish in the first yearand that increased significantly in the second or third years were

a

b

Fig. 4. Post-fire changes in mean canopy cover of reseeder and resprouterRestionaceae in a) dryland and b) seep areas. Four species of Restionaceae couldnot be assigned to a regeneration type but collectively constituted an average ofonly 0.07% canopy cover and were excluded from the analysis. Bars indicatestandard error.


Erica exleeana (pb0.05 for cover and density in the secondyear — Tables 1 and 2), Ischyrolepis capensis (pb0.01 forcover in the second year), Lobelia setacea (pb0.01 for coverand density in the second year), Microdon dubius (pb0.05 forcover in the third year and density in the second year) andSyncarpha vestita (pb0.05 for cover in the third year). Therewas a significant increase in density of Adenandra villosa and incover of Isolepis marginata only from the first to the third years(pb0.05 for both species). Species that showed no significantdifferences in cover or density (pb0.05) from the first yearonwards were Indigofera glomerata, Osteospermum polyga-loides, Roella triflora, Tetraria bromoides and Ursiniapaleacea.

The typical fire ephemeral response was best shown byItasina filifolia which was observed to produce a large numberof seedlings in the first year, with mean plant density decreasingsignificantly by two orders of magnitude and cover by 99% inthe second year (pb0.01 for density and cover). In the third yearboth density and cover remained at low levels. Other fireephemerals that declined significantly in the second and/or thirdyears were Corymbium africanum (pb0.05 for cover in thesecond year), Pseudoselago spuria (pb0.01 for cover in thethird year and for density in the second and third years) andTritoniopsis dodii (pb0.05 for density in the second year).

A species that was also observed to produce seedlings mostprominently in the first year was D. divaricata (densityN1000per 100 m2 Table 2) but such did not follow the pattern of fireephemerals. The significant decline in cover (pb0.05) anddensity (pb0.01) of D. divaricata in the second year wasfollowed in the third year by a significant increase in cover(pb0.05) while density remained at a low level. The decline indensity of R. decurrens was of intermediate magnitude(pb0.05) with no significant decrease in cover. A similarpattern was found in Leucadendron laureolum which after theinitial flush in the first year density and cover decreasedsignificantly in the second year (pb0.01 for density, pb0.05 forcover), with density in the third year falling significantly again(pb0.01) while cover tended to increase. Despite insignificantdeclines in density in Metalasia densa, cover increasedsignificantly in the third year (pb0.05).

The large flush of seedlings (densityN2000 per 100 m2) ofR. decurrens in the first year was followed by a significantdecline in the second year (pb0.05) but with no concomitantdecline in cover. There was no significant change in density and



Table 1Post-fire response patterns in canopy cover of species which occurred in more than two-thirds of the sample plots.

Species Canopy cover (%) Relative meancanopy covera

(%)Year 1 Year 2 Year 3

Mean S.E. Mean S.E. Mean S.E.

ShrubsAdenandra villosa 0.06 0.03 0.03 0.01 0.06 0.02 0.1Aspalathus retroflexa 0.09 0.09 1.00 0.37 1.14 0.44 2.1Diastella divaricata subsp. divaricata 3.95 1.50 0.56 0.14 1.21 0.42 5.4Edmondia sesamoides 0.00 – 0.05 0.02 0.42 0.18 0.4Erica exleeana 0.02 0.02 0.12 0.04 0.43 0.24 0.5Erica imbricata 0.00 – 0.06 0.03 0.21 0.06 0.3Gnidia tomentosa 0.30 0.11 0.05 0.01 0.07 0.03 0.4Indigofera glomerata 0.14 0.07 0.10 0.04 0.05 0.02 0.3Leucadendron laureolum 1.13 0.27 0.36 0.07 0.45 0.08 1.8Lobelia setacea 0.02 0.02 0.22 0.06 0.39 0.19 0.6Metalasia compacta 0.00 – 0.07 0.02 0.15 0.06 0.2Metalasia densa 0.29 0.11 0.07 0.03 0.31 0.11 0.6Microdon dubius 0.02 0.02 0.06 0.02 0.11 0.03 0.2Osteospermum polygaloides 1.00 0.46 0.39 0.14 0.50 0.18 1.8Prismatocarpus sessilis 0.15 0.07 2.70 1.87 1.45 0.80 4.0Pseudoselago spuria 1.10 0.57 0.28 0.14 0.04 0.01 1.3Roella decurrens 3.31 1.40 3.85 1.94 3.12 1.20 9.6Roella triflora 0.26 0.08 0.29 0.09 0.44 0.17 0.9Stoebe cyathuloides 0.00 – 0.14 0.04 0.77 0.44 0.9Struthiola dodecandra 0.00 – 0.16 0.04 0.52 0.23 0.6Syncarpha vestita 0.28 0.11 0.23 0.08 0.85 0.33 1.3Thesium capitellatum 0.00 – 0.04 0.01 0.04 0.01 0.1Thesium strictum 0.00 – 0.04 0.01 0.19 0.06 0.2Thesium sp. 0.00 – 0.15 0.05 0.32 0.06 0.4Ursinia paleacea 0.05 0.03 0.13 0.09 0.05 0.01 0.2

GraminoidsElegia stipularis 0.00 – 0.33 0.11 0.53 0.20 0.8Hypodiscus aristatus 1.59 0.92 0.68 0.19 1.13 0.28 3.2Ischyrolepis capensis 0.01 0.01 0.32 0.09 0.44 0.18 0.7Ischyrolepis cincinnata 2.29 0.63 0.85 0.22 1.65 0.38 4.5Isolepis marginata 0.01 0.01 0.03 0.01 0.22 0.08 0.2Merxmuellera decora * – 0.09 0.01 0.12 0.05 0.2Pentaschistis curvifolia 0.61 0.14 0.01 0.01 0.02 0.02 0.6Staberoha cernua * – 0.34 0.13 0.32 0.10 0.6Tetraria bromoides 0.29 0.25 0.28 0.13 0.29 0.15 0.8Tetraria cuspidata * – 0.23 0.06 0.31 0.07 0.5Tetraria microstachys * – 0.26 0.07 0.26 0.08 0.5Thamnochortus cf. fruticosus * – 0.09 0.08 0.30 0.18 0.4Thamnochortus lucens 2.50 0.34 1.27 0.36 1.68 0.39 5.1

HerbsCorymbium africanum 1.80 0.81 0.11 0.05 0.15 0.08 1.9Itasina filifolia 1.06 0.54 0.01 0.00 0.01 0.00 1.0

GeophyteTritoniopsis dodii 0.07 0.04 0.02 0.01 0.01 0.00 0.1

Remaining speciesb 12.94 4.93 12.89 5.51 21.61 6.33 44.5

* Usually resprouter, but not detected in year 1. a Mean over three years. b Species occurring in fewer than two-thirds of sample plots.


cover in the third year. Similarly, the density of Gnidiatomentosa declined in the second and third years (both pb0.05)without any significant change in cover.

A group of graminoid species showed similar trends butcover of only Ischyrolepis cincinnata decreased significantly inthe second year (pb0.01) then increased significantly in thethird (pb0.01). Cover of Pentaschistis curvifolia and Thamno-chortus lucens decreased significantly in the second year(pb0.01) but the increase in the third was not significant. Thedecrease in cover of Hypodiscus aristatus in the second year


was not significant but the increase in the third was significant(pb0.01). Densities of graminoid species were not recorded.

Two species appeared to peak later than those above, namelyin the second year. Thus significant increases were found in thesecond year in Aspalathus retroflexa (pb0.01 for cover,pb0.05 for density) and in Prismatocarpus sessilis (pb0.01for cover) before density declined significantly in the third year(pb0.05 for both species).

Species that were slow to regenerate and started only in thesecond post-fire year and that increased significantly from the



Table 2Post-fire response patterns in plant density of selected species which occurred in more than two-thirds of the sample plots.

Species Plant density per 100 m2

Year 1 Year 2 Year 3

Mean S.E. Mean S.E. Mean S.E.

ShrubsAdenandra villosa 4.4 2.6 7.5 3.0 7.4 2.7Aspalathus retroflexa 3.1 3.1 15.1 4.4 7.4 2.3Diastella divaricata subsp. divaricata 1248.1 430.5 102.3 23.2 80.6 26.5Edmondia sesamoides 0.0 0.0 6.5 2.3 18.2 8.1Erica exleeana 31.3 31.3 222.6 93.3 126.8 85.6Erica imbricata 0.0 0.0 225.8 87.2 281.0 94.1Gnidia tomentosa 318.8 136.6 59.8 18.6 27.1 10.6Indigofera glomerata 45.6 25.6 12.7 4.2 4.3 1.4Leucadendron laureolum 266.1 76.6 55.7 8.5 38.7 6.3Lobelia setacea 3.4 3.1 31.7 8.9 26.8 10.0Metalasia compacta 0.0 0.0 49.2 16.2 44.3 18.3Metalasia densa 232.8 107.8 27.1 9.4 42.5 19.4Microdon dubius 1.9 1.9 8.1 2.3 7.4 2.1Osteospermum polygaloides 553.4 243.4 60.3 16.9 76.7 33.0Prismatocarpus sessilis 56.9 26.9 66.3 50.2 0.6 0.4Pseudoselago spuria 126.9 40.5 32.0 13.2 3.6 1.2Roella decurrens 2162.5 1224.8 367.2 132.4 340.1 124.3Roella triflora 196.9 78.1 83.1 21.7 97.9 38.5Stoebe cyathuloides 0.0 0.0 49.2 16.4 52.3 14.4Struthiola dodecandra 0.0 0.0 129.9 37.3 236.1 79.9Syncarpha vestita 107.8 43.5 43.2 14.8 74.0 39.9Thesium capitellatum 0.0 0.0 9.0 3.1 5.4 1.6Thesium strictum 0.0 0.0 2.7 0.9 12.8 3.7Thesium sp. 0.0 0.0 13.2 4.2 30.5 12.3Ursinia paleacea 16.9 8.9 20.4 12.3 12.1 4.5

HerbItasina filifolia 724.7 376.4 2.6 1.3 5.0 3.0

GeophyteTritoniopsis dodii 23.1 12.3 3.3 0.9 1.8 0.6


second to the third year were Edmondia sesamoides (pb0.01 forcover, pb0.05 for density), Erica imbricata (pb0.01 for cover),Struthiola dodecandra (pb0.05 for cover) and Thesiumstrictum (pb0.01 for cover and density). Other species thatwere also not found in the first year but did not increasesignificantly in the third year were Elegia stipularis, Metalasiacompacta, Stoebe cyathuloides, Thesium capitellatum andThesium sp. Known sprouter species not recorded in the firstyear are indicated in Table 1. Of these, the significant increasein cover of Thamnochortus cf. fruticosus was marginal(p=0.044) in the third year.

Very few seedlings of M. fimbriifolius and L. conocarpo-dendron were recorded in the study.

The responses of very few species appeared to be linked to firehistory. Canopy cover of the annual, Helichrysum indicum, wassignificantly higher on the short fire interval plots in the secondyear (pb0.01) and in the third year (pb0.05) (Fig. 6a). In contrast,cover of the reseeder shrub,D. divaricata, was greatly reduced onthese plots in the first post-fire year (pb0.05) (Fig. 6b).

5. Discussion

Although the fire was regarded to be a clean burn, the verylow number of recorded seedlings of M. fimbriifolius and L.


conocarpodendron may indicate that the fire was not of aparticularly high intensity, according to observations by Bond etal. (1990). Nevertheless, occurring in autumn, the burn wouldhave been expected to maximize seedling production of manyother species (Bond et al., 1984; Le Maitre, 1988b; Van Wilgenet al., 1992). Moisture conditions were favourable forgermination and establishment of seedlings after the fire, as aresult of subsequent, regular occurrences of rain; also, the rainin the first winter season was above average (Fig. 7).Germination is likely to have substantially augmented theflush of cover of resprouting fire-ephemerals in the first post-fire year. Fire stimulated germination has been seen as anevolutionary response to increased availability of the resourcesof light, space and nutrients (Le Maitre and Midgley, 1992).However, post-fire flushes of nutrients have been shown to beof relatively short duration with availability of nitrogendecreasing rapidly after nine months and phosphorous levelsreturning to pre-fire levels after only four months (Cowling etal., 1997).

Those species with a high density of establishing plants inthe first post-fire year experienced a sharp decline in the secondyear. These include four fire ephemeral species found in thisstudy. Our results show that two other known fire ephemeralspecies appear to peak later, not before the third year. S. vestita



a

b

Fig. 6. Patterns of response in canopy cover following the 2007 fire fora) Helichrysum indicum and b) Diastella divaricata according to those areasburnt previously in 1986 (short fire interval of five years) and areas not burnedpreviously for at least 16 years i.e. burnt at some time before 1975 (long fireinterval).


is viewed as a fire ephemeral (Brown, 1993) but in this study itscover increased significantly in the third year. E. sesamoides,which is prominent ‘in the few years after fire’ (Bean and Johns,2005), was first recorded in the second year and its cover anddensity increased significantly in the third year. The decline inplant density was not always followed by a decline in canopycover. Thus, the mass mortality ofD. divaricata seedlings in thesecond year was followed by another significant reduction indensity in the third year while cover increased significantly, as a

Fig. 7. Monthly rainfall recorded at Klaasjagersberg (Rainfall Station 0004734 2,which is 3 km from the study site) for the period of the study. The fire occurredearly in April 2007 before any rain had fallen in that month.


result of presumed rapid compensatory growth of the survivingplants in the third year. The significant decrease in cover of thegraminoid Ischyrolepis cincinnata in the second year followedby a significant increase in the third is difficult to interpret in theabsence of density data. Le Maitre (1987) suggested that a largedecline in plant density found in the second post-fire year ingranite fynbos was the result of density-dependent thinningafter establishment. By contrast, in a field study of Proteaceae,drought was apparently found to be responsible for most seedlingdeaths, especially in shallow soil over rock (Bond, 1984).Relatively dry conditionsmay have contributed to seedling deathson the current study site, especially on the commonly foundshallow soil over rock. In the second year, rainfall in the first halfof the normal rainy season (April to July) was about 30% belowthe mean rainfall for this period (Fig. 7). Possible effects ofherbivory on seeding mortality were not assessed in this study.Bond (1984) established that predationwas not a significant causeof post-emergence seedling mortality in the Proteaceae that hestudied. However, in two species of this family, post-fireherbivory on young seedlings was found to have a significantlynegative effect on seedling survival (Le Maitre, 1988a).

Most of the species of Erica were only recorded from thesecond year onwards, which may suggest a delayed regener-ation after fire (but see possible early detection limitationsbelow). All of these apparently later-appearing species of Ericawere obligate reseeders. Germination and establishment ofsome Erica species elsewhere in fynbos have been observed tobe delayed until more than a year after fire (Kruger and Bigalke,1984; Manders and Cunliffe, 1987), with one study reporting adelay of more than two and a half years after fire (Adamson,1935). Although Erica seed may be expected to be relativelyshort-lived in the soil seed bank (Holmes and Newton, 2004),this should not significantly affect the viable seed supply in thefirst post-fire year. Germination of a wide range of Ericaspecies, stored in ambient room conditions, starts to declineafter about three years and after 10 years germinate very poorlyor not at all (A. Hitchcock personal communication). Ericaceaehave fine seeds (b1 mg, le Maitre and Midgley, 1992) or veryfine seeds (b0.1 mg, Bond et al., 1999; Holmes and Newton,2004). Seedlings of fine-seeded species cannot emerge from soildepths ≥20 mm and are likely to be killed by heat of a highintensity fire at shallower depths (Bond et al., 1999). Mostericoid species store their seed in the upper layers of soil(Cowling et al., 1997) and are thus vulnerable to fire. Evidently,storage at greater depths would not ensure survival. It has beensuggested that the delayed regeneration of Erica species afterfire may depend on transport of seed to the burned site, or onother special germination requirements that are met only at alater stage (Kruger and Bigalke, 1984; Manders and Cunliffe,1987). The small size of seeds of most Erica species shouldpermit their relatively long-distance dispersal, especially in thehigh-wind environment of the current study site. In addition, itappears that some level of spatial pyrodiversity (Parr andAndersen, 2006) may be necessary to allow transfer from fertilesources to sterile areas. Spatial configurations of different firesand size of the burnt areas (Manders and Cunliffe, 1987) cancritically affect such transfers.




Other species with apparently delayed regeneration after fireinclude the reseeder Restionaceae under dryland conditions,i.e. outside seep areas. This delay may be a consequence ofcompetition. Sprouting species can have a competitive edgeover reseeders because of the ability of the former to re-establishrapidly after fire (Vlok and Yeaton, 2000). By contrast, thedelayed regeneration found in the reseeder hemiparasite, T.strictum, (Moore et al., 2010) may also result from a delay in therecovery of the host plant(s) after fire (see Kazanis andArianoutsou, 2004). Moreover, the presumed limited mobilityof seeds of this species due to dispersal through myrmecochory,may further delay post fire recovery (Johnson, 1992).

If germination of most fynbos annuals is triggered by smokeor charred wood (Keeley and Bond, 1997) then it would beexpected that annuals might be largely confined to the first post-fire year (see Kruger, 1979). However, the current study showsincreasing canopy cover of annuals from the first to third yearsafter fire. This may relate to the less inhibiting effect of therelatively open vegetation canopy that remained below 50%cover by the third post-fire year. This is in contrast to therelatively high cover (Kruger and Bigalke, 1984) or near canopyclosure (Adamson, 1935) found on some other sandstonefynbos sites by the third post-fire year. Indeed, canopy cover infynbos can reach 70 to 90% in only two years after fire (Kruger,1983). Nevertheless, cover of annuals and the number of annualspecies in this study remained relatively low, which is expectedin fynbos (Taylor, 1978; Kruger, 1979), even in the very openareas after fire (Adamson, 1935). It has been suggested that therarity of fire annuals in fynbos is linked to limitation by lowlevels of soil nutrients (Le Maitre and Midgley, 1992).

The role of smoke in fynbos has largely focussed on its useas a germination cue for horticultural purposes (Brown andBotha, 2004), yet its ecological role often appears speculative.Germination of certain reseeder species, such as E. sesamoides,is reported to be stimulated by smoke but appear to establish inonly the second year. It is possible that the timing of the possiblestimulatory effect of smoke may be long after the fire for somespecies. The smoke cue may be long-lasting and the smoke-derived residue persists in the soil for a considerable time (VanStaden et al., 2000) and hence may still be effective as agermination stimulant some time after fire. However, the activeconstituent of smoke is water-soluble and may be leached outby winter rains (Van Staden et al., 2000). The ecologicalsignificance of smoke remains unclear (Cowling et al., 1997).

The increase in species richness over the first three yearsafter fire agrees with the findings of Adamson (1935) andHoffman et al. (1987) and does not support the notion of amonotonic decline in number of species over this period. Ourresults also appear to bear out Kruger's (1983) suggestion of aninitial increase in fynbos species richness after the first post fireyear. Although the study of Privett et al. (2001) indicated asteady declining trend in species richness from the first year forseveral decades after fire, the data that they presented for thefirst three years are too variable to support a decline over thisearly post-fire period. Furthermore, Taylor (1984), whose datawere used almost exclusively for the first three post-fire yearsby Privett et al. (2001), stated that species richness peaks at


between two and five years after fire. A study in granite fynbosfound that plant species richness remained at least relativelyconstant over the first three post-fire years before gentlydeclining toward the sixth year after fire (Le Maitre, 1987).

We have previously indicated that the challenge in determiningfynbos species richness over the first three post-fire years isdetecting and identifying immature species, especially in the firstpost-fire year. Indeed, Kruger (1983) cautioned that the findings ofapparent increasing richness in the first few post-fire years byAdamson (1935) in fynbos and Posamentier et al. (1981) inAustralian heathland may be spurious owing to the difficulties indetecting or identifying plants in the early stages of regeneration.Musil and de Witt (1990) have also pointed to the difficulty indistinguishing some fynbos species in the juvenile stage in the field.Even under relatively favourable conditions, species richness ispotentially subject to errors of omission or identification at thespecies level (Archaux et al., 2009). In the current study 20 knownresprouter species (out of the total of 206 species) were not detectedin the first year. Their addition would result in a corrected richnessof 46 per 100 m2 in the first year. This is very similar to themean of47 species per 100 m2 that was found for 30 replicate plots of oneyear old fynbos after a fire on a sandstone site in the Swartberg(Vlok and Yeaton, 1999). Although it is likely that many short-lived seedling individuals had disappeared by the time of samplingin January (Cilliers et al., 2004), it is unknownwhether any species(including some annuals) may have been consequently missed onour site. It would require the addition of more than 10 such speciesto have been undetected in the first year for species richness in thefirst year to surpass that found in the second year. However, if themajority of very young establishing Erica species as well as someother reseeder species had been undetected in the first post-fireyear, a temporary decline in species richness cannot be ruled out.Nevertheless, from the second to third post-fire years, when specieswere more readily identifiable, the increase found in speciesrichness remained significant. Therefore, a short-term monotonicdecline in species richness after fire in fynbos cannot be supportedby the current study. It remains an open question to what extent theapparent long-term decline in fynbos species richness after firewould be offset, were seed banks to be included (see Holmes andCowling, 1997).

Vlok and Yeaton (1999) have suggested that fire history priorto the last fire may be important in determining diversity levels infynbos. However, we could not detect any trend in speciesrichness according to the prior short (five year) fire interval and theprior long (more than 16 years) fire interval. It might be arguedthat the significant positive association of the annual,Helichrysumindicum, with the prior short fire interval area relates to annualsbenefitting from more open vegetation. It begs the question,however, how this effect could have been sustained through asubsequent period of 17 years under presumed closed canopy. Thenegative association of the obligate reseeder,Diastella divaricata,with the short fire interval areas does not seem adequatelyexplained by increased survival risk through immaturity. Thus, inmost five-year old post-fire populations of D. divaricata, at least50% of plants would have flowered at least twice (A.G. Rebelo,personal communication). However, even with juvenile periods ofonly three to four years for some Proteaceae, it can take at least




another 10 years for adequate seed banks to build up (Van Wilgenand Forsyth, 1992). If seed production ofD. divaricata had indeedbeen adversely affected by the short fire interval, this myrmeco-chorous species with very short dispersal distances (Cowling andGxaba, 1990; Johnson, 1992) could take many years to repopulatean area. However, since fynbos normally requires four to five yearsbefore a sufficient fuel load has accumulated to carry a fire (Kruger,1983), the burn of the young five-year old vegetationmay not havebeen clean, and small pockets of unburned vegetation may haveremained as close sources of propagules. It is important to note thatany inferences from these prior fire interval areas suffer fromunavoidable pseudoreplication in that neither interdispersion ofsamples nor independent treatment requirements of samplingreplication are met. This means that despite the very similarphysical conditions on both prior burn areas, other factors could beresponsible for, or contribute to, the differences.

The current study has helped quantify fire responses ofdifferent plant forms and species in fynbos, in contrast to usefulyet purely descriptive accounts (for example, Trinder-Smith,2006). It has also shown that claims of a short-term monotonicdecline in species richness after fire in fynbos are not supported.One of the problems in comparing some results of differentpost-fire studies in fynbos may be linked to the poorrepeatability of the results, given the great variation incomposition caused by fire. Post-fire research carried out over30 years in the vicinity of the current study has shown highlyunstable species composition at local scale (Thuiller et al.,2007) with, for example, L. laureolum becoming locally extincton nearly 30% of sites but colonizing a similar proportion ofnew sites in the same period (Privett, et al., 2001). Conditionsfollowing a fire are very complex and change rapidly (see Bondand Van Wilgen, 1996). There is still much to be learned abouthow these conditions are reflected in interactions and levels oftemporal segregation of species in the early stages after fire, andwhat impact they may have on later successional development.

Acknowledgements

We thankRuthCozien,KhuzeMahlabegwane andGwenCurryfor assistance during the fieldwork. The research comprised a subproject within BIOTA Southern Africa Phase III incorporated in aResearch and Development Contract between the University ofHamburg (UHH) and the South African National BiodiversityInstitute (SANBI) sponsored by German Federal Ministry ofEducation and Research (Bundesministerium für Bildung undForschung (BMBF)) — Promotion number 01LC0624A2. Wethank SANParks for granting a research permit and providingaccess to the Cape of Good Hope section of the Table MountainNational Park.

References

Adamson, R.S., 1935. The plant communities of Table Mountain. III. A sixyears' study of regeneration after burning. Journal of Ecology 23, 44–55.

Archaux, F., Camaret, S., Dupouey, J.-L., Ulrich, E., Corcket, E., Bourjot, L.,Brêthes, A., Chevalier, R., Dobremez, J.-F., Dumas, Y., Dumé, G., Forêt,M., Forgeard, F., Lebret Gallet, M., Picard, J.-F., Richard, F., Savoie, J.-M.,Seytre, L., Timbal, J., Touffet, J., 2009. Can we reliably estimate species


richness with large plots? An assessment through calibration training. PlantEcology 203, 303–315.

Bean, A., Johns, A., 2005. Stellenbosch to Hermanus. South African wild flowerguide. Botanical Society of South Africa, Cape Town.

Bergstedt, J., Westerberg, L., Milberg, P., 2009. In the eye of the beholder: biasand stochastic variation in cover estimates. Plant Ecology 204, 271–283.

Bond, W.J., 1984. Fire survival of Cape Proteaceae — influence of fire seasonand seed predators. Vegetatio 56, 65–74.

Bond, W.J., 1997. Fire. In: Cowling, R.M., Richardson, D.M., Pierce, S.M.(Eds.), Vegetation of Southern Africa. Cambridge University Press,Cambridge, pp. 421–446.

Bond, W.J., Keeley, J.E., 2005. Fire as a global ‘herbivore’: the ecology andevolution of flammable ecosystems. Trends in Ecology and Evolution 20,387–394.

Bond, W.J., Van Wilgen, B.W., 1996. Fire and Plants. Chapman & Hall,London.

Bond, W.J., Vlok, J., Viviers, M., 1984. Variation in seedling recruitment ofCape Proteaceae after fire. Journal of Ecology 72, 209–221.

Bond, W.J., Le Roux, D., Erntzen, R., 1990. Fire intensity and regeneration ofmyrmecochorous Proteaceae. South African Journal of Botany 56, 326–330.

Bond, W.J., Honig, M., Maze, K.E., 1999. Seed size and seedling emergence: anallometric relationship and some ecological implications. Oecologia 120,132–136.

Brown, N.A.C., 1993. Seed germination in the fynbos fire ephemeral, Syncarphavestita (L.) B. Nord. is promoted by smoke, aqueous extracts of smoke andcharred wood derived from burning the ericoid shrub, Passerina vulgarisThoday. International Journal of Wildland Fire 3, 203–206.

Brown, N.A.C., Botha, P.A., 2004. Smoke seed germination studies and a guideto seed propagation of plants from the major families of the Cape FloristicRegion, South Africa. South African Journal of Botany 70, 559–581.

Cilliers, C.D., Esler, K.J., Boucher, C., 2004. Effects of alien plant managementand fire on soil seed banks and regeneration in the Cape Peninsula NationalPark, South Africa. South African Journal of Botany 70, 705–712.

Cowling, R.M., Gxaba, T., 1990. Effect of a fynbos overstorey shrub onunderstorey community structure: implications for the maintenance ofcommunity-wide species richness. South African Journal of Ecology 1, 1–7.

Cowling, R.M., Pierce, S.M., 1988. Secondary succession in coastal dunefynbos: variation due to site and disturbance. Vegetatio 76, 131–139.

Cowling, R.M., Holmes, P.M., Rebelo, A.G., 1992. Plant diversity andendemism. In: Cowling, R.M. (Ed.), The Ecology of Fynbos. Nutrients, Fireand Diversity. Oxford University Press, Cape Town, pp. 62–112.

Cowling, R.M., Richardson, D.M.,Mustart, P.J., 1997. Fynbos. In: Cowling,R.M.,Richardson, D.M., Pierce, S.M. (Eds.), Vegetation of Southern Africa.Cambridge University Press, Cambridge, pp. 99–130.

Cowling, R.M., Procheş, Ş., Partridge, T.C., 2009. Explaining the uniqueness ofthe Cape flora: incorporating geomorphic evolution as a factor for explainingits diversification. Molecular Phylogenetics and Evolution 51, 64–74.

Fehmi, J.S., 2010. Confusion among three common plant cover definitions mayresult in data unsuited for comparison. Journal of Vegetation Science 21,273–279.

Forsyth, G.G., Van Wilgen, B.W., 2008. The recent fire history of the TableMountain National Park and implications for fire management. Koedoe 50,3–9.

Haaksma, E.D., Linder, H.P., 2000. Restios of the Fynbos. Botanical Society ofSouth Africa, Cape Town.

Haarmeyer, D.H., Luther-Mosebach, J., Dengler, J., Schmiedel, U., Finckh, M.,Berger, K., Deckert, J., Domptail, S.E., Dreber, N., Gibreel, T., Grohmann,C., Gröngröft, A., Haensler, A., Hanke, W., Hoffmann, A., Husted, L.B.,Kangombe, F.N., Keil, M., Krug, C., Labitzky, T., Linke, T., Mager, D.,Mey, W., Muche, G., Naumann, C., Pellowski, M., Powrie, L.W., Pröpper,M., Rutherford, M.C., Schneiderat, U., Strohbach, B., Vohland, K., Weber,B., Wesuls, D., Wisch, U., Zedda, L., Büdel, B., Darienko, T., Deutschewitz,K., Dojani, S., Erb, E., Falk, T., Friedl, T., Kanzler, S.-E., Limpricht, C.,Linsenmair, K.E., Mohr, K., Oliver, T., Petersen, A., Rambold, G., Zeller,U., Austermühle, R., Bausch, J., Bösing, B.M., Classen, N., Dorendorf, J.,Dorigo, W., Esler, K.J., Etzold, S., Graiff, A., Grotehusmann, L., Hecht, J.,Hoyer, P., Kongor, R.Y., Lang, H., Lieckfeld, L.A.B., Oldeland, J., Peters,J., Röwer, I.U., September, Z.M., Sop, T.K., Van Rooyen, M.W., Weber, J.,




Willer, J., Jürgens, N., 2010. The BIOTA Observatories. In: Jürgens, N.,Haarmeyer, D.H., Luther-Mosebach, J., Dengler, J., Finckh, M., Schmiedel,U. (Eds.), Biodiversity in southern Africa: Patterns at local scale — theBIOTA Observatories. Klaus Hess Publishers, Göttingen & Windhoek.

Hoffman, M.T., Moll, E.J., Boucher, C., 1987. Post-fire succession at Pella, aSouth African lowland fynbos site. South African Journal of Botany 53,370–374.

Holmes, P.M., Cowling, R.M., 1997. Diversity, composition and guild structurerelationships between soil-stored seed banks and mature vegetation inalien plant-invaded South African fynbos shrublands. Plant Ecology 133,107–122.

Holmes, P.M., Newton, R.J., 2004. Patterns of seed persistence in South Africanfynbos. Plant Ecology 172, 143–158.

Johnson, S.D., 1992. Plant-animal relationships. In: Cowling, R.M. (Ed.), TheEcology of Fynbos. Nutrients, Fire and Diversity. Oxford University Press,Cape Town, pp. 175–205.

Kazanis, D., Arianoutsou, M., 2004. Long-term post-fire vegetation dynamics inPinus halepensis forests of central Greece: a functional group approach.Plant Ecology 171, 101–121.

Keeley, J.E., Bond, W.J., 1997. Convergent seed germination in South Africanfynbos and Californian chaparral. Plant Ecology 133, 153–167.

Kruger, F.J., 1979. South African heathlands. In: Specht, R.L. (Ed.), Heathlandsand Related Shrublands — Descriptive Studies. : Ecosystems of the World,9A. Elsevier, Amsterdam, pp. 19–80.

Kruger, F.J., 1983. Plant community diversity and dynamics in relation to fire.In: Kruger, F.J., Mitchell, D.T., Jarvis, J.U.M. (Eds.), Mediterranean-typeEcosystems. The Role of Nutrients. : Ecological Studies, 43. Springer-Verlag,Berlin, pp. 446–472.

Kruger, F.J., Bigalke, R.C., 1984. Fire in fynbos. In: Booysen, P. de V., Tainton,N.M. (Eds.), Ecological Effects of Fire in South African Ecosystems.Springer-Verlag, Berlin, pp. 67–114.

Le Maitre, D.C., 1987. Effects of season of burn on species populations andcomposition of fynbos in the Jonkershoek valley. South African Journal ofBotany 53, 284–292.

Le Maitre, D.C., 1988a. Effects of season of burn on the regeneration of twoProteaceae with soil-stored seed. South African Journal of Botany 54, 575–580.

Le Maitre, D.C., 1988b. The effects of parent density and season of burn on theregeneration of Leucadendron laureolum (Proteaceae) in the Kogelberg.South African Journal of Botany 54, 581–584.

Le Maitre, D.C., Midgley, J.J., 1992. Plant reproductive ecology. In: Cowling,R.M. (Ed.), The Ecology of Fynbos. Nutrients, Fire and Diversity. OxfordUniversity Press, Cape Town, pp. 135–174.

Manders, P.T., Cunliffe, R.N., 1987. Fynbos plant life histories, populationdynamics and species interactions in relation to fire: an overview. SouthAfrican National Scientific Programmes Report 135, 15–23.

Milberg, P., Bergstedt, J., Fridman, J., Odell, G., Westerberg, L., 2008. Observerbias and random variation in vegetation monitoring data. Journal ofVegetation Science 19, 633–644.

Moore, T.E., Verboom, G.A., Forest, F., 2010. Phylogenetics and biogeographyof the parasitic genus Thesium L. (Santalaceae), with emphasis on the Capeof South Africa. Botanical Journal of the Linnean Society 162, 435–452.


Musil, C.F., de Witt, D.M., 1990. Post-fire regeneration in a sand plain lowlandfynbos community. South African Journal of Botany 56, 167–184.

Parr, C.L.,Andersen,A.N., 2006. Patchmosaic burning for biodiversity conservation:a critique of the pyrodiversity paradigm. Conservation Biology 20, 1610–1619.

Posamentier, H.G., Clark, S.S., Hain, D.C., Recher, H.F., 1981. Successionfollowing wildfire in coastal heathland (Nadgee Nature Reserve NSW).Australian Journal of Ecology 6, 165–175.

Privett, S., Cowling, R.M., Taylor, H.C., 2001. Vegetation change in the Cape ofGood Hope Nature Reserve. Bothalia 31, 99–115.

Rebelo, A.G., Boucher, C., Helme, N., Mucina, L., Rutherford, M.C., 2006.Fynbos biome. In: Mucina, L., Rutherford, M.C. (Eds.), The Vegetation ofSouth Africa, Lesotho and Swaziland. : Strelitzia, 19. South AfricanNational Biodiversity Institute, Pretoria, pp. 52–219.

Simmons, M.T., Cowling, R.M., 1996. Why is the Cape Peninsula so rich inplant species? An analysis of the independent diversity components.Biodiversity and Conservation 5, 551–573.

Taylor, H.C., 1978. Capensis. In: Werger, M.J.A. (Ed.), Biogeography andEcology of Southern Africa. Junk, The Hague, pp. 171–229.

Taylor, H.C., 1984. A vegetation survey of the Cape of Good Hope NatureReserve. II. Descriptive account. Bothalia 15, 259–291.

Thuiller, W., Slingsby, J.A., Privett, S.D.J., Cowling, R.M., 2007. Stochasticspecies turnover and stable coexistence in a species-rich, fire-prone plantcommunity. PLoS ONE 2 (9), e938. doi:10.1371/journal.pone.0000938.

Trinder-Smith, T., 2006. Wild flowers of the Table Mountain National Park.South African wild flower guide 12. Botanical Society of South Africa, CapeTown.

Van der Niet, T., Johnson, S.D., 2009. Patterns of plant speciation in the Capefloristic region. Molecular Phylogenetics and Evolution 51, 85–93.

Van Staden, J., Brown, N.A.C., Jäger, A.K., Johnson, T.A., 2000. Smoke as agermination cue. Plant Species Biology 15, 167–178.

Van Wilgen, B.W., Forsyth, G.G., 1992. Regeneration strategies in fynbosplants and their influence on the stability of community boundaries after fire.In: VanWilgen, B.W., Richardson, D.M., Kruger, F.J., Van Hensbergen, H.J.(Eds.), Fire in South African Mountain Fynbos. : Ecological Studies, 93.Springer-Verlag, Berlin, pp. 54–80.

Van Wilgen, B.W., Bond, W.J., Richardson, D.M., 1992. Ecosystemmanagement. In: Cowling, R.M. (Ed.), The Ecology of Fynbos. Nutrients,Fire and Diversity. Oxford University Press, Cape Town, pp. 345–371.

Van Wilgen, B.W., Forsyth, G.G., de Klerk, H., Das, S., Khuluse, S., Schmitz,P., 2010. Fire management in Mediterranean-climate shrublands: a casestudy from the Cape fynbos, South Africa. Journal of Applied Ecology 47,631–638.

Vlok, J.H.J., Yeaton, R.I., 1999. The effect of overstorey proteas on plant speciesrichness in South African mountain fynbos. Diversity and Distributions 5,213–222.

Vlok, J.H.J., Yeaton, R.I., 2000. Competitive interactions between overstoreyproteas and sprouting understorey species in South African mountainfynbos. Diversity and Distributions 6, 273–281.

Edited by RI Yeaton



Early post-fire plant succession in Peninsula Sandstone Fynbos: … · 2012-02-24 · Early...

Documents

Transcript of Early post-fire plant succession in Peninsula Sandstone Fynbos: … · 2012-02-24 · Early...