Regeneration After Disturbances of Varying Severities - Reyes Et Al - JVS 2010

8/14/2019 Regeneration After Disturbances of Varying Severities - Reyes Et Al - JVS 2010

1/12


2/12

Catastrophic re has historically been the pro-minent large, severe disturbance driving standdynamics in Abies balsameaBetula papyrifera bor-eal mixed-wood forests of northeastern NorthAmerica (Bergeron et al. 2001). But with a recentclimate-induced decrease in catastrophic re fre-

quency (Bergeron et al. 1998), other smaller and/orless severe disturbance types, such as those causedby spruce budworm ( Choristoneura fumiferana )outbreaks, windthrow, and their interactions, canbecome more important in terms of shaping land-scape structure and species distribution patterns.Moderate-severity disturbances, in particular, cancause profound change, as they can affect 10 to 100 sof hectares of forest (Clinton & Baker 2000; Grayet al. 2000; Nagel & Diaci 2006), and can occur re-latively frequently in the absence of re. Sprucebudworm outbreak, for example, has occurred on a

20- to 50-year cycle in northeast North Americasince the end of the Little Ice Age ( $ 1850) (Blais1983; Bouchard et al. 2006). Note, however, that al-though being similar in spatial extent tocatastrophic re, unlike catastrophic re, moderate-severity disturbances leave the regeneration layerrelatively intact, and are thus less severe.

Successful establishment occurs where specieslife-history strategies are compatible with the pre-vailing establishment or regeneration conditionsgenerated by the disturbance regime (Grime 1977).Different species functional groups often show di-vergent responses to disturbance (Oliver 1981;Roberts 2004; Kennard & Putz 2005). Early-succes-sion, light-demanding species are usually associatedwith larger, more severe disturbances, whereas late-succession, shade-tolerant species often dominatethe regeneration layer after smaller, less severe dis-turbances (Frelich 2002; Gromtsev 2002). Speciesdiversity tends to increase after disturbance (Pe-terson & Pickett 1995), and is also affected byvariation in the amount of disturbance-caused mor-tality (Peltzer et al. 2000). Highest estimates arehypothesized to occur at intermediate levels (Con-nell 1978; Miller 1982).

Given the variability in species responses, theability to predict regeneration patterns after differ-ent types of natural disturbance causing varyingdegrees of mortality would signicantly contributeto our understanding of boreal mixed-wood standdynamics; further, although the early/late succes-sional species classication is useful to predictspecies abundance and diversity after canopy gapversus large, catastrophic disturbances, species re-sponses to moderate-severity disturbances remainunclear. Past research in boreal mixed-woods has

focused on either end of the disturbance impactspectrum (e.g., Leemans 1991; Bergeron & Danser-eau 1993; Kneeshaw & Bergeron 1998), with littleconsideration of patterns of recovery across a rangeof disturbance-caused mortality.

The objectives of our study are to examine and

compare woody species regeneration abundanceand diversity after natural disturbances causingvarying degrees of mortality. Specically, we aim to:(1) determine if signicant changes in relative abun-dances of the principal canopy species occur aftercanopy gap, moderate-severity, and large, cata-strophic re disturbances; and (2) examine thevariation in woody species diversity in relation tothe these disturbance-caused mortality classes.

Study Area

The study was conducted within A. balsamea B. papyrifera boreal mixed-wood stands located inthe Chaleur Bay region of the Gaspe Peninsula,Quebec, Canada (48 1 100 481 350N, 65 1 45 0 66 1 150W). Total surface area of the study area isapproximately 6480km 2 . These A. balsameaB. papyrifera stands are a part of a transitional borealmixed-wood bioclimatic domain bounded by borealforests to the north and north-temperate deciduousforests to the south. Consequently, the forest cano-py consists of a mix of evergreen conifer andbroadleaf deciduous species. A. balsamea , a shade-tolerant, evergreen conifer species characteristic of late-succession conditions in these forests, andB. papyrifera , a shade-intolerant broadleaf decid-uous colonizer, dominate the forest landscape.Species of lesser importance include the shade-tol-erant Picea glauca , Picea mariana , and Thujaoccidentalis , the mid-tolerant Acer rubrum , Betulaalleghaniensis , and Pinus strobus , and the shade-in-tolerant Populus balsamifera and Populustremuloides . The general transition of species aftercatastrophic re is from a broadleaf deciduous ca-nopy followed by a mixed coniferousdeciduous

canopy that can persist for several generations,which transitions to a conifer-dominated climaxcommunity in the absence of re (Bergeron 2000).

Climate for the region is humid continental.Mean minimum daily temperature in January is 15.8 1 C, while mean daily maximum in July is22.6 1 C. Mean annual precipitation is 984 mm, withapproximately 79% falling as rain (EnvironmentCanada 2004). General topography is rolling tomontane, as the region is within the northern limit of the Appalachian Mountain chain that runs from an

2 Reyes, Gerardo P. et al.


3/12

eastnortheast to a westsouthwest into the USA.Elevations range from 80 to 600 m (mean 285 ma.s.l.). Soils originate from glacial tills, glacio-lacus-trine, uvio-lacustrine, or alteration deposits(Robitaille & Saucier 1998). Soil drainage rangesfrom imperfect to rapid. Very few lakes are present

although several important rivers are interspersedthroughout the region.

The current re cycle, estimated to be between170 and 250 years (Lauzon et al. 2007), allows manystands to escape catastrophic disturbances for longperiods of time. Subsequently, canopy gap andmoderate-severity disturbances can have greaterimpacts on the ensuing regeneration composition if occurring more frequently than re. Return intervalestimates were 30 and 39 years for canopy gap andmoderate-severity disturbances, respectively (calcu-lated using methods described in Canham & Loucks

1984; Runkle 1992; Zhang et al. 1999). Return in-terval estimates were consistent with estimates fromassociated research undertaken within the easternboreal mixed-wood region of Quebec, Canada(Bouchard et al. 2006; De Ro mer et al. 2007).

Methods

Effects of variation in disturbance-caused mortalityon regeneration abundance and diversity

Stands were surveyed during the summer of 2004. Relative species composition of the tree cano-py can be quite variable among A. balsamea B. papyrifera boreal mixed-wood stands. To isolatedifferences in regeneration patterns due to variationin disturbance-caused mortality, we selected mature( ! 89-year-old) A. balsameaB. papyrifera borealmixed-wood stands having similar relative canopyspecies composition and density prior to dis-turbance, general topography, elevation, and soilmoisture and drainage regimes (as determined fromvarious Government of Quebec forest inventory andland classication maps). Thus, we sampled from

mesic sites having 75:25 conifer to deciduous treespecies ratios prior to disturbance, being wellstocked, occurring on relatively at terrain, andhaving good to rapid drainage. Ground-truthing of sites (quantifying live and dead canopy trees withinthe disturbed area) veried that conifers representedat least 75% of the tree canopy density prior to dis-turbance, of which A. balsamea accounted for atleast 60% of all conifer trees. B. papyrifera was themost abundant deciduous species, accounting for atleast 60% of all deciduous canopy trees. Sampling

protocols for each disturbance-caused mortalityclass are described below.

Canopy gap disturbances

Canopy gap disturbances were differentiatedfrom moderate-severity disturbances according tothe spatial extent of the mortality caused to the treecanopy. Disturbances affecting between 0.0004 and0.2 ha of contiguous tree canopy were classied ascanopy gap disturbances. One to three transects of variable length (approximately 100 to 400 m) wererandomly established within each of ve matureA. balsameaB. papyrifera boreal mixed-wood stands( ! 89 years old) to quantify vegetation regenera-tion abundance and diversity after canopy gapdisturbances. Canopy gaps were caused by a num-

ber of disturbance agents (including mortalitycaused by the spruce budworm, windthrow, and se-nescence). Canopy gaps wherein the cause of gapformation could not be determined (i.e., no sign orpresence of gap makers) were not included in ourstudy. Transects started and ended at least 30 mfrom the nearest logging road or forest edge, andfollowed a straight trajectory. A total of 46 gapswere sampled. Various gap characteristics werequantied within each canopy gap that traversed thetransect line. Length and width of each gap weremeasured using the longest northsouth distanceaffected by gap disturbance in conjunction with thelongest distance in an eastwest direction that tra-versed the northsouth vector. Gap size wasdetermined using the following equation: area 5p (length width)/4 (Runkle 1992). Gap age (i.e.,time since disturbance) was determined using an-nual ring counts measured at ground level from veshade-intolerant Prunus pensylvanica or B. papyr-ifera seedlings established within each gap, as it wasexpected that shade-intolerant species establishedsoon after gap formation. In the absence of thesespecies, timing of release of A. balsamea was de-termined using the same methods, as we assumed

that release from suppression of A. balsamea oc-curred within a few years after gap formation(Fraver & White 2005; Metslaid et al. 2005). Tree,sub-canopy tree, and shrub regeneration density (allindividuals 8 cm dbh) was quantied along thenorthsouth vector used to measure gap length,within an area 2 m of each side of the vector. Thus,sampling effort was a function of gap size. Mean gapsize was 0.004 ha, ranged from 0.0005 to 0.02 ha,with more than half the gaps smaller than the mean,necessitating the use of a different sampling protocol

Disturbance effects on woody vegetation abundance and diversity 3


4/12

from moderate-severity disturbances and cata-strophic re described below.

Moderate-severity disturbances

Spruce budworm outbreak, windthrow, and in-teraction disturbances (dened as a stand affectedby spruce budworm outbreak followed by wind-throw prior to complete canopy recovery) fall withinour moderate-severity disturbance class when caus-ing more than 0.2ha of continuous canopymortality. An upper spatial limit was not initiallyrmly specied, as moderate-severity disturbancescan be quite extensive, differing with catastrophicre primarily in relation to effect on the regenera-tion layer. Nonetheless, the largest disturbed areaexamined was 95 ha (mean 5 24 ha, smallest 5 0.03ha). Eleven stands affected by moderate-severitydisturbances were sampled. Three to six, randomlyplaced 20 20 m quadrats were examined withineach disturbed stand. A total of 43 quadrats weresampled. All quadrats were at least 40 m from thenearest intact forest edge and 30 m from the nearestlogging road, to avoid edge effects. In every quadrat,regeneration density was determined for all tree,sub-canopy tree, and shrub species using a nestedplot design: density of seedlings o 1-m tall were tal-lied in a 2 10-m area, those between 1- and 2-m tall,in a 510-m area, and regeneration 4 2-m tall and

8 cm dbh, using the entire 20 20-m area. Time

since disturbance was determined with the samemethods used for canopy gap disturbances.

Large catastrophic re disturbances

Stand-level effects of catastrophic re can behighly variable (e.g., Lampainen et al. 2004). Unlikecanopy gap and moderate-severity disturbances,catastrophic re can affect stands in young, earlysuccessional stages. Moreover, divergent succes-sional trajectories have been shown to occur whenstand age differed prior to disturbance (Heinselman

1973; Johnstone & Chapin 2006). To make mean-ingful comparisons with the other disturbance-caused mortality classes, we limited our sampling toseverely affected areas (extensive mortality to bothcanopy and regeneration layers) that were matureA. balsameaB. papyrifera stands prior to a cata-strophic re that burned 59 ha of forest in 1988.Note that we did not account for variation in sever-ity of catastrophic re damage to the organic ormineral soil layers. Given that we sampled 16 yearsafter re, most of the differences among sites have

become difcult to detect. Thus, we only ensuredthat we sampled in areas wherein 4 75% mortalityof the pre-re regeneration layer occurred. This wasveried in the eld by determining the age of the re-generation from annual ring counts at ground levelof ten of the largest seedlings within each quadrat.

Pre-disturbance stand compositions and stand ageclasses were determined using Government of Que-bec forest inventory maps of the region producedprior to 1988, and were veried in the eld withidentication (mostly to species, but at least to con-ifer versus deciduous) and annual ring counts of remnant stumps. Tree and shrub regeneration den-sity was assessed in ten, 20 20-m quadrats using themoderate-severity disturbance protocol.

AnalysesEffects of variation in disturbance-caused mortalityon regeneration abundance and diversity

Canonical ordination analyses were used to ex-amine the role of variation in disturbance-causedmortality (canopy gap versus moderate-severity ver-sus large catastrophic re) on the response of regenerating tree, sub-canopy tree, and shrub species,in terms of absolute densities, with CANOCO 4.02software (ter Braak & Smilauer 1998). Disturbancetypes (spruce budworm outbreak, windthrow, inter-action, or other) were included in analyses as dummyvariables, to examine responses of vegetation to eachof these factors. An initial run using detrended corre-spondence analysis (DCA) showed data to have lineardistribution (all gradient lengths were o 2.2 standarddeviations for axes 1 to 4). Therefore, redundancyanalysis (RDA), a direct gradient approach, wasused to examine disturbance-caused mortality, dis-turbance type, and regeneration abundance anddiversity relationships. The forward selection optionwas implemented to rank the importance of eachenvironmental variable, and to remove any environ-

mental variables that did not signicantly contributeto the observed variation. Monte Carlo permutationtests were used to evaluate the signicance of eachenvironmental variable (at a 5 0.05 to enter or stay inmodel, 200 permutations for each run). While canopygap, moderate-severity, and large catastrophic redisturbances signicantly affected regeneration abun-dance and diversity (all P o 0.05), none of the variousdisturbance types had signicant effects (all P 4 0.05).Thus, only outputs comparing relationships betweeneach disturbance-caused mortality class and re-



5/12

generation composition and abundance are shown.A set of analyses separating regeneration sizeclasses (seedlings o 1-m tall, those between1- and 2-mtall, and regeneration 4 2-m tall and 8 cm dbh)were also run. Because our results did not appreciablydiffer from those with regeneration size classes

pooled, only results showing pooled size classes areshown. Further, only sites that had undergone dis-turbances within a few years of one another wereincluded in analyses to ensure that differences in spe-cies composition and abundance among disturbance-caused mortality classes were not a result of largediscrepancies in time since disturbance. Thus 23, 33,and 10 of the canopy gap, moderate-severity dis-turbance, and catastrophic re plots, respectively,were included in this analysis. Times since disturbancefor each canopy gap, moderate-severity, and cata-strophic re disturbance were 15, 19, and 16 years,

respectively.

Species richness and diversity in relation to variationin disturbance-caused mortality

Differences in species richness ( R ) amongseverity classes were examined by randomly select-ing 10 plots from each of the canopy gap (twoplots from each site) and MSD (one plot eachfrom 10 of the 11 sites) classes and analyzing theseagainst the 10 catastrophic re plots. Species rich-ness (R ), dened as the total number of tree, sub-canopy tree, and shrub species, and Shannonsdiversity index:

H 0 S pi ln pi 1were determined for each plot, and comparedamong disturbance-caused mortality classes usingGLM procedures (SPSS 10.0 1999). The Student NewmanKeuls multiple range test ( a 5 0.05) wasused for post hoc comparisons. Data transformationwas unnecessary for R as data met the parametricassumptions of normality and homogeneity of var-iances among treatment groups (Zar 1996), whereasH 0 data were ln transformed prior to analyses.

Results


Variation in disturbance-caused mortality re-sulted in distinct changes in woody species abundanceand diversity. Canopy gap and moderate-severitydisturbances caused minor changes to the relative

abundance and diversity of canopy tree species in re-lation to pre-disturbance conditions ( ! 75% coniferoverstory), whereas substantial changes were ob-served after catastrophic re (Fig. 1a and b). Mostsub-canopy tree and shrub species increased in abun-dance after catastrophic re (Fig. 1c and d). The rst

two axes in RDA explained 26 and 5% of the totalvariance in the species data, respectively (Fig. 2).Catastrophic re was the environmental variablemost strongly correlated with axis 1 (canopy gap: r 5 0.43, moderate-severity disturbance: r 5 0.24,catastrophic re: r 5 0.91), while canopy gap andmoderate-severity disturbances were more stronglycorrelated to axis 2 (canopy gap: r 5 0.63, moderate-severity disturbance: r 5 0.69, catastrophic re:r 5 0.12). Early-succession, post-re species segre-gated to the right along axis 1 whereas late-successional, shade-tolerant species pooled to the left

with the canopy gap and moderate-severity dis-turbances. The second axis mainly differentiated mid-to high-shade-tolerant species (in the upper leftquadrant of the ordination diagram) from mid- tolow-shade-tolerant species (in the lower half of theordination diagram) between canopy gap and mod-erate-severity disturbances.

A. balsamea , a shade-tolerant conifer species,was strongly associated with both canopy gap andmoderate-severity disturbances (Fig. 2). A. balsameaaccounted for 92% and 81% of all seedlings aftercanopy gap and moderate-severity disturbances,respectively (Fig. 1a and b). B. papyrifera , a shade-intolerant deciduous species, dominated the treeregeneration after catastrophic re, accounting for54% of all seedlings established (Fig. 1b). Othershade-intolerant and mid-tolerant tree species suchas Pinus strobus , Populus balsamifera and Populustremuloides clustered to the right, along axis 1 in or-dination space, and thus were more stronglyassociated with catastrophic re.

The shade-intolerant Corylus cornuta and Pru-nus pensylvanica dominated the sub-canopy tree andshrub components after catastrophic re, but werepresent at only very low densities after canopy gap

and moderate-severity disturbances. Rubus spp.,which are rapidly colonizing, shade-intolerantshrubs, were abundant after both moderate-severitydisturbances and catastrophic re, and absentafter canopy gap disturbances. Salix spp. estab-lished only after catastrophic re, while the mid-tolerant Acer rubrum and the high shade-tolerantA. pensylvanicum pooled towards the upper left inordination space, indicating greater abundance aftercanopy gap disturbances. A. spicatum was able tosuccessfully regenerate under all post-disturbance



6/12

conditions, suggesting that its distribution is medi-ated by other factors more than variation in theamount of disturbance-caused mortality. Coryluscornuta , Rubus spp., and A. spicatum exhibitedaggregated distributions, and produced densethickets that precluded establishment of other spe-cies. Nemopanthus mucronata , Sorbus spp., andAmelanchier spp. pooled to the left along axis 1, butmore towards moderate-severity disturbances alongaxis 2.


Twenty-two tree and shrub species were observedwithin the system (Table 1). Tree species richness ( R)was lowest after moderate-severity disturbances andgreatest after catastrophic re (Table 2). Sub-canopytree and shrub R and total woody plant species R washighest after catastrophic re. Greater R after cata-strophic re relative to either canopy gap or moderate-severity disturbance is attributable to the inux of

early-successional, shade-intolerant species. Shan-nons diversity index ( H 0) for canopy trees was similaramong disturbance-caused mortality classes (Table 2).Sub-canopy tree and shrub H 0 and total woody plantspecies H 0 were lower after moderate-severity dis-turbances relative to either canopy gap or catastrophicre disturbances. Tree species compositions were mostsimilar after canopy gap and moderate-severitydisturbances, while sub-canopy tree and shrub re-generation was most similar after moderate-severitydisturbances and catastrophic re disturbances. Thus,moderate-severity disturbances shared commonalities

with both extremes of the disturbance-caused mortal-ity gradient, but for different forest regenerationlayers. Irrespective of the amount of disturbance-caused mortality, canopy tree regeneration was domi-nated by one or two species, whereas sub-canopy treeand shrub regeneration was more evenly distributedamong component species (Fig. 1). Five species estab-lished exclusively after catastrophic re, one aftermoderate-severity disturbances, while no tree, sub-ca-nopy tree, or shrub species was exclusive to canopygap disturbances (Table 1).

d e n s

i t y

( / 1 0 0 m

2 )

100

300

700

900

0

200

800

ba

r e l a t i v e

d e n s

i t y

0.0

0.2

0.4

0.6

0.8

1.0

d e n s i

t y ( / 1 0 0 m

2 )

0

20

40

60

80

100

120

140

160c d

r e l a t i v e

d e n s

i t y

0.0

0.2

0.4

0.6

0.8

1.0

Fig. 1. (a-d) Total and relative tree and sub-canopy tree and shrub regeneration density after canopy gap disturbance,moderate-severity disturbance (MSD), and large catastrophic re in A. balsameaB. papyrifera boreal mixed-wood forests.Letters after species name indicate shade tolerance: H 5 high, HM 5 high to mid, M 5 mid, and L 5 low shade tolerance.Shade tolerance levels were derived from Humbert et al. (2007).



7/12

Discussion


Distinct regeneration patterns were observed aftercanopy gap, moderate-severity, and catastrophic redisturbances. Marked changes from pre-disturbance

conditions, wherein shade-tolerant conifer speciesdominated the tree community, only occurred aftercatastrophic re. The abundance of advance re-generation facilitated cyclical regeneration patternsafter canopy gap and moderate-severity disturbances,while destruction of most of the regeneration layerafter catastrophic re allowed for the establishment of more shade-intolerant species. For example, theshade-intolerant B. papyrifera dominated the treecommunity after catastrophic re, but was onlya minor component after canopy gap and moderate-

severity disturbances. B. papyrifera densities were si-milar and low after canopy gap and moderate-severitydisturbances, respectively, despite large differences inthe spatial extent among many of the canopy gap andmoderate-severity disturbances examined here (dis-turbances ranged from 0.0005 to 95ha).

Given that the moderate-severity disturbancesexamined in this study caused more than 75% con-

tiguous canopy mortality, and that the spatial extentof some moderate-severity disturbances was largerthan that after catastrophic re, it would be reason-able to expect greater changes in the tree communityafter such disturbances (e.g., Miller 1982; Peterson2000). Yet, despite causing major structural chan-ges, relative tree species composition was notsubstantially altered from pre-disturbance condi-tions. This was the case irrespective of the type of moderate-severity disturbance, be it spruce bud-worm outbreak, windthrow, or their interaction,

1.0 +1.0

1.0

+1.0

catastrophicfire

canopy gapdisturbance

moderate-severity

disturbance

cor

salbF Pb

pru

Pt

rub

sam

yB

die

wP

Ce

vib

mM

ame

sor

rMwS

wB

pM

nem

bS

a x i s 2 ( 1 6 . 8 % )

axis 1 (83.2 %)

Fig. 2. Redundancy analysis (RDA) of species density in relation to variation in disturbance-caused mortality. The rst twocanonical axes in RDA explained 26% and 5% of the cumulative variance in the species data, respectively. Species codes areas follows: bF 5 Abies balsamea , bS 5 Picea mariana , Ce 5 Thuja occidentalis , mM 5 Acer spicatum , Pb 5 Populus balsamifera , Pt 5 Populus tremuloides , pM 5 Acer pensylvanicum , rM 5 Acer rubrum , wB 5 Betula papyrifera , wS 5 Picea glauca ,wP 5 Pinus strobus , yB 5 Betula alleghaniensis , ame 5 Amelanchier spp., cor 5 Corylus cornuta , die 5 Diervella lonicera ,pru 5 Prunus spp., nem 5 Nemopanthus mucronata , rub 5 Rubus spp., sal 5 Salix spp., sam 5 Sambucus spp., sor 5 Sorbusspp., and vib 5 Viburnum edule . Shade tolerance: high, high to mid, mid, mid to low, low.



8/12


9/12

regeneration for long periods of time until a canopydisturbance occurs (Morin & Laprise 1997). A. bal-samea advance regeneration, in particular, alreadyoccupied much of the available growing space.Many shade-intolerant tree species in our studyregion that can regenerate from seed require ex-

posed mineral soils or recently burned organic soilsto establish; e.g., Populus tremuloides , Pinus bank-siana (Zasada et al. 1992). Seed dispersal constraints(Le Page et al. 2000) and low-pre-disturbance den-sities of canopy tree species that can regenerate viaroot suckering may also have played a role (e.g.,Populus spp.); moreover, seeds of the canopy treespecies in the region do not persist in the soil seedbank for more than a few years (Frank & Safford1970; Greene et al. 1999), suggesting that thesoil seed bank had little role in tree regenerationpatterns.

High densities of coarse woody debris can affectseedling establishment (Reyes & Kneeshaw 2008).Downed logs in our sites inhibited germination byphysically covering potential establishment sites,and reduced light availability to the surface. Con-versely, downed logs can also act as nurse sites whensufciently decomposed (Stevens 1997; Iijima et al.2007). However, in our study, advance regenerationhad already grown over the majority of these po-tential regeneration sites, negating any potentialbenets (such as reduced competition) that nursesites can provide. Species regenerating from seedmay thus be limited to establishing on exposed mi-neral soil resulting from tree uprooting afterwindthrow, and in areas with lower densities of ad-vance regeneration and coarse woody debris (Reyes& Kneeshaw 2008). The importance of advance re-generation on post-disturbance dynamics has alsobeen reported elsewhere (Leemans 1991; Ban et al.1998; Grassi et al. 2004). Nevertheless, the decid-uous component in mixed-wood stands is oftenmaintained after moderate-severity disturbance(Bergeron 2000; De ry et al. 2000). Although B. papyrifera did not increase in relative abundance aftermoderate-severity disturbances in our study, the

species maintained its importance in the canopy.Thus, moderate-severity disturbances may also actto slow successional trajectories towards late-suc-cession conditions or to stabilize the currentlandscape structure over many generations (Berger-on 2000; Woods 2004; Papaik & Canham 2006).

While tree regeneration was similar after cano-py gap and moderate-severity disturbances,considerable differences in the sub-canopy tree andshrub community were observed. Differences in es-tablishment success of sub-canopy tree and shrub

species between canopy gap and moderate-severitydisturbance are related to different life-historycharacteristics of the component species and theavailability and use of specic resource require-ments. Certain species such as Prunus pensylvanicaand Rubus spp. were quick to respond to ephemeral

resources, and could take better advantage of subtlemicroenvironment differences between canopy gapand moderate-severity disturbances. Rubus spp., inparticular, are disturbance specialists that can re-main dormant for many years in the seed bank, anddo not require exposed mineral soils for germination(Lautenschlager 1991; Palmer et al. 2000). Further,nitrate (NO 3 ) pulses generally occur in forest soils 1to 2 years after more severe disturbances (e.g., re,soil scarication) (Truax et al. 1994), which cantrigger germination of dormant Rubus spp. seeds(Jobidon 1993). Greater abundance of intolerant

sub-canopy tree and shrub regeneration within lar-ger gaps has been observed elsewhere in the borealmixed-wood region (Kneeshaw & Bergeron 1998).


Our study supports the view that groups of species differing in important life-history character-istics exhibit different responses across a range of disturbance-caused mortality. However, for therange of natural disturbances occurring in easternboreal mixed-woods, our results for species richnessand diversity of the woody plant community did notfollow Connells intermediate disturbance hypoth-esis (1978). Also, in contrast to Miller (1982), largespatial differences between canopy gap and moder-ate-severity disturbances did not equate tosignicant increases in species richness or diversity.Richness and diversity estimates were, in fact, simi-lar or lower after moderate-severity disturbancesrelative to both disturbance-caused mortality ex-tremes. Our results show that the intermediate-disturbance hypothesis is not always observed whencatastrophic re is the primary disturbance. Schwilk

et al. (1997) obtained comparable results afterres in Mediterranean climate shrublands, ndinglowest plant diversity estimates at intermediate refrequencies. Mackey & Currie (2001) and Sasakiet al. (2009) also show that the unimodal patternof diversity characteristic of the intermediate-disturbance hypothesis is not always realized inother terrestrial systems.

Mean species richness estimates and Shannonsdiversity index were unable to account for differ-ences in the actual species present among the



10/12

disturbance-caused mortality classes. Distinct chan-ges in the sub-canopy tree and shrub communityoccurred when transitioning from canopy gap andmoderate-severity disturbances, while several spe-cies were unique to catastrophic re. Proliferation of disturbance specialists and shade-intolerant species,

along with persistence of shade-tolerant, late-suc-cessional species, resulted in higher species richnessestimates after catastrophic re. The absence of several mid-tolerant and shade-intolerant speciesafter canopy gap disturbances suggests that some of these species would be extirpated from the systemwithout periodic, moderate-severity or catastrophicre disturbances. Thus, variation in disturbance-caused mortality is crucial to creating habitat di-versity, allowing different species functional groupsto persist across the landscape. Further, change inthe vegetation community from canopy gap to the

catastrophic re range of disturbances is not a line-ar, continuous species replacement process.Moderate-severity disturbances have distinct attri-butes, and subsequently effects, on regenerationabundance and diversity. The difculty in char-acterizing disturbances within the moderate-severityrange is that vegetation responses have importantsimilarities (and differences) to both disturbance-caused mortality extremes, depending on the vege-tation layer examined.

Conclusions

Natural disturbances can have lasting and dis-tinct effects on the landscape. Expected changes inglobal climate will have considerable effects onmortality, recruitment patterns, and disturbance re-gimes. While re cycles are expected to shorten andthe incidence and severity of insect disturbances ispredicted to increase in much of the boreal region(Stewart et al. 1998; Volney & Fleming 2000), east-ern North American boreal mixed-wood re cycleshave been lengthening (Bergeron et al. 2001; Lauzon

et al. 2007). Current return interval lengths for thevarious disturbances in A. balsameaB. papyriferaboreal mixed-wood forests is consistent with a dis-turbance regime characterized by relatively frequentpartial events and rare catastrophic events. Giventhat canopy gap and moderate-severity disturbancesdo not appreciably change relative species composi-tion of the canopy tree layer from pre-disturbanceconditions, convergence towards late-successionalforest conditions is promoted throughout the land-scape. The projected increase in time intervals

between catastrophic res may lead to reduced di-versity within the system.

Acknowledgements. Jean-Francois Liquidrano-Gagnon,Isabelle Nault, Mathieu Bouchard, David Saucier, Jona-

tan Belle-Isle, Steve Bujold, Julie Messier, and MaudeBeauregard were important contributors to data collec-tion (and integral in either keeping me from losing mysanity in the eld or accelerating the process). Ad-ditionally, this study would not have been possiblewithout nancial and/or technical support from TEM-REX, NSERC-CFS, and the SFMN.

References

Ban, Y., Xu, H., Bergeron, Y. & Kneeshaw, D. 1998. Gapregeneration of shade-intolerant Larix gmelini in old-

growth boreal forests of northeastern China. Journal of Vegetation Science 9: 529536.Bergeron, Y. 2000. Species and stand dynamics in the

mixed woods of Quebecs southern boreal forest.Ecology 81: 15001516.

Bergeron, Y. & Dansereau, P.R. 1993. Predicting thecomposition of Canadian southern boreal forest indifferent re cycles. Journal of Vegetation Science 4:827832.

Bergeron, Y., Richard, P.J.H., Carcaillet, C., Gauthier, S.,Flannigan, M. & Prairie, Y.T. 1998. Variability in refrequency and forest composition in Canadassoutheastern boreal forest: a challenge for sustainableforest management. Conservation Ecology 2: 6,Available at: http://www.consecol.org/vol2/iss2/art6/.

Bergeron, Y., Gauthier, S., Kafka, V., Lefort, P. &Lesieur, D. 2001. Natural re frequency for theeastern Canadian boreal forest: consequences forsustainable forestry. Canadian Journal of ForestResearch 31: 384291.

Blais, J.R. 1983. Trends in the frequency, extent, andseverity of spruce budworm outbreaks in easternCanada. Canadian Journal of Forest Research 13:539547.

Bouchard, M., Kneeshaw, D. & Bergeron, Y. 2006. Forestdynamics after successive spruce budworm outbreaksin mixedwood forests. Ecology 87: 23192329.

Canham, C.D. & Loucks, O.L. 1984. Catastrophicwindthrow in the presettlement forests of Wisconsin.Ecology 65: 803809.

Clinton, B.D. & Baker, C.R. 2000. Catastrophicwindthrow in the southern Appalachians:characteristics of pits and mounds and initialvegetation responses. Forest Ecology and Management 126: 5160.

Connell, J.H. 1978. Diversity in tropical rain forest andcoral reefs. Science 199: 13041310.

De Ro mer, A.H., Kneeshaw, D.D. & Bergeron, Y. 2007.Small gap dynamics in the southern boreal forest of

http://www.consecol.org/vol2/iss2/art6/http://www.consecol.org/vol2/iss2/art6/


11/12

eastern Canada: do canopy gaps inuence standdevelopment? Journal of Vegetation Science 18:815826.

De ry, S., Be langer, L., Marchand, S. & Co te , S. 2000.Succession apre s e pide mie de la tordeuse desbourgeons de le pinette ( Choristeneura fumiferana )dans des sapinie res bore ales pluviales de secondevenue. Canadian Journal of Forest Research 30:801816.

Environment Canada. 2004. Canadian climate normalsor averages 1971-2000 . Available at: http://climate.weatherofce.ec.gc.ca/climate_normals/index_e.html(accessed July 2007).

Frank, R.M. & Safford, L.O. 1970. Lack of viable seeds inthe forest oor after clearcutting. Journal of Forestry68: 776778.

Fraver, S. & White, A.S. 2005. Identifying growthrelationships in dendrochronological studies of forestdisturbance. Canadian Journal of Forest Research 35:16481656.

Frelich, L. 2002. The disturbance regime and itscomponents. In: Birks, H.J.B. & Weins, J.A. (eds)Forest dynamics and disturbance regimes: studies fromtemperate evergreendeciduous forests . pp. 1543.Cambridge University Press, New York, NY, US.

Frelich, L.E. & Reich, P.B. 1999. Neighbourhood effects,disturbance severity, and community stability inforests. Ecosystems 2: 151166.

Gleason, H.A. & Cronquist, A. 1991. Manual of vascular plants of Northeastern United States & AdjacentCanada . 2nd ed. New York Botanical Garden, NewYork, NY, US.

Grassi, G., Minotta, G., Tonon, G. & Bagnaresi, U. 2004.Dynamics of Norway spruce and silver r naturalregeneration in a mixed stand under uneven-agedmanagement. Canadian Journal of Forest Research 34:141149.

Gray, D.R., Re gnie re, J. & Boulet, B. 2000. Analysis anduse of historical patterns of spruce budwormdefoliation to forecast outbreak patterns in Quebec.Forest Ecology and Management 127: 217231.

Greene, D.F., Zasada, J.C., Sirois, L., Kneeshaw, D.,Morin, H., Charron, I. & Simard, M.J. 1999. Areview of the regeneration dynamics of NorthAmerican boreal forest tree species. Canadian Journal of Forest Research 29: 824839.

Grime, J.P. 1977. Evidence for the existence of three

primary strategies in plants and its relevance toecological and evolutionary theory. AmericanNaturalist 111: 11691194.

Gromtsev, A. 2002. Natural disturbance dynamics in theboreal forests of European Russia: a review. SilvaFennica 36: 4155.

Heinselman, M.L. 1973. Fire in the virgin forests of theBoundary Waters Canoe Area, Minnesota.Quaternary Research 3: 329382.

Humbert, L., Gagnon, D., Kneeshaw, D. & Messier, C.2007. A shade tolerance index for common understory

species of northeastern North America. Ecological Indicators 7: 195207.

Iijima, H., Shibuya, M. & Sato, H. 2007. Effects of surfacelight conditions of fallen logs on the emergence andsurvival of coniferous seedlings and saplings. Journal of Forest Research 12: 262269.

Jobidon, R. 1993. Nitratefertilization stimulates emergenceof red raspberry ( Rubus idaeus L.) under forest canopy.Nutrient Cycling in Agroecosystems 36: 9194.

Johnstone, J.F. & Chapin, F.S. III 2006. Fire intervaleffects on successional trajectory in boreal forests of northwest Canada. Ecosystems 9: 268277.

Kennard, D.K. & Putz, F.E. 2005. Differential responsesof Bolivian timber species to prescribed re and othergap treatments. New Forests 30: 120.

Kneeshaw, D. & Bergeron, Y. 1998. Canopy gapcharacteristics and tree replacement in thesoutheastern boreal forest. Ecology 79: 783794.

Lampainen, J., Kuuluvainen, T., Wallenius, T.H.,Karjalainen, L. & Vanha-Majamaa, I. 2004. Long-

term forest structure and regeneration after wildre inRussian Karelia. Journal of Vegetation Science 15:245256.

Lautenschlager, R.A. 1991. Red raspberry ecology and theeffect of raspberry and other forest brush on whitespruce growth . Maine Agricultural ExperimentStation, Misc. Rep. No. 360, 7pp.

Lauzon, E., Kneeshaw, D.D. & Bergeron, Y. 2007. Forestre history reconstruction (1680-2003) in the Gaspesieregion of eastern Canada. Forest Ecology and Management 244: 4149.

Leemans, R. 1991. Canopy gaps and establishmentpatterns of spruce ( Picea abies (L.) Karst.) in two old-growth coniferous forests in central Sweden. PlantEcology 93: 157165.

Le Page, P.T., Canham, C.D., Coates, K.D. &Bartemucci, P. 2000. Seed abundance versus substratelimitation of seedling recruitment in northerntemperate forests of British Columbia. CanadianJournal of Forest Research 30: 415427.

Mackey, R.L. & Currie, D.J. 2001. The diversity disturbance relationship: is it generally strong andpeaked? Ecology 82: 34793492.

Metslaid, M., Ilinsson, T., Nikinmaa, E., Kusmin, J. &Jogiste, K. 2005. Recovery of advance regenerationafter disturbances: acclimation of needlecharacteristics in Picea abies . Scandinavian Journal of

Forest Research 20 (Suppl 6): 112121.Miller, T.E. 1982. Community diversity and interactionsbetween the size and frequency of disturbance.American Naturalist 120: 533536.

Morin, H. & Laprise, D. 1997. Seedling bank dynamics inboreal balsam r forests. Canadian Journal of ForestResearch 27: 14421451.

Nagel, T.A. & Diaci, J. 2006. Intermediate winddisturbance in an old-growth beechr forest insouthwestern Slovenia. Canadian Journal of ForestResearch 36: 629638.

http://climate.weatheroffice.ec.gc.ca/climate_normals/index_e.htmlhttp://climate.weatheroffice.ec.gc.ca/climate_normals/index_e.htmlhttp://climate.weatheroffice.ec.gc.ca/climate_normals/index_e.htmlhttp://climate.weatheroffice.ec.gc.ca/climate_normals/index_e.htmlhttp://climate.weatheroffice.ec.gc.ca/climate_normals/index_e.htmlhttp://climate.weatheroffice.ec.gc.ca/climate_normals/index_e.html


12/12

Oliver, C.D. 1981. Forest development in North Americafollowing major disturbances. Forest Ecology and Management 3: 153168.

Palmer, M.W., McAlister, S.D., Arevalo, J.R. &DeCoster, J.K. 2000. Changes in the understoryduring 14 years following catastrophic windthrow intwo Minnesota forests. Journal of Vegetation Science11: 841854.

Papaik, M.J. & Canham, C.D. 2006. Species resistanceand community response to wind disturbance regimesin northern temperate forests. Journal of Ecology 94:10111026.

Peltzer, D.A., Bast, M.L., Wilson, S.D. & Gerry, A.K.2000. Plant diversity and tree responses followingcontrasting disturbances in boreal forest. ForestEcology and Management 127: 191203.

Peterson, C.J. 2000. Damage and recovery of tree speciesafter two different tornadoes in the same old growthforest: a comparison of infrequent wind disturbances.Forest Ecology and Management 135: 237252.

Peterson, C.J. & Pickett, S.T.A. 1995. Forestreorganisation a case study in an old-growth forestcatastrophic blowdown. Ecology 76: 763774.

Reyes, G. & Kneeshaw, D. 2008. Moderate-severitydisturbance dynamics in Abies balsameaBetula spp.forests: the relative importance of disturbance typeand local stand and site characteristics on woodyvegetation response. Ecoscience 15: 241249.

Rich, R.L., Frelich, L.E. & Reich, P.B. 2007. Wind-throwmortality in the southern boreal forest: effects of species, diameter and stand age. Journal of Ecology95: 12611273.

Roberts, M.R. 2004. Response of the herbaceous layer tonatural disturbance in North American forests.Canadian Journal of Botany 82: 12731283.

Robitaille, A. & Saucier, J.P. 1998. Paysages re gionaux duQue bec me ridional . Les Publications du Que bec, QU,CA.

Runkle, J.R. 1992. Guidelines and Sample Protocol forSampling Forest Gaps . USDA Forest Service PacicNorthwest Research Station General TechnicalReport PNW-GTR-283, 44pp.

Sack, L. & Grubb, P.J. 2002. The combined impacts of deep shade and drought on the growth and biomassallocation of shade-tolerant woody seedlings.Oecologia 131: 175185.

Sasaki, T., Okubo, S., Okayasu, T., Jamsran, U., Ohkuro,

T. & Takeuchi, K. 2009. Management applicability of the intermediate disturbance hypothesis acrossMongolian rangeland ecosystems. Ecological Applications 19: 423432.

Schwilk, D.W., Keeley, J.E. & Bond, W.J. 1997. Theintermediate disturbance hypothesis does not explainre and diversity pattern in fynbos. Plant Ecology 132:7784.

Splechtna, B.E., Gratzer, G. & Black, B.A. 2005.Disturbance history of a European old-growthmixed-species forest A spatial dendro-ecologicalanalysis. Journal of Vegetation Science 16: 511522.

SPSS Inc. 1999. Professional base system software forstatistical analysis (v.10.0) . SPSS Inc., Chicago, IL, US.

Stevens, V. 1997. The ecological role of coarse woodydebris: an overview of the ecological importance in BC forests . Research Branch, BC Ministry of Forests,Victoria, BC, Canada. Working paper 30/1997, 26pp.

Stewart, R.B., Wheaton, E. & Spittlehouse, D.L. 1998.Climate change: implications for the Boreal forest. In:Legge, A.H. & Jones, L.L. (eds.) Emerging Air Issues for the 21st Century: The Need for MultidisciplinaryManagement, Proceedings of an International SpecialtyConference, Sep. 2224, 1997, Calgary, AB . pp. 86 101. Saskatchewan Research Council, Saskatoon, CN.

ter Braak, C.J.F. & Smilauer, P. 1998. CANOCOReference Manual and Users Guide to CANOCO forWindows: Software for Canonical Community

Ordination (Version 4.02) . Microcomputer Power,Ithaca, NY, US.

Truax, B., Gagnon, D., Lambert, F. & Chevrier, N. 1994.Nitrate assimilation of raspberry and pin cherry in arecent clearcut. Canadian Journal of Botany 72: 1343 1348.

Turner, M.G., Dale, V.H. & Everham, E.H. III 1997.Crown res, hurricanes, and volcanoes: a comparisonamong large-scale disturbances. BioScience 47: 758 768.

Volney, W.J.A. & Fleming, R.A. 2000. Climate changeand impacts of boreal forest insects. Agriculture,Ecosystems & Environment 82: 283294.

Webb, S.L. & Scanga, S.E. 2001. Windstorm disturbancewithout patch dynamics: twelve years of change in aMinnesota forest. Ecology 82: 893897.

Woods, K.D. 2004. Intermediate disturbance in a late-successional hemlocknorthern hardwood forest.Journal of Ecology 92: 464476.

Zar, J.H. 1996. Data Transformations. In: Snavely, S.L.(ed.) Biostatistical analysis . 3rd ed, pp. 277284.Prentice-Hall, NJ, US.

Zasada, J.C., Sharik, T.L. & Nygren, M. 1992. Thereproductive process in boreal forest trees. In:Shugart, H., Leemans, R. & Bonan, G. (eds.) Asystem analysis of the global boreal forest . pp. 85125.Cambridge University Press, Cambridge, UK.

Zhang, Q., Pregitzer, K.S. & Reed, D.D. 1999.Catastrophic disturbance in the presettlement forestsof the Upper Peninsula of Michigan. Canadian Journal of Forest Research 29: 106114.

Received 26 August 2009;Accepted 11 November 2009.

Co-ordinating Editor: Dr. Christoph Leuschner.


Regeneration After Disturbances of Varying Severities - Reyes Et Al - JVS 2010

Documents

Transcript of Regeneration After Disturbances of Varying Severities - Reyes Et Al - JVS 2010