Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Structure and regeneration of old-growth Thuja plicata stands near Vancouver, British Columbia Daniels, Lori D. 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-0381.pdf [ 2.54MB ]
JSON: 831-1.0075247.json
JSON-LD: 831-1.0075247-ld.json
RDF/XML (Pretty): 831-1.0075247-rdf.xml
RDF/JSON: 831-1.0075247-rdf.json
Turtle: 831-1.0075247-turtle.txt
N-Triples: 831-1.0075247-rdf-ntriples.txt
Original Record: 831-1.0075247-source.json
Full Text

Full Text

STRUCTURE AND REGENERATION OF OLD-GROWTH THUJA PLICATA STANDSNEAR VANCOUVER, BRITISH COLUMBIAbyLori D. DanielsB.Sc. University of Manitoba 1991A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FORTHE DEGREE OF MASTER OF SCIENCEinFACULTY OF GRADUATE STUDIESDepartment of Forestry (Forest Sciences)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAJune 1994© Lori Dianne DanielsIn presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)Department of_________________7/1The University of British (blumbiaVancouver, CanadaDate /,yy2ci /999DE-6 (2/88)11ABSTRACTIn many old-growth stands on submontane (elevation <ca. 600 m) sites in the Very WetMaritime Coastal Western Hemlock biogeocimatic subzone, Thuja plicata populations areconsidered to be declining. Thuja dominate the upper canopy of such stands, but are scarce orabsent in the lower canopies and understories, and are believed to be in the process of beingreplaced by the more numerous Tsuga heterophylla and Abies amabilis. To assess Thujapopulation dynamics and to test the null hypothesis that Thuja populations are not sustained inold-growth forests, this study analyzed the structure of eight old-growth stands located nearVancouver, British Columbia, in which Thuja, Tsuga and Abies form the tree layer.The effect of stand-level fire disturbances were evident in the composition, age, and sizestructures of three of the study stands. However, small-scale processes, indicative of the old-growth stage of stand development, dominated the dynamics of seven of the stands. Regardlessof stand histories, Thuja populations included trees of a range of sizes and ages and there wereindividuals in the understories in all study stands, indicating that the species is capable ofregeneration in old-growth forests. Size and age structure analysis revealed that regeneration ofThuja was sporadic and is hypothesized to be disturbance-related, whereas regeneration of Tsugaand Abies appears to have been continuous.Spatial pattern analyses of seedlings and saplings revealed aggregate distributions for thethree study species; the distribution of canopy trees was not different from random. Thedistribution of seedlings and saplings was highly associated with the location of exposed mineralsoil and decaying wood substrates, which indicated the importance of disturbance to thesuccessful establishment of all three species. Thuja and Tsuga regenerated most successfully onexposed mineral soil and decaying wood, while Abies regenerated most successfully onundisturbed forest floor and decaying wood. No consistent relationship was identified betweenthe distribution of Thuja regeneration and canopy tree density, an indirect measure of canopygaps, which suggested that canopy gaps may not be necessary to facilitate Thuja establishment.111Temporal differences in the regeneration niches of the three species were revealed by theanalysis of diameter distributions of seedlings and saplings. Seedling mortality of all threespecies was high, but mortality of established Thuja regeneration (diameter at ground level >4cm) appeared low, which implied that once established, the chance for successful recruitmentwas high. As few seedlings reached 4 cm in diameter, the recruitment and canopy density ofThuja remained low. This interpretation was consistent with the age structure analysis of canopytrees (dbh >10 cm) which suggested that mortality of Thuja trees was lower than that of Tsugaand Abies trees. The low recruitment - low density - low mortality population dynamic of Thujais consistent with the storage effect (i.e., long-lived, rare species may be sustained in spite of lowestablishment and recruitment).Increment cores from all live canopy trees were measured to provide ring-width seriesfrom which individual tree growth and stand development could be interpreted. Past gap eventswere identified by spatial pattern analysis of trees of similar age and time of release and throughcomparison of ring-width series of all Thuja and their neighbours. Comparison of tree growth byspecies, height class, and gap occurrence revealed interspecific differences in growth response tocanopy gaps. Tsuga and Abies appeared dependent on gaps to recruit to the upper canopy. Theirpopulations featured suppressed trees in the lower canopy, with low mean annual diameterincrements although they often had released multiple times, and trees with significantly highermean annual diameter increments in the upper canopy. The relatively low frequency of releasesin Thuja and its constant mean annual diameter increment among height classes suggested Thujawas not dependent on canopy gaps to gain the upper canopy. Differences in the growth patternsand growth response to gaps of the three study species might be one mechanism that enablestheir coexistence.It was concluded that Thuja populations were not in decline in the study area.Differences in life history characteristics, including longevity, age-specific mortality rates,recruitment success, and adaptations and response to the old-growth understory lightenvironment likely explain the coexistence of Thuja, Tsuga and Abies in the old-growth forest.ivThe relative importance of these attributes to population and stand dynamics and quantificationof these mechanisms and processes remain to be explored.VTABLE OF CONTENTSABSTRACT iiTABLE OF CONTENTS vLIST OF TABLES viLIST OF FIGURES viiACKNOWLEDGEMENTS viiiChapter 1. Introduction 1The Study of Forest Dynamics 1The Dynamics of Thuja plicata-Dominated Old-Growth Stands 3Objectives and Thesis Outline 5Chapter 2. Methods 7Study Area 7Study Design 8Data Collection 11Chapter 3. Stand Structure 14Introduction 14Methods 15Results 17Discussion 29Conclusions 34Chapter 4. Natural Regeneration 35Introduction 35Methods 36Results 40Discussion 49Conclusions 53Chapter 5. Canopy Tree Growth and Gap Dynamics 54Introduction 54Methods 55Results 57Discussion 68Conclusions 72Chapter 6. Conclusions 74Thujaplicata Dynamics 74Comparisons of Schools of Thought on Succession 75Future Studies 76Literature Cited 78Appendix A Mean percent cover of understory species 89Appendix B Summary of substrate utilization by seedlings and saplings 90Appendix C Survivorship of regeneration on different substrates 96viLIST OF TABLESTable 2.1 Location and description of old-growth forest study stands 10Table 3.1 Species composition (number per hectare) by canopy stratum in the eightstudy stands 18Table 3.2 Species-specific regression model of age (years) on dbh (cm) for the fourintensively studied stands 21Table 3.3 Summary of seeding and sapling age- and height-class distributions (numberper ha) 23Table 3.4 Comparison of (stems I ha) and recruitment required for canopy maintenanceof the three study species in the four intensively studied 25Table 4.1 Morisita’s index of dispersion (1d) for Tsuga heterophylla and Abies amabilisseedlings and saplings in the eight study stands 43Table 4.2 Summary of substrate utilization by seedlings and saplings 44Table 4.3 Success of regeneration on different substrates 45Table 5.1 Decision key used to determine a tree’s growth history 56Table 5.2 Values of Morisita’s index of dispersion (‘d) and probability values (p-values)calculated for modes and decadal classes with significantly aggregated trees(i.e., not randomly distributed) in the four study stands 59Table 5.3 Number of release periods observed among canopy trees (dbh >10 cm) infour study stands 64Table 5.4 Summary of gap occurrence over time in the four the study stands 65viiLIST OF FIGURESFigure 2.1 Location of study stands and areas near Vancouver, British Columbia 9Figure 3.1 Diameter distributions of canopy trees (dbh >10 cm) in the eight studystands 20Figure 3.2 Age distributions of canopy trees (dbh >10 cm) in the four intensivelysampled study stands 22Figure 3.3 Height distributions of trees (height >5 m) in the four intensively sampledstudy stands 27Figure 3.4 Values of Morisit&s index (1d) for canopy trees and subcanopy trees atdifferent quadrat sizes for the four intensively sampled study plots 28Figure 4.1 Values of Morisita’s index of dispersion (Td) for all Thuja plicata regenerationand Thuja regeneration differentiated by substrate at different quadrat sizesfor the four intensively sampled study stands 41Figure 4.2 Diameter distributions of regeneration (dbh l0 cm) in the eight studystands 47Figure 4.3 Age structures of Thuja plicata regeneration (dbh l0 cm) for the fourintensively sampled study stands 48Figure 5.1 Chronology of ingress, releases, and gap events for four study plots 58Figure 5.2 Map of tree locations and occurrence of gap events in stand Sl 60Figure 5.3 Map of tree locations and occurrence of gap events in stand S4 61Figure 5.4 Map of tree locations and occurrence of gap events in stand N 1 62Figure 5.5 Map of tree locations and occurrence of gap events in stand N4 63Figure 5.6 Comparison of the mean annual increments of canopy trees in four heightclasses 67VIIIACKNOWLEDGEMENTSI thank the authorities of the Greater Vancouver Water District for permission to conduct thisresearch. Financial support for this study was provided by the Greater Vancouver Water District,the Vancouver Forest Region of the British Columbia Ministry of Forests, and Fletcher-Challenge (Canada) Ltd., which sponsored my G.R.E.A.T. scholarship. I was supported by aFaculty of Forestry Bursary, a British Columbia Science Council G.R.E.A.T. scholarship, andreceived unending “in-kind” support from Edith and George Daniels. The support of all of theseagencies is gratefully acknowledged.I thank Dr. Karel Klinka, my supervisor, for the opportunity to conduct this research. It was Dr.Kihika’s curiosity about Thuja plicata that inspired this project and gave me the opportunity tospend many days exploring, and of course studying, the magnificent old-growth forests of coastalBritish Columbia. Our discussions and his advice, guidance, and support have been instrumentalboth to this thesis and to the tremendous learning experience that I have had while studying at theUniversity of British Columbia. I hope that in return I have provided some understanding ofThuja plicata forests and have answered some of Dr. Klinka’s questions.Thank-you to the members of my supervisory committee, Dr. Peter Marshall, Dr. Philip Burtonand Mr. Reid Carter, for their comments, criticisms and helpful suggestions. I would also like torecognize Dr. Jaroslav Dobry who has helped me to learn the art of dendrochronology.Special recognition go to Dave Andison and Steve Daniels. To Dave, for your assistance, advice,comments and debates, for your love and support, and for reminding me, at just the right times,to “live a little”. To Steve, for your hard work in the field and for your energy and enthusiasmfor the study of ecology which made our work in the forest enjoyable, even on those cool wetmesothermal days. I hope you share with me fond memories of the rattley-blue-Nissan,“wheelies”, humming bears, VacciniumMany other people assisted in the field and I am grateful to them all: Bernard Splechna, DavidNew, Jaroslav Dobry, Pal Varga, Bob Brett, Annemeike Smit, Martin Goosens and Meike Seele.I have special thanks to Bob Brett, Gordon Kayahara, Pal Varga, and the other “Klinkoids” foryour endless assistance with computers, energetic discussions and debates, and for not giving methe nerd award.And finally, to readers of my thesis, may the forest be with you.421dd)Lori D. DanielsJune 30, 19941Chapter 1. IntroductionThe Study of Forest DynamicsForest ecosystems are dynamic. The study of ecosystem dynamics involves examiningchanges in ecosystems through time and space and determination of the factors driving suchchanges. The concept of succession is integral to such studies and has played a central role in thescience of plant ecology (Peet and Christensen 1980).The focus in the study of forest dynamics has been historically on species compositionand changes in composition through time. The traditional (“holistic”) approach to understandingvegetation dynamics describes succession as “ecosystem development” (Clements 1916, Odum1969). According to this view, successional change occurs in waves; the process is controlled bythe vegetation itself. Colonization of a site alters its characteristics, with both the abiotic andbiotic components of the ecosystem modified by the establishment and growth of the colonizersor pioneer species. These changes can make the site unsuitable for perpetuation of the pioneers,but ameliorate conditions for other species and enable invasion by new species. A shift inspecies composition in this way subscribes to the concept of forest succession as relay floristics(Egler 1954) or facilitation (Connell and Slatyer 1977).Accompanying this progression of species and amelioration of their environment is theexpression of the ecosystem’s emergent properties (McIntosh 1981). Of key importance isstability, which is maximized as succession culminates in a climax community. The climax stateconsists of an undisturbed, relatively stable species composition, best adapted to the prevailingabiotic conditions and capable of self-perpetuation through time (Clements 1916, Tansley 1935,White 1979, Kimmins 1987).The species forming stands of the latter stages of ecosystem development are termedlate-successional, sub-climax, and climax species. The stability of their populations can beassessed by the shape of their age or size distribution curves. Those forming an inverse-J shape,with many individuals in the youngest or smallest classes and progressively fewer in the older2and larger classes, are considered stable (Leak 1965, Heft and Loucks 1976). The shape of thiscurve provides evidence that there is sufficient regeneration within the subcanopy to replacesenescent canopy dominants (Whipple and Dix 1979).In recent decades, the study of forest dynamics has shifted from the analysis ofcommunity composition to studies of the mechanisms and processes governing succession. Theemphasis of this “reductionist approach” is to examine and understand the adaptations andcharacteristics of species and populations (Drury and Nisbet 1973, Peet and Christensen 1980).These are then extrapolated to the community and ecosystem levels and integrated to interpretsuccession.Reductionists view ecosystems as assemblages of populations. According to this view,ecosystem succession is a population process, in which a sequence of these assemblages occupiesand dominates a site over time. The occurrence of a species within the sequence is determinedby its life history characteristics which are evolutionarily controlled, adaptive traits (Pickett1976, Connell and Slatyer 1977) that define the species’ responses to the biotic and abioticinfluences of its environment. These characteristics govern reproduction, establishment, growth,and mortality, for example (Peet and Christensen 1980). The arrangement of species within thesuccessional sequence is a consequence of the variation among species with respect to thesetraits.The prevalent influence of disturbance in many ecosystems and on vegetation dynamicsis now recognized (White 1979, Oliver 1981). In keeping with the reductionist view ofvegetation dynamics, disturbance, through its influence on populations, can have significantinfluence on community composition and succession. Pickett et al. (1987) claim that “the natureof disturbance must be considered when succession is to be understood” (p. 111) and that “forsome questions concerning successional mechanisms and causes, it will be necessary to gobeyond [current] models” (p. 12). Studies investigating the sources of environmentalheterogeneity that influence the patterns of species’ establishment and survivorship andcommunity composition are now widespread in the field of vegetation dynamics (e.g., Kozlowski3and Algren 1974, Oliver and Stephens 1977, White 1979, Turner and Franz 1985, 1986, Stewart1986a). In many forest ecosystems, small scale disturbances and canopy gaps are an importantsource of environmental heterogeneity (e.g., Henry and Swan 1974, Runkle 1981, Whitmore1982, 1989, Lertzman 1989) and have been demonstrated to be of considerable influence on thepopulation dynamics of trees (e.g., Fox 1977, Stewart 1986a, 1986b, 1986c, Peet and Christensen1987, Runkle 1981, Lertzman 1989, Veblen eta!. 1989).The Dynamics of Thuja plicata-Dominated Old-Growth StandsIn the context of this thesis, old-growth is defined on the basis of the process by whichtrees comprising the canopy are replaced (Oliver 1981, Runlde 1981). In old-growth stands,mortality of canopy dominants occurs on an individual or tree-by-tree basis. Canopy individualsover an area are gradually replaced through ingress and the recruitment of advanced regenerationor subcanopy trees (Watt 1947, Peet and Christensen 1987). Old-growth stands in which somecanopy dominants are relicts of the cohort establishing on the site after a catastrophic disturbanceare “transition old growth”; stands that contain only trees that established beneath an existingcanopy are “true old growth” (Oliver and Larson 1990).The processes of tree replacement in, and development of, old-growth stands areindependent of species composition and compositional changes (Oliver and Larson 1990). Achange in the stage of development within a stand (e.g., the shift to single tree replacementprocesses and the old-growth stage of development) does not necessitate a change in speciescomposition. Replacement of canopy trees can include species self-replacement, reciprocalreplacement among coexisting species, cyclical replacement, or a shift in species composition(succession)(White 1979).At submontane elevations (<ca. 600 m) in the wet cool mesothermal forests nearVancouver, the conifer species, Thuja plicata, Tsuga heterophylla, Abies amabilis, andPseudotsuga menziesii (taxonomic authorities according to Hitchcock and Cronquist 1973 andSchofield 1992) compose old-growth canopies on most sites. Although such old-growth stands4are common and cutting of these forests is widespread, knowledge about their structuraldevelopment and dynamic processes is limited (Oliver et a!. 1988). It is speculated that forestsin this region originated after catastrophic disturbances, such as fire or landslides. For example,Pojar et a!. (1991) state that Thuja-dominated stands “typically occupy steep slopes or sites thatregenerated following fire or have not been disturbed by windthrow for many centuries” (p. 98).In these stands, Pseudotsuga requires exposure and is incapable of regeneration in theunderstory (Klinka et al. 1990, Carter and Klinka 1992). It establishes in open environmentscreated by fire and, due to its stature and longevity, is able to persist and dominate old-growthstands (Munger 1940, Franklin and Hemstrom 1981, Kuiper 1988). Thuja, Tsuga and Abies areexposure-tolerant with vigorous growth in the open, while shade tolerant and well adapted to theunderstory environment (Klinka et al. 1990, Carter and Klinka 1992). Krajina (1965, 1969)considered Tsuga and Abies well adapted to the climate and edaphic conditions that prevail inthese forests and regarded them as climatic climax species. Thuja is believed to be self-perpetuating and capable of sustained domination of stands only on poorly drained or nutrient-rich and very rich sites and is considered an edaphic climax species (Krajina 1969, Franklin andDymess 1973). In old-growth stands on sites of poorer nutrient status or drier moisture regimes,Thuja’s dominance is attributed to its large size and longevity (Turner and Franz 1985, Keenan1993), as is the case with Pseudotsuga.Many characteristics are attributed to climax tree species. These include shade tolerance,longevity, relative large size at maturity, slow growth, and the ability to regenerate in theunderstory (Krajina 1969, Whittaker 1975, White 1979, Oliver and Stephens 1977, Minore 1983,Oliver et al. 1988, Graham et a!. 1988). All but the latter are well documented for Thuja (e.g.Minore 1983, Carter and Klinka 1992, Daniels et al. 1994). Size frequency distributions of thisspecies do not fonn an inverse-J shaped curve, as regeneration of Thuja beneath the canopy inold-growth stands is scarce or lacking, even where this species is a canopy dominant. This hasbeen interpreted as instability within the population (Schmidt 1955, Gregory 1957, Franklin andDymess 1973) and has provided evidence in support of the assertion that Thuja is not self-5perpetuating throughout the submontane cool wet mesothemial forest. The same criteria havebeen applied in classifying Thuja as a subclimax species in the interior portion of its naturalrange (Daubenmire 1966, Daubenmire and Daubenmire 1968, Habeck 1968, 1978).A preliminary study of the age and size distributions of Thuja trees (diameter at breastheight [dbh] 10 cm) in the study area did not support the assertion of Thuja decline. It revealedthat Thuja in the tree layer are broadly uneven-aged (Daniels et al. 1994). The age distributionswere characterized by distinct cohorts and indicated that regeneration was sporadic but continualover the long-term, a population structure that Whipple and Dix (1979) suggested was typical ofpopulations affected by disturbance. Recent examinations of the population dynamics of Thujaon Vancouver Island (Keenan 1993) and in Idaho (Parker 1986) also raise issue with its currentclassification as an edaphic climax species. Others have reported sustained animal and plantpopulations that do not form inverse-J age and size distribution curves (Pearl 1928, Deevey 1947,Whittaker 1953). In light of these findings, this study analyzes the population dynamics of Thujain association with Tsuga and Abies in old-growth forests in the cool wet mesothermal forests atsubmontane elevations in coastal British Columbia to test the null hypothesis that Thujapopulations are not sustained.Objectives and Thesis OutlineThis study employs a reductionist approach to re-evaluate the successional role of Thujain old-growth stands. It aims to describe the species composition, structure and dynamics ofeight old-growth stands and to resolve whether the populations present will be sustained withtime. Chapter 2 identifies the study area, the study design and methods. The age and sizestructures of the tree species populations are examined and compared to provide interpretationsof the dynamics of each stand (Chapter 3). The abundance and distribution of naturalregeneration (seedlings and saplings) of Thuja were investigated to characterize the conditionsunder which this species establishes beneath the forest canopy and to compare the regenerationniches of Thuja, Tsuga, and Abies (Chapter 4). Dendrochronological techniques were applied to6investigate the growth dynamics of canopy trees and to identify the processes controlling changewithin the canopy (Chapter 5). Chapter 6 is a summary of my conclusions about the dynamics ofThuja in the study area. It focuses on the life history characteristics and adaptations of Thuja andrelates these to the local disturbance regimes and the population age and size distributions ofthese old-growth forests.7Chapter 2. MethodsStudy AreaThe study was conducted in the Capilano and Seymour watersheds, an area of 37, 510 hamanaged by the Greater Vancouver Water District (GVWD) for water supply to the greaterVancouver municipalities. The study area was located within the Very Wet Maritime CoastalWestern Hemlock (CWHvm) biogeoclimatic subzone, which delineates the influence of a verywet, cool mesothermal climate (Klinka et al. 1991). Mean annual temperature of the region is8.2°C, with a mean temperature in the wannest month of 16.0°C and in the coldest month of 0.3°C. Annual precipitation averages 2787 mm, 752 mm of which falls during the summer growingseason. During the driest summer month, mean precipitation is 75 mm, with an average of onlyeight days of measurable precipitation (Klinka et a!. 1991, Canadian Climate Program 1993).Soils are steep-slope, coarse-skeletal, Ferro-Humic Podzols (Agriculture Canada ExpertCommittee on Soil Survey 1987) derived from colluvium and glacial till, and underlain bygranitic rocks.Stands in which Thuja, Tsuga, Abies, and Pseudotsuga compose the tree layer, withTaxus brevifolia in the lower canopy, are prevalent on submontane sites in the watersheds wherelogging has not disturbed the forest cover. The canopy species are long-lived and capable ofgrowing to large sizes. These stands are typically multi-layered with tall emergents, canopygaps, snags, and accumulations of large coarse woody debris on the forest floor. Most extantstands are considered old-growth due to these structural characteristics, the dominance of treesgreater than 200 years of age (Hamilton and Pojar 1991, Beese and Sandford 1992, BCMOF1992), and the replacement of canopy dominants by tolerant species occupying the understory.Understory species associated with these stands include: Vaccinium alaskaense, V.ovalfolium, Menziesiaferruginia, Blechnum spicant, Cornus canadensis, Hylocomiumsplendens, Rhytidiadeiphus loreus, and Rhytidiopsis robusta (Appendix A). Predominant siteassociations of the stands in this study are Tsuga heterophylla-Abies amabilis-Vaccinium (HwBa8Blueberry), Tsuga heterophylla-Abies amabilis-Blechnum spicant (HwBa-Deerfem), Abiesamabilis-Thuja plicata-Tiarella (BaCw-Foamflower), and Abies amabilis-Thuja plicata-Rubusspectabilis (BaCw-Salmonberry)(Banner et at. 1990).Study DesignThis examination of the dynamics of the watershed forests employed mensurativemethods. The time required to address questions about pattern and process in old-growth forestsusing manipulative methods was beyond the limits of this study. A “state-factor control” (VanCleve et at. 1991) framework was applied to permit a pseudo-experimental study of theecological processes within these forests. This approach defines five properties of a region:regional climate, biota, topography (elevation, slope, aspect), soils, and temporal development.These determine the type of ecosystems that exist within the region and the phenomena definingecosystem stmcture and function (Jenny 1941, Van Cleve et at. 1991). To “experimentally” testhypotheses, ecosystems were compared that were similar except with respect to one of the aboveproperties. The role of that property in ecosystem structure and function was explored.All study plots were established in forest stands in the East Capilano Creek valley of theCapilano watershed, except one plot in the Seymour watershed (Figure 2.1). By locating theplots in close proximity to each other, regional climate, biota, and soil parent material werenaturally controlled. All study plots were located on midslopes, with dominant slopes of 25 -45%, in an elevation range from 335 to 540 m. All had fresh or moist soil moisture regimes, andmedium or rich soil nutrient regimes (Klinka et a!. 1989). Stands located on opposite sides ofthe valley, three south-facing and four north-facing, were selected to examine the influence ofaspect (local climate) on the forest dynamics. The stand in Seymour watershed was selected tomatch these site characteristics and was south-facing (Table 2.1).9(a)(b)Figure 2.1. Location of study stands (a) and areas (b) near Vancouver, British Columbia.Squares representing plot locations in (a) are not to scale.10Table 2.1. Location and description of old-growth forest study stands; FF = forest floor, MS =mineral soil, DW = decaying wood, SMR = soil moisture regime, SNR = soil nutrient regime.Site Elevation Aspect Slope Ground Cover (%) Fire1 SMR2 SNR3 Site Association4(m) % FF DW MSSi 500 S 38 43 41 11 No F M HwBa-BlueberryS2 540 SSE 39 55 42 3 Yes M M HwBa-DeerfernS3 530 S 28 46 52 1 Yes F M HwBa-BlueberryS4 335 SSE 27 71 22 6 No M M HwBa-DeerfernNi 430 NNW 26 55 37 5 Yes F R BaCw-FoamflowerN2 440 NNW 31 37 46 13 tr. F M HwBa-BlueberryN3 465 NNW 32 56 40 3 tr. M R HwBa-DeerfemN4 475 NW 44 66 28 5 tr. F M HwBa-Blueberry‘Fire: Yes = presence of continuous charcoal horizon, tr. = isolated observation(s) of charcoal25j abbreviations: F = fresh, M = moist3SNR abbreviations: M = medium, R = rich4from Banner et al. 1990Temporal development, the fifth regional property defined above, is a function of thedisturbance history and subsequent stage of development of an ecosystem (Van Cleve et al.1991). As this study focused on old-growth forests, stand composition and dynamics wereassumed to be closest to equilibrium with the prevailing regional climate rather than a reflectionof past stand influencing disturbances. Five criteria were set to address the temporalcomparability of the study stands: (1) dominance by Thuja with individuals >200 years of age,(2) consistent species composition, (3) diverse stand structure, indicative of old-growth dynamicprocesses (Oliver and Larson 1990), (4) minimum variation in site quality, and (5) lack of11anthropogenic disturbance. These criteria reduced the variation in species composition andstructure among the study stands; however, the stage of temporal development could not becontrolled in the selection of study stands, but was determined in the course of the study.Ideally, replicates of stands of the same history (type of disturbance and time since the mostrecent disturbance) should be studied and compared to fulflll a statistically robust experimentaldesign.Data CollectionWithin each selected stand, a square 0.49 ha study plot was established. Each measured70 m by 70 m, composed of 49 contiguous 100 m2 subplots. This size was greater than the 0.25ha plots that Mueller-Dombois and Ellenberg (1974) suggest as sufficient to capture the variationin mature coniferous forests; however, larger plots were necessary to provide sufficient samplesof canopy trees to characterize stand size distributions (Lorimer 1985) and to assess thepopulation dynamics of Thuja in these stands. Plot location within the stands was subjective tomeet the study plot criteria stated above.Location, elevation, and aspect were noted for each plot. Ground cover was estimatedalong four line transects, two parallel to the slope and two perpendicular to the slope (afterRebertus et a!. 1991). The extent along each line of five ground cover types: water, coarsefragments, forest floor, decaying wood (at or within 7 cm of the surface), and mineral soil(exposed or covered by a layer of living bryophytes and litter of depth <7 cm) was recorded. Asoil pit and eight soil trenches (approximately 0.5 m deep and 0.5 m long) were systematicallylocated on each plot. Soil profile descriptions (Agriculture Canada Expert Committee on Soils1987) and evaluation of soil moisture regimes (SMRs) and soil nutrient regimes (SNRs) (Klinkaet at. 1989) were derived from the soil pits. The pits and trenches were examined for charcoal,with particular attention to the humus-mineral soil interface (Lorimer 1985, Table 2.1).Tree species were sampled at three scales within each plot. For all trees (dbh >10 cm)species and location (X-Y coordinates of the centre of each stem estimated to the nearest 0.1 m12with meter tapes and the flagged grid that demarcated the subplots) were noted. Diameter atbreast height (outside bark) was measured to the nearest 0.1 cm with a diameter tape. Heightclass, a measure of tree position in the canopy relative to its neighbours, was subjectivelyassessed: classes 1 and 2 = upper canopy, 3 and 4 = lower canopy. Vigour was assessed on ascale of 0 to 4 (after Luttmerding et al. 1990, Carter and Klinka 1992):0- individual dead1 - poor: individual not expected to survive more than 2 years; evidence of poor foliageretention, dieback of leading shoots, arrested growth or loss of apical tendency, chiorosis2- fair: individual showing some above described symptoms but continues to producenew foliage and annual increment3 - good: individual showing adequate height and diameter growth and none of the abovedescribed symptoms4 - excellent: individual showing height and diameter growth commensurate with sitequality and none of the above described symptomsTsuga and Abies saplings (height >130 cm, dbh l0 cm) were sampled within 16subplots measuring 10 m by 10 m, and seedlings (with mature needles [Franklin 1961], height130 cm), within 25 2 m by 2 m quadrats. All subplots and quadrats were systematically locatedwithin the plots. Due to the scarcity of Thuja regeneration (with mature foliage [Franklin 1961],dbh 10 cm), all individuals within the study plots were located and measured. On plots Si, S2and S3, the central 50 m by 50 m grid of the study plots provided sufficient sample sizes forThuja tallies.For all regeneration (seedlings and saplings), the species and location (subplot or quadratnumber, X-Y coordinates [see above]) were noted. Stem diameter at ground level was measuredto the nearest 0.1 mm with a digital micrometer. The vigour of each was subjectively assessed(see above). Average vigour beneath an old-growth canopy was considered fair. Each individualwas examined to determine its rooting substrate (forest floor, decaying wood, and mineral soil)and, for seedlings, to determine whether reproduction was from seed or by vegetative means.13Understory vegetation was assessed within the 2 m by 2 m quadrats. All non-treespecies were identified, their percent cover estimated and they were grouped by growth-form:shrubs, herbs and bryophytes (Appendix A).On four study plots (Si, S4, Ni and N4), the analysis of tree population dynamicsincluded height and age structures. Height of trees was calculated to the nearest 0.1 m fromclinometer measurements taken at distances similar to the height of the trees. Height of seedlingsand saplings was measured with a meter tape to the nearest centimeter and nearest 0.1 m,respectively.To determine tree age, increment cores were extracted from trees using the methods ofJozsa (1988). Due to time constraints, only trees in the central 50 m by 50 m grid of the studyplots were cored. Cores were sampled as close to the ground as possible. Stand Ni was loggedin February 1992; thus, numbered tags were attached to the base of canopy trees prior to loggingand the ages of the tagged trees were determined by counts of annual rings on the cut stumps thefollowing spring. For these trees, suppression and release periods were visually assessed and theradial length and ring counts of each period were recorded to compare and contrast growth rates.Ages were not adjusted to account for the height from which cores were taken nor the height ofstumps. Subsets of seedlings and saplings were selected for each species to represent the rangeof heights and diameters observed on each site. Stem disks extracted at ground level were usedfor age determination. Ring counts of the stem cross sections and increment cores wereconducted using a 40X dissecting microscope. The surface of cores and disks were prepared witha razor blade and zinc oxide was used to improve ring visibility.14Chapter 3. Stand StructureIntroductionStructural attributes of forests can be used to understand dynamic processes, to interpretdevelopment of stands, and to predict their future composition and structure (Oliver and Larson1990). Age and size (diameter) structures, the relative abundance of trees of different age andsize classes (Veblen 1986), are frequently used in forest studies (e.g., Whipple and Dix 1979,Ross et al. 1982, Stewart 1986a). Age structures reflect species and stand origins (Veblen 1986)and provide some evidence of past stand development by revealing even-aged stands and cohorts(Lorimer 1985). Mthough tree size is usually more easily determined than age, size may not besuitable for interpretation of stand development and past dynamics. Unless the relationshipbetween tree age and size is strong, interpretation of stand size structures can be problematic(Veblen 1986). The relationship between tree size and age has frequently been shown to be weakor sporadic, or variable among species and sites (Stewart 1986a, Daniels et al. 1994). Thisstudy’s first objective is to quantify the relationship between tree age and size on a species- andstand-specific basis and to compare the age, diameter and height structures of the tree populationsin eight old-growth stands. The aim is to assess the applicability of these structural attributes tothe interpretation of the dynamics of these stands.Regeneration ability of species within forests is manifested in stand structure (Veblen1986, Hofgaard 1993), in that the spatial (horizontal) distribution of trees can reflect dynamicprocesses (Veblen et al. 1980). Differences in pattern among species and trees of different sizesreflect unique regeneration needs and development processes. A second objective of this study isto characterize the spatial disthbution of trees of different sizes (height classes) to investigate thespatial scale at which dynamic processes are active in the study stands.15MethodsData AnalysisSize structures were analyzed by diameter and height frequency distributions in whichcanopy trees were grouped in 15 cm diameter-classes (from 10 cm) and 5 m height-classes (from5 m). Trees of dbh 10 cm and of height 5 m were tabulated. Age structures of canopy trees(dbh >10 cm) were analyzed using 50 year age-class frequency distributions. The number ofseedlings and saplings of each species in each age-class were tabulated as follows. To estimatethe number of seedlings within a given age-class, the percentage of seedlings of known age andwithin the age-class limits (e.g., 50 years for age-class 50) was multiplied by the total numberof seedlings. This procedure was repeated for each age class and for saplings in each age class.For each species and age class, the estimated number of seedlings and saplings were summed toproduce the species-specific age-class distributions for trees with dbh lO cm.Complete age counts of some trees were missed because countable cores could not beextracted, cores missed the pith by a large distance, or increment borers were not long enough toreach the pith of the stem. Age prediction models were derived to estimate total ages and theages of the portions of trees that were not sampled. For Tsuga and Abies trees, stand-specificequations of age on dbh and height were derived by forward and backward stepwise regressionanalysis (Zar 1984) using data from trees from which complete cores had been extracted. As fewcomplete cores were extracted from Thuja, regression equations for this species were derivedfrom data collected in a preliminary study of its age structure (Daniels et a!. 1994). Indicatorvariable analysis (Cunia 1972) was used to compare the age-dbh relationships of Thuja in the treelayer of four sites in the East Capilano Creek area and two sites in the Balfour Creek area, fromwhich two area-specific models were developed. For some models, both age and diameter werelogarithmically transformed (ln[agej = b0 + b1 x ln[dbhj) to meet the homogeneity of variancesassumption of regression analysis (Zar 1984). The standard errors of estimate (SEE) of thesemodels were back transformed to represent their variation in the original units (years).16A simple model was developed to investigate species’ longevity and density in relation topopulation dynamics. These two attributes are identified by Horn (1981), White eta!. (1985) andLertzman (1989) to potentially have significant influence on population and stand dynamics.The frequency of recruitment of each species was calculated as the ratio of potential longevity totree density in the upper canopy and was converted to a mean number of recruits per 50 years(mean number of “required” recruits to maintain observed canopy composition). These valueswere compared to the average number of canopy trees in each 50 year age-class, calculated fromthe age distributions. The mean ages of the oldest Tsuga and Abies in each of the intensivelysampled stands were 379 and 272 years, respectively, and represented the longevity of thesespecies. Longevity of Thuja was estimated as 765 years, the mean of the oldest Thuja on 15 sitesin the study area that were sampled by Daniels et al. (1994) in a preliminary study. Theseestimates of longevity were considered conservative, as they were derived from living trees. Thefrequency of recruitment was determined for each plot independently.The spatial distributions of upper and lower canopy trees, within the intensively sampledplots, were assessed to characterize the horizontal structure of each plot. Stewart’s (1986)adaptation of the nested-quadrat technique (Kershaw 1973, Greig-Smith 1983) was used to detectscales of pattern. Morisita’s (1959) index of dispersion (Id) was used to detect departure from arandom distribution:(n —1)iiN(N-l)where: q = number of quadrats, n1 = number of trees in the 1th quadrat, and N = total number oftrees in all quadrats.Morisita’s index equals 1.0 when the population is randomly distributed. Its numericalvalue is proportional to the degree of aggregation or dispersion of the population. It is <1.0 when17individuals are uniformly spaced and >1.0 when individuals are aggregated. The statisticalsignificance of the index is evaluated by a Chi-squared (X2)test statistic (Greig-Smith 1983):x2 =1, *(N_1)+q_Nwhere: q = number of quadrats, N = total number of observations in all quadrats. The calculatedtest statistic was compared to a critical X2 (a = 0.05) with q-1 degrees of freedom.Morisita’s index is independent of population density and sample size. These featuresmake it comparable among groups of individuals that are sampled with quadrats of the same sizeor within groups that were sampled at different quadrat sizes (Huribert 1990).The degree of aggregation among individuals is a function of spatial scale (Pielou 1977,Huribert 1990). To detect patterns of distribution within populations, Morisita’s index wascalculated for quadrats that measured 16, 64, 256 and 1024 m2, with each quadrat createdthrough a combination of 4 m x 4 m units while the square shape was maintained (Pielou 1977).The index was calculated only when sample sizes were 20.It was hypothesized that subcanopy trees would be aggregated where canopy stemdensities were low, indicative of establishment in canopy gaps (Veblen et a!. 1979, Stewart1986a). The relationship between lower canopy and upper canopy tree location was tested by 2 x2 contingency tables and the Chi-squared test statistic with Yate’s correction (Zar 1984) forsample sizes <500. Statistical tests were conducted only when the number of observations withineach cell of a contingency table was 5 (Mueller-Dombois and Ellenberg 1974, Zar 1984).For all statistical tests, significance was set at a = 0.05.ResultsSpecies CompositionThe canopies of all eight stands included emergent Thuja. This species was moreabundant in the canopy of south-aspect compared to north-aspect stands. It was sparse in thelower canopy, sapling and seedling strata of all stands, except in stand S3. Abies co-dominated18the lower canopy and was prolific in the understories of north-facing stands and in stand S4. Instands Si, S2, and S3, Abies was a minor component of the lower canopy and understory strataand occurred in the upper canopy of only one of these stands. Tsuga was abundant in mostcanopy strata and understories. Pseudotsuga was present in the upper canopy of stands S2 andS3. Taxus was a minor component of the lower canopy of the four south-facing stands, but wasnot found in north-facing stands (Table 3.1).Table 3.1. Species composition (number per hectare) by canopy stratum in the eight studystands.STRATUM STANDSPECIES Si S2 S3 S4 Ni N2 N3 N4CANOPYThujaplicata 53 31 33 27 16 16 4 47Tsugaheterophylia 43 41 76 108 63 89 41 37Abies amabilis 4 0 0 12 35 12 33 8Pseudotsuga menziesii 0 27 6 0 0 0 0 0SUBCANOPYThujaplicata 6 14 53 4 2 0 0 4Tsuga heterophylla 218 278 227 90 139 147 145 73Abiesamabiis 35 12 24 222 4 43 141 122Taxusbrevifolia 10 20 45 2 0 0 0 0SAPLINGSThuja plicata 49 64 72 0 0 11 2 8Tsugaheterophylla 1069 1200 1006 525 1144 825 1644 775Abiesamabilis 138 13 31 1301 81 689 781 1600SEEDLINGSThujaplicata 1303 1649 2743 426 121 508 366 302Tsugaheterophylia 25500 14800 29200 21400 11800 20900 26800 12000Abies amabiis 300 300 200 10600 14100 4300 12100 820019Diameter DistributionsThe diameter distributions of the tree species in seven stands (Si, S3, S4, Ni, N2, N3and N4) formed an inverse-J; the diameter distribution of stand S2 was relatively uniform (Table3.1, Figure 3.1). The structure of the former stands was dictated by the distribution of Tsuga.Abies also formed an inverse-J distribution, except in stand Ni where its distribution wasbimodal. In stands Si, S2 and S3, this species had little influence on stand structures, but itdominated the smallest diameter classes in stands S4, N3 and N4. The majority of Thuja haddbh’s >100 cm and were the largest trees in each stand, except where Pseudotsuga was present.The occurrence of Thuja trees, dbh >10 cm, was irregular except for stand N3 in which no Thujadbh 175 cm were present, and stand S3 in which the distribution of this species was relativelyuniform.The diameter distribution of stand S2 was distinguished from the others by the relativelyuniform distribution of trees with dbh’s >25 cm (Figure 3. ib). Abies was a minor component,present only in the smallest diameter-classes. Thuja and Pseudotsuga accounted for allindividuals with a dbh 1 15 cm.Relative to the other stands, Ni had fewer trees in the 25 cm class (Figure 3. id). Abieswas scarce in the sapling layer, relative to other north-aspect stands, and was absent from the 25and 40 cm classes, but well represented in the 55 to 100 cm classes. There was one recruit ofThuja in the 40 cm class; the other Thuja had the largest diameters in the stand.Age DistributionsNo significant differences were identified in the age on dbh relationships for Thujaamong the four sites in the East Capilano Creek area (F = 1.22, p = 0.3 11) and between the twosites in the Balfour Creek area (F= 0.08, p = 0.926); thus, two area-specific equations werederived to estimate the age of Thuja trees. Regressions of age on dbh and height revealedstatistically significant relationships for Abies and Tsuga, except Abies in stand S4, for which nolinear relation was identified. The variation associated with all equations was high (Table 3.2).Figure 3.1. Diameter distributions of canopy trees (dbh >10 cm) in the eight study stands (from(a) to (h): Si, S2, S3, S4, Ni, N2, N3, and N4). Trees are grouped in 15 cm classes from 10 cm.In the legend, TP = Thuja plicata, PM = Pseudotsuga menziesii, TH = Tsuga heterophylla andAA = Abies amabilis. Values for seedlings and saplings (dbh l0 cm) are given in Table 3.1 (p.2011L.I25 40 56 70 86 100 115 130 145 180 176 1901867O60-50-4030-20 -10o26 40 55 70 85 100 115 130 146 160 115 190iH1417060p50.40Ci) 30, 20-10-0(a)1457060-I! 60-40-30-20-10-0(c)8070606040Ci, SO20100(e)11726 40 68 70 86 100 116 130 145 160 178 190(b)(d)(f)(h)-70-6050-40-co 30-2010-0I flA2 40 65 70 86 100 115 130 146 160 175 190! !%fl -7060•504030- J1Jl,ii.. Iii(g) Upper Limit of Ctametor Class (cm)• TPPMAAO TH25 40 66 70 86 100 116 130 145 160176 190Upper Limit of arrte CLass (cm)18).21Table 3.2. Species-specific regression models of age (years) on dbh (cm) and height (m) for thefour intensively studied stands. Area-specific models were derived for Thuja; stand-specificmodels for Tsuga and Abies. Transformations were applied where required to meet theassumptions of this technique.StandlArea Model n R2 p-value SEE(years)Thuja plicataCapilano In (age) = 1.004 + 0.778 [in (dbh)] 164 0.73 <0.001 103Seymour ln(age)= 1.129+O.688[ln(dbh)] 78 0.66 <0.001 122Tsuga heterophyllaSi age=67.146+3.360(dbh) 45 0.50 <0.001 53S4 age = 104.699 + 2.149 (dbh) 39 0.34 <0.001 72Ni in (age) = 2.765 + 0.601 [In (dbh)] 43 0.73 <0.001 41N4 age = 79.447 + 3.723 (dbh) 22 0.71 <0.001 47Abier amabilisSi age = 6 1.341 + 8.906 (dbh) - 9.240 (height) 9 0.66 0.041 26S4 NONENi age = 52.416 + 1.864 (dbh) + 0.306 (height) 13 0.43 0.013 35N4 age=i01.022+2.681(dbh) 38 0.36 <0.001 46The age distributions of stands Si, S4, and N4 were considered inverse-J and that ofstand Ni, bimodal (Figure 3.2). Trees of diameter l0 cm were up to 171 years of age and werenot included in Figure 3.2, although their influence on the form of the age distribution wasincluded in the classification of these curves. Age distributions for seedlings and saplings aresummarized in Table 3.3. Tsuga populations formed an inverse-J with a maximum age of 328 to424 years, depending on the plot. The age distributions of Abies also formed an22807060504030(a)20100’100 150 200 250 300 360 400 460 600 660 600 660 700(b)I I I I I I I I I I100150200250300 360 400 450 500 550 500 00Upper Limit of Ae Class (years)Figure 32. Age distributions of canopy trees (dbh >10 cm) in the four intensively sampledstudy stands (from (a) to (d): Si, S4, Ni, and N4). Trees are grouped in 50 year age classes from50 years. In the legend, TP = Thuja plicata, TH = Tsuga heterophylla, and AA = Abiesamabilis. Values for seedlings and saplings (dbh iO cm) are given in Table 3.3 (p. 23).I . Ii100150200250300360400 450 600 560 600 660 10080’7060’:1 50’4030I-2010’0’(c)80706050’40302010ii lIk III I I I I I I I I I100 150200250 300 350 400 450 600 560 600 650 700Upper Limit of Age Class (years)i • I(d)I TPAAo TH23Table 3.3. Summary of seedling and sapling age- and height-class distributions (number / ha).Stand Upper Limit of Age Class (years) Height Class (m)50 100 150 200 1.3 1.3-5Thuja plicataSi 1295 4 0 0 1303 55S4 310 0 0 0 302 12Ni 121 0 0 0 121 0N4 426 0 0 0 426 0Tsuga heterophyllaSi 23678 2752 132 23 25500 1006S4 21705 203 17 0 21400 488Ni 12736 208 0 0 11800 1119N4 12284 439 52 0 12000 756Abies amabilisSi 313 75 38 13 300 119S4 8874 3445 294 0 10600 1888Ni 14159 22 0 0 14100 75N4 6093 3479 152 76 8200 1575inverse-J; however, this species had greater abundance in stands S4 and N4, having considerableinfluence on the age structures of these stands. The maximum age of Abies were 193, 327 and295 in stands Si, S4 and N4, respectively. In stand Ni, Tsuga lacked individuals in the 100 yearage-class and Abies was absent from the 100 and 150 year age-classes. Both species werepresent in the 200 to 300 year age-classes.24The age distributions of Thuja were distinct from those of Tsuga and Abies. Individualsin the understory were up to 89 years of age with most less than 50 years (Table 3.3). Canopytrees, regardless of their large size, were of a broad range of ages that were irregularly distributedamong size-classes. No more than two individuals within a 0.25 ha plot occurred within anysingle age-class. Only one distinct cohort of Thuja was identified, in which four individuals aged462 to 534 years were responsible for the bimodal age distribution in stand Ni.The results of the longevity-density model (Table 3.4) revealed that Tsuga required thegreatest number of recruits per 50 year period in order to maintain the observed upper canopyspecies composition. The low number of recruits required to maintain Abies in stands Si, S4 andN4 is presumably related to its low density in the study area. The higher density of Abies in theupper canopy of stand Ni was reflected in the number of recruits required to sustain itspopulation. The fact that fewer recruits of Thuja were required is a function of its longevity.This phenomenon was evident in stands Si and N4 where the upper canopy density of Thuja wasequal to or greater than that of Tsuga; however, the number of recruits to maintain the density ofThuja was less than the number required to sustain Tsuga.The second trend revealed by this model was that the mean number of Tsuga and Abiestrees per 50 year age-class was orders of magnitude greater than the minimum number of recruitsrequired to maintain their upper canopy density. Exceptions were Tsuga in stand S4 and Abies instand Ni; in both cases the canopy tree density was significantly higher than in the other studystands. For both species, this relationship appeared to be age-dependent, with the differencesbetween the number of required and observed recruits diminishing with successive age classesuntil the frequency was less than the mean required number of recruits. In contrast, the requiredand observed recruits of Thuja per 50 year class were similar for all four stands, and given theflat age distributions for Thuja, this relationship did not appear to vary with tree age.25Table 3.4 Comparison of the density (trees I ha) and the recruitment required for the canopymaintenance of the three study species in the four intensively studied stands.Stand Current Recruits ha -1 50 yrTree Densityin Upper Canopy “Required” to maintain Mean observed frequencyobserved density per 50 year age-classThuja plicataSi 46.94 3.07 2.57S4 26.53 1.73 2.86Ni 16.33 1.07 1.82N4 46.94 3.07 3.27Tsuga heterophylla51 42.86 5.65 28.50S4 108.16 14.27 17.00Ni 63.27 8.35 27.30N4 36.74 4.67 13.50Abies amabilisSi 4.08 0.75 10.00S4 12.24 2.25 25.50Ni 34.69 6.38 4.70N4 8.16 1.50 22.3026Height DistributionsThe height distributions of the intensively sampled plots were bimodal (Figure 3.3). Thebreaks in the distributions delineated the upper canopy from the lower canopy. The height-classdistribution of trees in the lower canopy formed an inverse-J, except in stand Ni where there wasa conspicuous absence of trees in the smallest height-classes.Thuja dominated the upper canopy on all sites. It rarely occurred in the lower canopy.Tsuga prevailed in the subcanopy of all plots. Abies co-dominated the lower canopy in stands S4and N4.Spatial DistributionsMorisita’s index, calculated for the upper and lower canopy strata, was not significantlydifferent from 1.0 in any stand nor at any spatial scale analyzed (Figure 3.4). Nor were spatialcorrelation between these strata detected (X2yates 3.84, p> 0.05 for all associations tested).That is, there was no evidence to reject the null hypothesis that stems were randomly dispersed atall scales.27117 I I 236— I80 8070 70ID60 60-50 5040 4030 30flR[I,i_______1015202530354045505560 1015202530354045505560(a) (b)I I I I I I I I I I I I I I I I I I80 8070 700,eo 6050 5040• 40•30 30H, R11015202530354045505580 1015202530354045505560Upper Limit of Height Class (m) Upper Limit of Height Class (m)(c) (d)Figure 3.3. Height distributions of trees (height >5m) in the four intensively sampled studystands (from (a) to (d): Si, S4, Ni, and N4). Trees are grouped in 5 m height-classes from 5 m.In the legend, TP = Thuja plicata, TH = Tsuga heterophylla, and AA = Abies amabilis. Valuesfor seedlings and saplings of height 5 m are given in Table 3.3 (p.23).282.0 2.01.5 1.50.0 I I I 0.0 I I(a) 4 B 12 16 24 (b) 4 8 16 32Quadrat Size (m) Quadrat Size Cm)2.0 2.0[0 [. ../(c) 4 8 16 32 (d) 4 8 18 32Quadrat Size (m) Quadrat Size Cm)Figure 3.4 Values of Morisit&s Index (Id) for canopy trees (solid symbols) and subcanopy trees(open symbols) at different quadrat sizes for the four intensively sampled study plots (from (a) to(d): Si, S4, Ni, and N4). Quadrat size is given as the side length of a square quadrat. Thedashed line at ‘d = 1.0 represents the index value for random spatial distributions; no indexvalues are significantly different from 1.0.29DiscussionStand DynamicsThe size and age distributions of the study stands are similar to those of other mixedtemperate forests (e.g., Oliver and Stephens 1977, Rebertus et al. 1991,) and Thuja-dominatedforests (Alabeck 1984, Turner and Franz 1985, Parker 1986, Franklin and DeBell 1988, Edmondset al. 1993, Keenan 1993, Inselberg 1993). In five of the stands (Si, S4, N2, N3, and N4), thediameter structures formed an inverse-J, indicative of continuous ingress and recruitment withinthe stand. There was no evidence of recent stand level disturbances; small-scale disturbancesincluded canopy tree mortality, windthrow and water courses beneath the canopy that providedabiotic heterogeneity within the ecosystem. These five stands were considered true old-growthas defined by Oliver (1981) and Oliver and Larson (1990).The effects of past stand-level disturbances were manifested in the species compositionsand structures of stands S2, S3, and Ni. A distinct charcoal layer in the soil profile of plots S2and S3 indicated that they originated after fire. The dominant Pseudotsuga were likely part ofthe stand initiating cohort and were greater than 525 years old, based on ring counts of incrementcores. The presence of charcoal in the root mounds of windthrown Thuja in stand S2 suggestedthat the co-dominant Thuja may also have been established at that time. Such establishment isconsistent with Thuja’s affmity for burned substrates (Klinka and Feller 1993), the paradigm offire-origin Thuja-dominated stands (Pojar et al. 1991), and speculation on the role of fire as afacilitator of Thuja ingress (Daniels et a!. 1994). The diameter structure of stand S2 reflected theinfluence of this disturbance. The canopy was composed of trees of a broad range of diameters,perhaps indicative of prolonged ingress and differential growth of individuals after the fire(Franklin and Hemstrom 1981, Beese and Sandford 1992, Kneeshaw 1993). The lower canopywas dominated by small diameter Tsuga and included some Thuja, Abies, and Taxus. Thissuggests that the canopy was closed in the past and had restricted ingress and recruitment.Recent windthrows may have opened the canopy and initiated regeneration of tolerant species tothe subcanopy strata, a dynamic process documented by Stewart (l986a, 1989) in fire disturbed30Pseudotsuga-Tsuga-Abies stands. Given the evidence of recent small-scale dynamic processesand the dominance by Pseudotsuga, stand S2 was considered a transition old-growth stand(Oliver 1981, Oliver and Larson 1990).Stand S3 likely originated at the same time as S2, but appeared to be at a different stageof structural development. It might have been affected by a more recent disturbance as the standadjacent to it originated after a fire about 120 to 140 years ago according to the age of canopydominant trees (GVWD 1991). Whether this stand was in a transition old-growth stage ormultiple-cohort stage of development (Oliver 1981, Oliver and Larson 1991) could not bedistinguished as the age structure was not determined for this stand.The general deficiency of subcanopy regeneration in stand Nl was attributed to firedisturbance. Charcoal in the soil profile and the tree ring patterns of scarred live trees and snagssuggested that a fire occurred Ca. 1810. Many trees survived, which is indicative of moderate tolow intensity fires (Morrison and Swanson 1990). The lack of trees in the youngest and smallestclasses of the age and size distributions can be attributed to mortality of the subcanopy in the fire(as noted by Bonnicksen and Stone 1981, 1982 and Stewart 1989), and to reduced ingress andrecruitment since the disturbance (sensu stem exclusion and understory reinitiation stages, Oliver1981). The scarcity of Abies, relative to Tsuga, in the sapling and lower canopy strata of all threefire disturbed stands is consistent with Stewart’s (1989) observations that Abies rarely regeneratesimmediately after fire.Population DynamicsThe size and age structures for Tsuga and Abies were consistent with those of tolerant,self-perpetuating species (Whipple and Dix 1979). The size structures of Thuja illustrated itsdominance in the stands, but emphasized its scarcity beneath the upper canopy; however, thepoor relationship between diameter and age for Thuja indicated that its size distributionsinadequately depict its temporal population dynamics (Stewart 1986a, Veblen 1986, Taylor andHalpem 1991). The presence of Thuja in the understory of all eight stands showed that this31species was capable of ingress and establishment beneath old-growth canopies and the agedistributions indicated that recruitment to the canopy strata, although intermittent, was persistentover time.The differences in population structures, described above, have been the basis forconsidering Tsuga and Abies climax species and Thuja populations non-sustaining (see Chapter1). These interpretations rest on two assumptions. The first assumption is that a large number ofjuveniles is necessary to maintain the adult component of a population for all species. In forestcommunities, this is analogous to a prediction of future canopy composition based on the currentunderstory composition (e.g. Horn 1975, Whipple and Dix 1979); the importance of the speciesmost numerous in the understory will increase in dominance in the future (Runkle 1981, White etal. 1985, Lertzman 1992). The second assumption is that the growth and development ofseedlings and saplings are essentially the same, regardless of species and micro-environmentconditions (Runlde 1981, White et al. 1985). This means that differences in species’ abundanceare believed to reflect only their dissimilar abilities to occupy the site. For example, in ThujaTsuga forests, the abundance of Tsuga saplings and lower canopy trees, relative to Thuja, hasbeen attributed to differences in their competitive abilities. Tsuga has been consideredcompetitively superior, presumably excluding Thuja from stands (Daubenmire and Daubenmire1968). This interpretation of the population structures and stand dynamics manifests thetraditional view of successional change in favour of Tsuga in old-growth stands.An alternative to the above scenario is that the observed differences in populationstructures can indicate differences in the rates of tree establishment, growth, or mortality (Runkle1981). Establishment and growth will be addressed in Chapters 4 and 5, respectively. Thisstudy did not document mortality directly, but if one assumes that the conditions under which thepopulation developed were relatively consistent such that the age distribution emulates lossesfrom the population rather than historical variation in ingress and establishment, then patterns ofmortality are implicit in age distributions (Hett and Loucks 1968, Knowles and Grant 1983).Parker and Peet (1984) caution against this assumption and point out that in some ecosystems32variation in ingress is more important than is mortality, a warning that may apply to this studyarea (Daniels et al. 1994). In the absence of long-term data, and given that five of the eight studystands were true old-growth (lacked evidence of stand influencing disturbances), mortality wasinferred from the age structures with recognition of the limitations of this method.An age frequency distribution skewed to the left, as with Tsuga and Abies populations,can indicate high mortality (Runkle 1981), while a relatively flat age distribution, like those ofThuja, can indicate low mortality (Stewart and Rose 1990, Keenan 1993). High mortality amongTsuga and Abies trees (dbh> 10 cm) could explain the age-dependent trend revealed by thelongevity-density model; low mortality of Thuja trees is consistent with the comparabilitybetween the “required” and observed recruits of this species (Table 3.4). Spies et al. (1990) andEdmonds et a!. (1993) report that in old-growth Pseudotsuga-Tsuga forests of Washington andOregon, losses from Abies and Tsuga populations surpass those of Thuja populations, whichsupports the speculation that interspecific differences in canopy tree-mortality rates exists. Likelongevity, low mortality of Thuja can be ascribed to its resistance to insect and pathogen damage,particularly in comparison to Tsuga and Abies (Minore 1983).Runkle (1989) describes the “storage effect” (Chesson and Warner 1981, Warner andChesson 1985), which states that a population will persist when reproductive success in “good”years surpasses decline in more frequent “bad” years. Great longevity “stores” the effects ofgood years and allows survival through periods of poor reproductive success. For species of lowdensity, the frequency of good years need not be high. The relative increase of low densityspecies in good years can be great; losses in bad years, small, particularly relative to losses inhigher density species. Lertzman (1989), in modeling the dynamics of Chamaecyparis-AbiesTsuga stands in the Mountain Hemlock (MH) zone concludes that the “storage effect” facilitatesthe coexistence of Chamaecyparis nootkatensis with Abies amabilis, Tsuga heterophylla, andTsuga mertensiana. He demonstrates that three factors enable the perpetuation of long livedspecies: (1) interspecific differences in mortality, (2) low density, and (3) environmentalheterogeneity. The observed demography of Thuja, and the above longevity-density model,33support the idea that Thuja is adapted to a low recruitment - low density - low mortalitypopulation dynamic, and like Chamaecyparis, could be perpetuated under the principles of thestorage effect. Moreover, these interpretations refute the assumptions that underlie the perceptionthat Thuja populations are not self-sustained.A second aspect of population dynamics that can complement the storage effect is thatspecies which depend on environmental heterogeneity to reproduce will exhibit great variation inthe abundance of their regeneration in time and space (Fox 1977), which increases their chance ofoccasional success (Runkle 1989). The discontinuous form of Thuja age distributions suggestspulsed or sporadic ingress/recruitment episodes. This type of age structure is indicative of one ormore of the following: habitat on the margin of the species’ ecological amplitude, climaticaberration, or disturbance followed by regeneration and stand development (Whipple and Dix1979, Lorimer 1980, Oliver 1981, Oliver and Larson 1990, Hofgaard 1993). The study area waswell within the ecological range in which Thuja growth is vigorous (Krajina 1969, Minore 1983).Historic and long-term climatic variation cannot be discounted as a source of influence on Thujadynamics; however, supporting data have not been investigated. Current knowledge ofpopulation dynamics for many forest ecosystems recognizes the prevalence and influence ofdisturbance events (White 1979, Whitmore 1982). By the definition of an old-growth forest, thedynamics of the study stands were driven by small scale disturbance and individual canopy treereplacements (Oliver and Larson 1990). The environmental heterogeneity (sensu Runkle 1981)associated with such disturbances is the most likely explanation for the observed age structures ofThuja. Others have speculated on the role of disturbance in the regeneration of Thuja (Schmidt1955, Gregory 1957, Habeck 1968, 1978, Parker 1979, 1986). I hypothesize that small scaledisturbance instigates ingress and recruitment of Thuja in old-growth forests and is integral to themaintenance of this species in the study area.34Spatial DynamicsThe spatial pattern of the subcanopy and canopy trees in the study stands did not provideevidence of regeneration in groups, nor of upper and lower canopy strata interactions. However,the bimodal form of the height distributions support the interpretation that competition to gainthe upper canopy stratum was great, indicative of gap replacement. Trees regenerating aftersmall scale disturbances in canopy gaps originate in clusters, but become less aggregated withgrowth due to density dependent mortality (Harper 1977). Given the advanced age of trees inboth the lower and upper canopy, it was possible that they had established as a cohort which wasno longer evident due to competition and mortality (Turner and Franz 1985). If this were thecase, spatial evidence of small scale disturbances and gap replacement processes may be revealedthrough examining understory structures and distributions and tree-ring patterns of canopy trees.ConclusionsOf the eight study stands, three had been disturbed by fire which affected their speciescomposition, size and age structures; five stands were considered true old-growth. Age structuresprovided the greatest information with respect to population dynamics and suggested differencesin the mortality rates of canopy trees of the three study species. The demography, longevity anddensity of Thuja are consistent with the concepts of the storage effect, which implies this speciesmay be sustained in spite of low establishment and recruitment. Inherent in this interpretationare interspecific differences in the mortality, establishment, and recruitment of the study species,which contradicts the assumptions that underlie the interpretation that Thuja populations are indecline. It is speculated that heterogeneity associated with small-scale disturbances alsoinfluences Thuja population dynamics, although this was not evident in the spatial distribution ofcanopy and subcanopy trees. The following chapters investigate the regeneration dynamicswithin the study stands to assess establishment and recruitment of the study species and toaddress the role of disturbance in population and stand dynamics in order to test the hypothesis ofa low-recruitment - low density - low mortality population dynamic for Thuja.35Chapter 4. Natural RegenerationIntroductionTo understand the regeneration dynamics of the study stands, the spatial and temporalpatterns of each tree species’ regeneration (seedlings and saplings), and the conditions underwhich they establish and survive beneath the forest canopy, need to be identified. Theseobjectives require knowledge about ingress, establishment of seedlings and saplings, and theirsurvival beneath the canopy (Noble and Slatyer 1980, Runkle 1981). This study examines thenatural regeneration of Thuja, Tsuga and Abies, with a focus on Thuja.Ingress, the entry of individuals into the population, requires a source of propagules andappropriate conditions for their germination and establishment (Harper 1977). Thuja reproducesasexually and sexually (Minore 1983). Asexual, or vegetative, methods include layering orvegling development (Schmidt 1955, Habeck 1968, 1978, Parker 1979), rooting of fallenbranches and branch development of fallen trees (Schmidt 1955, Minore 1983). Seed productionis prolific but variable, with mast crops every three to five years (Graham et a!. 1988). Klinkaand Feller (unpubi.) measured viable seed production of 74,730 seeds ha-’ and 1,046,150 seedsha -1 in mast years for Thuja in Thuja-dominated stands in the study area. Studies havedemonstrated that germination and success of seedlings aged 2 to 3 years was greatest undermedium to low light and on disturbed substrates (Fisher 1935, Soos and Walters 1963, Minore1972, Klinka and Feller 1993). In the CWHvm subzone, Thuja is considered very shade tolerant,but tolerant of exposure to full sunlight (Carter and Klinka 1992, Klinka et a!. 1990). Seedlingscan survive in the understory, are capable of release (Schmidt 1955, Graham 1982) and,hypothetically, of recruitment into the canopy.Despite available seed, ability to germinate on a variety of substrates, multiple methodsof reproduction, and tolerance of the subcanopy light environment, advanced regeneration ofThuja is generally scarce or may be absent (Table 3.1; Schmidt 1955, Gregory 1957, Keenan1993). This study is an in-situ examination of the natural regeneration of Thuja, in comparison36to Tsuga and Abies, in eight old-growth stands. It aims to describe and contrast the conditionsunder which regeneration of each species establishes and survives beneath the forest canopy. Theobjectives are: (1) to quantify the regeneration of Thuja from seed or from vegetative means, (2)to determine whether the abundance of Thuja regeneration is independent of local climate, (3) todetermine whether each of the three study species regenerates at random locations within thestudy stands, and (4) to determine whether each study species indiscriminately reproduces ondifferent substrates.MethodsData AnalysisThe abundance of regeneration was determined for each quadrat, subplot, and plot.Thuja regeneration was differentiated according to its origin (i.e., seed versus vegetativereproduction). The survival of vegetative reproduction could not be assessed as the origin ofsuch individuals could not be determined. Only data from regeneration derived from seed wasused in the analyses of age-size relationships, spatial patterns, and the effects of micrositecharacteristics.A regression model of age on diameter was developed for small Thuja (dbh lO cm)based on data from plots Si, S4, Ni and N4. Both age and diameter were logarithmicallytransformed (ln[age] = b0 + b1 x ln[diameterJ) to meet the homogeneity of variances assumptionof regression analysis (Zar 1984). The standard error of estimate of the model was backtransformed to represent the variation in the original units (years).The spatial distributions of Thuja regeneration on four plots (Si, S4, Ni and N4) wereassessed with Morisita’s (1959) index of dispersion (1d) and nested quadrats (Stewart l986a).Indices were calculated for all Thuja regeneration and separately for Thuja regeneration ondifferent substrates. For stand Si (area analyzed = 48 x 48 m), the quadrat sizes were 16, 64,144, 256, and 576 m2; for stands S4, NI and N4 (area analyzed = 64 x 64 m), the quadrat sizeswere 16, 64, 256, 1024 m2. Morisita’s index of dispersion was calculated only when sample37sizes were 20 and was evaluated with the Chi-squared (X2) test statistic (Greig-Smith 1983).The tendency for Thuja regeneration to occur in association with canopy trees in different heightstrata was investigated with 2 x 2 contingency tables and the Chi-squared test statistic with Yate’scorrection (X2yates) for sample sizes <500. Statistical tests were conducted only when thenumber of observations in each cell of a contingency table was 5 (Mueller-Dombois andEllenberg 1974, Zar 1984).Morisita’s index of dispersion was calculated for Tsuga and Abies on all plots at twospatial scales: seedlings sampled in 4 m2 quadrats and saplings sampled in 100 m2 subplots.For each plot, the frequency and relative frequency (proportional frequency) of Thuja,Tsuga, and Abies regeneration rooted in each substrate were determined. As the substrates weredisproportionately available in each stand (Table 2.1), the frequency distributions of regenerationamong substrates were not indicative of the degree of use of each substrate. To makecomparisons of the utilization of substrates within stands, the density (frequency / cover [%] ofsubstrate) and relative density (proportional density) of regeneration on each substrate werecalculated. Where the distribution of regeneration is dependent only on substrate availability,regeneration density and relative density will be approximately equal on each substrate. Wherefactors other than substrate availability influence distribution of regeneration, densities will vary,increasing where substrate is conducive to germination and establishment, and decreasing whereit is less suitable. Within each stand, the relative density of regeneration of each substrate wasranked from 1 to 3, representing high to low substrate use, respectively. Where relative densitieswere the same or within 0.01 of each other they were allocated an average rank (e.g., 1.5 for twospecies of the same relative density and of greater density than the third species). Where therewere fewer than 10 observations for a stand, ranks were assigned, but were not used in thesubstrate use summary compiled for all stands. Note that in this context, the terms “use” and“utilization” refer to the successful germination and establishment of regeneration on a givensubstrate.38Variation in the densities among stands is due, in part, to differences in the absoluteabundance of regeneration and in substrate covers. Representing density as a proportion (relativedensity) standardizes these differences and provides a measure of substrate use comparablewithin substrates among stands. Substrate use by Thuja was compared among stands andcontrasted with between-stand differences in substrate availability. Analysis of variance(ANOVA, Zar 1984) was conducted to compare the covers of forest floor, decaying wood, andmineral soil among the study stands. Those means determined to be significantly different (cc0.05) were compared by the Tukey method (Zar 1984). Comparison of the results from thesetests was made to determine whether substrate use varied with substrate availability, and toconsider whether suitable substrate was a limiting resource to regeneration of Thuja.In the above analyses of regeneration distribution and substrate use, it was assumed thatseed rain of each species was uniform throughout the stand and onto all substrata, although it isrecognized that differences in substrate elevation and texture can influence seed distribution(Lusk and Ogden 1992).The success of each species on each substrate was quantified. Large size was consideredindicative of success, thus it was assumed that the greater the proportion of tall individuals on asubstrate, the more conducive that substrate was to the species’ growth. As large individualsgrow from smaller ones, direct comparison of their abundance could be made (Parker 1986).Saplings were compared to seedlings for Tsuga and Abies. Due to insufficient numbers of Thujasaplings, individuals of this species were differentiated by height. On each site, the tallest onethird of observed individuals were designated large; the other individuals were considered small.Expected values were calculated for saplings (or large individuals) of each species within eachstand:Seedlings.Expected saplings,=* Observed Saplings,Seedlings1 1=139where: i = 1, 2, ...j substrates. Success of regeneration was assessed by a paired test ofproportions (Zar 1984, Khazanie 1986) in which the proportions of observed and expectedsaplings were compared. For each substrate on each site, the test statistic was calculated asfollows:xym n- +!)]m nwhere x = Expected saplings on a given substrate, y = Observed saplings on a given substrate,and m n = total saplings in the stand, and = (x + y) / (m + n). The test statistic wascompared to a critical Z-value of the standard normal distribution at a = 0.05. These tests ofsuccess assumed congruency in the availability of each substrate at the time of saplingestablishment relative to substrate availability during recent seedling establishment and to thecurrent cover of each substrate.The differences in relative densities of Thuja regeneration of vigour3 were comparedamong substrates within stands. These analyses were included to accommodate historical biasassociated with the assessment of seedling survivorship by size, and assumed that the vigour ofan individual reflected its growth under current microsite conditions in which substrate was thepredominant influence.Size structures of regeneration (dbh 10 cm) were analyzed by 0.5 cm diameter-classdistributions (from >0 to 10 cm). The data were standardized by converting the number ofobservations in each diameter class to the equivalent number that would have occurred if thestarting density of the >0 to 0.5 cm class had been 1000. Standardization of the data allowedcomparisons of the regeneration dynamics among species and stands (Hett and Loucks 1968,Begon and Mortimer 1986). Diameter-classes were used rather than age-classes for two reasons:(1) size was accurately assessed for all individuals, and (2) age was considered a poor indicator40of an individual’s success and chance of recruitment to successive size classes due to the shadetolerance of Thuja, Tsuga, and Abies (Krajina 1969, Klinka et a!. 1990, Carter and Klinka 1992).Competition can govern the success of regeneration (Hannon 1987, Rebertus and Veblen1993). Adams and Mahoney (1991) showed that competition has significant influence onplanted Thuja seedlings and others have documented the suppression by understory vegetation oftree regeneration in canopy gaps (Alabeck 1984, Stewart 1986b, Lertzman and Krebs 1991,Inselberg 1993, Keenan 1993). Attempts were made to quantify the potential for competitiveinteractions; however, this study was not designed to address this research question and the datawere too general to identify interactions that might have existed. Manipulative experimentsspecifically designed to assess the role of competition in regeneration dynamics are needed.ResultsSpatial DistributionsIn each stand, over 89% of Thuja regenerated from seed. This concurs with Keenan’s(1993) observations on Vancouver Island, but contrasts with the findings of Schmidt (1955) incoastal stands of British Columbia and Parker (1979, 1986) in Idaho. Figure 4.1 gives Morisita’sindex of dispersion profiles derived for the Thuja seedlings and saplings in four stands. Thujaregeneration was non-randomly distributed in all stands, at all spatial scales, except for 32 x 32 mquadrats in stand N4. Aggregation was greatest within 4 x 4 m quadrats and was inverselyrelated to quadrat size. In relation to substrate, aggregation was greatest on mineral soil, followedby decaying wood, then forest floor. Regeneration rooted in the forest floor in stand Si wererandomly distributed, but only in 16 x 16 m and 32 x 32 m quadrats.Where fewer than 20 individuals rooted into a given substrate, statistical analyses of theirspatial distribution were not conducted, but visual assessments of the data were made. In standsNi and N4, regeneration that moted into the forest floor appeared evenly distributed. In standsN4 and S4, the majority of individuals that rooted in exposed mineral soil (11 of 13 and 6 of 8individuals, respectively) grew together; in each stand they occupied one 4 x 4 m quadrat.41Quadrat Size (m) Quadrat Size Cm)(a) (b)20 35Quadrat Size Cm) Quadrat Size Cm)(c) (d)Figure 4.1. Values of Morisita’s index of dispersion (1d) for all Thuja pilcata regeneration (.)and Thuja regeneration differentiated by substrate (forest floor A, decaying wood = D, mineralsoil = 0) at different quadrat sizes for the four intensively sampled study stands (from (a) to (d):Si, S4, Ni, and N4). Quadrat size is given as the side length of a square quadrat. The dashedline at = 1.0 represents the index value for random spatial distributions.42Comparisons of the distributions of Thuja regeneration to the distributions of thesubcanopy and canopy trees revealed only three statistically significant associations. In stand 54,a positive association existed with trees of height-class 4 at 4 x 4 m (X2 = 3.99, p = 0.046) and anegative association existed with height-class I and 2 trees at 8 x 8 m (X2 = 4.24, p = 0.039). Anegative association with height-class 3 and 4 trees at 8 x 8 m was detected in stand Ni (X2 =6.26, p = 0.012).Tsuga regeneration was aggregated on all plots with seedlings more clumped thansaplings. Where samples for Abies were sufficient, analysis showed regeneration to beaggregated, except for saplings in stand S 1 (Table 4.1).Substrate AnalysesIn each stand, the three study species grew on all available substrates, but the majority ofThuja and Tsuga rooted in decaying wood and Abies most commonly rooted in undisturbedforest floor. The relative frequency of regeneration on mineral soil and decaying wood wasgenerally high in relation to their cover, whereas the frequency of regeneration on undisturbedforest floor was consistently lower than its cover (Appendix B).The relative density of regeneration on each substrate revealed variation in the degree ofuse of substrates among the study species (Table 4.2, Appendix B). Thuja effectively exploitedmineral soil, where it was available, and decaying wood. Tsuga seedlings occurred mostfrequently on decaying wood, and Abies seedlings, on decaying wood and undisturbed forestfloor. The saplings of the latter two species predominantly grew on decaying wood, but therelative density of saplings on this substrate was less than that of seedlings, and there weresubstantial increases in the relative density on mineral soil of saplings of both species.43Table 4.1. Morisita’s index of dispersion (Id) for Tsuga heterophylla and Abies amabilisseedlings and saplings in the eight study stands.Stand Tsuga heterophylla Abies amabilisSeedlings Saplings Seedlings Saplings(g = 25 @ 4m2) (g = 16 @ 100m2) (g = 25 @ 4m2) (g = 16 @ 100 m2)N Id p N Id p N Id p N Id pSi 276 2.87 <0.001 172 1.23 <0.001 3 NA1 NA 24 1.39 0.462S2 175 4.51 <0.001 227 1.34 <0.001 3 NA NA 2 NA NAS3 340 1.95 <0.001 161 1.41 <0.001 1 NA NA 5 NA NAS4 221 3.62 <0.001 92 1.66 <0.001 108 1.45 <0.001 347 1.25 <0.001Ni 124 1.79 <0.001 198 1.60 <0.001 146 3.15 <0.001 15 NA NAN2 216 4.67 <0.001 136 1.71 <0.001 43 1.52 0.005 120 3.91 <0.001N3 176 3.67 <0.001 273 1.31 <0.001 124 2.21 <0.001 138 1.16 <0.001N4 123 4.24 <0.001 125 1.73 <0.001 93 1.68 <0.001 260 1.41 <0.0011NA = insufficient sample size (n <20)The variation in substrate use between seedlings and saplings within species, isconsistent with the evaluation of substrate-specific regeneration success. Success of Thuja andTsuga was equal to or greater than expected on the forest floor and equal to or less than expectedon decaying wood. Success of Abies on both of these substrates was variable among stands.Success of all three study species was generally higher than expected on mineral soil (Table 4.3,Appendix C). Comparison of the distribution of vigorous Thuja (vigour 3) on differentsubstrates confirmed its successful exploitation of exposed mineral soil (Table 4.2, Appendix B).44Table 4.2. Summary of substrate utilization by seedlings and saplings. Substrate use (relativedensity) within stands was ranked from 1 (high) to 3 (low) and summarized to show overall useby each species. Only rankings from stands that included lO seedlings or saplings are included;n = the number of stands that met this criterion. FF = forest floor, MS = mineral soil, DW =decaying wood. Data from individual stands are shown in Appendix B.Substrate Frequency Distribution of Use1 2 3 Overalln = 8 Thuja plicataFF 0 0 8 3DW 3.5 4.5 0 2MS 4.5 3.5 0 1n = 3 Thuja plicata (vigour 3)FF 0 0 3 3DW 1 2 0 2MS 2 1 0 1n = 8 Tsuga heterophylla SeedlingsFF 0 3.5 4.5 3DW 7 1 0 1MS 1 3.5 3.5 2n 8 Tsuga heterophylla SaplingsFF 0 0.5 7.5 3DW 6 2 0 1MS 2 5.5 0.5 2n = 5 Abies amabilis SeedlingsFF 2.3 2.3 0.3 1.5DW 2.3 2.3 0.3 1.5MS 1.3 0.3 4.3 3n = 6 Abies amabilis SaplingsFF 1 2 3 3DW 3 3 0 1MS 2 1 3 245Table 4.3. Success of regeneration on different substrates. FF = forest floor, DW = decayingwood, MS = mineral soil. Data from individual stands are shown in Appendix C.Substrate Seedlings1 Saplings1 Saplings Success p value(Observed) (Expected)Thuja plicataFF 226 147 113.12 High 0.011(0.12)2 (0.16) (0.12)DW 1448 647 724.79 Low <0.001(0.79) (0.70) (0.79)MS 166 127 83.09 High <0.001(0.09) (0.14) (0.09)Tsuga heterophyllaFF 123 152 108.15 High 0.002(0.08) (0.12) (0.08)DW 1336 1062 1174.00 Low <0.001(0.90) (0.82) (0.90)MS 24 89 20.85 High <0.001(0.02) (0.07) (0.02)Abies amabilisFF 297 432 516.92 Low <0.001(0.61) (0.51) (0.61)DW 187 372 325.82 High 0.012(0.38) (0.44) (0.38)MS 6 49 10.24 High <0.001(0.01) (0.06) (0.01)1For Thuja plicata, “seedlings” = shortest two thirds of individuals with dbh 10 cm and“saplings” = tallest one third of individuals with dbh 10 cm. For further explanation see p. 39.2Proportions for each substrate for each species are indicated in parentheses.46For each substrate, cover was significantly different among stands (F = 8.83, 5.23 and4.18; p 0.001, 0.001, and 0.004 for forest floor, decaying wood and mineral soil, respectively).The relative density of Thuja on each substrate varied among study stands (Appendix B), butshowed no consistent relationship to substrate availability. For example, where the cover ofdecaying wood was high, frequency of regeneration rooted in decaying wood was high, butrelative density of such regeneration was variable. In contrast, where mineral soil cover washigh, both the frequency and relative density of regeneration was generally low, but variable.These variations in substrate use suggest that factors other than the amount of substrate coverhave significant influence on the quality of a microsite for Thuja regeneration.Understory StructuresThe diameter distributions of seedlings and saplings are given in Figure 4.2. Thesecurves indicated that the proportion of Thuja and Tsuga that recruited from one diameter-class tothe next was relatively consistent within and among stands. Sufficient samples of Abies wereavailable only in stand S4 and the north-aspect stands. In four of five of these stands, a higherproportion ofAbies recruited to successive diameter-classes than was observed for Thuja andTsuga.For each species, less than 1% of seedlings in the >0 to 0.5 cm class grew past the 4 cmdiameter-class, except Abies in stands S4, N2, and N4, where between 1 and 10% of Abiesrecruited to diameter classes >4 cm. For Tsuga and Abies, these percentages representednumerous saplings, potential recruits to the canopy. The distributions of Thuja regeneration wereless continuous. In five stands, saplings with diameters greater than 5 to 8 cm, depending on thestand, were sporadic, and three stands included no saplings with diameters >2.5 cm (Figure 4.2).Age counts for 215 Thuja seedlings and saplings from plots Sl, S4, Nl and N4 rangedfrom 5 to 89 years. The ages of 663 individuals were estimated by regression of age on diameter(j2 = 0.51, SEE = 8 years, p 0.001). The age structures constructed from the actual andestimated ages indicated relatively consistent ingress in each stand (Figure 4.3).4711llI[pIr)I(a)01 2 3 4 5 6 7 8 9 1030aI 1riri0 ri—r-rrrrrii’iiriiii(c) 1 2 3 4 5 6 7 8 9 100.2aI :1 2 3 4 5 6 7 8 9 10(e)I1 2 3 4 5 6 7 8 9 10(g) Upper Umit of Diameter Ciass (cm)(b)01 2 3 4 5 6 7 8 9 101,j.J,i II2,I.i1LII1J J1Ifr1FIL II1111A,Ifl,Iii2 3 4 5 6 7 8 9 10ddJULfl12 3 4 5 6 10(h) Upper Umit of Diameter Class (cm)Figure 4.2. Diameter distributions of regeneration (dbh 10 cm) in the eight study stands (from(a) to (h): Si, S2, S3, S4, Ni, N2, N3, and N4). Seedlings and saplings are grouped in 0.5 cmclasses to 10 cm. The data have been standardized and log10 transformed to visually depictclasses that included few observations (i.e., diameter >4 cm). In the legend, TP = Thuja plicata,TH = Tsuga heterophylla and AA Abies amabilis.31 2 3 4 5 6 7 8 9 102(d)(t)db IflE JUh BBu1ii0321•0I I [ I I TPD AAJ TH4880EC)0V.0Va)a)C)qI-0Ez250200150100500706050403020100580(a) (b)10152025303540455 )E00V-c.0Va)a)00.I—0a).0EDz706050403020100(c)5101520253035404550Upper Limit of Age Class (years)(d)5 101520253035404550Upper Limit of Age Class (years)Figure 4.3. Age structures of Thuja plicata regeneration (dbh 1O cm) estimated for the fourintensively sampled study stands (from (a) to (d): Si, S4, Ni, and N4).49DiscussionThuja plicata RegenerationThe presence of Thuja regeneration in the understories of all eight study stands indicatedthat this species was capable of ingress and establishment beneath old-growth canopies. Thesuccessional status and species composition of the study stands may not be as relevant to themaintenance of the Thuja populations as others have suggested (Daubenmire 1966, Daubenmireand Daubenmire 1968, Krajina 1969). The abundance of regeneration varied among plots, butwas not consistently related to local climate as determined by aspect.Thuja regeneration was aggregated, with maximum clusters at small scales. This spatialdistribution emphasizes the importance of discrete “safe-sites”, microsites conducive to thegermination and establishment of this species (Harper 1977). Conceptually, a microsite is theintegration of substrate, moisture, temperature, nutrients, light and growing space available to anindividual (Harper 1977, Klinka eta!. 1990, Oliver and Larson 1990). In this study, micrositeswere differentiated by substrate and illustrated the increasing importance of undisturbed forestfloor, decaying wood and exposed mineral soil for Thuja regeneration, as has been observed byothers (Schmidt 1955, Parker 1979, 1986, Turner and Franz 1985, 1986, Moeur 1992, Keenan1993).Associated with each substrate are a number of characteristics that might contribute tothe value of a microsite for Thuja regeneration. For example, windthrow mounds and coarsewoody debris provide micro-environments that are elevated above the forest floor which reducescompetition with established vegetation for light, growing space, nutrients, and moisture (Christyand Mack 1984, Parker 1986, Stewart l986b, Harmon and Franklin 1989, Oliver and Larson1990). When initially exposed, these substrates lack competing vegetation and can be readilyexploited by the light, abundant Thuja seeds. Resource availability to new colonizers is high,which gives them a competitive advantage over those that establish later (Harper 1977, Peet andChristensen 1987, Runkle 1989). The long-term success of regeneration on exposed mineral soil50is attributed to the texture and water-balance of this substrate, whereas survival on decayingwood is likely limited by its low nutrient status (Harmon 1987).The fact that decaying wood and mineral soil, the most important substrates for Thujaregeneration, are derived by disturbance attested to the significance of disturbance in Thujapopulation dynamics in the study stands. Differences in the use of mineral soil and decayingwood substrates were observed among stands, which may relate to inherent variation in thequality of these substrates. Both substrates are ephemeral by nature of their origin. Theirlocation and abundance within a stand vary in time and space and their properties and value as asafe-site for potential colonizers change with time (Christy and Mack 1984, Harmon 1987, Oliverand Larson 1990, Hofgaard 1993).The disturbances that affect the forest floor and provide suitable substrates for seedlingestablishment also affect the continuity of the forest canopy (Fox 1977, Stewart l986c, Hofgaard1993, Rebertus and Veblen 1993). Beneath the canopy, light is a limiting resource (Chazdon andPearcy 1991). Removal of part of the canopy creates a gap, which will increase light availabilityto microsites within its sphere of influence, depending on gap size and shape and the structure ofthe surrounding canopy (Chazdon and Pearcy 1991). A direct relationship between canopy coverand the distribution of tree regeneration has been illustrated in many forests where aggregationsof seedlings, saplings and lower canopy trees correspond to the location of canopy gaps, snags,or broken-top trees (e.g., Oliver and Stephens 1977, Veblen eta!. 1979, Stewart 1986a, 1986c,Leemans 1991, Donoso eta!. 1993). Similarly, Gregory (1957), Habeck (1968, 1978), Parker(1979, 1986) and Moeur (1992) speculate that successful regeneration of Thuja relates to lightavailability and, indirectly, to canopy structure.In this study, regeneration of Thuja in canopy gaps was evident only in stand S4, wherethe distribution of regeneration was negatively associated with that of height-class 2 trees at aspatial scale congruent with single-tree gaps (Lertzman 1989, Spies et al. 1990). The positiveassociation between Thuja regeneration and height-class 4 trees within 16 m2 plots might reflect51ingress and recruitment within small canopy gaps. Alternatively, the association may be anartefact of ingress and differential growth of individuals on localized substrates.On plot Ni, Thuja regeneration was negatively correlated to lower canopy trees (classes3 and 4) at 64 m2. Had the lower canopy trees established together, possibly in a gap, the presentmicro-environment beneath them could be one of high shade and intense competition, asdescribed by Rebertus and Veblen (1993) and Donoso et al. (1993) in other gap-regeneratedstands. Shade or competition in association with lower canopy trees could restrict the chance ofingress and establishment of Thuja and explain the negative correlation.The spatial distribution of Thuja regeneration on the study plots did not provide clearevidence of understory-canopy interaction, nor did regeneration appear influenced by gapdynamics. Parker (1986) and Turner and Franz (1985) drew the same conclusions for Thujaregeneration in old-growth stands in Idaho; however, their studies, like this one, used canopy treedensity as a surrogate for canopy cover. This method assumes that the crown width and cover ofall trees are approximately equal and that overlap of canopies among strata does not affect thelight environment in the understories; neither assumption may be accurate in the study stands(Stewart 1986b, Kuiper 1988, Lieberman et a!. 1989).Interspecic Comparison ofRegeneration NichesThe structural complexity of old-growth forests suggests that opportunities for treeregeneration within them are variable in time and space (Fox 1977). Micro-environmentheterogeneity could facilitate different responses and adaptations among the study species whichcould be manifested as niche partitioning (Forcier 1975, Grubb 1977). Based on thediscrepancies in population structures and abundance of regeneration of Tsuga and Abies relativeto Thuja, it was hypothesized that differences could exist among the regeneration niches of thesespecies (Grubb 1977). For example, the abundance of Tsuga in all understories and of Abies infive stands suggests that the conditions conducive to their regeneration were available throughoutthe space-time continuum in the forests of the study area. Conversely, the conditions for Thuja52establishment appear limited in thne and space (Lusk and Ogden 1992), as shown by the relativelack of seedlings and saplings of this species.The overlap in regeneration niches of the study species refuted this hypothesis. LikeThuja, Tsuga regenerated on disturbance-origin substrates and was prolific on decaying woodand successfully exploited mineral soils. Abies overlapped with Thuja and Tsuga by its growthon decaying wood substrates, but was distinguished from the other two species because itfrequently rooted in the forest floor. These regeneration patterns for Tsuga and Abies are similarto those observed in other stand dynamics studies of these species (Minore 1972, Christy andMack 1984, Stewart l986b, 1989, Harmon 1987, Keenan 1993).During seedling establishment, the effects of mortality and suppression of Thuja andTsuga appeared comparable, and were higher than these effects on Abies. Despite suppressionand high seedling mortality, abundant Tsuga and Abies recruited to the sapling and canopy strata,strata in which their mortality was continuous. In contrast, the distributions of Thuja advancedregeneration became irregular after a few centimeters in diameter and formed a flat,discontinuous curve, similar to the diameter distributions of canopy Thuja trees (Figure 3.2).Recruitment success of advanced regeneration appeared high, which was consistent with theapparent low mortality of canopy trees.These interspecific differences in recruitment and mortality might relate to differentresponses among the species to light and canopy gaps. Establishment and recruitment of Tsugaand Abies have been shown by Stewart (1986a) to be gap dependent processes in coastal standsin Oregon. If this is true in this study area, the abundance of regeneration of these two species inthe understory would provide a “seedling bank” able to respond rapidly and exploit resources thatbecome available in a gap (Leemans 1991, Keenan 1993). High competition and mortalitybetween successive strata would be expected. If regeneration of Thuja does not depend ondisturbance to gain the canopy, its risk of mortality would be relatively low (Christensen 1977)and it would not require a “seedling bank” to ensure its recruitment.53Although disturbance to the forest floor facilitated the establishment of Thuja seedlings;it remained inconclusive whether successful regeneration of Thuja required variation in theunderstory light environment, as in a canopy gap. Examination of the growth of Thuja seedlingsand saplings beneath canopy gaps of various sizes in contrast to the growth of those beneath aclosed canopy (Runkle 1992), in combination with manipulation of other microsite properties,could distinguish the influences of light aid microsite on Thuja regeneration success. Analternative approach would be to examine the growth patterns of Thuja in various strata of thecanopy and in relation to known canopy gaps, to assess this species’ response to variation inavailable light. This latter approach was followed and presented in Chapter 5.ConclusionsDisturbance facilitated the successful establishment of Thuja as was evident from theaggregated spatial distributions of regeneration, its affinity for decaying wood and exposedmineral soil, and the high survivorship on the latter substrate. Substrate availability did notappear to limit regeneration of this species, as ingress and seedling establishment werecontinuous in time. Significant overlap in the regeneration niches of Thuja, Tsuga, and Abieswere identified, in part, by similarities in their substrate use, with no evidence of spatial-temporallimitations to the regeneration of the latter two species.Differences in the regeneration dynamics of Thuja, Tsuga, and Abies in the study standswere related to post-establishment recruitment success and mortality. The low sapling andcanopy tree mortality and high recruitment success of Thuja advanced regeneration likelyaccount for its population age and size structures and the scarcity of its seedlings and saplingsrelative to those of Tsuga and Abies. The regeneration dynamics of Thuja, described by thisstudy, are congruent with the low recruitment - low density - low mortality population dynamicdescribed by the storage effect, hypothesized to describe Thuja population dynamics.54Chapter 5. Canopy Tree Growth and Gap DynamicsIntroductionCanopy gaps and the structural diversity of the canopy are sources of heterogeneitywithin understory light environments of old-growth stands. Beneath continuous canopy cover,light is a limiting resource to tree growth (Chazdon and Pearcy 1991); the light environment isone of reduced intensity, diffuse background radiation, and ephemeral sunflecks. Within canopygaps, irradiance is higher and more conducive to tree growth. In addition to this spatialheterogeneity, variation in the light environment occurs temporally with the formation andclosure of gaps of various sizes.In the study area, Thuja, Tsuga, and Abies are shade-tolerant and survive in low lightconditions (Krajina 1969); however, none of the three species is shade-requiring. When morelight becomes available, as in a canopy gap, all are capable of more rapid growth (Carter andKlinka 1992). Stewart (1986a, 1986b, 1989) demonstrated the importance of canopy gaps to theregeneration of Tsuga and Abies. The population structures and regeneration patterns of Thujadescribed in Chapters 3 and 4 can be interpreted to indicate that disturbance and gap dynamicsmay also facilitate regeneration of this species. Given the heterogeneity of the light environmentin old-growth forests, in combination with the potential responses of shade-tolerant trees tocanopy gaps (Canham 1988, 1989), it is hypothesized that differences in the growth response ofThuja, Tsuga, and Abies to the old-growth forest light environment are likely to exist. Growthpatterns of individual trees can be analyzed with dendrochronological techniques andcomparisons within and among species can be made because the growth response of shade-tolerant trees to changes in their light environment are reflected in their annual growthincrements and tree ring-width patterns (Lorimer 1985, Fritts and Swetnam 1986).This portion of the study examines the ring-width series of canopy trees (dbh >10cm)within four old-growth stands to evaluate the response of Thuja, Tsuga, and Abies to past canopygap events. The objectives are: (1) to compare and contrast the growth of canopy trees, and (2)55to determine whether the study species depend on gap events to gain entry to the upper canopy.Inferences are made based on concurrent changes in the ring-width series of neighbouring trees toprovide evidence of past gap events (Lorimer 1985, Payette et a!. 1990). Comparisons are madeof the frequency of release of trees of each species in relation to gap events and tree position inthe canopy.MethodsData AnalysisOf those trees sampled for age, all Thuja and four neighbours of each were selected forgrowth increment analysis. Attempts were made to sample one neighbour from each of fourcardinal directions about each Thuja. Where stem density in the vicinity of Thuja was great,additional trees were selected. The large size of and incidence of decay in canopy trees,particularly Thuja, often prevented the assessment of their total age and initial growth rates, inwhich case regression models (Table 3.2) were used to estimate these parameters.Annual ring widths were measured to the nearest 0.01 mm with a binocular microscopeand stage micrometer. The series derived from Thuja and Abies were crossdated (Schweingruber1989) to standard chronologies of these species in the study area (Dobry and Klinka 1993). Astandard chronology to which samples of Tsuga could be related had not been established. Thedata were transformed with a low pass filter (13-year weighted moving average) and a 5-yearmoving average (Fritts 1976) and plotted against time to identify periods of abrupt increase ingrowth. An increase of >50 % over the previous 10 years’ average annual increment that wassustained for 10 years was considered a release (after Lorimer and Frelich 1989). The year inwhich the release began was the date of release.It was assumed that the spatial influence of a canopy gap could be delineated by thelocation of trees with similar dates of establishment and release (Lorimer 1985, Payette et a!.1990). Thus, evaluating the ring-width series of individual trees provided an inductive methodfor analyzing the processes that affected stand structure and identifying canopy gaps (Table 5.1).56Table 5.1. Decision key used to determine a tree’s growth history. Note: Not all decisions aremutually exclusive.(la) Tree grew into the canopy continuously (2a)(ib) Tree entered the canopy after periodic suppression and releases (3a)(2a) All individuals of the same species have the same patterni.e., the species is adapted to prevailing conditions (adapted to subcanopyenvironment or always establishes in a gap)(2b) Only old or large trees of the species grew continuouslyi.e., the tree is part of the initial cohort on the site(2c) Small but old individuals of the species grew continuouslyi.e., the tree is suppressed and slow growing(2d) Neighbours of the tree (regardless of species) show this patterni.e., the tree established with its neighbours in gap(3a) Releases are synchronous in all individuals throughout the standi.e., the releases were caused by a wide scale, stand influencing force(3b) Releases are synchronous in all individuals of that species in the standi.e., the release were caused by a stand influencing force with a species-specificresponse(3c) Releases are synchronous among neighbours and/or concurrent with ingressi.e., the release is in response to a gapFrom age chronologies, ten year frequency distributions of dates of tree origins (year thatcorresponded to the total age), and dates of releases, periods (modes) of frequent ingress andrelease within each stand were identified. Spatial analyses were conducted to ascertain whetherpatterns in the spatial distribution of trees within each mode (all trees that originated or releasedwithin the time period demarcating the mode) were detectable and if even-aged patches existedwithin these uneven-aged stands (Duncan and Stewart 1991). Morisita’s (1959) index of57dispersion(1d) was calculated for nested quadrats (Stewart l986a) that measured 25 m2, 49 m,100m2, and 196 m2 to determine the scale(s) at which canopy gaps could have occurred. It wascalculated only for modes that included 20 trees and was evaluated with the Chi-squared teststatistic (Greig-Smith 1983). The dates of establishment and release of trees that formedsignificant aggregates were compared to determine whether their observed growth could reflect aresponse to a canopy gap (i.e., two or more spatially associated trees that established and/orreleased within ten years of each other).Mean annual diameter increments (dbh outside bark [cml I age [years]) were determinedfor all canopy trees sampled for age. ANOVA was used to compare the growth rates of eachspecies among height classes. Natural logarithmic transformations were applied to the Tsuga andAbies data in order to meet the homogeneity of variances assumption of this method (Zar 1984).The reported means and standard deviations of these two species were back transformed after thestatistical analyses were complete. Those means determined to be significantly different werecompared by the Tukey-Kramer method for unequal sample sizes (Ott 1984, Zar 1984). For allstatistical tests, significance was set at a = 0.05.ResultsCanopy Gaps and Tree ReplacementThe age chronologies for the four study stands revealed that both ingress and release hadoccurred over the past four centuries (Figure 5.1). The oldest trees in stands Si, S4 and N4 wereuneven-aged, and although modal classes that included establishment and release of trees wereidentified, none dominated the shape of the chronologies. These observations confirmed that allthree stands were true old-growth stands, as suggested by their diameter and age structures(Figures 3.1 and 3.2), and that small-scale dynamic processes had predominated over the periodsdepicted in this study. Stand Ni was differentiated from the other stands due to the multiple-cohort age-structure of the canopy trees that included a remnant cohort (Figure 3.2) and a modethat resulted from disturbance to the stand in the nineteenth century (Figure 5.1).5825 • . •201510I1550 1600 1650 1700 1750 1800 1850 1900 1950a,25 • •••a, ..2015•1550 1600 1650 1700 1750 1800 1850 1900 195025- :°20-15-10-Di 7?1550 1600 1650 1700 1750 1800 1850 1900 195025- • • • :20-15- S1550 1600 1650 1700 1750 1800 1850 1900 1950Year by Decades 1550 — PresentFigure 5.1. Chronology of ingress (solid bars), releases (open bars) and gap events (.) for fourstudy plots (from top to bottom: Si, S4, Ni and N4). For criteria defining a release, see p. 55.59Statistically significant aggregates of trees were identified on each plot at a range ofspatial scales (Table 5.2). Investigation of the ring-width patterns of trees within theseaggregates revealed that only some trees, although spatially significant, had established orreleased concurrently (within 10 years of each other) and represented a group that responded to acanopy gap. Conversely, comparing the ring-width patterns of Thuja and their neighboursindicated that synchronous establishment and release of trees was more common than revealed bythe spatial analyses. Used together, these two methods proved useful to identify past gap events.There were 12, 16, 4 and 18 gap events identified in stands Si, S4, Ni, and N4,respectively (Figures 5.1 to 5.5). Ring-width series from representative gaps (Figures 5.2, 5.3,and 5.5) are shown where at least two series depicting tree growth response to a gap event wereTable 5.2. Values of Morisita’s index of dispersion (Td) and probability values (p-values)calculated for modes with significant aggregated trees (i.e., not randomly distributed) in the fourstudy stands.Stand Years N Quadrat Size ‘d p-value(m2)Si 1871 - 1900 20 196 1.71 0.039S4 1751 - 1830 42 25 1.63 0.042S4 1751 - 1830 42 49 1.53 0.041S4 1921 - 1960 41 25 1.71 0.029S4 1921 - 1960 41 49 1.53 0.041Ni 1881 - 1850 28 25 2.59 0.019N4 1781 - 1850 24 196 1.72 0.019N4 1861 - 1900 25 100 2.29 0.00360E4-C.4-’-Qa-706050403020100•o 0o1oc:1.-cP0-0 : : 0:0‘f’ 0 : : 4& ctp.1........ ..0cP co 0 0c2°04I. i r°i___00 10 20 30 40 50 605432101850• Thujao Tsugaz AbiesO Spatially significantgapsO Visually identifiedgaps70Plot Width Cm)(a)(b)(c)(d)1870 1890 1910 1930 1950Year (A. amabilis)E2-c-o0)Cct2-C-I--o0)Ccc2-cD0)c1870 1890 1910 1930 19505 I I I I41850Year (I. heterophylla)54321?850 1950Year (T. heterophylla)Figure 5.2. Map of tree locations and occurrence of gap events in stand S 1(a). Filtered ring-width series depict the establishment of trees in 1878(b) and 1884 (c) and release of one tree in1893 (d) within the gap marked “1”. A third Tsuga established within the gap in 1880; its ring-width series was not measured. Gap 1 was spatially significant at 196 m2.1870 1890 1910 19306170 A I A 1 I I A:A oL° 0cx::AA40 $A .1°:i° A I30 *.Ao A • ThujaA /a\ OA0 Tsuga20“:OA Z Abieso :OA o Spatially significantA: A 0 gaps10 •• Visua1lyidentifiedA : : A: : AO: asAA d9 TA A A0 10 A______(a) 0 10 20 30 40 50 60 70Plot Width (m)(b) . 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990a:Year (A. amabilis)1L j’’’’:(c) 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990Year (A. amabilis)(d) 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990Year (T. heterophylla)(e) 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990Year CT. heterophylla)Figure 5.3. Map of tree locations and occurrence of gap events in stand S4 (a). Filtered ring-width series depict the release of trees in 1949 (b, c, and e) and 1950 (d) within the gap marked“1”. The trees depicted in (b), (c) and (d) were significantly aggregated at 100m2.62E-c00706050403020100Plot Width (m)• Thujao TsugaA AbiesO Spatially significantgapsVisually identifiedgapsFigure 5.4. Map of tree locations and occurrence of gap events in stand Ni. Filtered ring-widthseries of trees within gaps were not available as most gaps were delineated by ring counts andvisual estimates of periods of release on stumps measured in the field, rather than by ring-widthmeasurements.0 10 20 80 40 50 60 706370 1• 10 1 •i I:0 d4•:OO6050- .4O300 •Thuja0 : : OTsuga20.. z\ Abies° 1. 0 Spatially significant• 0: gapsI : • 0 Visually identified0 gaps0 A I(a) 0 10 20 30 40 50 60 70Plot Width (m)nE!1V4/ArvJ\N(b) 1775 1825 1875 1925Year (T. plicata)I I(c) j ?775 1825 18751925E Year CT. heterophylla)(d) 1850 1870 1890 1910 1930 1950Year (A. amabilis)I I II-o1.0 I -(e) 2’ 1850 1870 1890 1910 1930 1950Year (A. amabilis)Figure 5.5. Map of tree locations and occurrence of gap events in stand N4 (a). Filtered ring-width series depict the release of one tree in 1824 (b) and establishment of another in 1832 (c)within the gap marked “1”. The Abies in gap 1 established in 1833; its ring-width series was notmeasured. Gap 1 was spatially significant at 196 m2. The trees in the gap marked “2” releasedin 1883 (d) and established in 1886 (e). Gap 2 was spatially significant at 100 m2.64available. Most gaps were delineated by the concurrent establishment and release of trees ratherthan establishment alone; thus, few gaps represented even-aged cohorts. The proportion ofestablishments to releases observed within a gap varied with the length in time since the gapevent, with release of trees more common than ingress for more recent gap events. Evidence ofgaps was most complete for the past 200 to 250 years, the period for which releases were bestdocumented. The majority of releases were to Tsuga and Abies (Table 5.3); however, themajority of trees of these species were <250 years old, thus much information about gap eventsprior to Ca. 1740 was no longer available (Table 5.4).Table 5.3. Number of release periods observed among sampled canopy trees (dbh >10 cm) infour study stands. Species-specific proportions for each release class are provided in parentheses.Species Stand Number of Releases0 lto2 SumThuja plicataSi 4 1 0 5S4 5 3 0 8Ni 1 7 1 9N4 8 7 0 15SUM 18 (0.49) 18 (0.49) 1 (0.02) 37 (1.00)Tsuga heterophyllaSi 5 15 2 22S4 3 13 4 20Ni 1 19 4 24N4 2 13 3 18SUM 11 (0.13) 60 (0.71) 13 (0.16) 84 (1.00)Abies amabilisSi 1 2 0 3S4 2 10 1 13Ni 1 9 0 10N4 10 11 0 21SUM 14 (0.30) 32 (0.68) 1 (0.02) 47 (1.00)65Table 5.4. Summary of gap occurrence over time for the study stands.Stand Number of gaps per hectare in each centuryPrior to 1600 1601-1700 1701-1800 1801-1900 1901-PresentSi 0 0 0 20 16S4 0 4 8 20 32Ni 0 0 0 16 0N4 4 8 4 32 24Growth Trends ofindividual SpeciesEighteen of the 37 Thuja trees examined in this study had never released; 18 treesreleased once or twice and only one tree had released more than twice (Table 5.3). Given thelack of full ring-width series for many trees, due to their large size and the incidence of decay,these frequencies likely under-estimate the total number of releases. The partial cores that wereavailable for analysis represented the recent growth of these trees during which they may nothave responded to gaps due to their dominant canopy position where light is likely not a limitingresource (Veblen et al. 1989). In spite of this limitation, the majority of observed releases inThuja were concurrent with the establishment or release of their neighbours, except for those instand Nl. Complete growth histories, including tree origin in relation to extant neighbours, wereobtained for eight Thuja. Of these trees, four originated with and/or released in association withtheir neighbours. The other four trees originated independently of their neighbours and showedno significant releases. Comparison of the annual diameter increments of Thuja trees revealed nosignificant difference among the means for the four height classes (Figure 5.6; F = 0.92, p =0.405).In contrast to Thuja, two of 84 Tsuga trees grew to the upper canopy without detectedreleases; nine Tsuga had never released but were suppressed and restricted to the lower canopyand 73 trees located in both the upper and lower canopy had released one to five times (Table5.3). Half the releases in stands Si and N4 were concurrent with those of neighbours, more than6675% in stand S4, but less than 33% in stand Ni. All neighbours of each Tsuga were not includedin these investigations, thus the interaction of this species with its neighbours may be differentthan identified.The mean diameter increments of Tsuga varied among the height classes (Figure 5.6; F =36.88, p O.OO1). The mean increment was lowest for height class 4 trees that had never releasedand increased with successive classes to a maximum mean increment for height class 1 trees. Asthe majority of Tsuga had released at least once, it is likely that the trees in the upper canopy(height classes 1 and 2) had at one time been suppressed and had grown at rates comparable tothose of the lower canopy trees (height classes 3 and 4). Moreover, the annual increment ofTsuga, while in the upper canopy, must have been substantially greater than that of Thuja canopytrees, given that the mean diameter increment of upper canopy Tsuga was comparable to that ofThuja, but its increment in the lower canopy was significantly less than that of Thuja. Thedifference in diameter increment of Tsuga in the lower and upper canopy could explain thebimodal form of the height distributions (Figure 3.3) in which the “breaks” between the lowerand upper canopy could be the result of suppression of lower canopy trees that grew rapidly intothe upper canopy upon release.Thirty-three of 47 Abies had released at least once; 14 trees had never released, 13 ofwhich were suppressed and one had grown to the upper canopy of stand N4 (Table 5.3).Regardless of age and of frequency of release, Abies trees in stands Si, S4 and N4 occurredalmost exclusively in the lower canopy and more than 66% of these trees established and 75%released in conjunction with their neighbours. While this is a strong correlation, not allneighbours of each Abies were sampled. In contrast, Abies in stand Ni had established in the late1700’s, released in the 1850’s to 1880’s, and occupied the canopy strata.The mean annual increment of Abies varied among height classes (Figure 5.6; F = 36.38,p O.00i), as with Tsuga. In contrast to Tsuga, the differences in mean increment of Abies invarious height classes did not reflect changes in the growth rate of individuals as they advancedthrough the canopy, but were due to contrasts in the growth rate of Abies in different stands. The671Tp___________________________:2 Th____-Aa_ _ _ ____-Tp_ _ _3 Th__ _______ ____Aa I H -Tp -4 Th IAaI I I0.0 0.1 0.2 0.3 0.4 0.5Mean Annual Diameter Increment (cm)Figure 5.6. Comparison of the mean annual increments of canopy trees (Tp = Thuja plicata, Th= Tsuga heterophylla, Aa = Abies amabilis) in four height classes (1, 2, 3, and 4). Bars representstandard error. The dashed line approximates the overall mean for Thuja. Data are combinedfrom the four study stands.68low mean increment of trees in height classes 3 and 4 reflected the suppression of Abies in standsSi, S4, and N4. The high mean annual increment for Abies in the canopy reflected the growth ofthe trees in stand Ni that had released within 90 years of their establishment and were among thestand dominants.DiscussionStand HistoriesBy definition, the dynamics of old-growth forests are driven by single tree replacementsin small canopy openings (Oliver 1981, Oliver and Larson 1990). Gap dynamics is the processby which change in canopy structure provides opportunity for replacement of canopy trees (Watt1947, Whitmore 1982) through ingress and release of advance regeneration (Peet and Christensen1987). The predominance of such gap dynamic processes was evident in stands Si, S4 and N4.Senescence, suppression and disturbance can cause mortality of individuals (Spies et at.1990, Edmonds et at. 1993, Inselberg 1993) and affect canopy continuity and stand structure anddynamics. Wind, fire, pathogens and insects are prevalent forces that disturb the canopy offorests in the CWH zone (Eis 1962, Franklin and DeBell i988, Hendersen et a!. 1988, Morrisonand Swanson 1990, Edmonds et al. 1993, Stewart i986a). The impact of wind disturbance isgenerally localized and affects individuals or groups of two to three trees (Henry and Swan 1974,Morrison and Swanson 1990, Lertzman 1989, Spies et a!. 1990). The expression in populationstructures of high frequency, small magnitude fires can be indistinguishable from the effects ofwind disturbances. However, the lack of tree fire scars and charcoal in soil horizons in standsSi, S4, and N4, together with the occurrence of upturned root mounds, coarse woody debris,broken crowns and stems suggested that wind rather than fire caused the small scale gap eventsidentified in these stands.In contrast to stands Si, S4 and N4, the influence of fire dominated the dynamicprocesses and had a significant effect on the composition and structure of stand Ni. Itschronology of total and effective ages revealed a multiple cohort structure with increased ingress69and releases during the nineteenth century and a large mode in the 1890’s. This chronologylikely reflects a response to a Ca. 1810 fire and possibly the influence of multiple low intensitydisturbances to the stand after that event (Taylor and Halpem 1992). The ca. 1810 fire andsubsequent growth response of surviving trees resulted in a high proportion of upper canopy treesrelative to the lower canopy and substantially more Abies in the upper canopy than what wasfound in the other stands. As the post-fire canopy was relatively continuous, gap dynamics havenot predominated in this stand and it will not be considered further in this discussion.Gap DynamicsAlthough gap dynamics dominated the canopy tree replacement process of stands S 1, S4and N4, few spatially discrete, even-aged (total age) cohorts were evident among the canopytrees. These results do not comply with the spatial-temporal equilibrium concept hypothesizedfor forests derived through gap dynamics (Watt 1947). Such forests are described as composedof a mosaic of “stands” (cohort clusters) of different stages of structural development. The sizeof each stand or cohort cluster depends on the disturbance event that creates the canopy gap;species composition is determined by gap size among other physical attributes, tolerance ofavailable colonizers, and time since disturbance. Theoretically, the sum of such stands is anuneven-aged forest, comprised of many even-aged stands, that is in a state of structural andcompositional dynamic equilibrium (Bonnicksen and Stone 1982, Whitmore 1982).This spatial-temporal concept has been described as the space-time mosaic (Watt 1947),shifting mosaic steady state (Bormann and Likens 1979), mosaic of aggregations (Bonnicksenand Stone 1981, 1982), and metapopulations composed of local populations and patchdemography (Kohyama 1993). Many intimate a mosaic of this type in coastal old-growth stands,given the structure and gap regeneration modes of these forests (Brady and Hanley 1984,Stewart 1989, Lertzman 1991, Inselberg 1993, Keenan 1993).In the study stands, even-aged patches of trees derived in gaps were ephemeral. Theseedlings and saplings that aggregated beneath canopy openings conform to the spatial mosaic70concept, but their aggregated spatial pattern diminished as trees advanced to the canopy. As gapage (time since disturbance) increased, the density of regeneration decreased (Spies et at. 1990)due to suppression, competition, and mortality, which limited the number of trees that advancedto the canopy within each gap.Where gap events had initiated establishment, they also facilitated release of trees; thus,gmups of canopy trees that responded together were largely uneven-aged. The longevity andshade tolerance of Thuja, Tsuga, and AMes contributed to the complexity of the gap replacementprocess. Gap formation was slow due to the long lifespan of canopy trees and the tendency forgaps to form progressively (Spies et a!. 1990, Lertzman and Krebs 1991). As Canham (1989)and Lertzman and Krebs (1991) point out, recruitment of tolerant trees to the dominant canopystrata can depend on multiple gap events that occur over extended periods of time. As noted inother forests (Lertzman 1989, Veblen et a!. 1989, Veblen et a!. 1992), gaps in the study standsmanifested a restructuring of the vertical position of trees through response of suppressed trees,but did not always facilitate the entry of those trees to the upper canopy. In many cases,recruitment of trees was prolonged over centuries and involved multiple gap events (Table 5.3).Gaps and Species’ResponsesAlthough Thuja, Tsuga, and Abies are shade tolerant (Krajina 1969, Klinka and Carter1992), the regeneration of each species was not independent of gap events. All three speciesestablished most frequently in conjunction with the establishment and/or release of theirneighbours, which suggests canopy gaps were conducive to their initial success. This assertion isconsistent with the aggregated distribution of seedlings and saplings of all three species in thestudy stands (Chapter 4) and Stewart’s (1986) observations of successful regeneration of Tsugaand Abies trees beneath canopy gaps.These patterns of regeneration contrast with Whitmore’s (1982, 1989) description of theregeneration mode of tolerant species as establishing under the canopy in shade with continuousingress (Rebertus et a!. 1991, Lusk and Ogden 1992), whereas intolerant species regenerate only71in gaps. However, Whitmore’s (1982, 1898) classification has been criticized by Schupp et al.(1989) as it fails to recognize that the regeneration process is multi-staged, includinggermination, establishment, survival, and recruitment to various strata (Pickett et at. 1987). Theclassification focuses on the establishment stage and so can explain the coexistence of species ofdisparate tolerances (e.g., VanWagner 1978, Veblen et at. 1981, Veblen and Stewart 1982, Readand hill 1985, Stewart 1986c, Spies and Franklin 1989, Kneeshaw 1992, Koyama 1993), but itde-emphasizes differences in development and growth in the latter stages and cannot account forthe coexistence of species of similar tolerances and regeneration niches, as was the case in thestudy stands.The growth patterns of the tree species revealed different responses to canopy gaps.Tsuga and Abies appeared dependent on distinct gaps in the canopy cover in order to recruit tothe upper canopy with their advancement occurring most frequently after release fromsuppression and usually in association with gap events. The infrequent releases of Thujasuggested that it did not depend on canopy gaps to gain dominance in the canopy; however, thesmall sample size of full ring-width series makes this assertion tenuous.Canham (1988, 1989) distinguishes a range of responses to canopy gaps among shadetolerant species which accommodates both the heterogeneity of the light environment beneath anold-growth canopy (Lieberman eta!. 1989, Schupp eta!. 1989, Chazdon and Pearcy 1991,Lertzman and Krebs 1991) and the potential for differential adaptations among species. On oneextreme of this range are species with juveniles that persist in a suppressed state and exhibitsignificant growth only under the influence of gaps and increased light availability (Canham1988, 1989). The growth patterns of Abies indicated opportunistic growth consistent with thisresponse to gaps. The majority of Abies trees were in the lower canopy and highly suppressed,although they had released numerous times. The few trees of this species that occurred in theupper canopy released after a stand-level disturbance which may have created an environment inwhich light was not a limiting resource and enabled them to grow into the canopy withoutsubsequent suppression. Tsuga also conformed to this type of growth response; however, it72exhibited a moderate degree of suppression with more recruits to the upper canopy and fewertrees that released multiple times restricted to the lower canopy.The opposite extreme in Canham’s (1988, 1989) range of responses to canopy gaps isthe slow, consistent growth of juveniles that do not respond greatly to gap events. Over thecourse of their life, such individuals likely are influenced by and respond to gaps, although theyare not dependent upon them (Schupp et a!. 1989). The observed growth of Thuja suggestseither (1) it is adapted to the low light environment beneath the canopy, with a complacentresponse to gaps, as described above; or (2) the successful establishment of Thuja occurs only ingaps in which it grows to the upper canopy before gap closure when light becomes a limitingresource (Table 5.1, Decision 2a). Carter and Klinka (1992) and Wang et a!. (1993) demonstratethat Thuja maintains apical dominance and height increment under low light conditions, andinterspecific comparisons indicate Thuja is better adapted to grow in low light than are Tsugaand Abies (Carter and Klinka 1992, Klinlca et a!. 1992). These results, in combination with theinfrequent release of dominant Thuja and its consistent diameter increment in all height classes,suggest that this species is adapted to grow in low light beneath the old-growth canopy and doesnot depend on increased light in gaps to gain the upper canopy.Evidently, Thuja, Tsuga, and Abies respond differently to the light environment beneaththe old-growth canopy, which, in effect, partitions the light environment among them and couldenable their coexistence. However, growth rates, physiological and morphological responses togaps and release merit further study for all three species.ConclusionsGap dynamics have dominated the dynamic processes in the true old-growth standsinvestigated in this study for a number of centuries. The spatial-temporal distribution of canopytrees (dbh >10 cm) within these stands were unique, in that the trees that responded to individualgaps were not of similar ages. Even-aged cohorts could not be identified within these stands, ascould be done in other gap-derived stands. A Ca. 1810 fire in stand Nl had significant influence73on the stand composition, structure and dynamics, and demonstrated that stand-level disturbancesin the study area are not always stand replacing events.Interspecific comparisons of the diameter increments, and the numbers of releases forcanopy trees of different height classes in relation to canopy gaps, revealed unique growth trendsas the study species recruited into and advanced within the canopy. These results demonstratethat differences exist in these species’ responses to canopy gaps and support the hypothesis thatthe populations of all three species are sustained within these old-growth stands.74Chapter 6. ConclusionsThuja plicata DynamicsThis study does not support the current paradigm that Thuja populations are in decline inthe CWH zone except on nutrient rich sites. I conclude that Thuja was neither competitivelyexcluded by Tsuga and Abies (Daubenmire and Daubenmire 1968) nor poorly adapted to the old-growth stands of the study area (Krajina 1969), despite its peculiar population structures. Thisspecies has evolved unique life history characteristics that enable it to successfully exploit smallscale heterogeneity within old-growth forests, to survive in the subcanopy environment, and tocoexist with Tsuga and Abies.This study demonstrates conclusively that Thuja populations in the study area, althoughdominated by a small number of large long-lived individuals, were composed of trees of all sizesand ages. Although fire initiated the ingress of Thuja in some study stands, and a local Thujadecline was observed in one stand, there is no support for the claim that Thuja requires standreplacing fires to establish as a canopy dominant in these forests; indeed successful Thuja ingressand recruitment was confirmed in stands in which small scale disturbances and gap dynamics hadpredominated for centuries. The fact that Thuja existed in the understory of all study standsverified it is capable of regeneration within old-growth forests.Sporadic recruitment of Thuja to the canopy, exhibited in discontinuous population agestructures, is indicative of the influence of disturbance on Thuja population dynamics. Smallscale disturbances, creating suitable microsites, are important in the establishment of Thuja, asindicated in this study by the aggregated spatial distribution of Thuja seedlings and saplings andtheir affinity for disturbed substrates. Although the relationship between regeneration anddisturbance is not obligate, the high density and survival of Thuja on mineral soil and theabundance of individuals on decaying wood reiterate Thuja’s dependence on these growing mediafor successful regeneration.75Neither establishment nor recruitment of Thuja was associated with the disturbance tothe canopy or the occurrence of canopy gaps. Canopy tree stem density was irrelevant to thespatial distribution of Thuja regeneration and the advancement of Thuja to the upper canopy waslargely independent of gap events. It is concluded that Thuja can grow without the benefit ofgaps; however, this conclusion is tenuous due to small sample sizes.Thuja’s unique life history characteristics, manifest in its population structures anddemographics, were critical to understanding this species’ dynamics. Thuja’s longevity, highmortality during germination and establishment but low mortality during the sapling and canopytree stages of its life cycle, consistent growth rates, and apparent independence from gaps duringrecruitment, differentiate it from Tsuga and Abies. Integration of these attributes can explain thepopulation structures and dynamics of Thuja and were used to demonstrate that Thuja is self-sustaining and does coexist with Tsuga and Abies in the study area.Comparison of Schools of Thought on SuccessionWhite (1979) eloquently addressed the role of disturbance in vegetation dynamics. Heasked, “What shapes vegetation?” Two alternative responses are proposed: “[1] disturbance freeperiods of competition and succession or [2] repeated disturbance and environmental fluctuation”(p. 260). The same question may be asked of factors influencing the regeneration processes ofThuja. If the former response is true, then the processes and mechanisms of Thuja dynamicsshould be adequately explained by the current paradigm which is based on the holistic view ofsuccession. Re-examination of Thuja dynamics, applying a reductionist approach whileconsidering disturbance, as in this study, should then lead to a similar conclusion. However, theconclusions of this study correspond to the second response above. As hypothesized analogouslyto White (1979), the influence of disturbance within the study stands was revealed by aninvestigation of stand composition and structure in view of disturbance and related populationresponses. This study demonstrated fundamental differences between the holistic andreductionist approaches to succession, as manifest in their different interpretations of Thuja76dynamics. The holistic school, or current paradigm, perceives Thuja as an unstable speciesunable to sustain its population in old-growth forests of the Very Wet Maritime Coastal WesternHemlock subzone. The reductionist approach, taken in this study, describes Thuja as a speciesuniquely adapted to the old-growth environment, that regenerates by exploitation of small scaleheterogeneity.Future StudiesBased on the life history characteristics of Thuja and the types of disturbances observedin the study area, I have developed two hypotheses to describe Thuja-population and standdynamics. For each I have identified areas for further study.The first hypothesis is based on the storage effect (Ri.mlde 1989) and relates to theregeneration processes and growth and mortality rates of Thuja. The storage effect states thatregeneration success need not be frequent for long-lived species of low density. Those speciesthat rely on environmental heterogeneity to reproduce, as does Thuja for establishment, willexhibit great variation in regeneration abundance in time and space (Fox 1977, June and Ogden1978), which increases the probability of occasional success. The relative increase of lowdensity species in good years can be great, and the loss in bad years can be relatively small. Thisvariation is consistent with the intermittent nature of Thuja recruitment and population agestructures for Thuja observed by Daniels et a!. (1994). The fact that Thuja exhibits low juvenileand adult mortality and consistent annual diameter increment, regardless of tree height, suggeststhat its chance of future presence in the canopy is high once established. Few individuals wouldbe needed in space and time to sustain the canopy composition; thus, the understory componentof a population would reveal liffle about a species’ ability to gain the canopy (Stewart and Rose1990).To further test the fit of the storage effect to Thuja population dynamics, specifichypotheses could address the proposed high recruitment rates of advance regeneration and low77mortality of post-establishment individuals. To do so would require long term data on therecruitment rates and mortality rates of Thuja of different sizes, ages, and canopy positions.The second hypothesis is that the observed interspecific differences in populationstructure and mortality relate to different species’ adaptations to the light environment within old-growth forests. Unlike the storage effect, this hypothesis does not assume a longevity-densityrelationship for the study species. Rather, species dependent on gaps are assumed to rely onstochastic events to facilitate their growth. This implies prolonged periods of suppression forlarge portions of the population, with which is associated a relatively high chance of mortality(Cartham 1989). The abundance of regeneration in the understory provides a “reserve” ofindividuals to respond rapidly when a gap becomes available (Ogden 1985, Leemans 1991,Keenan 1993). This suggests an abundance of seedlings with declining numbers in successiveage or size classes throughout the population, as observed for Tsuga and Abies. A species thatdoes not respond greatly to gaps, such as hypothesized for Thuja, also must be tolerant of longperiods of suppression to be self-sustaining. Its mortality may be lower and/or it may not requirea large abundance of recruits to sustain the canopy.To test this hypothesis, it must be determined whether the recruitment of Thuja requiresa change in available light such as occurs within canopy gaps. This question requires betterknowledge of age- and/or size-specific growth and mortality rates, recruitment patterns, andgrowth response to ephemeral and diffuse light that is typical of subcanopy old-growthenvironments and the light environment in canopy gaps. Interspecific comparisons are advised.78Literature CitedAdams, D.L. and R.L. Mahoney. 1991. Effects of shade and competing vegetation on growth ofwestern redcedar regeneration. West. J. Appi. For. 6: 21-22.Agriculture Canada Expert Committee on Soils. 1987. The Canadian System of SoilClassjfication. 2nd ed. Agric. Can. Pi.ibl. 1646. 164 pp.Alaback, P.B. 1984. A comparison of old-growth forest structure in the western hemlock-Sitkaspruce forests of southeast Alaska. Pp. 219-226 in Meehan, W.R., T.R. Merrell, Jr., andT.A. Hanley (eds.) Proc. of the Fish and Wildlife Relationships in Old-growth ForestsSymposium held 12-15 April 1982, Juneau, Alaska.Banner, A., R.N. Green, A. Inselberg, K. Klinka, D.S. McLennan, D.V. Meidinger, F.C.Nuszdorfer, and J. Pojar. 1990. Site classification for coastal British Columbia. Ministryof Forests, Victoria, B.C. 2 pp.BCMOF (British Columbia Ministry of Forests). 1992. An old-growth strategy for BritishColumbia. Report by the Old Growth Task Force. B.C. Mm. of For., Victoria, B.C.Beese, W.J. and J.S. Sandford. 1992. Age of old-growth forests on Vancouver Island. B.C.Ministry of Forests Contract Report #98444, MacMillan Bloedel Ltd., WoodlandsServices Division, Nanaimo, B.C. 19 pp.Begon, M. and M. Mortimer. 1986. Population Ecology: A Unified Study ofAnimals and Plants.2nd ed. Blackwell Scientific Publications, London. 220 pp.Bonnicksen, T.M. and E. C. Stone. 1981. The giant sequoia-mixed conifer forest communitycharacterized through pattern analysis as a mosaic of aggregations. For. Ecol. Manage. 3:307-328.Bonnicksen, T.M. and E.C. Stone. 1982. Reconstruction of a presettlement giant sequoia-mixedconifer forest community using the aggregation approach. Ecology 63: 1134-1148.Bormann, F.H. and G.E. Likens. 1979. Pattern and Process in a Forested Ecosystem. SpringerVerlag, New York. 253 pp.Brady, W.W. and T.A. Hanley. 1984. The role of disturbance in old-growth forests: sometheoretical implications for southeastern Alaska. Pp. 213-218 in Meehan, W.R., T.R.Merrell, Jr., and T.A. Hanley (eds.) Proc. of the Fish and Wildlife Relationships in Oldgrowth Forests Symposium held 12-15 April 1982, Juneau, Alaska.Canadian Qimate Program. 1993. Canadian climate normals 1961 - 1990, British Columbia.Ministry of Supplies and Services of Canada, Ottawa. 134 pp.Canham, C.D. 1988. Growth and canopy architecture of shade-tolerant trees: response to canopygaps. Ecology 69: 786-795.79Canham, C.D. 1989. Different responses to gaps among shade-tolerant tree species. Ecology 70:548-550.Carter, R.E. and K. Klinka. 1992. Variation in shade tolerance of Douglas fir, western hemlock,and western red cedar in coastal British Columbia. For. Ecol. Manage. 55: 87-105.Chazdon, R.L. and R.W. Pearcy. 1991. The importance of sunflecks for forest understory plants.BioScience 41: 760-766.Chesson, P.L. and R.R. Warner. 1981. Environmental variability promotes coexistence in lotterycompetitive systems. Am. Nat. 117: 923-943.Christy, E.J. and R.N. Mack. 1984. Variation in demography of juvenile Tsuga heterophyllaacross the substratum mosaic. 3. Ecol. 72: 75-91.Christensen, N.L. 1977. Changes in structure, pattern and diversity associated with climax forestmaturation in Piedmont, North Carolina. Amer. Midl. Nat. 97: 176-188.Qements, R.F. 1916. Plant succession: an analysis of the development of vegetation. CarnegieInst. Pub. 242, Washington, D.C..Connell, J.H. and R.O. Slatyer. 1977. Mechanisms of succession in natural communities andtheir role in community stability and organization. Amer. Natur. 111: 1119-1144.Cunia, T. 1973. Dummy variables and some of their uses in regression analysis. Proc. IUFROSubject Group S4.02. 1: 1-146.Daniels, L.D., P.L. Marshall, R.E. Carter, K. Klinka. 1994. Age structure of western redcedar inthe tree layer of old-growth stands near Vancouver, British Columbia. Submitted toNorthwest Science.Daubenmire, R. 1966. Vegetation: identification of typal communities. Science 151: 291-298.Daubenmire, R. and J.B. Daubenmire. 1968. Forest vegetation of eastern Washington andnorthern Idaho. Washington State Univ. Agric. Exp. Sm. Tech. Bull. 60. Pullman, WA.lo4pp.Dobry, 3. and K. Klinka. 1993. Dendrochronology of Pacific silver fir, western redcedar, andyellow-cedar in the Greater Vancouver Water District. 1992/93 Prog. Rep to the GVWD,Watershed Management, Burnaby, B.C. 13 pp.Deevey, E.S. 1947. Life tables for natural populations of animals. Quart. Rev. Biol. 22: 283-3 14.Donoso, C., V. Sandoval, R. Grez, and 3. Rodriguez. 1993. Dynamics of Fitzroya cupressoidesforests in southern Chile. J. Veg. Sci. 4: 303-3 12.Drury, W.H. and I.C.T. Nisbet. 1973. Succession. 3. Arnold Arbor. 54: 33 1-368.80Duncan, R.P. and G.H. Stewart. 1991. The temporal and spatial analysis of tree age distributions.Can. J. For. Res. 21: 1703-1710.Edmonds, R.L., T.B. Thomas, and K.P. Maybury. 1993. Tree population dynamics, growth, andmortality in old-growth forests in the western Olympic Mountains, Washington. Can. J.For. Res. 23: 512-519.Egler, F.E. 1954. Vegetation science concepts. 1. Initial floristic composition, a factor in old-field vegetation development. Vegetatio 4: 412-417.Eis, S. 1962. Statistical analysis of several methods of estimation of forest habitats and treegrowth near Vancouver, B.C. UBC, Fac. of For. Bull. No. 4.Fisher, G.M. 1935. Comparative germination of tree species on various kinds of surface-soilmaterial in the western white-pine type. Ecology 16: 606-611.Forcier, L.K. 1975. Reproductive strategies and the co-occurence of climax tree species. Science189: 808-8 10.Fox, J.F. 1977. Alternation and coexistence of tree species. Amer. Natur. 111: 69-89.Franklin, J.F. 1961. A guide to seedling identification for 25 conifers of the Pacific Northwest.USDA For. Ser., PNW For. and Range Exp. Sm. Portland. 65 pp.Franklin, J.F. and D.S. DeBell. 1988. Thirty-six years of tree population change in an old-growthPseudotsuga-Tsuga forest. Can. J. For. Res. 18: 633-639.Franidin, J.F. and C.T. Dymess. 1973. Natural vegetation of Oregon and Washington. USDAGen. Tech. Rep. PNW-8., PNW For. and Range Exp. Sm. pp. 53-108.Franklin, J.F. and M.A. Hemstrom. 1981. Aspects of succession in the coniferous forests of thePacific Northwest. Pp. 212-229 in West, D.C., H. Shugart, and D.B. Botkin (eds.).Forest Succession: Concepts and Application. Springer Verlag, New York.Fritts, H.C. 1976. Tree Rings and Climate. Academic Press, New York. 565 pp.Fritts, H.C. and T.W. Swetnam. 1986. Dendroecology: A Toolfor Evaluating Variation in Pastand Present Environments. Laboratory of Tree-Ring Research. University of Arizona,Tucson, Arizona. 61 pp.Graham, R.T. 1982. Influence of tree and site factors on western redcedar’s response to release: amodeling analysis. USDA For. Serv. Res. Pap. INT-296, Inter. For. and Range Exp. Sm.Ogden, UT. 18 pp.Graham, R.T., R. L. Mahoney, and D.E. Ferguson. 1988. Regeneration and early growth ofwestern redcedar in the northern Rocky Mountains. Pp. 33-38 in Smith, N.J. (ed.)Western red cedar - does it have a future? Conf. Proc., Univ. of British Columbia, Fac. ofForestry, July 13-14, 1987.81Gregory, R.A. 1957. Some silvicultural characteristics of western redcedar of southeast Alaska.Ecology 38: 646-649.Greig-Smith, P. 1983. Quantitative Plant Ecology. 3rd ed. Univ. of California Press, Berkeley,CA. 359 pp.Grubb, P.J. 1977. The maintenance of species richness in plant communities. The importance ofthe regeneration niche. Biol. Rev. 52: 107-145.Greater Vancouver Water District (GVWD). 1991. Forest cover maps. Greater VancouverRegional District, Watershed Management. Burnaby, B.C.Habeck, J.R. 1968. Forest succession in the Glacier Park cedar-hemlock forests. Ecology 49:872-880.Habeck, J.R. 1978. A study of climax western redcedar (Thuja plicata Donn.) forest communitiesin the Selway-Bitterroot Wilderness, Idaho. Northwest Sci. 52: 67-76.Hamilton, E. and 3. Pojar. 1991. Old growth forests in British Columbia. Pp. 450-457 in Proc. ofthe XIX IURFO World Congress held 5-11 August, 1990 Montreal, Quebec.Harmon, M.E. 1987. The influence of litter and humus accumulations and canopy openness onPicea sitchensis (Bong.) Carr. and Tsuga heterophylla (Raf.) Sarg. seedlings growing onlogs. Can. 3. For. Res. 17: 1475-1479.Harmon, M.E. and J.F. Franklin. 1989. Tree seedlings on logs in Picea-Tsuga forests of Oregonand Washington. Ecology 70: 48-59.Harper, J.L. 1977. Population Biology ofPlants. Academic Press, New York. 892 pp.Henderson, J.A., D.H. Peter, R.D. Lesher, and D.C. Shaw. 1988. Forest plant associations of theOlympic National forest. USDA For. Serv. PNW Region, R6-ECOL-TP 001-88. 502 pp.Henry, J.D. and J.M.A. Swan. 1974. Reconstructing forest history from live and dead plantmaterial - an approach to the study of forest succession in southwest New Hampshire.Ecology 55: 772-783.Hett, J.M. and O.L. Loucks. 1968. Application of life-table analyses to tree seedlings in QueticoProvincial Park, Ontario. For. Chron. 44: 29-32.Hett, J.M. and O.L. Loucks. 1976. Age structure model of balsam fir and eastern hemlock. 3.Ecol. 64: 1029-1044.Hitchcock, C.L. and A. Cronquist. 1973. Flora of the Pacific Northwest. University ofWashington Press, Seattle. 730 pp.Hofgaard, A. 1993. Structure and regeneration patterns in a virgin Picea abies forest in northernSweden. J. Veg. Sci. 4: 601-608.82Horn, H.S. 1975. Markovian processes in forest succession. Pp. 196-211 in Cody, M.L. and J.M.Diamond (eds.) Ecology and Evolution ofCommunities. Harvard Univ. Press.Cambridge, Mass.Horn, H.S. 1981. Some causes of variability in patterns of secondary succession. Pp. 24-35 inWest, D.C., H.H. Shugart, and D.B. Botkin (eds.) Forest Succession: Concepts andApplication. Springer Verlag, New York.Huribert, S.H. 1990. Spatial distribution of the montane unicorn. Oikos 58: 257-271.Inselberg, A.E. 1993. Tofino Creek watershed: ecological values and interpretations (DraftReport). Contract Report for the B.C. Ministry of Forests, Victoria, B.C. 68 pp.Jenny, H. 1958. Role of the plant factor in the pedogenic functions. Ecology 39: 5-16.Jozsa, L. 1988. Increment core sampling techniques for high quality cores. Special Pub. No. SP30. Forintek Canada Corp, Vancouver, B.C. 26 pp.June, S.R. and 3. Ogden. 1978. Studies on the vegetation of Mount Colenso, New Zealand 4. Anassessment of the processes of canopy maintenance and regeneration strategy in a redbeech (Nothofagusfusca) forest. N.Z.J. Ecol. 1: 7-15.Keenan, R. 1993. Structure and function of western redcedar and western hemlock forests onnorthern Vancouver Island. Ph.D. Thesis, Dept. of Forestry, Univ. of British Columbia,Vancouver, B.C. 167 pp.Kershaw, K.A. 1973. Quantitative and Dynamic Plant Ecology. 2nd ed. American Elsevier, NewYork. 308 pp.Khazanie, R. 1986. Elementary Statistics In a World ofApplications, 2nd ed. Scott, Foresmanand Company, Glenview, fflinois. 562 pp.Kimmins, J.P. 1987. Forest Ecology. Macmillan Publishing Co., New York. 531 pp.Klinka, K., R.E. Carter, and M.C. Feller. 1990. Cutting old-growth forests in British Columbia:ecological consideration for forest regeneration. NW Enviro. 3. 6: 221-242.Klinka, K. and M.C. Feller. 1993. Regeneration of western redcedar in the Very Wet MaritimeCoastal Western Hemlock subzone. 1992/93 Progress Report, B.C. Mm. of Forests,Vancouver Forest Region, Bumaby, B .C. 11 pp.Klinka, K, V.3. Krajina, A. Ceska, and A.M. Scagel. 1989. Indicator Plants of Coastal BritishColumbia. Univ. of British Columbia Press. Vancouver, B.C. 288 pp.Klinka, K., 3. Pojar and D.V. Meidinger. 1991. Revision of the biogeoclimatic units of coastalBritish Columbia. Northwest Sci. 65: 32-47.83Klinka, K., Q. Wang, G.J.K. Kayahara, and B.A. Blackwell. 1992. Light-growth relationships inPacific silver fir (Abies amabilis) and subalpine fir (Abies lasiocarpa) Can. 3. Bot. 70:1919-1930.Kneeshaw, D. 1989. Tree population dynamics of some old sub-boreal spruce stands. M.Sc.Thesis. Dept. of Forestry, Univ. of British Columbia, Vancouver, B.C. 153 pp.Knowles, P. and M.C. Grant. 1983. Age and size structure analyses of Engelmann spruce,ponderosa pine, lodgepole pine, arid lhnber pine in Colorado. Ecology 64: 1-9.Kohyama, T. 1993. Size-structured tree populations in gap-dynamic forest - the forestarchitecture hypothesis for the stable coexistence of species. 3. Ecol. 81: 131-143.Kozlowski, T.T. and C.E. A1ilgren. 1974. Fire and Ecosystems. Academic Press, New York. 542pp.Krajina, V.3. 1965. Biogeoclimatic zones of British Columbia. Ecol. West. N. Am. 1: 1-17.Krajina, V.3. 1969. Ecology of forest trees in British Columbia. Ecol. West. N. Am. 2: 1-152.Kuiper, L.C. 1988. The structure of natural Douglas-fir forests in western Washington andwestern Oregon. Agric. Univ. Wageningen (Netherlands) Papers 88-5. 45 pp.Leak, W.B. 1965. The J-shaped probability distribution. For. Sci. 11: 405-409.Leemans, R. 1991. Canopy gaps and establishment patterns of spruce (Picea abies (L.) Karst.) intwo old-growth coniferous forests in central Sweden. Vegetatio 93: 157-165.Lertzman, K.P. 1989. Gap-phase community dynamics in a sub-alpine old growth forest. Ph.D.Thesis, Dept. of Zoology, Univ. of British Columbia, Vancouver, B.C. 167 pp.Lertzman, K.P. 1992. Patterns of gap-phase replacement in a subalpine, old-growth forest.Ecology 73: 657-669.Lertzman, K.P. and C.J. Krebs. 1991. Gap-phase structure of a sub-alpine old growth forest. Can.3. For. Res. 21: 1730-1741.Liebennan, M., D. Lieberman, and R. Peralta. 1989. Forests are not just Swiss cheese: canopystereogeometry of non-gaps in tropical forests. Ecology 70: 550-552.Lorimer, C.G. 1980. Age structure and disturbance history of a southern Appalachian virginforest. Ecology 61: 1169-1184.Lorimer, C.G. 1985. Methodological considerations in the analysis of forest disturbance history.Can. 3. For. Res. 15: 200-213.Lorimer, C.G. and L.E. Frelich. 1989. A methodology for estimating canopy disturbancefrequency and intensity in dense temperate forests. Can. 3. For. Res. 19: 65 1-663.84Lusk, C. and J. Ogden. 1992. Age stmcture and dynamics of a podocarp-broadleaf forest inTongariro National Park, New Zealand. J. Ecol. 80: 379-393.Luttmerding, H. A., D.A. Dennachi, E.C. Lea, D.V. Meidinger, and T. Void. 1990. DescribingEcosystems in the Field, 2nd Ed. MOE Manual 11, Ministry of the Environment,Victoria, B.C. 213 pp.McIntosh, R.P. 1981. Succession and ecological theory. Pp. 10-23 in West, D.C., H.H. Shugart,and D.B. Botkin (eds.). Forest Succession: Concepts and Application. Springer-Verlag,New York.Minore, D. 1972. Germination and early growth of coastal tree species on organic seedbeds.USDA For. Sew. Res. Pap. PNW-135. PNW For, and Range Exp. Stn., Portland, OR. 18pp.Minore, D. 1983. Western redcedar - a literature review. USDA For. Sew. Gen. Tech. Rep.PNW-150. PNW For. and Range Exp. Stn., Portland, OR. 70 pp.Moeur, M. 1992. Baseline demographics of late successional western hemlock/western redcedarstands in northern Idaho research natural areas. USDA For. Sew. Res. Pap. INT-456, mt.Res. Sm. Ogden, UT. 16 pp.Morisita, M. 1959. Measuring of interspecific association and similarity between communities.Mem. Fac. Sci. Kyushu Univ. Ser. E. (Biol.) 3: 65-80.Morrison, P.H., and F.J. Swanson. 1990. Fire history and pattern in a Cascade Range landscape.USDA, For. Sew. Gen. Tech. Rep. PNW-GTR-254, PNW Res. Stn.. Portland, OR. 77pp.Mueller-Dombois, D. and H. Ellenberg. 1974. Aims andMethods of Vegetation Ecology. JohnWiley and Sons, New York. 574 pp.Munger, T.T. 1940. The cycle from Douglas fir to hemlock. Ecology 21: 451-459.Noble, I.R. and R.O. Slatyer. 1980. The use of vital attributes to predict successional changes inplant communities subject to recurrent disturbances. Vegetatio 43: 5-21.Odum, E. P. 1969. The strategy of ecosystem development. Science 164: 262-270.Ogden, J. 1985. An introduction to plant demography with special reference to New Zealandtrees. N.Z.J. Bot. 23: 75 1-772.Oliver, C.D. 1981. Forest development in North America following major disturbances. For.Ecol. Manage. 3: 153-168.Oliver, C.D. and B.C. Larson. 1990. Forest Stand Dynamics. McGraw-Hill Inc., New York. 467pp.85Oliver, C.D., M.N. Nystrom, and D.S. DeBeil. 1988. Coastal stand silvicultural potential forwestern redcedar. Pp. 39-46 in Smith, N.J. (ed.) Western red cedar - does it have afuture? Conf. Proc., Univ. of British Columbia, Fac. of Forestry, July 13-14, 1987.Oliver, C.D. and E.P. Stephens. 1977. Reconstruction of a mixed-species forest in central NewEngland. Ecology 58: 562-572.Ott, L. 1984. An Introduction to StatisticalMethods and Data Analysis. 2nd ed. PW Publishers,Boston. 775 pp.Parker, T. 1979. Natural regeneration of western redcedar. M.Sc. Dissertation, Coil. of Forestry,Wildlife and Range Sciences, Univ. of Idaho, Moscow, Idaho. 50 pp.Parker, T. 1986. Ecology of western redcedar groves. Ph.D. Dissertation, Coil. of Forestry,Wildlife and Range Sciences, Univ. of Idaho, Moscow, Idaho. 187 pp.Parker, A.J. and R.K. Peet. 1984. Size and age structure of conifer forests. Ecology 65: 1685-1689.Payette, S., L. Filion and A. Deiwaide. 1990. Disturbance of a cold temperate forest as deducedfrom tree-ring patterns: the Tantare Ecological Reserve, Quebec. Can. J. For. Res. 20:1228-1241.Pearl, R. 1928. The Rate ofLiving. A.A. Knopf Inc., New York. 185 pp.Peet, R.K. and N.L. Christensen. 1980. Succession: a population process. Vegetatio 43: 13 1-140.Peet, R.K. and N.L. Christensen. 1987. Competition and tree death. BioScience 37: 586-595.Pickett, S.T.A. 1976. Succession: an evolutionary interpretation. Amer. Natur. 110: 107-119.Pickett, S.T.A., S.L. Collins, and J.J. Armesto. 1987. A hierarchical consideration of causes andmechanisms of succession. Vegetatio 69: 109-114.Pielou, E.C. 1977. Mathematical Ecology. Wiley, NewYork. 385 pp.Pojar, J., K. Klinka, and D.A. Demarchi. 1991. Coastal western hemlock zone. Pp. 95-111 inMeidinger, D.V. and J. Pojar (eds.) Ecosystems ofBritish Columbia. B.C. Ministry ofForests Special Report Series 6. Victoria, B.C.Read, J. and R.S. Hill. 1985. Dynamics of Nothofagus dominated rainforest on mainlandAustralia and lowland Tasmania. Vegetatio 63: 67-78.Rebertus, A.J., B.R. Bums, and T.T. Veblen. 1991. Stand dynamics of Pinusfiexilis-dominatedsubalpine forests in the Colorado Front Range. I. Veg. Sci. 2: 445-45 8.Rebertus, A.J. and T.T. Veblen. 1993. Structure and tree-fall gap dynamics of old-growthNothofagus forests in Tierra del Fuego, Argentina. J. Veg. Sci. 4: 641-654.86Ross, M.S., T. L. Shank, and D.Wm. Smith. 1982. Age-structure relationships of tree species ina Appalachian oak forest in southwest Virginia. Bull. Torr. Bot. Club 109: 287-298.Runkle, J.R. 1981. Gap regeneration in some old-growth forests of the eastern United States.Ecology 62: 1041-1051.Runkle, J.R. 1989. Synchrony of regeneration, gaps, and latitudinal differences in tree speciesdiversity. Ecology 70: 546-547.Runkle 1992. Guidelines and sample protocol for sampling forest gaps. USDA For. Sew. Gen.Tech. Rep. PNW-GTR-283, PNW Res. Stn., Portland OR. 44 pp.Schmidt, R.L. 1955. Some aspects of western red cedar regeneration in the coastal forests ofBritish Columbia. B.C. Dept. of Lands and Forests, Res. Note 29, Victoria, B.C. 10 pp.Schofield, W.B. 1992. Some Common Mosses ofBritish Columbia. Royal British ColumbiaMuseum, Victoria. 394 pp.Schupp, E.W., H.F. Howe, C.K. Augspurger, and D.J. Levey. 1989. Arrival and survival intropical treefall gaps. Ecology 70: 562-564.Schweingruber, F.H. 1989. Tree Rings: Basics andApplications ofDendrochronology. KluwerAcademic Press, Dordrecht, Holland. 276 pp.Soos, J. and 3. Walters. 1963. Some factors affecting the mortality of western hemlock andwestern red cedar germinates and seedlings. Fac. For., U.B.C., Res. Pap. 56. 12 pp.Spies,T.A. and J.F. Franklin. 1989. Gap characteristics and vegetation response in coniferousforests of the Pacific Northwest. Ecology 70: 543-545.Spies, T.A., J.F. Franklin, and M. Klopsch. 1990. Canopy gaps in Douglas-fir forests of theCascade Mountains. Can. 3. For. Res. 20: 649-658.Stewart, G.H. l986a. Population dynamics of a montane conifer forest, western Cascade Range,Oregon, USA. Ecology 67: 534-544.Stewart, G.H. 1986b. Forest development in canopy openings in old-growth Pseudotsuga forestsof the western Cascade Range, Oregon. Can. J. For. Res. 16: 558-568.Stewart, G.H. l986c. Forest dynamics and disturbance in a beech/hardwood forest, Fiordland,New Zealand. Vegetatlo 68: 115-126.Stewart, G.H. 1989. The dynamics of old-growth Pseudotsuga forests in the western CascadeRange, Oregon, USA. Vegetatio 82: 79-94.Stewart, G.H. and A.B. Rose. 1990. The significance of life history strategies in thedevelopmental history of mixed beech (Nothofagus) forests, New Zealand. Vegetatio 87:101-114.87Tansley, A.G. 1935. The use and abuse of vegetational concepts and terms. Ecology 16: 284-307.Taylor, A.H. and C.B. Halpem. 1991. The structure and dynamics ofAbies magnifica forests inthe southern Cascade Range, USA. J. Veg. Sci. 2: 189-200.Turner, D.P. and E.H. Franz. 1985. Size class structure and tree dispersion patterns in old-growthcedar-hemlock forests of the northern Rocky Mountains (USA). Oecologia 68: 52-56.Turner, D.P. and E.H. Franz. 1986. The influence of canopy dominants on understory vegetationpatterns in an old-growth cedar-hemlock forest. Am. Midi. Nat. 116: 387-393.Van Cleve, K., F.S. Chapin III, C.T. Dymess, and L.A. Viereck. 1991. Element cycling in taigaforests: state-factor control. BioScience 41: 78-88.Van Wagner, C.E. 1978. Age-class distribution and the forest fire cycle. Can. 3. For. Res. 8: 220-227.Veblen, T.T. 1986. Age and size structure of subalpine forests in the Colorado Front Range.Bull. Torr. Bot. Club 113: 225-240.Veblen, T.T. 1989. Nothofagus regeneration in treefall gaps in northern Patagonia. Can. 3. For.Res. 19: 365-371.Veblen, T.T., D.H. Ashton, and F.M. Schlegel. 1979. Tree regeneration strategies in a lowlandNothofagus-dominated forest in south-central Chile. 3. Biogeog. 6: 329-340.Veblen, T.T., C. Donoso Z., F.M. Schiegel, and B. Escobar R. 1981. Forest dynamics in south-central Chile. J. Biogeog. 8: 211-247.Veblen, T.T., K.S. Hadley, M.S. Reid, and A.J. Rebertus. 1989. Blowdown and standdevelopment in a Colorado subalpine forest. Can. 3. For. Res. 19: 1218-1225.Veblen, T.T., T. Kitzberger, and A. Lara. 1992. Disturbance and forest dynamics along a transectfrom Andean rain forest to Patagonian shrubland. 3. Veg. Sci. 3: 507-520.Veblen, T.T., F.M. Schiegel, and B. Escobar R. 1980. Structure and dynamics of old-growthNothofagus forests in the Valdivian Andes, Chile. J. Ecol. 68: 1-31.Veblen, T.T. and G.H. Stewart. 1982. On the conifer regeneration gap in New Zealand: thedynamics of Libocedrus bidwillii stands on South Island. J. Ecol. 70: 413-436.Wang, G.G., H. Qian, and K. Klinka. 1994. Growth of Thujaplicata seedlings along a lightgradient. Submitted to Canadian Journal of Botany.Warner, R.R and P.L. Chesson. 1985. Coexistence mediated by recruitment fluctuations: a fieldguide to the storage effect. Am. Nat. 125: 769-787.Watt, A.S. 1947. Pattern and process in the plant community. 3. Ecol. 35: 1-22.88Whipple, S.A. and R.L. Dix. 1979. Age structure and successional dynamics of a Coloradosubalpine forest. Am. Midl. Nat. 101: 142-158.White, P.S. 1979. Pattern, process, and natural disturbance in vegetation. Bot. Rev. 45: 229-299.White, P.S., M.D. MacKenzie, and R.T. Busing. 1985. A critique of overstory/understorycomparisons based on transition probability analysis of an old-growth stand in theAppalachians. Vegetatio 64: 37-45.Whittaker, R.H. 1953. A consideration of climax theory: the climax as a population and pattern.Ecol. Monog. 23: 41-78.Whittaker, R.H. 1975. Communities and Ecosystems. 2nd ed. Macmillan Pub. Co., Inc., NewYork. 385 pp.Whitmore, T.C. 1982. On pattern and process in forests. Pp. 45-59 in Newman, E.I. (ed.) ThePlant Community as a Working Mechanism. Blackwell Scientific Publications. London.Whitmore, T.C. 1989. Canopy gaps and the two major groups of forest trees. Ecology 70: 536 -538.Zar, J.H. 1984. Biostatistical Analysis. 2nd ed. Prentice-Hall, New Jersey. 718 pp.APPENDIX A: Mean percent cover (per quadrat) of understory species in eight old-growth study stands.89SPECIES PLOT:Si S2 S3 S4 Ni N2 N3 N4Gauitheria shalionMenziesia ferrugineaRubus speetabilisVaccinium ovalifo(iumVaccinium paivifoliumSum Shrubso 0.08 00 0.12 00.36 0 00.8 0.6 2.280.44 0.52 0.921.6 1.32 3.20 0 0 0 00.12 0 0.12 0.08 0.361.12 0.6 0 2.68 08.36 5.04 2.44 8 5.80.48 024 0.52 1.08 0.4810.08 5.88 3.08 11.84 6.64Athyrium felix-feminaBiechnum spicantBoykinia elataChimphila menziesiiClintonia unifloraCoptis asplenifoliaComus canadensisDisporum hookeriDryopteris expansaGallium trifk’,umGoodyera oblongifoliaGymnocarpium dryopterisLinnaea borealisLystichitum americanumMalanthemum dilatatumOpiopanax hor,idusPoystichum munitumRubus pedatusSambucus racemosaStreptopus amplexifoliusStreptopus roseusStreptopus spp.Tiarella trifoliataTiarella unifollataTiarella spp.Viola glabellaSum Herbs0.64 0.04 0.249.12 2.16 4.440.12 0 00 0 0.20.08 0.36 0.240 0 00.88 0.4 0.760 0 00.6 0.04 00.08 0 00 0.16 0.160.2 0.04 00.16 0.12 00.2 0 0.120.44 0 0.082.32 0 0.080.48 0.08 00.2 0.04 0.20 0 00 0 0.160.36 0.2 0.360 0 00.32 0 0.320.56 0 00 0.48 0.040 0 016.76 4.12 7.402 1.44 0.32 0.2 0.525.96 5.64 1.84 1.84 3.240 0 0 0 00 0 0 0 00.24 0.16 0.04 0.4 0.360 0 0 0.04 0.520.92 0.6 0.76 028 0.60.52 0 0 0 00.48 4.4 1.56 12 1.320 0 0 0 00 0 0 0 00 0 0.48 0.16 00 0.12 0.4 0 01.8 0 0 0 0.080 0.28 0 0.08 0.121.36 0.84 1.16 1.48 00 0.08 0 0 00.56 0 0.52 0.24 1.040 0.56 0 0 00.48 0.04 0 0 0.560 1.72 0 0.04 00.16 0 0.36 0.4 00 0.4 0.12 0 0.760.32 0 0 0 0.60.76 0.76 0.32 0.32 00.12 0 0 0 013.88 17.04 7.88 6.68 9.72Bathilophozia spp.Conocephalum conicumDicranum fuscescensDicranum scopariumHylocomium splendensIsothecium stoloniferumKlndbergia oreganaLycopodium clavatumPlagiochila porelloidesPlagiomnium insignePogonatum alpinumPoytrichum spp.Rhizomnium glabrescensRhytidiadelphus loreusRhytidiopsis robustaSphagnum girgensohnhiSum Bryophytes0 0 00.16 0 00.08 0 0.120 0 00 0.04 0.160 0 00.64 0 00.24 0 0.080 0 00 0 00.2 0 00 0 00.08 0 01.84 0.32 0.042.12 1.12 0.61 0 06.36 1.48 10.6 0 0.72 0 00 0 0 2.52 00.12 0 0 0 00.16 0 0 0 01.56 0 0.04 0.12 0.20.08 0 0 0 00.4 0 0 0.08 0.120 0.04 0 0 0.120.16 0 0.16 0.12 00 0 0.04 0 00 0.04 0 0 00 0 0 0 01.44 0.76 0.52 0.96 0.520 0 3.44 1.76 0.3615.28 7.76 4.28 4.68 6.724.04 0 0 0.32 0.5223.84 8.6 9.2 10.56 8.5690APPENDIX B: Summary of substrate utilization by seedlings and saplings on each study plot.(All regeneration)Plot Substrate Substrate Frequency Relative Density Relative SubstrateCover (%) Frequency Density Use (Rank)51FF 0.43 72 0.15 16725 0.10 3DW 0.41 318 0.67 770.44 0.48 1.5MS 0.11 86 0.18 754.39 0.45 1.5S2FF 0.55 75 0.15 135.62 0.04 3DW 0.42 386 0.75 925.66 0.30 2MS 0.03 53 0.10 1981.31 0.65 1S3FE 0.46 141 0.16 305.53 0.04 3DW 0.52 726 0.80 1405.61 0.19 2MS 0.01 42 0.05 5600.00 0.7754FF 0.71 23 0.17 32.62 0.05 3DW 0.22 107 0.79 495.94 0.72 1MS 0.06 9 0.07 160.00 0.23 2NiFE 0.46 1 0.02 2.19 0.00 3DW 0.37 21 0.40 56.45 0.09 2MS 0.05 30 0.58 600.00 0.91 1N2FE 0.37 10 0.04 26.79 0.04 3DW 0.46 224 0.91 488.55 0.82 1MS 0.13 ii 0.04 82.71 0.14 2N3FE 0.56 35 0.23 62.53 0.05 3DW 0.40 88 0.60 219.31 0.18 2MS 0.03 30 0.21 937.50 0.77N4FF 0.66 7 0.04 10.57 0.01 3DW 0.28 150 0.90 530.97 0.70MS 0.05 10 0.06 220.99 0.29 2For notation see Chapter 4. Methods and Results and Table 4.2Thuja plicata91APPENDIX B: Summary of substrate utilization by seedlings and saplings on each study pbt.Thuja pilcata (Vigour 3+)Plot Substrate Substrate Frequency Relative Density Relative SubstrateCover (%) Frequency Density Use (Rank)SiFF 0.43 7 0.26 16.26 0.22 3DW 0.41 18 0.67 43.61 0.56 1MS 0.11 2 0.07 17.54 0.23 2S2FF 0.55 2 0.05 3.62 0.01 3DW 0.42 30 0.70 71.94 0.15 2MS 0.03 11 0.25 411.22 0.84 1S3FF 0.46 14 0.21 30.34 0.03 3DW 0.52 47 0.70 91.00 0.10 2MS 0.01 6 0.09 800.00 0.87 1S4FF 0.71 3 0.38 4.26 0.10 (3)DW 0.22 4 0.50 18.54 0.46 (1.5)MS 0.06 1 0.12 17.78 0.44 (1.5)NiFF 0.46 0 0.00 0.00 0.00 (2.5)DW 0.37 0 0.00 0.00 0.00 (2.5)MS 0.05 1 1.00 20.00 1.00 (1)N2FF 0.37 0 0.00 0.00 0.00 (2.5)DW 0.46 8 1.00 17.45 1.00 (1)MS 0.13 0 0.00 0.00 0.00 (2.5)N3FF 0.56 1 0.20 1.79 0.15 (2)DW 0.40 4 0.80 9.97 0.85 (1)MS 0.03 0 0.00 0.00 0.00 (3)N4FF 0.66 0 0.00 0.00 0.00 (3)DW 0.28 5 0.83 17.70 0.44 (2)MS 0.05 1 0.17 22.10 0.56 (1)For notation see Chapter 4. Methods and Results and Table 4.2APPENDIX B: Summary of substrate utilization by seedlings and saplings on each study plot.Tsuga heterophylla SeedlingsPlot Substrate Substrate Frequency Relative Density Relative SubstrateCover (%) Frequency Density Use (Rank)SiFF 0.43 ii 0.05 25.55 0.04 3DW 0.41 219 0.92 530.59 0.84MS 0.11 9 0.04 78.95 0.12 2S2FF 0.55 2 0.01 3.62 0.01 2DW 0.42 134 0.99 321.34 0.99 1MS 0.03 0 0.00 0.00 0.00 3S3FF 0.46 72 0.24 156.01 0.09 3DW 0.52 214 0.73 414.33 023 2MS 0.01 9 0.03 1200.00 0.68S4FF 0.71 0 0.00 0.00 0.00 2.5DW 0.22 214 1.00 991.89 1.00MS 0.06 0 0.00 0.00 0.00 2.5NiFF 0.46 20 0.17 43.81 0.14 2DW 0.37 94 0.82 252.69 0.80 1MS 0.05 1 0.01 20.00 0.06 -N2FF 0.37 3 0.01 8.04 0.02 2DW 0.46 206 0.99 449.29 0.98 1MS 0.13 0 0.00 0.00 0.00 3N3FF 0.56 12 0.08 21.44 0.04 3DW 0.40 143 0.90 356.39 0.71 1MS 0.03 4 0.02 125.00 0.25 2N4FF 0.66 2 0.02 4.53 0.01 - 3DW 0.28 112 0.97 396.46 0.94 1MS 0.05 1 0.01 22.10 0.05 2For notation see Chapter 4. Methods and Results and Table 4.29293APPENDIX B: Summary of substrate utilization by seedlings and saplings on each study plot.Tsuga heterophyllaPlot SubstrateFF 0.43DW 0.41MS 0.11FF 0.55DW 0.42MS 0.03FF 0.46DW 0.52MS 0.01FF 0.71DW 0.22MS 0.06FF 0.46DW 0.37MS 0.05FF 0.37DW 0.46MS 0.1324 0.14 55.75 0.10127 0.74 307.69 0.5720 0.12 175.44 0.336 0.03 10.85 0.01165 0.88 395.68 0.3916 0.09 598.13 0.6027 0.17 58.50 0.03121 0.75 234.27 0.1213 0.08 1733.33 0.853 0.04 4.26 0.0172 0.86 333.72 0.679 0.10 160.00 0.3222 0.12 48.19 0.08153 0.83 411.29 0.649 0.05 180.00 0.2814 0.11 37.51 0.11109 0.83 237.73 0.699 0.07 67.67 0.200.16 73.25 0.070.79 515.89 0.540.05 375.00 0.3915 0.12 22.66 0.05108 0.87 382.30 0.901 0.01-22.10 0.053232323232SaplingsSubstrate Frequency Relative Density Relative SubstrateCover (%) Frequency Density Use (Rank)SiS233S4NiN2N3N43232FFDWMSFFDWMS41207120.560.400.030.660. notation see Chapter 4. Methods and Results and Table 4.2APPENDIX B: Summary of substrate utilization by seedlings and saplings on each study plot.Abies amabilis SeedlingsPlot Substrate Substrate Frequency Relative Density Relative SubstrateCover (%) Frequency Density Use (Rank)SiFF 0.43 0 0.00 0.00 0.00 (2.5)DW 0.41 3 1.00 7.27 1.00 (1)MS 0.11 0 0.00 0.00 0.00 (2.5)S2FF 0.55 0 0.00 0.00 0.00 (2.5)DW 0.42 3 1.00 7.19 1.00 (1)MS 0.03 0 0.00 0.00 0.00 (2.5)S3FF 0.46 2 0.67 4.35 0.69 (3)DW 0.52 1 0.33 1.92 0.31 (2)MS 0.01 0 0.00 0.00 0.00 (1)S4FF 0.71 73 0.73 103.55 0.45 2DW 0.22 27 0.27 125.14 0.55 1MS 0.06 0 0.00 0.00 0.00 3NiFF 0.46 85 0.62 186.20 0.57 1DW 0.37 52 0.38 139.78 0.43 2MS 0.05 0 0.00 0.00 0.00 3N2FF 0.37 22 0.51 58.94 0.56 1DW 0.46 21 0.49 45.80 0.44 2MS 0.13 0 0.00 0.00 0.00 3N3FF 0.56 68 0.56 121.48 0.33 1.3DW 0.40 50 0.41 124.61 0.33 1.3MS 0.03 4 0.03 125.00 0.33 1.3N4FF 0.66 47 0.59 71.00 0.32 2DW 0.28 30 0.38 106.19 0.48 1MS 0.05 2 0.03 44.20 0.20 3For notation see Chapter 4. Methods and Results and Table 4.29495APPENDIX B: Summary of substrate utilization by seedlings and saplings on each study plot.Abies amabilis SaplingsPlot Substrate Substrate Frequency Relative Density Relative SubstrateCover (%) Frequency Density Use (Rank)SiFF 0.43 7 0.43 16.26 0.19 2DW 0.41 10 0.41 24.23 0.29MS 0.11 5 0.11 43.86 0.52 3S2FF 0.55 0 0.00 0.00 0.00 (2.5)DW 0.42 2 1.00 4.80 1.00 (1)MS 0.03 0 0.00 0.00 0.00 (2.5)S3FF 0.46 1 0.20 2.17 0.22 (2)DW 0.52 4 0.80 7.74 0.78 (1)MS 0.01 0 0.00 0.00 0.00 (3)S4FF 0.71 188 0.59 266.67 0.23 3DW 0.22 112 0.35 519.12 0.45 1MS 0.06 21 0.07 373.33 0.32 2NiFF 0.46 4 0.31 8.76 0.28 2DW 0.37 9 0.69 24.19 0.72 1MS 0.05 0 0.00 0.00 0.00 3N2FF 0.37 52 0.47 139.32 0.43DW 0.46 47 0.43 102.51 0.32 2MS 0.13 11 0.10 82.71 0.25 3N3FF 0.56 58 0.47 103.62 0.19 3DW 0.46 56 0.45 139.56 0.25 2MS 0.03 10 0.08 312.50 0.56 1N4FF 0.66 122 0.48 184.29 0.26 2DW 0.28 132 0.51 467.26 0.67 1MS 0.05 2 0.01 44.20 0.06 3For notation see Chapter 4. Methods and Results and Table 4.296APPENDIX C: Survivorship of regeneration on different substrates on each study plot.Thuja pilcataSubstrate Site Seedlings Observed Expected Survival Z statSaplings Saplings Z(O.06)- 1.65FF Si 40 34 20 High 2.0752 57 23 29 NS -0.84S3 83 60 42 High 1.9854 7 16 4 High 3.16Ni 1 0 1 NS -0.72N2 5 5 3 NS 0.93N3 27 8 14 NS -1.33N4 6 1 3 NS -1.02DW Si 232 106 116 NS -1.19S2 279 129 140 NS -1.29S3 512 226 257 Low -3.03S4 83 27 42 Low -3.39Ni 12 10 6 NS 1.28N2 158 71 79 Low -2.14N3 61 31 31 NS 0.10N4 111 47 56 Low -2.20MS Si 63 28 32 NS -0.51S2 29 31 . 15 High 2.61S3 22 21 ii High 1.80S4 5 5 3 NS 0.94Ni 22 8 11 NS -1.11N2 4 8 2 High 1.95N3 18 14 9 NS 1.18N4 3 12 2 High 3.03For notation see Chapter 4. Methods and Results and Table 4.397APPENDIX C: Survivorship of regeneration on different substrates on each study plot.Tsuga heterophyllaSubstrate Site Seedlings Observed Expected Survival Z statSaplings Saplings z (0.05)=1.65FF Si ii 24 8 High 3.00S2 2 6 3 NS 1.1153 72 27 39 Low -1.69S4 0 3 0 High 1.75Ni 20 22 32 NS -1.47N2 3 14 2 High 3.13N3 12 41 20 High 2.92N4 3 15 3 High 2.87DW Si 219 127 157 Low -4.27S2 134 165 184 Low -4.00S3 214 121 117 NS 0.53S4 214 72 84 Low -3.59Ni 94 153 150 NS 0.36N2 206 109 130 Low -4.44N3 143 207 234 Low -3.28N4 112 108 120 Low -2.72MS Si 9 20 6 High 2.75S2 0 16 0 High 4.09S3 9 13 5 High 1.97S4 0 9 0 High 3.08Ni 1 9 2 High 2.31N2 0 9 0 High 3.05N3 4 12 7 NS 1.29N4 1 1 1 NS -0.05For notation see Chapter 4. Methods and Results and Table 4.398APPENDIX C: Survivorship of regeneration on different substrates on each study pkt.Abies amabilisSubstrate Site Seedlings Observed Expected Survival Z statSaplings Saplings z (0.05)= 1.65FF Si 0 7 0 High 2.89S2 0 0 0 (Low)S3 2 1 3 NS -1.49S4 73 188 234 Low -3.85Ni 85 4 8 NS -1.60N2 22 52 56 NS -0.58N3 68 58 69 NS -1.41N4 47 122 152 Low -2.69DW Si 3 10 22 Low -4.06S2 3 2 2 (NS)S3 1 4 2 NS 1.49S4 27 112 87 High 2.16Ni 52 9 5 NS 1.60N2 21 47 54 NS -0.91N3 50 56 51 NS 0.66N4 30 132 97 High 3.09MS 51 0 5 0 High 2.38S2 0 0 0 (NS)S3 0 0 0 (NS)S4 0 21 0 High 4.66Ni 0 0 0 (NS)N2 0 ii 0 High 3.40N3 4 10 4 NS 1.63N4 2 2 6 NS -1.55For notation see Chapter 4. Methods and Results and Table 4.3


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items