Flora, chorology, biomass and productivity of the Pinus albicaulis-Vaccinium scoparium association by Frank Forcella A thesis submitted in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE in BOTANY Montana State University © Copyright by Frank Forcella (1977) Abstract: The Pinus dlbieaulis - Vacoinium seoparium association is restricted to noncalcareous sites in the subalpine zone of the northern Rocky Mountains (USA). The flora of the association changes clinally with latitude. Stands of this association may annually produce a total (above- and belowground) of 950 grams of dry matter per square meter and may obtain biomasses of nearly 60 kg per square meter. General productivity and biomass may be accurately estimated from simple measurements of stand basal area and median shrub coverage for the tree and shrub synusiae respectively. Mean cone and seed productivities range up to 84 and 25 grams per square meter per year respectively, and these productivities are correlated with percent canopy coverage (another easily measured stand parameter). Edible food production of typical stands of this association is sufficient to support 1000 red squirrels, 20 black bears or 50 humans on a square km basis. The spatial and temporal fluctuations of Pinus albicaulis seed production suggests that strategies for seed predator avoidance may have been selected for in this taxon.  STATEMENT OF PERMISSION TO COPY In presenting this thesis in partial fulfillment of the're­ quirements for an advanced degree at Montana State University, I agree that the Library shall make it freely available for inspec­ tion. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by my major pro­ fessor, or in his absence, by the Director of Libraries. It is un­ derstood that any copying of this thesis for financial gain or for Z publication shall not be allowed without my written permission. Signature Date ,3/ /??? FLORA, CHOROLOGY, BIOMASS AND PRODUCTIVITY OF THE' . PINUS ALBICAULIS - VACCINIUM SCOPARIUM ASSOCIATION ■by FRANK FORCELLA A.thesis submitted in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE . BOTANY Approved: Chairperson, Graduate Committee î fead, Major Department Graduate wDean ' ! ' MONTANA STATE UNIVERSITY Bozeman, Montana May, 1977 iii PREFACE ■ This thesis has been divided into four segments or chapters. Though all chapters are interrelated, each has been written, as an individual entity with its own Introduction, Methods etc. Such a format has been followed for the convenience of the reader. Plant taxonomists and geographers will be interested mostly in Chap.'I; interests of forest mensurationists lie entirely in Chap. 2; sil­ viculturists and ecologists will find Chap. 3 useful, if not a- - musing; lastly. Chap. 4 is oriented toward wildlife biologists and general biologists. Within this thesis, retrieval of specific information by 'special interest, groups’ will be greatly facili­ tated by the format used. iv ACKNOWLEDGEMENTS The Department of Biology at Montana State University, under the chairmanship of Dr. J. Pickett, provided me with the follow­ ing: a financially lucrative 'teaching assistantshipr for 7 aca­ demic quarters (with options for two more); computer time; trav­ elling expenses for field research, • field trips and for the pre­ sentation of research papers at annual scientific gatherings; and lastly, general hospitality. The research presented in this paper was initiated through a $5000 grant from the U.S. Forest Service to Dr. T.W. Weaver. In turn. Dr. Weaver gave me nearly full, unhindered responsibility for the appropriate use of these funds. Beginning with my parents, there are several individuals who have greatly affected my personal development. Those who have contributed academically (and in other ways).include: Arthur Johnson, Fred A. Barkley, J.H. Rumely, H.N. Metcalf, S.J. Harvey and J. Major. I thank all these individuals. Of■course'in- this same regard, T.W. Weaver deserves special mention. If I have any (good) scientific/analytic qualities, a very large proportion of them can be attributed to my association with Dr. Weaver. Though ones' initial boyhood enthusiasum in the natural sciences may often have been spontaneous, its maintainence into adulthood is "O Lly f ■ ' ' ....( -'.-u ■ 1 ' .. V not neccessarily self-perpetuating; Dr. Weaver has the uncanny ability to sustain, in another individual, an insatiable enthus iasum — through his very own, perhaps. CONTENTS Title Page i Vita ii Preface iii Acknowledgements iv Contents vii,- List of Tables ix- List of Figures x. Abstract xi._ Chapter I. The Flora and Chorology of the V-inus atb-taccut-is, - Vacc-Lnium scopariwn association. I Introduction 2 Methods 3 Results 4 Lichens. 11 Disjunctions 12 Species Number 14 Management Implications 14 Chorology ' 15 Conclusions 18 Chapter 2. Biomass and Productivity of the Pinus albieautis - Vaeeinium seoparium Association. vi 20 Introduction • 21 The Pinus albicaulis - Vaccinium seoparium Association. ■- 22 Methods 24 Vegetation ' - .24 Biomass & Productivity • 25 . • Regressions 28 Results and Discussion 30 Vegetation ' 30 Regressions 31 Tree Mass 34 Understory Mass 36 Total Mass ■ 38 Tree Production 39 ,' Understory Production 41 Total Production 41 Stand Age vs. Mass & Production - 43 Conclusions 46 Chapter 3. Cone, and Seed Crops of the Pinus dtbicaulis - Vaeeinium seoparium Association. 50 . Introduction 51 Preliminary Observations 52 vii Methods . 53 viii Results 56 Vegetation 56 Cone Phenology 57 Predicting Cone Crops ,from Stand Characteristics 57 ' Biomass and Seed Number - 63 Comparisons with Other Forests 64 Localizing Cone Crop Variability 66 Conclusion and Discussion . 6 8 Chapter 4. Food Productivity in a Rocky Mountain Subalpine Forest Association. 72 Introduction 73, Methods ' ■ 74 Results and Discussion 76 Literature Cited (composite) 85 Appendix I. Estimating Bark Production. 96 Appendix 2. Stand Locations. 98 Appendix 3. Annual Cone Crops. 99 ix LIST OF TABLES la. Flora of the Pinus albioaulis -Vacainium scopavium association. 6 lb. Flora of the Pinus albioaulis - Vaaainium ■ seopapium association. S la. Flora of the Pinus albioaulis - Vaocinium scopariwn association. 9 2. Biomass and production statistics for Pinus albioaulis. 32 3. Biomass and production statistics for Pinus albioaulis. 33 4. Comparison of cone and seed productivities- in several Pinus forest types. 65 \ X LIST OF FIGURES , 1. Distribution of the Pinus atbioauli-s - VaooirLiwn ■ sco-pccrium association. 5 2. Biomass and production of Pinus aZbiocaiZis' in. re- • . Iation to stand basal area. ' 35 3. Biomass and production of Vaooiniwn sooparium in relation to median shrub coverage. 37 4. Biomass.and production of Pinus atbioautis stands in relation to stand age. 44 5. Tree density in Pinus dlbicaulis stands in rela-Li tion to stand age. 45 6. Inter- and intravegetational comparison of bio­ mass accumulation ratios. 48 7. Mean annual cone and seed production in Finns atbioautis stands as a function of stand canopy coverage. 61 8. Plant food energy available in Pinus atbioautis stands as a function of the canopy cover­ age of the producing plants. 77 ABSTRACT .. FLORA, CHOROLOGY, BIOMASS AND PRODUCTIVITY OF THE PINUS ALBICAULIS - VACCINIUM SCOPARIUM- ■' ASSOCIATION A ' ' The Pinus albiaaulis -.Vacoinium scopanium association is restricted to noncalcareous sites in the' subalpine zone of the northern Rocky Mountains (USA) . The flora of the association chang­ es clihally with latitude. Stands of this association may annually produce a total (above- and belowground) of 950 grams of dry mat­ ter per square meter and may obtain biomasses of nearly 60 kg per ■square meter. General productivity and biomass may be accurately estimated from simple measurements of stand basal area and median shrub coverage for the tree and shrub synusiae respectively. Mean cone and seed productivities range up.-to 84 and 25 grams per square meter, per year respectively, and these productivities are correl­ ated with percent canopy coverage (another easily measured stand parameter). Edible food production of typical stands of this as­ sociation is sufficient to support 1000 red squirrels, 20 black . bears or 50 humans on a square km basis. The spatial and temporal fluctuations of Pinus dlbicautis seed production suggests that strategies for seed predator avoidance may have been selected for in this taxon. ' \ CHAPTER I THE FLORA AHD CHOROLOGY OF THE PIBUS ALBICAULIS VACCINIUM SCOPARIUM ASSOCIATION 2 . INTRODUCTION Floristic variation within a plant association may indicate that the habitat of the association is not uniform, throughout, . and that two or. more plant.associations are being considered as one, the extreme case being that each community is an individual, ■ an association unto itself. Other interpretations- exist; summaries'- can be found in Major and Pyott (1965) and in several texts dealing with vegetation. In this chapter I shall describe the compositional variation of. the ,Pvnus aZb'loccut'ts - Vaecvnvian seo- ’parvum association, and relate some of the variation to one factor of vegetation formation (Major 1951), i.e., the flora from which the vegetation may have originated. ' ' The suggestion that a single plant association varies according to the flora available to it suggests that this association may ' exist in more than one floristic region. If regional climate and events during historical time determine floristic regions, one might conclude that these factors could act differentially within the association and thereby affect its variation. ' Alternatively, . if one assumes that a recurring mixture of plant species indicates a particular set of environmental conditions, and that the prob­ ability of two or more of these species concurrently evolving the same degree of ecotypic variation is low, then it follows that the . • . ■ ■ . 3 habitat within which this association of plant species exists is, ' more or less equivalent throughout (if it is integrated over ecologic time) . Thus floristic differences of communities with "identical" habitats must be a result of either the availability of their flora at the time of their establishment (Egler 1953) and/or through the remainder of their existence (Major 1951). METHODS Stands in Wyoming, Idaho and Montana (U.S.A.) with over­ stories dominated by Pinus albicaulis, understories dominated by Vaoainiim soopavium and lacking conspicuous populations of Abies iasiooavpa seedlings and/or layered shoots (i.e., Abies repro­ duction less than that of P= .albicaulis), and soils not stony or rocky (Soil Survey Staff 1976) enough to obviously affect the 2growth and distribution of plants, were sampled. Within a 600 m area (three, 6.67 x 30 m) in each of 29 stands, all vascular. plant species were collected (identified and filed at the Herbarium, Montana State University; Bozeman, Montana); and 2within ninety 2 x 5 dm frames in each 600 m area, the coverage of each vascular plant taxon was estimated. Foliose arboreal lichens were also collected, but not systematically. Nomenclature follows that of Hitchcock and Crpnquist (1973) for the Pacific 4 Northwest vascular plants, Munz and Keck (1968) for other vascular plants', and Hale (1969) for lichens. Taxonomic authorities not in the text are listed in Table I. 'RESULTS Four .species other than Pinus dthicaulis and VaocinivM sco- -parium were nearly ubiquitous in the sampled stands: the wide­ spread Carex vossii, Abies Zasiocarpa and Poa nervosa with.con­ stancies of 80, 90 and 70% respectively, and Amioa ZatifoZia (80% constancy, though absent from most Wyoming, stands) . ' The presence of these taxa lends some support (or degrees of freedom in a statistical sense) to the initial assumption of the imp rob- .. ability of two or more species concurrently evolving associated, ecotypes. In Table I, the flora and some other stand characteristics are provided in releve-' form. This table lists the stands in a latitudinal sequence, with adjustments to accommodate latitudi- nally-similar stands with widely separated longitudinal ordinates ‘ (cf. Fig. I). Stands have not been sorted according to their floristic similarities as is usually done in releve* analyses .(Mueller-Dombois and Ellenberg 1974). However, the taxa have been arranged to give the maximum impression of latitudinal change to 5 116° 114° 112° 110° 108° MONTANA 4 • •. WYOMING IDAH O 49° 4 7° 4 5° 4 3° 41° Figure I. The distribution of the Pinus albicautis - Vaooinivjn sooparium association. The numbers and their associated charac­ ters (e.g. n-, -n-) represent stand numbers and the three geo­ graphic/ floristic regions referred to in the text. The geograph­ ical extent of the stands is thought to represent the range of the association. The unhatched area inside the dotted line rep­ resents the gap in the distribution of Pinus ponderosa (from Little 1971). (There is no Stand 21.) Table la. Flora of the Pinus albicaulis - Vaaoinium scoparium association. Stand age (years),'% arboreal canopy coverage, % V. scoparium coverage are shown along with the taxa of the association with constancies greater than 15% (and which show limited distri­ butions in the association). The stands have been arranged in a latitudinal sequence from south to north (cf. Fig. I.). Assume varietal status of each taxon to be common Rocky Mt. variety or equivalent to specific epithet unless stated otherwise. Taxa with limited dis­ tributions in the association due to their actual narrow regional distributions are de­ noted by asterisks (*). Simple presence in the 600 m sampling area is represented by a '+', whereas '1-5' represent frequency classes from low to high. S 5 2 R 0 £ 0 s , 0 S & (V vR C"! v> S S £ 0 % R £ % & pStand Age (N r-t (V Percent Arboreal Coveraqe 63 63 71 63 49 54 52 73 97 31 30 69 47 56 53 40 73 51 61 54 37 70 66 — 32 22 13 18 50 Percent Vaccinium coverag> 7 21 16 5 25 59 14 10 9 27 29 46 54 55 43 64 75 75 74 59 36 37 16 23 23 30 32 Stand Number 16 17 18 19 20 09 IO Ol 02 28 27 29 13 14 15 U 12 06 07 08 04 05 03 30 26 25 24 23 22 Sedum lanceclatuin Torr. I I + . A chillea m illefo lium L. . I + I + Claytonia laneeolata Pursh . I + ' Solidago m ultivadia ta Ait, . + I + • + I I I I I C aB tillja r e x i fo lia Rydb. • • . + • I I + + AgoaeHs glouea (Pursh) Raf. ' + . I • + I I • + I . . PinuB eontorta la t i fo l ia Engelm. • • + + + + * As te r fo liaceus apHoua Gray . • + • , I I THaetum spieatum (L.) Richter . . I + , I I I • 2 • Lupinue argenteue Pursh I + T I ' + • I + ' 2 ' ' ' . I + ' I ' I * I P o ten tilla d iv e r s i fo lia Lehm, + I • * I + • ' t « I • # + 1 I + » * • » ' • ' • Am iaa cordifo l ia Hook, . 2 2 3 + I 2 I 2 + . 1 . 1 I + , A, la t i fo l ia Bong • • • I 3 I 2 3 ♦ 2 2 . 1 I I 2 2 2 2 3 * I I 2 + I ♦ + Prythrcniun graniKflonun Pursh 2 • 2 I • I I . + • • + + • Carex gcycH Boott I I + + ' . + 3 3 . . 2 2 3 I I 2 • • + ALomatium ousiakii (Wats.) Coult . & Rose I I Pediaularna groenlanclica Retz. Juncus parryi Engelm. AntonnaHa la m ta (Hook.) Greene aTmzuIo h itchcookit Hamet-Ahti I . 1 I . I 2 2 7 demonstrate that the flora of the association changes clinally with latitude, and that the flora of any stand is at least a partial consequence, of the floristic region in. which the stand . exists. Table la consists of those taxa. with constancies > 15%, and whose presence appears to be nonrandomly distributed within the association. When the within-association distributions of these taxa are compared to their general distributions listed in stan­ dard Floras for the Pacific Northwest (Hitchcock and Cronquist 1973, Davis'1952, Booth and Wright 1966, Shaw 1976 and Despain 1975), approximately 15% of them are distributionally restricted from attaining 100% constancy in the P. dlbie.aulrls - T7. scopan-ium association. Similarly, of those taxa.with between 5 and 15%. con- - stancies, and those with < 5% constancy; (Table lb, lc), about 1/2 of the former and 1/3 of the latter are distributionally restricted from ubiquity in. the association. -Taxai which exhibit no latitu­ dinal affinities (Table lc) are characteristically widespread in their general distributions. A few of the taxa in Table I deserve special mention. - The low glandular shrub, Leptodaety Ion' pungens, and the- similar but more ■■ cushion-like Avenaria aeuteata both have stiff spinulose leaves often found in desert-region plants, as indeed both are.. In P. Leweeia pygmea (Gray) Robbins. Aquilegia flavescene Wats. Viola nutallii Pursh Potentilla gvacilie brurmceccne (Rydb.) Hitchc Lomatiur, dieeeetun (Nutt.) Math. 6 Const. Antennaria racemooa Hook. Ôemorhiza purpurea (Coult. s Rose) Suksd. *Senecio craesulue Gray Valeriana dioica L. Sitanion hystrix (Nutt.) Smith tChinophila tweedyi (Canby S Rose) Hcnd. tArcnaria aculeata Wats. Sibbaldia proaumbene L. Heuchera cylindrica Dougl. Spiraea hetulifolia Pall. tXerophyllum tenax (Pursh) Nutt. tArabis hirsuta (L.) Sco?. Antennaria microphylla Rydb. tLarix luallii Pari. tPenstemon frutiaosus cerratus (Keck) Cronq. Phyllodoce empetriformis (Sw.) D. Don. 16 I ? 18 19 20 09 10 01 02 28 27 29 13 14 15 11 12 06 07 08 04 05 03 30 26 25 24 23 22 + I . . + « I + . + + + + + . . + I I + ♦ + . + + + + I + I + + + + + I + + + I I I - I + I + + ♦ * I • I + I I , + + , - + + . + Table lb. Flora of the Pinus albiaaulis - Vaoainium sooparium association; the taxa of the association with constancies be­ tween 5 and 15% (and which show limited distributions within the association). Table specifications follow those in Table la. 1 6 17 18 19 20 09 10 01 02 28 27 29 13 14 15 11 12 0 6 07 08 04 05 03 30 26 25 24 23 22 Abie8 laaiocarpa (Hook.) Nutt. I + I + I I T I + I I I + + I + . I I + I I I I + Carex voasii Boott + I I + I I a I I . I I I . + I I I + + . + I I + T 2 Voa nervosa (Hook.) Vasey I 2 2 I 2 I I I I I T , I . + + + + I + + + Kieracium araeile Hook. + + + I I 2 2 , + I I 2 + I I + I I Fpilobium angustifolium L. + 2 I + I +. I + + + . + I + I + Juniperus eommnia man tana hit. + + I + + + + + I + + I + Antennaria umbrinella Rydb. + I + I + I + + + + Pieea engelmannii Parry I + + + + + + + Pyrola eecunda obtueata Turcz. + + + Penstemon rydbergii A. Nels. 2 2 . Oryzopois exi-gua Thurb. + I Arenarna eongesta aephaloidea (Rydb.) Maguire . I + . . . . + , + I + + + . Erigeron peregrinus (Pursh) Greene . + + I + . Polygonum bistortoides Pursh . + Fragaria virginiana platypetala (Rydb.) Hall Ribes laaustre (Pers.) Poir. Thirty-two taxa have only single occurrences in the stands (constancy < SM. In the latitudinal sequence shown above, the stands with their respective taxa ore: 17 Krigcron pimilue Nutt.i 19 tMgrtensio Oiliata (Torr.) G. Don., Centiana amarclla L . , 20 Armaria obtusiloha ( I tydb.) F e m . , Selaginella dense eoopuloner. (!'axon) Tryon, Saxifrage hronahialic L.; 10 P e n a tc m n fr u t ic o r u a fruticoeus (Purs):) Greene; 2 tTrifloiicr. haydenii Porter; 28 Stipa occidr.ntaiis Thurb., tHapplopappua auffruticoaue (Nutt.) Gray; 27 Silene repene Pers.; 15 Coodyera oblongifolia Raf.; 7 Poa alpina L.r 5 4Facctn-LiCT; membranaceien Dougl.; 3 Valeriana edulie Nutt., Lloydia serotina (L.) Sweet; 30 Sypopitys rr.onctropa Crantz; 26 M s t e r stenoxeree Gray, Brorrus carinatus H . s A.; 25 tLeptodactylon pungens (Torr.) Nutt.; 24 Buncx paucifclius Nutt., Salix phylioifolia L., tVeroniaa cusiakii Gray, tLinanthastrun nuttallii (Gray) Ewan., tPenstemon flavcscenc Pennell, Hieraciun albtflorun Hook., Danthonia intermedia Vasey, Hyperioun fomosun nortoniae (Jones) Hitchc., Agooeria auriantiaca (Hook.) Greene, tSpiraea denoifolia Pall., Anaphalie margaritaoea (L . ) a , C H.; 23 Carex payeonie Clokey, P e a eandbergii Vasey, Feetuca ovina brevifolia (R, Br.) Wats.; 22 Campanula parryi Gray. Table lc. Flora of the Pinus albiccculis - Vaccinium seopariwn association; the taxa of the association exhibiting no latitudi­ nal affinities. Table specifications follow those in Table la. 10 ■ aVb'ioaut'is forests- these species are found in the southern Bitter­ root Mountains and the Salmon River Mountains of east and central Idaho. The dry finger-like intermountain valleys (Lemhi, Pahsimsroi and Lost River Valleys) which extend from the northern edge-of the Great Basin and abut these mountain ranges probably supplied the migratory path for - these species from- the deserts to P. O-Zbiaaul-Ls ' forests. To find either taxon in a mesic subalpine forest is surprising, but would have been much more so if that forest had been in central Montana, :, rather than east-central Idaho with, its direct connection to the.Great Basin. Three species of the P. albioauVLs- - 7. scopapium association are relatively narrow endemics: PenstemorTi flaveseens (Idaho Co., Idaho and Ravalli Co., Montana), CkionopfoLla tweedyi (central Idaho and adjacent Montana) and Aster stenomeres (central Idaho and adjacent Montana to northeastern Washington and southeastern British Columbia). Similarly, the typically alpine Trifoliim haydenii extends only as far north as southern Montana. It occurs in a xvrhitebark pine stand immediately adjacent to alpine meadows • and scree in the Madison Range (Gallatin Co.,. Montana) and would not be expected in P. albicaulis forests further north. Though the general regional occurrences of the above taxa are easily obtained from standard Floras, their equally important 11 intra-regiorial distributions :are not so readily available. For example. Arnica Zdtifotia and Carex■geyeri, which are prominent . in many Montana whitebark pine stands, are absent from the stands in the Wind River Mountains of Wyoming. This mountain range does support both taxa, but their populations are not as extensive as1 elsewhere.' In such cases, the chance of limited taxa reaching. P. atbicautis forests is . low. King (1977)" has noted this same phenomenon, but on a much smaller scale; the.ability of a plant to colonize ant mounds in British pastures is determined by its relative abundance and distance from the mounds. Superficially at least, in some regions of P. atbicautis forests there appears to be an "ecological" replacement of one taxon (life form) by another. Arnica cordifotia is generally prominent in those stands in which A. Zatifotia is not,, and Luzuta hitchcockii is relatively important in the Bitterroot Mountains where Carex geyeri is not. Lichens ' Both Letharia-vuVpina (L.) Eue and Hypogymnia vittata (Ach.) Gas. were widespread throughout the association; the former being much more prominent. Atectbria oregana Tuck, and TL amer- acana Mot. were confined to the northern-most stands. Both 12 . ' - Alectorias have limited distributions in the Rocky. Mountains which correlate with their presence in the whitebark pine forests'. Disjunctions Daubenmire (1975) has applied•the term "oceanic element" to taxa with distributions largely restricted to mar,itime-influenced climates of the Pacific Northwest (NW Montana ̂N Idaho, W Oregony W and NE' Washington and the adjacent,parts of Alberta and British Columbia) . I had considered both Xevo’QhyVhmi tenax and Luzula ' hitdhaochil to be strict oceanic elements, but. their actual dis­ tributions are in fact more extensive. Widely disjunct populations . of both species occur as far south as Teton ..Co. > ' Wyoming. (Shaw 1976, Pfister et al. 1974, Maule 1959). MenzieslafeTrugineo, Smith and Pinus montieola Dougl. (not in whitebark. pine forests) are other oceanic elements often found close to or associated with X. Ienax and L. hiteheockii. They also have disjunct distributions nearly identical to the-others (Hickman and Johnson 1969 and personal observations). Perhaps in past times, the paleoclimate was sufficiently different to support an. "oceanic" vegetation ' throughout the northern Rocky Mountains, as presently exists in NW Montana and N Idaho. Additional evidence for such, a maritime paleoclimate is the discovery of Taxus brevifolia Nutt, (an unquestionably oceanic species) wood remnants during archaeological. 13 excavations in the Yellowstone Valley, SW Montana (Arthur 1966■ the same valley presently supports, very localized populations of X. tenacc and M. fevTuginea). Radiocarbon dates, for the. Tascus' material were 5000 years. BP. . The early Holocene, epoch in the Rocky Mountains is thought to have been, cool and wet (Hansen- 1947); the Xero- or Altithermal interval began, about 7500 BP and lasted until the onset of Neoglaciation, ca. 4000 BP (Richmond 1970). Wells (1970) has suggested that the "Xerothermal" inter- ' val in the Laramie Basin of Wyoming was, wetter, not dryer, than present. Unless these plant disjunctions and excavations repre-. sent relict vegetation from the pre-Pinedale Glacial times, with the recession of Cordilleran ice (12,000 BP; Richmond 1970), a. ■ Pacific maritime climate and vegetation maij' have pervaded the entire northern Rocky Mountains. A subsequent cooling and drying trend in W Wyoming and SW Montana could not support a maritime vegetation, and extinctions and disjunctions resulted. High elevation bog pollen-profiles. in Yellowstone National Park .(Waddington and Wright 1974) are dominated by TP-inus oontovta from ca. 11,600 BP to present; an increase of Fieea engetmannii pollen at 5000 BP implies climatic cooling. That the W " Wyoming - SW Montana area is still subjected to a relatively cold climate - can be seen by the present gap in the distribution of Finns 14 ■pondevosa Laws. (Fig. I), a typically "warm" pine (Miroy 1967) . Curiously, the absence of P. ponderosa from this area correlates generally with the occurrence of the P. aVb-icauUs - F. scopaviim association. If those whitebark.pine stands with oceanic elements are omitted, then the correlation is nearly perfect. Species Number The number of species in the whltebark pine stands ranged from 6 in the oldest stand (640 years) to 33 in one of the youngest (33 years). There was a general trend in decreasing species number with stand age, but stands that were proximal tended to have similar species numbers despite age differences. Management Implications . Although whltebark pine forests receive relatively little resource management attention at the present time, it can be expected to increase rapidly. P. albioautds produce excep­ tionally large mast crops (Chapters 3, 4), and'such production may significantly affect the habits of .wildlife.. (Craighead 1976, Chapters 3, 4). Total net primary productivity in these forests 2 2 may exceed 900 g/m /yr, and standing crops may approach 60 kg/m (Chapter 2); economically, these figures are substantial. There are also some practical aspects involved with the ■ floristic distributional anomalies of whltebark pine forests. 15 The three dominant herbaceous species of the association are: Cavex geyeri, Avn-Cea Iat-CfolyCa and A. eovd-Cfol-Ca. These, taxa all have known forage value for both domestic and wild ungulates. The biomass and energy (kcal) per unit-area of each species can be readily predicted from their canopy coverages (measured sep­ arately; Chapter 4)= Further, as can be seen in Table I, the species are distributiohally limited within the association. If the 29 stands are split into three geographic/floristic regions (Fig. I; separations based on plant distributions and agglomera- 2 tive cluster analysis), the mean energy value per m for each species differs significantly between at least two regions; (t-test, p = 0.01; Chapter 4). In vegetation mapping, the P. aTb-Ccaulis - V. seoparium association, as a whole, would prob­ ably comprise a single cartographic unit. Knowledge of regional differences in forage availability within associations might prove valuable to resource managers. . . ■ Chorology The stands shown in Fig, I essentially outline the distri­ bution of the P. alb-Ccaulis - V. SeopavCum association. To the, north and northwest, Abies lasioeavpa, Lavix lyofliii and Yaceinium, membvanaeeum gain importance in whitebark pine forests. In Alberta, Canada (on acidic substrates), P. albieaulis occurs with 16 equal amounts of Picea engelmccnni-i and A. Tasi-occâ pa' in the' over- story-. Understory components always contain F» Saopariwni hut - it may be accompanied or dominated by ̂Vaoeiniian eaespitosvm. Michx., Ewpetnan nignan L., Dryas octopetaia L., Saiix aratica Pall., or Spiraea sp. In Banff National Park, I found one stand on dolostone totally dominated by P„ ' aibieav.Zis \ its under story, in. order of importance, consisted of BetuZa gZanduiosa. M-IsSxs.. > TotentiZZa frutiaosa L ., Linnaea borealis’L ., Shepherdia canadensis (L.) Nutt. Juniperus communis and Dryas octopetaZa. There were no Vacciniums in this stand, probably due to.its basic substrate. The eastern limit of the P. aZbicauZis V. scoparium associa­ tion is correlated with the eastern extent of acid-rock mountain ranges.in Alberta and Montana. Limestone ranges such as the Big Snowy Mountains (Montana) do not contain this association. The . eastern.limit in Wyoming is the Absaroka and Wind River Mountains; the granitic Big Horn Mountains, 170 km eastward, do have scattered populations of T. aZbicauZis (Hoffman 1976, D. Despain per. comm.'). To the south, the Medicine Bow Mountains (Wyoming), the■ Colorado Rockies and the Uirita Mountains of Utah all lack whitebark pine. That the southern limit of P. aZbicauZis coincides with the northern boundary of other edible large-seeded, grove-forming pines (P. eduZis Engelm., S Wyoming; P. monophyZZa: Torr. & Frem., • 17 S Idaho to California) may be more than coincidental. Forcella and Rumely (in prep.) hypothesize that prehistoric man carried seed of P. si-bi-vica L. (= P. aW'Loaut'ts) across Beringia. His dispersal of the energy-rich seed ceased when contact was made with native large-seeded pines. In far western Wyoming (the Wyoming Range) , P. . atbi-oauHs forests contain an understory of Ribes montigenum McClatchie (which forms conspicuous closed circles under the canopies of the rather widely spaced trees) and Bromus Oarinatusr To the northwest, in.the White Cloud Peaks and Sawtooth Mountains of central Idaho, P. dlbicaulis stands often support an understory of Artemisia tridentata Nutt, and/or a carpet of forks, Luginus ccrgenteus being the most prominent. Within the distributional limits of the P. albioaulis - F. soogarium association, there may be other associations which contain P. dlbicaulis. On limestone outcrops. Weaver and Dale (1974) mention a stand in which P. flexilis James and various forks associate with whitebark pine. I have seen such stands and others similar, but always including ArctostaphyZos uva-ursi (L.) Sprang. This type of community, with a distinctly different habitat (limestone), appears to have lumped with the P. dlbicaulis V. scoparivm association in the "habitat-type" classification of 18 • Pfister et al. (1974) and Reed (1976). Also, on what may he more mesic sites, Abies iasioocô pa shares the overstory with whitehark pine, and V. membvanaceum is often present in ,the understory.. It is possible that alternate plant associations (.Abies vs. Finns'} may exist on the same site at different times, the occurrence of either possibly being a function of its seed crop size at the time of stand establishment. Seed production of P. dlbicautis varies significantly from year to year (Chapter 3). Treeline form(s) of whitebark pine community occurs too;, its distinguishing feature is,oof course, the stunted growth and flagged structure of the trees (Daubenmire and Daubenmire 1968). Clausen (1965) speculates a genetic basis for the stunted P. alhicauVis of the Sierra Nevadan krummholz. CONCLUSIONS . The Finns albieaulis - Vaccininm seoparinm association is . limited to subalpine sites on non-calcareous substrates in western Wyoming, southwestern Montana and east-central'Idaho. Its floristic composition changes clinally with latitude; this does not necessarily imply a change in habitat. Nearly 25% ■ of the taxa which comprise the association are distributionally restricted from, occurring in all stands of the association. 19 This suggests that to some degree, the floristic composition of a stand is a function of the local flora available to it ̂ CHAPTER 2 THE BIOMASS AND PRODUCTIVITY OF THE PJMXS ALBICAULIS mVACCINIUM SCOPARIUM ASSOCIATION 21 INTRODUCTION ' Natural resource management has been and remains a significant problem in the western United States of.America. In at least some geographic areas, vegetation classification is a constructive pre­ cursor to effective management. ’ The values of a classification lie, ' in part, in their applicability to estimation of "site potential”, i.e., the potential taxa, productivity and stability of specific communities within the classificatory scheme. A significant por­ tion of the Pacific Northwest's climax vegetation has recently been classified. Pfister (1976) lists all completed and on-going classifi­ cations [excepting Hickman (1976)] in the western U.S.A. Now that these classifications are more or less complete, and with a management perspective in mind, what are the current and near future research priorities? In conjunction with continual reassess­ ments of the classifications,. I believe that they might include studies of (I) vegetation mapping, (2) vegetation change, (3) biomass and productivity and (4) nutrient cycling. That resource managers • need immediate, coherent data, or at least "best guesses" from the research groups on these subjects has been clearly stated in a recent Institute of Ecology Report [1974, 4(3):3]. Relatively little research has been published on the biomass and productivity of Rocky Mountain forests (Moir 1972, Johnstone 1971, 22 Whittaker and Niering 1975, Landis and Mogren 1975, and Hanley 1976). This paper deals with the above- and belowground biomass .and produc­ tivity of the Pvnus atbicaut-is Engelm. - Vaaei-U-Lim - Scopcai-Ium. Leiberg association. Since I studied 14 stands (stands. 1-14, Fig. I), I was able to determine the range of values expected "in typical stands, and relate some of the variance to stand basal area and age.. The mass and production of each stand was "not, however, studied in depth. Samples of stand components were taken from each stand and regression equations developed for prediction of mass and productivity, (ulti-. mately on a unit area basis). The validity of this approach is a function of the initial regression equation statistics. Though error is, of course, present in the equations, I feel that accurate predictions of biomass and reasonable estimates of productivity are possible. I hope that the results reported here will contribute to a basis for the management of these forests., The Pinus aZbieautis - Vaeeinizm scopariwn Association- An association, as I have used the word here, is the sum of easily recognized, floristically and structurally repetitive stands of vegetation. The term "habitat type" has been adopted as a. classificatory unit for land by several researchers in the Pacific Northwest. Though the inclusion of habitat into the definition of association (Flahault and Schroter 1910, in Braun-Blanquet 1964) 23 . " . . . is appealing, I have avoided its use in this report because the abiotic factors of an organism's environment are not easily docu­ mented by field observations - except circularly through floristics and physiognomy. The P. CLlbi-CauZis - V. scopaccium association is an. easily, recognized, geographically repeatable, subalpine forest of non- ■ calcareous substrates in the northern Rocky Mountains, ylts vege­ tation, soil, soil parent material and climate have been charac- ■ terized by Weaver and.Dale (1974) . The stands of this association are normally even-aged and may attain relatively great ages (600. years +). The site index (tree height from time of establishment) . ■is about 8 meters at 100 years of age... The understory is typically dominated by the low-growing' (10-30 cm) V. scoparium. Vegetational change appears to be cataclysmic rather than "successional" (cf. Loucks 1970). Fire scarred stumps and/or soil charcoal are always found inLthe stands. Avalanches and wind—throwing may also affect . vegetation cycling in this association. - . Tree seedling establishment under the arboreal canopy is meager Occasionally, the few Abies Vasiocavpa Hook, individuals in the .understory outnumber those of P,-albicaulis. These small Abies often result from the layering of an older tree. . Both Daubenmire ... . and Daubenmire (1968) and Pfister et ,al. (1974) recognize an A. 24 Zasiocarpa (P. aVoieauZis) - 7. seopavivm Habitat type. The dominance of P. ■atbioauZis in my stands might.be due to (I) a very slow (600 years +) successional replacement of P. aZbicauZis by A. Zasiooâ pa, (2) physical (e.g., climatic) limitation to Abies- on P. .aZbioauZis sites, or (3) biological limitations to Abies 'due to the ubiquity ■ (100% constancy) of the. Abies TPueciniastrnm ■ geoppertianum. ■(Boyce 1961, Faull 1939) in the P. dtbieauZis - 7. scoparium associa- - tion. Though P. atbieautis is not generally recognized for its mer­ chantable qualities (until recently, see Day 1967, Kaspar and Szabo 1970, Keenam et al. 1970), the P. atbieautis - 7. scoparium association is valuable for its aesthetic qualities, for wildlife food (large edible pine seeds, Vaeeinium fruit and assorted, forages; see Weckwerth 1971, Mealey 1975) and for watershed protection. SAMPLING METHODS ' Vegetation ■ Fourteen stands of the P. atbieautis - 7. scoparium association were sampled in 7 mountain ranges throughout southwestern Montana (the Madison, Tobacco Root, Elkhom, Big Belt, Little Belt, Castle and Absaroka mountain ranges). Each stand was sampled with three 6.67 x 30 meter (0.06 ha) plots which were placed parallel to the 25 slope of the stand. All trees within these 200 m" plots were tallied by species into 5 cm dbh size-classes (trees less than 1.35 m height were considered seedlings). Shrub and herb coverages were estimated. 2in 10% classes in sixty I m quadrats which were placed in two contiguous I x 30 m transects along.the axis .of. each plot. These transects were large enough to encompass the contagious distribution of the shrubs (Forcella 1975). The canopy coverage of the tree • layer was measured with a vertical periscope and was considered to ' be,equal to the percentage of ,30 points (at I m intervals) covered along the axis of each plot. The vascular plant species encountered within each plot were recorded, collected and filed at the Herbarium, ; Montana State University. - - Biomass and Productivity The aboveground portions’ of 34 P. CLVbiecmZ-Ls trees, 1-37 cm dbh, ■ 1.35-18.0 m height and representative.of the stands in which they grew, were felled, sectioned into component parts, arid weighed in the field with hand-held scales. The roots wider than I cm diameter of 9 trees .were excavated and weighed also. Sections of each com­ ponent part of the tree were wrapped in polyethylene and transported to the laboratory for annual ring width, ring number and wet weight/ dry weight analyses. Samples were oven dried at 60°C to constant weight. Wood (xylem) production of 25 of the trees was estimated by calculation of the mean annual parabolic volume increment and conver ' 3 ■sion to mass increment by the density factor of 0.42 g/cm (Beattie 1953, Keenam et al. 1970). Radial increments used, in calculating volume increment were the annual (1969-1973) means measured at 8 equidistant circumferential points, perpendicular to the axis of growth on dbh cross-sections. Radial increments of bole cross- sections above breast height were not analyzed consistently; where, they were measured, they compared favorably with those at 1.35 m. A sample of leafy twigs from each of 16 trees was divided into annual increments for determination of leaf production and.longevity Phloem and bark production were not measured, directly. The combined production of these two tissues was. estimated through a series of assumptions described in Appendix I. Briefly, there appears to be a linear relationship between bark mass and tree basal area (BA). . Hence changes in BA (annual increment) will have corresponding ' increments in bark mass (annual production). • 2At the end of the 1974 growing season, twenty-nine 0.5 m . quadrats of the understory vegetation were clipped from, several of the stands to estimate understory biomass and production. Twig - and leaf production of F. SaopcwyIwn were estimated by separation of the leaves and current twigs from the main shoots; these plant parts were then dried at 60°C and weighed. Diameter increments Tl of the numerous small (less than 2 mm diameter) woody shoots of this shrub are small; and because of the shrub's- low stature (10-30 cm), the mass increment of these stems is probably minute (cf. Andreyashkina and Gorchakovskii 1972) and was not. measured, but it was estimated from the work of Whittaker and Wpodwell (1969) as described later. The dead twig mass of W.. Soopca3-Ltm was measured separately. Herbaceous production was considered equal to its standing crop,though some ephemerals.were senescent at the sampling . ■ • \ • ' ' •date. For a simple,estimate of biomass of roots and rhizomes less 2than I cm diameter, eleven I m , quadrats from 5 stands were excavated to 0-10 and 10-40 cm depths. These soil depth-intervals were chosen because the rhizomes of V. saopavium are confined entirely to the upper 10 cm of soil, and root material was not readily apparent below 40 cm. All roots and/or rhizomes greater than I mm diameter •were sieved from the soil and separated by species. To estimate the mass of fine roots (<1 mm) the sieved soil was evenly redistributed, throughout the excavation and 1/16 of it (to both 10 and 40 cm depths) was transported to the laboratory for further sieving and. . washing; these roots were not separated by species. 28 Regressions, Estimates and Stand Totals Untransformed and logarithmic (base 10) transformations of the : biomass and production data were regressed against the easily measured plant parameters, dbh and shrub coverage. The estimates, obtained from the resulting regression equations were substituted for original stand V data (tree density by size class, or shrub coverage) to determine mass, and production on a unit area basis. For tree species other than P. aZb-icautiSy and for the few individuals of P. aZbZco,uZZs .greater- than 37 cm dbh. in the sample plots, the biomass regression equations of Weaver and Forcella (1977) were used. The productivity-dbh equations were extended somewhat beyond their data limits (P. aZbZoauZis, 1-37 cm dbh) to account for the trees mentioned above. . The independent variables used in the regressions were chosen for their ease of field and laboratory measurement and were not necessarily those that gave the best fit to the data. For example, the independent variable "dbh x tree height" gives a slightly better fit to regressions of mass and production than simple, "dbh". However, ■ dbh was used for correlation with these data because it was not feasible to make height measurements for every tree in every stand. After unit area mass and production were calculated for each ■ stand, the two sets of variables were regressed against stand basal area and arboreal canopy coverage, and median shrub coverage for the tree and shrub synusiae respectively. x There is an inherent bias in logarithmic, sum of least squares' regressions due to the skewed distribution of the (arithmetic) squared deviates around the mean regression line (Baskerville 1972, Beauchamp and Olson 1973). The.antilogarithmic conversion of the . regression equation to arithmetic units should thus result . in systematic .underestimation. Brownlee (1967, as cited by Baskerville 1973) provides a method for correcting the arithmetic estimates. "Corrected" estimates of P. ■'aTtricautis data were graphically com­ pared to the original regression estimates and to the original data. In poorly correlated regressions (leaf mass-dbh) the difference between the corrected and original estimates is large, but this is likely due to the high variation in mass of component parts of the largest sampled trees (Whittaker et al. 1974). In highly correlated regressions (aboveground mass-dbh), corrected ■ estimates appear to overestimate the original data significantly. Madgwick and Satoo (1975) compared the actual standing crops to "corrected - unbiased".logarithmic regression estimatesof very small forest stands. The corrected regression estimates consistently overestimated the actual standing crop (Ibid., Table'S, p. 1449). Corrected logarithmic regression estimates have not been used in this report and the resulting calculations might, therefore, be considered mean to minimal estimates. 29 30 RESULTS M D DISCUSSION " Vegetation ■ Fe dlbisCaulisS always provided over 75% (usually 8-5-100%) of the. basal area (BA) in the stands studied; other contributors were old. "wolf" trees of Finus cantovta IatifoZia Engelm. and infrequent individuals of Abies lasiooaz*pa (Hook.) Nutt, and Ficea engetmannii Parry. Mean BA's for the stands (3 plots/stand) range from 25 to 2108 m /ha (c.v.,20% or less). This range of BA is similar to that. of many mature Rocky Mountain'forests, studied by Daubenmire and . Daubenmire (1968) and Whittaker and Niering (1975), but smaller than. 2stands, of the Tsuga heterophyZZa series (100-500 m /ha, Daubenmire ■ • 2and Daubenmire 1968) and of the Pacific Coast fore.sts (100-300+ m / ha, Franklin- and Dyrness 1973, Westman and Whittaker 1975) . Canopy ■ coverage of the tree layer ranged from 36 to 95% (c.v. 25% or less) and was correlated with BA (r = 0.81, Table I). Such open canopies suggest "woodland" conditions, but as will be shown later,•tree densities and biomasses are suggestive of a forest association. Median coverage of Vaooinium sobparium ranged from 12 to 67% (cf. Weaver and Dale 1974). These shrubs are distributed in patches on the forest floor (Forcella 1975), and stands on steep (32%+) slopes support considerably less V. sooparium than those of more' . level topography (slope vs. coverage, r =.-0.5)., This lack of 31 shrub coverage, in small patches or entire stands, may be due to'. snow drifts (Knight 1975), snow movement, or solifluction. ■ ' Herb coverage was' generally less than 5% and was not noticeably correlated with shrub coverage, though the differing phenologies of the taxa may obscure any relationship. Herbaceous species with high constancies were: Aienica tatifotia Bong 93%, Caeeex vossii ■ ■ Boott 71%, Lupt-nus argentrus Pursh 64%, C. geyeri Booth -57%, Poa nervosa Vasey 57%, Eieraciwn graeile Hook.. 57%, Epilobiwn angus- tifolium L. 57%, Potentilla diversifolia diversifolia Lehm. 50% and A. cordifolia Hook.. 50%. The shrub, Juniperus communis- montccna Ait., with negligible cover, occurred in 57% of the stands. No annual species were observed in the stands. Regressions '■ * . ■ The biomass and productivity regressions generated for P. albieaulis are remarkably similar to those of taxa presented by . Whittaker et al. (1974), Whittaker and Woodwell (1968) and Weaver and Forcella (1977). The regression statistics, independent variables and number of data points for all regressions used are listed in Tables 2 and 3. The correlation coefficients range from 0.72 to 0.99 and all are significant at the 99% level. ' Better evalu­ ations of regression error are made with the standard error "e" Table 2. Individual tree and stand biomass and productivity regression statistics for Pinus dVbicaulis and associated species. Equations are in the forms: y = a + bx and y = a = b Iog10 R E G R E S S I O N S T A T IS T IC S D e p e n d e n t v a r ia b le ( y ) r e E n S lo p e (6 ) Y - in te r c e p t (a) I n d e p e n d e n t v a r ia b le s (x ) T r e e M ass (k g ): L e a v e s 0 .9 1 7 4 1 .4 9 9 3 2 0 .8 7 5 5 - 1 . 4 2 9 L o g [ d b h ( c m ) ] 2 W o o d 1 -1 0 c m 0 .8 5 2 2 1 .5 5 5 3 2 0 .6 7 4 2 - 0 . 4 3 1 W o o d 1 0 + cm 0 .9 .5 9 2 1 .3 7 5 2 8 1 .4 1 7 0 - 1 . 8 5 0 W o o d 1+ c m 0 .9 8 7 2 1 .2 3 8 3 2 1 .2 3 5 0 - 1 . 2 4 4 A b o v e g r o u n d 0 .9 8 7 6 1 .2 2 4 3 2 1 .1 8 9 0 - 1 . 0 8 2 B e lo w g r o u n d 1+ cm 0 .9 9 7 1 1 .1 5 1 0 9 2 .1 0 7 0 - 1 . 2 8 5 L o g d b h (c m ) W o o d V o lu m e ( m 3 ): 0 .9 9 1 6 1 .1 9 4 2 6 2 .9 0 1 0 - 4 . 5 7 6 " T r e e P r o d u c t iv ity (k g /y r ) : L e a v e s 0 .9 6 1 3 1 .3 3 0 17 1 .6 7 5 0 - 1 . 9 6 6 L o g d b h (c m ) B o le 0 .7 2 7 1 2 .9 5 6 2 6 2 .4 4 2 0 - 3 . 5 6 4 B o le V o lu m e ( m 3 ): 0 .7 2 7 1 2 .9 5 6 2 6 2 .4 4 2 0 - 6 . 1 8 3 S ta n d M ass (k g /m 2 ): L e a v e s 0 .9 4 7 4 0 .0 0 2 7 1 4 0 .0 2 7 0 - 0 . 1 3 2 S ta n d B asa l A rea ( m 2 /h a ) W o o d 1 0 + c m 0 .9 4 3 7 0 .0 3 8 6 - 14 0 .3 8 2 0 - 7 . 0 2 9 B ark 0 .8 4 7 9 0 .2 2 7 3 1 4 0 .0 1 6 1 0 .4 0 1 B e lo w g r o u n d 0 .9 9 1 1 0 .0 0 3 9 1 4 0 .1 0 1 0 - 0 . 4 5 3 T o ta l T re e 0 .9 6 7 8 0 .0 4 4 6 1 4 0 .5 9 4 0 - 7 . 1 5 2 S ta n d B o le V o lu m e ( m ’ / m 2 ): 0 .9 4 3 7 0 .0 0 7 1 1 4 . 0 . 0 0 0 9 - 0 . 0 1 6 »» 1» J» »» S ta n d p r o d u c t iv ity ( k g /m 2 /y r ) : L e a v e s 0 .9 8 1 5 0 .0 0 0 3 1 4 0 .0 0 4 5 0 .2 0 6 9 S ta n d B a sa l A r e a (m 2 /h a ) W o o d 1 0 + cm 0 .9 5 4 3 0 .0 0 0 2 14 0 .0 0 1 9 - 0 . 2 5 7 0 B e lo w g r o u n d 0 .9 1 0 6 0 .0 0 0 2 1 4 0 .0 0 1 3 0 .2 8 2 5 T o ta l T re e 0 .9 9 1 7 0 .0 0 0 3 1 4 0 .0 0 7 7 0 .2 3 2 1 W o o d 0 .9 6 5 9 0 .0 1 1 5 14 0 .0 6 7 8 - 0 . 1 2 1 0 L e a f M ass ( k g /m 2 ) S ta n d B a sa l A r e a ( m 2 /h a ) 0 .8 1 1 3 1 3 .5 0 0 0 2 8 0 .9 6 2 0 0 .1 7 1 0 C a n o p y C o v e r a g e (%) Table 3. Individual quadrat (I m2) and stand biomass and productivity regression statis­ tics for Vacoinium sooparium. Equations are in the form: y = a = bx. R E G R E S S IO N S T A T IS T IC S D e p e n d e n t v a r ia b le O') r e n S lo p e ( 6 ) Y -in te r c e p t ( c ) I n d e p e n d e n t v a r ia b le ( x ) M ass ( g /m J ): A b o v e g r o u n d 0 .9 3 8 7 2 0 .1 0 0 2 9 2 .1 9 8 0 - 2 8 . 9 5 0 0 % C o v era g e ^ B e lo w g r o u n d 0 .7 4 7 3 9 5 .8 5 0 11 4 .2 8 0 0 2 .7 5 1 3 „ « D e a d T w ig s 0 .8 9 0 4 4 .8 6 8 2 9 0 .3 6 4 7 - 6 . 3 7 3 0 " I P r o d u c t iv ity (g /m ^ /y r ) : „ L e a v e s 0 .8 9 5 1 2 .6 7 4 2 9 0 .2 1 5 7 - 1 . 5 3 3 0 % C o v e ra g e 5 C u r r e n t T w ig s 0 .8 3 0 3 2 .8 3 7 2 9 0 .1 6 9 8 - 0 . 5 5 2 8 ( t o b e p r o p o r t io n a l to a b o v e g r o u n d p r o d u c t io n ) M a ss ( g /m a ): A b o v e g r o u n d 0 .9 8 8 9 4 .5 5 3 14 1 .5 7 9 0 - 4 . 0 1 8 0 M ed ia n % C o v e ra g e q B e lo w g r o u n d 0 .9 9 2 1 9 .6 6 5 14 3 .9 8 8 0 - 7 . 6 4 4 0 Z T o t a l S h ru b < 0 .9 9 5 3 1 0 .3 8 0 14 5 .5 6 8 0 - 1 1 . 7 2 0 0 " P r o d u c t iv ity ( g /m ’ / y t ) : L e a v e s 0 .9 9 6 1 0 .2 9 8 14 0 .1 7 5 0 - 0 . 3 1 7 4 M ed ia n % C o v e ra g e C u r r e n t T w ig s 0 .9 9 6 3 0 .2 2 9 14 0 .1 3 9 0 0 .2 2 0 5 „ B e lo w g r o u n d 0 .9 8 5 0 2 .6 8 2 1 4 0 .7 9 7 5 0 .3 7 1 4 „ T o ta l S h ru b 0 .9 9 2 8 2 .5 8 3 14 1 .1 1 2 0 0 .2 4 4 0 ....... .............. ............... — • 34 for simple regressions -and the relative error of the estimate "E" for logarithmic regressions (Whittaker and. Woodwell 1968) . ■Tree Mass Tree size within the stands ranged from- seedlings to individuals with breast height diameters of 55-60 cm. The biomass of the largest tree was: aboveground 1367 kg, belowground 266 kg, leaves 112 kg and merchantable bole (10 cm + diameter) 961 kg. 2The arboreal- masses per m were calculated by summing the. esti­ mated masses of individual trees in the sample plots. The resulting summations were well correlated with stand basal area (Figure 2, Table 2). The fact that basal area is also correlated with canopy coverage (Table 2) suggests that aerial photographs might be used to estimate biomass (and productivity), at least in arboreal vege- tation with relatively open canopies. The range of arboreal aboveground mass in the 14 stands was 8.4 2 , to 47.2 kg/m . The ranges of stand’ volume, merchantable (10 cm +) 3 2bole mass, bark mass and leaf mass.,.were 12.3-80.8 dm /m , 5.0—33.7 2 2 2 kg/m , 0.7-2.0 kg/m and 0.7-2.8 kg/m , respectively (Figure 2). 2 2■ The variation (or e) of the "leaf mass/m - basal area" •• regression is small compared to those of total tree or bole wood regressions (Table 2), but simple "leaf mass/tree, - dbh" regressions have typically high variability. Might this indicate that the 35 TO TAL TR E E csj 40 B O L E WOOD BELO W G RO U ND L E A V E S ----XX XX X — XX— X- STAND BASAL AREA ( M2Z h a ) 20 40 60 80 100 STAND BASAL AREA ( M2ZHA) Figure 2, Arboreal biomass and productivity in relation to stand basal area in 14 Vinus albicaulis - Vaocinium soopariwn communities. 36 forest canopy is more internally consistent then, and somewhat independent of, its supporting woody structures? Leaf mass is well correlated with the current functional vascular tissue of • single plants and stands (Shinozaki et al. 1964). Leaf mass and wood production are correlated in P. alb'laaut'Cs stands also (Table 2). 2The mass of major roots (I cm +) range from 2.6-10.4 kg/m . The mean mass of P. albioautis roots 0.1-1.0 cm diameter in the 2 2 eleven I m x 0.4 m excavations was 0.4 kg/m (s.d. = 0.18). The mean mass of fine roots (less than I mm, species composition 2 • unknown) was 0.5 kg/m (s.d. = 0.18). The root mass of the two finer diameter classes was approximately evenly distributed between the two soil depths (0-10 and 10-40 cm). There was no correlation of root mass with the distance of the excavation to the nearest tree, or with the above- or belowground mass of V. soopari-um. . The root to shoot biomass ratio is consistently 1:4 for the tree synthesis Temperate forest ratios are generally 1:5, and woodlands 1:4 (Rodin. and Bazilevich 1967, Whittaker and Marks 1975). Understory Biomass 2Shrub coverage in the m quadrats ranged from 0 to 100%. Above- and belowground mass for a quadrat with 100% coverage was 191 and 37 TOTAL * * b e l o w g r o u n d U 30 ABOVEGROUND L E A V E S 400 rOTylt300 x* BELOW GROUND ABOVEGROUND C O V E R A G E (%) C O V E R A G E (%) Figure 3. The biomass and productivity of Vacoinium scoparium in relation to median shrub coverage in 14 Pinus dtbicaulis — Vaooin— iwn soopavium communities. Aboveground wood productivity has been estimated as described in the text. 38 2 .... '425 g/m , respectively. Herbaceous "standing crops (- production) 2varied, but were small and averaged less than 10. g/m „. • The mass of the shrub-herb layer (the mean of the averages' for 2three plots per stand) ranged from 23-109 g/m for the aboveground parts, and 20-264 g/m^ for the belowground (>L mm diameter) portions (Figure 3). The regression equations used for the mass of V. seopari-im appear to also predict the aboveground weight of V. ■myrti-ZZus > a similar species, in Finland (Malkoene 1975). The below- to aboveground ratio for this synusia is 2.0-2.5:1.0, which is nearly identical to the ratios for V-. VaocZZZans and GayZussaoia baooata (Whittaker and Marks' 1975). This ratio for the understory, however, is strikingly different from that of the tree layer; one might speculate that selection for these contrasting "storage" strategies involved ground fire, animal browsing and competition for light. Total Masses . . . . Total aboveground biomass for both the arboreal and shrub- 2 ■herb synusiae ranged from 8.5-47.4 kg/m . Total belowground mass 2 2was 3.7-11.3 kg/m . Total stand biomass ranged from 12.2-58.8 kg/m The older stands are surprisingly massive. Published reports of ■ • forest stands with greater masses of which we are aware are: 2Appalachian cove forests (Whittaker 1966).with 60 kg/m aboveground; 39 ' 2 tropical rain forests (Rodin and Baziievich 1967) with 75 + kg/m total; Thuja pl-Lcata and"Ab-Les grahdis stands in northern Idaho 2(U.S.A,.maritime influence) with up to 93 kg/m. (Hanley 1976); and mesic northeastern Pacific Coast Tsuga hetevoyhylla - Fi-cea ■ s-ttohens-ls (Grier and Logan 1975), Pseudotsuga Tnenzies-Li- and. Sequoia sempewivens (Franklin and Dyrness 1973) forests with aboveground ■ 2 ■ ■weights of 87s 177, and 270 kg/m , respectively. Tree Productivity " ' Annual productivity for the largest trees in the stands (55-60 3cm dbh) was: bole wood, 6 kg. (14 dm ) ; leaves, 10 kg; arid below­ ground (estimated to be directly proportional to aboveground pro­ duction on a- biomass basis), 3 kg. Annual leaf production was normally 12% of the total leaf mass for individual trees. Leaves lived for more than 13 years in many cases. About 30% of the leaf mass had dropped after 6 years age, and .85% after 8 years. Similar leaf longevity was reported for a subalpine stand of Abies maa?iesii and A. Veitehii in Hokaido' (Kimura 1963). Grime (1966) suggests that long leaf persistence is correlated with high environmental stress. Stand wood production, found by summing the productivities of 2trees present in sample plots, was 32-180 g/m /yr (= 0.08-0.42 3 2 2dm /m /yr). Bark production of P. atbicaulis was about 1-7 g/m /yr 40 (Appendix I). Leaf productivity for the tree layer ranged from 2133-475. g/m /yr, and the belowground productivity estimates were 2from 50-144 g/m /yr (Figure 2). I have estimated a 6-8 year pistillate cone and seed production sequence in 28 P. atbicauHs stands which include the 14 stands . discussed here (Chapter 3). Average cone and seed production for 2the 14 stands considered here range from 25-84 and from 8-27 g/m /yr respectivelyc Average annual cone production is better correlated with (and is probably a function of) stand canopy coverage than with any other stand characteristic measured, ' Though there is wide variation among the stands, cone and seed productivity represent, about 9 and 3% of total tree -production. Of total tree production (229-890 g/m /yr),.' about 55% (s.d. = 3.6) is devoted to the leaves; the leaf production to aboveground wood production ratio is about 3.5, and smaller stands have a slightly greater ratio than the larger stands'. . The funnelling of photosynthate to new leaves was also observed in BetuZd 'piibescens (60%) in Greenland (Elkington and Jones 1974), and the phenomenon may be characteristic for treeline forests. In contrast to the stress environment species, the generally slow growth of woody ' plant seedlings may be due to channelling of.photosynthate to wood rather than leaf production (Grime and Hunt 1975). In temperate 41 forest trees the photosynthetic product channelled into wood production also equals or exceeds that which is used for leaf, production (Weaver 1975, Rodin and Bazilevich.1967). Understory Productivity 2 - -•Maximum shrub productivity for a m quadrat (100% coverage) was: leaves and twigs> 37 g/yr; and belowground (estimated as with tree, roots), 81 g/yr. The.range of average productivities for the shrub synusia was: 2 gleaves, 3-12 g/m /yr; and current twigs, 2-10 g/m /yr (Figure 3). Stem wood increment was not measured in V.- soopccriun, but since biomass and productivity ratios for Gccytussao-ta baccata (Whittaker. and Woodwell 1969) and 7. scopar-ium are nearly identical for organ ■ systems measured on each, I based the productivity estimates on the branch wood and bark production to aboveground mass ratio of G. baceata (0.086). Productivity of stem tissue in 7. scopaviim, 2which lacks a. main stem, would then be 2-9 g/m /yr. Total above- 2ground woody production for the shrub is 4-19 g/m /yr. The below­ ground productivity estimates for the shrub layer range from 4-79 g/m2/yr. - ' ' Total Productivity ■ Total aboveground productivity of all synusiae ranged from. 2208-752 g/m /yr. Fine root productivity, not considered above, was 42 2 • * assumed to equal 10% of its mass, i.e., 51 g/m /yr. The range of - 2 'total belowground productivity was 168-267 g/m /yr. Thus total 2 stand productivity ranges from 381-951 g/m /yr (the above figures ' are not additive). Such productivities are relatively low for temperate forests, but.are probably typical for temperate woodlands (Whittaker and Marks 1975, Rodin and Bazilevich 1967). Kimura's 2(1963) subalpine Abies stand produced 1100 g/m /yr, and the arctic treeline Betula pubeseens 207 g/m^/yr (Elkington and Jones 1974). . The mean below- to aboveground productivity ratio for P. albieaulis trees is 0.25 (s.d. = 0.03); the more massive stands have ratios less than the mean. Weaver -(1975) has shown that root-shoot ratios decline as oak trees age. , Kimura (1963) reports a produc­ tivity ratio of 0.30 for a single Abieŝ stand, and Bray (1963) . calculates a ratio of 0.21 for arboreal species in general. The. ' belowground/aboveground productivity ratio for all stand components, including shrubs, fine roots, etc., is 0.50 (±.12); again, the more massive stands, especially those with little V. saopariton cover, are those with ratios less than the mean. The shrub-herb ■■ , and fine root components play a significant, role in stand dynamics. I ■ 'In young stands the productivity of the shrub-herb layer may approach 25% of total stand production,.though its corresponding biomass is normally less than 2% of the stand total. 43 Stand Age vs. Mass and Productivity In contrast to the good empirical "mass or productivity - basal area" relationship (Figure 2), stand age. is a relatively poor predictor of arboreal mass and production. (Figure 4). Cone production of- p„ atbi-caulis fluctuates widely between years (Chapter 3) and variable seed crops may. play an important role in the -initial establishment of a stand. Extremely high variation of tree density in stands of similar age may. be a result of this. Figure 5 incorporates the field data of Weaver and Dale (1974) .. with that of this study to show the possibility of 5-fbld differ­ ences in stand densities of trees greater than 1.0 cm dbh in 50 year old stands. Figure 5 also indicates that high density varia­ tions persist, though they progressively decrease to an age of about 400 years. Basal area, biomass and productivity are in part a function of tree density (Spufr 1952), and the high variation in density may partially explain the. poor mass - age ' relationship. The great mass of Stand 2 (240 years old) cannot, however, be explained by density alone. Analysis of increment cores from this stand reveal a high growth rate (annual radial increment greater than 1.5 mm) until, the year 1850. At that time there was an abrupt reduction in growth (annual radial increment 0.27 mm, s.d. = 0.08 mm) that has continued to the. present; growth 44 ^ 20 PRODUCTIVITY STAND AGE (years) Figure 4. The biomass and productivity of 14 Pinus albicaulis Vaocinium scoparium communities in relation to the age of the stand. The left and right Y-axes represent biomass and annual productivity respectively. 45 IOO X COCO O 33 ZCSCDm 3) H O XO $>I- COmmo 2 NI x 30 C A N O P Y C O V E R A G E (%) Figure 7. Cone mass and number per m and seed mass, number and energy (endosperm and em­ bryo) per m2 in relation to stand canopy coverage of 28 stands (Stands 1-29; there is no Stand 21) of the Pinus albioaulis - Vaaainium saoparium association. The numbers 1-3 refer to stand areas of less than 0.75 ha, 0.75-1.25 ha and greater than 1.25 ha respectively. 62 As stated above, the number of fallen cones (F) was correlated - 2with C (cf. Sarvas 1962). A multiple regression of C over A and F 2 provided the equation (r = 46%) : ' C = 0.450 + 0.00023 (A2) + 0.206 (F) Eq. 7 ■ Though possibly not as useful as Eq. 5, this regression might.be beneficial for field surveys. Further multiple regression analyses employing BA, stand age, stand slope,. elevation, aspect,, etc. increased attributal varia­ bility up to 60%. However, such complicated regressions soon- become intangible and make data collection unwieldy. Originally I had expected that the counts of newly—fallen cones ■ (or remnant cone axes) to equal the actual cone production of the previous year. I also expected that the "leader shoot cone pro- duction estimates" would equal the newly-fallen cone counts for the appropriate year. The two methods were not equal in their estimates, but they were correlated (both nonparametrically, r^ = 0.71 [Mosteller and Rourke 1973] and parametrically, r = 0.72). Newly-fallen cone counts underestimated the leader shoot estimates' • (though not badly) as can be seen from the regression/ coefficient and equation for cone production in 1974: C(1974) * 0.305 + 1.314 V ne„) ' Eq. 8 63 Mounds of old cone scales can often be found at the entrances of rodent dens. This suggests that newly fallen cones may, at least temporarily, be buried underground; thus: I) making fallen cone counts an underestimate of actual production, and 2) partially explaining the inequality of the two cone production methods. . Biomass and Seed Number . ■ * 2 ■ An average mature ovulate cone weighs about 23 grams (range 10-50 g); cones produced in good years tend to be larger than those of poor cone crop years. Single cones support about 75. scales of which a third are infertile (generally the apical and basal scales only) . Typical cones contain 75 seeds; each with . a mass of about 0.1' g. Within a cone, seed mass is generally 30% of the total,, but it may occasionally (and consistently) reach as high as 50% within a stand in a good cone year. Either of these two values is quite high when compared to the repro­ ductive effort in cones of other coniferous species (Smith 1970). On a unit area basis, average cone production ranges from 6 2 • 2 • ' to 84 g/m /yr, with seed mass being 2 to .25 g/m. /yr (= 20-250 seeds) between stands. Previously in Chapter 2, I determined net above- and belowground annual primary productivity (less cones) of 14 of the stands studied here; cone and seed production repre­ sent about 10 and 3%, respectively, of the total tree production. 64 Comparisons With Other Forests Comparing cone and seed crops of various forests is. difficult. Few investigations have asked the same initial question, used the same method for data collection, reported the same "type" of data or in the same units. However, Table 4 presents average cone.and seed crops for several Pinus forest types. Schopmeyer (1974) was consulted for.masses of seeds and cones when not provided, by the specific reference. The number of cones produced by P. albieautis forests appears to be smaller than that of other pine types, especially Diploxylons.. . However, cone mass, seed mass, seed number and the % seed mass' of ■ total cone mass, are all greater for P. albieautis.. P. cembra Sibevicay whose individual seed mass is 2~3 times that of P. albieautis, -may be an exception.. Its seed crops are of sufficient quantity that they are annually harvested for human consumption, but the only seed mass per unit-area figures available are those of Formosof (1933) , and these appear to be misprints. If Formosofs figures should read "grams" rather than "kilograms", they would then be 25-30 g/m ; which is at the upper limit of the P. albieautis range. / - ' ' ' • _ IJ Table 4. Comparison of cone and seed crops of several Pinus forest types„™ • ■ - • • ■ 2 . 2 forest Type'_____^ ■ Average Cones/m /yr' ■. Average Seeds/m /yr ■ Reference ' # . gfams ■ # grams 7P %&/ Pinus oontorta 4.7 28 i-— 268 0.6 . - 2 - Smith 1968, -70 P. aontoria - , Purshia tridentata . 4.8 . 24 . 120 0.4 2 Lotan 1967 P. oontorta - ■ Geranium fremontii. 8 .0 . 40 13 ' ___ ’ . Moir 1972 ' . ' \ P. oontorta - Vaooinitm myrtitlus 8.0 • 40 . 5 . _ Moir 1972 P. oontorta boZanderi 52 20 —i— — — - - West. MThit. 1975 P. sibirica **•— — 100 ’ ' 30 — Formosof 1933 P. sibirioa - . Vaocinitm myrtilZus 0.1-2.4 Boichenko 1970 P. syZvestris - CaZZuna vulgaris 1.5 9 __ 30 0.2 2 _ Sarvas 1962 P. syZvestris - Vacoinium myrtiZZus 3.0 ' . 18 60 0.4 ■ 2 Sarvas 1962. P. syZvestris - OxaZis aoetoseZZa 4.5 27 90 0.6 2 Sarvas 1962 P. monophyZZa Juniperus osteosperma 0.1-1.8 2-34 1-26 1-9 28 Forcella unpubZ. P. eduZis - Juniperus osteosperma 0 .8 -P . 8 ' — ' _ Forcella unpubZ. P. oembroides - Juniperus deppeanm 7 .3 1.7 . ' '35 IS Forcella unpubZ. P. I/ aZbioauZis - Vaooinitm sooparium . 0.3-3.6 ^6-84 Ranges in figures represent smallest and 10 20-250 2-25 I largest data provided. 30 3 2/ Percent total pro- ductivity except third and fourth forest types which are % aboveground productivity only. Sj Percent total cone mass. ' . 66 Localizing Cone Crop Variability ■ As stated earlier, extreme variation in cone crops occurs between years in any stand. But variation may also be expressed between: I) branches within a tree, 2) trees within a stand, 3) stands within a geographic, region, and 4) stands within the entire P- Olbicauli-S - V. seopâ i-um association (assuming the stands are representative) . To localize the source of variation in cone productionI per­ formed a nested analysis of variance, using the 5-branch/tree samples as the basic data, thereby allowing the comparison between large and small trees, young and old stands, etc. Annual variation of production was tested as an interaction in each "nest" of the , analysis. . Cone crops per branch within the 5-branch/tree- samples were not significantly different (p > 0.05) in 71% of the trees, but annual production, was significantly different in 60% of the same trees. Differences in cone production of individual trees within a stand were not significant (p > 0.05) .in 52% of the stands, but annual production was in 78%. The stands were divided into 3 geographic regions; though the . divisions were subjective, the floristics of the stands substantiate the separations. The regions are: I) the Wind River and Absaroka Mountains, 2) SW Montana,. and 3) the Bitterroot and Salmon River 67 Mountains (see Chapter I). Cone production of stands within each region was not significantly different (p > 0.05), but yearly pro­ duction was in each region. An overall analysis, of variance (the final nest) of all stands sampled within the P. OVbi-QauZjIs - V. saopariiMn association showed no significant difference (p > 0.05) both between stands and between years. It must be recalled that the analyses did not include the entire cone crops of the trees of stands, only the 5 potential cone- producing leader shoots per tree for 5 trees in. each stand. With this in mind,'the above results suggest that the ability to produce strobili: I) is inherently similar in branches of the same tree, 2) becomes dissimilar in trees of the same stand,, 3) is similar among stands within and between geographic regions (i.e., all stands have equal potential), and 4) the actual cone crops of trees and stands are subjected to annual fluctuations in cone produc­ tivity; the non-significant difference between years of stands within the entire association is probably due to compensating variations between geographic regions. This can be shown with factorial analysis where the geographic regions are "blocks" in the analysis. Again, there is no significant difference between stands or years, but there is between geographic regions. 68 CONCLUSION AND DISCUSSION The actual number of cones produced over a number of years in the Pinus albieauiis - Vacoinium scopanium association is probably a function of the number of leader shoots possessed by a tree or a stand. In turn, canopy coverage is in part a description of the numerical and areal extent of the leader shoots; and it is moder­ ately correlated with mean cone productivity. I used a relatively crude method of estimating stand canopy coverage (verticle peri­ scope, 90 points/stand). The summation of canopy areas of all trees in a stand would have been more appropriate--but time consuming. Air-photo interpretation (where possible) of stand canopy coverage may prove an even more useful technique for esti­ mating potential cone crops in whitebark pine forests. When twenty-five leader shoots from each of 28 stands were analyzed for a six to eight year cone production sequence, there was no statis­ tical difference between the stands (except temporally); suggesting ; that all stands, regardless of their age, size, etc., have similar potentials for cone production. A further suggestion is that the ■ actual difference in totaVcone production between the stands is a • consequence of total leader shoot number (canopy coverage). With regard to the management of these forests for cone production whether for silvicultural or wildlife measures, large (but not necessarily old) stands with high canopy coverage (canopy surface area may be even more appropriate) should be encouraged.' Older stands (200 + years) often do have more extensive canopies, and in conjunction with the fact that P. dlbioautis forests are slow growing but long-lived, older stands become even more valuable. Indeed, it may take a young 50-100 year old stand one hundred . to. two hundred years to simply double its mean short-term annual cone productivity. Pinus aWicaulis cone and seed production varied greatly both between stands and between years. Cone crops varied from zero to 2eight (0-184 g) per m /yr and seed crops from zero to 600.(0-60 g) 2per m /yr. There were approximately an equal number of good and • poor cone producing years in the stands sampled. The mean cone and seed crops, from a 6-8 year cone production sequence, ranged 2from 0.25-3.4 cones (6-84 g) per m /yr and 20-250 seeds (2-25 g) 2'per m /yr. Pinus albicaulis cone and seed crops are as great or greater in both mass and number than .those of many other pine species.. The large, often wingless seed of P. aibicaulis and other Haploxylon pines (Mirov 1967), which occur.in "stress" environments, not only provide a large energy source.for their respective embryos but also.for their predators (Maley 1975, Formosof 1933) , some of whom 69 70- may function in diaspora dispersal. Thus, for successful coloni­ zation and recolonization, trees such as P. CLlbyLoaulis may not only have to produce large energy-rich seeds, but also great numbers of them. This will: I) attract dispersal agents, 2) mitigate the destructive;influence of the dispersal agents, and■other predators, • and 3) insure successful germination and seedling establishment. Trophic levels are probably essential for the continual functioning of communities, and their indigenous species. Yet large consumer popu­ lations, if not controlled, could be devastating to producers. It seems reasonable (and often obvious)•that strategies for predator avoidance to have been selected for in prey species. Several insects are obligate seed predators of coniferous trees (Mattson 1971, Abrahamson and Kraft 1965, Radcliffe 1952). Mattson (1971) found the populations of seed-consuming insects in. Pinus vesinosa. ' to be correlated to the cone crop size of. the previous year - regardless of current cone abundance. Perhaps, one strategy for a tree, to avoid seed predation, is to fluctuate cone production (Smith 1970, Janzen 1971). Fluctuations should be of sufficient magnitude as to intermittently "over-" and "under-provision"'the predator population. If P. albioaulis seed crops are classified and defined as: I) "excellent", greater than one s.d. unit above the mean, 2) "good", more than average, and 3) "poor", less than 71 average; then excellent seed years should be preceded by an equal number of good and poor years. In my data,, excellent seed years were preceded by a significantly greater and lower than expected " • 2number of poor (21) and good (9) seed years, respectively (X = 4.8 p < 0.05). Variation in seed crops may be both spatial and temporal; they may be inherent within the tree (Rehfeldt et al- 1971, Shoulders ■ 1967, Varnel et al. 1967), climatically induced, or both. The particular strategy or group of strategies employed by a tree species is probably regulated in part by the number of predators and the degree of predation. . . . . •. CHAPTER 4 FOOD PRODUCTIVITY IN A SUBALPINE ROCKY MOUNTAIN FOREST ASSOCIATION 73 INTRODUCTION One of the benefits of considering vegetation as existing in discrete, more or less homogeneous repeatable units, is the ability. to make generalizations about them. These generalizations can take the form of management recommendations such as: cattle grazing potentials, timber harvests, road building and maintenance, rec­ reation feasibility and many others. The abstract vegetation units, or plant associations, are also prerequisite for constructing vegetation maps - which further facilitate management decisions. Some plant associations of a given region are particularly and consistently more valuable for certain uses than are others. For instance, the subalpine- Pinus aVb-icaul-is - Vaecinium seoparium association (whitebark pine - grouse whortleberry forests) of the . northern Rocky Mountains is an important component of the array of plant associations which comprise the environment of Unsus avetos (grizzly bear), U. amevicanus (black bear) and numerous other animals. ' The importance of this association to wildlife stems from its high production of high energy foods. If quantitative generalizations concerning the annual production of these foods within this association can be made, then I) the potential annual diet of the animal(s) is known, and 2) approximation of wildlife demographics are possible. 74 The vegetational characteristics of the P. Olb-Loautis.- V. sooparittm association have been generally described by Weaver and Dale (1974); as have above- and belowground productivity (Chapter 2), the spatial distribution of the understory shrubs (Forcella 1975), the annual production (and its variability) of ovulate, cones and seeds (Chapter 3) and the distribution of the association with its floristic variation (Chapter I). . . . In this chapter I shall report production and calorific, value summations for all plant foods, known to be of importance- for wild­ life, taken from 28 stands of P. dlbioauVLs -'7. scopariim associa­ tion (Figure I, excepting stand 30). The geographic range of these stands and of the association itself, extends from west-central Wyoming, through southwestern Montana to east-central Idaho. ■ METHODS ' Canopy coverage, the percent of ground surface covered by a vertical projection of the aerial parts of plants (Daubenmire 1959), makes an efficient parameter of plant biomass when it is estimated separately for each species occurring within a given habitat. Such estimates were made in 10% classes for each vascular plant species encountered within forty-five 2 x 5 dm frames in each of the 28 .sampled stands. Further sampling details are given in Chapters I, 2 75 ■ and 3. From several of these 2 x 5 dm plots, the aboveground por­ tions of all plants which were of known forage value and of sufficient quantity to be of.importance, were also clipped, oven-dryed (60oC) and weighed.■ (For sample numbers, see caption of Figure 8). ' Fruit production of V. Seopari-Vm was determined by picking, drying (60°C) • 2and weighing all berries which occurred in several I m plots. The coverage of this shrub was additionally estimated as above, in each.. . of these plots. Calorific values were obtained for all plant parts through standard bomb calorimetric procedures (Chertu Stn. Analytic Lab., M.S.U.). ■ Simple linear regressions were developed for both mass and 2kilocalories (annual production) of each species