Growth and Physiological Responses of Tundra Plants to Defoliation Author(S): Steve Archer and Larry L

Growth and Physiological Responses of Tundra Plants to Defoliation Author(s): Steve Archer and Larry L. Tieszen Source: Arctic and Alpine Research, Vol. 12, No. 4, Patterns of Vegetation and Herbivory in Arctic Tundra: Results from the Research on Arctic Tundra Environments (RATE) Program (Nov., 1980), pp. 531-552 Published by: INSTAAR, University of Colorado Stable URL: http://www.jstor.org/stable/1550499 Accessed: 08/05/2009 15:03

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=instaar.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

INSTAAR, University of Colorado is collaborating with JSTOR to digitize, preserve and extend access to Arctic and Alpine Research.

http://www.jstor.org ArcticandAlpine Research, Vol. 12, No. 4, 1980, pp. 531-552

GROWTH AND PHYSIOLOGICAL RESPONSES OF TUNDRA PLANTS TO DEFOLIATION

STEVE ARCHER* AND LARRY L. TIESZEN

Departmentof Biology AugustanaCollege SiouxFalls, SouthDakota 57197

ABSTRACT

In addition to extreme abiotic conditions, sults from this study support the hypotheses biotic factors such as grazing influence the that (1) carbon allocation patterns are a func- growth of tundra plants. Strategies of carbon tion of growth form and dictate plant re- assimilation, accumulation, and utilization sponses to defoliation; (2) maximum photo- must not only satisfy the requirements of a synthetic rates are a function of growth form rigorous physical environment but also must and are inversely related to leaf longevity; (3) simultaneously adjust to the impacts asso- the impact of grazing is reduced in plant ciated with herbivory. Representatives of four species with rapid leaf turnover and little sup- growth forms found in northern Alaska (a de- portive tissue; (4) defoliation results in an im- ciduous shrub, an evergreen shrub, a single- mediate adjustment in carbon allocation pat- shooted graminoid, and a tussock-forming tern within the defoliated plant; and (5) car- graminoid) were subjected to various defolia- bon allocation to maintenance tissue or long- tion regimes and their physiological and mor- lived productive tissue is accompanied by phological responses were documented. Re- mechanisms that reduce grazing.

INTRODUCTION

Until the recent discovery of petroleum re- plant-herbivore interactions. First, the soil serves, the most important use of arctic tundra and biological communities of tundra are rela- for man has been as grazing grounds for na- tively young and are nutritionally im- tive ungulates and domesticated animals poverished both in the Arctic (Ulrich and (Dahl, 1975). And yet our understanding of Gersper, 1978) and in alpine regions (Costin, grazing relationships, at least from the point 1967). Although their evolution is not known of view of the plant, is poorer than both that of well, tundra floras and faunas can be con- other ecological processes in tundras and of sidered depauperate (see Hoffman and Taber, herbivory in other biomes (see Tieszen and 1967) due perhaps to the recent and severe Archer, 1979). Pleistocene extirpations. Second, tundras There are several characteristics of tundra which evolved in the total absence of ungu- systems which have special implications for lates have a similar physiognomy to those with large grazers. Thus gross community structure has not been directly determined by large *Presentaddress: Range Science Department, mammalian herbivores, although they can be ColoradoState University, Fort Collins, Colorado important in determining species composition 80523. (Mark, 1965, 1969; Klotzli, 1977). Third, the 0004-0851/80/040531-22$03.30 ?1980, Regents of the University of Colorado S. ARCHER AND L. L. TIESZEN / 531 "extreme"nature of some environmental fac- growth form is characterized by a particular tors-low temperatures, low soil nutrients, allocation of nutrients which presumably and short growing seasons-all help to ac- allows successful exploitation of particular count for low values of primary production. habitats. Lewis and Callaghan (1975) and As a result, green biomass of high nutrient Tieszen and Wieland (1975) have described quality, although predictable in occurrence, is distinct growth strategies of meadow grasses, often a scarce resource (Figure 1). Fourth, tussock sedges, and fellfield forbs, and these plants which inhabit tundra systems have growth forms have been related to specific evolved a variety of adaptations which enable environmental complexes (Webber, 1978; them to successfully cope with the complex of Komarkova and Webber, 1980, this volume). short growing seasons, long winters, low tem- In addition to extreme abiotic conditions, peratures and light intensities, and limited however, plant growth and perhaps commu- availabilitiesof nutrients. nity structure in arctic tundra are also influ- Five growth forms of vascular plants (Table enced by biotic factors such as grazing. Tik- 1), all representedby perennial species that re- homirov (1959), Bliss (1975), Batzli (1975), produce vegetatively, illustrate the array of and Batzli and Brown (1976) have reviewed vascular plants common at Atkasook, Alaska, the environmental impacts, both positive and and represent different evolutionary responses negative, of herbivores on tundra plants and to the same environmental problems. The de- soils. Thus strategies of carbon assimilation, gree to which differences in growth form are accumulation, and utilization must not only related to functional differences is not thor- satisfy the requirements of a rigorous physical oughly documented; however, work of Bliss environment, but must simultaneously adjust (1962), Hedberg (1964), and Mooney and to the impacts associated with herbivory. Our Dunn (1970) has suggested that growth forms goals in this paper are to develop a quantita- have functional importance and that each tive understanding of the relationships be- Seasonality of Forage Availability and Quality

_ I I I I II I I I I I I 1. 100 .= 4

75- E A/ ^ 0o

cC - \ Total j\ 3 C 50 1 b=. 0 0

2 0

0 h._ cm 25- z,, 0

, , , , I, , t , , I I I I 0 [~~~~~~~~~~~ l vI Dec Feb Apr Jun Aug Oct Dec Month of Year FIGURE 1. Seasonal trend in aboveground vascular plant material and nitrogen concentration in wet meadow vegetation at Barrow, Alaska, illustrating the limited availability of high quality (green) plant biomass and the displacement in time of peak biomass and peak nitrogen concentration (Tieszen, 1972; Chapin et al., 1975).

532 / ARCTIC AND ALPINE RESEARCH tween gross carbon and biomass allocation, growth following herbivory were documented. leaf growth, and photosynthesis patterns in We expected the long-term survival of an indi- several tundra growth forms and to determine vidual plant to be determined largely by its how these patterns are affected by grazing. ability either to tolerate or to avoid damage, In a series of field experiments designed to and we addressed the following specific hy- document the physiological responses of tun- potheses: (1) carbon allocation patterns are a dra plants to simulated grazing we sought to function of growth form and dictate plant re- test the general hypothesis that carbon alloca- sponses to defoliation; (2) maximum photo- tion patterns, a function of life-form type, will synthetic rates are a function of growth form dictate plant responses to defoliation that and are inversely related to leaf longevity; (3) minimize impact on production and repro- the impact of grazing is reduced in plant duction. For several species, photosynthetic species with rapid leaf turnover and little sup- rates, leaf longevities, and potential for re- portive tissue; (4) defoliation results in an im-

TABLE1 Photosyntheticcapacities (mean ? 1 SE) of speciesrepresentingfive arctic growthformsa

Photosynthetic capacityb (mg CO2 * g dry wt-l * h-1) Species Species averagec Growth form average Single-shooted Graminoids 39.3 1. Arctagrostislatifolia 47 ? 7 2. Carexaquatilisd 34 ? 3 3. Carexbigelowii 37 ? 2 Tussock-forming Graminoid 27.0 1. Eriophorumvaginatumd 27 ? 2 Forbs 21.7 1. Petasitesfrigidus 21 ? 2 2. Polygonumviviparum 29 ? 3 3. Rubuschamaemorus 15 ? 1 Deciduous Shrubs 38.5 1. Betulanana 37 ? 4 2. Salixpulchrad 40 ? 3 Evergreen Shrubs 9.0 1. Cassiopetetragona Current leaves 9 1 Noncurrent leaves 10 1 2. Dryasintegrifolia Current leaves 10 1 3. Ledumpalustre ssp. decumbensd Current leaves 10 1 1-yr-old leaves 18 + 1 2-yr-old leaves 17 + 1 3-yr-old leaves 11 ? 1 4. Vacciniumvitis-idaea Current leaves 7 1 1-yr-old leaves 5 1 2-yr-old leaves 5 1 3-yr-old leaves 5 ? 1 aAdaptedfrom Johnson and Tieszen (1976). Nomenclature follows Hulten (1968). bMeasured at 15?C, high humidity, and saturating light. cSpecies averages are the mean of 12 to 16 replicates, except Arctagrostis,Petasites, and Polygonumfor which there were 6 replicates. dUsed for defoliation experiments.

S. ARCHER AND L. L. TIESZEN / 533 mediate adjustment in carbon allocation pat- lived productive tissue is accompanied by terns within the defoliated plant; and (5) car- mechanisms that reduce grazing. bon allocation to maintenance tissue or long-

METHODS

Lack of time precluded detailed observation usually removes herbage at uniform heights of several representatives of each vascular and percentages while grazing usually results plant growth form, so we selected one impor- in herbage removal at various heights between tant species from each of four growth forms plants and even within the same plant. (4) found in tundra near the Meade River at Herbivore preferences for particular plant Atkasook, Alaska, for our defoliation studies parts, age classes of plants and phenological (Table 1). Two populations of Salix pulchra, stages are difficult to simulate by clipping. (5) the deciduous shrub, were apparent in the In defoliation experiments herbage is cleanly Meade River area. One consisted of erect and uniformly removed while actual grazing shrubs about 1.2 m tall found along creek and browsing typically involves ripping, margins. The other Salixpopulation was pros- shredding, tearing, or stripping of the vegeta- trate, 20 to 30 cm tall, and was a codominant tion. Damage resulting from trampling and in Carexaquatilis-Salix pulchra-Sphagnum spp. soil compaction must also be considered. (6) communities. This latter population was used There is evidence which suggests that the re- in our studies. Carex populations were covery of grazed plants is enhanced by the ac- examined in nearly pure stands of Carextillers tion of growth-promoting substances in the lightly interspersed with Salix and located in a saliva deposited on the wounded plant tissue moist lake margin. Eriophorumand Ledumex- by insect and vertebrate herbivores (Reardon periments were conducted on high-center et al., 1974; Dyer and Bokhari, 1976; Rear- polygons at a site interlaced with a mosaic of don and Merrill, 1978). polygon centers, rims, and wet troughs. Rubus All primary defoliation treatments involved chamaemorus,Carex bigelowii, and Vaccinium the removal of all exposed photosynthetic tis- vitis-idaeaalso grew on the polygon centers. sue. In the graminoids, leaves were clipped For a more complete description of com- back to moss level. In the Eriophorumtreat- munity types see Komarkova and Webber ments three tussocks that were defoliated early (1980, this volume). in the growing season were covered with In our defoliation studies plant parts were opaque plastic (black bag) to prevent photo- artificially clipped in an attempt to simulate synthesis. When leaf production ceased, the an actual grazing event, because the amount tussocks were destructively sampled for the of herbage removed from plants could be stem base and sheath components. The dry quantified, controlled, and held constant over weight of these structureswas assumed to rep- time and between treatments. Plant responses resent the approximate minimum biomass could then be correlated to known frequen- necessary for tiller growth and maintenance cies, intensities, and types of defoliations. It is (see Archer and Tieszen, submitted, a, b, for important, however, to acknowledge that more details). For the shrubs two basic de- some major differences do exist between clip- foliation schemes were imposed: Type I de- ping and actual grazing, and these should be foliation involved the removal of all current considered when extrapolating results from year's production; Type II treatment involved defoliation experiments to actual grazing sys- the removal of 1- and 2-yr-old stems in addi- tems (White, 1973). For example: (1) Grazing tion to current year's production. For Ledum, may be more detrimental than clipping if it re- the evergreen, old leaves were also removed in moves herbage from some plants and not both treatments. Defoliations on all growth others; unclipped plants may compete advan- forms were imposed at periodic intervals tageously with clipped plants for resources. (2) throughout the growing season. Defoliation Grazing may be less detrimental to plant vigor regimes initiated on 30 June were designated than clipping if some tillers of a rhizome sys- early spring treatments; those initiated on 15 tem are left intact and ungrazed, thus allow- July, 30July, and 15 August were designated ing for a transferof nutrients and energy from late spring, summer, and fall, respectively. intact tillers to injured tillers. (3) Clipping For chronic treatments regrowth was removed

534 / ARCTIC AND ALPINE RESEARCH at 10-d intervals throughout the growing sea- Methodology used for determining photo- son. Recovery treatments refer to those plants synthetic rate, leaf longevities, and above- that were chronically defoliated for one grow- ground biomass allocation are given in John- ing season and not defoliated at all the subse- son and Tieszen (1976). quent growing season. RESULTS AND DISCUSSION

LEAF GROWTH DYNAMICS AND BIOMASS ALLOCATION The abruptness and brevity of the arctic growing season places severe restrictions on the timing of plant growth (Figure 1) and limits the amount of time during which even small amounts of high quality forage is available to most herbivores. The plants at Atka- sook (70?28'N) have approximately 75 d to initiate current growth, expand and develop photosynthetic tissue, reproduce (either vegetatively or sexually), and translocate nutrients below ground prior to autumn die-back. Dif- ferences in seasonal leaf growth dynamics and biomass allocation patterns were observed in the different growth forms and reflect the various strategies employed by arctic plants to l00' fix carbon in the short time allotted. E 4- / / o *4 30 76 6 In the Arctic, graminoids are inactive be- 9-0o Eiz/ neath the winter snow (Tieszen, 1974), but begin to photosynthesize within a few days of 0- melt-off (Tieszen, 1975). The seasonal pro- 0^ \ of leaf area accumulation for each of 350- gression - the life forms are in 40 ? .3 representative graphed 30- 2. 27 had exserted / / \ Figure By June, Eriophorum 20- A over of its normal of 60% complement photo- 10- 0/& 08 synthetic tissue and Carex,the single-shooted d graminoid, had exserted 26% of its leaf tissue. 80 Eriophorum Neither shrub possessed current photosynthe- 70 tic tissue at this point in the growing season. 60 Thus graminoid tillers exploited the early por- 50 \ ____A_-- # tion of the growing season to a much greater 40-. .___ \Leaf extent than did either of the shrub life forms. 30 This was especially noticeable for Eriophorum 20 whose growth form enhances soil warming, decomposition rates, and nutrient availability a30 1 1:-5 : 14 2 within the individual tussock. et al. 30 5 10 15 20 25 30 4 9 14 19 24 Chapin 4-June) July )< - observed that tussocks be- August (1979) Eriophorum Date came snow free sooner and that intratussock (1976) soils thawed faster, reached maximum sum- FIGURE2. Seasonal progression of leaf exsertion for mer temperature sooner, were 6 to 8?C each of the in this Data and were more stable than principal species study. warmer, thermally were derived from direct growth measurements on soils at comparable depths between tussocks. marked individuals at Atkasook. Leaf length in In addition to differences in the seasonality each species was highly correlated with leaf area of leaf production we noted large differences and biomass (ohnson and Tieszen, 1976, with in patterns of leaf development. Buds on both permission of Springer-Verlag; Archer and Ties- shrub life forms did not begin expanding until zen, unpublished).

S. ARCHER AND L. L. TIESZEN / 535 late once occurred I. 2 June. However, expansion 2a Betula nana leaves developed rapidly and concomitantly 0. 8- (Figure 2a, 2b) and by 17 July, within 20 d of O. emergence, all the leaves from Salix buds were .4-n . - . 1-. fully expanded. The leaves on Salix plants b Salix - pulchra typically had an extended maturation phase 1.6I (Lake Basin) followed by an abrupt senescence phase. 1.2! I Ledum, the shrub, also exhibited evergreen 0.8 I- this synchronous pattern of leaf growth. - Newly initiated leaves did not, however, at- 0.4 tain full development until nearer the end of a$ m H r- Carex aquatilis the growing season. Estimates of leaf mor- ._ .I tality suggest that 70% of the leaves initiated on Ledum die after two growing seasons and d that over 96% die after four seasons 0 Eriophorum vaginatum growing 0.1 (Johnson and Tieszen, 1976). The of leaf exsertion of ,r- sequential pattern e Ledum palustre the graminoids (Figure 2c, 2d) contrasts with 0.3 the synchronous growth and development of leaves in shrubs. Graminoid leaves exserted 0.2 late in the season became growing quiescent, 0.1 then resumed and completed growth the following season. Leaves of Carexand Eriophorum 'n -"" n tillers typically had long growth phases, a 4, L. ,. L short maturation and an extended > 'v Eo O0 > O0. E 0O 0 ( E E phase, .t senescence phase. While both Carex and Erio- -J o() J.J JQ) NJ NO A phorum initiated some leaves much earlier in the growing season than Salix and Ledum, FIGURE3. Biomass allocation (mean dry weight) neither achieved maximum leaf area until 8 for aboveground plant compartments for (a) Betula August, for while new leaves were growing, nana, (b) Salix pulchra,(c) Carexaquatilis, (d) Erio- leaves initiated the previous growing season phorumvaginatum, and (e) Ledumpalustre. Average allocations were estimated from a were senescing. Hence, the net increase in compartment of 40 biomass was small. In contrast, Salix attained sample aboveground individuals for Betula, and Ledumand 50 tillers for Carexand Erio- maximum leaf area by 17 July, some 20 d Salix, phorum.Determinations were obtained near sooner than Carexand Eriophorum(Figure 2). peak biomass (21 to 26 July) (ohnson and Tieszen, Allocation of biomass to aboveground tissue 1976, with permission of Springer-Verlag; Archer also differed markedly among life forms (Fig- and Tieszen, unpublished). ure 3). In both the deciduous and evergreen shrubs 80 to 90% of current aboveground biomass (at peak season) accumulated in the leaf ure 3). Nearly 84% of the total nonreproduc- compartment and 10 to 15% accumulated in tive, aboveground biomass in Salix was de- current stems. In Eriophorum, the tussock voted to stems as compared to 75% in Ledum. forming sedge, 96% of the aboveground bio- Neither Carex nor Eriophorum had any non- mass was contained in leaves. The remaining photosynthetic tissue to support or maintain. 4% was invested in reproductive structures. However, in our study area, less than 3% of CARBON ASSIMILATION the E. vaginatum tillers flowered (Archer and Differences in the photosynthetic rates and Tieszen, submitted, a). A given tiller may leaf longevities for representatives of each life produce leaves for 3 to 4 yr after which time a form also illustrated different growth strate- terminal inflorescence may be initiated. The gies. Photosynthetic rates ranged from 5 to 47 tiller then dies (Goodman and Perkins, 1968). mg CO2 * g dry wt-' * h-~ for 13 species repre- Carex also allocated nearly 100% of its above- senting five growth forms (Table 1). Photo- ground biomass to leaf tissue. synthetic capacities were relatively consistent Finally, both shrub types committed large within a growth form, with deciduous shrubs portions of biomass to supportive tissue (Fig- and single-shooted graminoids generally hav-

536 / ARCTIC AND ALPINE RESEARCH TABLE2 Averageduration of differentgrowth phases of leavesof arctictundra species as estimatedfromseasonal progression of dryweight of leavesa

Length of interval (growing season days) Current season Photosynthetic Previous Leaf capacity (mg CO2 * g Species seasons Growth Maturation Senescence longevity dry wt * h-') Graminoids Carexaquatilisb 25 17 26 17 85 34 + 3 Eriophorumvaginatum 27 29 24 15 95 27 ? 2 Deciduous shrubs Salixpulchra 18 34 12 64 40 + 3 Betulanana 17 33 11 61 37 + 4 Evergreen shrubs Ledumpalustre ssp. decumbens 120C 80 200 14 + 1 Vacciniumvitis-idaea 280C 80 360 6 + 1 aFromJohnson and Tieszen (1976). bGrowthdata from Tieszen (unpublished). CEstimatesfrom leaf mortality data for average leaf. ing the highest photosynthetic rates, followed using a leaf for up to three seasons after its ex- by tussock-forming graminoids, forbs, and sertion (ohnson and Tieszen, 1976). Carex evergreen shrubs. The high photosynthetic and Eriophorum,each with photosyntheticrates rates observed in deciduous shrubs were, how- and leaf longevities intermediate to the de- ever, offset by relatively short leaf longevities ciduous and evergreen shrubs, initiated (Table 2). Conversely, the low photosynthetic growth earlier in the growing season and ex- capacities observed in evergreen shrubs were ploited an intermediate proportion of the associated with extended leaf longevities. Pre- growing season. This interaction between sumably evergreen leaves must be retained for photosynthetic rates and leaf longevities can more than one growing season in order to pro- even be seen within graminoid growth forms. vide a positive carbon contribution to the The fact that Eriophorumhad exserted over plant. Apparently leaves with high photosyn- 60 % of its current growth biomass by 28 June thetic capacity contribute positively to the car- as compared to only 26% in Carex(Figure 2) bon budget of the plant in only one season. suggests that the early season growth in Erio- The graminoid growth forms in Table 2 had phorumenables it to compensate for its photo- intermediate photosynthetic rates and also ex- synthetic capacity, which was 20% lower than hibited intermediate leaf longevities. that of Carex(Table 2). Similarly, Vaccinium, Thus it appears that among our tundra which had a photosynthetic rate less than half species photosynthetic capacity is inversely re- that of Ledum,had an average leaf longevity lated to leaf longevity. As a result, even nearly twice as long. though photosynthetic capacity and leaf longevity patterns differed markedly between LEAF REPLACEMENTFOLLOWING DEFOLIATION growth forms, net carbon gain, when con- Recovery from a defoliation event depends sidered over the lifespan of a leaf, may be upon the reestablishment of a photosynthetic quite similar for these tundra species. Salix, surface. The potential for leaf replacement of with its high photosynthetic rate and syn- a given plant species is a function of and de- chronous leaf growth exploited the smallest, pendent upon the type of leaf-producing unit but most favorable, portion of the arctic grow- the plant has evolved. Each Salix and Ledum ing season. Ledum,on the other hand, com- shrub was composed of woody stems and pensated for its low photosynthetic capacity by leaders that supported many leaf-producing exploiting the entire growing season and by buds. Each Eriophorumtussock and Carexstand

S. ARCHERAND L. L. TIESZEN/ 537 was composed of one or more rhizome systems temperature and moisture than to day length that gave rise to leaf-producingtillers. or light intensity. Such observations are consistent with those of Laude et al. (1961) and Shrubs Plumb (1961, 1963). Tinklin and Schwabe Buds on Salix and Ledum bushes were (1970), however, found that in Ribesshoot tip initiated in late June from predictable loca- removal early in the growing season, while tions on their stems. By diagrammatically new shoots were still elongating, resulted in mapping a stem with its associated leaders we the outgrowth of several lateral buds. Later in recorded the positions of leaf groups and dor- the season, removal of the shoot tips became mant buds. In Salix 96% of the buds produc- less and less effective as a means of inducing ing leaf tissue were located on 1-yr-old stems, bud break. 3% were initiated from 2-yr-old stems, and Buds on woody plants may be classified as about 1% from 3-yr-old stems. The average terminal, axillary, or adventitious (see Berg Salixplant, considered in its entirety, initiated and Plumb, 1972). Each of these may in turn growth in less than 20% of its visible buds (an be classified as short- or long-shoot buds. individual plant is defined as all aboveground Long-shoot buds give rise to long shoots biomass issuing from a stem which goes below whose leaves are separated by distinct inter- ground). The potential, then, for a defoliated nodes and bear lateral buds, some of which shrub to replace photosynthetic tissue lost to develop into more long shoots while others herbivores is great, since a large number of produce short shoots. Internode elongation is leaf-producingunits are available. essentially lacking in short shoots, and lateral For a shrub to replace photosynthetic tissue buds on short shoots are either lacking or de- following a grazing event, however, the velop into flowers (Dahl and Hyder, 1977). mechanism or mechanisms regulating apical Terminal buds in Salix and Ledumwere of the dominance of a given stem or leader must be long shoot type and were responsible for 96 overcome. Bud initiation in Salix and Ledum and 90%, respectively, of the subsequent sea- was synchronous rather than periodic or se- son's leaf production. Axillary buds were quential. Undisturbed shrubs initiated buds in formed at the base of the leaf petiole soon after late June and all buds on the plant producing each leaf expanded. Typically, in Salix, five to current stem and leaf tissue had done so by six leaves were produced per bud initiated early July. Removing leaf tissue, while leav- (Johnson and Tieszen, 1976; Archer and ing stems and buds intact (a simulation of leaf- Tieszen, unpublished); hence five to six axil- chewing by some insects and rodents) did not lary buds were generated. On the average alter this pattern. That is, no additional buds about 40 % of these axillary buds (typically the were activated to replace the leaf tissue lost by two or three most distal on the stem) de- this type of defoliation. When the terminal 1- veloped into leaf and stem primordia in the and 2-yr-old stems were removed along with subsequent growing season. The remaining current growth, apical dominance of the axillary buds retained their position outside leader was overcome and buds lower on the the cambium during secondary growth and stem were activated. Leaf replacement was either remained as visible lateral buds or be- highly variable, and a minimum of 14 to 21 d came buried in the bark (as described by usually elapsed before leaf replacement was Church and Goodman, 1966, for Acer).Those initiated. Because of the lag time, shrubs de- lateral buds buried in the bark have been foliated in mid- to late July regeneratedphoto- termed suppressed buds (Kormanik and synthetic tissue too late to recoup the energy Brown, 1969) and are apparent only after an investment before the end of the season. anatomical examination of the stem. Adventi- The release of growth of lateral buds was tious buds, on the other hand, form irregu- complicated and may have been dependent on larly in the cambium of older plant parts environmental history, past growth perform- rather than in leaf axils near an apical meri- ance, or nutrient reserves. For example, our stem (Kozlowski, 1971). All of these bud types defoliation treatments resulted in extensive played an important, and quite different, role lateral bud growth in a cool and moist year in the resprouting of Salixand Ledumfollowing but did not in a subsequent warmer and drier defoliation. year. Thus it appears that the release of lateral Two defoliation treatments, one involving buds in these shrubs may respond more to removal of current growth only and the other

538 / ARCTIC AND ALPINE RESEARCH involving removal of all biomass back to 3-yr- the long and short shoot variety. Our observa- old stems, elicited quite similar morphological tions on Salixpulchra are similar to those made responses in the two life forms. In both Salix by Wilson (1970) on other shrub species. and Ledum lateral buds were released from The minimum damage to the plant needed dormancy following defoliation and new shoot to stimulate these highly productive, long- growth was initiated, but only after a highly shoot buds appears to be greater than that im- variable lag time. In both life forms the acti- posed by typical browsing activities. Vigorous vated lateral buds were predominantly the long shoots may be produced in response to short-shoot types. Those few that were of the trampling, especially during the winter and long-shoot type were weak, and survivorship early spring when stems are stiff and brittle. of the new shoots with next year's potential In general Salixcan readily transform short- leaf groups was quite low. Leaves associated shoot buds into long-shoot status when long with the activated buds were typically small shoots are removed (Dahl and Hyder, 1977), with a much reduced mass and surface area. but this was not a strong tendency in Salix In Salixthe lateral buds most often activated pulchraat Atkasook. Most buds activated after following defoliation were those positioned long-shoot removal were the short-shoot type; nearest the point of defoliation. Thus, when the long-shoot type were weak and generally only current growth was removed, the lateral winter-killed. buds activated were those on 1- and 2-yr-old The production of short shoots following stems; when growth was removed back to 3- defoliation (especially late-season defoliation) yr-old stems, the lateral buds on the 3- and 4- may be advantageous because they can com- yr-old stems were most often activated. plete elongation in half the time required by a Ledum, in contrast to Salix, initiated nearly long shoot (Kozlowski, 1971). This rapid 50% of its new shoots from main-stem tissue elongation of a short shoot may enable the (stems greater than approximately 5 to 7 yr shrub to exploit a longer portion of the re- old) regardless of the type of defoliation treat- maining growing season while requiring less ment. As with Salix nearly all new shoots were carbohydrate and nutrient investment than a the short-shootvariety. long shoot. In addition, Zimmerman and While removal of plant parts up to 2 yr old Brown (1971) have shown that downward led to the activation of mainly short-shoot, translocationof photosynthate in many woody lateral buds, the removal of tissue back to 5- to plants occurs only after shoot elongation has 7-yr-old stems resulted in the activation of been completed. Although we have no data re- suppressed lateral buds and/or adventitious garding root responses to top removal in buds which were buried in the cambium. It is shrubs, it is possible that the preferential known that stump sprouts may arise from growth of short shoots, with their more rapid either suppressed lateral buds or adventitious completion of elongation, facilitated the rees- buds (Zimmerman and Brown, 1971). Root tablishment of a critical root-to-shoot ratio sprouts, on the other hand, appear to always needed to maintain root respiration and nu- arise from adventitious buds (Kozlowski, trient uptake. 1971). While anatomical differences do exist between adventitious and suppressed buds Graminoids (Dahl and Hyder, 1977), we did not distin- Graminoid tillers exserted leaves sequen- guish between them for sprouts arising from tially and at staggered intervals over the the basal portions of defoliated plants. In course of the growing season. Leaves grew Ledum these activated buds were the short- from the base of the tiller and telescoped up- shoot variety. Salix, however, responded by wards until mature (see Youngner, 1972). By initiating long shoots as well as short shoots. exposing less than 100% of the leaf tissue pro- Although we have no quantitative data, our duced throughout the growing season, a field observations indicate that these basal, smaller investment of nutrients and energy is long-shoot buds produced current leaf, and vulnerable to loss by grazing at any one time. especially stem, tissue far in excess of the Following defoliation, leaf replacement can be growth observed in terminal long shoots rapid. In fact CO2 studies in the laboratory under normal conditions for one season. The with other arctic graminoids indicate that vigorous stems arising from the suppressed within 3 d of defoliation sufficient leaf area buds contained axillary buds, which were both can be restored so that net photosynthesis for

S. ARCHER AND L. L. TIESZEN / 539 an entire tiller becomes positive (Kingston vegetative Carex and Eriophorumtillers to con- and Tieszen, unpublished). These observa- tinue to produce leaf tissue regardless of the tions are similar to those made on temperate, time of season of defoliation. pasture species (Ward and Blaser, 1961; In summary, the efficiency of the replace- Davidson and Milthorpe, 1965, 1966a, ment of photosynthetic tissue following a graz- 1966b; Smith, 1975). Whereas removal of or ing event is ultimately a function of (1) the damage to the apical meristematic tissue in type of leaf producing unit the plant has graminoids normally results in a cessation of evolved and (2) the type of defoliation event, leaf production and death of at least the i.e., whether or not the apical meristem is aboveground portions of a tiller (Hyder, damaged. Leaf replacement is further gov- 1972), this was seldom a problem with Carex erned by the frequency, intensity, and timing and Eriophorum tillers. Because their apical of the defoliation events and by those abiotic meristems are typically positioned 10 to 15 factors that govern growth processes. mm below a layer of moss and soil, these graminoids usually avoid such injuries. The SHRUB RESPONSESTO DEFOLIATION culmless growth habits of each of these sedges There are several options open to a de- insures that, unless they are flowering, the foliated shrub with regard to the reestablish- apical meristems will not be elevated above ment of a new photosynthetic surface: (1) acti- the moss level. Only in years of extreme lem- vate a "normal" complement of buds which ming population densities or heavy trampling yield a normal leaf area or biomass, (2) acti- by caribou or musk-oxen will the moss and vate more buds than normal with fewer leaves soil cover be sufficiently disturbed to expose or less leaf tissue per bud, (3) activate fewer the growing points. This strategic location of buds than normal with more and/or larger the growing point, in conjunction with the se- leaves per bud, (4) invest less in current stem quential pattern of leaf exsertion, enables production and proportionately more in leaf

TABLE 3 Locationof new growth (% of activatedbuds) on Salix pulchra plantsfollowing defoliationin earlyJuly

Stem age Type of Percentage of plants defoliation 1 yr 2 yr 3 yr 4 yr 5 yr >5 yr initiating regrowth Type Ia 57 23 20 0 0 0 60 Type IIb 43 33 14 10 83 Control (not defoliated) 96 3.3 0.4 0 0 0 aRemoval of current leaves and stems on 20 plants. bRemoval of current growth plus 1- and 2-yr-old stems on 20 plants.

TABLE 4 Mean (+ 1 SE) responsesof Salix pulchrafollowing defoliationin earlyJuly 1976

Current growth biomass (mg dry wt per leaf group) % buds 1976 1977 Treatment initiated August June August Control 18 ( 1.1) 54.0 (+ 0.9)a 18.5 ( 0.3)a 72.4 (+2.1)a Type I 13 (1.8)b 7.2(+ 1.6)b 12.6 ( 0.4)b 61.0 (1.6)b Type II 10(+ 2.9)b 5.8(+ 1.1)b 11.9 ( 0.2)b 59.8( 1.4)b

a'bDifferent superscripts indicate significantly different means among treatments in each column (t- tests, p <0.05). About 20 plants in each treatment.

540 / ARCTIC AND ALPINE RESEARCH production or, in the case of a partially de- trast, plants defoliated the previous growing foliated plant, (5) increase the photosynthetic season did not complete leaf initiation until 6 efficiency and/or extend the longevity of re- August. The number of leaf groups per plant maining leaves. was comparable in control and defoliated Replacement of leaf biomass following de- shrubs, but the current growth per plant foliation was highly variable in Salix and ap- throughout the growing season was signifi- peared to be greatly influenced by environ- cantly lower in treated plants due to the re- mental factors. In preliminary experiments duced weight of each leaf group (Table 4). concluded in 1976, Salix plants recovered 92 The mean number of leaves produced per bud and 64% of their current-growth biomass in was 5.0 in both control and recovery plants. Type I and Type II treatments (see Methods), However, current stem length decreased from respectively. However, in similar treatments 8.9 mm in controls to 4.1 mm in recovery at similar points in the growing season in plants, a reduction of 54 %. 1977, Type I and Type II defoliated plants re- In Type I treatments 55% of the 1-yr-old covered only 40 and 20% of their current- stems died completely, as did the terminal half growth biomass, respectively. Weather pat- of an additional 24%. Forty percent of the 2- terns of the two seasons were quite different, yr-old stems died completely along with the the 1976 season being cool and moist and the terminal portions of the remaining 60%. 1977 season being warmer and drier (Haugen Nineteen percent of the 3-yr-old stems died and Brown, 1980, this volume). The ensuing along with the terminal half of an additional discussion will deal primarily with growth re- 10%. In nearly every case leaders died back to sponses from the 1977 growing season for the location of the first visible bud. In 40 % of which there is a more substantial and detailed the leaders this resulted in death back to 3rd yr data base. stems and in 19% to 4th yr stems. Defoliated Salixplants, regardlessof time of Leaf production the following season was season, reinitiated growth from terminal posi- reduced by 47 to 72% in defoliated plants tions on the remaining stems. Following a (Table 5). With the exception of the 15 June Type I defoliation in early July, 57% of the defoliation there was no difference in leaf pro- buds initiated arose from positions on 1-yr-old duction between Type I and Type II defolia- stems; 23 and 20% arose from 2- and 3-yr-old tion treatments. Generally, those plants de- stems, respectively. No buds developed on foliated on or after 15 July produced slightly older stem tissue. Forty percent of the plants more current-growth biomass the subsequent examined did not initiate any leaf production season than plants defoliated on or before 30 after defoliation. Following a Type II defolia- June. tion 43% of the buds initiated were on 3-yr- Ledumpalustre ssp. decumbens,the evergreen old stems; 33 and 14% arose from 4- and 5-yr- shrub, responded even more dramatically to old stems, respectively; 10% arose from older defoliation (Table 6), recovering 10% or less stems. Approximately 17% of the Type II of its current-growth biomass. Plants sub- plants did not break bud following defoliation (Table 3). 0 Defoliated of leaf biomass in 100- E Control '/7/ Although the replacement ar Salix was highly variable between plants I~ within a treatment and between treatments, 0 I some differences and trends were 0 50 significant o r7/I I e-c noted. Plants to I and II de- // subjected Type C, llz foliations responded similarly (Table 4). In 0 I both defoliation treatments a slightly lower ( '/"' Z/Z S r a X Z 7 17 percentage of buds initiated growth relative to July July 27July 6August control plants, and the leaf biomass per bud Date initiated was substantially lower than that of FIGURE 4. Seasonal course of leaf initiation in Salix controls. pulchra control plants (untreated) and plants de- Leaf initiation was delayed the following foliated the previous growing season. Percentage is spring in treated plants by up to 20 d (Figure based on the total number of buds producing in 4). All buds on control plants that were to leaves by the end of the growing season 20 initiate growth had done so by 17 July. In con- plants.

S. ARCHER AND L. L. TIESZEN / 541 jected to the Type II defoliation produced TABLE 5 slightly, but significantly, more leaf biomass Theeffect of timeof defoliationon mean than those experiencing a Type I defoliation. ( ? 1 SE) leafbiomass produced thefollowing In both Type I and Type II treatments, the seasonfor Salix pulchra greatest impacts occurred on plants defoliated prior to 1 July. More than 48% of the leaf Dry weight (mg) per leaf initiated defoliation were Date of groups following defoliations group (earlyJuly 1977) from buds on the main stem in a Type I de- Type I Type II foliation, as compared to 30% in a Type II de- (1976) foliation, and 11% on undefoliated plants. 15June 5.1 (+0.5)a 8.8 ( ? 1.1l)a,b After the second recovery season this percent- 30 June 7.2 ( ?0.2)b 7.2 ( 0.2)a age increased to 82 in Type I plants, one plant 15July 9.4( +0.6)c 9.4 ( + 0.4)b of the 20 had died, and leaf production in sur- 20 July 8.8(?0.3)c 9.5 ( + 0.3)b viving plants remained substantially below 15 August 8.6 (+0.4)c 9.9 ( + 0.4)b that of controls (Table 6). Controls 18.5 ( + 0.3)d 18.5 ( ? 0.3) For Ledumthe contribution of older leaves a to defoliation must also b,c,dDifferent superscripts indicate significantly plant growth following different means treatments in each column be considered. Hadley and Bliss (1964) found among that in several ericaceous leaves (t-tests, p<0.05). About 20 plants in each treat- species, pro- ment. duced in previous growing seasons had negative net photosynthetic rates, and they concluded that old leaves were more important as storage units than as photosynthetic units. Ledum to conserve and cycle nutrients However, subsequent work by Johansson and (Thomas and Grigal, 1976; Reader, 1978). Linder (1975) and Johnson and Tieszen In summary, it appears that none of the hy- (1976) has shown positive photosynthetic rates pothesized options for reduction of damage in older leaves of several ericaceous species. that were listed earlier were employed by arc- These positive photosynthetic rates suggest tic shrubs. Generally, a lower percentage of that old leaves may be important in carbon as- buds was initiated following defoliation. similation. Whatever their function, leaves Further, the initiation of these buds was de- produced in previous years on Ledumshrubs layed, and the biomass produced from them were important for the exsertion and growth was substantially lower than that of control of current leaves and stems. When old leaves plants. Although fewer resources were in- were removed and terminal buds left intact, vested in current stem production there was development of that year's activated buds was no apparent or obvious reallocation to leaves. delayed, and the biomass of the current sea- The number of leaves per leaf group remained son's leaves was reduced by 60%. These ob- unchanged following defoliation, and while servations on Ledum palustre ssp. decumbens photosynthetic rates of remaining or replace- agree with those made by Reader (1978) on L. ment leaves were not monitored, no extended palustressp. groenlandicumin Canada. The loss rejuvenation or postponement of senescence of old leaves by defoliation altered not only the was observed. Thus both shrub life forms carbon balance of the plant, by decreasing net were substantially damaged by defoliation. carbon assimilation, but also the ability of The effect on carbon balance would be differ-

TABLE 6 Mean ( + 1 SE) biomassof currentgrowth in Ledum palustre ssp. decumbens afterone and two seasonsof recoveryfromdefoliation Seasons after leaf initial defoliation Dry weight (mg) per group (dates measured)a Type I Type II Controls One (late August 1976) 2.0 (?0.2) 3.1 (?0.5) 31.9(+ 4.2) Two (late August 1977) 3.2 ( 0.6)

aPlants defoliated in earlyJuly 1976. About 20 plants in each treatment.

542 / ARCTIC AND ALPINE RESEARCH ent for each growth form since their patterns tomical configuration is similar to that of other of CO2 uptake differ (Johnson and Tieszen, graminoids. A single defoliation, removing 1976; Tieszen et al., in press). Clipping of leaves to moss level, stimulated leaf produc- dwarf evergreen shrubs was most harmful be- tion in Eriophorumduring the season of clip- cause of their slow replacement of leaves. Al- ping (Figure 6). Changes in dry weights of though our data did not reflect any strong sea- tiller components, indicating an expenditure sonal trends, Reader (1978) found that for of TNC reserves (Hickman and Pitelka, 1975) Ledumpalustre ssp. groenlandicumand Kalmia suggest that Eriophorumrelied primarily upon polifolia effects were reduced by postponing the sheath component for regrowth during the defoliation treatments from the end of the first 10 d following leaf removal. Thereafter, leaves' first growing season to the middle of reserves from stem bases were mobilized to their second growing season. Grazing of Salix supplement sheath reserves (Figure 6). Dry leaves may be especially damaging during weight losses of nonleaf plant parts reached a early and mid-season because leaf production maximum at 20 d following defoliation. Stor- is synchronized and because new sets of leaves age structures then began to reaccumulate are produced too late in the season to pay for weight, indicating that sufficient leaf tissue their investment. Birch (Betulapubescens), an- had been exserted to establish a positive car- other common arctic deciduous shrub, ap- bon balance. parently responds similarly to defoliation and Others have also found that light grazing its ability to survive seems to depend upon can stimulate production of leaves (Mc- large reserves gathered during warm years Naughton, 1976). The increased production (Kallio and Lehtonen, 1975). While we have noted in our experiments could be attributed no data on root and stem growth, it is likely to canopy effects, for the photosynthetic rates that defoliation also affects these growth of expanding leaves can increase if the light in- parameters adversely (Kramer and Wetmore, tensity reaching those leaves increases 1943; O'Neil, 1962; Kulman, 1971). Our data (Woledge, 1977). Also, a simulation model of indicate, then, that these species are highly in- physical processes affecting primary produc- tolerant of grazing, especially in the short arc- tion in tundra (Miller and Tieszen, 1972) pre- tic growing seasons. dicted that periodic reduction of standing dead shoots by grazers would increase plant GRAMINOIDRESPONSES TO DEFOLIATION production. Grazing experiments at Barrow, Alaska The accelerated growth of leaf tissue follow- (Mattheis et al., 1976), showed clearly that ing defoliation (Figure 6) can be attributed to Dupontiafisheri,a single-shooted arctic grami- a delayedsenescence of older leaves (Archerand noid, was quite tolerant of chronic grazing Tieszen, submitted, b). Such a strategy would pressure. Its high degree of tolerance resulted be important in the recovery of lost assimilate from two related factors: (1) an interdepend- following grazing, especially if the decline in ence between tillers which resulted in a flow of photosynthetic rate commonly associated with nutrients and photosynthate from intact tillers leaf aging (Leopold, 1961; Jewiss and to defoliated tillers via rhizomatous connec- Woledge, 1967; Ludlow and Wilson, 1971) tions and (2) high concentrations and large were delayed or postponed. In other plants de- standing crops of total nonstructuralcarbohy- foliation has been shown to prevent the nor- drates (TNC) which buffer the plant against mal decline in photosynthesis associated with acute and chronic defoliation (see White, leaf age (Maggs, 1964; Hopkinson, 1966; 1973; Trlica and Singh, 1979). In an exten- Gifford and Marshall, 1973) and to rejuvenate sion of these preliminary studies we followed photosynthesis in senescing leaves to rates leaf growth and tiller biomass following de- comparable to those of recently expanded new foliation in Eriophorumvaginatum and Carex leaves (Woolhouse, 1967; Hodgkinson et al., aquatilis.We will summarize here our under- 1972; Hodgkinson, 1974). The preferential standing of Eriophorum,the tussock-forming growth and maintenance of older leaves might cottongrass of the Arctic. benefit the damaged tiller in several ways: (1) Figure 5 depicts the general tussock and Rejuvenation of old leaves may be more eco- tiller pattern characteristicof E. vaginatum.Al- nomical than biosynthesis of new leaf mate- though rhizomes are very short, thereby ac- rial. (2) Maintenance of older, perhaps less counting for the tussock form, its basic ana- palatable leaves, may enable the plant to avoid

S. ARCHERAND L. L. TIESZEN/ 543 and/or minimize the impact of future grazing trients that are important in growth and ex- episodes (Kamstra et al., 1968; Smith and pansion of younger leaves, especially when Neales, 1977). (3) Older leaves appear to be nutrient levels are low (Davidson and Mil- the primary source of assimilates for below- thorpe, 1966a). ground structures such as roots (Williams, Cessation of root growth and substantialde- 1964; Wardlaw, 1968) and as such may be increases in root respiration and nutrient uptake strumental in overcoming the suppression of rates following defoliation have been demon- root extension and nutrient uptake observed strated in several glasshouse studies (Jameson, in most defoliation studies (Evans, 1972; Bok- 1963; Davidson and Milthorpe, 1966b; hari, 1977). (4) Fully expanded leaves have Evans, 1972; Bokhari, 1977) and have been been shown to serve as reservoirs of labile nu- interpreted as a consequence of a shift in allo-

FIGUIRE5. Diagrammatic representation of an Eriophorumvaginatum tussock and tillers. Apical meristems were protected by a moss layer 8 to 15 mm thick (a). Stem bases formed a nearly continuous layer at the moss-mineral soil interface (b). Depths indicate the regions where numbers of roots were estimated.

544 / ARCTIC AND ALPINE RESEARCH 'E o ..:~~ \C 13 E 0.60

12 O 0.55 -_CD 11 S 0.50

10 E* 0.45 aX Defoliated 9 L 0.40 Q) E 0?" Control 0- 8 ' 0.35 - 0) Roots 7 8O 0.30

40 d E 10o 35 U 8 LieDefoliated Leaves / 30 E 6 0 , ~^ Control 25 m 4 U/, 4) 20 2

5 10 20 25 30 Days 0 5 10 15 20 25 30 E 0 5 10 15 20 25 30 Days 0 July 1 July 31 July 1 July 31 Date (1977) Time

FIGURE 6. Short-term changes in mean ? 1 SE dry weights of stem base, sheath, and leaf compartments and weight per unit length of roots of Eriophorumfollowing clipping on 1 July (N = 80 for sheaths and stem bases; N = 50 for roots; N = 20 for leaves).

TABLE 7 Summaryof growthform characteristicsrelated to herbivory

Ability to Probability recover Amount of Photosynthetic Leaf of being Principal from secondary Growth form rate longevity eatena herbivores defoliation compounds Graminoid Single-shooted High Medium High Rodents and High Low Tussock-forming Medium Medium High ungulates High Low Deciduous Shrub High Short Medium Ungulates, Medium Medium insects, and rodents Evergreen shrub Low Long Low None Low High Forbs Medium Medium (?) Medium Rodents and Medium (?) Medium insects

aProbabilitybased on herbivore preferences.

cation of reserves to shoot growth and restora- the findings of Chapin and Slack (1979) who tion of a favorable shoot-to-root ratio. Our re- also observed an increase in weight specific sults showed that weight per unit length re- respiration and nutrient uptake in Eriophorum mained the same in Eriophorumroots following roots following top removal. Only after four a single defoliation. These results agree with defoliations, spaced at the 10-d intervals, did

S. ARCHER AND L. L. TIESZEN / 545 nutrient uptake rates decrease to below con- until the stem base had lost some 75% of its trol levels. This maintenance and stimulation normal weight in the most severe treatment of root activity is especially important in the (black bag). By this time the sheath had died Arctic where leaf growth following defoliation and leaf production ceased. However, even depletes nutrient reserves more than carbohydrate reserves a (Chapin, 1977). 75 +f- While the response of Eriophorumto a single 70 defoliation was increased leaf production at 65 Total the expense of belowground structures, mul- 60 tiple defoliation imposed at 10-d intervals for 55 up to two growing seasons resulted in de- 50 I- creased leaf production, further weight loss in 45 ? storage structures, and a curtailment of root 40 35 growth. Leaf growth response during the first 4- season of chronic defoliation was similar to 30 + 25 that of a single defoliation. During the subse- 4- 20 quent growing season, however, leaf length and were to to 15 weight depressed markedly 25 10 of 50% control values (Figure 7), depending 5 upon the date clipping was initiated. The decline in b leaf production was accompanied by 14 . -+ . substantial weight losses in stem base and 12 Leaves + sheath components (Figure 8c,d) and a severe 10 curtailment in root initiation and 8 + I_ _ I- elongation E (Figure 9). Overall weight losses continued 6 4

0 c FALL 18 (a) f 160 . -^ ! CONTROL 16

E 140 SUMMER 14 E Sheaths ^- SPRING 12 120 / /-/^--O ~LATE

EARLY SPRING 10 w100 8 80 6 +] 60 4 2 im ' d I I l 1 l I I I I I I 40 4- 35' (b) CONTROL t 18 30 Stem Bases _ 16 25 . E v 14 20 -

|12 FALL 15 -

10 SUMMER 10 U 20 5 SPRING i nI I i6 Black Fall Fall Control I Early Chronic Early Treatment: Bag Spring Chronic U 4 (1977) SpringRecovery Chronic Recovery 2 No. ofReclips: (11) (11) (8) (6) (5) (2) (0) I I I I I I I I I f 18 28 8 28 7 FIGURE8. Mean biomass ? 1 SE of total JUNE JULY AUGUST (a) tiller

DATE (1977) (not including roots), (b) leaf, (c) sheath, and (d) stem base compartments of Eriophorum vaginatum FIGURE 7. Effect of time of initiation of clipping on following various experimental clipping treatments (a) cumulative leaf length and (b) cumulative leaf which extended through two growing seasons. See weight in Eriophorum during the second season of text for details. N = 20 for leaf compartment and chronic clipping at 10-d intervals (N = 20). 80 for sheath and stem base compartments.

546 / ARCTIC AND ALPINE RESEARCH after complete defoliation at 10-d intervals for two full growing seasons (11 reclips) Erio- phorum stem bases were still 34% above this minimum level. Tillers defoliated for one and one-half and one season were 46 and 52% above this minimum. Note, however, that one full season of recovery was insufficient to re- store leaf to control levels growth (Figure 8). 3 75- Initiation of vegetatively produced daugh- 0 ter tillers was not significantly depressed by up to 2 yr of chronic defoliation. The weight of E 50- 0 each daughter tiller produced was, however, z reduced to 25% that of controls. Such low 3 ,. 3 tiller biomass suggests that the success of vege- 25- tative reproduction following chronic defolia- 4 tion might be quite low. Our results also sug- that under extreme n n gest, grazing pressure, 0 Ist Season 2nd Season populations of E. vaginatummay experience an indirect loss of sexual reproduction due to Leaf MaterialRegrown Following Defoliation lowered and a of stored plant vigor depletion FIGURE9. Effect of various treatments on and nutrient reserves. In this clipping energy regard mean + 1 SE number of roots initiated and Tikhomirov and Smirnov and Tokma- (a) (1959) number of roots to of 10 and kova extending depths (b) (1971) observed that lemming grazing 25 cm in the soil beneath an tussock. in several arctic (c) Eriophorum suppressed flowering grami- See text for details. Four tussocks per treatment noids. Direct losses of reproductive tissue due were excavated for the determinations at 10 and 25 to preferential grazing may also occur. Spetz- cm; 40 tillers per treatment were examined for root man (1959) and Kelsall (1968) noted that E. initiation. vaginatum was utilized by caribou, especially during the flowering period (see also White treated in the fall were only clipped twice. and Trudell, 1980, this volume), and Pearsall Late-season defoliations occurred at a time (1950) mentioned that deer and sheep graze when arctic sun angles and ambient tempera- young flowering shoots of Eriophorum. In an tures were decreasing and soil temperatures environment such as the Arctic, where suc- were at their highest. Decreased light intensity cessful sexual reproduction is already limited, and air temperatures reduce photosynthesis, such losses may be significant. while higher soil temperatures simultaneously Late-season defoliations appeared to be increase respiration of belowground struc- more detrimental to leaf production in subse- tures. In conjunction with the loss of carbohy- quent seasons than did early-season defolia- drates and nutrients removed with the leaves, tions. For example, the recovery from an which normally would be translocated below- early-season treatment was nearly that of the ground prior to dieback in the fall (Chapin et late-season treatment (Figure 8) despite the al., 1975), such circumstances may lead to fact that the plants starting treatment in early tillers entering winter dormancy with reduced spring were clipped five times and the plants carbohydrate and nutrient levels.

COMPARISONS AND CONCLUSIONS

Our results generally support the hypothe- mediate in Salix, the deciduous shrub. The ses stated in the Introduction. Perhaps the same trends held true for the subsequent sea- best comparison of the response of different son. Neither of the shrubs approached normal growth forms to defoliation is to compare the foliage production during the second recovery amount of leaf tissue produced by intact plants season. Poor replacement of photosynthetic after a single, early season clipping (Figure tissue in Salix was due largely to a decreased 10). Leaf replacement was lowest in Ledum, investment in biomass per leaf group or bud the evergreen shrub; highest in Eriophorumand rather than to a decrease in the number of Carex, the perennial graminoids; and inter- buds initiating growth. Although this type of

S. ARCHER AND L. L. TIESZEN / 547 data was not available for Ledum, observations 4.5 suggested that here too, decreased production a -- 4.0 was the result of a decrease in biomass per leaf " Depth: 0 cm group more than a reduction in the number of 3.5 leaf groups initiated. Leaf production in Carex 3.0 ^- if- -- was nearly normal during the first season 2.5 I .4 + (93 %) and dropped to 80% of normal during 2.0 the subsequent season. A nearly 25 % increase 1.5 in leaf production was observed in Eriophorum 1.0 during the first season owing to the preferen- 0.5 tial growth and maintenance of older leaves, the opening of the normally dense canopy 2.0 z b (Archer and Tieszen, submitted, b), and in- 1.5 10 cm - creased root respiration and nutrient uptake 1.0 Depth: following a single defoliation (Chapin and 0.5 Slack, 1979). During the following growing season foliage production was normal. 1.0

A summary of characteristics of the various 0.5 Depth: 25cmn ] forms and their to herbi- growth relationship Erly-Fal onc-ary-al-CntoCr is in Table 7. The Early Fall Chronic Early Fall Control vory given synchronous Treatment: exhibited shrubs enabled Spring Chronic 1977 SpringRecovery growth strategy by Chronic Recovery them to more the most favorable exploit fully No. of Reclips: (11) (8) (6) (5) (2) (0) portion of the arctic growing season. Such a strategy, however, exposes the total season's FIGURE10. Leaf material regrown during the first complement of photosynthetic tissue to loss by and second season following a single clipping of 20 grazing. Ledum, the evergreen shrub, had the plants in late June 1976. (1) Eriophorumvaginatum, lowest photosynthetic rate (Table 1) of the life (2) Carexaquatilis, (3) Salix pulchra,and (4) Ledum forms examined and was evidently dependent palustre. upon an extended leaf longevity to secure a positive carbon balance. The retention of leaves for more than one growing season likely requires the synthesis of compounds to protect erences to browsing by caribou and muskox them from herbivores (Cates and Orians, on Ledum and other evergreen shrubs, these 1975) and from overwintering stress (Mooney plants rarely occur in the diets of arctic herbi- and Dunn, 1970; Small, 1972). Even though vores near Atkasook (Batzli and Jung, 1980, these chemicals may represent a considerable this volume; Batzli and Sobaski, 1980, this expenditure, they enhance the competitive volume; White and Trudell, 1980, this abilities of the plant by reducing loss of photo- volume; MacLean, pers. comm., 1980). synthetic tissue to herbivores (Cates, 1975). Thomas and Grigal (1976) estimated that the These factors, in addition to the large con- evergreen Kalmia lost less than 1.5% of its sumption of assimilates necessary for year- total leaf area'to herbivores. Reader (1978) round maintenance (Penning deVries, 1975), examined Ledumpalustre ssp. groenlandicumand probably result in high growth costs for ever- two other arctic and subarctic ericads and esti- green leaves and may result in fewer resources mated that these plants had lost less than 5% available to invest in replacing tissue lost to an of their biomass to herbivores. He also ob- herbivore. Jung et al. (1979) have reviewed served that among the three ericads in his and described the array of antiherbivore com- study, the retention time for overwintering pounds present in arctic growth forms and leaves depended on their susceptibility to at- found that those evergreen shrubs which were tack by herbivorous insects. Species with the unpalatable to all rodents (Ledum, Empetrum, shortest leaf longevity lost more biomass to in- and Cassiope) contained the greatest variety of sect herbivores, but at the same time were chemical compounds. Ledum, then, appears to least affected by defoliation treatments. Thus, have evolved to avoid grazing rather than to by allocating carbon to support leaf mainte- tolerate it. nance and preservation, Ledum essentially Although Kelsall (1968) makes several ref- avoids growth problems associated with the

548 / ARCTIC AND ALPINE RESEARCH regeneration of leaf tissue following defolia- Salix, it has a much higher probability of being tion. grazed than Ledum. Salix does, however, have In contrast to the evergreen shrubs, de- high levels of secondary compounds relative to ciduous shrubs had photosynthetic capacities the graminoid growth forms (Jung et al., two to four times higher (Table 1). However, 1979). these high carbon fixation rates in deciduous Patterns of photosynthetic rates and leaf shrubs were offset by relatively short leaf lon- longevities observed in graminoids were inter- gevities (Table 2). Such leaves contribute to mediate to those of the evergreen and de- the net carbon balance of the plant in only one ciduous shrubs. Because of their large nu- growing season, but the synchronous pattern trient reserves and the telescopic exsertion of of leaf exsertion still exposes the entire season's leaf tissue from a protected apical meristem, photosynthetic tissue to potential loss to herbi- Carex and Eriophorum were able to respond vores. Such a loss of leaf tissue represents not rapidly to defoliation by replacing the lost tis- only a loss of photosynthetic input but also sue. By initiating leaf growth early in the sea- substantial amounts of nutrients, especially son graminoids were able to exploit an earlier nitrogen (Chapin, 1980, this volume). Salix portion of the growing season than were de- has apparently sacrificed leaf longevity, avoid- ciduous shrubs. The monocotyledons at Atka- ance of herbivores, and leaf replacement sook had the least amount of antiherbivore potential for larger photosynthetic returns. compounds of all the growth forms (Jung et While Salix was decidedly more tolerant than al., 1979) but compensated for this lack of Ledum to defoliation it was much less so than chemical defense by morphological and were the graminoid growth forms. And, while physiological mechanisms associated with the rate of carbon incorporation was highest in tolerance to grazing.

ACKNOWLEDGMENTS

We thank F. S. Chapin and D. Johnson for ments (RATE) and was funded by National active collaboration and G. Archer, M. Slack, Science Foundation grant DPP 76-80647 to and D. Smith for field assistance. Our re- Augustana College. Logistic support was pro- search was part of a coordinated program vided by the Naval Arctic Research Labora- called Research in Arctic Tundra Environ- tory, Barrow, Alaska.

REFERENCES CITED abundance, and foraging patterns of ground Archer, S. and Tieszen, L. L., submitted, a: The squirrels near Atkasook, Alaska. Arcticand Al- effect of chronic defoliation on foliage and root pine Research, 12: 501-510. production and biomass allocation in an Arctic Berg, A. R. and Plumb, T. R., 1972: Bud activa- tundra sedge (Eriophorum vaginatum). (Sub- tion for regrowth. In McKell, C. M., Blaisdell, mitted to Oecologia.) J. P., and Goodin, J. R. (eds.), WildlifeShrubs. Archer, S. and Tieszen, L. L., submitted, b: Their Biology and Utilization. U.S. Department of Growth responses of tundra cottongrass (Erio- Agriculture, Forest Service, General Technical phorumvaginatum) to defoliation. (Submitted to Report, INT-1, 279-286. Oecologia.) Bliss, L. C., 1962: Adaptation of arctic and alpine Batzli, G. 0., 1975: The role of small mammals in plants to environmental conditions. Arctic, 15: arctic ecosystems. In Golley, F. B., Petruse- 117-144. wicz, K., and Ryskowski, L. (eds.), SmallMam- , 1975: Tundra grasslands, herblands, and mals. Their Productivity and Population Dynamics. shrublands, and the role of herbivores. Geosci- International Biological Programme 5. Cam- enceandMan, 10: 51-79. bridge: Cambridge University Press, 243-268. Bokhari, U. G., 1977: Regrowth of western wheat- Batzli, G. 0. and Brown, J., 1976: RATE-The grass utilizing '4C-labeled assimilates stored in influence of grazing on arctic tundra ecosys- belowground parts. Plantand Soil, 48: 115-127. tems. Arctic Bulletin, 2: 153-160. Cates, R. G., 1975: The interface between slugs Batzli, G. 0. and Jung, H. G., 1980: Nutritional and wild ginger: Some evolutionary aspects. ecology of microtine rodents: resource utiliza- Ecology, 56: 391-400. tion near Atkasook, Alaska. Arctic and Alpine Re- Cates, R. G. and Orians, G. H., 1975: Succes- search,12: 483-499. sional status and the palatibility of plants to Batzli, G. 0. and Sobaski, S., 1980: Distribution, generalized herbivores. Ecology,56: 410-418.

S. ARCHERAND L. L. TIESZEN/ 549 Chapin, F. S., III, 1977: Nutrient/carbon costs as- 347-355. sociated with tundra adaptations to a cold nu- Gifford, R. M. and Marshall, C., 1973: Photosyn- trient-poor environment. In: Proceedings of the thesis and assimilate distribution in Lolium CircumpolarConference on NorthernEcology, Ottawa multiflorumLam. following differential tiller de- 1975. Ottawa: National Research Council of foliation. Australian Journal of Biological Science, Canada, Scientific Committee on Problems of 26: 517-526. the Environment, I-183-I-194. Goodman, G. T. and Perkins, D. F., 1968: The , 1980: Nutrient allocation and responses to role of mineral nutrients in Eriophorumcom- defoliation in tundra plants. Arctic and Alpine Re- munities. III. Growth response to added inor- search,12: 553-563. ganic elements in two E. uaginatumcommuni- Chapin, F. S., III and Slack, M., 1979: Effect of ties. Journal of Ecology, 56: 667-683. defoliation upon root growth, phosphate ab- Hadley, E. B. and Bliss, L. C., 1964: Energy rela- sorption and respiration in nutrient-limited tionships of alpine plants on Mt. Washington, tundra graminoids. Oecologia,42: 67-79. New Hampshire. Ecological Monographs, 34: Chapin, F. S., III, Van Cleve, K., and Chapin, 331-357. M. C., 1979: Soil temperature and nutrient Haugen, R. K. and Brown, J., 1980: Coastal-in- cycling in the tussock growth form of Eriophorum land distributions of summer air temperature vaginatum L. Journal of Ecology, 67: 169-189. and precipitation in northern Alaska. Arcticand Chapin, F. S., III, Van Cleve, L., and Tieszen, L. Alpine Research, 12: 401-412. L., 1975: Seasonal nutrient dynamics of tundra Hedberg, 0., 1964: Features of afroalpine plant vegetation at Barrow, Alaska. Arcticand Alpine ecology. Acta PhytogeographicaSuecia, 49: 1-144. Research, 7: 209-226. Hickman, J. C. and Pitelka, L. F., 1975: Dry Church, T. W., Jr. and Goodman, R. M., 1966: weight indicates energy allocation in ecological The formation and development of dormant strategy of plants. Oecologia,21: 117-121. buds in sugar maple. ForestScience, 12: 301-306. Hodgkinson, K. C., 1974: Influence of partial de- Costin, A. B., 1967: Alpine ecosystems of the Aus- foliation on photosynthesis, photorespiration, tralasian regions. In Wright, H. E. and Os- and transpiration by lucerne leaves of different born, W. H. (eds.), Arctic and Alpine Environ- ages. AustralianJournal of Plant Physiology, 1: 561- ments.Bloomington: Indiana University Press, 578. 55-87. Hodgkinson, K. C., Smith, N. G., and Miles, G. Dahl, B. E. and Hyder, D. N., 1977: Develop- E., 1972: The photosynthetic capacity of mental morphology and management implica- stubble leaves and their contribution to growth tions. In Sosebee, R. E. (ed.), RangelandPlant of the lucerne plant after high level cutting. Aus- Physiology.Range Science Series No. 4. Denver: tralian Journal of Agricultural Research, 23: 225- Society for Range Management, 257-290. 238. Dahl, E., 1975: Stability of tundra ecosystems in Hoffmann, R. S. and Taber, R. 0., 1967: Origin Fennoscandia. In Wielgolaski, F. E. (ed.), and history of holarctic tundra ecosystems, with Fennoscandian Tundra Ecosystems. Part II. Animals special reference to their vertebrate faunas. In and Systems Analysis. New York: Springer- Wright, H. E. and Osborn, W. H. (eds.), Arc- Verlag, 231-236. tic and Alpine Environments. Bloomington: In- Davidson, J. L. and Milthorpe, F. L., 1965: Car- diana University Press, 143-170. bohydrate reserves in the regrowth of cocksfoot Hopkinson, J. M., 1966: Studies on the expansion (Dactylis glomerata L.). Journal of the British Grass- of the leaf surface. VI. Senescence and the use- land Society, 20: 15-18. fulness of old leaves. Journal of Experimental , 1966a: Leaf growth in Dactylis glomerata Botany,17: 762-770. following defoliation. Annalsof Botany,N.S., 30: Hulten, E., 1968: Flora of Alaska and Neighboring 173-184. Territories.Stanford, California: Stanford Uni- , 1966b: The effect of defoliation on the car- versity Press. 1008 pp. bon balance in Dactylis glomerata. Annals of Hyder, D. N., 1972: Defoliation in relation to Botany,N.S., 30: 185-198. vegetative growth. In Youngner, V. B. and Dyer, M. I. and Bokhari, U. G., 1976: Plant-ani- McKell, C. M. (eds.), The Biology and Utilization mal interactions: studies of the effects of grass- of Grasses.New York: Academic Press, 304-317. hopper grazing on blue grama grass. Ecology, Jameson, D. A., 1963: Responses of individual 57: 762-772. plants to harvesting. BotanicalReview, 29: 532- Evans, P. S., 1972: Root growth of Loliumperenne 594. L. III. Investigations of the mechanisms of de- Jewiss, O. R. and Woledge,J., 1967: The effect of foliation-induced suppression of elongation. age on the rate of apparent photosynthesis in New Zealand Journal of Agricultural Research, 15: leaves of tall fescue (Festuca arundinaceaeSchreb.)

550 / ARCTIC AND ALPINE RESEARCH Annals ofBotany, N.S., 31: 661-671. Ludlow, M. M. and Wilson, G. L., 1971: Photo- Johansson, L. G. and Linder, S., 1975: The sea- synthesis of tropical pasture plants. III. Leaf sonal pattern of photosynthesis of some vascular age. Australian Journal of Biological Science, 24: plants on a subarctic mire. In Wielgolaski, 1077-1087. F. E. (ed.), Fennoscandian Tundra Ecosystems. MacLean, S. F., Jr., 1980: Personal communica- Part I. Plants and Microorganisms. New York: tion. Institute of Arctic Biology, University of Springer-Verlag, 194-200. Alaska, Fairbanks, Alaska 99701. Johnson, D. A. and Tieszen, L. L., 1976: Above- Maggs, D. H., 1964: Growth-rates in relation to ground biomass allocation, leaf growth, and assimilate supply and demand. I. Leaves and photosynthesis patterns in tundra plant forms roots as limiting regions. Journal of Experimental in arctic Alaska. Oecologia,24: 159-173. Botany,15: 574-583. Jung, H. G., Batzli, G. 0., and Siegler, D. S., Mark, A. F., 1965: Effects of management prac- 1979: Patterns in the phytochemistry of arctic tices on narrow-leaved snow tussock, Chinochloa plants. BiochemicalSystematics and Ecology, 7: 203- rigida. New ZealandJournal of Botany, 3: 300-319. 209. , 1969: Ecology of snow tussocks in the Kallio, P. and Lehtonen, J., 1975: On the eco- mountain grasslands of New Zealand. Vegeta- catastrophe of birch forests caused by Oporinia tion, 18: 289-306. autumnata(Bkh.) and the problems of reforesta- Mattheis, P.J., Tieszen, L. L., and Lewis, M. C., tion. In Wielgolaski, F. E. (ed.), Fennoscandian 1976: Responses of Dupontiafischerito lemming Tundra Ecosystems. Part II. Animals and Systems grazing in an Alaskan arctic tundra. Annals of Analysis.New York: Springer-Verlag, 174-180. Botany,N.S., 40: 179-197. Kamstra, L. D., Schentzel, D. L., Lewis, J. K., McNaughton, S. J., 1976: Serengeti migratory and Elderkin, R. L., 1968: Maturity studies wildebeest: facilitation of energy flow by graz- with western wheatgrass. Journal of Range Man- ing. Science, 191: 92-94. agement, 21: 235-239. Miller, P. C. and Tieszen, L. L., 1972: A prelimi- Kelsall, J. P., 1968: The MigratoryBarren-ground nary model of processes affecting primary pro- Caribou of Canada. Canadian Wildlife Service duction in the arctic tundra. Arctic and Alpine Re- Monographs, 3, 67-105. search, 4: 1-18. Klotzli, F., 1977: Wild und Vieh im Gebirgsgrass- Mooney, H. A. and Dunn, E. L., 1970: Photosyn- land Aethiopiens. In Tiixen, R. (ed.), Vegeta- thetic systems of mediterranean-climate shrubs tionand Fauna. The Hague: Cramer, 478-512. and trees of California and Chile. American Komarkova, V. and Webber, P. J., 1980: Two Naturalist, 104: 447-453. low arctic vegetation maps near Atkasook, O'Neil, L. C., 1962: Some effects of artificial de- Alaska. Arctic andAlpine Research, 12: 447-472. foliation on the growth of jack pine (Pinus bank- Kormanik, P. P. and Brown, C. L., 1969: Origin siana Lamb.). Canadian Journal of Botany, 40: and development of epicormic branches in 273-280. sweetgum. U.S. Department of Agriculture, Forest Pearsall, W. H., 1950: Mountains and Moorlands. ServiceResearch Paper, SE-54. 17 pp. London: Collins. 312 pp. Kozlowski, T. T., 1971: Growth and Developmentof Penning de Vries, F. W. T., 1975: The cost of Trees.Vol. I. New York: Academic Press. 443 maintenance processes in plant cells. Annalsof pp. Botany,N.S., 39: 77-92. Kramer, P. J. and Wetmore, T. H., 1943: Effects Plumb, T. R., 1961: Sprouting of chaparrel by of defoliation on cold resistance and diameter December after a wildfire in July. U.S. Depart- growth of broad-leaved evergreen. American ment of Agriculture, Forest Service, Pacific Southwest Journal of Botany, 30: 428-431. Forest and Range Experiment Station Technical Kulman, H. M., 1971: Effects of insect defoliation Paper, 57. 12 pp. on growth and mortality of trees. AnnualReview , 1963: Delayed sprouting of scrub oak after of Entomology, 16: 289-324. a fire. U.S. Department of Agriculture, Forest Ser- Laude, H. M., Jones, M. B., and Moon, W. E., vice, Pacific Southwest Forest and Range Experiment 1961: Annual variability in indicators of Station ResearchNote, PSW-1. 7 pp. sprouting potential in chamise. Journalof Range Reader, R. J., 1978: Contribution of overwinter- Management, 14: 323-326. ing leaves to the growth of three broad-leaved, Leopold, A. C., 1961: Senescence in plant de- evergreen shrubs belonging to the Ericaceae velopment. Science,134: 1727-1732. family. Canadian Journal of Botany, 56: 1248- Lewis, M. C. and Callaghan, T. V., 1975: Eco- 1261. logical efficiency of tundra vegetation. In Mon- Reardon, P. O., Leinweber, C. L., and Merrill, teith, J. L. (ed.), Vegetationand Atmosphere. L. B., 1974: Response of sideoats grama to ani- London: Academic Press, 399-433. mal saliva and thiamine. Journal of Range Man-

S. ARCHERAND L. L. TIESZEN/ 551 agement, 27: 400-401. World and the Plant Cover of the Tundra. Moscow: Reardon, P. O. and Merrill, L. B., 1978: Re- Botanical Institute, Academy of Sciences. sponse of sideoats grama grown in different (Translated by Issakoff, E. and Barry, T. W., soils to addition of thiamine and bovine saliva. Boreal Institute, University of Alberta.) 87 pp. Abstracts of the First International Rangeland Tinklin, I. G. and Schwabe, W. W., 1970: Lateral Congress,Denver, Colorado, p. 37. bud dormancy in the black-current Ribes nigrum Small, E., 1972: Photosynthetic rates in relation to (L.). Annals of Botany, N.S., 34: 691-706. nitrogen cycling as an adaptation to nutrient Trlica, M. J. and Singh, J. S., 1979: Transloca- deficiency in peat bog plants. CanadianJournal of tion of assimilates and creation, distribution Botany, 50: 2227-2233. and utilization of reserves. In Perry, R. A. and Smirnov, V. S. and Tokmakova, S. G., 1971: Pre- Goodall, D. W. (eds.), Aridland Ecosystems: liminary data on the influence of different num- Structure,Functioning, and Management. Vol. I. In- bers of voles upon the forest tundra vegetation. ternational Biological Programme 16. Cam- Annales Zoologici Fennici, 8: 154-156. bridge: Cambridge University Press, 537-571. Smith, D., 1975: Trends of nonstructural carbohy- Ulrich, A. and Gersper, P. L., 1978: Plant nu- drates in the stem bases of switchgrass. Journal trient limitations of tundra plant growth. In of Range Management, 28: 389-391. Tieszen, L. L. (ed.), Vegetation and Production Smith, P. R. and Neales, T. F., 1977: Analysis of Ecology of an Alaskan Arctic Tundra. New York: the effects of virus infection on the photosyn- Springer-Verlag, 457-481. thetic properties of peach leaves. AustralianJour- Ward, C. Y. and Blaser, R. E., 1961: Carbohy- nal of Plant Physiology, 4: 723-732. drate food reserves and leaf area in regrowth of Spetzman, L. A., 1959: Vegetation of the Arctic orchardgrass. CropScience, 1: 366-370. Slope of Alaska. U.S. Geological Survey Profes- Wardlaw, I. F., 1968: The control and pattern of sional Paper, 302-B, 19-58. movement of carbohydrates in plants. Botanical Thomas, W. A. and Grigal, D. F., 1976: Phos- Review, 34: 79-105. phorus conservation by evergreenness of moun- Webber, P. J., 1978: Spatial and temporal varia- tain laurel. Oikos, 27: 19-26. tion of the vegetation and its productivity, Bar- Tieszen, L. L., 1972: The seasonal course of row, Alaska. In Tieszen, L. L. (ed.), Vegetation aboveground production and chlorophyll distri- and ProductionEcology of an Alaskan Arctic Tundra. bution in a wet arctic tundra at Barrow, Alaska. New York: Springer-Verlag, 37-112. Arctic andAlpine Research, 4: 307-324. White, L. M., 1973: Carbohydrate reserves of , 1974: Photosynthetic competence of the grasses: a review. Journal of Range Management, subnivean vegetation of an arctic tundra. Arctic 26: 13-18. andAlpine Research, 6: 253-256. White, R. G. and Trudell, J., 1980: Habitat pref- , 1975: CO2 exchange in the Alaskan arctic erence and forage consumption by reindeer and tundra: Seasonal changes in the rate of photo- caribou near Atkasook, Alaska. Arctic and Alpine synthesis of four species. Photosynthetica,9: 376- Research, 12: 511-529. 390. Williams, R. D., 1964: Assimilation and trans- Tieszen, L. L. and Archer, S., 1979: Physiological location in perennial grasses. Annals of Botany, responses of plants in tundra grazing systems. N.S., 28: 419-426. In Johnson, D. A. (ed.), Special Management Wilson, B. F., 1970: The GrowingTree. Amherst: Needs of Alpine Ecosystems. Range Science Series, University of Massachusetts Press. 152 pp. 5. Denver: Society for Range Management, 22- Woledge, J., 1977: The effects of shading and cut- 42. ting treatments on the photosynthetic rate of Tieszen, L. L. and Wieland, N. K., 1975: Physio- ryegrass leaves. Annals of Botany, N.S., 41: logical ecology of arctic and alpine photosyn- 1279-1286. thesis and respiration. In Vernberg, F. J. (ed.), Woolhouse, H. W., 1967: The nature of senes- Physiological Adaptation to the Environment. New cence in plants. Symposia of the Societyfor Experi- York: Intext Educational Publishers, 157-200. mental Biology, Aspects of the Biology of Ageing, 21: Tieszen, L. L., Lewis, M. C., Miller, P. C., 179-213. Mayo, J., Chapin, F. S., III, and Oechel, Youngner, V. B., 1972: Physiology of defoliation W. C., in press: An analysis of processes of pri- and regrowth. In Youngner, V. B. and mary production in tundra growth forms. In McKell, C. B. (eds.), The Biology and Utilization Bliss, L. C. et al. (eds.), Tundra Ecosystems: A of Grasses. New York: Academic Press, 292-303. Comparative Analysis. Cambridge: Cambridge Zimmerman, M. H. and Brown, C. L., 1971: University Press. Trees. Structure and Function. New York: Tikhomirov, B. A., 1959: Relationship of the Animal Springer-Verlag. 336 pp.

552 / ARCTIC AND ALPINE RESEARCH