INFORMATION TO USERS This manuscript has been reproduced from the microfilm m aster. UMI films the text directly from the original or copy submitted. Thus, som e thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of th is reproduction is d e p en d en t upon th e quality of the copy subm itted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand com er and continuing from left to right in equal sections with small overlaps. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. ProQuest Information and Learning 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 800-521-0600 UMI RESPONSE OF SCOULER’S WILLOW (Salix scouleriana) TO MECHANICAL BRUSHING: IMPLICATIONS TO THE QUALITY OF WINTER BROWSE FOR MOOSE (Alces alces) by Roy V. Rea B.S., California State University, Stanislaus, 1992 THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in BIOLOGY ®Roy V. Rea, 1999 THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA March 1999 All rights reserved. This work may not be reproduced in whole or in part, by photocopy or other means, without the permission o f the author. 1^1 National Library of Canada Bibliothèque nationale du Canada Acquisitions and Bibliographic Services Acquisitions et services bibliographiques 395 Wellington Street Ottawa ON K1A0N4 Canada 385. rue WeWngton Ottawa ON K1A0N4 Canada Yeurm voowrM tm m Ourm NatnréUnne» The author has granted a non­ exclusive licence allowing the National Library of Canada to reproduce, loan, distribute or sell copies of this thesis in microform, paper or electronic formats. L’auteur a accordé une licence non exclusive permettant à la Bibliothèque nationale du Canada de reproduire, prêter, distribuer ou vendre des copies de cette thèse sous la forme de microfiche/film, de reproduction sur papier ou sur format électronique. The author retains ownershÿ of the copyright in this diesis. Neitho* the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author’s permission. L’auteur conserve la propriété du droit d’auteur qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation. 0-612-62492-7 CanadS APPROVAL Name: Roy V. Rea Degree: Master o f Science Thesis Title: RESPONSE OF SCOULER’S WILLOW (Salix scouleriana) TO MECHANICAL BRUSHING: IMPLICATIONS TO THE QUALITY OF WINTER BROWSE FOR MOOSE (Alces alces) Examining Committee: Chair: D. Max BIouw, Dean o f Graduate Studies — r— / »'^ ^ a e l P. Gillingham, jervisorj Associate Proféssor Katherine tzTarK^r, Associate Professor Heard, Adjunct Faculty External Examiner: Arthur L. Fredeen, Assistant Professor Date Approved: 2L. Il ABSTRACT We examined the effects o f the timing o f mechanical brushing on the nutritional value o f Scouler’s willow {Salix scouleriana) for moose {Alces alces) in winter. Brushing at different times during the growing season differentially altered morphological, chemical and phenological attributes o f willows that are considered to be important determinants o f the nutritional value o f browse for moose. Willows brushed in July produced browse o f higher value for moose (increased current annual shoot size with increased digestible protein, decreased tannin and lignin content and longer autumn leaf retention) in the first and second winter after brushing than willows brushed at other times during the growing season or unbrushed willows. We also investigated the effects o f brushing (without a timing component) on the value o f willow browse for moose for 2 to 5 w inters post-brushing in relation to willows growing in unbrushed strips retained for wildlife food and cover. Many o f the effects o f brushing on the value o f willows for moose were apparent for several years after brushing, although the degree to which brushing affected willows depended on site, year o f brushing and sampling year. The winter shoots of brushed willows were larger for 5 years, lower in tannin for 2 years, lower in lignin for 3 years, and lower in digestible energy and digestible protein for 5 years post-brushing. Willows also delayed leaf senescence for 3 to 4 years after brushing. Additionally, we examined the effects o f browsing and clipping on the value o f willow shoots produced in the year after shoot removal. Shoots produced by some willows in the year after browsing/chpping were unaffected by the previous year’s browsing/clipping intensity, but others increased in size when measured in the third to fifth winter after brushing. As the previous year’s clipping/browsing intensity increased. Ill willows often produced smaller shoots that were higher in lignin and tannin when analyzed in the second winter after brushing and produced shoots that were lower in digestible energy and digestible protein when analyzed in the third to fifth and fifth winter after brushing, respectively. Based on the results o f a browse survey conducted in the third spring following brushing, moose browsed the shoots o f brushed and unbrushed willows at similar diameters. Moose browsed past current-annual shoots and into older, less nuhitious stem materials more often when browsing the shoots o f unbrushed when compared to brushed willows. In areas being managed for the production o f conifers and where concerns for the value o f browse for moose exist, willows should be brushed in July. Because the effects o f brushing alter the nutritional value o f browse for moose for at least 5 years after brushing such changes should be considered in relation to managing wildlife strips and adjacent forest stands in moose wintering areas. IV TABLE OF CONTENTS Abstract..................... u Table o f Contents.................................................. iv List o f Tables...................................................................................................................... vi List o f Figures..................................................................................................................... viii Acknowledgements............................................................................................................ ix Preface........................ 1 Chapter 1 Chapter 2 Effects o f the timing o f mechanical brushing on the nutritional value o f Scouler’s willow {Salix scouleriana): implications for winter browsing by moose {Aloes alces)................................................ Abstract...................................................................................................... Introduction ...................................................................................... M ethods..................................................................................................... Study Area...................................... Experimental Design................................................................... Shoot Collections and Analysis.................................................. Measuring L eaf Senescence....................................................... Measuring L eaf Flush. ............................................................. Statistical Analyses...................................................................... Results................................................................................ Shoot Morphology......................................................... Shoot Chemistry.......................................................................... Leaf Senescence........................................................................... Leaf Flush..................................................................................... Age Regressions.......................................................................... Discussion................................................................................................. Shoot Morphology....................................................................... Shoot Chemistry.................. Leaf Senescence.......................................................................... Leaf Flush.................. Conclusions............................................................................................... 4 4 5 10 10 11 11 15 16 16 17 17 20 24 .27 29 29 29 32 38 40 41 Changes in the nutritional value o f Scouler’s willow {Salix scouleriana) following mechanical brushing: implications for moose {Alces alces).................................................................................. Abstract...................................................................................................... Introduction..................................... 44 44 45 Methods..................................................................................................... Study Area................................................................................... Experimental Design.................................................................. Measuring Leaf Senescence....................................................... Browse Survey............................................................................ Statistical A n iy se s..................................................................... Results....................................................................................................... Shoot Morphology....................................................................... Shoot Chemistry............................................................... Leaf Senescence ....................................................................... Browse Survey Results............................................................... Discussion................................................................................................ Shoot Morphology....................................................................... Shoot Chemistry......................................................................... Site Effects.................................................................................. Leaf Senescence.......................................................................... Implications for Moose.............................................................. Conclusions.............................................................................................. 48 48 50 52 53 53 55 55 58 58 61 63 63 66 68 70 71 75 Conclusions and Recommendations....................................................... 79 Literature Cited.................................................................................................................. 85 Appendix A Comparison o f Fibre in Distal Sections o f Willow Shoots.................. 101 Appendix B Mineral Elements in Shoots o f Brushed and Unbrushed Willows 102 Appendix C Differential Deer Browsing With Respect to Time o f Brushing 103 Chapter 3 VI LIST OF TABLES Table 1.1 Table 1.2 Table 1.3 Table 1.4 Table 1.5 Table 2.1 Table 2.2 Comparison o f morphological and chemical attributes of shoots from Scouier’s willow brushed on 14 June 1996 (n = 15) and unbrushed controls (n = 11) measured in the first winter following brushing. %DM indicates that values are expressed as a percent o f dry matter................... 18 Comparison o f morphological and chemical attributes o f shoots o f Scouler’s willow brushed in June or July o f 1997 as well as unbrushed controls measured in the first winter afier brushing (n = 30 for all morphological attributes and « = 15 for all chemical attributes). P < 0.001 for overall comparisons between brushed and unbrushed willows for all attributes. Means sharing a common superscript are not significantly different from each other as determined by Tukey’s HSD post-hoc comparisons........................... ......................................................... 19 Comparison o f morphological and chemical attributes o f shoots o f Scouler’s willow brushed in June, July or September 1996, measured 2 winters after brushing, and unbrushed controls, w = 15, 28, 28 and 18 for morphometric attributes o f June, July, September-brushed and unbrushed willows, respectively; w —15 for all other attributes. P < 0.001 for overall comparisons between brushed and unbrushed willows for all attributes. Means sharing a common superscript are not significantly different from each other as determined by Tukey’s HSD post-hoc comparisons..................................................................................... 21 Proportion o f Scouler’s willow in each brushing treatment, in the 1996 and 1997 treatment years, bearing predominantly green leaves during fall 1997. Proportions sharing common superscripts are not significantly different from each other.......................................................... 26 Proportion o f willows in each o f the 1996 brushing treatments flushing new leaves as observed on 4 different days in the spring o f 1997. Proportions sharing a common superscript are not significantly different from each other................................................................................................ 28 P values for comparisons between brushed and unbrushed plants for morphological attributes o f shoots across all sites by year since brushing. The clipping column indicates the effect o f the percentage o f shoot removal by clipping/browsing during the previous winter when treated as a covariate in the analyses. Values in parentheses represent slopes for significant effects o f the covariate............................................... 57 P values for comparisons between brushed and unbrushed plants for chemical attributes o f shoots across all sites by year since brushing. The clipping column indicates the effect o f the percentage o f shoot removal by clipping/browsing during the previous winter when treated as a covariate in the analyses. Values in parentheses represent slopes for significant effects o f the covariate................................................................. 60 vil Table 2.3 Proportion o f willows in each brushing treatment (n = 13) bearing predominantly green leaves as calculated at weekly intervals in the fall o f 1997 for 2 sites brushed in 1993 and 2 sites brushed in 1995. Proportions sharing a common superscript within a site are not significantly different from each other as determined by separate z tests............................................................................................................. 62 via LIST OF FIGURES Figure 1.1 Figure 1.2 Figure 2.1 Figure 2.2 Figure 2.3 Comparison o f the mean length o f compensatory shoots 6 weeks following brushing, for willows mechanically brushed at different times in the 1996 and 1997 growing seasons. /i = 31 for June and July 1997 brushing treatments; n = 30 for all other treatments. Willows brushed in September did not grow until the following growing season. Error bars indicate I SB.................... 23 Proportion o f Scouler’s willow within each brushing treatment bearing predominantly green leaves on 8 October 1996. Error bars represent 1 SE. Because all willows in the 30 July treatment retained predominantly green leaves, there is no SE associated with this proportion....................................................................................................... 25 Location o f study sites within the Vanderhoof Forest District. Sites are as follows: O Sackner Pit (B.C. Geographical Map Classification System-Forest district cutblock number: 93J002-001), 9 Triple Buck (93J002-002), © N orth Layton (93K019-007), O Sawmill Pit (93K019-030), 0 Huckleberry Cut (93K019-027), and © Waterlily Lake (93J0OI-OO7)........................................................................................ 49 Comparisons o f mean morphological attributes o f shoots fi'om brushed and unbrushed willows collected for 2 to 3 winters post-brushing for 1995-brushed sites (Huckleberry Cut, Waterlily Lake, Sawmill Pit) and for 3 to 5 winters post-brushing for 1993-brushed sites. * indicates a significant effect o f brushing when site effects are considered. Error bars = 1 SE. n is shown above each bar.................................................... 56 Comparisons o f mean chemical attributes o f shoots from brushed and unbrushed willows collected for 2 to 3 winters post-brushing for 1995brushed sites (Huckleberry Cut, Waterlily Lake and Sawmill Pit) and for 3 to 5 winters post-brushing for 1993-brushed sites. * indicates a significant effect o f brushing when site effects are considered. DM — dty matter. Error bars = 1 SE. n is shown above each bar..................... 59 IX ACKNOWLEDGEMENTS I would like to thank Gerd Erasmus and Lee Foster who posed the original questions that spawned this project. I am indebted to Mike Gillingham, my supervisor and the chair o f my graduate committee, who helped me hone and refine the research objectives and design into a project worthy o f a thesis. As my supervisor, Mike helped to shape not only the form o f my thesis but many aspects o f my graduate experience, which are too many to list, but for which I owe him a great deal. In addition to Mike, I would like to thank Kathy Parker for all o f her help, not only as a member o f my graduate committee, but also as an instructor fi'om whom 1 learned a great deal as her teaching assistant for 3 years. I also wish to thank Doug Heard for his membership on my graduate committee, interest in the project and help in imderstanding the implications o f my findings for moose. I would also like to acknowledge the contributions o f Mike Walters for his help at the beginning o f the project and Mark Bernards who helped to design and monitor the progress o f my project. Additionally, 1 wish to thank my external examiner. Art Fredeen for his review o f the thesis. Special thanks to my lab technicians Maria Deli, Christine Smith, Holly Cedervind and Vema Corbett, and to my lab/field technicians Susan Gibson, Bonnie Clayton, Jelmer Wiersma and Rorie Brown for all o f their help in collecting, processing and analyzing the multitude o f willow shoots sampled over the last 3 years. This research would not have been possible without start-up fimds from The Habitat Protection Committee o f the Prince George Regional office o f the B.C. Forest Service and 3 years o f support from Forest Renewal British Columbia. The Vanderhoof Forest District, to whom I am grateful, provided additional logistical support. I am indebted to the Ministry o f Forest’s Library in Victoria, Bill Faber, Sauli Harkonen and Olav Hjeljord for helping me obtain obscure articles that I am sure would have otherwise been impossible to get; to Pat Mickelson and my Dad, Ron Schneider o f the Ripon California Division o f the Fox River Paper Co. for the Tree-Free bamboo paper used to print the countless copies o f my thesis on; to Richard Crombie and Dave Dick for their innumerable hours o f chemical counselling; to Norwest labs in Lethbridge for the free standards; and to Barbara Strobl for the German and Roger Wheate and Verpi Nurmela for the Norwegian translations. Thanks to Paul Manwaring and my brother Marshall Schneider for their help and companionship in the field. I am indebted to my mother Julie and her husband Mike Ephrom for innumerable nights o f room and board. I wish to extend my gratitude to niy brothers Marshall and Matthew and their families and to Paul and Daryl and Grandma and Grandpa Cadle who also put me up and fed me many times while in Vanderhoof. 1 would like to thank Matt and Joanne Shaw for their inspiration, encouragement and timeless friendship. I would also like to thank my office mates and fiiends, Chris Johnson, Laine Cotton, Georgie Harrison and Barry Pierce for always being there. Most importantly I could not have done this in a sane fashion without the love and support o f my wife Michelle, who not only had to live with me through it all, but also often times fought blackflies, hailstorms and writer’s cramp by my side. To my grandfather Randall Cadle, who was always eager to hear if those brushing operations up on Blue Mountain were improving forage quality and likely to attract any big bulls, but who passed away before I could give him any truly conclusive answers, 1 dedicate this thesis. PREFACE Quickly transforming early serai forests into productive forest plantations is a primary goal o f the forest vegetation manager. To meet this objective, managers often employ stand-tending measures, that although they may be advantageous to conifer crops, reduce the growth and maturation o f other plant species. Brushing and weeding operations, for instance, are often used to eliminate or reduce the competitive effects o f early-successional trees and shrubs on conifers such as pine {Pinus spp.) and spruce {Picea spp.) in the early stages o f forest succession (Smith 1986). Brushing and weeding operations in North America are typically accomplished through aerial applications o f forest herbicides such as glyphosate (Roundup®, Monsanto Chemical Co. St. Louis, Mo.) and 2,4-D formulations (Biggs and Walmsley 1988). In areas where social objections relative to herbicide use exist, or where forest-worker unemployment is high, however, other means such as sheep {Qvis aries) grazing and motor/manual (mechanical) brushing operations may supplant the application o f herbicides to reduce brush. Mechanical brushing operations, which employ crews o f forest workers with handheld brushsaws, are particularly popular in some parts o f British Columbia. Several forest-worker crews, for example, brush and weed an average o f 1500 ha o f regenerating forests each growing season in the Vanderhoof district o f the British Columbia Forest Service (R.V. Rea, unpublished data) By altering the vegetative composition o f regenerating forests, mechanical brushing also alters the habitat suitability o f these areas for a variety o f animals in the years following treatment (Goodrum and Reid 1956). For example, such treatments reduce the abundance o f important winter browses for moose {Alces alces) such as willow {Salix spp.) and birch {Betula spp.; Harkonen et al. 1998). Such compositional changes may be short-lived due to intense vegetative or compensatory sprouting following mechanical damage (Coates and Haeussler 1986). The value o f compensatory growth for herbivores such as moose {Alces alces), however, is not completely understood because plants often respond to mechanical injury by altering their morphological and chemical composition (Rhoades 1985; Danell and Bergstrom 1989). Plants may also alter patterns o f leaf senescence and leaf flush when damage is severe due to changes in plant physiochemistry (Millington 1963; Kindschy 1989). How plants specifically respond to damage depends to a large extent on the timing o f damage (Whitham et al. 1991; Ouellet et al. 1994). Because mechanical brushing operations are performed throughout the growing season, the morphology and chemistry o f shoots in winter and/or the fall and spring leafing phenology o f recently brushed plants could depend on when plants are brushed. Because moose tend to select browse based on shoot morphology and chemistry as well as leafing phenology (Danell et al. 1994), such changes will likely affect the value o f browse for moose. Although mechanical brushing affects browse production in the years following treatment (Hjeljord and Gronvold 1988), the effects o f the timing o f brushing on the overall value o f browse for moose and how long these effects last are poorly understood. My thesis research, therefore, was designed to determine the effects o f the timing o f brushing on the value o f browse for moose and to determine how long such effects last. I focused my investigation on the response o f Scouler’s willow because this upland species o f willow is commonly eaten by moose in winter and is regularly brushed throughout central BC. In chapter 1 ,1 describe the results o f an experiment designed to determine the effects o f brushing time on the nutritional value o f Scouler’s willow for moose. I compared nutritional value o f willows (in terms o f the morphology and chemistry of winter shoots and leaf phenology) from brushed willows for 1 to 2 years following brushing, with unbrushed willows, and compared the value o f plants brushed at different times. In chapter 2 , 1 examine the effects o f mechanical brushing on these same plant attributes for 2 to 5 years post-brushing. Unlike chapter 1, however, my analysis did not include a time-of-brushing within season. Instead, I determined the effects o f brushing among 6 sites that were brushed at more o f an operational level throughout the growing season prior to the commencement o f my study (3 in 1993 and 3 in 1995). I also examined the effects o f clipping and browsing on the nutritional value o f shoots produced during the following year and compared the shoot diameter at the point o f browsing from shoots on brushed and unbrushed willows. In Chapter 3,1 summarize my findings from Chapters 1 and 2 and discuss the implications o f these findings for managing brush in areas where concerns for the quality of browse for moose exist. I make specific recommendations on when to brush willows for moose while maintaining vegetation management objectives and discuss the long term effects o f brushing on browse quality with respect to managing adjacent forest stands and unbrushed wildlife strips. CHAPTER 1. EFFECTS OF THE TIMING OF MECHANICAL BRUSHING ON THE NUTRITIONAL VALUE OF SCOULER’S WILLOW (So/ix scouleriana): IMPLICATIONS FOR WINTER BROWSING BY MOOSE {Alces alces). Abstract We^ examined the effects o f the timing o f mechanical brushing on determinants o f nutritional value o f Scouler’s willow {Salix scouleriana) for moose {Alces alces) in the first 2 winters after brushing. Brushing at different times in the growing season differentially altered morphological, chemical and phenological attributes known to affect the value o f browse for moose. We determined nutritional value o f browse on the basis o f the size, digestible energy, digestible protein, tannin and lignin content o f willow shoots in winter. Additionally, we included the phenology o f plant leafing in our estimate o f nutritional value. Willows brushed in July were o f higher value for moose in the first 2 winters following brushing than willows brushed in June, August or September or unbrushed willows. In the first winter after brushing, plants that were brushed in early July had shoots that were lower in lignin content, higher in digestible protein, and lower or not different in tannin content than shoots from either earlier-brushed or unbrushed willows (all P < 0.001). Willows brushed in early July also had shoots that were large (in terms o f basal diameter, length and mass), high in digestible energy, and displayed a longer delay in leaf senescence when compared to shoots from either earlier-brushed or unbrushed willows in the first winter after brushing. Willows brushed in late July, however, were not available in the first winter after brushing. In the second winter after ' The first person plural is used in the context o f the individual chapters o f the thesis which have been written in manuscript format and prepared for journal submission. The first person plural indicates and acknow ledges the contributions o f my supervisor M ike Gillingham as co-author o f these manuscripts and o f those involved with the technical assistance required for field and lab work. brushing, willows that were brushed in July had shoots that were lower in digestible energy and digestible protein but larger, and lower in tannin and lignin content and displayed a longer delay in leaf senescence when compared to shoots o f either earlierbrushed or unbrushed willows. Willows brushed after July also produced nutritious shoots in the second post-treatment year, but did not regenerate sufficient growth to be available for moose in the first winter after brushing. In addition, willows brushed in September displayed a delay in leaf flush in the first post-brushing spring. We recommend that willows should be brushed in July in areas that are managed for conifer production and that are also important to moose in winter. Introduction Early-successional plants such as birch {Betula spp.) and willow {Salix spp.) comprise the majority of the diet o f moose in fall and winter (Pierce 1983: Regelin et al. 1987; Harkonen et al. 1998). These deciduous shrubs and trees are highly palatable and nutritious for moose and other herbivores in relation to other plants that are available during winter (Chapin 1980; Salonen 1982; Pastor and Naiman 1992). Despite the importance o f willows and other deciduous plants to moose, these species are often considered weedy competitors by foresters managing early serai forests for the production o f conifers such as pine {Pinus spp.) and spruce {Picea spp.). Consequently, the growth o f these plants in conifer plantations is suppressed through the application of herbicides and/or mechanical brushing. Although both herbicide applications and mechanical brushing affect browse production (Hjeljord and Gronvold 1988), most research has focused on determining the impacts o f herbicides on the quality and abundance o f browse for moose (Soper et al. 1993; Hjeljord 1994; Raymond et al. 1996). Mechanical brushing, is used extensively in parts o f northern Europe (S. Harkonen, South Savo Game Management District, Juva, Finland, personal communication) and in many parts o f North America where the use o f herbicides is unpopular (Bernstein 1978; Conard 1984). Mechanical brushing is extremely labour intensive and time consuming when compared to herbicide applications at comparable landscape scales. Consequently, plants growing in areas targeted for brushing treatments are not brushed simultaneously, but rather throughout the growing season. This constraint results in plants being brushed during different stages o f phenological development and growth (Coates and Haeussler 1986) and likely influences the overall quality and availability o f browse for moose. Deciduous browse species such as willow and oak (Quercus spp.) respond to severe forms o f mechanical damage through vegetative regeneration (Gysel 1957; Kauppi et al. 1988; Tschaplinski and Blake 1994; Sennerby-Forsse and Zsuffa 1995). The timing o f damage and the ratio o f retained nutrients to lost nutrients, however, often influence such responses (Kays and Canham 1991; Whitham et al. 1991; Ouellet et al. 1994). If current-annual-shoot (measured fi'om the previous year’s growth scar to the shoot tip and hereafter termed “shoot”) loss occurs when nutrients are concentrated in the above­ ground structures being removed, the ability o f plants to respond in a compensatory manner (compensate for lost tissues through shoot regeneration; sensu Belsky 1986) may be weaker than if damage occurs when the majority o f nutrients have been translocated to below-ground storage structures (Kays and Canham 1991). Yellowpoplar {Liriodendron tulipiferd) and flowering dogwood {Comus florida), for example, produce more shoot biomass when heavily pruned during plant dormancy than when pruned during summer (Harlow and Halls 1972). Birch produces larger, more fibrous compensatory shoots that are lower in nutrients following winter rather than summer browsing by moose (Danell et al. 1994). The nutritive quality o f red stem ceanothus {Ceanothus sanguineus), ninebark {Physocarpus malvaceus), rose {Rosa spp.) and snowberry (Symphoricarpus spp.) for wild ungulates increases after simulated early summer browsing by sheep but decreases if browsing occurs late in summer (Alpe et al. 1999). Nutrient losses that are associated with summer tissue damage may lead to fineroot mortality, which fiurther decreases nutrient uptake firom soils and alters plant carbon/nutrient ratios (Herms and Mattson 1992). Changes in carbon/nutrient ratios can alter the production o f chemical deterrents and structural compounds such as tannins and lignin (Bryant and Kuropat 1980; Belovsky 1981; Herms and Mattson 1992) and may affect the digestibility o f plant tissues for herbivores (Bryant and Kuropat 1980; Hanley et al. 1992). Changes in the production and distribution o f plant hormones following damage at different times also accompany changing carbon/nutrient ratios (Wenger 1953; Phillips 1975). Such changes, acting alone or in concert, can affect the nutritional value o f browse tissues for herbivores such as moose. Because the phenology o f plant leafing depends largely on hormonal activity and plant nutritional status (Millington 1963; and Kramer and Kozlowski 1979), changes to plant chemistry following mechanical damage are likely to alter patterns o f leaf senescence and leaf flush in accordance with the timing o f damage. For example, birch trees display a delay in spring leaf flush when defoliated early in the previous growing season, but trees defoliated later in the year are typically less affected (Tuomi et al. 1989; Raitio et al. 1994). Defoliating willows in early summer also stimulates the production o f leaves later in the growing season that are higher in nitrogen, phosphorus and potassium (Chapin 1980). Late-summer and winter shoot removal may cause browse plants to display premature leaf flush in the spring (Larson 1975; Kindschy 1989) and delayed leaf senescence in fall (Willard and McKell 1978; Kindschy 1989; Smith 1986); how the timing o f shoot removal may affect such patterns is less clear. Moose select browse shoots based on shoot morphology and chemistry (Danell et al. 19856; Risenhoover 19876). Moose select the largest available shoots when browsing and select compensatory growth originating from previously damaged versus undamaged plants (Penner 1978; Machida 1979; Danell et al. 19856; Risenhoover 1987a). This selection is presumably due to morphological changes that allow for increased bite sizes and intake rates (Gross et al. 1993; Shipley et al. 1994) and/or chemical changes that make the shoots of damaged plants more palatable and easy to browse (Danell et al. 19856; Suter 1993). Because the ratio o f bark and buds (where most nutrients and chemical deterrents are stored) to woody cortex is relatively reduced in large shoots (Danell and Bergstrom 1985; Radwan and DeBell 1988; Schwartz 1992), moose may select shoots based on a trade-off between their nutritional requirements and chemical tolerance (Palo et al. 1992). Although moose do not select forages based exclusively on palatability or nutritive quality (Risenhoover et al. 1985; Risenhoover 19876: Jia et al. 1997), they tend to select forages that are relatively high in digestible energy and protein (Regelin et al. 1987; Schwartz et al. 1987). Moose also avoid a variety o f phenolic compounds such as lignin (Risenhoover 19876) and tannins (Bryant and Kuropat 1980; Bryant et al. 1983; Singer et al. 1994), which can contribute significantly to reductions in dry matter digestibility (Bryant and Kuropat 1980; Hanley et al. 1992). Selecting browses low in tannin and lignin content is particularly important to ruminants because tannins can chemically inhibit the proper functioning o f rumen microbes (Risenhoover et al. 1985) and lignin can reduce cropping and processing rates that regulate energy intake (Spalinger et al. 1986). Although the selection o f winter shoots by moose depends to a large extent on shoot morphology and chemistry, the selection o f browse in the early spring and late fall is influenced by the presence or absence o f leaves (Renecker and Hudson 1986; Danell et al. 1994). Moose tend to select early-greening forages and late-senescing plants during spring and fall, respectively (Schwartz et al. 1988a; Danell et al. 1994). Such plants are important to moose because they provide a readily accessible pool o f nutrients during times o f the year when nutritious foods are reduced (Blair et al. 1980; Chapin 1980). Understanding how the timing o f brushing affects the phenology o f leafing as well as the morphology and chemistry o f shoots will help resource managers better understand and manage for the effects of brushing on forage quahty and availability in important wintering areas for moose. In this chapter, we investigate how the timing o f mechanical brushing affects the nutritional value o f Scouler’s willow for moose in the first 2 years following brushing. We determined the nutritional value o f willow shoots on the basis o f size, digestible energy, digestible protein, tannin and lignin content o f dormant shoots. Willows that displayed delays in leaf senescence in the fall or premature leaf flush in the spring were considered to be o f better quality than those that did not. We chose Scouler’s willow for our study because it is a predominant upland willow on clearcuts in central British Columbia and because o f its local importance as winter browse for moose (Roberts 1986; Porter 1990). Specifically, we examined whether the timing o f brushing affected: (1) the morphology and chemistry o f regenerating or compensatory shoots for moose in the first 2 winters following treatment; (2) leaf senescence and the amount o f time that willow 10 leaves were available to moose in the first 2 fails following treatment; and (3) the timing o f the emergence o f willow leaves for moose in the spring following brushing. Methods The study area Our study area was located 20 km north-east of Vanderhoof, British Columbia, Canada (54°05’N, 123°55’W) in the sub-boreal spruce forest ecotype (Meidinger and Pojar 1991). The topography is rolling and the site elevation is ~ 800 m. Soils are sandy loam (D. Sommerville, Vanderhoof Forest District, Vanderhoof, B.C., Canada, personal communication). The climate is continental and characterized by seasonal extremes with cold winters and warm, moist summers. Mean annual precipitation is ~ 46 cm; snow fall averages ~ 200 cm and mean annual temperatures range from 1.7 to 5 °C (from 19611990; Atmospheric Environment Service 1993). The landscape is dominated by coniferous forests o f hybrid white spruce {Picea engelmannii xglauca) and subalpine fir {Abies lasiocarpa). Lodgepole pine {Pinus contorta var. latifolia) and trembling aspen {Popiilus tremuloides) pioneer secondary successional sites (Meidinger and Pojar 1991), as do several species o f upland willows (Porter 1990). Our experiments were conducted on a 10-year-old regenerating pine clearcut replanted in pine and ~ 14 ha in size. We selected this site because o f the abundance o f Scouler’s willow saplings o f similar size (2.0 to 2.5 m) and structure and because o f the proximity o f the site to an active ranching operation and recreational trailhead. Locating the site in an area o f moderate human activity minimized site utilization and browsing by moose. Moose density in the surroimding area at the time o f the study was ~ 0.5 animals per km" (D. Heard, Ministry o f Enviromnent Lands and Parks, Prince George, B.C., Canada, personal communication). 11 Experimental Design We selected individual willows for our experiment on the basis o f size (2 to 2.5 m tall) and form and identified them according to leaf characters, shoot morphology and catkin anatomy following Argus (1992). Each willow contained ~ 10 to 15 co-dominant main stems; willows were easy to delineate from one another because groups o f main stems were well spaced. We selected, numbered and tagged 120 willows in the spring o f 1996 and another 150 willows in the spring o f 1997. In the spring o f 1996, 30 willows were randomly assigned to 1 o f 4 brushing treatments spaced at 6-week intervals: 14 June, 30 July, 15 September and a control. In the spring o f 1997, an additional 30 willows were randomly assigned to each o f 5 brushing treatments: 1 June, 1 July, 1 August, 1 September and an unbrushed control. A swing saw was used to brush willows ~ 10 to 15 cm above the ground. Disk samples were collected from the 5 largest main stems on all 1996-brushed willows. We then assessed the age o f each willow from disk samples by counting the annuli to determine whether or not the age o f willow affected the morphological and chemical response o f plants to brushing. The growth response o f each willow within each brushing treatment was determined 6 weeks after brushing by measuring the length o f 5 randomly selected compensatory shoots and calculating their mean length. Shoot Collections and Analysis During the winter o f 1996-97 we collected shoots from 15 randomly selected willows o f the 14 June 1996 treatment and corresponding imbrushed controls. We also collected shoots from all 30 o f the 1 June, 1 July, and 1997 control willows during the winter o f 1997-98 to determine the effect o f brushing on willow shoots in the first winter following brushing. Shoots from the 30 July 1996, 15 September 1996, 1 August 1997, 12 and 1 September 1997 treatments did not grow sufficiently following treatment to be visible above mid-December snows and were not collected. In the second winter after brushing, we collected shoots from the 15 previously unanalyzed 14 June and 1996 controls as well as shoots from all 30 o f the 30 July and 15 September 1996-bmshed willows. We sampled by selecting every third shoot from a tree. If we did not obtain at least 150 g o f material, we repeated the sampling procedure until we had —150 g for analysis. To inhibit shoot metabolic activities following clipping, shoots were collected during the first 2 weeks of December o f both years, in sub-zero weather. We separated collections from each willow in the field into 1 plastic freezer bag for tannin analysis and 1 bag for all other analyses. Bags were then sealed and all materials were stored at -20 °C until analysis. Morphometric measures o f mass, length, basal and tip diameters were taken for all shoots collected from willows. If more than 30 shoots were collected from a particular willow, we randomly sub-sampled 30 shoots for morphometric measures. Following morphometric measurements, all shoot material for each plant was combined, cut to ~ lO-cm lengths and dried to a constant mass (±0.1 g) at 39 °C in a forced-draft drying oven (Despatch LAD series2-24-3 Minneapolis, MN). We then milled the dried material with a Thomas-Wiley (Swedesboro, NJ) mill using a 0.5-mm sieve screen and hand mixed the samples to homogenize them. Gross energy was determined with a bomb calorimeter (Parr model 1341, Moline, IL) using 0.75 to 1.0 g o f material and procedures outlined by the manufacturer. Gross energy values were corrected to dry mass by desiccating with anhydrous CaS 0 4 (WA Hammond Drierite Co., Xenia, OH.) for 24 h. Elemental nitrogen (EN) was determined with an elemental CHN analyzer (Carlo Erba, Na Series 2, Milano, Italy) following procedures outlined by the manufacturer and Pella and Colombo (1973). The elemental analyzer was calibrated using atropine (4.84 %N) and the National Institute o f Standards and Technology (NIST) standard #1573a (3.03 %N). Because elemental nitrogen approximates the nitrogen content o f a sample with the same accuracy and precision as total Kjeldahl nitrogen (TKN; Hellinga et al. 1998), we substituted elemental nitrogen for TKN in equations outlined in Hanley et al. (1992) for estimating digestible protein. Digestible dry matter (the fraction o f plant dry matter that is digestible by ruminants such as moose after accounting for the content o f structural compounds, e.g., cellulose, hemicellulose and lignin) was determined for use in computing in vitro digestible energy (gross energy x digestible dry matter). Digestible dry matter was computed using equations outlined in Hanley et al. (1992). Although equations for digestible protein and digestible dry matter were originally developed for deer {Odocoileus spp.), these equations can be used to provide useful approximations for large-sized cervids that secrete salivary tannin-binding proteins, such as moose (Hanley et al. 1992). To estimate the fibre fractions in our samples for use in determining digestible dry matter, we used a fibre refluxing/distillation apparatus (Labconco model 30006, Kansas City, MO). We used the neutral detergent procedure to determine total plant fibre (neutral detergent fibre or NDF), the acid detergent procedure to determine cellulose and lignin content (acid detergent fibre or ADF), and the acid detergent lignin procedure to determine the lignin content (acid detergent lignin or ADL; including cutin) o f our samples (Goering and Van Soest 1970). Sodium sulfite was omitted from the NDF procedure as recommended by Hanley et al. (1992) for the determination o f NDF from browse stems. Because the filtrate did not congeal, we also omitted the optional wash with hexane from the ADF procedure (Goering and Van Soest 1970). Because there 14 were no problems with filtration, asbestos was not used in the determination o f acid detergent lignin (Goering and Van Soest 1970). We standardized NDF and ADF protocols by using standard forage mix samples from Norwest Labs (Lethbridge, Alberta). Although acid detergent lignin is used to calculate digestible dry matter, we also separated out and reported this fraction alone because lignin is the main cell wall component limiting digestion (Robbins 1993) and moose are knowm to select forages lower in lignin (Risenhoover 19876). Because dormant twigs contain relatively small amounts o f tannin (Palo 1984; Robbins et al. 19876), we followed the recommendation o f Hanley et al. (1992) and did not apply the tannin correction factor in calculations for either digestible protein or digestible dry matter. Although crude tannin content was not quantified for use in digestibility determinations, we did determine the relative differences in tannin content between brushing treatments in order to help understand changes to one o f the chemical characteristics known to influence winter browse selection by moose. Tannin content was assessed using a radial diffusion protein precipitation assay that we modified from Hagerman (1987). We altered the procedure to reduce chemical usage and concentrate the extract, because winter twigs are known to contain relatively little tannin (Hanley et al. 1992). Frozen samples were submersed in liquid nitrogen, and then ground with a Thomas-Wiley mill and a 1-mm sieve screen that had been cooled with liquid nitrogen to avoid gumming. We homogenized the material by hand mixing and dispensed 200 pg into a 1.5-ml centrifuge tube. After the addition o f 500 pi o f 70% acetone, the sample was vortexed for 1 min and centrifuged at 7500 g for 5 min. Subsequently, we pipetted 60 pi o f supernatant into a collection tube. We then 15 resuspended the sample pellet in an additional 100 pi o f acetone, vortexed, centrifuged again and collected 100 pi o f supernatant in the collection tube. This last step was repeated 2 additional times resulting in a 360-pl sample that was 2.5 times more concentrated than extracts obtained using Hagerman's (1987) original procedures. Finally, 4, 8-pl aliquots o f extract were dispensed into agarose plates containing bovine serum albumin fraction V powder (BDH Chemical Co., Toronto, Ont.) following procedures outlined by Hagerman (1987). We prepared 3 replicate samples o f 200 pg for each plant. Each replicate sample o f extract was dispensed into 9 separate wells distributed across 3 separate plates. After incubating the plates for 96 h at 30 °C, we determined the diameter o f the precipitation ring; to allow for non-circular rings, 3 ring measurements were made ~ 120° apart with a digital calliper on each ring. The 3 replicate ring diameters were then averaged first across plates, then among the 3 replicate samples from each plant to determine the mean diameter for a plant. Because the square o f the ring diameter is proportional to the amount o f tannin in the sample (Hagerman 1987), we use the mean ring diameter squared as an index o f the amount o f tannin (hereafter referred to as “tannin content”) contained in the shoots sampled from each willow. Because this modified technique uses plant extract 2.5 times more concentrated than extracts derived using the original technique, however, tannin content reported here cannot be directly compared to values reported by others using extracts that were not concentrated. Measuring L ea f Senescence Differences in fall-leaf senescence were estimated by determining the predominant leaf colour o f each willow within each brushing treatment on a given date, in both the falls o f 1996 and 1997. We classified plants as having predominantly green. 16 yellow, brown or no leaves. We then compared the proportion o f plants having predominantly green leaves among brushing treatments. To assess stage o f leaf senescence in the fall o f 1996 for willows brushed in 1996, the predominant leaf colour o f each willow was estimated on a single day in mid-fall (8 October). We estimated stage o f senescence in the fall o f 1997 for willows brushed in 1996 and 1997 by estimating the predominant leaf colour o f each willow on a weekly basis from 5 September to 17 October, before fall frosts. Measuring L ea f Flush To examine differences in spring-leaf flush for all willows from the 1996treatment year, we monitored plants every 3 days in the spring o f 1997 and compared the proportion o f willows in each brushing treatment bearing newly flushed Leaves on each day. Willows were scored as leaf bearing when the bud scales o f at least 1 bud had separated and the expanding foUage was visible (Suzuki et al. 1988). We collected data from 11 May, at the first signs o f bud break, through 28 May when all willows within each brushing treatment had flushed the majority o f their leaves. Most o f the shoots were removed from all willows in the final winter (1997-98) o f the study for analysis so no measurements were made in spring 1998. Statistical Analyses We used linear regression (Sokal and Rohlf 1995) to test the effect o f age o f willows brushed in 1996 on shoot length 6 weeks after brushing, as well as shoot morphological and chemical attributes in the first 2 winters following brushing treatments. To test the differences in shoot length 6 weeks after brushing, as well as the differences in shoot morphology and chemistry in the first 2 winters post-brushing among treatments, we used an analysis o f variance (ANOVA; Sokal and Rohlf 1995) with 17 treatment time as a fixed factor. When appropriate, square root, log or inverse transforms (Tabachnick and Fidell 1996) were applied to normalize and/or homogenize the variance o f the shoot attribute being tested between treatments. I f the marginal distributions were negatively skewed (as was the case for digestible protein), we reflected the variable before applying the appropriate transform (Tabachnick and Fidell 1996). To reflect the variable, we took the largest value in the distribution and added 1 to it to form a constant; a new variable was then created by subtracting each value from this constant (Tabachnick and Fidell 1996). Homogeneity o f variances were tested using a Levene’s test (Milliken and Johnson 1984); a Kolmogorov-Smimov test (Zar 1984) was used to test for normality. Tukey’s HSD test (Zar 1984) was used for post-hoc comparisons for ANOVA's. We used a z test (Zar 1984) to analyze the differences in the proportions o f willows bearing predominantly green leaves on specific dates in the falls o f 1996 and 1997. We also used a z test to analyze the differences in the proportions o f willows leafing within each brushing treatment every third day during the spring o f 1997. All analyses were conducted using Statistica (StatSoft Inc. 1997). Results Shoot morphology Willows brushed during the summer o f 1996 had longer and heavier shoots with thicker basal diameters, but thinner shoot tips than unbrushed willows in the first winter after brushing (Table 1.1). Willows brushed during the summer o f 1997 also had longer, heavier and thicker shoots with thinner tip diameters than unbrushed willows in the first post-treatment winter (Table 1.2). In the second winter following brushing, willows brushed in 1996 continued to have shoots that were longer, heavier and thicker with 18 T able 1.1. Comparison o f morphological and chemical attributes o f shoots from Scouler’s willow brushed on 14 June 1996 (n = 15) and unbrushed controls (n = 11) measured in the first winter following brushing. %DM indicates that values are expressed as a percent o f dry matter. Shoot Attribute Brushed Unbrushed Mean SE Mean SE F P Length (cm) 49.8 2.7 13.3 1.6 113.9 <0.001 Mass (g) 4.93 0.54 0.75 0.12 137.3 <0.001 Basal Diameter (mm) 4.2 0.2 2.8 0.1 59.3 <0.001 Tip Diameter (mm) 1.5 0.1 2.0 0.1 27.1 <0.001 Lignin (%DM) O.ll 0.01 0.13 0.01 12.3 0.002 Tannin Content^ 82.3 3.6 96.1 4.2 6.3 0.019 Digestible Energy (kcal/g) 2.885 0.028 3.327 0.309 111.0 <0.001 Digestible Protein (%DM) 0.53 0.10 1.76 0.33 16.0 <0.001 t Content o f tannin is the square o f the diameter (in mm) o f the precipitation ring (see text). 19 T able 1.2. Comparison o f morphological and chemical attributes o f shoots o f Scouler’s willow brushed in June or July o f 1997 as well as unbrushed controls measured in the first winter after brushing {n = 30 for all morphological attributes and « = 15 for all chemical attributes). P < 0.001 for overall comparisons between brushed and unbrushed willows for all attributes. Means sharing a common superscript are not significantly different from each other as determined by Tukey’s HSD post-hoc comparisons. Shoot Attribute Brushed 1 June Unbrushed 1 July F Mean SE Mean SE Mean SE Length (cm) 54.4 2.2 25.5 1.6 9.9 0.8 211.1 Mass (g) 4.59 0.38 1.31 0.15 0.42 0.06 163.3 Basal Diam. (mm) 4.1 0.1 2.8 0.1 2.2 0.1 128.8 Tip Diam. (mm) 1.3“ 0.04 1.2" 0.02 1.6 0.04 26.9 Lignin (%DM) 0.12 0.01 0.10 0.01 0.16 0.01 84.7 Tannin Content 78.9* 4.8 80.8* 6.8 130.7 10.2 44.3 Digestible Energy (kcal/g) 2.821 0.009 3.168 0.052 3.254 0.105 39.4 Digestible Protein (%DM) 1.37“ 0.57 2.78 0.25 1.74“ 0.58 20.5 20 thinner tips than unbrushed willows (Table 1.3). Although the shoot morphology o f brushed and unbrushed plants differed in the first 2 winters after brushing, the magnitude o f difference in shoot morphology between brushed and unbrushed plants depended on the timing o f brushing (Fig. 1.1 and Table 1.3). In the year following brushing, willows brushed earliest in the growing season produced the longest shoots as indicated by data collected 6 weeks following brushing (Fig. 1.1; 1996: = 180.67; P < 0.001; 1997: P'lso —252.20; P < 0.001). Measurements taken from dormant shoots in the winter following the 1997 brushing treatments showed similar trends: willows brushed earliest in the previous growing season had the largest shoots in winter (Table 1.2). Following a full season o f post-treatment growth, willows brushed earliest in the summer o f 1996 no longer had the largest browse shoots in the second winter after brushing. Instead, willows brushed later in the 1996 season had larger shoots in the winter o f 1997 (Table 1.3). Shoot Chemistry Lignin, tannin, digestible energy and digestible protein content were lower in the shoots o f brushed when compared to unbrushed willows in the first winter after brushing in 1996 (Table 1.1). Plants brushed during the 1997 growing season also had shoots in the winter after brushing that were lower in lignin, tannin and digestible energy content, but higher or not significantly different in digestible protein than the shoots o f unbrushed controls (Table 1.2). Plants brushed in July 1997 had shoots that were lower in lignin, but higher in digestible energy and digestible protein than the shoots o f plants brushed in June o f the same year (Table 1.2). Although plants brushed in July 1997 had shoots with T able 1.3. Comparison o f moiphological and chemical attributes o f shoots o f Scouler’s willow brushed in June, July or September 1996, measured 2 winters after brushing, and unbrushed controls. // = 15, 28, 28 and 18 for morphonietric attributes o f June, July, September-brushed and unbrushed willows, respectively; w = 15 for all other attributes. P < 0.001 for overall comparisons between brushed and unbrushed willows for all attributes. Means sharing a common superscript are not significantly different from each other as determined by Tukey’s HSD post-hoc comparisons. Shoot Attribute Unbrushed Brushed 14 June 15 Sep 30 July Mean SE Mean SB Mean SE Mean SE Length (cm) 20.6 1.9 54.8 2.6 80.5 2.9 13.8 1.6 165.1 Mass (g) 1.17 0.19 6.70 0.54 12.82 0.84 0.63 0.10 164.6 Basal Diam. (mm) 2.7" 0.2 4.5 0.2 6.0 0.2 2.5" 0.1 112.5 Tip Diam. (mm) 1.4" 0.1 1.5" 0.1 1.6* 0.03 1.7" 0.1 6.2 Lignin (%DM) 0.15" 0.01 0.13* 0.003 0.13* 0.003 0.15" 0.003 12.2 Table 1.3 (continued) Shoot Attribute Unbrushed Brushed 14 June 30 July 15 Sep Mean SE Mean SE Mean SE Mean SE F Tannin Content 92.8 2.4 80.4" 2.5 78.3" 2.8 107.1 4.1 19.3 Digestible Energy (kcal/g) 2.962 0.035 2.796'' 0.030 2.726" 0.028 3.178 0.033 40.5 Digestible Protein (%DM) 1.87"'' 1.36 0.68" 0.10 0.10*" 0.34 2.00" 0.32 5.5 50 1996 E ü 40 O 30 ÛC O LU 20 LU X 10 co 14 JUNE 30 JULY 1 JUNE 1 JULY 1 AUG Fig. 1.1. Comparison o f the mean length o f compensatory shoots 6 weeks following brushing, for willows mechanically brushed at different times in the 1996 and 1997 growing seasons. « = 31 for June and July 1997 brushing treatments; n = 30 for all other treatments. Willows brushed in September did not grow until the following growing season. Error bars indicate 1 SE. N U)) 24 significantly lower tannin content than the shoots o f unbrushed willows, the tannin content o f these shoots was not significantly different than that found in the shoots o f June-brushed plants (Table 1.2) In the second winter after brushing, the shoots o f plants brushed at different times during 1996 were lower in tannin content and digestible energy than the shoots o f unbrushed controls, although July and September brushing treatments had shoots lowest in these chemical attributes (Table 1.3). Shoots produced following July and September brushing treatments had significantly lower concentrations o f lignin and digestible protein than the shoots o f unbrushed willows in the second winter after brushing. There were, however, no significant differences between shoots from the June-brushing and unbrushed willows with respect to lignin and digestible protein (Table 1.3). L e a f Senescence A higher proportion o f plants brushed during the 1996 treatment year had greener, younger leaves than the imbrushed controls on 8 October, in the first fall after brushing (Fig. 1.2). Similarly, a higher proportion o f plants brushed during the 1997 treatment year had greener leaves than imbrushed willows from 19 September through 10 October in the first post-treatment year (Table 1.4). Brushing treatments applied later in the growing season, during both years, retained the highest proportion o f plants with predominantly green leaves in the first October following brushing (Table 1.4). Although brushing treatments applied later in the 1997 treatment year had increasingly higher proportions o f plants with green leaves on 17 October 1997, plants brushed in July and August of the same year showed no significant differences in patterns o f leaf senescence 25 il LU LU Cû — I n = 28 1.0 coz h LU O ^ n = 28 0 .8 d a 06 il 0 .4 g S 0.2 CL 0 .0 n = 30 14 JUNE 30 JULY CONTROLS Fig. 1.2. Proportion o f Scouler’s willow within each brushing treatment bearing predominantly green leaves on 8 October 1996. Error bars represent 1 SE. Because all willows in the 30 July treatment retained predominantly green leaves, there is no SE associated with this proportion. 26 Table 1.4. Proportion o f Scouler’s willow in each brushing treatment, in the 1996 and 1997 treatment years, bearing predominantly green leaves during fall 1997. Proportions sharing common superscripts are not significantly different from each other. Measurement Date o f Brushing Date 1996 1997 1 June 1 July 1 Aug Control 14 June 30 July 15 Sep Control n=32 rt = 33 « = 20 rt = 32 n = 15 « = 30 n =30 n = 18 5 Sep 1.00“ 1.00“ 1.00“ 0.94“ 1.00' 1.00' 1.00' 0.67 12 Sep 1.00“ 1.00“ 1.00“* 0.91* 0.97'^ 1.00' 0.87" 0.43 19 Sep 0.91“ 1.00* 1.00“* 0.75 0.57 0.93' 0.83' 0.30 28 Sep 0.91“ 1.00* 1.00“* 0.59 0.37^ 0.70" 0.60" 0.23' 3 Oct 0.82 0.97“ 1.00“ 0.44 0.23' 0.53" 0.57" 0.20' 10 Oct 0.50 0.91“ 1.00“ 0.18 0.07' 0.30" 0.33" 0.03' 17 Oct 0.32“ 0.88 1.00 0.19“ 0.03' 0.23" 0.33" 0.03' 27 from 5 September through 10 October (Table 1.4). Plants brushed in July and August 1997, however, did show signs o f delayed leaf senescence when compared to June brushing treatments from 3 October through 10 October (Table 1.4). In the second fall after treatment, brushed plants displayed delays in leaf senescence longer than unbrushed plants from 5 to 19 September 1997 and plants brushed in July and September displayed delays in leaf senescence longer than plants brushed in June (measured on 19 September 1997; Table 1.4). In the second fall after brushing, willows brushed in July and September o f 1996 displayed delays in leaf senescence longer than June-brushed and unbrushed willows from 28 September to 17 October. Willows brushed in July and September o f 1996, however, did not show any differences in leaf senescence from 28 September to 17 October 1997; neither did Junebrushed and control willows (Table 1.4). L e a f Flush The timing o f mechanical brushing affected patterns o f leaf flush in the spring following treatment as evidenced by a delay in leaf flushing when brushing was performed late in the previous growing season (Table 1.5). In contrast, there was no effect on spring-leaf flush when brushing was carried out earlier in the year (Table 1.5). On 14 May 1997, willows brushed during the pre\ious September had a smaller proportion o f plants that displayed a break in dormancy than unbrushed controls or plants brushed earlier in the year. This late break in dormancy for September-brushed plants was different from all other brushing treatments and was apparent on 14 May ( p = 0.6, z = -4.472, P < 0.001), 17 May ( p = 0.7, z = -3.586, P < 0.001), and 20 May 1997 ( p = 28 Table 1.5. Proportion o f willows in each o f the 1996 brushing treatments flushing new leaves as observed on 4 different days in the spring o f 1997. Proportions sharing a common superscript are not significantly different fi'om each other. Proportion calculated on Date o f Brushing 14 June 30 July 15 Sep Controls « = 15 /I = 30 /I = 30 n = 17 11 May 0.13" 0.13" 0.00" 0.00" 14 May 1.00" 1.00" 0.20 1.00" 17 May 1.00" 1.00" 0.40 1.00" 20 May 1.00" 1.00" 0.80 1.00" 29 0.9, z = -1.826, P < 0.034). No other plants from any other brushing treatments displayed significant differences in the timing o f leaf flush due to brushing time. During the spring o f 1996 willows on this site flushed leaves between 21 May and 25 May. A ge Regressions Based on our examination o f disk annuh taken from plants brushed during the summer o f 1996, we determined that the average age o f willows in our study was 7.3 ± 1.21 years (range: 4 to 9 years). The age o f willows in the summer o f 1996 did not predict the length o f 6-week compensatory growth following brushing for the 14 June 1996 or 30 July 1996 brushing treatments (all P > 0.069). In addition, age did not predict any morphological or chemical attributes o f shoots from these brushing treatments in either the first or second winter following brushing (all P > 0.135). Discussion Shoot morphology Brushing affected the morphology o f Scouler’s willow shoots in the first 2 winters following brushing. Although we did not measure the effect o f brushing time on the production o f overall above-ground biomass, the most obvious effect o f mechanical brushing was a complete lack o f available browse immediately following treatment. The duration of this effect, when measured in terms o f compensatory shoot size, appeared to depend on when brushing occurred within the growing season. Willows brushed in June produced compensatory shoots that averaged 38 cm in length by the sixth week following brushing and 50 (1996) to 54 (1997) cm by the onset o f dormancy. In the first winter after brushing, the shoots o f willows brushed in early July averaged 26 cm long while plants from later treatments produced relatively smaller (<10 cm), thiimer shoots that were not visible above December snow packs. Therefore, depending on the timing of 30 brushing, willows may produce shoots that are significantly longer than the average shoot size o f unbrushed willows (9 cm; 1997 to 13 cm; 1996), or may have shoots too short to protrude above mid-winter snows to be accessible to moose in the first winter after brushing. All brushed willows, regardless o f when they were brushed, produced shoots in the second growing season after brushing that were larger (as measured in the second winter after brushing) than shoots produced by unbrushed willows during the same growing season. Plants brushed earliest in the previous year, however, produced the smallest o f shoots from all brushing treatments; those brushed latest in the previous year produced increasingly larger shoots (>80 cm long). Similar responses to the timing o f cutting occur in other early-successional trees and shrubs (DeBell and Alford 1972; Kays and Canham 1991) and are related to the concentration o f root reserves at the time of damage and the number o f remaining growing points to which those reserves can be allocated (Stafford 1990; Kays and Canham 1991). Changes in active meristem numbers and root reserves even appear to outweigh any changes in post-damage plant response due to the age at time o f brushing (Kays and Canham 1991). Indeed, willow age at time o f brushing (4 to 9 years) was not related to any o f the morphological or chemical attributes altered by the timing o f brushing in the present study. Plants damaged later in the growing season have relatively higher concentrations o f root reserves through fall and winter when compared to plants damaged earlier in the year (Gregory and Wargo 1986). Such plants will have less time to compensate for damage before dormancy onset, and will have fewer growing points in the following spring to which those resources may be allocated (Sennerby-Forsse and Zsuffa 1995). The allocation o f resources to relatively few growing points leads to the production o f 31 shoots in the year following damage that are larger than those o f earlier-brushed and unbrushed plants (Harrington 1984; Kays and Canham 1991). Plants brushed early in the growing season, on the other hand, tend to exhaust remaining reserves in the treatment year attempting to compensate for earlier damage (Kays and Canham 1991), which may help explain differences in willow compensatory responses between the first and second post-treatment growing season. As shoot length, mass and basal diameter increased in shoots following brushing, the diameter o f the shoot tip (apical bud) decreased. Because large shoots have proportionately more woody cortex to bud material (Danell and Bergstrom 1985), the small shoot tips o f large shoots may result fi-om an allocation o f plant resources to shoot cortex rather than meristem. This difference in morphology may be advantageous to moose browsing large shoots arising fi'om brushed plants because many o f the chemical deterrents produced by plants are concentrated in buds (Clausen et al. 1986; Herms and Mattson 1992) and a smaller bud at the shoot tip may have relatively smaller concentrations o f these chemicals. Although plants may produce thinner apical buds due to the reallocation of resources to stem materials, the significance o f a thinner apical bud for moose remains unclear. Shoot size is one o f the most important determinants o f food selection by moose (Danell et al. 19856; Risenhoover 1987a) because it influences bite size and consequently affects intake rates (Gross et al. 1993; Danell et al. 1994; Shipley et al. 1994). Cropping smaller bites requires more firequent interruptions o f chewing to refill the mouth than consuming larger bites and requires moose to spend more time and energy on cropping than processing their bites (Shipley and Spalinger 1992; Gross et al. 1993). Minimizing cropping time by selecting larger shoots may allow moose to maximize energy intake and 32 fitness while minimizing the amount o f time spent foraging away from mature forest cover. Reducing time spent away from cover is important to moose in areas where the threat from predators is high; minimizing the amount o f time required minimizes the vulnerability to predation (Weixelman et al. 1998). Reducing the time required to prehend forage may help explain why moose typically choose the largest stems (Danell et al. 19856; Schwartz 1992; Bowyer and Bowyer 1997) with the thickest basal diameters available (Haukioja and Lehtila 1992; Bowyer and Bowyer 1997) and select large compensatory shoots, such as lower stump or basal shoots originating from mechanically-damaged trees (W olff 1978; Danell et al. 19856; Bergstrom and Danell 1987a; Bergstrom and Hjeljord 1987; Bevins 1989; Suter 1993). The prehension o f large bites appears to be o f such overriding importance that moose will select larger shoots even if such shoots are relatively high in fibre and low in protein (Jia et al. 1997). Moose in north-central British Columbia even browse past current-annual shoots into less digestible 2- and 3-year old stem materials when browsing the smaller shoots o f unbrushed willows (see this study: Chapter 2), presumably in an effort to maintain larger bite diameters and higher intake rates (Shipley et al. 1994). Compromises between larger bite sizes and reductions in the protein and energy digestibility o f large shoots become key to understanding how the overall nutritional value o f browse for moose may change following brushing at different times during the growing season. Shoot Chemistry Although the energy and protein content o f shoots o f unbrushed willows in this study are comparable to willows in other studies (Wolfe et al. 1983; Eastman 1984; Bergstrom and Danell 19876; Risenhoover 19876), the large shoots o f brushed willows 33 were generally lower in digestible energy and digestible protein for 2 years after brushing. The shoots o f willows brushed in July, however, were higher in digestible protein and relatively high in digestible energy when compared to the shoots o f willows brushed at other times during the growing season or unbrushed controls in the first winter after brushing. Because the large shoots produced by willows brushed in July were high in digestible energy and digestible protein these shoots were high in nutritional value for moose, especially when the associated decreases in the tannin and lignin content o f these shoots and increases in leaf senescence were considered. Estimating the overall nutritional value o f large shoots with reduced digestible energy and digestible protein content from plants brushed at other times o f the year requires considering changes to a larger suite o f plant attributes altered by mechanical brushing. Digestible energy and digestible protein are frequently reduced in larger shoots due to the high proportion o f digestibility reducing structural materials such as cellulose contained within the shoot cortex (Eastman 1984). Reductions in digestible energy and digestible protein as well as mineral content are also known to occur with increasing shoot size in species such as birch, black cottonwood (Populus trichocarpa) and red alder {Alniis rubra). This “dilution o f nutrients” (sensu Herms and Mattson 1992) appears to be a function o f the ratio o f bark and bud material where most nutrients are concentrated, to the woody, fibrous cortex (Danell and Huss-Danell 1985; Radwan and DeBell 1988; Schwartz 1992) and may exist not only in shoots o f varying sizes but between different portions o f the same shoot. For example, digestibility decreased in Scouler’s willow (Appendix A) as well as other woody deciduous browse plants (Bailey 1967; Haveraaen and Hjeljord 1981; Palo et al. 1992; Shipley and Spalinger 1995) from the tip to the base o f the shoot where there is relatively more cortex to bark and bud material. This change 34 in bark-to-cortex ratio highlights nutritional trade-offs often faced by moose when selecting for shoots o f a larger size and is likely responsible, not only for decreased energy and protein digestibility, but for the reductions in tannin and Ugnin content we found in the shoots o f brushed willows. Concentrations o f phenolic compounds such as Lignin (Hanley 1982; Risenhoover 19876; Nellemann 1990) and tannin (Singer et al. 1994) are known to be important factors influencing browse selection in ruminants and are suspected to influence selection more than digestible energy (Bryant et al. 1983; Coley et al. 1985; Robbins et al. 1987a). Indeed, the selection o f browse by most vertebrate herbivores is determined more by the nature and quantity o f phenolic compounds than overall energy and protein content (Chapin 1980). Shipley and Spalinger (1992) report that tannin and lignin are at least as important as other chemical or qualitative properties o f shoots, and even the spatial distribution o f browse. Such characteristics appear to determine the handling time o f winter food items and can significantly affect the energy budget o f the animal (Spalinger etal. 1986). The concentration o f lignin in the shoots o f unbrushed willows was similar to the concentration o f lignin reported for other species o f willow (Oldemeyer et al. 1977; Eastman 1984; Bryant et al. 1985; Risenhoover 19876; Schwartz et al. 19886). Our results, however, indicate that mechanical brushing causes willows to produce shoots significantly lower in lignin than levels found in the shoots o f unbrushed willows. Other willows regenerating from mechanical cutting in summer were found to produce winter browse that was lower in lignin than undamaged willows (Nellemann 1990). Furthermore, the timing o f brushing appears to cause plants brushed later in the year to 35 have shoots lower or equal in lignin content to earlier-brushed plants when analyzed in both the first and second winters following brushing. Brushing later in the growing season (i.e., 1 July versus I June) reduces the effective length o f the growing season and consequently the amount o f time required by plants to compensate for damages. This truncation o f the growing season decreases the plants ability to form mature shoots and deposit structural materials such as lignin before the onset o f dormancy (Danell and Bergstrom 1985; Raitio et al. 1994). Such reductions in lignin may be important to moose because lignin specifically determines plant fibrosity and is inversely related to the relative breakdown rate o f plant material (Spalinger et al. 1986). Shoots with lower lignin may be more easily handled and processed by moose through a reduction o f the chewing effort and may allow for higher cropping rates and increased intake (Gross et al. 1993). Our index o f tannin derived firom a modified assay was used for comparative purposes and cannot be directly compared with values obtained by others using Hagerman's (1987) original technique. Despite the differences in methodology used to obtain data on changes in tannin content, our results agree with the findings o f Dutoit et al. (1990), Suter (1993) and Singer et al. (1994) who reported overall reductions in tannin content associated with increases in shoot size following various forms o f mechanical damage. For example, Suter (1993) demonstrated that large compensatory shoots originating from the stumps o f willows contained less tannin content than the smaller crown shoots o f plants on the same site. Because the production o f phenolic compoimds such as tannins and lignin represent a competition with other cellular processes for common metabolic substrates (Herms and Mattson 1992), reductions in these compounds following plant damage and 36 rapid compensatory growth can be expected. Our results indicate that although tannin content was always lower in the winter shoots o f brushed compared to unbrushed plants, tannin content was not always different between the shoots o f willows brushed at different times. Such responses suggest that changes in tannin content may be less affected by the timing o f bmshing than by brushing itself. Tannins, unlike lignin, may not affect handling time or intake rates but may inhibit animal performance through the chemical inhibition o f rumen microbes (Risenhoover et al. 1985). Tannins can also reduce amino acid and iron bioavailabiUty (Choo et al. 1981) and may work as toxins rather than digestion inhibitors (Mehansho et al. 1987). Moose possess salivary tannin-binding proteins to help neutralize the effects of dietary tarmins (Hagerman and Robbins 1993; Jimtheikki 1996). Despite this and the fact that winter browses contain relatively small amounts o f tannins (Hanley et al. 1992), moose avoid browsing winter foods higher in tannin content (Bryant and Kuropat 1980; Bryant et al. 1983; Suter 1993; Singer et al. 1994). Even the small amounts o f tannins found in dormant browses tend to quickly bind all tannin-binding proteins in moose sahva (Junetheikki 1996). Although the metabolic costs associated with the production of these tannin-binding proteins are poorly understood, reductions in protein turnover are less costly. Additionally, because moose are not able to bind tannins from all winter browses equally well (i.e. Salixpentandra; Junetheikki 1996), the ingestion o f unbound tannins could be costly for moose in terms o f reduced rumen microbe functioning (Risenhoover et al. 1985) or other deleterious effects. If, however, the costs associated with neutralizing dietary tannins are insignificant compared to potential gains in energy and protein from browsing materials higher in tannins, but also higher in digestible energy and protein, the importance o f lower tannin 37 content in the shoots o f brushed plants may be inconsequential (Hagerman and Robbins 1993). In addition, winter browse is often much lower in absolute tannin content than summer forages (Hagerman and Robbins 1993). Consequently, the 15 to 60 percent reductions (depending on when willows were brushed) in tannin found in the shoots o f willow in this study may be offset by the higher digestible energy o f these shoots. Although useful for understanding the effect o f brushing on shoot tannin content, our index does not provide a quantitative measure o f tannin that could be used to estimate the effect o f decreased tannin content on dry matter digestibility, iron or amino acid bioavailability. Even in the event that such quantitative measures had been made and used to calculate reductions in dry matter digestibility for instance (Hanley et al. 1992), the biological significance o f such measurements for moose would be suspect because not all tannins (as measured here with BSA) have physiological consequences for moose (Hagerman and Robbins 1993; Juntheikki 1996). Furthermore other phenolic compounds such as salicylic acids (SA; Raskin 1992) may be more important to the winter foraging ecology o f herbivores than tannic acids (Danell et al. 19856); just how concentrations of such compounds are altered by mechanical brushing or how moose respond to such changes are not presently understood. The majority o f our results agree with the findings o f others and indicate decreases in shoot chemical concentrations, digestibility and a number o f mineral elements (Appendix B) associated with increases in shoot size (Hjeljord et al. 1982; Owen-Smith and Novelli 1982; Vivas and Saether 1987; Sæther et al. 1989; Haukioja and Lehtila 1992). Some o f our results, however, contradict this relationship between shoot morphology and chemistry. In the first winter after brushing, willows brushed on 1 July showed high levels o f digestible protein and digestible energy when compared with other 38 brushing treatments from that year. These abnormal levels were a result o f elevated digestible dry matter, gross energy and elemental nitrogen and were higher than what would be expected from corresponding shoot sizes. The relatively high digestible protein and energy for the size o f shoots produced by willows brushed in July were likely a result o f damage-induced juvenile reversion and the relative age o f the regenerating plants as they entered plant quiescence. The shoots of these plants appear to be effectively younger than shoots o f earlier damaged or undamaged plants and appeared to subscribe to a lammas-form o f growth. Lammas shoots, which are produced by the bursting and elongation o f current-year terminal buds late in the growing season, (Kramer and Kozlowski 1979), are known to be relatively high in succulence and digestibility during the fall when compared with more mature shoots (Dietz 1971). Such shoots remain metabolically active, continue shoot production and elongation and continue to sequester nutrients from leaves late into the fall well past older plants (Kramer and Kozlowski 1979). Shoots growing from plants brushed later in the growing season, therefore, may not have had the time required to mature and deposit structural materials such as lignin or transport photosynthates to roots before fall frosts (Danell and Bergstrom 1985; Raitio et al. 1994), in contrast to plants brushed earlier in the year. Rapid freezing and the resultant trapping o f nutrients in these pre-dormant lateseason shoots may help explain some o f the chemical anomalies we detected when analyzing the dormant shoots o f July-brushed plants. L e a f Senescence Brushing resulted in a delay in leaf senescence as long as or longer than unbrushed plants in both the first and second falls after brushing. Plants brushed latest in the year showed a delay in leaf senescence in both the first and second post-treatment 39 falls as long as or longer than plants brushed earlier in the growing season. Such delays in leaf senescence following brushing, like patterns o f extended shoot growth, are strongly influenced by chemical and hormonal changes brought about in the plant from cutting (Kramer and Kozlowski 1979). Changes in leaf senescence have been attributed to a richer supply o f nutrients being supplied to a relatively smaller number o f extant shoots (Millington 1963; Danell et al. 1997) and plant rejuvenation. Such changes may help to compensate for losses in translocating resources from shoots to roots due to earlier damage, but generally last only as long as there is recurrent damage to the plant or until the root-shoot ratio has been re-established (Willard and McKell 1978). Because green, leafy forages are an important component o f many imgulate diets (Blair et al. 1980; Hanley 1982; Schwartz et al. 1988a; Danell et al. 1994), changes in the patterns o f leaf senescence following brushing may have important consequences for moose. The production and retention o f greener, younger secondary leaves on the shoots o f recently-brushed plants could be important for moose during the fall when tannins and lignin begin to concentrate in bark and senescing leaves (Palo 1984; Herms and Mattson 1992) and nutritious foods are generally reduced (Blair et al. 1980). Moose, as well as black-tailed deer {Odocoileiis hemionus columbianus), will selectively forage on darker, greener foliage over lighter green or yellowing foliage (Cowan 1945; Bergerud and Manuel 1968), presumably because greener younger leaves are typically higher in nutrients than yellowing or older senescing leaves (Kramer and Kozlowski 1979; Chapin 1980). Leaves are higher in nutritive quality than woody stems (Eastman 1984). Consequently, moose avoid browsing shoots as long as fall leaves persist (Regelin et al. 1987). Moose will even select and forage on leaf litter when twigs and other forages are 40 available in the fall landscape (Renecker and Hudson 1986), demonstrating the nutritional importance o f leaves in the fall diet o f moose. Because brushing from mid to latesummer delays leaf senescence in the first 2 falls after brushing, such treatments may provide moose with leafy materials late into the fall and help to enhance the quality o f plants for moose on late fall and early winter ranges. L e a f Flush Although plants cut during the winter are known to flush leaves early in the spring (Kindschy 1989), our results suggest that brushing plants in the early to mid-summer has no effect on leaf emergence in the first post-treatment spring. Fall brushing, however, may cause a delay in willows leaf emergence in the spring following brushing. Such delays can occur if stem and leaf material is removed before it is able to complete the export o f nutrients and other factors, such as hormones, necessary for spring growth from above to below-ground structures in the late fall (Larson 1978; Tuomi et al. 1989). Delays in spring leaf flush may protect previously damaged plants from early spring frosts, but may also reduce the availability o f leaves for moose foraging in the early spring. Although willows brushed in July did not delay leaf flush and, overall, produced browse o f better quality for moose in the first 2 winters following brushing, we did not measure the selection o f such plants over others brushed during June, August, September or unbrushed controls by moose. The study site was selected in an area where browsing by moose was minimal (<5%); this was done to reduce the confoimding effects o f browsing on plant responses to brushing. Consequently, we were unable to determine the selection o f shoots from plants brushed at different times by moose. We did, however, observe some browsing by deer {Odocoileus spp.) in late August o f 1997; browsing was 41 directed predominantly to willows brushed in June 1997 (determined during winter collections; Appendix C). Because the removal o f a small percentage o f shoots from a plant has a relatively small impact on overall plant response (Danell et al. 1994), we do not believe that browsing on experimental plants altered our results. Furthermore, browse damage by deer w^as unlikely to affect our results because deer browsed plants late in the final growing season o f the study, beyond the time that willows were able to respond to damage with compensatory growth. Although the use o f exclosures may have helped to reduce browsing on study plants, the percentages o f shoots removed by browsing likely had no effect on our overall findings. Although we do not present data on the relative use o f shoots from brushed and unbrushed willows, moose do heavily utilize browse plants in the years immediately following mechanical brushing (Hjeljord and Gronvold 1988) often times browsing up to 100% o f post-brushing current shoot growth (R.V. Rea, personal observations). Such selectivity for re-growth following mechanical brushing is also demonstrated by other ruminants such as cattle (JBos taunts) and deer (Powell and Box 1966; Siehert et al. 1980), presumably due to the increases in plant palatability, availability or accessibility associated with cutting; increases that, as the present work outlines, can be influenced by the timing o f mechanical brushing. Conclusions Mechanical brushing significantly reduces the availability o f deciduous biomass for herbivores such as moose. Although some vegetation managers retain unbrushed wildlife strips (see this study; Chapter 2) in areas being mechanically brushed in an effort to mitigate reductions in available biomass, other managers do not subscribe to such 42 strategies. In areas where wildlife strips are not retained then, brushing operations should be performed in July to ensure a supply o f nutritious browse for moose using these areas in winter. Willows brushed in July produced shoots that were o f higher value for moose in the first 2 years after brushing. Although brushing in late July did not allow willows to compensate sufficiently for shoots to be available in the first winter after brushing, the shoots o f willows brushed in early July were available above first year snows and more nutritious than the shoots o f earlier-brushed willows. Brushing in early July also corresponds well with the time suggested by vegetation managers to brush competing vegetation (i.e., willows) growing in conifer plantations because brushing at this time more effectively reduces the plants ability to compete for resources (Kays and Canham 1991). Scouler’s willow tended to show morphological responses to the timing o f brushing that are consistent with the response o f other browse plants to time o f cutting (Kays and Canham 1991). How other important browse plants specifically respond to brushing time, however, should be investigated more thoroughly because plant response can be influenced by factors such as environment (site; see this study; Chapter 2), species genetics, life history traits and plant morphology (Whitham et al. 1991; Danell and Bergstrom, 1989; Borkowska and Konopko, 1994). Although the timing o f brushing may affect some browse plants for no longer than 2 years after damage (Kays and Canham 1991), the long-term effects o f brushing time on willows appears to last much longer (see this study; Chapter 2). Mechanical brush control is practised on regenerating conifer plantations, utility rights-of-way, rail and/or roadsides as well as rangelands and unproductive forest areas 43 where vegetation control or browse production is desired. Although it has been a longheld belief that stimulating vegetative growth through cutting results in better browse for herbivores (Gysel 1957; Ferguson and Basile 1966; Roberts 1986), recent studies show that this is not invariably true (Bryant et al 1985; this study). Our results indicate that significant changes in the nutritional value o f willow may occur for at least 2 winters post-treatment, depending on the timing o f brushing. Although brushing time may affect different plant species to different degrees, determining the response o f willows to brushing time, in combination with understanding the response o f other species to time o f cutting, may help delineate a more generalized pattern o f plant response to damage. Understanding these patterns can then help to determine the consequences that such responses may have on the overall nutritional value o f these plants for moose. 44 C H A P T E R 2. CHANGES IN TH E N U TRITIONA L VALUE O F SC O U LER ’S W IL L O W (Salix scouieriana) FO LLO W IN G M ECHANICAL BRUSHING: IM PLIC A TIO N S F O R M O O SE (Alces alces). A bstract We examined the effects o f mechanical brushing on the nutritional value o f Scouler’s willow {Salix scouieriana) browse for moose {Alces alces) for 2 to 5 winters after brushing in relation to willows growing in unbrushed strips o f regenerating vegetation retained for wildhfe food and cover. We also examined the effects o f the previous year’s clipping and browsing on browse quality and compared how moose utilized willows growing in brushed and unbrushed areas in the third winter after brushing. The nutritional value o f willows varied among sites, but was altered by brushing for at least 5 winters post-brushing. The current annual shoots o f brushed willows tended to be larger, although lower in digestible protein and digestible energy, than shoots o f unbrushed willows for at least 5 years after brushing. Shoots o f brushed willows were also lower in tannin and hgnin content than the shoots o f unbrushed plants for 2 and 3 years post-brushing, respectively. Furthermore, brushed willows displayed a delay in leaf senescence for up to 4 years post-brushing. The shoots produced by willows were often unaffected by the previous winter’s cUpping/browsing intensity, but increased in size when measured in the third to fifth winter after brushing. Shoot size decreased with an increase in the previous winter’s clipping/browsing intensity when measured in the second winter after brushing. The tannin and lignin content o f shoots increased with an increase in the previous winter’s chpping/browsing in the second winter after shoot removals. With an increase in the previous winter’s clipping/browsing intensity, digestible energy and protein decreased in the third to fifth and fifth winter after brushing. 45 respectively. The average shoot diameter at the point o f browsing did not differ between the shoots o f brushed and unbrushed willows; the shoots o f imbrushed willows were browsed past current annual shoots and into older stem materials significantly more when compared to shoots firom brushed willows. Introduction Releasing conifers firom the competitive effects o f deciduous trees and shrubs through the use o f herbicides or by motor/manual removal (i.e., mechanical brushing) leads to changes in forest stands that can reduce browse biomass and cover for wildlife (Goodrum and Reid 1956; Conard 1984; Lautenschlager 1993). To reduce these impacts, resource managers occasionally retain strips o f regenerating forest within stands being managed for conifer growth that allow for some o f the needs o f wildlife (hereafter termed “wildlife strips”; Gysel 1957; Santillo 1994). Because the application o f systemic herbicides such as glyphosate (Roundup®, Monsanto Chemical Co. St. Louis, Mo.) can kill deciduous plants and reduce browse availability for as many as 9 years following treatment (Hjeljord 1994), the retention o f wildhfe strips is particularly important for ensuring a continued browse supply for animals such as moose wintering in regenerating forest stands (Santillo 1994). Although mechanical brushing does not affect the production o f browse as long as herbicides (Hjeljord and Gronvold 1988), it does reduce available biomass in the years immediately after brushing (Harkonen 1998) and can alter the nutritional value of regenerating browse for moose (see Chapter 1). Consequently, the retention o f wildlife strips also provides forage in areas where brush is being reduced mechanically, as well as chemically. These wildlife strips, however, also contain commercially important crop trees and are, therefore, eventually brushed once other deciduous shrubs and trees within 46 the site begin to recover from brushing. The amount o f time required for plants to recover from brushing and provide suitable post-brushing habitat, at least in terms o f browse quality for moose, however, is largely unknown. Brushing is known to alter the nutritional value o f current-annual-growth shoots (measured from the previous year’s growth scar to the shoot tip and hereafter termed “shoots”) of willows in the first and often second year after brushing. For example, brushing late in the growing season generally caused willow saplings to produce shoots that were larger and lower in tannin content, lignin and digestibility than the shoots o f unbrushed willows in the first 2 winters post brushing (see Chapter 1). Willows brushed in June, however, had shoots that were not significantly different in basal diameter, lignin and digestible protein content than the shoots o f unbrushed willows in the second winter following brushing, even though such differences existed during the previous winter. Feltleaf willows {Salix alexensis) cut or browsed in winter are able to compensate for damage, re-establish their root-to-shoot ratios and reassume a pre-damage form o f growth in as little as 3 years post-cutting (Bryant et al. 1985). Although the shoots o f willows may even begin to reassume attributes similar to those o f uncut plants in as little as 2 years following damage, (see Chapter 1), the length o f time required by plants brushed during the growing season to fully compensate for damage remains unclear. Because the amount o f time required for plants to fully compensate for damage is highly dependent on the intensity o f damage (Danell and Bergstrom 1989; Whitham et al. 1991), compensating for the complete removal o f above-ground biomass following summer brushing is likely to require several years o f post-damage compensatory growth (Chapin 1980). 47 Cutting and browsing willows in winter leads to increases in the concentration o f tannins as well as increases in proteins, carbohydrates, lipids, lignin and mineral elements such as phosphorus for several years following damage (Bryant et al., 1985). Increases in nitrogen, magnesium, potassium and phosphorus occur following summer clipping o f the sedge Eriophorium vaginutum (Ouellet et al. 1994). Heavy browsing and/or pruning also stimulates browse plants such as willow and birch (Betula spp.) to produce larger shoots that are higher in fibre content (W olff 1978; Borkowska and Konopko 1994; Danell et al. 1994), which are often lower in chemical deterrents to browsing, such as tannins (Danell and Bergstrom 1989; Dutoit et al. 1990; Singer et al. 1994). Other forms o f severe mechanical damage can likewise alter plant morphology and chemistry (Danell et al. 1987; Nellemann 1990; Stephenson 1995) as well as the phenology o f plant leafing in the years following injury (see Chapter 1; Danell et al. 19856; Kindschy 1989). Changes in plant physiochemistry following mechanical damage are a result o f changes brought about in the plant due to the loss o f apical dominance and plant rejuvenation (Bryant et al. 1985). Such changes lead to a reallocation o f plant resources from flower and fruit production to compensatory (sensu Belsky 1986) vegetative growth. This reallocation leads to changes in the nutritive quality o f plant tissues for herbivores (Danell et al. 1985a), but only persists imtil the effects o f juvenile reversion subside following compensatory growth and the root-to-shoot ratio rebalances (Millington 1963; Danell et al. 1997). Moose select browse based on the morphological, chemical and phenological characteristics o f plants (Schwartz 1992; Danell et al. 1994), many o f which are altered by mechanical brushing (see Chapter 1). The quality o f browse for moose is, therefore, likely to be affected by such damage-induced changes until plants have fully 48 compensated for brushing. Determining how long brushing affects the morphological, chemical and phenological attributes o f browse plants may help resource managers determine how long to retain wildlife strips before brushing them and when to schedule management strategies for adjacent forest stands where concerns for the quality o f browse for moose exist. Understanding these changes and how long they persist may also help to contrast some o f the effects o f mechanical brushing with those o f herbicide applications. To evaluate the effects o f brushing on the nutritional value o f browse and to determine how long such effects last, we selected a plant that is often brushed and is important as browse for moose in winter, Scouler’s willow (Roberts 1986; Porter 1990; Weixelman et al. 1998). We determined the effects o f brushing on I) shoot morphology, 2) shoot chemistry, and 3) patterns o f leaf senescence in the first 5 years following treatment. Additionally, we determined how moose used the shoots o f brushed and unbrushed willows in the third winter following brushing and determined if chpping and/or browsing affected the nutritional value o f willows in the year after shoot removal by clipping and/or browsing. M ethods The study area The study area comprised 6 similarly-classified (DeLong et al. 1993) regenerating pine clearcuts ~ 15 to 40 ha, near Vanderhoof, British Columbia, Canada (54°05’N, 123°55’W; Fig. 2.1). All sites were in the sub-boreal spruce dw3 biogeoclimatic zone characterized by open stands o f lodgepole pine with poorly developed shrub and herb layers, and a well-developed moss layer dominated by lichens (Meidinger and Pojar CIrcum L. Barlow Lk. Rd. 5 km, î> Fountain L. Clear Or. Expected L, Wonder L Stanley L. WelrberL. Moss Or, B r ^ d e Rd. Sturgeon Ft. Rd. British Columbia I L VANDERH( SacknerRd. Stony Cr, Fig. 2.1. Location o f study sites within the Vanderhoof Forest District. Sites are as follows: O Sackner Pit (B.C. Geographical Map Classification System-Forest district cutblock number: 93J002-001), O Triple Buck (93J002002), e North Layton (93K019-007), 0 Sawmill Pit (93K019-030), 0 Huckleberry Cut (93K019-027), and 0 Waterlily Lake (93J001-007). 50 1991). Soils on all sites are clay and/or sandy loam (D. Sommerville, Vanderhoof Forest District, Vanderhoof, B.C., Canada, personal communication). The terrain is flat to rolling and all sites are between 750 and 800 m. Three o f the sites (North Layton, Three Buck; Sackner Pit) were clear cut 12 to 15 years prior to the initiation o f the study and were mechanically brushed throughout the 1993 growing season (June to September). The other 3 sites (Sawmill Pit; Huckleberry Cut; Waterlily Lake) were clear cut 9 to 11 years prior to the beginning of the study and were mechanically brushed throughout the 1995 growing season. Wildlife strips left on the sites brushed in 1993 contained willows ~ 5 m tall at the beginning o f this study (winter 1995) while brushed willows on these sites were ~ 2 m tall. Unbrushed plants on the 1995 sites were ~ 4 m tall while brushed plants within these sites were ~ 1 m tall at the beginning o f the study. Moose densities in the surrounding area were ~ 0.5 animals per km ' at the time o f the study (D. Heard, Ministry o f Environment Lands and Parks, Prince George, B.C., Canada, personal commimication). Experimental Design During the winter o f 1995-96, we randomly selected 6 Scouler’s willow plants from mechanically-brushed areas within the plantations and 6 willows from the wildlife strips on each of the 3 sites brushed in 1993. We could not select willows from sites brushed in 1995 until the following spring, as brushed plants had not grown sufficiently following treatment to be visible above the snow. We randomly selected willows that were o f similar size and form that we had identified to species according to leaf characteristics, shoot morphology and catkin anatomy following Argus (1992). During the winter o f 1995-96 we randomly collected shoots from each o f the 6 brushed and unbrushed willows on the 1993 sites. During the winter o f 1996-97, we 51 collected shoots from the 6 brushed and 6 unbrushed willows from each o f the 6 (1993 and 1995) sites. To increase our ability to detect differences between brushed and unbrushed plants within and among sites, w e increased the sample size o f willows from each site to 13 brushed and 13 unbrushed in the spring o f 1997. We did this by randomly selecting an additional 7 willows from both brushed areas and wildlife strips. During the winter o f 1997-98 we collected shoot samples from all 13 brushed and unbrushed willows from each site. Each o f the original 6 willows within the brushed and unbrushed areas from each site was randomly selected for a clipping treatment. Two o f the 6 brushed and unbrushed willows were clipped to remove 33% of current annual shoots; 2 were clipped at 66% and 2 at 100%. Clipping intensities were imposed by chpping every third shoot, leaving every third shoot, and clipping all shoots, for 33, 66, and 100% removal o f shoots, respectively. Following winter clippings, we determined that some o f the winter shoots were unavailable for collections due to burial by snow; in addition moose had browsed other shoots after clipping treatments were imposed. Because these factors altered the targeted amount o f tissue removal intended for each specific plant, each willow plant was revisited again in the spring o f each year and the exact percentage o f shoot removal was determined. We determined percent shoot removal by counting all overwintering shoots and classifying them as clipped/browsed or undamaged and then dividing the number o f clipped/browsed shoots by the total number o f shoots on the plant. Consequently, the level o f clipping and browsing was subsequently considered to be a covariate in our analyses. 52 We collected all shoot samples while plants were dormant, in February o f 1996, and December o f 1997 and 1998 in sub-zero weather to inhibit post-sampling metabolic activities within the shoot tissues. Collections from each willow were separated in the field into separate plastic freezer bags for (1) tannin analysis and (2) all other analyses. Following collections, bags were sealed and then stored at -20 °C until analyzed. All shoots were weighed and measured for length, basal diameter and tip (apical bud) diameter. When more than 30 shoots were collected from a particular willow, however, we randomly sub-sampled 30 shoots for morphometric measures. Following morphometric measurements, all shoot material from each plant was combined, cut to ~ 10-cm lengths and dried to a constant mass (±0.1g) at 39 °C in a forced-drafr drying oven (Despatch LAD series 2-24-3; Minneapolis, MN.). Dried material was then milled with a Thomas-Wiley mill (Swedesboro, N.J.) and a 0.5-mm sieve screen, then homogenized by hand-mixing. We determined elemental nitrogen (EN), neutral detergent fibre (NDF), acid detergent fibre (ADF), acid detergent lignin (ADL) content, gross energy, digestible protein and digestible energy content using procedures outlined in Chapter 1. The amount o f tannins (hereafter referred to as “tannin content”) in willow shoots was estimated from an index o f tannin that was derived from the use o f a radial diffusion technique that we modified from Hagerman (1987; see Chapter 1). Measuring L e a f Senescence We assessed differences in leaf senescence in fall by estimating the predominant leaf colour o f each willow in the brushed and unbrushed areas o f each site on given dates, in both the falls o f 1996 and 1997. We classified plants according to procedures outlined in Chapter 1. Stage o f leaf senescence in the fall o f 1996 was determined for 12 brushed and 12 unbrushed willows randomly selected from each site by estimating the 53 predominant leaf colour o f each willow on a single day in mid-fall (8 October 1996) when most deciduous plants were in full fall colours. Stage o f senescence in the fall o f 1997 was assessed for 13 brushed and 13 unbrushed willows that were randomly selected from each site by estimating the predominant leaf colour o f each willow on a weekly basis from 5 September 1997 through 17 October 1997. Frost damage to leaves following the week o f 17 October 1997 precluded further classification o f stages o f leaf senescence on the basis o f predominant colour o f fohage. Browse Survey To determine how moose utilized the shoots o f brushed and unbrushed willows in winter we established a line transect ~ 600 m long in the spring o f 1998 through an unbrushed wildlife strip and brushed areas o f a block (Huckleberry Cut site) that had been brushed in 1995. We sampled the first 50 brushed and 50 unbrushed willows that we encountered as we moved dowm the transect that were within 5 m o f either side o f the transect. For each willow sampled we selected 5 browsed shoots by randomly selecting (using a random number table) a primary, secondary, tertiary, etc., branch until a current years shoot was selected. We then measured the diameter o f the shoot at the point of browsing and the basal diameter. According to bark characteristics we then determined whether or not each browsed shoot had been browsed beyond the current annual shoot and into older stem tissues. Measurements taken from shoots were subsequently averaged for each willow. Statistical Analyses To test the effect o f brushing on the morphology and chemistry o f willow shoots in the third winter post-brusfiing (the first winter o f the study) for the 3 sites brushed in 1993, we used a nested analysis o f variance (ANOVA; Zar 1984) with treatment (brushed 54 or unbrushed) nested within site. Following the first winter o f the study, all other morphological and chemical comparisons made between brushed and unbrushed plants, across the 6 sites, were made using a nested analysis o f covariance ANCOVA (Zar 1984) with the previous year’s percentage o f clipping and browsing used as a covariate. ANCOVA analysis was not used during the analysis o f the first winter’s data, as we did not collect information on the percentage o f shoots removed by moose in the year prior to collections. In the nested ANOVA/ANCOVA models we treated site as a random and treatment (brushed or unbrushed) as a fixed effect. When appropriate, square root, inverse or log transforms (Tabachnick and Fidell 1996) were applied to normalize data and/or homogenize the variance o f the shoot attribute being tested between treatments. Homogeneity o f variances were tested using a Levene’s test (Milliken and Johnson 1984). A Kolmogorov-Smimov test was used to test assumptions o f normality (Zar 1984). Tukey’s HSD test (Zar 1984) was used for post-hoc comparisons. A z test (Zar 1984) was used to analyze the differences in the proportions o f willows bearing predominantly green leaves on various dates in the fall o f 1996 and 1997. We used ANOVA to compare shoot diameters at the point o f browsing between shoots from brushed and unbrushed willows. We used a Kruskal-Wallis ANOVA by Ranks test (Kruskal and Wallis 1952) to determine differences in the number of shoots from brushed and unbrushed willows that had been browsed past current annual growth and into older stem materials. All analyses were conducted using Statistica (StatSoft Inc. 1997). 55 Results Shoot Morphology Mechanical brushing affected the shoot morphology o f willows for several years following treatment (Fig. 2.2). Although the length and mass o f willow (brushed and unbrushed) shoots varied among sites (Table 2.1), willows responded to mechanical brushing by producing shoots that were longer and heavier than the shoots o f unbrushed willows for at least 5 years post-brushing (Fig. 2.2, Table 2.1). The basal diameters of willow shoots also varied among sites (Table 2.1); brushed plants had shoots with thicker basal diameters in the second, third and fifth year following brushing (Fig. 2.2, Table 2.1). Shoot length, mass and basal diameter increased in the shoots o f brushed plants, but the tip diameter o f these shoots decreased in the second and third year following brushing. In the fifth year following brushing, shoot tips o f willows were significantly different among sites (Table 2.1). With an increase in the percentage o f shoots removed during the previous winter, both experimentally (clipping) and by moose (browsing), shoot mass, basal diameter and tip diameter decreased during the following year when analyzed in the second postbrushing winter. An increase in clipping/browsing during the previous winter, however, also led to an increase in shoot length and mass during the following year when analyzed in the third to fifth winters post-brushing (Table 2.1). The previous year’s clipping/browsing also caused an increase in basal diameter o f the following year’s shoots with increased clipping/browsing when analyzed in the third and fifth year postbrushing but had no effect on these attributes in other post-brushing years (Table 2.1). Year 2 Year 3 Year 4 Year 5 Year 2 Year 3 Year 4 Year 5 wmm BRUSHED I 1 UNBRUSHED E u O Z Hi 6 E 5 *6 E K 4 -I Ü Z 3 < a I 2 . E 13 13 66 13 8 I E 24 oc Hi 1! ui 66 s!?! 12 % 3 6 A l6 6 1213 13 13 I. ÛL H 1 ■ Fig. 2.2. Comparisons o f mean morphological attributes o f shoots from brushed and unbrushed willows collected for 2 to 3 winters postbrushing for 1995-brushed sites (Huckleberry Cut, Waterlily Lake, Sawmill Pit) and for 3 to 5 winters post-brushing for 1993-brushed sites, * indicates a significant effect of brushing when site effects are considered. Error bars = 1 SE, n is shown above each bar. ^ 0\ Table 2.1. P values for comparisons between brushed and unbrushed plants for morphological attributes o f shoots across all sites by year since brushing. The clipping column indicates the effect o f the percentage o f shoot removal by clipping/browsing during the previous winter when treated as a covariate in the analyses, Values in parentheses represent slopes for significant effects of the covariate. Shoot Attribute Length (cm) Year 2 Year 3 Year 3 Year 4 Year 5 (1995 Sites) (1995 Sites) (1993 Sites)^ (1993 Sites) (1993 Sites) Treatment Site Clipping Treatment Site <0.001 <0.001 0,110 <0,001 <0,001 Clipping Treatment Site Treatment Site <0,001 <0,001 <0,001 0,050 <0,001 <0.001 0,042 0,001 <0,001 <0,001 Basal Diam. (mm) <0.001 0,013 0,012 <0,001 <0,001 0.020 0,071 0,023 <0,001 Site 0,022 <0,001 <0,001 <0,001 <0,001 0,007 <0,001 0,014 0,001 0,001 <0,001 0,092 <0,001 0,223 0,041 0,254 0,268 0,029 0,036 (+0,006) (+0,011) <0,001 Clipping (+0,004) 0,025 0,014 (+0,008) (-0,005) Tip Diam, (mm) <0,001 (+0,012) (-0,024) Treatment (+0,016) (+0,105) Mass (g) Clipping 0,011 (+0,007) 0,099 0,502 0,858 0,914 0,832 0,211 0.006 0,519 (-0,007) t For the first winter’s sampling, we did not measure the previous winter’s browsing by moose. Consequently, the effect o f the covariate could not be assessed for these data. 58 Shoot Chemistry Mechanical brushing affected the chemistry o f Scouler’s willow shoots for winters following brushing (Fig. 2.3). Although tannin content in the shoots o f willow varied among sites from 3 to 5 years post-brushing (Fig 2.3), tannin content was lower in the shoots o f brushed when compared to unbrushed plants in the second year after brushing (Fig. 2.3, Table 2.2). The lignin content o f willow shoots also varied among sites for 2 to 5 years post-brushing, but was lower in the shoots o f brushed plants for as many as 3 years post-brushing (Table 2.2). The digestible energy content o f willow shoots was different between sites for at least 5 years post-brushing and was lower in the shoots o f brushed plants for 2, 3 and 5 years after brushing (Fig. 2.3, Table 2.2). Digestible protein also varied among sites for 2 to 3 years post-treatment and was lower in the shoots o f brushed when compared to imbrushed plants for at least 5 years postbrushing (Fig. 2.3, Table 2.2). As the previous year’s clipping and browsing intensity on willows increased, the tannin and lignin content o f shoots produced in the following year increased when plants were analyzed in the second year after brushing (Table 2.2). Increased clipping/browsing intensity, however, caused a decrease in digestible energy and digestible protein o f the following year’s shoots when analyzed in the third to fifth and fifth post-brushing years, respectively (Table 2.2). The previous year’s clipping/browsing had no effect on these attributes in other post-brushing years. L eaf Senescence A higher proportion o f brushed plants had predominantly green leaves when 160 1 Year 2 Year 3 Year 4 Year 5 5 5 140 ■ s 120 . I- 8 80 60-1 H Year 4 Year 3 0.25 Year 5 BRUSHED UNBRUSHED 2 0.20 • D aR 0.15 . z z Ü _l 0.10- Z z Year 2 5t 6 Z tu I— 100 - I 0.30 40-1 0 05. 20 0 4". 2 O g g 3 0 00. 3 _ I, uj z * 566 6 « 6 5 T6 2 U1 UJ i ' UJ o o „ Fig. 2.3. Comparisons o f mean chemical attributes o f shoots from brushed and unbrushed willows collected for 2 to 3 winters post-brushing for 1995-brushed sites (Huckleberry Cut, Waterlily Lake and Sawmill Pit) and for 3 to 5 winters post-brushing for 1993-brushed sites. * indicates a significant effect of brushing when site effects are considered. DM = dry matter. Error bars = 1 SB. n is shown above each bar. ^ T able 2.2. P values for comparisons between brushed and unbrushed plants for ehemicai attributes o f shoots across all sites by year since brushing. The clipping column indicates the effect o f the percentage o f shoot removal by clipping/browsing during the previous winter when treated as a covariate in the analyses. Values in parentheses represent slopes for significant effects o f the covariate. Shoot Attribute Tannin Content’ Year 2 Year 3 Year 3 Year 4 Year 5 (1995 Sites) (1995 Sites) (1993 Sites)’ (1993 Sites) (1993 Sites) Treatment Site Clipping Treatment Site Clipping Treatment Site Treatment Site Clipping Treatment Site Clipping 0,013 0,053 0,022 0,192 0,021 0,065 0,255 0,003 0,476 <0,001 0,201 0,625 0,003 0,255 0,244 0,191 0,072 0,031 0,003 0,333 0,014 0,056 0,519 0,036 0,177 0,018 0,004 0,010 0,913 <0,001 0,095 0,002 0,034 0,003 <0,001 <0,001 (+0,453) Lignin (% DM) 0,005 0,002 <0,001 (+0,001) Dig, Energy (kcal/g DM) <0,001 <0,001 0,203 Dig. Protein (% DM) 0,006 0,037 0,389 0,256 0,594 0,083 (-3.700) (-3,122) (-2,859) 0,018 0,008 0,035 0,061 0,744 0,045 0,078 0,022 (-0,010) t For the first year’s sampling, we did not measure the previous year’s browsing by moose. Consequently, the effect o f the covariate could not be assessed for these data. t Tannin content is the square o f the diameter (mm) o f the precipitation ring (see text). s 61 compared to unbrushed plants on 8 October 1996, in the second fall after brushing, on 2 o f the 3 sites brushed in 1995 (Sawmill Pit: p = 0.33, z = -1.68, P = 0.048; Huckleberry Cut: p = 0.67, z = -1.87, P = 0.031). Brushed plants from only 1 o f the 3 sites brushed in 1993, however, showed signs o f delayed leaf senescence in the fourth fall after brushing (Three Buck: p = 0.35, z = -1.74, P = 0.041). When we measured differences in the proportion o f brushed and unbrushed plants retaining green leaves on a weekly basis in the fall o f 1997, we found that brushed plants from 2 o f the 1995 sites (Huckleberry Cut and Waterlily Lake) showed a delay in leaf senescence in the third fall after brushing (Table 2.3). The third site (Sawmill Pit), however, showed no differences (all P > 0.061) in leaf senescence between brushed and unbrushed willows on any o f the sampling dates. Patterns o f leaf senescence in the fifth fall (1997) after brushing for plants brushed in 1993 were not quite as clear. O f the 3 sites brushed in 1993, only brushed plants in North Layton showed a delay in leaf senescence towards the end o f the observation period; unbrushed plants at Sackner Pit showed a delay in leaf senescence for the majority of the observation period (Table 2.3). At the Three Buck site, however, no differences (all P > 0.342) in senescence between brushed and unbrushed plants were observed throughout the observation period. Browse Survey Results The average shoot diameter at the point o f browsing was not significantly different between the shoots o f brushed (3.95 ± 0.58 mm) and unbrushed willows (3.83 ± 0.56 mm; F, gg= 1.144; P = 0.287) when analyzed in the third spring after brushing. The shoots of unbrushed willows were, however, browsed past current annual shoots and into older stem materials significantly more often (82.0%) when compared to the shoots Table 2.3. Proportion o f willows in each brushing treatment (/; = 13) bearing predominantly green leaves as calculated at weekly intervals in the fall o f 1997 for 2 sites brushed in 1993 and 2 sites brushed in 1995. Proportions sharing a common superscript within a site are not significantly different from each other as determined by separate z tests. Sampling Date: 1995 Sites 1993 Sites Sackner Pit North Layton Huckleberry Cut Waterlily Lake Brushed Control Brushed Control Brushed Control Brushed Control 19 Sep 1997 1.00" 1.00" 1.00» 0.93» 1.00 1.00 1.00 1.00 28 Sep 1997 0.75 1.00 0.80" 0.80" 0.94" 0.87" 1.00 0.77 3 Get 1997 0.69 0.93 0.53 0.20 0.94 0.47 1.00 0.77 10 Oct 1997 0.69 0.93 0.46' 0 .20'' 0.88 0.33 0.92^ 0.77^ 17 Oct 1997 0.56 0.87 0.40 0.13 0.75 0.33 0.58« 0.46« Note: No differences existed in leaf senescence between brushed and unbrushed willows on any o f the above days for the 1995 Sawmill Pit site or the 1993 Three Buek site. 63 o f brushed willows (15.6%; X \ ~ 66.67, P = <0.001). Discussion Shoot Morphology Although shoot morphology o f both brushed and unbrushed willows varied among sites, mechanical brushing caused willows to produce shoots each year that were larger than the shoots o f unbrushed plants for 4 to 5 winters post-brushing. Willows responded to brushing by producing compensatory shoots in the first and second year following damage that had smaller tip diameters, but were 4 to 5 times the length and mass o f shoots o f unbrushed willows growing in wildlife strips. In successive years, willows continued to produce current annual shoots that, on average, were larger than the shoots o f unbrushed willows, although the magnitude o f difference between shoots of brushed and unbrushed willows appeared to decline with time since brushing. By the fourth and fifth year post-brushing, the shoots of brushed willows, although still significantly larger, were on average, less than twice the size o f shoots fi'om willows growing in wildlife strips (Fig. 2.2). Mechanical damage firom clipping and browsing often affected the morphology o f shoots produced in the year following clipping/browsing (effects o f the covariate in comparisons among treatments). Generally, in the post-brushing years in which the effect o f clipping/browsing was significant (years 3 to 5; Table 2.1), an increase in the intensity (4-98% o f current annual shoots) o f clipping/browsing during the previous winter caused an increase in the size o f shoots produced during the succeeding year. In the second winter after bmshing, however, increased clipping intensity during the previous winter seemed to have the opposite effect: an increase in the previous winter’s clipping led to decreases in the size o f shoots produced in the following year. 64 Plants often respond to increased clipping intensities by producing larger shoots (Danell and Huss-Danell 1985; Bergstrom and Danell 1987a; Danell et al. 1994), but decreases in shoot sizes with increased clipping intensities may also occur when plants are severely stressed (Danell and Bergstrom 1989). This type o f response may help to explain why willows normally produced larger compensatory shoots with increased clipping intensity except when clipping was imposed in the first winter (following a single truncated growing season) after severe damage firom brushing. Vigorous vegetative regeneration and the production o f large compensatory shoots following mechanical damage (i.e., cutting, browsing) is a common growth strategy for deciduous shrubs and trees that pioneer early-seral sites (Coates and Haeussler 1986; Sennerby-Forsse andZsuffa 1995). Black cottonwood (Populus trichocarpa) saplings, for example, produce sprouts over 3 m tall and over 3 cm in diameter in the first 2 years following coppicing (DeBell and Alford 1972). This type o f growth response to damage is particularly adaptive for plants such as willows growing in young boreal forests where plant competition for canopy occupancy is intense (Aarssen and Irwin 1991; Hjalten et al. 1993) and where the highest herbivore pressures are exerted closest to the ground (Danell et al. 1987). The capacity to respond in this fashion appears to be particularly developed in plants that must frequently compensate for herbivory (Ouellet et al. 1994). Physiochemically, such responses have been attributed to an attempt by the plant to re-establish its disrupted root-to-shoot ratio (Kramer and Kozlowski 1979; Tschaplinski and Blake, 1994), and involve the allocation o f plant resources into shoot growth and crown formation in the years immediately following tissue removal. 65 Total biomass and the absolute number o f growing points on a plant decreases following removal o f stem tissues. This loss results in a disproportionate allocation o f plant resources to relatively fewer growing points which leads to the production o f fewer, but larger shoots (Bergstrom and Danell 1987a). As plants regenerate, branch and produce new meristematic tissues in the years after damage, however, the number o f above-ground growing points increases and the amount o f plant resources distributed to each new shoot becomes proportionately reduced (Danell et al. 19856). An increase in meristem numbers with time coupled with a decrease in resource allocation per meristem causes willows compensating from damage to produce more, but smaller shoots, eventually leading plants to reassiune a form o f growth and overall biomass production similar to that o f pre-damaged plants. Our data suggest that the shoots o f willow generally become smaller with time since brushing (Fig. 2.2). These data indicate that following the first 2 post-treatment years o f vigorous compensatory growth, plants produce shoots similar in morphology to those produced by more mature or unbrushed willows. Although the shoots o f brushed willows were still significantly larger than the shoots o f unbrushed willows in the fifth winter after brushing, decreases in the shoot size o f brushed willows with time since brushing suggest a trend in which both brushed (following several years of compensatory growth) and unbrushed plants eventually produce shoots o f similar size. As plants begin to regenerate and reassume a more mature, predamaged form o f growth in the years following brushing, previous differences in the chemistry and leaf phenology also begin to disappear in as little as 2 to 3 years post-brushing. 66 Shoot Chemistry Brushing promoted the growth o f large compensatory shoots that were lower in tannin, lignin, digestible energy and digestible protein for at least 2 winters after brushing (Fig. 2.3). Although in most cases the previous winter’s clipping and browsing did not affect the chemistry o f shoots produced in the succeeding year, clipping/browsing did affect shoot chemistry in some years. The tannin and lignin content o f willow shoots produced in the year after clipping, for example, increased with an increase in the previous year’s clipping/browsing intensity when analyzed in the second winter after brushing. As the previous year’s chpping/browsing intensity increased the content o f digestible energy and digestible protein found in the shoots produced in the year after chpping/browsing decreased when analyzed fi’om 3 to 5 years after brushing, respectively (Table 2.2). In those cases in which increased clipping/browsing intensity affected shoot chemistry, plant response was generally similar to bmshing (except for the positive effect on tannin and lignin content in the second post-brushing year). The added stress o f chpping/browsing on recently-bmshed willows may have promoted the production o f shoots that responded differently (in terms o f tannin and lignin content as well as morphology) to increased clipping intensity in the second post-brashing year. Plants stressed by repeated hare browsing following cutting, for example, may produce shoots that are higher in phenolic compounds such as tannins and Ugnin during winter (Bryant et al. 1985). Overall, however, we found that the tannin and lignin content remained lower in the shoots o f bmshed Scouler’s willow for 2 and 3 years following treatment, respectively, when these plants were compared to unbmshed willows growing in the wildlife strips. Reductions in the tannin and lignin content o f browse have also been 67 shown to occur following other forms o f mechanical damage (Dutoit et al. 1990; Nellemann 1990; Suter 1993; Singer et al. 1994). Reductions in the tannin and lignin content o f shoots following brushing are likely due to the allocation o f plant resources into shoot elongation over the production o f chemical defences, as the production o f tannins and lignin often compete with growth for common metabolic substrates such as phenylalinine (Herms and Mattson 1992). Alternatively, the inverse relationship between shoot size and phenolic content may be driven largely by tannin production rather than resource competition, because tannins are important antagonists and regulators o f giberellin hormones that promote stem growth and elongation (Kramer and Kozlowski 1979). The smaller shoots o f unbrushed plants may be due to the growth-inhibiting effects o f tannins, rather than tannin content being reduced due to the allocation o f resources towards growth and away from chemical defence. In either case, Scouler’s willow appears to compensate for mechanical damage by producing large compensatory shoots low in phenolic metabolites. This strategy, although not shared by all boreal plants (Bryant et al. 1983), allows for vigorous growth following damage and leads to increases in plant nutritional value with respect to chemical anti-herbivore defence (Herms and Mattson 1992). In addition to reductions in tannin and lignin content, brushing likewise resulted in reductions in the dry matter digestibility o f shoots produced following brushing. Such reductions were due to increases in structiual materials such as cellulose and hemicellulose (as determined by detergent analysis) apparently synthesized to support the growth o f large vegetative shoots. Reductions in dry matter digestibility resulted in decreases in digestible protein and digestible energy in the shoots o f post-brushed growth 68 when compared to unbmshed willows and were significant for at least 4 to 5 years following bmshing. Decreases in digestibility often occur in association with increases in shoot size and are likely a result o f the increasing proportion o f stmctural cortex to bark materials found in relatively large shoots (see Chapter I). Because homogenizing and analyzing extremely large shoots can reduce the bark to cortex ratio as well as the digestibility o f plant tissues, some authors have recommended that the more distal portions o f shoots (i.e. those that are utilized by moose), as well as whole shoots, be analyzed (Eastman 1984; Danell and Bergstrom 1985). Following such a recommendation may have revealed smaller differences between the digestible protein and energy (and possibly tannin) o f shoots from bmshed and unbmshed willows in otw study. Site Effects The shoots o f willows collected from different sites in the same post-bmshing winter often varied in morphology and/or chemistry within and between bmshing treatments (bmshed and unbmshed). The differences between the shoots o f bmshed and unbmshed willows, however, were similar among sites. Analysis o f these site effects revealed that 1 o f the 3 sites from each o f the 2 bmshing years appeared to be largely responsible for the effects o f site detected in the analyses. Bmshed and imbmshed willows from both the 1993-bmshed North Layton and 1995-bmshed Waterlily Lake sites produced shoots that were generally smaller with higher chemical contents than either of the other 2 sites bmshed in 1993 or 1995, respectively. Although differences in site elevation (~ 50 m; North Layton and Waterlily Lake) do not explain the differences detected in the shoots o f willows among sites, differences in yearly meteorological conditions between sites (i.e., localized frost pockets, amount o f precipitation, etc.) could 69 affect patterns o f plant growth and the nutritional value o f plant tissue for herbivores (Kramer and Kozlowski 1979; Bo and Hjeljord 1991). Variation in the distance between brushed and unbrushed willows across sites may have also contributed to site effects if variation in distances between such plants altered the ability o f both brushed and unbrushed willows to send and receive airborne chemical signals from nearby plants. Willows and other plants alter the emission o f certain volatile organic compounds into the air following mechanical injury (e.g., herbivore attack; Rhoades 1985). Some evidence suggests that nearby, unattacked plants can detect and respond defensively to such airborne substances by increasing phenolic concentrations and decreasing the nutritive quality o f plant tissues (Rhoades 1985). An increase in phenolic concentrations could, in tiun, inhibit shoot elongation and the production o f large shoots (Kramer and Kozlowski 1979). Plant proximity and/or root connectivity, site-to-site variation in soil nutrients, moisture and leaching as well as plant population genetics (Eastman 1984) may have also contributed to site effects. Because the nutritive quality of shoots from male and female willows can differ significantly (Danell et al. 1985a), differences in the sex ratios o f plants between sites may have also contributed to site effects. Whatever the cause, brushed and unbrushed willows growing on the Waterlily Lake and North Layton sites had relatively small shoots with increased chemical concentrations when compared to willows growing on the other 4 sites. Treatment effects, however, revealed the predictability o f plant response to brushing across all sites. These results agree with those o f Danell et al. (1997), who suggest that the morphological and chemical response o f plants to damage are o f a more generalized nature across the landscape and more important and less flexible than the external differences among various sites. 70 Replacing and increasing sampling units (willows) from year to year in an effort to reduce the loss o f inferential power prevented us from taking repeated measurements from all plants in consecutive years. Despite replacements and increases in samples, however, some plants were sampled consecutively across each year o f the study so that the effects o f clipping/browsing (as a covariate) could be assessed. These repeated measurements o f some willows throughout the study raised issues o f sampling independence. Clipping/browsing levels could not be maintained at a constant level across years for a given plant (as plants were not contained within enclosures and were therefore exposed to browsing by moose afrer clipping treatments). Consequently, we were unable to track individual willows (repeated measures analysis) across years and simultaneously include the changing effects o f clipping/browsing. Therefore, in order to assess the effects o f clipping/browsing on plant response to brushing, we did incur some loss o f independence among years. L e a f Senescence We detected yearly delays in fall leaf senescence in brushed willows for up to 4 years post-treatment on some sites. Differences in biotic, abiotic and/or meteorological factors among sites likely contributed to differences in senescence among brushing treatments between sites. Although we did not directly measure how much longer brushed plants retained green leaves over unbrushed plants, our results revealed significant differences in the proportion o f willows bearing green leaves between brushed and unbrushed willows for up to 3 weeks following the onset of leaf colouring. Delays in leaf senescence, as indicated by the retention o f green leaves in the fall, were significant in brushed willows and obvious on the landscape for up to 4 years after brushing. Similar delays in leaf senescence occurred in red willows (Salix lasiandra) in 71 response to beaver cutting (Kindschy 1989) and in little rabbitbrush (Chrysothamnus viscidiflonts) and snowberry (Symphoricarpos vaccinioides) in response to clipping (Willard and McKell, 1978). Delays in fall leaf senescence are thought to be a result o f plant rejuvenation and the reallocation o f existing nutrients to relatively fewer extant shoots (Millington 1963; Danell et al. 1997). Such delays generally only last as long as plants continue to compensate for lost shoot biomass (Willard and McKell, 1978), which according to the delays we detected in the leaf senescence o f Scouler’s willow, may be as little as 3 to 4 years post-brushing. Implications fo r Moose Generalist herbivores preferentially attack plants in which defences have been reduced by physical stress such as mechanical damage (Rhoades 1985). The African elephant {Loxodonta africand), for instance, prefers to browse on large compensatory stump shoots, and is even thought to manage the production and availability o f such shoots by selectively felling preferred browse trees (Jachmann and Bell 1985). Moose may or may not consciously farm compensatory growth (Danell et al. 1994), but nonetheless select the shoots o f plants that have been previously browsed or broken (Danell et al. 19856; Suter 1993; Singer et al. 1994) and select the largest shoots available (Fermer 1978; Risenhoover 1987a; Schwartz 1992; Bowyer and Bowyer 1997). Moose will even select large stump shoots (such as those produced following brushing) when available (W olff 1978; Bergstrom and Hjeljord 1987; Bevins 1989; Suter 1993), presumably due to changes in the morphology and chemistry o f compensatory growth that make them more palatable and easy to browse (Danell et al. 19856). Selecting large shoots also enables moose to increase bite size and intake rates per cropping effort (Shipley et al. 1994; Gross et al. 1993), which can facilitate maximizing energy intake 72 while minimizing the amount o f time spent foraging in the open, away from mature forest cover (Weixelman et al. 1998). Our data indicated that the moose browsed shoots at similar diameters, on average, when browsing the shoots o f brushed (3.95 mm) and unbrushed willows (3.83 mm). Others have described low variability in moose bite diameters on willows over wide geographic areas (Kielland and Osborne 1998). Close examination o f bark characteristics o f browsed shoots from unbrushed willows, however, revealed that the majority (82%) o f these shoots had been browsed well past current annual growth and into older 2 and 3 year old stem tissues. Moose browsed beyond the smaller current annual shoots and into older stems o f unbrushed willows presumably in an effort to maintain larger bite sizes (Shipley et al. 1994). A significantly smaller percentage (15.6%) o f shoots from brushed willows were browsed past current armual growth. Compensatory shoots produced following cutting are generally too large for moose to consume in a single bite (Huseby 1982). Consequently, moose generally do not browse past the current annual growth scar and into older stem tissues when feeding on larger stems. Although browsing the smaller shoots o f unbrushed willows alone would increase digestible energy and digestible protein intake, the incorporation o f older stem materials into each bite significantly reduces plant digestibility and nutritional value (Hjeljord et al. 1982; Wolfe et al. 1983; Vivâs and Sæther 1987). For this reason moose typically avoid browsing older stems (Danell 1983). Reasons other than overall quality per bite, however, may help to explain moose browsing beyond the current annual shoots o f unbrushed willows. Browsing in wildlife strips, for example, may afford more cover and protection from predators when compared to browsing in recently brushed and more 73 exposed areas o f the forest stand; such factors can significantly influence foraging strategies (Weixelman et al. 1998). Although moose typically select browse based on shoot size when other variables are controlled for, moose are also known to select browses that are lower in phenolic compounds such as lignin (Risenhoover 19876) and tannins (Bryant and Kuropat 1980; Bryant et al. 1983; Singer et al. 1994). Lignin and tannin content are often more important in determining browse selection by herbivores than the digestible energy (Bryant et al. 1983; Coley et al. 1985; Robbins et al. 19876). Lignin is virtually nondigestible and interferes with the microbial digestion of cellulose (Blair et al. 1980) and tannins can reduce protein and dry matter digestibility (Robbins 1987a; Robbins 1987b; Hagerman et al. 1992) or may act as toxins (Mehansho et al. 1987). Lignin content determines plant fibrosity and toughness (Spalinger et al. 1986) and may be the reason moose specifically avoid this component o f fibre (Risenhoover 19876; Nellemann 1990). Concentrations of lignin have a large influence on the nutritive quality o f shoots (Shipley and Spalinger 1992) because increases in lignin concentrations increase the handling time o f winter food items (Spalinger et al. 1986). Increases in handling time reduce the rate o f forage intake and are costly in terms o f the energy budget of the animal, because energy and nutrient intake are highly dependent on forage intake as well as nutritive quality (Spalinger et al. 1986). Although increased handling times and reductions in forage intake are, therefore, likely to occur when moose browse the more lignified, smaller shoots and older stems o f unbrushed willows, these problems are not likely to occur when moose browse the shoots o f brushed willows in the first 3 years post-brushing. 74 In addition to lignin, moose specifically avoid tannins (Bryant and Kuropat 1980; Bryant et al. 1983; Suter 1993; Singer et al. 1994) and m ay avoid other phenolic compounds such as salicylates (salicylic acid; Raskin 1992) when browsing willows, although the role o f such compounds in the winter foraging ecology o f the moose remain unclear. Moose appear to avoid tannin-containing browses even though they produce salivary-tannin-binding proteins to help neutralize the effects o f dietary tannins (Hagerman and Robbins 1993; Juntheikki 1996). Such avoidance may be related to the inability o f moose to bind all tannins firom all deciduous browses (including some willows; Junetheikki 1996). Although dormant browse stems contain relatively little tannin when compared to other forage items (Hagerman and Robbins 1993), these small amounts o f tannin are sufficient to bind all o f the tannin-binding protein from moose saliva (Junetheikki 1996). The production o f shoots with even slight reductions in tannin content, such as those produced in the second year after brushing (4-24 % on average; Fig. 2.3) might appear, therefore, to potentially increase the nutritional value o f willows for moose. Although we cannot estimate absolute tannin levels in winter browse, these 4-24% decreases likely represent very small absolute differences in tannin content between brushed and unbrushed willows. In addition ungulates appear to be able to compensate for even relatively high levels o f tannins with increased digestible energy intake (Merril 1994). Consequently, increased energy intake rates achieved by simultaneously cropping several smaller shoots fi'om unbrushed willows could outweigh apparent increases in the nutritional value o f shoots due to reduced tannin and lignin content. Estimating intake rates o f moose browsing brushed and unbrushed willows would be necessary to completely understand trade offs in absolute nutritional value. 75 Although shoots o f deciduous trees and shrubs constitute the majority o f the moose winter diet (Pierce 1983; Risenhoover 1989), moose select leaves rather than shoot materials in the fall and early winter as long as leaves are available (Regelin et al. 1987). Moose will even forage on fallen leaves in the fall (Renecker and Hudson 1986) even though they typically select the greenest foliage available (Bergerud and Manuel 1968). Because the availability o f green leaves in the fall landscape are considered important to moose and other herbivores (Chapin 1980; Schwartz et al. 1988a), delays in fall leaf senescence in the years following brushing act to increase the nutritional value o f willows for moose. Conclusions Brushing altered the overall nutritional value o f browse for moose for several years after brushing. Brushed willows may not have been available in the first winter after brushing (depending on the timing o f brushing; see Chapter 1), but these plants retained green leaves in the second fall (and likely the first; see Chapter 1) and had large shoots with reduced tannin and lignin content in the second winter following brushing. In the third winter following treatment, brushed willows had shoots with less lignin that were larger, but contained more structural compounds (i.e., cellulose), that reduced digestible energy and digestible protein compared to shoots o f unbrushed plants. Brushed willows continued to show signs o f delayed leaf senescence in the third fall after brushing. In the fourth and fifth years following brushing, reductions in lignin content and delays in leaf senescence were no longer apparent and browse shoots continued to remain lower in digestible protein and digestible energy but larger in size. The previous winter’s clipping/browsing intensity often had a significant effect on the nutritional value o f shoots produced by willows in the year following 76 clipping/browsing. Because the intensity and effect o f browsing on each plant varied by year, however, we were unable to assess the effects o f brushing on individual plants from year to year. The confounding effects o f browsing, therefore, reduced our ability to track individual plants and reduced our sampling independence from year to year. Although this likely had little bearing on our findings within years, limiting browsing variability from year to year through the use o f exclosures would allow for repeated measurements and the tracking o f individual plants across years. In turn, such an approach would allow for closer tracking o f the effects o f brushing from one post-brushing year to the next. Irrespective o f site, willows responded to brushing by producing shoots that were thicker, longer and heavier than the shoots o f unbrushed willows for several years after brushing. Willows also showed consistent reductions in the chemical contents o f large shoots produced after brushing, although the number o f post-brushing years that these reductions were apparent differed depending on the particular chemical attribute in question. Aside from reductions in overall available biomass, the most obvious effect o f brushing is an increase in the size o f shoots produced by willows in the years following brushing. Although large shoots are typically lower in digestible energy and protein, moose are particularly adapted for prehending large bites (e.g., willow shoots; Shipley et al. 1994) and select larger rather than smaller shoots regardless o f overall nutritional quality (Jia et al. 1997). Weixelmann et al. (1998) found no correlation's between the use o f Scouler’s willow and protein content or dry matter digestibility by moose and speculated on the potential role o f plant secondary compoimds in their diet selection. Despite these findings potential reductions in overall nutritional value due to decreased protein and energy digestibility following brushing could outweigh gains in overall value 77 due to increased shoot size, decreased tannin and lignin content and delayed leaf senescence. Cafeteria-style feeding trials are likely the most accurate way to predict moose preferences and intake rates and to understand trade offs faced by moose when selecting for the shoots o f brushed versus unbrushed willows. Furthermore, extensive browse surveys within experimentally controlled stands would help to delineate the use o f brushed and unbrushed areas without removing natural influences known to drive moose foraging strategies (i.e., snowdepth, predator threats, etc.). Combining such information with our present findings may help resource managers compare the effects o f brushing with herbicide applications and determine the overall impacts o f brushing on browse quality for moose. This, in turn, may aid managers in the development o f vegetation management strategies for wildlife strips and adjacent stands. Not all woody deciduous plants respond to mechanical damages (i.e., brushing) in the same manner as Scouler’s willow (Ouellet et al. 1994; Alpe et al. 1999). Other woody plants such as aspen (Populus spp.), birch {Betula spp.) and ash (Sorbus spp.), however, appear to respond to mechanical damage (at least morphologically) in a fashion similar to that o f willow (Coates and Haeussler 1986). For resource managers interested in the specific effects o f brushing on particular browse plants, further experimentation on browse species o f interest should be conducted because a combination o f species, site characteristics, and previous browse history, will likely influence plant response. Furthermore, the timing o f brushing (see Chapter 1) influences plant response and the resulting quality o f brow se for moose. Obviously each factor associated with a particular influence on plant response to brushing must be investigated to fully understand the overall implications of brushing on 78 the production and quality o f browse. Many questions regarding species-specific plant responses to mechanical brushing and the implications that brushing may have for animals other than moose remain unanswered. Despite this, resource managers can begin to use these results in combination with future findings to manage brush, not only for conifer production, but for wildhfe such as moose that are often found wintering in young managed stands. 79 CHAPTERS. CONCLUSIONS AND RECOMMENDATIONS Stand-tending activities such as mechanical brushing affect the availability and quality o f browse for moose in winter. The degree to which plants are affected by brushing, however, depends on several factors including plant species, the timing o f damage and site characteristics. These factors influence the ability o f plants to compensate for such damages. Mechanical brushing must, therefore, be considered in relation to moose ecology as well as conifer production. This research examined the effects o f mechanical brushing on the nutritional value o f an important browse for moose in winter: Scouler’s willow. We examined the effects o f the timing o f brushing on the nutritional value o f regenerating browse shoots for the first 2 winters after brushing and determined how the nutritional value o f shoots was affected in the first 5 years following brushing at the field level (without a timing component). Specifically, data were collected to help to identify: 1) the most appropriate time to brush to increase the nutritional value o f browse for moose and; 2) how brushing affected the nutritional value o f willows in the first 5 years immediately following brushing. To assess the effects o f the timing o f brushing on the nutritional value o f Scouler’s willow, we brushed willows at 6- and 4-week intervals during the growing seasons o f 1996 and 1997, respectively. We then recorded plant response in terms o f compensatory shoot growth, leaf senescence and leaf flush and collected samples from willows in the first and second winters following brushing to determine the effects o f different brushing treatments on the morphology and chemistry o f shoots. To determine the long-term effects o f mechanical brushing and the effects of clipping and browsing on the nutritional value o f willow following brushing, we compared the quality o f browse 80 produced by brushed and unbrushed willows on 6 sites firom 2 to 5 years post-brushing. We also conducted a browse survey to assess the utilization o f brushed and unbrushed willows. We considered the nutritional value o f willows for moose to increase if willows retained leaves longer and produced shoots that were large and low in tannin and lignin content. We also considered the absolute nutritive value o f shoots to increase with an increase in digestible energy and digestible protein content. Our research indicated that mechanical brushing altered morphological, chemical and phonological attributes of Scouler’s willow that are known to influence forage selection by moose for a number o f years following brushing. In general, willows responded to brushing by showing signs o f delayed leaf senescence and by producing large compensatory shoots lower in tannin, lignin, digestible protein and digestible energy than the shoots o f unbrushed plants in the years following brushing. The specific response o f willows to brushing depended on the timing o f brushing. Willows brushed in July compensated for brushing by producing new growth that, overall, was more nutritious for moose in the first 2 post-brushing years when compared to plants brushed at other times dturing the growing season or unbrushed controls. Plants brushed earlier than July did not delay leaf senescence as long and had shoots that were less nutritious than July-brushed willows in the first and second post-treatment year. . Although the shoots o f willows firom plants brushed after July were comparable in quality to the shoots o f July-brushed willows in the second post-treatment year, these willows had shoots that were unavailable in the first winter after brushing and showed delayed leaf flush in the first spring following brushing. July-brushed plants also had shoots larger and more nutritious for moose in terms of digestible energy and protein, taimin and 81 lignin content and showed delayed leaf senescence longer than unbmshed willows in the first 2 winters post-bmshing Although chemical, and to some extent morphological, attributes o f willows in both bmshed areas and unbmshed wildlife strips varied among sites, the trends in plant response to bmshing were similar among sites. Digestible protein and digestible energy were lower in the shoots o f bmshed when compared to unbmshed willows for 4 to 5 years following bmshing. The shoots o f bmshed willow were, however, lower in tannin and lignin content for 2 and 3 winters following bmshing, respectively. Additionally, bmshed plants produced shoots that were larger than the shoots o f unbmshed willows for at least 4 to 5 years following bmshing and showed delayed leaf senescence longer than unbmshed plants for up to 3 consecutive falls post-bmshing. The intensity of the previous winter’s clipping and browsing sometimes affected the nutritional value o f shoots produced during the following year. When the effect o f clipping/browsing was significant, the effects were generally similar (except in the second post-bmshing year) to the effect o f bmshing; shoot size increased and shoot chemistry decreased with an increase in the previous year’s clipping intensity. Because moose select browse in late fall and winter based predominantly on shoot size, as well as tannin and lignin content and the presence o f leaves, the changes occurring in willows following bmshing appear to increase the amount o f food normally selected by moose. O f the suite of shoot attributes altered by mechanical bmshing, increases to the size o f willow shoots in the years following mechanical bmshing may be the most important to moose. Reductions in the lignin and tannin content o f these shoots may help to increase their digestibility and palatability and delays in leaf senescence may provide moose with leafy, nutritious forage later into the fall; larger shoots allow moose to take 82 larger bites and reduce the occurrence o f browsing past current annual growth. In this way, large shoots facilitate an increase in the intake o f current annual growth and a reduction o f older, less nutritious 2- and 3-year-old materials into each bite that often occurs when moose browsed the smaller shoots o f unbrushed willows. Willows brushed in July produced large compensatory shoots, which were low in tannin and lignin and high in digestible protein relative to willows brushed at other times or unbrushed willows. Willows brushed in July also delayed leaf senescence and retained green leaves late in the fall. Although brushing (regardless o f timing) reduces overall available browse biomass, changes due to brushing in July led to increases in the overall nutritional value o f browse for moose in the first 2 years post-brushing. We, therefore, recommend brushing in July in areas where concerns for an uninterrupted supply o f browse for moose exist. The long term nutritional value o f post-brushing growth remains questionable due to reductions in the energy and protein digestibility o f compensatory growth. These reductions were associated with increases in the amount o f structural materials (i.e., cellulose) which we found in the large compensatory shoots produced following brushing. Although larger shoots aid moose in taking larger bites, understanding the trade offs between increases in shoot (bite) size and reductions in digestibility must be more closely investigated to completely understand the biological significance o f changes in browse quality for moose following brushing. Only after conducting cafeteria-style feeding trials and, determining in situ intake rates, will the actual value o f shoots fi-om brushed and unbrushed plants for moose become evident. Understanding these factors, in combination with the present findings, will help to provide resource managers with information necessary to manage wildlife strips and adjacent forest stands not only for 83 conifers but for the production o f browse for moose. Such information may also be useful to managers considering supplanting chemical methods o f brush control with mechanical methods. Our findings regarding the effects o f mechanical brushing on the nutritional value o f browse for moose is based solely on the response o f Scouler’s willow to brushing. Other research into plant morphological responses to cutting, however, suggests that willows respond to damage in a manner typical o f other early-successional woody plants. The findings o f this study and the implications to the nutritional value o f browse for moose are, therefore, likely to be applicable to plants other than willows growing in regions other than the SBS dw3 sub-zone. The application o f our findings to other areas is particularly pertinent when considering changes in shoot morphology following brushing; shoot size increased following brushing regardless o f site. Although changes in shoot chemistry often depended on site, the importance o f brushing to moose is likely more closely related to the production o f larger shoots for several years by brushed plants than changes in shoot chemistry in the years immediately following brushing. For this reason, the implications of brushing on the browse quality o f shoots for moose in winter are likely to be similar across several different areas. Further research into plant response to mechanical brushing will continue to strengthen our understanding o f the effects that brushing can have on the production o f browse for animals such as moose in the years following brushing. For example, because the intensity o f browsing on plants changed from year to year, tracking the effects o f brushing on individual plants from one year to the next was impossible. The construction of exclosures, in this case, could help to limit the variability o f tissue removal o f 84 individual plants between years and allow for repeated measures and an assessment o f brushing on individual plants from one post-brushing year to the next. Using such measures will help to increase our understanding o f how plants respond to such practices imposed at different times, across different plant species and geographical boundaries. This in turn, will aid in the formulation o f appropriate management strategies for brush control in areas where concerns for the quantity and quality o f browse for moose exist. Implementing such strategies will then allow for a more integrated approach to controlling vegetation in which considerations are made not only for conifer production, but for animals such as moose wintering and foraging in recently brushed stands. 85 LITERATURE CITED Aarssen, L.W. and D.L. Irwin. 1991. What selection: herbivory or competition? Oikos 60: 261-262. Alpe, M.J., J.L. BCingery and J.C. Mosley. 1999. Effects o f summer sheep grazing on browse nutritive quality in autumn and winter. Journal o f Wildlife Management 63: 346-354. Argus, G.W. 1992. Key to willows in the northern interior o f B.C. In Plants o f northern British Columbia. Edited by A. MacKinnon, J. Pojar and R. Coupe. Lone Pine Publishing, Vancouver, B.C. pp. 54-77. Atmospheric Environment Service. 1993. Canadian Climate Normals 1961-1990 British Columbia. Canadian Climate Program, Atmospheric Environment Service, Environment Canada, Downsview, Ontario. Bailey, J.A. 1967. Sampling deer browse for crude protein. Journal o f Wildlife Management 31: 437-442. Belevsky, G.E. 1981. Food plant selection by a generalist herbivore: the moose. Ecology 62: 1020-1030. Belsky, A.J. 1986. Does herbivory benefit plants? A review o f the evidence. American Naturalist 127: 870-892. Bergerud, A.T. and F. Manuel. 1968. Moose damage to balsam fir-white birch forests in central Newfoundland. Journal o f Wildlife Management 32: 729-746. Bergstrom, R.K. and K. Danell. 1987a. Effects o f simulated winter browsing by moose on morphology and biomass o f two birch species. Journal o f Ecology 75: 533544. 86 Bergstrom, R.K. and K. Danell. 19876. Moose winter feeding in relation to morphology and chemistry o f six tree species. Alces 22: 91-112. Bergstrom, R. and O. Hjeljord. 1987. Moose and vegetation interactions in N.W. Europe and Poland. Swedish Wildlife Research Supplement 1: 213-228. Bernstein, A. 1978. Using a chainsaw to control brush. Journal o f Forestry 76: 474-475. Bevins, J.S. 1989. Activity patterns o f moose on the Kenai Peninsula, Alaska, as influenced by season and food availability. M.Sc. Thesis. University o f Alaska, Fairbanks, Alaska, U S A. Biggs, W.G. and M.E. Walmsley. 1988. Research requirements to determine the effects o f silvicultural herbicides on wildlife habitat: a problem analysis. Canadian Forest Service and The Ministry o f Forests Research Branch, F.R.D.A. Report 040. Victoria, B.C. Blair, R.M., H.L. Short, L.F. Biukhart, A. Harrell and J.B. Whelan. 1980. Seasonality o f nutrient quality and digestibility o f three southern deer browse species. USD A Forest Service Research Paper. SO-161. Southern Forest Experiment Station, New Orleans, LA. Bo, S. and O. Hjeljord. 1991. Do continental moose ranges improve during cloudy summers? Canadian Journal o f Zoology 69: 1875-1879. Borkowska, A. and A. Konopko. 1994. Moose browsing on pine and willow in the Biebrza Valley, Poland. Acta Theriologica 39: 73-82. Bowyer, J. and R.T. Bowyer. 1997. Effects o f previous browsing on tlie selection o f willow stems by Alaskan moose. Alces 33: 11-18. Bryant, J.P., F.S. Chapin and D R . Klein. 1983. Carbon nutrient balance o f boreal plants in relation to vertebrate herbivory. Oikos 40: 357-368. 87 Bryant, J.P. and P.J. Kuropat 1980. Selection of winter forage by subarctic browsing vertebrates-the role o f plant chemistry. Annual Review o f Ecology and Systematics 11: 261-285. Bryant, J.P., G.D. Wieland, T. Clausen and P. Kuropat. 1985. Interactions o f snowshoe hares and feltleaf willow in Alaska. Ecology 66:1564-1573. Chapin, F.S. 1980. Effect o f clipping upon nutrient status and forage value o f tundra plants in arctic Alaska. In Proceedings o f the 2"** International Reindeer/Caribou Symposium. Edited by Reimers, E., Gaare, E. and S. Skjenneberg. Directorate for Wildlife and Fish, Trondheim, Norway. Choc, G.M., P.G. Waterman, D.B. McKey and J.S. Gartlan. 1981. A simple enzyme assay for dry matter digestibility and its value in studying food selection by generalist herbivores. Oecologia 49: 170-178. Clausen, T.P., P.B. Reichardt and J.P. Bryant. 1986. Pinosylvin and pinosylvin methyl ether as feeding deterrents in green alder. Journal o f Chemical Ecology 12: 21172131. Coates, D. and S. Haeussler. 1986. A preliminary guide to the response o f major species o f competing vegetation to silvicultural treatments. B.C. Ministry o f Forests Land Management Handbook No. 9. B.C. Ministry o f Forests Research Branch, Victoria, B.C. Coley, P.D., J.P. Bryant and F.S. Chapin. 1985. Resource availability and plantantiherbivore defenses. Science 230: 895-899. Conard, S.G. 1984. Forest vegetation management in British Columbia: problem analysis. Ministry o f Forests Research Branch, Victoria, B.C. 88 Cowan, I.M. 1945. The ecological relationships o f the food o f the Columbia black-tailed deer, Odocoileus hemionus columbianus (Richardson), in the coast forest region o f southern Vancouver Island, British Columbia. Ecological Monograph 15: 109139. Danell, K. 1983. Shoot growth o f Betula pendula and B. pubescens in relation to moose browsing. Alces 18: 197-209. Danell, K. and R. Bergstrom. 1985. Studies on interactions between moose and two species o f birch in Sweden: a review. In Proceedings-Symposium on PlantHerbivore interactions. Edited by F.D. Provenza, J.T. Flinders and E. Durant McArthur. U.S. Department o f Agriculture Forest Service Intermountain Research Station Publication, Ogden, UT. pp. 48-57. Danell, K. and R. Bergstrom. 1989. Winter browsing by moose on two birch species: impact on food resources. Oikos 54: 11-18. Danell, K., R. Bergstrom, and L. Edenius. 1994. Effects o f large mammalian browsers on architecture, biomass, and nutrients of woody plants. Journal o f Mammology 75: 833-844. Danell, K., T. Ehnqvist, L. Ericson and A. Salomonson. 1985a. Sexuality in willows and preference by bark-eating voles: defence or not? Oikos 44: 82-90. Danell, K., T. Ehnqvist, L. Ericson and A. Salomonson. 1987. Are there general patterns in bark-eating by voles on different shoot types from woody plants? Oikos 50: 396-402. Danell, K., E. Haukioja and K. Huss-Danell. 1997. Morphological and chemical responses o f mountain birch leaves and shoots to winter browsing along a gradient o f plant productivity. Ecoscience 4; 296-303. 89 Danell, K. and K. Huss-DanelL 1985. Feeding by insects and hares on birches earlier affected by moose browsing. Oikos 44: 75-81. Danell, K., K. Huss-Danell and R. Bergstrom. 19856. Interactions between browsing moose and two species o f birch in Sweden. Ecology 66: 1867-1878. DeBell, D.S. and L.T. Alford. 1972. Sprouting characteristics and cutting practices evaluated for cottonwood. Tree Planters Notes 23(4):l-3. DeLong, C., D. Tanner and M.J. Jull. 1993. The field guide for site identification and interpretation for the southwest portion o f the Prince George forest region. B.C. Ministry o f Forests Land Management Handbook No. 24. B.C. Ministry o f Forests Research Branch, Victoria, B.C. Dietz, D R. 1971. Nutritive value o f shrubs. In Wildland shrubs - their biology and utilization. Edited by C M. McKell, J.P. Blaisdell and J R. Goodwin. U.S.D.A. F.S. Tech. Rep. INT-l, 1972. Intermountain Forestry and Range Experiment Station, Ogden UT. pp. 289-330. Dutoit, J.T., J.P. Bryant and K. Frisby. 1990. Regrowth and palatabilty o f Acacia shoots following pruning by Afiican savanna browsers. Ecology 71: 149-154. Eastman, D.S. 1984. Seasonal changes in crude protein and lignin o f ten common forage species o f moose in north-central British Columbia. Alces 19: 36-70. Ferguson, R.B. and J.V. Basile. 1966. Topping stimulates Bitterbrush twig growth. Journal o f Wildlife Management 30: 839-841. Goering, H.K. and P.J. Van Soest. 1970. Forage fiber analyses. U.S.D A. Agriculture Handbook 379: 1-20. Agricultural Research Service, U.S. Government Printing Office, Washington, D C., U.S.A. 90 Goodrum, P.D. and V.H. Reid. 1956. Wildlife implications o f hardwood and brush controls. In Transactions North American Wildlife and Natural Resources Conference pp. 127-140. Gregory, R.A. and P.M. Wargo. 1986. Timing o f defoliation and its effect on bud development, starch reserves, and sap sugar concentration in sugar maple. Canadian Journal o f Forest Research 16: 10-17. Gross, J.E., L.A. Shipley, N.T. Hobbs, D.E. Spalinger and B.A. Wunder. 1993. Functional response o f herbivores in food-concentrated patches: tests o f a mechanistic model. Ecology 74: 778-791. Gysel, L.W. 1957. Effects o f different methods o f releasing pine on wildlife food and cover. Down to Earth 13(2): 2-3. Hagerman, A. E. 1987. Radial diffusion method for determining tannin in plant extracts. Journal o f Chemical Ecology 13: 437-449. Hagerman, A.E., C.T. Robbins, Y. Weerasuriya, T.C. Wilson and C. McCarthur. 1992. Tannin chemistry in relation to digestion. Journal o f Range Management 45: 5762. Hagerman, A.E. and C.T. Robbins. 1993. Specificity o f tannin-binding salivary proteins relative to diet selection by mammals. Canadian Journal o f Zoology 71: 628-633. Hanley, T.A. 1982. The nutritional basis for food selection by ungulates. Journal o f Range Management 35: 146-150. Hanley, T.A., C.T. Robbins, A.E. Hagerman and C. McArthur. 1992. Predicting digestible protein and digestible dry matter in tannin-containing forages consumed by ruminants. Ecology 73: 537-541. 91 Harkônen, S. 1998. Effects o f silvicultural cleaning in mixed pine-deciduous stands on moose damage to Scots pine {Pinus sylvestris). Scandinavian Journal o f Forest Research 13: /« press. Harkônen, S., R. Heikkila, W.E. Faber a n d . Pehrson. 1998. The influence o f silvicultural cleaning on moose browsing in young scots pine stands in Finland. Alces: 34: 409-422. Harlow, R.F. and L.K. Halls. 1972. Response o f yellowpoplar and dogwood seedlings to clipping. Journal o f Wildlife Management 36: 1076-1080. Harrington, C.A. 1984. Factors influencing initial sprouting o f red alder. Canadian Journal o f Forest Research 14: 357-361. Haukioja, E. and K Lehtila. 1992. Moose and Birch: how to live on low-quality diets. Tree 7(1): 19-22. Haveraaen, O. and O. Hjeljord. 1981. Improvement o f moose {Alces alces) winter range by cutting and fertilization o f birch {Betula pubescens var. tortuosa) in Gausdal Vestfjell. Scientific reports o f the Agricultural University o f Norway Vol. 60 NR. 20. Vardings Trykkeri, Sarpsborg, Norway. Hellinga, D., O. Oftedal and B. Henen. 1998. Nitrogen analysis: Comparison o f carbonhydrogen-nitrogen (CHN) gas analysis to Kjeldahl procedures. Proceedings o f the Comparative Nutrition Society pp. 79-82. Herms, D A . and W.J. Mattson. 1992. The dilemma o f plants: to grow or defend. The Quarterly Review o f Biology 67: 283-335. Hjalten, J., K. Danell and L. Ericson. 1993. Effects o f simulated herbivory and intraspecific competition on the compensatory ability o f birches. Ecology 74: 1136-1142. 92 Hjeljord, O. 1994. Moose {Alces alces) and mountain hare {Lepus timidus) use o f conifer plantations following glyphosate application. Norwegian Journal o f Agricultural Sciences 8: 181-188. Hjeljord, O. and S. Gronvold. 1988. Glyphosate application in forest-ecological aspects VI. Browsing by moose {Alces alces) in relation to chemical and mechanical brush control. Scandinavian Journal o f Forest Research 3: 115-121. Hjeljord, O., Sundstol, F. and H. Haagenrund. 1982. The nutritional value o f browse to moose. Journal o f Wildlife Management 46: 333-343. Huseby, K. 1982. Production o f winter grazing for moose after cultivation o f old flood forest. Report from the consulate for Wildlife in Northland province, Bodo, Norway, Nr 2. lip. Jachmann, H. and R.H.V. Bell. 1985. Utilization by elephants o f the Brachystegia woodlands o f the Kasimgu-National-Park, Malawi. Afiican Joiunal o f Ecology 23: 245-258. Jia, J., P. Niemela, M. Rousi and S. Harkônen. 1997. Selective browsing o f Moose {Alces alces) on Birch {Betula pendula) clones. Scandinavian Journal o f Forest Research 12: 33-40. Juntheikki, M-R. 1996. Comparison o f tannin-binding proteins in saliva o f Scandinavian and north American moose {Alces alces). Biochemical Systematics and Ecology 24: 595-601. Kauppi, A., P. Rinne and A. Ferm. 1988. Sprouting ability and significance for coppicing o f dormant buds on Betula pubescens Ehrh. stumps. Scandinavian Journal o f Forest Research 3: 343-354. 93 Kays, J.S. and C.D. Canham. 1991. Effects o f time and frequency o f cutting on hardwood root reserves and sprout growth. Forest Science 37: 524-539. Kielland, K. and T. Osborne. 1998. Moose browsing on feltleaf willow: optimal foraging in relation to plant morphology and chemistry. Alces 34: 149-155. Kindschy, R.R. 1989. Regrowth o f willow following simulated beaver cutting. Wildlife Society Bulletin 17: 290-294. Kramer, P.J. and T.T. Kozlowski. 1979. Physiology o f woody plants. Academic Press Inc. San Diego, California, U.S.A. Kruskal, W.H. and W.A. Wallis. 1952. Use o f ranks in one-criterion variance analysis. Journal o f the American Statistical Association 47: 583-621. Larson, M.M. 1975. Pruning northern red oak nursery seedlings: effects on root regeneration and early growth. Canadian Journal o f Forest Research 5: 381-386. Larson, M.M. 1978. Effects o f late season defoliation and dark periods on initial growth o f planted northern red oak seedlings. Canadian Journal o f Forest Research 8: 6772. Lautenschlager, R.A. 1993. Response o f wildlife to forest herbicide applications in northern coniferous ecosystems. Canadian Journal o f Forest Research 23: 22862299. Machida, S. 1979. Differential use o f willow species by moose in Alaska. M.Sc. Thesis. University o f Alaska, Fairbanks, Alaska, U.S.A. Mehansho, H., D.K. Ann, L.G. Butler, J.C. Rogler and D M. Carlson. 1987. Induction o f proline-rich proteins in hamster saliva glands by isoproterenol treatment and an unusual growth inhibition by tannins. Journal o f Biological Chemistry 262: 12344-12350. 94 Meidinger, D. and J. Pojar. 1991. Ecosystems o f British Columbia. Special Report No. 6. Ministry o f Forests Research Branch, Victoria, B.C. Merril, E.H. 1994. Summer foraging ecology o f elk in the Mount St. Helens blast zone. Canadian Journal o f Zoology 72: 303-311. Milliken, G.A. and D.E. Johnson. 1984. Analysis o f messy data: Volume I. Designed experiments. Van Nostrand Reinhold, Company. New York, New York, U.S.A. Millington, W.F. 1963. Shoot tip abortion in Ulmus Americana. American Journal of Botany 50: 371-378. Nellemann, C. 1990. Vegetation management to improve moose browse in interior Alaska. M.Sc. Thesis, Agricultural University o f Norway, As, Norway. Oldemeyer, J.L., A.W. Franzmann, A.L. Bnmdage, P.D. Ameson and A. Flynn. 1977. Browse quality and the Kenai moose population. Journal o f Wildlife Management 41: 533-542. Ouellet, J.P., S. Boutin and D C. Heard. 1994. Responses to simulated grazing and browsing o f vegetation available to caribou in the arctic. Canadian Journal of Zoology 72: 1426-1435. Owen-Smith, N. and P. Novelhe. 1982. What should a clever ungulate eat? American Naturalist 119: 151-178. Palo, R.T. 1984. Distribution o f birch {Betula spp.), willow {Salix spp.) and poplar (Populus spp.) secondary metabolites and their potential role as chemical defense against herbivores. Journal o f Chemical Ecology 10:499-520. Palo, R.T., R. Bergstrom and K. Danell. 1992. DigestibiUty, distribution o f phenols, and fiber at different twig diameters o f birch in winter. Implication for browsers. Oikos 65: 450-454. 95 Pastor, J. and R J . Naiman. 1992. Selective foraging and ecosystem processes in boreal forests. American Naturalist 139: 690-705. Pella, E. and B. Colombo. 1973. Study o f carbon, hydrogen and nitrogen determination by combustion gas chromatography. Mikrochimica Acta 5:697-719. Penner, D.F, 1978. Some relationships between moose and willow in the Fort Providence, N.W.T. area. M.Sc. Thesis. University o f Alberta, Edmonton, Alberta. Phillips, I.J. 1975. Apical dominance. Annual Review o f Plant Physiology 26: 341-367. Pierce, J.D. 1983. Food habits, movements, habitat use, and population parameters o f moose in north central Idaho and relationships to forest management. M.S. Thesis, University o f Idaho, Moscow, Idaho, USA. Porter, G. L. 1990. Willow species o f disturbed sites in the S.B.S. zone in North-Central B.C. FRDA handbook 004. Ministry o f Forests Research Branch, Victoria, B.C. Powell, J. and T. W. Box. 1966. Brush management influences preference values o f South Texas woody species for deer and cattle. Journal o f Range Management 19: 212-214. Radwan, M.A. and D.S. DeBell. 1988. Nutrient relations in coppiced black cottonwood and red alder. Plant and Soil 106: 171-177. Raitio, H., K. Paukkonen and A. Kauppi. 1994. Effects o f defoliation, nitrogen nutrition, and temperature on leafing and root carbohydrates o f birch seedlings. Canadian Journal o f Forest Research 24: 1914-1920. Raskin, I. 1992. Role o f Salicylic acid in plants. Annual Review o f Plant Physiology and Plant Molecular Biology 43: 439-463. 96 Raymond, K.S., F.A. Servello, B. Griffith and W.E. Eschholz. 1996. Winter foraging ecology o f moose on Glyphosate-treated clearcuts in Maine. Journal o f Wildlife Management 60:753-763. Regelin, W.L., C.C. Schwartz and A.W. Franzmann. 1987. Effects o f forest succession on nutritional dynamics o f moose forage. Swedish Wildlife Research Supplement 1: 247-263. Renecker, L.A. and R.J. Hudson. 1986. Seasonal foraging rates o f moose in aspen boreal habitats. Journal o f Wildlife Management 50: 143-147. Rhoades, D.F. 1985. Offensive-defensive interactions between herbivores and plants: their relevance in herbivore population dynamics and ecological theory. American Naturalist 125: 205-236, Risenhoover, K.L. 1987a. Winter foraging strategies o f moose in subarctic and boreal forest habitats. Ph.D. Thesis, Michigan Technical University, Houghton, Michigan, U.S.A. Risenhoover, K.L. 19876. Intraspecific variation in moose preference for willows. In Plant-herbivore interactions. Edited by F.D. Provenza, J.T. Flinders, E.D. McArthur. U.S. Forest Service. Intermountain Research Station General Technical Report INT-222. pp. 58-63. Risenhoover, K.L. 1989. Composition and quality o f moose winter diets in interior Alaska. Journal o f Wildlife Management 53: 568-577. Risenhoover, K.L., L.A. Renecker and L.E. Morgantini. 1985. Effects o f secondary metabolites fiom Balsam Poplar and Paper Birch on cellulose digestion. Journal o f Range Management 38: 370-372. 97 Robbins, C.T. 1993. Wildlife feeding and nutrition. Second edition. Academic Press, Inc. San Diego, California, U.S.A. Robbins, C.T., T.A. Hanley, A.E. Hagerman, O. Hjeljord, D.L. Baker, C.C. Schwartz and W.W. Mautz. 1987a. Role o f tannins in defending plants against ruminants: reduction in protein availability. Ecology 68: 98-107. Robbins, C.T., S. Mole, A.E. Hagerman and T.A. Hanley. 19876. Role o f tannins in defending plants against ruminants: reduction in dry matter digestion? Ecology 68: 1606-1615. Roberts, A. 1986. Willows o f the lower Skeena and Bulkley Valleys, and their relative importance as winter food for moose. B.C. Ministry o f Environment, Wildlife branch, Smithers, B.C. Sæther, B.E., S. Engen and R. Andersen. 1989. Resource utilization o f moose Alces alces during winter: constraints and options. Finnish Game Research 46: 79-86. Salonen, J. 1982. Nutritional value o f moose winter browsing plants. Suomen Riista 29: 40-45. Santillo, D.J. 1994. Observations on moose, Alces alces, habitat and use on herbicidetreated clearcuts in Maine. Canadian Field-Naturalist 108: 22-25. Schwartz, C.C. 1992. Physiological and nutritional adaptation of moose to northern environments. Alces Supplement 1:139-155. Schwartz, C.C., M.E. Hubert and A.W. Franzmann. 1988a. Energy requirements o f adult moose for winter maintenance. Journal o f Wildlife Management 52: 26-33. Schwartz, C.C., W.L. Regelin and A.W. Franzmann. 1987. Protein digestion in moose. Journal o f Wildlife Management 51: 352-357. 98 Schwartz, C.C., W.L. Regelin and A.W. Franzmann. 19886. Estimates o f digestibility o f birch, willow and aspen mixtures in moose. Journal o f Wildlife Management 52: 33-37. Sennerby-Forsse, L. and L. Zsuffa 1995. Bud structure and resprouting in coppiced stools o f Salix viminalis L.. S. eriocephala [Michx.] and Salix amygdaloides [Anders.]. Trees 9: 224-234. Shipley, L.A., J.E. Gross, D.E. Spalinger, N.T. Hobbs and B.A. Wunder. 1994. The scaling o f intake rate in mammalian herbivores. American Natmalist 143: 10551082 . Shipley, L.A. and D.E. Spalinger 1992. Mechanics o f browsing in dense food patches: effects o f plant and animal morphology on intake rate. Canadian Journal o f Zoology 70: 1743-1752. Shipley, L.A. and D.E. Spalinger. 1995. Influence o f size and density o f browse patches on intake rates and foraging decisions o f young moose and white-tailed deer. Oecologia 104: 112-121. Siebert, V.H., H. Forstamt, B. Sooden-Allendorf. 1980. Selection o f willow and aspen for establishment o f browsing copses. Allgemeine-Forstzeitschrifl 26: 706-708. Singer, F.J., L.G. Lauryl and R.C. Cates. 1994. Ungulate herbivory o f willow on Yellowstone’s northern winter range. Journal o f Range Management 47: 435-443. Smith, D.M. 1986. The Practice o f Silviculture. John Wiley and Sons, Inc. New York, New York, U.S.A. Sokal, R.R. and F.J. Rohlf. 1995. Biometry the principles and practice o f statistics in biological research. Third edition. W.H. Freeman Co. New York, New York, U.S.A. 99 Soper, R.B., R.L. Lochmiller, D.M. Leslie Jr. and D.M. Engle. 1993. Nutritional quality o f browse after brush management on cross timbers rangeland. Journal o f Range Management 46: 399-410. Spalinger, D.E., C.T. Robbins and T.A. Hanley. 1986. The assessment o f handling time in ruminants: the effect o f plant chemical and physical structure on the rate o f breakdown of plant particles in the rumen o f mule deer and elk. Canadian Journal o f Zoology 64: 312-321. Stafford, R.A. 1990. Allocation and modular demography o f feltleaf willow. M.Sc. Thesis. University o f Alaska, Fairbanks, Alaska, U S A. StatSoft, Inc. 1997. Statistica for windows. Version 5. StatSoft Inc., Tulsa, Oklahoma, U.S.A. Stephenson, T.R. 1995. Nutritional ecology o f moose and vegetation succession on the Copper River Delta, Alaska. Ph D. Thesis. University o f Idaho, Moscow Idaho, U.S.A. Suter, S. 1993. The morphology and chemistry o f two willow species in relation to moose winter browsing. M.Sc. Thesis. University o f Alaska, Fairbanks, Alaska, U.S.A. Suzuki, T., M. Kitano and K. Kohno. 1988. Lateral bud outgrowth on decapitated shoots o f low pruned mulberry (Morus alba L.). Tree Physiology 4: 53-60. Tabachnick, B.C. and L.S. Fidell. 1996. Using multivariate statistics. Third edition. Harper Collins College Publishers. New York, New York, U.S.A. Tschaplinski, T.J. and T.J. Blake 1994. Carbohydrate mobilization following shoot defoliation and decapitation in hybrid poplar. Tree Physiology 14: 141-151. 100 Tuomi, J-, P. Niemela, I. Jussila, T. Vuorisalo and V. Jormalainen. 1989. Delayed budbreak: a defensive response o f mountain birch to early-season defoliation? Oikos 54: 87-91. Vivas, H.J. and B.E. Sæther. 1987. Interactions between a generalist herbivore, the moose, and its food resources: an experimental study o f winter foraging behavior in relation to browse availability. Journal o f Animal Ecology 56: 509-520. Weixelman, D.A., R.T. Bowyer and V. Van Ballenberghe. 1998. Diet selection by Alaskan moose during winter: effects o f fire and forest succession. Alces 34: 213238. Wenger, K.F. 1953. The sprouting o f sweetgum in relation to season o f cutting and carbohydrate content. Plant Physiology 28: 35-49. Whitham, T.G., J. Maschinski, K.C. Larson and K.N. Paige. 1991. Plant responses to herbivory: the continuum fi'om negative to positive and underlying physiological mechanisms. In Plant-Animal Interactions. Edited by P.W. Price, T.M. Lewinsohn, G.W. Fernandes and W.W. Benson. John Wiley and Sons, Inc. New York, New York, U.S.A. pp. 227-253. Willard, E.E. and C.M. McKell. 1978. Response o f shrubs to simulated browsing. Journal o f Wildlife Management 42: 514-519. Wolfe, M.L., W.H. Babcock and R.M. Welch. 1983. Effects o f simulated moose browsing on Drummonds willow. Alces 19: 14-35. Wolff, J.O. 1978. Burning and browsing effects on willow growth in interior Alaska. Journal o f Wildlife Management 42: 135-140. Zar, J.H. 1984. Biostatistical analysis. Second edition. Prentice Hall Inc. Englewood Cliffs, New Jersey, U S A. 101 APPENDICES tr ADF LU CELLULOSE a ce LIGNIN 0 .0 5 0.00 0 -2 .5 2 .5 -5 5 -7 .5 7 .5 -1 0 1 0 -1 2 .5 12.5-15 DISTANCE FROM SHOOT TIP (cm) Appendix A. The percentages o f 3 forms o f fibre in the distal portions o f Scouler’s willow shoots collected in the winter o f 1997 as measured fi'om the shoot tip. To obtain these data 4 to 6 shoots were randomly sub-sampled firom shoots collected from 75 brushed and unbrushed willows. These shoots were sectioned, separated by section, combined and homogenized for determinations. Means are determined from analyzing the materials twice. ADF is acid detergent fibre. Error bars represent 1 SB. 102 Appendix B. Comparison o f biomass and mineral content in the shoots o f willow brushed in June 1996 (n = 15) in the first winter after brushing and unbrushed (n = 11) controls. Degrees o f fi'eedom for all comparisons were 1, 24. ppm indicates parts per million. Brushed SE Unbrushed SB F P Mass (g) 4.93 0.54 0.75 0.12 137.3 <0.001 Boron (ppm) 48.4 2.7 39.1 2.5 5.76 0.024 Calcium (%) 0.511 0.014 0.822 0.054 37.10 <0.001 Carbon (%) 48.75 0.20 51.04 0.15 74.19 <0.001 Copper (ppm) 4.34 0.40 3.20 0.22 4.38 0.047 Iron (ppm) 29.09 1.53 41.00 4.35 0.86 0.363 Magnesium (%) 0.117 0.007 0.116 0.009 0.01 0.938 Manganese (ppm) 79.92 5.75 108.35 7.31 9.60 0.005 Nitrogen (%) 0.758 0.018 0.970 0.057 16.04 <0.001 Phosphorus (%) 0.143 0.004 0.138 0.008 0.37 0.551 Potassium (%) 0.422 0.014 0.370 0.020 4.74 0.040 Sulphur (%) 0.056 0.002 0.70 0.005 8.33 0.008 Zinc (ppm) 76.91 2.89 102.51 6.74 14.76 <0.001 Shoot Attribute 103 0.6 X o X < LU LU U J z Q 0 .5 n = 32 Ss dg î§ O 00 LU 0-3 0.2 §1 o ë 01 n = 33 I- X IX n = 32 0.0 JU N E-BR U SH ED JU L Y -B R U SH E D U N B R U SH E D A ppendix C. Proportion o f willows in each brushing treatment that were brushed in 1997 showing signs o f fall deer browsing in the same year following mechanical brushing, z tests indicated that June-brushed willows were browsed more than Julybrushed ( p = 0.231, z = -3.306, P < 0.001) or unbrushed willows ( p = 0.219, z = -3.628, P < 0.001). The amount o f browsing was not significantly different ( p = 0.046, z = -0.564, P < 0.286), however, among July-brushed and unbrushed willow. Error bars represent 1 SB.